Datasets:

disi-unibo-nlp

/

SciLay

like 0

Follow

DISI UniBo NLP 11

10.3390/ani11123522

PMC8698120

Drug dosing in sea turtles is often extrapolated from other reptiles, mammals, and humans. A more accurate way to determine the appropriate dose of a particular drug is by performing pharmacokinetic studies in that species. Meloxicam is an anti-inflammatory and pain management drug commonly used in humans and a wide range of animals including sea turtles. The authors recently published a study on single-injection meloxicam pharmacokinetics in sea turtles. The current study is a continuation of the single-injection study with objectives to determine the appropriate frequency of meloxicam administration in Kemp’s ridley and green sea turtles. Further, we evaluated whether the drug accumulated in the blood after multiple injections and if it caused any side effects. The results show that Kemp’s ridley sea turtles should receive a dose of 1 mg/kg subcutaneously every 12 h, whereas in green turtles, this same dose should be used at a frequency of every 48 h. No adverse side effects or statistically or clinically significant changes to blood work parameters were noted. This study provides important information to enhance pain management in endangered sea turtles undergoing rehabilitation.

The objective of this study was to determine the pharmacokinetics and safety of multiple injections of meloxicam (MLX) administered subcutaneously (SQ) in Kemp’s ridley (Lepidochelys kempii) and green (Chelonia mydas) sea turtles. Based on results from a previously published single-injection study, a multiple-injection regimen was derived for the Kemp’s ridleys, which consisted of administering MLX at a dose of 1 mg/kg SQ every 12 h for 5 days, and for green turtles at a dose of 1 mg/kg SQ every 48 h for three treatments. Six turtles of each species were used for the study, and blood samples were taken at multiple time intervals. The terminal half-life after the last dose for the Kemp’s ridley sea turtles was calculated at 7.18 h, and for the green sea turtles at 23.71 h. Throughout the multiple injections, MLX concentrations remained above 0.57 µg/mL, a concentration targeted in humans for the analgesic and anti-inflammatory effects. No negative side effects or changes to blood parameters evaluated were observed during the study in either species. The results of this study suggest MLX should be administered SQ to Kemp’s ridley sea turtles at a dosage of 1 mg/kg every 12 h and in green sea turtles at a dose of 1 mg/kg every 48 h. The novelty of this work is that it is a multiple-injection study. Multiple injections were administered and produced concentrations that were considered therapeutic in humans, and the turtles did not have any adverse side effects. Furthermore, there were large differences in the pharmacokinetic values between green and Kemp’s ridley sea turtles.

1. IntroductionIn sea turtle rehabilitation, traumatic injuries such as limb amputations and shell and long bone fractures often occur secondary to boat strikes, predator attacks, and entanglement in fishing gear and often require surgery [1,2]. Fibropapillomatosis is common in green turtles presenting to rehabilitation centers and is often managed by CO2 laser tumor removal, which produces significant postoperative pain [3]. Significant information has been published on the capacity of reptiles to perceive pain from these types of injuries and surgical procedures [4,5]; thus, appropriate perioperative and postsurgical pain management in these cases is critical [1,2,3]. A reduced time to return to feeding has been reported in sea turtles after receiving analgesic drugs [6]. Until recently, pharmacokinetic data on analgesic drugs in reptiles including sea turtles were very limited, meaning dosage regimens were generally extrapolated from other animal species. Extrapolating drug dosages from mammals, birds, reptiles, or even different species of turtles to sea turtles may result in clinical failure or toxicity. Reptiles have a wide range of unusual behavioral, physiologic, and anatomic characteristics that may affect drug metabolism. Some recent studies are beginning to unravel the complexities of pain management in reptiles including sea turtles [7,8,9,10,11,12,13,14,15,16,17].Meloxicam (MLX) is used for pain management and to reduce inflammation in a variety of species [18]. It is a cyclooxygenase-2 selective non-steroidal anti-inflammatory drug (NSAID) that is metabolized by the liver, and the inactive metabolites are eventually excreted in the feces and urine [18]. Meloxicam pharmacokinetic studies have been performed in ball pythons (Python regius) [19], green iguanas (Iguana iguana) [10], and loggerhead sea turtles (Caretta caretta) [8,12]. A dose of 0.1 mg/kg MLX was administered intramuscular (IM) and intravenous (IV), and orally in both the loggerhead studies; plasma MLX levels considered to provide analgesia (0.57 µg/mL) in humans were not achieved, and measurable plasma levels were maintained for only a few hours. Similar results were found in red-eared sliders administered a dose of MLX at 0.2 mg/kg [9]. Anti-inflammatory effects of plasma MLX concentrations have been shown to be significantly variable among mammalian species studied. In humans, the range is 0.57 to 0.93 µg/mL [20], while in horses and dogs, the range is 0.13 to 0.19 µg/mL and 0.82 µg/mL, respectively [21,22]. Efficacy and pharmacodynamic studies for MLX have not been conducted in any turtle species; thus, in this study, we used 0.57 µg/mL as the plasma drug target level.A study was recently published by the authors on single-injection MLX pharmacokinetics in three species of sea turtles [14]. In Kemp’s ridley and green turtles, a dose of 1 mg/kg MLX was adminstered SQ. The half-life (T1/2) of MLX was 5.51 h in the Kemp’s ridleys but could not be determined in the greens. After a dose of 1 mg/kg of MLX administered SQ in Kemp’s ridley and green sea turtles, the maximum concentration (Cmax) for MLX was 6.76 µg/mL and 9.35 µg/mL, respectively. Measurable plasma concentrations in Kemp’s ridley sea turtles occurred for 48 h, while in green sea turtles, MLX was detected for 120 h. No adverse side effects were noted. In loggerhead sea turtles, the half-life of MLX administered SQ at a dose of 2 mg/kg was 2.99 h. The Cmax was 3.63 µg/mL, and measurable plasma levels only lasted for 24 h. Based on this study, a dose of 1 mg/kg of MLX admintered SQ to Kemp’s ridley and green turtles produced plasma concentrations greater than 0.57 µg/mL for 12 h and 120 h, respectively. In loggerhead sea turtles, plasma concentrations of MLX above 0.57 µg/mL were only maintained for 4 h, even at a higher dose of MLX of 2 mg/kg SQ. One of the authors (TMN) has found that MLX anecdotally shortens the return to feeding in green turtles after surgical removal of fibropapilloma tumors with CO2 laser surgery.The objective of this study was to establish the pharmacokinetics and safety of MLX after multiple SQ injections in Kemp’s ridley and green sea turtles. Because therapeutic levels in loggerheads were maintained for only a short period of time in the single-injection study, this species was not included in this study. Studies such as this are extremely valuable in understanding the appropriate dose and frequency of a drug. After single-injection drug administration, the plasma drug level rises above and then falls below the minimum effective concentration (MEC), resulting in a decline in the therapeutic effect. To maintain prolonged therapeutic activity, many drugs are administered in a multiple-injection regimen. The plasma levels of drugs administered in multiple injections must be maintained within the limits of the therapeutic window to achieve optimal clinical effectiveness. While single-injection pharmacokinetic studies provide a glance into the basic characteristics of drug deposition, multiple-injection pharmacokinetic studies are necessary to characterize the drug disposition in a manner that is consistent with its intended clinical use. Single-injection pharmacokinetic study parameters provide information for predicting the average concentration at steady state, if linear and time-invariant pharmacokinetics applies. However, these assumptions are commonly violated when the drug hepatic metabolism is saturated or when the drug clearance and volume of distribution change slightly with time, a situation that is common with metabolically eliminated drugs. Therefore, repeated administration studies are necessary to understand the relation between dose and the drug concentration profile at steady state and to confirm the predictions made from single-administration pharmacokinetic parameters.2. Materials and MethodsSix juvenile Kemp’s ridley and six juvenile green sea turtles were used in this study. All turtles were being rehabilitated at the Georgia Sea Turtle Center (GSTC) on Jekyll Island, Georgia. The health of the turtles was evaluated by conducting a physical examination and blood work which included complete blood counts, plasma biochemical panels, and protein electrophoresis (PEP). A small amount of heparinized whole blood was transferred to a microhematocrit tube and centrifuged to measure packed cell volume. Plasma total solids were measured by a refractometer. White blood cell estimates and differential counts were performed by the same person. Biochemistry profiles were performed on plasma samples using standard dry-slide determinations with a Kodak 700XRTM chemical analyzer by the Department of Pathology, University of Miami (Miami, FL, USA). Vitros Performance Verifiers I and II (Ortho Diagnostics, Rochester, NY, USA) were used to test the chemistry analyzer. Results from the solutions, representing high and low controls, were compared to Vitros ranges. The analyzer was also calibrated with Ortho Vitros reagents. Any ‘‘flags’’ or errors were fully investigated. The following blood values were measured: glucose, sodium, potassium, carbon dioxide, blood urea nitrogen (BUN), creatinine, BUN/creatinine ratio, phosphorus, calcium, uric acid, creatine phosphokinase (CPK), alanine aminotransferase (ALT), aspartate aminotransferase (AST), lactate dehydrogenase (LDH), lipase, amylase, cholesterol, glucose, and gamma-glutamyl transferase (GGT). Protein fractions were evaluated by plasma electrophoresis. Blood analyte data were reported as median (minimum, maximum). The blood parameters were tested by using the non-parametric Wilcoxon signed rank test. Values of p ≤ 0.05 were considered significant.All turtles used in the study were eating normally. The weight and straight carapace length of the Kemp’s ridley sea turtles ranged from 3.15 to 5.5 kg and 27.20 to 34.10 cm, and those of the greens ranged from 3.0 to 7.30 kg and 28.40 to 38.70 cm. All turtles used in the study were either considered to be a permanent captive or were nearing release. The sea turtles were housed in circular fiberglass tanks measuring 8 feet (2.43 m) × 8 feet (2.43 m) or 10 feet (3 m) × 10 feet (3 m). Salt water was maintained at a temperature between 74 and 77 °F (23 to 25 °C). A closed filtration system was used which included a protein skimmer, biological filter, bead filter, and ozone for disinfection. A diet of fish, squid, and crabs with multivitamin and calcium supplementation was fed to the Kemp’s ridley sea turtles. Green sea turtles were fed a variety of leafy greens, an in-house prepared gel diet, and multivitamin and calcium supplementation [14].Baseline MLX plasma levels were measured 2 weeks prior to MLX administration. All turtles received MLX (injectable 5 mg/mL, Putney Inc., Portland, ME, USA) SQ in the shoulder region. Injection sites were changed from right to left to avoid local reactions. For the Kemp’s ridleys, MLX was administered at a dose of 1 mg/kg SQ every 12 h for 5 days for a total of 10 injections. A volume of 0.3 mL of heparinized blood was collected from the external jugular (cervical sinus) at time 0, then the first injection of MLX was administered, then 0.3 mL of blood was collected at 12 h, and then next injection of MLX was administered at 12 h. MLX was administered SQ every 12 h for 10 treatments with a blood sample taken prior to each treatment. Blood was collected at 5, 15, 30, and 45 min, and 1, 4, 8, 12, 24, 48, and 72 h after the last injection of MLX was administered (96 h). The total blood volume taken for the 5-day period was 6.9 mL. Green turtles were administered 3 injections of MLX at 1 mg/kg SQ every 48 h. Sample collection volume was 0.3 mL for each time with collection at time 0, 48, and 96 h, and then after the 96 h injection, blood was collected at 5, 15, 30, and 45 min, and 1, 4, 8, 12, 24, 48, 72, 96, 120, and 144 h. The total volume of blood obtained was 5.1 mL over 6 days. A complete blood count, plasma chemistry panel, and protein electrophoresis were performed approximately 1 week after the study was completed to assess general health and, more specifically, kidney and liver function. Shortly after collection, the blood was centrifuged, and plasma was then placed in a cryovial and frozen at −80 °C until plasma MLX analysis took place. Dry ice was used to ship the plasma samples to the University of Tennessee College of Veterinary Medicine.The animal care staff at the GSTC evaluated each turtle after they received the injection for any changes in swimming, intensity of flipper movement, posture in the water, mentation, respiratory rate, appetite, activity level, and any other changes in behavior. Additionally, the turtles were carefully monitored for a four-day period after injections were completed. The injection site was examined immediately post-injection and prior to the next injection for any redness or swelling.A validated high-performance liquid chromatography (HPLC) system was used to measure plasma MLX levels [23]. The chromatography system was composed of a 2695 separations module and a 2487 ultraviolet detector (Waters, Milford, MA, USA). MLX was separated on an Xbridge C18 column (Waters, Milford, MA, USA) with a flow rate of 1 mL/min. The mobile phase was composed of glacial acid in water (pH0 and acetonitrile (50:50, v/v), and absorbance was measured at 360 nm. MLX samples were thawed, and 100 µL of plasma was transferred to a screw-top tube, followed by 15 µL of prioxicam (internal standard, 5 µg/mL), 100 µL of 1 M HCL, and 2 mL of chloroform. The tubes were vortexed for 60 s and underwent centrifugation for 20 min at 1040× g. The organic phase was removed and evaporated to dryness with nitrogen. Samples were reconstituted in 250 µL of mobile phase, and 100 µL was injected into the HPLC system. Untreated plasma taken from the turtles in this study was fortified with MLX to prepare standardized curves for plasma analysis. A linear concentration was produced ranging from 5 to 10,000 ng/mL. Calibration samples were also prepared the same way. An amount of 5 ng/mL was the lower limit of quantification. The average recovery of MLX was 95%, and the intra- and inter-assay variability ranged between 1.1 and 10%.Phoenix WinNonlin 7.0 (Certara Inc., Princeton, NJ, USA) was used for noncompartmental analysis of the MLX data. Area under the plasma concentration time curve (AUC0-∞) from time 0 to infinity, maximum plasma concentration (Cmax), elimination rate constant (λz), plasma half-life (t½), and time to maximum plasma concentration (Tmax) were the pharmacokinetic parameters determined. The parameter values were expressed as the mean of the individually estimated parameters, and the variability was reported as the standard deviation, except for the half-life, which was reported as the harmonic mean, and pseudostandard deviation.The pharmacokinetic parameters from this study were statistically compared to those from the single-injection study by the paired Student t-test. Pharmacokinetic parameters from green and Kemp’s ridley sea turtles in the multi-injection study were also compared. Data were analyzed by use of the skewness test to determine normality. All statistical analyses were performed by using Graphpad Prism (San Diego, CA, USA). Mean values were considered significantly different at p < 0.05.3. ResultsNo negative side effects were observed during the study in either species. Pre- and post-drug administration behavioral observations by experienced husbandry staff revealed no adverse behavioral or appetite changes in any turtle. Complete blood counts and plasma chemistry profiles were found to be within normal limits prior to the study and approximately 1 week after the study was completed. The null hypotheses showed that blood analytes (paired data) were not significantly different before and after MLX treatment (Supplementary Tables S1–S3). After the last dose for the Kemps, the terminal half-life, maximum concentration, and time to maximum concentration were 7.18 ± 2.21 h, 4.77 ± 0.26 µg/mL, and 0.75 ± 0.27 h, respectively. For the greens, the terminal half-life, maximum concentration, and time to maximum concentration were 23.71 ± 2.81 h, 9.03 ± 2.59 µg/mL, and 1.29 ± 1.35 h, respectively; additional pharmacokinetic parameters are listed in Table 1. There were no statistical differences in any of the pharmacokinetic parameters for either group of turtles. The concentration–time curve for the Kemps is shown in Figure 1, and that of the greens in Figure 2. Throughout the multiple dosing, MLX concentrations remained above 0.57 µg/mL, a concentration targeted in humans for the analgesic and anti-inflammatory effects of the drug.4. DiscussionThere are very limited publications on single- and multi-injection analgesic drug pharmacokinetics in reptiles including sea turtles. The purpose of this study was to determine whether multiple injections of MLX administered SQ to Kemp’s ridley and green sea turtles would lead to and maintain blood concentrations over the course of a typical therapeutic regimen consistent with those producing analgesia and anti-inflammatory effects in mammals. An additional goal was to ensure that MLX did not accumulate and reach plasma levels over multiple injections that would lead to adverse side effects. There were no significant changes to behavior, appetite, or blood parameters that were monitored during the course of the study. Furthermore, there were no injections site reactions observed during the course of the study. The dose, route, and frequency of administration for this study were based on a previously published study by the authors on single SQ injection of MLX in Kemp’s ridleys, green, and loggerhead sea turtles [14]. Based on the results from that study, it was determined that MLX does not reach therapeutic levels in loggerhead turtles even after doubling the dose (2 mg/kg); thus, this species was excluded from this study. A dose of MLX of 1 mg/kg administered SQ every 12 h was used in Kemp’s ridleys, and a dose of MLX at 1 mg/kg was administered SQ every 48 h in green turtles. Significant variation exists among mammalian species that have been studied for the concentration of MLX that causes the anti-inflammatory effects, ranging from 0.57 to 0.93 µg/mL in humans [20], 0.13 to 0.19 µg/mL in horses [21], and 0.82 µg/mL in dogs [22]. There have been no MLX efficacy or pharmacodynamic studies conducted in any turtle species; thus, we chose plasma levels that have been determined to be anti-inflammatory in humans for the target levels in this study. The results from this study suggest that administration of MLX SQ at 1 mg/kg every 12 h in Kemp’s ridleys maintained MLX plasma levels above 0.57 µg/mL prior to the 12, 24, 36, 48, 60, and 72 h time periods. Plasma levels were monitored at 5, 15, 30, and 45 min, and 1, 4, 8, 12, 24, 48, and 72 h, after the last dose (96 h). The peak concentration was 4.77 ± 0.26 µg/mL and remained above therapeutic levels for the next 12 h. For green turtles, plasma levels were maintained above therapeutic levels prior to the 48 (second) and 96 (third) h doses. Plasma levels were monitored at 5, 15, 30, and 45 min, and 1, 4, 8, 12, 24, 48, 72, 96, 120, and 144 h, after the 96 h dose. The peak concentration was 9.03 ± 2.59 µg/mL, and concentrations were maintained above therapeutic levels for 72 h after the last dose. There was no accumulation of the drug, behavioral or appetite changes in the turtles, or clinical pathology alterations after multiple injections in either species. The multiple-injection PK parameters calculated (Table 1) were fairly consistent with our previous single-injection study. There were no statistical differences when the parameters were compared. In the green turtles, terminal half-life was not determined in the single-injection study, but in this study, it was calculated to be 23.71 ± 2.81. There were large differences in the pharmacokinetic parameters between the green and Kemp’s ridley sea turtles. The terminal half-life (h) in green turtles was 23.71 ± 2.81, whereas in Kemp’s ridley turtles, it was 7.18 ± 2.21, indicating the pharmacological activity of MLX lasts much longer in green turtles. The elimination rate constant K (1/h) was 0.03 ± 0.003 in green turtles and 0.10 ± 0.03 in Kemp’s ridley turtles, indicating the drug was eliminated more rapidly in Kemp’s ridleys than green turtles. Cmax (µg/mL) was 9.03 ± 2.59 in green turtles and 4.77 ± 0.26 in Kemp’s ridley turtles, indicating MLX is maintained at a higher plasma concentration between doses in green turtles when compared to Kemp’s ridleys between doses. There were large differences in the values between green and Kemp’s ridley sea turtles for all pharmacokinetic values.Green sea turtles are herbivorous, feeding primarily on sea grass and algae and utilizing colonic microbial fermentation for digestion. The Kemp’s ridley, on the other hand, is carnivorous, feeding primarily on invertebrate prey. These dietary differences likely affect drug metabolism and pharmacokinetics [24]. The age classes of the two species in this study were similar. Gender was not determined for the turtles in this study, but the differences in pharmocokinetic parameters were so great that it is unlikely to be soley due to gender differences.5. ConclusionsThis study confirms that multiple SQ injections of MLX maintain plasma concentrations that are analgesic in mammals and safe for use in Kemp’s ridley and green sea turtles for the endpoints analyzed. The clinical impression of the authors suggests beneficial effects from this course of treatment, as assessed by return to feeding and improved behavior in sea turtles post-surgery after removal of fibropapilloma tumors with a CO2 laser. Controlled efficacy studies are needed to assess the analgesic effects of this drug more rigorously in sea turtles. In order to determine the appropriate therapeutic concentration, pharmacodynamic studies need to be performed in each species of sea turtle. It is common practice for veterinarians and rehabilitators to extrapolate drug doses from one sea turtle species to another; however, based on the results of this study, this may not provide accurate dosing and could lead to treatment failures. In future pharmacokinetic studies in sea turtles, the authors recommend evaluating multiple species of sea turtles.

animals : an open access journal from mdpi

10.3390/ani11123473

PMC8698173

Early detection accompanied by effective treatment is vital to minimise the negative impacts of lameness in dairy cows. Locomotion scoring is commonly used for detecting lameness but can be challenging to implement effectively in cows at pasture-based systems. One potential alternative detection is measuring foot skin temperature using an infrared camera. Data were collected from a 940-cow dairy farm in New Zealand with cows observed at two consecutive afternoon milkings. Locomotion scoring was undertaken at the first milking and thermal imaging of the hind feet at the second milking. As the locomotion score increased, mean foot skin temperature increased, showing that measuring temperature could be a useful alternative to locomotion scoring. However, the process needs to be speeded up and automated if it is to be used widely.

Lameness in cattle is a complex condition with huge impacts on welfare, and its detection is challenging for the dairy industry. The present study aimed to evaluate the association between foot skin temperature (FST) measured using infrared thermography (IRT) and locomotion scoring (LS) in dairy cattle kept at pasture. Data were collected from a 940-cow dairy farm in New Zealand. Cows were observed at two consecutive afternoon milkings where LS was undertaken at the first milking (4-point scale (0–3), DairyNZ). The next day, cows were thermally imaged from the plantar aspect of the hind feet using a handheld T650sc forward-looking infrared camera (IRT). The association between FST and locomotion score was analysed using a generalised linear model with an identity link function and robust estimators. ROC curves were performed to determine optimal threshold temperature cut-off values by maximising sensitivity and specificity for detecting locomotion score ≥ 2. There was a linear association between individual locomotion scores and FST. For mean temperature (MT), each one-unit locomotion score increase was associated with a 0.944 °C rise in MT. Using MT at a cut-off point of 34.5 °C produced a sensitivity of 80.0% and a specificity of 92.4% for identifying cows with a locomotion score ≥ 2 (lame). Thus, IRT has a substantial potential to be used on-farm for lameness detection. However, automation of the process will likely be necessary for IRT to be used without interfering with farm operations.

1. IntroductionLameness is a complex multifactorial condition characterised by an abnormal gait, pain, and discomfort. Research shows that, in addition to its major impact on dairy cow welfare [1,2,3], lameness is also responsible for substantial economic losses due to treatment costs [4], reduced milk production [5,6,7] and reproductive performance [8,9,10], and increased culling [10,11,12,13,14]. Therefore, early detection and treatment of lame cows are vital to minimise the pain and discomfort associated with lameness [15,16] and to reduce the risk of irreversible claw damage [17]. Thus, early intervention improves welfare and decreases the economic impact of lameness but requires active lameness detection. Locomotion scoring (LS) is the most used method for detecting lameness on dairy farms. An observer rates the cow with a discrete score based on assessing various features of gait and posture. Numerous LS systems have been developed for use on-farm; Schlageter-Tello et al. [18] identified 25 different LS systems that had been published in the peer-reviewed literature by 2014. These systems vary in the features they use. For example, while Manson and Leaver [19] included the ability of a cow to turn, Sprecher et al. [10] did not, with the opposite being true for arching of the back. LS systems also vary in the scale they use, ranging from a simple two-point system (sound or lame) [20] to as many as nine points [19,21]. The most commonly cited system is the 5-point (1 to 5) system proposed by Sprecher et al. [10] as reported by Schlageter-Tello et al. [18]. While this system has been used in New Zealand [22], the current industry standard scheme is a 4-point (0 to 3) system based on a similar system used in the UK [23]. The large number of different LS systems identified by Schlageter-Tello et al. [18] shows no consensus on the optimal system, with each system having its own advantages and disadvantages. One crucial problem is the subjective nature of LS, with both within- and between-observer variation being high, especially when training is limited [18,24,25,26]. Visual LS is also time-consuming and can require significant labour resources, especially in larger herds where the rate at which cows exit the milking parlour can add significant difficulties for LS. This issue led Ranjbar et al. [27], who investigated risk factors for lameness in Australian dairy herds, to record locomotion scores as a tally rather than at the individual cow level. Furthermore, in pasture-based systems where LS is usually undertaken outside as cows return to pasture after milking, environmental factors such as sunlight, wind, and rain can make LS more difficult for the observer and, therefore, less accurate. One potential alternative to visual LS is infrared thermography (IRT), a non-invasive technique that measures body surface temperature and produces a pictographic representation of the imaged structure [28]. The infrared camera absorbs infrared radiation and generates an image derived from the amount of heat produced. Each pixel in the produced image represents the recorded surface temperature of the anatomical region [29]. The images can be presented both in greyscale and colour. When presented in the greyscale or colour, white or red are the hottest region, whereas black or blue represent the coldest region [28,30]. Extremities’ and surface skin temperature mostly depends on blood perfusion and tissue metabolism rate [31]. Changes in blood flow can impact the amount of radiated heat and therefore be detected by IRT [28]. One of the key reasons for changes in tissue blood flow is the inflammatory process. This link between inflammation and tissue temperature has stimulated the use of IRT as a diagnostic tool for lameness. However, inflammation is not the only process affecting foot temperature. Other factors, such as individual animal variation, physiological state, environment, and even activity level, can influence foot temperature [32,33,34]. For example, foot temperatures measured at the coronary band are higher in early to mid-lactation (≤200 days in milk) compared to late lactation (>200 days in milk) [32,34,35]. All these factors need to be considered when evaluating the utility of IRT as a method of lameness detection in dairy cattle.Nevertheless, there is a definite potential for IRT to be used for both screening for lameness and monitoring after treatment. For example, Wood et al. [36] recorded foot temperature fortnightly at milking using a non-contact infra-red thermometer alongside LS. They found that foot temperature was highest when a cow was identified as lame. Treatment resulted in a marked reduction in foot temperature, with the lowest foot temperature recorded six weeks after treatment. They noted that this temperature was lower than the temperature recorded six weeks before treatment, suggesting that an inflammatory process had been present in the foot for at least six weeks before detecting lameness using LS.There is increasing interest in the use of IRT to detect lameness in dairy cows. This increase may be related to the reducing costs of IRT and continued technical advances, which has meant that IRT has become affordable [37,38]. However, almost all the published studies of lameness and IRT have been undertaken in housed cows rather than in cows kept permanently at pasture, the production system that predominates in New Zealand. Furthermore, as the environment influences foot temperature [39,40], animal activity [33], and the type of lesion that is likely to be causing lameness [41,42], the relationship between IRT and lameness may be different in pasture-based systems. Of the peer-reviewed studies looking at IRT and lameness detection, all the papers that include data from cattle at pasture also include data from housed cows. As far as the authors know, no peer-reviewed study has analysed pasture-based dairy cows data separately from housed dairy cows data. For example, some of the data evaluated by Rodríguez et al. [43] came from cows at pasture during spring and early summer (September to January). However, they also included data from cows housed for winter (June-August) and did not differentiate between the two groups. Similarly, Harris-Bridge et al. [44] included data from cattle that were allowed to graze during the day in spring and summer (March to August) as well as from cattle that were housed during the winter or which were permanently housed but did not include housing status as a variable in their analyses.Further data on the association between IRT and lameness are needed in cattle at pasture. This is particularly relevant for New Zealand, as cattle are kept at pasture and never housed on the great majority of farms. So, we hypothesised that measuring the foot skin temperature from the plantar aspect of the hind feet of dairy cows would predict higher locomotion scores (lameness). Therefore, this study aimed to evaluate the association between hindfoot skin temperature, measured using IRT and LS in New Zealand dairy cattle kept permanently at pasture. 2. Materials and Methods2.1. Animals and Farm LocationThe study was undertaken in February on a 940-cow dairy farm in the Tararua district of the North Island of New Zealand. The farmer was a client of the Massey Farm Practice and when informed about this project was interested in participating. The herd was a split calving herd with 480 cows calving in the Spring (July–October) and 460 cows in the Autumn (March–May). Most of the cows were Friesian, with approximately 20% Jersey and 5% Friesian cross Jersey. The cows were of mixed ages ranging from 2 to 10 years with four years on average. The cows were milked twice daily through a 60-unit rotary milking parlour. The milking herd was managed as two roughly equal groups, grazing separate paddock rotations and milking in succession. Lame cows were generally identified by farm staff and then presented for treatment by a veterinarian. Cows identified as lame by farm staff were kept in a separate “lame group” in paddocks near the milking parlour until their lameness had improved enough to return to their main herd section. The lame group was excluded from the present study as, to minimise walking and standing time, cows in the group were milked only in the morning session. Based on farm treatment records of 50 lameness cases over the lactation, the main causes of lameness were white line disease (54%), sole injury (16%) and foot rot (8%). No digital dermatitis was identified at any time.Data collection involved observation of cows at two consecutive afternoon milkings. Locomotion scoring was undertaken at the first milking, with IRT being used at the second.2.2. Locomotion ScoringIndividuals were identified by their ear tag number and locomotion scored as they exited the milking parlour by CWW using the DairyNZ lameness score [45]. This scoring system has been adapted from the Agriculture and Horticulture Development Board (AHDB) mobility score to create a system that can be used to score cattle when they are walking back to pasture after being milked [23]. The DairyNZ lameness score is based on assessing walking speed, walking rhythm, weight-bearing, back alignment, head position, stride length, and foot placement (Table 1). Prior to the study commencing in February 2018, CWW was trained in locomotion scoring. The training consisted of observing training videos created by DairyNZ [46] and AHDB [47], followed by supervised locomotion scoring on-farm (live cows) with a trained and experienced observer until the trainer was satisfied that the trainee could perform locomotion scoring effectively. Visit 1: The whole herd was locomotion scored as they exited the milking parlour after afternoon milking. Locomotion scores were recorded at the individual cow level; the score was not recorded if a cow could not be identified from its ear tag. The locomotion scoring evaluation area was a flat concrete surface about 20 m in length. This walking distance was enough to assess cows’ gait and posture attributes while exiting the milking parlour.2.3. Infrared ThermographyVisit 2: Infrared thermography (IRT) imaging was performed during the next afternoon milking using a handheld T650sc Forward-looking Infrared camera (FLIR Systems, Wilsonville, OR, USA). On this day, the recorded atmospheric temperature was 22 °C. The infrared camera employed in this study had the emissivity value set at 0.95 (this relates to the capability of the object or body to absorb and emit infrared radiation).During this visit, the speed of the rotary platform was reduced to allow routine herd pregnancy diagnosis to be undertaken. CWW performed infrared thermography imaging of the claws in the hind feet at the three-quarter point on the rotation towards the exit before cows were pregnancy tested by another veterinarian. With the observer stationary (at a distance of approximately 1 metre from the cow) and the platform rotating, a plantar image of both hind feet was obtained of every fourth cow and her identity recorded. No claw preparation was performed, feet were not washed before imaging. The foot images were later analysed using FLIR Tools software (FLIR Systems, Wilsonville, OR, USA). The surface temperature estimates were obtained from seven zones on each hind limb (Figure 1). The maximum temperature for each zone was used for analysis in line with previous infrared studies aimed at lameness detection in the cow [48,49]. 2.4. Statistical Data AnalysesAll data were analysed using SPSS version 25 (IBM Corporation, Armonk, NY, USA) except where stated. Descriptive statistics exploration was undertaken for each zone temperature measure. First, the normality of foot temperature was visually assessed using Q-Q plots and histograms. A generalised linear marginal repeated measures model was then used to evaluate the effect of the foot and zone within the foot on skin temperature. Foot (right or left hind) was the dependent variable, zone within foot the repeated variable and skin temperature the outcome variable. Covariance structure was identified using the Akaike information criterion. Residuals were checked for normality using Q-Q plots and histograms. Posthoc pairwise comparisons between marginal means were then used to compare between zones (with the Šidák correction for multiple comparisons [50]).The association between locomotion score and foot temperature was tested using six temperature measures (summarised in Table 2). The relationship between these temperatures and locomotion scores was explored using box plots. This identified significant heteroscedasticity when foot temperature was compared across locomotion scores. Therefore, the association between foot skin temperature and locomotion score was analysed using a generalised linear model with an identity link function and robust estimators [51]. Each temperature definition was analysed as the outcome variable with LS as the predictor variable. A receiver operator characteristic (ROC) curve analysis was then performed. Six curves were created, one for each definition with categorised locomotion score (Lame (locomotion score ≥ 2) vs. not lame (locomotion score < 2)) to establish the sensitivity and specificity of IRT to predict locomotion score ≥ 2. The area under the curve (AUC) and coordinates of the curve (CC) were used to assess a model’s predictive accuracy. In addition, optimal threshold temperature cut-off values were determined by maximising sensitivity plus specificity. The statistical package software MedCalc Version 19.5.1 (MedCalc Software, Ostend, Belgium) was then used to calculate positive and negative predictive values for those optimal cut-offs.3. ResultsData for both locomotion scoring and infrared thermography (430 thermograms, one per hind limb) were available from 215 cows from the 940-cow herd. 3.1. Effect of Foot and Foot Zone on Skin TemperatureThere was no evidence of a meaningful difference between feet in skin temperature; left and right foot mean temperatures were 33.37 °C (95% CI: 33.286–33.443) and 33.58 °C (95% CI: 33.512–33.657), respectively, with a mean difference of 0.21 °C (95% CI: 0.18–0.45). However, differences between zones were identified; the difference between the zone with the lowest mean temperature (zone 6) and the zone with the highest mean temperature (zone 4) was 1.11 °C (95% CI: 0.87–1.34).Although mean temperatures were higher for the zones on the lateral claw than the equivalent zones on the medial claw (see Table 3, Figure 2), these differences were small (between 0.02 and 0.1 °C).3.1.1. Infrared Thermography versus Locomotion Scoring Of the 215 cows with data from both infrared thermography and locomotion scoring, 86 had score 0 (40%), 99 had score 1 (46%), 27 had score 2 (12.6%), and 3 had score 3 (1.4%). Due to the low number of cows with a score of 3, the data from the cows with scores 2 and 3 were amalgamated as a score of ≥2. For all six temperature measures (Table 2), the temperature was higher for cows with a locomotion score of 1 than those with a score of 0 and higher for cows with a locomotion score of ≥2 than those with a score of 1. Since the results for all zones showed the same trend, data are presented for mean temperature (MT) and the hottest coronary band zone (CB) only. The remaining data are presented in Appendix A. For MT, the mean difference between cows with scores 0 and 1 was 1.24 °C (95% CI: 0.9–1.58), and between scores 1 and ≥2 cows, it was 1.06 °C (95% CI: 0.58–1.54). The equivalent figures for CB were 1.2 °C (95% CI: 0.84–1.56) and 0.98 °C (95% CI: 0.47–1.49). The boxplots for MT and CB temperature measures are presented in Figure 3 and Figure 4, respectively.Interpretation Figure 3 and Figure 4 and Appendix B Figure A1, Figure A2, Figure A3 and Figure A4: The central box spans the quartiles, and the line in the box denotes the median. The line extends from the box (whiskers) to 1.5 times the interquartile range. Observations more than 1.5 times the interquartile range from the median are plotted individually as possible outliers (asterisks).3.1.2. Association of Foot Temperatures and Locomotion ScoresThere was a linear association between individual cow locomotion score and foot skin temperature for all six temperature measures. The data for MT and CB are presented in Table 4 and Appendix A Table A1 for the other four temperature measures. For MT, each one-unit locomotion score increase (assuming LS ≥ 2 was LS = 2) was associated with a 0.944 °C (95% CI: 0.781–1.141) rise in mean temperature. For CB, every one-unit increase in locomotion score was associated with a 1.067 °C (95% CI: 0.883–1.289) increase in the hottest CB temperature. 3.1.3. A Receiver Operating Characteristic (ROC) AnalysisROC curves for MT and CB are presented in Figure 5 (see Appendix B, Figure A5 for other temperature measures). In addition, optimal threshold values, area under the curve, and calculated parameters for MT and CB temperature measures are summarised in Table 5 (see Appendix A Table A2, for the results for other temperature measures).4. DiscussionThe present study aimed to evaluate the use of infrared thermography (IRT) as a tool for detecting lameness in a pasture-based dairy herd against the widely used visual locomotion scoring. IRT has been previously employed to detect foot lesions [32,36,40,44,48] and is associated with locomotion scores in housed cows [43,52]. However, the current study used IRT to detect gait changes (higher locomotion scores) in cows kept permanently at pasture. 4.1. Feasibility of Infrared Thermography as a Method on New Zealand Dairy Farms For this study, thermal imaging of the plantar aspect of the foot was done alongside routine pregnancy diagnosis, without physical animal contact. However, even with the slowing down of the platform for pregnancy diagnosis, it was impossible to obtain an IRT image for every cow due to the time required to generate an image of the suitable quality of each foot. This is obviously a major limitation of the protocol, as to score an entire herd, IRT will need to be used on multiple occasions. However, in a pasture-based system, not all cows can be locomotion scored at one milking. The high flow rate of cows exiting the milking parlour makes it impossible to observe the gait of all cows and individually identify an observed cow’s number [27]. Nevertheless, cows that are not recorded as having a locomotion score are much more likely to have locomotion scores of 0 or 1 because it is much easier to detect and identify lame and severely lame cows exiting the milking parlour than cows with no or minor gait changes. Thus to be used as an alternative to locomotion scoring, IRT needs to be much faster than it is currently. An automated imaging process from a fixed point would be faster. However, there would be challenges regarding picture quality as there will be no camera repositioning if the foot is not in focus.One limitation of this study is that the imaging process only captured temperature measurements of hind feet. Although, in housed cattle, hind limb lameness accounts for more than 90% of dairy cows [53], in New Zealand, the proportion of lame cows with hindfoot lesions is lower (71 and 56% in cows and heifers, respectively [54]). In New Zealand, the proportion of lame cows with front foot lesions is higher (29 and 44% in cows and heifers, respectively [54]). Thus only measuring hind feet is likely to have reduced the sensitivity of IRT for detecting lameness. However, front foot lameness may increase foot skin temperature in the hind feet as the animals compensate for that lameness by increasing the weight borne by the hind limbs. In addition, the IRT process may not detect cows that are lame due to non-hoof lesions (e.g., lesions of the hock or stifle). However, non-hoof-related lesions cause ~5% of lesions in lame dairy cattle in New Zealand, only [54]. Nevertheless, in the pasture-based production system that predominates in New Zealand, the only feasible time for collecting IRT images is during milking, when it is impossible to obtain high-quality images of the forelimbs easily and quickly. Therefore, further research on more cows and farms is required to establish how best to address these challenges and apply IRT in cows kept at pasture.Cows’ feet were not washed before IRT, as Stokes et al. [48] found no clinically significant difference between cleaned and dirty feet when using IRT. Furthermore, cleaning the feet would have significantly increased the time taken to obtain IRT images, further decreasing the proportion of the herd which could be imaged per milking. Nevertheless, several researchers have cleaned the feet before IRT in their studies [48,55,56]. Hence further research on the value of washing before IRT in pasture-based cattle is needed, particularly whether the benefit of washing changes during the season as cow dirtiness changes.4.2. Skin Foot Temperature and Effect of Claw and ZoneIn the present study, we used maximum temperatures within each zone for all the analyses as recommended for detecting lesions with IRT [48,49,57]. In addition, both hindlimbs were evaluated together as lameness-causing foot lesions can occur across both hind feet with equal likelihood. We found that lateral claws had a higher mean temperature than medial claws, though differences were small (highest difference of 0.1 °C between zones 3 and 7) (Figure 2). Other studies have reported larger differences in temperature between lateral and medial claws [34,58]. For example, Nikkhah et al. [34] reported that the temperature difference between the coronary band and area above the coronary band was 5.2 °C and 4.2 °C for lateral and medial claws, respectively, while Wilhelm et al. [58] reported mean temperatures of 18.6 °C and 16.9 °C for lateral and medial claws, respectively. However, both these studies were on trimmed feet, and Nikkhah et al. [34] recorded the temperature of the dorsal wall while Wilhelm et al. [58] recorded the temperature of the solear surface. In contrast, a recent study by Gianesella et al. [56] reported higher medial claw temperatures than lateral claw temperatures in both healthy cows and those with claw lesions. They reported that the difference between medial and lateral claws was 2.3 °C and 2.1 °C for healthy and diseased claws, respectively.Our results showed that both lameness and claw zone position (medial or lateral) affected temperature. These findings suggest that the small observed temperature difference between claws may be related to hindlimb lateral claws being more prone to claw horn lesions [34,54,59] and thus may have reflected subclinical conditions of the claws, which were not yet apparent. Further research is required to test this hypothesis. The effect of the zone was the same across claws, e.g., zones 1 and 5 (coronary band) had higher temperatures than zones 2 and 6 (skin above the coronary band). These results are consistent with previous studies [32,34,60,61]. However, the mean temperature difference in the current study is small, with 0.48 °C being the largest difference between the coronary band and skin above the coronary band. Temperature measurements for the other zones evaluated in the current study have not been frequently reported. However, considering the claws, zones 3 and 7 (below the accessory digits) had a higher temperature than both zones 2 and 6 (above the coronary band) and zones 1 and 5 (coronary band). Zone 4 (interdigital space) also had a higher temperature than the other zones within the foot. The higher skin temperature measured in the interdigital space (zone 4) could be explained by anatomical features. This hairless area is highly vascularised, and skin from both claws meets at this point; therefore, friction could be generated between the two claws in this relatively confined area leading to a rise in skin surface temperature. However, it is also a potential location for diseases such as foot rot, interdigital dermatitis, interdigital hyperplasia, and digital dermatitis [62]. Therefore, recording the temperature of zone 4 may be a more specific means of detecting those infectious diseases than measuring the temperature of other zones. However, this needs further investigation under New Zealand conditions because the farm was free of digital and interdigital dermatitis during this study and had a very low prevalence of footrot, so it was not suitable for testing this hypothesis.4.3. Infrared Thermography as a Predictor of Locomotion ScoreLameness prevalence (scores ≥ 2) of the cows examined in the present study was 14%. Although the current study did not include the lame group, this finding is within the range of results reported by Fabian et al. [23], who reported that lameness prevalence on 59 dairy farms across New Zealand ranged from 1.2 to 36% (mean 8.1%). The present study revealed that median claw temperature increased as locomotion scores increased (Table 4). This is consistent with previous studies that have measured skin temperature in the same foot region as this study. For example, Lin et al. [52], who used a non-contact infrared thermometer on washed feet and measured the temperature of the skin in an area roughly equivalent to zones 2 and 6 in this study, reported that they were able to differentiate score 0 from score 1 and score 1 from ≥2 using their temperature measurements. However, not all results have been as clear. Rodríguez et al. [43], who used a thermal camera and measured skin temperature in the same area as Lin et al. [52], were able to separate cows with score 0 from cows with score 2 and score 3, but could not separate score 1 cows from cows with higher or lower locomotion scores. The mean skin temperatures recorded by [43] (after washing) were 20.2, 23.2, 24.8 and 25.9 °C for locomotion score 0, 1, 2, and 3, respectively, so this lack of differentiation may be a lack of power (as Rodríguez et al. [43] only had 30 cows per score group).The present study used a cut-off of 34.5 °C for the mean temperature of all 14 zones, and this cut-off value maximised sensitivity and specificity at 80 and 92.4%, respectively (Table 5 and Appendix A Table A2). Thus, this cut-off point and the values for sensitivity and specificity are higher than previously reported figures by studies of IRT and LS. For example, Rodríguez et al. [43], using a cut-off of 25.5 °C, reported a sensitivity of 46.7% and specificity of 89.7%, while Lin et al. [52], using a cut-off of 23.3 °C, reported sensitivity of 78.5% and specificity of 39.2%. The sensitivity and specificity are also higher, though not so clearly than previous studies evaluating IRT and clinical lameness. For example, Main et al. [40], using a cut-off of 25.25 °C, reported a sensitivity of 72% and specificity of 73%, while Stokes et al. [48], with a cut-off of 27.0 °C, reported a sensitivity of 80% and specificity of 73%. Nevertheless, even the high specificity reported in this study is not sufficiently high for IRT to be used as the sole screening method for identifying cows that require lameness treatment. In this population, the positive predictive value of IRT was <65%, i.e., 1/3 of cows predicted as having a locomotion score > 1 by IRT actually had a score ≤ 1. If cows identified as lame by IRT are progressed to lameness treatment without further observation such as LS, a significant amount of staff time will be wasted examining non-lame cows. If IRT were to be used for frequent, ongoing monitoring in a population (e.g., daily measurement), this issue would be even more significant as monitoring will reduce the number of unidentified lame cows. Therefore, increase the proportion of the herd that is not lame, and the proportion of false positives produced by IRT. For example, in a herd where no cows are lame, IRT will still identify on average 15 lame cows for every 200 cows examined (specificity = 92.4%). One caveat to this discussion is that the calculation of sensitivity and specificity used in this study assumes that locomotion scoring is a gold standard when it is known that LS does not have 100% specificity or sensitivity [21,63,64]. This could be addressed by using a latent class analysis which does not assume that either LS or IRT are a gold standard. However, the authors are not aware of any latent class analysis of LS, and such an analysis is beyond the scope of this paper.The comparison of specificity and sensitivity results highlighted the difference between the optimal temperature thresholds identified in this study and previous studies. Previous studies of IRT and LS and IRT and clinical lameness have identified a range of thresholds [36,40,43,44,48,52,55,56], but, as far as the authors are aware, this study’s optimal cut-off is the highest reported. The difference between the threshold in this study and the highest previously reported threshold is much greater than the differences between previous studies. Some of this can be explained by differences in the protocol; e.g., both Rodríguez et al. [43] and Lin et al. [52] measured skin temperature after washing, but washing reduces temperatures by up to 2 °C [48], not than the 8–10 °C difference seen between the thresholds identified in the current study and those of previous studies. Thus other factors must be influencing at least some of the differences. These may include ambient temperatures, e.g., studies in the UK [36,40,48,62] were undertaken during the autumn/winter/spring seasons, while our study was conducted during summer. However, it is also possible that a key reason for the difference is that the cows in this study were all kept at pasture, whereas previous studies were undertaken in housed cows. Perhaps the main difference between these two systems is that cows at pasture are much more active in contrast to housed cows. In particular, the cows in this study will all have recently walked from the grazing area to the milking parlour. There have been no published data on the impact of such activity on foot skin temperature, but it is likely to have increased blood flow and, therefore, skin temperature. Further research is needed to understand better how walking affects temperature as there is significant variation in distance walked from pasture to the milking parlour within and between farms.Nevertheless, even though we do not know exactly how walking distance affects foot temperature. It is another variable that needs to be considered when interpreting IRT results alongside other factors such as ambient temperature, recent rainfall, current weather, foot cleanliness, and lactation stage. All of these factors change day-to-day, and thus their effect could not be investigated in this study which was based on the analysis of results from a single timepoint. There are also likely to be significant differences between farms in many of these factors. It is thus likely that the optimal threshold temperature for IRT in cattle kept at pasture will not be consistent across farms or over time within a single farm. Further research across New Zealand on more farms for longer periods is required to identify how optimal IRT threshold changes and the key factors responsible for this change.If thermal scanning does become feasible as an on-farm lameness detection method, its use will probably have to be based on repeated measurements on individual cows over time (which will necessitate some form of automation). Furthermore, these repeated IRT results will have to be combined with multiple inputs from other sources (such as weather stations). This practice will create a large dataset that is best analysed using a machine learning type process that can deal with within and between farm heterogeneity (such as classification by analysis which has just been used for diagnosing mastitis from a similarly complex dataset [65]). 5. ConclusionsOur results demonstrated that the plantar aspect of the hindfoot could be easily thermally imaged for measuring the hindfoot skin temperature. Therefore, this location can be used for assessing the presence of foot-associated lameness-causing lesions as it can evaluate multiple anatomical areas, including coronary band, surface skin above CB, interdigital space, and surface skin below the accessory digit. Furthermore, the results of the present study show that such measurements can be used to distinguish between cows with different locomotion scores such as score 0 (sound–cows that do not need attention with regards to lameness), 1 (imperfect gait–cows that need close observation), and ≥2 (lame cows that need treatment). Therefore, IRT has a considerable potential to be used on-farm to screen for lameness. However, the specificity of IRT observed in the current study does not appear high enough for IRT to be used as an alternative to locomotion scoring [66]. In addition, automation of the process will likely be necessary for IRT to be used without interfering with farm operations. This automation will also open the way for repeated skin temperature measurements, resulting in more accurate lameness detection than single measurements, especially if the IRT data are combined with other inputs in a machine learning process.

animals : an open access journal from mdpi

10.3390/ani11082215

PMC8388406

Increasing fibrous feed digestibility while reducing methane (CH4) emission through manipulating rumen fermentation patterns to improve animal performance is the most critical challenge in the animal nutrition field. Nanotechnology has revolutionized the commercial application of nano-sized minerals in medicine, engineering, information, environmental technology, pigments, food, electronics appliances, biological and pharmaceutical applications, and many more. Therefore, animal nutrition scientists also resorted to using minerals and clays such as zeolite with different forms in feeding animals and evaluate this additive in animal performance. The natural zeolite clay is known for its high cation exchange capacity and adsorption characteristics that can modify ruminal fluid viscosity and binding capacity with ammonia (NH3-N). After evaluating the addition of zeolite in vivo and in vitro, results indicated that zeolite (natural and nano forms) maintained rumen pH, increased protozoa numbers, and improved propionate production. Medium supplementation level of the natural form of zeolite at 20 g/kg dry matter (DM) was the most efficient dose in reducing CH4 production, while the zeolite nano-form supplemented at 0.4 g/kg DM was the most effective dose in improving the organic matter (OM) degradability and reducing the NH3-N concentration compared to the control.

This study aimed to evaluate in vitro and in vivo dietary supplementation with different levels of natural or nano-zeolite forms on rumen fermentation patterns and nutrient digestibility. In the in vitro experiment, a basal diet (50% concentrate: 50% forage) was incubated without additives (control) and with natural zeolite (10, 20, 30 g/kg DM) or nano-zeolite (0.2, 0.3, 0.4, 0.5, 1.0 g/kg DM) for 24 h to assess their effect on ruminal fermentation, feed degradability, and gas and methane production using a semi-automatic system of in vitro gas production (GP). The most effective doses obtained from the in vitro experiment were evaluated in vivo using 30 Barki goats (26 ± 0.9 SE kg body weight). Goats were allocated into three dietary treatments (n = 10/treatment) as follows: control (basal diet without any supplementations), natural zeolite (20 g/kg DM diet), and nano-zeolite (0.40 g/kg DM diet). The in vitro results revealed that only the nano-zeolite supplementation form quadratically (p = 0.004) increased GP, and the level of 0.5 g/kg DM had the highest GP value compared to the control. Both zeolite forms affected the CH4 production, linear, and quadratic reductions (p < 0.05) in CH4 (mL/g DM), consistent with linear increases in truly degraded organic matter (TDOM) (p = 0.09), and propionate molar proportions (p = 0.007) were observed by nano zeolite treatment, while the natural form of zeolite resulted in a linear CH4 reduction consistent with a linear decrease (p = 0.004) in NH3-N, linear increases in TDOM (p = 0.09), and propionate molar proportions (p = 0.004). Results of the in vivo experiment demonstrated that the nutrient digestibility was similar among all treatments. Nano zeolite enhanced (p < 0.05) the total short-chain fatty acids and butyrate concentrations, while both zeolite forms decreased (p < 0.001) NH3-N compared to the control. These results suggested that both zeolite supplementation forms favorably modified the rumen fermentation in different patterns.

1. IntroductionThe application of feed additives in ruminant rations is one solution to improve the animal’s performance via manipulating ruminal fermentation patterns and improving nutrients utilization. Microbial fermentation of the dietary organic matter results in loss of gross energy and nitrogen. Enteric CH4 emission in ruminants represents a loss of up to 15% of gross energy of feeds; also, 75–85% of the nitrogen consumed by ruminants is excreted in the feces and urine [1]. Therefore, enhancing fibrous feed digestibility, reducing CH4 emission, and nitrogen excretion by ruminants have to improve their performance [2].Natural zeolite clay is composited of crystalline aluminosilicates and characterized by a high cation exchange capacity, high sorbent property that can modify ruminal fluid viscosity and binding capacity with NH3-N; therefore, it has been extensively used as a potential feed additive [3]. It also can capture ammonium ions, reducing the rate of their release and absorption from the rumen wall, and act as adsorbents for mycotoxins [4]. Besides, clinoptilolite of zeolite can enhance microbial ruminal fermentation by regulating ruminal pH to act as a pH-buffering agent [5].The literature reported that zeolite supplementation levels had been examined ranging from 1% to 9% of DM of ruminant diets [6,7,8]. Dietary supplementation with zeolite clay exhibited positive effects on nutrients digestion and growth performance of sheep [9]. Furthermore, zeolite positively affected animal health status and performance due to its characteristic sorbent properties that modify the ruminal environment [10,11]. Nanoclays and other nano-particles have been shown to specifically absorb mycotoxins through the gastrointestinal tract of ruminants [12]. Nanotechnology is one of the most promising applications of the twenty-first century. It can create new materials with unique properties, which change the physical and chemical characteristics of the molecules/element to have the potential to revolutionize agriculture sectors and has given birth to the new area of agro-nanotechnology, particularly in livestock production. Size reduction of materials to the nano range can increase their adsorption, absorption, and cation exchange capacity [13]. Comparative research studies of nano and natural zeolite supplementations on rumen fermentation patterns and nutrient digestibility are limited. Therefore, we hypothesized that the effects of nano zeolite on ruminal microbial activity might differ from its natural form. Therefore, the objective of this study was to investigate the in vitro dose–response effects of natural and nano-zeolite supplementations on ruminal antimethanogenic activity, fermentation end-products, and nutrient degradation. The most effective doses of both zeolite forms were evaluated in vivo for ruminal fermentation characteristics and nutrient digestibility.2. Materials and MethodsThe study was carried out at the Advanced Laboratory of Animal Nutrition and experimental farm Faculty of Agriculture, Alexandria University and Laboratory of Livestock Research Department of Arid Land Cultivation Research Institute, the City of Scientific Research and Technological Applications, Alexandria. All procedures following protocols were approved and authorized by the Institutional Animal Care and Use Committee of the Alexandria University (ALEXU-IACUC/08-19-05-14-2-22).2.1. Experimental Feed AdditivesNatural zeolite was commercially purchased from A & O trading company, Giza, Egypt. Zeolite is composed of a microporous arrangement of silica and alumina tetrahedra (Clinoptilolite) with general formula (Ca, K2, Na2, Mg)4 Al8 Si40 O96. 24H2O. The chemical composition and physical properties of zeolite in its natural form are according to the zeolite datasheet by the A & O trading company (Table 1).The nano-zeolite powder was prepared mechanically by a high-energy planetary ball mill (Retsch PM, Germany) [14]. The mechanical route was performed in a period of 6 h with a reverse rotation speed of 300 rpm and vial rotation speed of 600 rpm with a ball to powder ratio of 9:1 mass/mass. The particle size of the obtained nano zeolite was measured by N5 submicron particle size analyzer (BECKMAN COULTER, Brea, CA, USA), with a range of 3 nm–5 µm of particle size.To detect the distribution size and shape of zeolite nano-particles, the scanning electron microscope (SEM; Jeol JSM-6360 LA, 3-1-2 Musashino, Akishima, Tokyo, Japan) and transmission electron microscope (TEM; JEOL JEM-2100, 3-1-2 Musashino, Akishima, Tokyo, Japan) were used to provide three-dimensional images, which are very useful for understanding the morphological characters of the tested nanoparticles [15]. The sample was coated with gold to improve the imaging of the sample. The SEM was operated at a vacuum of the order of 10, and the accelerating voltage of the microscope was kept in the range of 10–20 kV (Figure 1). The TEM nano-particles’ shape and size were prepared by dropping approximately 10–15 µL of a dilute sample of ZnO-NPs on the top of the carbon-coated copper grid and left in the hood to dry (Figure 2). The particle size mean was 60.2 nm of the nano zeolite.To identify the functional groups of the prepared nano-zeolite form, the Fourier Transform Infra-Red Spectroscopy (FTIR) analysis was performed using an infrared spectrometer (Shimadzu FTIR-8400S, Nakagyo-ku, Kyoto, Japan) by employing the KBr pellet technique [16], as shown in Figure 3.The surface charge of the nano-zeolite was measured by zeta potential analysis using a Malvern ZETASIZER Nano series (Malvern, Worcestershire, England, United Kingdom) [17], under the following circumstances: temperature (°C) 25.0, count Rate (kcps) 347.4, measurement position (mm) 2.00, and attenuator 7.00. The zeta potential of the prepared nano-zeolite was −5.85 (mv), zeta deviation and conductivity were 63.8 (mV) and 0.00165 (mS/cm), respectively, as presented in Figure 4.2.2. Basal DietThe experimental basal diet (used in the in vitro and in vivo experiments) consisted of (g/kg DM) 500 g concentrate and 500 g berseem hay (Trifolium alexandrinum) formulated as a total mixed ration (TMR) diet to meet the nutrient requirements of lactating goats [18]. The AOAC [19] analytical procedures were used for dry matter (DM), organic matter (OM), crude protein (CP as 6.25 × N; by Kjeldahl technique), and ether extract (EE). Cell wall ingredients (neutral detergent fiber (NDF), acid detergent fiber (ADF), and lignin contents (ADL)) were determined sequentially by an Ankom 200 fiber analyzer unit (ANKOM Technology Corporation, Macedon, NY, USA) and expressed exclusive of residual ash as described by Van Soest et al. [20]. Concentrations of hemicellulose were calculated as NDF—ADF, and cellulose as ADF—ADL.The major ingredients and chemical composition of the experimental diet are presented in Table 2.2.3. The In Vitro ExperimentThe experimental treatments consisted of control (basal diet without supplementation), five supplemental doses of the nano-zeolite (0.2, 0.3, 0.4, 0.5, and 1 g/kg DM basal diet), and three doses of the natural zeolite (10, 20, and 30 g/kg DM), and were evaluated in vitro.2.3.1. Gas Production ProcedureMethod of the semi-automatic system of GP equipped with pressure transducer and a data logger (Pressure Press Data GN200, Sao Paulo, Brazil) as described by Bueno et al. [21] and adapted by Soltan et al. [22] was used to evaluate the dose–response effects of the experimental supplementations.Rumen contents were collected freshly from adult fasted slaughtered of three Egyptian buffalo steers at the slaughterhouse of Faculty of Agriculture Alexandria University. The slaughtered animals were fed ad libitum a diet consisting of 50:50 commercial concentrate mixture: clover hay (Trifolium alexandrinum L.) and had free access to fresh water. Rumen contents were collected and kept separately in pre-warmed containers (39 °C) under anaerobic conditions. To prepare the rumen inocula (n = 3), the rumen content of each animal was blended for 10 s, squeezed through three layers of cheesecloth, and kept in a water bath (39 °C) under CO2 until inoculation took place. The different ruminal inocula were used to prevent the unusual effects of rumen environmental conditions [23,24].Four analytical repetitions (4 bottles/inoculum/treatment) were used; two for the fermentation parameters and protozoal count, and the other two were for the determination of truly degraded organic matter (TDOM). Similarly, blank bottles (rumen fluid and buffer solution), and internal standard bottles (rumen inoculum, buffer solution, and clover hay) were prepared to correct for the sensitivity induced by the inocula [24,25].Samples (0.5 g) of the experimental supplemented diets were weighed into numbered bottles and were incubated with 45 mL of diluted rumen fluid (15 mL mixed rumen fluid + 30 mL of Menkes buffered medium) in 120 mL incubation bottles [24,25]. Bottles were then sealed immediately with 20 mm butyl septum stoppers (Bellco Glass Inc., Vineland, NJ, USA), mixed, and incubated in a forced-air oven (FLAC STF-N 52 Lt, Treviglio, Italy) at 39 °C for 24 h. The gas head-space pressure of all bottles was recorded at 3, 6, 9, 12, 24 h incubation using a pressure transducer and a data logger (Pressure Press Data GN200, Piracicaba, Sao Paulo, Brazil). The pressure of GP in all bottles at each measuring time was converted into volumes to calculate the total accumulative gas produced through 24 h [22].For CH4 determination through 24 h, one mL of gas of the bottle head-space was sampled by a syringe (med Dawliaico, Assiut, Egypt) at each gas pressure measuring time and accumulated in a 5 mL vacutainer tubes (BD Vacutainer® Tubes, Franklin Lakes, NJ, USA). Methane concentration was determined using a gas chromatograph (Model 7890, Agilent Technologies, Inc., CO 80537, Santa Clara, CA, USA); the separation conditions were in detail described by Soltan et al. [22]. The amounts of CH4 produced were calculated according to Longo et al. [26]. Net values of both GP and CH4 were corrected for the corresponding blank values.2.3.2. Rumen Degradability At the end of the incubation, all bottles were put in cold water (4 °C) to stop the microbial fermentation process. Determination of TDOM was carried out according to Blümmel et al. [27] by immediate addition of neutral detergent solution (70 mL) without heat-stable α-amylase and incubated in a forced-air oven at 105 °C for 3 h. The remains were filtered in clean pre-weighed crucibles, washed with hot water, and dried at 105 °C for 16 h, and allowed to be burned at 550 °C for 4 h. The TDOM values were calculated from the difference between the amounts of the incubated OM and those remaining non-degraded. The portioning factor (PF) was calculated as the ratio of TDOM (mg) and gas volume (mL) [27].2.3.3. Rumen Fermentation Characteristics Rumen pH was determined using a pH meter (GLP 21 model; CRISON, Barcelona, Spain) in all fermentation bottles. Protozoal count was microscopically determined and differentiated by Digital Zoom Video microscope (LCD 3D, GiPPON; Wanchai, Hong Kong) following the procedure described by Dehority et al. [28]. Individual short-chain fatty acids (SCFAs) concentrations were determined according to Palmquist and Conrad [29] and adapted to Soltan et al. [22] using gas chromatography (Thermo fisher scientific, Inc., TRACE1300, Rodano, Milan, Italy) fitted with an AS3800 autosampler and equipped with a capillary column HP-FFAP (19091F-112; 0.320 mm o.d., 0.50 μm i.d., and 25 m length; J & W Agilent Technologies Inc., Palo Alto, CA, USA). A mixture of known concentrations of individual SCFAs was used as an external standard (Sigma Chemie GmbH, Steinheim, Germany) to calibrate the integrator. Concentrations of ruminal NH3-N were measured colorimetrically using a commercial lab kit (Biodiagnostic kits, Giza, Egypt) [30].2.4. In Vivo Experiment2.4.1. Animals and Experimental DesignBased on the in vitro assay results, the most effective level of both natural and nano-zeolite was selected to evaluate their responses on apparent nutrients digestibility. Thirty female non-lactating Barki goats were randomly divided into three dietary treatments (n = 10/treatment) according to initial body weight (26 ± 0.9 kg SE bodyweight) as follows: control (the same control basal diet that was used in the in vitro experiment), natural zeolite (20 g/kg DM), and nano-zeolite (0.40 g/kg DM). Animals were fed their experimental diets ad libitum. Zeolite supplementation was orally administrated to ensure the complete dose was received.Goats were fed twice daily at 08:00 and 16:00 and allowed free access to fresh water throughout the experimental period. Animals were adapted to the experimental diets for 15 days, followed by 7 days as a collection period. 2.4.2. Rumen Fermentation ParametersSamples of rumen fluid (~30 mL) were collected using an esophageal probe 3 h after the morning feeding. The first 15 mL of the ruminal sample was discarded to avoid saliva contamination; all samples were then strained through three layers of cheesecloth and immediately subjected to ruminal pH using the same portable digital pH meter that was used in the in vitro assay. Ruminal individual SCFAs, total protozoa numbers, and NH3-N concentration were analyzed as described previously in the in vitro experiment. 2.4.3. Apparent Nutrients DigestibilityFresh fecal samples (~40 g each) were obtained daily from each goat at 09:00 and 17:00, about 1 h post-feeding. Apparent nutrient digestibility was determined in which acid-insoluble fiber was used as an internal marker based on the relative concentrations of these nutrients in the feed and feces [20]. These samples were pooled per goat and stored at −20 °C for later analysis. At the end of this period, all the fecal samples were dried in a forced-air oven at 60 °C for 72 h, ground to pass through a 1 mm screen, and chemically analyzed for DM, OM, EE, NDF, and ADF as described previously. 2.5. Statistical AnalysesAll results were analyzed using the general linear model procedure (PROC GLM) procedure of SAS [31]. The in vitro gas production experiment was performed in one run for all treatments. The analytical replicates were averaged before statistical analysis, with each inoculum being the statistical replicate; thus, the statistical number of replications of treatments (n = 3) are the true replications. Orthogonal contrast statements were designed to test the linear and quadratic responses of each dependent variable to the increasing concentrations of nano or natural zeolite. The results of in vivo experiment were subjected to analysis of variance using the following statistical model as Yi = μ + Ti + ei, where Yi = observations mean, μ = overall mean, Tj = treatment effect, and ei = residual error. Differences between the treatments were considered significant at (p < 0.05), and trends were accepted if (p < 0.10). Tukey’s procedure for multiple comparisons was used to detect differences among means of the in vivo experiment.3. Results3.1. In Vitro ExperimentThe effects of different levels of natural and nano-zeolite forms on ruminal GP, CH4, TDOM, and partitioning factors are presented in Table 3. The GP increased quadratically (p = 0.004) with increasing doses of nano-zeolite supplementations, while the natural zeolite did not affect the GP values. Linear reductions (p < 0.05) in CH4 production (related to the incubated DM and TDOM) consistent with tended increases (p = 0.09) in TDOM were observed by both zeolite form supplementations. The most significant CH4 reductions (49 and 15%) were achieved by supplementations of 20 g/kg DM natural zeolite, and 0.4 g/DM kg nano zeolite, respectively, compared to the control. Neither nano nor the natural form of zeolite supplementation affected the partitioning factor. The in vitro effects of natural and nano-zeolite forms on rumen protozoal count are presented in Table 4. Increases in the total protozoal count consistent with increases in Diplodinium sp. and Epidinium sp. were observed by nano zeolite (quadratic effect; p < 0.05) and natural zeolite (linear effect, p < 0.01) supplementations. Only natural zeolite supplementation increased linearly (p = 0.001) and quadratically (p = 0.02) the Eudiplodinium sp., while no effects were observed by nano zeolite treatments. Similarly, Isotricha sp. tended to be increased (linearly, p = 0.09, and quadratically p = 0.05) with the increasing levels of the natural zeolite, while neither Entodinium nor Ophryscolex sp. was affected by both zeolite supplementations. Quadratic increases (p < 0.05) in total SCFAs concentrations and acetate molar proportions by the natural zeolite, while no effects were observed by the nano zeolite form. Both zeolite forms linearly enhanced (p < 0.05) and tended to increase quadratically (p < 0.001) propionate to molar proportions. Ratio of C2:C3 declined linearly (p = 0.01) by nano-zeolite, and quadratic (p = 0.02) by natural zeolite supplementation. The ruminal pH was not affected by dietary levels of nano or natural zeolite, while only natural zeolite linearly decreased (p = 0.004) the NH3-N concentration (Table 5). 3.2. In Vivo ExperimentThe effects of zeolite type supplementation on ruminal fermentation characteristics and protozoal count are shown in Table 6. Nano-zeolite increased total SCFAs (p = 0.021) and butyrate (p = 0.001) concentrations compared to other treatments, while it decreased (p = 0.03) valeric molar proportion compared with the natural form of zeolite. Goats fed natural zeolite had an increase (p = 0.05) in ruminal pH compared with goats fed the control diet, while no differences were observed between both zeolite forms on ruminal pH.Both natural and nano-zeolite forms declined (p < 0.001) NH3-N concentration compared with the control. Moreover, both nano and natural zeolite increased (p < 0.001) ruminal Isotrica sp. populations compared with the control, while no differences were detected among the experimental treatments on the other protozoal populations.The digestibility coefficients of DM, OM, CP, and EE are shown in Table 7. Natural and nano zeolite supplemented diets did not affect the DMI and nutrients digestibility. 4. DiscussionBoth TEM and SEM images of the experimental nano zeolite indicated that the mechanical grinding of the natural zeolite reduced their particle size and successfully presented in the nano-scale. The zeta potential of nano zeolite was negative charges that favorably enhance the affinity of palygorskite with cationic matters, e.g., cationic dyes, and then enhance the adsorption capacity. Results of the FTIR of the nano form of zeolite indicated the high efficiency of the performed nano-particles; 20 well-defined peaks of zeolite functional groups were observed, and 12 of them were found in higher frequencies at 2126–4009, while only four peaks appeared in the lower frequency range (from 466 to 789). These physico-chemical properties of the zeolite nano-form may result in different effects in rumen fermentation compared to its natural form. Increases in GP values caused by nano-zeolite addition may indicate the higher efficiency of nano-zeolite to improve the ruminal microbial fermentation than the natural zeolite; this can be due to the large surface areas, large capacity for cation exchange, and high activities caused by the size–quantization effect [32]. Rumen CH4 production is strongly related to microbial fermentation extent; therefore, enhancements in GP and nutrient degradability can increase rumen CH4 emission [33]. Thus, such increases in the total GP caused by the nano zeolite may partly explain the low efficiency of the nano-zeolite to reduce CH4 production compared to the normal form. Reductions in CH4 production caused by nano or natural forms confirmed the anti-methanogenic activity of zeolite in this study. Zeolite may act as an alkalinizer and has a high capacity for H+ exchange at different pH ranges [34,35]. Therefore, zeolite can reduce CH4 emission by affecting rumen H+ exchange capacity and can also affect all the end fermentation characteristics. Most common CH4 inhibitors may adversely affect the ruminal nutrient degradability and/or microbial fermentation at doses that achieve desirable CH4 reduction [22]. In the in vitro study, CH4 reduction consistent with increases in TDOM and GP caused by nano zeolite supplementations may indicate that both zeolite types might benefit the alteration of ruminal fermentation pattern towards less CH4 production without adverse effect on feed degradability. This can be due to the catalytic activity of zeolite nano-particles which can increase some digestive fiber enzymes (such as amylase, α-amylase) to improve OM degradability [36]. Moreover, the literature reported that enhancing the rumen nutrient degradability is a typical action of zeolite through the buffering effect and maintaining the ruminal pH from the rapid decrease [36]. In the current study, both zeolite forms enhanced the TDOM in vitro and the pH in vivo, while no differences were detected in the nutrient digestibility in vivo. The reasons for this phenomenon are not clear, but it seems that the activity of zeolite is more efficient in the rumen than in the post-ruminal digestive tract parts. The current results are in line with Galindo et al. [37], who reported that zeolite could provide favorable conditions for the increase of cellulolytic rumen bacteria and subsequently increase the ruminal degradable organic matter. The lacking effects of zeolite supplementation on apparent nutrient digestibility are consistent with what reported by Câmara et al. [38], as total tract DM digestion was unaffected when zeolite supplemented at levels of 30–50 g zeolite/kg of dietary DM, while McCollum et al. [39] observed enhancements in ruminal digestion of OM and starch with supplementation of 25 g zeolite/kg of finishing diet.Reduction in CH4 can be achieved indirectly by decreasing protozoal abundance [22], but results of the current study indicated that CH4 reduction was consistent with increases in the total protozoal count, which is mainly related to the significant increases in Diplodinium sp., the highest number of protozoal species naturally found in the typical rumen conditions of ruminants. This may partly explain the high TDOM caused by both zeolite types, as Diplodinium sp. is known for the high efficiency for cellulose degradation, and consequently, H+ abundance [33,34,35,36,37,38,39,40]. Therefore, the current results indicated that the CH4 reduction was a result of the high rumen H+ exchange capacity of zeolite.According to some previous studies [40,41], ruminal pH stability provides more favorable environmental conditions for more microbial proliferation. In the present study, the observed increase in total protozoal abundance may be due to the practical stability of rumen pH within the normal range associated with the available energy (as SCFAs production) and nitrogen (as adequate NH3-N concentration) for more microbial protein synthesis. This explanation agrees with Dschaak et al. [42], who reported that the great affinity of zeolites for holding water and osmotically active cations could enhance ruminal microbial fermentation and osmotic activity that can regulate pH in the rumen by buffering against hydrogen ions of organic acids. Our results also confirmed that rumen pH plays an important role in the survival of rumen -ciliated protozoa [40,41].The SCFAs patterns of both zeolite forms declared the ability of nano zeolite to modify the microbial fermentation activity differently from its natural form. The in vitro experiment revealed that natural zeolite quadratically enhanced acetate concentration; consequently, the total SCFAs (as acetate is the main contributor of total SCFAs), while these were not caused by the nano form of zeolite. Additionally, the nano form of zeolite enhanced butyric concentration in the in vivo experiment compared with the natural zeolite form. These differences may confirm our suggested hypothesis that performing the nano form of the zeolite may affect their efficiency as feed additive differently from its natural form. Results of the in vitro assay showed that both zeolite forms enhanced propionate molar proportions concentration. These results, alongside decreases in acetate to a propionate ratio, might be due to shifting SCFAs production pattern from acetate toward more propionate production, which may explain that the fermentation process occurred in a more efficient manner where more hydrogen ion (H+) may be used by ruminal microbes to synthesize SCFAs (propionate) rather than CH4. Additionally, differences in fermentation patterns were observed by the in vivo and in vitro experiments using the same experimental dose. It seems that the time of collection of the ruminal samples (3 h post-feeding) of the in vivo assay, rather than the nutritive buffering solution used in the in vitro assays, may affect the obtained results.Both zeolite forms decreased NH3-N concentration in the in vivo assay, while it occurred only by the natural form in the in vitro experiment. Zeolite, as a cation exchanger, is capable of exchanging and holding the ammonium ion before its release by the sodium ion (Na+) present in the saliva that was entering the rumen [43,44,45]. In this regard, zeolite additive could exhibit a higher potential to sink hydrogen through its cation exchange capacity, which might be another possible explanation for zeolite-buffering properties. Lower ruminal NH3-N concentration with the addition of nano-zeolite indicated that zeolite was able to capture NH3 through the character of cation exchange capacity [46].The modulation of rumen fermentation patterns that occurred by both zeolite forms may be nutritionally advantageous for lactating and growing ruminants through enhancing ruminal OM degradability and propionate production [47].5. ConclusionsThe nano transformation of the natural zeolite positively affected the physico-chemical properties of the natural zeolite. Zeolite, whether in its natural or nano-form, was able to maintain rumen pH while reducing NH3-N concentration and affecting CH4 production without adverse effects on the apparent nutrient digestibility. Zeolites as clay minerals play a role in improving the rumen environment and fermentation end-products because of their buffering role. In both experiments, nano zeolite modified the SCFAs pattern differently from the natural zeolite. These results may suggest that the consideration of zeolite as a modifier of rumen fermentation was not only dose-dependent but also particle-size-dependent.

animals : an open access journal from mdpi

10.3390/ani11113312

PMC8614506

Whales and dolphins in managed-care and wild settings are exposed to human-made, anthropogenic sounds of varying degrees. These sounds can lead to potential negative welfare outcomes if not managed correctly in zoos or in the open ocean. Current wild regulations are based on generally broad taxa-based hearing thresholds, but there is movement to take other contextual factors into account, partially informed by researchers familiar with work in zoological settings. In this spirit, we present more nuanced future directions for the evaluation of acoustic welfare in both wild and managed-care settings, with suggestions for how research in both domains can inform each other as a means to address the paucity of research available on this topic, especially in managed-care environments.

Cetaceans are potentially at risk of poor welfare due to the animals’ natural reliance on sound and the persistent nature of anthropogenic noise, especially in the wild. Industrial, commercial, and recreational human activity has expanded across the seas, resulting in a propagation of sound with varying frequency characteristics. In many countries, current regulations are based on the potential to induce hearing loss; however, a more nuanced approach is needed when shaping regulations, due to other non-hearing loss effects including activation of the stress response, acoustic masking, frequency shifts, alterations in behavior, and decreased foraging. Cetaceans in managed-care settings share the same acoustic characteristics as their wild counterparts, but face different environmental parameters. There have been steps to integrate work on welfare in the wild and in managed-care contexts, and the domain of acoustics offers the opportunity to inform and connect information from both managed-care settings and the wild. Studies of subjects in managed-care give controls not available to wild studies, yet because of the conservation implications, wild studies on welfare impacts of the acoustic environment on cetaceans have largely been the focus, rather than those in captive settings. A deep integration of wild and managed-care-based acoustic welfare research can complement discovery in both domains, as captive studies can provide greater experimental control, while the more comprehensive domain of wild noise studies can help determine the gaps in managed-care based acoustic welfare science. We advocate for a new paradigm in anthropogenic noise research, recognizing the value that both wild and managed-care research plays in illustrating how noise pollution affects welfare including physiology, behavior, and cognition.

1. WelfareCetacean welfare is a topic of concern for the public, scientists, and policymakers. Welfare is defined as the well-being of an individual, which teeters between two opposing states, positive and negative, as the animal responds to its environment [1]. Each experience is able to tip the scale towards positive or negative outcomes, but welfare status should not be dichotomized as only “good” or “bad” [2,3]. The framework of animal welfare began with the principles of “The Five Freedoms” that include “freedom from” stimuli that initiate negative experiences, such as poor nutrition, poor environmental standards, poor health, negative behaviors, and poor affective states [4]. Welfare science has expanded these freedoms to those of the Five Domains.The Five Domains approach is composed of a mental state and four interacting physical states, which include nutrition, environment, health, and behavior [5,6,7]. The five domains additionally promote positive experiences through enrichment as well as freedom from negative experiences extending beyond preventing cruelty into positive welfare promotion [5,6,7]. Potential negative experiences, such as pain and suffering, are central concerns to an animal’s quality of life. Negative affective states or experiences can arise from both physical and cognitive stimuli [5]. While pain impacts the physical state of the animal, suffering includes pain plus negative affective states [3]. The impact of negative experiences depends on the intensity and duration of the event, with intensive stimuli possessing the potential to demotivate animals from exploring positive rewarding behaviors, or even attending to basic needs such as food or water [6]. Therefore, it is important to replace negative situational experiences with positive ones to shift the animal’s welfare state into the positive spectrum [6].Due to their reliance on sound, cetaceans are vulnerable to potential distress in environments where sound exceeds various threshold levels [8], thus exposing them to negative experiences and poor welfare outcomes related to the Five Domains. A negative acoustic environment can affect the other interacting domains. If an environment promoting negative welfare inhibits foraging behaviors by masking echolocation, the animal can suffer poor nutrition, and if the poor nutrition continues, poor health outcomes are experienced. Similarly, a negative acoustic environment can impact mental states due to chronic negative experiences that can impact the four physical domains by causing a chronic stress response.Both wild animals and those housed in managed-care potentially face acoustic welfare concerns. In managed care, acoustic welfare challenges may be influenced by the habitat itself [2,3], as animals are housed in two main types of environments: pools and natural lagoon enclosures [9]. Reverberations related to the reflective concrete properties in pool environments are suggested by some to be a potential source of noxious acoustic stimuli; however, forethought of pool design may help mitigate noise reflected off smooth concrete surfaces. This could include the use of new multilayer membrane materials that alternate the composition of solid and liquid materials [9,10]. In contrast, while natural features may minimize the reverberations experienced in pool-based aquaria, different acoustic welfare concerns may depend on the geographic location of a lagoon-based facility and the degree of exposure to anthropogenic sound sources, including recreational and commercial vessels [9].Acoustic welfare in managed care settings remains a poorly studied field. This has prompted regulatory bodies like the USDA/APHIS (Animal and Plant Health Inspection Service) to reference the wild literature to inform welfare standards under managed care, arguing that an understanding of natural behavior and ecology are key to informing appropriate welfare standards for cetaceans under human care. Furthermore, APHIS has, in the past, requested comment from the public on whether they should establish noise thresholds for each species under human care [11], and some groups have been keen to make tenuous claims about captive whale acoustic welfare, based on inappropriate comparisons [12]. While we believe that there are opportunities for both managed-care and wild domains of acoustic welfare science to inform each other, one must accurately assess the conditions of wild animals and compare how they may differ from, and relate to, their facility-based counterparts to provide meaningful recommendations. For example, the noise effects of a distant rollercoaster on a pool of orcas may be fundamentally different than seismic water guns (with amplitudes up to 201 dB re 1 μPa) on animals, with no barrier between them and that extreme sound source [13], as the sound profiles of these stimuli are very different. In other cases, animals in managed-care facilities and wild animals may experience similar acoustic welfare outcomes. This could include cetaceans housed in a sea pen experiencing the same levels of shipping and low frequency noise as other wild cetaceans within the same area. In a welfare context, no one facility type or setting (wild vs. non-wild) should be seen as inherently acoustically superior to another without a comprehensive evaluation of the mitigating factors present related to sound exposure.Acoustic welfare concerns across both toothed and baleen cetaceans have included extreme physiologically damaging outcomes, including deafness and death [14,15], and have been studied far more extensively than potential effects in managed-care settings. Despite these critical consequences to wild individuals, the discussion of these issues, welfare consequences were more centered around the conservation and preservation of stocks over the experiences of individual animals, with a few exceptions [16,17,18]. Welfare ethicists have expressed concern related to the difficulty of assigning and assessing values of well-being to wild animals outside of human care [2]. However, a welfare framework that centers on the experiences of individual animals may help identify sub-lethal effects on wild cetaceans and allow us to frame the discussion beyond just the ways that noise affects population numbers [18,19].The next step in managed-care acoustic welfare is to integrate the focus on the individuals under human care with the massive amount of data, resources and advanced techniques garnered for wild studies on anthropogenic noise effects on cetaceans in the wild. Furthermore, we recognize that behavioral data taken from wild studies can inform the welfare of animals under human care who live acoustically connected to the ocean [20]. Conversely, data collected on acoustic welfare in managed-care settings can inform wild studies with respect to acute responses and cognitive effects furthering the opportunity for controlled studies with protocols that are impossible in wild settings [20]. Therefore, we advocate for a synergistic paradigm to evaluate acoustic welfare in cetaceans, both in the wild and under human care such that each can inform the other when considering the acoustic welfare of all cetaceans.2. Cetacean Audition and Auditory ProcessingCetaceans evolved from terrestrial vertebrates that developed specialized ears shaped for sensitivity to their environment. There are three parts to cetacean ears: external ear canals that are fused in multiple species, the middle ear that amplifies sounds, and the inner ear, which is comprised of a cochlea that performs the analysis of spectral characteristics [21]. Historically, two pathways for sound conduction to the ear have been hypothesized: (1) the primary path along the pan bone region of the jaw, which deliver sound to the ossicles of the ear, and (2) an external path through the auditory meatus. Now, toothed whales are believed to conduct sound through the internal mandibular fat body pathways along the jaw [22,23]. These mandibular trumpet-shaped fat bodies have a low structural impedance that might be specialized for capturing different frequency signals, as well as amplifying sound [21]. Once sound has entered the ear, hair cells are stimulated in the cochlear membrane causing receptor potentials, which are communicated to nerve fibers and sent to an expanded auditory cortex for processing [22]. Each part of toothed whales’ auditory systems evolved for sound reception and amplification of sounds traveling in water, making them susceptible to rising levels of anthropogenic noise pollution [24].3. Anthropogenic Noise PollutionBoth wild animals and animals in managed-care living in natural lagoons can face similar anthropogenic noise environments. The Anthropocene ocean is a mixture of biological, geophysical, and anthropogenic sources, and the ratios of each sound type are rapidly changing [24]. Many industries harness sound to map out the ocean to extract valuable resources producing powerful vibrations, while vessels designed for leisure complicate the soundscape. This diversity of sound production generates noise that varies in frequency, amplitude, and potential harm.3.1. Categories of Anthropogenic Noise and Signature CharacteristicsAnthropogenic noise is divided into low-frequency (less than 1000 Hz), mid-frequency (1–20 kHz), and high-frequency (>20 kHz) categories [25]. Sources include, but are not limited to, vessel traffic, sonar devices, naval sonar, fish finders, offshore windmills, personal watercrafts, marine animal deterrent devices, dredging, and hydrocarbon drilling [24,25,26]. These sources are produced either continuously, or at intervals, with energy levels varying globally.Low-frequency sound is the most pervasive anthropogenic sound type, due to the long distances that it travels in uninterrupted deep water [25]. Low-frequency vessel noises are the most abundant sounds contributing to marine noise pollution [25,26,27]. There are numerous low-frequency anthropogenic sound sources, including but not limited to, vessel traffic, seismic exploration for hydrocarbon farming, and dredging [25]. Shipping traffic sound results from numerous sources on the vessel, including the hull of the ship, propulsion machinery, and the cavitation of the propeller. These factors define the individual acoustic signatures that typify each ship, but they are indistinguishable at long distances due to the ability of low-frequency sounds to amalgamate, creating broad spectrum peaks between 5 and 500 Hz [25]. Thus, vessel sounds may create different experiences and affect cetaceans differently, depending on the distance from the vessel, the signature of the ship itself, or the combined effects of the amalgamation of multiple vessel sources. Vessel traffic is a continuous noise source rather than a sharp onset noise. In shallow areas with high presence of vessel activity, low-frequency waves increase. However, shallow-water vessels create more concentrated noise within these coastal regions, ultimately raising ambient noise levels [24]. In areas of high vessel density, some coastal dolphin populations have altered their activity budgets. Indo-Pacific bottlenose dolphins in an urbanized estuary allocated less time for resting and more time for traveling at increased speeds [28]. Consequently, dolphins travelling at higher rates of speed who are allocating less time to rest will need to successfully forage more or suffer an energetic deficit and potentially a lower body condition composition and poorer welfare outcomes [29,30].Vessels may also come equipped with technologies that produce powerful sounds tied to seismic exploration airgun arrays used in hydrocarbon extraction and sonar arrays used in military and commercial applications. Seismic exploration produces sound from a charged air cannon at high energy levels downwards to probe the seafloor for hydrocarbon extraction. Seismic airgun arrays emit pulses at frequency levels of less than 1000 Hz, with upward frequencies above 15 kHz at sound pressure levels around 240 dB re Pa2s [26]. Airgun noises are often loud persistent noises that can penetrate an area for weeks or months [24]. These acoustic tools are a source of concern for many in the marine mammal community, given their power and persistence, and as of 2021, the National Marine Fisheries Service (NMFS) has issued new rules associated with marine mammal take authorizations in the Gulf of Mexico related to oil and gas exploration [31]. These regulations include: standard detection-based mitigation measures, including use of visual and acoustic observation to detect marine mammals and shut down acoustic sources in certain circumstances; a time-area restriction designed to avoid effects to bottlenose dolphins in times and places believed to be of particular importance; vessel strike avoidance measures; and monitoring and reporting requirements. The particular focus on bottlenose dolphins offers an opportunity for captive and managed-care led/funded studies to help inform compliance with this rule, as this species is the most ubiquitous under human care.Active sonar is produced across multiple sound frequencies and categorized across multiple levels due to their variety of practical applications at different frequency levels. Multi-categorical sonars are used globally at locations that are stationary as well as mobile, as part of ships. Low-frequency active sonars (LFAS) are the most far-reaching sonars that are used for broad surveillance; they emit frequency modulated and continuous wave components at 215 dB per projector, at frequencies ranging between 100 and 500 Hz [25,26]. Consequently, long-ranging low-frequency sounds such as LFAS sonar may impact larger numbers of cetaceans, including geographically isolated stocks, because it travels long distances [32,33]. LFAS may be a fruitful area of research, given its far-reaching nature, ability to affect individual and groups of cetaceans, and the serious repercussions of strong doses [34]. Mid-frequency active sonar (MFAS) used for anti-submarine warfare (ASW) is designed to locate objects at distances of a few hundred meters to a few kilometers. Military sonar exercises linked to multiple mass stranding events in multiple cetacean species and naval operations, including ASW, account for 9% of global beaked whale, Ziphiidae sp., stranding events [35,36]. Additionally, ASW exercises can induce temporary hearing loss in toothed whales [37]. It is also important to note that sonar activities are not limited to military applications; commercial sonars operate a narrow downfacing beam or multibeam between 3 and 200 kHz, designed to measure depth and map out profiles of the seafloor [25,38]. Sub-bottom profilers produce sound at source levels as high as 230 dB [38]. Multibeam deep-water mapping systems operate at high sound outputs (245 dB), but are oriented highly directionally [38]. Hydroacoustic sonars, also known as fish finders, primarily operate at 20–1000 kHz range [38]. Much of the focus has been on the effects of military-related sonar applications, while the effects of commercial sonar types are less prevalent in the literature, due to their linear nature and limited operating ranges. However, some commercial active sonar types produce incidental spectral peaks outside of their center frequency that could potentially be detected for hundreds of meters from the source [39]. Future studies could focus on the behavioral effects of commercial sonar, like fish finders and seabed mapping technologies that operate within currently defined safe hearing thresholds.Unlike sonar and seismic surveys, acoustic deterrent devices (ADDs) and acoustic harassment devices (AHDs) are used to purposefully control the predation behavior of aquatic food stocks. In regions of dense aquaculture, the use of acoustic deterrent devices and acoustic harassment devices is changing the aquatic soundscape and becoming a widespread contributor to anthropogenic noise pollution [40]. ADDs generate omnidirectional pings at lower decibel levels that oscillate between frequencies of 5–160 kHz at 150 dB to avoid habituation [25,38]. AHDs produce higher source level pings and frequency sweeps at 205 dB between 5–160 kHz [38]. Both ADDs and AHDs must avoid habituation responses because they are used to repel marine mammals to reduce loss of stock and by-catch at fisheries, and aboard fishing vessels. Although many of these management devices target pinnipeds, there are ADDs specifically designed to keep cetaceans from becoming by-catch in gill nets [41]. However, exposed cetaceans may suffer from temporary or permanent hearing loss, avoidance of habitat, loss of prey, and masked communication [42].3.2. Factors Influencing Anthropogenic NoiseLocations around the globe have varying compositions and rates of anthropogenic noise production. Within the last 50 years, ambient low-frequency sound increased as much as 32-fold along shipping routes, but low-frequency noise has not increased at the same rate globally [24]. In the Northwest Pacific, low-frequency noise has increased at a rate of 3 dB/decade over 60 years with the Indian Ocean not far behind varying at a growth rate of 2–3 dB per decade. In comparison, the Southeast Atlantic, Northeast Pacific, and Equatorial Pacific have shown slight decreases in low-frequency sound at the seafloor [43]. Additionally, areas of dense population and traffic close to the coast have higher ambient noise levels of low-frequency sounds [24]. While quieter shipping technologies have led to a decrease in areas (like the South Atlantic), where shipping rates remained constant, the dominant sound sources are now mid-frequency seismic air guns [43]. Finally, in high tourism regions, scenic vessels may pass dolphin populations as much as every 6 min during daylight hours [44].The propagation of anthropogenic noise varies by the depth of water, and the intensity of sound depends on the geophysical constitution of the location that it is produced in, as well as its source. In shallow water, there is greater noise pollution due to greater reflection off substrate and the shallow longitudinal wavelengths of low-frequency sounds. In noisy areas of the Ganges River, vessel sound pollution increased the ambient noise levels by 14 dB, with an average water column height decrease of 1.5 m during the dry season [45]. Although deeper natural river channels may act as a buffer to some of the increased vessel noise during times with low water levels, river dolphins require full use of shallow and deep areas for resting and foraging [45]. During louder times, marine cetaceans may retreat into deeper waters, but river dolphins and cetaceans restricted to shallow waters in sea-side lagoons or sea pens may not have such opportunities.In human care facilities, anthropogenic noise composition depends on multiple factors including, but not limited to, facility type, amount of exposure to the ocean, life support machinery, presence of shows, and location. Facilities that house marine mammals in pools do not have additional oceanic anthropogenic noise, but their enrichment devices and life support systems impact ambient noise levels [9]. Enrichment components such as wave machines, sprinklers, and bubble machines produce extra environmental noise from machinery and action. The life support systems contribute the most to structure-borne ambient sound with continuous noise produced from pump machinery [46,47]. Results on the decibel and frequency of life support systems have been mixed. One study found that ambient noise levels in the largest pool of the Monterey Bay Aquarium were 15–25 dB higher than the bay environment that it was simulating, increasing with proximity to the pump room [46]. More recently, research at the Georgia Aquarium dolphin exhibit found that overall noise levels in-air and underwater were minimally above lower hearing thresholds of bottlenose dolphins, Tursiops truncatus, and the fully operational (all pumps running) life support systems raised ambient noise levels by approximately 10 dB primarily in frequencies under 1000 Hz [47]. Ambient noise levels above 1000 Hz were not significantly different, with the life support machinery completely on or off, and ambient noise levels of 15,000 kHz near the bottlenose dolphins’ most sensitive hearing range exhibited no marked impact due to life support systems [47]. This indicates that life support systems may not be as much as a noxious sound source or initiator of negative affective states that lead to poor welfare, however, the sampling of soundscapes across facilities to ascertain noise levels is suggested to monitor acoustic welfare.Additionally, sound can be introduced by external speakers outside of the pool, as well as through underwater speakers. Some pool facilities use sound and music in public demonstrations, but those auditory soundtracks and special effects of shows did not contribute sound above life support system levels when speakers were not in the vertical plane of the water [47]. Given that dolphins participating in demonstrations spend a large portion of the show with their ears above water, this potential area of impacted welfare should be studied further. Each show has a different sound level, acoustic environment, and soundtrack selection that impact the composition of ambient levels. Managed-care facilities should also measure the sound levels at the locations where the dolphins are stationed to evaluate sound levels for the purpose of avoiding noise levels near temporary hearing shift-inducing zones (see Section 3.3 below). While anthropogenic sounds can certainly be damaging to marine mammals, introduced sound does not necessarily need to be thought of only as aversive. Through partnerships with scientists familiar with cetacean communication, acoustic enrichment opportunities exist (see Signature Whistle Playbacks).Facilities that are exposed to oceanic anthropogenic noise vary in composition, as some are pool facilities close to noise sources and others are lagoon facilities, with varying degrees of sonic exposure to the ocean. Some protected lagoons may experience less ambient oceanic noise than netted seaside sea pens that are not tucked behind natural or artificial barriers, as netting cannot buffer environmental acoustic noise. The ultimate determinant of acoustic welfare must consider many factors, including the ambient noise, as well as the habituation profiles that the animals display toward it. Hypothetically, a coastal lagoon-type facility near an industrial shipping channel with random exposures to low-frequency sonar pings may have greater welfare issues than an open sea pen with little surrounding vessel traffic. It all depends on the relatively unpredictable nuances of the environment. One cannot broadly say that one habitat is more ideal than another, without testing this from an acoustic and welfare perspective. Furthermore, constant predictable exposures outside of hearing thresholds may be habituated readily and may have minimal welfare consequences. Sound exposures will vary by time of year, predictability of vessel traffic, depth of waters, and location; therefore, each facility will experience unique ambient acoustics. Intensity of tourism and the predominant types of vessels around the facilities will also impact ambient noise in coastal facilities.3.3. Current RegulationsCurrent anthropogenic noise policy development uses many of the same sets of data and sources. For example, the Danish Center for Environment and Energy uses criteria established by Southall et al. [48] and Southall et al. [49] to define detrimental Type-I sounds as very fast onset with a short duration and large bandwidth, and less damaging P-Type sounds as typically narrow-bandwidth signals and species are assessed based on sound type [50]. Although, more recently, there have been calls to abandon the categorization of “pulse” and “non-pulse” stimuli, as sounds can have different characteristics at source and at distance, and replace this with sound classifications organized by industry type [20]. In the United States, the Marine Mammal Protection Act (1972) prohibits the purposeful and accidental harm or killing of marine mammal species [51]. The National Oceanic and Atmospheric Administration produces guidance on mitigating the effects of anthropogenic noise activities [52]. Their efforts are designed to prevent aquatic species are exposed to certain sounds that might harm monitored stocks [48]. Currently, there are three proposed cetacean hearing sensitivities: low-frequency, high-frequency (which is the most abundant group including bottlenose dolphins and killer whales), and very-high-frequency [48]. Noise exposure criteria are based on potential to induce temporary hearing loss or temporary threshold shifts (TTS) in behavioral trials. Groups are weighted with auditory functions and generic band-pass filter equations, then TTS onset is extrapolated via exposure functions. Permanent threshold shifts (PTS), or permanent hearing loss, are estimated from TTS growth rates [48]. The low-frequency and high-frequency groups’ initial TTS onsets are nearly identical at 178 and 179 dB re 1 μPa2s under water, while the temporary hearing loss threshold for the very-high-frequency group is lower at 153 dB re 11 μPa2s under water. Onset of PTS is 20 kHz above the initial start of TTS. High-frequency and very-high-frequency groups experience PTS onset at 198 and 199 dB re 1 μPa2s, while the very-high-frequency group’s threshold is lower at 173 dB re 1 μPa2s (see Table 1 for noise thresholds for common species in managed-care).From the TTS and PTS onset levels for each hearing group, potential welfare impact by hearing damage can be measured. However, group criteria are not based on cognitive impact, and there are species within the groups that have not been studied individually. In beluga whales, bottlenose dolphins, harbor porpoises, and finless porpoises, hearing sensitivities have been studied extensively. Approximately one third of the high-frequency hearing group have been measured for hearing sensitivities and, in contrast, only three species of very-high-frequency cetaceans’ (Phocoena phocoena, Neophocaena phocaenoides, and Kogia breviceps) hearing sensitivities have been measured, with each species indicating large differences in their upper frequency hearing limits [48]. Additional research on the very-high-frequency group’s individual species’ upper threshold limits and peak hearing sensitivity would greatly aid the study of acoustic welfare and anthropogenic noise impact.Recently, there has been a reevaluation of criteria used for assessing the severity of behavioral responses and predicting effects of anthropogenic sounds in animals exposed to anthropogenic noise, especially in the experimental domain. Originally not considered separately, response scores of cetaceans in managed-care and wild settings have been decoupled into multiple severity scales [20]. Under this paradigm, scores for managed-care subjects would be divided into tracks corresponding to responses that include trained behaviors, such as stationing, and untrained behaviors. Focusing on the effects of anthropogenic noise during trained behaviors provides a method to measure performance and the level that a noxious stimulus will supersede positive reinforcement [20]. All managed-care behavioral responses range from 0—no response, 1—just detectable response, 2—aversive or negative response, 3—aversion, to 4—sensitization, accounting for habituation (See Cognitive Issues section for more on habituation and sensitization effects) [20]. Wild animal disturbance, in contrast, would be scored on severity across three domains: survival, reproduction, and foraging. These disturbance effects range as scores from 0 (no response) to 9 (serious injury or mortality, exhaustion of energy sources, or failure to successfully reproduce during breeding season), not assuming habituation effects [20]. These response scales are intended to assess discrete exposure events, and therefore are NOT valid for evaluating chronic noise exposure outcomes. Coupled with threshold shifts, earlier versions of these severity assessment scores have been historically used to create step-wise threshold levels and broad taxa-based regulatory approaches [48,49], however there is movement away from broad all-or-nothing threshold categories and broad taxa based approaches, due to the difficulty of obtaining a sense of the welfare consequences of noise at the individual and species level, as well as the difficulty of assessing the role context plays when predicting consequences of discrete noise exposures [20]. Southall et al. (2021) consideration of multiple species in wild settings also applies to animals in managed-care in that facilities with mixed species groups can mitigate noise by adjusting acoustics stimulation to the thresholds of the most sensitive species.4. Consequences of Anthropogenic Noise4.1. Sub-Lethal Physiological ChangesAlthough temporary hearing shifts are limited in TTS zones, frequent sound levels from 17 dB to 50 dB above TTS hearing thresholds induce PTS for cetaceans in all hearing groups [48,50]. Twenty to thirty minutes of pulsating sounds at 173 dB were strong enough to induce TTS in wild bottlenose dolphins [53], which is lower than the high-frequency group’s TTS onset indicated above in the most recent extrapolated TTS. Mid-frequency active sonar produced at 203 dB induced TTS in bottlenose dolphins in as soon as 5 min of exposure [37], suggesting that the level of intensity above the safe hearing threshold impacts how quickly threshold shifts occur, and the duration of time may change the upper limits of safe hearing thresholds. Studies have also found that TTS can be induced at lower levels in beaked whales than previously reported for bottlenose dolphins [54]. Additionally, potential hearing shifts (i.e., TTS or PTS) depend on the positionality of the animal in respect to the sound source. Sounds between 2–30 kHz in bottlenose dolphins affected hearing thresholds differently depending on the location sound source, indicating directional sensitivities to sound sources, and the position of the cetacean’s head directed 180° away with their tail facing the sound source may provide some respite from aversive sounds [55]. This angular sensitivity to sound indicates that potential harm is affected by the incidental position of animals, and may be further complicated by sound that is stemming from multiple sources around whales or dolphins. This complexity presents an additional challenge to assessing anthropogenic noise induced harm among wild individuals, whereas their maintained counterparts in facilities can provide additional insight into this phenomenon through behavioral studies focusing on the way in which cetaceans orient their heads in respect to certain anthropogenic sounds.Hearing loss potentially causes changes in echolocation rates and click frequencies, depending on which frequencies are lost to the cetacean’s perception of sound. For echolocation to be successful, dolphins must be able to produce click trains and be able to receive those clicks back to process. Bottlenose dolphins with high-frequency hearing loss lower their click emission energy into frequency ranges that they can hear, and they increase how loud their clicks are [56]. Upper frequency loss in individuals was correlated with an approximate 0.2 kHz decrease in click center frequency for every 1 kHz of upper frequency lost. Each dolphin altered their individual click parameters and temporal click emissions in a different unpredictable pattern from other individuals with upper frequency hearing loss; the extent of hearing loss was not a predictable indicator of how much the dolphins reduced their click frequency [56]. No longitudinal studies have been performed focusing on adaptation of echolocation parameters on members of the high-frequency group with reduced upper hearing thresholds. Similarly, echolocation changes among individuals with upper frequency loss in the very-high-frequency group have yet to be explored. The effect of altered echolocation parameters on foraging success is unknown. Impaired hearing and reduced echolocation processing capabilities could dramatically affect the fitness of an impaired wild cetacean if those constraints impact one or more of the five domains of welfare. For example, altered acoustic parameters of echolocation and reduced hearing may limit foraging success and navigation, leading to poor affective states relating to nutrition, environment, and health domains.Additionally, stranding events are potential repercussions of hearing loss. Ketten (1995) analyzed potential repercussions of blast charges to cetacean ears, and predicted the repercussions of multiple levels of acoustic trauma, including hearing loss, mixed lethality zones, and death [33]. Mann et al. (2010) found that in three species of cetaceans (bottlenose dolphins, rough-toothed dolphins, short-finned pilot whales) some individuals showed significant hearing threshold shifts equivalent to severe (70–90 dB) or profound hearing loss (>90 dB) in humans [57]. However, Mann et al. (2010) also recorded multiple stranded cetaceans without hearing loss. Consequently, more research is needed on stranding events investigating hearing threshold shifts and tympanic tissues in stranded toothed whales. Stranding events lead to serious physiological repercussions that often lead to death without intervention.4.2. Stress EffectsIn the wild, and in certain experimental conditions in managed-care facilities [13], anthropogenic noise increases stress markers and stress effects. In the Ganges River, the cost of chronic anthropogenic activity decreased caloric intake by as much as 40% in noisy conditions, increasing metabolic stress on the endangered South Asian river dolphin populations, Platanista sp. [45]. Amplified anthropogenic sound intensified the metabolic deficit, doubling metabolic cost during times of quadrupled ambient noise levels [45]. Tagged whales exposed to controlled levels of low-frequency active sonar displayed reduced rates of deep diving associated with foraging, and exhibited potential extra metabolic stress [58]. When exposed to intense levels of seismic gun sources (>100 kPa), captive belugas exhibited increased epinephrine and norepinephrine stress hormones [13].Of course, acoustic welfare is not limited to the application of sudden or startling noises. Sometimes, we see decreased stress responses through the removal of a typically perpetual noise in what is a sonically intense environment. After the events of 11 September 2001, the Bay of Fundy experienced a 6 dB drop in low-frequency noise levels. North Atlantic right whales, Eubalaena glacialis, responded to the decrease in ambient sound levels with a significantly decreased level of expressed stress hormones in their fecal matter [59]. The Anthropause associated with the 2020 COVID-19 pandemic is another rare occasion of decreased human activity across the globe that has led to altered oceanic soundscapes and the opportunity to further study how chronic noise is affecting wild populations, including geographically isolated stocks, as well as the potential benefits of a repose from certain perpetual noises, such as cruise ships and other low-frequency vessel noises [60,61].In the managed-care welfare landscape, acoustic stress effects are probably one of the issues most conjectured about by opponents of the presence of animals under human care [12,62]. Papers from authors that discuss noise and stress struggle to separate acute versus chronic sources in captive settings, and often make broad comparisons to distantly related species [63]. Given the unique nature of cetacean hearing physiology and anatomy [64], and the limited amount of data that we have on what sounds that these animals are exposed to in a broad sense, this remains an open area of welfare research. This is especially true given that multiple studies have shown that baseline cortisol levels in captive bottlenose dolphins are no higher than in their wild counterparts, which could have implications for hypotheses related to chronic exposure and potential habituation processes [65,66,67,68].4.3. Acoustic Behavior and Masking EffectsMany anthropogenic noises fall within the range of cetaceans’ communication, causing masking effects that could indirectly reduce individual fitness. Baleen whales are especially vulnerable to this acoustic masking by low-frequency sound sources, due to their reliance on long distance communication to attract mates and repel competing males. Humpback whale, Megaptera novaeangilae, song primarily consists of frequencies below 1000 Hz, but they can reach upper harmonics in the range of 24 kHz [69]. The lower frequency notes of humpback song that are used to increase that animal’s fitness propagate across long distances, and are attenuated or outright canceled by overlapping noise sources, such as vessel traffic, LFAS, and hydrocarbon exploration [69,70,71]. High-frequency hearing groups experience acoustic masking and echolocation shadowing (i.e., areas of signal masking due to acoustic interference) of lower frequency and mid-frequency sounds too. For very-high-frequency species living in the Ganges River, cavitation noises produced by vessel traffic completely shadowed the broadband clicks of foraging dolphins [45].Cetaceans primarily use two strategies to counter acoustic masking: (1) they increase the amplitude of their vocalizations (Lombard effect) and (2) they shift signal characteristics of their vocalizations. These responses increase the likelihood that the callers are heard through increased volume (1) or occupying a frequency with less spectral clutter (2). In areas saturated with noise pollution, many species of cetaceans exhibit the Lombard effect, which is an increase in the volume of vocalizations in the presence of excessive noise [72,73]. Bottlenose dolphins in Sarasota Bay, Florida responded to temporary increased sound bouts by increasing the intensity of their overall vocalizations, but the shift was not congruent among signature and non-signature whistles [73]. The dolphins shifted the intensity of their signature whistles less than their non-signature whistles [73]. Bottlenose dolphins respond to whale watching vessels by increasing the frequency of calls with an upward shift by up to 1.99 kHz [72]. Separate populations of killer whales of the Puget Sound and off the coast of Iceland responded to potential masking risks by calling louder [72,74]. Some species battle acoustic masking by fundamentally altering the frequency and duration of their calls. In addition to the Lombard effect, killer whales also responded to vessel noise by modulating their vocalizations, by increasing the durations of their calls [75]. Over just a few decades, right whales have shifted their calls to higher frequencies and decreased calling during peak background activity [76]. When exposed to vessel, naval, and air gun noise pollution of right whales, fin whales, Balaenoptera physalus, and grey whales, Eschrichtius robustus, change fundamental spectral characteristics, including the frequency components of their calls [76,77,78]. These responses to acoustic masking may have an indirect fitness reduction on species, and not every species responds the same. Although these changes may not be large percentages in the total hearing and vocal production range of cetaceans, shifts in the frequency of whistle production decrease the propagation of calls, and may decrease the efficacy of communication increasing energetic efforts, impacting the nutrition and health domain.Some populations of cetaceans respond to the increased presence of noise pollution by changing their vocal behavior, including the modification of vocal production rates. Moreover, they display avoidance tactics. Changes in vocal production rates have been marked across many species. Beaked whales and right whales reduce call rates in high noise conditions. When harassed by operating vessels near Sado Estuary, Portugal, bottlenose dolphins mean overall call rates decreased, and the dolphins displayed significantly reduced click rates [79]. Bottlenose dolphins off Sarasota Bay, Florida, are exposed to vessels passing as close as 100 m every six minutes. When approached by vessels, these dolphins temporarily increased whistle production, then call rates decreased after the vessels left [44]. Disruptions in communication may possibly affect bonding, social structures, learning opportunities provided by mother to calves, and successful hunting, all of which in turn jeopardize the health of populations. Smaller cetaceans that are within the very-high-frequency hearing group have a shorter threshold for noise exposure, and it is not unreasonable to surmise that they experience increased negative welfare outcomes, due to increased levels of anthropogenic noise. Heaviside’s dolphins, Cephalorhynchus heavisidii, relax acoustic crypsis (an anti-predation acoustic modification where the sound producer operates at a higher or lower frequency than is detectable by predators) to increase echolocation range which may expose them to higher risks of predation by eavesdropping predators [80]. This reduction in acoustic crypsis may be further impacted by increased levels of anthropogenic noise. Studies on the impact of anthropogenic noise on very-high-frequency species kept in captivity, such as Commerson’s dolphins, Cephalorhynchus commersonii, would provide instrumental information for shy wild species that are difficult for researchers to study. For example, acoustic crypsis in the presence of anthropogenic noise may be studied in Commerson’s dolphins by setting up a simple playback experiment. Potential experiments focusing on acoustic welfare in very-high-frequency species using a playback methodology could focus on respiration rates (a common dependent measurement used in facilities to assess increased metabolic function due to potential stress), displacement from the sound source, or other welfare metrics.Rose and Parsons (2019) speculated that the influence of smooth walls and increased reverberations may cause dolphins to reduce rates of echolocation [81]. In this case, the argument is that the reverberations will cause dolphins to reduce their use or change the parameters of their echolocation. To date, this concern remains conjecture, as it has not been systematically assessed. To address this issue, a fruitful area of research might involve the question of masking. Investigations on any potential decrease in echo-rates under managed-care could compare the levels of echolocation between pool facilities and lagoon or sea pen facilities, where environments are less likely to cause possible masking reverberations. Research on echo-rates could be done under experimental conditions if one was concerned about habitat familiarity inhibiting echolocation in managed care pool settings. A match-to-sample design could be used to determine echolocation rates, where the subjects would be incentivized to use echolocation to solve the task, and variations on the willingness to use the sensory system could be interpreted as a potential acoustic concern in that habitat. It would be difficult to isolate whether reverberations or masking explain any decrease in echo production, but the research on production levels should be conducted first before assuming negative acoustic welfare in pool facilities based on currently available data, or the lack thereof. Furthermore, reduced echolocation rates are not necessarily a negative welfare outcome. Toothed whales have some control over self-sound exposure beyond just ‘echolocate’ and ‘no-echolocate’. Just as children are taught not to yell inside, dolphins too likely possess the ability to learn to modulate their signals to prevent their own negative welfare consequences. This ability should be considered when evaluating acoustic welfare as a function of echolocation production.4.4. Alterations in BehaviorOne of the most common behavioral responses to anthropogenic noise includes avoidance behaviors and altered distribution at times of increased anthropogenic noise presence. Bottlenose dolphins in the Dolphin Bay of Bocas del Toro archipelago are subjects of the largest whale watching operation in Panama. When exposed to vessel noise, the dolphins in Dolphin Bay abruptly changed activity states, which led to decreasing amounts of foraging and increased energy expenditures from avoiding the stimuli [82]. Bottlenose dolphins in the Spanish Mediterranean Sea preferred to inhabit areas of low recreational activities, however, the dolphins remained a large presence in areas where fishing-related activities persisted [83]. These patterns suggest that the dolphins prefer to be away from increased anthropogenic activity, but they are willing to experience human-made sound stresses when foraging is necessary. Dolphins in the Sicily Strait also follow fishing trawlers in this specialized foraging strategy [84]. Distribution changes in the water column, known as dive shifts, have been explored in multiple cetacean species, with each species responding using distinct strategies for each anthropogenic sound source. When exposed to LFA sonar, cetaceans that spend much of their time at the surface, such as killer whales and pilot whales, Globicephala sp., continue their shallow diving behavior and cease any deep diving foraging they would normally exhibit [58]. Sperm whales, Physeter macrocephalus, normally spend much of their time in dives foraging for large squids that live at deep depths. When exposed to LFA sonar, sperm whales’ dives also become abnormally shallow, which reduced time spent foraging, and long periods of exposure could lead to increased metabolic stress resulting in reduced body condition and increased stress to populations [58].Unfortunately, right now, we simply lack enough data to support many conclusions about the behavioral effects of external sounds in a managed care environment. However, one study on bottlenose dolphin social play found a reduction in play frequency during bouts of pool-adjacent construction work but not frequency of agnostic or sexual interactions [85]. Expanded studies on dolphin social play could be used to measure the degree of impact on behavioral responses to external anthropogenic sounds. Additionally, we can use surfacing behaviors or respiration rates described in the wild literature as metrics for welfare in managed-care settings. We have used surfacing inhibition behaviors in managed care settings to determine aversion to specific drone frequency playbacks in air (Bruck et al. unpublished). If the concern is that noise from music or fireworks during a public exhibition of the animals is aversive [12], then animal respiration rates should change. Respiration rates were used as a welfare measurement during some of the first captive playback sessions with signature whistles in the late 2000s [86]. If sound sources are localized, then one could evaluate habitat utilization as means to determine if sound stimuli are noxious. Since most facilities do not, as a rule, have an ‘ear’ on their animals, localized noxious noise stimuli may go undetected. Facilities could invest in hydrophones or use tracking tools to monitor their animal’s habitat use, to monitor these impacts, or partner with researchers to help in this assessment. This could be especially true if the stimuli are inconsistent, which makes these types of noises hard to habituate to from the animal’s perspective, and less likely to be noticed by marine mammal specialists and divers who may be on fixed schedules.4.5. Cognitive IssuesMultiple noise exposure events may have a non-linear effect on cognition and behavior due to habituation and sensitization effects. Sensitization manifests as a heightened response that increases each time the individual is exposed to the provoking stimulus [87]. Both the processes of habituation and sensitization interact to produce behavioral plasticity, with arousal invoking stimuli more likely to lead to sensitization [87,88]. While sensitization is a marked increase in response, habituation is characterized by a decrease in response to repeated stimuli presentations. Both habituation and sensitization to anthropogenic sound could be either beneficial or detrimental depending on the sound source because of each type of sound’s varying characteristics including frequency, duration, and intensity. Habituation response to sound might aid the individual by reducing distractions from survival critical behaviors and positive experiences, but it could also reduce survival responses to potential dangers. While a certain degree of sensitivity may aid a whale or dolphin in the avoidance of potential dangers, the added energetic expenditure and distraction from positive and rewarding experiences is a maladaptation. The degree to which these effects develop differ based on an individual cetacean’s experience with anthropogenic noise pollution sources. In the wild and in certain natural lagoon managed-care facilities, it is possible that cetaceans experience potentially startling external sounds. These potentially repeated, startle provoking stimuli could lead to habituation or sensitization [88].Recently, there has been some attention from researchers working primarily in the wild domain on the cognitive effects of anthropogenic noise. Southall et al. (2021) placed sensitization responses into the most severe spectra of behavioral response severity scores. Sensitization responses in managed-care settings may manifest in different ways depending on whether the dolphin is exposed to discrete anthropogenic noise in a training paradigm or not [20]. Sensitization responses during trained behaviors include: (1) breaks in stationing or avoidance of stations, (2) cessation of current activity to attack or charge sound source, (3) refusal to perform tasks over time even when offered primary reinforcement, (4) repeated aggressive episodes targeting trainers, conspecifics, or objects, (5) failure to recall when logging or in placement of bottom of the pool, and (6) retreating into a refuge space when available [20]. In untrained paradigms, sensitization behaviors include: (1) repeated displacement events towards subordinate conspecifics, (2) acts of aggression toward the sound sources and displacement of objects in the way, (3) retreating into a refuge for more time than the exposure of the sound, (4) negative anticipatory behavior, and (5) logging at the surface or bottom of their enclosure [20]. Southall et al. (2021) represent a framework for evaluating the problem but did not provide information on the pervasiveness of these types of responses. This framework can be expanded by focusing on sensitization/habituation responses in managed-care settings, where caregivers know the animals and the subjects can be revisited. Furthermore, while Southall et al. (2021) address a basic type of learning, researchers can further investigate cognitive effects by focusing on attention, complex learning, and memory with this paradigm. Preliminary data from an acoustic playback response experiment in bottlenose dolphins housed in inland facilities and coastal facilities indicated habituation and sensitization profiles that were unique for each anthropogenic noise [89]. When exposed to LFAS, dolphins housed in coastal facilities responded with increased look duration to the sound source, demonstrating a possible sensitization effect of LFAS [89]. Each anthropogenic sound has the possibility of affecting attention in a unique way because of inherent distinct characteristics of the sound itself (e.g., frequency, amplitude, modulation patterns, and duration). We advocate that future research utilize animals under managed-care to address the question of noise and cognition especially as it potentially relates to survival and reproduction, as well as sociality and cooperation in wild animals [86,90,91].Cetaceans may respond in a different manner over time due to the plasticity of the habituation and sensitization pathway. In Sarasota Bay, Florida, dolphins exposed to vessel noise waited for vessels to pass before resuming whistle production [44]. Vibratory pile driver noise produced at 140 dB was sufficient to cause cognitive distractions in Naval dolphins asked to perform an echolocation task, resulting in significant decreased target detection [92]. Additional cognitive impacts on learning and memory should be explored, including long-term studies of chronic noise exposure, as it affects the survivability and reproduction of these animals. For example, bottlenose dolphins learn hunting techniques as well as signature whistles of conspecifics, and remember the signals of social partners for decades [86]. Loss of attention can impact more than just foraging success or energy expenditures; dolphins rely on learning and memory to pass on survival critical behaviors. Fear generated from the same startle-invoking stimuli that lead to habituation/sensitization response may also cause pessimistic cognitive biases in non-human animals [93]. These ‘pessimistic biases’ result in greater aversion to ambiguous information and greater amounts of avoidance behavior similar to anxiety behaviors in humans [93]. They may also impact decision making pathways, creating physiological markers of stress due to stress’ role of preparing the individual to respond to potential threats [93]. Although this is an adaptive response in normal states, chronic risk aversion and sensitization to anthropogenic noise can impact fitness significantly, leading to even more stress and negative welfare. The impact of anthropogenic noise on cognition in marine mammals lacks in-depth and controlled studies and management of cetaceans would benefit from increased noise impact and vigilance studies.5. Monitoring Soundscapes in Managed Care Passive acoustic monitoring systems, or PAMS, are a vital tool in the monitoring of ambient noise levels (SPLs), acoustic response, and distribution changes of wild stock. PAMS have been used in conjunction with open-source programs such as PAMguard for over a decade [94]. PAMS are a powerful tool for researchers monitoring anthropogenic noise levels and acoustic responses in wild populations [95]. While recording and documenting soundscapes in both facilities and nature are invaluable, PAMS record without sending warning of any changes in the communication characteristics or anthropogenic noise levels of monitored facility’s population. Recently, Jones et al. [96] introduced an open-source Welfare Acoustic Monitoring System, WAMS, that monitors and alerts husbandry staff to sudden onsets of large quantities of signature whistles, which may indicate instances of negative experiences. WAMS provides a flexible interface for acoustic monitoring and is customizable for each distinct facility being studied [96]. An automated system of acoustic monitoring is the next step, but the system needs to be able to recognize types of vocalizations and baseline vocalizations for each facility. Each facility has different soundscapes; some have high levels of signature whistle production, while others are quieter with few or almost no signature whistles [97]. For WAMS to be the most beneficial to researchers and husbandry staff, it is necessary to understand the unique vocal characteristics of each facility in which it will be deployed. Calibration to each facility’s desired vocalization level, acoustic environment, species, and vocalization types of interest are important for the proper functioning of WAMS [96]. If WAMS could be programmed to alert facilities to instances where external sounds are above safe hearing thresholds before noise is capable of inducing TTS and combined with WAMs application to monitor for sudden onsets of signature whistles, WAMS could alert facilities to potentially negative welfare outcomes and be a powerful tool in the welfare management of cetaceans or other species sensitive to anthropogenic noise.6. ConclusionsAnthropogenic noise pollution is an intense and complex challenge that wild and possibly captive cetaceans face. The frequency of anthropogenic noise is increasing with the expansion of human activities causing wakes of effects on cetaceans. Because cetacean ears were adapted to conditions that were presumably quieter than current conditions and are quite sensitive, excessive levels of anthropogenic noise can have deleterious consequences, including death, hearing loss, acoustic masking, behavioral changes including acoustic behavior, and cognitive effects. Cetaceans rely on proper acoustic functioning to maintain social bonds and good body condition; thus, they are at an increased risk of poor welfare outcomes, due to the pervasive noise around them. Whales and dolphins display resilience and behavioral plasticity in response to many aspects of human sound production, yet many populations are still in decline [98,99,100,101]. Some areas of anthropogenic sound effects are well studied, such as hearing threshold shifts, the Lombard effect and acoustic masking, while little is known about the effects of anthropogenic noise on attention, learning, and memory for either captive or wild groups. Habituation profiles from preliminary data suggests that anthropogenic noise exposure does not uniformly affect attention, and possibly other aspects of cognition including learning and memory. Habituation and sensitization responses to external sounds in managed-care and wild settings should be considered when managing acoustic welfare. Future studies involving anthropogenic noise should focus on managing sources of poor acoustic welfare, stress effects based on heart rate or hormone levels, behavioral plasticity, cognitive effects of each anthropogenic noise type, and the behavioral biases such as avoidance of areas with high levels of anthropogenic noise and changes in swim patterns an individual cetacean has developed from exposure to noise related to human activity. Only when we have enough information on each sound type’s harmful effects on cognition and behavior will we be able to create robust responses to mitigate anthropogenic sound, reduce negative welfare outcomes relating to the Five Domains, and promote a positive acoustic welfare environment.

animals : an open access journal from mdpi

10.3390/ani11082381

PMC8388693

Indigenous cattle have extraordinary adaptation capability to diverse environments under low input production system. However, the population size is declining rapidly in Bangladesh due to massive imports of high yielding dairy breeds. The genetic diversity measures are important for assessing population architecture as well as for development of conservation strategies. The aim of this study was to investigate genetic variability and population structure of indigenous cattle genetic resources of Bangladesh using Illumina Bovine SNP50K BeadChip genotyped data. Similar to other zebu populations, low genetic diversity measures were found in Bangladeshi cattle populations. Our findings revealed their distinct genetic structure but showed low levels of genetic differentiation among the six indigenous cattle populations. Moreover, admixture and phylogenetic analysis highlighted historical gene flow among the studied populations. Altogether, our findings provide a comprehensive genomic information on indigenous cattle populations of Bangladesh that could be utilized in their future conservation and breeding research.

Understanding the genetic basis of locally adapted indigenous cattle populations is essential to design appropriate strategies and programs for their genetic improvement and conservation. Here, we report genetic diversity measures, population differentiation, and structure of 218 animals sampled from six indicine cattle populations of Bangladesh. Animals were genotyped with Illumina Bovine SNP50K BeadChip along with genotyped data of 505 individuals included from 19 zebu and taurine breeds worldwide. The principal component analysis (PCA) showed clear geographic separation between taurine and indicine lineages where Bangladeshi indigenous cattle clustered with South Asian zebu populations. However, overlapped clusters in PCA, heterozygosity estimates, and Neighbor-Joining phylogenetic tree analysis revealed weak genetic differentiation among the indigenous cattle populations of Bangladesh. The admixture analysis at K = 5 and 9 suggests distinct genetic structure of the studied populations along with 1 to 4% of taurine ancestry. The effective population size suggested a limited pool of ancestors particularly for Sahiwal and North Bengal Grey cattle. In conclusion, these findings shed insights into the genetic architecture of six indigenous cattle populations of Bangladesh for the first time and suggested as distinct gene pools without potential admixture with zebu or taurine populations.

1. IntroductionIndigenous cattle are an important livestock species in smallholder farming system of Bangladesh. They have potential contributions in rural livelihoods, nutrition security, organic agriculture, and socio-cultural and historic events. Moreover, rural farmers prefer indigenous cattle due to their better reproducibility and adaptation under low-input management practices [1,2]. In Bangladesh, the total heads of cattle are estimated to be 24.39 million [3]. Indigenous cattle are primarily categorized into five different varieties or types namely Red Chittagong (RCC), Pabna (PC), Munshiganj (MC), North Bengal Grey (NBG), and Non-descript Deshi (DES), those possess distinct coat color as well as have differences in their productivity and morphometric features [2]. These varieties have been developed in their breeding tracts due to farmers’ selection over the years particularly for milk production and phenotypes. In particular, milk production is comparatively higher in RCC, MC, and PC than NBG and DES [1,4]. In addition, Sahiwal (SL) breed were introduced in Bangladesh six decades ago as an improved zebu dairy cattle and is now sparsely distributed throughout the country.In Bangladesh, the indigenous cattle alone predominated until the late seventies of the twentieth century. Upgradation of indigenous cattle varieties has been practiced for the last five decades to boost up milk production through the introduction of high yielding temperate and tropical dairy breeds [5]. However, indiscriminate upgrading and crossbreeding programs without proper conservation efforts have led to the production of upgraded animals’ exponentially at the expense of indigenous cattle varieties of Bangladesh [2,6]. Thereby, the population size of indigenous varieties declined rapidly, whereas MC and PC are at risk of extinction now [1]. Considering the above stated scenarios, ex situ conservation programs on PC, MC, and RCC are on-going at different government institutional herds in limited scales for more than a decade. More importantly, comprehensive information of current genetic diversity and demographic process of those indigenous populations are essential to take the necessary steps for development of sustainable breeding programs and maintenance of unique gene variants.Genetic diversity illustrates the key aspects of differentiation among the individuals of a population that exists either at phenotypic or DNA level [7]. The measures of genetic diversity provide essential insight information on livestock populations conservation and improvement strategies, as well as their adaptation to certain environments [8]. The landscape and depth of bovine genomic research has got a new dimension with the advent of SNP chip data. Genome wide SNP50K chip have been implemented to investigate genetic diversity, population structure, level of inbreeding, admixture analysis, effective population size, linkage disequilibrium, migration events, genome-wide association studies, and detection of selection signatures in different cattle populations around the world [9,10,11,12]. Estimation of genomic breeding value and its prediction accuracy is another frontline avenue in the genetic evaluation process using SNP chip data [13]. On the other hand, more accurate diversity parameter estimates have become possible now through utilization of genomic information compared to traditional pedigree data. However, the indigenous cattle population of Bangladesh remained poorly studied especially at the molecular level. Earlier genetic characterization studies were carried out in RCC population using microsatellite markers [14] and mitochondrial DNA [15]. Analysis of SNP80 K indicine commercial chip data revealed weak genetic differentiation between RCC and DES cattle [16]. It is noted that the comprehensive genetic diversity study utilizing all indigenous cattle genetic resources of Bangladesh has not yet been performed using high density SNP markers. Therefore, the present study aimed to investigate the extent of genetic diversity, population structure, and differentiation among the six indigenous cattle varieties of Bangladesh in comparison with other reference populations or breeds.2. Materials and Methods2.1. Ethics ApprovalFor this study, blood samples were collected from institutional, private, and farmers’ herds under the approval of Ethical Committee of Bangladesh Agricultural University, Bangladesh (no. 1218/BAURES/2020/ESRC/AH/10).2.2. Animal Sampling and DNA ExtractionA total of 240 blood samples were collected from six zebu cattle populations of Bangladesh (Figure S1). Among them, five indigenous cattle varieties namely RCC (n = 89), MC (n = 26), PC (n = 45), NBG (n = 21), and DES (n = 22) are distributed in different agro-ecological regions, while SL (n = 15) breed are being predominantly available in milk potential areas. Blood sampling was performed from government or institutional herds (Central Cattle Breeding and Dairy Farm, Bangladesh Livestock Research Institute, Savar, Dhaka), university managed herd (Bangladesh Agricultural University, Mymensingh, Bangladesh), private dairy farm (American Dairy Limited, Gazipur, Bangladesh), as well as from 37 smallholders’ farms (Figure S1). Precautions were taken to avoid sampling from the related individuals. Genotype of the sampled individuals was ascertained through pedigree analysis as well as in-depth enquiry of the animal owners. Blood samples were collected from jugular vein using vacutainer containing EDTA as anticoagulant and were immediately transferred to the laboratory for DNA extraction. Genomic DNA was extracted from the whole blood using the Prime PrepTM DNA isolation kit (GeNet Bio Co. Ltd., Daejeon, Korea). The concentration and purity of isolated DNA were measured by the Nanodrop spectrophotometer (Model ND2000, Thermo Fisher Scientific, Wilmington, DE, USA) prior to genotyping.2.3. SNP Genotyping and Quality ControlDNA samples were genotyped using Illumina bovine SNP50 v.3 BeadChip with the help of the commercial genotyping service provider (TNT Research Co. Ltd., Seoul, Korea). The cattle SNP50 chip possesses 53,218 SNPs that uniformly span over the entire bovine genome. Genome Studio® software (Illumina, San Diego, CA, USA) plugin PLINK v.1.9 was employed for genotypes calling. SNP filtering was performed using PLINK v.1.9 [17] based on the following exclusion criteria: Minor allele frequency (MAF) <0.01 and call rate <0.90. SNP filtering based on the Hardy–Weinberg equilibrium (HWE) was not performed since we expected HWE deviations in some of the studied populations due to their small and possibly sub-structured population and genetic drift. Furthermore, SNPs assigned to sex chromosomes and those lacking genomic locations were excluded from the analysis. Apart from this, genotyping data of Korean Hanwoo (KPN), Chikso (CHK), and Jeju Black (JB) were made available by Dr. Seung Hwan Lee, Chungnam National University, Daejeon, South Korea. In addition, Central Asian cattle breeds Yianbian (YBH) and Mongol (MG) data were downloaded from dryad.org [18]. Genotype information of African taurine (N’Dama, ND and Oulmes Zaer, OUL) and Zebu cattle (Zebu Madagascar, ZMA and Sheko, SHK), and European taurine breeds Angus (ANG), Brown Swiss (BSW), Hereford (HFD), Holstein (HOL), Limousin (LMS), Guernsey (GNS), Santa Gertrudis (SGT), and Beefmaster (BMA) were used from the Bovine Hapmap consortium. Details on breeds or cattle populations and number of samples used in this study are presented in Table S1. The final dataset included 723 individuals from 25 cattle populations/breeds and 35,964 SNPs in the merged data file.2.4. Statistical Analysis2.4.1. Genetic DiversityTo assess the within population genetic diversity, the proportion of polymorphic loci (PN), observed (Ho) and expected (He) heterozygosity were estimated using R package hierfstat [19]. The distribution of minor allele frequency (MAF) was grouped into six different categories based on their frequency as rare alleles (0 < maf ≤ 0.05), intermediate alleles (0.05 < maf ≤ 0.10), and common alleles (0.10 < maf ≤ 0.50). Diversity indices were calculated from 45,861 filtered SNPs for six zebu cattle populations of Bangladesh.2.4.2. Population Structure and Genetic DifferentiationThe following approaches were employed to investigate the population structure among the indicine cattle populations of Bangladesh, as well as to assess their relationships with other cattle breeds distributed globally, principal component analysis (PCA), admixture analysis, pairwise genetic differentiation (FST), and maximum likelihood phylogenetic tree construction. Two different datasets were used in PCA and admixture analysis, 17,477 SNPs that passed the quality control threshold from 723 individuals belong to 25 global cattle populations (Table S1), while 18,481 SNPs were used for 264 individuals of eight zebu cattle populations. To perform PCA, PLINK v.1.9 was used to generate eigenvectors and eigenvalues, and the outputs were visualized using the R package SNPRelate [20] that demonstrate the relationships among the PC1, PC2, and PC3 coordinates. Furthermore, the population genetic structure assessment was performed using ADMIXTURE v.1.3 software [21], assuming a number of hypothetical population clusters (K) ranging from 2 to 9 and the output was visualized using R plots. The optimum number of K value was obtained from the lowest cross-validation (CV) error estimation. The pairwise FST and Nei genetic distances among the populations were calculated using the R package StAMPP [22]. The Neighbor Joining (NJ) tree was constructed using the SNPhylo pipeline [23] to know the evolutionary relationships and was illustrated using FigTree v.1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/, accessed on 7 April 2020).2.4.3. Linkage Disequilibrium and Effective Population SizeLinkage disequilibrium (LD) was used to examine the recombination events of linked SNPs in each population and was measured as the correlation coefficient (r2) between two loci. SNPs spanning from 0 to 500 Kb distance were included in this step. The effective population size (Ne) was calculated at different generations from the resulted LD value according to the equation suggested by Sved [24]. The estimates of LD and Ne were performed using PLINK v.1.9 and SNeP v.1.1 [25] with default parameters, respectively. The Ne estimates were plotted over the last 0 and 10,000 generations using R software v.3.3.1 (R Foundation for Statistical Computing, Vienna, Austria) to investigate the diversity trends.3. Results3.1. Intra-Population Genetic DiversityAfter quality control procedures, 218 samples and 35,960 SNPs were remained in the final dataset for downstream analysis of Bangladeshi cattle populations. Twenty-two animals were excluded from the analysis due to low call rate (<0.90). The genetic diversity measures are given in Table 1. Most of the Bangladeshi indigenous cattle manifested a high proportion of polymorphic loci (PN), varying from 0.668 in SL to 0.905 in RCC. The highest proportion of SNPs that belong to MAF category of ≤0.05 was found in RCC (0.272), whereas the lowest proportion was shown in Sahiwal breed (0.120), with overall mean of 0.197 across populations. The highest observed heterozygosity was observed in RCC (0.250 ± 0.180), while the lowest was in PC (0.211 ± 0.166), indicating higher diversity in RCC compared with other indigenous cattle populations. However, small differences were observed regarding heterozygosity measures (Ho and He) among the studied cattle populations. Minor allele frequency distributions in six oindigenous populations under study based on different categories are shown in Figure 1. The percentage of fixed SNPs (MAF = 0.00) ranged between 16.6 and 31.5% in RCC and SL cattle, respectively, with an average of 24.2% across populations. Distribution of minor allele under low frequency category (0, ≤0.05) was significantly higher in all cattle populations except SL, also depicted from Table 1. However, the proportion of MAF did not differ largely among the cattle populations for the remaining category of [0.05, 0.1], [0.1, 0.2], [0.2, 0.3], [0.3, 0.4], and [0.4, 0.5].3.2. Population Structure and Genetic DifferentiationThe genetic structure among the indigenous cattle population of Bangladesh and 19 selected cattle breeds across the world were assessed through PCA using 17,477 SNPs. Cattle populations were separated at sub-species and region level (Figure 2). The PC1 explained 8.96% of the total variance that clearly separated taurine breeds from the indicine breeds/populations. Among the investigated indicine breeds or populations three major clusters were observed, the first one concatenating all Bangladeshi cattle populations, while the second and third clusters enclosing South Asian (NEL and BRM) and African (ZMA and SHK) indicine cattle, respectively (Figure 2A). The PC2 accounted for 3.06% of the variation, segregated European taurine from East and Central Asian taurine breeds. In addition, African taurine (ND and OUR) occupied an intermediate position in between European dairy (BSW, HFD, HOL, and GNS) and beef (BMA and SGT) breeds (Figure 2A and Figure S2). More detailed analyses using the indigenous cattle population of Bangladesh, PC1 and PC2 illustrated only 2.66 and 1.43% of the total variations, respectively. All individuals dispersedly distributed without formation of any specific cluster. However, some individuals of RCC clearly isolated from all other Bangladeshi populations, as depicted by PC2 (Figure 2B and Figure S2).The heat map shows Nei genetic distance and pairwise FST values in upper and lower diagonals, respectively among the 25 cattle breeds or populations (Figure 3). The lowest genetic differentiation was observed in Bangladeshi populations that ranged between 0.00 to 0.03 and also showed similarity with South Asian (BRM and NEL) and African (SHK and ZMA) zebu cattle breeds. The lowest FST values among the zebu cattle populations reflected their close relationship to each other. As expected, the highest differentiation (pairwise FST = 0.16 and Nei FST = 0.13) was obtained between Bangladeshi indigenous cattle populations and taurine cattle breeds of Africa (ND and OUL), East Asia (KPN, CHK and JB), and Europe (ANG, BSW, HFD, HOL, GNS, and LMS). However, Central Asian (YBH and MG) and European beef breeds (BMA and SGT) showed moderate genetic differentiation (pairwise FST = 0.08 to 0.11 and Nei FST = 0.08 to 0.14) with respect to Bangladeshi indicine populations.The admixture analysis inferred clustering patterns among Asian zebu cattle populations based on shared ancestry (at K = 3) that separates Bangladeshi Indigenous cattle from NEL cattle (Figure 4). The optimal K-value was acquired (K = 4) based on the lowest cross-validation error (Figure S4) that highlighted SL cattle shared major ancestry with Bangladeshi populations, where NEL and BRM clearly separated from them. On average, the Bangladeshi cattle had 56.0 to 86.0% similar genetic background with SL but only around 1.0 to 5.0% similarity was observed with NEL and BRM cattle (Table 2). More specifically, RCC populations demonstrated two distinct genetic backgrounds where 61% was similar with South Asian zebu cattle and the remaining 39% was accounted for population specific. Unexpectedly, 1–4% taurine ancestry (Table 2) identified between HOL and indigenous cattle population of Bangladesh might be inter se breeding. When another 11 European dairy and beef breeds (ANG, BMA, BSW, CHA, GNS, HFD, HOL, JER, LMS, PMT, and SGT) each possess 17,477 SNPs included with Bangladeshi indigenous cattle populations, K = 9 was found as the optimum K value (Figure S4). The admixture analysis at K = 2 grossly separated Bangladeshi indigenous cattle populations to European cattle breeds. At K = 5 and 9 each breed created their own cluster where all Bangladeshi cattle populations formed a separate cluster having minimum admixture of taurine dairy breeds (Figure S3). However, four beef breeds (BMA, LMS, PMT, and SGT) showed a certain level of zebu inheritance, while three of them (CHA, LMS, and PMT) manifested almost a similar genetic background. The degree of admixture declined with the increment of ancestry number through inclusion of 17 European cattle breeds to the present dataset. The maximum likelihood based phylogenetic tree showed evolutionary relationships among the Bangladeshi indigenous cattle, starting with the DES cattle without forming any specific cluster (Figure 5). However, some sub-clusters were noticed in the case of RCC, PC, and SL populations. Moreover, this illustration suggests their weak differentiation along with the historical exchange of genetic materials among the studied populations.3.3. Analyses of Linkage Disequilibrium and Effective Population SizeThe extent of LD was assessed up to 500 kb using pairwise r2 for six zebu cattle populations separately. The r2 values both at shorter and longer genetic distances varied among the cattle populations. The most rapid LD decaying pattern was observed in shorter distances of up to about 30 kb (Figure 6A). In general, levels of pairwise LD across the genome dropped with the advancing distance between adjacent SNPs. SL cattle displayed the higher LD value across the genomic distance, while RCC had the lowest LD value. The LD values (average r2 in 200 to 500 Kb fragments) were 0.12, 0.08, 0.09, 0.07, 0.06, and 0.04 for SL, DES, NBG, PC, MC, and RCC, respectively. The effective population size was evaluated based on LD estimates (r2) from the recent generation to 10,000 generations ago (Figure 6B). All six cattle populations showed sharp declining trends in their Ne values over time. Among the populations, Ne ranged from 494 to 2058 animals 100 generations ago. However, in the recent past (until five generations ago), RCC had the highest Ne (108.29), while the lowest value (26.02) was found in SL and the intermediate estimates were found to be 47.68, 40.87, 33.69, and 46.64 in MC, DES, NBG, and PC, respectively, representing a narrow genetic pool in the studied populations.4. DiscussionIndigenous cattle populations are mostly random bred and non-selected genetic resources that harbor unique gene pools resulting from their adaptation to the local environment [13]. Hence, more diverse alleles are expected in those populations due to the absence of intense directional selection for production traits. In this study, genome-wide SNP data were analyzed to advance our knowledge on genetic architecture and diversity of Bangladeshi indigenous cattle in the worldwide population context.The results of this study depicted relatively low genetic diversity measures in terms of PN, Ho, and He in Bangladeshi cattle populations and are in agreement with the previous reports of Xu et al. [12], Mustafa et al. [26], and Zhang et al. [27]. Previous studies also reported relatively higher diversity measures in taurine breeds compared to their indicine counterparts using the same genotyping array [9,11,13]. In fact, the low representation of indicine breeds in the SNP genotyping array results in the ascertainment bias towards taurine breeds that disproportionate MAF distribution among the Asian and African indicine breeds or subpopulations [11,28]. In addition, SNP filtering from combined datasets of 25 cattle breeds/populations kept a low number of SNPs to be investigated. However, the SNP genotyping and quality control methods are comparable with the previous studies in different indicine cattle populations where minimum bias on their results have been reported [26,29]. Similar to the present findings, the average minor allele frequency (MAF) ranging from 0.11 to 0.23 that was observed in three South Asian zebu breeds (Gir, Sahiwal, and Nellore), 10 indicine breeds of Pakistan, and two Chinese indicine cattle Wenshan and Nandan [12,26,27]. In earlier studies, Uzzaman et al. [16] and Edea et al. [29] found a bit higher MAF (0.28 to 0.31) in two Bangladeshi (RCC and DES) and three Ethiopian (Begait, Guraghe, and Ogaden) zebu populations. In our study, the MAF distribution pattern was consistent with Chagunda et al. [11] but contradicted with the reports of Pérez O’Brien et al. [30] and Bejarano et al. [31] who found a higher proportion of common alleles (≥0.10) in taurine cattle as compared to indicine breeds. The higher percentage of low MAF associated with greater genetic diversity, affects LD distribution and extends within the population [30,32], has larger effects and better genomic predictive ability for quantitative traits in cattle [31], and supports the present findings.On the other hand, the estimated heterozygosity values of this study (average Ho = 0.22 ± 0.17 and He = 0.19 ± 0.12) were in agreement with most of the previous studies involving different indicine populations. Sharma et al. [9] reported Ho and He levels in Brahman, Nellore, and Gir breeds ranging from 0.20 to 0.22 and 0.15 to 0.18, respectively. The current results are also consistent with those reported by Zhang et al. [27] and Chagunda et al. [11] in Asian (Gir, Sahiwal) and African (East African Shorthorn Zebu) zebu cattle that ranged between 0.22 and 0.27. However, worldwide distributed taurine breeds presented relatively higher MAF (ranging from 0.23 to 0.31), PN (varies from 0.80 to 0.97), Ho (0.30 and 0.39), and He (0.24 to 0.38) compared to the present findings as reported by Sharma et al. [9], Kim et al. [10], and Upadhyay et al. [13].Several statistical approaches have been utilized to infer relationships among the cattle populations as well as to assign individuals in their respective populations. The principal component analysis (PCA) using the genotyped dataset of this study along with 19 worldwide distributed cattle populations’ data revealed that Bangladeshi indicine cattle clustered with South Asian zebu cattle and a bit away from African zebu populations. Earlier SNP-based studies in cattle and sheep demonstrated that first (PC1) and second (PC2) components of PCA clustered breeds or populations were based on their geographic origin [33]. Likewise, our PCA analysis demonstrated geographic origin and sub-species oriented clustering that depict the recent division of Bangladeshi cattle from their primary center of zebu domestication, the Indus Valley and is in accordance with the findings of Chen et al. [34]. The major influence of South Asian zebu to Bangladeshi indigenous cattle was also ascertained through the genetic composition analysis (Table 2), as well as from the cattle breeding history of Bangladesh. During the British colonial period, Indian zebu male mediated introgression had been employed to improve the small sized indigenous cattle of Bangladesh [15] that also supports the present SNP data-based findings.In our study, the clustering pattern in PCA (Figure 2B), pairwise FST and Nei genetic distance (Figure 3), and population STRUCTURE analysis (Figure 4) indicate weak differentiation among Bangladeshi cattle populations. The absence of well-defined but overlapped clusters in PCA is probably due to the historical gene flow among themselves. In fact, the indigenous cattle population was a large random bred population over the centuries. They have been categorized in the recent past based on coat color and morphometric features, where sufficient genetic differentiation has not yet been established through within population selective breeding [1,14]. However, one third of individuals belong to RCC clearly separated from other zebu populations that reveals a genetically distinct population within RCC. This finding is supported by the mtDNA based study of Bhuiyan et al. [15] who found two separate clusters in RCC population that might be due to ex situ conservation through closed herd nucleus breeding and thus created genetic drift in the isolated population. It is noted that blood sampling was performed both from in and ex situ individuals for the said populations. Similar to our findings, Ethiopian cattle populations had low FST values of 0.011 to 0.012 [35] using microsatellite markers and low-density taurine derived chip. However, higher FST values were observed between Gir and EASZ (0.11) breeds [11], Ethiopian and Asian zebu (0.07) populations [29], between Deoni and Ongole (0.12) breeds of India [36]. In addition, Shah et al. [37] reported the low FST estimate between Kankrej and Malvi cattle breeds of India (0.013) and is similar to our results. Earlier FST estimates by Sharma et al. [9] observed genetic closeness among Asian zebu breeds compared to European, African zebu, and taurine cattle and supports our findings. Taken together, the lowest genetic distance observed among the Bangladeshi indigenous cattle populations suggest that they did not differentiate well as an independent breed and might be due to their common ancestral origin and exchange of genetic materials in the recent past [29]. This phenomenon is also supported from the phylogenetic results (Figure 6), where the absence of population specific clusters suggests strong gene flow among each other.Similar to the PCA, consistent results were also obtained from the admixture analysis. Among the zebu populations, the BRM and NEL clearly separated from six Bangladeshi cattle starting at K = 3. The distinct genetic composition of BRM and NEL might be their geographic isolation and long-term selection for desired traits [18]. However, a small proportion of zebu admixture in the indigenous cattle variety of Bangladesh probably is due to their common ancestral origin. When SNP information of worldwide distributed cattle was included to present the dataset, clear divergence was noticed between Asian zebu (Bos indicus) and taurine (Bos taurus) cattle at all K values starting from K = 2 to 9. Our findings are in agreement with previous studies that support two independent domestication events that took place for two bovine species [18,38]. Moreover, taurine beef breeds mostly genetically admixed those possessed both indicine and taurine ancestry and is consistent with the findings of Decker et al. [18] and Zhang et al. [27]. However, the very low level of taurine admixture in Bangladeshi cattle population could be explained from their breeding history, where remnants of taurine introgression through indiscriminate breeding still exists.Based on the average r2 value, a slower LD decay across the distance in SL cattle indicates its heterozygosity deficit compared to other five zebu populations. The lowest LD in RCC reflects a more diverse population with weak directional selection. In addition, the sharp decline of LD in short distance (Figure 6A) is also an indication of high haplotype diversity in the studied populations. In our study, LD values decrease in a constant manner as the distance among markers increase and the similar trend was reported in cattle by Kim et al. [10] and Shin et al. [39]. Usually, LD characteristics are population or breed specific despite the fact that several factors are associated with LD patterns and scale within or between populations such as demographic history and population structure, sample size, marker type and density, MAF thresholds, selection and method of LD measurement [40]. The estimated effective population size (Ne) was relatively low in the studied population except RCC. However, our results are comparable with the findings of Zhang et al. [27] who found that Ne values varied from 10 to 259 in 17 Chinese indigenous cattle breeds. In addition, Ne of three Korean cattle populations (BH, BRH, and JB) until 13 generations ago were 83, 59, and 67, respectively [9] and agrees with this study, but higher Ne values as 260, 202, and 55 were reported by Kim et al. [10] for those aforementioned genotypes. Importantly, the methodological aspects could potentially affect Ne estimates from the LD pattern, particularly LD from r2 estimates. For instance, a small sample size leads to bias LD estimation and therefore, a minimum sample size of 55 was suggested for accurate LD estimation based on r2 values [41]. The lower Ne values of the present study might be associated with the relatively small sample size and low number of filtered SNPs included in the analysis. Altogether, LD patterns and Ne reflect various demographic and selection events, as well as bear significance to understand the genomic architecture of a breed or population [42].5. ConclusionsThis study provided an important glimpse of genetic diversity, population differentiation, and structure among indigenous cattle of Bangladesh for the first time using high density genome-wide SNP data. In this study, we found relatively low genetic diversity measures that were comparable to other zebu populations worldwide. Furthermore, Bangladeshi cattle population did not differentiate well as independent breeds, revealed by their low FST values. A certain amount of zebu and taurine admixture in the investigated populations might be due to their common ancestral origin and the cattle breeding history of Bangladesh. The lack of well-defined and overlapped clusters in PCA and phylogenetic tree highlight the historical gene flow among themselves and the absence of strong directional selection over the generations. Taken together, our findings reported herein the basic features of genomic architecture of Bangladeshi indigenous cattle populations that will aid in their future conservation and genetic improvement strategies and programs.

animals : an open access journal from mdpi

10.3390/ani13061043

PMC10044354

Prion diseases are fatal and incurable neurodegenerative disorders affecting both humans and animals. The development of in vitro cellular models from naturally susceptible species such as ruminants can be very useful for the study of prion disease mechanisms and the discovery of potential therapies. Our study shows for the first time how the culture, in the form of three-dimensional spheroids of ovine mesenchymal stem cells derived from bone marrow in growth and neurogenic conditions, makes these cells more permissive to prion infection, mimicking the prion toxicity occurring in these diseases. This three-dimensional system appears as a potential in vitro model for studying prion diseases in a microenvironment approaching in vivo conditions.

In neurodegenerative diseases, including prion diseases, cellular in vitro models appear as fundamental tools for the study of pathogenic mechanisms and potential therapeutic compounds. Two-dimensional (2D) monolayer cell culture systems are the most used cell-based assays, but these platforms are not able to reproduce the microenvironment of in vivo cells. This limitation can be surpassed using three-dimensional (3D) culture systems such as spheroids that more effectively mimic in vivo cell interactions. Herein, we evaluated the effect of scrapie prion infection in monolayer-cultured ovine bone marrow-derived mesenchymal stem cells (oBM-MSCs) and oBM-MSC-derived spheroids in growth and neurogenic conditions, analyzing their cell viability and their ability to maintain prion infection. An MTT assay was performed in oBM-MSCs and spheroids subjected to three conditions: inoculated with brain homogenate from scrapie-infected sheep, inoculated with brain homogenate from healthy sheep, and non-inoculated controls. The 3D conditions improved the cell viability in most cases, although in scrapie-infected spheroids in growth conditions, a decrease in cell viability was observed. The levels of pathological prion protein (PrPSc) in scrapie-infected oBM-MSCs and spheroids were measured by ELISA. In neurogenic conditions, monolayer cells and spheroids maintained the levels of PrPSc over time. In growth conditions, however, oBM-MSCs showed decreasing levels of PrPSc throughout time, whereas spheroids were able to maintain stable PrPSc levels. The presence of PrPSc in spheroids was also confirmed by immunocytochemistry. Altogether, these results show that a 3D culture microenvironment improves the permissiveness of oBM-MSCs to scrapie infection in growth conditions and maintains the infection ability in neurogenic conditions, making this model of potential use for prion studies.

1. IntroductionCell cultures have proven to be useful tools for a variety of applications in research. There are currently many established cell lines and primary cultures available and widely used for different purposes. The two-dimensional (2D) platforms in which flat monolayer cells are cultured are most commonly used for research in cell-based assays [1]. These 2D cell culture systems are accessible, convenient, and cost-effective [1]. However, various limitations are still of concern, including the failure to imitate the in vivo architecture and microenvironments. Compared with in vivo cells, 2D-cultured cells possess many different features, such as morphological characteristics, proliferation and differentiation potentials, cell–cell and cell-surrounding matrix interactions, and signal transduction [1,2,3].To overcome these drawbacks, three-dimensional (3D) cell culture platforms have emerged as a promising approach. Although the optimal 3D condition requirements vary between cell types, and the characteristic features of cells in 3D cultures differ in accordance with their types, these 3D culture systems have proven to be more realistic for translating cell-based assay findings to in vivo applications due to their ability to closely mimic the behavior of in vivo cells [4]. There are different 3D culture systems, including spheroids, organoids, hydrogel embedding, bioreactors, scaffolds, and bioprinting [1,5], with potential applications in drug discovery, disease modeling, and tissue engineering [6,7,8].Spheroids are spherical-shaped multicellular aggregates that can be formed from several types of cells, including mesenchymal stem cells (MSCs) [9]. They are able to mimic cell–cell and cell–matrix interactions more effectively than 2D cultures, but they lack the capacity to recapitulate the tissue organization exhibited in vivo [10]. In MSCs, spheroid formation enhances the characteristics of these cells by improving their stemness, facilitating the differentiation to multiple lineages, and delaying the in vitro replicative senescent processes [11,12,13].In the field of neurodegenerative diseases, 3D cell culture platforms appear as good models to reproduce different features of neurodegeneration linked to these diseases. In Alzheimer’s disease (AD), 3D hydrogel-embedded human neural stem cells [14,15], neural progenitor cells [16,17], and scaffold-encapsulated induced pluripotent stem cell (iPSC)-derived neural progenitor cells [18] can recapitulate amyloid-beta aggregation and accumulation of hyperphosphorylated tau, which are key hallmarks of this disease. Furthermore, a triculture model using neurons, astrocytes, and microglia, constructed in a microfluidic platform, also showed these AD hallmarks and allowed the study of microglia recruitment, neuroinflammatory response, and neuron/astrocyte damage [19]. Human iPSC-derived hippocampal spheroids [20] and brain organoids [21,22,23] have been developed, which also have the ability to mimic AD’s pathology and the potential to be used for screening therapeutic strategies. In Parkinson’s disease (PD), dopaminergic neurons [24,25] and SH-SY5Y neuroblastoma cells [26,27] cultured in matrigel-based platforms can model several PD features, such as α-synuclein accumulation and Lewy body-like inclusions. Midbrain organoids carrying the LRRK2 G2019S mutation have also been created to successfully model LRRK2-associated sporadic PD [28].Prion diseases, or transmissible spongiform encephalopathies (TSEs), are neurodegenerative disorders caused by a pathological misfolded protein derived from an innocuous cellular prion protein (PrPC) called PrPSc [29]. These diseases occur in humans and animals, and among the various types of TSEs, the one affecting sheep and goats, known as scrapie, was the first to be discovered. It is considered a good model for studying different disease aspects in these pathologies [30,31,32]. Regarding 3D culture methods, only two studies using human cerebral organoids have been reported. In these studies, iPSC-derived human cerebral organoids were able to uptake and propagate sporadic Creutzfeldt–Jakob disease (sCJD) prions [33], and also responded to an anti-prion compound by showing delayed prion propagation [34].A previous study conducted by our group evaluated the effect of scrapie prion infection in ovine bone marrow-derived mesenchymal stem cells (oBM-MSCs) cultured in growth conditions and subjected to neurogenic differentiation [35]. In this study, oBM-MSCs initially took up PrPSc, but they were not able to maintain it over time. MSC-derived neuron-like cells, on the contrary, absorbed and maintained stable PrPSc levels throughout the duration of the culture [35]. These results were obtained using a 2D culture system approach that may not exactly mimic the physiological response of these cells to prion infection. Therefore, in the current study, we decided to assess the effect of scrapie prion infection in oBM-MSCs-derived spheroids cultured in growth and neurogenic conditions, analyzing their ability to maintain prion infection and the impact of this infection on cell viability.2. Materials and Methods2.1. Bone Marrow Extraction and Ovine Mesenchymal Stem Cell Isolation and CultureA bone marrow sample was obtained from an adult female rasa aragonesa sheep of two years of age who carried the ARQ/ARQ genotype for the PRNP gene. This sheep belonged to a flock from the Center of Encephalopathies and Emerging Transmissible Diseases (CEETE; University of Zaragoza), maintained for research purposes. After the sedation with xylazine and the local anesthesia with lidocaine, bone marrow aspirate was collected from the humeral head as previously described [36]. All procedures were approved by the Ethical Committee for Animal Experiments from the University of Zaragoza (project license PI44/18) and were in accordance with the Spanish Policy for Animal Protection, RD53/2013, and the European Union Directive 2010/63.MSC isolation from the bone marrow aspirate (3 mL) was performed following the previously described protocol [36,37,38]. This protocol is based on the separation of the mononuclear fraction after density gradient centrifugation in Lymphoprep (Atom) and further isolation thanks to the ability of MSCs to adhere to plastic. After isolation, cells were expanded up to passage 2 in a basal medium consisting of low glucose Dulbecco’s modified Eagle’s medium (DMEM, Sigma-Aldrich, St. Louis, MO, USA), supplemented with 10% fetal bovine serum (FBS), 1% L-glutamine (Sigma-Aldrich), and 1% streptomycin/penicillin (Sigma-Aldrich).All the subsequent experiments were performed using this oBM-MSC culture. The number of technical replicates used in each experiment is shown in Supplementary Table S1.2.2. Ovine Mesenchymal Stem Cell CharacterizationThe minimal criteria to characterize MSCs are their plastic-adherence capacity in standard culture conditions and their ability to differentiate to mesodermal lineages (adipocytes, osteoblasts, and chondroblasts) in vitro [39].Adipogenic, osteogenic, and chondrogenic differentiation was evaluated in vitro. The differentiation into mesodermal lineages was performed using specific commercial differentiation kits for osteogenic (StemPro® Osteogenesis Differentiation Kit, Gibco, Life Technologies, Waltham, MA, USA) and chondrogenic (StemPro® Chondrogenesis Differentiation Kit, Gibco, Life Technologies) differentiation. For adipogenic differentiation, basal medium supplemented with 1 μM dexamethasone, 500 μm 3-Isobutyl-1-methylxanthine (IBMX), 200 μM indomethacin, and 15% of rabbit serum was used [38,40]. Subsequently, adipogenic differentiation was confirmed by 0.3% Oil Red O (Sigma-Aldrich) staining after 8 days of differentiation; Alcian Blue G dye (1: 1 in methanol) (Sigma-Aldrich) was used to confirm chondrogenic differentiation at day 15; and osteogenesis was verified with 2% Alizarin Red S staining (Sigma-Aldrich) after 21 differentiation days.2.3. Formation and Culture of Ovine Mesenchymal Stem Cell SpheroidsFor spheroid formation, 96-well Nuclon Sphera low-attachment plates (ThermoFisher Scientific, Waltham, MA, USA) were used. In each well, 45,000 cells were seeded to form a unique spheroid per well. The cells were seeded in droplets, and afterwards, 200 μL of basal medium was added into each well. To ensure a correct formation of the spheroids, the plates were incubated for 5 days at 37 °C with 5% CO2 without changing the medium. Afterwards, the medium was changed every 48–72 h. Seven days after formation, the spheroids were stabilized and ready to perform the subsequent assays.2.4. Neurogenic Differentiation of Ovine Mesenchymal Stem Cells and Spheroids2.4.1. Neurogenic DifferentiationOvine BM-MSCs were seeded at 5000 cells/cm2 and differentiated into neuron-like cells using a HyClone AdvanceSTEMTM Neural Differentiation Kit (ThermoFisher Scientific). After 24 h of culture under neurogenic conditions, cells showed a neuron-like morphology, which was monitored by optical microscopy. The pick of neurogenic differentiation was observed at 72 h, as previously described [36]. To maintain the cells in the differentiation state, the medium was changed every 48 h.For the neurogenic differentiation of spheroids, after they were stabilized, the same neural induction kit was used as that for the 2D cultures, following the same guidelines for medium change.2.4.2. Nissl Bodies StainingTo verify the correct neurogenic differentiation, neuron-like cells and spheroids were stained with 1% Cresyl Violet Solution (abcam) to detect Nissl bodies, which are granular structures present in neuronal cell bodies and composed of rough endoplasmic reticula rich in RNA [41]. The staining was performed after 3 days of neurogenic differentiation.2.4.3. Expression Analysis of Neuronal MarkersThe expression of a set of neuronal markers was assessed by real-time quantitative PCR (RT-qPCR) in both neuron-like cells and spheroids. These markers were NEFM (neurofilament medium chain), MAP2 (microtubule-associated protein 2) and TUBB3 (tubulin beta 3 class III) [37]. HPRT (hypoxanthine phosphoribosyltransferase) and G6PD (glucose-6-phosphate dehydrogenase) were used as housekeeping genes [37].For total RNA extraction from 2D cell cultures and further retrotranscription to cDNA, the Cells-to-cDNATM II kit (ThermoFisher Scientific) was used. RNA from each of the two technical replicates was retrieved from two wells of a P6 culture plate.Total RNA was extracted from 42 pooled spheroids using the Direct-zolTM RNA Miniprep kit (Zymo Research, San Diego, CA, USA). Two technical replicates were analyzed. The quality and quantity of these RNA samples were checked with a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and a Qubit 4.0 fluorometer (Life Technologies, Carlsbad, CA, USA). cDNA was then obtained using the qScriptTM cDNA SuperMix (QuantaBio). All procedures were performed following the manufacturer’s instructions.Gene expression was quantified by RT-qPCR using the Fast SYBRTM Green Master Mix (Applied Biosystems, ThermoFisher Scientific, Waltham, MA, USA) in a QuantStudio 3 Real-Time PCR instrument (Applied Biosystems). A dissociation curve was performed after every RT-qPCR reaction to confirm the amplification of a single amplicon. The primer design in different amplicons avoided genomic DNA amplification. The comparative quantification of the results was standardized by the 2−∆∆Ct method [42], using the geometric mean of the HPRT and G6PD housekeeping genes as a normalizer.2.5. Scrapie Inocula and Infection of 2D Ovine Mesenchymal Stem Cell Cultures and Spheroids in Growth and Neurogenic ConditionsTwo different inocula were created using central nervous system (CNS) samples: one from a healthy sheep (negative control) and one from a classical scrapie-infected sheep, both carrying the ARQ/ARQ genotype. The inocula were preserved at the tissue bank of the CEETE, University of Zaragoza. The presence/absence of PrPSc in the CNS tissues was confirmed with two rapid diagnostic tests (Prionics-CheckWestern blotting; ThermoFisher Scientific and Idexx HerdChek; IDEXX, Westbrook, ME, USA) as previously descried [43]. CNS homogenates and microbe safety tests were prepared as described in previous works [35].The effect of prion infections and their ability to replicate prions were investigated in 2D oBM-MSC cultures and spheroids in growth and neurogenic conditions. Ovine BM-MSCs were seeded at 5000 cells/cm2, and each spheroid was generated from 45,000 oBM-MSCs. For each condition, three groups were established: positive (infected with scrapie inoculum), negative (with inoculum from healthy sheep), and control (without inocula). For infection, basal medium was substituted by inocula diluted 1:10 in DMEM medium (10% FBS, 1% L-glutamine and 1% streptomycin/penicillin) for the 2D oBM-MSCs and spheroids in growth conditions, and in HyClone medium for the 2D oBM-MSCs and spheroids subjected to neurogenic differentiation. Cells and spheroids were maintained in this medium for 48 h to analyze cell viability and prion replication ability. Afterwards, the medium was changed twice a week.2.6. PrPSc Detection2.6.1. ELISA (Enzyme-linked Immunosorbent Assay)To test the ability to replicate PrPSc of 2D oBM-MSCs and spheroids inoculated with scrapie-infected brain homogenate in growth and neurogenic conditions, the amount of pathogenic protein was quantified by ELISA at 2, 5, and 8 dpi using the EEB-Scrapie HerdCheck kit (IDEXX) and following the manufacturer’s recommendations. A previous study conducted by our group showed that this kit was suitable for the sensitive detection of PrPSc in oBM-MSC cultures [35]. oBM-MSCs were seeded at 5000 cells/cm2 in 6-well plates, and the retrieval of the cells was performed by means of trypsinization and subsequent centrifugation to obtain 3 technical replicates per condition and time (dpi). Spheroids formed from 45,000 cells were also collected, and 3 spheroids/technical replicates were analyzed per condition and time (dpi).2.6.2. Immunocytochemistry of Infected SpheroidsThe presence of PrPSc in spheroids was also confirmed by immunocytochemistry at 5 dpi. Spheroids in growth and neurogenic conditions were fixed in 4% paraformaldehyde and stained with 0.2% eosin. Each spheroid was then included in a 1% agarose matrix; subsequently, the matrices containing the spheroids were embedded in paraffin. Afterwards, paraffin-embedded spheroids were cut to obtain 3-micrometer-thick slices. Hematoxylin–eosin staining was performed to evaluate the integrity of the spheroids after their inclusion in paraffin.For immunocytochemistry, after deparaffination and rehydration, spheroid sections were digested with 4 µg/mL of proteinase K for 15 min at 37 °C, then subjected to antigen retrieval with citrate buffer (pH 6.0) for 10 min at 96 °C in a PTLink (Dako). Afterwards, endogenous peroxidase activity was blocked using a precast solution (Dako Agilent, Glostrup, Denmark). Sections were then incubated for 30 min at room temperature with the primary antibodies L42 (1:500, R-Biopharm, Darmstadt, Germany) and F89 (1:2000, R-Biopharm, Darmstadt, Germany). Omission of the primary antibody served as a background control for nonspecific binding of the secondary antibody. Subsequently, the sections were incubated with an anti-mouse enzyme-conjugated EnVision polymer (Dako Agilent, Glostrup, Denmark) for 30 min at room temperature. Diaminobenzidine (DAB, Dako Agilent, Glostrup, Denmark) was used as the chromogen. The spheroid sections were assessed and photographed using a Zeiss Axioskop 40 optical microscope (Zeiss, Jena, Germany) and AxioVision Rel.4.7 software.2.7. Cell Viability AssayAn MTT assay was performed to evaluate cell viability and early prion toxicity in oBM-MSCs and neuron-like cells at 2, 5, and 8 days post-inoculation (dpi). Inoculum removal was performed at 2 dpi considering the day of infection with the inocula as day 0. Cell viability in oBM-MSCs and neuron-like cells was studied in positive, negative, and control cultures at 2, 5, and 8 dpi, coinciding with the day of inoculum removal, 2 dpi. Cells were seeded in 96-well plates at 5000 cells/cm2, and 8 technical replicates were analyzed for each condition and time (dpi). The MTT assay was carried out as previously described [35].The viability after prion infection was also analyzed in spheroids in both growth and neurogenic conditions by MTT assay at 2, 5, and 8 dpi in 3 conditions (positive, negative, and controls). Spheroids were formed from 45,000 cells per spheroid in 96-well Nuclon Sphera low-attachment plates (ThermoFisher Scientific), and 4 technical replicates were analyzed for each condition and each time point (dpi). An MTT assay was performed using the same reagents as in oBM-MSCs and neuron-like cells, but the incubation times were different: spheroids were incubated at 37 °C for 24 h with MTT solution, and for 4 h at room temperature while protected from light with HCl solution [44].In both cases, differences in cell viability were assessed with Student’s t-test, defining p < 0.05 as a statistically significant difference.3. Results3.1. Ovine Mesenchymal Stem Cell Differentiation into Mesodermal LineagesTo characterize oBM-MSCs, their ability to differentiate into adipocytes, chondrocytes, and osteocytes was assessed.Cells under adipogenic conditions presented red cytoplasmatic lipid droplets (Figure 1a,b). In chondrogenic conditions, cells aggregated in nodule-like formations (Figure 1c,d) and red-stained calcium deposits were found in osteogenic conditions (Figure 1e,f).3.2. Spheroid FormationAs shown in Figure 2 and Figure 3, after 7 days of culture, the size of the spheroids was stabilized and 1 spheroid per well was formed. The sizes of the final spheroids ranged between 392 and 427 μm.3.3. Neurogenic Differentiation3.3.1. Ovine Mesenchymal Stem Cell 2D CulturesThe differentiation of oBM-MSCs into neuron-like cells was assessed by Nissl bodies staining in cells subjected to neurogenic differentiation, as well as by analyzing the expression of three neuronal markers: NEFM, MAP2, and TUBB3.The presence of Nissl bodies in the soma of neuron-like cells was confirmed using Cresyl Violet staining (Figure 4). Regarding the expression of neuronal markers, TUBB3 and MAP2 displayed higher expression levels in neuron-like cells in comparison to the cells in growth conditions. NEFM, conversely, showed lower expression levels in cells subjected to neurogenic differentiation (Table 1). Overall, the presence of Nissl bodies and the increased expression of two neuronal markers confirmed the neurogenic differentiation of oBM-MSCs.3.3.2. SpheroidsThe evaluation of neurogenic differentiation in spheroids was carried out in the same way as in oBM-MSCs: Cresyl violet staining was performed along with the expression analysis of NEFM, TUBB3, and MAP2 neuronal markers.Together with the presence of Nissl bodies, spheroids under neurogenic differentiation conditions displayed a kind of prolongation on its contour that could be associated with neuronal morphology (Figure 5). Moreover, in neuron-like differentiated spheroids, the expression of the three neuronal markers was increased, albeit slightly, compared to the spheroids in basal conditions (Table 2). Therefore, the neurogenic differentiation of spheroids was also confirmed.3.4. Levels of PrPSc after Prion Infection3.4.1. Ovine Mesenchymal Stem Cells and Neuron-like Cells Infected with ScrapieThe PrPSc signal was measured after the infection with scrapie brain homogenate in oBM-MSCs and neuron-like cells at 2, 5, and 8 days post-inoculation. As shown in Figure 6, the levels of PrPSc in oBM-MSCs cultured in 2D decreased over time, whereas in neuron-like cells, the levels were stably maintained throughout the study period, with a slight decrease at 8 dpi.3.4.2. Spheroids in Growth and Neurogenic Conditions Infected with ScrapieThe levels of PrPSc detected by ELISA were also analyzed in scrapie-infected spheroids in growth and neurogenic conditions. Unlike 2D oBM-MSCs, spheroids in growth conditions showed stable levels of PrPSc over time (Figure 7a). In spheroids cultured in neurogenic conditions, a slightly decrease in the PrPSc signal was found at 5 dpi, but later, t 8 dpi, PrPSc levels were recovered a (Figure 7b). The presence of PrPSc in scrapie-infected spheroids was also confirmed by immunocytochemistry (Figure 8).3.5. Viability after Prion Infection3.5.1. oBM-MSC and Neuron-like Cell 2D CulturesThe effect of prion infection on cell viability was assessed at three infection times. oBM-MSCs infected with scrapie-positive inoculum showed increased viability compared to the cells infected with negative inoculum and the non-inoculated controls at 2 dpi, whereas at 5 dpi, a significant decrease in viability was observed in scrapie- and negative-infected cells compared to the controls. At 8 dpi, the levels of viability in the cells infected with positive and negative inocula were recovered, and were similar to the non-inoculated controls (Figure 9a).In neuron-like cells, an initial increase in cell viability was also observed in scrapie-infected cells along with a significant decrease at 5 dpi. Conversely to oBM-MSCs, this decrease was maintained at 8 dpi (Figure 9b).3.5.2. Spheroids in Growth and Neurogenic ConditionsCell viability was also assessed in spheroids in both growth and neurogenic differentiation conditions.In growth conditions, spheroids infected with the negative inoculum displayed increasing cell viability over time. The ones infected with positive inoculum, however, showed an increased cell viability at 5 dpi, but afterwards, at 8 dpi, the viability decreased greatly (Figure 10a).In neurogenic conditions, spheroids infected with the negative inoculum also showed high levels of cell viability at the three infection times. Contrary to what was observed in spheroids in growth conditions, spheroids infected with scrapie-positive inocula displayed increments in cell viability over time (Figure 10b).4. DiscussionThree-dimensional in vitro culture systems have arisen as a novel approach for cell cultures. These systems are able to more effectively mirror the microenvironment and interactions of the cells in vivo, making them useful tools for creating in vitro disease models that reliably reproduce disease-inherent characteristics. Our group previously evaluated the effect of scrapie prion infection in oBM-MSCs cultured in 2D in basal and neurogenic conditions [35]. In order to determine whether the three-dimensional conditions induce changes in the way in which these cells react to prion infection, in the present study, we assessed the effect of scrapie prion infection in oBM-MSCs and oBM-MSC-derived neuron-like cells, cultured in two-dimensional monolayer conditions and as spheroids, analyzing the ability to maintain or propagate PrPSc and possible prion toxicity that could affect cell viability.Mesenchymal stem cells were the first to be characterized. Their plastic adhesion ability and the tri-lineage differentiation confirmed the nature of these cells [37,39]. The ability to differentiate in vitro into neuron-like cells was also confirmed by morphological analysis in 2D cultures and Nissl staining in both 2D cultures and spheroids. However, the expression of neuronal markers was different between 2D and 3D conditions. Although we cannot discard a potential effect of the RNA isolation method on these differences, in both 2D and 3D cultures, basal and differentiating conditions were treated with the same methodology and the amplification conditions did not allow the amplification of genomic DNA that could possibly cause contamination. It has been reported that undifferentiated human MSCs from different sources, including bone marrow, are able to express different neuronal markers including TUBB3, NEFM, and MAP2, which would explain the slight increase in the expression of some neurogenic markers between basal and neurogenic conditions. This fact implies that the expression analysis of neuronal markers should be complemented with other techniques to evaluate the neural differentiation of MSCs, such as morphological changes and specific stains [45,46], as we already did. The overexpression of MAP2 was much lower in differentiated spheroids than in 2D cultures; variations in the expression levels of MAP2 and TUBB3 between 2D and 3D cultures have also been described in the neural induction of neural progenitor cells obtained from human iPSC [47].Murine bone marrow-derived MSCs can be infected and propagate the Fukuoka-1 human prion strain [48,49], and a persistent propagation of the mouse-adapted variant CJD and Fukuoka-1 strains has been described in spleen-derived murine stromal cells [50]. In accordance with our previous study [35], we observed a decline in PrPSc levels in 2D cultures of oBM-MSCs infected with scrapie throughout the duration of the culture, whereas neuron-like cells displayed steady PrPSc levels over time. These results confirm that oBM-MSCs cultured in two-dimensional conditions are less permissive to scrapie infection compared to neuron-like cells, which seem to be able to uptake and maintain scrapie infection over time.Regarding 3D in vitro models for prion diseases, only human cerebral organoids have been described to uptake and propagate sCJD prions [33]. Within MSC-derived spheroids, a specific microenvironment more similar to in vivo conditions is formed in which the diffusion inwards of nutrients and gases, and outwards of metabolic wastes, is limited [51,52]. Moreover, a decreased cell size, cell cycle quiescence and reduced energy metabolism have been found in MSC spheroids compared with MSCs in monolayer conditions [51,53,54]. The differences in the microenvironment of MSCs in 2D and 3D conditions could explain the distinct results observed between oBM-MSCs infected with scrapie on a culture plate and scrapie-infected spheroids in growth conditions. Contrary to 2D cultures, spheroids in growth conditions were able to maintain stable levels of PrPSc over time. On the other hand, like neuron-like cells, spheroids in neurogenic conditions also took up and propagated PrPSc levels. These results indicate that the 3D microenvironment makes oBM-MSCs more permissive to prion infection when cultured in growth conditions, and maintains the ability of neuron-like cells to absorb and propagate scrapie prions.Although primary neuronal cultures show toxicity after prion propagation in vitro that varies in a strain- and neuronal type-specific manner [55,56], murine MSCs are able to propagate prions in growth conditions without signs of toxicity [48,49,50]. In our previous study, oBM-MSCs and their neuron-like cell derivatives, infected with brain homogenates of healthy and scrapie-infected sheep, displayed increasing levels of cell viability throughout the duration of the culture in comparison to non-inoculated cells, suggesting that brain inocula may contain factors that stimulate oBM-MSC proliferation [35]. Similarly, in the present study, scrapie-infected oBM-MSCs showed higher cell viability levels than the non-inoculated controls just after removing the inocula (2 dpi); however, we did not observe this increase in cells inoculated with negative inoculum. It seems that MSC division is more stimulated with scrapie brains. In fact, it has been described that MSCs can migrate to brain lesions caused by prions, and this migration is mediated by different chemoattractive factors [57,58,59]. Similarly, neuron-like cells infected with scrapie displayed an increase in cell viability only after being in contact with the inocula, confirming our previous data. However, in this study, toxicity was observed afterwards. Therefore, the effect of brain inocula in oBM-MSC proliferation seems to vary between cultures. This variation could be explained by the cellular heterogeneity found in MSCs, in which donor, tissue source, culture environment, isolation methods, and passage can affect the phenotype [60,61], making distinct cultures react slightly different to CNS homogenates. However, we have to keep in mind that inocula of both negative and positive brains are also different. The amounts of initial PrPSc could be different, which would affect the reaction of MSC to infection. Nevertheless, the early response to scrapie infection was the same in both studies, an increment of cell viability, which means a higher proliferation potential either in growth or neurogenic conditions when cultured on plates after being in contact with scrapie tissues. The decrease in cell viability observed afterwards could be a consequence of prion toxicity.Ovine MSC spheroids inoculated with brain homogenates of healthy and scrapie-infected sheep displayed different viability patterns than the monolayer-cultured cells. The negative inoculum increased the viability of spheroids in both growth and neurogenic conditions; however, this increase was not observed in spheroids infected with scrapie, which remained at viability levels similar to the controls. Compared to monolayer-cultured MSCs, human MSC spheroids show higher cell survival and enhanced cell yield and stemness [62,63,64]. The enhanced cell viability of spheroids seems to be mediated by the induction of autophagy and the suppression of reactive oxygen species (ROS) [64]. Remarkably, an impairment of the autophagy process has been described in scrapie disease [65,66]. The decreased viability observed in scrapie-infected spheroids in growth conditions, with respect to those inoculated with negative brain material could be due to a dysfunction of the autophagy mechanism caused by prion infection that could counteract the positive effect exerted by neurotrophic factors in the brain. These results suggest that in most cases, 3D conditions improve the viability of inoculated oBM-MSCs, as infected spheroids at least maintain the viability observed in controls, but prions exert their toxicity by limiting the growth potential of spheroids stimulated with neurotrophic factors.5. ConclusionsIn conclusion, oBM-MSC-derived spheroids in growth and neurogenic conditions seem to be able to uptake and propagate scrapie prions and mimic prion toxicity, making this three-dimensional approach a potential in vitro model to study prion disease mechanisms and therapeutics in a more in vivo-like environment. Nevertheless, further studies using a larger number of cultures are still necessary to confirm the reproducibility of this in vitro model, along with the analysis of the PrPSc profile generated in both oBM-MSCs and spheroids in growth and neurogenic conditions.

animals : an open access journal from mdpi

10.3390/ani12010124

PMC8749737

Testicular cryopreservation enables the maintenance of reproductive potential, the creation of germplasm banks and the transport of genetic material between different regions. This biotechnology represents the only possibility of preserving the fertility of prepubertal animals that have already died or that need to undergo gonadotoxic treatments. Despite advances in the use of cryopreserved testicular fragments, protocols that can be used in the clinical routine of dogs and cats have not yet been established. Due to the great importance of the topic, the objective of this review is to provide an overview of the subject, approaching the main works on testicular cryopreservation in dogs and cats.

The increased interest in breeding dogs and cats and their use as models for other canids and felids demand research to improve reproductive techniques. Among them, testicular cryopreservation stands out. Testicular cryopreservation enables the maintenance of reproductive capacity and allows the establishment of germplasm banks for several species of commercial value or at risk of extinction. Furthermore, it enables the transport of genetic material among different regions. It is noteworthy that this biotechnology represents the only possibility of preserving the fertility of prepubertal animals that have died, so it has great importance in the propagation of the genetic material of animals. The spermatogonia present in the testes can be cultivated in vitro and the sperm obtained can be used in artificial reproduction programs. Although advances have been achieved with the use of testicular fragments to obtain viable and functional germ cells, the establishment of protocols that can be used in clinical routine have not been concluded yet. The testicular cryopreservation process can be carried out through techniques such as slow freezing, fast freezing and vitrification. However, the protocols used for the canine and feline species are still in the experimental phase. Given the importance of the topic, the aim of this review is to draw a profile of the subject approaching the main works on testicular cryopreservation in dogs and cats.

1. IntroductionTesticular cryopreservation enables the maintenance of reproductive capacity [1,2,3,4] and allows the implantation of germplasm banks for several species of commercial value or even those at risk of extinction [5]. Moreover, it provides the transport of genetic material among different regions [2]. It is noteworthy that this biotechnology represents the only possibility of preserving the fertility of prepubertal animals that have died, and it thus has great importance in the propagation of their genetic material [6].Spermatogonia present in the testes can be cultivated in vitro, and the sperm obtained in this manner can be used in artificial reproduction programs. Although advances have been achieved with the use of testicular fragments to obtain viable and functional germ cells, the establishment of protocols that can be used in clinical routine has not been concluded yet [7].The testicular cryopreservation process can be carried out using techniques such as slow freezing, rapid freezing and vitrification. However, the protocols used for dogs and cats are still in the experimental phase.2. Canine SpeciesThere are very few published works about testicular cryopreservation in dogs. The pioneer work found on this subject in dogs dates back only to 2016 [8]. The main goal of this study was to evaluate the capacity of two different cryopreservation approaches (slow freezing versus vitrification) for testicular preservation in adult dogs. Table 1 provides a compilation of works about testicular cryopreservation in dogs.2.1. Cryopreservation MethodTesticular cryopreservation represents an important tool to support assisted reproduction techniques [14]; in view of this, the immediate use of this biological material is not always applicable. Testicular cryopreservation can be performed using the conventional freezing method or vitrification [8,15]. The conventional method is characterized by slow freezing, whereas vitrification is characterized by being an ultra-fast freezing process [16].When comparing the protocols of slow freezing and solid-surface vitrification, the latter revealed samples with a larger area composed of tubular compartment, tubular lumen, seminiferous epithelium and tunica propria. Furthermore, the intertubular compartment showed Leydig cells with normal morphology and typical characteristics of steroidogenic cells. These results showed that solid surface vitrification in association with dimethyl sulfoxide and trehalose was effective for histology and sperm ultrastructure preservation in adult dogs; however, viability assessment was not performed [8].Recently, an article on testicular cryopreservation was published comparing the classical slow freezing versus needle-immersed vitrification methods in grey wolf (Canis lupus) testicular fragments [13]. The wolf is an animal of the same family as the dog. In this work, it was observed that the slow freezing method was better than the needle-immersed vitrification method in the morphological preservation of germ cells in this specie of wolf.Testicular cryopreservation can be done from cell suspension or by fragments [17]. Cell suspension is a laborious and harmful method for cell proliferation and differentiation, and thus work with the testicles of dogs and cats has been carried out with fragments.2.2. Fragment SizeAlthough a fragment’s size can influence the action of cryoprotective agents [18], thus far there are no studies with the aim of evaluating whether the size of the canine testicular fragment has an influence on the success of cryopreservation. On the other hand, such studies have already been carried out for cats.In previous works, fragment sizes of 3 to 5 mm3 [8] and 0.4 × 0.4 × 0.4 cm [9,10,11] were used in testicular cryopreservation protocols for dogs.2.3. CryoprotectantsDimethyl sulfoxide (DMSO) and glycerol (GLY) were tested separately using conventional freezing for testicular cryopreservation in adult dogs. The associations between dimethyl sulfoxide/ethylene glycol (EG) and glycerol/ethylene glycol were investigated in the vitrification process as well. A dimethyl sulfoxide/ethylene glycol combination provided better results than a combination of glycerol/ethylene glycol [11].In conventional freezing, cryoprotectants were used separately. On the other hand, vitrification was performed using combinations of different cryoprotectants. This association in vitrification showed better results in preserving the seminiferous tubules compared to conventional freezing with a single cryoprotectant [11].In our laboratory, the combination of cryoprotectants for testicular vitrification in prepubertal dogs is being evaluated to see which association best preserves testicular histological architecture. Thus far, preliminary results indicate that the EG/GLY and DMSO/EG associations are the ones that best preserve testicular integrity following the testicular vitrification process in prepubertal dogs (Figure 1). In all groups, the seminiferous tubules presented poorly developed germinal epithelium, with good distinction between spermatogonia and Sertoli cells, although the EG/GLY association was inferior to the other groups. In all vitrified groups, there was some degree of disorganization of the histological architecture, as demonstrated by the greater separation of the basement membrane compared to the control. All groups were similar to the control regarding basement membrane retraction and nuclear visualization. Regarding nuclear condensation, EG/GLY did not differ from the control, nor from the DMSO/EG, which in turn was inferior to the control and similar to DMSO/GLY [19].2.4. Thawing/WarmingThere was no work found in the literature about testicular cryopreservation in dogs with the purpose of testing different warming or thawing temperatures. In the works in which freezing was performed, three standard thawing protocols were carried out: (1) The cryopreserved testicular cells were thawed at 37 °C for 1 min [12]; (2) Straws were immersed into a water bath (37 °C) until the ice melted; the sealed end of the straws was cut, and the material drained into 2ml of the first thawing solution at 37 °C and incubated for 1 minute before materials were washed in the second solution at 37 °C for 5 min [8]; (3) Cryotubes were placed at room temperature for 30 s, then each fragment was placed in a warming medium (modified TCM 199 plus 20% FCS and 1M sucrose) for 10 min [9,10,11].For post-vitrification warming, different protocols were adopted in the two studies found. The second was the same employed for thawing testicular fragments in the work of Santos (2018): (1) Cryovials containing vitrified testis fragments were kept at room temperature for approximately 30 s, then filled with 10 mL of solution 1 (Dulbecco’s modified Eagle’s medium nutriente mixture F-12—DMEM/F12 + 20% fetal bovine serum—FBS v/v + 100 mM trehalose) and the fragments were transferred into the same solution in a Petri dish at 37 °C with constant swirling for 2 min. Fragments were then transferred to a second dish containing solution 2 (DMEM + 20% FBS v/v + 50 mM trehalose) at room temperature for 1 min. All pieces then were washed twice in 5 mL DMEM + 20% FBS v/v each time with gentle swirling for 10 min at room temperature [8]; (2) Cryotubes were placed at room temperature for 30 s, then each fragment was placed in a warming medium (modified tissue culture medium 199—TCM 199 plus 20% FBS and 1 M sucrose) for 10 min [11].2.5. Post-Xenotransplantation Spermatogenesis or In Vitro CultureProtocols for sperm development through xenotransplantation have already been tested, and it could be an interesting alternative to test these protocols with cryopreserved testicular fragments. It was found that testicular fragments from immature animals xenotransplanted to immunocompromised mice were able to respond to endogenous gonadotropins, thus allowing complete differentiation of sperm capable of fertilization. Other studies have been carried out with this specific approach [20].In this way, xenotransplantation of testicular fragments of two-month-old domestic dogs was carried out in the subcutaneous space of mice. After 13 months of culture, sperm was retrieved from 5 of the 29 fragments that were implanted [14].Leaving the field of testicular fragments, testicular cells from Belgian Shepherd Malinois dogs four or five months in age were frozen, thawed after three months and cultivated in StemPro-34 medium. Then, colonies derived from germ cells and somatic testicular cells conjugated with extracellular matrix were transplanted into immunodeficient mice. The transplanted cells colonized the recipient testes, forming seminiferous tubules mainly composed of Sertoli cells and some germ cells. It was concluded that the StemPro-34 medium with dimethyl sulfoxide was optimal for the cryopreservation of canine testicular cells; the characteristics of the germ cells were maintained in the culture of colonies derived from germ cells cultivated after thawing, and the colonies derived from germ cells from transplanted dogs were able to colonize recipient mouse seminiferous tubules [12].3. Feline SpeciesTable 2 shows a compilation of the works found in the literature regarding testicular cryopreservation in cats.3.1. Cryopreservation MethodTesticular fragments from pubescent cats were used to compare slow freezing and rapid freezing, using dimethyl sulfoxide, ethylene glycol, glycerol and propanediol as cryoprotectants. The group subjected to slow freezing had significantly lower percentages regarding the degree of intact sperm plasma membrane when compared to the control group, which tended to be less than that of the two-step freezing technique (45.9% vs. 60.3% vs. 55.0%). On the other hand, the proportion of testicular spermatozoa with fragmented DNA was not different among the two freezing techniques and fresh testicular fragments. The fertilizing capacity of sperm was also demonstrated; after thawing, in vitro fertilization of mature oocytes was carried out and there was a progression to the blastocyst stage in 14% of cleaved embryos. This rate was similar to the one in control group [22].3.2. Fragment SizeFragment size considerably influences the action of the cryoprotective agent during freezing and/or vitrification [18]. Considering that testicular fragment size can be an important factor in cryopreservation, studies were carried out to test the influence of fragment size on feline testicular quality post-cryopreservation.Fragments with 0.5 cm3 of volume were compared to 0.3 cm3 fragments. Larger fragments (0.5 cm3) had less harmful effects on germ cells than smaller ones (0.3 cm3) [23].DNA damage was evaluated and apoptosis rates were estimated in testicular fragments obtained from cats after thawing. The values of these variables were compared regarding the type of cryoprotectant used (3% propanediol and 3% glycerol) and the size of the testicular fragment (0.3 cm3 and 0.5 cm3). The evaluation with acridine orange indicated that glycerol was more effective than propanediol in preserving DNA in cryopreserved 0.5 cm3 fragments. The results of the histomorphological evaluations indicated greater cellular integrity based on the evaluated criteria among the non-cryopreserved germ cells for both fragment sizes. The values of these variables decreased after cryopreservation, with no differences regarding the size of the stored fragment or cryoprotectants. In this work, immunohistochemical analysis using caspase-3 was conducted to investigate whether cellular apoptosis in the cytoplasm, nuclei or sperm occurred. After cryopreservation, both fragment sizes had similar percentages of samples with caspase-3 staining, with no particularly relevant findings for either cryoprotectant. This analysis showed differential staining between the cytoplasm and nuclei of seminiferous tubular cells along with the staining of spermatozoa heads at different intensities. The immunohistochemical results indicate that there was marked damage to all tissue fragments [26].3.3. CryoprotectantsThe use of different cryoprotectants such as glycerol and propanediol has been proposed in the rapid freezing of cat testis. However, the histomorphological characteristics are better-preserved using glycerol [23].Due to the need for high concentrations of cryoprotectants for vitrification, the association of cryoprotectants can be an alternative to reduce both the amounts of these concentrations as well as their harmful effects. In this way, the two-by-two association of the cryoprotectants dimethyl sulfoxide, ethylene glycol and glycerol were tested in the solid surface vitrification of testes of prepubertal cats. The combination of dimethyl sulfoxide/glycerol was the one that provided better conservation of the morphology of the seminiferous tubules (i.e., the smallest number of tubules seminiferous with separation and retraction of the basement membrane) and higher percentage of cell proliferation potential after warming [24].When using the technique of vitrification in cryotubes with testicular fragments from prepubertal domestic cats, the effect of different combinations of cryoprotectants on cell viability after warming the fragments was evaluated. The dimethyl sulfoxide/glycerol combination was the only one to present a cell survival rate equal to that of the fresh group [25].3.4. Thawing/WarmingComprehending the importance of warming and reanimation conditions is essential to improve the survival of post-vitrification testicular cells. The structural and functional testicular properties of prepubertal domestic cats were studied after vitrification followed by two warming protocols (directly at 37 °C or with a pre-exposure of 5 s at 50 °C) and three reanimation moments (immediately, 24 h and 5 days post-warming). Preservation of the seminiferous tubule structure was better with warming at 50 °C for 5 s, and somatic and germ cell survival was greater compared to direct warming at 37 °C for one minute. Short-term in vitro culture (for resuscitation) proved that cell composition and functionality were better preserved when heated for a short period of time at 50 °C [7].In another work, three different warming temperatures (50, 55 and 60 °C for 5 s) were tested using the fresh testicular fragments as a control (Figure 2A). The aim of this work was to evaluate the possible influence of different warming temperatures on the structure, metabolic activity, composition, and cellular functions of the vitrified testicular fragments of prepubertal cats [27]. The authors concluded that vitrified testicular fragments from prepubertal cats have better preserved morphology, morphometry and viability when warmed at 50 °C (Figure 2B) compared to 60 °C (Figure 2C). This work sought to contribute to the knowledge regarding warming of testicular fragments after vitrification, reinforcing the importance of this step to the germplasm preservation process.3.5. Cryopreserved Testis XenograftXenotransplantation has been used as a form of ex situ cell culture in different species. This technique allows the obtaining of sperm cells from prepubertal animals. However, this method is still restricted to the experimental field.In this way, slow freezing with dimethyl sulfoxide of testicular fragments from prepubertal and pubescent cats was carried out, followed by subsequent xenografts in immunosuppressed mice. After 10 weeks, it was observed that material from prepubertal cats showed an increase in the amount of stem cells and the presence of seminiferous tubules compared to that of pubescent cats, demonstrating that the testis of prepubertal animals had greater spermatogenic potential than that of pubescent animals [21].4. New ApproachesDifferent cryopreservation techniques invariably lead to cellular cryoinjuries. To deal with these questions, researchers have been inspired to use anhydrobiosis to develop room temperature storage methods. Anhydrobiosis is the temporary suspension of vital activities that enable an organism to tolerate long dehydration. This is a natural process used by several small organisms to resist dry conditions. It is based on the properties of trehalose, which reaches a high concentration of solute without molecular mobility during dehydration, thus avoiding intra- and extracellular degradation.Based on the principle of anhydrobiosis, one work has been presented using the cat as a model to develop future preservation at nonfreezing temperatures. The aim of this study was to characterize changes in histology, DNA integrity, and viability of adult testis versus prepubertal individuals during microwave-assisted drying. The results demonstrated for the first time that normal morphology, incidence of degeneration, DNA integrity and viability of testis remained at acceptable levels during microwave-assisted drying for 20 min. Overall, prepubertal testis appeared to be more resilient to microwave-assisted desiccation than adult testis [28].Although the few studies on canine and feline testicular cryopreservation have been carried out using testicular fragments, in 2020, a German team published a paper on cryopreservation of a cat testicular cell suspension. The aim of this work was to establish a cryopreservation protocol for testicular cell suspension of domestic cats to be implanted in endangered feline species, applying two concentrations of dimethyl sulfoxide (7.5 and 15%) and performing a slow and a fast freezing protocol. The best protocol was obtained with slow freezing using 7.5% dimethyl sulfoxide, resulting in a mean cell survival rate of 45.4 ± 9.1% [29].In 2021, a published work suggested epididymal vitrification as an alternative to the conservation of the male gamete as a means of preserving individual reproductive potential. The purpose of the study was to determine the effect of the vitrification of epididymal cauda by comparing the effects of glycerol and ethylene glycol on epididymal sperm quality post-vitrification. The authors observed that epididymal tail vitrification appears to be a suitable method for long-term storage of cat sperm, especially if the procedure is performed with ethylene glycol as the cryoprotectant [30].5. Final ConsiderationsAll cryopreservation techniques have risks of damaging cell structures, resulting in decreased cell viability. Despite the advances obtained, there is still no standardized optimal technique for cryopreservation, which remains an area of open study.After reviewing some existing works on testicular cryopreservation in dogs and cats, it is observed that there is a divergence between the degrees of evolution in these two species. Undoubtedly, testicular cryopreservation in cats is more advanced than in dogs.However, as interest in the area of cryobiology increases, especially testicular cryopreservation, the development and improvement of this area will certainly expand, contributing not only to the increased reproduction of dogs and cats, but to its application in wild canids and felids as well.

animals : an open access journal from mdpi

10.3390/ani11041038

PMC8067757

Supplementation of the poultry diet with plant extracts rich in polyphenolic compounds could improve the performance of animals as well as the oxidative stability of their derived meat. The present study evaluated the efficacy of cornelian cherry extract (CCE) on the expression of genes controlling glucose transporters and different assays regulating the oxidative stability of frozen, stored meat over a long period of time (90 days of storage). The results indicated that the addition of 200 mg/kg of CCE to the diet could improve the growth rate and antioxidant status of broiler chickens and thus increase their productivity. Moreover, polyphenolic compounds rich in CCE can act as antioxidant agents to increase the shelf-life extension of frozen, stored poultry meat. Finally, supplementation with CCE could increase the total concentration of phenolic compounds in poultry meat offered to human consumers.

The use of natural plant extracts in poultry feed could improve their productivity as well as the oxidative stability of stored derived meat. The roles of cornelian cherry extract (CCE) in growth, cecal microbes, and meat antioxidative markers of broiler chickens were evaluated. A total of 500 Ross 308 broiler chicks were fed diets supplemented with CCE (0, 50, 100, 200, 400 mg/kg of diet) for 38 days. The highest levels of weight gain and feed utilization were observed in a group fed 200 mg/kg of CCE. Maximum upregulation of glucose transporters—1 and 2 and sodium-dependent glucose transporter genes—were found in the group fed 200 mg/kg of CCE. Lactobacilli and Bifidobacterium colonization increased as the CCE levels increased. The greatest upregulation of antioxidant genes (glutathione peroxidase, catalase, and superoxide dismutase) in breast meat was observed in groups fed CCE (200 and 400 mg/kg). Dietary CCE significantly delayed the lipid oxidation of breast meat compared with that of the control group. The total phenolic content, 2,2-Diphenyl-1-Picrihydrzyl (DPPH) radical scavenging activity and reducing power in meat improved with higher levels of CCE. Dietary CCE improved the growth, performance of broilers, and meat antioxidant stability after 90 days of storage.

1. IntroductionThe intention for the widespread use of phytonutrients in the poultry industry is ultimately associated with the growing discouragement of the use of antibiotics in feed. Recently, natural, active, plant-derived compounds have been gaining great importance, because of their ability to enhance poultry growth performance by improving nutrient digestibility, increasing the concentration of nutrient transporters, sustaining a healthy gut environment, and improving the quality of their products [1]. A phytogenic diet has also been reported to produce changes in the cell membrane permeability, resulting in a higher absorption rate of micronutrients from the small intestine [2]. Additionally, herbal and medicinal plant additives might have the capacity to control intestinal pathogenic bacteria and improve the beneficial intestinal microbiota [3,4,5] due to their antimicrobial, fungicidal, antiviral, anticoccidial, and antioxidant properties [6,7]. Furthermore, plant extracts that are rich in polyphenols can be effective for preserving meat and their products against oxidative deterioration, pathogen growth, and bacterial spoilage [8].On the other hand, modern large-scale broiler production prompts stressful conditions such as high-stocking density, heat stress, immunological challenges, handling, feed quality, and transportation [9]. These stressors can enhance reactive oxygen species (ROS) production and interrupt the balance between the antioxidant defense systems and oxidation inside a bird’s body, causing oxidative stress [10]. The harmful effect of oxidative stress can be reduced by the dietary inclusion of antioxidants [5]. The use of natural plant-derived compounds rich in polyphenolic compounds can improve the antioxidative status of the living birds and increase the oxidative stability of their derived meat [11].Among these natural compounds is cornelian cherry (Cornus mas) extract (CCE), which is composed of several active compounds, including five anthocyanins: delphinidin 3-galactoside, cyanidin 3-rhamnosylgalactoside, cyanidin 3-galactoside, pelargonidin 3-rhamnosylgalactoside, and pelargonidin 3-galactoside [12]. Ursolic acid is an important constituent of CCE [13,14,15,16] possessing antioxidant and antibacterial properties. Additionally, it contains vitamin C, trace minerals, organic acids, pectintriterpenoids, iridoids, pectins, and tannins that range within the safe standards levels of food [17,18,19]. It is rich in flavonoids such as quercetin 3-O-rhamnoside, quercetin 3-O-rutinoside, and quercetin 3-O-glucuronide [20] and phenolic compounds such as caffeic acid, caffeoylquinic acids, p-coumaric acid, and ellagic acid [21]. Furthermore, CCE has been shown to have antimicrobial [22] anti-inflammatory, and antioxidant activity [23] as well as a hypoglycemic effect [24]. Moreover, the high content of iridoids, such as cyclopentanopyran, found in CCE provides pharmacological properties such as antibiotic and anti-inflammatory effects [25].Poultry meat is highly susceptible to quality deterioration by lipid oxidation during storage, leading to a decrease in nutritive value and the production of a high content of lipid oxidation products [26]. The oxidative stability of poultry meat is influenced by birds’ diets and dietary inclusion of CCE with an abundant amount of polyphenolic bioactive compounds that have been demonstrated to scavenge free radicals and chelate metal ions, helping to increase the oxidation resistance of meat. Moreover, the application of CCE to broiler breast meat was associated with a lower thiobarbituric acid reactive substances (TBARS) value and an increased meat shelf life [15].Digested carbohydrates, protein, and lipids are transferred into the body by certain transporter proteins located in the brush border enterocytes of the small intestine [27,28]. These include GLUT1 and GLUT2, which are responsible for monosaccharide transportation (glucose, galactose, fructose, and mannose) across the intestinal membrane [29]. The greater expression of transporter-encoding genes leads to a higher flood of nutrients into the intestinal cells and, subsequently, into the blood [30].Cornelian cherry extract can play an important role in chickens’ intestinal health and meat quality owing to its active principle content. Thus, this work investigated different mechanisms and provides new data about the effects of cornelian cherry extract on growth performance, glucose transporters, gut microbes, and meat oxidative stability in broiler chickens.2. Materials and MethodsThe management practices and procedures followed animal welfare, ethical norms, and guidelines of the Institutional Animal Care and Use Committee of the Faculty of Veterinary Medicine at Zagazig University.2.1. Birds, Diets, and Experimental DesignA total of 500 male Ross broiler chicks (ROSS 308), on the day of hatching (initial body weight 45.8 ± 1), were purchased from a commercial hatchery. Chicks were weighed and randomly divided into five treatment groups with 10 replicate pens containing 10 birds each. The study was organized at the Faculty of Veterinary Medicine, Zagazig University, Egypt. The experimental protocol was accepted by the ethics committee of the Institutional Animal Care and Use Committee of Zagazig University, Egypt. All animal experiments were done according to the recommendations described in “The Guide for the Care and Use of Laboratory Animals in scientific investigations” to ensure their welfare, maintain their rights, and cause minimal stress. All chicks were housed in the same environmental and sanitary conditions all over the experimental period. Birds were raised in floor pens with wood shavings (bird density: 10 broilers/m2) in an environmentally controlled room. The photoperiod in all experimental pens was maintained at 23 L:1 D h for the first 3 days, followed by 20 L:4 D h until the end of the experiment. The relative humidity ranged from 65 to 75% throughout the trial. During the 1st week, the room temperature was initially set at 33 °C and then gradually decreased until the final temperature of 23 °C was reached. The control starter (d 1–10 d) and grower-finisher (d 11–38) diets were formulated to cover the nutrient requirements of Ross broilers according to the nutritional specifications of ROSS [31]. All birds were allowed access to water and feed ad libitum. The birds were offered a basal diet supplemented with (0 (control), 50, 100, 200, and 400 mg/kg diet of cornelian cherry extract (CCE). The quantities of feed ingredients and the chemical composition of the control diet are listed in Table 1. The proximate analysis of the feed ingredients was done according to the standard procedures of the Association of Official Agricultural Chemists [32]. Asiatic cornelian cherry extract was obtained from Shaanxi Sinuote Biotech Co. Ltd. China. The extract was collected by water alcohol extraction with an extraction ratio of 10:1. The HPLC analysis of the extract based on the manufacturing company was 100 g of cornelian cherry extract containing 203 parts per thousand (PPT) iridoids (consisting of 85% loganic acid), 2.8 PPT ellagic acid, 8.9 PPT anthocyanins, and 4.1 PPT flavonols such as quercetin 3-glucuronide, kaempferol galactoside, and kaempferol glucoside.2.2. Growth Parameters and Digestibility TrialThe body weight and feed intake (FI) were estimated during the starter period (d 1–10) and grower–finisher period (d 11–38) to calculate the body weight gain, FI, and feed conversion ratio for the whole experimental period (d 1–38). The apparent nutrient digestibility was determined with titanium oxide. At 38 days of age, titanium oxide was added to experimental diets at a rate of 5 g/kg diet. The excreta from each replicate pen was collected every 8 h for five days and analyzed for dry matter, crude protein, ether extract, and crude fiber according to the Association of Official Agricultural Chemists [32]. The titanium oxide content in the diets and excreta was analyzed spectrophotometrically after acid digestion in accordance with the method presented by Short et al. [33] The apparent digestibility coefficient of nutrients was calculated in accordance with the equation presented by McDonald [34]. Apparent nutrient digestibility = 100 − [100 × (Indicator content (diet)/Indicator content (feces) × Nutrient content (feces)/Nutrient content (diet)](1)2.3. Sample Collection and Analytical ProceduresAt the end of the experimental period, randomly selected birds were weighed and slaughtered.For serum biochemical measurements, 3 mL blood samples were collected from each bird and then centrifuged for 15 min at 2000 rpm. Clear serum samples were kept at −20 °C until further biochemical analysis.For meat chemical composition analysis, meat samples were collected from the breast and thigh and then stored at −20 °C until chemical analysis.For the meat antioxidant analysis, breast meat samples were frozen immediately until the total phenolic content (TPC) and thiobarbituric acid reactive substance (TBARS) content were analyzed and the 2,2-Diphenyl-1-picrihydrzyl (DPPH) assay and Ferric reducing antioxidant power (FRAP) were conducted at 7 and 90 days of storage at −20 °C.For the molecular analysis, breast meat samples were collected and stored at −80 °C until the analysis of antioxidant-related genes. The small intestine (jejunal part) was separated and the digesta was squeezed out from it and rinsed 3 times in PBS (NaH2PO4, 1.47 mmol/L; Na2HPO4, 8.09 mmol/L; and NaCl, 145 mmol/L). One square centimeter of the distal jejunum immediately before the Meckel’s diverticulum was dissected and kept in Trizol reagent at −80 °C until the analysis of nutrient transporter encoding genes.2.4. Serum Biochemical AnalysisSerum aspartate aminotransferase (AST), and alanine aminotransferase (ALT), total cholesterol (TC), triglycerides (TGs), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), very-low-density lipoprotein cholesterol (VLDL-C), creatinine, uric acid, total protein, globulin, and albumin concentrations were determined using commercial diagnostic kits (Spinreact Co., Santa Coloma, Spain).2.5. Chemical Copmostion of MeatThe dry matter, crude protein, fat and ash contents of breast and thigh meat were analyzed according to the Association of Official Analytical Chemists (AOAC) [32].2.6. Antioxidant Potential of Broiler MeatBreast meat samples (5 g) were mixed with phosphate buffer (20 mL; pH 7.4) and glycerol (20 mL; 20%) and homogenized and filtered to ensure they were free from connective tissues.2.6.1. Total Phenolic Content (TPC)The total phenolic content of breast meat was measured in accordance with [35]. Briefly, a previously prepared meat sample (100 μL), distilled water, 2.5 mL of 95% ethanol (500 μL), and 50% Folin–Ciocalteu reagent (250 μL) were mixed. This mixture settled for 5 min, and then 5% Na2CO3 (500 μL) was added. The mixture was rotated in a vortex meter and left in a dark place for 1 h. The absorbance of samples was measured at 725 nm via a spectrophotometer. The quantity of TPC in meat was measured as Gallic acid equivalents (milligrams of gallic acid per 100 g of meat sample).2.6.2. 2,2-Diphenyl-1-picrihydrzyl (DPPH) AssayThe DPPH scavenging capacity of the meat sample was calculated as described by [36]. The DPPH A solution (0.25 mM) was formed in methanol. Each sample (100 mL) was mixed with 100 μL of DPPH solution and maintained for 30 min at 25 °C in a dark place, and the sample absorbance was read at 517 nm by a spectrophotometer. The scavenging activity percentage of DPPH in the meat was determined with the following equation: Scavenging activity of DPPH (%) = 1 − [A1 − A2] × 100(2)Blank absorbanceA1 = Sample absorbance; A2 = Blank absorbance2.6.3. Ferric Reducing Antioxidant Power (FRAP)FRAP in meat was detected according to Oyaizu [37]. A 200-μL homogenized meat sample was mixed with sodium phosphate buffer (500 μL). After that, the prepared solution was maintained in a water bath for 20 min at 50 °C and then centrifuged for 10 min with trichloroacetic acid (2.5 mL) and ferric chloride solution (100 μL). The spectrophotometric absorbance of the samples was measured at 700 nm. The FRAP was estimated as μmol/Fe2+/g meat.2.7. TBARS AssayLipid oxidation in breast meat was evaluated based on the malondialdehyde (MDA) content, as the MDA concentration in breast meat was determined as previously described by [38]. Briefly, perchloric acid (27 mL) was added to 5 g breast meat samples and then homogenized and filtered. Supernatant samples were mixed with thiobarbituric acid (2 mL) and incubated for 20 min in a water bath (100 °C). Consequently, direct cooling and centrifugation were done for 15 min, and the absorbance was measured by a spectrophotometer at 532 nm. The values are expressed as milligrams of malondialdehyde per kilogram of meat.2.8. Real-Time PCR to Assess Nutrient Transporter Encoding GenesTotal mRNA was extracted from jejunum and breast meat samples (n = 10 per treatment) using Trizol reagent (TaKaRa Biotechnology Co. Ltd., Dalian, Liaoning, China). The isolated RNA was treated with the RNeasy Mini Kit; Qiagen, Cat. No. 74104 according to the manufacturer’s guidelines. The quantity and purity of the total RNA were determined by a NanoDrop ND-8000 spectrophotometer (Thermo Fisher Scientific, Waltham, USA). Complementary DNA (cDNA) was obtained by reverse-transcription of isolated RNA samples using RevertAidTM H Minus kits (Fermentas Life Science, Pittsburgh, PA, USA). One microliter of this cDNA was mixed with 2× maxima SYBR Green PCR mix (12.5 μL) and RNase free water (10.5 μL), and then, 0.5 μL of each forward and reverse primer for the selected genes was added. The primers’ sequences of genes encoding the glucose transporter and antioxidant enzymes are described in Table 2. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was selected as a reference gene.2.9. Microbiological AssaySpread plate counting method for different microbes: One gram of cecal content was mixed with 9 mL phosphate-buffered saline and vortexed for 1 min. Samples were serially diluted in sterile diluents (0.5 g/kg peptone water in distilled water). De Man, Rogosa, and Sharpe (MRS, CM1153, Oxoid, Basingstoke Hampshire, UK) agar medium was used for the enumeration of Lactobacilli (MRS, CM1153, Oxoid), Bifidus selective agar was utilized to determine the Bifidobacteria content (BSM-Agar, 88517, Sigma, St. Louis, MO, USA), and violet-red bile glucose agar (VRBG, CM485, Oxoid) was used to determine the Escherichia coli content. After incubation under appropriate conditions for each group of bacteria (72 h at 37 °C under anaerobic conditions for lactobacilli and Bifidobacteria and 48 h at 39 °C under aerobic conditions for E. coli), colonies were counted on the plates, and the microbial population was expressed as log10 CFU/g of cecal content.2.10. Statistical AnalysisThe data analysis was conducted using the general linear model (GLM) procedure of Statistical Package for the Social Sciences, software (SPSS), after confirming the homogeneity among experimental groups using the Levene test and the normality using the Shapiro–Wilk test. Tukey’s test was used to test for significant differences between the mean values. All results are expressed as the standard error of the mean (SEM), and the statistical significance was set at p < 0.05. Cecal colony-forming unit (CFU) data were converted to log10 CFU numbers before analysis. The fold change was measured by the following equation: (B–A)/A where the lowest value is A and the highest value is B. Relative fold changes in the expression of target genes were calculated by the 2−ΔΔCt method using the GAPDH gene as an internal control gene to normalize the target gene expression levels [39].3. Results3.1. Growth Performance and Nutrient DigestibilityThe growth performance parameters of the total growing period and nutrient digestibility are shown in Table 3. The body weight gain (BWG) was significantly greater (p < 0.05) in groups fed 100, 200, and 400 mg/kg of CCE when compared with the control group. Moreover, the highest BWG was observed in the group fed 200 mg/kg of CCE (increased by 9% in comparison with the control group). The feed conversion ratio (FCR) was improved in all groups fed CCE supplemented diets at different levels, and the biggest improvement in FCR was detected in the group fed diets supplemented with 100 or 200 mg/kg CCE. Concerning the nutrient digestibility, the groups fed 100 or 200 mg/kg CCE showed higher dry matter (DM) contents and CP digestibility levels in comparison with diets containing other levels of CCE, while the control group showed the lowest level of DM digestibility. No significant difference in the digestibility of CF was observed among the different groups.3.2. Serum Biochemical ParametersData regarding the impact of CCE on serum biochemical parameters after 38 days are shown in Table 4. The levels of serum AST, ALT, creatinine, uric acid, total protein, albumin, globulin, TGs, HDL-C, and VLDL-C were not affected by dietary CCE (p > 0.05). However, total cholesterol and LDL-C concentrations were significantly reduced (p < 0.05) in the group supplemented with 400 mg/kg of CCE in comparison with the other groups.3.3. Chemical Composition of Meat:The dry matter, crude protein, fat, and ash contents both breast and thigh meat were not affected (p > 0.05) by dietary CCE, as shown in Table 5.3.4. Gene Expression of Glucose TransporterThe mRNA expression levels of the jejunal nutrient transport genes (GLUT1, GLUT2, and SGLT-1) are presented in Figure 1. The expression levels of glucose transporter genes (GLUT-1, GLUT-2, and SGLT-1) were significantly upregulated in groups fed CCE, and the highest level was found in the group supplemented with 200 mg/kg of CCE (p < 0.05) with fold changes of 2.55, 2.32, and 2.36, respectively. The aforementioned upregulation was declined in the group supplemented with 400 mg/kg of CCE. Of note, GLUT1 expression in the group supplemented with 50 mg/kg of CCE showed no significant difference as compared with the control group. Moreover, SGLT-1 expression was significantly upregulated in all groups supplemented with CCE, and the maximum level of upregulation was found in the group supplemented with 200 mg/kg of CCE.3.5. Gut MicrobiotaThe gut microbiota data are presented in Table 6. The mean cecal populations of beneficial lactobacilli, bifidobacteria significantly increased after dietary supplementation with 100, 200, or 400 mg/kg of CCE when compared with the control group (p < 0.05). The population of E. coli in cecal samples significantly decreased (p < 0.05) as the CCE level increased, and the lowest reduction of E. coli was observed with diets supplemented with 200 or 400 mg/kg of CCE.3.6. Antioxidant-Related GenesThe expression patterns of selected antioxidant-related genes (glutathione peroxidase, GPx, superoxide dismutase, SOD and catalase, CAT) are presented in Figure 2. The highest expression levels of GPX were observed in groups supplemented with 400 mg/kg of CCE followed by the groups fed 200 mg/kg of CCE then 50 and 100 mg/kg of CCE. The mRNA expression of GPx was significantly upregulated (p < 0.05) in the groups fed 200 or 400 mg/kg of CCE compared with groups fed 50 or 100 mg/kg of CCE when compared with the control group. The mRNA expression level of SOD was significantly increased as the CCE level increased when compared with the control. The highest expression levels of catalase were observed in groups supplemented with 200 or 400 mg/kg of CCE followed by the groups fed 50 or 100 mg/kg of CCE when compared with the control group.3.7. Antioxidant Potential of Breast MeatData from the analysis of the total phenolic content (TPC), 2,2-diphenyl-1-picrihydrzyl (DPPH) assay, and ferric reducing antioxidant power (FRAP) are presented in Table 5.3.7.1. Total Phenolic Content (TPC) in Breast MeatAfter a short storage period (7 days), the TPC content of breast meat was significantly increased with an increasing level of dietary CCE. After a long storage period (at day 90), the highest values of TPC were observed in the meat of bird groups fed 200 or 400 mg/kg of CCE.3.7.2. The Free (DPPH) Radical Scavenging ActivityThe DPPH activity significantly increased (p < 0.05) in breast meat from groups fed dietary CCE in a dose-dependent manner, even after 90 days of storage.3.7.3. FRAP Reducing ActivityThe capacity of the breast myofibrillar protein to reduce Fe3+ to Fe2+ was greater in the meat of groups fed an increased level of dietary CCE. This capacity was increased after 90 days of storage by 1.1 and 1.7 times, respectively, in the meat from groups fed 200 and 400 mg/kg of CCE when compared to the control group.3.8. Lipid PeroxidationLipid peroxidation, determined as the concentration of MDA in breast meat after 7 and 90 days of storage, is presented in Table 7. At day 7 (short term storage), the MDA level was significantly reduced (p < 0.05) in all groups supplemented with CCE, and the lowest levels were observed in the groups fed 200 and 400 mg/kg of CCE when compared with the control group. At day 90 (long-term storage), all groups supplemented with CCE, except the group fed 50 mg/kg, showed reduced levels of MDA when compared with the control group. The greatest reduction in MDA was observed in the group fed 400 mg/kg of CCE. Moreover, the meat MDA content in the group supplemented with 400 mg/kg of CCE was decreased by up to 59.5% and 68.6% after 7 and 90 days of storage, respectively, when compared with the control (p < 0.05).4. DiscussionNatural plant extracts have been shown to improve the performance of birds by augmenting nutrient utilization and bacterial modulation in the gastrointestinal tract as well as improving the meat quality and oxidative stability. Among these natural products, polyphenols gained growing interest due to their numerous functional properties. Cornelian cherry extract (CCE) has been reported to have an abundance of polyphenolic compounds. The current study demonstrated that CCE could be used to improve the growth rate of broilers by controlling nutrient utilization and absorption and the gut microbiota, and increasing the shelf-life storage of poultry meat by boosting the oxidative stability of meat. In the present study, supplementation of CCE has been shown to improve the feed efficiency of birds by lowering the overall FCR by nearly 5% in the group supplemented with 100 and 200 g/kg CCE. Herein, CCE supplementation has been shown to improve the feed efficiency of birds by lowering the overall FCR by nearly 5% in groups supplemented with 100 or 200 g/kg CCE. The present results are consistent with the results of other researchers [40], who reported that the dietary inclusion of different phenolic compounds for broilers had a positive impact on growth performance parameters. Plant-rich phenolics could promote the growth performance of broilers [40,41] through their potential to improve the antioxidant status of the gut [42]. Furthermore, the application of green tea extract, which is rich in catechins, at concentrations of 100 or 200 mg/kg, in feed has been found to boost the growth performance of broiler chickens [43,44]. Additionally, Herrero-Encinas et al. [45] showed that dietary supplementation with olive extract rich in polyphenolic compounds significantly improved the weight gain and feed conversion ratio of broilers. Similarly, the body weight gain of broiler chickens improved after feeding with grape seed proanthocyanidin extract [46]. Additionally, a sugarcane-served polyphenol mix had a positive effect on the growth performance of broilers and modulated the negative effect of heat exposure [47]. Moreover, the consumption of berries, which have numerous bioactive phenolic compounds, was shown to upregulate the expression of growth-related genes such as insulin-like growth factor binding proteins [48]. Furthermore, certain phytogenic derived agents could improve gastrointestinal barrier function and nutrient absorption [49,50]. Sarker et al. [51] showed that feeding with Cornus officinalis had no adverse effect on the growth rate in broilers.On the other hand, the concentrations of serum AST and ALT indirectly reveal the liver health status and increases in their levels are considered markers of liver damage [52]. In addition, the function of kidney can be estimated via the decrease or increase in of urea and creatinine serum levels. Herein, serum concentration of AST, ALT, uric acid, and creatinine were not affected by dietary CCE and were within normal range, which indicates healthy liver and kidney functions in both control and CCE supplemented groups. Moreover, the current study revealed that higher levels of dietary CCE (400 mg/kg) significantly reduced total cholesterol and LDL-C levels in serum. Similarly, Zhang et al. [53] specified that the feeding of broiler chickens with Chinese bayberry leaves, which are rich in phenolic compounds, significantly reduced the serum cholesterol concentration. The presence of a higher concentration of proanthocyanidins in sorghum, which has antioxidative properties, was also reported to be associated with cholesterol-lowering [54]. Furthermore, the higher concentration of anthocyanins in cornelian cherry powder had a hypercholesterolemic effect in rats via augmenting peroxisome proliferator-activated receptor (PPARα) protein expression and controlling reactive oxygen species (ROS) production and, subsequently, the inflammatory process [55]. Additionally, serum protein and globulin concentrations were not significantly different between treatments and were within the normal range [56]. Similarly, supplementation of broiler chickens with polyphenol extract did not affect their serum protein and globulin concentrations [57]. In addition, supplementation with CCE had no significant effects on the chemical composition (DM, CP, EE and Ash) of meat; these results are in accordance with Gopi et al. [40].Additionally, the enhanced growth performance of the broilers could have resulted from increasing levels of probiotic bacteria such as lactobacilli and Bifidobacteria. Moreover, the population of these beneficial bacteria increased in groups supplemented with higher levels of CCE. These positive effects could be related to the role of phenolic-rich cornelian cherry extract on the intestinal microflora, leading to an increase in the concentration of beneficial bacteria (probiotic effect) or inhibiting the growth of pathogenic species (antimicrobial effect) [58]. In agreement with our results, Bifidobacterium and Lactobacillus have been shown to be the most widely used probiotic bacteria, exerting health-promoting properties, such as the maintenance of gut barrier function [59]. Additionally, an increase in probiotic bacteria, such as Lactobacillus, is accompanied by a decrease in the concentration of pathogenic E. coli, in accordance with authors [60,61] who stated that Lactobacillus can quantitatively inhibit the adherence of pathogenic E. coli. The potential effect of phenolic compounds in CCE on the gut microbiota may result from modulation of the bacterial population by acting as prebiotics and enriching the beneficial bacteria [62]. The antimicrobial properties of polyphenols are of primary significance and inhibit biofilm formation in the gut by suppressing harmful bacteria [63]. In vitro testing of the antimicrobial activity of cornelian cherry extract demonstrated its inhibitory effects against Staphylococcus aureus and Escherichia coli [64]. Caffeic acid, present in CCE, has been described as a potential inhibitor of the growth of E. coli and Clostridium [65].Furthermore, flavonoids can change the microbiota ecosystem through their bacteriostatic or bactericidal properties [66]. In addition, blueberry flavonoids can inhibit the activity of E. coli, which reduces the integrity of the intestinal barrier as a key mechanism of its pathogenesis [67]. Moreover, feeding with dietary polyphenol-rich grapes was shown to significantly increase the Lactobacillus population in the ceca of broiler chickens [4].In the small intestine, glucose transporter 1 (GLUT1), recognized as part of solute carrier family 2 (SLC2A1), facilitates glucose transport across the apical surface of the enterocytes, whereas glucose and fructose transport across the basal side of the enterocytes and into the blood circulation is facilitated by GLUT2 [68]. In the current study, the inclusion of CCE in the broiler diet upregulated the expression of glucose nutrient transporters such as GLUT1, GLUT2, and SGLT-1, which are responsible for the transportation of fructose, galactose, mannose, and glucosamine. This can be attributed to the presence of anthocyanins in CCE that are characterized by α-glucosidase inhibitor activity and the capacity to combine with and activate peroxisome proliferator-activated receptor gamma (PPARγ) [69]. The activation of PPARγ accelerates lipid metabolism and glucose uptake by increasing the actions of insulin in glucose utilization in animals [70]. The enhanced expression of GLUT2 can regulate the digestion and absorption of poultry by modulating food consumption by controlling the feedback signal to the brain [71]. Additionally, cornelian cherry extract administration can initiate the uptake of glucose by tissues such as muscle and participate in increasing muscle mass in poultry [72,73]. Moreover, the upregulation of GLUT1 and GLUT2 can enhance their absorptive functions and increase the final body weight of broiler chickens via increasing nutrient transporter expression in the small intestine [74,75]. Additionally, plant-rich phenolics could promote the growth performance of broilers [41] through their potential to improve the antioxidant status of the gut [42]. Antioxidant capacity is a critical factor in poultry health that affects meat quality after slaughter. The radical scavenging ability is linked to the rich polyphenolic compound composition [76]. Animals have developed effective methods of protection against oxidative stress. SOD can eliminate superoxide anion free radicals, and GSH-PX and catalase can catalyze hydrogen peroxide decomposition [77]. Lipid peroxidation is produced by high levels of free radicals, and it leads to an increase in the content of MDA, the end product of lipid oxidation [78]. Levels of thiobarbituric acid reactive substance (MDA) are biomarkers for assessing the lipid peroxidation degree [79]. To the best of our knowledge, there are no data available on the impact of CCE supplementation on the expression of genes encoding the antioxidant enzymes in broilers. In the current study, the expression of antioxidant-related genes (SOD, GPX, and catalase) was upregulated in breast meat by increasing the CCE level. Chickens’ breast meat that was not supplemented with CCE had elevated MDA levels and, as a result, lowered oxidative stability, compared with groups supplemented with CCE. Thus, dietary supplementation with CCE had a postmortem effect of decreasing the rate of lipid oxidation by decreasing the MDA content in breast meat, even after 90 days of storage. Similarly, the total antioxidant capacity and phenolic content in the broilers’ breast meat improved following dietary supplementation with pomegranate peel extract [80]. Additionally, previous studies showed that lipid oxidation in chicken meat was reduced by feeding with dietary antioxidants, such as plant extracts rich in phenolic compounds [81,82]. A similar positive effect of dietary phenolic compounds on TBARS in breast meat was detected [83]. The higher antioxidant capacity of CCE may be related to the higher polyphenol and flavonoid contents [84], which augment the ability to scavenge radicals. Parallel results support the idea that lipid oxidation could be prevented by fortifying flavonoid antioxidants in animal feed [85]. Furthermore, supplementation of the broiler diet with blackcurrant-extract-rich-polyphenolic compounds enhanced the oxidative stability of their meat after 90 days of frozen storage [86].Additionally, the reduced TBARS value was useful for lengthening the shelf time of meat products and improving the meat quality [11]. Additionally, the increased total phenolic content in the breast meat of broilers fed increased levels of CCE compared with broilers fed the control diet indicated a higher total antioxidant capacity [87], as these phenolic compounds are able to scavenger free radicals [88]. Besides this, supplementation with dietary CCE improved the DPPH scavenging activity of breast meat, especially at higher levels, even after 90 days of storage. The higher scavenging activity of DPPH indicated an increased antioxidant content in broiler meat, which has the potential to provide one proton to produce a stable DPPH2 compound, thus scavenging the free radicals [89]. Additionally, Cerit et al. [90] reported that cornelian cherry fruits have greater DPPH radical scavenging and ferric-reducing activity.Moreover, the highest FRAP values were measured in meat enriched with higher levels of supplemental CCE. Faiz et al. [91] showed decreased TBARS values and better activity against 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radicals, associated with a dose-dependent increase in the total phenolic compound concentration detected in chickens’ meat after they received different levels of citrus waste. Furthermore, Jang et al. [92] stated that the oxidative stability of the breast meat of chickens fed a diet enriched with Coptis chinensis extract was mainly attributed to the higher concentration of polyphenolic compounds. The antioxidant content present in broiler meat tended to convert ferric ion (Fe3+) to ferrous ion (Fe2+) by providing one electron. Higher metal-chelating potential after supplementation with CCE protected tissues from damage resulting from oxidation. Similarly, a higher metal ion reducing capacity was observed in meat enriched with natural antioxidants in contrast with a control treatment group [93]. Similarly, cornelian cherry extract can reduce 20.41 μmol of Fe2+/g of solution [13].5. ConclusionsSupplementation with cornelian cherry extract, which is rich in antioxidants, improved the growth performance of broiler chickens via several mechanisms, such as increasing the favorable probiotic populations and lowering the concentration of harmful bacteria, such as E. coli spp. The modulation of genes expression responsible for glucose absorption and antioxidant enzymes indicates that CCE can play an effective role in the previously mentioned molecular mechanisms. CCE can scavenge free radicals, thereby improving the antioxidant capacity and lipid peroxidation of poultry meat without affecting its chemical composition. The results demonstrate that the application of dietary CCE (200 mg/kg) is recommended in chickens’ diets to boost their growth performance, health, and meat shelf stability during long periods of frozen storage.

animals : an open access journal from mdpi

10.3390/ani13050790

PMC10000167

Incorporating warm-season grasses into traditional cool-season grass equine rotational grazing systems can increase pasture availability during hot, dry months and bridge the “summer slump” forage gap. The objective of this study was to evaluate the impacts of this pasture management practice on the equine microbiome and to explore relationships between the fecal microbiota, forage nutrients, and metabolic responses of grazing horses. Results of this study indicate that distinct changes in microbial community structure and composition occur as horses adapt to different forages and that shifts in the microbial community were most influenced by forage non-structural carbohydrates and crude protein, rather than fiber. Interrelationships were found between these nutrients, glycemic responses, and Akkermansia and Clostridium butyricum. These bacteria were also found to be enriched in horses adapted to warm-season grasses. While the results of this study suggest that integrating warm-season grasses may not offer substantial metabolic benefits in healthy adult horses, this study did reveal new insights and targets for future research necessary to better understand the function of Akkermansia and Clostridium butyricum in the hindgut microbiome of grazing horses and possible roles in modulation of equine metabolic health.

Integrating warm-season grasses into cool-season equine grazing systems can increase pasture availability during summer months. The objective of this study was to evaluate effects of this management strategy on the fecal microbiome and relationships between fecal microbiota, forage nutrients, and metabolic responses of grazing horses. Fecal samples were collected from 8 mares after grazing cool-season pasture in spring, warm-season pasture in summer, and cool-season pasture in fall as well as after adaptation to standardized hay diets prior to spring grazing and at the end of the grazing season. Random forest classification was able to predict forage type based on microbial composition (accuracy: 0.90 ± 0.09); regression predicted forage crude protein (CP) and non-structural carbohydrate (NSC) concentrations (p < 0.0001). Akkermansia and Clostridium butyricum were enriched in horses grazing warm-season pasture and were positively correlated with CP and negatively with NSC; Clostridum butyricum was negatively correlated with peak plasma glucose concentrations following oral sugar tests (p ≤ 0.05). These results indicate that distinct shifts in the equine fecal microbiota occur in response different forages. Based on relationships identified between the microbiota, forage nutrients, and metabolic responses, further research should focus on the roles of Akkermansia spp. and Clostridium butyricum within the equine hindgut.

1. IntroductionIntegrating warm-season grasses into traditional cool-season grass rotational grazing systems can provide productive pasture for grazing horses during the hot, dry months of the “summer slump” period [1,2]. Despite reported benefits for pasture yield [2,3] the impacts of this practice on equine metabolic health and the hindgut microbiome have not been previously investigated.Warm- and cool-season grasses have different mechanisms for storage of soluble carbohydrates [4], with non-structural carbohydrate (NSC = sugars + starch + fructans) concentrations in cool-season grasses typically greater than that of warm-season grasses [1,5]. These differences in NSC content are of interest in equine management as current feeding recommendations for horses with existing metabolic dysfunction include limiting dietary NSC concentrations [6,7,8]. Thus, lower-NSC warm-season grasses have been suggested as an alternative forage source [1,9]. Feeding supplemental concentrate higher in NSC has been shown to lower insulin sensitivity in horses [10,11,12], but over the relatively smaller range of NSC concentrations observed in forages, potential benefits of limiting NSC intake are less clear [13,14,15]. However, glycemic and insulinemic responses of horses grazing low-NSC warm-season grasses have not been extensively evaluated.In addition to the potential metabolic effects, transitioning horses between forage types may also have implications for equine gastrointestinal health. Diet has been identified as a dominant factor shaping the community structure of the gut microbiota both in humans and across animal species including horses [16,17]. However, prior studies on the influence of diet on the equine hindgut microbiome have focused primarily on concentrate vs. concentrate or concentrate vs. forage diets [18,19,20,21]. Recent studies have demonstrated that type of hay (alfalfa vs. grass) and transitions between hay and pasture grass can impact equine cecal or fecal microbial community composition [22,23]. Overall, few studies have evaluated the hindgut microbiome of grazing horses and only one previous study has been conducted in horses grazing cool- vs. warm-season pasture grasses [24].A number of studies have begun to explore associations between the equine hindgut microbiota and metabolic health [25,26,27,28], but there is a lack of information on the interplay between forage nutrients, the hindgut microbiome, and metabolism of grazing horses. Biddle et al. [28] explored relationships between abundance profiles of fecal microbial taxa, feed types (pasture, hay, or hay supplemented with concentrate), and blood analytes including circulating glucose and insulin, finding negative correlations between insulin and over 50 taxa. Studies conducted in mouse models have conclusively demonstrated that changes in diet influence host metabolism in a microbiome-dependent manner [29,30]. Thus, there is a need for future research to better understand potential roles of the gut microbiota in modulating metabolism of grazing horses and the influence of specific forage nutrients on these interactions. While shifts in hindgut microbial communities have been documented in pastured horses over time, little is known regarding relationships between these changes in bacterial composition and forage nutrient profiles or metabolic responses of grazing horses. Therefore, the aims of this study were to characterize shifts in glucose metabolism and the fecal microbiota of horses adapted to different forage types and to explore relationships between forage nutrients, microbial composition, fecal metabolites, and metabolic responses of grazing horses.2. Materials and MethodsResearch was conducted in 2018 at the Ryders Lane Environmental Best Management Practices Demonstration Horse Farm (Rutgers, The State University of New Jersey; New Brunswick, New Jersey, NJ, USA). Weather data for the study period and historical averages are presented in Table S1 [31]. 2.1. Grazing SystemsTwo separate 1.5 ha integrated warm- and cool-season rotational grazing systems were utilized in this study. Grazing system design and management were as published in previous companion studies [2,24]. In brief, warm-season grass pasture sections contained either Wrangler bermudagrass [BER; Cynodon dactylon (L.) Pers.; Johnston Seed Company, Enid, OK, USA] or Quick-N-Big crabgrass [CRB; Digitaria sanguinalis (L.) Scop.; Dalrymple Farms, Thomas, OK, USA]. Mixed cool-season grass sections contained Inavale orchardgrass [Dactylis glomerata (L.)], Tower tall fescue (endophyte-free) [Lolium arundinaceum (Schreb.) Darbysh.], and Argyle Kentucky bluegrass [Poa pratensis (L.)] (DLF Pickseed, Halsey, OR, USA). Grazing management was guided by established best management practices [32].2.2. Animal and Grazing ManagementUse of animals in this study was approved by the Rutgers University Institutional Animal Care and Use Committee protocol #PROTO201800013. Eight adult Standardbred mares (age: 18 ± 0.71 yr; body weight (BW): 537 ± 17 kg; body condition score (BCS): 5–7)) were used in this study, with horses grouped by BW and BCS (Henneke Body Condition Score scale [33]. Horses were then randomly assigned to each system (n = 4 horses system−1). Prior to the study, oral sugar tests (OST) were administered to screen horses for impaired insulin sensitivity [34]. Insulinemic responses of all horses were found to be normal (peak insulin ≤45 mIU/L) [7]. Animal care and management have been previously detailed in companion studies [1,24]. Horse condition measures over the course of the study can be found in Table S2.2.3. Fecal and Blood Sample CollectionManual grab fecal samples were collected (0800 h) rectally from horses after 21 d adaptation to the initial hay diet in the spring (HAY-SP), cool-season grass pasture in the spring (CSG-SP), warm-season grass (WSG)—either BER or CRB during the “summer slump” period, and again following a return to cool-season grass in the fall (CSG-FA) and a final cool-season grass hay at the end of the grazing season (HAY-FA). See Figure S1 for a diagram of experimental design and sampling protocol. Due to delayed establishment and grazing of BER, only 17 days of grazing were possible prior to sample collection. Upon collection, samples were immediately placed on ice, transported to the laboratory, and then stored at −80 °C.To determine impact of forage type within integrated systems on glycemic and insulinemic responses of grazing horses, OST [34] were conducted following adaptation to CSG-SP, WSG, and HAY-FA. On the evening prior to each collection period, horses were confined to dry lots at 2000 h with no access to either hay or pasture. Following the overnight fast, horses were moved into the barn facility at 0800 h. Two consecutive baseline blood samples were collected (15 min apart) by venipuncture at 0800 and an oral dose of Karo Syrup (0.25 mL/kg BW; modified dosing as utilized by Jacob et al. [12]) was immediately administered after the second baseline sample. Subsequent blood samples were collected at 30, 60, 90, 120, 180, and 240 min following Karo Syrup administration. Whole blood was collected into sodium heparin Vacutainer tubes, which were inverted several times before being placed on ice. Due to logistics of distance and travel time (~10 min from the barn to the laboratory), samples were transported to the laboratory after the 120, 180, and 240 min samples. In the laboratory, whole blood was centrifuged at 3700 rpm for 7 min. Following centrifugation, plasma was harvested from the Vacutainer tubes and aliquoted into microcentrifuge tubes, which were then stored at −80 °C.2.4. Forage SamplingRepresentative hand-clipped forage samples were also collected (0800–1000 h) on three days per period for analysis of nutrient composition. Pasture samples were collected according to previously published procedures [1,24,32] and dried (60 °C; 36 h minimum) in a Thelco oven (Precision Scientific, Chicago, IL, USA). Samples were ground (1 mm) and submitted to a commercial laboratory (Equi-Analytical Laboratories, Ithaca, NY, USA) for analysis by near-infrared spectroscopy. The mean nutrient composition of hay diets and pasture forages are shown in Table 1.2.5. Analysis of Plasma Glucose and InsulinPlasma glucose was analyzed by colorimetric assay (Glucose C-2, Wako Chemicals, Richmond, VA, USA), with the commercial kit adapted for microplate assay following manufacturer instructions. Plasma insulin was evaluated using an enzyme-linked immunoassay (Mercodia Equine Insulin ELISA, Mercodia, Winston-Salem, NC, USA) previously validated in horses [35]. Inter-assay and intra-assay coefficients of variation for glucose were 4.0% and 2.9%, respectively. Inter-assay and intra-assay coefficients of variation for insulin were 7.8% and 3.4%, respectively.2.6. Fecal Sample AnalysesFecal pH was measured in duplicate with a handheld Accumet pH meter (Fisher Scientific; Waltham, MA, USA) using a previously published protocol for preparation of fecal slurries [24,36]. Short-chain and branched-chain fatty acid (SCFA and BCFA, respectively) concentrations were determined by GC-MS analysis of fecal samples. The SCFA analyzed included acetate, propionate, butyrate, and valerate. The BCFA analyzed included isobutyrate, isovalerate and isocaproate. Sample preparation was performed according to a previously published protocol [37]. In brief, frozen fecal samples were weighed and deposited in bead tubes over dry ice. Feces were then resuspended in 1 mL of 0.5% phosphoric acid per 0.1 g of sample and tubes were beaten for 5 min at 22.5 rpm in a cold case. Samples were again frozen at −80 °C until analyzed by Gas Chromatography and Mass Spectrometry system GC-MS (Agilent Technologies, Santa Clara, CA, USA). Prior to GC-MS analyses, thawed fecal suspensions were re-homogenized and centrifuged (10 min at 17,949× g), with the aqueous phase extracted using diethyl ether in a 1:1 volume to volume ratio. Before analysis, 2-methyl hexanoic acid (Thermo Fisher Scientific, Dallas, TX, USA) was added to the organic phase extract as an internal standard. Samples were analyzed in duplicate, with independent extractions for each replicate. The specifications of the GC-MS system and analyses as well as data acquisition and procedures for SCFA/BCFA quantitation have been previously described by Honarbakhsh et al. [37] and Garcia-Villalba et al. [38].Quick-DNA Fecal/Soil Microbe Kits (Zymo Research; Irvine, CA, USA) were used for DNA extraction (in triplicate). The highest yielding replicate (quantified with a Qubit 2.0 Flourometer [Invitrogen; Carlsbad, CA, USA]) was submitted to a commercial laboratory (RTL Genomics; Lubbock, TX, USA). Amplification of the V4-V5 region of the 16S rRNA gene was conducted using region specific primers (515F/926R) [39]; sequencing was conducted by Illumina MiSeq.2.7. Sequence and Statistical AnalysisSequence and statistical analyses were performed in QIIME 2 (Quantitative Insights into Microbial Ecology, v. 2020.8) [40] and R (v. 4.0.2) [41]. Network mapping was conducted in Cytoscape (v. 3.8.0) [42]. Animal was considered the experimental unit. Quality and chimera filtering of forward reads was conducted using DADA2 (read length = 185) [40,43,44]. Mafft and FastTree (q2-phylogeny plugin) were used to create trees for diversity analyses [45,46,47]. The lower quartile of Amplicon Sequence Variants (ASV) based on absolute abundance were removed (minimum frequency = 16; minimum samples = 4). The feature table was rarefied to a minimum sampling depth of 10,600 prior to α- and β-diversity analyses. The α-diversity metrics were analyzed by Kruskal–Wallis tests [48,49,50,51,52]. The β-diversity metrics analyzed included Weighted and Unweighted UniFrac by permutational ANOVA (PERMANOVA) [53,54,55,56,57,58,59]. Benjamini and Hochberg FDR adjustments for multiple pairwise comparisons were applied for all diversity analyses. Permutational multivariate analysis of dispersion (PERMDISP) was used to test homogeneity of dispersion [60]. To further explore differential abundances, ASV were then grouped into bacterial co-abundance groups (BCG) based on abundance profiles using Sparce Cooccurrence Network Investigation for Compositional Data (SCNIC) [61,62]. A random forest classifier with nested cross validation was applied to determine if forage type could be predicted based on BCG composition [63,64]. Features (BCG) were removed from the model based on importance scores generated by the random forest classifier in an iterative process (serial reduction with increments of 5) until the point at which model accuracy began to decline [65]. Relative abundances of remaining BCG and any uncorrelated ASV retained in the model following feature reduction processes were then analyzed by linear discriminant analysis effect size (LEfSe) to identify BCG specific to each forage type, with significance set at an LDA score >2.0 [66]. Taxonomy was then assigned using the latest SILVA database (SSU 138) [63,67,68,69,70].Glucose and insulin response variables as well as SCFA/BCFA concentrations and fecal pH were analyzed by mixed model ANOVA in R, with grazing system, forage type and their interactions set as fixed factors and horse as the random factor in the initial model. The area under the curve (AUC) was calculated based on the trapezoid rule as the positive incremental AUC utilizing a published macro [71] in SAS (v.9.4 SAS Institute, Cary, NC, USA). Proxies for insulin sensitivity (fasting glucose-to-insulin ratio (FGIR); reciprocal of the square root of insulin (RISQI)) and insulin secretory response (modified insulin-to-glucose ratio (MIRG)) were calculated using baseline (fasting) values as previously described [72,73]. The relationship between forage nutrients and SCFA/BCFA fermentation metabolites (as well as pH) and fecal microbial community composition was then explored using random forest regression with nested cross validation to determine if nutrient concentrations and/or metabolite concentrations could be predicted based on bacterial abundance profiles. Spearman correlations between BCG and forage nutrients and fecal metabolites were analyzed in R. Relationships between forage nutrients/fermentation metabolites and BCG were then visualized in Cytoscape. Spearman correlations between metabolic variables and the BCG, as well as between forage nutrients and metabolic responses were also evaluated in R. For all variables analyzed by mixed model, model residuals were analyzed for normality using the Shapiro–Wilk test. Log, square root, or inverse data transformations were applied where appropriate for non-normal data. Means were separated using Tukey’s method. When analyzing pairwise comparisons of transformed data, means and standard errors were back-transformed to the original variable scale following application of Tukey’s method with the delta method. For all analyses which generated p-values, results were considered significant at p ≤ 0.05, with trends considered at p ≤ 0.10. Data for variables analyzed by mixed model are presented as means ± SEM. Overall analysis of microbiome, fecal metabolite, and glucose/insulin data did not reveal differences by grazing system. Therefore, grazing system was removed from models and results for combined data are presented with n = 8.3. Results3.1. Initial 16s rRNA Sequence AnalysisA summary of 16S rRNA gene sequencing reads before and after quality and chimera filtering as well as after filtering of low abundance features in the 40 samples analyzed for this study is shown in Table 2. There were 1264 distinct ASV in the final dataset (taxonomy can be found in Table S3).3.2. Diversity AnalysesAll α-diversity metrics evaluated, including the Shannon Diversity Index, Faith’s Phylogenetic Diversity, Pielou’s Evenness, and Observed ASVs, differed by forage (Kruskal–Wallis tests with Benjamini and Hochberg FDR adjustments; p < 0.03; Figure 1a–d). Shannon Diversity was greater when horses were adapted to WSG than CSG-SP (p < 0.05), and there was a trend for greater diversity when adapted to HAY-SP compared to CSG-SP (p < 0.08) and WSG vs. CSG-FA (p = 0.09). However, the only differences in evenness (Pielou’s Evenness) were trends for greater evenness in WSG and CSG-SP than in CSG-FA (p < 0.06). Conversely, richness (Observed ASVs) was greater in horses adapted to WSG, CSG-FA, and HAY-FA than CSG-SP (p < 0.02), but did not differ between horses adapted to HAY-SP and CSG-SP. Similarly, phylogenetic differences (Faith’s Phylogenetic Diversity) were also found between CSG-SP and subsequent forages (WSG, CSG-FA, HAY-FA; p < 0.02). Principal coordinate analysis of β-diversity metrics including Weighted and Unweighted UniFrac did not reveal distinct clustering by forage (Figure 2a,b). However, statistical analysis by PERMANOVA (with Benjamini and Hochberg FDR adjustments) found significant differences in these measures (p ≤ 0.02). Subsequent PERMDISP analysis confirmed that these differences were not due to differences of variance or dispersion within groups. Unweighted UniFrac differed for all pairwise comparisons of forages (p < 0.02), with the exception of WSG vs. CSG-FA and WSG vs. HAY-FA for which there were trends for differences (p < 0.10). In contrast, there were no significant differences in any pairwise comparisons of forages for Weighted UniFrac. The percent of variation explained by PC1 was almost 3 times greater for Weighted than for Unweighted UniFrac, indicating an influence of abundance profiles in addition to phylogenetic differences. 3.3. Differential AbundanceApplication of SCNIC identified 333 BCG, with 224 individual ASV remaining ungrouped. Iterative reduction of features (BCG) based on random forest model importance scores revealed that model accuracy increased through reduction to the top 65 features (0.90 ± 0.09). Further feature reduction to the top 25 features (based on importance scores) did not impact random forest model accuracy (0.90 ± 0.09 with the top 25 features retained). All retained features were BCG; no ungrouped ASV remained in the reduced feature set. These 25 BCG were retained for further analysis, with 11.54% of the total microbial community abundance represented by the reduced 25 BCG feature set. The strength of the random forest model accuracy score indicated that forage type could be predicted based on bacterial composition, and that, conversely, bacterial community composition was influenced by forage. Subsequent Linear discriminant analysis Effect Size (LEfSe) analysis conducted on features retained from the random forest classification modelling identified 6 BCG as markers of HAY-SP, 3 BCG enriched in CSG-SP, 5 BCG for WSG, 5 BCG for CSG-FA, and 5 BCG for HAY-FA (LDA > 4.0; p < 0.03). Forage-specific BCG markers as well as taxonomic classifications of subsequent linear discriminant analysis Effect Size analysis conducted on features retained from the random forest classification modelling identified 6 BCG as markers of HAY-SP, 3 BCG enriched in CSG-SP, 5 BCG for WSG, 5 BCG for CSG-FA, and 5 BCG for HAY-FA (LDA > 4.0; p < 0.03). Forage-specific BCG markers as well as taxonomic classifications of individual ASV within each BCG are presented in Table 3, Table 4 and Table 5.Numerous taxa were represented in BCG markers for multiple forages. At the family level, the BCG markers for HAY-SP contained twice as many ASV mapped to the Lachnospiraceae family as any other forage (CSG-SP: 0; WSG: 2; CSG-FA: 3; HAY-FA: 2 ASV). The family Oscillospiraceae also had ASV members of BCG markers for all forages but CSG-SP (HAY-SP: 3; WSG: 3; CSG-FA: 2; HAY-FA: 2 ASV). At the genus level, ASV assigned to Christensenellaceae R-7 group were present in BCG markers of both hay diets (HAY-SP: 2; HAY-FA:1 ASV) and WSG (2 ASV); ASV assigned to the NK4A214 group of Oscillospiraceae were among members of BCG identified as markers of HAY-SP (2 ASV), WSG (1 ASV), and CSG-FA (1 ASV). The BCG markers of both hay diets and CSG-FA each included an ASV mapped to Rikenellaceae RC9 gut group, and BCG markers of HAY-SP and WSG each contained ASV assigned to Fibrobacter and Papillibacter. The BCG markers of HAY-SP and CSG-FA contained ASV within Catenisphaera and Lachnospiraceae UCG-009. Amplicon sequence variants mapped to the genus Clostridium sensu stricto 1 were found in BCG markers of CSG-SP and WSG. The BCG markers of WSG and CSG-FA included ASV assigned to Bacteriodales RF16 group and the Family XIII AD3011 group of Anaerovoracaceae. The BCG markers of CSG-FA and HAY-FA each contained an ASV assigned to Coprostanoligenes group, Lachnospiraceae XPB1014 group and Marvinbryantia, and ASV assigned to Treponema were assigned to BCG markers of WSG and HAY-FA. Amplicon sequence variants assigned to Ruminococcus, Pseudobutyvibrio, Anaerovibrio, probable genus 10 of Lachnospiraceae, and the p-251-o5 genus and family of the order Bacteroidales were only present in BCG markers identified for HAY-SP; no other ASV assigned to these taxa were found in BCG markers of other forages. An ASV within Bacteroidales BS11 gut group was the only taxa specific to BCG markers of CSG-SP. Amplicon sequence variants mapped to the genera Akkermansia, Mogibacterium, and the Hallii group of Lachnospiraceae as well as UCG-005 metagenome of Oscillospiraceae and Clostridium butyricum were only found in BCG markers of WSG. The BCG markers of CSG-FA contained ASV assigned to Alloprevotella, Erysipelatoclostridium, Bacteroidales UCG-001, the WCHB1-41 genus, family and class within the class Kiritimatiellae, and Denitrobacterium detoxificans; these taxa were not identified in BCG markers of other forages. Taxa specific to only BCG markers of HAY-FA included ASV from the Incertae Sedis genus of Ethanoligenenaceae as well as Streptococcus.3.4. Fecal pH and Fermentation MetabolitesFecal pH differed by forage type (mixed model ANOVA with Tukey’s post hoc adjustment; p < 0.0001). Fecal pH was greater in horses adapted to WSG (7.56 ± 0.18) and HAY-FA (7.57 ± 0.18) than in HAY-SP (6.73 ± 0.18), CSG-SP (6.58 ± 0.18), or CSG-FA (6.53 ± 0.18; p ≤ 0.02; Figure 3). Fecal concentrations of short-chain fatty acids (SCFA) and branched-chain fatty acids (BCFA) also differed by forage (mixed model ANOVA with Tukey’s post hoc adjustment; p < 0.001; Table 6).Unlike fecal pH, differences in SCFA and BCFA were primarily between pasture forages and the hay diets, with horses adapted to WSG often intermediate (numerically) between cool-season pasture and hay diets. Total BCFA were greater in CSG-SP, WSG, and CSG-FA than for either HAY-SP or HAY-FA (p < 0.05). Total SCFA were greater in CSG-SP and CSG-FA than HAY-SP or HAY-FA; WSG did not differ from CSG-SP and CSG-FA or HAY-SP, but was greater in comparison to HAY-FA (p < 0.002). Similarly, acetate was greater in CSG-FA than HAY-SP or HAY-FA, while WSG only differed from HAY-FA (p < 0.02). Acetate concentrations for CSG-FA also differed with HAY-FA (p = 0.0003), but there was only a trend for a difference between CSG-SP and HAY-SP (p = 0.09). Fecal butyrate was lower for HAY-FA than when horses were adapted to any of the pasture forages (p < 0.03), but there was no difference between HAY-SP and any other forage. Fecal propionate was lowest in horses adapted to HAY-FA (p < 0.002), but propionate concentrations were also lower for HAY-SP than either CSG-SP or CSG-FA (p < 0.03). Propionate did not differ between horses adapted to WSG and all other forages. Fecal valerate was lower for horses adapted to HAY-FA than CSG-SP or CSG-FA (p < 0.03) but did not differ from WSG or HAY-SP. Horses adapted to HAY-SP also had fecal valerate concentrations lower than CSG-SP or CSG-FA (p ≤ 0.0002). Isobutyrate was greater for all pasture forages in comparison to both HAY-SP and HAY-FA (p < 0.02). Isovalerate was greater in CSG-SP and CSG-FA vs. HAY-SP and HAY-FA (p ≤ 0.002), while concentrations in horses adapted to WSG were once again intermediate. Isocaproate was detected in all fecal samples, but concentrations were negligible and below the limit of quantification (<1.0 ug g feces−1).3.5. Relationships between Fecal Microbiota and MetabolitesRandom forest regressors were applied to determine if fecal pH and metabolite concentrations could be predicted based on microbial composition. Random forest regression did not support a strong influence of bacterial composition of the full microbial community on these fecal variables. Relatively weak predictive accuracy was found for isovalerate, valerate, hexanoate, and heptanoate (model R2 and p-values are shown in Table 7). Major SCFA including acetate, butyrate, and propionate could not be predicted based on BCG abundance profiles.However, when random forest regressors were applied to only the top 25 BCG identified as most predictive of forage type, model accuracies improved for all fecal variables with the exception of valerate (Table 7). While predictive accuracy was still relatively weak, all metabolites (and pH) could be predicted with statistically significant accuracy (p ≤ 0.001).Individual correlations were found between 19 of these BCG and at least one fecal metabolite or pH (Spearman correlation; rs ≥ |0.30|; p ≤ 0.05; Figure 4), with most BCG correlated with multiple fecal variables. There were positive correlations between BCG_300 and total SCFA and total BCFA as well as with acetate, butyrate, propionate, valerate, isobutyrate, and isovalerate, while this BCG was negatively correlated with hexanoate, heptanoate, and pH. The ASV members of this BCG were assigned to the genera Lachnospiraceae XPB1014 and Streptococcus. There were also positive correlations between both BCG_9 and BCG_259 and total SCFA and BCFA in addition to acetate, butyrate, propionate, valerate, isobutyrate, and isovalerate; these BCG were negatively correlated with heptoanate. These BCG included multiple ASV assigned to Lachnospiraceae as well as ASV within the genera Christensenellaceae R-7 group, Rikenellaceae RC9 gut group, Papillibacter, and the NK4214 group of Oscillospiraceae. Positive correlations were also found between BCG_173 and BCG_201 and total SCFA and BCFA, acetate, propionate, valerate, isobutyrate, and isovalerate, with negative correlations between these BCG and both hexoanate and heptoanate. These BCG included ASV assigned to Treponema, Rikenellaceae RC9 gut group, the NKA214 group of Oscillospiraceae, and the p-251-o5 and F082 genera and families of Bacteroidales. Total SCFA and BCFA as well as acetate, butyrate, propionate, isobutyrate, and isovalerate were positively correlated with BCG_113, while hexanoate was negatively correlated. This BCG contained ASV assigned to taxa including Synergistaceae, Bacteroidales RF16 group, Papillibacter, Christensenellaceae R-7 Group, Mogibacterium, and Clostridium butyricum.Seven BCG were positively correlated with total BCFA and some combination of propionate, valerate, isobutyrate, and isovalerate, while negative correlations were found between these BCG and hexanoate and/or heptanoate. These BCG included ASV members assigned to the family level for Anaerovoracaceae, Oscillospiraceae, and Lachnospiraceae. Other ASV within these BCG were assigned to Lactobacillus equigenerosi and the genera Akkermansia, Fibrobacter, Treponema, Sphaerocheata, Phasolarctobacterium, Christensenellaceae R-7 group, Lachnoclostridium, Cellulosilyticum, Marvinbryantia, Coprostanoligenes group, Lachnospiraceae XPB1014, Lachnospiraceae UCG-009, Erysipelatoclostridium, Clostridium sensu stricto 1, and Bacteroidales BS11 gut group as well as Family XIII AD3011 group of Anaerovoracaceae, the UCG-004 genus within Erysipelatoclostridiaceae, the UCG-002 and UCG-005 genuses of Oscillospiraceae, and the UCG-010 genus and family of Oscillospirales.Fecal pH was positively correlated with five BCG (in addition to BCG_300) (rs ≥ |0.30|; p ≤ 0.05), which contained ASV from taxa including Anaerovoracaceae, the Family XIII AD3011 group of Anaerovoracaceae, Akkermansia, Christensenellaceae R-7 group, Fibrobacter, Treponema, Catenisphaera, Alloprevotella, Bacteroidales RF16 group, Marvinbryantia, Coprostanoligenes group, Bacteroidales UCG-001, the NK4A214 group of Oscillospiraceae, and the WCHB1-41 genus, family, and order within Kiritimatiellae. Fecal pH was negatively correlated with BCG_94. This BCG included ASV assigned to Prevotella, Sarcina, and Clostridium sensu stricto 1.3.6. Relationships between Fecal Microbiota and Forage NutrientsApplication of random forest regressors demonstrated that forage nutrient concentrations could be predicted based on bacterial community composition (BCG composition of the full microbial community), including water-soluble carbohydrate (WSC) and NSC at a predictive accuracy > R2 = 0.50 and crude protein (CP) with a predictive accuracy of R2 = 0.41 (p < 0.0001; Table 7). Weaker, but still statistically significant, predictive accuracy was found for digestible energy (DE), acid detergent fiber (ADF), neutral detergent fiber (NDF), ethanol-soluble (ESC), and starch (p < 0.04).Similar to results for fecal metabolites, when random forest regressors were applied only to the top 25 BCG identified as most predictive of forage type, model accuracies improved (Table 7). Forage CP, NSC, and WSC remained as the nutrients most accurately predicted by the bacterial composition of this subset of BCG, all with R2 > 0.60 (p < 0.0001). Model predictive capacity with the reduced feature set was also above R2 = 0.40 for starch; weaker, but statistically significant, predictive accuracy was found all other nutrients (p < 0.0001). Subsequent correlation analysis found 23 BCG (of the 25 in the reduced feature set) were correlated with at least one forage nutrient (Spearman correlation; rs ≥ |0.30|; p ≤ 0.05; Figure 5). Nine BCG were correlated with ADF, NDF, CP, and DE. In all cases, opposite correlations were seen between ADF/NDF and CP/DE (i.e., if ADF and NDF were positively correlated with a BCG, CP and DE were negatively correlated). There were distinct ASV from the taxa Christensenellaceae R-7 group and NK4A214 group of Oscillospiraceae that were members of separate BCG which were positively and negatively responding to these nutrients (i.e., these taxonomic classifications were found across all positive and negative responder groups). The BCG negatively correlated with ADF/NDF, and thus positively correlated with CP/DE included ASV members assigned to Clostridium butyricum and genera Papillibacter, Lachnoclostridium, Cellulosilyticum, Fibrobacter, Treponema, Catenisphaera, Bacteroidales RF16 group, Synergistaceae, Mogibacterium, Rikenellaceae RC9 gut group, the Hallii group of Lachnospiraceae, the p-251-o5 genus and family of Oscillospiraceae, and the F082 genus and family of Bacteroidales, as well as multiple ASV assigned only to the family level for Lachnospiraceae. The BCG positively correlated with ADF/NDF and negatively correlated with CP/DE included ASV within Sphaerochaeta, Phascolarctobacterium, Bacteroidales UCG-001, Rikenellaceae RC9 gut group, Coprostanoligenes group, the UCG-002 genus within Oscillospiraceae, and the UCG-010 genus and family within Oscillospirales. Four additional BCG were correlated with some combination of ADF/NDF and CP/DE. The BCG negatively correlated with ADF/NDF, and thus positively correlated with CP/ DE contained ASV members assigned to additional taxa including Ruminococcus, Anaerovibrio, Pseudobutyvibrio, Lachnospiraceae UCG-009, probable genus 10 of Lachnospiraceae, and the WCHB1-41 genus, family and order of Kiritimatiellae. The BCG positively correlated with ADF/NDF, and negatively correlated with CP/DE included ASV within Bacteroidales BS11 gut group and Clostridium sensu stricto 1. Forage ADF and NDF were also negatively correlated with BCG_300 (Streptococcus and Lachnospiraceae XPB1014), but this BCG was not correlated with either CP or DE. Forage CP was also positively correlated with BCG_35, for which there was no relationship with ADF, NDF, or DE, but for which there was also a negative correlation with NSC and WSC. The ASV members of BCG_35 were assigned to taxa including Akkermansia, Fibrobacter, Treponema, Christensenellaceae R-7 group, and Family XIII AD3011 within Anaerovoracaceae.Forage NSC and WSC were correlated with 13 and 12 BCG, respectively (rs ≥ |0.30|; p ≤ 0.05). In addition to the above-mentioned taxa from ASV within BCG_35, ASV within BCG negatively correlated with both NSC and WSC were assigned to Lactobacillus equigenerosi and genera including Lachnoclostridium, Cellulosilyticum, Catenisphaera, Marvinbryantia, Erysipelatoclostridium, Mogibacterium, Lachnospiraceae XPB1014 group, Lachnospiraceae UCG-009, Coprostanoligenes group, UCG-005 metagenome within Oscillospiraceae, the Hallii group within Lachnospiraceae, the NK4A214 group of Oscillospiraceae, the Incertae Sedis genus of Ethanoligenenaceae, the WCHB1-41 genus, family, and order within Kiritimatiellae, as well as Denitrobacterium detoxificans. Additional ASV were assigned only to the family level for Synergistaceae, Lachnospiraceae, Oscillospiraceae, and Anaerovoracaceae. Forage NSC and WSC were positively correlated with BCG_94 (Clostridium sensu stricto 1, Prevotella, and Sarcina). Forage ESC was correlated with 9 of the same BCG as NSC and WSC, but also had a negative correlation with BCG_173, which contained multiple ASV assigned to Lachnospiraceae, probable genus 10 within Lachnospiraceae, Pseudobutyvibrio as well as to the p-251 genus and family of Bacteroidales. Starch was positively correlated with BCG_113 and BCG_300 and negatively correlated with two BCG, BCG_43 and BCG_48, containing ASV from Sphaerochaeta, Phascolarctobacterium, Christensenellaceae R-7 group, Coprostanoligenes group, Bacteroidales UCG-001, Rikenellaceae RC9 gut group, the NK4A214 group of Oscillospiraceae, the UCG-002 genus within Oscillospiraceae, and the UCG-010 genus and family within Oscillospirales; these BCG were also among those positively correlated with ADF/NDF and negatively correlated with CP/DE. Relationships between fecal variables and forage nutrients were also evaluated through correlation analysis (Figure 6). Acetate, butyrate, propionate, valerate, isobutyrate, isovalerate and total SCFA and BCFA were positively correlated with DE (Spearman correlation; rs ≥ 0.62) and CP (rs ≥ 0.41), but were negatively correlated with NDF (rs ≤ −0.48) and ADF (rs ≤ −0.57; p ≤ 0.008). There was also a weaker positive correlation between these fecal metabolites and starch (rs ≥ 0.39; p ≤ 0.01). Total BCFA, isobutyrate, and isovalerate were negatively correlated with NSC, WSC, and ESC (rs ≤ −0.37; p ≤ 0.02), with a weaker negative correlation between valerate and NSC and WSC (rs ≤ −0.34; p ≤ 0.03). Hexanoate and heptanoate were negatively correlated with DE, CP, and starch (rs ≤ −0.39; p ≤ 0.01), but were positively correlated with NDF, ADF, NSC, WSC, and ESC (rs ≥ 0.37; p ≤ 0.02). Conversely, fecal pH was positively correlated with ADF (rs = 0.47; p = 0.002), but negatively correlated with DE, NSC, WSC, ESC, and starch (rs ≤ −0.32; p ≤ 0.04).3.7. Blood Samples and Glycemic/Insulinemic ResponsesPlasma glucose responses to OST administration differed by forage (mixed model ANOVA with Tukey’s post hoc adjustment; p ≤ 0.01). The AUC for glucose was lowest when horses were adapted to WSG (42.4 ± 6.8 mg/dL*h) vs. CSG-SP (70.0 ± 6.8 mg/dL*h) and HAY-FA (76.3 ± 6.8 mg/dL*h; p ≤ 0.03; Figure 7a). Peak plasma glucose was also lower for WSG (110 ± 3 mg/dL) in comparison to HAY-FA (123 ± 3 mg/dL; p = 0.01), and there was a trend for lower peak plasma glucose for WSG than for CSG-SP (120 ± 3 mg/dL; p = 0.06; Figure 7b). Forage did not, however, impact OST insulin responses. Neither AUC (CSG-SP: 36.1; WSG: 29.1; HAY-FA: 32.7 ± 6.4 mIU/L*h) or peak plasma insulin (CSG-SP: 28.8; WSG: 25.5; HAY-FA: 26.6 ± 3.3 mIU/L) varied by forage (Figure S2). Fasting plasma insulin (CSG-SP: 3.69; WSG: 5.04; HAY-FA: 5.20 ± 0.64 mIU/L) and glucose (CSG-SP: 81.4; WSG: 78.7; HAY-FA: 81.3 ± 1.4 mg/dL) also did not differ by forage (Figure S3). There were trends for differences by forage in proxies for insulin sensitivity including the fasting glucose-to-insulin ratio (FGIR) and reciprocal of the square root of insulin (RISQI; mixed model ANOVA with Tukey’s post hoc adjustment; p ≤ 0.10; Figure S4). Differences in FGIR and RISQI were limited to WSG (FGIR: 25.2 ± 16.4; RISQI: 0.58 ± 0.04) vs. HAY (FGIR: 16.4 ± 2.1; RISQI: 0.46 ± 0.04; p = 0.08). However, the modified insulin-to-glucose ratio (MIRG), a proxy for the insulin secretory response, did not vary by forage (Figure S4).3.8. Relationships between Glucose Metabolism and the Fecal MicrobiotaGlucose and insulin dynamics were only assessed after three of the forages (CSG-SP, WSG, and HAY-FA), and this smaller dataset precluded the use of random forest regression modelling for these variables. As the primary metabolic differences in grazing horse metabolism were AUC and peak plasma glucose in response to OST administration, correlation analysis was conducted to explore relationships between these metabolic variables and BCG. Only one BCG (of the 25 BCG in the reduced feature set) was correlated with AUC (rs = −0.48; p = 0.04), with members of BCG_124 including ASV within Papillibacter, Christensenellaceae R-7 group, and Synergistaceae. Four additional BCG were correlated with peak plasma glucose. Positive correlations were found between peak glucose and BCG_51 and BCG_305 (rs ≥ 0.43; p ≤ 0.04). These BCG included ASV mapped to genera including Lachnospiraceae XPB1014, Lachnospiraceae UCG-009, Erysipelatoclostridium, Catenisphaera, and Fibrobacter as well to the family level for Anaerovoracaceae. Conversely, BCG_113 (Clostridium butyricum, Papillibacter, Bacteroidales RF16 group) and BCG_259 were negatively correlated with peak glucose (rs = −0.44; p = 0.03). Members of BCG_259 included ASV assigned at the family level to Lachnospiraceae in addition to the Hallii group within Lachnospiraceae. While relationships between AUC and peak plasma glucose and forage nutrients were identified through correlation analysis (Spearman correlation; rs ≥ |0.41|; p ≤ 0.04; Figure 7c), these metabolic variables were not correlated with any fecal variables including SCFA, BCFA, and pH. Forage NSC, WSC, and ESC were positively correlated with AUC (rs ≥ 0.53; p ≤ 0.007) and peak glucose (rs ≥ 0.41; p ≤ 0.04). There was also a negative correlation between CP and glucose responses to OST administration (rs ≤ −0.50; p ≤ 0.01). Forage DE, NDF, ADF, and starch, however, were not correlated with AUC or peak glucose.4. Discussion4.1. Forage Type and the MicrobiomeWarm-season grasses can be utilized to bridge the “summer slump” forage gap in cool-season grass grazing systems [1,2], and differences in NSC between these forage types could have implications for equine metabolic health [1,5,9]. However, there is limited information on the impacts of grazing warm-season grasses on the equine hindgut microbiome as well as potential associations between forage nutrients, the hindgut microbiome, and equine metabolic responses. Therefore, this study aimed of to characterize the fecal microbiota of horses adapted to different forage types and to explore relationships between forage nutrients, microbial composition, fecal metabolites, and glycemic responses of grazing horses. Results of this study clearly demonstrated that shifts in fecal microbiome structure and species composition occur as horses are adapted to different forages within an integrated warm- and cool-season grass rotational grazing system. This was supported by statistical differences in both α- and β-diversity. Furthermore, random forest classification modelling was able to predict forage type based on microbial community composition, indicating the influence of forage type on the fecal microbiome. Finally, forage-specific BCG were identified through LEfSe analysis, but the 25 BCG identified as most predictive of forage type through random forest modeling and subsequently analyzed by LEfSe represented only ~10% of the total fecal microbiota across all forages. This indicates that distinct and identifiable shifts in microbial composition do occur as horses adapt to different forage types. However, the majority of the microbiome (~90% in the current study) is resistant and/or resilient to potential perturbations induced by transitioning among forage types with different physical and chemical properties, including between cool-season and warm-season grass pastures.The BCG identified as markers specific to HAY-SP included multiple ASV assigned to taxa to which fibrolytic and butyrate-producing functions have been previously ascribed [74,75,76]. The BCG specific to this forage contained a heavy representation of ASV within the Lachnospiraceae family as well as Fibrobacter, Ruminococcus, and Pseudobutyvibrio. Prior studies have also reported increased prevalence of Lachnospiraceae in horses fed hay vs. pasture [19,77]. Conversely, ASV assigned to Anaerovibrio were also identified in BCG markers of HAY-SP. Increases in this genus in response to abrupt inclusion of dietary starch have been documented [78] as well as in temporal proximity to oral administration of oligofructose in experimental laminitis induction models [79,80]. The co-occurrence of bacteria in these taxonomic groups could potentially be reflective of the nutrient composition of HAY-SP, which was the highest in both fiber and NSC of all forages. Co-occurrence of these bacteria may also reflect cross-feeding relationships between bacteria and/or metabolic plasticity of bacterial populations. These factors may also account for unexpected associations revealed by analysis in the current study, such as ASV assigned to Streptococcus within BCG markers of HAY-FA, despite the relatively low NSC content of this forage. However, Zhu et al. [77] also reported a lower abundance of species within Streptococcaceae in horses maintained on pasture vs. horses fed hay or silage [77].Overall, the fecal microbiota of horses adapted to cool-season pasture was characterized by bacteria capable of utilizing rapidly fermentable fibers, which could explain the greater fecal SCFA concentrations for CSG-SP and CSG-FA. Capacity for hemicellulose fermentation and SCFA production have been previously documented in the Bacteroidales BS11 gut group [81], which was identified only in BCG markers of CSG-SP. Equine studies have found increased relative abundance of Bacteroidales BS11 gut group in horses fed barley vs. a hay diet [82] and in horses presenting with colic [83,84]. For CSG-FA, BCG markers included genera such as Alloprevotella and Erysipelatoclostridium, which are also associated with degradation of fermentable fibers [85,86,87]. The BCG markers of CSG-FA also included ASV assigned WCHB1-41 clade within Kiritimatiellae. Prior studies have found enrichment of the phylum Kiritimatiellaeota in the fecal microbiota of horses with equine metabolic syndrome and insulin dysregulation [26,27]. In contrast to results of the present study, Fitzgerald et al. [27] reported increased abundance of this phylum in response to a change from pasture to a hay diet in both healthy and insulin dysregulated ponies, and Ericsson et al. [88] similarly found increased abundance of the class Kiritimatiellae when horses with equine metabolic syndrome were transitioned from pasture to a hay diet. These conflicting results reinforce that nutritional composition of specific forages, rather than form of forage alone, need to be considered when evaluating interstudy effects of diet on the equine microbiome.A number of interesting relationships were found between BCG identified as markers specific to WSG and fecal variables, forage nutrients, and horse glucose metabolism. Warm-season grasses are characteristically lower in soluble carbohydrates than cool-season grasses [1,5], and of all forages evaluated in the current study, NSC and WSC were lowest in WSG. Four of the five BCG markers of WSG were negatively correlated with forage NSC and WSC, including BCG_35, which contained ASV assigned Akkermansia as well as Fibrobacter, Treponema, Christensenellaceae R-7 group, and the Family XIII AD3011 group of Anaerovoracaceae. Akkermansia has been linked to metabolic health via regulation of inflammatory responses and has also been explored for probiotic use in animal species [89,90,91,92]. Lindenberg et al. [93] recently demonstrated immune (and inflammatory) modulation by specific hindgut microbiota in horses, including Akkermansia spp. Furthermore, supplementation with mannanoligosaccharides and fructooligosaccharides led to increased abundance of Akkermansia muciniphilia in a subsequent study [94], and improvements in insulin sensitivity have been previously documented in horses supplemented with low doses of fructooligosaccharide [95,96]. Markers of WSG also included BCG_113, which contained an ASV assigned to Clostridium butyricum. Clostridium butyricum is a prolific butyrate producer that has also been investigated for probiotic use due to its capacity to promote anti-inflammatory responses and improve gut barrier function and metabolic health [97,98,99]. BCG_35 was positively correlated with CP as well as isobutyrate and total BCFA, which reflects the proteolytic capacity of this BCG. BCG_113 was also positively correlated with total SCFA and all three major SCFA (acetate, butyrate, and propionate) in addition to isobutyrate, isovalerate, and total BCFA. This BCG was also negatively correlated with peak plasma glucose. These findings suggest that this bacterial group including Clostridium butyricum could play a role in modulation of equine metabolic health. While limited connections between the gut microbiota and metabolic responses were observed in the current study, three of the five BCG correlated with AUC and peak plasma glucose were WSG-specific BCG markers. Further research is necessary to determine if other WSG species or varieties would produce similar effects as those observed in the current study.While Akkermansia and Clostridium butyricum have been investigated in mice and other species, comparatively little research has been conducted to understand the function of these bacteria in the equine hindgut. The relationships found between ASV in these taxa and forage nutrients, fecal metabolites, and equine glycemic responses in addition to associations with low-NSC warm-season grasses in the current study support further research to determine the role of these bacteria, factors including dietary interventions that can promote prevalence in the hindgut, and potential probiotic applications.4.2. Forage Nutrients and the MicrobiomeResults of this study also confirmed the strong influence of dietary nutrients on the equine microbiome and provided insights into the complex relationships between forage nutrients and the gut microbiota. The concentrations of several nutrients could be predicted based on microbial community composition. Furthermore, regression model accuracy improved when the reduced feature set identified through prior forage classification modeling was utilized for prediction of nutrient concentrations. This indicates that differences in forage nutrients were driving the shifts in equine fecal microbial communities as horses adapted to various forages within the integrated grazing system. Results of this study revealed the influence of soluble carbohydrates including NSC and WSC on fecal microbial community composition. In contrast, fiber was not identified as a key nutrient shaping differences in microbial communities. Forage NSC and WSC concentrations could be predicted based on microbial community composition with an accuracy of R2 = 0.61 and R2 = 0.67, respectively, while predictive accuracy for ADF and NDF was < 0.40 (R2). Random forest modeling also revealed that in addition to soluble carbohydrates, the gut microbiota were also influenced by CP (R2 = 0.62). The impact of CP on the gut microbial community was interesting, but unsurprising, as two-thirds to three-fourths of total tract nitrogen digestion and absorption occurs in the equine hindgut [100], and many microbial species are capable of proteolytic fermentation [101]. Prior studies in horses have found distinct effects of high-fiber vs. lower-fiber diets on the hindgut microbial community, with benefits of increased fiber including greater microbial diversity [102], reduction of lactic-acid producing bacteria [101,103], and less-acidic hindgut pH [104,105]. However, these previous studies have been primarily conducted in horses fed concentrate vs. forage-based diets in which there was broad variation in fiber concentrations between treatments. In the current study, horses were maintained on a forage-only diet throughout, and there were relatively high concentrations of NDF and ADF found across all forages. It is possible that only minimal shifts occur in bacterial populations most responsive to fiber above a certain concentration threshold, and that these populations remained relatively stable throughout the study (and thus fiber concentration was not able to be predicted with strong accuracy by random forest modeling). However, it should be noted that the concentrations of NSC, WSC, and CP in forages is low in comparison to that of fiber. Additionally, in contrast to fiber, these nutrients are all subject to digestion in the equine foregut, and only a fraction of the total soluble carbohydrates and CP ingested would be available for bacterial fermentation in the hindgut. Thus, only a small amount of these nutrients were capable of exerting influence on gut microbial communities. Digestibility was not evaluated in the current study, however, and a more in-depth analysis of total digestibility and digestibility of specific nutrients would be necessary to fully understand the interactions between forage nutrients and the gut microbiota.4.3. Fecal pH and Fermentation MetabolitesFecal pH is commonly utilized as a marker of microbial activity in the equine hindgut [20,103]. Differences in fecal pH in the current study indicated that functional changes occur in the equine gut microbiota, in addition to shifts in microbiome structure and composition, as horses adapt to different forages. Fecal pH was highest in horses adapted to WSG and HAY-FA, which were lower in NSC than CSG-SP, CSG-FA. Accordingly, fecal pH was negatively correlated with forage NSC. Lower fecal pH is broadly associated with hindgut dysfunction in horses [106,107], and thus the higher fecal pH in horses adapted to WSG in comparison to cool-season pasture could suggest some benefit to grazing horses on warm-season grass pastures. However, mean pH was above 6.5 for CSG-SP, CSG-FA, and HAY-SP, which is within ranges previously documented in healthy forage-fed horses [33,108]. Therefore, while statistically significant, the difference in fecal pH between horses adapted to WSG vs. cool-season pasture may not be of physiological relevance.Differences in fermentation metabolites across forages did not mirror results for fecal pH, as differences between forages for SCFA and BCFA were primarily between pasture forages and the hay diets. Numerically, SCFA and BCFA concentrations in horses adapted to WSG were intermediate between the cool-season pasture (greatest) and hay (lowest), but these differences were not statistically significant. Furthermore, while statistically significant, the predictive accuracy of random forest regressors for fecal metabolites and pH were lower in comparison to those found for forage nutrients. Numerous studies have noted that while the gastrointestinal tract harbors a phylogenetically diverse community, many unrelated microorganisms are capable of performing similar functions [109,110,111,112,113]. The relatively poor predictive accuracy of regressors for fecal metabolites in the current study is likely due to this functional redundancy within the equine microbial community. Functional redundancy has been previously suggested as a contributing factor in stability of the gut microbial communities [111,113] including within the equine microbiome during transitions between warm- and cool-season grasses [24]. 4.4. Glucose and Insulin MetabolismWhile there were distinct shifts in the fecal microbiota across forages in the integrated system, forage type exerted minimal effects on glucose metabolism. A large body of research, primarily conducted in mouse models, has established that diet (and dietary intervention) modulates metabolic health in a microbiome-dependent manner [29,30,114]. However, shifts in gut microbial structure and function may precede changes in metabolic outcomes [115]. The adaptation period utilized in the present study was of a similar duration as used in prior studies evaluating the microbiome of forage-fed horses [22,23,116], and is considered sufficient for stabilization of microbial communities [95]. However, longer treatment periods are often required to assess metabolic adaptations to diet [13,103,117,118]. The lower AUC and peak plasma glucose in horses adapted to WSG in the current study does suggest that glucose may be cleared more rapidly in horses adapted to this forage, but a longer adaptation period would be necessary to confirm these results. Regardless, glucose and insulin responses to the OST for all forages evaluated in the current study were within normal and previously reported ranges [12,34], indicating that substantial metabolic changes are unlikely within the context of integrated rotational grazing management even with the observed changes in the fecal microbiota, fecal metabolites, and pH. Additionally, the lack of correlation between any fermentation metabolites and glucose responses, as well as the relatively small number of BCG correlated with AUC and peak plasma glucose, suggest that any difference in glycemic responses of horses in the current study may not have been heavily dependent on shifts in the hindgut microbiome.4.5. Additional ConsiderationsSeasonal and environmental factors should be considered when interpreting results of this study. Prior studies have reported seasonal changes in microbial diversity and species composition [119,120,121] but also lacked a true seasonal control, and thus, the effect of seasonality on the equine hindgut microbiome requires further investigation. Seasonal variance in insulin sensitivity has been documented in grazing horses [15,122], but when fed controlled diets without nutrient fluctuations seen in pasture forages, minimal seasonal differences in circulating glucose and insulin and insulin sensitivity have been found in healthy horses [15,123,124]. Environment and management factors beyond season can also impact pasture forage nutrient composition. Characterizing the microbiome of a larger number of grazing horses over multiple years and in multiple locations/regions would allow a more robust evaluation of the impacts of cool- versus warm-season grasses on the equine hindgut microbiome.Finally, it should be noted that the analytical approach in this study differs from more conventional taxon-based analysis. Rather, this study implemented a guild-based approach, grouping individual microbial ASV by co-abundance to subsequent analysis of abundance profiles [123,124,125,126,127]. This strategy has been previously utilized as an alternative to grouping bacteria by taxonomy [24,29,30,92,121]. Substantial genetic variation is possible within taxa, even at the species level, and therefore bacteria with similar taxonomic assignments may not represent a functionally homologous group [29,92,127]. Results of the current study also illustrate this concept, as ASV with the same assigned taxonomy were found in BCG characteristic of different forages as well as in separate BCG that positively and negatively responded to forage nutrients and with both positive and negative relationships with fecal metabolites and grazing horse metabolism.5. ConclusionsIn conclusion, distinct shifts in equine fecal microbial community structure and composition occur as horses adapt to different forages within an integrated warm- and cool-season grass rotational pasture system, but a substantial impact of this management practice on glucose metabolism in healthy adult grazing horses is unlikely. Forage NSC, WSC, and CP were the most influential nutrients driving these shifts in microbial composition. Results of this study underscore the potential for relatively small amounts of NSC to influence hindgut microbial composition and also that protein utilization may be an important ecological niche within the microbiome of forage-fed horses. Fecal BCFA and SCFA concentrations were higher in horses adapted to all pasture forages versus hay, but in comparison to forage nutrients, bacterial community composition did not have as strong an impact on fermentation metabolites, likely reflecting functional redundancy of the microbial community. The guild-based analytical approach utilized in this study also identified key relationships between specific bacterial groups associated with adaptation to warm-season grass pasture and forage nutrients, fecal metabolites, and equine glycemic responses to administration of oral sugar tests. Relationships identified in this study revealed new insights and targets for future research necessary to better understand the function of Akkermansia spp. and Clostridium butyricum in the hindgut microbiome of grazing horses, as these bacteria may play a role in modulation of equine metabolic health.

animals : an open access journal from mdpi

10.3390/ani11051268

PMC8145500

The development of mammary gland is directly related to the productivity of dairy animals. Some studies showed that feeding the enhanced plane of nutrition at pre-weaning stage are advantageous to the development of mammary gland. However, regulators which are involved in this biological process remain largely unknown. In this work, we have identified some long intergenic non-coding RNAs (lincRNAs) in mammary parenchyma (PAR) and mammary fat pad (MFP) of heifer calves under different levels of nutrition at pre-weaning stage by using the published RNA-seq database. Furthermore, those putative lincRNAs, which were highly correlated with these key protein-coding genes in mammary gland development, were highlighted. Our results not only confirmed the advantages of feeding calves with enhanced feeding plane in pre-weaning stage, but also provided fundamental base for further research on the biological processes of mammary gland development.

Enhanced plane of nutrition at pre-weaning stage can promote the development of mammary gland especially heifer calves. Although several genes are involved in this process, long intergenic non-coding RNAs (lincRNAs) are regarded as key regulators in the regulated network and are still largely unknown. We identified and characterized 534 putative lincRNAs based on the published RNA-seq data, including heifer calves in two groups: fed enhanced milk replacer (EH, 1.13 kg/day, including 28% crude protein, 25% fat) group and fed restricted milk replacer (R, 0.45 kg/day, including 20% crude protein, 20% fat) group. Sub-samples from the mammary parenchyma (PAR) and mammary fat pad (MFP) were harvested from heifer calves. According to the information of these lincRNAs’ quantitative trait loci (QTLs), the neighboring and co-expression genes were used to predict their function. By comparing EH vs R, 79 lincRNAs (61 upregulated, 18 downregulated) and 86 lincRNAs (54 upregulated, 32 downregulated) were differentially expressed in MFP and PAR, respectively. In MFP, some differentially expressed lincRNAs (DELs) are involved in lipid metabolism pathways, while, in PAR, among of DELs are involved in cell proliferation pathways. Taken together, this study explored the potential regulatory mechanism of lincRNAs in the mammary gland development of calves under different planes of nutrition.

1. IntroductionThe mammary gland is a complex organ, distinguishing mammals from other animal species, undergoing through a series of developmental changes at different physiological stages starting from embryo to pubertal stage following reproductive stage, also known as mammogenesis. It is composed of different types of cells like parenchyma cells, mammary fat pad (MFP), fibroblasts and vascular endothelial cells [1]. Parenchyma (PAR) is the key tissue in synthesis and secretion of milk, while, the mammary fat pad (MFP) tissue is used to provide the protection and support for PAR [2]. MFP is necessary for development of the secretory epithelium and provides signals that mediate ductal morphogenesis and potentially alveolar differentiation. Moreover, these MFPs are susceptible to dietary changes [3]. Although, mammary gland ducts’ elongation and branching mainly occur in the pubertal period. But, the pre-weaning stage of mammary gland played an important role in heifer’s future milk yield [4,5]. Previous studies showed that the EH and R nutritional levels have different impacts on the mammary gland mass and composition of PAR and MFP in the pre-weaning stage. A number of genes have been identified that are involved in mammary gland development under different dietary planes, like, EGF, FGF2, IGF-1, etc. [6]. Still, there is a dire need to identify novel genes and their interaction in different tissues of the mammary gland.Long non-coding RNAs (lncRNAs) have been defined as transcripts of length ≥200 nt that lacks protein-coding potential [7]. According to the genomic location and context, lncRNAs are divided into five classes, including intergenic, sense, antisense, intronic and bidirectional non-coding RNAs and vast majority is long intergenic non-coding RNAs (lincRNAs) [8,9]. LincRNAs have diverse different features form messenger RNA (mRNA) and exercise functions such as chromatin modifications and transcriptional regulation in nucleus, and also implied in post-transcriptional regulation in cytoplasm [10]. Besides, some of the lincRNAs have been confirmed to be the key regulators and biomarkers in the development of mammary gland, like, nuclear-enriched abundant transcript 1 (NEAT1), pregnancy induced noncoding RNA (PINC) and zinc finger NFX1-Type containing 1 (znfx1) antisense RNA 1(ZFAS1), etc. [11,12,13]. However, those lincRNAs, which might be involved in mammary development under different nutrient supply in pre-weaning period of Holstein calves are yet to be known [14,15,16,17].In this study, we identified the lincRNAs in MFP and PAR of pre-weaning Holstein heifers, under enhanced (EH) and restricted (R) plane of nutrition [18]. A total of 534 transcripts originating from 434 gene loci were identified as putative lincRNAs. The function of these putative lincRNAs’ was predicted by analyzing neighboring genes and significantly correlated differentially expressed genes (DEGs) in MFP or PAR under EH and R nutrition supply. The present study broadens the knowledge of lincRNA annotation in bovine as well as facilitates future research about the mammary gland development.2. Materials and Methods2.1. Experiment Design and Library ConstructionThe experiment was previously published by Vailati-Riboni et al. [18]. Briefly, 12 Holstein heifer calves (6.0 ± 2 d old, 39.0 ± 4.4 kg) were divided into two groups under the same forage and feeding management conditions. Feeding was started at the end of 4th week and both of treatments were reduced to 50% at 8th week, to induce weaning. During the trials, both milk replacers were fed in two equal portions twice daily at 06:00 and 17:00 h and calves were provided with drinking water supply. Total RNAs was extracted from the MFP and PAR after the calves were euthanized and their whole mammary glands were removed and dissected. The RNA-seq cDNA libraries were constructed using the Illumina TruSeq Stranded mRNA Sample Prep kit. The single-end read library construction following the manufacturer’s instructions with mRNA enrichment.2.2. DatabasesA total of 22 single-read RNA-seq data were downloaded from NCBI Sequence Read Archive (SRA) database. The Bos taurus UMD3.1.1 reference genome FASTA file and the Gene Transfer Format (GTF) file were downloaded from the ensembl website (ftp://ftp.ensembl.org/pub/release-98/fasta/bos_taurus/dna/ (accessed on 1 June 2014)) and (ftp://ftp.ensembl.org/pub/release-98/gtf/bos_taurus/ (accessed on 1 June 2014)). The UniRef90 (UniProt Reference Clusters) database was downloaded from the UniProt website (http://www.ebi.ac.uk/uniprot/database/download.html (accessed on 1 January 2019)). Moreover, non-redundant reference sequence (RefSeq) NR data was downloaded from (ftp://ftp.ncbi.nih.gov/blast/db/ (accessed on 1 January 2016)). The Homo sapiens and Mus musculus over.chain file (converting genome coordinates intermediate files) downloaded from (https://hgdownload.soe.ucsc.edu/goldenPath/bosTau8/liftOver/ (accessed on 1 October 2014)). The bed file downloaded from (http://asia.ensembl.org/biomart/martview/ (accessed on 1 June 2014)).2.3. Alignment and Assembly of RNA-Seq Data Reads were aligned to Bos taurus reference genome (UMD3.1.1) by Hisat2 (version2.1.0, Iowa State University, Ames, IA, USA) with default parameters [19]. Mapped reads were assembled and 22 assembled transcripts files (GTF format) of four groups were then merged into a non-redundant transcriptome by StringTie (version 1.3.5, Johns Hopkins University, Baltimore, MD, USA) [20].2.4. Identification and Characterization of Putative LincRNAsLincRNAs are the intergenic transcript which have been defined as transcribed non-coding RNAs ≥ 200 nucleotides. Based on this, our pipeline to identify lincRNAs has been shown in (Figure 1a).We used non-redundant transcriptome to identify lincRNAs: (1) only the “u’’ class as candidate linRNAs by gffcompare (version 0.10.6, Johns Hopkins University, Baltimore, MD, USA), which represented the unknown, intergenic transcripts, were retained [21]; (2) transcripts were selected to do further analysis (having exon ≥ 2 and length ≥ 200 bp); (3) We used coding potential calculation (CPC) tool (version -0.9-r2, Tsinghua University, Beijing, China) to calculate the coding potential of transcripts in both strands, CPC < 0 were retained [22]; (4) we used command “transeq” and “hmmerscan” of HMMER (version 3.2.1, HHMI Janelia Fam Research Campus, Ashburn, VA, USA) tool to translated the retained transcript sequence into six possible protein sequences, which had a significant hit in the Pfam database (E-value < 1 × 10−5) were removed [23]; (5) we compared retained transcript sequences with NCBINR and UniRef90 database by BLAST(version v2.6.0 +, National Center for Biotechnology Information, Bethesda, MA, USA), with similarity to known proteins (E-value < 1 × 10−5) were removed [24]; (6) those transcripts, having fragments per kilobase of transcript per million mapped reads (FPKM) score more than 0.1 at least one sample were retained. In addition, we found that 234 out of 534 putative lincRNA sequences were highly similar to the known lncRNA by comparing the NONCODE database (http://www.noncode.org (accessed on 1 January 2018)) [25]. Kolomogorv–Smirnov (KS) test was used for the characterization of putative lincRNAs and protein-coding genes. Furthermore, UCSC website liftOver tools (http://genome.ucsc.edu/cgi-bin/hgLiftOver (accessed on 2 July 2015)) and the over.chain file were used to compare evolutionary conservation of putative lincRNAs and protein-coding genes.2.5. Express Analysis mRNA and Putative LincRNAs in MFP and PARGene expression levels were estimated based on read counts through feature count (version 1.6.4) software [26]. R package DEseq2 was used to conduct differential expression tests between the two groups. And it was considered differentially expressed, if the value of |log2(Fold-change)| ≥ 1 and padj ≤ 0.05 [27]. In addition, unpaired t-test was used to identify differentially expressed lincRNAs (DELs).2.6. Neighboring Gene Identification and Correlation Analysis of DELs of MFP and PARWe identified the neighboring gene (<100 kb) of putative lincRNAs by Bedtools (version 2.29.0, University of Virginia School of Medicine, Charlottesville, VA, USA) [28]. Pearson correlation coefficient as applied to calculate the correlation between DELs and neighboring gene or DEGs. In addition, only putative lincRNAs’ function was inferred according to the pathway analysis result of the genes, which were adjacent to these DELs or significantly correlated with DELs. 2.7. GO Ontology and Pathway AnalysisIn order to predict these putative lincRNAs’ function, the KOBAS (version 3.0) (http://kobas.cbi.pku.edu.cn/kobas3/?t=1 (accessed on 5 July 2011)) was used in order to conduct the pathway analysis [29]. p-value of pathway and enrichGO less than 0.05 were considered statistically significant.2.8. Quantification of LincRNAs through QRT-PCRPCR primers for three randomly selected lincRNAs were designed by the oligo7 program (Supplement Table S1). Total RNA was extracted from each target tissue of bovine by using TRIzol reagent (Invitrogen, California, CA, USA) as per the instruction of manufacturer. The Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH) served as the endogenous control gene. QRT-PCR was performed with SYBR® premix Ex Taq II (Tli RNaseH Plus) (2×) (TAKARA, Kyoto, Japan) to assess the expression level of these three lincRNAs and the results were calculated using the 2−ΔΔCt [30].3. Results3.1. Identification and Characterization of the Putative LincRNAsA total of 22 RNA-seq data were collected from Gene Expression Omnibus (GEO) database. These data represent all the transcripts of MFP and PAR tissues from pre-weaning Holstein heifer calves, which were treated with restricted milk replacer (R, 0.45 kg/day, including 20% crude protein, 20% fat), and enhanced milk replacer (EH, 1.13 kg/day, including 28% crude protein, 25% fat) to identify lincRNAs in these tissues related to energy supply [18]. By using HISAT2, approximately 579.8 of 600.1 million clean reads were mapped on the Bos taurus reference genome (UMD3.1.1) (Supplement Table S2). Then 97,541 non-redundant transcripts were reconstructed for each sample by merged command of the stringtie software. The results of gffcompare showed that 9131 transcripts were intergenic transcripts. Finally, 534 transcripts originated from 434 gene loci were identified as the putative lincRNA based on their protein coding abilities (Figure 1a) (Supplement Table S3). By comparing with the NONCODEv5_cow database (http://www.noncode.org (accessed on 6 January 2018)), we found that 234 out of 534 transcripts have high similarity with known lincRNAs (Supplement Table S4). The other 300 transcripts were considered as novel lincRNAs (Figure 1b). More than 30 lincRNAs were enriched in BTA 3 (Bos taurus autosome 3), followed by BTA 5, 2, 10 and 11 (Figure 1c).There were many differences between lincRNAs and protein-coding genes, including different exon number, length of transcript, expression level and evolutionary conservation, etc. [31] to characterize putative lincRNAs, these were compared with those protein-coding transcripts, which have been annotated in Ensembl database. The characteristics of protein-coding genes and putative lincRNAs were revealed by Kolomogorv–Smirnov (KS) test. These results showed that the average transcript length of these putative lincRNAs was significantly shorter when compared with the protein-coding genes (mean value 967 bp vs. 2058 bp, respectively; p-value < 2.2 × 10−16). (Figure 1d). Meanwhile, we found that the average exon number of the putative lincRNAs was less than protein-coding genes (mean value 2.4 vs. 9, respectively; p-value < 2.2 × 10−16) (Figure 1e). Most of the putative lincRNAs had two exons. Fragments per kilobase of transcript per million (FPKM) analysis results showed that the expression level of putative lincRNAs was lower than protein-coding gene (mean value 0.77 vs. 31.57, respectively; p-value = 9.508 × 10−13) (Figure 1f). After converting genome coordinates, we found that 3.7% and 0.56% of exonic regions of putative lincRNAs in Bos taurus have orthologous regions in Homo sapiens and Mus musculus. And, 10.6% and 5.6% of exonic regions of protein-coding genes in Bos taurus have orthologous regions in Homo sapiens and Mus musculus. It can be implied that lincRNAs showed lower evolutionary conservation than protein-coding transcripts both in Homo sapiens and Mus musculus (Figure 1g). Taken together, these putative lincRNAs showed shorter length, less exon number, lower expression and evolutionary conservation than protein-coding transcripts.Previous studies showed that lincRNAs were remarkably tissue-specific than mRNA [32]. To confirm these identified lincRNAs, we randomly picked four novel lincRNAs to compare the expression level in mammary gland and other tissues by using the quantitative reverse transcription polymerase chain reaction (QRT-PCR). The results showed that these lincRNAs were highly expressed in mammary gland (Figure 2a–d).3.2. Expression Analysis and Functional Prediction of DELs in MFPA total of 79 DELs (61 upregulated, and 18 downregulated; EH vs R) were identified in response to different energy levels in MFP by using the unpaired t-test (Figure 3a). LincRNAs have been found to act as cis-element to participate in the transcriptional regulation of neighboring genes (<100 KB) [33,34]. According to the position of the genome, 261 genes (including 217 protein-coding genes) were found near those DELs (Supplement Table S4). These genes were involved in glucose metabolism and DNA repair signaling pathway (Figure 3b, upper yellow panel). Three of these nearby genes were also identified as DEGs in MFP (Table 1).Besides, lncRNA could also act as trans-elements to regulate distant genes [35,36]. Those genes having similar co-expression patterns could be used to predict the putative lincRNA function. Therefore, we calculated the correlation between DELs and DEGs in MFP by using the Pearson correlation coefficient method (Supplement Table S5). A total of 488 DEGs (including 15 transcription factors) were found to be negatively or positively correlated with these 79 differentially expressed lincRNAs (p-value < 0.05). These DEGs were enriched in metabolism and signal transmission-related signaling pathways (Figure 3b, lower blue panel). Among them, peroxisome proliferators-activated receptors (PPARs) signaling pathway having the highest log2(p-value). Further analysis matched these DEGs with DELs to a total of 11,536 pairs. For example, TCONS_00016005 was positively correlated with apolipoprotein C3 (APOC3) and angiopoietin-related protein 4 (ANGPTL4), which were both enriched in PPAR signaling pathway and cholesterol metabolism pathway (Figure 3c).3.3. Expression Analysis and Functional Prediction of DELs in PARBy using the same method, 86 differentially expressed lincRNAs (54 upregulated, 32 downregulated; EH vs R) in mammary parenchyma (PAR) were identified (Figure 4a). There were 281 genes (including 256 protein-coding genes) near these DELs (Supplement Table S4), including 19 DEGs (Table 1). These neighboring genes were mainly enriched in various metabolism pathways (Figure 4b, upper yellow panel). A total of 1453 DEGs (including 92 transcription factors) were found to be negatively or positively correlated with these 86 differentially expressed lincRNAs (Supplement Table S5). These DEGs were enriched in immune, cell proliferation, cell cycle and signal transduction pathways (Figure 4b, lower blue panel). Further analysis matched these DEGs with DELs to a total of 47,099 pairs. For example, TCONS_0062567 and TCONS_0091815 were positively and TCONS_0066624 was negatively correlated with Insulin Like Growth Factor I (IGF-I), which were enriched in MAPK signaling pathway, PI3K-Akt signaling pathway, Focal adhesion, Ras signaling pathway and epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor resistance pathway and involved in the mammary gland development.4. DiscussionSome studies showed that feeding heifer claves with restricted milk replacer was the right strategy to save rearing cost and increase milk production [37,38]. With the development of research, it has been confirmed that increased nutrient supply would benefit the development of mammary gland ultimately leading towards increased milk yield [39,40,41]. Furthermore, nutrient supply can affect the total protein, total DNA, total fat, protein and fat concentration changes in MFP and PAR. In order to reveal this molecular mechanism, a number of transcriptomic studies have been reported [42,43]. However, it is still largely unknown. Over the last decade, more and more studies showed that lncRNAs were the key regulators in various biological processes, and most of them were lincRNAs [10]. Although, PINC, NEAT1 and ZFAS1 were reported to be involved in mammary epithelial cell proliferation. Which lincRNAs are involved in the mammary gland development under different feeding regimes remained largely unknown. In this study, we have identified 534 putative lincRNAs from 22 samples from two tissues (MFP and PAR) and two treatments (EH, 1.13 kg/day, including 28% crude protein, 25% fat; R, 0.45 kg/day, including 20% crude protein, 20% fat), by using published high throughput RNA-seq data [18]. Results of characterization for these putative lincRNAs in our study are consistent with previous reports [44,45] which were also confirmed by QRT-PCR and it was concluded that most of the putative lincRNAs were highly expressed in mammary tissues.In MFP tissue, those DELs neighboring genes and significantly correlated DEGs were enriched in various metabolism and signal transmission processes. Such as, PPAR signaling pathway, Calcium signaling pathway and Cytokine-cytokine receptor interaction, etc. Especially, the PPARs signaling pathway played a key function in preadipocyte proliferation [38,39]. In addition, we also found that these neighboring genes and significantly correlated DEGs were associated with lipid metabolism, which played a functional role in adipocyte tissue like, angiopoietin-related protein 4 (ANGPTL4), apolipoprotein C3 (APOC3), etc. [46,47]. And, these DEGs were downregulated in MFP under the restricted supply. Therefore, we speculated that these DELs might regulate neighboring genes and are significantly correlated with DEGs to promote MFP development under the enhanced nutrients supply. In PAR tissue, those DELs’ neighboring genes and significantly correlated DEGs were enriched in metabolism-related and cell proliferation-related pathways. Such as, PI3K-Akt signaling pathway, JAK-STAT signaling pathway and Ras signaling pathway, etc. Moreover, TCONS_0062567 was negatively correlated with Insulin Like Growth Factor I (IGF-I), while, TCONS_0091815 and TCONS_0066624 were positively correlated with IGF-I, which were considered as the important regulators in the lactation process [48]. Some other significantly correlated DEGs play indispensable role in mammary gland development like, Epidermal Growth Factor Receptor (EGFR) and Cyclin D2 (CCND2), etc. [49,50]. And, these DEGs were downregulated in PAR under the R supply suggesting that these DELs may promote the PAR development under the EH supply.Mammary gland development is a complex process, including various tissues and cell types [1,51,52]. Both MFP and PAR play important roles in various stages of mammary gland development. In this study, only lincRNAs with ploy A tail were identified for this biological process. Functional prediction was accomplished for DELs highly correlated DEGs. However, further research can be conducted to explore specific functions and regulated mechanisms for these lincRNAs.5. ConclusionsIn summary, we identified 79 and 86 DELs in MFP and PAR of pre-weaning heifer calves under the enhanced and restricted nutrient supply. These putative lincRNAs may influence metabolism, cell proliferation and tissue interaction in MFP and PAR by positive or negative regulation to promote the mammary gland development and tissue communication under the enhanced nutrition supply. In this study, only those putative lincRNAs were studied which were highly correlated with these key protein-coding genes in mammary gland development. Moreover, it is suggested to assess the functional analysis of these putative lincRNAs by experiment.

animals : an open access journal from mdpi

10.3390/ani13091444

PMC10177455

In Switzerland, reindeer are exclusively kept in captivity. The aim of the present work was to evaluate and summarize management and feeding practices, and to examine the prevalence of endoparasite infections in Swiss reindeer. A total of 67 animals from eight different farms and zoos were evaluated. On two visits to the farms and zoos, a standardized questionnaire was completed by the breeders and/or animal husbandry managers, the animals were weighed, and fecal samples were collected. All reindeer were fed roughage ad libitum and supplementary feed for reindeer or other browsers, with different compositions in each herd. The prevalence of gastrointestinal strongyles was 68.6%, with reindeer in zoos having a lower prevalence than reindeer from private farms. This study presents an overview on husbandry, feeding, and endoparasite prevalence in captive reindeer in Switzerland and provides basic data for breeders and veterinarians dealing with this deer species.

The aim of the present work was to provide an overview of management and feeding practices, and the prevalence of endoparasite infections in captive Swiss reindeer. On two visits to eight farms or zoos, a standardized questionnaire was completed. A total of 67 reindeer were weighed, and fecal samples were collected. The primary management concerns voiced by owners/managers were feeding and successful breeding. All reindeer were fed roughage ad libitum and supplementary feed for reindeer or other browsers, with different compositions in each herd. Males over two years of age weighed from 60 kg up to 127.5 kg, whereas females had a body weight from 53.5 kg to 86.5 kg. The prevalence of gastrointestinal strongyles was 68.6% (46/67), with reindeer in zoos having a lower prevalence (36%; 9/25) than reindeer from private farms (88%; 37/42). Capillaria sp., Strongyloides sp., and Trichuris sp. were detected in lower prevalences (<24%) and were also more frequent in private farms. Intestinal protozoa, as well as fluke and tapeworms, were not detected in any herd. This study provides an overview on husbandry, feeding, and endoparasite prevalence in reindeer in Switzerland and should be of help for breeders and veterinarians dealing with this animal species.

1. IntroductionReindeer or caribou (Artiodactyla, Cervidae, Rangifer tarandus) are deer species within the genus Rangifer [1]. Reindeer are the only semi [2] or fully [3] domesticated deer species on a large scale worldwide, and the only deer species with a circumpolar distribution [3].Keeping captive deer is a relatively new branch of Swiss agriculture, and only small numbers of reindeer are privately kept [4]. There is no individual registration requirement for reindeer in Switzerland. The number of reindeer must only be reported to the respective canton every two years. In Spring 2021, there were a total of 78 reindeer in Switzerland, spread across nine herds, including four herds belonging to zoos. Since 1992, captive deer have been classified as farm animals. Those who keep deer in captivity must acquire a cantonal husbandry permit and meet technical and professional qualification standards according to Swiss law. In Switzerland, data on the management and feeding of reindeer in captivity are scarce. Reindeer are browsers and graze on leafy plants and short leafy grasses but have a limited ability to digest long fibers rich in cellulose [2]. Depending on the climate and season, their daily food intake is approximately 2–3% of their body weight [2]. A combination of leafy forbs, seasonal green forage, or commercially available dried lichen and a pelleted feed formulated for browsers with a beet pulp base has proven adequate for reindeer in zoos [5]. In Fennoscandia, commercial reindeer feed contains different cereal grains; however, they are mainly intended for winter-feeding and therefore cover the nutritional needs of the animals during that time of the year. In North America, barley and oats are primarily used as cereal grains in reindeer pellets. The nutritional composition of feed rations for reindeer is highly variable, as some may also be produced for special purposes and adapted needs, depending on the season or category of reindeer [2].Like other ruminants, reindeer are hosts to a variety of endoparasites that may affect their health and productivity [6]. In their natural habitat, the prevalence of subclinical, low-intensity mixed gastrointestinal parasitic infections is high [2,7,8,9]. Common gastrointestinal parasites affecting reindeer are Eimeria spp., gastrointestinal nematodes such as strongylids, Capillaria spp., Trichuris spp., Skrjabinema spp., and cestodes such as Moniezia spp., among others [10]. There are limited data on the prevalence of parasitic infections in captive reindeer, especially for reindeer in moderate climate zones. Veterinarians regularly educate clients and producers about appropriate nutrition, animal health, and preventative measures. Information on the health management of captive reindeer and the local prevalence of various endoparasites is valuable. Especially given the zoonotic potential of some animal endoparasites, such information is also important for animal caretakers, veterinarians, and human health care professionals. This is the first study on reindeer husbandry in Switzerland. 2. Materials and MethodsIn Spring 2021, all nine reindeer keepers (i.e., five private owners and four zoological institutions) in Switzerland were contacted to participate in this study, and eight of them accepted the invitation; one farmer declined it due to time constraints.The participating owners or zoo veterinarians were asked to complete a questionnaire with 135 questions (Supplementary Material). The questionnaire included a general part with questions about the farm structure and management, and a specific part with questions about the owner’s personal interest in keeping reindeer, feeding practices, pasture use, and health status of the herd (including parasitological status, medical history including diagnosed diseases, and losses). In addition, individual coproparasitological analyses of all reindeer in the study were carried out between 5 September 2021 and 30 September 2021. To obtain fecal samples, animals were physically restrained for less than one minute, and rectal fecal samples were taken. The same approach was undertaken for each visit in six herds. In only two zoos were fresh fecal samples collected from the ground right after defecation. Each animal was weighed on a portable scale (TRU-TEST EziWeigh 1) and body condition score (BCS), ranging from one (low) to four (high) was assessed according to Laaksonen and Nieminen (2005) [11]. For animals that could not be captured, BCS was assessed visually, and they were not weighed. Fecal samples were refrigerated, transported to the laboratory, and examined within a week after collection and always by the same person. All fecal samples (n = 67) were examined using a combined sedimentation/flotation technique with a saturated zinc chloride solution (specific gravity 1.45) for cestode and nematode eggs, and coccidia using at least five grams of feces, and also by a sedimentation technique for trematode eggs [12]. If nematode eggs were found by the sedimentation/flotation method, a quantitative McMaster technique with a detection limit of 50 EpG (=eggs per gram feces) was performed (n = 44) [12]. In five of the sixty-seven samples, there was not enough material left to perform a McMaster. Fecal samples from which sufficient material was available (n = 53) were additionally examined for lungworms using the Baermann–Wetzel technique [12]. In one herd, a larval culture was performed. Parasite identification was performed based on morphological criteria according to Deplazes et al., 2021; van Wyck and Mayhew 2013; and Tryland et al., 2018 [2,12,13].Aliquots for 65 of 67 fecal samples were stored at −20 °C for coproantigen detection using commercial kits. In two samples, there were not enough material left to include them in this procedure. Samples were thawed for 24 h, and then analyzed for Giardia duodenalis coproantigens using a commercial ELISA (MegaELISA®GIARDIA, MEGACOR Veterinary Diagnostics, Austria), and for Cryptosporidium parvum antigens using a rapid immunochromatographic test (FASTest®CRYTO, MEGACOR Veterinary Diagnostics, Austria). Microsoft Excel 365 for Windows (Microsoft Corporation, Redmond, WA, USA) was used to analyze data and to create the results. Exact binomial 95% confidence intervals were calculated with the Sample Size Calculators by UCSF Clinical and Translational Science Institute [14]. The questionnaire was analyzed by its categories. Answers were summarized and associated with the herd and its individuals. No statistical tests were performed. Prevalence was considered as the number of positive samples/total number of samples. Data subsets were created based on the age of the animals, their origin: zoo or private farm, their sex, and their parasite infections.This study was carried out in accordance with the Swiss animal welfare legislation and was authorized by the involved cantonal animal welfare committees (approval number BE 11/2021). All participating reindeer keepers gave written consent for all examinations and for the publication of the results thereof. All collected data, as well as the subsequent evaluations, were processed confidentially.3. Results3.1. Reindeer PopulationIn autumn 2021, a total of sixty-seven reindeer were distributed on eight participating herds in Switzerland, consisting of seven Eurasian tundra reindeer (Rangifer tarandus tarandus) and one Eurasian wild boreal forest reindeer (Rangifer tarandus fennicus) herds. The average herd size was 8.15, the smallest with two animals and the biggest with nineteen. Four of the eight animal stocks were located in zoos and the rest were private keepings. The location of these herds ranged in altitude from 300 to 1300 m above sea level. The participating zoos have kept reindeer since 1935 and the private breeders since 2009 to 2018, respectively. The age and sex distribution of the 67 Swiss reindeer is shown in Table 1. The whole population consisted of 41 females. Half of the twenty-six male animals were infertile—nine/thirteen were castrated and four/thirteen had a temporary chemical castration. Chemical castration was achieved with a contraceptive implant containing 4.7 mg deslorelin per animal (Suprelorin® ad us. vet., Virbac AG, 8152 Opfikon, Switzerland). Castrated animals only occurred in private keepings. All zoos had one adult intact male within their herd and one zoo had a male calf. All other intact males belonged to private herds (eight/thirteen).Most animals (22/67) of the study population were between one and four years of age. Seven out of thirteen intact males were between one and four years old. Most of the males older than four years were castrated (seven/nine). There were nine calves (younger than one year). In autumn 2020, there were twenty-seven females and one male of fertilizing age. A third of the productive population reproduced in between autumn 2020 and autumn 2021.3.2. Purpose of KeepingThe main reason for keeping reindeer were educational purposes (four/eight); these were all located in zoos. Other purposes for keeping were animal rental for Christmas markets, walks, photoshoots; sponsorships; or as a pure leisure pursuit. The facility with the boreal forest reindeer also aims at breeding for the conservation of the species (EAZA Ex Situ Program). None of the animal owners lived solely on the income from keeping or breeding reindeer.3.3. HusbandryAll animals were kept in a permanent enclosure throughout the year. The height of the fence varied from 1.2 m up to 2.8 m. All the private keepings and one zoo had fencing of at least 2 m in height. All the herds had access to at least one stable with a concrete floor. In winter, seven of eight herds had parts of their barn floor bedded with straw or sawdust. One herd’s stable floor was lined with coffee bean shells covered with straw all year round. Four of eight herds’ enclosures comprised pasture, while the remaining four had ground with stones, gravel, or marl; three had wood chips and two a forest floor. All the herd enclosures had some hard floor with concrete at least in the barn; one herd additionally had pavement. Enclosures were cleaned once a day in six herds, twice a day in one herd, and every second day in one herd. In half of the herds (four/eight), the whole enclosure was cleaned, in three herds just the most frequented sites, and in one only the stable was cleaned. The artificial water facilities comprised fountains (four/eight), basins or buckets (two/eight), and automatic or self-drinkers (two/eight) in or near the barn, and a few farms had a stream or creek in the enclosure (three/eight). In winter, one herd was subjected to complete snow cover over the whole winter, three herds had occasional snow in winter, and four herds had snow only sporadically.3.4. FeedingThe feeding management in each participating herd is shown in Table 2. Six of the eight herds were fed twice a day with supplementary feed. One and occasionally a second herd were fed three times a day. Only one herd was fed once a day. Roughage was fed ad libitum in all herds, but only three herd keepers could indicate the amount fed per day, so that the consumed amount could be estimated. Roughage was primarily hay of the second cut (75%), coarse hay (25%), haylage (25%), and hay chaff (13%). The coarse hay was fed together with the haylage. Five herds had the availability of grass in at least one season of the year. The feed residues were estimated by the owners or animal keepers. The estimations vary from 10% up to 50% feed residues. Most of the owners/keepers added that it varies depending on the season. Pelleted feed was offered in all the herds daily (see Table 2). All herds received pelleted supplementary feed formulated for reindeer or browsers (e.g., Browser 3699, Granovit, Kaiseraugst, Switzerland). The amount of supplementary feed per animal varied from 1.2 to 5.8 kg/d.3.5. Animal Health and ManagementThe challenges in reindeer husbandry mentioned by reindeer owners/keepers more than once were feeding or changes in feeding (seven/eight); keeping males in rut, breeding season (four/eight); the management of calves (three/eight); and the management of endoparasites (three/eight). Management challenges that were mentioned once included handling wild animals; the enclosure being a stagnant pasture; the keeping of the animals within the enclosure and suitable ground; the ability of the animal to disguise clinical signs; the lack of knowledge about how a healthy animal presents; and the lack of veterinary knowledge.In all of the privately kept herds, the reindeer were used to being haltered and were occasionally walked. In two zoos, animals were not handled, including those from which fecal samples were collected from the ground. In five of the eight herds, hoof care was performed as needed when hooves were too long, while in three herds, hoof care was not performed except when animals presented lame.The most mentioned health problems in the questionnaire were diarrhea, lameness, and young animal problems (weak calves, sudden death of calves). Other health problems mentioned were the acute death of animals, endoparasites including two cases of Haemonchus contortus infection, ectoparasites, intestinal obstruction, alopecia, dental problems, babesiosis, pneumonia, enterotoxaemia caused by Clostridium perfringens, antler injury, emaciation, pharyngeal congestion, myiasis of the antlers, eye inflammation, and infertility. All these health problems did not necessarily result in the death of the animals.Most diseases causing death were confirmed by a veterinarian, but not all by a pathological examination at necropsy. The most mentioned known death causes in the questionnaire were enterotoxemia, intestinal obstruction, and euthanasia because of brachygnathia inferior or because of trauma due to forkel injury (injury by antlers).The mean body weight and range (min–max) in different sex and age groups are shown in Table 3. Fifty-six animals were weighed in six herds. Among the male calves younger than one year, the lightest calf weighed 7.5 kg and was born only two weeks before the weighing took place. In each group, weights varied by at least 20 kg between the highest and lowest weight. In the group of adult males (over two years of age), the lightest animal was less than half the weight of the heaviest. The same applies to animals younger than one year and in females that were one year old.3.6. Parasitological ExaminationsGastrointestinal strongyles (GIS) were the most frequently recorded parasites both at the animal and herd levels (>50%). The prevalence of herds with Trichuris sp. or Strongyloides sp. infections was the same (two out of nine herds; 22.2%), but the prevalence of infection at the animal level was higher for Trichuris sp. (seven/sixty-seven; 10.4%) than for Strongyloides sp. (two/sixty-seven; 3%). One third of the herds had animals with Capillaria sp. infections, and the prevalence at the animal level was 23.9% (16/67 animals). Neither Fasciola hepatica or Dicrocoelium dendriticum eggs, nor larvae of any lungworm species, were detected by the sedimentation and Baermann–Wetzel techniques, respectively, in any of the 67 fecal samples. Neither Eimeria spp. oocysts nor tapeworm eggs were detected by the sedimentation/flotation technique. All the samples tested negative by ELISA for Giardia duodenalis and by the FASTest® for Cryptosporidium parvum coproantigens.The prevalence and the 95% confidence intervals of the detected endoparasite species in Swiss reindeer at both the animal and at the herd levels are shown in Table 4. This table does not consider mixed infections. The prevalence of parasite infections in the entire study population, including samples in which no parasite eggs were found and samples with mixed infections, is shown in Figure 1. It shows that 27% of the study population (n = 18) did not shed parasite stages or did so in such low numbers that they could not be detected. The largest proportion of the study population had a single evidenced infection with GIS (48%; n = 35). All the mixed infections included infection with GIS besides other parasite species. Most of them had an additional infection with Capillaria sp. (9%). The second most common mixed infection was a triple infection with GIS, Capillaria sp., and Trichuris sp. (6%). This was followed by a dual infection with GIS and Trichuris sp. (4%). Only one animal had a mixed infection with GIS and Strongyloides sp. (1%). In one herd, a larval culture was performed and 92% of the third-stage larvae found were identified as Haemonchus contortus. When the population was divided into two housing groups (zoos vs. private keepings), the prevalence of nematode infections was higher in privately kept animals than in zoos for all the parasites listed in Table 4 (Figure 2). For GIS infections, the confidence intervals did not overlap between zoos and private farms. However, the confidence interval overlapped for Trichuris sp., Capillaria sp., and Strongyloides sp. Figure 2 does not include mixed infections. The results of the McMaster technique, which was performed with forty-four samples derived from eight herds, are shown in Table 5. The highest GIS burden was found in a privately-owned reindeer with 13,200 EpG. The highest GIS burden found in zoo animals was 50 EpG. In five of the seven animals in which Trichuris sp. eggs were detected, a McMaster was performed, revealing egg burdens of 100 (Trichuris) EpG in one animal and <50 EpG in the four remaining reindeer. In addition, a McMaster was also performed for nine of the twelve animals infected with Capillaria sp., and egg counts between 50 and 400 EpG in five of the animals and <50 EpG in the remaining ones were obtained. Anthelmintics were used in every herd. All the zoos used the active substance fenbendazol in pellets or in a powdered form. The private herds used active substances such as fenbendazole, moxidectin, doramectin, monepantel, or a combination of levamisole and triclabendazole. In one herd, toltrazuril was administered once a year. 4. DiscussionThe aim of this study was to investigate the farm structure and management, feeding practices, endoparasite prevalence, and the general health situation of reindeer in Switzerland. The actual total Swiss reindeer population is small with 78 animals. Compared with other captive wild hoof stock, it corresponds to 0.6% of the fallow and red deer population (13,237 fallow and red deer) in Switzerland (https://www.sbv-usp.ch/fileadmin/sbvuspch/04_Medien/Agristat_aktuell/2021/Aktuell_AGRISTAT_2021-09.pdf) (accessed on 28 March 2023). The number of reindeer herds in Switzerland is 1.45% of registered captive deer farms (nine of six hundred and eighty-nine farms) with an average herd size of 8.15 animals and a range of two to nineteen animals. Swiss reindeer herds are smaller than captive Swiss deer herds. Although reindeer have a life expectancy of up to 20 years, the current population consisted mainly of animals less than 10 years (76%), and only one animal was over 15 years old. Most of the older animals were imported from Germany or Sweden, which provided age information at the time of import. The sex distribution of the Swiss reindeer population in this study is similar to that of Swiss South American camelids reported by Hengrave Burri et al. (2005) [15]. However, there are slightly fewer intact male reindeer (10%) and slightly more castrated males (13%), with an additional 6% of reindeer temporarily infertile due to chemical castration. With nine intact breeding males, there are more males than herds. However, the risk of inbreeding is high with a total population below 100 animals. This can only be solved by new imports. The similarity in sex distribution between South American camelids [15] and reindeer could be due to their similar use in private husbandry. In addition to walks, reindeer are also used for photo shoots and the rental of the animals to Christmas markets; in advertising, etc., which implies a close relationship between humans and animals; and as trained reindeer for handling. During the Christmas season, male reindeers are in rut, which makes them difficult to handle. This explains why half of the male population is castrated or temporarily infertile.The substrate in the enclosures varied. For captive deer, a partial hard floor is recommended for good hoof abrasion, e.g., gravel or stones in the most frequented areas. However, several breeders with this type of floor have had lameness in their reindeer herd, citing sprains or sole abscesses due to entered stone. The feeding of the reindeer varied among herds. The dietary composition changed with the vegetation period and the seasonal availability of the feed components, whereas the composition of supplementary feed at an individual herd level largely depended on ingredient availability, preferences, and the convenience of the individual owner. Reindeer are classified as intermediate mixed ruminant feeders [16,17]. In their original habitat, they live either as semi-domesticated reindeer that still roam free on natural pastures most of the time or as wild individuals, so they feed on natural forage and migrate between summer and winter areas [2]. In winter, when feed is scarce, reindeer have the advantage to include a large portion of lichens in their diet compared with other herbivores in their natural environment. Lichens present a valuable energy source due to their high content of easily degradable carbohydrates [18]. However, only two of the participating herds in Switzerland supplemented with lichens daily because lichens either need to be collected by hand in nature or imported from northern countries. Moreover, the lichens represented only a minor dietary component in captive Swiss reindeer. Although the overall dry matter (DM) intake could not be determined in the present study, other research findings indicate that reindeer fed mainly on lichens have a lower total DM intake resulting in a loss of body mass [19,20]. When winters become harsher and an icy snow cover prevents reindeer from reaching the lichen by digging, northern reindeer owners are forced to increasingly provide their animals with supplementary feed and hay [2]. In summer, they need to compensate for the restricted nutrient intake from winter with rapid growth and the replenishment of body reserves [2]. The natural extreme environmental changes between winter and summer require an immense seasonal adaptability of the animals in their natural habitat. However, the results of this study show that their seasonal diet in Swiss captivity does not change substantially throughout the year, except for the inclusion of pasture and browse. Roughage was fed throughout the entire year. In herds with pasture access in spring and summer, the available grass of the pasture was also added to the diet during these seasons. Most of the herds (seven/eight) had access to high quality hay (second or third cut) or similar energy and nutrient rich hay as a roughage source. Reindeer are especially delicate as their digestive system cannot handle large quantities of fibers and therefore relies on roughage with a higher digestibility. Excessive crude fiber leads to the accumulation and gradual stretching of the rumen, but it remains undigested [2]. The appetite stays, but the lack of energy causes emaciation and, eventually, the death of the animal, even though the rumen is full of grass or hay [2]. However, the condition can be avoided by supplying a more easily digestible diet [2]; this is a reason why reindeer are fed additional supplementary feed to their roughage.It is noticeable that the estimated remaining feed varied greatly between herds. This is likely caused by the estimation instead of the measurement of the remaining feed by the owners and keepers. However, constant access to roughage is crucial for ruminants, and all owners and keepers stated that roughage was fed ad libitum. The amount and composition of supplementary feed differed between the different Swiss herds. Zoos fed a greater amount of fresh ingredients (1.2–4.6 kg per animal per day), such as vegetables or fruit, than private farms (0–0.3 kg per animal per day). This is probably due to the better availability of vegetables and other perishable foods in zoos that are simultaneously required for other species. Furthermore, the amount of supplementary feed was higher for reindeer in zoos than in private husbandries. In general, there is a risk of over conditioning reindeer in captivity due to inadequate feeding. Reindeer’s basal metabolic rate (BMR) in winter was assumed to be 293 kJ/kg0.75 [21]. Especially in spring and summer, they need an energy rich diet that allows them to replenish body reserves that were mobilized during previous feed shortages. It is noteworthy that almost all herds received supplementary feed more than once a day (Table 2). For a ruminating species, such as reindeer, it is important to reduce the temporal accumulation of carbohydrates in the rumen to avoid rumen acidosis. A more frequent distribution of non-structural carbohydrate rich feed sources, such as the supplementary rations more than twice a day, would be even more desirable. In addition, most of the reported cases of intestinal obstipation occurred in one herd (herd number two) that only provided the supplementary feed once a day. The only other documented case of intestinal obstipation was because of bezoars in herd three (unpublished data). Even though the data are small, these findings indicate that it is important to spread the daily ration of supplementary feed on multiple feedings during the day, especially as pelleted feed can swell when in contact with fluids. The large amount of pelleted feed in the supplementary ration combined with stress could be a risk factor for intestinal obstructions.Feeding was considered the most challenging aspect for owners, keepers, and zoo veterinarians when it came to dietary composition, amount, and frequency. This is illustrated through the varied feeding management approaches between farms. Reindeer must be considered a species whose dietary habits and nutritional needs may not be met through the offered feed sources available in Switzerland. Each farm or zoo chooses a different, individual approach that is based on their own empirical experience. It is important to assess the nutrients and minerals in captive reindeer and compare them with data available for samples collected from wild reindeer; this will help establish feeding recommendations for captive reindeer. Keeping a stag was considered difficult, especially during the rut, by at least 50% of the survey respondents. However, all herds had at least one stag or used to have one, because of breeding purposes. In 2020, of the herds that had a stag for breeding (six/eight), only three herds had live offspring, with a total of nine calves. The management of newborns and calves was also considered challenging, especially for zoos who reported that it is difficult to produce healthy offspring. When offspring is born, it is difficult to keep them alive, because they are sometimes born weak or underweight.Of all the reported deaths, 23/41 were calves younger than six months of age. This suggests difficulties in producing live offspring and having them thrive through the first year of life. In 2020, 40% of the population were females old enough to breed and had a sexually mature stag in their herd, resulting in a total of 27 females able to reproduce. However, only nine calves were alive in autumn 2021 suggesting that one third of females were not able to reproduce successfully. Pregnancy rates in wild populations normally exceed 60% and in adult domestic flocks it may approach 100% [22]. Even though the pregnancy rate was not determined in this study, the reproduction rate seems rather small. Body weights varied by sex and age (Table 3). In another study, calf weights ranged from 33 to 45 kg at five to five-and-a-half months of age [21]. In this study, the mean body weight would be similar (34.9 kg for females and 34.7 kg for males) when the two youngest calves, that were born in late summer, are excluded. The body weights of older females ranged from 30.5 kg to 86.5 kg, with females weighing less than 60 kg in all age groups. The optimal criteria for reproduction and calf production cannot be met if the live weight of the females does not exceed 60 kg [23,24]. It is important to further investigate whether the low body weights may have caused the poor calving rate, especially since the animals from two zoos that had problems with reproduction could not be weighed. The body weights of male animals also varied substantially when compared with three-year-old males that weigh between 88 kg and 97 kg [25]. However, the number of animals weighed in this study was small and must be interpreted with caution. For future studies concerning calving rates, we recommend weighing cows prior to the rutting season to then compare with the reproductive success the following year. Furthermore, the heavy burden of parasites should also be considered if fertility is low. Other reasons, such as management factors regarding herd size, herd composition, or bulls, should also be considered. Reindeer can harbor a variety of endoparasites. The prevalence of endoparasite infections in this study population (n = 67) was 73% or 49/67 (Figure 1). Although protozoa such as Cryptosporidium spp., Giardia duodenalis or, or Eimeria spp. may infect reindeer [9], we did not find any of these protozoa in our study. This is in contrast to caribou from Northern Alaska where the overall prevalence of Cryptosporidium oocysts was 6.1% in [26]. On the other hand, in Northern Norway, no Cryptosporidium oocysts were found [6]. Cryptosporidiosis, especially C. parvum infection, is a common diarrheic disease in bovine calves, and the infection can also be transmitted to humans and other animal species. Giardia duodenalis was not detected in any of the samples from this study, which is in agreement with the findings in caribou from Northern Alaska [26]. In contrast, in 2021, a prevalence of 5% of G. duodenalis cysts was detected in reindeer from Northern Norway [6]. Giardia duodenalis has a high prevalence in various species of young animals and it may be a cause of diarrhea [12,27]. Although only G. duodenalis would be expected to be present in reindeer, different Cryptosporidium species have been described in this deer species [28,29]. The commercial tests used in this study were designed to detect coproantigens of G. duodenalis and C. parvum, but it is not known if the used commercial kit would be able to detect coproantigens from other Cryptosporidium species described in reindeer besides C. parvum.Intestinal coccidia species are usually specific to their host species. Eimeria spp. are also frequently found in semi-domesticated reindeer calves, whereas the prevalence in adult wild reindeer is low [9]. Nevertheless, the prevalence of Eimeria spp. was 51% in reindeer calves in 2019 on Fennoscandia with the higher density of reindeer considered a risk factor for shedding oocysts [10]. Interestingly, there were no Eimeria spp. or Isospora spp. found in this study. This was unexpected, especially when considering the higher density of captive animals compared with wild populations. Possible reasons are the small number of calves distributed in only three herds in 2021, or the import of reindeer into Switzerland who were free of intestinal coccidia and, therefore, no infection source for young animals was present in this country.Compared with protozoa, more gastrointestinal nematodes were found in the Swiss reindeer population. The highest prevalence of 68.9% (CI 95%: 56.2–79.4) was identified for GIS. This prevalence is very similar to the prevalence of 75.6% reported in 2019 for reindeer calves in Fennoscandia [9]. An abomasal nematode reported in reindeer is Haemonchus contortus [30]. Haemonchus contortus infections result in blood loss and consequent anemia. Cases of H. contortus infections have been reported in reindeer in Switzerland (C. Luginbühl personal communication). In addition, coprocultures and the differentiation of third-stage larvae were performed on one farm during this study, confirming a heavy burden of H. contortus in all animal age groups. The findings from this farm were also confirmed by the Swiss Consulting and Health Service for Small Ruminants (Niederönz, Bern, Switzerland) (unpublished data). The clinical signs of haemonchosis in the farm were diarrhea, emaciation, pale mucous membranes, and submandibular edema in a late stage. In this same herd, the highest EpG numbers were found. The animal, which was hardest hit by clinical signs had a count of 7100 EpG in our fecal analyses on samples taken one week before the clinical signs were strongest. The herd was treated with levamisol and triclabendazol and the animal fully recovered. A McMaster performed two weeks after deworming showed an almost 100% efficacy of the anthelminthic treatment. Another parasite of ruminants that occurs in the small intestine is Capillaria sp. In this study, the prevalence of Capillaria sp. was 23.9% (CI 95% 14.3–35.9), which was lower than in other studies that found a prevalence of 60% in semi-domesticated reindeer calves, while the frequency was much lower in their adult counterparts [31]. In addition to being a frequent parasite in reindeer calves, Capillaria spp. appear to be a predominant parasite in reindeer zoo enclosures [32]. It is interesting to note that only one of the three Swiss herds with Capillaria sp. infections was a zoo. Two of these herds, including the zoo, had calves in 2021. Although these parasites shed eggs all year-round, they appear to be most prevalent during the cold winter months [9], which may be one reason why prevalence was not higher in Switzerland, where samples were collected in early fall. Capillaria sp. has also not been associated with specific clinical signs in ruminants [33], yet it was documented that severe infection can lead to enteritis and diarrhea [34]. Commonly known as “whipworms”, Trichuridae is another parasite family that colonizes the intestines of many mammalian species. Infection with Trichuris spp. can lead to acute or chronic inflammation in the caecum [9]. Although Trichuris spp. are not commonly observed in reindeer in Europe [35,36], at least five Trichuris spp. have been reported from Russian reindeer herds [37]. The results of this study showed a prevalence of 10.4% (CI 95% 4.3–20.4) in Swiss reindeer. Compared with a study in Fennoscandia, where the prevalence of Trichuris sp. in reindeer calves was only 0.6%, the prevalence in Switzerland seems to be high [9]. In wild moose, infections with these parasites are not considered a threat [35], but they can cause bloody diarrhea, especially in young animals [38]. Furthermore, it is thought that Trichuris spp. may be a component of the Wasting Syndrome Complex, a condition involving chronic diarrhea and the loss of body weight and condition [39,40]. There is also a report of a reindeer’s death, which was infected with “whipworms”, in a Finnish zoo [9]. However, Figure 1 shows that all animals with a Trichuris sp. infection had a co-infection with GIS or GIS and Capillaria sp. Furthermore, the only two herds where animals shed Trichuris sp. eggs were herds that had calves. Five out of the seven Trichuris infections were recorded in calves younger than one year old. The other two animals were one-and-half and eleven-and-a-half years old. It is evident that these two animals with 13,200 EpG and 10,750 EpG excreted the most GIS eggs per gram of feces of the entire study population (Table 5). The difference in prevalence of parasite infection in zoos and in private farms is shown in Figure 2. Why reindeer in zoos seem to be less frequently infected with nematodes is unclear. However, zoos have fewer calves (three/nine) and thus fewer reservoirs, but the number of calves in the entire study population was small, and the only zoo with calves had no infections with GIS, only some infections with Capillaria sp. in its calves. Another reason could be that reindeer in private keepings have more pastures in their enclosure, from which they also feed. These pastures could be an ideal place for the parasites to complete their life cycle and (re-)infect their hosts. Reindeer without pastures may spend less time feeding from the ground, where they also defecate, and therefore have less chance of (re-)infecting themselves. Another argument could be that more areas, sometimes even the entire enclosure, are cleaned more frequently in zoos than in private keepings resulting in a reduced parasite burden. 5. ConclusionsThis study is the first to evaluate captive reindeer husbandry, feeding, and management practices in Switzerland. The findings that GIS are more common in some herds will help reindeer keepers and veterinarians to better plan diagnostic and preventative strategies, especially concerning management practices. However, it would be of value to further investigate this subject over a longer period of time to see how parasite burden changes throughout the year and to focus on management factors to determine which are crucial.

animals : an open access journal from mdpi

10.3390/ani11113141

PMC8614288

During emergencies, people’s decision-making and actions are strongly influenced by their relationship with their animals. In emergency management, a holistic approach is needed which recognises the important interrelationships between animal welfare, human well-being, and the physical and social environment. It is also vital to break down barriers of collaboration between individuals, organisations, and the community. One Welfare, a concept with human–animal-environment interdependencies at its core, provides a framework to achieve this. Successful implementation of a transformative change will require positive strategies to deal with challenges and to ensure that animals are truly integrated into emergency management, not just included as an aside.

Responding to emergencies requires many different individuals and organisations to work well together under extraordinary circumstances. Unfortunately, the management of animal welfare in emergencies remains largely disconnected from emergency management overall. This is due predominately to professional silos and a failure to understand the importance of human–animal-environment (h-a-e) interdependencies. One Welfare (OW) is a concept with these interrelationships at its core. This paper argues that by adopting an OW framework it will be possible to achieve a transdisciplinary approach to emergency management in which all stakeholders acknowledge the importance of the h-a-e interdependencies and work to implement a framework to support this. Acknowledging that such a transformational change will not be easy, this paper proposes several strategies to overcome the challenges and optimise the outcomes for animal welfare emergency management (AWEM). These include legislation and policy changes including h-a-e interface interactions as business as usual, improving knowledge through interprofessional education and training, incorporating One Welfare champions, and recognising the role of animals as vital conduits into communities.

1. IntroductionIn many countries animals are increasingly included in emergency management legislation and policy with specific organisations delegated responsibility for animal welfare in emergencies, including the development of animal inclusive emergency management plans and is termed AWEM. [1,2,3,4,5,6]. Despite this progress, issues persist, such as, animal welfare response being disconnected from the official overall emergency response [7]. Inadequate AWEM responses often result in increased risk behaviours of animal owners [7]. This can then result in failure to meet the intended outcomes of protecting public safety and wellbeing, animal welfare, food security, and biosecurity [1,4,7,8,9]. Improving AWEM requires a shift from an inclusive approach to one in which animals are fully integrated into emergency management [7]. This demands a transdisciplinary approach to emergency management in which all stakeholders acknowledge the importance of the h-a-e interdependencies and implement a framework to support this [10]. Terms such as multidisciplinary, interdisciplinary, and transdisciplinary are frequently used interchangeably to discuss efforts that involve several disciplines. A multidisciplinary approach draws on the knowledge of different disciplines, but disciplines that work independently considering an issue, and their perspectives typically remain unchanged. With an interdisciplinary approach, knowledge is shared between disciplines but work and perspectives continue to be largely rooted in independent disciplines. A transdisciplinary approach involves diverse stakeholders providing complementary perspectives and contributing unique expertise to search for ‘whole of problem’ solutions that ‘transcend’ their own discipline [11,12].OW is a concept that describes the interrelationships between animal welfare, human wellbeing, and the physical and social environment with the aim of creating a platform to enhance the understanding of, and response to, the complexities of the h-a-e interrelationships [13,14]. OW acknowledges that h-a-e interrelationships transcend the expertise and boundaries of any one organisation and seeks to transition from traditional management by individual sectors towards an interdisciplinary approach [15,16,17]. It is a way of breaking down barriers between agencies, individuals, sectors, and the community [18]. OW evolved from the One Health (OH) concept of structured collaboration and coordination between multiple health science professions to attain optimal human, animal, and environment health systems [19,20,21,22,23]. OH has been cited as an efficient and sustainable governance approach to address complex health issues [16]. It is argued that although OH has sparked an evidence-based body that goes beyond individual disciplines, the strong health focus lacks a vision of a set of social, cultural, economic, and environmental outcomes whose interdependence is similarly acknowledged, thus creating a significant gap in welfare-focused approaches [14,24]. To address this gap, OW was developed as a broader concept to embrace an interdisciplinary approach [21,25,26]. OW is better suited to the emergency management context as it is not solely focused on health and, by encompassing social, economic, environmental, and cultural interdependencies, is more holistic.While little is known about how the new OW concept might be applied in practice, the experience from OH implementation provides valuable insights to inform the implementation of an OW framework in emergency management. Although positive steps towards achieving an integrated OH approach have occurred over several decades, there continue to be challenges in its implementation [16,19,24,27,28]. These include lack of awareness and understanding of OH across the sectors required to implement the approach, insufficient “whole-of-government” policy prioritisation and funding to support OH, lack of integrated education and training programmes, reliance on the leadership of a few individuals and institutions, and the translation of global OH to local-level efforts and outcomes [16,18,19,24,27]. Dos S. Ribeiro et al. explain: “the challenges in stakeholder collaboration relate to the fact that a multidisciplinary team of scientists [and practitioners] working together but within their own silo is not enough for the knowledge co-creation proposed in OH innovations” [24] (p. 3).OW provides a framework to address emergency management silos that negatively impact on AWEM, but challenges, similar to those faced with OH, are likely. Strategies will be needed to address these, and to provide practical examples and evidence of the value of OW. While OW has been mentioned in AWEM literature, this has largely been limited to comments about the need to include animals by considering them alongside people [8,14,29,30]. There has been no detailed consideration of the potential for OW to promote interdisciplinary collaboration through a transdisciplinary approach. This paper addresses that gap by arguing for the full integration of animals in emergency management and discusses how that could be implemented.Additionally, this paper synthesises and extends a suite of conceptual frameworks, scholarly work around implementation, research findings, and practice-based knowledge into an integrative framework (Figure 1). Figure 1 illustrates several core strategies aligned with the complex interdependencies between humans, all animals (companion, production, and wildlife), and the social (the social environment is depicted as a marae, the social and spiritual hub of a Māori community, a place where people gather for meetings, celebrations, funerals, and other important events) and physical environments. The strategies that have been tested and refined within the New Zealand emergency management system include legislation and policy changes including: h-a-e interface interactions as business as usual, improving knowledge through interprofessional education and training, incorporating OW champions, and recognising the role of animals as vital conduits into communities.2. Legislation and Policy: From Animal-Inclusive to IntegratedIn New Zealand and elsewhere, the failure of emergency management legislation and policy to reflect the h-a-e interdependency has artificially compartmentalised human, social, and animal welfare and created barriers to collaborative practice [7,13,31,32]. However, given that community wellbeing and safety is at the heart of emergency management, it is vital to break down the sectoral partitioning that exists between emergency management agencies, emergency service organisations, AWEM support agencies, and other sectors of activity.Animal welfare legislation, separate from emergency management and emergency service legislation, has limited application when it comes to collective planning for and management of animal welfare in disaster situations [33]. As a result, fragmented policies and plans set different goals and standards for various agencies, sometimes in direct conflict. For example, law enforcement agencies have the power to temporarily close a road by placing cordons, if they believe that there is a danger to the public. However, cordons may be counterproductive if animal owners engage in risky behaviours, such as breaking cordons to gain access to their animals [4]. Separating owners from their animals also means that they are unable to fulfil their duty of care (providing food, water, and shelter) as required under animal welfare legislation [8,33]. Without consultation, the wider consequences of measures, such as the placement of cordons, are not generally factored into the decision-making.If emergency management legislation and policy is to be fit-for-purpose, it must have multilateral coherence with other relevant legislation, be consistent across Acts and policy, and reflect h-a-e interdependencies to enhance social protection [18,34,35]. Generally, however, legislation and policy are not agile enough to reflect the complex needs of communities with changes being made in a reactive manner, rather than proactively, which only serves to increase sector fragmentation [27,34,35,36].Legislation and policy must respond to evolving societal needs and foster a holistic approach to emergency management that acknowledges h-a-e interdependency. This can only be achieved when animals are not simply ‘included’ in emergency management legislation and policy, but are fully integrated and done so in a manner that reflects a “whole-of-person” and “whole-of-society” approach [7,10,12,22,27,34,37,38]. For example, the insistent call for animal welfare plans to be in place as a mechanism to protect animals in emergencies is problematic. Plans are only of value if they are co-created with stakeholders who experience the h-a-e interface, build capacity and capability, and integrate h-a-e across all pillars of emergency response and recovery. For example, an animal plan that includes animal evacuation considerations will be constrained and ineffective if the overall evacuation plan does not directly reference animals or is not aligned with the animal plan. By focusing solely on animals, silos are created and the intended outcome of protecting public safety, animal welfare, the economy, and biodiversity is not achieved. Legislation and policy must reflect a transdisciplinary view of emergency management in which h-a-e interdependencies are an integral component. This is a critical step in addressing interagency fragmentation, competition, and breaking down barriers created by the bureaucratic division of responsibility, which is counter-productive to the desired goals of emergency management.Flexible, innovative practices implemented during the Eastern Bay of Plenty flood [7] and the 2019 Pigeon Valley wildfire responses [39] in New Zealand illustrate the positive potential of an OW approach to emergency management. During these emergencies, an animal welfare coordinator was included in the emergency operations centre (EOC) and incident management team (IMT), which assisted with better integration of the animal welfare response with an overall better response, facilitated access to valuable information and intelligence, offered opportunities to advocate for the inclusion of animals in decision-making, and articulated the consequences of decision-making on animal welfare and animal owner behaviour. An animal welfare coordinator also became part of the cordon management team, which created opportunities for amalgamated animal response teams to be granted emergency access to cordoned areas to assess animals and address any immediate needs. The inclusion of multiple and more diverse perspectives within pivotal decision-making teams, such as the EOC, was a departure from previous practice and illustrates how the OW approach can serve as a catalyst to transdisciplinary consequence management [40]. In this way, response and recovery environments provide a unique opportunity to test, refine, and develop ideas and provide practice-led evidence in support of legislative and policy changes.3. Normalising the Presence of h-a-e Interface Networks and RelationshipsEmergencies are complex and fluid. They require a diversity of people, disciplines, and organisations to share expertise, perspectives, and resources, and to collaborate and form a common goal of addressing acute challenges facing communities [41]. Strong collaborative networks based on existing, trusted relationships support better decision-making processes and actions during an emergency [42]. Yet, many of those encountering the human–animal interface during an emergency meet for the first time in a highly stressful environment which is not conducive to developing the interprofessional trust and understanding necessary for positive transdisciplinary relationships [2,7,43,44].In the absence of pre-existing relationships, emergency management organisations are frequently required to act as brokers between agencies during a response [45]. However, networks created under such conditions are rarely long-lasting, resilient, or cost-effective with previous siloed ways of working coming to the fore [46]. Ideally, organisations need to forge direct links with one another so that their interactions are normalised into business-as-usual activities, rather than relying on emergency management organisations to broker collaborative networks during emergencies. In this way, business-as-usual interactions can lead to more positive and trusting relationships and networks that underpin meaningful collaboration during times of crisis.Disasters, while affecting whole communities, share some impact characteristics with other crises and extreme stressors experienced at the h-a-e interface, such as domestic violence, animal hoarding, structural fires, and animals requiring technical rescue [15,26,47,48,49]. These incidences often transcend the expertise and/or jurisdiction of any one organisation and should involve agencies such as emergency services, law enforcement, human services, animal control, environmental health, animal charities, and veterinary professionals [43,49,50]. In practice, however, agencies generally work in silos, and do not deploy simultaneously, and if they do it is usually ad hoc due to informal individual relationships [7,51]. This siloed approach leads to unsafe practices [7,52,53], inadequate resourcing and capability and, in the case of animal hoarding, a high recidivism rate [17,49,54].For example, while fire services frequently experience h-a-e interactions with the presence of animals in structural fires, motor vehicle accidents, and with entrapped animals requiring rescue, they do not routinely work with animal organisations. Such incidences occur more frequently than floods, fires, and earthquakes, yet the presence of animals creates the same complexities [48,52,55,56].Safe and successful resolution of emergency situations involving animals requires emergency services personnel to have specialist training, skills, and equipment, and a collaborative multidisciplinary team [4,48,57,58]. However, emergency services did not traditionally have the skills or experience to respond to the presence of animals in such incidences, have lacked relationships with animal organisations, and relied on an element of luck which resulted in a high risk of serious harm to the animal, owner and responders, and significant reputational risk [44,52]. It is equally important that supporting animal agencies have an understanding of, and align with, emergency management systems and practices [44].Internationally, the approach to entrapped animal rescue has been increasingly transdisciplinary and collaborative with the inclusion of interprofessional education and training, simultaneous deployment of animal expert resources by emergency services, utilisation of complementary skills and resources, and inclusion of the presence of animals as part of the incident risk assessment [48,52]. This has resulted in better response outcomes with a decreased risk to responders and animal owners, reduced mortality of animals, created a better understanding of individual disciplinary skill sets, roles, and responsibilities, and more efficient and effective responses and opportunities for transdisciplinary teams [48,55,59]. Regular interactions between fire services and animal organisations over ‘routine’ incidences would cement the relationships needed during an emergency, such as a wildfire, earthquake, or flood event. This would enable them to move from novel relationships being brokered by emergency management organisations to normalising the relationships needed for AWEM. The transition to collaborative transdisciplinary teams requires a novel approach to identifying and aligning the synergies between an organisations’ core business and h-a-e interdependencies. Focusing on the synergies between diverse positions can improve access to high-quality information and incentives within the network, increase confidence, reliability and integrity, and develop a sense of trust and reciprocity between partners [60].4. Interprofessional Education and TrainingA key step in moving emergency management systems from fragmentation to a position of strength and unity is to become a collaborative practice-ready emergency management workforce [41,61]. There is clear evidence that interprofessional education enables effective collaborative practice through attitude change, greater understanding of the roles and responsibilities of others, increased awareness of barriers across professions, and increased awareness of the importance of professional collaboration [61,62,63].Interprofessional education and training is delivered in health and aviation sectors and is recognised as an essential element of successful transdisciplinary collaborative practice, noting that just working with others in scenarios or temporary teams is not enough to build an effective collaborative practice-ready workforce [16,18,19,22,28,61,62,64,65,66].Despite individuals being expected to possess the skills, knowledge, and attitudes necessary to work together in interprofessional emergency management teams, education and training has largely occurred in professional silos or assumed to be learned during scenarios [67]. Emergency management and emergency service disciplines are taught almost exclusively in isolation from other emergency responders who experience the h-a-e interface. Animals in emergencies are generally not included in emergency management, emergency services, veterinary, animal science, agriculture, or environmental science curricula [68]. The lack of interprofessional education and poor curriculum integration amplifies differences in organisational cultures and reinforces barriers to effective interagency collaboration.If collaborative practice is to become the norm, change will be needed in attitudes, systems, and operations. By embedding interprofessional education and collaborative practice into legislation, accreditation requirements and/or registration criteria, policy-makers and government leaders can champion change and endorse interprofessional collaboration [61]. Leaders, who choose to contextualise, commit, and champion interprofessional education and collaborative practice, position their emergency management system to strengthen disaster risk governance, a priority of the Sendai Framework for Disaster Risk Reduction 2015–2030 [69]. Endorsed by the United Nations General Assembly in 2015, the framework aims to substantially reduce disaster risk and losses in lives, livelihoods, and health and in the economic, physical, social, cultural, and environmental assets of people, businesses, communities, and countries [69]. It recognizes that the State has the primary role to reduce disaster risk, but that responsibility should be shared with other stakeholders, including local government, the private sector, and other stakeholders [70].The potential for working relationships in emergency management to be in flux due to the rise and complexity of emergency events, new occupations, professionalisation of emergency management, and challenges to the historically dominant single profession makes the present an ideal time to promote collaborative practice and implement interprofessional education.Interprofessional education and training should promote complementary approaches to emergency management and include core interprofessional practice competencies across disciplines whilst being sympathetic to the conceptual and practical differences.5. One Welfare ChampionsThe shift to an OW emergency management framework will require a significant shift with stakeholders expected to operate in a context of collaborative practice with different organisational cultures coming together and working towards a common goal. The success of such a challenge, requiring commitment over a long period of time [24], is critically influenced by the presence of strong and innovative champions of change [24,35,71,72]. Transformation of new cross-sector ideas and concepts without transdisciplinary champions rarely have the impetus needed for successful implementation [73]. Successful champions have the courage to break down barriers, take risks, and broker opportunities for collaborative practice whilst creating engagement and trust among all stakeholders [24]. They exhibit influence, ownership, physical presence during interactions, persuasiveness, and a participative leadership style [72,74,75]. Facilitating a process to improve the current state to a desired future level [2,74,75], champions use their mana (in Māori, mana refers to a person’s prestige, authority, control, power, influence, status, spiritual power, and charisma, https://maoridictionary.co.nz/word/3424, accessed on 1 October 2021), knowledge, resources, and influence to help navigate the complex socio-political maze within their organisations. Using language that their discipline understands, they address resistance to new ideas and build organisational coalitions [72,76].OW champions will need to work within and across organisations experiencing the h-a-e interface before, during, and after emergencies. This includes all levels of government, emergency services (police, fire), emergency responders (emergency management officers, response teams, lifelines, and utilities), defence forces, human service organisations, veterinary professionals, human and animal welfare charities, primary sector, and geological and environmental practitioners.Champions for OW must understand the challenges their organisations experience due to h-a-e interdependencies, know how to connect between multiple actors, domains and levels, forge change through transdisciplinary collaborative practice, and be able to shift the focus from business-as-usual to innovative practice. Collectively, champions need to be able to identify and utilise strengths, opportunities, and comparative advantages of disparate disciplines and organisations whilst engendering co-creation central to OW initiatives [24,29,72].If OW champions, representative of all levels within organisations and all key stakeholders, are engaged from the outset through the entire process, initiatives can be planned, designed, and implemented in a collaborative manner across all sectors so that deeper and sustainable change can be achieved for AWEM and emergency management.6. Animals as a Conduit to the CommunityDuring emergencies, human decision-making and actions are strongly influenced by their relationship with their animals [6,55,56,58]. Given that over 60% of urban and 90% of rural households in developed countries own animals [4,5,6], a failure in emergency management to acknowledge this is a failure to consider the whole person, and the measures intended to protect human life and wellbeing may, in fact, be counter-productive [7].During disasters, animal attachment can pose a risk to human safety but conversely, it can be leveraged through community engagement strategies to increase disaster preparedness [59,77]. Shifting the balance from the negative influence of the human–animal bond to more positive outcomes will require a transdisciplinary approach to develop innovative strategies to engage with and motivate animal owners to better prepare for and respond to disasters. Animal support organisations have a unique opportunity to create connections with people due to their mutual connection with animals. This indirect social benefit of animal ownership can bridge the gap between people and facilitate coordinated, cooperative actions for mutual benefit [15,78]. In emergencies, people cannot always express their emotional needs, and innovative engagement strategies are needed to ensure equitable access to available resources [55]. Asking about animals can be an ‘icebreaker’ in a social setting and provides a means of building rapport [36] and trust between newly acquainted people [78]. This mutual, relatively ‘safe’ topic of conversation can elicit important information about relationships and family functioning [79] and can be used to help individuals establish social connections. This is particularly true for vulnerable or hard-to-reach members of the community, such as the elderly, people affected by a mental illness or drug dependencies, refugees, indigenous people, women, single parents, people with disabilities, and the homeless [55,80]. Though socially isolated and/or vulnerable individuals may possess the skills to function on a day-to-day basis, they may lack the resilience to cope with a crisis situation. Where connectedness to others within a community may be lacking, the support people feel from animals can strengthen the emotional resilience of socially isolated individuals [81]. During the recovery phase of disasters when communities can become fragmented and social support networks are lost, there can be a rise in the number of socially isolated and/or vulnerable people. It is during this time that people need appropriate psychosocial assistance, but many may not know where to seek help or believe that they are not entitled to it [82]. As many animal owners are motivated to look after their animals before themselves, and have high trust and pre-existing relationships with animal service organisations, these organisations play a valuable role in the psychosocial recovery of communities, such as the role of navigators. A navigator’s role is to link people to other agencies providing support services and to help people navigate their way through the complex support systems that activate in emergencies, such as financial assistance, temporary accommodation, psychosocial services, and insurance support. Animal welfare support organisations may not be aware of support services available or have training in psychological first aid; therefore, navigators require collaborative multiagency, multidisciplinary support, which includes experts in psychosocial recovery. Animal ownership offers a unique opportunity for communicating with people and motivating them to engage in resilience building behaviours that promote survival for themselves and their animals, and facilitate recovery from a disaster. Animal service organisations can act as conduits for hard-to-reach community members and have an important role to play as recovery navigators. It is important to note, however, that as front-line workers in emergencies, those providing animal welfare support may experience mental health consequences as a result of exposures to secondary trauma [83,84,85,86] and may require support themselves. 7. ConclusionsEmergency management requires many different individuals and organisations to work together towards common goals under extraordinary circumstances. Unfortunately, professional silos and a lack of understanding of the importance of h-a-e interdependencies means AWEM remains largely disconnected from emergency management overall. This paper argues that by adopting an OW framework and creating a sustained practice of engaging and partnering with others across the h-a-e interface, AWEM responses can be greatly improved. However, as seen with OH, there will be significant challenges moving from theory to practical implementation. To overcome the challenges and optimise the outcomes for AWEM, several key strategies are suggested.Effective interdisciplinary collaboration is vital if OW is to work in practice, yet current legislation and policy incorporate much that is counter to this. Change at this level is vital and must underscore all other change initiatives. Agencies involved with the human–animal interface during an emergency must be supported to develop positive transdisciplinary relationships through business-as-usual interactions, rather than waiting until a crisis, which is not conducive to building trust and understanding. Interprofessional education and training also has a critical part to play in improving interdisciplinary collaboration by increasing knowledge of the multiple roles and responsibilities within AWEM, and by developing skills for functioning as practice-ready interprofessional teams.The implementation of an OW framework is transformational change. Engaging committed and skilled change champions representative of all stakeholders and levels within organisations will be essential across all phases of implementation. Finally, the human–animal bond should be harnessed as a valuable conduit for communication and engagement with communities and significant stakeholders central to the OW concept.

animals : an open access journal from mdpi

10.3390/ani11102891

PMC8532714

Cellular reproduction is a key part of the apicomplexan life cycle, and both mitotic (asexual) and meiotic (sexual) cell divisions produce new individual cells. Sexual reproduction in most eukaryotic taxa indicates that it has had considerable success during evolution, and it must confer profound benefits, considering its significant costs. The phylum Apicomplexa consists of almost exclusively parasitic single-celled eukaryotic organisms that can affect a wide host range of animals from invertebrates to mammals. Their development is characterized by complex steps in which asexual and sexual replication alternate and the fertilization of a macrogamete by a microgamete results in the formation of a zygote that undergoes meiosis, thus forming a new generation of asexual stages. In apicomplexans, sex is assumed to be induced by the (stressful) condition of having to leave the host, and either gametes or zygotes (or stages arising from it) are transmitted to a new host. Therefore, sex and meiosis are linked to parasite transmission, and consequently dissemination, which are key to the parasitic lifestyle. We hypothesize that improved knowledge of the sexual biology of the Apicomplexa will be essential to design and implement effective transmission-blocking strategies for the control of the major parasites of this group.

The phylum Apicomplexa is a major group of protozoan parasites including gregarines, coccidia, haemogregarines, haemosporidia and piroplasms, with more than 6000 named species. Three of these subgroups, the coccidia, hemosporidia, and piroplasms, contain parasites that cause important diseases of humans and animals worldwide. All of them have complex life cycles involving a switch between asexual and sexual reproduction, which is key to their development. Fertilization (i.e., fusion of female and male cells) results in the formation of a zygote that undergoes meiosis, forming a new generation of asexual stages. In eukaryotes, sexual reproduction is the predominant mode of recombination and segregation of DNA. Sex is well documented in many protist groups, and together with meiosis, is frequently linked with transmission to new hosts. Apicomplexan sexual stages constitute a bottleneck in the life cycle of these parasites, as they are obligatory for the development of new transmissible stages. Consequently, the sexual stages represent attractive targets for vaccination. Detailed understanding of apicomplexan sexual biology will pave the way for the design and implementation of effective transmission-blocking strategies for parasite control. This article reviews the current knowledge on the sexual development of Apicomplexa and the progress in transmission-blocking vaccines for their control, their advantages and limitations and outstanding questions for the future.

1. IntroductionCellular reproduction is a key part of the cellular (and organismic) life cycle. Both mitotic (asexual) and meiotic (sexual) cell divisions produce new individual cells. The general definition of sexual reproduction refers to the process where DNA combination of two cells creates new offspring through the fusion of two different cells (syngamy), nuclear fusion (karyogamy), and the final process of homologous chromosome crossover (meiosis). Sexual reproduction is widespread in all eukaryotic branches of the tree of life, and it is the predominant mode for recombination and segregation among all major groups of eukaryotes [1,2]. Sex evolved early in eukaryotic evolution, appearing approximately 850 million years ago, at least 200 million years after bacteria evolved [3]. The predominance of sexual reproduction in eukaryotes indicates that it has had a considerable success during evolution, and it must confer profound benefits, considering its significant costs relative to asexuality [4]. This success is due to meiosis, in which the ability to recombine the genomes of individuals passes a unique mixture of parental genes onto the next generation and makes selection highly efficient. Sex generates genetic variation, repairs DNA breaks and prevents accumulation of mutations [5,6].The last eukaryotic common ancestor (LECA) had all of the features associated with modern eukaryotes: sexual reproduction through meiosis and genetic material inside a nucleus. It was thus capable of full “meiotic sex” [3,7]. Certain eukaryotes lost the ability to reproduce sexually, such as species of Daphnia and aphids [2,8], but ancient asexuality appears to be rare in eukaryotes [2,9]. The presence of genes encoding the meiotic machinery in early diverging protists implies that sexual reproduction arose early in eukaryotic evolution [7]. The term “protists” refers to those eukaryotes that are never multicellular, essentially any eukaryotic organism that is not an animal, plant, or fungus [10,11]. Certain lineages may be more closely related to animals, plants, or fungi than they are to other protists; however, the grouping is used for convenience, and the clade is divided in groups such as the SAR supergroup (Stramenopiles, Alveolata, and Rhizaria), Archaeplastida, Excavata (mostly unicellular flagellates), Amoebozoa, Hacrobia, Hemimastigophora, Apusozoa and Opisthokonta [12]. Sex is well documented in many of these groups; however, direct observations of sexual processes are missing for a majority of protist species, and in certain cases, entire protist lineages cannot be considered as “sexual” with any certainty yet. In many different taxa of parasitic protists, sex and meiosis are linked to transmission [8]. In certain well-studied clades from highly diverse lineages of amoebozoans, apicomplexans (Alveolata), and kinetoplastids (Excavata), sex is described to be induced under the (stressful) condition of having to leave their host, and either gametes or zygotes are transmitted to a new host and eventually disseminated which is key to the parasitic lifestyle [13]. In parasitic species that form cysts or oocysts, these resistant forms that are easily spread in the environment are often in the transmissible stages, and they may have occurred in early eukaryotes and may be linked to the evolution of sex [3,14].The phylum Apicomplexa is a major group of protists, which includes species that cause important human and animal diseases worldwide. It is a monophyletic, extremely large and diverse group with more than 6000 named, and possibly, thousands of undescribed species. Five apicomplexan groups with medical and veterinary importance are gregarines, coccidia, haemogregarines, haemosporidia and piroplasms [12,15]. With few exceptions of non-parasitic (commensal or mutualist) species, Apicomplexa are obligatory parasites, and potentially every vertebrate and the majority of invertebrates host at least one species. It includes the causative agents of e.g., malaria, piroplasmosis, cryptosporidiosis, toxoplasmosis and coccidiosis. The genera Babesia, Theileria, Cryptosporidium, Eimeria, Toxoplasma, Cystoisospora and Neospora cause major diseases of veterinary importance linked to economic losses in the affected livestock. These parasitic diseases impair animal health, reproduction and growth and lead to economic losses with regard to decreased production, treatment costs, morbidity and mortality [16,17]. The most cost-effective strategies for the control, including prevention, elimination or eradication of pathogenic diseases are vaccines [3]. Compared to other pathogens such as viruses and bacteria, there is an eminent lack of antiparasitic vaccines; only two anti-nematode vaccines, one anti-tick vaccine and several antiprotozoal vaccines are commercialized for domestic animals [18]. Among all the antiprotozoal vaccines available, the anticoccidial vaccines for poultry are well developed. By contrast, no vaccine is available for protozoal diseases of humans [19,20], with the exception of two candidates for malaria, with promising results in early clinical trials [21,22,23].Over the past decade, knowledge on the general biology of Apicomplexan parasites has increased considerably, and has helped to explore host-parasite interactions, vulnerabilities of important parasites within their developmental cycle, host immune reaction and development of immunity, all of which is of great importance for the development of control strategies. In addition, major advances in the current knowledge on parasite transmission biology has increased our understanding on the epidemiology and risk factors for infection. Apicomplexan parasites as a group infect a wide host range of animals from invertebrates to mammals (although the host range of a single parasite species can be restricted to a single host species). This development is frequently characterized by highly complex steps in the life cycle in which asexual and sexual replication alternate, and the fertilization of a macrogamete by a microgamete results in the formation of a zygote that undergoes meiosis, forming a new generation of asexual stages [24]. Depending on the taxonomic subgroup, development either is restricted to single or few host species, and the parasite concludes the whole endogenous life cycle in the same host species (e.g., in the coccidian genera Eimeria and Cystoisospora), or a host switch is required for the sexual development and completion of the life cycle (e.g., Plasmodium). The sexual stages constitute an important bottleneck in the life cycle of these parasites, as they are obligatory for the further development of transmissible stages. Consequently, the sexual differentiation process and the characterization of the gametocytes have become attractive targets for both basic research on cell biology and applied vaccine development. Sexual development of apicomplexan parasites leads to the production of gametocytes, which are characterized morphologically and by the expression of stage-specific genes; however, the initiation of gametogenesis is thus far only poorly understood [25,26]. Sexual development in different apicomplexan taxa varies greatly and still needs further research to provide sufficient details for translational and applied research [27].Here, we review the current knowledge on the sexual development of Apicomplexa parasites and the progress in research on transmission-blocking vaccines for their control, their advantages and limitations, and possible approaches and outstanding questions for the future. 2. Sexual Reproduction in ApicomplexaWith few exceptions, Apicomplexa are obligatory parasites. Most members have a complex life cycle with significantly different forms (stages) characteristic for each of the main taxonomic groups of the phylum, and they fully depend on their hosts throughout most of their life cycle. The terminology used to describe these various life cycle stages varies between families, but the generic developmental cycle is divided into three distinct phases with alternating asexual and sexual multiplication—sporogony, merogony and gamogony—including four different basic cell types—sporozoite, merozoite, gametes (haploid types) and the zygote (diploid type) [24,28]. The release of sporozoites from the oocyst or sporocyst marks the initiation of the infection process, and host cells are invaded. The gliding sporozoite targets and invades the cells by using its eponymous apical complex, transforms into a trophozoite and immediately initiates a sequential series of asexual reproduction steps (referred to as merogony), resulting in the development of merozoites. These ultrastructurally resemble the sporozoite and infect further host cells, resuming asexual division to produce a new generation of merozoites. The merogonic cycle can be summarized as repeated events of invasion, replication and host cell egress of merozoites, and it precedes gamogony [29]. The final generation of merozoites is considered to be already sexually committed, despite morphological resemblance with the asexually reproducing preceding stages [30]. Although the trigger and the process of sexual differentiation of apicomplexan parasites are still obscure, they appear to be genetically programmed and not directly dependent on environmental changes [25]. In the following phase, the gamogony, some merozoites become female macrogametocytes (macrogamonts), while the majority transforms into male microgametocytes (microgamonts). The life cycle eventually proceeds to the fusion of a motile flagellated microgamete with a large and immobile macrogamete, with fertilization leading to the formation of a diploid zygote [28]. The apicomplexan zygote finally develops into the sporozoite. The process is termed sporogony and is characterized by cell divisions that can vary in numbers between taxa, with several rounds of meiosis and mitosis, resulting in the formation of infectious haploid sporozoites. Upon the transmission of sporozoites to the next host, the life cycle is completed (see Figure 1) [29,31].The life cycles of different taxonomic groups vary in the number of hosts involved and the type of cellular invasion. In monoxenous species, the complete endogenous development occurs in a single host and frequently in a single cell type or tissue. By contrast, in heteroxenous species it involves different hosts and generally also different types of tissue. The life cycle of haemosporidia, piroplasms and coccidia is comprised of asexual multiplication, i.e., merogony and sporogony, and sexual development, gamogony [32,33], while most gregarines only practice gamogony and sporogony [34]. Although most Apicomplexa exhibit this overall general life cycle, the details can vary between species as outlined in more detail in Table 1.The phylum is divided into two classes, the Conoidasida with the subclasses Coccidiasina, Gregarinasina and Cryptogregarinasina (and a number of orders contained therein) and the Aconoidasida which includes the order Haemosporidia and Piroplasmida [15,28,35,36].2.1. Class Conoidasida2.1.1. Subclass GregarinasinaGregarines are the primitive lineage of the Apicomplexa. They have monoxenous life cycles with several exceptions (e.g., Nematopsis spp. utilize two hosts, crustaceans and mollusks) [37]. Most cycles comprise exclusively gamogony and sporogony, only few species include merogony [38]. Many gregarines do not exhibit intracellular stages and the sexual and asexual cycles occur extracellularly in the intestinal and coelomic cavities of their invertebrate hosts, setting the gregarines apart from most other apicomplexan taxa which infect vertebrates. The life cycle of most gregarines starts with the ingestion of the sporocyst by the host from which sporozoites are released and transform to trophozoites that attach to the host cell. Eugregarines infect the intestines, coeloms, and reproductive vesicles, neogregarines infect mainly the host tissues and archigregarines infect the intestines of the host [38]. There, the trophozoite increases in its size and consequently breaks the host cell. The developmental step that precludes gregarine sexual reproduction is called syzygy, the association of two mature trophozoites after they detach from the host cell. The two assembling cells are committed to develop into male and female gamonts, both unflagellated. Mature gametes are generally similar in shape and size and are produced in equal amounts [28]. Micro- and macrogametes mainly pair end-to-end before they fuse to form a zygote, but some have species-specific orientations such as head-to-head, tail-to-tail or head-to-tail [39]. The zygote forms a protective envelope, which later becomes the oocyst wall. Finally, it will undergo meiosis to produce new sporozoites, which will infect a new host [34,40]. Archigregarinida and Eugregarinida do not display merogony, while different types of merogony are reported in Neogregarines [39,41]. Although there is abundant literature about their life cycle, morphology and ultrastructure, gregarines are almost unstudied at the genomic/transcriptomic levels and have been the subject to only few biochemical analyses [40].2.1.2. Subclass CryptogregariaIn contrast to the poorly studied gregarines, genomic studies applying a variety of molecular tools have greatly advanced our understanding of the biology of the genus Cryptosporidium [42], including sexual development. This genus was recently classified as the sole member of the new subclass Cryptogregaria, as it resembles the gregarines in many developmental aspects [43,44,45]. Cryptosporidium has a monoxenous life cycle consisting of several sexual and asexual stages. In Cryptosporidium, merozoites arising from type-II-meronts multiply sexually to produce gamonts. Macrogamonts develop further into large single-celled immobile macrogametes, whereas sixteen smaller single-celled microgametes without flagella develop from a microgamont, which bursts open to release the mature microgametes. Since 2002, when the first study by Hijjawi and colleagues demonstrated the presence of gamont-like extracellular stages in the life cycle of Cryptosporidium, several investigations tried to elucidate the origin of these stages [46]. To achieve fusion of micro- and macrogametes, both must appear at the same time and be compatible with each other. The microgametes fertilize the macrogametes, producing zygotes, which mature into oocysts [47]. During sexual differentiation, the expression of genes responsible for fundamental cellular processes changes widely. Cryptospoiridium parvum microgametes express the homologue of hapless2, a class II membrane fusion protein that is required for gamete fusion in Apicomplexa, similar to most eukaryotic organisms [48,49]. Macrogametes show an increased expression of conserved eukaryotic factors of meiotic recombination (DMC1, Spo11, HORMA and HOP2), meiosis-associated DNA repair, cell cycle regulation factors and genes involved in oocyst wall formation and wall persistence [50,51]. The oocyst wall forming protein COWP1 is expressed both in macrogametes and oocysts of C. parvum [52], and similar homologues also exist in other Apicomplexa [53,54,55].2.1.3. Subclass CoccidiasinaThe order Eucoccidiorida includes a large number of pathogens of vertebrates with varying life cycles, host specificities and host roles [28], but all undergo sexual reproduction within specific hosts. However, the two suborders Adeleorina and Eimeriorina differ greatly in their life cycles; Adeleorina develop via syzygy while Eimeriorina produce independent gametes. Of the many genera and species of the order, the sexual development has been studied in only several, including the model species T. gondii [56,57,58], certain species of the genus Eimeria [25,59,60,61], and recently, Cystoisospora suis [53]. Sexual development of T. gondii occurs exclusively in the feline intestine [62], while Eimeria species can display distinct host and tissue tropisms, such as small and large intestines, stomach, the gallbladder or bile ducts, liver or gonads [63,64,65,66], but in most species, gamonts are formed in the intestinal lumen of the definitive host after merogony is completed, and the final generation of merozoites is considered to be sexually committed [67]. The early gamonts are already differentiated into micro- and macrogamonts, which are both immobile and morphologically similar in shape and size [67]. In the microgamont, the nucleus divides several times, and the newly formed nuclei locate to the periphery of the cell. Microgamete morphology varies among the Coccidia. The larger, unflagellated and immobile macrogamete contains a single nucleus larger than the small nuclei that are formed in the microgamont. The number of micro- and macrogametes that are formed from the gamonts varies greatly between coccidian taxa, but microgametes are usually greater in number than macrogametes [68,69]. The microgamete approaches the macrogamete and upon fusion of the two cells a zygote is formed. This is followed by several divisions of the zygote and simultaneous formation of the oocysts wall. The oocyst is the environmentally resistant stage that is either retained in the final host or discharged and disseminated in the environment [67,70].Eimeria spp. complete their whole life cycles in the same host, and most species develop in the epithelial tissue of intestinal and adjacent (e.g., bile duct) tissues [71,72]. After a genetically fixed and taxon-specific number of merogony cycles [25,73], two distinct sexual stages occur [24]. Macrogametes are immobile and remain intracellular, whereas the biflagellated microgametes can move freely to find a compatible macrogamete [74,75]. The egress process of microgametes from host cell and fertilization still must be elucidated, only few studies show microgametes in contact with macrogametes [61,76], and since microgametes have not cell-invading machinery, this could suggest it occurs extracellularly in the gut lumen [25,54]. Eimeria tenella transcriptomic analysis showed an upregulation of over 800 transcripts involved in gametogenesis, identifying numerous sexual stage-specific genes. Many of these transcribed genes are intrinsically involved with gamont biology, related to axoneme and flagellum formation, locomotion, gamete membrane fusion, DNA condensation and oocyst wall formation [25,54,77].In Toxoplasma gondii, macrogametes have an oval shape, a single nucleus, endoplasmic reticulum, mitochondria, lipid bodies, amylopectin inclusions, and two types of wall-forming bodies. Microgametes are elongated; they display a condensed nuclear chromatin, two flagella in the anterior region, and a mitochondrion located at the base of the flagella [78,79]. The fusion of micro-and macrogamete leads to the formation of an oocyst that is released into the intestinal lumen of the definitive host. Proteomic and transcriptomic studies uncovered hundreds of genes expressed uniquely in gametes, suggesting their potential role in gametocyte biology [78]. Recent studies can also show that upregulated genes in the sexual stages of T. gondii and C. suis (which is closely related to T. gondii and has sexual stages of similar morphology [80,81]) are those coding for proteins already highlighted as playing critical roles in the sexual biology of other Coccidia, including oocyst wall formation, microgamete motility and fertilization [56]. In Toxoplasma, Cystoisospora, Eimeria and Cryptosporidium, genes coding for oocyst wall proteins are already upregulated in macrogamonts, while proteins involved in axoneme-flagellar formation and motility and fusogen proteins are restricted to microgametes [82,83,84]. Cystoisospora suis also produces gamonts which develop further into clearly differentiated spherical macrogametes and microgametes. The flagella in C. suis are positioned on opposite sides, which might also affect microgamete movement on the search for a macrogamete whereas the flagella in Eimeria or Toxoplasma present a different morphology where the flagella are localized in the anterior region [25,78,85]. After in vitro merogony in epithelial host cells, C. suis also can continue gamogony in a host cell-free environment, indicating that gamete production and fusion occur extracellularly, as previously indicated for C. parvum [47,86]. The general development pattern of the suborder Adeleorina is highly complex, and at the species level, details in the life cycles often greatly differ [35,87]. All species in this suborder from gametes by syzygy. This involves the association of (often motile) gamonts prior to the formation of functional gametes and fertilization [28]. Despite this resemblance with the gregarines, Adeleorina are currently classified as Coccidia, and have similarly differentiated life cycles [88,89]. Members of the family Haemogregarinidae (common heteroxenous parasites of cold-blooded, less frequently warm-blooded vertebrates [90]) are characterized by their ability to invade different organs, cell types and hosts. They develop two morphologically different forms of meronts (micro- and macromeronts), which mainly infect erythrocytes and develop in these cells to sausage-shaped gamonts which subsequently give rise to differentiated micro- and macrogamonts, lying in syzygy in the same host cell. Macrogametes are larger and often immobile, whereas microgametes seem to be smaller, but occur in larger numbers. Haemogregarinidae show a large morphological diversity in microgametes, from unflagellated, monoflagellated, and biflagellated forms. Fusion of gametes (which occurs in invertebrates as definitive hosts) leads to formation of sporozoites enclosed in a sporocyst surrounded by an oocyst for continuation of the life cycle [90]. In contrast to Haemogregarinae, intracellular gametocytes of Hepatozoon (as a member of the Hepatozoidae) are found in the blood circulation either in erythrocytes or in leukocytes [91,92], and in all Klossiella species described to date, gametogony and sporogony occur in renal tissues [35].2.2. Class Aconoidasida2.2.1. Order HaemosporidaIn vertebrates, haemosporidian parasites form gamonts in blood cells after asexual development [28,89,93], either in tissue (families Haemoproteidae and Leucocytozoidae), in tissues flowed by erythrocytes (family Plasmodiidae) or in leucocytes (family Garniidae) [94,95]. When the infected blood cells of the vertebrate host are ingested by an arthropod, the immature gamonts develop into mature gametocytes. Microgametes often have up to eight flagella, whereas macrogametes are immobile. The fertilization is extracellular, the zygote (termed ookinete due to its motile nature), elongated and mobile, penetrates the epithelial gut until reaching the basal lamina, and the oocysts are formed [96]. After one or more sporogonies, infective sporozoites are formed, which are injected into the vertebrate host when the hematophagous arthropod feed [97,98]. The most representative genus of this order is Plasmodium. Malaria parasites display asexual reproduction twice in the vertebrate host, in liver hepatocytes (pre-erythrocytic merogony) and in red blood cells (blood-stage merogony), and once in the mosquito (sporogony) [99]. The essential sexual stage occurs at the transmission from vertebrate to insect. Certain asexual erythrocytic parasites develop into either male or female gametocytes within erythrocytes. These newly formed gametocytes undergo gametogenesis within the midgut of the mosquito where fertilization occurs [100]. The mature gamete types must fuse to form an ookinete, which penetrates the epithelial layer of the midgut of the mosquito to form an oocyst [98,101]. Plasmodium falciparum does not have chromosomes that determine the sex of the gametocytes, but all merozoites derived from a single committed meront become either micro- or macrogametocytes. Generally, more macrogametocytes than microgametocytes are produced. Plasmodium shows sexual commitment before the appearance of gametocytes [26]. Gametocyte sex allocation is therefore not only regulated by early stage-specific transcriptional factors, but also by protein phosphorylation [102]. Sexual commitment in P. falciparum is a well-described process. Recent work has shown that sexual differentiation is controlled by the expression of a master regulator, the transcription factor AP2-G [103]. Sexually committed cells are thereby allocated to the gene expression underlying sexual differentiation, by the activation of a transcriptional feedback loop that drives AP2-G expression up [27,104,105]. Other genes of the ApiAP2 family are the major transcript factors in Plasmodium development and especially HP1, HDA2 are also responsible for the epigenetic regulation during sexual commitment [27]. Activation of AP2-G expression triggers expression of early gametocyte genes including Pfs16, Pfg27/25, Pfg14.744, Pfg14.745 and Pfg14.748 [106]. HAP2/GCS1, P16, P48/45, P230, and FACT are transcribed in microgametes in Plasmodium. In the macrogametes, PKG and P47 are the main transcribing factors [98]. Furthermore, the homeodomain protein 1 (HDP1) was found to be an essential regulator of gene expression during the sexual differentiation of Plasmodium. HDP1 binds DNA during Plasmodium development and is tightly associated with chromatin in early gametocytes [107].2.2.2. Order PiroplasmidaPiroplasms are characterized by the eponymous pear-shaped intracellular stages in host blood cells. As mentioned above, merogony can occur in a variety of cells, and gamogony and sporogony occur in the gut and salivary glands of arthropods [28,108]. After merogony, rupture of host blood cell releases the merozoites which can undergo an undetermined number of rounds of merogony in the blood. Asexual division continues until merozoites can transform into gamonts [109]. Similar to Plasmodium, sexual multiplication of the parasite starts by gametocytes appearing in the host’s red blood cells. During the blood meal of the tick host, gametocytes develop into gametes that mature in the tick midgut lumen [110,111]. Macrogametes are large spherical immobile cells, whereas mobile microgametes form the families specific “spiky-rayed-stages”, which are sometimes referred to as pseudopodia-bearing gametes. Theileria spp. have anisogametes clearly distinguishable by light microcopy, unlike in Babesia spp. which do not differentiate into micro-and macrogametes, but equally formed isogametes, but still form two gametes populations, which differ in their shape and cytoplasm density [112,113,114]. During gamete fusion, a filamentous structure between gamete membranes is formed and subsequently a finger-like protrusion of one gamete penetrates the opposite one [109]. Gamete fertilization then gives rise to a zygote that penetrates the tick peritrophic matrix of the tick’s epithelial cells. Inside these, the zygote undergoes a meiotic division and results in the formation of kinetes, which are released into the tick’s hemolymph. The kinetes of Theileria spp. directly invade salivary glands (primary kinetes), while those of Babesia spp. parasites are subjected to two series of asexual multiplication in various tick tissues, and subsequent secondary kinetes invade the tick salivary glands [115]. Sporogony starts after kinete invasion of salivary gland cells, and the sporont is formed which has a polymorphous syncytium. The sporont later evolves into a multinucleated meshwork referred as a sporoblast, which is dormant during tick ecdysis. Maturation of the parasite sporoblast starts after tick attachment to the host and results in sporozoites being released into tick saliva [108]. Little is known about the expression of sexual stage-specific genes in piroplams. However, genes of the ApiAP2-family show changes in their expression level during the merogony of T. annulata. It hints to an involvement in the transmission of T. annulata to the tick vector, hence beginning gametogenesis [116]. An analysis of regulatory domains responsible for life cycle transitions in Babesia bovis, Babesia microti and Theileria equi show a high level of conservation between species and indicate that the transcriptional factor Api-AP2 is involved in the transition of the merozoites to gamonts [117]. Furthermore, Babesia bovis lacks 6-Cys sexual-stage genes, normally important for the development of sexual stages [118]. 3. Breaking the Cycle—Targeting Sexual Stages of Apicomplexan Parasites for InterventionIn most apicomplexan species of medical and/or veterinary relevance, sexual differentiation produces dimorphic sexual stages, male and female gametes, those will fuse and for some, the fertilization culminates in the development of an oocyst [24]. Only a small proportion of asexual stages (merozoites) will differentiate into gametocytes, and therefore this step is to be considered a bottleneck of development—not only in Plasmodium [48,119] but in all Apicomplexa that produce gametocytes. The gametocytes are the transmissible stages to invertebrate definitive hosts (for blood-dwelling Apicomplexa) or leads to oocyst formation (for gut-dwelling Apicomplexa) [24,28]. In each developmental step, a reduction of the population size occurs, also from the formation of gametocytes to oocyst development, resulting in low numbers of infective parasites compared to the multiplying endogenous stages preceding them. For example, in Plasmodium, it has been estimated that out of the thousands of gametocytes that a female Anopheles mosquito typically ingests with a blood meal, only 50–100 develop into ookinetes and only several of them survive to form the infected oocysts. The lowest parasite numbers occur during the oocyst stage, but when each oocyst releases its sporozoites and merogony is initiated, the number of parasites again drastically increases. For this reason, apicomplexan sexual stages constitute good targets for vaccination, since their numbers are lowest in comparison to the other stages that develop throughout the parasite’s life cycle, and consequently, they are easier targets for the immune system [101]. As stated previously, the life cycles of the most important Apicomplexa are elucidated; however, the process of differentiation into the sexual stages or the fertilization process remains largely unknown, and such information would be essential to develop and implement control strategies that suppress the passage of parasites from one host to the next. Among the many approaches to control protozoan diseases, transmission-blocking strategies, developed to interrupt the life cycle using different mechanisms for halting transmission to the vertebrate hosts, appear to be the most logical type of prevention. To achieve a transmission blocking strategy, it is essential to consider the key elements of the parasite’s life cycle and to identify its weak points. Such strategies can be based on different mechanisms; one is the generation of transmission blocking vaccine with the induction of high antibody titers against the specific proteins of sexual stages. These antibodies will bind to the surfaces of the gametes to prevent/block the progression of parasite development. They do not directly protect immunized individuals but specifically block the transmission [120]. Another strategy is the identification of chemotherapeutic transmission blocking drugs that affect specifically the sexual stages [121,122]. A reduction in gametocytes can result in reduced transmission, and both transmission-blocking vaccines and transmission-blocking drugs may have the potential for eradication of a disease. In recent years, the accessibility to the transcriptomes and proteomes of sexual stages provides a useful tool and support the definition of sex-specific gene transcripts and protein candidates for vaccination, some of which are exclusively expressed in either immature or mature gametocytes [25,48].Coccidia are characterized by the environmental oocyst as the final result of the fusion of gametocytes; thus, the reduction of this final stage would lead to a decrease in environmental contamination, and with this, the infection of new hosts. Interruption of the parasites’ life cycle before the formation of oocysts has been shown to abrogate oocyst formation. Over the past 40 years, the precise knowledge of the sexual reproduction of Eimeria genus resulted in the commercialization of a transmission-blocking vaccine against different Eimeria species in order to prevent coccidiosis in chickens. Wallach and coworkers, in the 1980s, were the first group to report immunization of chicken with Eimeria maxima gamont-specific proteins recognized by serum IgY collected from chickens that had recovered from E. maxima infection. In following studies, two glycosylated tyrosine-rich proteins of the wall-forming bodies of E. maxima, 56 and 82 KDa in size, were identified as antigens conferring protection because of a strong serum antibody response, not only against E. maxima but also against other Eimeria species of chicken due to their high degree of conservation across the genus [123,124,125,126]. In 2002, the commercial vaccine CoxAbic® using the native gametocyte antigens of E. maxima, Gam56 and Gam82, was commercialized. In clinical trials, CoxAbic® reduced oocyst shedding of the three major species of Eimeria (E. maxima, E. tenella, and E. acervulina) in broiler chicken by 50–80%. The assumption was that the vaccination with these gam proteins stimulated an immune response that blocked the construction of the oocyst wall [127]. The production of this vaccine based on affinity purification of native gametocyte proteins from parasites, however, was expensive and time-consuming. Since then, several proteins (molecular weights of 14, 22, 30, 56, 82 and 230 kDa) associated with E. maxima, E. tenella, and E. necatrix gametocytes have been proposed as potential vaccine targets to induce immunity [128,129,130,131,132,133]. The most recent study using Gam proteins involve the recombinant rEtGam22. Vaccination of chicken with EtGam22 significantly elicited both Th1 and Th2 cytokine mediated immune responses and induced protection against E. tenella and E. maxima-infected chickens [129]. Taken together, these studies indicate that Gam proteins are potent immunogens for the use as vaccines against chicken coccidiosis, as they induce a diverse and robust immunity [134].Regarding male gametocytes and their role in protection, monoclonal antibodies against microgametocytes of E. tenella reduced the oocyst formation in vitro more than 50% but the mechanism of inhibition was not investigated in detail [60], and this approach was not followed further. Oral application of sera containing E. tenella gamont-specific monoclonal antibodies significantly reduced oocyst output and cecal lesions in passively immunized chicken but the direct role of fertilization inhibition in immune-mediated protection was not explored further [135]. A new approach in the era of genetic manipulation technologies is the production of knockout (KO) parasite strains to be used as genetically attenuated live vaccines [136]. Recently, a HAP2-deficient T. gondii strain was created using the CRISPR/Cas9 approach. HAP2/GCS1 is a male fertility factor; it is a conserved protein expressed in male gametocytes, which was originally identified in Arabidopsis thaliana and later in Plasmodium spp. [49,137,138]. The HAP2-deficient T. gondii generated parasites that could infect cats and start asexual multiplication but failed to complete fertilization and undergo meiosis, which completely inhibited oocyst formation and excretion. Inoculation with HAP2-deficient T. gondii tissue cysts to cats completely prevented oocyst excretion, simultaneously inducing immunity to oral challenge with tissue cysts from wild-type T. gondii. This demonstrated impressively that interfering with fertilization can completely block transmission of this parasite and that the development of a transmission-blocking vaccine is feasible for coccidia [56].Independent of oocyst formation, gamete formation is also crucial for the development of downstream infectious stages (sporozoites) in Haemosporida and Piroplasmida. In Plasmodium, numerous sexual-stage specific proteins with antigenic properties and novel enzymes that putatively regulate functions during sexual-stage development have been suggested as candidates for vaccination or drug targets. Different surface proteins of macrogametes, such as Pfs48/45 and Pfs230 (both playing a role in male gamete fertility) have been shown to be immunogenic [139]. Zygote surface proteins expressed only post-fertilization in the mosquito host, such as Pfs25 and Pfs28, have also been considered as antigen candidates for transmission-blocking vaccines. In animal studies, Pfs25 vaccine candidates induced equal or greater serum transmission-blocking activity compared to other antigens or antigen combinations, and hence Pfs25 has been the focus of clinical trials published to date [139,140]. Ongoing trials are now examining the activity of Pfs230 and other transmission-blocking vaccine candidates, and it is expected that at least some of them will enter the clinical phase in the coming years [141,142]. Other studies proposed HAP2 protein as a candidate for a transmission-blocking vaccine in Plasmodium. When the hap2 gene is absent or mutated, the zygote formation is completely blocked, indicating its relevance in this event [143]. Antibodies targeting HAP2 inhibited P. berghei transmission in vivo by 58.9%, and anti-PvHAP2 antibodies reduced infection prevalence by 50% in P. vivax [144,145]. In addition, the glycolytic pathway has been proposed as a target for intervention as it is the source of energy for flagellar beat and hence mobility of the microgamete [146]. Motility through flagellar action is a unique feature of this stage [74], and the flagellum was proposed as a target in earlier works on immune intervention in Eimeria [135] and in Plasmodium [147,148]. In addition, a range of kinases, surface proteins, signaling and regulatory proteins of Plasmodium involved in sexual stage development, differentiation and zygote formation have been proposed as candidates [139]. In the piroplasm Babesia, specific antibodies against HAP2 significantly decreased zygote formation in vitro [149,150].4. Concluding RemarksDiseases caused by apicomplexan parasites, both in domestic animals and in humans, represent a significant health and economic problem in affected populations. Despite the availability of treatment against many of these diseases, there is an urgent need for vaccination strategies due to poor drug efficacy (mostly during the early stages of the infection), serious side effects of treatment, and the increasing resistance of parasites against parasiticides [20,151]. Considering these issues, a preventive vaccine appears essential for sustainable disease control. Vaccinating against apicomplexan parasites, however, is a complex task. The difficulties in the development and implementation of such vaccines include the intriguing complexity of the parasites’ life cycles, involving different hosts, stages and affected tissues, and the existence of poorly accessible arthropod vectors in many species, but also the lack of in vitro culture technologies and suitable in vivo animal models for large-scale screening. In certain taxa, the life cycle is not known in sufficient detail to study effects of intervention on parasite development. An additional factor is that the complex interactions between the parasites and the host’s immune system have co-evolved to create (at least in most cases) a balance between host and parasite. As a result the apicomplexan protozoa are frequently capable of evading host immune responses by targeting privileged intracellular sites, immune evasion mechanisms or antigenic variation during its development [19].Reproduction is critical for perpetuating a species; however, the modes by which cell division can occur are diverse. Fusion of gametes of opposite sex (or mating cell type) to form a zygote is the defining moment of the cellular sexual reproduction. With few exceptions, sexual reproduction appears to be a hallmark of apicomplexan development. During sexual reproduction, genetic recombination occurs, having a significant impact on parasite genomics. Recombination almost always produces novel genotypes, which ultimately results in an increase in genetic variability that may enhance the fitness of the parasites in an ever-changing environment (e.g., facilitating evasion of the host’s immune system and accelerating the spread of mutations conferring resistance to drugs), and reduces the accumulation of deleterious mutations. Previously published studies already proposed that targeting the sexual stages should primarily inhibit the formation of fertile oocysts thus acting as a transmission-blocking vaccine [25]. For example, using the Gam proteins as vaccination to target the oocyst wall formation in Eimeria reduced oocyst shedding in the host, and the vaccines that incorporate surface antigens of the macrogametes of Plasmodium in order to induce antibodies that kill parasites in the mosquito bloodmeal and interrupt parasite transmission through the vector significantly reduced oocyst formation. Monoclonal antibodies against microgametes of E. tenella could reduce oocyst formation [60], HAP2 KO parasites fail to complete fertilization and zygote formation in Toxoplasma [56] and antibodies against HAP2 significantly decreased zygote formation in vitro [144,149,150]. Current research on transmission-blocking vaccines seeks to increase the degree and the durability of functional activity through the combination of different antigens, engineered immunogens, and the optimization of vaccine delivery system and adjuvants. It is required to achieve a sufficient adaptive response to maintain high levels of antibodies over time, as well as broad coverage of antigen variations, to achieve immunity. In addition, transmission-blocking vaccines must have an exceptional safety profile, as they do not confer a direct benefit to the individual. Previous findings indicate that inhibiting either (or both) the fertilization of macrogametes by microgametes (which is necessary for the production of next-generation sporozoites) and the oocyst wall formation can effectively interfere with the parasite’s developmental cycle. Targeting such stages may well be an effective approach to apicomplexan parasite control in the future.

animals : an open access journal from mdpi

10.3390/ani13081353

PMC10135138

Most animals blend well into their natural environment, which protects them from predators. However, in captive environments, animals are often exposed, which can lead to stress. Research suggests that animals are more difficult to detect in complex environments. We tested background preferences in the Gouldian finch, which occurs in two main head color morphs in the wild. Birds were tested in groups of four in either the same head color (black or red) or mixed head color (two black and two red) pairings. One half of the cage had a simple background pattern, the other a complex background pattern. We measured the time spent in front of each background after 10 days in the cage (phase 1), after which backgrounds were swapped (left to right), and assessed preferences again on day 17 (phase 2). Birds preferred the simple background, particularly in phase 1. However, females initially chose the simple background but used both backgrounds in phase 2, whereas changes in males were not significant. Both color morphs preferred the simple background in phase 1, with the black-headed birds moving to the complex background in phase 2. Results indicate that background preferences differ among individuals, which should be considered when designing enclosures.

Most wild animals camouflage well into their environment, providing protection from predators, whereas captive animals often contrast with their background. This can cause stress for the animal, which may perceive it as being exposed. Theory suggests that prey is more difficult to detect in front of complex backgrounds; hence, animals should prefer complex over simple backgrounds. We tested this in the polymorphic Gouldian finch by providing a complex background pattern in one half of the flight cage and a simple background pattern in the other half for 10 days (phase 1). Patterns were then swapped and presented for another week (phase 2). Groups of four birds consisting of either pure black-headed or red-headed or mixed head color (two black-headed and two red-headed) pairings were tested. Gouldian finches spent significantly more time in front of the simple background in phase 1 but not in phase 2. Specifically, females preferred the simple background in phase 1 significantly more than males. Moreover, red-headed birds consistently perched in front of the simple background, whereas black-headed birds used both backgrounds, particularly in phase 2. Results indicate that background preferences differ between sexes and morphs, which should be considered when designing backgrounds. Moreover, natural habitat preferences need consideration.

1. IntroductionOne of the main anti-predator strategies is camouflage, with animals blending into their environment. As most predators use vision to detect prey with a preference for color cues [1,2] (but see [3]), animals’ coloration has evolved to match their background [4,5,6,7,8]. However, in captive environments, there is often a mismatch between an animal’s color and its environment, which can be perceived by the animal as being vulnerable and under stress.Background matching is one of the most common camouflage strategies [8,9,10,11,12,13,14]. Camouflage is achieved by matching the average background [11,15,16] or specific aspects of the background (e.g., leaves) [17] and exploiting the perception and cognitive mechanisms of predators [18]. Reducing the contrast against the background has been shown to decrease detectability and hence predation [6,8,19,20,21]. It can affect population densities with higher densities in areas of better camouflage efficiency [16], support species divergence with recently diverged species showing a close match to the environment [20], and positively affect the invasion success of new habitats with better background matching. Moreover, it can prolong foraging [22], reduce social harassment, or facilitate hidden copulations [23]. While it is usually prey that camouflages to avoid predation, predators also camouflage to get close to prey [24] or avoid mobbing [19].Considerable attention has been paid to color polymorphic species, where individuals in a population differ in their appearance, which can affect their camouflage [25,26]. For example, in the color polymorphic cichlid Amphilophus labiatus, morphs differed in their ability to match backgrounds, which coincided with their frequency [27]. Likewise, in Tawny owls (Strix aluco), the gray morph was less conspicuous against a snowy background and tree trunks, resulting in higher survival in snowy winters [19,21]. In other species, morphs match different backgrounds, affecting their microhabitat selection [28,29].In captivity, animals often contrast against their background. The inability to camouflage with the background has been shown to cause stress and negative thigmotaxis in European cuttlefish (Sepia officinalis) [30] and stress, atypical behavior, and weight loss in African clawed frogs (Xenopus laevis) [31]. When given the choice, hermit crabs (Pagurus bernhardus) prefer shells that match the background [32], desert tortoises (Gopherus agassizii) preferentially rest next to rocks where they are more difficult to detect [10], and Gouldian finches (Chloebia gouldiae) prefer green over white backgrounds, which matches their color better [33]. Both 2D [30,33] and 3D backgrounds (e.g., [31]) are effective. This indicates that welfare is improved in captive animals when they can blend in with the background [31]. However, finding the right background may be difficult given that individuals may differ in their coloration (males vs. females, young vs. adults, color polymorphic species). One solution may be to increase the complexity of the enclosure.Complex environments increase the background noise against which predators must detect prey [18]. While background matching is still important [4,7,23], complexity has been shown to reduce predation [4,34,35,36,37,38,39]. Moreover, Dimitrova and Merilaita [40] found that complex backgrounds reduce predation for matching and non-matching prey, whereas Maend et al. [41] demonstrated that complex backgrounds eliminate size-dependent predation. In a captive environment, Jacky dragons (Amphibolurus muricatus) preferred perches with more complex patterns over simply patterned ones [14]. In contrast, Gouldian finches preferentially perched in front of simple green backgrounds and avoided complex patterned backgrounds, possibly due to their open habitat preference [33].Here we re-investigate preferences for simple and complex backgrounds in the Gouldian finch by controlling for the overall appearance of the backgrounds to account for their open habitat preference. Gouldian finches are color polymorphic in both sexes, with a ratio of about 70% black-headed birds, 30% red-headed birds, and less than 1% yellow-headed birds in the same population [42]. They inhabit the open savannah woodland of Northern Australia [43] and are listed as endangered by the Australian Government, with just 2500 birds left in the wild due to habitat destruction [44]. Gouldian finches are one of the most popular bird species among aviculturists [45], and self-sustaining populations have been maintained in captivity since the Australian export ban in 1960 [46]. Research has shown that the Gouldian finch’s head color indicates their personality. Red-headed birds are more aggressive [47,48,49] but take less risk in potentially dangerous situations and are less explorative of changes in their environment than black-headed birds [49,50]. However, red-headed birds enter unfamiliar environments faster and accept novel food quicker than black-headed birds [51,52]. Earlier experiments on background matching have shown that Gouldian finches prefer a simple green background over a patterned multi-color background [33]. However, it was unclear whether this was potentially affected by the denser appearance of the patterned background. As the Gouldian finch is an open habitat specialist [53,54], they may have chosen the appearance of openness over complexity.In the current study, we controlled for density effects and offered two types of background—a complex and a simple—to investigate (a) whether Gouldian finches prefer complex backgrounds when density effects are controlled for, (b) how preferences may change over time, and (c) how head color, age, and sex affect preferences.2. Materials and MethodsTwenty-four Gouldian finches were part of this study, consisting of 12 females (6 black-headed and 6 red-headed) and 12 males (6 black-headed and 6 red-headed), ranging in age from one to seven years. All birds originated from 11 bird breeders purchased over the years. Birds were housed in groups of six birds mixed in age, sex, and head color in flight cages (120 cm × 80 cm × 100 cm; L × W× H). The cages contained three wooden walls plus wire mesh on the front and ceiling. The cages were furnished with natural twigs and perches. Food and water were attached to the front, with a bath on the floor. The food was a seed mixture of Blattner Amadine Zucht Spezial, Blattner Astrilden Spezial, and Blattner Rote Mannahirse (Blattner Heimtierfutter, Ermengerst, Germany). Blattner bird grit and eggshells were provided in separate pots. Birds were kept at a temperature of 24 °C and 51% humidity and provided with a full spectrum light source with a light:dark cycle of 13:11 h.2.1. Experimental ProcedureExperiments took place in an experimental room adjacent to the housing room, which contained six similarly sized flight cages (120 cm × 70 cm × 100 cm) with three wooden walls and a front and wire mesh ceiling. Cages were arranged so that birds could hear but not see each other and contained two upper perches, left and right, in the cage and a lower perch close to the bottom, which was only used by the birds when going for a bath. Food and water were provided in the middle of the wire mesh front. Each cage had two background patterns. Half of the rear wall and the adjacent side wall were covered with a simple background pattern, and the other half of the rear wall and adjacent side wall were covered with a complex background pattern (Figure 1). Backgrounds were designed in Adobe Photoshop 2018. Both background patterns consisted of vertical stripes alternating between broad (7.9 cm) and narrow (1.9 cm) stripes to control density [34]. The colors were specifically chosen to avoid green, red, and black as the backs of the birds were mostly green and the latter two differed between the color morphs. This allowed specifically testing for simple vs. complex patterns, excluding color matching, which is difficult in this colorful species. As the finches have tetrachromatic vision, including in the ultraviolet range [48], matching background colors to their plumage is further hampered. The simple background alternated broad light blue (#7d8beb) stripes with white stripes, whereas the complex background had a pattern of broad yellow (#e2eb0c), purple (#cd4fff), and brown (#5d4028) stripes separated by narrow white stripes. The lightness of backgrounds was compared using the Lab Color mode in Adobe Photoshop. The simple background had a lightness of 60 for each stripe, whereas the complex background had an overall lightness of 60 calculated as the average of yellow, brown, and purple lightness (90 + 30 + 60/3). The backgrounds were printed on 60 × 40 cm paper and attached to the walls. The position of the complex and simple backgrounds (on the left or right side of the cage) alternated between cages. A video camera was mounted in front of each cage. Overall, four cages were used at a time.The experiments consisted of two phases: phase 1 lasted from day 1 to day 10, after which backgrounds were swapped (left-right) in each cage, and phase 2 lasted from day 10 to day 17. This allowed for controlling of side preferences. Birds were tested in groups of four same-sex individuals of different ages and head color combinations. For each sex, we had one group each of pure black-headed, pure red-headed, and mixed head colors (two birds of each head color). Groups were composed of birds, ideally from different holding aviaries, to mix birds from different groups. As age did not have an effect in the earlier study on background preferences [33], we did not control it. We tested 16 birds (distributed across four cages) in one go (batch one), followed by the remaining 8 birds (batch 2).Birds were released into the experimental cage and allowed to habituate to the new cage environment for nine days. On day 10, birds were video recorded (GeoVision 1480, GeoVision Inc., Taipei, Taiwan) for one hour starting at 9:00 a.m. (end of phase 1). Backgrounds were swapped at 1:00 p.m. the same day, and birds were released back into the cage at 2:00 p.m., with another hour of video recording commencing at 2:00 p.m. (the start of phase 2). A third one-hour recording was performed on day 17 starting at 9:00 a.m. (the end of phase 2), after which the birds were moved back into their housing aviaries. The cages were then cleaned and prepared for the second batch of birds, which were moved in at 2:00 p.m. the same day.2.2. AnalysisAll analyses were performed with SPSS version 26. All data are available in the supplementary Table S1. From the videos, we extracted the total amount of time each individual spent on the perch in front of each background on day 10 (end of phase 1), again on day 10 after swapping the backgrounds (start of phase 2), and on day 17 (end of phase 2). Then the proportion of time spent in front of the simple background in relation to the overall time spent on the two perches was calculated for each of the three recording days using the formula (t(simple) − t(complex)/Σ[t(simple) + t(complex)]). Proportions were arcsine transformed to meet requirements or ANOVAs. Out of the 24 birds, two were excluded from analysis, as one had died after the first day due to unknown reasons and the other one perched primarily on the feeder.In the first step, we evaluated potential side preferences by inspecting the proportion of time spent in front of the simple background at the end of phase 1 (day 10) and directly after the swap at the start of phase 2 (day 10). Out of the six groups tested, all the birds in one group remained on the same side of the testing cage as before (i.e., showing side preferences). This group (consisting of three birds; see above) was excluded from all further analyses.In the second step, background preferences were analyzed (n = 19). The time spent in front of the simple and complex backgrounds was compared on the last day of phase 1 (day 10 before the swap) and the last day of phase 2 (day 17). Here, the actual time spent on each perch was compared rather than the proportions. As the data in phase 2 were not normally distributed, Wilcoxon tests were used.Finally, it was investigated whether preferences changed over time (habituation) and whether this was affected by head color, age, or sex. Arcsine transformed proportions were used for this analysis comparing the proportion of time spent in front of the simple background on the last day of phase 1 (day 10) and the last day of phase 2 (day 17) using repeated measures ANOVA. The within-individual factor was phase (days 10 and 17), whereas head color, sex, and age class (4 classes: 1 year: n = 5, 2 years: n = 6, 3–4 years: n = 6, >4 years: n = 5) were the between-individual factors, which were included as main factors. Age classes were categorized in this way to have meaningful sample sizes. Two-way interactions between the within-individual factor and the between-individual factors are routinely included in the repeated measures ANOVA. Greenhouse-Geiger adjustments were used to account for deviations in the covariance matrix. Based on the outcome, we additionally tested for changes in background preferences between days 10 and 17 within sexes using paired t-tests, which could not be performed as part of the repeated measures ANOVA.2.3. EthicsThe experiments have been in accordance with The Association for the Study of Animal Behavior (ASAB) ethical guidelines [55] and non-invasive in nature. Experiments have been approved by the University Ethics Committee (CMH_RM/2018-12).3. ResultsGouldian finches significantly preferred the simple background over the complex background at the end of phase 1 (day 10; Wilcoxon Signed Rank test: n = 19, U = −2.696, p = 0.007; Figure 2). This preference had disappeared by the end of the second phase (day 17; U = −0.845, p = 0.398; Figure 2), primarily due to increasing variance in the data.The repeated measures ANOVA investigating changes in preference over time and the effects of head color, sex, and age showed that the interaction between habituation x and sex was significant (repeated measures ANOVA: n = 19, F = 7.292, p = 0.018). Females spent significantly more time in front of the simple background at the end of phase 1 (day 10) as compared to the males (t = −2.906, p = 0.012), whereas the sexes did not differ significantly in their background preference from each other at the end of phase 2 (day 17; F = 1.531, p = 0.150; Figure 3). However, the paired t-test comparing preferences on days 10 and 17 for each sex separately indicated that females showed a significant change in preference from spending much more time in front of the simple background in phase 1 to spending similar amounts in front of both backgrounds or even a bit more time in front of the complex background in phase 2 (paired t-test: t12,1 = 2.734, p = 0.019; Figure 3). In contrast, changes in preferences among males were not significant (t7,1 = −1.553, p = 0.171; Figure 3).Moreover, head color had a significant effect (F = 5.187, p = 0.040), with black-headed birds overall showing no clear preference for either background, whereas red-headed birds preferred the simple background. These differences were primarily driven by a trend among the black-headed birds to spend more time in front of the complex background at the end of phase 2 (day 17; t = −1.944, p = 0.074; Figure 4) as compared to the red-headed birds. No significant differences between head colors were present at the end of phase 1 (day 10; t = −0.730, p = 0.478; Figure 4). No other factors had an effect (Table 1).4. DiscussionOverall, Gouldian finches preferred the simple over the complex background, although considerable dynamics were observed among sexes and head colors. Sexes initially differed in their background preference, with males using both backgrounds or slightly preferring the complex one, whereas females clearly preferred the simple background in phase 1. In contrast, the sexes did not differ significantly in phase 2, although females significantly changed their preferences using both backgrounds in phase 2. Likewise, red-headed birds preferred the simple background throughout the study, whereas black-headed birds used both backgrounds, particularly in phase 2.The study specifically tested whether Gouldian finches prefer complex over simple backgrounds. While background matching with respect to colors is an important aspect of camouflage [6,8,19,20,21], we did not try to match background colors with plumage colors due to the potentially different perception of colors by the birds (tetra-chromatic vision [48]). However, the complexity of backgrounds alone has been shown to reduce predation [40]. Nonetheless, the colors might have affected background preferences. The colors in the complex pattern (yellow, violet) resembled plumage colors (belly and breast) more than the blue color of the simple pattern. One might have expected that this would further increase the preference for the complex pattern (better camouflage), which was not the case. We did not measure whether the backgrounds reflected any ultraviolet (UV). If so, birds were expected to settle in front of the background, emitting less UV, as the plumage of Gouldian finches does not reflect UV except for the turquoise band in the neck [48]. While we cannot comment on the potential UV reflectance of the backgrounds, it is unlikely that there was an effect as in both studies (Perkovic and Mettke-Hofmann [33] and this study), the Gouldian finches preferred the simple background despite the very different colors used. It is difficult to imagine that, by chance, the complex background had more UV reflectance in both studies. Finally, any differences in preferences between sexes or morphs are unlikely to be caused by the UV reflectance of the backgrounds, as sexes and morphs do not differ consistently in their UV reflectance [48]. We, therefore, believe that the differences in background preference are primarily due to the complexity of the background patterns. However, future studies should consider UV and possibly different degrees of color matching.Gouldian finches showed a clear preference for the simple background, although this preference became non-significant in phase 2 due to increased variation. This parallels earlier findings, where background density was not controlled for [33]. However, in this earlier experiment, background preferences did not change, suggesting that the density effect of the background is one factor affecting background choice. Nonetheless, controlling for density still resulted in an overall preference for the simple background. Only a few other studies have investigated responses to complex backgrounds, with Jacky dragons preferring complexly patterned perches [14] and least killifish (Heterandria formosa) females choosing complex backgrounds under predation [36]. Gouldian finches live in relatively simply structured open savannah woodland [43], and the simple background seems to capture this better than the complex background. The preference for simple backgrounds aligns with the spatial exploration patterns of the Gouldian finch. They made more approach attempts (reflecting a stronger approach-avoidance conflict) in dense habitats as compared to simple novel environments before entering these habitats, supporting their open habitat preference [51]. This highlights the importance of considering the natural habitat of a species when choosing backgrounds. However, not all individuals consistently preferred the simple background.The sexes differed in their preferences, which were modulated by different phases. In phase 1, males used both backgrounds equally, with a slight preference for the complex background, whereas females showed a clear preference for the simple background. Males are more colorful than females in the Gouldian finch and are likely to be more exposed to predation due to their higher conspicuousness [25,56,57,58]. Choosing a more complex background makes it more difficult for predators to detect them [40,41]. Males might have perceived the complex background as safer, increasing their welfare. The perception of requiring more cover may be particularly strong when new to an environment, as better camouflaged prey is less predated in new environments [20,59]. An indication for this might be that males spent more time in front of the complex background in phase 1 than in phase 2. However, this change was not significant and requires further investigation, giving the birds potentially more time to adapt. One would expect that once fully habituated to their environment, their natural preference for simple habitats would overcome their desire for higher protection. The perception of dangerousness also affected background preferences in killifish. Both males and females preferred matching over non-matching backgrounds, but only when under predation threat [36]. Moreover, females preferred complex over matching backgrounds in more dangerous situations [36].Alternatively, males are more explorative than females in the Gouldian finch [49] and may have explored both backgrounds initially before settling on their preferred simple background. However, the first preference test was performed after 10 days, and one would expect that any exploration of a new environment would have ceased by that time. Interestingly, females started to spend more time in front of the complex background at the end of phase 2. The less conspicuous females may have felt less exposed in the novel environment and settled for their simple natural habitat preference in phase 1. Indeed, females are known to take more risk at waterholes, which was linked to their lower conspicuousness attracting fewer predators [60]. They might have relied on their lower conspicuousness in the current experiment, too. However, over time, they seemed to have explored the complex background, resulting in a significant change in preference with equal usage and even a slight preference for the complex background in phase 2. The contrasting background preferences between males and females indicate different camouflage requirements for males and females, which should be considered in enclosure design. It might also point towards different microhabitat uses in the wild [28,29]. For example, different sexes may prefer different parts of a tree. This would be an interesting aspect to investigate in the wild. Different camouflage requirements have also been shown in other species [36,61].Finally, head color morphs differed in their background preference, with black-headed birds overall showing no preference for either background, whereas red-headed birds clearly preferred the simple background throughout the entire experiment. This is surprising as red-headed birds are more conspicuous [62], and the complex background would provide more camouflage for them [7,23], particularly as the main predators of Gouldian finches are raptors [62]. However, it is consistent with earlier findings in this species, where red-headed birds consistently preferred the simple background [33]. Again, their simple habitat preference seems to override camouflage advantages. Looking at the black-headed birds more closely, they initially also preferred the simple background but changed their preference for the complex background in phase 2. Black-headed birds are more interested in exploring changes in their familiar environment than red-headed birds [49,50]. The swap in backgrounds from left to right may have motivated the black-headed birds more to explore the complex background than the red-headed birds, which resulted in a trend for opposing background preferences at the end. Differences in microhabitat selection that aid their camouflage are well known in color polymorphic species [19,21,28,29] and should be considered when using backgrounds.It is surprising how long the birds needed to explore the complex background, which raises the question of whether this reflects the time to habituate to the environment or their dislike of the complex background. On the one hand, the males in this experiment did not avoid the complex background in phase one, indicating that the complex background is not generally disliked. On the other hand, when this early preference for the complex background reflects the need to camouflage better in an unfamiliar environment, this supports the lengthy habituation. This would be in line with the late exploration of the complex background of the black-headed birds and the females. However, studies on vigilance show habituation to a similar novel environment within four days [63,64], which would speak against such a lengthy habituation. In the current experiment, the only difference in the experimental cages were the added backgrounds. It seems that this 2D change in the environment constituted a major change for the animals, resulting in prolonged habituation. This indicates the importance of backgrounds but also shows that backgrounds must be carefully considered. More research is needed in this area.5. ConclusionsGouldian finches preferred simple backgrounds; however, preferences differed between sexes and head color morphs, with in part opposing trends over time. Changes in preferences were likely linked to risk perception (higher in novel environments in males) and exploration later in the trial (black-headed birds and females). Considering diverting background preferences is an important aspect when designing enclosures. Specifically, a mix of backgrounds could be used to account for different camouflage requirements. Moreover, it would allow birds to choose a background reflecting their current risk perception, thereby reducing fear and increasing wellbeing. However, the results also show that natural habitat preferences require consideration, as simple, open habitat preferences seemed to have overridden the desire to camouflage better. In summary, 2D backgrounds can help improve welfare by providing choice when a mix of backgrounds is used, which addresses changing or differing camouflage requirements. However, some consideration of habitat preferences is needed.

animals : an open access journal from mdpi

10.3390/ani11092553

PMC8471125

Soybean meal and palm oil are important protein and energy sources in European livestock production, respectively. In the next decades, the demand for these feedstuffs is supposed to increase as the world population and its demand for meat and dairy products increases. Alternatives to replace those feedstuffs are necessary. It is necessary to promote low-input, local, and circular production systems: in this sense, adopting feeding systems that use cheaper and local alternative feedstuffs represents a good strategy. One of them is the product obtained after a simple pressing process for the production of biofuels. Cold pressing usually produces oil that can be used as biofuel and a cake rich in fat and with high oil quality but with lower protein content than the conventionally solvent-extracted cakes. Therefore, the availability of cold-pressed cakes can represent an example of integration between the industry and livestock production. The objective of this study was to assess the suitability of formulating cold-pressed rapeseed cake (CPRC) in a dairy cows’ concentrate as a substitute for conventional feedstuffs. Feeding CPRC has the advantage of slightly improving the milk fatty acid profile and consumer acceptance. In conclusion, CPRC can replace conventional feedstuffs without detrimental effects on milk production and composition.

The aim of this trial was to assess the effect of feeding a concentrate including cold-pressed rapeseed cake (CPRC) on productive performance, milk quality and its sensory properties, ruminal biohydrogenation, and bacterial communities. Eighteen cows were paired, and two experimental diets (control vs. CPRC) were distributed within the pair. Concentrates were iso-energetic and iso-proteic and contained similar amounts of fat. The average days in milk, milk yield, and body weight of the animals were (mean ± SD) 172 ± 112 d, 585 ± 26 kg, and 25.4 ± 6.2 kg/d, respectively. The experiment lasted for 10 wk. Feeding CPRC resulted in lower ruminal saturated (p < 0.001) and higher monounsaturated (p = 0.002) fatty acids. Feeding CPRC increased Ruminococcus, Prevotella, and Entodinium but decreased Blautia; p-75-a5; undefined genera within orders Clostridiaceae and RF39 and within families Christensenellaceae, Lachnospiracease, and Ruminococcaceae; and fungi from the phylum neocallimastigomycota. The milk fatty acid profile was characterized by a lower n6:n3 ratio (p = 0.028). Feeding CPRC did not affect the milk yield, milk quality, or fat corrected milk (p > 0.05). Feeding CPRC improved the overall milk acceptability (p = 0.047). In conclusion, CPRC affected some microbial taxa, modified the biohydrogenation process, and improved the milk fatty acid profile and consumer acceptance without detrimental effects on milk production and composition.

1. IntroductionSoybean meal and palm oil are important protein and energy input sources in European livestock production. In the next decades, the demand for these feedstuffs is supposed to increase as the world population and its demand for meat and dairy products increases [1]. The creation of massive plantations to produce either soybean meal [2] or palm oil [3] means that wild forests have been replaced with monocultures. Burning of forests to make space for soybean or palm also has social and environmental impacts [2]. Deforestation removes the local economy of these regions, reduces biodiversity, and pollutes the environment, being recognized as a major contributor to the emission of greenhouse gases [2]. In addition, the impact of the transportation of these feedstuffs from their origin has to be kept in mind in the context of the actual need for more sustainable farming systems.The EU has been supporting farmers to adopt or maintain practices that contribute to fulfill environmental and climate goals through the “green direct payment” (or “greening”), which rewarded, among others, crop diversification. This policy is expected to continue through the agricultural practices included in the eco-schemes of the new CAP reform. In recent decades in Europe, this fact has led to an increase in landing of different oilseeds, such as sunflower or rapeseed. In this context of increased oilseed landing, alternative uses have been proposed. One example has been a simple pressing process for the production of biofuels in local areas using cold-pressing. Cold presses are normally mechanically operated and often involve a screw device that is tightened against the paste to extract the oils. In this process, therefore, neither solvents nor heat are applied to help in the extraction. This process usually produces oil that can be used as biofuel and a cake rich in fat and with high oil quality but with lower protein content than the conventionally solvent-extracted cakes. As a consequence, an innovative local agricultural production chain based on the cultivation of oilseeds has been developed. Livestock production systems in the same area require protein and energy supplements from the market. Therefore, the availability of cold-pressed cakes can represent an example of integration between the industry and local livestock production.Feeding studies with cold-pressed rapeseed cake (CPRC) have shown that its use as a lipid supplement in ruminant diet is susceptible to reducing the extent of ruminal biohydrogenation (BH) and to modifying the ruminal fatty acid (FA) profile towards reduced saturated FA content and increased mono- and polyunsaturated FA levels, as was documented in meat [4], dairy sheep milk [5,6], and cheese [7]. These changes occur with no detrimental effect on the rumen in vitro fermentation process [8,9], apparent digestibility [10], or productive performance [5,6], presenting this cake as a very promising alternative to soybean cake and palm oil. However, there is a lack of studies covering the potential use of CPRC on dairy cow diets’ and its effect on milk production and quality.The effect of dietary lipids on nutrients degradability and digestibility as well as on animal products FA profile is well known. These relationships are mainly mediated by the toxicity of unsaturated fatty acid (UFA) on many microorganisms, especially fibrolytic bacteria [11]. Nevertheless, no recent studies have elucidated the specific effect of CPRC, rich in UFA, on ruminal populations so far. To better understand how CPRC affects the ruminal ecosystem, sequencing methods are promising to be implemented, offering detailed information about the microbial complexity and functionality.Therefore, in the present study, we used locally produced oil-rich CPRC in the formulation of a concentrate for dairy cows and we hypothesize that the oil present in this cake, rich in UFA, could modify the ruminal microbial communities and therefore ruminal biohydrogenation process, leading to a shift in milk FA profile towards an improved n3:n6 ratio. Moreover, we also expect a change in milk sensory characteristics due to the expected more unsaturated milk FA profile [12]. For this reason, the objective of the current trial was to assess the effect of feeding a concentrate including CPRC on productive performance, milk composition and its sensory properties, ruminal biohydrogentation process, and bacterial communities.2. Materials and Methods2.1. Animals and TreatmentsThe trial was conducted at the experimental research farm of Fraisoro Farm School (Zizurkil, Spain). Cows were in loose housing conditions. A total of 18 cows were used; 10 Holstein (H) and 8 Brown Swiss (BS). The average days in milk (DIM), body weight (BW), number of parity, and milk yield of the cows before starting the experiment were (mean ± SD) 172 ± 112 d, 585 ± 26 kg, 1.9 ± 1.1, and 25.4 ± 6.2 kg/d, respectively. Animals were paired by taking into account the breed, number of parity, DIM, and mean milk yield during a 2-week covariate period. At the end of the covariate period, cows were randomly assigned (within pair) to the CTR or CPRC concentrate for the experimental trial. Concentrates were fed for 10 wk; the first 2 wk were for adaptation to the diets, and during the last 8 wk, measurements were taken. Concentrates were formulated to be iso-proteic and iso-energetic and to provide similar amounts of fat, following the NRC [13] recommendations (Table 1). Table 2 shows the FA profile of the two experimental concentrates. Cows within a pair received the same quantity of concentrate (7 ± 1.3 kg/d), but they had ad libitum access to a basal forage ration. Concentrate was offered in individual buckets three times per day.Cows were milked with an automatic milking system (AMS, DeLaval, 2004) machine with free access to the AMS for 22.5 h/d (1.5 h for cleaning of the system). Milking intervals were set to a minimum of 6 h from the previous milking. Nevertheless, if a milking failure occurred, cows would be granted permission to be milked again immediately. During the day, for any particular cow, when the time passed since last milking was more than 12 h, the cow would be forced to visit the AMS.2.2. Sampling and MeasurementsDaily milk production was recorded individually at each milking by the AMS. On the last day of the covariate period and during weeks 2, 4, 6, 8, and 10 of the experimental period, milk samples were taken from the AMS at each milking and stored with azidiol (3.3 mL/L) at 4 °C for fat, protein, and lactose determination (ILL, Lekunberri, Spain). Additional milk samples were taken at each milking on weeks 4 and 9. Then, these milk samples were bulked by animal and day on a milk production basis and were stored at −20 ± 5 °C with azidiol (3.3 mL/L) for FA composition analysis (LIGAL, Mabegondo, Spain). Offered basal forage and orts were sampled on a daily basis, and concentrate feeds were sampled weekly to characterize their chemical composition.In week 9 of the experimental period, rumen samples were collected over two consecutive days for analysis of the FA profile and for DNA extraction for the study of the ruminal microbial community. Sampling was performed at 00:00 and 12:00 h on the first day and at 06:00 and 18:00 h on the second day. Ruminal samples were collected from each dairy cow using a stomach tube (18 mm diameter and 160 cm long) connected to a mechanical pumping unit (Vacuubrand ME 2SI, Wertheim, Germany). The ruminal content was filtered using four layers of sterile gases. For FA profile analyses, a 100 mL pool was made for each cow with 25 mL of the liquid fraction of each ruminal extraction. For DNA extraction in the study of the ruminal microbial community, another 100 mL of each ruminal extraction was saved into a container. All samples were immediately stored frozen at −20 ± 5 °C until analysis.In the last week of the trial, a composite milk sample (36 kg) from each treatment was taken in stainless steel milk cans for subsequent sensory analysis.2.3. Sample Handling and Laboratory Procedures2.3.1. FeedBasal forage and concentrates were dried in a forced-air oven (60 °C/48 h) and ground through a 1 mm sieve. The samples were analyzed for dry matter (method 934.01) and N (method 984.13) content following AOAC [14]. Neutral detergent fiber was determined by the method of [15] with the use of an alpha amylase but without sodium sulfite and was expressed free of ash. The acid detergent fiber, expressed exclusive of residual ash, was determined by the method of [16]. The ether extract content was determined without hydrolysis by the automated soxhlet method (Soxtec System HT 1043 Extraction Unit, Madrid, Spain) using hexane for 6 h as solvent. The starch content was measured by polarimetry [17].Fatty acid methyl esters (FAME) of lipid in both concentrates were prepared in a 1-step extraction-trans-esterification procedure using chloroform and 20 mL/L sulfuric acid in methanol [18]. Methyl esters were separated and quantified with a gas chromatograph (Agilent 7890A GC System, Santa Clara, CA, USA) equipped with a flame-ionization detector, a 100 m fused silica capillary column (0.25 mm i.d., 0.2-μm film thickness; CP-SIL 88, CP7489, Varian Ibérica S.A., Madrid, Spain), and hydrogen as the carrier gas (207 kPa, 2.1 mL/min). The total FAME profile in a 2 μL sample volume at a split ratio of 1:50 was determined using the temperature gradient program described in [18]. Peaks were identified based on retention time comparisons with commercially available standard FAME mixtures (Nu-Chek Prep., Elysian, MN, USA; and Sigma-Aldrich, Madrid, Spain).2.3.2. Rumen Fatty Acid Profile AnalysisRumen FA profile determinations were performed as described in [19]. Briefly, lipid in 200 mg freeze-dried rumen samples was extracted with a mixture of hexane and isopropanol (3:2, vol/vol; [18]) and converted to FAME by sequential base-acid catalyzed transesterification [20]. The total FAME profile was analyzed by gas chromatography using the same chromatograph and temperature gradient program utilized for the analysis of FA in feeds, but isomers of 18:1 were further resolved in a separate analysis under isothermal conditions at 170 °C [18]. Peaks were identified based on retention time comparisons with the same FAME mixtures used for the analysis of feeds and other commercially available standards (from Nu-Chek Prep.; Sigma-Aldrich; and Larodan, Solna, Sweden), cross referencing with chromatograms reported in the literature (e.g., [18,20]), and based on a comparison with reference samples for which the FA composition was determined based on gas chromatography analysis of FAME and gas chromatography-mass spectrometry analysis of corresponding 4,4-dimethyloxazoline derivatives [20].2.3.3. Rumen DNA ExtractionRumen samples were thawed for 10 h at refrigeration temperature (5 ± 3 °C) and squeezed using four layers of sterile gases to separate between solid (particle size smaller than the diameter of the stomach tube) from liquid digesta phases. Liquid digesta phase was separated into planktonic organisms and bacteria associated with the liquid fraction. The solid phase was separated between associated and adherent fractions following the methodology described in [21]. The four fractions obtained were lyophilized and composited to obtain a unique sample with the four fractions represented proportionally (on dry matter basis). DNA extraction was performed using the commercial Power-Soil DNA Isolation kit (Mo Bio Laboratories Inc, Carlsbad, CA, USA) following manufacturer’s instructions. The extracted DNA was subjected to paired-end Illumina sequencing of the V4 hypervariable region of the 16S rRNA [22] and of the V7 region of the 18S rRNA genes. The libraries were generated by means of Nextera kit. The 250 bp paired-end sequencing reactions were performed on a MiSeq platform (Illumina, San Diego, CA, USA). The bacterial and archaeal communities were grouped as OTUs (Operational Taxonomic Units) based on 16S rRNA similarities and protozoal and fungi on 18S rRNA similarities. Data processing was performed using QIIME (v.1.9.0): Quantitative Insights Into Microbial Ecology software package [23]. Sequences were clustered as operational taxonomic units (OTUs) of 97% similarity using UCLUST [24]. OTUs were checked for chimeras using the RDP gold database and assigned bacterial and archaeal 16S RNA taxonomy using the Greengenes database [25], whereas protozoal and fungi 18S rRNA genes were aligned against the 18S SILVA database [26]. Alpha and beta diversity metrics were calculated using the QIIME pipeline.2.3.4. MilkMilk fat, protein, and lactose contents were analyzed by near-infrared spectroscopy (Foss System 4000, Foss Electric, Hillerød, Denmark; ILL, Lekunberri, Spain). To analyze the milk FA profile, milk fat extraction was carried out according to ISO 14156 [27], methylated according to ISO 15884 [28], and analyzed using gas chromatography. The upper phase was injected into a gas chromatograph (Varian 3800) equipped with a capillary column (Cp-sil 88 to over 50 m) and the FID detector. Working conditions were set according to the standard [29]. The carrier gas, nitrogen with a pressure of 14 psi, was used, and the injector temperature was 250 °C. Temperature program proposed by Kramer et al. [30] was used: 4 min at 45 °C, then an increase in temperature of 13 °C per minute up to a temperature of 175 °C (27 min), and an increase in temperature of 4 °C per minute up to 215 °C (35 min).2.3.5. Pasteurized Milk Perceptibility and Sensory PropertiesRaw milk was pasteurized at 72 °C for 30 s using a continuous plate heat exchanger (ATA Tecnología Alimentaria, Irun, Spain). A triangle test was performed to analyze the consumers’ ability to distinguish differences between samples for the attributes of appearance, flavor, odor, texture, and overall acceptability. Forty untrained panelists evaluated four milk samples per treatment in private booths. The panelists were served 2 sets of samples in which the reference was either milk from the CTR or CPRC diet. In every set, one sample was identical to the reference and one was different. For each sample set, the panelists had to identify the sample that tasted the same as the reference. The acceptance test was carried out using a non-trained sensory panel of women and men, regular consumers of cow milk. A 9-point line scale was used, with 1 being the lowest and 9 being the highest score, for each of the measured attributes.2.4. Calculations and Statistical AnalysisMilk fat, protein, and lactose concentrations were calculated as weighted average of milking data: 3.5% fat corrected milk (FCM) was calculated as 0.4318M + 16.23F, with M being milk production (kg) and F being milk fat (kg).For statistical purposes, each cow was considered the experimental unit (n = 18). Milk yield, FCM, milk fat and protein contents, and milk fat and protein yield were analyzed by a MIXED model for repeated measures using the MIXED procedure of SAS software [31], assuming a covariance structure fitted on the basis of Schwarz’s Bayesian information model fit criterion. Yjklmn= µ +Covj+Tk+Pl+Wm+CPn+εjklmn where Y is the dependent variable, µ is the mean values for each treatment, Cov is the initial record (covariate), T is the fixed effect of the concentrate used, P is the fixed effect of the pair, W is the fixed effect of the week (week 3–week 10), C(P) is the random effect of cow within pair, and ε is the residuals. Least squares means for treatments are reported.Rumen FA concentrations were averaged by cow. The rumen and milk FA concentrations were analyzed using the previous statistical model but without considering covariate or repeated measures. The sensorial data (n = 40) were analyzed using the previous statistical model but without considering the effect of the week. Treatment means were separated using a Tukey test except for rumen and milk FA profile, where Bonferroni adjustment was used. Significant effects were declared at p ˂ 0.05.Relative abundances (RA) of bacterial and eukaryote taxa at the phylum, family, and genus level were analyzed using the MIXED procedure [31], according to the following model:Yjkl = µ + Tj + Pk + C(P)l +εjkl where Y is the dependent variable, µ is the mean values for each treatment, T is the fixed effect of the concentrate used, P is the fixed effect of the pair, C(P) is the random effect of cow within pair, and ε is the residuals. Residuals were checked for normality with either the Shapiro–Wilk or Kolmogorov–Smirnov tests using [31], and the data were transformed (log, square-root, and reciprocal transformation) when necessary until the residuals followed a normal distribution.Significant differences between experimental groups’ bacterial and eukaryote community composition were analyzed by analysis of dissimilarity (ADONIS) with 999 permutations. The significant fold changes of the OTUs were tested using DESeq2 [32] and filtered by the false discovery rate value.To investigate the associations between ruminal FAs and bacterial taxa, a regularized canonical correlation analysis (rCCA) was carried out with the package mixOmics (v6.6.2) [33] in R (v3.5.1) [34]. To perform the rCCA analysis, the correlation values between the RA of bacterial genera and each ruminal FA proportions were computed to calculate a similarity matrix. A clustered image map was inferred using a similarity matrix obtained from the rCCA. A threshold of R = 0.45 was used to obtain the relevant components.3. Results3.1. Rumen Fatty Acid CompositionDiet containing CPRC changed the rumen saturated fatty acid (SFA) profile (Table 3). The main changes were significant decreases in the proportions of C12:0 (p < 0.001), C14:0 (p = 0.019), and C16:0 (p < 0.001) and increases in the proportions of C17:0 (p = 0.005), C18:0 (p = 0.007), C19:0 (p = 0.020), C20:0 (p < 0.001), and C22:0 (p < 0.001). The total SFA was lower in the CPRC experimental group (p < 0.001). Escriba aqui la ecuación.The diet containing CPRC resulted in an increase in total rumen MUFA (p = 0.002), cis MUFA (p = 0.028), and trans MUFA (p = 0.002; Table 4). Shifts were mainly characterized by an increase in the proportions of C18:1 cis-9 (p = 0.028), C18:1 trans-11 (p = 0.005), and C18:1 trans-13-14 (p = 0.020) but without changing the C18:1 trans-10/trans-11 ratio (p = 0.377).The total ruminal PUFA contents were not affected (p = 0.829) by the diet containing CPRC (Table 4). The experimental concentrate with CPRC did not affect C18:2 cis-9 trans-11 CLA, C18:2 trans-9 cis-11 CLA, C18:2 trans-11 trans-13 CLA, and C18:2 trans-10 cis-12 CLA contents (p > 0.05). As a consequence, the diet with CPRC did not result in increased CLA proportions (p = 0.430). The proportions of long-chain n-3 PUFA (p = 0.157) and n-6 PUFA (p = 0.803) were not altered in the CPRC experimental group. As a consequence, the diet with CPRC did not alter the n-6:n-3 ratio (p = 0.263).3.2. Ruminal Microbial CommunityThe main phyla were Bacteroidetes (50.7%) and Firmicutes (33.2%). Within the Bacteroidetes, the most abundant families were Prevotellaceae (42.4%), undefined families within the order of the Bacteroidales (4.2%), and (Paraprevotellaceae) (1.2%). The dominant families of Firmicutes were Lachnospiraceae (11.4%), Ruminococcaceae (7.3%), undefined families within order of the Clostridiales (7.3%), and Veillonellaceae (5.5%) (Figure 1).The experimental concentrate with CPRC did not influence the bacterial or Eukaryote species richness as expressed by different diversity indices, such as chao1 or Shannon (Table 5). The beta diversity and the statistical test performed with ADONIS revealed no differences in bacterial (p = 0.186) community and a tendency in eukaryote (p = 0.063) community between experimental concentrates.Among the different bacterial phyla (Supplementary Materials Table S1), the diet containing CPRC only significantly decreased Tenericutes (p = 0.038) and tended to decrease Plantomycetes (p = 0.056).At the family level (Supplementary Materials Table S2), the diet with CPRC decreased RA of the undefined families within the order Clostridiales (p = 0.016), Christensenellaceae (p = 0.033), and the undefined families within the order RF39 (p = 0.024) and tended to decrease RA of Lachnospiraceae (p = 0.08), Ruminococcaceae (p = 0.08), and Pirellulaceae (p = 0.055).At the genus level (Supplementary Materials Table S3), the diet with CPRC increased RA of Ruminococcus (p = 0.047) and decreased RA of Blautia (p = 0.045); p-75-a5 (p = 0.013); undefined genus within the orders Clostridiales (p = 0.016) and RF39 (p = 0.024); and those within families Christensenellaceae (p = 0.033), Lachnospiraceae (p = 0.011), and Ruminococcaceae (p = 0.034) and tended to decrease Clostridium (p = 0.098), Shuttleworthia (p = 0.0506), Pyramidobacter (p = 0.072), and undefined genus within the family Pirellulaceae (p = 0.055).At the OTU level, the OTU belonging to the genera Prevotella and Ruminococcus were enriched when the animals were fed the experimental concentrate with CPRC, whereas in the ruminal content of the animals fed the CTR concentrate, an enrichment in OTU of undefined genera within the order Clostridiales was observed (Figure 2).The most abundant Eukaryote phyla (Figure 3) in both experimental groups (CPRC and CTR) were Ciliophora (42 and 34%), Ascomycota (33% and 32%), and neocallimastigomycota (4 and 11%).Among the different Eukaryote phyla (Supplementary Materials Table S1), diets with CPRC only significantly decreased neocallimastigomycota (p = 0.016) compared with the control (Figure 4).Regarding the genera belonging to Ciliophora phylum, diets containing CPRC only significantly increased the RA of genus Entodinium (p = 0.039; Table 6). At the OTU level (Supplementary Table S4), some OTUs belonging to the genera Entodinium, undefined genus within subclass Trichostomatia, and Ophryoscolex were enriched in the rumen of animals fed the experimental concentrate with CPRC, whereas in the ruminal content of the animals fed the CTR concentrate, enrichments in OTUs of the undefined genus within subclass Haptoria, Ophryoscolex, and undefined genus within family Neocallimastigaceae were observed.The associations between rumen FA and bacterial taxa were represented by a clustered image map (Figure 5) inferred from the rCCA analysis. Genera Ruminococcus, Anaerovibrio, Butyrivibrio, Bulleidia, Methanosphaera, SHD231, Mogibacterium, and Methanobrevibacter and undefined genera within the families Veillonellaceae and Coriobacteriaceae were positively correlated with the total MUFA and some BH intermediates (C18:1 trans4, C18:1 trans-5, C18:1 trans 6-7-8, C18:1 trans-9, C18:1 trans-10, C18:1 trans-12, C18:1 trans13-14, C18:1 trans-15, C18:1 trans-16, C18:1 cis-11, C18:1 cis-9, C18:2 trans-11 cis-15, and C22:1 cis13) while these FA were negatively correlated with the genera Clostridium, Succiniclasticum, Lachnospira, Blautia, Pyramidobacter, Pseudobutyrivibrio, and Coprococcus; undefined genera within order Clostridiales and RF39; and undefined genera within families S247, Ruminococcaceae, and Lacnospiraceae.Genera Clostridium, Succiniclasticum, Lachnospira, Blautia, Pyramidobacter, Pseudobutyrivibrio, and Coprococcus; undefined genera within orders Clostridiales and RF39; and undefined genera within families S247, Ruminococcaceae, and Lacnospiraceae were positively correlated with total SFA, concretely with C12:0, C14:0, C16:0, and 13-oxo C18:0, while these FA were negatively correlated with genera Ruminococcus and Anaerovibrio and undefined genera within family Veillonellaceae.Genera Prevotella, Treponema, YRC22, CF231, Ruminococcus, and Anaerovibrio and undefined genera within families Succinivibrionaceae, Paraprevotellaceae, and S247 were positively correlated with long chained saturated fatty acids, concretely with C17:0, C18:0, C19:0, C20:0, and C22:0, while these FA were negatively correlated with genera p-75-a75, Butyrivibrio, Clostridium, Coprococcus, Blautia, and Shuttleworthia; undefined genera within families Chistensenellaceae, RF16, Ruminococcaceae, and Lacnospiraceae; and undefined genera within orders RF39 and Clostridiales.3.3. Milk Fatty Acid CompositionThe proportions of most short and medium chain SFA were not modified by the dietary treatments (Table 7), except for C13:0, which was increased in the milk fat of CPRC-fed cows (p = 0.043).The diet with CPRC did not modify the milk proportions of C18:1 cis-9 (p = 0.628), C18:1 cis-11 (p = 0.427), or C18:1 trans-11 (p = 0.650) but increased C18:1 trans-6 (p = 0.001), C18:1 trans-10 (p = 0.034), and C18:1 trans-12 (p = 0.043). The total MUFA (p = 0.495), cis MUFA (p = 0.633) and trans MUFA (p = 0.062) were not modified when the CPRC diet was fed.As shown in Table 8, feeding a diet with CPRC did not affect the total PUFA in milk (p = 0.625). The use of CPRC did not affect milk fat C18:2 cis-9 trans-11 CLA (p = 0.834) or total CLA (p = 0.711) but reduced C18:3n-6 (p = 0.043) and increased C18:3n-3 (p = 0.008) and C20:1n-9 cis-11 (p < 0.001) proportions. The milk ratio of PUFA:SFA did not differ between treatments (p = 0.507), whereas the n6:n3 ratio was lower in the CPRC experimental group (p = 0.028).3.4. Milk Yield and Milk CompositionThe milk composition in terms of crude fat (p = 0.100), crude protein (p = 0.203), or lactose (p = 0.556) proportions did not differ in the experimental group fed a diet with CPRC compared with the control group (Table 9). Similarly, feeding a diet with CPRC did not affect the yields of milk (p = 0.304), FCM (p = 0.679), crude fat (p = 0.633), crude protein (p = 0.616), or lactose (p = 0.485).3.5. Pasteurized Milk Perceptibility and Sensory PropertiesIn the triangle test, consumers were able to differentiate between the milk of the CTR and CPRC groups (p < 0.001). Feeding a diet with CPRC enhanced the overall acceptability by 0.43 points out of 9 (p = 0.047) and by improving the flavor by 0.52 points out of 9 (p = 0.021). Appearance, odor, or texture were not perceived being as different (p > 0.05; Table 10)4. DiscussionThe proportion of total SFA and specifically short/medium-chain FA (C12:0, C14:0, and C16:0) in the ruminal liquid mimicked that of the diets and is in agreement with the changes observed in other in vitro studies using CPRC [9]. Moreover, feeding a concentrate with CPRC induced some relevant changes related to the ruminal FA BH process. Although the total SFA decreased in rumen fluid in diets with CPRC, the C18:0–C22:0 proportions increased. The main FAs present in the experimental concentrates were C18:1 cis-9 (23.4 vs. 41.0 g/100 g FA for CTR and CPRC, respectively) and C18:2 cis9 cis12 (35.0 vs. 32.7 g/100 g FA for CTR and CPRC, respectively). These FA were subjected to a BH process in the rumen carried out by ruminal bacteria that ended up in the formation of C18:0 [35]. The higher proportion of ruminal C18:0 found with CPRC can be due to the fact that the CPRC diet provided greater amounts of C18 UFAs compared with the CTR diet. Other authors have observed the same trends using CPRC [9] and cold-pressed sunflower cake, also rich in C18 UFAs [36].Plant lipid sources rich in UFA have been related to an increase in the C18:1 production in the rumen [37,38]. The extent of rumen BH of C18 UFAs is known to vary between 58 and 100% [39]. However, it is important to highlight that the final reduction step of UFA to C18:0 is considered rate limiting, and therefore, C18:1 intermediates (mainly C18: trans-11) can accumulate and flow out of the rumen, mainly when excessive amounts of UFA are ingested [40,41]. Considering the higher ruminal proportions of total MUFA, especially C18:1 cis-9, and total trans-MUFA, especially C18:1 trans-11, this seemed to be the case when feeding a diet with CPRC in the present study.Moreover, an effect of the type of fat present in the CPRC on the microorganisms involved in the BH process cannot be precluded. Huws et al. [42] proposed that uncultured bacteria belonging to genera Anaerovorax, Prevotella, Lachnospiraceae Incertae Sedis, Ruminococcus, Butyrivibrio, Pseudobutyrivibrio, Tanerella, unclassified Bacteroidales, Clostridia and Clostridiales, Ruminococcaceae, Lachnospiraceae, Prevotellaceae, and Porphyromonadaceae might be implicated in ruminal C18:1 trans-11 formation. Other authors also stated that other genera including genus Ruminococcus, as one of the most prevalent in the rumen, are involved in ruminal C18: trans-11 formation [43]. In this sense, we observed an increase in the RA of genera Ruminococcus and some OTUs of genus Prevotella with CPRC, whereas Clostridium and the undefined genus within family Lachnospiraceae RA decreased with CPRC. However, although C18:1 trans-11 ruminal concentrations increased with diets containing CPRC, no direct relationship of any specific bacterial genus was observed with C18:1 trans-11 in the present study. Although bacterial species involved in ruminal C18:1 trans-10 formation are not well known, some authors observed ruminal formation of this FA by Ruminococcus albus [43,44]. In agreement with these observations we observed an increased RA in genus Ruminococcus in the CPRC experimental group, and this genus presented a positive relationship with C18:1 trans-10 concentrations in the rumen contents in the clustered image map.Regarding the last step of ruminal BH, although Butyrivibrio proteoclasticus is the only bacterial species known to reduce C18:1 FA to C18:0 [35,45], non-cultivated Butyrivibrio, Pseudobutyrivibrio, and other unknown Lachnospiraceae strains could play a role in the final BH step [46]. In the current study, only the RA of Blautia (family Lachnospiraceae) and undefined genera within family Lachnospiraceae were decreased in ruminal contents of cows fed a diet with CPRC. Furthermore, there was a negative correlation between these taxa and trans C18:1 intermediates, potentially suggesting that these genera were involved in ruminal 18:0 formation though minor BH pathways [47]. The RA of the order RF39 was also decreased in cows fed with CPRC and was negatively correlated with 18:1 isomers, which agrees with the results observed in [48] when supplementing a fat rich in PUFA to goats. These authors hypothesized that genera within this order might also be implicated in ruminal 18:0 formation, a hypothesis that is also supported by our results. Another alternative explanation is that feeding CPRC reduced the biohydrogenating activity of B. proteoclasticus instead of its RA, but to test this hypothesis, metatranscriptomic assays should be performed and are far from the objective of the present study.The use of a diet with CPRC seemed not to affect the first steps of the BH pathway of C18:2 in the rumen, since neither the main intermediate C18:2 cis-9 trans 11 CLA proportion nor the proportions of other minor alternative intermediates proportions were altered in the rumen [43]. This may be explained by the great extent of ruminal BH that happens with linoleic acid (up to 95%; [49]). For the C18:3 BH process, none of the main intermediates of the first stages of the BH process seemed to be affected by the diet with CPRC. However, some alternative pathways seemed to be affected. Dewanckele et al. [43] showed that a minor intermediate pathway for BH of C18:3 in the rumen was the hydrogenation and isomerizarion to C18:2 cis-12,cis15 and C18:2 trans-12,cis-15 and the posterior hydrogenation to some C18:1 isomers (C18:1 cis-11, C18:1 cis-12, C18:1 trans-12, C18:1 cis-15, C18:1 trans-15, and C18:1 trans-16), which were finally hydrogenated to C18:0. We observed that some of these intermediates (C18:1 cis-11, C18:1 trans-12, C18:1 cis-15, and C18:1 trans-16) increased in the rumen of cows of the CPRC group. This would be related to an inhibitory effect of the lipids present in the CPRC on the last step of BH of these FA to C:18:0. As mentioned before, some unknown Lachnospiraceae strains might play a role in the final BH step [46]. In this sense, we observed a negative relationship of these intermediates with the RA of Blautia (family Lachnospiraceae) and the undefined genera within family Lachnospiraceae, and we also observed that the RA of these genera was decreased in the ruminal contents of cows fed a diet with CPRC.The contribution of protozoa and fungi in the rumen to the BH process has been reported as negligible and mainly associated with activity of protozoa ingested bacteria [50,51]. However, it is recognized that rumen protozoa contain proportionally more UFA than rumen bacteria and thus could play an important role in increasing CLA or C18:1 trans-11 ruminal proportions and can contribute in a significant way to the flow of UFAs leaving the rumen [52,53,54]. This is in agreement with the increased RA of some rumen ciliates and the increased ruminal C18:1 trans-11 concentration in the ruminal content of CPRC experimental group. Conversely, some authors reported decreased ciliate protozoa when rapeseed oil was included in the diet of sheep [55]. However, the level of inclusion and the physical form of the fat supplement could play a role in the effect towards protozoa population. In addition, Newbold et al. [56] suggested that, although high dietary lipid concentration is toxic to protozoa, the antiprotozoal effect of fat depends on the FA composition, with medium chain FA being more effective in reducing ciliates than PUFA.In the present paper, increasing the dietary UFA content with the use of CPRC in the cow ration did not have a great effect on microbial populations diversity (alpha and beta diversity indices) but led to changes in some bacterial and eukaryotic taxa. However, CRPC partially replaced other ingredients in the concentrate. In this regard, differences not only in the FA profile but also in the chemical composition of both concentrates evaluated cannot be ignored and might also contribute to explain some subtle differences in ruminal microbial populations. It was observed that these changes could modify the BH process in the rumen. However, changes observed in ruminal FA profile had a slight reflect on milk FA profile. Opposite to those changes observed in rumen contents, a diet with CPRC did not reduce proportions of total SFA in milk, probably due to a compensation of the observed lower short chain FA in the rumen with de novo synthesis of these FA in the mammary gland. This is in agreement with the results observed by other authors on sheep milk [6] but differs from the results observed by [5] with sheep and by [57] with dairy cows and with the idea that including long chain UFA in the ration decreases milk short and medium chain FA through inhibition of de novo synthesis in the mammary tissue [58,59]. The abundance of C18:1 cis-9 and PUFA in plant lipids is known to alter the distribution of trans FA in milk fat [60], and in agreement with our results, supplementation with canola or rapeseed has been previously related to the increases in milk trans FA concentrations [57,61]. However, other authors have observed no changes [58,62]. This inconsistency could be partially explained by the physical form of the fat supplement. Givens et al. [58] observed that the physical properties of the rapeseed supplement were crucial to observing important changes in the milk FA profile. While rapeseed oil or rapeseed milled increased milk C18:1 isomers (cis and trans), diets containing whole rapeseeds resulted in minor changes, highlighting the key role of the bioavailability of lipids.Although we observed an increase in ruminal proportions of C18:1 trans-11, which is known as the main precursor of C18:2 cis-9 trans-11 CLA synthesis in mammary tissue, the proportion of these CLA isomer was not increased in the milk of the CPRC group. Pascual et al. [6] also observed no effect of feeding CPRC on milk C18:2 cis-9 trans-11 CLA proportions, but these authors observed a clear increase in the milk C18:1 trans-11 proportions. Moreover, this is in disagreement with previous studies where rapeseed-based feeds increased the milk proportions of C18:2 cis-9 trans-11 CLA and other CLA isomers [57,58,62]. Regarding n3 FA, our results agree with previous studies that have pointed out that supplementing with CPRC, rich in C18:3-n3, increases milk long-chain n-3 FAs [5,6].Finally, regarding production performance, no detrimental effects of using CPRC as a UFA rich lipid source in dairy cow rations was observed, which is consistent with other studies with dairy sheep [5,6] or beef cattle [4]. Moreover, the changes observed in the milk FA profile did not affect the milk sensory quality in a negative manner. In dairy rations, one key factor for the practical use of new feedstuffs, especially those rich in lipids, is to ensure that the final product’s taste remains pleasant and free of off-flavors. In the present study, as mentioned, not only was there no negative effect but a better flavor and overall acceptability was observed for milk from CPRC-fed cows compared with the control. Flavor is known as one of the key attributes for acceptability, and among the variables affecting milk flavor, fat is pointed out as one of the most important ones [63], so even slight changes observed in the milk FA profile seemed to be enough to affect milk flavor in a positive way. Other authors have observed no effect of feeding CPRC on sheep curd [6] or cheese [7] sensory properties, whereas the authors of [64] observed similar results in dairy cattle milk when feeding cold-pressed sunflower cake rich in UFA.This study provided a new insight into the effects of using CPRC as an alternative lipid supplement in dairy cows’ diets on ruminal BH of dietary FA and ruminal microbial communities and how the changes exerted in the rumen influence productive performance, milk FA profile, and milk sensorial quality.5. ConclusionsIn conclusion, a diet with CPRC affected some microbial taxa at the rumen level, modified the fatty acid biohydrogenation process, and resulted in a slight improvement in the milk fatty acid profile and consumer acceptance without detrimental effects on milk production and quality.

animals : an open access journal from mdpi

10.3390/ani11051455

PMC8159077

South American camelids (SACs) constitute the greatest livestock wealth of the Andean populations. Approximately half a million people from the high Andean areas are dedicated to the breeding of SACs as their main activity. In general, infectious diseases, particularly diarrheal infections, cause high morbidity and mortality in offspring and adult animals. In the study, we demonstrated that multiple virus pathogens circulate among neonatal SACs, and coinfections from other viruses might be common among SAC crias. We also demonstrated, for the first-time anywhere, the circulation of mammalian orthoreovirus in SACs or camelids. Diarrheal infections can potentially impact livestock productivity, which translates into serious economic losses for the Peruvian SAC industry, especially within rural communities, directly impacting their livelihood. Better knowledge of the infections that afflict these animals will enable the implementation of measures for the prevention and control of pathogens, the results of which will ultimately be reflected in improving the quality of life of these communities.

Enteric infections are a major cause of neonatal death in South American camelids (SACs). The aim of this study was to determine the prevalence of enteric viral pathogens among alpacas and llamas in Canchis, Cuzco, located in the southern Peruvian highland. Fecal samples were obtained from 80 neonatal alpacas and llamas and tested for coronavirus (CoV), mammalian orthoreovirus (MRV), and rotavirus A (RVA) by RT-PCR. Of the 80 fecal samples analyzed, 76 (95%) were positive for at least one of the viruses tested. Overall, the frequencies of positive samples were 94.1% and 100% among alpacas and llamas, respectively. Of the positive samples, 33 (43.4%) were monoinfected, while 43 (56.6%) had coinfections with two (83.7%) or three (16.3%) viruses. CoV was the most commonly detected virus (87.5%) followed by MRV (50%). RVA was detected only in coinfections. To our knowledge, this is the first description of MRV circulation in SACs or camelids anywhere. These data show that multiple viruses circulate widely among young alpaca and llama crias within the studied areas. These infections can potentially reduce livestock productivity, which translates into serious economic losses for rural communities, directly impacting their livelihoods.

1. IntroductionSouth American camelids (SACs) constitute the greatest livestock wealth of Andean populations. The raising of alpacas and llamas is a crucial economic activity in Andean regions, enabling the production of fibers; primarily those of alpacas, which have a high value in international markets because of their fine texture. The llama, due to its size and physical strength, is also used as a pack animal and plays an important role in rural transportation. Peru possesses three million alpacas, the largest alpaca population in the world, and has a herd of approximately one million llamas. Most of these animals are found in the departments of the southern highlands, particularly in Puno and Cuzco. Over 80% of Peru’s alpacas and the entire population of llamas are kept by small producers who lack adequate infrastructure for conducting both production and marketing activities of their products [1,2,3]. According to the Peruvian Ministry of Agrarian Development and Irrigation [1], Cuzco is the second most important center for alpaca and llama breeding in Peru, with the province of Canchis being the main producer. Unfortunately, neonatal mortality rates among alpacas and llamas in this province are 30% and 25%, respectively [1]. Diseases cause great losses due to both mortality and decreased productivity. Enteropathies are a major cause of neonatal death in SACs. The most significant pathogens causing diarrhea in SACs are Coccidia, Cryptosporidium spp., Giardia spp., Salmonella spp., Escherichia coli, Clostridium spp., rotavirus, and coronavirus [4,5,6,7,8,9,10,11]. Despite the vital importance of alpaca and llama breeding to the Peruvian economy, studies on enteric infections, particularly viral infections, in SACs are incipient. A better understanding of the epidemiology of enteric infections is essential to develop preventive measures. The aim of this study was to determine the prevalence of enteric viral pathogens—specifically coronavirus (CoV), mammalian orthoreovirus (MRV), and rotavirus A (RVA)—associated with infections among alpacas and llamas in Canchis, Cuzco, located in the southern Peruvian highland.CoV and RVA have already been identified as important diarrheal pathogens in neonatal SACs. On the other hand, MRV has not been associated to enteric infection. However, in a previous study, during the characterization of RVA strains detected in feces of alpacas and llamas collected in February 2014 in the province of Canchis in the state of Cusco, Peru, three viral strains containing 10 segments of double-stranded RNA (dsRNA) suggestive of MRV were isolated in an MA-104 cell culture [12]. In this study, we describe the characterization of these strains and investigate the prevalence of this virus in these animals.2. Materials and Methods2.1. MRV Isolation and CharacterizationThe protocol used for isolating MRV in cultures of African green monkey cells (MA-104) has been described elsewhere [12]. Viral culture supernatants were examined using polyacrylamide gel electrophoresis (PAGE) followed by silver nitrate staining [13]. Based on the PAGE profile suggestive of the MRV genome, specific primers were designed and used for amplification and sequencing of the MRV sigma 1 gene (Supplementary Table S1). The amplified genomic segments were sent for sequencing to Macrogen Inc. (Seoul, Korea). Overlapping sequences were assembled and edited using SeqMan, EditSeq, and MegAlign in the Lasergene software package (DNASTAR, Madison, WI, USA). Phylogenetic analysis was performed with MEGA software (version 7.0.14). Dendrograms were constructed using the maximum likelihood method based on the GTR-G model. Statistical significance was estimated by bootstrap analysis with 1000 pseudoreplicates. The sequences were compared to reference MRV strains obtained from GenBank (https://www.ncbi.nlm.nih.gov/nucleotide/, accessed on 5 April 2021). Sequences generated in this study were deposited in GenBank under the accession numbers MN200219, MN200220, and MN200221.2.2. Fecal SamplesEighty fecal samples were collected from neonatal alpacas (n = 68) and llamas (n = 12), all one to five weeks old, with (n = 43; 35 alpacas and 8 llamas) and without (n = 37; 33 alpacas and 4 llamas) diarrhea, from three camelid breeding areas of the districts of Marangani (Silly and Quisini) and Sicuani (Pataccalasaya), province of Canchis, department of Cuzco, Peru. Samples were collected during the birthing season of alpacas and llamas, in January and February 2015. Fecal samples were collected directly from the rectum of animals and kept at −20 °C until processing at the Laboratory of Veterinary Virology and Immunology, Universidad Nacional Mayor de San Marcos, Lima, Peru.2.3. Viral RNA ExtractionThe extraction of viral RNA samples required an initial clarification process. Fecal suspensions were prepared at 10–20% in PBS (pH 7.2) and then centrifuged at 2500× g for 5 min. Supernatant was filtered using a 0.22 μM diameter filter. Viral RNA was extracted from clarified fecal supernatants using TRIzol™ LS Reagent (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s recommendations.2.4. Viral Detection and IdentificationSamples were submitted to RT-PCR and nested PCR using specific primers (Table 1). Reverse transcription polymerase chain reaction (RT-PCR) and nested PCR were performed using the GoScript™ reverse transcription system and GoTaq® Green Master Mix (Promega, Madison, WI, USA), respectively. The genomic RNAs were subject to one cycle of reverse transcription (5 min at 25 °C followed by 45 min at 42 °C), and one step of 2 min at 95 °C followed the cycles of PCR. For CoV testing, the primers Cor-FW and Cor-RV, which amplify a 251-bp fragment of the RNA-dependent RNA polymerase (RdRp) gene of any CoV, were used for the first round of PCR, according to the protocol described previously [14]. Five microliters of the amplicons produced in the first round of PCR were used for BetaCoV detection, using the primers Beta.CoV.F and Beta.CoV.R, to generate a 227-bp fragment of polymerase-encoding gene. The PCR conditions were as follows: 3 min at 95 °C; followed by 35 cycles of PCR, each consisting of 40 s at 94 °C, 1 min at 57 °C, and 40 s at 72 °C; and the final extension step for 10 min at 72 °C. BetaCoV-positive samples were subsequently analyzed by nested PCR for the identification of the subgenus Embecovirus (formerly known as BetaCoV lineage A, BetaCoV 1, or bovine-like CoV (BCoV-like)), using the first-round PCR products and specific CV2U and CV2L primers [15] that amplify all Embecovirus strains except for HCoV-HKU1 (AY597011) and ChRCoV (KM349742), with a predicted product of 136 bp. The PCR conditions were the same as described above for BetaCoV. MRV detection was performed by one round of RT-PCR using the primers MRV-FM and MRV-RM. The primers used for MRV RT-PCR detection target the conserved genomic segment that encodes the RdRp. Selection was carried out in silico by aligning all MRV genomes deposited on GenBank, representatives of all serotypes (MRV1-MRV4), with a predicted product of 181 bp. The specificity of the primers was tested by amplifying the cell-cultured MRV strains and confirming the results by sequencing analysis. Sequences generated from the RdRp 181-bp amplicons were also deposited in GenBank under the accession numbers MN200216, MN200217, and MN200218. The PCR conditions were as follows: 3 min at 95 °C; followed by 40 cycles of PCR, each consisting of 40 s at 94 °C, 1 min at 45 °C, and 1 min at 72 °C; and the final extension step for 5 min at 72 °C. RVA was detected using primers that amplify the NSP5-encoding gene, as described previously [12,16]. All primers were synthesized by IDT (Integrated DNA Technologies, Coralville, IA, USA). Reactions were performed in a Bio-Rad T100 PCR thermal cycler (Bio-Rad Laboratories, Hercules, CA, USA). The PCR products were separated by 1.5% (w/v) agarose gel electrophoresis, stained with ethidium bromide, and visualized under UV light. A 100-bp DNA ladder (Promega, Madison, WI, USA) was used to determine molecular size.To validate the PCR assays, positive controls were used for each of the studied viruses, which consisted of viral strains isolated in cell culture whose identification was previously confirmed by sequencing analysis. Strains AlpCoV-SA44 and AlpCoV-HN (GenBank accession numbers KX266949 and KX266944, respectively), both BetaCoV, subgenus Embecovirus, were used as CoV-positive controls. For MRV, strains SA44-Alpaca, H8-Alpaca, and SL7-Llama (GenBank accession numbers MN200216, MN200217, and MN200218, respectively), identified as MRV-1, were used. RV strains RVA/Alpaca-tc/Per/ SA44/2014/G3P(30) (GenBank accession number KT935485) and RVA/Human-wt/Per/FRNM/2014/G3P(8)(14)(40)(50) (GenBank accession number KY972105), belonging to rotavirus species A (RVA), were used as positive controls.3. Results3.1. MRV Characterization by PAGE and SequencingAnalysis of the migration profile of the dsRNA genome of three viral strains (SA44, H8, and SL7) isolated in MA-104 cells revealed an electrophoretic migration profile compatible with MRV (Figure 1). A sequence of 1069 bp of the viral attachment protein sigma 1-encoding gene was obtained for the alpaca (SA44 and H8) and llama (SL7) strains. They showed a nucleotide identity of 96.7% to 99.3% with bat strains, 96.7% to 96.9% with human strains, and 96.5% with mink MRV strains. Phylogenetic analysis identified those strains as serotype 1 (MRV1) (Figure 2).3.2. Virus DetectionOf the 80 fecal samples, 76 (95%) were positive for at least one of the screened viruses. Overall, the frequencies of positive samples were 94.1% and 100% among alpacas and llamas, respectively (Table 2). Positive samples were detected in all three sites of collection. Viral infections were observed among both symptomatic and asymptomatic animals, with no statistically significant difference (p = 0.1048). Specifically, among the 68 samples from alpacas, 26 were from animals with diarrhea and all tested positive. Of the 42 samples from alpacas without diarrhea, 38 tested positive. Among the 12 samples from llamas, 9 and 3 were from animals with and without diarrhea, respectively, and all tested positive.Of the positive samples, 33 (43.4%) were monoinfected, while 43 (56.6%) had coinfections of two or three viruses (Table 2). The community with the highest rate of coinfections was Silly, at 69.7% (23/33) among alpacas and 41.7% (5/12) among llamas, followed by Quisini and Pataccalasaya, at 47.8% (11/23) and 33.7% (4/12) among alpacas, respectively (Table 2). CoV was the most commonly detected virus in all studied communities, being present in 95.3% (n = 29) of monoinfections and 95.3% (n = 41) of coinfections. MRV was found in 12.1% (n = 4) of monoinfections and 83.7% (n = 36) of coinfections. RVA was only found in coinfections (37.7%; n = 16) (Table 2).3.3. Identification of CoV Genus and SubgenusOf the 70 samples positive for CoV, 66 (94.3%) were identified as BetaCoV. Of these, 16 (22.9%) belonged to the subgenus Embecovirus, with the majority of these samples (15/16) obtained from alpacas. The genus was not identified for 4 (5.7%) and 71.4% of CoV and BetaCoV samples, respectively (Table 3).4. DiscussionThere are few reports on enteric viral infections in SACs, particularly in Peru [6,8,9,10,17,18]. The majority of published studies were conducted in research centers with semi-intensive production systems with good nutritional and health management. Our study focused on small producers in rural communities with extensive production systems and inadequate health management training, who are responsible for more than 80% of the SAC herd in Peru. In Peru, the SAC breeding area is geographically located over 3800 m above sea level. This presents an extreme climate and poor forage for animal species that are not typical of the Andean plateau, although it is possible to observe the rare presence of traditional livestock (cattle, sheep, horses, and pigs). An estimated one and a half million people from the high Andean areas were engaged in the breeding of SACs as their primary economic activity in the year 2000. However, the per capita income in these camelid-producing areas is the lowest in the country [3,19,20].The economic conditions of peasant communities such as those studied here, which are dedicated to the breeding of SACs, are chaotic due to extreme poverty. Their situation is becoming even more critical due to the lack of drinking water, vital hygienic services, and healthcare. These shortages are reflected in the health of their animals, which suffer greater disease burdens due to inadequate handling and lack of prophylactic programs. The animals are managed in extensive systems that use swampy areas watered by streams or rivers. Infectious and parasitic diseases are major limiting factors in the production of these animals. In general, infectious diseases cause high morbidity and mortality in both offspring and adult animals, which translate into serious economic losses for rural communities, directly impacting their livelihood [4,6,21].Our data showed a wide circulation of the surveyed viruses among alpacas and llamas in the three studied communities. CoV and RVA have already been identified as important diarrheal pathogens in neonatal SACs [6,10,17,22]. MRV, on the other hand, has been associated primarily with mild respiratory and enteric infections in mammals. Yet in the last decade, MRV has been associated with upper respiratory tract infections, encephalitis, and diarrhea [23,24,25,26,27,28,29]. In a previous study, we accidentally isolated three viral strains whose genome presented an electrophoretic profile similar to MRV [12]. The characterization of these isolates confirmed their identification as MRV1, with a high nucleotide identity with MRV1 strains from bats, humans, and mink (Figure 2). Therefore, we investigated the presence of MRV in new fecal samples of alpacas and llamas in an attempt to determine whether the detection of these strains was a rare event or whether this virus circulated frequently among these animals. In this study, MRV was the second most prevalent virus, detected in 50% of samples. In the Americas, this virus has been associated with diarrhea in calf and deer in the USA [24,27] and in humans in Brazil [30]. To our knowledge, this is the first description of MRV circulation in SACs or camelids anywhere. The high prevalence of infected animals suggests that this virus is well adapted to the environmental conditions of the Peruvian Andean highlands and local SAC herds.Previous studies conducted in the Andean areas of the Peruvian highlands have shown the circulation of CoV in alpacas on farms with semi-intensive systems and in communities with extensive systems [6,8,22], with infection rates of 18.3% to 53.3%. On the other hand, in alpacas reared under semi-intensive systems in research centers, CoV was detected in 26.8% of animals [17]. The variation in rates observed in different studies probably reflects differences in methodologies, the breeding system employed, and the sanitary conditions of each location. However, it is clear that this virus is widespread in Peruvian alpaca herds. Studies on CoV in llamas have not been reported in Peru. However, previous studies have disclosed CoV infections in outbreaks of diarrhea in llamas in the USA [7,21]. We observed a CoV infection rate of 91.7% among these animals. Nucleotide sequences of alpaca CoV strains exhibited high identity with circulating bovine strains (EmbeCoV, formerly BetaCov 1), suggesting a possible bovine origin of these viruses [10,16,30,31]. The majority of CoV strains detected in alpacas (72.7%) and llamas (90.1%), although belonging to the Betacoronavirus genus, do not belong to the Embecovirus subgenus. These data suggest the circulation of different genera and subgenera of CoV among these animals.RVA has already been associated with outbreaks of diarrhea among alpacas and llamas in the Peruvian highlands with high mortality rates [6,9,22,32]. Interestingly, in this study, RVA was only detected in coinfections. High rates of coinfections involving CoV and RVA have been described among SACs previously [6,22].Virus-positive samples were detected among animals with and without diarrhea. The occurrence of asymptomatic infections in these animals could be due to the persistence of maternal immunity, or to the fact that as these pathogens are circulating widely in the herd, animals could have been infected before the study period and may have developed immunity, thus not presenting with diarrhea. Asymptomatic shedding is a source of contamination favoring the environmental persistence of these pathogens.One of the most impressive findings of this study was the extremely high prevalence of coinfections (56.6%). This can be explained by (i) the high sensitivity of the detection technique used in the study, (ii) the high circulation of these viruses among herds, and (iii) inadequate sanitary conditions for managing the animals, allowing the dissemination of pathogens. Unfortunately, we do not have information that allows us to infer the impact of coinfections on clinical presentations. However, it has been demonstrated that coinfections probably increase the severity of diarrhea in alpacas [6].5. ConclusionsThe raising of alpacas and llamas in Peru is conducted in a particular ecological niche by rural communities and is characterized by the close contact of shepherds with animals as well as failures in sanitary and prophylactic programs due to lack of economic resources. These factors facilitate interspecies transmission of viruses such as CoV, MRV, and RVA, triggering possible zoonotic or anthropozoonotic infections [32,33,34]. Consequently, epidemiological surveillance is essential to prevent and control the emergence or re-emergence of new viral genotypes and variants with zoonotic potential.

animals : an open access journal from mdpi

10.3390/ani13111805

PMC10252018

This study investigated the effect of low molecular weight sodium alginate (LMWSA) on the growth and health of witheleg shrimp (Litopenaeus vannamei). Further, we tested whether LMWSA can alleviate the negative impacts of cadmium. This study showed that this additive could improve the feed conversion ratio (FCR) and antioxidant parameters. While cadmium suppressed the antioxidant system parameters in the control group, these parameters were not decreased in those fed dietary LMWSA.

Decreasing low molecular weight can improve the digestibility and availability of ingredients such as sodium alginate. This study aimed to test the four dosages of low molecular weight sodium alginate (LMWSA) (0%: Control, 0.05%: 0.5 LMWSA, 0.10%: 1.0 LMWSA, and 0.2%: 2.0 LMWSA) in whiteleg shrimp (Litopenaeus vannamei) (3.88 ± 0.25 g) for eight weeks. After finishing the trial, shrimp were exposed to cadmium (1 mg/L) for 48 h. While feed conversion ratio (FCR) improved in shrimp fed dietary 2.0 LMWSA (p < 0.05), there was no significant difference in growth among treatments. The results showed a linear relation between LMWSA level and FCR, and glutathione S-transferase (GST) before; and malondialdehyde (MDA), glutathione (GSH), GST, and alanine transaminase (ALT) after cadmium stress (p < 0.05). The GST, MDA, ALT, and aspartate transaminase (AST) contents were changed after stress but not the 2.0 LMWSA group. The survival rate after stress in 1.0 LMWSA (85.23%) and 2.0 LMWSA (80.20%) treatments was significantly higher than the Control (62.05%). The survival rate after stress negatively correlated with GST and ALT, introducing them as potential biomarkers for cadmium exposure in whiteleg shrimp. Accordingly, the 2.0 LMWSA treatment had the best performance in the abovementioned parameters. As the linear relation was observed, supplementing more levels of LMWSA to reach a plateau is recommended.

1. IntroductionFreshwater scarcity has been one of the most significant barriers to aquaculture development [1]. As a result, mariculture and coastal aquaculture have emerged as promising candidates for providing food for the next century [2]. This type of aquaculture now accounts for more than 55% of total aquaculture output (122 million tonnes) in 2020 [1]. Whiteleg shrimp (Litopenaeus vannamei) was the most-produced shrimp species in 2020, with 5.8 million tonnes [1]. This shrimp species dominates the production of coastal aquaculture and is an important source of foreign exchange income in many developing countries [1]. One of the most important themes for transforming Asian aquaculture (which produces more than 88% of total production) is biosecurity and disease control. The best suggestion has been the stocking of postrave shrimp sourced from pathogen-free broodstock and using novel feed additives to increase animal resistance to challenges [1]. Therefore, any additive that can improve growth and shrimp resistance against any challenge can be steps toward aquaculture sustainability. There are many challenges in the continuous development of this species’ aquaculture, such as heavy metal pollution, most importantly, cadmium (Cd).Cd pollution has become a serious global environmental concern since this metal can be easily accumulated in the food chain and aquaculture production tissues and eventually negatively affects human health [3]. This heavy metal is relatively more soluble than others, causing high mobility in the food webs, and can exist for a long term in an environment as it is non-biodegradable [4]. Many studies have reported that Cd pollution in shrimp tissues in different parts of the world, such as the Persian and Oman Gulfs [5,6], Turkey [7], Vietnam [8,9], Bangladesh [10], China [11,12], India [13,14], and Mexico [15,16]. In terms of aquaculture species, Cd can negatively affect shrimp health and growth [17,18,19]. Some other studies tested different supplements to alleviate Cd toxicity. For example, probiotics in Nile tilapia (Oreochromis niloticus) [20], turmeric (Curcuma longa), and black pepper (Piper nigrum) in African catfish (Clarias gariepinus) [21], chitosan, or vitamin C alone or in a combination in common carp (Cyprinus carpio) [22], Chinese parsley (Coriandrum sativum) in rainbow trout (Oncorhynchus mykiss) [23], and onion (Allium cepa) in Nile tilapia [24] declined this metal toxicity.Alginic acid is a naturally occurring, edible polysaccharide found in brown algae, and its combination with sodium makes sodium alginate. Some studies have used this additive to improve growth and immunity in aquatic species as a prebiotic [25] and boost the antioxidant system [26,27,28,29,30,31,32]. Further, sodium alginate has been used to improve broodstock reproductive performance, larval survival [33], shelf life [34], and fish sperm preservation [35]. Lower molecular weight increases solubility and fermentation [36] and is potentially more accessible and digestible for aquatic species. Few studies tested low molecular weight sodium alginate (LMWSA) in fish to improve immune parameters, resistance against bacteria, and growth [25,37,38,39].To the best of our knowledge, no study tested the LMWSA on shrimps; further, no study in aquatic species tested its alleviation effect on heavy metal toxicity. We hypothesised that this additive could improve shrimp response to heavy metal exposure by boosting shrimp health. Therefore, this study was designed to examine how LMWSA can improve growth performance, survival rate, body composition, antioxidant response, and haemolymph parameters of whiteleg shrimp.2. Material and Methods2.1. Animal Ethical StatementThe national ethical framework for animal research in Iran and its guidelines [40] that were adopted from the Declaration of Helsinki (1975) and the Society for Neuroscience Animal Care and Use guidelines (1998) approved this study [40] to optimise handling and minimise animal stress.2.2. Experimental DietsThe LMWSA supplement for this experiment was provided from Thailand, and the process of making LMWSA was explained elsewhere [41]. The commercial diet (Beyza Feed Mill 21 (Beyza 21 Manufacturing Company, pellet size:1.8–2.2 mm)) was powdered. We added powdered LMWSA to diets in four dosages, including Control, 0.5 LMWSA (0.5 g LMWSA per kg diet), 1.0 LMWSA (1 g LMWSA per kg diet), and 2.0 LMWSA (2 g LMWSA per kg diet). These dosages were chosen based on earlier studies in fish [25,37,41], as no study has been done in shrimps. After the milled diets became homogeneous by adding warm water, the resulting mixture was compressed by a meat grinder (Electrokar EC-1, Tehran, Iran) to form pellets with a 2 mm diameter. Then, pellets were spread out on a tray and dried in an oven to ≥90% dry matter at 60 °C for 24–48 h. After drying, the feeds were packed in suitable packages and kept at 4 °C [2]. The chemical compositions of experimental diets are presented in Table 1.2.3. Shrimp and Husbandry TrialThis experiment was done at the Laboratory of Aquatic Research (Persian Gulf University, Bushehr, Iran). Post larvae shrimp in stage 12 were purchased from a local farm and were fed with starter diets (Beyza Feed Mill 21 (Beyza 21 Manufacturing Company, Shiraz, Iran), pellet size: 1.8–2.2 mm)) for 20 days. Then, shrimp was distributed to experimental tanks and fed with a Control diet for two weeks. Two hundred and forty whiteleg shrimp (3.88 ± 0.25 g) were stocked into 12 tanks (300 L) (20 shrimp per tank, triplicate). Tanks were filled with filtered and disinfected (chlorine, 10 ppm) seawater (40 ± 0.6 ppt), and about twenty percent of water was exchanged daily. The average temperature and pH were 30.0 ± 1.1 °C, 7.5 ± 0.5, ammonia-nitrogen 0.09 ppm, and the natural photoperiod was applied. During eight weeks of the feeding trial, shrimp were fed with the diets three times a day (8:00, 13:00, and 18:00 h) at 3% of body weight. 2.4. Growth PerformanceAt the end of the feeding trial, all shrimp were fasted for 24 h and were then anesthetised with ice-cold water. Growth and feeding performances were evaluated by the following parameters: Specific growth rate (SGR) (% day−1) = 100 × (Ln [mean final body weight] − Ln [mean initial body weight])/time (days)) Weight gain (WG) (%) = 100 × ([mean final body weight − mean initial body weight]/mean initial body weight) Daily weight (g/day) = (mean final weight (g) − mean initial weight (g))/time (day)Feed conversion ratio (FCR) = dry feed intake (g)/wet weight gain (g) Survival was calculated as follows: Survival (%) = 100 × (final shrimp number/initial shrimp number)2.5. Biochemical Composition AnalysisThe analysis of the proximate composition of shrimp was performed using the AOAC standard methods [42]. Briefly, the dry matter was measured gravimetrically after oven drying of homogenized samples for 24 h at 105 °C (AMB50; ADAM, Milton Keynes, UK). Crude protein (N × 6.25) was determined by the Kjeldahl procedure using an automatic Kjeldahl system (BÜCHI, Auto-Kjeldahl K-370; Flawil, Switzerland). Crude lipid was determined by ether extraction using Soxhlet (Barnstead/Electrothermal, Knutsford, UK), and ash content was determined after incineration in a muffle furnace (Finetech, Shin Saeng Scientific, Paju-si, Gyeonggi-do, Republic of Korea) at 550 °C for 6 h.2.6. Haemolymph CollectionSeven shrimp from each tank were sampled to measure antioxidant parameters and serological enzymes. Hemolymph was collected directly from the cardiac sinus of shrimp with sterile syringes and transferred to centrifugal tubes on ice. Tubes were maintained at 4 °C overnight and then centrifuged at 1500× g rpm for 10 min at 4 °C. The hemolymph supernatant serum was collected and used for further analysis. 2.7. Antioxidant Enzyme Activity Malondialdehyde Evaluation and Serological EnzymesFor evaluating the activity of liver antioxidant enzymes, hepatopancreas was quickly dissected and washed in ice-cold phosphate buffer (pH = 7.4) and immediately frozen in liquid nitrogen, then stored at −80 °C until homogenate preparation. Hepatopancreases were homogenised in ice-cold 100 mM phosphate buffer by using a homogeniser for 30–45 s. The tube was then centrifuged at 12,000× g for 30 min at 4 °C. Supernatants were collected and stored at −80 °C [43]. Glutathione S-transferase (GST) was measured by the absorbance increase at 340 nm, resulting from the conjugation of reduced glutathione (GSH) and 1- chloro-2,4-dinitrobenzene as described by [44]. The formation of malondialdehyde (MDA) was determined via the thiobarbituric acid method [45] with some modification. Briefly, to 100 µL homogenate, we added 1400 µL of 15% trichloroacetic acid dissolved in hydrochloric acid (0.25 N) and then added 14 µL of 2% butylated hydroxytoluene in methanol and mixed well. The mixture was heated in a 100 °C water bath for 15 min, then cooled to room temperature and centrifuged at 12,000× g for 5 min. Absorbance was measured in the supernatant at 532 nm, and we calculated the MDA concentration in samples based on the standard curve. Results expressed as nanomoles MDA formed per milligram protein. The GSH values were measured with the Ellman method [46]. For each sample, 100 μL of the sample was mixed with dithiobisnitrobenzoate (DTNB) and PBS. After an incubation of 5 min, the absorbance was read at 412 nm. The value of GSH was expressed as nmol/mg protein. The concentrations of haemolymph activities of aspartate aminotransferase (AST) and alanine transaminase (ALT) were determined using an automatic microplate reader (Synergy 2 Biorad) and Pars Azmun kit (Tehran, Iran).2.8. Statistical AnalysisThis experiment was conducted in a completely randomised design with four treatments and three replications. All data were analysed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA). Normality and homogeneity of variance were tested initially using the Kolmogorov–Smirnov and Levene tests, respectively. For providing a comprehensive analysis, we applied orthogonal polynomial contrasts to determine if the LMWAS level had linear and/or quadratic relations with measured parameters [47]. Further, two-way ANOVA was done with the effect of diet and stress. When the interaction was significant, we unpacked the original data. When the interaction was not significant, we compared the main effects in pooled data). The data before and after stress was compared with an independent sample t-test. Data are presented as means ± standard deviation, and differences were considered to be significant at p < 0.05.2.9. Cd Challenge TestAfter eight weeks, ten shrimp from each experimental tank was distributed to Aquarium (40 L) for the challenge test. Acute toxicity of Cd as medium lethal concentration (LC50) values for whiteleg shrimp after 48 h was 1.30 mg/L [48]. For this experiment, we selected 1 mg/L as the Cd level for 48 h challenge. The solutions of metal were prepared with CdCl2 (Sigma-Aldrich Co., St. Louis, MI, USA) dissolved in distilled water to obtain a stock of 1 mg/L solution. The experimental metal mixture solution was obtained by adding the appropriate volume of each stock solution to the tanks. After finishing the exposure, hemolymph taken from the ventral sinus of five shrimp/aquariums was sampled with a 1 mL sterile syringe for further analysis. Shrimp were not fed during exposure. Shrimp survival percentages were recorded daily, and dead organisms were immediately removed from the aquaria.3. Result3.1. Growth Performance, Survival Rate, and Proximate CompositionThe FCR was significantly higher in the control group than in those fed the 2 g LMWSA/kg diets. Results show that FCR has a linear relation with LMWSA levels in the diets (Table 2) (p < 0.05). There was no significant difference in growth performance and survival rate among groups. Results indicated that ash content had a quadratic relation with LMWSA levels in diets (p < 0.05) (Table 3). However, there was no significant difference in protein, lipid, and moisture contents among groups.After Cd stress, the control group had the most mortality rate (p < 0.05), and there was a quadradic relation between LMWSA level and survival rate (Table 4).3.2. Antioxidant ActivitiesThere was a linear and quadratic relationship between GST and LWMSA levels, and with increasing its content in diets, GST went up (p < 0.05) (Figure 1). The results of two-way ANOVA indicated that the effect of stress on GSH was significant, and GSH decreased by stress (Table 5). The effect of diet on GSH was significant, and those fed dietary 2.0 LWMSA had higher levels compared to 0.5 LWMSA and control treatments. The interaction effect for MDA and GST was significant, and the original data was unpacked (Figure 1). Before stress, there was no significant difference in MDA levels. After stress, all parameters were changed, and GST and MDA decreased with increasing levels of LWMSA in diets (p < 0.05). The most important results were that the Control group had significantly higher values of GST and MDA parameters after stress than before, while the same results were not observed for shrimp-fed dietary LWMSA in dosages more than 0.1% (p < 0.05).3.3. Haemolymph EnzymesThe results of two-way ANOVA indicated that the effect of stress on AST was significant, and this parameter increased by stress. The interaction effect for ALT was significant, and the original data was unpacked (Figure 2). Results indicate before stress, there were no significant differences in hepatopancreas ALT enzyme activities. However, after stress, whiteleg shrimp fed dietary LMWSA had a lower value of these enzymes compared to the Control (p < 0.05) (Figure 2). With stress, this parameter had higher values in the control and 0.5 LWMSA groups but not others.4. Discussion4.1. Growth Performance, SurvivalR, and Proximate CompositionThe growing demand for novel supplements in aquaculture has attracted the attention of researchers to find alternative additives. Low molecular weight polysaccharides have recently been introduced as novel prebiotics [49,50]. Many studies on plant polysaccharides found that low molecular mass or hydrolysed oligosaccharides improved colonic persistence and increased fermentability by gut microflora [51,52,53,54].Although several studies investigated the effect of sodium alginate on improved growth, immunity, and health of shrimps [28,29,30,33,55,56], no investigation tested the effect of the size of this supplement on animal performance. In the few studies available in aquaculture, tilapia-fed dietary LMWSA had higher growth performance, immunity, and resistance against bacterial challenge compared to the control [41]. The result of the present study indicated that whiteleg shrimp FCR has a linear relation with LMWSA levels in the diets (Table 2) (p < 0.05); therefore, more levels should be tested. The highest level in this study was 0.2%, while in tilapia, 0.3% of this supplement positively affected growth. However, the growth and SGR of whiteleg shrimp fed dietary LMWSA were not positively affected by this supplement. The reason might be that different species respond differently, and higher dosages should be tested to reach a plateau. Another possible hypothesis is that probably digestive system and microbiota of shrimp were not improved by changing the molecular weight. Unlike this study, some works showed improved growth by feeding animals with normal sodium alginate [56]. More research is required to determine the optimum dosage of this supplement in aquatic species.There was no significant difference in survival rate among groups showing that LMWSA did not positively or negatively affect this parameter. Additionally, it can be said that shrimp were farmed in good condition, and the survival rate was higher than 89% in all treatments. After Cd stress, there was a quadradic relation between LMWSA level and survival rate; and the Control group had the most mortality rate (p < 0.05). This result clearly shows that this supplement improved whiteleg shrimp’s ability to tackle Cd stress. Similarly, other studies indicated that adding this supplement to tilapia diets improved the post-challenge survival rate against Streptococcus agalactiae [41]. In the present study, the higher survival rate might be due to the improvement of antioxidant defenses by the LMWSA. Similarly, sodium alginate improved the survival rate of common carp against Edwardsiella tarda infection [57].While both internal (age, gender, and size) and external factors (water quality, season, and geographical location) affect the proximate body composition of aquatic species, the diet is most likely responsible for most of the changes [58]. The results of this study indicated that ash content had a quadratic relation with LMWSA level (p < 0.05) (Table 3). However, there was no significant difference in protein, lipid, and moisture contents. No change in the proximate composition of Malaysian Mahseer (Tor tambroides) with feeding sodium alginate was observed [32]. Decreasing protein, lipid, and ash contents by feeding tilapia with sodium alginate was also reported [59]. Further, when sea bream (Sparus aurata) was fed sodium alginate, lipid contents in the body were elevated. As was observed, there was a wide variety of responses in different species with feeding sodium alginate, and it is hard to make any solid conclusion. 4.2. Antioxidant ActivitiesMeasuring parameters such as MDA, GSH, and GST can be reliable markers of the antioxidant system and eventually shows the health status of animals. The Scopus database shows that more than 1900 articles have used these parameters to monitor aquatic species’ health in the last ten years. Imbalanced oxidative activities can bring superoxide and H2O2 radical damage, which antioxidant enzymes protect cells from occurring [60]. Cd toxicity causes oxidative stress in aquatic species [18,61,62]; therefore, measuring this parameter helps to understand how shrimp was affected by this stress. There is no study on shrimp related to the effect of sodium alginate on antioxidant parameters. However, when Asian sea bass (Lates calcarifer) were fed a 1% LMWSA diet, GST and MDA levels were significantly increased [26], reflecting improvement in the general health status of fish. Previous studies proved that most prebiotics stimulates the synthesis of the antioxidant enzymes such as glutathione. For example, crude polysaccharides [63], Ganoderma lucidum polysaccharides [64] in shrimps, and galactooligosaccharide in rainbow trout [65] resulted in this improvement. The many prebiotics’ effects on the antioxidant system were reviewed elsewhere [66,67,68,69]. The results of current data align with those studies and show that this supplement improved the antioxidant system of shrimp and caused less change in this system after Cd stress. The improved antioxidant system was in line with a higher survival rate in 1.0 LMWSA and 2.0 LMWSA groups (Table 4). Sodium alginate is a well-known strong antioxidant [70,71] and we observed this effect clearly in our study. More studies are required on shrimp to illustrate the potential improvement of the antioxidant system by LMWSA.4.3. Haemolymph EnzymesSerological enzymes such as AST and ALT are frequently examined to monitor the physiological status of aquatic species under different stressful or nutritional situations. Consistent with the growth data, there was no specific trend or change in ALT and AST enzymes, illustrating that the shrimp was in good condition. When shrimp were exposed to Cd stress, ALT and AST levels in the Control group elevated, but not in other treatments showing that those fed supplements had more stability in these enzymes. The survival rate in the control group was lower as well, and it can be hypothesised that if animals can control and have fewer changes in the antioxidant system and serological enzymes, they are more likely to able to cope with stress better and have a higher survival rate as was observed in LMWSA groups. More studies are required in this area to properly understand how liver enzymes can be changed by stress in shrimps. Many past studies have reported increased liver enzymes in response to various stresses [72,73,74,75,76]. More research is required to demonstrate the effect of Cd stress on the physiological status of shrimps.4.4. Correlation between Measured ParametersIn the present study, a positive correlation between SGR and moisture contents (60%) and a negative correlation with ash (−68%) in the body was observed (p < 0.05) (Table 6). It can be hypothesised that the higher weight of the shrimp was due to higher moisture content. Further, FCR had negative correlations with LMWSA level (−58%) and GST (−70%) (p < 0.05). A negative correlation means lower FCR, showing that this supplement positively affected FCR. Further, the LMWSA level had positive correlations with GST (64%) and GSH after stress (71%); and a negative correlation with MDA after stress (−67%), GST after stress (−73%), and ALT after stress (−67%) (p < 0.05). It shows that LMWSA strongly affected the antioxidant system and serological enzymes in shrimp. These enzymes correlated with each other as well. For example, GSH after stress had negative relationships with MDA (−59%) and GST (−58%) (p < 0.05). Antioxidant parameters greatly correlated with serological enzymes showing that these two physiological systems are closely related to each other in whiteleg shrimp. In this way, GST after Cd stress had a positive correlation with ALT (84%) and AST (72%) (p < 0.05). MDA after Cd stress also had a positive relation with ALT (59%). Similarly, the same trend in these data was observed when whiteleg shrimp were fed Cd-polluted diets [77] and oxidised fish oil [78]. Interestingly, the survival rate after Cd stress had a negative correlation with enzymes such as MDA (−50), GST (−68%), ALT (−59%), and AST (−37%), which for GST and ALT were significant (p < 0.05). It is further evidence that these parameters can be indicators of survival rate after stress in whiteleg shrimp. The same trend was observed earlier, and stressed fish had higher values of these parameters [79,80].5. ConclusionsConclusively, LMWSA did not increase the growth rate but improved feed conversation efficiency. Regarding growth performance, the survival rate after Cd stress, antioxidant response, and serological enzymes, shrimps fed dietary 2.0 LMWSA had the best performance. No alteration in MDA, GST, ALT, and AST before and after Cd stress for the 1.0 LMWSA and 2.0 LMWSA groups caused shrimp to have a higher survival rate than the Control. After the Cd challenge, lower MDA, GST, ALT, and AST values were observed in the 2.0 LMWSA group. As there was a linear relationship between these parameters and LMWSA levels, supplementing more levels of this additive to diets is recommended to reach the optimum level. Further, the effect of LMWSA on Cd bioaccumulation in shrimp should be tested.

animals : an open access journal from mdpi

10.3390/ani11061829

PMC8235084

It has been reported that a high crude protein diet could reverse the diet-induced lipid accumulation in the liver of mice and rodents. However, in vivo data supporting a functional role of a high crude protein diet on hepatic lipid metabolism-associated genes and proteins in weaned piglets is not available. In the present study, we aimed to provide a mechanistic insight into alterations in the hepatic lipid lipogenesis, lipolysis, oxidation, and gluconeogenesis in response to different dietary crude protein levels. Our results demonstrated that dietary crude protein could regulate hepatic lipid metabolism through regulating hepatic lipid lipogenesis, lipolysis, oxidation, and gluconeogenesis. The result indicated an important role of dietary crude protein in regulating hepatic lipid metabolism in weaned piglets.

Amino acids serve not only as building blocks for proteins, but also as substrates for the synthesis of low-molecular-weight substances involved in hepatic lipid metabolism. In the present study, eighteen weaned female piglets at 35 days of age were fed a corn- and soybean meal-based diet containing 20%, 17%, or 14% crude protein (CP), respectively. We found that 17% or 20% CP administration reduced the triglyceride and cholesterol concentrations, while enhanced high-density lipoprotein cholesterol (HDL-C) concentration in serum. Western blot analysis showed that piglets in the 20% CP group had higher protein abundance of hormone-sensitive triglyceride lipase (HSL) and peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α), as compared with other groups. Moreover, the mRNA expression of sterol regulatory element binding transcription factor 1 (SREBPF1), fatty acid synthase (FASN), and stearoyl-CoA desaturase (SCD) were lower in the 17% or 20% CP group, compared with those of the piglets administered with 14% CP. Of note, the mRNA level of acetyl-CoA carboxylase alpha (ACACα) was lower in the 17% CP group, compared with other groups. Additionally, the mRNA level of lipoprotein lipase (LPL), peroxisome proliferator-activated receptor alpha α (PPARα), glucose-6-phosphatase catalytic subunit (G6PC), and phosphoenolpyruvate carboxykinase 1 (PKC1) in the liver of piglets in the 20% CP group were higher than those of the 14% CP group. Collectively, our results demonstrated that dietary CP could regulate hepatic lipid metabolism through altering hepatic lipid lipogenesis, lipolysis, oxidation, and gluconeogenesis.

1. IntroductionThe liver, a pivotal organ in the metabolism, plays a central role in regulating lipid homeostasis [1,2]. The functional role of liver in lipid metabolism is to absorb free fatty acids from blood circulation, hereafter free fatty acids were re-esterified into triglycerides (TG), and then were secreted back into the blood circulation within triglyceride-rich very-low-density lipoprotein [3,4]. Therefore, intrahepatic lipid levels are predominantly affected by a balance of lipid metabolism, including lipogenesis, lipolysis, oxidation, secretion, and gluconeogenesis [5]. Hepatocytes, the main liver parenchymal cells, could synthesize triglyceride from fatty acids (FA) and store triglycerides primarily in the form of lipid droplets [6]. Excessive accumulation of TGs within the liver leads to the onset or progress of liver metabolic syndromes, such as nonalcoholic fatty liver disease (NAFLD), nonalcoholic steatohepatitis (NASH), and fibrosis [7], which are becoming a global health problem in adults and children [8]. An understanding of the underlying mechanisms responsible for TG accumulation in the liver would benefit the development of therapeutic strategies for patients affected by the liver diseases [6]. Of note, accumulating evidence has demonstrated that nutritional status especially dietary crude protein (CP) level could alter liver lipogenesis, lipolysis, oxidation, transportation, and gluconeogenesis, and contribute to physiological functions [9,10].Animals fed a high CP level diet is associated with excretion of nitrogen, which is a major contributor to environmental pollution [11]. Therefore, low CP diets supplemented with crystalline amino acids have been extensively investigated as an effective strategy to reduce product cost, nitrogen excretion, and diarrhea in weaned piglets [12,13]. Over the past two decades, growing studies have prompted us to investigate the effects of low CP diet supplemented with or without amino acids on growth performance, intestinal metabolism, and gut microbiota [14,15,16]. Additionally, previous study has shown that high dietary CP significantly reduce adipocyte size, fat percentage, and backfat thickness in finishing pigs [11]. However, in vivo data supporting an impact of dietary CP level on hepatic lipid metabolism in weaned piglets, an animal model for studying human nutrition, are limited [17].Amino acids serve not only as building blocks for proteins synthesis, but also as substrates for the synthesis of low-molecular-weight substances [18]. Indeed, emerging evidence showing that amino acids play a key role in preventing or ameliorating tissue TG accumulation in humans and mice [19,20]. For instance, L-arginine stimulates the expression of peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α), nitric oxide synthase, heme-oxygenase, and AMP-activated protein kinase (AMPK). Moreover, L-arginine enhances lipolysis, the oxidation of glucose and fatty acids, and inhibits fatty acid synthesis at the whole-body level in rats [21,22]. Additionally, isoleucine has been reported to reduce TG concentrations in liver and muscle, and markedly enhance the mRNA expression of peroxisome proliferator-activated receptor alpha (PPARα) and uncoupling protein (UCP) in mice [20]. Also, relatively high CP diets supplemented with branched chain amino acids (BCAAs) have been reported to reduce the incidence of obesity and related metabolic disorders in rats [23,24]. In mammals, BCAAs are substrates for the synthesis of monomethyl branched chain fatty acids (mmBCFAs), which might implicate in and contribute to outcome of a lean liver [25].Based on these findings, the objective of this study was to investigate the effect of CP level on the serum TG, cholesterol (CHO), and hepatic lipid metabolism-related gene expression in weaned piglets.2. Materials and Methods2.1. Piglets and Experimental DesignThe experimental procedures were approved by Institutional Animal Care and Use Committee of China Agricultural University. A total of 18 crossbred healthy female piglets (Duroc × Landrace × Yorkshire) were weaned at 28 days of age. After 7 days of adaptation, piglets with the initial body of 9.57 ± 0.64 kg were randomly assigned to one of three groups (n = 6/group). Piglets were fed a corn- and soybean meal-based diet with different dietary CP levels (14%, 17%, and 20%). The piglets had free access to feed and water throughout the 45-day experimental period. The diets were formulated based on nutrient requirements of National Research Council (NRC; 2012) for the piglets. Additionally, lysine, methionine, tryptophan and threonine were supplemented to diets. Other amino acids and EAA/NEAA ratio of diets were detected and shown in Table 1.2.2. Serum and Liver CollectionAt the end of 45-day trial, jugular venous blood samples were obtained from each piglet. Serum was obtained in centrifugation 1500× g for 10 min after incubation at room temperature for 1 h. Serum was separated and stored at −80 °C in tubes for further analysis [26]. Then, piglets were killed by euthanasia as previous described [10]. Liver was quickly isolated and stored at −80 °C for the later analysis. From each liver, 18 tissue samples (each approximately 25 mm × 25 mm × 15 mm) were collected to represent left medial.2.3. Triglyceride, Cholesterol, HDL-C, and LDL-C DeterminationConcentrations of triglyceride and cholesterol in liver and serum were determined by using a commercial assay kits from Nanjing Jiancheng Biochemistry (Catalog no. A110-1-1 and A111-1-1; Nanjing, China). Concentrations of high-density lipoprotein cholesterol (HDL-C) and Low-density lipoprotein cholesterol (LDL-C) in serum were measured by using a commercial assay kits from Nanjing Jiancheng Biochemistry (Catalog no. A112-1-1 and A113-1-1; Nanjing, China).2.4. Urea DeterminationConcentrations of urea in serum were determined by using a commercial assay kit purchased from Nanjing Jiancheng Biochemistry (Catalog no. C013-2-1; Nanjing, China) based on previous describe [27].2.5. Quantitative Real-Time PCRTotal RNA was isolated from the liver by using the Trizol reagent (CWBio Biotech Co., Beijing, China), and then the RNA was reverse transcribed into cDNA with a high-capacity cDNA archive kit (Applied Biosystems, CA, USA) according to the manufacturer’s protocol. The gene expression was examined by quantitative real-time PCR as previously described [28,29]. Primer sequences used for genes were designed using Primer 3 web software. The primer sequences (5′–3′) of sterol regulatory element binding transcription factor 1 (SREBPF1), fatty acid synthase (FASN), acetyl-CoA carboxylase alpha (ACACα), stearoyl-CoA desaturase (SCD), lipoprotein lipase (LPL), peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PPARα), glucose-6-phosphatase catalytic subunit (G6PC), phosphoenolpyruvate carboxykinase 1 (PKC1), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were listed in Table 2. The primer sequences (5′–3′) of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGCR), microsomal triglyceride transfer protein (MTTP), apolipoprotein B (ApoB), carnitine palmitoyltransferase 1B (CPT1B), carbohydrate response element binding protein (ChREBP), and Leptin were listed in Table S1. Fold change of mRNA level was calculated using the standard 2−ΔΔCt method, and GAPDH was used as a reference gene for normalization [30].2.6. Western Blot AnalysisFrozen liver was homogenized in a ceramic mortar with liquid N2 and then lysed in the ice-cold radioimmunoprecipitation assay lysis buffer containing 50 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 1% NP-40, 0.1% SDS, 1.0 mmol/L phenylmethanesulfonyl fluoride (PMSF), 1.0 mmol/L Na3VO4, 1.0 mmol/L NaF, and protease inhibitor cocktail (Roche, Indianapolis, IN, USA). Equal amounts of proteins (50 μg) were separated on 10% SDS-PAGE gels after determination of protein concentration with a bicinchoninic acid protein assay kit (Huaxingbio, Beijing, China), and then proteins were transferred to PVDF membranes (PVDF, Millipore, Billerica, MA, USA). The membranes were blocked in 5% fat-free milk solution for 1 h at 25 °C and then were incubated with indicated primary antibodies overnight at 4 °C. Antibodies against GAPDH (Catalog no. sc-59540), HSL (Catalog no. sc-25843), and PGC-1α (Catalog no. sc-13067) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). After that, the members were incubated with a horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. The protein bands were incubated with an enhanced chemiluminescence kit (1:5000; Huaxingbio, Beijing, China) and developed by using the ImageQuant LAS 4000 mini system (GE Healthcare, Piscataway, NJ, USA). Density of protein bands was quantified by using the Image-Pro Plus 6.0 software (Media Cybernetics, CA, USA).2.7. Statistical AnalysisAll data were presented as means ± SEMs and were subjected to a one-way analysis of variance (ANOVA; SAS Version 9.1). Differences between group were determined by using the Tukey’s multiple comparisons test. Differences at p < 0.05 were considered significant.3. Results3.1. Effects of Dietary Crude Protein on Triglyceride and Cholesterol Concentration in Liver and SerumAs illustrated, 17% or 20% CP administration reduced the concentrations of TG (Figure 1A,B) and CHO (Figure 1C,D) in liver and serum when compared with those of piglet administered with 14% CP (p < 0.05). Concentrations of CHO in liver and serum were lower in 20% CP, as compared with 14% CP groups (Figure 1C,D). Additionally, no difference was observed in serum and liver TG concentration between 20% CP and 17% CP group (p > 0.05). Similarly, no difference was observed in liver CHO concentration between 14% CP group and 17% CP group (p > 0.05). Notably, piglets in the 17% or 20% CP group had a higher concentration of HDL-C (Figure 1E) in serum, when compared with 14% CP group (p < 0.05). Moreover, LDL-C concentration in serum was not affected by CP levels (p > 0.05).3.2. Effects of Dietary Crude Protein on Hepatic Lipogenesis Genes ExpressionmRNA expression of SREBPF1 and the downstream of its targets, such as FASN, ACACα, and SCD were determined. As shown in Figure 2, 17% or 20% dietary CP downregulated the mRNA expression of SREBPF1 (Figure 2A), FASN (Figure 2B) and SCD (Figure 2C) when compared with those of piglet administered with 14% CP (p < 0.05). Consistently, 17% CP dramatically reduced the mRNA expression of ACACα (Figure 2D) when compared with 14% CP group (p < 0.05). However, no difference was observed between 14% CP and 20% CP group (p > 0.05). Taken together, our results showed that dietary CP regulated the master genes expression which involved in hepatic lipogenesis and downstream targets genes.3.3. Effects of Dietary Crude Protein on Hepatic Lipolysis and Lipid OxidationTo investigate the protein abundance of HSL (Figure 3A) in response to different CP levels, western blot analysis was performed to measure liver protein abundance of HSL in weaned piglets. Protein abundance of HSL (Figure 3A) in liver was enhanced (p < 0.05) in piglets fed with 17% or 20% CP diet, compared with those of piglets fed with 14% CP diet. Consistently, mRNA expression of LPL (Figure 3C) was significantly upregulated by 20% CP when compared with 14% or 17% CP groups (p < 0.05). Additionally, protein abundance of PGC-1α (Figure 3B), mRNA expression of PPARα (Figure 3D) were significantly enhanced in 17% or 20% CP group when compared with 14% CP group (p < 0.05). Additionally, relative high dietary CP enhanced the mRNA expression of HMGCR (Figure S1A), MTTP (Figure S1B), ApoB (Figure S1C), CPT1B (Figure S1D), and ChREBP (Figure S1E) when compared with 14% CP group. However, no difference was observed in the mRNA expression of Leptin (Figure S1F). These results indicated that high dietary CP reduced liver TG concentration through the enhancement of lipolysis and lipid oxidation.3.4. Effects of Dietary Crude Protein on Serum Urea Concentrations and Hepatic GluconeogenesisThe concentrations of urea in serum (Figure 4A), as a major metabolic endpoint product of amino acids, were reduced (p < 0.05) in the 17% CP group when compared with the 14% CP group. However, 20% dietary CP enhanced (p < 0.05) the urea concentration in serum when compared with those of piglets fed with 14% or 17% CP diet (Figure 4A). To further understand the mechanism of dietary CP on liver lipid metabolism, mRNA expression of G6PC and PKC1 involved in gluconeogenesis were determined. mRNA expression of G6PC (Figure 4B) and PKC1 (Figure 4C) were upregulated in 20% CP group when compared with the 14% or 17% CP group (p < 0.05). However, no differences were observed in the mRNA expression of G6PC and PKC1 between the 14% and 17% CP groups (p > 0.05).4. DiscussionBase on previous study, it has been well demonstrated that low dietary CP could reduce product costs, nitrogen excretion and diarrhea [31]. Mounting evidence indicates that low dietary CP reduced the growth performance by affecting intestinal morphology, digestive enzymes and gut microbiota in weaned piglets [11,13]. Furthermore, accumulating evidence suggest that high dietary CP reduced mRNA expression of glycolysis enzymes (GK, L-PK) and lipogenesis enzymes (ACACα, FASN), and upregulated mRNA expression of gluconeogenesis enzymes in rats [11,32]. To our knowledge, the effects and the underlying mechanism of dietary CP level on lipid metabolism in the liver of weaned piglets, the best model for studying human nutrition, have not been elucidated [1,2,33]. In the present study, we found that high dietary CP reduced TG concentration in liver and serum through regulating hepatic lipogenesis, lipolysis, oxidation, and gluconeogenesis (single data points are provided in the Appendix A).Enhanced TG concentration in liver and blood circulation, reduced β-oxidation in the liver, and/or reduced synthesis or secretion of apolipoproteins are major determining factors in the development of liver metabolic syndrome [1]. Furthermore, accumulation of lipid infiltration leads to hepatic steatosis [34]. In the present study, TG and CHO concentrations (Figure 1A,B) in liver and serum were reduced in the 17% CP groups compared with piglets fed with 14% dietary CP. However, no difference was observed between the 17% and 20% CP group in TG concentration. Correspondingly, the reduced TG concentration was associated with downregulated mRNA level of SREBPF1, FASN, ACACα, and SCD (Figure 2). Similarly, previous studies have shown that high protein intake reduces hepatic lipid accumulation and plasma TG concentration in rats and humans [35,36], alleviates steatosis [37], and reduces body weight [38]. SREBPs is a master regulator of lipo- and sterol- genic gene transcription on lipid homeostasis of vertebrate cells [8,39]. ACACα plays a key role in the provision of the malonyl-CoA substrate for fatty acids biosynthesis. Downregulation of SREBPF1, FASN, and ACACα at mRNA level as observed in the present study is in agreement with previous studies showing that high CP diet reduced the expression of the lipogenic genes (ACACα, FASN, and SREBPF1) in rats or diet-induced obese rats [33,36]. SCD is a limiting enzyme in the synthesis of monounsaturated fatty acids required for normal rates of synthesis of TG, cholesterol esters, and phospholipids [40]. Recent studies indicated that SCD1-/- mice have lower concentration of hepatic TG and cholesterol esters, and are resistant to the liver steatosis [40,41]. Consistently, mRNA expression of SCD in the present study was dramatically downregulated with the rise of dietary CP which is consistent with the previous study [36].Accumulation of TG in the liver is dependent not only on lipogenesis, but also on lipolysis, oxidation, transportation, and gluconeogenesis [42]. We next determined the proteins and genes expression implicated in lipid metabolism, including HSL, PGC-1α, PPARα, LPL, G6PC, and PKC1. Interestingly, our results revealed that 20% dietary CP level markedly enhanced HSL protein abundance and mRNA expression of LPL (Figure 3A,C) when compared with 14% or 17% CP groups, suggesting that the enhancement of lipolysis by relative high dietary CP [43,44]. This finding was in agreement with the previous study [32]. The protein abundance of PGC-1α and mRNA expression of PPARα were also determined in the liver. As expected, our data indicated that relative high CP diet was associated with enhanced lipid oxidation [36].G6PC and PKC1 are important molecules involved in gluconeogenesis and glycogenolysis. In the present study, we demonstrated that 20% CP enhanced G6PC and PKC1 mRNA expression when compared with the 14% or 17% CP groups [45]. In line with an enhanced mRNA expression of G6PC and PKC1, a pronounced enhancement in the circulating concentration of urea in serum were observed in piglets fed with 20% CP diet. Urea concentration in serum represents the utilization efficiency of protein, also conversion of amino acids into glucose [46,47,48]. In agreement with our findings, previous studies show that PCK1 expression was upregulated ~3-fold by a high CP diet [10,32]. However, the result of G6PC expression in our study is inconsistent with previous study [10]. The discrepancy between ours and previous study might be explained by using different animal models. Amino acids serve not only as building blocks for proteins, but also as substrates for the synthesis of low-molecular-weight substances which contribute to the hepatic lipid metabolism [18]. Furthermore, growing studies have demonstrated that some amino acids play a key role in preventing or ameliorating tissue TG accumulation [20,49]. L-arginine, BCAAs, lysine, threonine, and low-molecular-weight peptides have been reported to reduce liver TG concentration by modulating lipid metabolism and lipid metabolism-related gene expression [50,51].5. ConclusionsCollectively, our results provided important evidence for a dietary CP level in regulating hepatic lipid metabolism in weaned piglets. A dietary CP level could regulate hepatic lipid metabolism through regulating hepatic lipid lipogenesis, lipolysis, oxidation, and gluconeogenesis.

animals : an open access journal from mdpi

10.3390/ani11072098

PMC8300421

High-fiber agriculture by-products, which can enhance animal performance and health, have the potential to be used as feed additives. Before using high-fiber agriculture by-products, it is necessary to pay attention to the problem of anti-nutritional factors and contamination due to mycotoxins. Solubility and fermentability are the keys that mainly affect fiber availability. In recent years it has been pointed out that fiber as an animal feed or feed additive does not seem to be as unfeasible as previously thought. Instead, dietary fiber and other functional compounds, such as polyphenol and flavonoids, could enhance health, antioxidant capacities, and stabilize the microbiota in animals. In addition, high-fiber agriculture by-products are a suitable and inexpensive source of fiber and their proper use may reduce costs of animal feeding. Scientists must integrate characteristics and appropriate usage analysis to jointly evaluate the effects of different fiber compositions on those animals. Based on this foundation, animal producers should be encouraged to use high-fiber agricultural by-products as animal feed and feed additives.

With the increase in world food demand, the output of agricultural by-products has also increased. Agricultural by-products not only contain more than 50% dietary fiber but are also rich in functional metabolites such as polyphenol (including flavonoids), that can promote animal health. The utilization of dietary fibers is closely related to their types and characteristics. Contrary to the traditional cognition that dietary fiber reduces animal growth, it can promote animal growth and maintain intestinal health, and even improve meat quality when added in moderate amounts. In addition, pre-fermenting fiber with probiotics or enzymes in a controlled environment can increase dietary fiber availability. Although the use of fiber has a positive effect on animal health, it is still necessary to pay attention to mycotoxin contamination. In summary, this report collates the fiber characteristics of agricultural by-products and their effects on animal health and evaluates the utilization value of agricultural by-products.

1. IntroductionThe world’s demand for food is increasing, and this includes bulk cereals including wheat, rice, corn, and soybean [1]. However, crop production is accompanied by large amounts of agricultural by-products. The by-products produced during agricultural production mainly include the stems, hulls, leaves, brans, and roots of plants. In the past, these agricultural by-products were mostly incinerated or composted to be used as fertilizer, but they also caused the problem of air pollution and inefficient use of resources [2]. The demand for animal protein is also rising, with a concomitant increase in the demand for animal feed. This further increases the requirement for common feed crops and their by-products. However, with increasing awareness of environmental protection, agricultural waste needs to be properly treated [3]. Agricultural by-products from different plant sources may also contain dietary fiber (such as cellulose and hemicellulose), starch, crude protein, oligosaccharides or vitamins, and other nutrients [1,3]. Phenols and flavonoids, the most important phytochemicals, are often included in the fiber [3,4].Carbohydrates, one of the main sources of energy for animals, can be roughly divided into two broad categories. The first category is the source of energy for animals, such as starch, glucose, and sucrose, which are decomposed by enzymes produced by animals. The second category is the dietary fibers cellulose, chitin, and other fibers which may be fermented by the microorganism. In the past, scientists and animal producers considered that the role of dietary fiber in feed is mainly to dilute protein or energy [5]. However, studies in recent years have often pointed out the complex role of fiber in animal feed. High-fiber and low-fat diets can reduce the incidence of cardiovascular disease and promote intestinal health [6]. In addition, dietary fiber content positively correlates with microorganisms such as Bifidobacterium, Roseburia, and Eubacterium rectale, and these microorganisms are the primary microorganisms secreting short-chain fatty acids (SCFA) which could provide energy for intestinal epithelial cells [7,8]. When there is a single source of carbohydrates, it can easily cause a decrease in the diversity and richness of the intestinal microbiota [9]. It even affects mucosal secretion and the health of intestinal epithelial cells.Fiber, like Pennisetum grass, a kind of common forage species, can also be used as a medium for mushroom cultivation [3,10]. Previous studies showed those mushroom wastes compost (high-fiber agricultural by-products) can reduce the inflammatory response and increase the antioxidant capacity of serum and liver of animals and enhance the growth of intestinal villi when added in the animal’s diet [3,6,10]. Moreover, adding dietary fiber (as potato pulp, sugar beet pulp, and pectin residue) can also promote the gastric emptying rate and alter the satiety of animals, the author’s results can also help sows maintain the shape of feces and reduce the incidence of diarrhea by water-retaining contents [11]. However, the sources and types of fibers also greatly limit the availability of fibers [12]. Fibers with too high a molecular weight or degree of polymerization may be difficult to digest and be utilized by the animal’s gut microbiota in a short time [13]. Too many soluble non-starch polysaccharides (NSP) may lead to excessive fermentation by intestinal microorganisms and reduce animal production performance [14]. Likewise, improperly preserved fibers may also become a breeding ground for the growth of fungi and cause the accumulation of mycotoxins. Fortunately, fiber can be fermented under controlled conditions through probiotics or enzymes, thereby reducing the damage caused by anti-nutritional factors or toxins in the fiber to animals [15,16,17].Overall, although fiber is generally beneficial for animal health, one must pay attention to the characteristics of the fiber, including solubility, fermentability, or anti-nutritional factors, as these affect the timing or availability of fiber. In vitro, fermentation of fibers with probiotics (Saccharomyces cerevisiae and Aspergillus oryzae) and/or enzymes (phytase) can form postbiotics with fermentation products and increase fiber availability [17]. This article aims to discuss the availability and possible harm of high-fiber agricultural by-products in animal feed.2. Potential Utilization of Feed Crop By-ProductsIn the feed industry of animal husbandry, corn, soybean, and wheat are the main feed sources. To produce these feed ingredients, tons of by-products, such as straw, hull, and bran, also need to be produced. Corn grains only account for about 20% [18] and soybeans account for about 30% of the dry whole plants. Accordingly, the residues are burnt or used as a material for compost, not only exacerbating air pollution but also causing waste of resources [10,15]. A wastage of these residues is not consistent with the concept of sustainable development. On the other hand, the concept of animal welfare is rising, and therefore, the producers have to try to enhance the health and decrease stress when rearing animals. Fiber increases sow satiety, decreases aggressive behavior, and improves antioxidant capacities [10,11], fulfilling the requirements of animal welfare. Therefore, feed crop by-products could be a high potential feed ingredient of the animal feed industry.3. Composition of FiberFiber from food sources is closely related to animal health; depending on the source, fiber can be divided into composite dietary fiber or a purified prebiotic. According to Prasad and Bondt [12], dietary fiber is defined as “non-digestible polysaccharides largely composed of complex carbohydrates”. The International Scientific Association for Probiotics and Prebiotics (ISAPP) defines a prebiotic as “a substrate that is selectively utilized by host microorganisms conferring a health benefit” [19]. In the field of animal nutrition, scientists classify fibers based on their biochemical characteristics into acid detergent fiber (cellulose and lignin), neutral detergent fiber (hemicellulose, cellulose, and lignin), and dietary fiber (non-starch polysaccharides) (Figure 1).Nevertheless, fermentability and solubility are also some of the concerns of animal nutritionists. The fermentability of fiber is related to the microbial composition of the intestine and short-chain fatty acids, while the solubility is related to the satiety and fecal configuration of animals [11,12]. Insoluble fibers generally include cellulose and lignin, while soluble fibers include inulin and pectin [12]. Interestingly, most soluble fibers are more fermentable than insoluble fibers. Therefore, soluble fiber (3% pectin) may also promote excessive growth of microorganisms (total count of bacteria and E. coli) and increase chyme viscosity [14].Considering cost constraints, it is more difficult to use highly processed and purified prebiotic as feed additives for animal production; on the contrary, dietary fiber is very suitable. However, dietary fiber (consisting of approximately 40–50% cellulose, 20–30% hemicellulose, 10–25% lignin, and about 35% pectin) may contain dozens or even hundreds of different carbohydrate bonding patterns, as well as complex compositions such as phenolic, minerals, and vitamins [20]. Therefore, most of the research on fiber is still mainly aimed at prebiotics. Research on dietary fiber indicates that it is difficult to explain which kinds of fibers have positive or adverse effects, and only a simple distinction can be made according to the source of dietary fiber. Although there is almost no relevant research reported, to understand the efficacy of different fibers in complex dietary fiber, metabolomic techniques could assist scientists in understanding the changes in animal digestive state after involving dietary fiber [7,21].4. Partial Plant Phytochemicals and Anti-Nutrition Factors in High-Fiber By-ProductsIn addition to fiber, plants also contain a variety of secondary metabolites, including antioxidant peptides, phenols, flavonoids, phytic acid, trypsin inhibitors, and lectins [22,23]. These plant-derived secondary metabolites are rich in antioxidant, anti-inflammatory, and antibacterial activity, but may also reduce the animal’s absorption efficiency to nutrients such as minerals or amino acids [22]. Phytochemicals have been shown to have positive antioxidant benefits toward animals in terms of favored growth performance, production quality, and enhanced endogenous antioxidant systems, possibly by directly affecting specific molecular targets or/and indirectly as stabilized conjugates affecting the metabolic pathways [11,13,23]. Accordingly, dissecting the antioxidant effects and the underlying mechanism of dietary phytochemicals is an important area. Therefore, much attention is being focused on a new wave of nutrigenomics [6,10,23]. Several studies have been dedicated to understanding and formulating mechanistic pathways by which these naturally derived substances could alter the fate of cells, particularly the antioxidant properties of phytochemicals have been implicated as stress-alleviation agents [11,23,24]. Besides phytic acid, trypsin inhibitors, and lectins, there are also anti-nutritional factors commonly found in plants [24,25] that cannot be ignored while using high-fiber agricultural by-products as animal feed or feed additive.4.1. PhenolPhenols are common components in plants, and their basic unit structure is a benzene ring connected with an OH-. In general, the phenolic compounds found in plants vary from plant to plant. Some of the well-known phenols include catechins, chlorogenic acid, caffeic acid, and quinine [26]. These phenolic compounds improve the antioxidant capacity of animal serum and reduce the degree of animal inflammation [27]. On the other hand, some phenolic compounds can also inhibit the production of pathogenic bacteria such as grape extract, which was effective at inhibiting antibiotic-resistant Staphylococcus aureus and E coli, including methicillin-resistant S. aureus, with minimum inhibitory concentrations (MIC) ranging from 0.3 to 3.0 mg/mL, by chelating minerals necessary for the microorganism survival, or perforating the microbial cell membranes [28]. However, some plants also contain potentially toxic phenolic compounds, such as Pyrogallol, a simple phenolic found in green tea, which was shown to cause hepatic damage when administered at 100 mg/kg in rats. Serum enzymes including aspartate aminotransferase and alanine aminotransferase (ALT) as well as malondialdehyde (MDA) were increased, suggesting that free radical formation and pro-oxidant toxicity played a role. Therefore, excessive supplementation may cause animal poisoning, shock, and reduced nutrient absorption in animals [29].4.2. FlavonoidsFlavonoids are a more functional classification of phenolic compounds, and their basic unit structure is 2-phenyl-1,4-benzopyrone. The classic flavonoids include estrogen, anthocyanin, and catechins [30]. The function of flavonoids is similar to that of phenols. Among them, catechin, the classic flavonoid in green tea, can increase the antioxidant capacities, and also promote fat metabolism in animals, thereby reducing low-grade inflammation caused by excessive accumulation of fat [30,31]. Catechins can be found not only in tea but also in Pennisetum [3]. Phenols or flavonoids, the plant secondary metabolites (functional phytochemicals) which have the ability to induce the expression of antioxidant/phase II enzymes, appear to have a major role in acting as modifiers of signal transduction pathways to elicit its cytoprotective responses through suppressing stress-induced protein activation and enhancing Kelch-like ECH associating protein 1 (Keap1), a cytoskeleton binding protein, and dissociation from nuclear factor (erythroid-derived 2)-like 2 (Nrf2) in response to stressors. Therefore, suppression of abnormally amplified oxidation signaling and restoration of improperly working systems, as well as the activation of antioxidant machinery could provide important strategies for prevention of oxidative stress and augmentation of antioxidant defense in animals [3,15].4.3. GossypolGossypol is a type of di-sesquiterpene aldehyde, mainly found in cotton seeds [32,33]. Besides having high antioxidant and antitumor activity, gossypol can reduce the inflammatory response by decreasing the NF-κB expression, and also has high neurotoxicity and reproductive toxicity [33,34,35]. Therefore, the residual amount and use of gossypol in plant raw materials are highly valued. Owing to its drought-tolerant properties, cotton is particularly suitable for planting in arid areas [36] leading to the production of cottonseed oil and cottonseed meal. However, as an agricultural by-product, cottonseed meal contains gossypol in addition to its high fiber content. Using cottonseed meal, such as Hy-line Brown layer or Shanshui White duck, will leave gossypol in the eggs, reduce the yolk quality, and destroy duck liver cells [37,38]. In contrast, research on a ruminant (Odocoileus virginianus) indicated the accumulation of gossypol in the white-tailed deer serum when using whole cottonseed supplements but did not affect white-tailed deer health and reproduction [39]. Overall, the utilization of cottonseed meals should be strictly controlled and the hazards of cottonseed phenol residues on animal health and food quality must be closely monitored.4.4. Tannic AcidTannic acid is a water-soluble polyphenolic substance and is found naturally in plants. Tannic acid has antioxidant, anti-inflammatory, and antibacterial activities, but may also reduce the digestive enzyme activity of animals, hinder mineral absorption, and exhibit cell membrane toxicity [40,41]. A higher tannic acid addition also reduced the feed intake of broilers [42,43].Using drinking water to ingest 10 g of tannic acid in poultry greatly reduces the utilization of amino acids, especially methionine, histidine, and lysine, and D-xylose [43,44]. Tannic acid also reduces plasma iron levels and the performance of weaned piglets [45]. Mansoori and Modirsanei [46] further pointed out that tannic acid reduces the effect of the anti-coccidial vaccine. In contrast, Tonda et al. [47] indicated that the addition of 0.5 g/kg gallnut tannic acid could decrease the number of Eimeria spp. Overall, tannin is regarded as a plant-derived anti-nutritional factor. However, because Kubena et al. [42] reported the antibacterial activity of tannic acid, Cengiz et al. [48] added some (2 g/kg feed) tannic acid to animal feed and found that it could reduce the negative effects of barley NSP. Therefore, if the appropriate amount of tannic acid is found through animal trial for evaluation, it will have a positive effect on animal health.4.5. Phytic AcidsAbout 50–80% of plant phosphorus is stored in phytic acid, mainly in the form of inositol hexakisphosphate (IP6) [49]. There are different types of phosphoric acid depending on the number of phosphates attached to inositol [49,50]. Phosphoric acid on inositol can combine with minerals or amino acids of valence 2 or 3 and reduce the absorption rate of nutrients and enzyme activity of animals [49,50,51,52]. In addition, the involvement of too much phytic acid may also lead to poor phosphorus utilization by animals, causing the accumulation of phosphorus in stool and this may cause pollution to the environment [53]. Fortunately, phytic acid can be degraded by phytase. Currently, most commercial phytase is produced from fungi or Escherichia coli [50]. Adding phytase to animal feed can reduce the impact of phytic acid on animal digestibility [54].4.6. Trypsin InhibitorTrypsin inhibitor is a well-known anti-nutrition factor in plant-based ingredients, which could reduce the degradation of protein and thereby reduce the nutrient absorption of animals [55]. Fortunately, trypsin inhibitors can be reduced or destroyed by heat and reducing agents [55]. Therefore, the process methods significantly affect the quality of the feed source containing a high amount of trypsin inhibitor, such as that of soybean, chickpea, and buckwheat [56]. Avilés-Gaxiola et al. [57] reported that while reducing agents, such as L-cysteine, can decrease the activities of trypsin inhibitor (up to 89.1%), treating the raw ingredients by heat and reducing agents could be more effective (up to 99.4%). Another study showed that polyphenols extracted from tea, especially epigallocatechin gallate and epigallocatechin, can suppress the activities of trypsin inhibitors [58].4.7. LectinLectins are commonly produced plant proteins and can chelate the carbohydrate on the glycoprotein, thus they can condense and destroy cells [59]. Ricin, one of the most well-known lectins, is extracted from Ricinus communis and causes animal death [59]. Many common feed ingredients also contain lectins, including corn, soybeans, and wheat [59]. The toxicity of lectin is mainly based on the “degree of lectin resistance to proteolytic degradation” [59]. Accordingly, the toxicity of lectins can also be utilized for removing hazardous waste in animals. Therefore, for the utilization of lectins, their physicochemical properties (including hemagglutination activities, inflammatory, antibacterial, and antifungal activity) as well as dosage must be considered and evaluated [60].5. Effect of Fiber Addition on Animal Production Performance and MicrobiotaAnimals cannot effectively degrade dietary fiber, so the fiber ingested by the animal initially undergoes preliminary fermentation and degradation through the intestinal microbiota before being used as an animal’s nutritional source. When digesting fiber, the composition of microbiota might therefore alter [7]. For instance, excessive intake of soluble NSP may lead to the increase of E. coli and Clostridium, which negatively impact animal performance [14]. Conversely, insoluble dietary fiber may absorb harmful substances and excrete them [12]. Animals have different sensitivities in the gut microbiota at different growth stages, therefore, animals with different life cycles should be discussed separately. In this review, we mainly discuss the effect of dietary fiber on the changes of animal intestinal microbes and production performance. Table 1 lists the effect of different fiber replacement or addition on animal growth performance and microbial microbiota in the ileum.5.1. Effects of Fiber on BroilersMateos et al. [66] indicate that the dietary fiber addition could improve poultry health, including increase the gizzard weight, nutrition digestibility, and intestine morphology; however, the fiber addition should not be over 3% in broiler diet. Kermanshahi et al. [14] indicated that the 3% cellulose addition in the diet of broilers (Ross 308) had similar growth performance, intestinal morphology, microbe composition, and serum characteristics in comparison to the control group during the experimental period (0 to 14 days). However, the 3% pectin and carboxymethyl cellulose addition would increase the number of E. coli and thereby decrease the growth performance and gut health.5.2. Effects of Fiber on SwinePigs have different nutritional requirements in different stages of their life cycle. For piglets during lactation, sow milk is the most important source of energy; therefore, it is important to maintain stable production of sow milk and ensure that every piglet consumes milk [65,67]. Piglets face huge environmental changes and stress when weaning, while for the piglets that have just been weaned, there are significant changes in the type of food ingested, so diarrhea is prone to occur [45,64]. Feeding piglets after weaning provides an appropriate amount of insoluble fiber (as inert fiber) to avoid diarrhea caused by accumulation of undigested nutrients and help piglets to restore intestinal function [64,65]. Moreover, it is necessary to maintain stable feeding and improve the physiological health of pigs during their late growth period or during sow pregnancy. Therefore, adding dietary fiber to the sow diet could increase feed intake during lactation, thereby increasing the number of weaned piglets, weaned piglet weight, and average daily gain. Feeding high-fiber diets in late pregnancy may also reduce the stillbirth rate by reducing delivery time. In addition, it can alleviate constipation in pregnant and lactating sows [67,68].The addition of 40 mg/kg chito-oligosaccharide can improve the carbohydrate composition of sow milk, and therefore increase litter size, survival rate, and total litter weight of both at 12 and 21-day-old pigs [67]. In total, simultaneous addition of 1% soluble (inulin) and insoluble (lignocellulose) fiber could significantly increase the feed conversion rate of weaning pigs (24 to 52-day-old); however, adding soluble fibers or insoluble fibers singly is less effective [64]. The total short-chain fatty acids increase by about 30% in all 1% fiber addition groups compared to the control group, thereby enhancing the tight junction (TJ) expression in piglet ileum. Nevertheless, the sodium-glucose cotransporter-1 gene expression in piglet jejunum increased by 2-fold in the 1% insoluble fiber addition group compared to that in the control group [64]. Moreover, Naya et al. [68] indicated that the addition of soluble fiber could decrease tail biting in 3 to 10-week-old pigs. According to the above-mentioned results, insoluble fiber is better than soluble fiber in promoting the growth performance of pigs. However, adding both soluble and insoluble fibers is more effective than adding insoluble fibers singly.6. Effect of Functional Components of Fibers on Animal Physiology6.1. Fibers Antioxidant and Anti-Inflammatory Responses in AnimalsPrevious studies have discussed in detail the inflammation and antioxidant mechanism of phytochemicals on animal immune regulation and antioxidant capacity [10,69,70]. Briefly, when animals are subjected to environmental stress, such as heat stress or pathogens infection, stimulation causes an inflammatory response and increases oxidative stress in animals [71]. Stimulated by pathogenic bacteria, the animal initiates an immune response leading to a cytokine storm [71]. However, excessive inflammation can reduce animal performance and even lead to death [10]. On the other hand, when the oxidative pressure is too high, animals are not able to eliminate the damage caused by free radicals to cells or organs [69]. Among them, the animal’s antioxidant system is mainly regulated by the liver, so the antioxidant capacity is also related to liver performance [3]. Applying phytochemicals or botanical compounds to the feed could promote intestinal health, reduce the inflammatory response, and enhance the antioxidant capacity of the animal [10]. The mechanisms are mainly the nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and nuclear factor kappa B (NF-κB) for they are respectively the key transcription factors involved in oxidative stress and inflammation for elucidating the underlying signal transduction pathways. Therefore, phytochemicals can regulate these transcription factors leading to the improvement of oxidative status, the heme oxygenase-1 (HO-1) gene is found to be crucial for Nrf2-mediated NF-κB inhibition. Hence, proper fibers as phytochemicals (likely 0.5–1% mulberry leaves addition in laying hens) with such modulatory effects should be used to explore the possible crosstalk in oxidative stress and immunomodulation in animals [69,70,71].6.2. Fibers Satiety in AnimalsObesity caused by overfeeding of sows is not conducive to piglet production, because it may increase the production time of sows, and suffocate oversized piglets in the vagina, reduce the number of births, and overall life cycle of sows [72]. Therefore, during the sow’s pregnancy, the breeder limits the feed intake of the sow [73]. However, failure to get enough satisfaction may lead to a more stereotyped and aggressive behavior of the sow [73,74].To improve the satiety of sows, scientists have proposed that high-fiber feed can be given to dilute the total energy in the feed [74,75]. Although dietary fiber cannot be digested by endogenous enzymes in animals, gut microbes degrade dietary fiber into short-chain fatty acids, further regulating animal feeding patterns and gut health [76]. According to its characteristics, fiber can be either soluble or fermentable. After dissolving, soluble fiber increases the viscosity of chyme, thereby increasing the transit time of chyme [11]. Therefore, compared with insoluble fiber, soluble fiber can improve the satiety of animals [11]. The satiety of food in the liquid phase is higher than that in the solid phase [11]. Fermentable fiber can produce more short-chain fatty acids, stimulate animals to produce antimicrobial peptides, and further adjust intestine health [76].Giving high-fiber (totally dietary fiber about 28.2%), low-energy diets (including mainly 24.4% soybean hulls) to sows in group cultures can improve their feeding time and health, and reduce aggressive and stereotyped behavior [73,74]. Providing higher amount of fiber (7.5% crude fiber consisting of 20% Alfalfa meal and 52% corn in the lactation diet) can also improve the welfare and improve the production performance of sows [77] and, therefore, be beneficial to the health of pregnant sows.7. Fibers Can Increase Animal PerformancesThe physiological response of fiber addition to animals has been discussed in detail in the earlier sections. In this review, we discuss further how fiber addition and application could improve animal performances. Traditionally, fiber is considered to be an anti-nutritional factor in feed and has a negative effect on animal palatability and production performance [5]. In contrast, many studies have repeatedly pointed out that adding fiber can improve animal performance [78,79,80,81,82,83]. Herein, the review discusses more the role of fiber in increasing animal production performance by promoting intestinal health, immune regulation, and changing fat metabolism patterns. Table 2 lists the effect of different fiber supplements on digestibility, health, or production of animals.7.1. Intestinal Health and Immune RegulationBoth mice and poultry studies have pointed out that high fiber intake can increase the performance and thickness of the intestinal barrier of animals, including TJ and mucosal proteins [7,16,85,86,87]. The addition of prebiotics (autoclaved drinking water supplements with 1.0% oligofructose-enriched inulin (w/v)) can also increase the viscoelasticity of mucosal proteins [88]. A robust intestinal barrier can increase the distance between pathogenic bacteria in the intestinal cavity and intestinal epithelial cells and reduce the potential destruction of intestinal epithelial cells by pathogenic bacteria [7,85]. With the increase in the distance between pathogenic bacteria and intestinal epithelial cells, the inflammatory response of animal intestines decreases [7,85]. On the other hand, intestinal stability is related to the health of animals. Especially for economic animals, the efficiency of nutrient absorption is affected by intestinal villi, which may be damaged by any environmental stress [88,89]. The addition of fermented fiber can be used to suppress the number of pathogenic bacteria or inflammation, thereby reduce the damage, and increase the length of intestinal villi [3,16].In addition to the villus height, the motion of circular muscles in the intestine can drive the surface convection of the chyme. This movement is different from peristalsis and segmentation in mammals, a reflex type of motility [90]. The depth of the unstirred water layer is largely determined by the length of villi during the motion of the circular muscles which have various microbiota and can cause damage to the upper villus [89,90]. In some cases, such as in wheat or barley, which have a thicker unstirred water layer, there will be greater lumen viscosity and reduced oxygen transfer from mucosa to the intestinal cavity, thus increasing the activity of anaerobic bacteria [90,91]. However, the higher villus length can strengthen the agitation of chyme and alleviate the above situation.7.2. Digestibility Adjustment in AnimalsThe quality of ingested feed is one of the most important environmental factors that affects animal production performance. Besides the energy density of feed and the composition of anti-nutritional factors, the digestibility of feed by animals is also an important indicator [82]. Higher digestibility can reduce the waste of nutrients in feed and reduce environmental pollution due to animal waste. Giving higher fiber can generally increase the digestive capacity of animals and can stabilize the composition of excreta more quickly [82]. However, excessive fiber addition (more than 40% peach palm (PP) meal replacement for maize in goats, the NDF corrected for ash and protein (NDFap) is 40.1% and acid detergent fiber (ADF) is 20.2% in 40% DM level of PP meal substitution group) may still cause a decrease in palatability, and the effect of digestibility is also related to the source of fiber [91]. Choi and Kim [82] and Navarro et al. [92] indicated out that although soluble fiber can be fermented in the intestine and produce short-chain fatty acids, insoluble fiber seems to increase the digestibility of animals.Although intestinal health is closely related to the overall health of the animal, de Nanclares et al. [93] showed a positive correlation of the digestibility of pig with the enzyme activities but not with intestine morphology. Similarly, Liu et al. [94] pointed out that wheat bran has higher neutral detergent fiber (NDF), acid detergent fiber (ADF), and crude protein digestibility than soybean hull of animals, while exogenous addition of NSP-degrading enzyme could enhance nutrient utilization.Chen et al. [58] also indicated that 1% addition of both soluble (inulin) and insoluble (lignocellulose) fiber could increase the digestibility of about 30% of dry matter, crude protein, and organic matter on weaning piglets. However, 5% cellulose, xylan, or β-glucan supplement was found to decrease the fecal digestibility of weaning pigs [65]. In growing pigs (21.3 ± 1.0 kg), 5% inulin supplement would decrease the digestibility of dry matter, NDF, and carbohydrates; however, the digestibility of ether extract increased. In contrast, although 5% carboxymethyl cellulose sodium supplement showed a similar result on the digestibility of ether extract compared to the inulin supplement group, the digestibility of other nutrients did not decrease, and even increased the NDF.In the animal model of poultry, 2.5 or 5% supplement of oat hull, which is insoluble fiber, enhanced the gizzard weight and decreased the pH value in the intestine [83]. Fiber addition could further decrease the negative effects of pelleting and increase nutrient digestibility [83]. Likewise, Jiménez-Moreno et al. [95] also indicated that 2.5 or 5% supplementation with oat hull, rice hull, and sunflower hull did not decrease the growth performance in 21-day-old poultry. However, an increase in fiber supplements could improve water intake [95]. Overall, insoluble fiber can induce the digestibility of nutrition, thereby enhancing growth performance both in poultry and swine.7.3. Fat Metabolism and Muscle GenerationIn addition to promoting intestinal health, adding fiber can also reduce the amount or pattern of animal fat accumulation by promoting animal fat metabolism and reducing mRNA expression of animal fat synthesis genes [96]. Okrathok and Khempaka [62] showed that 1% cassava pulp could decrease the cholesterol content in the breast, thigh, liver, and serum in broilers. Similarly, in previous research reported by our team, we showed that addition of 0.5–2% mushroom waste compost could enhance adipolysis both in the liver and adipose tissues in broilers [10]. Moreover, the level of oxytocin, which promotes muscle formation, also increased in the 1–2% mushroom waste compost addition group compared to the control group by about 3.4 to 3.7-fold [10].8. Fermented Fiber8.1. Solid-State and Liquid FermentationsDuring fermentation, microorganisms use specific substrates as energy sources to decompose, metabolize, or produce metabolites. Therefore, in general, the substrates fermented by microorganisms usually have bacterial proteins, secondary metabolites, small molecule peptides, and carbohydrates [97]. If the bacterial species used are GRAS (generally regarded as safe) probiotics and a nontoxic substrate, the fermentation products can be used as animal feed additives. According to different modes, fermentation can be either solid fermentation or liquid fermentation.Solid-state fermentation uses a solid substrate, with adjusted moisture or pH, and probiotics are added to ferment at a specific temperature. The fermentation duration depends on different substrates or microorganisms [97]. Because of the low moisture content of solid-state fermentation, filamentous fungi or yeasts are generally considered suitable for growth [97]. In addition, microorganisms such as yeast or Aspergillus oryzae, with high activities of carbohydrate catalyzing enzymes are also used to ferment high-fiber substrates, such as wheat bran, straw, and lupin flour [17,25,98]. The high-protein soybean meal is suitable for fermentation using microorganisms with high protease activity such as Lactobacillus spp. and Bacillus subtilis [99]. After fermentation, cellulose and lignin content decrease, while hemicellulose and extractable functional metabolites increase slightly [3,53]. However, other crude ingredients, such as crude protein and minerals, are concentrated [53]. After probiotic fermentation, the anti-nutritional factors usually present in plants also decrease [97]. Fermented fiber also increases the number of probiotics in the intestinal tract and enhances the antioxidant capacity in animals. These characteristics increase the utility and benefits of fermented fibers [16,97].In liquid fermentation, the water content in the fermentation substrate is much higher than the dry matter, or the solid substrate is directly immersed in the culture broth for fermentation [4]. The advantage of liquid fermentation is that the fermentation degree is relatively uniform and can easily adjust the nutrient composition and pH value in the culture broth while promoting the production of specific functional metabolites [4]. However, the functional components after fermentation are easily diluted by the culture solution and are not easily dried.8.2. Stage Fermentation and Co-FermentationIn recent years, the probiotic mode of fermentation has emerged. Because probiotics are not a panacea, a probiotic may only be good at secreting enzymes related to fiber or protein. Therefore, this problem can be solved through two-stage fermentation. Although there are only a few reports on two-stage fermentation for animal feed, this technology has been used in the brewing, biochemical, and decontamination industries for decades [100,101,102]. Two-stage fermentation can reduce the restrictions on the use of probiotics and also provide complementary effects. Scientists have also formulated fermented substrates with multiple probiotics at the same time; this could also enhance the quality of fermentation, although there may also be mutual inhibition effects.Besides using probiotics for two-stage fermentation, enzymes and probiotics can also be used for co-fermentation [17]. Although probiotics can survive on many substrates, they may also encounter non-degradable anti-nutritional factors such as phytic acid, which is secreted in large amounts mainly by fungi or E. coli [50]. When probiotics ferment fibers, they release complex metabolites enveloped in fibers, including phytic acid and phenols [3,17]. When phytic acid sequesters minerals or amino acids, the growth rate of probiotics decreases [52]. However, when probiotics and phytase simultaneously ferment plant raw materials, probiotics can release phytic acid in the fiber, and phytase can further degrade phytic acid [17]. The released minerals or amino acids can further promote the growth of probiotics. The above cycle of co-fermentation of plant raw materials with probiotics and phytase will be more effective than using one of the substances alone [17].Whether two-stage fermentation or co-fermentation can increase the availability of fiber needs to be assessed carefully. In addition to releasing the secondary metabolites of plants and reducing the content of anti-nutritional factors, the fermented fiber is also rich in probiotic metabolites, including bacterial proteins and short-chain fatty acids [31,103]. The bacterial protein can be used as one of the animal’s nutrient sources to promote animal growth. Short-chain fatty acids can lower the pH value in the intestine, maintain the health of intestinal epithelial cells, reduce inflammation, and provide animal energy [104]. Menconi et al. [105] also pointed out that the addition of organic acids can reduce the cost of producers and increase profits. A detailed discussion on the impact of short-chain fatty acids on animal health has been reported by Xu et al. [104].9. Evaluation of the Use of Agricultural By-ProductsDespite the mentioned benefits, the use of agricultural by-products, considering economic benefits or animal health, may still pose some risks and problems that must be addressed. Agricultural by-products are mostly rich in water. High-fiber and high-moisture substances are easily contaminated by mold and produce toxic mycotoxins [106], which decrease animal performance or even lead to death [107,108]. Therefore, the rapid drying of agricultural by-products is an important issue. However, because the agricultural by-products are not valued by producers, additional drying of agricultural by-products are not expected. To solve this problem, we put forward two suggestions. First, we encourage producers to discover and value the agricultural by-products and collect and dry them quickly after their production. We also suggest that agricultural by-products be collected quickly for adjusting water or pH, followed by fermentation with probiotics in a controlled environment to increase their utility value [4,16,17]. This approach will save the cost of drying the by-products once; however, it requires sufficient expertise to proceed.On the other hand, because the agricultural by-product fiber structure is varied and complex, we still do not know the extent to which these composite fibers contribute to animal health. Considering that not every fiber may be suitable for improving animal health, the composition, and content of fiber affect the composition of intestinal microbiota, and intestinal microbes are closely related to virus composition in the animal gut [21,109]. Therefore, we suggest an integrated assessment of the effects of different agricultural by-products on animal health and their correlations through emerging metabolomics, microbiome, and virome analysis [7,21,109,110].10. ConclusionsAlthough the world’s food demand is rising, people are pursuing a better quality of life while also noticing that the earth’s environment needs to be protected. Food is a rigid requirement of life. Since the production of agricultural by-products is inevitable, we expect their use as animal feed or feed additives after appropriate treatments. Studies in recent years have repeatedly pointed out that fiber as an animal feed or feed additive does not seem to be so unfeasible as previously thought. Instead, dietary fiber and other functional compounds, such as polyphenol and flavonoids, could enhance health, antioxidant capacities, and stabilize the microbiota in animals. In addition, agricultural by-products are a suitable and inexpensive source of fiber; they are not only inexpensive, but their proper use can also reduce costs of waste disposal and animal feeding. Scientists must integrate metabolomics, microbiome, and virome analysis to jointly evaluate the effects of different fiber compositions on animals. Overall, we recommend that animal producers be encouraged to use high-fiber agricultural by-products as animal feed and feed additives.

animals : an open access journal from mdpi

10.3390/ani11072094

PMC8300101

The European bison (Bison bonasus) is an endangered species which faces a number of health threats. One potentially dangerous disease is paratuberculosis, which can cause diarrhea and cachexia in animals and is a potentially dangerous disease for humans. The aim of this study was to conduct a serological survey of paratuberculosis in Polish bison herds. Of the tested 165 European bison, three were found to be positive, indicating that paratuberculosis is not currently an urgent problem in this population. However, as the appearance of symptomatic paratuberculosis in a single European bison subpopulation could be fatal for the restitution of the species as whole, further random checks are needed.

The European bison (Bison bonasus) is an endangered species which faces a range of health threats. As little is known of exposure of European bison to paratuberculosis caused by Mycobacterium avium spp. paratuberculosis, known to cause losses in cattle, the aim of the present study was to conduct serological survey in Polish bison herds. Between September 2018 and February 2021, blood samples were collected from 165 European bison from different regions of Poland. Samples were taken whenever the animals were immobilized (e.g., putting on telemetry collars) as well as from any dead animal. The serum samples were tested with ELISA. Three individuals, originating from different captive herds, were found to be seropositive. In conclusion, it was found that Mycobacterium avium subsp. paratuberculosis infections are not currently a problem in European bison, especially in free-range herds.

1. IntroductionThe European bison (Bison bonasus) is an endangered species which faces a wide range of population management problems [1,2] as well as environmental and health threats [3,4,5]. More than a quarter of the world’s population lives in Poland. Currently, the total number of European bison in Poland is 2316 (212 captive and 2104 free-living) (Bialowieski Park Narodowy—European Bison Pedigree Book issues edited after (bpn.com.pl, accessed on 1 June 2021)) and the population is gradually increasing. However, this increasing density favors the spread of infectious diseases, and hence there is a pressing need to implement active protection measures and carefully monitor the spread of any potential threats [6].Among bacterial diseases, the most dangerous threat in recent years for European bison has been the outbreak of bovine tuberculosis (BTB) which, in this species, is mainly caused by Mycobacterium caprae. However, atypical mycobacteria have also been recently isolated from European bison lymph nodes [7]. The most common emerging atypical mycobacterium in cattle seems to be Mycobacterium avium spp. paratuberculosis (MAP) [8]. Not only does it cause serious economic losses in livestock production, but is also a potentially zoonotic agent [9]. MAP is an etiological agent of paratuberculosis (Johne’s disease) which has been noted in a number of types of livestock, as well as wild animal species [10,11,12]. Infections usually persist in a subclinical state for several years, and when clinical signs occur, they usually concern the digestive system (diarrhea) [12].Arguably the most attractive way to monitor disease, especially in wildlife, is the serological survey; it is a relatively cheap and fast method which allows for the simultaneous examination of many individuals, and the performance of retrospective studies. It should be emphasized that the diagnosis of paratuberculosis may be more complicated in the case of infection with other mycobacteria, including those belonging to the Mycobacterium tuberculosis complex [13]. Additionally, it should be emphasized that, while ELISA testing seems to be suitable for monitoring disease in wild populations, this is not the case for individual animals, especially when both IgM and IgG antibodies are not checked and when samples are collected during one-time sampling. Such testing is associated with a risk of false positive results, if obtained in small numbers.The aim of the study was to conduct a serological survey of paratuberculosis in the European bison (Bison bonasus) population in Poland and to figure out where there is a need for continuous monitoring of the disease.2. Materials and Methods2.1. Animal SamplingThe research was conducted as a part of the “Complex project of European bison conservation by State Forests” project in Poland; it is a comprehensive study of the cross-section of the Polish European bison population over several years. Samples were collected ante mortem whenever pharmacological immobilization of animals was performed for another purpose, such as when moving them to another bison center or putting on telemetry collars. All the procedures were carried out with high respect of animal welfare by qualified veterinarians in accordance with Polish law. Post mortem samples were collected during culls approved by the decision of Polish General Directorate for Environmental Protection or Regulation of Polish Ministry of the Environment, or from animals found dead.Between September 2018 and February 2021, blood samples were collected from 165 European bison from free ranging (n = 63) and captive herds (n = 102) (Figure 1). Four out of the six free-living populations in the country were tested (Knyszyńska Forest (1/214), Borecka Forest (16/128), Bieszczady Mountains (15/707), and Białowieska Forest (31/715)). Of the captive herds, the following numbers of animals was tested in the largest centers: Pszczyna (34/56), Białowieża (16/27), Niepołomice (14/22), and Gdańsk Oliwa (2/13). Animals were tested in other captive herds, but in these, only fewer than 10 animals are kept. The sex division was as follows: 79 males and 85 females, with one individual not being identified (partly eaten carcass). The animals ranged in age from a few months to 23 years; however, 14 individuals were of unknown age (Table 1). Blood samples were collected ante mortem (after pharmacological immobilization) or from the jugular vein (v. jugularis externa), tail vein (v. caudalis mediana), or, in case of post mortem collection, from the heart. Blood was collected with a 1.2 mm-diameter needle into serum tubes with clot activator and centrifuged. The obtained serum samples were stored in the freezer at −20 °C.2.2. Diagnostic AnalaysisAfter defrosting, the serum samples were tested with a Mycobacterium paratuberculosis Test Kit for Cattle PARACHEK® 2 (Prionics AG, Schlieren-Zurich, Schlieren, Switzerland). The test is designed and standardized for cattle. However, ELISA tests for cattle are also commonly used in European bison [14,15] due to the close phylogenetic connection between those two species and the lack of tests designed specifically for European bison. The procedure was carried out in accordance with the manufacturer’s instructions. The absorbance of each well was read at a wavelength of 450 nm with an EPOCH spectrophotometer (BioTek Instruments Inc., Winooski, VT, USA). For interpreting the results, the same cut-off values were used as those given by the manufacturer of the test for cattle.3. ResultsParatuberculosisFrom tested serum samples (n = 165), only three (1.81%) were found to be positive in the test for M. avium spp. paratuberculosis antibodies. All seropositive animals originated from captive herds: a one-year-old male from Pszczyna herd (56 herd members), a three-year-old male from Warsaw Zoo (seven members), and a 10-year-old female from Smardzewice Breeding Center (six members). Although other ruminants are kept in Pszczyna Herd and Warsaw Zoo, they do not have direct contact with the European bison. Interestingly, the seropositive bison from the Smardzewice Breeding Center had been confirmed to be Mycobacterium caprae-positive in other studies [7].4. DiscussionOur findings indicate that antibodies against M. avium ssp. paratuberculosis are rarely present in European bison, indicating that they do not have much contact with this pathogen. Three seropositive animals were detected, but unfortunately their feces were not available to confirm or exclude infection by PCR. The present study is the first cross-sectional study of paratuberculosis in European bison population. The collected material is unique, particularly considering that the European bison is a wild and endangered species. Previous research was limited to 23 young individuals from one region of Poland, and no positive result was obtained [16]. As it might be expected, the three seropositive individuals identified in the present study came from captive herds, which is in line with previous scientific reports indicating that MAP infections are mainly associated with captive herds and zoo animals [17]. Although MAP infection is known to occur in farmed cervids, recent studies have examined the problem in zoos, in which the problem is believed to be underestimated. Infection has been noted in both ruminants and other species [18,19].To the best of our knowledge, this is the first recorded detection of antibodies to MAP in European bison. However, cases of MAP infection have been noted in the closely related American bison (Bison bison) [20,21,22]. Furthermore, reports suggest that ruminants are the main reservoir of paratuberculosis [12], which places the European bison in the high-risk group for potential MAP infection. In the case of large ruminants, MAP affects the digestive system, with the most noticeable clinical sign in cattle being intense, watery diarrhea, not susceptible to treatment. It can be assumed that similar symptoms would be seen in European bison. While it is possible that vaccination may be used to manage an outbreak in a captive herd (see, e.g., [23]), the disease could lead to a rapid population reduction in the case of free-ranging herds. Some wild ruminant populations have demonstrated a high seropositivity, e.g., 17.64% in sika deer (Cervus nippon) in China [24], and 30.16% in Spanish red deer (Cervus elaphus hispanicus) [25], while much lower levels have been found in others, e.g., from 1.9% to 12.2% in free ranging cervids in Norway, depending on species [26].In Poland, the largest group of ruminants are cattle. Although the prevalence of MAP-positive herds is estimated to be rather low in Poland, low numbers of cases are nevertheless regularly reported [27,28]. As disease transmission between species is relatively easy, e.g., through sharing common pastures, which is commonly known in Polish European bison herds [29], paratuberculosis should be viewed from a broader perspective, i.e., as a threat to all wildlife, including the European bison. Furthermore, such transmission should also take into account the economic aspects or potential public health implications [12].The diagnosis of paratuberculosis is complicated, and ELISA is considered to be the most specific method to detect antibodies; however, higher sensitivity can be gained with fecal PCR, culture, and pathology. Nevertheless, both methods are equally effective and are used comparably frequently [12]. According to the manufacturer, the ELISA kit used in the present study has a test specificity of 99% for cattle, with a detection rate approaching 80% while the subject is still asymptomatic; however, the manufacturers also state that the history of the cattle herd can affect the calculations. While it would be reasonable to assume that the test would demonstrate similar specificity and sensitivity in European bison as in cattle, some studies have found ELISA tests to perform significantly less well in animals other than the target species [30].Another limitation of this study is the small number of samples taken from most of the herds, which prevents any accurate of seroprevalence. However, given the endangered nature of European bison population, it is extremely difficult, or even impossible, to randomly select a given number of individuals from all herds in a country. Therefore, the best method for monitoring the contact of population with pathogens could seem to be collecting material whenever an opportunity presents itself. Such large-scale serological investigations are also convenient in other, unendangered wildlife species, as material can be collected not only during planned immobilization, but also during hunting. However, it should be noted that serological tests are not the best method for examining individual animals, especially valuable animals, such as those from the zoo.It cannot be ruled out that, due to the potential immunological cross-reactivity, for example, following contact with environmental mycobacteria, our results may have been false. Furthermore, such isolated single positive results should be interpreted with great caution in serological surveys. Interestingly, in one out of the three seropositive European bison, the MTBC complex was [7]. It cannot be excluded that this result was a false positive, and occurred as a result of MTBC infection. While little data exist on the impact of MTBC infection on the result of paratuberculosis testing, it is commonly known that M. avium subsp. paratuberculosis can influence the diagnosis of tuberculosis [31].5. ConclusionsOur findings indicate that Mycobacterium avium subsp. paratuberculosis infections does not present a major problem, especially in free-range herds. It seems that there is currently no need for continuous monitoring of this disease in the Polish European bison population.

animals : an open access journal from mdpi

10.3390/ani11082425

PMC8388701

Ruminal microorganisms, especially bacteria, play a vital role in utilizing fibrous material in ruminants. The yak is a bovid on the Qinghai-Tibet Plateau that traditionally only grazes natural pasture all year. During lactation, energy intake of yaks is often well below requirements, and yaks lose body weight. Today, to mitigate body weight losses during lactation, suckling yaks are often offered supplementary feed. This study examined the effect of dietary supplements on rumen bacteria in lactating yak. The yaks were offered supplementary concentrate feed (C), rumen-protected Lys and Met (RPA), or both (RPA+C). The ratio of the relative abundance of Firmicutes to Bacteroidetes in RPA+C was greater than in the RPA group, while there was no difference between C and RPA+C. The intakes of supplements resulted in a number of alterations in the abundances of bacteria at the genus level. When supplemented with C, yaks increased the concentration of ruminal total volatile fatty acids (VFAs), acetate, and butyrate. These results demonstrate that supplementary feed: (1) alters the composition of rumen microbiota and VFAs of lactating yaks; and (2) can be used to manipulate the composition of rumen microbiota.

Traditionally, yaks graze only natural pasture all year round without supplements. Forage intake of lactating yaks is below energy and protein requirements, even in the summer, and suckling yaks lose a substantial amount of significant body weight. Today, to mitigate the loss in body weight, supplementary feed is being offered to lactating yaks. However, the effects of supplementary feed on ruminal bacterial communities in lactating yaks is unknown. In the current study, we examined the effect of supplementary feed on ruminal microbiota, using 16S rRNA sequencing, and on volatile fatty acids (VFAs). Twenty-four lactating yaks of similar body weight (218 ± 19.5 kg) and grazing natural pasture were divided randomly into four groups and received different supplements: (1) rumen-protected amino acids (RPA); (2) concentrate feed (C); (3) RPA plus C (RPA+C); and (4) no supplements (control-CON). The concentrations of total VFAs, acetate, and butyrate were greater (p < 0.05) when supplemented with concentrate feed (C and RPA+C) than without concentrate feed (CON and RPA). Bacteroidetes (B) and Firmicutes (F) were the dominant ruminal bacterial phyla in all groups. The ratio of relative abundance of F:B in RPA+C was greater than in the RPA group, while there was no difference between CON and RPC (interaction, p = 0.026). At the genus level, the relative abundances of Absconditabacteriales_SR1, Bacteroidales-RF16-group, Bacteroidales_BS11_gut_ group, Prevotellaceae, and Rikenellaceae_RC9_gut_group were lesser (p < 0.05) with supplementary concentrate feed (C and RPA+C) than without concentrate feed (CON and RPA), whereas Butyrivibrio_2 and Pseudobutyrivibrio were greater (p < 0.05) with supplementary rumen-protected amino acids (RPA and RPA+C) than without rumen-protected amino acids (CON and C). These results demonstrate that supplementary feed: (1) alters the composition of rumen microbiota and concentrations of ruminal VFAs in lactating yaks; and (2) can be used to manipulate the composition of rumen microbiota.

1. IntroductionThere are approximately 16 million yaks (Poephagus grunniens) worldwide, with 95% in China, mainly on the Qinghai-Tibetan Plateau (QTP) [1]. They are raised between 3000 and 5000 m above sea level and are well adapted to the harsh conditions of the QTP. Yaks provide meat [2] and milk [3,4] for food, dung for fuel, and wool for clothes [5], and they serve as a cultural symbol for Tibetans [6]. They are also important in maintaining stability of the alpine ecosystem.It was reported that the meat and the milk of yak are of better quality and have higher protein content and polyunsaturated fatty acid concentrations than cattle from the lowland, for example, Holstein cattle [2,3]. However, meat and milk production in yaks is low. The reason, at least in part, is that yaks traditionally graze only natural pasture all year round without supplements [7] and, hence, are often exposed to the harsh environment in the cold season with insufficient energy and nutrient intake. As a result, yaks generally lose body weight during the harsh winter. Today, to mitigate the losses in body weight and to improve meat and milk production, supplementary feed is being offered [3].Yaks generally calve between April and June, when forage is scarce and the nutrients are inadequate. This is a time when lactating yaks require extra energy and nutrients to provide milk for the sucking calf. Therefore, lactating yaks are typically in negative energy balance [8], and intake of amino acids, in particular lysine (Lys) and methionine (Met) [9], is insufficient while grazing with calves, even during the summer season. Feed supplementation has become an effective management option for these yaks.Diet is an important factor that influences the rumen microbiome. Ruminal microorganisms ferment feedstuffs, especially carbohydrates (fibers and starch), that provide precursors for energy in ruminants [10]. Supplementary protein fed to pregnant heifers enhanced bacterial populations involved in hemicellulose and pectin degradation and in ammonia assimilation [11]. Abundances of the genera in Firmicutes decreased in Holstein heifers when supplemented with lysine (Lys) and methionine (Met) [12], while abundances of Fibrobacter succinogenes, Ruminococcus albus, Ruminococcus flavefaciens, and Methanogen archaea increased in beef cattle when supplemented with concentrate feed [13].The effect of supplementary feed on rumen microbiota of grazing, lactating yaks still remains unknown. The aim of this study was to fill this gap by examining the effect of supplementary concentrate feed and rumen-protected (RP)-Lys plus RP-Met on rumen microbiota and fermentation. Although the majority of rumen-protected amino acids avoided ruminal fermentation, portions of the rumen-protected Lys and Met were released into the rumen [14,15]. In an earlier study, microbial protein synthesis increased in response to supplementary rumen-protected Lys and Met [16].2. Materials and MethodsAll procedures in this study were approved by the Animal Care Committee of Lanzhou University (Protocol number: LZU 201805010).2.1. Study SiteThis study was done during the summer from July to August 2018 at Wushaoling Yak Research Facility of Lanzhou University (102°51.7′ E, 37°12.4′ N, altitude 3154 m above sea level), located in the northeastern part of the Qinghai-Tibetan Plateau. The dominant plant species in summer were Potentiaal fruticosa, P. viviparum, Juncus himalensis, Deschampsia cespitosa, Festuca ovina, Saussurea amara, Carex atrofusca, and C. moorcroftii. The annual mean temperature was −0.1 °C with a peak in July (average: 11.3 °C).2.2. Animals, Diet, Experimental Design, and ManagementTwenty-four lactating yaks with similar body weight (218 ± 19.5 kg) were divided randomly into four groups (n = 6 per group) and received either supplementary: (1) 15 g/day RP-Lys and 5 g/day RP-Met (RPA); (2) 1.2 kg/day concentrate feed (40% corn, 20% soybean meal, and 40% highland barley meal; C); (3) RPA and C (RPA+C); or (4) no supplements (control-CON). The rumen bypass of the two rumen-protected amino acids in the present study was ≥ 800 mg/g (data from supplier). The compositions and the chemical components of the experimental diets are presented in Table 1. The different diets were offered for 32 days, after which time, samples were collected.The lactating yaks had free access to pasture all day, and the supplementary feed was offered once daily at 18:00. The calves ran with the dams during the day and could suck milk freely. The dams were separated from the calves overnight between 18:00 in the evening and 08:00 the next day, when they were hand-milked after the calves initiated milk letdown. The milking time was about 5 min for each yak.2.3. Sample CollectionOn days 1, 15, and 30, three random pasture samples (100 g DM each) of mixed forage were collected according to Ren et al. [17]. In addition, 100 g of concentrate feed were collected. All samples were stored in self-sealing plastic bags at −20 °C for later analysis.Before the yaks were released for grazing in the morning on day 32, approximately 200 mL of rumen fluid were collected from each one using a flexible oral stomach tube (Anscitech, Wuhan, China), as described by Shen et al. [18]. The tube was cleaned thoroughly between sample collections, and the first 50 mL of fluid were discarded to minimize saliva contamination. The rumen fluid was strained through 4 layers of cheesecloth, snap-frozen immediately with liquid nitrogen, and stored at −80 °C for later analysis.2.4. Feed AnalysesForage and concentrate feed samples were weighed, air dried for 48 h, and ground to pass through a 1 mm screen. Then, dry matter (DM) content of concentrate feed and forage samples was determined by drying samples for 24 h at 105 °C in a forced-air oven (AOAC 1990), and organic matter (OM; AOAC 1990) [19] was determined as loss in dry weight upon complete combustion at 600 °C for 6 h in a muffle furnace. Total N content of concentrate feed and forage was measured by the micro-Kjeldahl N method, and crude protein (CP) was estimated as 6.25 × N. Neutral detergent fiber (NDF, assayed without a heat stable amylase and expressed inclusive of residual ash) and acid detergent fiber (ADF) of feed were estimated sequentially by a fiber analyzer (Ankom 2000; ANKOM Technology, Fairport, NY, USA), following Van Soest et al. [20] and Robertson et al. [21], respectively.2.5. Rumen Fermentation VariablesRumen fluid was centrifuged at 800× g and 4 °C for 15 min, and the supernatant was used to determine the concentrations of volatile fatty acids (VFAs). Ruminal VFAs were analyzed by gas chromatography (SP-3420, Beifenrili Analyzer Associates, Bejing, China) with a capillary column (AT-FFAP: 30 m × 0.32 mm). The procedure was started at an initial temperature of 90 °C, increased to 120 °C at a rate of 10 °C/min, held at 120 °C for 3 min, increased from 120 °C to 180 °C at the same rate, and held for 5 min [16].2.6. DNA Extraction, 16S rRNA Gene Amplification, and SequencingTotal DNA of rumen bacteria was extracted using the E.Z.N.A ® kit (Omega Bio-tek, Norcross, GA, USA), according to the protocol of the manufacturer. The concentration and the purity of the extracted DNA were determined from the 260/280 nm ratio (1.8 to 2.2) using a NanoDrop 2000 UV-vis Spectrophotometer (Thermo Scientific, Wilmington, DE, USA), and the integrity of the extracted DNA was assessed using 1% agarose gel electrophoresis. All extracted DNA samples were frozen at −80 °C for further analysis.The PCR amplification and the bioinformatic analysis of samples were done by Shanghai Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). The procedures were as follows: initial denaturation at 95 °C for 3 min, followed by 27 cycles of denaturing at 95 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 45 s, and single extension at 72 °C for 10 min, ending at 4 °C. The PCR mixtures contained 4 μL of 5× TransStart FastPfu buffer, 2 μL of 2.5 mM dNTPs, 0.8 μL of forward primer (5 μM), 0.8 μL of reverse primer (5 μM), 0. 4 μL of TransStart FastPfu DNA Polymerase, 10 ng of template DNA, and ddH2O up to 20 μL. The PCR reactions were done in triplicate. The PCR product was extracted from 2% agarose gel and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to the manufacturer’s instructions and quantified using Quantus™ Fluorometer (Promega, Madison, WI, USA).The V3–V4 of bacterial 16S rRNA gene were amplified using primers 338F (5′-ACTC CTACGGGAGGCAGCAG-3′) and 806R (5 -GGACTACHVGGG TWT CTAAT-3′). The bacterial 16S amplification and quality filter, cluster, and analysis of 16S rRNA sequencing data followed Liu et al. [22]. After amplification, purified amplicons were pooled in equimolar and paired-end sequenced on a platform (Illumina, Miseq PE300 platform/NovaSeq PE 250, San Diego, CA, USA) according to standard protocols by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). Data were analyzed using the free online Majorbio Cloud Platform (www.Majorbio.com (accessed on 28 December 2020)).2.7. Statistical AnalysesTwo factors were examined in the 2 × 2 factorial design study: (1) rumen-protected AA with two levels (0 and 15 g/day PR-Lys plus 5 g/day RP-Met); and (2) concentrate feed with two levels (0 and 1.2 kg/day). Data were analyzed by two-way analysis of variance using SAS 9.2 (SAS Inst. Inc., Cary, NC, USA) with the model: Y = μ + C + AA + (C × AA) + E, Y = dependent variable; μ = treatment mean value; C = effect of concentrate; AA = effect of rumen-protected amino acids; C × AA = interaction between concentrate and rumen-protected amino acids; E = residual error. When the C × AA interaction was significant, means between yaks with and without supplementary C in the same RP Lys and RP Met treatment groups were separated by t-test. A p < 0.05 was accepted as the level of significance.3. Results3.1. Rumen Fermentation ParametersThe ruminal concentrations of total VFAs, acetate, and butyrate were greater (p < 0.05) in yaks with supplemented concentrate feed than those without (Figure 1). Neither supplementary RPA nor concentrate (C) feed affected the ruminal concentration of propionate, and there was no interaction between RPA × C (p > 0.05).3.2. Collective Sequencing Data SummaryA total of 1,269,178 raw reads were generated from the 24 rumen fluid samples. After data processing, which consisted of filtering, quality control, assembling pared-end reads, and removal of primers, chimeras, and low-confidence singletons, 1,268,641 high quality reads with a high-quality coverage of 99.5% were used. A total of 2762 operational taxonomic units (OTUs) were detected based on 97% nucleotide sequence identity analysis among reads.In total, 2225 OTUs (Figure 2) were shared, accounting for 85.8%, 87.2%, 88.6%, and 86.7% of the total OTUs for CON, RPA, C, and RPA+C groups, respectively. The numbers of OTUs specific to CON, RPA, C, and RPA+C were 23, 14, 15, and 15, respectively.Alpha diversity indices, including Sobs, Shannon, Simpson, Chao, and coverage, were not affected by either rumen-protected amino acids supplements, concentrate supplement, or their interaction (p > 0.05; Table 2). However, ACE decreased when supplemented with concentrate (p = 0.046).3.3. Bacterial Community Composition in the Rumen FluidSix phyla were identified in the rumen fluid of the four treatments. Bacteroidetes (B; 61.5%) and Firmicutes (F; 29.3%) were the dominant phyla, with lesser abundances of Tenericutes (3.09%), Spirochaetes (1.96%), Patescibacteria (1.55%), and Proteobacteria (1.31%). There was no difference in the abundance of any phyla among the yak groups (p > 0.05; Table 3, Figure 3); however, the C and the RPA+C groups had a greater unclassified phyla (p = 0.043) than groups not receiving concentrate supplement. The ratio of F:B was greater in the RPA+C group than the RPA group, while the CON and the RPA groups did not differ (interaction, p = 0.026).A total of 22 genera of bacteria were identified in the rumen fluid in the four treatments (Table 4, Figure 4), with Prevotella the most abundant (20.4%) followed by Rikenellaceae_RC9 (9.5%) and F082 (7.6%). The relative abundances of Absconditabacteriales_SR1 (p = 0.005), Bacteroidales_RF16_group (p = 0.006), Bacteroidales_BS11_gut_group (p = 0.031), Prevotellaceae (p = 0.013), and Rikenellaceae_RC9_gut_group (p = 0.017) were lesser when supplemented with concentrate feed than when not supplemented; however, the abundances of Butyrivibrio-2 (p = 0.041) and Pseudobutyrivibrio (p = 0.022) were greater in yaks when supplemented with RPA than when not supplemented. There was no interaction detected.4. Discussion4.1. Rumen Fermentation ParametersThe increased concentration of ruminal VFAs with supplementary concentrate feed in the present study was consistent with previous studies in dairy goats [23] and Holstein-Friesian crossbred heifers [24]. This was due to the increased fermentable substrate, mainly starch, with the concentrate feed. In general, with an increase in non-fiber carbohydrate intake, there is an increase in concentration of ruminal propionate [25]. However, this did not occur in the present study, as the concentration of acetate was greater in yaks with supplementary concentrate than those without, but concentrate did not affect the concentration of propionate. The reason was probably, at least in part, due to the specific rumen microbiome in grazing yaks, which was reported to contain a greater proportion of fiborolytic bacteria than cattle [26,27]. Supplementary rumen-protected amino acids did not affect the ruminal concentration of VFAs, which is in agreement with a previous study [28].4.2. Effects of Supplementary Feeds on the Bacterial Community of the Rumen FluidSupplementing concentrate feed and/or rumen-protected Lys and Met had no effect on the alpha diversity of the rumen bacterial community, which was in agreement with Anderson et al.’s study on steers [29]. However, yaks with supplementary concentrate feed had the lowest ACE value, which was also reported in cattle grazing tropical pasture and supplemented with concentrate [13]. In a previous study on yaks, ruminal bacterial diversity increased with an increase in forage intake because of the complexity of dietary ingredients [4]. In the present study, we reasoned that the forage intake decreased in yaks when offered supplementary concentrate, which could explain the concomitant decrease in diversity of the ruminal bacterial community.Accounting for 87.1% of the total detected OTUs, Bacteroidetes and Firmicutes were the dominant phyla in the rumen bacterial community, which indicates that they play an important role in the bacterial ecology and the degradation of substrates in grazing, lactating yaks. Similarly, Bacteroidetes and Firmicutes constituted the majority of bacteria in cattle [30], dairy cows [31,32], growing yaks [33] and other lactating yaks [28].The F:B ratio is influenced by a number of factors, including dietary composition, animal species and breed, and physiological stage of the animal. It was reported that the F:B ratio generally increased when the proportion of forage intake increased [34]. However, the F:B ratio in the present study (average: 0.48) was lower than in lactating yaks raised in feedlots (average: 0.55) [28]. In the present study, the F:B ratio increased with supplementary concentrate (C and RPA+C), which was in agreement with Hu et al., who reported that increasing dietary energy increased the F:B ratio [31]. Consequently, both an increase in the proportion of forage and an increase in dietary energy can increase the F:B ratio under certain conditions.Prevotella-1 was the most dominant genera in the phyla of Bacteroidetes, followed by Rikenellaceae_RC9_gut_group, which was in agreement with previous findings in yaks [26,31,35], sheep [36], and dairy cows [37]. Zhao et al. [28], however, reported that Rikenellaceae_RC9_gut_group was the most dominant genus in the phyla of Bacteroidetes in indoor lactating yaks, followed by Prevotella-1 [28].Pseudobutyrivibrio is a butyrate-producing, fibrolytic bacteria [38], and the C group had the lowest relative abundance of this genus. This suggested that yaks in the C group had a lower proportional intake of natural forage than the control yaks and, consequently, a lesser need of fibrolytic bacteria, which could explain this difference. Yaks receiving rumen-protected amino acids (RPA and RPA+C) had higher relative abundances of Butyrivibrio-2 and Pseudobutyrivibrio than yaks without supplementary rumen-protected amino acids (CON and C), which was consistent with the greater ruminal concentrations of butyrate. Butyrivibrio_2 was reported to increase energy and nitrogen supplies [39].The relative abundance of the genus uncultured_o_Absconditabacteriales (SR1) was lesser in yaks with supplementary concentrate feed than those without. Absconditabacteriales was linked with inflammatory bowel diseases [40,41] and periodontal diseases in humans [42]. In addition, Absconditabacteriales was associated with an unhealthy udder, and it was suggested that this genus could be used as a biomarker to assess mastitis in dairy cows [43,44]. Bacteroidales RF16 group was more abundant in the control yaks than the other three groups, which was in agreement with Xue et al. [45] and Liu et al. [22], who reported the greatest abundance in grazing yaks without supplements. An earlier study reported that the relative abundance of Bacteroidales_BS_11_gut_group was greater in yaks grazing natural pasture than yaks supplemented with concentrate feed [22]. Bacteroidales_BS_11_gut_group specializes in fermenting active hemicellulose monomeric sugars (e.g., xylose, fucose, mannose, and rhamnose) to short-chain fatty acids (e.g., acetate and butyrate), which are important for ruminant energy [46].5. ConclusionsSupplementary rumen-protected amino acids had no effect on the diversity or the richness of the rumen bacterial community, but supplementary concentrate feed reduced the diversity. The supplements altered the relative abundance of dominant bacteria at the genus level but not at the phylum level. Supplementary concentrate feed decreased the relative abundance of Abscoditabateriales-SR1, Bacteroidales_RF16_group, Bacteroidales_BS11_gut_group, Prevotellaceae, and Rikenellacae RC9_ gut_group, whereas supplementary rumen-protected amino acids increased the abundances of Butyrivibrio_2 and Pseudobutyrivibrio. Collectively, Prevotella was the most abundant genus in the phylum Bacteroidetes. The present study demonstrated that supplementary feed: (1) alters the composition of rumen microbiota of lactating yaks; and (2) can be used to manipulate the composition of rumen microbiota. Data from the present study provide a basis for future research on supplementary feeds in grazing, lactating yaks on the Qinghai-Tibetan Plateau.

animals : an open access journal from mdpi

10.3390/ani11061506

PMC8224660

Oxidative stress in mammalian sperm can be induced during cryopreservation; therefore, it is important to inhibit oxidative stress to maintain sperm motility during cryopreservation. The present study was performed to investigate the effects of supplementing apigenin (AP), astragalus polysaccharide (APS), or their combination on an extender for bull semen freezing. Sperm viability, motility, and motion parameters; acrosome integrity and membrane integrity; endogenous antioxidant indices; and reactive oxygen species (ROS) and malondialdehyde (MDA) levels were evaluated after semen thawing. Our results indicated that 0.2 mmol/L AP or 0.5 mg/mL APS improved the cryopreservation of bull sperm; moreover, their combination in the extender could further improve the protective effects on bull sperm post-thaw.

The purpose of this study was to determine the effects of apigenin and astragalus polysaccharides on the cryopreservation of bovine semen. Apigenin, astragalus polysaccharides, or their combination were added to a frozen diluent of bovine semen. Afterwards, Computer Assisted Semen Analysis (CASA), membrane functionality, acrosome integrity, mitochondrial integrity, CAT, SOD, GSH-Px, MDA, and ROS detection were conducted. The results showed that adding 0.2 mmol/L AP or 0.5 mg/mL APS could improve the quality of frozen sperm. Compared to 0.2 mmol/L AP alone, the combination of 0.2 mmol/L AP and 0.3 mg/mL APS significantly increased the total motility (TM), average path distance (DAP), straight line distance (DSL), average path velocity (VAP), curvilinear velocity (VCL), wobble (WOB), and sperm CAT and SOD levels (p < 0.05), while reducing the ROS and MDA levels (p < 0.05). These results indicated that the addition of 0.2 mmol/L AP or 0.5 mg/mL APS alone has a protective effect on the freezing of bovine semen. Compared to the addition of 0.2 mmol/L AP, a combination of 0.2 mmol/L AP and 0.3 mg/mL APS could further improve the quality of frozen semen.

1. IntroductionArtificial insemination (AI) accounts for an increasing proportion of cattle reproduction [1]. The development of AI involves increasingly stringent requirements on the quality of frozen semen [2]. Although the quality of semen after thawing has improved in the past few decades, about 50% of sperm are fixed or destroyed due to freezing and thawing, resulting in a reduced fertilization ability [3]. The oxidative stress caused by the production of reactive oxygen species (ROS) during the freezing and thawing process was found to be the main reason for the decreased semen motility after thawing. As mammalian sperms lack part of the cytoplasm, their antioxidant capacity is not sufficient to resist the lipid peroxidation (LPO) caused by the ROS-induced polyunsaturated fatty acids that are bound to sperm phospholipids [4]. Due to the antioxidant enzymes in semen, such as glutathione peroxidase (GSH-Px), catalase (CAT), and superoxide dismutase (SOD), concentrations are very important. As a mammalian sperm cytoplasm has a relatively high concentration of polyunsaturated fatty acids (PUFA), it lacks sufficient natural antioxidant reserves. Therefore, the addition of natural antioxidants to the frozen semen dilution has become necessary to improve sperm motility after thawing. Antioxidants can balance ROS and reduce LPO levels to improve the freezing resistance of semen and improve the quality of melted semen. Antioxidants have been widely used in frozen semen [4,5,6]. Plant-derived compounds are considered to be important sources of antioxidants because they produce almost no side effects and are easily available in nature [7].Flavonoids are widely present in nature. These compounds act as free radical scavengers and antioxidants, and have antimutagenic, anti-inflammatory, and antiviral effects [8,9,10,11]. Apigenin (AP) (APEBIO Co., Ltd., Houston, Texas, USA), a flavonoid, is widely distributed in a plant [10]. Compared to that of other structurally related flavonoids, it has a lower inherent toxicity to normal cells [12,13]. AP penetrates cells and binds to DNA to form an apigenin-DNA complex, protecting the DNA from oxidative stress in the cell [14]. Recent studies have demonstrated that adding AP to a frozen diluent of swine semen could increase sperm motility and the antioxidant enzyme content and reduce the concentration of the LPO product malondialdehyde after thawing [2]. Astragalus polysaccharide (APS) (Shanghai Yuanye Biological Co., Ltd., Shanghai, China) can be considered as an antioxidant as there is research that shows that APS can inhibit mitochondrial damage by scavenging ROS [15]. Recent studies have found that APS could reduce the oxidative damage of cardiomyocytes caused by free radicals and lipid peroxides as well as reduce membrane damage. Furthermore, APSs have recently been found to have antibacterial and antiviral properties and can be used as immune enhancers [16]. Due to the various properties of APS, it can be used for the cryopreservation of pig semen, and studies have shown that it can improve semen vitality and antioxidant levels [17].Different antioxidants have a synergistic effect on the removal of active oxygen [18,19]. AP and APS, as two different antioxidants, have been shown to improve semen quality through different action pathways and action times [17,20]. However, neither of the two substances have been used before in the freezing of bovine semen. The purpose of this study was to explore the effects of AP, APS, and their combination on the freezing of bovine semen.2. Materials and Methods2.1. Experimental DesignIn this study, pooled semen was extended using a Tris extender with different levels of AP (0, 0.2, 0.4, 0.6, and 0.8 mmol/L) and APS (0, 0.1, 0.3, 0.5, and 0.7 mg/mL) or a combination of AP and APS (0.2 mmol/L AP, 0.2 mmol/L + 0.1 mg/mL, 0.2 mmol/L + 0.3 mg/mL, 0.2 mmol/L + 0.5 mg/mL, and 0.2 mmol/L + 0.7 mg/mL). Each experiment was repeated at least three times.2.2. AnimalsThis study was conducted at Jilin Agricultural University in P.R. China. Experimental Animal Welfare and Ethics Committee of Jilin Agricultural University approval was gained, the number is 20200803001. Four bulls (4 years old) were used for semen collection. The uniformity of the feed, housing, and light conditions were ensured. The bull feed was composed of 45% corn, 32% soybean cake, 5% wheat bran, 5% rice bran meal, 6% soybean germ meal, 2% molasses, and 5% bull special premix. In terms of nutritional indicators, the bull feed had approximately 20% crude protein, 13% moisture, 7% crude ash, 0.6–0.7% calcium, 0.6–0.7% total phosphorus, and 0.8–1% sodium chloride. The bulls had free access to water and salt, and no additional antioxidants were present in the feed.2.3. Bull Semen CollectionSemen samples were collected using an artificial vagina twice a week per bull, for 12 weeks. The criteria for cryopreservation were as follows: the semen samples were sent to the laboratory within 30 m. The semen was held in a water bath at 37 °C while the sperm concentration was estimated using a calibrated spectrophotometer and the motility of sperm was subjectively evaluated using microscopy, at a concentration of at least 1 × 109 spermatozoa/mL, a sperm motility ≥70%, and an abnormality ≤15%. Healthy ejaculates were used in the experiments. After the initial assessment, the semen samples were mixed to eliminate individual differences.2.4. Basic ExtenderThe basic extender was slightly modified from that described by Tarig et al. [21], as shown in Table 1. All of the chemicals were placed into a capacity bottle and double evaporative water was used to stabilize the capacity to 100 mL, a magnetic stirrer was used to stir for 2 h, then 20% egg yolk and 3% glycerin were added and stirred again for two hours to completely dissolve. The basic diluent was divided evenly and AP or APS of different concentrations were added.2.5. Semen ProcessingAfter the quality evaluation, the samples were diluted and incubated in a water bath at 37 °C for 30 min for the full absorption of AP and APS by the sperm membrane. After that, the samples were packed in 0.25 mL straws (IMV Co., Ltd, L’Aigle, Paris, France) with 8 × 106 sperm/straws, then cooled from 37 °C to 4 °C for 2 h, as previously described, and subsequently cooled from 4 °C to −140 °C for approximately 8 min, using a turbo freezer (Minitube Co., Ltd, Munich, Bavaria, Germany). After that, the straws were transferred to a liquid nitrogen tank (−196 °C) and stored for 1 month before detection.2.6. Evaluation of Post-Thawed Sperm2.6.1. Computer Assisted Semen AnalysisThe kinematic parameters of sperm were analyzed using a sperm analyzer (developed jointly by Hamilton and IMV, IVOS II, 10871, 6 October 2017). Two straws were thawed by immersion in a water bath at 37 °C for 30 s. Each sample was analyzed at least three times. Five microliters were detected for each straw, and four fields were randomly examined. The total motility (TM), average path velocity (VPA), straight line velocity (VSL), curvilinear velocity (VCL), amplitude of lateral head displacement (ALH), beat/cross frequency (BCF), straightness (STR), linearity (LIN), wobble (WOB), distance of average path (DAP), straight line distance (DSL), and curvilinear distance (DCL) were recorded.2.6.2. Acrosome IntegrityThe sperm acrosome integrity was slightly modified from that described by Masoudi et al. [22]. The sample was centrifuged, and the resultant sperm pellet was obtained. It was then equilibrated in 96% ethanol for 10 min. Afterwards, the sperm was placed on a glass slide and fluorescein isothiocyanate-conjugated pea lectin (PSA-FITC) (sigma) was added. The slides were incubated for 20 m, and glycerol was added to fix the sperm onto the slide. At least 5 fields of view were observed through a fluorescence microscope (400× magnification) (Cytation 5 imaging reader, Bio Tek Co., Ltd, Winusky, Vermont, USA). The percentage of sperm with abnormal acrosomes was recorded by counting a total of 200 sperm under visualized microscopically. (see Figure 1).2.6.3. Plasma Membrane IntegrityThe method of measuring sperm plasma membrane integrity used was slightly modified according to the method in an article published by R.A. Harrison et al. [23]. Using the double staining method of carboxy fluorescein diacetate (CFDA) (Andy Forno BiotechnologyCo., Ltd. Wuhan, China) and propidium iodide (PI) (Coolaber), 0.46 mg of CFDA was dissolved in 1 mL of dimethyl sulfoxide and 0.5 mg of PI was dissolved in 1 mL of normal saline. It was stored at −20 °C and protected from light. Afterwards, 20 μL of CFDA and 10 μL of PI were dissolved in 1 mL of PBS with a pH of 7.4 and 0.01 mol/L. A 20-microliter semen sample was then taken and an 80-microliter staining solution was added. The cells were incubated at 37 °C for 10 min and washed with PBS. After washing, a 10-microliter sample and was taken and at least 5 fields of view were observed through a fluorescence microscope (400× magnification) (Cytation 5 imaging reader, Bio Tek Co., Ltd, Winusky, Vermont, USA). The percentage of sperm with abnormal acrosomes was recorded by counting a total of 200 sperm under visualized microscopically. (see Figure 2).2.6.4. Mitochondrial ActivityThe method of measuring sperm mitochondrial activity was slightly modified according to the method in an article published by Shahverdi et al. [24]. The mitochondrial activity of sperm was evaluated using double fluorescent staining with rhodamine 123 (Rh123) (Coolaber) and PI. Rh123 was dissolved with Me2SO and stored in the dark, while PI was dissolved in a physiological saline. The semen was then mixed with Rh123 and PI, incubated for 20 min, taken out and centrifuged to remove the supernatant, and then was added to PBS to mix the semen. Afterwards, the sample was centrifuged again to wash away the impurities and at least 5 fields of view were observed through a fluorescence microscope (400× magnification) (Cytation 5 imaging reader, Bio Tek Co., Ltd, Winusky, Vermont, USA). The percentage of sperm with abnormal acrosomes was recorded by counting a total of 200 sperm under visualized microscopically. (see Figure 3).2.6.5. Endogenous Antioxidant Indices Detection in the Frozen–Thawed SemenThe determination of various endogenous antioxidant enzymes (superoxide dismutase, glutathione peroxidase, and catalase) was carried out using an enzyme-linked immunoassay kit (Shanghai Enzyme Biotechnology Co., Ltd., China), according to the manufacturer’s instructions. The sample to be tested was mixed with a sample diluent. After incubation, the enzyme-labeling reagent and stop solution were added in sequence. The resulting mixture was then placed into a microplate reader for analysis.2.6.6. MDA and ROS Concentration Determination in Post-Thawed SemenA bull MDA and ROS ELISA assay kit was used to determine the concentrations of MDA and ROS (Shanghai Enzyme Biotechnology Co. Ltd., China). The bovine semen was centrifuged repeatedly to destroy the sperm’s structure. Next, 40 μL of diluent was added to the sample, and placed onto a shaker to mix the semen and diluent thoroughly. It was then incubated in a 37 °C incubator for 30 min, washed with a washing solution 5 times, 50 μL of enzyme-labeled reagent was added, incubated again in a 37 °C incubator for 30 min, and washed 5 times after equilibration. Two kinds of diluent shake were then added and mixed well. The entire procedure was performed for 10 min at 37 ℃ in the dark. After adding the stop solution, the absorbance was measured using a microplate reader at a wavelength of 450 nm, and the content of MDA and ROS was calculated according to the standard curve.2.7. Statistical AnalysesAll results were expressed as mean ± SEM. The mean values of the sperm characteristics, movement characteristics, GSH-Px, CAT, SOD, ROS, and MDA concentrations, acrosome integrity, mitochondrial activity, and membrane integrity were compared using Statistical Product and Service Solutions (SPSS 22.0; SPSS, Chicago, IL, USA) Duncan’s multiple range test by ANOVA procedure; p < 0.05 was considered statistically significant.3. Result3.1. Effect of Different Concentrations of AP and APS on Several Kinematic Parameters of Bovine Semen Samples after Thawing3.1.1. Effect of Different Concentrations of AP on Several Kinematic Parameters of Frozen Bovine SemenThe results are shown in Table 2. Adding different concentrations of AP could improve the kinematic parameters of frozen semen. A concentration of 0.2 mmol/L significantly improved the kinematic ability of the TM, DAP, DSL, VAP, VSL, LIN, and WOB (p < 0.05). Simultaneously, 0.2 mmol/L AP relieved the DCL and VCL levels (p < 0.05). Although adding 0.4 mmol/L AP also had a certain effect, it was not as good as the state of the sperm when 0.2 mmol/L was added.3.1.2. Effect of Different Concentrations of APS on Several Kinematic Parameters of Frozen Bovine SemenThe results are shown in Table 3. Adding different concentrations of APS improved the kinematic parameters of semen freezing. When the concentration was 0.5 mg/mL, the sperm’s TM, DAP, DSL, VAP, VSL, LIN, WOB levels were significantly increased, and the DCL and VCL levels were simultaneously relieved (p < 0.05). Although the addition of 0.5 and 0.7 mg/mL APS was also significantly different from the control group in some kinematic parameters (p < 0.05), generally, the best effect was achieved when 0.5 mg/mL APS was added.3.2. Effects of AP and APS at Different Concentrations on Sperm Plasma Membrane Integrity, Acrosome Integrity, and Mitochondria Activity after Thawing3.2.1. Effects of AP at Different Concentrations on Sperm Plasma Membrane Integrity, Acrosome Integrity, and Mitochondria ActivityThe results of plasma membrane, acrosome integrity, and mitochondria activity assays following a supplementation with 0, 0.2, 0.4, 0.6, and 0.8 mmol/L AP are summarized in Figure 4. The mitochondria activity was significantly higher in the 0.2 mmol/L group (p < 0.05), and both the plasma membrane and the acrosome integrity were also improved in the AP 0.2 mmol/L (p < 0.05) group, when compared to the others.3.2.2. Effects of APS at Different Concentrations on Sperm Plasma Membrane Integrity, Acrosome Integrity, and Mitochondria ActivityThe results of plasma membrane, acrosome integrity, and mitochondria activity assays following a supplementation with 0, 0.1, 0.3, 0.5, and 0.7 mg/mL APS are summarized in Figure 5. The mitochondria activity, plasma membrane, and acrosome integrity were significantly higher in the 0.5 mg/mL group (p < 0.05).3.3. Antioxidant Enzymes, ROS, and MDA Content in Semen after Thawing3.3.1. Effect of AP on the Antioxidant Enzyme Level in SpermThe results of the CAT, GSH-Px, and SOD assays after semen cryopreservation in 0, 0.2, 0.4, 0.6, and 0.8 mmol/L AP are shown in Figure 6. The CAT, GSH-Px, and SOD levels in the AP 0.2 mmol/L groups retained better activity (p < 0.05) than the other groups.3.3.2. Effects of APS on the Antioxidant Enzyme Level in SpermThe results of the CAT, GSH-Px, and SOD assays after semen cryopreservation in 0, 0.1, 0.3, 0.5, and 0.7 mg/mL APS are depicted in Figure 7. The CAT, GSH-Px, and SOD levels in the APS 0.5 mg/mL group were higher (p < 0.05) than those in the other treatment groups.3.3.3. Effects of the Addition of Different Concentrations of AP on the Oxidation Products of Bovine SemenThe results of the ROS and MDA assays after sperm cryopreservation in 0, 0.2, 0.4, 0.6, and 0.8 mmol/L AP are shown in Figure 8. The ROS level was significantly reduced in the AP 0.2 mmol/L group (p < 0.05) compared to that in the others, and this group also demonstrated a significant reduction in MDA levels (p < 0.05).3.3.4. Effects of the Addition of Different Concentrations of APS on Oxidation Products of Bovine SemenThe results of the ROS and MDA assays in sperm samples cryopreserved in 0, 0.1, 0.3, 0.5, or 0.7 mg/mL APS are shown in Figure 9. Both the ROS and MDA levels were reduced in the APS 0.5 mg/mL groups compared to those in the others (p < 0.05).3.4. Effect of the Combined Use of AP and APS on Several Kinematic Parameters of Bovine Semen Samples after ThawingEffects of AP and APS on the Kinematic Parameters of Frozen Bovine SemenThe results are shown in Table 4. A combined addition further improved the kinematic parameters of frozen semen compared to that with 0.2 mmol/L AP alone. A combination of 0.2 mmol/L AP and 0.3 mg/mL APS significantly improved the TM, DAP, DSL, VAP, and WOB kinematic abilities. It also simultaneously relieved the VCL level (p < 0.05).3.5. Effects of the Combined Use of AP and APS on Sperm Plasma Membrane Integrity, Acrosome Integrity, and Mitochondria Activity after ThawingEffect of the Combined Use of AP and APS on Plasma Membrane Integrity, Acrosome Integrity, and Mitochondria ActivityA combined AP and APS treatment further improved the mitochondria activity, sperm plasma membrane integrity, and acrosome integrity in frozen semen compared to an addition of 0.2 mmol/L AP alone (Figure 10). A combination of 0.2 mmol/L AP and 0.3 mg/mL APS significantly improved the mitochondria activity, sperm plasma membrane integrity, and acrosome integrity of the semen (p < 0.05).3.6. Antioxidant Enzymes, ROS, and MDA Content in Semen after Thawing3.6.1. Effects of the Combined Use of AP and APS on Antioxidant Enzymes in SpermCombined supplementation resulted in a further improvement in the CAT, GSH-Px, and SOD activity in frozen semen when compared to 0.2 mmol/L AP supplementation alone (Figure 11). A combination of 0.2 mmol/L AP and 0.3 mg/mL APS significantly improved the SOD and CAT activity in these samples (p < 0.05).3.6.2. Effect of the Combined Use of AP and APS on ROS and MDA LevelsThe combination of AP and APS further reduced both the ROS and MDA levels in frozen semen samples compared to 0.2 mmol/L AP alone (Figure 12), and a combination of 0.2 mmol/L AP and 0.3 mg/mL APS significantly reduced the ROS and MDA levels in these semen samples (p < 0.05).4. DiscussionAs oxidative stress affects the quality of frozen semen, adding antioxidants has become the main method used to improve sperm quality after thawing. Our results show that adding AP and APS could improve the quality of frozen semen. AP and APS were added in combination, and the quality of semen was further improved, compared to when they were added separately.Bacterial contamination is one of the main factors causing the quality of frozen semen to decline. Recent research has shown that AP has antibacterial, antiviral, antifungal, and antiparasitic properties [25]. Consistent with the results of this study, we found that AP could reduce bacterial contamination and improve the kinematic parameters and abnormal semen movement in frozen semen. MDA is the final product of LPO. Low levels of MDA can improve the quality of frozen semen. A recent study found that AP could increase the CAT content in pig semen and reduce MDA levels [2]. In this study, it was found that adding AP could improve the CAT in frozen bovine semen and contribute to reducing the content of MDA. Overproduction of ROS can cause oxidative stress. Research shows that apigenin can show the chemopreventive potential of cancer cells by regulating the intracellular accumulation of ROS in lung cancer cells and the expression of antioxidant enzymes [26]. In this study, it was found that AP could reduce the ROS content in frozen semen and increase the expression of various antioxidant enzymes. It was also observed in previous studies that treatment with AP attenuated the reduction in superoxide dismutase activity and glutathione levels and reduced the levels of reactive oxygen species and malondialdehyde in rat hepatic stellate cells [27]. These findings are similar to the results of this study. Oxidative stress can cause sperm motility to decrease after thawing. An increasing number of studies have shown that the imbalance between the production of ROS and the level of antioxidants in sperm is the main cause of oxidative stress. As the main source of ROS is the mitochondria, mitochondrial damage caused by oxidative stress will lead to changes in the level of ROS in sperm, leading to the loss of sperm motility after thawing [28]. Mitochondria are the main energy supply sites for cells. Recent studies have found that, compared to the control group, adding APS to semen diluent can significantly increase the mitochondrial membrane potential of boar semen and reduce the level of ROS and malondialdehyde content [18]. In this study, it was found that the mitochondrial integrity of frozen semen was improved after the addition of APS, and the MDA and ROS levels were reduced. Previous research found that another way for sperm to gain energy in a diluted solution is through glycolysis. APS can provide energy for sperm in this way. It protects the mitochondria by removing reactive oxygen species and increasing the antioxidant capacity, and ultimately improves sperm energy metabolism [19]. Our results observed that the content of antioxidant enzymes such as SOD in thawed semen has been improved by the addition of APS. This result may be due to astragalus polysaccharides protecting the mitochondria of semen and improving the energy metabolism of semen. In some studies, it was found that APS could protect cells by improving cell viability, reducing apoptosis, and inhibiting the production of inflammatory cytokines [29]. APS has a protective effect on cells and can inhibit inflammatory cytokines, but its protective effect on bovine semen is unknown. It was observed that after thawing, compared to the control group, the group with APS added exhibited an increase in the vitality of sperm. Some studies have shown that APS can reduce the expression of Aspergillus toxin A, promoting apoptosis-related proteins and pro-inflammatory cytokines [30]. APS can reduce the expression of pro-inflammatory cytokines, and we found that APS can improve the inflammatory damage caused during freezing and thawing. Through observation under a microscope sperm analyzer, it was found that the multiple kinematic parameters of sperm were improved compared to the control group. The circle movement was also relieved. In some studies, APS provided effective protection in various disease models related to oxidative stress for heart, brain, kidney, intestine, liver, and lung damage [31]. A previous study showed that the use of APS in vitro can enhance sperm motility and improve testicular toxicity and has proven its potential to treat male infertility [32]. APS can not only improve the quality of semen but also has great potential in the treatment of infertility. Similar to previous studies, it was found that the addition of APS could increase the acrosome and plasma membrane integrity and other indicators of semen by adjusting the content of various antioxidant enzymes in frozen semen.Some studies have shown that different antioxidants have a synergistic effect on the removal of active oxygen. This is because they have different binding speeds with active free radicals and different ways of action [33]. Due to the different stability that results from a combination of antioxidants, the synergistic effect of two different antioxidants can further improve the various indices of semen after thawing compared to the addition of a single antioxidant alone. As expected, our study found that combined addition of antioxidants significantly improved the quality of semen compared to a single addition, and the index with no significant difference was also improved compared to a single addition. This result confirms the feasibility of adding antioxidants in combination. In some studies, a combination of curcumin and dithioerythritol can improve the integrity of semen plasma membrane and increase the expression of glutathione peroxidase [34]. It has also been found in recent studies that, AP, as a ketone antioxidant similar to curcumin, can be used in combination with ferulic acid to improve the quality of pig semen. APS is a plant polysaccharide that is mild and non-irritating. Some studies have shown that the combined use of APS and polysaccharide peptides can immunosuppress and immunomodulate lung cancer in mice [35], resulting in low toxicity and no mutagenesis. Upon combining AP and APS, it was observed that the combination of the two antioxidants could improve the quality of semen without other side effects. DNA integrity was vital for evaluating the spermatozoa quality. Assessing the integrity of DNA is valuable for the identification of seriously damaged spermatozoa. Our experiment was mainly focus on evaluating the effects of two antioxidants on sperm motility, kinematic parameters, and various antioxidant enzyme indexes after freezing and thawing. Therefore we ignored the assessment of DNA integrity and we did not detect DNA integrity. Ignoring the detection of this intrinsic indicator is also a limitation of this experiment. In addition, adding these substances could improve the vitality after freezing; however, this does not imply that they can improve the fertilization ability, because there is no measurement of the conception rate or other indicators, which is a limitation of this research.5. ConclusionsIn conclusion, adding AP and APS alone could significantly improve the kinematic parameters of sperm, acrosome integrity, plasma membrane integrity, and mitochondrial activity. However, a combined addition could further improve the kinematic parameters of sperm, the integrity of the acrosome, plasma membrane, and mitochondria. These results may be due to the addition of AP and APS alleviating the level of ROS and MDA as well as increasing the enzymatic activity of CAT, SOD, and GSH-Px.

animals : an open access journal from mdpi

10.3390/ani13081318

PMC10135255

A global decline in the population of bee pollinators is regarded as a potential threat to species extinction and global food security. Mite infestation plays a vital role in contributing to the collapse of bee populations. However, the correlation between bee population and mite infestation remains unclear. This study investigated Tropilaelaps mercedesae mite infestations to the larval, pupal, and crippled adult stages of honey bee Apis mellifera, the relationship between mite infestation rate and injury numbers for each of bee larvae and pupae, and the relationship between mite infestation rate and population size per beehive. Mite infestations occurred in all developmental stages of the honey bees, and it was pronounced in the abdomens and the antennas of the honey bees. Mite infestation rate was positively correlated with the number of injuries per bee in each of the larvae and pupae and negatively correlated with the population size per beehive. Overall, our findings suggested that the use of a large population size of beehives can reduce the infestation rate. It also provided important information about the adaptation of mite/antibacterial immune competence of honey bees to different life stages and the breeding stock of bees for hygienic behaviors resisting mite infestations.

Tropilaelaps mercedesae, one of the most devastating parasitic mites of honey bee Apis mellifera hosts, is a major threat to honey products by causing severe damage to honey bee colonies. Here, we recorded injury numbers caused by T. mercedesae to different body parts of the larval, pupal, and crippled adult stages of honey bee A. mellifera. We evaluated the relationship between infestation rate and injury numbers per bee for both larvae and pupae. We also noted the total bee numbers per beehive and examined the relationship between the infestation rate and population size. T. mercedesae infested all developmental stages of honey bees, with the highest injury numbers in the abdomens of bee pupae and the antennas of crippled adult bees. Although larvae received more injury numbers than pupae, both infestation rate and injury numbers decreased as the larval stage progressed to the pupal stage. The infestation rate increased as the population size per beehive decreased. This study provided new perspectives to the understanding of changes in the effects of T. mercedesae infestations on different developmental stages of honey bees. It also showed useful baseline information for screening honey bee stock that might have high defensive behaviors against mite infestation.

1. IntroductionBee pollinators, especially Apis species, are the key players in the crop yield process and important vectors in maintaining the natural balance of ecosystems. A consistent decline in the populations of honey bees (also known as Apis species) and other bee species has been demonstrated worldwide, causing a potential threat of species extinction and a risk to global food security [1,2,3,4]. Parasite infestation is one of the most critical factors that have led to bee population decline worldwide [5,6]. Previous studies have postulated that parasitic mites in the genus Tropilaelaps and Varroa are the main factors that severely damage honey bee colonies in the continent of Asia [7,8,9,10,11,12,13].Tropilaelaps, a genus of parasitic mites in the family Laelapidae, is often regarded as a major threat to honey products by causing severe damage to honey bee colonies in Asia [11,14,15,16,17]. Tropilaelaps (e.g., T. mercedesae) mites originally parasitized the brood of the Asian giant honey bee Apis dorsata [9,18]. These mites have transferred to the European honey bee A. mellifera [9,19] and have been found to infest both temperate and tropical populations of A. mellifera in Asia [8,9,10,20]. In general, the life cycle and reproductive strategies of T. mercedesae are similar to that of Varroa mites. They both feed on the brood of honey bees and are vectors for the Deformed Wing Virus (DWV) [9,21]. However, unlike Varroa mites, T. mercedesae mites have smaller size, more rapid locomotion, and a greater reproductive rate [9,20,22,23], and because of these, their population growth can be even faster than those of Varroa mites [20,24]. Moreover, the life span of T. mercedesae mite is shorter than that of Varroa [16,25,26]. Thus, T. mercedesae mites spend most of their life in the capped brood cells of their host and infest honey bees by sucking out the hemolymph of developing honey bees in the colony [9,27], and thus they consume the body resources of bees. Previous studies suggested that the major impact of Tropilaelaps mite infestation is caused by the mite itself, reducing bee host immune responses (e.g., [28]). Honey bee’s hemolymph is composed of several nutrients (e.g., carbohydrates and proteins) and different hemocyte types which are crucial for the immunity of honey bees [29,30,31,32]. However, A. mellifera lacks behavioral mechanisms to defend against T. mercedesae mite infestations [16]. For this reason, like Varroa destructor, T. mercedesae mite infestations often cause both abnormal brood development and brood death in honey bee A. mellifera such as deformed pupae and adults (e.g., stunting, damaged abdomens/antennas/legs/wings), parasitic mite syndrome, and colony health [8,19,21,27,33]. These effects are regarded to be developed through mites feeding on the hemolymph of the bee and also by spreading honey bee viruses, particularly DWV [34,35]. Mite infestations have also been found to reduce protein concentrations, longevity, and weight in the pupal stage of honey bees [8,27,36]. A previous study suggested that either a decrease in carbohydrate and fat contents or an increase in protein content and defense mechanisms of the immune systems of A. mellifera was accompanied by honey bees at different developmental stages from the larval to adult stages [37]. Therefore, mite infestation is likely to vary across different life stages of the bees depending on their nutrient compositions and defense mechanisms of immune systems against the infestation. If nutrient compositions and defense mechanisms or hygienic behaviors against mite infestations are weak in the early developmental stage of the honey bees, the infestation rate (the proportion of infested bees per hive) caused by T. mercedesae should be higher in bee larvae than in bee pupae and adult bees. However, the potential for variations in infestations caused by T. mercedesae in the different developmental stages of the honey bees is rarely studied.It has been postulated that honey bees infested by T. mercedesae during the early developmental stage enhance viral proliferation in the beehive through long exposure to the virus and the stress on susceptibility to viral infection rates [38]. Thus, variations in mite infestation rates may influence the number of injuries of honey bees across different developmental stages, which would lead to the spreading of the viruses and further affect the colony’s health and honey products. Moreover, injury numbers may vary among different body parts of the bees, as each body part is structured with different organs and functions differently [8]. A previous study on Thailand populations of A. mellifera suggested that the number of injuries was positively correlated with the number of actively feeding mites in both the fifth instar larvae and pupae of the bees. Among different body parts of the bees, the highest injury numbers were found in the abdomens and the antennas, whereas the lowest injury numbers were found in the thoraxes and the mouth parts [8]. However, it remains unknown whether the T. mercedesae infestation rate is correlated with the injury numbers on the brood and the population size (the total number of bees per hive) of A. mellifera bees. It is postulated that abiotic factors (e.g., temperature and humidity) have a great impact on the population level, survivorship, and fecundity of mites, such that high temperatures in the colonies could reduce mite growth [39,40]. A previous study also noted that bees with large colonies increased the temperature level of the colonies when compared to those with small colonies [41]. In addition, the hygienic behaviors of the bees have been found to influence mite infestations [40]. Therefore, bees with small population sizes, relative to those with large population sizes, are expected to have a higher infestation rate and injury numbers in the body parts of the individual bees due to the potential for the reduction in relative temperature and humidity, and hygienic characteristics of the colonies. If bee pupae had the highest number of injuries in the abdomens and antennas [8], the greatest number of injuries should also occur in the abdomens and antennas of adult bees and the lowest in the thoraxes and mouth parts, as they both have almost identical morphological features, however, this remains unclear.In the present study, we documented the number of injuries (both fresh wounds and scars) caused by T. mercedesae to the bee larvae (i.e., fifth instar larvae and prepupae) and different body parts (i.e., abdomens, antennas, thoraxes, mouth parts, legs, and wings) of the pupal and crippled adult stages of A. mellifera. We also examined the T. mercedesae mite infestation rate, the correlation between mite infestation rate and the number of injuries per bee for both larvae and pupae and the relationship between mite infestation rate and bee population size. We hypothesized the highest injury numbers in the abdomens and the antennas of the bees, the lowest injury numbers in the thoraxes and mouth parts of the bees, a positive correlation between mite infestation rate and injury numbers for both bee larvae and pupae, and a positive relationship between the infestation rate of mites and the population size of honey bee A. mellifera.2. Materials and Methods2.1. Study Sites and Source of Honey Bee SamplesSamples of the larval and the pupal stages of honey bees (Apis mellifera) from hives were collected at Chan Chawa in the Chiang Rai province of Northern Thailand and that of newly emerged crippled adult bees from hives were collected in Chiang Mai city of Northern Thailand. Both field and laboratory observations of injury numbers of the bee larvae and pupae were performed between May and July 2017, while that of the adult honey bees was evaluated between February 2017 and March 2018. In this study, we used a queen-right Langstroth hive of a honey bee colony consisting of eight to ten movable frames. Each frame contained different developmental stages of bees such as 5th instar larvae, prepupae, white-eyed pupae, pink-eyed pupae, purple-eyed pupae, older pupae, and crippled adult bees. Fifth instar larvae and prepupae were defined as the larval stage of honey bees, whereas white-eyed pupae, pink-eyed pupae, purple-eyed pupae, and older pupae were defined as the pupal stage. We used fifteen randomly selected beehives for bee larvae and pupae as well as eight randomly selected beehives for adult bees to examine injury numbers. Before the study, all selected beehives were previously infested with mites. Injury numbers of larvae and pupae were examined at the laboratory of the School of Life Science at Mae Fah Luang University, while that of adult bees were explored at Bee Protection Laboratory at Chiang Mai University. All beehives and the observation of mite infestations at each developmental stage of bees were monitored on the same day or one close to the day to minimize variations in the number of mite infestations among bees or beehives.2.2. Examination of Population Size, Injury Numbers, and Infestation Rate in Larvae and PupaeWe examined the population size of adult honey bees following the method developed by Delaplane et al. [42]. Briefly, we visually estimated the percentage of adult bees on both sides of a comb of all individual frames in each colony. All percentages were pooled as a total percentage, then the population size was calculated by multiplying the total percentage of adults by bee numbers per fully occupied side of the Langstroth.To observe injuries caused by Tropilaelaps mites to the larvae and different body parts (i.e., abdomens, antennas, legs, mouth parts, thoraxes, and wings) of the pupae, we selected one frame from the middle Langstroth hive for each of the fifteen beehives. From each selected brood frame, we randomly collected 35 infested larvae and pupae and recorded their injury numbers and locations under a stereo microscope. In this study, fresh wounds (a fresh injury to the skin or body tissue of the honey bees) or scars (a mark left on the skin or within body tissue where a wound has not healed completely, and fibrous connective tissue has developed) caused by T. mercedesae mites to any parts of a bee’s body was defined as injury. Before the examination of injured bees, all infested larvae and pupae were extracted from brood cells with blunt-pointed dissection forceps (Union Science Ltd., Muang Chiang Mai 50200, Thailand), taking extreme care to avoid damage to any part of their body. All signs of wounding in developmental stages were considered as damaged/injured brood which then was removed by destroying the surrounding wax cell and lifting the brood with the forceps. Moreover, the wounds on the broods were confirmed by immersing the brood in 0.4% trypan blue solution (Invitrogen, Carlsbad, CA, USA) for 30 min at room temperature according to the method developed by Ganbar and Engels [43]. This chemical solution stains the damaged epidermal cells of the bees caused by Tropilaelaps mites, enabling us to detect injury locations and numbers of both larvae and pupae. Fresh integumental wounds appeared as blue spots, while wounds that had healed and appeared brown to black were referred to as scars (Figure S1). The examination of injuries was conducted under a stereo microscope. Then, the number of injuries (both fresh wounds and scars) was recorded according to the corresponding location.We justified our initial determination of mite infestation rate by uncapping 300 randomly selected worker brood cells per colony. Regardless of the number of mites in the cell, it was considered as one infested cell. The infestation rate was then calculated by dividing the total number of infested cells by the total number of both infested and uninfested cells for each frame in each beehive. The brood cells were collected from two to three randomly selected frames per beehive. A total of 35 frames from 15 beehives were used for the determination of the infestation rate.2.3. The Observation of Injuries on Crippled Adult Honey BeesSymptomatic bees with wing deformities or with normal wings that were small and sluggish [44] were observed in each of eight randomly selected beehives before the examination of injury numbers caused by T. mercedesae mites to newly emerged honey bees. We randomly collected 243 newly emerged crippled/mite-infested adult bees from eight monitored beehives using blunt-tipped forceps, taking extra care to avoid damage to bee bodies. We then recorded injury numbers that occurred in different body parts of crippled adult bees under a stereo microscope (an illustration of injury types in different body parts of the honey bees is shown in Figure S2).2.4. Scanning Electron Microscope (SEM)Based on the observation of injuries at different developmental stages of honey bees, we randomly selected Tropilaelaps-infested pupae and crippled adults for scanning electron microscope experiments. The pupal and adult tissues (i.e., wings and antennas) of honey bees were fixed into 1.0 mM glutaraldehyde in cacodylate-buffered, pH 7.2 for two hours. The fixed samples were vacuumed for better penetration for 2 h and then rinsed in the buffer for 30 min. After stepwise dehydration in the ethanol, we applied the critical point for CO2. The dried pupal and adult tissues were mounted and sputtered with gold palladium. The photographs of injured tissues were taken using the scanning electron microscope (SEM: JSM-IT300) [45].2.5. Data AnalysesAll statistical analyses were performed in R version 4.1.2 (R Development Core Team, 2022), and data were expressed as the mean ± standard deviation. Independent sample t-tests were used to test for significant differences in injury numbers either between the larval and pupal stages of honey bees or between 5th instar larval and prepupal stages of the bees. A generalized linear model (GLM) with Poisson errors was used to test the differences in the number of injuries among body parts (abdomens, antennas, thoraxes, legs, mouth parts, and wings) and types of injury (fresh wounds and scars) for bee pupae. We divided neither body parts nor injury types when analyzing injury numbers of the bee larvae. The interaction between the body parts and injury types was also included. When significant differences were found, differences between levels of each effect were analyzed using multiple comparisons of means with Tukey contrasts using a “glht” function in the “multcomp” package [46]. We used GLM with Binomial errors to test for the relationship between the infestation rate and injury numbers for each of the larval and pupal stages as well as the relationship between infestation rate and population size. Kruskal–Wallis tests were employed to compare injury numbers among different body parts of the adult honey bees. We did not find any injuries on the thoraxes and the mouth parts of adult bees and only a single injury was detected on the legs from a single honey bee. We, therefore, excluded the data for thoraxes, mouth parts, and legs from statistical analysis to increase statistical power.3. Results3.1. Bee Population, Number of Injuries in Larvae, Pupae, and Crippled Adult Honey BeesThe population size of honey bee Apis mellifera ranged from 19,712 and 32,472 per hive. The mean number of injuries in fifth instar larvae (13.16 ± 8.61, n = 18) was much lower than that of injuries in prepupae (31.59 ± 21.20, n = 88) (t = −3.616, df = 104, p < 0.001). Injury numbers varied among body parts of the bee pupae (Wald chi-square = 339.740, df = 3, p < 0.001), with the highest number of injuries in abdomens (2.80 ± 3.96, n = 174), followed by antennas (0.46 ± 0.97, n = 174), legs (0.13 ± 0.84, n = 174), thoraxes (0.05 ± 0.45, n = 174), and mouth parts (0.03 ± 0.02, n = 92), respectively. The wings of the bee pupae had no injury. When the injury of the bee pupae was further divided into fresh wounds and scars, the overall mean number of fresh wounds (0.69 ± 2.47, n = 348) was significantly more than that of scars (0.13 ± 2.23, n = 348) (Wald chi-square = 30.395, df = 1, p < 0.001) although their interaction was significant (Wald chi-square = 39.785, df = 3, p < 0.001). Except for fresh wounds (2.89 ± 3.57, n = 87) and scars (2.71 ± 4.34, n = 87) in the abdomens of the pupae, the antennas, the thoraxes, and the legs of the pupae received significantly higher numbers of fresh wounds than that of scars (0.81 ± 3.57 for fresh wounds and 0.26 ± 0.72 for scars, n = 87 for antennas; 0.21 ± 0.61 for fresh wounds and 0.01 ± 0.11 for scars, n = 87 for thoraxes; and 0.48 ± 1.13 for fresh wounds and 0.04 ± 0.18 for scars, n = 87 for legs).We found that 113 (46.50%) out of 243 crippled adult honey bees were injured. A significant difference was observed among the injury number of antennas, abdomens, and wings (Kruskal–Wallis: R2 = 958.66, df = 1, p < 0.001; Figure 1). The mean injury number in the antennas (0.61 ± 0.79, n = 243) was significantly more than that in the abdomens (0.07 ± 0.41, n = 243) and wings (0.07 ± 0.36, n = 243). There was no significant difference in the number of injuries between the abdomens and the wings of the bees.3.2. The Relationship between Infestation Rate and Injury Numbers in Larvae and PupaeA significant relationship between mite infestation rate and injury numbers was observed for both bee larvae and pupae (Table 1). The infestation rate increased as the number of injuries increased for both larvae and pupae (Figure 2). The overall mean infestation rate was 14.26 ± 8.87 (n = 35) for larvae and 1.91 ± 1.44 (n = 35) for pupae. The infestation rate was significantly higher in larvae than in pupae (t = 8.126, df = 68, p < 0.001).3.3. The Relationship between Infestation Rate and Bee Population Per HiveThere was a negative relationship between the population size of honey bee Apis mellifera and the infestation rate of Tropilaelaps mercedesae mites (Table 2). The infestation rate decreased as the number of honey bee individuals per colony increased (Figure 3).4. DiscussionTropilaelaps mites are widespread in honey bee Apis mellifera populations across many parts of Asia [20,27], and their infestations persist even in regions where winter can be harsh and brood production is limited [15,47]. These mites can have a high ability to reproduce in infested colonies [20,23]. In this study, unlike Varroa mites which only begin feeding after the sealed larvae consumed their larval food [48], but similar to a previous study [8], we found T. mercedesae mite infestations in all life stages of honey bees. Of these, the number of injuries to prepupae, also known as mature larvae, was significantly more than that to fifth instar larvae. This result was inconsistent with our hypothesis, where we expected a decreased number of wounds as the larvae progress to the adult stage. We postulated that increased injury numbers in the prepupal stage of honey bees could be related to feeding strategies on nutrients. For instance, the bee prepupae may have been exposed for a longer period for feeding on nutrient sources, in situations where food sources in the capped brood cells are limited. Another possible scenario is that poor hygienic behaviors [49,50,51], a mechanism of mite and disease resistance, in the pupal stage of honey bees might have caused a great number of injuries (Kitiphong et al. [10] and references therein). Consequently, the mites could infest the pupae when workers uncapped the brood cells, thereby affecting the cycle of feeding strategies.Mite infestations in the capped brood cells of honey bee workers might also influence the emergence development and damage adult foraging ability such as crippled wings [21,27,44]. However, consistent with our prediction, we found a higher number of injuries in the bee larvae than in the bee pupae. Patcharin et al. [8] suggested that most injuries in mature larvae were healed when they molt during the white-eye pupal stage. In this study, we included both fresh wounds and scars when counting the number of injuries in the pupae, however, we still obtained a much lower number of wounds in the pupae compared to the number of wounds in the larvae. This result implies indirect support for the idea that bees might have developed mite or disease resistance as they progressed to the pupal stage. Nevertheless, we could confirm the reduction in the number of injuries accompanied by the development of honey bees from the larvae stage to the pupae stage of A. mellifera.An increase in mite infestation rates can be expected where the presence of injuries on honey bees caused by parasitic mites is relatively high because these mites have rapid locomotion and reproductive rates [12,23]. Indeed, a previous study has reported that increased numbers of injuries are positively correlated with increased numbers of actively feeding mites [8]. Consistent with this trend, we found that the number of injuries in the bee larvae and pupae was positively correlated with the infestation rate of T. mercedesae. The more mites present, the higher number of injuries occurred in the bees. Specifically, the infestation rate was pronounced in the larvae, suggesting that changes in injury numbers could influence mite infestation rates at the larval and pupal stages of honey bees. In contrast, the infestation rate decreased as the honey bee population increased, showing the importance of bee colonies in response to mite infestations. However, it is important to remember that the infestation rate at the population level was evaluated during one specific time in this study. The present result might have been different if we based our estimation of brood infestations over a longer time because changes in the infestation rate might be effectively influenced by temporal variation in various mechanisms such as the density and hygienic behaviors of honey bees and the development of other unknown parasitic mechanisms (e.g., DWV) infesting mites/bees in the honey colonies. A previous study has suggested that bees with large colonies, relative to those with small colonies, have a greater density and movement (i.e., walking, resting, nursing, hive maintenance, worker maintenance, in festoon, and foraging) of workers and the ability to increase the temperature in the colonies [41]. Other studies have also reported that the development and population densities of mites are highly sensitive to weather conditions [39,40]. For example, the temperature was found to be crucial for the pest population levels in means of mite growth, survivorship, and fecundity of mites [39], while the hives’ location, humidity, and the beeline’s hygienic characteristics were also important for mite falls [40]. These patterns imply the fact that variations in mite infestation rates among bee colonies with different population sizes might be related to changes in the density and movement of workers, hygienic characteristics of the bees, and the temperature in bee colonies, which might subsequently influence colony defense mechanisms against mite/virus infestations. However, empirical investigations are needed to validate these hypotheses.Although most injuries that occurred in mature larvae might have healed when they progressed to either the pupae stage or the adult stage, the combined effect of pre-capped and post-capped brood injuries might have damaged various body parts of the older pupal and adult stages of honey bees including wings, mouth parts, antenna, thoraxes, abdomens, and legs. In this study, we found a great number of injuries in the abdomens of the pupae. This result supports the previous finding [8] that injury numbers in the abdomens of the bee pupae of A. mellifera were significantly higher than that in other body parts of the honey bees. Injuries caused by Tropilaelaps mites to the pupal stage of the honey bees resulted in permanent injuries at the adult stage of the crippled honey bees, and it was commonly found in the bees’ antennas, wings, and abdomens. Of these, we found that the highest number of injuries was found in the antennas. Antennas are important paired sensory organs of workers and drones of adult honey bees, particularly responsive to stimuli, touch, and odor [52]. Therefore, antennal segment deformation may be affected by the behavioral strategies of honey bees. Tropilaelaps mite infestations showed a negative effect on the olfactory learning, flight ability, and homing ability of A. mellifera [12]. Gao et al. [12] reported that mushroom bodies of Tropilaelaps mite-infested honey bees significantly increased when compared with uninfested honey bees, which may be related to a lower learning ability in the infested honey bees. In general, the mushroom body of insect brains is associated with antennal lobes containing primary olfactory neuropils [53]. As a result, the survival rate of the Tropilaelaps-infested honey bees can be much lower than that of uninfested honey bees [8,27]. In general, bees have two sets of wings that work together for flight performance [54] and make the air move and sound through the antennas to detect pheromones [54,55,56]. The result from our study showed that Tropilaelaps mites cause overt symptoms of wing deformities resulting in emerging honey bees that are unable to fly. Therefore, the colony was found to be less likely to survive as the number of infested honey bees increased. The degree of variations in the subsequent infestation among different body parts of the older and adult stages of honey bees could also be the cause and consequence of DWV infestation on either mites or bees in addition to the potential for variations in the preference of parasitic mites for different nutrient sources provided by different body parts of the bees. Unfortunately, we do not know whether variations in injury numbers among different body parts of A. mellifera were due to DWV or mite itself, or both. Nevertheless, our present findings provide a unique opportunity to investigate variations in the presence of mite feeding on different body parts across different developmental stages of honey bees.5. ConclusionsIn summary, the present study revealed that T. mercedesae mites infest throughout the entire development stage of A. mellifera honey bees. Specifically, the number of injuries was relatively high in the abdomens and the antennas. The presence of injuries and the decrease in infestation rates as the larvae progressed to the adult stage might be attributed to changes in the availability of nutrient contents, hygienic behaviors, and other unknown defense mechanisms against mite/virus infestations in the colony as the bees aged. This is the first study on the relationship either between T. mercedesae mite infestation rates and the injury numbers at different developmental stages or between the infestation rate and population size of honey bee A. mellifera. The advantage of this study is that it provides baseline information about the infestation rates of T. mercedesae mite in the larval and pupal stages of honey bees and the locations of injuries to both the pupal and adult stages of honey bees. It also provided important information about developing the population size of beehives when developing pest management programs for mites. Future studies could focus on how mechanisms that influence variation in the preference of T. mercedesae mites, coupled with other parasitic mites and viruses, among different body sizes of different bee species at different developmental stages and the abilities of the immune systems of each life stage of honey bees in response to such infestations to deepen our understanding of pest management for mites.

animals : an open access journal from mdpi

10.3390/ani11071905

PMC8300402

Dried distillers grains (DDG), a corn by-product of the ethanol industry, are a common feedstuff in cattle diets. Distillers can be used in diets as an effective source of protein and can reduce risk of acidosis by reducing the highly fermentable carbohydrate of starch. However, the price and availability of distillers grains are variable based on droughts, energy prices, and other factors. When inclusion of distillers grains in diets becomes uneconomical or is not possible, diets may decrease in crude protein and increase in starch-altering rumen fermentation parameters and feeding behavior. The effect of distillers grains on cattle growth performance and carcass traits has been studied extensively, but little is known about how distillers inclusion affects feeding behavior. Therefore, the objective of this experiment was to determine the effect of low inclusion levels of DDG on feeding behavior in heifers consuming a high-moisture corn-based diet in the finishing phase. This study demonstrated that low inclusion levels of distillers grains has little effect on growth performance, but can alter feeding behavior and reduce variability in feeding behavior traits.

The objective was to determine the effect of low inclusion levels of dried distillers grains (DDG) on feeding behavior in heifers consuming a high-moisture corn-based diet in the finishing phase. Simmental × Angus heifers (N = 90; 323 ± 50 kg) were fed for 180 d. Heifers were blocked by initial body weight (BW) into two groups, stratified by sire, and assigned to 15 pens with six heifers each. Pens were randomly assigned to one of three dietary treatments: 0% DDG inclusion (0DG), 7% DDG inclusion (7DG), or 14% DDG inclusion (14DG). Treatments did not affect (p > 0.59) BW, average daily gain, and gain:feed. Although there was a treatment × time effect (p = 0.05) for dry matter intake (DMI), with 0DG having greater DMI during the last 70 d, no differences in overall DMI were detected. Treatment affected (p < 0.01) bunk visit duration and head down duration, with 7DG and 14DG having less minutes per day. Bunk visit frequency (p = 0.02) was less variable for heifers fed 14DG and DMI tended (p = 0.08) to be less variable for both distillers treatments. While dietary inclusion of DDG has minimal effects on overall heifer performance, low levels of DDG inclusion can affect feeding behavior and intake variation.

1. IntroductionCorn is the primary feedstock used to produce ethanol in North America through the dry milling process, which results in a coproduct known as dried distillers grains (DDG). During the distilling process, starch is removed from corn through yeast fermentation, resulting in a coproduct that is more concentrated in fat, fiber, and protein levels [1]. Dried distillers grains, a common feedstuff in cattle diets, have been used as an effective protein source, while reducing acidosis by reducing the highly fermentable carbohydrate of starch. However, DDG availability can become an issue as prices are susceptible to droughts, energy prices, and other factors [2]. In the spring of 2020, the availability of DDG was significantly decreased due to the impact of the COVID-19 outbreak. Pandemic travel restrictions and implementation of the Phase One trade deal caused a decline in ethanol production and DDG [3]. Without cost-effective protein sources such as DDG, protein levels in diets can be reduced to less than 12.5%. Decreased DDG also typically translates to greater corn inclusions, which increases dietary starch and acidosis risk. Cattle are also being fed to heavier weights [4] with more days on feed compared with the last time diets with little to no DDG were fed in the 1990s [1]. Starch inclusion levels can affect rumen function, digestibility, and, overall, change feeding behavior [5]. The long-term effects distillers have on feeding behavior and variation in intake are yet to be evaluated. Therefore, the objective of this study was to determine the effect of low inclusion levels of DDG on feeding behavior in heifers consuming a high-moisture corn-based diet in the finishing phase.2. Materials and MethodsExperimental animals were managed according to the guidelines recommended in the Guide for the Care and Use of Agricultural Animal in Agricultural Research and Teaching (Federation of Animal Science Societies, 2010). All experimental procedures were approved by the Institutional Animal Care and Use Committee of the University of Illinois (IACUC #17292). 2.1. Experiment Design, Animals, and ManagementNinety Simmental × Angus heifers were utilized for the current study. The experiment was conducted as a randomized block design. Heifers were blocked by initial body weight (BW) into 2 groups, stratified by sire, and assigned to 15 pens with 6 heifers/pen. Pens were randomly assigned to 1 of 3 dietary treatments (Table 1): (1) 0% dried DDG inclusion (0DG), (2) 7% DDG inclusion (7DG), or (3) 14% DDG inclusion (14DG). Dried distillers grains were replaced with high-moisture corn, soybean meal, and urea (Table 1). Dietary crude protein (CP) ranged from 12.33 to 13.45% as CP decreased slightly, with lesser DDG to reflect typical changes observed in the feedlot industry. Pens (4.88 × 9.76 m) had concrete slats covered with rubber mats. During the adaptation phase, heifers were transitioned using the same diet and fed in concrete bunks for 17 d. Heifers were then fed in GrowSafe bunks (GrowSafe Systems Ltd., Airdrie, AB, Canada) to determine individual feed intake and feeding behavior. Heifers were transitioned to a new crop source of high moisture corn on d 102–104 and were fully transitioned to new high moisture corn on d 105.Heifers were weighed on consecutive days upon start of the trial and before slaughter. Individual weigh dates were recorded on d 57 and 127. A 4% pencil shrink was applied to all weights to account for gut fill. Heifers were implanted 36 d prior to the start of the trial with a Component TE-IH with Tylan implant (80 mg trenbolone acetate, 8 mg estradiol USP, and 29 mg tylosin tartrate; Elanco) and on d 62 with a Component TE-200 with Tylan implant (200 mg trenbolone acetate, 20 mg estradiol USP, and 29 mg tylosin tartrate; Elanco). Heifers were fed Optaflexx 45 (ractopamine hydrochloride, Elanco) at 300 mg per heifer per day the last 28 days of the trial. On d 181, heifers were transported approximately 300 km to a commercial abattoir and were humanely slaughtered under USDA inspection. Hot carcass weight (HCW) was recorded immediately postharvest. Individual camera data were provided by Tyson Foods Inc. for determination of yield (formula derived from USDA, 1997) and quality grades.All heifers were fed in concrete bunks prior to the trial, then switched to GrowSafe bunks at the start of the trial for determination of individual feed intake feeding behavior. Feeding behaviors and intake were monitored continuously in the GrowSafe system. Data were removed on individual days if less than 85% of the consumed feed was assigned to heifers or if the bunk was empty for more than 12 h. Meal criterion and meal analysis for each animal were determined using the Meal Criterion Calculator software version 1.8.7154.27227 (MCC; http://nutritionmodels.tamu.edu (accessed on 30 March 2021), verified 3 August 2019). Meal criterion is defined as the minimum non-feeding event interval between bunk visits before the next bunk visit is considered part of a new meal [6]. Uninterrupted bunk visit events are defined as back-to-back bunk visits for an animal at the same feed bunk [6]. Feed behavior variability was determined by calculating the standard deviation of each feed behavior trait for an individual animal for the entire experiment.2.2. Sampling and Analytical ProceduresFeed ingredient samples were collected biweekly throughout the trial. Equal portions of each ingredient in each period were composited. Composites were dried in a 55 °C oven and ground through a Wiley mill (1-mm screen, Arthur H. Thomas, Philadelphia, PA, USA). Composites were analyzed for dry matter (DM; 24 h at 105 °C), neutral detergent fiber (NDF) and acid detergent fiber (ADF; using Ankom Technology method 5 and 6, respectively; Ankom200 Fiber Analyzer, Ankom Technology, Macedon, NY, USA), crude protein (CP; Leco TruMac, LECO Corporation, St. Joseph, MI, USA), ether extract (EE, Ankom method 2; Ankom Technology), and organic matter (OM; 600 °C for 12 h; Thermolyte muffle oven Model F30420C; Thermo Scientific, Waltham, MA, USA). animals-11-01905-t001_Table 1Table 1Diet composition and nutritional value.Item0DG7DG14DGIngredient, % DM inclusion High moisture corn69.5863.8658.14Corn Silage13.3413.3413.34Soybean Meal2.861.43--DDGS 1--7.1514.30Supplement Ground corn6.9066.9256.940Limestone1.4771.6201.763Urea0.9530.7910.634Trace mineral premix 20.0970.0970.097Rumensin 900.0160.0160.016Fat0.0800.0800.080Supplement MGA 30.0130.0130.013Ground corn4.6774.6774.677Chemical analysis, % DM Dry matter65.2865.9466.70Organic matter92.3492.0991.92Crude protein12.3312.8513.45Neutral detergent fiber11.2012.4613.74Acid detergent fiber5.545.816.09Ether extract3.343.884.42NEm 4, Mcal/kg1.981.981.98NEg 4, Mcal/kg1.461.461.461 Dried distillers grains plus solubles; 2 Supplement contained 2 8.5% Ca, 5% Mg, 7.6% K, 6.7% Cl, 10% S, 0.5% Cu, 2% Fe, 3% Mn, 3% Zn, 278 mg/kg Co, 250 mg/kg I, 150 mg/kg Se, 2205 KIU/kg Vit A, 662.5 KIU/kg Vit D, 22,047.5 IU/kg Vit E. 3 MGA 200, Zoetis, Parsippany, NJ; 4 Calculated net energy for maintenance (NEm) and net energy for gain (NEg) values from National Academies of Sciences, Engineering, and Medicine [7].2.3. Statistical AnalysisThe experimental design was a randomized block with pen as the experimental unit. The MIXED procedure of SAS 9.4 (SAS Inst. Inc., Cary, NC, USA) was used for all statistical analysis. The model included fixed effects of block, diet, time, and the interaction of diet and time. Pen nested in the treatment was included as a random effect [8]. Start BW and the most appropriate expected progeny differences (EPD) estimate were used as a covariate for BW, ADG, G:F, DMI, and all carcass traits to help account for initial and inherent genetic differences. When there was not an appropriate EPD, sire was included as a fixed effect. A repeated measures statement was used for time effect to analyze BW, ADG, G:F, and DMI using an autoregressive (1) covariate structure based on model fit statistics. When analyzing variation in feeding behavior traits, individual animal average for the feed behavior was used as a covariate. Cook’s D was used to identify outliers. When residuals were not normally distributed for several feeding behavior traits, the Box–Cox procedure was used to determine the appropriate transformation. Least squares means have been back-transformed for ease of interpretation. Significance was declared at p ≤ 0.05 and tendencies were discussed at 0.05 < p ≤ 0.10.3. Results3.1. Growth Performance Overall, BW, ADG, and G:F (Table 2) did not have a diet × time or diet effect (p ≥ 0.11). Although, DMI was also not affected by diet (p = 0.29), there was a diet × time effect observed (p < 0.05). On d 56–126, there was a tendency (p = 0.09) for heifers fed 0DG to have a greater DMI when compared with 7DG and 14DG. On d 126–180, heifers fed 0DG also had a greater (p < 0.01) DMI than 7DG and 14DG. 3.2. Feeding BehaviorFeeding behavior traits including bunk visit duration and head down did have a diet effect (p < 0.01; Table 3). Heifers fed 0DG had increased bunk visit durations and head down duration compared with 7DG and 14DG. Heifers fed 14DG had less variation (p = 0.02) in their bunk visit frequency throughout the trial. Both diets including DDG also had a tendency (p = 0.08) to have decreased variation in DMI. Heifers fed 14DG also had a tendency (p = 0.06) to have less variation in meal frequency. However, heifers fed 0DG had a tendency (p = 0.07) to have less variation in meal duration. There was no diet effect (p ≥ 0.72) for meal frequency, meal duration, meal length, meal size, and eating rate. There was also no diet effect (p ≥ 0.13) for variation in behavior traits, including bunk visit duration and head down duration. 3.3. Carcass TraitsCarcass data (Table 4) reveal that heifers fed 7DG had smaller (p = 0.02) longissimus muscle (LM) area. No other diet effects were present (p ≥ 0.23) for hot carcass weight, dressing percentage, 12th rib fat thickness, yield grade, marbling, or kidney, pelvic, or heart fat. 4. DiscussionThe availability and price of coproducts such as DDG are a primary factor driving inclusion in feedlot diets. Determining how to optimally use DDG to alter feeding behavior and feed intake variation has not been studied extensively. In contrast, its effect on growth and performance is well-known. In the current study, no heifer overall performance advantages were observed with the low inclusion levels of DDG. Similar to the current study, ADG and G:F were not different for heifers consuming diets with 0% and 13% DDG [9]. When increasing DDG inclusion up to 40%, a quadratic ADG response was observed with the greatest ADG at a 20% inclusion [10]. However, most distillers titration studies substitute DDG with corn and urea and also use diets with greater crude protein levels than the current study. Corn processing methods may also affect growth results. When comparing the use of high moisture corn and rolled corn in distillers diets, ADG is increased and G:F is improved for cattle fed high-moisture corn [11]. When substituting DDG with high-moisture corn, soybean meal, and urea, inclusions of DDG at 15% or less may not offer any performance advantages.A meta-analysis [1] demonstrated that increasing DDG up to 40% resulted in a quadratic effect for DMI, with intake maximized at 30% DDG inclusion. A direct comparison of 0 and 15% DDG inclusions indicated no difference in overall DMI [12]. However, the 15% DDG diet did increase DMI during the initial 28 d of the finishing phase [12]. In contrast, DMI was increased for heifers fed 0DG in the final 70 d of the current study. Most studies, including those in the meta-analysis [1], only analyze DMI for the whole feeding period rather than in time intervals. Therefore, it is difficult to compare the current findings with the existing literature on how DMI is affected throughout the finishing phase.Despite the importance of coproducts in cattle feeding diets, little is known about how they alter feed intake behaviors. Feeding behaviors are important to understand feed intake regulation, feed efficiency, and health status [6]. Specifically, cattle with greater G:F spend less time at the bunk, have a greater eating rate, and have less variation in feed intake [13], which may play a part in feeding behavior responses to distillers inclusion levels due to decreased starch in the diet. When comparing diets without distillers to those with 16% DDG, distillers inclusion increased meal size and eating rate as cattle decreased the number of meals and time at bunk [14]. Importantly, these steers were only fed for 122 d, which was a shorter period of time compared with the current study. Although there were no differences in meal size or eating rate in the current experiment, there was a decrease in time spent at the bunk with inclusion of distillers similar to the described study. Crude protein in the diets of the study previously described was also much greater with 0% distillers inclusion at 16% CP due to the use of alfalfa mixed haylage as the forage source and the 0DG diet of the current study only having 12.33%. Steers were also used in the study previously described, whereas the current study used heifers. However, there is no current research evaluating feeding behavior differences between steers and heifers consuming DG. Variation in feed intake over time can be managed by several strategies and is an important factor for feedlot growth performance and efficiency [15]. However, variation in feed intake and behaviors is rarely reported. Decreasing DDG inclusion would typically lead to greater dietary starch and risk for acidotic bouts. Thus, it was hypothesized that 7DG and 14DG would have less variation in feed intake and feed behaviors during the finishing phase. Feed intake and variation in feed intake among days can be used to detect ruminal acidosis. Decreased DMI and increased variation in intake are often a sign of subacute acidosis [5]. As hypothesized, there was a tendency for DMI variation to be increased in heifers fed 0DG. Variation in bunk visit frequency duration was also decreased with heifers fed 7DG and 14DG. As with any evaluation of an ingredient such as DDG, the observed effects cannot be solely attributed to DDG inclusion because of a concomitant decrease in dietary starch. Thus, other strategies to reduce dietary starch may also decrease variation in DMI and feeding behaviors.Previous research on carcass data has reported that DDG inclusions of 0–40% may not affect carcass traits. When comparing 20% to 40% distillers grain inclusion levels, HCW, LM area and marbling were not affected by inclusion levels [16]. Similarly, no differences between carcass characteristics in steers fed 0%, 15% or 25% DDG have been found in other studies [10]. Similar results were found in the current study except for LM area, which was decreased in heifers fed 7DG, despite final BW and HCW being similar and genetic differences being accounted for by the EPD for LM area. However, LM area reduction has also been observed and unexplained in other wheat and triticale distillers grains studies [17,18].5. ConclusionsHeifers were fed different inclusion levels of DDG to evaluate the effect of low DDG inclusion on feeding behavior compared to a high starch control diet. There was no treatment effect for growth performance. While heifers fed 0DG had greater DMI on d 126–180 than 7DG and 14DG, there was no treatment effect on overall DMI. Feed intake behaviors were affected by treatment as bunk visit duration and head down duration were decreased for heifers fed 7DG and 14DG compared with 0DG. Overall, the results indicated that inclusion of distillers has minimal impact on overall performance and intake but affect some feeding behavior traits. Inclusion of DDG also decreased variation in bunk visit frequency and tended to decrease variation in DMI. Even at low dietary inclusion rates of 7% and 14% of the diet, DDG can modify and reduce variation in feeding behavior in feedlot heifers consuming a high-moisture corn-based diet.

animals : an open access journal from mdpi

10.3390/ani13111732

PMC10251812

The predation and/or dispersal of Quercus seeds by rodents play an important role in the creation of the tree species. Using the tagging method, we measured the effects of density, storage method, and seed size on rodents’ predation and dispersal behavior. We found that seed survival had a significant negative density-dependent effect; that litter cover and soil burial increased seed in situ time and survival; that in situ feeding rates were significantly higher for small seeds than for large seeds; and that large seeds fed significantly further after dispersal than small seeds. These findings provide insights into the ecological characteristics of Quercus tree regeneration and shed light on the coexistence between rodents and different-sized seeds.

The predation and/or dispersal of Quercus seeds by rodents play an important role in the creation of the tree species. The present study examined the effects of community habitats on the predation and dispersal of Quercus wutaishanica seeds by rodents. We released seeds with densities set at 2, 4, 8, 16, and 32 seed square meter with litter cover, soil burial, and bare ground in the Liupan Mountains National Nature Reserve in the Ningxia Hui Autonomous Region, northwest China. The results showed that (1) the litter cover and soil burial significantly increased the seed survival probability compared with bare ground treatments, especially the predation in situ (PIS) (p < 0.05). Both the scatter hoarding (SH) and larder hoarding (LH) for litter cover and soil burial were significantly increased compared with bare ground (p < 0.05). (2) The large seeds are preferentially predated after dispersal and their long-distance dispersal (>5 m) was significantly greater than that of small seeds (p < 0.05), while small seeds are more likely to be preyed on in situ or during short-distance dispersal (<3 m). (3) The Q. wutaishanica seed predation by rodents increased at a high density rather than at a low density, indicating a negative density-dependent predation. These findings provide insights into the ecological characteristics of Quercus tree regeneration and shed light on the coexistence between rodents and different-sized seeds.

1. IntroductionRodent-mediated seed predation plays an important role in seedling establishment, population dynamics, and species coexistence [1,2]. When the seeds mature and scatter, small rodents hoard higher quantities of seeds in case of food shortages [3]. The behavior of rodents in hoarding food usually takes two forms: (1) larder hoarding, which refers to animals that store all their food in one or a few caches or nests; (2) scatter hoarding, which means that rodents store food in many caches with a small amount of food in each cache [4]. The former could lead to catastrophic consequences once food is stolen, and, on the contrary, the latter may be more beneficial for the natural regeneration of plants [5,6,7]. Even if some of the hoarded food is eaten by predators or thieves, some seeds may be forgotten or scattered, allowing for successful seedling establishment [8,9].The Janzen–Connell theory predicts that seed density may change the predators’ selection preferences [10,11]. High-density seeds may cause large amounts of predation by animals or infection by specific pathogens, thereby reducing the chance of seedling establishment [12,13]. Conversely, the farther the seed is from the parent tree or the lower the seed density, the greater the chance of escape and the more conducive it is to seedling building [13]. Moreover, it is likely that the direction, as well as the intensity, of selection preferences including predation, dispersal, and hoarding will vary with the seed size [14,15,16,17]. Generally, large seeds may disperse farther and are likely to be buried by predators, which is conducive to seedling establishment, while small seeds may be eaten by animals due to their lower nutritional benefits, which are not worth the longer investment, and the distance to be transported is shorter [2,18].Aside from the seeds’ size and density, both the predators’ selection preferences and the fate of seeds are highly susceptible to heterogeneous habitats [3]. There is evidence that the seeds of some trees are easier to find on bare ground than buried [19,20]; thus, the litter cover and soil burial could reduce the risk of seed predation and thereby increase the chance of seedling establishment [20]. Moreover, the positive effect of burial is also reflected: (1) to mitigate the impact of adverse environmental factors (e.g., high temperature and low moisture), or (2) shape a mechanical barrier by changing the chemical environment of the seeds, which contributes to seed germination and seedling establishment [20].Quercus wutaishanica is the dominant species of the warm, temperate, deciduous, and broad-leaved forest in northwest China [3]. The Q. wutaishanica seed varies in seed size at both interspecific and intraspecific levels, which in turn, affects the hoarding and predation preferences of rodents [20]. Previous studies of Q. wutaishanica have shown that seed predation, dispersal, and hoarding by small mammals prior to germination are dependent on seed size [2,20], but density-dependent seed predation has not been well studied, especially when the seeds are in different states. Here, we describe a field experiment focusing on the Q. wutaishanica seeds in three states to explore the influence of seed sizes and their density on rodents’ hoarding and predation behaviors. We expected that (1) rodent-mediated seed dispersal on bare ground may be faster than litter cover and soil burial. (2) Seed size effects on hoarding and predation preferences of rodents. (3) Rodent’ hoarding and predation behaviors dependent on seed density.2. Material and Methods2.1. Study SitesThis study was conducted in the Liupanshan National Natural Reserve (35°15′ N to 35°41′ and 106°09′ to 106°30′ E) in the Ningxia Hui Autonomous Region, northwest China. The area at the edge of the northern agriculture–pastoral ecotone and the semi-humid to semi-arid region in the warm temperate zone. It is hot and rainy in the summer, while dry and cold in the winter. The total annual precipitation is about 767 mm with 60% precipitation from June to September, and the annual evaporation is about 1426 mm [3,20]. The annual mean temperature is 5.8 °C, with the extreme high and low temperatures being about 30 °C (July) and about −26 °C (January) [3,20]. The soil type is mainly grey cinnamon. The main woody vegetation includes trees such as Q. wutaishanica, Tilia paucicostata, Pinus tabulaeformis, Betula platyphylla, etc., and shrubs including Rubus spp., Crataegus kansuensis, Amygdalus davidiana, Cotoneaster multiflorus, etc. The herb mainly comprised Brachypodium sylvaticum, Carex tristachya, Phlomis umbrosa, etc. Shrublands and secondary forests of Q. wutaishanica are the main types of vegetation at the study site [3,20]. Q. wutaishanica seeds matured from August to September. Natural regeneration of this species in this area was often influenced by rodents, and several rodent species, including Niviventer confucianus, Mus musculus, Sciurotamias davidianus, Apodemus agrarius, and Microtus fortis [3,20].2.2. Seed CollectionQ. wutaishanica seeds were collected from Mt. Liupan on 18 September 2017. In order to remove the damaged seeds (worm-eaten and moldy), they were stored in the laboratory for a week, and then seeds were separated based on their sizes (the weight of large and small seeds is (mean ± standard deviation (SD), 3.05 ± 0.38 g) (n = 100) and (1.46 ± 0.27 g) (n = 100), respectively.2.3. Marking and Labeling of SeedsA 7.5 cm long, 0.8 mm diameter copper wire was fixed at a pink plastic label with size of 2 cm × 1.2 cm (containing the mass of the copper wire (0.17 ± 0.002 n = 100) g). Each label was written in pencil with relevant information such as plot location, seed size, seed density, and habitat (bare ground, litter cover, and soil burial), so as to search for and record the fate of the seed [21]. We used an electric drill with a drill bit of 1 mm to make a small hole at the end of the seed away from the germ, and then secured the seed with the prepared label [21,22].2.4. Experimental DesignOn 8 April 2018, in the secondary forest of Q. wutaishanica in Mt. Liupan Nature Reserve, five transects (90 m × 20 m: length × width) were selected. Each transect was considered as density treatments with ratios set as 2 seed Sq. m, 4, 8, 16, and 32 Sq. m, respectively. We set up three plots (2 m × 2 m, each plot was equally spaced ~25 m) as three bare ground treatments from the bottom to the top of the hillside in the midline area of each transect. Furthermore, three litter cover treatments with a size of about 2 m × 2 m are set at five meters from the bare ground as three replicates, and the thickness is about 2 cm. In each experimental transect, we selected three plots buried with 2 cm soil, which had similar sizes to the bare ground and soil burial treatments (each 5 m away from litter cover treatments) (Figure 1). The large and small seeds (1:1) were evenly released at each plot. Total seeds: 5 densities (2, 4, 8, 16, and 32) × plot area (4 Sq. m) × 3 states (bare ground, litter cover, and soil burial) × 3 repetitions = 2232 seeds.2.5. Field InvestigationWe recorded the number of seeds that were preyed on, dispersed, and buried on the 1st, 2nd, 3rd, 5th, 15th, 30th, and 60th days of release plots. We searched and recorded the seed label in transects, and measured the distance from the seed to the falling point using a tape. Here, we recorded the number of seeds retained, the number of in situ preying, the number of preying after dispersal, and the number of stored seeds after dispersal. Dividing the fate of seeds into six types: retained in situ (RIS), predation in situ (PIS), scatter hoarding (SH), larder hoarding (LH), predation after dispersal (PAD), and lose (L).2.6. Data AnalysisWe used Cox regression to analyze the seed survival probability for habitats, densities, and seed sizes during different observation periods. Cox regression was fitted using “Surv” function in the “survival” package. We constructed generalized linear mixed-effects models (GLMMs) for seed retention rate remaining in the experimental transect after the end of the observation period, with fixed factors for the seed sizes, densities, and states and random effect terms for plots and observation period. The random-effects term of plot and observation period were included to allow us to measure variations of plot-to-plot and different observation period, which are not explained by fixed factors. GLMMs were fitted using “glmer” function (family = “beta”) in the “lmer4 package” [23]. To test the impact of the interaction between seed size, density, and state on seed retention rate, we used the “anovo function”. These analyses were performed with R 4.2.0 (R Development Core Team; http://r-project.org; accessed on 2 November 2022).One-way ANOVA with SPSS (Version 21.0) was used to evaluate the effect of the seed states, sizes, and densities on seed fates, including RIS, PIS, SH, LH, PAD, and L. This method was also used to determine the influence of seed states, sizes, and densities on the dispersal distance, and the least significant difference method (LSD) was used to detect differences. The normality test and standardization of data are conducted in SPSS 21.0. We used SigmaPlot (Version 12.5) to make all figures.3. Results3.1. Quercus Wutaishanica Seed DynamicsCox regression analysis showed that rodents prey on seeds the fastest on bare land, and the predation of rodents was significantly inhibited after soil covering (Wald = 12.56, p = 0.045) (Figure 2a). During the observation period, the survival probability of small seeds was significantly higher than that of large seeds (Wald = 7.78, p = 0.032) (Figure 2b). The survival probability was significantly different among different densities (Wald = 24.59, p = 0.005) (Figure 2c), and from high to low, it was 2, 4, 8, 16, and 32 Sq. m, respectively.3.2. Quercus Wutaishanica Seed FatesThere is no significant difference between litter cover and soil burial for the RIS (Figure 3a). Significant differences were observed for the PIS and SH in different treatments (p < 0.05) (Figure 3a). Soil burial and litter cover treatments were significantly higher than bare ground for the LH (p < 0.05). Soil burial and litter cover treatments were significantly higher than bare ground for the PAD (p < 0.05). There is no significant difference in the lost seeds among different treatments (Figure 3a).In terms of the RIS, there were no significant differences between the large and small seeds. The PIS for large seeds was significantly lower than that for small seeds (p < 0.05) (Figure 3b). In contrast, the SH for large seeds was significantly higher than that for small seeds (Figure 3b) (p < 0.05). There were no significant differences in the rates of LH, PAD, and L for large and small seeds (Figure 3b).The RIS and PIS of seeds decreases with increasing seed density (Figure 3c). In contrast, the SH, LH, and PAD of seeds increase with increasing seed density (Figure 3c). Seed loss rates were significantly lower at 2 seeds Sq. m and 32 seeds Sq. m than at the other three densities (p < 0.05) (Figure 3c).3.3. Dispersal DistanceSeeds were carried near the release point (<1 m), and the soil burial treatment was significantly higher than those placed in bare ground and litter cover (p < 0.05) (Figure 4a). Seeds were carried 1–3 m to the release point, and the litter cover was significantly higher than the bare ground and soil burial treatments (p < 0.05) (Figure 4a). Seeds were carried 3–5 m to the release point, and there was no significant difference between the three placement methods (Figure 4a). Seeds were carried 5–10 m and >10 m to the release point, and the bare ground treatment were significantly higher than those from the soil burial and litter cover treatments (p < 0.05) (Figure 4a).Near the seed release point (<1 m), we observed that the retention rate of small seeds was significantly higher than that of large seeds (p < 0.05) (Figure 4b). Within 1–5 m of the seed release point, there was no significant difference in the seed retention rate between different-sized seeds (Figure 4). When the dispersal distance is 5 m away from the release point, the retention rate of large seeds is significantly higher than that of small seeds (p < 0.05) (Figure 4b).The percentage of seeds carried relatively close to the release site (>1 m, 1–3 m, and 3–5 m) increased with the increasing placement density (Figure 4c). Conversely, the percentage of seeds carried relatively far from the release site (5–10 m and >10 m) decreased with increasing placement density (Figure 4c).The fixed effects, including seed sizes, densities, and states, as well as their interactions, explained 32.25% of the seed retention rate, and the densities and states significantly influenced the seed retention rate (p < 0.05). The random effects, including plots and observation time, explained 22.58% and 15.89% of seed retention rate, respectively (Table 1).4. Discussion4.1. Microhabitat Affects the Predation and Dispersal Behavior of RodentsOak seeds rely mainly on internal rodents for dispersal and then for seed germination, seedling establishment, and seedling survival to ensure successful regeneration [24,25]. However, the heterogeneous habitat may affect the distribution of rodents and then affect their foraging activities, ultimately affecting the dispersal and survival of seeds [26]. In this study, we found that litter cover and soil burial significantly increased the seed retention ratio compared with bare ground treatments (Wald = 12.56, p = 0.05) (Figure 2a, Table 1), especially the PIS (p < 0.05) (Figure 3a). This may be the result of the trade-off between predation risk and profit [26]. Given that the cost of capturing prey is high, predators should increase their foraging success by focusing on prey species that are easier to find [27]. However, covering litter and burying soil will increase the time rodents need to search for seeds making them more vulnerable to predation [3]. Conversely, rodents could detect food quickly when the obstacle affecting rodents’ foraging activities is removed, thus reducing the cost of hunting prey to a certain extent [3].It is noteworthy that no matter what seed states, rodents prefer to consume seeds in situ, and the seed percentage of PIS was higher than the RIP, SH, LH, PAD, and lose. The foraging decision of rodents is the result of the comprehensive weighing of many factors [28]. The rodent predation risk is increased if the seed handling time is prolonged [29]. Therefore, rodents choose to prey on the seeds in situ in order to improve the net income per unit time [30]. Furthermore, we found that the retention percentages of litter cover and soil burial treatments were significantly higher than bare ground for the PAD (p < 0.05). This may be another trade-off between the time of searching for seeds and the risk of being preyed by others; in other words, if rodents need to invest a lot of energy in their search for seeds, there is no doubt that predation after dispersal is a safe behavior strategy [30]. Meanwhile, the positive effects of litter and soil are reflected in the hoarding behavior of rodents. For example, compared with bare ground, the SH and LH were significantly increased (p < 0.05) (Figure 3a). Rodents could preferentially store or hoard seeds when the cost is greater than the hoarding, because the former means investing more time and energy, losing the opportunity to prey on other foods, and increasing the risk of being predated by natural enemies [2,31]. In the evolutionary sense, the benefit from rodents hoarding seeds is far greater than the cost of the seeds that they prey on. Because once the seeds escape from predation or are forgotten, they may germinate successfully; even some discarded or partially consumed seeds after being hoarded may potentially establish seedlings [22,32].4.2. The Preference of Rodents for Seeds of Different SizesThe optimal foraging theory predicts that predators prioritize catching large seeds with sufficient nutrients [33]. The detection of Q. wutaishanica seed selection by rodents in our study was consistent with the theoretical expectation (Figure 2b). This supports the previous findings of Zhang et al. and Celis-Diez et al. [2,20]. Rodents can estimate the seeds’ qualities, in order to maximize the energetic intake and/or to minimize the searching time. Seed predators preferentially consume large seeds that contain greater nutrient [24,25]. By extension, rodents prefer to prey on large seeds after dispersal, while small seeds are more likely to be preyed on in situ (Figure 3b), ultimately resulting in different sizes of seeds having the same dispersal fitness [2].In addition, dispersal distance can also reflect rodents’ preferences for different seed sizes [34]. We found that the retention rate of small seeds was significantly higher than that of large seeds near the seed release point (<1 m) (p < 0.05), and most of the dispersal seeds were concentrated within 5 m, while the retention rate of large seeds was significantly higher than that of small seeds when the dispersal distance was 5 m away from the release point (p < 0.05) (Figure 4b). This result is similar with that reported by Xiao et al., who found that most of the seeds of Q. variabilis are transported within 6 m by rodents [21]. Our finding also conforms to the prediction results of the rapid sequestering hypothesis [7]; that is, the seed with lower nutrients is stored near the resource pool, while the seed contained higher energy is stored far away from the resource pool [4].4.3. Negatively Density-DependentIn general, rodents search for food by their acute sense of smell [12]. Stronger olfactory signals from high-density food seed release sites are conducive to successful search by rodents, and once predators find one food cache site, they may be more likely to search for the next one in the adjacent area. [1,35]. We found that the rodents’ foraging is negatively density-dependent (especially, the preference for the Q. wutaishanica seed is higher at a high density than at a low density, while the RIS and PIS of seeds decrease with increasing seed densities. In contrast, the SH, LH, and PAD of seeds increase with increasing seed densities (Figure 2c and Figure 3c and Table 1). Negatively density-dependent foraging was also found in community levels, and these similar studies speculated that this predation was most likely to occur when a certain prey type was more abundant [36,37]. For example, when glabrous plants were abundant, hairy plants incurred less herbivory by Phaedon brassicae [38], the storage of Juglans regia and Castanea mollissima with similar apparent perceived value by scattered storage animals was affected by their abundance [39]. The negatively density-dependent foraging of rodents may be driven by the number of prey individuals [40]. In other words, high-density seeds emit higher concentrations of seed odor and provide more olfactory cues to rodents, whereas low-density seeds would be forgotten or ignored by the rodents. [41,42]. Of course, the satiety effect should not be ignored either, especially at high densities. At that time, seeds are consumed more by rodents, but those that are not attacked due to satiation can survive [43]. Therefore, seed predation mediated by rodent shows a negatively density-dependent result in our study, which could alleviate the loss of rare seeds and promote seeds to escape. As a result, they are more likely to successfully germinate and establish seedlings.In this study, we found that, after taking into consideration the differences among plots and observation time, the random effects could explain 32.25% of the seed retention rate (Table 1), which implied that other factors besides predation mainly drove seed dynamics, even though rodent-mediated seed predation seems very important in determining the short-term seed fates. This is because (1) our study focused on the overall effects across all plots on seed predation and dispersal. However, the microhabitat surrounding the seed included things such as slope position and slope degree, which may affect the predation behavior of rodents [3]. (2) Some rodents prefer to prey on one particular type of prey while others prefer to prey on alternative prey, or size selection may even be related to the size of the rodents [20]. (3) We have not ruled out the influence of secondary dispersal, especially for seeds of different characteristics. Because there is considerable variation in the seed structure, both among and within species as well as within individuals [2]. Although some studies concerning how this variation affects the primary dispersal of seeds have been conducted, little is known about the impact of this variation on secondary dispersal, by any vector; hence, understanding the relationship between the seed structures and secondary dispersal could reveal the adaptations shown by individual plants [1].5. ConclusionsThe results of our study suggest that litter cover and soil burial could prevent rodents from preying on seeds, while this protective effect only delayed the time for rodents to search for seeds to some extent. Once rodents find the release site, the seeds will be quickly preyed on or transported. Of course, scatter hoarding, forgotten after dispersal, or partially consumed seeds, may germinate and establish seedlings. Moreover, rodents prefer to prey on large seeds after dispersal, while small seeds are more likely to be preyed on in situ, which has implications for the potential evolution of different-sized seeds. More importantly, we found that rodent foraging is negatively density-dependent; that is, the preference for the Q. wutaishanica seed is higher at a high density than at a low density. However, rodent-mediated seed predation is a complex ecological process that includes a series of successive stages, and in particular, the secondary dispersal driven by rodents may be more complex due to the complexity of different study areas and the diversity of predators; therefore, future studies should especially focus on the relationship between secondary dispersal and density dependence.

animals : an open access journal from mdpi

10.3390/ani11102814

PMC8532937

The current growing social awareness of animal welfare has led to the development of welfare indicators, which are effective tools for assessing each of the integrated aspects of this multidisciplinary issue. Hence, skin diseases have been suggested as potential general health indicators for use in cetaceans. Particularly cetacean poxvirus causes distinguishable hyperpigmented “ring” or “tattoo” lesions that affect cetaceans both in the wild and in managed facilities. However, most studies have analyzed these characteristic lesions through visual appraisal, while only a few have implemented diagnostic methods to corroborate the presence of the virus. To this end, skin biopsies are usually the sampling method selected, although they are considered to be an intrusive procedure. In this study, we analyzed sloughed skin sampled with cytology cell samplers (CCSs) in 12 tattoo-like lesions from two free-ranging cetaceans stranded in the Canary Islands. We employed two different DNA extraction methods and compared the effectiveness of the device with that of biopsies. All the lesions resulted positive for cetacean poxvirus, obtaining reliable data from the use of this device. Thus, CCS is considered to be a promising non-invasive tool for further assessing skin diseases in cetaceans, particularly those under human care, without affecting their welfare.

Poxvirus-like lesions are widely used as a potential health indicator in cetaceans, although for this application, corroboration of Poxvirus skin disease is imperative. Aiming to address skin biopsies intrusiveness, a preliminary investigation of a non-invasive skin sampling procedure to molecularly detect CePV-1 in 12 tattoo-like-lesions from two free-ranging stranded cetaceans in the Canary Islands was performed. Skin lesions were brushed with cytology cell samplers (CCSs) and placed into 1.5 mL microcentrifuge tubes with 1 mL of RNAlaterTM Stabilization Solution. For factual comparisons, DNA extractions from sloughed skin obtained with CCS and biopsies from the same lesions were accomplished with DNA Tissue Kit STM (QuickGene, Kurabo, Japan). Moreover, a second DNA extraction from sloughed skin with DNeasyTM Blood and Tissue Kit (Qiagen, Inc., Valencia, CA, USA) was performed to ascertain kit suitability for CCS. Molecular detection of CePV-1 was performed through a real-time PCR. As a result, a 91.7% and 83.3% rates of positivity were obtained with biopsies and CCS through Quickgene, respectively, compared to the rate of 100% using CCS with Qiagen. Accordingly, CCS is a reliable non-invasive sampling device to obtain sufficient genetic material to be analyzed for CePV-1 in tattoo-skin-lesions as well as for other purposes in cetaceans under human care.

1. IntroductionCetaceans’ skin is considered a multidimensional feature that can provide a wealth of information, forming the basis of research in a wide number of studies covering a broad range of scientific branches [1,2,3,4,5,6,7,8]. Hence, this tissue has been used for long-term health assessments, enabling us to gain a closer look at the health status of marine mammals and aquatic ecosystems [9,10,11,12]. For instance, skin diseases have been suggested to be triggered by exposing both free-ranging and human-managed cetaceans to continuous aberrant conditions, resulting in compromised immune system function and a consequent increase in susceptibility to disease [13,14,15,16].Cetacean poxvirus (CePV) is one of the most widely reported skin diseases [17,18,19,20,21]. Recently, it has been classified into two groups: cetacean poxvirus 1 (CePV-1), which affects both free-ranging and human-managed odontocetes, and cetacean poxvirus 2 (CePV-2), which infects mysticetes [22,23]. In cetaceans, this cutaneous disease displays characteristic lesions which are recognized as hyperpigmented “ring” or “tattoo” lesions, with the latter being referred to as tattoo skin disease (TSD) [20]. Regarding the unanimous consensus that clinical signs of disease can be indicative of compromised health, these distinguishable skin manifestations have been considered as a potential general health indicator in cetaceans [13,24,25]. Despite being broadly described, this viral skin disorder is still unassigned within the Chordopoxvirinae subfamily due to the limited genomic information on it. One of the main reasons for this is that CePV has mainly been identified through visual appraisal [25,26,27,28], with few studies having used diagnostic assays to correctly detect and therefore determine the presence of this pathogen in poxvirus-like lesions in cetaceans [22,24,29,30,31].The detection of CePV in tattoo-like lesions in cetaceans under human care is necessary in order for these lesions to be applied as an “animal-based” health indicator [32,33,34,35]. As most research fields in which skin is the focus of study, skin biopsies are the method of choice to either molecularly or histologically diagnose skin diseases in cetaceans in managed facilities [36,37,38,39,40,41]. Nevertheless, the increasing awareness of welfare in cetaceans has prompted attempts to develop dynamic methodologies for safe handling and sampling with the aim of minimizing the risk of compromising their well-being. As with wild individuals, in managed facilities some researchers have highlighted the fast turnover time of cetaceans’ skin [42,43,44,45,46], proposing the collection of sloughed skin of animals’ bodies as an advantageous non-invasive method that could potentially be an alternative to biopsies [5,47,48]. Notwithstanding the aforementioned points, research on the use of these emerging non-intrusive skin sampling techniques in cetaceans under human care is still scarce, with their effectiveness having been poorly explored in research studies.Correspondingly, as managed facilities are actively committed to the advancement of scientific research while maintaining ethical responsibility, efforts to create innovative sampling methodologies and improve the standards of practice during these procedures should be encouraged [49,50,51]. Thus, stranded cetaceans could plausibly be used in model studies, providing an opportunity to perform preliminary investigations [52,53,54]. Additionally, their use would enable protocol adjustments to resolve possible misgivings and achieve feasible results that could be reproduced in cetaceans under human care.Accordingly, the aim of the present study is to validate a potential non-invasive skin sampling device using sloughed skin to molecularly detect cetacean poxvirus (CePV) in tattoo-like lesions by comparing its sensitivity and effectiveness with that of skin biopsies obtained from stranded cetaceans in the Canary Islands.2. Materials and MethodsOn 21 February 2021, a juvenile male common bottlenose dolphin (Tursiops truncatus) (Case 1), 240 cm in length, was found stranded and dead at Abades, Arico, Tenerife, Canary Islands, Spain (28°09′00″ N, 16°25′00″ W). On 17 April 2021, a juvenile female Atlantic spotted dolphin (Stenella frontalis) (Case 2), 150 cm in length, was found stranded and dead at Playa San Juan, Guía de Isora, Tenerife, Canary Islands, Spain (28°10′47″ N, 16°48′45″ W). Based on their anatomic parameters, both animals showed a moderate body condition [55] and the carcasses were in a good state of preservation (code 2/5) [56,57,58,59]. Due to their exceptional states of preservation, neither the refrigeration nor freezing of either of the animals were required prior to their necropsies. Thus, standardized necropsies [60] were performed on each dolphin the day after they were found. Throughout the external examination during necropsies, several skin lesions affecting the rostral and lateral areas of both animals were observed. As a result, two lesions from Case 1 and ten lesions from Case 2 were described, photographed, and measured before their later collection. Each skin lesion from both animals was split to retain a portion at −80 °C, while the remainder was first sampled with a sterile cytology cell sampler (CCS) (Deltalab, Barcelona, Spain) and later correctly identified and preserved at the same temperature as the other portion. The skin sampling procedure using these CCSs consisted of gently brushing the surface of the lesions to obtain sloughed epidermis, which adhered into the bristles of the brush. Then, all the CCSs were placed into 1.5 mL sterile RNAse- and DNAse-free microcentrifuge tubes with a safe lock (Thermofisher Scientific, Madrid, Spain), in which 1 mL of RNAlaterTM Stabilization Solution (Thermofisher Scientific, Madrid, Spain) had previously been added. Subsequently, the bristles of the CCS stayed embedded in the RNAlaterTM, while the plastic stems were cut to the level of the microcentrifuge tubes’ tops with a pair of scissors to allow the closure of the vials, using the safe lock to avoid unexpected openings (Figure 1). Due to the genomic stabilization capacity of the RNAlaterTM solution, microcentrifuge tubes were stored at room temperature until their subsequent molecular analysis, which was performed within 1 working week [61].After accomplishing the procedure previously explained, the rest of both necropsies were performed by sampling and collecting representative tissues of all the major organs and lesions for subsequent analyses in order to proximate the most plausible cause of death/stranding, as routinely performed [58,59]. Hence, all samples were stored in a 10% neutral buffered formalin fixative solution for histologic and immunohistochemical analysis, whilst few of them were preserved at −80 °C until processing for biomolecular studies.Approximately 0.5 g of each fresh-frozen skin samples from both animals was mechanically macerated in lysis buffer and subsequently centrifuged, later progressing to simultaneous DNA/RNA extraction using the DNA Tissue Kit STM (QuickGene, Kurabo, Japan). Considering that an initial sample of ≤0.5 g is required to correctly perform genomic extraction with this method, some modifications in the manufacturer protocol were necessary in order to accurately extract the DNA/RNA from the fresh skin samples collected with CCS. They were first agitated using a vortex for 15 s at maximum speed to ensure the detachment and mixture of a great part of the epidermal crust adhered among the bristles into the 1 mL RNAlaterTM solution. After this, the tips of the CCS were removed, preserving the acquired RNAlater–sloughed skin mixture in the vials. With the aim of obtaining an approximate amount of 5000 μL of macerates from each sample, some adaptations in the proportions of the components were made. Therefore, instead of adding 4500 μL of 0.1% diethylpyrocarbonate (DEPC)-treated water and 500 μL of 1× lysis buffer as accomplished with biopsy samples, 3600 μL and 400 μL from each component, respectively, were applied apart from the 1000 μL RNAlater–sloughed skin mixture. Finally, all macerates were centrifuged (2500 rpm for 15′ at 4 °C) and supernatants were collected to continue with their genomic extraction. DNA/RNA extraction was achieved from each macerated sample (N = 24) in a QuickGene Mini 80 nucleic acid isolation machine (QuickGene, Kurabo, Japan) according to the manufacturer’s instructions with some modifications: an RNA carrier (Applied BiosystemsTM, Thermo Fisher Scientific, Waltham, MA, USA) was added during the lysis step, as previously indicated [62].The molecular detection of CePV-1 was performed using a 1-step real-time polymerase chain (q-PCR) method to amplify a conserved region (150 bp) of the DNA polymerase gene by using the degenerate primer sets designed by Sacristán et al. [63] (Odontopox-F: 5′-CARGAAATMAAAAAGAARTTTCCATC-3′, and Odontopox-R: 5′-ACGTTCTGTTAARAAYCGTCTTAGTA-3′). The thermocycler profile was set for initial denaturation at 95 °C for 5 min, followed by 40 amplification cycles, each compromised of a denaturation step at 95 °C for 15 s, an annealing step at 60 °C for 30 s, and an elongation step at 72 °C for 30 s. The final cycle was composed of an extended elongation, which was performed at 72 °C for 7 min [29]. A melting curve step was added at the end of the reaction. The thermal cycler employed was a MiniOpticonTM Real-Time PCR System (Bio-Rad Laboratories, Irvine, CA, USA). Adequate non-template negative controls (nuclease-free water) for both extraction and amplification as well as extraction-positive and amplification-positive controls previously confirmed by our group were included.The PCR products from positive lesions were purified using a commercial kit (Real Clean Spin kit 50 Test-REAL), and then sequenced using Sanger DNA sequencing (Secugen S.L., Madrid, Spain). The amplicon identities were confirmed with BLAST (www.ncbi.nlm.nih.gov/blast/Blast.cgi/ (accessed on 4 June 2021)).In order to compare the effectiveness of the DNA extraction from the skin samples collected with CCS using the QuickGen kit method, a second extraction using DNeasyTM Blood and Tissue Kit (Qiagen, Inc., Valencia, CA, USA) was undertaken. In this instance, it was necessary to unfreeze each of the halves which had previously been scraped with CCS from the 12 skin samples collected from Cases 1 and 2. Hence, the same skin sampling procedure using CCS explained above was performed repeatedly. The CePV-1 positive control was a 0.025 g biopsy that had previously been confirmed by our group using the real-time PCR method [63] and sequencing amplicons (unpublished sequencing results) previously described. Once 1 mL RNAlater–epidermis mixtures were obtained, they were subsequently subjected to a high centrifugation speed for 5 min (14,000 rpm). On some occasions, this step had to be repeated as many times as necessary to obtain enough pellet. Consequently, approximately 0.025 g of skin pellet precipitation was collected from each sample by removing practically all the supernatant. To verify that the maximum sample weight specified by the manufacturer’s instructions for correctly performing the DNA extraction had not been exceeded, all the vials were weighted. Thus, precise weights of samples were acquired by subtracting the weight of an empty vial (≈1.078 g). Afterwards, all samples were ready for the DNA extraction to proceed following the manufacturer’s instructions. The importance of incubating the biopsy CePV-1 positive control, which had previously been cut into small pieces, for at least 30 min with continuous 15 s high-speed vortexing every 5 min during incubation for complete lysing must be noted. Upon the completion of the extraction, DNA products were tested with the same real-time protocol as that mentioned above (see Appendix A for more details). 3. Results3.1. Macroscopic Findings3.1.1. Case 1In total, Case 1 (Figure 2) showed two lesions that could be attributable to CePV. Of both lesions, the most remarkable was a 5 × 3.5 cm serpiginous and stippled light grey tattoo-like lesion located on the ventral right corner of the oral cavity (Figure 2A). The other lesion (Figure 2B) showed an oval and depressed shape, which was observed on the melon of the common bottlenose dolphin.3.1.2. Case 2Just like Case 1, Case 2 (Figure 3) presented compatible CePV lesions. In a multifocally manner, tattoo-like lesions with different evolution stages were randomly distributed and affected many areas of the skin. Three of them (Figure 3A–C) were the characteristic persisting ring lesions, delimited with black edges and showing a black and stippled pattern at the center. One of these lesions (Figure 3C) showed blistering across half of its center. Another two lesions (Figure 3D,E) were observed on the tip and melon of the dolphin, respectively. They were lighter in color and featured a barely visible black margin, corresponding to the lesions in the healing process. On one of the flanks of the spotted dolphin, a ring lesion that was black in color with pale edges was observed (Figure 3F). Close to it, a lesion that appeared very similar to this last one, apart from its pale, irregular, and raised center, was observed (Figure 3H). The lesions observed at the ventral part of the animal (Figure 3G,I) were irregular, light grey, and blurred, being hardly perceptible and without delimiting margins. On the peduncle, there was a remarkably large lesion affecting almost all the entire length (Figure 3J). This lesion was irregular in shape and black, and featured a pale grey pin-hole pattern along its center.3.2. Molecular FindingsThe results of the molecular findings are compiled in Table 1. Of the 12 cutaneous lesions sampled using biopsies taken from both individuals, which were previously submitted for Quickgene DNA/RNA extraction, 11 were positive for CePV-1. More specifically, from Case 1, both lesions were positive; meanwhile, in Case 2, of the 10 lesions tested, nine were positive. Only one lesion (Figure 3J) presented an abnormal amplification curve with a RT-PCR cycle threshold value (Ct) of 11.66 without melting temperature. For the purpose of confirming this lesion as negative to CePV-1 and to prove that the PCR product was neither too concentrated nor overloaded with inhibitors leading to incorrect PCR interpretations, it was diluted into 10-fold serial dilutions up to 10−3. In this way, we sought to stablish a better sensitivity and quantification dynamic range. However, the real-time PCR detected neither of the dilutions of the PCR product from this lesion.Regarding the samples collected with sterile CCS from these 12 lesions, different results were obtained depending on which genomic extraction method was used. Hence, with Quickgene, 10 lesions from both cases were found to be CePV-1 positive: both lesions from Case 1 (Figure 2A,B) and eight lesions from Case 2 (Figure 3A–F,I,J). Comparing these results with those obtained with tissue sampling, it can be observed that, in both sampling methods, the same lesions from Case 1 were found to be positive for cetacean poxvirus. Nevertheless, the same outcome was not observed in Case 2, in which a lesion that was found to be negative when sampled with a biopsy (Figure 3J) was found to be positive when sampling using CCS. However, using this latter sampling method, two other lesions which were found to be positive for CePV-1 when collected using a biopsy (Figure 3G,H) were not detected. Conversely, through Qiagen, all lesions from Case 2 were found to be positive for poxvirus, in addition to the other two lesions from Case 1. Therefore, making a general comparation from these results with the other obtained by employing a different genomic extraction method, we observed that using the same sampling procedure (CCS), the two lesions that were not detected (Figure 3G,H) were both found to be positive with Qiagen. Moreover, with this last genomic extraction kit, the negative tissue sample from Case 2 was also found to be positive when CCS was employed. Thus, from a broad-based assessment, we could observe that with Quickgene, a slightly better sensitivity was acquired when samples were collected with biopsy rather than with CCS. Yet, an improvement on these results was gained when applying both the CCS sampling method and the Qiagen extraction kit.Comparing the Ct values from positive lesions between both different sample collections and genomic extraction methods a remarkable range of the Ct values was observed, with 15.17 and 37.10 being the minimum and the maximum values, respectively. Considering the theoretical correlation in which it was established that low Ct values correspond to a high viral loads and vice versa, it is observed that the lesion 3A from Case 2 presented the maximum viral load in both sampling techniques and extraction methods. In addition, the same sigmoidal correlation was observed between the lesions that had the lowest viral load extracted with Quickgene. Thus, lesion 3H, which was sampled by biopsy, had a Ct value of 35.37, with it being undetected when using the cytology cell sampler. However, these results did not concur with the ones obtained with Qiagen, with lesion 3G being the one which presented the least viral load, with a Ct value of 37.10.The sequence similarity searching from the DNA polymerase sequences of CePV-1 obtained from all the positive lesions of both cases in this study was performed with BLAST. We could identify that, in both cases, the sequences revealed high percentage homologies of 100%, 99%, and 98% with the already uploaded nucleotides sequences under the GenBank accession numbers of MF458199, KU726612, and MH005249, respectively.4. Discussion4.1. Sampling Methods and DNA Extraction ProtocolsIn the present study, the molecular detection of cetacean poxvirus from two free-ranging cetaceans stranded in the Canary Islands was achieved, with different results being obtaining depending on the sampling method used and the genomic extraction kit employed.Aiming to reproduce and extrapolate the respective sampling procedures used for cetaceans under human care, skin samples were attained in fresh conditions. The use of sterile CCS enabled us to obtain an acceptable quantity of sloughed skin from all poxvirus-like lesions for posterior genomic extractions. Macroscopically, the load of sloughed skin that adhered to the bristles of CCS was determined by the size of the lesions, with it being possible to gain more epidermal crust from larger samples than was possible from smaller ones. This resulted in it being easier to obtain epidermal crusts from lesions with a larger volume due to it being easier to rub their surfaces than ones with a smaller size, with it being necessary in the latter to scrape the outer layer more times to gather sufficient desquamating epidermal debris. Accordingly, the time required to finally acquire sloughed skin was found to be approximately 1 min in all attempts. Nevertheless, as was recently reported by Bechshoft et al. [5], enough skin cannot always be obtained from scraping alongside the flanks of bottlenose dolphins in managed care when using a rubberized scraper. This may be due to the high metabolic and mitotic activity which affected skin undergoes, leading to a continuous removal of epidermis, in contrast to healthy skin [17,18]. In either case, the variation in the quantity of sloughed skin obtained between each sample did not lead to further complications for the DNA extraction, as each of the protocols was standardized.Since biopsy is the current method of choice for collecting skin samples, we decided to compare the genomic yield gained from this sampling method with that obtained through CCS by using a DNA extraction kit that was specifically suitable for use with tissue samples [64]. Therefore, through Quickgene, we attempted to contrast the reliability and effectiveness of both sampling procedures in order to gain enough genetic material from the skin lesions.Furthermore, Qiagen was used for a second DNA extraction from sloughed skin obtained from the same poxvirus-like lesions. According to the manufacturer’s protocol, Qiagen is suitable for purifying DNA from very small amounts of starting material, ensuring high-quality yields from nonstandard samples and considering 0.025 g as the maximum weight [65]. Therefore, the purpose of this second genomic extraction was to corroborate the point mentioned above, comparing the genome extraction from the sloughed skin that was collected via CCS through both kits. Thus, this study not only attempted to show which of the two sampling methods obtained more sensitive results, but also attempted to ascertain which of the genomic extraction protocols is more appropriate for use with the proposed non-invasive sampling method. In contrast to the first extraction, which was carried out through Quickgene, the undeliberate unfreezing of skin lesions from both animals had to be conducted to repeat the skin sampling procedure with CCS. This feature is recognized to have an undesirable impact on the quality of DNA preservation, apart from not serving as a standard operating procedure if it is intended to be extrapolated in cetaceans kept in managed facilities. Despite this, the DNA extraction was carried out while considering the latter facts regarding the further interpretation of the molecular results. In both the CSS sampling procedures, the sloughed skin was embedded in RNAlaterTM Stabilization Solution. The purpose of the use of this reagent is that it can serve as a transport medium in situations in which samples cannot be immediately processed or frozen, as often happens in managed facilities, where samples are normally sent to external laboratories.4.2. q-PCR Molecular Results from CePV-1 Positive LesionsThe visual diagnosis of TSD was confirmed in all the samples tested in the present study. Nevertheless, the results differed when using the different tissue sampling methods and DNA extraction kits.Focusing on the juvenile bottlenose dolphin (Case 1), both lesions were found to be positive when employing both sampling and genomic DNA extraction protocols. Lesion 2A presented a typical serpiginous irregular pattern and was delimited with black borders. In the literature, these lesions are considered to represent the acute phase of infection [24]. On the other hand, the other lesion, 2B, would have been hard to detect if it were not for its depressed and oval-shaped appearance. This lesion is considered to be in an advanced stage of the infection [18]. Through Quickgene, it can be observed that both the biopsy and CCS sampling methods were effective for both lesions. However, it is evident that lesion 2A showed a higher viral load than lesion 2B, with the biopsy sample showing a better Ct value (23.65) than the sloughed sample (25.46). Nevertheless, both values indicate a considerable viral load when in terms of poxvirus infections. Thus, there is a correlation between the macroscopic findings, since 2A was considered to be in an initial stage and due to the viral load. Regarding the other positive lesion, 2B, it was also successfully extracted using both sampling methods. However, in this case it was the sloughed sample that presented a better Ct value (27.93) compared to the biopsy (31.80). In addition, these molecular results are also correlated with the advanced stage of the lesion. The Qiagen extractions of sloughed skin collected from both lesions gained good DNA genomic yields, to such an extent that each lesion presented even better Ct values than those found using the Quickgene extraction, with Ct values of 20.85 and 26.70, respectively.In the Atlantic spotted dolphin, different evolution stages of 10 tattoo-like lesions were observed, coinciding with a wide range of Ct values. Macroscopically, the first three lesions (3A–C) were typical rounded lesions with a stippled pattern in the center, representing the early stage of the infection [24]. Lesions 3A and 3C were larger in size than 3B and also presented considerably more dark pinpoints at the center. Moreover, lesion 3C presented slightly raised margins and half of its center was blistered; both features could be used to identify acute phases of the infection [18]. Regarding the Quickgene biopsy genome extractions, the Ct values obtained from these three tattoo-like lesions were 15.65, 18.08, and 16.42, respectively, with these lesions having higher viral loads than all the others tested in the present study. The same pattern can be observed in the molecular results from the CCS, with Ct values of 17.25, 20.80, and 19.04. In this case, the biopsy samples gathered better genomic yields than the sloughed samples when using same extraction kit. Concerning the DNA extraction yield with Qiagen from the sloughed skin, very similar results are obtained. Regarding these first three lesions, in this case the biopsy samples gained better Ct values from 3A and 3B, with only the result for the third lesion, 3C, being improved with the use of Qiagen.Within the other seven lesions from specimen 2, the variations in molecular values between the sampling methods through Quickgene were significant. At first glance, the lack of CePV amplification on the three lesions can be noticed. Regarding 3J, we were not able to amplify poxvirus DNA using the biopsy sample. Conversely, the same lesion was amplified when using the CCS sampling technique. Our first impression was that the PCR product from the biopsy sample was overloaded, leading to amplification faults. However, once they were diluted into serial dilutions, none of them were found to be positive for CePV. These results might suggest that an inappropriate genomic DNA extraction procedure was used for this lesion, or that incorrect sampling was achieved due to selecting an area from the lesion without viral content. Whatever the case, the Ct value from the sampling of this lesion with CCS was 33.20, indicating a low viral load, a feature which might have also influenced the result obtained with the biopsy. The other two lesions from which DNA was extracted that did not have amplified poxvirus sequences were 3G and 3H. In this case, both lesions were sampled with CCS. This might have been due to their low viral loads (the samples presented values of 31.79 and 35.37 from biopsy sampling, respectively), meaning that they were undetected when collected from sloughed skin. On the other hand, and corroborating the above point, lesion 3G was barely visible macroscopically and featured a dark-grey area, appearing to be an almost healed tattoo-like lesion [33]. However, contrary to what might be expected, lesion 3H, which was compared to 3G in a prior evolution stage, presented a lower viral load than the other lesion. When analyzing the molecular results obtained through Qiagen for these three lesions, it is possible to reach more reasonable conclusions. The DNA extracted from sloughed skin from lesions 3G, 3H, and 3J using this kit were found to be positive for CePV-1. The Ct values obtained were 37.10, 35.66, and 31.88, respectively. In this case, an expected clear correlation between the low expression level of Ct values and macroscopical findings can be observed. Interestingly, the Ct results for lesion 3J were more favorable when using this genomic extraction protocol employing CCS as a sampling method. Accordingly, as has been reported, Ct values of above 35 in q-PCR are not considerable and should not be interpreted as marginally positive. However, the melting curves obtained from these PCR products were identical for all positive samples, and negative controls did not produce any product. Due to this, they were sequenced to confirm their specificity, leading to poxvirus DNA polymerase sequences being obtained.The rest of the lesions (Figure 3D–F,I) were all in different stages of regression, with black margins being less evident or disappearing and the lesions becoming lighter in color [20]. The four of them presented low viral loads when using both genome DNA extractions, with slightly better Ct values being obtaining with the biopsy compared to with CCS through Quickgene. The other genomic extraction protocol obtained barely weakened molecular results compared with both sampling methods.4.3. Validating Cytology Cell Samplers as a Reliable Non-Invasive Method to Sample Skin LesionsIn attempting to determine the effectiveness of the CCS, different percentages of positive results were obtained when comparing the sampling and genomic extraction methods. In this manner, considering that all 12 lesions were determined to be CePV-1 positive through Quickgene, the effectiveness of detecting this virus in skin lesions was 91.7% and 83.3% when using biopsy and CCS, respectively. From this, it can be deduced that sampling with skin biopsies has an 8.4% accuracy. Comparing the Ct values of both sampling methods when using Quickgene as a genome extraction kit, it is evident that the lesions sampled with CCS require more amplification cycles in order to cross the positivity threshold. This is reflected in the negative Ct values of the lesions sampled with CCS (Figure 3G,H). Accordingly, as mentioned before poxvirus lesions in healing stages with low viral loads might lead to CCS losing a certain amount of sensitivity. However, these results could be improved by the use of Qiagen with CCS, which had a 100% success rate. Considering physical status as an important aspect of an animal’s overall wellbeing, detecting cetacean poxvirus in tattoo-like lesions is important in order to correctly corroborate an animal’s condition. Hence, such characteristic lesions could generally serve as an indicator of disease progression, thus correlating them with the health state of the animal. In cetaceans under human care, the presence of confirmed poxvirus lesions could potentially be used as a visual health parameter, especially when combined with the advantage of applying CCS as a non-invasive skin sampling procedure which is unlikely to negatively interfere with the welfare of animals.Having access to two specimens with positive poxvirus lesions was very significant in our quest to validate CCS as an effective viral skin sampling method. The exceptional states of preservation of the animals was crucial in developing the present skin sampling protocol in order to be used for cetaceans under human care. In addition, despite the limitations of the sample size, the fact that all 12 lesions were found to be positive for cetacean poxvirus through CCS was outstanding and reaffirms the need to prove their efficacy in cetaceans under managed care.In summary, this pilot study on stranded animals has served as an opportunity to validate the use of sterile CCS for the diagnosis of poxvirus skin disease. The skin sampling procedure making use of sterile CCS can be considered to be a promising method for the detection of cetacean poxvirus, with accurate results for animals in managed care. Furthermore, we can additionally consider their implementation for sampling sufficient genetic material for other multiple areas of study, not limiting their applicability in poxvirus skin disease. Additionally, this leads to the idea that there should be a rigorous discussion as to whether biopsies are truly the best sampling method for detecting pathogenic microorganisms such as viruses in cetaceans under human care. This needs to be balanced against the potential stress and risk caused to the individual by the handling and sampling processes. However, further investigation is needed to address the uncertainties involved and ensure the potential of the use of this non-invasive method in cetaceans in managed facilities.5. ConclusionsIn the present study, we demonstrated the reliability of the use of CCS for the detection of cetacean poxvirus, comparing the results with those of biopsy samples. These findings will be highly significant for validating the further use of this device as a non-invasive method for assessing viral skin lesions in cetaceans under human care and carrying out visual health assessments.

animals : an open access journal from mdpi

10.3390/ani13061021

PMC10044630

Alpha-2 agonists are commonly used sedatives in horses. They are known for their inhibitive effects on gastrointestinal motility, which limits their use in horses with colic. Dexmedetomidine belongs to the α2 agonist drug class, and studies in human patients have reported that it may enhance gastrointestinal function instead of inhibiting it. Therefore, the aim of this study was to investigate the effect of dexmedetomidine on intestinal smooth muscle function in horses. To evaluate this in varying degrees of intestinal damage, tissue samples were taken from 12 horses prior to and during the disruption of small intestinal blood flow (pre-ischaemia and ischaemia), as well as following the reinstatement of blood supply (reperfusion). We found that the circular smooth muscle (CSM) contractility was not affected by ischaemia, whereas the longitudinal smooth muscle (LSM) showed an increase in both spontaneous and nerve mediated contractile activity. The addition of dexmedetomidine caused a decrease in the spontaneous contractile activity of CSM, but an increase in that of LSM. During ischaemia, dexmedetomidine also mildly increased the nerve mediated contractile activity. These results may indicate a stimulatory effect of dexmedetomidine on small intestinal contractility. However, the influence of dexmedetomidine administration on intestinal motility in vivo needs to be further investigated.

α2 agonists are frequently used in horses with colic, even though they have been shown to inhibit gastrointestinal motility. The aim of this study was to determine the effect of dexmedetomidine on small intestinal in vitro contractility during different phases of ischaemia. Experimental segmental jejunal ischaemia was induced in 12 horses under general anaesthesia, and intestinal samples were taken pre-ischaemia and following ischaemia and reperfusion. Spontaneous and electrically evoked contractile activity of the circular and longitudinal smooth muscles were determined in each sample with and without the addition of dexmedetomidine. During a second experiment, tetrodotoxin was added to determine if the effect was neurogenic. We found that the circular smooth muscle (CSM) contractility was not affected by ischaemia, whereas the longitudinal smooth muscle (LSM) showed an increase in both spontaneous and induced contractile activity. The addition of dexmedetomidine caused a decrease in the spontaneous contractile activity of CSM, but an increase in that of LSM, which was not mediated by the enteric nervous system. During ischaemia, dexmedetomidine also mildly increased the electrically induced contractile activity in LSM. These results may indicate a stimulatory effect of dexmedetomidine on small intestinal contractility. However, the influence of dexmedetomidine administration on intestinal motility in vivo needs to be further investigated.

1. IntroductionMotility disorders of the small intestine are associated with high mortality rates in horses [1,2,3]. The most common manifestation is a post-operative ileus following colic surgery for strangulating intestinal lesions, with an initial neurogenic and subsequent inflammatory phase inhibiting intestinal motility [4,5]. Many factors can contribute to reduced gastrointestinal (GI) motility, such as intestinal manipulation or anastomosis and the use of opioids or sedatives [6,7,8]. Even though α2-adrenergic agonists can inhibit intestinal motility, the use of these drugs for sedation or analgesia during the perioperative period of colic surgery may be inevitable in some cases. Looking more closely at the evidence available for the inhibition of intestinal motility by α2 agonists, experimental trials in healthy horses have shown that xylazine, detomidine, and romifidine decrease the myoelectrical and/or mechanical activity of the small intestine in an experimental setting [9,10,11,12,13,14]. Other reported GI effects of this drug class are decreased borborygmia following detomidine administration [15] and dose-dependently reduced intestinal smooth muscle contractions caused by xylazine and detomidine addition in vitro [16]. A peripheral α2-adrenoceptor antagonist Vatinoxan was shown to prevent (me)detomidine mediated decrease in borborygmia, indicating that this hypomotility is most likely mediated through activation of the peripheral α2-adrenoceptor [15,17].Medetomidine and its active D-isomer dexmedetomidine are more selective α2 agonists that are not as commonly used in horses, yet are known produce effective and safe sedation [18,19,20]. These drugs have also been shown to affect intestinal function, with medetomidine decreasing intestinal circular smooth muscle contractility in vitro [16], and decreased borborygmia being documented the first hour after dexmedetomidine administration in horses and donkeys [18,21].Remarkably, clinical studies in humans have found that dexmedetomidine may enhance intestinal motility peri-operatively or during the continuous sedation of patients in intensive care. These studies reported a variety of improved clinical GI parameters such as increased intestinal sounds and decreased passage time with dexmedetomidine administration compared with treatment with saline as placebo, lidocaine, morphine, and propofol [22,23,24,25,26,27]. Different mechanisms of action have been proposed for this phenomenon, such as the attenuation of inflammation and ischaemic injury or vagal stimulation [28,29,30]; however, there is no direct evidence available to validate these theories. Experimental trials investigating this phenomenon have yielded conflicting results. Studies performed in laboratory animals have reported both inhibitive and stimulating GI effects of dexmedetomidine [29,31,32,33,34], and one study in healthy human individuals found decreased GI passage [35].With uncertainty surrounding the topic of dexmedetomidine and its effect on intestinal motility, this brings up the question of how dexmedetomidine influences GI motility in horses. If dexmedetomidine could elicit a comparable stimulation of GI function, as seen in humans, this could offer an alternative to the commonly used GI inhibiting α2 agonists. Therefore, the objective of this study was to investigate the effect of dexmedetomidine on the in vitro contractility of equine jejunal smooth muscle during pre-ischaemia, ischaemia, and reperfusion. The second objective was to compare the response of circular smooth muscle (CSM) and longitudinal smooth muscle (LSM), as these play a different role in intestinal motility and may show different responses to the addition of pharmacological substances, as shown in other species [36]. Further objectives were to investigate if the effect is mediated by the enteric nervous system (ENS), and to assess how ischaemic injury affects intestinal contractility. We hypothesized that dexmedetomidine addition would negatively affect contractility, and that this would be more pronounced in CSM. Furthermore, we hypothesized that this would be independent of ENS, and that ischaemic injury of the intestine would further decrease contractility.2. Materials and Methods2.1. AnimalsTwelve horses were subjected to experimental segmental jejunal ischaemia under general anaesthesia. The intestinal tissue samples taken during the different ischaemia phases were used for this controlled in vitro experimental trial. Seven mares, two geldings, and three stallions were selected for this terminal study because of severe musculoskeletal issues. The horses were warmbloods with a mean age of 15.9 ± 7.4 years and weight of 547 ± 53 kg. Physical examination, blood cell count, and faecal egg count were performed to assess the general health of the horses prior to the experiment. At least 2 weeks prior to surgery, the horses were stabled at the facilities of the equine clinic of the University of Veterinary Medicine Hannover, with free access to hay and water. On the day of the trial, food but not water was withheld for 6 h prior to anaesthesia.2.2. Surgical ProcedureFollowing premedication with 5 µg/kg dexmedetomidine (Dexdomitor, Orion Corporation, Espoo, Finland), general anaesthesia was induced (0.05 mg/kg diazepam, Diazedor, WDT eG, Garbsen, Germany; 2.2 mg/kg ketamine, Narketan, Vétoquinol GmbH, Ismaning, Germany). Isoflurane (Isofluran CP, CP-Pharma GmbH, Burgdorf, Germany) in 100% oxygen combined with a continuous rate infusion of dexmedetomidine (5 µg/kg/h) was used to maintain general anaesthesia. Direct blood pressure measurements were performed in the facial artery, and the mean arterial pressure was maintained between 60 and 80 mmHg by administrating lactated Ringer’s solution (RingerLaktat EcobagClick, B. Braun Melsungen AG, Melsungen, Germany) and dobutamine (Dobutamin-ratiopharm 250 mg, Ratiopharm GmbH, Ulm, Germany) to effect. A pre-umbilical median laparotomy was performed in dorsal recumbency. After checking the intestines for pre-existing abnormalities, a jejunal segment located 7 m oral to the ileocecal fold was taken as control (pre-ischaemia sample). Forty-five minutes after induction, segmental ischaemia was induced by occlusion of the mesenteric arteries and veins with umbilical tape in 2 m of aboral jejunum that was located 1 m oral to the ileocaecal fold. In six horses, no flow ischaemia was induced by complete tightening of the ligatures. In the other six horses, low flow ischaemia was implemented with 80% reduction in intestinal blood flow under monitoring with Laser Doppler fluxmetry (O2C, LEA Medizintechnik GmbH, Giessen, Germany). Ischaemia was maintained for 120 min, and prior to removal of the ligatures, another tissue sample was taken from the ischaemic segment (ischaemia sample). Subsequently, reperfusion was initiated, followed by resection of the reperfusion sample after 120 min of reperfusion. The horses were euthanized with 90 mg/kg pentobarbital administered intravenously (Release 50 mg/mL, WDT eG, Garbsen, Germany) after the final sample was taken.2.3. Organ Bath2.3.1. In Vitro Experiment—Day 1The jejunal samples were placed in a modified Krebs−Henseleit buffer (117.0 mmol/l NaCl, 4.7 mmol/l KCl, 2.5 mmol/l CaCl2, 1.2 mmol/l MgCl2, 1.2 mmol/l NaH2PO4, 11.0 mmol/l Glucose, and 25 mmol/l NaHCO3) at 4 °C and transferred to the Institute for Physiology and Cell Biology. For each time point (pre-ischaemia, ischaemia, and reperfusion) from all horses (n = 12), eight smooth muscle tissue strips including CSM and LSM of equal size (1 × 0.5 cm) and weight were prepared from each jejunal sample. Consequently, there were two tissue strips used as the control samples and two tissue strips used to test the effect of dexmedetomidine for each type of smooth muscle tested (Figure 1). In the analyses following the experiment, the values of these duplicates were averaged.The tissue strips were placed into an organ bath filled with 12 mL of modified Krebs−Henseleit buffer (117.0 mmol/l NaCl, 4.7 mmol/l KCl, 2.5 mmol/l CaCl2, 1.2 mmol/l MgCl2, 1.2 mmol/l NaH2PO4, 11.0 mmol/l Glucose, and 20 mmol/l NaHCO3) with constant aeration with 95% O2 and 5% CO2, and mounted on an isometric force transducer (Hottinger Baldwin Messtechnik, Darmstadt, Germany). The initial tension of the muscle strips was adjusted to 2 g, which corresponds to 20 mN. Isometric contractile forces of smooth muscle tissues were continuously measured using a chart recorder (4.8 kHz/direct current; Spider 8 chart recorder, Hottinger Baldwin Messtechnik) until the end of the experiment at 120 min, and data were collected with data acquisition software (catmanEasy software, version 1.01, Hottinger Baldwin Mess-technik). The viability and response of the enteric nervous system and smooth muscle cells was tested with electrical field stimulation (EFS) with pulses of 0.5 ms during 10 s with 10 Hz and 30 V at 65, 95, and 110 min (Figure 1). At 80 min (t80), 1 μM dexmedetomidine hydrochloride (Dexdomitor, Orion Corporation) was added to the organ bath chambers. Spontaneous contractile activity was determined based on baseline tension (g), frequency (peak/min), amplitude (mN), and mean active force (MAF; mN) at t75–80 and t105–110 (Figure 1). These time points were selected for the analysis, after visual inspection of all of the results, because this time frame included representative spontaneous intestinal motility before and following dexmedetomidine addition. In the dexmedetomidine-treated LSM samples, these values were also determined at t~85, because the visual inspection of the original traces indicated an intense response in the LSM directly following the addition of dexmededetomidine. Induced contractile activity as a response to EFS was expressed in amplitude (mN), which was quantified at t65 and t110.2.3.2. In Vitro Experiment—Day 2The full thickness intestinal tissue of the pre-ischaemia samples was stored in the refrigerator at 4 °C overnight, and on day 2 of the experiment, the tissue strips of smooth muscle were freshly prepared. The aim of this experiment was to test the response to dexmedetomidine with and without tetrodotoxin (TTX) to determine if the changes elicited by dexmedetomidine were nerve dependent. This part of the experiment was only performed with the pre-ischaemia samples of eight horses. The vitality of the tissues was tested by EFS. Then, 1 μM dexmedetomidine was added to all of the samples, with and without 1 μM TTX. The time schedule was comparable to that of the experiment on day 1, with the addition of TTX at t70 and the addition of dexmedetomidine at t80. Successful inhibition was verified by a lack of response to EFS.2.4. Data AnalysisA power analysis was performed before commencing the study (G*Power 3.1.9.1s). To detect a difference of 10 mN in amplitude between the treatment groups with a standard deviation of 5 mN, based on a power of 0.8 and alpha of 0.05 through the use of a two-tailed unpaired two sample t-test, a total sample size of 12 horses was required.Statistical analysis and graph design were performed using commercially available software (Excel 2016, Microsoft, Redmond, WA, USA; Graphpad Prism 9.4.1, Graphpad Software Inc., San Diego, CA, USA). Data were presented as mean (±standard deviation). Data were tested for normal distribution using a Shapiro−Wilk test. The values for basal contractile activity and response to EFS were compared between the groups subjected to low flow and no flow ischaemia using a two-tailed unpaired two sample t-test. As there were no significant differences between the horses subjected to the different low-flow and no-flow ischaemia types, the results of both groups were combined for further statistical analysis. Statistical significance was set at p < 0.05.Differences between the different ischaemia phases at t75–80 were compared using a repeated measures ANOVA with Greenhouse–Geisser correction of the p-values, followed by a post hoc Tukey test for multiple pairwise comparisons. To evaluate the development over time in the organ bath with or without dexmedetomidine, the basal contractility was compared between t75–80 and t105–110 and the response to EFS between t65 and t110 using a two-tailed paired t-test. A comparison between the samples treated with and without dexmedetomidine was performed for each time point using a two-tailed unpaired two sample t-test.3. Results3.1. Ischaemia ModelThere were no significant differences between the groups subjected to low-flow and no-flow ischaemia. Therefore, the results of the two ischaemia groups were pooled for further data presentation and analysis.3.2. Basal Contractile Activity3.2.1. Ischaemia PhasesComparing the different ischaemia phases at t75–80, ischaemia and reperfusion did not affect the contractility of the CSM in any of the tested variables. In contrast, the ischaemic and reperfused LSM showed a significantly higher frequency, amplitude, and MAF compared with the pre-ischaemic values (Figure 2).3.2.2. Development over Time without the Addition of DexmedComparing the different time points t75–80 and t105–110 of the organ bath experiment, CSM showed an elevated amplitude in the ischaemia samples and an elevated MAF in the reperfusion samples at t105–110 (Figure 3). LSM showed a significant increase in amplitude over time in all of the samples. LSM frequency was elevated in the ischaemia samples only, whereas MAF increased over time in the ischaemia and reperfusion samples.3.2.3. Effect of Dexmedetomidine AdditionEvaluating the initial response to dexmedetomidine addition at t~85, all of the tested variables of the basal contractile activity were significantly increased in the LSM of all ischaemia phases, whereas no significant initial response of CSM was observed.Comparing the basal contractility at t105–110 between the samples with and without dexmedetomidine addition, the pre-ischaemia and ischaemia CSM exhibited a significantly lower frequency in the dexmedetomidine-treated samples (Figure 4). Furthermore, the ischaemia CSM showed a significantly lower MAF with dexmedetomidine. Dexmedetomidine did not affect the frequency of LSM. In contrast, LSM amplitude and MAF were significantly higher in the dexmedetomidine-treated samples of all of the ischaemia phases (Figure 4).3.3. Electrical Field Stimulation3.3.1. Ischaemia PhasesThe amplitude of EFS that induced the contractile activity of CSM was comparable during the different ischaemia phases. In LSM, the ischaemia and reperfusion samples exhibited higher amplitudes than the pre-ischaemia samples (Figure 5).3.3.2. Development over Time without the Addition of DexmedCSM and LSM showed a significant change in response to EFS comparing t110 with t65, with an increased amplitude in all of the ischaemia phases (Figure 6).3.3.3. Effect of Dexmedetomidine AdditionThere were no significant differences in response to EFS of CSM for any of the ischaemia phases. In LSM, the amplitude was significantly higher in the dexmedetomidine-treated ischaemia samples (p = 0.02) compared with the untreated ischaemia samples (Figure 7). During the other ischaemia phases, this did not reach statistical significance (p = 0.059 and p = 0.80 for pre-ischaemia and reperfusion, respectively).3.4. Day 2 of the In Vitro Experiment—Dexmedetomidine with and without TTXAll of the tissue samples used in these experiments showed vitality, as they exhibited spontaneous motility and they responded to EFS. Comparing the basal contractile activity at t105–110 of the pre-ischaemia samples that were treated with dexmedetomidine with and without the addition of TTX, there were no significant differences between these groups (Figure 8).4. DiscussionThe main finding of the study was that the in vitro addition of dexmedetomidine caused a decrease in the spontaneous contractile activity of CSM, but an increase in that of LSM. The in vitro addition of dexmedetomidine also stimulated the nerve-mediated contractile activity of LSM, yet to a lesser extent. Therefore, we partially rejected the first hypothesis that dexmedetomidine addition would negatively affect contractility. The changes in basal contractile activity were not affected by the addition of TTX, indicating that this was not mediated by ENS, confirming the second hypothesis. The contractility of CSM was not affected by the different phases of ischaemia. In contrast, LSM showed an increased basal and EFS induced contractile activity following ischaemia and reperfusion, leading to the rejection of the third hypothesis.The stimulating effect of ischaemia on the spontaneous contractile activity of LSM found in the current study is in contrast with studies performed in various species reporting that ischaemia reduced the intestinal contractility of the affected and neighbouring intestinal segments [29,37,38,39,40]. Discrepancies between some of these studies and the current investigation could be attributed to differences between species, in which the SM layer was investigated, or for the variance in ischaemia duration (15 min vs. 2 h). The increase in LSM contractile activity may be caused by a stress response with sympathetic activation leading to initial hypermotility. Circulating catecholamines could affect SM directly, considering that norepinephrine has been shown to elicit excitatory effects in the human colon when binding at α1-adrenoceptor D subtypes [41]. Furthermore, the stress-related activation of intestinal corticotropin-releasing factor (CRF) receptors could mediate the stimulation of motility [42]. Regarding the clinical occurrence of a postoperative ileus in the horse, there is only sparse evidence for the pathophysiologic mechanism in this species [5]. In general, rodent models have shown that sympathetic stimulation can lead to decreased motility in this disease entity [43].Dexmedetomidine increased the spontaneous and nerve-mediated contractile activity of LSM. This response differed from the effect on the circular layer, where it decreased CSM contractile activity. This corresponds with previous investigations in horses and rats that have shown a varying response to α2-adrenoceptor agonists in the different muscle layers, with inhibited or unchanged CSM contractility, as well as unchanged or enhanced LSM contractility [36,44,45]. These differences may be caused by asymmetric innervation of CSM and LSM with possible variation in α-adrenoreceptor distribution, and may facilitate a reciprocal contractile activity of the two muscle layers [46,47].In the current study, TTX addition did not change the effect of dexmedetomidine, hence it can be concluded that the observed effect on the spontaneous contractile activity was not mediated through ENS. In line with our results, other studies have also shown that spontaneous contractile activity in equine jejunum and ileum was independent of ENS [16,48]. The effect of dexmedetomidine in this context could be elicited through direct stimulation of receptors on the smooth muscle itself [49]. In the rat ileum, α2 adrenoceptors were shown to be negligible in direct muscular effects of adrenergic agonists. However, this may not be the case for other species, considering the major interspecies differences in α2-adrenoceptors’ subtypes and distribution [45,50]. There are no data available on the distribution of α2-adrenoceptor subtypes in horses, complicating direct comparison and the extrapolation of the results obtained in other species. Apart from direct stimulation of the smooth muscle, other possible explanations for the observed effect on the spontaneous contractile activity may include the activation of adrenoreceptors on the platelet-derived growth factor receptor α+ cells (PDGFRα+ cells), which are electrically coupled with smooth muscle cells in the intestine [41]. Alternatively, it could influence the pacemaker activity of the interstitial cells of Cajal (ICC), which can be independent of ENS [51]. However, the presence of α adrenoceptors on intestinal ICC has not been clarified completely [41,52], and there are no studies available investigating this in horses. Another possible mechanism of action of dexmedetomidine could be an effect on the enteric glia Ca2+ response [31], which alters the membrane potential of the smooth muscle cell.The increased nerve mediated contractile activity may be caused by dexmedetomidine enhanced stimulation of α receptors located on cholinergic neurons of ENS [45,53]. Another possible mechanism of action may be the activation of nonadrenergic imidazoline receptors by dexmedetomidine [54], considering that presynaptic imidazoline receptors have been indicated as a pathway to attenuate ENS response by modulating intestinal cholinergic neurotransmission [54]. However, this could not be confirmed in guinea pig ileum, revealing that imidazoline-like compounds only modulated cholinergic neurotransmission through the interaction with presynaptic α2 adrenoceptors instead of imidazoline receptors [53]. Therefore, this theory is less plausible, yet again, with the limitation of the extrapolation of experimental data between different species.Reviewing the previous studies investigating dexmedetomidine and other α2 agonists on in vitro contractility, both enhanced and inhibited contractility have been reported. Xylazine, detomidine, and medetomidine reduced spontaneous and/or evoked contractility in horse jejunum, guinea pig duodenum, and mice ileum [16,55,56]. On the other hand, two studies in the rat ileum found that dexmedetomidine increased the amplitude of spontaneous contractions without affecting the frequency [34], and that dexmedetomidine reduced the ischaemia induced inhibition of contractility [29]. Differences between experiment duration, species, and the small intestinal segments that were used, complicate the exact comparison of the studies. Furthermore, different tissue preparations were used, with full thickness intestinal samples as well as combined CSM and LSM layers, and not all of the studies investigated both the CSM and LSM contractile activity.This study found increased in vitro contractility of the LSM with dexmedetomidine treatment, which conflicts with two in vivo studies that found decreased gut sounds following dexmedetomidine administration in horses [18,21]. This illustrates the fact that in vitro contractile activity does not necessarily translate to an increased motility in vivo. Regarding our results, this difference may be due to CSM playing a more dominant role than LSM in equine small intestinal motility, considering the relative thickness of CSM compared with LSM. Apart from the local effects of dexmedetomidine on the intestinal smooth muscle, its influence on other parts of the intestine or organs may also be of relevance for its ultimate effect on GI motility. The anti-inflammatory properties of dexmedetomidine may reduce the leukocytic inhibition on motility [28,30], and its reduction in ischaemia reperfusion-related injury of the mucosa [57,58] could indirectly ameliorate both the neurogenic and inflammatory inhibition of motility. Moreover, dexmedetomidine can alter the intestinal perfusion, with different authors reporting both improvement and deterioration of intestinal circulation [59,60,61]. Considering that pain is a significant motility inhibiting factor, the analgesic effects of α2 agonists may also be of significance in this matter [13,62]. Finally, there are many differences between the in vitro situation compared to in vivo that influence the extrapolation of the results, such as the migrating myoelectrical complex, luminal filling, intraluminal pressure, and an interaction between the intact muscle layers and the myenteric plexus [46,51,63,64].A limitation of the study is the use of systemic dexmedetomidine during anaesthesia as premedication and as continuous rate infusion during anaesthesia. This was selected because it is considered unethical to induce general anaesthesia without proper premedication. The continuous rate infusion during the maintenance of anaesthesia was elected as part of the analgesic protocol and as isoflurane sparing anaesthetic management. Consequently, dexmedetomidine was elected to rule out the influence of any other anaesthetics. Any effect would be present in both groups and most likely be temporary, considering that rapid wash-out has been described in organ baths [65]. The measurement of dexmedetomidine levels in the intestinal tissue prior to the organ bath experiment or at the end of the in vitro experiment could have given more information on the relevance of this in vivo dexmedetomidine CRI. The administration of dexmedetomidine prior to ischaemia could have ameliorated the intestinal injury by pharmacological preconditioning, which could have influenced the effect of the ischaemia model [57]. On the other hand, it is unlikely that this affected the comparison of the treatment groups in the contractility measurements, considering this affect would be present in both groups. Furthermore, we confirmed the presence of significant intestinal injury during histological examination of these samples, as published elsewhere [66].Comparing the in vitro applied concentration of dexmedetomidine with clinical doses, the dexmedetomidine plasma concentrations reached after commonly applied doses of 3.5–5 µ/kg are around 5 ng/mL [18,19], corresponding to approximately 25 nM. This is lower than the concentrations used in the current study, and thus represents another limitation of this work. For this in vitro study, a concentration was elected within the lower range of those previously described in in vitro studies in other species. Even though the investigated drug concentrations are not directly transferable to the clinical situation, the current observations serve as a first step in elucidating the mechanism of dexmedetomidine induced changes on the small intestinal contractility in horses. Further limitations of this study include the small sample size, the individual variation between the horses, and the relatively short reperfusion time.5. ConclusionsThe contractile activity of CSM was not affected by ischaemia, yet the LSM contractility increased following ischaemia and reperfusion. Dexmedetomidine inhibited the spontaneous contractile activity of CSM, whereas it stimulated that of LSM. This was not mediated by the enteric neurons, possibly indicating a direct effect on the smooth muscle cells. Dexmedetomidine also mildly increased the electrically induced contractile activity in LSM, which most likely indicates a differential distributionof the α-receptors between the two muscle layers. How the inhibitive effect on CSM and the excitatory effect on the LSM translate to motility in the clinical patient remains unclear, yet CSM may play a more dominant role in peristalsis due to its thickness in relation to that of LSM. Many other factors such as pain, inappetence, or the primary disease may also influence the final effect of dexmedetomidine on the case outcome. To fully comprehend the impact of dexmedetomidine on motility, investigations in perfused ex vivo models or in vivo techniques may be performed in future studies to clarify the many remaining questions. If dexmedetomidine could elicit a comparable stimulation of GI function as seen in man, this could offer an alternative to the commonly used GI inhibiting α2 agonists.

animals : an open access journal from mdpi

10.3390/ani12010107

PMC8749518

With this study we present our therapeutic strategy for cats with purulent fluid accumulation in the thorax. In addition to the systemic administration of antibiotics, the aim of the therapy is always the drainage of the purulent fluid from the thorax. For this purpose, we use a particular small-bore chest drain. The first aim of our study is to assess the efficacy and complication rate of our drainage. The second objective is to evaluate two treatment groups regarding their disease outcomes. We were able to show that our small-bore chest drain is similarly effective to the traditionally used large-bore drains. At the same time, we had a very low drain-associated complication rate. We detected no difference between the treatment groups and, thus, no effect on survival by early placement of bilateral drains into the thoracic cavity or lavage of the thoracic cavity with a heparinised solution. Our study supports the theory that drainage of purulent fluid from the thoracic cavity in cats can be performed with small-bore drains with good results and minimal risk of complications.

First-line therapy for cats with pyothorax consists of intravenous antibiotics, drainage of the septic pleural effusion and closed-chest lavage. Large-bore thoracostomy tubes are traditionally used for drainage, but case series indicate a comparable efficacy using small-bore tubes. In this retrospective study, we describe a new technique of sheath-guided small-bore (6 F) thoracostomy tubes in cats with pyothorax and evaluate their efficacy and complications. Additionally, we compare outcomes between two treatment groups. Placement and use of the small-bore thoracostomy tubes described here has a low complication rate of 4% (3/67 tubes), and 53% (24/45) of the cats could be treated with thoracostomy tubes and closed-chest lavage according to the protocol. The success rate is reduced by 18% (8/45) due to deaths caused mainly by sepsis, 16% (7/45) due to structural diseases requiring surgery and a further 14% (6/43) due to lavage failures that could only be cured after additive therapy (thoracotomy or fibrinolysis). The long-term prognosis was very good, with a survival rate one year after discharge of 94% (30/32). We detected no effect on survival by early placement of bilateral thoracostomy tubes or closed-chest lavage with a heparinised solution. In conclusion, therapy of pyothorax with small-bore thoracostomy tubes is as successful as therapy with large- or medium-bore tubes.

1. IntroductionPyothorax is a life-threatening disease. The overall prognosis for cats with pyothorax is fair to good. Cats survive to discharge in 46% to 78% of cases [1,2,3,4,5]. The mortality rate is highest during the first 48 h of hospitalisation [1,2,6].Treatment modalities include medical management and surgery. A combination of drainage of the septic pleural effusion via thoracostomy tubes and intravenous (IV) administration of a broad-spectrum antibiotic is currently the first-line treatment. Surgical management is recommended in 4% to 6% of cases because of intrathoracic abscesses [1,2,5] detected by diagnostic imaging and in 5% to 9% of cases because medical management fails [1,2,3,4].For medical management, single or multiple thoracentesis or placement of a thoracostomy tube with and without intermittent closed-chest lavage are described [1,2,3,4]. Traditionally, large-bore thoracostomy tubes (14–16 F) are placed with a rigid trocar under anaesthesia [4,5,7]. Complications associated with large-bore thoracostomy tubes in cats with pyothorax include anaesthetic complications, pneumothorax, lung laceration and subcutaneous fluid leakage [4,5].In addition to the classical large-bore tubes, the use of a 10 F trocar-based thoracostomy tube placed under general anaesthesia is published [3]. Alternatively, the use of a small-bore (6 F) wire-guided thoracostomy tube is reported in 2 case series of cats with pyothorax [8,9]. The most common complication with these small-bore tubes is a failure to drain, as a result of kinking [8].The aim of this retrospective study is to describe a new technique of sheath-guided small-bore (6 F) thoracostomy tubes in a large series of cats with pyothorax and to evaluate the efficacy and complications of this technique. As a second objective, we compare the outcome between 2 groups with different management strategies.2. Materials and MethodsCats with spontaneous pyothorax treated via small-bore thoracostomy tubes between February 2006 and September 2018 were included. Primary surgical care (foreign bodies on initial diagnostic imaging or oesophageal perforation) led to exclusion.Two groups were compared based on different management strategies.Cats in the Standard Group (February 2006 to July 2010) initially had a unilateral thoracostomy tube placed and waited for clinical stabilisation over the next few hours to decide if a second thoracostomy tube was needed. A second tube was placed on the contralateral side of the thorax if the cats showed no improvement in their dyspnoea, despite oxygen supplementation and unilateral drainage, after 3–12 h. Cats in the Standard Group received closed-chest lavage with a balanced electrolyte solution. Cats in the Intensive Group (August 2010 to September 2018) had bilateral thoracostomy tubes placed as early as possible to completely drain the septic pleural effusion. A second tube was placed if sonography showed residual pleural effusion on the contralateral side of the thorax after the placement of the first tube, allowing for the safe placement of a second tube. Cats in the Intensive Group received closed-chest lavage with a heparinised balanced electrolyte solution. Medical records were reviewed for information regarding signalment, medical history, physical examination findings, diagnostic work-up, treatment, complications and outcomes. Follow-up information was obtained by re-examination or by contacting the owner or referring veterinarian by email or telephone. The diagnosis of pyothorax was based on a purulent to pyogranulomatous pleural exudate with evidence of bacteria in cytology and/or culture. The distribution of the thoracic effusion was assessed by dorsoventral radiography or sonography of the thoracic cavity. Clinical and haematological findings were evaluated with regard to systemic inflammatory response syndrome (SIRS) criteria (Table 1) at presentation; the diagnosis of SIRS was made if ≥3/4 of the markers were positive [10].Stabilisation included a combination of IV balanced electrolyte solution, supportive oxygen and general supportive care (e.g., analgesia and antiemetics), as needed. All cats received IV broad-spectrum antibiotics. Antibiotic selection was empirical and determined by the clinician in charge of the case.A 6 F sheath-guided thoracostomy tube (Drainage 6 F S00124-7, Walter Veterinaer-Instrumente e. K., Baruth/Mark, Germany; Figure 1a) was inserted in all cats. The site and location of maximum fluid accumulation was determined by ultrasound. Sedation with an opioid (0.2–0.4 mg/kg butorphanol IV/subcutaneous (SC) or 0.005–0.01 mg/kg buprenorphine IV) was induced in all cats. The thoracic wall was clipped and aseptically prepared. Local anaesthesia (0.5 mL lidocaine 2% SC) was administered to all cats. If necessary, additional sedatives were administered at the time of thoracentesis (0.125–0.25 mg/kg midazolam IV and 2.5–5.0 mg/kg ketamine IV). The thorax was punctured with a 20-gauge venous catheter. A thoracentesis was performed via this catheter. Then, a 0.018-inch guidewire was inserted through the catheter. After a stab incision, a 6 F valved vascular sheath with a dilator (Prelude Radial Sheath Introducer 6F PSI-6F-7-018, Merit Medical Systems, Inc., South Jordan, UT, USA; Figure 1b) was inserted into the thorax. The dilator and guidewire were removed and the thoracostomy tube was inserted through the sheath. After removing the sheath, an airtight Luer Lock adapter was attached. The patency of the tube was checked. The tube was fixed with a purse-string suture, followed by a finger trap and a simple interrupted suture about 5 cm apart. Then, the pleural cavity was drained as much as possible via the thoracostomy tube.The decision for a second tube on the opposite side of the thorax was determined differently in the 2 groups (see group definition above).A chest dressing and pet cone were applied to protect the tubes from contamination and manipulation by the cat.After at least 12 h, closed-chest lavage with a warmed (approx. 37 °C) balanced electrolyte solution (10 to 20 mL/kg per cycle and tube) was performed two to four times daily. Pure balanced electrolyte solution (Sterofundin ISO, B. Braun Melsungen AG, Melsungen, Germany) was used in the Standard Group, whereas 10 IU/mL of unfractionated heparin (UFH; Heparin-Natrium Braun 25.000 I.E./5 mL, B. Braun Melsungen AG, Melsungen, Germany) in balanced electrolyte solution was used in the Intensive Group.In some of the cats in both groups, ampicillin was additionally instilled intrathoracically after lavage (500 mg/chest side q 12 h; Ampicillin-ratiopharm 1.0 g, ratiopharm GmbH, Ulm, Germany).Further interventions were implemented on a case-by-case basis, such as thoracotomy for severe lung pathology documented on diagnostic imaging or the use of local fibrinolytics (alteplase 0.25 mg/chest side q 24 h; Actilyse 10 mg, Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach an der Riss, Germany or streptokinase 10,000 IU/chest side q 24 h; Streptase 250 000 I.E., CSL Behring GmbH, Marburg, Germany) if lavage was ineffective due to an overly viscous effusion or a lack of improvement.The thoracostomy tube was electively removed when fluid production had decreased to less than 5 mL/kg/day and there was clinical, radiographic or cytological improvement. The cats were discharged with oral antibiotics for at least 1 month.Failure of closed-chest lavage was defined as the need for additional therapy by intrathoracic fibrinolysis or thoracotomy.Tube-related complications (misplacement, blockage or accidental removal) were additionally analysed.Survival in the early phase (≤48 h), short-term survival (survival to discharge), hospitalisation time (days), long-term survival (survival 1 year after discharge) and recurrence rate were evaluated.The safety of heparin therapy was assessed by comparing the drop in haematocrit during treatment in both groups.All statistical analyses were performed using commercially available statistical software (GraphPad Prism, Prism 9 for Windows, Version 9.1.2, San Diego, CA, USA). Data were tested for normal distribution using D’Agostino–Pearson omnibus tests and visual inspections of the data. For descriptive purposes, continuous variables are reported as means ± standard deviations or medians and ranges, depending on the data distribution. Group comparisons were made using t-tests, Mann–Whitney tests and Fisher’s exact tests. Values of p < 0.05 were considered significant.3. ResultsOverall, 45 cats received medical management of their pyothorax with small-bore sheath-guided thoracostomy tubes. Of these, 20 cats were included in the Standard Group, and 25 cats were included in the Intensive Group.3.1. Patient Data and Diagnostic Work-UpSignalment, medical history and physical examination findings are summarised in Table 2. There were 30 Domestic Shorthairs, 6 Maine Coons, 2 Persians, 2 crossbreeds and 1 British Shorthair, Russian Blue, Balinese, Somali and Kurile Bobtail. Of these, 28 were males (4 intact males), and 17 were females (1 intact female). The median age of the cats was 5 years (range 1 to 18 years).Main owner concerns were dyspnoea 80% (36/45), lethargy 76% (34/45) and hyporexia 91% (41/45). The median duration of symptoms before presentation was 7 days (range: 1 to 70 days). Of the cats, 82% (37/45) were already pretreated with antibiotics.The median respiratory rate at the time of presentation was 60 breaths per minute (range: 32 to 124 breaths). The mean heart rate was 171 (±35) beats per minute, and the median body temperature was 38.5 °C (range: 33.3 to 40.8 °C). There were no significant differences between the groups in terms of signalment, medical history or physical examination. All cats were diagnosed radiologically with bilateral effusion.Complete blood count (CBC), clinical chemistry and pleural fluid analysis, as well as the incidence of certain laboratory abnormalities are shown in Table 3. There was no significant difference in the laboratory parameters between the Standard Group and the Intensive Group.SIRS criteria were available for 96% (43/45) of the cats. SIRS was diagnosed in 40% (17/43) of these cats. In the Standard Group, 50% (9/18) of the cats were diagnosed with SIRS compared to 32% (8/25) in the Intensive Group. This was not a significant difference between the groups (p = 0.3443).Aerobic and anaerobic cultures of the thoracic effusion were performed in 89% (40/45) of the cats. No bacteria could be cultured in 3 cats. All 3 cats were pretreated with antibiotics. In 17 cats, only 1 bacterial species was cultured. In 12 cats, 2 bacterial species were detected; in 6 cats, 3 bacterial species were detected and in 2 cats, 4 bacterial species were detected. The detected bacterial species are listed in Table 4.3.2. TreatmentThe following antibiotics were administered: β-lactam antibiotics (amoxicillin/clavulanic acid and ampicillin) in 91% (41/45) of cats; fluoroquinolones (marbofloxacin and enrofloxacin) in 80% (36/45) of cats; metronidazole in 49% (22/45) of cats; amikacin in 4% (2/45) of cats and cefovecin in 4% (2/45) of cats. There was no significant difference between groups in the use of β-lactam antibiotics (p > 0.9999) or fluoroquinolones (p = 0.1573). Metronidazole was administered significantly more frequently in the Standard Group than in the Intensive Group (16/20 cats and 6/25 cats, respectively; p = 0.0003). Ampicillin was administered intrathoracically more often in the Standard Group than in the Intensive Group (18/20 cats and 6/25 cats, respectively; p < 0.0001).3.3. Complications and OutcomesThe population and the development over time are shown in Figure 2. In the early phase (≤48 h), 6 cats died due to the severity of their disease. In the Standard Group, 1 cat was euthanised due to septic shock with a secondary acute kidney injury. Another 3 cats died in this group due to septic shock. In the Intensive Group, 2 cats died due to septic shock. The survival rate in the early phase (≤48 h) was 87% (39/45) in all cats. The survival rate in the Standard Group was 80% (16/20), which was not significantly different (p = 0.3830) from the survival rate in the Intensive Group of 92% (23/25).After the initial stabilisation phase, structural changes requiring surgery were detected by diagnostic imaging in 16% (7/45) of cats. In the Standard Group, 2 cats were diagnosed with lung lobe abscesses. Of these, 1 cat was euthanised (day 6) at the owner’s request, and the other cat was successfully treated by thoracotomy (day 4). In the Intensive Group, 1 cat had a retroperitoneal abscess. This cat was euthanised (day 3) at owner’s request. Another 4 cats in this group required thoracotomies due to lung lobe abscesses (median: day 8, range: 4 to 11 days).In the course of treatment, 3 more cats died. In the Standard Group, pyothorax of 1 cat initially improved. After removal of the tubes, a severe thoracic effusion (transudate) occurred. At this timepoint, a reticular nodular lung pattern was detected on chest radiographs. This cat died acutely (day 11) and a necropsy was not allowed by the owner. In the Intensive Group, 1 cat was euthanised (day 13) at owner’s request after thoracotomy with lung lobe resection, subtotal pericardiectomy and partial pleurectomy, because the lung function did not improve after 48 h of controlled mechanical ventilation. Histopathological examination revealed chronic purulent pericarditis and pleuritis and multifocal atelectasis with purulent bronchopneumonia. Another cat in the Intensive Group was euthanised (day 5) at the owner’s request because of suspected septic encephalopathy with a brainstem lesion (score 11/18 on the modified Glasgow Coma Scale, assigned to the category with guarded prognosis [11]). Histopathology of the brain was unremarkable.Forty-three cats survived the first twelve hours after tube placement, so lavage was performed. Failure of closed-chest lavage occurred in 14% (6/43) of the cats. In the Standard Group, 1 cat had a thoracic effusion that was too viscous for adequate drainage, so streptokinase was successfully used intrathoracically (from day 8). In the Intensive Group, lavage failed in 5 cats. Of these, 2 cats had thoracic effusion that was too viscous, so alteplase was successfully used intrathoracically (from day 3 and day 6, respectively). Two cats did not improve sufficiently. In 1 of them, the amount of effusion and the number of cells in the effusion increased during the treatment so that alteplase was successfully instilled intrathoracically (from day 9). The second cat showed persistent bacteria in the effusion and septations on ultrasound examination (day 7). The cat underwent thoracotomy in another clinic. One cat developed lung lobe torsion during closed-chest lavage, so this cat underwent lung lobe resection (day 9).Overall, 53% (24/45) of the cats were successfully treated with small-bore thoracostomy tubes and closed-chest lavage alone. The 23 cats that were treated according to medical recommendations were discharged after a median of 10 days (range: 7 to 23 days). One cat was discharged early (day 7) at the owner’s request for financial reasons. In the Standard Group, 60% (12/20) of the cats were discharged after a median of 10 days (range: 7 to 23 days). Similarly, in the Intensive Group, 48% (12/25) of the cats were discharged after a median of 10 days (range: 7 to 15 days). The difference in survival to discharge with lavage alone and in length of hospitalisation was not significant between groups (p = 0.5503 and p = 0.4995, respectively). The hospitalisation of the cat in the Standard Group with closed-chest lavage and thoracotomy lasted 7 days. The cat in the Standard Group with closed-chest lavage and intrathoracic fibrinolysis was hospitalised for 14 days. The hospitalisation of the 5 cats discharged alive from the Intensive Group after closed-chest lavage and thoracotomy lasted a median of 17 days (range: 11 to 20 days). The hospitalisation of the 3 cats in the Intensive Group treated with closed-chest lavage and intrathoracic fibrinolysis lasted 14, 14 and 16 days.The short-term survival rate (survival to discharge) was 76% (34/45) for all cats. In the Standard Group, 70% (14/20) of the cats survived to discharge and in the Intensive Group 80% (20/25) survived. The difference was not significant (p = 0.5001). One cat from the Standard Group was lost to follow up after discharge.Thirty-three cats had a follow-up for a median of 1222 days (range: 14 to 4645 days).The recurrence rate was 9% (3/33). Only 8% (1/13) of the cats in the Standard Group showed a relapse (day 14). One cat was euthanised due to financial limitations and concurrent FIV (feline immunodeficiency virus) infection. In the Intensive Group, 10% (2/20) of the discharged cats showed a relapse. The early discharged cat relapsed on day 14 and was euthanised for financial reasons. The second cat relapsed on day 67 and was successfully treated again according to the intensive protocol. This cat had a disease-free follow-up of 1205 days.One cat was lost to follow-up 295 days after discharge.After discharge, 94% (30/32) of the cats survived at least 1 year. In the Standard Group, 92% (12/13) of the cats survived more than 1 year and in the Intensive Group 95% (18/19). There was no difference (p > 0.9999).Of the 43 cats with a follow-up, 70% (30/43) survived at least 1 year. In the Standard Group, 63% (12/19) of the cats survived and in the Intensive Group 75% (18/24). There was no significant difference (p = 0.5095).3.4. Tube Placement and ComplicationsIn the 45 cats, a total of 59 thoracostomy tubes were placed for initial chest drainage. In the Standard Group, 20 thoracostomy tubes were placed in 20 cats, while in the Intensive Group, 39 thoracostomy tubes were placed in 25 cats. All 59 tubes functioned in terms of drainage of pleural exudate. In 1 cat from the Intensive Group, the perforations of the tubes were partially subcutaneous, so the tube was judged as non-functional for lavage. The tube was removed. The cat was then successfully treated with 1 tube.During the course, a second tube was placed to empty the thorax on the opposite side in 7 cats in the Standard Group (median: 2 days, range: 1 to 11 days) and in 2 cats in the Intensive Group (day 2 and day 3). In 2 cats of the Intensive Group, a tube had to be replaced. In 1 cat, the tube was blocked (day 6). Another cat was able to pull the tube (day 6) itself. Thus, a total of 67 tubes were in place during the treatment. Of these, 27 tubes were placed in 20 cats of the Standard Group and 40 tubes in 25 cats in the Intensive Group. No complications were observed in the Standard Group. As mentioned above, in the Intensive Group, 1 tube was misplaced (partly subcutaneously), 1 tube was blocked and 1 tube was pulled by the cat. This resulted in a complication rate of 4% (3/67 tubes).3.5. Effect of the Treatment ProtocolSeventeen of the cats received bilateral tubes within 24 h of admission and 82% (14/17) survived. The other 28 cats received only 1 tube (n = 23) or the second tube (n = 5) was placed after 24 h. Survival to discharge in this group was 71% (20/28). The effect of early bilateral tubes on survival was not significant (p = 0.4927).As mentioned before, lavage could be performed in 43 cats. Lavage with heparinised balanced electrolyte solution was used in 23 cats and 87% (20/23) survived. Lavage with balanced electrolyte solution without UFH was used in 20 cats and 70% (14/20) survived. The effect of different lavage solutions on survival was not significant (p = 0.2634)The haematocrit of the cats decreased with a mean of 0.06 l/l (±0.06) during the treatment. The mean decrease in the Standard Group was 0.09 l/l (±0.06, mean day 6 ± 2, n = 11) and in the Intensive Group 0.04 l/l (±0.05, mean day 7 ± 2, n = 19). The decrease was significantly lower in the Intensive Group (p = 0.0326).4. DiscussionThe current study is the first to demonstrate the efficacy and complication rate of small-bore sheath-guided thoracostomy tubes for the treatment of pyothorax in cats. The more intensive protocol with the early goal of bilateral tubes and closed-chest lavage with UFH shows no significant effect.There is no established standard for the treatment of pyothorax in cats. Parenteral antibiotics are recommended as a starting point, while their administration into the pleural cavity is no longer recommended [12]. In addition, in most studies a closed-chest lavage is performed [1,2,3,4,5,6], which is also recommended in the current guidelines [12].Different techniques have been described for the placement of thoracostomy tubes for the treatment of pyothorax in cats. In most clinical studies, large-bore tubes (14–16 F) are placed over a rigid trocar under general anaesthesia [1,4,5,6]. Reported complications with these systems include death during general anaesthesia in 10% [4], pneumothorax in 11% [5], lung laceration in 17% [5] and subcutaneous fluid leakage in 6% of cases [5]. An experimental study demonstrates that small-(8 F) and large-bore (16–20 F) tubes are similarly effective in removing fluid or air from the pleural space of canine cadavers [13]. Two studies have been published in cats with pyothorax in which tubes of comparable size (personal communication) were placed over a trocar [3] (10 F) or through a needle [2] (8 F). The trocar technique required general anaesthesia in all cases [3]. The large needle for the 8 F tubes is sharp and may cause injury to the lungs or intrathoracic vessels. In addition, the needle has a larger outer diameter (10 F) than the tube (8 F), which may promote subcutaneous fluid leakage. A wire-guided technique is described for small-bore tubes (6 F) in 2 case series of cats with pyothorax (n = 8 [8] and n = 10 [9]). These tubes can be placed under sedation using an introducer needle (14 Ga or 18 Ga included in the kit) and a 0.035 inch guidewire [8]. Inadequate drainage effectiveness due to kinking of these tubes is described in 14% of cases (2/14 tubes) [8]. A possible cause could be the thin wall (0.4 mm) in relation to the large inner diameter (1.2 mm) [14].The sheath-guided tube technique described here is a combination of the advantages of the different tube systems. Anaesthetic complication, lung laceration and fluid leakage did not occur in our cats because the tube system is applicable under local anaesthesia with sedation and because a small (20 Ga) introducer needle, a tiny guidewire (0.018-inch) and a thin-walled sheath are used. Another advantage is the valved sheath, which reduces air entry during placement. Consequently, none of our cats developed a clinically relevant pneumothorax during tube placement, but due to the retrospective nature of our study, small amounts of pneumothorax could not be excluded. In addition, the sheath guidance helps to ensure optimal positioning of the tube. The tube used in this study is constructed for wound drainage with suction, therefore it has a relatively thick wall (0.5 mm). The thick wall (0.5 mm) in relation to the small inner diameter (1 mm) could result in high kink resistance [14]. Nevertheless, due to its multiple fenestration (20 side holes), the tube has a very good drainage performance, which also reduces the risk of blockage. Therefore, medical treatment failure due to tube complications, such as malposition or blockage, occurred in only 1 cat each, and kinking did not happen. One cat removed the tube itself, resulting in a low tube complication rate of 4% (3/67 tubes) in our study. In the literature, such tube complications are described in 11% [4] to 17% [5] of cases with large-bore tubes, 4% [2] to 14% [3] with medium-bore tubes and in 29% [8] with small-bore tubes.The goal in the Intensive Group was to clear as much as possible of the effusion early in the course of treatment and to improve lavage by using UFH. The reason for using UFH was to prevent the occlusion of thoracostomy tubes by fibrin clots [15]. Closed-chest lavage with UFH has been described in dogs and was associated with a higher long-term (12-month) survival rate [16]. UFH increases fibrinolysis to a small extent, as has been shown in an experimental study in rabbits [17]. It also has an anti-inflammatory effect that prevents the accumulation of leukocytes, as, for example, in ischaemic brain injuries [18]. Heparin is a glycosaminoglycan. A synthetic glycosaminoglycan showed a significant reduction in total cell count, total protein concentration and proinflammatory cytokine concentrations in peritoneal lavage fluid in a model of acute peritoneal inflammation in mice [19].We observed no difference in the complication rate between the Standard Group and the Intensive Group. The survival rate also did not differ between the treatment groups. This can be explained by several reasons. First, the small number of cats included in the study and the severity of the disease, which leads to a high mortality rate in the first 48 h [6], independent of the different treatment strategies. In addition, communication between both sides of the thorax has been described in cats [20,21], so both sides of the thorax may be adequately drained and lavaged via 1 tube. Single or multiple thoracentesis with parenteral antibiotic treatment has also been described to be curative [1,4].The overall survival rate for cats treated with sheath-guided tubes is 76%. This is comparable to survival rate in studies using large-bore tubes (14–16 F; 46% to 78%) [1,4,5] or medium-bore tubes (8–10 F; 65 and 72%) [2,3] and to a case series using small-bore tubes (6 F, 88%) [8]. Survival is mainly influenced by 3 factors: death from severe inflammation in the first 48 h and, less frequently at a later stage, structural disease requiring surgical intervention and, finally, tube failure with a need for additional intervention such as thoracotomy or fibrinolysis. Added to this are the financial concerns of the owner, which can lead to losses at all 3 of the aforementioned levels.In the current study, the acute mortality rate (≤48 h) of 13% is comparable to the rates in other studies of 17% to 22% [1,2]. Pyothorax can lead to sepsis in cats, which is likely the cause of the high mortality rate in the first 48 h [1,22]. In cats, sepsis is defined as meeting SIRS criteria and having a documented source of infection [10,22]. Overall, 40% of cats were SIRS positive with no significant difference between the Standard and Intensive Groups. Future treatment strategies should include more aggressive treatment of this problem.Another factor is the structural diseases detected by diagnostic imaging. In our cat population, 16% had structural changes, which is more than twice as many cats as in other studies (4% to 6%) [1,2,5]. Surgical intervention for these patients has a good prognosis in all studies, but sometimes owners refuse surgery because of emotional or financial reasons.Medical treatment fails due to lavage failure in 5% to 9% of cases in studies of cats with pyothorax [1,2,3,4]. In these cases, thoracotomy is usually recommended. In our cat population, lavage failure occurred in 14% of cats. One of these cats required a thoracotomy for lung lobe resection due to lung lobe torsion. Another cat underwent thoracotomy in an external clinic due to lack of improvement. Four other cats received intrathoracic fibrinolysis due to an overly viscous effusion or insufficient improvement. Inflammation leads to an accumulation of extravascular fibrin, as local coagulation is increased while fibrinolysis is decreased. Therefore, intrapleural fibrinolytic therapy has been investigated in humans with empyema [23]. Following this, we have successfully used fibrinolysis at an extrapolated dose in all 4 cats. The use of a higher dose of UFH or earlier use of fibrinolysis may be a treatment option in the future.A side effect of lavage with heparinised balanced electrolyte solution might have been a bleeding tendency. In both groups, there was a decrease in haematocrit during hospitalisation. Hospital-acquired anaemia is common in cats [24,25]. It is thought that anaemia of inflammatory disease is the most common cause of this finding [26]. The Intensive Group with UHF in the lavage solution had a slightly lower drop in haematocrit, making a bleeding tendency unlikely. A possible reason for the difference could be a small variation in dehydration status between the 2 groups at presentation. For a more objective assessment of the bleeding risk, in the form of an iatrogenic coagulopathy, the measurement of clotting times and/or antifactor Xa activity would be helpful. This should be considered, especially when planning a possible higher UFH dose in the future.One cat developed lung lobe torsion during lavage. This is rarely described in cats with other types of pleural effusion (especially chylothorax) [27] and should be considered, particularly if the thoracic effusion changes composition.The relapse rate in our study was 9%, which is similar to that reported in other studies (5% to 8%) [1,2,4,6]. When the analysis included cats that required retreatment, the long-term survival (survival 1 year after discharge) was comparable, with 92% in the Standard Group and 95% in the Intensive Group. Another cat study reported a similar long-term survival rate of 97% [2].Our investigations have some limitations. First of all, the retrospective character of the study and the large time frame (change in treatment over time) may influence our results. Furthermore, the low statistical power likely affects our results. Although pyothorax is named as the third most common cause of pleural effusion in cats [28,29,30], the condition is rare. This is also reflected in the number of cases in published studies. Despite long observation periods, the number of cases is predominantly less than 40 cats. A prospective study, with case numbers that allow an adequate power of evaluation, is therefore only feasible as a multicentre project. However, in order to plan and justify a meaningful study protocol, an evaluation of our previous data is essential.5. ConclusionsBased on our data, we can show that the therapy of pyothorax with small-bore thoracostomy tubes is comparably successful as the therapy with large- or medium-bore tubes.In addition, we did not observe an increased complication rate due to bilateral tube placement or closed-chest lavage with a heparinised lavage solution.Cats treated according to the intensive protocol could be discharged slightly earlier compared to the standard protocol. This group also appeared to have a higher percentage of surviving cats. Larger studies are needed to test the validity of these results.

animals : an open access journal from mdpi

10.3390/ani12070933

PMC8996929

Calves could be reared using automatic computer-controlled feeding machines or in a foster cow system, where one cow usually nurses two calves. The unrestricted access of calves to fresh milk in the foster cow system might have a beneficial effect on their health and later fattening performance and, ultimately, meat quality. Therefore, the aim of this study was to compare the effects of these rearing systems and fattening intensities on bioactive compounds content in beef originated from young bulls. An analysis of different rearing systems and fattening intensity levels indicates that natural calf rearing should be followed by intensive fattening to produce beef with optimal sensory attributes and a high nutritional value.

The study was aimed at determining the effects of the rearing system and intensity of fattening on beef physicochemical properties and sensory quality, fatty acid composition, and mineral compounds and vitamins concentration. The study was conducted using meat from 38 young, crossbred bull calves, which were reared with nurse cows (C) or were fed milk replacer (R). In the study, intensive (Int) or semi-intensive (SInt) fattening system were applied. The bulls were slaughtered at the age of 560 days and samples of the longissimus lumborum (LL) muscle were collected. Meat from C bulls was juicier (p < 0.05) and had a higher concentration of conjugated linoleic acid (CLA), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA), as well as zinc (Zn), iron (Fe), and α-tocopherol, compared with meat from R bulls. The Int system resulted in the intramuscular fat (IMF) content increase (p < 0.01) and reduced shear force (p < 0.05), compared with the SInt system. Meat from Int bulls had a better eating quality and a higher monounsaturated fatty acid (MUFAs), Zn, and Fe; however, it had a lower proportion of polyunsaturated fatty acids (PUFAs) and α-tocopherol concentration than meat obtained from SInt bulls.

1. IntroductionIn many countries where cattle are raised mostly for milk, to inseminate a part of cows and heifers in dairy herds, a beef bulls’ semen is usually used [1,2]. Several-day-old, crossbred calves being the progeny of dairy cows (or heifers) and beef bulls are sold to specialized in cattle fattening farmers. On such farms, calves are most often reared using automatic computer-controlled calf feeding machines [3]. However, those machines are expensive and the group housing of calves is associated with an increased risk of disease transmission. Therefore, alternative calf rearing systems are being sought [4]. One of these alternatives is a foster cow system [5], where one cow usually nurses two calves, and unrestricted access to fresh milk affects beneficially in later fattening results and, ultimately, beef quality [6,7]. Reddy et al. [8] demonstrated that nutritional stimulation in the early life stages of calves, which involved the administration of readily digestible high-energy diets, might result in metabolic imprinting demonstrated in the carcass and meat quality modification, especially backfat thickness and intramuscular fat scores.The bases for beef production are male young bulls or steers, which are grazed intensively or semi-intensively, depending on the technology adopted [2]. In the semi-intensive system, animals are fattened using a fodder with a high proportion of silage and a small addition of concentrates, which extends the fattening time. This prolonging fattening period increases the animals age at slaughter, which may result in deterioration of the sensory quality of beef, i.e., its tenderness [9]. When grazing bulls, better results are obtained by using intensive feeding [10]. Bull calves reach slaughter maturity early and achieve optimal final body weight, and thus burden the environment to a lesser extent [11]. Moreover, bulls are a better choice than steers due to a higher feed efficiency and larger growth [12]. When environmental issues are discussed, it seems that there is a trend towards sustainable produced beef, especially that which is ‘grass-fed’. However, as it was reported by McGee et al. [12], the intensively fattened bulls, which received grass silage and concentrate showed lower greenhouse gases emission in relation to live, carcass, and meat weight gain.Experiments conducted in our research team focus on the combined impact of the rearing system and intensity of feeding on the growth rates, carcass value, and meat quality from young bulls. Results regarding the growth rates of Polish Holstein–Friesian × Hereford calves, fattening performance and carcass characteristics were described in detail by Nogalski et al. [10]. It was concluded that a suckling system with nurse cows is more beneficial than using a milk replacer distributed from automated stations. Those nursed calves were healthier and had a higher body weight gain, and therefore were more suitable for fattening than calves which received milk replacer. Crossbred Polish Holstein–Friesian × Hereford bulls fattening indices and carcass traits were positively affected by an increased concentrate share in the fodder. However, the study [10] did not comprise the evaluation of meat quality. Thus, this study was aimed at analyzing the rearing system and the intensity of fattening effects on the physicochemical and sensory attributes of the meat, fatty acid composition, and concentration of mineral compounds and vitamins in meat from young, crossbred bulls. A hypothesis that natural rearing and intensive fattening of bulls contribute to increased concentration of bioactive compounds in beef, making it more valuable from a nutritional perspective, was tested in this study.2. Materials and Methods2.1. CalvesThe study was conducted using young, crossbred bulls (n = 38) produced by inseminating Polish Holstein–Friesian cows with Hereford bulls’ semen. Until 150 days of age, those calves were raised naturally with nurse cows (C) or were fed milk replacer (R). During the rearing period, C calves turned out to be healthier—they had a better survival rate and a higher average daily body weight gain (by 0.15 kg) than R calves. After a 30-day transition period, the animals were fattened intensively (Int) or semi-intensively (SInt). The feeding period lasted from 181st day to 560th day of age. All details about calf rearing and bull fattening were provided in a previous paper [10]. The authors of the study obtained an approval for conducting the experiment from the University of Warmia and Mazury in Olsztyn Ethics Committee for Animal Experimentation (Decision No. 121/2010).2.2. Slaughter and Meat SamplingAfter termination of the fattening period, young, crossbred bulls were transported to a slaughterhouse, where they rested in individual boxes for 15 to 20 h; unlimited access to water was provided. The slaughter procedure was conducted in accordance with European Commission Regulation [13]. No electrical stimulation was applied to the carcasses. After 48 h, the pH48 value was measured in the longissimus thoracis (LT) muscle, between the 10th and 11th thoracic vertebrae (HI 8314 pH-meter with FC 200 combined electrode; Hanna Instruments, Olsztyn, Poland) [14]. Before the determinations, the pH-meter was calibrated with the use of pH 7 and pH 4 buffers. Ninety-six hours after the slaughter, the carcasses were dissected and approx. 1000 g samples of the longissimus lumborum (LL) muscle were collected from the right half-carcass of each animal (from the 1st to 3rd lumbar vertebrae). The muscles were transported to a laboratory under refrigerated conditions (delivery time of approx. 1 h) and kept at refrigerated temperature (4 ± 1 °C) overnight. Next, the muscles were divided and subjected to analyses.2.3. Determination of Physicochemical Attributes of MeatPrior to analyses, beef muscles were trimmed from external fat and individually comminuted using a meat grinder (ZMM4080, Zelmer S.A., Rzeszów, Poland) with 3 mm mesh. Dry matter was determined by drying meat samples at 103 ± 2 °C (UF55, Memmert GmbH+Co. KG, Schwabch, Germany) to a constant weight. The concentration of protein was determined according to the Kjeldahl method [15], using Büchi Labortechnik AG (Flavil, Switzerland) equipment; fat was determined according to the Soxhlet method using the Buechi B-811 extraction system, with hexane as a solvent [16] in oven-dried samples, and ash content was determined by ashing samples in quartz crucibles in a muffle furnace (FCF 22SM, Czylok, Poland) for 16h [17]. Drip loss in raw whole meat according to the Honikel method [18] and cooking loss—assessed on whole meat cooked to 75 °C [18]—were determined using a AKA 2200 (AXIS, Gdańsk, Poland) scale; drip loss and cooking loss indicated the water holding capacity (WHC) of muscle tissue.LL color was evaluated in the CIE L*a*b* system [19] on the surface of freshly cut meat after 60 min of blooming using the Konica Minolta CR-400 (Sensing Inc., Osaka, Japan) (with a 2° view angle, D65 illuminant). Measurements were conducted in triplicate at randomly selected points and lightness (L*), redness (a*), and yellowness (b*) values were recorded. Chroma (C*) was calculated according to the formula C* = (a*2 + b*2)0.5, whereas hue angle was calculated according to the formula h° = atan (b*/a*)·180°/(Π).A 2.5 cm thick steak (approx. 200g) obtained from each LL muscle was cooked in an individual plastic pouch in a water bath (W415E, Laboplay, Bytom, Poland) at 80 °C until a temperature of 75 °C was reached to determine the Warner–Bratzler shear force (WBSF, N). WBSF was measured on cuboid samples (10 × 10 mm, approx. 40 mm long, n = 5 from each steak) with the use of the Instron 5942 universal testing machine (Instron, Norwood, MA, USA) equipped with a V-shaped shear blade with a triangular aperture of 60° (load 500 N, head speed 200 mm/min). All details of the procedure were provided in [14].2.4. Sensory EvaluationSamples for sensory evaluation were prepared by cooking a 2.5 cm thick steak (approx. 200 g) obtained from each LL muscle in an individual plastic pouch in a water bath at 80 °C until an internal temperature of 75 °C was reached. Immediately after the termination of the thermal treatment, the samples were evaluated in accordance with Standard PN-ISO 4121 [20], by a six-person team trained and experienced in sensory evaluation of meat. A detailed description of the method was provided in Modzelewska-Kapituła et al. [14]. A total of seven sensory analysis sessions were performed, a maximum of six meat samples being assessed per session; the same panelists took part in all sessions. Panelists evaluated each LL sample in terms of typical beef aroma intensity (1, imperceptible; 5, extremely intense), juiciness (1, extremely dry; 5, extremely juicy), tenderness (1, extremely tough; 5, extremely tender), and typical beef taste intensity (1, imperceptible; 5, extremely intense).2.5. Determination of Fatty Acid Composition, Mineral Compounds, and Vitamin A and E ContentFatty acid composition was determined according to PN-EN ISO 5508 [21] and PN-EN ISO 5509 [22] standards. To obtain fatty acid methyl esters, the modified Peisker method [23] was used. The fatty acids were determined by gas chromatography, using the Varian CP 3800 system (Varian, Palo Alto, CA, USA). All the details of the analysis were provided in the paper by Nogalski et al. [24]. Saturated fatty acids (SFAs), unsaturated fatty acids (UFAs)—including monounsaturated fatty acids (MUFAs)—and polyunsaturated fatty acids (PUFAs) were reported as the relative percentage of total fatty acids along with the following calculated ratios: UFA/SFA, PUFA/SFA, and n-6/n-3 PUFA.The minerals such as potassium, sodium, magnesium, zinc, and iron were determined using an atomic absorption spectrometer (Candela, Warsaw, Poland) according to the method described in Nogalski et al. [24].The content of vitamins A (retinol) and E (α-tocopherol) was determined based on the applicable standards [25], with slight modifications [21], using a high-performance liquid chromatography (HPLC) using chromatograph 920-LC (Varian, Palo Alto, CA, USA), equipped with a Polaris C18-A (Agilent Technologies, Santa Clara, CA, USA) silica column). The separated compounds were identified by two detectors in tandem (UV-visible photodiode array detector and fluorescence detector). The analysis was conducted in duplicate.2.6. Data AnalysisData were analyzed using Statistica 13.3 software [26]. To determine if the rearing system (C and R) and fattening intensity (Int and SInt) affected meat attributes, the least squares method was used. To conduct the analysis, the following model was created:Yijk = μ + Ai + Bj + (AB)ij + eijk where Yijk is the analyzed parameter value, μ is population mean, Ai is the effect of rearing system (1, 2), Bj is the effect of fattening intensity (1, 2), (AB)ij is the rearing system x fattening intensity interaction, and eijk is random error.To study the similarities between the treatments obtained from the nurse cow (C) and milk replacer (R) rearing systems, fattened intensively (Int) or semi-intensively (SInt), a cluster analysis was conducted using only these variables, which were affected by either rearing or fattening system. The data set included: fat content, WBSF, the proportion of conjugated linoleic acid (CLA), eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), docosahexaenoic acid (DHA), PUFAs, n-3, n-6/n-3 ratio, and Fe, Zn, and α-tocopherol contents. In the analysis, a single linkage between treatments and Euclidean distance were applied.3. Results3.1. The Influence of Rearing System and Intensity of Fattening on the Proximate Composition, Physicochemical, and Sensory Quality of MeatRearing system did not affect (p > 0.05) beef proximate composition (Table 1), nor its physicochemical properties (pH, color, WHC and WBSF, Table 2.) On the contrary, the intensity of feeding affected fat content in LL muscles, which was higher (p ≤ 0.01) in the Int system than in the SInt system (Table 1). However, no differences in the remaining components of meat tissue (dry matter, protein, and ash), as well as pH48 between Int and Sint raised bulls’ meat, were noted (p > 0.05).The meat from Int bulls was lighter (L*), with a higher contribution of redness (a*) and yellowness (b*), but the noted differences were not significant (Table 2). This resulted in insignificant differences in chroma and hue angle between treatments. The experimental factors, such as rearing system and intensity of feeding, had no influence on the WHC of meat, which was expressed in the lack of significant differences in drip and cooking losses between the treatments (Table 2). The results of an instrumental evaluation of meat tenderness indicated that the meat of Int bulls was characterized by lower WBSF values than the meat of SInt bulls, and the effect of fattening intensity was significant (p ≤ 0.05) (Table 2).Feeding levels also exerted a significant effect on the sensory attributes of the LL muscle (Table 3). Meat from Int bulls received higher scores, for all analyzed attributes, than meat from SInt animals. Only meat juiciness was affected by the method of calf rearing, and C bulls’ meat was juicier than R bulls meat (Table 3).3.2. The Influence of Rearing System and Intensity of Fattening on the Longissimus Lumborum Fatty Acid CompositionThe fatty acids concentrations in the IMF of the LL muscle were affected by both rearing system and fattening intensity (Table 4). The natural suckling system (C) and a better health status of calves contributed to an increase in conjugated linoleic acid (CLA), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) concentrations, whereas SInt fattening increased the concentrations of docosapentaenoic acid (DPA) and total PUFAs, while decreasing MUFAs proportion. Moreover, the natural suckling system (C) increased n-3 fatty acids proportion (p ≤ 0.05), thus decreasing the n-6/n-3 ratio (p ≤ 0.01).3.3. The Influence of Rearing System and Intensity of Fattening on the Content of Macronutrients, Micronutrients, and Vitamins in MeatCalf rearing system significantly affected (p < 0.05) the content of the analyzed micronutrients in the longissimus lumborum (LL) muscle (Table 5). C bulls’ meat showed significantly higher concentration of zinc (Zn) and iron (Fe) than meat from R bulls. The natural suckling system (C) also contributed to an increase in α-tocopherol concentration in meat. Meat from C bulls had higher (p ≤ 0.05) vitamin E content than meat from MF bulls. The concentrations of α-tocopherol, Zn, and Fe in meat were affected by fattening intensity. Beef from the SInt system showed a higher (p ≤ 0.05) vitamin E concentration compared with beef from the Int system, whereas Zn and Fe concentrations were higher in the meat from Int bulls.3.4. Cluster Analysis ResultsThe cluster analysis clearly showed the two clusters differed with fattening intensity: animals from intensive fattening differed from those fattened semi-intensively (Figure 1). This suggests that although rearing system is important and affected some of the quality attributed to the meat, the fattening system strongly affects beef quality, including fat content and the proportion of valuable fatty acids, minerals such as Fe and Zn, and vitamin E content.4. Discussion4.1. The Proximate Composition and Physicochemical and Sensory Characteristics of the Longissimus LumborumIn this study, the intensive fattening system (Int) resulted in a higher IMF content in the LL muscle compared with semi-intensive system (SInt). The results of this study are similar to those reported by Mezgebo et al. [27], who also found that using high levels of ground grain in cattle fodder increased the IMF content in beef. A similar was noted by Therkildsen et al. [28]. The IMF content at the beginning of the growth period in calves is likely to play a crucial role in shaping the IMF content after finishing. The phenomenon can be explained by the fact that adipose tissue starts to accumulate fat in the early weaning period [29] and the higher IMF content in this life stage, the higher level of IMF at the end of fattening and the same in meat. The results of this experiment are also consistent with those reported by Modzelewska-Kapituła and Nogalski [30], who noted that although intensity of feeding had an impact on the IMF content, dry matter and protein content remained unaffected in the infraspinatus muscles of Polish Holstein–Friesian bulls. On the other hand, Schoonmaker et al. [31] found that the IMF content increased in the muscles of early-weaned steers when they were fed ad libitum with rations containing a high proportion of concentrate during the growing stage. Nevertheless, the effect did not occur when steers were offered the same diet in the finishing period and the rates of fat deposition were slower. In this study, no effect of calf rearing system on the IMF content in LL muscle was noted, which is in agreement with the results reported by Greenwood et al. [32]. In their study, no impact of rearing system and BW gain on the IMF content in beef was shown.The values of WBSF, which indicate the tenderness of muscle tissue, depend on the structure of both intramuscular connective tissue and myofibrillar proteins, which in turn are affected by muscle type and feeding system [14]. It was shown that animal feeding is a key factor which affects the biological reactions which proceed in muscle cells, such as muscle protein turnover [33]. In this study, the meat of Int bulls was characterized by lower WBSF values, compared with bulls reared in the SInt system. These results are different from those reported by Therkildsen et al. [28] and Cox et al. [34]. In the study by Therkildsen et al. [28], an increase in feed energy value had no influence on the WBSF values or tenderness of young bulls’ meat. Similarly, Cox et al. [34] reported that diet (forage vs. grains) had no effect on the loin WBSF values.Intramuscular fat (IMF) indicates the quantity of fat in muscles and is visible, described as fat spots or deposits (so called marbling) in some beef carcass elements. IMF content positively influences the sensory attributes of meat, such as juiciness, flavor, and tenderness [14,29,35], in addition to WBSF values [36]. This was also noted in the present study, where the higher IMF concentration in Int bulls’ meat corresponded with an increase in the desirable sensory properties of the meat. Moreover, meat with the best sensory quality (group Int) had a lower value of WBSF. Meat from C bulls was juicier, which could result from a higher growth rate of nursed calves [10]. Bispo et al. [37] found that nursed calves were characterized by higher BW gains and their meat was juicier, compared with calves that could not suckle from their mothers. In contrast, Hennessy et al. [38] demonstrated that samples of the longissimus lumborum et thoracis muscle originated from calves showing lower growth rates before weaning had a better tenderness than those from animals which had higher growth rates. In this study, meat from C calves had a somewhat higher IMF content than meat from R calves (0.09% and 0.2%, a non-significant difference), which could partially explain the observed difference in meat juiciness.4.2. The Quality of Fat in the Longissimus Lumborum MuscleThe natural suckling system and a better health status of calves [10] contributed to the elevated levels of value from human nutrition fatty acids such as CLA, EPA, and DHA. A high share of SFAs and a low proportion of PUFAs in the fat result from the hydrogenation of dietary fat by ruminal microbiota [39]. For optimal results, beef producers should decrease the concentrations of SFAs in fat and/or increase the content of PUFAs, especially n-3 fatty acids [40]. In this study, the natural suckling system caused an increase in the proportion of n-3 fatty acids. The findings of De la Fuente et al. [41] suggest that artificially reared lambs have a lower rumen functionality, which affects their absorption of fatty acids synthesized in the rumen via the anaerobic microbial fermentation of fiber and starch [42]. In a study by Osorio et al. [43], ewes’ milk had a beneficial influence on the concentrations of n-3 fatty acids and the n-6/n-3 PUFA ratio in the IMF of lamb meat, compared to milk replacer. Wielgosz et al. [4] reported that diseases and infections during the rearing period negatively affected IMF composition in calves. In addition, in this experiment, diseases that occurred in the pre-weaning period might have influenced the post-weaning health status of calves and, consequently, the fatty acid profile. In this study, SInt bulls’ meat had a lower IMF content, but the composition of the fat was more favorable compared to that from Int bulls, due to the fact that semi-intensive fattening increased DPA and total PUFAs contents in meat.4.3. The Content of Macronutrients, Micronutrients, and Vitamins in MeatAs a result of natural rearing, Zn and Fe contents increased in meat, most likely because calves could obtain high-quality milk by suckling from nurse cows. Bispo et al. [6] reported that the earliest weaned calves produced meat with a lower nutritional value. The contents of minerals in beef may be associated with the fat content, which was shown in a study Williams et al. [44], where Zn, Fe, P, Na, and K concentrations were negatively correlated to carcass fatness [44]. A similar result was noted in this study, where higher Zn and Fe contents were noted in the meat with a lower IMF content. From a nutritional perspective, the meat from C bulls was a rich source of Zn, satisfying 52% and 23% of the RDA (Recommended Dietary Allowance [45]) in adult females and males, respectively, when consumed at the amount of 100 g (the weight of cooked meat).In this study, both the natural suckling system and semi-intensive fattening had a beneficial influence on the α-tocopherol concentration in meat. This can be explained by the fact that in the meat of animals fed mainly roughage, which is used in semi-intensive fattening, higher concentrations of vitamin E are noted [46]. At the same time, the roughage feeding also causes elevated proportions of long-chain unsaturated fatty acids (DPA and total PUFAs). It should be noted that, in the meat which has an increased content of PUFAs, there is a higher risk of detrimental lipid oxidation and off-flavor development during shelf-life. However, the lipid oxidation in meat might be inhibited by high vitamin E content [47]. Therefore, the simultaneous increase in PUFA and vitamin E is highly desirable in light of beef shelf-life. Antioxidants, such as vitamin E, not only prevent those undesirable changes in lipids, but slow down metmyoglobin formation, in addition to the same unfavorable changes in meat color [47]. To exert the inhibition, a minimum of 3–4 µg of α-tocopherol/g of fresh beef is needed [48]. In our study, the concentration of α-tocopherol was within this range, excluding beef from bulls fed with milk replacer fattened intensively; therefore, a positive effect on lipid oxidation and color stability might be expected. However, to verify this hypothesis, further studies are needed.5. ConclusionsIn conclusion, both rearing and feeding systems affected beef quality; however, the influence of feeding system was stronger than rearing system. The advantage of suckling system over milk replacer was shown in this study; therefore, it is recommended to raise calves using less valuable cows for nursing. Meat from bulls raised using nurse cows, compared with meat from bulls, which received milk replacer, was juicier and the concentration of functional fatty acids (CLA, total n-3 PUFA, including EPA and DHA), in addition to Zn, Fe, and α-tocopherol, was higher. Intensive fattening, compared with semi-intensive fattening, contributed to the IMF increase in the LL muscle and a reduction in the values of WBSF. Meat from intensively reared bulls was more desirable in terms of sensory quality, and had higher concentrations of MUFAs, Zn, and Fe. Although the beef obtained from semi-intensive fattening had more desirable fatty acid composition in terms of PUFAs and DPA proportions, it had also a lower intramuscular fat concentration; therefore, it will not probably be a much better source of those components in a human diet, compared with the meat from intensive fattening. An analysis of different rearing systems and fattening intensity levels indicates that natural calf rearing should be followed by intensive fattening to produce beef with optimal sensory attributes, delivering health benefits to consumers.

animals : an open access journal from mdpi

10.3390/ani13061069

PMC10044481

Vegetable oil inclusion in fish diets is a common practice, but their effects in the oxidative status at intestinal level, which would affect animal health and welfare, are poorly understood. In the present study, we compared the effects of different dietary treatments containing soybean oil alone or in combination with other vegetable oils in sea bream. Overall, the results revealed that the blend of soybean and linseed oils negatively affects intestinal integrity as it triggered high oxidative stress that could not be counteracted by the high levels of antioxidant enzymes. However, the addition of palm oil to the previous mixture of vegetable oils makes it possible to maintain low oxidative stress, preserving the intestinal health of the animal. In conclusion, this study demonstrates the importance that the mixture of vegetable oils in a diet can have on the intestinal health of sea bream.

Fish oil is commonly replaced by vegetable oils in sea bream diets, but little is known about their effects on intestinal health regarding oxidative stress biomarkers. The negative effects of lipid peroxidation on digestive mucosa could have consequences in animal nutrition and welfare. In this study, five isonitrogenous (46%) and isolipidic (22%) diets with 75% of vegetable oils inclusion were evaluated: soybean oil (S) alone or different mixtures containing soybean oil with linseed (SL), linseed and rapeseed (SLR), linseed and palm (SLP), and linseed, rapeseed, and palm (SLRP). Gilthead sea bream juveniles were fed twice a day for 18 weeks. Pyloric caeca and proximal intestine samples were collected 24 h post feeding for lipid peroxidation (LPO), antioxidant enzyme activities (SOD, CAT, GPx, GST, and GR) and gene expression analyses. Pyloric caeca presented larger unhealthy changes in oxidative status than proximal intestine. Although SL-fed fish showed the highest antioxidant activities, they were unable to cope with LPO that in pyloric caeca was 31.4 times higher than in the other groups. Instead, SLP fish presented the best oxidative status, with low LPO levels, antioxidant enzyme activities, and gene expression. In summary, between the vegetable oils dietary mixtures tested, SPL would maintain better intestinal health.

1. IntroductionFor the sustainability of aquaculture, it is necessary to replace fish oil (FO) with alternative ingredients that maintain the quality, health, and welfare of the fish, while allowing to reduce production costs [1]. This need arises from the high prices of FO due to its use in nutraceutical and agricultural industries because of its health properties [1]. Vegetable oils (VO) are primarily used to replace FO in feed formulation and among the most used vegetable oils, we can find soybean oil, which has 64% of polyunsaturated fatty acids (PUFAs), mainly linoleic acid (C18:3n-6) (57% of the total fatty acids) and rapeseed oil, rich in monounsaturated fatty acids (MUFAs) (57%), mainly oleic acid (C18:1n-9) [2]. Among the VO, there is also linseed oil, very rich in n-3 fatty acids with 75% of PUFA, mainly α-linoleic acid (18:3n-3), moderate content of MUFA (16%) and low in saturated fatty acids (SFA, 9%), but with a high market price, which limits its use in fish feeds [2]. Moreover, we can also find palm oil, rich in SFA and MUFA (50 and 40% of total fatty acids, respectively), its use being limited indeed by its amount of SFA [2]. Therefore, although FO replacement is an unavoidable necessity, an imbalance in the dietary fatty acid profile could negatively affect fish health and welfare, especially in marine species such as gilthead sea bream. In this species, eicosapentaenoic (EPA, 20:5n-3) and docosahexaenoic (DHA, 22:6n-3) acids among others, are essential fatty acids required for proper development and growth [2,3,4,5], and thus, a minimum amount of FO would be necessary in those diets to meet the fish requirements concerning these fatty acids.It is well known that FO replacement by VO affects directly the gastrointestinal tract by changing the membrane fatty acid composition due to the fast turnover of enterocytes, endangering their primary functions: digestion and absorption [6,7]. The effect of VO is very dependent on dietary factors such as the type of both FO and VO incorporated, the level of FO substitution, the presence and availability of other nutrients (particularly antioxidants) [8] and the duration of feeding. Moreover, it is also influenced by other non-dietary factors such as the fish species, fish size, and life cycle stage, as well as environmental conditions, mainly temperature [9]. However, the use of blends of different VO can reduce or avoid the deleterious effects on growth performance in fish fed diets with less than 7% of fish meal and FO sources [10,11,12].When delving into the effects at the intestinal level, dietary changes could also provoke morphological alterations that impair intestinal functionality, including paracellular permeability, and epithelial transport functions, specifically, membrane digestive enzymes, amino acids and glucose transporters activities and diffusion rates, thus affecting nutrient utilization [9,13]. In this sense, VO inclusion led to a supranuclear lipid droplet accumulation in Artic charr (Salvelinus alpinus) fed a diet that included linseed oil [14] and in rainbow trout (Oncorhynchus mykiss) fed a diet including soybean oil [15,16], probably due to a reduction in lipoprotein synthesis as it was described in gilthead sea bream [6,17]. Nevertheless, the total replacement of FO by soybean oil in a diet for Atlantic salmon (Salmo salar) did not affect enterocyte lipid accumulation [18]. Moreover, dilated intercellular spaces were found in the intestine of sea bream fed 60 and 80% of soybean oil inclusion reflecting an impaired transit capacity through the lamina propria [17]. Thus, VO inclusion can directly affect the gastrointestinal tract at different levels: membrane composition, structure, integrity, and function. Furthermore, the oxidative status of this organ could also be affected. The formation of reactive oxygen species (ROS) is a natural process caused by cellular metabolism itself and its interaction with the environment. Dietary changes of both lipid content as fatty acid degree of unsaturation and length [19] could modify ROS formation due to an imbalance between production and removal. It is known that an increase in ROS triggers DNA damage, enzyme inactivation, protein oxidation, and lipid peroxidation (LPO) [20]. This could compromise the physical barrier function at intestinal level by causing negative effects on membrane structure, fluidity and permeability [21] and thus, negatively affect fish welfare [22,23,24].Antioxidants are the main cellular mechanisms to fight ROS and maintain physiological status [20]. Cells have developed enzymatic antioxidant mechanisms against oxidative stress [25], which include superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), glutathione S transferase (GST), and glutathione reductase (GR), which is coupled with GPx and GST and recycles oxidised glutathione (GSSH) [20]. Moreover, organisms also have non-enzymatic antioxidants with low molecular weight that directly quench ROS, such as vitamins, carotenes, and glutathione (GSH) among others [20,26].Specific studies on oxidative status in pyloric caeca and proximal intestine are scarce despite that are multiple external and internal triggers inducting oxidative stress at intestinal level [21,22,27,28,29,30,31,32]. To our knowledge, two studies in this area have been conducted in gilthead sea bream, an important species for Mediterranean aquaculture [22,30]. Castro et al. [22] tested the effect of total replacement of FO by a blend of palm, linseed, and rapeseed oils (30:50:20) in fish fed high-protein diets (50–66% from fishmeal) in whole intestine and they found a protective role of VO against LPO in fish fed diets containing VO versus fish fed a FO diet. Moreover, that study revealed that LPO and antioxidant status were tissue specific, since liver and intestine LPO levels and antioxidant activities showed different profiles. Magalhaes et al. [30] tested diets with the same blend of VO but supplemented with arachidonic acid (ARA, 20:4n-6), EPA and DHA and their results revealed the importance of the ratio n-6/n-3 long-chain (LC)-PUFA content in the proximal intestine oxidative status. In this sense, some other studies have demonstrated that appropriate levels of n-3 LC-PUFA could improve antioxidant capacities against oxidative stress in salmonids [23,33,34].The present study seeks to investigate if the dietary inclusion of different VO including soybean oil alone or combined and resulting in different n-6/n-3 ratios, could improve the antioxidant status and LPO, both in the pyloric caeca and in the proximal intestine of gilthead sea bream, looking for a healthy impact on fish of the FO substitution by VO while improving the sustainability of the aquaculture sector.2. Materials and Methods2.1. Experimental DietsSkretting ARC (Norway) formulated and produced five isonitrogenous (46%) and isolipidic (22%) diets (Table 1) in which 25% of included oils was FO and 75% of VO. One diet contained only soybean oil as VO (S diet), in the second diet, soybean and linseed oils were blended (SL diet). In another two other diets, a combination of three VO was used, where soybean and linseed oils were mixed with either rapeseed or palm oils (SLR and SLP diets, respectively) with a similar amount of soybean oil in both of them. Finally, in the last diet, a blend of all VO was included (SLRP diet).All diets, except diet S, presented high and similar levels of n-3, achieving the optimum levels for this species. Moreover, as soybean oil diminished, the amount of n-6 fatty acids also decreased, producing diets with different n-6/n-3 ratios; however, at the same time, the amount of MUFAs increased due to the inclusion of other VO showing the diets’ different fatty acids profiles (Table 2). All diets contained 88% of vegetable protein from soya concentrate (30%), corn gluten (15%), fava beans (6%), wheat gluten (3.8%), sunflower meal (3%), as well as wheat (7%) and fishmeal (15%).2.2. Fish and Feeding TrialThree hundred and eight sea bream (81.73 ± 0.36 g) were randomly distributed in a semi-recirculating saltwater system of 14 fiberglass tanks (500 L; 22 fish per tank) and acclimatized for 11 days at the Institut de Recerca i Tecnologia Agroalimentàries (IRTA, La Ràpita, Spain) facilities. The photoperiod followed natural changes (11:24 to 10:29 h of daylight), according to the course of the trial (October-February), and temperature was maintained at 21.9 ± 0.85 °C. During the 18-week trial, fish were fed ad libitum the corresponding diet twice a day (at 8 a.m. and 14 p.m.). Triplicate tanks were used for all experimental conditions, except for the S group where they were in duplicate. The present experiment is part of a more complex study from which data on growth performance have been already published [35].2.3. Sampling Procedures and Samples PreparationsAt the end of the growth trial, 3 fish per tank were anaesthetized (MS-222, Sigma, Madrid, Spain), measured, weighed, and sacrificed by severing the spinal cord 24 h post feeding (n = 9 from triplicate groups and n = 6 from duplicate one). The Ethics and Animal Care Committee of the University of Barcelona approved all procedures following the European Union assigned principles and legislation (permit number DAAM 8982).Samples of pyloric caeca and proximal intestine (PI) were collected, rapidly frozen in liquid nitrogen and maintained at −80 °C. Pyloric caeca and proximal intestine samples were homogenized in buffer solution (Tris-HCl, 50 mM, pH 7.5) using rapid vibration (6500 rpm; 3 × 20 s with three breaks of 20 s; 4 °C) in a Precellys Evolution® Homogenizer combined with Cryolys® as a cooling system (Bertin Technologies, Montigny-le-Bretonneux, France). Next, homogenates were centrifuged for 15 min (2400 rpm; 4 °C; Eppendorf, 5418R) and supernatants were stored at −80 °C until the analyses were performed.2.4. Oxidative Stress Markers AnalysisLPO were determined according to Uchiyama and Mihara [36]. Briefly, homogenized samples were thawed on ice and mixed with HCl 0.024 N and thiobarbituric acid 0.06 M (TBA) solution at pH 7.0. The mixture was heated to 95 °C for 10 min and then kept in ice and darkness for 5 min. Next, 300 µL of cold butanol were added and samples centrifuged for 10 min at 2000 rpm and 4 °C (Eppendorf, 5418R). The concentration of malondialdehyde (MDA) was recorded by fluorescence using a Tecan infinite 200 spectrofluorometer (Tecan Trading AG, Männedorf, Switzerland) with a 515/548 nm (excitation/emission) filter. A calibration curve with MDA in the range of 0–10 µM MDA was used to calculate the MDA concentration and the results were expressed as nmol MDA per mg of protein.The ferricytochrome C method was used to determine total SOD (EC 1.15.1.1) activity according to Mccord and Fridovich [37], with some modifications, with xanthine/xanthine oxidase as the source of superoxide radicals. Homogenized samples were thawed on ice and then mixed with xanthine oxidase (0.5 IU·mL−1) and 200 µL of the reaction buffer (50 mM potassium phosphate buffer at pH 7.8, 0.1 mM EDTA, 0.095 mM cytochrome C, and 0.015 mM xanthine). One unit of activity was defined as the amount of enzyme necessary to produce 50% inhibition of ferricytochrome C reduction rate and normalized by mg of protein. CAT (EC 1.11.1.6) activity was determined according to the method described by Aebi [38], with some modifications. Homogenized samples were thawed on ice and mixed with the reaction buffer containing: 50 mM potassium phosphate buffer at pH 7.0 and, 10 mM H2O2 freshly added and the disappearance of H2O2 was measured at 240 nm. For GPx (EC 1.11.1.9) activity determination, samples were reacted with a 50 mM potassium phosphate buffer (pH 7.2), 1.33 mM EDTA, 2.66 mM sodium azide, 40 mM GSH, 4 U·mL−1 GR, 2 mM NADPH, and 1 mM H2O2, and the rate of NADPH oxidation in the coupled reaction with glutathione reductase was determined at 340 nm according to Bell et al. [39]. GST (EC 2.5.1.18) was evaluated as previously described by Habig et al. [40] with some modifications. The assay determined the formation of an adduct between the oxidant agent 1-chloro-2,4-dinitrobenzene (CDNB) and GSH as an increase in DO at 340 nm. The reaction buffer contains: 100 mM potassium phosphate buffer (pH 6.5), 0.1% Triton X-100, 50 mM GSH and 40 mM CNDB. GR (EC 1.6.4.2) activity was determined according to Carlberg and Mannervik [41] by measuring the NADPH consumption rate. Homogenized samples were thawed on ice and reacted with a 0.1 mM potassium phosphate buffer (pH 7.5), EDTA 1 mM, NADPH 0.66 mM, and GSSG 3.25 mM and NADPH oxidation was measured at 340 nm. All substrates, reagents, coenzymes, and purified enzymes were from Sigma and Bio-Rad Laboratories, Inc. All enzymatic analyses were performed at 25 °C ± 0.5 °C using a Tecan M200 spectrophotometer (Tecan Trading AG, Männedorf, Switzerland). CAT, GPx, GR, and GST enzymatic activities are shown as U per mg of protein, where one unit is defined as the amount of enzyme required to transform 1 μmol of the substrate per minute, under the assay conditions.The protein concentration in homogenates was determined by the Bradford method [42] using bovine serum albumin as a standard.2.5. RNA Extraction and cDNA SynthesisTotal RNA extraction was performed from 30 mg of pyloric caeca or proximal intestine tissue samples in 1 mL TRIzol® reagent solution (Applied Biosystems, Madrid, Spain) following the manufacturer’s instructions. RNA concentration and purity were determined with a Nanodrop 2000 (Thermo Scientific, Alcobendas, Spain). RNA integrity was checked with a 1% agarose gel stained with SYBR-Safe DNA gel stain (Life Technologies, Alcobendas, Spain). To eliminate all genomic DNA, 1 µg of total RNA were treated with DNase I (Invitrogen, Alcobendas, Spain) following the manufacturer’s recommendations before cDNA synthesis. Finally, reverse transcription was performed using the Transcriptor First Strand cDNA synthesis kit (Roche, Sant Cugat del Vallès, Spain) following the manufacturer’s instructions, using anchored-oligo(dT)15 and random hexamer primers.2.6. Real-Time Quantitave-PCR (qPCR)Gene expression (mRNA) analyses were performed by qPCR in a CFX384 real-time system (Bio-Rad, El Prat de Llobregat, Spain), according to the requirements of the MIQE guidelines [43]. The antioxidant genes examined in pyloric caeca and proximal intestine, all previously validated for gilthead sea bream [44,45,46] were the following: superoxide dismutase 1 and 2, sod1 and sod2, respectively; catalase, cat, GPx mitochondrial, gpx1, and cytosolic, gpx4, gr and gst; and three reference genes (beta actin, β-actin, elongation factor 1 alpha, ef1α, and ribosomal protein S18, rps18).The analyses were performed in triplicate using 2.5 µL of iTaq Universal SYBR Green Supermix (Bio-Rad, El Prat de Llobregat, Spain), 250 nM of forward and reverse primers (Table 3) and 1 µL of diluted cDNA for each sample in a final volume of 5 µL. The reactions consisted of an initial denaturation step of 3 min at 95 °C, 40 cycles of 10 s at 95 °C, 30 s at 60 °C, followed by an amplicon dissociation analysis from 55 to 95 °C at 0.5 °C increase each 30 s [35]. Prior to the analyses, a dilution curve with a pool of samples was run to determine the appropriate cDNA dilution for each gene, as well as confirm the specificity of the reaction, and the absence of primer-dimers. The expression levels of each gene were calculated by the Pfaffl method [47], relative to the gene expression or geometric mean expression of the most stable reference gens analysed depending on the tissue (ef1α for pyloric caeca and ef1α and rps18 for proximal intestine), using the Bio-Rad CFX Manager 3.1 software.2.7. Statistical AnalysesThe tanks were used as biological replicates in the analysis of biometric parameters [35], and the individual fish for the antioxidant enzyme activities and gene expression analyses. Prior to individual fish analysis, tank effect was checked for each parameter, but significant differences were not observed. All data were tested for normality by the Shapiro–Wilk test and homogeneity of variances by Levene’s test. If normality was met experimental values were compared using a one-way analysis of variance (ANOVA), and differences among means were tested for significance using a post hoc Tukey’s multiple range test. When the test for normality failed, the Kruskal–Wallis non-parametric test was used followed by Mann–Whitney U test. The significance level was set at p < 0.05. The software used for statistical analysis was SPSS (IBM-SPSS Statistics v.25.0, SPSS Inc., Chicago, IL, USA) and the one used for graphic representation was GraphPad 7.00 (GraphPad Software Inc. San Diego, CA, USA).3. Results3.1. Lipid PeroxidationAfter 18 weeks of growth, levels of LPO (Table 4) were measured 24 h post feeding in pyloric caeca and proximal intestine of gilthead sea bream fed the experimental diets. In all experimental groups, LPO was higher in proximal intestine that in pyloric caeca; showing SLP group the LPO lowest levels in pyloric caeca and fish fed SLRP in proximal intestine.Sea bream fed SL diet showed the highest levels of LPO in both intestinal regions. In this group, an increase in LPO by 98% in pyloric caeca and 71% in proximal intestine was found compared to fish fed S diet. However, the SL group presented a high individual variability, as it can be inferred from the high standard error of the mean. Thus, analyzing the results for this group more deeply, we found two very different subgroups of fish (Table 4). The first one (n = 6), called L-SL, presented a similar LPO amount and profile as the rest of the experimental treatments, with higher LPO levels in proximal intestine versus pyloric caeca. However, the second subgroup (n = 3, one fish per tank), called H-SL, showed towering levels of LPO in both intestinal regions compared to the rest of the groups. Since these differences suggested an unequal dietary adaptation inside this group of fish, the subgroups were considered from now on along the results section.The addition of rapeseed or palm oils to the diets reduced LPO levels in both intestinal regions in comparison to SL-fed fish, being more evident in pyloric caeca (more than 95%) than in proximal intestine, where the reduction was associated mainly by the inclusion of palm in the diet (more than 75%) (Table 4). 3.2. Antioxidant Enzyme ActivityPyloric caeca and proximal intestine antioxidant enzyme activities are shown in Table 5. Regarding SOD activity, no differences between intestinal segments were found, except for SLRP-fed fish that showed higher activity in pyloric caeca versus proximal intestine. Conversely, CAT, GST, and GR activities were higher in proximal intestine for all experimental conditions, although these differences disappeared in H-SL fish group where the enzyme activities showed a rise in pyloric caeca (Table 5). Finally, GPx activity presented a different regionalization pattern depending on the dietary treatment, being significantly higher in pyloric caeca from S and H-SL groups and in proximal intestine from SLR, SLP and SLRP fish (Table 5). Furthermore, when comparing the different enzymatic activities, we found that CAT activity was 102 higher than GST and 103 times higher than GPx at both intestinal regions (Table 5). These data on antioxidant activities were scaled and placed in a radial chart to better visualize the differences between dietary treatments (Figure 1). In pyloric caeca (Figure 1A, Table 5), the highest antioxidant activities were found in fish fed H-SL diet, followed by L-SL group, according to their unequal dietary adaptation. Instead, when incorporating palm oil to the blend of soybean and linseed oils (SLP group), the activity levels of SOD and CAT were significantly lower and different from the other experimental conditions, except for the S group (Figure 1A, Table 5). Conversely, the incorporation of rapeseed oil to the soybean and linseed oils blend induced CAT activity and showed intermediate levels of activity for SOD in pyloric caeca, while SLRP-fed fish presented the contrary, an increased SOD and intermediate CAT activities. Moreover, fish fed diets containing more than 50% of soybean oil (S and SL groups) presented significantly higher GPx activity than those fed with lower amounts of that oil (SLP, SLR, and SLRP), suggesting that soybean oil could trigger GPx activity (Figure 1A, Table 5). All groups showed similar GST activity (mean 1.35 ± 0.07 U × mg prot−1), except for H-SL and L-SL groups, that was higher. GR activity in pyloric caeca also presented the highest activity levels in H-SL fish (Figure 1A, Table 5).In the proximal intestine, and similarly to what was found in the pyloric caeca, dietary linseed oil inclusion increased SOD activity by 62.5% in L-SL and 99.1% in H-SL fish versus fish fed S diet, whereas the inclusion of palm oil in the diet (SLP dietary group) downregulated the activity up to levels similar to those found in the S group (Figure 1B, Table 5). This significant downregulation of SOD activity was not seen in the SLR group. CAT activity of proximal intestine was similar for all experimental groups except SLRP, which showed the lowest values for CAT as well as SOD activities, and the highest values for GST. In this intestinal region, contrary to what happened in pyloric caeca, no changes were found in GPx activity between dietary treatments. Interestingly, in proximal intestine, a clear differential pattern in GR activity was detected when comparing L-SL and H-SL fish, being in the latter significantly lower than in the former (Figure 1B, Table 5). H-SL animals showed GR activities similar to those found in the SLP group, representing a decrease of the activity with respect to the rest of experimental conditions (Table 5).3.3. Gene ExpressionGene expression of antioxidant enzymes at intestinal level 24 h post feeding was slightly modulated by the dietary treatments (Table 6). In this sense, in pyloric caeca, the expression of sod2, gpx1, and gpx4 was upregulated in all dietary treatments compared to SLP group that showed the lowest values. Moreover, in proximal intestine only gpx4 was modified showing the S group the highest value and the L-SL and the SLRP the lowest (Table 6).3.4. Antioxidant RatiosSince the activity of the antioxidant enzymes investigated can be affected by each other, either because the products of the activity of one enzyme are the reagents of the other or because they use the same substrate, the relationships between them were also calculated (Table 7). For all experimental groups, the (CAT+GPx)/SOD ratio was higher in the proximal intestine than in the pyloric caeca; in contrast, the ratio (GST+GPx/GR) was higher in the pyloric caeca, except for the H-SL group that showed similar activity ratios in both intestinal segments (Table 7). The dietary linseed oil inclusion caused a significant decrease in the (CAT+GPx)/SOD ratio in both intestinal segments of L-SL group versus diet S-fed fish; however, in H-SL animals, this decrease was only observed in proximal intestine. Moreover, in this intestinal segment, SLR animals showed a lower ratio than S, SLP and SLRP fish (Table 7). Regarding the CAT/GPx ratio, lower values were found in pyloric caeca from S and SL groups. Dietary palm oil inclusion (SLP group) contributed to increase this activity by 61%, whereas the SLR blend triggered a 254% increase in this ration in comparison to that observed in S, L-SL and H-SL fish. SLRP sea bream had intermediate values for this enzymatic ratio, between those observed for SLR and SLP animals, probably due to the lower palm oil content of this diet versus the SLP diet.Instead, in proximal intestine, lower CAT/GPx ratios were found in sea bream fed diets containing rapeseed oil (SLR and SLRP) (Table 7). The highest levels in the (GST+GPx)/GR ratio for both intestinal segments were found in SLP-fed sea bream. In pyloric caeca, high levels of this ration were also measured in L-SL and S groups, being significantly lower for H-SL fish. In proximal intestine, the lowest values for the (GST+GPx)/GR ratio were observed in S and L-SL groups.4. DiscussionROS are generated by aerobic cellular metabolism in small quantities and most of the time, the intestine responds adequately against oxidative stress [48]. Nevertheless, external factors such as nutrition stress by both high-fat or high-carbohydrate diets could exacerbate the ROS production inducing tissue oxidative damage [20], negatively affecting intestinal functionality and health [48]. LPO is the hallmark of oxidative stress, and at intestine level is closely related with dietary lipid composition [49,50], since it depends on the amount of lipid and its length and unsaturation degree [19]. Accordingly, the inclusion of VO in fish feed formulation contributed to the reduction of intestinal LPO in comparison to sea bream fed FO diets [22]; on the contrary, plant protein diets contributed in sea bass to its increase in comparison with fish fed a fishmeal diet [51]. In the present study, LPO levels in SL group versus fish fed S diet were exacerbated, in accordance with previous results found by Magalhaes et al. [30] in sea bream fed diets supplemented with EPA and DHA, pointing to an inversely relationship of LPO levels with n-6/n-3 ratio. Nevertheless, this effect was not found in the rest of the groups fed diets including linseed oil, SLR, SLP and SLRP, which contained similar n-3, but lower levels of n-6, suggesting an effect on LPO of both the n-6/n-3 ratio and the n-6 content in the diet. In addition, the present results also showed an effect of dietary UFA/PUFA ratio on LPO in fish fed diets containing similar amounts of n-3 fatty acids. In this sense, fish fed the SLR diet, with a similar UFA/SFA ratio to that present in the SL-fed group, showed higher levels of LPO in pyloric caeca and proximal intestine than those found in SLP-fed sea bream with the lower UFA/SFA ratio, whereas fish fed SLRP diet showed globally intermediate LPO levels according to their halfway ratio.Furthermore, the SL group of fish presented the highest LPO levels in both intestinal segments, but also showed high individual variability, since not all the fish fed this diet had high LPO levels, especially in the pyloric caeca region. This led to consider dividing this group into two subgroups with a different LPO profile and antioxidant activity in the pyloric region, which suggested an unequal adaptation to the administered diet. On one side, H-SL group presented exacerbated LPO levels and high antioxidant enzyme activities and on the other side, L-SL group, showed low LPO and moderate levels of SOD, CAT, and GR activities. Moreover, GST activity that is associated with a xenobiotic pathway was only regulated in the SL subgroup that presented the highest LPO, pointing to a higher reaction to toxics or allergens of the H-SL fish, mainly in pyloric caeca. In this regard, several studies showed a higher susceptibility to peroxidation of intestine versus liver due to its high turnover and dietary allergens or toxics exposure [52,53,54].In addition, intestinal antioxidant enzyme activities also showed a modulation related to dietary lipid composition in an attempt to cope with the LPO generated to maintain a healthy oxidative status. In this sense, in the present study, an increase in SOD, CAT, GPx, and GR activities was found in pyloric caeca in the H-SL fish when compared with the other experimental groups and the L-SL fish. This suggests that this dietary VO blend triggers superoxide anion production in response to the high LPO generated in the H-SL group, stimulating SOD and in turn the enzymes involved in the second line of defence (CAT and GPx) as well as GR. Accordingly, antioxidant enzymes’ gene expression was generally highest in the H-SL group. On the other hand, pyloric caeca from SLP fish showed the lowest peroxidation levels, in agreement with the low gene expression and activity of antioxidant enzymes. The differences in the profile of antioxidants’ activities found between intestinal segments in SLRP-fed sea bream versus the other experimental groups could also be related to the different LPO levels found, which were higher in pyloric caeca than in proximal intestine. In this sense, low SOD and CAT activities were detected in proximal intestine from SLRP-fed sea bream, whereas their activities were high and intermediate in pyloric caeca.As previously mentioned, the mixture of soybean and linseed oils (SL diet) is not the most appropriate for sea bream since LPO can be triggered, which did not occur when palm and/or rapeseed oils were included in the mixture of VO. Additionally, the dietary fatty acids profile affects, in addition to LPO levels, the activity of antioxidant enzymes. Specifically, SOD activity, the first line of defence against oxidative stress, was highly modulated by the dietary lipid profile, with the S, SL, SLP, and SLR groups showing a similar change in pattern and magnitude of activity in both intestinal segments. In this sense, dietary inclusion of rapeseed oil and/or especially palm oil seems to promote the reduction of SOD activity whose increase could be related with the dietary addition of linseed oil. CAT and GPx constitute the second line of defence, removing H2O2 produced by SOD [20]. CAT acts in high oxidative stress conditions [20,55] and its activity was low only in pyloric caeca of fish fed diets with palm oil inclusion (SLP and SLRP) and in proximal intestine in SLRP-fed sea bream. Instead, GPx copes with oxidative stress in basal conditions [20,55]. Regarding this antioxidant enzyme, low levels of activity were found in pyloric caeca from sea bream fed with the SLP diet and especially in those fed with SLR and SLRP diets, which were different from those found in S and SL sea bream. These results pointed to a downregulation in GPx activity by the presence of rapeseed oil and an upregulation by soybean oil levels above 50% in pyloric caeca, whereas no modulation was found in proximal intestine.Among the studies conducted at intestinal level on oxidative stress, this is the first, to our knowledge, in which the pyloric caeca and proximal intestine are analysed at the same time. The present results showed that LPO was higher in proximal intestine than in pyloric caeca, except in SL fish, and antioxidant activities levels were also different between these intestinal regions. Thus, the anatomical location of pyloric caeca, anterior to proximal intestine, and their specific functions could in part explain the differences found. Pancreatic and bile juices are released in pyloric caeca and in then alkaline digestion starts [56]. Moreover, this region is subjected to a higher enterocyte renewal rate than proximal intestine [57] and presents a characteristic mixing and retrograde contractile activity [58]. These characteristics contribute to slowing down the intestinal transit favouring the chyme to be retained in the pyloric caeca for a short time until it gradually passes into the proximal intestine [59,60], where digestion will finish and nutrient absorption will take place. An intestinal regionalization has been previously described in studies related to digestive enzyme activities, diet, and temperature [61,62,63,64,65,66,67], showing how pyloric caeca and proximal intestine functions were differently modulated. In the present study, a clear dietary modulation of most of the antioxidant enzymes studied in pyloric caeca and only SOD in proximal intestine was observed. This idea was also reinforced by the low gene expression changes found. Moreover, in proximal intestine, the antioxidant capacity of CAT, GPx, GST, and GSH recycling by GR were generally higher than in pyloric caeca, regardless of the diet administered. Thus, data suggest the attempt to preserve the epithelium of proximal intestine, which has a lower rate of cell renewal, mature enterocytes [57,68] and has greater digestive and absorptive capacities than those found in the pyloric caeca region [56]—even going so far as to affect the proper function of the region of the pyloric caeca in the case of H-SL animals. Thus, this group reached very high levels of LPO in pyloric caeca that would lead to a higher cellular renewal rate, which would prevent proper cell maturation for the performance of digestive and absorptive functions, possibly aiming to maintain as low as possible peroxidation levels in the proximal intestine. This idea agrees with other studies where the effects of antinutritional factors and enteritis in Atlantic salmon were analysed [69,70,71]. In fact, antioxidant ratios calculated in the present study also support this possible regulation since the (CAT+GPx)/SOD ratio was higher in proximal intestine; whereas, the (GST+GPx)/GR ratio, a pathway related with xenobiotics and GSH reduction, was higher in PC, being both processes directed to maintaining the functionality of proximal intestine.GR is involved in GSSH recycling to GSH. The latter is a non-enzymatic antioxidant that directly quenches ROS or indirectly acts as a substrate for both GPx and GST [20], being a mechanism extensively used by the intestine due to their susceptibility to oxidation and the turnover rate of enterocytes [53]. In this context, in H-SL fish, higher GR activity was found in pyloric caeca than in proximal intestine according to the high LPO levels that were accompanied by a higher activity of GPx and GST. In this group, the difference in GR activity between both segments could be due to a different use of GSH to remove ROS. Instead, the low GR activity found in fish fed the SLP diet may be due to the low LPO levels found in both pyloric caeca and proximal intestine.Considering the possible benefits and disadvantages of the diets, it is well known that fish fed a diet containing only soybean oil present lipid accumulation in enterocytes that could led to an impairment in the digestion process [6,15,16,17]. Our results showed that the SL diet also had also negative effects in the gut since it induced, in some animals, exacerbated levels of LPO in both intestinal regions; revealing their inability to adapt to the diet. Moreover, fish fed with the SLR diet presented higher oxidative damage in proximal intestine versus SLP and SLRP fed sea bream. Since the proximal intestine area is principally involved in the processes of completion of digestion and nutrient absorption, the SLR diet would not be the most recommended for this species. Regarding the last two groups, SLP and SLRP, the main differences were that in the latter, the highest oxidative damage was found in pyloric caeca, a region involved in lipid absorption [56], and presented an increase mainly in SOD activity which would lead to a greater energy expenditure to cope with oxidative stress. On the contrary, the SLP diet would be the optimal one considering the intestinal redox status, since SLP-fed fish presented low and moderate activities in all the antioxidants enzymes studied, but that allowed them to also keep low LPO levels, therefore suggesting a lower energy expenditure need to cope with dietary induced LPO at intestinal level.5. ConclusionsIn summary, this study shows for the first time the interest of the pyloric caeca region in coping with the oxidative stress generated by the inclusion of VO in fish diets. Moreover, it also highlights the importance of the proportion of different VO in these mixtures, being in this case the inclusion of linseed oil a factor to consider in dietary formulation since it can negatively affect the oxidative status of the animal at intestinal level. To conclude, among all the diets evaluated, the SLP diet induced in sea bream the best intestinal oxidative status, with low LPO levels and antioxidant enzyme activities.

animals : an open access journal from mdpi

10.3390/ani11061568

PMC8227775

With the continuous improvement in the progress of poultry industry, a better understanding of the avian immune system is necessary. A prolonged holding period (36–72 h), along with a delay in access to feed and/or water post-hatching, has a negative influence on performance, intestinal histomorphology, and the immune system development of chicks. Therefore, the present study aimed to investigate the effect of early feeding with different diet composition or delayed feeding during the first 72 h of chick’s life on the expression of immunity-related genes and histomorphology of digestive and lymphoid organs of layer-type chicks. Early nutrition post-hatching had no negative effect on the development of the lymphoid and digestive organs in chicks. Histomorphological examination revealed an increase in cortex and cortex:medulla of thymus and bursa in the early fed groups compared to the fasted ones, with resultant impacts on the primary lymphoid organs. Higher germinal center areas and white pulp of the spleen were recorded in the early fed chicks, implying augmented proliferation and maturation of B cells in the secondary lymphoid organs. In the liver, a strong positive reaction to Best’s carmine stain in the early fed groups, indicating that the liver of these chicks had numerous glycogen granules in the cytoplasm of hepatocytes. The expression levels of splenic-immunity related genes were up-regulated in most of the early fed chicks at 14 day of age. Our findings suggested that early feeding post-hatch can modify the splenic-immunity related genes and modulate the histomorphology of the digestive (liver and proventriculus) and lymphoid organ in layer-type chicks during the early life post-hatching.

Early feeding post-hatching (EFPH) can impact the immune response and modify the immunity-related gene expression. Therefore, we aimed to examine the effects of EFPH with different diets composition versus fasting during the first 72 h of chick’s life on the histomorphological structures of the liver, proventriculus, central and peripheral lymphoid organs, and immunity-related genes in layer-type chicks during the brooding period. A total of 400 chicks were randomly allotted into 4 groups with 4 replicates each. The experimental groups during the first 72 h of life were: feed and water deprivation (control, T1), feeding a starter layer diet (20% CP and 11.84 MJ/kg ME, T2), feeding a starter layer diet contained 3% molasses in its composition (20% CP and 11.81 MJ/kg ME; T3), and feeding a starter broiler diet (23% CP and 12.68 MJ/kg ME, T4). After the first 72 h of chick’s life, all chicks were fed ad libitum the T2 diet. EFPH had no negative effect on the development of the lymphoid or digestive organs in chicks. Greater relative weights of the spleen and bursa of Fabricius (p < 0.05) were observed in the early fed chicks compared to control at day 14 of age. Histomorphological examination revealed an increase (p < 0.05) in thymus cortex and cortex:medulla in the T3 and T4 groups compared to the fasted ones at day 28 of age. Pelicae height, follicular width, cortex, and cortex:medulla of bursa were improved (p < 0.01) in the fed groups compared to fasted chicks, with resultant influences on the primary lymphoid organs. Compared to control, higher germinal center areas and white pulp of the spleen (p < 0.05) were recorded in the early fed chicks, implying augmented proliferation and maturation of B cells in the secondary lymphoid organs. In the liver, a strong positive reaction to Best’s carmine stain in the early fed groups, indicating that the liver of these chicks had numerous glycogen granules or greater glycogen density in the cytoplasm of hepatocytes. There was a significant enhancement in the proventriculus mucosal and gland thickness, as well as fold height (p < 0.05) in the early fed chicks. The expression levels of splenic Toll-like receptor 2, interleukin 4, tumor necrosis factor α, and interferon gamma were up-regulated (p < 0.01) in most of the early fed chicks (T2, T3, and T4) compared to fasted ones at 14 day of age. In conclusion, EFPH could modify the splenic-immunity related genes and modulate the histomorphology of the digestive (liver and proventriculus) and lymphoid organs in layer-type chicks during the brooding period.

1. IntroductionDuring the last two decades, a significant enhancement has been made on the functional characteristics of layer- and meat-type poultry through genetic selection; accordingly rising the prevalence of early chick mortality, possibly due to immunosuppression and a reduced resistance to infections. With the continuous improvement in the progress of poultry industry, a better understanding of the avian immune system is necessary. The immune system development of the newly hatched chicks is influenced by numerous factors, one of the most critical factors is the early feeding post-hatch (EFPH) [1,2,3,4]. Newly hatched chicks are commonly pulled out when the majority of them have hatched, causing chicks to be deprived of feed or water for 24–72 h. In addition, other common management practices such as sex determination, counting, vaccination, and transportation are accountable for the delayed feeding post-hatch (PH) [4,5].Prolonged holding period (36–72 h) along with a delay in access to feed or water have negative influences on chicks because of dehydration and energy depletion [4,6]. Feed deprivation (FD) PH could adversely impact the growth performance, intestinal histomorphology, the rate of nutrient absorption, and the immune response, both in broiler- and layer-type chicks [3,4,7,8,9,10]. Metabolic changes caused by early FD could possibly diminish the lymphoid organs growth, and further increase the susceptibility to diseases and compromise the immune response [2,4,6], resulting in a greater production cost of birds. The intestinal development, in particular villi length and number of enterocytes, as well as the lymphoid organs growth occur more rapidly after the first 7 days PH with intensive early development up to 21–35 days PH in broilers [11,12] and between 14 and 42 days PH in layers [9], accordingly, chicks during the first week PH have an insufficient immune response due to the insufficiency of the immune systems development [1,4,9]. The immune system of newly hatched birds (broilers and layers), in particular the mucosal immune system, necessitates feed intake for prompt development and maturation [1,6,9,10]. In layers, Simon et al. [9] reported that early-fed chicks had higher live weight from day 3 through day 35 PH, greater relative weight of spleen (until day 49 PH) and bursa of Fabricius (day 6 and day 9 PH), and an earlier onset of ileal IgA expression (day 9 vs. day 14 PH) than FD chicks. The application of nutritional strategies can efficiently stimulate the early development and maturity of the immune system and boost the innate immunity of chicks [9,11].The yolk sac of the hatchlings has high protein and fat contents but little carbohydrate content [13]. Due to the fat not playing a part in the metabolic synthesis of glucose, gluconeogenesis is mostly originated from body reserves and yolk sac protein [14]. Newly hatched chicks have low efficiency in utilizing nutrients, and the inclusion of highly digestible energy sources such as simple glucose, sucrose, or glucose-based materials in their early diets is believed to improve the performance of chicks [5]. These materials are considered as a good source of glucose and are efficiently utilized by newly-hatched chicks for their abrupt glucose demands PH [14]. Moreover, partial substitution of starch with simple sugars (4–8%) in pre-starter diets potentially enhanced the chick’s performance and can be applied to mitigate the adverse effects of delayed access to feed PH [15]. Sugarcane molasses is cheap, easily available in Egypt, and contain approximately 56% total sugars (sucrose, fructose, and glucose). Molasses can be serving as a source of energy in the poultry feeds and was used to improve the palatability and feed intake of chicks with a dietary recommended level not more than 5–7% in the mash chicks’ diets [16]. On the other hand, it was documented that increasing dietary protein (24.6% vs. 21.4%) and digestible amino acids (114% vs. 100% of cobb recommendation) in the pre-starter diets (day 1 to day 10 of age) greatly enhanced the growth performance with significant changes in the intestinal weight and histomorphology of broilers [17]. Early nutrition on easily digestible carbohydrates, protein or amino acids either in ovo or in the chicks’ early diets were applied to decrease the demand for gluconeogenesis, meet the needs for rapid growth, enhance the intestinal development, and augment the immune response during the early PH period [3,5,7,17,18,19,20].EFPH of chicks is a stimulus to the development and functioning of the immune system, the primary and secondary immune organs, during the early stages of life [2,4,8] and the health status of chicks on long-term [20]. After activation, the naive T cells is differentiated either into T helper 1 (Th1) cells (cellular immunity and pro-inflammatory function), or Th2 effector cells release cytokines that enhance humoral immunity [21]. The development of these cells begins during the embryogenic stage and continues PH. Recently, it has been reported by Song et al. [22] that the lowest point of the immunity in broiler chickens appeared within day 6 to 13 of age, whereas the highest levels of the immune status occurred within 30 and 34 days of age, and the key indicators of this pattern were blood and splenic cytokines. In layer-type chickens, the intestinal cytokine expression levels were enhanced between day 14 and 42 PH compared to day 3–4 PH [9,23]. In both breeds (broiler and layer chicks), EFPH has been recorded to increase the weights of lymphoid organs and enhanced the immune response PH when compared with the FD chicks [1,4,6,8,9,18]. Furthermore, the impacts of feed and water access versus fasting in layer-type chicks are mostly studied [9,10], ignoring the effects of EFPH or FD on the histomorphology of digestive (liver and gizzard) and lymphoid organs as well as the effect of diet composition [9,10].Very few studies have been performed to assess the effect of early feeding or feed and water deprivation PH (FWDPH) on the growth performance and intestinal morphology in both broiler and layer-type chicks but to our knowledge little information is existing in the literature about the effect of EFPH with different diets composition versus FWDPH on the expression of immunity-related genes and histomorphology of digestive (proventriculus and the liver) and lymphoid organs (bursa of Fabricius, spleen, and thymus gland) of layer-type chicks. Therefore, the present study aimed to determine the effect of EFPH with different diet composition or DF during the first 72 h of chick’s life on the expression of immunity-related genes and histomorphology of digestive and lymphoid organs during the brooding period of layer-type chicks. The hypothesis of the present study was that delayed feeding PH will influence the early development of digestive and lymphoid organs and alter the expression of immunity-related genes of chicks.2. Materials and Methods2.1. Ethical ApprovalThe procedures adopted in this trial were performed according to the guidelines as approved by the Animal Ethics Committee of the Faculty of Agriculture, University of Menoufia, Menoufia, Egypt (No. 4/2017).2.2. Hatching Eggs, Experimental Treatments and ManagementThe experimental design was previously published in Abou-Elnaga and Selim [7]. Eggs from Norfa layer flock (White Leghorn × Fayoumi × White Baladi) weighing between 40 and 45 g were collected and incubated in a hatchery (Poultry Research Farm, Faculty of Agriculture, Menoufia University, Menoufia, Egypt). On day 18 of incubation, eggs were moved to the hatchery. A group of 400 hatched chicks (200 males + 200 females) was used in the current trial. Hatching trays were divided with partitions and 4 different treatments were applied using a completely randomized design. Each treatment had 4 replicates with 25 chicks per replicate. Treatments were fasting (control, T1; no feed or water), feeding a layer starter diet (20% CP and 11.84 MJ/kg ME; T2), feeding a starter diet containing 3% molasses (20% CP and 11.81 MJ/kg ME; T3), and feeding a starter broiler diet (23% CP and 12.68 MJ/kg ME, T4) during the first 72 h post-hatch. Immediately, all chicks after fasting or feeding the experimental diets for the first 72 h of their life were fed layer starter mash diet (T2) ad libitum for a period of 42 days. All chicks were moved from the hatchery after 72 h PH and kept in the floor pens. The ingredients and the proximate chemical analysis of the diets used in the current trial were published previously in Abou-Elnaga and Selim [7]. All chicks were reared under the same standard management practices and routine vaccination during the experimental period.2.3. Lymphoid and Digestive Organs HistomorphologyAt 14 and 28 days PH, 8 chicks (both the sex) from each group (2 per replicate), within the average live weight, at each time point were killed by cervical dislocation. The weight of liver [7], proventriculus, bursa, spleen, and thymus were recorded and expressed as per cent of live weight. The routine histological techniques were performed according to Bancroft and Stevens [24]. Tissue samples were rapidly placed in 10% neutral buffered formalin for approximately 48 h, then tissue processing occurred firstly by transferring samples to the dehydration process through ascending grades of alcohols from 70% to 100%, following by the clearing process through passing in a clearing agent as methyl benzoate then embedding in paraffin wax at melting point 56 °C in a hot air oven. The paraffin tissue blocks were cut by using a rotatory microtome. All samples were sectioned at 4–6-μm thickness. Sections were mounted on a clean glass slides and stained with Hematoxylin and Eosin (H and E) for the general histological examination and Best’s carmine stain for detection the distribution of glycogen in the liver tissue [24]. After glass slides being dried, cross sections from tissue samples (proventriculus, liver, bursa of Fabricius, thymus, and spleen) were examined using a light microscope. The photomicrographs from the selected specimens were taken using Leica digital camera connected with bilocular microscope for better illustration of the findings. The Morphometric measurements were performed according to Madej et al. [12], Sikandar et al. [25], and Sayrafi and Aghagolzadeh [26]. In proventriculus, height of the proventriculus fold, thickness of mucosa, diameter of compound gland, and Pelicae height were determined. In bursa of Fabricius, follicular width and thickness of follicular cortex and medulla were calculated. In thymus, thickness of cortex and medulla were measured. In spleen, the thickness of area of white pulp and red pulp were performed. Splenic germinal center areas were measured as described by Madej et al. [12] and Sikandar et al. [25]. To estimate the hepatic glycogen density in the cytoplasm, a scale was performed ranging from 1 (low glycogen density) to 4 (very high glycogen density), depending on the staining of glycogen granules with the Best’s carmine stain. Four histologists (masked to the groups) issued the scores.2.4. Immunity-Related Gene ExpressionEight chicks from each treatment group at each time point (2 chicks per replicate) were selected and killed with cervical dislocation for the gene expression analysis. Splenic samples were dissected carefully and frozen at −80 °C until further analysis. Splenic immunity was performed by a quantitative measurement of the mRNA expression of interleukin 4 (IL4), interferon gamma (IFNγ), tumor necrosis factor α (TNFα), and Toll-like receptor 2 (TLR2) on day 14 and 28 of age. The total RNA extraction, preparation, and cycling conditions for real-time polymerase chain reaction (PCR) were determined as described previously by Bhanja et al. [3]. The oligonucleotide sequences of the primers used in the current study were listed in Table 1. Amplification curves and cycle threshold (CT) values were measured using Stratagene MX3005P software (Agilent Technologies, Inc., Santa Clara, CA, USA). The reference housekeeping gene was 28S rRNA. Relative mRNA expression was calculated using the 2(−ΔΔCt) method, and the results were recorded as a fold change [27].2.5. Statistical AnalysisNormal distribution of the data was determined with the Kolmogorov–Smirnov test. The obtained data from the effect of early nutrition with different diets and fasting on the digestive and lymphoid organs histomorphology and splenic gene expression at two time points were subjected to One-way Analysis of Variance using GLM procedures of SPSS software (SPSS, version 21, SPSS Inc., Chicago, IL, USA). Statistically different means were separated using Duncan’s multiple range test at p < 0.05.3. Results3.1. Histomorphology of Lymphoid Organs3.1.1. Thymus GlandThere was non-significant (p > 0.05) effect of early nutrition PH on the relative weight of thymus or its medulla thickness, both at day 14 and day 28 of age (Figure 1 and Table 2 and Table 3). The histomorphological structure of the thymus gland was normal in all treatments and at both time points (Figure 1). Thymus gland is composed of incompletely separated lobules. Each thymic lobule is divided into an outer dark staining cortex and an inner light staining medulla due to the increased numbers of T-lymphocytes and other cortical cells than that in medulla. Intensive growth of thymic lobules was detected up to day 28 PH (Figure 1). A significant effect of early nutrition PH (p < 0.05) was noticed for the cortex of thymus, indicated that the T3 group at day 14 of age and the T2, T3, and T4 groups at day 28 of age had the greatest cortex thickness compared to the fasted ones. The cortex to medulla ratio was higher (p < 0.05) in the T3 and T4 groups at day 28 of age than the fasted and T2 groups (Figure 1 and Table 3). However, there was a non-significant difference in the cortex to medulla ratio among the treatment groups at day 14 of age (Table 2).3.1.2. Bursa of FabriciusHistomorphological characteristics of bursa of Fabricius at day 14 and 28 PH of layer-type chicks are presented in Figure 2 and Table 2 and Table 3. The histological structures of bursa were normal in all groups and at all time points (Figure 2). Bursa of Fabricius is consisted of folded mucosa lined with pseudostratified columnar epithelium. The mucosal folds are filled with numerous polyhedral follicles extended in lamina propria and the submucosa, which is separated by connective tissue, each follicle is composed of lymphatic tissue and divided into cortex and medulla. At day 28 PH, bursa is characterized by the presence of mucoid cysts in between the bursal epithelium (Figure 2). Greater relative weight of bursa (p < 0.01) was recorded in chicks fed the T2, T3, and T4 diets at day 14 of age compared to the fasted ones (Table 2). EFPH had significant effect (p < 0.01) on the Pelicae height and follicular width of bursa, characterized by higher values in the chicks fed on the T2, T3, and T4 diets at day 28 of age. Greater cortex thickness (p < 0.01) was recorded in the T2, T3, and T4 groups than in the fasted group at day 14 of age, while, it was only higher (p < 0.001) in the T4 group at day 28 of age than other treatments. There was non-significant difference in the bursal medulla among the treatment groups. The cortex to medulla ratio was higher (p < 0.01) in the chicks fed on the T4 diet than the other treatment groups at day 28 of age.3.1.3. SpleenThere was non-significant (p > 0.05) effect of EFPH on the splenic red pulp among the treatment groups (Figure 3 and Table 2 and Table 3). Splenic relative weight was greater (p < 0.05) in the T3 and T4 chicks at day 14 of age, while, there was non-significant difference in the splenic relative weight among the treatment groups at day 28 of age (Figure 3 and Table 2 and Table 3). Germinal center became noticeable on day 14 and 28 of age, nearby the arteries and typically obviously distinguished from the surrounding structures (Figure 3). Early nutrition PH increased the splenic white pulp and germinal center area in the early fed chicks at day 14 (p < 0.05) and 28 (p < 0.001) of age compared with the fasted group (Figure 3 and Table 2 and Table 3).3.2. Liver and Proventriculus Histomorphology3.2.1. The LiverAt 14 and 28 days PH, the hepatocytes appeared pyramidal in shape (Figure 4). The histomorphological examination of the liver was normal in all groups at both time points (Figure 4). The cytoplasm was more acidophilic without cytoplasmic vacuoles in addition to aggregations of lymphocytes were mostly aggregated in three regions in the liver such as around the portal area surrounding the blood vessels, around the central veins, and between the hepatic plates. The hepatocytes were arranged in two-cell-thick hepatic plates. The hepatic sinusoids were lined by flat endothelial cells and Von Kupffer cell. With Best’s carmine stain, Figure 4 revealed that the T2, T3, and T4 treatment groups had a strong positive reaction to the stain and significantly greater glycogen density score, both at day 14 and 28 of age, characterized by more reddish dots in the cytoplasm of hepatocytes, suggesting greater hepatic glycogen density (Figure 4).3.2.2. ProventriculusHistomorphological characteristics of proventriculus at day 14 and 28 PH of layer-type chicks are shown in Figure 5 and Table 2 and Table 3. The wall of proventriculus is consisted of mucosa which thrown into grossly folds, and it is composed of columnar cells while lobules of simple tubular gland consisted from cuboidal to low columnar secretory cells located in propria-submucosa, then surrounded by muscle and serosal layer (Figure 5). Greater relative weight of proventriculus was recorded in chicks fed the T3 and T4 diets compared to the FD ones (p < 0.05). The effect of early nutrition on the proventriculus histomorphology indicated that chicks fed the T2, T3, and T4 diet recorded the highest mucosal fold height at 14 day of age compared to fasted ones (p < 0.01), while the T4 chicks had the greatest mucosal fold height at day 28 of age (p < 0.001) compared to other treatments. At day 14 of age, there was a significant increase (p < 0.001) in the proventriculus mucosal thickness in the T3 and T4 groups compared to the other groups. A significant effect of early nutrition PH was observed on the proventriculus submucosal gland thickness, indicated higher values (p < 0.05) in the T2, T3, and T4 chicks compared to fasted ones at day 14 and 28 of age.3.3. Immunity Related Gene ExpressionThe expression levels of splenic TLR2 and INFγ were up-regulated (p < 0.01) in most of the early fed chicks (T2, T3, and T4) compared to fasted ones (CON, T1) at 14 days PH (Figure 6). The mRNA expression of splenic IL4 and TNFα was up-regulated (p < 0.05) in the T3 and T4 groups compared to the T2 and fasted birds at day 14 of age (Figure 6). At 28 days PH, there was non-significant difference in the mRNA level of splenic TLR2, INFγ, IL4 or TNFα among the treatments (Figure 6). The expression levels at day 28 PH were non-significantly up-regulated (p > 0.05) in the fasted birds.4. DiscussionEarlier studies on broilers [1,2,8,18] and layers [9,10] proposed that EFPH can influence their early immune system development, especially the relative lymphoid organs weights, and the immune response. However, to our knowledge little information is existing in the literature about the effect of EFPH or FD on the histomorphology of the liver, proventriculus, and lymphoid organs in layer-type chicks. Therefore, the current study aimed to determine the effect of EFPH versus FWDPH on the early development of the lymphoid (weak immune status) and digestive organs in layer-type chicks. This is because the immune organs are matured and fully developed (strong immune status) during the grower period of layers [9,28].Thymus, bursa of Fabricius, and spleen are the main originators of the immune cells involved in the cellular and humoral immunity. The development and maturation of these organs appear more effectively in healthy birds than in diseased ones, and their growth can reveal the function and response of the immune system [1,2,6,12,25]. In the current study, EFPH increased the relative lymphoid organs weight, particularly the spleen and bursa of Fabricius, which is in agreement with previous studies in broiler- [2,6,8] and layer- [9] type chicks. In contrast, Shinde et al. [10] reported non-significant difference in the thymus, bursa, and spleen relative weights between the FD and early-fed layer-type chicks at 36 h, 7 day, and 14 day PH. Greater indexes of lymphoid organs in birds are commonly determined as indicators of the augmented proliferation of B and T lymphocytes, which implies better immunity [25]. Thymus is the main site of thymocytes, particularly T-lymphocytes, proliferation and maturation, bursa of Fabricius is essential for B-cell proliferation, and the spleen, the largest peripheral lymphoid organ, is included in the immune response [28,29]. The histomorphological structure of the thymus gland, in particular the cortex and medulla, can be affected by various stimulus [30]. The differences in the cortex and medulla have a linkage to the development, maturation, and function of the immune system [12,25,31]. In the current study, the histological structure of thymus was normal in all the experimental groups and at both time points, and in agreement with previous studies [12,25,31]. EFPH on the T3 and T4 diets increased the cortex of thymus, both at day 14 and 28 of age. In the cortex, there is an intensive proliferation and maturation of T lymphocytes, while the medulla is dominated by mature CD4+ or CD8+ cells [12]. The greater cortex in the T3 and T4 groups may suggest a faster proliferation and maturation of T-lymphocyte [32]. The noticeable increase in the cortex to medulla ratio in these early fed groups at day 28 of age revealed more mature CD4+ and CD8+ T cells migrate from the cortex to the medulla of thymus than in the fasted ones. These cells are known to recognize pathogens and can travel to the secondary lymphatic organs through the blood circulation [12]. It was reported that intensive development of thymus lobules was recorded between day 7 to day 21 PH [12]. Overall, these results suggested that EFPH with diets rich in protein and energy contents or contain easily available carbohydrates may modify the early development of the thymus in layer-type chicks. Madej et al. [12] reported that in ovo-injection of probiotic and/or synbiotic resulted in greater cortex and higher cortex to medulla ratio of thymus than non-injected group. To the best of our knowledge, no previous research is available in literature emphasizing the effect of EFPH on histomorphology of thymus to which we may compare our findings. It has been reported that the thymus gland is highly susceptible to FD, energy or amino acid deficiencies, which induce a rapid decrease in cellularity and weight [4,33]. Nutrient deprivation for long period may cause a reduction in the numbers of CD4+ and CD8+ T cells [4]. Furthermore, higher concentrations of CD4+ and CD8+ cells were observed in the early fed chicks compared to the FD ones [1].The development of bursa occurs by the formation and colonization of the medulla during the embryonic stage, while the cortex develops after hatching [34]. Post-hatching, the B cells development takes place in the appearance of gut constituents such as antigens [35]. These antigens could possibly play a key role in the proliferation and/or differentiation of B cells [12,35]. The findings of the current trial showed that EFPH on the T2, T3, or T4 diets increased the bursa cortex and follicular width PH, suggesting an enhancement in the migration of B cells from the medulla to the cortex, which leading to acceleration in the follicular cortex development in the fed groups when compared with the fasted ones. The higher follicular width of the bursa suggested an increase in the B-lymphoblasts count, and as a result the formation of B-lymphocytes for antibodies [36]. Furthermore, EFPH could enhance the humoral immunity of chicks PH owing to the availability of nutrients essential to produce antibodies [36]. Juul-Madsen and Sorensen [1], Dibner et al. [2], and Bar Shira et al. [6] proposed that EFPH provides the birds with the essential nutrients, impacts the endogenous hormones or immunomodulators levels, and modifies the gut antigen repertoire, which in turn augments the differentiation of immune cells such as B lymphocytes. Supporting evidence was obtained by Simon et al. [9] who observed that EFPH resulted in an earlier onset of IgA expression in the early-fed (day 9 PH) compared with FD (day 14 PH) layer-type chicks. They suggested an antigenic stimulation through feed intake could be contributed to the B cell development [9]. Furthermore, the microbial sequence may have been postponed because of a delay in feed intake [9,20]. Another possible explanation was that fasting can stimulate the corticosteroids secretion and other metabolic changes, which in turn can diminish the proliferation of the immune organs [2]. The splenic index is directly correlated with the immune response [37]. The spleen has two main parts; red pulp and white pulp. The red pulp is mainly consisted of red blood cells, whereas white pulp is predominantly composed of lymphocytes. Histological structure of the spleen in the present trial was normal in all groups at both time points and was in accordance with the earlier studies [12,25]. In the spleen, the germinal center area is well-known as an essential histological structure of the spleen stimulation [12,25]. It was recorded that the splenic germinal center area was noticed with clear histomorphological structure at day 14, 21, and 35 of age [11,12]. In the current study, an increased white pulp and germinal center area of the spleen along with a greater spleen index in the early fed groups when compared with the fasted chicks, suggesting that EFPH may have stimulated the splenic B cells proliferation, which in turn augmented the immune response [12,25]. In support of this, EFPH, particularly the T3 and T4 groups, up-regulated the splenic gene expression of IL4, TLR2, TNFα, and INFγ, which could enhance B cells proliferation and differentiation. The B and T lymphocytes have a key role in the immune response into which native T cells differentiate into Th1 cells (cellular immunity) and Th2 cells (humoral immunity) [21]. It has been recorded that glucose is an important source of energy for Th2 cells proliferation [38] and was observed to be decreased in fed deprived chicks for 24 h or more PH [39]. Shinde et al. [10] reported that the humoral and cellular immune responses was considerably improved in the FD layer-type chicks for 6, 12, and 24 h than those FD for 36 h PH. Moreover, Bakyaraj et al. [40] and Bhanja et al. [41] found that in ovo injection of amino acid and glucose enhanced the cellular and humoral immunity, as well as up-regulate the IL4, TLR2, TNFα, INFγ, and TNFα genes in chicks during the PH life, indicating their role in the development of the immune system of chicks. In agreement with our findings, Tamboli et al. [8] reported down-regulation of the splenic IL6, INFγ, TNFα, and TLR2 in the fasted birds compared to the early fed ones. Moreover, toll-like receptors (TLRs) was known to have a key role in innate immunity. Following the infection, TLRs initiate a signaling cascade for the cytokines production and up-regulate the co-stimulatory molecules [42]. In the current trial, the splenic TLR2 expression was up-regulated in the early fed chicks PH during the early stage of life. On contrary, Shinde et al. [10] observed up-regulation of splenic cytokines (TLR2 and IL6) as the FD period increased in layer chicks, and they did not observe any connection between in vivo immune response and immunity-related gene expression.In birds, the liver has a wide range of metabolic and homeostatic functions [43]. At day 14 and 28 PH, the histological examination of the liver revealed the presence of lymphatic cell aggregations that are variable in number and size, and with an uneven scattering in the parenchyma. It has been reported that the hepatic lymphatic aggregations in birds are considered as a part of the peripheral lymphoid tissue that is a key constituent of the lymphatic system due to birds do not have distinctive lymph nodes [44,45]. Lymphoid tissue was recognized in non-immune organs in birds, including liver, pancreas, kidney, endocrine glands, gonads, and the central nervous tissue [43,44,45]. It is still unknown that are these lymphoid tissues a burst of lymphomatosis or a normal reaction of the avian immune system to antigens [46]. Little is known about the liver of layer-type chicks, particularly under the influence of early nutrition or FDPH and thus this point needs to be further explained in future studies.The liver had an intense positive reaction to Best’s carmine stain and a greater glycogen density score in the early fed groups of this study, indicating that the liver of these chicks had numerous glycogen granules in the cytoplasm of hepatocytes. The liver glycogen concentration (majority) and muscle (minority) varies during the embryonic stage [47]. One day PH, both the heart and liver glycogen contents decline markedly to 40 and 16% of pre-hatching levels, respectively [48]. The glycogen contents in the liver persist at a lower level for several weeks PH and rise to the adult concentrations at 4 months of age [47]. Hepatic glycogen is constantly being formed and break down, its content varied depending on the carbohydrate (gluconeogenic substrates) intake and the glycogenolysis [47]. In the current study, EFPH or FWDPH may influence the carbohydrate metabolism of the liver in layer-type chicks, however the exact mechanism of action was not investigated. In line with the current findings, Kornasio et al. [49] observed that administration of carbohydrate before hatch via in ovo feeding combined with EFPH maximized the carbohydrate availability and elevated liver glycogen content. Furthermore, the early fed chicks had more than 30-fold increment in the hepatic glycogen content compared to the FD chicks [49].The proventriculus is the glandular stomach where hydrochloric acid and digestive enzymes are produced for the digestion process of feeds. In the present study, the histomorphological structure of proventriculus was normal and coincide with previous studies [26,50,51]. Our data revealed that the mucosal fold thickness, proventriculus mucosal thickness, and diameter of compound tubular gland lumen were greater in the fed groups PH than the fasted group. It has been reported that the relative weight of proventriculus increased gradually from day 1 to day 28 of age in birds [52]. Increased thickness of the proventriculus mucosa and gland are important indicators of the proventriculus development and suggesting that early feeding PH increased the production of the gastric juice, which is essential for digestion [53]. This effect has been accompanied by a rapid growth rate of the studied chicks as was reported in the previously published data of our study [7]. Previous researches have recorded the impacts of PHFD on the relative digestive organs’ weights of chickens, including the liver, proventriculus, gizzard, intestinal weight, and length [7,54]. It can be assumed that PHFWD delays the digestive organs development, particularly the proventriculus and the liver. In support of this, Maiorka et al. [55] observed that the FD chicks had lower relative proventriculus and gizzard weights after 48 h and 72 h of fasting compared to the fed ones.5. ConclusionsThe findings of this study reveal that EFPH with the experimental diets had no negative effect on the development of the immune and digestive organs. FWDPH reduced the cortex and cortex:medulla in the thymus and dampened the growth of the cortex, follicular width, and Pelicae height in bursa of Fabricius, with resultant influences on the primary lymphoid organs. Early nutrition PH enhanced the formation of germinal center areas in the spleen of early fed chicks, implying augmented proliferation of B cells in the secondary lymphoid organs. Early nutrition PH, particularly with diets rich in protein and energy contents or contain easily digestible carbohydrate, had an intense positive reaction to Best’s carmine stain in the liver of these groups, indicating that the liver of these chicks had numerous glycogen granules in the cytoplasm of hepatocytes. Improved proventriculus mucosal and gland thickness, as well as fold height were observed in the early fed chicks at day 14 of age, suggesting an increase in the production of the gastric juice, which is essential for digestion. The expression levels of splenic TLR2, IL4, TNFα, and INFγ were up-regulated in most of the early fed chicks compared to fasted ones at day 14 of age. In conclusion, EFPH could modify the splenic-immunity related genes and modulate the histomorphology of the lymphoid and digestive organs (liver and proventriculus) in layer-type chicks during their early life PH.

animals : an open access journal from mdpi

10.3390/ani11113087

PMC8614410

Animal-vehicle collisions are the major cause of rescue and need for hospitalization in wildlife referral centers. Radiography is used to assess the traumatized animal and is a rapid means to evaluate various traumatic pathologies. Clinicians can exploit radiography when making rapid decisions about whether to euthanize or treat an animal. We evaluated data (reasons for rescue, diagnosed lesions, and outcome) from a population of hospitalized wildlife ungulates and we investigated the benefits of using radiography as a diagnostic tool.

Animal-vehicle collisions are the major cause of rescue and need for hospitalization in wildlife referral centers. Clinicians need easy-to-use tools to make rapid decisions about whether to euthanize or treat injured animals. The aim of the study was to evaluate the data (reasons for rescue, diagnosed lesions, and outcome) from a hospitalized population of wildlife ungulates and investigate the benefits of using radiography as a diagnostic tool. Data were collected from three wildlife referral centers in Tuscany (Italy). The following information was collected for each animal: reason for hospitalization, clinical examination, radiographic examination, definitive diagnosis, and outcome. A chi-squared test was used to assess the benefits of radiography in detecting different traumatic lesions. Prevalence was reported according to the reason for hospitalization, definitive diagnosis, radiographic diagnosis, and outcome. The main reason for hospitalization was traumatic lesions due to vehicle collisions and 71.1% of the animals did not survive. Radiography was more useful in patients with traumatic axial skeletal lesions and/or multiple traumas with respect to traumatic appendicular skeleton lesions. Our results show that radiography is a useful diagnostic technique for assessing wildlife emergencies and it could help the clinician in making medical decisions.

1. IntroductionIn the last few decades, the population of wildlife ungulates has increased exponentially in Western Europe [1,2], including Italy, where our study took place [3,4]. This should be seen as part of a wider and general phenomenon, related to deep changes in the environment where these species live and interact with the human population [1].In Tuscany (central Italy), this increase was the result of the reduction in agricultural practices in hilly and mountainous areas, along with an increase in woodland or forested areas, changes in agricultural practices (i.e., the proliferation of winter cereals), less live-stock husbandry, changes in hunting and management (including controlled culling and reintroduction) practices, and a warm climate [5]. The increase in the wildlife ungulates population led to the expansion of animals in nearby areas leading to more interactions between humans and wildlife ungulates [5].Tuscany has wide expanses of rural areas crossed by extensive road networks and urban areas. The conflict between wildlife and human activities has led to a sharp rise in deer–vehicle collisions [3,4,6,7,8,9]. Some studies have shown that cervid movement is the main factor influencing collision risk and frequency, but also that deer–vehicle accidents are related to habitat, climatic, and traffic characteristics, as well as predation, hunting, and disturbance effects [10,11,12,13].The animal-vehicle collisions represent a direct cause of death for wildlife mammals every year, and the main cause of wildlife rescues and admission to specialized veterinary hospitals (VHs) for first aid [3,4], with a wide range of different species involved [13,14,15,16]. Animal-vehicle collisions not only affect wildlife populations, but also endanger humans [17]. Deer-vehicle collisions have been associated with ecological, social, and economic consequences, such as property damage, deer loss, and human injury and death [17].The specialized wildlife referral centers rescue and provide first aid to injured animals. To ensure the welfare of individuals, a proven process needs to be followed that enables the clinician to make a rapid decision about euthanasia or clinical recovery of animals [6,7,8,9]. The decisions should be made quickly, ideally within 48 h of admission, in order to prevent unnecessary suffering or casualties in captivity [6,7,8,9].An accurate clinical general veterinarian examination represents the starting point and sometimes it is followed by laboratory examinations, radiology, and ultrasonography [6,7,8,9]. Only approximately 40% of wildlife casualties, across all species and ages, are suitable for release back into the wild. Although radiology is frequently used to assess traumas in small animals [18], and several studies have evaluated its utility for various traumatic pathologies [19,20,21,22,23], to the best of our knowledge, there have been no studies on using radiology in wildlife ungulates to assess traumatic injuries.The aims were: (1) to assess data collected by three wildlife referral centers on the reason for rescue, the diagnosed lesions, and the outcome of a cohort of roe deer and fallow deer in Tuscany (Italy); (2) to assess the benefit of using radiology as a diagnostic tool during emergencies, investigating the feasibility of using the clinical diagnosis alone or associated with radiology.2. Materials and Methods2.1. Data CollectionIn this retrospective study, medical records for 2015–2020 were collected and analyzed from a cohort of rescued roe deer (Capreolus capreolus) and fallow deer (Dama dama), from three centers in Tuscany (Italy) specializing in giving first aid to wildlife. The deer were rescued in multiple areas of Tuscany (Pisa, Grosseto, Siena, Florence, and Arezzo), and from different municipalities. Details on rescue areas are reported in Table 1.The following information was included in the survey: (1) reason for hospitalization, (2) outcome, (3) clinical diagnosis, and (4) radiographic diagnosis, if performed. When radiographic examination was needed, it was performed under sedation or general anesthesia in a clinical setting in order to reduce the stress and handling time. All the animals that had undergone a radiographic examination were assessed in two of the rescue centers involved in the study. The radiographs were acquired using a high-frequency digital radiography system (MAXIVET 400 HF, Multimage s.r.l., Cavaria, Varese, Italy and Ida 9 G, ISOMEDIC, Somaglia, Italy). The clinical examinations and radiographs were performed under sedation or general anesthesia to reduce the stress and handling time. The same anesthetic protocol was followed in both the rescue centers. Deep sedation was obtained by the association of dexmedetomidine 8 ± 1.3 mcg/kg, ketamine 2 mg/kg, and midazolam 0.2 mg/kg given intramuscularly. During sedation, mask oxygen was administered to all the subjects. General anesthesia was performed using propofol 2 mg/kg IV, and endotracheal intubation was performed only in the case of respiratory depression.Based on the reason for hospitalization, the lesions of the deer were grouped as follows: (1) vehicle collisions (certain or assumed), (2) entrapment in nets/fences, (3) combine harvesters, gunshot, and predation.Based on the outcome, the deer were classified as follows: (1) survived (released/given custody), (2) died (spontaneously/euthanized), (3) unknown.According to the clinical diagnosis and the radiographic diagnosis (if performed), the deer were grouped into the categories shown in Table 2.2.2. Statistical AnalysisThe prevalent reason of hospitalization, the outcome, the clinical diagnosis, and radiographic diagnosis (if performed) were identified for each category of roe deer, fallow deer, and the total number of animals enrolled in the study.In order to verify the feasibility of using the clinical diagnosis alone or associated with radiology, the animals that had undergone both clinical and radiographic diagnosis were divided into two sub-groups: (a) sub-group A: traumatic skeletal injuries of the appendicular skeleton: in this group, forelimb and/or hindlimb fracture/luxation were included; (b) sub-group B: traumatic skeletal injuries of the axial skeleton and/or multiple trauma: in this group, vertebral fracture/luxation, pelvic fracture/diastasis, and multiple traumas were included. Furthermore, a chi-squared test was used to verify the differences between the two groups. The statistical analysis was performed using Graph Pad Prism (San Diego, CA, USA), and the significance was set at p < 0.05.3. ResultsA total of 1135 records were assessed, of which 1070/1135 (94.3%) were roe deer and 65/1135 (5.7%) were fallow deer. The main reason for hospitalization was traumatic lesions due to vehicle collision (certain or assumed) both in roe and fallow deer, as shown in Table 3.Regarding the outcome, most of the roe and fallow deer died spontaneously or were euthanized, as shown in Table 4.Clinical diagnosis pointed out a higher prevalence of multiple trauma both in roe and fallow deer, followed by the other traumatic skeletal lesions, as described in Table 5.X-rays were performed in 163 out of 1135 (14.4%) ungulates, of which 145/163 (89.0%) were roe deer and 18/163 (11.0%) were fallow deer. In 121/163 animals, the radiographic exam highlighted traumatic skeletal lesions, whereas in 42/163 patients, no traumatic skeletal lesions were detected. Table 6 shows the results of the prevalence of the radiographic diagnosis in the ungulates with traumatic skeletal lesions. The results show a similar distribution within categories, but forelimb fracture/luxation in roe deer was the least represented 8/145 (5.5%) and hindlimb fracture/luxation in fallow deer had the highest prevalence 5/18 (27.8%).Table 7 shows the results of the prevalence of the outcome in the ungulates with traumatic skeletal lesions. Most of the roe and fallow deer died spontaneously or were euthanized; within survived ungulates (8/121, 6.6%), 7/8 were multiple traumatic ungulates (2/7 fallow deer and 5/7 roe deer), and 1/8 hindlimb luxation in roe deer.Table 8 shows the results on the accordance between clinical and radiographic diagnosis in the group A and B, expressed as number and proportions (n/%) of the roe deer, fallow deer, and total population.A chi-squared test showed statistically significant differences (p < 0.0001) between groups A and B (Figure 1), considering the accordance or non-accordance between clinical and radiographic diagnosis.4. DiscussionOur retrospective study analyzed the data collected from a cohort of 1135 roe deer and fallow deer admitted to three rescue centers. The first aim was to assess the reason for rescue, the clinical and radiographic diagnosis (if performed), and the outcome.Our results showed that the main reasons for rescue and hospitalization were traumatic injuries and the most represented was trauma caused by certain or assumed collisions with vehicles. A study about rescued roe deer carried out in Emilia-Romagna (Italy) showed similar findings, with a prevalence of 71.4% of patients hospitalized for trauma [24]. Our results are in line also with previous reports [3,4] in which the prevalence for rescue was evaluated in a more limited geographical area in Tuscany (municipality of Pisa). Pacini and colleagues [4] reported a prevalence of 71% of deer emergencies due to road accidents.Our results are also in line with studies in other countries in Europe [25,26,27]. Accidents involving roe deer represent the majority of the wildlife collisions with vehicles in Lithuania [25,26] and Poland, where over half of the traffic incidents (66%) involving wildlife were collisions with roe deer [27]. Additionally, in a study performed in Switzerland about causes of mortality and morbidity in roe deer, the main diagnoses of non-infectious problems were traumas (61%), including blunt trauma due to traffic accidents [28]. In the UK, animal-vehicle collisions represented 37% of the adult badger casualties admitted to wildlife hospitals [29].Several authors have underlined that animal–vehicle collisions are usually unreported, and that accurate records are lacking [24,30,31,32], leading to an incorrect evaluation and monitoring of the current situation. Moreover, different ecological factors, such as density, areas with different landscapes, climates, and population structures influence the probability of deer having a car accident [33].Recording the number of car accidents involving animals combined with the numbers of rescued wildlife would allow the monitoring of the wildlife population [33]. Thus, in this light, our study could be used to promote surveillance and monitoring as part of national and international wildlife health surveys.Regarding the clinical diagnosis, the majority of deer enrolled in this study were affected by traumatic skeletal injuries and/or multiple traumas. This finding is in line with previous studies [34,35] which reported skeletal fractures as being the most common traumatic injuries, in particular related to animal–vehicle collisions [35].Most of the animals hospitalized died spontaneously or were euthanized (71.9% of roe deer and 69.2% of fallow deer). Our results are in line with previous studies that reported high mortality of rescued wildlife animals [4,30] and a low number of subjects suitable for release back into the wild (40%) [36].Our results on the outcome are likely related to the reason for hospitalization, in line with others [2,34]. In fact, the deer population evaluated in this study was mostly affected by severe traumatic injuries caused by vehicle collisions. In our study, radiographic diagnosis identified severe traumatic skeletal lesions (e.g., vertebral fracture/luxation, hindlimb/front limb fracture) that could not be successfully treated and therefore with no possibility of full rehabilitation, and thus the best option was euthanasia [37]. Only a small group of ungulates, mostly affected by multiple trauma (6.6%), were suitable for release back into the wild.In one of the three centers included in the study, the medical records were sometimes incomplete; thus, it was not possible to know the outcome of all the patients included, and, in these cases, the outcome was classified as “unknown”. This could represent a limit for the study.Our second aim was to verify the effectiveness of radiography compared to clinical diagnosis alone. We did not evaluate the impact of performing a radiographic examination on the outcome. Clinicians need to make rapid decisions about whether to euthanize or hospitalize animals [38,39]. The decisions should be made quickly, ideally within 48 h from admission, in order to prevent subsequent unnecessary suffering in captivity [38,39]. An accurate veterinarian examination is the starting point and is sometimes followed by laboratory tests, radiology, and ultrasonography examination [40].In small animals, radiography is frequently used to assess veterinary traumatized patients [18]; however, its utility has not been studied in wildlife ungulates. Our findings showed that radiography is more useful in animals affected by traumatic axial skeletal lesion and/or multiple trauma (group B) with respect to the traumatic appendicular skeleton lesions and traumas (group A). We found that the non-agreement between clinical and radiographic diagnosis was 95.1% in group B and 32.5% in group A. This finding agrees with a study performed in feline trauma patients in which whole-body radiographs were used to detect thoracic, abdominal, pelvic, and spinal injury [22]. Appendicular skeleton lesions could, perhaps, be diagnosed with a clinical evaluation alone (severe lameness, swelling, pain, deformity, abnormal mobility, or crepitus at the affected site) because the affected site is easier to localize and assess [41].Thoracic and abdominal ultrasound (US), and computed tomography (CT) are also useful diagnostic imaging techniques for traumatized animals [42,43,44,45]. The US assessment of the thorax and abdomen is reported to be a rapid and accurate method to detect traumas in dogs [42]. CT is considered the gold standard for the evaluation of acute canine spinal trauma [44]. It also appears to be more sensitive than ultrasound and radiography in the identification of thoracic pathologies in traumatized patients (i.e., pleural fluid, pulmonary contusion), but further studies are needed [45]. Thoracic or abdomen US or CT were not performed in the ungulates enrolled in this study; thus, a comparison between different diagnostic imaging procedures for the diagnosis of traumas was not possible.In other European countries, trauma has been described as one of the major causes or contributing causes of death in roe deer [46,47,48], and clinicians need to make a rapid decision about the euthanasia or clinical recovery of these animals [38,39]. Based on our results, radiography can help to identify traumatic lesions of the spine, of the pelvis, or multiple skeletal traumas, which indicate that the animal will probably not have a reasonable chance of survival upon release. Regarding the traumatic lesions of the appendicular skeleton, our results showed that a clinical evaluation could be sufficient for the diagnosis; however, in our opinion, radiography helps to correctly classify the type of lesion (e.g., fracture vs. luxation) and is essential for orthopedic surgery [49].5. ConclusionsOur results indicate that radiography examination is a useful diagnostic technique for assessing pathologies that are not clinically evident in rescued wild animals. We believe that the use of radiography is essential in deer emergencies with a history of traumatic injuries and is a key means to make a diagnosis and rapidly decide on the best treatment.

animals : an open access journal from mdpi

10.3390/ani12010114

PMC8749897

The meat production of Hu sheep affects the mutton supply, and how to increase the meat production is the primary focus of genetic breeding. Therefore, in order to explore the key factors affecting the proliferation and differentiation of Hu sheep skeletal muscle cells, we performed functional verification of miRNA at the cellular level. Our findings are helpful to clarify the molecular regulation mechanism of proliferation and differentiation of skeletal muscle cells, which will benefit the molecular breeding of Hu sheep.

The growth and development of skeletal muscle require a series of regulatory factors. MiRNA is a non-coding RNA with a length of about 22 nt, which can inhibit the expression of mRNA and plays an important role in the growth and development of muscle cells. The role of miR-22-3p in C2C12 cells and porcine skeletal muscle has been reported, but it has not been verified in Hu sheep skeletal muscle. Through qPCR, CCK-8, EdU and cell cycle studies, we found that overexpression of miR-22-3p inhibited proliferation of skeletal muscle cells (p < 0.01). The results of qPCR and immunofluorescence showed that overexpression of miR-22-3p promoted differentiation of skeletal muscle cells (p < 0.01), while the results of inhibiting the expression of miR-22-3p were the opposite. These results suggested that miR-22-3p functions in growth and development of sheep skeletal muscle cells. Bioinformatic analysis with mirDIP, miRTargets, and RNAhybrid software suggested IGFBP3 was the target of miR-22-3p, which was confirmed by dual-luciferase reporter system assay. IGFBP3 is highly expressed in sheep skeletal muscle cells. Overexpression of IGFBP3 was found to promote proliferation of skeletal muscle cells indicated by qPCR, CCK-8, EdU, and cell cycle studies (p < 0.01). The results of qPCR and immunofluorescence experiments proved that overexpression of IGFBP3 inhibited differentiation of skeletal muscle cells (p < 0.01), while the results of interfering IGFBP3 with siRNA were the opposite. These results indicate that miR-22-3p is involved in proliferation and differentiation of skeletal muscle cells by targeting IGFBP3.

1. IntroductionHu sheep is a Chinese local breed, famous for its high reproductivity worldwide. Today, improving meat production of Hu sheep is necessary for the sheep industry in China. Probing the molecular mechanism underlying the proliferation and differentiation of skeletal muscle cells can provide useful clue for this problem.MiRNA is a non-coding RNA with a length of about 22 nt. The seed region of miRNA can inhibit transcription and translation by targeting and binding mRNA, hence affecting cell proliferation, apoptosis at different developmental stages [1]. To date, several important miRNAs have been identified for regulating muscle growth and development, such as miR-205, miR-126, miR-60, miR-75, miR-133, miR-499, etc. [2]. MyoD1, a marker gene of muscle differentiation, is regulated by miR-1 and miR-206 [3,4]. MiR-192 targets myogenic regulator RB1, inhibiting the proliferation and promoting the differentiation of Hu sheep skeletal muscle cells [5]. MiR-128 participates in the regulation of the CDS region of myostatin and inhibits the proliferation and promotes differentiation of C2C12 cells [6]. MiR-143, miR-696, miR-34b, etc., are involved in the growth process of skeletal muscle cells, including proliferation and differentiation [7,8,9]. However, there are still a lot of miRNAs without functional verification in terms of muscle growth and development in sheep.MiR-22-3p has been found as one of the differentially expressed miRNAs during Hu sheep muscle development [10]. A recent study showed that miR-22-3p treats fibrous cataract by targeting HDAC6 [11]. Another report indicated miR-22-3p inhibits proliferation and promotes differentiation of porcine skeletal muscle cells [12]. Studies related to C2C12 cells showed that miR-22-3p inhibits proliferation and promotes differentiation of C2C12 cells, meanwhile promoting the transition from fast-twitch to slow-twitch [13,14]. Although quite a few research works have studied on miR-22-3p, there are no reports on skeletal muscle cells in Hu sheep.The IGFs are related to muscle growth and development. As IGF binding proteins, the IGFBPs also play an important role in biology process. IGFBP2 induces the proliferation and invasion of glioma cells through the β1/ERK signaling pathway, indicating that IGFBP2 can be used as a potential therapeutic target for gliomas [15]. IGFBP1 and IGFBP2 are regulated by insulin, which affect glucose tolerance, participate in glucose metabolism and lipid metabolism, and have a therapeutic effect on obesity [16]. Studies have shown that IGFBP3 can be used as a therapeutic target for lung adenocarcinoma metastasis to the brain [17]. Increasing expression of IGFBP3 could promote the formation of endothelial bone in rats fed with eleutherococcus extract mixture (EEM) [18]. The enhanced transcription of IGFBP3 accumulates the abundance of IGF1, which affects the growth and metabolism of mice [19]. Under repeated acute stress, the expression levels of IGFBP3 and IGF1 in pig blood were elevated, and the IGF system was activated at this time, indicating that IGFBP3 is involved in acute physiological stress response, inflammation pathways, and energy metabolism pathways [20]. Although there are many studies on IGFBP3 in cancer and metabolic pathways, its roles in the growth and development of skeletal muscle in Hu sheep are not yet known. In this study, we speculated that miR-22-3p could have a regulatory effect on the growth and development of skeletal muscle cells in Hu sheep.In summary, to explore the function of miR-22-3p regulating skeletal muscle cells of Hu sheep, we carried out qPCR, CCK-8, EdU, cell cycle, and immunofluorescence studies and found overexpression of miR-22-3p inhibited differentiation and promoted proliferation of skeletal muscle cells. We identified IGFBP3 as one of its target gene by Dual-luciferase reporter system assay. Our results suggested that miR-22-3p plays a significant role in the growth process of skeletal muscle cells by targeting IGFBP3 in Hu sheep, making miR-22-3p as a molecular marker for breeding.2. Material and Methods2.1. Ethics StatementAll experimental procedures were strictly in accordance with the management measures of experimental animals in Jiangsu Province (License Number: 45). All animal procedures used in this study were approved by the Ethics Committee for Animal Experiments of Yangzhou University (No. 202103279) and were performed in accordance with the Guidelines for Animal Experimentation of Yangzhou University (Yangzhou, China).2.2. Cell CulturePrimary skeletal muscle cells were isolated from three 56-day fetal sheep according to the previous method [21] and cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, Grand Island, NE, USA) supplemented with 20% fetal bovine serum (FBS, Gibco, Grand Island, NE, USA) and 1% Penicillin streptomycin mixture 100× (Solarbio, Beijing, China). HEK293T cells were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in DMEM supplemented with 10% FBS and 1% Penicillin streptomycin mixture 100×. Cells were cultured in 37 °C with 5% CO2.2.3. Plasmid Construction and Cell TransfectionThe coding region of IGFBP3 and pcDNA3.1 (+) vector were double digested with restriction enzymes Xho I and EcoR I (Takara, Kusatsu, Shiga, Japan), and the 3′UTR of IGFBP3 and psi-check2 plasmid (Youbio, Changsha, China) were double digested with restriction enzymes Xho I and Not I (Takara, Kusatsu, Shiga, Japan). Enzyme products were recovered by MiniBEST Agarose Gel DNA (Takara, Kusatsu, Shiga, Japan), connected by solution I ligase (Takara, Kusatsu, Shiga, Japan) and transformed using Trelief™ 5α Chemically Competent Cell (Tsingke, Nanjing, China). The recombinant vectors were extracted using an Endofree Mini plasmid kit II (Tiangen, Beijing, China) Extraction kit, followed by confirmation with enzyme digestion sequencing. The small interference sequences of IGFBP3 were designed and synthesized by Genepharma Co., Ltd. (Suzhou, China).When cells grew to 60% confluence, we transfected inhibitor and mimics of miR-22-3p, siRNA, or recombinant vector of IGFBP3 to perform interference or overexpression assays with jetPRIME transfection reagent (Polyplus, New York, NY, USA), with at least three replicates, respectively. After 24 h of transfection, the corresponding experiment was carried out. Specific sequences were shown in Table 1.2.4. RNA Preparation and qPCRWhen cells grew to 80% confluence, cellular RNA was extracted with reference to TRIzol total RNA extraction reagent (Beyotime, Shanghai, China). Reverse transcription was followed by a one-step reverse transcription kit (Tiangen, Beijing, China). Products were used to conduct qPCR using a CFX96 fluorescence quantitative instrument according to the 2× TSINGKE® Master qPCR mix (Tsingke, Nanjing, China) reagent manual. Primers were synthesized by Sangon biotech Co., Ltd. (Shanghai, China). Specific sequences were shown in Table 2.2.5. CCK-8 Cell Counting Kit AssaySkeletal muscle cells were cultured in 96-well plates until 60% cell confluence. Transfection was performed with 24-h incubation. Each group had 6 parallel wells, and CCK-8 Cell viability was tested at 0 h, 24 h, 48 h, and 72 h. Specific steps of CCK-8 reagent (Beyotime, Shanghai, China,) were as follows: to each well was added 10 μL of CCK-8 reagent, and cells were incubated in 37 °C for 2 h. Finally, cells were measured by a microplate reader (OD value at 450 nm).2.6. Cell Cycle Kit AssayWhen cells grew to 80% confluence, they were re-seeded onto 6-well plates (1 mL/well). Transfection was performed under 60% cell confluence, followed by 24h incubation. Cells were digested with 0.25% trypsin and centrifuged at 1000× g for 5 min with 1 mL PBS. Each well was added 1 mL pre-cooled 70% ethanol, mixed by pipetting, and was fixed at 4 °C for 24 h. Then, cells were rinsed by pre-cooled PBS and centrifuged. Next, a solution of propidium iodide was prepared according to the Beyotime Cell Cycle Kit (Beyotime, Shanghai, China). To each tube of cells was added 0.5 mL staining solution, incubated at 37 °C for 30 min in the dark, and detected by flow cytometric.2.7. 5-Ethynyl-2′-Deoxyuridine (EdU) AssayWhen cells grew to 80%, they were digested with 0.25% trypsin, added appropriate complete medium to prepare cell suspension, then distributed cells in 96-well plates (100 μL/well), and placed in 5% CO2, 37 °C constant temperature incubator. When they grew to 60%, we performed transfection. Each group had 6 parallel wells, and the EdU test was performed when the cells grew to 80%. Specific methods refer to the instruction of EdU cell proliferation detection (Ribobio, Guangzhou, China) and observation with a fluorescent inverted microscope.2.8. Immunofluorescence StainingWhen cells grew to 80%, they were distributed to 24-well plates (500 μL/well). When they grew to 60%, we performed transfection. At the same time, cells were cultured in DMEM supplemented with 2% horse serum (Solarbio, Beijing, China) and 1% Penicillin streptomycin mixture 100 ×. There were 3 parallel wells in each group, and the cellular immunofluorescence test was carried out after 4–5 days of induced differentiation. Specific steps of immunofluorescence staining were as follows: The cells were washed twice with PBS, 3–5 min each time, and to each well was added 300 μL 4% paraformaldehyde (Solarbio, Beijing, China); they were fixed for 30 min. Then, they were washed in the same way and permeated with 0.5% Triton X-100 (Solarbio, Beijing, China) for 20 min, and added 300 μL of 1% BSA (Solarbio, Beijing, China) to block for 1 h. Next, each well added 200 μL primary antibody MyH3 (1:400) (Affinity, Changzhou China) at 4 °C overnight, then added 200 μL secondary antibody (1:1000) (Abconal, Wuhan, China), incubating for 2 h. Finally, each well added 200 μL DAPI (5 μg/mL) (Beyotime, Shanghai, China) for staining in 5 min. Fluorescence inverted microscope was used to observe cell differentiation and myotube formation.2.9. Dual-Luciferase Reporter System AssayWhen HEK293T cells grew to 80%, they were distributed to 24-well plates (500 μL/well). When they grew to 80%, we performed transfection. Each group was in 3 parallel holes. Groups were designed as follows: miR-22-3p mimics-NC + IGFBP3-wild, miR-22-3p mimics + IGFBP3-wild, miR-22-3p mimics -NC+ IGFBP3-mutant, miR-22-3p mimics + IGFBP3-mutant. After 48 h of transfection, the cells were tested according to the Dual Luciferase Reporter Assay Kit (Vazyme, Najing, China). Specific steps were as follows: To each well was added 100 μL 1 × Cell Lysis Buffer, pipetted into a 1.5 mL centrifuge tube, centrifuged at 12,000× g for 2 min. Next, to each well was added 100 μL of Luciferase Substrate to the microplate, absorbing 20 μL of the above cell lysate, immediately detecting the activity of the Firefly luciferase reporter gene. To the above reaction solution was added 100 μL Renilla substrate, immediately detecting the activity of the Renilla luciferase reporter gene. The ratio of Renilla fluorescence value to firefly fluorescence value is the relative luciferase activity, and the ratio is statistically analyzed with the ratio of the control well.2.10. Statistical AnalysisResults were presented as the Mean ± SEM. One-way analysis of variance was used to perform variance analysis and significance test. Results were considered significant at p < 0.05 and highly significant at p < 0.01.3. Results3.1. miR-22-3p Regulates the Proliferation of Skeletal Muscle Cells in Hu SheepWe used miRBase software to get the sequences of miR-22-3p across diverse species and found that it was highly conserved (Figure 1a). In order to study the effects of miR-22-3p in Hu sheep skeletal muscle cells, we transfected miR-22-3p mimics into the sheep skeletal muscle cells. The qPCR results showed that after overexpression of miR-22-3p, the mRNA expression of proliferation marker genes PCNA, CDK2, and cyclin D1 were significantly lower than that of the control group (p < 0.01) (Figure 1b). The proliferation status of sheep skeletal muscle cells was detected by CCK-8 reagent, and the OD450 value of the miR-22-3p group was significantly reduced (p < 0.01) (Figure 1c). The results of EdU staining showed that the number of EdU positive cells was significantly reduced compared with the control group (p < 0.01) (Figure 1d,e). In addition, the results of flow cytometry showed that the number of cells in S phase was significantly lower than that of the control group (p < 0.01) (Figure 1f,g). Next, we transfected the miR-22-3p inhibitor into the cells and found that the expression levels of CDK2, cyclinD1, and PCNA increased compared with the control group (p < 0.01) (Figure 1h). The results of CCK-8 showed that the OD450 of the cells increased (p < 0.01) (Figure 1i). In addition, the number of EdU positive cells increased significantly (p < 0.01) (Figure 1j,k). At the same time, the results of flow cytometry showed that the number of cells in S phase was significantly higher than that of the control group (p < 0.01) (Figure 1l,m). The above results indicated that miR-22-3p could inhibit the proliferation of skeletal muscle cells in Hu sheep.3.2. miR-22-3p Regulates the Differentiation of Skeletal Muscle Cells in Hu SheepAfter transfection of miR-22-3p mimics and NC, the differentiation marker genes MyoD and MyoG were significantly increased as indicated by qPCR (p < 0.01) (Figure 2a). At the same time, the immunofluorescence results showed that the number of MyH3 positive myotubes was more than that of the control group (Figure 2c). Next, we transfected the miR-22-3p inhibitor and NC into cells, and the differentiation marker genes, MyoD and MyoG, were significantly reduced (p < 0.01) (Figure 2b). Immunofluorescence results showed that the number of MyH3 positive myotubes was less than that of the control group (Figure 2d). These results indicated that miR-22-3p could promote the differentiation of skeletal muscle cells in Hu sheep.3.3. miR-22-3p Regu’Lates the Expression of IGFBP3We used mirDIP and miRTargets software to predict the target genes of miR-22-3p. The result was confirmed by RNAhybrid (Figure 3a). We focused on IGFBP3 because the function of this gene in sheep skeletal muscle cells is unknown yet, and it is differentially expressed during skeletal muscle development (unpublished RNA-seq data). At the same time, the mirDIP bioinformatic software predicted IGFBP3 as target with a high score. The tissue expression profile showed that IGFBP3 had the highest expression in skeletal muscle cells (Figure 3b). The results of the dual luciferase reporter vector showed that the fluorescence activity is reduced (Figure 3c). After transfection of miR-22-3p mimics and NC, the expression of IGFBP3 was significantly reduced (p < 0.05) (Figure 3d).3.4. IGFBP3 Regulates the Proliferation of Skeletal Muscle Cells in Hu SheepWe constructed the overexpression vector of IGFBP3 and transfected it into skeletal muscle cells. The qPCR results showed that the expression of key genes CDK2, cyclin D1, and PCNA for cell proliferation increased significantly (p < 0.01) (Figure 4a). CCK-8 showed that the OD450 of the cells was significantly higher than that of the control group after transfection 24 h (p < 0.01) (Figure 4b). In addition, the number of EdU positive cells increased significantly (p <0.01) (Figure 4c,d). The results of flow cytometry showed that the number of cells in S phase was significantly higher than that of the control group (p < 0.01) (Figure 4e,f). Next, we interfered IGFBP3 mRNA with siRNA. The qPCR results showed that the expression level of cyclin D1 was extremely significantly reduced (p < 0.01), and the mRNA expression of CDK2 and PCNA was significantly reduced (p < 0.05) (Figure 4g). CCK-8 showed that the OD450 value of the interference group was significantly reduced after 48 h and 72 h compared with the control group (p < 0.01) (Figure 4h). EdU staining results showed that the number of positive cells was significantly reduced (p < 0.01) (Figure 4i,j). In addition, the results of flow cytometry showed that the number of cells in S phase was significantly lower than that of the control group (p <0.01) (Figure 4k,l). These results indicated that IGFBP3 could regulate proliferation of skeletal muscle cells.3.5. IGFBP3 Regulates the Differentiation of Skeletal Muscle Cells in Hu SheepTransfecting of IGFBP3 overexpression vector and NC, we found that the differentiation marker gene MyoD was significantly reduced (p < 0.05), and MyoG was extremely significantly reduced compared with the control group (p < 0.01) (Figure 5a). At the same time, the immunofluorescence results showed that the number of MyH3 positive myotubes was less than that of the control group (Figure 5c). Next, we transfected the IGFBP3 siRNA into skeletal muscle cells, and the differentiation marker genes MyoD and MyoG were significantly increased (p < 0.01) (Figure 5b). Immunofluorescence results showed that the number of MyH3 positive myotubes was more than that of the control group (Figure 5d). In summary, these results indicated that IGFBP3 could regulate the proliferation of skeletal muscle cells in Hu sheep.Through cell proliferation and differentiation verification experiments, we found that overexpression of miR-22-3p could inhibit proliferation and promote differentiation of skeletal muscle cells by targeting IGFBP3 in Hu sheep. These findings may benefit the understanding of the mechanism of growth and development of skeletal muscle cells at miRNA levels.4. DiscussionMiRNA was first found in nematodes and was named lin-4, and another miRNA named let-7 was found, both of which were about 22 nt in length. These miRNAs can bind to the 3′UTR of mRNA, thereby regulating the development of nematodes [1]. In the following years, miRNAs and multiple potential mechanisms of their binding were discovered in thousands of species, opening up a new world in the field of scientific research. MiRNA is mainly processed and produced in the cytoplasm and transported to the nucleus. It can be used in transcription activation. Studies have revealed that miR-744 functions in the transcription initiation site of Cyclin B1 under Ago1 participating [22]. There are also related reports that miRNA can target the 5′UTR region of mRNA and inhibit mRNA expression. The classic miRNA regulatory network aims to inhibit mRNA expression by targeting the 3′UTR region of mRNA. In colorectal cancer tissues, tumor transcription factor Jun inhibits the transcription of miR-22, and miR-22 targets TIAM1. Reducing the expression of miR-22-3p inhibits the proliferation of cancer cells [23,24,25]. In livestock, this classic regulation mechanism has also been reported. Research shows that miR-1 and miR-206 target PAX7, i.e., inhibiting the proliferation of skeletal muscle by inhibiting the expression of these two miRNAs [26], thereby regulating muscle growth and development.Although many miRNAs have been reported in domestic animals, there are still a lot of miRNAs that require further verification in terms of muscle growth. Since the number of myotubes is constant before birth, we focus on miRNAs with low expression before birth and high expression after birth. MiR-22-3p has been detected in the longissimus dorsi muscle of sheep and the expression of miR-22-3p is the lowest at 60 days of pregnancy and the highest at 360 days after birth [10]. Cancer literatures proved that miR-22-3p targets SP1, inhibits the expression of downstream genes CCND1 and BCL2, thereby inhibiting the growth of liver cancer cells, and the low expression of miR-22-3p is associated with metastatic liver cancer [27,28]. Hsa-miR-22-3p serves as the target of DGCR5, reducing its expression to inhibit the occurrence of lung cancer [29]. It is documented that the transfection of miR-22-3p inhibits the proliferation of skeletal muscle cells and promotes differentiation in porcine skeletal muscle cells [12]. Overexpression of miR-22-3p inhibits the proliferation of C2C12 cells (mouse myoblast) and promotes the differentiation of muscle fibers, promoting the transition of mouse C2C12 myotube fiber type from fast-twitch to slow-twitch [13,14]. However, the function of miR-22-3p regulating skeletal muscle cells in Hu sheep has not been reported yet. What is exciting is that we found that the mature sequence of miR-22-3p is completely conserved among different species, which further supported our assumption. Thence, we carried out experiments in Hu sheep skeletal muscle cells.In order to explore the function of miR-22-3p in Hu sheep skeletal muscle cells, we adopted qPCR, CCK-8, EdU, cell cycle, and immunofluorescence assays. Through the preliminary verification by qPCR, the proliferation marker genes of the miR-22-3p mimics transfected group were significantly reduced, indicating that overexpression of miR-22-3p inhibited the proliferation of skeletal muscle cells. PCNA, CDK2, and cyclin D1, the well accepted proliferation marker genes, were used in study to detect whether skeletal muscle cells are in a proliferating state [30]. However, it has been reported that CDK4, cyclin E2, and E2F1 are used as pygmy killer whale skin fibroblasts proliferation genes [31], which may be due to the genes in specific expression of different species and different tissues. However, we still want to continue to explore the influence of the downstream genes of miR-22-3p on the growth and development of skeletal muscle cells. Through qPCR and immunofluorescence, we found that the differentiation marker genes were significantly up-regulated after overexpression of miR-22-3p. MyoD and MyoG, differentiation markers, have also been used in the research of fetal bovine skeletal muscle [30]. These results indicated that overexpression of miR-22-3p inhibited the proliferation and promoted differentiation of skeletal muscle cells. However, we still want to explore the influence of the downstream regulatory elements of miR-22-3p regulating the growth and development of skeletal muscle cells in Hu sheep.Bioinformatic analysis with mirDIP, miRTargets, and RNAhybrid software predicted the targeting relationship between downstream genes and miR-22-3p. Previous study has identified HDAC6 as a downstream target gene of miR-22-3p in lens epithelium cells [11]. Here, we predicted it through miRTargets software, but its score is not high. What is exciting is that we found IGFBP3 has a high score in miRTargets, mirDIP, and RNAhybrid software, and the seed sequence of miR-22-3p completely binds to the 3′UTR region of IGFBP3, so we focused on IGFBP3. A large number of literature reports that IGFs are indispensable in the growth and development of skeletal muscle. Studies have shown that IGFs promote the proliferation and differentiation of muscle cells, as well as regulate each other with MRF and MyoD to improve muscle hypertrophy and regeneration [32,33]. Studies showed that IGFBP1 activates ERK1/2 pathway to promote the proliferation of smooth muscle cells (SMCs) by regulating IGF1 [34]. IGFBP6 inhibits the expression of IGF2 and activates the MAPK pathway to promote muscle differentiation, and the activation of this pathway does not require IGF-1R or insulin receptor (IR) to participate [35]. These findings reveal that the IGFBPs are involved in the growth and development of muscles. Based on this, we speculated that IGFBP3 has a similar effect on Hu sheep skeletal muscle. The tissue expression profile preliminarily certificated that IGFBP3 is highly expressed in skeletal muscle cells. Moreover, a large number of research reports illustrate that IGFBP3 is currently only studied in diseases. IGFBP3 activates the XBP1/IGFBP3/MMP-9 pathways to regulate the invasion and metastasis of non-small cell lung cancer (NSCLC) cells [36]. Studies have shown that Vi4 and miRNA-185-5p competitively combine with IGFBP3 to promote neuronal cell proliferation and reduce the risk of neonatal hypoxic ischemic encephalopathy (HIE) [37]. In view of the fact that no research related to muscles has been carried out, we made a bold attempt to deeply explore the effect of IGFBP3 in skeletal muscle cells of Hu sheep. Experiments were carried out by qPCR, CCK-8, EdU, cell cycle, and immunofluorescence studies, and these studies indicated that overexpression of IGFBP3 promoted proliferation and inhibited differentiation of skeletal muscle cells in Hu sheep.5. ConclusionsIn summary, our results indicated that overexpression of miR-22-3p inhibited proliferation and promoted differentiation of skeletal muscle cells by targeting IGFBP3 in Hu sheep. The expression activity of miR-22-3p is low, which indicates that the Hu sheep muscle cell proliferation efficiency is high, and miR-22-3p is used to identify the inflection point of Hu sheep muscle growth. Our findings are helpful to clarify the molecular regulation mechanism of proliferation and differentiation of skeletal muscle cells, which will benefit the molecular breeding and theoretical basis to mutton producers of Hu sheep.

animals : an open access journal from mdpi

10.3390/ani11082449

PMC8388784

Contagious agalactia is a multi-faceted disease affecting small ruminants worldwide. It is caused by four different Mycoplasma (sub)-species. In the absence of highly efficient vaccines, its control relies mostly on antibiotic treatment. Tetracyclines are one of the main families used, as they are cheap and often ensure clinical recovery, if not microbial clearance. However, some isolates have shown lowered susceptibilities even without the mutations in the target gene known to result in resistance. We suspected that an active efflux mechanism could be responsible for such lower-susceptibility phenotypes. Using various techniques, we demonstrated that most of the strains we studied did exhibit a capacity to actively extrude various substances including tetracyclines. This might contribute to low resistance profiles.

Contagious agalactia is associated with mastitis, keratoconjunctivitis, arthritis, pneumonia, and septicemia in small ruminants in countries with large dairy industries worldwide. The causative agents belong to four (sub)species of the Mycoplasma genus that have remained essentially susceptible to antimicrobials, including to the widely-used tetracycline family. However, some clinical isolates have been detected that show increased minimum inhibitory concentrations of tetracyclines, although they do not harbor the mutation in the 16SrRNA gene usually associated with resistance. The present work aimed to assess whether efflux pumps, infrequently described in mycoplasmas, could participate in the observed moderate loss of susceptibility. General efflux mechanisms were measured (i) using the fluorescence property of ethidium bromide when accumulated intracellularly and intercalated in the mycoplasma genomes, its active extrusion resulting in a temperature-dependent decrease in fluorescence and (ii) monitoring the growth inhibition of mycoplasmas by subinhibitory concentrations of tetracycline with or without reserpine, a known inhibitor of efflux in other bacteria. Both methods revealed non-specific efflux phenomena in most of the isolates tested, although their efficacy was difficult to quantify. This property could contribute to the acquisition of mutations conferring resistance by maintaining intracellular concentrations of tetracyclines at subinhibitory levels.

1. IntroductionContagious agalactia (CA) is an infectious syndrome affecting small ruminants worldwide and especially countries in Southern Europe with large dairy industries. Its main clinical signs are mastitis, keratoconjunctivitis, and arthritis, but others such as pneumonia and septicemia have also been reported [1]. The negative economic impact of CA comes essentially from milk production loss and morbidity in young animals [2]. CA can be caused by four different Mycoplasma (sub)species: M. agalactiae (Ma), the primary causative agent isolated from Spanish sheep and goat herds [3], and M. mycoides subsp. capri (Mmc), M. capricolum subsp. capricolum (Mcc), and M. putrefaciens (Mput), which belong or are related to the M. mycoides cluster [4] and are seldom isolated from sheep. Mmc is the main etiological agent responsible for clinical caprine CA in France [5,6].There are no commercially available vaccines with demonstrated efficacy [7], and thus controlling CA mainly relies on biosafety strategies (culling or isolation of infected animals, hygiene measures, etc.) and chemotherapy [8]. Among the few antimicrobials with marketing authorization for small ruminants, the convenient broad-spectrum, low-cost tetracyclines are often used [9]. The activity of oxytetracycline against CA-causing mycoplasmas has been demonstrated in vitro in several studies, but isolates with increased minimal inhibitory concentration (MIC) suggest emerging resistance in the field [6,8,10,11,12,13,14].In bacteria, resistance to tetracyclines has been associated with (i) ribosomal protection proteins, (ii) target mutation corresponding to mutations in the 16S rRNA (rrs) genes, (iii) efflux, or (iv) drug enzymatic inactivation [15,16]. Tet(M)-related ribosomal protection was shown to result in high tetracycline MIC values (MIC ≥ 8 µg/mL) in two human mycoplasmas, namely M. hominis and Ureaplasma spp. [17,18,19], but this mechanism has never been evidenced in animal mycoplasmas [20]. In the bovine pathogen Mycoplasma bovis, tetracycline resistance was correlated to hot-spot mutations in the 16S rRNA encoding genes [21,22]. In its closely related counterparts in small ruminants, M. agalactiae, some strains with decreased susceptibility did not harbor any significant binding site alterations [10]. Efflux pumps able to extrude tetracyclines were firstly characterized in Escherichia coli in 1980 [23] when more than 30 tetracycline pumps have been described in Gram-negative and -positive bacteria [15,16]. However, studies assessing efflux pumps in Mycoplasma spp. are scant and focus mainly on fluoroquinolones [24,25,26] or macrolides [27]. There are no previous reports on tetracycline efflux in Mycoplasma spp.In a previous study, we evidenced that MICs at up to 4 µg/mL in different CA-associated clinical isolates were not always correlated with the presence of hot-spot mutations in the rrs genes, or mutations in the rspJ gene (30S ribosomal subunit protein S10) [10]. The aim of the present work was to assess whether efflux pumps could participate in the observed moderate loss of susceptibility. General efflux mechanisms were first measured indirectly using the fluorescence property of ethidium bromide (EtBr) intercalated between the bases of the genomic DNA. EtBr is known to (i) cross the cytoplasmic membrane in bacteria, (ii) accumulate intracellularly, generating an EtBr-polynucleotide fluorescent complex, and (iii) in cases of resistance, be extruded by both proton-motive force and ATP-dependent systems [28]. The efflux was then assessed by monitoring the growth inhibition of Mycoplasma spp. by tetracycline at subinhibitory concentrations in the presence or absence of reserpine, a plant-derived alkaloid known to inhibit tetracycline efflux in other bacterial models [29]. Both methods demonstrated the existence of an efflux, but its quantification in different isolates was difficult to standardize.2. Materials and Methods2.1. Mycoplasma IsolatesA total of 32 Mycoplasma spp. isolates were collected from contagious agalactia outbreaks (mastitis, arthritis or pneumonia) or from bulk tank milk during inspections in France and Spain (Table 1). They comprised 17 M. agalactiae and 15 strains of the so-called M. mycoides cluster or related (5 Mmc, 4 Mcc, and 6 Mput). Their identification, oxytetracycline MICs (from 0.25 to 8 µg/mL) and their corresponding 16S rRNA and rpsJ genotypes were characterized in a previous study [10].2.2. Efflux Assay by EtBr (EtBr)–Agar MethodThis agar-based method relies on the ability of EtBr to cross the cytoplasmic membrane, accumulate in mycoplasma cells, and fluoresce under UV light when intercalated in DNA. Active efflux systems result in a temperature-dependent depletion of intracellular EtBr and hence a decrease in the fluorescence. Our assays were adapted from the work of Martins et al. on other bacterial models with some modifications [30,31]. Briefly, two-fold EtBr dilutions ranging from 0.2 to 6.4 µg/mL were prepared in PPLO–agar (Indicia Production, St Genis L’Argentière, France). Plates were inoculated with spots of 1 µL of each isolate at 107–108 cfu/mL for M. agalactiae and 106–107 cfu/mL for M. mycoides cluster or related strains, and were incubated at 37 °C with 5% CO2 for 72 h to obtain confluent colonies. One set of plates (Set 1) was then incubated in the same conditions for another 24 h. The other set (Set 2) was stored at 4 °C. After the total 96 h incubation, fluorescence was examined under UV light using a transilluminator (Gel DocTM XR System, Bio-Rad, Marnes-la-Coquette, France), with an exposure time of 1 s, and plates were photographed.2.3. Monitoring of EtBr Efflux by FluorometryThis technique was adapted from previously described protocols for other bacteria or other mycoplasmal species [26,32,33]. Briefly, mycoplasma isolates were grown in 2 mL of PPLO broth (Indicia, France) until they reached the stationary phase (~109 cfu/mL), centrifuged to obtain a cell pellet (10,000× g, 20 min, 20 °C) and resuspended in 2 mL of PBS 1X with EtBr at 10 µg/mL and reserpine at 20 µg/mL to inhibit potential efflux. These cells were then incubated at 37 °C for 30 min with agitation to facilitate the uptake of EtBr. Strains were then centrifuged again (10,000× g, 15 min, 20 °C) and the supernatant was discarded. Cells were resuspended in 1 mL (for M. agalactiae strains) or 2 mL (for M. mycoides cluster or related strains) of PBS 1X. For each condition studied, 200 µL of this cell suspension were placed in a well of a Greiner black, flat-bottom, 96-well plate (Sigma-Aldrich, Saint-Quentin-Fallavier, France) to measure fluorescence every minute for 1 h. Bacterial efflux activity was assessed in different conditions: (i) with no energy source (only PBS 1X), (ii) adding energizers such as pyruvate (0.5% m/V final concentration) or glucose (0.5% m/V final concentration), and (iii) at 37 °C versus 25 °C to reduce mycoplasma metabolism. Measurements were made using the CLARIOstar® plate reader (BMG LABTECH, Champigny s/Marne, France) with excitation and emission wavelengths of 360 nm and 590 nm, respectively, as recommended by the EtBr supplier (Sigma-Aldrich, Saint-Quentin-Fallavier, France). Focal height and gain were automatically adjusted before initiating the reading process. The maximal fluorescence activity reached after EtBr uptake was normalized to 1 (i.e., 100%). Energizers were added after the first 5 min and at the end of the experiment (60 min), and SDS (Sodium Dodecyl Sulfate, 1% m/V final concentration in each well) was added to lyse cells and release the remaining EtBr. Analyses were repeated three times for each strain and condition.2.4. Growth in the Presence of Subinhibitory Concentrations of OxytetracyclineTen isolates (5 M. agalactiae, 3 Mmc and 2 Mcc) were grown in the presence of subinhibitory concentrations of oxytetracycline (one fourth of their MIC) with or without reserpine (20 µg/mL) at 37 °C using a protocol adapted from Kovacevic et al. [34]. Briefly, mycoplasma cells were inoculated at 104 cfu/mL in PPLO broth supplemented with 0.5% (m/V) pyruvate in different wells of a sterile 96-well plate (final volume in each well 200 μL). When necessary, reserpine and/or oxytetracycline were added. The effect of reserpine alone on bacterial growth was also assessed. The plates were sealed and incubated with agitation at 37 °C. Colony counts were performed at t = 0, 16, 24, 40, 48, 64, 72, 88, and 96 h of incubation using a multi sample inoculator to deposit 1 μL from each well onto an agar plate (Mast Uri® Dot, Mast Diagnostic, Amiens, France). To obtain countable values, 1/100 and 1/20,000 dilutions were made in sterile PBS 1X. Each experiment was repeated twice.3. Results3.1. Efflux Assay by EtBr (EtBr)–Agar MethodFigure 1 illustrates the different patterns of colonies fluorescence using increasing concentrations of EtBr and different incubation schemes. Some isolates of Mput, not shown in Figure 1, were inhibited at EtBr concentrations of 1.6 μg/mL, suggesting the MIC was reached, and were not further analyzed (noted “inhibited” in Table 1). The appearance of pixel saturation resulted in a red coloration of bacterial spots. This saturation increased as expected with increased concentrations of EtBr. In the absence of red saturation for 8 M. agalactiae and 2 Mput strains, the fluorescence was not interpreted (n.i., not interpreted in Table 1), as the white fluorescence was difficult to distinguish from the bacterial spot background. Most of the Mmc and Mcc strains showed saturated colonies at 1.6 µg/mL of EtBr at 37 °C, whereas the M. agalactiae strains did so at 6.4 µg/mL except for strain L16160 and Ag9 (3.2 µg/mL). This could result from a better EtBr intake capacity and/or a lower efflux for Mcc and Mmc strains. For all these strains, the fluorescence level (read as the relative proportion of saturated pixels) was compared between plates transferred to 4 °C and those left at 37 °C, after a 72 h-intracellular accumulation of EtBr at 37 °C. The 9 M. agalactiae isolates with saturated pixels were divided into two groups: one with a moderate almost equivalent saturation at 4 °C and 37 °C at 6.4 µg/mL of EtBr (Ag26, F10671, Ag10, L16156) and one with a markedly high saturation at 6.4 µg/mL of EtBr at 4 °C, but not at 37 °C (L16160, 5632, L15242, L16086, Ag9, noted as “+++” in Table 1). For most of the Mcc and Mmc strains, except for strains Mcc F10261 and cap3, the efflux was blocked at 4 °C, resulting in a saturation of fluorescence as low as 0.4 µg/mL.This suggests that most but not all of the strains express temperature-dependent systems capable of extruding EtBr [30,31]. A semiquantitative evaluation of the efflux system efficacy was done based on both the EtBr concentrations resulting in saturation at 37 °C (the higher the concentration, the more efficient the system) and the intensity difference between 4 °C and 37 °C (the greater the difference, the more efficient the efflux) (Table 1).3.2. Monitoring of EtBr Efflux by FluorometryTo quantify EtBr efflux more accurately, a real-time detection method was developed to monitor the loss of fluorescence of EtBr previously accumulated in Mycoplasma cells and bound to DNA and RNA. The effect of the incubation temperature (25 °C versus 37 °C) and presence of energizers (glucose or pyruvate) was also tested. Figure 2 illustrates the changes in fluorescence in each of the studied conditions for six representative strains. Supplementary Figure S1 (M. agalactiae) and Figure S2 (M. mycoides susbp. capri; M. capricolum subsp. capricolum and M. putrefaciens) give the results obtained for all the isolates studied. The maximum intensity reached at the beginning of the experiment was normalized to 1. In the absence of any energizer, the fluorescence was shown to decrease to [0.66–0.83] at 37 °C and only to [0.84–0.96] at 25 °C, which could result from temperature-dependent diffusion or low-level efflux (Figure 2). Adding an energizer resulted in an immediate drop in fluorescence, more marked with pyruvate than with glucose (tested only for subspecies of, or related to, the cluster M. mycoides), followed by a slower fluorescence decrease to reach [0.28–0.66] at 37 °C and [0.52–0.72] at 25 °C after 60 min, with large differences between the different isolates (Figure 2), indicating the presence of temperature-dependent efflux pumps. Interestingly, when available, the kinetics for fluorescence decrease were roughly parallel in the presence of pyruvate and glucose (e.g., see strain Mcc cap19 or Mput Put13), suggesting a similar mode of action for these two energizers.The final addition at t = 60 min of SDS released the remaining intracellular EtBr and reduced fluorescence to [0.2–0.3] for all the strains studied. The efflux capacity observed in the presence of an energizer varied widely between strains. An attempt was made to convert the observed kinetics into semiquantitative data for comparison between strains (Table 1). The total, i.e., immediate and slower, drop in fluorescence after adding the energizer, was considered. It was calculated by subtracting the residual fluorescence after 60 min in the presence of the energizers from the fluorescence at t = 60 min with PBS alone, at both 25 °C and 37 °C. Most of the strains were shown to be able to actively extrude EtBr, more efficiently at 37 °C than at 25 °C and in the presence of an energizer, and for isolates of the M. mycoides cluster, the maximum drop in fluorescence was obtained with pyruvate at 37 °C. In this condition, the maximum drop in fluorescence ranged between −0.14 and −0.52. For instance, Mcc Cap19, F10261 or M. agalactiae F11129, Ag28 and Ag304 were able to expel almost 100% of EtBr at 37 °C in the presence of pyruvate, and the addition of SDS resulted only in a slight further drop in fluorescence. By contrast, M. agalactiae L4212c and Ag26 had poor efflux capacity: half of the fluorescence was still present after 60 min at 37 °C in the presence of pyruvate. No poor efflux capacity was observed within strains of the M. mycoides cluster.However, these observations did not correlate well with the categorization obtained using the EtBr–agar method (Table 1). M. agalactiae strains with an intermediate (++) or high (+++) efflux activity on agar screening showed a mean loss of fluorescence of 0.38 +/− 0.02 (n = 5, [0.34–0.40]) while those ranked (+) showed a very variable loss, with a mean of 0.31 +/− 0.11 (n = 4, [0.18–0.42]). A comparison was also performed for uninterpretable strains, for instance with strain Ag28, which did not become fluorescent in any of the conditions tested in the EtBr–agar method, yet showed the highest active efflux activity of the studied M. agalactiae strains (Figure 1 and Table 1). Likewise, Mcc and Mmc strains, with an intermediate (++) or high (+++) efflux activity, showed a loss of fluorescence of 0.40 +/− 0.06 (n = 7, [0.34–0.52]) not different from strains ranked (+) or (+/−) (0.39 and 0. 48 respectively). EtBr-based methods thus demonstrated a general efflux capacity, temperature sensitive and energy-driven, that varies among strains, albeit more universal within the cluster M. mycoides, but in our hands, they did not adequately quantify this efflux.3.3. Growth in the Presence of Subinhibitory Concentrations of OxytetracyclineThis part of the study was designed to assess whether the general efflux observed with EtBr was also able to extrude oxytetracycline and could contribute to increased MIC values (Table 1). Nine isolates with increased oxytetracycline MICs of 2, 4, and 8 µg/mL, respectively, and with intermediate (++) or high (+++) EtBr efflux were selected. Strain Mcc Cap19 was also added as a control, because it showed high EtBr efflux capacity but had a low MIC of oxytetracycline (0.25 µg/mL). These isolates were grown in the presence of oxytetracycline at a concentration corresponding to ¼ of the MIC with or without reserpine (20 µg/mL). The growth inhibition was compared to the growth in PPLO medium alone. In an independent experiment, we verified that reserpine (20 μg/mL) alone had no effect on mycoplasma growth (Figure 3). The bacteriostatic effect of oxytetracycline at ¼ of the MIC was either marked, with a least two time points showing a log2 delay in cfu/mL (Ma Ag304, Ma L16160, Ma 5632, Mmc F10751 and Mmc F9545), or moderate (Mcc 10621, Ma Ag28, Mmc LC54) to nil (Mmc LC54, Ma Ag316, Mcc cap19). In experimental conditions when growth inhibition by oxytetracyline failed, the presence of reserpine had either no effect (Mcc cap19, which was the low MIC control) or a potential enhancer effect on the bacteriostatic property of tetracycline (AG 316), suggesting that an efflux was responsible for the absence of the oxytetracyline effect. When a growth inhibition by oxytetracycline was observed, two responses to reserpine were noted: (i) the growth inhibition remained unchanged (Mcc 10621, Maga 5632, Mmc 9545) or (ii) it was clearly enhanced (at least two time points with one log difference) (Ag 304, Ag 28, Mmc LC54), sometimes only after 48 h of growth (Ag16160, Mmc 10751).This demonstrates that the inhibition of efflux pumps clearly potentiated the bacteriostatic effect of oxytetracycline, which was stronger and longer-lasting in the presence of reserpine. Semiquantitative marks (Table 1) were assigned by dividing the maximum difference observed between the growth curves obtained with and without reserpine for Ag304 (3.5 log10) to delimit four levels of inhibition, ranging from less than 0.9 log10 (+/−) to more than 2.6 log10 (+++). The efflux capacities estimated using EtBr fluorescence and using growth inhibition kinetics were comparable for M. agalactiae strains Ag304, Ag28, L16160, Ag316, and for Mmc strains LC54 and F10751, but not for strains Ma 5632, Mmc F9545, and Mcc 10621. For the low susceptibility control (Mcc cap19), the results were not interpreted, because the oxytetracycline concentration used was potentially too low (0.0625 µg/mL at ¼ of the MIC) to observe any inhibition.4. DiscussionIn a previous study, we observed a moderate decrease in susceptibility to oxytetracycline in some Mycoplasma strains belonging to species involved in contagious agalactia syndrome of small ruminants, M. agalactiae, Mmc and Mcc [10]. Only Mput did not show any in vitro increased tetracycline MICs, which was recently confirmed [6]. Three main mechanisms, namely target protection, target modification, and efflux extrusion, are known to result in tetracycline resistance in bacteria. Target protection has so far never been observed in animal mycoplasmas [20]. Tetracycline resistance associated with target mutations in the 16S rRNA genes and in the rpsJ gene coding for the ribosomal protein S10 is well described for mycoplasma of animal origin, and such mutations were observed for field isolates of M. agalactiae, Mmc and Mcc harboring a MIC of oxytetracycline ranging from 2 to 8 µg/mL [10]. However, several isolates with similar increased tetracycline MIC showed no mutation in the rrs and/or rpsJ genes. We hypothesized that efflux mechanisms could be involved in the loss of tetracycline susceptibility for these specific isolates [10].Bacterial efflux pumps are classified into five families: the adenosine triphosphate (ATP)-binding cassette (ABC) superfamily, the resistance-nodulation-division (RND) family, the small multidrug resistance (SMR) family, the major facilitator superfamily (MFS), and the multidrug and toxic compound extrusion (MATE) family [35]. Four of them have been described as able to extrude tetracycline (SMR, MFS, RND and ABC families). Efflux in mycoplasma has seldom been evidenced except for fluoroquinolones in M. hominis [26] and in Mmc [24], and in M. pneumoniae for macrolides [27], all three studies pointing to an ABC-type efflux pump, able to also extrude unrelated compounds such as EtBr. A multi-drug efflux transporter of the MATE family has also been suggested in M. bovis though not formally demonstrated [36]. In general, efflux pumps in bacteria confer cross-resistance to a wide range of entities, including dyes, detergents, disinfectants, and antimicrobials [37]. Dyes such as EtBr are of particular interest as they can be easily used as markers of an efflux phenomenon. Other ways to identify efflux pumps include their inhibition and how it impacts on resistance to antimicrobials. The plant-derived alkaloid reserpine is known to inhibit both ATP binding cassette and major facilitator efflux transporters, although its exact mechanism of action is still unclear. Convergent results from different experiments are often necessary to demonstrate the presence of efflux pumps with certainty. For instance, in Staphylococcus aureus, the reserpine screen failed to identify overexpression of one or more MDR efflux pump genes, as shown by qRT-PCR [38].In this study, we assessed the presence of an efflux in field isolates of mycoplasmas involved in CA syndrome using two different EtBr-based assays, together with a growth inhibition experiment with oxytetracycline in the presence or absence of reserpine. The EtBr agar plate method was first assessed as it allows a fast, easy screening of many bacterial isolates. However, for some of the Mycoplasma isolates tested, the results were not always interpretable owing to either a growth inhibition of mycoplasma colonies in the presence of EtBr in the agar medium (e.g., the presence of EtBr, at 0.4 µg/mL in the agar plate, resulted in a complete growth inhibition for 4 out of 6 M. putrefaciens isolates) or to a fluorescence too low to be readable. The last effect could result from (i) a partial growth inhibition limiting colony development and/or (ii) a default in the accumulation of EtBr within the cells, because Mycoplasma colonies were grown on a complete medium containing energy sources that could potentiate efflux systems in the first 72 h of incubation. The second hypothesis was supported as transfer of plates at 4 °C for a further 24 h incubation blocked the efflux, resulting in a higher fluorescence than their counterparts maintained et 37 °C. This phenomenon was marked for 12 isolates and was more moderate for 6 other isolates, whatever their species, suggesting a temperature-sensitive efflux system.A more dynamic EtBr-based assay relying on efflux kinetics in PBS was also used. Unlike the agar plate method, no cell growth was expected in the presence of EtBr (as the contact time was only 30 min) and EtBr accumulation within the cells was maximized by using reserpine to inhibit efflux. This helped to harmonize the quantity of EtBr within the cells before starting the observation of the EtBr release time and thus the fluorescence decrease. With this method, all our isolates gave readable results and showed an efflux of EtBr energized by adding pyruvate, a known energy source for mycoplasmas. The efflux was slower when the temperature went from 37 °C to 25 °C, but it varied widely between strains within each species. We failed to establish a coherent link between the efflux level and the tetracycline MIC values. The clearest example was that of strain M. capricolum subsp. capricolum Cap19 with an oxytetracycline MIC of 0.25 µg/mL, but a marked EtBr efflux in both agar plates and the fluorometry assays (Figure 1 and Figure 2).Taken together, the EtBr experiments evidenced efflux mechanisms in mycoplasma species involved in CA, with a high inter-strain variability, not correlated to the overall oxytetracyline susceptibility.Potentiation by adding reserpine of growth inhibition due to oxytetracycline was assessed for nine representative isolates selected for their increased MICs (2–8 µg/mL) in comparison with a susceptible Mcc control isolate (Mcc cap19, MIC of 0.25 µg/mL). For three isolates, adding of reserpine had a low to nil effect on growth inhibition (Mcc 10621, Maga 5632, Mmc 9545). However, isolates Ma 5632 and Mmc 9545 showed a capacity to extrude EtBr, suggesting that there might be some specificity in the entities transported by the efflux pumps or that some pumps are not inhibited by reserpine. Of the six remaining isolates (4 Ma and 2 Mmc), reserpine enhanced the growth delay due to oxytetracycline, with different intensities between isolates and sometimes time-dependently (isolates Ag16160 and Mmc F10751). Interestingly, the effect of reserpine in our experimental conditions was maximum with isolate Ag304, with an MIC of 2 µg/mL and no rrs gene mutation. Approximately half the effect was observed for other isolates of MIC 2 to 8 µg/mL, with (Ag316, Mmc LC54, Ag28), or without (Ma 16160, Mmc 10751) rrs or rpsJ mutations. Our observations suggest that the efflux system in mycoplasma allows tetracycline resistance up to 4 µg/mL and might facilitate the acquisition of a target mutation for higher resistance. This is consistent with the fact that the M. putrefaciens species, which has remained so far mostly susceptible to tetracycline [6], exhibits no or a weak efflux. On the experimental basis proposed here, it would be interesting to further test the contribution of efflux in other Mycoplasma species with high tetracycline MICs and the corresponding mutations in the target genes, such as M. bovis.Early studies suggested that without a rigid peptidoglycan-based cell wall or actin-based cytoskeleton structure, mycoplasma cells could regulate their cell volume and adapt to different osmotic environment through active solute extrusions [39]. The effect would be to osmotically balance the colloid osmotic effect and work against osmotic lysis. Glucose-enhanced pumps were notably described as extruding Na+ ions. Hence, the efflux might rely on a mechanism not specific to antimicrobials and very variable in expression between strains in time, depending on the nutritional environment. Its contribution to resistance in vivo, with very variable environments depending on the colonized body niche remains to be demonstrated.5. ConclusionsWe evidenced potential efflux mechanisms in mycoplasma species involved in CA, with a high inter-strain variability. The efflux efficacy was not always comparable for the ethidium bromide dye versus oxytetracycline and was not always correlated to oxytetracyline susceptibility, suggesting non-specific mechanisms. The contribution of efflux pumps to the general antimicrobial resistance in Mycoplasma spp. now needs to be studied.

animals : an open access journal from mdpi

10.3390/ani13081340

PMC10135258

European sea bass is a species with high economic and societal value in the Mediterranean due to its intensive use in aquaculture. However, it is a species characterized by high cortisol levels that show high variation. The present systematic review and meta-analysis collected and examined all the published data on circulating cortisol in this species. The aim of the study was to analyze all published values in order to provide normal values and ranges of plasma cortisol in this species, both in basal and post-acute stress conditions. Results revealed a very high between-study heterogeneity, while it also calculated the pooled levels of cortisol and their confidence intervals for both basal and post-stress conditions. Moreover, results were analyzed based on various parameters that can potentially affect cortisol levels, including technical, such as assay type and rearing unit, as well as biological, such as body size and anesthesia, influences.

Background: European sea bass is a species characterized by high and dispersed cortisol levels. The aim of the present study was to analyze all published data on basal and post-acute stress cortisol levels in this species. Methods: For this systematic review and meta-analysis the Web of Science and Scopus databases were searched for papers reporting plasma or serum cortisol levels in E. sea bass, without language or date restrictions. Data were extracted directly for the reported results and were analyzed separately for basal and post-acute stress levels, as well their standardized mean differences (SMD) using random-effects meta-analyses. Results: Of 407 unique records identified, 69 were eligible. Basal cortisol levels had a pooled effect of 88.7 ng mL−1 (n = 57), while post-acute stress levels were 385.9 ng mL−1 (n = 34). The average SMD between basal and post-stress was calculated to be 3.02 (n = 22). All analyses had a high between-study heterogeneity. Results for basal and post-stress levels were affected by the assay type and anesthesia prior to blood sampling. Conclusions: Cortisol levels in E. sea bass are higher than most studied fish species and display large heterogeneity. Application of stress led to elevated cortisol levels in all studies examined. In all cases, sources of between-studies heterogeneity were identified.

1. IntroductionCortisol is the major stress hormone in fish [1,2,3]. It is the final product of the action of the Hypothalamus–Pituitary–Interrenal tissue (HPI) axis, and it has been reported to respond with increased concentrations to various types of stress in order to regulate stress responses [1,2,3]. Apart from controlling the stress responses, cortisol is a regulatory hormone for both metabolism and osmoregulation in fish [3]. Therefore, it is a crucial hormone in the physiology and biology of fish.European sea bass, Dicentrarchus labrax, is a fish species with high economic value due to the fact that it is one of the main marine aquaculture fish species in the Mediterranean. Although widely cultured, this species shows high cortisol responses to stress compared to other species widely cultivated in the Mediterranean, such as gilthead seabream, Sparus aurata, and meage, Argyrosomus regius [4,5], as well as high variation in cortisol concentrations in basal (pre-stress) and post-stress conditions [6]. A decade ago, Ellis et al. (2012) [6] reported that a high between-study variation in the basal and post-stress cortisol concentration can be observed in this species, proposing five possible sources of variation. Recent research has shown that most of these sources indeed can add variation, while other factors remain untested still (Table 1).1.1. Rationale for Meta-AnalysisAs discussed earlier, cortisol variation in E. sea bass is large, both in the same study (within population [8]) and between different studies. This makes the generalization of conclusions based on a single measurement impossible, since under such conditions it is hard to define accurate reference values. However, cortisol measurement is an important indicator of the physiological status of fish in terms of stress, osmoregulatory, and metabolic regulations [2]. A meta-analysis of data from different independent studies can provide a quantitative statistical way to combine their results. Especially in cases with data that show high heterogeneity, such as cortisol in E. sea bass, a meta-analysis using random effects models allows for important conclusions to be drawn. Moreover, certain meta-analysis statistical tools provide the means to incorporate other effects, such as environmental (husbandry, water quality etc.), biological (body weight), and technical (method used to quantify cortisol), in the analysis. For instance, the assay type used to measure cortisol in each study (ELISA, RIA, HPLC) can be included in a sub-group meta-analysis to provide information on whether the assay type affects the outcome of the study.1.2. ObjectivesThe objective of the present study was to provide normal values and ranges of plasma cortisol in E. sea bass, both in basal (i.e., without experimental exposure to any stressors) and post-acute stress conditions, as well as quantify the standardized mean difference between basal and post- stress (both acute and chronic) cortisol levels. Moreover, this study aimed at investigating the effects of (1) cortisol measurement assay type, (2) type of rearing unit, (3) use of anesthesia during the blood sampling, (4) water parameters such as temperature, salinity, dissolved oxygen, pH, (5) fish body weight, (6) stocking density, and, in the cases of post-stress levels, (7) the time after stress that the blood sampling took place.2. Materials and MethodsThe Preferred Reporting Items for Systematic reviews and Meta-Analyses Statement (PRISMA) guidelines to plan, implement, and report this systematic review and meta-analysis have been followed in this study [25]. The PRISMA checklist is available in Supplementary Table S1.2.1. Identification of StudiesThe databases Web of Science and Scopus were assessed using the search terms (“European sea bass” OR “sea bass” OR dicentrarchus OR labrax) AND (cortisol OR glucocorticoid OR corticosteroid) to find peer-reviewed articles reporting cortisol levels in E. sea bass, until the date of search (7 March 2022). Figure 1 shows a flow diagram that summarizes all stages of the systematic review process, including the numbers of studies identified at each stage and any reasons for exclusion. This workflow has not been published and no protocol was prepared.2.2. Eligibility CriteriaOut of the 407 research items retrieved from the database search after duplicate removal, the criteria used for screening the articles were that: (Scr1) the study concerned E. sea bass, (Scr2) examined fully developed fish, therefore excluding larvae, and (Scr3) reported plasma, or serum cortisol levels with their concentration. At this stage, 330 research items were excluded, and the remaining 77 were subjected to more detailed inspection for eligibility, including only studies that: (Eli1) provided information on the number of animals used in each experimental group, (Eli2) provided information regarding the dispersion of the data, either as Standard Deviation (S.D.) or as Standard Error of the Mean (S.E.M.), (Eli3) used the individuals and not the tanks as experimental units and, therefore, presented the results as an average of individuals and not tanks, and (Eli4) it was possible to attribute the data to control or post-stress conditions. After this stage, 8 research items were excluded ([16,26,27,28,29,30,31,32]; Table S2), and the remaining 69 items were used in the meta-analysis.2.3. Data ExtractionAll qualitative information of the studies, such as assay type, rearing system, fish size, and water parameters, were retrieved from the text and tables of the research items. Data from the research items were extracted either from tables or graphs reporting the mean value ± S.D. or S.E.M. In the latter case, the software ImageJ was used for image analysis based on measuring the length of the y-axis and the length of the projections of the cortisol mean ± S.D. or S.E.M. In cases where various post-stress time points were presented, the time of the peak response was used. Moreover, when more than one control group was presented and in order not to violate the assumption of independency of data for the meta-analysis by analyzing data from the same research item more than once, the control groups were pooled. All cortisol data were analyzed as ng mL−1. The vast majority of the studies reported this measuring unit (62 out of 69 studies) or its derivatives ug/dL, ug/mL and ng/dL (4 out of 69 studies). There were 3 studies that reported the results as nmol L−1 (nM), and their data were converted to ng mL−1 by multiplying with the conversion factor 0.36245.2.4. Coding of DataEach study was coded for quantitative and qualitative data. Quantitative data included (1) fish body size, (2) water temperature, (3) dissolved oxygen, (4) water salinity, (5) water pH, and (6) stocking density. The qualitative data included (7) assay type used to measure cortisol, (8) rearing system, (9) anesthesia type, and, (10) in cases of post-stress samplings, total time between the application of stress and sampling, defined as classes (e.g., 0–30 min, 30–60 min etc.).In cases where the quantitative data were reported as a range, the mean value was calculated. For instance, in the study by Tintos et al. 2006 [33] where the body weight was reported to range between 15–20 g, the weight was recorded as the mean between the two values, i.e., 17.5 g.2.5. Statistical AnalysisAll statistical analysis was performed in RStudio [34], using the packages “meta” [35], “dmetar” [36], and “tidyverse” [37]. Since considerable between-study heterogeneity was expected, a random-effects model was used to pool effect sizes. The heterogeneity variance τ2 was calculated using the restricted maximum likelihood estimator, while the confidence interval around the pooled effect was calculated using the Knapp–Hartung adjustments [38].Specifically, for the basal and post-stress analysis, the pre-calculated effect sizes were analyzed under the “metagen” function, while analysis of the standardized mean differences (SMD) and their 95% confidence intervals (CIs) between basal and post-stress levels were analyzed under the “metacont” function, using Hedges method to calculate the SMD due to the small number of subjects in most studies [35,39]. Subgroup analysis for the qualitative treatment data was performed using the “byvar” argument and was based on calculating different τ2 for each subgroup, and subgroups were tested for significant differences using the Q test. Meta-regression on the quantitative influence data was performed using the “metareg” function and risk of bias was assessed by Egger’s test using the “metabias” function. The forest plot was created using the “forest” function.3. Results3.1. Study CharacteristicsThe final outcome of the literature search was 69 peer-reviewed studies that concerned circulating cortisol levels in E. sea bass and provided sufficient information on the number of animals used and the population mean values and dispersion (Figure 1). Out of these, 35 studies reported only basal, 12 only post-stress, and 22 both basal and post-stress cortisol concentrations, thus making a total of 57 studies reporting basal levels and 34 studies reporting post-stress levels. Summary characteristics for the research items included in the meta-analysis are presented (Table 2). All 69 studies had reported the weight of the fish.3.2. Basal LevelsFrequency distribution analysis on the basal cortisol levels resulted in a positively skewed distribution (Figure 2a). The pooled effect size was calculated at 88.7 ng mL−1 [95%-CI: 65.5–109.8], with a high between-study heterogeneity (Table 3). Egger’s test showed that no significant publication bias was present in the dataset (intercept = 13.5, df = 55; p = 0.525).Subgroup analysis based on the assay type revealed differences between groups. Due to the low number of studies using HPLC and chemiluminescence/electrochemiluminescence assays, an analysis including only the studies using ELISA and RIA assays was performed and showed that the difference between them was significant (Q1 = 5.36; p = 0.021), being higher in ELISA than RIA. The use of anesthetic also had a significant effect on cortisol (Q5 = 12.69; p = 0.026), while no differences were observed between the rearing unit systems (Q3 = 1.41; p = 0.703).Meta-regression analysis between effect sizes and quantitative characteristics showed that none of the examined parameters, i.e., fish body weight, water temperature, dissolved oxygen concentration, salinity, pH, and stocking density, affected the results.3.3. Post-Stress LevelsFrequency distribution analysis on the post-stress cortisol levels resulted in a positively skewed distribution (Figure 2b). The pooled effect size was calculated at 385.9 ng mL−1 [95%-CI: 310.8–460.9], with a high between-study heterogeneity (Table 4). Egger’s test resulted in a marginally significant publication bias in the dataset (intercept = 6.4, df = 32; p = 0.047).Subgroup analysis based on the assay type revealed differences between groups (Q1 = 7.76; p = 0.005), being higher in ELISA than RIA. The use of anesthetic also had a significant effect on cortisol (Q3 = 17.13; p < 0.001), excluding the “none”, “ice/cold water”, and “decap/blow to head” groups from the analysis due to the small number of studies in each group. On the other hand, no differences were observed between rearing unit systems (Q2 = 5.36; p = 0.069), excluding the ponds due to their small number. Time after stress, excluding the “>240” group due to the low number of studies, showed a significant effect (Q4 = 9.99; p = 0.041). Finally, none of the quantitative parameters was related to the effect sizes when the respective meta-regression analysis was performed.3.4. Standardized Mean Difference between Basal and Post-Stress CortisolTo assess the difference between basal and post-stress cortisol, 22 studies that included both pre- and post- exposure to acute stress data were used. The pooled SMD was calculated to be 3.02 [95%-CI: 2.46–3.58] (Table 5). The between-study heterogeneity was lower than the ones in the previous analysis, but still significantly large (I2 = 81.7%, τ2 = 1.12). Egger’s test using the Pustejovsky and Rodgers modification to avoid false positive results that arise with the classical Egger’s test on SMDs [94], resulted in a significant publication bias in the dataset (intercept = 4.7, df = 20; p = 0.010).It is obvious that stress had an overall effect on cortisol, a result that was observed in every study (Figure 3). The high heterogeneity between the studies can also be observed (Figure 3).Subgroup analysis based on the assay type revealed no significant differences between groups (Q1 = 0.03; p = 0.870), being similar between studies using ELISA rather than RIA assays. No differences were observed between rearing unit systems, excluding ponds due to the low number of studies in this group (Q2 = 0.11; p = 0.945). The same was true for the use of anesthetics, analyzing only phenoxyethanol and MS222 (Q1 = 0.65; p = 0.421). Finally, time after stress had no significant effect on cortisol response (Q2 = 1.56; p = 0.458) excluding the “0–30”, “120–240”, and “<240” due to low number of studies. Finally, meta-regression analysis between effect sizes and quantitative characteristics showed that none of the examined parameters, i.e., fish body weight, water temperature, dissolved oxygen concentration, salinity, pH, and stocking density, affected the results.4. DiscussionThe study of circulating cortisol concentration in E. sea bass is intriguing due to the fact that this species shows high basal and post-stress levels of cortisol, as well as high variation both within the same population and between different studies [4,6,8]. In fact, out of the studied teleost species, E. sea bass is among the ones with the highest reported cortisol levels, together with the chub, Leuciscus cephalus, the latter having been characterized as a cortisol resistant species [95]. This high stress susceptibility has been suggested to be co-responsible for disease outbreaks in this species [96].In this context, a systemic review of the published cortisol levels of E. sea bass could assist in better understanding whether cortisol levels in this species are indeed high as well as to define possible sources of variation between studies. This meta-analysis led to the conclusion that a high between-studies heterogeneity exists in both basal and post-stress concentrations. The reported basal concentrations were calculated to have a pooled effect size of 88.7 ng mL−1 with an 95% confidence interval between 65.5–109.8 ng mL−1, while post-stress concentration had a pooled effect size of 385.9 ng mL−1 [95%-CI: 310.8–460.9].Subgroup meta-analysis revealed some interesting findings. However, these findings should be interpreted with care since there are constraints in the use of subgroups meta-analyses. The most important ones are the small number of studies in a subgroup and the high between-study heterogeneity, since both reduce the statistical power of the analysis. In order to reduce the “small number of studies” effect in the current analyses, subgroups with lower than three studies were excluded from subgroup meta-analysis. On the other hand, although high heterogeneity was observed in the current study, there are no available tools to mitigate its effect on the subgroups analysis.Having the above constrains in mind, one of the major factors that seemed to affect the heterogeneity was the assay type. In both basal and post-stress levels, the pooled effect size of studies using ELISA assays was significantly higher than studies using RIA assays. Ιt is generally accepted that RIA has a higher efficiency in measuring cortisol compared to ELISA. In many studies, RIA assays are considered as more accurate when it comes to the analysis of fish cortisol [97], human salivary cortisol [98], as well as mice [99] and bird [100] corticosterone. However, there are also studies in mammals that show equal results between ELISA and RIA [101], or even better performance in the ELISA assays [102]. Therefore, it is difficult to definitely conclude which assay type is more accurate in reporting cortisol levels, but the current study supports the notion that the cortisol assay type should be taken into careful consideration when designing a study and when interpreting the results. On the other hand, when the standardized mean differences between basal and post-stress cortisol levels were analyzed, no difference between assay types was observed. This result indicates that although ELISA assays tend to over-estimate cortisol levels, they do so in a similar manner in basal and post-stress concentrations. In other words, both assay types record the magnitude of the response in the same way although ELISA overestimates the absolute values.The rearing system, on the other hand, seemed not to affect cortisol levels. The most commonly used systems were the open-flow and the RAS, consisting of approximately 3/4 of the total number of studies. To the best of our knowledge, there are no published studies in E. sea bass to directly compare fish welfare between these rearing systems, though the effects of increased stocking density seem to be the same in fish reared in RAS and open flow systems [55,78]. What has been shown to affect welfare in this species is the size of the rearing unit, in either open flow tanks in larval stages [103] or sea cages during on-growing [76].Regarding means of anesthesia, most studies used chemical anesthetics, mainly phenoxyethanol, followed by MS222. In both basal and post-stress conditions anesthesia treatment significantly affected the results, showing lower values in decapitated or percussively blown-in-the-head fish. It is well known that E. sea bass is a species with a rapid cortisol response [75], and, therefore, immediate killing does not allow cortisol to rise. In that way, minimum cortisol levels are reported when using this method. In basal conditions, the highest levels were reported in fish that were not under anesthesia, i.e., they were conscious, during blood sampling. When anesthetics were used, phenoxyethanol and MS222 resulted in lower levels than the other anesthetics indicating that, when E. sea bass is sampled for cortisol levels under anesthesia, it is preferable to use one of the aforementioned anesthetics. However, it should be noted that studies assessing direct comparisons between conscious and chemically anesthetized fish, using either phenoxyethanol or clove-oil [59] and MS222 [65], have not reported such differences.In terms of magnitude of the stress response in relation to the time after stress, it is known that, in E. sea bass, cortisol starts to rise at least 6 min after the application of stress [75], reaching maximum levels at 60 to 120 min post-stress when recovery starts to take place [4,7,14]. Grouping of studies based on the post-stress time at which fish were sampled revealed a similar outcome although it should be noted that the differences were not significant. However, this synthesis of results reflected the typical, more-or less, time-course cortisol response of this species [4,5,14,75], even though different stressors, in terms of nature, intensity, and duration were used.It is acknowledged that there are some limitations in the conclusions of the current meta-analysis. As mentioned before, the first is due to the high between-studies heterogeneity. This is a result of various reasons, including the different aims of the studies, the different assays used, the rearing methods, temperature, and so on. The second lies to the fact that E. sea bass responds very fast to handling [75], and it is, therefore, difficult to ascertain that the reported basal levels have been obtained under similar sampling stress between studies. This is similar to the source of variation #4 proposed by Ellis et al. [6] (Table 1). Third, there was a scarcity of information in environmental data that could affect the cortisol response, such as water temperature, salinity, pH, and dissolved oxygen concentration. Finally, the circadian rhythm [7,19] the seasonality [5,18], and the sex of the fish [104] are additional potential sources of variation in cortisol levels which are hardly taken into consideration—and subsequently not reported—in most published studies, and, therefore, the current study could not include them in the analysis.5. ConclusionsIn conclusion, taking into consideration the limitations discussed above, the current meta-analysis examined 69 studies and calculated a pooled effect for basal and post-stress cortisol levels for E. sea bass. A high between-studies heterogeneity was recorded, with the factors assay type and anesthesia affecting cortisol levels and adding variance to the results. Moreover, a significant effect of acute stress on cortisol levels was observed in all studies examined. On the contrary, no association between cortisol and fish body weight or environmental conditions such as water temperature, dissolved oxygen, salinity pH, stocking density, or the rearing unit was observed. Finally, although it was not possible to directly test for genetic differences between fish, seasonality, circadian rhythms, and sex due to the lack of data in the published studies, the high between-study heterogeneity indicates that these factors may be additional factors causing variation between studies in the examined species.

animals : an open access journal from mdpi

10.3390/ani11123601

PMC8698060

Microalgae synthesize a wide variety of bioactive compounds, such as polysaccharides, PUFAs, and pigments. In addition, microalgae are known for positively influencing growth and intestinal parameters in birds and thus have been contemplated as potential feed additives. Nevertheless, several species widely used in aquaculture have not received much attention in terrestrial animal nutrition. Therefore, in this investigation, we used biomass from Tysochrysis lutea, Tetraselmis chuii, and Porphyridium cruentum to evaluate its safety and usefulness for improving intestinal architecture, body weight, and selected meat quality parameters. Dietary administration of the different biomasses showed positive effects regarding intestinal architecture, which were associated with the observed body weight. Furthermore, the thawing weight loss of fillets decreased after supplementation with T. chuii. Hence, these outcomes suggest that dietary microalgae might be considered as a promising bio-friendly alternative for feed additive production.

This pilot investigation aimed at studying the feasibility of using a low dose (0.2%) of dietary microalgae as a means of improving intestinal morphometry, body weight, and selected meat quality parameters in broilers. A total of 72 one-day-old ROSS 308 male chicks were randomly separated into four groups; three experimental pens in which the birds were fed with biomass from Tysochrysis lutea, Tetraselmis chuii, and Porphyridium cruentum over 30 days and a control group. T. chuii and P. cruentum had a positive effect with regard to body weight. In treated animals, duodenal and ileal sections showed characteristic tall and thin villi, with serrated surfaces and goblet cell differentiation. In both sections, values of the villus-height-to-crypt-depth ratio were increased by microalgae ingestion. The thawing weight loss of fillets was reduced in T. chuii-fed animals. The positive effects exerted by T. chuii and P. cruentum on intestinal architecture were associated with the improved body weight. Arguably, these outcomes exhibit the potential of using these species to enhance growth performance in broiler chickens by promoting gut homeostasis and thus nutrient absorption.

1. IntroductionIn animal husbandry, antibiotics are principally used to prevent, control, and treat diseases. However, they have also been extensively used as growth promoters [1,2]; this practice has been banned in Europe and the U.S. [3,4] but is still widespread in other regions, especially in rural areas, where farmers tend to depend more on antibiotics. Various countries lack legislative measurements to control antibiotic use. Moreover, farmers are not appropriately trained due to the lack of educational programs. Countries with deficient and ill-administrated health systems will suffer the greatest problems when coping with antibiotic resistance [5,6]. Diverse alternatives for promoting animal growth and well-being have been studied, especially in poultry production. For instance, probiotics are known for enhancing growth performance and immune parameters [7,8,9,10,11]. Likewise, extracts from various plants have proved useful for such purposes [12,13,14]. Algae administration has also yielded positive outcomes with regard to animal performance and immunity [15,16].Algae constitute a heterogeneous group of photosynthetic organisms present in a variety of sizes, shapes, and colors (which include red, green, and golden brown). They are divided into two principal groups: macroalgae, often referred to as seaweeds, and microalgae [17]. Algae have been incorporated into fish feed in farms and hatcheries. In particular, microalgae are considered important as they synthesize valuable compounds, such as polyunsaturated fatty acids, polyphenols, essential amino acids and proteins, pigments, and various polysaccharides [18]. In fact, microalgae-enriched feed has proved useful for promoting growth performance in broiler chickens before [19,20,21,22]. Nonetheless, most experiments have made use of Chlorella sp. and Arthospira (ex. Spirulina) [18,23]. Effects of other, less common species have also been reported, including Schizochytrium and Amphora sp. [24,25]. In addition, the effects of microalgae administration on intestinal morphometry have not been extensively documented. One study showed that feed enriched with 2.5% Chlorella improves intestinal integrity by positively influencing villus height and crypt depth [17].Various species produced for aquaculture have yet to be tested as components of poultry feed. In golden-brown microalgae, Tysochrysis lutea belongs to a group of auspicious marine haptophytes with high lipid content and the production of polyunsaturated fatty acids. Moreover, this species is a rich source of the pigment fucoxanthin, which has nutraceutical and pharmaceutical applications [26]. In green algae, Tetraselmis chuii is known for its straightforward cultivation and high nutritional value. This microalga accumulates high quantities of important compounds, such as pigments, proteins, and polyunsaturated fatty acids [27]. Among red microalgae, Porphyridium cruentum is considered a promising organism for the food industry because it can synthesize and secrete high amounts of sulfated polysaccharides as well as lipids and polyunsaturated fatty acids [28], which are recognized as important antioxidant, anti-inflammatory, and cytotoxic agents [29]. Consequently, in this pilot study, we aimed at evaluating the effects of dietary microalgae mainly on intestinal morphometry but also on body weight and selected meat quality parameters in broiler chickens, using biomass from species traditionally used in aquaculture: T. lutea, T. chuii, and P. cruentum.2. Materials and MethodsThe study was carried out in the Experimental Centre for Animal Research of the Veterinary Medicine Faculty, Central University, Ecuador. The facility is located in the parish of Uyumbicho, situated 23 km southeast of Quito. Experiments were conducted following the guidelines for poultry management provided by the Agency for the Regulation and Control of Phytosanitary and Animal Health (AGROCALIDAD, technical resolution n. 0017). The Ethics Committee on the Use of Animals in Research and Teaching of the San Francisco de Quito University (USFQ) revised and approved the related protocols (reference number: 2020-008).2.1. Microalgae BiomassThe freeze-dried microalgae powder of T. lutea, T. chuii, and P. cruentum was purchased from Necton S.A., Olhão, Portugal (https://necton.pt/ accessed on 2 June 2021). Tysochrysis lutea belongs to the phylum Haptophyta; these organisms are known to contain both chlorophylls a and c. The presence of the carotenoid accessory pigment fucoxanthin is partly responsible for their often-golden-brown color [26]. This biomass possesses crude protein (40%), crude fat (6%), and crude ash (23%). T. chuii belongs to the phylum Chlorophyta; these organisms are characterized by the presence of chlorophylls a and b, along with the carotenoids violaxanthin, antheraxanthin, zeaxanthin, neoxanthin, and lutein [27]. This biomass possesses crude protein (35%), crude fat (5%), and crude ash (30%). P. cruentum belongs to the phylum Rhodophyta, which contains chlorophyll a as well as α- and β-carotene, lutein, and zeaxanthin. Additionally, these organisms produce water-soluble pigments known as phycobilins [28,29]. This biomass possesses crude protein (35%), crude fat (5%), and crude ash (31%). The components of these microalgae suggest a potentially high nutritional value for birds.2.2. Experimental SetupA total of 72 one-day-old male broiler chicks (ROSS 308) were weighed (50 ± 5.8 g on average) and then randomly divided into four groups. Each experimental group was set up in an individual pen (3 m × 3 m), divided into nine subgroups (1 m × 1 m). Each subgroup contained two birds; the subgroups were housed separately from the others. For body weight measurements, both animals were weighed and the averaged values were used for analyses. For sampling, one chicken was selected per subgroup (N = 9). The subgroup was considered the experimental unit, as each of the screened birds was independently allocated to treatment conditions and experimental intervention and could not influence others on the measured outcome [30]. All chickens were fed basal diets free of probiotics, antibiotics, or anticoccidiostats with nutritional components of broilers for starter (1–11 days) and finisher (12–30 days) diets (Table 1). Feed and water were available ad libitum throughout the entire experiment, which lasted 30 days. Animals were sacrificed at this age as the main objective of the research was to assess the influence of treatments on intestinal morphometry before the onset of critical growth stages (weeks 5–6). This study was not aimed at determining growth and meat quality parameters in detail. Nonetheless, the chicken body weight was registered as a mean to monitor the health of animals during the trial. Moreover, cooking and thawing loss of fillets were also determined. The first group were fed with pelleted basal diet and thus considered the control. Birds from groups 2–4 were fed an enriched diet with freeze-dried microalgae powder. All experimental diets were provided with the same dose of biomass (2g/kg); group 2 (TL) was supplemented with T. lutea, group 3 (TC) with Tetraselmis chuii, and group 4 (PC) with Porphyridium cruentum. Chickens were raised on a floor covered with hardwood shavings. The ambient temperature was maintained between 30 and 32 °C during the first week and gradually decreased by 3 °C every week, i.e., on day 7 (27–29 °C), 14 (24–26 °C), 21 (21–23 °C), and 28 (19–21 °C). The relative humidity was kept between 50 and 60%. During the first seven days, the birds were exposed to a regime of 23 h of light (intensity 30–40 Lux) and 1 h of dark, which was followed by one regime of 21 h of light (intensity 5–10 Lux) and 4 h of dark, until the end of the experiment. Housing and environmental conditions abided by the ROSS Broiler Management Guide [31]. 2.3. Body WeightAll animals were weighed, and the averaged values were used for analyses. This was carried out on day 5 and then on days on which the rearing temperature was modified (7, 14, 21, and 28). Additionally, weights were registered on days 29 and 30.2.4. Morphometrical AnalysesOn day 30, all animals were weighed and one bird per subgroup was electrically stunned and euthanized by bleeding. From the intestine, the loop of the duodenum and the tract before the ileocolic junction were collected. Segments were fixed with a 10% formalin solution for 48 h. Then, samples were dehydrated by serial washes using ethyl alcohol (70–100%), diaphanized using xylol, and finally embedded in paraffin blocks. These blocks were sliced in three longitudinal sections of 5 μm thick blades using a rotary microtome (Leica RM2235, Wetzlar and Mannheim, Germany) and stained with hematoxylin and eosin (HE staining). Images were captured and processed using the Motic Images Plus 2.0 (Motic, Hong Kong, China). Villus height and crypt depth were determined per segment on uninjured villi, six at least, this was performed four times for a total of 24 readings per animal. The villus choice was based on the presence of an intact lamina propria. The villus-height-to-crypt-depth ratio was calculated as hitherto described [8]. The number of goblet cells per 100 intestinal epithelial cells in six intact villi was determined using the aforementioned program, as described previously [32].2.5. Meat Quality ParametersAt the end of the experiment, nine animals per group were sampled for their meat quality and six 10 g fillets from the Pectoralis major of each euthanized bird were dissected and cleaned. For cooking loss (three fillets), samples were placed in thermotolerant plastic bags and put in a water bath until 70 °C was reached; this was maintained for about 10 min. Samples were cooled on ice to 5 °C and then weighed. For thawing loss (three fillets), breast samples were dried, trimmed, and stored at –18 °C. A week later, thawing of the frozen fillets was carried out at 5 °C for 1 day and then the final weight was determined. The differences between initial and final weights were considered the values of loss. The values of fillets were averaged for statistical analyses. These procedures were performed as previously described [24].2.6. Statistical AnalysesFirst, Shapiro–Wilk’s test and Levene’s test were used for assessing normality and homogeneity of variance, respectively. For normally distributed and homoscedastic data, a one-way analysis of variance was used for determining significant differences between experimental groups, along with a Tukey post hoc test. For normally distributed and heteroscedastic data, Welch’s ANOVA and Welch’s t-test were applied. For non-normally distributed and homoscedastic data, Kruskal–Wallis test and Mann–Whitney U test (Wilcoxon rank sum test) were used. In the latter case, since the distribution is non-symmetrical, the mean is affected as a measure of the central tendency of distribution. Thus, medians were used because they reflect the center of distribution more appropriately in such conditions. Pearson’s r correlation coefficient was performed to identify potential relationships between the indicators. Analyses were carried out in MATLAB® version 9.9.9341360 (MathWorks, Natick, MA, USA) (R2016a).3. Results3.1. Effects of Diet on Body WeightCompared to control conditions, no significant differences regarding body weight were observed during the first two weeks in microalgae-fed groups. From day 28, chickens of the T. chuii group were heavier than those fed a basal diet. At the end of the experiment, administration of T. chuii and P. cruentum proved to significantly increase chicken body weight compared to control conditions, by 16% and 13%, respectively (Table 2). 3.2. Histological Findings and Small Intestine Morphometric MeasuresIllustrative micrographs of duodenum and ileum in 40× magnification are shown in Figure 1. Sections of duodenal and ileal segments of untreated birds exhibited short and thick villi with limited crypt development, whereas those of microalgae-fed animals appeared healthier, with tall and thin villi, along with active crypts with increased size and rows of enterocytes. In sections from the duodenum, tall and thin villi displayed serrated surfaces and broad tips related to multiplying enterocytes. In ileal sections, proliferation of the epithelium and goblet cell differentiation could be observed. Chickens fed with microalgae biomass displayed an increase in villus height and crypt depth in the duodenum, except for those supplemented with T. chuii. No differences were found regarding the villus-height-to-crypt-depth ratio (Table 3). In ileal sections, the height of villi was augmented in all experimental treatments, whereas only P. cruentum increased the crypt depth. A higher ratio of villus height to crypt depth was detected in all microalgae-enriched groups (Table 3). Pearson’s r coefficient analysis showed a positive correlation between the weight of chickens and the villus-height-to-crypt-depth ratio (0.977; p < 0.05). This suggests that the increment in body weight could be associated with healthier intestinal epithelia. In duodenal sections, goblet cells were found in higher numbers in microalgae-enriched groups than in control conditions, whereas in ileal sections, only P. cruentum induced a positive effect (Table 3).3.3. Meat Quality ParametersCooking loss of the not fully developed pectoral muscles was not altered by microalgae supplementation, but there was improvement in thawing loss observed in animals fed with T. chuii compared to the control, TL, and PC groups (Table 4).4. DiscussionMicroalgae are known to be a reliable source of nutrients and have long been exploited in aquaculture [18]. These organisms have also been integrated into animal nutrition, although to a lesser extent [18,23]. Previous studies have shown that the inclusion of dietary microalgae in feed could improve growth parameters in broiler chickens [20,21,22]. The majority of experiments regarding microalgae have been carried out using Spirulina platensis and Chlorella vulgaris [18,23], although effects of other, less common species have also been reported, including Schizochytrium sp. and Amphora coffeaformis [24,25]. Information regarding supplementation with T. lutea, T. chuii, and P. cruentum in broilers is scarce, with only one study demonstrating that dietary supplementation with Porphyridium sp. reduces feed intake although it does not affect body weight [33]. Therefore, we wanted to test the safety and effectiveness of these microalgae, commercially produced for aquaculture, in improving intestinal morphometry, body weight, as well as selected meat quality parameters in broiler chickens. T. chuii and P. cruentum are acknowledged for improving growth performance in shrimp and fish [34,35]. The present outcomes reveal that dietary administration of these species of microalgae had an overall positive effect on chicken body weight. For microalgae, published results suggest a biomass incorporation rate of 2%, or between 1 and 5%. However, such rates are not deemed suitable for mass application due to the costs [18]. Here, we used a dietary incorporation rate of 0.2%, as we wanted to test the effects of a lower dosage. Previous data showed an increase in body weight, compared to control conditions, in birds fed with 0.1% C. vulgaris (8%) and A. coffeaformis (7.4%) [24]. Similarly, dietary administration of 1% Chlorella raised the body weight by 3.5% [22]. These values are lower than those found in the present study, where 0.2% of T. chuii and P. cruentum augmented the body weight by 16% and 13%, respectively. Another study showed that supplementation (7%) with a product derived from Schizochytrium increased growth by 22% [36]. Evidently, the quantity of biomass used influences the extent of microalgal effects on growth. T. lutea administration has not led to an improvement in weight parameters in fish [37]. Similarly, in the present study, the group where the feed was enriched with T. lutea did not show any differences when compared with the control group. Inclusion of microalgae in feed has not always resulted in higher body weight average values in birds [24,25]. In any case, the positive results with regard to growth performance exhibited here were supported by morphometrical analyses of the small intestine. Macroalgae-enriched feed has proven useful for ameliorating intestinal architecture in broiler chickens [15,16]. However, the influence of microalgae administration on intestinal health has not been documented extensively. It has been demonstrated that inclusion of Chlorella in diets increases villus height and crypt depth in jejunal and ileal sections [38,39]. Another study showed an improvement in these parameters after administration of extracts enriched with A. platensis [40]. Our results are in line with the mentioned studies, as all treatments exerted a positive effect on intestinal integrity. In fact, T. lutea, Tetraselmis sp., and P. cruentum are known for improving intestinal morphology and health in fish [35,41]. In duodenal and ileal sections, villi were characteristically taller and thinner in microalgae-treated birds than in untreated animals, whereas crypt depth was only improved in chickens fed with P. cruentum. All microalgae species improved values of the villus-height-to-crypt-depth ratio in ileal sections. Deeper crypts have been associated with a swift regeneration of the villi, which increases their overall height [42]. Longer villi along with a higher ratio of villus height to crypt depth are considered crucial indicators of gastrointestinal health, as they are linked to a greater capacity of nutrient absorption [43,44]. Positive morphometry was correlated with an increase in body weight. Various studies have shown that the administration of probiotics ameliorates the integrity of intestinal epithelial cells, which leads to a more efficient absorption of nutrients, which ultimately improves growth performance [7,8,45,46]. The current outcomes demonstrate that dietary supplementation with the tested microalgae was beneficial for an improvement of intestinal epithelial architecture, which could have led to a more proficient intake of food and thus increased body weight. These physiological conditions would allow birds to enhance nutrient absorption during critical stages of growth (weeks 5 and 6).The goal of this preliminary study was to evaluate the safety and usefulness of using the screened species as feed additives. The biomasses used in this study are known to contain fat, protein, and oligosaccharides (Section 2.1). It has been shown that ω-3 fatty acids are capable of improving intestinal morphology in broilers [47]. T. chuii produces hexadecatetraenoic acid (C16:4ω-3), stearidonic acid (C18:4ω-3), and oleic acid (C18:1ω-9), while P. cruentum also accumulates ω-3 fatty acids, such as docosahexaenoic acid (C22:ω-6) and eicosapentaenoic acid (C22:ω-5) [47,48]; these compounds are known for their high nutritional value for poultry production [47]. As green algae, T. chuii contains xanthophylls such as lutein and zeaxanthin, along with β-carotene [27]; the latter two are known to increase the height of villi in the small intestine of broiler chickens [49]. Plant extracts containing lutein proved to activate intestinal nutrient absorption by increasing the surface of the villi [50]. Similarly, P. cruentum produces β-carotene, lutein, and zeaxanthin. These microorganisms accumulate various phycobilins, including phycoerythrin, which is responsible for their red color [28,29]. These compounds are known for exerting growth promoting effects and improving intestinal morphology in broiler chickens [51]. Oligosaccharides are reported to enhance intestinal structures [52]. In particular, Porphyridium sp. synthesizes important amounts of structural polysaccharides and exopolysaccharides. These compounds are recognized for inducing morphological modifications in the small intestine as well as augmenting the number of goblet cells in the mucosal layer of rats [53]. However, monogastric animals cannot digest such oligosaccharides, so they reach the colon and are fermented by a diverse group of microorganisms [54]. Microbial fermentation produces short-chain fatty acids (e.g., butyrate) that play an important role in intestinal epithelium development [55]. These are considered important sources of energy and are notorious for stimulating mucus production and secretion [56]. It could be argued that the results observed in this study maybe be related to the aforenoted compounds, which are known for stimulating intestinal and growth parameters. Further work must be emphasized on producing extracts for compound acquisition and concentration.The oligosaccharide side chains on mucins can interact with bacterial adhesins and therefore prevent their reaching and damaging the epithelium [56]. Goblet cells make part of the epithelial cells lining the villi. These cells synthesize the major component of mucus, mucin 2 (Muc2) [57]. Maintenance of the mucus layer of the intestinal epithelium is crucial for nutrient transport, protection, and lubrication [58]. Plant-derived compounds, probiotics, and prebiotics have proven useful for ameliorating intestinal histology, including goblet cell count, in untreated chickens and in chickens challenged with bacterial pathogens [12,13,59]. Here, we showed that microalgae administration promotes goblet cell differentiation in both duodenum and ileum. This seems relevant in the context of a bacterial infection, since goblet cells and their products make part of the first line of defense against undesired microorganisms. Therefore, these microalgae species could be applied with the aim of reducing the impact of pathogenic bacteria, which has been demonstrated for other algae in the context of a Salmonella infection [60,61]. Data on meat quality parameters showed that microalgae did not modify cooking loss, but T. chuii administration proved to reduce thawing loss compared to control conditions. These results are comparable to those of a previous study showing no differences in cooking loss but an improvement in thawing loss when using a microalgae-derived product [20]. Another report showed that S. platensis administration did not alter either parameter [62] whereas dietary (0.1%) C. vulgaris, S. platensis, and A. coffeaformis reduced cooking loss but did not alter thawing loss [24]. Results appear related to the species of microalgae used. In any case, dietary administration does not seem to have a negative impact on the analyzed parameters.Microalgae represent a novel area of relevance for animal production, especially with regard to nutrition and health. These photosynthetic organisms synthesize and accumulate several important compounds valuable for improving intestinal architecture. This is of prime importance, not only because nutrient absorption efficiency could be enhanced, but also because intestinal epithelium serves as a physical barrier against invasive pathogens. In addition, biotechnology has proved useful for producing important biomolecules by helping genetically engineer the plastid genome of the model alga Chlamydomonas reinhardtii. For instance, potential vaccines against common diseases such as the avian influenza virus and infectious bronchitis virus (IBV) have been produced in such a way [63,64]. This opens up the possibility of oral delivery of vaccines by enriching feed with dried microalgae, thereby reducing the requirements for purification and cold-chain transportation [65]. Undoubtedly, microalgae biotechnology will permit the design of bespoke strains with superior productivity that could be applied in animal nutrition.5. ConclusionsThe present outcomes suggest that the screened species, especially T. chuii and P. cruentum, have the potential to be used as feed additives in nutrition. Administration of dietary microalgae (0.2%) proved useful in ameliorating intestinal morphology and epithelial barrier; this was associated with improved body weight. Furthermore, T. chuii supplementation reduced thawing loss of dissected fillets. These results provide important insights into the current understanding of microalgae use in poultry production. However, further investigation must be carried out for refining our knowledge of the properties of these microorganisms and assessing the effects not only on animal welfare and nutrition but also on quality products.

animals : an open access journal from mdpi

10.3390/ani11082267

PMC8388511

Anecdotal accounts abound of pet dogs predicting their owner’s epileptic seizures by becoming attentive and by demonstrating attention-seeking behaviours, but no scientific study has investigated the veracity of these claims. Here, we explored this phenomenon, by assuming the presence of seizure-associated odours and then recording the reactions of a cohort of pet dogs to the emergence of such odours, apparently coming from their non-epileptic owners. Using two specially designed pieces of apparatus called the Remote Odour Delivery Mechanism (RODM), we separately delivered epileptic seizure-associated odours and nonseizure associated odours and video-recorded the reactions of the dogs to each. We found that all the dogs demonstrated more affiliative behavioural changes when confronted by seizure-associated odours, compared with their response to control odours. Our results support the view that untrained dogs detect a seizure-associated odour and are in line with the findings of the emerging literature, which attests that those epileptic seizures are associated with a unique volatile organic signature.

Epilepsy is a debilitating and potentially life-threatening neurological condition which affects approximately 65 million people worldwide. There is currently no reliable and simple early warning seizure-onset device available, which means many people with unstable epilepsy live in fear of injury or sudden death and the negative impact of social stigmatization. If anecdotal claims that untrained dogs anticipate seizures are found to be true, they could offer a simple and readily available early warning system. We hypothesized that, given the extraordinary olfactory ability of dogs, a volatile organic compound exhaled by the dog’s epileptic owner may constitute an early warning trigger mechanism to which make dogs react by owner-directed affiliative responses in the pre-seizure period. Using 19 pet dogs with no experience of epilepsy, we exposed them to odours that were deemed to be characteristic of three seizure phases, by using sweat harvested from people with epilepsy. The odours were delivered to a point immediately under a non-epileptic and seated pet dog owner’s thighs. By altering the alternating odours emerging from sweat samples, captured before seizure, during a seizure and after a seizure, and two nonseizure controls, we were able to record the response of the 19 pet dogs. Our findings suggest that seizures are associated with an odour and that dogs detect this odour and demonstrate a marked increase in affiliative behaviour directed at their owners. A characteristic response of all 19 dogs to seizure odour presentation was an intense stare which was statistically significant, (p < 0.0029), across the pre-seizure, seizure and post-seizure phases when compared to control odours of nonseizure origin.

1. IntroductionInternational surveys investigating whether untrained pet dogs can predict the onset of seizures in humans with epilepsy, have reported that they demonstrate behavioural changes which resemble attention seeking. The behaviours include intense staring; maintaining close-proximity; excessive panting; paw lifting; vocalisation; increased locomotor activity; licking; yawning and scratching [1,2,3,4,5,6,7,8]. While some authors contend behaviours of that nature might be indicators of physiological canine stress [9], others, such as [10,11,12], conjecture that stress in dogs is more likely to be demonstrated in activities such as cringing; crouching; freezing; hiding; shaking or barking; growling; baring teeth; snapping and lunging [13]. In the [8] international survey of untrained reactions of dogs to seizures, affiliative behaviours were the most commonly reported, with fearful or aggressive behaviours observed in less than 2% of dogs.Here we investigate the hypothesis that seizures are associated with distinct Volatile Organic Compounds (VOCs) and that these are the trigger mechanism for behavioural changes seen in untrained pet dogs. The proposition for the existence of distinct seizure-associated VOCs is supported by reports of physiological changes and excessive electrical activity in the brain preceding epileptic seizures [14,15,16]. Previous research indicates that these are instigated by the autonomic nervous system and the hypothalamic-pituitary-adrenal axis (HPA) and cause an increase in heart and respiration rates [17,18,19,20]. VOCs, which dissolve in the blood and saliva are exhaled as part of the respiratory process, or effused as sweat emanations [7,21,22,23,24,25,26,27,28,29]. VOCs have previously been shown to be indicators of a range of diseases including the presence of cancers, cholera, cystic fibrosis, diabetes, gut diseases, heart allograft rejection, heart disease, liver diseases, pre-eclampsia, renal disease, TB and congestive heart failure in older patients and COPD [24,26,30,31,32,33].Given the extraordinary scent capabilities of dogs [34,35,36,37,38,39], it is logical to hypothesize that VOCs may also act as a trigger mechanism for alterations in the behaviours of untrained pet dogs at the onset of epileptic seizure. Dogs are known to communicate not just between themselves but also with humans, using strategies which include visual, tactile, auditory, and olfactory signals [4,40]. They are known to gaze intensely when trying to draw the attention of their human care provider to a difficult to reach object [41,42,43,44], and can understand that a pointed finger indicates a direction they should follow [45,46]. Therefore, if the many anecdotal reports of untrained pet dogs anticipating seizures and apparently attempting to communicate that to their caregiver, are true, a seizure-specific VOC stimulus for such behavioural change seems feasible.The emergence of three recently published studies provides compelling evidence to support this hypothesis; for example, [7] report that medical detection dogs were successfully trained to alert on seizure-specific odour(s) and later [28] attest that data retrieved from SIFT-MS chemometric analysis illustrated the presence of pre-seizure-associated VOCs. Similarly, [29], delineate that seizure detection dogs demonstrated a predictive ability for seizures of up to 90 min prior to a seizure event.The aim of this current study was to explore, at a fundamental level, whether dogs with no previous training for epilepsy detection and having never previously witnessed an epileptic seizure would show changes in behaviour when exposed to odours associated with human epileptic seizures and apparently emerging from their owners.2. MethodsA repeated measures design experiment was conducted in which 19 recruited non-epilepsy dog-owner dyads from a local dog training club were subjected to a series of odours from sweat samples from three volunteer people with epilepsy and sweat samples from two non-epileptic (control) volunteers. Because this experiment included working with human volunteers, all methods were performed in accordance with the relevant guidelines and regulations.2.1. Ethical NoteEthical approval for the study was given by the Research Ethics Committee, School of Biological Science Queens University Belfast, To conduct nonlicensed animal research using 19 pet dogs which were exposed to seizure and control sample odours and their reactions monitored and recorded to those odours, Ref No. QUB-BS-AREC-19-001.Ethical approval was also granted to obtain sweat samples from three epilepsy patient volunteers and three control volunteer participants and for 19 dog owners to accompany their dogs and sit passively while the reaction of their dogs to seizure and control odours were being recorded. REF: 04/19/PowellNR1.All samples were provided by participants who had completed Informed Consent Forms. The passive observation and recording of the dogs’ responses to seizure and control odours were conducted in accordance with relevant ‘Arrive’ guidelines and regulations, (https://arriveguidelines.org, (accessed on 1 September 2019)).2.2. Volunteers for Sweat SamplesAn internet search generated a list of the principal English-speaking epilepsy Charities in the UK and Ireland and was extended internationally to involve volunteers from a broad cultural basis, thereby minimizing the risk of location bias. Appeals were also made for participants via the social media channels of Queen’s University Belfast. Charities were informed that the aim of the study was to scientifically investigate reports of pet dogs apparently anticipating seizure onset in their owners. To facilitate this research project, volunteers had to have medically diagnosed epilepsy, experience frequent seizure events (daily, weekly) and own pet dogs that could predict seizure onset.This yielded three epilepsy volunteers from different areas of the UK and Ireland, each of whom owned a dog that demonstrated pre-seizure awareness behaviour and remained with its owner throughout their seizure. Warning times varied from 10 min to 60 min. Early warning behaviours by these dogs enabled the acquisition of pre-seizure samples which would otherwise have been difficult to acquire. All three epilepsy volunteers, were females within the age range 21–55:–Volunteer A—a mature female living with daily absence and tonic-clonic seizures. Cause of epilepsy unknown.–Volunteer B—a mature female, also experiencing daily absence and tonic-clonic seizures. Cause of epilepsy unknown.–Volunteer C—an adult female with the genetic Lennox–Gastaut syndrome and severe developmental and learning issues and experienced several daily recurring tonic-clonic seizures. She was cared for by her parents.Control samples were provided by two people who did not have epilepsy, had no contact with people living with epilepsy, and were dog owners. It was felt that both groups should own dogs to avoid potentially confounding factors regarding a dog/no-dog effect.The epilepsy and control groups were asked not to alter their normal bathing/showering habits for sampling purposes because it was felt this would more accurately reflect the everyday conditions within which their own dogs reacted to seizures. Both groups were given written instructions on how to capture and store the apocrine sweat samples using sterile gauze pads (see Appendix A and Appendix B). Previous studies exploring the efficacy of bio-medical detection dogs, attempted to capture VOCs in exhaled breath and sweat taken from the back of the neck and hands [7,47]. However, apart from the difficulties in maintaining breath sample integrity, reservations have been expressed about the efficacious nature of sweat samples from hands and neck [27,48]. Thus, this study elected to use sweat samples taken from the axillae as the most appropriate vehicle for the detection of biomarkers [27]. For those with epilepsy, trusted others were invited to assist in harvesting samples during the course of the three seizure phases (one sample for each seizure phase) –pre-seizure taken when their untrained pet dog characteristically indicated that a seizure was imminent,–seizure sample harvested immediately while a seizure was occurring–post-seizure taken 6 h after a seizure episode to allow time for potential seizure-associated odours to dissipate.2.3. Sample Collection and StorageAll samples were stored at 4 °C since this method has been found to retain VOCs without negative impact for up to 4 weeks [7,26,28,29,30,49,50,51]. To minimize the risk of degradation of VOCs, the airtight glass vials were recapped between each sample presentation and only 3–4 dogs were tested each day [29].2.4. Study ParticipantsCanine participants and their owners were recruited in part from friends and acquaintances, and from a dog Training Club in Lisburn Co. Antrim (Lisburn, Northern Ireland), following a routine training club night, when time was set aside for the research project to be explained. Those present were invited to participate with their dogs in an investigation which would explore whether epileptic seizures are accompanied by a distinct odour. They were also informed that if evidence could be found for the existence of such an indicator, it could offer profound benefits for the safety of people who have difficulty managing their epilepsy. This appeal yielded a sample of 19 dogs of varying breeds, ages, and of both genders, whose owners were mainly female (16) (male, 3) with an age range across both genders of 18–66+ (Table 1). We hypothesized that, if pet dogs owned by people with epilepsy anticipate seizures by showing owner-directed behavioural changes, then pet dogs owned by non-epileptic people, could also demonstrate an attention-seeking behaviour on encountering seizure-associated odours apparently coming from their owners. In designing this experiment, a method was needed to convince pet dogs that their non-epileptic owners were about to experience an epileptic seizure. One possibility was to place gauze squares containing seizure-associated sweat samples into the pockets or socks of owners, but this proposal was rejected because of concerns about residual odours remaining in clothing [38]. Therefore, to prevent contamination issues, two specially designed pieces of apparatus called remote odour delivery mechanisms, (RODM, Figure 1), were used to deliver target odours from a separate laboratory to the dog owner’s location. Contamination issues would no longer be a problem because only target odours would be delivered from samples stored separately, which, after each trial, could easily be vented externally, leaving no residual trace.Each RODM consisted of a pump attached to a scent container with an outlet pipe which delivered odours a distance of 6 metres from a different room with closed door. Validation of the RODMs was completed in a separate earlier investigation using operational police drugs dogs and game flushing field trial champion dogs [8]. The results validated that all the dogs recognized their specific target substances and responded to them as they had been trained with no lack of performance [8]. To further reduce risk of contamination, one RODM was used to deliver experimental odours and the other for controls with no seizure-associated odour. Thus, the RODM offered a simple but effective solution to the issues of cross contamination and in future may prove useful in preventing the risks associated with experimental procedures involving volatile or toxic materials.2.5. Data CollectionDog characteristics recorded were breed, sex, age-group and years owned. Indicators of seizure odour response by the 19 untrained dogs were those which were most frequently reported in other studies of this topic; intense staring at the owner, close-proximity to the owner, and pawing or nudging the owner [1,2,3,4,5,8,48,52,53,54,55,56,57,58].Here, close proximity to the owner was measured as being within one metre as delineated by visual reference to marks which were already present on the floor of the laboratory, one to the left and one to the right of the owner. In studies of seizure alerting behaviours in dogs, owners were almost unanimous in believing their dogs’ pre-seizure behavioural changes were a warning mechanism [1,2,3,5,8]. Thus, while it is acknowledged that other potential behavioural responses such as avoidance, stress or aggression might also have been included in this study, they were excluded because they accounted for less than 2% of reported behaviours in several surveys of untrained canine seizure alerting activity.All trials of the dogs were conducted ‘blind’, thus, neither the principal researcher nor the dog owners, had any knowledge of the sequences of odour presentation. The dogs’ responses were recorded in seconds and were measured over five trials, each of which lasted three minutes with breaks of two minutes between trials, to minimize the fatigue factor on the dogs [25]. Thus, including the 3 min habituation time at the beginning of a test series, each test session took approximately 28–30 min to complete. Recordings were by HP laptop installed video camera with a backup provided by an iPhone tablet camera. Analysis of the video footage was made without knowledge of the sequence of sample odours used to prevent unconscious confirmation bias [59]. Because only three behaviours were being monitored in this study, automated video tracking software was considered unnecessary.2.6. Experimental ProcedureTo minimize risk of pseudoreplication, prior to testing, one epilepsy volunteer from the three available was chosen as the initial source of samples for the day’s tests. This was done by means of a simple toss of the coin (following the procedure explained in ‘Test’) by the research assistant, who did not share the outcome with any other research member. The samples used included two controls and three seizure phase odours comprising pre-seizure, seizure, and post-seizure, (thus, 9 seizure samples in total). The samples from each of the three volunteers were assigned numbers:–Volunteer A: pre-seizure, 1, seizure, 2, post-seizure, 5–Volunteer B: pre-seizure, 3, seizure, 4, post-seizure, 6–Volunteer C: pre-seizure, 10, seizure, 11, post-seizure, 12–Control 1: 7–Control 2: 8An online randomizer was used to produce several groups of five odour presentations for each of the epilepsy volunteers, consisting of their three seizure-associated sweat samples and two control sweat samples. The odour samples were re-used among the 19 dogs in a randomized fashion and were re-capped after each three-minute exposure. Thus, all the dogs were exposed to 9 seizure-associated samples but not necessarily all from the same person.Individual dog owners and their dogs were assigned specific times to attend the testing laboratory at Queen’s University Belfast, and on their arrival and before any testing began, each dog was allowed to familiarize itself with the test area. Each dog then underwent a series of randomly delivered odour presentations via the RODM (see above) to an area beneath the dog’s owner who was seated in the middle of the test area and was asked not to engage with their dog. It was recognized that this lack of response was not natural and carried some risk of negatively impacting the dogs’ normal behaviour, but, had they been allowed to engage with their dogs as normal, they may have inadvertently influenced their dogs’ responses. The RODMs individually delivered the experimental and control odours following the sequences chosen by the research assistant.The target behaviours being measured were stands or sits and stares at their owner—time eye contact, (TEC), which in one published work accounted for 70.8% of all attention seeking behaviours reported by over 130 surveyed dog owners [4]. The second attention seeking behaviour used in this study was time near owner (TNO), which [4], report accounted for (64.6%) of ASBs. The third response was time pressing close (TPC), which translates to nudging or pawing, accounting for (60.0%) of the behaviours identified in their survey [4].Pre-test: On arrival each dog was individually brought to the test room which measured 10.5 m × 8 m × 4 m (Figure 1) and was given 3 min to habituate to the surroundings prior to the trial beginning.Test: At the start of each day’s trials, the research assistant tossed a coin to determine which of the three volunteers’ samples would begin that day’s test, (A = 1, 2, 5, 7, 8/B = 3, 4, 6, 7, 8/C = 10, 11, 12, 7, 8). Thus, the coin was tossed three times using H to denote the winning value heads and T to denote the losing value Tails. The choices were A, B, C. If, after three tosses the outcome was HTT, then A wins. If the outcome was THT, then B wins and if the outcome was TTH, then C wins. In the event of a TTT or HHH outcome, the coin was tossed again until an uneven result emerged.Six sets of randomized combinations of 5 sample sequences, were created for each of the three volunteers, for example, the sequences of sample odours for Volunteer A were, 1, 2, 5, 7, 8; 8, 2, 1, 5, 7; 7, 8, 2, 5, 1; 5, 2, 8, 1, 7; 2, 5, 8, 7, 1; 2, 1, 7, 5, 8. Once the initial toss of the coin had determined which of the three epilepsy volunteers would be used to begin that day’s trials, subsequent odour deliveries from volunteer one, two or three, were chosen at random by the research assistant. Between each odour presentation, the dogs were removed from the test area to rest for two minutes, while the room and RODMs were being ventilated as per the times shown below.The two RODM scent chambers and pumps, one to deliver experimental odour and the other to deliver control odour, each at separate times, were kept in a separate laboratory to prevent risk of contamination. In addition, fresh sterile gloves were worn each time samples were handled. Each RODM (Figure 2) consisted of a new aquarium pump (5 W, 240/50 Hz Air pump 200 delivering 200 L/h) and connected to a re-sealable 3.6 L plastic watertight storage keg (UN approved, meaning, they are suitable for a range of applications such as transportation of chemicals to food or water storage). The storage keg had a resealable open top wide mouth and had inlet and outlet valves fitted at opposite sides. Pumps and kegs were connected to each other by 4 mm (internal diameter) plastic aquarium hose. The outlet pipes were 15 m aquarium tubing (4 mm) each with a nonreturn valve at the end. The outlet tubes were placed one under each thigh of the participant with the tube ends just showing at the inside of each leg. It was expected that this arrangement would increase the likelihood of the emerging scent samples remaining close to the participant’s body.During exposure to each odour, the times spent by the dogs on each of the 3 behavioural responses, detailed above, were recorded. The lengths of time for which the pumps ran while delivering scent samples and during the system flush sequences were calculated thus:–Pump delivers 200 Lt in 60 min = 18 s/1 L–Airtight keg volume = 3.6 Lt–Time to clear 3.6 Lt = 3.6 × 18 s = 64.8 s = 1 min approx.–Time to run sample scent before introducing dog = 1 min.–Time to run sample scent, dog in room = 3 min–Flush time after trial (directed outside window dog out of room) = 1 min.–Total time for each scent sample = 5 min–Total number of samples per dog = 5 samples–Total time needed for each dog = 5 × 5 = 25 min–+3 min initial habituation time = 28 min/dog2.7. Statistical AnalysisAll data analysis was performed using the statistical package IBM Corp. Released 2016. IBM Staistics for Windows, Version 24.0. Armonk, New York, IBM Corp.SPSS, (Version 24) IBM, Corporation, Armonk, NY, USA). Data from the two control odours, (C1, C2) were combined and averaged to give a mean for each of the three selected behavioural responses, time near owner combined control, time in eye contact combined control, and time pressing close combined control. The dogs’ three behaviours, time near owner (TNO), time in eye contact with owner (TEC), and time pressing close to owner (TPC) were then measured across all three seizure-related odours and comparisons drawn with their responses to the combined controls.Visual inspection of histograms and use of Kolmogorov–Smirnov tests revealed the behavioural data were not normally distributed, thus indicating the need for nonparametric statistics. Friedman tests were used to examine whether each of the three behavioural responses (TNO, TEC, TPC) differed across the four treatment conditions, (pre-seizure, seizure, post-seizure and control). Wilcoxon signed rank tests were then used to make pairwise comparisons between each of the pre-seizure, seizure, post-seizure, with the mean control where significant differences were found.3. Results3.1. Demographic InformationAs can be seen in Table 1, nineteen owners (16 female, 3 male), from a wide age range volunteered for this study, most of whom were female. The majority of the participating dogs were female, of pedigree status, and under 8 years of age. Most owners stated their dogs had been with them for 12 months or less and their main reason for acquiring a dog had been companionship. None of the dogs had witnessed an epileptic seizure and none of the dog owners had epilepsy.3.2. Behavioural Responses to Seizure-Related and Control OdoursTNO—Time near owner differed across the four treatments, pre-seizure, seizure, post-seizure and combined control, (Friedman test, X2 (3) = 10.5, p = 0.015). The median levels for all four odour responses measured in seconds were respectively, 37.0, (IQR = 43), 47.0, (IQR = 71), 53.0, (IQR = 76), and 34.5, (IQR = 27) (Figure 3). More specifically, Wilcoxon paired comparisons of behavioural changes to the seizure-related odours with the combined control odour revealed: TNO pre-seizure, Z = −1.53, p = 0.13; TNO seizure, Z = −3.1, p = 0.002, TNO post-seizure, Z = −2.8, p = 0.005. Thus, dogs spent more time near their owner during the delivery of the seizure and post-seizure odours compared to the control condition, with no difference for the pre-seizure condition.TEC—Time in eye contact: the dogs engaged in significantly more eye contact with their owners across the three seizure odours than they did with the combined control odours, (Friedman test, X2 (3) = 11.8, p = 0.008); median levels for TEC pre-seizure, seizure, post-seizure and TEC combined control were, respectively, 3.0 (IQR = 7.0), 7 (IQR = 9), 8.0 (IQR =11) and 2.5 (IQR = 5) (Figure 4).Wilcoxon signed rank tests comparing response to seizure-related odours and the combined control revealed, TEC pre-seizure: Z = −2.3, p = 0.028; TEC seizure, Z = −2.2, p = 0.022 TEC post-seizure Z = −2.9, p = 0.004. Thus, dogs spent more time engaging in eye contact with their owners during the delivery of each of the three seizure related odours compared to the controls.TPC—Time pressing close: was found to differ across the four treatments (Friedman test, X2 (3) = 8.4, p = 0.038); median levels for TPC pre-seizure, seizure, post-seizure and combined control were 4.0 (IQR = 7), 6.0 (IQR = 18); 1.0 (IQR = 8) and 3.0 (IQR = 5) (Figure 5). Wilcoxon signed rank tests comparing the responses of the dogs to the seizure-related odours and the combined control revealed, TPC pre-seizure, Z = −1.17, p = 0.24; TPC seizure, Z = −2.5, p = 0.013, TPC post-seizure, Z = −0.44, p = 0.66. Thus, dogs spent more time pressing close to their owner during the delivery of the seizure odour compared to control, with no difference for the other seizure related odours.4. DiscussionThis study explored the propensity of pet dogs to anticipate and respond to human epileptic seizure onset apparently emanating from their owners. Each dog was exposed to odours from three phases of seizure and their reactions compared with exposure to control odours. It was hypothesized that when seizure-odours were apparently coming from their owners, the 19 untrained pet dogs would demonstrate significant behavioural changes concomitant with attention-seeking activities. In line with this prediction, all nineteen pet dogs engaged in a significant increase in attention-seeking behaviours on detecting odours from seizure associated sweat samples compared with control odours. These findings have since been strengthened by a study which reports that seizures are associated with specific VOCs, which can be detected by trained seizure alert dogs more than an hour prior to a seizure’s manifestation [29]. In a separate investigation, further support has emerged from reports that a distinct VOC seizure-related profile has been detected by ion flow tube spectrometry (SIFT-MS) up to four hours in advance of a seizure [28]. Interestingly, [60] contend that a seizure-specific odour has been identified which is ‘predominantly of menthone’ [60] (p. 8), and this is also emitted by non-people-with-epilepsy who are experiencing a fearful situation and is believed to be an alarm pheromone. Taken together therefore, the results of this study and those of the published articles referred to, provide compelling support for the hypothesis that untrained dogs respond to epileptic seizures. This is an important finding because it offers the means to achieving a simple and reliable protocol for training dogs to warn people of an impending epileptic event, thus meeting a profound and long held wish for some form of pre-seizure warning device, held by those who live with difficult to control seizures [15,61]. It also holds out the hope, not only to improve patient safety and well-being but also of promoting ‘therapies aimed at rapidly treating seizures (and) be able to abort seizures through targeted therapies’ [62].The behavioural responses of the dogs varied during the different seizure-phase sample presentations, but at this stage it is unclear whether the dissimilarities were an outcome of the phase of seizure, the sample taking procedure used, or the nature of the behaviour itself. For example, the ‘time spent near the owner’ (TNO) reaction differed significantly from controls during the seizure and post-seizure odours, but not during the pre-seizure odour presentation. It is possible this may have been the result of a sampling error by the volunteer or because there was an insufficient concentration of odour to meet the threshold for that response. For instance, the pre-seizure sweat sample capture depended upon recognition by the owner or family member, that their dog was displaying his/her innate ‘warning’ of an impending seizure event. Thus, a misinterpretation of the dog’s actions prior to prodrome, could have resulted in a less than optimum sample. A further anomaly was the unexpected similarity of the dogs’ reactions to the seizure and post-seizure odours, despite the post-seizure sample having been taken six hours after seizure. This outcome appears to contradict the findings of two recent studies [28,29]. The former report 3–4 dogs distinguished between seizure and inter-seizure samples (3 h post-seizure), with a probability of 93.7% [29]. Equally, [28] delineate inter-seizure samples, analyzed by SIFT-MS chemometric analysis taken 6 h post seizure, as having a significantly different VOC profile to that of seizure odour. This inconsistency may be understood, however, by considering the potential impact of environmental differences in sampling. For example, the volunteers in both the [29] study and that of [28] were all patients receiving treatment for epilepsy in medical centres with excellent ventilation and well-established clinical attention to cleanliness and hygiene. In the period following a seizure, therefore, it is conceivable that seizure scent could have quickly dissipated or that the patients may have been given a change of clothes or been moved to a different ward. In any of those circumstances, residual seizure odour would have been unlikely to linger, thus implying that the post-seizure samples used by those two studies, [29], study and [28], more accurately represented that phase of seizure. On the contrary, the sampling procedures used by the three epilepsy volunteers in this current study were not clinically derived, having been secured in the volunteers’ own homes before, during and after a seizure event. It is therefore conceivable that the 6-h post seizure samples had been contaminated by residual seizure odour which might explain why the 19 dogs in this current study gave positive responses to them instead of ignoring them as had been expected. Future research in this area, where clinical sampling of seizure associated samples may not be possible, would be advised to encourage volunteers to change their clothing between seizure stages, open windows to ventilate the room or even move to a different room, and, to allow a longer post-ictal sampling period, perhaps in the order of 12 h.On the other hand, the behavioural response described as time of eye contact (TEC) by the dogs was found to be the most prolific of all three behaviours, having significantly increased across the three seizure phases, pre-seizure, seizure, and post-seizure, when compared to controls. This outcome is consistent with the findings of [4], who report that staring is the most common (70.8%) attention seeking mechanism engaged in by dogs. It also accords with [52], who found that gazing by dogs is often a request for help in the context of unsolvable tasks and reflects the findings of [58,63], who contend that gazing may have been a successful coping strategy in the past. Perhaps in the context of seizure odours, regardless of seizure-phase, the TEC response becomes the dogs’ preferred strategy in the face of what might be seen as an insurmountable problem [63].The third behavioural response considered in this study was time pressing close (TPC), comprising of nudging or pawing their owners. The data suggest that this was significantly increased during detection of the seizure-phase odour compared to control, but no difference was found between controls and pre-seizure or post-seizure phase odours. These disparities are difficult to explain but might be understood in the context of TPC being the least frequently reported of the three attention-seeking behaviours. It is also the behaviour which is reportedly discouraged as most annoying by many dog owners [4]. That said, there is currently no way of knowing whether the potency of odour, regardless of seizure-phase, might fall short of a necessary threshold for evoking any particular behavioural expression, and would therefore indicate the need for further exploration.In this investigation, seizure associated odours were presented to pet dogs using a novel method which was designed to deceive the dogs into believing the seizure-related odours were emanating from their owners. Thus, the current study demonstrates support for the hypothesis that a seizure-related olfactory trigger mechanism evokes spontaneous seizure alerting behaviour in pet dogs. This does not, however, exclude the possibility that other trigger stimuli may also exist, and, whilst beyond the scope of this study, it is acknowledged that other researchers have hypothesized the existence of a sensitivity in dogs to electromagnetic changes [21,64,65,66]. That being the case, it is conceivable that dogs may also be responding to electromagnetic signals which are associated with pre-seizure physiological changes [67,68,69,70,71,72] and invites further investigation. Thus, it is also conceivable that dogs which anticipate seizure onset may be responding not only to olfactory stimulation but also to minute variations in the body’s electromagnetic field which accompany the onset of epileptic seizure.5. LimitationsIt is acknowledged that a larger sample of seizure volunteers would have strengthened our findings, but despite a widespread appeal, we were unable to obtain any more than the three. They were, however, all from different countries, thereby reducing potential cultural bias.All three seizure volunteers were female which was governed by participant response and therefore possible gender bias cannot be dismissed.Seizure-odour degradation was a potential limitation which was addressed by recapping scent containers between each individual test and by restricting testing 3–4 dogs each day.6. ConclusionsThis study set out to scientifically examine anecdotal reports of seizure anticipation behaviours by untrained pet dogs and hypothesised that the trigger mechanism for these reported activities might be some form of odour that is unique to seizures. Consistent with this hypothesis, the results of this investigation provide compelling support for the contention that pet dogs anticipate epileptic seizures, consistent with an innate response to the impact of an olfactory trigger mechanism. These findings reflect those of more recent research into this subject. We have also provided compelling evidence for the contention that this olfactory biomarker(s) is directly associated with epileptic seizures, across their three phases, pre-seizure, seizure, and post-seizure. Although the nature of the biomarker remains unexplained, the findings have significant implications for developing a programme of targeted training for seizure prediction dogs with the possibility of a reduction in accidents and injury caused by unexpected seizure occurrences. The insights gained from this study may also be of assistance in improving the sense of self-worth of people living with a challenging epileptic condition while at the same time improving their quality of life and their sense of independence.

animals : an open access journal from mdpi

10.3390/ani11123605

PMC8698036

Suckling lamb meat is the secondary product of the Mediterranean traditional dairy sheep industry. Similar to the main production, i.e., milk, lamb meat contributes to the emission of greenhouse gases (GHG), whose main portion is represented by enteric methane produced by the lamb dams. Such an emission, although limited in quantitative terms, should be mitigated by appropriate feeding or compensation techniques. Among all the sources of variation of meat lamb emissions, sex of the lamb and type of lambing (single or twins) showed the largest effect.

The aim of this study was to estimate the methane-linked carbon footprint (CF) of the suckling lamb meat of Mediterranean dairy sheep. Ninety-six Sarda dairy ewes, divided into four groups of 24 animals each, were assigned to 2 × 2 factorial design. The experiment included the suckling lamb feeding system: traditional (TS), in which lambs followed their mothers on pasture during grazing time, vs. separated (SS), in which lambs remained indoors, separated from their mothers during the grazing time. Each group was divided into high (HS) and low (LS) supplemented ewes (600 g/d vs. 200 g/d of concentrate). The estimated CH4 emission of the ewes, calculated per kg of body weight (BW) gain of the lamb during the suckling period, was then converted to CO2eq with multiplying factor of 25. The TS lambs showed lower methane-linked emissions than SS ones (p < 0.05). The sex of lambs affected their methane-linked CF, with males having lower (p < 0.05) values than females. Twins displayed much lower methane-linked CF than singles (4.56 vs. 7.30 kg of CO2eq per kg of BW gained), whereas the level of supplementation did not affect greenhouse gases (GHG) emission. Interaction displayed lower and not-different GHG emissions for both indoor- and outdoor-reared twins. In conclusion, the methane-linked CF of the suckling lamb meat can be reduced by maintaining the traditional lamb rearing system and by improving flock prolificacy.

1. IntroductionDairy lamb is a secondary product of dairy sheep farms, and it is consumed mostly in Mediterranean countries [1]. It represents a niche product appreciated by consumers for its nutritional and organoleptic characteristics, due to both the young slaughtering age (4–6 weeks of age) and the quality of maternal milk obtained mainly by grazing natural pastures [2]. Moreover, the suckling lamb meat is an interesting source of fatty acids of nutritional importance [3] and it is particularly suitable for children’s diets, especially in the weaning phase [4,5].Growing concerns of European citizens about the environmental impact of animal productions require that foods must also guarantee sustainability, especially in terms of climate-altering gas emissions [6]. As a consequence, the number of livestock life cycle assessment (LCA) studies has considerably increased in the last two decades [7]. Most of this research deals with the environmental sustainability of beef and pork production. Carbon footprint (CF) of lamb meat has received less attention [7], and studies were carried out mostly on the quantification of environmental performance of the heavy lamb with values ranging hugely from 2.8 [8] to 38.45 [9] kg CO2eq per kg of live weight (LW) [8,9,10,11,12,13,14,15,16,17,18]. All research was carried out in Oceania (most of the studies), in Europe, US, China, and Chile [19]. Most of them quantified greenhouse gases (GHG) emissions by comparing different farming systems, from pasture to zero-grazing [13], from lowland to hill farms [14], from conventional to organic [10], or considering different forage species in pasture-based flock management [20]. Recently, methane production of fattening lambs was predicted by intramuscular fatty acid profile [21].Most of these studies consider 1 kg of live weight (LW) as functional unit and, since sheep farms produce two or three coproducts (milk, meat, and wool), economic or a biophysical allocation are generally used to distribute the overall impacts between them [19]. Furthermore, the greater amount of impact in this meat production system occurs at the farm level (90% [9,22]), and it depends mostly on enteric methane emissions whose relative contribution to total CF ranges from 58% [13] to 80%% [23]. CO2eq emissions from purchased feeds, energy, and fuel, and N2O emissions from soil and manure management, contribute to the total impact in a lesser proportion. All of these studies deal with meat sheep lamb production, whereas, to the best of the authors’ knowledge, the CF of suckling dairy lambs has not been estimated. As recently reviewed by Battacone et al. [3], suckling lambs are fed exclusively maternal milk from birth to slaughter. Thus, this type of production does not require additional inputs than those demanded by their mothers. For these reasons, CF of suckling dairy lamb should be probably lower than values available in the literature.Feeding technique has been demonstrated to be effective in reducing the CF of milk in dairy cows [24,25,26], goats [25,27], and sheep [25,26,27]. Methane emission per unit of milk or meat has continuously decreased during the last decades, and it is expected to continue this trend [28]. Thus, it is reasonable to hypothesize that feeding techniques implemented to reduce the environmental impact of dairy ewes could also influence the CF of suckling lamb meat. In fact, the feeding regimen influences ewe dry matter intake (DMI) and, consequently, the methane yield (the principal GHG produced by sheep). In this work the CF of suckling lambs under different management systems and mother feeding was estimated by considering exclusively the CO2eq derived from the CH4 emitted by the ewes during the suckling period. This choice assumed that methane emissions related to gestation and replacement can be totally attributed to milk production, which is the main activity of the farm, with an allocation of 100% of all other emissions to this production. Such a strong assumption was made in order to simplify calculations and to make the comparison between different production systems easier, without altering the overall impacts of sheep dairy farms. Aim: The aim of this work was to estimate the methane-linked carbon footprint (CF) of the suckling lamb meat of Mediterranean dairy sheep.2. Materials and MethodsThe CF of suckling lambs was estimated using data collected in a commercial dairy sheep farm located in the northwest of Sardinia (Italy). The animal protocol was carried out in compliance with the EU and Italian regulation on animal welfare. For this study, no animals were specifically killed for experimental purposes; however, data at commercial slaughter were collected. University ethics approval was also not required. 2.1. Experimental Procedure: Animals and DietThe experiment involved ninety-six Sarda nursing ewes (body weight (BW): 46.33 ± 0.40 kg; mean ± standard error) who were monitored with their lambs for a period of 28 days. The trial started immediately after lambing. In Sardinian dairy sheep farming, there are two lambing seasons: autumn for pluriparous, and spring for primiparous. In this trial, only pluriparous ewes were chosen, so lambing was concentrated at mid-November.A sample of 100 ewes lambing with parturition occurring within two days were selected from the flock and serially numbered. Then, 96 animals were extracted and randomly assigned to four groups of 24 animals each in a 2 × 2 factorial design: (a)Traditional system with high supplementation (TS-HS), in which mothers were followed by suckling lambs during the grazing time and they received a high dose of supplement (600 g/d of concentrate).(b)Traditional system with low supplementation (TS-LS), in which mothers were followed by suckling lambs during the grazing time and they received a low dose of supplement (200 g/d of concentrate).(c)Separated system with high supplementation (SS-HS), in which mothers were not followed by suckling lambs during the grazing time (suckling lambs remained indoors) and they received a high dose of supplement (600 g/d of concentrate).(d)Separated system with low supplementation (SS-LS), in which mothers were not followed by suckling lambs during the grazing time (suckling lambs remained indoors) and they received a low dose of supplement (200 g/d of concentrate).The ewes grazed daily on a lush pasture for 6 h (9:30 a.m. to 15:30 p.m.). The concentrate was offered during two daily meals. In addition, all ewes had ad libitum access to hay during the night. The chemical composition of feeds offered is showed in Table 1.The newborn lambs (n = 44 females and 52 males) were fed exclusively maternal milk throughout the whole experimental period (28 days). At 28 days of age, they were weighed and then slaughtered in an authorized commercial abattoir.2.2. Measurements and SamplingDuring the experimental period, BW of ewes was measured weekly by using an electronic scale. Lamb weight was measured at birth and then once a week until slaughter. Average daily gain was calculated.Individual milk yield was measured on the two consecutive days after slaughter (two times per day, at morning and evening milking) to confirm the estimation of milk produced by the dams in function of daily growth of lambs. Individual milk samples (n = 384; 96 per treatment) were also collected and analyzed for chemical composition. Samples of grass, hay, and concentrate were collected weekly for chemical analysis.2.3. Chemical AnalysesMilk samples were analyzed for fat, protein, lactose (infrared method; Milkoscan 4000, Foss Eletric, Hillerød, Denmark), urea content (enzymatic-colorimetric method based on Berthelot reaction; Chemspec 150, Bentley Instruments Inc., Chaska, MN, USA), and somatic cell count (SCC, flow-cytometry method; Fossomatic 5000, Foss Electric, Hillerød, Denmark).Feed samples were ground with a Hammer mill by using a 1 mm screen, and then analyzed for DM, CP (Kjeldahl method; AOAC International, [29]; method 988.05), NDF, ADF, ADL (including termostable-amylase and following the method of Van Soest et al. [30]), ether extract (EE; Soxlet, AOAC International, [31]; method 920.39), and ash (AOAC International, [29]; method 942.05) after drying at 105 °C.2.4. Carbon Footprint Assessment, System Boundary, Functional Unit, and Allocation MethodThe CH4-linked carbon footprint was calculated within a cradle to farm gate system boundary considering maternal enteric CH4 emissions and milk suckled by the lambs as the main emissions hotspots. Data collected during the experimental trial were used to estimate maternal DMI [32] as follows:I = −0.545 + 0.095 MW + 0.65 FPCM + 0.0025 BWC(1) where I = DMI in kg/head day−1; MW = metabolic weight (BW0.75) in kg; BWC = bodyweight change in g/day; FPCM = fat (F = 6.5%) and protein (p = 5.8%) corrected milk (M) in kg which, in turn, was calculated as: FPCM = M (0.25 + 0.085F + 0.035P) (kcal/kg = 1047) (2) where M = milk yield in kg; F and p = fat and protein concentration in %, [33].Milk suckled by the lamb was estimated at 5.376 kg/kg of BW growth (BWG), arranging the Pulina et al. [34] equation which estimates the daily milk production of dams (M) in g/day as function of BWG (in g/day) and MBW of lambs (in kg):M = 140.6 + 4.52 BWG − 0.705 MBW (3)Methane emissions were then estimated by using the equation 10.21 of the Intergovernmental Panel on Climate Change (IPCC) guidelines for national GHG inventories [35]:EF = GE(Ym/100)/55.65(4) where EF = emission factor, kg CH4 head−1; GE = gross energy intake, MJ head−1 day−1; Ym = methane conversion factor, per cent of gross energy in feed converted to methane; the factor 55.65 (MJ/kg CH4) is the energy content of methane. The estimated CH4 emission of the ewes was then expressed in terms of CO2eq where 1 kg CH4 = 25 kg CO2eq in accordance with the global warming potential of emissions defined by the IPCC guidelines. Finally, the CH4-linked CF was calculated considering 1 kg of BW gain (during the suckling period) as functional unit (FU) and applying no allocation factor.2.5. Statistical AnalysisEwe DMI data were analyzed with the following linear model:DMI = µ + G + C + P + 1st_inter + ε where DMI (in kg) is the total 28 days intake during the suckling period, G is the lamb management (TS vs. SS), C is the supplement level (HS vs. LS), P is the kind of lambing (single vs. twins, no triplets were admitted to the experiment), and 1st_inter are the first order interactions between the couples of experimental factors.Birth weight, slaughter weight, average daily gain (ADG), and CH4-linked CF of lambs were analyzed with the following linear model: Y = µ + G + C + P + S + 1st_inter + ε where Y is the dependent variable, G is lamb management (TS vs. SS), C is the supplement level (HS vs. LS), P is the kind of lambing (single vs. twins, no triplets were admitted to the experiment), S is the sex of lamb, and 1st_inter are the first order interactions between the couples of experimental factors.Differences between means were detected with Tukey test, and significative level was declared for p < 0.05 [36].3. Results and DiscussionMilk production of ewes in the first control after lamb slaughter were 1.36 ± 0.074 kg/d (mean ± standard error) for TS-HS group, 1.19 ± 0.075 kg/d for TS-LS group, 1.28 ± 0.074 kg/d for SS-HS group, and 1.32 ± 0.075 kg/d for SS-LS group. They were slightly lower than those estimated by using Equation (3) (TS-HS: 1.68 ± 0.095 kg/d, TS-LS: 1.46 ± 0.069 kg/d, SS-HS: 1.32 ± 0.092 kg/d, SS-LS: 1.35 ± 0.068 kg/d; mean ± standard error) because milking normally produces slightly less milk than suckling. 3.1. Suckling Lambs Performance at Birth and after the Suckling PeriodData on birth body weight, body weight at slaughter, and ADG are shown in Table 2. The first-order interactions between the couples of experimental factors were never significant (p > 0.05), except for lamb management x type of lambing interaction for kg CO2eq/kg lamb BW gain. Mother’s supplement level did not affect lamb birth weight. The lamb management significantly affected body weight at slaughter (p < 0.01) and the ADG (p < 0.001). TS Lambs exhibited higher slaughter weights and ADG compared to SS lambs, probably because they could suckle more times a day from their mother than SS. Such a larger amount of milk received by TS lambs compensated the higher energy expenditure for movement and thermoregulation needed to follow the mothers during grazing.Concerning the type of lambing, our data evidenced that single-born lambs had higher birth weight (p < 0.01), birth weight at slaughter (p < 0.001), and ADG (p < 0.01) compared to twins. These results are comparable to those observed in previous studies in which single lambs showed higher live weight than twin lambs both at birth and later on [37,38], and tended to grow faster than those with lower live weight at birth [38]. Regarding the sex, male lambs had higher birth weight (p < 0.01), birth weight at slaughter (p < 0.01), and ADG (p < 0.01) compared to female lambs. Generally, at birth, males show higher birth weight than females [39,40,41]; this difference has been reported for several breeds, and it seems to persist during the life [42]. Other studies did not find differences between sexes at birth, but they evidenced that male lambs tend to grow faster than females [38]. 3.2. Carbon Footprint of Suckling LambsIn the present study, CH4-linked CF of suckling lamb varied between 4.56 and 7.30 kg CO2eq/kg lamb BW gain, respectively. A comparison with other studies is quite difficult because of the lack of research on the estimation of CF of suckling lambs and also because of the methodological heterogeneity among studies. However, from our data, it is possible to observe a strong relationship between CH4-linked CF of suckling lamb and BW gain of lambs (Figure 1), in agreement with the relationships between GHG emissions and the kg of LW or BW gain observed in sheep meat breeds [43,44,45,46].The estimation of total DMI of ewes and CH4-linked CF of suckling lambs are reported in Table 3. Ewes of TS showed higher (p < 0.01) total DMI as a consequence of their higher milk production, caused by the higher frequency of suckling activities of the lambs. In fact, lambs following their mothers all day around matched their ethological behavior, compared to lambs kept indoors that had the possibility to suckle only during the night. Increasing the frequency of suckling in TS lambs caused more frequent udder-emptying, which stimulates milk secretion [47]. The lower CH4-linked CF of TS lambs compared to SS (p < 0.05) could be therefore explained by the fact that the higher DMI of mothers is diluted in a greater BW gain of their offspring (Figure 2).The level of supplement used did not affect the DMI of ewes or the CH4-linked CF of suckling lambs. In this sense, very few studies have evaluated the effect of concentrate supplementation in animals under grazing conditions. Commonly, a negative relationship between CH4 (g/kg DMI) and level of concentrate in the diet has been reported in beef cattle [48] and lambs [49], and the use of concentrates has been proposed as a valid mitigation strategy for ruminants [50]. In the current study, the results are likely explained by the lack of substitution effect of supplement on pasture, because of the similar nutrient composition of grass and supplement. This is in agreement with a study conducted in dairy cows under high-quality grazing conditions where the increase in concentrate supplementation resulted in a simultaneous increase in enteric methane emissions and milk production, and so considering that methane emissions are expressed per unit of milk yield, the effect of supplement on GHG mitigation was not evident [51].Previous studies carried out in lambs evidenced that varying the proportion of concentrate did not affect CH4 emissions in lambs fed a basal diet composed of high-quality forages [52]. This suggests that the positive reduction of CH4 per kg of DMI due to an increase of concentrate amount in the diets could be, in part, counterbalanced by the higher intake of high-quality forages. The findings of the present study suggest that high-quality pasture could act in the same way as concentrates in reducing CH4 emissions. In fact, several studies conducted on beef cattle [53] and sheep [54] farmed in grazing systems with different pasture quality evidenced that high-quality grass can reduce CH4 emissions per unit of DMI (CH4/kg DMI) in comparison with low-quality grass. The positive effect of high-quality pasture on the reduction of enteric CH4 emissions is due to the lower content of NDF, to the high content of CP, and to the higher digestibility [55]. The quantity of fresh forages offered can also have an effect: in fact, CF is lower in high, rather than in low, productive grazing systems [9]. Thus, the improvement of pasture (both in quality and quantity) can be considered a good mitigation strategy to implement at a farm level for reducing enteric methane emissions of ruminants under grazing condition [50,55]. In addition, the use of pasture contributes to reducing the consumption of off-farm feeds and the management of pasture increases soil’s carbon sequestration [23].Twin-lambing ewes showed higher (p < 0.001) DMI than ewes with single lambs, probably because of their higher milk production. It is widely assessed that ewes with twins produce more milk than ewes with single lambs [56]. This is due to both the action of the placental lactogenic hormone, whose secretion is proportional to the weight of the placenta and that stimulates greater mammary growth [57], and to the more frequent and complete emptying of the mammary gland [58,59]. As twin-lambing ewes produced more milk, the CH4-linked CF of their lambs was markedly lower than that of single lambs; this result is also due to sharing of the maintenance requirements of mothers between the twins. The relation between CH4-linked CF of suckling lambs and lamb BW gain in twins compared to single-born lambs is shown in Figure 3.The CH4-linked CF of twins did not change with the management system, while that of single lambs was lower when they followed the mother on pasture, as evidenced by the significant interaction (p < 0.05; Figure 4).The sex of lambs affects their CH4-linked CF, with males having lower (p = 0.038) values than those of females, due to their highest growth rates which dilute the CH4 emission of the mother into a higher BW at slaughter, as previously shown in Figure 1.3.3. Practical ImplicationsSince Sarda sheep are fed mainly on pasture, our data suggest that the environmental impact of suckling lamb meat production can be reduced by improving flock prolificacy and maintaining the traditional lamb rearing system. However, these results are affected by the high nutritional value of pasture and of its large availability. The improvement of forage quality (through the evaluation of the best phenological stage) and forage type (in terms of botanical composition) can be considered as valid mitigation strategies to reduce livestock emissions [26]. Pasture-based systems in the dairy sheep industry are an important tool to mitigate the GHG impact for the capacity of grasslands to sequester C in soil as well as for the ability to provide ecosystem service and animal welfare [26,60,61,62].Considering the high diffusion of agro–silvo–pastoral systems in Mediterranean countries where dairy sheep are farmed, new findings on carbon sequestration in soil under pasture management demonstrate that the GHG emissions for suckling lamb meat production can be compensated annually by few m2 of undisturbed pastureland [63].4. ConclusionsGrowing concerns about GHG emissions among consumers are driving supply chains to reduce their impacts until the net zero goal is achieved. Not even niche productions, such as dairy lambs linked to traditional pastoral systems, widespread in the Mediterranean area escape this logic. This work, which evaluated only methane emissions from lactating ewes as representative of the GHG impact of dairy lamb meat production, evidenced that the type of suckling management, but not the ewe concentrate level, affected the CH4-linked CF of lamb meat. Specifically, traditional suckling techniques resulted in a lower CH4-linked CF of lambs compared to one in which the lambs were separated from their mother during the grazing period. Moreover, a high twinning rate of the flock can be an effective option for reducing the GHG impact.To conclude, this paper provides the first data on the estimation of environmental impact of the suckling lamb meat production in the Mediterranean region and suggests that to reduce the environmental impact of suckling lamb production systems, lambs could be raised with traditional suckling technique and should be twins.Some agronomic and livestock practices can be linked to mitigate the GHG impact of dairy sheep industry.

animals : an open access journal from mdpi

10.3390/ani11071875

PMC8300211

Saccharin sodium (SS) is one of the commonly used artificial sweeteners (AS) in the food industry, but the mechanisms mediating the physiological effects of sweeteners in the gut-brain axis is still unclear. The aim of this study was to explore the regulatory effect of SS on the microbiota-gut-hypothalamus axis on guinea pigs. We found that SS treatment may alter the growth and glucose metabolism of guinea pigs by activating sweet receptor signaling in the gut and growth hormone-releasing peptide (GHRP) hormone secretion. Besides, SS treatment increased the abundance of Firmicutes and Lactobacillasceae-Lactobacillus in the ileum, and subsequently increased levels of lactic acid and short-chain fatty acids (SCFAs). Adding 1.5 mM SS to drinking water alters the growth of guinea pigs by regulating the microbiota-hypothalamus-gut axis. This conclusion has theoretical implications for the comprehensive assessment of the biological effects of appropriate SS in the food industry.

The effects of saccharin, as a type of sweetener additive, on the metabolism and development of mammals are still controversial. Our previous research revealed that saccharin sodium (SS) promoted the feed intake and growth of guinea pigs. In this experiment, we used the guinea pig model to study the physiological effect of SS in the microbiota-gut-hypothalamus axis. Adding 1.5 mM SS to drinking water increased the serum level of glucose, followed by the improvement in the morphology and barrier function of the ileal villus, such as SS supplementation which increased the villus height and villus height/crypt depth ratio. Saccharin sodium (SS) treatment activated the sweet receptor signaling in the ileum and altered GHRP hormone secretion. In the hypothalamus of SS and control (CN) group, RNA-seq identified 1370 differently expressed genes (796 upregulated, 574 downregulated), enriching into the taste signaling transduction, and neuroactive ligand–receptor interaction. LEfSe analysis suggested that Lactobacillaceae-Lactobacillus was the microbe with significantly increased abundance of ileum microorganisms in the SS-treated group, while Brevinema-Andersonii and Erysipelotrichaceae-Ilebacterium were the microbes with significantly increased abundance of the control. Furthermore, SS treatment significantly enhanced the functions of chemoheterotrophy and fermentation of ileal microflora compared to the CN group. Accordingly, SS treatment increased levels of lactic acid and short-chain fatty acids (acetic acid, propionic acid and N-valeric acid) in the ileal digesta. In summary, drinking water with 1.5 mM SS activated sweet receptor signaling in the gut and altered GHRP hormone secretion, followed by the taste signaling transduction in the hypothalamus.

1. InstructionArtificial sweeteners (AS) are widely used in food products and soft drinks. Saccharin has good hydrolysis and PH stability, which can be completely absorbed and quickly metabolized [1]. Early intervention studies revealed that the chronic replacement of dietary sugar with AS can reduce energy intake and body weight via reducing calorie intake [2]. It is suggested that AS induces taste receptor activation in the intestine and adaptively regulates the expression of glucose transporters (SGLT-1/GLUT2), which are closely related to the glucose homeostasis [3]. Therefore, many researchers insist that AS ingestion stimulates hunger and consequently increases food intake via activating sweet taste receptors in the small intestine [4,5]. The paradoxical association between the consumption of AS and weight gain mainly focuses on the feed intake, glucose absorption, and gut flora. The importance of nutrient-induced brain-gut signaling in the regulation of animal metabolism and food intake has become increasingly obvious. However, the mechanisms mediating the physiological effects of sweeteners in the gut-brain axis are still unclear.Gut microflora not only influences intestinal function, but also plays an important role in regulating brain-gut axis signals [6]. An earlier study found that saccharin intake reduced the ratio of anaerobic/aerobic microorganisms in the gut of rats using the first-generation DNA sequencing technology [7]. However, which specific microorganisms are involved is still unclear. Short-chain fatty acids (SCFAs), deriving from bacteria-dependent hydrolysis of fibers, are essential to intestinal epithelial cells, which can modulate their proliferation and differentiation, to influence gut motility and to strengthen the gut barrier functions as well as host metabolism [8]. Furthermore, SCFAs are predicted to have an essential role in microbiota-gut-brain crosstalk [9]. Therefore, exploring the regulatory effects of AS on mammalian growth and metabolism requires comprehensive consideration from the perspective of the microbiota-gut-brain axis. Diversity in sweet taste sensation among mammal species occurs also at the level of the gut [10]. Since guinea pig is sensitive to external stimuli, it is a suitable animal model for AS research. What is more, the diet of guinea pig is mainly composed of plant-derived fiber, which makes the content of SCFAs in the intestine extremely vulnerable [11]. Mallett et al. suggested that adding saccharin in the diet could alter the ratio of SCFAs in the cecum of mice and inhibit the hydrolysis of amylase [12]. Therefore, we hypothesized that saccharin sodium might regulate gut development and energy intake through SCFAs. In this experiment, we aim to explore the regulatory effect of saccharin sodium (SS) on the microbiota-gut-hypothalamus axis, which can provide a theoretical basis for studying the effect of AS on adolescent development.2. Materials and Methods2.1. Animal, Diets, and ManagementTwo feeding trials were conducted for 28 days: a control group which received normal water and a SS group which received water with 5 mM SS solution. The dose of SS was designed the same as in the previous experiments [13]. A total of 12 (4-week-old) female Harley-white guinea pigs (Cavia porcellus) with a body weight of 240.7 ± 7.7 g were housed in the Laboratory Animal Research Center of Zhejiang Chinese Medical University (Hangzhou, China) and in accordance with guidelines approved by the National Laboratory Animal Management Regulations (SYXK(ZHE)-2018-0012) and the Institutional Animal Care and Use Committee of Zhejiang Chinese Medical University (IACUC- 20181224–14). Only female guinea pigs were chosen in this study to observe specific effects on female animals. The environmental conditions were set the same as in the previous study [14]. The room temperature was kept at 21 to 23 °C, relative humidity of 30–40%. The treatments were randomly assigned to six replicates each. All guinea pigs were fed the same meal every day in the morning and body weight data were collected every 7 days.2.2. Sample CollectionAt the end of the experimental periods, all guinea pigs were euthanized by CO2 anesthesia. All guinea pigs were in the luteal phase on the day of sampling to ensure consistent physiological conditions. Blood was collected immediately from the heart and serum was then separated by centrifugation at 1118 g for 10 min at 4 °C. Serum biochemical indices, including TG (triglyceride), ALT (glutamic-pyruvic transaminase), AST (glutamic-oxalacetic transaminase), TC (total cholesterol), CRE (creatinine), ALP (alkaline phosphatase), UA (uric acid), GLU (glucose), CHO (cholesterol), and TP (toll protein) were measured using assay kits (Unicel DXC 800, Breaa, CA, USA). After the hypothalamus and ileum mucosa were collected, the samples were repeatedly washed with DEPC sterilized water and stored in liquid nitrogen and then stored at −80 °C for RNA-Seq and RT-qPCR. The hypothalamic tissue samples were homogenized with a low-temperature high-throughput grinder to extract RNA. The sample collection of ileum followed methods used in previous study [15].2.3. Analysis and Observation of Intestinal MorphologicalThe ileum segments were fixed in 4% paraformaldehyde for hematoxylin-eosin (HE). The averaged villus height and crypt depth were measured by using ImagePro Plus software version 6.0 (Media Cybernetics, Rockville, MD, USA). For each slide, morphological analyses and observation were performed at magnifications of 50 times.2.4. Lactic Acid and SCFA Detection in the Ileal DigestaThe samples were added into 2 mL of water (1:3 phosphoric acid aqueous solution) and vortex and homogenate samples for 2 min. Then, 1 mL of ether was added in the samples to extract for 10 min, 4000 rpm/min and centrifuge for 20 min under 4 °C, followed by adding 1 mL of ether into solutions to extract and centrifuge for 10 min at 4000 rpm. The two extracts were combined and volatilized to within 1 mL and analyzed for further detection. The concentration of lactic acid (9–5000 U/L) was measured using commercial assay kits (A020-2-2) purchased from Nanjing Jiancheng Bioengineering Institute, Nanjing, China. SCFAs were quantified using a gas chromatography/mass spectrometry analysis (GC/MS, Thermo Fisher ISQTM LT, Thermo Fisher Scientific, Waltham, United State). The SCFAs were separated on a TG WAX capillary column (30 m × 0.25 mm × 0.25 μm) using helium (He) as a carrier gas at a flow rate of 1 mL/min. Column temperature: 100 °C (5 min) −5 °C/min −150 °C (0 min) −30 °C/min −240 °C (30 min). Fatty acid profiles were measured and are expressed in milligrams per gram of total SCFA in the ileal digesta.2.5. RNA IsolationHypothalamus and ileal mucosa samples were randomly selected for RNA isolation following methods in previous studies [15]. Total RNA was extracted using Trizol reagent (15596018; Invitrogen LifeTechnologies, Carlsbad, CA, USA) according to the manufacturer’s protocol with the nucleic acid/protein quantitative measuring instrument (Bio-Rad Laboratories, Inc., Hercules, CA, USA). For each individual hypothalamus sample, 3 μg of RNA were used for samples preparation and a NEBNext1 UltraTM RNA Library Prep Kit for Illumina1 (New England Biolabs, Inc., Ipswich, MA, USA) as sequencing libraries was used according to the manufacturer’s protocols. The cDNA was selected at 250–300 bp in length and PCR products were purified by performing AMPure XP system (Beckman Coulter, Beverly, MA, USA). The quality of the library was evaluated on the Agilent Bioanalyzer 2100 system (Agilent Technology Co., Ltd., Santa Clara, CA, China).2.6. RNA-Seq AnalysisThe raw data in FASTQ format are processed through internal Perl scripts, and clean data are obtained by removing low-quality data or adapter sequences from the raw data (FASTX-Toolkit). The filtering criteria are adapters, low-quality reads (Qphred <= 20, phred = −log10(e), e(base-calling error rate) = 0.01) at the 3′end, reads with fuzzy N bases, rRNA sequences shorter than 20bps. In order to allow for less than two base mismatches, all double-end data from the two groups were compared with the guinea pig reference genome (Cavia porcellus 3.0, Scaffolds). The reads per kilobase per million (RPKM) were used to calculate gene expression intensity. For this study, if RPKM in one group is >0.3, but is <0.3 in the other group, the gene is only expressed in one of the two groups. The DESeq2 R package (1.16.1) were used to analyze the differential expression of the two groups and the resulting P values needed to be adjusted to control the false discovery rate by using the Benjamini-Hochberg method [16]. The genes were defined as differently expressed genes (DEGs) when the gene expression levels had a p value <0.05, fold change >1.5 (increased 1.5 times compared with the control) or 0< fold change <0.67 (decreased 1.5 times compared with the control).2.7. Genes IdentificationGene ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) were used for performing enrichment analysis of the significant overrepresentation of GO items and KEGG pathway categories.2.8. qRT-PCRAlthough RNA-Seq is the core technology of gene expression profiling, qRT-PCR is still the preferred technology for verification. qRT-PCR were performed because qRT-PCR validation of microarray involves hybridization to a glass slide and the dynamic range of microarrays has long been known to be constrained compared to qRT-PCR. When RNA-Seq data are based on a small number of biological replicates, performing qRT-PCR on more samples and focusing on some interesting goals is a good way to verify RNA-Seq results and construct studies. Ten DEGs were randomly selected for performing qRT-PCR. Total mRNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and was reverse-transcribed using the cDNA reverse transcription kit (TaKaRa, Dalian, China) with gDNA Eraser. The one-step real-time RT-PCR (ABI7500; Applied Biosystem, Carlsbad, CA, USA) was performed using SYBR Premix Ex TaqTM and was used to quantify the mRNA expression levels. Supplementary Data 1 show the primer sequence (Primer Express 3.0.1 software). The relative quantitation of a given gene expression was adjusted as the housekeeping gene, β-Actin, using the 2−ΔΔCT−method [17], and normalized to the control.2.9. 16 S Ribosomal DNA Gene SequencingFive samples of ileal digesta were chosen for 16S rDNA gene sequencing analysis and a QIAampDNA Stool Mini Kit (Qiagen Inc., Valencia, CA, USA) was used for DNA extraction, following methods used in previous studies [15]. Barcode-specific primers (16S V4, 515F, GTGCCAGCMGCCGCGGTAA; 806R, GGACTACHVGGGTW TCTAAT) were used to amplify different regions of the 16S rRNA gene. Quantitative Insights Into Microbial Ecology software (v1.7.0, http://qiime.org/ (accessed on 21 December 2020)) was performed to analyze the sequences. The quality control removes the linker sequence contained in the reads, and removed the low-quality reads with a base error rate less than 0.01 (Qphred <= 20). Clean reads refer to the sequence that is finally used for subsequent analysis after filtering the chimera. UPARSE software (v7.0.1001) was used to analyze sequences and they were clustered into operational taxonomic units (OTUs) at a similarity level of 97%. Each operational taxonomy unit (OTU) will have a representative sequence and use the ribosome database item classifier to perform classification. UCHIME was used to identify and remove chimeric sequences. The most abundant sequence in each OTU is designated as representative sequences and ribosomal Database Project (RDP) classifier was used to classify. Ace, Chao, Shannon, and Simpson diversity indices were used to calculate the Alpha calculation index of community richness. Beta diversity was assessed by principal component analysis (PCA) and the significance of separation was analyzed by ANOSIM (R, v2.15.3). The functional potential of bacteria communities was predicted by using Tax4Fun analysis.2.10. Statistical AnalysesData were presented as the mean ± SD or mean ± SEM (for gene expression). The significant differences were evaluated using test in SPSS 11.0 for Windows. Comparisons of the relative abundances of microbia between the two groups were performed by unpaired Student’s t-test. Significant difference was declared when p < 0.05. The significance level for difference was p < 0.05. The Tax4Fun analysis package was used to compare the sequencing data with the bacterial metagenomics database (SILVA Reference data) for microbial function prediction. The analysis process refers to previous experiment [18]. LEfSe (LDA Effect Size) is used to analyze high-dimensional biomarkers (genes, pathways, and taxa), which can figure out the biological correlation and biomarker with statistical difference between groups.3. Results3.1. Serum Biochemical IndexesWe first assessed the effect of SS on the serum metabolic indices of guinea pigs (Table 1). The SS group resulted in poorer (p < 0.05) UA and GLU than the CN group. However, SS supplementation had no significant effects on serum levels of TG, ALT, AST, CRE, ALP, CHO, and TP. The gastrointestinal tract participates in brain–gut interactions through brain–gut peptides, mainly including growth hormone-releasing peptide (GHRP), glucagonlike peptide 1 (GLP-1), cholecystokinin (CCK), peptide-YY (PYY). In the current study, adding SS to drinking water increased the serum level of GHRP (p < 0.05) with no significant effects on PPY, GLP1, and CCK.3.2. Morphological Analysis of the Ileum VillusAs is shown in Figure 1, the villus height and villus height/crypt depth ratio were raised by supplementing with SS (p < 0.05, Figure 1a,c), along with the decreased crypt depth compared with the CN group (p > 0.05, Figure 1b). SS supplementation improved the ileal villus density compared with the CN group based on the photomicrographs of the ileum cross section (Figure 1d).3.3. Gene Expressions of Taste Receptors, Glucose Transporter and Tight Junction in the Ileal MucosaTaste receptor type 1 member 2/3 (T1R2/3) is the sweet receptor expressing in the ileum. In the present experiment, SS treatment upregulated the mRNA expressions of T1R2, T1R3, and downstream genes (PLCβ2, phospholipase Cβ2; TRPM5, transient receptor potential cation channel subfamily M member 5; Figure 2, p < 0.05). Furthermore, SS treatment exerted significant main effects on upregulating mRNA expressions of solute carrier family 5 member 1 (SGLT1, p < 0.05), but with no effects on GLUT2. The tight junction protein is the major indicator to evaluate the intestinal barrier function. The mRNA expressions of tight junction protein 1 (ZO1) and claudin 1 (CLDN1, p < 0.05) in the ileal mucosa of SS groups were much higher than those in the CN group. However, there were no significant effects of SS on claudin 2 and occludin.3.4. The Levels of Lactic Acid and SCFAs in the Ileal DigestaAs we can see in Figure 3, adding SS to drinking water significantly increased the level of lactic acid in the ileal digesta (p < 0.05). Furthermore, a significant increase of acetic acid, propionic acid, and N-valeric acid was similarly observed in the ileal digesta of SS-treated guinea pigs compared with the CN group (Table 2, p < 0.05). Meanwhile, the levels of isobutyric acid (p = 0.069) and isovaleric acid (p = 0.092) in the ileal digesta of the SS group were higher than that of the CN group. There are no significant effects of SS on the levels of N-butyric acid and N-hexanoic acid in the digesta.3.5. Effects of the Saccharin Sodium InterventionSequence data showed that SS supplementation had an impact on the microbial composition of ileal digesta. Each group tested 6 samples, sequence data with >25,000 were considered to be included, which resulted in 5 samples for each group for final analysis. Furthermore, the number of average raw sequence reads detected in each group exceeded 85,411 (Supplementary Data 2). In order to study the species composition of each sample, OTUs (operational taxonomic units) clustering was conducted with 97% identity for the effective tags of all samples, which yielded a total of 6231 OTUs for the entire dataset (Supplementary Data 2). There is no major difference between the diversity indices (Shannon and Simpson) and richness estimators (Ace and Chao) of ileal digesta microbiota in the CN and SS groups (Figure 4a). The PCA analysis indicated that the ileal microbiota composition exhibited significant changes after SS treatment, with a significant separation between the CN and SS groups (t-test, p < 0.05) (Figure 4b).The bacterial composition was evaluated at different taxonomic levels. The dominant bacterial groups were Firmicutes, Bacteroidetes, Actinobacteria, Spirochaetes, Euryarchaeota, and Proteobacteria at the phylum level. The abundance of Firmicutes tended to decrease in the SS group in comparison with the CN group (Figure 5a,b, p = 0.078). In contrast, the abundance of Bacteroidetes in the CN group is lower than that in the SS group (Figure 5a,b, p = 0.129). The relative abundance of Muribaculaceae (p < 0.05) and Lactobacillaceae (p < 0.01) dramatically increased in the SS group when compared to the CN group (Figure 5c,d) at the family level. In contrast, SS treatment significantly decreased the relative abundance of Erysipelotrichaceae and Eubacteriaceae (Figure 5c,d, p < 0.05). At the genus level, the relative abundance of Lactobacillus was significantly increased after SS treatment, along with a decrease in the relative abundance of Ileibacterium (Figure 5e,f, p < 0.01).As is shown in Figure 5g,h, Bacilli, Lactobacillales, Lactobacillaceae, and Lactobacillus were the biomarkers in the SS group distinguishing them from the CN group. Meanwhile, Brevinema-andersonii, Brevinemataceae, Brevinema, Brevinematales, Eubacteriacae, ilebacterium, Clostridia, Clostridiales, Erysipelotrichaceae, Erysipelotrichia, and Erysipelotrochales more widely exist in the digesta of the CN group compared with the SS group.Tax4Fun analysis was used to evaluate the changes in the presumptive functions of the ileal microbiota of guinea pigs. Figure 6a shows that the top 25 predicted microbial functions at the second level of the GO enrichment. Chemoheterotrophy, fermentation, methanogenesis, hydrogenotrophic_methanogenesis, methanogenesis_by_CO2_reduction_with_H2, dark_hydrogen_oxidation, nitrate_reduction, animal_parasites_or_symbionts, mammal_gut, ureolysis have top predicted functions with rich Taxa. SS treatment significantly enhanced the functions of chemoheterotrophy and fermentation of ileal microflora compared to the CN group (Figure 6b, p < 0.05). The abundance of GO terms related to ureolysis was significantly reduced (Figure 6b, p < 0.05). The top abundant microbial pathways enriched into second-level functional categories were about cellular process, environmental information processing, genetic information processing, human diseases, metabolism, organismal system (Figure 6c). Among them, cell growth and death, signaling molecules, carbohydrate and lipid metabolism, nervous system, immune system, and endocrine system were significantly enriched after SS supplementation.3.6. Hypothalamic RNA Sequencing Data and Identification of DEGs in the Hypothalamus between GroupsTo study the gene expression profiles in the hypothalamus, we performed RNA-seq analysis of SS-treated guinea pigs and the controls. Note that 52−64 million (M) clean reads from eight RNA-Seq libraries were received to obtain more references to get concise reads with a high proportion of mapped reads ranging from 89.54% to 91.31% after removing the low-quality and adaptor sequences. Most mapped reads were located within an exon (67.01−71.05%), and fewer mapped reads were located within the introns (4.62−6.43%) and intergenic (22.52−26.87%) regions (Supplementary Data 3). These results demonstrated that, compared with the hypothalamus transcriptomes between the SS and CN groups, these 8 libraries had high quality and high coverage of the guinea pig genome.In the hypothalamus of the SS and CN groups, we identified 1370 DEGs, among which 796 genes were upregulated and 574 genes were downregulated (Figure 7c, |Fold change| >1.5, p value < 0.05). All the information of annotated gene and DEGs are shown in Supplementary Data 4. There were 610 upregulated DEGs and 394 downregulated DEGs identified in the KEGG and GO databases using Omicbean analysis software. PCA and hierarchical clustering analysis revealed that 3 SS samples were clustered together and were distinctly different from the clustering of 3 controls (Figure 7a,b). These data demonstrated that SS exhibited unique gene expression profiles in comparison with that of controls.3.7. Function Enrichment Using DEGs of Hypothalamus RNA-SeqTo further characterize the enriched functions, we performed Gene Ontology (GO) analysis using the upregulated DEGs (Supplementary Data 5). Figure 8a indicates the top 10 enriched terms of biological process, cell component and molecular function at all the GO annotated levels. Importantly, many of these enriched GO terms are closely related to nervous system development (GO:0007399), cell–cell signaling (GO:0007267), neuron part (GO:0097458), synapse (GO:0045202), protein binding (GO:0005515) and cation binding (GO:0043169). At the 6th level (Figure 8b), we can see in detail that DEGs were mostly enriched into trans-synaptic signaling (GO:0099537), generation of neurons (GO:0048699), brain development (GO:0007420), plasma membrane bounded cell projection morphogenesis (GO:0120039), synaptic vesicle exocytosis (GO:0016079), synaptic vesicle localization (GO:0097479), cell morphogenesis involved in neuron differentiation (GO:0048667), cellular protein modification process (GO:0006464), organic hydroxy compound transport (GO:0015850), ion transmembrane transport (GO:0034220), secretion by cell (GO:0032940) monoamine transport (GO:0015844), regulation of plasma membrane bounded cell projection organization (GO:0120035), cation transport (GO:0006812), regulation of catecholamine secretion (GO:0050433), and positive regulation of cell development (GO:0010720). KEGG annotation was used to identify the function of enriched DEGs in the signaling pathways. Figure 8c indicates that DEGs were annotated to 13 KEGG pathways in the hypothalamus (p value_adjusted < 0.05). Among them, cAMP signaling pathway, MAPK signaling pathway, Ras signaling pathway, phosphatidylinositol signaling system, and neuroactive ligand–receptor interaction were closely related to signal transduction and signaling molecule interaction. Furthermore, insulin secretion, axon guidance, cocaine addiction, taste transduction, nicotine addiction, phosphatidylinositol signaling system, and amphetamine addiction have the most number of DEGs, indicating that these pathways play important roles in the endocrine system, nervous system, sensory system, and development. In addition, SS treatment enhanced renin secretion, which was related to the hypothalamic thirst by angiotensin stimulation. We validated the expression of 9 genes in these related pathways using real-time qRT-PCR and demonstrated that Cacna1c, Gabra5, Hcn4, Htr1b, Kcnb1, Scn2a, Bdnf, Creb, and Slc18a2 were increased in SS compared with the control.Protein–protein Interaction Analysis for all upregulated DEGs using Cytoscape bioinformatics was performed (Figure 8d, Supplementary Data 5). The major biologic functions r were significantly related to the categories’ neuroactive ligand–receptor interaction, MAPK signaling pathway, Ras signaling pathway, insulin secretion, axon guidance, cocaine addiction, taste transduction, nicotine addiction, cAMP signaling pathway, and Amphetamine addiction. Among them, Cacnalc, Creb1, Bdnf, GSK3b, Scn2a, Grin2a, Slit1, Abl1, Oxtr, Agtr1, Gnrhr, Chrm5, Tacr3, and Grp are related to the most abundant genes, which can help explain why SS treatment can activate taste-signaling transduction.4. DiscussionSaccharin sodium (SS), as a type of AS, is widely used in human food. Under normal dietary conditions, AS can regulate energy balance through the neuroendocrine system [19]. However, it is unclear whether SS has direct or indirect effects on animal metabolism and growth. The relevant data published by our team revealed that adding 1.5 mM SS to drinking water increased the feed intake and body weight gain of guinea pigs [14]. In the present study, SS treatment significantly increased the serum level of glucose in comparison with the CN group, this increase may affect the energy intake and body weight of guinea pigs. Previous research suggested that the diet with 5% AS caused a significant increase in blood glucose and metabolic disorders in rats [20,21,22]. In the current experiment, SS treatment had no significant effect on the serum levels of TG, CHO, AST, and ALT, indicating that the increase in blood glucose was still within the tolerance range of guinea pigs, and induced no obviously metabolic disorders and abnormal organ function. This difference might be related to the different dose of SS and the tolerance of guinea pigs from other animals.Sternini et al., have proved that the taste receptor exists in the intestinal endothelial cells and endocrine cells, which can regulate the intestinal cavity and food intake [23]. In the present experiment, SS treatment significantly increased the transcriptional expressions of T1R2, T1R3, and downstream genes (PLC-β2, TRPM5). It is demonstrated that sweet receptor T1R2/T1R3 can increase the glucose absorptive capacity in response to AS through regulating SGLT1 expression [24]. Consistently, drinking water with SS significantly upregulated transcriptional levels of SGLT1 in the ileum, which could explain why blood glucose in the serum increased. The presence of AS can initiate gut-brain signaling via the release of gut hormone and concomitant activation of vagal afferents [25]. Importantly, SS treatment increased the serum level of GHRP. GHRP, as the only type of appetite-stimulating brain gut peptide, is known to promote gastric acid production, gastric motility and emptying. Furthermore, as an endogenous ligand of growth hormone secretagogue type I receptor (GHSR), GHRP can stimulate the pituitary gland to release growth hormone and promote appetite [26]. Therefore, SS treatment might increase the feed intake and body weight through activating sweet receptor signaling and brain–gut hormone secretion.The hypothalamus is at the core of homeostatic control, which integrates signal input involved in eating behavior [27]. Ventromedial hypothalamic area (VMH) is known as the “satiety center”, expressing orexigenic factors [28]. In the current study, adding SS into the drinking water promoted hypothalamus nervous system development, trans-synaptic signaling, generation of neurons, synaptic vesicle exocytosis, synaptic vesicle localization and sensory system development. Furthermore, SS treatment enhanced expression of genes involved in this processes which was closely associated with signal transduction and signaling molecule interaction. ARC neurons, expressing the neurotransmitter NPY and agouti-related peptide (AGRP), signal to stimulate feeding [29]. Importantly, SS treatment upregulated the expression of AGRP in the hypothalamus in comparison with the control. Furthermore, SS treatment upregulated the transcriptional levels of CREB1, GSK3B, SLC18A2, and Cacna1c, which were enriched into dopaminergic synapse process. Therefore, SS treatment can activate taste signaling transduction, and enhanced neuroactive ligand–receptor interaction through altering the RNA profile in the hypothalamus.As mentioned above, the optimal adding level of SS had not resulted in obviously metabolic disorders in guinea pigs. Energy intake is not the only one impact factor impacted by SS treatment, there must be other regulatory effectors. The gut microflora is an important participant in the brain–gut axis system, known as the “second brain”, mediating neural signaling exchange and growth development [30]. In the present study, the addition of SS increased the abundance of Firmicutes in the ileum, while reducing the abundance of Bacteroidetes. Bacteroidetes and Firmicutes are the dominant bacteria in the mammal gut. Several studies have suggested that the abundance of Firmicutes in the gut is directly proportional to body weight gain, while Bacteroides is the opposite, which is consistent with the growth phenotype in this experiment [31,32]. Therefore, we concluded that SS might promote the growth of guinea pigs by regulating the gut microflora. What we stress here is that SS treatment increased the family abundance of Muribaculaceae and Lactobacillus in the ileum, followed by the increased level of lactic acid. Sufficient evidence revealed that Lactobacillus can maintain the micro-ecological balance, inhibit pathogen proliferation in the intestine, and improve immune function [33,34,35]. Bacteria in the family S24-7 (phylum Bacteroidetes) are dominant in the gut microbiota of mouse and have been detected in the intestine of other animals [36]. The latest research indicated that S24-7 plays an important role in complex dietary carbohydrate degradation, which is consistent with the elevated blood glucose after SS treatment in this experiment. In contrast, adding SS to drinking water reduced the abundance of Erysipelotrichaceae and Eubacteriaceae, among which Ileibacterium decreased significantly. In particular, the family Erysipelotrichaceae is emerging as a group of bacteria that may affect host metabolism and inflammatory diseases, and closely related species have been associated with obesity or protection from weight gain on a high-fat diet [37,38,39]. LEfSe analysis suggested that Lactobacillasceae-Lactobacillus was the biomarker of ileum microorganisms in the SS-treated guinea pigs, while Brevinema-Andersonii and Erysipelotrichaceae-Ilebacterium were the biomarkers of the CN group. This is consistent with the abundance analysis of microflora above. It is further suggested that SS treatment could promote the proliferation of Lactobacillus, and reduce the abundance of Ilebacterium in the ileum, which plays an important role in regulating the balance of gut microflora. Gut microbiota will influence liver, brain, and even the metabolism of muscle tissue, finally affecting the whole energy metabolic network in the host (Schroeder and Bäckhed 2016). In the present research, KEGG pathway analysis suggested that SS treatment influenced cell growth and death, signaling molecules, carbohydrate and lipid metabolism, the nervous system, immune system, and endocrine system. In addition, the SS group exhibited significantly less abundance of GO terms related to ureolysis. This was consistent with the decrease of serum urea in SS-treated guinea pigs. Therefore, gut microflora plays an essential role in the regulation of growth and metabolism after SS treatment.Gut microbiota is an essential barrier for the host and plays vital roles in the digestion and fermentation of carbohydrates, intestinal villi development, and immune response [40]. The metabolites produced by the bacterial fermentation of diets, namely SCFAs, play an essential role in linking host nutrients to intestinal homoeostasis maintenance. Furthermore, due to their neuroactive properties and their influence on other gut-brain signaling pathways, including the immune and endocrine systems, SCFAs may directly or indirectly participate in communication along the microbiota-gut-brain axis [9]. Mallett et al. suggested that the diet with 5% or 7.5% saccharin increased digesta content in rats, followed by the increased level of lactic acid [11]. In the current experiment, adding SS to drinking water significantly increased the lactic acid level in the digesta, which is consistent with the increase of Lactobacillus in the ileum and fermentation process enriched by GO analysis. As we all know, Lactobacillus produces organic acids, enzymes and acidophilins by fermentation process to improve nutrient potency and growth performance [41,42]. Further detection revealed that SS treatment increased the content of acetic acid, propionic acid, N-valeric acid, N-butyric acid, and N-hexanoic acid in the ileal digesta. Acetate is a fermentation product for most gut bacteria, while butyrate and propionate are produced by more specific bacterial species. Propionate is mainly involved in the process of gluconeogenesis [43], while acetate and butyrate are mainly involved in lipid biosynthesis [44]. It is reported that the amounts of acetate and propionate correlate positively with Bacteroidetes within Firmicutes [45,46]. This is consistent with the change of ileal microflora in our experiment. Through glycoside hydrolases, polysaccharide lyases, and carbohydrate esterases, gut-associated bacterial communities associated with the intestine can decompose and ferment complex carbohydrates into SCFAs [46]. Butyrate is produced from acetate, lactate, amino acids, and various carbohydrates via glycolysis. In the present study, the change of butyrate in the digesta is consistent with that of lactic acid and acetic acid, which are associated with the enriched process of complex carbohydrate degradation. Therefore, SS treatment could promote the production of lactic acid and SCFAs by regulating gut microflora. Besides, SCFAs act as specific G protein-coupled receptor (GPR) signaling molecules to regulate glucose metabolism [47]. This also explains the consistency of changes in ileal SCFA content and serum biochemical indexes after SS treatment. Furthermore, previous study indicated that SCFAs can decrease pH of the gut and restrain the colonization and proliferation of some pathogens [48]. In this current study, adding SS to drinking water significantly inhibited the abundance of pathogenic bacteria in the ileum. Therefore, adding 1.5 mM SS to drinking water is beneficial to improve the intestinal health of guinea pigs.Previous studies demonstrated that SCFAs play an essential role in improving the morphology and function of epithelial cells [8]. For example, butyrate, as the main energy source of intestinal cells, is metabolized by hindgut cells. Accordingly, we found that SS supplementation increased the villus height and villus height/crypt depth ratio. The increase in the intestinal villi height can reflect the increased number of mature villi cells and absorptive capacity. The high value of villus height/crypt depth ratio indicates that the intestine has strong digestive and absorption capacity. Previous studies suggested that SS treatment could produce more SCFAs by regulating gut microflora and improve intestinal villi morphology. Similarly, the dietary fiber is fermented, resulting in SCFAs, which promote proliferation of the mucosal epithelium and villus height [49]. Previous studies indicated that SCFAs (acetate, propionate, and butyrate) produced by the microbiota as final metabolites can improve gut barrier function [50]. Supplementing butyrate to diet can inhibit the disruption of the intestinal epithelial barrier induced by high-fat diet by upregulating CLDN1 gene expression [51]. Similarly, SS treatment significantly increased the mRNA expressions of zonula occludes protein, including ZO1 and CLDN1, which can enhance the barrier function of intestinal epithelial cells. Therefore, SS supplementation might enhance intestinal barrier function via regulating SCFAs.5. ConclusionsIn conclusion, the microbiota-gut-hypothalamus axis plays an important role in the regulatory effect of SS on the growth and glucose metabolism of guinea pigs. SS treatment activated sweet receptor signaling in the gut and altered GHRP hormone secretion, followed by the taste signaling transduction in the hypothalamus. Importantly, SS treatment increased the abundance of Firmicutes and Lactobacillasceae-Lactobacillus in the ileum, followed by the increased levels of lactic acid and SCFAs. Therefore, adding 1.5 mM SS to drinking water is beneficial to promote the growth of guinea pigs through regulating the microbiota–hypothalamus–gut axis. This finding is of theoretical significance for comprehensively evaluating the biological effects of appropriate levels of saccharin in food and drinks.

animals : an open access journal from mdpi

10.3390/ani11123542

PMC8698146

Fresh egg is a very fundamental food ingredient to consumers’ daily life, while consumers in Taiwan have faced barriers to effective decision-making for eggs they needed, including the increasing consumer concerns about food safety, traceability, and the increasing demand for animal welfare production. Consequently, Taiwanese supermarkets are eager to identify what types of fresh egg attributes would fit to their customers. Further, the effect of unrealistic options in the choice experiment design has a limited understanding. This study aims to investigate what factors affect customers’ decision-making to purchase fresh eggs in Taiwan. The findings of this research suggest that the effect of unrealistic options in the choice experiment design may mislead the results. Thus, the confirmed results showed that Taiwanese consumers are willing to pay a premium for attributes for animal welfare, traceability, farm brand, and brown color egg. Especially, the attribute of animal welfare reveals a significant effect on consumers’ decision-making. This provides an important hint to government policymakers and egg industry in Taiwan to better understand issues encompassing the diversity of attributes associated with fresh egg products offered in supermarkets.

Eggs are the crucial component of daily meals for almost everyone in Taiwan, while the multi-attributes of fresh egg products generate the challenges of marketing and promotions in supermarkets. This study analyzes the market segmentation and consumer willingness-to-pay (WTP) for fresh egg attributes (i.e., color, traceability, animal welfare, brand, and price). In particular, the effect of the unrealistic choice set is considered in this study. The data collection was distributed near markets, schools, and train stations across Taiwan from July to September in 2020. A total of 1115 valid responses were collected, and the Latent Class Model was used. Results show that fresh egg products in supermarkets reveal a strong preference for animal welfare label with the highest WTP, which is about 64.2 NT$ (≈US$ 2.29). Furthermore, traceability label, farm brand, and brown-color egg still exhibit positive WTP of about 33.4 NT$ (≈US$ 1.19), 32.6 NT$ (≈US$ 1.16), and 32.5 NT$ (≈US$ 1.16) in supermarkets, respectively. However, including the unrealistic choice set can potentially alter the final outcomes, and it provides a good example for researchers who may have the same situation. This research helps to know more about the complexity of attributes for fresh egg products in supermarkets, so marketers would be able to adopt the effective marketing strategies for fresh egg products in supermarkets.

1. IntroductionThe egg business in Taiwan has witnessed fast expansion, with strong and rising market demand for fresh egg products in Taiwan [1]. This is shown by the growth of egg sales in Taiwan from 4% to 22% between 2015 and 2018 [2]. Further, the egg consumption per capita among Taiwanese consumers is substantial, approximately 322 eggs per person per year [3]. This implies that Taiwanese consumers consume almost one egg every day. There is a high demand for eggs because eggs are considered as one of the main staple products that can provide moderate calories (about 150 kcal per 100g) and high protein for diets [4]. Not only are eggs a good source of nutrients and protein [5,6], but also eggs can play an important agent for the human body in receiving key nutraceutical elements for the dietary [7,8,9,10]. Thus, the global guideline advocates that eggs should be consumed on a regular basis as part of a healthy diet [11]. In short, eggs are regarded as a vital element of the daily diets of all consumers [11].Since eggs played a prominent role in consumers’ daily lives [12], the marketing strategies in supermarkets have attempted to promote many different attributes for fresh egg boxes to attract consumer attention [13,14]. Thus, it is imperative to identify consumer preference and their purchasing decision for fresh egg boxes. Previous studies [15,16,17,18,19] had focused on freshness, visual features, and prices. However, more attributes (i.e., color, brand, animal welfare, traceability, organic, and nutrition label) of fresh eggs have been discussed in other countries, but not in Taiwan [20,21,22,23].Consumer preferences of fresh eggs may have significant differences between nations [24]. According to Li (2013) [25], Chinese consumers prefer eggs with longer shelf life and the best-before-date information. Indonesian and Indian consumers exhibit a similar preference when it comes to purchase eggs. The findings show that these consumers pay more attention on price attribute [11,26]. Further, consumers in the United States (U.S.) prefer eggs with animal welfare and organic labels [27,28]. Additionally, the majority of European consumers are more likely to buy and pay extra for eggs with animal welfare labels [29,30]. Therefore, the attributes of egg products may have a significant impact on consumers in different countries, such as China, Indonesia, the U.S., and European countries.In general, fresh eggs in supermarkets have a series of product attributes, such as brand (providing by farm brand and private brand, e.g., Carrefour), animal welfare (giving an animal-friendly environment to enhance animal welfare for laying hens, e.g., cage-free, etc.), traceability or traceable agricultural product (providing barcodes on egg products to assist consumers in tracing and tracking the product information), and certified agricultural standard (giving a certificate that guarantees eggs are fresh, clean, food safety, etc.) [3,14,31,32]. Do more product attributes embedded on fresh eggs mean more selling? Marketers believe that product attributes would potentially satisfy consumer preferences [33]. Since more product attributes do not necessarily mean more positive influences on consumer preferences, sometimes too much information may impede decision-making and may even decrease the consumer’s desire to purchase [34,35]. Therefore, more in-depth investigation is required to determine which product attributes should be provided on fresh egg products to appropriately entice Taiwanese consumers to purchase.Since there are a series of fresh egg attributes that will be examined and the willingness to pay (WTP) for fresh egg attributes is intended to be estimated, the choice experiments would be appropriate to implement in this study [36]. When choice experiment studies are conducted, there may exist an unrealistic situation that may not be in accordance with the actual market situations [37]. In addition to the survey design of choice experiments, the evidence of unrealistic choice sets is rare to be discussed. Since the design of choice sets for fresh egg attributes contains several unrealistic choice combinations, whether the unrealistic choice sets lead to a potential bias should be considered and discussed. Hereinafter, this study further tests the including and excluding the unrealistic choice sets to see if there is a potential bias that makes the estimator failure. Hence, this study will be able to provide more literature for the unrealistic choice sets in choice experiment research method.2. Materials and MethodsA choice experiment (CE) supported by a structured questionnaire was used in this research to evaluate consumer preference, market segmentation, and willingness-to-pay (WTP) for various eggs and attribute variables in Taiwanese supermarkets. Additionally, the CE is one of the methods that can be used to reveal respondents’ preferences for different conditions, which are often restructured by researchers [38,39,40]. The target respondents in this study are the fresh egg buyers in Taiwan. The data in this study was collected through online and offline questionnaires. Finally, the data were examined using Latent Class Model (LCM), which is a statistical modeling tool that could assist researchers in calculating the prospective consumer segments and the WTP [41]. The LCM was used to ascertain which fresh egg attributes (i.e., animal welfare, brand, traceability, price, and color) should be prioritized in Taiwanese supermarkets, as well as the potential markets.2.1. Questionnaire Design and SampleIn general, the methods of the CE, Conjoint Analysis (CA), and a Contingent Valuation Method (CVM) are often adopted to assess the stated preference data. The CVM method has been extensively used to estimate WTP for a specific product attribute, as well as a product that does not yet exist on the market [42]. However, the CVM is ineffective when estimating the consumer WTP with a bundle of attributes. Thus, in assessing consumer WTP for individual attribute parameters, the CE and CA are more appropriate compared to CVM [43]. However, the CE and CA still have some differences, such as the theoretical foundation. For the CE, the researchers asked respondents to select one from two or more alternatives in the questionnaire, while the researchers adopt the CA by asking the respondents to rate or rank the alternative [40]. In addition, the actual decision-making behavior is more closely to choose rather than rating or ranking. Thus, the CE method would be more appropriate to construct empirical studies for the choice paradigm. In short, the CE method is utilized to achieve the objective in this study.Sample serves a critical role in research because it is utilized to make inferences about a population. Fresh egg buyers could be very diverse, but the majority of fresh egg consumers are those who are frequently visiting traditional markets, supermarkets, or hyper-supermarkets [14]. Those consumers may also be in charge of food grocery purchases for the family. Therefore, the sampling method should pay extra attention to selection bias [44]. This study especially focuses on fresh egg attributes on animal welfare, brand, traceability, price, and color, so the suitable respondents should be identified.The questionnaire design in this study consists of four portions: screen questions, shopping background, the choice set questions of the CE method, and the social demographic questions. The sampling methods used in this study were offline and online. In order to fully control the selection bias and proceed with the quantitative analysis [45,46,47,48], both offline and online respondents were asked to fill out a survey link that was managed by SurveyMonkey, Inc. Thus, the survey link gave two screening questions to confirm whether respondents are targeted in this study. The first screening question is: “Are you the one in charge of grocery shopping in your family?” The second screening question is: “Have you bought any fresh raw eggs at least once in the last six months?” Respondents who are either from offline or online were able to be identified via the screening questions. The offline method was distributed in public places, such as markets, front-gate of schools, parks, and public transportation stations, and the online method was distributed through the Facebook advertisement. In order to encourage potential respondents to join the survey, a total of 150 pieces of 7–11 gift cards (valued 100 NT$ (≈US$ 3.57)/each) were provided as a lucky draw to get more responses. A total of 1555 respondents participated in our survey sampling event from July to September in 2020, while only 1115 observations are valid in this study.2.2. Choice ExperimentsThe CE was conducted with fresh egg products in this work, as many Taiwanese consumers regularly consume approximately 322 eggs every year [3]. Hereinafter, the hypothetical discrete choice experiments in this study would determine what consumers need in the market, and their WTP for different fresh egg product attributes [49]. To mitigate the hypothetical bias, respondents in this research were encouraged to pay for the product that they would select in actual market circumstances [50]. In the CE, respondents were asked to make a choice out of different attribute alternatives [50]. In the scenario, the respondents must choose one of the options in the choice sets; however, if none of the options was of interest, they could also select a none-of-these (NOT) or no-buy option.A no-buy option must be included in the CE model to make the decision of food purchase more realistic [51]. In addition, based on the Random Utility Theory (RUT), a respondent in the CE must choose one of many options that will maximize their utility [52]. Lancaster (1996) [53] reported, in consumer theory, the maximum utility of the product is derived from several product attributes. Thus, it is crucial to investigate the utility of each attribute in the product.If compared to other WTP analysis methods (e.g., auctions and contingent valuation method), the advantage of the CE is that the CE format would be more similar to an actual market situation [54]. Further, this study adopts the “D-efficient” design to minimize the determinant of the covariance matrix for the parameter estimator. However, the results from the D-efficient design often produce some unrealistic combinations of attributes that contradict with the actual market situations [55]. When unrealistic choice sets exist, it is worthy to examine the differences. Thus, this study would like to investigate whether the unrealistic choices may reduce respondents’ interest in our research and cause potential biases to enlarge the estimator variance in this empirical study.2.3. Attribute and Level SettingsOne of the benefits of the CE model is that researchers could analyze consumer preferences for a product in the market [56]. The CE model also assigns consumer trade-offs amongst multi-attributes concepts in order to understand consumer preference [11]. There were eight choice sets produced by the R software at the time of designing the CE in this research. Before inputting the choice sets into the questionnaire, these eight choice sets were reviewed with industrial experts. There were two choice sets that indicated an unrealistic situation in light of real market circumstances. Thus, it is conceivable that if these eight choice sets can be sufficiently examined by the R software, what influence would be generated when an unrealistic choice set is included or excluded in the data analysis? As a result, this research will also examine the effect of including and excluding unrealistic CEs in the analytic model.Additionally, this part will also discuss the design of the attributes and level settings used in this study. Next, in order to emulate a real decision-making scenario, the five attributes (i.e., brand, color, traceability, animal welfare label, and price) are often presented in supermarkets. The five attributes with different levels are shown in Table 1. The attribute of brand is defined as farm brand and private brand (i.e., Carrefour). The Carrefour private brand is the only private brand attempting to promote their fresh egg emphasizing cage-free eggs, which is similarly linked to the concept of animal welfare egg [3,57]. The color attribute has two levels, which are brown and white.Since the Taiwanese government has enhanced food safety by promoting traceability in fresh egg products [14], the attribute of traceability is split into two levels: provided QRcode for traceability information and no QRcode for traceability information. Provided QRcode for Traceability may assist consumers to access the product information such as the farm owner’s name, location, management methods, the date of harvest, etc. Only when the chicken farm meets the government regulations will it be able to provide a QRcode for traceability. Further, the attribute of animal welfare label is defined as: provided animal welfare certified label and no animal welfare certified label. The animal welfare label is a certificate that requires an authorized third party to assess egg production meeting the regulations.In order to evaluate consumer preferences, price is the most important factor [58]. However, the prices of fresh egg products are often different from each other in markets. In the design of the choice set, it is crucial to set the price to avoid the amount of level effect [59]. Based on market research in Taiwanese supermarkets, to avoid the amount of the level effect, the four different pricing levels were determined for one box of egg (i.e., 10 eggs): 60 NT$ (≈US$ 2.14); 95 NT$ (≈US$ 3.39); 130 NT$ (≈US$ 4.64); and 165 NT$ (≈US$ 5.89).The R software with Support.CEs package was used to assist the development of fractional design by selecting a profile that balances the independent effects of all egg attribute effects. Theoretically, there are 64 possible combinations that can be utilized to create the CE for this study. Following that, the total eight choice sets were finalized and created by using the “rotation.design” and “questionnaire” functions contained in the Support.CEs package. The example of a choice set is shown in Figure 1, which provides three options (option A, B, and C) and each with various attribute combinations. If respondents would not like option A or B, then they could choose option C to refuse both scenarios.After the set-up of choice sets, it is recommended that the choice sets should be discussed with an industrial expert to see whether the choice sets do not exist in an irrational scenario. When the choice set is excessive or unreasonable, it may influence respondents’ interest and engagement with an irrational decision as well [60,61]. Therefore, an interviewed discussion with the expert who serves in the National Animal Industry Foundation in Taiwan was implemented in June 2020. There were two choice sets identified as irrational situations, and the example is shown in Figure 2. Since a box of fresh eggs only with private brand attributes is priced at 165 NT$ (≈US$ 5.89), it was indicated by the expert that the scenario would never happen. Thus, these types of unrealistic choice sets are considered and placed as the last scenario for the CE testing.2.4. Theoretical and Econometric Model UsedIn general, each fresh egg product has differences in its physical characteristics, such as product packaging, size, and shell color [62]. Moreover, specific attributes contained in fresh egg products may also influence consumers while making a purchase decision [53,63]. Therefore, in order to analyze the market segmentation as well as to assess consumer preferences and the WTP for fresh egg attributes in this study, the Random Utility Theory (RUT) is adopted to describe the consumer utility of product attributes [64,65,66,67]. In addition, the RUT can play a crucial role in supporting the CE model based on the fundamental principles of economic theory [53,68].The RUT is the model that can help researchers to capture the mobility choices of consumers. In RUT, consumers are assumed to be decision-makers, and they can often maximize their utility based on their choices or preferences [69]. Thus, in this study, it is assumed that consumers are rational and decide to optimize their utility based on the price restriction and the budget that consumers have. Further, each consumer is requested to select an alternative that was given in CE and expressed in vector notations in line with the mathematical model as:(1)Uin=βk Xijk+εij where Uin is the utility from the alternative situation “n” for the participant “i”. The homogeneous factor of the coefficient among participants can be indicated by “β” and the “Xijk” represents the “k” attributes that are used in the alternative “j” for the “i” participant. The “εij” presents the random residuals, which is an unknown deviation for the participant “i”.2.5. Econometric ModelsThe egg-purchasing scenario in a hypothetical market was simulated using discrete choice experiments (DCE). The DCE model was selected in this study because it can assist the researchers in observing consumer preferences for fresh egg products as well as product attributes and is a well-known model among researchers [70,71]. The DCE has also been shown to be capable of predicting consumer behavior through simulating consumers’ purchasing decisions [36].Respondents in this study were requested to choose an alternative fresh egg product that they would like to purchase or select from the scenario in the CE. Further, to express the consumer preferences, participants were forced to make a trade-off among different attributes and levels of fresh egg products in CE. Taiwanese consumers are heterogeneous in buying fresh egg products as assumed in this work [72] and different in error variation. Thus, it is crucial to consider individual preference and heterogeneity in the modeling process. Nowadays, several statistical models have been used to generate heterogeneous preferences in the DCE, such as the Latent Class Model (LCM) and the Mixed Logit Model (MLM) [73].The LCM was used in this research to evaluate the heterogeneity of consumer preferences and market segmentation [39] for the attributes contained on fresh egg products. In order to create the CE, five attributes were selected, namely brand, color, traceability, animal welfare egg label, and price. In the LCM, participants were assumed to belong to a group with a particular probability Cis for s = 1, …, S (where Cis > 0, while ∑Cis = 1; and S indicates the total number of groups). Hence, the probability of each class membership of the segment from the LCM will be explained in the statistical formula below:(2)Cis=expαλS∑s=1SexpαλS where the vector for specific market segmentation in the LCM is represented by λS; besides this, it is also assumed that the scale factor α = 1. Thus, each respondent only has the probability of being part of a certain segment in the LCM [74].To analyze the LCM, respondent i’s preference probability for alternative scenario (hypothetical market) j in the CE t can be demonstrated as below [38,39,40]:(3)Pijt=∑s=1SCis=expβsXijt∑j=1JβsXijtThe maximum likelihood approach is used in this study to analyze the LCM. Moreover, this model not only can estimate preferences for different consumer groups but can also provide the probability of each group or class share for each consumer group. The LCM has a function of classifying respondent i into a group class s and assumes that the stochastic error term in the membership probability function is i.i.d. (i.e., independent and identically distributed) across respondents and groups or classes.In prior research, it has been shown that consumers have preferences that may be classified into classes or groups [39]. Therefore, the LCM model was chosen for this research since it can compute consumer preferences for each class and assume that consumer preferences differ by segments [39]. Given the membership in class S, the probability of membership in each latent class S can be calculated as the following formula:(4)πcexpSc+y′c Zi∑s=1Sexp(Sc+y′c Zi) where Zi is the vector of variables describing respondent i, while y denotes the vector of related parameters to be estimated, and Sc denotes a class-specific constant. Only S−1 sets of coefficients may be recognized in the estimate for identification purposes. For the sake of identification, the vectors yc and Sc are both set to zero for one arbitrary class S. In addition, if the covariate results are significant, the information provided in the study findings may be utilized to explain segment membership.After the LCM results are obtained, it is essential to calculate the WTP in each market segmentation for each attribute used in the LCM. The way to analyze the WTP in the LCM is to divide each attribute’s coefficient by the coefficient of the price variable. The WTP calculation can be expressed as below:(5)WTP=−2 β attribute levelβ price3. Results3.1. Sample DistributionRespondents who are often in charge of grocery food purchasing for their families were targeted in this study. Particularly, respondents who had purchased fresh eggs in the past six months from supermarkets were focused on. According to the sample size calculation under the condition of the 95% confidence level [75,76], the minimum sample size is 385 observations in this study. However, although Green and Srinivasan (1978) [77] recommend that the minimum sample size is at least 100 observations to provide a credible estimate, this study, overall, collected 1115 valid respondents for this study. The sample summary of description and statistics is provided in Table 2.Results show in Table 2 that females constituted the largest group of respondents (about 75%). Under the set-up of screening questions, this implies that most respondents in charge of grocery food purchases are female, which is identified with previous literature findings [78]. The sample’s average age is approximately 40 years old. Further, the average education year is about 15 years, which means that most respondents roughly have a bachelor’s degree education level. About 40% of respondents stated that they have a child at home. Among those occupation categories, about 14% of respondents are housewives, about 10% work in the manufacturing industry, and approximately 19% work in the service sector, while almost 6% of respondents are retired. The purchasing egg background of respondents exhibits differently in each market channel. Results reveal that respondents often go to supermarkets to purchase fresh eggs (averagely about 2.3 times in a month), while traditional markets and hypermarkets show about 1.78 and 1.11 times in a month, respectively. Thus, supermarkets are the major market channel for consumers to buy the fresh eggs they need.3.2. The Determination of the Class Number in the LCMThe LCM was selected in this research to analyze the DCE choice data as well as to estimate and identify specific consumer segments for sundry attributes of eggs in supermarkets. When the LCM is utilized, the first step is to identify the number of classes. In order to decide the appropriate number of classes, the information criteria [79] are adopted in this study, such as the AIC (i.e., the Akaike Information Criteria; [80], the BIC (i.e., the Bayesian Information Criteria; [81], and the Log-Likelihood ratio index [82,83]. Swait (1994) [84] further suggested that the lowest value of AIC and BIC and the highest value of Log-likelihood would be able to identify the appropriate number of classes. Table 3 demonstrates the computational outcomes of AIC, BIC, and Log-Likelihood for both excluding and including unrealistic choice set models. The marginal changes in values of AIC and BIC from class 3 to 7 are very modest compared to the changes between class 2 to class 3. Further, the estimated values of the LCM for 4–7 classes are starting to deteriorate. This means that adding a class segment from 4 to 7 classes will have a potential effect, which indicates that more classes are not appropriate. In short, after trying to run LCM with several classes as well as based on the value of the Log-likelihood, AIC, and BIC, the appropriate number of classes to estimate these two LCM models is 3 classes.3.3. Consumer Preferences on Fresh Egg Attributes and the ClassificationIn order to examine how each attribute influences consumers’ preferences, the LCM is adopted and examined in Table 4. The including unrealistic choice set model is particularly tested, and the results of AIC, BIC, and Log-Likelihood [79] reveal that the excluding unrealistic choice set model has a better goodness-of-fit than the including unrealistic choice set model. Therefore, this study will adopt the results of the excluding unrealistic choice set model as final determination, while the results of the excluding unrealistic choice set model are treated as the comparison purpose.In this research, the LCM divides egg consumers into three categories and demonstrates that the majority of results are identical. Moreover, the researcher may examine the class share to compare classes in the LCM. As shown in Table 4, class 1 is a minority group with a market share of roughly 12.5%, class 2 is a middle group with a market share of approximately 28.3%, and class 3 is the majority group in this research with a market share of approximately 59.1%. The segmentation results indicate that there are three distinct consumer groups for fresh egg products in Taiwan supermarkets. The results in Table 4 show that the attribute of importance greatly differs among the 3 classes.The consumers in segment 1, which represents a minority group (12.5%), show a significant disdain for almost all egg characteristics examined in this research, including farm brand and traceability. This indicates that Taiwanese consumers in this category are unconcerned with fresh egg product characteristics such as farm branding and traceability. Additionally, consumers in this segment choose private brands (carrefour) over farm brands. In addition, the estimated parameters of the color attribute have a substantial negative significance, indicating that Taiwanese consumers in this class prefer white eggs to brown eggs. The reason Taiwanese consumers favor white eggs may be related to the pricing of white eggs, which are often less expensive than brown eggs in Taiwan [85,86,87].Respondents in class 2 (28.3% of respondents) express high preferences for attributes (brand, color, traceability, and animal welfare), but they consider the price as a monetary constraint. In this class, Taiwanese consumers prefer brown eggs as well as fresh egg products that have the farm’s brand and traceability code, as they think that the fresh egg products are safer and more reliable. Moreover, Taiwanese consumers in this category place a high premium on animal welfare labels, indicating that Taiwanese consumers prefer fresh egg products that have an animal welfare label. Thus, if we provide information about animal welfare, traceability, and farm brand on brown eggs, it may stimulate Taiwanese consumers to buy fresh egg products in Taiwanese supermarkets. Interestingly, the price variable in this class shows negative significance, indicating that although they prefer fresh egg products with animal welfare, brand, and traceability attributes, they also still prioritize price as a mandatory factor when purchasing eggs, and if the price offered is too high for them, they would not buy it as well. Therefore, it is essential to estimate how much consumers in class 2 are willing to pay for the attributes of fresh egg products.The largest class, which is class 3 with 59.1% of the respondents, seems to prefer white eggs, farm brand label, traceability, and animal welfare label. The respondents in this class are nearly identical to class 2, in that they are more likely to have farm brand labels, traceability, and animal welfare labels. This may imply that by implementing the strategy of adding farm brand label, traceability, and animal welfare attributes into fresh egg products, it is possible that at least 87.4% percent of Taiwanese consumers will be enticed to buy fresh egg products derived from a combination of class share in classes 2 and 3.Furthermore, the LCM membership effect is utilized to ascertain the characteristics of respondents inside each class. Table 4 indicates that the estimated parameters in the membership (socio-demographic) have an estimated value of 0, as class 3 membership is designated as the reference group in the LCM. In order to provide additional information about the effect of socio-demographic factors on the main attributes (price, brand, color, traceability, and animal welfare), this study also performs a post-estimation to predict and demonstrate the socio-demographic characteristics for each of the three classes in Table 5.Table 5 displays the probability of class membership in each class in the LCM. Since this research focuses on the outcomes of excluding unrealistic choice sets, thus, Table 5 focuses exclusively on the findings of class membership probability in the excluding unrealistic choice sets. In terms of covariance, the estimated parameters of the female variable in Table 4 have a 10% significance with a negative coefficient, indicating that the majority of class 1 respondents are male. This is also shown in Table 5 by the findings of the LCM prediction of sociodemographic probability. It indicates that the class 1 has a higher proportion of male respondents if compared to the class 3. On the other hand, it can also be said that female respondents are relatively higher in the class 3. Regarding the age factor, respondents in the class 1 are relatively higher than those in the class 3. The education factor reveals that respondents who are in the class 3 on average have higher education than those in the class 2. The hypermarket factor does show the significant difference between the class 1 and 3.3.4. The Estimation of WTP for Fresh Egg AttributesFollowing Equation (5), the consumer WTP for each fresh egg attribute in the CE can be calculated and presented in Table 6. The superiority of using the LCM to analyze WTP is not only able to estimate the points for each class as well as the parameters for the entire model, but the estimation results can also be corrected with the class probability value. As explained in Table 4, the estimation outcome is focused on the excluding unrealistic choice set model, so the results of the including unrealistic choice set model is compared in Table 6.Results of the including unrealistic choice set model in Table 6 reveal that many WTP calculations for each attribute display a significant level. However, some WTP calculations remain questions, especially for those who are in the largest consumer group. It shows that attributes of farm brand, color, and traceability have negative WTP for the largest consumer group. Since the WTP calculation is compared with the price attribute, it implies that the price attribute is the most important attribute in the largest consumer group. The middle group in Class 2 shows that consumers prefer animal welfare over price, so it shows 128.0 NT$ (≈US$ 4.57) more for animal welfare attribute. Further, other attributes of farm brand, brown-color eggs, and traceability also show that consumers in the middle group are willing to pay about 65.8 NT$ (≈US$ 2.35), 62.7 NT$ (≈US$ 2.24), and 51.7 NT$ (≈US$ 1.85) for attributes of traceability, brown-color eggs, and farm brands, respectively. However, consumers in the smallest consumer group of Class 1 would like to pay about 22.2 NT$ (≈US$ 0.79) for the traceability attribute. Although the results of the including unrealistic choice set model may represent any potential situation, the final WTP results would not adopt it from this model in order to not over-explain the outcomes in this study.Furthermore, the excluding unrealistic choice set model results indicate that class 2, which is the middle group (28.3% of respondents), is the only segment for which the estimated parameters are significant in terms of mean willingness-to-pay (MWTP). According to Table 6, Taiwanese consumers want to pay about 32.6 NTD (≈US$ 1.16) more for farm brand labels than private brand labels such as Carrefour. The result indicates that farm brand is becoming more well-received in Taiwan, as Taiwanese consumers are starting to pay more attention and support local farmers in Taiwan. Additionally, Taiwanese consumers are willing to pay an additional 32.5 NTD (≈US$ 1.16) for brown eggs to white eggs. Hereinafter, Taiwanese consumers are willing to pay an over 33.4 NTD (≈US$ 1.19) premium for fresh egg products with traceability labels. On average, the highest WTP of the egg attributes is constituted by animal welfare at 64.2 NTD (≈US$ 2.29).Although the majority group did not show a significant positive WTP in Table 6, they showed a positive response to the farm brand, traceability, and animal welfare variables in Table 4. The estimated parameters for the farm brand, traceability, and animal welfare attributes in class 3 did not show the positive correlation with MWTP, due to Taiwanese consumers’ lack of knowledge about those attributes. Hence, government assistance is needed to promote the attributes of farm brand, traceability, and animal welfare; thus, Taiwanese consumers will be more familiar with those attributes. Furthermore, as the results indicate that brown eggs with farm brand, traceability, and animal welfare attributes have a positive response from Taiwanese consumers, this can be further confirmed and provide a hint to the Taiwanese government and egg industry that by providing farm brand, traceability, and animal welfare attributes on fresh egg products, it may influence Taiwanese consumers to purchase fresh eggs.4. DiscussionThe present research used DCEs to ascertain which attributes may affect Taiwanese consumers’ decisions to buy fresh eggs in Taiwanese supermarkets, and five attributes of fresh eggs were considered: price, brand, color, traceability, and animal welfare. This researcher conducted an unrealistic choice experiment on CE to evaluate the impact of the unrealistic choice set on the estimated findings. Thus, this research will serve as a reference for including or excluding the unrealistic choice experiment for the CE design. Furthermore, this study also contributes significantly to the Taiwan government, egg business, and policymakers’ in terms of heterogeneous consumer preferences for egg attributes in Taiwanese supermarkets. Using the LCA method, this research found three unique groups of Taiwanese consumers, each with its own distinct set of preferences and WTP for fresh egg attributes, as well as distinct sociodemographic and behavioral traits.As shown by the AIC, BIC, and Log-likelihood values in Table 3, the data set without unrealistic options is more suitable than the data set with unrealistic choices as it has a greater Log-likelihood and lower AIC and BIC values [79]. Moreover, Table 3 also shows that 3 classes in the LCM gave the best fit since the marginal changes in AIC and BIC values between classes 3–7 are extremely small in comparison to the changes between classes 2 and 3. Additionally, the model’s estimated value for classes 4–7 have begun to degrade, resulting in an unstable AIC and BIC value. To summarize, the optimal number of classes for estimating LCM models is three.Table 4 presented and contrasted the findings of LCM analysis on two distinct data types in order to determine the most appropriate data set for assisting the researchers in accurately analyzing and interpreting this study’s results. Further, this study will determine whether or not an unrealistic choice set has an effect on the LCM estimate outcome. The goodness of fit value generated by the LCM analysis may be evaluated [79] to determine which data sets are more fitted and perform better with the model in this research. The AIC, BIC, and Log-likelihood values in Table 4 suggest that excluding the unrealistic choice set model is more appropriate than excluding the unrealistic choice set model.The results of attributes in Table 4 show that most outcomes are similar to each other. The class shares of the excluding unrealistic choice set model reveal that the class-3 is the biggest consumer group with about 59% market share, while the middle and small groups of class share exhibit about 28% and 12%, respectively. It means that there roughly can present three types of consumer groups for fresh egg products in supermarkets. The attribute preferences of the largest consumer group show that they prefer farm brand, white-color eggs, traceability, and animal welfare. However, the middle-group consumers prefer lower prices, farm brand, brown-color eggs, traceability, and animal welfare; and the small-group consumers are particularly like the private brand of supermarkets, white-color eggs, and no traceability label. This implies that there still is a small consumer group that does not really prefer any attribute.In addition, the results of the including unrealistic choice set model show that the largest consumer group with about 45% market share, while the middle and small groups of class share reveal about 40% and 15%, respectively. The composition of market shares in the including unrealistic choice set model is similar to the excluding unrealistic choice set model. If the traceability is compared in the small consumer group, then it received a different result. It implies that the smallest consumer group reveals a positive preference for traceability. Further, the results of attributes in the largest consumer group show that consumers prefer higher prices, farm brand, white-color eggs, and traceability. This is a bit of a contradictory outcome in price, since it implies that consumers will tend to buy more eggs if the price is higher. This outcome may reflect the cause of unrealistic choice sets. Moreover, if the unrealistic choice sets exist, it may lead to an expected outcome like this. Therefore, it is an important finding in this study that provides a good example if future studies adopt the CE method. However, the results of including unrealistic choice sets are not adopted as the final outcome in this study.Following Equation (4), Table 4 also shows how each social demographic variable and shopping background contribute to each class share. The largest consumer group of the class-3 is compared. Results of the excluding unrealistic choice set model show that higher age male consumers who usually purchase fresh eggs from non-hypermarkets are more likely to be in the smallest consumer group (i.e., the class-1). It also means that these consumers prefer attributes of private band, white-color eggs, and no-traceability when comparing to those who are in the largest consumer group. Although previous studies [85,86,87] mentioned that the preference for white eggs is because it is cheaper than brown eggs, consumer preferences of white eggs in the class-1 do not link to the price attribute. It can be confirmed that some consumers may still prefer white eggs. The reason Taiwanese consumers prefer white eggs may correlate with the eggs price as usually, the price of white eggs in Taiwan tends to be cheaper than brown eggs. Consumers with lower education would prefer attributes of a lower price, farm brand, brown-color eggs, traceability, and animal welfare if compared to a reference group. Thus, it can be identified that younger female consumers with higher education who usually shopped at hypermarkets tend to be in the largest consumer group. This also corresponds to previous findings [14,31], that there is a potential market trend on-farm brand, traceability, and animal welfare for the majority of consumers.The final WTP results will not contain an unrealistic option set model to avoid over-explaining the WTP findings of this study, even though the results may reflect every possible scenario. Thus, in order to account for the WTP findings in this study, the researcher concentrates only on excluding the unrealistic choice set model.The WTP results of the excluding unrealistic choice set model reveal that only the middle consumer group of class-2 shows a significant level with a positive sign. This implies that consumers in the middle group (28.3% of market share) are more likely to pay more for these attributes than the price attribute; in other words, consumers are willing to pay more for the attributes of farm brand, brown-color eggs, traceability, and animal welfare. Particularly, consumers in the middle group would like to pay about 32.6 NT$ (≈US$ 1.16) more for farm brand labels if compared to the private brand label (i.e., Carrefour). This indicates that fresh eggs with farm brands are receiving more attention than the private brand in supermarkets. Further, regarding the egg-color attribute, consumers are willing to pay more about 32.5 NT$ (≈US$ 1.16) for brown-color eggs if compared to white eggs. This result also corresponds to previous studies [15,21] that consumers prefer to purchase brown eggs over white eggs due to the impression of health concerns and quality issues. However, this result represents the middle consumer group, so it is still not contradictory to the consumers who may prefer the white-color eggs in the small group of class-1.In addition to the traceability attribute, consumers are willing to pay about 33.4 NT$ (≈US$ 1.19) for fresh egg products with traceability labels in supermarkets. Since consumers in Taiwan are more concerned about food-product originality and safety [88], this study also corresponds to the argument of whether traceability is important in food product labeling. However, the animal welfare attribute presents a higher WTP than any other attribute. This implies that the animal welfare attribute is the most important among these attributes. Consumers in the middle group are willing to pay about 64.2 NT$ (≈US$ 2.29) for fresh eggs with animal welfare attributes in supermarkets. This finding also corresponds to a previous study [13] that consumers do care for animal welfare. Therefore, this study confirms that there are potential markets for animal welfare, traceability, farm brand, and brown-color eggs in supermarkets in Taiwan.Since the price attribute of the largest consumer group in Table 4 did not show a significant level in the excluding unrealistic choice set model, it may lead to the WTP calculation of the largest consumer group in Table 6 having not shown a significant level in the excluding unrealistic choice set model. This implies that these attributes in the largest consumer group are not identified in the WTP calculations. In other words, consumers in the largest group may focus on other attributes as their preferences. However, regardless of the WTP calculations, consumers in the largest consumer group still care about farm brand, white-color eggs, traceability, and animal welfare attributes.5. ConclusionsFresh egg products in supermarkets are getting more diversified by promoting different qualities in eggs, such as animal welfare, traceability, farm brand, color of eggs, etc. Since eggs are one of the fundamental elements for daily diet and consumers do care about what they eat, this study attempts to ascertain consumer preferences via the WTP and market segmentations. The CE method was utilized to estimate the market segmentations as well as the most important attributes that may influence consumer preferences. Further, this study also compares the results of including and excluding the unrealistic choice sets in the estimation of the LCM model. According to the values of the goodness of fit, the excluding unrealistic choice set model reveals better goodness of fit than the including unrealistic choice set model. Indeed, the overall outcomes of these two models present differently. Moreover, it is strongly recommended to other studies if the CE method with the choice set situation is adopted; the existing unrealistic choice sets should pay extra attention. With the indication of AIC, BIC, and Log Likelihood, this study only adopts the outcomes of the excluding unrealistic choice set model.The findings in this research contribute to a better understanding of what motivates Taiwanese consumers to purchase fresh eggs in supermarkets. Among the major attributes, i.e., animal welfare, traceability, farm brand, and brown-color eggs, each attribute reveals a positive preference for certain segments of Taiwanese consumers. Especially, consumers who prefer animal welfare eggs are willing to pay up to about 64.2 NT$ (≈US$ 2.29), which is about twice above from the WTPs of other attributes, i.e., traceability, farm brand, and brown-color eggs. This presents that the animal welfare factor is a hot topic in supermarkets now. The overall results of this study convey strong signals to the egg industry, stakeholders, government, and policymakers about the market potential for brown-color egg, animal welfare, traceability, and farm brand attributes in supermarkets.Several limitations in this study should be addressed. First, only five attributes are considered in this study, while other potential factors (i.e., nutritional facts, Halal, freshness, etc.) are omitted. Second, this study only focuses on supermarkets in Taiwan, while other major markets, such as traditional markets, are not considered in this study. In order to provide more comprehensive information, a further examination is needed, so government and policymakers would be able to make the relevant policies that encompass the entire consumer demand.

animals : an open access journal from mdpi

10.3390/ani11113264

PMC8614513

It is of great interest to quantify adaptive evolution in human lineage by studying genes under positive selection, since these genes could reveal insights into our own adaptive evolutionary history compared to our closely related species and often these genes are functionally important. We used the great apes as the subjects to detect gene-level adaptive evolution signals in all the great ape lineages and investigated the evolutionary patterns and functional relevance of these adaptive evolution signals. Even the differences in population size among these closely related great apes have resulted in differences in their ability to remove deleterious alleles and to adapt to changing environments, we found that they experienced comparable numbers of positive selection. Notably, we identified several genes that offer insights into great ape and human evolution. For example, SOD1, a gene associated with aging in humans, experienced positive selection in the common ancestor of the great ape and this positive selection may contribute to the aging evolution in great apes. Overall, an updated list of positively selected genes reported by this study not only informs us of adaptive evolution during great ape evolution, but is also helpful to the further study of non-human primate models for disease and other fields.

Alleles that cause advantageous phenotypes with positive selection contribute to adaptive evolution. Investigations of positive selection in protein-coding genes rely on the accuracy of orthology, models, the quality of assemblies, and alignment. Here, based on the latest genome assemblies and gene annotations, we present a comparative analysis on positive selection in four great ape species and identify 211 high-confidence positively selected genes (PSGs). Even the differences in population size among these closely related great apes have resulted in differences in their ability to remove deleterious alleles and to adapt to changing environments, we found that they experienced comparable numbers of positive selection. We also uncovered that more than half of multigene families exhibited signals of positive selection, suggesting that imbalanced positive selection resulted in the functional divergence of duplicates. Moreover, at the expression level, although positive selection led to a more non-uniform pattern across tissues, the correlation between positive selection and expression patterns is diverse. Overall, this updated list of PSGs is of great significance for the further study of the phenotypic evolution in great apes.

1. IntroductionAdaptive evolution has been reported to be associated with many phenotypic changes in humans [1,2,3]. The Hominid, known as the great ape, has experienced various adaptive evolutionary innovations, such as significant sexual dimorphism [4], increased body mass [5], and increased brain volume correlated to high-order cognitive ability [6]. Identifying genes targeted by adaptive evolution (also known as Positively Selected Genes, PSGs) will advance our understanding of the underlying genetic basis of evolution. Although many studies have investigated PSGs in humans or primates, the detection of positive selection relies on the accuracy of orthology, the quality of assemblies, models, and alignments [7,8,9,10]. The improved branch-site model built in PAML is a commonly used model to detect positive selection [11]. In the branch-site model, branches in the tree are classified into foreground branches, in which some sites have been targeted by positive selection (the dN/dS of sites > 1), and background branches, in which no positive selection occurs [12]. Then an LRT is performed to compare an alternative model in which some sites undergo positive selection on the foreground branches with a null model that does not. Bakewell and colleagues [13] used this improved branch-site model to identify PSGs in the human and the chimpanzee, with rhesus as the outgroup. They performed analysis based on old, less-complete genome assemblies without filtering for the low confidence PSG candidates, which will cause potential false negatives. Lee et al. [14] performed the latest genome-scale positive selection test in primates to identify high-confidence PSGs with a stringent threshold. However, their analysis was based on the site model, which cannot detect the branch-level positive selection signals [15], and they did not use the latest upgraded genome assemblies [16].In this study, utilizing the orthologs and gene families produced by the latest genome assemblies and gene annotations, we searched for high-confidence positive selection across all sites of all orthologous genes in all great ape lineages, and showed that the great ape species experienced comparable numbers of positive selection; even the differences in population size among these closely related great apes have resulted in differences in their ability to erase deleterious alleles and to adapt to changing environments. We also observed that the majority (63%) of the PSGs are from multiple-gene families, suggesting various positive selection resulted in the functional divergence of duplicates. Based on these PSGs, we investigated their functional contribution to great ape evolution and their expression pattern. We also tested human disease adaptations among our PSGs. Overall, this study provides a timely update of PSG candidates that might have contributed to the adaptive evolution of great apes.2. Materials and Methods2.1. Identification of Orthologs in Six PrimatesWe applied Orthofinder [17] among six primates and identified 14,758 one-to-one orthologs (Table 1). After excluding 361 ortholog groups containing gene models with in-frame stop codon or whose length of CDS was not a multiple of three, we retained 14,397 one-to-one orthologs and used them in downstream analysis. The six primate genomes included were human (H. sapiens, GRCh38.p13), chimpanzee (P. troglodytes, Clint_PTRv2), gorilla (G. gorilla, Kamilah_GGO_v0), orangutan (P. abelii, Susie_PABv2), gibbon (N. leucogenys, Asia_NLE_v1), and rhesus (M. mulatta, Mmul_10).To ensure the coherence of the annotation resource, we collected all the corresponding annotations from the NCBI database based on the latest assembly version (Table 1). We downloaded respective NCBI gene sets and kept the longest transcript for each protein-coding gene. For humans, we only kept the protein-coding genes from primary assembly to avoid redundant gene models.2.2. Building the Gene FamilyTo determine the gene families, we extracted coding sequences (CDSs) according to the gene annotations and genomes of corresponding species and then translated these coding sequences into protein sequences. Next, we conducted an all-against-all blastp search (blast-2.2.26; E-value < 10−7) [18] of these protein sequences in the six primate species [19]. Finally, protein sequences were clustered into gene families based on identity using hcluster_sg (hcluster_sg -m 750 -w 0 -s 0.34 -O) in Treefam. This gene family information was used to classify orthologous groups into different gene families. We classified gene families with multiple members from one or more species as multigene families, and gene families with only single-copy genes as single-gene families.2.3. Positive Selection of GenesWe screened orthologous genes to explore the patterns of positive selection in great apes: H. sapiens, P. troglodytes, Hominini, G. gorilla, Homininae, P. abelii, and Hominidae. We first extracted the CDS from genomes based on protein-coding gene annotations and translated CDS into protein sequences. We then aligned these protein sequences using GUIDANCE (guidance.pl --program GUIDANCE --seqType aa --msaProgram PRANK --MSA_Param “\+F “; v2.02) [20], back-translated protein alignments into CDS alignments based on the original CDS, and fed these as the input to the improved branch-site model in PAML software package v4.9j [11,12]. We labeled all terminal branches and inner branches of great apes as foreground branches separately to perform multiple branch-site model tests for each ortholog group. To exclude potential artifacts, the final PSGs were determined by the following steps: (1) p-values were computed using the LRT test based on the output from PAML software and only genes with p-value < 0.05 were regarded as PSG candidates and used in the following analysis; (2) the PSG candidates with potential sites for selection (Bayes Empirical Bayes (BEB) posterior probability > 0.95) with gaps in 5 upstream or downstream amino acids were filtered out to exclude the false positive PSGs caused by alignment gaps (see an example in Figure S1); (3) filtered out PSG candidates with positively selected sites whose GUIDANCE alignment column score (range from 0 to 1) was less than 1.0 (the highest column score that ensures high-quality column alignment), (4) and excluded the false positive PSGs potentially caused by genetic drift (K-value < 1 and p-value < 0.05) by running RELAX from the Hyphy package in PSG candidates [21].2.4. Population Analysis of PSGsWe used population SNPs datasets (in VCF format) of great ape species to investigate whether the positively selected sites of PSGs were fixed in these lineages. For humans, chimpanzees, and orangutans, we downloaded their corresponding SNPs dataset from the Ensembl database (Ensembl release 104). For gorillas, whose SNPs dataset was not available in the Ensembl database, we downloaded the published data by Prado-Martinez et al. [22]. Since the reference genome assemblies of chimpanzee, gorilla, and orangutan SNPs datasets were different from the assemblies used in this study, we converted the coordinates of these VCF files by LiftoverVcf in GATK (v4.1.4.1) based on the chain files between different assemblies produced by LASTZ [23] followed by chaining and netting [24] using scripts from the UCSC genome browser source code.We defined a positively selected amino acid site as fixed in a population if all the allele frequency of these alleles located in this codon > 0.95 (in humans whose allele frequency information was available) or it has no alternative allele in this codon (in chimpanzees, gorillas, and orangutans where the allele frequency information was not available). A PSG was regarded as fixed if it contained at least one fixed positively selected amino acid site.2.5. Expression Analysis of PSGsTo investigate the expression pattern of PSGs, we used gene expression profiles (Strand-specific RNA-seq of 13 human tissues from Michael Snyder’s lab for the ENCODE project) in the Expression Atlas database (https://www.ebi.ac.uk/gxa/home, last accessed 18 April 2021). We used log2(TPM + 1) as the proxy to measure the differences in the expression patterns (tissue specificity or expression level) of PSGs and their most similar paralogous non-PSGs, which were paralogs with the highest similarity to PSGs. We compared tissue specificity and expression level of the human gene expression pattern between PSGs and non-PSGs. We used the putative τ [25] as a proxy of the tissue specificity of a gene and used mean expression level across all tissues as a proxy of expression level.2.6. Association Analysis of PSGs in Great ApesTo test whether PSGs tend to be associated with disease, we downloaded disease-associated genes from OMIM [26] and performed all the statistical analysis with R.3. Results3.1. Comparable Number of Genes Had Experienced Positive Selection among Great Ape SpeciesThe orthologous relationships of all gene families allowed us to investigate the forces of positive natural selection on genes derived from the common ancestors in different great ape lineages. Using an improved branch-site likelihood method [11], we searched for positively selected genes (PSGs) at all evolutionary branches in great apes. We first identified 14,397 one-to-one orthologous genes that evolved from the common primate ancestor using the gibbon and the rhesus as the outgroup. After filtering out false positives potentially caused by alignment error and genetic drift (Materials and Methods), our analyses revealed that 211 orthologs had experienced positive selection in at least one evolutionary branch (Table 2 and Table S1). Of these genes, 208 were positively selected at one evolutionary branch and 3 were positively selected in two evolutionary branches. Overall, the great ape species harbor comparable PSGs (31~40 PSGs): from 31 PSGs in gorillas (the fewest among great ape lineages), 39 PSGs in humans, to 40 PSGs in chimpanzees (the largest among the four tested extant great ape species) (Table 2; Figure 1). Compared to terminal branches, the inner branches have fewer PSGs. Based on the published population SNPs datasets (Materials and Methods) [22,27], we found most of the PSGs (204 out of 211 PSGs) contain at least one amino acid changes that have been fixed in the population (454 out of 484 positively selected sites, Table 3).Bakewell et al. 2007 used the branch-site likelihood model but old annotation and orthologous data, and identified 154 and 233 PSGs in humans and chimpanzees, respectively [13]. These numbers decreased to 39 in humans and 40 in chimpanzees in our analyses. We found that 11 human PSGs and 9 chimpanzee PSGs have been reported by previous study. Many of the PSGs (77/154 in the human and 97/233 in the chimpanzee) could be attributed to the different input caused by different annotation versions and ortholog assignment and problematic gene models (Table 4). Among these, 58 PSGs in humans and 75 PSGs in chimpanzees in the previous study were not listed in the current annotations. Additionally, several previously reported PSGs (19 in humans and 25 in chimpanzees) were not included in our orthologs. This was because of different species sampling (16 in humans and 22 in chimpanzees) and low confidence gene models with in-frame stop codon or the length of CDS was not a multiple of three in these ortholog groups (three in humans and three in chimpanzees) and thus these orthologs were not included in this analysis.Beyond different input concerns, 66 human PSGs and 124 chimpanzees PSGs in Bakewell’s list were not identified as no PSGs in this study because of different data processing method. Specifically, a total of 25 human PSGs and 96 chimpanzee PSGs did not pass our LRT test. Moreover, 38 human PSGs and 27 chimpanzee PSGs on Bakewell’s list did not contain any positively selected sites with BEB posterior probability greater than 0.95. Next, 3 PSGs in humans and 27 PSGs in chimpanzees in Bakewell’s list were filtered out because their positively selected sites were biased. Apart from these, we detected 25 human PSGs and 30 chimpanzee PSGs which were not identified by Bakewell et al. This may be caused by the more complete assembly or our denser species sampling and thus increases the power of detection of PSGs.We observed that 73% (154/211) of PSGs were from multigene families (Figure 1). In principle, once a new gene is duplicated from its ancestral copy, especially by DNA level duplication in which often provides a functional promoter to express the duplicated gene [28], it introduces a redundant function to the parental copy, which would result in relaxed selective constraint in one copy and be ultimately lost through pseudogenization in most cases [29]. Thus, we investigated whether genes with duplicated copies are less likely to be positively selected by testing whether PSGs are less enriched in multiple-gene families. We found the number of PSGs in multiple families was not significantly less than PSGs in single-gene families (p-value range from 0.186 to 0.992) (Table 5), suggesting imbalanced positive selection resulted in the functional divergence of duplicates. These results are consistent with the hypothesis that gene duplication events can also offer new genetic materials for selection [30].3.2. The PSGs Contributed to Functional Evolution of Great ApeTo explore how PSGs contribute to functional evolution in great apes, we first performed gene enrichment analyses with Metascape [31], and found that PSGs in humans and ancestral great ape lineage are enriched in essential biological functions such as positive regulation of cell junction assembly and the superoxide metabolic process (Figure S2). We then investigated the functional contributions of PSGs by focusing on specific genes in great apes. We found that 12 SLC genes were positively selected in great apes. For example, SLC39A6, a gene of the SLC39 gene family, was positively selected in humans. SLC39 transporters primarily serve to pass zinc into the cytoplasm and play critical roles in maintaining cellular zinc homeostasis [32]. Homozygous knockout of SLC39 family genes cause neurodegeneration growth retardation, morphological defects, and abnormal neurogenesis in mice [33,34]. The positive selection of SLC39A6 may be associated with the distinct neurogenesis in humans and correlated to high-rank cognition ability [35,36]. We also found two leukocyte antigens (CD36 and CD3E) under positive selection in the common ancestor of the great ape. Interestingly, CD36 has been reported related to malarial resistance in humans [37] and mutations in CD36 are associated with malaria susceptibility [38] and protection against malaria [39], indicating that the positive selection of CD36 may be associated with malarial resistance evolution in great apes, although the difference in malarial resistance between great ape and other primates has yet to be investigated.We found four PSGs that may be associated with aging in great apes. For example, SOD1, which encodes superoxide dismutase 1, responsible for destroying free superoxide radicals in the body [40], was detected to have been targeted by positive selection in the common ancestor of the great apes at multiple sites (Met 3, Gln 51, Ser 113; Figure 2). SOD1 contributes to the senescence as an important player in cellular senescence by catalyzing superoxide radicals (O2._) to H2O2 and O2 [41] and mediating the p53 pathway [42], which are both involved in the cellular senescence process [43], and the overexpression of this gene in fruit flies extended their lifespan [44]. This indicates the positive selection of this gene may be associated with the longer lifespan of great ape species compared with other primates [45] (Table S2).We also found several PSGs that may contribute to great ape functional evolution. In humans, positive selection was detected in CA14 at Lys 204 (Figure 3). Carbonic anhydrase (CA) is a large multigene family that contains 15 paralogs and is associated with reversible hydration of carbon dioxide in the primate. In this gene family, positive selection was only detected in one copy, CA14. CA14 catalyzes conversion between carbon dioxide and carbonic acid and bicarbonate ions in humans [46]. The CA14 maintains a high expression level in the central nervous system in normal human adults [47] and CA14 may also play an important role in modulating excitatory synaptic transmission in the brain [48], indicating that positive selection in CA14 may contribute to the nervous system evolution in humans.3.3. The Expression Pattern of Positively Selected GenesWe further investigated whether positive selection in humans affects the gene expression pattern in tissues by comparing PSGs in humans with their most similar paralogous copies in multigene families (Materials and Methods). Although the expression patterns between PSGs and their closest paralogous copies were different, we found that there are no uniform expression pattern shifts between these two copies, i.e., there is no clear correlation between the positive selection and tissue specificity or expression level (Figure 4 and Figure S3). For tissue specificity among the 23 gene pairs (PSG and its closest paralogous non-PSG), PSGs had higher tissue specificity than non-PSGs in 10 gene pairs (Figure S3a). For example, ZFAND4 encodes zinc finger AN1-type-containing 4 and serves as a marker to predict metastasis and prognosis in oral squamous cell carcinoma [49], while ZFAND4 was strictly expressed in the testis; its paralogous copy ZFAND5 is a ubiquitously expressed gene and was highly expressed in other tissues such as the brain, lung, and testis (Figure 4b), indicating the genes not under positive selection harbor wide tissue and positive selection may eliminate the tissues ubiquity of ZFAND4. In the remaining 13 gene pairs, non-PSGs had higher tissue specificity than PSGs.Among the expression levels of 23 gene pairs, the PSGs had higher mean expression levels across multiple tissues than non-PSGs in 11 gene pairs (Figure S3b). Taking the aforementioned SLC39A6 as an example, SLC39A6 (PSG) was expressed in all examined tissues (highest in brain), while its paralogous copy SLC39A10 had a lower expression level across multiple tissues except in the spleen (Figure 4c). These results indicate that the association between selection and expression is divergent among different PSGs. Among the remaining 12 gene pairs, the non-PSGs had higher mean expression levels across multiple tissues than PSGs.3.4. PSG and Disease EvolutionThere was a hypothesis suggesting that PSGs tend to be associated with disease because the current environment of humans is substantially different from that of earlier hominins and thus previous adaptive mutations may become deleterious nowadays [50,51]. We tested this hypothesis using our newest PSGs and latest OMIM data [26]. In contrast to the results of the previous study [13] which found some support for the hypothesis, we found that there is no significant association between the human PSGs and human disease-associated genes (Table 6) and thus found no evidence to support that hypothesis.4. DiscussionWe detected 211 PSGs at all evolutionary branches in great apes. Compared with the previous study [13] using the same branch-site model [11], we identified 39 and 40 PSGs in humans and chimpanzees after filtering low-quality aligned sites and false positives caused by genetic drift, which ensure high confidence in PSGs. It should be noted that we could not identify all PSGs in all great ape evolution, especially those selected alleles that have not been fixed. Previous reports showed that the number of PSGs of chimpanzees is much greater than that of humans [13]. By contrast, we found that the great ape species experienced comparable numbers of positive selection. This could be caused by the update of annotation and assembly and our stricter filters for PSGs limiting the false positive PSGs and thus making the number of PSGs less than before, especially in chimpanzees whose genome then was poorly assembled.Previous study showed that gene duplication results in relaxed selective constraint in one copy and is ultimately lost through pseudogenization in most cases [29]. However, we found that genes in multigene families, which experienced multiple rounds of gene duplication, can also be subject to positive selection, suggesting imbalanced positive selection resulted in the functional divergence of duplicates. We did not find a consistent expression pattern shift in PSGs and its closest paralogous copy (Figure 4 and Figure S3). We found positive selection led to a more non-uniform pattern across tissues. SLC39A6 represents an interesting case given its high expression level in the human brain and its expression level is higher than its closest paralogous copy in multiple tissues. Previous studies showed that paralogs always show different expression profiles and are more tissue specific [52,53] and the evolutionary rate of a gene is negatively correlated with the expressional level [54,55]. This could be attributed to the fact that we only compared the expression level between PSG and non-PSG gene pairs instead of investigating the association between gene expression and evolutionary rate in all the genes [56], and thus this sampling led to different results. Interestingly, given that many PSGs are associated with functional and phenotypic changes in humans (e.g., the PSG CA14 may be associated with metabolic evolution in humans), it would be fascinating to validate how these PSGs contribute to the phenotypic evolution in future studies.We also found several genes are associated with aging. For example, SOD1, which plays important roles in senescence [41], experienced positive selection in the common ancestor of the great ape. Positive selection of this gene may be associated with the extended lifespan in great ape species compared with other primates (Table S2). Antagonistic pleiotropy (AP) hypothesis argues that the genes benefitting early life withstand more active positive selection [57], while these genes may impair late life and cause senescence. The pleiotropy role of this gene may be a case of AP hypothesis. Previous studies found that the AP hypothesis was supported by recently evolved enhancers in humans [58]. It will be interesting to test the AP hypothesis in coding regions based on population data in further studies. Unlike the previous study that found that PSGs in humans are enriched with human disease-associated genes [13], we found no evidence to support that conclusion in our analysis. This could be partly attributed to the update of genes in OMIM, which has grown from 847 in 2007 to 2301 nowadays, and the more stringent filters for PSGs we applied.5. ConclusionsThis study performed comparative evolutionary analysis of the latest assembled primate species to identify PSGs during great ape evolution. Among great ape species, the numbers of positively selected genes are comparable, even the differences in population size among these great apes should have resulted in differences in their ability to remove deleterious alleles and to adapt to changing environments. We also uncovered that more than half of multigene families exhibited signals of positive selection, suggesting that imbalanced positive selection resulted in the functional divergence of duplicates. We did not find a consistent shift in PSGs compared to their closest copies, implying that positive selection led to a more non-uniform pattern across tissues. PSGs in the human and Hominidae contribute to the neural transporter, antigens, and aging evolution. In addition, our results did not support the hypothesis that PSGs tend to be previously adaptive but deleterious nowadays.

animals : an open access journal from mdpi

10.3390/ani12060788

PMC8944783

Femoral head necrosis (FHN) and other locomotor problems cause severe impacts on the poultry industry due to huge economic losses and reduced animal welfare. Femoral head separation (FHS), the initial phase of FHN, is usually a subclinical condition characterized by the detachment of articular cartilage from the bone. In this study, we aimed to identify genes and biological processes involved with FHS in broilers. A better understanding of the FHS molecular mechanisms can help to develop strategies to reduce this condition in chickens. Here, we described several genes that have their expression altered in the articular cartilage and femur when normal and FHS-affected animals were compared. Furthermore, genetic variants were found differing between the studied groups. Therefore, performing an integrated analysis of these datasets, we were able to detect genes and variants related to FHS in chickens. Some of them, such as SLC4A1, RHAG, ANK1, MKNK2, SPTB, ADA, C7 and EPB420 genes were highlighted and should be further explored to validate them as candidates to FHS and FHN in chickens and possibly in humans.

Femoral head separation (FHS) is usually a subclinical condition characterized by the detachment of articular cartilage from the bone. In this study, a comprehensive analysis identifying shared and exclusive expression profiles, biological processes (BP) and variants related to FHS in the femoral articular cartilage and growth plate in chickens was performed through RNA sequencing analysis. Thirty-six differentially expressed (DE) genes were shared between femoral articular cartilage (AC) and growth plate (GP) tissues. Out of those, 23 genes were enriched in BP related to ion transport, translation factors and immune response. Seventy genes were DE exclusively in the AC and 288 in the GP. Among the BP of AC, the response against bacteria can be highlighted, and for the GP tissue, the processes related to chondrocyte differentiation and cartilage development stand out. When the chicken DE genes were compared to other datasets, eight genes (SLC4A1, RHAG, ANK1, MKNK2, SPTB, ADA, C7 and EPB420) were shared between chickens and humans. Furthermore, 89 variants, including missense in the SPATS2L, PRKAB1 and TRIM25 genes, were identified between groups. Therefore, those genes should be more explored to validate them as candidates to FHS/FHN in chickens and humans.

1. IntroductionThe poultry industry has experienced huge growth in recent decades, due to the advances in new research and technologies, which has brought benefits to the industry and producers. Chicken meat is one of the most consumed proteins in the world, due to the low cost, for being considered cost-effective and healthy, and because it is exempted from cultural impediments [1]. The constant investment in genetic improvement, nutrition and management has generated a better performance of the broilers, with slaughter age reduction, better feed conversion and less water consumption, always ensuring the animal welfare [2].The genetic breeding programs have brought numerous benefits for poultry production, such as the increased body weight, greater thoracic musculature, and reduction of abdominal fat. In the last 50 years, chicken growth rates have risen more than 300%, from 25 g per day to approximately 100 g per day [3,4]. However, some negative consequences, such as locomotor problems and immunological and metabolic disorders have appeared due to the selection for fast-growing [3,5].Locomotor problems cause severe impacts on the poultry industry, affecting about 6% of the animals in commercial flocks, which results in huge economic losses [6]. Furthermore, animals with bone disorders have their welfare affected because they cannot eat and drink water properly, being susceptible to opportunistic agents [7]. The bacterial chondronecrosis with osteomyelitis (BCO), also known as femoral head necrosis (FHN), is the most prevalent leg disorder in broilers [8]. The real incidence of FHN in broilers is difficult to estimate since most of the lesions often remain subclinical, and its etiological basis is not fully understood [9]. However, some authors have shown 20% of FHN incidence in chickens [10]. FHN occurs in the proximal part of the chickens’ femur head, starting with the femoral head separation (FHS), which consists in the separation of the growth plate (GP) from the articular cartilage (AC) [9,11]. This occurs due to the high growth rates and the mineralization deficiency of the chondrocytes that cause damage to the bone, which could allow for the colonization by opportunistic bacteria featuring the BCO [12].Since the etiology of FHN remains unclear, studies are being conducted to clarify its pathogenesis and the molecular mechanisms involved, and to discover genes with a potential role in triggering FHS and subsequent BCO in the head of the femur and tibia [10,13,14,15,16,17]. Although most of the previous studies describe biological processes (BP) involved with this condition in bone tissue, there are no studies evaluating gene expression profiles between bone and articular cartilage of normal and FHS-affected chickens. The femur GP or physis is a point of huge stress during the broilers growing phase. The GP tissue presents a cartilaginous matrix and sequential layers of long columns of chondrocytes in different maturation stages [12]. Above the GP is the AC that consists of chondrocytes and extracellular matrix (ECM), which is mainly composed by proteoglycan, glycosaminoglycan, and collagen fibers [18]. The understanding on how these tissues are interacting might be one way to improve the knowledge of molecular mechanisms involved with FHS. Therefore, in this study, the femoral growth plate and articular cartilage transcriptomes from normal and FHS-affected broilers were compared to identify common and exclusive molecular mechanisms and differentially expressed (DE) genes between groups, and polymorphisms highlighting possible new candidate genes involved with FHS in chickens.2. Material and MethodsTranscriptomes used in the present study were obtained from a project that evaluated femoral head separation (FHS) in the growth plate and articular cartilage, published respectively by Peixoto et al. [15] and Hul et al. [17], which are available in the SRA database BioProject # PRJNA350521 for articular cartilage (AC) and PRJNA352962 for growth plate (GP). Briefly, 16 AC and GP samples from 35-day-old male broilers (Cobb500), previously characterized as normal (4 AC and 4 GP) or FHS-affected (4 AC and 4 GP), were used in this study. At the time of collection, the normal samples had a good adhesion between the AC and GP, and the FHS-affected samples had the AC easily detached from the GP. For the normal group, the AC was separated from the GP using a scalpel. These tissues were submitted to RNA extraction using Trizol and a cleanup using RNeasy mini kit (Qiagen, Hilden, Germany). RNA quality was assessed in 1% agarose gel and in Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA). Samples with an RNA integrity number (RIN) higher than 8 were used to construct the RNA-Seq libraries. Eight samples from each tissue (4 normal and 4 FHS-affected), were prepared for RNA sequencing. For GP, the libraries were prepared from 2 µg of RNA using the “TruSeq RNA Sample Prep Kit v2” (Illumina, San Diego, CA, USA), while for AC the “TruSeq Total Stranded Sample Preparation” (Illumina, San Diego, CA, USA) kit was used. Libraries were sent to the Functional Genomics Center, ESALQ, University of São Paulo, Piracicaba, São Paulo State, Brazil, for sequencing in Illumina HiSeq2500 equipment (2 × 100 bp, Illumina, San Diego, CA, USA), with samples of each tissue in a different lane.2.1. Bioinformatics, Differential Expression Analyses and Functional AnnotationThe baqcom pipeline (https://github.com/hanielcedraz/BAQCOM, accessed on 31 December 2021) was used for quality control (QC) and mapping. QC for each sequencing file was performed with the Trimmomatic v. 0.38 [19] to remove short reads (<70 bp), low-quality reads (QPhred < 20) and adapter sequences. AC and GP sequences were mapped against the chicken reference genome (GRCg6a, Ensembl 101) using the STAR software v 2.7 [20], and reads were counted with HTSeq-counts [21]. The EdgeR package [22] was used to identify the DE genes between normal and affected groups in each tissue separately, considering differentially expressed those with false discovery rate (FDR) ≤ 0.05 after correcting for the Benjamini–Hochberg (BH) multiple tests [23]. Positive and negative log2 fold-change corresponded to DE genes upregulated or downregulated, respectively, in FHS-affected compared to the normal group. A heatmap was generated to visualize patterns of expression across samples from both tissues using R. The expressed genes were annotated using the Biomart database (https://www.ensembl.org/biomart, accessed on 31 December 2021). Visualization and Integrated Discovery (DAVID 6.8) [24] and Panther (http://pantherdb.org/, accessed on 31 December 2021) [25] databases were used for assessing functional profiles of DE genes based on the biological processes (BP), cellular components (CC), molecular functions (MF) and metabolic pathways categories of gene ontology (GO). The enrichment analyses were performed using the chicken genome information available in those databases.2.2. qPCR ValidationA relative quantification using qPCR was performed to confirm the expression profile of eight chosen DE genes in the GP using eight samples from the four normal and four from the FHS-affected group. The genes selected were solute carrier family 4 member 1 (SLC4A1), Rh- associated glycoprotein (RHAG), ankyrin 1 (ANK1), MAP kinase interacting serine/threonine kinase 2 (MKNK2), spectrin beta, erythrocytic (SPTB), adenosine deaminase (ADA), complement C7 (C7) and erythrocyte membrane protein band 4.2 (EPB42). Gene sequences were downloaded from Gallus gallus on Genebank (http://www.ncbi.nlm.nih.gov/gene/, accessed on 31 January 2022) and Ensembl (www.ensembl.org, accessed on 31 January 2022). Primers were designed in exon–exon junctions using the Primer-Blast online tool [26], and their quality was evaluated in the NetPrimer online software (http://www.premierbiosoft.com/netprimer/, accessed on 31 January 2022) (Table 1). For AC, the analysis was also performed in eight samples for five genes (SPTB, MKNK2, SLC4A1, C7 and EPB42), since the RHAG, ANK1 and ADA genes have already been confirmed in the analyzed population by Hul et al. [17]. Ribosomal protein 4 (RPL4) and Ribosomal protein 30 (RPL30) genes were used as reference for GP, and Ribosomal protein 5 (RPL5) and Ribosomal Protein Lateral Stalk Subunit P1 (RPLP1) for AC, as described by Peixoto et al. [15] and Hul et al. [17], respectively. The cDNA was synthetized using SuperScript III First-Strand Synthesis SuperMix (Invitrogen, Waltham (MA), USA), and qPCR reactions were carried out in QuantStudio 6 (Applied Biosystems, Waltham (MA), USA) equipment. The reactions had a final volume of 15 μL with 1×GoTaq qPCR Master Mix with BRYT Green (Promega, Madison (WI), USA), 0.13 µM of each primer and 2 μL of cDNA. Cycle threshold (Ct) mean for each replicate sample was obtained and normalized using the geometric mean of the reference genes, calculating 2−ΔCt [27]. Statistical analyses for group comparisons were performed using Mann–Whitney–Wilcoxon test in the R environment, considering DE as the genes with p ≤ 0.05.2.3. Integrated AnalysisTo identify the main molecular mechanisms involved in FHS, the common and exclusive DE genes in the AC and GP datasets were obtained, and the David [24] database was used to find BP involved with FHS based on the gene ontology database. The REVIGO tool [28] was used to reduce and highlight the most abundant BP in our data. Furthermore, gene networks with the common DE genes between the two tissues and those obtained with DE genes only detected in AC or GP were constructed using the STRING database [29].Furthermore, we used the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/gds, accessed on 9 November 2020) to search datasets with the terms “femoral head necrosis” and “epiphysiolysis” to better understand the role of the main DE genes involved with FHS across different species. This search returned two main datasets: identification of potential biomarkers for improving the precision of early detection of steroid-induced osteonecrosis of the femoral head (GSE123568) and gene expression profile of hip cartilage with necrosis of femoral head (GSE74089). The first one evaluated the gene expression on blood from 40 patients (10 normal and 30 affected with femoral head necrosis) and the second one, the hip cartilage in 24 patients (12 normal and 12 affected with femoral head necrosis). DE genes were obtained using the GEOR tool from the GEO dataset (https://www.ncbi.nlm.nih.gov/geo/info/geo2r.html, accessed on 9 November 2020). The list of DE genes from the two datasets obtained in the GEO database and the DE genes found in the GP and AC from our study were submitted to the InteractiVenn (http://www.interactivenn.net/, accessed on 10 November 2020) [30] to find common DE genes among the experiments.2.4. Polymorphism Identification Using the RNA-Seq DataThe Genome Analysis Tool kit 3.6 (GATK) [31] was used for polymorphism identification. We followed the best practices guidelines standard parameters for transcriptome variant analysis available on the GATK website (https://gatk.broadinstitute.org/, accessed on 1 February 2021). The Picard tools 2.5 (https://broadinstitute.github.io/picard/index.html, accessed on 17 May 2017) was used to generate the genome index, to assign read groups and mark duplicates. In the GATK, the CIGAR strings determination (SplitNCigarReads), qualities reassigning mapping, base recalibration, variants calling, and filtering were performed. To minimize the false positive variants, the following filters were used in GATK to filter and select variants: FS > 30.0, MQRankSum < −12.5, SNPcluster considering 3 variants in a 35 bp window, QD < 5.0, MQ < 50.0, GQ < 5.0, QUAL ≥ 30.0, ReadPosRankSum < −8.0 and DP ≥ 100.0. Furthermore, variants with less than 10 reads per sample were removed from the downstream analysis. Once the polymorphisms were identified in the GP and AC datasets, a filter was performed to obtain the variants that differed between normal and FHS-affected groups. The Variant Effect Predictor (VEP) tool [32] was used for variant annotation, effect and consequence predictions using the Gallus gallus genome (GRCg6a) with the Ensembl 103 annotation database version. Furthermore, the Stringdb [29] and EnrichR [33,34] databases were used to verify the interactions between DE genes and variants identified in genes.3. Results3.1. RNA-Sequencing, Mapping and CharacterizationThe RNA sequencing generated about 345 million paired end reads for all samples (N = 16), and after the quality control, approximately 306 million reads (88.78%) remained for further analyses (Supplementary File S1: Table S1), with an average of 16,355,490 reads for GP and 26,819,697 for AC. A clear separation between the samples from the two tissues, as well as between the normal and affected groups can be verified in the heatmap (Supplementary File S2: Figure S1). As expected, most of the genes were characterized as protein coding (95%), and the remaining 5% were classified in IncRNA, pseudogenes, miRNAs and other noncoding RNAs (Table 2).3.2. Differentially Expressed Genes in AC and GP TranscriptomesUsing the Ensembl annotation 101, in the AC transcriptome, 12,182 genes were identified and 106 genes were DE; as well, 99 (93.44%) were upregulated and seven (6.56%) were downregulated in the AC of the FHS-affected compared to the normal group (Supplementary File S1: Table S2). Regarding the GP tissue, 12,632 genes were expressed, where 324 were DE between the analyzed groups. From those, 174 were upregulated (53.7%) and 150 (46.3%) downregulated in the affected compared to the normal group (Supplementary File S1: Table S3).Comparing the DE genes between the AC and the GP, 36 DE genes were common to both transcriptomes (Table 3). From these 36, 33 were upregulated and three were downregulated in the FHS-affected group in both transcriptomes. From the 70 DE genes remaining in the AC data, four were downregulated and 66 were upregulated in the FHS-affected broilers (Table S2). Finally, 141 genes were upregulated and 147 were downregulated in the affected group when considering the DE genes exclusive to the GP tissue (Table S3).The enrichment analysis performed based on the common DE genes between the two tissues showed that the main biological processes found were related to ion transport, inflammatory and defense response, homeostasis, protein biogenesis and immune response processes (Figure 1).The most relevant BP for the AC tissue was the defense response to bacterium, cell chemotaxis and regulation of cell motility (Table S4). Conversely, the regulation of lipid metabolic process, cytokine-mediated signaling pathway, T cell activation, inflammatory response, interleukin-1 beta production, small GTPase-mediated signal transduction, and negative regulation of chondrocyte differentiation and cartilage development were the BP most prominent on GP tissues (Table S5).A gene network was constructed based on the 36 common DE genes. From those, 23 were recognized by Stringdb tool, where it was possible to observe a main network of grouped genes related to ion transport and translation factors (Figure 2). The transferrin (TF) gene is involved in antimicrobial activity, while some genes such as Hemoglobin subunit gamma-2 (HBG2) and Rh-associated glycoprotein (RHAG) are involved in oxygen and ion transport. Another group connected the PTPRC, CCL10, ISG12-2 and SAMD9L genes that are most related to the immune response BP (Figure 2).Another two gene networks were constructed based on the DE genes exclusive for AC or for GP. For AC, 61 genes were recognized by the Stringdb tool, which highlighted a group of genes related to the response against bacteria, with all genes involved in this BP being more expressed in the FHS-affected group (Figure 3). The second network was constructed with 266 GP DE genes, and the processes related to chondrocyte differentiation and cartilage development stood out, grouping genes that had the lowest expression in the affected group (Figure 4).3.3. qPCR ValidationIn the qPCR analysis, from the eight candidate genes evaluated in GP, four (C7, EPB42, MKNK2 and SPTB) were DE. The MKNK2 and SPTB genes were also DE in the AC. Furthermore, all evaluated genes were upregulated in the FHS-affected group (Figure 5A,B), having the same expression profile and therefore confirming the RNA-Seq results.3.4. Comparison of the DE Genes from AC and GP Tissues with Different DatasetsAfter the differentially expressed analysis, we compared the DE genes from the AC and GP tissues with other datasets obtained in the GEO database, as previously described. According to the GEOR analysis, the comparison performed between normal and FHN-affected individuals (GSE123568) evinced 5250 DE genes in the blood of those individuals, while 5978 DE genes were found in the hip cartilage of normal and FHN-affected individuals (GSE74089). Considering these four datasets, four genes were DE in all studies: SLC4A1 (Solute Carrier Family 4 Member 1), EPB42 (Erythrocyte Membrane Protein Band 4.2), SPTB (Spectrin Beta, Erythrocytic) and ANK1 (Ankyrin 1) (Figure 6A), all of them related to erythrocyte and ion transport functions.We also performed a comparison using DE genes from cartilage (chicken and human) and bone datasets to verify the presence of shared genes in the tissues affected by FHN/FHS. In this comparison, eight shared genes were found: the four previously described (SLC4A1, EPB42, ANK1 and SPTB) and four additional genes: CCL26 (Chemokine (C-C motif) ligand 26), ADA (Adenosine Deaminase), SLC25A37 (Mitoferrin-1) and MKNK2 (MAPK Interacting Serine/Threonine Kinase 2), all involved with immune response (Figure 6B).3.5. Variant Identification and AnnotationA total of 89,823 variants (SNPs and InDels) were found in the analyzed samples with the GATK tool. Out of those, 89 differed between the normal and FHS-affected groups (Supplementary File S1: Table S6). These 89 variants were processed in the VEP tool and approximately 3% from all of them (three) were first described in the current study, while 86 (~97%) have already been identified in chickens. One of the three novel SNPs was located in the 5′ UTR region of the NET1 (Neuroepithelial Cell Transforming 1) gene, another SNP overlapped the downstream region of two genes, CCM2 Scaffold Protein (CCM2) and NAC Alpha Domain Containing (NACAD, ENSGALG00000050931—predicted), and the other was a missense variant in the Staphylococcal Nuclease and Tudor Domain Containing 1 (SND1, ENSGALG00000054297—predicted). Variants were mainly classified as downstream gene variant, synonymous variant, upstream gene variant, intron variant, 3′ prime UTR variant and missense variants (Figure 7).Out of the described coding variants, 90% was classified in synonymous variants, 10% was missense (Supplementary File S1: Table S7). Considering the four missense variants, three were previously described in the genes PRKAB1 (rs733851944), SPATS2L (rs737569658) and TRIM25 (rs314175171), with a Sorting Intolerant from Tolerant (SIFT) score of 0.36 and 0.16. The novel variant (chr1:1049560) in the SND1 had a SIFT = 0, being predicted as deleterious.After identifying the variants that differ between the normal and affected groups in the RNA sequences, a search was performed to verify if some of the variants were present in the DE genes in both AC and GP tissues previously found between normal and FHS-affected samples. None of the variants were found in the common DE genes; however, two polymorphisms were found in the DE genes in the GP: a SNP (rs737937191) in the downstream region of the LOXL2 gene and the other in the amino acid position 52 of the SND1 (ENSGALG00000054297) gene.Some SNPs that differed between normal and affected broilers were found in genes that have already been described in genomic regions associated with related phenotypes in the chicken QTLdb (https://www.animalgenome.org/cgi-bin/QTLdb/GG/index, accessed on 31 December 2021), such as a SNP in the RPS24 gene, located in a QTL region associated with tibia volume (135,882, Animal QTLdb) (Supplementary File S1: Table S8). Furthermore, some of the genes that have variants differing between the normal and affected groups, such as Solute Carrier Family 25 Member 3 (SLC25A3), Yip1 Domain Family Member 3 (YIPF3), DEAD-Box Helicase 55 (DDX55), Exportin 5 (XPO5), RNA Polymerase I And III Subunit C (POLR1C), R3H Domain And Coiled-Coil Containing 1 (R3HCC1), Transmembrane P24 Trafficking Protein 2 (TMED2), Ras Responsive Element Binding Protein 1 (RREB1), charged Multivesicular Body Protein 7 (CHMP7), DNA Polymerase Eta (POLH) and Lysyl oxidase homolog 2 (LOXL2) have already been associated with waist-to-hip ratio adjusted for body mass index (BMI) in humans (Supplementary File S1: Table S9).The interactions between the DE genes and genes with variants were verified through a gene network, where it was possible to observe that those genes have physical or functional interactions and were grouped in five main subnetworks (Figure 8). For example, the ADA gene was DE in AC and GP and grouped with several genes, which have SNPs differing between normal and affected groups, such as the TRIM25, LUM, MYH9, RPS24 and NACAD.4. DiscussionThe mechanical stress caused by the overload of the skeleton that modern chickens sustain due to the high growth rates has been recognized as one of the main causes of locomotor problems [11,12,35]. Among these locomotor problems, the FHN/BCO is a disorder that has a large impact on the poultry production due to the huge economic losses and its negative effect on animal welfare [8,9,15]. One of the biggest issues to study FHN is its early detection, since the animals do not show clinical signs often, being visible only after slaughter [16].There are few studies related to FHN and other bone integrity problems in chickens, especially those approaching the molecular mechanisms involved with these conditions. Previous studies from our group have shown some biological processes (BP) and genes associated with FHN/BCO in different ages and lines of chickens [14,15,16,17]. Paludo et al. [13] pointed out that the downregulated expression of RUNX2 and SPARC genes may be associated with reduced vascularization and poor bone mineralization, increasing the risk of skeletal problems in chickens. The impaired collagen formation, connective tissue cell adhesion, and bone extracellular matrix (ECM) were also identified as predisposing factors for BCO and other leg disorders [14,15,16,17]. The translocation of bacteria from blood to bone has also been suggested as a secondary condition [16], whereas the bacterial profile of the microbiome in the blood of animals affected with BCO was different from healthy ones [36]. However, most of these studies focus only on the bone tissue, and there are no studies evaluating the articular cartilage and the femoral head jointly. Therefore, our study aimed to find common DE genes and BP between those tissues as well as genes exclusively DE in each tissue (AC and GP), and some of these genes were validated by qPCR (Figure 5). We also identified variants in the AC and GP transcriptomes between normal and FHS-affected broilers through an integrated analysis of the RNA sequencing datasets previously published by Peixoto et al. [15] and Hul et al. [17].In the global transcriptome characterization of the GP and AC tissues, about 12,000 genes were expressed in both tissues. From those, 11,141 were shared and 891 and 441 genes were exclusively to the GP and AC, respectively. As expected, due to the RNA sequencing methodology used, approximately 95% of the genes were coding, although lncRNAs, miRNAS, mitochondrial and miscRNAs were also identified in the analyzed datasets (Table 2).A total of 36 DE genes were shared by both tissues, and 91.6% of the common genes were upregulated in the FHS-affected group (Table 3). Some of those genes have already been highlighted in other studies [14,16,17]. However, when the common BPs were evaluated, most of them were related to ion transport, defense response and those related with tissue homeostasis (Figure 1).Two main branches were observed in the common DE genes network (Figure 2): one was composed of the TF, HBG2, ANK1, SLC25A37, EPB42, RHAG, ADD2, KEL, MKNK2 genes and another by the PTPRC, CCLI10, ISG12-2 AND SAMD9L genes. In general, the first group of genes are related to blood type and to cartilage regeneration, and the second is related to inflammatory response [15,37]. Moreover, a set of variants was identified differing between normal and affected groups, where none of them were located in the common DE genes. It was observed that some of these polymorphisms were in genes that are co-expressed or co-regulated with DE genes (Figure 7). The possible role of some of these genes in the FHS are discussed below.The transferrin (TF) was in the main branch of the gene network and its primary function is to transport iron to tissues that require this mineral [38], being involved in defense against systemic infection [38]. There are two forms of transferrin in birds: ovotransferrin, found in oviducts, and serum transferrin, secreted by the liver. Serum transferrin may stimulate cell proliferation and is regulated by iron levels, while ovotransferrin has a bacteriostatic role independent of iron levels [39,40]. Furthermore, the TF has a function in endochondral ossification, being considered the angiogenic molecule released by the hypertrophic cartilage [41]. TF presented strong interactions with the HBG2 (Hemoglobin Subunit Gamma 2) gene, which is involved in the transport of oxygen from the lungs to the peripheral tissues, where the beta chain is a component of adult hemoglobin A and D [8]. HBG2, as the TF, was upregulated in the FHS-affected broilers (Table 3). The gamma globin genes (HBG1 and HBG2) are expressed in several tissues, such as fetal liver, spleen and bone marrow [42]. Chicken has multiple types of globins, and erythropoiesis occurs in two waves: one primitive that acts in blood cells during early embryonic development, and the other producing definitive erythrocytes in late embryonic and post-hatching development [43]. In the bone and cartilage, globin function is not clear, and no information is available on its involvement with FHS in chickens.The SLC25A37 (Solute Carrier Family 25 Member 37) works as a solute transporter located in the internal mitochondrial membrane, importing iron, which is essential for the synthesis of mitochondrial heme and iron–sulfur clusters [44]. In general, solute carrier proteins are essential to transport a wide set of molecules, such as ions, amino acids and vitamins to the tissues. Recently, it has been demonstrated that they are sensitive to extracellular levels of phosphate [45] and also regulate the pH of the cells [46]. In humans, several bone morphogenic proteins such as BMP13, BMP14 and BMP15 were responsible for regulating genes of the SLC family [47].The ANK1 gene encodes a binding protein of the cytoskeleton, helping to bind other membrane proteins to the actin–spectrin cytoskeleton [48]. It also acts on contact activation, maintenance and proliferation of specialized membrane domains [48]. Multiple ankyrin isoforms with different affinities for various target proteins are expressed in a tissue-specific manner, regulated by development. The ANK1 is usually found in erythrocytes, but it has already been found in the brain and muscles [49]. In our study, the high expression of ANK1 can possibly affect the actin structure of the cytoskeleton, changing the structural integrity of the femoral articular cartilage and contributing to the occurrence of FHS. In addition, its expression is related to cell damage, which can be a consequence of FHS, because when FHS starts, positive regulation can act as a sign of attempt to combat the progression of this condition. This gene is also involved in the upregulation of inflammatory cytokines in osteoarthritic lesions [50]. The ANK1 gene is regulated by the EPB42, another gene that was upregulated in the FHS-affected group. The ANK1, EPB42 and RHAG (Figure 2) are closely related to some genes that have SNPs differing between the two analyzed groups, such as TMED2, DDX55 (Figure 7) and POLRC1.The MKNK2 (Serine/threonine-protein kinase 2) gene interacts with MAP kinase and may respond to environmental stress and cytokines. Among its related pathways are the interleukin 1 signaling pathway. MAP/ERK signaling acts as a mediator of the suppressive effects of IFN-gamma in hematopoiesis [51]. The MKNK2 is not normally expressed in rats, but appears in stressful situations, showing its adaptive function and can also be a signal for cellular apoptosis [52]. Bringing it to our study, this gene may be involved in the initial phase of FHN with the function of initial signaling.Another DE gene, the MYH15, has also been associated with the adrenergic signaling process in cardiomyocytes, in cases of imbalance in the O2 supply and removal of CO2 from tissues [53]. This gene could be associated with FHS as a consequence of local inflammation.The EPX gene is released during the immune response by eosinophils, having a cytotoxic effect on cells [54,55] and bactericidal activity [56]. The eosinophil peroxidase, an extremely cytotoxic molecule that has anti-inflammatory and pro-inflammatory properties, regulates inflammation by combat invading microorganisms [57,58]. In chickens, the EPX gene has been studied as a biomarker for inflammatory events in the gastrointestinal tract [59] and has been related to leg disorders [15,17]. The EPX upregulation in broilers affected with FHS may indicate the immune system response to inflammation and, in severe cases, may be related to local necrosis.The extracellular matrix (ECM) is a stable structural component that is located under the epithelium and close to connective tissue [60], and is responsible in providing support to the tissues and organs throughout the body. It acts in biochemical processes of the body assisting in the differentiation, morphogenesis and homeostasis of tissues [61]. The upregulated genes ADA and RHAG also participate in the process of metabolic glycosaminoglycans (GAGs) and aminoglycans involved in the metabolism of ECM [62], indicating that the body tries to repair the damage caused by FHS through tissue remodeling. Furthermore, the ADA enzyme acts as an endogenous regulator of the adaptive immune system, especially on proliferation and differentiation of T lymphocytes, regulating cell metabolism and triggering several physiological effects on cell proliferation [63,64].Inflammation is an essential component of the immune system; however, an excess of inflammation can cause tissue damage [65]. The ADA gene also acts as a sensor providing information to the immune system regarding tissue damage, protecting cells from excessive tissue damage associated with inflammation [66]. The upregulation of ADA has a regulatory role in immune responses, acting on the activation and regulation of lymphocyte and neutrophil levels [64,67]. In our study, the ADA was upregulated in AC and GP in the FHS-affected group (Table 3), possibly due to the large tissue inflammation that occurs in the affected animals. The presence of specific variants in this gene between the studied groups also highlights that those genetic mechanisms could be involved with FHS in chickens. The ADA gene is in a branch of the gene network with several genes that have SNPs with different genotypes between the normal and affected groups, such as TRIM25, RPS24, NACAD, and others (Figure 7). The TRIM25 and ADA are important genes to the RNA machinery [66,68], and the involvement of the ADA in bone metabolism has been observed in humans [69]. Deficiency in ADA expression has led to a reduction of bone volume and downregulation of RANKL expression, while ADA upregulation can lead to skeletal abnormalities such as scapular spurring [70,71].The ADA and IFI6 genes also play an important role in the regulation of apoptosis, which is an essential physiological mechanism in the development and tissue homeostasis [14,72]. The IFI6 gene, also known as ISG12, regulates cellular metabolism during the differentiation of osteoblasts and apoptosis [14,73]. The upregulation of the IFI6 gene may be related to a causative factor, stimulating apoptosis in the articular cartilage, leading the animal to be more susceptible to FHN.Furthermore, once the exclusively DE genes in the GP were evaluated, the main biological processes were those related to chondrocyte differentiation and cartilage development, grouping three main genes that had the lowest expression in the affected group: CHADL, GDF5 and PTHLH (Figure 6, Table S5). Chondroadherin-like (CHADL) is a member of a family of collagen-associated small leucine-rich proteins (SLRPs) responsible for signaling of differentially regulated collagen fibrils during development, homeostasis, or pathogenesis. CHADL is expressed in cartilaginous tissues, influences collagen fibrillogenesis and negatively modulates chondrocyte differentiation [74].In humans, increased GDF5 gene expression is related to joint tissue remodeling. After joint damage, its high expression in chondrocytes is observed, both in the new cartilage recovery tissue and in the adjacent damaged cartilage [75]. The PTHLH gene promotes the proliferation of chondrocytes and inhibits their hypertrophy and terminal differentiation. In patients with osteoarthritis, this gene is highly expressed in chondrocytes and synovial fluid [76].Among the BP of the AC, the response against bacteria can be highlighted, with all genes involved in this BP being more expressed in the FHS-affected group (Figure 1). Among them, cathelicidin-2 (CAMP) and cathelicidin-3 (CATH3) stand out, which have a broad spectrum of antimicrobial activity and the ability to limit inflammation by inhibiting the activation of Toll-like receptor 2 (TLR2) and TLR4 [77].These results suggest that the growth plate is not fully efficient in the process of recovery and remodeling of damaged cartilage. Conversely, articular cartilage can partially control bacterial infection and the intensity of inflammation that can intensify tissue damage. Other mechanisms should be evoked.In addition to the identification of the DE genes, biological processes and SNPs found in our study, another interesting result presented here was the comparison of the chicken transcriptomes with other datasets. When we compared the DE genes obtained in the AC and GP tissues with other gene expression datasets obtained in the GEO, we found four genes in common, SLC4A1 (Solute Carrier Family 4 Member 1), EPB42 (Erythrocyte Membrane Protein Band 4.2), SPTB (Spectrin Beta. Erythrocytic) and ANK1 (Ankyrin 1) (Figure 6A), all of them related to erythrocyte and ion transport functions. In the second comparison, using the DE genes of bone and cartilage datasets, besides the four previously described genes (SLC4A1, EPB42, ANK1 and SPTB), four additional genes were found (CCL26, ADA, SLC25A37 and MKNK2), all involved with immune response. Furthermore, two genes, the CCL26 and SLC4A1, have not been enriched in the gene network. CCL26 (chemokine (C-C motif) ligand 26) is from the cytokine family of secreted proteins involved in immunoregulatory and inflammatory processes, displaying chemotactic activity for normal peripheral blood eosinophils and basophils, and it may contribute to the eosinophil accumulation in atopic diseases. Finally, the SLC4A1 gene encodes a protein that is part of the anion exchanger family and is expressed in the plasma membrane of erythrocytes, involved in the transport of carbon dioxide from the tissues to the lungs. Furthermore, this gene is located in QTL regions previously described as associated with femur mineral content and femur weight, which reinforces its importance as a good candidate gene to FHS.Based on the functions and the high expression of the genes that were DE in both tissues, we had two hypotheses: the first is that the high expression of these genes may be a consequence of the FHS, due to cell damage and from the attempt to combat the progression of this condition. The second hypothesis is that the high expression and possible mutations of these genes cause FHN, as they can alter the structure of the actin cytoskeleton, affecting the structural integrity of the femoral articular cartilage, contributing to the occurrence of FHS. Therefore, using the two datasets in this study, we could highlight common genes that were regulated in the two tissues, as well as those that were exclusively DE in each one. With this information, new genes involved with FHS in broilers were described, where most of them have never been cited as candidates to this condition in chickens. We have also shown conserved mechanisms among huma and chicken osteonecrosis. Furthermore, genetic variants were identified; some of them were described for the first time in this study, and some of them differed between the normal and FHS-affected groups and could be further investigated as molecular markers for this condition. Finally, our results can also contribute to understand the onset of this condition in other species since the molecular mechanisms seem to be at least in part conserved among them.5. ConclusionsWith the integrated analysis of gene expression of the two tissues and variant identification, we were able to detect strong candidate genes and SNPs related to FHS/FHN in chickens. Comparing the AC and GP DE genes with the other datasets related to FHN allowed us to identify genes possibly involved with FHN, evincing the shared mechanisms across different species. Finally, the role of SLC4A1, RHAG, ANK1, MKNK2, SPTB, ADA, C7 and EPB420 genes should be more explored in order to validate them as candidates for FHS/FHN.

animals : an open access journal from mdpi

10.3390/ani11092696

PMC8472390

Heterozygosity-rich regions (HRRs) are regions of high heterozygosity, which can harbor important genes associated with key functional traits such as immune response and disease resilience. Runs of homozygosity (ROH) are contiguous homozygous segments of the genome, which can be informative of the population’s history, structure, demography events, and overall genetic diversity. We first detected factors impacting the identification of ROH and HRR in worldwide sheep populations, which were artificially selected for specific purposes or under natural conditions. We also identified common regions of high homozygosity or heterozygosity among these populations, where a diversity of candidate genes with distinct functions might indicate differential selection pressure on these regions in breeds with different trait expression. Moreover, we evaluated a tool commonly used in the corporate environment, making use of the business intelligence (BI) concept to support managers in the decision-making process, which allowed us to combine results from multiple analyses and create visualization schemes integrating different information. Our findings and proposed tools contribute to the development of more efficient breeding strategies and conservation of genetic resources in sheep and other livestock species.

In this study, we chose 17 worldwide sheep populations of eight breeds, which were intensively selected for different purposes (meat, milk, or wool), or locally-adapted breeds, in order to identify and characterize factors impacting the detection of runs of homozygosity (ROH) and heterozygosity-rich regions (HRRs) in sheep. We also applied a business intelligence (BI) tool to integrate and visualize outputs from complementary analyses. We observed a prevalence of short ROH, and a clear distinction between the ROH profiles across populations. The visualizations showed a fragmentation of medium and long ROH segments. Furthermore, we tested different scenarios for the detection of HRR and evaluated the impact of the detection parameters used. Our findings suggest that HRRs are small and frequent in the sheep genome; however, further studies with higher density SNP chips and different detection methods are suggested for future research. We also defined ROH and HRR islands and identified common regions across the populations, where genes related to a variety of traits were reported, such as body size, muscle development, and brain functions. These results indicate that such regions are associated with many traits, and thus were under selective pressure in sheep breeds raised for different purposes. Interestingly, many candidate genes detected within the HRR islands were associated with brain integrity. We also observed a strong association of high linkage disequilibrium pattern with ROH compared with HRR, despite the fact that many regions in linkage disequilibrium were not located in ROH regions.

1. IntroductionRuns of homozygosity (ROH) are contiguous homozygous segments of the genome, which can arise from the mating of two related individuals that transmit identical haplotypes to their offspring [1]. Long ROH segments are often associated with recent inbreeding, while short ROH are linked to ancient inbreeding, owing to the higher probability of recombination events occurring as the number of generations increases [2]. Thus, ROH analyses are paramount for estimating genetic diversity metrics such as ROH-based inbreeding coefficient (FROH), i.e., the ratio of the total length of an individual’s autosomal genome in ROH to the total length of the autosomal genome covered by single nucleotide polymorphism (SNP) [2]. FROH tends to be more accurate than pedigree-based inbreeding coefficients and enables the identification of specific genomic regions with greater inbreeding [3]. The identification of ROH regions also contributes to the characterization of population history, structure, and demographic events [4], and further reveals the selection signatures that are characterized by fixation of alleles under high selection pressure on a population [5,6], with a subsequent increase in homozygosity in regions around these alleles [7,8].Runs of homozygosity have been extensively studied across many species for the quantification of inbreeding [2,3,9,10,11,12], detection of selection signatures [13,14,15], and comparison of statistical methods and identification parameters [4,16,17,18,19]. Runs of heterozygosity, most appropriately defined as heterozygosity-rich regions (HRRs) [20], represents a more recent concept [21], and is not as well described in the literature as ROH [20,22,23,24]. HRRs can also provide insights about population structure and demographic history [24], and these HRRs may harbor important loci for key functional traits such as immune response, survival rate, fertility, and other fitness traits [25]. To the best of our knowledge, there are no reported studies characterizing HRRs in sheep populations.The process of sheep domestication (Ovis aries) started in the Fertile Crescent, approximately 11,000 years ago [26]. Nowadays, sheep are raised across the globe under divergent environmental conditions. While some breeds have been artificially selected for certain purposes (e.g., meat, milk, or wool), other populations have evolved without direct human interventions. The characterization of ROH patterns on populations selected for specific or divergent purposes could reveal genomic regions of predominant homozygosity related to the fixation of certain alleles associated with the traits under selection. Furthermore, HRRs on these populations may be an indicator of regions associated with important fitness traits [25]. Purfield et al. [27] analyzed the genome of sheep from six meat breeds to identify selection signatures using ROH and two complementary methods, Fixation Index and hapFLK [28], and observed regions under putative selection that frequently overlapped with high ROH regions. Dzomba et al. [29] also characterized the distribution of ROH islands in 13 South African sheep breeds and 31 worldwide sheep populations, which enabled the identification of common and unique ROH islands across populations.Data visualization is a critical step in genomic data analytics for proper interpretation of the findings. Plotting results instead of looking at tabular data frequently provides additional insights into the patterns and trends of the results. Several tools have been developed to visualize genomic data, which can be challenging owing to the structure and complexity of the data [30]. Furthermore, the volume of data that usually results from independent analyses may present an additional challenge for the integration and comparison of such results, which are usually projected in distinct static images. Business intelligence (BI) is a concept used in the corporate environment to support managers in the decision-making process by enabling informed decisions based on data [31]. Distinguishing features of BI tools include the possibility of creating dynamic data visualizations and integrating distinct data sources. Consequently, BI tools provide a great opportunity for combining results from different analyses and navigating through parameters more quickly, such as by changing the exhibited chromosome or breed in a click.In this study, we chose 17 worldwide sheep populations of eight breeds, which were intensively selected for different purposes (meat, milk, or wool), or locally-adapted breeds. Our main objectives were as follows: (1) to identify and characterize factors impacting the detection of ROH and HRR in sheep breeds selected for different breeding goals; (2) to evaluate the feasibility of using a BI tool to visualize and filter the observed results, integrating multiple types of information in a single visualization, such as ROH islands and previously-identified quantitative trait loci (QTL) or linkage disequilibrium (LD) pattern; and (3) to compare different parameters for the identification of HRR, with the aim of providing some basis for future studies of this nature in sheep and other livestock populations.2. Materials and Methods2.1. Genotypic Data and Quality ControlThe genotypic data used in this study were made publicly available by the International Sheep Genomics Consortium [32], and downloaded from the public online WIDDE database (http://widde.toulouse.inra.fr/widde/, accessed on 15 March 2021). Genotypes were obtained using the Illumina® OvineSNP50 BeadChip, and 1186 sheep from 17 worldwide populations from eight breeds were chosen for this study (Table 1). These populations were classified based on their main breeding purpose, such as meat (Lacaune, Suffolk, and Texel), milk (Churra, East Friesian Brown, and Lacaune), wool (Merino), or locally-adapted (Soay and Tibetan). The Lacaune breed comprised two populations (selected for meat or milk), while Merino, Suffolk, and Texel comprised distinct populations sampled in various countries (Table 1). The following filtering criteria were applied in the quality control (QC) for the entire dataset (all populations together): (i) markers on non-autosomal chromosomes, (ii) with missing call rate >0.05, and (iii) individuals with missing call rate >0.05 (although no individuals were removed). Data were not pruned for minor allele frequency (MAF) nor LD, as removing fixed alleles was not desired, and LD pruning has distinct effects on the detection of ROH depending on the population [19]. After QC, 46,095 markers were kept, and no individual was excluded from the dataset. QC was performed using the PLINK software v1.9 [33].2.2. Data Integration and VisualizationThe Business Intelligence Software Tableau V. 2020.1.14 (https://www.tableau.com, accessed on 12 September 2021) was used under the Student License to integrate data obtained from the previously described databases and analyses. As one of its key features is the ability to integrate data from multiple sources, the format of the data flat files was minimally rearranged according to the Tableau requirements, using the R software [34]. All visualizations presented in this paper were created on the BI tool, and screenshots were taken to exhibit the visualizations with certain filters applied, which were mentioned in the respective figures. The visualizations created on the Tableau platform are divided in dashboards, where one can observe the distribution of the ROHs, HRRs, and islands in the chromosomes, as well as QTLs, LD pattern, and additional information. One can zoom in the visualizations and apply different filters, such as population, chromosome, position, type of island (ROH or HRR), LD (measured as r2), and others. Hovering the mouse over the visualizations exhibits extra information. The entire file, with visualizations and data integrated from all the analyses, can be accessed in the Supplementary Materials section (File S1).2.3. Detection of Runs of Homozygosity and Heterozygosity-Rich RegionsThe R package detectRUNS [35] was used to detect both ROH and HRR, through the “Sliding Windows” method. In order to identify ROHs, the following criteria were applied: (i) the minimum number of SNPs in an ROH was 20; (ii) the minimum length of an ROH was 1000 kb (equivalent to ~20 SNPs); (iii) the minimum marker density was set to one SNP every 70 kb; (iv) the maximum gap within an ROH was 250 kb; (v) the size of the SNP window was set to 20 SNPs (according to Meyermans et al. [19], an appropriate size for the SNP window is equal to the number of SNPs in the smallest ROH); (vi) a maximum of one missing and one heterozygous SNP were allowed within an ROH and SNP window; and (vii) the window threshold of 0.05 was kept as the package default. The population average of the proportion of ROH coverage in each chromosome was calculated as (LROHCHR/N)/LCHR, where LROHCHR is the sum of the total length of the chromosome covered by ROH of all individuals in a population, N is the number of individuals in a population, and LCHR is the length of the chromosome calculated as the position of the last marker minus the position of the first marker on the chromosome in base pairs.For the detection of HRR, different scenarios were evaluated (Table 2), and the following criteria were applied: (i) the minimum number of SNPs in a HRR ranged from 5 to 10; (ii) the minimum length of a ROH ranged from 10 to 400 kb; (iii) the minimum marker density was set to one SNP every 70 kb; (iv) the maximum gap within a HRR was 1000 kb; (v) the size of the SNP window was equal to the minimum size of a HRR; (vi) the maximum number of homozygous SNPs allowed within a HRR and SNP window ranged from one to three; (vii) the maximum number of missing SNPs allowed within a HRR and SNP window was one or two; and (viii) the window threshold was kept as the default (0.05).2.4. Definition of ROH and HRR IslandsThe ROH and HRR islands were defined specifically for each population, following the methodology described by Purfield et al. [27]. The R package DetectRuns [35] was used to obtain the proportion of times each SNP fell inside a run in each population, which corresponded to the locus homozygosity or heterozygosity in the respective population. In order to define the ROH and HRR islands, the top 0.999 SNPs of the percentile distribution of the locus homozygosity or heterozygosity range within each population were selected, determining different thresholds of within ROH/HRR frequency for SNPs to be included in the islands. From these top frequency SNPs within each chromosome, markers with a distance further than 250 kb from the previous (according to the max gap defined for the ROH detection) were identified as the start of a new island, and the minimum and maximum positions of SNPs in each island were assigned as the start and end of each island, respectively.2.5. Identification of Regions in Strong Linkage DisequilibriumThe degree of LD was calculated as r2 for each population individually, using the PLINK (v1.9) software [33]. A different QC was applied to remove SNPs with MAF lower than 0.05 in each population and reduce bias in the LD estimation. Markers with a missing call rate higher than 0.1 were also excluded. The PLINK default threshold of r2 > 0.2 was used to select the analyses output.2.6. Gene Annotation, Gene Ontology (GO), and KEGG Pathway Enrichment AnalysesThe regions defined as ROH and HRR islands were annotated for gene content from the Ensembl database [36] within the coordinates using the R package GALLO [37]. These genes were further analyzed for gene ontology (GO) terms and metabolic pathway information from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [38]. The GO and KEGG pathway enrichment analyses were carried out with the R package WebGestaltR version 0.4.4 [39], using the method over-representation analysis (ORA), separately for eight subsets of genes, corresponding to the genes identified within the two types of island (ROH and HRR) and the four breed groups (meat, milk, wool, and adaptation). The model organism selected was Homo sapiens, as it is genetically closer to Ovis aries, which was not available. As the gene IDs could not be directly used to perform the analyses, only genes with gene symbol information were used. In order to obtain missing gene symbols, two methods were used: (i) the protein IDs were identified through the ovine genes’ Ensembl IDs and these protein IDs were used to obtain the gene symbols on Uniprot Kb; (ii) using the ovine Ensembl IDs, the Entrez IDs of the genes were obtained and the Rambouillet v1.0 database was used to annotate the orthologous genes in the bovine (ARS-UCD1.2) and human (GRCh38.p13) databases. Only orthologous genes with over 70% of similarity with the sheep sequence were maintained for further analyses. After matching the recovered gene symbols with the respective gene IDs, the GO and KEGG pathway enrichment analyses were performed. The GO terms and pathways were considered enriched after a multiple testing correction using a 5% false discovery rate (FDR).2.7. QTL AnnotationQTL information was obtained from the Animal QTL Database [40] for the OAR 3.1 assembly, and plotted against the ROH and HRR information to identify economically relevant regions previously described in the literature within the relevant genomic regions. Only QTL identified based on SNP markers were included through the implementation of Tableau filtering criteria (map type = genome).3. Results3.1. Runs of HomozygosityA total of 80,639 ROH were identified in the 1186 genomic samples analyzed (67.99 ± 47.32 ROH per sample). ROH lengths ranged from 1000 kb (minimal detectable size) to 50,908 kb, identified in the East Friesian Brown population. Figure 1a shows the average length of the genome covered by ROH, and Figure 1b shows the average number of ROH for each population. ROH were classified by length, in classes of 1 to 2, 2 to 4, 4 to 8, 8 to 16, and over 16 Mb. Australian Poll Merino had the lowest average coverage of ROH (98,308 kb), and Lacaune (meat) had the lowest average number of ROH (34.7), while Soay had both the highest average coverage of ROH (484,828 kb) and the highest average number of ROH (188.4). The profile of ROH varied greatly among populations (Figure 1). The percentage of ROH in length classes for all populations is shown in Supplementary Table S1. All populations had more ROH between 1 and 2 Mb than in the other classes (Figure 1 and Table S1), and Tibetan had the highest proportion among populations (63.36%), while also having a considerable proportion of ROH longer than 16 Mb (2.13%). East Friesian Brown had the highest proportion of ROH longer than 16 Mb (3.74%), while Scottish Texel had the lowest (0.12%).Figure 2 provides an example of population averages of the proportion of ROH coverage in each chromosome for five populations. The results for all the other populations are presented as Supplementary Material (File S1). Australian Poll Merino and Tibetan had low percentages of ROH coverage on all chromosomes (less than 10%), while East Friesian Brown and Soay had over 20% in some chromosomes. However, ROH longer than 16 Mb were more frequently detected in the East Friesian Brown than in the Soay population (Figure 1). Scottish Texel had relatively few ROH longer than 16 Mb, whereas Tibetan, also with a low percentage of ROH coverage, had a greater amount of long ROH (mainly located in chromosomes OAR7, OAR16, OAR22, and OAR25). The distribution of ROH across chromosomes also varied among populations, as some had a significant amount of ROH coverage on a given chromosome and others had a small amount on that same chromosome.Plotting the ROHs of each individual enables the visualization of patterns in the distribution and size of ROH for the population, in a complementary manner to the previous figures. Figure 3 shows the distribution of ROH on OAR2 for five sheep populations, as an example (Supplementary File S1 for all populations). Hovering the mouse over each run exhibits a tooltip with specific information, such as start and end positions, number of markers, and length (File S1). New Zealand Texel, German Texel, Scottish Texel, and Soay presented clear patterns of runs on the region from 107,000 kb to 120,000 kb (Figure 3 and File S1). The different proportions of short and long ROHs are again evident when comparing East Friesian Brown and Soay populations, despite both having a similar ROH coverage on the chromosome OAR2 (Figure 2). Furthermore, we can visualize gaps between close runs as a repetitive pattern on different individuals and populations, which could be otherwise considered as single longer runs.3.2. Heterozygosity-Rich Regions Detection ScenariosFor the detection of HRRs, nine scenarios with different parameters (Table 2) were tested on the entire dataset of sheep populations. In scenarios 1 to 3, the minimum length allowed for an HRR was 400, 250, and 10 kb, respectively. Scenarios 4 to 6 were similar to scenario 2, with a maximum number of homozygous allowed in an HRR and SNP window of 2, 1, and 1, respectively, and a maximum number of missing SNPs allowed in an HRR and HRR window of 2, 2, and 1, respectively. Scenarios 7 to 9 were similar to scenarios 1 to 3, with a minimum number of SNPs allowed and an SNP window of five SNPs.Figure 4 presents the total number of HRRs (Figure 4a), the maximum length in Mb (Figure 4b), and the minimum and maximum number of SNPs within an HRR (Figure 4c) for each scenario. The number of HRRs detected increased as the minimum length allowed for an HRR was reduced, and it decreased as the minimum number of SNPs, SNP window size, maximum number of homozygous, and missing SNPs allowed were reduced (Figure 4a). The maximum length of the detected HRR was reduced when the minimum number of SNPs and the SNP window size were reduced. The impact of the number of homozygous SNPs allowed on the maximum length of the detected HRR was not so clear, and the reduction in the number of missing SNPs allowed did not affect the maximum length of the detected HRR (Figure 4b). The minimum number of SNPs in an HRR remained the same as the minimum allowed in the parameters for all scenarios, except for scenario 7, where it was one SNP larger (Figure 4c). The maximum number of SNPs in an HRR varied from 16 to 19. Scenarios with a smaller minimum number of SNPs in an HRR had shorter HRRs in terms of the maximum number of SNPs, and the effects of reducing the number of maximum homozygous allowed were not clear. Reducing the number of missing SNPs allowed had no effect on the maximum number of SNPs in an HRR (Figure 4c).Scenario 2 was chosen to carry out the identification of HRR islands. In this scenario, the average number of HRR per animal identified across populations was 139.59, the average length was 459.894 kb, and the average of the total length was 64,198 kb. Table 3 presents each population average of total HRR length (KB) per animal and number of HRR per animal. Soay is the population with the least number of HRRs per animal (104) and total length (47,897.39 kb) per animal, and Australian Suffolk was the population with the largest number of HRRs (154) per animal and total length (70,704.99 kb) per animal (Table 3). The number of SNPs within an HRR ranged from 10 to 18, with 76.85% of the runs composed by the minimum number of SNPs (10), and only one run composed by 18 SNPs (Figure S1).3.3. Presence of Linkage Disequilibrium on ROH and HRR IslandsThe fact that LD plays a role in the formation and maintenance of ROH throughout generations has been reported by many authors [2,41,42,43]. In the case of ROH islands, which refer to short runs that are present in a representative portion of the population, LD may present an even stronger influence. For the purpose of investigating the association between LD and the presence of ROH and HRR islands, the islands were plotted against SNPs’ pairwise calculation of r2. To avoid the overlap of longer regions in LD by shorter regions within closer SNPs, we applied different filters in Tableau, restricting the minimum LD distances (length) and r2, thus filtering the information shown in the visualizations.The plots of regions in LD against both ROH and HRR islands in chromosomes where common regions for more than one population were identified are presented in Figure 5 and Figure 6, respectively. We applied the same filters in both visualizations, first allowing only LD calculated between SNPs from 0.5 to 1 Mb apart and with r2 > 0.5, and later setting r2 > 0.9. In Figure 5a, we can observe that regions with relatively strong LD (r2 > 0.5) spanning over 500 Kb frequently overlap with ROH islands; however, not all ROH islands fall upon such regions. In Figure 5b, only regions in very strong LD (r2 > 0.9) are exhibited, and such regions tend to be located close to where regions in ROH islands common to more than one population were identified. In the case of HRR islands, however, their occurrence seems to be independent from the regions in LD (Figure 6a,b).We also observed the extent of LD within and close to ROH and HRR islands shared among populations. As an example, Figure 7 shows the approximation of the region between 29 and 43 Mb in OAR6, where a common ROH island was detected for up to seven sheep populations. As the r2 threshold increases, it is possible to visualize which regions are in stronger LD for each population. We considered a minimum length of 250 kb for the calculation of r2, given that the objective of this analysis was to observe blocks of LD, and the presence of small adjacent segments of LD could lead to a misrepresentation of the extent of LD in longer distances. In Figure 7a, only three populations presented at least a fraction of the detected islands free of LD blocks stronger than 0.3, showing that most of the islands in this region were under the influence of some amount of LD. However, considering LD blocks stronger than 0.5, the number of regions drastically decreased (Figure 7b), and only three of the seven populations presented LD blocks stronger than 0.9 within the islands detected on this region (Figure 7d).3.4. Identification of ROH and HRR Islands and Gene AnnotationThe ROH and HRR islands were defined as the SNPs within a run present on a percentage of the population above a certain threshold, defined as the 99.9% quantile of the distribution of each population. Scenario 2 was chosen as the base scenario for HRR detection. Fifty-seven ROH islands and 115 HRR islands were identified from the 17 sheep populations, after excluding HRR islands with four or less SNPs. This criterion was applied to the HRR islands in order to avoid small regions that may have resulted from the adoption of less stringent parameters for HRR detection. The list of all detected ROH and HRR islands from each population is presented in Supplementary Tables S1 and S2, respectively. The largest ROH island was found in the New Zealand Texel population, on OAR2 between 109,132 and 111,301 kb, with a length of 2,169,874 kb. The Australian Merino population had the shortest ROH (OAR4; 47,347,002–47,397,772), with a length of 50,770 kb (Table S2). The longest HRR was found in the Scottish Texel population (OAR21; 231,657–956,070), with a length of 724.413 kb, while the shortest HRR was found in the Australian Industry Merino population (OAR18; 16,667,614–16,791,775), with a length of 124,161 kb (Table S3).There were 898 candidate genes identified within the ROH and HRR islands, from which 577 had gene symbols identified in the Ensembl database. Fifty-nine gene symbols were retrieved in total, from which one was exclusively identified from Uniprot Kb, eight were orthologous only to cattle genes, and seven to human. GO terms and pathways associated with genes identified within the ROH and HRR islands were tested for evidence of functional enrichment within the group and type of island where they were identified. Table 4 presents the GO terms enriched and the respective islands associated; all other GO and pathways can be found in Table S4. There were eleven enriched GO terms in total, five in adaptation HRR, one in milk HRR, and five in wool ROH. No pathway was enriched.Regions in ROH or HRR islands common to two or more populations are described in Table 5 and Table 6, respectively. We identified regions in four chromosomes (OAR2, OAR6, OAR10, and OAR11) where ROH were frequent in more than one population. OAR2 harbored the greatest number of such regions (Table 5). Five regions were identified as common regions in HRR islands, and four contained at least one gene (Table 6).3.5. Overlap of Known QTL with ROH and HRR IslandsIn an effort to investigate whether there is overlap of ROH or HRR islands with previously reported QTL, data from the Sheep QTL database (https://www.animalgenome.org/cgi-bin/QTLdb/OA/index, accessed on 8 August 2021) were plotted against ROH and HRR islands. Figure 8 shows a region on OAR6 between 15 and 80 Mb, harboring a few islands from eight sheep populations. On the top of the image, all types of QTL are selected. Applying QTL type filters, it becomes clearer which QTLs are overlapping with each island. Health association QTLs overlap with an HRR island in the Australian Suffolk population, as well as an ROH island from the Tibetan population and very close to other ROH islands. Meat and carcass association, milk association, and production association QTLs were found to overlap with ROH islands from the Australian Poll Merino, Australian Suffolk, Chinese Merino, Lacaune Meat, Lacaune Milk, Merino Landschaf, and Merino de Rambouillet populations. A milk association QTL also overlapped with an HRR island from the Merino de Rambouillet population. These QTLs are reported as being related to traits such as mean corpuscular hemoglobin concentration, pneumonia susceptibility, and fecal egg count (health association); bone area, fat weight in carcass, total fat area, and dressing percentage (meat and carcass association); milk fat yield in 180 days, and curd firming time (milk association); and body weight and total bone weight (production association).4. DiscussionIn this study, we evaluated the use of a BI Software to integrate data obtained from different databases and analyses, regarding ROH and HRR detected in worldwide sheep populations. The use of the BI concept allowed us to dynamically visualize outputs from different analyses, as well as apply filters to efficiently select specific populations, chromosomes, and parameters and focus on the interaction between the studied phenomena. Furthermore, we would like to emphasize that, although the genotypic data used in this study were collected from multiple flocks [32], and sizes of the samples were taken into consideration when selecting the populations to be included herein, any conclusions drawn from the present study should be carefully considered along with other studies that used different data sources and a considerable sample size, in order to avoid any chances of misrepresentation of the populations. Moreover, the visualization method implemented in this study could also be applied to future studies.All sheep populations included in this study presented more than 45% of their detected ROH between 1 and 2 Mb, the shortest ROH length class defined. Many studies also reported the prevalence of ROH in the shortest length category for several sheep breeds [27,44,45,46,47]. It has been reported that modern populations of sheep usually present higher effective population sizes (Ne) and SNP diversity than cattle populations [11,27,32,48], which could be related to the prevalence of short over long ROH in sheep. Moreover, Ferenčaković et al. [17] reported that the use of low-density SNP chips for the detection of ROH may lead to an overestimation of the number of ROH shorter than 4 Mb.Nosrati et al. [48] detected on average 50.38 ROH in individuals from the same Soay population used in the present study, which corresponds to roughly one-quarter of the runs detected herein (188.4). This divergence in the results could be attributed to the differences in the detection parameters, such as higher values of minimal number of SNPs in an ROH (40) and maximal gap between adjacent SNPs (1 Mb), as well as lower SNP density (100 kb/SNP). Our results suggest that setting a low minimal number of SNPs (20) and maximal gap (250 kb), and higher SNP density (70 kb/SNP) when using a low-density SNP chip may lead to the break of runs in regions of lower SNP density, as illustrated in Figure 3, creating an overestimation of the number of runs and an underestimation of the percentage of long runs. On the other hand, Dzomba et al. [29] applied similar parameters as in the present study, with a higher minimum number of SNPs per run (30), a lower density (100 kb/SNP), and used the method Consecutive Runs. The authors reported higher averages of the number of ROH per animal per population (considering the same populations used in the present study). We have also tested the effects of applying a 0.01 MAF filter, which had almost no effects on the overall results and caused the break of some runs. Therefore, we decided not to prune the data for MAF. Besides the Sliding Windows and Consecutive Run approaches implemented by Detect Runs, there are other software and methods that could also be used for the detection of ROH, and might lead to different results.The distribution of ROH in length classes (Figure 1a), chromosomes (Figure 2), and positions (Figure 3) showed an obvious differentiation between populations. ROH has been shown to be non-randomly distributed across the genomes, instead they reflect the occurrence of demographic events and selection pressure for different objectives [4]. The East Friesian Brown and Soay populations showed a similar total ROH length, which leads to similar inbreeding levels. However, the percentage of long ROH was much higher in the East Friesian Brown population, indicating recent inbreeding events. The Soay population was raised in isolation on the Soay Island for hundreds of years [49], and inbreeding was probably frequent when the first individuals arrived on the island, hence the high number of small runs. The three Texel and the two Lacaune populations presented similar averages of total length and number of ROH within each breed, while the two Suffolk and the six Merino populations showed a significant divergence on these metrics (Figure 1), which might indicate that the processes of selection in different countries can be more differentiated for some breeds than for others.Few studies have been conducted with the aim of characterizing HRR in livestock, and only one has attempted to identify factors impacting HRR detection, using a low-density SNP chip [23]. Furthermore, most of the studies on HRRw used high-density SNP chips [21,22,24], which have been shown to require other parameters than low density SNP chips for ROH detection [17,41]. The same is most likely true for the identification of HRRs. In this study, we set the minimum number of SNPs within an HRR at 5 or 10, which is lower than the number used for ROH (15) because HRRs are usually reported as being shorter than ROH [23]. The same difference in the parameters was observed in other studies [20,21,24]. We observed that changing the minimal number of SNPs and window size from 10 to 5 did not increase the number of HRRs detected; in fact, the number and length of HRRs detected decreased. This could be related to the fact that we used the sliding window approach, and the reduction in the window size may have had an interaction with the other parameters, such as number of missing and homozygous SNPs allowed, causing the HRR to break even shorter. We also tested allowing different numbers of homozygous (1 to 3) and missing (1 or 2) SNPs within an HRR. Biscarini et al. [23] reported that allowing only one homozygous SNP reduced the number of detected HRR and increased its average size when compared with allowing two homozygous SNPs, while increasing this number to up to five caused both metrics to increase. We observed a similar effect in our data—when reducing the number of homozygous allowed from three to two, the number of HRR detected was reduced and the length increased, and reducing it to one caused both metrics to decrease.Scenario 2 was chosen as the best scenario for the detection of HRR islands, for presenting a high number of HRRs and a satisfactory maximum HRR length, when compared with the other scenarios. The average number of HRRs detected per animal (139.59) was higher than that detected by other authors in turkey (57.80), cattle (9.87), and horse (52.17) populations [20,23,24], and similar to the number detected by Ferenčaković et al. [22] in a cattle population (122.52). Most of these studies reported the detection of higher numbers of ROH than HRR; however, our results showed the opposite. We hypothesized two reasons: (1) misadjustment of parameters for the detection of HRR, or (2) the sheep genome of the populations analyzed presents small and frequent HRR. Therefore, further research is needed as a means to further test these hypotheses, using different parameters and methods for the detection of HRRs. The use of a higher density SNP chip could also provide further insights.Kijas et al. [32] reported the inbreeding coefficient (F) calculated for each of the populations used in this study, and the populations with the lowest F, such as Chinese Merino (0.08), Australian Suffolk (0.08), and Australian Poll Merino (0.09), presented higher average numbers of HRRs and total HRR length per individual (Table 3), while populations with the highest F, such as Soay (0.33), East Friesian Brown (0.26), and Irish Suffolk (0.22), presented the lowest HRR metrics (Table 3). When comparing the average numbers and total length of ROH (Figure 1) and HRR (Table 3) for the populations, a negative correlation between them was also observed.With the purpose of investigating the occurrence of LD within ROH and HRR islands, we plotted the results from pairwise SNP calculations of r2 against the islands, applying filters on Tableau of minimal LD length and r2 values. This approach was shown to be effective because, differently from other studies, the LD values could be presented directly and not through summarizations such as average r2 per bins of distance (e.g., Mastrangelo et al. [50]). Moreover, when observing LD within the islands, we could identify the minimum amount of LD present and visualize the location of the LD blocks within the islands, instead of calculating the r2 between the first and last SNPs (e.g., Mastrangelo et al. [51] and Purfield et al. [27]), which could overshadow the presence of stronger LD between closer SNPs within the island.Using the approach described above, we observed that most of the regions in ROH islands identified in more than one population (Table 5) were located in regions with some extent of LD (r2 > 0.2), with few exceptions where no LD was detected in some portion of the islands, even when allowing the minimal LD length (0 bp) and r2 (0.2). The presence of stronger LD within the islands varied depending on the chromosome and the population, and some populations showed more overall LD than others. Interestingly, some regions showed a strong LD (r2 > 0.9) in blocks over 250 Mb long across many populations, such as the region around 112 Mb in OAR2, and no islands were identified in such regions. New Zealand Texel presented LD over 0.9 in blocks within the region of 118,497–121,331 kb, and no island within the region (File S1). These findings could indicate poor identification of ROH islands, but also that the presence of strong LD in certain regions does not always result in an increase in homozygosity.The regions detected as ROH islands for two or more sheep populations in the present study spanned across populations selected for different purposes. Abied et al. [52], using data from the OARv4.0 assembly, detected candidate regions on OAR2, OAR6, and OAR10 for five Chinese sheep breeds. Gorssen et al. [53], analyzing 100 populations from the same public database used in this study, identified islands in the same region of OAR6 (around 38 Mb) identified herein, for 15 populations. This region was a common island for four of the six merino populations we analyzed, including the Chinese Merino and the two Lacaune populations (meat and milk). He et al. [46] also identified an ROH hotspot on this region in a Chinese Merino population, and reported the influence of NCAPG/LCORL, genes associated with calving ease and fetal growth in cattle [54,55], body size in mammals [56,57,58], and reduced subcutaneous fat thickness in cattle [58]. A few QTLs within or very close to the region were associated with body weight (7), bone area (2), and milk fat yield. Taken together, these results suggest that this region on OAR6 is important for multiple traits, which could be beneficial for meat, wool, and milk production.The region from 109 Mb to 119 Mb on OAR2 harbored ROH islands from six different populations, including breeds selected for meat, milk, and wool. Moreover, a great number of genes with distinct functions were observed within this region, such as CLCN3, a gene involved in several basic cellular functions, and that was shown to reduce the inflammatory response induced by a high-fat diet in mice [59]; HPF1, associated with early embryonic development in zebrafish [60]; PMS1 and ERCC3, identified as candidate genes in a genomic footprint for dryland stress adaptation in Egyptian fat-tail sheep [61]. Purfield et al. [27] reported the region between 115.48 and 126.34 Mb on OAR2 as the ROH hotspot with the most occurrences and as under putative selection in breeds selected for meat (i.e., Texel), but not for Suffolk. In our study, the Texel and the Suffolk populations did not share common islands, in agreement with Purfield et al. [27], who reported a significant differentiation between these breeds. The QTLs observed within 109 Mb and 119 Mb on OAR2 were mostly related with horn type (21); meat color (1) and texture (1); and health traits, such as fecal egg count, platelet count, mean corpuscular volume, and hemoglobin level. These results also indicate that a variety of traits are impacted by this region, thus harboring ROH islands for different selection groups.We identified three genes (BIN1, MYO7B, and GAS7) in common ROH islands that were associated with terms related to muscle development and enriched in the wool group: Actin Cytoskeleton (GO:0015629) Actin Binding (GO:0003779), Contractile Fiber (GO:0043292), and Motor activity (GO:0003774). BIN1 and MYO7B were detected in a region in OAR2 shared by Chinese Merino, Merino Landschaf, and Scottish Texel. B1N1 is involved in muscle cell differentiation [62]. It was reported by Purfield et al. [27] as a candidate gene in Texel, and by Al Kalaldeh et al. [63] as a candidate gene in a GWAS study for parasite resistance in Australian sheep. GAS7 was identified in a different region, located on OAR11 and shared by Australian Industry Merino and Australian Suffolk. This gene is expressed in the central nervous system and associated with motor activity and muscle fiber composition [64].Furthermore, Australian Industry Merino and Australian Suffolk shared a region on OAR11 where two genes (PIK3R5 and STX8) were previously detected in a putative selection region in Swiss sheep [45], and are associated with body size [65,66]. DHRS7C and NTN1, also detected within this region, were reported as being related to enhanced muscle performance [67] and body size [65,66], respectively. QTLs detected within this region are associated with body height, average daily gain, milk yield, and milk fat yield. According to Safari et al. [68], there are moderate positive correlations between live weight at various ages and wool traits. They suggested that a greater need for both wool and meat products led sheep breeders to combine these two traits, as well as quality and disease resistance, into their breeding objectives. Other authors also endorsed the selection of Merino flocks for meat and carcass traits [69,70] and disease resistance [71]. Therefore, we suggest that the need to improve a variety of traits led breeds with distinct selection purposes to present a higher homozygosity in certain common regions, described herein as well as in other studies, where these distinct traits would be improved.No gene nor QTL were detected within the region shared by Australian Industry Merino, Australian Merino, Chinese Merino, and Tibetan populations in OAR11 (41,526–42,049 kb), which may indicate the need for better annotation of the sheep genome, or that this region contains distal regulatory elements, such as silencers or enhancers. Fewer common genomic regions were identified in HRR islands than in ROH islands. From those, two regions contained identified genes. Australian Merino, Australian Poll Merino, and Chinese Merino shared a region in OAR8 (89,939–90,351 kb), which contains TCTE3, a gene previously described as a candidate influencing congenital diaphragmatic hernia [72] and sperm motility and morphology [73]. Three protein-coding genes (ERMARD, PHF10, and WDR27) detected within this region were previously reported in a study about structural brain abnormalities in humans, and only ERMARD and PHF10 were considered as plausible candidates [74]. Furthermore, it was reported that heterozygous variants in ERMARD (C6orf70) are associated with brain anomalies and syndromic dominant forms of periventricular nodular heterotopia in humans [75,76]. WDR27 was also detected as a candidate for insomnia [77].The other common region in HRR with detected genes was identified on OAR21 (400–926 kb) and was shared by Australian Industry Merino, Australian Suffolk, German Texel, Lacaune (meat), New Zealand Texel, and Scottish Texel. CEP295 and MED17, genes identified within this region, are responsible for building centrioles [78,79] and for the transcriptional activation of lipogenic genes in response to insulin [80], respectively. VSTM5, also identified within this region, codes a protein responsible for the regulation of neuronal morphogenesis and migration during cortical development in the brain [81].A common region in HRR was observed in OAR13 (34,513–34,530 kb) for Merino de Rambouillet and New Zealand Texel. Despite no annotated genes being detected within this region, two QTLs were identified nearby. A QTL for milk fat yield was detected within the region in HRR island exclusive of New Zealand Texel (34,254.2–34,530.07 kb), and a QTL for average daily gain was detected outside, but near the HRR island detected in the Merino de Rambouillet (34,513.4–34,887.99 kb). A QTL for milk fat yield was also identified near an HRR island detected in Australian Suffolk, Churra, and Lacaune (milk) in OAR26 (43,609–44,004 kb).5. ConclusionsIn this study, we detected ROH and HRR islands in worldwide sheep populations. The parameters applied for the identification of ROH resulted in an inflation in the number of short ROH owing to the fragmentation of longer ROH. We also characterized HRRs, which had not yet been reported in sheep, and provided comprehensive knowledge about the effects of changing the parameters for HRR detection using the Sliding Windows approach. Our findings suggest that HRRs in sheep are small and frequent, and further studies using a higher density SNP chip are suggested. Regions in high LD were more closely located from ROH than HRR islands, and many regions in LD were not in ROH. Candidate genes and QTLs identified within common regions in ROH islands for different populations were related to a variety of production traits (e.g., body wight, milk fat yield, and meat color), while genes identified within common HRR islands may play a fundamental role in the survival of these individuals, as many of them are involved in brain integrity. The integration and visualization of genomic data from worldwide sheep populations, after applying filters to highlight the key results from independent analyses, allowed us to better understand structure, distribution, and LD pattern in ROH and HRR regions, as well as to identify candidate genes, QTLs, and related phenotypes.

animals : an open access journal from mdpi

10.3390/ani13060990

PMC10044568

Flow cytometry (FC) is the recommended technique for assessing sperm quality. In comparison with fluorescence microscopy (FM), FC supports analyses of much larger sperm populations and generates more reliable results. However, FC is not always accessible, and sperm populations are often evaluated with the use of FM. In the present study, FC and FM were used to assess the functionality of various organelles in European red deer epididymal spermatozoa stored in liquid nitrogen. Spermatozoa were collected from the epididymides of hunter-harvested European red deer stags. The epididymides were stored in a refrigerator (2–4 °C) for 12 h before analysis. The study demonstrated that the refrigerated storage of the epididymides for 12 h had no significant effect on the sperm quality before cryopreservation, but it significantly influenced the percentage of early necrotic sperm after thawing. The results of FM and FC assays differed significantly, excluding in the assessment of the plasma membrane integrity. However, the results of both assays revealed significant correlations between the examined variables, except for mitochondrial activity. The study demonstrated that the spermatozoa from epididymides chill-stored for 12 h can be used for cryopreservation. Fluorescence microscopy and FC are equally reliable techniques, but FM was more useful for evaluating mitochondrial activity.

Thawed spermatozoa, sampled post mortem from the fresh epididymides of European red deer and epididymides stored for up to 12 h at 2–4 °C, were evaluated by fluorescence microscopy (FM) and flow cytometry (FC). The sperm samples were extended and cryopreserved. The sperm motility (CASA), sperm viability (SYBR+/PI-), acrosome integrity, mitochondrial activity, apoptotic changes, and chromatin stability were assessed. Sperm were analyzed by FM before cryopreservation, and by FM and FC after thawing. Epididymal storage time (for 12 h) had no significant effect (p > 0.05) on the examined variables before cryopreservation. After thawing, the storage variants differed (p ˂ 0.05) in the percentage of apoptotic sperm (FM and FC) and DNA integrity (FC). The results of FM and FC differed (p ˂ 0.05) in all the analyzed parameters, excluding SYBR+/PI. Significant correlations (p ˂ 0.01) were observed between the sperm viability, acrosome integrity, and the percentage of non-apoptotic spermatozoa, regardless of the applied technique. In FM, the above parameters were also significantly correlated with mitochondrial activity. The study demonstrated that European red deer spermatozoa stored in the epididymides at 2–4 °C for 12 h can be used for cryopreservation. Both techniques were equally reliable, but FM was better suited for evaluating mitochondrial activity whereas FC was more useful in the evaluation of DNA fragmentation.

1. IntroductionSperm collected from hunter-harvested wild-living animals is a valuable source of genetic material that can be used for the conservation and genetic improvement of animal populations in breeding farms [1,2,3]. Sperm can be cryopreserved in liquid nitrogen to preserve its fertilizing potential for a long period of time (many years), which has important implications for species conservation [4,5,6].Ejaculated sperm is the most recommended for reproductive purposes, but it is not easy to acquire, and epididymal sperm can also be used [7]. In wild-living animals, post-mortem sperm sampling from the epididymides is the easiest sperm-harvesting method, which, unlike electroejaculation, does not require specialist equipment or complex procedures [8,9]. The epididymal sperm for reproductive purposes can be stored in a liquid state or cryopreserved [4,10,11]. However, epididymal spermatozoa cannot always be collected directly from hunter-harvested animals because the hunting grounds are usually situated in remote locations. As a result, the sperm are harvested several hours after the hunt. The collected material is transported to a laboratory and stored in a refrigerator at a temperature of 0–6 °C and 80–85% humidity. The storage conditions can significantly affect the viability of the epididymal sperm, and further research is needed to address this aspect.Previous studies have shown that the sperm of many animal species, including mice [12], rams [13], dogs [14], and bulls [15,16,17], is resistant to low storage temperatures in the range of 0–6 °C. Iberian red deer spermatozoa can be stored in the epididymides for up to several days at a temperature 4–5 °C [7,18] before cryopreservation [19,20]. These observations suggest that European red deer spermatozoa intended for cryopreservation should also tolerate temperatures in the range of 2–4 °C.Sperm quality is evaluated with the use of various laboratory methods, where fluorescence techniques, motility analyses, and morphology assessments are most recommended. These assays are conducted with the use of a fluorescence microscope (FM) or a flow cytometer (FC). Each technique has its strengths and weaknesses [21]. Fluorescence microscopy is a time-consuming method that should be performed by a qualified and experienced observer, and it involves time limits and a small number of cells [22]. In turn, FC supports accurate and simultaneous assessments of various cell structures [23,24]. Flow cytometry is highly recommended in clinical trials because it supports rapid and objective analyses of large cell samples [22,25]. Flow cytometry is also recommended for fluorescence analyses of sperm quality [23]. However, flow cytometers are expensive and not readily accessible. Most laboratories are equipped with fluorescence microscopes, but only small sperm samples (200–300 cells) are analyzed under an FM. According to many researchers, these limitations undermine the reliability of the results.In view of the above, the aim of this study was to evaluate different structures in cryopreserved epididymal spermatozoa of European red deer, collected directly from fresh epididymides and from epididymides stored for 12 h at a temperature of 2–4 °C, with the use of FM and FC. Sperm motility and motility parameters were determined, and fluorescence analyses of sperm viability, acrosomal membrane integrity, mitochondrial membrane potential, apoptotic changes, and DNA status were performed. Spermatozoa were subjected to FM only before cryopreservation, and to both FM and FC after thawing. The same fluorochromes were used in FM and FC.2. Materials and Methods2.1. AnimalsRed deer stags were hunted in accordance with the harvest plan for each species of game animals, in the in the Forest District of Nowe Ramuki (Region of Warmia and Mazury, Poland), in September and October (during the rutting season), in accordance with the Polish Hunting Law. Experimental material was collected post mortem, and it consisted of the testes and epididymides that were stored in the scrotum until sperm collection. The sperm samples were collected from the epididymides of 16 European red deer stags (aged 4 to 11 years, with body weight of 140 kg to approx. 180 kg).2.2. Sampling and Cryopreservation of SpermatozoaSpermatozoa were collected from the tail of the epididymis (within 2 h post mortem, variant I) and from epididymides refrigerated for 12 h (2–4 °C, variant II). Spermatozoa were sampled according to a previously described procedure [26]. In the collected samples, the sperm motility was assessed using the computer-assisted semen analysis (CASA) system and the sperm concentration was evaluated using a Bürker counting chamber (Equimed-Medical Instruments, Cracow, Poland).The spermatozoa were cryopreserved according to the procedure described by Martínez-Pastor et al. [27], with some modifications. In the first trial, the samples were diluted with a commercial freezing extender (Andromed, Minitub GmbH, Tiefenbach, Germany) to a concentration of 400 × 106 cells/mL and transferred to a refrigerator (4 °C) for 2 h to equalize the temperature. Then, the samples were diluted with the same extender to a concentration of 160–200 × 106 cells/mL and left to stand for 1 h. The samples were then packed in 0.25 mL straws (IMV, L’Aigle Cedex, France) and frozen in nitrogen vapor (4 cm above liquid nitrogen) for 10 min. The straws were transferred to liquid nitrogen, where they were kept for at least 1 year. The samples were thawed by placing the straws in water (65 °C) for 6 s.2.3. Sperm Analysis2.3.1. Sperm MotilitySperm motility was evaluated using the CASA system (Hamilton Thorne Sperm Analyzer IVOS version 12.2l; Hamilton Thorne Biosciences, MA, USA). Sperm samples were diluted 1 : 100 (fresh) and 1 : 5 (fresh diluted sperm, frozen-thawed) in phosphate-buffered saline (PBS) to a concentration of around 30–50 × 106 cells/mL. A 5 μL aliquot of the sample was placed in a pre-warmed Makler counting chamber (Sefi-Medical Instruments Ltd., Haifa, Israel) and evaluated at 37 °C. In each sample, spermatozoa were analyzed in five fields of view selected randomly by the computer. The analyses included the determination of: percentage of total motile sperm (TMOT), percentage of progressive motile sperm (PMOT), and parameters characterizing sperm movement: VAP, VSL, VCL, ALH, BCF, LIN (VSL/VCL ratio × 100 %), and straightness (STR, VSL/VAP ratio × 100 %).The CASA system settings recommended by Hamilton Thorne for gazelle/deer [26] were used in the analyses. Progressive motility was defined as the percentage of spermatozoa with a VAP > 75.0 μm/s and an STR > 80%.2.3.2. Fluorescence AssayThe sperm characteristics were evaluated by the fluorescence method before and after cryopreservation. The Olympus BX41 Fluorescence Microscope, which offers ultraviolet (330–385 nm), blue (460–490 nm), and green (510–550 nm) excitation wavelengths, was used for fluorescence analyses. Stained samples were analyzed at 600 × magnification. A minimum of 300 cells per slide were examined in each aliquot.The viability (plasma membrane integrity) was assessed using SYBR-14 and propidium iodide (PI) fluorescent probes (Live/Dead Sperm Viability Kit; Life Technologies Ltd., Grand Island, NY, USA) according to a previously described method [26]. The diluted sperm samples (200 μL) were supplemented with 2 μL of 1 mM SYBR-14 solution and 2 μL of PI (2.4 μM in Tyrode’s salt solution), and incubated at 37 °C in the dark for 10 min. Sperm with green heads (SYBR-14+/PI−, live sperm with integral membranes) and sperm with red heads (dead sperm) were identified.The acrosomal status of spermatozoa was evaluated using fluorescein isothiocyanate-labeled peanut (Arachis hypogaea) agglutinin (FITC-PNA; Life Technologies Ltd., Grand Island, NY, USA) with a PI solution. The method was described in earlier reports [11,28]. Next, 1 µL of JC-1 (1 mg JC-1/mL anhydrous dimethyl sulfoxide, DMSO) was added to diluted sperm samples (200 μL), and they were incubated at 37 °C for 15 min (to identify all sperm cells under the fluorescence microscope). Then, 2 μL of FITC-PNA solution and 2 μL of PI were added to the samples, which were incubated again for 5 min at 37 °C. The FITC-PNA working solution was prepared by dissolving 2 mg of FITC-PNA in 1 mL of PBS. Four sperm populations were observed under the microscope: non-stained spermatozoa in the head region with fluorescence in the midpiece were classified as live sperm with intact acrosomes (FITC-PNA−/PI−); sperm with green-red fluorescence in the head (FITC-PNA+/PI+) were classified as early necrotic, acrosome-reacted sperm; sperm with a red head were considered as dead; and a small population of spermatozoa had only green staining of the acrosomal cap (acrosome-reacted sperm) (Figure 1A).The mitochondrial activity was evaluated by examining the mitochondrial membrane potential (MMP) of sperm with the use of JC-1 fluorochromes (Life Technologies Ltd., Grand Island, NY, USA) with PI, according to a previously described method [11,28]. Diluted sperm samples (200 μL) were incubated with 1 μL of JC-1 solution for 15 min at 37 °C. Next, 1 μL of PI was added to the samples, which were incubated again for 5 min at 37 °C. The analysis revealed the presence of sperm with orange midpieces (with active mitochondria, high MMP) and sperm with green midpieces or unstained, often with red fluorescence in the head region (low MMP) (Figure 1B).The DNA integrity was assessed using acridine orange, as previously described by Partyka et al. [29] with some modifications. Sperm samples (50 μL) were subjected to acid denaturation (30 s) by adding 200 μL of a lysis buffer (Triton X-100 0.1% (v/v), NaCl 0.15 M, HCl 0.08 M, pH 1.4). Subsequently, 600 µL of acridine orange solution was added to the samples, which were incubated for 3 min in the dark. The fluorescence analysis revealed the presence of sperm with green heads (with integral DNA) and sperm with orange or red heads (with DNA fragmentation; DFI) (Figure 1C,D).The Vybrant Apoptosis Assay Kit #4 (Life Technologies Ltd., Grand Island, NY, USA) was used to assess apoptosis and membrane integrity according to a previously described method [11,28]. First, to visualize all cells under a fluorescence microscope, the diluted samples (200 µL) were stained with 1 µL JC-1 (as described above). Then, 2 μL of YO-PRO-1 and 2 μL of PI were added to the samples, which were incubated at 37 °C for 5 min. Four sperm populations were distinguished in the fluorescence analysis: viable sperm (YO-PRO-1−/PI−), sperm showing apoptotic-like changes (YO-PRO-1+/PI−), sperm with early necrotic changes (YO-PRO-1+/PI+), and dead/necrotic sperm (YO-PRO-1−/PI+), according to a previously described staining model [11].2.3.3. Flow-Cytometry AnalysisFlow-cytometry analyses were performed in a Guava EasyCyte 5 (Merck KGaA, Darmstadt, Germany) cytometer. Fluorescence was induced by an argon ion 488 nm laser. Data were acquired using the GuavaSoft™ 3.1.1 software (Merck KGaA, Darmstadt, Germany). Non-sperm events were gated out based on scatter properties and excluded from analyses. A total of 10,000 events were analyzed for each sample. For the flow-cytometry analysis, the stained sperm samples were extended to a concentration of around 5–10 × 106 spermatozoa/1 mL. The same fluorochromes were used in fluorescence flow-cytometry analyses and in the microscopic analysis.To assess the plasma membrane integrity, spermatozoa were stained with SYBR-14 and propidium iodide (PI) according to the protocol described by Partyka et al. [29]. The diluted samples (500 μL) were stained with 3 μL of SYBR-14 and 3 μL of PI. SYBR-14-positive sperm with green fluorescence and PI-negative sperm were classified as live sperm cells with intact plasma membranes.Sperm acrosome status was assessed with lectin PNA from Arachis hypogaea conjugated with Alexa Fluor® 488 (Life Technologies Ltd., Grand Island, NY, USA). The diluted semen samples were combined with 10 μL of PNA working solution (1 μg/mL) and incubated for 5 min at room temperature in the dark. After incubation, the samples were washed, and 3 μL of PI was added before the fluorescence analysis [29].Sperm mitochondrial activity was determined by JC-1 and PI staining. A 3 mM stock solution of JC-1 in DMSO was prepared. Diluted sperm samples (500 μL) with a concentration of 50 × 106 cells/mL were combined with 0.67 μL of JC-1. Before analysis, the probes were incubated for 20 min, in the dark, at 37 °C [30]. Sperm emitting orange fluorescence were categorized as cells with high mitochondrial potential (HMP), whereas sperm emitting green fluorescence were classified as cells with low mitochondrial potential (LMP).Apoptosis and sperm viability were evaluated with YO-PRO-1 (25 μM solution in DMSO) and PI fluorochromes. Sperm were diluted with PBS to a concentration of 50 × 106 cells/mL. A diluted sample of 1 mL was stained with 1 μL of YO-PRO-1 (with a final concentration of 25 nM) and 1 μL of PI [31]. In the FC analysis, fluorescence was measured with a FL-2 sensor and a 575 nm band-pass filter to detect PI, and with a FL-1 sensor and a 525 nm band-pass filter to detect YO-PRO-1.Acridine orange was used to assess DNA integrity, as described above, with minor changes. The sperm suspension (100 μL) was denatured with a lysing solution (200 μL) for 30 s, and then AO solution (600 μL) was added. The samples were analyzed after 3 min of incubation. Spermatozoa with normal double-stranded DNA displayed green fluorescence and were regarded as the main population. Spermatozoa with increased red fluorescence indicative of DNA fragmentation (DFI) were located to the right of the main population.2.4. Statistical AnalysisThe results were processed with the use of a general linear model (GLM) and ANOVA in the Statistica program (v. 13.3, StatSoft Incorporation, Tulsa, OK, USA). The assumption of normality was checked using the Shapiro–Wilk test, and the homogeneity of variance was assessed with Levene’s test. The data that were not normally distributed were transformed before statistical analysis. The percent data were arcsine transformed. The results were presented as means ± standard error of the mean (SEM). Tukey’s post-hoc test was used to compare the means, and the statistical significance was set at p < 0.05. The relationships between sperm motility and fluorescence parameters were examined by calculating the Pearson correlation coefficient.3. ResultsThe sperm motility analysis revealed that the compared storage variants had no significant effect (p > 0.05) on sperm motility parameters both before cryopreservation and after thawing (Table 1). In both variants, cryopreservation induced a significant (p ˂ 0.05) decrease in the values of TMOT, PMOT, VAP, VSL, VCL, and ALH. No significant changes were found in the values of BCF, STR, and LIN before and after thawing.The FM assay revealed that cryopreservation significantly influenced all examined sperm structures, excluding DNA status, in both storage variants (Figure 2 and Figure 3). The variant where sperm were stored in the epididymides for up to 12 h before cryopreservation did not induce significant changes in sperm cell structures, relative to the variant where sperm were sampled directly from fresh epididymides and cryopreserved.A similar relationship was observed in thawed sperm, excluding the percentage of normal spermatozoa without apoptotic changes (FC), the percentage of apoptotic-like spermatozoa (FM and FC) (Figure 3A,B), and DNA fragmentation (DFI, FC) (Figure 3E,F).In the thawed sperm, the percentage of spermatozoa with intact plasma membranes was the only parameter that was not significantly (p > 0.05) affected by the applied fluorescence technique. The results of FM and FC assays differed significantly (p ˂ 0.05) in the following parameters: the percentage of spermatozoa with high mitochondrial membrane potential (Figure 2C,D), acrosome integrity (Figure 2E,F), apoptotic changes (Figure 3A–C), and chromatin stability (Figure 3E,F). In the analysis of mitochondrial activity, the percentage of spermatozoa with a high mitochondrial membrane potential was significantly higher in FM than in FC, regardless of the storage variant (Figure 2C,D). The percentage of sperm with intact acrosomes was significantly higher in FC than in FM (Figure 2E,F). The percentage of non-apoptotic sperm was also significantly higher in FC than in FM (Figure 3A). However, the percentage of apoptotic-like and early necrotic spermatozoa was higher in FC than in FM, irrespective of the storage variant (Figure 3B,C). In the DNA status analysis, a significantly (p ˂ 0.05) smaller percentage of spermatozoa with fragmented DNA was identified in FC than in FM (Figure 3E,F).The correlations between sperm motility and the analyzed fluorescence parameters in FM and FC are presented in Table 2 and Table 3, respectively.The correlation analysis for FM revealed significant (p ˂ 0.05) correlations between the percentage of motile sperm (TMOT) versus acrosomal membrane integrity, the percentage of non-apoptotic sperm, and the percentage of sperm with high mitochondrial membrane potential (HMP). The parameter HMP was also correlated (p ˂ 0.01) with plasma membrane integrity, acrosome integrity, and the percentage of non-apoptotic sperm. Progressive motility was significantly (p ˂ 0.05) correlated with plasma membrane integrity.The FC assay also revealed a positive correlation (p ˂ 0.05) between TMOT versus acrosome integrity, and between PMOT versus plasma membrane integrity and the percentage of non-apoptotic spermatozoa. Plasma membrane integrity was highly positively correlated (p ˂ 0.001) with acrosome integrity, and positively correlated (p ˂ 0.01) with the percentage of non-apoptotic sperm. In turn, the HMP was not significantly (p > 0.05) correlated with the remaining parameters in the FC assay. No significant correlations were found between the DNA status and the remaining parameters.4. DiscussionThe study demonstrated that the storage of European red deer spermatozoa in the epididymides at a temperature of 2–4 °C for up to 12 h did not significantly decrease sperm viability or functionality in comparison with the variant where spermatozoa were sampled directly from fresh epididymides. Sperm motility, motility parameters, plasma membrane integrity, mitochondrial activity, and apoptotic changes were similar in both storage variants. These results confirm the previous observations made in Iberian red deer, and indicate that red deer spermatozoa are relatively resistant to cold shock [7,18,19]. Interestingly, even small fluctuations in temperature in the range of 2–4 °C did not lead to a significant decrease in the quality of spermatozoa stored in the epididymides for 12 h. Possibly, the storage time was too short to induce significant changes in the sperm function. In other studies, significant changes in the motility and structure of spermatozoa stored in the epididymides at a temperature of 4–5 °C were reported only after 24 h [32] or even 48 h [19,20]. However, in the present study, the epididymides were stored in the scrotum, whereas in the cited studies, the epididymides and the testicles were removed from the scrotum, placed in plastic bags, and stored in beakers filled with water at a temperature of 4–5 °C. The above factor could have played a role in preserving high sperm quality before cryopreservation.Despite considerable advances in freezing and thawing methods, cryopreservation always leads to damage in cell structures, which alters sperm motility, motility parameters, and decreases semen fertilizing ability [33,34]. In the current study, significant changes in all the evaluated parameters were also observed, including the plasma membrane integrity, acrosome integrity, and mitochondrial membrane potential, which decreased the TMOT and, in particular, PMOT values that are influenced by the VAP and VSL values.After thawing, most of the analyzed variables did not differ significantly between storage variants, which could be attributed to similar sperm quality before cryopreservation. The quality of sperm cells before cryopreservation significantly influences their resistance to freezing damage [19]. The significant differences between the tested storage variants were observed only in the percentage of apoptotic spermatozoa (FM and FC) and DNA fragmentation (DFI, FC). This observation suggests that sperm stored in the epididymides at a temperature of 2–4 °C were more susceptible to oxidative damage and apoptotic changes during cryopreservation. In another study, cryopreservation induced apoptosis in bovine spermatozoa [35]. In boar spermatozoa, freezing and thawing stimulated an excessive generation of reactive oxygen species (ROS), exerted a negative effect on the cell membrane, mitochondrial activity, and chromatin stability, and accelerated apoptosis [36].Fluorescent microscopy and FC are widely used to assess sperm structures and semen quality. In the present study, the applied fluorescence technique significantly affected the values of most variables, which could be attributed to numerous factors, including the number of sperm cells. A total of 10,000 sperm cells were evaluated by FC, but only 300 spermatozoa were assessed by FM. It should also be noted that sperm sampled from the tail of the epididymis can be contaminated with blood cells, epithelial cells, or tissue fragments, which can compromise the differentiation of stained spermatozoa by FC [22]. Forward-scatter (FSC) and side-scatter (SCC) gating strategies can be applied to eliminate these contaminants based on differences in the size and complexity of the debris and spermatozoa, but this approach does not always generate reliable results [23]. Flow cytometry is the most recommended technique for the fluorescence assessment of cells, but FM is also an accurate and reliable method, regardless of the origin of sperm cells (ejaculate, epididymides, or testicles) [22].The compared fluorescence techniques produced similar results only with regard to the viability of sperm stained with SYBR-14/PI. Similar observations were made by Merkies et al. [37], where sperm viability evaluated with the same combination of fluorochromes was similar in both FC and FM.The greatest differences in the results of FM and FC were noted in the percentage of sperm with high mitochondrial membrane potential (HMP) and DNA integrity in thawed spermatozoa. JC-1 staining supports the identification of two or three sperm populations with different mitochondrial membrane potential [23]. Sperm with a medium MMP (green-orange fluorescence) were most difficult to identify in both FM and FC, which could have affected the results. In the FM analysis, sperm with a medium MMP were difficult to distinguish from sperm with HMP (orange fluorescence) and were classified as spermatozoa with HMP. In turn, in the FC analysis, these sperm were classified as spermatozoa with low MMP (Figure 2D). These difficulties contributed to significant differences in HMP values between FM and FC, irrespective of the storage variant. According to some authors, JC-1 is the most sensitive and reliable fluorescent probe for assessing MMP in FC [38,39], but Martínez-Pastor et al. [23] observed that this probe can be highly sensitive to the staining conditions and that controls should be established to select flow-cytometer settings.The percentage of sperm with intact acrosomal membranes and the percentage of non-apoptotic sperm were lower in FM than in FC, probably due to differences in the number of the analyzed sperm cells. These differences can be also attributed to a lower detection sensitivity (depending on the observer’s visual processing skills), a smaller number of analyzed cells, and the risk of fluorescence decay during the readout [22]. This problem was particularly evident in the analysis of the percentage of early necrotic sperm, which was significantly lower in FM than in FC. The FM assessment revealed a higher percentage of dead than early necrotic sperm, which influenced the final percentage of non-apoptotic spermatozoa.The percentage of sperm with fragmented DNA was higher in FM than FC, which could be attributed to the difficulties in distinguishing between the green (sperm with normal DNA) and orange fluorescence (sperm with fragmented DNA) in FM. The fluorescence decay in the samples analyzed under a microscope could be another reason. When the samples for DNA analyses are exposed to laser light for a long period of time, sperm cells are probably rapidly degraded. They change color from green to orange, which is indicative of DNA fragmentation. The fluorescence excitation can lead to the formation of highly reactive photochemical products (reactive oxygen species (ROS), heat, and DNA damage), which are influenced by many factors, including excitation wavelength, intensity, and exposure time [40,41,42,43]. Flow-cytometry measurements are conducted rapidly (in less than 1 min), and this technique seems to be more appropriate for evaluating DNA fragmentation than FM. Sample preparation, as well as the time and conditions of denaturation and staining, can also significantly affect DNA integrity [23,44]. In other studies, the results of DNA fragmentation analyses also differed considerably between FM and FC [21,45].Despite the considerable differences in the values of most analyzed variables, the results of FM and FC were subjected to a correlation analysis that revealed similar relationships between the sperm viability, acrosome integrity, the percentage of normal sperm without apoptotic changes, and sperm motility. Similar observations were made in other studies, where spermatozoa were assessed by FM [19,38] and FC [46,47]. Therefore, both techniques can be effectively used to assess sperm cell structures. The results of the correlation analysis indicate that FM was a more reliable method than FC in determining the percentage of sperm with high MMP.The present study revealed that both fluorescence techniques have certain limitations that should be taken into consideration when assessing the various sperm structures.5. ConclusionsThe results of this study indicate that spermatozoa obtained post mortem and stored in the epididymides for up to 12 h (at 2–4 °C) can be used for cryopreservation. Fluorescence microscopy and FC are equally reliable techniques for assessing various sperm structures. However, the FM method proved to be more useful in the assessment of mitochondrial activity, whereas the FC assay was more useful in the evaluation of DNA fragmentation.

animals : an open access journal from mdpi

10.3390/ani12030389

PMC8833707

Samples from an African lion cub in the Greater Kruger National Park area (South Africa), which could not walk, were tested for a gene mutation that causes one type of muscle weakness in domestic cats. The cause of the muscle weakness is believed to be genetic, but our study showed that the mutation that is found in similarly affected domestic cats was not present in the cub. Genetic diseases are more common in inbred animal populations, so this condition needs to be further evaluated to assist in the conservation of these magnificent creatures.

Polyphasic skeletal muscle degeneration, necrosis and mineralization of skeletal muscle was diagnosed in eight juvenile free-ranging lions (Panthera leo), from five different litters in the Greater Kruger National Park area that were unable to walk properly. A detailed investigation was not possible in free-ranging lions, so the cause could not be determined. The cases resembled hypokalemic polymyopathy in domestic cats with muscle weakness. A candidate-gene approach previously identified a nonsense mutation in the gene coding for the enzyme lysine-deficient 4 protein kinase (WNK4) associated with the disease in Burmese and Tonkinese cats. In this study, we sequenced all 19 exons of the gene in one case, and two control samples, to identify possible mutations that may be associated with polymyopathy in free-ranging lions. Here, no mutations were detected in any of the exons sequenced. Our findings indicate that the WNK4 gene is not a major contributor to the condition in these lions. Further studies into the pathogenesis of this condition are needed to inform conservation policies for this vulnerable, iconic African species.

1. IntroductionSevere skeletal muscle weakness resulting in difficulties in standing and walking was observed in eight free-ranging lion (Panthera leo) cubs (<10 months old) from five separate litters in several locations in the Greater Kruger National Park area (KNP), South Africa, between 2015 and 2020. The cub number, date of death, litter, age and sex of the lion cubs are shown in Table 1.Cubs 1–3 and 5 were found dead; the remaining cubs were euthanized. Cubs 7 and 8 had been abandoned by the pride. The cubs were variously described as having paralyzed front legs (Cubs 3 and 4), crawling (Cub 6) and reluctant to walk, finding it difficult to stand (especially on the hind limbs), and creeping along with all four limbs flexed (Cubs 7 and 8). No traumatic lesions were found; the cubs had reduced skeletal muscle mass. Full post-mortem examinations (after euthanasia) of the cases showed multiple pale tan tracts (1–5 cm dia) in many skeletal muscle masses including the semi-membranous, pelvic girdle, intercostal, occipital, masseter and periocular muscles (Figure 1).The histological changes in each case were similar: severe polyphasic skeletal muscle fiber degeneration, necrosis, regeneration and mineralization. Scattered individual or clusters of myofibers were variably swollen, hypereosinophilic and fragmented with variable loss of striations, and karyolysis or large multiple nuclei in rows interpreted as regeneration (Figure 2). Satellite cell hypertrophy and a few myofibers contained deeply basophilic particulate material interpreted as dystrophic mineralization.The cause of the clinical signs and lesions could not be determined. Detailed clinical examination and serum biochemistry were not possible in these cases because the animals were found dead or, in the case of live animals, park policy is to only examine and treat human-made injuries such as those caused by snares or vehicular accidents. In addition, the cubs died in locations over 400 km from a veterinary laboratory.Polymyopathy may have many causes in domestic cats [1]. Muscle fibers were not of varying sizes, myofibers were not hyperplastic and did not show splitting and endomysial fibrosis were absent. Therefore, muscular dystrophy caused by alpha-dystroglycan, beta-sarcoglycan and laminin deficiencies and congenital myotonia were thought to be unlikely, but could not be ruled out. Three cubs were female, ruling out X-linked muscle dystrophy. No evidence of glycogen storage disease was present in any tissues. Frozen muscle sections were not available to check for congenital nemaline myopathy, although the myofiber atrophy seen in this condition was not present. Lesions compatible with toxoplasmosis were not seen in the skeletal muscle, or in any other organ. Vitamin E and/or selenium deficiency are uncommon in carnivores and were also thought to be unlikely in free-ranging lions. Toxic myopathies were thought to have rather presented as outbreaks in lions of varying ages, whereas in this series only cubs were affected.Skeletal muscle weakness has also been observed in two South-African-born six-month-old white lion cubs housed in a North American zoo [2]; in these animals, extensive clinical diagnostic investigation suggested that the clinical syndrome was similar to hypokalemic polymyopathy in domestic cats. This possibility was explored in this study.Hypokalemic periodic paralysis in humans is a rare disorder that was first described by Musgrave in 1972. The disorder is usually diagnosed in late childhood or in the teenage years [3]. The condition results in periodic extreme muscle weakness, which results in an inability to move muscles in both the arms and legs lasting from hours to days. Triggers include a cold environment, stress, fasting, rest after exercise, a viral illness, certain medications and/or pregnancy. The severity of attacks varies between individuals and can occur daily, weekly, monthly or only rarely [4]. Both inherited and acquired causes of hypokalemic periodic paralysis have been identified. Mutations identified in the Calcium Voltage-Gated Channel Subunit Alpha1 S (CACNA1S), Sodium Voltage-Gated Channel Alpha Subunit 4 (SCN4A) and Potassium Inwardly Rectifying Channel Subfamily J Member 2 (KCNJ2) that encode the voltage-gated channels in muscle membranes that generate membrane potentials has been found associated with the disorder in humans [5]. Furthermore, missense mutations in Ryanodine receptor type 1 (RYR1), which promotes release of calcium within myofibrils, were recently identified in patients with periodic paralysis [6]. In addition, acquired hypokalemic periodic paralysis has been associated with thyrotoxicosis (high-circulating blood thyroid hormone levels).In animals, hypokalemia resulting in skeletal muscle weakness was first reported in domestic cats (Felis catus) by Eger et al. [7]. Since then, it has been reported in several countries and has been shown to be due to an autosomal recessive trait that affects young Burmese and Tonkinese cats. Affected kittens display varying degrees of weakness in the neck, thoracic limb girdle and appendicular muscles [8]. Thus far, the disease has been linked to a single nonsense mutation, producing a premature stop codon in the serine-threonine kinase gene (‘with no kinase 4′, WNK4) that codes for lysine-deficient 4 protein kinase [9]. The enzyme is expressed mainly in the kidney, while the pancreas, bile duct, brain, epididymis and skin show lower enzyme levels [9,10,11]. WNK4 is involved in the regulation of complex pathways in the renal distal convoluted tubule, that in turn regulate the activity of all major sodium and potassium transporters [12]. Thus, altered function in WNK4 causes loss of potassium in the urine in kittens, which in turn results in symptomatic hypokalemia. Potassium is necessary for the muscle contraction, so hypokalemia results in muscle weakness.This project sought to investigate if the 2899C<T mutation of the WNK4 gene is present in African lions, to determine whether or not the polymyopathy syndrome in free-ranging lions could be due to hypokalemic polymyopathy.2. Materials and MethodsThis study included one case sample from an affected free-ranging lion (Cub 6) based on necropsy and histological examination, and two unaffected lion samples stored in the SANBI Biobank as controls. Genomic DNA was extracted from tissue samples using the Quick-DNA Miniprep Plus Kit (Zymo Research, Irvine, CA, USA) following the manufacturer’s protocol. In order to determine concentration and purity of the extracted DNA, a Nanodrop ND-1000 Spectrophotometer (Inqaba Biotech, Pretoria, South Africa) was used. Amplification was conducted using fourteen published WNK4 primer sets developed by Gandolfi et al. [8]. The primers used the published human WNK4 exons (GenBank number: NM_032387.4) to identify cat exonic sequences in the available cat genome assembly. Primers developed from domestic cat are often used to genotype other feline species, including lions [13,14]. Thus, this published primer set was used in this study to amplify all 19 exon regions of the WNK4 gene in lion. All Polymerase Chain Reaction (PCR) cycling was conducted in a MiniAmp Plus (Thermo Fischer Scientific, Carlsbad, CA, USA) thermocycler in a final reaction volume of 15 µL containing 6.25 µL Ampliqon Taq DNA Polymerase Master Mix RED (Ampliqon, Odense, Denmark); 0.5 µL of the forward and reverse primers (Thermo Fischer Scientific, Carlsbad, CA, USA); 5.25 µL double distilled water (ddH2O) and 2.5 µL of the template DNA. The PCR amplification conditions were as follows: one cycle at 95 °C for 5 min followed by 35 cycles at 95 °C for 30 s, 55–62 °C for 30 s and 72 °C for 30 s; a final extension step of one cycle at 72 °C for 10 min. The PCR product was then run on a 2% agarose gel to ensure that amplification was successful and then purified by adding 0.25 µL Exonuclease 1 and 1 µL FastAP enzymes (Thermo Fisher Scientific, Carlsbad, CA, USA) to the PCR product. The purification reaction was run for one cycle at 37 °C for 15 min followed by one cycle at 85 °C for 15 min in a MiniAmp Plus (Thermo Fischer Scientific, Carlsbad, CA, USA). Cycle sequencing reactions were completed using the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Carlsbad, CA, USA) using the Sanger chain termination method. Sequencing was conducted in a final reaction volume of 10 µL containing 0.7 µL BigDye, 2.55 µL of BigDye Terminator v3.1.5 × Sequencing buffer, 0.75 µL double distilled water (ddH2O), 1 µL of primer and 5 µL of the PCR product. Sequencing reactions were completed for both the forward and reverse direction with the cycling conditions as follows: one cycle at 94 °C for 2 min; 40 cycles at 85 °C for 10 s, 53 °C for 10 s and 60 °C for 2 min 30 s. The cycle sequencing product was purified using the BigDye XTerminator Purification Kit (Thermo Fisher Scientific, Carlsbad, CA, USA) and sequences were visualized using the 3500 Genetic Analyzer (Thermo Fisher Scientific, Carlsbad, CA, USA). Forward and reverse were aligned to create a consensus sequence and all sequences were manually trimmed in BioEdit [15]. Additionally, sequences were aligned in MEGA7 [16]. Lastly, sequences were visually inspected checked for ambiguous peaks in BioEdit [15]. All sequence reads were obtained using standard Sanger sequencing methods and resulting sequence traces were submitted to nucleotide blast and were blasted against the Coding Sequence (CDS) of WNK4 from Burmese cats (GenBank accession number: JQ522971), in order to ensure that the correct region of WNK4 was obtained. Accession details for DNA sequences used in the study have been submitted to Genbank (GenBank accession numbers: OK513187 and OK513188).3. ResultsA mutation (c.2899C>T) is reported to cause a premature stop codon (CAG > TAG) in Burmese cats with hypokalemic polymyopathy (Figure 3A). WNK4 sequences were successfully amplified for all primers sets tested, indicating that this gene is highly conserved. To examine a possible association between alterations in the WKN4 gene and susceptibility to polymyopathy syndrome in a free-ranging lion cub, we sequenced all 19 exons in three samples (one case and two controls). Alignment of sequences in MEGA7 did not identify any nucleotide differences between the case or control samples. In addition, visual inspection did not detect any ambiguous peaks. Thus, genotype analysis demonstrated that the c.2899C>T mutation was not present in either the case or control samples (Figure 3B). Further, mutations were not found in any of the exons sequenced in the coding regions of the WKN4 gene in either case or control samples.4. DiscussionThe diagnosis of the cause of polymyopathy in these lion cubs remains uncertain, given the limited investigation that was possible in free-ranging lion cubs. The sporadic presence of the syndrome and absence of clinical signs in adult animals is strongly suggestive of a genetic cause with recessive inheritance and/or incomplete penetration. It is possible that the polymyopathy syndrome is multi-factorial and contributing factors may include both environmental and genetic factors as described in humans with hypokalemic periodic paralysis [4].We found no evidence of mutations in the WNK4 gene, but cannot rule out the possibility of altered WNK4 gene expression. In South Hampshire sheep with neuronal ceroid lipofuscinoses (Batten disease), a causative mutation in the Ceroid Lipofuscinosis Neuronal 6 (CLN6) gene was absent; however, quantitative real-time PCR revealed that CLN6 mRNA expression was one-third lower in affected South Hampshire sheep compared to control sheep [17]. The authors attributed this finding to an unidentified mutation in or near the CLN6 gene resulting in lower mRNA expression. Thus, although the WNK4 gene mutation identified in domestic cats was absent in the lions, other alterations in or near the gene may result in downregulation of the transcription of the gene or mRNA stability resulting in hypokalemia and polymyopathy.An alternative hypothesis is that mutations in other genes may be involved in the pathogenesis of the polymyopathy syndrome. Mutations in the CACNA1S, RYR1, SCN4A gene have been reported to be associated with hypokalemic periodic paralysis (HOKPP) in humans [5]. In addition, mutations in KCNJ2 have been found associated with non-familial HOKPP, and with thyrotoxic periodic paralysis in humans [18]. Although polymorphisms in these genes have not been found to be associated with hypokalemia in domestic cats [19], their involvement in this polymyopathy syndrome in free-ranging lions cannot be currently be ruled out.This study only examined one animal with polymyopathy. A genome-wide association study with a large number of case and control samples may identify further possible genes associated with the disorder in the future. This technique refers to the screening of markers of genetic variation known as single-nucleotide polymorphisms (SNPs) distributed throughout the genome in cases and controls to identify susceptibility regions for complex diseases. This method has been used in the analysis of several human diseases and is currently available for several domestic species such as cattle (Bos taurus), dogs (Canis lupus familiaris), sheep (Ovis sp.), pigs (Sus sp.), horses (Equus caballus) and chickens (Gallus gallus domesticus). Although this is a recently established method in domestic animals, it has been successfully used to identify disease-causing genes as well as genetic mechanisms of quantitative traits [20]. The disadvantages of this method are that it requires a reference genome sequence where SNPs have been mapped on the genome; many cases and controls must be tested and that it only detects a region that is associated with disease. Further analysis would be needed to identify the exact SNP involved in the disease. Due to these limitations, little to no studies have used this method in wildlife species.Additional methods to identify variants associated with inherited diseases in non-model organisms include whole-genome sequencing (WGS) or whole-exome sequencing. These technologies are becoming cheaper, easier and can provide large volumes of data [21]. Whole-genome sequencing refers to a comprehensive analysis of the entire genome and has been used to identify gene mutations associated with early-onset progressive retinal atrophy in African black-footed cats (Felis nigripes) [22]. Whole-exome sequencing, on the other hand, is more cost effective than WGS and involves sequencing the complete set of protein-coding regions (exons). This method has identified deletions in the sarcoglycan genes in Boston terriers that were found to be associated with limb-girdle muscular dystrophy [23].Interactions between genetic and environmental effects can be complex and may result in variation in expression of the disease phenotype. In addition, the disease may be polygenic, thus an as-yet-unidentified gene or combination of genes may be responsible for polymyopathy in lions. Novel technologies such as next-generation sequencing should make it possible to perform case/control studies that could eventually shed more light on the etiology of polymyopathy. However, sample sizes would need to be increased and pedigree analysis conducted, which present significant challenges in free-ranging lion populations. These studies may be easier to perform in captive lions.Low genetic diversity leading to inbreeding is associated with higher mortality, lower fecundity, developmental instability, more frequent developmental defects and greater susceptibility to disease [24]. For example, Burmese and Tonkinese cat breeds, which are particularly susceptible to hypokalemic polymyopathy, have low heterozygosity and high inbreeding coefficients [25]. Throughout their range, lion populations have declined by more than 42% over the past 21 years, coinciding with the rise of a European colonial presence [26], which has led to increased rates of inbreeding in southern African lions [27]. Inbreeding is further exacerbated in lion populations in South Africa that are confined to small, enclosed reserves [28]. Although reduced genetic diversity may be present in southern African lions, the WNK4 gene appears not to be affected (at least in this small number of cases).5. ConclusionsThese cases highlight the inherent difficulties in investigating and documenting disease in free-ranging felids. No mutations were found in the coding regions of the WNK4 gene in a free-ranging lion cub from KNP with polymyopathy. Further investigation is needed to determine serum potassium and creatinine kinase levels in affected cubs and whether or not the condition in fact has a genetic basis. Although, non-genetic nutritional, toxic and metabolic causes of polymyopathy are unlikely in free-ranging animals, and typically affect animals of varying ages, such causes need to be additionally considered. This study therefore supports further investigation into polymyopathy in free-ranging African lion populations. The worldwide prevalence of this disorder should be determined, particularly in captive lions, which can be more easily evaluated than their free-ranging counterparts.

animals : an open access journal from mdpi

10.3390/ani11071945

PMC8300365

Many viruses, including human immunodeficiency virus 1, influenza virus, or Rift Valley fever virus, cause cell damage by generating reactive oxygen species and altering redox homeostasis. However, cells have developed various antioxidant mechanisms. We assumed that small ruminant lentivirus (SRLV), which has been found to infect sheep and goats worldwide can also disrupt the homeostasis of animals. SRLV target organs are the joints, lungs, brain, and the udder. To our knowledge, no information exists on the influence of SRLV infection on the oxidative processes occurring in goats. Understanding the influence of viral infection on oxidative stress may help develop novel antiviral treatments. Our study aimed to examine the effects of SRLV infection on oxidative stress biomarkers in the serum of dairy goats during lactation. No differences in any studied parameter at any stage of lactation were found between infected and uninfected goats. On the other hand, significant differences in almost all investigated parameters were found between stages of lactation, regardless of the infection status of goats. In conclusion, asymptomatic SRLV-infected goats do not reveal any apparent dysfunctions in serum oxidative stress biomarkers compared to their uninfected counterparts. The only changes in oxidative stress biomarkers observed during lactation appear to reflect the metabolic effort associated with milk production and developing pregnancy.

The present study examines the effects of natural infection by small ruminant lentivirus (SRLV) in the two most common goat breeds in Poland, i.e., Polish white improved and polish fawn improved. It focuses on biomarkers of oxidative stress, oxidatively modified proteins and antioxidant defenses, ceruloplasmin level as an acute phase protein, and the activities of antioxidant enzymes in the goat serum. It was conducted on 24 goats divided into two equal groups: one SRLV-seropositive (SRLV-SP) and another SRLV-seronegative (SRLV-SN). Both groups were identical in terms of breed and parity. Despite infection, the SRLV-SP goats demonstrated no symptoms of caprine arthritis-encephalitis. In addition, the SRLV-SP goats did not reveal pronounced dysfunctions in oxidative stress biomarkers in the serum compared to the SRLV-SN animals. However, both groups demonstrated elevated levels of the aldehydic and ketonic derivatives of oxidatively modified proteins during the lactation period. In addition, both groups retained a high total antioxidant capacity in serum despite the decrease of enzyme antioxidant defenses throughout the 200-day lactation period.

1. IntroductionThe optimal level of reactive oxygen species (ROS) in the organism is controlled by the cellular antioxidant protection (AOP) system, comprising enzymatic and non-enzymatic elements. When the AOP system is insufficient, the organism is subjected to increasing levels of oxidative stress, resulting in a cascade of pathological processes, with extremely negative consequences for the organism. Viral infections are also associated with ROS generation and thus increased oxidative stress [1].Many viruses have been demonstrated to cause cell damage by generating ROS and altering redox homeostasis [1,2,3]. Isaguliants et al. [2] reported a direct correlation between the ability of human immunodeficiency virus 1 (HIV-1) proteins, such as Tat and reverse transcriptase, to induce oxidative stress and their immunogenicity. In addition, infection by virulent influenza virus (IV) is characterized by heavy cellular infiltration and severe lung pathology, and both conditions are accompanied by oxidative stress and matrix metallopeptidase 9 (MMP-9) production [1]. Flaviviridae virus infections are also known to cause oxidative stress, affecting both the life cycle of the virus and the cellular metabolism. Furthermore, antiviral inflammatory signaling pathways are activated following infection [4]. Rift Valley fever virus (RVFV) infection leads to an increase in ROS production in liver cells due to the presence of the viral protein NSm (small cytosolic protein of RVFV, a major virulence factor in the mammalian host) in the mitochondria. Interestingly, the associated increase in cytokine and pro-apoptotic gene expression caused by infection was reversed with antioxidant treatment [5]. In addition, infection with dengue virus (DENV) leads to the accumulation of NADPH oxidase (NOX)-dependent intracellular ROS [6]. Conversely, a reduction of ROS level by chemical or genetic inhibition of the NOX complex weakens the innate immune responses to DENV infection and facilitates viral replication. ROS were also found to be essential in driving mitochondrial apoptosis in infected dendritic cells. ROS appear to play a critical role in stimulating the innate immune response to the virus and promoting apoptosis of human cells infected with DENV. Simultaneously, antioxidant pathways that are regulated by nuclear factor erythroid 2-related factor 2 (Nrf2) were activated to maintain redox homeostasis.Because viruses cause an imbalance in the cellular redox environment, it can result in various responses depending on the virus and the cell type, including cell signaling, antioxidant defenses and reactive species generation. For example, ROS produced during cellular metabolism play an important role as signaling messengers, and can also stimulate inflammatory signaling pathways via protein kinases, transcription factors, and increased genomic expression of proinflammatory factors [7]. To counteract the oxidative effects deriving from the high chemical reactivity of ROS, and to maintain redox homeostasis, cells have evolved antioxidant mechanisms [8]. However, the presence of ROS is needed for the activation of cells providing antimicrobial immunity, i.e., neutrophils and macrophages, and the production of proinflammatory cytokines [7,8].Although varying in their ability to induce ROS, viruses employ a common pathogenic pathway to defend against oxidative stress [4]. One such universal strategy employed by viruses to trigger antioxidant responses is the modulation of Nrf2 signaling [9]. The cellular antioxidant defense system acts under both physiological and pathological conditions to protect against the harmful actions of ROS and maintain cellular homeostasis. One component of the defense system comprises endogenously produced enzymatic antioxidants, the major ones being superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx) [8].One virus that appears to disturb the homeostasis of goats is small ruminant lentivirus (SRLV), which has been found to infect many populations of sheep and goats worldwide [10,11,12]. Its wide prevalence makes it one of the most significant causes of decreased dairy goat production among small ruminants [10]. However, contradictory results have been obtained regarding milk yield and composition among infected goats [13,14,15,16]. The presence of the virus was first confirmed in the Polish population in 1996. This may have been due to the extensive import of Saanen and Alpine goats following 1990 [17]. The SRLV form a genetically diverse group, comprising five main genotype groups (A-E) and many more subtypes. Being in the Retroviridae family, SRLV integrates into the host DNA during replication. The virus causes caprine arthritis-encephalitis (CAE) and maedi-visna disease (MVD) in sheep [18]. Interspecies transmission has been confirmed [19]. Direct evidence of mixed infections with SRLV from four genetic groups has been found in both sheep and goats [20]. In the first stage of infection, SRLV infiltrates the dendritic cells of the respiratory tract, mucous membranes, or intestine. These cells migrate to the lymph nodes, where, in turn, monocytes are infected. The viral mRNA is transcribed into DNA to form a provirus, which is integrated into the host genome. Following this, the infected cells leave the lymph node and can cause systemic infection. Infection of bone marrow cells, myeloid stem cells, or myeloid precursors leads to persistent infection [21,22]. The target organs of infected monocytes are the joints, lungs [23], and the udder. In addition, in kids, the central nervous system can also be infected; however, this is extremely rare [10,24].Information on the immune processes occurring in organisms infected with SRLV is still limited; however, some studies indicate that both innate and acquired immunities are involved in the host response [12,25,26]. However, to our knowledge, no information exists on the influence of SRLV infection on the oxidative processes taking place in goats. Understanding the influence of viral infection on oxidative stress in different stages of lactation may help to develop new antiviral treatments.Under both physiological and pathological conditions, the cellular antioxidant defense system acts to protect against the harmful actions of ROS and maintain cellular homeostasis. One component of the defense system comprises endogenously produced enzymatic antioxidants, the major ones being superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx), as well as glutathione reductase (GR), and various proteins with pronounced antioxidative properties, such as ceruloplasmin [8]. The level of oxidative stress can be determined according to the level of lipoperoxidation based on the use of 2-thiobarbituric acid reactive substances (TBARS) as a marker.Therefore, the aim of our study was to examine the effects of natural SRLV infection on selected biomarkers of oxidative stress, including TBARS, oxidatively modified proteins (protein carbonyl derivatives of amino acids, aldehyde derivatives, AD and ketone derivatives, KD) and antioxidant defenses (total antioxidant capacity (TAC), ceruloplasmin (Cp) level). The goats used in the study belong to the two most common breeds in Poland: Polish white improved (PWI) and Polish fawn improved (PFI).2. Materials and Methods2.1. Animals and Experimental DesignAll procedures involving animals were performed according to the Guiding Principles for the Care and Use of Research Animals and were approved by the III Local Ethics Commission (Warsaw University of Life Science; Permission No. 31/2013).The study was carried out throughout lactation in 24 dairy goats, selected from a herd of 50 does (Central Poland), belonging to two goat breeds: Polish white improved (PWI) and Polish fawn improved (PFI). The goats were fed in groups according to a system developed by the Institute National de la Recherche Agronomique (INRA) of France and adopted by the Research Institute of Animal Production (IZ PIB), Poland [27]. The basal diet consisted of maize silage and concentrates. Fresh hay was administrated ad libitum each afternoon. Water and saltlicks were also available ad libitum. The feed samples were analyzed for the content of basic components using standard methods [28]. The data on the ingredients and nutrient composition of the basal diet are shown by [29].All goats in this herd were regularly tested serologically for the presence of antibodies against SRLV using a commercial ELISA test (ID Screen MVV/CAEV indirect screening test, IDVet Innovative Diagnostics, Grabels, France). Testing was performed twice a year, in June and December, for 20 years [30]. This testing was continued during the study period to eliminate any newly infected goats from the experimental group. The presence of the virus was also confirmed by isolation [20]. RT-qPCR analysis, performed according to Brinkhof et al. [31] in blood, found that the virus load was below detection levels, despite the presence of antibodies (Bagnicka—personal communication). This could explain the lack of observed symptoms of CAE in the infected goats.The animals remained under the constant care of a veterinarian, who assessed the possible occurrence of infection and clinical signs of any other diseases. In addition, each goat was also clinically examined by certified specialists throughout the study (Diplomates of the European College of Small Ruminant Health Management—co-authors JK and MC). One group of animals used in the study were SRLV-seronegative (SRLV-SN, n = 12), i.e., free of SRLV antibodies in the blood, while the other group was SRLV-seropositive (SRLV-SP, n = 12) at least for two years before the start of observation. However, as it was mentioned above, all goats were asymptomatic, without any clinical sign of arthritis. All goats were between their third and eight lactations, with completed somatic development, to avoid any additive influences on homeostasis. The SRLV-SP and SRLY-SN groups were equally represented in terms of parity (3rd, 4th, and >4th), and breed (PWI vs. PFI). However, the only difference between these two breeds identified by transcriptomic analysis (~50 K) was found for the Capra hircus agouti signaling protein (ASIP), which is responsible for the coat color [31]. Moreover, in a study of the entire Polish active goat population, [32,33] did not report any differences between those two breeds in milk yield, fat, protein, or lactose content; however, they did indicate a higher somatic cell count in the milk of the PFI than the PWI goats. The animals were maintained in pens for 12 goats, and they remained in the herd at the end of the experiment and were not euthanized.2.2. SamplesBlood samples were taken collected six times during lactation (the first sampling several hours after parturition) from the jugular vein into sterile 9 mL S-Monovette tubes with clot activator (Sarstedt, Nümbrecht, Germany). All sampling procedures were performed by a veterinarian. After centrifugation, the serum samples were removed and frozen at -20 °C and stored until analysis.2.3. Biochemical AssaysTBARS assay. The level of lipid peroxidation was determined by quantifying the concentration of TBARS according to the method for determining the malondialdehyde (MDA) concentration [34]. This method is based on the reaction with thiobarbituric acid (TBA) in an acidic pH at 90–100 °C. In the TBA test reaction, MDA or MDA-like substances (produced during lipid peroxidation) and TBA react, with the production of a pink pigment with a 532 nm absorption maximum. Protein carbonyl derivative assay. The level of oxidative modified proteins (OMPs) was evaluated by the reaction of protein carbonyl derivatives (ketone-2,4-dinitro-phenylhydrazone) in serum with 2,4-dinitro-phenylhydrazine (DNFH). The rate of protein oxidative destruction was estimated as described by Zaitseva and Shandrenko [33]. DNFH was used to determine the carbonyl content in soluble and insoluble proteins. Carbonyl groups were determined spectrophotometrically based on the difference in absorbance at 370 (aldehydic derivatives, OMP370) and 430 nm (ketonic derivatives, OMP430).Assay of superoxide dismutase activity. Superoxide dismutase (SOD) activity was assessed in an alkaline medium (pH 10.0) by its ability to dismutate superoxide produced during quercetin auto-oxidation as described by Kostiuk et al. [34]. The activity of SOD was expressed in units per mL of serum.Catalase activity assay. Catalase (CAT) activity was determined as described by Koroliuk et al. [35]. Briefly, CAT activity was evaluated spectrophotometrically by measuring the decrease of H2O2 in the reaction mixture at the wavelength of 410 nm. One unit of CAT activity is defined as the amount of enzyme required for decomposition of 1 mmol H2O2 per min per L of serum.Measurement of glutathione reductase activity. Glutathione reductase (GR) activity in the serum of goats was analyzed according to Glatzle et al. [36]. The GR activity was assayed spectrophotometrically by measuring NADPH2 reduction and was expressed as μmol NADPH2 per min per mL of serum.Assay of glutathione peroxidase activity. Glutathione peroxidase (GPx) activity was determined according to Moin [37] by detecting the non-enzymatic utilization of GSH as reacting substrate. GPx activity was measured at an absorbance of 412 nm after incubation with 5,5-dithiobis-2-nitrobenzoic acid (DTNB). GPx activity was expressed as μmol GSH per min per mL of goat serum.Ceruloplasmin level assay. The ceruloplasmin (Cp) level in the serum was assayed spectrophotometrically as described by Ravin [38] in 0.4 M sodium acetate buffer (pH 5.5), and 0.5% p-phenylenediamine at 540 nm. Ceruloplasmin was expressed in mg per L of serum.Measurement of total antioxidant capacity (TAC). The TAC level in the serum was estimated spectrophotometrically according to Galaktionova et al. [39]. The TAC level was determined by measuring the TBARS level after ferrum/ascorbate induced by Tween 80 oxidation at 532 nm. The level of TAC in the sample (%) was calculated against the absorbance of the blank sample.2.4. Statistical AnalysisResults are expressed as mean ± SEM (standard error). All variables were tested for a normal distribution using the Kolmogorov-Smirnov and Lilliefors tests (p > 0.05). The homogeneity of variance was checked using Levene’s test. The significance of the differences between the serum biomarkers of oxidative stress from the SRLV-SN and SRLV-SP goats was established by analysis of variance, with the model including the stage of lactation (1 to 6), the health state of the goats (SRLV-SN vs. SRLV-SP), breed (PWI vs. PFI), and number of lactation (3rd, 4th, or more than 4th). The ANOVA Friedman test and Kendall’s coefficient of concordance were also conducted with STATISTICA 13.3 software (Tibco Software Inc., Palo Alto, CA, USA). The significance of any differences between the SRLV-SN and SRLV-SP groups regarding the levels of lipid peroxidation, carbonyl derivatives, and antioxidant enzyme activities was examined using Student’s t-test. Differences were considered significant at p < 0.05 or p < 0.01 [40].3. Results3.1. Oxidative StressNo differences were found between SRLV-SN and SRLV-SN goats for any studied parameter at any studied stage of lactation (Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8 and Figure 9); however, differences were found between stages of lactation for almost all studied parameters. In addition, the direction of the changes was the same in both the SRLV-SN and SRLV-SP groups (Table S1).The level of TBARS, i.e., the by-product of free radical-induced lipid peroxidation, in the serum of SRLV-SP (p < 0.000) and SRLV-SN (p = 0.000) goats during lactation (on days 1, 7, 30, 60, 140, and 200) is presented in Figure 1. In both groups, the highest level of TBARS on was observed on day 7 after delivery, with a lower level observed on day 1 and the lowest on day 140. No differences between groups were observed in any stage of lactation.No differences in the levels of either OMP, i.e., aldehydic and ketonic derivatives, were observed between the two groups (p > 0.05). Nevertheless, OMP levels were influenced by stage of lactation (p < 0.01) (Table S1), namely their levels were found to increase over the course of lactation.The highest levels of the ketonic derivates were observed on day 60 (only RLV-SP group) and day 200 of lactation, while aldehydic derivatives peaked on day 140 (only SRLV-SN group) (p < 0.05), with similar tendencies observed in both groups (Figure 2 and Figure 3). The stage of lactation also influenced TAC (Table S1). A significant decrease in TAC was observed at the end of lactation compared to day 1 for both groups (p < 0.05) (Figure 4).Significant differences in serum Cp concentration were observed between the stages of lactation in both the SRLV-SN (p = 0.000) and SRLV-SP goats (p = 0.023) (Table S1). The changes occurred in parallel for both groups. Cp concentration was the lowest on day 30 after parturition compared to day 1 (Figure 5). 3.2. Antioxidant Enzyme ActivitySerum SOD was influenced by the stage of lactation in both SRLV-SN (p = 0.000) and SRLV-SP goats (p = 0.004) (Table S1). In both groups, SOD activity was lowest on day 60 (peak of the lactation). This level then increased on day 140 and was fully restored on day 200 compared to day 1 (Figure 6).Both groups demonstrated similar changes in CAT activity during lactation (SRLV-SN p = 0.010; SRLV-SP p = 0.015) (Table S1). The highest activity was observed until day 7 after parturition. CAT then underwent only small changes between early lactation (day 30) and peak (day 60), full lactation (day 140), and the end (day 200) of lactation as compare to the first day (Figure 7).Serum GR activity remained stable in both SRLV-SN (p = 0.252) and SRLV-SP goats during lactation (p = 0.227) (Table S1, Figure 8).The serum GPx activity was also influenced by the stage of lactation (Table S1). In the SRLV-SN goats, the activity was lowest on day 60 of lactation. This value then increased and remained stable until day 200 (p < 0.05). In the SRLV-SP goats, activity was high at the first stages of lactation, peaking on day 30 after parturition (p < 0.05), then falling almost 2-fold by day 60 and remaining stable until day 200 (Figure 9); however, no differences were found between groups.4. DiscussionThe SRLV-SP and SRLY-SN groups were equally represented in terms of parity (3rd, 4th, and >4th) and breed (PWI vs. PFI). However, in our earlier studies, the only difference between these two breeds identified by transcriptomic analysis (~50 K) was found for the Capra hircus agouti signaling protein (ASIP), which is responsible for the coat color [41]. Moreover, in a study of the entire Polish active goat population, Bagnicka et al. [42,43] did not report any differences between those two breeds in milk yield, fat, protein, or lactose content; however, they did indicate a higher somatic cell count in the milk of the PFI than the PWI goats.ROS and reactive nitrogen species (RNS) are byproducts of aerobic metabolism in various cell compartments, including the mitochondria and ER. They play key roles in the maintenance of redox homeostasis during both normal physiological functions, and in numerous pathophysiological states [4,44]. ROS generation can cause various cellular consequences depending on their overall concentration at steady-state levels and on their site of generation [45]. Unbalanced ROS production and scavenging contributes to oxidative stress. Despite having a negative influence on health, this stress can also be used to treat clinical conditions, such as cancer, with a certain degree of clinical success [46]. As a natural defense system, the cell regularly produces antioxidants, or adapts other mechanisms, to eliminate the ROS and thus maintain the balance of antioxidation and oxidation processes [4]. ROS and RNS overproduction can result in various destructive effects by disrupting the antioxidant defenses and impairing cellular integrity and functionality [47]. In vivo, free radicals are primarily responsible for chemical modifications and damage to proteins (aggregation and denaturation), lipids (peroxidation), and carbohydrates. They also induce changes to the DNA structure, leading to mutations or cytotoxic effects [48].Very little, if anything, is currently known about the relationship between known biomarkers and oxidative stress in SRLV-SP goats. The most significant finding of the present study is that the SRLV-SP and SRLV-SN goats appear to demonstrate no difference in the levels of oxidative stress biomarkers. However, it should be stressed that throughout the period of infection, the SRLV-SP goats were asymptomatic for CAE, as well as other diseases, including mastitis. This suggests that animals without clinical symptoms of CAE, with low virus loads, do not suffer any oxidative stress and their welfare is not disturbed. However, as oxidative stress is related to a number of aspects of the pathogenesis of viral etiologic agents, this was a surprising finding. Viruses affect the cellular redox balance by increasing oxidants, such as superoxide and nitric oxide, and inhibiting the synthesis of antioxidant enzymes, such as SOD, CAT, and GPx [8]. A study of SRLV by Mdurvwa et al. [49] found, in contrast to our results, that the antioxidant potential of serum from SRLV-infected goats of various age groups demonstrated significantly higher catalase activity than serum from healthy controls. In addition, this activity increased over time following infection with SRLV. However, similar to our results, no differences in total SOD or GPx activity were observed, although Cu, Zn-SOD levels were elevated in the infected goats. The authors also reported a positive correlation between serum catalase activity and hydrogen peroxide (H2O2) scavenging activity [49]. Although a transient increase was observed in lactate dehydrogenase (LDH), no correlation was observed between increased serum catalase activity and LDH activity. The authors also note decreased oxyradical production in SRLV-infected goats. This may be due to the increase in serum catalase, a scavenger of endogenous free radicals, such as H2O2. Nevertheless, Santos et al. [50] showed that, as in the present study, no pronounced dysfunctions in blood or milk polymorphonuclear leukocytes or in monocytes/macrophages were observed in naturally SRLV-infected goats.However, some changes in oxidative biomarkers have been observed in other viral diseases. For example, Balikci et al. [51] compared blood oxidative stress biomarkers (MDA and nitric oxide), antioxidant levels (SOD and GPx), and acute-phase proteins activity (haptoglobin and serum amyloid A) between aborting and non-aborting goats with border disease (BD) caused by border disease virus (BDV). Both the infected non-aborting and infected aborting groups demonstrated a decrease in GPx and SOD activities and an increase in MDA, NO, haptoglobin, and serum amyloid A levels compared to the non-infected group. Additionally, the aborting goats displayed significantly higher MDA, NO, haptoglobin, and serum amyloid A levels, and lower SOD levels than the non-aborting groups. The oxidant-antioxidant balance has also been found to be disrupted in sheep with peste des petits ruminants (PPR), which in turn can cause further oxidative damage. Nisbet et al. [52] examined the changes in the biomarkers of free oxygen radicals and antioxidant activity, i.e., MDA, GPx, and SOD, in sheep with PPR. Their findings indicate that the PPR-positive group demonstrated a significantly higher mean MDA level and lower mean GPx and SOD activities compared to controls.Our present findings are, in general, in agreement with the those in our previous study on cytokine and acute-phase protein (APP) gene expression [25,26], at both the mRNA and protein level, in the blood cells/sera of SRLV-SP but CAE-asymptomatic and SRLV-SN goats. Although Reczyńska et al. [26] found a higher concentration of serum amyloid A (SAA) in blood serum of SRLV-SP goats, this may also indicate that viral multiplication was promoted, since this APP may inhibit antibody production and could stimulate the differentiation of monocytes to macrophages, an essential step in viral multiplication. In addition, Jarczak et al. [25] report a decreased concentration of the pro-inflammatory cytokines IL-1α, IL-6, and INF-β in the sera of SRLV-SP goats. Both these studies indicate that the systemic immune system of infected goats is impaired, thus preventing them from fighting the disease. As it was stressed above, those findings are consistent with the results of the present study—SRLV appears to avoid the systemic immune system and does not cause oxidative stress. They are also consistent with Tian et al. [53], who note that the concentration of cytokines undergoes a series of changes under prolonged stress in humans. Constant stress continues to increase the pro-inflammatory cytokines, which finally cause inflammation and may lead to various diseases. This may mean that the animals involved in the study are in the early stage of the stress caused by infection. All of our studies, both the present study and previous ones, were conducted on asymptomatic goats. Even so, the levels of several oxidative stress biomarkers changed over the course of lactation. In dairy cows, oxidative stress has been closely studied during the transition period and early lactation [54,55,56], during which, a loss of overall antioxidant potential can be observed. This phenomenon is connected with increased metabolic demands and leads to a weakening of the antioxidant defense of dairy cows, accompanied by a greater risk of metabolic diseases [55]. In addition, Piccione et al. [57]. reported that oxidative processes increase at the end of lactation in sheep. Of the various markers of protein oxidation, the best studied is protein carbonyl formation, which increases in tissues and organs during oxidative stress [58,59]. In our study, the levels of both aldehydic and ketonic derivatives of oxidatively modified proteins underwent some fluctuations during lactation. The aldehydic derivatives increased from a low level at the beginning of lactation to peak on day 140 in SRLV-SN goats, while the ketonic derivates peaked on day 200 in both groups (Figure 2 and Figure 3). However, some studies suggest that the accumulation of protein oxidation products increases with age [58]. The increase in the level of ketonic derivatives observed on day 60 of lactation in the SRLV-SP goats is probably associated with the peak of lactation, as well as gaps between nutrient supply and demand. Lactating animals use their own energy reserves to cover their demand as the dietary intake alone is insufficient. The fact that a high level of ketonic derivates was only observed in the serum of SRLV-SP goats may mean that the burden for an infected organism was greatest during the highest milk production period. Under these conditions (i.e., day 60 of lactation in SRLV-SP goats), the energetic metabolism becomes destabilized: the energy loss exceeds the physiological capacity of the animal’s feed intake. ROS-induced activation of lipid peroxidation, occurring in response to hormonal changes during the lactation process, significantly intensifies oxidative modifications in proteins, observed in the accumulation of aldehydic and ketonic derivatives. On the other hand, the serum level of ketonic derivatives peaked on day 200 in both groups, which may indicate that the udder was being prepared for drying off. This period is very critical for all high-yielding dairy animals, including goats. This first stage of the lactation process is associated with an increase in milk yield; however, in the same periods of lactation, increased levels of both aldehydic and ketonic derivatives were observed (Figure 2 and Figure 3). Free radical-induced oxidative stress and intensification of protein metabolism results in oxidative modification of proteins because the goat organism does not receive the necessary amount of energy from the feed and uses the energy stored in the body. During the lactation period, serum TBARS level was found to decrease in both SVRL-SN and SVRL-SP goats after 30 days and remained stable for the rest of the study (Figure 1). High levels of lipid peroxidation were observed on day 1 and day 7 of the experiment, and these can be associated with the physiological changes in the organism of goats related to the postpartum period. This might also be associated with the transition period between colostrum and milk production, which also places a considerable burden on the animal, mainly due to the rapid increase in milk production. The TBARS level was slightly lower on day 60, but still high. This is probably connected with the peak of lactation, i.e., the highest milk production. In contrast, full lactation i.e., on day 140 and day 200, was associated with low TBARS levels, and this was probably related to the physiological stabilization of the lactation process. In goats, changes in metabolism can decrease milk production, thus probably allowing animals with significant energy loads to cover their needs related to lactation.TAC level remained high throughout lactation; however, a decrease was observed on day 200 (Figure 4). This reduction is probably associated with the initiation of the drying off process. Under these conditions, the free radical processes and antioxidant defenses become destabilized. The accumulation of aldehydic and ketonic derivatives of oxidatively modified proteins, together with the depletion of antioxidant defenses, causes a significant decrease in TAC, and increase of oxidative stress. The observed changes may reflect the dysfunction of the body under these conditions.The changes of the antioxidant defenses observed under these conditions in both SRLV-SN and SVRL-SP goats during lactation may reflect changes in the level of oxidative stress biomarkers. This increase in stress, resulting from an increase in the production of free radicals, may induce the activation of the antioxidant defense in tissues. Our present findings reveal a decrease in the activities of antioxidant enzymes (SOD, CAT, and GPx) after 60 days of lactation (Figure 6, Figure 7 and Figure 9). However, in contrast to our results, several studies indicate that the activities of SOD and GPx were reduced in some viral infections. Nisbet et al. [52] highlight that GPx and SOD activities were significantly lower in sheep with peste des petits ruminants than uninfected controls. Some clinical studies also report lower CAT and SOD activities in the blood of hepatitis B virus-infected patients [60]. However, SRLV infection does not appear to influence the activities of antioxidant enzymes in the asymptomatic stage of the disease. Then again, they were found to be influenced by the stage of lactation. Increased levels of protein damage biomarkers can lead to changes in SOD, CAT, and GPx activities. Their reduction suggests that the antioxidant defenses remain inadequate, probably due to increased oxidative-induced protein modification. The decrease in antioxidant enzyme activities observed in the current study may be due to oxidative damage occurring in the cellular structures as a result of the inability to fully detoxify free radicals. As SOD, CAT, and GPx are involved in the conversion of radicals into less effective metabolites, these changes, coupled with an increase in the levels aldehydic and ketonic derivatives of oxidatively modified proteins, confirm the presence of oxidative stress during the lactation period, both in the SRLV-SN and SVRL-SP goats.Ceruloplasmin (Cp) has been shown to exhibit antioxidant functions, which have a beneficial effect under several pathological conditions [61,62]. Cp catalyzes the oxidation of ferrous (Fe2+) into ferric ions (Fe3+), with the process involving four of the sic copper atoms associated with Cp. Dioxygen can also act as an electron acceptor in the absence of any insufficiently reduced ROS, such as O2− or H2O2. This can also determine the binding of iron to transferrin and ferritin [63]. Pro-inflammatory agonists of the acute phase reaction, such as IL-6 and TNF-α, increase hepatic cell Cp synthesis by a transcriptional mechanism [64]. Cp may act as an antioxidant and as an acute-phase reactant [65,66]. Some Cp regulation was observed to the end of the current study; however, although no complete recovery of Cp level was observed by day 200, compared to the first day after parturition, the final level was the same as that on day 7 (Figure 4). The elevated Cp level may have manifested as a response to systemic inflammation [66] and its high level just after parturition is connected probably with the high burden associated with delivery. Cp level was found to decrease on day 30 of lactation, indicating that the perinatal period is over. Therefore, a key finding of our study is that pronounced elements of oxidative stress appear to be absent from infected but asymptomatic animals. We propose this as an area for further research, by including symptomatic animals or animals with a very high viral load in studies. This will help identify the primary links between the effects of viral infection on biomarkers of oxidative stress and the mechanisms initiated by the virus under these conditions.5. ConclusionsIn conclusion, the goats naturally infected with SRLV but without clinical sign of CAE did not reveal any pronounced dysfunctions in serum oxidative stress biomarkers compared to uninfected animals. The only changes in oxidative stress biomarkers observed during lactation probably reflect the burden of goat metabolism caused by perinatal or drying off periods, or milk production. The two groups demonstrated similar biomarkers of oxidative stress in each stage of lactation, which may mean that SRLV does not trigger the immune system and does not cause oxidative stress in all SRLV-SP goats, but the asymptomatic ones. The next step of the study will be to examine samples derived from goats with clear clinical CAE symptoms and detectable virus loads.

animals : an open access journal from mdpi

10.3390/ani12030339

PMC8833459

The greatest challenge of long-distance live transport of fish is high mortality caused by stress and oxidative damages. In this study, the effects of lemon balm (Melissa officinalis L., MO) on the stress response of sea bass were evaluated. Sea bass were treated with different concentrations of MO (10, 20, 40 mg/L, respectively) and were transported for 72 h in transport boxes. The results of this study indicated that the level of cortisol, glucose, lactic acid, heat shock proteins, catalase, myeloperoxidase, glutathione peroxidase, uric acid, and urea nitrogen of samples treated with MO were lower than the control. The sample treated with 40 mg/L MO showed higher antioxidant capacity. In conclusion, the effect of MO on alleviating stress responses was similar to MS-222 and eugenol.

This study was conducted to enhance the viability and alleviate the oxidative stress response using MO for sea bass during live transport. Six experimental groups were designed, and the effects of the physiological responses of MO were evaluated in comparison with MS-222 and eugenol. The physiological stress levels, proprotein convertase subtilisin/kexin type 9 (PCSK-9), antioxidant enzyme activities, and kidney parameters of blood serum were determined. It was found that cortisol level, glucose (Glu), lactic acid (LD), heat shock proteins (HSPs), catalase (CAT), myeloperoxidase (MPO), glutathione peroxidase (GSH-Px), uric acid (UA), and urea nitrogen (BUN) in the MO-treated samples were lower than that of the control (133.72 ng/L); however, the total antioxidant capacity (T-AOC) was higher after 72 h of the simulated live transport. The ability to resist oxidative stress increased along with the increase in the MO concentration in the water during live transport, which was similar to the results of MS-222 and eugenol treatment. In conclusion, MO, acting as a kind of novel sedative and anesthetic, can be used to improve the oxidative system and survival rate during live transport. The results of this study provide a reference for enhancing animal welfare and anti-oxidative stress ability, reducing mortality and the stress response during live fish transport.

1. IntroductionLive fish transport is becoming increasingly important in aquatic product trade and fish are put in bags or containers and then transported alive [1]. However, fish are vulnerable to aberrant environmental stress factors and display robust neuroendocrine and physiologic parameter changes and a stress response [2,3]. Many studies have been conducted on environmental stress of live fish, such as tissue hypoxia stress [4,5], water quality stress [6,7,8], salt concentration and temperature stress [9,10], physical handling stress, etc. [11]. Therefore, the toughest challenge during live fish transportation is minimizing the stress response and sensitivity induced by environmental changes to obtain a high survival rate. The stress response is a complex regulatory mechanism that occurs during live fish transport, which can be divided into three levels [12]. The first stress response is a neuroendocrine reaction, in which the hormones of fish are released along with an increase in cortisol, adrenaline, and other hormones. Nestor et al. [13] found that capture, transport, and high stocking density caused high cortisol levels during live fish transport. The secondary stress response consists of changes in energy metabolism, blood physiology, and biochemistry and immune regulation caused by neuroendocrine reactions. The tertiary stress response affects performance [14]. Wang et al. [15] showed that the function of immune and antioxidative parameters of juvenile tiger puffer (Takifugu rubripes) were affected after short-term live transport and recovered at 168 h.At present, the common methods used for live transport include optimization of transportation technology, upgrade of facilities, and addition of anesthetics or sedatives to the transport water to improve the efficiency of live transport [16,17,18]. Some anesthetics, such as tricaine methanesulfonate (MS-222) and eugenol, have been developed. Essential oils have been used in live fish transport and have a similar efficiency to chemical anesthetics [19,20]. The addition of anesthetics and essential oils to transport water can temporarily block sensory (afferent) nerve conduction to relieve pain, discomfort, struggle, and physical energy consumption induced by stress [21]. Lemon balm (Melissa officinalis L., MO) is also known as American mint and wild bergamot. The main ingredients of lemon balm are aldehydes, terpenes, and phenolic compounds, such as citral, rosmarinic acid, and flavonoids [22,23]. Its mechanism of action involves rosmarinic acid (RA), a compound that inhibits γ-aminobutyric acid (GABA) transaminase activity and slows the degradation of GABA. RA is able to cross the blood–brain barrier and acts on GABA in the brain, thereby maintaining the stability of GABA concentrations in living organisms [24]. At present, some scholars have applied plant essential oil to transport water in live fish transport. Boaventura et al. [25] confirmed that lophiosilurus alexandri (Siluriformes, Pimelodidae) can be stabilized by treatment with 90 and 150 mg/L essential oil of Ocimum gratissimum L. (EOOG) in transport water, resulting in a decrease in the metabolic rate of fish and an increase in the antioxidant stress ability. Khumpirapang et al. [20] and Rodrigues et al. [19] developed Alpinia galanga oil (AGO) and canela-amarela (Nectandra grandiflora) essential oil as novel anesthetics for use during live fish transport. Essential oil is a promising natural source of alternative sedatives and anesthetics for fish. In the present study, sea bass (Lateolabrax japonicus), an edible marine fish, were used as experiment subjects. This study evaluated the negative effects of live fish transport on sea bass by investigating the influence of the addition of MO and two kinds of anesthetics to the transport water. Sea bass were transported in cold water (12 °C) and treated with different concentrations of MO, MS-222, and eugenol. The objective of this study was to exploit a novel and effective sedative and anesthetic to enhance animal welfare and the survival rate of live edible marine fish and reduce the stress effects and oxidative damage.2. Materials and Methods2.1. Preparation of Sea BassThis experiment followed the principles and guidelines established by the animal care and use committee of Shanghai Ocean University (SHOU-DW-2021-67). Sea bass (500 ± 120 g, 39 ± 1 cm) were purchased from the local market in Luchao Port town (Shanghai, China) and transported to the laboratory in live fish transport boxes within 30 min. The temperature in the facility from which the fish were purchased was 20~22 °C. All of the fish were healthy and unhurt, and fish were temporarily cultured for 36 h without feeding [26]. The temporarily cultured water had the following parameters: the density of temporary culture was 20 g/L, the temperature of the water was 20~22 °C, the salinity was 16, the dissolved oxygen was 4~6 mg/L, and the pH was 7.5~8.5. After temporary culture, the temperature was dropped from 20~22 °C to 12 °C at a rate of 3 °C/h. The temperatures used in this study are within the natural tolerance range of this species [3,18,27]. When the temperature of the transport water reached 12 °C, the live transport experiment started.2.2. Live Transport Experimental DesignThe optimum concentration was obtained from a pre-experiment, and the MO and eugenol solution were prepared in a mixture of alcohol and tween-80 (200:1, v/v). Six groups were used in this study. The concentrations of MO in the transport water were 10 (10 MO), 20 (20 MO), and 40 (40 MO) mg/L; MS-222 was 30 mg/L (MS-222) [28]; and eugenol was 20 mg/L (Eugenol) [29], respectively. The character of MS-222 was weakly acidic, and the pH value range of the transport water after adding 30 mg/L MS-222 was 7.0~8.0. The sea bass samples in transport water to no agents were added were considered as the control. Sea bass were stocked and transported in live transport boxes with the same volume of cold water (12 °C) to which different concentrations of MO and two kinds of anesthetics were added to in advance [30]. The volume of the boxes was 108 L. The sea bass underwent simulated transport for 72 h and the density of live transport was 250 g/L. The water parameters were the same as the temporarily cultured water. The number of fish in each box per group was 30, and a total of 180 fish were used in the experiment. MO was purchased from Gaodao essential oil trading company (Chongqing, China). The main components were citral (44.9%), geraniol (21.1%), citionella (15.4%), citronellol (6.3%), rosmarinic acid (4.1%), and D-limonene (2.3%). The simulated transport was as follows: 1 h on a B-level road (80 km/h) →4 h on an A-level road (100 km/h) →1 h on a B-level road (80 km/h), repeated 12 times. Three sea bass were randomly selected for sampling after 2, 4, 6, 8, 10, and 12 cycles. After simulated transport for 72 h, the sea bass samples were recovered in cultured water at room temperature. Eventually, the survival rates were determined [16,31]:Survival rate=Survival numberSample number×100%2.3. Pretreatment of SamplesThe sea bass were stunned with ice water for 15 min and then killed. Ice did not touch the fish during this period. The blood of sea bass was taken from the tail vein without anticoagulant. The blood was stored at 4 °C for 2 h, and then centrifuged at 10,614× g, 4 °C for 5 min, and the liquid supernatant (serum) was collected. The serum was stored at −80 °C before use for index determination.2.4. Determination of the Physiological Stress LevelGlucose and lactic acid in serum were measured using commercial kits (Jiancheng Bioengineering Institute, Nanjing, China). The methods used to detect cortisol and heat shock proteins was enzyme-linked immunosorbent assay (ELISA), and determination of heat stress protein and cortisol antibody-coated pore plates, respectively. The absorbance (OD value) was measured at 450 nm with a microplate reader, and the sample concentration was calculated. Commercial fish ELISA kits for cortisol and heat shock proteins were supplied by Jiancheng Bioengineering Institute (Nanjing, China);2.5. Determination of Proprotein Ponvertase Subtilisin Kexin Type 9 (PCSK-9)The method used to detect PCSK-9 was enzyme-linked immunosorbent assay (ELISA) and determination of alanine aminotransferase (PCSK9) coated on a pore plate. The absorbance (OD value) was measured at 450 nm with a microplate reader, and the sample concentration was calculated. Fish ELISA kits for PCSK-9 were supplied by Jiancheng Bioengineering Institute (Nanjing, China).2.6. Determination of Antioxidant Enzyme ActivitiesCatalase activity (CAT) in the serum was calculated by measuring change in the yellow complex produced by the interaction of hydrogen peroxide and ammonium molybdate. The vitality unit is expressed in U/mL. The activity of myeloperoxidase (MPO) was calculated by the yield of the yellow compound produced by hydrogen donor anisidine, and the vitality unit is expressed in U/mL. CAT and MPO were quantified by commercial kit assays (Jiancheng Bioengineering Institute, Nanjing, China). The consumption of reduced glutathione (GSH) in an enzymatic reaction was measured and the activity of glutathione peroxidase (GSH-Px) was calculated by the speed of the enzymatic reaction. Glutathione peroxidase (GSH-Px) activity was determined by referring to Boaventura et al. [25]. Total antioxidant capacity (T-AOC) in serum was measured using commercial kits (Jiancheng Bioengineering Institute, Nanjing, China).2.7. Determination of Kidney Function IndexUric acid (UA) and blood urea nitrogen (BUN) in serum were determined by Jia et al. [32].2.8. Statistical AnalysisThe one-way ANOVA-Duncan test program in SPSS 21.0 software was used for multiple comparisons. The Levene test was used to check the homogeneity of the samples before applying Duncan. The results are expressed as means ± SD, and the significance threshold was 0.01. Origin software was used to make graphs.3. Results and Discussion3.1. Survival Rates of Sea BassThe survival rates during and after long-distance live transport of sea bass were recorded and are presented in Table 1. The survival rates of the fish treated with MO or anesthetics were higher than that of the control (50%) after 12 h of recovery. The survival rate of samples in the 20 MO and 40 MO groups was 96% after transport and recovery. The survival rate of samples in the MS-222 group was 100% and the survival rate of samples in the eugenol group was 95%. No fish from the 20 MO, 40 MO, and MS-222 treatment groups died during the period of recovery.3.2. Physiological Stress LevelAs shown in Figure 1C, cortisol in sea bass samples increased at the beginning and then decreased during live transportation. Cortisol levels significantly increased from 0 to 12 h. The level of cortisol reflects the stress response of fish [33]. The increase in the cortisol level suggests that fish’s neuroendocrine system exerted a primary response and secreted a large amount of cortisol during the initial stage of environmental stress. Then, the level of cortisol decreased with transport time, which indicated that fish gradually adapted to the initial stress during live transport. However, the accumulation of excreted waste increased the stress response of sea bass. The fish presented a secondary stress response after transport for 60 h and the stress indices reached a peak value after transport for 72 h, which is consistent with the research of Wang et al. and Vanderzwalmen et al. [34,35], who found that MO-treated samples demonstrated lower cortisol levels than other samples during live transport. The cortisol level decreased dramatically after 12 h of recovery, and the cortisol level of the MO treatment groups was the lowest in comparison with those of the other groups with no significant difference.Muscle and liver store energy in the form of glycogen, and blood in the form of glucose, which are important energy substances required for the various life activities of fish. Blood glucose levels can affect fish’s stress response by modulating cortisol release and glucose homeostasis to affect glycogen metabolism and gluconeogenesis [36]. Lactic acid is produced by glycolysis in the presence of an insufficient oxygen supply. The stress response evoked during live fish transportation caused anaerobic muscle activity and the formation of lactic acid both in the muscle and blood [37]. It can be observed from Figure 1A, B that the glucose and lactic acid levels of each treatment groups increased along with the increase of the transport time and reached a peak value after 72 h of live transport and then decreased sharply after 12 h of recovery. The lactic acid level of 40 MO returned to the 0 h level with no significant difference after 12 h of recovery. This trend could be ascribed to the increase in the glycogen level to combat the stress response and the body performing anaerobic respiration to produce large amounts of lactic acid induced by the low temperature. At the same time, the increase in the serum cortisol level promoted the generation of glycogen, resulting in an increase in serum glucose. These results are close to the values presented by Liu et al. and Zhao et al. [38,39]. It was observed that the glucose and lactic acid levels of the control were the highest and those of the sea bass in the group treated with 40 MO were the lowest during the whole simulated transportation. The glucose level in the serum gradually decreased along with the increase in the MO concentration and the glucose level of sea bass in 40 MO was lower than those of other groups. Similar trends were encountered for the change in the lactic acid level for all treatment groups, with differences between the values. It was indicated that the energy consumption of sedated and anesthetized fish was lower in comparison with that of the control, and the levels of glucose and lactic acid in the samples of the 40 MO treatment group were the lowest in comparison with those of other groups.The heat shock protein level is low and stable for fish under normal conditions; however, the heat shock protein level increases to different degrees to protect the body tissues and organs from damage when fish are exposed to stress [40]. Figure 1D shows an increase trend in heat shock proteins among each treatment groups during the whole live transport process. After transport for 72 h, heat shock proteins in the control displayed the highest level (135.20 ng/L) with no significant difference to that of the eugenol group (131.86 ng/L) while the 40 MO groups showed the lowest level (122.91 ng/L) with no significant difference with that of the 20 MO group (125.53 ng/L) and MS-222 group (124.41 ng/L). As compared with the control, the fish stress response of the groups treated with MO and anesthetics was relatively light. With the increase in the MO concentration in the transport water, the production of heat shock proteins was relatively low during live sea bass transport.The sedated and anesthetized fish showed a lower physiological stress response than that of the control, especially fish in the 40 MO treatment group, which showed a minimum change in the serum stress indexes. The application of MO to the transport water mitigated the stress response and improved the antistress ability of sea bass.3.3. PCSK-9PCSK-9 is also known as neural apoptosis-regulated convertase (NARC1) and is a newly identified subtilase belonging to the peptidase S8 subfamily. Additionally, PCSK9 is suggested to participate in immunoregulation by modulating cytokine production and indirectly affecting the apoptosis rate and blood cholesterol levels. Activated PCSK-9 plays a role in the promotion of apoptosis [41,42,43]. It can be observed from Figure 2 that the PCSK-9 activity in each group increased dramatically after 12 h of transport, which indicated that fish suffered stress responses and cell apoptosis during the initial simulated transport and PCSK-9 activity was activated in the organisms. There was a significant increase in PCSK-9 activity in the control (39.42 ng/L) after 72 h of transport due to the persistent stress response. The activity of PCSK-9 decreased dramatically after 12 h of recovery. The results suggested that a decrease in the PCSK-9 activity indicated alleviation of the stress response for sea bass after 12 h of recovery; however, the stress response was not completely eliminated. It is proved that both anesthetics and MO can be used to protect the body from injury, and the results of this study demonstrate the efficacy of anesthetics and sedatives with MO.3.4. Antioxidant Enzyme ActivitiesThe activity of antioxidant enzymes in fish was activated in order to resist lipid peroxidation in the stress environment. SOD converts O2− to O2 and H2O2 first, and then CAT catalyzes H2O2 to harmless H2O and O2 at random. As an important peroxidase decomposition enzyme, GSH-Px catalyzes other harmful peroxides in the body and reduces toxic peroxides to non-toxic hydroxyl compounds, to protect the structure and function of cell membranes of living organisms [44,45]. MPO is a marker of neutrophil function and activation, and the peroxidase activity of MPO generates hypochlorite to kill microbes and inactivate inhibitors of lytic enzymes. Next, neutrophil leukocytes are released to degrade material in their vicinity [46]. Therefore, these enzymes make up the body’s antioxidant defense system.It can be observed from Figure 3A–C that the CAT, GSH-Px, and MPO activity in the samples of each group showed a slow growth trend during live transport and presented a sharp increase in CAT and MPO activities in the control (p ≤ 0.01) after 72 h of transport, which indicated that oxidative stress displayed an exponentially increasing trend along with the increase in the transport time. However, Karu et al. [47] and Zeng at al. [48] found that the CAT activity in fish blood did not change significantly and the activity of CAT increased in fish liver. These changes may indicate a specific ability of fish to protect the tissue and organs of the body under low-temperature conditions. After transport for 72 h, the CAT, GSH-Px, and MPO activities of samples in the 40 MO group showed the lowest level in comparison with the other groups, and the antioxidant efficiency was positively correlated with the concentration of MO. The GSH-Px activity in the samples treated with MS-222 demonstrated a sharply increasing tendency to the highest level (855.14 mol/L) after 72 h of transport. The enzymatic activity in the samples treated with 40 MO (642.16 mol/L) presented lower activity than the samples treated with anesthetics (MS-222: 855.14 mol/L, eugenol: 685.41 mol/L) and the control (830.27 mol/L). This may be because anesthetized fish were woken up for anesthetic consumption and presented a stress response, or different individuals showed various oxidative stress responses. After 12 h of recovery, the antioxidant enzyme activities in each group demonstrated an obvious downward trend. The enzymatic activity of the samples in the 40 MO group was the lowest (340.54 mol/L) compared with the samples in the other groups, which meant that the addition of 40 mg/L MO to the transport water alleviated the stress effects on live fish more quickly during the recovery period and enabled fish to gradually recover to the state before transportation. Live transportation under low temperatures can inhibit enzyme activity in fish to some extent. Moreover, the application of MO and anesthetics demonstrated an excellent antioxidant enzyme activity effect under low temperatures.The total antioxidant capacity (T-AOC) refers to the total antioxidant level composed of various antioxidant substances and antioxidant enzymes to protect cells and the body from oxidative stress damage caused by reactive oxygen species. The total antioxidant capacity can be used to evaluate the antioxidant capacity of biologically active substances [49]. Figure 3D indicates that the level of total antioxidant capacity decreased along with the increase in the simulated transport time. The total antioxidant capacity of the control declined dramatically with a significant difference compared with the other treatment groups (p ≤ 0.01) during live transport. The total antioxidant capacity of sea bass decreased with the increase in the MO concentration in the transport water during live transport. After 12 h of recovery, the total antioxidant capacity level of samples in the 40 MO group reached the highest level (0.59 U/mL) in comparison with the other groups. Thus, it was found that the addition of MO to the transport water had sedative and antioxidant stress effects during live fish transport and protected fish from oxidative stress. The recommended concentration of MO is 40 mg/L.3.5. Kidney Metabolism IndexChanges in the uric acid level can fully reflect the conditions of metabolism and the immune ability of animals. Serum urea nitrogen is filtered out by the glomerulus and is the main product of nitrogen organic matter and protein metabolism. However, the kidney is susceptible to invasion by various factors and reduction of the waste excretion function, leading to harmful toxins entering the body that cannot be excreted to the outside of the body normally. This leads to an increase in uric acid and urea nitrogen in the blood [50,51]. Figure 4 depicts the changes in uric acid and urea nitrogen in the different treatment groups during live transport of sea bass. It is obvious that the uric acid and urea nitrogen levels of the samples in the control were higher than those of the other groups (p ≤ 0.01) while the addition of MO and anesthetics to the transport water reduced this increase for fish to different degrees. The growth rate of uric acid and urea nitrogen slowed down along with the increase in the MO concentration. The uric acid level of samples in the 40 MO group (200.59 μmol/L) and the urea nitrogen level of samples in the MS-222 group (5.90 mmol/L) were at the lowest level, but no differences were observed among the 10 MO, 20 MO, 40 MO, and MS-222 groups after transport for 72 h. The application of MO and MS-222 had similar effects in alleviating kidney metabolism damage during live fish transport. After 12 h of recovery, the uric acid and urea nitrogen levels in fish showed decreasing trends, and the uric acid level of samples in the 40 MO group (93.77 μmol/L) reached the level of samples at 0 h (48.66 μmol/L) with no significant difference. The urea nitrogen level of samples in the 10 MO (5.48 mmol/L), 20 MO (5.23 mmol/L), 40 MO (5.06 mmol/L), MS-222 (5.17 mmol/L), and eugenol (5.29 mmol/L) groups reached the level of samples at 0 h (4.87 mmol/L) with no significant difference. To summarize, as discussed in this section, it was proved that the addition of 40 mg/L MO to the transport water played an important role in the antistress response of live fish.4. ConclusionsThe addition of anesthetics and sedatives to transport water is effective and traditional methods keep fish alive during long-distance transport. In this study, sea bass were selected as the research subject, and the effects of live transport on its survivability and oxidative stress injury were investigated. Samples treated with MO and anesthetics were protected from oxidative damage compared to the control. The survival rates of sea bass in the 10 MO, 20 MO, 40 MO, MS-222, and eugenol treatment were higher (80%, 96%, 96%, 100%, 95%, respectively) than the control (50%). The 40 MO sample presented lower levels of serum stress indices, cell apoptosis, and kidney metabolism damage. This study preliminarily explored the effects of MO on reducing stress responses of and physiological changes in live sea bass. The application of MO to the transport water enhanced the antistress ability and survivability during live transport. It is essential to explore the anesthetic and sedative mechanism of essential oil and to enhance animal welfare and minimize the physical stress effects during live transport.

animals : an open access journal from mdpi

10.3390/ani13111778

PMC10251824

Feline calicivirus (FCV) is a highly contagious virus found in cats and a cause of upper respiratory and oral infections. Typical clinical signs of FCV include nasal discharge, gingivitis, and stomatitis. FCV is also able to affect the joints of cats, resulting in lameness. In this study, we monitored a small outbreak of FCV limping disease in two household cats. The transmission between the two animals likely occurred indirectly via virus shed in the environment from the respiratory tract. The findings of this study highlight the need for the adoption of adequate prophylaxis measures to prevent the transmission of highly transmissible infectious diseases.

Feline calicivirus (FCV) is a common viral pathogen found in domestic cats. FCV is highly contagious and demonstrates a high genetic variability. Upper respiratory tract disease, oral ulcerations, salivation, and gingivitis–stomatitis have been regarded as typical clinical signs of FCV infection. Ulcerative dermatitis, abortion, severe pneumonia, enteritis, chronic stomatitis, and virulent systemic disease have been reported more sporadically. Limping syndrome has been also described either in naturally or experimentally FCV-infected cats. In this study, we monitored a small outbreak of FCV infection in two household cats, in which limping disease was monitored with a 12-day lag time. The complete genome sequence was determined for the viruses isolated from the oropharyngeal and rectal swabs of the two animals, mapping up to 39 synonymous nucleotide mutations. The four isolates were sensitive to low pH conditions and trypsin treatment, a pattern usually associated with viruses isolated from the upper respiratory tract. Overall, the asynchronous pattern of infections and the results of genome sequencing suggest that a virus of respiratory origin was transmitted between the animals and that the FCV strain was able to retain the limping disease pathotype during the transmission chain, as previously observed in experimental studies with FCV strains associated with lameness.

1. IntroductionFeline calicivirus (FCV) belongs to the genus Vesivirus, included in the Caliciviridae family [1]. FCV is an icosahedral virus with a positive-sense, single-stranded RNA genome approximately 7.7 kb in length [2]. The FCV genome contains a virus-encoded protein at the 5′ end, a poly-A tail at the 3′ end, and three adjacent open reading frames (ORFs). ORF1 encodes a polyprotein that is proteolytically cleaved into nonstructural proteins, including a viral protease and the RNA-dependent RNA polymerase. ORF2 encodes the capsid precursor protein, which is post-translationally cleaved into the leader capsid protein and the mature capsid protein VP1. ORF3 encodes the putative minor capsid protein VP2 [3,4].FCV is an important infectious and endemic pathogen and mostly affects cats under one year of age. FCV infection usually causes mild self-limiting clinical manifestations with high morbidity and low mortality [5,6]. Clinical signs can evolve as acute or subacute infection of the upper respiratory tract and oral cavity (stomatitis and oral inflammatory syndrome) [7,8]. FCV infection is rarely associated with ulcerative dermatitis, abortion, severe pneumonia [2], chronic stomatitis, or virulent systemic disease (VSD) [9]. Vaccination is indicated for all cats [5], although the efficacy of FCV vaccines is affected by viral genetic and antigenic variability [10]. Current vaccines are either mono-valent, including only a single FCV strain (F9 or 255), or bi-valent, including two strains (G1 and 431), but they do not seem to cross-protect effectively against all the field strains [11].Limping syndrome is also a rare outcome of FCV infection. This syndrome has been described either in naturally or experimentally infected kittens [12,13] and older cats [14]. The peculiar clinical signs are stiffness, hyperesthesia, mild joint pain, and muscle soreness. The pathogenesis of this syndrome likely involves immune complexes [15], and FCV has been isolated from affected joints [12]. FCV antigens have been detected in the joints of cats experimentally inoculated with either a field or a vaccine virus [15]. Thickened synovial membranes and the intra-synovial collection of fluid have been observed. Pyrexia is commonly present, and some cats show concurrent signs of respiratory disease with or without oral ulceration [12].In this paper, we report a small outbreak of limping disease observed in two household cats. The animals underwent asynchronous infection with a 12-day lag time, suggesting sequential infection between the two animals and, therefore, an intrinsic ability of the FCV strain to induce limping.2. Materials and Methods2.1. Clinical Cases and Sample CollectionA 7-month-old female domestic shorthair (DSH) cat #460/20-1 (cat-1) with outdoor access was presented in November 2020 (day 0) to a veterinary clinic in Andria, Apulia region, Italy. Upon clinical examination, the cat showed lameness, with painful joints, fever (41 °C), oral lesions (painful oral ulcers, stomatitis, glossitis), anorexia, depression, and mild respiratory signs. The clinical signs had started 2 days before presentation to the veterinary clinic (days 2 and 1) (Table 1).On day 9, a cohabiting male DSH cat #460/20-2 (cat-2) of the same litter displayed similar clinical signs. Both cats were unvaccinated. Serum biochemistry for both cats (at day 7 for cat-1 and day 16 for cat-2) revealed remarkable elevations of the serum amyloid-A protein (240.9–257.1 ug/mL; reference interval [RI] 0.1–0.5 ug/mL) and decreased levels of iron in the blood (21–26 ug/dL; RI 55–152 ug/dL). Serum electrophoresis evidenced an increase in alpha1 (2.2–2.6%; RI 0.8–2.0%), alpha2 (24.9–30.0%; RI 8.0–20.0%), and beta globulins (16.7%; RI 7.0–14.0). These data were indicative of acute inflammation.After treatment with cefovecin (Convenia, Zoetis; 8 mg/kg/day), prednisolone trimethyl acetate (1 mg/kg/day), and fluid therapy, the cats recovered within 15 days.Oropharyngeal swabs and rectal swabs (OSs and RSs, respectively) were collected on days 0, 7, and 14 from cat-1 and on days 9, 16, and 23 from cat-2. Blood and serum samples were collected during hospitalization on day 7 from cat-1 and day 16 from cat-2 (Table 1).The collected samples were sent to the infectious diseases Section of the Department of Veterinary Medicine of the University of Bari with a suspected FCV infection based on clinical presentation.2.2. Nucleic Acid Extraction from SamplesThe collected swab samples were processed by homogenization at 10% w/v in Dulbecco’s minimal essential medium (D-MEM). The supernatant was separated by centrifugation at 2500× g for 10 min. Nucleic acids were extracted from the supernatant of swab homogenates, sera, and blood samples using the QIAamp® cador® Pathogen Mini Kit (Qiagen S.p.A., Milan, Italy), according to the manufacturer’s instructions.2.3. Molecular Screening for FCVReverse transcription (RT) of RNA extracts was carried out using the GeneAmp® RNA PCR kit (Life Technologies Italia Applera Italia, Monza, Italy). FCV was identified using a quantitative RT-PCR (RT-qPCR) assay based on TaqMan technology targeting the ORF1 region of FCV [16]. A total of 10 μL of the cDNA was added to 15 μL of a master reaction mix containing 0.6 μmol/L of each primer (FCV for GTTGGATGAACTACCCGCCAATC; FCV rev: CATATGCGGCTCTGATGGCTTGAAACTG) and 0.1 μmol/L of the probe (FCV- prob: [FAM] -TCGGTGTTTGATTTGGCCTG- [BHQ1]).To confirm the presence of FCV RNA, the samples were screened by a one-step RT-PCR assay using the SuperScript One-Step RT-PCR kit (Invitrogen, LifeTechnologies, Milan, Italy) and the forward (Cali1 5′-AACCTGCGCTAACGTGCTTA-3′) and reverse (Cali2 5′-CAGTGACAATACACCCAGAAG-3′) primers, which amplified a 926-bp fragment corresponding to the conserved regions A and B of the ORF2 region of FCV [17].A nested PCR was performed on a 1:10 dilution of the one-step RT-PCR products using the Hot Master Taq DNA Polymerase (Eppendorf) with the forward (Cali3 5′-TGGTGATGATGAATGGGCATC-3′) and reverse (Cali4 5′-ACACCAGAGCCAGAGATAGA-3′) primers, which amplified a 477 bp portion of the ORF2 gene [17].The PCR amplicons were purified by the Qiaquick PCR Purification Kit (Qiagen GmbH, Hilden, Germany). Samples with sufficient DNA concentrations (>10 ng/μL) were used for Sanger sequencing at Eurofins Genomics (Milano, Italy). Analysis of the sequences was carried out using BLAST (http://www.ncbi.nlm.nih.gov, accessed on 1 April 2023) and FASTA (http://www.ebi.ac.uk/fasta33, accessed on 1 April 2023) web-based tools.2.4. Molecular Screening for Other Pathogens The DNA extracts obtained from OS were screened by PCR assay for the feline herpes virus (FeHV) [18]. Screening for the feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV) was carried out with an immunochromatographic assay (Virbac Test Speed DUO FeLV/FIV, Italy) and confirmed using a nested PCR and RT-qPCR, respectively, on blood and serum samples [19,20].2.5. Cells and VirusCrandell–Rees feline kidney (CRFK) cells were cultured at 37 °C in a 5% CO2 atmosphere in D-MEM. The same medium was used for subsequent experiments. The viruses used in this study included FCV strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020.2.6. Virus IsolationOS and RS samples from cat-1 and cat-2 were immersed in 1.5 mL of D-MEM and centrifuged at 5000× g for 5 min. The supernatant was then treated with antibiotics and inoculated onto CRFK cell monolayers.Viral growth was monitored daily for the onset of the cellular cytopathic effect (cpe) and cell supernatants were tested by RT-qPCR. Samples with evidence of cpe at the first cell passage were subsequently titrated.2.7. Viral Titration Ten-fold dilutions (up to 10−9) of each supernatant were titrated in quadruplicates in 96-well plates containing CRFK cells. The titer was calculated by the end-point dilution method after 72 h of incubation at 37 °C.2.8. Seroneutralization (SN)SN assays were performed on the sera of cat-1 and cat-2, collected on days 7 and 16, respectively. Serial 2-fold dilutions of heat-inactivated sera (56 °C for 30 min) were mixed in 96-well microtiter plates with 100 Tissue Culture Infectious Doses (TCID50) of homologous strains isolated from cat-1 and cat-2.After 45 min of contact at 37 °C, 20,000 CRFK cells per well were added. The neutralization titer of the sera was evaluated after 3 days of incubation.2.9. Evaluation of Susceptibility to pH, Trypsin, and Bile SaltsThe in vitro susceptibility to low pH, trypsin, and bile salt treatment of FCVs was investigated and compared with control strains, using protocols previously described [21].2.10. Full-Genome Amplification The complete viral genomes of the OSs and RSs of strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 were generated using consensus primers p1277 (GGCCGCCGGGTTATTGTAAAAGAAATTTGAGACAA) and p1278 (CCGAAGTTGGGGGGGTTTTTTTTTTTTTTTTTTTTTTTTTTCCCTGGGGTTAGGCGCA), binding at the terminations of the FCV genome [21]. Briefly, the RNA was reverse transcribed with primer p1278 using the SuperScript III First-Strand cDNA synthesis kit (Invitrogen Ltd., Milan, Italy). PCR was then performed with TaKaRa LA PCR Kit Ver. 2.1 (Takara Bio, Tokyo, Japan) with primers p1277 and p1278. The amplicons were gel-purified by the Qiaquick Gel Extraction Kit (Qiagen GmbH, Hilden, Germany).2.11. Oxford Nanopore SequencingThe preparation of the DNA library for sequencing was carried out using the Ligation Sequencing Kit (SQK-LSK110), following the manufacturer’s guidelines. DNA was quantified with the Fluorometric Qubit dsDNA High Sensitivity Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). Quality control analysis was assessed on the DNA libraries with the High Sensitivity DNA kit of the Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA, USA). Adapters were added prior to library loading on a Flow Cell R9.4.1 and sequencing was performed using a MinION Mk1c sequencer (Oxford Nanopore Technologies, ONT, Oxford, UK). All purification steps were carried out using AMPure XP beads (Agencourt, Beckman Coulter, Brea, CA, USA) according to the SQK-LSK110 sequencing protocol. For the sequencing of the libraries, the NC_12hr_sequencing_FLO-R9_SQK-LSK110 program was run and MinKNOW Software v.4.0.1 (ONT, Oxford, UK) was used for the base calling of raw sequence data.2.12. Sequence and Phylogenetic AnalysisThe total paired reads obtained were checked for quality, trimmed, and assembled to reference FCV strains using the Minimap2 plugin implemented in Geneious Prime software version 2022.0.1 (Biomatters Ltd., Auckland, New Zealand).Full-genome FCV sequences obtained from the NCBI database were aligned using the software Multiple Alignment using the Fast Fourier Transform (MAFFT) plugin implemented in Geneious Prime software version 2022.0.1 (Biomatters Ltd., Auckland, New Zealand). Multiple alignments of the full amino acid (aa) sequence of the capsid and its hypervariable E region were also inspected to identify hallmark mutations based on a review of the literature [22]. The appropriate substitution model settings for the phylogenetic analysis and estimation of selection pressure on coding sequences were derived using “Find the best protein DNA/Protein Models” implemented in the freely available online MEGA X version 10.0.5 software (https://www.megasoftware.net/, accessed on 1 April 2023). The evolutionary history of the nucleotide sequences was inferred by using the maximum likelihood method, six-character states (general time-reversible model), a discrete gamma distribution, and a proportion of invariable sites to model evolutionary rate differences among sites (six categories), supplying statistical support with 1000 replicates.The evolutionary history of the protein sequences was deduced by implementing a maximum likelihood method and a Jones–Taylor–Thornton (JTT) distance model correct for multiple substitutions based on the model of aa substitution and described as substitution-rate matrices; moreover, a discrete gamma distribution and a proportion of invariable sites were used, providing statistical support with 1000 replicates.Phylogenetic analyses using other evolutionary models (Bayesian inference, neighbor-joining) were performed to compare the topology of phylogenetic trees. Similar topologies were observed with slight differences in bootstrap values at the nodes of the tree, thus retaining the maximum likelihood tree.2.13. Analysis of the Hypervariable Region ESeven remarkable residue positions of the hypervariable region E in the capsid region were previously found to be statistically significant for pathotype differentiation [22]. The aa positions were 438, 440, 448, 452, and 455 in the N-HV part of region E, 465 in the central conserved part, and 492 in the C-HV part of region E. Each aa could be described by a set of nine properties, i.e., hydrophobic, positive, negative, polar, charged, small, aromatic, aliphatic, and proline, as previously described [22]. Multiple correspondence analysis (MCA) was performed based on 14 peculiar aa properties of the seven residue positions of the hypervariable region E of strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 identified in this study. These aa properties were compared with those of a set of 61 FCV strains retrieved from the GenBank database: 4 limping strains, 4 vaccine strains, 36 classical oral respiratory disease (ORD) strains, 6 enteric strains, and 11 VSD strains. All strains were identified or isolated between 1958 and 2019 from several countries worldwide (Table 2).The MCA was carried out using XLSTAT software (Data Analysis and Statistical Solution for Microsoft Excel, Addinsoft, Paris, France 2017).2.14. GenBank Sequence SubmissionThe complete genomic sequences of the FCV strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 identified in the OSs and RSs of cats were deposited in GenBank under the accession numbers OP626899, OP626900, OK428795, and OP626901, respectively.3. Results3.1. Molecular Investigation, Virus Isolation, and TitrationThe two cats tested positive with the RT-qPCR assay for FCV. In detail, cat-1 presented an increasing viral load in the OS, ranging from 1.4 × 103 RNA copies/mL on day 0 to 6.6 × 104 RNA copies/mL on day 14. Viral loads in the RS were 8.1 × 101 RNA copies/mL on day 0, 3.0 × 102 RNA copies/mL on day 7, and 1.0 × 100 RNA copies/mL on day 14.Cat-2 presented an increasing viral load in the OS ranging from 8.1 × 101 RNA copies/mL on day 9 to 1.1 × 103 RNA copies/mL on day 23, whilst the RS presented a decreasing viral load ranging from 5.81 × 101 RNA copies/mL on day 9 to 1.0 × 100 RNA copies/mL on day 23 (Figure 1).Samples that tested positive for FCV by RT-qPCR were subjected to a nested RT-PCR protocol that amplified a short diagnostic genome fragment (477 nucleotides, nt) of the ORF2 region of FCV. Samples that yielded visible PCR products under gel visualization were subjected to Sanger direct sequencing. The sequences from the OSs and RSs of cat-1 and cat-2 shared 98.7–100.0% nucleotide (nt) identities with each other. By FASTA and BLAST analyses, the sequences displayed the highest nt identities (87.0 to 87.2%) to FCV strain Case 9 (GenBank accession no. KP862871). Furthermore, all the samples analyzed in this study tested negative for FeHV, FeLV, and FIV, ruling out mixed infections.By visual inspection of the CRFK cell monolayers inoculated with the OSs and RSs collected from cat-1 and cat-2, a cpe referable to calicivirus replication was observed after 24 h. FCV was isolated at the first cell passage from the OSs and RSs collected from both the cats on day 0 and day 9, respectively. The OS and RS of cat-1 presented a titer of 3.25 and 1.75 log10 TCID50/50 µL, respectively, whilst the OS and RS of cat-2 displayed a titer of 1.50 and 1.75 log10 TCID50/50 µL, respectively. A second passage on CRFK cells was performed and virus titers increased to 7.0–7.25 log10 TCID50/50 µL. The viruses were stored at −80 °C and then used for the evaluation of phenotypes and for SN.3.2. Sequence Analysis of FCV The 7687-nt complete genome sequences of the strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 were reconstructed from the OSs and RSs of cats. Sequence analyses predicted three ORFs in the genomic sequence of the strains by comparison with other FCVs retrieved from the GenBank database. ORF1 was 5292 nt long (nt 16–5307) and encoded a polyprotein of 1763 aa. ORF2 was 2007 nt in length (nt 5310–7316) and encoded for a capsid protein of 668 aa. ORF3 was 321 nt long (nt 7313–7633) and encoded a protein of 106 aa. Strains 460.20-1/ITA/2020 to 460.20-2/ITA/2020 collected from OSs and RSs shared a 99.5–100% nt identity to each other at the full-genome level, whilst the identity of the deduced aa was 100% for polyproteins and the capsid proteins VP1 and VP2. The strains identified in this study displayed the highest nt identity (equal to 81.3%) to the FCV isolate CH-JL4 (KT206207) in the full-genome; this was followed by the FCV strain GXNN01-19 (MZ712023) in the ORF1 portion (82.6%), the FCV strain 182/2015/ITA (MT008247) in the ORF2 portion (80.6%), and the FCV strain KP331/2020/TH (MZ064640) in the ORF3 portion (90.8%).Upon phylogenetic analysis based on the full genome sequence, there were several polyphyletic clades, and the strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 clustered along with FCVs of ORD, VSD, and enteric pathotypes, although not tightly (Figure S1). The overall nt identity of the full genome sequences of strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 ranged from 76.0% to 81.3% for other FCV strains.The complete capsid sequences of FCV strains with different pathotypes were retrieved from the databases and used to perform the phylogenetic analysis. In the tree (Figure 2), the strains identified in this study clustered with FCVs displaying ORD pathotypes. A close genetic relatedness to the lameness-associated FCV strain 2280 was also evidenced (Figure 2). Strains 460.20-1/ITA/2020 and 460.20-2/ITA/2020 displayed an overall aa identity of 83.5–92.2% in the capsid region for the other FCVs used for the phylogenetic analysis.In the phylogenetic tree based on the hypervariable region E, the FCV strains identified in this study formed a clade with ORD and enteric strains; limping-associated FCV strains 1466-1 and 1466-2, identified in cats with polyarthritis in Northern Italy [23]; and the lameness-associated strain 2280, isolated in 1982 in Canada [13] (Figure 3).3.3. MCA An initial MCA was carried out on the aa properties of the matched sequences, considering these properties as categorical variables, regardless of the associated pathotype. The expressed percentage of variance was 26.86% for the first axis and 15.34% for the second axis. On the first axis, MCA was able to discriminate between VSD strains and limping, enteric, vaccine, and ORD strains, although the pathotype was not involved in the set of variables (Figure 4). The abscise values were subjected to an ANOVA (analysis of variance) test: the mean value for the abscises of ORD, limping, enteric, and vaccine sequences was significantly different from the mean value for the abscises of VSD sequences (F-test, p < 0.0001), substantiating the segregation of ORD, enteric, limping, and vaccine strains from VSD sequences.Nevertheless, some ORD, enteric, and limping strains remained undifferentiated (i.e., located near the center of the graph) or closer to VSD strains; a unique VSD strain (strain Georgie) was located along with ORD, limping, enteric, and vaccine strains (Figure 4).A second MCA was performed by adding the “pathotype” as a further variable to the aa properties. The results confirmed those of the first analysis. The uniformity of both analyses confirmed that the aa properties were relevant factors for discriminating between VSD and non-VSD pathotypes.On the basis of 14 peculiar aa properties from 7 residues, the limping strain identified in this study displayed 4 properties from 4 residues (438, 448, and 452 in the N-hypervariable part of region E and 465 in the central conserved part) that were significantly associated with the limping pathotype (Table 3).According to the MCA, the limping FCV strain could display a composition of the following properties: polar aa in position 438, small aa in position 448, small aa in position 452, and non-hydrophobic aa in position 465 (Table 3).3.4. SNThe serum of cat-1 collected on day 7 showed an antibody titer of 1:128; the serum of cat-2 collected on day 16 showed an antibody titer of 1:32.3.5. Evaluation of Susceptibility to pH, Trypsin, and Bile SaltsThe results of the in vitro evaluation of the effects of pH, trypsin, and exposure to bile salts are reported in Table 4. In the acid lability test, both strains showed a loss of infectious titer of 4.0 log10 TCID50. When analyzing susceptibility to trypsin, the two isolates showed a 3.5–4.0 log10 TCID50 reduction. Exposure to bile salts apparently did not affect virus infectivity, with a reduction ranging from 0 to 0.25 log10 TCID50 3.2.4. DiscussionFCV markedly varies genetically, antigenically, and phenotypically as a result of a relentless process of evolution based on the accumulation of punctate mutations, persistent infections, and recombination [24]. The existence/distinction of two distinct FCV disease phenotypes, referred to as classical (or respiratory) and VSD (or hypervirulent) forms or pathotypes has now been largely accepted [25,26]. In addition, FCV may also be associated with other peculiar clinical manifestations, such as enteric disease in cats [21,27,28] and limping syndrome [12,14]. Thus far, it has been difficult to decipher the mechanisms of phenotype variation and find genetic hallmarks of this epiphenomenon [22,29]. However, experimental infections have been able to reproduce the VSD form [30], classical upper tract respiratory disease [13,31], enteritis [27], and limping disease [13], suggesting that different FCV strains have different propensities to cause specific disease syndromes [32].Although lameness was described with FCV infection as early as 1960, it was largely ignored in subsequent studies until the early 1980s [14]. Lameness has since been reported with a number of FCV strains, including modified live vaccine strains [12,15,33,34,35]. In this study, we monitored a small outbreak of limping disease in two unvaccinated 7-month-old household cats. The animals underwent asynchronous infection with a delay of 12 days. Clinical signs included fever, limping, and oral ulcers, whilst the respiratory signs were mild. Both the animals exhibited antibodies 8–10 days after the onset of the disease. Upon genome analysis of the isolates obtained from the OSs and RSs, the viruses detected in cat-1 and cat-2 were highly similar to each other (≥99.5% nt identity) and all 39 nucleotide mutations were found to be non-effective. Virus 460.20-1 RS/ITA/2020 exhibited the highest number of nt mutations (n = 34 for virus 460.20-1 OS/ITA/2020, n = 39 for virus 460.20-2 RS/ITA/2020, and n = 34 for virus 460.20-2 OS/ITA/2020), whilst the other three isolates differed only by 0–5 nt mutations in the consensus genome sequences. Since there is evidence of some phenotype differences between respiratory and enteric FCV isolates [21], the four isolates obtained from cat-1 and cat-2 were tested in terms of their resistance to pH, trypsin, and bile, revealing a pattern of low resistance to pH and trypsin treatment, which was suggestive of a respiratory disease origin. Finally, the neutralizing antibody titer was 3-fold higher in cat-1 than in cat-2; this was likely due to the fact that cat-2 was sampled at an earlier stage of infection (9 days after the onset of clinical signs for cat-1 versus 7 days after the onset of clinical signs for cat-2). Since the cats were not vaccinated, we hypothesized that the antibody response was due to serum conversion, although this was not assessed with serum samples collected at other time points.Overall, the chronology and sequence data suggest sequential infection of the two animals, with cat-2 being infected by cat-1, likely with a virus derived from the respiratory tract. Curiously, in both animals, the virus from the enteric tract seemed prone to a higher degree of diversification, whilst the isolates made from the oropharyngeal tract of the two animals 9 days apart were virtually identical in the consensus sequence. As cat-1 was hospitalized and, since then, the two cats were kept separated, and considering that the incubation of FCV can range between 2 and 10 days [36], we hypothesize that the infection of cat-2 occurred by indirect contact, i.e., by exposure to the respiratory virus of cat-1 shed in the household, rather than by direct contact with cat-1, despite the owner of the animals being warned and instructed to properly clean and disinfect the household environment. Due to the lack of a viral envelope, FCV is highly resistant to inactivation [37] and FCV infection can occur through contact via fomites [38]. This hypothesis is consistent with the high genetic conservation of the respiratory isolates of cat-1 and cat-2.A limitation of our study is that we did not carry out experimental infections in cats to reproduce the observed clinical form with the FCV isolates made from the two animals. Several experiments have been carried out to investigate the limping phenotype using different FCV strains and different routes of infection, thus making it difficult to compare the data from the literature. The FCV strain F65, used in several experiments to study FCV-associated lameness, was isolated from an outbreak of lameness and oral disease in a household of unvaccinated cats [35]. Interestingly, the findings observed in the outbreak of febrile lameness described in our study seem to mirror experimental data observed with strain F65 [12]. Twenty-four hours after the intra-articular inoculation of a group of cats with strain F65, the virus was identified in the oropharyngeal cavity of co-housed animals. Additionally, clinical signs of lameness appeared in three out of four contact cats from 5 to 9 days after infection, and lasted for 1 to 5 days [12], suggesting horizontal transmission by a natural route of infection and conservation of the lameness phenotype. In our study, limping appeared in cat-2 with a delay of 12 days and lasted for 11 days, thus suggesting a transmission chain. This would also suggest that the propensity of some FCV strains to cause lameness is maintained over cycles of infections among animals and can be, as in our case, the prominent clinical sign.Experimental infections with FCV have explored patterns of infection and the onset of the lameness pathotype after oro-nasal exposure [13]. Oro-nasal infection with a pneumotropic FCV strain (FCV 255) and a lameness-associated isolate (FCV 2280) in kittens failed to induce severe upper respiratory signs, but both FCV isolates caused oral ulcers and lameness. However, oral ulcers were more prevalent, and lameness and depression were more pronounced in animals infected with strain 2280. A decrease in lymphocyte counts was also noted only in kittens infected with strain 2280. Kittens with signs of lameness also had increased blood levels of the alpha-1-acid glycoprotein (a₁-AG), which is used to monitor arthritis in humans [39].In one study, neither intra-nasal infection with the respiratory strain A4 nor subcutaneous inoculation with the vaccine strain F9 was able to induce lameness in unvaccinated cats or in cats vaccinated twice with strain Webster 2113, with the second dose administered 1 month before starting the experiment. Thickened synovial membranes and minimal histological changes were observed only in three cats, whilst the FCV antigen was identified in 14 joints of five cats inoculated with either the attenuated strain F9 or the field strain A4 7 days before euthanasia, but not in animals inoculated 4 days before euthanasia, although the virus could not be isolated from any joints [15]. These findings, whilst confirming the possibility of the localization of FCV in the joints, also demonstrated that the limping pathotype relies on intrinsic characteristics of the FCV strains.Although the molecular bases of FCV pathotypes have been deciphered, attempts have been made to identify hallmarks able to differentiate VSD from classical respiratory FCVs. A comparison of the capsid region E sequence, which interacts with cell receptors, has identified seven key residues (at positions 438, 440, 448, 453, 455, 465, and 492) that are seemingly significant for pathotype differentiation [22]. When using MCA with Brunet’s criteria, strain 460.20/ITA/2020 showed residue properties typically found in respiratory FCVs (Figure 4). A similar pattern was also observed for other FCV strains associated with limping syndrome, i.e., 1466-1, 2280, F65, and LLK. A limit of this MCA analysis [22] is that it was originally conceived with the specific purpose of differentiating hypervirulent FCVs from classical ORD strains. Therefore, additional residues potentially involved with other pathotypes were not investigated. Accordingly, MCA was tentatively performed, considering strains with non-VSD and non-ORD pathotypes, i.e., enteric, limping, and vaccine strains. In this analysis, 4 aa residues were found to be significantly associated with the limping pathotype. This finding deserves further evaluation using a wider collection of sequences of limping strains.Additionally, recent studies based on in vitro recombination have revealed that replacing the ORF3 (p30) of strain 2280 with the p30 of strain F9 can affect virulence in vivo. This has been related to the ability of strain 2280 p30 to downregulate the expression of Interferon Alpha and Beta subunit 1 (IFNAR1) and inhibit the phosphorylation of STAT1 and STAT2 in the JAK-STAT signal transduction pathway [40]. Accordingly, genetic tracts outside the ORF2 of FCV could play a role in the pathogenesis of FCV and influence pathotype expression.Likewise, in the phylogenetic analysis based on the full-genome nt sequence, on the aa sequences of the capsid and the hypervariable region E, various limping-associated FCV strains were clustered in polyphyletic clades. Noteworthily, however, in the capsid-based phylogenies, our limping-associated FCV strains clustered in a group that also included limping-associated FCV strains from Northern Italy isolated in 2017 [23] and the lame-associated strain 2280 isolated in 1982 [13]. This could suggest a common ancestry and retention of the lameness phenotype among these FCV strains, although the massive genetic heterogeneity and lack of precise metadata for some FCV sequences available in the databases evidently challenge this interpretation of the data. For instance, in our case, lameness was the prominent clinical sign, but fever and oral lesions were also present and cat-1 was initially presented to the hospital seeking orthopedical intervention. In other cases, however, lameness could be perceived as a minor clinical sign and the clinical picture could be attributed to respiratory or systemic forms, thus losing relevant metadata.5. ConclusionsIn conclusion, we identified a small outbreak of FCV infection with a predominant limping phenotype in unvaccinated cats. In the outbreak, we observed a transmission chain and, based on genome sequencing, the animal-to-animal transmission likely occurred indirectly, with the virus shed by the respiratory tract via contaminated fomites in the household. These findings, whilst reinforcing the notion that the limping syndrome is a reproducible phenotype peculiar to some FCV strains, also stress the need for the adoption of adequate prophylaxis measures to prevent the transmission of highly transmissible infectious diseases. Since vaccines for FCVs are available, the immunization of susceptible cats should always be considered a priority.

animals : an open access journal from mdpi

10.3390/ani11071847

PMC8300397

Sea stars are iconic marine invertebrates and are important for maintaining the biodiversity in their ecosystems. As humans, we interact with sea stars when they are used as research animals or displayed at public or private aquaria. Molecular research requires fresh tissues that have thus far been considered to be of the best quality if collected without euthanasia. This is the first paper describing a method to euthanize sea stars that still allows for sampling of high-quality tissue that can be used for advanced research. Since it can be difficult to tell if an invertebrate has died, it is important to use a two-step method where the first step makes it non-responsive and the next step ensures it has died. Sea stars were placed in a solution of magnesium chloride until they no longer reacted to having their underside or oral surface tapped. Sea stars did not seemingly find the magnesium chloride solution unpleasant. After they were no longer reactive, the sea stars were immediately sampled. Tissue from their digestive glands was extracted and interpreted to be of adequate quality for molecular research techniques. This study promotes animal welfare while maintaining high-quality tissue sampling for molecular research.

Sea stars in research are often lethally sampled without available methodology to render them insensible prior to sampling due to concerns over sufficient sample quality for applied molecular techniques. The objectives of this study were to describe an inexpensive and effective two-step euthanasia method for adult common sea stars (Asterias rubens) and to demonstrate that high-quality RNA samples for further use in downstream molecular analyses can be obtained from pyloric ceca of MgCl2-immersed sea stars. Adult common sea stars (n = 15) were immersed in a 75 g/L magnesium chloride solution until they were no longer reactive to having their oral surface tapped with forceps (mean: 4 min, range 2–7 min), left immersed for an additional minute, and then sampled with sharp scissors. RNA from pyloric ceca (n = 10) was isolated using a liquid–liquid method, then samples were treated with DNase and analyzed for evaluation of RNA integrity number (RIN) for assessment of the quantity and purity of intact RNA. Aversive reactions to magnesium chloride solution were not observed and no sea stars regained spontaneous movement or reacted to sampling. The calculated RIN ranged from 7.3–9.8, demonstrating that the combination of animal welfare via the use of anesthesia and sampling for advanced molecular techniques is possible using this low-cost technique.

1. IntroductionSea stars are iconic marine invertebrates and are important for maintaining biodiversity in their ecosystems [1,2,3]. Sea stars are commonly kept under human care for research and for display at public and home aquaria. Sea stars are appealing as research subjects as they possess many unique traits, including impressive regenerative capabilities [4,5]; mutable collagenous tissue capable of rapid nervously mediated changes in tensile strength, which is of great interest to the biomedical industry [6]; adhesives from tube feet [7,8]; and deuterostome-type development similar to vertebrates that makes them a valuable invertebrate model for embryonic development [9,10]. Common sea stars (Asterias rubens Linnaeus, 1758) have been used extensively in research as they are easy to collect due to their abundance in the eastern and western northern Atlantic Ocean; they are commonly found from the intertidal zone down to 650 m, and adapt well to life under human care [11,12]. Sea stars may need to be killed due to the deterioration of their health, overpopulation, their status as an invasive or damaging species in some regions, or for research endeavors. The collection of sea star samples in research often requires lethal sampling due to the amount of tissue required for analyses and thus the invasive nature of sample collection. Due to concerns over adequate tissue quality, sea stars used in research are currently often lethally sampled without any prior method rendering them insensible.The terminology used to describe the act of ending an animal’s life is based on the application of the activity. Euthanasia is used to describe the act of ending the life of an animal using a method that minimizes or eliminates pain and distress. Slaughter is the act of killing animals to harvest them for consumption. Killing is typically used to denote ending an animal’s life in a way that minimizes distress but is not being performed to end an animal’s suffering, as in true euthanasia [13]. The term euthanasia applies to the work presented herein since the goal was to minimize harm to the animal during sampling. For a method to be considered euthanasia, it should achieve rapid unconsciousness and death, minimize stress, and be reliable, reproducible, and irreversible [14]. It is important to note that methods that do not cause rapid death or that result in trauma prior to loss of consciousness cannot be considered euthanasia. Unacceptable methods for invertebrates include removal from water to die by desiccation (i.e., “dewatering”), freezing, or immersion in caustic chemicals (i.e., placing directly in tissue fixative) as a solo or first step [13].One challenge with ensuring the humane death of sea stars is verification of death. Methods used for verification of death in vertebrate animals include auscultation, electrocardiogram, Doppler ultrasound, and pulse oximetry. However, these techniques are not feasible in echinoderms which do not have pumping structures detectable outside the organism. Methods that can be useful in invertebrate species include a lack of movement, loss of response to stimuli, and flaccidity [13]. In echinoderms, permanent loss of response to stimuli and lack of tube feet suction to substrate can be considered as markers that the animal is insensible and perhaps dead; however, these overlap with anesthesia in echinoderms [15]. To confirm death, a secondary technique should be used that is unsurvivable. The American Veterinary Medical Association Guidelines for the Euthanasia of Animals 2020 Edition recommends a two-step approach for euthanasia of aquatic invertebrates where the first step results in chemical induction of non-responsiveness and is followed by a second step that destroys the brain or major ganglia physically or chemically. The first step can include immersion in magnesium salts, clove oil, eugenol, or ethanol followed by the second step of immersion in 70% alcohol or 10% formalin, or physical methods including pithing, freezing, and boiling [13]. Recommendations for euthanasia of echinoderms include immersion in 7.5% to 8% magnesium chloride (MgCl2) or buffered MS-222 (tricaine methanesulfonate) at 1–10 g/L [16]. Immersion in 7–8% MgCl2 has been reported in sea stars of various life stages for relaxation prior to sampling for evaluation of gross anatomy and histology [17,18,19,20,21,22,23]; however, the process has not been described in detail and none has been performed for molecular techniques.Pyloric ceca, also known as digestive glands, are where absorption and storage of nutrients occur in sea stars [24]. Common sea stars undergo a yearly cycle where the pyloric ceca expand and store glycogen during the non-reproductive time in the summer to fall, then lyse to provide nutrients to support gonadal growth over the winter to spring as the gonads are active [25]. Evaluating the RNA in pyloric ceca can inform about toxins [26] and environmental stressors [27] and provide reference genes [28]. To determine RNA quality via the RNA integrity number (RIN), the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) uses microfluidic technology, combining traditional gel electrophoresis and laser-induced fluorescence detection in the same platform. RIN scores vary from 1–10; samples with a RIN of 10 have no RNA degradation while samples with a RIN of 1 are highly degraded [29,30,31,32,33]. The assessment of RIN is considered an essential step for the evaluation of quantity and purity of RNA before using advanced molecular techniques (e.g., RT-PCR).The objectives of this study were to (1) describe an inexpensive and effective two-step euthanasia method for adult common sea stars and to (2) demonstrate that high-quality RNA samples useful for molecular diagnostics can be obtained from the pyloric ceca of MgCl2-immersed sea stars. The two-step euthanasia technique reported here is immersion in 75 g/L (7.5%) MgCl2 until the sea star is non-responsive followed by a physical method of quick dissection.2. Materials and Methods2.1. Sea Star Collection and HusbandrySea stars are not currently covered by research oversight guidelines in the United States and thus no Institutional Animal Care and Use Committee (IACUC) review was performed. Sea stars were handled and housed in accordance with guidelines for aquatic species published in the Guide for the Care and Use of Laboratory Animals, 8th edition [34].Adult common sea stars (Asterias rubens) (n = 15, 8 females, 7 males) were purchased from a commercial collector (Ocean Resources Incorporated, Sedgwick, ME, USA) in the subtidal zone off the coast of Sedgwick, Maine. Immediately following collection, the sea stars were packaged in perforated plastic bags and a small amount of water in a heavy-weight plastic bag for overnight shipment to the University of Florida. On arrival, sea stars were visually examined, photographed, weighed (g), and contoured diameter (cm) was measured. Contoured diameter was defined as the distance in cm from the longest tip of one ray across the central disc to the tip of the ray directly across using a flexible tape measure across the contoured aboral surface of the sea star at rest. Sea stars were slowly acclimated to temperature and water chemistry over 6 h with ~10% water addition per hour then group housed in a static renewal 454 L (120 gallon) round fiberglass tank in artificial seawater (Instant Ocean, Blacksburg, VA, USA) mixed according to the manufacturer’s specifications. A large air stone connected to compressed air was provided in the center of the tank for aeration. Sea stars were housed in a facility with an air temperature of 25 °C and 16:8 photoperiod automatically maintained with overhead fluorescent lighting on a timer. Sea stars were housed and euthanized in water approximating the temperature of the Atlantic Ocean near the location of sea star collection as reported by buoy I01 of the Northeastern Regional Association of Coastal and Ocean Observing Systems (NERACOOS, http://www.neracoos.org/, accessed on 26 March 2021).Water quality was tested daily for ammonia, nitrite, and nitrate with a portable colorimeter (DR 900, Hach, Loveland, CO, USA), alkalinity with a drip test kit (Salifert Worldwide, Duiven, Holland), salinity with a refractometer (model RHS-10ATC, Aquatic Eco-Systems Incorporated, Apopoka, FL, USA), and pH and dissolved oxygen with a handheld meter (ProDSS, YSI Incorporated, Yellow Springs, OH, USA). Water temperature was monitored with a max–min thermometer reset daily (Brannan Thermometers & Instrumentation, Cumbria, UK). Water quality was maintained from (min–max) 0.00–0.54 mg/L for ammonia, 0.000–0.027 mg/L for nitrite, 0.0–1.2 mg/L for nitrate, 10.2–10.4 dKh for alkalinity, 33–34 ppt for salinity, 8.09–8.13 for pH, 9.06–10.49 mg/L for dissolved oxygen, and 12.0–12.9 °C for temperature. A 10% water change was performed at least every 72 h with a 25% water change performed as indicated when ammonia or nitrite were above thresholds (ammonia >0.25 mg/L and nitrite >0.1 mg/L).Sea stars were each fed one frozen clam (Ocean Nutrition, Newark, CA, USA) every 48–72 h. Sea stars were maintained under the laboratory conditions described above and observed daily for two weeks before sampling.2.2. Sea Star Euthanasia and SamplingSea star sampling was completed over two days. Each individual sea star was visually examined immediately prior to sampling and was included in the study if it had fed since arrival, had no evidence of visible lesions, and was responsive to external stimuli (e.g., tapping the ambulacral groove with forceps or moving away from conspecifics in the tank). Sea stars were weighed and contoured diameter was measured a second time at time of sampling. A baseline response to stimuli was noted for each individual by tapping the ambulacral groove with a pair of forceps which resulted in tube foot movement and ambulacral groove closure. Sea stars were placed in a 7.5% MgCl2 solution of 75 g magnesium chloride hexahydrate (Fisher Scientific, Waltham, MA, USA) dissolved in 1 L de-ionized water chilled to 12 °C. The approximate cost of 75 g of MgCl2 is USD 5. Response to tactile stimulation was measured once per minute by quickly removing the sea star from the water and tapping the ambulacral groove with forceps. Sea stars were immediately re-immersed if they were still responsive. Behavior of the sea stars was assessed by continuously observing the sea stars during immersion. Possible aversive reactions to MgCl2 based on clinical evaluation, including mucus production, arm curling, stomach eversion, and escape behaviors, were considered for documentation if observed. The time from initial submersion to lack of response to tactile stimulation was recorded. Sea stars were left in the MgCl2 solution for 1 min past complete loss of response to tactile stimulation for the purpose of going beyond anesthetic effects of MgCl2 immersion. Once sea stars were rendered insensible, they were immediately dissected with sharp scissors at the junction of the aboral and oral body walls. Samples of cardiac stomach, pyloric ceca, gonad, body wall, and tube feet were removed, cut into small pieces (3–4 mm), and immediately flash frozen in liquid nitrogen. The aboral body wall was sampled with a 6 mm punch biopsy (Miltex®, Integra, Princeton, NJ, USA). All samples were stored in sterile cryogenic vials with no additives (Corning Incorporated, Corning, NY, USA) at −80 °C until processing. RNA was extracted within 18 days of sampling and stored at −80 °C until analysis. Extracted RNA samples were analyzed within 166 days of sampling, which is well within the time frame of known stability for RNA integrity for up to 15 years if frozen at −80 °C [35].To determine the sex of each individual sea star, a squash preparation of the gonad was made at the time of sampling. The slides were air dried, stained with Wright–Giemsa (Harleco; EMD Millipore, Billerica, MA, USA), and evaluated via light microscopy.Samples of artificial seawater from the sea star tank and 75 g/L MgCl2 solution were analyzed for ion composition and osmolality. Chloride was measured using Standard Methods 4110 [36] with a Dionex System (ThermoFisher Scientific, Waltham, MA, USA). Cations (Na+, Mg2+, Ca2+, K+) were measured using EPA method 6010D [37] with an Ametek SpectroBlue ICP-OES (Spectro Analytical, Kleve, Germany). Osmolality was measured using a vapor pressure osmometer (Wescor Inc., Logan, UT, USA).2.3. RNA ExtractionTen individuals were randomly chosen with a random number generator for RNA extraction due to costs and time limitations associated with sample preparation and quality analysis. Total RNA was extracted from pyloric ceca with RNA Stat-60 reagent (Tel-test, Friendswood, TX, USA) as previously described [38]. Briefly, 0.1 g of frozen tissue was homogenized in Stat-60 with a tissue grinder. Chloroform was added and the mixture was centrifuged at 20,000× g for 15 min at 4 °C. This extraction was repeated with the upper aqueous phase. Following the second centrifugation, total RNA was precipitated overnight with isopropanol. After centrifugation, the pellet was washed with 75% ethanol, air dried, and resuspended in 20 µL of RNAsecure™ (ThermoFisher Scientific, Waltham, MA, USA) following the manufacturer’s protocol to inactivate RNases. The concentration of nucleic acid in each sample was determined using a NanoDrop microvolume spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA). Samples were diluted with RNAsecure™ to bring the RNA concentration to <500 ng/μL, then treated using the TURBO DNA-free™ Kit (ThermoFisher Scientific, Waltham, MA, USA) per the manufacturer’s protocol to remove DNA contamination. Total RNA quality was evaluated using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) and RNA 6000 Nano Kit per the manufacturer’s protocol.2.4. Statistical AnalysesStatistical analyses were conducted in R (version 4.0.4, R Foundation for Statistical Computing, Vienna, Austria) using the RStudio integrated development environment (version 1.4.1106, RStudio Public Benefit Corporation, Boston, MA, USA). A significance value (α) of p < 0.05 was used to determine statistical significance in all tests. For all analyses, a non-parametric test was performed as the data did not meet the assumptions for parametric analyses. Kruskal–Wallis H-tests were performed using the Kruskal.test function in the stats package to evaluate differences in size and weight between male and female stars at sampling. A Kruskal–Wallis H-test was also performed to determine if there was a difference in time to lack of response to stimuli between male and female sea stars. To evaluate if size had an impact on time to lack of response to stimuli, sea stars were divided into two groups: above the median diameter (>14.2 cm) and the median diameter and below (≤14.2 cm), then a Kruskal–Wallis H-test was used to determine if there was a difference between the groups.3. Results3.1. Study AnimalsOn arrival, sea star diameter ranged from 10.2–14.6 cm (mean: 12.5 cm) and weight ranged from 32.9–69.2 g (mean: 51.7 g) (see Table 1). All sea stars were acclimated for 16–17 days. During this time, sea stars appeared clinically normal, behaved normally, and were observed to feed at least once prior to sampling. All 15 sea stars met the inclusion criteria.At the time of sampling (16–17 days after arrival), sea star diameter ranged from 10.3–16.0 cm (mean: 13.9 cm) and weight ranged from 39.4–77.2 g (mean: 59.7 g). There were seven male and eight female sea stars sampled. There were no significant differences in diameter (p = 0.30) or weight (p = 0.06) between male and female sea stars at sampling.3.2. Sea Star EuthanasiaNo sea stars exhibited aversive reactions to immersion in MgCl2. The ambulacral groove of sea stars remained open during immersion, but tube feet consistently lost response to tactile stimulation. The mean time to lack of response to stimuli was 4 min (range: 2–7 min). No sea stars regained spontaneous movement or response to stimuli during sampling. It took approximately 5 min to completely sample each sea star after removal from the MgCl2 solution. Male and female sea stars did not show a significant difference in time to lack of response to stimuli (p = 0.31). There was no significant difference in time to lack of response to stimuli between larger (diameter >14.2 cm) or smaller (diameter ≤14.2 cm) sea stars (p = 0.81).The ion composition and osmolality of artificial seawater from the sea stars’ tank and MgCl2 solution are reported in Table 2.3.3. RNA QualityTotal RNA was extracted from 10 randomly selected pyloric cecum samples. The RIN ranged from 6.8 to 9.8 (mean: 8.7). A gel-like image and electropherogram of intact total RNA are shown in Figure 1. Sampling and RIN data for all sea stars are provided in Table 2.4. DiscussionThis study provides proof of concept that the application of animal welfare considerations to molecular research with sea stars is possible. The proposed two-step method of euthanasia with immersion in MgCl2 followed by a physical method should quickly render the animal insensible without suffering or distress.The quick relaxation of sea stars after immersion in MgCl2 and lack of adverse reactions or regaining responsiveness indicate that immersion in 75 g/L MgCl2 is an acceptable first step in the euthanasia of common sea stars. If samples are not being collected for molecular techniques, sea stars could be left in the MgCl2 solution for a longer period of time as immersion for longer than 30 min is recommended in the American Veterinary Medical Association Guidelines for the Euthanasia of Animals 2020 Edition [13]. To minimize stress, sea stars should be euthanized in a solution with a temperature similar to the temperature they are housed at. The osmolality of the tank water and MgCl2 solution were similar (809.0 and 802.0 mmol/kg, respectively) which would also minimize potential stress for immersed animals. Unsurprisingly, the concentration of chloride and magnesium were higher in the MgCl2 solution than the tank water. Potential stress effects on sea stars resulting from these overt biochemical differences are unknown.Magnesium chloride has been used to anesthetize or “narcotize” invertebrates for decades [39]. Many species of invertebrates have been anesthetized, which would also serve as the first step of euthanasia, with MgCl2, including scallops [40], oysters [41,42,43], queen conch [44], and sea urchins [45]. Magnesium chloride as the first step of euthanasia, followed by destruction of the central nervous system, has been reported for cephalopod mollusks [46]. Immersion in MgCl2 (142.25 g/L) was effective as the sole step for euthanasia in jellyfish [47]. To euthanize sea stars, immersion in MgCl2 has been followed by immersion in 5–10% buffered formalin [20,23]. A similar euthanasia technique of immersion in MgCl2 followed by dissection has been performed; however, those samples were then placed into fixative for histology [17,21], put into 1% glutaraldehyde to photograph and measure [19], or weighed to determine seasonal cycles in pyloric ceca and gonads [22]. One of the limitations of this study is that an air bubbler was not used in the MgCl2 tank. This likely did not have an effect on the sea stars in this study as they were immersed for a short time and sea stars can tolerate hypoxia. The median sublethal oxygen concentration for echinoderms was reported as 1.22 mg O2/L [48] and they can survive for several days at 0.2 mg O2/L [49]. However, it is recommended to use an air bubbler for all anesthetic and euthanasia solutions for aquatic animals since hypoxia is not an approved euthanasia method [13].The mechanism of MgCl2 in invertebrates is unknown, although it is thought to act on the post-synaptic membrane at the neuromuscular junction [50]. Increased magnesium inhibits acetylcholine, inhibiting post-synaptic potentials and causing decreased muscle fiber membrane excitability [51]. In vertebrates, MgCl2 enhances the effects of non-depolarizing muscle relaxants [52]. Historically, there has been a debate on if MgCl2 blocks nerve transmission and neurotransmitter release or if it acts solely as a neuromuscular blocking agent [53]. Recent literature has shown that the former appears to be true; cephalopod mollusks immersed in MgCl2 showed loss of both afferent and efferent signals which indicates the animals had true anesthesia and would not feel subsequent procedures or sampling [46]. There have been no investigations on the impact of MgCl2 on neurotransmission in adult echinoderms to date.The second step in a two-step method of euthanasia should ensure a rapid, non-reversible death. The secondary method described herein is a physical method of quick dissection. Other acceptable secondary methods include immersion in 70% alcohol, immersion in 10% formalin or physical methods including pithing, freezing, and boiling [13]. The secondary step can be altered depending on the ultimate goal for sample analyses. For example, if histology is desired, placement in 10% formalin as a second step would facilitate both killing the sea star and sample fixation. However, formalin results in crosslinks which chemically modify RNA and preclude isolation and evaluation of RNA [54]. When possible, the second step in a two-step method of euthanasia should result in destruction of the central nervous system. The echinoderm circumoral nerve ring and radial nerve cords are considered the central nervous system [55,56]. However, there is no evidence that these are coordinating structures, rather the circumoral nerve ring around the mouth appears to provide connections between radial nerve cords which extend down the arms of sea stars [57]. Common sea stars do not have nervous tissue that is visible with the naked eye which precludes the ability to pith them. However, our thorough sampling method presumptively resulted in the rapid destruction of the majority of nervous tissue. In invertebrate species with a visible nervous system center (e.g., decapod crustaceans and cephalopod mollusks), the authors recommend rapid destruction of the central nervous system in the secondary step to ensure nervous transmission ceases quickly.When assessing methods appropriate for euthanasia of invertebrates, it is important to bear in mind that invertebrates are a highly diverse group of animals which make up more than 95% of the species on Earth [58]. Aquatic invertebrates have a wide range of nervous system complexity, from sponges which have no true nervous tissue but are able to respond to stimuli [59] to cephalopod mollusks which are arguably the most complex and have more than half a billion neurons [60]. One of the most common euthanasia methods in use is an overdose of anesthetics. Anesthetics commonly used with aquatic invertebrates include buffered MS-222 (tricaine methanesulfonate), ethanol and MgCl2 [61]. However, it is important to ensure the animal does not find the anesthetic immersion aversive. Recent evidence has shown that some animals, including zebrafish and sea snails, may find immersion in MS-222, even buffered, to be noxious [62,63]. This appears to be true for sea stars as well; small spine sea stars (Echinaster spinulosus) displayed behavioral indicators that they found immersion in 0.8 g/L buffered MS-222 aversive, including mucus production, arm curling, and stomach eversion (unpublished data).Invertebrate welfare has received increasing attention following the inclusion of cephalopod mollusks in the European Union Directive 2010/63/EU in 2010 [64,65]. At the forefront of the conversation is the ability of invertebrates to feel pain [66,67]. Although there is mounting evidence for invertebrate pain perception, it is not unanimously agreed upon in the scientific and veterinary communities. For the purpose of considering animal welfare, the authors recommend following the precautionary principle and ensuring that care is taken to prevent potential pain and suffering in invertebrate species during euthanasia and sampling processes. To maintain the social license to use invertebrates in research, the ethics must also be considered. The two largest ethical dilemmas identified in invertebrate research by Drinkwater et al. [68] are collection of individual animals and euthanasia practices. The current study represents a small step towards greater inclusion of additional invertebrate species into consideration in animal welfare regulations by building up data on humane euthanasia practices.The Agilent Bioanalyzer reports RNA quality as a RIN score on a scale from highly degraded (RIN = 1) to completely intact (RIN = 10) [29,30,31,32,33]. There is currently no consensus on a RIN cutoff for sufficient RNA quality in a given sample, with proposed numbers ranging from 3.95 [69] to 8 [70]. The most commonly reported acceptable RIN is 5 or greater. This number has been found to be acceptable as a basis for downstream applications such RT-PCR [71,72] and gene expression analysis [73]. However, the RIN was not a good predictor of microarray performance [74]. All samples in this study had a minimum RIN > 5 (minimum 6.8) which indicates these samples would likely be useful for many analyses. Pyloric ceca were chosen as the tissue for extraction as they were the most abundant tissue collected and is a common target for RNA isolation in sea stars [26,27,75]. We were able to extract high-quality RNA (RIN > 8) from eight out of 10 pyloric cecum samples. The two samples with RIN < 8 initially did not have a RIN calculated by the Agilent Bioanalyzer due to critical errors in the instrument. The anomaly threshold for these errors were manually adjusted per the Agilent Bioanalyzer Troubleshooting Guide [76] and a RIN was calculated. It is unlikely that immersion in the MgCl2 solution resulted in a lower RIN since these two individuals were immersed for 3 and 4 min, which was below and at the median time of 4 min, respectively. Factors which could have influenced the RIN include pre-analytical variables such as sampling time and method as well as RNA extraction technique. A single person performed all RNA extractions to minimize variation, however, errors can occur in the extraction process. One limitation of this study is that the isolated RNA was not used for downstream analyses, such as qPCR, microarrays, or transcriptomics, for further assessment of RNA quality. However, given the high quality of RNA based on RIN, it can be presumed that these samples would be useful for a variety of molecular analyses. Samples from sea stars immersed in MgCl2 may also be useful for techniques not involving nucleotides, such as metabolomics. Jellyfish euthanized in MgCl2 provided useful samples for NMR-based metabolomics [47]. To further determine if MgCl2 immersion had an effect on RNA quality, paired molecular tests (i.e., qPCR or transcriptomics) between immersed and non-immersed sea stars should be performed. Since RIN values provide an assessment of total RNA integrity, further studies will be necessary to address potential effects of anesthetic procedures on gene expression profiles.5. ConclusionsThe two-step euthanasia method described herein was quick, inexpensive, and effective, and allowed for the collection of high-quality RNA samples. The authors recommend immersion of sea stars in 75 g/L MgCl2 as the first step of euthanasia as sea stars were quickly rendered insensible. The secondary step of euthanasia may vary depending on the goals of euthanasia and sampling; quick dissection with sharp scissors, as described here, was a physical method that allowed for sample collection and ensured death of the individual. This method provides a practical way to improve the animal welfare of a sea star species commonly used in research settings and provides a springboard for future studies. Future investigations could evaluate the efficacy of higher and lower MgCl2 concentrations and the optimal immersion time.

animals : an open access journal from mdpi

10.3390/ani13101604

PMC10215897

An extensive body of literature has been produced on the effects of transport-related factors on the welfare of adult pigs, while less attention has been paid to weaned piglets. Understanding how to safely transport weaned piglets is very important as they present their own set of challenges and risks. Despite this gap in the scientific literature produced to date, transporting piglets immediately after weaning is a common practice and every year millions of weaned piglets are transported on long journeys to reach the growth-finishing farms. The combination of the stress of weaning and transport can lead to negative effects on the health and welfare of piglets. The review will provide an in-depth discussion of the main factors that affect the welfare of weaned piglets during transport and will highlight the need for more research on key factors affecting the welfare of piglets during transport aimed at providing scientific evidence to the existing recommendations and/or regulations.

The purpose of this review is to present the best available scientific knowledge on key effects of pre-transport and transport factors influencing the response of piglets to transport stress and post-transport recovery. To date, research on piglet transportation particularly focused on the effects of season (i.e., heat and cold stress), vehicle design features (ventilation type and deck/compartment location), space allowance and transport duration, and piglet genetics. More specifically, in this review the effect of transport duration has been dealt with through its impact on death rate, behaviour and physiological response, and feeling of hunger and thirst. Based on the available literature, clear conclusions can be drawn on the vulnerability of piglets to heat stress during transport. Both short and long transportation have an effect on piglet welfare, with effects being biased by the genetic background, ambient conditions and vehicle design. Further studies investigating the impact of factors such as vehicle design, truck stocking density and environment, piglet genetic background, and weaning age are needed.

1. IntroductionPigs are transported for sanitary and market reasons throughout their lives, from weaning to slaughter [1]. Transport is a multi-factorial stressful event where a combination of factors, i.e., vehicle design, ambient conditions (temperature, humidity, vibrations, and noise), journey conditions (duration, space allowance, and driving quality), and handling at loading and unloading, rather than a single one is responsible for the animal’s well-being [2,3].Although the transport of pigs has largely been investigated over the past years, most studies have focused on finishing pigs destined for the abattoir, while less attention was addressed to newly weaned piglets and weaners [4,5]. The first transports which happen in pigs’ life actually take place soon after weaning, when newly weaned pigs are transferred for a short distance to nursery barns, while weaners (around 30 kg) are transported over short or long distances from the nursery barn to reach a different farm where they will be reared until they reach a market weight [1,6]. Newly weaned piglets are mainly transported for biosecurity reasons, with the aim of relocating them to a different site, reducing the risk of transmission of infectious agents. On the other hand, weaners are usually transported as sow herds, which are located in more remote, biosecure regions, while grow-to-finish barns are located near feed production and slaughter facilities [5]. In both pig types, transport immediately after weaning can exacerbate the negative effects of weaning stress due to prolonged fasting, as well as increasing the risk of diarrhea and delayed growth [7,8,9].Weaning and related practices represent a particularly stressful phase in the life of pigs [10] as they strongly affect the metabolic state of the animals, triggering in some cases disruptions in intestinal barrier function and permeability [10,11]. Disruption in the homeostasis of the intestinal mucosa can lead to the translocation of bacteria, antigens and toxins to other tissues, leading to states of inflammation, and, in severe cases, bacterial infections. Due to the diet changes and stress associated with this phase, weaning can also lead to energy, lipid and protein metabolism dysfunctions, affecting kidney and liver functions [12]. Weaning is a stressful phase from a social perspective too as the piglet is separated from the mother and littermates and mixed with unfamiliar individuals. This social stress can have effects on its behavioural development as well as increasing cortisol production, the main stress-induced glucocorticoid, with effects on innate and adaptive immunity [13]. In addition, piglets are growing animals with immature intestinal mucosa, which puts them at greater risk of developing disease states. This deadly cocktail of factors makes weaning one of the most critical periods in pig management.According to EU Council Directive 2008/120/EC [14], in the European Union, no piglets shall be weaned from the sow at less than 28 days of age, unless the welfare or health of the dam or piglets would otherwise be adversely affected. In practice, in most commercial pig farms, weaning is anticipated to be at three weeks of age. Intra-EU trade of piglets involves more than 15 million heads per year, with exports mainly from Denmark and The Netherlands and mostly heading to Germany, Poland, Belgium, Spain and Italy [15]. In the intra-EU trade, distances between the origins and destinations are generally within 24 h of travel as the countries exporting and importing piglets are predominantly those in central Europe, as well as Denmark, France and the northern regions of Spain and Italy [1]. Despite the relevance of this trade to the EU, studies on the effect of transport on piglets have been mainly carried out in Canada and USA, where piglets, also called Iso-Wean piglets, are weaned around 17 days after birth and the transport occurs immediately after weaning [7,8,9,16,17,18].Despite these differences between countries in age at weaning, transporting piglets immediately after weaning is a common practice. However, the combination of the stress of weaning and transport can lead to negative effects on the health and welfare of piglets [19]. The present review aims at analysing and discussing the main factors that affect the welfare of weaned piglets during transport and their post-transport recovery at the farm of destination.2. Materials and MethodsThe literature search was performed using Scopus® database (www.scopus.com) in November 2022 and then repeated in February 2023. The used search terms were (WEANERS OR “WEANED PIGLETS”) AND TRANSPORT AND WELFARE. Given the limited number of articles being found (only 13 peer-reviewed publications), no limitation was set in the time span of the search. Grey literature was also searched using Google Scholar database based on the same search terms. An additional assessment of bibliographical items resulted in further identification of relevant sources of information, which were then included among the list of documents of interest. The bibliographic material—mostly peer-reviewed resources and in some instances government publications—was restricted to primarily English language and non-English publications which included an English abstract, where relevant information could be retrieved. Once relevant publications were identified, they were filed and categorized for further evaluation. Figure 1 shows the number of studies and documents found and retained at each step of the literature review process.3. Before Departure from the Farm: Weaners’ Health and ManagementAnimals’ Health Status and Fitness for Transport: An Ounce of Prevention Is Worth A Pound of Cure!Quoting Grandin, “it is impossible to assure good animal welfare during transport if the animal is unfit” [20]. Fitness for transport is defined as the animal’s ability to withstand transportation without compromising its welfare [21]. A proper assessment of fitness for transport is essential to prevent suffering in animals and avoid deaths during transport or in the hours following the arrival [21]. Poor fitness for transport was the major source of the high DOA rate (0.3%) reported by Valkova et al. [22] based on the analysis of a 10-years retrospective database of piglet transports in the Czech Republic. EU Council Regulation (EC) No 1/2005 mandates that piglets must not be transported if they are new-born with yet unhealed navels, if they present severe wounds or prolapses, if they are unable to move or walk unaided and without pain, or are aged less than three weeks. In the latter case, piglets aged less than three weeks must only be transported less than 100 km [23]. Other health conditions that may make a piglet unfit for transport include the presence of respiratory distress [24]. Therefore, pigs showing lameness, hernias, injuries and wounds, abnormal discharges, diarrhea, and breathing difficulties need to be carefully inspected before being transported as, based on the severity of these states, their ability to cope with transport-related stressors may be impaired [24].Among these clinical signs, hernias (i.e., the common term used for umbilical outpouchings) are commonly observed in piglets [25]. Umbilical outpouchings can be a clinical sign of various abnormalities, which can range from the less serious (cysts or mild hernias) to more problematic states, such as abscesses and severe hernias that can become strangulated. Depending on the size, location, and underlying pathology, umbilical outpouching can be considered a problem that hinders the normal mobility of pigs and their ability to balance during transport practices [26]. In addition, vibrations and possible loss of balance during on-road transport could cause the rupture of abscesses and cysts, or in cases of serious hernias, their strangulation or rupture. An association between presence of hernias at departure from the farm and increased mortality (×1.3) has been reported in a survey of adult pigs transportation [27]. Unfortunately, there are no data on the incidence of hernias and related transport mortality in piglets.Depending on the breed and sex, umbilical hernia affects between 1.6 and 7.8% of piglets [28]. Increased incidence of hernias may be due to several genetic and environmental factors, including abnormal behavioural patterns among piglets, such as belly nosing and navel sucking [29]. Herniated animals may be an even more attractive target for the perpetrators of belly nosing and suckling, exposing such animals to increased risks during travel and in the days immediately following the start of weaning, when these behaviours tend to peak [30].Prolapses are mentioned in EU Council Regulation (EC) No 1/2005 as a factor that makes an animal unfit for transport [23]. However, there are no real indications on the size beyond which a hernia should be considered dangerous. According to the guidelines published by the Consortium of the Animal Transport Guides Project [24], pigs are unfit for transport if they show umbilical outpouchings wider than 15–20 cm and with sores. However, besides not having been validated for its effect on the welfare of pigs during transport by any controlled study, it is not specified whether this recommendation for maximum size applies to all pig ages and categories, or whether it is only applicable to adult pigs. Given the smaller size of piglets when compared with adult individuals, it is indeed reasonable to assume that the limits for severe umbilical outpouching size should be lower in piglets.Traumatic injuries and bruises may be among the factors that make a piglet unfit for transport, especially if they result in obvious blood loss or a state of pain during limb movement. Limb lesions seem to be quite frequent in piglets, with a high prevalence of limb abrasions and sole bruises and erosion resulting from their housing on hard concrete floors [31]. Valkova et al. [32] evaluated the prevalence of traumatic injuries on the carcasses of piglets sent to slaughter and reported a higher prevalence of limb injuries in piglets than finisher pigs (0.15 vs. 0.10%). However, this difference could also be related to the fact that piglets are not a pig category normally slaughtered in the Czech Republic [32,33], and therefore those sent to slaughter are culled animals that were excluded from the growth phase due to health problems and slower-than-normal growth performance. Although most injuries and bruises may not be serious, piglets with limb lesions may be less able to balance themselves during transport and thus more prone to slips and falls resulting in injuries and lesions that can get infected and possibly turn into abscesses and bursitis.Other clinical conditions that could adversely affect the welfare of piglets during transport include pathological states, such as diarrhea and weakness. Emaciated and weak piglets may not be able to survive stress, feed deprivation and dehydration that occur during the first hours of the weaning process and during transport. Apart from the information taken from recommended practices for adult animals [24], the lack of scientific evidence does not allow us to associate the incidence of transport-related losses with different body and health conditions and their monitoring at the departure from the farm in piglets. Therefore, it is reasonable to think that the mortality observed during the weaning stage may in some cases be the result of the additive effect of weaning and transport of likely unfit piglets.The genetic background can also contribute to the response to transport stress in piglets. A transport study showed that piglets heterozygous for the Ryanodine Receptor 1 (RYR1) or Halothane (HAL) gene (i.e., piglets displaying the TC genotype, alias HALNn, for the C1843T RYR1 mutation) presented a greater physiological response to transport, as shown by the increased blood albumin concentrations and total white blood cell and neutrophil counts, when compared with the CC homozygous (alias HALNN) [34].Although they are not science-based, the recommendations or regulations for the monitoring of fitness for transport must be also applied to piglets. Sick, herniated, or emaciated piglets should not be transported, but should be kept, when possible, in the herd of origin to be treated and left to heal before being shipped to the growing site.4. Factors Influencing Piglets’ Response to On-Road Transport4.1. Environmental Conditions during Transport4.1.1. Thermoregulation PrinciplesPigs are homeotherm mammals, and thus, like other homeotherm animals, produce several physiological, endocrinological, and behavioural responses to thermoregulate and maintain constant their core body temperature. A stable core temperature can be maintained only when heat production and heat loss are balanced [35]. Thermoregulation responses vary depending on how far the ambient, skin, and body core temperatures are from the animal’s biophysical requirements. Thermal balance is indeed dependent on a combination of these temperatures, and the animal’s tolerance to different ambient temperature ranges which varies by age, sex, and breed. Thermal homeokinesis is a steady state where the internal body temperature of a homeotherm animal is kept constant at the normal core temperature level with little additional energy expenditure [35]. Thermal homeokinesis is maintained when ambient temperatures and environmental parameters are within the range of Thermal Comfort Zone (TCZ). The concept of TCZ was first hypothesized based on the human perception of the thermal environment [35] but is now also applied to livestock species. Translated into animal welfare, the TCZ is the range of environmental parameters (mainly temperature) where the energetic and physiological efforts for thermoregulation are minimal and within which an animal expresses satisfaction with the thermal environment, and does not need to change its behaviour to cope with the environment [35,36,37]. Outside the TCZ, the animal starts to experience thermal discomfort, which drives thermal-related behaviours (e.g., huddling, posture adjustments, searching for shaded places, etc.) that anticipate autonomic thermoregulatory mechanisms [35,36,37]. The TCZ is comprised within a wider range of ambient temperatures, namely the Thermoneutral Zone (TNZ), which is defined as “the range of ambient temperature at which temperature regulation is achieved only by control of sensible (dry) heat loss, i.e., without regulatory changes in metabolic heat production or evaporative heat loss” [38]. The TNZ boundaries are represented by the Lower Critical Temperature (LCT and the Upper Critical Temperature (UCT). Temperatures (absolute or perceived) below the LCT lead to cold stress in homeotherm animals. Below LCT, metabolic heat production increases, as the animal attempts to keep body core temperature in an acceptable range using shivering (irregular frequent muscle contractions), vasoconstriction, and in some animal species also by activating brown adipose tissue catabolism [39]. When environmental temperature and humidity values are below the LCT and the physiological responses activated by the body are not able to maintain or restore an acceptable core temperature, the animal enters hypothermia. Hypothermia is associated with several organ failures, cardiovascular dysrhythmias, such as ventricular fibrillation and pulmonary edema. The central nervous system’s electrical activity is also noticeably diminished [40] and, when hypothermia is severe and prolonged, multiple organ failure and death may occur [41].Temperatures (absolute or perceived) above the UCT lead to heat stress in homeotherm animals. At this point, physiological, endocrinological and behavioural responses are activated to counteract the increase in core body temperature. The energy expenditure to activate these responses increases; thermoregulation is mainly sought with increased water evaporation from the body surface (thermal sweating) or the respiratory ways mucosa (thermal panting) [42]. These responses are often combined with other endocrinological and physiological processes aimed at maintaining a stable core body temperature. When heat stress is prolonged over time, animals experience health issues, infertility, decreased growth and production, decreased immune system efficiency, and cellular and mitochondrial oxidative damage [43]. When the body’s ability to thermoregulate becomes disrupted, overheating (hyperthermia) and, in the most serious cases, organ failure and death may occur [41].4.1.2. Piglet Thermal Needs during TransportEU Council Regulation (EC) No 1/2005 mandates that vehicle interior temperatures during animal transport must remain between 5 °C and 30 °C, with an acceptable variation of ±5 °C. It also states (Chapter 2) that transported animals must be protected from extreme temperatures and must not suffer [23]. However, the acceptable range of temperatures shown in this law is not sufficient to ensure the protection of all animals from suffering and extreme environmental conditions as it does not consider the differences in the perception of TNZ between animal species, determined by species categories, age, weight, sex, genetics and rusticity [44,45].In pigs, the science-based recommended upper limit of TNZ is considerably lower than that indicated by the current EU regulation. Depending on the pig categories, UCT appears to vary between 19 °C for boars and 30 °C for piglets (8–30 kg liveweight [26,44]). However, such ranges have been observed in pigs kept on farms, whereas very little work has been done to establish the TNZ ranges of weaners during transport.As mentioned in the previous paragraph, both temperatures below the LCT and above the UCT can jeopardize the welfare conditions of pigs. While in market and heavier pigs (such as sows and boars) high temperatures pose a serious welfare risk during transport, piglets have a better tolerance to higher temperatures, having a range of TNZ more shifted towards higher temperatures [26,44]. This is related to the fact that these young individuals still do not have a thick subcutaneous fat layer that can impair their heat-dissipating capacity, and have a greater surface-to-body mass ratio when compared to adult individuals [46,47]. Furthermore, unlike other animal species, piglets do not possess brown adipocytes [47,48], which are functional in preventing heat dissipation (non-shivering thermogenesis), and therefore their thermoregulation relies exclusively on huddling behaviours and shivering when exposed to cold temperatures [48].Most publications reporting the effects of ambient temperature on the behaviour and physiology during transport, and post-transport performance of piglets were produced between 2005 and 2014 [16,18,49,50,51,52,53]. In the study by Lewis et al. [53], early-weaned piglets (17 ± 1 days of age) were transported for 0 h, 6 h, 12 h or 24 h during three seasons, summer (22.1 °C to 30.3 °C), fall (3.4 °C to 12.8 °C) and winter (−2.8 °C to 3.2 °C). Ear skin and rectal temperatures during transit were higher during the summer than during fall or winter. No significant effect of season was found on ear skin temperature, weight loss or average daily gain during the first 7 days and on day 14 post-transport. However, piglets transported in winter had lower rectal temperatures during 12–24 h transports, suggesting that they were unable to compensate when temperatures inside the truck dropped for longer periods, such as in transports lasting between 12 and 24 h. The cold stress experienced by these piglets during the winter journeys was also confirmed by the reduced social interactions between conspecifics and the increased frequency of resting behaviour during the journey reported in a previous study [16]. In a similar study, Wamnes et al. [51] found that early-weaned pigs transported for 6, 12 or 24 h in winter (within vehicle temperature ranging from −4 °C to +14 °C) had greater weight loss and lower average daily gain than piglets transported in summer (within vehicle temperature ranging from 14 °C to 25 °C) during the first 8 days and on days 10, 12 and 14 post-transport. This increased weight loss was explained by the incremented energy expenditure and muscular activity (with shivering) performed to maintain homeostasis. However, even transport during summer appeared to affect the welfare of piglets as, during the first 4 days after transports of 6, 12 and 24 h, they spent more time lying and showed longer drinking bouts than piglets transported in winter or not transported [51,52]. These results may suggest that journeys longer than 6 h at temperatures around 25 °C may lead to exhaustion and dehydration in piglets, probably caused by an increased amount of energy spent to increase heat loss through panting. The effect of season on the behaviour, physiology and mortality rate of transported weaned piglets has also been investigated in other studies [6,8,18,49,50]. However, due to the experimental design applied in these studies, the effects of season were biased by the additive effect of transport duration.A recent study has also shown the effect of higher temperature-humidity index (THI) values (THI > 85) on behavioural and physiological parameters in piglets of about 25 kg liveweight (68 days old) transported in a double-decked, open/passively-ventilated vehicle [54]. A difference in THI values was recorded between decks, with the bottom deck presenting higher THI values compared with the top deck (>83 vs. <82). In this study, relative humidity (RH) more than temperature, which did not differ between decks, contributed to THI variation. Piglets transported in the bottom compartments (THI: approx. 86) showed higher mean rectal temperatures and respiratory rate, as well as greater blood cortisol concentrations, when compared with piglets transported in the top deck compartments. These signs of thermoregulation issues (starting from THI > 85) may be explained by the poor ventilation rate within the bottom deck resulting in the production of an oversaturated atmosphere (RH > 80%) in this vehicle location. Based on the results obtained under the experimental conditions of this study, a UCT of 30 °C with a maximum RH of 80% were finally recommended to ensure the thermal comfort of piglets during transport [54].The scientific literature regarding the effects of different temperature and RH values on the welfare of piglets during transport is particularly scant. Furthermore, since thermal needs are closely dependent on the weight and growth stage of the animals, the differences in weaning age of piglets between studies make the identification of TNZ values for this pig type even more complex. The almost total lack of comparative studies between different breeds and genetic lines also needs to be highlighted. It is indeed reasonable to think that some autochthonous breeds reared in geographical locations with tropical climates are more able to withstand higher temperatures and RH values. To the best of our knowledge, there is only one study that compared heat stress resistance (based on respiratory rate and rectal temperature) between Large White and Creole piglets (25 kg liveweight) [55]. Based on the results of this study, Creole piglets had higher UCT values compared with Large White subjects (UCT = up to 34 °C vs. 32 °C, respectively), which confirms the contribution of the genetic component to the heat stress resistance of pigs.4.2. Space AllowanceThe microclimate (temperature and RH) inside the vehicle is also closely dependent on other variables related to transport conditions, including loading density (kg/m2) or space allowance (m2/kg). Research has shown a positive relationship between the size of the load and the within truck temperature that could increase up to 7 °C for every two pigs added to the load [56]. New recommendations for loading density/space allowance calculated following the allometric equation (K-value based on space need for standing/sternal lying, and semi- and fully recumbent position) suggested by Petherick and Phillips [57] have been proposed for the transport of piglets, finishers and sows, gilts and boars in recent EFSA opinion [26].EU Council Regulation (EC) No 1/2005 [23] establishes that pigs transported by road must be provided with sufficient space to lie down or stand up in their natural position. Depending on the meteorological conditions, the floor surface area per animal may also be increased by a maximum of 20% [23].While the effect of space allowance on the welfare and carcass and meat quality of slaughter pigs has been largely investigated [3,58,59,60], very little is known about the effect of floor space in the truck on the welfare of transported piglets.The few available results would indicate a limited effect of space allowance on behaviour and the physiology of weaned piglets during transport. Results from physiological studies range from no effect on blood stress indicators (i.e., cortisol, aspartate aminotransferase, creatine kinase, and glucose concentrations), immune response (i.e., neutrophil chemotaxis and phagocytosis) and performance measures (i.e., body weight gain and lesion scores) [18,50,61], to increased lymphocyte-to-neutrophil ratio (NL) in piglets transported at lower space allowance (0.50 m2/pig vs. 0.06 and 0.07 m2/pig) during summer [49]. However, this latter effect recorded in one-hour transportation trials would appear to be more closely related to the combined effect of high ambient temperatures (ranging from 26.1 °C to 30.3 °C) and higher stocking density during transport on piglets’ heat stress condition. The effects of space allowance on piglet welfare during transport have been mostly shown in terms of changes of position during transport, with piglets transported at smaller space allowances (0.05 m2/pig vs. 0.06 and 0.07 m2/pig) standing/rearing more and lying less [18,49,50]. Increased standing/rearing or sitting behaviors may be considered as an indication of competition for space that may be prevented by applying a minimum space allowance of 0.06 m2/pig [49].The few available results show that, unlike adult pigs [3], overcrowding does not particularly affect the welfare of piglets during transport. The reason for this difference may be two-fold. First, in piglets, overcrowding-related increased temperatures may not result in as severe heat stress as in adult pigs, since younger pigs require higher environmental temperatures (higher UCT) to be in thermal comfort (see Section 4.1.2). Second, spatial proximity and lying in full body contact is a common behavior in transported young pigs, which tends to decrease as piglets grow [62].In agreement with these observations, the minimum floor space requirements suggested for piglets during transport by the recently published EFSA scientific opinion are 0.13, 0.20, and 0.26 m2/pig for piglets of 10, 20, and 30 kg live weight, respectively [26]. These minimum space allowance requirements are similar to those indicated for piglets transported by air in EU Council Regulation (EC) No 1/2005 [23].4.3. Transport DurationEU Council Regulation (EC) No 1/2005 [23] limits journey duration for pigs to a maximum of 8 h, except when several additional requirements for the vehicle and animal’s needs are met (i.e., vehicles have obtained a type II transporter authorisation). All pigs may be transported for a maximum period of 24 h, except for unweaned piglets, which after 9 h of travel must rest for at least one hour before travelling for a maximum of another 9 h. After this journey time, animals must be unloaded, fed and watered and be rested for at least 24 h. Travel duration has been identified as a priority welfare issue for a long time in Europe and North America. Based on the conclusions of a comprehensive literature review of swine transportation, both long and short transports may result in poor welfare outcomes for pigs [5].Over the years, the impact of transport duration on the welfare of piglets has been assessed through the study of its effects on mortality (dead-on-arrival, DOA) rate, behavioural and physiological responses, hunger, and dehydration/thirst.4.3.1. Effects on DOAsSimilarly to adult pigs, DOA rate has been used as an indicator of piglets’ welfare during transport, even if the available results do not allow the identification of hazards specifically responsible for it. Averós et al. [6] surveyed the factors affecting the number of weaned piglets found dead after commercial transport from different farms. Information related to 58,682 piglets during 109 journeys in different EU countries was collected at the end of each journey using questionnaires. Piglets had been weaned at 21 to 28 days of age and were transported at 85 to 100 days of age. Overall, the DOA rate was 0.07%, with dead piglets being reported in 13.8% of the loads. The duration of the journey, ranging from 0.3 to 69 h, and the mean outside temperature, ranging from 0 °C to 38 °C, showed a significant interactive effect, with a gradual increase in the predicted number of DOA with increasing journey duration and outside temperature. The provision of drinking water during the journey reduced the number of dead piglets by almost 2.5%, with an estimated mortality rate at 30 °C of about 3% in eight-hour journeys without drinking water available that dropped to about 0.15% in similar journeys where piglets were provided with drinking water [6]. In this study, an interactive effect on DOA rate in piglets was found between transport duration and vehicle design [6]. When piglets were transported for more than 8 h, mechanically ventilated vehicles exponentially decreased the risk of deaths, with an estimated reduced incidence of dead piglets (approx. 1%) compared with 24 h transports at an outside temperature of 30 °C using passively ventilated vehicles (>8%).Based on data obtained from 78 loads of weaned piglets (79,715 piglets of approx. 6 kg), including outside and within vehicle temperature, travel distance and time, stocking density, and DOA rate by compartment, Harmon et al. [63] reported an effect of the interaction compartment temperature (ranging from −5 °C to 27 °C) × travel time (average of 8.51 h, ranging from 3.4 to 12.3 h) on DOA rate (0.03%, ranging from 0 to 1.11%), which increased with increasing ambient temperature and transport time. However, a difference was observed in the mortality rate among vehicle compartments during the different seasons. In particular, the mortality rate tended to be higher in winter for the piglets transported in the lower vehicle compartments. The situation reversed during the hot season, when the mortality rate tended to be higher in the upper compartments, probably due to the higher internal temperatures [63]. However, the authors cautioned about the meaning of these results given the very small number of mortality events, with only 10 loads out of 78 (13%) presenting dead piglets on arrival.More recently, Golightly et al. [17] failed to find an association between mortality rate (0.06%) in weaned piglets undergoing shorter or longer duration (<3 h and >30 h, respectively) commercial transport in the summer.Based on the published research, it therefore appears evident that the DOA rate is mainly determined by the conditions under which piglets are transported more than the duration of the journey. In particular, high ambient temperatures (e.g., 30 °C) associated with the unavailability of drinking water during transport and the use of passively ventilated vehicles appears to be a major contributor to the increased DOA rate in piglets.4.3.2. Effects on Behavioural and Physiological ResponseLewis and Berry [16] have examined the effects of the season (summer, fall and winter) on the behaviour of early-weaned piglets during and immediately after transports of different duration (0 or control, 6 h, 12 h, or 24 h) and carried out without supplemental heat (in winter), feed and water. Overall, in transit piglets spent more time lying than standing (75.6 vs. 21.6% of the time). As transport time increased, the percentage of time piglets spent standing significantly decreased, while lying time increased. Resting behaviour, in terms of lying in lateral or sternal recumbency, whether asleep or awake, during transport increased with transport time from 59.8% (1–12 h) to 91.5% (13–24 h), while standing behavior decreased from 36% (first 12 h period) to 7.4% (second 12 h period). The authors speculated that the increased lying behaviour may have been associated with either fatigue or huddling behaviour. The latter assumption may be justified by the fact that this pattern was more defined in winter and fall implying cold as a causal factor. Furthermore, sitting behaviour was more common during the first 12 h of transport (2.8 %) than during the second 12 h of transport (0.3%). Fighting behaviour was infrequent during the first 6 h of transport, but increased significantly in summer compared with winter, which may indicate that the establishment of the dominance hierarchy may have been reduced or halted in the colder transport environment. In this study, it was also shown as an effect of transport itself on piglet fatigue on arrival at the farm, with higher levels of post-transport resting in transported (81.4%) compared to non-transported (control) piglets (77.5%). Such effect was exacerbated by the time in transit as the percentage of time spent resting on the day following the arrival increased after 6 h (80.2%), 12 h (82.2%) or 24 h (81.9%) of transport when compared to control piglets (77.5%).Magnani et al. [64] examined the effects of long transport (14 h) and ambient temperature and humidity conditions on behaviour and blood parameters of post-weaned piglets (8–12 kg liveweight, approx. 35 days of age). The journey time was divided into three periods of about 4 h each. Similar to Lewis and Berry [16], resting (lying on the belly or in the lateral position with or without head movements) increased with transport duration and, in particular, during the last period of the journey. The increased resting behaviour over time can be considered an indicator of the progressive habituation of piglets to the transport environment. However, this adaptation was not observed in the weaned piglets transported at THI > 72, which rather tended to keep a standing position during the journey, probably in an attempt to dissipate more body heat [64].Garcia et al. [65] compared unweaned piglets (control) with piglets undergoing weaning (two groups, with and without water and feed for 32 h) and newly weaned piglets transported for 32 h with or without the provision of feed and water. Piglet behaviour was recorded for all treatment groups during transport and 24 h after transport. During the 32 h journey, weaned piglets transported with or without the provision of feed and water spent 77% and 81% of their time lying, respectively, while the control group only spent 65% of their time lying. Untransported weaned piglets provided or not provided with feed and water spent 81% and 87% of their time in the lying posture, respectively. Immediately after transport, newly weaned piglets transported without access to feed and water spent less time lying (20%) compared to weaned piglets that were not transported with free access to feed and water (79% of those piglets were lying) [65]. Overall, the results of this study showed that transportation and weaning have a negative effect on behavior of pigs, especially when feed and water are not provided during transport [65,66].Both short and long transport duration have an impact on the physiological response of piglets to transportation. Averós et al. [34] studied the effect of journey duration (0.6 vs. 8.3 h) on stress levels as assessed by the analysis of blood creatine kinase (CK) and lacto-dehydrogenase (LDH) activities, both indicators of physical fatigue, in 136 weaned piglets transported from a nursery to a growing-finishing farm. In this study, blood CK and LDH activities increased, particularly after the short journeys. Similar results after transport were recorded by Golightly et al. [17] who found that piglets transported for a shorter duration (<3 vs. >30 h) presented increased blood CK levels, although the values were within the normal physiological range for pigs, and greater serum cortisol and N:L ratios.In a more recent study, Golightly et al. [67], when comparing two different weaning managements associated with transports of two different durations without food and water, observed that the group of piglets weaned 6 days before being transported for a long time (30 h) were hungrier and thirstier on arrival and lay down less in the pen during the first two post-transport days. In contrast, the other group composed of pigs weaned immediately before short transports (<3 h) lay down more upon arrival at the farm, but also showed more severe ear and body lesions, likely resulting from more intense episodes of aggression. This latter result might have been biased by the different handling of piglets on arrival at the farm, with piglets transported for shorter time being mixed in larger groups (25 pigs) than the other transport group, which was instead housed in smaller group pens (mostly 15 pigs each). There is evidence that fighting rate is greater in larger than smaller groups in pigs [68]. In addition, pigs subjected to short transport showed a lower growth rate during the first two post-transport days when compared to the group of pigs transported for longer time. Besides post-transport fighting activity, this effect could be due to the effect of short transport, whether combined or not with weaning stress. A number of studies showed the effects of shorter transportation on the fatigue condition of piglets on arrival at the farm due to lack of time to recover from the stress of handling at loading and settle and rest completely in the truck [16,34,69].The effects of longer journeys on blood parameters in piglets are not clear, ranging from no effect on blood CK and N:L ratios after 14 and 32 h, respectively [64,65], to greater blood CK activity in female piglets transported with the provision of feed and water for 32 h [65].4.3.3. Effects on HungerAccording to the EU Regulation 1/2005, piglets must be provided with only water during the road journey [23]. The fasting period before loading should be carefully planned in relation to the expected length of the journey, to avoid hunger, weight loss, and death [6,23,64]. The recommended fasting period for piglets before transport is 5 h [21]. However, this recommendation should be validated scientifically, as longer fasting periods (up to a maximum of 20/24 h) seem to decrease the risk of DOA in piglets [6].Berry and Lewis [7] ran two simulated transport trials to study the effects of journey duration (0 h or control, 6 h, 12 h, and 24 h) and ambient temperature (20 °C, 25 °C, 30 °C, and 35 °C) on the post-transport performance of early-weaned piglets. The interaction transport duration × temperature affected liveweight variation during the first 24 h after transport, with piglets either transported for 24 h at high transport temperatures (30 °C and 35 °C) or for 6 h at 20 °C and 35 °C presenting the greatest weight loss in comparison with control groups. Such effects were still present up to 7 days post-transport, indicating the difficulty of young animals to recover from transport stress.Other Canadian transport studies reported (1) greater weight loss and slower post-transport weight recovery after 20 min transport compared with 6 h transport, likely resulting from a reduced motivation to feed and drink following transport [51], (2) 7.1% weight loss after 24 h of transport with weight loss increasing with increasing transit time [52] and (3) lower body weight on arrival after >30 h compared with <3 h transport (5.6 vs. 6.2 kg), but this difference disappeared after 3–4 days (6.6 vs. 6.3 kg; [17]).Piglets transported for 32 h without access to feed and water have been reported to lose more body weight (approx. −9%) compared to non-transported piglets with feed and water access [65]. In a similar study, Garcia et al. [66] reported no significant losses in body weight of piglets transported up to 24 h, independently from the availability of feed and water. However, when the journey was extended to 32 h, piglets transported without feed and water or only with feed lost more weight compared to non-transported control piglets and piglets provided with feed and water or only water (−7.4 and −7.5% and −5.7 and −5.5%, respectively).Information concerning the variation of blood indicators of hunger in weaned piglets during transport is scarce. Studying the effect of short (0.6 h) and long (8.3 h) journey duration on weaned piglets, Averós et al. [34] found a significant decrease in serum glucose concentrations in pigs transported for 8.3 h vs. 0.6 h. Magnani et al. [64] also reported increased blood urea levels as a result of the prolonged fasting in weaned piglets transported for 14 h.Although feed withdrawal for more than 6 h is associated with decreased blood glucose levels and increased blood free fatty acids (FFA) and urea concentrations, a fasting period prior to transport is highly recommended to decrease the risk of DOA in piglets [6]. Furthermore, the reported results suggest that water withdrawal has a greater effect on weight loss and risk of DOA than feed withdrawal in transported piglets.4.3.4. Effects on Thirst and DehydrationThe EU Regulation 1/2005 clearly states that pigs may be transported for a maximum period of 24 h, provided they have continuous access to water [23] to prevent thirst and dehydration. Thirst is expressed by increased visits to the drinker, drinking bouts, water intake, and reduced latency to drink after a period of water deprivation. There is no evidence of piglets drinking during transports (14 h; [64]) but increased drinking behaviour in early-weaned piglets after 24 h transport has been reported [16,52].Elevated total plasma protein and albumin concentrations and hematocrit levels are indicators of dehydration as a result of transport and the associated feed and water deprivation [70]. In most studies, increased blood levels of total plasma protein, albumin and/or hematocrit percentage have been reported after long journeys (from 8.3 to >30 h; [7,17,34]). However, other studies either reported increased blood total plasma protein and albumin concentrations in weaned pigs after a short journey (60 min; [46]) or slight to no variation in blood hematocrit percentage and albumin in weaned piglets transported for 14 and 32 h, respectively, regardless of whether water and feed were provided [64,65].5. Environmental Enrichment in the TruckMany studies assessed the efficiency of within truck enrichment strategies to improve comfort and decrease stress during transport in market weight pigs [71,72,73]. In piglets, this assessment is only limited to a couple of studies. Roldán-Santiago et al. [74] evaluated several physiometabolic responses to stress in weaned piglets of different age (8, 15 and 22 days) transported at a space allowance of 1.2 m2/piglet for 1 h using trucks either with bedded floors or without. Bedding consisted of a 15 cm-thick layer of straw. Regardless of the age, piglets transported without straw bedding showed higher blood hematocrit and glucose concentrations, and decreased blood pH and pO2 upon arrival at the farm when compared with piglets transported on straw bedded truck floor. The positive effect of straw may not only be related to the increased comfort around resting, but also to its property as chewable material allowing the expression of rooting [75] and curbing fighting or other negative social interactions [76,77].A more recent study evaluated the potential positive effect of exposing piglets to music 5 days before transport and during a short transport (approx. 70 min trip) and providing chewable material (toys made with a plastic hose) and lavender aroma (sachets containing 20% lavender) on their skin surface temperature, respiratory rate and behaviour during transport [78]. Music (60 dB maximum sound pressure) consisted of vehicle sounds piglets could listen to at the farm in order to become familiar with the truck environment and pop-rock sound being played at the farm and during transport. Both toys and lavender sachets were hung along each truck compartment side. Overall, enriched piglets showed a lower respiratory rate before transportation compared to control ones. Piglets exposed to vehicle sounds and pop-rock music spent most of their time lying during transport, and were less aggressive and more playful during and after transport, respectively, when compared with piglets subjected to the other treatments. In contrast, an increased mounting behaviour was observed in piglets exposed to the toy and aroma treatments. This behaviour may have been indicative of their attempts to reach the toys and the lavender sachets. However, these results would deserve a further validation under more variable on-farm familiarization/training periods and transport conditions, in terms of duration and climate.Figure 2 summarizes the major factors affecting the welfare of piglets during transport, highlighting the consensus between studies and the gaps in knowledge arising from the existing literature review.6. ConclusionsAmong pig categories, piglets are the most transported pig type after slaughter pigs. More than 14 million piglets are transported from northern European countries to other EU member states every year [79]. In Canada, approximately 18 million weaned piglets are transported domestically (J. Clark, Canadian Pork Council, personal communication) as well as approximately 400,000 newly weaned piglets (<7 kg) have been exported from Western Canada alone to the USA in 2020 [80]. Despite this extended trading activity, it is surprising to acknowledge the considerable gap in the scientific literature on the effects of transport and related factors on piglets’ welfare and post-transport growth performances.Weaning is one of the most sensitive phases in pig farming as it represents a phase of change in both the diet and the environment in which newly-weaned piglets live. Since transport is combined with the beginning of this phase, it is reasonable to assume that at least part of the data concerning morbidity and mortality of pigs during and immediately after weaning may be due to the additive effect of these events. This association of stressors does not ease the evaluation of the contribution of feed and water deprivation to piglets’ response to weaning and transport stress. The different ages at weaning (from 17 to about 30 days of age), and therefore in liveweights, may help explain the significant discrepancy in the results between studies. Furthermore, although the results showing market weight response to transport stress, for which a much larger body of research has been carried out, cannot be representative of the interpretation of piglets’ welfare during transport, it can be generally concluded that, similarly to adult pigs, piglets are particularly vulnerable to heat stress and both long and short transport duration, with related effects being confounded by genetic background, fitness condition before transport, ambient conditions [4], transport vehicle used and lack of drinking water.Future studies should consider morbidity, postures, body temperature and mortality rates in piglets during transport and post-transport growth performance focusing on factors, such as vehicle design, within truck stocking density and environmental factors (i.e., temperature, RH, noxious gases concentrations, noise and vibrations), piglet genetic background, and weaning age. This information is vital to develop science-based recommendations for optimal minimum transport conditions ensuring the physical comfort of weaned pigs during transport. Furthermore, given the close relationship between stocking density and thermal needs of piglets during transport, the new EU recommendations based on calculated allometric equations for floor space density must be validated by scientific research in order to make sure they are adjusted to the piglets’ thermal needs under variable external and within truck environmental conditions during the journey.

animals : an open access journal from mdpi

10.3390/ani12020176

PMC8773241

There are several conditions and diseases considered painful to cattle. One reason for the inconsistency in pain recognition and thus pain relief in cattle is the inadequate ability to identify and assess pain. In fact, both increased and/or reduced daily lying time can be indicative of pain in cattle. This review helps to properly interpret pain in cows through behavioural activity patterns and explores whether pain relief is capable to restore their normal activity.

The main conditions and diseases considered painful in dairy cows are mastitis, lameness, calving (including dystocia and caesarean section) and metritis. The cattle literature reports that deviation from normal daily activity patterns (both increased and/or reduced daily lying time) can be indicative of painful conditions and diseases in cows. This narrative review discusses on how pain due to several health conditions in dairy cows modifies its activity pattern and explores if non-steroidal anti-inflammatory drugs (NSAIDs) are capable of restoring it. Divergent outcomes may differ depending upon the painful cause, the severity and the moment, and consequently its interpretation should be properly explained. For instance, cows with clinical mastitis reduced their time lying and increased the number of lying bouts and stepping due to pain caused by the swollen udder when cows are lying. However, lame cows show longer lying times, with a lower number of lying bouts and longer and more variable lying bouts duration, as compared to non-lame cows. When the relationship between painful disorders and daily activity patterns is studied, factors such as parity, bedding type and severity of disease are important factors to take into consideration. The potential benefits of the NSAIDs treatment in painful health disorders depend upon the type of drug administered, its dosage and administration mode, and the time of administration relative to the painful health disorder. This narrative review can be used as a tool to properly interpret and grade pain in cows through behavioural activity patterns and proposes directions for future investigations.

1. IntroductionRecent reviews have focused on the assessment and alleviation of pain in management procedures in cattle, such as castration and/or dehorning [1,2,3]. Other reviews focused on the assessment and management of pain associated with calving [4], lameness [5] and/or surgical pain in cattle [6]. Special emphasis has been placed on pain evaluation in dairy cattle based upon pain-specific behaviours [7] and daily behavioural patterns in sick dairy cows [8]. In addition, lying time and its association with welfare in cows has recently been reviewed [9]. Although several studies that focused on health conditions used activity patterns (especially lying time) as a pain indicator, its outcome interpretation has not been previously reviewed. This narrative review discusses how pain due to several health conditions in dairy cows modifies their activity patterns and explore whether non-steroidal anti-inflammatory drugs (NSAIDs) are capable of restoring it.This work consists of a narrative review. All searches were performed using keyword search terms (cattle or dairy cows and pain or discomfort) in Science Direct and Scopus, including all peer-reviewed scientific articles published before May 2021. The review focuses on mastitis (or udder inflammation), lameness (or hoof trimming), calving (or parturition or dystocia) and metritis. Activity was reported as lying time, number of lying bouts, lying bout duration or number of steps. Different methodologies of activity recording were considered: direct observations, video images observations and automatic sensors of activity. NSAIDs including meloxicam, flunixin meglumine, ketoprofen, carprofen and acetylsalicylic acid were considered.2. The Sensory and Emotional Experience of PainIt is generally accepted that animal welfare comprises physical and mental health [10] and includes several aspects such as absence of thirst, hunger, discomfort, disease, pain, injuries and stress, as well as the possibility to express normal behaviours [11]. As a consequence, one of the essential components of good welfare is the recognition and control of pain.Humans and vertebrates share similar neuroanatomical structures associated with pain perception, namely, nociceptors, nociceptive pathways and processing areas in the central nervous system (CNS) [12,13]. As a consequence, it is well established that farm animals are capable of experiencing pain.Pain is defined by the International Association for the Study of Pain (IASP) as an unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage [14]. By this definition, pain is a subjective experience which requires the activity of structures in the brain to be perceived. This is in contrast to nociception, which is defined as the encoding and processing of noxious stimuli in the nervous system. Pain has a fundamentally important protective role, alerting us to threats and providing an impetus for the preservation of the integrity of the body. However, in the context of veterinary treatment of illness or injury, pain may become an unwanted consequence that is generally not useful and compromises the welfare of animals. Pain can be developed following tissue damage, inflammation and nerve injury [15]. Pain is usually transitory, lasting only until the noxious stimulus is removed or the underlying damage or pathology has healed (acute pain). However, chronic pain may persist for months and can be even more detrimental to animal welfare [16]. Chronic pain is characterized by hypersensitivity to potentially painful stimulation and is clinically manifested as spontaneous pain (pain in the absence of any stimulation), as well as hyperalgesia (an exaggerated response to a noxious stimulus) and allodynia (the presence of a pain response to a non-noxious stimulus, such as a gentle touch) [17].Nociceptive pain may also be classified according to the site of origin (a terminology that will be used in the present review). Visceral pain results from the activation of nociceptors of the thoracic, pelvic or abdominal viscera. Visceral pain is diffuse, difficult to localize and often referred to a distant, usually superficial, structure. Superficial somatic pain is initiated by activation of nociceptors in the skin or other superficial tissue, and is sharp, well-defined and clearly located [18,19].The pain experienced by cattle may be the result of management procedures such as castration, dehorning and tail docking, or a variety of conditions and diseases considered painful. According to different surveys of farmers and veterinarians [20,21,22,23], the main conditions and diseases considered painful in dairy cows and reviewed in the present work are mastitis, lameness, calving (including dystocia and caesarean section) and metritis. Many of these conditions occurred on the farm without the use of anaesthetics and/or analgesics for pain management [1]. Veterinarians are expected to be able to diagnose, score and treat pain in cattle. A large variability in pain scores for individual diseases and pain conditions in cows, as well as in analgesic treatment practices in cows, have been documented [21,24]. One reason for the inconsistency in pain recognition and thus pain relief in cattle is the inadequate ability to identify and assess pain [25]. In addition, economic and legislative considerations may be significant factors, too [26].3. Pain Identification and Assessment in CowsThere are at least two reasons that explain why assessing pain in ruminants is particularly challenging. Firstly, as pain is a subjective personal experience, it is very difficult to objectively measure it scientifically and validate it in animals. Secondly, from an evolutionarily perspective, cattle are considered a prey species, and are inclined to avoid showing pain, even when exposed to harmful stimuli. This stoicism makes identification of pain, and therefore disease, a challenging task [27,28].Pain indicators should be reliable (repeatable), valid and sensitive. Reliability is tested to ensure that the indicator is repeatable between (inter) and within (intra) observers and is not affected by observer bias or observer subjectivity. Validity ensures that the indicator is measuring pain and not related to any other state such as fear, anxiety or general sickness behaviour. A pain indicator should be sensitive, meaning that it allows the detection of different levels of pain and that it increases linearly with the severity of pain [29]. In addition, indicators for on-farm use should be readily applied, inexpensive, non-invasive and provide immediate (not retrospective) information [30].Ideally, pain indicators should be validated by a “gold standard”. In humans, the self-reporting of pain has been the most commonly used method of pain assessment and validation both in clinical practice and research [31,32]. Perceived pain in animals cannot be directly measured or based upon other measures; it can only be inferred [32]. Since validation is based upon the measurement of an indicator against a “gold standard”, pain indicators must be thoroughly evaluated using experimental studies. The majority of studies used several measures simultaneously, with different experimental approaches, to evaluate the potential value of these measures as indicators of pain [32]. Pain assessment in animals has tended to use four approaches: measures of general indices, physiological indicators, behavioural indicators [28] and, recently, facial expressions [33]. Some studies integrate those measures into a graded scale in order to fully assess the impact of a painful condition or event on the individual (e.g., [7]). Pain assessment based on behaviour has received increasing attention and is the most commonly used parameter to assess pain in farm animals [34]. Three types of behaviours, useful for pain evaluation in farm animals, have been proposed [7,28]: (1) pain-specific behaviours; (2) avoidance-of-pain behaviours; and (3) a change in certain behaviours that the animals are very motivated to perform.Pain-specific behaviours encompass a promising method of assessing pain in farm animals (e.g., [30]). This approach has been used extensively in rodents to assess pain in order to provide the best possible treatment and refine painful techniques [35]. For instance, decreased burrowing behaviour and voluntary wheel running, weight bearing/gait, abnormal postures, paw flinches and ultrasonic vocalizations have been systematically observed in multiple rat models of chronic pain [36]. However, pain-specific behaviours are usually obtained by direct and continuous observations by trained observers (e.g., [7,29]), thus making it time-consuming.Avoidance-of-pain behaviours assess the reaction of animals upon being manipulated and is a commonly used method to assess pain. These indicators are considered valid and reliable as long as the reaction is scored in a standardized way [37]. Pain sensitivity has been quantified using mechanical (e.g., [38]) or thermal stimulation of a hind leg or the udder (e.g., [39]), mainly to assess pain in cows suffering from mastitis. These methods measure the nociceptive threshold, defined as the minimum stimulus necessary to elicit a pain response. When a stimulus is applied to a painful site, a cow responds with avoidance behaviour such as kicking, leg lifting or tail flicking [40]. Lower nociceptive threshold values indicate that there is increased pain.A change in certain normal behaviours, such as daily activity patterns, including lying behaviour, has been used extensively in the cattle literature to assess pain caused by several conditions. However, up to now, behavioural activity patterns have not been properly interpreted as a pain indicator in cows. Several electronic data loggers are widely available and can be used to measure lying behaviour accurately, including the total time spent lying down, the number of lying bouts and the duration of each bout for individual cows [41]. The automatic recording of behavioural activity (lying, standing and walking) can be achieved using a variety of sensor systems; for example, mercury tilt switches, three-dimensional accelerometers embedded sensor technology and automatic local position systems [42,43].Daily activity patterns indicative of pain in cows have been investigated in experimental studies by (a) animals as their own control: observing animals before, during and after a painful process or procedure; (b) comparing animals thought to be in pain to the controls (pain-free animals); (c) observing animals thought to be in pain and receiving effective analgesia, as compared to a placebo; and (d) observing whether indicators of pain increase with the severity of pain. Ideally, a combination of these methods would be used to test the validity of behavioural pain indicators [30].The main behavioural indicator of activity patterns used in cows is the time spent lying down. Other related behavioural indicators of activity patterns used are the frequency of lying bouts (i.e., a transition from standing to lying), the duration of individual bouts and activity or number of steps. Healthy lactating cows lay down 11.0 ± 2.1 h/d in 9 ± 3 bouts/day, with a bout duration of 88 ± 30 min/bout. These values ranged from 9.5 to 12.9 h/d, 7 to 10 bouts/d and 65 to 112 min/bout across farm means [44]. However, lying behaviour varies considerably among dairy systems, with the shortest duration often in pasture systems and the longest usually in tie-stalls (see [9] for further revision).Lying is a behavioural need for dairy cattle. Lying is considered to be a higher priority than eating and social contact when opportunities to perform these behaviours are restricted [45]. As a consequence, lying behaviour has been identified as an element that can be used to measure a cow’s welfare status [45,46,47]. Lying behaviour has been identified as a sensitive measure of animal comfort (e.g., [48]) and to assess pain caused by several conditions in cows (e.g., [49]). Deviation from normal lying behaviour (both increase or reduction) can be indicative of pain in cows. Longer lying times (within a normal range) are generally indicative of increased welfare [9]. However, resting for a long time with few lying bouts can also indicate illness. Sickness behaviour is a well-organized adaptive response to enhance disease resistance and facilitate recovery from disease [50]. Sickness behaviour includes changes in behaviour and physiology. In fact, tissue injury and infection unleash the signals necessary for immune activation. Inflammatory mediators released during immune activation are initiated by a group of chemical mediators known as cytokines [51]. Interleukins are a particularly important group of compounds within cytokines and they mediate the acute-phase response by initiating fever, synthesis of acute-phase proteins and the immune response originating in lymphoid tissue, among others [51]. Beside this, they act on the brain (hypothalamus), which initiates anti-inflammatory mechanisms and some of the main behavioural effects involved in sickness behaviour. Sickness behaviour can include lethargy, apathy, somnolence, fever, loss of appetite and thirst, reduced social interaction and a decrease in general physical activity, among other symptoms. Lying is often increased in ill animals [50,52], where reduced activity is considered beneficial for energy conservation and maintenance of the febrile response [53].Overall, in order to properly understand how a painful condition can affect lying behaviour in cows, different aspects have to be considered. Firstly, the origin, location, intensity and source of pain have to be identified. Secondly, irrespective of pain, it is necessary to identify if the cow shows some or several symptoms associated with sickness behaviour.Daily activity patterns are affected by different factors others than pain. Studies assessing pain in cows should standardize other factors affecting daily activity patterns or at least consider them as influencing factors. Daily activity patterns can be influenced by individual, environmental, housing system and management factors. The main individual factors described are parity [54], social ranking [55], days in milk (DIM) [56] and milk yield [48]. The main environmental factors described are seasonal changes [57,58], including heat stress [59]. The main housing-system factors described are dairy systems [60,61,62,63], stall dimensions [64], stall features [65,66], stall surface [67] and bedding material [68]. The main management factors described are stocking density [69,70,71], grouping strategies [72], feeding management [73] and time spent away from the pen for milking [65]. In summary, other factors are known to have an important effect on daily activity patterns (especially lying behaviour). Clearly, these factors have to be taken into consideration when an experimental study assessing pain through daily activity patterns is done and caution is required in extrapolating findings to other conditions.4. Pain Management in Dairy CowsPain relief is based on treating inflammation and other systemic effects that commonly accompany inflammation and slow any further tissue damage. The use of corticosteroids may not be appropriate because of their immunosuppressive and metabolic side effects [74,75]. Numerous NSAIDs are available and licensed for use in farm animals, demonstrating their safety and efficacy. Over the last decade, there has been an increase in their use, probably because there is increasing evidence to support the benefits in painful diseases such as lameness, mastitis and metritis [26]. NSAIDs are commonly used in animals to reduce inflammation (anti-inflammatory), reduce pain (analgesic) and decrease overall body temperature (antipyretic). Incorporating NSAID therapy into treatment protocols for a variety of painful conditions should improve the welfare of animals and, consequently, decrease the related economic losses [40,76]. When a NSAID is administered prior to a specific painful condition or disease (instead of administering NSAIDs during or after the painful moment), animals returned to a normal physiologic state more quickly [77]. NSAIDs act by inhibiting the cyclooxygenase enzyme (COX), which in turn prevents prostaglandin synthesis [40]. It is known that while COX-1 acts mainly on housekeeping physiological functions, COX-2 is normally activated in specific conditions such as inflammation [78]. Therefore, the therapeutic effect of NSAID drugs is mostly derived from their ability to inhibit COX-2, while most of their side-effects (gastrointestinal irritation, renal toxicity and inhibition of blood clotting) are caused by the inhibition of COX-1 [79]. The most frequently cited NSAIDs used for painful conditions in cattle are flunixin meglumine, meloxicam and ketoprofen [1,21,80]. Meloxicam is known to be a preferential COX-2 inhibitor [81], having a targeted action against inflammatory processes and optimal safety profile in cattle. Both ketoprofen and flunixin meglumine have a prevalent activity on COX-1. The half-life elimination of flunixin meglumine is 3 to 8 h, and ketoprofen has a short half-life of 0.42 h, and therefore requires extra dosing. In turn, the elimination half-life of meloxicam is found to be 24–26 h [82]. Other NSAIDs used in cattle are salicylate (COX-1 and -2) and carprofen (COX-2 selective and well-tolerated in cattle) [83]. Furthermore, it is important to note that with most NSAID treatments, milk discarding is mandatory for at least 24 h to prevent contamination with drug residues (see [83] for further review).5. Pain Associated with MastitisMastitis is an inflammation of the mammary gland that may be caused by a number of different bacteria. Mastitis is a multifactorial disease in which the environment, the pathogens and the host (cow) interact with each other. Despite being a major animal welfare and economic problem in dairy cattle production [84], mastitis is usually underestimated by farmers [85]. The incidence rate of clinical mastitis ranges from 13 to 40 cases/100 cows per year in different countries and housing types [84].From several surveys performed using farmers and veterinarians, mastitis due to Escherichia coli (E. coli) in cows was given a high score for pain (from 7 to 9 out of 10). However, mild mastitis (milk clots only) received lower painful scores (from 1 to 3 out of 10) [21,24,86]. During clinical mastitis, the concomitant inflammation of the udder, increased intramammary pressure and increased external pressure (e.g., of the adjacent limb on a swollen udder) induce pain [40]. It is accepted that cows can also experience pain in mild or moderate cases of clinical mastitis. Even subclinical mastitis is accompanied by increased levels of bradykinin, a peptide that mediates the inflammation related to mastitis [87]. Additional knowledge on the levels of pain experienced by cows, depending on the phase of the disease (especially during the remission phase), is needed. In dairy cows, after udder inoculation with E. coli, de Boyers des Roches et al. [88] suggested that cows may have experienced discomfort in the pre-clinical phase (0–8 h post-inoculation) and pain in the acute phase (12–24 h post-inoculation), but neither discomfort nor pain in the remission phase (32–80 h post-inoculation). Nevertheless, Fogsgaard et al. [89] showed in dairy cows with naturally occurring mastitis, that local clinical udder signs as well as behavioural changes persisted throughout one week after antibiotic treatment. Severe cases of mastitis lead to hyperalgesia and allodynia [90]. A hyperalgesic state lasting at least four days has also been described in cows with mild or moderate mastitis [91].5.1. Alteration in Activity Patterns as a Result of Pain Due to MastitisAlthough lying time usually increases during illness and it is thought to help the animal to save energy, most studies observed that cows with clinical mastitis (experimentally induced with endotoxin or E. coli or naturally occurring) reduced their time lying and increased stepping (Table 1). The most likely explanation for the reduced lying time is pain caused by the swollen udder when cows with mastitis are lying [92,93]. It may be possible that lying through direct contact with a hard surface causes more pain to the udder than having the udder hanging freely while standing [94]. In fact, the cows increased their standing time simultaneously with the local swelling of the udder quarter [92], and time spent lying was negatively correlated with rectal temperature and blood cortisol during the first 12 h after intramammary infusion [95]. Moreover, the distance between the hocks when the cow is standing is wider in cows with mild and moderate mastitis, when compared to healthy cows, suggesting that there has been a change in the hindleg stance of the cows as a result of the inflamed udder [96]. If lying is painful for the cow, and thus reduced, the cow might become increasingly frustrated and restless over time. The increased stepping and increased frequency of lying bouts [89] may be indicative of such motivational conflict and, hence, of coping difficulties in response to the disease. The increased stepping may be also a pain-avoidance behaviour [97], as an altered stance has earlier been shown in cows suffering from mastitis [96].Recent descriptions of altered behaviour during experimentally induced clinical mastitis in cows have shown normalization within a few days after diagnosis [92,93]. However, Fogsgaard et al. [89] reported that clinical mastitis occurring naturally might last at least 10 days (one week after antibiotic administration).Nonetheless, the differences among activity measurements observed after a Streptococus uberis (S. uberis) challenge in cows (increased lying time and reduced stepping) reflect a difference in host mammary responses to different pathological agents [98]. Similarly, other studies did not show differences in the activity or resting behaviours, presumably because other factors such as stage of lactation or housing conditions (free-stalls vs. tie-stalls) are not completely comparable.In addition, some authors observed that cows with mastitis spent [93] or tended to spend [92] less time lying on the side of the affected udder quarter. Pain or discomfort when lying on the affected udder quarter would cause the cow to spend more time lying on the other udder side, as an avoidance behaviour. Other studies did not report any difference in lying time depending on the affected udder side [94,99,100,101], suggesting that cows seem to have a preference as to which side they lie upon, regardless of the infection status of their mammary gland [100].5.2. The Effect of NSAIDs on Activity Patterns in Cows with MastitisIn general, the use of NSAIDs has been shown to decrease rectal temperatures, decrease signs of inflammation, restore rumen motility, increase the eating time and dry-matter intake and reduce heart rates in experimentally induced mastitis cases, as compared with their untreated counterparts [27]. However, the amount of time cows spent lying was not affected by flunixin meglumine following experimentally induced mastitis [95,100]. It can be expected that NSAIDs would reduce udder inflammation and increase the lying time. The reason why flunixin meglumine did not show these expected effects remains unclear. Several factors, such as the short half-life elimination of flunixin meglumine [83] or delayed administration after challenge (from 4 h to 24 h) [95,100], could have influenced the result. Additional research including the effect of other NSAIDs on experimentally induced mastitis on lying time should include a closer administration before or after challenge and/or a repeated dose administration. To our knowledge, no study reports the effect of NSAIDs on lying time in naturally occurring mastitis in dairy cows.animals-12-00176-t001_Table 1Table 1Summary of changes in activity patterns in cows (lying time, lying time on quarter affected, number of lying bouts, lying bout duration and activity) caused by mastitis. Studies are summarized according to how they have been assessed in relation to pain using the following methods: cows studied as own control (before, during and after mastitis) and comparing cows with mastitis to the control. Moment (in days, d, or hours, h, related to challenge or naturally occurring mastitis; D0 = day of challenge or diagnosed natural mastitis) and type of change (↑ increase; ↓ decrease; NS = non-significant; NE = non-evaluated) are given comparing the painful moment or cows in pain due to mastitis versus the control moment or control cows.Studies ClassificationReferenceOrigin of Pain Due to MastitisControl Moment or Control CowsLying Time (h/d)Lying Time on Quarter Affected (h/d)No. Lying Bouts (No./d)Lying Bout Duration (min/d)Activity (No. Steps/Day)Cows as own control (before, during and after mastitis)[94]Challenge (D0) with E. coliPrechallenge (D-1, D-2)↓ (from +4 h to +7 h)NSNSNSNEPostchallenge (D+1, D+2)NSNSNSNSNE[92]Challenge (D0) with E. coliPrechallenge (D-1)↓ until +20 h↓ (t)NSNS↑[93]Challenge (D0) with E. coliPrechallenge (D-1, D-2)↓ (D+2)↓ (D+2)NENENEPostchallenge (D+1, D+2)↓ (D+2)↓ (D+2)NENENEMastitis vs. control[95]Challenge with E. coli + PBS 4 h post-challengeIntrammamary infusion with PBS + PBS 4 h post-challengeNS (from 0 h to +24 h)NENENENE[100]Challenge with E. coli + PBSNon-challenged:Infusion with PBS↓ (from +3 h to +6 h)NSNSNSNS[98]Challenge with S. uberis + PBSUninfected↑ (D+4)NENSNE↓ (from D+1 to D+3)[89]Cows with naturally occurring clinical mastitisClinically healthy control cows↓ (from D0 to D+2)NE↑ (from D+3 to D+10)NE↑ (from D+6 to D+10)Significant differences are established at p ≤ 0.05, and tendency (t) at p ≤ 0.1. PBS = phosphate-buffered saline (water-based salt sterile solution that helps to maintain a constant pH).6. Pain Associated with LamenessLameness is characterized by an abnormal gait caused mainly by an injury, a disease, or a dysfunction of one or more feet and/or limbs. Lameness is a multifactorial disease associated with management, equipment features, nutrition and genetic selection hazards [102]. Lameness compromises the welfare of affected animals and results in economic losses due to reduced milk yield and fertility and increased treatment cost [103]; however, lameness is usually underestimated by farmers [104]. Under-recognition of lameness occurs in all cases, slight lameness being the most affected. The major effect of this under-recognition of lameness is not that lame cows are not treated but that treatment is delayed more than 3 weeks, especially of less severely lame cows [105]. Lameness prevalence ranges from 5.1% to 54.8% [106], with the highest rates observed in herds housed in tie-stalls [107] and free-stalls [108,109,110]. Although the effect between free-stall and tie-stall housing on lameness incidences is not clear [107], several studies conclude that lameness in tie-stalls system are underestimated due to the difficulty of assessing it [111,112]. The use of a Lameness Scoring System (LCS) on a regular basis is the most effective means of identifying lameness in cows. A LCS describes the gait properties to classify the severity of lameness on a numerical scale [104]. The five-point LCS is one of the most frequently employed methods in lameness detection (score 1 = normal; score 2 = mildly lame; score 3 = moderately lame; score 4 = lame; and score 5 = severely lame) [113].Several surveys performed using farmers and veterinarians concluded that different diseases in cows that can lead to lameness are scored as a painful condition. For instance, interdigital necrobacillosis, digital dermatitis and swollen hock received the following painful scores on a scale of 10: 8.2, 7.0 and 5.3, respectively, by farmers, and 7.5, 6.3 and 5.4, respectively, by veterinarians [23]. In another survey, Canadian veterinarians scored acute lameness as 6.4 and chronic lameness as 5.2 on a scale of 10 [20]. Claw lesions remain among the major causes of lameness in dairy cows. Injuries of the claws or the limbs cause painful bruises, which lead the cow to change its locomotion due to pain avoidance. Tissue damage or infections of the claw are enough to induce sickness behaviour (e.g., reduced feed intake and lethargy), as inflammatory cytokines are released. Furthermore, non-inflammatory reasons for lameness lead the cows to change their behaviour independently of inflammatory cytokines [8,114]. Hyperalgesia has been observed at the time of lameness detection and one month later, suggesting that clinically lame cows suffer from chronic pain [5]. Furthermore, cows with moderate lameness can also experience pain and pain can appear before the cow becomes clinically lame [115].6.1. Alteration in Activity Patterns as a Result of Pain Due to LamenessMost of the studies report that lame cows have longer lying times, with a lesser number of lying bouts and longer and more variable lying bouts duration, as compared with non-lame cows (Table 2). For example, Solano et al. [116] showed that cows with a lying time ≥ 14 h per day, ≤5 lying bouts per day, bout duration ≥ 110 min/bout or standard deviations of bout duration over 4 d ≥ 70 min had 3.7, 1.7, 2.5 and 3.0 odds of being lame, respectively. Lame cows are less active (in terms of number of steps per hour) than non-lame cows [117], whereas Chapinal et al. [118] did not find this difference. Similarly, hoof trimming (a preventive procedure that reduces the incidence of lameness) also provokes increased time spent lying on the day of hoof trimming, and for up to 4–5 weeks afterwards, and a reduced activity on the day of hoof trimming in both lame and non-lame dairy cows [119,120,121,122].The severity of foot lesions (mild, moderate or severe) had a more pronounced effect on reducing the daily activity levels of cows than did the type of lesion present (digital dermatitis, sole ulcer or white-line disease). In fact, increased time spent lying down is associated with increased locomotion scores [123], and increased locomotion scores are associated with increased weight-shifting behaviour, a marker of pain in lame cows [118]. Severely lame cows present a longer daily lying time and lying bouts duration than do moderately lame cows [23,114,123]. Lesions that cause the greatest increase in lying behaviour are digital dermatitis and sole ulcers, which have longer lying bouts duration and, consequently, spend more time lying down per day [23,124]. For instance, cows with sole ulcers spent > 1 h/d more lying down, and their mean lying bout duration is 20 min longer, when compared to cows without hoof lesions [119].These findings suggest that the pain associated with ulcers may reduce the willingness of a cow to stand up once she is lying down [124]. Tissue injury and infection, which are common elements of lameness in dairy cows, lead to immune system activation and the behavioural modifications intrinsic to sickness response. In fact, in addition to changes in locomotion and activity, lame cows display decreased food intake and altered social behaviour, such as increased self-grooming behaviour [125]. Nevertheless, lying time per se does not seem to be a potentially useful indicator for detecting moderate lameness in dairy herds [126].Although most studies report an increase in lying time in lame cows, Cook et al. [67] report a decrease in lying time in lame cows allocated in free stalls with sawdust. Other studies report no differences or a tendency of an increase in lying time for lame cows [123,126]. A potential reason for the lack of association between lying time and lameness could be that severely lame cows are not included in some studies (e.g., [126]). In addition, in lame cows, lying behaviour is specially affected by bedding type. For instance, lame cows spent more time standing in stalls than did non-lame cows, but this difference was greater with mattress stalls, as compared with sand stalls [67,127]. Severely lame cows in deep-bedded stalls lie down longer than do non-lame cows, whereas in mattress herds, no difference in lying time was observed [123]. Similarly, moderately lame cows had a reduced lying time, when compared with non-lame cows in herds with a mattress, an effect not observed in herds with sand [68]. In summary, increased lying time due to lameness may be more difficult to identify in mattress stalls, whereas lame cows in bedded stalls are more prone to maintain their normal resting behaviour. Clearly, when the relationship between lameness and lying behaviour is studied, bedding material and the inclusion of severely lame cows are important factors to take into consideration, and special caution is required in extrapolating findings to other studies [118]. In addition, high variability between lying time in cows [44] could hinder finding differences between lame and non-lame cows using activity measures alone. Lame cows show contradictory extreme lying times, suggesting that the effect of lameness on lying behaviour must consider the pain suffered in both rising and lying movements. The interaction with the surface during both rising and lying movements may appear as difficult (and painful) in lame cows. Some lame cows may find it difficult (and painful) to stand, and thus lie down for longer (in total and per each lying bout), whereas others may find it difficult (and painful) to lie down, and thus show a lower number of lying bouts and stand in the stall for longer [68].animals-12-00176-t002_Table 2Table 2Summary of changes in activity patterns in cows (lying time, number of lying bouts, lying bout duration, standard deviation of bout duration and activity) caused by lameness. Studies are summarised by comparing cows with lameness to the control (including degree of severity). Lameness assessment score (from 1 = non-lame to 3 or 5 = severely lame) and housing system and bedding type are documented. Type of change (↑ increase; ↓ decrease; NS = non-significant; NR = non-reported; NE = non-evaluated) is given comparing cows in pain due to lameness versus the control cows.ReferenceLSCCriteria of Painful Moment or Cows in Pain Due to LamenessCriteria ofControl Moment or Control CowsHousing System and Bedding TypeLying Time (h/d)No. Lying Bouts (No./d)Lying-Bout Duration (min/d)SD of Bout Duration (min)Activity (No. Steps/h)[116]5-point scaleScore ≥ 3Score 1 and 2Free-stalls(different types of bedding)↑↓↑↑NE[68]4-point scaleScore 2Score 1Free-stalls(mattress + sand)NS↓NSNENEScore 3Score 1↑ (mattress)NS (sand) NENE[126]5-point scaleScore 3 or 4Score 1Free-stalls(straw)NSNS↑NENE[67]3-point scaleScore 2 or 3Score 1Free-stalls(mattress + sawdust)↓NENRNENEScore 2 or 3Score 1Free-stalls (mattress + sand)NSNENRNENE[127]4-point scaleScore 3Score 1 and 2Free-stalls (mattress + sawdust; and sand)↓NENENENE[124]5-point scaleD: score = 3.0H: score = 2.7U: score = 3.3NL: score = 2.9Free-stalls(sand)D: NS H: NSU: ↑D: NSH: NSU: ↓D: NS H: NS U: ↑NENE[118]5-point scaleScore > 3Score ≤ 3Free-stalls (sand)↑(t)NS↑NENS[128]4-point scaleScore 2 or 3TrimmingScore 0 or 1NRNSNSNSNENEScore 2 or 3Trimming + foot blockScore 0 or 1NR↑NSNSNENE[123]5-point scaleL: score ≥ 3SL: score = 4NL: score ≤ 2Free-stalls (deep-bedded)L: ↑ (t)SL: ↑L: NSSL: NSL: NSSL: ↑L: NSSL: ↑NEFree-stalls (mattress)L: ↑ (t) SL: NSL: NS SL: NSL: NS SL: NSL: NS SL: ↑(t)NE[117]5-point scaleScore ≥ 3Score ≤ 2Free-stalls(bedding not described)NENENENE↓Significant differences are established at p ≤ 0.05, and tendency (t) at p ≤ 0.1. SD = standard deviation; LSC = Lameness Scoring System; D = dermatitis; H = haemorrhages; U = ulcers; NL = non-lame; L = lame; SL = severely lame.6.2. The Effect of NSAIDs on Activity Patterns in Cows with LamenessThe major part of studies assessing the effect of NSAIDs on lameness and lying behaviour are field trials performed around the moment of hoof trimming (Table 3). Flunixin meglumine has shown substantial acute analgesia in the induced lameness model, as shown through improvement of the gait score and increased pressure on mat systems on the affected foot and claw [129,130]. However, in field trials, NSAIDs have shown variable results, with a mild improvement in the locomotion score and nociceptive thresholds [5].To our knowledge, only one study assessed the effect of NSAIDs on induced lameness in cattle (by an intra-articular injection of amphotericin B) and recorded lying behaviour as an indicator of pain relief [129]. Schulz et al. [129] reported that flunixin-treated steers spent less time lying than did their control counterparts during the early post-induction lameness. Regarding field trials, Offinger et al. [131], studying cows severely lame from septic arthritis, showed that meloxicam treatment improved gait scores after surgery. In addition, cows treated with meloxicam showed a reduced daily lying time and a higher number of steps per hour, as compared with control cows. On the contrary, other field trials showed no significant differences [118,128], only a tendency [119] or results with little clinical relevance [121] when evaluating the effect of NSAIDs on lameness and lying behaviour and/or activity. These inconsistent outcomes between induced lameness and field trials may be a result of clinical heterogeneity present in different cases of clinical lameness in field trials and/or a low sensitivity of indicators to detect analgesia effects on moderate lameness [5]. Additional studies are needed to further investigate the potential analgesia effect of NSAIDs in mitigating the pain associated with lameness, both in induced lameness models and field studies. These studies should accurately assess the degree of severity of lameness and, if possible, to identify the most plausible cause of lameness. Lying behaviour can be included as one of the measures to assess pain in lame cows, but other indicators, such as pressure mats and a locomotion score, should be included.7. Pain Associated with CalvingCalving itself (and especially dystocia) is a cause of pain [4]. Dystocia can be defined as a difficult parturition resulting from prolonged spontaneous parturition or prolonged or severe assisted extraction [132]. Calving normally occurs between 30 min and 4 h from the appearance of the amnion to the birth of the calf [133]. The two main causes that can produce calving longer than normal are foeto-pelvic disproportion and foetal malpresentation. From several surveys, dystocia was ranked by veterinarians and farmers as one of the most painful conditions of cattle, obtaining a score from 7 to 8 out of 10 [21,22,86]. Most dairy cows experience a physiological subacute inflammation during the first days after calving [134]. If the degree of inflammation exceeds a certain level, it becomes harmful, impairing normal lactation [83]. In fact, around the time of birth, the levels of acute-phase proteins (such as haptoglobin and serum amyloid protein) increase considerably in response to inflammation, tissue damage and, thus, pain [135]. Injury, trauma and inflammation associated with calving (particularly during dystocia calving) can have important negative effects on health, welfare and productivity [4].Caesarean sections in dairy cattle are carried out in approximately 1–2% of calvings [136,137]. Following a caesarean section, cows may experience pain of somatic and visceral origin, arising from the surgery itself and failed attempts at vaginal delivery [4,137,138,139]. Caesarean sections have been rated at 9 on a 10-point pain severity scale by UK veterinary surgeons [21], suggesting that pain associated with a caesarean section is a significant welfare concern. However, very little is known about the pain experienced by cattle following a caesarean section.7.1. Alteration in Activity Patterns as a Result of Pain Due to CalvingThe activity of cows increases quite dramatically prior to calving. On the day of calving, cows spent approximately 2 h less time lying down and spent an average of 31.4 min less per lying bout, as compared with the days before calving (e.g., [140]). The number of lying bouts were more than doubled from Day-2 to Day-1 prepartum [141]. Activity such as number of steps per day on the day of calving ranges from 2847 to 5424 steps/day [135,140] and represents an increase from 57% [135] to 62% [140], when compared to baseline days (7 days before or 7 days after calving, respectively). Although this behavioural pattern might be a part of the normal calving situation, it has been suggested that this increased restlessness may be due to pain caused by uterine pressure and cervical dilatation [4,142]. Alternatively, and in a complementary way, increased activity may also be explained by their natural instinct to distance themselves from the herd to find a secluded place to give birth [143,144].Deviation from normal activity behaviour around calving (either increased or reduced activity) are indicative of additional pain during the calving process. Cows that delivered stillborn calves have greater prepartum activity than do those that delivered live calves, probably indicating additional pain before calving [145]. However, cows that experienced dystocia spent more time lying, had a fewer number of steps, had more lying bouts and had shorter lying bouts duration than did eutocic cows from calving to 7 DIM [146]. Similarly, cows who delivered twins have reduced activity during prepartum, as compared with cows delivering singletons [145]. In those two latter cases, the reduced prepartum activity observed in dystocic cows or cows delivering twins suggest that dystocic cows are subjected to additional pain before calving (Table 4). Primiparous cows show a higher number of steps [135,140] and reduced total time lying [49], when compared to multiparous cows around calving. Apart from the lack of experience of the dams at first parturition, the majority of studies confirm that primiparous females have longer durations of parturitions and the degree of effort associated with it is usually greater than in multiparous females [133]. In order to better correlate pain due to calving with lying behaviour, the degree of difficulty at calving should be considered. Currently, pain scales used to grade dystocia are very subjective and complicate attempts to compare or merge findings between farms and countries [147]. Future studies should incorporate pain scales more objectively at calving, for instance, including physiological indicators of inflammation to properly grade the degree of pain at calving.Activity patterns around calving can be easily affected by other factors than pain. Such confounding factors (e.g., metabolic strain of lactation, management, adaptation to new calving environment, early calf-mother separation, etc.) can all influence the activity-patterns expression of post-partum pain [137].7.2. The Effect of NSAIDs on Activity Patterns in Cows with CalvingThe potential benefits of the treatment of periparturient cows with NSAIDs depend upon the type of drug employed, its dosage and administration mode, and the time of administration relative to calving. Overall, milk yield and composition are not affected by NSAIDs administration after calving. Nonetheless, the impact of NSAIDS on milk yield and composition may vary depending upon cow inflammation levels. Indeed, the administration of salicylate increases milk production of cows with high haptoglobin levels [148,149]. Positive effects of NSAIDs that preferentially inhibit COX-2 activity, such as meloxicam, on milk production seem to be plausible in a large-scale study [150]. It has been reported that NSAIDs that preferentially inhibit COX-1 activity, such as flunixin meglumine, can trigger side effects such as an increased risk of retained placenta, culling or metritis [151], reduced feed intake and increased rectal temperature after calving [152]. On the contrary, COX-2-preferential NSAIDs displayed possible benefits to health and performances such as a lower culling rate, a lower risk of mortality on the farm, a reduced incidence of mastitis and an increased pregnancy rate at first insemination [148,150,153].According to studies carried out to date, lying time, number of lying bouts and lying bouts duration were not significantly affected by NSAIDs treatment around calving (Table 5). However, the activity (in terms of number of steps per day) was altered by NSAIDs treatment around calving, although it shows divergent results among studies. Eutocic cows treated with acetylsalicylic acid [146] and primiparous eutocic cows treated with meloxicam [135] after calving coincided in that NSAID treatment increases activity during the first 2 DIM, when compared to the controls. NSAIDs could lead to a reduction in the mediators of inflammation (cytokines and prostaglandins) responsible for the expression of sickness behaviour, such as lethargy and depression, observed in inflammatory states [154,155]. In a similar way, Newby et al. [156] reported that meloxicam administered after calving in cows increased the number of feeding visits and total time spent feeding, suggesting some pain alleviation after calving. On the contrary, Ref. [157] reported that meloxicam administered either before or after calving decreased activity post-calving in the Jersey breed (but not in Holstein) and in cows with dystocia (defined as cows that calved in >70 min). The authors speculate that administration of meloxicam to dystocic animals reduced activity by alleviating inflammation, allowing the cow to rest more easily. Similarly, Barrier et al. [137] observed that beef cows treated with meloxicam prior to a caesarean section spent more time lying and had more lying bouts following surgery, as compared to cows receiving a placebo. This last result concurs with previous research examining behaviour following abdominal surgery [137,158,159].animals-12-00176-t004_Table 4Table 4Summary of changes in activity patterns in cows (lying time, number of lying bouts, lying bout duration and activity) caused by calving. Studies are summarized according to how they have been assessed in relation to pain using the following methods: cows studied as their own control (before, during and after calving) and comparing cows with dystocia to eutocic calving. Moment (in days, D; D0 = day of calving) and type of change (↑ increase; ↓ decrease; NS = non-significant; NE = non-evaluated) are given comparing the painful moment or dystocic cows versus the control moment or eutocic cows.Studies ClassificationReferencePainful Moment or Cows in Pain Due to CalvingControl Moment or Control CowsLying Time (h/d)No. Lying Bouts (No./d)Lying Bout Duration (min/d)Activity (No. Steps/d)Cows as own control (before, during and after calving)[135]Around calving (from Day-1 to +2)Post-calving (from D+3 to D+7)NENENE↑[140]Calving day (24 h before calving)Pre-calving (from -24 h to D-4 to D-11)↓↑↓↑[58]Post-calving(D+4, D+5)Post-calving(from D+5 to D+28)↓NENENEDystocia vs. eutocia[146]Calving ease score of 3 (assistance from 2 or more people) or score of 4 (mechanical traction assistance)Calving ease score of 1 (unassisted) or 2 (assistance from 1 person)↑↑↓↓[135]Easy manual pullUnassistedNENENENS[157]Cows that calved in >70 minCows that calved in ≤70 minNS↑ (D0)↓ (D+2)↑ (D+1, D+2,D+4, D+5)↓ (Holstein D+1, D+4, D+5 and Jersey D0, D+2)Significant differences are established at p ≤ 0.05.8. Pain Associated with MetritisClinical metritis is an inflammatory process that affects all layers of the uterus during the early postpartum period in lactating dairy cows [161,162]. This condition is characterized by the presence of an abnormally enlarged uterus with a fetid red-brownish uterine discharge, with or without systemic signs of illness such as depression, anorexia, decreased milk yield and pyrexia within 21 d after parturition [49,162,163]. The incidence of clinical metritis ranges from 15% to 20%, although it could be higher, affecting up to 40% of cows, depending on the herd and lactation stage [164,165,166,167]. Metritis has been associated with reduced milk production, impaired reproductive performance and increased risk of culling during lactation [168,169,170,171,172,173]. Besides the negative economic implications, clinical metritis is a welfare concern. From several surveys performed using farmers and veterinarians, acute metritis in cows was given from 4 to 6 out of a 10-point score for pain [21,22,24,86]. In fact, it is known that cows with clinical metritis have shown physical signs of pain such as back arching at rectal examination, as compared with cows without clinical metritis, suggesting that clinical metritis may be associated with visceral pain [174].8.1. Alteration in Activity Patterns as a Result of Pain Due to MetritisAlteration in activity patterns due to metritis appears to be more pronounced in primiparous cows than in multiparous ones. Primiparous cows diagnosed with clinical metritis spent more time lying (10.1 h/day vs. 9.2 h/day, respectively) during one week after metritis diagnosis, when compared to primiparous cows without metritis [49]. In addition, reduced activity has been reported in cows diagnosed with metritis (as measured with neck-mounted accelerometers [145,175]). In grazing dairy cows, Sepúlveda-Varas et al. [176] also found that primiparous cows diagnosed with more than one health disorder (retained placenta, mastitis or metritis) spend more time lying and tended to have longer lying bouts in the days following calving, as compared with healthy cows. When studying multiparous cows, no differences were reported in terms of daily activity patterns between healthy cows and those with metritis [49,177]. During the early postpartum period, primiparous cows show an exacerbated inflammatory response (increased haptoglobin and serum amyloid A) when compared to multiparous cows [135,178]. This suggests that the immune system of primiparous cows may be more reactive to inflammatory stimuli and may lead to an increase in circulating cytokines. As cytokines play a key role in the sickness behaviour, the increased time lying associated with clinical metritis should be more evident in primiparous cows than in multiparous ones [49].Metritis seems to affect a cow’s activity pattern before metritis gets diagnosed. Neave et al. [54] found that cows later diagnosed with metritis spent less time lying and had fewer lying bouts in the two weeks before calving. During the three days before diagnosis, cows with metritis had fewer lying bouts and a longer bout duration, as compared with healthy cows. In addition, cows with metritis spent more time fully standing in the lying stall and had more failed lying events during the three days before metritis diagnosis, when compared to healthy cows [179]. No association between the occurrence of metritis or prepartum activity (measured as arbitrary units) was observed by Liboreiro et al. [145].Altogether, those studies indicate that metritis may produce visceral pain [174], which may be exacerbated during lying movements. Visceral pain may lead to somatic hyperalgesia (e.g., [180]), making skin areas sensitive to touch during lying movements and when lying. Future studies should benefit from the inclusion of hyperalgesia measurements in dairy cows with metritis.8.2. The Effect of NSAIDs on Activity Patterns in Cows with MetritisThe benefits of NSAIDs as a supportive treatment of metritis in addition to antibiotics show some discrepancies. Several studies reported that cows treated with NSAIDs after metritis diagnosis improve their reproduction performances by showing increasing pregnancy rates [181] and shortened calving-to-first-oestrus intervals [182]. However, carprofen administered alone [183] or NSAIDs administered in conjunction with antibiotics [181] to cows with metritis did not improve milk yield. Several studies also showed divergences in the effects of NSAIDs on uterine involution or recovery from uterine infection [182,184,185,186].Only one study on lying behaviour reports the effect of meloxicam administered as a complementary treatment to antimicrobial therapy as metritis treatment [187]. In multiparous cows, complementary meloxicam treatment was associated with decreased lying times during five days after treatment, but there was no treatment effect for primiparous cows. Lying bouts duration was not affected by treatment, and the number of lying bouts tended to show a lesser increase in cows treated with meloxicam, as compared with cows that received the placebo [187].9. Concluding Remarks and Potential Further ResearchThis narrative review shows that pain due to several health conditions in dairy cows can modify daily activity patterns, depending upon the cause, the severity and the moment.Cows with clinical mastitis reduced their time lying and increased the number of lying bouts and stepping. These activity pattern changes are due to pain caused by the swollen udder when cows are lying and to frustration from not being able to rest as long as necessary. Additional research is needed to observe whether their activity pattern changes with different severity of mastitis (including subclinical mastitis). Two studies have investigated the effect of flunixin meglumine on experimentally induced mastitis and reported that daily activity patterns were not affected by treatment. Additional research including the effect of other NSAIDs on experimentally induced mastitis on lying time includes a closer administration before and after challenge and/or a repeated dose administration. Additional research is guaranteed assessing the effect of NSAIDs on daily activity patterns in naturally occurring mastitis in dairy cows.Overall, lame cows show longer lying times, with a lower number of lying bouts and longer and more variable lying bouts duration, as compared to non-lame cows. Daily activity patterns are a potential useful indicator for severe lameness but show limitations for detecting moderate lameness. When the relationship between lameness and daily activity patterns is studied, bedding material and the inclusion of severely lame cows are important factors to take into consideration. Two studies using flunixin meglumine and meloxicam confirm that analgesia used in experimentally induced lameness and severe lameness show the welfare of cows is increased by increasing the lying time and decreasing stepping. However, inconsistent or non-clinically relevant outcomes have been reported for other field studies. Additional research on assessing lameness using daily activity patterns is recommended by reducing the clinical heterogeneity of lameness and assessing its degree of severity. Measures aimed to detect central sensitization due to chronic lameness are recommended.It is assumed that the increased activity around calving is due to pain. Deviation from normal daily activity patterns around calving (either increased or reduced) are indicative of additional pain due to dystocia or a cow delivering a stillborn calf. When pain at calving and daily activity patterns are studied, factors such as parity and the degree of difficulty at calving should be taken into consideration. Additional research studies that include an objective definition of severity of calving (for instance, including physiological indicators of inflammation) are needed. Overall, stepping is altered by NSAIDs treatment around calving, although it shows divergent results among studies. The potential benefits of treating periparturient cows with NSAIDs depend upon the type of drug employed and its dosage and administration mode, as well as the timing of administration relative to calving. NSAIDs that preferentially inhibit COX-2 activity should be preferred for use around calving as they are less likely to have the side effects described.Primiparous cows with metritis spent more time lying down and have reduced activity, whereas multiparous cows reduced their number of lying bouts. These activity pattern changes are associated with sickness behaviour and avoidance of pain during lying movements. Further research on clinical metritis and daily activity patterns has to take into consideration an objective definition of the severity of the metritis (including whether metritis compromises the cows systemically or only locally) and the inclusion of measures of hyperalgesia. We encourage researchers to perform more studies regarding the effects of the NSAIDs on daily activity patterns in cows with metritis.When the relationship between painful disorders and daily activity patterns is studied, factors such as parity, bedding type and severity of disease are important factors to take into consideration. In addition, future investigations should include indicators of sickness behaviour and hypersensitivity (hyperalgesia and allodynia) in order to properly interpret daily activity-pattern changes due to painful disorders in dairy cows. The potential benefits of the NSAIDs treatment in painful health disorders depend upon the type of drug administered, its dosage and administration mode, and the time of administration relative to the painful health disorder.

animals : an open access journal from mdpi

10.3390/ani11071923

PMC8300104

Herpesviruses (HVs) are a large family of DNA viruses infecting animals (including insects and mollusks) and humans. Cetaceans can be also infected by HVs presenting different range of lesions, from dermatitis to meningoencephalitis, or being asymptomatic. Several studies have addressed the question of HVs in cetaceans, although no previous systematic survey of HV in beaked whales (BWs) (Ziphiidae family) has been previously performed. The family Ziphiidae, which includes 22 species in 6 genera, is one of the most widespread families of cetaceans, with a strict oceanic habitat pattern. Beaked whales, Cuvier’s BW in particular, are one of the deepest diving whales and are of particular interest because of a notable relationship between military operations employing mid-frequency sonar and the mass stranding of BWs in different geographic areas, including the Canary Islands. In this study, we analyzed 55 BWs (294 samples) stranded in the Canary Islands from 1990 to 2017 by molecular methods (conventional nested polymerase chain reaction). Our results showed that 8 BWs were infected by HVs, although only three animals displayed lesions indicative of active viral replication. Phylogenetic analysis suggests that HV-BW sequences are species-specific, although more studies are needed to better address this question.

Herpesviruses (HVs) (Alpha- and Gammaherpesvirinae subfamilies) have been detected in several species of cetaceans with different pathological implications. However, available information on their presence in beaked whales (BWs) is still scarce. In this study, a total of 55 BWs (35 Ziphius cavirostris and 20 animals belonging to the Mesoplodon genus) were analyzed. Samples (n = 294) were obtained from BWs stranded along the coasts of the Canary Islands (1990–2017). Molecular detection of HV was performed by means of a conventional nested PCR based on the DNA polymerase gene. Herpesvirus was detected in 14.45% (8/55) of the analyzed BWs, including 2 positive animals from a previous survey. A percentage positivity of 8.57% was found within the Cuvier’s BW group, while the percentage of positivity rose to 25% within the Mesoplodon genus group (three M. densirostris, one M. europaeus, and one M. bidens). All the obtained sequences from this study belonged to the Alphaherpesvirinae subfamily, from which three are considered novel sequences, all of them within the Mesoplodon genus group. In addition, to our knowledge, this is the first description of HV infection in Gervais’ and Sowerby’s BWs. Three out of eight HV-positive BWs displayed histopathological lesions indicative of active viral replication.

1. IntroductionThe family Ziphiidae is one of the most widespread families of cetaceans, with a strict oceanic habitat pattern. The Canary Islands, due to its volcanic origin and oceanic location, hold resident and transient populations of Beaked whales (BWs), with records of six different species. Beaked whales, Cuvier’s BW in particular, are one of the deepest diving whales. Depths of more than 2900 m and dive durations of over 2 h have been recently recorded in Cuviers’ BWs during single breath-hold dives [1]. Most of the distribution information of BWs is based on stranding records. They face threats from entanglement in fishing gear, ingestion of marine debris, and ship collision, among others [2,3,4]. They are also of particular interest because of a notable relationship between military operations employing mid-frequency sonar and the mass stranding of BWs in different geographic areas, including the Canary Islands [5,6,7,8]. Natural pathologies affecting these species include verminous arteritis by Crassicauda spp. [9], brucellosis [10,11] and virosis (morbillivirus and herpesvirus (HV)) [12,13,14,15,16,17,18]. The number of publications concerning HV infection in cetaceans has been increased in the past decades. However, very little information is available regarding the presence of these viruses in BWs.The Herpesvirales order is split into three families: Alloherpes-, Herpes-, and Malacoherpesviridae. The Herpesviridae family is divided into three subfamilies (Alpha-, Beta-, and Gammaherpesvirinae), 13 genera, and 107 species [International Committee on Taxonomy of Viruses (ICTV) (https://talk.ictvonline.org/taxonomy/ (accessed on 3 September 2020)) [19]. A single, linear, double-stranded DNA characterizes these viruses, which can cause immunosuppression [20] and latent infections [21,22]. Moreover, HV can infect a wide range of hosts (mammals, birds, reptiles, amphibians, fish, mollusks, and insects) [23].In cetaceans, several pathological findings associated with HV infection (alpha- and gamma-herpesvirus) have been documented, although in some cases, HV-related lesions may not be present in the infected animals [24]. Specifically, alphaherpesvirus has been related to fatal systemic infections [25], lymphoid necrosis [17], interstitial nephritis [18] and encephalitis and meningoencephalitis [26,27,28]. Gammaherpesvirus, however, has been described as mainly associated with mucocutaneous, skin, and genital lesions [16,29,30,31,32,33], although it has been recently described the first detection of gammaherpesvirus in the central nervous system of several striped dolphins (Stenella coeruleoalba) stranded in the Cantabrian Sea, Spain [34].Herpesviruses have been found in many cetacean species, which are summarized in Table 1.There are some documented cases of HV coinfection in cetaceans. Herpes- and morbillivirus co-infection have been reported in striped dolphins from the Mediterranean and Atlantic coasts [24,35,45] and from the Canary Islands [26]. Some cases of herpes- and papillomavirus coinfections have also been reported in Atlantic bottlenose dolphins from the Atlantic coast of the USA and Cuba [46,47].To date, only seven HV-BW sequences, five alphaherpesviruses, and two gammaherpesviruses are available in GenBank. Concerning the alphaherpesviruses, most of them (3/5) have been detected in BWs stranded in the Canary Islands (two Cuvier’s BW (GU066291 and KY680659) and one Blainville’s BW (JN863234)). The only other BWs in which the alphaherpesvirus has been detected are three Cuvier’s BWs, two stranded in the Mediterranean Sea in 2012 (KP995682 and KP995685) and the other stranded on the Atlantic coast of the Spanish mainland in 2015 (KY680659).The present study aims to detect HVs (novel or already known) in samples from stranded BWs in the Canary Islands and to correlate this positivity with histopathology to identify lesions compatible with the HV infection. In addition, phylogenetic relationship between the obtained sequences and all the available ones, within the Herpesviridae family, detected in cetaceans will be performed.2. Materials and MethodsIn this study, 294 samples from 55 BWs stranded along the coasts of the Canary Islands, from November 1999 to May 2017, were analyzed. These BWs included 35 Cuvier’s BWs and 20 specimens belonging to the Mesoplodon genus. This last group consists of 2 Sowerby’s BWs (Mesoplodon bidens), 7 Blainville’s BWs, 10 Gervais’ BWs (Mesoplodon europaeus), and 1 True’s BW (Mesoplodon mirus). Two of these animals have been already published previously (CET 243 and CET 294) [17,18]. Adults were more highly represented, while males and females were present in a similar proportion, in both groups of BWs. This information and other biological parameters (stranding epidemiology (type, location, and date) and life history data (species, age category, and sex)) are summarized in Table 2 and Table 3. Five codes of conservation condition were established [48]: Code 1 (extremely fresh carcass, as an animal that has recently died or euthanized), Code 2 (fresh carcass), Code 3 (moderate decomposition), Code 4 (advanced decomposition), and Code 5 (mummified or skeletal remains).All the animals were submitted to a complete standardized necropsy [48,49], and representative tissue samples were collected for further analysis. For the histopathological study, collected samples were fixed in a 10% neutral buffered formalin solution, processed, and embedded in paraffin blocks, which were sectioned at 5 µm and stained with hematoxylin and eosin (HE). The slides were then visualized in an optical microscope with the objective to find HV-associated lesions. For molecular analysis, collected samples were frozen at −80 °C.A wide range of tissue samples, according to availability in each case, were analyzed for the presence of HV DNA by PCR: lung (16.33%; (48/294)), kidney (15.65%; (46/294)), brain (12.93%; (38/294)), skin (12.59%, (37/294)), liver (12.24%; (36/294)), spleen (9.86%; (29/294)), mesenteric lymph node (8.16% (24/294)), skeletal muscle (6.12%; (18/294)), intestine (2.04%; (6/294)), prescapular lymph node (1.36%; (4/294)), mediastinal lymph node [0.68%; (2/294)], thyroid gland (0.34%; (1/294)), thymus (0.34%; (1/294)), palate (0.34%; (1/294)), and esophagus (0.34%; (1/294)). In addition, blood was analyzed in one animal (0.34%; (1/294)) (Table 2 and Table 3).Thawed samples were mechanically macerated in a lysis buffer and centrifuged. DNA/RNA extraction was simultaneously carried out from each 300 μL macerated sample by pressure filtration, by means of a QuickGene R Mini 80 nucleic acid isolation instrument, with the DNA Tissue Kit S (QuickGene, Kurabo, Japan) according to the manufacturer’s instructions with some modifications: An RNA carrier (Applied BiosystemsTM, Thermo Fisher Scientific Waltham, MA, USA) was added during the lysis step, as previously published [50].A panherpesvirus conventional nested polymerase chain reaction (PCR) was performed for HV detection, amplifying a fragment of the DNA polymerase gene of the Herpesviridae family of about 200 bp [51]. Two negative controls (non-template) for extraction and amplification and an amplification-positive control (known herpesvirus DNA previously obtained in our laboratory) were included in each protocol. Horizontal gel electrophoresis, in 5% agarose containing GelRed® (Biotium, Inc. California, USA), was performed for 5 µL of the obtained amplicons from the second PCR. Purification of PCR products was carried out using a Real Clean spin kit (REAL®, Durviz, s.l.,Valencia, Spain) to perform sequencing (the Sanger method).Furthermore, a reverse transcription real-time polymerase chain reaction (RT-qPCR) based on SYBRN® Green dye (Bio-Rad Laboratories, Inc., California, CA, USA) [52] was performed for Cetacean morbillivirus (CeMV) detection only in those animals that were positive for HV.The obtained HV sequences were compared with similar sequences retrieved from GenBank via a Blast search with the blastn algorithm (www.ncbi.nlm.nih.gov/blast/Blast.cgi/(accessed on 3 May 2021)) [53]. ClustalW was used to perform the HV multiple sequences alignment using MEGA X software (Pennsylvania, PA, USA) [54]. To construct the phylogenetic nucleotide tree, 73 alphaherpesvirus sequences were retrieved from GenBank. To root the phylogram, nine gammaherpesvirus sequences were used as the outgroup. The best substitution model for the nucleotide phylogenetic tree analysis was selected based on its lowest BIC score (Bayesian Information Criterion). Accordingly, a phylogenetic tree was constructed using the Maximum Likelihood Method and the Tamura 3-parameter with a discrete Gamma distribution to model the evolutionary rate differences among sites (5 categories (+G, parameter = 0.7779)). Bootstrap resampling (1000 replicates) was used to assess the reliability of the tree.The nucleotide sequences were translated to the deduced amino acid sequences (70 alphaherpesvirus and 9 gammaherpesvirus sequences retrieved from GenBank). The deduced animo acid phylogenetic tree was built using the best substitution model based on its lowest BiC, a Maximum Likelihood Method and the Jones Taylor Thornton matrix-based model with a discrete Gamma distribution to model the evolutionary rate differences among sites (5 categories (+G, parameter = 17,401)). A bootstrap consensus tree for 1000 replicates was also performed.A consensus tree was computed accepting the default 50% cut-off value (nodes supported by <50% of bootstrap replicates are collapsed), as previously proposed [55]. Only those bootstrap values equal or greater than 70% were considered valid among the remaining nodes (>50%). Furthermore, all sequences obtained from this study were submitted in GenBank, whose number accessions are from MZ066758 to MZ066765.3. ResultsEight out of the fifty-five analyzed BWs (14.45%) had a positive result of the PCR test. Specifically, HV was detected in adult individuals of three Ziphius cavirostris and five animals belonging to the Mesoplodon genus (three M. densirostris, one M. europaeus, and one M. bidens), from which two had been previously published [17,18].Moreover, these viruses were detected in 15 out of 294 (5.1%) analyzed samples. HV-positive samples were, by decreasing frequency, as follows: lungs (5/48, 10.41%), kidney (5/46, 10.89%), brain (2/38, 5.26%), liver (1/36, 2.77%), spleen (1/29, 3.44%), and prescapular lymph node (1/4, 25%).3.1. Molecular FindingsHerpesvirus was detected in 3 out of 35 Cuvier’s BWs (8.6%). Specifically, Cases 3, 5, and 8 were positive for HV in one or more tissues (Table 4) (Figure 1A). Case 3 (CET 294), an adult female in advanced decomposition, showed positivity for HV in spleen and lung samples [17]. In Case 5 (CET 771), an adult female in good state of preservation (code 2), HV was found in brain tissue; in Case 8 (CET 855), an adult male in moderate decomposition, the positivity was found in the lung. Thus, a total of 4 samples out of 168 (2.4%) were positive for HV in the Ziphius cavirostris group. By decreasing frequency, HV was detected in lung (2/30, 6.6%), brain (1/21, 4.8%), and spleen (1/17, 5.9%) samples (Table 4). The size of the new sequences obtained in the present study, (Cases 5 and 8) ranged from 210 to 234 bp.Five out of twenty animals (25%) were positive for HV within the Mesoplodon genus group (Cases 1, 2, 4, 6, and 7) in one or more tissues (Figure 1B–D). Specifically, a total of 11 samples out of 126 (8.7%) were positive for HV. By decreasing frequency, HV was detected in kidney (5/17, 29.4%), lung (3/18, 16.6%), liver (1/18, 5.5%), brain (1/17, 5.9%), and prescapular lymph node (1/1, 100%) samples (Table 4). HV was found in lung and kidney samples from Case 1 (CET 243), a very fresh (live stranded individual, which subsequently died) adult male of M. densirostris [18], Case 4 (CET 379), an adult male in code 2 of M. bidens, and Case 7 (CET 852), an adult female in code 2 of M. densirostris. Herpesvirus was detected in just one tissue (kidney) in Case 2 (CET 259), an adult female in code 2 of M. europaeus. In Case 6 (CET 824), an adult male in code 2 of M. bidens, HV was found in liver, prescapular lymph node, kidney, and brain tissues. The size from the obtained six new sequences from our study (Cases 2, 4, 6, and 7) ranged from 194 to 234 bp. None of the HV-infected animals were positive for CeMV. Phylogenetic analysis showed that all the sequences obtained from both groups of BWs belonged to the Alphaherpesvirinae subfamily.3.1.1. Nucleotide IdentityEight sequences obtained from this study were new; two were previously published [17,18]. Nucleotide similarities are summarized in Table 5 and Table 6. The criteria to considerer a novel sequence are that the sequence are ≥100 bp long and has a <90% identity to the reference genome [56]. Based on this, the sequences from Cases 2 and 4 and the lung sample from Case 7 (7a) can be considered novel, as presented a 78.35%, 89.32% and 78.30%, similarity, respectively. Sequences from the kidney and lung in Case 1 (JN863234) were considered novel when they were published [18], although they currently show a high percentage of identity with a sequence detected in the brain of a striped dolphin stranded in the Mediterranean Sea in 2011. A novel sequence was obtained from the kidney in Case 2 (MZ066758); the sequence most closely related to this novel sequence was a previously published sequence detected in the prescapular lymph node of a Cuvier’s BW stranded in the Mediterranean Sea in 2012. Sequencing and further comparison with GenBank records showed a novel sequence highly related to cetacean alphaherpesvirus in the lung and spleen samples from Case 3 (GU066291) when they were published [17]. However, these sequences are currently identical to four previously described sequences in striped dolphins. A novel sequence was amplified from the lung and kidney in Case 4 (MZ066759), showing the highest similarity with sequences amplified from the same organs in Case 1 from our study. Two different sequences were obtained in Case 6, one sequence from the liver, prescapular lymph node, and kidney (6a) (MZ066761) and another from the brain (6b) (MZ066761). Two different sequences were also obtained from Case 7, one from the lung (7a) (MZ066763) and other from the kidney (7b) (MZ066764). Sequence 7a is considered novel.The nucleotide phylogenetic analysis (Figure 2A) showed that sequences from Cases 1 (kidney), 4 (lung and kidney), 6a (liver, prescapular lymph node, and kidney), 6b (brain), and 8 (lung) clustered together in a clade supported by a bootstrap value of 64, with two subclades: one containing sequences from Cases 1, 4, 6a, and 6b and a sequence from a striped dolphin stranded in the Mediterranean Sea in 2011 (KP995684) (84 bootstrap value) and another containing two sequences, one from Case 8 and another from a Cuvier’s BW stranded in the Mediterranean Sea in 2012 (KP995682) (79 bootstrap value).The sequence from Case 7 (lung) did not cluster with any sequence within the nucleotide phylogenetic tree, nor did the sequence from Case 2 (kidney), even if both of them belong to a large clade (62 bootstrap value) containing all the cetaceans alphaherpesvirus published until now.However, the sequence from Case 3 (lung and spleen) clustered with sequences detected in striped dolphins stranded in the Mediterranean and Atlantic coasts (92 bootstrap value); sequences from Cases 5 (brain) and 7 (kidney) clustered together (97 bootstrap value) and with a sequence (blowhole swab) from a beluga whale (MF678601) (100 bootstrap value).3.1.2. Amino Acid IdentityBlast analyses of translated amino acid sequences showed similar results to those of the nucleotides. However, some differences were observed for Cases 1, 4, 6a, and 7a. Case 1 showed the highest similarity (79.13%, 100 QC) with a sequence detected in a beluga whale (ANG08598). Sequences from the lung and kidney in Case 4 were very similar (95.59%, 87% QC) to a sequence detected in the brain of a striped dolphin stranded in the Mediterranean Sea in 2011 (ALP00300), and to a sequence detected in a Cuvier’s BW stranded in the Mediterranean Sea in 2012 (ALP00298) (94.12%, 87% QC). The sequence from the prescapular lymph node, liver, and kidney in Case 6 (6a) was very similar (97.06%, 99 QC) to sequence ALP00300. The lung sequence from Case 7 showed the highest similarity (72.73%, 94% QC) with a sequence detected in the skin of a Cuvier’s BW stranded in the Mediterranean Sea in 2012 (ALP00292). Amino acid similarities are summarized in Table 5 and Table 6.Phylogenetic analysis (Figure 2B) showed that the tree based on deduced amino acids consists of 45 alphaherpesvirus branches and a root that contains nine gammaherpesvirus sequences. All the obtained sequences from our study take part of a large polyphyletic clade within which seven sequences (Cases 1, 2, 4, 6a, 6b, 7a, and 7b) did not form sub-clades with any sequences of any of the previously identified HV in cetaceans; while sequences from Case 5 (brain) and Case 8 (lung) clustered together with a sequence obtained from the prescapular lymph node of a Cuvier’s BW stranded in 2012 in the Mediterranean Sea (ALP00298) (bootstrap value of 81). Finally, the sequence from Case 3 clustered together with sequences obtained from striped dolphins stranded in the Mediterranean Sea and the Atlantic coasts of Spain (mainland and Canary Islands) and the Cantabrian Sea (54 bootstrap values).3.2. Gross and Histopathological FindingsNone of the HV-positive animals showed gross lesions associated with Herpesvirus infection, except Case 7, which presented several ulcerative and well-defined round skin lesions.At the histopathological level, three animals presented lesions attributable to HV infection, one within the Ziphius cavirostris group and two within the Mesoplodon genus group. Specifically, Case 3 (Z. cavirostris) presented diffuse lymphoid and splenic necrosis with intranuclear inclusion bodies in monocytes [17]; Case 1 (M. densirostris) [18] and Case 7 (M. densirostris) displayed similar lesions, characterized by membranous glomerulonephritis and lymphoplasmacytic interstitial nephritis at the cortico-medullary region, moderate multifocal interstitial and tubuloepithelial necrosis with the presence of intranuclear inclusion bodies within tubuloepithelial cells in the renal medulla; mainly in the blood capillary network defined as the vasa recta of the kidney (vasa rectae renis) (Figure 3).Additional findings in the other positive HV-BWs were as follows: mild multifocal parasitic bronchopneumonia with the presence of intraluminal nematodes in Cases 1 and 4; generalized lymphadenopathy and lymphoplasmacytic interstitial nephritis in Case 4; severe parasitic nephritis and verminous mesenteric arteritis by Crassicauda sp., a moderate multifocal suppurative bronchopneumonia, and moderate multifocal suppurative lymphadenitis in Case 5; periductal fibrosis with lymphoplasmacytic pericolangitis in the liver, as a result of severe parasitic infection, in Case 6; the presence of severe hyaline membranes and mild multifocal interstitial bronchopneumonia in Case 8.In addition, Case 2 also displayed several skin lacerations in the caudal peduncle produced by fishing tackles and diffuse congestion and hemorrhages in the lungs, liver, kidneys, and adrenal glands, and Case 6 showed active stranding related lesions, consistent with skin lacerations, a hypercontraction of muscle fibers, and hyaline globules within hepatocyte cells.4. DiscussionThis study represents the first systematic survey of HV infection in cetaceans from the Ziphiidae family. Beaked whales are found in all oceans and are of particular interest because they are one of the deepest diving whales and there is a proven relationship between several mass stranding events of BWs and military operations employing mid-frequency sonar [1,7,8].We have detected HV in 14.45% (n = 8) of the surveyed BWs (n = 55). A percentage positivity of 8.57% (3/35) was found within Cuvier’s BW group, while the percentage of positivity rose to 25% (5/20) within the Mesoplodon group. Previously published prevalence of HV infection in stranded cetaceans include: 5.3% in the Western North Pacific (Japan) [40], 3.7% in Brazil [44], 7.8% in Portugal [35], and 41.2% in Cantabria, Spain [34]. Most of the references of HV infection in BWs are case reports: two alphaherpesviruses in a Cuvier’s BW and a Blainville’s BW stranded in the Canary Islands [17,18] and one gammaherpesvirus in a Blainville’s BW stranded in the USA [16]. The detection of a gammaherpesvirus in a Stejneger’s BW from the Japanese coasts took part of a survey of 76 stranded cetaceans that include 4 BWs (one Stejneger’s BW and three Blainville’s BWs) [40]. In addition, three sequences from the brain, skin, and prescapular lymph nodes of Cuvier’s BWs are available in GenBank (KY680659, KP995685, and KP995682).All the sequences obtained from this study belonged to the Alphaherpesvirinae subfamily in contrast to similar previous studies, in which both alpha and gammaherpesvirus were detected [27,34,35,40,44].The Ziphiidae includes 22 species in 6 genera, being the second largest family of cetaceans after the Delphinidae [57]. Herpesvirus sequences detected in BWs from our study should be then analyzed by species. Specifically, sequences from Cuvier’s BWs are not novel sequences, displaying higher homologies (100–98.06%) with sequences from striped dolphins stranded in the Mediterranean Sea and central and northeast Atlantic coasts (the Canary Islands and Portugal) (MG437217, KY680657, KY680656, and KJ156331) [26,34,35] (Case 1) and with a sequence detected in other Cuvier’s BW from the Mediterranean Sea (KP995682) (Cases 5 and 8). The sequence KP995682 can be considered novel since it shows the highest homology (83.98%) with the sequence detected in the brain of a striped dolphin stranded in the Mediterranean Sea in 2011 (KP995684), which in turn showed the highest homology with the sequence from Case 1 in our study (95.5%) and with the sequence KP995682 (83.98%). This relationship could indicate that HV transmission has occurred between these proximal regions, as previously suggested for other alphaherpesviruses [26,35,44] and between these two different species. Cuvier’s BWs are found in most oceans and seas worldwide (in temperate, subtropical, and tropical waters), and have the most extensive range of all BWs species, although the seasonality and migration patterns of this species are still unknown [58].Regarding the Mesoplodon genus, sequences from the M. densirostris species are, in general, highly (97.4–93%) related to a sequence detected in the brain of a striped dolphin stranded in the Mediterranean Sea in 2011 (KP995684) (Cases 1, 6b, and 7b). As we mentioned before, this sequence, even if detected in a striped dolphin, showed its highest homologies with sequences detected in members of the Ziphiidae family, suggesting the idea of HV transmission from BWs to the striped dolphin species rather than the contrary. The only sequence from Case 7a is considered novel within the M. densirostris species in our study, being related to the sequence obtained from the penile skin lesion of a beluga whale stranded in St. Lawrence Estuary (Canada) (KF155406). This beluga sequence was also considered novel and tentatively named beluga whale herpesvirus [36]. The relationship (geographical and interspecies) between these two sequences is, for the moment, unknown. Blainville’s BWs are little-known members of the Ziphiidae, living in tropical to temperate waters worldwide. There is little information on the abundance of Blainville’s BWs worldwide, although they are considered to have the most extensive distribution of any whale in the Mesoplodon genus [59].Two novel sequences were obtained from M. europaeus (Case 2) and M. bidens (Case 4) species. Sequence from Case 2 was related to sequence KP995682, detected in a Cuvier’s BW from the Mediterranean Sea; as most of the Cuvier’s BWs from our study; while sequence from Case 4 was related to sequence from Case 1 (JN863234), detected in a Blainville’s BW stranded in the Canaries and taking part of this study. Interspecific interactions between these four species of BWs could explain HV transmission within the Ziphiidae family. Gervais’s and Sowerby’s BWs are little-known members of the Ziphiidae. Both species of BWs are distributed throughout the Atlantic Ocean, although it is unknown if they undertake seasonal movements or migrations [60,61]. The presence of both BWs species in the Canary archipelago is reported as sporadic [62,63].Regarding the nucleotide phylogeny of the reported HV-BW sequences from our study (n = 10), most of the sequences (80%, 8/10) clustered with a sequence previously identified in a Cuvier’s BW, but also from other species, such as striped dolphins or beluga whales. Specifically, the sequence from the lung in Case 7 (7a) clustered with the sequence from Case 5 and with a sequence detected in the blowhole swab of a beluga whale (MF678601) in the nucleotide tree. The phylogenetic analyses showed that the virus isolated from this beluga whale grouped with members of the genus Varicellovirus, in the subfamily Alphaherpesvirinae, and it was tentatively named Monodontid alphaherpesvirus 1 [64]. Thus, although most of the HV-BWs sequences from our study are quite host-specific, as previously suggested for the members of Herpesviridae [65], it seems that there is a possible interspecific transfer of these viruses. The obtained results from the amino acid phylogenetic analysis showed, however, that sequences from animals within the Mesoplodon genus group (70%, 7/10) did not cluster with any previously identified sequence. This discrepancy between trees could be due to the presence of too few characters in the aminoacid dataset, being the phylogenetic signal for the tree reconstruction, in accordance, too small. Sequences from the Cuvier’s BW group (n = 3) clustered with a sequence previously detected in a Cuvier’s BW (20%, 2/10) or a striped dolphin (10%, 1/10). However, apart from this study, there are no alphaherpesvirus sequences from animals within the Mesoplodon genus available in GenBank. More and larger sequences will be needed to better understand the species specificity of HV-BWs. In conclusion, from the phylogeny analyses, it can be observed that the obtained BW sequences from this study are more often closely related to each other and occasionally with sequences from other cetacean species, specifically striped dolphins and beluga whales. However, the short partial sequences of the catalytic subunit (UL30) of the DNA polymerase available from this and other studies, only allow for subfamily identification [66]. Further studies are needed to better understand the phylogenetic relationship between these and other BW sequences within these species.In addition, to our knowledge, this is the first report of HV infection in two species of BWs: Gervais’ and Sowerby’s BWs, respectively.Two different sequences were obtained within the same animal in Cases 6 and 7. Coinfection with different viral strains seems to be a common feature of HV infection in cetaceans as it has been previously reported in several studies [24,26,30,35].Most HV-BWs in our study were detected in the lung and/or kidney, representing a percentage of 21.3% considering all the analyzed samples. However, if we consider only positive samples, this percentage increased to 66.7%. Moreover, HV was detected in the lung in 62% of the positive BWs (5/8) and the kidney in the same proportion. However, if we consider the two groups of BWs, HV was detected in the kidney in 100% of HV-positive animals within the Mesoplodon group of our study. A minor proportion of samples were positive for HV, specifically the brain, prescapular lymph node, liver, and spleen, being the first identification of an HV DNA sequence in a liver sample from a BW.The capacity of HV to cause disease is uncertain displaying the broad pathogenic and epidemiological features of the disease. In cetaceans, there have been descriptions from asymptomatic cases of HV infections to systemic and/or central nervous system infections. No gross or histopathological lesions attributable to the virus infection were observed in most of the positive samples in our study (62.5%). In addition, despite the presence of skin lesions compatible with herpesvirus infection in Case 7, no HV DNA was detected in the corresponding sample. All the HV-infected BWs from our study were adults, although only three out of eight (37.5%) displayed histopathological lesions indicative of active herpesviral replication, consistent with previous publications [17,18]. It is well known that the role of the viral factors in the course of a infection is both determined by viral and host factors, including the immune status itself or in combination with the age of the individuals [67,68]. In this study, we described multifocal interstitial nephritis, tubuloepithelial necrosis, and the presence of several intranuclear inclusions in Case 7, lesions that are very similar to those previously described in other M. densirostris stranded in the Canary Islands 13 years earlier [18]. The animal from Case 3 (CET 294) presented an advance stage of decomposition (code 4), which could partially impair DNA integrity for PCR detection, as previously published [35]. However, HV was detected in lung and spleen samples, which showed severe histopathological lesions and large number of intranuclear inclusion bodies indicative of early stages of the infection and high viral load, allowing molecular detection and identification of the virus. In a similar way, HVs were detected in code 4 cetaceans, specifically two mysticeti species stranded in the Mediterranean Sea [42] and in an Atlantic spotted dolphin stranded in Brazil [44].5. ConclusionsThis research describes the presence of HV in BWs stranded in the Canary Islands over a 19-year period (1999–2017) by molecular methods. Our results showed a prevalence of positive BWs of 14.45% (8/55), representing the first systematic survey of this pathogen in BWs. Three out of eight HV-positive BWs displayed histopathological lesions indicative of active viral replication, which is in concordance with the latent period of most herpesviruses. However, HVs are also capable of causing severe disease in association with other pathogens, such as CeMV. No CeMV infection was detected in any of the HV-positive BWs, highlighting the potential disease-causing capacity of these viruses as primary pathogens. Eight new HV sequences were detected in this study, which were analyzed and compared to all HV existing sequences in cetaceans. Most of these sequences did not cluster with any other sequences in the amino acid phylogenetic trees, indicating a possible species-specificity in BWs; although testing for clustering and host specificity would need more tailored analyses. In addition, three novel sequences of a partial fragment of the conserved DNA polymerase of HVs are described, all of them within the Mesoplodon genus group. To our knowledge, this work is the first to describe herpesvirus infection in two species of BWs: Gervais’ and Sowerby’s BWs.

animals : an open access journal from mdpi

10.3390/ani12040465

PMC8868563

Environmental noise influences the behavioral patterns of animals. However, few quantitative studies have evaluated the effects of ship noise on wintering waterbirds in lakes. In this study, the effects of ship noise simulated by noise playback at different intensities and interference distances on the behaviors of the bean goose, a wintering waterbird species, were evaluated. Sensitivity to noise was higher in small populations than in large populations. Noises of >70 dB at distances of <100 m and >80 dB at <200 m clearly altered the flight patterns of bean geese. This study provides insight into the tolerance of endangered and protected waterbirds to environmental noise and may guide the development of strategies to minimize the impact of ship noise.

Wild animals are vulnerable to environmental noise. In wetlands, wintering waterbirds are easily disturbed by ship noises; however, the behavioral changes of waterbirds in response to different levels of noise are unclear. We simulated the acoustic environment created by ship movement to investigate the effects of ship noise on foraging, vigilance, and flight behaviors of the wintering bean goose (Anser fabalis). In particular, we used a noise playback method to simulate the acoustic environment created by ship operations at various noise levels (i.e., background noise <50 dB, 60, 70, 80, 90, and 100 dB), distances from the noise (i.e., short <100 m, medium 100–200 m, and long distances 200–300 m), and noise duration (i.e., short 0–1 min, medium 2–3 min, and long 4–5 min). Results indicated that the noise intensity and interference distance had obvious influence on the bean geese behavior, but the noise duration had no effect. Smaller populations (N ≤ 30) were more sensitive to noise interference. As the noise level increased, the frequency of foraging behavior decreased and the frequencies of vigilance and flight behaviors increased, particularly above 70 dB. For noises >70 dB at short disturbance distances and >80 dB at medium disturbance distances, flight behavior increased significantly. These findings suggested that ships should keep a distance of more than 200 m from waterbirds to reduce noise interference.

1. IntroductionNoise is a very common form of disturbance with harmful effects on many species, e.g., causes stress responses of animal behaviors [1,2], reducing foraging efficiency, and social communication [3,4]. To minimize noise interference, animals can change temporal or spatial patterns of behavior [5,6], change the intensity of behaviors, or adopt alternative behaviors [7].Behavior is a common indicator of the tolerance of birds to noise disturbance [8,9]. Noise can affect the availability of sound information to birds and alter foraging and vigilance [10]. For example, in captivity, zebra finches (Taeniopygia guttata) spend more time on vigilance in noisy areas than in quiet areas when feeding, resulting in a low foraging efficiency [6]. In response to simulated traffic noise, when at 80 dB, the thrushes (Garrulax canorus) showed obvious retreat behavior, then gradually distributed to the region with low noise intensity [11]. Most of these studies have been conducted in laboratories with caged birds, thus this approach may ignore important ecological factors and thereby the results may not be generalizable to field conditions. Meanwhile, through field experiments, when simulating railway noise the black-necked crane (Grus nigricollis) gradually moves away from the noise source at approximately 60 dB of noise, with an escape distance of approximately 60–80 m from the noise source [12]. In addition, highway noise causes avoidance behavior of black-necked cranes, and the average avoidance distance is about 135.18 m; the closer to the highway, the easier it is to be startled [13]. Further, when cars drive by at a nearer distance of 150 m from the road, mallard (Anas platyrhynchos) are severely affected by the noise and show flight behavior, while at a farther distance of 360 m, swan geese (Anser cygnoides), bean geese (Anser fabalis), and grey crane (Grus grus) were only slightly disturbed [14]. There are also research findings that show the influence area of road noise on different birds was as far as 189.63 m, but the avoidance distance of birds had no relationship with population size. However, there are other findings that demonstrate that population size has an effect on the vigilance behavior of birds. For example, when there is no noise, the proportion of vigilance individuals of hooded crane (Grus monacha) decreased with population increase. Moreover, when hooded crane population sizes increased 90 individuals, the percentage of vigilance individuals dropped to a minimum, and remained essentially unchanged [15]. Most of these studies focused on field observations and research conducted on roads and railways on the behavior of birds, or combined with habitat change, population size, human interference, and other forms of interference [16]. However, few studies have evaluated bird behavioral responses to ship noise. In particular, quantitative studies of the influence of noise intensity on the behaviors of wintering waterbirds based on simulated ship noise are lacking. At present, many wintering waterbirds depend on the wetlands, since the disturbance caused by ship noise is affecting their habitat and living environments. Therefore, it is necessary to investigate the ecological impact of ship noise on overwintering waterbirds.Recent research has shown that container ships, oil tankers, ferries, and bulk carriers are the biggest sources of shipping noise at present and in future vessel traffic services [17,18]. Some ship equipment can emit a high level of airborne noise [19]. On a typical ship, there are ubiquitous 24 h-day high noise environments, with noise levels ranging from 87 to 102 dB [20]. Noise pollution from ships with frequent traffic, such as fishing and patrol ships [21,22], has become one of the main types of environmental pollution in rivers, lakes, and other water areas, and has gradually become one of the main hazard sources that harm the water environment and destroy the water ecosystem. Therefore, it is necessary to conduct simulated ship noise experiments to evaluate the impact of ship noise pollution by observing the behavior patterns of overwintering waterbirds.Bean geese are a common waterbird species wintering in the lakes of the middle and lower Yangtze River floodplain with the largest populations and widest distribution [23]. Their habitats are mainly in the shallow waters of lakes, lakeside meadows, and farmlands and are often mixed with other swimming birds (e.g., spot-billed ducks (Anas zonorhyncha) and mallards), waders (e.g., little egrets (Egretta garzetta), and herons (Ardea cinerea)). In this study, we focused on its behavioral responses to ship noise with the hope of providing a scientific basis for shipping management. We simulated ship noise and observed the effects of the ship noise source, distance, and duration on the behaviors of bean geese. We then analyzed the relationships between noise-related parameters and the flight and other behaviors. Finally, the lowest noise intensity and distance were determined, and the ecological impact of noise on waterbirds was evaluated.2. Materials and Methods2.1. Ethics StatementThe ship noise simulation experiment did not involve bird catching or hunting. Approval was obtained from the local wildlife protection departments. The research process complies with current Chinese laws.2.2. Study Site SelectionThe Shengjin Lake National Nature Reserve (30°15′–30°30′ N, 116°55′–117°15′ E) in Anhui Province is an important gathering area for waterfowl in the middle and lower Yangtze River floodplains [24,25]. Every October, a large number of bean geese arrive at the Shengjin Lake for the winter seasons. The number reaches the annual maximum in mid-early December and decreases from March to April of the following year. With the approval of the National Reserve Management Office, the research site was the farmland area of the Lianhe Village in the experimental area of the reserve. The study area had a large number of shrubs up to 1 m, providing good cover conditions for noise playback equipment and observers. The site was far from residential areas, with no human activities, vehicle traffic, and other disturbances. Daytime ambient noise met level one acoustic environmental quality standard (55 dB).2.3. Typical Ship Noise Recording and Playback DesignTo obtain ship noise for the experiment, a Philips VIR8800 recording pen (Shenzhen Jinghua Electronics Co., LTD, Shenzhen, China) and a high-precision noise tester (AS844+) (Shenzhen Xima Yinghao Trading Co., LTD, Shenzhen, China) were used to monitor the noise of 1000 t ships on Nanfei River, Hefei city, Anhui Province. The equivalent sound level of 5 min at 1 m distance from the exhaust cylinder was recorded synchronously by noise meter. The test method refers to the HJ 640-2015 “Technical Specification for Environmental Noise Monitoring Routine, Monitoring of Urban Environment Noise”. According to the field measurement, the noise level at 1 m of the upstream full load ship of 1000 t class was 97.4 dB; therefore, the noise was selected as the noise source for simulations. Adobe Audition CS6 audio editing software (Adobe Systems Incorporated, San Jose, CA, USA) was used to select the typical time of noises, in addition to exhaust cylinder noise, and to eliminate other sudden sounds, such as whistles. When listening through headphones, the recording was rechecked and 2 h of noise was edited to simulate the noise in the field experiment.2.4. Distance between the Noise Source and the Distribution of Bean GeeseBean geese remain mostly in the paddy field or forage on the beach. When observers approached, they flew away. Therefore, the principle of triangulation was adopted to calculate the nearest distance between the animals and noise source (Figure 1). The three vertices were bean geese distribution area (point A), noise equipment (point B) and reference point (point C). An electronic Total Station was used to measure the two interior angles, α and β, of vertex B and C in the triangle, and measure the distance L between point B and point C, as well as calculate the distance, D, between point B and point A. By using this method, noise equipment could be arranged in the area far from the distribution area of bean geese, which could avoid the influence on the waterbirds as much as possible and meet the requirements of this experiment. The calculation formula is as follows: D=L×sinβsin(α+β)2.5. Noise Disturbance and Behavioral ObservationsA preliminary experiment was conducted from December 9 to 11, 2019. Frequent activity areas of bean geese were first identified and the main behaviors during the wintering period were observed. We found that if there was no human disturbance, the foraging and resting behavior of bean geese generally began in the morning and continued into the evening in the same area. If disturbed by traffic, honking, or pedestrians, the geese would stay away from, or fly out of, their original area. By simulating ship noise in the field, when the noise level reached 100 dB and the horizontal distance was more than 300 m, the researchers heard the noise below the background value (dB < 50).The formal experiment was conducted from December 12 to 28, 2019. Under normal conditions, the observation time was 8:30 to 17:30, or when the bean geese flew away from the grassy beach. Before the experiment, an area with a high concentration of bean geese was selected, and the population size was in the range of 10–60, far from areas with substantial human interference, such as village roads, to ensure that the subject was only disturbed by the noise of a single ship noise, and no other disturbances. The noise equipment SAST/A60 (Shenzhen Xianke Enterprise Group, Shenzhen, China) was placed in a suitable position near the typical area, and the ship sound was oriented towards the study area. Environmental background noise was measured in the test area. Owing to the reduced noise sources in the study area in the winter, the equivalent sound level was measured for 5 min, revealing that the background noise in the test area was lower than 50 dB. Noise playback [26,27,28] was used to simulate the sound of ship operations. Six noise level gradients were set: control group (ZR, background noise) and experimental group (60, 70, 80, 90, and 100 dB) obtained by adjusting the volume of the audio equipment 1 m away from the playback device. The behavior observation time was 30 min (i.e., continuously playing 6 gradients of noise for 5 min each) or the time taken for all observed objects to exhibit flight, and the observations were made at three equal intervals in each noise period (i.e., at 0, 2, and 4 min). In the experiment, the observers and the noise equipment were hidden, and kept a certain distance from the subjects with no contact. Before playing the noise, relatively concentrated bean geese were selected to ensure that all geese could be observed. Then, the instantaneous scanning method was used to ensure that the behavioral responses of all experimental subjects could be observed. Binoculars (SWAROVSKI EL 8.5 × 42) (Swarovski (Shanghai) Trading Co., LTD, Shanghai, China) and monoculars (SWAROVSKI ATS80HD+20-60) (Swarovski (Shanghai) Trading Co., LTD, Shanghai, China) were used to observe and record the six types of behaviors in bean geese [29]: foraging (head lowering, moving slowly, pecking at ground food, approaching or stopping at the water surface, and drinking several times), grooming (bending neck toward the back, moving beak back and forth to sort out feathers), rest (bending legs together, sticking chest and abdomen on the ground or the water surface, neck shrinkage), moving (a body displacement, two legs moving frequency to speed up, moving over a long distance in a short time), warning (neck elongation, head scanning the surrounding environment or watching the interference source), flight (panic, constantly emitting cries, flapping wings quickly, spreading wings, flying in the opposite direction to the interference source or moving over a long distance and stopping). The observation time was 30 min when six noise level gradients were played, or all observed objects had time to exhibit flight. After the experiment was completed, a new population of bean geese was observed following the same methods.The disturbance distance was divided into short (<100 m), medium (100–200 m), and long distances (200–300 m). In actual observations, noise transmitted to the bean geese at greater than 300 m was close to the level of natural noise. The duration of interference was divided into three stages: T1 (0–1 min), T2 (2–3 min), and T3 (4–5 min). The population sizes of bean geese were set to N (0 < N ≤ 60), N1 (N ≤ 30), and N2 (30 < N ≤ 60) [29]. In the process of recording bean geese behaviors, data for other sources of interference were eliminated, such as vehicles, horns, and barking. In the natural environment, the activity range of bean geese is large, and they may disperse. Therefore, groups with good visibility and high concentrations were selected as the observation objects. Individuals exhibiting each behavior were counted as follows:X=nN×100%where X represents the proportion of bean geese exhibiting a certain behavior, n represents the number of bean geese exhibiting a certain behavior, and N represents the total number of bean geese observed.A total of 30 noise simulation experiments were carried out, and a total of 495 behavioral observations were collected (population size N), including 251 (population size N1) and 244 (population size N2).2.6. Statistical AnalysisSPSS 24.0 was used for all data processing and statistical analyses. The K-S test was used to verify whether the data conformed to a normal distribution. The number of bean geese population was a fixed factor. Then, univariate ANOVA and multivariate ANOVA in the general linear model were used to analyze the percentage of individuals exhibiting each behavior (foraging, vigilance, and flight) and the relationships with the noise intensity and disturbance distance, as well as their interaction with different population size. The SNK (Student-Newman-Keuls) test was used to compare the differences between the pairs of influencing factors and the behaviors of bean geese.3. Results3.1. Relationship between Foraging Behavior and Influencing FactorsGeneral linear model analysis of univariate ANOVA was used to evaluate foraging behavior with respect to noise intensity, interference distance, and interference duration. The model test results were available, and it was found that when the population size was N, there was a significant difference of noise intensity on foraging behavior (F = 6.302, p < 0.001, df = 5). When for N1, there also was a significant difference (F = 12.899, p < 0.001, df = 5). When for N2, there was no significant difference (F = 1.441, p = 0.222, df = 5). Meanwhile, there was no significant difference of interference distance (F = 0.024, p = 0.976, df = 2) and interference duration (F = 2.091, p = 0.102, df = 2) on foraging behavior.It was concluded that noise intensity had influence on the foraging behavior of the small population N1. The SNK test is shown in Figure 2. It could be seen that when the population size was N, the frequency of foraging behavior was its lowest at 100 dB (i.e., approximately 30.41%), at 70, 80, and 90 dB it was moderate (i.e., 49.93%) for only half of the population, and at ZR and 60 dB was the highest (i.e., 60.42%), which was twice of that for 100 dB. When the population size was N1, the frequency of foraging behavior was at its lowest at 100 dB (i.e., 30.53%), while at 80 and 90 dB it was moderate (i.e., 42.88%) for only half of the population, and at ZR, 60, and 70 dB it was highest (i.e., 60.42%), which was twice of that for 100 dB. However, N2 had no significant effect on foraging behavior. Therefore, with an increase in noise, the foraging ratio lower at N1 (smaller populations) was more obvious than that at N2 (larger populations).3.2. Relationship between Vigilance Behavior and Influencing FactorsGeneral linear model analysis of univariate ANOVA was used to evaluate vigilance behavior with respect to noise intensity, interference distance, and interference duration. The model test results were available, and it was found that when the population size was N, there was a significant difference of noise intensity on vigilance behavior (F = 10.431, p < 0.001, df = 5). When for N1, there was a significant difference (F = 18.516, p < 0.001, df = 5). When for N2, there was also a significant difference (F = 2.64, p = 0.031, df =5). The SNK test is shown in Figure 3. The model test results were available, and it was found that when the population size was N, there was a significant difference of interference distance on vigilance behavior (F = 5.378, p = 0.006, df = 2). When for N1, there was a significant difference (F = 8.48, p < 0.001, df = 2). When for N2, there was also a significant difference (F = 3.697, p = 0.03, df = 5). The SNK test is shown in Figure 4. Meanwhile, there was no significant difference of interference duration on vigilance behavior (F = 0.228, p = 0.776, df = 2).General linear model analysis of multivariate ANOVA was used to evaluate vigilance behavior with respect to noise intensity and interference distance. The model test results were available, and it was found that there was no significant difference of noise intensity and interference distance on vigilance behavior (F = 0.678, p = 0.744, df = 10).It can be seen that when the population size was N, the frequency of vigilance behavior was at its highest at 100 dB (i.e., 18.15%), while ZR and 60 dB was at its lowest (i.e., 5.96%), which was one third of that for 100 dB. When the population size was N1, the frequency of vigilance behavior was highest at 100 dB (i.e., 18.51%), while at ZR it was lowest (i.e., 5.92%), which was one third of that for 100 dB. When the population size was N2, the frequency of vigilance behavior was highest at 100 dB (i.e., 18.66%), while at ZR it was lowest (i.e., 7.04%), which was half of that for 100 dB. It could be seen that when the population size was N, the frequency of vigilance behavior was highest at close and medium distances (i.e., 12.67%), and at long distances was at its lowest (i.e., 5.96%), which was half of that for close distances. When the population size was N1, the frequency of vigilance behavior was highest at close and medium distances (i.e., 12.99%), and at long distances was at its lowest (i.e., 8.44%), which was half of that for close distances. When the population size was N2, the frequency of vigilance behavior was at its highest at close and medium distances (i.e., 14.11%), and at long distance was at its lowest (i.e., 6.23%), which was half of that for medium distances.3.3. Relationship between Flight Behavior and Influencing FactorsGeneral linear model analysis of univariate ANOVA was used to evaluate flight behavior with respect to noise intensity, interference distance, and interference duration. The model test results were available, and it was found that when the population size was N there was a significant difference of noise intensity on flight behavior (F = 24.389, p < 0.001, df = 5). When the population size was N1, it had a significant difference (F = 43.58, p < 0.001, df = 5). When the population size was N2, it also had a significant difference (F = 8.178, p < 0.001, df = 5). The SNK test is shown in Figure 5. The model test results were available, and it was found that when the population size was N, there was a significant difference of interference distances on flight behavior (F = 3.909, p = 0.022, df = 2). When the population was N1, it had significant difference (F = 9.78, p < 0.001, df = 2). When the population was N2, it had no significant difference (F = 0.362, p = 0.697, df = 5). The SNK test is shown in Figure 6. Meanwhile, there was no significant difference of interference duration on flight behavior (F = 3.427, p = 0.089, df = 2).It can be seen that when the population size was N, the frequency of flight behavior was highest at 100 dB (i.e., 33.76%), at 80 and 90dB were medium (i.e., 20.42%), and at ZR and 60 dB were lowest (i.e., 0.00%). When the population size was N1, the frequency of flight behavior was highest at 100 dB (i.e., 33.77%), at 80 and 90dB were medium (i.e., 20.42%), and at ZR was lowest (i.e., 0.00%). When the population size was N2, the frequency of flight behavior was highest at 100 dB (i.e., 39.54%), at 70, 80, and 90dB were medium (i.e., 24.32%), and at ZR was lowest (i.e., 0.00%). It also can be seen that when the population size was N, the frequency of flight behavior was highest at close distances (i.e., 15.00%), while at long distance was lowest (i.e., 8.28%), which was half of that for close distances. When the population was N1, the frequency of vigilance behavior was highest at close distances (i.e., 16.31%), while at long distance was lowest (i.e., 6.72%), which was half of that for close distances. However, when the population was at N2, there was no significant effect on flight behavior. Therefore, with an increase in noise, the flight ratio was higher at N1 (a smaller population) and was more obvious than that at N2 (larger populations).General linear model analysis of multivariate ANOVA was used to evaluate flight behavior with respect to noise intensity and interference distance. The model test results were available, When the population size was N, there was a significant difference of noise intensity and interference distance on flight behavior (F = 2.555, p = 0.007, df = 10). When the population was N1, there was a significant difference of noise intensity and interference distance on flight behavior (F = 5.923, p < 0.001, df = 10). When the population was N2, there was no significant difference of noise intensity and interference distance on flight behavior (F = 1.887, p = 0.063, df = 10). It can be seen that for noises >70 dB at short disturbance distances and >80 dB at medium disturbance distances, the flight behavior increased significantly, and with the increase of noise flight frequency gradually increased (Figure 7). However, there was no significant change in flight behavior at long distance.4. DiscussionNoise can affect the availability of sound information to birds and alter their foraging and vigilance behaviors [10]. Our research found that, in the natural environment, one or two bean geese in each population were responsible for guarding, while the others foraged or rested on the grassland. This is consistent with other studies, where, under no noise, the most frequent behavior was foraging (63.8%), followed by resting (15.6%), and other behaviors (such as alarm and flighting, standing, and walking) [30]. When the simulated noise was 60 dB, the distribution of behaviors was similar to that of bean geese without noise. Up to 70 dB, we observed that vigilance is always observed in bean goose populations, and as the number of individuals increases, foraging efficiency decreases [31], suggesting that noise may be sufficient to cause a stress response in birds [4]. Further, at this noise level, very few individuals showed retreat or flight behaviors, indicating that bean geese have the ability to adapt to or tolerate mild noise stimulation. When the intensity is 75 dB, birds showed an increased vigilance and reduced foraging behavior [26,32]. Moreover, as the intensity increased to 80 dB, the number of individuals showing foraging behavior decreased gradually, while the frequency of vigilance behaviors continued to increase. Some individuals moved away from the observed population, resulting in retreat or flight. A former study has shown that the thrush shows greater retreat behavior at 75.4 dB than 80 dB [11], and this may be due to the fact that the birds were raised in captivity, with a different living environment from the natural environment. However, at our field site, the noise was attenuated, to a certain extent, at various distances; accordingly, the noise intensity resulting in flight was higher than the actual noise intensity.Noise can keep birds away from certain sites. The initial proximity to humans can affect birds, resulting in flight [10]. Vehicle noise within 100 m of a road has a great impact on bird behavior, of which flight (retreat) accounts for 11% of behaviors [33]. We found that ship noises at close distances (<100 m) have a greater effect on the behavior of bean geese than noises at longer distances, resulting in reduced foraging activities and increased time spent on guard. The noise source was continuous and fixed, which may result in a longer-term effect on birds than those of moving sources, such as vehicles. Studies have shown that black-necked cranes maintain distances of 135.18 m from highways [12] and mallards are seriously affected within 150 m from roads [14]. At medium distances (100–200 m), bean geese continued to feed on grass; however, vigilance behavior increased. A portion of individuals exhibited obvious preparation for flight, while others maintained foraging behavior or remained vigilant, revealing that there is intraspecific variation in noise tolerance [34]. Unlike individuals with high tolerance to noise, those with low tolerance will identify weak stimuli as risks and show behavioral responses [35]. In addition, there are differences between waterbird species in their tolerance to noise [32]. In experiments, noise not only directly affects target subjects, but also unintentionally affects other organisms in the ecosystem [36,37]. For example, playing noise can cause agitation among livestock in nearby residential areas, and the results of the experiment can be affected by the indirect effects of interactions with other species, particularly in mixed populations of great white-fronted geese (Anser albifrons), tundra swans (Cygnus columbianus), and hooded cranes. Christoph et al. [21] observed this phenomenon, in which swan geese were frightened by barking and flew to other habitats, resulting in a significant difference in the distribution between conditions with barking and no barking (p < 0.001). However, with an increase in the distance from the noise source, the number of bean goose individuals exhibiting foraging behavior within 200 m remained basically unchanged, while the individuals exhibiting vigilance behavior showed a slight decrease and no or very few individuals exhibited flight. Bean geese and greylag geese (Anser anser) [14] located 360 m from the highway are also only slightly disturbed by noise.The flight distance of birds increases as noise interference increases [38]. In the Haizhu Lake area, Vervet spoonbills (Platalea minor) produced flight at a noise value of 61.6 ± 5.2 dB when disturbed by humans [39]. When disturbed by simulated noise in farmland habitats, the bean geese exhibit flight reached 70 dB. It is possible that Vervet spoonbills in the Haizhu Lake area are more prone to flight in response to various disturbances, such as noise and human activities. Research also suggests that hooded cranes show vigilance behaviors in response to noise and move away from the noise sources [15], as well as when ambient noise increases by approximately 60 dB, black-necked cranes gradually escape to the side to a distance approximately 60–80 m away from the noise source [13]. They are more sensitive to human interference and less vigilant to traffic noise, such as motor vehicles or ships [13]. Bean geese live in a natural environment, avoiding interference from human activities, and are sensitive to noise. In a study of the impact of highway noise, for noises >60 dB at short distances (<100 m) and noises <60 dB at long distances (>300 m) on bird diversity indicated that the richness and encounter rate of birds in close range decreased significantly and the species composition was significantly different [40]. Our results also indicated that 70 dB at a short distance and 80 dB at a medium distance result in the departure of some individuals from the observed group of bean geese, and they began to move away from the noise source until they retreated to a safe distance to forage and rest, while at a long distance, the population and behavior of bean geese did not change significantly. The noise value measured in this study was relatively high when the geese exhibited flight. The difference among studies may be related to the indirect influence of vehicle traffic and human disturbances [21], the group size of social foragers [6], or a difference in sensitivity to noise among species [11].5. ConclusionsWith an increase in noise, the frequency of foraging behavior of bean geese wintering at Shengjin Lake decreased, while the ratio of vigilance to flight behaviors increased. Additionally, sensitivity to noise was higher in small populations (N ≤ 30) than in large populations. For 70 dB at a short distance and 80 dB at a medium distance, the frequency of flight increased significantly; accordingly, these values can be considered thresholds for the influence of simulated ship noise on the behavior of bean geese. These results suggested that ships should maintain a distance of more than 200 m from waterbirds to reduce noise disturbance. Since different waterbirds have different sensitivities to noise, behavioral research and analyses of more waterbirds should be conducted, in order to establish a corresponding database and provide a scientific basis for the protection of waterbirds. Our study has certain limitations. First, our experiment only considers the noise intensity, not the frequency. Second, the noise duration in our simulations was not long enough to detect more behavioral changes. These limitations should be clarified in future studies.

animals : an open access journal from mdpi

10.3390/ani11071955

PMC8300271

Alpacas and llamas are domesticated species of New World camels. If the mare dies or produces insufficient colostrum or milk, information about the composition of colostrum and milk is needed to formulate suitable substitutes to adequately supply the crias. Milk composition in alpacas has been sparsely studied. In this study colostrum samples were taken daily during the first four days after parturition and milk samples were obtained monthly during the first four months of lactation. The samples were analyzed for their composition. The fat and lactose content are lowest on the day of birth and then increase, the protein content decreases during the first four days. Over the next four months, these contents do not change significantly. The results can be used for the development of colostrum and milk replacers.

Although alpacas are not used for milk production a detailed knowledge on the composition of the colostrum and milk is needed for development of colostrum and milk replacers. The aim of the present study was to measure the concentration of fat, protein, lactose, and minerals in alpaca colostrum and milk. Colostrum samples were taken daily over four days after parturition from 20 multiparous alpaca mares. Milk samples were obtained monthly, during the first four months of lactation from 17 alpacas. Composition of colostrum and milk differed in numerous indicators. The concentrations of fat and lactose increased from day 1 (0.5%, 4.0%) to day 4 (5.3%, 5.0%), protein decreased from 20.4% on day 1 to 8.3% on day 4. In milk these three indicators did not change during the lactation. Minerals have been little studied in alpaca colostrum and milk in the past, many of which had the highest concentrations in colostrum immediately after birth. The results of the present study do not support that goat’s milk is the preferred substitute for feeding crias. This study contributes to the knowledge of the composition of alpaca colostrum and milk which can be of particular use in developing replacers.

1. IntroductionAlpacas, along with llamas, are one of the domesticated species of New World camelids [1,2]. Although alpacas and llamas are not used for milk production knowledge on the composition of the colostrum and milk is needed for development of colostrum and milk replacers. Such replacers are required for crias which have to be hand reared if their mothers have died or are not producing sufficient quantity of milk. Currently milk from other species, mostly cattle or goat or milk replacers designed for calves, lambs, and kids are used [3].In contrast to numerous studies in several dairy species only a small number of studies have been reported on the composition of milk from South American camelids (SAC) [4,5,6,7,8,9]. The studies on colostrum composition in alpacas have mainly reported on immunoglobulin concentration [10,11,12]. Additionally, studies on alpaca which is a dominant SAC breed in many countries are mostly based on a small number of animals [13,14,15].One study measured colostrum constituents taking samples 48 h after parturition, comparing alpacas kept in different altitudes (18 animals at sea level and 24 animals at 4400 m above sea level). Later in the same study milk samples were obtained monthly from the first into the fifth month of lactation. However, to obtain sufficient volume for analyses (60 mL) the researchers had to pool the samples of three animals, reducing the effective number of samples by a third. Colostrum fat, protein, and lactose concentrations did not differ between the habitats, but milk constituents were influenced by lactation status [16]. The content of fat, protein, casein, and minerals (calcium, phosphorus, magnesium, potassium, sodium, and zinc) were measured in the milk of eight alpaca mares taken on the 30th and 60th lactation day in Italy [17]. Chad [18] studied the composition of alpaca milk (calcium, phosphorus, potassium, magnesium, sodium, sulfur, lactose, fat, protein, and urea) using milk samples of 11 alpaca mares taken on two Californian farms. The samples were taken every week during the first 25 weeks of lactation. To our best knowledge studies on trace elements in colostrum or milk from SAC have not been published so far.The aims of the present study were to measure the concentration of fat, protein, lactose, macro and trace minerals in alpaca colostrum and milk at different stages of lactation. Further the composition of colostrum and milk should be compared to information given in the literature for SAC and other species.It was hypothesized that:The composition of colostrum from alpacas changes within the four day colostral period after parturition.The composition of milk changes according to the lactation status in alpaca mares.The composition of alpaca colostrum and milk differs in certain indicators from ruminant milk which supports the development of species specific colostrum and milk replacers.2. Materials and Methods2.1. Animals and Sampling ProceduresThe project was discussed and approved by the institutional ethics and animal welfare committee in accordance with GSP guidelines and national legislation (Ethic Code ETK-21/11/2016).2.1.1. Alpacas for Colostrum SamplingThe colostrum samples were taken from 20 alpaca mares kept in three smaller alpaca farms in the district Bruck-Mürzzuschlag in Styria (Austria). The farms were in close proximity having almost identical weather, husbandry, and feeding conditions. All available pregnant multiparous alpaca mares of the foaling season (June–September) were included in the study. The animals belonged to the Huacaya breed and were between four and nine years old. They were pastured in the Austrian Alps at an altitude between 700 and 1100 m above sea level on a calcareous source rock. The animals were additionally offered hay ad libitum. Colostrum samples were taken on four consecutive days after foaling, the first samples (day 1) were taken on the day of birth between two and four hours after parturition. Crias were not separated from their mothers, but the teats were closed with a tape at least two hours before sampling. After that, the entire amount of colostrum was milked by hand.2.1.2. Alpacas for Milk SamplingThe milk samples were obtained from lactating alpacas four times at monthly intervals from all 17 alpaca mares on a farm in the Alps in South Tyrol (Northern Italy). Mothers and crias were not separated the whole time, but only for two hours before sampling. These animals (15 Huacaya and two Suri) were comprised of ages between three and 10 years. Thirteen of the mares were multiparous and four had given birth to their first cria. The alpacas grazed on alpine meadows between 1320 and 1550 m above sea level on a calcareous source rock.2.1.3. Sample PreparationFor sampling the mares were restrained by their owners using a halter and were milked by hand. The udders were examined for conspicuous redness, swelling, and induration to exclude animals with clinical mastitis. Colostrum or milk from all four udder quarters was obtained in similar volume. Samples obtained were examined for visible milk changes that would indicate mastitis. After milking the sample container was immediately transferred to a cooled polystyrene box (4 °C) (Henry Schein Medical Austria GmbH, Vienna, Austria). Within one hour the samples were divided into sub samples in Eppendorf® Safe-Lock microcentrifuge tubes (volume 2.0 mL, Eppendorf Austria GmbH, Vienna, Austria) and in one Greiner tube (15.0 mL volume, Greiner Bio-One GmbH, Kremsmünster, Austria) and stored at −18 °C and shipped to the laboratories.2.2. Laboratory Analyses2.2.1. Analyses of Milk Fat, Protein, and Lactose ConcentrationAnalyses of milk and colostrum fat, lactose, and protein concentrations were performed at the laboratory of the Institute of Food Hygiene, Veterinary Faculty at Leipzig University applying standardized laboratory methods as described by the German Industry Standard (Deutsche Industry Norm, DIN). The fat content was determined using the Weibull–Berntrop gravimetric method [19]. Due to the limited volume of colostrum or milk the analysis of fat concentration could only be performed as a single measurement. The content of lactose was analyzed by the UV lactose/D-galactose method (Roche Diagnostics®, Mannheim, Germany) and the protein content by the Kjehldahl method [20].2.2.2. Analyses of Macro and Trace MineralsThe milk and colostrum samples were analyzed at the Vet Med Labor GmbH, IDEXX Ludwigsburg (Germany). The elements sulfur (S), phosphorus (P), sodium (Na), and potassium (K) were analyzed by inductive coupled plasma optical emission spectrometry (ICP-OES) using Vista-Pro device (Varian Inc., Palo Alto, CA, USA) and the elements lithium (Li), boron (B), magnesium (Mg), aluminum (Al), calcium (Ca), manganese (Mn), iron (Fe), cobalt (Co), nickel (Ni), copper (Cu), zinc (Zn), arsenic (As), selenium (Se), strontium (Sr), molybdenum (Mo), cadmium (Cd), tin (Sn), barium (Ba), thallium (Tl), lead (Pb), and uranium (U) by the inductive coupled plasma mass spectrometry (ICP-MS) device Aurora M90 (Bruker Daltonics, Bremen, Germany). The coefficients of variation of the analyses are shown in Table 1 for each measured parameter.2.3. Statistical AnalysesSince not all data were normally distributed (Kolmogorov–Smirnov Test), all the indicators are presented as median and first/third quartile. Composition of milk and colostrum was compared using the nonparametric Mann–Whitey U Test. Further analyses were performed using log transformed data. The fat, protein, lactose, and mineral element concentrations were compared over time separately for colostrum and milk using a mixed linear model (measurement repetition). The individual alpaca mare was considered as a random effect, while the age of the animal, the day in milk and the lactation number were fixed effects. The Bonferroni test was applied as a post-hoc test. Microsoft Excel 2010 and the SPSS Statistics Version 24 (IBM Corp., Armonk, NY, USA) were used for statistical analysis. The significance level was set at 5%.3. Results3.1. Colostrum and Milk VolumeA total of 77 colostrum samples were obtained from the 20 mares during four days postpartum. The three missing samples were due to one animal dying on day 3 and one animal developed a clinical mastitis on day 4. The volumes which could be obtained varied between 12 and 28 mL with a median of 20.8 mL. There was no difference between the days in the obtainable colostrum volume.Overall, 61 milk samples were obtained (17 in the first month, 16 in the second month, 16 in the third month, 12 in the fourth month). The hand milking of the alpaca mares became generally more difficult with increasing duration of lactation and in some animals in advanced lactation only very small volumes or no milk at all could be obtained. One animal had to be excluded since it developed a clinical mastitis in lactation month 2, while in four animals no milk could be obtained in month 4. The volumes which could be obtained varied between 0 and 25 mL with a significant difference between the milk volumes of the months 1 and 2 (median 14.0 mL and 12.5 mL) and the volumes of month 3 and 4 (median 8.0 mL and 5.5 mL).3.2. Fat, Protein, and Lactose Concentrations in Colostrum and MilkThe colostrum concentrations of fat, protein, and lactose are shown in Figure 1 and Table 2. The concentration of fat in colostrum increased significantly from day 1 to day 4; at day 4 the fat concentration was already similar to the fat concentration in milk during later lactation (Figure 2, Table 3). The lactose concentration on the day of parturition (day 1) was lower in comparison to days 2, 3, and 4. In contrast the colostrum protein concentration decreased substantially over the colostral period; however, the protein concentration on day 4 was significantly higher in comparison to the concentrations in milk later during lactation (Figure 2, Table 3).The concentrations of milk fat, protein and lactose during four months into lactation are shown in Figure 2 and Table 3. None of the three indicators differed over the sampling period.3.3. Mineral Concentrations in Colostrum and MilkThe concentrations of macro and trace elements are presented in Table 4 (colostrum) and Table 5 (milk). The concentration of Ca, P, and Mg decreased during the colostral period, having the highest concentration at the day of parturition (day 1). The same condition was present in a number of trace elements (Fe, Cu. Zn, Sr, Ba, Co, Ni, S) showing the highest concentrations immediately after parturition (Table 4).During later lactation, milk calcium was the only macro element for which concentrations decreased during the four months of lactation. A comparison of mineral concentration between colostrum and milk is shown in Table 6. The concentration of numerous macro and trace elements differ significantly. Additionally, for comparison upper concentration limits for drinking water (Austria) and concentration for llama and alpaca milk reported in the literature are provided.4. DiscussionAs hypothesized, the composition of fat, protein, and lactose concentration in colostrum changed substantially during the colostral period. The decreasing concentration of protein and the increase of fat concentration mirrored that of other animal species [21,22,23]. The substantial decrease in colostral protein concentration can be attributed to the decrease in immunoglobulin concentration as described by Mößler [24] in alpacas and occurs in most other mammals [25,26,27,28,29,30]. For comparison Parraguez [16] took colostrum samples 48 h after parturition from alpaca mares in two regions of Chile. Six pooled alpaca colostrum samples taken in a herd kept at sea level in Patagonia had concentrations of 2.71 ± 0.6% fat, of 9.24 ± 0.5% protein, and of 5.33 ± 0.1% lactose whereas in eight pooled samples from the High Altiplano region (4400 m above sea level) concentrations of 4.80 ± 1.2% fat, of 9.84 ± 0.6% protein, and of 4.41 ± 0.1% lactose were measured. These measurements at 48 h after parturition are comparable and similar to the measurements at day 2 and 3 of the present study. The concentration of fat and lactose in colostrum at day 1 of the present study was also similar to the concentration measured in llama colostrum (fat 0.75 ± 0.25%, lactose 4.12 ± 0.46%) taken between four and 12 h after parturition [8]. However, the protein concentration in alpaca colostrum in the present study appeared to be higher in comparison to llamas at the same time (16.79 ± 1.64%) [8]; since the sample sizes are rather small in both studies (20 alpacas and nine llamas) it is impossible to draw conclusions on species differences. The composition of colostrum milked at day 4 after parturition is similar considering numerous indicators to milk later during lactation.Chad [18] measured on average 3.68 ± 1.32% fat, 4.53 ± 0.78% protein and 6.00 ± 0.48% lactose in milk of 11 alpacas over 25 weeks into lactation. The minor differences to the present study might be due to the fact that they also included animals in the first week of lactation meaning they used some colostrum samples and they applied a different method (infrared spectrometry) for measurement which has been validated for cow milk. In a study covering five months of lactation performed in two regions of Chile Parraguez [16] found in six pooled alpaca milk samples in Patagonia (sea level) concentrations of 2.6 ± 0.5% fat, of 6.5 ± 0.3% protein and of 5.2 ± 0.5% lactose and in eight pooled samples from High Altiplano (4400 m above sea level) concentrations of 3.8 ± 0.6% fat, of 6.9 ± 0.3% protein and of 4.4 ± 0.5% lactose. The milk composition in Chile was also similar to the present study.In contrast in comparison to llama milk in which Morin [7] in the USA found a fat concentration of 2.7 ± 1.0%; the fat concentration in alpaca milk in the present study was substantially higher. Additionally, the reported lactose (6.5 ± 0.5%) and protein (3.4 ± 0.4%) concentrations in lamas in the study by Morin [7] also differed substantially from the indicators in the present study in alpacas. Since the used laboratory methods were similar possible reasons for the differences are not obvious, however the diet might have had a substantial influence. The results for llama milk composition of a German study [8] (fat 4.70 ± 0.81%, protein 4.23 ± 0.23%, lactose 5.93 ± 0.27%) and an Argentinian study [9] (fat 4.55 ± 0.66%, protein 4.33 ± 0.17%, lactose 6.34 ± 0.34%) were much closer to the results of the present study in alpacas.Comparing fat, lactose, and protein concentration of alpaca milk to goat milk (fat 3.9%, protein 3.3%, lactose 4.2%) and cow milk (fat 3.8%, protein 3.4%, lactose 4.7%) it seems that there is no firm rationale for the widely accepted opinion in the breeder community that goat’s milk is the preferred option for feeding alpaca crias [31]. Our clinical experience also does not support this.The mineral content of alpaca colostrum and milk found in the present study is difficult to compare to other studies in SAC. It seems that there are no studies available on the mineral content of SAC colostrum. Further, it was observed that minerals are obviously stored in the udder before parturition as the concentration of a number of elements was highest in the first colostrum and decreased within the colostral period (Table 4). Only sparse information is available especially for trace elements in milk (Table 5). In a study by Morin [7] in 1995 the concentrations of a number of trace elements (B, Co, Mo, Sn, As, Cr, Cd, Hg, Pb, Se, Tl) were below the detection limits of the methods used at that time. Over 25 years later these technical limitations no longer exist; however, comparisons between different measurement methods are always difficult to draw.Additionally it has to be considered that the trace element concentrations found in colostrum and milk are influenced by the availability of those minerals in feed. The study found some substantial differences in a number of elements between colostrum and milk. However, since the alpacas herds in which colostrum and milk samples were obtained are kept in different areas of the Alps with similar but not identical conditions it remains unclear which part of the differences is caused by the lactation status or by the different feed supply.A limitation of the present study was that the trace and macro minerals were studied only in two regions in the Alps which were geologically similar. Since the source rock has an influence on the mineral content of milk, the results may therefore differ in other regions with different geological conditions. Further studies in geologically different areas on larger sample sizes would provide more reliable data.Summarizing the findings, the present study contributed to the establishment of reference ranges for these indicators, albeit there were some differences between fat, protein, lactose, and macro element concentrations between the present study and information from different literature sources. Taken together, they provide guidance for replacement colostrum and milk from other sources including milk replacers for motherless reared crias.5. ConclusionsThe components fat, protein, and lactose change significantly in the first four days of lactation and remain at a constant level during the further months of lactation. Concentrations of numerous macro and trace elements also differ significantly over lactation especially during the first four days after parturition. The composition of colostrum and milk is substantially different from cow or goat milk. It appears that there is no rationale for the widely accepted opinion that goat milk is the preferred option for feeding motherless alpaca crias.

animals : an open access journal from mdpi

10.3390/ani11041118

PMC8069800

Long-non-coding RNAs (lncRNAs) are transcripts of more than 200 nucleotides, which lack protein-coding potential. LncRNAs have been well characterized in many organisms, but their functions in insects have not been well deciphered. Advancement of high-throughput technologies has enabled the sequencing of genomes and transcriptomes of several insects, which has led to the identification of many important lncRNAs in insects. Characterization of lncRNAs and their regulatory roles in insects may provide insights of novel pest control strategies. Through this comprehensive review, we present an overview of insect lncRNAs, their identification as well as their function in insects of different orders. Toward the end of the review, we highlight the role of lncRNAs in insect developmental processes and discuss the future prospects of lncRNAs in insects.

Only a small subset of all the transcribed RNAs are used as a template for protein translation, whereas RNA molecules that are not translated play a very important role as regulatory non-coding RNAs (ncRNAs). Besides traditionally known RNAs (ribosomal and transfer RNAs), ncRNAs also include small non-coding RNAs (sncRNAs) and long non-coding RNAs (lncRNAs). The lncRNAs, which were initially thought to be junk, have gained a great deal attention because of their regulatory roles in diverse biological processes in animals and plants. Insects are the most abundant and diverse group of animals on this planet. Recent studies have demonstrated the role of lncRNAs in almost all aspects of insect development, reproduction, and genetic plasticity. In this review, we describe the function and molecular mechanisms of the mode of action of different insect lncRNAs discovered up to date.

1. IntroductionEukaryotic genomes are known to produce a huge array of RNA molecules differing in their size, abundance, and protein coding ability [1]. Only 2% of RNA is translated to make proteins; other RNA sequences that do not code for any protein are called non-coding RNA (ncRNAs) [2]. While the proteins are considered as the major trans-acting regulators, they are limited by the organizational complexity in forming a regulatory network. This is partially explained by the alternative splicing in the pre-mRNA of protein-coding genes and by the post-translational modification of proteins [3]. A paradigm shift has occurred in recent years with the identification and functional characterization of an increasing number of ncRNAs. The knowledge of ncRNAs has gained a huge momentum with the discovery of RNA interference (RNAi) and the role of ncRNAs in gene silencing phenomena [4]. This has led to the identification of various types of small ncRNAs, which guide the protein complex to the target mRNA to influence its translation. The ncRNAs include ribosomal RNA (rRNA), transfer RNA (tRNA), small ncRNAs (sncRNAs), and long ncRNAs (lncRNAs) (Figure 1). The discovery of rRNA and tRNA dates back to the 1950s, whereas sncRNAs and lncRNAs are relatively recent discoveries [5]. As per our current understanding, molecular size is the distinguishing feature between sncRNAs and lncRNAs [6]. SncRNAs, which have a size <200 nucleotides include microRNA (miRNA), small interfering RNA (siRNA), small nuclear RNA (snRNA), and PIWI-interacting RNAs (piRNAs), whereas the ncRNAs of size >200 nucleotides belong to the category of lncRNAs and are present in almost all the eukaryotic organisms [7].2. Long Non-Coding RNAs (lncRNAs)LncRNAs can be categorized based on the region of the genome from which they are transcribed: (1) Sense lncRNAs overlap the exonic regions of another transcript produced from the same strand; (2) Antisense lncRNAs are located on the complementary strand of the sense strand; (3) Intergenic lncRNAs (lincRNAs) are those produced from the DNA between two genes (Intergenic regions); and (4) bidirectional lncRNAs are transcribed simultaneously to coding transcripts at the opposite strand [9] (Figure 2).LncRNAs display many common characteristics of mRNAs [10]. The majority of lncRNAs are transcribed by RNA polymerase II and are post-transcriptionally modified [11,12]. They undergo 5′ capping, 3′ polyadenylation, and splicing, but lack poly-A tails [13]. Compared to mRNAs, lncRNAs contain fewer exons, expressed in low abundance and are generally tissue-specific [14]. There are certain databases that are specifically designed for lncRNAs such as NONCODE [15], lncRNAdb [16], lncRNBase [17], DeepBase [18], etc. Even though lncRNAs have been discovered from many insect species, most of our knowledge about the functional aspect of insect lncRNAs comes from studies in Drosophila melanogaster. Besides Drosophila, lncRNAs have also been reported in Anopheles gambiae [19], Apis mellifera [20], Bombyx mori [21], and Tribolium castaneum [22,23]. With the advent of next-generation sequencing (NGS) technologies, large numbers of lncRNAs have been identified, but most of them remain functionallyunvalidated [24]. A total of 11,810 lncRNAs (6250 lincRNAs) were identified in the lepidopteran, B. mori [21], which was almost double the lncRNAs identified in many other insect species (3844 lincRNAs from Plutella xylostella [25], 2059 lincRNA from Anopheles gambiae [19], 3482 lincRNAs from Aedes aegypti [26], and 1514 lincRNAs from Apis mellifera [20]). This difference could perhaps be attributed to their genome size as well as various annotation strategies, which need bona fide checks. As the lncRNAs are known to regulate several biological processes, viz cell cycle progression, cellular differentiation, development, disease mechanism, metabolism and immune response, understanding the mechanism of action of lncRNAs is very important to regulate the gene expression and epigenetic silencing through heterochromatin formation, histone modulation, and DNA methylation, which are associated with it [27,28,29,30]. Cumulatively, lncRNAs are an important player in regulating gene expressions in several physiological, pathological, and immunological processes [31]. Insect lncRNAs display tight temporal and spatial expression patterns and play important roles in several developmental regulations like sex-determination, immunity, and morphogenesis [32]. They are also known to be involved in determining certain insect behaviors like sleeping, foraging, courtship, etc [33]. A recent study by Stork et al. suggests the existence of 5.5 million insect species, where only about one million species have been characterized to date [34,35]. Furthermore, the least attention has been given to lncRNA studies considering the vast diversity of insects. In this review, we have attempted to summarize lncRNAs and their functions in different insect species studied to date [36].3. Identification and Functional Characterization of lncRNAsThere are at least tens of thousands of lncRNA present in the human genome as well as in the genomes of non-human primates [37]. Similarly, thousands of lncRNAs are also found in other vertebrates, invertebrates, insects, and plants. Furthermore, this number is continuously increasing because of the advancement in sequencing technologies [38]. Since the list is so extensive, no clear orthologous lncRNAs between these groups have been reported. Compared to mRNA, the expression of lncRNA is more tissue specific and its evolution is much faster than mRNA [39]. This is evident by the observation that there is great similarity between the protein-coding genes and miRNA genes of human and mouse, whereas most human lncRNAs do not have any homologs in mice and vice-versa [40]. Interestingly, LncRNAs that are discovered through greater sequencing efforts are less conserved than those previously discovered through shallow sequencing methodologies. Higher sequence conservation exists between exons of lncRNAs compared to that between introns of protein-coding genes [41]. Taking all of the present information about lncRNAs into account, it can be concluded that identifying conserved regions of lncRNAs is a challenging task that would require better prediction tools [42]. For example, it is evident that the evolutionary conservation of long intergenic noncoding RNAs (lincRNAs) might result in the presence of coding chunks and may be biased; as such conserved sequences are significantly reported across all eutherians/mammals [43]. As a result, the functional characterization of lncRNAs and the mechanisms underlying these transcripts lack good precision. Given this paradigm, there is a need for the dissemination of bona fide prediction tools as we hardly know whether insect lncRNAs, if not pest lncRNAs, influence gene expression. Nevertheless, there are reports wherein the transcriptomic changes in plants such as rice take place, but their impact is on the DEGs interacting with lncRNAs alone, suggesting their subtle response to farm chemicals [44]. Altogether, this is also augmented by the fact that insect lncRNAs are mis-annotated by the lack of such annotation tools.Since, lncRNAs are relatively less conserved across species, investigating their biological importance as well as their mechanism of action is a difficult task. The discovery of lncRNAs and their functions to date suggest their roles in different biological processes via gene regulation by serving as molecular signals, guides, decoys, and/or scaffolds (Figure 3) [45]. LncRNAs can bind with DNA, RNA, and proteins to regulate gene expressions both at the transcriptional as well as post-transcriptional level. There are different computational approaches for the identification of lncRNAs from the genome as well as transcriptome data, investigation of lncRNA functions, and its associated regulatory networks. Although identifying lncRNAs is out of the question from whole exome sequencing, recent reports including ours have shown how the third generation sequencing technology is used to screen the lncRNAs from exomes [46,47]. Nevertheless, the lncRNAs are ascertained as differentially expressed gene (DEG) counts and many of them turn out to be up/downregulated. These could be from 5′-UTR or intergenic or regulating from the intron-exon boundaries. They are validated using downstream approaches like qRT-PCR or further validation checks. A detailed characterization and potential identification of long read/short-read sequencing were beyond the scope of this review even as potential lncRNAs could be screened from either of these technologies [48].RNA sequencing is widely used to identify lncRNAs where total RNA is used to generate raw reads, which are then assembled into transcripts either with or without a reference genome and finally, the resulting transcripts are annotated [49]. There are many pipelines available for the identification of lncRNA. Thousands of lncRNAs have recently been identified in Drosophila melanogaster using such approaches, which provides a platform for the structural and functional exploration of lncRNA in other invertebrates. Identification of lncRNA started from the Functional ANnoTation Of the Mammalian genome (FANTOM) project, which established a comprehensive platform for the comparative analysis of lncRNAs in mammalian transcriptomes [50]. This was followed by some experimental methods such as tiling arrays and chromatin immunoprecipitation studies. To predict lncRNA–dsDNA interactions, tools like triplextor and Longtarget have been used. LncRNA can be integrated with Hoogsteen hydrogen binding into the main groove of the DNA duplex where Hoogsteen bonding is weaker than the Watson–Crick bonds. Different experimental approaches have been utilized to study lncRNA and DNA interactions. The formation of lncRNA DHFR and dsDNA triplex was demonstrated using electrophoresis mobility shift, and the binding of lncRNA Fendrr to dsDNA was determined by in vitro pull-down experiments [51,52]. Although the mechanism of RNA-dsDNA triplex formation is still not well illustrated, it is clear that the lncRNA–DNA interaction offers a potent mechanism for gene regulation. Similarly, there are different bioinformatic tools to predict RNA–RNA interactions such as RNAplex [53], RNAup [54], and intaRNA [55]. To investigate lncRNA–protein interactions, RPI-Pred [56], lncPRo [57], and RPIseq [58] can be used. There are some tools that can also predict the binding sites of RNAs or proteins such as BindN [59], RNAProB [60], PPRint [61], etc. When the proteins introduced by lncRNAs are methylation-related enzymes, they can induce promoter CpG island methylation or demethylation. If histone-modifying enzymes are imported by lncRNAs, histone changes can lead to gene expression, transcriptional silencing, or repair of DNA and genomic labeling. The neighboring (cis) or distal protein-coding genes (trans) can be regulated by lncRNA [62]. The LncRNAs from one chromatin can bind to another chromatin, like LncRNA HOTAIR, transcribed on chromatin 12 from the HoxC locus and suppresses transcription in the HoxD locus [63]. All these data suggest the complicated roles of lncRNA in gene regulation. In order to gain a deeper insight into the interactions between lncRNA and different biological molecules, extensive experimental validation along with bioinformatics analysis is required. Our understanding of gene regulation by lncRNA is in its infancy, but seeing the large number of lncRNAs studies, we can expect better outcomes in this field in the near future. LncRNAs have been well studied in mammals and are known to be involved in many important biological processes, but it is poorly characterized in non-model organisms such as insects. Excitingly, research focused on insect lncRNAs has increased dramatically in recent years, which suggests their role in insect development, anti-viral defense, and insecticide resistance mechanism. The existing lncRNA studies are limited only to some of the model insects; hence there is a need for the expansion of lncRNA studies in a more diverse range of insects. The discovery and function of lncRNAs in different insects are described below:3.1. LncRNAs in Drosophila MelanogasterAs in the case of other insect specific pathways like sex-determination, metamorphosis, etc., insect model Drosophila melanogaster was the first to be attempted for the extensive studies of lncRNA [64,65]. More than 100 lncRNAs are estimated to be coded by the Drosophila genome, most of which are expressed during embryonic development and are spatially restricted to the developing central and peripheral nervous system [66,67]. Knockdown studies in Drosophila have also confirmed the role of lncRNA in spermatogenesis and male fertility [68]. The function of lncRNAs in Drosophila covers development, behavior, stress resistance, gender identification, and dosage compensation [69]. Several testis-specific LncRNAs are required for nuclear condensation and sperm individualization during gonadal development in Drosophila [70]. Remodeling of spermatids through chromatin condensation (by replacement of histone proteins with protamines) is essential for the removal of excess cytoplasm required for sperm individualization. Testis of lncRNA (CR44455/6, CR45542, and CR44420) mutant Drosophila was found to have spermatids with scattered, round, and un-condensed nuclei [71]. Crumpled nucleus phenotype (similar to that in case of protamine mutants) was also displayed by mutants of lncRNA TS1 and CR43484. Furthermore, RNA sequencing confirmed the role of CR43484 lncRNA in regulating the expression of several testis-specific genes. This included both protein-coding genes and lncRNA, suggesting the involvement of testis-specific lncRNAs in late spermatogenesis through transcriptional regulation (Wen et al., 2016).This is similar to the case of other functionally characterized lncRNAs like Paupar and Pantr [72,73].A conserved yellow-achaete intergenic lncRNA (yar) regulates sleep behavior in Drosophila [74]. yar lncRNA is cytoplasmic and expressed during the mid-embryogenesis [75]. yar resides within the neural gene cluster, upstream to the yar locus is yellow gene (y-encodes a secreted protein required for cuticle coloration and male sexual behavior) and downstream is acheate gene (ac-encodes one of four related bHLH transcription factors of the achaete–scute complex (AS-C) required for proper development of the central and peripheral nervous system [76,77]. There are different lncRNAs that regulate miRNA activities by acting as a sponge to downregulate them. To further investigate the link between yar and Drosophila miRNAs, sequences encompassing the yar exons were submitted to a web-based tool, Microinspector. Using miRBase, which includes both predicted and confirmed miRNAs, 33 miRNA seed matches corresponding to 19 confirmed miRNAs, were identified within yar exons. Similar experiments with the yar intron resulted in the identification of 36 miRNA seed matches corresponding to 25 confirmed miRNAs. Further studies are needed to investigate the interaction between yar and miRNAs and its functional significance. Phenotypic analyses of Drosophila null mutants of yar suggested its requirement in sleep maintenance and homeostasis. These mutants had reduced night time sleep as a result of reduced sleep bout length. Additionally, there was no increase in the daytime sleep in yar null mutants, indicating the loss of sleep homeostasis in these flies and suggesting the involvement of yar in sleep regulation. Interestingly, one of the miRNA seed matches within the yar exon corresponds to miRNAs from the miR310 cluster (predicted and experimentally confirmed miRNAs). No match was found in similar analyses using intronic sequences of yar or exonic sequences corresponding to three other genes (y, ac, and GAPDH2). Loss of miRNAs 310 to 313 alters synaptic transmission at the larval neuromuscular junction, with no effect on viability or fertility. These findings are consistent with the possibility that yar might participate in a regulatory circuit that influences levels of miRs within the brain, which may have the capacity to contribute to synaptic homeostasis [74]. Comparative sequence analysis of the yar locus (with one of the conserved sequence motifs bound to the yar promoter) suggested its conservation in different Drosophila species, representing 40–60 million years of co-evolution. The similar timing of yar expression in D. virilis and D. melanogaster further suggests conservation in the transcriptional regulation of yar [78].Nuclear lncRNAs, roX1 and roX2, are essential for the process of dosage compensation in Drosophila [79]. Both of these lncRNAs (roX1 and roX2) form the male-specific lethal (MSL) ribonucleoprotein complexes. It was observed that synchronous removal of lncRNA-rox1 and roX2 decreases X-chromosome localization of the MSL complex. Interestingly, both these lncRNAs independently play an important role in the dosage compensation process where roX1 is the most abundant one and its loss results in the reduced expression of X-chromosome genes. Loss of roX2 leads to MSL independent upregulation of genes [80,81]. Similarly, Xist is a lncRNA involved in dosage compensation in humans where it regulates chromatin modification and expression of specific genes. Sexlethal (Sxl) is the master regulator of the sex-determination process in Drosophila [82]. The dose sensitive early promoter of Sxl (i.e., SxlPe) senses the number of X chromosomes (one X vs. two X) and gets transcribed only in females (i.e., with XX sex chromosome composition) [83]. Since sex determination and dosage compensation processes are linked and lncRNAs are involved in the dosage compensation process in Drosophila, researchers have also speculated on the role of lncRNAs in the activation of SxlPe [84]. Drosophila lncRNAs, heat-shock RNA-omega (hsr-omega/ω), and CR34335 are involved in the process of cellular aging [85]. hsr-ω, present in the form of nucleoplasmic omega speckles, is essential for accumulating heterogeneous nuclear RNA binding proteins (hnRNPs) [86]. hsr-ω functions as a hub for the coordination of transcriptional regulators and hnRNPs, which impact many cellular responses such as apoptosis [87]. The 93D, or hsr- ω (heat-shock RNA-omega), locus of Drosophila is known to be involved in transcriptional and translational activities. Initially, this gene was named 93D as it is present in the 93D cytogenic region of the polytene chromosome of D. melanogaster, but was later renamed as hsr-ω (heat shock RNA-omega). This gene was compared in three different Drosophila species: D. hydei, D. melanogaster, and D. pseudoobscura. Although the structure of this locus is highly conserved in all three Drosophila species, the primary base sequence had diverged rapidly between them. In all three species, hsr-ω consists of a unique region in the transcription unit at 5′ site, which consists of two exons and one intron [88]. Overall, three transcripts (two nuclear and one cytoplasmic) are produced from this locus, which do not have any significant protein-coding capacity. This locus is developmentally active in nearly all cell types and is essential for the viability of flies. Its induction during heat shock is independent of the other members of the heat shock gene family [89]. The other selective inducers act on this locus through separate response elements and hsr-ω activity has a characteristic effect on transcription/turnover of the heat shock induced hsp70 and the alpha-beta transcripts in Drosophila melanogaster [90]. Research has shown that lncRNA-hsr performs a crucial function in thermo tolerance to cope with heat stress [91]. Upon temperature shock, the nullisomy, RNAi, or overexpression of lncRNA-hsrω imply lethality in most embryos and first or third-instar larvae [92]. Three-day-old null fly lncRNA-hsr assemblies have poor projections after heat shock, whereas both down- and upregulation of lncRNA-hsr assemblies reappear during the recovery from heat shock [93,94].The Drosophila maternal effect gene oskar encodes the protein oskar and has distinct roles in germ line determination and posterior abdominal segment differentiation [95]. There are two isoforms of the Oskar protein, which are formed as a result of translation of two different in-frame start codons of oskar mRNA. Out of these two, the short oskar (139–606aa) is absolutely required for germ cell development whereas the long oskar (1–606aa) is essential for tight anchoring of the germplasm to the posterior oocyte cortex. Nonetheless, oskar RNA plays a major role during early Drosophila oogenesis via a translation-independent mode that acts as lncRNAs [96]. Most oskar RNA levels have been decreased by a sterile phenotype due to early oogenesis. Moreover, expression of the oskar 3′ UTR is sufficient to recover the eggless defect of the RNA null mutant independent of protein [96]. Previously, the localization of Staufen, an RNA-binding protein, within the oocyte is interdependent with that of oskar mRNA [97]. In the oskar null mutant, the Staufen protein fails to transport from the nurse cells into the oocyte. Expression of the oskar 3′ UTR alone is sufficient to restore Staufen accumulation in the oocyte. This reveals that the mutual interdependence of Staufen and oskar RNA in their localization during oogenesis is mediated by the interaction of Staufen with the oskar 3′ UTR [98]. Another possibility is that this non-coding function is mediated partly through sequestration of the translational regulator Bruno, which binds to Bruno response elements in its 3′ UTR [99]. Another important lncRNA in Drosophilamelanogaster is acal, which is involved in negative regulation of Jun-N-terminal kinase (JNK) signaling and maintaining cell stretching and closure [100]. Bereft and bxd, other important lncRNAs, are involved in the development of extra sensory organs [101] and are also involved in growth and development, respectively [102,103]. Iab-8 represses the homeotic gene abd-A and its knockdown leads to male and female sterility [104]. Another important lncRNA in D. melanogaster is CRG, which is involved in locomotor activity and climbing ability [105]. There is a controlled expression of lncRNA-CRG in the central nervous system (CNS) from the embryonic to the adult stages. This lncRNA shows high sequence conservation across the 12 Drosophila species. LncRNA-CRG is located downstream of the Ca2+/calmodulin-dependent protein kinase (CASK) and partially overlaps with the 30 UTR of CASK, where CASK is a behavior-related coding gene. All these lncRNAs and their properties are summarized in Table 1.3.2. LncRNAs in Apis MelliferaHoney bees have socio-economic importance for both mankind and nature. The division of labor in honey bees has been well studied, which represents an excellent coordinated work of society [106]. Worker bees have an age-dependent division of labor; young worker bees perform the duties of nursing the brood and old ones are involved in foraging (collection of nectar and pollen) [107]. This transition is thought to be regulated by the interaction of yolk protein, vitellogenin (Vg), endocrine factor, juvenile hormone, and the biogenic amine, octopamine. Furthermore, microarray analysis revealed the preferential expression of certain genes (foraging, malvolio, Hormone-like Receptor in 38 (HR38)) in the forager bee brain. LncRNAs have been found to function in developmental processes in honey bees [108]. Enhanced expression of a non-coding RNA, Nb-1(Nurse bee brain-selective gene-1) was found in the brain of nurse bees compared to that in forager and queen bee brains in normal colonies [109]. Nb-1 is specific only to some hymenopteran species. Nb-1 is involved in age-dependent transition in worker bees by regulating the synthesis and secretion of octopamine and juvenile hormones [109]. Other lncRNAs reported in the brain of honey bees are KS-1, AncR-1, and kakusei. KS-1 (Kenyon cell/small- type preferential gene-1) is a 17 kb nuclear transcript that is preferentially expressed in Kenyon cells of the honey bee brain [110]. Kenyon cells are subtypes of inter neurons in mushroom bodies (MB), which is an essential and regulatory part of the insect brain. Ks-1 positive neurons were found to be greater in drones than in queens between the lateral calyx and the optic lobes of the brain, suggesting their involvement in drone specific brain functions. Overall nucleotide sequence of Ks-1 was found to be conserved among the honey bees [111]. AncR-1 is another lncRNA that was discovered in honey bee brains and is also expressed in sex and secretory organs. Localization of AncR-1 and Ks-1 transcripts in a distinct portion of a single neural nucleus suggests their involvement in distinct neuronal functions in the brain [112]. The kakusei locus produces both constitutive- and inducible-type variants of the kakusei transcript in the brain of worker bees [113]. Some kakusei transcripts are produced as a result of increased neural activity whereas some are constitutively expressed (i.e., independent of neural activities). Since, there is neither an alternative splicing (within the kakusei locus) nor extended transcription from the kakusei cDNA sequence, constitutive-type kakusei variants might be produced by differential transcription initiation or termination within the anterior part of the kakusei cDNA sequence. Both kinds of transcripts are localized predominantly in the neural nuclei and might be related to different nuclear functions [113,114]. LncRNAs, lncov1 and lncov2, are intronic lncRNAs (located within the first introns of their host genes) that are expressed in honey bee ovaries. The host gene for lncov1 is ‘LOC726407′ (with unknown function) and lncov2 is fringe, a homolog of an important Drosophila developmental gene, with the latter, found to be over-expressed in larval queen ovaries and was involved in JH-dependent maintenance of developing ovarioles in the early fifth instar of queen larvae [115]. The expression of lncov1 was found to be maximum in the fifth instar larvae, exactly before they enter into metamorphosis. This coincides with the autophagic cell death in the larval worker ovary, which indicates the role of lncov1 in autophagic cell death during larval to pupal metamorphosis.Apart from various species of honey bees, two significant species—Apis mellifera (Western honey bee) and Apis cerana (Asian honey bee)—play crucial roles in development, social behavior, and disease transmission, and also serve as the significant pollinators of economically important crops [125,126]. A total of 2470 lincRNAs, encoded by 2376 gene loci in the A. cerana genome and a total of 1514 lincRNAs in the A. mellifera genome were identified in silico [20]. These lincRNAs were presumed to be associated with functions like hormone signaling, metabolism, association with diseases, and role in gene modulation. The comparative analysis of lincRNA between two sister species resulted in high conservation among them, and only 5% were found to be unique to each species. Both species showed tissue-specific expression of lncRNAs; in the case of A. cerana, more lincRNA were expressed in the fat body and antenna while in A. mellifera, expression was visible in the ovary and brain [20]. These results were considered to be the proponent of the role of lincRNA in the major metabolic and hormone signaling pathways in insects. Chen et al. performed strand specific whole transcriptome RNA-sequencing of control and Nosema ceranae infected midgut samples of Apis mellifera ligustica workers. This study resulted in the identification of a total of 6353 lncRNAs out of which 4749 were conserved lncRNAs and 1604 were novel lncRNAs. Interestingly, these lncRNAs did not show any similarities with other known lncRNAs in other species, though there was some structural similarity with counterparts in mammals and plants. Additionally, 27 discovered lncRNAs were harboring eight known miRNA precursors. This study gave a basis for understanding the host–pathogen interaction in Apis mellifera and investigating the roles of lncRNAs associated with this process [127].3.3. LncRNAs in Aedes AegyptiAedes aegypti (the yellow fever mosquito) is a major vector of arboviruses including dengue, Zika, and Chikungunya viruses [128,129]. A recent study in Aedes aegypti has led to the identification of 4689 novel lncRNA transcripts, which includes 2064 intergenic, 2076 intronic, and 549 antisense lncRNAs [130]. Genome-wide analysis and developmental profiling of these newly identified lncRNAs suggests that a subset of lncRNAs shows maternal inheritance and early embryonic expression [130]. One of the probable reasons for implicating the involvement of lncRNA in developmental regulations is their proximal location to the genes involved in various developmental processes. High expression of few lncRNAs was also found in the ovary of blood-fed mosquitoes, which lasts up to 12 h of embryonic development, suggesting their maternal supply [131]. The temporal expression and maternal inheritance of lncRNA in A. aegypti account for developmental transition in mosquitoes [132].3.4. LncRNAs in Anopheles GambiaeJenkins et al. identified 2949 lncRNAs from different life stages of malaria mosquito Anopheles gambiae using RNA-Seq data [18]. These authors observed that most (2059) of the identified lncRNAs were intergenic, whereas some (108) were anti sense and some (782) were mapped within the intronic region of protein-coding genes [18]. These lncRNAs displayed very little sequence conservation compared to protein-coding genes in different Anopheles species. Interestingly, the secondary structural features for many lncRNAs were found to be conserved. While the lncRNAs are thought to be the potential targets of epigenetic regulation and controlling vector-transmitted infectious diseases, they are also known to impede the ncRNA targets in vector insects. The evolution of lncRNA secondary structures tends to follow the concept of ‘pervasive transcription’ (i.e., most regions of the genome are transcribed, even those that failed to encode proteins or functional ncRNAs) [133].3.5. LncRNAs in Bombyx MoriDomesticated silkworm, Bombyx mori is an economically important insect [134]. Besides being the primary producer of silk, it is also a lepidopteran model insect that has attracted researchers in answering several biological questions. Silkworm displays several sexually dimorphic behaviors that are associated with the insect brain. A lncRNA, Fben-1(female brain expressed noncoding RNA-1)is preferentially expressed in the cells surrounding the mushroom bodies in the brain of female silkworm, suggesting the involvement of Fben-1 in certain neural or developmental/sexual functions in females [135]. Fben-1 is located ~6 kb upstream of the fruitless (fru) gene in the B. mori genome. LncRNA Bmdsx-AS1is expressed abundantly in B. mori testis and is involved in the sex-specific alternative splicing of pre-mRNA of its doublesex (Bmdsx) gene [136]. doublesex (dsx) has been found to be conserved in all the insects studied to date [137]. The pre-mRNA of dsx is alternatively spliced to produce sex-specific transcripts, which ultimately make sex-specific proteins responsible for all kinds of sexual dimorphism in insects [138].Bmdsx-AS1 lncRNA interacts with BmPSI (B. mori P element somatic inhibitor) via a splicing factor hnRNPH [118]. BmPSI is a male-specific protein that promotes male-specific splicing of Bmdsx pre-mRNA by binding to the CE1 sequence (exonic splicing silencer) present in the fourth exon of Bmdsx, leading to its skipping [139]. Another lncRNA of B. mori, iab-1, is coded by the intergenic region between Bmabd-A and Bmabd-B in the Homeobox (Hox) cluster of the silkworm. At least seven alternatively spliced iab-1 lncRNAs are produced and are expressed at specific developmental stages. The siRNA mediated downregulation of iab-1 leads to the death of larvae, suggesting its involvement in certain essential physiological processes [119]. A recent study showed the expression of lncRNAs in response to infections. B. mori nucleo-polyhedrovirus (BmNPV) is threatening to economic growth and the stability of sericulture business. The BmNPV infection in silkworms leads to the expression of 4450 lncRNA [140]. In-silico analysis suggests the target of these differentially expressed lncRNAs to be the genes involved in ubiquitin mediated proteolysis, endocytosis, and lysosomal pathways [141].3.6. LncRNAs in Plutella XylostellaResearchers have identified 3844 lincRNA from Plutella xylostella (a deadly pest of cruciferous plants), which has developed field resistance to all classes of insecticides. Some of the identified lincRNAs were found to serve as the precursor of small ncRNAs. These lincRNAs can be linked to the development of resistance to a large variety of insecticides in P. xylostella as differential expression of lincRNA was observed in the larvae of P. xylostella that were resistant to insecticides like organophosphate, phenylpyrazole, and Bt endotoxins [24,142]. Interestingly, most of the presumptive lincRNAs were found to be over expressed whereas some were found to be repressed in deltamethrin resistant larvae. Previous studies have shown the differential expression of lncRNAs with respect to various stress factors. Recently, a study compared the dsRNA induced lncRNAs in three insect species: Helicoverpa armigera, Plutella xylostella, and Tribolium castaneum. A total of 3463 H. armigera, 6245 P. xylostella, and 3067 T. castaneum differentially expressed lncRNAs were identified with respect to dsRNA induction [121].3.7. LncRNA in Planthopper InsectsWell-known crop pests of Asia include Soagetlla furcifera (white-backed planthopper, WBPH), Nilaparvata lugens (brown planthopper, BPH), and Laodelphax striatellus (small brown planthopper, SBPH). With the help of NGS, genomes of all three planthoppers were sequenced, resulting in the identification of many functional genes [143,144,145]. From this study, 2439 lncRNA transcripts were reported in BPH [146], 5790 lncRNA transcripts were identified in SBPH [147], and 1852 lncRNA transcripts were found in WBPH [148]. Interestingly, these lncRNAs from different insects shared similar sequences due to their close genetic relationships. These lncRNAs were found to play distinct roles in fecundity, virulence, and developmental regulation.4. Challenges and Future PerspectivesDuring the past few years, studies on lncRNAs have steadfastly gained momentum, albeit the insect group remains unexplored. Despite the large diversity of lncRNAs, the study of lncRNA is either limited to few model insects or is at the elementary level of research in other insects. Given the fact that the lncRNAs possess a greater potential to interact with other genetic materials like DNA, RNA, or protein molecules, it is likely that gene expression studies would yield possible insights into understanding their biological phenomena. However, due to the scarce information about the functional aspects of lncRNAs, there is a need to resolve the notion of the hidden non-coding potential of lncRNA. In the past decade, many technologies and strategies were tailored to identify and characterize lncRNAs. Continual efforts in the direction of annotating different kinds of lncRNA has allowed for the discovery of their biological roles, although the significant challenges of filtering out ncRNAs from coding RNA and the lack of proper bioinformatics tools to identify multifunctional RNA still exist. The development of highly sensitive and specific techniques is required to identify interactions between lncRNAs and other molecules (chromatin, RNA, protein) or locating the molecules found in proximity to lncRNA in order to determine their regulatory functions. The insufficient knowledge and limited potential of in silico tools to predict the mechanisms and functional domains within an lncRNA also poses a major task. Previously, it was considered to be the junk part of the genome, but later came to light as a key regulator in chromatin remodeling, DNA methylation, and RNA editing. Apart from epigenetic regulation and chromatin remodeling, lncRNAs are also known to serve developmental processes, immunity responses, etc. Hence, lncRNAs can potentially be used in therapeutics. For example, in Drosophila, few lncRNA perform a vital role in neurogenesis and spermatogenesis; this point marks their importance in countering any fetal abnormalities [66,71]. The involvement of lncRNA in cellular processes like oncogenesis and tumor suppression may provide an opportunity of developing lncRNA as therapeutic targets of cancer [149]. Likewise, in mammals, lncRNAs exhibit varying expression levels upon severe acute respiratory syndrome coronavirus (SARS-CoV) viral infection, regulating the innate immune response [150]. Similarly Zhou et al. also accounted for evidence regarding the association of Arid2-IR lncRNA with renal fibrosis and renal inflammation [151]. One particular importance of lncRNA is their involvement in host–virus interaction and conferring insecticide resistance in insects [28,152]. Due to the significant heterogeneity of lncRNAs, they can be potentially used in improving insect control strategies as well as in predicting the mechanisms underlying the developmental processes as they are commonly designated as time-specific tuners by governing the time of developmental transitions.5. ConclusionsThe complex patterns of expression and regulation of the entire eukaryotic genome are in the hands of ncRNAs. The ncRNAs participate in almost all sorts of biological processes including epigenetic control of traits, transcriptional, and post-transcriptional regulations. Interestingly, a growing number of ncRNAs have been identified in insects primarily from the RNA-Seq of RNA libraries or transcriptomics studies. Small ncRNAs such as miRNAs and piRNAs have been well studied compared to lncRNAs. In this review, we discussed the importance of lncRNAs in different insect orders. Keeping in mind the rationale that insects are the most abundant and diverse group of animals and play an important role in different ecosystems aside from significantly contributing to the population dynamics, their influence on the ecological and economic environments of humans is undebated. Studies on RNAs and their role in gene regulation using insect models will open windows to understanding the immune response and disease dynamics. Although there are currently many studies on lncRNAs in insects, all these studies are still at the preliminary stages and focus solely on model insect species. One particular aspect, which could be promising, is to compare RNA at the cross-sectional level. Furthermore, insects are particularly ectothermic and global warming has pushed them to newer territories where human interaction has increased many fold. In this scenario, ncRNAs, particularly lncRNAs, are a promising solution to novel diseases. An essential aim of developing our understanding of lncRNAs is to create environmentally-friendly and efficient pest control strategies for agricultural pests. Therefore, lncRNAs have great potential to be used as targets in pest control strategies in the future.

animals : an open access journal from mdpi

10.3390/ani13111810

PMC10252065

Recent reports focusing on the extent of plastic pollution have shown that petroleum-based products can currently be found in most marine species, which is a consequence of the immense production and use of plastics. The severe contamination of plastic nano-/microparticles (NPs/MPs) mainly results in immediate negative outcomes, such as organic impairments and tissue damage, as well as long-termed negative effects, such as developmental retardation and defects, chronic inflammation, oxidative stress (OS), metabolic imbalance, mutagenesis, and teratogenesis. We aimed to correlate the possible toxic effects of plastic NPs/MPs in zebrafish models, by focusing on OS and developmental processes, and the size, shape, and doses of the NPs/MPs. We found that plastic NPs/MPs toxic effects could be observed during the entire developmental span of zebrafish in close correlation with OS-related changes. The decreased antioxidant enzymatic defense due to plastic NPs/MPs exposure and accumulation suggests important neurodevelopmental negative outcomes (cognitive abnormalities, neurodevelopmental retardation, and behavioral impairments) and neuronal effects, such as impaired digestive physiology.

Recent reports focusing on the extent of plastic pollution have shown that many types of fibers and polymers can now be found in most marine species. The severe contamination of plastic nano-/microparticles (NPs/MPs) mainly results in immediate negative outcomes, such as organic impairments and tissue damage, as well as long-termed negative effects, such as developmental retardation and defects, chronic inflammation, oxidative stress (OS), metabolic imbalance, mutagenesis, and teratogenesis. Oxidative responses are currently considered the first line molecular signal to potential toxic stimuli exposure, as the oxidative balance in electron exchange and reactive oxygen species signaling provides efficient harmful stimuli processing. Abnormal signaling or dysregulated ROS metabolism—OS—could be an important source of cellular toxicity, the source of a vicious cycle of environmental and oxidative signaling-derived toxicity. As chemical environmental pollutants, plastic NPs/MPs can also be a cause of such toxicity. Thus, we aimed to correlate the possible toxic effects of plastic NPs/MPs in zebrafish models, by focusing on OS and developmental processes. We found that plastic NPs/MPs toxic effects could be observed during the entire developmental span of zebrafish in close correlation with OS-related changes. Excessive ROS production and decreased antioxidant enzymatic defense due to plastic NPs/MPs exposure and accumulation were frequently associated with acetylcholinesterase activity inhibition, suggesting important neurodevelopmental negative outcomes (cognitive abnormalities, neurodevelopmental retardation, behavioral impairments) and extraneuronal effects, such as impaired digestive physiology.

1. IntroductionDue to the immense production and use of plastics, residual and non-recyclable plastic nano-, micro-, and macroscopic particles have been found to accumulate in water, sea food, fish, and birds, eventually posing significant threat to human health [1,2,3,4]. Recent reports have shown that more than 1900 different plastic items (categorized by polymer composition) were found in seafood in the last year [5], while many fibers and polymer types were found in most marine species [6].Considerable efforts have been made to describe the potential toxicity and mechanism of action of microscopic plastic particles, especially of polystyrene, mainly using animal models, such as rodents and fish. Thus, it has been shown that plastic nano-/microparticles (NPs/MPs) tended to accumulate in the lungs following inhalation from environmental sources [7], or in the gastrointestinal tract of some fish species from waters polluted with plastic NPs/MPs [8]. The severe intoxication of plastic NPs/MPs mainly resulted in immediate negative outcomes, such as organic impairments and tissue damage, as well as long-termed negative effects, such as developmental retardation and defects, chronic inflammation, OS, metabolic imbalance, and mutagenic and teratogenic potential [7,8,9,10,11].In modern molecular toxicology, oxidative response is considered the first line molecular signal to toxic stimuli exposure [12]. While the modulatory activity of oxidative/reducing pathways and reactive oxygen species (ROS) signaling provides oxidative balance, the impairments that occur in processing and responding to harmful stimuli, as a result of abnormal signaling or dysregulated ROS metabolism, could be an important source of cellular toxicity [13,14]. In this context, OS could provide the molecular premise for the vicious stress cycle including both external (chemical, biological, and/or physiological) and oxidative signaling-derived toxicity [14,15,16]. As chemical environmental pollutants, plastic NPs/MPs are no exception to causing such toxicity, as a recent in vitro study on human brain cell lines demonstrated that exposure to polystyrene (PS) MPs is correlated with OS in a cause-effect relationship that is modulated by overproduction and accumulation of ROS [17].By relation to oxidative response, plastic NPs/MPs toxicity has also been reported in several animal models, such as mice manifesting increased OS and inflammation in brain tissues [18,19], and mussels showing low enzymatic antioxidant defense in gills and digestive glands [20], and OS-mediated immunotoxicity [21], while fish exhibited severe OS in cerebral tissues [22]. PS, polypropylene (PP), polyethylene (PE), and poly (methyl methacrylate) (PMMA), solely or in mixture, have all been shown to cause similar toxic effects on oxidative metabolism and development [23,24,25]. Given our previous experience in zebrafish toxicology studies and the current interest in plastic NPs/MPs pollution, our aim was to describe, summarize, and correlate the possible toxic effects of plastic NPs/MPs at different stages in the life cycle using an animal model, focusing on the operating system and developmental processes. Efforts have also been made to hypothesize several possible OS markers useful in the assessment of plastic NPs/MPs poisoning.Initially, zebrafish models were used in anticancer therapy research [26]. Once their genetic and phenotypical characterization was uncovered, their potential use for a wider range of research applications was outlined. Moreover, the physiological advantages of zebrafish contributed to their fast emergence in animal model research. The low breeding costs, simple maintaining conditions, and respective ethical regulations also contributed to their demonstrated versatility. Regarding the potential of zebrafish model in neurotoxicological and neurodevelopmental studies, Kalueff et al. [27] suggested that their value resided in the cognitive abilities, sleeping patterns, and brain tissues cellular morphology. Moreover, they described the similarities between the general macro-organization of the zebrafish and rodents’ brains, despite the increased evolutionary distance between the two animal classes. Recently, our group also thoroughly described these aspects and suggested that diverse measurable behavioral patterns are available to be assessed as early as pre-hatching stages of life [8,28]. Similarly, considering that the blood-brain barrier is functional from 36 hpf (hours post fertilization), zebrafish embryos could be used for neurotoxic effect evaluation for various molecules that are capable of penetrating the blood-brain barrier [29]. In this context, it is also valuable to mention that zebrafish reach the peak of cognitive development in 90 dpf (days after fertilization) through a process that progresses rapidly from the embryonic stage, larvae (72 hpf), juveniles (30 dpf), until the adult stage [30]. While behavioral and molecular models are gradually developing during this period, the range of biomarkers has becoming more complex and provides a powerful assessment tool for aggression, sleep, locomotion, memory, and social preferences [9,31].The molecular patterns of oxidative balance develop concomitantly with the behavioral patterns. Despite the fact that the zebrafish brain is less complex than rodent and human brains, it is a major source of reactive oxygen species and is thus the most susceptible tissue to oxidative stress damage, one of the most common preserved traits. The effects of OS on the cerebral tissues and central nervous system were previously extensively discussed by our group while describing new pathophysiological patterns or bringing additional evidence to support the emergent OS pathological hypothesis in neurophysiology [9,32,33,34,35,36,37,38,39,40]. In this context, the evaluation of enzymatic and non-enzymatic antioxidants and OS effects on cellular and molecular structures could be observed in zebrafish in a tissue-specific manner but is not restricted to it [41].Another advantageous feature of zebrafish is their completely sequenced genome and their facile manipulation by successfully using the innovative molecular techniques, such as the CRISP/Cas9 editing tool [42]. Moreover, 70% of the 26,206 zebrafish protein-coding genes have at least one orthologue in humans [43].2. Plastic Nano-/Microparticles Accumulation and Oxidative Stress ResponsePlastic MPs are solid masses of plastic that come from particles (size less than 5 mm) that consist of mixtures of polymers and various additives. These plastic MPs are either prefabricated or are derived from the gradual degradation of plastic materials in the environment. Plastic NPs (size less than 1 µm) exhibit a colloidal behavior which makes them more reactive and toxic than plastic MPs and their counterparts. This is mainly due to their very small size that facilitates their ingestion by animals and their penetration through biological membranes [44]. From a toxicological point of view, both plastic NPs and MPs have been reported to exhibit bio-accumulative behavior mainly in the soft tissues and to commonly predispose to OS disregarding the developmental stage [45]. For instance, PS NPs were found to accumulate in the brain, gills, blood, liver, and digestive tract of zebrafish immediately after hatching (Figure 1) and cause oxidative DNA damages and developmental malformations [46].However, it has been shown that the chorionic barrier could prevent plastic MPs accumulation, when they are smaller than 100 nm. Duan et al. [47] recently described the accumulation dynamics of plastic NPs in zebrafish development. Starting with 72 hpf, plastic NPs were absorbed through oral intake and gills, while at 96 hpf, they were easily found in blood circulation, and gradually accumulated in the liver and digestive tract. Further bioaccumulation was observed at 120 hpf in the brain, eyes (<50 nm PS NPs), intestinal tract, and outer epidermis (>50 nm PS NPs) of the zebrafish larvae, as plastic NPs were able to penetrate the blood-brain barrier [47,48]. Similarly, Parenti et al. [49] reported that the PS NPs tended to accumulate in the intestinal tract and later migrate to other tissues in a size-dependent manner, as it was shown in 72 to 120 hpf embryos exposed for 48 h to 500 nm PS NPs. Furthermore, mild OS following short-term exposure to PS NPs of 500 nm has also been reported in zebrafish embryos: significantly increased SOD activity, decreased COX activity, but no relevant CAT and GPx activity changes [49]. Despite these, PS NPs have been shown to induce apoptosis in brain tissues of zebrafish embryos, developmental malformations, and excessive ROS activity [46].The plastic MPs accumulation in zebrafish adults has been thoroughly described as variable, multifactorial, and size/shape dependent. While the smaller plastic MPs (i.e., 5 μm PS MPs) have been reported to mainly accumulate in gills, liver, and gut, the larger ones (i.e., 20 μm PS MPs) tend to accumulate only in gills and gut [50] following short-termed exposure (96 h). This could suggest that the circulatory system and the caliber of blood vessels plays an important role in the plastic MP transport though the body.Also, it has been reported that longer exposure times could further cause more severe molecular response in the tissues they accumulate in (Figure 2). Gu et al. [51] showed that both small (i.e., 100 nm) and larger (i.e., 5 μm) PS MPs accumulation in the intestines during a 21-day chronic exposure could lead to significant OS and sustained inflammatory response, as suggested by intense mucus secretion. Similar effects were also reported by Qiao et al. [52] showing, impaired intestinal lipid metabolism correlated with intestinal inflammation in a comparable exposure design (21 days exposure of zebrafish adults to 5 μm PS MPs). In this context, significant OS and tissue structure changes were reported consequently to liver antioxidant enzymes activity impairment (SOD and CAT) [50,52,53] (Figure 2).Qiao et al. [53] observed that fibers, fragments, and beads tend to exhibit different accumulation patterns: plastic microfibers tend to cause more severe intestinal toxicity (mucosal damage, inflammation, microbiota’s dysbiosis, and intestinal metabolism disruption), as compared to micro-fragments and microbeads.Furthermore, Pitt et al. [54] reported that the plastic NPs/MPs toxicity could extend beyond the individual effects, as they showed that plastic particles could be passed to offspring through maternal and paternal transfer. Plastic MPs accumulation in the chorion was described in F1 embryos, even if only one parent exhibited plastic particles intoxication. The accumulation in the yolk sac of the embryo was higher in co-parental transfer, as early as at 48 hpf for maternal transfer and at 72 hpf for paternal transfer. In this case, OS changes were reported in a sex-dependent manner. While muscular glutathione reductase (GR) was significantly decreased in both sexes, brain GR levels were significantly lower in females than in males, and conversely, in the gonads. Additionally, brain GPx activity was higher in the female zebrafish, while no significant changes were observed in CAT activity, as compared to male zebrafish [54] (as summarized in Table 1).3. Plastic Nano/Microparticles Co-Exposure and Oxidative Stress ResponseConsidering the mixed exposure patterns that occur in the environmental conditions, extensive studies have shown that the interaction of plastic MPs with various environmental pollutants are needed to describe the correlations between co-exposure to multiple pollutants and their effects on development and molecular response. While many previous studies thoroughly described the effects of the main contaminants on development and OS, it was suggested that their interaction with other plastic NPs/MPs is based on aggregation, adsorption and transformation, which could lead to either synergistic or antagonistic effects, or harmful potentiating effects [63]. In this way, it has been reported that the 96 hpf co-exposure of plastic MPs with copper (Cu) could induce morphological malformations and abnormalities of the body structure, disruption of retina layers, and OS in embryos [64]. However, the decrease in SOD activity was dependent on Cu concentration, suggesting that MPs co-exposure could lead to potentiating or cumulative toxicological effect. Despite that, compensatory cellular responses to the increased ROS levels (increased GPx and CAT activities, and increased GSH levels) were observed following low and intermediate Cu concentrations and co-exposure. Thus, these results could suggest that following extensive exposure periods (14 dpf), plastic MPs could act as Cu vehicles, increasing the toxicological effects. All in all, plastic MPs solely or co-exposed with Cu could induce increased mortality rates, neurotoxicity (acetylcholine pathways modulation), and OS, in a suggestive synergistic manner (via bioavailability increase effect).Similar synergistic effects were reported in short-term co-exposure (24 h) of PS nanobeads (50 nm) combined with gold (Au ions 1 μg/mL), which lead to an increased mortality rate, altered development, reduced hatching rate, and increased production of ROS in zebrafish larvae [65]. Regarding this aspect, Lee et al. [65] suggested that gold-induced ROS production was synergistically aggravated in the presence of PS nanobeads.Long term exposure (three weeks) to PS microbeads combined with cadmium (Cd) in 18-weeks-old zebrafish adults led to increased plastic MPs bioaccumulation in the gills, liver, and gut [66]. Also, significant histological changes and plastic MPs concentration-dependent OS were reported in the tissues targeted by accumulation. GSH levels and SOD activity showed dose-dependent and co-exposure-dependent changes in the gills and gut. However, the most significant OS-related toxicological effects were observed for SOD activity in the gills of the Cd-treated fish, suggesting that plastic MPs could be implicated in Cd transport, but not in potentiating its toxicity. However, previous studies have shown that Cd is able to pass through the blood stream barrier and is predisposed to trigger ROS-mediated neurological impairments [67] without the assistance of plastic MPs.Furthermore, it has been suggested that the toxicological effects of co-exposure to plastic MPs and other contaminants could further escalate when combined with other factors, such as exposure to ethylhexyl salicylate (EHS), one of the major organic UV filters commonly found in the environment. In this regard, Zhou et al. [68] reported significant OS changes (decreased SOD and CAT activities), suggesting that PS NPs could potentiate the pro-oxidative effects of EHS and its bioaccumulation in the offspring leading to mild OS. Subsequent increases in ROS and MDA content indicated that PS NPs could synergize the oxidative toxicity of EHS in offspring [68].The co-exposure to various other contaminants has been shown to alter the antioxidant enzymatic defense and the oxidative balance. Increased MDA levels were found in adult zebrafish bodies following a 5-day exposure to PS NPs, metal oxides NPs, and polycyclic aromatic hydrocarbons, as compared to respective single treatments. On the other hand, decreased CAT activity was observed in both single and co-exposed groups, as compared to control. Similar effects were also reported in PS/Cu oxide or Zn oxide NPs and polycyclic aromatic hydrocarbons, in a potentiating manner [69] (as summarized in Table 2). This could suggest that the adsorption of toxic contaminants by plastic particles can alter the bioaccumulation and toxicity profile.4. Behavioral Analysis and Oxidative Stress ResponseRegarding the possible neurobehavioral effects of plastic NPs/MPs, recent studies have reported that PS NPs exposure alters larval behavior, as suggested by swimming hypoactivity in larvae [56] and impaired angle turning behavior in embryos [49]. Also, the latter study evaluated OS response following PS NPs exposure in embryos and reported significantly decreased SOD activity. Moreover, it has been shown that the negative effects of PS NPs exposure (similar types and sizes) could go even further, as Brun et al. [71] observed altered gene expression in glucose metabolism, cortisol secretion, and OS, correlated with impaired locomotor and behavioral patterns in zebrafish larvae, all of which could clearly suggest neurobehavioral toxicity. While the developmental neurotoxicity of plastic NPs was described, defective brain or nervous system functions in correlation with OS, such as increased CAT and GPx activity and ROS overproduction have been observed in zebrafish larvae [55].However, the most frequent and constantly observed molecular response regarding OS following PS NPs intoxication was SOD activity impairment in all developmental stages. This pattern has been correlated with the reported developmental defects and behavioral impairments [49,50,52,53]. Regarding the possible molecular patterns of PS NPs toxicity, Pedersen et al. [56] described that the behavioral hyperactivity during dark cycles is correlated with dysregulated neurological and neuromuscular signaling and modulation pathways, generally occurring in neuropsychiatric disorders. For example, they reported impaired expression profiles of the SLC6A1 gene, which is commonly found in ADHD models. Moreover, PS MPs lead to impaired locomotor activity (as suggested by the altered larval swimming behaviors observed in the free-swimming test) and to subsequent upregulated OS-related genes expression (i.e., CAT gene) [55].Similar patterns were observed in zebrafish adults; Mak et al. [60] reported that abnormal behaviors, such as epileptic seizures, and erratic movement were observed following PE microbeads exposure. These impairments were generally correlated with genetic defects, illness, toxic agents’ exposure, or aging in adult individuals, while in larvae they were commonly associated with neurological phenotypes [72]. Moreover, Mak et al. [60] reported caudal fin deformation in adult zebrafish exposed to PE microbeads, which is regularly an important marker for morphoanatomical alterations that are suggestive of neurological seizures, tissue remodeling, pain, and/or inflammatory processes, all of which could indicate neurotoxic and neurodegenerative effects in adult individuals [73].Taking into consideration that previous reports describing the significant correlations between affective disorders and OS [36], reduced aggressivity, impaired predator avoidance behavior, and social behavior associated with increased stress and anxiety has also reported in zebrafish adults following PS NPs exposure [62]. Furthermore, it has also been shown that PS NPs exposure resulted in impaired acetylcholine metabolism, mitochondrial chain, and excessive ROS production. Additional evidence on the correlation between the plastic MPs exposure and anxiety-like behavioral impairments was highlighted by Sarasamma et al. [62], who showed that the levels of some of molecules implicated in socio-affective behavior (oxytocin, vasopressin, serotonin, dopamine, and melatonin) modulation were impaired following PS NPs exposure.In this way, the previously presented results could suggest that constant alterations in zebrafish behavioral patterns could be established as a result to plastic MPs exposure, and could be further correlated with several molecular changes occurring in brain signaling, toxic agents’ degradation, and OS. However, further studies regarding these aspects are needed, as the patterns were shown to be dependent on the type, size, and doses of plastic particles used in exposure. Also, the molecular mechanisms underlying the behavioral impairments and their correlation with OS mechanisms could be further studied in the context of plastic NPs/MPs intoxication, since the correlation between two out of the three components of this multifactorial interaction has already described [9,32,33,34,35,36,37,38,39,40].5. Developmental Anomalies and Oxidative Stress ResponseThe developmental changes and impairments are among the most important evidences regarding the toxicological effects of chemical or environmental factors, with plastic NPs/MPs being no exception. In this way, the toxicological effects of plastic MPs effects on zebrafish development have been seen at every developmental stage. Recent studies have predominantly shown body malformations, such as pericardial and yolk sac oedema, abnormal tail, and axial curvature in zebrafish embryos [46,64]. Also, typical of toxicological evaluations, exposure to plastic MPs has been shown to induce increased mortality rates and delayed hatching, correlated with increased OS in the larval head region [47]. Furthermore, these changes suggested that the adsorption of PS NPs and MPs through the outer surface of chorion could lead to changes in chorion mechanical properties, further predisposing to embryonic hypoxic microenvironment. Regarding this aspect, Duan et al. [47] recently presented additional evidence that shows that the chorion property changes could be correlated with impaired heart rate, blood flow speed, and a slower hatching rate of the embryos, thus significant developmental toxicity.The recent studies also showed that OS could be co-occurring with developmental defects as a result to NPs/MPs exposure. In this way, glutathione reductase (GR), GPx, and CAT activities have all been shown to be altered in developmentally impaired offspring, not necessarily in a sex-correlated manner [54]. Also, Chen et al. [58] reported both larvae body length reduction and increased CAT and GPx activities following PS NPs exposure. The correlation between the two different components was suggested by the changes occurring in the acetylcholine pathway impairments (as observed in the reduction of acetylcholine esterase activity). In this context, they suggested that the developmental abnormalities were significantly correlated with excessive ROS production, impaired CAT and GPx activity, and neurotoxic effects. As bradycardia was also observed in embryos originating from parents exposed to PS NPs [54], the developmental defects could also be the result of parental transfer.A similar neurotoxic response was observed in the impaired swimming behavior in adults and in reduced larvae locomotor activity [49,56]. Different responses to stress and energy regulation networks, combined with hyperactivity, especially at night, could counteract potential motor dysfunctions and neurodegenerative effects to promote a hyperactive phenotype [57]. Antioxidant activity may be increased or inhibited under chemical stress depending on the intensity and duration of exposure.Furthermore, the inhibition of acetylcholine esterase activity serves as a prominent biomarker of neurotoxicity since it causes severe neurotransmission impairment. As a consequence, the accumulation of acetylcholinesterase at synaptic gaps could induce hyperstimulation and ultimately death from respiratory or cardiac failure [74]. The sensitivity of acetylcholine esterase activity to various chemicals, including emerging pollutants in the environment such as nanomaterials, suggests the usefulness of this biomarker in providing an integrative measure of overall neurotoxic risk [75]. PS MPs could also cause gene expression modulated neurotoxic responses in the adult brain, including the inhibition of acetylcholine esterase activity in brain and liver, increased ROS activity, intense lipid peroxidation, and decreased CAT, SOD, and GPx activities in brain and liver [61].6. ConclusionsThe developmental and molecular toxic effects of plastic NPs/MPs depend on several factors, such as their size, shape, and doses. Each developmental stage can exhibit different patterns of behavioral and molecular responses following plastic NPs/MPs exposure. Developmental impairments, such as morphoanatomical defects and lower hatching rates, and molecular responses, such as upregulation of SOD, GPx, and CAT activity, have been commonly found in embryos exposed to plastic NPs/MPs. In zebrafish adults, the toxic effects of plastic NPs/MPs are similar and include significant OS (increase in ROS production, higher GPx, CAT activity), as well as behavioral and neurosignaling impairments (altered locomotion and socio-affective behavior correlated with OS and decreased AChE activity). Despite some variations in plastic NPs/MPs toxic effects throughout the individual development of zebrafish models, similar patterns of OS were generally seen. However, the SOD activity changes have often been reported to be dependent on the size of plastic NPs/MPs, suggesting that OS is relevant in the context of evaluating the molecular toxic effects of plastic NPs/MPs.

animals : an open access journal from mdpi

10.3390/ani11030653

PMC8001368

The family Canidae, composed of dog-like species such as wolves, foxes, and jackals, demonstrates a significant variety in reproductive biology. In general, female canids experience very long periods of ovarian inactivity during the year; however, there are diverse patterns with regard to seasonality between species, as well as within an individual species depending on geographic region or housing status. Understanding of these differences is critical to the development of assisted reproductive technologies for canid conservation efforts. This review summarizes the current knowledge of canid reproduction, including reproductive cyclicity, seasonal breeding, male sperm traits, and recent developments in assisted reproductive technologies for canids.

The reproductive physiology of canids is unique compared to other mammalian species. Specifically, the reproductive cycle of female canids is characterized by extended periods of proestrus and estrus followed by obligatory diestrus and protracted ovarian inactivity (anestrus). Although canid reproduction follows this general pattern, studies have shown variations in reproductive biology among species and geographic regions. Understanding of these differences is critical to the development of assisted reproductive technologies including estrus induction, gamete rescue, and embryo production techniques for canid conservation efforts. This review summarizes current knowledge of canid reproduction, including estrus cyclicity, seasonality, and seminal traits, with the emphasis on species diversity. The application of reproductive technologies in wild canid conservation will also be discussed.

1. IntroductionIncluding the domestic dog, the canid family consists of 37 extant species, five of which are listed as “endangered” or “critically endangered” by the International Union of Conservation of Nature (IUCN) (Figure 1a). Furthermore, 1/3 of wild canid species are in decline due to increased anthropogenic activities, such as agricultural development, whereas two species, including the coyote (Canis latrans) and golden jackal (Canis aureus) are abundant with increasing population trends (Figure 1b).Reproductive sciences play critical roles in species conservation and management [1,2]. Specifically, advances in the understanding about species’ reproductive biology may permit the development of reproductive technologies that are useful for ensuring genetic and demographic viability of ex situ wildlife populations [1]. Such knowledge can also assist in the development of strategies to control overpopulated wildlife species [1]. Yet, reproductive biology has not been studied in more than 90% of mammals on the planet, and the knowledge gained from “well-studied” species have highlighted remarkable diversity in reproductive mechanisms, even among related species [3]. For example, within the Genus Canis, the domestic dog (Canis familiaris) displays non-seasonal reproduction, whereas their wild cousins, including grey wolf (Canis lupus), red wolf (Canis rufus), and coyote are seasonal breeders, and dingo [4] (Canis lupus dingo) exhibits variable reproductive seasonality. Though the domestic dog serves as a critical baseline for the reproductive physiology of the taxon as a whole, investigations on species-specific differences are critical for the development of reproductive technologies for managing endangered canid populations. Comparison of reproductive seasonality, puberty, and parental behaviors across Canidae have previously been reviewed [5]. However, there are also several aspects of domestic dog reproduction that as-yet elude understanding, including the mechanisms of estrus resumption and egg maturation prior to fertilization. In this review, we summarize current knowledge of canid reproductive biology, with the emphasis on species diversity, and the application of reproductive technologies in wild canid conservation.2. Canid Reproductive BiologyThe current knowledge about reproductive biology of the canid family is mostly gleaned from the studies in the domestic dog [6,7,8]. Generally, female canids are monestrus with the reproductive cycle characterized by an extended period of proestrus and then estrus (~1 week each). The estrous period is characterized by an estrogen peak that coincides with rising circulating progesterone concentration, prior to ovulation [6,8,9,10,11,12,13]. Estrus is followed by diestrus, a luteal phase averaging 2 months in duration irrespective of pregnancy [6,14]. Diestrus is succeeded by anestrus, an extended (2–10 months) interval of ovarian quiescence [6]. The reproductive cycle of most wild canids is consistent to that of the domestic dog where diestrus is followed by an extended period of ovarian inactivity. However, peripubertal bush dogs (Speothos venaticus) housed in zoological institutions can undergo multiple estrous cycles without the anestrous period [15]. Seasonal polyestrous cycles have also been shown in captive dholes (Cuon alpinus) and New Guinea singing dog (Canis dingo hallstromi) where females enter estrus when the prior breeding attempt does not result in pregnancy [16,17,18]. While female domestic dogs and most wild canids ovulate spontaneously in the absence of a male, there is evidence that the Island fox (Urocyon littoralis) [19] and maned wolf (Chrysocyon brachyurus) [20,21] are induced ovulators and require the presence of a male for estrus and/or ovulation to occur. To date, the mechanisms of induced ovulation in these two canids have not been elucidated. However, there has been evidence suggesting that an olfactory mechanism involving volatile chemicals excreted in the urine may play roles in this phenomenon. Specifically, female maned wolves that had visual access to a male as well as the ability to contact his urine scent marks deposited on the shared fence line ovulated, but other individuals housed at the same facility with only visual contact to males failed to ovulate [20]. Analysis of volatile organic compounds excreted in maned wolf urines revealed differences in between males and females, and several compounds that were more abundant in the males have been reported as semiochemicals in other mammals [22].Like other mammals, the canid’s ovarian cycle is regulated by the hypothalamic–pituitary–gonadal axis [6,23,24,25]. Prior to proestrus, there is an increase in gonadotropin releasing hormone pulses from the hypothalamus which, in turn, stimulate follicle stimulating hormone and luteinizing hormone (LH) release from the anterior pituitary [6]. The increase in pituitary hormone pulses initiates follicular growth that stimulates gonadal steroidogenesis [26]. Continued rise in estradiol during proestrus triggers an LH surge that is followed by ovulation ~60 h later in dogs and 2 days in farmed Arctic foxes (Vulpes lagopus) [25].2.1. Reproductive SeasonalityWith the exception of the Basenji, whose annual estrus is controlled by decreasing day length [27], the domestic dog exhibits non-seasonal reproduction with males producing sperm year round while females enter estrus once or twice a year [6]. Unlike their domestic counterpart, most wild canids are seasonal breeders with the onset of a breeding season dependent on latitudinal location and/or variety of environment factors (e.g., rainfall and food availability; Table 1) [14]. The grey wolf becomes reproductively active in response to increasing day-length [14]; however, breeding date of this species shifts 22 days with each 10° increase in latitude [28]. Other North American canids, such as the red wolf [29], coyote [30], red fox (Vulpes Vulpes [31]) and island fox [19] also are reproductively active as day-length begins to increase. For South American canids, the time of breeding season varies among species. Specifically, the maned wolf breeds in the fall to early winter when day-length decreases [32]. The time of reproductive season in the crab-eating fox (Cerdocyon thous) varies depending on location [11,33]. Crab-eating foxes in Brazil breed once a year in winter and give birth in spring [11,33], whereas animals living in Argentina can reproduce twice a year when food is abundant [33]. On the opposite spectrum, the bush dog (Speothos venaticus) shows no reproductive seasonality and can breed year-round [34], and the New Guinea singing dog has a strict breeding season which does not shift with latitude [18]. Captive male wild dingos (Canis familiaris dingo) display no reproductive seasonality [35], whereas females experience distinct March–July breeding season in Central Australia, likely in response to draught periods [36].To date, factors controlling reproductive seasonality have not been fully eludicated. It has been suggested that the intervals of ovarian cycle of Basenji and wild canids are part of an endogenous circannual cycle that is influenced by photoperiod [23]. Evidence supporting this assertion is that translocation of canids (maned wolves, African wild dogs, red foxes) from the Southern to Northern hemisphere, or vice versa, shifts the breeding season by 6 months [13,14,37]. The pineal gland and its hormone, melatonin, has been indicated as playing the central role in mediating neuroendocrine responses to changing daylight in seasonal breeders [38]. Melatonin-induced advancement in reproductive season has been reported in silver foxes (Vulpes vulpes) [39], raccoon dogs (Nyctereutes procyonoides) [40], and male Arctic foxes [41], but not in female farmed Arctic foxes [41]. Furthermore, exposing domesticated silver (i.e., red fox) males to short day light from February (natural breeding season) to June (non-breeding season) can prolong spermatogenesis [42]. However, pinealectomy fails to alter reproductive seasonality in both male and female grey wolves, suggesting that other pathways or factors also play roles in regulating reproductive seasonality in wild canids [43].The linkage between reproductive seasonality and rain fall/food availability has been shown in some wild canids. For example, the Ethiopian wolf (Canis simensis) exhibits a strong association between reproduction and rainy season [44,45]. In this species, females give birth at the end of rainy season and the young become independent at the end of the dry season when prey availability is high. For the African wild dog (Lycaon pictus), strong seasonal reproduction is observed in animals living at latitudes from 7 to 25° S (e.g., Bostswana and Zimbabwe); however, dogs living at latitudes <2° (e.g., Kenya) do not exhibit a clear pattern of reproductive seasonality [46]. In seasonal breeder populations, births normally occur during the cool, dry season with no association with the variations in total rainfall [46]. Yet, only female African wild dogs exhibit strict seasonality in reproductive activity while males can produce sperm year-round, albeit with poor quality during the non-breeding season [13]. Some small canid species, such as the fennec fox (Vulpes zurda), bush dog, and crab-eating fox may exhibit variations in the number of reproductive cycle per year between animals living ex situ and in situ. Specifically, under a controlled environment in captivity, females of these canid species can cycle twice per year compared to once per year for counterparts living in the wild [12,33].Spermatogenesis occurs year-round in aseasonal canids. However, spermatogenesis only takes place during breeding season in canids that are strict seasonal breeders [14,47,48]. In canids that breed seasonally, there are also temporal variations in testosterone concentration with the hormone level beginning to rise prior to breeding season and the peak concentration coinciding with maximum sperm production [30,40,49,50,51].2.2. Gametogenesis and EmbryogenesisThe female gamete of canids exhibits several unique characteristics compared to that of other carnivore species. Canid oocytes contain a large amount of cytoplasmic lipids compared to other mammalian species [74]. Lipid yolk bodies first appear in the cytoplasm of the primary oocyte and the amount of these lipid bodies increases throughout the entire process of oogenesis, giving its dark appearance which is distinct from that of other mammalian species [75,76,77,78,79,80]. Analysis of total lipids extracted from dogs of various breeds revealed that intracellular lipids include saturated fats, triglyceride, cholesterol, phospholipids, and glycolipid, and that types and amounts of lipid fractions are consistent among individuals within and between different breeds [77]. However, the distribution of lipid bodies varies among oocytes recovered from different reproductive stages [78]. Specifically, oocytes recovered during the follicular phase displayed a diffuse pattern of lipid bodies while those retrieved during anestrus or luteal period exhibited peripheral or perinuclear distribution [78]. Analysis of chemical lipid composition revealed that phosphatidylcholines containing 34 carbons, especially phosphatidylcholine (34:1), are the most abundant phospholipids in the dog oocyte [81]. This phosphatidylcholine also has been found to be among the most abundant lipid ions in cat, human, sheep, and cattle oocytes, although variations in unique ions with specific m/z values corresponding to individual lipid species and ion abundances among species have been observed [81,82].Unlike other mammalian species, canid oocytes ovulate at an immature stage and require up to 48–72 h to completing nuclear maturation within the oviduct [79,80,83,84]. However, nuclear maturation interval can be slightly varied among species. Specifically, indirect evidence indicate that silver fox oocytes may complete maturation within 24 h as high conception rates can be observed within 24 h after ovulation in this species [85]. In the domestic dog, the oocyte remains fertile 4–5 days after nuclear maturation (i.e., 6–7 days post ovulation) [86].Like other mammalian species, canid spermatogenesis is controlled by the hypothalamic–pituitary–gonadal axis [87]. In the domestic dog, sperm can be recovered when the males reach sexual maturation at around 6–8 months of age [87,88], with optimal sperm production observed at 15–16 months old in beagles [87]. However, sperm production may not occur until the males approach 2 years old in grey wolves [14]. Sperm maturation occurs in the epididymis and the gametes acquire fertilizing ability when they reach the cauda region [89]. Studies in the domestic dog have shown that the entire process of spermatogenesis takes 62 days and sperm spend 15 days in epididymal transit [87]. Following ejaculation, dog spermatozoa can survive in the female reproductive tract for up to 7 days [86].Although in vitro studies have indicated that dog sperm can penetrate immature oocytes [90], in vivo fertilization does not occur until 83 h post-ovulation (or ~35 h after the LH surge), even in the presence of spermatozoa [84]. Information generated from an in vitro study has indicated that dog metaphase II oocytes may require an additional 12 to 24 h to fully acquire developmental competence, or 5–6 days following the LH surge [91]. Fertilization occurs at the middle or distal part of the oviduct in foxes [83] and dogs, respectively [84]. Two-pronuclei zygotes can be observed 92 h after ovulation (~6 days after the LH surge) in the dog, and between 29 and 73 h post-mating in the raccoon dog [80]. The duration during which canid early stage embryos remain within the oviduct may vary among species [80,84,85,86,92]. Specifically, domestic dog embryos remain in the oviduct until the 16-cell to morula stage, and migrate into the uterus 10–12 days post-LH surge and develop to the blastocyst stage 12–13 days post LH-surge [93]. In the farmed Arctic fox, embryos remain in the oviduct for 6–8 days after mating and enter the uterus at the morula stage [85]. However, red fox and raccoon dog [80] embryos enter the uterus at the 8–16 cell stage, 4–6 days post mating [85].3. Assisted Reproductive TechnologiesWhen a genetically valuable individual dies before successful breeding, or in situations where physical or behavioral obstacles prevent the production of offspring, it is of critical importance to rescue and preserve the genetic potential of the individual for later re-introduction into the population. Assisted reproductive technologies include, but are not limited to, monitoring of and manipulation of estrus, gamete rescue and preservation, artificial insemination (AI) and embryo production technologies such as in vitro fertilization (IVF) and intracytoplasmic sperm injection, or even cloning (i.e., somatic cell nuclear transfer), and embryo transfer. Some of these technologies have been recently reviewed in the domestic dog [94]. Here, we focus our review on recent advancements in these technologies in wild canids.3.1. Genome RescueAs previously noted, our ability to collect mature sperm and eggs from the vast majority of canid species is hampered by their strict seasonality and unique oocyte maturation system. Sperm can be collected relatively easily in non-seasonal breeders, or during breeding seasons. This has been accomplished via epididymal sperm collection following castration, electroejaculation, urethral catheterization, or, for some more tractable species, trained manual collection. Epididymal sperm collection involves the isolation of the cauda epididymis and vas deferens and either retrograde flushing or cutting into the epididymis and incubation in warmed medium to allow motile sperm to swim out. Electroejaculation, which has been the gold standard for many wild canid species, is accomplished via electrical stimulation of the prostate. Urethral catheterization, while necessitating animal anesthesia like electroejaculation, it is more simple, involving the passing of a catheter into the urethra 30 s to 1 min to allow capillary action to pull reserve sperm into the catheter.Method utilized has significant impact on both sperm quality and concentration. While not the focus of the current review, in the tractable domestic dog, sperm collected manually has been shown to maintain motility for longer than sperm collected via electroejaculation [95]. Manual collection is regularly applied in dog artificial insemination breeding efforts, and even epididymal sperm collection [96] has resulted in live births for the domestic dog following artificial insemination. Though AI success has yet to be reported, urethral catheterization has yielded high concentration and viability sperm in the domestic dog as well [97].In wild canids, electroejaculation allows for the collection of good quality and volume of semen in a variety of species (Table 2). Notably, the fluids from the prostate may also help support the functionality of the collected sperm, though it does not appear to improve post-thaw sperm metrics in the domestic dog [98]. However, urine contamination in the sample is a significant issue in electroejaculation, and the procedure requires specialized equipment, expertise, and anesthesia. Conversely, urethral catheterization has been applied successfully to collect low-volume but high-concentration and motility sperm in red wolves [99], African wild dogs [100], and maned wolves [100], with minimal urine contamination and seminal fluid. Recently comparison of both techniques was conducted for red wolves, and determined that urethral catheterization was valuable as a means to evaluate male fertility, but did not allow for the collection of sufficient sperm for future assisted reproduction technologies [99]. Further, while digital manipulation is likely to produce the best quality sperm sample overall, this method requires training of the animals, and thus is not suitable for species or individuals that are part of reintroduction efforts. Manual sperm collection has been described in farmed foxes [101] as well as racoon dogs [102], crab-eating foxes [103], and maned wolves [104].While significant variation in number and motility of sperm collected within the various techniques is apparent, this is partly a function of differences in collection time/season, animal age/health/dominance statuses, and technique (i.e., different stimulation protocols for electroejaculation). Nevertheless, no obvious patterns are evident between small, fox-like canid output and the larger bodied wolves for either of the highlighted metrics (Table 2). This may partly be due to the inverse relationship between body and testis size in mammals [105], which has previously been noted to explain the larger relative gonad weight of the crab-eating fox than that of the maned wolf [106].Thus far, epididymal sperm collection has been described only for the domestic dog [117], farmed Arctic fox [118] (note: Ultrastructural study, did not report sperm concentration or motility), and grey wolf [115]. Unsurprisingly, while this technique allows for the rescue of motile sperm, it is limited in use by animal age (post pubertal only) and breeding season status (where applicable). Motile epididymal sperm have been collected from African wild dogs and maned wolves (Table 3, unpublished data). Briefly, testes were removed during necropsy or via castration, wrapped in saline-soaked gauze, and shipped overnight on ice. The following day, the vas deferens and cauda epididymis were isolated and motile sperm allowed to swim out into warmed (37 °C) phosphate buffered saline, then evaluated for concentration and motility. Interestingly, sperm presence is not always strictly correlated with season, as maned wolf sperm have been collected in June and August, outside the September to February, normal breeding season for the Northern hemisphere. Overall, sperm counts and motility are lower than electroejaculated samples from the same species (For comparison, refer to Table 2). However, the fact that the testes were collected primarily for management purposes (i.e., from surplus young or post-reproductive males), or following animal death (i.e., from old or sick animals) may also have contributed to the reduced values compared with electroejaculated sperm from healthy males in reproductive prime. While these epididymal concentrations would not be sufficient for artificial insemination via vaginal deposition, they are potentially enough for use in in vitro fertilization or, once developed in the future, oviductal AI.On the female side, collection of immature oocytes can be done following spay or animal death, typically by mincing the outer surface of the ovary to release oocytesObtaining competent oocytes is significantly more challenging. In the Mexican grey wolf, aspiration of oocytes from post-spay ovaries during a natural breeding season or following estrus stimulation protocols have been performed [119]. To our knowledge, in situ oocyte aspiration has not been reported in canids, including the domestic dog, due to the additional challenge of the ovarian bursa, a fat pad encompassing the canid ovary and through which the oviduct runs. Oocytes can be collected surgically in dogs via laparotomy and retrograde flushing of the oviduct [120]. However, less-invasive, laparoscopic oocyte aspiration techniques utilized in other taxa (such as that which allowed for the collection of oocytes for the recent successful cheetah in vitro fertilization [121], will require significant additional fine-tuning to get through the ovarian bursa without damaging the future reproductive potential of the animal. Still, laparoscopic ovariectomies are becoming increasingly common in domestic veterinary practice [122], and continued advances in this area will support future applications to endangered canid assisted reproductive technology development.Owing to the challenge of obtaining competent gametes outside of breeding season (i.e., the majority of the year for most canid species), there is interest in growing sperm and oocytes in vitro. Ideally, even if the animal was pre-pubertal or not in breeding season, gonadal tissues can be incubated under appropriate hormonal and growth factor conditions to produce mature gametes. For example, murine neonatal testicular tissue has been cultured on an agarose gel block to produce mature sperm in vitro [123] which in turn produced live offspring. Currently, the majority of research in gonadal tissue culture has focused on the domestic dog [124,125,126,127,128,129,130,131] as a model/starting point for developing the technologies in endangered canid species. Using culture conditions developed for the dog [124], we have cultured ovarian follicles isolated from an African wild dog ovary (Figure 2, unpublished data). Specifically, small, growing ovarian follicles were mechanically isolated from surrounding ovarian tissue, then encapsulated in 3D hydrogel for structural support and incubated individually in minimum essential medium-based growth medium supplemented with 1 µg/mL porcine FSH and 100 ng/mL recombinant human Activin. Isolated early antral (diameter at isolation: >230 µm) and small antral stage (>500 µm) follicles increased in diameter and displayed development of antral cavities (a necessary development for the production of a maturation-competent oocyte, and denoted by * in Figure 2) over 14 days culture. Cavity development, as well as the increased responsiveness of early antral stage follicles over antral stages in the current culture system, is similar to what has been observed with domestic dog follicles [124].Nevertheless, ovarian and testicular tissue culture has not yet advanced to a stage where mature, fertilizable gametes can be grown from large mammalian species. In the female, this is likely due to the prolonged incubation necessary to grow large mammalian follicles from the earliest (primordial) stage through to preovulatory, which is estimated to take about 100 days in species like the dog, cow, and human compared with ~20 days in the laboratory mouse [132]. Further, while mouse preovulatory follicles reach the final diameter of ~0.5 mm [132], the ~25 µm diameter dog primordial follicles [133] grow to an average preovulatory size of 3.5 mm [134]. This over 100× expansion in size comes with significant changes in physical, hormonal, and nutrient requirements which are challenging to recapitulate in vitro. As such, recent research has focused on the development of more “biomimetic” culture systems. One primary example is the implementation of bioreactor and microfluidic chip culture methods, wherein culture medium is designed to “flow” through the tissue culture chamber to better mimic nutrient availability and waste exchange rates that the tissue would be exposed to via circulation in vivo. On the female side, a prototype microfluidic ovary-on-a-chip has been developed for the in vitro culture of domestic dog ovarian tissue and isolated follicles [127]. Dynamic culture for ovarian tissues is ongoing in the dog and other large-mammalian models to fine-tune this and other novel culture methods to support in vitro folliculogenesis.On the male side, this organ-on-a-chip technology has been used to support murine spermatogenesis in vitro for 6 months, including producing live offspring from sperm developed in the system [135]. There have been no published reports of similar technology being applied to domestic dog testicular tissue culture; however, grafting of canine testicular tissue into immunodeficient mice has been utilized to grow dog tissue under more “natural”/in vivo conditions [136]. In that study, progression of spermatogenesis was observed in tissues from immature but not adult donors, including presence of elongated spermatids after 4–8 months grafting.3.2. CryopreservationThere have been significant efforts to preserve canid sperm via cryopreservation. Techniques typically have either involved loading extended sperm with cryoprotectants into a straw prior to cooling and submersion in liquid nitrogen (so-called “closed” cryopreservation systems because cells do not come into direct contact with liquid nitrogen), or the pelleting of sperm in indentations in dry ice, followed by plunging into liquid nitrogen directly (“open” cryopreservation). There is an vast body of literature on sperm cryopreservation in the domestic dog, which have been previously reviewed [137]. More recently, research has focused on developing sperm cryopreservation protocols that do not utilize animal-based proteins, such as the gold-standard egg yolk which protects cells against cold-shock [138]. The rationale for developing egg yolk-free media is the risk of disease or microbial transmission and subsequent additional challenges with exportation regulations, as well as the biological variability inherent with the use of an animal-sourced reagent. Common egg yolk substitutes include soy lecithin [139,140,141] and polyvinyl alcohol [142], which have been comparable to yolk based extenders with regard to post-thaw survival, membrane protection and motility.Thus far, all methods employed with wild canid sperm cryopreservation have utilized egg yolk in the extender solution, but cryoprotectant solutions have varied between species. For example, glycerol was superior to dimethyl sulfoxide (DMSO) in the cryopreservation of red wolf sperm [99], yet the opposite was true for the maned wolf [110]. In the domestic dog, DMSO is toxic [143], and therefore the majority of work has utilized glycerol as a starting-point for developing sperm cryopreservation in wild canids, including for the African wild dog [107,108,144], and grey and Mexican grey wolves [111]. Though application of artificial insemination in wild canids is yet limited, cryopreserved grey wolf [145] and farmed red fox sperm [146] have been successfully used to produce live offspring, highlighting the significant potential value to genetic management in endangered canid species.Interestingly, though initial sperm survival following warming is fairly par-for-the-mammalian-course, both African wild dog [107,144] and red wolf studies [99,147] have noted rapid declines in sperm motility/viability in the first two hours of post-thaw incubation. This is not necessarily due entirely to damage resulting from the freezing process, as freshly collected red wolf sperm also experience a rapid loss in viability when maintained at near-physiological temperatures [147]. Recently, the influence of extracellular vesicles (EVs) from domestic dog oviducts on red wolf sperm post-thaw was evaluated. EVs are membrane-bound, nano- and micro-meter sized particles secreted from cells which carry protein, RNA, and DNA messages to neighboring cells. Understanding of the role(s) EVs play in modulating communication between gametes and embryos and the reproductive tract are only now being elucidated (see reviews [148,149]). The presence of domestic dog oviduct-derived EVs during thawing and incubation supported maintenance of red wolf sperm motility over ≥ 2 h incubation [150]. It was postulated that several proteins identified in the EVs with known sperm-motility and cell stress-modulating functions may have been responsible for the positive effects. While more investigations are necessary, these early results are promising for the development of an improved method to cryopreserve sperm in wild canids.Comparatively less research has been done on wild canid oocyte preservation, although progress on the development of domestic dog ova cryopreservation has been reported [151,152]. This is in no small part due to the challenge of obtaining canid oocytes as well as the ability to in vitro mature the gamete, a step that is necessary for the evaluation of the cryopreservation success. As an additional complication, the lipid-richness of canid oocytes described previously has been suggested to increase oocyte sensitivity to chilling injuries [152]. Nevertheless, there has been one report of oocyte vitrification in the Mexican grey wolf wherein oocytes were equilibrated in a 7.5% ethylene glycol and DMSO solution prior to vitrification with a CryoTop (open) system in 15% ethylene glycol, 15% DMSO, and 0.5 M sucrose [119]. Oocyte viability post-thaw, as determined via propidium iodine staining, was similar to fresh controls. Unfortunately, owing to the lack of established in vitro maturation protocols in canids, the functionality of warmed oocytes has yet to be determined.Domestic dog embryos (2–16 cell stages) vitrified in both open [153,154] and closed systems [91,155] have resulted in live births. Birth rates from these early stage embryos range from 6–36% based on embryo stage and location of transfer (i.e., uterine versus oviduct). This latter point is not trivial. Transfer of 2- to 16-cell stages to their physiologically-appropriate location (i.e., the oviduct, see section Gametogenesis and Embryogenesis) is technologically challenging, currently requiring laparotomy to access the oviduct through the dog’s bursal fat pad [156]. Standing/non-surgical transfer of embryos into the uterus has been successful in the domestic dog [153]; however, although transfer of fresh zygotes to 4-cell stage embryos to the uterine horns can produce pregnancy, the success rates are lower than those achieved from transferring advanced stage embryos (5.7 versus 23.5% for 4-cell to 8-cell embryos, per [157]). As such, there is interest in developing cryopreservation protocols for blastocysts which in turn would permit uterine transfer. Unfortunately, dog blastocysts have been shown to respond poorly to cryopreservation [153,155,158], typically displaying either abnormal morphology or reduced cell viability post-warming. In another study, dog blastocysts slow frozen with glycerol as a cryoprotectant remained intact post-thawing, whereas >80% exposed to ethylene glycol cryoprotectant ruptured [159]. To our knowledge, there have been no reports of live births from embryo cryopreservation in wild canid species; however, both an open vitrification system and programmable freezer method have been applied to Arctic fox embryos which resulted in implantation following embryo transfer [160]. Additional work is necessary to both improve blastocyst cryopreservation methods and determine the embryo cryopreservation technique(s) that will translate well to endangered canid assisted reproduction efforts.As gametes and embryos are not always available and gonadal in vitro maturation techniques are not yet developed, testicular and ovarian tissue preservation holds high potential for genome preservation in canid species (See reviews [161,162,163]). Tissue cryopreservation/vitrification is complicated by the large mass, challenging consistency of CPA exposure and cooling/warming rates within the sample, as well as the heterozygous cell populations with potential differences in cryotolerance [164]. Nevertheless, the ability to successfully bank gonadal tissue in this way, combined with tissue transplantation or future in vitro culture technologies, would allow for the re-infusion of genetics long-term.Again, most of the research in the area of gonadal tissue preservation for canids has been done in the domestic dog and taking advantage of xenografting to assess post-warming functionality of tissues. Ishijima et al. [165] evaluated the developmental potential of vitrified ovarian tissues after grafting into the ovarian bursa of immunodeficient mice and observed cell proliferation and morphologically normal ovarian follicles in grafts 4 weeks after transplantation. More recently, needle vitrification was compared directly with slow freezing [166]. The researchers found dog ovarian tissues vitrified in ethylene glycol, DMSO, and polyvinylpyrrolidone demonstrated follicle activation and progression to primary/secondary stages of development 9 weeks after grafting into immunodeficient rats [166]. For male domestic dogs, dissociated testicular cells have been vitrified in a commercial medium for stem cell culture (StemPro-34) then transplanted into chemically-castrated testes of immunodeficient mice after warming [167]. Transplanted cells formed seminiferous tubules containing donor germ cells 20 weeks after transplantation. There is one report of testicular tissue cryopreservation in grey wolves [168], wherein slow freezing (closed system) with 15% DMSO (SF-D), or a combination of 7.5% ethylene glycol and 7.5% DMSO (SF-ED), or needle immersion vitrification (open system) with 15% ethylene glycol, 15% DMSO, and 0.5M sucrose (NIV) all similarly maintained germ cell density compared with fresh (DDX4 marker, Figure 3). However, tissue architecture and levels of apoptosis post-thaw were improved in SF-ED compared with DMSO alone or NIV, indicating slow freezing with a combination of CPAs better supports viability of cryopreserved canid testicular tissues [168]. While significant work is yet needed to optimize these protocols for wild canids, ideally banked gonadal tissues can be either grafted into donors or grown in vitro to produce mature gametes in the future as a means of genetic rescue.3.3. Estrus InductionThe ability to control canid estrus cycles would allow for the precise control of estrus and/or ovulation, which is necessary for optimal timing of artificial insemination or oocyte pickup. As with earlier discussed techniques, the majority of work in this area has been done in domestic dogs (reviewed in [169,170,171]). A variety of approaches to induce canine estrus have been reported, including administration of exogenous gonadotropins, dopamine agonists, and gonadotropin releasing hormone (GnRH) and its analogs. The latter approach works via acting on pituitary to release follicle stimulating hormone and luteinizing hormone, which promotes ovarian follicle development. Interestingly, prolonged exposure results in reproductive suppression and as such has been applied as a reversible contraceptive [172]. In short exposures (at least 7 days [173]) to GnRH agonists, domestic dogs in anestrus can display high rates of ovulation (up to 100% [174]). However, if treated during diestrus [174] or early anestrous [175], both ovulation rates and pregnancy success are significantly reduced compared with mid- or late-anestrus treatment.The GnRH agonist deslorelin has also been used to induce estrus in several wild canid species during their natural breeding seasons. In grey wolves, estrus induction was achieved in at least 3/5 animals, and subsequently resulted in pregnancies [176]. In the coyotes, deslorelin implanted females display increased steroid hormone levels and have achieved successful pregnancies via artificial insemination with grey wolf sperm [177]. In the maned wolf, deslorelin implants stimulated ovarian activity, but ovulation only occurred in females paired with males or treated with an injection of luteinizing hormone to induce ovulation [20]. Deslorelin-treated Mexican grey wolves also had increased numbers of oocytes collected from their ovaries compared to those in natural cycles [119]. Importantly, administration of deslorelin to non-breeding season (i.e., mid-late anestrus, in October) coyotes resulted in transient elevations in steroid hormone concentrations, reflecting an ovarian response, and the display of sexual behaviors [178]. Such a technique could be valuable to obtain additional breeding opportunities in strictly seasonal species. While no pregnancies were achieved, copulatory ties were observed in two of the six coyote pairs. Though these data suggest additional, out-of-season breeding opportunities might be made possible in managed canids, treated females subsequently displayed suppressed reproductive behaviors during the following breeding season. In the study this behavior change did not significantly impact reproductive success, but it was noted that such shifts could have long-standing impacts on social hierarchy. In sum, while studies in the dog and wild canid demonstrate that estrus induction is possible, improved understanding of the physiology of anestrus is needed to improve on existing techniques. Further, the behavioral implications of out-of-season or significantly shifted estrous cycles must be considered in animal management and breeding efforts.3.4. Artificial InseminationWhile artificial insemination (AI) is commonly applied to domestic dog and fox breeding, there have been only a handful of successful reports of live offspring from AI in wild canids (see reviews [48,179]). For example, Asa et al. [176] reported the birth of grey wolf pups following estrus induction via deslorelin implant and deposition of 750 × 106 sperm with high motility in the cauda vagina. As previously noted, this concentration may be feasibly obtained by electroejaculation in many canid species, but would not likely be possible utilizing sperm collected via urethral catheterization unless multiple collections were combined. Evaluation of frozen–thawed sperm concentrations necessary for intrauterine AI has previously been performed in farmed foxes [101]. The authors concluded that, taking into account sperm loss due to cryodamage and reduced fertilizing capacity, the minimum number of sperm necessary to achieve normal pregnancy rates and litter sizes is at least double what would be needed if using fresh sperm. This previous work in the dog and fox have laid the groundwork for successful implementation of the sperm-deposition technique, but the primary challenge in adapting AI for additional non-domesticated canid species is the need for intensive estrus monitoring and/or consistent estrus induction protocols. However, with increased sperm collection and banking efforts across Canidae, reports of additional success with this technique in other species is likely forthcoming.3.5. In Vitro Oocyte Maturation and FertilizationAs noted earlier, the mechanism of canid oocyte delayed maturation in vivo is not yet fully understood, and as such it has been challenging to recapitulate the process in vitro. In vitro maturation involves the incubation of groups of oocytes in maturation media (typically TCM199) containing a protein source (e.g., bovine serum albumin), follicle stimulating hormone and luteinizing hormone or eCG/hCG, with incubation at 37–38 °C and 5% CO2 for 48–72 h. Rates of meiosis resumption and oocytes reaching Metaphase II varies depending on size of the source ovarian follicle [180] and animal age [181], but not reproductive stage of the donor [182]. Unfortunately, a protocol for producing consistently high rates of mature oocytes from the domestic dog has not been achieved. Still, embryos have been produced from in vitro matured oocytes in the domestic dog [183,184,185,186], and live offspring have been produced via IVF with in vivo matured oocytes [82,151]. In the latter, in vivo matured oocytes were collected from the oviducts of domestic dogs 6 days after the LH surge, and fertilized with 100,000 sperm that had been incubated under capacitation-stimulating conditions for ~3 h [82]. As yet, intracytoplasmic sperm injection has not produced offspring in the domestic dog.Though obtaining in vivo matured oocytes in endangered canids is not yet technically feasible, and development of consistent in vitro maturation methods are still ongoing, translation of this technology to wild canid conservation efforts has been initiated in sperm capacitation studies. Recently, the in vitro survival and function of red wolf sperm in several different gamete handling medium was assessed [187]. It was found that the media previously utilized for domestic dog sperm capacitation and in vitro fertilization both supported red wolf sperm survival and early capacitation during extended incubation at physiological temperatures. This is promising for translating the success of IVF in the domestic dog to endangered canids, and also for improving in vitro incubation methods for wild canids in general, which would aid in the evaluation of handling and cryopreservation methods as they are being developed.3.6. CloningCloning, or somatic cell nuclear transfer, involves the transfer of somatic cell nuclei into enucleated oocytes and then activation to resume development. While there are some commercial entities using this technique to clone domestic dogs, efficiency of producing cloned embryos and live offspring via cloning is ~4% [153]. Grey wolves have been cloned using fibroblasts (collected and cultured from both donor ear cells and post-mortem abdominal skin) and enucleated, in vivo matured domestic dog oocytes [154,155]. Still, this technology overall is likely a long way from being applied to most endangered canid species’ conservation efforts. This technique would primarily be valuable in re-infusing genetics from founder individuals into the population, which is only feasible if properly preserved somatic cells exist from these animals and one can obtain mature canid oocytes. Some biobanks, like the Frozen Zoo at the San Diego Zoo institute for Conservation are poised to take advantage of somatic cell nuclear transfer and induced pluripotent stem cell technologies [156]. Further, efforts to improve in vitro maturation success in canid oocytes, which would support the advancement of many of the assisted reproductive technologies described here, are ongoing.4. ConclusionsThere is an incredibly variety of reproductive strategies within the Canidae family, from mechanisms of breeding season initiation to ovulation induction, despite the majority of members being under-studied. Though assisted reproductive technologies have advanced for the domestic dog and select wild canid species in recent years, understanding of each species’ unique traits will be necessary to implement techniques for population management in the future.

animals : an open access journal from mdpi

10.3390/ani11051355

PMC8150830

According to European Food Safety Agency (EFSA), human campylobacteriosis is the most commonly diagnosed zoonotic disease in the EU. In 2018, the Member States reported 246,571 cases (30% increase since 2015). For years, poultry meat and poultry products have been considered a main source for human infections. In 2018, the highest occurrence of Campylobacter spp. was detected in chicken (37.5%) and turkey meat (28.2%). Considering this situation, there has been ongoing discussion regarding the potential strategies to minimize the level of Campylobacter spp. colonization in poultry and therefore in humans. In 2018, EFSA Panel on Biological Hazards indicated that use of feed and water additives is the second most likely strategy that can be successful in minimizing Campylobacter spp. colonization rate in broiler chickens. One of these water and feed additives are probiotics—living microorganisms which, when supplemented in the right dose, have a positive effect on microbial ecosystem of the host gut by ensuring a favorable balance between commensal and pathogenic microflora. In this review paper, the authors present current results of the studies concerning the potential use of probiotics as a preventive measure of Campylobacter spp. infection, under laboratory conditions and at a chicken farm level.

Campylobacter spp. are widely distributed microorganisms, many of which are commensals of gastrointestinal tract in multiple animal species, including poultry. Most commonly detected are C. jejuni and C. coli. Although infections are usually asymptomatic in poultry, poultry meat and products represent main sources of infection with these bacteria to humans. According to recent EFSA report, campylobacteriosis is the most commonly reported zoonotic disease. In 2018, EFSA Panel on Biological Hazards indicated that use of feed and water additives is the second most likely strategy that can be successful in minimizing Campylobacter spp. colonization rate in broiler chickens. One of those feed and water additives are probiotics. From numerous research papers it can be concluded that probiotics exhibit plenty of mechanisms of anti-Campylobacter activity, which were evaluated under in vitro conditions. These results, to some extent, can explain the efficacy of probiotics in in vivo studies, although different outcome can be observed under these two laboratory conditions. Probiotics are capable of reducing Campylobacter spp. population count in poultry gastrointestinal tract and they can reduce carcass contamination. Potential modes of anti-Campylobacter activity of probiotics, results of in vivo studies and studies performed at a farm level are widely discussed in the paper.

1. IntroductionBacteria from the genus Campylobacter spp. are Gram-negative, widely distributed microorganisms, many of which are detected as commensals of the gastrointestinal tract (GIT) in multiple animal species, including poultry, domestic, and wild birds. Campylobacter spp., of which the two most common species are C. jejuni and C. coli, are thermophilic and microaerophilic bacteria that find favorable conditions for colonization in the GIT of all warm-blooded animals and birds, especially the latter due to their higher body temperature (approximately 41 °C) than other animal species [1].Since the isolation of the genus Campylobacter from Vibrio spp. in 1963, Campylobacter spp. [2] infections have become the most important cause of foodborne bacterial gastroenteritis in humans in many developed countries [3]. For years, there have been an ongoing discussion and number of studies were performed in order to establish and evaluate potential strategies to overcome these issues. In this review paper, we have paid special attention to the potential of probiotics, which together with other food and water additives were jointly classified as the second most likely strategy capable of minimizing Campylobacter spp. prevalence in poultry by the European Food Safety Agency (EFSA) Panel on Biological Hazards [3].2. Campylobacter spp. in Poultry—Colonization, Carcass Contamination, and PrevalenceCampylobacter spp. colonizations are more common in domestic than in free-living birds [4]. Although these bacteria do not pose a serious threat to birds per se, or to large-scale poultry production in general, colonizations with them are extremely important because poultry products represent the main source of Campylobacter bacteria in humans, and also because human campylobacteriosis has for many years been the most frequently diagnosed zoonotic enteropathy, surpassing even salmonellosis [3,5].As thermotolerant microorganisms, Campylobacter spp., find favorable conditions for colonization in the gastrointestinal tract of birds, which have a higher body temperature than other animal species. The predominant species detected in birds include C. jejuni and C. coli, the latter having mainly been isolated from turkeys, but C. lari bacilli have also been sporadically isolated from chickens [6,7,8,9,10].The colonization of the GIT of poultry with Campylobacter spp. usually proceeds without any distinct clinical manifestations and affects the small intestine, the cecum, and the cloaca. The consequences of colonization have been demonstrated to depend on the chicken breed and age. Humphrey et al. [11] have shown that the varying extents of the appearance of clinical symptoms in various chicken breeds can be ascribed to differences from breed to breed in the immune system responsiveness to infection with Campylobacter spp. They have also demonstrated that in some chicken breeds, the prolonged inflammatory reaction in the intestinal mucosa is due to a lack of interleukin (IL) 10 production, which, in turn, leads to diarrhea. Some works have also indicated that a clinical course of campylobacteriosis is most often noted in young birds [12,13] and manifests with enteritis and watery diarrhea, sometimes with mucus and blood in the excreta. This condition can lead to poorer body weight gains, reduced feed conversion ratio, and ultimately to differences in bird growth within the flock. Similar symptoms including diarrhea and lower body weight gains were observed in young turkeys [5]. In addition, cases of campylobacteriosis reported in flocks of laying hens weakened laying performance and egg hatchability [14].The incubation period of campylobacteriosis in the cases with diarrhea was reported to range from 2 to 5 days [14]. The colonization of the intestines by these bacteria primarily takes place in the jejunum, then in the ileum, and finally in the cecum [15], where their population number peaks [16], and they can be detected and excreted in feces for a prolonged period [1]. Campylobacter spp. reaches higher concentration in distal parts of avian GIT [1]. For instance, the concentration of bacteria in the crop was significantly higher than in the gizzard [17], which results from growth pH requirements.Even though Campylobacter spp. can be detected in the liver and other internal organs, deep muscles, and blood of infected birds [18], it is believed that the majority of incidences of contamination with these bacteria found in bird carcasses result from the contact of live birds or carcasses during slaughter with a contaminated external environment in a slaughterhouse or on the production farm. Hue et al. [6] demonstrated that the level of carcass contamination with Campylobacter spp. correlated directly with the degree of intestine invasion by these bacteria. In addition, Berrang et al. [19] demonstrated that the number of Campylobacter spp.-positive breast skin sponge samples increased after bird defeathering during slaughter from 1 (prior to defeathering) to 120 out of the 120 tested. Additionally, after defeathering, the Campylobacter spp. population count reached log10 4.2 colony forming units (CFU) per sample. The same authors recorded an increase of Campylobacter spp.-positive breast skin samples from 0 to 13 out of 120 tested samples from an experimental group of slaughtered birds, the cloacae of half of which were plugged with tampons and sutured closed. According to these authors, an increase in the recovery of Campylobacter after defeathering can be related to the escape of contaminated feces from the cloaca during the process. Other risk factors increasing the likelihood of poultry carcass contamination include cross-contamination during transport, scalding, plucking, evisceration, and chilling operations [20,21,22]. Moreover, Campylobacter can survive on the surface of equipment used for bird slaughter despite cleaning and sanitizing, and the persistence of the bacteria can contribute to cross-contamination of carcasses during the slaughter process [23].It has been demonstrated, that only 35 Campylobacter spp. CFU are sufficient to establish colonization in the bird gastrointestinal tract [24]. The transmission rate of Campylobacter jejuni was estimated to be 2.37 ± 0.295 infections per infectious bird per day. It means that in a flock consisting of 20,000 broilers, 95% of birds will become infected within 4.4–7.2 days after colonization of the first broiler [24]. The same study showed that the mean age at which birds become infected with Campylobacter spp. was 21 days of life. Based on selected papers published after the year 2000, it can be concluded that Campylobacter spp. prevalence in poultry flocks ranged from 3.5% to 71.5% [25,26]. Besides the immediate threat to consumers, such a widespread prevalence of Campylobacter bacteria in the poultry population poses an additional risk. Namely, that in a given location where large-scale production takes place, the constant presence of these microorganisms and the widespread use of chemotherapeutic agents facilitates the selection of Campylobacter spp. strains resistant to antimicrobials, which, unfortunately, translates to the results returned in monitoring studies [27,28]. For example, Woźniak and Wieliczko [29,30] showed an increase in the percentage of enrofloxacin-resistant Campylobacter strains isolated from poultry in Poland from 52.1% in 1994 to 93.6% in 2008. A similar trend was noted by these authors regarding resistance to tetracyclines. An additional disturbing aspect is the emergence at the beginning of the 21st century of multi-resistant Campylobacter strains that were not found in the 1990s [29,30]. These bacteria can pose a direct risk to consumers since these strains could be passed to humans via the food chain or by direct contact and they can additionally act as donors of antibiotic resistance genes to other bacteria.In the light of the above and taking into account the EU policy to reduce the antibiotic usage in animal husbandry, it is worth emphasizing, that the use of antibiotics is not considered as a preventive option for Campylobacter spp. infections in poultry [3].3. Transmission and Prevention of Campylobacter spp. in PoultryVertical route of infection with Campylobacter spp. in poultry is recognized not to be of great importance to Campylobacter spp. prevalence, despite these bacteria having been isolated from different parts of the reproductive system of laying hens [31,32,33]. This hypothesis is confirmed by studies which corroborated Campylobacter spp. genotype differences in the offspring of Campylobacter-positive breeder flocks [34,35].Infection of birds with Campylobacter spp. occurs mainly horizontally through the gastrointestinal tract. The main sources and factors increasing the risk of infection in poultry include contaminated litter, rodents, flies, farm staff, other farm animals kept on or near the production farm, inadequately long production breaks, insufficient washing and disinfection of facilities, contamination of water and surrounding lands, proximity of Campylobacter-positive flocks, and thinning [3,36]. It is worth noting that in many cases, bacteria of the genus Campylobacter are present in livestock facilities even before birds are settled there. They were detected in dust and drinking water in poultry houses, which had been washed and disinfected immediately before the delivery of the chicks for rearing [36].The findings from farm practice and location indicate the capacity of biosecurity measures to reduce the frequency and degree of Campylobacter spp. infection in poultry. However, taking into account the multitude of underlying factors of infection, it is reiterated increasingly often that these precautions should be comprehensive and multidimensional to raise the probability of achieving their goal. According to the recent report of the EFSA Panel on Biological Hazards regarding the control options for Campylobacter in broilers in primary production, the strategies most likely to be successful in minimizing the rate of infection and the prevalence of Campylobacter spp. in poultry products include vaccination, use of feed and water additives, discontinuation of thinning, employment of only a few and only well-trained staff, elimination of drinkers that allow standing water, addition of disinfectants to drinking water, hygienic anterooms, and designation of one set of tools per broiler house [3].4. Campylobacteriosis Cases in HumansAn EFSA report states that the number of registered and confirmed European cases of human infection with Campylobacter bacteria in 2015 exceeded 190,000 [5]. Additionally, EFSA assumed that the approximate number of actual cases of infection could be as high as 9 million (because only 1 in 45 cases is confirmed in laboratory testing) [5]. The costs incurred in Europe associated with the decline in livestock productivity due to infection with Campylobacter bacteria together with the costs of treating infections with these bacteria in humans, have been estimated at approximately 2.4 billion EUR per year [5]. Additionally, the number of confirmed cases of infection with Campylobacter genus bacteria is increasing every year. According to the latest report by the EFSA and the European Centre for Disease Prevention and Control (ECDC), campylobacteriosis is the most commonly reported zoonotic disease in the EU. In 2018, the Member States reported 246,571 cases (an approximate 30% increase over the number in the 2015 report) [3].The widespread prevalence of Campylobacter spp. in the animal population carries the risk of contamination of food products such as raw meat, milk, and water. The available literature data show that poultry and poultry meat are considered to be a common source of Campylobacter spp. infections; however, beef and pork products are also emphasized to contribute to the unfavorable epidemiological situation of campylobacteriosis in the human population. Considering the 2015 data on campylobacteriosis [5], broiler chickens and products made of them accounted for nearly 22.5%, eggs and egg products for 6.12%, but milk or cattle and beef products for less than 4.1% of cases each [5]. In the 2018 EFSA report, the highest occurrence of Campylobacter spp. was detected in chicken meat (37.5%) and turkey meat (28.2%) [3].5. Benefits from Using Probiotics in PoultryProbiotics are living microorganisms which, when supplemented in the right dose, have a positive effect on the microbial ecosystem of the host gut by ensuring a favorable balance between commensal and pathogenic microflora [37].The bacteria most often used as probiotics include those from the following genera: Bacillus, Bifidobacterium, Enterococcus, Lactobacillus, Pediococcus, and Streptococcus. Probiotics are also produced from selected species of fungi and yeasts (e.g., Saccharomyces).The main beneficial effects of probiotics relate primarily to their raising of feed digestibility and bioavailability, stimulation of the immune system, improvement of health, and provision of superior organoleptic properties and chemical composition of carcasses [38,39,40,41,42,43,44,45,46].One of the first reports on the beneficial effects of probiotics in poultry comes from 1973, when Tortuero [47] noticed an improvement in weight gain coinciding with the use of L. acidophilus in chicks during the first 5 days of life. In addition, in this experiment, the group receiving the probiotic was characterized by Lactobacillus dominance among the gastrointestinal microflora and by the simultaneous reduction in enterococci population. Considering the enterococci population in the probiotic-treated group, Tortuero [47] obtained a result similar to that in the group of birds receiving the antibiotic instead of the Lactobacillus culture in the same period. In addition, Nurmi and Rantal [48] showed that protection against Salmonella infantis could be obtained in broiler chicks by the per os administration of bacterial flora isolated from the intestines of healthy adult chickens. This concept was later referred to as competitive exclusion.Probiotic bacteria are able to inhibit the growth of pathogenic microflora in the gastrointestinal tract of birds. This is due to the depletion of nutrients in the environment, the blocking of target receptors for pathogens on the surface of epithelial cells, or the production of natural antibacterial agents known as bacteriocins [49]. Probiotic bacteria also exhibit strong immunomodulatory effects, improving the local immune mechanisms in the gastrointestinal tract. For example, their regular and occasional uses in poultry have been shown to have an immunostimulating effect on interferon production; activities of macrophages, heterophiles, lymphocytes, and natural killer (NK) cells; and the production of specific antibodies [38,39,41,42,44]. In addition, it was previously concluded that probiotics exert a non-specific effect on the stimulation of the gut-associated lymphoid tissue (GALT), but as antigens with relatively low immunogenicity, they do not contribute to the excessive development of the inflammatory reaction nor activate the immunological mechanisms aiming at their complete elimination [43]. Through these properties, probiotics enhance the responsiveness of the immune system to an infecting pathogen [43]. These phenomena induced by probiotics minimize the risk of colonization or limit the population size of a wide range of microorganisms potentially pathogenic to poultry [7,10,13,41,48,50,51,52,53,54].6. Probiotics and Poultry Campylobacter spp. Infection and ColonizationVarious systems are used to assess probiotic efficacy in minimizing the consequences of infections with Campylobacter spp., and they have been previously reviewed [10,37,55]. In this review article, we would like to present the general scope of probiotics’ modes of action against Campylobacter from molecular, in vitro, and in vivo studies and in conclude the work to present the results obtained in field experiments performed under commercial broiler farm conditions. General modes of probiotic anti-Campylobacter activity are summarized in Table 1.After entering the avian GIT, Campylobacter spp. use different mechanisms to establish their population in the gut environment, like motility, chemotaxis, adhesion, intracellular infection, and the capability to synthesize entero- and cytotoxins, as reviewed by Mohan [10]. Given the set of mechanisms, the bacteria can exploit, most research works addressing the choice of potential probiotic bacteria have been focused on the evaluation of their potential to overcome these properties of Campylobacter spp.7. In Vitro StudiesMost of the experimental models applicable to the in vitro assessment of probiotic efficacy refer to their ability to inhibit the proliferation potential of Campylobacter spp. in co-cultured assays or their colonization potential in cell cultures.Contemporary scientific research has in the most past assumed the use of lactic acid bacteria (LAB; particularly various selected strains of Lactobacillus genus) as probiotics against Campylobacter jejuni infection. From a number of pertinent studies, it can be concluded that LAB alone, or in the mixture with different carbohydrate prebiotics, can decrease the C. jejuni growth rate by 4–8 log10 CFU/mL after 24–72 h of co-culturing [65,66]. The main mechanism of the inhibitory effect of LAB against Campylobacter growth has been described in those studies as acidification of the medium by producing lactic and acetic acids [56,57,58,62]. In an assortment of studies, the strongest antagonism towards Campylobacter was exhibited by L. salivarius and L. reuteri. It has also been confirmed in a study by Dec et al. [56], who evaluated the probiotic potential of 46 Lactobacillus isolates from chicken feces or cloacae against C. jejuni and C. coli. Singled out again in this research, L. salivarius and L. reuteri evinced the highest anti-Campylobacter activity, and results indicated that it was the reduced pH of the supernatant from Lactobacilli culture that played the key role in inhibiting Campylobacter growth. The cited authors also highlighted that all the isolated Lactobacillus strains were capable of producing hydrogen peroxidase, but the production rate did not correlate with the level of anti-Campylobacter properties [56], which has been also demonstrated previously by Campana et al. [67], and it is a result of catalase production by these Campylobacter species [67].Both acidic and neutralized lactobacilli supernatants have been shown to inhibit C. jejuni growth to comparable levels. This finding suggests that other bioactive factors could also contribute to pathogenic bacteria growth inhibition. Messaoudi et al. [60] described L. salivarius isolates that were capable of producing bacteriocins and exhibited high anti-Campylobacter activity. Bacteriocins are small proteins produced by bacteria including LAB, which enable them to inhibit the growth of other bacteria in the environment. The inhibitory effect of bacteriocins against Campylobacter spp. growth was confirmed in both in vitro and in vivo studies [61]. Bearing this in mind, it seems that bacteriocin production and acidification can play equally important roles in inhibiting the growth of pathogenic bacteria.It has been demonstrated that the mixture of bacterial stains of the Lactobacillus genus that have been individually confirmed to inhibit the growth of Campylobacter in in vitro studies may not display additive properties under the same conditions, which has been suggested in the in vivo studies. A recent study by Taha-Abdelazi et al. [62] evaluated the ability of five different Lactobacillus strains (L. salivarius, L. johnsonii, L. reuteri, L. crispatus, and L. gasseri) to inhibit Campylobacter jejuni growth. In this study, the authors confirmed the efficacy of all strains and reported the highest inhibitory activity for L. salivarius. However, its activity was comparable to the activity of a mixture of all five strains tested. This suggests some limitations in the possibility of direct translation of in vitro study results into in vivo study findings or may indicate that probiotics used in a mixture can be antagonistic to each other.Another step toward identifying the mechanisms of probiotic action against Campylobacter is made via investigations exploring changes in the molecular properties of pathogenic bacteria and their capability of adhering to or invading the target cells. However, the applicability of these investigations to poultry campylobacteriosis is less than perfect because of their use of human and not avian cell lines, as the latter are commercially unavailable [55]. This limitation coupled with the only selectively aspectual observation of the pathogen–host interaction that is possible can hinder extrapolation of their results into likely results under farm conditions.However, it has been demonstrated that L. salivarius, L. johnsonii, L. crispatus, and L. gasseri significantly reduced the expression of virulence-related genes in C. jejuni after 24 h of co-incubation with probiotic bacteria. This down-regulation involved C. jejuni genes responsible for both motility (flaA, flaB, and flhA) and invasion (ciaB), which correlated with less extensive invasion by C. jejuni of Caco-2 human intestinal epithelial cell line by over 80% [62]. Using the same cell line, it has been demonstrated that L. acidophilus was capable of reducing both adhesion of 10 various human-origin C. jejuni strains to the intestinal epithelial cells and these cells’ invasion by the bacteria [61]. In this experiment, the authors used different schemes of probiotic treatment of Caco-2 cells, which were referred to as exclusion (probiotic treatment prior to infection), competition (probiotic treatment and infection performed simultaneously), and displacement (probiotic treatment after the infection). In all three strategies, the authors reported diminished adhesion and invasion of most of the C. jejuni strains used, but the most prominent results were achieved in the displacement test where the adhesion was reduced by 11–53% and the invasion by 11–52% [67]. These results could be explained by competitive exclusion where the LAB block the adhesion sites on epithelial cells for C. jejuni and/or by probable bacteriocin production by the L. acidophilus probiotic.The adhesion of probiotic bacteria to epithelial cells as one of their mechanisms of anti-Campylobacter activity has been suggested previously [10]. Properties of probiotic bacteria that impart their adherent capacity include their high hydrophobicity as well as specific binding to cells by surface adhesins. Wang et al. [68] demonstrated that the high hydrophobicity of L. casei strain ZL-4 correlated with high inhibiting activity against the adhesion of C. jejuni to intestinal cells of culture HT-29 and the invasion of these cells by the bacteria. In another study, Dec et al. [56] demonstrated that approximately 98% (45 out of 46 tested) of different Lactobacilli strains of chicken origin that possessed activity against C. jejuni and C. coli of chicken origin displayed very high hydrophobicity.Recently, in parallel to works investigating the effect of probiotics on the growth and properties of Campylobacter spp., probiotic bacteria have also been tested for their immunomodulatory and immunostimulatory properties. Despite the studies conducted so far on the specificity of poultry immune system response to infection with Campylobacter spp. being few, their results enable a much better perception of the potential efficacy of the tested probiotics under field conditions. In a recent study performed by Taha-Abdelazi et al. [62], the stimulation of macrophages with either a single species or a species mixture of heat-killed lactobacilli (L. salivarius, L. johnsonii, L. reuteri, L. crispatus, and L. gasseri) enhanced C. jejuni phagocytosis. At the same time, macrophages exposed to lactobacilli had increased expression of interferon-γ, IL-1β, IL-12p40, and IL-10 genes. Furthermore, L. salivarius, L. reuteri, L. crispatus, and lactobacilli mixture increased the expression of the co-stimulatory CD40, CD80, and CD86 molecules in macrophages [62]. Additionally, it has been shown that B lymphocytes are involved in C. jejuni clearance and decrease the shedding of the bacteria [15]. Co-stimulatory molecules are known to participate in the cascade of antigenic signal transduction and activation of both T and B cells. Therefore, it may be speculated that the LAB mixture can enhance both the non-specific and specific immune responses against Campylobacter.8. In Vivo StudiesAt this stage, it can be summarized that probiotic bacteria possess and display, under in vitro conditions, multiple mechanisms of anti-Campylobacter activity. In vivo tests allow for a more comprehensive evaluation of probiotics and their efficacy in the highly variable and complex environment of the chicken GIT. These tests can be divided into those in which the efficacy of probiotics is assessed against natural Campylobacter spp. infection and those assaying how they act against artificial infection.The choice of probiotic bacteria for experimental and commercial needs is governed by strict criteria of safety, functionality, and technological feasibility. One of the basic commercial criteria is that they should have confirmed efficacy in the target animal species [68]. Although most often the primary selection and determination of the basic properties of probiotic strains are based on in vitro studies, most of the criteria described above can only be assessed using in vivo tests. It is worth mentioning here that not always can an effect in one research strategy be directly replicated in another. Robyn et al. [69,70] demonstrated that the Enterococcus faecalis strain MB 5259 that was capable of inhibiting C. jejuni growth in in vitro tests was unable to do so in in vivo experiments, irrespective of the inoculum volume. On the other hand, in these studies E. faecalis was capable of colonizing the chicken GIT despite its impairment during passage through the tract which the in vitro analysis showed [69,70].Fritts et al. [63] demonstrated that the Calsporin® probiotic comprising Bacillus subtilis (strain C-3102) administered to birds in feed from the day of hatching to slaughter (at 42 days) was capable of reducing the extent of broiler carcass Campylobacter contamination in the course of natural infection. During the experiment, the probiotic-treated and control birds were kept in isolated pens and separated from each other in a way that prevented cross-contamination with probiotic bacteria. In two experiments, the authors recorded a 6.5% reduction in mean Campylobacter spp. CFU/mL after probiotic administration. In the same study, the population counts of Salmonella and E. coli decreased in carcass samples of the probiotic-treated group of chickens. The results of those studies were also confirmed for the ceca of broilers, as Guyard-Nicodème et al. [71] demonstrated a decrease in C. jejuni count in birds given Calsporin® after artificial C. jejuni infection. On the other hand, Garcia-Hernández et al. [72] did not record any reduction of the C. jejuni population count in the ceca of chickens artificially infected at 14 days of life after a similar treatment with probiotic B. subtilis of the DSM17299 strain.Saint-Cyr et al. [49] demonstrated that L. salivarius of the SMXD51 strain given to birds on the first day of life and afterwards every 2–3 days was capable of reducing the C. jejuni count after artificial infection at 11 days of life. The bacterial populations in the cecal content on the 14th and 35th days of the birds’ life declined by 0.82 and 2.81 log10, respectively. This was also confirmed by Ghareeb et al. [13], who treated the birds with a multispecies probiotic comprising Enterococcus faecium, Pediococcus acidilactici, Bifidobacterium animalis, L. salivarius, and L. reuteri at doses of 2 or 20 mg/bird via drinking water from the first day of life and noted up to 5.81 and 5.85 log10 reduction of C. jejuni CFU/g in the cecal content after the birds were artificially infected on the day of hatching. These data may suggest that a combined preparation may display higher anti-Campylobacter activity than single strain probiotics.The works by Saint-Cyr et al. [49] and Ghareeb et al. [13] cited above are only selected examples confirming the anti-Campylobacter activity of probiotics, while this efficacy has also been confirmed in many other scientific studies [13,54,73,74]. In contrast, studies have also reported the inefficacy of probiotics in this context [69,75,76]. Nevertheless, as in the case of in vitro studies, distinctly different results may also be obtained in in vivo studies evaluating the benefits of probiotics against poultry campylobacteriosis under controlled experimental conditions. On the other hand, it is known that the results of in vitro tests will not always translate into comparable results under in vivo conditions [69,70]. This may be due to differences in the bacterial strain used as a probiotic, the final composition of the probiotic, its dose and application pattern, the Campylobacter strain used in the experiment, experimental conditions, the age and breed of the birds, and therefore different outcomes of probiotic activity. Additionally, it is worth emphasizing that in vitro studies do not take into account the complexity and variability of birds GIT environment as well as the interactions between probiotic, Campylobacter spp., and GALT that may occur in vivo.9. On Farm StudiesIn the case of experiments carried out under commercial poultry farms conditions, an additional set of variables that were not included in laboratory experiments should be taken into account and they include, i.a.: differences in farm infrastructure; breeding practices; epidemiological status of the farm/region; biosecurity regime; season of the year; quality, health, and immunological status of chicks; preventive vaccination program; antibiotic treatment administered to birds; and the use of other water and feed additives, which may directly affect either the probiotic or the Campylobacter spp. population and others. Therefore, it is not surprising that currently there are scarce research works describing field experiments on the anti-Campylobacter activity of probiotics in poultry. Although, Śmiałek et al. [64] demonstrated that a multispecies probiotic comprising an LAB mixture (L. lactis, Carnobacterium divergens, L. casei, and L. plantarum) and Saccharomyces cerevisiae given to birds in feed for the entire production cycle was capable of reducing the Campylobacter spp. population count in broiler ceca and feces. In this case, the Campylobacter spp. count was approximately 10 times lower in experimental chickens than in the control chickens. The authors also recorded no growth of Campylobacter spp. from samples of pectoral muscles and overlying skin after slaughter of the probiotic-treated birds, while 50% of the control bird sample cultures were positive, with a mean Campylobacter spp. count of 4.15 × 101 CFU/g. Those researchers compared the results of this experiment to the results of control production cycles in which the population number of Campylobacter spp. from equivalent samples from the same farm and the same chicken houses was monitored. The production cycle prior to the experiment and two production cycles after the experiment were used as these controls. In all of them, the Campylobacter spp. population count was significantly higher in the chicken house in which the probiotic was used during the main experiment. The pre-experiment production cycle results allowed the most appropriate chicken house to be selected in which the probiotic was used and the full set of determined dependencies enabled the authors to conclude that the probiotic used as a feed additive in the study was capable of reducing the extent of Campylobacter spp. invasion in the GIT and the contamination level in the birds’ environment, which subsequently contributed to the improved hygienic parameters of the analyzed poultry carcasses [64].10. Final ConclusionsThe use of probiotics is one of the strategies that can be implemented in order to minimize the risk of infection and the level of colonization with Campylobacter spp. in GIT of chickens and humans. It has been demonstrated that numerous potential probiotic strains, or probiotic strains mixture, are capable of inhibiting Campylobacter spp. growth in in vitro studies. In the nearest future, a special attention should be paid to in vivo studies. The results of those studies would enable the evaluation of in vitro studies results, evaluation of probiotics modes of action as well as the verification of target probiotic composition, which in turn would result in optimization of probiotic formula in order to establish commercial products with the highest anti-Campylobacter activity in poultry.Additionally, it is worth to emphasize the lack of knowledge into the immunomodulatory properties of probiotics, in the context of both: general immunity enhancement as well as the specific anti-Campylobacter immunity stimulation.Considering the list-topping position of poultry and poultry products for many years as potential sources of Campylobacter spp. to humans, it should be expected that in the near future, adequate administrative programs to alleviate the peril in this situation will be implemented. Considering the above it is necessary to conduct appropriate studies under farm conditions in order to evaluate the efficacy of probiotics in minimizing the risk of Campylobacter spp. infection and infection rate. In these studies, probiotics should be considered as one of the variable elements of multidimensional biosecurity program.

animals : an open access journal from mdpi

10.3390/ani13071211

PMC10093065

Since there has been an increase in the frequency of extreme weather events (i.e., heavy rains, heat stress) due to climate change, the interaction between feeding and management issues and the required facilities to alleviate environmental effects on animal performance has become relevant. Although there is extensive literature on the implementation of confined vs. mixed (grazing + mixed ration) feeding strategies in dairy systems, most are short-term studies and there is little information on their effects on cow full lactation performance, associated with environmental exposure stress throughout their productive cycle, which depends on their calving season. This manuscript determines and interprets its effects on milk production and composition and energy balance (i.e., body condition score, non-esterified fatty acids, and beta-hydroxybutyrate) during a full lactation in two calving seasons, addressing whole-herd (extensive to whole-farm) feeding and management issues. The results demonstrate that outdoor soil-bedded milk production systems, when well-managed, have a very high milk production potential that could equate to the productive response of improved infrastructure systems (i.e., a compost-bedded pack barn with cooling capacity) under moderately unfavorable environmental conditions (i.e., infrequent heavy rains), but in worse situations (i.e., severe heat waves and frequent heavy rains), performance could be compromised.

Environmental exposure during confinement and feeding strategy affects cow behavior, nutrient utilization, and performance. Milk production and composition, body condition score, non-esterified fatty acids, and beta-hydroxybutyrate were determined during a full lactation in cows submitted to (a) grazing + partial confinement in outdoor soil-bedded pens with shade structures (OD-GRZ); (b) grazing + partial confinement in a compost-bedded pack barn with cooling capacity (CB-GRZ); or (c) total confinement (same facilities as CB-GRZ) and fed TMR ad libitum (CB-TMR). Autumn (ACS) and spring (SCS) calving season cows were used for each treatment, except for CB-TMR (only SCS). In ACS, treatments did not differ in any variable, possibly due to mild weather. In SCS, milk production was higher in CB-TMR than CB-GRZ, which in turn produced more milk than OD-GRZ. Differences coincided with heat waves and/or heavy rains (similar grazing conditions and mixed ration DM intake). Milk fat, protein and lactose yield, protein content, and BCS were higher in CB-TMR, without differences between CB-GRZ and OD-GRZ. Cows in OD-GRZ had impaired energy metabolism. Under moderately unfavorable environmental conditions (ACS), when well-managed, OD-GRZ systems could equate to the productive response of CB-GRZ. However, in worse climatic conditions (SCS), performance could be compromised, especially when compared to TMR systems.

1. IntroductionCows in totally confined milk production systems fed total mixed rations (TMR) have higher intake, production, and energy balance than those in pastoral systems that harvest their forage. Nevertheless, mixed systems (grazing + a MR; MS) can capture some benefits of confined systems while maintaining relatively low feeding costs, which could improve the economics of totally confined systems [1,2,3].Intensification of pasture-based milk production systems through stocking rate increments results in higher inclusion of supplements in the diet. Indeed, use of conserved forage in conjunction with concentrates in a MR results in higher dry matter (DM) intake and higher milk and solids production [1,2]. This production strategy involves cows being out of the paddock 40 to 60% of the time in an area where feed supplements are provided [3]. Furthermore, high pasture grazing intensity associated with high stocking rates leads to longer periods without access to pasture in order to recover forage in the system, thereby generating the need for facilities to confine the cows [4].Calving distribution through the year has important effects on the annual milk re-mittance pattern [5] and, consequently, on milk industry supply. Autumn/winter calved cows achieve more total milk production than those calved in spring. They also have longer lactation and a shorter calving to conception interval [6]. However, in Uruguayan winters, a lower pasture growth rate and heavy rains can prevent pasture access, resulting in less directly harvested forage inclusion in the diet of autumn calving cows (ACS) during early lactation. This results in higher feed supplementation and more time in confinement facilities. Although spring calving cows (SCS) achieve higher pasture DM intake and lower production costs, cows experience the challenge of maintaining normothermia in addition to the stress of early lactation [7], which can compromise welfare and productivity. Each dairy system must be evaluated for the opportunities and weaknesses of each calving strategy and adopt the one that best suits its productive objectives.When confinement time becomes considerable, the negative effects of exposure to environmental conditions on cow comfort, behavior, and performance are accentuated [8]. During heat stress (HS), energy demand for the thermoregulation of immune system hyperactivation diverts the energy supply to the mammary gland [9,10]. This, together with reduced nutrient absorption and lower udder nutrient uptake [11,12], causes milk production decreases of up to 35% [10,13]. On the other hand, under adverse winter conditions, with cold and mud in feeding and resting areas, cows spend more time standing and less time lying [14,15]. Lying time is critical not only because it is associated with rumination time [16], but it is also important for cows to achieve a sufficient amount of quality rest time [17]. Lying deprivation during confinement periods has an impact on behavior during next period activity, which in MS corresponds to grazing [15,16,18]. In addition, cow resistance to moving through muddy areas can cause cows to eat less overall, with fewer larger meals, which negatively impacts ruminal fermentation [19]. As a consequence, nutrient utilization and performance are impaired.There is scarce literature on pasture-based cow performance under different housing conditions and environmental exposure stress (i.e., cold, mud, or heat stress) at different stages of lactation, depending on calving season strategy. Research on the effects of expo-sure to the environment on intensive grazing systems is not only original but also highly relevant from an economic and environmental perspective for the intensification of dairy systems in the southern hemisphere [4]. Although the implementation of confined vs. MS feeding strategies has research antecedents, the existent literature refers to short-term studies [1,2,20,21,22,23,24], and according to our knowledge, no research has been conducted throughout the productive cycle. Furthermore, compost-bedded pack barns are novel facilities to house animals that allow cows freedom to move and a soft place to lye, thus improving animal welfare while enabling manure recycling, consequently diminishing environmental pollution and productive costs [25,26].The objectives of this study were to: (a) measure and evaluate the effect of different levels of environmental exposure on the performance of cows consuming MR + grazing in two different strategies of calving season (autumn and spring); (b) compare high and low environmental exposure MS with a 100% confined TMR system when detrimental heat stress effects are most pronounced (i.e., SCS strategy). It was hypothesized that milk cows partially confined in a compost-bedded pack barn with a cooling capacity would improve full lactation milk and solids production and energy balance when compared to cows partially confined in outdoor soil-bedded pens with shade structures. Further, a totally confined system would improve milk and solids production, as well as energy balance, compared to MS-fed cows, obtaining more contrasting responses when compared to cows housed in outdoor soil-bedded pens than when compared to cows housed in a compost-bedded pack barn. 2. Materials and Methods 2.1. Cows and Experimental DesignA total of 80 Holstein cows (2.8 ± 1.25 lactations, 640 ± 85 kg body weight; BW) were used in 2 calving experiments, 1 in each calving season, conducted at the Estación Experimental Dr. M. A. Cassinoni (EEMAC) of the Facultad de Agronomía (Paysandú, Uruguay) of the Universidad de la República (UdelaR). Experimental periods lasted from March 2019 to January 2020 for ACS cows and from August 2019 to May 2020 for SCS cows. The experimental protocol was evaluated and approved by the Comisión Honoraria de Experimentación Animal (CHEA) from UdelaR (Montevideo, Uruguay). All cows were managed equally during the dry and prepartum periods and confined and fed a pre-partum TMR for ~3 weeks before expected calving dates. Autumn-calving cows had calving dates of 18 March 2019 ± 14.5 days and spring-calving cows had calving dates of 16 August 2019 ± 8.2 days. Cows were blocked by BW, lactation number, pre-calving body condition score (BCS) and expected calving date, randomly assigned to treatment, and grouped in 4 pens of 4 cows each (i.e., 16 cows/treatment).Treatments consisted of: (a) ACS and SCS cows subjected to 8 h in a grazing paddock + supplemental MR in outdoor soil-bedded pens with shade structures during confinement (OD-GRZ; high environmental exposure MS); (b) ACS and SCS cows subjected to 8 h in a grazing paddock + supplemental MR in a compost-bedded pack barn with cooling capacity during confinement (CB-GRZ; low environmental exposure MS); (c) SCS cows subjected to a totally confined system with cows in the same facilities as CB-GRZ but fed a TMR twice daily ad libitum (CB-TMR). 2.2. Management and Feeding The CB-TMR and CB-GRZ cows were confined in a fully roofed, compost-bedded pack barn divided into pens, each containing four cows. Compost-bedded pack area was of 13.5 m2 per cow and was continued by a concrete area of 6.7 m2 per cow with access to a feed bunk per pen with a length of 0.75 m per cow and automatic drinkers to ensure ad libitum access to water. A 15 cm layer of new, fresh bedding material (i.e., rice husks and wood chips) was supplied every 15–20 days. Compost was superficially labored twice a day with a chisel plough in order to remove water vapor, allow oxygen entry, and maintain small, homogeneous particles. The concrete area was cleaned three times a week by trawling with a tractor carrying rubber. The barn had cooling capacity, with continuous-operation fans and sprinklers with automatic operation over 25 °C. The OD-GRZ cows were confined in outdoor soil-bedded pens made up of 2 paddocks, alternately occupied according to soil moisture and surface deterioration, with an area of 48 m2 per cow. Paddocks had a slight slope for water and manure runoff. Each pen had shade structures of 4.8 m2 per cow (nylon roofed at 4.5 m height with a slope of 15%), close to automatic drinkers (same as previous). Feeders were located at the other end of the paddocks, with a length of 1.10 m per cow and a feeding area of 10 m2 per cow.The milk parlor was built 100 m from pens in order to minimize cow activity and long waiting periods during milking.The TMR/MR were the same for all treatments. It varied over time according to available conserved forage as well as market-available grains and by-products for the commercial concentrate, with a total of 7 combinations used in the 15 months of the experiments (Table 1). Diets were formulated based on the recommendations of the National Research Council (NRC [27]) for 620 kg cows producing 45 L/d of 4% fat-corrected milk. In MS, both pasture and MR were considered nutritionally balanced diets, with MR used as a pasture complement (limited amounts) to achieve the desired DM intake, which was dependent on current pasture stock of the system. The MS (i.e., OD-GRZ and CB-GRZ) were high stocking rate systems (i.e., 2.5 lactating cows and/or 1550 kg BW/ha grazing platform), where cows had 8 h of daily access to weekly grazing plots, if allowed by weather conditions and/or grazing platform available herbage mass (HM). Both grazing treatments accessed different grazing plots (all pens of the same treatment together) with similar herbage allowances (HA). From March to October 2019, cows grazed between 7 a.m. and 2 p.m., while from November 2019 to April 2020, cows grazed from 6 p.m. to 2 a.m., in order to minimize heat stress and its negative impact during grazing. The pastures used were: tall fescue (Festuca arundinacea), lucerne (Medicago sativa) + orchard grass (Dactylis glomerata), oat (Avena sativa), annual raygrass (Lolium multiflorum), and soybean (Glycine max). Forage management and chemical composition are summarized by season in Table 2. Herbage mass was determined weekly using the double sampling technique [28] and then calculated pasture growth in order to adjust HA, taking into account sward condition (i.e., number of leaves or axillary buds) and available HM in the total grazing platform. Supplementation with TMR/MR was adjusted to ensure cow requirements and productivity goals were met and to achieve the appropriate pasture rotation length depending on pasture growth rate. 2.3. Data Collection, Measurements and EstimatesClimatic conditions (i.e., ambient temperature, relative humidity, rain) records were obtained from the meteorological agency of the experimental station. Heat stress was calculated using a temperature humidity index (THI) as: (1.8 × ET + 32) − (0.55 − 0.55 × RH/100) × (1.8 × ET − 26.8), where ET is environmental temperature and RH is relative humidity [29]. It was considered a mild heat wave when 2 of the 3 following criteria occurred at least 3 days in a row: daily THI average > 72, maximum daily temperature was >32 °C and/or minimum daily temperature was >23 °C. When all three conditions oc-curred simultaneously, it was considered a severe heat wave [30]. Cows were milked at 4 a.m. and 5 p.m. during spring/summer and at 3 a.m. and 4 p.m. during autumn/winter in order to minimize heat stress and its negative impact during grazing, as previously mentioned. Individual cow production was recorded at each milking. Milk samples were collected weekly from calving to 90 days in milk (DIM), biweekly from 91 to 180 DIM, and then monthly to the end of the lactation to determine milk fat, protein, and lactose levels (MilkoScan FossElectric FT2®, Drachten, The Netherlands).The offered and refused TMR/MR were measured weekly, as well as sampled, weighed, and oven-dried at 55 °C for 48 h to calculate dry matter intake. Samples were also analyzed for crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF), according to AOAC [31]. Total N for CP estimation used the Kjeldahl method of AOAC [32], which involves sulfuric acid digestion with subsequent distillation and titration. NDF used α-amylase, and, as for ADF, an ANKOM200 Fiber Analyzer (ANKOM Tech. Corp., Fairport, NY, USA) was used. Pasture was also sampled and chemically analyzed weekly.Daily pasture DM intake (kg DM/cow) was estimated by energy balance according to NRC [27], as the kg of pasture necessary to provide the remaining energy to achieve cow net energy (NE) requirements did not come from MR. Cow NE requirements were estimated as the sum of maintenance, pregnancy, and milk production requirements, taking into account energy contributed or required from losing or gaining BW and BCS [29]. Maintenance requirements were calculated as 80 kcal of NE/kg BW0.75, with a 20% increase for grazing activity in mixed systems [27]. Cow BW was measured monthly for this purpose. The amount of NE/kg BW was calculated according to actual BCS, adjusted for when it was used to support milk production or body deposition [29]. The BCS was assessed biweekly until 120 DIM and monthly from 120 to 305 DIM based on a 5-point scale, according to [33]. Pregnancy requirements were calculated from 190 to 279 days of gestation as: NEL (Mcal/day) = (0.00318 × D − 0.0352) × (CBW/45)/0.218, where D is day of gestation and CBW is calf birth BW in kilograms. Milk NEL concentration for productive requirements (i.e., energy in milk) was calculated from milk production and composition as milk NEL (Mcal/kg) = 0.0929 × Fat% + 0.0547 × CP% + 0.0395 × Lactose%. The NEL provided by pasture and TMR was calculated according to NRC [34] as Pasture NEL (Mcal/kg) = 2.149 − (0.0223 × ADF) and TMR NEL (Mcal/kg) = 1.909 − (0.017 × ADF).Blood samples were collected biweekly until 120 DIM and monthly from day 120 to 210 DIM, by venipuncture of the coccygeal vein, using 10 mL Vacutest® tubes (Vacutest Kima, Arzergrande, Italy) with heparin. Refrigerated samples were centrifuged at 3000× g for 15 min, and plasma was stored at −20 °C until analysis. Non-esterified fatty acids (NEFA) and beta hydroxybutyrate (βHB) concentrations were determined spectrophoto-metrically using commercial kits (Wako NEFA-HR (2) from Wako Pure Chemical Indus-tries Ltd., Osaka, Japan, and Oxidase/Peroxidase, UREA/BUN-UV, Ureasa/Glutamate De-hydrogenase from BioSystems SA, Barcelona, Spain, respectively).animals-13-01211-t001_Table 1Table 1Composition, chemical analysis, and nutritional value of mixed diets fed to lactating cows (% dry matter). From February-19June-19July-19August-19October-19March-20April-20 To May-19August-19October-19February-20April-20May-20Experiment ACS 1ACS 1ACS 1ACS 1ACS 1 SCS 2SCS 2SCS 2SCS 2Ingredient Forage Corn silage24.635.335.3--26.023.0 Sorghum silage---37.537.5-- Lucerne silage-----9.08.0 Ryegrass silage21.1---6.59.08.0 Fescue hay-2.06.06.5--- Commercial concentrate mix 354.362.758.856.056.056.061.0Dry matter (DM)43.451.859.755.950.437.041.0Nutrient NEL (Mcal/kg DM) 41.631.681.681.681.651.641.64 Crude Protein15.915.116.416.515.814.616.7 Neutral Detergent Fiber33.131.028.029.529.636.334.4 Acid Detergent Fiber16.513.613.513.515.115.715.9 Ether Extract3.83.54.13.84.63.13.5 Starch 22.027.020.018.017.017.018.01 Autumn calving season. 2 Spring Calving Season. 3 Based on ground corn grain, wheat bran, soybean meal, sunflower meal, cottonseed meal, canola meal, rumen inert fat, urea, yeast, and minerals. 4 Net Energy of lactation, calculated as 1.909 − (0.017 × ADF) according to NRC [34].animals-13-01211-t002_Table 2Table 2Forage management and chemical composition in each season in mixed systems with low (CB-GRZ) or high (OD-GRZ) environmental exposure. HerbageChemical Composition Allowance 1Mass 2DMNEL 3CP 4NDF 4ADF 4Autumn 2019CB-GRZ16.3237834.71.459.762.032.2OD-GRZ15.1241132.41.4713.857.929.5Winter 2019CB-GRZ17.9221322.71.6019.447.922.0OD-GRZ19.7242921.71.6418.247.820.0Spring 2019CB-GRZ23.8319824.01.5414.551.525.5OD-GRZ23.5302224.51.6516.751.219.3Summer 2020CB-GRZ26.5380435.71.5614.541.024.6OD-GRZ26.7356630.31.5517.940.125.1Autumn 2020CB-GRZ21.1249738.01.6122.443.121.6OD-GRZ19.8233737.81.6021.344.522.31 Expressed as kg DM/cow/day. 2 Expressed as kg DM/hectare, estimated at ground level. 3 Calculated as (3.2 − 0.028 × ADF) × 0.62 [35]. Expressed as Mcal/kg DM. 4 Expressed as % of DM.2.4. Statistical Analysis Data were analyzed using the MIXED procedure of SAS (SAS Institute Inc., Cary, NC, USA) with the model: Yij = µ + Ti + WOSj + Bk+ Ti × WOSj + eijk, where Yij is the response variable, Ti is treatment, WOSj is week of study (WOS), Bk is block fixed effect, Ti × WOSj is treatment by WOS interaction effect, and eijk is residual error. Treatments were compared with each other. The cow was considered the experimental unit for milk production and composition, BCS, and metabolite concentrations, while the pen was the experimental unit for TMR and pasture DM intake. Data were analyzed as repeated measures over time. Equidistant distribution was considered for milk production (daily) as well as DM intake (weekly), while uneven distribution was considered for milk solids production, BCS, NEFA, and βOHB. Values at calving were included in the model as co-variables to compare BCS, NEFA, and βOHB between treatments throughout lactation. Each calving season was analyzed separately. Mean comparisons were performed by Tukey-Kramer analysis. Mean differences were considered significant if p ≤ 0.05. Results are shown as least square means ± standard error of the mean (SEM). 3. ResultsThe daily average THI and percentage of the days in which THI was <68, between 68 and 72, and >72, and times of heavy rains (i.e., >50 mm/week during at least 3 consecutive weeks or >80 mm in a week) through the 15 months of both experiments are shown in Figure 1. Table 1 shows the ingredients and chemical composition of the 7 TMR/MR fed in both experiments (ACS and SCS). Table 2 shows HA and HM during each season of the experiment, as well as the chemical composition of all MS.3.1. Autumn Calving SeasonHeavy rain occurred during WOS 10, 15, 23, 32 to 35, 41 to 43, and 45. Mild heat waves occurred in WOS 37, 40, and 45, and a severe heat wave occurred at WOS 43 (Figure 1). The MR intake averaged 11.7 ± 3.4 kg DM/cow/day, while pasture was 7.2 vs. 7.3 ± 3.2 kg DM in CB-GRZ and OD-GRZ during periods without total confinement (Table 3). Cows were confined during WOS 6–7, 14–15, and 42–43 due to low HM. During these periods, DMI averaged 21.5 ± 4.3 kg/cow/day.All milk production response variables were affected by WOS, but none were affected by treatment (Table 4). Milk production had a T × WOS interaction (Figure 2). Although fat and protein content and protein had overall T × WOS interaction effects, there were no specific WOS in which the treatment outcomes differed.animals-13-01211-t004_Table 4Table 4Milk production and composition (means ± SEM) per cow in autumn and spring calving seasons in confined (CB-TMR) and mixed systems with low (CB-GRZ) or high (OD-GRZ) environmental exposure during complete lactation. Treatment p-Value CB-TMRCB-GRZOD-GRZSEMTRTWOSTRT × WOSAutumn calving season L/cow/day -26.226.50.160.14<0.01<0.01 Fat%-3.553.570.040.74<0.01<0.01 kg/d-1.021.010.020.58<0.010.07 Protein%-3.403.440.020.21<0.01<0.01 kg/d-0.980.980.010.80<0.01<0.01 Lactose%-4.824.870.030.28<0.010.06 kg/d-1.421.410.030.73<0.010.14Spring calving season L/cow/day 35.9 a27.4 b26.0 c0.17<0.01<0.01<0.01 Fat%3.59 a3.55 ab3.43 b0.040.03<0.01<0.01 kg/d1.34 a1.04 b1.01 b0.02<0.01<0.01<0.01 Protein%3.29 a3.17 b3.19 b0.040.01<0.01<0.01 kg/d1.23 a0.96 b0.95 b0.02<0.01<0.01<0.01 Lactose%4.89 a4.78 b4.87 a0.04<0.01<0.01<0.01 kg/d1.82 a1.44 b1.44 b0.02<0.01<0.01<0.01a,b,c Means within season with different superscripts differ (p < 0.05). TRT—treatment; WOS—weeks of study; TRT × WOS—interaction.The BCS patterns during complete lactation adjusted for BCS at calving (3.05 ± 0.27) were affected by WOS for all treatments (Figure 3) but did not differ between treatments (2.67 vs. 2.69 ± 0.02).Plasma levels of NEFA and βHB were only affected by WOS (Figure 4). The mean levels throughout WOS 2 to 30 were 0.35 ± 0.13 mmol/L of NEFA and 0.72 ± 0.07 mmol/L of βHB.3.2. Spring Calving SeasonHeavy rain occurred during WOS 12 to 15, 21 to 23, 25, 30, 34, and 37. Mild heat waves occurred in WOS 17, 20, 25, 28, and 32 to 34, and severe heat waves occurred in WOS 23 and 29 (Figure 1).The MR intake averaged 12.3 ± 2.3 kg DM/cow/day, while pasture averaged 6.8 vs. 5.8 ± 2.5 kg DM in CB-GRZ and OD-GRZ, not including periods of total confinement (Table 3). Cows were confined during WOS 22–23, 35–38, and 43–44, always as a consequence of low HM, averaging 20.0 ± 1.4 kg DM/cow/day. The totally confined system for SCS cows averaged a daily TMR intake of 25.7 ± 2.8 kg DM/cow/day. All production responses were affected by WOS treatment and had a T × WOS interaction (Table 4). Milk production was higher in CB-TMR than MS, and CB-GRZ produced more milk than OD-GRZ. Confinement system cows produced more milk than both MS in all WOS except 4, 36, and 37, whereas CB-TMR equaled CB-GRZ and was higher than OD-GRZ. The MS groups were similar, except in WOS 23, 29, and 37, when CB-GRZ milk production was higher than OD-GRZ. Fat content was higher in CB-TMR than OD-GRZ and intermediate in CB-GRZ. Fat, protein, and lactose yields, as well as protein content, were higher in CB-TMR compared to both MS, which were similar. Lactose content was higher in TMR and OD-GRZ than CB-GRZ. While fat content had a T × WOS interaction in SCS, no treatment differences occurred in any WOS. Protein content was only higher in CB-TMR than OD-GRZ and CB-GRZ in WOS 14. Lactose, protein, and fat yield were higher overall in CB-TMR than OD-GRZ and CB-GRZ.The BCS was affected by WOS treatment and had a T × WOS interaction (Figure 3). The confined system had a higher BCS than CB-GRZ and OD-GRZ (p < 0.01), which did not differ (2.76 vs. 2.62 and 2.59 ± 0.02, respectively). At WOS 15 and 26 onwards, CB-TMR had a higher BCS than MS. During WOS 40, OD-GRZ had a lower BCS than CB-GRZ, which in turn was lower than CB-TMR, with a BCS of 2.67, 2.87, and 3.15 ± 0.05, respectively (p < 0.01).Plasma NEFA was only affected by WOS (p < 0.01, Figure 4, mean 0.35 ± 0.03 mmol/L during the first 30 WOS). The plasma βHB was affected by treatment (p = 0.01), as well as WOS and the T*WOS interaction (p < 0.01). The OD-GRZ had the highest plasma βHB throughout, while the confined system had the lowest mean and CB-GRZ was intermediate (0.91 vs. 0.69 vs. 0.77 ± 0.05 mmol/L, respectively). During WOS 10, OD-GRZ had higher βHB plasma levels than CB-TMR, while CB-GRZ was intermediate (0.91 vs. 0.51 and 0.66, ± 0.15 mmol/L respectively, p < 0.01). At WOS 15, OD-GRZ had the highest βHB plasma levels of all WOS and treatments, 1.93 mmol/L, without differences between CB-TMR and CB-GRZ. In WOS 21, OD-GRZ βHB plasma levels declined and were intermediate between CB-GRZ and CB-TMR (0.94, 1.07, and 0.65 mmol/L, respectively, p < 0.05).4. Discussion4.1. Autumn Calving SeasonIn the autumn calving season, the MS subjected to different levels of environmental exposure did not differ from each other in any analyzed variable (i.e., milk production, milk composition, energy metabolism). Although June, October, and December rainfall values exceeded historical averages, during the rest of the year they were below historical values (the cumulative value of the 3 months was 598 vs. 307 mm, and the remaining months were 965 vs. 1369 mm for experimental station and the national historical average, respectively, Instituto Uruguayo de Meteorología) [36]. In addition, heat stress only occurred for 2 weeks at the end of the study at an advanced stage of lactation, when cows are less susceptible to heat stress due to lower intake and production, and therefore lower metabolic heat output [11]. Although the OD-GRZ treatment was more exposed to environmental conditions than the CB-GRZ, good maintenance of the infrastructure in the feeding and resting areas (i.e., cleaning after periods of rain, mound construction), as recommended [37], in addition to correct shade sizing [7] and ad libitum access to fresh water, likely mitigated the negative effects of such exposure. The open-air conditions to which the OD-GRZ treatment cows were subjected were judged to be better than those on commercial Uruguayan dairy farms. In general, commercial dairy farm animal facilities maintenance is less frequent and rigorous than in our study, such that cows face longer and muddier conditions, which are detrimental to their well-being and performance [14,16]. As well, the number of cows per pen is usually much higher than that used in our study, with higher deposition of excreta and higher ground pressure, causing higher moisture and surface deterioration.The seasonal HM and HA values, as well as paddock access time, indicate that cows had no l imitations [38,39] to reaching forage harvest levels close to 11 kg DM/cow/day [40]. Herbage allowance reached in our MS, characterized by high stocking rates and supplementation, was 38.2 vs. 38.5 for CB-GRZ and OD-GRZ, respectively. It should be taken into account that pasture DM intake values were estimated according to the difference between the energetic contribution of the TMR and estimated production and maintenance requirements, plus a fixed extra maintenance cost of 20% for grazing activity, making it an approximate estimate that allows values to be compared to other studies. Notwithstanding, milk fat content did not differ between groups. In general, increasing pasture inclusion causes a higher acetic:propionic ratio in the rumen [20,41], thereby increasing precursors for de novo milk fat synthesis as opposed to diets with a higher inclusion of concentrate. This result confirms the lack of difference in estimated pasture intake and pasture inclusion diet between our treatments [42].In ACS, the lack of treatment differences with no T × WOS interaction in BCS is consistent with similar levels of indicators of energetic metabolism (i.e., NEFA and βHB). Calving BCS was at the lower limit of what is considered desirable (3.03 and 3.00 ± 0.27 for both MS) according to Roche et al. [43]. However, cows mobilized tissue to a nadir of 2.58 in WOS 8 and 12, a value that is not critically low, which is supported by NEFA values, which decreased to <0.6 mmol/L in WOS 6, stabilizing in 0.16 ± 0.05 mmol/L between WOS 14 and 30.4.2. Spring Calving SeasonIn SCS mixed systems, differences occurred in milk production during full lactation among treatments subjected to different levels of environmental exposure. This was mainly due to the cumulative numerical differences between WOS 20 and 37, which was summer. As MR quantity (15.4 ± 3.9 kg DM) and composition, as well as HM and HA, were similar between treatments during this period (3783 ± 2610 kg DM/ha and 28.4 ± 15.9 kg DM/cow, respectively), the lower cow performance in the OD-GRZ treatment (1.5 L less than CB-GRZ) could be due to differences in environmental exposure. It was during this period that the 6 mild waves and the 2 severe heat waves were concentrated, and, during WOS 23, both treatments were in confinement, consuming equal amounts of MR. Thus, it seems that the productive difference was due to differences in environmental exposure, as the first severe heat wave occurred in that WOS. The next significant difference in WOS 29 coincides with the 2nd severe heat wave. At this time, cows were under similar grazing conditions: 16 vs. 18 kg DM/cow of HA and 2150 vs. 2500 kg DM/ha of HM for OD-GRZ and CB-GRZ, respectively. The third significant difference in milk production between MS and WOS 37 was observed. Moreover, during WOS 36 and 37, CB-GRZ milk production did not differ from CB-TMR, but it did from OD-GRZ. Although the conditions to rate WOS 36 as a heat wave were not reached, it averaged a THI of 73.2, which means that cows suffered heat stress. In addition, there was a heavy rain event in WOS 37 that could have affected milk production, as moisture content is aversive to cows’ willingness to lie down. The fact that both MS were in total confinement entails that OD-GRZ cows could not trade off confinement resting behavior deprivation during grazing time [16,18], impairing their welfare and performance.The higher energy demand due to the activation of adaptation mechanisms [9,10,44], together with the reduction of nutrient absorption and lower udder nutrient uptake [45] cause milk production to decrease by up to 35% [7,13]. Thus, milk production differences between levels of environmental exposure were ~5%, a lower difference that may be due to this being a long-term study that encompasses a full lactation, with part of the experiment occurring in thermoneutral conditions. During summer, the mean difference amounted to 10%, with peak differences of 20% in the weeks where heat waves occurred (i.e., WOS 23 and 29). No differences in milk fat and protein content occurred between MS subjected to contrasting environmental exposure, although milk lactose content was higher in OD-GRZ. Smith et al. [46] showed increased milk fat content (3.5 to 3.7%) and decreased milk protein content (3.2 to 3.1%) in heat stressed cows from mild to moderate HS without changes between moderate and severe HS. This is consistent with Wheelock et al. [9], who did not report an effect on milk fat or protein content in heat-stressed cows. In contrast, Cowley et al. [47] observed that HS had a strong influence on milk protein concentration but no effect on that of milk fat or lactose, hypothesizing that HS could provoke specific downregulation of mammary protein synthesis. Quantifying direct physiological effects on milk production is difficult, and inconsistent results have been reported, as it is also strongly affected by behavioral factors that affect nutrient ingestion and absorption, which is even more complex in grazing conditions. The numerically lower pasture DM intake and its inclusion in the diet in OD-GRZ, and therefore higher TMR inclusion, suggests a relatively higher starch passage to the small intestine and higher glucose availability in the mammary gland in comparison to CB-GRZ, which consumed more herbage, which could explain the higher milk lactose content [22,41].The treatment with the highest environmental exposure expressed an energy metabolism imbalance, evidenced by higher values of βHB between WOS 15 and 19 (~85 to 125 DIM), indicating situations of subclinical ketosis (>1.2 mmol/L [48]), which resulted in lower BCS recovery at the end of the study, compared to treatments less exposed to the environment (CB-GRZ) and/or with a higher feeding level (CB-TMR). Climatic conditions during WOS 12–15 were associated with cold rather than heat stress, as the ambient temperature aver-aged 18.5 °C and heavy rain occurred. Cows subjected to wet surfaces exhibit less lying and a lower quality of rest [16], in addition to the extra energy expended on thermoregulation [49]. The lack of a NEFA surge in OD-GRZ cows coupled to β-hydroxybutyrate may have occurred because of differential tissue utilization of NEFA due to increased physical activity [50,51]; as muscle shivering for thermoregulation [52], less lying and resting time [16], more walking time, and standing in postures that may reduce the amount of surface area exposed to wind and rain [14,49]. These metabolic differences allowed CB-GRZ cows to have different BCS from OD-GRZ cows at WOS 40 of almost a fourth point on the scale of Edmonson et al. [33].The CB-TMR treatment had 6.4 and 7.1 kg more DM intake than CB-GRZ and OD-GRZ, respectively, which reached pasture inclusion levels of 35.9 and 31.4% of the di-et [27]. In consequence, the confined group exceeded milk production by 31% and 38% compared to the MS of CB-GRZ and OD-GRZ, respectively, and reached an average of 28% more milk solids production than the MS, while maintaining higher and more stable production levels throughout the full lactation compared to systems including grazing. Indeed, MS production levels sharply declined (~31.0 to 23.5 L) from WOS 11 to 18, throughout spring, as daily PA incremented from 18 to 27 kg DM/cow and supplementation decreased from 15 to 11 kg DM/cow. Once summer began, PA diminished again, and therefore supplementation increased, showing a production improvement although not recovering initial values. The early difference in milk production in favor of CB-TMR (~8 L at 30 DIM) leads to the assumption that a short confinement at the beginning of lactation (i.e., the first 3 weeks) with ample TMR supply in conditions of low environmental exposure might be a management alternative in MS in order to maintain high production levels during full lactation.The higher milk production of CB-TMR vs. MS cows was due in part to higher DM intake, in agreement with other studies that indicate a higher DM intake in more nutrient-rich diets [1,2,42]. Bargo et al. [20] studied the performance of cows in early to mid-lactation during 21 weeks in spring consuming 100% TMR or 32% pasture + 68% MR and housed overnight in a free-stall barn and found that totally confined cows had higher DMI and milk production compared to MS (26.7 vs. 25.2 kg DM and 38.1 vs. 32.0 kg milk/day for each treatment, respectively), which represented an advantage of 19% in milk production during the study. Otherwise, neither milk fat nor true protein content differed between treatments [20]. Similar results were obtained by White et al. [53] in a 4-year study that showed 11% more milk production per lactation in cows fed TMR than MS cows but found no differences between treatments for milk fat or protein content in pasture + con-centrate (occasionally fed with pasture hay or haylage) compared to confined cows fed TMR. Vibart et al. [54] showed that increasing the pasture proportion of the diet caused a quadratic decrease in DM and CP intake, which resulted in lower milk production in cows consuming 35% of the diet as pasture compared to TMR-fed cows (32.7 vs. 36.6 kg/d) dur-ing mid-lactation, housed in free-stall barns in an 8-week study during spring, although there were no differences when consuming pasture during autumn at up to 41% of the diet inclusion of pasture. Again, no differences occurred in milk fat content between feeding systems (3.68 vs. 3.31%) or in CP (2.86 vs. 2.84%) with increased pasture proportion. Salado et al. [24], in a 9-week study in autumn-winter during early lactation, found no change in milk fat (3.88%) or protein (3.43%) content as TMR proportion increased from 25 to 100% of the diet, with gradual increases in productive levels in response to higher DM and energy intake as pasture proportion in the diet decreased. Subsequently, Salado et al. [55] observed higher milk production (33.7 vs. 32.3 kg/d) during early to mid-lactation in TMR fed cows compared to 75% TMR + 25% oat pasture during autumn–winter, with similar milk fat content (3.90%) but higher milk protein content (3.53 vs. 3.47%). In our study, milk protein content was higher in TMR-fed cows than MS-fed cows, in accordance with the latter experiment and with expectations, as DM and energy intake are associated with milk protein content [23]. Contrary to mentioned previous reports, milk fat content was lower in OD-GRZ cows relative to CB-TMR cows, although it was similar between CB-TMR and CB-GRZ cows. High environmental exposure during HS alters ingestive patterns, prompting less frequent and longer meals at a higher intake rate [11], thereby altering rumination patterns and saliva supply to the rumen [56,57]. Finally, fewer ruminal contractions and decreased blood flow to the rumen epithelium in response to HS [58] impair ruminal stability, fermentation, and nutrient absorption [59], which may cause a lower contribution of fatty acids to support the synthesis of milk fat in the mammary gland in OD-GRZ cows. In SCS, calving BCS was below recommended values for successful transition to early lactation [43], with treatment means of 2.80, 2.83, and 2.72 ± 0.21 for CB-TMR, CB-GRZ, and OD-GRZ, respectively. For the confined system, the BCS nadir was at WOS 10 (2.58 ± 0.03), while MS continued losing BCS until WOS 16, with a minimum of 2.48 ± 0.03. The confined treatment had a change in BCS slope from WOS 26 onward, which corresponds to 164 DIM, when BCS started recovering, although it was slow compared to that reported for these systems [43], probably due to the low BCS at calving. However, the confined treatment overcame the negative energy balance in less time than the MS, supported by lower levels of βHB in the OD-GRZ cows, and started increasing BCS earlier than in both MS, although there were no differences in plasma NEFA between treatments. The final BCS of MS cows was more than a quarter point lower in the scale [33] than the CB-TMR treatment cows, staying below 3.00 until the end of the study. These results are not consistent with those of Bargo et al. [20], where the final BCS did not differ between CB-TMR and MS, probably because this experiment started at 110 DIM, when the period of greatest lipid mobilization had already occurred. Fajardo et al. [2] did not find differences in BCS when comparing feeding strategies, similar to our results, since the experiment was completed in the 1st 10 weeks of lactation, when cows on both treatments were still in early lactation, as in our study.5. ConclusionsIn autumn calving cows, CB-GRZ and OD-GRZ failed to differentiate from each other in any measured response. This was likely due to the lower rainfall compared to the historical average, limited and mild heat stress only at lactation’s end, and good infrastructure design and maintenance in the feeding and resting areas. Spring-calving cows in a fully confined TMR system had the highest milk production and overcame the early lactation negative energy balance in less time than MS. In grazing spring-calving cows, lower milk production and worse indicators of energy metabolism (i.e., higher βHB and lower BCS recovery by the end of the study) showed that heat stress impaired the performance of grazing cows with high environmental exposure. Overall, results show that outdoor soil-bedded milk production systems, when well-managed, have a very high milk production potential that could equate to the productive response of improved infrastructure systems (i.e., a compost-bedded pack barn with cooling capacity) under moderately unfavorable environmental conditions, but in worse situations (i.e., the presence of medium or severe heat waves and heavy rain), performance could be compromised.

animals : an open access journal from mdpi

10.3390/ani11092602

PMC8472721

African swine fever (ASF) is a devastating viral disease of both wild boar and domestic pigs. Historically, the disease was mainly found in Sub-Saharan Africa. However, after the introduction of ASF into Georgia in 2007, the fatal disease spread to many European and Asian countries. In the absence of vaccines or treatment options, early detection of disease incursions is of paramount importance to limit the impact on animal health and pig industry. Thus, the biological characteristics of circulating virus strains must be known and communicated to practitioners and official veterinarians. Along these lines, the ASFV strain found in Belgium in 2018 was further characterized for its disease course in young and subadult domestic pigs. In general, clinical and pathological findings were in line with previous experiments utilizing highly virulent ASFV genotype II strains. However, in one of our experimental infections, four out of eight subadult domestic pigs showed milder signs and recovered, which was unexpected and points to an age dependency of clinical signs that could impact the early recognition of ASF incursions. We hope that communication of the available data will help practical and official veterinarians in the field to detect ASF as early as possible and thus minimize its impact.

African swine fever (ASF) is one of the most important and devastating viral diseases in wild boar and domestic pigs worldwide. In the absence of vaccines or treatment options, early clinical detection is crucial and requires a sound knowledge of disease characteristics. To provide practitioners and state veterinarians with detailed information, the objective of the present study was to characterize the ASF virus (ASFV) isolate “Belgium 2018/1” in subadult and weaning domestic pigs. To this end, two animal trials were performed. Trial A included eight subadult domestic pigs and trial B five weaner pigs. In general, clinical signs and pathological lesions were in line with previous studies utilizing highly virulent ASF genotype II viruses. However, in trial A, four subadult domestic pigs survived and recovered, pointing to an age-dependent outcome. The long-term fate of these survivors remains under discussion and would need further investigation.

1. IntroductionAfrican swine fever (ASF) is a highly contagious and devastating disease of Suidae. The causative agent, African swine fever virus (ASFV), is a large double-stranded DNA virus, which belongs to the genus Asfivirus in the Asfarviridae family [1].African swine fever has been endemic in many Sub-Saharan African countries and in Sardinia for many decades. However, after the introduction into Georgia in 2007, the disease spread to numerous eastern European countries and reached the European Union (EU) in 2014 with the first outbreaks in wild boar in the Baltic States and Poland [1,2]. In September 2018, the first case of ASF in wild boar was documented in Belgium [3]. The ASFV isolate associated with this outbreak, “Belgium 2018/1” [1,4], belongs to p72 genotype II and has shown high virulence in European wild boar [5]. Thus, it was comparable to other genotype II strains that are circulating in Europe and Asia, showing almost 100% lethality in animals of all age classes and sexes [6]. Clinical signs associated with such an infection include high fever, depression, inappetence, and respiratory distress [2,7]. Under experimental conditions, the animals showed the first clinical signs starting at 3–5 days post infection (dpi) [6], and animals developing an acute lethal disease course died within 7–13 dpi. Pathomorphological changes included enlarged, haemorrhagic lymph nodes; reddening of tonsils; congestion of spleen or splenomegaly; petechiae in different organs such as the kidney, colon, or urinary bladder; and lung and gall bladder wall edema [7]. Experimentally, the clinical course of ASF mostly has been studied in younger pigs (8–12 weeks), but there are some indications that the clinical course of ASF could be age-dependent under certain conditions [8].In the absence of vaccines or other treatment options, early clinical detection is of paramount importance and requires detailed knowledge of clinical signs and pathological changes. Thus, biological strain characterization is important to inform farmers, practitioners, and state veterinarians involved in disease control [9].Here, we report on the experimental inoculation of eight subadult domestic pigs and five weaner pigs for further characterization of the ASFV strain “Belgium 2018/1” and assessment of the influence of age on the clinical course and survival rate.2. Materials and Methods2.1. Experimental DesignThe study comprised two animal experiments (trials A and B), that were performed to assess virulence and pathogenesis of genotype II ASFV from Belgium (ASFV strain “Belgium 2018/1”), and to collect suitable reference materials. Trial A was performed at Sciensano, Brussels, Belgium and trial B was carried out at the Friedrich-Loeffler-Institut (FLI), Greifswald-Insel Riems, Germany.2.2. Animal Trials2.2.1. Trial AThe study comprised eight ASFV and ASFV antibody negative domestic pigs (Hypor hybrid × Pietrain breed) from a conventional farm weighing about 20 kg (10 weeks old). They were housed in a group at Sciensano in Animal Safety Level 3 facilities on slatted floors with water and food ad libitum for the duration of the experiment. The experiment was authorized by the Ethical Commission of Sciensano under N°20190614-01 and approved by the Biosafety commission. Upon arrival, the animals were randomly marked with ear tags starting from 1 to 8.As these animals were part of another trial, each animal received an intramuscular injection of 1 mL of saline solution (mock injection) after one week acclimation. A second mock administration was performed in the same manner 23 days after the first injection.At 49 days post mock injection and day 0 for this experiment, the subadult animals (18 weeks of age, weighing between 60 and 80 kg) were inoculated nasally using a small nebulizer (1-mm spray opening) fixed on a syringe to drip the infectious dose into the nostrils. Each animal received 4 mL virus suspension (2 mL per nostril) containing 1 × 104.3 hemadsorbing units (HAU)/mL of ASFV “Belgium 2018/1”. The inoculum was cultivated on PBMC following standard procedures (EURL protocol: https://asf-referencelab.info/asf/images/ficherosasf/PROTOCOLOS-EN/SOP-ASF-VI-1REV2018.pdf (accessed on 28 August 2021). Following the initial protocol that had been drafted assuming acute disease courses, the experiment was terminated at 18 dpi. Due to the lack of high containment capacities and organizational reasons, the study could not be extended beyond 18 dpi.Upon inoculation, body temperature and clinical parameters of all animals were assessed daily based on a harmonized scoring system as previously described [2]. Evaluated parameters are anorexia, recumbence, skin hemorrhage/cyanosis, swelling, breathing/coughing, ocular discharge, digestive trouble, and temperature. Based on the severity of the clinical signs, zero to three score points were awarded per parameter. The sum of points was recorded as the clinical score (CS). Temperatures higher than 40.5 °C were considered as severe fever, whereas temperatures higher than 39.7 °C (average +3 SD) were considered as mild fever. Animals reaching the humane endpoint of three subsequent days with severe fever (>40.5 °C), 9 score points, or were suffering unacceptably without reaching the endpoint score, were killed by electrocution and exsanguination.Blood and serum samples were collected prior to inoculation, 3, 7, 10, 14, and 18 dpi at the day of euthanasia. Necropsy was performed on all animals, and at the same time, tissue samples (lymph nodes, spleen, tonsil, lung, liver, and kidney) and blood (EDTA, serum) were collected.2.2.2. Trial BThe study comprised five ASFV and ASFV antibody negative domestic weaner pigs (German Landrace × Large White) from a conventional farm weighing 20 to 25 kg. They were kept in the high containment facility (L3+) of the FLI. The animal experiment was approved by the competent authority (Landesamt für Landwirtschaft, Lebensmittlsicherheit und Fischerei (LALLF) Mecklenburg-Vorpommern, Rostock, Germany) under reference number LALLF 7221.3-2-011/19. Upon arrival, all animals were ear-tagged individually with numbers from 16 to 20. Over the course of the trial, the animals were fed a commercial pig feed with hay cob supplement and had access to water ad libitum. The animals were kept in one group and received species specific stable enrichment.After an acclimatization phase, the animals were inoculated oro-nasally with 2 mL virus suspension containing 2 × 104.6 HAU/mL ASFV “Belgium 2018/1”. The experiment was carried out until 9 days post-infection, when all animals had reached the humane endpoint as defined above. Upon inoculation, clinical scoring was performed as described for trial A. The same endpoint definitions were applied. Animals reaching the humane endpoint were euthanized through intracardial injection of embutramide (T61, Merck, Darmstadt, Germany) after deep sedation with tiletamine/zolazepam (Zoletil®, Virbac, Carros, France), ketamine (Ketamin 10%, Medistar, Ascheberg, Germany) and xylazine (Xylavet® 20 mg/mL, CP-Pharma, Burgdorf, Germany) or ketamine (Ketamin 10%, Medistar, Ascheberg, Germany) and azaperone (StresnilTM 40 mg/mL, Elanco, Bad Homburg, Germany). Blood samples were collected prior to inoculation and at the day of euthanasia. At necropsy tissue samples (lymph nodes, spleen, tonsil, lung, liver, and kidney) and blood (EDTA and serum) were collected from all animals.2.3. Virus Inoculum2.3.1. Trial AThe inoculum, ASFV “Belgium 2018/1”, was isolated by the Belgian national reference laboratory (NRL) for ASF at Sciensano from a wild boar carcass found in the Belgian municipality Etalle (Luxembourg region) [3]. The isolate belongs to genotype II and is closely related to strains circulating in eastern Europe [4] and beyond.For the animal trial, cell culture supernatant was prepared on porcine peripheral blood monocytic cells (PBMCs) with a final titer of approximately 1 × 104.3 HAU/mL. The titer was confirmed by an end-point back titration of the inoculum and calculated using the Reed and Muench method [10]. Titers were expressed as the amount of virus causing hemadsorption in 50% of infected cultures (HAU 50/mL).2.3.2. Trial BThe inoculum, ASFV “Belgium 2018/1”, was shipped from the Belgian NRL for ASF at Sciensano to the NRL for ASF in Germany at the FLI. For animal trial B, culture supernatant was prepared with a final titre of approximately 1 × 104.6 HAU/mL. The titre was confirmed by an end-point back titration of the inoculum and calculated as described in trial A.2.4. Cells for Virus Titration2.4.1. Trial AAll virus titrations and haemadsorption tests were carried out using PBMC-derived macrophages according to the protocol of the European Union Reference Laboratory for ASF in which harvesting of the PBMC by buffy coat had been replaced by separation on SepMate™ PBMC Isolation Tubes (STEMCELL Technologies, Vancouver, BC, Canada) with Ficoll-Paque™ PLUS Media (GE Healthcare, Chicago, IL, USA).2.4.2. Trial BAll virus titrations and hemadsorption tests were carried out using PBMC-derived macrophages. PBMCs were obtained and treated as previously described [11].2.5. Pathology2.5.1. Trial AFull autopsy was performed on all subadult domestic pigs infected with the ASFV strain “Belgium 2018/1”. Pigs were investigated macroscopically, and all lesions documented.2.5.2. Trial BFull autopsy was performed on all domestic weaner pigs infected with the ASFV strain “Belgium 2018/1”. Pigs were evaluated macroscopically according to a scoring system published by Galindo-Cardiel et al. [12] with slight modifications [6].2.6. Processing of Samples2.6.1. Trial ASerum samples were obtained from native blood through centrifugation at 2.500× g for 20 min at 20 °C. Aliquots were stored at −80 °C until further use.Tissue samples were collected during necropsy and stored at −80 °C. Fragments of about 100 mg tissue were homogenized in 1 mL phosphate-buffered saline (PBS) with 2 metal beads using a TissueLyser II (Qiagen® GmbH, Hilden, Germany) for 2 × 2 min at 25 Hz before nucleic acid extraction.2.6.2. Trial BSerum samples were obtained from native blood through centrifugation at 2.500× g for 20 min. Aliquots were stored at −80 °C until further use.Tissue samples were cut into pea-sized fragments during necropsy and were stored at −80 °C for future use. One fragment was homogenized in 1 mL phosphate-buffered saline (PBS) with a metal bead using a TissueLyser II (Qiagen® GmbH, Hilden, Germany) for 3 min at 30 Hz before virus isolations (haemadsorption tests) and qPCRs were performed.2.7. Pathogen Detection—Nucleic Acid Extraction and Real-Time PCR2.7.1. Trial ADetection of viral genome was done in blood, serum, and tissues using real-time PCR (qPCR). For qPCR, viral nucleic acids were extracted from blood and serum using the IndiMag Pathogen Kit (Indical Bioscience, Leipzig, Germany) on the Indimag48® extraction platform (Indical Bioscience, Leipzig, Germany) and for tissue, the High Pure Viral Nucleic Acid Kit (Roche Applied Science, Penzberg, Germany) was used. All qPCRs were performed using the primers and probes for ASFV and endogenous gene beta-actin published by Tignon et al. [13] in AgPath-ID™ One-Step RT-PCR Master mix (Applied Biosystems, Foster City, USA) as described in Schoder et al. [14]. All PCRs were performed using a LC480® cycler (Roche, Basel, Switzerland). Results of the qPCR were recorded as quantification cycle (Cq) values. Using a dilution series of an in-house ASFV DNA standard, the genome copies in the respective samples were determined. For generation of the ASFV standard, p72 gene from Lisbon60 strain (genotype 1) was amplified and cloned in pCR2.1 (Invitrogen, Carlsbad, CA, USA) for further multiplication in competent E. coli. The plasmid was extracted with Plasmid Plus Maxi Kit (Qiagen, Hilden, Germany) and linearized by restriction with BamHI. Subsequently, the DNA concentration was determined by spectrophotometry using a Nanodrop 2000 c (Thermo Fisher Scientific, Waltham, MA, USA) and the exact number of DNA molecules was calculated using an online tool (http://www.molbiol.edu.ru/eng/scripts/0107.html (accessed on 25 March 2020).2.7.2. Trial BPrior to real-time PCR analysis, viral nucleic acids from all samples were extracted using the NucleoMag VET kit (Macherey-Nagel, Düren, Germany) on the automated KingFisher 96 flex platform (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s recommendations. Subsequently, nucleic acids were analyzed using the qPCR protocols published by King et al. [15] and Tignon et al. [13] on a Biorad CFX real-time cycler (Bio-Rad Laboratories, Hercules, CA, USA). For each qPCR, a quantification cycle (Cq) value was determined. Using a dilution series of the same standard as described in trial A, which was provided by Sciensano (Belgium), the genome copies in the respective samples were determined.2.8. Antibody Detection2.8.1. Trial ASera were tested in three commercially available antibody ELISAs. In detail, ASFV p72-specific antibodies were detected using the INGEZIM PPA COMPAC ELISA (Ingenasa, Madrid, Spain), the ID Screen ASF competition ELISA (IDVet, Grabels, France) for antibodies against p32, and the ID Screen® African Swine Fever Indirect (IDVet, Grabels, France) for antibodies against p32, p62 and p72. The tests were carried out according to the manufacturer’s instructions.In addition, serum samples were tested in the indirect immunoperoxidase test according to the standard protocols provided by the European Union Reference Laboratory for ASF with slight modifications regarding the virus strain (Lisbon60 ASFV strain adapted on Vero cells). Over the study period, results were recorded in a qualitative way (positive/negative). Sera taken upon necropsy were end-point titrated in log2 steps to obtain semi-quantitative antibody titers.2.8.2. Trial BSerum samples were tested as described for trial A. Slight modifications concerned the virus strain. Here, a cell culture adapted variant of genotype II ASFV “Armenia 2008” was used for the indirect immunoperoxidase test.2.9. Statistical AnalysisInitial data recording and analyses (comparison of mean values, transformation of values) were done using Microsoft Excel 2010 (Microsoft Germany GmbH, Munich, Germany).GraphPad Prism 8 (Graphpad Software Inc., San Diego, CA, USA) was used for graph creation.3. Results3.1. Clinical Findings3.1.1. Trial AFollowing nasal inoculation, all animals developed unspecific clinical signs starting from day 4 post inoculation (pi) (see Figure 1). The signs included general depression, lack of appetite and unwillingness to stand up, reduced mobility, tremor, reddened skin, cyanosis on the ears and snout, hunched-up back, and respiratory distress. The highest score was reached at 8 and 9 dpi with a maximum of 7 points.The onset of fever was observed as early as 4 dpi and the climax of the illness was observed around 8–9 dpi (see Figure 2). The peak of fever was observed at 7 and 8 dpi with the highest number of severe feverish animals. Afterwards the intensity of the fever decreased but remained present in a mild form until 14 dpi. No temperature increase was observed during the infection period for animal #3 despite the presence of other clinical signs. The average number of days with fever was as follows: for severe fever 2.75 +/− 2.25 and for mild fever 5.5 +/− 3.85.Ethical euthanasia was performed between 7 and 14 dpi on animals with clinical scores between 6 and 7 and showing persistent severe fever (3 subsequent days) and/or very poor reactivity. Starting from 12 dpi, the remaining animals started to recover. At the end of the experiment (18 dpi), the surviving pigs, four out of eight, presented no more clinical signs.3.1.2. Trial BFollowing oronasal inoculation, all animals developed severe, unspecific clinical signs starting from day five pi (see Figure 1). The signs included general depression, lack of appetite and mobility, hunched-up back, ataxia, and respiratory distress. The onset of fever was observed at 5 dpi (see Figure 2). On 9 dpi, all animals reached the humane endpoint except one (pig 17), which died overnight. The clinical scores ranged from 5.5 up to 11.The survival curve from both animal trials is presented in Figure 3.3.2. Pathomorphological Findings3.2.1. Trial AAnimals were euthanized either during the experiment for ethical reasons (severe fever for more than 3 consecutive days or high clinical score or poor health condition) or at the end of the experiment (18 dpi). At necropsy, various pathological patterns were observed: from asymptomatic to typical ASF lesions of varying severity: generalized hemorrhagic lymphadenopathy especially of the gastrohepatic lymph nodes, congestion of the spleen, and multiple hemorrhages in several organs, particularly in the kidneys (Table 1).3.2.2. Trial BFive ASFV “Belgium 2018/1” infected domestic pigs reached the humane endpoint at 9 dpi (n = 4) or died spontaneously (n = 1) and were submitted to necropsy. At gross pathologic investigation, all infected pigs revealed typical lesions indicative of ASF. All pigs showed severely hemorrhagic enlarged lymph nodes with the gastrohepatic and renal lymph nodes mainly affected. Renal petechiae were present in all pigs and were mainly confined to the renal cortex and to a lesser extent to the medulla. Four out of five pigs had ascites. Marked sanguineous effusion was present in two pigs, while the other two showed accumulation of serous fluid. Likewise, rather mild, serous to sanguineous pleural effusion was present in three animals. Bruises of variable size appeared in four pigs. In individual cases, mild to moderate perirenal and gall bladder wall edema, mucosal petechiae in the urinary bladder, bilateral cyanosis of the ears, mild multifocal pulmonary consolidation, and alveolar edema were observed (Table 2).3.3. Pathogen Detection3.3.1. Trial APrior to infection, all animals were tested negative for ASFV nucleic acids by qPCR in blood samples. After infection, the presence of the virus was first detected in blood samples of 3 animals collected at 3 dpi. All remaining animals were detected ASFV PCR positive at the following sampling time point (7 dpi), except one animal (#3) which became positive 10 dpi. Once detected positive, the animals remained positive until death by ethical euthanasia or at the end of the experiment (18 dpi).When detected at 3 dpi, the virus load was low (<25 copies/run). However, at 7 dpi and until 14 dpi, the virus load in blood was at a maximum with 1.29 × 103 to 6.80 × 104 copies/reaction (Table 3).At the end of the trial, viral genome was detected in blood, serum, and all tissue samples (spleen, lung, lymph nodes, tonsil, kidney, and liver). Highest loads of viral genomes (>1 × 103 to 1 × 104 copies/reaction) were found in blood, serum, and tissues from animals presenting severe clinical signs, whereas low viral genome loads (<500 copies/reaction) were detected in the blood of animals that survived the infection (Table 4).3.3.2. Trial BPrior to inoculation, all animals were tested negative for ASFV, ASFV antigen, and viral nucleic acids.At the end of the trial, viral genome was detected in blood and all tissue samples (spleen, lung, lymph nodes, tonsil, kidney, and liver). Highest loads of viral genome were found especially in blood, serum, and spleen (Table 5).3.4. Antibody Detection3.4.1. Trial APrior to inoculation, all animals were tested negative for ASFV antibodies (Ab) by ELISA tests.After infection, seroconversion against ASFV p72 was demonstrated using the INGEZIM PPA COMPAC ELISA (Ingenasa, Madrid, Spain) in the five still remaining animals at 10 dpi (two positive and three doubtful results) (Figure 4). With the ID Screen ASF Competition ELISA (IDVet, Grabels, France), which detects p32, all remaining animals were positive from 10 dpi, except one which already showed a questionable result at 7 dpi. By testing with the ID Screen ASF Indirect ELISA (IDVet, Grabels, France), three animals showed a questionable result at 10 dpi, one was negative, and one showed a positive antibody result from 10 dpi onwards. From 14 dpi and onwards, all four remaining animals were seropositive in all three ELISA assays.With the confirmatory indirect immunoperoxidase test, seroconversion was observed starting from 7 dpi (qualitative result). At the end of the experiment (18 dpi), the surviving animals presented antibody titers between 2560 and 5120 (semi-quantitative result).3.4.2. Trial BNo antibodies were detected in the sera prior to inoculation or at the end of the trial on 9 dpi with all ELISAs (Figure 4). With the indirect immunoperoxidase test, however, the animals showed semi-quantitative antibody titers between 160 and 620 at 9 dpi.4. DiscussionAfrican swine fever virus has entered the European Union and many countries in eastern Europe and Asia, threatening animal health and agriculture alike. In the absence of a licensed vaccine or other effective treatment options, prevention through farm biosafety and the earliest possible detection of outbreaks are of utmost importance. Early detection is only assured if clinical signs are recognized promptly and interpreted correctly by farmers and practitioners. Thus, well-trained livestock farmers and veterinarians in practice and subsequently the competent authorities play a crucial role here. Basic data to inform these key persons can be obtained from animal experimental work, such as that carried out in the National Reference Laboratories (NRLs) as part of their tasks. These studies are not infrequently conducted to generate and archive relevant and well-characterized sample materials for validation and harmonization of diagnostic methods at the national level (Regulation (EU) 2017/625, Article 101), but provide space to map issues of pathogenesis and virulence of locally relevant viral variants. In this context, the present study was conducted in close cooperation with the Belgian and the German NRLs with the aim to further characterize the ASF virus isolate “Belgium 2018/1” in domestic pigs. Thereby, the focus was laid on the clinical courses and pathomorphological changes in different age classes of animals. While various studies with genotype II strains have been conducted in weaner pigs, studies with older animals are scarce. Although there are indications that no age dependency is involved in highly virulent strains [16], different courses have been described for moderately virulent strains [8]. In the study published by Post et al. [8], age had a marked effect on disease outcome, while the inoculation dose was secondary. The latter is consistent with studies describing similar, often lethal, courses with different doses of highly virulent virus strains [11].In our combined study, inoculation of weaner pigs did not hold any surprise. After an incubation period of five days, which is in line with previous findings [17,18], all young animals developed an acute lethal disease course with high fever, general depression, anorexia, ataxia, and respiratory distress. All animals reached the humane endpoint or had died acutely by 9 dpi. It is noteworthy that again the clinical signs were severe but rather unspecific, leading to many differential diagnoses that could be relevant in the field. Necropsy findings were in line with previous studies using ASFV “Armenia 2008” [6]. Viral genome was found in organs, blood, and serum of all animals irrespective of the disease course. In line with previous studies, blood, spleen, and liver showed the highest copy numbers in the majority of animals. However, genome loads reflected the clinical course and timepoint of sampling. While viral genome loads in spleen were roughly 10,000 genome copies per run for animals sampled between days 7 and 9, less than 10 genome copies were found in recovering pigs at 18 dpi.The inoculation of the older pigs took a different, rather unexpected turn. Although the Belgian virus hardly differs at the sequence level from the highly virulent strains from Armenia and Georgia [4], which showed rather age-independent courses in previous studies [16], only a 50% lethality was recorded in the present study until the end of the experiment at 18 dpi. Clinical signs were milder and pathomorphological findings reflected the clinical course (severe to absent lesions). It must be mentioned, however, that the capacity-related observation period of 18 days limits a reliable and complete statement on the final clinical outcome. Previous studies have shown intermittent and recurring viremia after 18 dpi [2]. In addition, it cannot be excluded from the clinical scoring and virus detection that at least one animal (#3) got infected only by contact (not following inoculation) and thus was slightly delayed. Nevertheless, the lack of findings in the pathological-anatomical examinations and low genome loads in organs indicate that the remaining animals, including animal #3, were true survivors or showed only a sub-clinical infection due to the lack of lesions. Survivors were rarely seen when using highly virulent ASFV strains of genotype II; however, they were reported. As an example, Gallardo et al. [2] reported on the survival of one out of eight pigs inoculated intramuscularly with the Lithuanian ASFV strain “LT14/1490”. The respective pig showed weak and intermittent peaks of viremia, and viral DNA could be detected in tissues at 38 dpi. However, no seroconversion was observed, which was different in our study where all survivors seroconverted. Animals surviving the acute phase were also reported by Walczak et al. [9]. In this study, two animals inoculated with the Polish ASFV “Pol18_28298_O111” strain (pig one got 1000 HAU and pig two got 500 HAU) developed chronic disease courses after a delayed incubation period (pig one 12 dpi: clearly visible clinical signs, like joint swelling and minor breathing disorders, moderate fever, and constant low virus load value in blood but without pathological lesions. Pig two 20 dpi: showed only moderate fever and enlarged submandibular lymph nodes) but had to be euthanized 24 and 32 dpi, respectively. While it is obvious that animals may survive, the long-term fate and epidemiological role of these survivors are still discussed controversially and need further long-term studies to allow final conclusions. While survivors may eventually recover completely, longer term virus excretion will impact on transmission dynamics. In a previous long-term study with a moderately virulent genotype I strain, virus was isolated from recovering pigs up to day 63 post infection [19].Apart from age, application route and dose, the general health and immune status, genetic background (hybrid breed), and concomitant infections could impact on the clinical outcome and thus explain our observations. Indeed, due to the size of the sub-adult pigs, inoculation had to happen in standing position rather than in dorsal recumbency, which was used for the weaner pigs. This could have influenced the amount of virus that reached the tonsils. Adding to that, the dose is reduced when compared to body weight. However, as mentioned earlier, several studies showed severe courses upon low-dose infection and did not report a significant impact of the viral dose on the final outcome [2,9,11]. Whether an even lower dose would have led to more survivors or just less infected pigs could be debated. Considering the study reported by Pietschmann et al. [11], we could rather expect lower infection rates. The route itself can also play a role in the efficiency of infection. While trial A was conducted with nasal inoculation, trial B was done with oro-nasal inoculation. In this context, Howey et al. [20] reported on the variable efficiency of intranasopharyngeal (INP) and intraoropharyngeal (IOP) inoculations, especially using lower doses (102 HAU). In this study, oropharyngeal infection was less efficient. However, infection was confirmed in all animals of our study upon either nasal or oral infection. Given that the full range of clinical outcomes and swift seroconversion in survivors were seen, a dose-related impact cannot be excluded but does not seem likely. Considering that both studies employed clinically healthy, commercial pigs, the impact of the general health status seems rather small.With regard to the genetic background, differences in susceptibility are seen rather frequently in indigenous pig breeds in Africa [21,22]. Yet, our study involved only widely used pig breeds of Europe that may not differ markedly in their susceptibility, even if a different hybrid breed was used in trial A than in trial B. As genotyping of pigs was not carried out, no final conclusion on this aspect can be drawn.Recently, the gut microbiota was discussed as an important factor for ASF susceptibility. Interestingly, fecal microbiota transplantation from warthogs to domestic pigs resulted in higher resistance of the latter [23]. While we cannot rule out such factors, we do not assume a high impact of the gut microbiota in our study.The variability of clinical signs and outcomes in our study points to both a certain age-dependency and biological variability that has to be taken into account when communicating clinical signs and typical disease outcomes.5. ConclusionsTaken together, we saw a variable clinical picture when considering all age classes of animals that could cause problems in the clinical evaluation of ASF under field conditions and in early warning scenarios. The final fate of the here observed survivors could not be addressed under the limited time scheme of our study and should be part of further long-term studies with older pigs.The outcome of our study highlights the need for swift and reliable laboratory diagnosis, even when only mild to moderate clinical signs are detected. Easy and fast genome detection by qPCR did not pose any problems even with the surviving animals.We hope that communication of the available data will help practical and official veterinarians in the field to detect ASF as early as possible and thus minimize its impact. The study outcome clearly underlines once again that clinical courses can be highly variable and non-specific. For this reason, exclusion diagnostics of ASF by sensitive qPCR methods should be routine, especially in light of the current situation with ASF outbreaks in several countries including Germany.

animals : an open access journal from mdpi

10.3390/ani11030722

PMC8000480

In this study, we characterize the influence of short-term (4 days) heat stress on Holstein cows during early lactation. The use of indicators, such as production performance, physiological variables, blood parameters, micro-RNA expression, and metabolomes, in heat-stressed cows during early lactation—which is a high-stress phase—may provide insights into how to deal with the level of damage to dairy cows, through appropriate nutritional and management strategies. We identify that short-term heat stress has a negative effect, to some extent, on feed and water intake, rectal temperature, heart rate, blood hematology and metabolites, milk characteristics, miRNA expression in milk, and metabolomics in blood.

This study aims to characterize the influence of short-term heat stress (HS; 4 day) in early lactating Holstein dairy cows, in terms of triggering blood metabolomics and parameters, milk yield and composition, and milk microRNA expression. Eight cows (milk yield = 30 ± 1.5 kg/day, parity = 1.09 ± 0.05) were homogeneously housed in environmentally controlled chambers, assigned into two groups with respect to the temperature humidity index (THI) at two distinct levels: approximately ~71 (low-temperature, low-humidity; LTLH) and ~86 (high-temperature, high-humidity; HTHH). Average feed intake (FI) dropped about 10 kg in the HTHH group, compared with the LTLH group (p = 0.001), whereas water intake was only numerically higher (p = 0.183) in the HTHH group than in the LTLH group. Physiological parameters, including rectal temperature (p = 0.001) and heart rate (p = 0.038), were significantly higher in the HTHH group than in the LTLH group. Plasma cortisol and haptoglobin were higher (p < 0.05) in the HTHH group, compared to the LTLH group. Milk yield, milk fat yield, 3.5% fat-corrected milk (FCM), and energy-corrected milk (ECM) were lower (p < 0.05) in the HTHH group than in the LTLH group. Higher relative expression of milk miRNA-216 was observed in the HTHH group (p < 0.05). Valine, isoleucine, methionine, phenylalanine, tyrosine, tryptophan, lactic acid, 3-phenylpropionic acid, 1,5-anhydro-D-sorbitol, myo-inositol, and urea were decreased (p < 0.05). These results suggest that early lactating cows are more vulnerable to short-term (4 day) high THI levels—that is, HTHH conditions—compared with LTLH, considering the enormous negative effects observed in measured blood metabolomics and parameters, milk yield and compositions, and milk miRNA-216 expression.

1. IntroductionEnvironmentally induced hyperthermia is a global-scale threat to the dairy industry in many ways, with its effects including economic loss, animal health issues, and productivity. The adverse effects of heat stress (HS) on the productivity of dairy animals, in terms of milk yield, composition, and quality, have been well-documented [1]. Despite advances in cooling systems and environmental management [2], HS continues to negatively affect the diversity of dairy production characteristics. As such, lactating dairy cows under HS experience limited energy intake and, thus, are unable to meet the demands of their bodies for maintaining milk production and health, resulting in reduced milk yield and quality and leaving the animals susceptible to diseases [1,3,4]. The harm of HS is particularly more important in high-yield dairy cows, due to their higher susceptibility to stressors which periodically happen during early stage of lactation [5,6,7].The temperature humidity index (THI) has exclusively been employed as an index of HS status in cattle [3]. However, short-term and long-term exposure to HS may reflect variations in responses of the animals. Furthermore, THI is based only on the air temperature and humidity [3] and, as such, is not a direct biomarker of metabolic alterations in response to HS. Thus, alterations in physiological parameters, such as rectal temperature, heart rate, and respiratory dynamics, reflect the degree of HS in cattle [8], as well as feed intake [1,4], as the first indicator of HS and, consequently, its influence on milk yield and characteristics [1]. Feed intake can establish certain blood metabolite alteration changes during HS. Besides those, metabolic profiling (known as metabolomics) has increasingly been used in clinical pharmacology and is an ideal tool for the acquisition of the several thousand metabolite alterations that are applied in the relationship between endogenous metabolite metabolism and body metabolism [9]. Metabolomics has been applied in cow investigations, in order to predict the risk of diseases [10], and has been used for biomarker and pathway discovery in some metabolic diseases in cows [10]. Given this review, metabolomics analyses provide a powerful platform for the identification of animals and humans associated with pathophysiological alterations resulting from exposures to specific environmental factors [10]. However, the metabolic changes related to short-term HS during the early lactation stage in dairy cattle still remain unclear. The activation process of the hypothalamus–pituitary–adrenal axis enhances the production and circulation of cortisol in the animal body. It is the primary indicator that ruminants can be identified when they are stressed [11]. Additionally, haptoglobin is one of the acute phase protein indicators commonly identified to animal health and inflammatory responses [11]. Hence, measuring these two correlated hormones in biological samples will allow better understating the mechanism behind the effect of short-term HS in dairy cows. HS response is a complex molecular process that involves the transcriptional and post-transcriptional regulation of stress-related genes. Acute environmental change initiates the heat stress response at the cellular level. Furthermore, microRNAs (miRNAs) was recognized as an important regulator of gene expression beyond the transcriptional stage and various biological reactions such as development, apoptosis, differentiation, and viral infection [12]. Additionally, miRNAs regulate numerous functions of bovine mammary epithelial cells (bMECs), which play a putative role in milk secretion [13]. In humans, it has been documented that miR-216 can regulate cell apoptosis through repressing target genes in several cancer cells [13]. Likewise in this study, we aimed at exploring the potential involvement of miR-216 in the milk of heat stressed cows. Furthermore, to better deal with the consequences of HS, we aimed to characterize the short-term effects of HS in terms of not only the alterations of productive performance characteristics but also the narrative response of metabolomics and gene expression.Therefore, the objective of this study was to characterize short-term HS (4 day) in Holstein dairy cows using altered indicators of metabolomics, milk miRNA-216 and characteristics, and blood metabolites.On the other hand, our traditional knowledge regarding HS effects is limited to long-term HS, in which the animals were housed outdoors or underwent uncontrolled environmental housing conditions, where the temperature and humidity were subject to many fluctuations over time, such as from being severe around noon to cooling in the evening. These phenomena may compromise our understanding of HS effects, due to disparities in environmental and housing situations, thus misleading the characterization of a framework of HS influence. In this regard, the outcome of this study provides insights towards better understanding how dairy cows exposed to chamber-controlled short-term HS can be characterized, not only through production performance parameters but also using newly presented indicators, such as metabolomics and miRNA analyses.2. Materials and Methods2.1. Experimental DesignThis study was conducted at Konkuk University experimental farm, Republic of Korea. The experimental procedure was evaluated and approved by the Institutional Animal Care and Use Committee at Konkuk University (KU19121). Among the herd of multiparous early Holstein lactating cows, eight healthy animals with very similar conditions of days in milk (DIM = 40 ± 9 day; p > 0.05), milk production (milk yield = 30 ± 1.5 kg/day; p > 0.05) and parity (1.09 ± 0.05; p > 0.05) were chosen for this study. These criteria allowed homogeneity among the cows in the groups, and thus, any small changes could define reliably. The cows used for this specific experiment were assigned to two groups with distinct temperature and humidity index (THI) levels (of ~71 and ~86, respectively). The reason for choosing these two THI values (71 and 86) was that a THI of 71 represents the thermoneutral zone, while a THI of 86 represents severe HS; thus, we ensured that cows were under two distinct environmental conditions and not the range in between. As we had only four chambers (5 m × 5 m × 5 m), the two sets of THI levels were not conducted simultaneously. Hence, the first four chambers were used for the THI = 71 group, in order to provide the group with a low-humidity, low-temperature (LTLH) environment (period 1). Then, the second run was carried out with the other group of cows—the THI = 86 group—which were provided with a high-temperature, high-humidity (HTHH) environment (period 2). Each period, the environment inside the chamber was adjusted, in order to maintain the specified temperature and humidity for 3 days during the adaptation period. After the adaptation period, the temperature and humidity were controlled, for the cows to undergo HS for 4 days. The temperature was controlled using an air conditioner (CSVR-Q118E, Carrier corporation, Thailand), and the humidity was justified with the specified temperature and humidity using a humidifier (DE-9090UH, Zhongshan Xinhao Electrical.,LTD, Guangdong, China) and a dehumidifier (EDHA11W3, WINIA) throughout the experiment. Regarding Korean photoperiod conditions, light was provided from 09:00 h to 19:00 h. Experimental diets were fed twice daily, at 08:00 h and 14:00 h. All cows were fed common basal diets throughout the entire experimental period, according to NRC nutrient requirements (2001; Table 1). Basal diet and amino acid (AA) compositions are summarized in Table 1. Fresh water was provided five times daily during the experiment.2.2. Temperature Humidity IndexThe temperature humidity index (THI) was calculated using the following equation: (1.8 × Tdb + 32) − [(0.55 − 0.0055 × RH) × (1.8 × Tdb − 26)] [8,14], where Tdb is the dry-bulb temperature (°C), and RH is the relative humidity (%). The temperature and humidity inside the chambers were controlled using an automatic computerized system. In the automatic chambers, the ambient temperature and humidity were fixed at 25 °C and 35–50%, respectively (low-temperature, low-humidity, LTLH), and 31 °C and 80–95%, respectively (high-temperature, high-humidity, HTHH).2.3. Individual Animal Sampling and Analysis2.3.1. Feed and Water IntakeDuring the experiment, each cow was provided with feed and water twice a day (08:00 and 14:00 h) individually, while the remaining amount of feed (FI) and water (WI) intake were measured daily using a scale machine (GL-6000S Series, G-Tech International Co., LTD., Uijeongbu-si, Korea). 2.3.2. Milk YieldEach early lactating cow was milked twice daily by portable machine (PMM 1B EPV170, Italy) at 05:00 and 17:00 h. The amount of the milk was recorded using a scale (GL-6000S Series, G-Tech International Co., LTD.) after each milking.2.3.3. Milk Composition AnalysisDaily milk samples (05:00 and 17:00 h) were mixed and sub-sampled bi-weekly for milk composition analyses. The milk was sampled in a 50 mL tube and analyzed immediately. The samples were analyzed using a milk scanner FT1 (Foss Alle 1 DK-3400 Hilleroed, Denmark) for protein, fat, lactose, solid-not-fat (SNF), somatic cells, milk urea nitrogen (MUN), acetone, beta-hydroxybutyrate (BHB), beta-casein, mono (MUFA)- and poly (PUFA)-unsaturated fatty acids, saturated fatty acids (SFA), total fatty acids (TFA), milk protein and fat yield, 3.5% fat-corrected milk (3.5% FCM), and energy-corrected milk (ECM). Milk protein and fat yields, 3.5% FCM, and ECM were calculated by multiplying the milk yield from protein and fat composition of the milk of an individual.2.3.4. Blood ProfilesBlood samples were harvested from the bovine jugular vein at 14:00 h on the 3rd day (adaptation period) and on the 7th day of the experiment (heat stress period). For blood hematology, analyses of white blood cell (WBC), lymphocyte (LYM), monocyte (MON), granulocyte (GRA), red blood cell (RBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular volume (MCV), red cell distribution width (RDWc), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), platelet count (PLT), mean platelet volume (MPV), plateletcrit (PCT), and platelet distribution width (PDWc) were conducted in ethylene-diamine-tetra-acetic acid (EDTA)-treated vacutainer (Becton-Dickinson, Franklin Lakes, NJ, USA) tubes, using an HM2 (VetScan HM2 Hematology System) machine. For the analysis of blood metabolites, the extracted blood samples from jugular venipuncture were transferred into non-heparinized vacutainers (BD Vacutainer, Plymouth, UK). Serum samples were obtained from blood after centrifugation at 2000× g for 15 min at 4 °C. The serum was isolated for analysis of metabolites, including albumin, blood urea nitrogen (BUN), calcium (CA), r-globulin, glucose (GLU), glutamic-oxalacetic transaminase (GOT), r-GT, magnesium (MG), non-esterified fatty acid (NEFA), inorganic phosphorous (IP), and total protein (TP). A total of 500 µL of serum from each sample was separated in a 1.5 mL tube (Eppendorf AG, Hamburg, Germany) and stored at −80 °C in a deep freezer, for further analysis using the analyzer (U9280-0002, Mississauga, NB, Canada). 2.3.5. Stress HormoneThe sample was extracted from the serum tube, separated by centrifuge, and 500 µL of the supernatant was sampled into each 1.5 mL tube (Eppendorf AG, Hamburg, Germany). The serum samples were analyzed using a cortisol ELISA kit (MBS701325, MyBioSource) and a haptoglobin ELISA kit (MBS739905, MyBioSource). The concentration of hormones, including cortisol and haptoglobin, in blood was measured using an analytical ELISA machine (PMT49984, BioTek Instruments Synergy Korea Ltd., Winooksi, VT, USA) measuring absorbance at a wavelength of 450 nm. The inter- and intra-assays of coefficients of variance for cortisol were 10% and 10%, while those for haptoglobin were 10% and 8%, respectively.2.3.6. Physiological ParametersHeart rate (HR, bpm/min), skin temperature (ST, °C), and rectal temperature (RT, °C) were measured on the third and seventh days of each period, at 09:00 h. In order to measure skin temperature, a machine (S60, Caterpillar FLIR camera, Vernon Hills, IL, USA) was used to measure on the back, hips, and mammary glands. Rectal temperature measured using another machine (TES 1300, K Type thermometer, Taipei, Taiwan), which inserted into the rectum and held for 1 min. Heart rate was measured using a stethoscope on the neck side of each individual for 1 min.2.3.7. MicroRNAMilk samples were transferred into five 50 mL tubes containing 0.5M EDTA at pH 8.0 (324503, AMRESCO, Bala Cynwyd, PA, USA), and incubated at 4 °C for 30 min. The samples were then centrifuged at 2700× g for 10 min at 4 °C. Thereby, the supernatant was removed from the skim milk layer. Afterwards, five 50 mL tubes were transferred into one 50 mL tube (12150, Taeshin Bio Science, Korea). After repeating the procedure three times, 1 mL of 1× PBS was added into the 50 mL tube and then filtered using a 200 µm (43-50200-03, Pluristrainer, Germany) membrane filter. Then, the tubes were centrifuged at 2700× g for 15 min at 4 °C. After removing the supernatant, 1× PBS was added and then transferred to a 2 mL tube. The milk sample was recovered and stored at −80 °C until further analysis. The isolation of miRNA from milk was performed using a mirVana™ miRNA isolation kit (AM1560, Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s protocol. qPCR was performed using a TaqMan™ Fast Advanced Master Mix kit (4444557, Thermo Fisher Scientific). Thermal cycling was conducted according to the manufacturer’s recommended protocol, and all experiments were performed in duplicate. The TaqMan microRNA assays used in this study and their IDs were as follows: miR-216b (ID: 002326), miR-92a (ID: 000431).2.3.8. Metabolomics AnalysisExtraction of Serum SampleExtraction of serum samples from blood was conducted by adding 450 µL of cold methanol to 150 µL of serum in an optimal ratio of 1:3. Each serum sample was homogenized (with a frequency of 30) using a mixer mill (Retsch GmbH & Co, Haan, Germany) for 10 min and stored at −20 °C for 1 h. After that, centrifugation of the sample was performed at 13,000 rpm for 10 min at 4 °C. The supernatants were then passed through an additional 0.2 µm PTFE filter and transferred to Eppendorf tubes. The supernatant was completely dried with a speed vacuum machine. The dried serum sample was dissolved by methanol and syringe filtration (0.2 µm), prior to GC-TOF-MS analysis. For the GC-TOF-MS analysis, two-step chemical induction was performed for the sample. First, an oximation process using the GC-MS analysis protocol was carried out with 50 µL of metoxyamine hydrochloride (20 mg/mL in pyridine, 90 min, 30 °C), followed by silylation using 50 µL of N-Methyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA) (30 min, 37 °C). GC-TOF-MS AnalysisThe GC-TOF-MS system was operated according to a previous study [15]. This was performed using an Agilent 7890B GC system (Agilent Technologies, Palo Alto, Santa Clara, CA, USA) and a Leco TOF Pegasus BT mass spectrometry (LECO, St. Joseph, MI, USA). A DB-5MS capillary column (30 m length, 0.25 mm i.d, 0.25 µm film thickness; J &W Scientific, Folsom, CA, USA) was used for helium gas flow of 1.5 mL/min. Afterwards, 1 µL of the sample derivatized using the GC-MS analysis protocol was injected into the split mode (1:10) for analysis. After 2 min of operation in an oven set to 75 °C, the temperature was increased to 300 °C, at a rate of 15 °C/min, and maintained for 3 min. To collect in electron ionization (EI) mode, mass data was collected using an EI method with ionization energy of 70 eV and a mass scan range of 50–600 m/z at an acquisition rate of 20 spectra/s. The ion source temperature and injector were set at 230 °C and 250 °C, respectively.2.4. Statistical Analysis Pre-processing and statistical analysis of metabolomics data were carried out in accordance with a previous study [15]. Raw data collected through the GC-TOF-MS were converted to CDF (NetCDF) files with the LECO CROMA TOF software (version 4.44; LECO Corporation, accessed on 3 July 2020). For alignment, retention time correction, we used the online METALIGN software package (http://www.metalign.nl, accessed on 12 July 2020). Then, the final data were converted into an excel file. The data were placed in a three-dimensional matrix, utilizing information regarding peaks, peak areas, and sample names. The SIMCA-P+ software (version 12.0, Umetrics, Umea, Sweden, accessed on 20 July 2020) was used for multivariate statistical analysis. Unsupervised principal component analysis (PCA) was conducted, in order to investigate the general aggregate status and trends of other groups in all samples. After that, a supervised orthogonal partial least-squares discriminant analysis (OPLS-DA) model was used, in order to identify metabolomics that significantly differed, by maximizing the metabolomics changes between different groups. Variables were differentiated by variable importance in the projection (VIP) values. The statistical analyses were conducted based on the GLM procedure of SAS (Studio Version, SAS Institute Inc., Cary, NC, USA). All data were subjected to student’s t-tests (two-tailed) for comparisons between the means of the different groups. For the traits that were repeatedly measured, such as milk yield and behavioral data, the repeated measure analysis was performed, while the distribution of animals into treatments was considered as a random effect. Initial milk yield, compositions, and blood metabolites of the lactating cows showed no significant differences after the initial statistical analysis. Therefore, the continuous variable (covariate) was removed from the model. Analysis of covariance was utilized for correcting treatment means, controlling the experimental error, and increasing precision. Variance and covariance assumption structures, including AR(1), UN, CS, ANTE(1), TOEPH, ARH(1), and so on, were tested; the covariance structure that resulted in the lowest values for the Akaike information criteria was selected for the final analysis, due to its good fit to our design. Differences between the two subsets of data were considered statistically significant at p-values less than 0.05, while values between 0.05 and 0.10 were considered to indicate a significant trend tendency. 3. Results3.1. Feed and Water IntakeFI was significantly decreased (p < 0.05) by 10 kg/day in HTHH, compared to LTLH. Water intake showed no statistical difference (p > 0.10) between the two groups, but the HTHH group demonstrated a numerical increase of 14 kg/d (Table 2). 3.2. Physiological IndicatorsRectal temperature showed a significant increase (p < 0.05) in the HTHH group, compared to the LTLH group. In addition, heart rate was higher (p < 0.05) in the HTHH group than in the LTLH group (Table 3). 3.3. Blood Hematology and Metabolite ProfileUnder HTHH treatment, blood RBC was tended to increase (p = 0.072). In contrast, the treatments did not influence WBC, LYM, MON, GRA, HGB, HCT, MCV, RDWc, MCH, MCHC, PLT, MPV, PLT, and PDWc in the blood (p > 0.10, Table 4).Analysis of the blood metabolite profile in early lactating Holstein cows showed no effects of THI levels between the LTLH and HTHH groups, where GLU, NEFA, BUN, TP, albumin, r-globulin, CA, IP, MG, CHO, and GOT did not differ between the two groups (p > 0.10, Table 5).3.4. Milk Yield and Compositions3.4.1. Milk YieldMilk production was about 10 kg/day lower (p < 0.05) in the HTHH group, compared to that in the LTLH group (Table 6). 3.4.2. Milk CompositionsUnder HTHH treatment, MFY, 3.5% FCM, and ECM all decreased (p < 0.05); however, MUN and MPY only tended to decrease (p > 0.05). In addition, there were no differences (p > 0.10) in protein, fat, lactose, SNF, somatic cells, acetone, BHB, beta-casein, MUFA, PUFA, SFA, and TFA between the two groups (Table 6).3.5. Stress HormonesWe examined the stress hormone changes in the blood, including cortisol and haptoglobin levels, both of which were increased (p < 0.05) in the HTHH group, compared to the LTLH group (Table 7). 3.6. MicroRNA Gene ExpressionThe abundance of miR-216 gene expression was affected by low- and high-temperature and humidity, being higher in the HTHH group (p < 0.05) than in the LTLH group (Figure 1). 3.7. MetabolomicsMetabolic raw data of sera were collected for a multivariate statistical analysis. The status of metabolites in the LTLH or HTHH groups was determined in the score plot for PCA from serum (Figure 2). The distribution of metabolite parameters in serum (R2X = 0.359, R2Y = 1.000, Q2 = 0.803, p = 0.0105) showed a clear separation in the OPLS-DA model (Figure 2). According to the OPLS-DA analysis, potential metabolic markers were separated from the serum, based on a value of importance in the projection higher than 1.0 (VIP > 1.0). As a result, valine, methionine, phenylalanine, tyrosine, tryptophan, lactic acid, 3-phenylpropionic acid, 1,5-anhydro-D-sorbitol, myo-inositol, and urea were found to be decreased, while fructose and 1-monostearin were increased significantly (p < 0.05) in the HTHH group, compared with that in the LTLH group (Figure 3). For the pathway analysis, the online MetPa system (METABOANALYST 4.0, http://www.metaboanalyst.ca/, accessed on 16 January 2021) was used. Metabolites with significant changes were introduced into this online system, in an attempt to generate the metabolome view list. Bos taurus was selected for pathway analysis in the model organism interface. Targets were selected based on both impact value (not below 0.1) and p-value (no more than 0.05). Figure 4 shows the map regarding relevant metabolic pathways, while the function pathway results are presented in Table 8. As shown in Table 8, phenylalanine, tyrosine, and tryptophan biosynthesis (p < 0.001) and phenylalanine metabolism (p = 0.004) pathways were highly downregulated. 4. DiscussionA decrease in FI is a very usual phenomenon in response to heat stress (HS) [1,4]. Under HS conditions, the amount of energy expended by the cow to maintain homeothermy increases (e.g., 20% more at 35 °C, compared to 20 °C). Panting also increases the maintenance requirement by 7–25% under HS. Therefore, the FI must increase to cover this additional energy cost [16]. However, during HS, FI decreases; this means that the energy status of the cow gets a double hit—greater energy costs to try to maintain homeothermy and lower energy intake [17]. In this regard, it is not surprising that milk production decreases. Indirectly, however, HS-decreased FI provokes consequences in connection with various physiological, metabolic, and blood parameters, in an attempt to lessen the effects of HS by activating homeostatic mechanisms across the body of the animal [5,8,18]. In this study, the mechanism behind the significant decrease in FI can indicate the direct effect of HTHH conditions on animals and the partial inability of cows to dissipate the excess of heat from their body, and thus, less intake will help them to reduce heat production (i.e., metabolic heat production and physically generated heat, known as heat increment) in their body. These results are in alignment with numerous previous studies that have claimed similar phenomena in response to HS [18]. Water intake has a high correlation with FI in animals and humans [1,5]. A normal phenomenon occurring in attempt to deal with HS is that cows tend to intake much more water than under normal conditions, in order to accommodate their body to dissipate heat by evaporation and by alterations in blood circulation. Additionally, under HS conditions, the cow loses water through its skin and respiration, in order to minimize the rise in body temperature [19]. Although HS may cause higher water consumption, on the other hand, decreased FI could also alleviate this water intake. In this study, HTHH did not seem to indicate the influence of FI in increasing water intake; however, water intake was partially induced by HTHH, showing a numerical increase. There is also the possibility that a larger sample size and longer period of HTHH exposure may have allowed us to observe a significant difference in water intake, which should be further studied. Several studies have shown that physiological indicators such as heart rate (HR) and rectal temperature (RT) are the foremost induced by short-term HS exposure in beef calves [8] and dairy cattle [20]. Kim et al. [8] have pinpointed that HR and RT are closely associated indicators, in response to HS, and are the most sensitive markers to be elevated. Therefore, HR and RT are likely to change in animals under HS, which is in agreement with the present results.The body processes stressful information and elicits a response, depending on the degree of stress [21]. The body’s autonomic nervous system is broken down into the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS). In times of stress, the SNS is activated. The SNS is responsible for a cascade of hormonal and physiological responses [21]. The hypothalamus subsequently activates the SNS, and the adrenal glands release a surge of catecholamines, such as epinephrine [21]. This results in effects such as increased heart rate and respiratory rate. As the body continues to perceive stress, the hypothalamus activates the HPA axis. Cortisol is released from the adrenal cortex, which allows the body to continue to stay on high alert. Acutely, catabolic mechanisms of cortisol provide energy to the body [13]. The higher cortisol levels observed in HTHH group were in line with the higher heart rate in the corresponding group in this study. Kim et al. [8] provided a correlation study regarding the relationship between cortisol levels and blood and physiological parameters, including HR and RT. They mentioned that HR is correlated with the concentration of blood cortisol; as such, it has been utilized as an index for the regulation of animal body homeostasis [22]. Furthermore, it has been reported that the rise in blood pressure is associated with increased HR [23]. RT is also an important indicator for the homeostasis regulation of body temperature. The positive correlation between cortisol levels and RT may provide evidence that there exists a metabolic relationship between cortisol levels and RT. A previous study has suggested that the concentration of serum cortisol is a sensitive indicator of HS, and there is a significant correlation between cortisol and RT [24]. Therefore, the significant correlations between HR, RT, and the concentration of serum cortisol can be used to determine HS severity through physiological parameters in high-yielding dairy cows. Given the above review, the reasons behind increased HR and RT in the HTHH group, compared with the LTLH group, can be elucidated. Blood hematology encompasses several immune factors, including WBC, LYM, MON, and GRA, which are expected to be suppressed as a result of stress and an increase in cortisol, due to adverse correlations [25]. HS can cause immunosuppression in ruminants by inhibiting rumination [26], thereby leading to more chances of disease occurrence in the animals [25]. However, in the current study, most of the hematological parameters (except for RBC, which showed a decreased tendency) were not affected by short-term HS exposure. The unaffected parameters could be attributed to either the short time of HS exposure (only 4 day) [27,28] or low sample size in this study, due to natural fluctuations in blood parameters. In line with the present RBC results, a previous study has shown increases in the fraction of RBCs, including erythrocyte number, hematocrit value, and hemoglobin content, in HS group cows [29]. The tendency for higher RBC has been pinpointed, by Kumar and Pachauri [30], to be due to RBC release from the spleen and/or changes in erythrocyte stimulating factor (ESF) release, which is governed by the relationship between the oxygen demand of tissues and the amount of oxygen carried by the blood [30], in an attempt to dissipate heat from the body. In this study, the tendency of increase in RBC was also confirmed through higher serum haptoglobin, which we discuss, in detail, later (with respect to the hormonal effect on the HTHH group). On the contrary, Morar and Hutu [31] have reported that RBC, Hct, and Hb were decreased significantly in dairy cows under HS. Casella et al. [32] revealed the reduction in RBC, Hct, and Hb to be associated with a hemodilution effect due to increased water consumption in an attempt to facilitate evaporative cooling under HS [30]. However, we did not observe a significant increase in water intake in the HTHH group.Blood metabolites may be directly (e.g., by FI [18]) or indirectly influenced by HS, in an attempt to reduce the deleterious effect of HS by activating body hemostasis mechanisms [8,18]. In this study, the serum metabolite profile, including GLU, NEFA, BUN, TP, ALB, r-GLU, Ca, P, MG, CHO, and GOT, showed no significant difference between two groups, LTLH and HTHH. As HS was accompanied with decreased FI, we expected to observe changes in some of the aforementioned profiles. However, contrary to our expectation, there were no changes in any of the metabolites. We could speculate some hypotheses behind this unchanged profile. One hypothesis could be the short application of HS, which may have compromised the ability of the body to challenge, with the first line of defense, against HS; while body lipolysis, gluconeogenesis, and other pathway activation or metabolite changes require time to elaborate [33]. Another hypothesis could be attributed to the low sample size (e.g., due to fluctuations in blood metabolites). It is well-known that blood metabolite alterations in response to stress conditions are subject to fluctuations, particularly when the size of the experimental unit is low [18]. In order to lessen these fluctuations, a larger sample size is recommended, in order to elaborate the variation effects within the data. In this way, we could possibly find significant differences in some blood metabolite profiles in future research. A natural phenomenon in response to HS is a decrease in FI (direct effect), where such a decrease may alternatively cause a decrease in milk yield and some of its characteristics (indirect effects of HS). One study has established that 35% of decrease in milk yield was due to decreased FI, whereas 65% was governed by the direct physiological effect of HS [4,25]. However, decreases in FI can be improved by feeding the cows early in the morning and at night. In these cooler periods of the day, cows can consume up to 80% of their total daily DM intake [16]. However, in cases where the night and morning times still exceed the upper critical THI (of 72)—such as the situation in this study, where cows were exposed to constant 24 h HS—the amount of feed consumed will not compensate for the greatly depressed intake during the day. Beyond THI = 69, each point increase in THI can cause milk reduction of 0.2 kg [1,25]. In other words, for each 1 °C raise in air temperature above the thermal comfort zone, an 0.85 kg decrease in FI occurs, which causes a milk yield decline of approximately 36% [1,4,25]. Given the above discussion, the significant decrease in milk yield of HTHH group in this study could be speculatively explained.HS not only may lessen milk yield, but can also negatively affect milk constituents—particularly in high yielding dairy cows [25]. HS is widely responsible for a decline in milk fat, mainly due to higher concentrate ration and less fiber content or consumption of ration and, consequently, a disruption in fat synthesis in mammary glands due to increased body temperature [34]. The decline in milk protein content [13] may be due to specific down-thermoregulation activity of mammary protein synthesis [25]. These assumptions could explain the decline in milk fat and the tendency for decreased protein content in the HTHH group of this study. A raise in THI compromises the ability of dairy cows to dissipate excessive heat from the bodies [5], resulting in physiological changes such as reduced milk fat and protein contents [13,34]. Lower FI and thus less protein consumption aligned with decreased milk protein can also explain the decreased tendency of MUN. The mechanism may also rely on the lower urease activity in the wall of rumen and disruption in rumination [25,26], due to higher THI in HTHH group. Energy-corrected milk (ECM) determines the amount of energy in the milk, based upon milk fats and proteins, adjusted to 3.5% fat and 3.2% percent protein. Given the definition of ECM, it is obvious that the ECM showed a significant decrease in HTHH group, compared with the LTLH group, due to the lower milk fat and protein. The reasons for other milk constituents to remain unaffected in the HTHH group are unknown. Cortisol is the first hormone to look at in blood, saliva, or hair, when assessing stress situations, such as HS. Circulating cortisol has been shown to be a very sensitive index of heat stress, heralding the onset of poor tolerance of severe climates [5]. A high-temperature environment as a source of stress triggers a series of stress responses of the body [18]. The cortisol level adapts to the adverse environment, which is the evaluated index for the degree of stress and plays an extremely important role in the body [18]. Exposure to HS will shortly induce the production and release of cortisol from adrenal glands into the blood stream, in effect triggering a flood of glucose, which provides an immediate energy source for the body use. Given the above review, and as expected, the higher cortisol concentration in HTHH group was the result of short-term HS exposure in the corresponding animals. This result is consistent with numerous other studies, whether animals were exposed to short- or long-term HS [35,36]. It is worth noting that, as cortisol is one of primary responses of body to HS, it is not surprising to see its increase in less than an hour of stress exposure, and thus, cortisol increases can reflect acute stress conditions, which was observed under the conditions of this study. Haptoglobin is an acute phase protein produced by the liver, which the body uses to clear free hemoglobin (found outside of red blood cells) from circulation [8,22,35]. In other words, haptoglobin is a hemoglobin-binding protein which prevents oxidative damage by utilizing free hemoglobin and is integral in the formation of the haptoglobin–hemoglobin complex [36]. Hemoglobin is the iron-containing protein complex that transports oxygen throughout the body. It is normally found within red blood cells (RBCs) [35]. Haptoglobin binds to free hemoglobin in the blood. This forms a haptoglobin–hemoglobin complex, which is rapidly cleared out of circulation by the liver such that it can be broken down and the iron recycled. When an increased number of RBCs are damaged and/or break apart (hemolysis), they release their hemoglobin into the blood, increasing the amount of free hemoglobin in circulation, which is consistent with the obtained result of increased tendency of RBC, as hemoglobin carriers, in this study. On the other hand, when large numbers of RBCs are destroyed, haptoglobin levels in the blood will temporarily decrease, as the haptoglobin is used up faster than the liver can produce it [22,35]. A decrease in the amount of haptoglobin may be a sign of a condition that is causing red blood cells to be destroyed or to break apart [35,36], which was not the case in the present study. When the binding capacity of haptoglobin is exceeded, the free hemoglobin level in circulation goes up, which may cause tissue damage and/or organ dysfunction due to oxidative stress by free hemoglobin. In agreement with the presented results, an earlier study in beef calves showed higher haptoglobin supported by higher cortisol concentrations in blood [8]; in addition, another study [24] has reported an increase in serum haptoglobin in response to physical stress in cattle.Thau et al. [21] explained the mechanism behind higher gene expression due to stress conditions as follows: Steroid hormones, such as cortisol, are primary messengers. They can cross the cytoplasmic membrane due to their fat-soluble properties. Cell membranes are composed of phospholipid bilayers, which prevent fat-insoluble molecules from passing through. Once cortisol passes through the cell membrane and enters into the cell, it binds to specific receptors in the cytoplasm. In the absence of cortisol, the glucocorticoid receptor binds to a heat shock protein (HSP) 90 chaperone protein in the cytosol. The binding of cortisol to the glucocorticoid receptor dissociates HSP90. The cortisol–receptor complex then enters the nucleus of the cell and affects gene transcription. In addition, MicroRNAs (miRNAs) are small single-stranded non-coding RNA, which repress post-transcriptional gene expression that can be altered by cortisol via targeting HSPs to modulate HS responses in dairy cattle [37]. Kumar et al. [38] reported increased miRNA expression due to summer HS in Tharparkar and Sahiwal cattle. Subsequently, Kishore et al. [39] reported higher expression of HSP40 transcript in Holstein Friesian, compared to Sahiwal cows, during summer HS [37]. After a short-term HS (2 h) exposure, Shandilya et al. [40] also reported induced mRNA expression of HSP40 and HPS70 in fibroblasts of zebu cattle. Taken together, the increase in miRNA expression in the HTHH group, compared to the LTLH group, in this study can be explained.In our study, serum metabolic pathway analysis showed that the phenylalanine, tyrosine, and tryptophan biosynthesis and phenylalanine metabolism pathways were downregulated. Phenylalanine, tyrosine, and tryptophan are aromatic amino acids (AAA) which belong to the α-amino acid family for protein synthesis [41]. Phenylalanine, tyrosine, and tryptophan have been reported to play regulatory roles under heat stress, through their co-expression network [41]. In addition, AAAs play the role of precursors for numerous metabolomics related to protecting against stress, including melatonin, alkaloids, auxin, and phenolic compounds [41]. AAAs play important roles in the metabolic processes of microflora in all animal bodies. The AAAs and metabolites derived from them also play integral roles in the health of animals [41]. TAT1 is a T-system AA transporter, which plays an essential role in transporting AAAs. This transporter has been reported to be reduced in the chest and ileum of chickens affected by heat stress of 35 °C [42]. Phenylalanine had reduced levels in the brain and liver of chickens, as well as embryos, after exposure to heat stress of 38 °C. Phenylalanine is converted to a phenylamine neurotransmitter in response to the PLP-dependent aromatic enzyme decarboxylase [41]. In addition, it is often converted to tyrosine in the animal body, which is synthesized for epinephrine, dopamine, and norepinephrine neurotransmitters. Phenylalanine is also in charge of the biosynthesis of bacterial cell walls for inhibiting mureidomycins [43] and antibodies classes [44]. Tyrosine is changed to p-hydroxyphenylacetic acid by a mixture of bacteria and protozoa cultures, then converted to p-cresol [41], which plays an important role in the production of antioxidants [41]. Tyrosine is the precursor of the catecholamine neurotransmitters—dopamine and norepinephrine—which can administrate the behavioral, physiological, and neurochemical consequences, under a cold or heat stress environment, by adjusting the release of norepinephrine, thus demonstrating the role of tyrosine in protecting against the adverse effects of heat or cold stress [45]. In addition, when tyrosine is activated to its thiol ester form, it can be attached to the enzyme modular thiosteraease enzyme for use in antibiotic synthesis [41]. Tryptophan is the precursor for synthesis of serotonin, tryptamine, the neurohormone melatonin, and enzyme cofactors, which act as neurotransmitters [41]. A study in steers has shown that tryptophan supplementation can slowly increase in RT in response to acute heat stress through an increase in brain 5-HT, followed by a presumable increase in evaporative heat loss from the skin surface in cattle [46]. Among others, serotonin is involved in the melatonin synthesis metabolic process, which regulates growth activities in response to various biological stresses, such as pathogens, environmental toxins, and extreme temperature [46]. In a previous study, myo-inositol has been reported to be involved in glucose uptake and insulin signaling regulation, as well as adipogenesis regulation [9]. In the northern drosophila fly, high levels of myo-inositol were observed under a cold environment, which decreased in a warm or hot environment [47]. In our current study, we also found the myo-inositol level to be significantly lower in the HTHH group; however, we did not find any significant change in the pathway analysis. The accumulation of glycine betaine can reduce the effect of HS and improve productivity in lactating dairy cows [48]. A previous study has suggested that methionine supplementation can protect proteins from degradation by upregulating genes related to protein synthesis and decreasing genes related to protein breakdown [49]. In our current study, we observed the downregulation of serum glycine and methionine, which indicated that the heat stress in the HTHH group may have adverse effects in early lactating cows. However, from the pathway analysis, we did not find any change related to phenylalanine metabolism, inositol phosphate metabolism, glyoxylate and dicarboxylate metabolism, cysteine and methionine metabolism, glycine, serine and threonine metabolism, and tryptophan metabolism, which implies that these parameters had limited adjustment in the pathway analysis. As per the aforementioned review, the downregulation of these two pathways and their final impact metabolomics resulted in negative modulation in immune parameters and biological polymers (e.g., proteins, muscle cells, and so on) due to HTHH conditions, even when only exposed in the short-term (4 d). Thus, we postulate the importance of the negative association observed between HTHH and the final products of the resulting pathways in this study.5. ConclusionsIn this study, we successfully carried out metabolomics analyses in an attempt to characterize the influence of short-term HS on early lactating cows. According to the metabolic candidates, strong relationships with HTHH conditions were detected through the observed changes in the mentioned metabolic pathways. The results are worth further large scale investigation, in order to identify potential biological markers that can be used to accurately monitor HS conditions, as well as to develop a basis to further explain the physiological mechanisms underlying the metabolic pathway changes induced by HS.

animals : an open access journal from mdpi

10.3390/ani13091554

PMC10177112

I trained an American black bear in human care to choose different response buttons when presented with an image of either a highly preferred or a less preferred food item. The bear learned to choose the appropriate response button when presented with the preferred food item at above chance levels and differentiated between the use of the buttons appropriately. However, she did not reach a high level of performance with the less preferred food item even after over 1000 trials, suggesting that performing a conditional discrimination on the basis of preferences may be challenging for black bears. However, the work presented here represents the first attempt to train a bear to indicate her relative preferences using something like a Likert scale commonly used with humans to indicate their preferences and could be a valuable welfare tool for animals in human care. Similar work with gorillas suggests that bears are as capable as great apes in learning such tasks and would also benefit from this type of technical enrichment.

A preference scale for use by nonhuman animals would allow them to communicate their degree of liking for individual items rather than just relative preferences between pairs of items. It would also allow animals to report liking for images of objects that would be difficult to directly interact with (e.g., potential mates and habitat modifications). Such scales can easily be presented using touchscreen technology. Few zoos have used touchscreen technology for species other than nonhuman primates. I present a description of efforts taken to create such a scale for use with a single zoo-housed American black bear (Ursus americanus). Although the bear did not reach a high level of proficiency with assigning preferred and non-preferred food items to categorical responses of “like” and “dislike,” she was able to learn how to use the like and dislike buttons differentially for a single preferred and less preferred food item and she selected the correct response button for the preferred item at above chance levels. These data contribute to our limited understanding of black bear cognition and suggest that conditional discriminations may be difficult for black bears. This finding can inform continued efforts to create a simpler tool for nonhumans to communicate their preferences to human caregivers in a more nuanced way than is currently possible. More generally, the current study contributes to the growing body of work supporting the use of touchscreen technology for providing enrichment to less studied species like bears.

1. IntroductionWhile touchscreens are becoming increasingly common for enrichment or research purposes in zoo-housed nonhuman primates, the number of other species provided with this level of technical enrichment remains extremely small. Bears are widely recognized to be highly intelligent and curious animals that could benefit from more complex and dynamic enrichment. Although there is a paucity of work describing bears’ visual abilities, early work suggested that black bears discriminated various hues from grey, having difficulty with only red–green discriminations [1]. In addition, bears have been successfully trained to make categorical discriminations between stimuli presented on touchscreens (e.g., brown bears, Ursus arctos, Bernstein-Kurtycz et al., personal communication; American black bears, [2,3,4,5,6]; Malayan sun bears, Helarctos malayanus [7], and polar bears, Ursus maritimus (Jeremiasse et al., personal communication). Not only do bears appear to enjoy the stimulation provided from interacting with trainers through touchscreen training, the use of the computer presents researchers with novel ways to communicate with the bears. Computers have often been used in zoological settings to provide enrichment—most typically for nonhuman primates [8,9,10] but also for other species like parrots [11]—in the form of games, puzzles, or auditory enrichment. Computer interfaces have also been used to conduct assessments of animal well-being [9,12]. Computer interfaces can also be used to present images of foods and other objects, which subjects can then indicate their preferences for. For example, researchers recently presented a tablet to a Goffin’s cockatoo (Cacatua goffiana) so that the bird could select symbols representing various items, activities, or interactions. Their results suggested that the single cockatoo subject could use the tablet effectively to request objects and interactions that presumably had positive effects on her well-being [13]. The current study aimed to provide a means for a bear to symbolically communicate preferences for the first time.Understanding individual preferences is critical for optimizing an animal’s environment and ensuring positive welfare. Preferences can inform habitat planning, husbandry, enrichment, and food provisioning [14]. Traditionally, preferences have been assessed indirectly by measuring degree of engagement with different enrichment items, foods, and environmental features etc. (e.g., [15,16]); objects that trigger different events like sounds [10], approach, and avoidance behavior (e.g., [17]); or efforts exerted to obtain access to space, social companions, or objects [18,19,20]. Preferences have been assessed more directly with forced choice tests between pairs or groups of real objects (e.g., [21,22,23,24,25]) or choices of symbols representing options like sounds [26]. These methods assess relative preferences among pairs or groups of choices, but fail to provide a more nuanced assessment of amount of liking (e.g., this object is liked to some degree compared to this item that is liked very much). Importantly, only objects that can be safely presented for investigation can be used in assessments involving real objects. In addition, paired-choice tests require the repetition of multiple pairings across items, which can be time consuming and can result in satiation when assessing preferences for foods or rewarding the individual for their choices. I attempted to develop a novel method to assess the degree of liking for various elements of the environment presented in pictorial form in an American black bear. This scale would ultimately allow her, and other nonhumans, to indicate preferences for food, enrichment, care staff, environments, sounds, and other stimuli that are not physically present at the time of assessment. Ultimately, I wished to be able to assess preferences for unfamiliar and previously inexperienced stimuli, such as planned habitat changes, possible mates, or even images representing more abstract concepts such as natural environments. Notably, preferences can be assessed in a single trial using this method once the animal understands the meaning of the end-points. I began training the bear to use such a nonverbal animal preference scale (NAPS) using images of foods for which her relative preferences could be determined.In humans, preference scales are commonly encountered in product assessment, customer satisfaction surveys, and research into attitudes, beliefs, and personality traits. Such measures typically take the form of Likert scales [27], which allow respondents to indicate relative preference for items or agreement with ideas. One of the advantages of this type of scale is that respondents are able to indicate when they do not like an item at all rather than being forced to choose between equally preferred or non-preferred items. Rather, items are presented one at a time with a rating scale that has end-points representing a spectrum of agreement (e.g., from “strongly dislike” to “strongly like”). Using paired-choice tasks, an item might never be selected because it is less preferred than the other options, but it would not be possible to determine whether this item may also be liked rather than disliked. Preference scales have been widely adopted for research in multiple disciplines due to their flexibility [28]. Although there has been only one other known attempt to use such a scale with nonhumans, which I conducted concurrently with a bachelor group of gorillas (Gorilla gorilla gorilla, [29]), nonverbal versions have been used with human children [30,31] and clinical patients [32]. In these cases, verbal scale endpoints are replaced with intuitive images such as facial expressions to denote degree of liking [33], level of pain [34,35], and mood [36]. Therefore, although admittedly more complex and abstract relative to existing methods for assessing preferences, it seemed desirable and feasible to adopt similar methods to train nonhumans to use such a scale.It should be noted that even pictorial Likert scales require verbal instruction, and a recent meta-analysis reveals that children below the age of five years cannot reliably use self-report measures of health outcomes. Furthermore, children below eight years of age may not be able to use a scale with more than two response options [37]. Therefore, training nonhumans to understand the construct of a sliding scale of preferences posed several challenges, not the least of which was deciding upon scale end points that might intuitively reflect “dislike” and “like.” Training animals to understand task requirements without verbal instructions is not a novel challenge, but does necessitate a prolonged period of training prior to administering the test in contrast to the type of one-off assessments conducted with human respondents. Second, because constructs of liking or preferences might rely on explicit self-knowledge, use of a preference scale may depend upon a degree of abstraction beyond the grasp of most nonhuman animals. There is no existing work that suggests that nonhumans can accurately report on their own (or others’) preferences when directly asked, and indeed, it has been noted that it would be difficult to learn signals of others’ preferences when they are discordant from our own, especially inconsistently so [38]. However, as a starting point, effective use of the scale could be acquired through a simpler process of association between items that evoke a particular visceral response (e.g., disgust, excitement) and different operational responses (use of the different response buttons). A process of generalization might support the appropriate use of response buttons associated with negative and positive feelings toward novel stimuli. Thus, use of the NAPS could be assumed to measure relative preferences for categories or objects that can be represented physically regardless of whether the subject explicitly represents the items as “things I dislike and things I like”.Although stimuli to be rated could be presented in any modality perceived by the organism, use of a touchscreen system is most suitable for visual or auditory stimuli. Successful implementation of a visual NAPS requires that subjects understand the correspondence between pictures and their real-life referents. Many species have demonstrated picture–object correspondence (for review see [39]; e.g., in pigeons, columbidae [40]; in kea, Nestor notabilis [41]; macaques, Macaca silenus [42]), including the black bear that is the subject of the current study [2]. This apparently widespread ability supports the computer touchscreen methodology used here. However, another challenge with the NAPS is that stimuli, both in training and testing, must be subject-specific. To train subjects to use the NAPS, it is necessary to train them to understand what the different response buttons represent using stimuli for which the researchers already know the subject’s preference. These buttons must be presented at the extreme ends of a spectrum (i.e., spatially) so that responses representing intermediate levels of preference can be added later to allow a more nuanced scale of preference. Once appropriate use of the most extreme response buttons is established for items for which preferences are known, researchers can introduce the use of intermediate buttons, and finally, begin assessment of preferences for novel items. Here, I presented the bear with images of food items based on her preferences as indicated by her care staff to train her on the use of the scale.I conducted a simple validation of the food preferences indicated by the care staff by presenting the bear with a set of images of preferred versus less-preferred food items on a touchscreen in a two-alternative forced-choice task. As with gorillas tested previously [29], it was expected that the bear would spontaneously select images of the preferred items. As expected, the bear selected the images of preferred over less preferred foods at above chance levels even when items belonging to the preferred and less preferred categories were continuously changed, which was done to ensure the generalizability of the concept. Others have provided a similar validation of the use of pictorial stimuli to assess food preferences in other species with many of these subjects showing generalization of choices to novel food images within the same categories of preference (e.g., sloth bears, Melursus ursinus, [43]; lion-tailed macaques, Macaca silenus, [42]; gorillas, [18,44,45]; Japanese macaques, Macaca fuscata, and chimpanzees, Pan troglodytes, [45]). Early work with black bears showing their stable food preferences [15] and my own previous work with black bears making natural category discriminations using a touch-screen (e.g., [4,6,46]) made me optimistic that this new black bear subject would become proficient in communicating her preferences for items presented visually using the NAPS. Furthermore, there is some existing evidence that great apes—at least those that have received some symbolic/language training—can appropriately use symbols representing “bad” and “good” [47], and can use pictures to communicate desires [48] and my previous work with bears and apes suggested that they were capable of representing similar levels of abstraction compared to great apes [3,4,5,6,46,49,50]. Ultimately, I wished to present the task with a 5-point scale, but given the difficulties of gorillas trained previously [29] with the use of a neutral response button, I began training the bear with only the two extreme (non-preferred and preferred) response buttons. Had she demonstrated proficiency with these two end-points, I would have gradually added in additional response buttons along the spatial continuum.2. General MethodThe studies reported here were approved by the IACUC of Oakland University (Protocol #12082) and the Animal Welfare and Management Committee of the Detroit Zoological Society.2.1. SubjectOne wild-born female American black bear (Migwan, Basel, Switzerland), age 11 years at the beginning of the study, participated in this study when she resided at the Detroit Zoo, Royal Oak, MI, USA. Migwan was rescued from the wild at a very young age and rehabilitated due to injuries. She was housed individually. Although experimentally naïve, Migwan had participated in husbandry training. For example, she was target trained using positive reinforcement, including clicker training.2.2. MaterialsAll experiments were programmed in Real Practice or Inquisit 3.0 (millisecond.com) and presented on a Panasonic CF-19 Toughbook or an Asus Aspire One Laptop projected to a 19” VarTech Armorall capacitive touch-screen monitor. The touchscreen monitor was affixed to the front of a rolling LCD cart. The touchscreen monitor was secured flush to the front of the steel mesh with bungee cords to secure the screen in place so that Migwan could touch the screen with her tongue through the gaps in the mesh. The care staff member and researcher always tested the touchscreen from the bear’s side of the mesh prior to letting the bear into the indoor testing habitat. The laptop sat on a shelf on the cart behind the touchscreen (Figure 1). The experimenter stood against the back wall of the indoor area behind the cart and did not interact with the bear during trials. The care staff member placed the food rewards into a PVC tube affixed to the steel mesh to deliver food rewards for correct responses without any direct contact. This staff member always stood to the same side of the touchscreen during trials and did not direct attention to the bear or the laptop.Stimuli used in the experiments were non-copyrighted photographs downloaded from various websites or images drawn in Microsoft Paint. These stimuli included images of various foods, and two-dimensional shapes such as circles, squares, triangles, and misshapen objects drawn in blue, yellow, red, and green. Food items used to reward correct responses composed a minimal proportion of the bear’s daily diet (e.g., almonds, biscuits, raisins, grapes).2.3. General ProcedureThe research took place in a non-public area of Migwan’s indoor habitat. She participated in testing three afternoons a week at around 13:00 h between April to September in 2014 and 2016. Migwan did not participate in testing from October to March as she was in a state of torpor during the colder months. During the spring and summer months of 2015, Migwan participated in other tasks including a picture–object correspondence test [2] and an ambiguous cue affective bias task [3]. She also participated in a novel judgement bias task that was conducted simultaneously from April to September 2016 [5]. Testing took about 10–15 min each test day, and Migwan completed 4–5 sessions of testing each day. Participation in the tasks was entirely voluntary. Testing for the day ended when Migwan had consumed an appropriate number of rewards as determined by the care staff. A flowchart of the experimental phases is presented in Figure 2.If Migwan selected the correct stimulus, a pleasant auditory beep was emitted, the touchscreen turned white, and the care staff member assisting with the trials placed a small food reward down a PVC chute affixed to the mesh. If Migwan selected an incorrect stimulus, there was no audio feedback, the touchscreen turned black, and there was a 500 ms inter-trial interval.2.4. Phase 1 TrainingMigwan had already been trained to target and to station by her caretakers using positive reinforcement. Prior to beginning the study, she was trained by her care staff to station in front of the touchscreen without it being turned on. She was rewarded for targeting to a familiar target by touching it with her nose. Once she was reliably touching the target positioned right in front of the screen, the care staff removed the target and rewarded Migwan for touching the blank screen with her nose. This training took a period of approximately one week.To train Migwan to use the touchscreen, I first presented her with a two-alternative forced-choice task where she was presented with two stimuli drawn in Microsoft Paint: a yellow square on a white background and a blue circle on a black background. She was reinforced for selecting the blue circle and not reinforced for selecting the yellow square. The idea was to create a positive association with the blue circle and not with the yellow square so that these would be intuitive response buttons for the end-points of the scale, with the blue circle representing “like or preferred” and the yellow square representing “dislike or less preferred.” The two stimuli filled most of the screen. The response button covered 80% of the stimulus so that Migwan had to touch the center of the stimulus and could not activate it just by nudging the edge of the stimulus. She was reinforced only if she used her tongue or nose to contact the touchscreen, not for using her paw. Migwan participated in, on average, four sessions a day, three days a week between April and July, 2014. The stimuli were presented in 20-trial sessions with the side of the correct stimulus (the blue circle) counterbalanced within the session.In each trial, the stimuli appeared simultaneously and disappeared when one of them was selected. If Migwan selected the blue circle, a tone sounded, the screen turned white, the care staff member placed a food reward in the PVC tube affixed to the mesh, and the next trial commenced after 500 ms. If she selected the yellow circle, there was no sound, the screen turned black and she received no food reward. The inter-trial interval was the same. The criterion was set to four consecutive sessions at 80% correct or better (i.e., 16/20 correct responses) or two consecutive sessions at 90% correct or better (i.e., 18/20 correct responses).2.5. Food Preference AssessmentsTo assess a spontaneous preference for images of preferred foods, I again used a two-alternative forced-choice procedure. Sessions included 20 trials and were identical to the training task described above except for the stimuli used. Migwan completed 39 sessions of this task. On each trial, a food indicated by the care staff to be preferred by Migwan was randomly paired with beets, lettuce, or carrots (on sessions 5 and 6), which were foods identified by care staff as being least preferred by Migwan. An image from the preferred category was randomly paired with an image from the non-preferred category on each trial and presented in random order. Table 1 indicates which food items were presented in each category on each session along with the number of trials on which each food image was presented within a session. One photo was used for each of these food types. Changes in the food items presented were made to test the generalizability of Migwan’s preferences. The side the non-preferred foods were presented on was counterbalanced within sessions with the constraint that they could not appear more than three times consecutively on the same side of the screen. Migwan was rewarded if she selected one of the presumed preferred foods and not if she selected the presumed non-preferred foods. 2.6. Phase 2 Training ContinuationI next presented Migwan with a session of 20 trials of photographs of beets paired with the image of the blue circle (with side counterbalanced), where she was rewarded only for touching the blue circle to reinforce the idea of the blue circle as something positive and the beets as something not positive.I then presented a single 20-trial session where the blue circle was paired with images of the preferred foods. As expected, she performed at chance, choosing the blue circle only 10 times, suggesting that she perceived the blue circle as equally positive, or likely to lead to reward, as the images of the preferred foods.I then presented Migwan with five additional sessions of the blue circle paired with the yellow square to ensure she was still performing at criterion with the training stimuli.2.7. Phase 3 NAPS TrainingI created a computer program in Inquisit v. 3 that presented an image of a less preferred, or a preferred food in the center of the screen that, once touched, prompted the appearance of a response button in the top left (yellow square) or top right (blue circle) of the screen. The food image remained on the screen, centered in the bottom half of the screen once the response buttons appeared (Figure 3). Each response button took up about 30% of the top half of the screen. These sessions consisted of 10 trials (5 with beets and 5 with grapes). Only one image was used to represent each food type (beets for the less preferred item and grapes for the most preferred item) and it was the same image used during the food preference assessments described above. Migwan was trained to associate images of beets with the yellow square response button and images of grapes with the blue circle response button. On each trial, there was only one available response button. When Migwan selected that button, a beep sounded, the screen turned white and the care staff member placed a food reward in the PVC chute. The next trial began once Migwan had touched the image of the food and the subsequent response button with her nose or tongue. Migwan completed 7 sessions of this phase.2.8. Phase 4 NAPS TrainingIn Phase 4 of Training, sessions consisted of 10 trials, which were the same as above except that both response buttons appeared simultaneously on every trial and Migwan was rewarded only if she chose the correct one (Figure 4). Migwan completed six sessions of this phase.2.9. Phase 5 NAPS TrainingBecause I had successfully trained Migwan to associate the blue circle with reward, she was understandably reluctant to choose the yellow square even on trials when it would have been the correct response (i.e., when beets were presented). Therefore, I trained her to select the yellow square when beets were presented in this phase. This phase consisted of 10-trial sessions in which beets were always the food stimulus (always the same image as used previously) and both response buttons were presented simultaneously after she had touched the image of the beets (Figure 4). She was rewarded only for touching the yellow square/dislike button. Criterion was set to 80% correct responding for four consecutive sessions. Migwan received 13 sessions of this phase and then testing went on hiatus for fall torpor.2.10. Testing HiatusWhen I resumed testing in April 2015, I focused instead on a picture–object correspondence task, which validated the use of two-dimensional images to represent objects for Migwan to rate [2]. I also presented her with a judgement bias test to assess affect changes across seasons [3]. In the spring of 2016, I returned to training Migwan in the current task. I trained her in several simpler conditional discrimination tasks using simple shapes (green triangle, red oval) rather than non-preferred and preferred foods to validate her ability to perform a conditional discrimination, before returning to the version of the task involving foods in the fall of 2016. Migwan learned to select a novel grey square response button conditional on being presented with a green triangle and to select a novel purple circle response button in response to being presented with a red oval. It took her 49 sessions (490 trials) to reach criterion and she successfully transferred at above chance levels to different images of the same shapes and colors as the original training stimuli. However, it took her 38 and 46 sessions to reach criterion again with the transfer shape and color stimuli, respectively.2.11. Phase 6 NAPS TrainingHaving established that Migwan could learn this conditional discrimination task with less abstract decision rules, I returned to the task of training her to respond differentially to preferred and less preferred foods almost two years later. I presented Migwan with a version of the NAPS in which carrots were presented as the less preferred food and grapes were presented as the preferred food. I switched from beets to carrots as the less preferred foods to maximize the difference in appearance of the two presented foods as both beets and grapes were of a similar purplish color, and upon the suggestion of her care staff who noted that Migwan no longer preferred carrots relative to other foods from her daily diet. I verified that Migwan did not select carrots until all other foods were selected when presented with a handful of foods from her regular diet in her water trough. I also used the newly trained response buttons so that the dislike button was a gray square and the like button was a purple circle within a black background to mitigate against Migwan’s retained preference for the blue circle as the like button. The locations of the stimuli remained the same with the food appearing in the center of the bottom half of the screen and the dislike button appearing on the top left and the like button appearing on the top right.Each session consisted of 10 trials: 5 in which carrots were presented and 5 in which grapes were presented, in random order. Both response buttons appeared simultaneously on the screen after the food item was selected by Migwan. She was rewarded for selecting the dislike button if carrots were shown and the like button if grapes were shown. Migwan completed 112 sessions of this phase between 4 August and 30 September 2016 before testing was halted. Testing took place three times a week at 13:00 h. Migwan simultaneously participated in a novel test of judgement bias during this time [5]. Migwan moved to another facility in 2017 and could no longer be tested.3. ResultsAnalyses were conducted using SPSS v. 28. Alpha was always set to p = 0.05.3.1. Phase 1 TrainingMigwan reached criterion in 35 sessions (approximately 700 trials with some sessions missing some trials).3.2. Food Preference AssessmentInitially, Migwan had a strong left side bias. She eventually met criterion with two consecutive sessions at 90% correct by session 33 but I continued testing her with additional minor changes to the composition of the food items and she reached criterion again with four consecutive sessions at 90% or better by her 39th session (780 trials). Overall, she chose the preferred foods at levels above chance determined by a one sample Wilcoxon signed rank test (Z = 4.434, p <0.001). Her performance improved across sessions, as can be seen in Figure 5. There was a significant difference in performance between the first and last halves of the testing sessions, Wilcoxon, Z = −3.732, p < 0.001.3.3. Phase 2 Training ContinuationMigwan chose the blue circle 13 times on the single session in which it was paired with beets. In the single session, in which it was paired with preferred food items, she selected it 50% of the time. Across the five sessions in which it was paired with the yellow square, Migwan chose it on 80% or more trials on all but a single session, where she chose it 11 times.3.4. Phases 3 and 4 NAPS TrainingIn Phase 3, only the correct response button appeared on each trial so Migwan was 100% correct on all 7 sessions. In Phase 4, Migwan chose the dislike response button only once across six sessions, so her performance was at chance.3.5. Phase 5 NAPS TrainingOn the first four sessions, Migwan chose the dislike button correctly 50% of the time. However, by the end of 12 sessions, she had met the criterion, responding at 80% or more for four consecutive sessions. She was accidentally given a 13th session, on which she also scored 80% correct. Testing went on hiatus for torpor after this phase.3.6. Phase 6 NAPS TrainingMigwan completed 112 sessions. She responded equally quickly to touch photos of carrots (M = 2672.87, SD = 15,881.30) and grapes (M = 2596.05, SD = 18,051.841, p = 0.93 with a Wilcoxon signed ranks test).Her average performance across all sessions was 56.61% correct, which was significantly above chance, (binomial test, N = 64, p < 0.001). She did not reach criterion; however, she had a run of three sessions at 80% correct between sessions 75 and 77 and she missed meeting criterion by a single trial by the 107th session. When comparing her performance on the first half of sessions (M = 55.36, SD = 1.33) to performance on the last half of sessions (M = 57.86, SD = 1.67), she showed little improvement. A Wilcoxon signed rank test confirmed no significant difference in performance between the first half and last half of sessions, Z = −0.123, p = 0.092. Figure 6 shows her performance across blocks of 4 sessions (40 trials).Migwan’s performance on trials where carrots were shown was not significantly different from chance, M = 0.48, SD = 0.500, binomial p = 0.488, but her performance was significantly above chance when grapes were shown, M = 0.67, SD = 0.471, p < 0.001.I also conducted Chi square tests of independence to test whether the food item presented on that trial was significantly associated with selection of the different response buttons. The likelihood of choosing a particular response button was significantly associated with the food that was shown, X2 = 26.066, p < 0.001. Migwan was more likely to choose the dislike button for carrots and the like button for grapes, as can be seen in Figure 7.To test whether the latencies to respond were a function of the response button chosen (dislike, like, referred to henceforth as “response”) and correctness of the response (henceforth “correct”), I used a generalized linear model (GLM) with a gamma distribution and a log link function. I included response, correct, and their interaction as fixed effects in the model. Response significantly predicted response time, X2 = 34.092, p < 0.001, but correct did not, X2 = 0.139, p = 0.709. However, response interacted with correct to predict response latencies, X2 = 6.918, p = 0.009. The difference in response latencies for correct and incorrect responses was more pronounced if the dislike button was selected. In this case, Migwan was quicker to select dislike when it was the correct response (M = 952.99, SEM = 36.920) compared to when it was the incorrect response (M = 1082.40, SEM = 50.204). She showed the opposite pattern when choosing the like response. With the like response, she was faster to respond incorrectly (M = 1221.11, SEM = 45.816) than correctly (M = 1334.46, SEM = 43.864). These data appear in Figure 8.4. DiscussionI report on the first attempt to train a bear to use symbols to communicate her preferences. Members of many other species, including nonhuman primates [47,51], domestic dogs [52], dolphins [53] parrots [54], and a cockatoo [13] have shown the ability to use symbols to communicate to varying degrees. Despite lofty intentions of training a black bear to use a touchscreen to communicate her preferences for two-dimensional stimuli, the task, which depended upon a conditional discrimination, proved very difficult to train, as it had been for three gorillas trained in parallel [29]. This was somewhat surprising as the bear had previously outperformed the gorillas in two conditional discrimination tasks used to assess judgement bias [3,5,49,50]. Furthermore, conditional discrimination tasks have been mastered by individuals of various species, including pigeons [55], rats [56], octopuses, cuttlefishes [57], squirrel monkeys [58], and chimpanzees [59], so the task should not have been beyond Migwan’s capability. Notably, the previous tasks involved associations between stimuli defined merely by shape and color and different response outcomes, whereas the current study aimed to test associations between broad abstract categories of preferred and less preferred foods. A construct concerning preferences is highly abstract and there is no existing evidence that nonhumans can represent a concept of their own or others’ preferences. Because the preferences of others can inconsistently match or differ from our own, it is likely a challenging construct for nonverbal organisms to represent [38]. However, there is some evidence that at least three language-trained apes appropriately used lexigrams representing “good” and “bad” and applied them in a manner that was appropriate and consistent with their human caretakers’ notions of good and bad behaviors [47]. There is also growing evidence that nonhuman mammals are capable of internally generating hedonic experiences in the absence of an external stimulus [60], making the use of a Likert scale for reporting preferences for symbolically represented aspects of their environment feasible. Furthermore, the current study progressed only to the point of training Migwan to associate a particular response with one preferred and one less preferred food, so should not have been conceptually more abstract than the previous studies. Unfortunately, due to Migwan’s move to another facility, I was unable to continue testing her. I had initially aspired to train Migwan to use a five-point scale indicating a more nuanced sliding scale of preferences for items for which I did not already know her preferences, but I was unable to reach this goal.However, there are some promising data from this project. First, Migwan did learn to perform at above chance levels in the NAPS presenting only a single preferred and less-preferred food item, and she missed our somewhat arbitrary criterion level of performance by only a single trial. Thus, one could conclude that she acquired the discrimination. However, she performed with greater accuracy when presented with an image of the preferred food—grapes. This may not be surprising given that, in this final phase of training, I had replaced the previously trained less preferred beets photo with a photo of carrots. Carrots were less consistently presented as a member of the “less preferred food” category across all of the training presented here. One of the major limitations of this study is that I do not have data from systematic preference tests verifying the care staff’s indication of Migwan’s preferences, although I did conduct informal assessments by presenting multiple food items in the water trough and I observed that Migwan did not eat beets when presented as rewards and did not choose carrots when presented alongside other options.When presented with forced-choice tests of preferred versus non-preferred foods, Migwan did not spontaneously select the images of preferred foods at above chance rates unlike two of three gorillas [44], lion-tailed macaques [42], and two sloth bears [43] tested in food preference assessments with images of foods. However, Migwan did choose the preferred foods at above chance levels across all 39 sessions, and did learn to select them to a criterion of 80% over the course of testing, suggesting that she may have formed categories for “preferred foods” over “less preferred foods.” It is less likely that she merely memorized which food photographs were associated with reward because I changed the food photographs periodically. However, it is true that beets, carrots, and lettuce were the only foods used as non-preferred foods so she may have simply learned not to select those images. I did not have the opportunity to test generalization to other photographs or to other preferred and less preferred foods to verify that she had formed such categories.To further corroborate the conclusion that Migwan could learn to use the NAPS, she showed differential use of the response buttons dependent upon which food item had been presented on that trial, as did one of the gorillas tested in a similar procedure [29]. However, she used the buttons more accurately when the preferred grapes were shown, and did not clearly differentiate her use of the dislike and like buttons when the less preferred carrots were shown. It is possible that our initial training, in which I selectively rewarded Migwan for choosing the “like” button in order to reinforce its association with something positive, biased Migwan to the like button even when I changed its image in the final phase of training (the spatial location remained the same). It should be noted, though, that I did not train the gorillas selectively with the like icons used in their training, and they also struggled to learn this task with two of the three gorillas receiving many more trials in the initial training phase than the 1120 trials Migwan received [29]. Furthermore, although Migwan chose the like button more often, she did not choose it as often when it was incorrect as she did when it was correct. It was also not the case that every error involved inappropriate selection of the like button. Migwan also mistakenly chose the dislike button sometimes when grapes were presented. That she was above chance overall but did not show marked improvement across trials suggests that she had some spontaneous understanding of the task when I resumed testing in the fall of 2016, but did not develop an abstract conceptualization of the conditional discrimination nature of the task.As with the gorillas who also struggled with this version of the task [29], I interrupted training on the NAPS with preferred and less preferred foods to present what I imagine to be a simplified conditional discrimination task using two-dimensional shapes of two colors (a green triangle and a red oval) associated with two new response buttons. Eventually, Migwan met a learning criterion. She met criterion more quickly with shape cues compared to color cues, which was in contrast to the gorillas who matched the colors more accurately. This is interesting because human children show a bias to attend to shape over color whereas chimpanzees tested in the same relational matching task showed the opposite bias and performed better on color matching trials [61]. These differences aside, both Migwan and two of the gorillas were able to learn the conditional discrimination task when presented with arbitrary shapes rather than images of food that were linked to their own preferences, suggesting that the mechanics of the task itself are not beyond their abilities, but that responding on the basis of their own preferences may be too abstract and require too much training to become a practical tool for use with animals in human care. This suggests that other procedures like token exchange [62] or the use of a progressive ratio reward schedule [63,64] to assess motivation to obtain rewards might have greater potential as a tool to assess animal preferences. In particular, although I had ultimately planned to present a 5-point Likert scale with five response buttons, adding more than two response buttons may be too challenging for nonhuman subjects [29], as human children cannot reliably use a 3-point scale until the age of eight years according to a recent meta-analysis [37].Another limitation of the present study was the limited number of images used to represent the categories of preferred and non-preferred foods. Subjects show more robust transfer following training with a large number of exemplars representing categories, although training with multiple exemplars may slow acquisition of a category [65]. Training Migwan that she would not be rewarded for selecting the yellow square that represented the dislike button slowed her acquisition of the NAPS. In future, I would instead use images of items that held differential appeal for the like and dislike buttons. Although Migwan did quickly (i.e., within 120 trials) reach criterion in a task where selecting this button was always correct, she remained slightly biased toward the use of the like button even when the less preferred food image was presented and even when the images for the response buttons were replaced. That she did learn to use the dislike button, and to use it more often when it was rewarded (i.e., when the less preferred carrots were presented) indicates her flexibility in updating prior learned reward contingencies.It is possible that presenting Migwan with other tasks in the intervening periods and breaking from training during torpor may have interfered with her reaching criterion levels of performance in this task. However, Migwan came very close to passing the admittedly somewhat arbitrary criterion. Furthermore, it should be noted that Migwan’s performance was quite exceptional in other tasks presented to her over the same period. In fact, she outperformed gorillas on several similar cognitive tasks [3,5,49,50]. Therefore, her ability to perform accurately was not generally hampered by the presentation of multiple tasks during the same period of testing. In fact, she demonstrated remarkable flexibility in switching between tasks.5. ConclusionsAlthough this particular attempt to develop a NAPS suffered from several limitations, it is my hope that other researchers are inspired to improve on these methods. Developing such a tool would provide a valuable new method in which nonhumans could communicate their degree of liking for various items in a single trial, including for things that cannot be physically presented. The use of a NAPS, once trained, allows an assessment of stability of preferences over time without numerous repetitions of pairs of items over many trials. This may be especially appealing for assessing food preferences when foods cannot be presented repeatedly due to satiety or other factors. Notably, the basic idea for the NAPS can be extended to the presentation of auditory, olfactory, or tactile stimuli and response buttons can be presented in other forms other than touchscreen buttons. Thus, the basic paradigm could be easily modified to suit various species and modalities of presentation. However, use of the touchscreen allows for random presentation of stimulus items and the recording of both responses and latencies to respond. Both measures provided some indication here as to how Migwan was understanding the task. Although the present study was motivated by the desire to develop a novel welfare tool, it also provides some insight into black bear cognition. Bears are still quite understudied with regard to their cognition. Migwan’s performance in this study suggests that she can learn conditional discriminations, but that black bears, similar to other nonhumans, may not represent categories that are defined by unobservable features, such as relative preferences. Understanding such fundamental differences in how humans and nonhumans conceptualize their worlds will allow us to fully appreciate the uniqueness of other intelligent species and improve our abilities to provide them with the most appropriate stimulation while they are in our care.

animals : an open access journal from mdpi

10.3390/ani11092676

PMC8466034

Heat stress during the dry period of dairy cows reduces milk yield in the following lactation. Factors such as altered mammary metabolism could impact yields and alter milk composition, including milk protein. We sought to determine if exposure to dry period heat stress would influence mammary milk protein metabolism during the subsequent lactation. Objectives were to first determine the impact of dry period heat stress on milk protein yields and secondly characterize the amino acid and protein profiles in the mammary tissue, milk, and blood to elucidate potential carry-over impact of dry period heat stress on systems that participate directly in milk protein metabolism (i.e., mTOR). We found that heat stress during the dry period reduces milk yield, protein content, and protein yield in the subsequent lactation. The plasma amino acid profile and mammary amino acid transporters are altered in dry period heat-stressed cows, and mammary mTOR signaling proteins are differentially expressed as well. It appears that dry period heat stress impacts mammary metabolism with consequences on milk yield and protein content. The continuous production of high-quality and -quantity milk is vital as a sustainable source of protein in the face of rising global temperatures.

Dry period heat stress impairs subsequent milk production, but its impact on milk protein content and yield is inconsistent. We hypothesize that dairy cow exposure to dry period heat stress will reduce milk protein synthesis in the next lactation, potentially through modified amino acid (AA) transport and compromised mTOR signaling in the mammary gland. Cows were enrolled into heat-stressed (dry-HT, n = 12) or cooled (dry-CL, n = 12) treatments for a 46-day dry period then cooled after calving. Milk yield and composition and dry matter intake were recorded, and milk, blood, and mammary tissue samples were collected at 14, 42, and 84 days in milk (DIM) to determine free AA concentrations, milk protein fractions, and mammary AA transporter and mTOR pathway gene and protein expression. Dry matter intake did not significantly differ between treatments pre- or postpartum. Compared with dry-CL cows, milk yield was decreased (32.3 vs. 37.7 ± 1.6 kg/day) and milk protein yield and content were reduced in dry-HT cows by 0.18 kg/day and 0.1%. Further, dry-HT cows had higher plasma concentrations of glutamic acid, phenylalanine, and taurine. Gene expression of key AA transporters was upregulated at 14 and 42 DIM in dry-HT cows. Despite minor changes in mTOR pathway gene expression, the protein 4E-BP1 was upregulated in dry-HT cows at 42 DIM whereas Akt and p70 S6K1 were downregulated. These results indicate major mammary metabolic adaptations during lactation after prior exposure to dry period heat stress.

1. IntroductionHigh ambient temperatures and relative humidity negatively impact dairy production, costing the U.S. dairy industry nearly $1 billion annually due to lower milk yield of lactating cows alone [1]. Physiological heat stress occurs when elevated ambient temperature and humidity push an animal past the upper critical temperature (UCT) limit of the thermoneutral zone. To acclimate, animals adapt physiology and behavior to reduce heat production and increase heat loss [2]. Lactating dairy cattle are particularly susceptible to hyperthermia due to high metabolic rates and production demand. Heat stress responses are initiated above skin-surface temperature of 35 °C or a temperature-humidity index (THI) as low as 68 [3,4,5]. Initial behavioral and physiological responses in dairy cattle include reduced feed intake and energy diversion away from production, such as reduction in milk yield or impaired milk component synthesis, including milk protein [6,7,8]. Hyperthermia in lactating dairy cows leads to decreased milk protein synthesis beyond the anticipated losses from reduced dry matter intake alone [8,9,10,11]. The reduction in overall milk protein content is attributed in part to several physiological and molecular systems, including a shift in blood flow away from splanchnic tissues and the mammary gland to peripheral tissues [12,13], reduced availability of total and individual amino acids (AA) for milk protein synthesis [9,14,15], and alterations in cell-signaling activity of pathways regulating milk protein synthesis such as the mechanistic target of rapamycin complex 1 (mTORC1) [10,16]. Indeed, in vitro studies of high incubation temperature on bovine mammary epithelial cells (bMEC) indicate a downregulation of genes involved in AA utilization and protein transcription, increased gene expression of AA transporters, and impaired mTORC1 kinase activity, all possibly leading to a reduction in milk protein synthesis [16,17]. Milk protein fractions are also altered under heat stress [11,18].While dry, non-lactating cows generate less metabolic heat and have a higher UCT to their thermoneutral zone than lactating cows [19], heat stress during the dry period can still negatively impact milk production in the subsequent lactation. Relative to cows cooled with fans and soakers, cows heat-stressed during the dry period produce an average of 3.6 kg/day less milk even when all cows are provided active cooling after calving [20]. Lack of dry cow cooling could cost the dairy industry about $800 million annually due to decreased milk yield in the dam alone [21]. However, this does not account for lost premiums from reduced milk protein percentage, as the impact of dry period heat stress on milk protein fraction and content is unclear. Past studies report conflicting results; cows exposed to dry period heat stress show no difference, slight decreases, or even moderate increases in milk protein content or yield in the next lactation [22,23,24]. The objectives of the present study were to measure alterations in milk protein content and yield after exposure to dry period heat stress and to determine any carry-over effects on milk protein synthesis regulation during lactation, specifically AA transport and mTORC1 signaling. We hypothesized that exposure of dry cows to heat stress would reduce milk protein content and yield across the next lactation by impairing the capacity for milk protein synthesis through altered AA availability and reduced mTORC1 signaling.2. Materials and Methods2.1. Animals and Experimental DesignThis study occurred between May to December 2016 at the University of Florida Dairy Unit (Hague, FL, USA) with a herd of multiparous Holstein cows as previously described by Dado-Senn et al. 2019 [25]. All procedures were approved by the UF Institutional Animal Care and Use Committee (Protocol #201508730). Cows were dried off at 46 days before expected calving date according to Dairy Unit standard operating procedures and randomly assigned to treatments based on previous lactation mature-equivalent milk production and parity (1.7 ± 0.8 vs. 1.8 ± 1.3 lactations for cooled vs. heat-stressed cows, respectively). The two treatments applied for the duration of the dry period were heat stress (dry-HT; n = 12) with access to a shaded, sand-bedded free-stall barn or cooling (dry-CL; n = 12) with access to shade plus water soakers and fans. Dry period treatment occurred from May to September in a subtropical climate. Heat stress was considered achieved when the temperature-humidity index (THI) exceeded 68 [4]. Barn ambient temperature and relative humidity were recorded every 15 min across the dry period with Hobo Pro series Temp probes (Onset Computer Corp., Pocasset, MA, USA) to calculate THI [26,27]. The THI during the dry period averaged 76.1 ± 3.7 over the duration of the dry period treatment. During the dry period, cows were fed a standard total mixed ration (TMR), and individual dry matter intake (DMI) was recorded using a Calan gate system. After calving, during the entire lactation all cows were fed a lactating cow TMR and housed as a group in shaded, free-stall barns with free access to water soakers and fans (i.e., actively cooled), where THI averaged 64.4 ± 7.9. Postpartum DMI was measured up to 42 days in milk (DIM). Cows were milked twice daily according to standard operating procedure. 2.2. Physiological Measures and Milk YieldDuring the dry period, respiration rate (RR) was measured thrice weekly at 14:00 h by counting flank movements per minute and rectal temperature (RT) was recorded twice daily at 07:30 h and 14:30 h. Post-calving RR and RT were recorded at 14:00 h every 7 days until 84 DIM. Daily milk yield, component concentration (i.e., fat, protein, fat:protein ratio, and lactose), and component yield were obtained from AfiFarm Dairy Herd Management Software using milk meters and near-infrared spectroscopy, respectively (Afimilk Ltd., Kibbutz, Afikim, Israel) up to 210 DIM. The range for accuracy for this system is between 2–6% for fat, 2–5% for protein, and no range given for lactose [28]. Colostrum yield and composition were also measured; colostrum was considered the milk collected at the first milking after calving only (i.e., 0 DIM).2.3. Milk and Blood Profile AnalysisAfter calving, milk samples (from n = 6 cows per treatment) were collected at 14, 42, and 84 DIM at approximately 11:00 h using standard DeLaval milk collection kits. Milk was stored at −20 °C until analysis. Samples were shipped to the University of South Dakota Dairy Manufacturing laboratory to determine milk protein profile percentages using high-performance liquid chromatography according to standard operating procedures of the laboratory [29]. Blood samples (from n = 5 cows per treatment) were collected at 14, 42, and 84 DIM at approximately 15:00 h from coccygeal vessels into sodium-heparinized vacutainers (BectonDickinson and Co., Franklin Lakes, NJ, USA). After collection, blood samples were promptly placed on ice and centrifuged at 3000× g for 20 min within 1 h after collection. After centrifugation, plasma samples were frozen at −20 °C until analysis. Samples were analyzed by the Experiment Station Chemical Laboratories at the University of Missouri for free amino acid analysis using cation-exchange chromatography (cIEC-HPLC) coupled with post-column ninhydrin derivatization and quantitation [30]. All subsampling was collected from a subset of the n = 12 cows per treatment used to measure physiological measures and milk yield. 2.4. Mammary Biopsies Mammary biopsies were collected on 14, 42, and 84 DIM from a subset of cows that remained the same across sampling (from n = 6 cows per treatment), according to the methods of Farr et al. with modifications previously described in Dado-Senn et al. 2019 [25,31]. After sedation and sterilization, a stainless-steel biopsy tool attached to a drill was inserted into an incision in the mammary gland to cut a core of parenchymal tissue. The tissue was immediately washed in sterile saline, trimmed, and sectioned, and stored in RNAlater at −20 °C or flash frozen and stored at −80 °C. 2.5. RNA Isolation and qRT-PCRTotal RNA was extracted from 49.7 ± 5.9 mg of mammary tissue (n = 6 per treatment per time) using the RNeasy Mini Kit (catalog #74104, Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. RNA concentration was determined on Qubit 2.0 Fluorometer (ThermoFisher, Invitrogen, Grand Island, NY, USA). RNA purity (A260/A280) for all samples averaged 1.9 ± 0.5. A total of 1 μg RNA from each sample was used to synthesize cDNA using the iScript cDNA synthesis kit (Bio-Rad Laboratories, CA, USA) and diluted 1:5 in ddH2O. Gene expression was measured by quantitative real-time PCR (qRT-PCR) with the CFX96 Touch Real-Time PCR Detection System (Bio-Rad). Genes selected for analysis included known bovine mammary gland amino acid transporter genes solute carrier (SLC) 1A1, 1A5, 3A2, 7A1, 7A5, and 36A1 and bovine mTOR pathway target genes unc-51 like autophagy activating kinase 1 (ULK1), ribosomal protein S6 kinase beta-1 (p70 S6K), ribosomal protein S6 (rpS6), protein kinase B (Akt), and eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1). Reaction mixtures were completed as previously described [32] and cycling conditions were: 1 cycle for 3 min at 95 °C, 50 cycles of 10 s at 95 °C, and 30 s at 60 °C, followed by melt curve measurement from 65 °C to 95 °C in 0.5° increments for 5 s. Positive and negative non-template controls were added to each PCR plate. Each sample was assessed in duplicate and the %CV between the duplicates was <2%. Genes were validated before use (Supplementary Figure S1). Primer sequences for the validated genes were obtained from the literature [33] or specifically designed to span exon-exon junctions to minimize the potential of amplifying genomic DNA using Primer3 software (Supplementary Table S1). Either the geometric mean between three housekeeping genes (ribosomal protein S9, RPS9, ubiquitously expressed prefoldin-like chaperone, UXT, and eukaryotic initiation factor 4F, EIF4; AA transporter genes) or UXT alone (mTOR genes), selected based on literature review [34] were used to calculate the relative gene expression using the method 2-ΔΔCt, with dry-CL as the reference group [35]. Only UXT was used for mTOR genes due to similarity of RPS9 and EIF4 to mTOR genes measured. 2.6. Protein Extraction and Western BlottingFrozen tissues were lysed using a Mini-Beadbeater-24 (BioSpec Products, Bartlesville, OK, USA) and 1 mm glass beads in RIPA buffer containing 50 mM Hepes, 40 mM NaCl, 2 mM EDTA, 1.5 mM Na3VO4, 50 mM NaF, 10 mM Na4P2O7, and 10 mM C3H7Na2O6P, supplemented with Halt™ Proteases and Phosphatases Inhibitor Cocktail (#1861282, Thermo Scientific™, Waltham, MA, USA). Proteins were isolated by centrifugation for 15 min at 18,000× g. Lysate protein concentration was determined by bicinchoninic acid assay (#71285, Millipore Sigma, Darmstadt, Germany) and standardized to 1.5 mg/mL in sodium dodecylsulfate (SDS) sample buffer (Laemmli, Bio-Rad #161-0747). Thirty micrograms of protein were denatured at 95 °C for 10 min, and separated by SDS polyacrylamide gel electrophoresis on either 16% or 8% Novex™ Tris-glycine gels (ThermoFisher Scientific, Waltham, MA, USA) for 35 min at 200 V. Proteins were transferred to nitrocellulose membranes (60 min at 20 V) and membranes blocked with Odyssey® Blocking Buffer (LI-COR Biosciences, Lincoln, NE, USA) for 60 min. Proteins were probed overnight against target primary antibodies from Cell Signaling Technology (Danvers, MA, USA): 4E-BP1 (#9644), Actin (#4970), Akt (#4060), and p70 S6K1 (#2708). All primary antibodies were diluted 1:1000 in blocking buffer. Primary antibodies were probed against anti-rabbit or anti-mouse horseradish peroxidase-linked secondary antibodies (Cell Signaling Technology, Danvers, MA, USA) diluted 1:2000 in blocking buffer. Protein signaling was detected in an Odyssey FC imaging system (LI-COR Biosciences, Lincoln, NE, USA) and chemiluminescence signal was quantified in Image Studio software (LI-COR Biosciences, Lincoln, NE, USA). 2.7. Mammary Tissue AA ConcentrationFor tissue AA concentration, 60.9 ± 27.4 mg of mammary tissue was mixed with PBS and a known amount of universally labeled 13C AA mix as an internal standard. The tissue was bead-homogenized, centrifuged at 12× g for 10 min, and the supernatant was deproteinized with 0.5 M perchloric acid. Amino acids were derivatized following EZ-Faast kit instructions (CN KH0-7338, Phenomenex, Torrance, CA, USA). The tissue concentration of the derivatized AA was measured in a Shimadzu 2020 liquid chromatography mass spectrometer (Shimadzu, Kyoto, Japan) as previously reported [36]. 2.8. Statistical AnalysesStatistical analysis was conducted in SAS v. 9.4 (SAS Institute, Cary, NC, USA). Continuous data (i.e., physiological measures and milk yield) were analyzed by generalized linear mixed models using PROC MIXED with fixed effects of treatment, time (day or week-in-milk as repeated measures), and their interaction. Data collected at 14, 42, and 84 DIM (i.e., milk and blood profiles and mammary tissues analyses) were analyzed by generalized linear mixed models using PROC MIXED with fixed effects of treatment, time (14, 42, and 84 DIM), and their interaction. Colostrum variables were analyzed in PROC MIXED without repeated measures and ID within treatment was considered random. The covariate analysis used was the first-order autoregressive covariance structure (AR-1). Residuals were tested for normality, and data were log-transformed as needed. Significance was declared at p ≤ 0.05 and tendency was declared at 0.05 < p ≤ 0.10. p-values listed in text are for the main effect of treatment (i.e., TRT) unless otherwise stated. Data are presented as least squares means (LSM) ± standard error (SE) unless otherwise stated. 3. Results3.1. Physiological Measures and Milk YieldResults related to environment, DMI, and vital responses were previously reported [25,37]. Briefly, dry period (i.e., prepartum) DMI was not statistically different between treatments (9.8 vs. 11.9 ± 1.2 kg/day for dry-HT vs. dry-CL, respectively, p = 0.24), but RR and RT were higher in dry-HT relative to dry-CL cows (p < 0.01, and p < 0.01, respectively). This indicates that dry period heat stress abatement in the cooled group was successful in reducing physiological thermal indices. After calving, (i.e., postpartum, during the subsequent lactation) active cooling was provided to both groups, and DMI (16.8 vs. 18.0 ± 0.8 kg/day, p = 0.32), RR, and RT (p > 0.43) were similar between treatment groups. Dry-HT cows yielded 5.4 kg/day less milk compared to dry-CL cows across 210 DIM (p = 0.03; Figure 1A), but colostrum yield did not significantly differ (3.7 vs. 4.6 ± 0.6 kg; p = 0.36). Milk yield differed over time (p (DIM) < 0.01), but there was no interaction between treatment and week in milk (p (TRT × DIM) = 0.34, Supplementary Figure S2). Colostrum did not significantly differ in protein content between treatment groups (Figure 1B). Dry-HT cows had decreased colostrum protein yield by 0.08 kg, but this difference was a tendency (p = 0.10; Figure 1C). Dry-HT cows experienced a 0.09% overall reduction in milk protein content and a 0.18 kg/d loss in protein yield across 210 DIM compared to dry-CL, despite provision of active cooling to both groups during lactation (p < 0.04; Figure 1B,C). Milk protein content and yield varied across 210 DIM for both groups, but there was no significant interaction between treatment and time (Figure S2). Colostrum fat content and yield were decreased by 0.9% and 0.12 kg in dry-HT cows compared to dry-CL cows (p < 0.03; Figure 1D,E). Conversely milk fat content was higher in dry-HT cows (p = 0.05; Figure 1D), though milk fat yield tended to be reduced compared to dry-CL cows (p = 0.08; Figure 1E). Similar to fat content outcomes, colostrum fat: protein ratio tended to be reduced in dry-HT cows (p = 0.08; Figure 1F), but the milk fat: protein ratio was significantly greater for dry-HT cows compared with dry-CL cows (p = 0.002; Figure 1F). Colostrum lactose content did not significantly differ, and milk lactose content only tended to increase in dry-HT cows relative to dry-CL cows (p = 0.08; Figure 1G).3.2. Milk Protein Profile and Plasma and Tissue AA Milk protein fractions did not significantly differ between treatment groups, and there was no treatment by DIM interaction (p > 0.16; Table 1). Similarly, there were no statistical differences in mammary tissue AA concentration (p > 0.18; Supplementary Figure S3). Venous AA profile differed slightly between treatments, but of the 36 free AA and metabolites measured, 15 did not differ signficantly over DIM, treatment, or interaction (p > 0.10) and 15 were changed over time only (p (DIM) ≤ 0.10; Supplementary Table S2). However, dry-HT cows had higher overall concentrations of glutamic acid and phenylalanine (p ≤ 0.04) and tended to have higher taurine concentrations (p = 0.06; Table S2, Figure 2A–C) during lactation relative to dry-CL cows. There was a treatment by DIM interaction for lysine whereby dry-HT cows had lower concentrations of lysine at 42 DIM relative to dry-CL cows (p (TRT × DIM) = 0.02; Table S2, Figure 2D). There were tendencies for an interaction for tryptophan and aspartic acid; whereby dry-HT cows had elevated tryptophan concentrations at 42 DIM but lower aspartic acid concentrations at 14 DIM (p (TRT × DIM) ≤ 0.10; Table S2, Figure 2E,F). 3.3. Mammary Tissue AA Transporter Gene ExpressionAll mammary AA transporters evaluated herein, except for SLC36A1, had a treatment by DIM interaction whereby dry-HT cows had upregulated expression of AA transporters at either 14 or 42 DIM only (p (TRT × DIM) ≤ 0.03; Figure 3A, Supplementary Table S3). The treatment by DIM interaction was a stastitical tendency for SLC1A5 and SLC7A5 (p (TRT × DIM) ≤ 0.08). More specifically, dry-HT cows had increased mRNA expression of SLC1A1, SLC3A2, and SLC7A1 at 42 DIM relative to dry-CL cows (p ≤ 0.05; Figure 3A). The transporters SLC1A5 and SLC7A5 were significantly upregulated and tended to be upregulated, respectively, at 14 DIM in dry-HT cows relative to dry-CL cows (p = 0.04 and p = 0.10; Figure 3A). There was also a DIM effect for SLC1A5, 3A2, 7A1, and 7A5 (Supplementary Table S3). 3.4. Mammary Tissue mTOR Gene and Protein ExpressionThere were few differences between treatments in mammary expression of mTOR pathway genes. There were no significant effects for treatment, but there was a treatment by DIM interaction for rpS6 whereby dry-HT cows showed upregulated expression of rpS6 at 84 DIM (p (TRT × DIM) = 0.04; Figure 3B). All genes had a significant effect of day (p < 0.01). In contrast with gene expression results, dry period heat stress increased mammary protein abundance of 4E-BP1 at 14 DIM (p = 0.02; Figure 4). In addition, dry period heat stress significantly reduced mammary protein abundance of p70 S6K1 at 42 DIM and of p70 S6K1 and Akt at 84 DIM (p ≤ 0.05; Figure 4). A visual depiction of the relative expression of AA and mTOR pathway genes (i.e., ΔCt) and proteins (i.e., abundance) for each treatment group can be found in Supplementary Figure S4.4. DiscussionHyperthermia during lactation impairs both milk production and milk protein synthesis due to direct and indirect alterations to a number of biological systems [5,9,10]. While extensive research demonstrates decreased milk production due to dry period heat stress, there is conflicting evidence on its effect on milk protein synthesis and composition in the subsequent lactation [22,23,24]. The current study explores the impact of dry period heat stress on milk protein content and determines the carry-over impact on mechanistic pathways involved in milk protein synthesis during lactation, specifically AA availability and mTOR signaling. During the dry period, RR and RT were elevated in the cows exposed to heat stress relative to those provided heat stress abatement despite similar THI, indicating successful heat stress abatement for the dry-CL group, consistent with previous studies [22,23]. Notably, there was no significant difference in dry period DMI between treatment groups, which is dissimilar to past studies where dry period heat-stressed cows consumed on average 13% less than their dry period cooled counterparts [20,38]. Upon calving, all cows were provided heat abatement across the duration of the lactation. The dry period CL and HT cows had similar DMI and comparable RR and RT within thermoneutral ranges as lactating cows, suggesting that the heat stress abatement during lactation was successful in both groups and neither group was experiencing a significant heat load [39]. This lack of difference in postpartum DMI, concurrent with the subsequent lactation, is consistent with previous studies of dry period (i.e., prepartum) heat-stressed versus cooled dams [23,38,40]. Thus, we suggest that alterations during lactation related to nutrient partitioning and signaling can be partially attributed to heat stress exposure during the dry period.After calving, dry period heat-stressed cows produced 5.6 kg/day less milk. This lower production is within the range of previous findings, though the above literature review averages; these report that dams heat-stressed during late-gestation produced 3 to 7.5 kg/d less milk compared with dams provided access to shade, fans, and soakers, even when both groups were cooled during lactation [20,22,23]. While the colostrum protein content did not differ between treatments, colostrum protein yield and milk protein content and yield were significantly decreased in cows exposed to dry period heat stress. This decline in milk protein concentration could be costly to dairy producers, as most U.S. milk marketing orders employ a multiple component pricing system that compensates producers based on milk protein, as well as fat and other solids. Indeed, Bailey et al. found that if milk protein concentration fell one standard deviation below the mean, milk value would be reduced by approximately 7% or $0.82/cwt [41]. Analysis of economic losses due to dry period heat stress has yet to take into account any impact on milk composition [21], thus warranting further modeling and investigation to more accurately predict financial losses. Interestingly, while milk protein content was significantly decreased during lactation after dry period heat stress exposure, lactose content did not differ, and fat content and fat: protein ratio surprisingly increased; however, fat yield was decreased as a consequence of reduced milk yield. The increase in milk fat concentration and fat: protein ratio is inconsistent with previous studies in heat-stressed lactating and dry cows that found either no difference or decreased milk fat concentrations in heat-stressed groups [8,11,22,23,42]. Indeed, an RNA-Seq analysis of mammary metabolism of mid-lactation dairy cows under heat stress found overall downregulation of genes and pathways related to mammary tissue lipid metabolism [43]. Thus, the nature of our results necessitates further research, such as assessing dilution effects or differentiating between de novo versus pre-synthesized fatty acids, which is beyond the scope of the current analysis. The current study did not find any difference in milk protein fractions, while past research in lactating dairy cows reports alterations in milk protein profile upon direct exposure to high temperatures or across different seasons. In particular, studies show an overall lower casein concentration, a decrease in the proportion of αs2-casein and increase in the proportion of αs1-casein, or a reduction in αs-casein and β-casein concentrations under heat stress or in the summer months [11,18]. The phosphorylation of α- and β-caseins requires ATP [44], thus these authors postulated that the reduced energy and protein availability under heat stress may contribute to the alterations in protein fraction [18]. Dry period heat stress does not have a significant impact on DMI during lactation with little-to-moderate impact on postpartum NEFA, BHBA, and glucose concentrations [40,45,46]. Interestingly, a study assessing seasonal heat stress during late gestation found elevated NEFA and reduced plasma insulin and glucose during lactation; however, the cattle were unable to maintain thermoneutrality postpartum despite the provision of active cooling, which may have influenced metabolic outcomes [46]. Herein, the carry-over impact of dry period heat stress may not have influenced metabolic energy availability to a degree that would alter milk protein fraction as seen in heat stress during lactation. In the present study, the lack of difference in postpartum DMI during lactation lends credence to the idea that additional physiological and molecular systems may be regulating milk protein metabolism after exposure to prepartum heat stress. However, the numerical reduction in prepartum DMI during dry period heat stress or more nuanced alterations in energy partitioning not found in the present study could also contribute to this. Another possible contributor is reduced availability of AA and metabolites for milk protein synthesis. In the current study, systemic plasma Glu, Phe, and taurine had overall increased concentrations in dry period heat stressed cows across early and peak lactation. Previous studies in lactating cows show both increases [47] and decreases [9,10,43] in AA availability in hyperthermic conditions. Consistent with the current results, Guo and others found an increase in the concentration of blood Glu but a decrease in Lys concentration in heat stressed lactating cows [47]. Glutamic acid supports the local immune function and is required for heat-shock protein structural and functional integrity [48,49,50]. Although not directly studied in bovine heat stress studies, Phe supplementation in vitro increased the expression of HSP70 mRNA in bovine renal epithelial cells [51], and Phe metabolism assists in the suppression of T-cell immune responses [52]. Similarly, taurine supplementation in chronically heat-stressed poultry has been shown to promote heat shock protein expression and enhance protein synthesis [53,54]. Thus, upregulation of these AAs in circulation might play a role in assisting in the increased immune and heat shock responses associated with heat stress, both during the dry period and during lactation [43,55]. Further, our results show interaction effects; for instance, plasma Lys, a limiting essential AA, tended to decrease specifically at 42 DIM (i.e., roughly peak milk production) in dry-HT cows compared with dry-CL cows. Lysine is one of the most limiting AA for milk protein synthesis [56], and Lys stimulates milk protein synthesis, partly by promotion of amino acid transporter B (0+) and activation of the mTOR pathway in bMEC [57]. Though, the plasma AA measured herein represents the systemic AA profile and thus caution should be taken when inferring its specific impacts on mammary metabolism. Notably, there is considerable interest in infusing limiting AA to improve milk and protein yields [58], but additional infusion of Lys (as well as Met and branched-chain AA) has been shown to alter milk protein content with no impact on milk or protein yields during heat stress in lactating cows [59]. Thus, supplementing dry cows with the essential AA may aid in modulating milk protein content but might not be an effective tool to combat the impact of dry period heat stress on milk yield. Further research should test this experimentally. The amino acid transporter genes SLC7A5 and SLC1A5 had increased expression at 14 DIM in dams heat-stressed in late gestation relative to cooled dams. Later at 42 DIM, transporter genes SLC1A1, SLC3A2, and SLC7A1 were also upregulated in dry period heat-stressed cows. A comprehensive study by Bionaz and Loor [33] demonstrates that the transporter genes measured herein are responsible for encoding a variety of proteins that are responsible for active pumping of AA into mammary tissue (SLC1A1 and 1A5) or the counter-transport of cationic AA (SLC3A2, 7A1, 7A5). Of the transporters upregulated around peak lactation (i.e., 42 DIM) in the current study, SLC1A1 encodes EAAT3 for the transport of AA such as Glu. The gene SLC3A2 assists in guiding LAT1 to the cell membrane though synthesis of heavy polypeptide chain 4f2hc, leading to the transport of essential AA like Phe and Met. The protein CAT1 is encoded by SLC7A1 and transports Lys and Arg. We suggest that this orchestrated upregulation of genes encoding for AA transporters in dry-HT cows at early and peak lactation may be a compensatory mechanism of the mammary gland to increase mammary AA availability and consequently drive milk protein synthesis and cellular growth. Kaufman et al. [16] measured the same set of AA transporter genes in a bMEC in vitro direct heat stress model. They also found an upregulation in the gene expression of SLC1A1 and SLC3A2 after 12 h of 41.5 °C heat exposure and proposed a similar mechanism–increased AA transport activity could be a cellular adjustment to maintain AA uptake, protein synthesis, and bMEC mass [16,60]. Interpretation relative to the present study requires caution, however, as the dairy cows herein were exposed to heat stress during their dry period and not to direct hyperthermia during lactation. Along with an increase in AA transporter gene expression, Kaufman and collaborators reported a decrease in phosphorylation of the insulin transductor and mTORC1 activator Akt and of the mTORC1 downstream substrate rpS6 [16]. Phosphorylation of Akt was also reduced in skeletal muscle of gilts upon short term heat stress [61]. In the present study, dry period heat stress reduced Akt protein abundance at 84 DIM. It also reduced protein abundance of mTORC1 substrate and rpS6 kinase p70 S6K1 at 42 and 84 DIM. Intriguingly, mTORC1 inhibition stimulates the expression of the mRNA translational repressor 4E-BP1, which stimulates mRNA expression of AA transporter genes SLC3A2, SLC7A1, and SLC7A5 [62], potentially as a coping mechanism of AA deficiency. Herein, dry period heat stress tended to increase 4E-BP1 abundance at 14 DIM, confirming previous in vitro results at the mRNA level [63]. Finally, though not measured in the present study, systemic circulating insulin could have played a role in milk protein metabolism via the mTORC1 pathway. Insulin activates mTORC1 signaling and mTORC1 kinase activity, and in turn, mTORC1 controls insulin signaling and sensitivity to modulate biological functions such as cell division and growth and protein translation [64]. While dry period heat stress does not influence prepartum insulin metabolism, it has been shown to promote elevated circulating insulin and glucose in early lactation [45]. Finally, it should be noted that the limited sample size in the present study (n = 6 per treatment for molecular analysis and n = 12 per treatment for physiological measures and milk and component yields) could contribute to the lack of significant differences observed in dry matter intake and milk protein profiles. Thus, results and further application should be interpreted with caution. 5. ConclusionsOur results indicate an orchestrated response of the lactating mammary gland to prior dry period heat stress exposure, in part mediated by the mTORC1 pathway. This includes the upregulation of AA transporter gene expression to possibly overcome AA deficiency, along with a repression of milk protein production, which extends from early lactation to mid-lactation. Factors affecting plasma or mammary tissue AA concentrations during lactation in response to dry period heat stress require additional investigation. More research in this area is necessary to ensure production of high-quality and -quantity milk as a sustainable source of protein in the face of rising global temperatures.

animals : an open access journal from mdpi

10.3390/ani11092638

PMC8469073

Mountain hares in Scandinavia are classified as Near Threatened in the Norwegian and Swedish Redlists assessing the risk of species extinction. This is due to a possible population decline witnessed during the last decades in Scandinavia. Competition between large herbivores such as moose, red deer, roe deer on one hand and hares on the other, is one of several hypotheses that has been put forward to explain this decline. In a cafeteria trial (providing several types of forage to determine food preference) we investigate the effects of previous moose winter foraging on the food selection (i.e., amount consumed, bites per minute and bitediameter) of downy birch and goat willow by captive hares. We find that hares do not differentiate among levels of previous moose foraging on downy birch but have larger bite diameters of goat willow earlier eaten on by moose, compared to plants not fed on by moose. Thus, effects of moose on hare winter food quality seem to be limited. We highlight the need for studies focusing on (1) effects of previous moose foraging using wild hares in a natural experimental design, and (2) effects of moose foraging on available hare food at a landscape scale during winter.

Mountain hares (Lepus timidus) in Scandinavia are classified as Near Threatened in the Norwegian and Swedish Redlists. This is due to a possible population decline witnessed during the last decades in Scandinavia. Competition between large herbivores and mountain hares is one of several hypotheses that has been put forward to explain this decline. In a cafeteria trial we investigate the effects of previous moose (Alces alces) winter browsing on the food selection (i.e., biomass consumed, bites per minute and bitediameter) of downy birch (Betula pubescens) and goat willow (Salix caprea) by captive mountain hares. We find that mountain hares do not differentiate among previous browsing levels of downy birch, but have larger bite diameters of goat willow earlier browsed by moose, compared to non-browsed plants. Thus, effects of moose on mountain hare winter food quality seem to be limited. We highlight the need for studies focusing on (1) qualitative effects of moose browsing using wild mountain hares in a natural experimental design, and (2) quantitative effects of moose browsing on available mountain hare forage at a landscape scale during winter.

1. IntroductionIn Scandinavia, mountain hares (Lepus timidus), have shown possible population declines over the past decades [1,2,3] (www.viltdata.se accessed on: 27 August 2021) and were classified as Near Threatened in the Norwegian Redlist in 2015 and again in 2021 (www.artsdatabanken.no accessed on: 27 August 2021), as well as the Swedish Redlist in 2020 (www.artfakta.se accessed on: 27 August 2021). Several hypotheses have been put forward to explain this decline including climate change, land use change, parasites, predation, and competition [3]. Climate change, color mismatch and increased predation is currently a topic of general research interest, and we have earlier shown that abundance of mountain hares is negatively affected by an interaction between reduced snow cover duration and abundance of generalist predators; red fox (Vulpes vulpes) and pine marten (Martes martes) [1]. However, the mountain hare may not only be negatively affected by top-down effects, but also bottom-up through interactions with other species from within the herbivore guild [4]. On a landscape scale, moose (Alces alces) is known to reduce abundance of deciduous trees e.g., [5,6], and mountain hares are negatively associated with moose presence at a habitat patch scale [7]. Thus, in areas of high density, moose may be an important competitor towards mountain hares [8].Herbivore species are not expected to compete if they either differ in digestive systems [9], or in body size [10]. Hence, one would expect moose and mountain hares to have large ecological niche separation in terms of diet overlap and potential for food competition. However, moose [11] and mountain hare [12] consume several of the same plant species (i.e., birch (Betula spp., salix (Salix spp.) and aspen (Populus tremula), especially during winter. Herbivore species’ food preferences vary along a gradient from bulk feeding of low nutritious forage to selective feeding of high-quality forage see [13], and references therein. On one hand, hindgut fermenters (such as the mountain hare) have less efficient digestive system compared to ruminants (such as the moose). Due to their relatively inefficient digestive systems, hind gut fermenters select plant quantity over quality. Small herbivores on the other hand, are in general more selective, targeting higher food quality compared to large-bodied herbivores [13]. Ironically, due to this interaction between body size and digestive systems—a large ruminant and a small hindgut fermenter may compete over food sources. Indeed, previous studies have shown mountain hares to be competitively inferior to larger herbivores such as the roe deer (Capreolus capreolus) [14]. Hulbert and Andersen [14] did not detect any feeding-height separation between the two species, however mountain hares switched to smaller bite diameters when in sympatry with roe deer, likely leading to higher concentrations of harmful plant secondary metabolites. The authors suggest that this switch may lead to reduced survival of mountain hares in the presence of roe deer. Large herbivores such as moose and roe deer may therefore limit mountain hare densities, especially in areas of high cervid density. These negative impacts may either be transferred through reduced forage quantity, i.e., deciduous winter browse [5,6] or forage quality through chemical changes in plants [14,15].To investigate the latter competitive pathway of reduced food quality we conducted cafeteria trials with captive mountain hares, using downy birch (Betula pubescens) and goat willow (Salix caprea), two species which are consumed by moose and mountain hares in winter. For both moose and mountain hares, downy birch is the staple food, while goat willow is among the highly preferred plant species [11,12,15,16]. We predicted that mountain hares would prefer unbrowsed downy birch due to a lower concentration of plant secondary metabolites [17]. For goat willow we predicted that mountain hares would prefer previously browsed shoots, as the related tea-leaved willow (S. phylicifolia) is known to produce shoots with higher biomass and lower levels of chemical defense as a response to moose browsing [18].2. Methods2.1. Sample CollectionThis study was carried out in Østerdalen (61° N, 11° E), Southeast Norway along a gradient in moose density and browsing pressure. In January 2008, we collected downy birch and goat willow branches (approximately 30–40 cm long), after inset of winter dormancy but without signs of recent browsing. We here consider branches consisting of several years’ growth, while shoots are one year’s growth. All branches were collected from trees <2.5 m height. Any collected shoots with signs of current moose browsing were discarded. Shoot coloration and bark structure may be used to separate the most current growing season from previous years growing seasons. Browsing levels were estimated as a percentage of available shoots browsed during the previous browsing season. For downy birch, we collected branches without (0%), intermediate (30% < 60%), and high levels of previous moose browsing (>90%). We also collected goat willow without browsing (0%) and more than 30% previously browsed shoots. Browsing levels were determined based on morphology of the branches. After browsing, lateral shoots tend to form leaving a dry «stump» where the moose bit off the shoot last winter. We specifically targeted the mentioned percentage intervals as to avoid overlapping of categories—thus, this approach should be robust towards any slight over or underestimation of percentage browsing. We originally aimed for three categories of goat willow, like that of downy birch, but high moose densities and limited availability of goat willow prevented us from doing so. For the control (no browsing) we collected branches from two military facilities (Mil I and Mil II) within the Østerdalen valley, which were fenced more than 30 years before the experiment. The other treatments were collected just outside the two military facilities as well as two additional areas (Imsdalen and Kopppangskjølen) known to have high moose presence and browsing pressure in winter [19,20]. All four sites lie within the boreal forest and have similar forest composition dominated by Norway spruce (Picea abies), Scots pine (Pinus sylvestris), interspersed with downy birch, silver birch (B. pendula), Salix. spp., aspen, rowan (Sorbus aucuparia) and grey alder (Alnus incana). Samples were collected at these four sites depending on availability. Thus, for Imsdalen we had the categories “willow browsed”, “birch intermediate”, and “birch high”. For Koppangskjølen we had “birch intermediate” and “birch high”, For Mil I we had “birch no browsing”, “willow no browsing”, “birch intermediate” and “willow browsed”. For Mil II we collected “birch no browsing”, “willow no browsing”, and “willow browsed”.2.2. Cafeteria TestFourteen captive mountain hares were housed in cages with ad lib. access to water, rabbit pellets, and branches from various deciduous trees. The mountain hares were part of a breeding facility for restocking natural populations. No permit was required under the Norwegian Food Safety Authority as the experiment was part of the natural feeding of the animals. Mountain hares were kept indoors in an unheated barn. For two days prior to the start of the experiment, mountain hares were fed downy birch and goat willow, in order to habituate them to the material to be used in the experiment.Mountain hares were deprived of food (12 h) over the night before the experiment started. We provided the mountain hares with bundles of branches, weighing approximately 100 g (±0.01 g), for 90 min. All 14 mountain hares were given all five categories/species of branches in randomized order, over a six-day period. Immediately after finishing the trial, we measured bite diameter and shoot base diameters (to the nearest 0.1 mm) of the remaining material using callipers. Bundles and all shoot residues on the cage floor were weighed and we calculated wet biomass consumed. We weighed control bundles of goat willow and downy birch not used in the experiment but treated similarly. Mean weight loss due to evaporation was 0.10 g and 0.2 %; thus, we regarded all weight loss as biomass consumed.Additionally, mountain hares were filmed during the cafeteria trial, and we registered time spent feeding, and number of bites. Some of the recordings failed to document all mountain hare activity during the trial and were thus excluded from further analysis.2.3. Statistical AnalysisThe dataset consisted of branches from four different areas, the means of response variables varied among the different areas (Table 1); however, the 95% CI overlapped for tree species as well as browsing levels. Thus, these differences were not considered for further analysis, but our results should reflect general patterns as the material was collected from several sites.For the two species goat willow and downy birch we ran separate mixed effects ANOVAs with grams consumed, bites per minute and bitediameter as response variables, and moose browsing intensity as explanatory variable. We included mountain hare individual as a random term, to account for individual variation among mountain hares. Analyses were performed in R version 4.0.4 [21].3. ResultsFor downy birch we found no effects on amount consumed (Figure 1A), bites per minute (Figure 1B) or bite diameter (Figure 1C); thus, mountain hares did not differentiate among level of previous browsing from moose (Table 2). For goat willow we only found an effect of treatment on bite diameter (Figure 1C), with mountain hares having larger bite diameter for goat willow previously browsed by moose (Table 2). For the variables biomass consumed (Figure 1A) and bites per minute (Figure 1B) we found no effect of treatment (Table 2). Overall, biomass of goat willow consumed was higher than that of downy birch (Figure 1A).4. DiscussionFor downy birch we found no effect of previous moose browsing on mountain hare preference. In a similar field-based study by Danell and Huss-Danell [17], mountain hares also did not differentiate between moderate and low previous moose browsing on birches (B. pubescens and B. pendula). For goat willow we found an effect on bite diameter only, where mountain hares have larger bite diameters on goat willow previously browsed by moose. Herbivores are known to select for large shoot base diameters [22], and our results corresponds to findings by Stolter (2008) in the closely related tea-leaved willow that is known to respond to moose winter browsing by producing thick shoots, which again is preferred by moose the following winter.Plants may respond to browsing by tolerance or avoidance. Avoidance of herbivory may either be escaping browsing altogether or physical and chemical defense [23]. Chemical defense may be constitutive (i.e., omnipresent) or induced as a response to damage by e.g., herbivores [23,24]. Regrowth after moose browsing on the tea-leaved willow produced shoots with lower concentration of plant secondary compounds (Stolter 2008). This counter-intuitive response led to a positive feedback loop, where moose rebrowsed tea-leaved-willow that were browsed during the previous browsing season (Stolter 2008). Although we did not investigate chemical composition in the treatments, it is possible that such a positive feedback loop with decreased levels of plant secondary compounds following moose browsing is the case also with goat willow in the current study.The lack of a strong effect of previous moose browsing in the current study could be due to several reasons: (1) The mountain hares used in the experiment are captive; thus, their food selection and preferences may not be reflecting that of wild mountain hares. (2) Many plant secondary compounds act as digestive inhibitors e.g., [25]. Thus, if the mountain hares of our study have unlimited access to high quality rabbit pellets, any response in the plants to moose browsing leading to increased concentration of digestive inhibitors may be swamped by the mountain hares’ unlimited access to high quality food.5. ConclusionsIn recent years, several studies have suggested competition between lagomorphs and cervids. In Italy, European hare (L. europaeus) and roe deer are known to increase their diet overlap in winter [26]; however, they seem to reduce the degree of competition by having low spatial overlap on the landscape scale [27], similarly to that of moose and mountain hares in Northern Sweden [7].From our results and those of others [17] we suggest that if there is a competitive relationship between moose and mountain hares, this is not likely caused by qualitative changes in mountain hare winter forage, but rather quantitative changes as a result of moose reducing available mountain hare winter forage in the landscape. Future studies should investigate (1) any qualitative effects (i.e., impacts on plant secondary compounds) of moose browsing using wild mountain hares in a natural experimental design, and (2) the possible alternative competitive route of quantitative effects of moose on available mountain hare forage at a landscape scale during winter.

animals : an open access journal from mdpi

10.3390/ani11092541

PMC8466186

The objective of this study is to examine the effects of T-2 toxin (T-2) and green tea powders (GTP) on growth performance, hematology, and pathology parameters in Brown Tsaiya ducklings (BTDs) and Kaiya ducklings (KDs). T-2 toxin shows a strong and differential toxicity in growth suppression, as well as abnormalities in the hematological and pathological parameters of BTDs and KDs. We found that GTP could potentially prevent T-2-induced poor growth performance and improve some hematological parameters. Moreover, BTDs were more sensitive than KDs in terms of responses to T-2 toxicity and GTP detoxification.

A 3-week feeding trial in a 3 × 2 × 2 factorial design was conducted with three concentrations (0, 0.5, and 5 mg/kg) of T-2 toxin (T-2) and two levels (0% and 0.5%) of green tea powder (GTP) supplements used in the diets of female brown Tsaiya ducklings (BTDs) and Kaiya ducklings (KDs), respectively. Breed had a significant effect on the growth performances and the relative weights of organs and carcass. In general, the growth performances of KDs were better than BTDs. The relative weights of organs and carcass of BTDs were typically heavier than those of KDs; however, the breast of KDs was heavier than those of BTDs. Both ducklings received 5 mg/kg of T-2 blended in the diet showed lower feed intake and body weight gain (BWG) in the second and the third week. The diet containing 5 mg/kg of T-2 and 0.5% GTP improved the BWG compared to those fed the diet supplemented with 5 mg/kg of T-2 without GTP in BTDs. Ducklings fed the diet containing 5 mg/kg of T-2 induced hypocalcemia and hypomagnesemia, as well as decreased concentrations of creatine phosphokinase and alkaline phosphatase. The concentrations of blood urea nitrogen (BUN) and glutamate oxaloacetate transaminase (GOT) were increased in KDs and BTDs fed the diet containing 5 mg/kg of T-2 without GTP, respectively. However, duckling diets containing 5 mg/kg of T-2 with 0.5% GTP lowered concentrations of BUN and GOT in the blood plasma of KDs and BTDs, respectively. The diet containing 5 mg/kg of T-2 increased the relative kidney weight but decreased the relative breast weight of ducklings. Enlarged gizzards and reduced relative leg weights were observed in BTDs fed the diets containing 5 mg/kg of T-2. In summary, BTDs are more sensitive than KDs in responding to T-2 toxicity and GTP detoxification. Green tea powder has detoxification ability and could potentially mitigate T-2 toxicity on BWG, BUN, and GOT in ducklings.

1. IntroductionMycotoxins are secondary toxic metabolites produced by various mold species [1,2]. One of the most studied mycotoxins is trichothecenes, which has more than 200 derivatives [3]. The major chemical features responsible for the biological activities of trichothecenes are the 12,13-epoxy ring and the variable structures of side-chain branches, which determines the toxicities and characteristics among the different trichothecenes [4,5,6]. Trichothecenes are classified into four types (A, B, C, and D) according to their side-chain structures [6]. Within them, Type A includes T-2 toxin (T-2) and HT-2 toxin (HT-2), and type B includes deoxynivalenol (DON) and nivalenol (NIV). These two types of trichothecenes contribute the major problems of mycotoxin contamination in ingredients and feeds, mainly by T-2 and DON. T-2 toxin is a Fusarium-derived trichothecene, which can inhibit protein synthesis [7], induce lipid peroxidation [8], and damage the organs and gastrointestinal system in animals [9]. T-2 toxin can induce poor growth performance, but considerable differences exist between chickens and ducks in dosage, age, and response time. Kutasi et al. [10] reported that 1-day-old white Pekin ducklings decreased body weight by consuming feed contaminated only with 0.6 mg/kg of T-2 for 4 weeks, whereas in 1-day-old chicks, it took 3 weeks to show the reduced body weight after consuming feed contaminated with 4 mg/kg of T-2 [11]. It is known that waterfowls have more unsaturated fatty acids (UFA) in their body tissues than chickens [12]. Given that mycotoxin T-2 is fat-soluble, it can accumulate in animals’ bodies for months and make waterfowls more susceptible to the UFA-related damage than chickens [13]. Fernye et al. [14] also indicated that waterfowls were exceptionally sensitive to T-2. It has been reported that energy metabolism can be inhibited by T-2 in poultry [15]. Additionally, the basal metabolic rate differs significantly between meat-type and laying-type animals during their brooding period [16]. It is not known whether meat-type ducklings and egg-type ducklings are similarly sensitive to T-2 toxicity.The chemical and physical properties of T-2 include high heat stability [4], high molecular weight, and low polarity [17]; therefore, the efficacy of detoxifying T-2 via thermal inactivation or absorbent binding is relatively limited [18]. T-2 toxin exhibits its toxicity mainly by inducing excessive production of free radicals, which trigger lipid peroxidation in animals [19]. It has been well-known that antioxidants, such as ascorbic acid, tocopherol, and selenium, can neutralize superoxide anion, reactive oxygen species (ROS) scavengers, free radicals, and lipid peroxidation by T-2 [20]. It is also reported that the antioxidants detoxified T-2 toxicity in poultry [21,22].Recently, there has been increasing interest in finding natural antioxidants from plant phytochemicals to protect animals against free radicals [23,24]. Green tea is one of the most widely consumed beverages in Asian countries, such as China, Japan, India, and Taiwan. Green tea powder contain many functionally active substances, including polyphenols, catechins, alkaloids, and polysaccharides [25], which have attracted considerable attention as animal feed additives in recent years [26,27]. The antioxidant capacity of GTP comes from green tea polyphenol compositions [28], which include catechin (C), catechin gallate (CG), epicatechin (EC), epicatechin gallate (ECG), epigallocatechin (EGC), epigallocatechin gallate (EGCG), gallocatechin (GC), and gallocatechin gallate (GCG) [29]. The green tea polyphenol scavenges ROS and free radicals through several proposed mechanisms, including depolarization of electrons, formation of intramolecular hydrogen bonds, and rearrangement of molecular structure [30,31]. They interrupt oxidative reactions by chelating free copper and iron, which also catalyze the formation of ROS in animals [32]. Similarly, the antioxidant nature of green tea polyphenol is beneficial in preventing lipid metabolism disorders and reducing DNA damage [33,34,35]. Serving as an antioxidant, it has also been shown that GTP could detoxify the hepatotoxicity and genotoxicity of AFB1, fumonisin B1 (FB1), and citrinin [29,36,37] in different animal species. However, the effect of GTP on detoxifying T-2 remains elusive.Brown Tsaiya ducks (BTDs) and Kaiya ducks (KDs) are endemic and popular laying- and meat-type duck species, respectively, in Taiwan. However, no reports on their response to T-2 toxicity are currently available. Green tea has long been recognized for its antioxidative potential in many aspects, which are thought to be beneficial in scavenging cellular ROS toxicity, a partial contributor associated with the T-2 toxicity. Therefore, the objectives of the present study are to test the effects of trichothecene T-2 along with GTP as a dietary supplement in duckling rations on the growth performance, plasma biochemical parameters, and certain representative organs weights of BTDs and KDs.2. Materials and Methods2.1. Animal Care and UseThis in vivo study was conducted in strict accordance with the guidelines recommended and approved by the Institutional Animal Care and Use Committee (IACUC) of National Chung Hsing University (Approval number: IACUC-100-41).2.2. Production and Analysis of T-2 ToxinT-2 toxin was produced by the fermentation of corn powder using Fusarium trincintum, which was selected from mold-contaminated wheat. Corn was inoculated with F. trincintum (approximately 108 conidia/mL) conidia via dispersion and placed in Erlenmeyer flasks. After culturing at 14 °C in the dark for 28 days [38], the culture was dried at 65 °C for 2 days. The dried culture was crushed into powder using a shredder (RT-UF26, Rong Tsong, Taichung, Taiwan). Quantification of T-2 in inoculated corn powder was performed by high-performance liquid chromatography (HPLC) following the extraction, purification, and analysis based on the protocol established in Visconti et al. [39] with minor modification. Briefly, standards and sample extracts were injected into an HPLC pump (L2130, Hitachi, Tokyo, Japan) with a fluorescence detector (L-2485, Hitachi, Tokyo, Japan) using an auto-sampler (L-2200, Hitachi, Tokyo, Japan) set to an excitation wavelength of 381 nm and an emission wavelength of 470 nm. The flow rate of the mobile phase was set at 1.0 mL/min. A binary gradient was applied as follows: The initial composition of the mobile phase was established; 70% acetonitrile/30% water for 5 min, after which the acetonitrile content was increased to 85% in 10 min and kept constant for 7 min. Finally, the acetonitrile was decreased to 70% in 1 min and kept constant for 7 min. Acetonitrile was HPLC grade and purchased from Sigma (Saint Louis, MO, USA). The HPLC MightysilTM RP-C18 column (4.6 mm × 250 mm, 5 μm, Kanto Chemical, Tokyo, Japan) was used in the analysis. The concentration of T-2 in the corn powder was 230 mg/kg.2.3. Compositions and Antioxidative Capacity of Green Tea Powder2.3.1. Analyses of the Chemical Composition and Catechin ConcentrationGreen tea powder was provided as a commercial feed additive by Leshan Yujia Tea Development Co., Ltd. (Sichuan, China). Dry matter was analyzed according to AOAC methods [40]. Crude fiber, neutral detergent fiber, and acid detergent fiber were determined by using a fiber analyzer (ANKOM 200/220, ANKOM Technology, Macedon, NY, USA) according to Fay et al. [41]. Green tea polyphenols (C, CG, EC, ECG, EGC, EGCG, GC, and GCG) were determined by HPLC, as modified by Zhang et al. [42]. Briefly, the GTP sample (3 g) were dissolved with 150 mL water in water bath at 80 °C for 5 min for the extraction of polyphenols. The final solution was aliquoted to 150 mL and filtered (pore size 0.45 μm) for HPLC analysis. For the HPLC analysis of catechin, standards and samples were injected into an HPLC pump using an autosampler with an ultraviolet (UV) detector (L-2400, Hitachi, Tokyo, Japan) set to 280 nm. The flow rate of the mobile phase was 1.0 mL/min. The mobile phase consisted of water with 1% (v/v) formic acid (Saint Louis, MO, USA) and acetonitrile with linear gradient elution. The acetonitrile level was increased from 4% to 18.7% in 42 min, and then the acetonitrile level was decreased from 18.7% to 4% in 1 min. The analytical HPLC column used was MightysilTM RP-C18 column. The chemical compositions and catechin concentrations of GTP are shown in Table 1.2.3.2. Antioxidative Capacity of Green Tea PowderThe scavenging activity of 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical was determined according to the method described by Villano et al. [43]. Briefly, a 25 mg/L solution of DPPH (Alfa Aesar, Haverhill, MA, USA) radical solution in methanol (Merck, Darmstadt, Germany) was prepared, and 1.95 mL of this solution was mixed with 50 μL of extract solution (100 to 4000 μg GTP/mL distilled water). The solution was then mixed by a vortex mixer (Vortex-Genie 2, Scientific Industries, Bohemia, NY, USA) and left for 90 min at room temperature in the dark. The absorbance (A) was measured at 515 nm by using a spectrophotometer (Smart Spec Plus, Bio-Rad, Hercules, CA, USA). Ascorbic acid (vitamin C, Vit. C, Sigma-Aldrich, Saint Louis, MO, USA) was used as a positive control. This activity was given as the percent of DPPH scavenging and was calculated as follows:DPPH scavenging (%) = [(Acontrol − Asample)/Acontrol] × 100(1) where Asample is the absorbance of sample containing GTP and Acontrol is the absorbance of the sample without GTP.Total reducing power was determined according to the method described by Yildirim et al. [44]. Briefly, GTP (100 to 4000 μg) was dissolved in 1 mL of distilled water, then mixed with 1 mL of 0.2 M phosphate buffer (pH 6.6) and 1 mL of 1% potassium ferricyanide (Sigma-Aldrich, Saint Louis, MO, USA), after which the mixture was incubated at 50 °C for 30 min. Next, 1 mL of 10% trichloroacetic acid (Sigma-Aldrich, Saint Louis, MO, USA) was added, and the mixture was centrifuged at 3000× g for 10 min. Finally, 1 mL of upper layer solution was mixed with 1 mL of distilled water and 0.1 mL of 0.1% ferric chloride (Sigma-Aldrich, Saint Louis, MO, USA). The solution was then mixed by a vortex mixer and left for 10 min at room temperature in the dark. The absorbance was measured at 700 nm by a spectrophotometer. Vitamin C was used as a positive control. The absorbance values correspond to the capacity for metal ion reduction.Ferric chelating activity was determined according to the method described by Dinis et al. [45]. Briefly, the extracts (100 to 4000 μg) in 1 mL of distilled water were mixed with 3.7 mL of methanol and 0.1 mL of 2 mM ferrous chloride (Sigma-Aldrich, Saint Louis, MO, USA) and then left for 30 s. Next, 0.2 mL of 5 mM ferrozine (Sigma-Aldrich, Saint Louis, MO, USA) was added. The solution was then mixed by a vortex mixer and left for 10 min at room temperature in the dark. The absorbance was measured at 562 nm by a spectrophotometer. Ethylenediaminetetraacetic acid (EDTA, Sigma-Aldrich, Saint Louis, MO, USA) was used as a positive control. The calculation of ferric chelating activity was shown as follows:Ferric chelating activity (%) = [(Acontrol − Asample)/Acontrol] × 100(2) where Asample is the absorbance of sample containing GTP and Acontrol is the absorbance of sample without GTP.2.4. Experimental Designs and Feeding Management of DucklingsDuring the 3-week brooding period of BTD and KD, all ducklings were fed the same basal diet supplemented with various levels of T-2 and/or GTP as follows: basal diet without T-2 and GTP (control group), basal diet with GTP (0.5% GTP), low T-2 diet (0.5 mg T-2/kg), low T-2 diet with 0.5% GTP, high T-2 diet (5 mg T-2/kg), and high T-2 diet with 0.5% GTP. The T-2 concentration was designated to show the phenotypic T-2 toxicity and the experimental period was the length of the regular brooding period [10,11,46]. Moreover, 120 1-day-old female BTDs were obtained from the hatchery of the Department of Animal Science, National Chung Hsing University (Taichung, Taiwan). Another 120 1-day-old female KDs were obtained from Yilan Livestock Research Institute (Yilan, Taiwan). Ducklings of each type were randomly divided into six groups with four replicates per group (five ducklings in each replicate). The ducklings were fed ad libitum, with free access to clean water under light and heat supplied 24 h a day at an ambient temperature of 28 °C. To avoid synergistic effects caused by the contaminations of other mycotoxins, the basal diet was analyzed for AFB1, DON, FB1, HT-2, ochratoxin A (OTA), T-2, and zearalenone (ZEN) by HPLC. All of these mycotoxins had concentrations below the limit of detection (LOD < 10 μg/kg). The nutrients and chemical compositions of basal diet are shown in Table 2. The fungal powder of F. trincintum was mixed with the mycotoxin-free corn in the basal diets and adjusted to 0.5 or 5 mg/Kg of T-2 concentrations in the diets according to the experimental design. The GTP was additionally supplemented into the T-2 containing basal diets according to the experimental design. All duckling feeds were stored at 4 °C until use.2.5. Growth PerformancesDuring the experimentation period, body weights of day-old ducklings were recorded individually prior to the onset of the experiment and at the end of each week to calculate daily body weight gain (BWG, g/bird/day). Feed intake was recorded daily. The formulation for feed intake was as follows:Feed intake = (Feed supplied − Feed remained)/Number of ducklings per replicate(3)Feed conversion ratio (FCR) was calculated as feed intake (g/bird/day)/BWG.2.6. Relative Weights of Organs and Carcass, and Plasma Biochemical ParametersWhen the ducklings were 3 weeks old, 12 ducklings were randomly chosen from each group and sacrificed by cervical dislocation. The heart, gizzard, liver, kidney, breast, leg, and tibias (left and right) of ducklings were sampled and weighed. The relative weights of organs and carcass were calculated as follows:Relative organ or carcass weight = (Weight of organ or carcass/Live weight of duckling) × 100(4)Blood samples were collected into heparinized tubes by cardiac puncture. Plasma was obtained after the removal of red blood cells by centrifugation at 1500× g for 15 min. All plasma samples were stored at −20 °C until analysis. The concentrations of blood urea nitrogen (BUN), uric acid (UA), creatinine (CREA), creatine phosphokinase (CPK), glutamate oxaloacetate transaminase (GOT), glutamate pyruvic transaminase (GPT), alkaline phosphatase (ALK), cholinesterase (CHE), total protein (TP), albumin (ALB), and globulin (GLO) in plasma were determined using an automated clinical chemistry analyzer (Automatic Analyzer 7150, Hitachi, Tokyo, Japan), according to the manufacturer’s instructions. The levels of calcium (Ca) and magnesium (Mg) in plasma were determined with flame atomic absorption spectroscopy (Atomic Absorption Spectrophotometer Z-5000, Hitachi, Tokyo, Japan).2.7. Statistical AnalysisA randomized complete block design (RCBD) with a 3 × 2 × 2 factorial arrangement was designed. Treatment groups consisted of three T-2 concentrations (0, 0.5, and 5 mg/kg), two GTP levels (0% and 0.5%), and two duckling breeds (BTD and KD). Data were statistically analyzed using general linear models (GLMs) (PC-SAS® ver. 9.2, 1995) following factorial treatments in a spilt-plot design, in which the two breeds of ducklings nested in each of the 24 pens were regarded as blocks.The mathematical model is as follows:Yijkl = μ + Ri + Tj + Gk + Bl + (TG)jk + (TB)jl + (GB)kl + (TGB)jkl + εijkl(5) where Yijkl = the observed response of duckling breed in a pen; μ = the overall mean; Ri = the effect of the ith block (pen); Tj = the fixed effect of T-2 concentrations; Gk = the fixed effect of GTP level; Bl = the fixed effect of duckling breed; (TG)jk = the interaction effect of T-2 concentrations × GTP levels; (TB)jl = the interaction effect of T-2 concentrations × duckling breed; (GB)kl = the interaction effect of GTP levels × duckling breeds; (TGB)jkl = the interaction effect of T-2 concentrations × GTP levels × duckling breeds; and εijkl = the residual error when duckling breed nested in a pen are regarded as an experiment unit, εijkl∩N (0, δ2ε). Means of the T-2 concentrations, the GTP levels, and the two duckling breeds were compared by using Tukey’s testing, and the significance level was set at p < 0.05.3. Results3.1. Antioxidative Capacity of the Green Tea PowderIn this study, GTP was used as an antioxidant, and its antioxidative capacity was first tested as shown in Figure 1. The results showed that the DPPH scavenging activity, reducing power, and ferric ion chelating activity of GTP (100–2000 μg/mL) appeared to be lower than those of vitamin C and EDTA. Moreover, the activities of DPPH scavenging and ferric chelating were equivalent to those of the control (vitamin C and EDTA) at a high concentration (4000 μg/mL).3.2. Growth PerformanceBreed had a significant effect on the feed intake, BWG, and FCR throughout the experimental period (p < 0.0001, Table 3). Daily feed intake and BWG in KDs were higher than in BTDs, but FCR in KDs was lower than that in BTDs. When T-2 concentration was 5 mg/kg, lower feed intake and BWG were observed in duckling when compared to those fed 0 mg/kg of T-2 in the second and third week. The diet contained 0.5% GTP increased BWG and improved FCR in ducklings in the third week. There was an interaction between T-2 concentration and duck breed (p < 0.05), wherein the BTDs fed diets contained 5 mg/kg of T-2 had lower feed intake (first week) when compared to the BTDs fed 0 mg/kg of T-2 (Supplementary Materials Table S1). Both ducklings had lower BWG in 5 mg/kg of T-2 treatment in the third week. Only BTDs fed the diet containing 5 mg/kg of T-2 with 0.5% GTP improved the BWG compared to those fed the diet 5 mg/kg without GTP (Table S2). The poor BWG of Kaiya was not improved by GTP in the third week. Additionally, there was an interaction between GTP level and duck on feed intake (second week) and BWG (first week). This interaction induced the significant difference of feed intake between BTDs and KDs in the second week. The diet supplemented 0.5% GTP that increased the feed intake of BTDs, in contrast, it decreased the feed intake of KDs (Table S3). In addition to feed intake, 0.5% GTP-induced KDs had lower BWG in the first week.3.3. Plasma Biochemical ParametersUnlike the growth performances, breeds had significant effects on BUN, CREA, UA, GOT, GPT, ALB, GLO, and TP (Table 4) in plasma biochemical values. The concentrations of CREA, UA, GOT, GPT, ALB, GLO, and TP in the plasma of KDs were lower than those in BTDs, but BUN in KDs was higher than that in BTDs. The concentrations of Ca, Mg, BUN, CPK, and ALK were affected by dietary T-2 concentrations (p < 0.05, Table 4). The concentrations of Ca, CPK, and ALK in the groups treated with 0.5 and 5 mg/kg of T-2 were lower than those without T-2 treatment. Magnesium concentration decreased in the group administered 5 mg/kg of T-2. In contrast to Mg concentration, BUN activity was increased in the blood plasma of the duck fed the diet containing 5 mg/kg of T-2. When the diet was supplemented with 0.5% GTP, lower concentrations of BUN, UA, CHE, GOT, and GPT were observed in the plasma of the ducklings compared to those fed without GTP; however, supplementation with 0.5% GTP increased the plasma concentrations of Ca, GLO, and TP. The BUN concentration of BTD in the group administered 5 mg/kg of T-2 without GTP supplementation was higher than the group not administered T-2 and GTP (Table S4). In addition, KDs fed the diet containing 5 mg/kg of T-2 and 0.5% GTP improved their BUN level compared to those fed with the same diet without GTP. Moreover, CPK activity was also affected by the interactions between T-2 and GTP (p < 0.0001, Table S5). Only the ducklings fed the diet without GTP had decreased CPK activity with the increasing T-2 concentrations. Glutamate oxaloacetate transaminase activity was not only associated with the effects of GTP and breed, but it was also under the influence of the two-way interactions between T-2 levels and breed, as well as the three-way interactions of T-2, GTP, and breed (Table 4). Furthermore, GOT activity of BTDs in the group that was administered 5 mg/kg of T-2 without GTP supplementation was higher than the group that had no T-2 treatment (Table S6). Moreover, BTDs fed the diet containing 5 mg/kg of T-2 but supplemented with 0.5% GTP had a reduced GOT level compared to those fed with the same level of T-2 but without GTP supplementation. Additionally, there was an interaction between GTP level and duck breed on ALK activity, which had a significant difference between BTDs and KDs. The diet supplemented with 0.5% GTP decreased ALK activity of BTDs, but increased ALK activity of KDs (Table S7).3.4. Relative Weights of Organs and CarcassBreed had a significant effect on the relative weights of all organs and carcass examined (p < 0.0001, Table 5). In general, organs of BTDs were heavier than those of KDs, but the breasts of KDs were heavier than those of BTDs. The relative weights of gizzards and legs were significantly affected by the effects of T-2 (p < 0.05) levels and an interaction between T-2 and breed (p < 0.001) was detected. The relative gizzard weight was increased in BTDs in both T-2-treated groups (Table S8). However, the relative leg weight was decreased only in BTDs treated with the highest (5 mg/kg) T-2, compared to the 0 mg/kg of T-2 group. Unlike the BTDs, the relative weights of gizzard and leg were not affected by T-2 concentrations in KDs. It is interesting that T-2 increased relative kidney weight but reduced relative breast weight in the 5 mg/kg of T-2-treated ducklings (Table 5), while those fed with the diet supplemented with 0.5% GTP lowered relative weights of heart, liver, and kidneys. In addition, relative left tibia weight was higher in the ducklings with 0.5% GTP supplementation. There was an interaction between GTP and duck breed (p < 0.05); the BTDs fed diets supplemented with 0.5% GTP had higher relative tibia weight (both left and right) when compared to that without GTP (Table S9). In contrast, the relative tibia weights were not affected by GTP supplementation in KDs.4. Discussion4.1. Antioxidative Capacity of Green Tea PowderGreen tea is a traditional Asian beverage and is known for its antioxidative capacity. The current study investigated the antioxidative capacity of GTP by determining the DPPH scavenging capacity, ferric ion chelating activity, and reducing power. A previous study [47] indicated that a positive correlation exists for other antioxidant capacity methods, such as DPPH-scavenging capacity, ferric chelating activity, and reducing power with green tea polyphenols. The GTP of the current study possessed more DPPH scavenging and reducing power than the GTP of other studies. By contrast, ferric ion chelating was lower [23,48]. Several studies have indicated that diets containing 0.5% GTP improved the growth performance, blood biochemical parameters, and the intestinal traits in broilers [49,50]. In addition, our previous tests showed that GTP at 4000 μg/mL had nearly 100% of DPPH radical scavenging activity and ferric ion chelating activity. The concentration of 4000 μg/mL was equivalent to 0.4% GTP supplementation in the duckling feed. Similar to the allowance of T-2 level used, 0.5% but not 0.4% GTP was decided to compensate potential experimental errors in the present study.4.2. Effects of T-2 Toxin and Green Tea Powder on the Growth Performance of DucklingsIn the present study, KDs (meat-type) demonstrated higher feed intake and BWG, and better FCR than BTDs (laying-type). As expected, breed type had a significant impact on the growth performance throughout the experimental period due to the genetics. It has been suggested that meat-type poultry have higher oxidative phosphorylation efficiency than laying-type poultry in their skeletal muscle mitochondria [51]. The basal metabolic rate of meat-type poultry is lower than that of laying-type poultry starting from hatching day until reaching 500 g of body weight [15]. These factors might help explain their better FCR and growth rates. It is known that T-2 could interfere the energy metabolism in poultry [16]. When the ducklings fed the diet containing 5 mg/kg of T-2, lower feed intakes and BWG were observed during the second and the third weeks. Furthermore, another study has suggested that T-2 could suppress the protein synthesis, BWG, and feed intake of chicken and piglets [16]. Interactions between T-2 and breed that decreased the feed intake only in BTDs fed the diet containing 5 mg/kg of T-2 in the first week. In other words, BTDs were more susceptible to T-2 toxicity than KDs in the growth performance. To our best knowledge, no study has been conducted and reported such differential sensitivity between meat- and egg-type ducklings in response to T-2 toxicity. It appears that such diverse physiologic responses are mainly due to the intensive selection and breeding of KDs, although more studies are required. Different from the feed intake and BWG, the FCR of both BTDs and KDs remained unaffected by T-2 toxicity, which is similar to the T-2 effects on White Roman geese [52]. Besides the genetic difference, the reasons behind the indiscernible adverse effects of T-2 toxicity on the growth performance of these ducklings remain unexplained. Nevertheless, supplementation of GTP in duck diets had a significant interaction with breed. The differential breed effects further intensify the GTP inhibition on feed intake and BWG in KDs. A similar phenomenon was also found in broilers, to which 1% GTP was fed, which resulted in reduced feed intakes [53].It has also been demonstrated that green tea polyphenols could alter the phosphorylation efficiency of mitochondria in rodents [54]. Additionally, our study demonstrated that GTP supplementation also inhibited the toxicity of T-2 (5 mg/kg) in terms of BWG only in BTDs. Therefore, GTP could serve as a detoxicant to protect BTDs against the T-2 toxicity, perhaps in poor growth conditions.4.3. Hematological Alterations after T-2 Toxin Challenge Followed by GTP SupplementationHepatic and renal tissues are rich in metabolism-associated enzymes. Damages to these organs often lead to the release of enzymes into the bloodstream, resulting in the elevation of enzymatic activities in peripheral circulation [55,56]. Our analyses showed that T-2 altered plasma biochemical parameters, including hypocalcemia, hypomagnesemia, and low concentrations of CPK and ALK in the blood plasma of both KDs and BTDs. These effects were supported by previous studies, in which blood Ca and Mg concentrations of the intoxicated chicken (treated by 4 mg/kg of T-2) were significantly decreased [57,58]. Other studies have also shown that T-2 induced hypocalcemia by decreasing 1,25(OH)2D3 receptors in small intestine mucosa [59,60]. When animals show symptoms of hypomagnesemia, parathyroid hormone will be insufficiently produced and hypocalcemia could be further aggravated [61], because Mg is required for the secretion of parathyroid hormone [62].Alkaline phosphatase plays an important role in the calcification of cartilage and bone via the increased concentration of inorganic phosphates by hydrolyzing phosphate-esters for bone mineralization [63]. Previous studies have demonstrated that diets containing T-2 reduce ALK concentrations in poultry [64,65]. Administration of T-2 can cause hypocalcemia, likely due to a decrease in Ca absorption and inactivation of ALK in small intestine mucosa; in turn, a pathological syndrome, such as spongy bones or osteoporosis, could be observed [60].Creatine phosphokinase plays an essential role in the energy metabolism of all tissues, particularly in skeletal and cardiac muscles [66]. In contrast to the present study, Edrington et al. [67] indicated that growing broilers fed with T-2 diet (6 mg/kg) for 3 weeks showed no difference in CPK concentrations. It is likely that such different responses in CPK concentrations between ducklings and growing broilers are species-specific (ducks vs. chicken).Blood urea nitrogen and GOT are two markers that indicate renal injury and liver damage, respectively [68]. In the present study, the treatment of 5 mg/kg of T-2 increased BUN and GOT of the blood plasma in KDs and BTDs, respectively, suggesting the malfunction in both kidneys and livers of the ducklings. It has been reported that broilers that received 3 mg/kg of T-2 for 5 weeks displayed higher concentrations of peripheral GOT and BUN [69]. Another study on broilers treated with only 927 μg/kg of T-2 for 3 weeks caused their bile tract lesions [70]. A similar response was also observed in Japanese quails fed with 4 mg/kg of T-2 for 5 weeks, and an elevation of GOT concentration was detected, with no changes in BUN concentrations [64]. When the diet was supplemented with 0.5% GTP after T-2 treatment (5 mg/kg), an effective improvement of the abnormal BUN and GOT concentrations in the KDs and BTDs, respectively, was observed. Previous studies have indicated that GTP could effectively reduce BUN and GOT concentrations [28,71,72]. Other biochemical parameters, such as the concentrations of GLO and TP, were also increased in ducklings fed a diet supplemented with 0.5% GTP. These results were consistent with previous studies in rats and rainbow trout [73,74]. In general, TP and GLO are involved in several physiological processes, including the transport of ions, hormones, and lipids [75]. The increased concentrations of TP and GLO indicated that GTP might also have played a role in regulating certain physiological mechanisms in ducklings.4.4. Effects of T-2 Toxin and Green Tea Powders on the Relative Weights of Organs and CarcassIn terms of relative organ weight, T-2 induced gizzard enlargement in BTDs, which is supported by previous study in chicken [57]. Furthermore, a decrease in relative breast and leg weights of BTDs was observed when fed with relatively high level of T-2. It might be caused by the altered metabolisms of essential amino acids and the inhibition of protein synthesis by T-2 toxicosis [76,77]. We found that relative kidney weights increased in ducklings fed the diets that contained 0.5 and 5 mg/kg of T-2. However, other studies in broilers [57,67] demonstrated that their relative kidney weights did not differ from the control group, when broilers were fed with 4–6 mg/kg of T-2 for 3 weeks. The differential responses in relative kidney weight between ducklings and broilers remain undetermined. Additionally, it is evident that GTP increased the relative tibia weights (both right and left) in BTDs, which is supported by previous studies [78,79]. Although more in-depth studies are required, the present study found that GTP enhanced the strength of the tibia and was likely also beneficial to growth and bone development.4.5. Detoxification and Antioxidative Activities of Green Tea PowderIt has been reported that T-2 can cause oxidative stress by producing ROS [39,80]. Studies have shown that GTP or green tea extract possesses a strong antioxidant activity, reducing power and scavenging capability on ROS and free radicals [29,81]. The antioxidative effects of GTP are mainly due to its polyphenolic compounds, particularly for the large amount of epigallocatechin (EGC) and epigallocatechin gallate (EGCG) [47]. Epigallocatechin and EGCG, in addition to their antioxidative activity, have also been indicated to possess detoxicant properties against trichothecenes in vitro [82,83]. Other green tea polyphenolic compounds, such as C, CG, EC, ECG, GC, and GCG, also had both antioxidative and DNA protection properties [37]. Several studies have indicated that GTP has the capacity to ameliorate hepatotoxicity and genotoxicity caused by mycotoxins including AFB1, FB1, or citrinin [29,36,37]. In this study, the improved BWG and GOT or BUN level also confirmed that dietary supplementation of GTP could partially detoxify T-2 toxicity in ducklings.5. ConclusionsIn the present study, we have demonstrated that BTDs and KDs had differential responses to T-2 toxicity and GTP detoxification. More sensitive responses to T-2 and GTP supplementation in BTDs were demonstrated in their growth performances, blood plasma parameters, and some organ weights, as well as carcass weights. Moreover, GTP specifically improved the T-2-caused toxicity, including the reversal of hepatic or nephrotic indexes, such as GOT and BUN, in the T-2-treated ducklings. For practical feeding and management of waterfowls, the use of GTP as their dietary supplement can be an option for detoxifying animal consumption of trichothecene contaminated feeds or ingredients.

animals : an open access journal from mdpi

10.3390/ani12070879

PMC8997082

Fox tapeworms are pathogens that need two hosts to complete their life cycle. The reproductive part of the cycle takes place in foxes and the intermediate step usually occurs in rodents. Muskrats can also act as the intermediate host for fox tapeworms. This increases the total number of potential hosts which can also increase the number of pathogens in the environment. In Belgium, where muskrats are non-native, this is one of the ways in which an invasive alien species can have a negative effect on its new environment. From 2009 to 2017 all muskrats caught in Flanders and across the border with Wallonia and France were collected and dissected to estimate how often they were infected with fox tapeworms. Visual examination of the livers of 15,402 muskrats revealed 202 infected animals (1.31%). In Flanders, we found 82 infected animals out of 9421 (0.87%). The percentage of infected animals did not increase during the research period. All the infected animals in Flanders were found in municipalities along the Walloon border. We did not observe a northward spread of EM infection from Wallonia to Flanders.

One way in which invasive alien species affect their environment is by acting as pathogen hosts. Pathogens limited by the availability of the native host species can profit from the presence of additional hosts. The muskrat (Ondatra zibethicus) is known to act as an intermediate host for the fox tapeworm (Echinococcus multilocularis). From 2009 to 2017, 15,402 muskrats caught in Flanders and across the border with Wallonia and France were collected and dissected with the aim of understanding the prevalence of this parasite in muskrats. Visual examination of the livers revealed 202 infected animals (1.31%). Out of the 9421 animals caught in Flanders, we found 82 individuals (0.87%) infected with E. multilocularis. No increase in prevalence was observed during this study. All of the infected animals in Flanders were found in municipalities along the Walloon border. We did not observe a northward spread of E. multilocularis infection from Wallonia to Flanders. We hypothesise that the low prevalence is the result of the reduced availability of intermediate hosts and the successful control programme which is keeping muskrat densities in the centre of the region at low levels and is preventing influx from other areas. Our results illustrate that muskrats are good sentinels for E. multilocularis and regular screening can gain valuable insight into the spread of this zoonosis.

1. IntroductionInvasive Alien Species (IAS) represent a significant source of human-mediated introduction of pathogens to a new host or region [1]. They can facilitate the spread of infectious diseases, either by acting as an alternative host for native pathogens or by bringing in pathogens from their native range. These pathogens can qualify as emerging infectious diseases (EIDs), some of which form a significant threat to public health [2]. When an invasive species becomes an alternative host for native parasites this leads to either spill back, where the presence of the additional host increases disease impacts in native species [3], or dilution, where more diverse host communities inhibit the spread of parasites [4]. Well known examples of diseases disseminated by IAS or exotic pet trade are chytridiomycosis in amphibians, squirrel pox, and crayfish plague [5,6,7].The muskrat (Ondatra zibethicus) is a rodent native to swamps and wetlands in North America and an invasive alien species (IAS) in Europe. The species was first introduced in 1905 in Europe as a furbearer [8] and has spread to suitable habitats in most European countries where they cause damage to dykes and agricultural crops. Muskrats can also be carriers of diseases such as tularemia and leptospirosis and they can carry parasites, such as tapeworms, most commonly Taenia taeniaeformis [9]. Because of their economic, human health, and environmental impacts, muskrats are managed in several European countries but most intensively in Belgium and the Netherlands. Management practice consists mostly of systematic trapping to reduce muskrat densities, lowering the levels of muskrat damage [10]. In Flanders, an intensification of the trapping effort since 2000 resulted in a spectacular decline of the muskrat population, as shown by the number of muskrats caught per year: from over 60,000 animals in 2001 to just over 5000 in 2020 [11].Fox tapeworm (Echinococcus multilocularis) (EM) is a parasite of which the adult stage lives in the intestines of foxes and occasionally other carnivores [12]. The adult tapeworm is only a few millimeters in size and consists, on average, of five proglottids or segments [13]. Fox tapeworms incubate for 26 to 29 days, after which they produce up to 100,000 eggs per day [14]. This can last up to four months [15]. Once the eggs are released into the environment, they can be ingested by intermediate hosts. Several mammal species can act as intermediate hosts for EM, including the European beaver (Castor fiber), brown hare (Lepus europaeus) and coypu (Myocastor coypus) [16]. However, voles (Arvicola sp., Microtus sp.) and muskrats are considered the main intermediate hosts in Europe [15,17].Muskrats ingest the infective eggs through contaminated food and vegetation. The degree to which the species is infected indicates the presence of eggs in the environment and thus the potential risk to humans. It can be considered as a bioindicator for the presence of EM in new areas [18,19]. Moreover, muskrats show a greater chance of infection than the naturally present intermediate hosts and can therefore function very well as sentinel species [20].Echinococcus multilocularis can develop into a life-threatening infection in humans, alveolar echinococcosis (AE), and is a species of great concern in the context of public health. Humans can act as an aberrant host once they become infected with the eggs of the parasite through contact with infected animals or their faeces. One of the infection routes is contaminated food such as mushrooms, wild berries, or vegetables from gardens. Most infections occur in people active in forestry, horticulture, agriculture, or hunting, but also pet owners [15,21,22]. The eggs will develop to alveolar hydatid cysts, mainly in the liver, leading to AE. Due to the strong increase in red fox (Vulpes vulpes) populations in Europe at the end of the last century [23,24,25], the concern for this serious parasitosis also grew [13,15]. In 2020 there were 115 confirmed cases of AE in the EU, 10 of those were from Belgium [26].Southern Germany, Switzerland, and Alpine Austria are historically considered the endemic area for this infection in Europe [27]. This area has been expanding to the north, west, and east since the 1970s [28]. For example, Combes et al. [29] observed a strong westward spread of the fox tapeworm in red foxes in France. In the Dutch province Limburg, a northward expansion was found moving at a speed of 2.7 km per year [30]. Vervaeke et al. [31] also found a northwestward spread of infected foxes in Belgium, starting from the south of Belgium. However, they also observed that in 2000–2002 this spread stabilized at the regional border. Other studies have also confirmed this difference in the prevalence of EM in foxes between Flanders (northern Belgium) and Wallonia (southern Belgium): Wallonia has a prevalence between 5 and 25%, while the prevalence in Flanders ranges between 0 and 1% [32]. Most infected red foxes found in Flanders are found close to the regional southern border [33,34].From 1994 onward, muskrats from Flanders have been autopsied with the aim of understanding their demography. In 2008, we first detected EM in a muskrat from Lessines (Wallonia, Belgium), just across the border with Flanders. Therefore, from 2009 onward, all available muskrats were collected and examined with the aim of answering the following questions:What is the prevalence of EM in muskrats captured in Flanders?Does this prevalence show an increase over time?Does the geographical range of EM infected muskrats expand?2. Materials and MethodsMuskrats were systematically collected by the Flanders Environment Agency, the main trapping organization in Flanders, in the period from 2009 to 2017. Additionally, muskrats caught just across the border in Wallonia and France were also examined, since this was an area of particular interest. During this period, 17,497 muskrats were caught by the Flanders Environment Agency (Figure 1). In the same period, 15,402 muskrats were dissected and scored for EM. Thus, 12% of all specimens caught during this period were missing. These were animals that had decayed too much or animals to which no trapping location could be assigned.All reported infection rates are based on dissected animals caught in Flanders between 2009 and 2017 unless stated otherwise. More general results, i.e., monthly weight distribution, are based on the total dataset collected between 1994 and 2018.Animals were stored at −20 °C pending autopsy to neutralize the protoscoleces and/or germinative cells of the EM metacestode. Animals were then cleaned of dirt (mud, leaves, stones, etc.) and excess water in the fur. Each animal was weighed, sexed, and examined for injuries or particularities. We checked the abdominal cavity for the presence of macroparasites. This involved screening the liver surface for irregularities, discolourations, and lesions. We scored the EM presence as positive, negative, or inconclusive. In most cases the presence of EM was clearly recognizable, however, when this was not the case, samples of the presumed infected liver tissue were taken for confirmation and examined under a light microscope to check for the presence of protoscoleces.Data were grouped by year, sex, month, and weight, and the significance of observed differences between groups was tested using a chi-square test or a Fisher’s Exact Test if subgroups were too small. The data were analysed using R [35].3. ResultsVisual inspection of 9421 livers revealed 82 infected animals (0.87%, CI 0.70–1.01%). In all regions combined, we found 202 infected animals out of 15,402 (1.31%, CI 1.14–1.50%), with no infections occurring in the 2477 animals trapped in France. The parasite mainly occurred in the western part of the Flanders-Wallonia border (Figure 2) and was found in muskrats from this area yearly. As of 2013, positive cases were also found in the eastern part of the border, with the most eastern observation occurring in 2017 in Tongeren. All cases were found along the administrative border with Wallonia, no northward spread was observed.The overall prevalence of EM in muskrats fluctuated around an average of 0.9% (range 0.2–1.9%). Visual interpretation of these results showed no up- or downward trend over the years (Figure 3) and there was no significant correlation between year and prevalence (cor = −0.34, CI = −0.82–0.41, p = 0.36).3.1. Sex Ratio and Infection RateWe find a slightly higher number of infected animals among males compared to females (Figure 4), however there was no significant effect of sex on infection rate (Χ2 = 1.21, df = 1, p-value = 0.27).3.2. Weight/Age Distribution and Infection RateThe rate of infection in muskrats under 500 g was particularly low in comparison with heavier/more adult animals (Figure 5). The number of inconclusive cases also increases with weight (Figure 5). There was a significant effect of weight on infection rate (p-value < 0.005).3.3. Infection Rate and MonthThe highest percentage of infected muskrats was found in spring (Figure 6). We found a significant effect of month on infection rate (Χ2 = 144.89, df = 11, p < 0.005). Given the seasonal reproduction of the muskrat, the age structure of the population changes during the year. The percentage of young animals increases drastically from May on, while fewer young are born in autumn. The spring population consists mainly of subadults and a smaller fraction of older animals (>1 year) (Figure 7). This could explain the seasonal difference in prevalence.4. DiscussionThe prevalence of EM in Flemish muskrats fluctuated around 0.9%, indicating low infection rates in the region. The prevalence also remained fairly constant throughout the sampling period (2009–2017). Echinococcus multilocularis positive muskrats were mainly caught in the western part of Flanders, yet additional positive cases emerged in the eastern province of Limburg. Overall, in Flanders, the parasite was found only in muskrats caught close to the regional border.Using body weight as a proxy for age (based on our findings in Figure 7), older muskrats were found to have higher infection rates than younger animals. From weaning, a young muskrat runs the risk of becoming infected with tapeworm eggs. The probability of infection increases with the time an animal is present in an infected area. After ingestion of EM eggs, the likelihood of detecting cysts in the liver also gradually increases. Based on these two arguments, an increase in the risk of a visible infection can therefore be assumed with increasing age. We also found the number of inconclusive cases increasing with weight. This may be due to an accumulation of other parasites in the liver. For example, it appears that there is also a positive correlation between the body weight of the muskrat and the degree of infection with T. taeniaeformis (correlation = 0.49, p < 0.001) (unpublished results). The fact that no infected animals with a weight >1500 g were found could be explained by the fact that this is a very small cohort (75 animals), that it might be harder to distinguish EM infection from other sometimes massive infections with T. taeniaformis, or that infected muskrats have a higher mortality risk.The univariable approach followed here to describe patterns of infection across years, season, and demographic parameters (sex, age) of muskrats inherently represent limitations. Although based on an extensive dataset of autopsied muskrats, in reality, the epidemiology of echinococcosis is complex, and demographic, spatial, and temporal variables might exhibit multiple interaction effects contributing to observed infection status [36]. Additionally, the effect of muskrat management has not been explicitly tested. Risk assessment for this zoonosis also ideally needs to consider all intermediate and potential end hosts, including foxes and humans. Further multivariate analysis is needed to reveal such complex interaction patterns.Whether the parasite is permanently present in Flanders or whether most (infected) animals are migratory muskrats that have contracted the infection in Wallonia remains open for discussion. The second option seems more likely as infection rates for both foxes and muskrats are much higher in Wallonia, especially south of the Meuse River [12,32]. Muskrats also have a higher chance of reproducing and living long enough in the environment to ingest and develop the infection due to lower trapping efforts in Wallonia [37]. Muskrats migrate mainly in the spring as young adults when they look for a mate, with males migrating further than females [38]. More males are indeed caught in Flanders. We do find a similar sex ratio for additional observations from Wallonia, however, these observations were all made very close to the Flanders-Walloon border and could therefore also experience a high influx of migratory animals. When looking at muskrats collected more to the south of Wallonia [39] we see a different, more equal sex ratio (e.g., 53.1% male, 46.9% female). On the other hand, little is known about the symptoms and mortality of EM infection in muskrats. Therefore, it is unclear how likely it is that an infected animal will migrate over long distances. Examination of the natural intermediate hosts along both sites of the Flemish Walloon border could reveal this.Vervaeke et al. [31] predicted that the expansion of fox tapeworm seen in Wallonia would continue northwards in Flanders, but this was not reported in more recent studies [33,34]. Genetic research by Knapp et al. [40] showed that the observed changes in distribution seen all over Europe were mostly the result of increased attention for the disease rather than an actual geographic expansion. This leaves the question of why the infection has never entered Flanders, especially with the reoccurrence of the red fox in recent decades. A possible explanation is that there are too few suitable intermediate hosts in Flanders to complete the cycle. Although few concrete figures are available on this, the hypothesis that Microtus spp. are relatively rare in Flanders [34] seems justified. According to the Flemish Red List of mammals [41], similar to the two bigger vole species, three of the four smaller vole species (Short-tailed field vole (Microtus agrestis), European pine vole (Microtus subterraneus), and common vole (Microtus arvalis)) are classified in the ‘near threatened’ category. In Wallonia, on the other hand, almost all small vole species are still common and there are no indications of a decline [42]. It is striking in this regard that these species are not the staple food of the fox in Flanders, as is generally the case [43]. Instead, the brown rat (Rattus norvegicus) is the primary prey species [34,44]. We hypothesize that successfully keeping the muskrat population down has also kept the potential pool of intermediate hosts small, thus helping to prevent the spread of the fox tapeworm.Due to targeted, intensive control, muskrats no longer occur all over Flanders. The vast majority of the animals are caught at the regional border. Since the fox tapeworm could still be spread across Flanders in foxes and other intermediate hosts, monitoring the fox tapeworm in muskrats alone is therefore not the best proxy to gain insight into the general occurrence of the tapeworm. However, since muskrats are considered as a very good sentinel species for EM, regular screening of the muskrats in Flanders and especially near the Flemish Walloon border could still be valuable to gain insight into a possible spread of this zoonosis.5. ConclusionsFlanders currently has a low prevalence of EM in muskrats and there has been no increase in prevalence during this study. The infected animals were all located in the regional border zone. Our results illustrate that muskrats are good sentinels for E. multilocularis and regular screening can gain valuable insight into the spread of this zoonosis.

animals : an open access journal from mdpi

10.3390/ani13091445

PMC10177288

Broilers were fed increasing levels of supplemental choline chloride in diets where methionine was minimally reduced (0.15%), reared under summer conditions for 41 days, and then processed. Supplemental choline improved broiler feed efficiency by reducing feed intake without altering body weight gain. Increasing dietary choline concentrations also increased carcass yield, breast yield, and the incidence and severity of wooden breast. It can be concluded that when broilers are reared under high environmental temperature and fed diets with reduced methionine, supplemental choline chloride can positively impact growth performance and carcass yields.

Choline has been demonstrated to partially substitute methionine in broiler chicken diets due to their interconnected biosynthesis pathways. Yet, research on the impacts of dietary choline supplementation on modern strains of high-yielding broilers is limited. The objective was to evaluate the effect of increasing additions of choline chloride on the performance and carcass characteristics of broilers fed reduced methionine diets and reared under summer environmental conditions. Ross 708 x Yield Plus male broilers were reared for 41 days on used litter in floor pens (n = 2232; 31 birds per pen). Birds were fed one of six corn and soybean meal-based, reduced methionine diets containing 0, 400, 800, 1200, 1600, or 2000 mg of added choline chloride per kg of feed. Diets were provided in three phases. On day 43, 10 birds per pen were processed. Increasing dietary choline resulted in similar body weight gain, reduced feed intake, and improved feed efficiency. Choline chloride supplementation linearly increased both breast and carcass yields while concomitantly increasing the incidence and severity of wooden-breast-affected fillets. These results indicate that supplementing reduced-methionine broiler diets with choline chloride during high environmental temperatures may improve feed efficiency and increase carcass and breast yields but may also increase wooden breast.

1. IntroductionCholine is a nutrient similar to an amino acid that functions as an important precursor for cell membrane lipids, neurotransmitters, and methylated compounds. Its role as a methyl donor in the biosynthesis of compounds containing methyl groups is important for the production and availability of methionine (Met) [1,2]. Choline enters the s-adenosyl methionine (SAM) pathway when irreversibly oxidized to betaine. The methyl groups in betaine are then passed to a derivative of folic acid that is later converted to Met. SAM is formed from Met and acts as a high-energy methyl donor that is important for many anabolic reactions. In cases where choline is insufficient to fuel the regeneration of Met, Met is recovered from protein breakdown in the body [3].Choline can be synthesized de novo from ethanolamine in all nucleated animal cells, specifically in the liver where phosphatidyl choline is needed for lipoprotein synthesis. [4] However, de novo choline synthesis alone does not appear to be sufficient, and choline concentration in the body depends heavily on the presence of adequate Met, betaine, and folic acid [5,6]. For this reason, it is difficult to assess dietary choline requirements without also controlling SAM pathway intermediates.Animals rely on dietary sources of preformed choline to make up for poor de novo synthesis. Fatty liver development that is caused by a deficiency in dietary choline [7] has been implicated as a potential way that inadequate choline impacts the performance of production poultry as choline plays a role in lipid transport and metabolism [8,9]. Reduced fat content in the liver has been observed in laying hens and broiler breeders fed supplementary choline [10,11]; however, additional choline did not alter the egg production or feed efficiency of broiler breeders. Supplementing 1200 mg of dietary choline per kg of feed to Ross 308 broiler chicks caused an increase in feed intake but no change to feed conversion when dietary Met concentrations were above broiler nutrient recommendations [12]. Jahanian & Ashnagar [13] reported reduced feed intake and improved FCR in Ross 308 broilers fed additional choline.Heat stress is a common issue facing the broiler industry, specifically during the summer months in the Southeastern United States. Heat stress negatively impacts growth performance, increases mortality, and can alter carcass characteristics [14,15]. Betaine, which can be directly oxidized from choline, is thought to protect against heat stress due to its osmoregulatory qualities that can help decrease cellular dehydration [16]. Mahmoudi et al. [17] evaluated the effects of replacing dietary Met with choline and betaine in Ross 308 broilers exposed to heat stress. These authors did not observe differences in average daily feed intake or body weight gain in birds fed either choline, betaine, or a combination of the two, thus suggesting that Met can be replaced with choline and betaine in heat-stressed broilers. While it has been observed that supplementing broiler diets with betaine is more efficient than choline at increasing betaine concentrations in the liver where choline is converted to betaine [18], Kpodo et al. [19] reported similar plasma betaine concentrations in broilers fed diets supplemented with either choline or betaine. This suggests feeding either choline or betaine can have similar metabolic benefits as they result in comparable increases in circulating betaine.Feeding increasing additions of choline chloride on top of the basal choline provided by a corn and soybean meal-based diet to high-yielding, large-frame Ross 708 x Yield Plus broilers did not previously result in improvements in growth or carcass traits when dietary Met was adequate and rearing temperatures were maintained according to bird comfort [20,21]. Therefore, the aim of the present study was to evaluate the effects of increasing concentrations of supplementary choline chloride on growth performance and carcass characteristics of broilers fed diets with reduced Met under summer rearing conditions.2. Materials and MethodsAll procedures involving live birds were reviewed and approved by the Auburn University Institutional Animal Care and Use Committee (PRN 2020-3749).2.1. Diet FormulationSix experimental corn and soybean meal-based diets were manufactured in the Auburn University Poultry and Animal Nutrition Center Feed Mill. Diets were formulated according to 2019 Aviagen Nutrition Specifications [22] with the exception of a 15% reduction in digestible Met. Choline chloride was supplemented at either 0, 400, 800, 1200, 1600, or 2000 mg per kg of feed on top of the intrinsic concentration of choline provided by the ingredients of the basal diet as used in previous experiments [20,21]. Diets were fed in 3 phases: crumbled starter from d 0 to d 15, pelleted grower from d 16 to d 28, and pelleted finisher from d 29 to d 41 (Table 1). All diets were third-party tested for total choline content using ion chromatography (Table 2; Eurofins Nutrition Analysis Center, Des Moines, IA, USA).2.2. Broiler HusbandryDay-old Ross 708 x Yield Plus male broiler chicks were transported to the Auburn University Charles C. Miller, Jr. Poultry Research and Education Center in Auburn, AL. Chicks were randomly allotted into 72 pens (12 replicates per diet; n = 31 birds per pen), and pen weights were obtained. Chicks were placed into floor pens (0.075 m2 per bird) on used litter in a plenum-style poultry house. Both water and feed were provided ad libitum. Ambient temperature set points started at 34.4 °C at the time of chick placement and decreased in temperature to reach a final set point of 26.7 °C on day 21 through the remainder of the study. Birds received 23 h of light from 0 to 3 days of age, 20 h of light from 4 to 7 days of age, and 16 h of light from day 8 onwards. Light intensity was set to 30 lux from 0 to 7 days of age, 10 lux from 8 to 14 days of age, and 5 lux from day 15 through the end of the study.2.3. Broiler Growth Performance and Carcass Part YieldsMortality-corrected feed intake (FI), body weight gain (BWG), and feed conversion ratio (FCR; FI:BWG) were calculated using pen weights on day 0 and individual bird body weights on days 15, 28, and 41. Mortality was recorded on a twice daily basis to account for body weight gained by birds that died before the end of a feeding phase. On day 43, 720 birds (n = 10 birds per pen) were processed at the Auburn University Fortenberry Poultry Processing Plant following an 8-hour fast. Birds were selected for processing based on proximity to the median individual day 41 body weight of each pen. Individual fasted, live body weights of selected birds were recorded prior to slaughter, and hot carcass weights were recorded immediately following evisceration. Carcasses were then chilled in a static water bath for 3 h, and cold carcass weights as well as abdominal fat pad weights were determined. Carcasses were deboned the next day by professional commercial deboners. Carcass parts were separated, and weights of the following parts were recorded: boneless and skinless breasts (Pectoralis major) without rib meat, tenders (Pectoralis minor), bone-in and skin-on wings, bone-in and skin-on drumsticks, and bone-in and skin-on thighs. Hot carcass yield was calculated as a proportion of the fasted, live body weight, and carcass part yields were calculated as proportions of the cold carcass weight.2.4. Wooden Breast and White Striping ScoringBreast fillets were assigned Wooden Breast (WB) and White Striping (WS) severity scores using a 4-point scale (0 = normal, 1 = mild, 2 = moderate, and 3 = severe). WB was evaluated based on the degree of palpable hardness of each fillet using a tactile examination of both sides of the deboned breast. Visual evaluation of WS was conducted concurrently to ascertain the proportion of both sides of the breast containing white striations as previously described [23]. Fillets with a score of 0 were free of defects and considered normal, fillets with a score of 1 were up to ¼ affected, fillets with a score of 2 were up to ½ affected, and fillets with a score of 3 were greater than ½ affected. All fillets were scored by one pre-trained evaluator. Scores for both WB and WS are presented as a proportion of all breasts within the treatment group and as a mean score for each added choline chloride treatment.2.5. Statistical AnalysisA randomized, complete block experiment with 12 replicate pens per treatment was designed. Replicate pens represented each of the 6 dietary treatments (0, 400, 800, 1200, 1600, and 2000 mg of added choline chloride per kg of feed). The GLIMMIX procedure of SAS software, ver. 9.4, was employed to analyze all data as a one-way ANOVA. Polynomial orthogonal contrasts were also evaluated. The pen served as the experimental unit with location as the blocking factor. The fixed effect was the supplemental choline chloride dietary treatment, and degrees of freedom were corrected using the Satterthwaite adjustment. The events/possible events syntax with a binomial distribution and R-side covariance structure in SAS was used to evaluate proportional data (e.g., mortality and carcass yields). The PDIFF option was used to conduct all least squared means separations; means were declared different when p < 0.05, and tendencies were declared when 0.05 < p ≤ 0.10.3. Results3.1. Growth PerformanceAs shown in Table 3, supplementing broiler diets with choline resulted in improved feed intake during the finisher and overall growth period. This change in feed intake was observed without altering BWG. Changes in FCR do not appear to follow a specific trend across treatments during the starter and grower phase. However, FCR was reduced in broilers fed 800 or more mg of added choline chloride per kg of feed during the finisher phase. Orthogonal contrasts revealed that the reduction in FCR as choline supplementation increases overall from day 0 to 41 of age was linear (p < 0.0001). Mortality was not impacted by dietary choline supplementation (Table 4).3.2. Carcass CharacteristicsTable 5 reports data obtained from processing at day 43 of age. Hot Carcass HC yield was calculated as a proportion of d 43 fasted live body weight (FLBW). HC yield increased as choline supplementation increased. Dietary treatments of 1200 or more mg of additional choline chloride per kg of feed resulted in the heaviest breast weights and greatest breast yields, thus generating a 78 g improvement when comparing the control with the 2000 mg. Conversely, supplementing 2000 mg of choline chloride decreased abdominal fat pad yield compared with birds fed no additional choline. Similar effects were observed with the reduction in wing, thigh, and drumstick yields as dietary choline increased.3.3. Wooden Breast and White StripingThe incidence and severity of WB and WS myopathies were assessed and displayed in Table 5. The incidence of severe WB, or breasts assigned a score of 3, increased as supplementary choline concentrations increased. A greater number of breasts from birds not fed additional choline were assigned a score of 0 and lacked signs of WB development than all other groups. This follows a similar pattern to breast yield. The incidence of severe WB appears to increase as breast yield increases. Average WB scores also increased with supplemental choline inclusions in the diet; however, average WB scores were below 2 for all treatments. The incidence of WS score 3 tended to increase as supplemental choline increased. Otherwise, no differences were observed in the incidence or severity of WS among dietary treatments.4. DiscussionAdding supplemental choline to broiler diets may spare low dietary concentrations of Met as shown by the observed improvements in feed efficiency as choline concentration increased without altering BW gain. This is in agreement with a previous study that found supplementing choline in a Met-deficient diet results in a similar growth performance to broilers fed sufficient concentrations of Met [17]. The combination of reduced Met and high-temperature conditions likely amplified the positive impacts of dietary choline on feed efficiency as Gregg et al. [20,21] did not observe changes in growth performance of broilers fed increasing concentrations of choline chloride when dietary Met was sufficient and rearing temperature was optimal. However, supplementation of dietary choline has not been previously demonstrated to alleviate losses in body weight gain associated with heat stress [24]. This suggests that the Met sparing effect of dietary choline may have more of an impact on broiler performance in the present study than any potential impacts choline has on growth during heat stress.The observed increase in breast yield combined with a reduction in wing, thigh, drumstick, and abdominal fat yield suggests that additional dietary choline may aid in diverting more nutrients to the growth of lean muscle than the less-desirable carcass parts and adipose tissue. This is similar to results reported by Waldroup et al. [25] who observed improvements in breast yield with added dietary choline, but did not observe an interaction between choline and Met concentrations in the diet. Jahanian & Ashnagar [13] studied the lipotropic effects of adding choline to broiler diets. These authors reported an increase in carcass yield and a reduction in abdominal fat in birds fed additional choline. Interestingly, they also noted a reduction in analyzed fat content of the breast muscle in choline-supplemented broilers. This is consistent with the reduction in abdominal fat pad yield observed in broilers supplemented with the greatest concentration of choline in the present study. In swine, dietary choline was found to improve fatty acid oxidation and reduce circulating free fatty acids and triglycerides [26]. Therefore, additional dietary choline may promote leaner broiler carcasses due to its role in lipid metabolism.The greatest increases in both WB severity and breast yield occurred with the supplementation of an additional 1200 or greater mg of choline chloride per kg of feed. This increase in WB incidence and severity is consistent with current knowledge of WB as the development of the myopathy strongly correlates with increased growth rate and breast yields [27,28]. Therefore, it is likely that the observed increases in WB incidence can be attributed to improvements in breast yield as there is no known mechanism for the role of choline in the development of WB. Increasing levels of dietary choline appear to protect against a reduction in feed efficiency caused by high environmental temperatures utilized in this experiment. Heat stress is understood to result in diminished growth performance in broilers [14]. Previous literature shows that decreases in WB incidence in heat-stressed birds is related to a reduction in growth performance and breast yield [29,30]. Based on the results of the present study, increasing additions of dietary choline may counteract the impacts of heat stress on growth performance, specifically reduced feed intake and poor feed efficiency. This would negate the reduction in WB commonly observed under heat stress conditions, therefore helping to explain the increase in WB incidence with increased choline when reared under high-temperature conditions.5. ConclusionsFeeding additional choline to high-yielding broilers improved overall feed efficiency, and supplementing 1200 or more mg of choline chloride per kg of feed increased carcass and breast yields when Met was insufficient. This evidence supports the supplementation of choline chloride in high-yielding broiler diets where Met is limited or as a partial substitute for Met. The observed improvements in breast yield resulting from choline supplementation appear to lead to an increase in WB severity. Given the osmoregulatory qualities of betaine and its ability to be oxidized to choline, dietary choline supplementation may also be beneficial in improving growth efficiency in broilers reared during summer months when heat stress is common.

animals : an open access journal from mdpi

10.3390/ani11030863

PMC8003097

The amount of electromagnetic field (EMF) in the environment emitted by electrical and electronic devices, mobile phone masts, or power lines is constantly increasing. Honey bee can be exposed to the EMF in the environment, and the influence of this factor on bees is still under consideration. Studying the impact of EMF on honey bees can give valuable information about whether it poses a threat to them. The honey bee is an important pollinator, playing a significant role in maintaining biodiversity and food production. Our research showed that a 50 Hz electric field at various intensities reduced the number of occurrences of walking, contacts between individuals, and self-grooming, and increased the activity of proteases, which are involved in the immune system response.

The effect of an artificial electromagnetic field on organisms is a subject of extensive public debate and growing numbers of studies. Our study aimed to show the effect of an electromagnetic field at 50 Hz and variable intensities on honey bee proteolytic systems and behavior parameters after 12 h of exposure. Newly emerged worker bees were put into cages and exposed to a 50 Hz E-field with an intensity of 5.0 kV/m, 11.5 kV/m, 23.0 kV/m, or 34.5 kV/m. After 12 h of exposure, hemolymph samples were taken for protease analysis, and the bees were recorded for behavioral analysis. Six behaviors were chosen for observation: walking, flying, self-grooming, contact between individuals, stillness, and wing movement. Bees in the control group demonstrated the highest number of all behavior occurrences, except flying, and had the lowest protease activity. Bees in the experimental groups showed a lower number of occurrences of walking, self-grooming, and contacts between individuals than the control bees and had significantly higher protease activity than the control bees (except that of alkaline proteases in the 23.0 kV/m group).

1. IntroductionThe amount of electromagnetic field in the environment emitted by electrical and electronic devices, mobile phone masts, or power lines is constantly increasing [1]. The effect of the artificial electromagnetic field on organisms is the subject of extensive public debate and growing numbers of studies. The influence of the electromagnetic field on the honey bee has also been a topic of various research projects. The honey bee as an element of the environment is constantly exposed to various stressors, including electromagnetic fields of various frequencies and intensities.50 Hz is a widely used power frequency in most countries [2,3]. If a honey bee flies at a height of about 2 m above ground in an open space near a power line it is exposed to an E-field with an intensity of 10–12 kV/m. If high obstacles appear in the honey bee’s way, it flies about five or more meters above the ground so it is exposed to an E-field with an intensity of 5–7 kV/m [3,4,5].Bees have been proved to avoid feeding places exposed to a static electromagnetic field (1.5 kV/m) [6]. The success of the foraging of bees was limited by exposure to a low-frequency electromagnetic field [7]. A 60 Hz electromagnetic field > 150 kV/m caused wing, antennae, and body vibrations [8]. Migdał et al. [9] show that bees exposed to a 50 Hz electric field (E-field) at various intensities changed the activity of the bees. Bindokas et al. [10] show that the exposition of the honey bee to conductive tunnels increased mortality. Moreover, an electromagnetic field has an impact on the honey bee’s physiology by modifying pupal development (mobile phone radiation) [11], increasing oxygen consumption (static field 1.4–2.8 kV/m) [12], or changing biochemical parameters (50 Hz E-field at 5.0 kV/m, 11.5 kV/m, 23.0 kV/m, and 34.5 kV/m for 1, 3,6, and 12 h) [13,14]. Thus, previous studies indicate that an electromagnetic field may be one of the threats to honey bees.Numerous environmental factors, including anthropogenic ones, have an influence on the activity of the honey bee’s immune system. Broadly understood environmental pollution weakens the bee’s immune system and reduces colony health [15,16]. Two types of immunity can be distinguished in honey bees: individual and social. Individual immunity includes anatomical barriers, cellular and humoral immunity, while social includes behavioral immunity [17,18].One of the important individual barriers is proteases, which occur both inside the honey bee organism and on the surface of its body [14]. The proteolytic system in the honey bee (Apis mellifera L.) organism is involved in crucial processes, such as protein digestion, receptor activation, the release of hormones, and activation of the zymogens [19,20,21,22]. Among their various functions, these enzymes play a significant role in the activity of the immune system. They are one of the basic lines of defense against pathogens [23].Behavioral immunity consists of, among others, hygienic instinct, bee fever, and absconding. Bees show grooming, which includes self-grooming, and social-grooming (allo-grooming). This behavioral complex helps to reduce, for example, Varroa destructor infestation [16,24].Our study aimed to show the effect of an electromagnetic field at 50 Hz and variable intensities on honey bee proteolytic systems and behavior parameters after exposure for 12 h.2. Materials and Methods2.1. BeesQueens originating from the same mother-queen colony were inseminated with the semen of drones from the same father-queen colony. Ten mother queens were randomly selected and kept in isolators with empty Dadant combs (435 × 300 mm) for egg-laying. Each queen was kept in a separate bee colony. On the 20th day of bee development, the combs with the already sealed worker bee brood were transferred to an incubator with constant conditions (temperature of 34.4 °C ± 0.5 °C and relative humidity of 70% ± 5%) for emerging without adult bees. The combs were transported at the same time and put together in one incubator. Feed (honey and bee bread) was provided ad libitum.2.2. Experimental DesignOne-day-old workers were randomly placed in 50 wooden cages (20 × 15 × 7 cm). Each cage contained 100 workers and two inner feeders with a 50% sucrose solution. Bees were fed ad libitum. The bees were divided into four experimental groups which were exposed to the following 50 Hz E-field intensities: 5.0 kV/m, 11.5 kV/m, 23.0 kV/m, or 34.5 kV/m for 12 h, and the control group. The control group was not treated by the artificial E-field; in this group, the bees were under the influence of an electromagnetic field < 1.00 kV/m. Each group consisted of ten cages. The group name was the E-field intensity to which the bees were exposed. The control group is marked with the letter C.2.3. E-Field GenerationA homogeneous 50 Hz E-field was generated in the exposure system in the form of a plate capacitor with the distance of 20 cm between two electrodes constructed as a squared cage made out of wire mesh, as per Migdał et al. [9]. In most countries, 50 Hz is a widely used power frequency [25]. The E-field intensity and the homogeneity in the test area were verified by an LWiMP accredited testing laboratory (certification AB-361 of Polish Centre for Accreditation) using an ESM-100-m No. 972,153 with calibration certificate LWiMP/W/070/2017, dated 15/02/2017 and issued by the accredited calibration laboratory PCA AP-078. The measurements were done at points of a 10 × 10 × 5 cm3 mesh inside an empty emitter. The stability of the electric field was maintained by permanently monitoring the voltage applied to the exposure system using a control circuit. The field intensity was fixed at 5.0 kV/m, 11.5 kV/m, 23.0 kV/m, or 34.5 kV/m. Changes in the homogeneity and stability of the E-field intensity were no higher than ±5% in the emitter, to which the bees were exposed during the whole experiment.2.4. Protease AnalysisHemolymph samples were collected from 100 bees randomly taken from each group. The hemolymph was taken after exposure by removing the antennae of a live bee using sterile tweezers as per Migdał et al. [26]. The hemolymph sample was collected in sterile glass capillaries with a volume of 20 μL end-to-end without anticoagulant. The prepared capillaries were placed in 1.5 mL Eppendorf tubes filled with 150 μL of 0.6% NaCl. The test tubes were placed on the cooling block during this procedure. The prepared tubes were transferred to a cryo-box and then frozen at −80 °C [27]. Determinations of the acidic, neutral, and alkaline protease activities were done according to the Anson method [28] modified by Strachecka and Demetraki-Paleolog [16]. The activities of acidic proteases were assayed in a buffer of 100 mM glycine-HCl at pH 2.4, neutral ones in a buffer of 100 mM Tris-HCl at pH 7.0, and alkaline ones in a buffer of 100 mM glycine-NaOH at pH 11.2 using the method described by Łoś and Strachecka [27]. The samples of hemolymph were collected immediately after the end of exposure to the E-field.2.5. Behavior AnalysisTwenty-one bees were randomly taken from each group and were placed in a behavioral assessment station made of glass, with a height of 20 cm and a diameter of 40 cm. Observations were conducted with the use of recorded material (offline). Three bees were recorded at the same time for 360 s (60 s for adaptation to location change and 300 s for analysis) using a SONY HDR-CX240E camera (Lund, Sweden). Recorded videos were transferred to a computer on which Noldus Observer XT 9.0 software was installed. Six basic behaviors were selected for observation, i.e., walking, self-grooming (self-cleaning of the body surface, cleaning of antennae, and cleaning of proboscis), flying (between the walls, the bottom and the lid of the container), stillness (time when the bee remained motionless), contact between individuals (including trophallaxis and allo-grooming), and wing movement (exposed Nasonov’s gland).For behavioral analysis using the Noldus Observer XT 9.0 software, a project with a mutually exclusive type of behavior was used (observation for each bee was done separately and only one bee was observed at any one time). The project did not use behavior modifiers in the form of changing conditions or interfering with the insects, as all individuals were assessed under the same conditions. Independent variables in the form of age, body condition, and damage were excluded, as per Migdał et al. [9].For analysis, we chose the average duration of behavior (how much time, on average, bees from one group spent on the behavior) and the number of individual behavior occurrences (how many times during the observation individuals from the group displayed the behavior). The recording of the bees was immediately after the end of exposure to the E-field.2.6. Data AnalysisThe normality of the data distribution was analyzed using the Shapiro–Wilk test. The statistical significance of data between groups was determined by the Kruskal–Wallis test and Dunn’s post hoc rank sum comparision using the package “pgirmess” for “kruscalmc” function. For all tests, RStudio [29] was used with a significance level of α = 0.05.3. Results3.1. Protease AnalysisIn all experimental groups, the level of protease activity was higher than in the control group (Figure 1). Differences between all experimental groups and the control group were statistically significant except for the 23.0 kV/m group in the case of alkaline proteases (Table 1).An intensity of 5.0 kV/m increased the activity of acidic proteases by 78%, neutral by 74%, and alkaline by 40% compared to the control group. Bees treated with an intensity of 11.5 kV/m were characterized by 63% higher activity of acidic proteases, 61% higher neutral protease activity, and 5% higher alkaline protease activity in comparison to the control bees. An intensity of 23.0 kV/m caused an increase of acidic protease activity by 142%, neutral protease activity by 125%, and alkaline protease activity by 4% compared to the control group. Bees exposed to an E-field with an intensity of 34.5 kV/m had 261% higher acidic protease activity, 74% higher neutral protease activity, and 27% higher alkaline protease activity compared to the control bees.3.2. Acidic ProteasesThe highest activity of acidic proteases was recorded in bees treated with an intensity of 34.5 kV/m, while the lowest was in the control group. Among the experimental groups, the least influence on acidic protease activity was for an intensity of 11.5 kV/m (Figure 1). All differences between the groups were statistically significant (Table 1) (p-value < 2.2 × 10−16).3.3. Neutral ProteasesAn intensity of 23.0 kV/m caused the highest increase of neutral protease activity. Control bees were characterized by the lowest activity of neutral proteases. Among the experimental groups, the least influence on neutral protease activity was for an intensity of 11.5 kV/m (Figure 1). The activity of neutral proteases in bees from the 5.0 kV/m and 34.5 kV m groups did not differ significantly (Table 1) (p-value < 2.2 × 10−16).3.4. Alkaline ProteasesThe highest alkaline protease activity was recorded within bees exposed to an E-field with an intensity of 5.0 kV/m while the lowest was in the control bees. Among the experimental groups, the least influence on alkaline protease activity was for an intensity of 23.0 kV/m (Figure 1). The control group and the 23.0 kV/m group did not differ significantly (Table 1). Changes in the activity of alkaline proteases between the experimental groups and the control group were smaller compared to acidic and neutral protease activity (p-value < 2.2 × 10−16).3.5. Behavior AnalysisBees in the control group and the 5.0 kV/m group displayed all six behaviors (Figure 2 and Figure 3). For the 11.5 kV/m, 23.0 kV/m, and 34.5 kV/m groups, all behaviors were observed except wing movement. The number of stillness and wing movement observations was too small to show statistically significant differences between the groups (Table 2 and Table 3).3.5.1. Number of Behavioral OccurrencesThe most frequently shown behavior in all groups was walking and the least stillness or, in the case of the 5.0 kV/m group, wing movement, which was noticed only once (Figure 2, Table 2). Bees in the control group demonstrated the highest number of all behavior occurrences, except flying, which was most often displayed by the 34.5 kV/m group (p-value = 0.0001). Differences in walking occurrences between the control group and all experimental groups were statistically significant (Table 2) (p-value < 6.96 × 10−7). The 23.0 kV/m group displayed the lowest number of flying, walking, and self-grooming occurrences. Contact between individuals was displayed least frequently by the 34.5 kV/m group and stillness by the 11.5 kV/m group. The number of occurrences of contact between individuals decreased with increasing intensity; however, a statistically significant difference occurred only between the control group and the 34.5 kV/m group (Figure 2, Table 2) (p-value = 0.0032).3.5.2. Time Spent Performing BehaviorsBees in the control group, the 5.0 kV/m, 11.5 kV/m, and 34.5 kV/m groups, on average, spent most of the time walking (Figure 3). The average time spent walking by the control group statistically differed from the other groups (Table 3) (p-value = 3.69 × 10−8). In the 23.0 kV/m group, the bees remained still most of the time. In all groups, the bees spent the least time flying (p-value = 0.0001). The control group displayed the shortest average duration of all behaviors, except wing movement. Walking was displayed the longest by bees in the 23.0 kV/m group, self-grooming by bees in the 34.5 kV/m group, flying by bees in the 5.0 kV/m group, stillness by bees in the 23.0 kV/m group, and contact between individuals in the 5.0 kV/m group.4. Discussion4.1. Protease AnalysisProteases are enzymes that catalyze the hydrolysis of peptide bonds which link amino acid residues [30]. They occur in all types of organisms—eucaryotic, procaryotic, and viruses [31]. Proteases are involved in physiological reactions, e.g., digestive, apoptosis, blood clotting. The material of our research was the hemolymph in which serine proteases occur [32]. This group of proteases plays an important role in regulatory and signaling processes, digestion, transport, and degradation of the damaged protein. Serine proteases are involved in the correct action of insect resistance barriers and antioxidant systems. They are responsible, among others functions, for melanization, wound healing, and phagocytosis stimulation by taking part in phenol pro-oxidase cascade activation [33]. Based on the optimal pH in which these enzymes are active, proteases can be classified as alkaline (basic), neutral, or acid.In our present study, treating bees with an E-field at 50 Hz and an intensity of 5.0, 11.5, 23.0, or 34.5 kV/m for 12 h caused an increase in the level of proteases in comparison to the control group (Figure 1, Table 1). The activity of proteases did not increase with increasing electromagnetic field intensity. In our previous study [13], bees exposed to an E-field with identical parameters for 1, 3, and 6 h had higher acid and neutral protease activity than the control group. Regarding alkaline proteases, their activity was higher only in those bees treated with an E-field with an intensity of 23.0 or 34.5 kV/m. Changes in the protease activity in the honey bee organism after long exposure (12 h) to an E-field still persist. The increased activity of proteases in hemolymph was also noticed after treating bees with bromfenvinphos [34]. The authors assumed that Varroa treatment with bromfenvinphos markedly suppresses the honey bee biochemical defense levels. We hypothesized that increased protease activity after E-field exposure could also cause such effects.Changes in protease activity occurred after treating bees with curcumin, coenzyme Q10, or caffeine. The addition of curcumin and coenzyme Q10 in sugar syrup caused the decreased activity of proteases in hemolymph [35,36]. Caffeine caused an increase of neutral protease activity and a decrease of alkaline and acidic protease activity [37]. Bees treated with Q10, curcumin and caffeine lived longer than control bees [35,36,37]. Based on these studies, it can be assumed that these substances have a potentially positive effect on longevity by decreasing protease activity. Strachecka et al. [15] assumed that the activity of proteases on the body surface differs in polluted and clean environments.Additionally, after treating bees with imidacloprid, a decrease in acidic and alkaline protease activity and an increase in neutral protease activity was noticed regardless of the doses of insecticide (5 or 200 ppb) [38].It is difficult to clearly state what effect the increased activity of proteases has on bees’ immunity. Decreases and increases of protease activity occur in healthy bees and are connected with the age of the insect. The increase of acidic, neutral, and alkaline protease activity can be noticed until the age of 18–20 days and decreases after this time [34].4.2. Behavior AnalysisBehavior plays a significant role in insect immunity. Bees as social insects have evolved mechanisms of individual and social behavioral defense that can minimalize the presence of pathogens, pests, and parasites [17,18]. Factors that threaten bees often affect insect behavior and change their activity. These phenomena can influence disease susceptibility by affecting behaviors related to immune responses, like self-grooming (even when the threatening factor does not alter the immune response of the bee).In our study, treating bees with an E-field at 50 Hz and an intensity of 5.0, 11.5, 23.0, or 34.5 kV/m for 12 h caused a reduction in the number of occurrences self-grooming, contact between individuals, and walking while increasing the average time spent on the behavior in comparison to the control group (Figure 2 and Figure 3). Contact between individuals is a significant behavior in pheromone transmission and any disorder of this behavior results in a modification of the relationships between individuals and thus community functioning. Self-grooming is an important trait that contributes to the defense against pests, pathogens, and parasites. Thus, if an E-field changes the behavioral pattern of the honey bee, it can indirectly affect the honey bee’s immune system.Since bees in the control group changed their behavior more often than the experimental bees, the average time spent on an individual behavior was shorter in the control group (Figure 2 and Figure 3). Bees in the experimental groups were less active (changing their behavior less often). Our previous study shows that treating bees with an E-field with the same parameters for 1, 3, and 6 h caused a similar behavioral change. Bees in the experimental group were cleaning themself and displayed contact between individuals less frequently than the control bees [9]. Our present study shows that bees after longer exposure (12 h) to an E-field still displayed modified behavioral patterns compared to the control bees.Only a few publications have evaluated the impact of an E-field on honey bee behavior; thus, a comparison of the results to other factors was necessary. Changes in honey bee activity were observed in studies on the effect of pesticides on bees, in particular, neurotoxins, which affect neural processes by affecting the conduction of the electrical signals between nerve cells (neurons). We can hypothesize that an E-field can also change the processing and conduction of nerve impulses. Depending on the dose, imidacloprid may induce an increase or a decrease in activity. The lowest dose used in the studies by Lambin et al. [39] (1.25 ng per bee) caused an increase in motor activity, while higher doses (2.5–20 ng per bee) caused a decrease. In sublethal doses, cypermethrin, tetramethrin, and tau-fluvalinate reduced the motor activity in honey bees. In our present study, the E-field caused a reduction in the number of occurrences of most behaviors (Table 2, Figure 2). Control bees were more active—they changed behavior more often (41 times on average during the whole observation time) than bees in the experimental groups (from 23 to 35 times). James and Xu [40] assumed that by influencing the motor activity of bees, neurotoxic insecticides can affect disease resistance even not affecting individual immunity.Imidacloprid provoked problems with coordination, convulsions, excessive agitation, or stillness in honey bees [39,41,42]. Exposure for 12 h to an E-field caused stillness to be displayed by the bees, but the number of occurrences of this behavior was too small to show significant differences (Table 2 and Table 3). Even so, it is worth paying attention to the fact that the control bees were still for 5.72 s, while bees in the experimental groups were motionless for from 14.07 s (34.5 kV/m group) to 47.71 s (23.0 kV/m group) (Table 3). Bee stillness after contact with imidacloprid results from the disturbance of impulse conduction caused by a binding of this insecticide with acetylcholine receptors [39]. Morfin et al. [43] found that chronic sublethal exposure to clothianidin affected the proportion of bees grooming intensively and, based on RNAseq, found an effect on pathways linked to neural function, which could be related to the bees’ ability to perceive external stimuli. It is possible that E-field and neurotoxic insecticides can cause changes in the bee nerve impulse transmission; however, the mode of action is probably different. Bees exposed to esfenvalerate and permethrin spent, respectively, 43% and 67% less time in social interaction compared to control bees [44]. Bees in our present work treated with an E-field with an intensity of 34.5 kV/m or 11.5 kV/m spent on average 47% less time in contact between individuals than control bees, while bees in the 23.0 kV/m group spent on average 8% less time on this behavior. Bees treated with an E-field with an intensity of 5.0 kV/m spent on average 262% more time in contact between individuals. Nevertheless, bees in the control group displayed this behavior more often than the other groups (Figure 2 and Figure 3; Table 2 and Table 3).Chronic exposure to thiamethoxam significantly impairs a bee’s ability to fly: It reduces the flight time (−54%), flight distance (−56%), and its average speed (−7%) [45]. According to our present study, bees in the control group spent the least time flying, although, the differences between the groups, except the 5.0 kV/m group, were not statistically significant (Figure 3, Table 3). Regarding the number of flying occurrences, the 34.5 kV/m group had the highest number of this behavior occurrence (Figure 2). The difference between this group, the control group, and the 23.0 kV/m group was not statistically significant. The bees in the 5.0 kV/m and 11.5 kV/m group flew significantly fewer times than bees in the 34.5 kV/m group (Table 2). In conclusion, E-fields cause changes in bee behavior, most often by reducing the number of occurrences of individual behavior.4.3. Behavior and Protease Analysis ComparisonIn our present study, the level of protease activity increased in the experimental bee organisms (Figure 1). Experimental bees also displayed a reduced number of behavior occurrences in comparison to control bees (Figure 2). In our previous study, bees that were treated with an E-field for 1, 3, or 6 h have higher activity of neutral and acidic proteases than control bees. In the case of alkaline proteases, only bees treated with intensities of 23.0 or 34.5 kV/m have a statistically significant higher activity of this enzyme than control bees [13]. Our previous behavioral research shows that bees in the experimental groups have a reduced number of contacts between individuals and self-grooming occurrences [9].Bees’ immune system response consists both of individual and social immunity, which includes anatomical barriers, cellular and humoral immunity, and behavioral immunity [17,18]. Serine proteases which occur in hemolymph are responsible for melanization, wound healing, and phagocytosis stimulation [33].Based on our present and previous studies [9,13], it can be concluded that bees after E-field exposure are characterized by higher protease activity and reduced contact with other individuals and clean themself less often. Protease activity and behavior parameters analysis can give valuable information about the effect of an E-field on the bees’ immunity. Changes in these parameters may indicate the interaction of behavioral immunity and protease activity, which are designed to protect honey bee’s organisms against environmental stressors (pesticides, pathogens, etc.)5. ConclusionsThe amount of artificial electromagnetic field in the environment is constantly increasing, thus the honey bee is exposed to this factor. In our study, bees in the control group demonstrated the highest number of all behavior occurrences, except flying, and had the lowest activity of all types of proteases. Bees in the experimental groups showed a lower number of walking, self-grooming, and contact between individual occurrences than control bees and had higher protease activity than control bees. Our results show that an E-field is potential harmful factor to the honey bee. However, we do not know if the changes in behavior and protease activity of the honey bee after E-field exposure persist and for how long. It would be important to investigate behavior parameters and biochemical markers at different time intervals after the end of exposure to an E-field. It can be helpful to determine the interaction between the biochemical marker activity and behavioral parameters. Such an observation could provide a better understanding of the immune response of the honey bee exposing to environmental stressors.

animals : an open access journal from mdpi

10.3390/ani12040422

PMC8868386

Pseudomonas aeruginosa is an opportunistic pathogen of dogs and cats able to cause both local and systemic infections. This bacterium is widespread in the environment, resistant to unfavorable conditions, and may spread between humans and other mammals. Its virulence and transmission rely on various virulence factors including those responsible for biofilm formation. Biofilm is defined as a complex biological system that is composed of exopolysaccharides, proteins, extracellular DNA, and biomolecules. Extracellular polymeric substances are the main ingredients of biofilm, accounting for 90% of its total biomass. In this study we analyzed the prevalence of five virulence genes involved in biofilm formation (pelA, pslA, ppyR, fliC and nan1) in 271 P. aeruginosa isolates obtained from dogs and cats. All animals had clinical symptoms of P. aeruginosa infection. In dogs, the strains were isolated from the external auditory canal, respiratory tract, and skin. In cats, the strains were isolated from the nasal cavity, external auditory canal, and skin. Biofilm-forming strains accounted for 90.6% of P. aeruginosa isolates from dogs and 86.4% from cats. The most commonly identified virulence factor gene was ppyR (97.4%). The fliC and pslA genes were detected in 62.4% and 60.1% of the study population, respectively, whereas nan1 and pelA genes were found in 45.0% and 38.7%, respectively. Prevalence of the virulence factor genes was not significantly different between dogs and cats. Given that the ability to form biofilm is related to the antibiotic resistance of P. aeruginosa, our results indicate potential candidates for biomarkers assisting in selection of the most effective treatment for P. aeruginosa infections.

Pseudomonas aeruginosa is an ubiquitous bacterium and opportunistic pathogen that plays an important role in nosocomial infections. The presence of virulence factors and the biofilm-forming ability of this species contributes to a high risk of treatment complications. In this study, we examined the biofilm-forming ability and the prevalence of five virulence factor genes (pslA, pelA, ppyR, fliC, and nan1) in 271 P. aeruginosa isolates (212 from dogs and 59 from cats). Biofilm-forming ability was detected in 90.6% of isolates in dogs and 86.4% of isolates in cats. In P. aeruginosa isolates from both species, the most prevalent virulence factor gene was ppyR (97.2% in dogs and 98.3% in cats), followed by pslA (60.8% and 57.6%), fliC (60.4% and 69.5%), nan1 (45.3% and 44.1%), and pelA (40.1% and 33.9%, respectively). In dogs, a significantly higher proportion of biofilm-forming P. aeruginosa strains possessed the fliC gene compared to non-biofilm-forming strains (p = 0.015). In cats, a significantly lower proportion of biofilm-forming strains had the nan1 gene compared to non-biofilm-forming strains (p = 0.017). In conclusion, the presence of fliC gene and the absence of nan1 gene could be indicators of biofilm-forming ability of P. aeruginosa.

1. IntroductionPseudomonas aeruginosa (P. aeruginosa) is a ubiquitous Gram-negative bacillus. It is also an opportunistic pathogen that occurs on the skin and mucosal membranes of humans and other mammals. In dogs and cats, P. aeruginosa causes skin, systemic and urinary tract infections [1], as well as ulcers, hemorrhagic crusts, erythematous papules [2], otitis externa [3], conjunctivitis [4], rhinosinusitis [5], periapical tooth abscesses [6], and periodontal disease [7]. These infections are consistent with the “One Health” concept, because of the bacterium persistence in the environment, and possible transmission between humans and other mammals.Biofilm is defined as a complex biological system that is composed of exopolysaccharides, proteins, extracellular DNA, and biomolecules [8]. Extracellular polymeric substances (EPS) are the main ingredients of biofilm, forming 50–90% of total biofilm biomass [3]. P. aeruginosa with biofilm-forming ability colonize both biotic and abiotic surfaces. The increasing number of experiments on this subject confirm that chronic infection is caused by bacterial biofilm formation, rather than the planktonic form of bacteria, which causes acute infection [9]. The quorum-sensing system controls the ability of flagellated microorganisms to transform from the planktonic into the biofilm form [8].The various mechanisms involved in biofilm formation are related to the production of virulence factors, which are encoded by established genes. Crucial components of the biofilm structure are the exopolysaccharide containing components Psl and Pel. Pel is required to form biofilm on solid surfaces; on the other hand, Psl regulates adhesion to solid surfaces and initiates biofilm growth by developing new microcolonies [10,11]. The pslA and pelA genes participate in the formation of the carbohydrate-rich structure of the biofilm matrix. Moreover, after inactivation of the ppyR gene, the Psl operon is suppressed and a decline in biofilm formation occurs. The ppyR gene encodes a putative transmembrane protein. The fliC gene is a crucial component of flagella production, due to subunit protein-encoding-flagellin type B [12]. Flagella are highly immunogenic, and are essential for inflammation development [13]. Flagellin P. aeruginosa strains releases epithelial TLR5 signaling NF-kB. Flagellum mediated-swimming is one of several types of motilities in P. aeruginosa strains. The mutants of P. aeruginosa that do not have flagella have poor colonization ability [12].Another virulence factor is neuraminidase, which is an extracellular factor involved in the implantation of P. aeruginosa. Neuraminidase may upregulate a number of potential bacterial receptors and it might release terminal sialic acid residues from sialylated gangliosides [14,15]. Wolska et al. [16] obtained a statistically significant difference in the adhesion of buccal epithelial cells in P. aeruginosa strains that contained one neuraminidase encoding gene (nan1) versus strains in which nan1 was not detected.In this study we aimed to analyze the prevalence of five virulence factor genes (pelA, pslA, ppyR, fliC, and nan1) that could affect biofilm formation in P. aeruginosa strains from dogs and cats.2. Materials and MethodsThis study included 271 isolates collected from dogs (n = 212) and cats (n = 59) in the Lower Silesia, Poland, between 2017 and 2020. Male-to-female ratio was close to 1. The age of animals ranged from 2 months to 18 years (median of 7 years) and was significantly higher in dogs than in cats (p = 0.001). Pedigree individuals accounted for 80% of dogs, whereas the domestic shorthair breed predominated among cats (Table 1). All animals had clinical symptoms of P. aeruginosa infection, which was confirmed in microbiological tests. In dogs, the external auditory canal was the most common collection site, followed by the respiratory system, and skin. In cats, most of P. aeruginosa strains were isolated from the nasal cavity. These strains were significantly more often isolated from the external auditory canal, skin, and appendages of dogs, and from the respiratory tract of cats (Table 1).2.1. IsolatesThe isolates of non-fermentative gram-negative bacilli were transferred to basic media. All isolates were cultured on a Columbia Blood Agar Base Thermo Scientific OXOID with 5% sheep blood and MacConkey Agar (Thermo Scientific OXOID, Gdańsk, Poland). The isolates were then checked for oxidase production (Oxidase Detection Strips MICROGEN MID-61g GRASO BIOTECH, Starogard Gdański, Poland). Isolates were frozen with Brain Heart Infusion Broth (BHI) (Thermo Scientific OXOID, Gdańsk, Poland) supplemented with 30% glycerol.2.2. DNA ExtractionOxidase positive isolates were cultured on a Columbia Blood Agar Base with 5% sheep blood (Thermo Scientific OXOID, Gdańsk, Poland) and were incubated overnight at 37 °C. A few colonies from each strain were added to 200 μL distilled water, shaken, and boiled for 20 min. The samples were then frozen at −20 °C for 5 min. After deactivation, the suspensions were centrifuged at 13,000 rpm for 3 min. The supernatant was preserved as template DNA.2.3. PCR AssayAfter pre-selection, isolates with P. aeruginosa identification were confirmed by polymerase chain reaction (PCR) based on amplification of two outer membrane lipoproteins genes, oprI and oprL. OprL encodes lipoprotein specific for P. aeruginosa, and oprI encodes lipoprotein which occur in both other fluorescent pseudomonads and P. aeruginosa [17].The DNA was amplified in a thermocycler (Bio-Rad, Marnes-la-Coquette, France), using modified methods. The methods and primers are presented in Table 2.Genes were amplified using 2.5 µL buffer (MgCl2 at 20 mM concentration), 0.2 µL Taq DNA Polymerase (5 U/L) (Thermo Fischer Scientific, Vilnius, Lithuania), 0.2 µL dNTP Mix 10 (Thermo Fischer Scientific, Vilnus, Lithuania), 0.2 µL of each specific primer created by Genomed S.A. (Warsaw, Poland) (Table 2), and 2 µL template DNA. The solution was increased to a 25 µL volume with sterile water.Electrophoresis was performed in 2% Agarose gel with Midori Green (NIPPON Genetics EUROPE, Düren, Germany), and Marker 2 (A&A Biotechnology, Gdańsk, Poland) was used as a DNA ladder and detected by UV transillumination (Bio-Rad, Marnes-la-Coquette, France). Genetic material of P. aeruginosa ATCC 27853 strain was used as a positive control, and Mili Q water as a negative control.2.4. Microtiter Plate MethodThe microtiter plate method (MTP) was performed following the method of O’Toole et al. [20]. Pure colonies of strains were cultured on Columbia Agar with 5% sheep blood, after overnight incubation at 37 °C. A few colonies were diluted in a BHI broth, and incubated at 37 ℃ overnight. This culture was made equal with 0.5 McFarland standard, and was then diluted at a ratio of 1:100 in fresh a BHI medium. Then, 200 µL of the solution was added to a 96-well sterile microtiter plate. Each sample of P. aeruginosa isolate was replicated in eight wells. The fresh BHI broth was used as a negative quality control. Plates were incubated for 24 h at 37 °C. Then, the plate was turned over and the broth was shaken out. Next, water was gently added to each well and shaken out. This protocol was repeated twice. Then, 250 µL of a 0.1% solution of Cristal Violet, was added to each well. The microtiter plate was incubated at room temperature for 10 min, and was then washed three times by submerging it in a tube of water, and shaking it. The plate was left upside down for a few hours to dry completely [20]. Biofilm was quantified using the modified MTP method [21]. To each well, 250 µL of 95% ethanol was added to solubilize the crystal violet. Absorbance was measured (Spark 10M) at 590 nm. Sterile BHI was used as a negative control. P. aeruginosa ATCC 27853 with the capacity to form biofilm was used as a positive control. Biofilm-formation ability was considered as positive at a cut-off level 0.269. We determined cut-off arbitrarily by the mean for the negative control (culture medium, 0.149) plus two standard deviations (0.06). Levels of biofilm production were established based on following classification criteria: weak biofilm formers: 0.269 < A590 < 0.538 (2 × negative controls); moderate biofilm formers: 0.538 < A590 < 1.076 (4 × negative controls); strong biofilm formers: 1.076 < A590 (6 × negative controls).2.5. Statistical AnalysisCategorical variables were expressed as counts in groups and percentages from the study population and compared between groups using the maximum likelihood G-test or Fisher’s exact test if an expected count in any cell of the contingency table was below 5. Proportions were compared between ordinal classes (strength of biofilm-forming ability) using the χ2 test for trends [22]. The 95% confidence interval (CI 95%) for proportions (prevalence) was calculated using the Wilson score method [23]. Numerical variables were presented as the median, interquartile range (IQR), and range, and compared between groups using the Mann–Whitney U test. The same test was used to compare ordinal variables (strength of biofilm-forming ability) between groups. Correlations between ordinal and numerical variables were analyzed using the Spearman’s rank correlation coefficient (Rs). All tests were two-tailed. A significance level (α) was set at 0.05. Statistical analysis was performed using TIBCO Statistica 13.3 (TIBCO Software Inc., Palo Alto, CA, USA).3. ResultsBiofilm-forming strains accounted for 90.6% (CI 95%: 85.9%–93.8%; n = 192) of P. aeruginosa isolates from dogs, and 86.4% (CI 95%: 75.5%–93.0%; n = 51) of P. aeruginosa isolates from cats. Proportions of biofilm-forming strains with weak (n = 64; 26.3% of 243 biofilm-forming strains), intermediate (n = 85; 35.0%), and strong (n = 94; 38.7%) biofilm-forming ability were similar, and their distribution was not significantly different between dogs and cats (p = 0.893) (Figure 1).The prevalence of biofilm-forming strains did not significantly differ between males and females (dogs: 90.2% vs. 91.0%, respectively, p = 0.838; cats: 91.2% vs. 80.0%, respectively, p = 0.265) or among collection sites either in dogs (p = 0.092) or in cats (p = 0.999) (Table 3).The strength of biofilm-forming ability was neither different between sexes (p = 0.731 in dogs and p = 0.934 in cats) nor correlated with animals’ age either in dogs (Rs = 0.11; p = 0.148) or in cats (Rs = 0.29; p = 0.059).The most commonly identified virulence factor gene was ppyR (n = 264; 97.4%, CI 95%: 94.8%–98.7%). The fliC (n = 169; 62.4%, CI 95%: 56.5%–67.9%) and pslA (n = 163; 60.1%, CI 95%: 54.2%–65.8%) genes were detected significantly less often, whereas nan1 (n = 122; 45.0%, CI 95%: 39.2%–51.0%) and pelA (n = 105; 38.7%, CI 95%: 33.1%–44.7%) genes were observed in the lowest proportion of isolates. Prevalence of the five virulence factor genes was not significantly different between dogs and cats (Figure 2).The only virulence factor genes whose prevalence significantly differed between biofilm-forming and non-biofilm-forming P. aeruginosa strains were the fliC gene in dogs (significantly more often detected in biofilm-forming P. aeruginosa strains; p = 0.015) and nan1 gene in cats (significantly less often detected in biofilm-forming P. aeruginosa strains; p = 0.017) (Table 4).Only the prevalence of nan1 gene appeared to be significantly linked to the strength of biofilm-forming ability. A significantly lower proportion of strains with strong biofilm-forming ability than strains with weak and moderate biofilm-forming ability had nan1 gene (p = 0.022) (Figure 3).4. DiscussionBiofilm formation is the most important virulence factor of P. aeruginosa and is often responsible for failures of the antibiotic treatment. Antibiotic resistance is higher among biofilm-forming strains compared to strains without this property. Commercial methods for testing antibiotic susceptibility rely on the minimum inhibitory concentration (MIC). However, this approach is only feasible for the planktonic phase of bacteria. Consequently, if clinicians do not know which phase of bacteria (planktonic versus biofilm-forming) is being examined, MIC indication may be misleading. Proper selection of antibiotic therapy in P. aeruginosa infection with the strain bearing virulence factor requires the minimum biofilm eliminating concentration (MBEC) be determined or additional biofilm removing substance be applied [24]. Analysis of cases with biofilm-forming Pseudomonas spp. strains shows that the MBEC should be determined [24,25]. Microbiological methods verifying P. aeruginosa biofilm-forming ability are time-consuming and require specific equipment and trained laboratory staff. Our study aimed at finding an easier and faster way to confirm P. aeruginosa biofilm-forming ability. We analyzed genes which could be related to biofilm forming based on the available literature.Our study showed that only two virulence factor genes (fliC in canine strains, and nan1 in feline strains) could be linked to the biofilm-forming ability. The remaining genes occurred similarly often in both types of P. aeruginosa strains—the ppyR gene in virtually all examined strains, fliC and pslA in slightly more than a half, whereas nan1 and pelA were in less than a half of examined strains.Milivojevic et al. [26] and Sharma et al. [5] previously showed that P. aeruginosa isolates had biofilm forming ability in 93% and 89% of dogs, respectively. This phenomenon was supported by our study, in which 86% and 91% of isolates in cats and dogs were biofilm producers, respectively. In our study, 34.9% and 33.9% of isolates in dogs and cats, respectively, were strong biofilm-producers. However, we are not able to compare the strength of biofilm-forming ability with our research because Milivojevic et al. [26] used PAO1 as a cut-off value of the MTP test, while we established the cut-off value arbitrarily. Results of biofilm-forming ability in humans and animals in the study by the mentioned authors were similar. This conclusion supports the necessity of controlling the occurrence and biofilm-forming ability of P. aeruginosa in companion animals.In contrast to our study, Pye et al. [27] classified only 40% of P. aeruginosa samples isolated from dogs as biofilm formers. This discrepancy with our results might be attributed to the use of lysogeny broth (LB) medium. Wijesinghe et al. [28] showed that, compared to LB medium, BHI performed better at analyzing in vitro biofilm growth. A difference between Pye et al. [27] and our results may be associated with the period of time when the studies were carried out as the genetic content of P. aeruginosa strains collected could be different between the two studies.The presence of pelA, pslA, ppyR, fliC, and nan1 genes has been previously associated with biofilm formation in human medicine [18,19,26]. However, the fact that they are also detectable in isolates from companion animals implies that they are universal. Samad et al. [29] obtained a similar prevalence for the pelA gene (44%) to our study, with 40% of P. aeruginosa samples from dogs harboring it. Higher prevalence might be associated with the origin of collected samples because Samad et al. [29] collected both clinical and environmental samples. In our study, only samples from companion animals were collected.Ertugrul et al. [30] detected fliC gene in 60% of samples, and this result is similar to ours. We detected fliC in the P. aeruginosa samples of 60% and 70% of dogs and cats, respectively. We showed that the detection of two genes in dogs and cats could be used to determine P. aeruginosa biofilm-forming ability. The presence of fliC in dogs and the absence of nan1 in cats was significantly associated with biofilm formation. This latter finding could result from unknown elements in the quorum sensing mechanism for P. aeruginosa infection in cats. It could be related to nan1 gene occurrence and the planktonic form of P. aeruginosa that is responsible for acute infection. Lanotte et al. [14] showed that, although not statistically significant, the prevalence of the nan1 gene rose in cases with worse clinical status. The authors detected nan1 in the P. aeruginosa strain of 57% of patients with cystic fibrosis with good and excellent clinical status, and in 71% of patients with poor or weak clinical status.Our study showed that cats were closer to the cystic fibrosis human model compared to dogs, because 83% of strains originated from the respiratory tract, compared to only 17% for dogs. We compared our results regarding nan1 and fliC genes with human data because no studies on these genes in companion animals have so far been published. To our knowledge, this study is the first to report the detection of nan1 and fliC genes in dogs and cats. However, our results on cats are at odds with the result obtained by Soong et al. [15] in humans, as a positive correlation between biofilm formation and neuraminidase production was observed in this study. This result might be attributed to the contribution of initial colonization in the airway. Thus, different neuraminidase might be produced under genetic control in particular species.5. ConclusionsOur study indicates that analysis of P. aeruginosa biofilm-forming ability may be based on detection of selected virulence factor genes. The development of a simple marker for quick PCR gene detection could help support the treatment of P. aeruginosa infections both in human and animal medicine. It has to be stressed that this is a preliminary study, and further investigations and analyses are essential to introduce these results into medical practice.

animals : an open access journal from mdpi

10.3390/ani11082444

PMC8388652

The knowledge of the parasite fauna in a given territory is the basis for a successful control. Publications in Germany and abroad showed that muskrats are suitable intermediate hosts for Echinococcus multilocularis, a small tapeworm of red foxes and other carnivores with zoonotic potential. After the first detection of this tapeworm in the Brandenburg state of Germany, research started to investigate the distribution of this parasite in final hosts, but the question of identifying infections in intermediate hosts remained. Introduced more than 100 years ago from north America, muskrats were shown to be suitable intermediate hosts for this parasite. In own investigations, 130 muskrats were examined for internal parasites and eleven endoparasites were found. Examination showed that muskrats trapped in the Barnim district of Brandenburg are final hosts for seven intestinal trematodes and are intermediate hosts for four tapeworms of carnivores. The larval stage of E. multilocularis was not detected.

The muskrat is a neozoon species that has occupied many countries of continental North Europe after its introduction from north America as fur animals. Due to its burrowing activity it damages river and canal banks and structures of flood control. For this reason, the eradication of this alien species is recommended. Muskrats are also of parasitological interest since they can act as suitable intermediate hosts for Echinococcus multilocularis. On the other hand, little is known on the other helminths that infect muskrats. A total of 130 muskrats of different age groups trapped in different habitats in the Barnim district of the Brandenburg state by a professional hunter were examined for parasites and seven trematodes (Echinostoma sp., Notocotylus noyeri, Plagiorchis elegans, Plagiorchis arvicolae, Psilosostoma simillimum, P. spiculigerum, Opisthorchis felineus and four larval cestode species (Hydatigera taeniaeformis, Taenia martis, Taenia polyacantha, Taenia crassiceps) were detected. Larval stages of E. multilocularis were not found. O. felineus was found for the first time in muskrats in Germany. All the named parasites were present in Europe prior to the introduction of muskrats. With a prevalence of 48.9%, Strobilocercus fasciolaris, the larval stage of the cat tapeworm, H. taeniaeformis, was the most frequent parasite found in adult muskrats.

1. IntroductionThe muskrat (Ondatra zibethicus) is a medium sized semiaquatic, herbivorous rodent which is native to North America and inhabits wetlands, river banks, irrigation channels, lakes, ponds, coastal areas and estuaries. To be protected from predators, muskrats place the entrance to their burrow below of the water level. In shallow water, muskrats construct lodges using plant material. Three female and two male muskrats had been introduced to Bohemia in Czech Republic as fur animal and for hunting purposes. The current muskrat population in Germany is based on these five animals (according to other sources five pairs of muskrats were released [1]). Due to its reproduction potential and the migration behavior of this animal species, muskrats had occupied territories in southern Germany already in 1928. Already two years later, muskrats were seen in territories that today belong to Thuringia, Saxony, Saxony-Anhalt and Brandenburg [2]. It is documented that muskrats were also released in England but were eradicated later on. Papers on helminth parasites of muskrats in Karelia [3] and in Siberia [4] suggest that muskrats were also introduced to Russia. Muskrats have also occupied territories in northern Mongolia, north-eastern China, North Korea and Honshu Island of Japan [1].Due to its burrowing activity (it damages river and canal banks and structures of flood control) feeding damage in field crops and feeding on the protected freshwater pearl mussel and other mussel species, the muskrat is listed by the Inventory of Alien Invasive Species in Europe [5] as one of the worst invasive species in Europe and is recommended for eradication by the Bern Convention on the Preservation of European Wild Plants and Animals and their Natural Habitats [6].In Germany, foxes, otters, minks, polecats, birds of prey and nowadays raccoons and raccoon dogs are the main predators and this predator-prey-relation reflects on the parasite fauna of muskrats.Investigations on Echinococcus multilocularis in the Brandenburg state of Germany in the 1990s revealed that this zoonotic tapeworm can be found in red foxes [7] but the involvement of small mammals as intermediate hosts in the Brandenburg state was not investigated. Already Abuladze [8] mentioned findings of larval stages of. E. multilocularis in muskrats in the 1950th in Russia and listed 42 other intermediate host. Although muskrats are not the first choice of food for red foxes, they might be an indicator for the occurrence of the fox tapeworm since the lifespan of muskrats is longer than that of voles and mice the preferred food of foxes.The aim of this paper was to examine the helminth fauna of muskrats in the Barnim district of the Brandenburg state of Germany under special attention to cestode metacestodes.2. Materials and MethodsThe Barnim district has the highest share of surface waters (5.8%) of all rural districts of the Brandenburg state. Large lakes (Werbellinsee, Grimnitzsee, Parsteinsee) and a number of smaller lakes and cut off meanders (Alte Oder) are relicts of the last ice age.The northern part of the district has in addition, a number of melioration ditches and canals with slow running water as well as the flood plains of the river Oder. All these are suitable habitats for muskrats.A total of 130 muskrats that were trapped with special body grip and baited vintage traps by a professional muskrat hunter in various habitats of the Barnim district were available for our research. Classification into age groups (adults, subadults, juveniles) was done by the hunter (Table 1).The origin of muskrats was allocated to three major habitats (1. lakes, 2. cut-off meanders of the Alte Oder, including adjacent melioration canals and 3. running waters: river Oder and Finow canal) (Table 2).Most of the carcasses were kept in a deep freezer until examination. In fully defrosted carcasses abdominal and thoracic cavities were opened by a scalpel and scissors and macroscopically examined for the presence of bladderworms. Special attention was paid to the liver as it is the preferred location of E. multilocularis larval stages. Cysts at the surface of the liver were removed and opened. The liver was then sliced into 0.5 cm slices to exclude parasites in deeper parenchyma levels. After removal of internal organs, stomach and caecum were separated, opened by scissors and its content was emptied in 1 L glass cylinders and mixed with 500 mL tap water. Mesenterium was removed from small intestines and the intestinal lumen was three times flushed with 60 mL of normal saline. The rinsing liquid was collected in a 1 L glass cylinder. Contents of stomach, small and large intestines were allowed to settle for 30 min and supernatant was discharged. Water was added and this procedure was repeated until the supernatant became transparent. The sediment was transferred into Petri dishes and checked under a stereoscopic microscope.For the photographic depiction, temporary mounts were produced. For this, cestode larvae or their anterior ends (in case of strobilocerci) were put for 20 min into 40 °C artificial gastric juice on a magnate stirrer and after neutralization with phosphate buffer solution, few drops of bile were added to provoke the evagination of the scolex. The scolex was cut off right posterior to the suckers, was placed between two slides in apical position and was fixed and dehydrated in rising alcohol concentrations. Glycerin was used to clear the preparation. Staining with lacto-carmine was also done but did not reveal additional morphological structures. Trematodes were stained in lacto-carmine, dehydrated in rising alcohol concentrations and cleared in clove oil. Since echinostomatid trematodes tend to lose spines, for the depiction of the collar, an unstained specimen obtained from a freshly trapped muskrat was used.3. ResultsOf the 130 examined carcasses, 73 contained parasites. Thirty-eight muskrats harbored one parasite species. Two and three different parasite species were found in 27 and 7 muskrats, respectively, and four parasite species were found in a single animal only. Examination of 57 carcasses did not reveal parasites. In 28 (=53.8%) juveniles, 14 (45.1%) subadults and 15 (26.3%) adults no endoparasites were detected.The endoparasite fauna of muskrats in the Barnim district of the Federal State of Brandenburg consisted of seven trematodes: Echinostoma sp. (Figure 1 and Figure 2), Notocotylus noyeri (Figure 3), Plagiorchis elegans (Figure 4), Plagiorchis arvicolae (Figure 5), Psilotrema simillimum (Figure 6), P. spiculigerum, Opisthorchis felineus (Figure 7) and four larval cestodes (Hydatigera taeniaeformis (Figure 8 and Figure 9), Taenia martis (Figure 10 and Figure 11), Taenia polyacantha (Figure 12 and Figure 13), Taenia crassiceps (Figure 14) (Table 2). Larval stages of E. multilocularis were not detected. All trematodes except the bile duct parasite O. felineus inhabited the small intestine. Strobilocerci of H. taeniaeformis were located in the liver while other metacestodes were found between intestinal lopes in the abdominal cavity.Of the 73 positive muskrats, 43 were infected with one, 22 with two, 7 with three and one with four helminth species. A total of 40 animals showed Taeniidae metacestodes. Of the 30 cases of liver strobilocercosis, 23 occurred in adult, four in subadult and three in juvenile hosts. In two adult muskrats H. taeniaeformis larvae were found in combination with T. martis larvae. Eight muskrats (five adults, two subadults and one juvenile) harbored T. martis. In addition to the above mentioned mixed infection with H. taeniaeformis strobilocerci, another animal showed a combination of T. martis and T. polyacantha. T. polyacantha and T. crassiceps metacestodes were found in four and one animal, respectively. All these five muscrats were classified as adults.4. DiscussionThe parasite spectrum of muskrats in Europe is relatively poor compared to that of North America. In publications on muskrat parasites from Canada and four US states 35 to 54 different endoparasite species were listed [9]. Examination of helminth fauna of muskrats in the state of Saxony-Anhalt in Germany (n = 80) showed the presence of 15 and 13 different species, respectively [10,11]. In the present material from the Brandenburg State eleven different helminth species were found.Despite the presence of E. multilocularis in final hosts in the Brandenburg state of Germany [7,12,13,14], its larval stage was not detected in muskrats in our material.The role of muskrats as intermediate hosts of E. multilocularis in Germany has been shown in Württemberg (southern Germany) [15] in an area that was known to be endemic for this parasite in foxes [16]. Larval stages of E. multilocularis were also found in muskrats hunted in Baden Wuerttemberg [17], Lower Saxony [18], in North Rhine Westphalia [19,20]. In addition, in France, Belgium and the Netherlands muskrats were found to be infected with E. multilocularis [21,22,23,24,25].In our material, larval stages of E. multilocularis were not detected. Low prevalence of the adult cestode in final hosts and a relative low number of examined hosts might be the reason.In the Brandenburg State E. multilocularis in red foxes is unevenly distributed. Prevalence in the high endemic focus reached 24% while in the surrounding low endemic area prevalence dropped to 5% [7]. Examination of raccoon dogs revealed a similar picture [13].Taeniidae metacestodes were most often diagnosed in adult muskrats and there were also few cases of combination of two Taeniidae species. Thus, it is quite unlikely that muskrats ingest taeniid eggs accidently.In accordance with other sources, Strobilocercus fasciolaris, the larval stage of the cat tapeworm, Hydatigera taeniaeformis, was the most frequent parasite in muskrats (Table 3). Apart from sources mentioned in Table 4, S. fasciolaris in muskrats in Germany was previously found by other authors [1,10,26,27,28]. Strobilocerci terminate with a small liquid filled bladder [29]. In Schleswig-Holstein, nearly all 670 examined muskrats were infected [30]. The pseudo segmented S. fasciolaris is located in up to cherry sized cysts in the liver parenchyma. Relaxed, it can reach a length of up to 459 mm. The scolex is armed with 30–34 rostellar hooks. Larger and smaller hooks measured 384–420 μm and 240–270, respectively.Since muskrats spend most of the time in water, feed mainly on plants growing in water or on bank vegetation (bulrush, iris, sedges, reed grasses, water lilies and others) and only seldom leave the aquatic habitat, there is so far no explanation for the relative high metacestode infection rate. It cannot be excluded that muskrats cover their demand in minerals by ingesting carnivore feces. Other additional sources of calcium and phosphorus would be shells of snails, mussels or crabs. Cats may patrol the banks of water but avoid to enter the water and most probably do not prey on muskrats. Most of the infected muskrats in our material were trapped in lake habitats (Table 2). In Germany, H. taeniaeformis is the most frequent feline cestode. In a survey on parasites of feral cats in the Barnim district, between 6 and 30% harbored this tapeworm [31].Cysticercus talpae, the larval stages of T. mustelae (syn. T. tenuicollis) is often situated in 3–5 mm oval shaped thin-walled cysts under the liver capsule of rodents and insectivores. Less often they can be found in the body cavity, under the skin or in the kidneys. The rostellum of the scolex is equipped with 44–50 or more hooks. Large and small rostellar hooks have nearly the same length, 170–190 and 210 μm, respectively. Apart from muskrats, C. talpae was found in bank voles, yellow necked mice, harvest mice, short tailed and common voles in Germany [32]. Final hosts of T. mustelae are weasels and other members of mustelids. This bladder worm was not found in own material.Other metacestodes listed in Table 4 are situated in the body cavities not causing visible host reactions. Cysticercus longicollis, the larval stage of T. crassiceps, were found in the thorax of an adult muskrat trapped in one of the lakes. This larval stage multiplies in the intermediate host by budding. C. longicollis is a 2–4 mm long egg shaped, thin walled bubble with an invaginated scolex. Rostellar hooks in numbers between 30 and 36 were arranged in two circles and measured 180–197 and 130–151 μm, respectively. Rostellar hooks of both types had strikingly long blades. Budding toke place at the larger end, opposite of the scolex. The infected muskrat harbored 37 fully developed bladder worms, some of them were with buds. Previously, C. longicollis was detected also in a swelling between skeleton muscles [11]. Canids were listed as main and martens, badgers and cats as accidental final hosts for T. crassiceps [33]. In the Brandenburg state, T. cassiceps was found in raccoon dogs and red foxes in a prevalence of 2 and 5%, respectively [14].T. polyacantha is another canid specific cestode with a large spectrum of rodents, including muskrats, as intermediate hosts. The larval stage of T. polyacantha was up to 10–12 mm long, with an invaginated scolex that bared up to 60 rostellar hooks arranged in two circles. The prevalence of T. polyacantha in final hosts, red fox and raccoon dog hunted in the neighboring Uckermark district added up to 40 and 13%, respectively [14].T. martis, also known under its synonyms T. intermedia, T. melesi, T. sibirica, is a cestode of mustelids. Its larval stage with a white, bilaterally flatted, elongated body with frizzed margins was situated in the abdominal cavity without any viewable host reaction. The invaginated scolex was armed with 28 rostellar hooks arranged in two circles. Larger hooks were 175–195 μm and smaller hooks were 130–145 μm long, respectively. Both hooks stroke due to comparably small blades and relatively strong roots. Apart from muskrats, the larval stage of T. martis was found in Germany also in beavers [34]. In Germany, the adult cestode was previously found in stone martens [15], otters [35] and beech martens [36].Allegedly, larvae of T. pisiformis and tetrathyridia of Mesocestoides sp. that were found in the liver parenchyma [10] most probably, were misdiagnosed early stages of S. fasciolaris, since C. pisiformis is usually found in the mesenterium and Mesocestoides tetrathyridia are located free in abdominal and thoracic cavities of intermediate hosts [15]. In our opinion, these stages could also be C. talpae.As far as we know, the recent muskrat population in Germany go back to five to ten specimen that were released in Bohemia [1] and examination showed that the trematode, Quinqueserialis quinqueserialis, seems to be the only species that has traveled with infected muskrats from America to Europe and was able to establish the life cycle in the new habitats. Qu. quinqueserialis is a frequent parasite of muskrats in America and planorbid snails of the genus Gyraulus are the intermediate hosts. Experimental infections showed that at least 15 rodents are susceptible for an infection and produced fertile trematodes [37]. In Germany, this notocotylid species was found in muskrats in Saxony-Anhalt [10,11] and Lower Saxony [38]. In addition, apart from muskrats, Qu. quinqueserialis was also found in a brown rat in Saxony-Anhalt [39]. It has not been detected in our material from Barnim district.All other diagnosed trematodes were already present in Europe prior to the introduction of muskrats and were obtained from other hosts.Two species of the genus Psilotrema, P. simillimum and P. spiculigerum were the most frequent helminth found in muskrats of the Barnim district. Findings of Psilotrema spp. were more frequent in muskrats originating from lakes and cut off meanders of the Alte Oder (Table 2). Both species are primarily intestinal trematodes of water fowl and were originally described from white eyed pochards [40]. Smew, Pallas’s gull and non-fish eating anseriform birds were later listed as final hosts [41,42]. Both Psilotrema species were also found in European water voles and muskrats in the European part of the former Soviet Union [43]. P. simillimum differs morphologically from P. spiculigerum by strikingly larger pharynx and ventral sucker compared to the oral sucker. Species differentiation in deep frozen carcasses were difficult and for this reason both species were combined as Psilotrema spp. in Table 2. Prosobranch snails of the genus Bithynia act as intermediate hosts for both fluke species [41], cercariae encyst on water plants at the surface of the water and are ingested accidentally by the final hosts. It has to be underlined that B. leachi is a very rare snail species in Germany but was frequently seen in the Finow canal of Barnim district, while B. tentaculata is a ubiquitous species [44]. In addition, other sources [10,45,46] reported findings of P. pharyngeatum and P. marki in muskrats. Both names are junior synonyms for P. simillimum and P. spiculigerum, respectively [47].Plagiorchis elegans is a trematode with a wide spectrum of final hosts. P. elegans was found in six lizard [48] and 25 bird species belonging to different orders as final hosts [42]. It also occurs in 18 rodent species including the muskrat [41]. More than ten Plagiorchis species were named as muskrat parasites. However, the validity of many of those species is doubtful because experimental studies showed that this species showed high morphological variability depending on final hosts [49,50,51,52]. Freshwater snails, Lymnea stagnalis and Stagnicola palustris are first and a variety of aquatic insect larvae and crustaceans act as second intermediate hosts [53,54]. Final hosts become infected when the accidently ingest infected larval stages of lake flies and other small water arthropods. P. elegans was more often diagnosed in muskrats from lakes and cut off meanders (Table 2).The other species of the genus Plagiorchis, P. arvicolae, differs strikingly form P. elegans by its ellipsoid shape and massive vitellaria that cover most of the body. Most of the publications on this parasite species originate from former USSR where it was found in 11 rodent species in Russia, Belarus, Ukraine, Georgia, Azerbaijan, Armenia and Kazakhstan [55]. In addition, muskrats were mentioned as final hosts although P. arvicolae is a typical parasite of the northern water vole. P. arvicolae has lymnaeid snails as first and aquatic insect larvae (caddisfly, chironomids, dragonflies, damselflies) as second intermediate hosts. Metacercariae were also detected in their sporocysts [43,55].Notocotylus noyeri is another trematode that was more often found in the northern water vole and is less frequent in other vole species [43]. It is one of the three Notocotylus species in Europe that occurs in small rodents. All other European species of this genus Notocotylus have birds as final hosts [56,57]. The life cycle of N. noyeri was studied in Germany and planorbid snails (Bathyomphalus contortus, Anisus leucostomus and A. vortex were identified as intermediate hosts. Metacercariae encyst on solid items, under natural conditions on water snail shells [58]. N. noyeri can be considered as accidental parasite of muskrats since previously it has not been found in muskrats.Contrary to Psilotrema spp. and P. elegans where most of the cases were detected in muskrats trapped in lakes and cut off meanders, seven of the 12 Echinostoma cases were found in animals that originated from running waters. Until now, the species inventory of the genus Echinostoma is disputed [59]. Echinostoma flukes armed with 37 spines on their head collar (E. revolutum group) contain 56 nominal species which are morphologically similar and difficult to distinguish. Müller [10] named E. armigerum as the species infecting muskrats in Germany. However, E. armigerum and also E. coalitum and E. callawayensis that were described as muskrat parasites in America are synonyms to E. trivolis, a species that occurs only in North America [60]. More recent molecular investigations confirmed that E. trivolis is the only Echinostoma species in muskrats in America [61]. In a previous study on muskrat helminths in the German State of Saxony-Anhalt [11], the detected Echinostoma species was named E. echinatum based on research work conducted in Bulgaria [62]. However, the validity of E. echinatum is disputed [63] and also in recent times not dissolved [59]. Various freshwater snails act as first intermediate hosts for Echinostoma spp. and final hosts become infected when ingesting metacercariae located in mussels, snails and amphibians.Opisthorchis felineus from bile ducts of an adult female muskrat trapped in a cut off meander of the Oder River is a remarkable finding because O. felineus is a parasite of fish eating mammals including humans. It was the first record of this parasite in muskrats in Germany. The muskrat as host for O. felineus is mentioned in the Russian literature [43] as a non-specific parasite for rodents. Only one original source reported the finding of O. felineus in one out of 72 examined muskrats in Siberia [64].So far, O. felineus was detected only in the north-eastern part of Germany and this fact is related to the restricted occurrence of its intermediate host, B. leachi. In the Barnim district, O. felineus was found in 26% of feral cats [30] and in 7.6% of red foxes [65]. The prevalence of O. felineus in foxes in western Brandenburg and Berlin was 8.9% [66] and 17.7% [67], respectively. Apart from cats and red foxes, this zoonotic parasite was found in the Brandenburg state also in dogs [68], raccoon dogs [14] and otters [35]. Final hosts get infected by ingesting raw cyprinid fish where metacercariae are located in fins, gills and flesh. In examination of 802 cyprinids caught in waters of Berlin and Brandenburg, the highest prevalence (>80%) of opisthorchiid metacercariae was found in id, bleak and silver bream [69].No nematodes were detected in our material. The nematode spectrum of muskrats in Eurasia is poor and consists only of 10 species that are specific for terrestrial rodents [70]. With the exception of Trichinella spiralis, infection happens when eggs (Hepaticola hepatica, Trichuris muris, Syphacia obvelata) or larvae (Heligmosomoides laevis, Heligmosomum costellatum, longistriata spp., Trichostrongylus retortaeformis) are ingested. Infective larvae of Strongyloides ratti enter the host through the skin and T. spiralis infections are a result of feeding on carrion.Previous examination of the helminth fauna of muskrats in Germany revealed Trichuris sp. [10,11,38], Trichostrongylus retortaeformis and Heligmosomum polygurum [10] and Ascaris sp. [10,11] in low prevalence.5. ConclusionsMuskrats are herbivorous animals in the first place. While infection with Psilotrema spp. and P. elegans happens when muskrats accidently swallow infected small water insect larvae or crustaceans when feeding on water plants, infections with Echinostoma sp., N. noyeri, P. arvicolae and O. felineus happens when they feed on mollusks, amphibians or on dead fish as emergency food.All the intestinal trematodes are related to water and are primarily parasites of other hosts. It has to be mentioned that except Taeniidae metacestodes, no other helminths (anoplocephalids, ascarids, trichurids and trichostrongylids) with a terrestrial life cycle were found. This indicates that the muskrats in the Barnim district mainly search for food in the water.However, muskrats have to leave the aquatic environment to get infected with metacestodes of carnivore Taeniidae. The larval stage of H. taeniaeformis was the most frequent bladder worm not only in own material but also in other publications. The reason behind might be insufficient mineral content in water and marsh plants that forces muskrats to eat carnivore feces that contain a higher amount of calcium and phosphorous.

animals : an open access journal from mdpi

10.3390/ani11030739

PMC8001425

Triggering of poultry capacity to resist challenge stressors could be vital for animal performance and health. Diet may serve as a tool for modulating animal response to oxidative stress. Within the context of a balanced diet, certain feed additives of plant origin, such as phytogenics, may confer additional cytoprotective effects. As gut health is a prerequisite for animal performance, this work delved into advancing our knowledge on dietary and phytogenic effects on the capacity of the poultry gut to counteract oxidative stress. Study findings showed that a reduction in dietary energy and protein intake by 5% primed important antioxidant responses especially upon phytogenic addition. The new knowledge could assist in devising nutritional management strategies for counteracting oxidative stress.

The reduction in energy and protein dietary levels, whilst preserving the gut health of broilers, is warranted in modern poultry production. Phytogenic feed additives (PFAs) are purported to enhance performance and antioxidant capacity in broilers. However, few studies have assessed PFA effects on a molecular level related to antioxidant response. The aim of this study was to investigate the effects of administering two dietary types differing in energy and protein levels (L: 95% and H: 100% of hybrid optimal recommendations) supplemented with or without PFA (−, +) on gene expressions relevant for antioxidant response along the broiler gut. Interactions of diet type with PFA (i.e., treatments L−, L+, H−, H+) were determined for critical antioxidant and cyto-protective genes (i.e., nuclear factor erythroid 2-like 2 (Nrf2) pathway) and for the total antioxidant capacity (TAC) in the proximal gut. In particular, the overall antioxidant response along the broiler gut was increased upon reduced dietary energy and protein intake (diet type L) and consistently up-regulated by PFA addition. The study results provide a new mechanistic insight of diet and PFA functions with respect to the overall broiler gut antioxidant capacity.

1. IntroductionGut health biomarkers associated with the regulation of the antioxidant response and inflammation currently attract a lot of scientific attention [1,2,3]. In particular, one of the most important regulators of antioxidant response and inflammation is the transcription factor nuclear factor erythroid 2-like 2 (Nrf2) [4]. Transcription factor Nrf2 is a basic leucine zipper-containing transcription factor that is regulated by Kelch-like ECH-associated protein-1 (Keap1) and activates phase II/detoxifying enzymes and more than 100 genes through the antioxidant response element (ARE). These genes include NAD(P)H:quinone oxidoreductase 1 (NQO1), glutathione S-transferase (GST), heme oxigenase-1 (HO-1), glutathione peroxidase (GSH-Px), catalase (CAT), superoxide dismutase 1 (SOD1), glutamate cysteine ligase (GCL), glutathione-disulfide reductase (GSR) and the thioredoxin/peroxiredoxin system [5,6,7,8].In particular, Keap1 binding with various inducers (e.g., phytogenic compounds) leads to transcription of many cyto-protective genes [9,10]. Briefly, CAT and SOD1 are antioxidant enzymes that directly react with radical species, whereas GPX and GSR regenerate oxidized antioxidants [11]. In addition, NQO1 engages a two-electron transfer to diminish quinones to hydroquinones preventing the production of free radical oxygen intermediates [12]. Moreover, GST catalyzes the conjugation of GSH with xenobiotics and protects cells against reactive oxygen metabolites [13]. On the other hand, PRDX1 is proven to be a functional enzyme adjusting cell growth, differentiation and apoptosis [14]. Finally, TXN operates along with PRDX1 as reductase in redox control, preserves proteins from oxidative aggregation and inactivation, supports the cells confront various environmental stresses (e.g., ROS, peroxynitrite, arsenate) and regulates programmed cell death via denitrosylation [15].Research evidence highlights that activation of the Nrf2/ARE signaling pathway could be regarded as beneficial for effectively counteracting oxidative stress in animals and humans. In this respect, contemporary research tries to elucidate the role of dietary components for animal and human health and well-being [16]. In particular, while there is evidence that reduced energy and protein intake could be beneficial for adult humans [17,18], yet the role of energy and protein intake in the activation of the Nfr2 pathway [19] is still limited. On the other hand, accumulating evidence demonstrates that inclusion of various phytogenic feed additives (PFAs) may regulate the Nrf2/ARE pathway in a manner perceived as beneficial for human and animal health [3,6,20,21,22].The aim of this study was to generate new knowledge on the effects of dietary energy and protein levels with or without PFA addition on the modulation of the Nrf2/ARE signaling pathway in the broiler gut mucosa. For the purpose of the study, the expression of critical genes belonging to the Nrf2/ARE pathway was profiled along the chicken broiler gut. In addition to the gene expressions, the antioxidant capacity of the intestinal mucosa was assessed biochemically.2. Materials and Methods2.1. Animals and Experimental TreatmentsFor the purpose of the experiment, 540 one-day-old male Cobb 500 broilers vaccinated at hatch for Marek, Infectious Bronchitis and Newcastle Disease were obtained from a commercial hatchery. Birds were allocated to 4 experimental treatments for 6 weeks. Each treatment had 9 floor replicate cages of 15 broilers each. Each replicate was assigned to a clean floor cage (1 m2), and the birds were raised on rice hulls litter. The temperature program was set at 32 °C at week 1 and gradually reduced to 23 °C by week 6. Heat was provided with a heating lamp per cage. Except for day 1, an 18 h light to 6 h dark lighting program was applied during the experiment to ensure adequate access to feed and water.A 2 × 2 factorial design was used with diet specifications and PFA addition as the main factors. A three-phase feeding program with starter (1 to 10 d), grower (11 to 22 d) and finisher (23 to 42 d) diets was followed. In particular, for each growth phase, two diet types () were formulated to meet 95% and 100% of optimal Cobb 500 metabolizable energy (ME) and protein (CP) specifications, stated as L and H, respectively. The PFA used contained a blend of compounds such as carvacrol, thymol, carvone, methyl salicylate and menthol (Digestarom® Biomin Phytogenics GmbH, Stadtoldendorf, Germany). Diets were in mash form, based on maize and soybean meal and were supplemented with coccidiostat. Throughout the experiment, feed and water were available ad libitum.The calculated chemical composition per kg of the basal diets (L vs. H) was as follows. For the starter diet: AMEn (11.97 vs. 12.60) MJ; crude protein (204.3 vs. 215.0) g; lysine (12.5 vs. 13.2) g; methionine + cysteine (9.4 vs. 9.9) g; threonine (8.2 vs. 8.6) g; calcium 9 g; available phosphorus 4.5 g. For the grower diet: AMEn (12.27 vs. 12.92) MJ; crude protein (185.3 vs. 195) g; lysine (11.3 vs. 11.9) g; methionine + cysteine (8.6 vs. 9.0) g; threonine (7.5 vs. 7.9) g; calcium 8.4 g; available phosphorus 4.2 g. For the finisher diet: AMEn (12.59 vs. 13.26) MJ; crude protein (175.8 vs. 185.0) g; lysine (10.0 vs. 10.5) g; methionine + cysteine (7.8 vs. 8.2) g; threonine (6.8 vs. 7.1) g; calcium 7.6 g; available phosphorus 3.8 g.Depending on diet type (L and H) and PFA supplementation (0 and 150 mg/kg of diet), the four experimental treatments were: L− (95% of optimal ME and CP requirements with no PFA supplementation), L+ (95% of optimal ME and CP requirements with PFA supplementation), H− (100% of optimal ME and CP requirements with no PFA supplementation) and H+ (100% of optimal ME and CP requirements with PFA supplementation).The experimental protocol was in compliance with the current European Union Directive on the protection of animals used for scientific purposes [23,24] and was approved by the relevant national authority (Department of Agriculture and Veterinary Policy, General Directorate of Agriculture, Economy, Veterinary and Fisheries). Birds were euthanized via electrical stunning prior to slaughter.2.2. Broiler Growth Performance ResponsesBroiler performance parameters such as body weight gain (BWG), feed intake (FI), and feed conversion ratio (FCR) were evaluated for the entire duration of the experiment (42 days) (Table 2).2.3. Organ SamplingAt 42 d of age, 9 broilers per treatment were randomly selected and the duodenum, jejunum, ileum and ceca samples were excised carefully and immediately snap frozen in liquid nitrogen and subsequently stored at −80 °C for further analyses.2.4. Molecular Analyses2.4.1. RNA Isolation and Reverse-Transcription PCRΤhe central section of duodenum, jejunum, ileum and the whole ceca were exposed and the luminal digesta was ejected. Then, the segments without digesta were washed completely in 30 mL cold phosphate buffered saline (PBS)–ethylene diamine tetra-acetic acid (EDTA; 10 mmol/L) solution (pH = 7.2), and the mucosal epithelium was taken off with a micro-slide to a sterile Eppendorf type tube. Eventually, the total RNA from the duodenal, jejunal, ileal and caecal mucosa was obtained as reported by the manufacturer’s protocol from Macherey-Nagel GmbH & Co. KG, Duren, Germany, by handling NucleoZOL Reagent. RNA quantity and quality were ascertained by spectrophotometry with the use of NanoDrop-1000 by Thermo Fisher Scientific, Waltham, United Kingdom.DNAse treatment was exercised due to the removal of contaminating genomic DNA from the RNA samples. Ten micrograms of RNA was diluted with 1 U of DNase I (M0303, New England Biolabs Inc, Ipswich, UK) and 10 μL of 10x DNAse buffer to a final volume of 100 µL upon the inclusion of DEPC water, for 15–20 min at 37 °C. Before the DNAse inactivation at 75 °C for 10 min, EDTA should be added to a final concentration of 5 mM to protect RNA from being degraded during enzyme inactivation. RNA integrity was examined by agarose gel electrophoresisFrom each sample, 500 ng of total RNA was reverse transcribed to cDNA by PrimeScript RT Reagent Kit (Perfect Real Time, Takara Bio Inc., Shiga-Ken, Japan) according to the manufacturer’s recommendations. All cDNAs were afterwards stored at −20 °C.2.4.2. Quantitative Real-Time PCRThe following Gallus gallus genes were examined: nuclear factor erythroid 2-like 2 (Nrf2), kelch-like ECH associated protein 1 (Keap1), catalase (CAT), superoxide dismutase 1 (SOD1), xanthine oxidoreductase (XOR), glutathione peroxidase 2, 7 (GPX2, GPX7), heme oxygenase 1 (HMOX1), NAD(P)H quinone dehydrogenase 1 (NQO1), glutathione S-transferase alpha 2 (GSTA2), glutathione-disulfide reductase (GSR), peroxiredoxin-1 (PRDX1), thioredoxin (TXN), glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and actin beta (ACTB). Suitable primers were designed using the GenBank sequences deposited on the National Center for Biotechnology Information and US National Library of Medicine (NCBI) shown in Table 1. Primers were checked using the PRIMER BLAST algorithm for Gallus gallus mRNA databases to ensure that there was a unique amplicon.Real-time PCR was accomplished in 96-well microplates with a SaCycler-96 Real-Time PCR System (Sacace Biotechnologies s.r.l.,Como, Italy) and FastGene IC Green 2x qPCR universal mix (Nippon Genetics, Tokyo, Japan). Every reaction included 12.5 ng RNA equivalents along with 200 nmol/L of forward and reverse primers for each gene. The reactions were incubated at 95 °C for 3 min, accompanied by 40 cycles of 95 °C for 5 s, 59.5 to 62 °C (depending on the target gene) for 20 s, 72 °C for 33 s. This was tailed by a melt curve analysis to check the reaction specificity. Each sample was determined in duplicates. Relative expression ratios of target genes were calculated according to [25] adapted for the multi-reference genes normalization procedure according to [26] using GAPDH and ACTB as reference genes.2.5. Biochemical AnalysesTotal Antioxidant Capacity of Intestinal MucosaTotal antioxidant capacity (TAC) was determined using the oxygen radical absorbance (ORAC) assay [27] to evaluate the hydrophilic antioxidants [28]. Appropriately diluted mucosal samples from duodenum, jejunum, ileum and caecum in phosphate-buffered saline (PBS) were used, and the ability to delay the decay of phycoerythrin fluorescence under the presence of 2,2′-azobis (2-methylpropionamidine) dihydrochloride (APPH) used as oxidant was compared with that of trolox (6-hydroxy-2,5,7,8 tetramethylchroman-2-carboxylic acid) used as an anti-oxidant standard. Data were expressed as concentration of trolox equivalents (TE) (mmol/L of serum).2.6. Statistical AnalysisExperimental data were tested for normality using the Kolmogorov–Smirnov test and found to be normally distributed. Data were analyzed with the general linear model (GLM)–general factorial ANOVA procedure using diet type (L, H) and PFA addition (NO and YES) as fixed factors. Statistically significant effects were further analyzed, and means were compared using Tukey’s honestly significant difference multiple comparison procedure. Statistical significance was determined at p ≤ 0.05. All statistical analyses were performed using the SPSS for Windows Statistical Package Program (SPSS 17.0, Inc., Chicago, IL, USA).3. Results3.1. Growth Performance ResponsesSignificant interactions between diet type and PFA were found for BWG (PD×P = 0.001) and FCR (PD×P = 0.024). In particular, broilers of treatments H- and H+ had higher BWG and lower FCR values compared to treatments L− and L+, while broilers of treatment L+ had better BWG and FCR compared to treatment L−. Moreover, broilers fed diet type H showed higher (PD < 0.001) BWG, FI (PD = 0.046) and lower FCR (PD < 0.001) compared to broilers fed diet type L. In addition, PFA inclusion significantly increased BWG (PP = 0.002) and improved FCR (PP = 0.043) for the whole experiment (Table 2).3.2. Profile of Selected Gene Expression along the Intestine3.2.1. DuodenumIn the duodenal mucosa, significant interaction (PD×P = 0.017) was shown (Figure 1) between diet type and PFA inclusion for HMOX1 gene expression levels, with broilers of treatment (L−) having lower relative gene expression compared to the other treatments.Moreover, as shown in Table 3, diet type significantly affected (P < 0.05) relative gene expression of GPX2 (PD = 0.043) and HMOX1 (PD = 0.017) with broilers fed diet type L showing higher expression compared to broilers on diet type H. In addition, diet type affected the relative gene expression of TXN (PD = 0.017) with broilers fed diet type H showing higher expression levels compared to broilers fed diet type L. In addition, PFA inclusion significantly (P < 0.05) up-regulated relative expression levels of Keap1 (PP = 0.001), CAT (PP = 0.035), SOD1 (Pp = 0.019), HMOX1 (Pp = 0.001), NQO1 (Pp = 0.001), GSR (Pp = 0.041), PRDX1 (Pp = 0.019) and TXN (Pp = 0.035). Gene expression of Nrf2, XOR, GPX7 and GST was not significantly affected (P > 0.05) by PFA inclusion of diet type3.2.2. JejunumIn the jejunal mucosa, diet type significantly affected GPX2 (PD = 0.005) and PRDX1 (PD = 0.002), with broilers fed diet type L showing higher expression levels compared to broilers fed diet type H, as presented in Table 4. On the other hand, relative gene expression of Nrf2, Keap1, CAT, SOD1, XOR, GPX7, HMOX1, NQO1, GST, GSR and TXN was not significantly affected (P > 0.05) neither by diet type nor PFA inclusion3.2.3. IleumIn the ileal mucosa, significant interactions between diet type and PFA inclusion were noted for GPX2 (PD×P = 0.007) and HMOX1 (PD×P = 0.024) as shown in Figure 2 and Figure 3. In particular, the highest relative expression level of GPX2 was found on the L+ treatment, whereas on HMOX1, treatment H+ had the higher gene expression level compared to the other treatments.In addition, diet type significantly affected (PD = 0.006) the expression of CAT, with broilers fed diet type L showing higher expression levels compared to broilers fed diet type H. Moreover, as displayed in Table 5, PFA inclusion, significantly up-regulated (PP < 0.001) the relative gene expression of GPX2. Finally, relative gene expression of Nrf2, Keap1, SOD1, XOR, GPX7, NQO1, GST, GSR, PRDX1 and TXN was not significantly affected (P > 0.05) neither by diet type nor PFA inclusion3.2.4. CecaIn the cecal mucosa, as shown in Table 6, diet type significantly affected (P < 0.05) the gene expression levels of Keap1 (PD = 0.014), GPX2 (PD = 0.003), GPX7 (PD = 0.032) and PRDX1 (PD = 0.006), with broilers fed diet type L showing higher expression levels compared to broilers fed diet type H. Moreover, PFA inclusion significantly up-regulated (PP = 0.041) relative gene expression level of GST. However, relative gene expression of Nrf2, Keap1, CAT, SOD1, XOR, HMOX1, NQO1, GSR and TXN was not significantly affected (P > 0.05) neither by diet type nor PFA inclusion3.3. Total Antioxidant Capacity (TAC) along the IntestineSignificant interactions between diet type and PFA inclusion for TAC were noted in the duodenal (PD×P = 0.024) and ileal (PD×P = 0.007) mucosa as shown in Figure 4 and Figure 5, respectively. In particular, treatment L+ had higher TAC compared to the other treatments. In addition, diet type significantly affected TAC in jejunal mucosa (PD = 0.001) with broilers fed diet type H having higher TAC compared to broilers fed diet type L.Finally, as presented in Table 7, PFA inclusion significantly increased TAC in the jejunal (PP < 0.001), ileal (PP = 0.033) and cecal (PP = 0.032) mucosa.4. DiscussionA deeper understanding of the effects of dietary energy and protein levels on broiler gut function and health is still warranted. In this study, overall BWG and FCR were better in chickens receiving diet type H compared to diet type L (Table 2). These results were in line with previous studies regarding effects on performance and relevant biological responses [29,30,31,32]. However, the topic of energy and protein reduction on metabolic pathways related to cyto-protection via the Nrf2/ARE pathway is still scarce in broilers. On the other hand, there are indications for beneficial effects of dietary energy and protein reductions on the Nrf2/ARE pathway for adult humans [10,11].In this study, overall body weight gain and FCR were improved in chickens receiving PFA, especially in the case of diet type L, whereas chickens in treatment L+ were better compared to L− (Table 2). Similar results have been previously reported [1,29,31]. Phytogenic compounds beyond their benefits for growth performance, nutrient digestibility and meat antioxidant capacity in broilers [1,29,31,33] are currently gaining attention for their functional role on critical elements of gut barrier integrity and inflammation [34,35,36]. Interestingly, emerging evidence reveals that PFA may modulate beneficially the Nrf2/ARE pathway in the broiler gut [3,37].Therefore, the present study aimed to generate new knowledge regarding the effects of dietary energy and protein levels in conjunction or not with PFA supplementation on the stimulation of antioxidant and cyto-protective enzymes, via the activation of Nrf2/ARE pathway. For this reason, this study has employed a powerful analytical palette of gene coding for cytoprotective factors and enzymes in order to profile diet and PFA effects along the broiler gut. The genes and factors studied include a number of phase-2 proteins and antioxidant gene components of the Nrf2 signaling pathway (i.e., Nrf2, Keap1, CAT, SOD1, XOR, GPX2, GPX7, HMOX1, NQO1, GST, GSR, PRDX1 and TXN).In this study, the overall feed intake did not differ significantly between the experimental treatments. Therefore, according to diet specifications, birds on the low specification diet (i.e., treatments L and L+) had indeed lower overall ME and CP intake by approximately 5%, compared to birds on the high dietary specification (i.e., treatments H and H+).Data analysis revealed interactions between diet type and PFA inclusion for HMOX1 and GPX2. In particular, the low diet specs in combination with PFA supplementation up-regulated the GPX2 gene in the ileal intestinal segment. Meanwhile, HMOX1 in duodenum had shown a significant up-regulation in all treatments compared to L−. However, in ileum relative gene expression of HMOX1 was increased only in H+ treatment. Regarding the interactions between diet and PFA, the biological significance of a single gene changes (HMOX1) by its own is questionable and could be considered with the findings about GPX2 gene and TAC results in ileum (Table 5) that indicate PFA role in improving antioxidant capacity. Concerning the segment dependence for gene responses, PFA constituents had been shown to be mainly absorbed in the proximal gut (e.g., stomach and duodenum) [38]. As it was observed in this study, the examined genes that were up-regulated in duodenum upon PFA addition were above 60% (8/13 genes).Moreover, reduced energy intake has been shown to intensify the repairment of DNA systems, advocate the elimination of damaged proteins and oxidized lipids and increase antioxidative defense mechanisms in humans [16], rats and monkeys [39,40]. The results of the present study support the modulatory effects of lower ME and CP intake towards an improved broiler anti-oxidative status. In particular, the low diet specs up-regulated the expression of genes relevant for cyto-protection (i.e., Keap1, GPX2, GPX7 and PRDX1) mainly at the cecal level. In addition, benefits of reduced ME and CP diet specs for other gut health biomarkers (i.e., TLR, tight junctions) have also been shown previously in broiler ceca [36].In this study, PFA supplementation resulted in Keap1 up-regulation in the duodenum. This could be considered relevant for overall gut inflammation management since Keap1, besides its active participation in the Nrf2 pathway [41], also inhibits NF-κΒ via binding to its activator protein Ikkb [42]. Up-regulation of Keap1 and down-regulation of NF-κΒ have recently been shown in the case of a PFA dose response study in broilers [3].The PFA inclusion in this study up-regulated CAT and SOD1 expressions in the duodenal mucosa. Increased CAT and SOD1 activity have been shown in broiler blood following PFA addition [43], whereas [44] observed a significant up-regulation in these two enzymes with oregano essential oil (carvacrol) addition in porcine small intestinal epithelial cells. Furthermore, a significant up-regulation of SOD1 expression in duodenal, jejunal and cecal mucosa has been shown in a PFA dose response study in broilers [3].Irrespective of diet type, the administration of PFA up-regulated NQO1 in the duodenal mucosa. In another study, an up-regulated NQO1 expression in the duodenum was also shown following PFA supplementation [3]. NQO1, as mentioned earlier, catalyzes the two-electron mediated reduction of quinones to hydroquinones, which is commonly proposed as a mechanism of detoxification [12].Furthermore, in this study, PFA inclusion resulted in increased GSR, PRDX1 and TXN expressions in duodenal mucosa. Although there are no other relevant studies to compare directly, PRDX1 up-regulation has also been reported in the case of other gut function modulating additives such as mannan-oligosaccharides in young broilers chickens [45].From all the above, it appears that there are differences regarding the intensity and intestinal site specificity of PFA modulation of the Nrf2 pathway components, between various studies [3,36,37]. It is possible that phytogenic composition and inclusion level, as well as the absorption and metabolism kinetics of phytogenic active components within the birds, could possibly account for the differences between studies. However, the required knowledge on these topics is still rather limited.The effects of PFA on the broiler intestine at a molecular level have been recently shown to correlate with increased intestinal mucosa TAC [3]. Similarly, in this study, PFA inclusion resulted in increased TAC in the jejunum, ileum and ceca. Overall, both studies above provide evidence for PFA cytoprotective and anti-oxidative potential at the broiler intestine.5. ConclusionsIn conclusion, this study has confirmed our previous findings on performance [1,31] and provided new knowledge for the effects of diet ME and CP specs on host antioxidant response. We found that a more intense priming of host antioxidant response was seen in birds fed the diets with ME and CP reduced by 5% of the recommended optimal dietary specifications for the broiler genetic line used. Moreover, beyond the known PFA benefits for performance [1,31], the study results have highlighted the PFA potential for host antioxidant protection, detoxification and inflammation management at intestinal level. Interestingly, it was shown that when PFA was used in conjunction with the low specifications diet, the cyto-protection potential at the intestine was maximized. From a human perspective, and given the higher than 60% homology between the Gallus gallus and the human genome [46], study findings could also be relevant for human gut health as contemporary dietary recommendations for reduced food intake and use of plant bioactive compounds increase in popularity.

animals : an open access journal from mdpi

10.3390/ani11082443

PMC8388624

Workability traits are a group of functional traits that affect the economics of dairy production and are increasingly included in selection indexes. The most important of them include milking speed and temperament. The aim of this study was to estimate genetic and phenotypic parameters of workability traits. The estimation was carried out by considering two approaches: the first using pedigree data and the second using pedigree and genomic data. The obtained results indicate that workability traits belong to low heritable traits and are positively correlated genetically and phenotypically, which means a possibility of their effective improvement in the population, taking into account that the genomic information of sires did not have a significant effect on the estimated genetic parameters.

Heritabilities of workability (WT) traits—milking speed (MS) and temperament (MT)—as well as genetic and phenotypic correlations between these traits in the population of Polish Holstein-Friesian (PHF) cows were estimated. The estimation of genetic parameters was performed twice: first with the use of pedigree data; and second with the use of pedigree and genomic data. Phenotypic data from routinely conducted MS and MT evaluations for 1,045,511 cows born from 2004 to 2013 were available; the cows were evaluated from 2011 to 2015. The main dataset was reduced based on imposed restrictions (e.g., on age of calving, stage of lactation and day of first trial milking). The dataset prepared in this manner comprised 391,615 cows. It was then reduced to daughters of 10% randomly selected sires for computational reasons. Finally, for genetic parameter estimation, 13,280 records of cows were used. The linear observation model included additive random effects of animal, fixed effects of herd-year-season of calving subclass (HYS) and lactation phase, fixed regressions on cow age at calving and the percent of HF breed genes in the cow genotype. Heritabilities estimated based on pedigree data were 0.12 (±0.0067) for MS and 0.08 (±0.0063) for MT, the genetic correlation between MS and MT was estimated at 0.05 (±0.0002) and the phenotypic correlation coefficient was estimated at 0.14 (±0.0004). The inclusion of genomic information of sire bulls had no clear effect on the size of the estimated WT genetic parameters. The heritabilities of MS and MT were 0.11 (±0.0065) and 0.09 (±0.0012), respectively. The genetic and phenotypic correlation coefficients were 0.07 (±0.0003) and 0.12 (±0.0005), respectively. The sizes of the obtained heritabilities of WT and of the genetic and phenotypic correlation between these traits indicate the possibility of effective population improvement for both WT traits.

1. IntroductionIn a review of selection indexes in countries with developed cattle breeding, Miglior et al. [1] found that the importance of functional traits in dairy cattle breeding has increased significantly since the early 2000s. The group of functional traits, which includes workability (WT) traits, are traits that are not directly related to milk yields but affect the profitability of milk production by reducing its costs. In recent years, the importance of WT, such as milking speed (MS) and temperament (MT), has been increasing in cattle breeding programmes. MS can be defined as the cow’s ability to milk in a short time, while MT is the cow’s behaviour and ease of handling during milking [2,3]. The main aim of estimating the breeding value of WT is to identify sire bulls from which daughters are born with undesirable phenotype in terms of MS or MT [4]. Several studies have shown that WTs were a determining factor on culling cows from the herd [5,6,7,8]. WTs are considered as a very important group of traits, especially in herds with Automatic Milking Systems (AMS) [9]. Rupp and Boichard [10] found that very fast milking cows tend to produce elevated somatic cell counts in milk, indicating health problems such as the occurrence of subclinical or clinical mastitis. Cows should be selected for a moderate milking speed.Heritabilities for MT published in the literature typically range from 0.05 to 0.12 [11,12,13]. Heritabilities for MS range from 0.05 [14] to 0.22 [15]. In 2011, the team of Sewalem et al. [13] estimated the genetic parameters of MS and MT in a population of Canadian Holstein-Friesian cows. The genetic correlation coefficient was 0.25 and the phenotypic correlation coefficient was 0.10 [13]. A positive genetic correlation between MS and MT proves that cows characterized by an average or calm MT during milking gave up milk faster and within a shorter time, and this time period is longer in the case of more nervous cows.In recent years, more and more countries in the world with highly developed HF cattle breeding have included WT into their national breeding value (WH) evaluation systems and selection indexes. However, to be able to estimate WH for this group of traits, it is necessary to know the genetic parameters. Currently, no genetic parameters of WT have been estimated in Poland. However, since 2006, data concerning phenotypic evaluation of this trait group have been collected routinely as part of the performance control system conducted by breeders’ associations. Based on the collected data, it became possible to undertake studies on the estimation of genetic parameters of WT in this work. Moreover, by using an additional source of information contained in the genome, it became possible to implement a new methodology of WH estimation based on genomic prediction. Currently, there are no papers in the scientific literature on the possibility of a more in-depth estimation of genetic parameters of WT, i.e., taking into account genomic information. Therefore, the aim of this study was to estimate genetic parameters of MS and MT and phenotypic and genetic correlations between these traits by using both pedigree and genomic data.2. Materials and Methods2.1. DataThe study material consisted of MS and MT phenotypes from 1,045,511 Polish Holstein-Friesian cows born between 2004 and 2013. The phenotypic evaluation of WT cows was carried out between 2011 and 2015. The dataset was created from the SYMLEK system database belonging to the Polish Federation of Cattle Breeders and Dairy Farmers (Warsaw, Poland). WTs are measured in Poland by trained classifiers by using the subjective scoring method that is developed in accordance with the recommendations of the International Committee for Animal Recording (ICAR, Stockholm, Sweden) and implemented since 2006 [16]. WTs are evaluated on a scale point, MS on a scale of 1–5 points from very slow to very fast milking cows and MT on a scale of 1–3 points from slow to excitable or even aggressive (Table 1). The WT evaluation is performed once by trained classifiers for assessment on the second test milking day in the first lactation.The basic dataset included cow numbers, dates of birth and first calving, percent of HF breed genes and phenotypic scores for WT and parental numbers. Next, the sizes of the half-sister groups and of the herd-year-season of calving (HYS) subclasses were calculated, and their distributions were made. The half-sister group was taken as the daughter cows of the same sire and the HYS subclass was defined as the group of cows calved in the same farm, year and season, with the summer calving season covering April to September and the winter calving season covering October to March. A total of 85,172 HYS subclasses and 16,720 half-sibling groups were created. Approximately 40% of the study population consisted of cows with at least 10 half-sisters.The distribution of half-sibling groups and HYS subclasses was presented in the Table 2 and Table 3.Next, the date of phenotypic evaluation of workability traits was added to the basic dataset. Age at first calving was calculated from birth and calving dates. Calving dates and dates of phenotypic evaluations were used to calculate the day of lactation when the evaluation of WT was performed. Based on the day of evaluation, the cows were assigned to eight 10 day lactation stages covering the period from day 21 to day 100 of lactation. The next step was to reduce the size of the basic dataset by introducing the following restrictions: Cows with phenotypic evaluation of WT traits made later than the 100th day of lactation were removed, and WT evaluations made during trial milking performed earlier than 26 days after the first trial milking were discarded. Cows from farms where no variation in MS or MT was observed were also removed. The dataset prepared in this manner (Dataset A) counted 391,615 cows.Based on the data contained in the genotype database of the National Research Institute of Animal Production, a dataset was created containing the genotypes of 2228 bulls whose daughters were evaluated in Poland for WT (Dataset G1). Each bull was genotyped using the Illumina BovineSNPv2 BeadChip, 54609 SNP, and genotype data were stored in the cSNP database system. Next, the imputation findhap software was used to generate genotypes dataset [17]. As a final dataset, we used 43772 SNPs with MAF > 0.001.2.2. Estimation of Genetic Parameters between MS and MTA multitrait linear model of observation including the following fixed and random effects was used to estimate the genetic and phenotypic (co)variance components of WT:y = Xβ + Za + e(1) where y is a vector of observed traits MS/TM; X is the incidence matrix of fixed effects; β is a vector of fixed effects (HYS, lactation stage, regressions on percent of HF genes and regression on calving age); Z is the incidence matrix of random effects; a is a vector of additive genetic random effects; and e is a vector of random error effects.In advance, for the estimation of the genetic and phenotypic (co)variance components of WT traits, additional restrictions were imposed on dataset A. Only the daughters of cows of 10% randomly selected sires were left in the dataset, and HYS subclasses consisting of less than 10 cows were removed. This dataset (Dataset B) of 13,280 cows was treated as the basic dataset for estimating (co)variance components of genetic and phenotypic WT traits. Then, a smaller dataset (Dataset G2) was created, containing the genotypes of 319 sire bulls from Dataset G1, which were sires from dataset B. The (co)variance components were estimated by using programs implemented in the BLUPF90 package [18] that are freely available for research purposes. The programs used were Renumf90 and Gibbs2f90 with the Bayesian Method implemented [19]. The convergence diagnostic and analysis of the posteriori distribution was conducted using the postgibbsf90 procedure included in Blupf90 package, developed by Prof. Misztal’s team at the University of Georgia (Georgia, GA, USA) During the estimation of the (co)variance components, 100,000 parameter samples were generated using a Bayesian approach via Gibbs sampling algorithm. The first 20,000 samples were removed as so-called burn-in samples, and then every fifth sample was recorded. From the subsequent samples, the heritabilities of WT traits and the genetic and phenotypic correlation coefficients between them were calculated. All the steps involved in estimating genetic parameters were repeated, taking into account in the calculations conventional data and standard pedigree information enriched with information contained in the collection with genotypes of sires (Dataset G2). The above estimations were carried out based on methodology of single-step genomic BLUP (ssGBLUP), which is one of the most frequently used methods for incorporating genomic information into genetic evaluations in livestock [20,21]. The ssGBLUP permits the use of all animals in the database by combining pedigree, genomic information from genotyped animals (usually a small portion of animals in the pedigree) and phenotypes to estimate genomic breeding values (GEBV).Genetic and phenotypic parameters were calculated using variance components estimated with 2 different relationship matrices: pedigree-based A (BLUP) and realized H (ssGBLUP). The inverse of matrix H that combines pedigree and genomic relationships was constructed according to Aguilar et al. [21]. Cow and bull pedigrees contained two generations of ancestors, i.e., parents and grandparents, in both approaches.3. Results3.1. Phenotypic Characteristics of Workability TraitsThe first stage of the study was to perform a detailed characteristic of the data, including the calculation of means and standard deviations and the percentage distribution of each grade for both WT by year of birth. The results of this stage of the work are presented in Table 4.The average scores of both WT were close to the optimum in intermediate values: these values were 3.05 points for MS and 1.97 points for MT. In the following years of the birth of cows, two trends can be observed. Between 2004 and 2008, the average scores in terms of both WT were decreasing, after which the trend changed to an increasing one. Until the last analysed year of birth (2013), the phenotypic scores were increasing quite regularly, finally slightly exceeding the optimal value for each trait. The value of the coefficient of variation CV (about 20%) indicates that the right arm of the distribution of phenotypic scores in terms of WT descends more smoothly than the left one.3.2. Characteristics of Polish Sire PopulationThe characteristics of the Polish population of PHF sires, depending on the country of origin and the method of import to Poland, are presented in Table 5. The total number of bulls included in the study was 29,578, of which 10,260 were males of domestic origin and the rest came from abroad. Among the bulls imported to Poland, the highest number came from Germany (8430) and the Netherlands (4530), and the lowest number came from France (1507) and Canada (458).3.3. Genetic Parameters of WT Estimated from Pedigree Data with and without Genomic DataThe genetic parameters of workability traits estimated based on pedigree data and pedigree and genomic data of cows are presented in Table 6.First stage was to estimate parameters based on pedigree data. In the next stage of the study, the data used for estimating the values of WT genetic parameters were enriched with information derived from the genotypes of sires. The genetic parameters of WT were estimated using information enriched from the genotypes of sires. The genotypes of 319 bulls collected in the database of the National Research Institute (Dataset G2) were included in the calculations.Heritability estimated with genomic information was 0.11 for MS and 0.09 for MT. Next, the estimated genetic and phenotypic correlation coefficients between these traits were 0.07 and 0.12, respectively. The values of these parameters are slightly different compared to the parameters estimated without using genomic information on sire bulls. The differences between heritabilities are 0.01 and between correlations are 0.02.The descriptive statistics (mean and Monte Carlo error) and highest posterior density (HPD) regions of the a posteriori distribution for all co(variance) estimates are presented in Table 7.4. DiscussionKnowledge of genetic and phenotypic parameter, i.e., heritability and coefficients of genetic and phenotypic correlations between traits enables the optimization of selection and increases the accuracy of the prediction of its results. Typically, genetic and phenotypic correlations are used to predict changes in the value of one trait during improvement of another trait or, when the values of one trait are difficult to measure, it can be inferred from its relationship with another trait that is easier to determine.The accurate estimation of genetic parameters of functional traits such as WT, for example, has become possible due to the increase in computational capabilities of computers, mathematical method development and increasingly sophisticated genetic parameter estimation software. The (co)variance components of functional traits were estimated by using multivariate linear observation models [22]. The authors of this paper did the same by performing genetic parameter estimation using the latest version of software developed by Prof. Misztal’s team [18]. In this work, the (co)variance components were estimated using the Bayesian method via the Gibbs sampling algorithm [23,24]. The Gibbs sampling is a method for calculating a complex posterior distribution as a steady state measure of a Markov chain. One of the problems of inference from Markov chain generation is that there will always be areas of the target distribution that have not been covered by the finite chain. Therefore, when using this method, it is important to verify the convergence and to assess the posterior distributions [25].In our study, the convergence diagnosis (results not shown) was analyzed by using the Geweke method [26], using the algorithm implemented on the above-mentioned software. As a result, it was found that convergence was achieved for all parameter estimates because the obtained values of the convergence diagnostic test were less than one [18]. Presented in our work, the highest posterior density (HPD) region provides the interval that includes 95% of samples and is a measure of reliability. Moreover, the HPD can be applied to nonsymmetric distributions [27]. For all estimates of variance components, the HDP intervals did not include zero. Nevertheless, most of phenotypic covariance estimates showed different results, and the lower limit of intervals was less than zero. Generally, the slightly larger HDP regions for the (co)variance components were found for the estimates obtained from the genomic approach (pedigree and genomic data) than from the conventional approach (pedigree data), which suggests a slightly higher accuracy of parameter estimates when using conventional dataIn this study, a two-trait animal model based on a linear model was used for the observations including both WT. The coefficient of heritability estimated in this work for MS was 0.12 (±0.0067). This result is similar to the results obtained in the study of Sewalem et al. [13], who obtained a heritability coefficient for MS equal to 0.14. A similar value for this parameter, 0.15, was obtained by Boettcher et al. [28] and Zwald et al. [29] obtained a slightly lower value of 0.11. In other works, higher heritability estimates for MS were obtained. Meyer and Burnside [14] obtained a heritability of 0.21 for the Canadian Holstein-Friesian population, while Potočnik et al. [30] estimated heritability at 0.25 by using a univariate model. The heritability of MS for Brown Swiss cattle estimated by Wiggans et al. [31] was 0.22. In a similar study to the one conducted in this paper, an Italian team of researchers performed the estimation of genetic parameters and breeding values in a population of Simmental cows [32]. The estimated heritability for milkability (very similar trait to MS) was 0.12 (± 0.01). The result of this analysis showed that genomic information could improve the accuracy of breeding valuesSimilar values of heritability coefficients obtained in our study may be a result of the population structure of bulls in Poland: As much as two-thirds of the population consisted of bulls imported from different countries, with the vast majority from world leading countries of cattle breeding. Such a population structure shows that the genetic variability in Poland is similar to that in the whole population of HF cattle.Rensing and Ruten [33] estimated the genetic parameters of workability traits for Holstein-Friesian cattle by using an objective method of collecting information from milking equipment capable of measuring cow milking time. The MS heritability coefficient estimated in this manner was even higher at 0.28. The heritability of this trait, estimated based on different multivariate models applied to the first three lactations using data measured by MS recording devices in the German Simmental breed population, ranged from 0.21 to 0.40 [34]. On the other hand, a similar heritability to that published in the previously cited works, i.e., estimated using phenotypic data from electronic devices in a Hungarian Holstein-Friesian breed population, was estimated by Amin [35]; the heritability coefficient obtained by the author was 0.20. An extension of the data collected from AMS is the study by Kliś et al. [36]. The authors found that longer milking duration along with simultaneous feed consumption in the milking robot had a beneficial effect on MT. The cows were calmer and milked more easily, which also influenced the increase in milk yield.Sewalem et al. [13] concluded from their study that the coefficient estimates may vary due to the phenotypic assessment method and analytical methods used. In addition, these authors showed that the range of heritability estimates obtained in different populations may be smaller provided that more objective methods of phenotypic evaluation of MS are used. The definition of MS in the national cattle performance monitoring program is based on the subjective evaluations of classifiers, and as a result the heritability coefficients for this trait obtained in this study are probably smaller than the results of heritability estimates published by Rensing and Ruten [30] and Dodenhoff and Emmerling [31].The heritability 0.08 estimated from this work for MT was lower than that of MS, and this result is consistent with those published in the literature. The same exact heritability for MT was presented by Sewalem et al. [12]. A slightly lower heritability coefficient 0.04 was presented by Kramer et al. [37]. By contrast, a slightly higher heritability coefficient of MT estimated by Sewalem et al. [13] was equal to 0.13 when using the single-coefficient model and 0.20 when using the two-trait model. A similar heritability MT 0.22 for an Australian population of HF cows was reported by Visscher and Goddard [15].The inclusion of bull genotype data had little effect on the value of heritability estimates for MS and MT; however, slightly higher estimates of genetic and phenotypic correlation coefficients between MS and MT were obtained. It is worth noting that the genetic correlations between these traits are small but positive, meaning that if the daughters of sires are milking faster, they are also generally slightly more excitable. The positive genetic correlation 0.247 (±0.075) and phenotypic correlation 0.10 between MT and MS were estimated by Sewalem et al. [13]. Wethal et al. [38] estimated genetic correlations between milking speed, temperament and leakage within milking system. The correlations were slightly higher in AMS for all combinations of traits and estimates were larger than standard errors. The genetic correlations showed absolute values ranging from 0.15 to 0.88. The genetic correlation estimated in the system milking parlour, such as in our work, was 0.16 (±0.03).The small changes in genetic parameters after WT after the enrichment of pedigree data with genomic information can probably be explained by their structure. In this study, data including phenotypic values of cows and genotypes of their sires were used. The results of the analysis of cow genotypes were not available; consequently, there was a large disproportion between the amount of information coming from both sources. This data structure is not conducive for a high accuracy in genetic parameter estimates. In their study, Dehnav et al. [39] found that a negligible amount of information on genotyped females can improve data structure and genetic parameter estimates. Similar conclusions were presented by Cesarini et al. [32]. Furthermore, only cows with their own phenotypes and sires with a relatively large number of daughters with phenotypes were used in the calculations. These conditions were probably the main reasons for the very similar estimates of the WT genetic parameters that were obtained with the use of pedigree data and of pedigree and genomics.In practice, low coefficients of heritability of a trait indicate the possibility of its effective genetic improvement in a selected animal population (direct selection), but can simultaneously mean a slower genetic gain than in the case of highly heritable traits. One solution is the indirect selection for other moderate and highly heritable traits that are strongly and positively genetically correlated with low heritable traits. Therefore, the continuation of these studies should be the estimation of genetic and phenotypic correlation coefficients between WT traits and other genetically improved production and functional traits.5. ConclusionsIn conclusion, the obtained coefficients of heritability and genetic and phenotypic correlations between WT indicate a possibility of effective improvement of the population in terms of this group of traits.The analysis of the values of coefficients of genetic parameters of WT estimated on the basis of pedigree and genomic data showed that the enrichment with information derived from the genotype of bulls did not bring significant changes in the values of the heritability and correlation coefficients. However, further studies with larger data sets included genotypes of cows are recommended.

animals : an open access journal from mdpi

10.3390/ani11030886

PMC8003775

The use of date palm pollen ethanolic extract (DPPE) is a conventional approach in improving the side-effects induced by Doxorubicin (DOX).DPPE mitigated DOX-induced body and heart weight changes and ameliorated DOX-induced elevated cardiac injury markers. In addition, serum cardiac troponin I concentrations (cTnI), troponin T (cTnT), and N-terminal NBP and cytosolic (Ca+2) were amplified by alleviating the inflammatory and oxidative injury markers and decreasing histopathological lesions severity. DPPE decreased DOX-induced heart injuries by mitigating inflammation, fibrosis, and apoptosis through its antioxidant effect. To reduce DOX-induced oxidative stress injuries and other detrimental effects, a combined treatment of DPPE is advocated.

Doxorubicin (DOX) has a potent antineoplastic efficacy and is considered a cornerstone of chemotherapy. However, it causes several dose-dependent cardiotoxic results, which has substantially restricted its clinical application. This study was intended to explore the potential ameliorative effect of date palm pollen ethanolic extract (DPPE) against DOX-induced cardiotoxicity and the mechanisms underlying it. Forty male Wistar albino rats were equally allocated into Control (CTR), DPPE (500 mg/kg bw for 4 weeks), DOX (2.5 mg/kg bw, intraperitoneally six times over 2 weeks), and DPPE + DOX-treated groups. Pre-coadministration of DPPE with DOX partially ameliorated DOX-induced cardiotoxicity as DPPE improved DOX-induced body and heart weight changes and mitigated the elevated cardiac injury markers activities of serum aminotransferases, lactate dehydrogenase, creatine kinase, and creatine kinase-cardiac type isoenzyme. Additionally, the concentration of serum cardiac troponin I (cTnI), troponin T (cTnT), N-terminal pro-brain natriuretic peptide (NT-pro BNP), and cytosolic calcium (Ca+2) were amplified. DPPE also alleviated nitrosative status (nitric oxide) in DOX-treated animals, lipid peroxidation and antioxidant molecules as glutathione content, and glutathione peroxidase, catalase, and superoxide dismutase activities and inflammatory markers levels; NF-κB p65, TNF-α, IL-1β, and IL-6. As well, it ameliorated the severity of histopathological lesions, histomorphometric alteration and improved the immune-staining of the pro-fibrotic (TGF-β1), pro-apoptotic (caspase-3 and Bax), and anti-apoptotic (Bcl-2) proteins in cardiac tissues. Collectively, pre-coadministration of DPPE partially mitigated DOX-induced cardiac injuries via its antioxidant, anti-inflammatory, anti-fibrotic, and anti-apoptotic potential.

1. IntroductionDoxorubicin (Adriamycin®), an anthracycline chemotherapeutic medication, has been effective against several types of malignancies since the 1960s [1,2]. It is the most valuable cytotoxic medication approved by oncologists in tandem with other anti-tumor medications or radiation and surgery [3]. It is highly potent and effective against solid tumors, i.e., breast, lung, bladder, gastrointestinal, thyroid, testicular, and ovarian carcinoma [4,5]. It is also used for treating hematological cancers, i.e., Hodgkin’s and non-Hodgkin’s lymphoma and pediatric leukemia [2,3]. Two proposed mechanisms for DOX antineoplastic effects have been reported [6]. The first one is through DNA chelation as DOX interacts with DNA, inhibits topoisomerase-II progression, and hinders DNA repair, which triggers DNA damage and cell death [7]. The second mechanism includes reactive oxygen species creation (ROS) and oxidative stress induction [8]. In vivo, DOX is metabolized into an unstable semiquinone, which is transformed back to DOX in a reaction that discharges ROS and reactive nitrogen species (RNS), causing lipid peroxidation, cell membrane, DNA, and proteins damages, and prompts apoptotic pathways of cell downfall to kill cancer cells [8,9]. Genes that can regulate this pathway include those involved with the oxidation outcome (xanthine oxidase, NADH dehydrogenases, and nitric oxide synthases) and those which disable free radicals, involving catalase (CAT), superoxide dismutase (SOD) and glutathione peroxidase (GPx) [9,10]. Nevertheless, such effects are not selective for cancer cells alone as the same mechanisms can also affect healthy cells [11,12].It has been reported that DOX administration can induce structural and functional cardiac alterations, i.e., ventricular distension, diminished output, systolic and diastolic disturbance [13,14], congestive heart failure (CHF), left ventricular remodeling, and cardiomyopathy [15,16]. The detailed mechanisms behind DOX-induced cardiac injury have not been elucidated, but it is possibly involved with several paths. Previous studies reported that DOX-induced cardiotoxicity involved the production of oxidative ROS [17]. As DOX enters the body, it binds tightly to cardiolipin present in the inner mitochondrial sheath [18], accumulates in mitochondria, and affects the electron transport chain creation of ROS and RNS [19]. Later, they aggravate mitochondrial and cellular membrane damage and diminish the antioxidant defense system [20,21], subsequently leading to cell apoptosis [21].Mitochondrial damage can also initiate intracellular Ca2+ imbalance [22], which further affects the apoptosis paths and causes myocardial cell death [23]. DOX also interferes with iron regulation [24], up-regulates NF-κB expression, which consequently causes the release of pro-inflammatory cytokines, i.e., tumor necrosis factor-alpha (TNF-α), interleukin-1 beta (IL-1β), and interleukin-6 (IL-6), and triggers vascular and cardiac inflammatory reaction [25] and exaggerates their downstream apoptotic pathways [26]. In addition, oxidative stress activates several pro-fibrogenic factors, which enhances the accumulation of extracellular matrix and development of cardiac fibrosis [27], remodeling [28], and eventual cardiac dysfunction [29]. DOX is still in use. To remain an efficient anticancer medication, it is essential to find appropriate new therapeutic agents to serve as adjuvants to mitigate DOX-induced cardiotoxicity. Several therapeutic strategies were developed to minimize DOX-induced oxidative injury, inflammation, DNA damage, and apoptosis. However, most of them interfere with DOX’s therapeutic effects, limiting their clinical use for cardio-protection against DOX-induced cardiotoxicity [30] and up till now no specific effective and safe drugs have been found.Phytomedicine is one of the strategies that focus on chemical substances naturally present in plants to improve health conditions and prevent, manage, and treat many diseases. Plant phenolics are natural antioxidant agents that embrace an electron that forms comparatively stable phenoxyl radicals and, consequently, interrupt the redox reactions within the cells [31]. They were also found to activate a cellular redox defense mechanism by stimulating endogenous antioxidant fractions [32] and keep the cells from xenobiotic oxidative stress, DNA impairment, and apoptosis [33]. Much attention has been given to using plant chemicals as a defensive strategy to resolve cardiotoxicity triggered by DOX [34,35,36].Date palm pollen (DPP) is a powder formed from date palm (Phoenix dactylifera L.) male reproductive cells. It has been utilized by the initial Chinese and the primeval Egyptians as a regenerating factor and worldwide as a dietary supplement [37,38]. Yearly, approximately one thousand tons of DPP are created in Arabic areas [39]. DPP is rich in many health-promoting factors, i.e., flavonoids and volatile unsaturated fatty acids [40,41]; that have strong antioxidant properties in scavenging free radicals [41,42]. In addition, DPP has anti-inflammatory, anti-coccidial, aphrodisiac, anti-apoptotic actions, and is a hepatoprotective agent [43,44,45,46]. Egyptian DPP has been proven to have a vast range of nutritionally and biochemically bioactive constituents, i.e., essential and non-essential amino acids, nucleic acids, different carbohydrates, trace elements, minerals, and vitamins. It also contains important phenolic compounds, including gallic, caffeic, coumaric, cinnamic, ferulic acids, catechin, rutin, quercetin, and naringenin propyl gallate. It also contains saturated (arachidic, capric, lauric, myristic, palmitic, and stearic) and unsaturated (arachidonic, linoleic, linolenic, oleic and palmitoleic) fatty acids, ω3, ω6 [47] and a lot of enzymes and cofactors [38,48]. Furthermore, Egyptian DPP has estrogenic substances, i.e., estriol, estradiol (E2), and estrone, which were recognized to alleviate male subfertility problems through their gonadotrophic activity [49]. This study was intended to explore the potential ameliorative effect of date palm pollen ethanolic extract (DPPE) against DOX-induced cardiotoxicity and the mechanisms underlying it.2. Materials and Methods2.1. Chemicals, Kits, and ReagentsDoxorubicin HCl (Adricin®) injectable solution was procured from EIMC United Pharmaceuticals (Badr City, Cairo, Egypt). Commercially available kits for the measurement of ALT, AST, LDH, CK, CK-MB, GSH, GPx, CAT and SOD pursuits, and NO and MDA contents were obtained from Biodiagnostic Co. (Cairo, Egypt). Rat specific ELISA kits for cTnI and cTnT levels were gained from MyBiosource, Inc. (San Diego, CA, USA). ELISA kit for NT-proBNP was obtained from CUSABIO (Hubei, China). A commercially available colorimetric kit for Ca+2 was purchased from Elabscience Co. (Houston, TX, USA). Rats-specific ELISA kits for IL-6, IL-1β, and TNF-α were bought from BD Biosciences (San Jose, CA, USA). Masson’s trichrome was purchased from (Sigma Aldrich, St. Louis, MI, USA). NF-κB p65 total ELISA Kit, hydroxyproline colorimetric assay kit, hematoxylin and eosin stain (H&E), rabbit polyclonal anti-TGF-β1 antibody (Product# ab25121), rabbit polyclonal anti-cleaved caspase-3 antibody (Product# ab4051) and rabbit monoclonal anti-Bax antibody E63 (Product# ab32503) were purchased from Abcam Co. (Cambridge Science Park, Cambridge, UK). Rabbit polyclonal anti-Bcl-2 antibody (Product# PA5-27094) was obtained from Thermo Fisher Scientific Co. (Waltham, MA, USA).2.2. Date Palm Pollen Grains Collection and Ethanolic Extract PreparationDate palm pollen grains were gathered from Phoenix dactylifera L. in March 2020 from El-Beheira, Egypt. They are authenticated at the Department of Botany, Faculty of Science, Alexandria University. After collection, the pollen grains were dissected from the bark and washed with water, dried with air, and ground at room temperature using a grinder to fine powder kept at 4 °C until use.Two hundred grams of DPP powder was extracted twice with 1600 mL of 80% ethanol for 24 h at room temperature. The extract was filtered in a Buchner funnel and then centrifuged at 5000 radius centrifugation force (RCF) for 30 min. The obtained supernatant was evaporated at 40 °C in a rotary evaporator under vacuum till complete dryness; then, the final dry extract and stock solution was preserved in dark glass bottles in the refrigerator at 4 °C for further analysis. The DPPE was re-dispersed in distilled H2O and orally intubated to treated rats using an intragastric tube at the time of experimentation.2.3. Acute Oral Toxicity of DPPE “Median Lethal Dose, LD50”Acute toxicity trial was carried out following the guidelines of the Organization for Economic Co-operation and Development [50] to evaluate the acute oral hazard of DPPE. The ‘Limit Test’ in the up and down procedure (UDP) was conducted to reduce the overall animals’ suffering. A maximum of 5 male Wister albino rats per group was administered sequentially with DPPE up to a test dose of 5000 mg/kg bw. Twenty-five adult male Wister albino rats (180 ± 10 g bw) were allocated randomly into 5 groups (5 rats each) and were acclimatized for 7 days. The rat groups were fasted overnight and the next morning, freshly prepared DPPE was orally administrated to groups 1–5 at doses of 1, 2, 3, 4, and 5 g/kg bw, respectively. The limit test involved dosing an animal with up to 5 g/kg bw. If the animal managed to survive, two extra animals were dosed. If both animals stayed alive, the LD50 was supposed to be higher than the limit dose, and the test was finished. The rats were observed every 2 h for 24 h, and again at 48 and 72 h to record any behavioral changes, signs of toxicity, and mortalities. The survived animals were observed for any delayed toxic signs or death for the next 14 days.2.4. Animals Experimentation and SamplingForty adult male Wister albino rats weighing approximately 190 ± 10 g (10 weeks-old) were purchased from the Medical Research Institute, Alexandria University, Egypt. The rats were kept in stainless-steel boxes at controlled environment “temperature 25 ± 5 °C and humidity 55 ± 5%” with a 12 h light/dark cycle and free access to standard rat feed (El-Nasr Co., Cairo, Egypt) and water for 2 weeks before the experiment to follow-up normal growth and behavior. The animals were given humane treatment in compliance with the Institutional and National Procedures for the Care and Use of Experimental Animals (NIH). They were declared by the Local Committee of the Faculty of Veterinary Medicine, Alexandria University (Ethical Committee Approval Number: 2020/013/59) and ethical approval of Taif University (42-0081).After acclimatization, the rats were randomly distributed into 4 equal groups (10 rats each). Group I (CTR) rats weighing approximately (192 ± 3.6) were orally intubated with 1 mL distilled water using a stomach tube daily for about 4 weeks. Group II (DPPE-treated) rats weighing approximately (190 ± 8.1) were orally intubated with DPPE at a dose of 0.5 g/kg bw daily for 4 weeks. Groups I and II were also intraperitoneally injected with 0.5 mL isotonic saline solution six times over the last two weeks of the experiment. Group III (DOX-treated) rats weighing approximately (195 ± 5.2) were orally intubated with 1 mL distilled water for 4 weeks and were DOX injected intraperitoneally at a dose of 2.5 mg/kg bw six times over the last two weeks of the experiment [51] with an accumulative dose of 15 mg/kg bw. Group IV (DPPE + DOX) weighing approximately (194 ± 4.5) rats obtained DPPE and DOX at the same dosage used in groups II and III. DPPE was administrated to rats an hour before DOX administration (Figure 1).At the end of the experimentation, the animals were only allowed free access to water and fasted for 12 h. After that, they were weighed, and blood samples were obtained just before euthanasia from the retro-orbital plexus of the inner eye canthus under diethyl ether anesthesia. The collected blood was centrifuged for 10 min at 3000 rpm, and then the resulting sera samples were kept at −20 °C for further analysis. Subsequently, rats were euthanized by cervical dislocation. The heart was rapidly harvested, rinsed with saline, dried, weighted, and dissected. The cardiac specimens were immediately frozen and kept at −80 °C. In an ice-cold phosphate buffer saline, the frozen samples were thawed and homogenized. (0.1 M pH 7.4) utilizing a homogenizer with a Teflon pestle and then centrifuged at 5000× g for 15 min. at 4 °C. Aliquots of the supernatant were frozen at −80 °C for the chemical analysis. In neutral buffered formalin 10% solution, other heart specimens were immediately fixed for the histopathological and immunohistochemical assessment.2.5. Assessment of the Body, Heart, and Relative Heart WeightsAt the end of the experimentation, in each rat, the body and heart weight were recorded. The relative heart weights (RHW) were estimated using the following formula: RHW=Heart weight (g)Bodyweight (g)×1002.6. Assessment of Cardiac Injury Biomarkers and Cytosolic Calcium (Ca+2)The serum ALT, AST, LDH, CK, and CK-MB levels were estimated as instructed by the manufacturers. The serum cTnI, cTnT, and NT-proBNP (a marker of heart failure) concentrations were also measured using the corresponding enzyme-linked immunosorbent assay (ELISA) kits using ELISA Plate Reader (Bio-Rad, Hercules, CA, USA). The supernatant obtained by centrifugation of cardiac tissue homogenate was used to evaluate the concentration of cytosolic Ca2+ using the Ca2+ colorimetric assay kit, as instructed by the manufacturer.2.7. Estimation of Cardiac Nitro-Oxidative Stress and Lipid PeroxidationThe concentration of nitric oxide (NO) was assessed in the supernatants of the cardiac homogenates based on the enzymatic reduction of nitrate to nitrite. For nitrite detection, the colored azo dye product “Griess reaction” was spectrophotometrically monitored at 550 nm absorbance [52]. The levels of MDA [53] and GSH [54] and the activities of GPx [55], CAT [56], and SOD [57] were spectrophotometrically estimated in the cardiac tissue homogenates. The total protein content was also assessed [58].2.8. Assessment of Inflammatory MarkersThe cardiac total NF-κB p65, TNF-α, IL-1β, and IL-6 were measured in the supernatant obtained by centrifugation of cardiac tissue homogenate using the corresponding rat-specific ELISA kit following the manufacturer’s protocols.2.9. Estimation of Hydroxyproline ContentBriefly, about 100 mg of the right ventricle was homogenized in double-distilled water. In a tightened screw-capped polypropylene vial, the tissue homogenates were mixed with conc NaOH (10 N) and then boiled for an hour at 120 °C. The alkaline lysate was ice-cooled, neutralized to pH 7.0, and centrifuged to get off supernatants. The hydrolysates were hot air-dried, chloramine T-oxidized, and finally reacted with Ehrlich’s reagent. The resultant colored product was measured at 560 nm absorbance, and the amount of hydroxyproline content was detected by comparing it with a standard curve [59].2.10. Histopathological Assessment and Semi-Quantitative Scoring ApproachCardiac samples were immediately fixed in phosphate-buffered formalin (10%, pH 7.4) after necropsy for 24 h, then were handled using the conventional paraffin embedding method. The 5 μm thick pieces were cut and placed on slides, deparaffinated in xylene, and rehydrated using decreasing concentrations of ethanol. One set of slides was hematoxylin and eosin (H&E)-stained for the routine histopathological setting. An additional set was Masson’s trichrome-stained for detecting the amount and distribution of collagen fibers [60]. Stained sections were blindly examined using light microscopes and photographed using a digital camera at a magnification of 400× (Nikon Corporation Co., Ltd., Tokyo, Japan).To convey the occurrence and severity of the histopathological lesions, a semi-quantitative scoring approach was used. In each animal group, seven H&E-stained slides (one slide/rat) were examined, and 10 random fields per slide were used for grading the various pathological lesions in a blinded fashion. The severity of pathological lesions was asessed according to the percentage of tissue affected in the entire section as None (−): normal histology with zero immersion of the inspected field, Mild (+): 5–25% of the tested field was involved, Moderate (++): 26–50% of the inspected field was involved, Severe (+++): >50% of the examined field was applied. The incidence represented the number of lesion rats per total examined.2.11. Immunohistochemical AssessmentAccording to Hsu, et al. [61], the immunodetection was assessed using four overlapping paraffin-embedded cardiac tissue sections. Sections were sliced at 4 µm thicknesses utilizing a microtome and put-on slides that are positively charged. Then, the sections were deparaffinized, rehydrated in xylene, then in different graded ethanol solutions and underwent antigen repossession using sodium citrate buffer (10 mM, pH 6.0) in the microwave at 105 °C for 10 min. Then, the activity of endogenous peroxidase was inhibited with 3% H2O2 for 10 min.; the non-specific proteins were blocked with 5% goat serum for 30 min at room temperature. The cardiac tissue slices were washed thrice in Dako Tris-buffered saline (TBS) and then incubated with the specific rabbit primary antibodies: polyclonal anti-TGF-β1 (dilution 1/200), anti-cleaved caspase-3 (dilution1/100), monoclonal anti-Bax (dilution1/250), and anti-Bcl-2 (dilution 1/100) at 4 °C overnight. In the negative control sections, normal IgG was substituted for the primary antibodies at the same concentration and antibody species. Following PBS washing, the tissue sections were incubated for an hour with goat anti-rabbit biotin-labeled secondary antibody, rinsed in PBS for 2 min. Then the sections were incubated with streptavidin-horseradish peroxidase reagent (VECTASTAIN1 Elite ABC kit, Vector Laboratories, Inc., Burlingame, CA, USA) at 37 °C for 20 min then washed with rinsing buffer and incubated with 3,3-diaminobenzidine tetrahydrochloride (DAB Substrate Kit, Thermo Fischer Scientific, Rockford, IL, USA) as the chromogen to developed peroxidase reaction. The sections were finally objected to Mayer’s Hematoxylin to augment the nuclear staining and mounted with di-poly cysteine xylene (DPX). All slides were assessed blindly and photographed using a digital camera.2.12. Histomorphometric AssessmentThe assessment was performed using H&E, Masson’s trichrome, and immunostained cardiac sections (one section from each rat and seven per group). The digital images (ten different fields per section at ×400 magnification power) were blindly analyzed using image analysis software (ImageJ Version 1.47, National Institutes of Health, Bethesda, MD, USA, wayne@codon.nih.gov. The ten values were averaged in each animal, and the average was used as individual sampling data.Using H&E-stained sections, the cross-sectional area of cardiomyocytes was assessed in the left ventricular wall. Fifteen cardiomyocytes with a visible nucleus and intact cellular membrane were selected per field for the measurement and analysis [62].Using Masson’s trichrome-stained sections [63], the collagen volume fraction (CVF %) and perivascular collagen area (PVCA %) percentage were estimated as per the following formulas: CVF (%)=Collagen areaTotal area×100; PVCA (%)=area occupied by the collagen/total area of the vessel section×100 .Using images of immunostained slides, the area percentage (%) of TGF-β1, cleaved caspase-3, Bax, and Bcl-2 immunopositive cardiomyocytes were estimated as area percent (%) across ten different fields/sections [64].2.13. Data AnalysisOne-way Variance Analysis evaluated the numerical data estimation (ANOVA)test using SPSS data analysis software (Version 21; SPSS Inc., Chicago, IL, USA) and summarized it as means ± (SEM). Tukey’s post-hoc test was used to ascertain the statistical difference between experimental groups. * p < 0.05 was set as statistically significant.3. Results3.1. Median Lethal Dose, Mortality, and Survival RatesRats of DPPE groups did not exhibit any behavioral changes, toxic side-effects, or even mortalities after 24 h and 14 days post-treatment. Thus, dosing was ceased at 5 g/kg bw. Consequently, the LD50 of DPPE was evaluated to be more than 5 g/kg bw. The CTR and DPPE group did not exhibit any mortality all over the experimental period. However, the DOX group showed a scruffy appearance and exhibited 30% mortality (three dead rats), and DPPE + DOX group exhibited 10% mortality (one dead rat) (the data not shown).3.2. Body, Heart, and Relative Heart Weights and Myocyte Cross-Sectional AreaAs demonstrated in Table 1, CTR and DPPE groups exhibited a statistically non-significant change in the body weight, heart weight, RHW, and myocyte cross-sectional area. Conversely, the DOX group displayed a significant reduction in body weight (≈0.86-fold) and a substantial rise in the heart weight, RHW, and the myocyte cross-sectional area (1.6, 1.7, and 1.52-fold, respectively) compared to CTR values. Nevertheless, DPPE + DOX group exhibited a non-substantial rise in body weight (≈1.08-fold) and a considerable reduction in heart weight (≈0.76-fold), RHW (≈0.71-fold), and the myocyte cross-sectional area (≈0.73-fold) compared to DOX group values. In comparison with the CTR group values, DPPE + DOX group expressed a non-significant reduction in body weight (≈0.93-fold), a significant rise in the heart weight (≈1.22-fold), a non-significant increase in the RHW (≈1.2-fold), and a substantial increasing in the cardiomyocyte cross-sectional area (≈1.22-fold).3.3. Cardiac Injury Biomarkers and Cardiac Cytosolic Calcium (Ca2+)As demonstrated in Table 2, the CTR and DPPE groups disclosed a statistically non-significant change in the activities of ALT, AST, LDH, CK, CK-MP, and the levels of cTnI, cTnT, NT-pro BNP, and cardiac Ca2+. Meanwhile, the DOX group exhibited a significant increase in ALT (≈1.52-fold), AST (≈1.42-fold), LDH (≈2.49-fold), CK (≈2.8-fold) and CK-MP (≈2.03-fold) activities, and cTnI (≈5-fold), cTnT (≈3.44-fold), NT-pro BNP (≈2.38-fold), and cardiac Ca2+(≈1.74-fold) levels as compared to CTR values. In contrast, DPPE + DOX group demonstrated a significant reduction in ALT (≈0.79-fold), AST (≈0.85-fold) LDH (≈0.73-fold), CK (≈0.7-fold) and CK-MP (≈0.72-fold) activities, and cTnI (≈0.57-fold) fold, cTnT (≈0.54-fold), NT-pro BNP (≈0.63-fold) and cardiac Ca2+ (≈0.82-fold) levels, as compared to DOX values. In addition, this group showed a significant increase in the activities of ALT (≈1.2-fold), AST (≈1.22-fold), LDH (≈1.83-fold), CK (≈1.97-fold), and CK-MP (≈1.48-fold), and the levels of cTnI (≈2.89-fold), cTnT (≈1.84-fold) NT-pro BNP and cardiac Ca2+(≈1.43-fold), as compared to the CTR values.3.4. Cardiac Nitro-Oxidative Stress and Lipid PeroxidationAs demonstrated in Table 3, the CTR and DPPE groups revealed a non-significant change in the concentrations of NO, lipid peroxidation marker (MDA), and antioxidant parameters (GSH level and GPx, CAT, and SOD activities). Meanwhile, the DOX group disclosed a statistically substantial increase in the quantities of NO (≈3.61-fold), MDA (≈1.86-fold), and a significant reduction of the GSH level, and GPx, CAT, and SOD pursuits (≈0.46, 0.31, 0.54 and 0.31-fold, respectively) compared to the CTR values. Quite the opposite, the DPPE + DOX group demonstrated a considerable drop in the levels of NO (≈0.58-fold) and MDA (≈0.69-fold) and a significant rise in the GSH level and GPX, CAT, and SOD activities (≈1.67, ≈2.18, ≈1.41 and ≈2.28-fold, respectively) equated to the DOX group values. Compared to the CTR group values, DPPE + DOX group showed a significant rise in the levels of NO (≈2.11–fold), MDA (≈1.29–fold), and a considerable reduction in the GSH level, and GPx, CAT, and SOD pursuits (≈ 0.77, 0.68, 0.77 and 0.72-fold, respectively).3.5. Inflammatory MarkersAs shown in Table 4, the CTR and DPPE groups displayed a statistically non-significant change in the NF-κB p65, TNF-α, IL-1β, and IL-6 concentrations. Meanwhile, the DOX group exhibited a significant rise in the quantities of NF-κB p65 (≈2.63-fold), TNF-α (≈2.35-fold), IL-1β (≈1.74-fold), and IL-6 (≈1.85-fold) equated to CTR values. Conversely, DPPE + DOX group disclosed a significant reduction in the concentrations of NF-κB p65 (≈0.72-fold), TNF-α (≈0.63-fold), IL-1β (≈0.82-fold), and IL-6 (≈0.74-fold) compared to the DOX group values. Compared with CTR values, DPPE + DOX group exhibited a significant rise in the NF-κB p65 (≈1.9-fold), TNF-α (≈1.48-fold), IL-11β (≈1.43-fold), and IL-6 (≈1.36-fold) levels.3.6. Cardiac Hydroxyproline ContentAs demonstrated in Table 4, the CTR and DPPE groups displayed a non-significant change in the hydroxyproline concentrations. Meanwhile, the DOX group exhibited a substantial rise of about 2-fold when equated with the CTR quantity. In contrast, the DPPE + DOX group displayed a significant reduction (≈0.83-fold) compared to DOX values and a considerable enhancement (≈1.6-fold) compared to the CTR value.3.7. Histopathological Results and Lesions ScoringFigure 2 demonstrated the histomorphological results of H&E-stained cardiac tissue sections. Table 5 also illustrated the prevalence and severity of the identified pathological lesions in various treatments.Heart tissues from the CTR (Figure 2a) and DPPE (Figure 2b) groups revealed normal histoarchitecture with well-organized and branched cardiac myofibers. The cardiomyocytes were closely arranged with oval centrally located nuclei, eosinophilic cytoplasm, and cross striations. In addition, minimal interstitial connective tissue and few fibroblasts were noticed. Meanwhile, the DOX group exhibited moderate to severe histological alterations and high lesion scores (Table 5), wherein disoriented cardiac myofibers with wavy appearance were evident. Furthermore, myocardial degenerative changes such as sarcoplasmic vacuolization (Figure 2c), myofibrillar flocculation, and fragmentation were noticed. Many cardiomyocytes showed Zenker’s degeneration. Meanwhile, others exhibited Zenker’s necrosis. Myocytolysis (Figure 2d) and multifocal zones of myocardial necrosis combined with infiltrations of mononuclear inflammatory cells (Figure 2e) were obvious. Additionally, hyperemia of interfibrillar blood vessels (Figure 1f), perivascular edema with inflammatory cell infiltrations (Figure 2g), intramyocardial edema, fibrin deposition, and focal areas of hemorrhage were noticed. There were interfibrillar infiltrations of active fibroblasts with a hypertrophic nucleus and marked myocardial and perivascular fibrosis. On the contrary, the DPPE + DOX group displayed a marked enhancement in cardiac tissue structure and integrity. Nevertheless, they were not identical to the CTR limits. Compared with the DOX group, the previously noted lesions were less in severities and distribution in the DPPE + DOX group (Figure 2h, Table 5).3.8. Masson’s Trichrome Staining and Histomorphometric FindingsAs illustrated in Figure 3, the cardiac tissue sections of the CTR (Figure 3a,e) and DPPE (Figure 3b,f) groups displayed normal spreading of greenish delicate collagen fibers in between the cardiomyocyte fibers and around the intramyocardial coronary vessels. They also revealed non-significant changes (p > 0.05) in the mean CVF% (Figure 3i) and PVCA% (Figure 3j). Conversely, the DOX group (Figure 3c,g) exhibited an apparent increase in the amount of collagen fiber deposition as well the mean CVF % (≈8.5-Fold) and PVCA% (≈3.09-fold) as compared to the CTR values (Figure 3i,j, respectively). However, the DPPE + DOX group revealed a marked reduction in collagen fiber deposition (Figure 3d,h). Meanwhile, the mean CVF % and PVCA% showed significant (p > 0.05) decrease with approximately 0.61-fold and 0.68-fold, respectively, when linked to DOX group values and significant (p > 0.05) increases with approximately 5.24 and 2.11-fold, correspondingly when equated to the CTR values (Figure 3i,j, respectively).3.9. Immunohistochemical AnalysisFor the expression of TGF-β1 in the cardiac tissue of the control groups, the CTR (Figure 4a) and DPPE (Figure 4b) groups revealed a normal expression of TGF-β1 (brown color). Both groups showed no significant alterations in the mean immune-stained area % of TGF-β1 (Figure 4e). Conversely, DOX-treated rats exhibited a noticeable increase in TGF-β1 expression (Figure 4c) with a substantial (p < 0.05) rise of the mean immune-stained area% (≈2.6-fold), as associated with the CTR group value (Figure 4e). However, DPPE + DOX-treated rats exhibited a conspicuous decrease in TGF-β1 expression (Figure 4d) with a substantial (p > 0.05) decline in mean immune-stained area% (≈0.7-fold) as compared to the DOX group value and a significant (p > 0.05) increase (≈1.84-fold) concerning CTR group quantity (Figure 4e).Referring to cleaved caspase-3, Bax, and Bcl-2 expressions, the control (Figure 5a, Figure 6a and Figure 7a) and DPPE (Figure 5b, Figure 6b and Figure 7b respectively) groups revealed weak representation of cleaved caspase-3 and Bax. In addition, both groups showed a robust expression of Bcl-2 with dispersed, intensely brown stained immune-reactive cardiomyocytes. There were no significant alterations in the mean area% of cleaved caspase-3, Bax, and Bcl-2 immuno-stained cells (Figure 5e, Figure 6e and Figure 7e, respectively). Meanwhile, DOX-treated rats’ cardiac tissues displayed moderate to strong expression and immune-staining of cleaved caspase-3 (Figure 5c) and Bax (Figure 6c) and weak expression and immune-staining for Bcl-2 (Figure 7c). Relative to the CTR group values, the mean area% of immune-stained cells showed a substantial (p > 0.05) rise for cleaved caspase-3 (≈8.95-fold) and BAX (≈2.37-fold) and a significant decrease for Bcl-2 (≈0.3-fold) (Figure 5e, Figure 6e and Figure 7e, respectively).The cardiac tissues of DPPE + DOX-treated rats exhibited weak expression and immune-staining of cleaved caspase-3 (Figure 5d) and Bax (Figure 6d). It also showed moderate to strong expression and immune-staining of Bcl-2 (Figure 7d). Constantly, the mean immune-stained areas % in DPPE + DOX group showed a significant (p < 0.05) reduction for cleaved caspase-3 (≈0.39-fold) and Bax (≈0.72-fold) and a considerable increase for Bcl-2 (≈1.9-fold) as associated with the DOX treated group values (Figure 5e, Figure 6e and Figure 7e, respectively). Meanwhile, they showed a substantial (p < 0.05) surge for cleaved caspase-3 (≈3.55-fold) and Bax (≈1.72-fold) and a significant decrease for Bcl-2 (≈0.59-fold), as equated with the CTR group (Figure 5e, Figure 6e and Figure 7e, respectively).4. DiscussionAnthracyclines, including doxorubicin, play a crucial role in chemotherapy for the medication of numerous solid organ tumors and hematologic malignancy [12,16]. However, dose-based cardiotoxicity of anthracyclines is frequently reported to limit their therapeutic efficacy [5,65,66]. Since the DOX cardiotoxic effects are generally irreversible, searching for new protective approaches that could interrupt DOX-induced pathogenic events and confer protection against its cardiotoxicity should be developed [67]. Currently, the handout study is nearly the first to verify the meliorative potential of DPPE on DOX-induced cardiotoxicity. DOX cardiotoxicity is the ultimate obstacle to be solved to enhance its clinical usage [68]. In this work, rats were treated with 15 mg DOX/kg bw as a cumulative dose to mimic its chronic cardiotoxicity, as seen in clinical therapies [69]. Several pathways participate in DOX-induced cardiotoxicity.However, the main mechanism involved is ROS generation, which causes peroxidation of lipids and depletion of antioxidant enzymes [35]. The cardiac tissue has many mitochondria because it needs much energy, making it more susceptible to DOX toxicity [67]. They have a high DOX affinity because their inner membrane encompasses cardiolipin, an anionic phospholipid with a high binding affinity to cationic DOX [67,70]. In mitochondria, cytochrome p450 reductase, xanthine oxidase, and NADPH dehydrogenase have to convert DOX into a semiquinone radical that interacts with molecular oxygen forming superoxide anion and an extra ROS [67]. Since free radicals act as the main contributor to DOX-induced cardiotoxicity, antioxidant compounds are known to be possible protective and therapeutic agents [71,72]. Dexrazoxane is the only synthetic medication used for cardiotoxicity prevention in clinical conditions [73,74]. DPP contains polyphenols and flavonoids with effective antioxidant and anti-inflammatory potentials that clarify its prospective use in many diseases. It also has antimicrobial, anti-coccidial, anti-apoptotic, and hepatoprotective potential [43,44,46]. Similarly, DPP is used as an anti-toxicant [42] and provides a cardio-preventive ability against isoproterenol-triggered myocardial infarction [42]. Its effect against DOX-induced cardiotoxicity has never been investigated.The DOX group showed a scruffy appearance and exhibited 30% mortality. This result was parallel to those reported by Wu, et al. [75]. However, in DPPE + DOX-treated animals, there was relatively low mortality, reflecting that pre-cotreatment with DPPE might improve the survival of DOX-intoxicated rats. Additionally, the DOX persuaded a significant reduction in body weight [76,77], which might be due to decreased appetite, reduced protein synthesis, mucositis, and/or inadequate assimilation of nutrients [78,79,80]. The improved body weight in DPPE + DOX-treated animals mirrored the protective effects of DPPE. Cardiomyopathy caused by DOX is a shift from myocardial hypertrophy to heart failure [81]. In animals, myocardial hypertrophy is mainly assessed by measuring heart index weight [82]. In the current work, the DOX caused an increment in heart and relative heart weights and cardiomyocyte cross-sectional area, which indicated ventricular hypertrophy [75,81,83,84,85]. However, pre-cotreatment with DPPE revealed a reduction in the previous parameters, suggesting the ability of DPPE to maintain the normal integrity of cardiomyocytes. The cardiac enzymes ALT, AST, CK, CK-MB, LDH, CK, LDH, cTnI, and cTnT are a dynamic bioindicator of myocardial injury [35]. The DOX-induced cardiac damage was evident through the substantial rise of cardiac injury biomarkers: ALT, AST, LDH (not very specific biomarkers), CK, CK-MB, cTnI, and cTnT (more specific and sensitive biomarkers) activities, reflecting cardiomyocyte membrane disruption and extensive cardiomyocyte damage [59,76,86]. However, they were reduced following pre-cotreatment with DPPE, suggesting its competency to maintain the normal integrity of cardiac muscle and to inhibit DOX-induced myocardial damage [42]. An earlier study showed that antioxidant compounds could decrease cardiac function biomarkers in DOX-intoxicated rats [35].The N-terminal pro-brain natriuretic peptide is a peptide produced to control blood pressure fluid equilibrium. It is liberated from the heart following ventricle volume expansion and/or pressure overload [87]. A large amount of NT-pro BNP is released into the blood during cardiac insufficiency, so it is considered a sensitive biomarker of congestive heart failure [88]. It is also a valuable predictor in patients with anthracycline chemotherapy as a vital biomarker of left ventricular dysfunction [89]. Herein, the DOX-induced, a dramatic increment of serum NT-pro BNP level, demonstrating that it can cause acute cardiac failure [35,90]. However, pre-cotreatment with DPPE lowered NT-pro BNP serum level, suggesting that DPPE may defend the heart from DOX-induced toxicity and cardiac damage.The generation of large quantities of ROS and O2− in DOX metabolism results in DNA and mitochondrial injury, therefore enhancing lipid peroxidation in the cell membrane and increased MDA levels in cardiac cells [91]. In turn, free radicals released in response to DOX can interfere with the balance between oxidative and antioxidants agents, followed by depletion of the endogenous myocardial antioxidant compounds (GSH) and enzymes (GPx, SOD, and CAT) [77]. Herein, DOX-intoxicated rats exhibited increased oxidative and nitrosative stresses, indicated by an increment of MDA and NO, and a substantial decline in antioxidant enzyme activity [76,92]. The recorded decline in the antioxidant enzyme activity could be attributed to their utilization in the fight against oxidative stress [35]. Remarkably, pre-cotreatment with DPPE diminished MDA and NO levels, and improved antioxidant activity in cardiac tissue, suggesting the antioxidant ability of DPPE against DOX-induced oxidative/nitrosative stress. Furthermore, DPPE expressed antioxidant activity and defensive mechanisms by restoring oxidative stress/antioxidant balance in several toxic modules [40,49,93]. Many earlier studies of phytochemical or antioxidant elements have demonstrated their ability to reduce lipid peroxidation and improve the value of antioxidant markers in DOX cardiotoxicity [21,35,71,76,94,95]. The mechanism for protecting the DPPE may include de-activating potentially toxic metabolites and free radicals and potentiation of antioxidant paths [96]. Other investigators claimed that DPPE incorporates considerable amounts of flavonoids, phytosterols, and carotenoids [42,97], which are antioxidants with redox activities acting as reducing agents ROS/NOS quenchers [98].It is well established that increased free radicals output with excitotoxicity and lipid peroxidation accelerates inflammatory conciliators’ synthesis and thus activates the inflammatory response in the cardiac tissue [99]. NF-κB, a transcription factor, is involved in cell survival, inflammation, and immune responses. NF-κB p65 modulates the inflammatory responses, whereas its translocation to the nucleus enhances transcription of the pro-inflammatory cytokines, i.e., TNF-α, IL-1β, and IL-6 [100,101]. In turn, they provoke leukocyte infiltration into the myocardium and aggravate inflammatory injury [102]. Consistently, DOX stimulated NF-κB p65, TNF-α, and IL-1β production in cardiac tissues, which can also cause cardiomyocyte apoptosis by increasing Ca2+ store of heart muscle cells [26]. As well, DOX provoked a substantial rise in NF-κB level and correspondingly induced an increment in TNF-α and IL-1β quantities, reflecting enhanced inflammatory responses [71]. However, DPPE pre-cotreatment generated a substantial decline in NF-kB’s cardiac contents and the related downstream pro-inflammatory cytokines, TNF-α and IL-1β, indicating its potential to suppress the initiated inflammatory cascade.Disruption of Ca2+ homeostasis is another pathway implicated in DOX-generated cardiac toxicity. It is documented that DOX cardiotoxicity is accompanied by a high overload of Ca2+ in cardiomyocytes, resulting in inadequate contraction and interference with Ca2+ regulation, thus triggering ROS generation and leading to cell dysfunction [103,104,105]. In the present work, DPPE was found to mitigate oxidative stress-mediated Ca2+ overload in DOX-challenged cardiac tissues. This may be attributed to DPPE’s antioxidant properties, which suppress ROS generation and consequently reduce such high Ca2+. Moreover, ROS and O2− exaggerated cardiac dysfunction and mitochondrial damage induced in DOX therapy [17]. The formation of the Fe-anthracycline complex catalyzes the transformation of H2O2 to OH• radicals, resulting in severe cytoskeleton injury and plasma membrane disruption followed by myofibril loss, sarcoplasmic reticulum dilation, and myocardial necrosis [24,106].The findings of the cardiac tissues histopathological analysis supported the biochemical interpretations. Due to enlargement of the sarcoplasmic reticulum, DOX induced several degenerative changes in heart tissue, including myocardial hyalinization, and sarcoplasmic vacuolization [107], wavy myocardial fibers flocculation, and fragmentation. In addition, multifocal areas of myocardial necrosis, myofibrillar loss, inflammatory cell infiltrations, myocardial fibrosis, hemorrhage, vascular congestion, and interfibrillar edema were observed [71,74,100,108,109]. On the contrary, cardiac tissues of DPPE plus DOX-treated animals showed a marked improvement in cardiac tissue structure and integrity. According to the histomorphometric analysis, the previously declared lesions were less in incidence and severity. These remarks were comparable with other experiments that verified the potential of DPPE as anti-myocardial damage [42]. The literature reviewed that DPPE contains bioactive substances, including estradiol [110], stigmasterol [111], β-sitosterol [112], carotenoids lutein [113,114], δ-tocotrienol [115], and isorhamnetin [116], which have potential cardioprotective activities.Fibrosis is a reparative reaction to DOX-induced cardiotoxicity [85]. The necrotic or apoptotic cardiomyocytes are replaced by overproduced collagen by fibroblast. However, it contributes to heart rigidity and instability [85,117]. DOX-induced cardiac fibrosis is based on the inflammatory and growth factors signaling paths regulated by TGF-β1. Increased oxidative stress and the subsequent antioxidant depletion and lipid peroxidation trigger tissue inflammation and necrosis and enhance tissue fibrogenesis progression [35,59,118]. TGF-β1 is another key factor in the regulation of collagen production in DOX-induced cardiomyopathy. The pro-fibrogenic cytokine, TGF-β1, is a proliferation-mediated fibrotic protein produced by cardiac myofibroblast and is responsible for cardiomyocyte hypertrophy, apoptosis, and fibrosis. It may stimulate cardiac fibroblastic hyperplasia, increased production of type I and III collagen fiber and fibronectin, and cause increased extracellular matrix and decreased extracellular matrix degradation through inhibition of collagen enzyme release [28,119]. In the current work, DOX elevated the cardiac hydroxyproline level, the major component of fibrillar collagen [59]. Additionally, DOX-induced remarkable hypertrophy of fibroblast increased collagen deposition, and fibrosis was further confirmed by Masson’s trichrome staining of the cardiac tissues [120]. DOX also increased the TGF-β1 expression in cardiac tissues [85,121,122,123]. However, DPPE pre-cotreatment induced a marked reduction in collagen fiber deposition between cardiac muscle fibers and around the intramyocardial coronary vessels, hydroxyproline content, and TGF-β1 expression in cardiac tissues. Therefore, DPPE has a potential aptitude to maintain the normal integrity of cardiac muscle and inhibit DOX-induced myocardial damage, which attenuated fibrosis development via modulation of fibrogenic genes.The major regulators of apoptosis are Bcl-2 family members, which involve pro-apoptotic protein (Bax, caspase-3) and anti-apoptotic (Bcl-2) proteins [74]. During apoptosis, the Bcl-2 expression declines, while Bax and caspase-3 expressions rise [124]. Bax stimulation ensures cell damage by forming a pore in the mitochondrial membrane, leading to poly (ADP-ribose) polymerase cleavage and mitochondrial cytochrome-c induction, which mediate apoptosis [125]. Meanwhile, the Bcl-2 inhibits apoptosis by inhibiting mitochondrial permeability transition [126] in cardiomyocytes protecting mitochondrial structure and function [124]. Caspases are essential parts of the apoptotic process. Opened mitochondrial pores lead to mitochondrial cytochrome-C release, and activated caspase-3 triggers proteolytic degradation of cellular components death [127]. Herein, the immunohistochemical staining of myocardial tissues showed that DOX caused an increase in the cleaved caspase-3 and Bax expressions and decreased Bcl-2 expression, which reflected apoptosis’s role in DOX-induced cardiomyopathy [76,86,128]. Nevertheless, pre-cotreatment with DPPE reduced cleaved caspase-3 and Bax and enhanced Bcl-2 expressions, implying that DPPE’s anti-apoptotic activity could conserve myocardial integrity and mitigate myocardial damage.5. ConclusionsCollectively, for the first time from these observations, it is indicated that date palm pollen ethanolic extract displayed an effective cardioprotective potential against doxorubicin-induced cardiac myopathy. The antifibrotic and anti-apoptotic mechanisms of DPPE were attributed to suppressing cardiac oxidative/nitrosative damage, pro-inflammatory cytokines production, and fibrogenic and apoptotic gene expressions, thereby reducing myocardial myopathy and detrimental structural alterations. Accordingly, DPPE is highly recommended as an adjunct to avert the toxic side-effects caused by doxorubicin.

animals : an open access journal from mdpi

10.3390/ani12050659

PMC8909719

The aims of this study were (1) to evaluate the prevalence of lameness, dirtiness of the body surface, and abomasal disorders of slaughter cattle; and (2) to determine the association between these welfare indicators and animal-related factors (e.g., housing type, carcass weight, and transportation and waiting duration of the animals). In contrast to dirtiness (level of contamination of the body surface, also referred to as cleanliness) and the prevalence of abomasal disorders, the determined lameness prevalence was very low. The husbandry of cattle was identified as a significant influencing factor for both the dirtiness and occurrence of abomasal disorders of slaughter cattle.

Three cattle welfare indicators (lameness, dirtiness, and abomasal disorders) were evaluated in 412 slaughter cattle in a cross-sectional study in Austria. The aims of this study were (1) to evaluate the prevalence of lameness, dirtiness of slaughter cattle, and abomasal disorders; and (2) to determine the association between these welfare indicators and animal-related factors (e.g., housing type, carcass weight, transportation and waiting duration of the animals). The lameness prevalence was 0.73%, the abomasal disorders prevalence was 52.43%, and 88.59% of all cattle were contaminated. The latter result indicates that the cattle were kept in a dirty environment. The occurrence of abomasal disorders was associated with cattle housing systems (p ≤ 0.00) and slaughter weight (p = 0.03). The odds for abomasal disorders were 28.0 times higher for cattle housed on slatted flooring compared to cattle kept in a tethered system. The chance for occurrence of abomasal disorders was 3.6 times higher for cattle with a low carcass weight compared to cattle with a high carcass weight. Furthermore, significant associations were found between dirtiness (also referred to as cleanliness or contamination) and husbandry system, sex, and breed. Cattle housed in deep litter boxes had 40.8 times higher odds of being contaminated compared to cattle in a tethered housing system. Cows (odds: 32.9) and heifers (odds: 4.4) had higher odds of being contaminated with feces compared to bulls, whereby female calves (odds: 0.09) and male calves (odds: 0.02) had significantly lower odds of being contaminated. Furthermore, the breeds Brown Swiss (odds: 0.26) and Holstein-Friesian (odds: 0.14) had a significantly lower chance of being contaminated compared to Simmental cattle. Other collected factors, such as production system, transportation duration, life days of the cattle, average daily weight gain, carcass classification, and fat coverage, showed no association with the collected welfare indicators. The study presented here indicates that welfare indicators evaluated for slaughter cattle are suitable to assess cattle welfare, and improvements in husbandry may positively impact both the abomasal physiology and cleanliness of cattle.

1. IntroductionMore than 70 cattle welfare indicators are described in the literature [1]. These 70 cattle welfare indicators can be assigned to four main categories: morphometric, behavior-specific, physiologic, and meat-quality-affecting indicators. Besides the evaluation of animal welfare indicators on farms, several studies have assessed animal welfare issues in abattoirs [2,3,4,5]. A benefit of using ante- and postmortem indicators, including meat inspection, is the ability to collect data on animal welfare and on food safety from different farms [6]. Data regarding lameness, injuries, emaciation, and cleanliness are considered important animal welfare indicators at the slaughterhouse [7].Lameness is one of the most important cattle welfare issues [8,9]. Locomotion of cattle can be categorized based on the severity of lameness by a five-point scoring system [10]: (1) clinically normal, i.e., cattle walks and stands with level-back posture; (2) in contrast to score 1, the animal shows an arched-back posture while walking; (3) a moderate lameness is detected if an arched-back posture is observed while the cattle is walking and standing; (4) in contrast to score 3, an arched-back posture is observed all the time; and score (5) indicates a severe lameness characterized by an inability of the cattle to bear weight on one or more of their limbs/feet [10].Abomasal ulcers and lesions are the most common cause of digestion disorders in cattle of all ages and are important welfare indicators; however, the number of published prevalence studies is low [11,12]. Ulcers lead to pain, loss of production, and death in severe cases [13,14]. Abomasal lesions can be classified, based on distinct variations in clinical signs, between type I (superficial lesions of the abomasal mucosa) and II (deep lesions of the abomasal mucosa), categorized as non-perforating abomasal ulcers, and type III (perforating ulcers with acute, circumscribed peritonitis) and IV (ulcerations with diffuse peritonitis [15]), classified as perforating ulcers [12]. Risk for the occurrence of abomasal ulcers are nutritional factors, destruction of mucosa by high-concentrate feedstuff that decrease pH in the abomasum, mineral imbalance, stress, comorbidities, and medical treatments. For instance, straw is described to damage the mucosa if fed as solid roughage in veal calves [12]. An overview study of identified risk factors for abomasal ulcers was recently published [16].Fecal contamination of animals represents a hygienic problem during the slaughter process [17]. Dirty cowhides result in higher bacterial loads on concerned carcasses and represent a high risk for cross contamination during the slaughter process [18]. Besides hygiene, animal health, and food safety concern, contaminated animals might indicate poor on-farm management due to dirty flooring and litter, as described previously [19]. A five-step evaluation system is used to assess the level of the cleanliness of cattle developed by the Food Standard Agency (London, UK) [20]. Score 1 includes all clean and dry animals; score 2 incorporates all animals with a slight contamination; score 3 covers all animals with moderate contamination; score 4 indicates a high contamination of the slaughter animal; and score 5 includes all animals with extremely high contamination with feces, resulting in a prohibition of slaughter.The aims of the present study were (1) to evaluate the prevalence of the animal welfare indicators: lameness, dirtiness (also referred to as cleanliness or contamination) of the body surface, and abomasal ulcers and lesions of slaughter cattle; and (2) to estimate the association between the animal welfare indicators and animal-related factors, such as housing type, production system, transportation and waiting duration, sex, carcass weight, and classification.2. Materials and MethodsThe present study was carried out as a cross-sectional study at one slaughterhouse in Austria. The abattoir is EU certified for slaughter of cattle, pigs, and horses. One official veterinarian (J.B.) carried out the sampling at the slaughterhouse on 19 days from 27 July to 30 November 2020. Data were collected regarding (i) lameness, (ii) abomasal disorders, and (iii) dirtiness of slaughter cattle (dependent variables). We focused on these three factors because pre-study observations of the abattoir showed a wide range of dirtiness levels, although no emaciations and injuries were observed. Abomasal disorders were chosen as a factor because the occurrence seems to be far underestimated, measured by a scarcity of scientific studies. In contrast, lameness and dirtiness are commonly used as welfare indicators in scientific studies. Furthermore, these three indicators were also chosen for practical reasons, as one observer scored lameness, dirtiness, and abomasal disorders.Scoring of lameness was performed on hard ground after the animals arrived at the slaughterhouse based on the five-point scoring system described in [10]. Additionally, all cattle were scored regarding their cleanliness level in the waiting room of the abattoir and after stunning based on the five-point scoring system of the Food Standard Agency [20]. All abomasa were collected and inspected for the presence of lesions. After removing the abomasum from the gastrointestinal tract, it was opened on the side of with greater curvature, and stomach content was washed out. Abomasal lesions were characterized based on the four-point scoring system described in [15,21].Animal-related metadata (independent variables), such as breed, sex, housing type, transportation and waiting duration, daily weight gain, and production system (see detailed description in Table 1 and Figure 1), were collected for each slaughtered cattle from transport and health certificates, slaughter protocols, and the national cattle database. Information about the type of housing for each individual animal was gathered by interviewing the driver of the cattle transport, who picked cattle up on farms. Table 1 gives an overview of all collected data, the scale and distribution of the collected data, whether the data were included in the statistical analysis, and information about reclassification of some data for the statistical analysis due to low occurrence of some scores. For instance, the five categories of the abomasal lesion score were reduced to two categories (No: free of abomasal lesions; Yes: not free of abomasal lesions). Furthermore, low data variability within the variable “lameness” was determined (see Section 3). Therefore, the statistical analysis was performed for two response variables only, i.e., “abomasal lesions” and “cleanliness level” of the cattle. The “abomasal lesions” and “cleanliness level” of the slaughter cattle were analyzed in two different models due to different scales of the dependent variables: a binary scale of the recorded abomasal lesions (No or Yes) and an ordinal scale of the recorded cleanliness level (i.e., clean, low, medium, and highly contaminated with prohibition of slaughter).A mixed binomial logistic model (BGLMM) was performed based on Formula (1) to determine the association between the independent variables (xi: metadata), considered fixed factors (also known as predictor variables), and abomasal lesions alterations (yi), considered the dependent variable (also known as response variable). The farm and individual cattle were included as random factors (λf,c). A stepwise, backwards factor-selection approach using log-likelihood ratio tests was applied to compare models with different included fixed-factor combinations (with a threshold of 0.05) and to consider only the most relevant fixed factors in the backwards-fitted BGLMM final model [22]. The BGLMM final model was checked for overdispersion and collinearity using the variance inflation factor (VIF). An issue of collinearity exists when VIF is larger than 10, i.e., the variables cannot independently predict the value of the response variables in case of correlation between predictor variables [23]. To determine the goodness of fit and the predictive accuracy of the BGLMM model, a data partition was conducted in the training (60% of the dataset) and testing dataset (40% of the dataset). In this context, a confusion-matrix, representing the matches and mismatches between predictions and actual results, was calculated to determine the accuracy, sensitivity, and specificity, as well as the positive and negative predictive values for abomasal lesions. (1)yi={1 if the i-th cattle has abomasal lesions0 otherwise yi = ß0 + Σßixi + λf,c + E(2)To identify associations between “cleanliness level” (Y) of the cattle and the independent variables (i.e., animal-related metadata as predictor variable, also known as fixed factors), a cumulative link mixed model (CLMM) with adaptive Gauss–Hermite quadrature approximation was used. Because other farm specific herd management factors may also contribute to the cleanliness level of cattle, we considered the “farm (n = 97)” (bf) and the “individual cattle identification number (n = 412)” (bc) as random effects in the CLMM [24] based on the assumption of normally distributed random effects (see formula 2). The odds for the clean, low, medium, and high contamination of the slaughtered cattle were modelled by the CLMM based on formula 2, where θj describes the flexible threshold for category j, xi specifies the fixed meta factors, i is the index of all observations, and j = 1, …, J is the index of the response categories (J = 4; clean, low, medium, and high contamination). In the first step, we calculated a full initial CLMM model (i.e., considered all fixed meta factors, xi, in Table 1) by subsequent application of a drop function with a Chi-squared (likelihood ratio) test to include the significant xi fixed factors in the final model [25]. The correlation between the final predictor variables in the model were assessed with Carmer’s V (φc) (φc: 0 = no correlation/association; 1 = high correlation/association). In the second step, we calculated the random effects, bf and bc, via conditional modes with 95% confidence intervals based on the conditional variance and the expected probability (including 95% confidence intervals) for the cleanliness level using the ggpredict function (see Figure S2). logit(P(Yi ≤ j)) = θj − ß1xi,1 − … − ßn,xi,n − uibf,c;(3) bf,r~N(0, σ2f) r =1…97; bc,m~N(0, σ2c) m = 1…412(4)In both models, the significance level was set to p < 0.05. The models were implemented using the packages ‘ordinal’, ‘ggeffects’, ‘lme4’, ‘car’, ‘caret’, ‘lattice’, ‘ROCR’, ‘pROC’, and ‘LMERConvenienceFunctions’ in the R (Version 4.0.5) statistical computing environment.3. ResultsData were collected from 412 slaughtered cattle representing 97 cattle farms. The lameness prevalence in the study presented here was 0.73% (n = 3) of all slaughter cattle (i.e., including slight (score 2) and moderate (score 3) lameness). No cattle were identified with severe lameness (scores 4 and 5). Thus, the majority of the analyzed slaughter cattle had no lameness (99.27%; score 1 (n = 409)). The frequency of deep abomasal ulcers was low (i.e., 7.78% (n = 32; type 2) and 0.24% (n = 1; type 3), respectively) compared to cattle free of lesions (47.81% (n = 197; type 0)) and cattle with superficial ulcers of the abomasal mucosa (44.17% (n = 182; type 1)). Types 1, 2, and 3 were cumulated into one group, and abomasal lesions were categorized as a binary variable for statistical analysis (no: free of abomasal mucosa lesions and ulcers (type 0); yes: not free of abomasal mucosa lesions and ulcers (types 1, 2, and 3)). Overall, the prevalence of lesions of the abomasal mucosa was 52.43%.In total 11.40% of the cattle were assigned to cleanliness level 1 (clean), 60.92% to level 2 (low contamination), 22.83% to level 3 (medium contamination), and 4.85% to level 4 (high contamination). No slaughtered cattle were assigned to the level 5 (slaughtering would have been prohibited due to hygienic deficiencies; see Table 1). In total, the contamination prevalence of the slaughter cattle was 88.59%.Figure 1 illustrates the scorings of the slaughter cattle regarding “lesions of the abomasal mucosa” and “cleanliness” at the slaughterhouse. The average transportation and waiting duration of the slaughter cattle was 7.03 h (min: 0.05 h; max: 29.43 h). The average live days was 393 days (min: 60 days, max: 1981 days), the mean carcass weight was 262 kg (min: 58 kg, max: 762 kg), and the average daily weight gain of slaughter cattle was 0.66 kg (min: 0.16 kg, max: 1.37 kg). These animal-related data were recategorized for statistical analysis as low, medium, or high. Table 1 shows the frequency distribution of all collected data of the slaughter cattle.After the backwards factor-selection approach and after excluding “sex” as a fixed factor due to collinearity with slaughter weight, the final BGLMM model included a total of three fixed factors: “housing type”, “carcass classification”, and “slaughter weight”. The frequency of these three factors in relation to the binary abomasal classification is shown in Figure 2a. Significant associations were determined between abomasal lesions and “housing type” and “slaughter weight”, respectively. No association was identified between abomasal lesion and “carcass classification” (Table 2). The odds for occurrence of abomasal lesions were 28 times higher for cattle housed with slatted flooring compared to a tethered housing system. The odds for abomasal mucosa lesions compared to no occurrence of abomasal mucosa lesions were 3.69 times higher for animals with low slaughter weight class (<150 kg) compared to cattle with a high slaughter weight (>300 kg). The mean accuracy of the model in cross validation was 68.0% (see ROC analysis, Figure S1). The latter model result indicated that 32.0% of the abomasal lesion cases would be misclassified by using these three predictor factors.After applying the dropping function, the final CLMM included three fixed factors, i.e., “housing type”, “sex”, and “breed”. No correlation between these three factors was identified (i.e., φc: 0.47 between “housing type” and “sex”; φc = 0.38 “housing type” vs. “breed”; and φc = 0.30 “sex” vs. “breed”). All three factors were identified as significant influencing factors on the cleanliness level. The relative frequency of these three factors stratified by “cleanliness level” is shown in Figure 2b. Considering the “housing type” of the cattle, “deep litter boxes” (odds: 40.82) and “calves in group housing on straw bedding” (odds: 7.04) had significantly higher odds of contamination compared to cattle kept in a “tethered housing system”. Table 3 shows that the odds for contamination were significantly lower for “Brown Swiss” (odds: 0.26) and “Holstein-Friesian” (odds: 0.14) compared to “Simmental” cattle. The odds for contamination were significantly higher for “cows” (odds: 32.95) and “heifers” (odds: 4.40) and lower for “female calves” (odds: 0.09) and “male calves” (odds: 0.22) compared to “bulls”. The estimated variance components for the random effects, farm (σ2f) and cow (σ2c), were 0.1 and 0.0, respectively (Figure S2 shows the associated random effect of the single farms). The probability for a specific contamination level stratified by “housing type”, “sex”, and “breed” is shown in Figure S3. The expected probability for clean cattle was more than 50% for both female calves and male calves across all housing types and breeds, whereas the probability was less than 1% for “high” contaminations.4. DiscussionAnimal welfare is a growing demand of the modern consumer. People ask for food produced from animals kept under animal-friendly and fair conditions, understood as the treatment, transport, and slaughter of animals without causing pain, suffering, lesions or severe fear of the animals [24]. To improve the situation for the animals, two approaches are common today. The first is to evaluate the environment and husbandry, and the second is to focus on the animal itself by observing the impact of the environment on behavior, wellbeing, and health. Animal welfare indicators describe how animals cope with their production and husbandry system. This study describes three animal welfare indicators in slaughter cattle. One official veterinarian scored all animals regarding these factors (lameness, contamination level, and abomasal lesions); therefore, we assume no interobserver bias in the present study. Data were collected at one abattoir in Austria. The animals came from 97 holdings; therefore, a huge variety of farm individual factors had to be taken into account. Animal and farm-specific data were collected from transport certificates, the national cattle database, and by interviewing the transporter of the individual animals. Animal and farm-specific data, including findings from official meat inspection and data from the independent meat classification company, were defined as possible risk factors for the three predefined animal welfare indicators. The study results show a significant association between contamination level and husbandry system, sex, and breed, as well as significant associations between abomasal lesions and housing type and slaughter weight.The first part of the study describes the prevalence of lameness, contamination, and abomasal lesions in slaughter cattle. A total lameness prevalence of 0.7% was observed, which can be compared to the results of studies by Fjeldaas et al. [28] describing 1.1% lame cattle in suckler herds, which is lower than the 2.0% documented by the University of Nebraska for US feedlot cattle [29]. The observers’ expertise in scoring lame cattle was high [24,30], so we assume that even slightly lame animals could be detected. This point is crucial, as farmers often underestimate the prevalence of lameness in their herds [31]. Slaughter animals in this study were mainly young bulls and calves, a group of animals that is far less affected by lameness compared to dairy or cull cows [32,33]. An on-farm lameness scoring before loading would be helpful to determine the prevalence in a more accurate way; however, highly lame animals would not arrive at the slaughterhouse because of legal restrictions concerning transport of these animals. Contamination of the slaughter animals’ bodies was scored by using a five-level system. The results for cleanliness are in line with those of another recent study from Austria [17] and a study from Serbia [34] and show that slaughter animals are often kept in a dirty and therefore unhygienic environment. This indicates strong deficits in animal husbandry and management on farms [19]. The overall prevalence of abomasal lesions was 52.4%, whereas 60.3% of the bulls, 43.1% of female calves, and 55.6% of male calves had abomasal lesions in the study at hand. Similar results were reported by Hund et al. [11], although higher prevalences were reported, ranging from 70.0 to 93.0% [16]. In total, 26.8% of slaughtered heifers in this study showed abomasal lesions, almost the same percentage as that reported by Jensen et al. [35] of 24.9%. Transport and waiting time from loading on the farm to stunning was calculated. The time of loading has to be documented by the farmer who accompanies the cattle to their destination. The time of stunning was documented by the observer of this study (JB). The average transportation and waiting duration was 7.03 h (min: 0.05 h; max: 29.43 h). The average transport and waiting time for slaughter cattle transported by the farmer himself was 2.14 h, compared to 7.45 h for cattle transported by an external company. The duration of slaughter animal transport within Austria is limited to 8 h by national regulations [36]. The total transport time could not be determined because some of the animals had been unloaded at collection points during their journey. Some of the animals were unloaded at the slaughterhouse during the night before slaughter. Nevertheless, the total transport and waiting time of slaughter animals from loading to stunning as reported in this paper is long; however, in this study, no association was determined with respect to the considered welfare indicators. Slaughter cattle therefore have to cope with stress, new environments, and unknown situations during that time. Preferably, slaughter cattle should be transported in small groups by farmers to avoid or mitigate negative experiences.The second part of the study, i.e., model results on the occurrence of alterations of the abomasal mucosa, showed that cattle kept on slatted flooring had a significantly higher chance of developing ulcers compared to cattle kept in tie stalls. This fact might show that cattle on slatted flooring have limited space to sidestep in their boxes and develop more stress, even though daily weight gain and production data are not affected. Slaughter cattle with lower carcass weight showed significantly more alterations of abomasal lesions. Therefore, we assume that either cattle with abomasal ulcers are slaughtered earlier when they weigh less due to lower production performance or abomasal ulcers recover during later fattening periods. Furthermore, cattle housed on deep litter flooring have the highest chance of being the dirtiest. These results are in line with those of a study from Italy comparing welfare and cleanliness of finishing bulls [37]. The reason for this finding might be that most deep-litter flooring systems are managed poorly, i.e., are not well cleaned. Sex also turned out to have a significant influence on cleanliness. This result was biased by the fact that mainly bulls and calves were observed in this study, with only a few heifers and cows. Besides that, bulls were mainly kept on slatted flooring, whereas heifers were kept on deep-litter flooring. The factor of breed had a significant influence on cleanliness as well, showing that Brown Swiss and Holstein-Friesian had a higher chance of being cleaner. This effect was mainly influenced by the lower age of Brown Swiss and Holstein-Friesian calves compared to Simmental.5. ConclusionsIn contrast to dirtiness and the prevalence of abomasal disorders, the determined lameness prevalence was very low. The husbandry of cattle was identified as a significant influencing factor for both the dirtiness and occurrence of abomasal disorders of slaughter cattle. Gathering data on animal welfare indicators at slaughterhouses is useful to determine potential influencing factors on the side of production.

animals : an open access journal from mdpi

10.3390/ani12040416

PMC8868451

Meat adulteration and fraud encompasses the deliberate fraudulent addition or substitution of proteins of animal or plant origin in edible products primarily for economic gain. The mitochondrial 16S ribosomal (rRNA) gene was used to identify species that are present in pure and processed meat samples. The meat samples were sequenced using an Illumina sequencing platform, and bioinformatics analysis was carried out for species identification. The results indicated that pork was the major contaminant in most of the meat samples. The bioinformatics pipeline demonstrated its specificity through identification of species specific and quantification of the contamination levels across all samples. Food business operators and regulatory sectors can validate this method for food fraud checks and manage any form of mislabeling in the animal or plant protein food ecosystem.

Processed meat is a target in meat adulteration for economic gain. This study demonstrates a molecular and bioinformatics diagnostic pipeline, utilizing the mitochondrial 16S ribosomal RNA (rRNA) gene, to determine processed meat product mislabeling through Next-Generation Sequencing. Nine pure meat samples were collected and artificially mixed at different ratios to verify the specificity and sensitivity of the pipeline. Processed meat products (n = 155), namely, minced meat, biltong, burger patties, and sausages, were collected across South Africa. Sequencing was performed using the Illumina MiSeq sequencing platform. Each sample had paired-end reads with a length of ±300 bp. Quality control and filtering was performed using BBDuk (version 37.90a). Each sample had an average of 134,000 reads aligned to the mitochondrial genomes using BBMap v37.90. All species in the artificial DNA mixtures were detected. Processed meat samples had reads that mapped to the Bos (90% and above) genus, with traces of reads mapping to Sus and Ovis (2–5%) genus. Sausage samples showed the highest level of contamination with 46% of the samples having mixtures of beef, pork, or mutton in one sample. This method can be used to authenticate meat products, investigate, and manage any form of mislabeling.

1. IntroductionMeat species identification is a subject that has received special attention worldwide, mainly due to the increased incidence of fraudulent practices that have been reported [1,2,3,4]. These reports have led to consumers demanding the accurate identification and labelling of meat products [5]. Incidents of meat species substitution include contamination of a product with a cheaper-priced protein. For instance, replacing Grade A beef with rejected horse meat, replacing mutton with a lower grade of beef, or replacing mutton with pork [3,4]. The addition of plant proteins, such as grain by-products or soyabeans, to meat products like beef patties and sausages has also been reported [6]. Meat species substitution is common in processed meat products that are difficult to accurately identify morphologically once processed into value-added products. For instance, pork is intentionally added to beef products to reduce production costs [7]. Once the two different meats are minced or ground it is difficult to identify them using the naked eye. Meat adulteration predominately occurs in ground meat products [8].Consumers have a right to purchase meat products that are correctly labelled for reasons of health (allergies), religious belief, individual preference and ethics [9]. Therefore, there is need for the accurate identification of meat species in processed meat products. Food labelling regulations require that ingredients in food products are accurately declared to consumers [1]. The governing organizations in South Africa have issued new legislation to encourage clarity and the accurate explanation of food products, in response to consumer demand. These are the controls linked to Advertising and Labelling of Foodstuffs (R. 146/2010) comprised of a compulsory ingredient list on food labels [10] and the Consumer Protection Act (R. 147/2009), which prevents unfair marketing and business practices and provides an improved standard of consumer information [11]. The food regulation standard in Europe requires that meat products should be accurately labeled with information that includes the composition and percentage of ingredients included in the products [12].Techniques for meat species identification need to be reliable, rapid and cheap enough for routine applications. In the past, the identification of meat species has been conducted using protein-based methods, which entail different immunological, chromatographic, and electrophoretic methods [4,13]. However, the disadvantages of protein-based methods are that proteins are denatured by heat, salt and pressure, making protein-based methods unsuitable for the identification of species in seasoned, cured or dried meat, and meat patties [14,15]. Protein-based methods are also inaccurate in identifying species that have a close phylogenetic relationship due to cross-reactivity, for instance in poultry species [1,13,16,17] deoxyribonucleic acid- (DNA) based methods are now preferred in place of protein-based methods, because DNA is more stable during heating and less likely to be disturbed during food processing [13].Mitochondrial DNA (mtDNA) is commonly used in the identification of meat species, as it can be extracted, undamaged, from cooked and processed meat products [18,19]. Mitochondrial DNA is commonly used in species identification, since mtDNA occurs in multiple copies (an average of 1000 per cell), can withstand heat, salt, and pressure and can discriminate closely related species due to its high rate of evolution [20]. Several mtDNA genes have been used in meat species identification, such as cytochrome b, cytochrome c oxidase subunit 1 (COI), NADH dehydrogenase subunit 2 and 5 (ND2 and ND5), ATPase 6 and 8, mitochondrial 16S, and 12S ribosomal RNA (rRNA) genes [21]. Mitochondrial DNA-based methods that have been used for species identification in the past are polymerase chain reaction (PCR) [22], PCR-restriction fragment length polymorphism (PCR-RFLP) [23], species-specific PCR [24,25], DNA hybridization [17], multiplex PCR [26], and real-time PCR [17,27]. The limitations of previous DNA-based methods are that species-specific PCR methods were used and, therefore, these studies targeted specific species as opposed to having a universal method that targets any species. Species-specific methods are advantageous when species in a sample are known, however, a universal method is a better approach for investigating multiple and unsuspected contaminations in meat products. Recently, there have been advances in DNA-based methods, specifically in DNA sequencing technologies. Next-Generation Sequencing (NGS) is a method that can generate sequenced data from degraded DNA, and one that can produce large amounts of sequenced data at a low cost and with minimum errors [28]. Furthermore, essentially no prior information of species is needed, making NGS technology a non-species-specific method [28]. With the NGS method, mtDNA barcoding genes can be sequenced using universal primers, without knowledge of which species are present and without targeting specific species. This, then, enables the identification of every species present in a sample, as against merely the suspected/hypothesised ones [17]. However, the database used in the identification of species needs to be comprehensive, such that it contains a large number of species to enable accurate identification.Meat species contamination has been reported in the South African meat sector [1,2,29]. A routine universal diagnostic method that can be used by laboratories needs to be developed, as the methods used to date, in South Africa, have been species-specific. Meat producers can use this method to authenticate their products and gain consumers’ confidence in the products they will purchase. Mitochondrion carries extra-chromosomal genetic material and contains high copy numbers as compared to single copy nuclear genes. Therefore, mitochondrial DNA is the preferred analytical tool in forensic, molecular, and zoological experiments. The objective of this study was, therefore, to develop a universal and robust diagnostic molecular and bioinformatics pipeline that can utilize the mitochondrial 16S ribosomal RNA (rRNA) barcoding gene to identify processed meat product mislabeling/contamination using NGS. Universal mitochondrial 16S rRNA primers will be used in this study to identify different species, including those in mixed samples. Meat suppliers can possibly implement the current method to authenticate their products, and the food industry may also use this method to reveal any form of mislabeling that may be present.2. Materials and Methods2.1. Collection of Pure Meat Samples to Confirm the Reliability of the Species Identification PipelinePure meat samples were collected as controls to verify the use of the 16S rRNA gene in the molecular and bioinformatics pipeline developed. This was done primarily to test the specificity, sensitivity, and ability of the pipeline to be used as a diagnostic method. Nine unprocessed pure meat samples from nine different species were collected from a local butchery in Pretoria (South Africa), transported in an icebox and stored at −20 °C. The pure meat samples were placed in separate plastics upon collection, transportation, and storage, to avoid any unintentional cross-contamination. These species were pig, cattle, sheep, chicken, turkey, goat, ostrich, duck, and kangaroo.2.2. The Collection of Processed Meat Samples for Species IdentificationMeat products were randomly collected from processing plants and retail outlets in the Gauteng and Free State provinces in South Africa for the species identification test. A total of 155 samples from the meat value chain were collected and analyzed. Four different categories of processed meat products were collected for analysis, namely, minced meat (49), burger patties (35), biltong (28), and raw sausages (43). Some samples included information on which species they were produced from, and, of these, 22 were beef mince, 20 were beef patties, 17 were beef biltong, and 21 were beef sausages. All samples were transported in an ice box and stored at −20 °C.2.3. DNA ExtractionGenomic DNA from the pure meat samples used for the verification test was ex-tracted manually from 40 mg of each meat sample. A Macherey–Nagel NucleoMag Tissue kit for DNA purification from cells and tissue (Macherey–Nagel, Düren, Germany) was used for DNA extraction according to the Genomic DNA from Tissue user manual. The pure DNA was stored at −20 °C while awaiting further analysis. Thirteen two-species DNA mixtures of known species and composition were artificially mixed (Table 1). Two ratios were used for the DNA mixtures, 1:1 (50%:50%) and 0.9:0.1 (90%:10%). The artificially mixed samples were used to test the specificity of the 16S universal primers, by confirming the origin of the known species in the 1:1 ratio mixture. The ratio of 0.9:0.1 was used to test the sensitivity of the pipeline, using, as a metric thereof, the smallest amount of DNA that the pipeline could correctly identify at an affordable cost. Each mixed ration had three replicates. The concentration of the DNA of each pure meat sample used was normalized to two different concentrations, 25 ng/µL and 5ng/µL prior to running of the PCR. The differences in DNA concentration are explained in the PCR step. Genomic DNA for the species identification of the samples collected from the meat value chain was extracted from 300 mg of each processed meat sample, using a Hamilton Microlab Star automated liquid handler (Hamilton Inc., Reno, NV, USA). A Macherey–Nagel NucleoMag Tissue kit for DNA purification from cells and tissue (Macherey–Nagel, Düren, Germany) was used for DNA extraction according to the Genomic DNA from Tissue user manual. The DNA concentration of the meat value chain samples was between 28–467 ng/µL prior to PCR testing. The quantification of DNA for all samples was checked using the Qubit® fluorescent dye method, and gel electrophoresis was used to assess the quality of the starting material. A ratio of A260/A280 was used to access the purity of all extracted DNA.2.4. PCR Amplification of the Mitochondrial 16S rRNA GenePolymerase chain reaction (PCR) for the mitochondrial 16S rRNA gene was performed using universal mammalian primers designed by [17] and tailed with Nextera adapters (Table 2). Thermal cycling was performed in a Labnet MultigeneTM Gradient Thermal Cycler (Woodridge, IL, USA) at a final volume of 50 µL. All the ratio mixtures had a normalized DNA concentration of 50 ng/µL in the PCR run. This was done so that if one ratio mixture did not identify the contained species, it would not be due to the DNA having a lower concentration. The PCR for the 50%:50% (1:1) ratio mixture contained 25 µL of 2× Hot start PCR mastermix, 5 µL of each forward and reverse primer (1 mM final concentration), 13 µL RNase-free water and 1 µL of 25 ng/µL DNA template of each species. This brought the total amount of DNA template for the 50%:50% ratio mixture to 2 µL and the concentration to 50 ng/µL. The PCR for the 90%:10% (0.9:0.1) ratio mixture contained 25 µL Kapa HiFi Hotstart Readymix (Roche, NY, USA), 5 µL of each forward and reverse primer (1 mM final concentration), 12.2 µL RNase-free water, 1.8 µL of the 25 ng/µL DNA template for the species with a ratio of 90% and 1 µL of 5 ng/µL DNA template for the species with a ratio of 10%. This brought the total amount of DNA template for the 90%:10% ratio mixture to 2.8 µL and the concentration to 50 ng/µL. Similar to the DNA extraction process, sterile tips and PCR tubes were not reused and the pipettes and work-bench area were disinfected with 70% ethanol between analyses.The PCR for the samples from the meat value chain contained 25 µL of Kapa HiFi Hotstart Readymix (Roche, NY, USA), 5 µL of each forward and reverse primer (1 mM final concentration), 13 µL RNase-free water, and 2 µL of DNA template. The PCR conditions for all samples were as follows: denaturation at 95 °C for 3 min, followed by 30 cycles of 90 °C for 20 s, 65 °C for 30 s, 72 °C for 30 s, and finalization at 72 °C for 5 min. The PCR products for the mitochondrial 16S rRNA gene were 186 bp in length. The PCR products were viewed in 2% agarose gels in 1× tris-acetate-EDTA (TAE) buffer at 90V for 45 min. The amplified products were visualized under ultra-violet light in a transilluminator. Purification of PCR products was performed using a Qiagen MiniElute® PCR purification kit (Qiagen, Germany) according to the manufacturer’s protocol. Quantification of the purified samples was done using the Qubit® fluorescent dye method. A ratio of A260/A280 was used to access the purity of all extracted DNA. The purified products were stored at 4 °C prior to sequencing.2.5. Library Preparation and Illumina MiSeq SequencingPrior to sequencing, library preparation was performed using the 16S Meta-genomics Sequencing Library Preparation kit, according to the manufacturer’s protocol (Illumina, Inc, San Diego, CA, USA). Quality control of the sample library and quantification of the DNA library templates was performed. Quantification of DNA was done using Qubit® fluorescent dye method. The library size distribution was checked using a High Sensitivity DNA chip. Thereafter, the indexed libraries were normalized, pooled and loaded onto an Illumina MiSeq reagent cartridge using MiSeq reagent kit v3 and 600 cycles. The paired end 2 × 300 bp sequencing was run on an Illumina MiSeq sequencer at 0.2 × coverage at the Biotechnology Platform, Agricultural Research Council, Onderstepoort, South Africa. The DNA from pure meat samples were each sequenced individually prior to artificially mixing the DNA, to confirm the origin of each meat type.2.6. Bioinformatics and Data AnalysesPrior to species identification, quality control, adapter removal, decontamination, and error correction of the raw sequence data was done using BBDuk (version 37.90; https://jgi.doe.gov/data-and-tools/bbtools/bb-tools-user-guide/bbduk-guide/, accessed on 20 July 2020). All available mitochondrial genomes (10,788) were downloaded from the NCBI RefSeq database (https://ftp.ncbi.nlm.nih.gov/refseq/release/mitochondrion, accessed on 20 July 2020) [30]. Filtered reads were aligned to the complete mitochondrial genomes using BBMap v37.90 [31], (https://www.osti.gov/biblio/1241166-bbmap-fast-accurate-splice-aware-aligner, accessed on 20 July 2020) for species identification and the average fold coverage across a mitochondrial reference genome was used for further analysis. The average fold of each sample was exported into Microsoft Excel to calculate the percentage of the number of reads that aligned to a reference, over how much of the reference was covered in a sample. The percentage average fold was used to determine samples that were contaminated and uncontaminated.Statistical analysis was initially performed on the pure and artificially mixed samples using R software, to determine whether our pipeline would work in a controlled environment. The percentage deviation from the expected composition within the pure (100%) and artificially mixed (50%:50% and 90%:10%) samples was also determined, and these values were used to calculate the mean (using absolute values), median, standard deviation, and variance. Bar plots, plotting the percentage composition by species present in all samples, were constructed, showing samples that were contaminated and uncontaminated. A chi-square proportion test was used to determine whether there was a significant correlation between two categorical variables, i.e., contaminated and uncontaminated meat samples. A p-value was determined as a result of the number of contaminated versus uncontaminated. Associations between contaminated versus uncontaminated meat were considered statistically significant only for p-values ≤ 0.05. Cramer’s V test, which measures how strongly two categorical fields are associated, was also performed. The confidence interval (CI) was set at 95% and the number of samples observed (nobs) was also determined. The workflow of the molecular and bioinformatics pipeline is shown in Figure 1.3. Results3.1. Identification of Meat Species Using Pure DNA from Known Meat TypesPaired-end reads, with a length of 300 bp, were sequenced using the Miseq sequencer. Each sample had an average of 156,863 reads, before quality filtering, and 134,230 reads after quality filtering. Nine pure meat samples with two replicates each were analysed, for a total of 18 analysed pure samples. The reads obtained mapped to the corresponding pure meat species, with a similarity of 98% and above for all meat types (Figure 2). Besides identifying the expected genera, traces of other meat species were observed (Table S1). The beef, mutton, and pork meat samples had reads with an average fold of 99% for the Bos, Ovis, and Sus genus, respectively. The chevon, chicken, and duck meat samples had reads with an average fold of 98% for the Capra, Gallus, and Anas genus, respectively. The turkey and kangaroo meat samples had reads with an average fold of 99% for the Meleagris, Struthio, and Macropus genus, respectively. One of the ostrich samples was contaminated with beef, as it showed a proportion with reads that mapped to the Bos genus (Figure 1). Based on these results from the controls, samples whose highest percentage average fold was less than 98% were considered contaminated.3.2. Identification of Meat Species Using Pure DNA from Known Meat Types Artificially Mixed at a Ratio of 1:1The pipeline identified all the meat types whose DNA were in the 1:1 ratio mixture. However, some positive and/or negative deviations were observed from the expected 50:50 percentages (Table S2). The beef (50%) and kangaroo (50%) DNA mixtures had reads with an average fold of 73% and 25% for the Bos (cattle) and Macropus (kangaroo) genus, respectively (Figure 3). The chevon (50%) and mutton (50%) DNA mixtures had reads with an average fold of 35% and 63% for the Capra (goat) and Ovis (sheep) genus, respectively. The chicken (50%) and turkey (50%) DNA mixtures had reads with an average fold of 48% and 50% for the Gallus (chicken) and Meleagris (turkey) genus, respectively.The duck (50%) and ostrich (50%) DNA mixtures had reads with an average fold of 69% and 29% for the Anas (duck) and Struthio (ostrich) genus, respectively. The pork (50%) and beef (50%) DNA mixtures had reads with an average fold of 51% and 48% for the Sus and Bos genus, respectively. The greatest deviations for the 1:1 (50% of each species) ratio were between beef (Bos): kangaroo (Macropus) where beef meat had an overestimation of 23% and kangaroo meat had an underestimation of 24%. The lowest deviations were between pork (Sus): beef (Bos), where pork had an overestimation of 1.5% and beef had an underestimation of 1.9% (Table S2).3.3. Identification of Pure DNA from Known Meat Types Artificially Mixed at a Ratio of 9:1The pipeline showed that all the meat types whose DNAs were in the 9:1 ratio mixture were identified. Some positive and/or negative deviations were observed from the expected 90:10 percentage ratio (Table S3). The beef (90%) and pork (10%) DNA mixtures had reads with an average fold of 92% and 6% for the Bos (cattle) and Sus (pig) genus, respectively (Figure 3). The chicken (90%) and duck (10%) DNA mixtures, the reads had an average fold of 82% and 17% for the Gallus (chicken) and Anas (duck) genus, respectively. The duck (90%) and chicken (10%) DNA mixtures had reads with an average fold of 96% and 3% for the Anas (duck) and Gallus (chicken) genus, respectively. The duck (90%) and ostrich (10%) DNA mixtures had reads with an average fold of 86% and 13% for the Anas (duck) and Struthio (ostrich) genus, respectively. The goat (90%) and sheep (10%) DNA mixtures had reads with an average fold of 81% and 17% for the Capra (goat) and Ovis (sheep) genus, respectively. The ostrich (90%) and duck (10%) DNA mixtures had reads with an average fold of 81% and 18% for the Struthio (ostrich) and Anas (duck) genus, respectively. The pork (90%) and beef (10%) DNA mixtures had reads with an average fold of 91% and 8% for the Sus (pig) and Bos (cattle) genus, respectively. The mutton (90%) and chevon (10%) DNA mixtures had reads with an average fold of 92% and 6% for the Ovis (sheep) and Capra (goat) genus, respectively. The lowest deviations for the 9:1 (90% and 10%) of each species ratio were between pork (Sus): beef (Bos). Pork meat had an overestimation of 1.2% and beef meat had an underestimation of 1.5%. The greatest deviations were between ostrich (Struthio): duck (Anas). Ostrich meat had an underestimation of 8.8% and duck meat had an overestimation of 8.5% (Table S3).3.4. Percentage Deviation from Expected Percentages in the Pure and Artificially Mixed SamplesThe percentage deviation (expected–observed) was determined and used to calculate the descriptive statistics (mean, median, standard deviation, variance, minimum value, and maximum value) of the pure and artificially mixed samples (Table 3). Measures of central tendency were described by the mean and median values, while measures of variability were described by standard deviation, variance, and minimum and maximum values. The range within the artificially mixed samples was 23.59% and 3.19% for the pure samples. The higher spread of values in the artificially mixed samples lead to a higher mean value, and that, in turn, resulted in a higher standard deviation and variance that was further from zero. In contrast, the smaller spread of values within the pure samples lead to a lower mean value that resulted in a lower standard and variance that was closer to zero.3.5. Identification of Species and Species Contamination of Processed Meat Collected from Processing Plants and Retail Outlets3.5.1. BiltongThe pipeline demonstrated that all the 11 biltong samples that had not specified from which species they were from were uncontaminated and were essentially beef (Bos genus) (Figure 4). The biltong samples, however, contained trace contaminants of other species (Table S4).The pipeline also demonstrated that 14 out of the 17 samples labelled as beef biltong were uncontaminated and were also essentially from beef (Bos genus) (Figure 5). The contaminated samples, descending from most contaminated, were predominantly composed as follows: Sample 16: Bos (cattle) (57.5) and Sus (pig) (37.4); Sample S: Bos (cattle) (93.6) and Ovis (sheep) (4.9); and Sample 117: Bos (cattle) (97.9) and Bubalus (Buffalo) (0.7). The major contaminants of the labelled beef biltong products were pork (Sus) and mutton (Ovis). Sample 16 contained beef (Bos) (57.5) but was contaminated with pork (Sus) (37.4) and mutton (Ovis) (4.6) (Table S5).3.5.2. MinceThe pipeline demonstrated that 23 out of 27 mince samples that had not specified from which species they originated were uncontaminated and were essentially from beef (Bos genus) (Figure 6). The five contaminated samples, descending from the most contaminated were predominantly composed as follows: Sample 65: Bos (cattle) (83,2) and Ovis (sheep) (16.4); Sample 78: Bos (cattle) (95,3) and Sus (pig) (3.8); Sample 4: Sus (pig) (97.7) and Bos (cattle) (2.2); and Sample 34: Bos (cattle) (97.7) and Ovis (sheep) (1,8). Sample 4 was evidently pork mince (predominantly Sus genus) contaminated with beef (Bos genus), while the rest were beef (Bos genus) mince contaminated with either pork (Sus genus) or mutton (Ovis genus) (Table S6).The pipeline also demonstrated that 20 out of the 22 samples labelled as beef mince were uncontaminated and were essentially from beef (Bos genus) (Figure 7). The contaminated descending from the most contaminated were predominantly composed as follows: Sample 183: Bos (cattle) (93.3) and Sus (pig) (6.1); and Sample 17: Bos (cattle) (97.0) and Ovis (sheep) (2.0). The major contaminants of the labelled beef mince products were pork (Sus) and mutton (Ovis). Two samples, S158 and S99 had traces (0.1% and 0.5%, respectively) of the Homo (human) genus (Table S7).3.5.3. PattiesThe pipeline demonstrated that 13 out of the 15 patty samples that had not specified which species they are from were uncontaminated and were essentially from beef (Bos genus) (Figure 8). The two contaminated samples, descending from the most contaminated, were predominantly composed as follows: Sample 15: Sus (pig) (59.6) and Bos (cattle) (40.0), and Sample 48: Bos (cattle) (91.8) and Sus (pig) (7.7). Sample 15 was a pork patty (Sus genus) contaminated with beef (Bos genus), while sample 48 was a beef (Bos genus) patty contaminated with pork (Sus genus) (Table S8).The pipeline also demonstrated that 13 out of 18 samples labelled as beef patty were uncontaminated and were essentially from beef (Bos genus) (Figure 9). The contaminated samples descending from the most contaminated were predominantly composed as follows: Sample 122: Bos (cattle) (64.5), Ovis (sheep) (34.3); Sample 112: Bos (cattle) (81.7) and Ovis (sheep) (17.7); Sample 179: Bos (cattle) (93.4) and Ovis (sheep) (6.2); Sample 113: Bos (cattle) (93.4) and Ovis (sheep) (5.6); and Sample 136: Bos (cattle) (94.9) and Sus (pig) (4.4). The major contaminants of the labelled beef patty products were mutton (Ovis) and pork (Sus). Sample 122 contained beef (Bos) (64.5) but was contaminated with mutton (Ovis) (34.3) (Table S9).3.5.4. SausagesThe pipeline demonstrated that 14 out of the 21 sausage samples that had not specified which species they were from were uncontaminated and were essentially from beef (Bos genus) (Figure 10).The seven contaminated samples descending from the most contaminated were predominantly composed as follows: Sample 15: Bos (cattle) (37.1) Sus (pig) (38.5) and Ovis (sheep) (23,7); Sample 83: Bos (cattle) (74.7) and Sus (pig) (24.7); Sample 26: Bos (cattle) (91.0), Ovis (sheep) (5.2) and Sus (pig) (3.0); Sample 159: Ovis (sheep) (91.4) and Bos (cattle) (7.0); Sample 5: Sus (pig) (94,8) and Bos (cattle) (5,2); Sample 79: Bos (cattle) (95.9) and Ovis (sheep) (3.6); Sample 4: Sus (pig) (97.4) and Bos (cattle) (2.7); and Sample 68: Bos (cattle) (97.9), Bubalus (buffalo) (0.8) and Rupicapra (goat antelope) (0.6). Sample 15 was a mixed sausage made up of a substantial amount of beef (Bos), pork (Sus) and mutton (Ovis). Sample 159 was a mutton sausage contaminated with beef (Bos genus), while sample 4 and 5 were pork (Sus genus) sausages contaminated with beef (Bos genus (Table S10).The pipeline also demonstrated that 7 out of 21 (33%) samples labelled as beef sausage were uncontaminated and were essentially from beef (Bos genus) (Figure 11). The major contaminants of labelled beef sausage products were mutton (Ovis) and pork (Sus). Two samples, Sample 15: Bos (cattle) (18.1) and Sus (pig) (78.1) and Sample 135: Bos (cattle) (38.8) and Sus (pig) (60.1) can be considered as a mislabeled sample because the Bos (beef) genus represented a smaller percentage than the predominant Sus (pork) genus (Table S11).3.6. Proportion Test of Two Categories (Contaminated vs. Not Contaminated) Using a Chi-Square TestP-values of p = 1.62 × 10−4, p = 4.24 × 10−10 and p = 1.95 × 10−9 were determined as a result of the chi-square test for the number of contaminated versus uncontaminated pure, artificially mixed and retail samples, respectively (Figure 12). There was 6% contamination in the pure samples and there was no statistically significant level of contamination. However, there was 100% and 26% contamination in the mixed and retail samples, respectively. The p values therefore indicate a significant level of contamination in the artificially mixed and retail samples. The overall p value for all three sample groups was p = 1.85 × 10−18. Cramer’s V association was 0.62, confidence interval (CI) was set at 95% (0.48, 0.75) and the number of samples observed (nobs) was 209 (Figure 12).The retail samples were then statistically analysed, according to the different meat types. There was no contamination observed in the biltong samples that had not specified which species they were from and the p-value (p = 0.001) shows that there was no statistically significant level of contamination (Figure 13). There was contamination observed in the mince (15%), patties (13%), and sausage (38%) samples that had not specified which species they were from. The p-values for the mince (p = 2.56 × 10−4) and patty (p = 0.005) samples indicate that there was a statistically significant level of contamination, however, there was no statistically significant level of contamination in the sausage (p = 0.275) samples (Figure 13).There was 18%, 9%, and 29% contamination in the beef biltong (p = 0.008), beef mince (p = 1.24 × 10−4), and beef sausage samples (p = 0.050), indicating a statistically significant level of observed contamination. However, regardless of finding 28% contamination in the beef patty samples, there was no statistically significant level of contamination (p = 0.059) (Figure 13). The overall p value for all the retail samples was p = 1.02 × 10−5. Cramer’s V association was 0.48, confidence interval (CI) was set at 95% (0.21, 0.55) and the number of samples observed (nobs) were 152 (Figure 13).4. DiscussionThe main aim for this work was to develop a diagnostic pipeline for species identification in meat samples, including the identification of species in artificially mixed samples from different mammalian species. The mitochondrial 16S rRNA marker used in this study has proven, in earlier studies, to have the power to detect individual species and even distinguish between closely related species [17].A pipeline, using the mitochondrial 16S rRNA gene and NGS, was established to initially identify known meat types from pure DNA that were not mixed with any other meat types. To achieve this, pure DNA from the respective meat types was used. The overall results demonstrated the ability of our protocol to identity pure DNA that is not mixed with other meat types, as all meat types from the known pure DNA were identified and mapped to the corresponding genera. The reads obtained mapped to the corresponding genus of each pure meat sample, with a similarity of ≤98% relative abundance for all meat types and with minor deviations of less than 2% relative abundance. One ostrich sample that was contaminated with beef DNA was a result of human error in the lab.Contamination can either be intentional or unintentional. Intentional contamination occurs when deliberately adding a cheaper material to a product for economic gain. Unintentional contamination is the mistaken introduction of something into a product. This usually occurs through cross-contamination from the use of the same equipment amongst different products [3].In our experiment, the initial aim was to test the pipeline using 100% pure meat samples. There was no intention to assess contamination with other meat types. However, a contamination of less than 2% from other meat types was observed, which could be attributed to (i) trace amounts of other species having occupied abattoirs, butcheries, and retailers that slaughter, process, and sell multiple species’ meat and use the same equipment for their processing, or (ii) a lack of maintenance of sequence databases could affect the stringency or sensitivity of species identification pipelines if new information is not added to a given database.Overall, the initial verification test enabled us to determine some of the sensitivity thresholds of the pipeline. A threshold of 1% (w/w) for undeclared meat species in meat products was set by the Food Safety Authority (FSA) and Department for Environment Food and Rural Affairs (Defra) in Europe, after horse and pig DNA was identified in beef products [32]. Based on our analysis, a threshold of 2% is more practical when considering cross-contamination and database-stringency factors. However, a threshold of 2% carries implications for consumers who are mindful of their diet for religious reasons. The Jewish and Muslim communities prioritize the traceability and authenticity of the meat they consume, because they only consume meat from ritually slaughtered animals in accordance with their beliefs [33]. Not declaring the meat species composition in a product violates their rights as consumers.Having established the workflow and thresholds, the pipeline was further used to identify meat types from artificially mixed DNA at ratios of either 1:1 or 9:1. The aim of this part of the experiment was to simulate the conditions of retail market and determine whether the pipeline could identify species in mixed DNA. The 1:1 ratio simulated retailers that intentionally contaminate meat products and do not try to hide it. This type of contamination mainly occurs for economic gain, by intentionally adding a cheaper product to the primary product [1,7,29]. The pipeline demonstrated that the DNA of all meat types in the 1:1 ratio mixture were identified. However, some deviations were observed from the expected 50:50 percentages, and minute traces of species not included in some ratio mixtures. The major deviations were observed amongst the mutton: chevon, duck: ostrich and beef: kangaroo ratio mixtures. The deviation in the mutton: chevon ratio mixture may have been due to sheep and goats having similarities between their genomes, since they originate from the same family, Bovidae, and sub-family Caprinae. Previous research has demonstrated that sheep and goats evolved from the same ancestor Rupicaprids (goat antelopes) in the Pleistocene era [15,34]. This may have resulted in an over or underestimation in the mutton: chevon ratio mixture. Similarly, ducks and ostriches have similarities in their genomes as they are both Aves species. Previous comparative cytogenic work has suggested that there is a preserved sequence homology between the Z and W chromosomes in ducks and ostriches, since recombination was suppressed [35]. Furthermore, research has also demonstrated that the ostrich IgM isotype has a 66% and 63.1% sequence identity with the Cα and Cµ genes of the duck, respectively [35]. The overestimation of beef meat and underestimation of kangaroo meat in the beef: kangaroo ratio mixture may have been due to the cattle genome being sequenced more than the kangaroo genome. The sequencing of the bovine genome was initiated in 2002 [36] and has continued, with several other works since published [37,38,39,40,41]. The kangaroo genome, on the other hand, was only first sequenced in 2011 [42,43], even though the benefits of sequencing the kangaroo genome where initially discussed and published in 2003 [44]. Further research on sequencing the kangaroo genome has been published [45,46], but there is a clear indication from the number of published articles that the kangaroo genome has been sequenced less than the cattle genome. Therefore, this may have resulted in an overestimation of cattle reads in the beef: kangaroo ratio mixture. The ever-decreasing cost of sequencing, coupled with increased efforts in sequencing non-conventional livestock species such as the kangaroo, will improve on the composition and quality of databases, which will also improve on the accuracy of the methods developed to date.The 9:1 ratio simulated retailers that also intentionally contaminate meat products but try to hide it. This type of contamination occurs in situations where retailers intentionally add the fat or trimmings of certain meat species, such as pork, to improve the sensory value of some products [31]. Similar to the results of the 1:1 ratio mixtures, the pipeline also managed to identify all the meat species whose DNA were in the 9:1 ratio, with some deviations from the expected 90:10 percentages. Descriptive statistics for the pure and artificially mixed samples were analysed. The standard deviation and variance indicated how close an observed value in a dataset is to the mean. A dataset with a smaller spread of values results in values closer to the mean, yielding smaller variance and standard deviation. In contrast, if a dataset has a wider spread of values, this results in values that are further from the mean, yielding a larger variance and standard deviation [47]. The pure samples had lower standard deviations and variances that were closer to zero, meaning the values in the dataset had a smaller range and mean value. In contrast to the pure samples, the artificially mixed samples had higher standard deviation and variance values. This was brought about by a higher range within the dataset. The higher range of values may have been a result of the under and overestimation of expected percentages in the ratio mixtures. The use of mitochondrial genes in species identification has been used due to mitochondria having a mutation rate that is 10-fold higher than that of nuclear genes, allowing for the discrimination of closely related species. Mitochondria is also abundant, with thousands of copies of DNA per cell, in comparison to nuclear genes that have single copies per cell [48]. However, the presence of several copies of mitochondrial DNA in a single cell can lead to either an underestimation (−70%) or overestimation (+160%) of species’ DNA content [6]. It has also been previously reported that there is a difference in binding efficiency of the universal primers for different species, resulting in a difference in the amplification efficiency and, therefore, leading to a large degree of error in quantitative analysis [32]. The quantitative accuracy in meat species identification can be improved through the use of genes that have a single copy, the introduction of correction factors for primer amplification efficiency, designing degenerate primers, and controlling the number of amplification cycles and the amplification conditions [32].The analyses of retail meat samples showed that beef was the main species found in most samples, since their reads predominantly mapped to the Bos genus. The samples that had indicated which species they were from on their product labels predominantly mapped to the Bos (cattle) genus (90% and above), confirming their origination from beef, as stated on the labels. There was, however, evidence of contamination and mislabeling of the pork (Sus) and/or mutton (Ovis) meat observed in most samples, but no mention of the presence of any other species on the labelling. All the unspecified biltong samples predominantly mapped to the Bos genus, however, there were minor traces of te Sus (pig) and Ovis (sheep) genera of less than 2% relative abundance, hence we concluded that there was no intentional contamination in the non-specified biltong samples. The beef biltong samples showed that pork had the highest percentage of contamination, with one of the samples having as high as 36% of the reads mapping to Sus genus. The contamination seemed intentional and for economic gain, as it is not practical to mistakenly add 36% of a different meat species, especially if it is a meat type that has been previously reported to have a cheaper purchase price [1]. Bottaro et al., 2014 [49] also reported of addition of low-valued meat and fat, such as pork, to high-valued meat, such as beef, as a form of intentional meat contamination for the purposes of economic gain.Similarly, specified and unspecified mince samples showed that the Bos (cattle) genus was predominant, with few samples contaminated with pork or mutton. There were a few contaminated mince samples that were contaminated with either mutton or pork. The percentages of pork found in the mince samples were between 2–3% relative abundance. These contamination percentages of may not necessarily have occurred intentionally for economic gain, since they were found at low percentages. Rather, the contamination may have been due to cross contamination from equipment not properly cleaned in operations that process multiple species [29]. Similarly, in a South African study, Tembe, Mukaratirwa, and Zishiri, 2018 [29] concluded that the contamination of processed meat products was unintentional, and that the contamination may have been due to the use of the same equipment for processing different species. Regardless of the low contamination percentages in our study, the presence of pork has negative consequences to consumers who choose not to consume pork due to health reasons [7]. The consumption of meat with undeclared allergens may cause an allergic reaction to certain consumers [7,50,51]. Previously, in the United States, allergy prevalences of 73%, 58%, and 41% to beef, pork and chicken, respectively, were reported in 57 patients suspected of being allergic to meat [52]. The patty samples were also mainly from beef meat, with samples predominantly mapping to the Bos (cattle) genus. Pork was the main contaminate in the unspecified patty samples. One sample (Sample 48) had 7% pork in it, possibly a case of intentionally adding pork to a patty sample to improve its sensory and oxidative properties. The addition of pork meat or lard to processed meat products has been previously reported [33], to improve the sensory properties and oxidative reactions of such processed meat products. Patty Sample 85 was intentionally mixed with two meat types, pork and beef, as 59% of its read mapped to the Sus (pig) genus and 40% mapped to the Bos (cattle) genus. A case like this demonstrates intentional contamination for economic gain (unless if specified on the product label), since the production costs of pork are cheaper than those of beef [1], resulting in pork being cheaper to purchase. The major contaminant of labelled beef patty products was mutton (Ovis). This was unexpected, as mutton has a higher market price than beef. One of the reasons for substituting more expensive meat such as mutton for a cheaper meat such as beef may be due to the use of unmarketable trimmings from more expensive meat types [7]. It is possible that intentional contamination with unmarketable mutton occurred in the contaminated beef patty samples, as one of the samples had as high as 34% of its reads map to the Ovis (mutton) genus.Our results demonstrated that pork and mutton were the main species that were contaminated in the sausage samples that had reads that predominately mapped to the Bos (beef) genus. However, there were samples of mutton and pork sausages contaminated with beef at percentages of 2–7% relative abundance. Some authors [21,49] have indicated that the contamination of beef, in some meat products, may be from the addition of non-fat powdered milk to increase the overall yield and taste of the product. This may have been the case with the mutton and pork samples contaminated with beef. There was one sample (Sample 15) that was intentionally mixed with three meat types, beef, pork, and mutton, as the sample had 37%, 38%, and 23% of reads map to the Bos (cattle), Sus (pig), and Ovis (sheep) genus, respectively. This may have been a case of contamination for economic gain, because mutton and beef have a higher market price than pork [29]. The overall results of the specified and unspecified meat products indicated that pork was the main contaminate. Surowiec et al., 2011 [53] previously reported on undeclared pork and chicken in processed meat products such as burger patties and sausages and suggested that it could be from mechanically recovered meat (MRM), usually produced from pork and chicken carcasses. According to Surowiec et al., 2011 [53], the addition of MRM, which is normally found in a paste-like form, represents a source of cheap protein in processed meat products, such as deli meats, burger patties, and sausages. This practice is, however, illegal in most countries, including South Africa [1]. Furthermore, failure to declare the presence of other meat species in ingredients lists betrays consumer rights and has negative implications for consumers allergic to such contaminants and consumers whose religions observe dietary restrictions [7].Similarly, in a South African study [1], undeclared pork and mutton were found in minced meat, burger patties, and raw sausages labelled as beef, pork meat was the main undeclared meat type found in these meat samples. In another South African study on meat species’ substitution, undeclared beef, pork, and lamb were found in commercially labelled wildlife meat products [2]. More recently, another South African study revealed the presence of meat contamination in the province of KwaZulu-Natal [29]. A high proportion of beef and mutton products were contaminated with pork and chicken. Undeclared species in the above-mentioned studies and in ours reflect that there is still a presence of meat adulteration in the South African meat market. Judging from the results we have observed from the retail samples, there seems to be intentional contamination for economic gain. There is need to improve product labelling so as to indicate every species within a meat product so that consumers can make informed decisions. Some major retailers in South Africa, such as Food Lover’s Market, Pick and Pay, and Checkers, now mention the presence of multi-species on their meat product’s labels. For example, a sausage sample, today, might be labelled as 70% beef, 20% pork, and 10% Water. This type of clarity in labelling assists consumers who prefer to avoid certain species for allergenic, religious, or ethical reasons.Statistics indicate that beef has the highest gross value as compared with other meat species produced in South Africa, with an average of R 23.5 billion per annum [54]. There was a slight decrease in cattle production between 2017–2018, due to farmers in South Africa not having enough cattle to slaughter. This led to an increase in beef market prices, as herds were replenished, which in turn decreased the consumption of beef and beef products. Consumers opted for cheaper alternatives, such as chicken and pork [54]. The gross production of mutton in South Africa is an average of R 4.57 billion per annum [55]. There was, however, a decline in sheep production from 2017 due to stock theft, which led to an increase in demand and subsequent shortages in the supply of mutton [55]. This shortage in supply, coupled with high production costs, has led to high market prices of mutton in the South African meat market. Beef is more readily available, hence it is the main species processed into value-added products in South Africa. However, high production costs result in a higher purchasing price for beef. Mutton also has a higher purchasing price than the other meat types, mainly because it has higher production costs and is not readily available on the market. Chicken and pork have lower production costs in comparison with beef and mutton and this has led to them having a lower purchasing price and, therefore, being more frequently fraudulently added to higher-value products labelled as beef or mutton, for economic gain.5. ConclusionsIn conclusion, the current paper presents a universal diagnostic molecular method for the identification of meat species. The method used the mitochondrial 16S rRNA gene, which has demonstrated its variability from the results of the phylogenetic analysis in our previous study. The verification experiment identified all species present in the known DNA mixtures, proving the accuracy of the pipeline in the species identification in the processed meat samples that were collected. Meat suppliers can possibly implement the current method to authenticate their products, and the food industry may also use this method to reveal any form of mislabeling that may be present within meat productsi.

animals : an open access journal from mdpi

10.3390/ani12050604

PMC8909040

Rumen fermentation, the level of hormones and major metabolites in the blood, and tissue heat production, as well as milk composition, were studied in high-producing cows with different milk fat content. It was found that low milk fat is associated with neither a lack of acetate formation in the rumen, nor a change in the hormonal profile. The result is an increase in the availability of cows for metabolizable energy, the likelihood of ketosis, and the extension of their productive use.

In order to clarify the mechanism of the depression of milk fat formation and preserve the health of animals, the aim of the research was to study the characteristics of rumen digestion, energy metabolism, and milk composition in high-producing dairy cows with high and low levels of milk fat that are fed the same diet. Two groups of cows with normal milk fat content (3.94 ± 0.12; n = 10) and low milk fat content (2.95 ± 0.14, n = 10) contained in the same diet were identified. Gas exchange (O2 uptake and CO2 output) was studied in cows and blood samples, rumen contents (pH, NH3-N), and VFA and milk (fat, protein, and fatty acid composition) were collected and analyzed. It was determined that cows with low fat milk are more efficient at using the metabolized energy of their diets due to the tendency to have a decrease in the proportion of heat production (by 6.2 MJ; p = 0.055) and an earlier start of a positive energy balance. At the same time, the fat content in milk did not depend on the level of hormones in the blood or on the formation of acetate in the rumen. An analysis of the duration of the productive use of cows on this farm (n = 650) showed that the number of lactations was inversely correlated with the level of fat in milk (r = −0.68; p < 0.05, n = 1300). These results indicate the advantages of cows that can reduce the fat content of their milk in the first months of lactation.

1. IntroductionMaintaining the normal physiological state of animals is the key to their long-term productive use and the guarantee of obtaining high-quality milk. Only a healthy cow is able to provide high productivity. Therefore, the health of the cow and her productivity are inextricably linked. In recent years, the number of diseases of alimentary etiology has increased significantly. The main breakdown occurs in the first months of lactation, when the feed consumption potential lags behind the growth rate of milk production [1].As a result, there is an increased mobilization of fat depots to reduce the negative energy balance in the body of cows. Therefore, in the first phase of lactation, the use of concentrated feeds in large quantities is practiced. The main negative aspect when feeding high-concentration diets is a decrease in rumen pH, which causes a change in rumen metabolism accompanied by the acidification of the contents of the rumen, a violation of microbial activity, a decrease in acetate formation, a drop in milk fat, and a number of other negative consequences associated with subacute rumen acidosis, or SARA (acidosis, laminitis, and infertility). The economic losses associated with SARA consist of a shortage of dairy products, a decrease in quality, the costs of veterinary measures, losses associated with the death or forced slaughter of animals, a decrease in reproduction functions, complications at the hotel, etc. For the health of the herd, the most dangerous impact of SARA is subclinical acidosis, for which there are no obvious signs, and the consequences manifest after a significant delay. A diet in which the rumen has a pH of medium to lower than 6.0 can cause the depression of fat formation, and the addition of buffers to correct milk fat depression is an effective means to increase pH and milk fat in such situations [2].An inadequate amount of fiber results in a low pH in the rumen. Low rumen pH affects the ratio of trans-10 fatty acids produced by biohydrogenation and potentially inhibits the complete saturation of trans-18: 1 fatty acids to stearins. We found that an increase in trans-10 fatty acids in milk fat and a decrease in milk fat appear only when diets with a low level of roughage (low rumen pH) supplemented with vegetable oil (a source of polyunsaturated fatty acids (PUFA)) are fed [3]. A direct effect of rumen pH was shown when the addition of buffers to a high-concentration diet reduced duodenal trans-fatty acids (TFA) flow and increased fat [4]. The addition of high levels of PUFA to a normal roughage diet has been ineffective in causing a depression of milk fat [5]. Diets formulated with adequate fiber to maintain a normal pH and those that also limit PUFAs in the feed as potential TFAs will contribute to a normal percentage of fat in milk [5].A reduction of energy deficiency in the first stage of lactation is achieved by the use of special fat additives that do not affect microbial fermentation in the rumen [6,7]. Reducing the energy deficit in cows at the beginning of lactation can be achieved by reducing the release of energy from milk by reducing the fat content in it. For these purposes, additives containing mixtures of positional and geometric isomers of octadecadienoic acid with conjugated bonds (conjugate linoleic acid (CLA)) have been studied and used [8,9,10,11]. It has been demonstrated that feeding such additives protected cows from the effects of rumen microorganisms and reduced the fat in milk, depending on the dose, from 36 to 62%, and, at the same time, did not affect milk yield [12]. Diets high in starch and low in fat, which can cause subacute rumen acidosis, may temporarily reduce milk fat synthesis [13].In order to clarify the mechanism of the depression of milk fat formation and preserve the health of animals, the aim of the research was to study the characteristics of rumen digestion, energy metabolism, and milk composition in high-productivity cows with high and low levels of milk fat that are fed the same diet.2. Materials and Methods2.1. Experiment Design and Animal ManagementThe studies were carried out on cows of the Holstein breed during their second lactation at the end of the first phase (80 days) using milk from the previous lactation (8–9 thousand kg of milk). Based on the data from the controlled milking, milk composition, lactation day, fatness, and lactation number, two groups of cows (10 heads each) were found in the same feeding group (a group of highly productive cows), but with different fat contents of milk (according to the results of the last controlled milking (Table 1)). At the same time, by origin (assigned bull-producer), the cows in the groups did not differ.The feeding ration for all groups of cows was the same and was set in the form of a feed mixture consistent with the level of milk production, live weight, and day of lactation, and was within the limits of the allowed content of individual nutrients (Table 2). The cows’ diets were represented by silage corn (20 kg), haylage of perennial grasses (8 kg), hay grass (0.5 kg), mixed fodder (16 kg), soybean meal (1 kg), fresh beer pellet (6 kg), and molasses (1 kg).Cows were also fed compound feed consisting of wheat (20%), barley (33%), oats (5%), maize (13%), sunflower meal (25%), salt (1%), mineral additives (1%), and premix (1%).Salt plus trace mineralized salt and vitamins were available ad lib and contained 38.5% NaCl and 61.5% of a trace mineral and vitamin mix. Trace mineral and vitamin mix contained 0.11% Mn, 0.14% Zn, 0.05% Fe, 0.025% Cu, 0.0027% I, 0.0024% Co, and 0.0007% Se. It also contained 650,000 IU of vitamin A, 250,000 IU of vitamin D3, and 2500 IU of vitamin E per kg of mix.2.2. Sample Collection and Measurement2.2.1. Rumen Fluid Collection and MeasurementSamples of rumen liquid were obtained 3 h after morning feeding using an oral gastric tube. The initial 100 mL of rumen fluid was discarded to avoid saliva contamination, and, finally, 100 mL of rumen fluid was filtered by four layers of cheesecloth and then collected. The rumen fluid pH was immediately measured using a pH electrode (OHAUS Starter ST2100-B) after collection. Immediately after rumen fluid collection, a 10 mL aliquot of ruminal fluid supernatant was preserved by adding 1 mL of 25% metaphosphoric acid for volatile fatty acid (VFA) determination, and another 50 mL aliquot of untreated ruminal fluid was kept for other analysis.The NH3-N concentration was measured using a microdiffusion method [14] in Conway dishes. Laboratory studies of the enzymatic activity of the microflora of the animal rumen were carried out using glass capillaries (in vitro). Cellulolytic activity was evaluated in accordance with the Henderson method [15] by the difference in the weight of the threads before and after incubation. The amylolytic ability of microorganisms in relation to “pure” nutrients (a 20% solution of potato starch) was studied. The amount of enzymatic activity was judged by the decrease in starch concentration in the solution after a 24 h incubation period. Quantitative indicators of rumen microbiocenosis were determined in the Goryaev chamber for 15–20 min from the time of collection, and the number of ciliates and the total number of bacteria in 1 mL of the contents were counted according to standard procedures [16].Samples of ruminal fluid were analyzed for VFA using a gas chromatograph (Cvet–800, Rus) equipped with a Hromaton N-AWDMCS (Chech) fused-silica capillary column (1.5 m × 3 mm i.d.). The detector temperature was 250 °C, the hydrogen carrier gas flow to the detector was 30.0 mL/min, airflow was 300 mL/min, and the flow of nitrogen makeup gas was 30.0 mL/min. A volatile free-acid mixture standard (MTY-1075 Matreya LLG, USA) was used for VFA determination.2.2.2. Indicators of Energy Metabolism in CowsHeat production was measured using the method of indirect calorimetry in respiratory trials using face masks [17]. The study of pulmonary gas exchange by the mask method allowed us to obtain a number of indicators characterizing pulmonary respiration, gas exchange, and tissue energy costs, such as: the volume of pulmonary ventilation per unit of time, in liters; respiratory rate per minute; exhalation capacity, in liters; the concentration of absorbed oxygen from inhaled air, as a percentage; the concentration of carbon dioxide in exhaled air, as a percentage; and the total volume of absorbed oxygen and released carbon dioxide during the experiment. Based on these data, the respiratory coefficient was calculated, the caloric value of one liter of oxygen consumed was determined, and the energy costs in the animals were calculated. Immediately afterward, the exhaled gases collection analysis of 02 and C02 was carried out on a gas analyzer-chromatograph (AHT-TI, Rus). The gross energy of the diets and milk samples was determined using an adiabatic bomb calorimeter (ABK-1, Rus).2.2.3. Plasma SamplesBlood samples were obtained 3 h after morning feeding by puncturing the tail artery, and then transferring the samples into vacutainers with sodium citrate. Plasma samples were obtained from the blood samples after centrifugation at 1500× g for 15 min at 4 °C. The plasma was then preserved at −20 °C until analysis. In the blood plasma, the glucose (GLU) content was determined with the help of an enzymatic colorimetric glucose oxidase method (using the panel of reagents from the firm Soared Diagnostic, SPb RUS), the concentration of non-esterified fatty acids (NEFA) was determined by means of enzymatic colorimetric method kit (Rendox, Crumlin, UK), β-hydroxybutyrate was assessed (with a set of reagents from Rendox), VFA was analyzed using a gas chromatograph (Cvet–800, Rus), urea was analyzed (with a set the reagents “Urea 450” with diacetylmonooxime from the firm Lahema), α-amine nitrogen [18] was assessed, and triacylglycerols were anylyzed by an enzymatic colorimetric method (with the panel of reagents from the firm Soared Diagnostic, SPb RUS). In the blood plasma samples, the concentration of insulin, thyroxine, triiodothyronine, and cortisol were determined by an enzyme immunoassay using commercial kits (DRG, Marburg, Germany).In the blood plasma, the activity of pyruvate carboxylase was determined by the method described by Scrutton [19] and by lactate dehyrogenase LDH, wherein a set of Lahem’s company is used when NAD is introduced into the reaction medium to determine the rate of conversion of lactate to pyruvate. The amount of acetoaldehyde was determined at 340 nm on a Specol-11 spectral colorimeter (Carl Zeiss, Oberkochen, Germany) after preliminary preparation. Using a chromatograph (Millichrom, RUS), the amounts of α-tocopherol, retinol, and thiamine were defined.2.2.4. Milk Composition AnalysisSamples of milk from three consecutive days were combined according to yield for each day and for each cow on the 50th, 80th, 100th, 140th, 170th, and 200th days of lactation. The milk was sampled in a 50 mL tube and analyzed immediately. The samples were analyzed using a milk analyzer (Lactostar, GmbH, Munich, Germany) for protein, fat, lactose, and urea nitrogen content.2.2.5. Fatty Acid AnalysisLipids were extracted from the milk according to Folch [20] by using methanol–chloroform (2:1), and then transmethylating. For analysis of FAs, a gas chromatograph (Cvet – 800, Rus) equipped with a Hromaton AWYMDS fused-silica capillary column (3.5 m × 3 mm i.d.) was used. The inlet and flame-ionization detector temperatures were 200 °C, and a 1-μL injection volume was used. The hydrogen carrier gas flow to the detector was 30 mL/min, airflow was 300 mL/min, and the flow of nitrogen makeup gas was 30 mL/min. Fatty acid peaks were identified by using a fatty acid methyl ester standard (SUCPM-47885, Supelko; Bellefonte, PA, USA).2.3. Statistical AnalysisThe obtained results were statistically analyzed using the general linear models (GLM) procedure, adapted from the IBM SPSS Statistics 11.5 user’s guide, with one-way ANOVA according to the following model: Yij = µ + xi + eij, where Yij is the dependent variable, µ refers to the mean, xi is the effect of treatment, and eij is the experimental error. Data are expressed as LS means and SE, unless reported otherwise. Differences were assumed to be significant if p ≤ 0.05 and p > 0.05, but p < 0.10 was considered a trend. The correlation analysis between the characteristics of the ruminal fluid and the amount of lactation between the fat of milk was performed using the Spearman procedure via the two-tailed test. p ≤ 0.05 were considered statistically significant and 0.05 < p < 0.1 was considered a trend difference.3. Results3.1. Ruminal Fermentation ParametersThe study of enzymatic and microbiological processes in the rumen of cows with different milk fat content showed that the indicators of rumen digestion corresponded to the characteristics of the cows’ diets. There were no signs of subacute acidosis (Table 3). The study of enzymatic and microbiological processes in the rumen of cows with different milk fat content showed that the indicators of digestion in the rumen corresponded to the characteristics of the cows’ diets. At the same time, cows with low milk fat content showed a tendency to have decreased pH (p = 0.09), fibrolytic activity (p = 0.07), a significant decrease in buffering capacity (p < 0.05), and an increase in the level of propionate (p < 0.05) compared to cows with normal milk fat content. The study of the correlation dependencies of rumen metabolism and milk fat content showed a direct dependence on pH (r = 0.62; p < 0.05), acetate level (r = 0.64; p < 0.05), and the ratio of acetate to propionate (r = 0.82; p < 0.05). An inverse relationship was found for propionate (r = −0.73; p < 0.05).3.2. Biochemical Parameters of the Blood of the CowsThe analysis of the biochemical parameters of the cows’ blood showed that the level of blood metabolites corresponded to the stage of lactation, the cows’ productivity, and the characteristics of the cows’ diets (Table 4). Analysis of the blood biochemical parameters showed that the level of blood metabolites corresponded to the stage of lactation, the cows’ productivity, and the cows’ dietary characteristics (Table 4). In the blood of cows with low milk fat content, there was a significant decrease in the concentration of non-esterified fatty acids (p < 0.05), with a higher level of VFA (p < 0.05). This indicates a higher rate of mobilization of the body’s fat depots to ensure the synthesis of milk components.3.3. Indicators of Energy Metabolism in CowsAnalysis of the data on energy exchange showed that cows with low milk fat tended to release less energy with milk (p = 0.1), and instead spend it on heat generation (by 6.2 MJ; p = 0.055) (Table 5). As a result, cows with low milk fat had a positive energy balance earlier, as evidenced by a significant (26.1 MJ p = 0.035) level of energy retention. During the observation of the fatness of cows from the 50th to the 80th days of lactation, it was determined that cows with normal milk fat content lost 1.5 points of fatness, while cows with low milk fat content lost only 1 point (p = 0.43), with unreliable differences.3.4. Milk Yield and CompositionsDuring sampling (days 80–82 of lactation), productivity indicators and milk composition significantly (p < 0.05) differed between the groups, though only for milk fat (Table 6). A significant difference was noted between the indicators only for the fat content in milk. Analysis of the milk of the various groups showed that all milk had no deviations in sanitary and hygienic standards, or in technological properties.Fatty Acid Composition of Milk LipidsThe determination of the fatty acid composition of milk lipids showed that only an increased content of fatty acids with an odd number of atoms (C15, C17) was revealed in the milk of cows with a low milk fat content (p < 0.05) (Table 7).4. DiscussionIt is generally believed that the formation and concentration of fat in the milk of dairy cows is influenced by diet [21,22,23,24,25]. In our experiment, the cows of both groups received the same diet; however, the fat content of milk was very different. The study of the subsequent milk productivity showed that by the 140th day of lactation, the fat content of the milk of the cows of both groups did not change (Table 8).Lower milk fat content means that less energy is required for milk synthesis. Romo et al. [26] noted that dietary trans-C18:1, which reduces milk fat percentage, significantly decreased milk energy output by 3.4 Mcal/d and numerically increased energy retained in tissue by 3.4 Mcal/d when compared with cis-C18:1. This reciprocal change between milk fat synthesis by the mammary gland as compared to fat deposition in adipose tissue has been observed for various dietary situations which cause milk fat depression [21,27,28].Reducing the fat content of milk in the first phase of lactation by the temporary artificial blocking of fat synthesis is used in practical cattle breeding to prevent excessive loss of body weight and preserve reproductive function. For this purpose, feed additives based on conjugated fatty acids are used. In such earlier experiments carried out [29], it was possible to reduce the release of energy with milk by 13.3% and to reduce the negative energy balance by 6.6 MJ. In the current experiment, the energy yield with milk decreased by 13.2%, and the energy retention increased by 26.1 MJ (p < 0.05) (Table 5).Cows with low milk fat had an earlier onset of positive energy balance and a shorter service period (109.5 ± 19.8 vs. 139 ± 18.4 days; p = 0.194). Thus, a decrease in the volume of synthesis of fatty acids in the mammary glands, as the most energy-intensive process, allows cows to reduce metabolic stress and maintain health.Analysis of the duration of productive use of the cows on this farm showed that the number of lactations was inversely correlated with the level of fat in milk (r = −0.68; p < 0.05, n = 1300). Special experiments aimed at assessing the long-term effects of feeding diets that reduce fat content have not been conducted, although it is noted that there are enough examples of apparently healthy herds that produce milk with a fat content in the range of 2.5 to 3.0%, which indicates that the level of milk fat is not necessarily incompatible with “healthy” cows [22].Milk fat depression caused by high concentrate, low forage diets usually occurs within a few days following dietary changes and is characterized by a substantial reduction in both yield and percentage of milk fat. Usually, a change in milk fat content (decrease) is associated with an increased consumption of concentrates and, thus, a low pH value of the rumen contents, a low proportion of acetate, and a decrease in the biohydrogenation capacity of the rumen microflora. High concentrate diets result in increased rumen production of propionate and increased hepatic rates of gluconeogenesis, which in turn increase the pancreatic release of insulin. In our research, we have studied the possible causes of milk fat reduction. Most of the proposed mechanisms that influenced milk fat content were associated with of a limited supply of lipid precursors in VFA content and a corresponding change in VFA metabolism [30]. Therefore, it was believed that low fat in milk is associated with a lack of acetate formation in the rumen as a precursor for the synthesis of fatty acids in the mammary gland because acetate and β -hydroxybutyric acid (BHBA) account for almost all of the carbon in the de novo synthesis of fatty acids by the mammary gland under normal situations, as well as during milk fat depression. However, the content of acetate in the contents of the rumen did not decrease (6.03 mmol/dL and 6.65 mmol/dL in the two groups, respectively) (Table 3). The blood acetate level in cows with low milk fat did not differ from those in cows with normal milk fat (Table 4).Another precursor for the formation of milk fat is BHBA. In our experiments on the rumen content, the concentration of butyric acid in cows with low milk fat was only slightly lower than in cows with normal milk fat (1.42 ± 0.02 versus 1.32 ± 0.02; p > 0.05) (Table 3). At the same time, the content of BHBA in the blood was also slightly lower (p > 0.05) (Table 4). The use of radioisotopes has shown that acetate content does not become deficient during the depression of fat formation [31]. It was also concluded that a BHBA deficiency for milk fatty acid synthesis is not a causative factor in the decrease in milk fat production [32]. Results of the present study also indicate that the reduction in milk fat content was not due to a decrease in the production of acetate and BHBA.Another possible mechanism for reducing fat in milk is related to an increase in insulin in the blood and a deficiency of vitamin B [27]. It was suggested that the decrease in milk fat previously observed with hyperinsulinemic-euglycemic clips, or with glucose [33] or propionate infusions was the most likely consequence of the ability of insulin to inhibit lipolysis, thereby limiting the availability of preformed fatty acids mobilized from body reserves in the mammary glands [34]. The involvement of thyroid hormones in the regulation of milk fat production has been reported [35]. However, neither vitamin [36] nor insulin [37] injections affected milk fat. We found neither cortisol nor thyroid hormones in the content of insulin, nor did we find them in the content of water or fat-soluble vitamins (Table 4). The glucogenicinsulin theory of milk fat depression assumes that tissues compete for nutrient use and that insulin causes nutrients to be diverted to adipose tissue at the expense of the mammary gland, as was observed in our experiments, however, without changing the level of insulin in the blood.Studies of the past years confirm that the depression of fat in milk related to the change in the rumen biohydrogenic processes (the polyunsaturated fatty acids (PUFA) present in the fat of the diet are hydrogenated rumen bacteria), and not to changes in the rumen VFA. Under normal feeding conditions, very few unsaturated fatty acids reach the small intestine. The precursors of PUFA in the diets of dairy cows are the linoleic acid (C 18:2) and linolenic acid (C18:3) contained mainly in plant lipids. A high concentration diet led to an increase of TFA in milk [26,38]. It has been demonstrated that conjugated linoleic acid (CLA) and trans-18:1 fatty acids (LC) resulting from the incomplete biohydrogenation of dietary polyunsaturated fatty acids (PUFA) in the rumen [3,39,40], or abomasal infusion [26,38,41], can significantly alter the synthesis of milk fat. Acetyl CoA carboxylase (ACA) has been identified as a limiting enzyme for the synthesis of fatty acids in the mammary glands [42]. The measured enzyme activity, together with tRNA for ASA, fatty acid synthetase, and sterol CoA desaturase in the mammary glands of cows, were markedly reduced in cows fed a diet that caused a depression of fat formation [9,39]. Unfortunately, no measurements were carried out in our experiments on CLA in rumen fluid, and, most importantly, CLA in rumen fluid was not altered by the diet intake. Therefore, it can only be assumed that the noted shifts in scar fermentation (a tendency to have lower pH (p = 0.09), a significant increase in propionate (p = 0.02), and a significant decrease in buffering capacity (p = 0.04)) could lead to a decrease in bio-hydrageneration capacity. However, this is only our guess. When milk fat is depressed, short-chain fatty acids (<C16) are primarily reduced. This indicates that trans fatty acids reduce the fat content in milk due to de novo synthesis. In our experiments, the fatty acid composition of milk fat did not differ in the two groups of cows. There was only a slight unreliable decrease in short and medium chain fatty acids in the milk of cows with low milk fat (54.4 ± 2.00 and 56.8 ± 1.48; p > 0.35) (Table 7). An increased content of fatty acids with an odd number of atoms (C15 and C17) was revealed in the milk of cows with a low fat content of milk (p < 0.05) (Table 7). An increase in the composition of milk fat with its low content of C14, C16, and C17 was also noted in other works [43].In our experiment, cows of both groups received the same diet; however, the fat content of the milk was very different. Apparently, there are additional factors that allow some cows to withstand the feed load for ruminal microbiocenosis or a different response. The increased production of trans-10 and cis-12 CLA in the rumen does not provide a universal explanation for the decrease in milk fat during diet-induced milk fat loss, suggesting that other biohydrogenation intermediates may also be involved [24]. Further research is needed to characterize the structure and function of other biohydrogenation intermediates and to consider the contribution of broader changes to rumen lipid metabolism to provide a more universal explanation for diet-induced low-fat milk syndrome (milk fat depression).5. ConclusionsIn our experiment, cows of both groups received the same diet; however, the fat content of the milk was very different. A decrease in the release of energy with milk led to an increase in the deposition of body reserves, as a result of which the availability of milk synthesis increased, and the service period decreased. This was due to the metabolizable energy being directed to body reserves instead of milk fat. The study of the functioning of the scar and the concentration of the main metabolites and hormones in the blood did not reveal significant changes that would explain the decrease in fat excretion with milk. It was found that the low milk fat content was not associated with low acetate formation in the rumen and hormone levels. Further study of this issue is required, taking into account the formation of CLA.

animals : an open access journal from mdpi

10.3390/ani13050853

PMC10000239

Beluga and sevruga are two highly valuable sturgeon species from the Acipenseride family in Iran. In recent years, research has been focused on commercial rearing of these species. A very important aspect in the sturgeon farming industry is the development of formulated compound diets for promoting growth. However, the ability of fish to digest compound diets is mostly related to the existence of the digestive enzymes in different parts of the gastrointestinal tract. In gastric species, protein digestion is conducted along the gastrointestinal tract by several proteases such as pepsin, trypsin, and chymotrypsin. Trypsin, as an alkaline protease, is able to hydrolyze protein residues and peptides to release free amino acids and small peptides for intestinal absorption; therefore, the activity of trypsin has been widely used as a valuable indicator of digestive capacity in fish. In this work, we aimed to characterize trypsin from beluga and sevruga for the first time. The results of our study show that the physicochemical and biochemical properties of trypsin from beluga and sevruga were in agreement with data reported in bony fish and may be considered a preliminary step to design in vitro tests for the assessment of protein digestibility in these primitive species.

This work aimed to determine the physicochemical and biochemical properties of trypsin from beluga Huso huso and sevruga Acipenser stellatus, two highly valuable sturgeon species. According to the results obtained from the methods of casein-zymogram and inhibitory activity staining, the molecular weight of trypsin for sevruga and beluga was 27.5 and 29.5 kDa, respectively. Optimum pH and temperature values for both trypsins were recorded at 8.5 and 55 °C by BAPNA (a specific substrate), respectively. The stability of both trypsins was well-preserved at pH values from 6.0 to 11.0 and temperatures up to 50 °C. TLCK and SBTI, two specific trypsin inhibitors, showed a significant inhibitory effect on the enzymatic activity of both trypsins (p < 0.05). The enzyme activity was significantly increased in the presence of Ca+2 and surfactants and decreased by oxidizing agents, Cu+2, Zn+2, and Co+2 (p < 0.05). However, univalent ions Na+ and K+ did not show any significant effect on the activity of both trypsins (p > 0.05). The results of our study show that the properties of trypsin from beluga and sevruga are in agreement with data reported in bony fish and can contribute to the clear understanding of trypsin activity in these primitive species.

1. IntroductionBeluga (Huso huso) and sevruga (Acipenser stellatus) are among the most important species of sturgeon fish (Acipenseridae) inhabiting the Caspian Sea with a high demand for products such as caviar, meat, skin, and cartilage [1,2]. Today, sturgeons are considered a vulnerable group of fish species for different reasons such as overfishing for production of meat and caviar, water pollution, and destruction of their natural habitats [3,4]. Therefore, in recent years, researchers have focused their studies on restocking and commercial rearing purposes. According to the Iranian Fisheries Organization report, the aquaculture production of sturgeon has increased from 363 t in 2009 to 2516 t in 2020 [5]. A very important aspect in sturgeon farming industry, affecting its production efficiency and long-term sustainability, is the development of formulated compound diets for promoting growth and product quality. However, the ability of fish to digest compound diets is mostly related to the existence of the digestive enzymes in different parts of the gastrointestinal tract [6]. Digestive enzymes reflect the capability of digestion in the organism under study and thus indicate the nutritional status at different stages of growth [7,8]. Herein, analysis of digestive enzymes activity is regarded as a biochemical procedure which may contribute to generate valuable information for understanding the physiology of digestion in fish [9,10]. This important issue can also help to define the requirements of fish for essential nutrients such as proteins, lipids, or carbohydrates [11]. Among macronutrients, dietary proteins are key nutrients for fish growth, since proteins are the building blocks of muscle cells and organs. They occur in a great array of forms, and their nutritional value depends on their amino acid composition. As Moraes and Almeida, 2020 [12] reviewed, the use of dietary proteins depends on a wide array of functional, biochemical, and genetic species-specific characteristics such as the age of organisms; the range of environmental factors (pH, dissolved oxygen, and ammonia levels); the amino acid profile of dietary protein; the digestible usable dietary energy levels; and the presence of antinutritional factors, among others.Protein digestion is conducted along the gastrointestinal tract by several proteases, with specific actions on the polypeptide chain. In gastric species, pepsin, trypsin, and chymotrypsin are the most important proteolytic enzymes in fish [13,14,15]. As shown by Nolasco-Soria, 2021 [16], trypsin in combination with other alkaline proteases and peptidases such as chymotrypsin, aminopeptidases, and carboxypeptidases complete the acid predigestion conducted in the stomach, hydrolyzing protein residues and peptides to release free amino acids and small peptides for intestinal absorption; therefore, the activity of trypsin has been widely used as a valuable indicator of digestive capacity in fish, as well as a useful biomarker for its nutritional and physiological condition [17]. According to surveys conducted in various species of fish, trypsin participates in activating trypsinogen and other zymogens in the intestine and plays an effective role in protein degradation of the consumed diet in the carnivorous fish up to 40–50% [14,18,19,20,21]. Furthermore, trypsin quantification is essential for the design of in vitro digestibility protocols of feed ingredients and for the formulation of high digestible compound feeds for aquaculture fish species [22].Hence, a better understanding of the properties of trypsin is necessary to generate valuable information for protein degradation in the fish digestive tract. The characterization of trypsin, especially its physicochemical and biochemical properties, has been thoroughly studied from the intestine of various fish including grass carp (Ctenopharyngodon idellus), spotted goatfish (Pseudupeneus maculatus), grey triggerfish (Balistes capriscus), skipjack tuna (Katsuwonus pelamis), smooth hound (Mustelus mustelus), and Brazilian flounder (Paralichthys orbignyanus) [23,24,25,26,27,28]. The activity of trypsin among several sturgeon species has been mostly studied during larval ontogeny in the members of the genus Acipenser such as A. transmontanus [29], A. fulvescens [30], A. oxyrinchus [31,32], A. baerii [33], A. persicus [34], A. nacarii [35,36], A. stellatus [37], and genus Huso such as H. huso [38]. However, there are no studies evaluating the physicochemical and biochemical characteristics of trypsin in sturgeons; thus, this study attempts to characterize trypsin from beluga and sevruga, as two of the main sturgeon species from the Caspian Sea, for the first time.2. Materials and Methods2.1. Fish SamplesViscera from five specimens of beluga and sevruga (8.0 ± 0.4 kg; 95 ± 8 cm) were obtained from Saei sturgeon rearing center, Sari, Mazandaran, Iran. Fish were fed a commercial diet (crude protein 46%, crude lipid 16%, ash 8.5%, crude fiber 2.5%, and moisture 9%, Mazandaran Animal & Aquatic Feed Company, Semeskandeh Olya, Iran) and kept in fasting condition for 72 h before sampling. Those samples were packed in polyethylene bags, placed in ice with the sample/ice ratio of approximately 1:3 (w/w), and directly transported to the laboratory. Upon arrival, the intestine was separated from the rest of the collected viscera, washed with cold distilled water (4 °C), pooled, and stored at −80 °C for further analysis.2.2. Preparation of Intestinal Crude Extracts for Trypsin CharacterizationThe frozen intestine of beluga and sevruga was partially thawed in the refrigerator at 4 °C for 2 h. The samples were then cut into small pieces and homogenized in 50 volumes of 50 mM Tris–HCl buffer (pH 7.5, 10 mM CaCl2, 0.5 M NaCl) by a tissue homogenizer (Heidolph Diax 900, Sigma Co., St. Louis, MO, USA) at 4 °C for 2 min. The homogenate was then filtered with a cheese cloth to separate the floating fat phase and centrifuged for 45 min at 4 °C at 14,000× g by a refrigerated centrifuge (Hettich Benchtop Centrifuge Rotina 420R, Berlin, Germany). The resulting supernatant from each sample was collected, defined as intestinal crude extract (ICE), and then used throughout this study.2.3. ReagentsEDTA (Ethylenediaminetetraacetic acid), Pepstatin A, PMSF (phenylmethanesulfonyl fluoride), and sodium cholate were obtained from Molekula Co (Gillingham, UK). BAPNA (Nα-benzoyl-DL-arginine-ρ-nitroanilide hydrochloride), ß-mercaptoethanol, BSA (bovine serum albumin), iodoacetic acid, saponin, SBTI (soybean trypsin inhibitor), TLCK (N-ρ-tosyl-L-lysine-chloromethyleketone), and TPCK (N-tosyl-L-phenylalanine chlorom ethyleketone) were purchased from Sigma Chemical Co (St. Louis, MO, USA). Molecular weight marker (PM 2700) was obtained from SMOBIO Technology, Inc. (Hsinchu, Taiwan).2.4. Trypsin AssayTo measure the enzyme activity in ICE, BAPNA was used as a substrate at a concentration of 1 mM in 50 mM Tris–HCl, 20 mM CaCl2 (pH 8.5) according to the method of Erlanger et al., 1961 [39]. Each ICE (25 μL) was mixed with the prepared substrate (1250 μL) and incubated at 55 °C for 20 min. The reaction was terminated by adding 30% (v/v) acetic acid (250 μL) to the mixture and followed by measuring the trypsin activity at an absorbance of λ = 410 nm using a spectrophotometer (UV-1601, Shimadzu, Kyoto, Japan). One unit of activity was defined as 1 μmol of ρ-nitroaniline released per min and calculated with the following equation [40]:Trypsin activity unit/mL=Absorbance 410 nm ×1000× mixture volume mL8800× reaction time min×0.025 where 8800 (cm−1 M−1) is the molar extinction coefficient of ρ-nitroaniline measured at λ = 410 nm.2.5. Protein AssayThe concentration of protein in both ICEs was determined at λ = 750 nm by using BSA (1 mg mL−1 as a standard) and Folin–Ciocalteau reagent according to the Lowry et al., 1951 [41] method.2.6. Characterization of Trypsin by SDS-PAGE ElectrophoresisSDS-PAGE electrophoresis was performed to determine the protein pattern in ICEs from both sturgeon species [42]. Each ICE was mixed at 2:1 (v/v) ratio with sample buffer (62.5 mM Tris–HCl pH 6.8, 2% SDS (w/v), 10% (v/v) glycerol, 0.3% (w/v) bromophenol blue and 5% (v/v) ß-mercaptoethanol) and boiled for 10 min. Thereafter, the ICEs (with protein concentration of 15 µg) were loaded onto the gel made of 4% stacking gel and 12% separating gel and the electrophoresis was run at a constant current of 15 mA using a vertical electrophoresis system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). After the run, protein bands present in the gel were stained with 0.1% Coomassie Brilliant Blue (G-250) in methanol (35%) and acetic acid (7.5%) and unstained in methanol (35%) and acetic acid (7.5%).Casein-zymography was performed after electrophoresis for detection of proteases in both ICEs as described by Garcia-Carreno et al., 1993 [43]. Both ICEs were submitted to native-PAGE electrophoresis in a same manner of SDS-PAGE where samples were not boiled, and SDS and reducing agent were removed. After the run, the gel was immersed in 50 mL of a casein solution (20 mg mL−1 in 50 mM Tris–HCl, pH 7.5) for 1 h at 4 °C with gentle agitation to allow diffusion of the casein into the gel. Thereafter, the gel was transferred to another solution (50 mL) containing casein (20 mg mL−1 in 50 mM Tris–HCl, pH 8.5, 10 mM CaCl2) for 20 min at 55 °C with continuous agitation. The gel was then stained with 0.1% Coomassie Brilliant Blue (R-250) in methanol (35%) and acetic acid (7.5%) and unstained in methanol (35%) and acetic acid (7.5%). The presence of proteolytic activities in both ICEs was indicated by the appearance of clear zones on the blue background of the gel, which meant that casein was digested by the targeted protease in these areas.To reveal the trypsin present in both ICEs, the inhibitory activity staining was used after the submission of both ICEs to native-PAGE electrophoresis as described by Ahmad and Benjakul [44] with a slight modification. After the run, the gel was immersed in 30 mL of an SBTI solution (1 mg mL−1 in 50 mM Tris–HCl, pH 8.5, 10 mM CaCl2) for 30 min at 4 °C to allow diffusion of SBTI into the gel. Thereafter, the incubation of gel was performed for 40 min at 55 °C and followed by washing in cold distilled water and staining with 0.05% Coomassie Brilliant Blue (R-250) to appear inhibitory zones, indicating the presence of the trypsin in both ICEs. The molecular weight of the trypsin that appeared in both ICEs was estimated using wide-range molecular weight markers (PM2700, SMOBIO, Hsinchu, Taiwan) by calculating the trypsin Rf in comparison with those of protein markers.2.7. Optimum Temperature and ThermostabilityTo determine the optimum temperature for trypsin activity, the activity of this alkaline protease was measured in both ICEs at different temperatures, including 10, 25, 35, 45, 50, 55, 60, 65, and 70 °C after 20 min of incubation at pH 8.5, using 1 mM BAPNA as a substrate. For the thermostability test, both ICEs were incubated at the above-mentioned temperatures for 30 min and then cooled in an ice bath for assay of residual activity of the enzyme at pH 8.5 as described by Zamani et al., 2014 [40].2.8. Optimum pH and StabilityDifferent buffers in the pH range of 4.0 to 11.0 (50 mM acetic acid–sodium acetate for pHs 4–6; 50 mM Tris–HCl for pHs 7–9 and 50 mM glycine–NaOH for pHs 10–11) were used for determining the optimum pH for trypsin activity. Both ICEs were used, and they were incubated using1 mM BAPNA as a substrate after 20 min of incubation at 55 °C at different pHs. For the pH stability test, the remaining activity of the trypsin from each ICE was measured using 1 mM BAPNA as a substrate at 55 °C after being incubated at the above-mentioned pHs for 30 min [40].2.9. Effect of InhibitorsSeveral protease inhibitors (0.01 mM pepstatin A, 0.05 mM SBTI, 1 mM iodoacetic acid, 2 mM EDTA, 5 mM TLCK, 5 mM TPCK, 5 mM ß-mercaptoethanol, and 10 mM PMSF) were prepared in the relevant solvents and incubated with an equal volume of each ICE at room temperature for 15 min. The remaining activity of the enzyme was then measured by 1 mM BAPNA as a substrate (at 55 °C, pH 8.5) and the percent inhibition was calculated according to the method of Khantaphant and Benjakul, 2010 [45]. The trypsin activity of control was measured in the same manner without the presence of inhibitors and scored to 100%. 2.10. Effect of Metal IonsTo investigate the effect of metal ions (5 mM) on the trypsin activity of both ICEs, univalent (K+, Na+) and divalent (Ca2+, Cu2+, Zn2+ and Co2+) cations were dissolved in 50 mM Tris–HCl (pH 8.5) and then incubated with an equal volume of each ICE for 30 min at room temperature. The residual activity of the enzyme was determined using 1 mM BAPNA as a substrate at 55 °C and pH 8.5 [40]. The enzymatic activity of control was assayed without the presence of metal ions and taken as 100%.2.11. Effect of Surfactants and Oxidizing Agents The effect of surfactants (anionic: SDS and sodium cholate; non-ionic: saponin and Triton X-100, all at 1%) and oxidizing agents (sodium perborate at a concentration of 1% and H2O2 at three concentrations of 5%, 10%, and 15%) on the trypsin activity was measured by incubation of the above-mentioned surfactants and oxidizing agents with an equal volume of each ICE for 1 h at 40 °C. The residual activity of the enzyme was then determined at 55 °C and pH 8.5 using 1 mM BAPNA as a substrate. The assessment of control enzymatic activity was conducted in a similar condition in the absence of chemicals and scored to 100% [25].2.12. Statistical AnalysisThis study was conducted on the basis of a completely randomized design, and a one-way ANOVA was used for data analysis using SPSS package 22.0 (SPSS Inc., Chicago, IL, USA). All experimental assessments were performed in triplicate, and data was expressed as the mean ± standard deviation (SD). The comparison of means was carried out by Duncan’s multiple range tests with a statistical significance at p < 0.05.3. Results and DiscussionFish proteases such as trypsin have been the main objective of many of studies, but it is difficult to compare the results obtained in different species because the data are affected by the use of many different methodologies, the state of feeding of experimental animals (fed vs. starved fish), and the type of enzyme preparation (intestinal tissues alone vs. intestinal extracts with the intestinal content) [46].3.1. Protein Pattern, Casein-Zymographyand Inhibitory Activity Protein pattern of ICE from beluga and sevruga is depicted in Figure 1a. Based on results obtained from the SDS-PAGE, a number of proteins with different molecular weights were shown in the ICE of both sturgeon species. The major bands of each ICE appeared between molecular weights comprised between 10 and 60 kDa.Casein-zymogram can be used as a highly sensitive and fast assay method for detecting nanograms of protein. The protease activity of both ICEs was demonstrated by casein-zymography as illustrated in Figure 1b. The clear bands, showing the presence of protease, appeared on the gel with different molecular weights. Based on the zymogram pattern, proteases present in ICE from beluga were observed in the range of molecular weights between 19 to 35 kDa, while those in the ICE of sevruga ranged from molecular weights of 19 to 45 kDa.Inhibitory activity staining for detection of trypsin in ICEs is depicted in Figure 1c. Results showed that a single band for each ICE clearly appeared on the gel with a molecular weight of 27.5 and 29.5 kDa for sevruga and beluga, respectively. In general, trypsins have molecular weights in the range of 20–30 kDa [47]. In particular, different molecular weights for trypsins have been reported in various fish species such as 21.7 kDa for mrigal carp [48], 23.2 kDa for common kilka [40], 23.5 kDa for pirarucu [49], 24 kDa for small red scorpion fish [50], 21 and 24 kDa for liver of albacore tuna [51], 24 kDa for catfish [52], 24.4 kDa for gulf corvina [53], 25 kDa for Monterey sardine [54], 26 kDa for common dolphinfish [55], 27 kDa for zebra blenny [56], 28.8 kDa for sardinelle [57], 29 kDa for Atlantic bonito [58], 38.5 kDa for tambaqui [59], and 42 kDa for skipjack tuna [60]. However, several reasons such as different habitat and climate, autolytic degradation, and genetic variation among fish species may explain why trypsins from various sources have different molecular weights [60,61].3.2. Optimum Temperature and ThermostabilityEnzymes are one of the main biological macromolecules and their maximum activity depends on an optimum temperature to make them functional. Figure 2a revealed that optimum temperature of the trypsin in ICE prepared from beluga and sevruga was found to be 55 °C, although 92.70% of the maximum activity of the enzyme was still maintained at 60 °C for both sturgeon species. However, an obvious decrease in the trypsin activity of both ICEs was observed at temperatures above 60 °C, probably due to thermal inactivation of this enzyme caused by protein unfolding [40]. Similar optimum temperature (55 °C) was recorded for trypsins in skipjack tuna [26], gilthead seabream [46], sardinelle [57], and silver mojarra [62]. Optimum temperature of trypsin for both sturgeon species was higher than values reported for cold-water fish such as Atlantic cod (Gadus morhua) [63], grey triggerfish [25], lane snapper (Lutjanus synagris) [64], and Japanese sea bass (L. japonicus) [65] indicating optimum temperatures over a range of 40–45 °C. These differences could be attributed to the temperature of the fish habitat or experimental conditions used in assessments [66]. The optimum temperature for enzyme maximum activity may be interesting for comparative physiological studies, even though such data offer limited information on enzyme activity under normal rearing conditions. Although fish trypsins are mostly unstable at temperatures higher than 40–50 °C, their thermal stability is well known to be at temperatures below 40 °C [67]. Trypsin thermal stability from ICE of beluga and sevruga is displayed in Figure 2b. As can be observed in this figure, the stability of both trypsins was highly maintained up to 50 °C with a remaining activity of 90.2% and 91.7% for sevruga and beluga, respectively. A gradual decrease in the activity of both trypsins was recorded at 55 °C, whereas enzymatic activity sharply decreased at 60 °C. After heating the ICEs at 70 °C, the relative activities for both trypsins were only about 0.9% and 1.6% of their initial activity for sevruga and beluga, respectively. These results were in accordance with those of sardinelle, common kilka, mrigal carp, and pirarucu, which were exhibited to be stable up to 50 °C [40,48,49,57]. The trypsins from beluga and sevruga showed to be more stable at high temperatures in comparison with those reported for the Monterey sardine, chinook salmon, bluefish, Tunisian barbell, and common dolphinfish that the enzymatic activity was rapidly lost at temperatures above 40 °C [54,55,58,68,69]. In general, thermostability of the trypsin enzyme might vary by some factors such as fish species and experimental conditions [23,70]. From a biological point of view, it is difficult to deduce any advantage for beluga and sevruga in possessing proteases showing different resistances to heating, since the normal temperature of water rarely exceeds 21–24 °C. Nevertheless, from a biotechnological perspective, it may be interesting to have information about active and easily denaturalizable proteases potentially useful in the feed industry [71].3.3. Effect of pH on Trypsin Activity and StabilityThe results observed from the effect of pH on the activity and stability of trypsin from beluga and sevruga are illustrated in Figure 3. Trypsins from both species had a maximal activity at pH 8.5 (Figure 3a). Trypsin activity was dramatically reduced at pH values ranging from 4.0 to 5.0. Our results showed that the stability of trypsin from both sturgeon species was highly preserved at pH values comprised between 5.0 and 11.0 with activity values of 75% for beluga and 80% for sevruga. The high ranges of pH may change the net charge and conformation of an enzyme and inhibit to bind to the substrate properly, resulting in the abrupt loss of enzymatic activity [60,72]. Trypsins are mainly known to have more activity within a range of pH values comprised between 7.5 and 10.5 [46,73]. The optimum pH (8.5) recorded for trypsin in both sturgeon was similar with results reported for trypsins from the brownstripe red snapper viscera and the albacore tuna hepatopancreas [45,51], whereas both trypsins indicated the lower optimum pH than those recorded for the Japanese sea bass, gilthead seabream and common dentex, and the pirarucu [46,49,65]. However, optimum pH may differ depending upon the experimental conditions such as concentration and type of substrate, temperature, and type of metal ions [60]. For instance, Martinez et al., 1988 [74] showed that the trypsin from pyloric caeca of the anchovy had the optimum pHs of 8.0 and 9.5 for the hydrolysis of BAPNA and casein, respectively.The effect of pH on the trypsin stability in both ICEs is displayed in Figure 3b. The stability of both enzymes was considerably retained between pH 6.0 to 11.0. Trypsin activity was lost about 63.15% and 69.26% at pH 4.0 in sevruga and beluga, respectively, while the loss of the enzyme activity at pH 5.0 was recorded by 34.13% and 35.84% for sevruga and beluga, respectively. However, trypsin in ICE of sevruga and beluga lost only 1.19% and 1.51% of its activity at pH 8.5, respectively. Similar behavior was reported for trypsins from common kilka [40], common dolphinfish [55], and zebra blenny [56] which retained 80–100% of the activity at pH ranges of neutral and alkaline. The high catalytic activity of the trypsin is observed in alkaline pHs, and its stability at a particular pH may be linked to the net charge of the enzyme at that pH [75].3.4. Effect of Inhibitors on Trypsin Activity The sensitivity of protease enzymes to various inhibitors is a valuable tool for their proper functional characterization [55]. Based on their nature, inhibitors can be classified into two classes: chemical inhibitors and protein inhibitors [76]. A trypsin inhibitor is a type of serine protease inhibitor that reduces the biological activity of trypsin thereby rendering it unavailable to bind with proteins for the digestion process [77,78]. Therefore, it can be considered important to characterize the effect of different inhibitors on the trypsin activity.Table 1 shows the effect of different inhibitors on the activity of trypsin from beluga and sevruga. As it is shown in this table, a serine protease inhibitor such as PMSF inhibited 39.11% and 36.29% of the trypsin activity in sevruga and beluga ICE samples, respectively. Both enzymes were completely inhibited by trypsin specific inhibitors such as SBTI and TLCK, while a chymotrypsin-specific inhibitor (TPCK) did not show any inhibitory effect on their enzymatic activity (p > 0.05). Furthermore, a metalloproteinase inhibitor (EDTA) and a disulfide bond reducing agent (ß-mercaptoethanol) had a partial inhibitory effect on trypsin activity in both sturgeon species, although the inhibition rates varied between both species. In particular, trypsin from sevruga was inhibited by ß-mercaptoethanol (25.33%) and EDTA (23.55%) more than those in beluga (22.84 and 21.06%, respectively) where no significant difference was shown between both species (p > 0.05). Results from EDTA indicate the high dependence of trypsin activity from both sturgeon species on divalent cations [46]. However, an aspartic proteinase inhibitor (Pepstatin A) and a cysteine proteinase inhibitor (iodoacetic acid) exhibited a negligible inhibitory effect on the trypsin activity of both species. Similar results have been observed in other fish species [40,51,56]. For instance, Khangembam and Chakrabarti, 2015 [48] reported that trypsin activity from the digestive system of mrigal carp was inhibited by SBTI and TLCK. SBTI is a single polypeptide chain that acts as a reversible competitive inhibitor of trypsin and forms a stable, enzymatically inactive complex with trypsin, resulting in reduction of the enzyme availability [79]. TLCK is an irreversible inhibitor of trypsin and trypsin-like serine protease that deactivates these enzymes through the formation of a covalent bond with histidine residue in the catalytic site of the enzyme and blocks the active center of the enzyme for binding to substrate [80]. As reported by several authors [40,49,56], PMSF strongly inhibited the activity of trypsin from viscera of common kilka, pirarucu, and zebra blenny, respectively, whereas TPCK had no effect on the enzyme activity from common kilka [40]. The trypsin activity from the intestine of common dolphinfish was partially inhibited by β-mercaptoethanol and EDTA [54], while the trypsin activity from liver of albacore tuna was not reduced in the presence of pepstatin A and iodoacetic acid [51].3.5. Effect of Metal IonsMetal ions have a key role in the activity regulation of many enzyme-catalyzed reactions [81]. Our results on the effect of metal ions on the trypsin activity in beluga and sevruga are detailed in Table 2. No significant effect on the activity of both enzymes was found in the presence of univalent cations Na+ and K+ (p > 0.05). The enzymatic activity of trypsin in both species was significantly reduced by divalent cations Cu2+, Zn2+ and Co2+, whereas Ca2+ significantly enhanced the activity of both trypsins (p < 0.05). Similar results on the effect of Ca2+ on the trypsin activity were also observed in common kilka, common dolphinfish, and zebra blenny [40,55,56]. The attachment of Ca2+ to the active site of serine proteases such as trypsin not only increases the stability of the enzyme structure, but it also protects the enzyme from self-digestion [66,69,82]. The enzymatic activity of trypsin in common dolphinfish was reduced by 82% and 81% by Zn2+ and Cu2+, respectively [55], while 100% of enzymatic activity of tryspin from zebra blenny was lost in the presence of Zn2+ and Cu2+ [56]. In common kilka [40], no inhibition was observed in the trypsin activity in presence of Na+ and K+. Differences in percent inhibition might be linked to species diversity, environmental adaptations and feeding habits of fish [83].3.6. Effect of Surfactants and Oxidizing Agents Surfactants are the most widely used groups of compounds today, with wide application in industry and household. These are unique substances that contain hydrophobic and hydrophilic moieties within their molecule and find enormous applications in biology. Oxidizing agents such as sodium perborate and H2O2 are also used in the detergent industry as bleaching agent. Surfactants and oxidizing agents may be harmful or even toxic to aquatic organisms. These compounds can penetrate through tissues and bind to biomolecules, such as enzymes, causing changes in cellular activity [84]. As reviewed by Rubingh, 1996 [85], surfactants can influence the activity of enzymes in two ways. Firstly, by binding to the enzyme, surfactants can influence intrinsic enzyme properties such as the secondary and tertiary structure or flexibility, and thereby, affect its ability to serve as a catalyst. A less direct, but equally important, way in which surfactants affect enzyme activity is by changing the environment in which the enzyme functions. It is well-known that SDS disrupts non-covalent bonds within and between enzymes, denaturing them, and resulting in the loss of their native conformation and function [86], whereas saponin, Triton X-100, and sodium cholate are the non-denaturing surfactants [87,88,89]. Hence, characterizing the effect of these chemical compounds on trypsin activity is of relevance for proper characterizing its activity.The results on the effect of various surfactants and oxidizing agents on the trypsin activity in sevruga and beluga are shown in Table 3. A significant increase in the activity of both trypsins was observed after incubation for 1 h at 40 °C in the presence of surfactants tested, including saponin, sodium cholate, and Triton X-100 at final concentrations of 1% (p < 0.05). Both trypsins were highly unstable against sodium dodecyl sulfate (SDS), in which trypsins from sevruga and beluga significantly lost about 94% and 97% of their activity in the presence of 0.1% SDS, respectively (p < 0.05). Similar results were found in trypsins of other fish species in the presence of saponin, sodium cholate, Triton X-100 and SDS [40,56]. The obtained results on the effect of oxidizing agents on both trypsins showed that the enzymatic activity was reduced in the presence of sodium perborate (1%) in sevruga and beluga by 22.23% and 24.37%, respectively. The activity of both enzymes was also decreased significantly with an increase in H2O2 concentrations from 5% to 15%, as described in Table 3 (p < 0.05). Trypsin from sevruga showed significantly higher activity than trypsin from beluga in the presence of H2O2 ranging from 5% to 15%, indicating that trypsin from sevruga was more tolerant to H2O2 than trypsin from beluga. The biochemical and structural properties of enzyme can affect its ability as a catalyst in presence of oxidizing agents [85]. These results showed that trypsins from sevruga and beluga were more stable against H2O2 than trypsins from grey triggerfish and zebra blenny [25,56], whereas most proteases have shown to be unstable in the presence of oxidizing agents like hydrogen peroxide [25].4. ConclusionsThe results of our study indicated that trypsin from intestine of beluga and sevruga had similar properties to trypsins from bony fish. The enzyme had an optimum temperature of 55 °C and thermal stability was maintained over 90% up to 55 °C. This alkaline protease had an optimum pH of 8.5 and showed to be tolerant in the pH range of 6.0 to 11.0 in both studied sturgeon species. The molecular weight of trypsin for sevruga and beluga was estimated to be 27.5 and 29.5 kDa, respectively, as data from inhibitory activity staining indicated. Both trypsins were inhibited by main specific inhibitors, SBTI and TLCK. Additionally, the enzymatic activity of trypsins was still detected after 1 h in the presence of surfactants and oxidative agents. Information provided in this manuscript related to trypsin activity for beluga and sevruga based on changes in the activity of this alkaline protease based on a tested range of pH and temperature values, and presence of potential inhibitors, ions and cations may be considered a preliminary step to design in vitro tests for the assessment of protein digestibility in these species.

animals : an open access journal from mdpi

10.3390/ani13101654

PMC10215608

In our research, candidate genes related to sheep-milk production were revealed by a genome-resequencing analysis and a genome-signal-selection analysis, and a RT-qPCR experiment was performed to prove the expression levels of these candidate genes, the results showed that the FCGR3A gene’s expression level had a significant negative relationship with sheep-milk production.

Natural selection and domestication have shaped modern sheep populations into a vast range of phenotypically diverse breeds. Among these breeds, dairy sheep have a smaller population than meat sheep and wool sheep, and less research is performed on them, but the lactation mechanism in dairy sheep is critically important for improving animal-production methods. In this study, whole-genome sequences were generated from 10 sheep breeds, including 57 high-milk-yield sheep and 44 low-milk-yield sheep, to investigate the genetic signatures of milk production in dairy sheep, and 59,864,820 valid SNPs (Single Nucleotide Polymorphisms) were kept after quality control to perform population-genetic-structure analyses, gene-detection analyses, and gene-function-validation analyses. For the population-genetic-structure analyses, we carried out PCA (Principal Component Analysis), as well as neighbor-joining tree and structure analyses to classify different sheep populations. The sheep used in our study were well distributed in ten groups, with the high-milk-yield-group populations close to each other and the low-milk-yield-group populations showing similar classifications. To perform an exact signal-selection analysis, we used three different methods to find SNPs to perform gene-annotation analyses within the 995 common regions derived from the fixation index (FST), nucleotide diversity (Ɵπ), and heterozygosity rate (ZHp) results. In total, we found 553 genes that were located in these regions. These genes mainly participate in the protein-binding pathway and the nucleoplasm-interaction pathway, as revealed by the GO- and KEGG-function-enrichment analyses. After the gene selection and function analyses, we found that FCGR3A, CTSK, CTSS, ARNT, GHR, SLC29A4, ROR1, and TNRC18 were potentially related to sheep-milk-production traits. We chose the strongly selected genes, FCGR3A, CTSK, CTSS, and ARNT during the signal-selection analysis to perform a RT-qPCR (Reale time Quantitative Polymerase Chain Reaction) experiment to validate their expression-level relationship with milk production, and the results showed that FCGR3A has a significant negative relationship with sheep-milk production, while other three genes did not show any positive or negative relations. In this study, it was discovered and proven that the candidate gene FCGR3A potentially contributes to the milk production of dairy sheep and a basis was laid for the further study of the genetic mechanism underlying the strong milk-production traits of sheep.

1. IntroductionThe sheep is one of the earliest domesticated-farm-animal species and has experienced evolution and domestication over thousands of years [1]. Dairy sheep are traditionally farmed in southern Europe (France, Italy, Spain, Greece), central Europe (Hungary and the Czech and Slovak Republics), eastern Europe (Romania and Ukraine), and countries in the Middle East, such as Turkey and Iran [2]. Organized sheep-breeding programs were developed from at least the 1960s [3]. The most efficient selection scheme for local dairy sheep is based on the pyramidal management of the population, with the breeders of the nucleus flocks at the top, and pedigree, official milk recording, AI (artificial insemination) breeding, controlled natural mating, and breeding-value estimation (i.e., BLUP) are carried out to make genetic progress. In 2013, dairy small ruminants accounted for a minor part of the total agricultural output in France, Italy, and Spain (0.9 to 1.8%) and a larger part in Greece (8.8%) [4]. In these European countries, the dairy-sheep industry is based on local breeds and crossbreeds raised under semi-intensive and intensive systems, and it is concentrated in a few regions. While with the development of dairy-sheep breeding and the emergence of dairy products emerging, dairy ruminants have become a major part of European agricultural income [5,6]. The average flock size varies from small to medium (140 to 333 ewes/farm), and the average milk yield ranges from low to middle (170 to 500 L/ewe) [7], showing substantial space for improvement in relation to cows’ milk [8]. Furthermore, sheep milk has higher protein, fat, lactose, total non-fat solids, and ash contents and a higher nutritional value than cows’ and goats’ milk [9], which makes it suitable for processing into various types of dairy products. Most sheep milk is sold to industries and then processed into traditional cheese products, most of which are made into Protected Denomination of Origin (PDO) cheeses for gourmets. The animals’ udder health is also critical in sheep-milk quantity and quality. Mastitis is an inflammation of the mammary gland that is usually caused by pathogens, mainly bacteria, which develop in the udder tissue after infections in the teat canal. It is one of the most prevalent and costly diseases in the dairy industry due to the significant reductions in milk production and physical harm that it causes [10,11,12]. In previous studies, some quantitative trait loci (QTLs) and SNPs associated with SCC (somatic cell counts) that serve as markers of mastitis were identified based on linkage-disequilibrium analyses of different dairy-sheep breeds [10,13,14], leading to significant improvements in ewes’ ability to resist mastitis [15,16].The dramatic decreases in the costs of whole-genome sequencing (WGS) and RT-qPCR experiments on animals have made it possible to scan the complete genomes of thousands of animals. Using genome information makes it possible to explain parts of the total genetic variance that are difficult to measure, such as low-heritability, sex-limited, and postmortem traits. To help clarify the link between the genome sequence and real data of important traits, this research was designed to gain a better understanding of sheep lactating genes by comparing the SNP-variant information on different sheep breeds with significantly different dairy-production characteristics. These breeds significantly differ at two dairy levels, since there are high-milk-yield (East Friesian sheep, 700 kg/lactation, Dairy Meade sheep, 500 kg/lactation and Awassi sheep) and low-milk-yield sheep breeds (Hu sheep, small-tailed Han sheep, and Churra sheep) [17]. To this end, we performed a whole-genome sequencing (WGS) analysis of these sheep breeds, as we found that in previous studies, little research was performed using whole-genome sequencing on these dairy sheep populations. First, we re-sequenced the crossed breed of Dairy Meade sheep and small-tailed Han sheep. The information on this sheep breed will be uploaded to a public genome-sequence database for researchers to use freely (NCBI). In our research, a genetic-structure analysis was performed to establish the relationships and geographical distances between 10 sheep breeds, and then genome-wide-signal scans were carried out to identify significant genes associated with sheep-milk production. The genes validated by the RT-qPCR results and their relationship with milk production represent significant progress in our knowledge of the gene-expression levels of lactating sheep and, thus, provide potentially valuable information for future dairy-sheep studies.2. Material and Methods2.1. AnimalsOne hundred and one sheep samples were sequenced in this study. This step was approved by the Animal Ethics Committee of China. In total, 41 sheep-ear samples were provided by the M-Natural Animal Husbandry Technology Company, including 36 sheep with high levels of milk production and 5 sheep with low levels of milk production. The high-milk-yield populations included Dairy Meade (DM) sheep (11), the crossbreed of Dairy Meade sheep and small-tailed Han sheep DM (F1) (15), and the crossbreed of DM (F1) and small-tailed Han sheep, DM (F2) (10). The low-milk-yield population consisted of small-tailed Han sheep (STHS) (5). The sequence data of the 60 remaining sheep were downloaded from (https://www.ncbi.nlm.nih.gov/sra, accessed on 15 March 2022). These sheep were from three high-milk-yield breeds, East Friesian (EFR) (10), Awassi (AWS) (2), and Dairy Meade (9), and four low-milk-yield breeds: Hu (HS) (14), Fin (FS) (9), Suffolk (SFK) (10), and Churra (CS) (6) (Table 1).2.2. Whole-Genome-Resequencing AnalysisWe sequenced 41 sheep genomes at an average depth-of-coverage of 10X, which contained 4 sheep breeds. To this end, genomic DNA was extracted from ear tissue using the Wizard® Genomic DNA purification kit (Promega, A1125, Madison, WI, USA). The A260/A280 ratio using Nano-Drop ND-2000 (Thermo Fisher Scientific, MA, USA) was used to check the quality and integrity of the extracted DNA and agarose-gel electrophoresis was used to check its quality. Good-quality DNA from each collected sample was sequenced on the MGI-SEQ2000 platform from the Beijing Compass Biotechnology Company (Beijing, China). The satisfactory sequence data obtained were used to characterize individual genomes at a minimum of 10X depth of coverage, with more than 30 G of raw data from each sample. After quality control, data were used to perform genome mapping and SNP calling. Resequencing data from a further 60 sheep, downloaded from NCBI, were analyzed by Plink parameters on Linux (San Francisco, CA, USA) (Version V1.90). The details are listed in Section 2.3. All high-quality double-trimmed read pairs were aligned against the reference assembly Oar v.4.0 genome [29] (https://www.ncbi.nlm.nih.gov/assembly/GCA_000298735.2, accessed on 15 March 2022) using BWA software (https://sourceforge.net/projects/bio-bwa/files/, accessed on 15 March 2022) (version: 0.7.12). Paired-end reads that mapped to exactly the same position on the reference genome were deleted by the Mark Duplicates in Picard (picard-tools-1.56, at http://picard.sourceforge.net, accessed on 15 March 2022). Additional realignment of indels and SNPs was performed by using the Genome Analysis Toolkit (GATKV4.0) (https://gatk.broadinstitute.org/hc/en-us, accessed on 15 March 2022) [30] and sequence alignment (SAM tools) [31]. The GATK was used to identify SNP variation in each individual sample. The SNPs that did not meet the following criteria were excluded: (1) SNP call rate > 99.6%; (2) minor-allele frequency > 0.01; (3) missing rate is lower than 0.1; and (4) all loci followed the Hardy–Weinberg rule; (5) linkage-disequilibrium loci were excluded (R2 < 2). All SNPs were annotated using the Bio Mart 4.0 (https://asia.ensembl.org/info/data/biomart/index.html, accessed on 15 March 2022) [32] based on the gene-reference genome provided by the Oar v.4.0 genome from NCBI.2.3. Population Structures and Phylogenetic AnalysisPopulation structures among all sheep breeds were investigated using a total of 59,864,820 high-quality SNPs. The VCF (Visual Component Framework) files were transformed in the PLINK format by using VCF Tools (https://vcftools.sourceforge.net/, accessed on 15 March 2022) [33], and the SNP filtering was undertaking under the following conditions: minor-allele frequency (MAF) greater than 0.05, call-out rate higher than 0.9, missing-genotype rate > 0.05, and Hardy–Weinberg rules > 1 × 106. Pruning was carried out within windows of 50 SNPs and in 5 steps, using the indep-pairwise 50 5 2 parameters in PLINK [34] (plink --file qc --indep 50 5 2 --chr-set 27 --recode --out qc_prune) to obtain linkage-disequilibrium SNPs. This provided a total of 4,751,642 unlinked SNPs for PCA, ADMIXTURE, and neighbor-joining-tree-construction analysis. The PCA was conducted using the GCTA software (https://yanglab.westlake.edu.cn/software/gcta/, accessed on 15 March 2022) [35]; the first two components were plotted using with the R program gg plot package (https://www.r-project.org/, accessed on 15 April 2022) (version 4.1.0). Individual ancestry origins were predicted using replicates in the ADMIXTURE (http://software.genetics.ucla.edu/admixture/download.html, accessed on 15 April 2022) (version 1.3.0) [36], and assuming 2 to 4 ancestral populations (K), with the smallest CV error as the correct result. Neighbor-joining tree was constructed using the PHYLIP v3.69 (https://evolution.genetics.washington.edu/phylip.html, accessed on 15 April 2022) [37] based on the pairwise genetic distance matrix revealed by the PLINKv1.9 (plink1 --sheep --file prune --cluster --distance-matrix –out), and then MEGA7 (https://www.megasoftware.net/, accessed on 15 April 2022) [38] was used to construct the phylogenetic trees.2.4. Genome-Wide Signal-Selection Scan and Gene AnnotationWe used all SNPs that passed quality control to detect signatures of selection in high- and low-milk-yield sheep within 101 sheep genomes. Genome-wide signal-selection analysis was performed by using FST fixation indices (population-differentiation value), (θπ), the nucleotide-diversity ratio (θπ-HMY/θπ-LMY), HMY (high milk yield), LMY (low milk yield), and the transformed heterozygosity score (ZHP). The window-based ZHP method was calculated by the formula ZHP = (Hp − μHp)/σHp, where μ is the overall average signal value of heterozygosity and σ is the standard deviation of all windows of each group. High-milk-yield group included 57 samples from five breeds (Dairy Meade sheep = 20, DM (F1) = 15, DM (F2) = 10, East Friesian sheep = 10, and Awassi sheep = 2), while low-milk-yield group included 44 individuals across five breeds (STHS = 5, Hu sheep = 14, Fin sheep = 9, Churra sheep = 6 and Suffolk sheep = 10). The FST fixation, ZHp [39], and log2θπ (high/low), were calculated within 100-kb sliding windows and 10-kb steps to obtain overlapping regions. The variation regions within the highest 5% of all three statistics were considered as candidate selections, and then candidate genes were annotated by a genomic-database search and annotating engine, Bio Mart (https://asia.ensembl.org/info/data/biomart/index.html, accessed on 15 March 2022) [40].2.5. Gene Ontology and Kyoto Encyclopedia of Genes and GenomesFunctional enrichments for Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using g: Profiler (https://biit.cs.ut.ee/gprofiler/gost, accessed on 1 June 2022) for all selected genes. The enrichment significance was assessed using a condition for g: SCS ‘Set Counts and Sizes’, and the p value was set at <0.05.2.6. Validation of RT-qPCR ExperimentWe randomly selected 11 Lacaune sheep in from a farm in Belgium (https://lesfauveslaineux.wordpress.com/le-troupeau/, accessed on 1 September 2022) to determine their genes’ expression levels. These were frequently selected in our research, since Lacaune sheep are among the most famous and heavily produced dairy sheep in France, and imported into Belgium. This farm efficiently recorded the milk-production data from different milk lactations and information on newly lambing ewes. Fresh milk was sampled from 11 newly lambed ewes. From each ewe, 50 mL of milk was sampled in non-RNA se tubs. All these steps complied with the requirements of European Animal Welfare Committee. The milk was stored at 4 °C during transport from farm to laboratory and processed immediately for somatic cell isolation. Subsequently, 50 uL of 0.5 M EDTA was added for each 50-mL milk sample. The samples were then centrifuged for 10 min at 2000× g, after which the supernatant containing cream and skin milk was removed, followed by washing in 10 mL PBS. Samples were then centrifuged a second time, the maximum amount of supernatant was removed, and the lysed samples were stored at −80 °C. The RNA was extracted according to the procedure of the isolation of RNA from non-fibrous tissue (Promega Inc., Madison, WI, USA), after which concentration and A260/A280 of RNA were tested; the standards were concentration >10 ng/ul and 2.2 > A260/A280 > 1.8. Satisfactory RNA was used to perform reverse transcription in accordance with Protocole reverse transcription (Promega Inc., Madison, WI, USA). The conditions of reverse transcription were as follows: 25 °C, 5 min; 42 °C, 60 min; 70 °C, 15 min, 4 °C, 20 min; 4 cycles in total. The cDNA was stored at −20 °C. Quantitative real-time PCR was performed using DNA, premiers were designed to amplify target genes and reference genes, reference genes were selected from NCBI database, which were similarly expressed in all tissues, the genes’ premier pairs for RT-qPCR were designed by Eurogentec CO., (Brussel, Belgium), and RT-qPCR analyses were performed using Rotor Gene 6000 system (Corbett) and SYBR green (Gembloux, Belgium). Each amplification reaction contained 2 μL of cDNA and 18 μL of SYBR master mix (Thermo Fisher Scientific, MA, USA), including 500 nM of primers at a final volume of 20 μL. Reactions were performed as follows: 95 °C for 5 s, 60 °C for 30 s, 72 °C for 45 s, for 40 cycles in total, with data collection at the annealing step.2.7. The Relationship between Gene Expression and Milk ProductionCandidate genes from RT-qPCR experiment were used to perform a linear correlation analysis between genes’ relative expression values and milk production. The milk-production recordings were from Belgian farm with the same Lacaune sheep as those in 2022 (Data S4), as well as the production data from 2015 to 2021 (Data S5). We used the BLUPF90+ (http://nce.ads.uga.edu/html/projects/programs/, accessed on 15 March 2023) program to obtain estimated values (similar to estimated breeding values) of milk-production traits of each ewe from 2015 to 2021. Usually, higher estimated values meant the animal had stronger production traits, so we used these values to conduct a linear analysis with ΔCT gene values and actual milk production from 2022, as well as the estimated values from period of 2015–2021 to show their relationships. Animal model is needed in BLUPf90+. In our research, the animal model was as follows: yijk = Ai + Sj + βxijk + eijk, where yijk is observed milk production, Ai is fixed effect of lactation days, lactation number, birth year, and milking times, Sj is random effect of animals, βxijk is regression coefficient of variance of 1.0, and eijk is residual effect. The estimations were given after data analysis using BLUPF90+ program.3. Results3.1. Genomic Variants and Principal Component AnalysisWe combined the genome sequences of 101 high- and low-production dairy sheep. The genomes were mapped and the SNPs were called using PLINK. After the joint calling and quality control, we detected 4,751,642 SNPs in total. To study the population stratification, three different approaches were employed, of which PCA is the first to be discussed here. The intention behind the use of PCA was to reduce the dimensionality of the genomic relationship matrix so that individuals (and breeds) were separated along different principal components (PCs). The first three principal components were able to explain most of the variation: PC1 explained about 50.26%, PC2 explained about 23.51%, and PC3 explained about 14.95%. Principal component 1 (PC1) differentiated the breeds in this study into three main clusters (Figure 1A): Cluster1 (high-yield breeds, EFR, DM, DMF1, and DMF2); Cluster2 (Mongolian low-yield breeds, HS, STHS, CS, and AWS) and Cluster3 (low-yield breeds, FS and SFK). However, overlap was observed among the DM sheep, DMF1 and DMF2, within Cluster 1. The reason for this may have been that these sheep have extremely close genetic relations. The PCA separated the different sheep breeds in this study, with SFK the most clearly separated due to its longer distance from and non-blood relations with the other sheep. In the PCA plots, the high-yield breeds were closer to the Mongolian sheep than to the low-yield sheep, and in the high-yield cluster, the EFR sheep were clearly separated from the DM sheep. The separation of the clusters corresponded to the geographical origins and blood mixes of these sheep breeds, and may have also revealed their genetic distance.3.2. Neighbor-Joining-Tree AnalysisIn this study, the phylogenetic neighbor-joining tree divided the high- and low-production groups into nine different clusters (Figure 1B). The nine main branches were as follows: DM, DMF1, DMF2, EFR, CS, FS, HS, STHS, and SFK. According to the tree, we observed that the DM and DMF2 sheep branches were close to each other and that some of the DM, DMF1, and DMF2 sheep were mixed together, indicating their genetic similarity and probable sharing of a single ancestral line. The DM sheep is the hybrid offspring of EFR breeds and New Zealand sheep [41], so it was the second-nearest to the EFR sheep. Our phylogenetic results showed two breeds (DMF1 and DMF2) in the different sub-branches, which might have been due to the hybridization between the DM breeds and the STHS breeds. For the HS and STHS sheep, all the Mongolian sheep had similar genetic information, and their branches were close to each other. Similar to the PCA, there was a close genetic distance between the FS and SFK sheep, since there are shorter geographical distances between them. In the NJ-tree picture, the Awassi breed was not clearly separated, possibly due to its small numbers. It should be noted that the results of the phylogenetic analysis were reliable due to the large number of SNPs from the 10 sheep breeds used in this study.3.3. Ancestor AnalysisTo further explore the population structure among individuals from the different breeds in this study, a model-based hierarchical clustering analysis was undertaken for K-values of 2–7 (with K the user-defined number of biological ancestral populations). The cross-validation (CV) estimates revealed the K-value of 3 to be the best fit for the nine populations; it demonstrated the lowest CV error among all the K values. The bar plot of the ADMIXTURE results from the K-values of 2 and 3 is presented in Figure 1C. The analysis with K = 2 separated the high-milk-yield sheep breeds from the low-yield breeds. When K = 3, the FS and SFK were separated from the other breeds in the low-yield group. The DM and EFR sheep had the highest milk yield among all the populations, the Mongolian CHS, HS, and STHS sheep had average milk production [26], and the FS and SFK sheep produced the least milk. It is known that FS and SFK sheep are clearly separated from the other two sheep populations, and this was validated by the results obtained from the PCA analysis and the admixture analysis.3.4. Signal-Selection Analysis and Gene OntologyWe scanned the genomes of 101 high- and low-milk-yield sheep for signals of positive selection with different traits. To achieve this, we calculated three complementary statistics along the sheep reference genome Oar v.4.0 (https://www.ncbi.nlm.nih.gov/assembly/GCA_000298735.2, accessed on 15 March 2022) using 100-kb-long sliding windows and 15-kb step sizes. The first statistic was the population-differentiation index FST, for identifying genomic regions with different allelic frequencies between high and low groups. The second statistic measured the differences in nucleotide diversity between the two groups (θπ(high/low)), with high or low θπ values indicating positive selection of different milk-production sheep, respectively. The third measure was also used to calculate the selected candidate SNPS using the index of the transformed heterozygosity score (ZHP) (Figure 2). With regard to the limiting of the false-positive identifications, we considered the 5% most heavily overlapping regions from all three scans, which provided a total of 995 overlapping SNP regions and 553 protein-coding genes and small RNA. The selection candidates identified in the high- and low-yield groups are provided in Data S1. Most of the protein-coding genes were significantly enriched using the functional gene ontology and KEGG categories related to the protein binding, RNA binding, molecular transport, and cytokine-receptor activity (p = 9.23 × 10−3) (Figure 3, Data S2). Importantly, through the gene annotation, the three highest 5% genomic regions providing the strongest signatures of selection containing FCGR3A, CTSK, CTSS, ARNT, GHR, and SLC29A4 were selected as the strongest annotation genes by comparing the high- and low-milk-yield groups.3.5. RT-qPCR ResultsWe chose the four genes that were most frequently selected by the signature-selection analysis shown in Figure 2, which were FCGR3A, CTSK, CTSS, and ARNT, according to the NCBI database and previous studies. We chose GAPDH gene as reference gene to perform RT-qPCR validation. The premier sequence is listed in Table 2. All the gene-expression values were derived from 11 ewes; the CT and ΔCT values of each gene are listed in Data S3, with ΔCT values meaning the relative expression of each gene.3.6. The Relationship between Gene Expression and Milk ProductionThe milk-production data were sourced from the Belgian farm. They covered the average daily milk production from four months in 2022: September, October, November, and December (Data S4). A linear analysis of the ΔCT values of each gene was performed using the daily average milk production from 2022 and values estimated using the program of BLUPf90+ (Data S5), and the correlations between daily milk production and the estimated values are also shown in Figure 4. The results showed that the FCGR3A gene had significant correlations with milk production and estimate values. The higher ΔCT values meant lower expression values during the lactation period. It was revealed that the FCGR3A gene had a significantly negative relationship with milk production and the estimate values (R2 = 0.8452, R2 = 0.7382); furthermore, as predicted, the relationship between actual milk production in 2022 and the estimate value was positive (R2 = 0.6988. The CTSK, CTSS, and ARNT did not show significant differences between their expression values, the milk production values, and the estimate values, but the correlation coefficients between actual milk production in 2022 and the estimate values for the 2015–2021 milk production were positive (R2 = 0.6988, R2 = 0.628, R2 = 0.628).4. DiscussionIn this study, we sequenced the genomes of 101 sheep with high and low levels of milk production. Using a neighbor-joining-tree analysis, we divided these sheep into 10 groups according to their genetic relationships and background, and then, using a principal component analysis and a structure analysis, these 10 groups were divided into three clusters: high-milk-yield sheep (EFR, DM, DMF1, DMF2), Mongolian low-milk-yield sheep (HS, CS, AWS, STHS), and low-milk-yield sheep (SFK and FS). The scanning of the genomes of high- and low-yield sheep breeds revealed that the FCGR3A ARNT, CTSK, CTSS, and GHR genes were the strongest candidates. According to the RT-qPCR experiment and the analysis of the relationship between these candidate genes’ ΔCT values and milk production, the FCGR3A gene had a significant relation with Lacune sheep milk production. All of these genes have also been reported to be highly expressed during the lactation period in cattle [42] and buffalo, reflecting the similar biological functions of these genes when they are expressed in lactating mammary glands. The GO and KEGG analyses showed that most of these genes were significantly enriched for mammary-gland-specific GO terms (cytokine-receptor activity and protein binding), as well as establishing the cellular functions and protease bindings. These genes were found to have high and low levels of milk production, milk protein, milk fat, milk lactose, cheese traits, somatic cell counts, etc. For some of these traits, it is not easy to find a direct relationship with milk quantity. This was probably the reason why we did not find a significant correlation coefficient between the CTSK, CTSS, or ARNT expression values and milk production. Furthermore, as we found in many publications, some of the most significant genes in our research play a crucial role in sheep-mastitis traits and in the immune-response process, which improves milk production. Sheep-mammary-gland mastitis is also a common source of economic losses on sheep farms, so it is also important for us to focus on some of the genes related to mammary-disease resistance.Candidate Genes Potentially Associated with Some Milk TraitsThe FCGR3A (Fc fragment of IgG, low-affinity IIIa, receptor) gene is considered a novel and promising candidate for relieving stress, inflammation, and disease [43], as well as dairy-cattle mastitis. The binding of FCGRs to the Fc region of immunoglobulins mediates a variety of immune functions, such as antigen presentation, the clearance of immune complexes, the phagocytosis of pathogens, and cytokine production [44], they work together to resist viral injection to improve udder health and, thus, increase the production of milk. It is reasonable to suggest that lactation periods cause the increase in expression of a large number of genes, resulting in improvements in performance, as mastitis or Staphylococcus aureus infections often occur during lactation [45,46]. With regard to CTSS (Cathepsin S)-encoding protease, Sodhi et al. (2021) found a higher expression of most of the cathepsin genes (CTSS, CTSD, and CTSK) during the mid-to-late lactation stages, emphasizing their potential roles in milk synthesis, since the expressions of most of the proteases were higher during peak lactation [47]. The higher expression of cathepsin-encoding genes during late lactation stages could be attributed to the fact that CTSS plays a crucial role in mammary-gland involution [48]. Previous studies found the role of the CTSK gene in influencing cheese traits, as it is a protease-encoding casein that increases milk protein [49]. Similar results were found when evaluating the expression patterns of important protease-pathway-associated CTSK genes derived from different lactation stages in Sahiwal cows and Murrah buffalo. In both breeds, the RNA-expression levels of these genes were higher in the late lactation stages than in the early lactation stages. Lactation induces bone loss in order to provide sufficient calcium in milk, and to prevent this from taking place, the CTSK gene elevates its expression to increase osteocyte numbers, in order to maintain the balance of bone and milk calcium [50,51]. This proves that the CTSK gene plays an important role in milk-calcium traits. Furthermore, a differential expression analysis of Churra and Assaf sheep allowed us to notice some genes that were significantly and differentially expressed between the two breeds. These genes were mainly associated with protein-protease activity. Furthermore CTSK was differentially expressed and was selected as a candidate gene associated with cheese traits in this study [49]. These findings confirmed previous results, which highlighted the importance of the expression of genes encoding for some proteases in sheep milk. The aryl hydrocarbon receptor nuclear translocator (ARNT) can interact with the AHR aryl hydrocarbon receptor). The AHR is restricted to the cytoplasm in its unbound state [52,53]. Once activated, the AHR is combined with the nucleus of the ARNT and forms an active complex with the AHR nuclear translocator (ARNT) to alter the expressions of target genes [54]. It was reported that an association between AHR activation and ARNT causes changes in milk production [55]. More specifically, pregnant mice exposed to TCDD (an ARNT agonist) in vivo produced lower levels of the milk proteins, β-casein and whey protein [56]. The GHR is a regulator of developing growth and, as a growth hormone, it has important effects on carbohydrate, protein, and lipid metabolism. In cattle, mutations in GHR have been associated with milk yield and composition in Ayrshire, Holstein, and Jersey cattle. Dettori et al. (2018) reported that variations of the ovine GHR gene might affect milk-quality traits in Sarda sheep [57]. The rs55631463 GHR SNP genotype affects milk fat and protein yield, and the rs411154235 SNP is associated with lactose content in milk, as it encourages the transformation of glucose into glycogen to help glycogen cellular deposits [58].In our study, except for FCGR3A, the strongest gene candidates, CTSK, CTSS, ARNT, were not found to be significantly relevant to milk-production quantity. This was potentially due to the small number of Lacune sheep selected to validate their function. Since our objective was to detect milk-yield genes, not milk-composition genes, it was difficult to directly find the relationships between these genes, which may participate in milk-protein or milk-fat synthesis. Thus, further measures need to be applied in the future to prove that these genes are correlated with other sheep-milk traits. In particular, Dairy Meade should be tested, since this breed, or crosses in which it is involved, comprised a high proportion of the high-milk-yield group.5. ConclusionsIn conclusion, our research is the first attempt to report a genome-signature-selection scan revealing genes that are associated with high- and low-milk-yield sheep breeds by using whole genome sequencing. Some of the most significant candidate genes associated with milk yield were identified through a combination of genome-signature-selection scanning and RT-qPCR experimentation. In the high- and low-yield groups, 553 genes were detected that were enriched with significant protein-receptors-combing pathways. Under selective pressure, we selected the four genes, FCGR3A, CTSS, CTSK, ARNT, that were most likely to be related to milk-production traits, in order to perform a correlations analysis, in which FCGR3A was found to have a significant relationship with daily milk yield, as it participates in the process of immune reaction. Furthermore, since higher gene-expression values mean severer mastitis, they are negatively correlated with milk production. These results were supported by the correlation between the FCGR3A ΔCT values and the estimated values of the corresponding ewes. Therefore, these results could provide a better genetic perspective on the phenotypic differences between different-milk-yield groups for similar studies. Our findings provide an insight into the dynamic characterization of sheep-mammary-gland gene expression, and the identified candidate genes can provide valuable information for future functional characterization, as well as contributing to a better understanding of the genetic mechanisms underlying the milk-production traits in sheep.

animals : an open access journal from mdpi

10.3390/ani11113213

PMC8614306

Perilla is an edible oil crop containing high levels of polyunsaturated fatty acids (PUFA), such as alpha-linolenic acid (ALA) and omega-3. The omega-3 helps in mitigating the risk of cardiovascular disease in humans. For industrial use, perilla seed is extracted for virgin oil, which is generally achieved by mechanical screw pressing. This process generates perilla cake that contains a fat content around 9–10%. In this study, we examined the effect of the supplementation of perilla cake in the pig diet on productive performance, pig carcass characteristics, meat quality, and fatty acid composition in fat tissue, and meat. The pig performance improved in terms of average daily gain after supplementation with perilla cake; however, the pig carcass and meat quality was unchanged. Moreover, the perilla cake supplement increased PUFA fatty acids and reduced the proportion of saturated fatty acids and unsaturated fatty acids in pork. The overall outcome of this study provides an alternative source of novel raw material for functional feed additives in livestock production.

The objective of this study was to determine the effect of perilla cake (PC) supplementation in a growing pig diet on overall growing performance, meat quality, and fatty acid profile. A total of 24 barrow grower crossbred pigs (Large White × Landrace) × Duroc with an initial average body weight of 26.33 kg were fed with a basal diet supplemented with PC at 0%, 5%, and 10% in (PC0, PC5, and PC10, respectively) for 12 weeks. At the end of the experimental period, pigs were slaughtered to determine carcass traits and meat quality. Back fat, abdominal fat, and longissimus dorsi (LD) muscle were collected to investigate fatty acid composition. The results show that the average daily gain (ADG) in the PC10 significantly increased. However, PC supplementation did not influence carcass traits and meat quality except the color as described by lightness (L*). Dietary PC supplementation significantly increased the α-linolenic acid (ALA, C18:3 cis-9, 12, 15), whereas n6/n3 ratio decreased significantly in all tissues investigated. Thus, it can be concluded that the supplementation of PC in growing pig diet is a potential way to increase the fatty acid composition to that required for healthier meat.

1. IntroductionPerilla (Perilla frutescens L.) is an annual herbaceous plant in the Lamiaceae family. It has been cultivated widely in China, India, Japan, Korea, Thailand, and in many other Asian countries as a source of edible oil, protein, and fiber [1] with various biological properties such as antiviral [2,3], anti-inflammatory [4], and antioxidant [2,5] effects. Perilla oil consists of 90.60% total unsaturated fatty acids, 17.90% monounsaturated fatty acid, and 72.70% polyunsaturated fatty acids [6]. The latter is in the form of omega-3 polyunsaturated fatty acids (PUFAs), specifically α-linolenic acids (ALA) 55.00–64.00% [7,8,9] and omega-6 and omega-9 fatty acids [7]. Presently, the production of perilla seed in Thailand accounts for roughly 272 tons/year of refined oil, and meal (ca. 60%) with low fat content, around 1–2% [10,11]. In addition, the screw press method is one of the popular techniques for extracting perilla oil, which yields perilla cake (PC) as a biomass. This cold pressing method, nonetheless, yields as much as ca. 8–14% of the available oil in the cake [12]. Souphannavong et al. [11] described that the PC contained crude protein (CP) 31.54%, ether extract (EE) 10.52%, and more importantly, high ALA ca. 55.97%. Thus, considering the functional ingredients present, it is interesting to use this biomass in animal diets to improve the overall quality of the livestock.Pork is the most widely eaten meat in the world, but typical feeding practices lead to poor meat quality, as defined by a high omega-6 (n6) to omega-3 (n3) fatty acid ratio and low n3 fatty acid [13]. The conventional farmed pork also contains saturated fatty acids [14], which have adverse effects on human health. The ALA are essential for the normal growth and development of humans and animals [15,16], involved in the evolution of brain activity and the nervous system, and play an important role in the prevention and treatment of cardiovascular diseases, inflammatory diseases, and cancer [17,18]. Consumption of n3 PUFA from terrestrial animal products is mainly limited to the intake of ALA [19]. The main sources of long-chain n3 PUFA are marine fish, seafood, and fish oil [20]. Sadly, over 66.8% of the world’s adult population has very low intake of n3 PUFA as they are not able to access seafood [20]. One way to achieve the recommended daily intake of the n3 PUFA omega-3 fatty acids is to consume meat and meat products, and research is currently searching for the functional ingredients that can be supplemented in feed, thereby improving the n3 fatty acid composition in the meat [13,21,22]. For this reason, PC with high ALA content was a possible candidate for feed supplementation to improve the level of n3 PUFA in pigs. In previous work, Cui et al. [23] reported that chickens fed with perilla oil diets exhibited higher contents of α-linolenic acid (C18:3n3), DHA (22:6n3), polyunsaturated fatty acids, and n3 fatty acids, as well as a lower n6/n3 ratio. Oh et al. [24] also reported that feeding broilers with 2% perilla seed meal in their diet could improve growth performance, meat quality, and the fatty acid composition of thigh meat; specifically, the omega-3 fatty acid (7.55%) was higher than the control diet (6.64%). Hadi and Sudiyano [25] described that adding perilla seed meal to the diet of ducks increased average daily gain and omega-3 fatty acids (0.94%). Furthermore, Peiretti et al. [21] reported that perilla seeds supplementation in the rabbit diet increased α-linoleic acid and polyunsaturated fatty acid contents in the longissimus dorsi (LD) muscle. There is currently no study on the efficiency of supplementation of PC in pig diet in improving n3 PUFA. Consequently, the objective of this study is to investigate the effect of PC supplementation in the growing pig diet on the productive performance, meat quality, and fatty acids profile.2. Materials and MethodsThis study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Agricultural Animals in Research and Teaching. The experimental protocols were reviewed and approved by the Animal Care and Use Committee of Chiang Mai University (2560/AG-0001) prior to the experiment.2.1. Animals and ManagementThis experiment was performed at the Mea Hia Agriculture Resource Demonstrative and Training Center, Faculty of Agriculture, Chiang Mai University, Chiang Mai, Thailand. A total of 24 barrows grower crossbred pigs (Large White × Landrace) × Duroc with an initial average body weight of 26.33 kg were used. The pigs were randomly allotted into three groups (n = 8 per group) and each individual pig was penned in an area of 2.0 m2 on the concrete floor, with drinking water and a feeding trough provided [26]. The granulated feed and water were given ad libitum. Prior to the experiment, all pigs were dewormed and vaccinated.2.2. DietsThe three diets were based on maize, broken rice, rice bran, fish meal, and soybean meal, supplemented with a vitamin–mineral premix. The perilla cake was supplemented as 5% and 10% in feed formula. The diets included perilla cake supplementation in basal diet at 0% (PC 0 or control), 5% (PC 5), and 10% (PC 10) (Table 1), formulated to contain equal concentrations of metabolizable energy, crude protein (CP), minerals, and vitamins to meet the requirements for growing pigs according to the National Research Council (NRC) [27]. In addition, the measured fatty acid compositions of the three diet formulars are presented in Table 2.2.3. Sample CollectionThe body weight of the individual pigs was taken at the end of weeks 0, 4, 8, and 12. In addition, feed intake was recorded daily to calculate the productive performance, including average daily feed intake (ADFI), average daily gain (ADG), and feed conversion ratio (FCR). At the end of the experiment, all animals were off feed, with access to water, for 12 h prior to slaughter, and were transported to the Huay Kaew Slaughterhouse, Department of Livestock Development, Chiang Mai, Thailand within 20 min. After a minimum of 6 h resting time, pigs were euthanized via electrical stunning and exsanguination. All experimental procedures were carried out following the good manufacturing practices of the abattoir Thai Agricultural Standard TAS 9004-2004 [28]. Individual hot carcass weights were recorded. The longissimus dorsi muscle, abdominal fat, and back fat were collected from the right half-carcass, stored in a plastic bag at −20 °C for chemical analysis.2.4. Assessment of Carcass TraitsAfter chilling at 4 °C for 24 h post-mortem, carcasses were weighed for chill carcass weight. Carcasses were dressed according to Thai and USDA styles into four lean cuts (picnic, boston, loin, and ham) [29]. The carcass dressing percentages were calculated by dividing the carcass weight by the live weight, obtained after fasting. The loin eye area was measured by tracking the surface area of the 10th rib of the LD muscle according to Santos et al. [30]. Point counting was over a 1 cm2 plastic grid (PCGP 1 cm2) made from a graph paper sheet with a transparent plastic sheet as a copy. The sum of the squares was performed to obtain the total area. In addition, the backfat thickness was measured at the 11th rib (including the skin) with a vernier caliper in cm of the LD muscle, according to Álvarez-Rodríguez and Teixeira [31].2.5. Assessment of Meat QualityThe LD muscle was cut into 2.54 cm-thick slices [29]. Then, all samples were vacuum-packaged and stored at −20 °C until further analysis. The LD muscle underwent proximate analysis for percentages of moisture, ash, ether extract, and crude protein according to the association of official analytical collaboration (AOAC) [32]. The average pH value was determined in the LD muscle at 45 min and 24 h postmortem using a Testo 205—pH electrode/NTC temperature measuring instrument (Testo, Lenzkirch, Germany). Meat color was determined at 48 h postmortem on the surface area of the LD muscle after blooming for 1 h at 4 °C and fluorescent lighting at 2600 lumen [33]. The color values including L* (lightness), a* (redness), and b* (yellowness) were taken using a Minolta Chroma Meter Model CR-400 (Minolta Camera, Osaka, Japan). Then, the drip loss of the LD muscle was measured by preparing approximately 30 g of sample in a plastic bag and storing it at 4 °C for 24 h [34]. The water extruded from the sample was removed and the sample was weighed.2.6. Fatty Acid (FA) AnalysisThe diets, backfat, abdominal fat, and LD of growing pigs were analyzed and the fatty acid (FA) profile was analyzed according to the method of Chaiwang et al. [29]. Lipids were extracted from diets, backfat, abdominal fat, and LD by the Soxhlet extraction (method 920.39). Fatty acid methyl esters were prepared as described by Morrisson and Smith [35]. Gas chromatography–flame ionization detector (GC–FID) analysis was accomplished using the Shimadzu model GC-2030 (Kyoto, Japan) equipped with a 0.25 mm × 100 m × 0.25 µm wall-coated fused wax capillary column (RT-2560, RESTEK, Bellefonte, PA, USA). Helium was used as the carrier gas. Injector temperatures were held at 250 °C. Oven temperature programming was increased from 100 °C, held for 4 min, increased from 100 to 240 °C at a rate of 3 °C /min, and then held at 240 °C for 20 min. Injector volume was 1 µL, and the flame ionization detector temperature was 250 °C. Chromatograms were processed using the Lab Solution (Shimadzu, Kyoto, Japan). Identification was accomplished by comparing the retention time of peaks from samples with those of fatty acid methyl ester (FAME) standard mixtures (Food Industry Fame Mix, RESTEK, Bellefonte, PA, USA).2.7. Statistical AnalysesThe statistical analyses were performed using the SPSS software package (version 23.0 for Window, SPSS Inc., Chicago, IL, USA). Analysis of variance (ANOVA) was used to evaluate the effects of the PC supplementation in the diet on productive performance, carcass trait, meat quality, and fatty acid composition. Differences in means among treatment groups were determined using Duncan’s multiple range test, with a p < 0.05 indicating statistical significance.3. Results3.1. Chemical Composition and Fatty Acid Profile of PCThe chemical composition and fatty acid contents of PC are shown in Table 1 and Table 2. In the present study, PC was rich in crude protein (31.54%) and fat (10.52%), while it had a relatively high crude fiber content (24.43%). The addition of various levels of PC in the growing pig diet decreased linoleic acid (LA; C18:2 cis-9, 12) and significantly increased α-linolenic acid or ALA (C18:3 cis-9, 12, 15) content, while the n6/n3 ratio decreased with increasing PC inclusion levels (p < 0.05) (Table 2).3.2. Productive PerformanceThe effect of PC supplementation in the growing pig diet on productive performance is presented in Table 3. The results show no significant difference among the treatments on the final weight, even though pigs in the PC10 group had a higher final weight than they did in other groups (p > 0.05). During weeks 0–4 of the experimental period, PC10 was the significantly highest FCR compared with PC0 and PC5 (2.81 vs. 2.56 and 2.50, respectively). In weeks 5–8 of the experiment, we found that ADFI of the PC10 group was significantly higher than the other groups during this period, and for the overall period (p < 0.05). In the last period of the experimental study (weeks 9–12), pigs fed with PC10 exhibited significantly higher ADG than the other groups (p < 0.05) However, in the overall experimental period, the PC10 group exhibited significantly higher ADFI and ADG (p < 0.05).3.3. Carcass Traits and Meat QualityThe supplementation of PC in the growing pig diet did not affect pig carcass traits, including slaughter weight, carcass percentage, hot carcass weight, chill carcass weight, carcass length, and backfat thickness (p > 0.05) (Table 4). The meat quality, as described in terms of drip loss percentage, pH at 45 min and pH at 24 h, was not affected by the PC supplementation, except in terms of the meat color, especially the lightness (L*). The PC10 group expressed the lowest L* value (48.42) compared with the other groups. In addition, we found that the PC supplementation in the growing pig diet did not influence the chemical composition of LD muscle, including moisture, crude protein, ether extract, and ash (p > 0.05) (Table 5).3.4. Fatty Acid Profiles in Backfat, Abdominal Fat, and Longisimus dorsiTo explore the effect of the supplementation of PC in the growing pig diet on the FA composition in pigs, we collected the tissues at three positions viz. back fat, abdominal fat, and LD muscle, and determined the FA profiles. We found that the supplementation of PC did not affect the fatty acids in the saturated fatty acid (SFA) and monounsaturated fatty acid (MUFA) compositions (p > 0.05) in the backfat (Table 6). However, the LA (C18:2 cis-9, 12) content in backfat significantly increased when the PC supplementation was at 10% (PC10). Moreover, PC supplementation correlated with a significantly higher ALA (C18:3 cis-9, 12, 15) compared with PC0 in backfat (p < 0.05). Similarly to the backfat, PC supplementation did not influence the SFA and MUFA compositions in the abdominal fat (Table 7). On the other hand, we found that the ΣSFA was significantly lower in the PC10 group than it was in the PC0 group. In addition, the ALA was increasingly raised with the PC supplementation levels (p < 0.05). For the LD fat, we found that the PC supplementation in the diet significantly decreased the SFA composition (p < 0.05) (Table 8), including palmitic acid (C16:0), stearic acid (C18:0), behenic acid (C22:0), and tricosylic acid (C23:0). Conversely, PC supplementation increased the unsaturated fatty acid composition significantly (p < 0.05), especially LA and ALA. Furthermore, the ΣSFA in LD fat was significantly lower in PC5 and PC10 compared with PC0. In addition, this study indicates that the supplementation of PC in the diet significantly increased ΣMUFA and ΣPUFA in LD fat (p < 0.05).The ratios of ΣMUFA/ΣSFA and ΣPUFA/ΣSFA are illustrated in Figure 1. The supplementation of PC in the diet did not affect ΣMUFA/ΣSFA and ΣPUFA/ΣSFA in backfat. The ΣPUFA/ΣSFA ratio was significantly higher in the PC10 group in abdominal fat and LD fat (p < 0.05). Moreover, the n6/n3 ratio was lower (p < 0.05) in all tissues when supplementation was used in the diet (Figure 2).4. DiscussionGenerally, perilla is used for oil production as a rich source of omega-3 polyunsaturated fatty acids, specifically alpha-linolenic acid [7]. Perilla cake is a by-product of the perilla seed oil cold pressing process, and contains high protein content (31.54%), natural fatty acids, and dietary fiber [11]. In this study, we found that the EE content in PC was 10.52% which was in agreement with the findings of Shrikanta Rao [12], who reported that mechanical screw presses are relatively inefficient for edible oil recovery, leaving ca. 8–14% of the available oil in the cake. Additionally, the EE in the PC was higher than that of perilla meal, which was reported to contain ca. 1.08% of EE [38]. Souphannavong et al. [11] found that total PUFA was as high as 70.91%, which mostly consisted of the ALA (55.97%). Presently, the concern for food intake relating to health issues has led to the demand for functional foods such as diets rich in omega-3 fatty acids [39]. When pigs are fed diets with n3 fatty acids, pork and pork products could be recognized as a functional food with new health-promoting properties [40]. However, the use of PC in the diet should be considered at suitable supplement levels, because PC possesses relatively high fiber (24.43%) even if there is a high CP. Arjin et al. [26] explained that the high level of fiber in the diet leads to a low digestibility of nutrients. Therefore, the level of PC supplementation used in this experiment was in accordance with our previous study, in which the supplementation of PC in the growing pig diet was suggested to be no more than 10% [11]. In the present study, PC supplementation did not affect the final weight during the experimental period. However, in weeks 0–4 of the experimental period, the PC10 group exhibited significantly higher FCR levels than the other groups. There was a relationship between lower ADG and high ADFI in this period that led to the increase in FCR. Moreover, in the overall experimental period, the PC10 group was significantly higher in ADFI and ADG than in other groups (p < 0.05). Our results agree with several studies that reported that the supplementation of n3 PUFA (linseed) in the diet significantly increased indicators of productive performance, such as ADFI, ADG, and FCR [41,42,43].The dietary treatments in this study did not affect the carcass traits, including slaughter weight, carcass percentage, hot carcass weight, chill carcass weight, carcass length, and backfat thickness (p > 0.05). These results agree with the reports of Tratarkoon et al. [44], Ivanovic et al. [45], and Bertol et al. [46], who reported that the supplementation of different fat sources did not affect carcass traits. In terms of the meat quality, PC supplementation in growing pig diets did not affect meat quality factors such as drip loss, pH at 45 min, pH at 24 h, and chemical composition. These finding were in line with de Tonnac and Mourot [47], who reported that the supplementation of n3 PUFA in the diet did not influence meat quality, especially pH and drip loss in the finishing pig. In addition, we noticed the higher range of pH at 24 h in LD. This indicates the dark firm dry (DFD) which occurs due to heat stress, as the experimental period was conducted during summer in Chiang Mai (March to June), Thailand with a high temperature range ca. 30-39°C. Regarding this, Adzitey and Nurul [37] explained that when animals were exposed to chronic or long-term stress before slaughtering, the DFD meat can occur. The examples of chronic stress are long distance transportation, long hours of food deprivation, and overcrowding of animals in the lairage over a long period of time [37]. However, the pH 45 min and pH 24 h were not significant different between the treatments. Nonetheless, we found that the color of meat, particularly lightness (L*) in PC10, was significantly lower than that in other groups. Tartrakoon et al. [44] explained that the increased high content of unsaturated fatty acids in the diet affects meat color, especially L*. The low L* values also indicate a low fat content [29]. Moreover, Filho et al. [48] showed that water holding capacity had a positive correlations with lightness (L*), despite the absence of correlations between L* and pH 45 min and the low negative correlations of L* with pH 24 h. Together with the pH fall, the denaturation of myofibrillar and sarcoplasmic (myoglobin) proteins and the expulsion of the water from the myofibrils towards the extracellular space during rigor mortis may lead to structural changes that increase light scattering, making the meat paler (greater L*) [48,49,50,51].In the present study, the supplementation of PC in diet altered the fatty acid composition in pig tissues, including backfat, abdominal fat, and LD, mainly unsaturated fatty acids. Kouba and Mourot [52] reported that fatty acid composition in animal products is influenced both by the biosynthesis of fatty acids in animal tissues and by the lipids present in feedstuffs consumed by livestock. The effect of nutrition is more important in monogastric animals than in ruminants, because ruminants are capable of hydrogenating FA in the rumen. Moreover, Wojtasik et al. [19] explained that it was possible to change fatty acid content and the relationship between fatty acids that belonged to the n6 and n3 series in pig tissues, by applying an appropriate source of fat in the diet. In this study, the supplementation of PC in the diet significantly increased LA and ALA in all analyzed tissues. The results agree with the report of Nuernberg et al. [53], that the diet supplemented with 5% olive oil or 5% linseed oil significantly increased the relative content of linolenic acid and long-chain n3 fatty acids in lipids of muscle and backfat in pigs. The LA and ALA cannot be synthesized de novo because it lacks ω-6 (Δ12)-desaturase and ω-3 (Δ15)-desaturase, and, thus, they are essential dietary FA for humans and other mammals [20,54,55]. The PC was relatively rich in ALA (55.97%) content [11]. Therefore, we believe that this is one reason to induce high ALA content in tissues. However, the deposition of ALA depends on various factors, such as feed and type of tissue. Sobol et al. [56] found a higher ALA deposition in subcutaneous fat than in meat. They also reported that the deposition of ALA in pigs ranged from 50.4 to 69.0% in pigs weighing around 60–105 kg [56]. PC supplementation in the growing pig diet resulted in total ΣPUFA higher than in PC0 group for all analyzed tissues, but only LD was significantly different (p < 0.05), and similar results were presented by several studies; pig diet containing high n3 fatty acid contributes to a significantly higher ΣPUFA in various tissues [42,56,57,58]. We found that both LA and ALA were mainly fatty acids of ΣPUFA composition in the analyzed tissues. Generally, LA and ALA are metabolic precursors of long-chain n3 PUFA [20], including eicosapentaenoic acid (C20:5 or EPA) and docosahexaenoic acid (C22:6 or DHA). The biosynthetic pathway involves successive desaturation and chain elongation steps in the endoplasmic reticulum and a β-oxidation process localized in the peroxisome [20,59,60]. We found in this study that the supplementation of PC in the growing pig diet had no influence on DHA content in all analyzed tissues. This result agrees with Smink et al. [61], who reported that high ALA intake increased the long-chain ω-3 PUFA content, except that of DHA, in grower pigs. The increase in unsaturated fatty acids in the PC supplementation diet affected the ratio of ΣMUFA/ΣSFA significantly, being increased in abdominal fat, and increased the ΣPUFA/ΣSFA ratio in LD when dietary supplementation of PC was 10% in the diet. Our experimental results agree with Wojtasik et al. [19], in that the supplementation of n3 polyunsaturated fatty acids in diet influenced the increase in the ΣPUFA/ΣSFA ratio in LD muscle and subcutaneous fat. However, De Smet et al. [62] demonstrated that the ΣPUFA/ΣSFA ratio was influenced by genetics first, then followed by nutrition (mainly the overall fat level of the animal and intramuscular fat content). In addition, Cui et al. [23] reported that the supplementation of perilla seed oil induced a significantly reduced ΣPUFA/ΣSFA ratio in chicken breast muscle. In this study, the supplementation of PC 10% in the growing pig diet increased the ΣPUFA/ΣSFA ratio in backfat, abdominal fat, and LD to a higher level than that recommended by the UK Department of Health, and they recommended a ratio of ΣPUFA/ΣSFA in pork of 0.4 as a minimum [13]. Furthermore, the n6/n3 ratio was reduced in the investigated tissues when the level of PC supplementation was increased. The result showed the proportion of n6/n3 ratio between PC0 and PC supplementations (PC5 and PC10) ranged between 2.6 and 3.6 that was lower than the recommended ratio by the UK Department of Health (the n6/n3 recommended ratio is 4:1) [13]. Similar findings were obtained by Wojtasik et al. [19] and Okrouhlá et al. [63], who found that the increased n3 fatty acid content in the diet decreased the n6/n3 PUFA ratio. The decrease in the n6/n3 ratio means that the supplementation of PC in the diet increased the n3 fatty acid content in the pig. A previous explanation was that the fatty acid composition and ΣPUFA/ΣSFA and n6/n3 ratios in pig tissues are strongly influenced by the fat source in the diet [64]. This study found that a diet containing high ALA by PC supplementation had a significantly preferable influence on the fatty acid content in pig tissues, by increasing the level of ALA in tissues. The level of n3 fatty acid or ω-3 are important in mitigating the risk of cardiovascular diseases. Generally, the adult population has a very low intake of ω-3 PUFA due to limited access to the appropriate food sources, mainly seafood [20,65]. The improvement in pork quality by increasing n3 fatty acid content is necessary to help consumer meet the minimum nutritional requirement. Therefore, PC is a potential functional ingredient to enhance n3 fatty acid in pork. More importantly, the outcomes of this research encourages bio-circular green economic policy through biomass valorization.5. ConclusionsBased on the results of this study, it can be concluded that the supplementation of perilla cake in the growing pig diet improved product performance, in particular, the average daily gain, without affecting carcass traits and meat quality, except the lightness. At the same time, the supplementation of this biomass in pig diet elevated the fatty acid compositions in backfat, abdominal fat, and the longissimus dorsi muscle. In addition, perilla cake supplementation enhanced the polyunsaturated fatty acid content, especially C18:3n3, in their tissues, as well as the ΣPUFA/ΣSFA and n6/n3 ratios. On the basis of the obtained results, perilla cake has the potential to be used in pig diet to enhance pork quality, as an alternative functional meat that has a high n3 fatty acid content for consumer health concerns. The mechanism for changing fatty acid deposition and its effects on the health status in pigs requires further investigation.

animals : an open access journal from mdpi

10.3390/ani13050946

PMC10000074

The embryonic period, together with puberty and pregnancy, are known as the three main stages of mammary gland development. The development of the mammary glands is slowed during the embryonic period due to factors such as inadequate nutrition, which directly affect the development of the mammary glands and lactation after birth. However, the impact of embryonic nutrition on fetal mammary gland development is often unnoticed. We investigate the effect of nutritional intake on embryonic mammary gland development by administering different levels of nutritional restriction to female mice during gestation. Contrary to common belief, we found that mild maternal nutritional restriction contributes to mammary gland development in the offspring. Mammary gland dysplasia is not obvious until maternal nutritional restriction reaches 70% of the normal intake. Further embryonic mammary gland development studies can be performed based on our level of maternal nutritional restriction. In addition, the use of mice as model animals can also provide a reference for dairy farming, where nutrition should not be excessive during the gestation period of the cow; otherwise, it affects the mammary gland development of the offspring.

We aimed to investigate the effect of different levels of nutritional restriction on mammary gland development during the embryonic period by gradient nutritional restriction in pregnant female mice. We started the nutritional restriction of 60 female CD-1(ICR) mice from day 9 of gestation based on 100%, 90%, 80%, 70% and 60% of ad libitum intake. After delivery, the weight and body fat of the offspring and the mother were recorded (n = 12). Offspring mammary development and gene expression were explored by whole mount and qPCR. Mammary development patterns of in offspring were constructed using Sholl analysis, principal component analysis (PCA) and regression analysis. We found that: (1) Mild maternal nutritional restriction (90–70% of ad libitum intake) did not affect offspring weight, while body fat percentage was more sensitive to nutritional restriction (lower at 80% ad libitum feeding). (2) A precipitous drop in mammary development and altered developmental patterns occurred when nutritional restriction ranged from 80% to 70% of ad libitum intake. (3) Mild maternal nutritional restriction (90% of ad libitum intake) promoted mammary-development-related gene expression. In conclusion, our results suggest that mild maternal nutritional restriction during gestation contributes to increased embryonic mammary gland development. When maternal nutritional restriction reaches 70% of ad libitum intake, the mammary glands of the offspring show noticeable maldevelopment. Our results help provide a theoretical basis for the effect of maternal nutritional restriction during gestation on offspring mammary development and a reference for the amount of maternal nutritional restriction.

1. IntroductionNutritional challenges that occur during gestation, a critical period for embryonic growth and development, may lead to alterations in the physiological development and metabolism of the offspring after birth [1]. The most possible nutritional challenges during gestation are undernutrition and overnutrition, which can affect the health of both the fetus and the maternal body [2]. In particular, malnutrition during pregnancy, which still exists in underdeveloped regions, as a global problem, has important implications for the healthy development of the mother and the newborn [3]. These nutritional damages can cause permanent adjustments in the embryonic physiological state and organ development by inducing genetic changes in the proliferation/differentiation pathways during embryonic development [4,5]. Current research on nutritional restriction during pregnancy has focused on the placenta, brain and other organs that affect fetal survival [6,7], with little attention paid to fetal mammary gland development.The embryonic period, puberty and pregnancy are known as the three main stages of mammary gland development. The embryonic development of the mammary gland begins in many mammals at mid-gestation [8]; for mice, with a gestation period of 19–21 days, the mammary gland initiates development on day 10 of embryonic life [9]. The embryonic mammary glands are formed by a bilaterally multilayered ectodermal stripe from the forelimb bud to the hindlimb bud on the ventral surface of the embryo, referred to as the mammary line [10]. At day 11.5 of the mouse embryonic stage, these milk lines form five visible pairs of placebos. These placebos then become embedded in the mammary mesenchyme. At day 15–16 of the mouse embryonic stage, primary bud formation invades the secondary mammary mesenchyme and begins to develop a branching morphology [11]. Before birth, the mammary gland consists of a small ductal tree with a dominant duct and 10–15 branches embedded in the nascent fat pad [12]. The basic mammary duct system that forms at this time arises in the absence of hormonal input and remains essentially quiescent until puberty [10]. This basic ductal system forms the framework from which the mammary glands develop further during puberty and pregnancy to form the mature mammary glands [13]. If the development of the embryonic mammary glands is slowed at this period due to nutritional deficiencies and other factors, it directly affects the development of the mammary glands after birth and may even affect the amount of milk produced during lactation [14,15].Although the embryonic stage is the initiation of mammary gland development, nutritional regulation has remained less studied for this stage of mammary gland development. Since puberty is considered a critical window for nutritional regulation, most nutrition-related research has focused on this period [16,17]. The impact of embryonic nutrition on fetal mammary gland development is often unnoticed. Although the mammary gland that develops during the embryonic period is considered a basic ductal system, it has the ability to produce milk, known as neonatal milk or switch’s milk [10]. This indicates that the mammary gland is already equipped with basic lactation functions after birth, and if there are problems with the development of the mammary gland during the embryonic period, these basic functions are affected. Terminal end buds, an important structure in the extension of the mammary ducts during puberty, form only at the tips of the elongated ducts which are based on branches generated during the embryonic period [9,10]. Multipotent mammary stem cells (MASCs) from embryonic mammary gland formation are the source of MASCs and progenitor cells required for mammary duct development during puberty and alveolar luminal formation during pregnancy [9]. These results all suggest that there is a connection between embryonic, pubertal and gestational mammary development, and that mammary gland damage caused by nutritional fluctuations received during the embryonic period further affect mammary gland development after birth.The embryonic stage is the period of initial mammary gland formation, when the mammary gland gradually begins to expand through proliferation and differentiation in multipotent MASCs [18]. Many genes associated with mammary stem cells during the embryonic period have been shown to influence the future developmental fate of the mammary gland. The Axin2 gene was found to have the ability to allow cell regeneration in mammary gland transplantation assays, and the expression of the Axin2 gene during the embryonic period has been shown to be associated with future development in the ductal cell lineage [18]. During this period, Wnt5a has also been shown to be required for normal development of the mammary ducts [19]. In addition, MASCs marker genes such as Sox10, Procr, ELF5 and Aldh1a1 were identified by knockout studies and regulate key functions of mammary gland development [20,21,22,23]. After birth, MASCs become lineage-restricted with some becoming progenitor cells and contributing to the development of the mammary gland base or lumen [9]. Thus, nutritionally induced changes in embryonic mammary development may continue to affect mammary development in adulthood. The effect of altered nutrient levels on mammary stem cells has been demonstrated in previous studies with cells and adult mice [24,25]. Maria Theresa E. Montales et al. found that angiotensin in food affects the number of mammary cell-like/progenitor cells [24]. Omar M. Rahal et al. speculated that diet-regulated hormonal signaling could influence MASC self-renewal [25]. Studies on the effects of nutritional restriction on mammary stem cells have had mixed results, with one study suggesting that nutritional restriction attenuates mammary stem cell viability and inhibits mammary gland development [26], while another study suggests that nutritional restriction induces the self-renewal of mammary stem cells [27]. These results may be due to differences in the amount of nutrient limitation. Mild nutrient limitation mediates the restoration of stem cell self-renewal capacity through nutrient and energy-sensing pathways [27]. When nutrient limitation exceeds the regulatory level of the cells, apoptosis and necrosis of stem cells can occur due to nutrient deficiency.Compared to nutritional treatment after birth, nutritional treatment for the embryonic period is more difficult and requires nutritional interventions for the maternal body. The most common approach in studies of fetal undernutrition is accomplished through food or caloric restriction of the mother during gestation [28]. The mammalian placenta has evolved mechanisms that help buffer the fetus from short-term fluctuations in maternal diet and energy status [29]. In order to avoid this buffering mechanism, most of the studied protocols reduce maternal nutritional intake to 50–60% of the normal amount, exerting a significant impact on fetal growth and development through high levels of food restriction [7]. Moderate or low levels of food restriction may better mimic the clinical features of malnourished women, but few studies have investigated the effects of moderate food restriction during pregnancy on embryonic development.In addition to maternal nutritional interventions, the smaller size of the embryonic mammary gland presents challenges for the study of mammary gland development. The most visual method of viewing mammary gland development is the whole mount, a method of viewing a three-dimensional overview of the mammary gland, which provides a dense ductal epithelial structure within the complete mammary gland [30]. The whole mount requires the complete mammary gland to be isolated from the skin of the mouse and spread out as naturally as possible, which is more challenging for embryonic and newborn mice. In earlier studies, the results of the whole-mount analysis were difficult to quantify and were only used as a display image in the studies [31]. The complex ducts of the mammary gland in puberty can be evaluated in terms of the area covered and the denseness of the ducts observed visually. However, this unquantifiable observation is difficult to evaluate in the primary mammary gland, which has only 10–15 branches at birth. Jason P. Stanko et al. reported the use of Sholl analysis, an ImageJ plug-in for neuronal analysis, to quantify whole-mount results of the mammary gland [32]. The Sholl analysis creates a series of concentric rings based on a custom center (origin of the mammary duct) and extends to the most distal portion of the branch (enclosing radius). The Sholl analysis plug-in calculates the number of intersections that occur on each ring and then returns a Sholl regression coefficient (k), which is a measure of the rate of decay of the epithelial branches. In Sholl analysis, the sum inters (N) is the number of intersections of multiple concentric circles centered on the primary ducts with the ducts, reflecting the complexity of the mammary gland. The sholl regression coefficient (k) is a measure of the distal mammary branch complexity, which is close to 0, indicating more complex and well-developed distal mammary branches. Branch density is calculated using the formula N/MEA. Sholl analysis provides a valid quantitative measure of mammary branch complexity and has become a reliable method for studying mammary gland development. Mammary gland development in embryonic mice can be evaluated through a combination of fine dissection and whole-mount and Sholl analysis.Different levels of maternal nutritional restriction may have different effects on embryonic mammary gland development due to different maternal nutritional buffering and stem cell responses to nutrition. To investigate this, we established a pattern of nutritional restriction on mammary gland development during embryonic period by setting 100%, 90%, 80%, 70% and 60% diet intakes for female mice during pregnancy. The objective of our study was to investigate the effect of maternal gradient nutritional restriction on mammary gland development in offspring and provide a reference for the amount of maternal nutritional restriction. 2. Materials and Methods2.1. Animals and Experimental DesignSixty female 8-week-old CD-1(ICR) mice were provided by Vital River Laboratory Animal Technology Co., (Beijing, China) and mated with males of similar age. Each male mouse was put in a cage with 1 female mouse. Mating of mice was demonstrated by the presence of vaginal plugs. Female mice were individually housed after the discovery of the vaginal plugs and recorded as day 0 of gestation. All mice in our study were fed commercially available irradiated sterile growth and reproduction diets for experimental mice (SFS9112, Xietong Biotechnology, Yangzhou, China). To reduce the impact of nutritional restriction on early embryonic growth, it began on the ninth day of pregnancy. Pregnant mice were divided into five groups (n = 12): the 100% group was fed ad libitum (control group), and the 90%, 80%, 70% and 60% groups were fed 90%, 80%, 70% and 60% of the ad libitum food weight daily, respectively. The ad libitum group was mated one day earlier and their intake was used as the basis multiplied by 90%, 80%, 70% and 60% as the feed intake for the gradient nutrient restriction. The weight of the mice was recorded daily. The number of litters as well as the weight of the female mice and offspring were recorded on the day of delivery.2.2. Body Fat Percentage AssayOn the day of delivery, whole body image and body fat percentage were evaluated in vivo using dual-energy X-ray absorptiometry (DEXA) on an InAlyzer (Medikors Co., Seongnam, Republic of Korea). Female mice and female offspring were anesthetized using isoflurane (RWD, Shenzhen, China) and placed on a scanner bed and operated according to the instructions. After in vivo imaging, female offspring mice were euthanized using CO2.2.3. Collection and Preservation of Mammary GlandsThe mammary glands were removed immediately after euthanasia, the #4 inguinal mammary glands were placed on slides and immersed in Carnoy’s solution for whole mount, and the other mammary glands were stored at −80 °C for real-time quantitative polymerase chain reaction (qPCR).2.4. Mammary Whole MountThe mammary glands were fixed in Carnoy’s solution (60% absolute ethanol, 30% chloroform, 10% glacial acetic acid) for 4 h and then placed sequentially in ethanol at 100%, 70%, 50% and 10% concentrations for 15 min each. After soaking in deionized water for 5 min, the mammary glands were stained using carmine alum solution (1 g carmine alum, 2.5 g aluminum potassium sulfate in 500 mL dH2O) for 4 h. The stained mammary glands were soaked for 5 min using distilled water, then sequentially soaked in 70%, 95% and 100% alcohol, each concentration for 15 min. The mammary glands were placed in xylene for 12 h for transparency and then sealed with neutral resin. Whole-mount slices of mammary glands were sectioned for image acquisition using an upright microscope (Nikion, Japan).2.5. RNA Extraction and qPCRRNA from offspring mammary glands was extracted using RNA-easy Isolation Reagent (R701-01, Vazyme Biotech, Nanjing, China) according to the instructions. RNA quality was evaluated by 1% agarose gel electrophoresis, while the purity of the total RNA was determined by NanoDrop 2000 (NanoDrop, ThermoFisher Science, Waltham, MA, USA). The genomic DNA was removed from each RNA sample and reverse-transcribed into cDNA using an Evo M-MLV Mix Kit (Accurate Biology, AG11728, Hunan, China). Then qPCR was performed using a SYBR Green Premix Pro Taq HS qPCR Kit (Accurate Biology, AG11701, Hunan, China) with a LightCycler 96 Instrument (Roche, Basel, Switzerland). The reaction program was set to pre-denaturation at 95 °C for 30 s, followed by denaturation at 95 °C for 5 s and extension at 60 °C for 30 s, for a total of 40 cycles, with each reaction repeated 3 times. The primer sequences are shown in Table 1. The amplification efficiency and the specificity of the amplified products of each primer pair were verified using standard curves and melting curves, respectively. The mRNA expression of each sample was normalized relative to the expression of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Relative gene expression levels of each target gene were analyzed using the 2−ΔΔct method.2.6. Statistical AnalysisWe performed Sholl analysis on mammary whole-mount results according to the method described in a previous study [32]. Mammary gland whole-mount analysis was performed using ImageJ 2.1 software, and the Sholl analysis plugin 4.0.1 for ImageJ was used for Sholl analysis. The distance from the primary ducts to the most distal end of the mammary epithelium (enclosing radius) and the mammary epithelial area (MEA) were measured using ImageJ. The Sholl analysis was performed with the primary duct as the center, the enclosing radius as the ending radius and a radius step size of 0.02 mm. Since the mammary ducts in newborn mice are less developed and farther away from the mammary lymph nodes, the area occupied by the mammary lymph nodes was not calculated in Branch density.Body weight, body fat, litter size, Sholl analysis results and gene expression were analyzed using one-way ANOVA in the ad libitum feeding, 90%, 80%, 70% and 60% groups. One-way ANOVA was performed using IBM spss 25 (Armork, NY, USA), with Sidak correction for multiple testing. Body weight, body fat, litter size and gene expression data were presented as the mean ± the standard deviation (SD). Principal component analysis (PCA) was performed on enclosing radius, MEA, sum inters and k from the results of Sholl analysis in all groups. PCA was performed using the FactoMineR and factoextra packages in R4.2.1, and PCA biplot figures were generated. The enclosing radius of each group in the Sholl analysis results were regressed against MEA, sum inters and k. Linear regression analysis of the mammary Sholl analysis was performed using simple linear regression in GraphPad Prism software 9.1.0 (San Diego, CA, USA).3. ResultsAfter nutrient restriction management, a significant difference in body weight was observed in mice from day ten of pregnancy, and the difference persisted until the end of gestation (p < 0.05; Figure 1A). After parturition, the adult female mice showed a significant decrease in body weight compared to the control group (p < 0.05), except for the 90% group (p > 0.05; Figure 1B). However, there was no significant difference in body fat percentage in adult female mice after parturition (p > 0.05; Figure 1C).When the nutritional intake was only 60% of the normal intake, a significant decrease in litter size was observed compared to the control group (p < 0.05; Figure 2A), while the individual offspring weight was significantly lower than that of the other groups (p < 0.05; Figure 2B). The body fat percentage of the offspring was significantly higher in the control group than in the 80%, 70% and 60% groups (p < 0.05; Figure 2C,D).The mammary whole-mount images are shown in Figure 3A. For the enclosing radius, a significant increase was observed in the control group compared to the 70% and 60% groups (p < 0.05), while a significant increase was observed in the 90% group compared to the 60% group (p < 0.05; Table 2). MEA did not differ in the control, 90% and 80% groups (p < 0.05), while it was significantly lower in the 70% and 60% groups than in the former three groups (p < 0.05; Table 2). Sum inters were significantly higher in the control, 90% and 80% groups than in the other two groups (p < 0.05; Table 2). The k of 70% and 60% were significantly higher than the other three groups (p < 0.05; Table 2). Branching density was not significantly different among the groups (p > 0.05; Table 2). Consistent with the results in Table 2, an identifiable change in mammary gland development was observed from the 80% group to the 70% group in Figure 3B. Figure 3C shows the number of intersections of each concentric circle with the mammary ducts in the Sholl analysis. The control group had the longest duct extension distance. At a radius of 0.5 mm, the control, 90% and 80% groups reached the highest number of intersections with a similar peak, all higher than the 70% and 60% groups.In order to further investigate the reasons for the dramatic decline in offspring mammary development from the 80% group to the 70% group, we performed a PCA (Figure 3D) of the mammary whole-mount results (enclosing radius, MEA, sum inters and k). After dimensionality reduction, the data points in the control, 90% and 80% groups were nearer to each other, forming visible distance differences with the 70% and 60% groups, indicating that a massive reduction in mammary gland development in the offspring occurs when the maternal nutritional limit is reduced from 80% to 70%. The PCA bipartite plot shows the scores and loadings of the first two components (dim1 and dim2), revealing the projection of the observed indicators on a space with dim1 and dim2 as axes. In our study, the indicators of mammary gland development were explained by 79.1% of dim1 and 10.2% of dim2, respectively. The variable with the highest weight in the first principal component is the enclosing radius, indicating that the main reason for the difference in distance from the 80% to the 70% group in the mammary glands was the change in enclosing radius. A positive correlation between enclosing radius and MEA and sum inters and a negative correlation with k are presented in the PCA biplot.To analyze the effect of the enclosing radius, which has the highest weight in PCA, on the pattern of mammary gland development in the offspring, we performed a regression analysis of the whole-mount results (Figure 4). In the regression analysis of the enclosing radius with MEA, the 90% and 60% groups had larger slopes compared to the control group, while the 80% and 70% groups had smaller slopes. In the regression analysis of the enclosing radius versus sum inters, as maternal nutritional restriction increased, the slope first increased in the 90% group, then gradually decreased in the 80% and 70% groups and then showed an increase in the 60% group. In the regression analysis of the enclosing radius versus k, the slope of each group is less than the control group, with the 60% group having the lowest slope.We analyzed the expression of development-related genes (Sox10, Axin2, Elf5, Lgr5, Wnt5a, Aldh1a1, Procr), mammary basal cell marker genes (K5), mammary luminal cell marker genes (K18), estrogen (ERα,ERβ) and progesterone receptor (PR) genes in the mammary glands (Figure 5) by one-way ANOVA followed by a Sidak multiple-comparison test. In the 90% group, Sox10 expression was significantly higher than in the other four groups (p < 0.05), and Elf5 was significantly higher than in the control and 60% groups (p < 0.05). Sox10 was significantly lower in the control group than in the 90%, 70% and 60% groups (p < 0.05), and Axin2 was significantly higher in the control group than in the 60% group (p < 0.05). Aldh1a1 was significantly higher in the 80% group than in the 60% group (p < 0.05). The expression of K5 was significantly higher in the control group than in the 80%, 70% and 60% groups (p < 0.05). In the 60% group, ER1 was significantly lower than in the 90% group (p < 0.05) and ER2 was significantly lower than in the control group (p < 0.05). The expression of other genes did not differ significantly among the groups (p > 0.05).4. DiscussionNutritional deficiencies during gestation cause irreversible effects in fetal organs [33], but nutritional deficiency research on embryonic mammary gland development remains vacant. The impairment of mammary gland development at this phase may directly lead to delayed fetal mammary gland development in adulthood [9]. The small size of the mammary gland, which is difficult to observe, and the buffering through the placenta, which reduces the impact of nutritional fluctuations in the embryo, present challenges for the study of mammary gland development during this period. We performed a quantitative study of the mammary glands using whole mount combined with Sholl analysis and further analyzed the developmental pattern of the mammary gland by PCA and regression analysis. Through maternal gradient nutrient limitation, we established a pattern of offspring mammary gland development and revealed stem cell-related gene expression through a gradient reduction in maternal nutrient intake from 100% to 60% during gestation. The main findings of the study were: (1) Mild maternal nutritional restriction (90–70% of ad libitum intake) did not affect offspring weight, while body fat percentage was more sensitive to nutritional restriction (lower at 80% ad libitum feeding). (2) A precipitous drop in mammary development and altered developmental patterns occurred when nutritional restriction ranged from 80% to 70% of ad libitum intake. (3) Mild maternal nutritional restriction (90% of ad libitum intake) promoted mammary development-related gene expression.Inadequate nutrition during pregnancy can have an impact on maternal and fetal health, most notably in the form of weight loss [34]. In our study, differences in body weight of female mice emerged from the tenth day of gestation after nutritional restriction. After delivery, maternal mice in the 80% group lost significant body weight, while body fat percentage was not affected. For offspring, body fat percentage decreased first when nutritional intake was 80% of ad libitum, and weight loss occurred when it was 60%. Our result is similar to a previous study, which found no significant change in offspring birth weight during gestation for a maternal restriction to 75% ad libitum feeding [28]. A reduction in offspring body weight occurs when nutritional restriction reaches 60% or less of the ad libitum intake [34,35]. It seems that embryonic body fat percentage is more susceptible than body weight when faced with nutritional constraints. Mammary ducts and epithelium need to be embedded in the mammary stroma for growth, which is composed of homogeneous adipose tissue [36]. In studies on obesity, there is a strong association between mammary fat pads and obesity [37]. Although mammary fat pad and body fat have not been studied in studies on nutritional restriction, the possibility exists that a decrease in whole body fat percentage may affect mammary fat pad development.To assess mammary gland development, we performed a Sholl analysis on the mammary glands of offspring with different levels of nutritional restriction. Based on the Sholl analysis reported in the previous study [32], we innovatively performed PCA analysis and regression analysis on the results of the Sholl analysis to explore the developmental pattern of mammary glands. We found a dramatic decrease in mammary gland development when nutritional restriction was dropped from 80% to 70% of ad libitum intake. In contrast, there was no significant difference in the effect of normal feeding versus 90% and 80% of ad libitum feeding on mammary gland development. We hypothesize that the dramatic delay in mammary gland development may be due to the buffering of embryonic nutrients by the placenta as a “nutrient sensor” [29]. For maternal nutritional restriction to 90% and 80% of normal intake, the buffering mechanism in the maternal body mitigates the effect of nutrition on fetal mammary development, and as the intake decreases to 70%, the maternal buffering limit is exceeded, resulting in delayed mammary development. Inconsistently with our results, the nutritional intake of sows being restricted to 70% of ad libitum intake had no effect on the weight of mammary parenchyma, fat content, protein content and DNA content of the offspring [38]. The weight, DNA content and other methods used in their study to evaluate the mammary glands do not provide a complete view of the development of the mammary ducts compared to the whole mount. In addition, the species may also be responsible for the discrepancy between their results and our findings.From the results of the PCA analysis, we determined that the variable with the greatest weight is the enclosing radius. This suggests that the dramatic decrease in mammary development from the 80% to the 70% group was mainly caused by changes in the enclosing radius. To analyze the developmental pattern of the mammary glands, we performed a regression analysis of the enclosing radius with MEA, sum inters and k. Our study found a positive regression relationship between the enclosing radius and MEA and sum inters and an inverse regression relationship with k. This suggests that as the distance of the terminal duct from the primary duct increases, the mammary gland will cover a larger area with more complex branching, while the terminal decay will be slower. Similar to our results, the similar trend of the mammary longitudinal extension distance with the mammary epithelial area was found in a previous study [39].We found that compared to controls, mild nutritional restriction (90% of ad libitum intake) had larger regression coefficients in regression analyses with MEA and sum inters and smaller regression coefficients with k. This suggests that offspring with mild maternal nutritional restriction have better potential for mammary gland development. When nutrition was restricted to 80% of ad libitum feeding, the regression coefficients of the enclosing radius and MEA reflected reduced mammary epithelial area growth potential, despite no difference in mammary gland developmental indicators compared to the control group. Interestingly, the regression coefficients of MEA, sum inters and k all showed greater absolute values when the nutritional restriction was 60% of the ad libitum intake. At this level of nutritional restriction, the enclosing radius already showed a significant shortening and, therefore, mammary gland development slowed down.Sox10, Axin2 and Elf5 have been shown to function as key genes in embryonic mammary gland development. Sox10 is expressed in fetal mammary gland stem cells during embryonic mammary gland development and plays a central role in mammary gland development [40,41]. Axin2, a target gene of the Wnt/β-catenin pathway, has been used as a marker of functional stem cells in the mammary gland in a lineage-tracing approach [18]. Elf5 is required for the proliferation and differentiation of mammary epithelial cells in embryonic mouse mammary glands [42]. We found that mild nutritional restriction (90% of ad libitum intake) increased the gene expression of Sox10 and Elf5, suggesting a positive effect on mammary gland development. In a previous study on dietary control, alternate-day nutritional restriction was proven to increase the activity of tissue-specific stem cells and had positive implications for life extension [43]. Combined with our regression analysis of whole-mount results, our results suggest that mild maternal nutritional restriction does not impair offspring mammary development and may even increase offspring mammary growth potential by increasing the expression of stem cell-related genes. In addition, 60% of ad libitum feeding reduced Axin2 expression, suggesting that high levels of nutritional restriction inhibit mammary stem cell development and mammary gland development. Consistent with our results, in a study of high levels of maternal gestational nutritional restriction (50% of ad libitum feeding), the ability to differentiate neural progenitor cells was decreased [44]. These results suggest that stem cell activity in the embryonic mammary gland is related to the level of maternal nutritional restriction, with mild nutritional restriction contributing to stem cell-associated gene expression and high nutritional restriction inhibiting them. K5 is a known marker gene in the myoepithelial/basal layer of the mammary gland [45]. Our study shows that a decrease in K5 gene expression occurs in the basal lamina of the mammary glands when nutrition is restricted to 80% of ad libitum feeding. Combined with regression analysis, our results showed that the expansion potential of mammary basal and mammary gland area was affected by maternal nutritional restriction up to 80% of ad libitum feeding, despite no significant difference in the results of whole-mount analysis.Embryonic mammary gland development is considered to be hormone-nondependent, and previous studies have demonstrated that embryonic mammary glands are able to develop in mice lacking estrogen (ER-α and ER-β) and progesterone receptors [9,46,47]. After birth, especially during puberty, the mammary glands are stimulated by these hormones to develop rapidly. Estrogen is required for the branching of the mammary ducts during puberty, and estrogen and progesterone are required for lobuloalveolar development during pregnancy. In our study, ER-β receptor expression appeared to be reduced when nutritional restriction reached 60% of the ad libitum intake. This suggests that high levels of maternal nutritional restriction may affect the development of offspring mammary estrogen receptors whose impairment may have further effects on mammary development during puberty.5. ConclusionsIn conclusion, our results suggest that mild maternal nutritional restriction (90% of ad libitum intake) during gestation contributes to increased embryonic mammary gland development. When nutritional restriction ranges from 80% to 70% of ad libitum intake, mammary gland development decreases dramatically, and changes in developmental patterns occur.

animals : an open access journal from mdpi

10.3390/ani11082324

PMC8388487

Cryopreservation of oocytes can cause high oxidative stress, reduce the quality of vitrified-warmed oocytes, and seriously hinder the application of oocyte cryopreservation technology in production and medicine. In this work, we found for the first time that melatonin can exert antioxidant effects through receptors and regulate the Nrf2 antioxidant pathway to respond to oxidative stress of vitrified-warmed oocytes, thereby improving both oocyte quality and the potential for subsequent development. The results illustrated the molecular mechanism of melatonin’s antioxidant effect in vitrified-warmed oocytes and provided a theoretical basis for the application of melatonin in the cryopreservation of oocytes. These findings are of great significance for the further application of oocyte cryopreservation technology to production and assisted reproduction in the future.

Previously it was reported that melatonin could mitigate oxidative stress caused by oocyte cryopreservation; however, the underlying molecular mechanisms which cause this remain unclear. The objective was to explore whether melatonin could reduce oxidative stress during in vitro maturation of vitrified-warmed mouse germinal vesicle (GV) oocytes through the Nrf2 signaling pathway or its receptors. During in vitro maturation of vitrified-warmed mouse GV oocytes, there were decreases (p < 0.05) in the development rates of metaphase I (MI) oocytes and metaphase II (MII) and spindle morphology grades; increases (p < 0.05) in the reactive oxygen species (ROS) levels; and decreases (p < 0.05) in expressions of Nrf2 signaling pathway-related genes (Nrf2, SOD1) and proteins (Nrf2, HO-1). However, adding 10−7 mol/L melatonin to both the warming solution and maturation solutions improved (p < 0.05) these indicators. When the Nrf2 protein was specifically inhibited by Brusatol, melatonin did not increase development rates, spindle morphology grades, genes, or protein expressions, nor did it reduce vitrification-induced intracellular oxidative stress in GV oocytes during in vitro maturation. In addition, when melatonin receptors were inhibited by luzindole, the ability of melatonin to scavenge intracellular ROS was decreased, and the expressions of genes (Nrf2, SOD1) and proteins (Nrf2, HO-1) were not restored to control levels. Therefore, we concluded that 10−7 mol/L melatonin acted on the Nrf2 signaling pathway through its receptors to regulate the expression of genes (Nrf2, SOD1) and proteins (Nrf2, HO-1), and mitigate intracellular oxidative stress, thereby enhancing in vitro development of vitrified-warmed mouse GV oocytes.

1. IntroductionOocyte cryopreservation has been used in genetic selection, the preservation of germplasm resources and scientific research [1,2]. More importantly, oocyte cryopreservation, combined with advanced biotechnologies, can avoid some of the ethical restrictions inherent in human-assisted reproduction techniques [3,4,5]. However, vitrification and warming procedures inflict cryodamage in oocytes and affect their cell membrane, organelles, DNA, and other structures [6,7,8]. Mitochondrial damage also causes the excessive accumulation of reactive oxygen species (ROS), resulting in high oxidative stress which will further damage mitochondria and other structures, affect cell signal transduction, and decrease developmental potential of vitrified-warmed oocytes [6,9]. Therefore, reducing the cellular oxidative stress can improve the quality of vitrified-warmed oocytes.The NF-E2-related factor 2/Antioxidant response element (Nrf2/ARE) pathway, one of the important antioxidant response pathways in cells, is regulated by the active factor Nrf2 [10]. Normally, Nrf2 binds to Kelch-like ECH-associated protein-1 (Keap1) in the cytoplasm and is inactive. However, when intracellular oxidative stress increases, electrophiles and quinones bind to the cysteine residues of Keap1, resulting in conformational changes and the dissociation of Nrf2, enabling it to enter the nucleus, bind to the ARE sequence, and induce the expression of downstream antioxidant proteins [11], including Heme Oxygenase-1 (HO-1), Recombinant Glutamate Cysteine Ligase, and Modifier Subunit (GCLM) and Superoxide Dismutase (SOD) [12,13], with important roles in anti-oxidation and anti-apoptosis [14,15,16]. Cryopreservation can significantly decrease the expression of Nrf2 in porcine oocytes [17], contributing to increased oxidative stress in vitrified-warmed oocytes. However, it remains unclear whether cryopreservation can impact the Nrf2 signaling pathway and related proteins during in vitro maturation of mouse GV oocytes. Additionally, the possible relationship between changes in expression of these proteins and ROS-inflicted oxidative stress remains to be explored.Melatonin (MT), an important antioxidant, can regulate the Nrf2 signaling pathway in cells [18]. Through the Nrf2 signaling pathway, melatonin, can antagonize oxidative injury induced by manganese (Mn) in the striatum of mice [19]; protect against early brain injury in a subarachnoid hemorrhage model in mice [20]; treat experimental allergic encephalomyelitis in mice [21]; attenuate acute kidney ischemia/reperfusion injury in diabetic rats [22]; prevent lipopolysaccharide- (LPS) induced epithelial–mesenchymal transition in human alveolar epithelial cells [23]; promote porcine embryo development [24]; and inhibit oxidative stress and apoptosis in cryopreserved ovarian tissues [25]. Although previously it has been shown that melatonin can reduce excessive ROS in oocytes induced by cryopreservation in mice [26,27,28,29], bovine animals [30], and humans [31], the underlying mechanism through which melatonin might regulate the Nrf2 signaling pathway remains to be elucidated.Melatonin receptors, MT1 (MEL-1A) and MT2 (MEL-1B), are mainly distributed on the cell membrane [32,33,34,35]. Melatonin can have an antioxidant role in cells through receptors, such as by protecting against cisplatin-induced ovarian damage in mice [36] and delaying the fertility decline in female animals [37]. Melatonin can also regulate the intracellular Nrf2 signaling pathways through receptors, preventing the senescence of canine adipose-derived mesenchymal stem cells [38] and promoting the integrity of the blood–brain barrier in methamphetamine-induced inflammation in the primary microvascular endothelial cells of rat brains [39]. In preliminary studies, we determined that there were melatonin receptors on the membrane during in vivo maturation of mouse GV oocytes. Melatonin can reduce the oxidative stress level during the in vitro maturation of vitrified-warmed GV oocytes, but it remains to be studied whether this role is performed through its receptors.It has been reported that Brusatol, a quassinoid from the seeds of Brucea sumatrana, can specifically inhibit the transcription and translation of the Nrf2 protein and can inhibit the function of Nrf2 pathway [40,41]. Luzindole (N-acetyl-2-benzyltryptamine) is considered an antagonist of melatonin receptors and is usually employed to study the signaling pathways involved in the action of melatonin in cells [42,43]. Therefore, our objective was to explore whether melatonin can reduce oxidative stress during in vitro maturation of vitrified-warmed mouse GV oocytes through the Nrf2 signaling pathway or its receptors.2. Materials and MethodsUnless otherwise stated, all chemicals were commercially available from Sigma-Aldrich (St. Louis, MO, USA). The experimental mice were maintained and handled in agreement with the requirements of the Animal Ethical and Welfare Committee (AEWC) of Sichuan Agricultural University, as well as in accordance with the relevant laboratory animal management regulations (Ministry of Science and Technology of China, Beijing, June 2004) and with the approval of the Animal Health and Use Committee of Sichuan Agricultural University College of Animal Science and Technology (No. DKYB20081003).2.1. Oocyte CollectionFemale ICR mice (n = 298), 8–10 wks old, were purchased from Chengdu Dashuo Experimental Animal Co., Ltd. and maintained at 18–25 °C and 50–70% humidity, with 14 h of light and 10 h of darkness. After a 2-wk adaptation period, each mouse was intraperitoneally injected with 10 IU pregnant mare serum gonadotropin (PMSG, NingBo Second Hormone Factory, Ningbo, China). After 44–48 h, the mice were subjected to a sudden death by separating the neck from the base of the skull. Ovaries were taken out, placed in a 37 °C M2 solution, sliced under the stereomicroscope with syringe needles, and GV oocytes with obvious germinal vesicles were selected for experiments.2.2. Oocyte Vitrification and WarmingOpen-pulled straws (OPS) were prepared as described [26,44]. In brief, the straws (0.25 mL) were heat-softened and pulled manually, to obtain straws approximately 3 cm long, 0.10 mm inner diameter, and 0.15 mm outer diameter.Vitrification solutions were (1) 10% dimethylsulfoxide (DMSO) and 10% ethylene glycol (EG) in phosphate-buffered saline (PBS) medium (referred as ED), and (2) 15% EG, 15% DMSO, 300 g/L Ficoll, 0.5 mol/L sucrose and 3 g/L bovine serum albumin (BSA) in PBS medium (referred as EDFS30).Oocytes were vitrified by the OPS method. Firstly, the oocytes were equilibrated in ED for 30 s, and then transferred to EDFS30 for 25 s. Finally, the oocytes were precisely arranged at the end of the OPS (8–10) and quickly dropped into liquid nitrogen. For warming, the OPS was removed from liquid nitrogen and the end was quickly placed in warming solution (0.5 mol/L sucrose solution) equilibrium for 5 min, then transferred to M2 solution and rinsed three times. All manipulations were carried out at 37 °C on a temperature control stage attached to a stereomicroscope (SMZ1500, Nikon, Tokyo, Japan).2.3. Oocyte Culture and In Vitro MaturationThe GV oocytes were randomly assigned into following groups. Control group (Con): fresh oocytes matured directly in vitro; vitrification group (Vit): oocytes were vitrified and then matured in vitro; vitrification + melatonin group (Vit + MT): on the basis of vitrification group, 10−7 M melatonin [26] was added to the warming and maturation solutions (M16); vitrification + melatonin + brusatol group (Vit + MT + Bru): on the basis of vitrification group, 10−7 M melatonin was added to the warming and maturation solutions, and 50 nM Brusatol (Nrf2 inhibitor) [24,45] was added to the maturation solution; and vitrification + melatonin + luzindole group (Vit + MT + Luz): on the basis of vitrification group, 10−7 M melatonin was added to the warming and maturation solutions, and 10−7 M luzindole (Melatonin receptors inhibitor) [46] was added to the maturation solution. Then, fresh and vitrified oocytes were rinsed and placed in M16 medium. After 8 h, oocytes were stained with DAPI (Vector Laboratories Inc., Burlingame, CA, USA), and the chromosomes that were arranged in an orderly manner on the equatorial plate under a fluorescence microscope (BX53F, OLYMPUS, Tokyo, Japan) were in the MI stage. Similarly, 14 h later, the oocytes of each group were observed under a stereomicroscope, and the first polar body (PB-I) was extruded in the MII stage. The development rates of MI and MII oocytes in each group were calculated (Figure 1).2.4. Spindle Morphology and ClassificationFor evaluation of spindle morphology and classification, MI (or MII) oocytes were fixed, permeabilized, incubated with FITC-anti-α-tubulin antibody, stained with DAPI, and observed according to our previous report [26]. The spindle configuration was graded as described [47]. In brief: Grade 0 = no spindle detected; Grade 1 = severely diminished spindle, less than 50% of the normal spindle in size; Grade 2 = mildly diminished spindle, larger than 50% of the normal spindle; Grade 3 = equivalent to normal spindle size and shape (fusiform spindle); Grade 4 = equivalent to normal spindle size and shape (barrel-shaped spindle).2.5. Measurement of Intracellular ROS and Glutathione (GSH) LevelsTo measure intracellular ROS levels, MI (or MII) oocytes were incubated in M2 solution containing 20 µM 2, 7-dichlorodihydrofluorescein diacetate (H2DCFDA, Invitrogen, Carlsbad, CA, USA) for 30 min (37 °C, 100% humidity, and 5% CO2 concentration), and then washed three times in M2 solution containing 3 g/L BSA for 5 min each. Finally, oocytes were placed under a fluorescent microscope and measured under 460 nm excitation with a filter, and fluorescence images were recorded as TIFF files. After deducting background value, fluorescence intensities were quantified using Image J software (Version 1.48; National Institutes of Health, Bethesda, MD, USA) [26].Intracellular GSH levels were measured by M2 solution containing 10 µM 4-chloromethyl-6, 8-difluoro-7-hydroxycoumarin (Cell-Tracker Blue, Invitrogen, Carlsbad, CA, USA) with a filter at 370 nm excitation, and other procedures as described for assessment of ROS, were carried out.2.6. Quantitative Polymerase Chain Reaction (Q-PCR)In each group, total complementary DNA (cDNA) was obtained from oocytes (n = 20–25) at MI and MII stages using TransScript-Uni Cell to cDNA Synthesis SuperMix for Q-PCR (TransGen Biotech, Beijing, China). Then, cDNA was quantified by Q-PCR using TransStart Tip Green qPCR SuperMix (TransGen Biotech, Beijing, China) on a CFX Connect Real-Time Detection System (Bio-Rad, Hercules, CA, USA) under standard conditions. Three replicates were performed for this assay, and the relative mRNA expression levels were obtained using the 2−Δ∆Ct method, with Gapdh was used as a reference gene for normalization [48]. Primer information is detailed in Table 1.2.7. Immunofluorescent StainingThe MI (or MII) oocytes were fixed in 4% (w/v) paraformaldehyde for 30 min, permeabilized in PBS with 1% Triton X-100 (v/v) for 20 min, and blocked with 1% BSA for 1 h sequentially. Then they were exposed to primary antibody at 4 °C overnight, washed three times for 5 min each in wash buffer (PBS containing 0.01% Triton X-100 and 0.1% Tween 20), and then stained with fluorescently labeled secondary antibodies and incubated at 37 °C for 1 h. After incubation, oocytes were washed three times in washing buffer for 15 min. Finally, the oocytes were placed on a clean glass slide, stained with DAPI and observed under a fluorescent microscope [49]. Slides for quantitative analysis were photographed with a fluorescent microscope under the same fluorescence parameters and magnification, and images recorded as TIFF files. Fluorescence intensities were quantified using Image J software after deducting the background value.Antibodies and dilution ratios used in immunofluorescence were as follows: Nrf2 antibody, 1:200 (Proteintech, USA, 16396-1-AP); GCLM antibody, 1:300 (Proteintech, 14241-1-AP); HO-1 antibody, 1:300 (Proteintech, 66743-1-Ig); MEL-1A antibody, 1:200 (SC-390328, SantaCruz); and MEL-1B antibody, 1:200 (SC-398788, SantaCruz).2.8. Statistical AnalysisStatistical analyses were performed by SPSS statistical software (v. 22.0; IBM, Chicago, IL, USA) to make a one-way ANOVA followed by a post hoc Fisher’s least significant difference (LSD) test. Data were expressed as the mean ± standard error and all experiments were repeated three times. For all analyses, p > 0.05 means no significant difference, whereas p < 0.05 means statistically significant.3. Results3.1. Melatonin Promotes In Vitro Development of Vitrified-Warmed Mouse GV Oocytes through Nrf2 ProteinThe development rates of GV oocytes to MI and MII were lower (p < 0.05) than in the corresponding control group after vitrification (Table 2). However, when melatonin was added, the development rates of GV oocytes to MI and MII were higher (p < 0.05) than the corresponding vitrification group, and not different from the control groups. In contrast, when the vitrification group was co-treated with melatonin and Brusatol, the development rates of GV oocytes to MI and MII were lower (p < 0.05) than in the vitrification + melatonin and control groups, and there was no difference (p > 0.05) from the vitrification group. Therefore, melatonin promoted in vitro development of vitrified-warmed mouse GV oocytes; however, when Nrf2 protein was specifically inhibited, melatonin failed to promote maturation.3.2. Melatonin Improves the Spindle Morphology Grades of MI and MII Oocytes from the Vitrified-Warmed Mouse GV Oocytes through Nrf2 ProteinImmunofluorescence staining of intracellular α-tubulin protein was used to observe the spindle morphology of GV oocytes as they developed to the MI and MII stages (Figure 2A). The spindle morphology grades of MI and MII oocytes in the control group were 3.00 ± 0.04 and 3.02 ± 0.04, respectively; however, the average grades of MI and MII oocytes in the vitrification group were 2.08 ± 0.03 and 1.96 ± 0.04, respectively, lower (p < 0.05) than the corresponding stages in the control group. With melatonin supplementation, the average grades of MI and MII oocytes were 2.37 ± 0.02 and 2.45 ± 0.03, respectively, better (p < 0.05) than the corresponding vitrification group, but still lower (p < 0.05) than the control group. However, when the vitrification group was co-treated with melatonin and Brusatol, average grades of MI and MII oocytes were 2.09 ± 0.02 and 1.97 ± 0.06, lower (p < 0.05) than the corresponding vitrification + melatonin and control groups, and not different (p < 0.05) from the vitrification group. Therefore, melatonin promoted the spindle morphology recovery of vitrified-warmed mouse GV oocytes during in vitro maturation, but when the Nrf2 protein was specifically inhibited, melatonin failed to improve spindle morphology.3.3. Melatonin Regulates Oxidative Stress of MI and MII Oocytes from the Vitrified-Warmed Mouse GV Oocytes through either Nrf2 Protein or Melatonin ReceptorsWhen GV oocytes developed to the MI and MII stages, ROS and GSH levels were measured to evaluate intracellular oxidative stress. The ROS levels of the MI and MII oocytes in the vitrification group were higher (p < 0.05) than that of the corresponding control group (Figure 3A,B). However, when melatonin was added, intracellular ROS levels (MI and MII) were lower (p < 0.05) than the vitrification group, but still higher (p < 0.05) than the control group. Furthermore, when the vitrification group was co-treated with melatonin and Brusatol, intracellular ROS levels (MI and MII) were higher (p < 0.05) than the corresponding vitrification + melatonin and control groups. The GSH levels of MI oocytes in the vitrification group were lower (p < 0.05) than in the control group (Figure 3C,D). With addition of melatonin, GSH levels of MI and MII oocytes were higher (p < 0.05) than the corresponding vitrification group, even at MI stage, and higher than the control group. Finally, when the vitrification group was co-treated with melatonin and Brusatol, GSH levels of the MI and MII oocytes were lower (p > 0.05) than the corresponding vitrification + melatonin and control groups, but not different (p > 0.05) from the vitrification group. Therefore, melatonin, acting through Nrf2 protein, regulated the oxidative stress level of vitrified-warmed mouse GV oocytes during in vitro development through the Nrf2 pathway.During in vitro development of vitrified-warmed mouse GV oocytes, melatonin can reduce intracellular ROS levels and increase GSH levels. Here, we further explored whether melatonin has a role through its receptors. Melatonin receptor 1 (Red, Figure 4A) and melatonin receptor 2 (Green, Figure 4B) were expressed on the membranes of GV, MI, and MII oocytes. Melatonin decreased (p > 0.05) ROS levels of the MI and MII oocytes after vitrification; however, when the vitrification group was co-treated with melatonin and luzindole, intracellular ROS levels (MI and MII) were higher (p < 0.05) than the vitrification + melatonin group, but still lower (p < 0.05) than the vitrification group (Figure 5A,B). The addition of melatonin increased (p < 0.05) intracellular GSH levels (MI and MII) after vitrification (Figure 5C,D). However, when the vitrification group was co-treated with melatonin and luzindole, there was no difference (p > 0.05) in GSH levels in the MI and MII oocytes compared to the corresponding vitrification + melatonin group. Therefore, during in vitro development of vitrified-warmed mouse GV oocytes, when melatonin receptors were inhibited, the ability of melatonin to scavenge ROS was reduced, but the effect on GSH levels was not significant, indicating that melatonin exerted a certain antioxidant protective role through its receptors.3.4. Melatonin Regulates Nrf2 Pathway-Related Genes and Proteins of MI and MII Oocytes from the Vitrified-Warmed Mouse GV Oocytes through Melatonin ReceptorsA qPCR analysis of the mRNA of Nrf2, GCLM, SOD1 and SOD2 genes during in vitro development of GV oocytes (Figure 6) produced several outcomes. When GV oocytes of the vitrification group developed to MI and MII, the expressions of Nrf2 and SOD1 genes were lower (p < 0.05) than the corresponding control group, and melatonin up-regulated (p < 0.05) expression of Nrf2 and SOD1 genes in MI oocytes and Nrf2 gene in MII oocytes. Although vitrification did not affect (p > 0.05) the expression of GCLM and SOD2 genes, the expressions of these two genes were increased (p < 0.05) after melatonin was added. When the vitrification group was co-treated with melatonin and Brusatol, except for the SOD1 gene in MII oocytes, the expression levels of genes (Nrf2, GCLM and SOD2) were lower (p < 0.05) than in the corresponding vitrification + melatonin group. Therefore, vitrification disrupted the expression of Nrf2 signaling pathway-related antioxidant genes (Nrf2, SOD1) and in vitro development of vitrified-warmed mouse GV oocytes, although exogenous melatonin restored the expression of these genes; Finally, when the Nrf2 protein was specifically inhibited, the ability of melatonin to restore Nrf2 and SOD1 genes decreased.The protein expressions of Nrf2, GCLM, and HO-1 during the development of mouse GV oocytes in vitro were detected by immunofluorescence staining. We found that the Nrf2 protein was expressed in mouse GV oocytes throughout the development process in vitro (Figure 7, Green fluorescence). In GV oocytes, the Nrf2 protein was mainly located in germinal vesicles, and MI and MII oocytes were more concentrated around the chromosome. When the GV oocytes of the vitrification group developed to the MI and MII stages, the Nrf2 (Figure 8A,B) and HO-1 (Figure 8C,D) proteins were lower (p < 0.05) than the corresponding control group. When melatonin was added, these two proteins were higher (p < 0.05) than in the corresponding vitrification group, and there was no difference (p > 0.05) from the control group. However, when the vitrification group was co-treated with melatonin and Brusatol, these two proteins were lower (p < 0.05) than in the corresponding vitrification + melatonin and control groups, and there was no difference (p > 0.05) from the vitrification group. The GCLM protein of MI oocytes in the vitrification group (Figure 8E,F) was significantly lower than the control group. When melatonin was added, the upregulation of the GCLM protein was not significant. Furthermore, when the vitrification group was co-treated with melatonin and Brusatol, the expression of the GCLM protein in MI oocytes was lower (p < 0.05) than the vitrification + melatonin and control groups and there was no difference (p > 0.05) from the vitrification group. Furthermore, there was no difference (p > 0.05) among all groups in the expression of the GCLM protein in MII oocytes. Therefore, vitrification down-regulated the expression of Nrf2 and HO-1 during the development of GV oocytes in vitro, and the addition of melatonin restored the expressions of these proteins. In addition, when the Nrf2 protein was specifically inhibited, the stimulatory effect of melatonin on the HO-1 protein disappeared.Next, we further explored whether melatonin regulated genes (Nrf2 and SOD1) and proteins (Nrf2 and HO-1) through receptors. The GV oocytes that were vitrified and co-treated with melatonin and luzindole (Figure 9) developed to the MI stage, but the gene expression levels of Nrf2 and SOD1 were lower (p < 0.05) than the vitrification + melatonin group. After the development to the MII stage, the gene expression levels of Nrf2 and SOD1 were not decreased (p > 0.05), but there was no difference (p > 0.05) from the vitrification group, indicating that the receptors’ inhibition also down-regulated the expression of these genes. As for the GCLM and SOD2 genes, there was no difference between the co-treatment group and the only group containing melatonin at MI and MII stages. When GV oocytes were vitrified and co-treated with melatonin and luzindole (Figure 10A,B), the levels of Nrf2 proteins in the MII oocytes were lower (p < 0.05) than in the vitrification + melatonin and control groups. There were no (p > 0.05) regulation effects in the MI oocytes and no difference (p > 0.05) from the corresponding vitrification group in MI and MII oocytes. However, when the vitrification group was co-treated with melatonin and luzindole, levels of HO-1 protein (Figure 10C,D) in the MI and MII oocytes were lower (p > 0.05) than in the corresponding vitrification + melatonin and control groups, with no difference (p > 0.05) from the corresponding vitrification group. For the GCLM protein, the co-treatment group was lower than the melatonin group at the MI stage, but there was no difference between the two groups at the MII stage.In summary, when melatonin receptors were inhibited, the regulatory effects of melatonin on the Nrf2 pathway-related genes (Nrf2, SOD1) and proteins (Nrf2, HO-1) during development of vitrified-warmed mouse GV oocytes were weakened, indicating that melatonin regulated the Nrf2 signaling pathway through its receptors.4. DiscussionThe cryopreservation of mouse [27], bovine [50], pig [51], and human oocytes [52] increases oxidative stress, although exogenous melatonin reduce oxidative stress in vitrified-warmed oocytes [30,53]. The Nrf2 signaling pathway is an important intracellular antioxidant signaling pathway, critical for maintaining redox balance [31,54]. Melatonin is a regulatory molecule of the Nrf2 protein [18]. In this study, after the Nrf2 protein was specifically inhibited by Brusatol, the potential of melatonin to reduce ROS levels decreased during in vivo development of vitrified-warmed mouse GV oocytes, and the in vitro maturation rate of oocytes was lower than that of melatonin per se. Therefore, melatonin acted on Nrf2 to regulate intracellular ROS and promote in vitro development of vitrified-warmed oocytes. Moreover, we found that the development rates and GSH levels of the Nrf2 protein-inhibited group were not significantly different from the vitrification group and the ROS levels were even higher, indicating that the function of melatonin was completely inhibited and melatonin may have played a role mainly through the regulation of the Nrf2 protein in oocytes. Moreover, melatonin can exert antioxidant protection through its receptors by antagonizing oxidative stress-induced mitochondrial dysfunction in retinal pigmented epithelium cells [55], reducing sodium fluoride-induced human hepatotoxicity [56], and improving the developmental competence of oocytes from juvenile goats [46]. In this study, when luzindole, a melatonin receptor inhibitor, was added to the vitrification group, the ability of melatonin to scavenge ROS decreased during in vivo development of vitrified-warmed mouse GV oocytes. However, it still had a significant antioxidant capacity compared to the vitrification group. Therefore, melatonin exerted a degree of antioxidative protection through the receptors in vitrified-warmed oocytes. Melatonin, as a small liposoluble molecule, easily penetrates through the cell membrane and enters into the cells, so melatonin may also enter into the oocyte directly through the osmotic pathway in order to function. High oxidative stress can cause errors in spindle formation [57,58], and melatonin can protect spindle configuration by reducing oxidative stress [26]. In this study, similar results also appeared: the vitrified-warmed mouse GV oocytes produced excessive ROS and the spindle morphology grades of MI and MII oocytes decreased during in vitro development. However, after adding melatonin, ROS decreased and spindle morphology grades recovered. In addition, when the Nrf2 protein was inhibited, the antioxidant effect of melatonin disappeared, and the spindle could not be effectively protected. Therefore, we inferred that melatonin may have acted on the Nrf2 protein to mitigate spindle damage caused by vitrification.Cryopreservation can damage or destroy mRNA in oocytes, such as HSP90, P34cdc2 and Cyclin B in sheep [59]; CD9 in mice [60]; and SOD1, Mfn2, Bax and BCL in pigs [61]. In this study, vitrification down-regulated the expression of genes related to the Nrf2 signaling pathway (Nrf2, SOD1), which indicated that vitrification could damage the mRNA structure of related genes in the Nrf2 pathway or accelerate their degradation. Vitrification had no significant effect on GCLM and SOD2 genes, indicating that these genes were more stable and perhaps either produced a strong response to cold stimulation or had differences related to the timeliness of their functions [62]. Melatonin can enhance the expression of genes related to the intracellular Nrf2 pathway in response to oxidative stress [25,35,63]. Similarly, in this study, melatonin restored levels of Nrf2 and SOD1 genes during in vitro development of vitrified-warmed mouse GV oocytes. Therefore, we inferred that melatonin may further activate the expression of Nrf2 signaling pathway-related genes or inhibit mRNA degradation, reduce oxidative stress, and promote in vitro maturation of vitrified-warmed mouse GV oocytes. Furthermore, melatonin can regulate the mRNA levels of Nrf2 signaling pathway-related genes through receptors in response to damage [38,64]. In this study, after melatonin receptors were inhibited, the mRNA levels of Nrf2 and SOD1 genes in the MI oocytes were significantly decreased, and Nrf2 and SOD1 mRNA levels in MII oocytes were also down-regulated. Therefore, melatonin can also regulate genes related to the Nrf2 signaling pathway through receptors during in vitro maturation of vitrified-warmed mouse GV oocytes.Cryopreservation can destroy some proteins in oocytes, e.g., porcine Nrf2, human aquaporin 1 (APQ1) [31], goat CDK1 [65], and mouse Mad2 [66], etc. After the vitrification of porcine COC oocytes, proteomics analysis revealed that 59 proteins were down-regulated [67]. Similarly, in this study, the vitrification of mouse GV oocytes resulted in a significant down-regulation of Nrf2 and the downstream antioxidant protein HO-1 was disturbed in MI and MII during in vitro maturation. The disorder of protein expression caused by cryopreservation may be directly caused by damage to the protein structure, or lead to the degradation of the corresponding mRNA, which ultimately affects basal protein levels.During in vitro maturation of vitrified-warmed mouse GV oocytes, 10−7 M melatonin restored the expression levels of Nrf2 and HO-1, indicating that melatonin can reduce the intracellular oxidative stress by protecting proteins related to the Nrf2 signaling pathway. Similarly, melatonin-enhanced Nrf2 and downstream proteins in response to injury stress on porcine COC oocytes [35], rat hippocampus [68], human alveolar epithelial cells [23], and rat bladders [69], etc. Melatonin can regulate Nrf2 signaling pathway-related proteins to respond to stress through its receptors [38,39]. In this study, when melatonin receptors were inhibited, melatonin could not effectively restore the expression of Nrf2 and downstream HO-1 during in vitro development of vitrified-warmed GV oocytes, which indicated that melatonin receptors were nodes of melatonin’s antioxidant mechanism. This was apparently the first report that melatonin regulated the Nrf2 signaling pathway through the receptors of vitrified-warmed mouse oocytes.Based on the above, we speculate that during the in vitro maturation of vitrified-warmed mouse GV oocytes, melatonin may exert an antioxidant role through the MT1/MT2-Nrf2 pathway. Similarly, melatonin can also enter oocytes through osmosis, because when the receptors are completely inhibited, the regulation of melatonin on Nrf2 pathway-related genes and proteins is partially reduced. Either way, the Nrf2 protein is an important node of melatonin action.5. ConclusionsVitrification of mouse GV oocytes could alter the expression of Nrf2 signaling pathway-related genes (Nrf2, SOD1) and proteins (Nrf2, HO-1), disrupt the redox system, damage spindle structure, and decrease oocyte development potential during in vitro maturation. The addition of 10−7 M melatonin may act on the Nrf2 signaling pathway through its receptors, promote the expression levels of genes and proteins to return to normal, reduce the oxidative stress and spindle damage of oocytes induced by vitrification, and increase the in vitro maturation rate of oocytes (Figure 11).

animals : an open access journal from mdpi

10.3390/ani13111850

PMC10251938

Pet dogs are more prone to exhibit challenging behaviors than ever before. Dog trainers are increasingly tasked with helping pet owners resolve behavior issues, not just teach their charges good manners. The interventions used by professionals to help ameliorate behavior complaints must be evidence-based and include the effectiveness of the intervention, how the intervention is perceived by the learner, and how the intervention affects the learner’s quality of life before, during, and after behavior intervention procedures. The objective of this paper is to review literature from multiple scientific disciplines and demonstrate how concepts from applied behavior analysis and the animal welfare sciences can be used together to ensure that the animal undergoing intervention experiences good welfare during the training process.

Social validity refers to the social significance and acceptability of intervention goals, procedures, and outcomes. Animal practitioners, who are often guided by the principles of ABA, lack the benefit of verbal participants (at least with respect to target animals) with which to assess a client’s needs and preferences. The study of a learner’s welfare is useful for determining areas where intervention is needed or how the learner feels about an intervention that is underway. Three tenets of animal welfare measurement include physiological function, naturalistic behavior, and affect, where affect refers to private events, including emotions, which are a function of the same variables and contingencies responsible for controlling public behavior. The development of new technologies allows us to look “under the skin” and account for subjective experiences that can now be observed objectively. We introduce the reader to tools available from the animal welfare sciences for the objective measurement of social validity from the learner’s perspective.

1. IntroductionScientist practitioners have an ethical responsibility to possess a comprehensive understanding of behavior, without which behavior analysis principles may be inadequately applied. It is also the duty of scientist practitioners to understand how other behavioral sciences contribute to the application of behavior analysis, educate the public about a learner’s needs in order to change public opinion and action [1], and disseminate knowledge about interventions that are both effective and ethical. Animal practitioners, who are often guided by the principles of applied behavior analysis (ABA), lack the benefit of verbal participants (at least with respect to target animals) with which to assess a client’s needs and preferences. To do so, animal practitioners (trainers, behavior consultants, and animal behaviorists) incorporate tools developed by animal welfare scientists [2,3] to assess how animal’s feel about their own circumstances and overall quality of life (QOL). Research about the domestic dog provides insight into their cognition, including but not limited to their sensitivity to, perception of, and relationships with humans, interpretation of human vocal and physical cues, discrimination learning, executive functioning, spatial and visual processing, memory, sense perception, umwelt, and the way cognition changes throughout development and aging [4,5,6,7,8,9,10,11,12,13,14,15]. This research is critical for our understanding of the salience of different stimuli to the domestic dog [16,17], without which it would be impossible to design living environments in which they can succeed and thrive, nor learning environments that they find reinforcing. When designed from a multidisciplinary perspective including cognition, ethology, welfare, and behavior analysis, interventions can modify interactions between animals and their conspecifics or humans [18,19,20,21], resolve unwanted behaviors [22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37], temporarily increase enrichment usage, and expand the learner’s behavioral repertoire [38] while ensuring good welfare. When an intervention’s scope does not consider how the intervention impacts the learner, interventions focus on the elimination of behavior via punishment and aversive control [39], leading to outcomes that include fear, anxiety, pain, stress, aggression, and a negatively impacted dog–handler relationship [18,32,34,40,41,42,43,44,45]. Effectiveness is important, especially as it pertains to social validity [46,47,48]. However, the learner’s experience of the intervention matters, particularly as it relates to their choice to participate in any procedure [49,50].Until recently, the exploration of canine behavior interventions relied on group design rather than the robust technologies of behavior change developed by behavior analysts [51,52]. Much of our understanding of training success comes from owner questionnaires, surveys, or between subject studies that focus on measures of central tendency (means, medians) rather than individual differences, and often do not meet best practices for skill acquisition and behavior change, or fail to report on metrics including operational definitions, criteria setting, generalization, maintenance, treatment integrity, and interobserver agreement. Many interventions are under-developed and fail to report adequately on what was carried out and what components contributed to the intervention’s success or failure; much of the animal intervention literature is therefore evidence-inspired rather than evidence-based [53]. For a review of the common ways behavior interventions are designed, implemented, and reported in the animal literature, see [53] (Table A2).Over the past 10 years, functional analysis (FA) was used to examine the variables responsible for various problem behaviors in companion dogs prior to the implementation of function-based treatments. Analyses and their function based interventions successfully treated jumping up behavior [23,31], stereotypic behavior [29], resource guarding [27], kennel aggression and leash pulling in a shelter environment [37], demonstrated that access to the owner functions as a reinforcer for pet dogs [54], and showed that owners are capable of implementing FA and treatment of mouthing behavior [35]. Others demonstrated the success of interventions grounded in behavior analysis, using baseline measures to demonstrate treatment effectiveness [22,24,26,28,32,55]. As the body of literature pertaining to dog welfare and behavior interventions increases, so does the commitment to honoring the human–animal bond. It is easier to encourage the use of interventions that are both function and welfare focused when those interventions improve the lives of both animals and the humans who care for them. Yet the measurement of an animal’s welfare before, during, and after intervention, which is necessary to ensure that an intervention is having only a positive effect on the learner, is yet to become standard practice. The objective of this paper is multifold: (1) to introduce the applied behavior analysis community to tools available from the animal welfare sciences for the objective measurement of affect, (2) to illustrate the importance of programming for true choice to determine how an animal feels about an intervention that is underway, (3) to propose that measurement of a learner’s welfare be incorporated into data collection and analysis before, during, and after behavior intervention, and (4) to encourage multidisciplinary collaboration.2. Animal WelfareWelfare describes the state of an animal at a given point in time. The study of a learner’s welfare is useful for determining areas where intervention is needed or how the learner feels about an intervention that is underway. In this vein, the “Five Freedoms” model [56,57,58] was introduced to alleviate suffering experienced by captive farm animals and provided an early framework for identifying stimuli that negatively impact animal welfare and cause negative affective states. Affective states are subjective states that are experienced as pleasant or unpleasant rather than hedonically neutral [58]. In the “Five Freedoms” model, welfare was measured by individual animals’ freedom from experiences such as thirst, hunger, discomfort, pain, injury, disease, fear, and distress, and the freedom to express “normal” behavior (i.e., species-typical or otherwise individually desired responses) [56,59]. This preliminary model labeled welfare “good” if the animal was not suffering and had the ability to cope with their environment. This early foray into animal welfare led to interventions that helped move the needle from poor to improved welfare states and paralleled the concurrent behavior analytic shift toward focusing on what skills a learner needs to acquire rather than which responses need to be eliminated [60]. The “Five Domains” model, an extension of the “Five Freedoms” model, was developed to assess the compromised welfare of animals used in research, teaching, and testing [61]. The “Five Domains” model outlines five areas of potential welfare compromise (i.e., nutrition, environment, health, behavior, and the mental states arising from these factors). This framework is consistently updated, with measures of positive welfare and the ability to thrive now included [62,63,64,65]. Most recently, Mellor et al. [2] amended the domain ‘behavior’ to ‘behavioral interactions’, expanding the domain to include behaviors that contact not only the environment and conspecifics, but also humans themselves. Domains 1–3 of the “Five Domains” model allow for the systematic identification of external circumstances, and the internal physical states and affects associated with them, of survival–critical behaviors related to nutrition, the environment, and health. Domain 4 allows for the systematic identification of situation-related behaviors and the effects associated with them, and highlights the importance of agency. For example, barren environments, inescapable sensory impositions, and choice restraints are labeled as welfare compromising, where free movement, exploration, and foraging are labeled as welfare enhancing. Domain 5 allows for the systematic identification of a learner’s mental state as it is brought about by the previous four domains; constraints on an environment-focused activity are associated with depression, where congenial sensory inputs are associated with gratification. This framework mirrors the assertion made by radical behaviorists that responses, including private events, are caused by the environment. Each domain is to be assessed in relation to the animal’s “species-specific behavior, biology, and ecology considered in relation to their specific physical, biotic, and social environment” [66]. The model can be used to identify a wide range of welfare states, grade welfare compromise and enhancement, enable and monitor interventions when corrective action is required, engender empathy towards learners through the identification of the wide range of positive and negative welfare-relevant experiences that can be identified, and facilitate the consideration of quality of life when the model is used repeatedly over time. The “Five Domains” were used for animal welfare management, particularly in zoo settings, and there is no reason why this model could not be incorporated into any setting in which a learner is under the care and control of others. This is especially true when the learners’ behavior is being targeted for intervention. Just as human beings “accept no limits to their own wellness” and “cannot be too well” [67], setting the events for optimal welfare should be the goal of any individual responsible for the care of any learner. Animal welfare is an increasingly public subject, with the perception of an animal’s welfare used by the public to determine whether a procedure or system is acceptable. The same environments and responses may be deemed acceptable in one scenario or unacceptable and targeted for modification and intervention in another. Ethics guides decisions about what others feel is tolerable for the organism. This can be in alignment or at odds with what the organism values and needs, with welfare compromise likely when the values of the organism and others are at odds. For example, beagles are known for their tenacity for following a scent. When used for a human’s purpose (i.e., sniffing out bed bugs, bombs, or drugs), sniffing behaviors are valued. When a pet owner is attempting to walk their beagle on a leash, incessant pulling is not valued and may lead to frustration and frustration-related human behaviors, such as yelling or yanking on the leash. Where social validity historically placed the values and concerns of others above the learner and was therefore measured subjectively, welfare is both evidence and values based [68]. The lens through which welfare is viewed, and by which welfare measuring technologies are developed, are also influenced by an individual’s philosophy and morality, which are shaped by the mores of their society, tribe, family, and personhood. This highlights the importance of ensuring that welfare is measured objectively, with what the learner values included within the measurement tool. While the movement toward “freedom to” and away from “freedom from” can be seen in the recent animal welfare literature [3,65,65,66], veterinarians, who are the primary source of information for most pet owners, currently consider the ability to cope as the parameter by which to assess if a companion animal has “good” welfare (American Veterinary Medical Association, n.d.-a). Yet the ability to cope does not define or guarantee good welfare. Coping is defined as having control of mental and bodily stability [69] or having the ability to adapt, where adaptation is the use of regulatory systems, with their behavioral and physiological components, to help an individual cope with its environmental conditions [70]. The inability to cope indicates welfare compromise, and as such frames veterinarians to focus on suffering reduction, which is a noble goal that aligns with the pathological framework from which they work [71]. The inability to cope with the environment is associated with reduced behavior variability, reduced species-specific behaviors, and the presence of stereotyped or abnormal behaviors [51]. Failure to cope is reflected in reduced fitness measured by shorter life expectancy, increased intervals between breeding periods, and fewer offspring [70], which are not metrics that veterinarians can easily measure. Physiological measures of stress, including temperature, heart rate, and respiratory rate can be used alongside visual inspection of the pet’s body condition [72], pain scales [73,74], and body language, but veterinarians must rely on owners conveying their concerns about their pet’s behavior, as QOL assessments for clinical practice are used only during end-of-life care to guide euthanasia decisions [75,76,77]. These surveys ask pet owners to rate easily operationalized concepts, including hurt, hunger, hydration, hygiene, and mobility. They also ask pet owners to rate happiness and “more good days than bad”, which are subjective and difficult to quantify without more robust, operationalized definitions or the inclusion of more systematic metrics. While they provide a guide for tracking an animal’s decline from terminal medical diagnosis to euthanasia, they do not include baseline measures when the animal is healthy, nor robust positive welfare indicators, so are insufficient for identifying how the welfare of apparently healthy companion animals could be improved through interventions or for guiding decisions regarding behavioral euthanasia.Outside of the veterinary field, updates to welfare assessments aim to assess the animal’s QOL, where QOL is “the emotions of the dog both positive and negative, physical fitness and health, and the ability and opportunity to perform natural behaviors” [3]. The aim of welfare assessments is to facilitate the creation of environments and interventions that promote a good QOL, or a life worth living [66]. Doane and Sarbeno [78] proposed a model for dog welfare based on the five domains model, excluding the first two domains, and adding a sixth domain to evaluate the QOL of the dog owner. The authors found it was possible to construct a reliable questionnaire using questions from three places; a modified Canine Behavioral Assessment and Research Questionnaire (C-BARQ) [78,79]; a validated tool used in applied clinical settings to assess the severity of problem behaviors in dogs, a dog QOL questionnaire [76], and an owner QOL questionnaire [80]. One hundred eighty-five questions examined the typical behavior of the dog, including behaviors associated with fear, anxiety, aggression, excitability, attention, attachment, touch/pain sensitivity, and trainability, in addition to questions about the owner’s emotions, social and physical quality of life and stress, and the happiness, physical functioning, hygiene, and mental status of the dog. This pilot study demonstrated that this questionnaire may be suitable for further development as a tool for dog welfare assessment, where scores could illustrate areas of decreased welfare for both the dog and their caretaker, and when repeated aid in the assessment of how a behavior intervention impacts stakeholders. However, in order to create a tool that is not only accurate, but also likely to be adopted, further consideration should be given to the number of questions asked and time needed to complete and review the questionnaire.2.1. Welfare and EnvironmentMeasuring response allocation allows us to see how much time is spent, or the percentage of time spent, performing a response. Response allocation changes when an environmental change results in increased or decreased time spent performing the response. The occurrence of an event (the environmental change) is an inducer, which is a stimulus that occasions an (induced) activity. Unlike a discriminative stimulus or an elicitor, a close temporal relation is not required between the inducer and its response. A phylogenetically important event (PIE) is an unconditioned inducer that directly affects survival and reproduction. Thus, a PIE induces PIE-related activity, which includes operant activity that is related to the PIE as a result of the contingency [81]. Fitness-reducing PIEs will induce defensive activities that remove or mitigate danger, where fitness-enhancing PIEs induce fitness-enhancing activities. Some examples include seeing a predator and hiding and seeing prey and hunting. Similarly, induction can help explain ‘instinctive drift’ (i.e., misbehavior), as observed by Keller and Marion Breland [82], where non-reinforced behavior interfered with operant learning. For example, the presence of food elicits behaviors that are not targeted, and are not reinforced by food, yet persist and “strengthen”. When picking up coins, raccoons engaged in food washing behavior, which was not reinforced and interfered with the goal of having the raccoons deposit coins in a piggy bank. Food is phylogenetically important in that it is critical for survival. Hence, food is called a “primary” reinforcer, and is also an inducer [81,83,84]. In this instance, the coins induced the raccoons to dip the item into the box and bring them back out as they would with food into water, before rubbing the items together. The problem persisted in such a manner that the raccoons allocated all their time to food washing, and Breland and Breland could not get the raccoons to drop the coins into the box. Induction is related to allocation through contingency, which links an environmental event to an increase or decrease in activity. Response allocation is related to welfare due to the importance of response variability to welfare, with enclosure use and behavioral diversity measurement used to assess how environmental variables impact an animal’s quality of life [64,85,86]. Welfare is compromised when an organism spends too much time engaging in one activity to the detriment of other behaviors that are necessary for survival and wellbeing. A few reasons this may occur in dogs includes environmental deprivation [66], unwillingness to exit the home or interact with humans or conspecifics due to fear of the environment or something in it [87], neophobia, which causes the animal to avoid a novel objects or environments, preventing exploration [88], and stereotypies, which are repetitive or sustained goal-focused behaviors that do not change much from one situation to another and which are not a normal part of the ethogram (i.e., behavioral inventory) of the organism within the given context [89]. Some stereotypies include tail spinning [90], flank sucking [91], and fly snapping [92], which can be caused by or cause medical issues [92,93]. Welfare is compromised when the percentage of time allocated to one behavior or class of behaviors prevents the organism from engaging sufficiently in rest, play, exploration, and social interaction. The identification of inducers that promote behavior that support welfare enhancement, and the elimination of inducers that promote behavior correlated with welfare compromise, including but not limited to behaviors related to fear, anxiety, and stereotypy, is paramount in order to design an environment that supports a good quality of life. Additionally, variables responsible for changes in response allocation in natural settings must be identified so they can be replicated in restricted settings, where applicable [94]. In captive settings that are naturally restricted, humans control all resources, which are commodities or opportunities to perform specific activities in order to obtain an objective [95]. Behavioral repertoires and time budgets vary based on circumstances including environmental conditions and resource distribution [95]. In this scenario, “income” is a variable (for example, time or energy) that limits the resource acquiring response. Behavioral needs are requirements that must be met to allow for the organism’s effective functioning [95]. To determine if a behavior is one that “needs” to be performed, welfare compromise must be observed after the learner is blocked from performing that behavior [96]. An analogous operant, such as button pressing, can be trained to test the “price” of a resource, where a response, such as the number of button presses performed, could be used to determine the price of some commodity, or what an individual is willing to “spend” on that resource [97,98]. For example, Feuerbacher and Wynne [54] used “depressing a large button light” and “pawing a backstop” as analogous operants for responses that typically lead to dog owner access; these responses were used in a functional analysis to establish that access to owners is functionally reinforcing. While they required the dogs to perform the response only once, if singular button pressing was put on extinction, the number of button depressions performed could be used to determine how valuable access to an owner is. If welfare compromise was observed after access to the owners was not granted, or if the button was removed, responses that lead to owner access could be labeled behavioral “needs”. Blocking a learner from performing a behavior via environmental arrangement or response extinction is only one way that humans, who are an unavoidable part of a companion animal’s environment, cause poor welfare. The behavioral needs of dogs that have been identified include running, resting, playing, exploring, and positive, consistent interactions with humans; the environment should be of adequate size and complexity for these purposes [99]. For companion animals, who were genetically modified by humans, human environments comprise dogs’ natural ecological niche, and dog–owner attachment is functionally analogous to human infant–mother attachment [100,101]. Dogs have been observed wagging their tails more when granted access to human contact compared to access to conspecifics [102], prefer petting over food when the petting is provided by their owners in unfamiliar contexts [103], play more with conspecifics when receiving owner attention [104], and can communicate with owners via “showing” behavior [12]. Dogs and humans have a mutualistic relationship, with dogs bred to serve human purposes including hunting, guarding, and herding [105]. Further, humans and dogs support each other in stressful situations [106]. An oxytocin-mediated positive loop is facilitated and modulated by dogs and their owners gazing at each other [107]. Dogs even have a muscle responsible for raising the inner eyebrow intensely, which increases paedomorphism and may trigger a nurturing response in humans [108]. This muscle is not found in wolves.All living beings that underwent natural selection are adapted to environments that allow them to thrive by every metric: affective, health, and behavior [58]; but what about animals that exist due to artificial selection? Over the last 200 years, dog breeding goals favored form over function and the behavior and personality in each breed is now highly variable [109,110,111]. Pet owners may not think of their pets, who are part of the family, as being held in captivity. Nevertheless, pets are restricted to the environment and resources provided by their humans. The natural behaviors of the domestic dog are ones that exist within their habitat, the human environment; yet dogs have widely diverse phenotypes and breeds are highly differentiated to display a wide variety of behaviors best suited to a large range of human settings [111]. The captive condition that is the pet home may not allow for the evolved adaptations of companion animals to match the challenges of their current circumstances, causing poor welfare when natural behavior no longer leads to beneficial consequences. Where herding behavior may be valuable on a farm, it is not appreciated when directed towards humans in an apartment! While pet owners’ homes should elicit natural behaviors that produce reinforcement for the dogs that occupy them, many modern pet homes lack the ability to meet the needs of companion dogs without purposeful construction, enrichment, and intervention. Bamberger and Houpt [112] found that the number of dogs exhibiting unwanted behaviors including aggression increased between 1991 and 2001, and veterinary behaviorists report experiencing an increase in the severity and complexity of their average patient (Christensen, E., personal communication, 20 October 2022). Some unwanted pet behaviors become stuck in a “sick social cycle” [113] with their owners’ behaviors. Fear and anxiety are reported as increasingly common behavioral disorders in dogs [87,112], with a prevalence ranging from 26.2 to 44% [43,87,114,115]. Fear and anxiety disorders seriously compromise the welfare of dogs and may lead to chronic stress, relinquishment by the owner, and euthanasia [87,116,117,118,119,120,121,122]. In zoos, wildlife centers, safari parks, and sanctuary settings, habitats are designed for the animals specifically and staff must adhere to basic requirements. While the level to which this is conducted is outside the purview of this paper, it is of benefit to both the animal and to conservation, research, and education efforts that animals in captive settings have good welfare [123,124,125]. Human dwellings are not built with pets’ needs in mind, and no welfare requirements exist for companion animals. However, we have an ethical responsibility to provide captive animals, including our pets, with environments that promote a good quality of life [124].Dogs that find themselves in shelters are afforded even less autonomy. When an environment does not intrinsically meet an animal’s needs, enrichment should be supplemented. Enrichment typically refers to “inputs” or manipulanda, which provides for species-typical needs by meeting adaptive relevance [126]. Enrichment is the contingency between a response and stimulus or event, where the enrichment produces an observable, measurably improved state of well-being for the animal [37]. Kiddie and Collins [127] developed a scoring system to assess how enrichment impacted the welfare of kenneled dogs in a rehoming center (i.e., shelter). The authors compiled an ethogram of behaviors associated with anticipation, play, and relaxation, which were further categorized by whether the responses were provoked or unprovoked by human interaction (approached, touched, engaged in play, and physical examination). Dogs were put into four groups, and sorted by length of stay (more or less than 30 days) and environment (enrichment vs. no enrichment). Enrichment included removing the dog from the kennel, encouraging the dog to make body contact, giving a massage, speaking to the dog in a soothing voice, grooming with a soft brush, and 5 min of clicker training. They found that enrichment increased welfare scores and the assessment itself had good content evidence and criterion validity, but poor internal consistency. Staff were trained to promote response evidence of construct validity; however, interobserver agreement (IOA) data were not taken. Training can be enriching for dogs by “affording learning opportunities, and learning is considered to be enriching” [38,128]. Training may also teach skills that afford access to additional reinforcers. For example, successfully training dogs to “come when called” allows them the freedom to be off leash to choose where to go, who to interact with, and when and where to sniff. With this skill, owners can call their dogs away from livestock, wildlife, or any dangerous or inappropriate situation and release them when it is safe to do so. In this scenario, the response “coming when called” could be considered a behavioral cusp [129], as it is a requisite skill for being safely off leash, affording access to new environments, new opportunities to engage in species-typical behaviors associated with “feeling good” [130], and contact with new reinforcers, including but not limited to access to variable scents. Additional examples of enrichment include food puzzles from which animals must manipulate an object to obtain food [131], chews and toys meant for gnawing and dissecting [132], sniffing, which induces positive judgment bias [133,134,135], and play with humans or conspecifics [136]. Enrichment was shown to positively impact behavior welfare indicators by increasing behavior variability and exploration and decreasing stress-related behaviors. The ability to perform natural behaviors that are afforded through enrichment may have a regulating effect on the dog’s nervous system, and may be a setting event for other behaviors that human stakeholders find valuable, such as resting calmly at home. Enrichment strategies are purported to be most effective when targeting the primary sensory abilities of the species concerned [137]. Dogs rely on olfaction as their primary sense [130], and olfactory enrichment modifies the behavior of both pet and shelter dogs [133,134]. After exposure to cloths scented with ginger, coconut, vanilla, and valerian, dogs displayed reduced levels of vocalization and movement, with coconut and ginger also causing an increase in sleeping behavior [133]. By contrast, dogs spent more time moving and barking upon exposure to peppermint and rosemary [138]. Preliminary, non-peer reviewed research demonstrated that when on a walk, sniffing behavior lowers dog heart rates independent of their walking activity, with sniffing intensity positively correlated with heart rate reduction [139]. With increased freedom came increased sniffing, as dogs would sniff 280% more on a long leash compared to a short leash, and 330% more off leash compared to on a short leash. Given that dogs allocated proportionately more time to sniffing in relation to their freedom, we can assume that the opportunity to sniff is of value to dogs; further research is warranted to determine if sniffing is one of a dog’s behavioral needs. 2.2. Environment Complexity and ChoiceA balance must be found between predictability, which has been shown to reduce anxiety [140,141], and the variability and novelty afforded by more diverse environments and variable stimuli that allow for animal-driven choices [126]. Increased habitat complexity, environmental enrichment, contact with appropriate social groups, and training sets the occasion for species-typical behaviors and increases behavioral diversity, which is one measure of an organism’s welfare [64]. A large repertoire of behaviors allows for an increased chance of survival, as well as multiple routes to access primary reinforcement and escape, which supports the idea that programming to expand response classes may improve quality of life. Animals who are stressed or medically unwell have lower behavioral diversity, and as such, behavioral diversity may be an indicator of positive welfare [64]. As diversity can include behaviors associated with both positive and negative emotional valence [94], an increase in behavioral repertoire only points to an improvement in welfare if new responses lead to greater contact with positive reinforcement. For example, dogs in a kennel environment displayed more behavioral diversity compared to dogs in a home environment, and spent more time standing and locomoting, and less time lying down [142]. However, kenneled dogs also spent less time resting, and five of the 29 dogs spent more time panting, which indicates an increase in distress and a reduction in welfare [142]. While behavioral diversity reduces the likelihood of behavioral restriction and increases the likelihood that we are meeting the behavioral needs of that individual, the type of responses and their consequences must be considered. When animals have an inability to engage in behaviors that they are motivated to perform, there is a welfare compromise. Restriction of highly valued behaviors contributes to behavioral and physiological manifestations of stress, as was demonstrated by Glavin et al., [143] who found that movement restriction (response blocking) caused increased heart rate in rats. Miller et al. [64] suggest considering “appropriateness” of responses and propose that husbandry programs be developed around outcome-based metrics that lead to experiences that are meaningful to that species. To do so, they suggest (a) identifying and outlining species-specific behaviors; (b) identifying the stimuli and environment necessary to elicit or evoke species-specific responses, as well as completing a task analysis to identify response components; (c) list the adaptations that allow the animal to execute the behavior; (d) determine behavioral outcomes and how they will be measured; and (e) identify the practice, structure, or techniques that will be used to achieve behavioral goals. This leads animals to have a complete experience relative to their adaptations, increases the opportunity for animal-driven choices, and reduces the need for inputs or manipulanda provided by caretakers as supplemental enrichment [64,126].Animals have been shown to prefer free choice over forced choice, which may be due to free choice improving survival odds in ancestors (phylogeny) or individuals learning that preferred activities and items are more often available in free choice conditions (ontogeny) [144]. The environment should be designed to induce behaviors associated with good welfare, where sensory stimuli exist on a temporal schedule that reflects what the animal would experience in their natural setting and variability promotes positive experiences. For example, foraging behaviors change throughout the year based on the season and availability of different types of food, number of resource sites, types and amounts of food available at various resources sites, distance between resources sites, and competition for resources [126,145]. While it is easy for caretakers to vary the amounts and types of foods available, it is more challenging to determine how to set the environment to promote a wide variety of foraging behaviors and ensure that the foraging setup is appropriately challenging so that the activity is fun, not frustrating. Using two Treat & Train® [146,147] food delivery apparatus which functioned as resources sites, Salzer and Reed [145] examined the foraging behaviors of companion dogs by using a free operant arrangement in a daycare setting, changing the variable–time schedules of food delivery. They found that the dogs distributed themselves to maximize the delivered resources. However, given that the food was delivered on a variable interval schedule that was not also contingent on foraging behaviors, and the delivery apparatus were freely visible in a barren daycare setting, conclusions are limited. The utilization of a compound schedule of reinforcement, combining both responses and time, where responses targeted are foraging behaviors or analogous operants for foraging behaviors, would allow more information about how to use this system in a home environment to provide for more varied experiences in a restricted setting.2.3. Genuine ChoiceNeither freedom nor coercion exist in duality, but instead lay on opposite ends of a spectrum. Given any environment, an individual has a limited number of response options available based on their skillset, setting events, and present contingencies in effect. In behavior analysis, freedom is the availability of alternative responses, where at least two well defined behavioral alternatives are necessary for genuine choice to exist [148,149,150]. Choice responses are a function of antecedent, behavior, consequence (the ABCs), and their histories and the ABCs of alternative available patterns of responding and their histories [148,149,150,151]. In an effort to ensure that captive organisms have genuine choice, all contingencies currently available must be identified and the response costs and consequential benefits of alternative behavior patterns weighed. The freedom an individual has can be measured as a degree of freedom represented by n contingencies −1 [148,149,150]. If there is only one contingency in effect, this is represented by 1 − 1 = 0, demonstrating that the individual is not free, as there is no choice available. If there are two contingencies in effect, this is represented by 2 − 1 = 1 and the individual has one choice available or has one degree of freedom. In this way, freedom lies on a spectrum, and we can make our learners “more free” by teaching additional responses and providing additional reinforcers, and setting the environment so our learners can use their skills to access reinforcement in their natural environment. Each additional behavior available to an individual increases genuine choice, or the ability to choose, without coercion, from equally possible contingencies that are simultaneously available only if those contingencies exist in the environment (for a discussion of the definitions of the terms control and freedom by Skinner, Baum, Catania, and Goldiamond that are “consistent with the epistemological assumptions of radical behaviorism”, see de Fernandes and Dittrich, 2018 [148]). Alternatively, it is possible to make our learners “less free” by providing reinforcement contingent on only a small number of responses. Worse, if a behavior that is required is one that is inherently aversive to the learner, we are requiring a forced choice behavior [152]. For example, in “nothing in life is free” [153,154], dogs are taught to perform a skill, such as a sit, to earn everything that they need, including food. In this scenario, where some trainers require the dog to perform one or two specific skills to access reinforcers necessary to maintain life, the dog is lacking genuine choice. If the dog sits naturally on their own, while having 0 or 1 degree of freedom is not ideal, teaching the dog to sit is not causing additional harm. However, if the dog never sits of their own accord due to their morphology, an injury, or other circumstance, asking the learner to sit may cause pain or other aversive conditions for that animal in addition to the reinforcement they receive. Forced-choice behaviors are necessarily coercive, and as such should receive a negative value when degrees of freedom are measured. In the instance where a dog must sit for their dinner, but sitting is painful, the dog would have −1 degrees of freedom. Unfortunately, illness and injuries happen and are not always immediately apparent. Programming for multiple response options reduces the chance that a well-meaning handler inadvertently forces their dog into this unpleasant, coercive situation. An individual’s degrees of freedom could be an excellent welfare metric. In circumstances where a learner is living in a restricted setting (institutionalized, kenneled, or even living in a pet home), others are responsible for teaching skills, arranging the environment, providing opportunities, establishing the conditions that make consequences critical, and ensuring contingencies that provide access to those consequences are available. When multiple contingencies are equally possible, the learner has true choice [148,150]. In a scenario where a dog is required to perform any of a wide array of skills to access reinforcers necessary to maintain life [155], choice exists. The identification of both behavioral needs and forced choice behaviors is necessary for an objective assessment of the social validity of any behavior selected for intervention; therefore, degrees of freedom should be used alongside other indices to measure welfare. 2.3.1. Choice and InterventionsHanley et al. [156] demonstrated that given a choice, participants with severe problem behavior (self-injury, aggression) preferred participating in a functional communication training (FCT; a differential reinforcement (DR) procedure in which an individual is taught an alternative response that results in the same class of reinforcement identified as maintaining problem behavior) plus punishment treatment package over FCT or punishment alone. The authors provided a choice to participants by presenting three microswitches that when pressed, functioned as the first step in a concurrent chain leading to different treatments. This created a person-centered intervention, and allowed for the objective measurement of the social validity of the treatments themselves. In addition to participants’ preference for this treatment, the combination treatment was demonstrated to be most effective. Rajaraman et al. [157] replicated and extended Hanley et al. [158] by incorporating an enhanced choice model to offer participants multiple choice-making opportunities, which included (a) participating in treatment involving differential reinforcement, (b) “hanging out” with noncontingent access to putative reinforcers, or (c) leaving the therapeutic space. Participants overwhelmingly chose to participate in treatment, which was successful, demonstrating that it is possible to eliminate dangerous problem behavior that previously escalated during attempts at physical management, while providing participants with the agency and control that is necessary for good welfare. In addition, by giving participants the choice to participate, Rajaraman et al. demonstrated how social validity can be placed at the forefront of a treatment package and objectively measured. The authors implemented this enhanced choice model in both an outpatient clinic and a specialized public school, and generalized skills across teachers, classrooms, and time periods, providing further evidence that skill-based treatment provided through an enhanced choice model can produce socially meaningful behavior change. Finally, Ramirez [159] demonstrated how to reduce the task refusal of a beluga whale by implementing a choice protocol. The whale was previously an active participant in training sessions but was no longer performing difficult and husbandry behaviors such as presenting her tail for a blood draw. After hypothesizing that the whale lost trust in the younger trainers on their care team, Ramirez provided the whale with the opportunity to perform a low-effort behavior (target) for reinforcement at any time during a training session, even if another behavior was requested. This gave her a way to say “no” when asked to perform difficult skills, and ensured there was always an alternative behavior she could offer at any time to earn reinforcement. The whale’s degrees of freedom were increased from 0 to 1, and her refusal to do behaviors was reduced from 38% to 2%. While these three examples demonstrate how behavior intervention plans can be designed with the concept of genuine choice in mind [156,157], designing a complex environment to provide the animal with multiple biologically relevant options from which to choose [124], allowing learners to choose from multiple intervention options [156], providing an additional way for learners to earn reinforcement during interventions beyond meeting target criteria [159], and giving learners the ability to opt out of the learning environment altogether [157] are not yet common practices. 2.3.2. Control over Pain The need for control is biologically motivated, adaptive for psychological functioning and physical health, and imperative for survival and wellbeing [160]. Choice is necessary for an individual to have a perception of control, but there is a difference between the perception of control and making choices [160,161]. The perception of control over a stressor inhibits autonomic arousal, and a vast majority of humans view having a choice as positive [160,161]. Passivity is the default, unlearned response to a prolonged aversive event that can be overcome when a learner learns how to control the aversive event [162]. Crombez et al. [163] demonstrated that having control over pain reduces vigilance in humans. When attempts to avoid pain were blocked, the participants persisted in their avoidance attempts, tried harder, narrowed their focus of attention to avoiding pain, experienced higher fear of the impending pain stimulus, and had reduced performance on a secondary task. Researchers recommended that future studies include measurement of psychophysiological responses (skin conductance and heart rate) instead of using participants’ self-reported pain levels, which suggests that this type of research could be replicated with animals if these measures are used.Operant learning can cause a pain modulation effect. Lee et al. [164] taught 21 human participants to select from low or high pain cues, which would be followed, respectively, by mildly or severely painful stimuli. After this training, participants who selected low pain cues expected to receive mildly painful stimuli, but instead received the high pain stimuli. Participant’s pain ratings were significantly reduced when receiving the high pain stimulation after selecting the low-pain cue, compared to receiving the high pain stimulation after selecting the high pain cue. This may explain why animals who are given a choice to participate in medical handling do so even when the handling is occasionally painful. Many repetitions of pain-free trials, in addition to the choice component of the intervention, may have a similar pain modulating effect as what Lee et al. [164] demonstrated with humans. 2.3.3. Pain during InterventionIn 2004, Schilder and van der Borg [45] examined the behavioral effects of the use of a shock collar for working guard dog training using Malinois, Malinois cross, German Shepherds, and one Rottweiler. Methods used for training were determined by the trainers, and the authors report that dogs who were not trained with shock were trained with prong collars, beatings, kicks, and choke collar corrections. The most frequent administration of shock occurred after a dog was asked to do something and before the dog had the opportunity to respond to the request. Dogs were most often shocked for not “obeying” a ‘let go’ cue, heeling ahead of the handler, biting a criminal at the wrong moment, and long duration latency when told to ‘heel’. After observing changes to body language that indicate fear and distress, the authors concluded that training itself is stressful. However, all dogs observed were trained with strategies that include the application of pain, or at the very least, discomfort. The authors examined the dogs’ responses during sessions where no shock was applied and found that training with shock leads to reduced welfare post-training. In addition, behavioral responses suggested that shocked dogs made the association between presence of their owner and shock, and as such, the welfare of shocked dogs is reduced in the presence of their owner. The natural conclusion is not that training itself is stressful, but that training with pain causes stress and reduced welfare. As the dogs used for this study were working dogs in formal training programs, we can assume that these dogs were bred for this work. Dogs bred for working are selected to have the morphology and temperament necessary for their work. When breeding is carried out well, working dogs are genetically predisposed to performing working tasks that they find reinforcing [165]. For example, untrained Malinois puppies bite and hold objects while they are lifted off the ground, while Labrador retriever puppies do not do this when presented with the same stimulus [166]. This natural inclination toward the tasks being trained may provide a buffer from the effects of harsh training methods, as reactions to potentially stressful events depend on their meaning for the individual [167]. It is possible, but not yet demonstrated, that when pain is applied to a learner who may not have a genetic inclination toward a behavior being trained, pain may have more severe consequences. The degree to which dogs exhibit reduced welfare is positively correlated with the level of aversive control of the operant procedure used [34]. Available information suggests that the use of aversive-based methods in training is correlated with stress-related behaviors during training, fear, aggression, elevated cortisol levels, and reduced owner interaction, all of which indicate welfare was compromised [168]. Many of the completed studies are survey-based, with populations of interest that are police and/or laboratory dogs that do not represent the spectrum of breeds and temperaments of companion animals. Most studies examined the use of a shock collar specifically [168], did not include objective descriptions or measurements of training methodology, and lacked data about procedural integrity and interobserver agreement. While the data suggest that use of a shock collar is correlated with welfare compromise, we are unable to conclude that all aversive-based methods are associated with welfare compromise. Further, while most dogs demonstrated reduced welfare after experiencing harsh training methods, some did not. Many factors may contribute to this finding, including but not limited to the dog’s breed [99,109,165,169,170], physical conformation [99], temperament [99], sex [171], age, [36], physiological measures, medical conditions including behavioral diagnosis [172], rearing and learning history [18,33,34,40,99], overall quality of life before, during, and after training, owner attachment and caregiving style [172,173,174,175], dog’s motivation to perform the skill being trained, if the skill being trained with shock ultimately allowed for greater access to reinforcement, level of shock or type of aversive, number of aversive applications, inclusion of reinforcement, training methodology (appropriate criteria setting, correct timing, etc.), and procedural integrity. In military dogs, suspicion of previous rough handling, along with less time spent with the handler, was associated with fear and aggression, while dogs that lived with their handlers were more social [174]. Further, the obedience of military dogs was greater, fewer bites to military staff were reported, and dog welfare was improved when the dogs lived with their handlers and practiced a sport. 3. Welfare and Private EventsTo work towards a good quality of life, it is essential to understand concepts such as “subjective experiential states” or “situation-related affective states” [3,66] which refer to private events, including emotions, which are a function of the same variables and contingencies responsible for controlling public behavior [176,177,178]. Behaviors are associated with subjective experiential states such as pain and pleasure [3,179,180]. Negative affective states include “breathlessness, thirst, pain, hunger, nausea, dizziness, debility, weakness, sickness, anxiety, fear, frustration, anger, helplessness, loneliness, and boredom”, where positive affective states include “feeling energized, engaged, affectionately sociable, maternally rewarded, nurtured, secure, protected, excitedly joyful and/or sexually gratified” (see [66], Figure 2). Positive behaviors indicate anticipation, decision making processes, problem solving, investigation, preference indications, and agency, where negative behaviors indicate anxiety, fear, pain, or boredom [94]. In order to evaluate dogs’ affective states, researchers explored the relationship between mood states and stimulus appraisal [181], the way that emotionality impacts information processing to create cognitive biases [182], and how cognitive bias may indicate emotional valence [183]. Cognitive bias testing was used to determine how training methods affect dogs’ optimism/pessimism [34]. Dogs demonstrate behavioral laterality in response to emotional stimuli [184,185,186,187,188]. For example, there is a relationship between high levels of canine facial asymmetries and emotional and physiological distress, with dogs displaying higher levels of facial asymmetry when approached by an unfamiliar human compared to when approached by their owner or when they are alone [189]. When reunited with their owners, dogs have been shown to move their left eyebrow [190]. When presented with their owner, an unfamiliar person, and a cat, dogs wag their tails to the right, but when presented with an unknown, dominant (as established through unspecified behavioral tests) dog, they wag their tails to the left [191]. This finding was supported by Siniscalchi et al. [190] who found that dogs respond to conspecifics differently in relation to the conspecifics wagging laterality; when seeing other dogs who wag with a left-biased tail, dogs experienced increased heart rate and displayed anxious behaviors (i.e., lowering of body posture, paw lifted). This demonstrates not only that motor lateralization bias should be considered for measuring emotional valence, but that dogs are able to detect emotion in conspecifics when lateralization is present. While affective states may be considered constructs when viewed through the lens of applied behavior analysis, they are helpful concepts from which operational definitions can be created for individual assessments of captive animals, including companion animals. The topography of a response differs based on the context, which informs how an organism is feeling; lying down on the couch looks different than holding a down stay on a platform during an agility trial. The former is looser, with relaxed musculature, different ear and tail positions, and a larger percentage of the body in contact with the resting surface. The latter is tighter, with eyes fixed on the handler and all four limbs tucked underneath the body, as well as muscle tension sufficient to quickly launch the dog into the next response in the chain. Similarly, a dog who is comfortable having a picnic in the park with their owner might lay down on the grass in a relaxed manner as described above, where a dog who is uncomfortable in that environment would hold their body upright, primed for action should a threat appear. Further analysis into the dog’s facial expression can help the delineation of arousal from emotional valence [108,189,192,193,194]; in an agility context, a dog may be aroused and excited where in the park they are aroused and worried, as is elucidated by eye shape (round, almond, and squinting), pupil dilation, sclera visibility, the presence or absence of creases in facial muscles, tension in and position of the commissure and ears, or changes to any of the above. A two-dimensional model can be used to describe the affective state of animals by differentiating between emotional valence and arousal. Arousal can range from low to high, and valence ranges from positive to negative. There is a reciprocal relationship between the identification of affective states and the identification of contingencies, where the identification of one allows for the inference of the other [176]. Layng [176] proposed that private emotions are contingency indicators or descriptors which function as tacts of consequential contingencies, providing insight into their identification. Further, he posits that Panksepp’s seven types of emotions [195] describe different patterns of reinforcement. Seeking is associated with nearing an occasion for reinforcement, anger/rage is associated with removal of the other, fear is associated with the removal of oneself or the other, lust is associated with nearing an occasion for a sexual encounter, care is associated with removing distress signals, panic and grief are associated with nonspecific distancing, and play is associated with nearing reciprocal social or activity related consequences. The subcortical brain regions that “light up” during imaging when a human is experiencing an emotion exist in all mammals [196], suggesting that these brain regions, and subsequently emotional states, are homologous across species. Emotions occur with and are determined by the contingency and are accompanied by autonomic, physiological changes, which may be necessary to meet the requirements of that contingency to aid survival [176]. Behaviors start as operants, and when those behaviors contribute to survival and reproduction, the morphological and physiological structures necessary for the behavior are inherited by offspring [176,197]. Potentiating variables may not elicit respondent behavior associated with emotion, but instead might evoke canalized operants that produced reinforcement in the past [176]. These concepts echo the idea that feelings cannot be separated from other biological mechanisms when individuals are trying to cope with their environment [95]. Both contingency analysis and the “Five Domains” model provide frameworks for the assessment of environmental variables that influence mental state. Welfare scientists now tackle the challenge of how to measure and quantify an affective state (or covert responses and private events; [198,199]), an ability that Skinner predicted [200]. Many researchers looked at the correlations between physiological stress responses and the overt behavior of dogs during various conditions to infer affective state. Physiological, endocrine, and neural measures included immunological markers, salivary cortisol, urinary oxytocin, heart rate (HR), heart rate variability (HRV), body temperature, and respiration rate. These biological measures were compared to both dog body language (a dog’s verbal behavior) and responses such as approach and avoidance to determine if biological and behavioral responses consistently correlate with one another or are, on their own, accurate measures of the third variable, the animal’s affective state. 3.1. Canine Verbal Behavior and Emotional ValenceIt is necessary to determine the extent to which an animal’s behavior consistently conveys emotional valence both between and within subjects across conditions and time. Given that all responses, once elicited, contact the environment, it is critical to determine if responses exist that are not affected by environmental consequences and whose appearance always elucidates an underlying emotional response. Behaviors linked to physiological indicators of acute stress include startling, lethargy/decrease in activity levels, increase in activity levels, loss of appetite, reduced playing, hunched posture, low tail position, paw lifting, yawning, ears held low, trembling, snout/lip licking, lowered body positions, vocalizing, panting, increased salivation, repetitive behaviors, increased activity, ‘nosing’, increased urination, decreased drinking, increased salivation, increased vigilance, reduced responsiveness to humans or previous reinforcers, increased self-grooming, self-mutilation, coprophagy, hiding, destruction, change in elimination patterns, aggression, repetitive behaviors, and a change in the rate or frequency of these behaviors, in particular a lower threshold for showing any of these behaviors over time, which indicates chronic stress [99,201]. In reaction to receiving a shock, dogs demonstrated the following behaviors (in order of decreasing response frequency): lowering their ears, emitting a high-pitched yelp, flicking their tongues, lowering their tails, lifting their front paws, squealing, exhibiting “characteristic head movement” and avoidance behaviors, scream barking, crouching, snapping at their owners, or exhibiting no reaction [45]. After subjecting dogs to six different aversive stimuli, including sound blasts, short electric shocks, and a falling bag while measuring heart rate, salivary cortisol, and behavioral responses, Beerda et al. [201] found that with the exception of very low body posture, which was correlated with elevated salivary cortisol, the correlation between behavioral and physiological stress parameters were not significant. When the aversive stimuli included the presence of the experimenter (i.e., the dog is pushed down and held for 20 s), high levels of oral behaviors were observed. The authors indicate that oral behaviors functioned as a signal of submission or appeasement rather than a signal of stress; however, it is not clear how these constructs could be abstracted from one another. Camps et al. [172] retroactively examined 12 clinical cases to examine the features of pain-related aggression. Owners reported that dogs who display aggression after becoming painful exhibit a reduction or complete lack of warning prior to an aggressive display (labeled “impulsiveness”) compared to dogs who display aggression before or without becoming painful [172]. No relationship was found between cause of pain and the topography of aggressive behavior; however, dogs who were not aggressive prior to their painful condition showed aggression as a result of manipulation context more frequently. While [172] reported that pain-related aggression is a primary problem in only 2–3% of dogs who are referred to a behavioral specialist, Dinwoodie et al. [43] examined 963 dogs whose owner reported at least one aggressive response and reported 15% of dogs had an underlying medical problem. After reviewing 100 caseloads, Mills et al. [202] state a conservative estimate of behavior cases that involve pain to be 33–80% and postulate an under-reporting of the ways pain can be associated with problem behavior.Appeasement signals are incompatible with aggressive behavior and are “context-and response-dependent sequences which are part of a ladder leading to threats or overt aggression when ignored” [203,204]. Lip licking and gaze aversion are two intraspecific appeasement signals that may also serve as an appeasement signal in dog–human communication [203]. While the term “appeasement” is mentalistic, these signals happen frequently in threatening and conflict-ridden situations [203,205], and when negatively reinforced, escalation to more dangerous distance-increasing responses are not necessary and do not occur. When not reinforced, these signals may be followed by overt aggression, including snapping and biting [205,206]. The most common behavioral changes observed before a bite include holding the body low with ears in a non-neutral position, head turning, and panting, with stiffening, staring, frowning, and snapping occurring closest to the bite [206]. Head turning, staring, and snapping decrease in frequency, plateau, or fluctuate directly before a bite occurs, with growling and “restrained” behavior observed immediately preceding a bite. Camps et al. [172] reported that dogs who displayed aggression after acquiring a painful condition were more “impulsive”, meaning they displayed minimal or no warning signals prior to an attack. In situations that experimenters assume to be exciting (successful problem-solving results in reward delivery), dogs demonstrated increased tail wagging and overall activity; in situations experimenters assume to be frustrating (reward delivery is unpredictable and independent of a dog’s behavior), dogs chewed the operant device available to them [102]. The relationship between consequence and behavior was consistent across reward types, which included food, social contact with a familiar human, and social contact with other dogs. The mean latency to enter the training area was negatively correlated with mean tail wags; if the dogs moved quickly to begin their task, their tails wagged more frequently. Tail wagging was historically viewed as an indicator of arousal, while tail position, which was not recorded, was historically viewed as an indicator of emotional valence [45,189,190,207,208]. The number of tail wags per second, influenced by expected reward type, was correlated with positive affective state. A less mentalistic interpretation is that tail wagging rate may be correlated with the positive value of an expected consequence. No condition included a predictable aversive consequence, so an inverse relationship between wagging rate and the negative value of an expected consequence is yet to be demonstrated, and tail wagging lateralization, which was found to correlate with emotional valence [189,190,191], was not reported. No correlation was observed between reward type and chewing frequency when the reward delivery was unpredictable. In a similar study examining facial expression, Bremhorst et al. [194] found that ear adduction was more common when a high-value food reward was delivered 5 s after the food delivery apparatus was approached, while blinking, lips parting, jaw dropping, licking nose, and flattening ears were more common when the food was withheld for 55 s. While flat ears were previously shown to signal fear, within the context of food being withheld, they may also signal frustration.In an aim to assess a reliable measurement of fear responses in pet dogs, Ogata et al. [209] measured behavioral changes, heart rate, and body temperature during a respondent aversive conditioning procedure and found that both a remote-controlled spray collar and the conditioned aversive stimulus (a buzzer) induced significant heart rate and body temperature increases. In the first minute after the spray was presented, dogs exhibited lowered tails, running, whining, freezing, looking towards the sound, oral behaviors, panting, and jumping. Dogs in the treatment group continued exhibiting these behaviors when the buzzer was presented alone. Importantly, each dog displayed different behavioral responses both across and within subjects (see [209], Table 1), demonstrating that dogs’ behavioral responses cannot be accurately assessed at the group level. As found in previous work, behavior changes were not consistently correlated with physiological responses and autonomic changes after conditioning were found in dogs whose behavior remained consistent, indicating that autonomic reactions can be assessed more objectively than behavioral ones. Topographical InterpretationThe noninvasive measurement of animal emotions should be standard practice for the continual monitoring of an animal’s welfare quality. The most simple, noninvasive, and frequently used metric of an animal’s emotion is their facial expression and body language. After hypothesizing that emotional identification accuracy would be lowest for the Doberman, a breed with darker coloration, Bloom et al. [193] exposed Malinois, Doberman, and Rhodesian Ridgebacks to contexts that would elicit happiness, sadness, anger, fear, and disgust and took their photographs. When viewing the photos, participants were able to successfully identify these emotions at a rate significantly higher than chance. The lowest mean correct responses belonged to the Rhodesian Ridgeback, the only dog included with floppy ears, which may alter the appearance of threat. When viewing photos of dogs who did or did not receive artificial tears, participants assigned more positive scores to the photos with artificial tears [210]. Together with the finding that dogs’ tear volume increases both when they are reunited with their owners and when an oxytocin solution is applied to dogs’ eyes, these results suggest that emotion-elicited tears can facilitate human–dog emotional connections [210]. A qualitative behavior assessment (QBA) was developed to identify an animal’s emotional expressiveness based on the assumption that an animal’s behavior is dynamic and psychologically expressive. Adverbs are assigned to behaviors; these become a generated list of terms used for scoring observed responses. When multiple rather than singular environments are used for a QBA, the terms used to describe the animal’s emotions diversify, suggesting that QBA is sensitive to an animal’s circumstances and capable of capturing a wide repertoire of emotions [211]. This lends credence to Layng’s [176] assertion that humans know that emotional state is related to environmental circumstance, even when identifying the emotions of another species. The differences observed between shelter and owned dogs outline some of the ways in which welfare compromise is observed in shelter environments [211,212,213]. Pet dogs were labeled as more relaxed both in home and novel environments; they rest more compared to shelter dogs and exhibit fewer lip licks. Behaviors associated with increased stress (i.e., paw lifting, “displacement behaviors” such as digging/drinking, vocalization) occur more frequently in shelter compared to home environments, with dogs that remained in the shelter for more than 30 days labeled as more cautious [213]. Terms generated in QBAs have broad dimensions similar to each other, and are semantically consistent, with the exception of the types and frequency of terms used to describe sociability, fearfulness, and boredom [211]. Arena et al. [212] found that when terms generated by laypeople are modified by experts (in the fields of dog personality, behavior, welfare, and QBA methodology), and tested for interobserver reliability (labeled “mean clip score”) before being finalized for use, laypeople observers needed training to increase interobserver agreement. Expert definitions were not consistently operationalized and did not describe topography, but were “a brief depiction containing both qualitative and quantitative elements that together should illustrate the meaning of the term” (see [212], Table 4). This suggests that experts’ definitions of terms may differ from layperson understanding, and while terms that laypeople generate are consistent, they may not be accurate. This comes as no surprise, given our human propensity for anthropomorphism. For example, Horowitz [214] demonstrated that the body language and facial expressions that owners typically interpret as “guilt” are actually the dog’s response to being scolded independent of their engagement in disallowed behaviors. If training for IOA is included, QBA can function as a valid welfare assessment protocol. Inquiries about their pets’ responses land with the professionals that pet owners have most frequent contact with: veterinarians. Unfortunately, Dawson et al. [215] demonstrated that 50% of veterinarians were not able to sufficiently identify aggression in either dogs or cats, where aggression is an intentional, overt, and potentially expensive distance-increasing behavior signifying an animal is highly motivated to avoid a stimulus and is not a subtle form of communication [89]. It follows that anyone unable to identify overt aggression would also fail to identify more subtle signs of communication associated with fear, anxiety, and stress. Most veterinary colleges continue to lack veterinary behavior and applied animal welfare programs, and as such, most veterinarians lack competence in these areas and hold beliefs that do not keep pace with published data on topics including behavioral medicine and intervention [216,217]. After interviewing the staff at thirty vet clinics about their behavioral welfare practices, Dawson et al. [215] examined videos of client–staff interactions and found discrepancies between interview response and veterinary staff–patient interactions, suggesting that veterinary staff are not implementing low stress handling practices, even when they are committed to and think they are doing so. While online training programs such as Fear FreeⓇ attempt to fill the gap, not all veterinarians have exposure to this continuing education, and further, the most important component to skill acquisition, the chance to rehearse and receive feedback [218], is not provided via the online medium. In C-BARQ questionnaires, 41% of owners reported that their dog displayed fearful behavior while at the vet, with 14% of owners labeling their dog’s fear as extreme [219]. Dogs who are fearful are experiencing poor welfare. Many dogs are experiencing welfare compromise during veterinary visits, with potential for further welfare compromise when access to veterinary care is restricted due to the animal’s behavior while at the vet [220,221]. The Fear FreeⓇ movement campaigned to educate veterinarians about best practice to reduce fear in the veterinary setting, where recommendations include creating a low-stress environment by taking steps to reduce visual access to other patients in the waiting area, installing sound-absorbing tiles, rubberizing floors, providing separate areas for dogs and cats, using non-threatening body language, using a touch gradient when handling patients, paying attention to and adjusting the care plan based on an animal’s body language, distracting with high-value food, using prophylactic medications when necessary to reduce pain, anxiety, and fear, and teaching animals to accept veterinary procedures (cooperative care) [220,222]. Unfortunately, learning about techniques to reduce fear does not necessarily translate to reducing fear in practice, as time constraints, owner non-compliance, and space limitations may impede veterinary staff’s ability to carry out practices recommended for improving their client’s welfare while at the clinic [215,223]. 3.2. Cooperative Care Cooperative care refers to teaching animals new skills for medical care and husbandry by giving the learner the option to leave, practicing skills frequently, and teaching the learner to accept variety. This husbandry training was pioneered by the students of B.F. Skinner (for a review, see [224]). Skilled practitioners recommend practicing painful procedures without the painful stimuli at least 100 times per each painful experience [225]. The recommendations of Ramirez [225] describe what behavior analysts would consider best practice by using positive reinforcement, skill building, and programming for generalization and maintenance. Inspired by this work, animal trainers used similar techniques with dogs. Patel [226] demonstrates how stationing on a mat and targeting a bucket with eye contact can be used to teach dogs to “opt in” to husbandry and medical handling. Bertillson and Johnson Vegh [227] developed multiple ways to incorporate “start button” and “stop button” behaviors with companion animals, where a start button behavior serves as a discriminative stimulus (or conditional discriminative stimulus) for the handler to start a procedure, and a “stop button” behavior is a discriminative stimulus for the handler to terminate the procedure. Whether labeled “The Bucket Game”, “Start Button Behaviors”, or simply cooperative care, these techniques aim to give the learner choice and control. While cooperative care training is becoming popular in the companion animal training sphere, there is still little empirical evidence for its use, as the effects of specific interventions on companion animal behavior in a veterinary setting were examined only twice [228,229]. Stellato et al. [228] evaluated if the implementation of a desensitization and counterconditioning program would reduce pre-existing veterinary fear in companion dogs. The treatment, which was designed by veterinarians, reflected the type of advice a general practice veterinarian might give to the owner of a fearful patient; the dog owners received only written and video instructions without an opportunity to practice or receive feedback. Treatment included gradually exposing their dog to handling of different body areas, beginning with least-invasive touch (i.e., “place hand beside their paw on the ground”) and advancing to the next level of handling when the dog was “calm” and not displaying signs of fear or discomfort (i.e., “touch their paw for 2–3 s.”). Desensitization steps did not include picking up the dog to place them on a table, examining them on a raised surface, examining with a stethoscope or otoscope, or inserting a thermometer into the dog’s anus; the most invasive type of touch that owners were instructed to use included moving hands around the body in a massage-like circular motion, which is not the type of handling that occurred during the veterinary exam. After four weeks, a small effect was shown for fear reduction in the treatment group; however, owner compliance to the intervention protocols was poor (suggesting poor social validity) and the intervention did not influence temperature, heart rate, respiratory rate, trembling, vocalizations, or the amount of encouragement needed before a dog would step on the scale. Given that desensitization to veterinary handling was incomplete, nor did it meet best practice for cooperative care training, we suggest that larger effects would be demonstrated after the completion of a well-designed desensitization program. Veterinarians are likely to engender greater success for their clients by working in conjunction with or referring their clients to animal trainers and animal behavior consultants who specialize in applied work. Wess et al. [229] taught dogs without a history of severe aggression toward veterinary staff (i.e., snapping, biting) to place both front paws onto and stand on a target; this was their “cooperation signal” (i.e., “start button” behavior, or discriminative stimulus for the handler to start the exam). When the dogs assumed this position, desensitization to veterinary handling began. If the dog stayed on target, they received treats. If they stepped off their target, both the examination and treats were terminated. A certified dog trainer coached participants and assessed dogs after 9–12 weeks of training and before the second veterinary visit, with results indicating that 77% of dogs exhibited moderate to very good training progress. This training was more comprehensive than Stellato et al. [228], with desensitization procedures listed commensurate with a full veterinary examination. However, the number of desensitization steps completed was not reported, and training did not include generalization to the veterinary clinic, restraint, or handling by an unfamiliar person. During testing, owners were bystanders only allowed to give their dogs treats at specified times. Even with these limitations, a stronger reduction in the mean HR between visits was related to improved tolerance of handling, and dogs whose owners performed the restraint during the exam had lower HR compared to dogs who were restrained by an unfamiliar assistant. Dogs with more pronounced trainer-rated improvement in tolerance of handling had reduced HR during visit two, suggesting that training success was correlated with lower autonomic arousal during handling. This study provides the first data available that may demonstrate a correlation between a reinforcement schedule (extinction) and a biological measure of emotional valence (HRV). Only one owner asked the veterinarian to stop when the dog stepped off target; for 21 of the 22 dogs, their “stop button” behavior was inadvertently placed on extinction during testing and “struggling” and “attempting to jump off the table” was reported. The authors suggest that the data indicate training had different effects on the dogs based on their previous tolerance to handling and training progress; however, given that transfer of training skills was poor, conclusions cannot be drawn between treatment success and any other measures. Additionally, latent variables not examined include the number of desensitization steps successfully completed, treatment integrity, training frequency, and consistency of stopping the exam when the dog was off target. 3.3. Physiological Measures and Emotional ValenceHR and HRV are commonly used metrics that open a window to affective state by assessing autonomic nervous system (ANS) activation and could offer tools for assessing the emotional state of animals [62]. Heart rate (HR) measures the number of times the heart beats per minute, while heart rate variability (HRV) measures the vagally mediated beat-to-beat changes in heart rate. Optimal regulation occurs when autonomic arousal matches performance requirements [230]. Changes to heart rate provide information about how much an individual is having to cope with a situation, with freeze responses correlated with low heart rate and fight or flight responses correlated with higher heart rate [70]. Arousal can be measured by HR, where affective state can be measured by HRV [229]. Higher HRV is correlated with more optimal autonomic, behavioral, and emotional regulation and lower HRV is correlated with poor autonomic, behavioral, and emotional regulation and poor coping responses [229,231,232,233]. Dogs with a history of conspecific aggression, owner-directed aggression, and bite histories have significantly lower baseline HRV [231,232,233]. Low baseline HRV is additionally correlated with increased stress and anxiety [231], suggesting that this type of objective measurement might be explored as an index of welfare. After measuring changes to both HR and behaviors during and after a respondent aversive conditioning procedure, Ogata et al. [209] found that heart rate rose consistently in response to both the unconditioned and a conditioned fear-eliciting stimuli, while behavioral responding did not change reliably either between or within subjects. This suggests that measures of physiological responses could provide more accurate insight into an animal’s emotional state above observing changes in public responding.Brugarolas et al. [234] used ECG and an electronic stethoscope to monitor HR and HRV during scent detection tasks. They observed that the highest instantaneous heart rate occurred at the beginning and end of each search, which indicates an increase in arousal. While the interpretation of patterns found were outside the scope of the study, the authors demonstrated that HR and HRV can be obtained during intervention to be interpreted as a welfare metric. Polar® human heart rate monitors were validated for use in dogs [229,231,235], allowing for an accessible way for pet owners, shelter staff, trainers, or veterinarians to measure HR and HRV, with alternate tools of measurement for use in pets becoming commercially available (i.e., PetPace collars or similar purchasable pet monitoring products).Significantly increased HR in dogs is correlated to types of animal-assisted interventions (AAI) and duration of travel [236], an increase in the appeasement gestures ‘tongue flick’ and ‘paw lift’ [237], barking and growling at a stranger [231], threatening encounters [106], aversive conditions including unexpected sound blasts, unexpected short electric shocks, unexpected falling bags, opening umbrellas, restraint [201], and a conditioned stimulus ‘sound buzzer’ after being conditioned with an unconditioned aversive stimulus, a spray collar [209]. Zupan et al. [238] demonstrated that exposure to stimuli confirmed to be positive resulted in changes to nine research beagles’ HRV parameters, suggesting that higher positive emotional valence in dogs is associated with parasympathetic deactivation. Katayama et al. [239] collected data via ECG and found that HRV changed when dogs were exposed to an appetitive condition (owner petting) and an aversive condition (owner leaving the dog alone in a novel space); the type of change observed could be used to determine the dog’s emotional response. Similarly, Gácsi et al. [106] found that dogs who made distress vocalizations during separation from their owners and who growled or barked at a threatening stranger (dogs labeled “reactive”) experienced a decrease in HRV during the threatening encounter. Together, this suggests that measuring HRV during behavior interventions could determine how the animal feels about that intervention. As interpretations of HRV measurements and their implications are not yet universal (for a similar review of HRV research that reaches different conclusions, see Polgár et al., 2019 [183]), future research on baseline HRV measures and HRV changes in dogs will allow us to better understand the suitability of these measures as an applied welfare metric during intervention, especially as developing technologies for the measurement of HR and HRV become increasingly available. We did not include all available measures in this review, as many are impractical for applied use, as they cannot be easily obtained nor immediately interpreted. For example, the relationship between cortisol:creatinine ratios (C/Cr) and stress were examined, and even though a negative correlation between C/Cr ratio and lip licking was discovered, the validity of this measure was questioned [142]. Cortisol is typically measured via saliva, urine, or hair; while taking samples and waiting for results may be a useful tool for a veterinarian measuring stress over time [240] and should be considered when available, it may be cost prohibitive and fails to immediately inform the applied practitioner about the animal’s welfare during the intervention underway. See Csoltova and Mehinagic [241] for a review of all available neurobiological and psychophysiological measures for the assessment of positive dog emotion.4. DiscussionAforetime, both ethologists and radical behaviorists focused on directly observable (overt) responses [200,242]. This set a precedent that persists to this day, with few studies measuring internal events that eluded direct measurement in the past. With the advancement of technology, we are now able to combine behavioral, physiological, and biochemical measures, including endocrine and neural measures, for the consideration of a learner’s emotional affect as one measure of their welfare [3,140,241]. Ogata et al. [209], Fraser [59], and Gácsi et al. [106] note that dogs’ behavioral responses and physiological reactions to the environment are individual and may not be accurately assessed at the group level, yet single subject design is not prolific among published welfare or training research. As behavior is an epiphenomenon that does not cause or explain the occurrence of behavior [58], single subject design [243,244,245] provides methods for isolating the environmental variables controlling a learner’s response, affect, and ultimately their welfare and quality of life. A complete understanding of the biological basis for stress, [220] ethological, niche related mechanisms [246], and group research that elucidates most animals needs and preferences [58,59] provides a robust framework from which to design habitats that maximize antecedent arrangements that elicit variable species-typical behaviors in order to improve animal welfare. This knowledge is necessary for creating hypotheses about the setting events, antecedents, and consequences responsible for individual behaviors that may need intervention and which can be tested with within-subject methodology. The development of new technologies allows us to look “under the skin” and account for covert responses that can now be observed objectively. In combination with single-subject design that allows for the measurement of individual behavior change, welfare metrics must be used to assess how a learner feels about the intervention underway and to determine how applied interventions affect the learner’s quality of life. To this end, we make the following recommendations: (1)Target responses selected for intervention should meet a behavioral need, or replace a forced-choice behavior, and increase the learner’s degrees of freedom. This prompts practitioners to carefully consider if they are using coercion to achieve a goal, even when positive reinforcement is programmed as reinforcement; (2)Operational definitions of target responses should include qualitative descriptions of responses that indicate positive affective state. This highlights an additional way to gauge how a learner is experiencing a contingency in effect in order to promote a positive learning experience. For example, while a learner may ultimately express frustration by mouthing on an object, as was observed by McGowan et al. [102], being aware of and looking for a more subtle change, such as the flattening of ears [194], would allow practitioners to stop and modify the intervention before frustration, and the responses that accompany it, which are not part of the target response, escalate;(3)Practitioners should improve the chance that an intervention is socially valid to the learner by ensuring that interventions (a) stem from educated hypotheses created based on both population and individual data, taking the learner’s adaptations, perception, and cognition into account, (b) use functional reinforcers evaluated by functional analysis and/or preference assessments, rather than presumed or contrived reinforcers, (c) program for choice, allowing the learner to choose to participate, (d) provide genuine choice within the intervention by providing the learner the opportunity to perform an alternate, non-target behavior to earn the same reinforcer available for performance of the target behavior, (e) improve the learner’s relationship to, or teach additional coping skills that can be used when the learner must come into contact with an unavoidable aversive stimuli that exists, and will exist, in their environment, where appropriate, and (f) program for contact with positive reinforcement, even when positive reinforcement is not used at the outset of an intervention. These recommendations set the occasion for learner-centered interventions and increase the likelihood that welfare is positive, or at least improved, in learning contexts;(4)To ensure that the intervention is having a positive impact on welfare, quality of life metrics that can be repeated across time should be completed before, during, and after the intervention, and affective state should be measured during the intervention using validated metrics, such as body language, HR, and HRV.While this paper focused on the behaviors of, interventions for, and welfare concerns of companion animals, namely pet dogs, the outlined recommendations may be applied to any learner. Where artificial conditions are inevitable for animals in captivity, humans who are neurodivergent also find themselves in suboptimal environments that disallow for their needs. Everything from lighting to social expectations are designed for neurotypical individuals. The responses used for managing overstimulation caused by the environment are considered “inappropriate” and often socially punished (for example, hand flapping) [247,248], while responses which may be forced-choice behaviors (e.g., eye contact; [249]) are generally promoted. For these reasons, some neurodivergent individuals may be prohibited from experiencing optimal welfare in public settings [247,250,251,252]. Behavior analysts “serve the status quo” [60] by promoting neurotypical behavior [249] that may actively cause harm. A lack of social validity from the client led to public outcries against ABA services, with some autistic adults (who received behavior analytic services as children) sharing that their experiences with ABA were abusive and led to trauma [253,254,255]. While these claims warrant further investigation [256], and it may be the quality of behavior analytic services implemented and not the technology itself that caused harm, if behavior analysts are going to take appropriate steps to “support clients’ rights, maximize benefits, and do no harm” [257], include ethics-based criteria for implementing treatments, and adopt a trauma-informed care model [258], the acceptability of target behaviors and their interventions must be centered around the interests of the clients themselves. To accomplish this, practitioners must have a comprehensive understanding of the cognitive abilities, perspective, and needs of the individuals with which they work.5. ConclusionsWe believe the recommendations outlined in this paper provide a guide for increasing the ethicality of behavior analytic interventions, no matter who the learner is that is being supported. These recommendations provide multiple technologies for the objective measurement of how a learner values various responses and resources, allows for more accurate inference to a learner’s affective state, and promotes the adoption of interventions that provide choice to the learner. Taken together, scientists and practitioners gain robust metrics from which social validity can be more accurately inferred. This is an important factor that cannot be underrepresented in the provision of ethical services [150,256,259] in order to promote a good quality of life.

animals : an open access journal from mdpi

10.3390/ani13111720

PMC10251852

This study evaluated 26 sequences of terrapins worldwide through COI DNA barcoding and phylogenetic analysis, which included 12 species and three families. Moreover, 16 haplotypes were found; they were either misidentified, or a potential cryptic species was determined between B. baska and B. affinis affinis. Thus, COI remains an effective barcode marker for the terrapin species.

Technological and analytical advances to study evolutionary biology, ecology, and conservation of the Southern River Terrapin (Batagur affinis ssp.) are realised through molecular approaches, including DNA barcoding. We evaluated the use of COI DNA barcodes in Malaysia’s Southern River Terrapin population to better understand the species’ genetic divergence and other genetic characteristics. We evaluated 26 sequences, including four from field specimens of Southern River Terrapins obtained in Bota Kanan, Perak, Malaysia, and Kuala Berang, Terengganu, Malaysia, as well as 22 sequences from global terrapins previously included in the Barcode of Life Database (BOLD) Systems and GenBank. The species are divided into three families: eight Geoemydidae species (18%), three Emydidae species (6%), and one Pelomedusidae species (2%). The IUCN Red List assigned the 12 species of terrapins sampled for this study to the classifications of critically endangered (CR) for 25% of the samples and endangered (EN) for 8% of the samples. With new haplotypes from the world’s terrapins, 16 haplotypes were found. The intraspecific distance values between the COI gene sequences were calculated using the K2P model, which indicated a potential cryptic species between the Northern River Terrapin (Batagur baska) and Southern River Terrapin (Batagur affinis affinis). The Bayesian analysis of the phylogenetic tree also showed both species in the same lineage. The BLASTn search resulted in 100% of the same species of B. affinis as B. baska. The Jalview alignment visualised almost identical sequences between both species. The Southern River Terrapin (B. affinis affinis) from the west coast of Peninsular Malaysia was found to share the same haplotype (Hap_1) as the Northern River Terrapin from India. However, B. affinis edwardmolli from the east coast of Peninsular Malaysia formed Hap_16. The COI analysis found new haplotypes and showed that DNA barcodes are an excellent way to measure the diversity of a population.

1. IntroductionTerrapins inhabit either freshwater or brackish water [1]. There is no clear taxonomic group for terrapins, which may be unrelated. Numerous species belong to the families of Geoemydidae and Emydidae [2]. The only terrapin species not in this group is the Pelusios seychellensis from Seychelles [3].The “Barcode of Life” Consortium is a global effort to conduct a molecular inventory of the planet’s biodiversity [4]. After it was demonstrated that the cytochrome c oxidase subunit I (COI) gene of the mitochondrial DNA (mtDNA) could be used to successfully identify North American bird species, such as Sturnella magna, Tringa solitaria, and Hirundo rustica [5], numerous other vertebrate COI barcodes have been developed [6,7,8]. Ref. [9] also reported that the COI marker was better for barcoding than sequences from the mitochondrial control region.Traditional taxonomy frequently fails to distinguish between the different terrapin species because they lack essential morphological characteristics. Currently, molecular methods are required to identify certain species [10,11]. A complementing tool to traditional taxonomy and systematics research, DNA barcoding allows for a more accurate understanding of the existing fauna around the world [12]. Especially in species with complicated, accessible anatomy, DNA barcoding is proposed as a method for quickly and readily identifying species using a short DNA sequence [12,13]. DNA barcoding has been used to identify freshwater turtles all over the world, even in Malaysia [14].Batagur affinis ssp. [15] is among 24 species of turtles found in Peninsular Malaysia [16] and Sumatra, Indonesia, and was initially believed to be conspecific with B. baska, a species native to the North (Bangladesh and India) [17]. According to [18], B. baska consisted of at least two heritably distinct species: B. affinis ssp. populations in the Kedah River systems and B. affinis affinis populations in the Perak River systems, both on the west coast of Peninsular Malaysia. In contrast, individuals in the Terengganu River basin were identified as B. affinis edwardmolli. According to [19], this species is one of the world’s 25 most endangered freshwater turtles and tortoises.B. affinis ssp. used to live in a large river in Southeastern Asia, including the Tonle Sap in Cambodia and the Mekong delta in Vietnam. However, many of its wild populations have been severely reduced or wiped out [20,21,22,23,24]. Batagur affinis ssp. is found only on the west coast of Peninsular Malaysia and is extinct in Sumatra, Indonesia [25,26].In contrast, the subspecies B. a. edwardmolli, located on the east coast of Peninsular Malaysia that once reached from Singapore to Southeast Asia, is now thought to have vanished from Vietnam, Thailand, Singapore, and Indonesia [23,26,27]. Currently, only Peninsular Malaysia and Cambodia are home to this species [18,23,24,28]. Moreover, according to [23], there are still populations of B. a. edwardmolli in Cambodia and along the east coast of Peninsular Malaysia. This implies that the Malaysian and Cambodian populations are the only ones whose genes have remained constant across the species’ range.Unfortunately, this study was carried out during a difficult period, namely the COVID-19 pandemic. Due to the Malaysian Movement Control Order (MCO), or lockdown, we were only permitted to gather four specimens of the Southern River Terrapin from Peninsular Malaysia by the Malaysian government authority. The samples are limited due to the conservation status of B. affinis ssp., which has been listed as critically endangered on the IUCN Red List since 2000 [16]. This study compares them to the other eleven terrapin species listed by [3,26] and accessed from the public database portal.In addition, we were the first to upload COI B. affinis ssp. sequences to the GenBank database portal. The objectives of this study were to determine if terrapin DNA barcoding could be used all over the world by comparing the unique COI sequences to other COI sequences that were already available from the Barcode of Life Data (BOLD) Systems and GenBank, and to analyse the phylogenetic relationships among terrapins, including the recently collected specimens from Malaysia.2. Materials and Methods2.1. Study SitesFour Batagur affinis ssp. individuals from two distinct population locations on the east and west coasts of Peninsular Malaysia were randomly chosen for this study, and the sampling was carried out in 2020 (Figure 1). The captive hatchling population at the Bota Kanan head-starting facility (BK; GPS coordinates: 4.3489° N and 100.8802° E) in Perak, Malaysia, provided the blood samples of B. affinis affinis (N = 1). The facilities were developed beside the Perak River, which is a habitat for the wild Southern River Terrapin population. There was no uncertainty regarding the genetic origin of that sample. In addition, blood samples from three wild B. affinis edwardmolli hatchlings (translocated eggs) were taken from a population in Bukit Paloh, Kuala Berang (KB; GPS coordinates: 5.0939° N, 102.7821° E), which is in Terengganu, Malaysia. According to [29], blood was drawn from the species using venipuncture methods through the internal jugular vein and subcarapacial venous plexus (SVP). In a 2 mL microcentrifuge tube, 1.5 mL of blood was preserved with 0.5 mL of EDTA in a 1:3 ratio before being kept at −20 °C. The Department of Wildlife and National Parks, Peninsular Malaysia, issued the study and field permit approval number, which is B-00335-16-20.2.2. DNA Isolation, PCR, and SequencingFor each sample, 200 µL of EDTA whole blood was used to extract the nucleic acids. After cell lysis and protein denaturation, DNA was extracted using the ReliaPrepTM Blood gDNA Miniprep System with binding column technology (Promega, Madison, WI, USA) according to the manufacturer’s instructions. The final volume extracted was adjusted to 200 µL based on the input volume of the EDTA whole-blood sample. Using the Thermo ScientificTM NanoDrop 2000 c spectrophotometer model ND-2000, the amount and purity of the extracted DNA samples were evaluated (Thermo Fisher Scientific, Waltham, MA, USA). After quantifying the extracted nucleic acids, the DNA samples were put onto a 1% (w/v) agarose gel with molecular markers. Electrophoresis was performed to assess the integrity and intactness of the high molecular weight DNA band.The cross-species primer derived from Painted Terrapin, Batagur borneoensis, was utilised for PCR. Ref. [30] made the “Tuntong” primer pair, which targets the COI marker gene. The forward primer (5-CGCGGAATTAAGCCAACCAG-3) and the reverse primer (5-TTGGTACAGGATTGGGTCGC-3) are designed. The COI gene fragment PCR amplification was carried out in a Go Taq Flexi PCR (Promega, Madison, WI, USA) reaction mixture containing 2 µL of DNA template, 0.4 µL of primers, 4 µL of 5× PCR buffer, 1.6 µL of 25 mM MgCl2, 0.4 µL of dNTPs, 0.2 µL of Taq DNA polymerase, and 11 µL of distilled water (ddH2O). Following an initial denaturation at 94 °C for 4 min, 35 cycles of denaturation at 94 °C for 45 s, annealing at 55 °C for 35 s, and extension at 72 °C for 1 min were performed, followed by a 10 min extension at 72 °C. Finally, the purified PCR products were forwarded to a local laboratory company (First BASE Laboratories Sdn Bhd) for Sanger sequencing of the COI gene of the mitochondrial DNA (mtDNA-COI). In addition, 17 COI sequences of terrapin were extracted from GenBank and downloaded, while five COI sequences of terrapin were extracted from the BOLD Systems. This analysis led to the discovery of four novel sequences (GenBank accession numbers: OL658844–OL658847) for 26 sequences (Table 1).2.3. DNA Barcode Sequence Quality Control Measures and AnalysisChromatograms displaying the nucleotide sequences of both DNA strands for each sample were created—trimmed chromatograms with more than 2% unclear bases and low-quality noisy sequences on both ends. The bidirectional reads were eliminated by benchmarking against a quality value greater than 40. The consensus sequences were obtained by combining the forward and reverse chromatograms in SeqScape, version 2.7 (Applied Biosystems), and comparing them with reference sequences from the NCBI nucleotide (NT) database using BLASTn [38,39]. Additionally, using our COI sequences in a BLASTn search of GenBank, the species that most closely matched our sequences were noted. The sequences’ accession codes and BOLD sequence identifiers were confirmed against GenBank and the BOLD Systems (Table 1). Using the BOLD Systems’ sequence analysis [40], the Kimura 2 Parameter (K2P) model was used to calculate the pairwise sequencing divergences for the distance analyses. MEGAX was used to find the polymorphic sites (PS) or variable sites [41].2.4. Analyses of Molecular Phylogenetics and Divergence TimesThe best-fitting evolutionary model for each sequence analysed was determined using the Akaike information criterion (AIC) with sample size correction implemented in jModelTest2 on XSEDE (2.1.6) [42]. The phylogenetic studies used models of sequence evolution selected as best with jModelTest2 for coding and non-coding sequences. maximum likelihood (ML) analyses [43] were performed. As a result, the alignments were carried out in MEGAX using ClustalW [41]. All sequences produced multiple alignments with the same length and beginning point. However, Jalview, Ref. [44], was used to accomplish various sequence alignments, functional site analyses, and web postings of alignments between B. affinis affinis and B. baska [45]. IQ-tree was used for phylogenetic reconstruction by [46] on XSEDE and [47] via the online CIPRES Science Gateway V.3.3 [48]. The trees were visualised in FigTree v1.4.4 [49].On the other hand, using the BEAST v2.6.6 tool, the phylogenetic tree topology and divergence dates were computed concurrently [50,51]. BEAUti 2 [52] was used to unlink the substitution models of the data partitions and implement the sequence evolution models selected with jModelTest2 as optimal. The “Clock Model” was set to a rigorous clock with uncorrelated rates, while the “Tree Model” was assigned to a Yule speciation process. The sequences were examined using a relaxed molecular clock model, which permits substitution rates to vary among branches based on an uncorrelated lognormal distribution [50]. We established the species tree before the Yule process. Two simultaneous assessments were conducted utilising Bayesian Markov Chain Monte Carlo (MCMC) simulations with a sampling frequency of 5000 for 100,000,000 generations. The nucleotide substitution model for ML was empirically set to TN93. Bootstrap analysis (1000 pseudoreplicates) provided branch support, and all other parameters were left at their default settings.After that, the phylogenetic trees were plotted using FigTree v1.4.4. To create the phylogenetic trees, the whole mitochondrial COI sequences of Batagur affinis (MTD042-21) and the out-group species Ophiophagus hannah (MH153655) were chosen from the GenBank online database [33,53]. Then, using the software DnaSP 6.12.03, we analysed the haplotype of each specimen [54,55,56]. A Median Joining (MJ) network analysis by [57] was performed with NETWORK 10.2.3. Results3.1. Taxonomic Range and Red List CoverageTable 1 contains the details on the taxa used in this study. The final data collection includes 12 species from the Testudines order, two previously unrepresented in the barcode database. One is not available in the BOLD Systems, and five were not sent to GenBank. We initially deposited our novel COI gene of the mitochondrial DNA (mtDNA-COI) samples (Batagur affinis ssp.) in the GenBank database portal.As a result, the IUCN Red List assigned the 12 species of terrapins sampled for this study to the classifications of least concern (LC) for 33% of the samples, critically endangered (CR) for 25% of the samples, vulnerable (VU) for 8% of the samples, and endangered (EN) and near-threatened (NT) for 17% of the samples (Figure 2).3.2. COI Divergence AssessmentAll 26 produced barcodes had sequence lengths of more than 503 bp with no indels or stop codons found. The nucleotide composition was as follows: 16.88% Guanine, 27.21% Cytosine, 27.5% Adenine, and 38.41% Tyrosine. GC Codon position 1 was 52.62% followed by GC Codon position 2 (43.21%) and GC Codon position 3 (36.46%). Almost all species (83.33%, ten species) were represented by dual specimens with a single specimen representing another species and five specimens representing another species (Table S1).The genetic divergences of the COI sequences within the order Testudines were studied at various taxonomic levels (Table 2). The genetic divergence rose with the taxonomic rank as expected. The hierarchical taxonomic relationship was directly associated with increased K2P genetic divergence. The conspecific K2P levels ranged from 0% to 2.14% with a mean of 0.68% (SE = 0.04). The mean K2P divergence amongst the congeneric species specimens was 5.49% (SE = 0.15; range 0–9.14%). The average K2P divergence between the specimens from various genera in the same family was 17.10% (SE = 0.03; range: 4.98–22.48%). This range, though they overlap, indicates intraspecific (S) and intragenus (G) distances (Figure S1).Deep intraspecific K2P divergences were identified in a Batagur baska (2.14%) that exceeded the conventional threshold distance of 2% [12,58] (Table 3). A barcode gap analysis revealed that practically all species represented by multiple sequences had a barcode gap (Figure 3). Notably, just one species, Batagur baska, had its maximum intraspecific and nearest neighbour distances (0%).3.3. Population RelationshipsThe nucleotide diversity at 199 nucleotide positions and transitions is approximately 55% saturated (Table S2). When all codon locations are analysed, transitions and transversions are displayed against the pairwise sequence divergence Tajima-Nei Method (TN84) for the terrapins utilising 503bp of the COI DNA barcode (Figure 4). DAMBE [59] uses these substitution models to perform various molecular phylogenetic analyses. DAMBE also includes functions for determining the optimum substitution models for particular sequences.The network had 16 haplotypes (Figure 5), which were confirmed with DNAsp 6.12.03 analysis (Table 1). Different haplotypes were found in Malaclemys terrapin, Emys orbicularis, Melanochelys trijuga, Trachemys scripta elegans, and Batagur affinis ssp. Furthermore, Batagur baska and Batagur affinis affinis shared a single haplotype (Hap_1), which was shown to be the most variable haplotype. The remaining haplotype only had two specimens and one species.4. DiscussionThis study examined 26 terrapin COI sequences from the order Testudines. The species are divided into three families: eight Geoemydidae species (18%), three Emydidae species (6%), and one Pelomedusidae species (2%) (Figure S2 and Table S3). Based on the IUCN Red List of the 12 species of terrapins, 25% were critically endangered (CR) and 8% were endangered (EN). The terrapins studied all inhabit fresh or brackish water [26]. Furthermore, “terrapin” refers to more or less aquatic, hard-shelled turtles [60]. Notably, refs. [3,26] identified 13 terrapin species worldwide but ignored a previously thought-to-be-extinct Seychelles black terrapin species (Pelusios seychellensis). However, a genetic analysis of the lectotype revealed that this terrapin is not extinct and is now known as Pelusios castaneus. Before the Zoological Museum Hamburg bought a private collection of specimens in 1901 [26,61], the specimens could have been mislabelled or mixed up.Therefore, the discovery of species-specific COI sequences allows for the identification of terrapin species using DNA barcodes to supplement taxonomy. This can also be used in the field when identifying lost nests or those caught as bycatch in fishing nets. When no other material is available, terrapin eggs or meat are used in the forensic investigations [4].Additionally, DNA barcoding holds excellent promise for species identification and other conservational genetic applications in terrapins, which are distinct in the evolutionary tree of terrapins for inhabiting the river realm and are well-known for their lengthy migrations. One of the main objectives of the DNA barcoding initiative, species identification, was accomplished using their COI sequences. Even though these ancient taxa have undergone relatively slow molecular evolution [62,63], diagnostic sites at the COI gene were found for all 12 species of terrapins. Ref. [9] found that the distance-based analysis of COI sequences always put members of the same species together, even though the phenetic methods required a total baseline sample for a correct assignment. Using distinct nucleotide combinations, unique COI barcodes were generated for each of the 12 previously defined terrapin species (Table S2). The diagnoses were reliable with species-specific haplotypes [9] (Table 1; Figure 5).If a phenetic technique based on a BLAST search was used without a comprehensive baseline sample, such as the one available in GenBank prior to this work, query sequences could be assigned to the wrong species. There were no Batagur affinis ssp. COI sequences in GenBank, for example, and a query on a Southern River Terrapin (B. affinis affinis) grouped it with a Northern River Terrapin (B. baska). The BLASTn search validated it, showing 100% similarity between the B. affinis-MTD042-21 COI sequences and B. baska-HQ329671 COI sequences (Table 1). So, Jalview’s alignment and visualisation (Figure 6) showed that the sequences of B. baska (GenBank Accession Number: HQ329671) and B. affinis (BOLD ID: MTD042-21) were very similar. Similarly, Emys orbicularis, a species with COI sequences in GenBank, may be confused with Emys trinacris or a cryptic species due to 98% identical COI BLASTn results (Table 1).Furthermore, in the BOLD Systems, the identical sequence of the Northern River Terrapin has two different BIN numbers (AAW2850 and ADX0374), which could be misinterpreted as Southern River Terrapin or a cryptic species.The detection of the so-called “barcode gap,” which can be measured by comparing the highest intraspecific distance with the minimum interspecific distance (also known as the nearest neighbour genetic distance), is one of the premises of DNA barcoding [64]. Moreover, DNA barcodes are helpful in the investigation of cryptic species [65], particularly those that appear similar but differ genetically [66]. A morphological species gap is strong evidence for species-level cryptic diversity [67]. On the other hand, the absence of a gap between two morphological species implies that they are different forms within the same species, or that they share ancestral polymorphism and/or hybridisation followed by introgression. In this case, it would be helpful to use a multigene (i.e., genomic) method to figure out the reciprocal taxonomic status of the two morphological species [68].Table 3 shows that the DNA barcoding method revealed possible hidden variety within a species while failing to discover a meaningful difference between two biological species (B. baska and B. affinis). Such findings demand additional taxonomic research. In comparison to the mean congeneric divergence (5.49%), the mean conspecific K2P divergence (0.68%) was eight times smaller. Thus, as predicted, there was less genetic diversity between the conspecific individuals than between the congeneric species. It makes sense that there would be a rise in the taxonomic levels and an increase in the genetic divergence [69]. Therefore, both mean genetic estimations are comparable to those that have already been noted. In most fish molecular analyses, the conspecific divergence was found to be 0.25–0.39%, while the congeneric divergence was found to be 4.56–9.93% [70,71,72,73,74].4.1. Population RelationshipsThis research began by examining the terrapins’ DNA barcodes and mitochondrial COI gene haplotypes worldwide. Some existing terrapins and sea turtles are reported to carry mitochondrial COI gene haplotypes [4,9,26,30]. Nonetheless, our study contributes significantly by discovering new sequences from previously unknown areas in Malaysia and around the world. Previous research employing the COI gene in DNA barcoding of terrapins and sea turtles identified 1–10 haplotypes [4,9,30]. This study revealed 16 haplotypes (Table 1; Figure 5) of terrapins from around the world. The BOLD Systems differ from those previously described in Bota Kanan, Perak, and Kuala Berang, Terengganu. Also, the novel B. affinis ssp. COI gene sequences from Malaysia were submitted to GenBank (Table 1). They may serve as a reference for future genetic research of populations. A more comprehensive analysis involving additional sites and samples will be necessary to find common haplotypes. Previous studies by [28,75] described the divergence of Batagur baska and Batagur affinis ssp. Our research checks the sequences between the Indian and Malaysian populations. Moreover, the sequences from the Malaysian specimens are novel, and we hypothesise that this population is exclusive to this region (Figure 1).Thus, clustering analyses and haplotype networks indicate that the three families are separated by four significant unique lineages (Figure 7). Figure 5 demonstrates that Hap_1 and Hap_16 are more closely related than other haplotypes. Hap_1 contains two B. baska specimens and two B. affinis affinis specimens, while Hap_16 contains three B. affinis edwardmolli specimens, which are in line with [14] that only found a haplotype in the Kuala Berang, Terengganu population; it has been proven that this is a random sampling, and we are not focusing on a clutch. In this case, it appears to be a cryptic species between B. baska and B. affinis affinis. We would need a more extensive set of genes and many markers from the nuclear genome [66,76,77] to decide if these groups should be called species or subspecies. Perhaps revision is required following the separation of B. baska and B. affinis ssp. by [28,75]. Even though it can be challenging to identify the morphological diagnostic features in morphologically cryptic species [78,79], the usefulness of such diagnoses may be in doubt [80]. We now recognise that cryptic species are relatively abundant [81,82] and widespread across most animal phyla [83,84]. Moreover, recent DNA research discovered cryptic species in many aquatic taxa [85], raising the possibility that aquatic biodiversity is higher and speciation possibilities have occurred more frequently than previously thought [86].In addition, using Bayesian analysis, the maximum likelihood phylogeny of the investigated dataset revealed coherent, monophyletic clustering of all studied species (Figure 7). On the phylogenetic tree, cohesion was also detected between the database reference sequences for the representative species and the created sequences. The species were classified according to their family with Geoemydidae being the most abundant. The evolutionary tree indicates that B. baska originated in India and is closely related to B. affinis affinis from Malaysia, which is supported as a potential cryptic species. Melanochelys trijuga is similar to the Persian Gulf’s Mauremys caspica, but the Malaclemys terrapin in North America is identical to Trachemys scripta elegans.4.2. Conservation StatusThe International Union maintains the Red List for biodiversity for the Conservation of Nature (IUCN). The IUCN is essential for guiding and igniting conservation and policy change activities; it is much more than a list of species and their states. The preservation of the natural resources that humans depend on is essential [87,88]. The IUCN Red List Categories and Criteria are designed to offer a clear framework for locating species in danger of going extinct globally. According to [87], species can be “Not Evaluated,” “Data Deficient,” “Least Concern,” “Near Threatened,” “Vulnerable,” “Endangered,” “Critically Endangered,” “Extinct in the Wild,” or “Extinct”.Nearly every nation with native species has its own conservation effort (Table 4). Three Batagur species of terrapin, B. affinis, B. baska, and B. borneoensis, are listed as having Critically Endangered (CR) status in Table 1. Moreover, B. affinis ssp. falls under the Extinct in the Wild (EW) category in Southeast Asian nations, including Indonesia, Singapore, Thailand, and Vietnam [23,25]. B. affinis ssp. is currently restricted to Malaysia and Cambodia. Ref. [89] also states that B. baska may be threatened in Thailand and Myanmar. Additionally, B. borneoensis was discovered in Brunei, Malaysia, and Indonesia, although it was virtually extinct in Thailand [90].5. ConclusionsIn conclusion, COI remains an effective barcode marker for terrapin species, contributing vital information that can be utilised to distinguish and identify genera and species. Compatibility with traditional taxonomy could provide a solid and dependable instrument for accurate species identification and biodiversity assessment facilitation. However, more markers and specimens from new sites should be added to the collection to more accurately compare terrapin populations. The detailed results provided fresh insights into the taxonomic classification of terrapins and revealed the existence of potential cryptic species. This investigation found compelling evidence of potential cryptic species between B. baska and B. affinis affinis. Our research shows that B. affinis affinis might be the same species as B. baska, but B. affinis edwardmolli might be its own species. However, further research is required. Therefore, the genomic and bioinformatics analysis of terrapins described here could serve as a reference for future global studies of this species and permit a more rational attempt to conserve terrapins. The proposed conservation units are based on the fact that phylogeny and phylogeography change over time and space.

animals : an open access journal from mdpi

10.3390/ani11092537

PMC8465824

Lecithin can not only provide energy to animals but also serves as an emulsifier and has the potential to enhance the utilization of dietary fat by animals. Thus, there is a need to elucidate the underlying mechanism of the positive effect in broilers. The present feeding trial aims to evaluate the effect of lecithin on broilers’ performance, meat quality, lipid metabolism, and cecum microbiota. The obtained results revealed significant improvements in broiler meat quality resulting from the lipid metabolism and microbiota that were affected by lecithin treatment. Consequently, it could be used in broilers’ diets for the aim of meat quality improvement.

The present study was conducted to evaluate the effects of lecithin on the performance, meat quality, lipid metabolism, and cecum microbiota of broilers. One hundred and ninety-two one-day-old AA broilers with similar body weights (38 ± 1.0 g) were randomly assigned to two groups with six replicates of sixteen birds each and were supplemented with 0 and 1 g/kg of lecithin for forty-two days. Performance and clinical observations were measured and recorded throughout the study. Relative organ weight, meat quality, lipid-related biochemical parameters and enzyme activities were also measured. Compared with broilers in the control group, broilers in the lecithin treatment group showed a significant increase in L* value and tenderness (p < 0.05). Meanwhile, the abdominal adipose index of broilers was markedly decreased in lecithin treatment after 42 days (p < 0.05). In the lipid metabolism, broilers in the lecithin treatment group showed a significant increase in hepatic lipase and general esterase values at 21 days compared with the control group (p < 0.05). Lower Firmicutes and higher Bacteroidetes levels in phylum levels were observed in the lecithin treatment group after 21 and 42 days. The distribution of lactobacillus, clostridia, and rikenella in genus levels were higher in the lecithin treatment group after 21 and 42 days. No statistically significant changes were observed in performance, relative organ weight, or other serum parameters (p > 0.05). These results indicate that supplementation with lecithin significantly influence the lipid metabolism in broilers at 21 and 42 days, which resulted in the positive effect on the meat color, tenderness, and abdominal adipose in broilers.

1. IntroductionPhospholipids, which are found in thousands of organisms, are key components of the cell membrane. Dietary intake of exogenous phospholipids provides the majority of phospholipids organisms need. Lecithin, which is mostly obtained from soybean, oilseed rape, and sunflower seed in a normal diet, is widely known to be an important transporter of lipids in organisms [1]. Commercially available products are mostly extracted from soybeans, and they are widely used in healthcare for humans and animals [2]. Lecithin can not only provide energy to animals but also serves as an emulsifier and has the potential to enhance utilization of dietary fat by animals [3]. The physiological effects of lecithin have been studied extensively in recent years, although public health recommendations regarding lecithin intake currently have no limit. Even so, lecithin has become a popular animal dietary supplement for increasing performance and nutrient utilization. The effect of lecithin on cholesterol reduction was validated in monkeys, hamsters, and many other species [4]. It was reported in a previous study that a diet supplemented with lecithin could increase the daily gain of nutrients and nutrient digestibility in animals [5]. Moreover, lecithin treatment in intestinal cell membranes alters the permeability of cell bilayers and resulted in the greater influx of micro- and macro-molecules across the cell membrane [6]. The positive effect of lecithin is also attributed to healthy gut improvement [7]. Therefore, supplementing exogenous lecithin emulsifiers in the diets has become very popular in poultry production.To date, limited studies have been well conducted with exogenous lecithin or emulsifiers, and inconsistent responses in broilers have been noted. Only a few studies on broilers reported that lecithin can maintain broiler performance with low energy diets [8]. Several studies reported that supplementation with lecithin to the diet of broilers improved growth performance [9]. Meanwhile, many studies of lecithin treatment in broilers show the positive effect on the improvement of apparent energy and nutrient utilization [10]. Contrarily, it has been reported that emulsifiers have no significant impact on the growth performance of broilers [11,12]. The different effect of lecithin or emulsifiers on broilers are attributed to the ingredients and test conditions. However, these studies have barely noticed the effect of lecithin on the gut microbial community of broilers. As the pivotal component of intestinal barrier, the composition and function of the gut microbiota is dynamic and affected by diet properties. Meanwhile, the gut microbiota has shown effect on lipid metabolism and lipid levels in blood and tissues [13]. Hence, we hypothesized that lecithin supplementation might have unknown effects on microbial communities in broilers, which might further affect the lipid digestibility and utilization in broilers. Therefore, the present study was conducted to evaluate the effects of lecithin on performance, meat quality, lipid metabolism, and the microbial community of broilers.2. Materials and Methods2.1. Birds, Diets, and ManagementA total of one hundred and ninety-two one-day male Arbor Acres (AA) broilers were randomly assigned to two groups (control and treatment) with six replicates of sixteen birds each. The birds of each replicate were reared in a single cage (2.4 × 0.6 × 0.6 m) with a wire screen floor. Water and feed were provided ad libitum, with the photoperiod set at 23 L:1 D throughout the study. The temperature in the broiler house during the first week was 32 to 35 °C, after which it was lowered by 1 °C every other day until it reached 27 °C. The study was conducted according to the Regulations of the Experimental Animal Administration issued by the State Committee of Science and Technology of the People’s Republic of China. The animal use protocol was approved by the Animal Care and Use Committee of the Poultry Institute at the Chinese Academy of Agriculture Science.Lecithin was derived from soybeans (PHOSPHOLIPON 90 g, Lipoid Co. Ltd., Ludwigshafen, Germany) with 94.3% purity, 1.3% nonpolar lipids, 1.2% lysophosphatidylcholine, 0.2% water, 0.17% tocopherol, and other sterols. The diet of the birds was formulated to meet or slightly exceed all nutrient requirements (Table 1) (NRC, 1994) [14], and it was provided in mash form to avoid degradation of lecithin. All of the birds were fed diets supplemented with 0 (control) and 1 g/kg (treatment) lecithin for 42 days.2.2. Growth Performance and Sample CollectionCage-side observations, which included recording changes in clinical condition or behavior, were made at least twice daily throughout the study. All macroscopic abnormalities in the birds or deaths throughout the whole experiment were recorded after necropsy. The body weights of the birds from different replicate were determined at the beginning, 21 days, and 42 days into the study. Feed consumption was recorded on a replicate basis at 21 and 42 days. Feed conversion was expressed as the grams of feed consumed per grams of weight gain. The average daily gain (ADG), average daily intake (ADI) and FCR were calculated at 1 to 21 days of age, 22 to 42 days of age, and 1 to 42 days of age.At days 21 and 42, 1 bird was randomly selected from each replicate (6 birds from each group) and weighed after 8 h of feed deprivation. Before the necropsy, 1.5 mL wing blood were centrifuged at 3500 g for 10 min so that serum could be collected for clinical blood chemistry and enzyme activity detection. After that, the birds were fed for 4 h and sacrificed by jugular bleeding, while the organs (liver, heart, spleen, and thymus) were removed and weighed. The abdominal adipose tissue of broilers were removed and weighed at 42 days. Breast muscle (both the pectoralis major and minor included) was collected from the right pectorals and stored at 4 °C for later analysis. The contents from both ceca were thoroughly mixed and stored at −80 °C for 16S rDNA amplicon sequencing analysis.2.3. Meat QualityThe meat samples were stored in 4 °C for 24 h before the detection. Indices of meat quality including pH, color, shear force, and drip loss were determined with the methods described previously [15,16].The pH value was measured with a portable pH meter (HI8424, Beijing Hanna Instruments Science & Technology Co. Ltd., Beijing, China) equipped with an insertion glass electrode calibrated in buffers at pH 4.01 and 7.00 at ambient temperatures. The measurements were made at the same location on individual breast and thigh muscle samples. The average pH value was calculated from 3 readings taken on the same muscle sample.Meat color was assessed with a chroma meter (CR-10, Minolta Co. Ltd., Suitashi, Osaka, Japan) to measure CIE LAB values (L* means relative lightness, a* means relative redness, and b* means relative yellowness). The tip of the colorimeter measuring head was placed flat against the surface of the muscle. The meat color was expressed using the CIELAB dimensions of lightness (L), redness (a), and yellowness (b). The higher L* values were lighter, higher a* values were more red, and higher b* values were more yellow. Drip loss was estimated by determining expressible juice using a modification of the filter paper press method. A raw meat sample weighing 1.0 g was placed between 18 pieces of 11 cm diameter filter paper and pressed at 35 kg for 5 min at 25 °C. The expressed fluid was determined as the change in the weight of the original sample. The water-holding capacity was calculated as the ratio of expressible fluid/total moisture content.A shear force test was done on the breast fillets using the razor blade method with an Instron Universal Mechanical Machine (Instron model 4411, Instron Crop., Canton, MA, USA). Meat samples were stored at 4 °C for 24 h and were then individually cooked in a water bath at 80 °C in plastic bags to an internal temperature of 70 °C. The samples were then removed and chilled to room temperature. Strips (1.0 cm (width) × 0.5 cm (thickness) × 2.5 cm (length)) parallel to the muscle fibers were prepared from the medial portion of the muscle and sheared vertically. Shear force was expressed in kilograms. Three values were recorded for each replicate sample and averaged.2.4. Lipid MetabolismBiochemical parameter detection was performed in serum immediately. Glucose (GLU), cholesterol (CHO), triglyceride (TG), low-density lipoprotein cholesterol (LDL), and high-density lipoprotein cholesterol (HDL) were measured using an Unicel DXC 800 (Beckman Coulter, Fullerton, CA, USA).Lipoprotein lipase (LPL) and hepatic lipase (HL) in serum were measured with colorimetric enzymatic methods using commercially available kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, Jiangsu, China). General esterase (GE) activity was calculated as the sum of LPL and HL.HL detection in liver was performed in liver supernatant. Approximately 0.1 g of liver sample was transferred into a 1.5 mL precooled centrifuge tube with 0.9 mL physiologic saline and homogenized into 10% homogenates (50 Hz for 3 min through tissue grinder, SCIENZT-48, Scientz Biotechnology Co., Ltd., Ningbo, Zhejiang Province, China). It was then centrifuged through a low-speed centrifuge (2500–3000 rpm/min for 10 min, DL-5M, Xiangyi Power Testing Instrument Co. Ltd., Changsha, Hunan, China). The test kit was the same as serum.2.5. DNA Extraction, PCR Amplification of 16S rDNA, Amplicon Sequence, and Sequence Data ProcessingMicrobial genomic DNA was extracted from 220 mg of cecal contents sample using a QIAamp DNA Stool Mini Kit (Tiangen Biotech Company Limited, Beijing, China) following the manufacturer’s instructions. Successful DNA isolation was confirmed by an A260/280 ratio ranging between 1.8 and 2.0 and by agarose gel electrophoresis.Based on previous comparisons, the V4 hypervariable regions of 16S rDNA were PCR amplified from microbial genomic DNA harvested from samples and were used for the remainder of the study. PCR primers flanking the V4 hypervariable region of bacterial 16S rDNA were designed. The barcoded fusion forward primer was 520F 5′-barcode + GCACCTAAYTGGGYDTAAAGNG-3′, and the reverse primer was 802R 5′-TACNVGGGTATCTAATCC-3′. The PCR conditions were as follows: one pre-denaturation cycle at 98 °C for 30 s, 25 cycles of denaturation at 98 °C for 15 s, annealing at 50 °C for 30 s, and elongation at 72 °C for 30 s, and one post-elongation cycle at 72 °C for 5 min. The PCR amplicon products were separated on 0.8% agarose gels and extracted from the gels. Only PCR products without primer dimers and contaminant bands were collected for sequencing by synthesis (Axygen Axy Prep DNA Gel Extraction kit, New York, NY, USA). Barcoded V4 amplicons were sequenced using the paired-end method by Illumina MiSeq (Sangon Biotech Company Limited, Shanghai, China) with a 600-cycle index read. Only sequences with an overlap longer than 10 bp and without any mismatch were assembled according to their overlap sequence. Reads that could not be assembled were discarded. Barcode and sequencing primers were trimmed from the assembled sequence [17,18].2.6. Statistical AnalysisIn this study, operational taxonomic unit (OTU) cluster analysis was used to classify the OTU sequences based on a 97% similarity criterion. The OTU abundance of each sample was generated at the genus level. The bacterial diversity is shown by the number of OTUs. The mean length of all effective bacterial sequences without primers was 280 bp. The abundance and diversity of microbiota were compared between each sample by calculating OTUs.A pen of birds was the experimental unit for performance parameters. For all other measurements we used individual birds from each replicate. All the data are presented as the means ± SEM. Statistical analyses were carried out with SPSS 18.0 for windows (SPSS Inc., Chicago, IL, USA). Differences between groups were tested with a t-test for independent samples. A p value less than 0.05 was considered to indicate statistical significance, and a trend was considered present at p < 0.10.3. Results3.1. PerformanceThe mean performance (ADG, ADI, and FCR) and mortality are shown in Table 2. No statistically significant differences were found among all performance parameters or mortality observed in either the treated group or the control group throughout the whole period (p > 0.05). The relative organ weight results are shown in Table 3. As shown in Table 3, no statistically significant changes were observed in relative organ weight of broilers in 21 or 42 days (p > 0.05). The abdominal adipose tissue of the control group at 42 days was significantly higher compared with the lecithin treatment groups (p < 0.05).3.2. Meat QualityThe meat quality parameters results can be seen in Table 4. The L values were markedly higher in the lecithin treatment groups compared with the control group (p < 0.05). Drip loss and b value were higher in the lecithin treatment group than in the control groups, while the p values were 0.062 and 0.078 separately. the tenderness value (shear force) was markedly lower in the lecithin treatment groups than in the control group (p < 0.05) at 42 days. No other parameters were significantly affected by dietary lecithin supplementation (p > 0.05) at 21 or 42 days.3.3. Lipid MetabolismThe results of the lipid-related biochemical parameters in serum are shown in Table 5, while the results of lipid-related enzyme activity in serum are shown in Table 6. As shown in Table 5, the GLU value at 42 days was significantly higher in lecithin-supplemented groups than in the control groups (p < 0.05), which confirmed that exogenous high lipid intake would increase the glucose metabolism in vivo. Meanwhile, the cholesterol level in serum of broilers at 21 days significantly decreased with lecithin supplementation, but no difference at 42 days.3.4. 16S rDNA Analysis of Bacterial CommunitiesShifts of cecal microbial community along with body development: The results shown in Figure 1 and Figure 2 describe the distribution of DNA sequences into phyla after 21 and 42 days. For the majority of phylum in intestinal content, Firmicutes was the most dominant phylum for all development stages. However, the abundance of Firmicutes (88.10%) in the control group was significantly higher than that in the lecithin treatment group (78.83%) after 21 days. Conversely, the Bacteroidetes level in the control group (5.34%) was significantly lower than that in the lecithin treatment groups (11.99%) after 21 days. Proteobacteria were observed at a higher level in lecithin treatment groups (11.99%) after 21 days. The results after 42 days were the same as after 21 days. The abundance of Firmicutes (82.26%) in the control group was significantly higher than that in the lecithin treatment group (70.10%) after 42 days. Conversely, the Bacteroidetes level in the control group (12.39%) was significantly lower than that in the lecithin treatment groups (26.19%) after 42 days.Principal components analysis (PCA) plots of samples from after 21 (A) and 42 (B) days: PCA results (Figure 3) illustrated the differences in distribution of microbial community. Based on the PCA analysis, the control group and the lecithin-supplemented group were significantly divided into two clusters. Combined with previous data, this result suggested that the lecithin treatment group had a significantly different gut micro-flora profile.Genus level significance analysis of the cecal microbial community after 21 and 42 days: Table 7 and Table 8 represent the abundance of selected genera (>0.1% in at least one sample) across all samples. It clearly showed that there were apparent differences in genus distribution between control and lecithin treatment groups. The proportions of lactobacillus, lactobacillus agilis, clostridia, and rikinellaceae were higher in the lecithin treatment group, whereas the proportions of prausnitzii, erysipelotrichi, lachnospiraceae, and alactolyticus were higher in the control group after 21 days, which is the same as 42 days.4. Discussion4.1. Growth Performance and Organ WeightThe result of this trial revealed that direct addition of high purity lecithin into the diet of broilers showed no significant effects on the performance or relative organ weight throughout the whole experiment period, which is the same as most previous studies. The energy requirements of commercial broilers is very high, while the fat source is limited in traditional diet. Exogenous supplementation of lipid becomes an inevitable trend in broiler production for better performance. However, the effect of lipid or emulsifier additives remain inconsistent. Through exogenous supplementation of phospholipids, lecithin had different effects on performance improvement in several reports. Previous reports showed that supplementation of lecithin in low energy and protein diets improved both performance and digestibility parameters in broilers [19], while some reported the effect was not significant [11]. Meanwhile, a previous study showed that the body weight of groups supplemented with lecithin at 21 days and 35 days were not significantly different compared to the control group [20]. Additionally, even more, lecithin supplementation in animal diet showed suppression effect in gastric emptying and resulted in vomiting and diarrhea in some reports [21]. Meanwhile, lecithin was used as exogenous emulsifiers to improve the utilization of fat and energy in weaning piglets [22] or broilers [19,23]. Fats, such as hydrophobic components, must aggregate to form micelles to be absorbed. Emulsifiers found in the digestive tract (mainly bile salts) naturally mediate this process and improve the formation of micelles [24]. However, most of the studies reported that no significant influences were observed on the N or energy digestibility. Due to the short digestive tract and unstable digestibility of the broilers, the effect of lecithin in increasing the fat digestibility seems not obvious [25]. That may explain the non-significant effect of lecithin on growth performance or organ weight in broilers.Overall, these results may be affected with the source of dietary lipids, the formation and addition amount of phospholipid product, the breed of the chickens and the duration of the trial. Improvement effect of lecithin on performance or organ weight may be more consistent and obvious in the low energy and protein diets in broilers.4.2. Meat QualityIn this study, dietary supplementation with lecithin had a positive effect on meat color, water-holding capacity, and tenderness at different periods, which may be due to the effect of lecithin supplementation in lipid metabolism. Zhao et al. [10] reported that lecithin supplementation could increase the transportation of lipids in the body and improve the deposition of fat in the muscle, which is consistent with our results. Meat characteristics, including tenderness, water-holding capacity, pH, and meat color, are important indices for evaluation of meat quality. Meat color attribute for the final evaluation and acceptance of a meat product in the consumption [26]. In this study, the color value of breast meat was increased by lecithin supplementation. Similar results were reported and indicated that the color value of meat also showed a trend in improvement with the increase of emulsifier dosage [25,27]. As the most important textural characteristic of meat, tenderness has great impact on the consumer experience of broiler meat. Shear force value in this trail was significantly increased by lecithin treatment, which is also related to the fat content in the meat [28]. The pH value showed no difference in this trail, revealing that lecithin treatment showed no effect on lactic content.4.3. Lipid MetabolismThe cholesterol level in serum of broilers after 21 days significantly increased with the lecithin supplementation. This trial provides evidence that serum cholesterol level of broilers remains constant at 42 days. Similar results were detected by Zhao et al. [10]. The reason may be the immature development of the digestive tract in chicks. Previous studies reported that lecithin supplementation could increase the duodenum development [29], and the effect is more obvious in the early stage. The results of those studies suggested that the effect of lecithin on the serum profile of broilers may be more efficient in the starter period [9]. Previous research on cholesterol metabolism using isotope tracer techniques indicated that the net balance of cholesterol homeostasis was relatively stable and resistant to plant sterol supplementation in chickens [30].Exogenous emulsifier can accelerate the emulsification of lipids in the small intestine and promote the activation of lipase. In addition, lecithin was reported to promote the secretion of endogenous bile acid, and further improve the utilization rate of fat [25]. This trial showed that the enzyme activity of HL and GE was significantly increased by supplementation with lecithin at 21 days. No other parameters were significantly affected by dietary lecithin supplementation at 21 or 42 days. Previous research reported that there is a positive correlation between the enzyme activities of lipase and the deposition of lipids across both in broilers and layer strains [31]. As the main content of lipoprotein, lecithin plays an important role in the lipid metabolism such as lipid transportation. Exogenous phospholipids can elevate the HDL level and decrease the cholesterol level in serum, resulted in the regulation of lipid deposition [19,23]. The lipid metabolism results of this study indicated that the lipid transportation and deposition was altered by lecithin supplementation. Exogenous lecithin accelerated the lipodieresis in liver and reduced the transport of lipids in serum, which resulted in fewer lipid depositions in the abdomen. Boontiam et al. [32] reported that lecithin supplementation increase the proportion of lipids that used in muscle formation. That may neatly explain the results of meat quality in this experiment. As the visual indicators of lipid deposition in broilers, tenderness, and abdominal adipose were affected in this trial through the addition of lecithin, which validates the prediction that lipid metabolism is affected by the lecithin addition.4.4. Microbial CommunityThe composition of intestinal microbiota is important for maintaining homeostasis of the gastrointestinal tract and the health of the host [33]. The intestinal microbial community has been recognized as a strong determining factor of host physiology, especially through its critical role in the digestion of feed in a host [34]. The 16S rDNA gene sequencing in cecal content shown that addition of lecithin altered the microbiota composition at the phylum level in broilers. PCA plots confirmed the results at the phylum level. In our study, Bacteroidetes, Firmicutes, Tenericutes, and Proteobacteria were the main bacteria found in the broiler intestinal flora, which was consistent with previous research [35]. More specifically, lower Firmicutes and higher Bacteroidetes levels were observed in the lecithin treatment groups, whereas higher levels of Bacteroidetes and Proteobacteria were observed in the control group. Studies comparing the gut microbiota between obese and lean animals showed that lower Firmicutes and higher Bacteroidetes levels were associated with the lean phenotype [36]. It can also be concluded from many studies that lower Firmicutes and higher Bacteroidetes levels lead to fewer lipid deposits in animals. Lactobacillus were observed as the bacterial genus which can cure enteritis, while the clostridia were reported as synergistic action with lactobacillus [37]. The higher distribution of lactobacillus and Clostridia in lecithin treatment group showed that the distribution of bacterial in the genus level may be helpful in the lipid absorption of birds. Therefore, the altered profiles of the cecal microbiota after 21 and 42 days was involved in the process of lecithin effect on broiler lipid metabolism.5. ConclusionsIn summary, by taking advantage of 16S rDNA sequencing, this study revealed that lower Firmicutes and higher Bacteroidetes in levels are noted in broiler ceca, responding to lecithin treatment. Lecithin supplementation improved the enzyme activity of HL and GE in serum, also reducing the abdominal adipose tissue in broilers. Finally, this study provides evidence that lecithin supplementation had a positive effect on meat color and tenderness in broilers.

animals : an open access journal from mdpi

10.3390/ani11123586

PMC8697937

The purpose of this study is to investigate the impact of COVID-19 on the revenue of the livestock industry and to study the challenges COVID-19 brings to the livestock industry. From the perspective of financial statements, we use the revenue function of listed Chinese livestock companies from 2015 to 2020 to quantitatively estimate the impact of the pandemic. The study results show that the COVID-19 pandemic has reduced the revenue of the livestock industry and has different implications on livestock enterprises of various sizes. Based on the above study results, we provide further suggestions for measures that governments and livestock companies can consider to reduce the impact of the pandemic, thereby increasing the level of sales of livestock products.

The COVID-19 pandemic has affected social order and people’s health and has also caused a heavy blow to the livestock industry, affecting animal management and welfare. The livestock industry is one of the main contributors to economic growth in many regions, and it is of great significance to people’s lives and regional economic growth. COVID-19 has reduced the livestock industry’s market as well as consumers’ opportunities to purchase livestock products, resulting in no sales or low sales of livestock or their products. The main purpose of this study is to consider the impact of the pandemic on the revenue of the livestock industry, and to study the challenges arising from the pandemic to the livestock industry. Based on the perspective of financial statements, we estimate the impact of COVID-19 through the translog revenue function of listed Chinese livestock companies from 2015 to 2020, and the study results show that the COVID-19 pandemic has reduced the revenue of the livestock industry, but the decline in revenue of large livestock enterprises is lower than that of small and medium-sized livestock enterprises. In the last two parts of this study, we make policy recommendations to livestock enterprises and the authorities.

1. IntroductionHow does the livestock industry affect the environment and society? On the one hand, the impact of animal husbandry on climate change is very significant [1,2]. The greenhouse gases produced by livestock account for more than 14.5% of the total global greenhouse gas emissions, exceeding all vehicles, trucks, trains, airplanes, and ships on the planet. The sum of emissions. In the past 40 years, more than 40% of the forests in Central America have been cut down, and 720,000 square kilometers of rainforest in South America have been cut down, all of which provide land for animal husbandry production and growing fodder (mainly soybeans). In addition, animal husbandry occupies one-third of the earth’s freshwater resources, 45% of the earth’s land resources, and one-third of the land is desertified because of animal husbandry. On the other hand, the livestock industry affects people’s diet and health [3], is the foundation of human well-being [4], and plays a vital role in the food supply, culture, and economy in rural and urban areas [5].Before the COVID-19 pandemic, African swine fever had a significant impact on China’s livestock industry, and more precisely, it had a substantial impact on China’s pig farming. Although African swine fever has brought panic, it has also brought about changes in China’s pig farming. Affected by African swine fever, China’s large-scale pig farming companies have expanded rapidly, and the concentration of the pig farming industry has increased. According to a report from the Ministry of Agriculture and Rural Affairs of China, most African swine fever outbreaks in China before 2020 occurred in small- and medium-sized pig farms. Due to small- and medium-sized pig farms’ low epidemic prevention and risk-bearing capabilities, they are increasingly withdrawing from the market under the pressure of the epidemic and government policy. However, large-scale pig farming companies can increase the rapid detection of African swine fever from aspects of pig farm management and equipment utilization and improve safety guarantees. In the market and government policies, large pig farming companies have used their capital to acquire or build new pig farms to accelerate their expansion. The industry concentration of China’s pig farming industry has rapidly increased. In June 2019, the Ministry of Agriculture and Rural Affairs of China and the Ministry of Finance of China jointly issued the “Notice on Doing a Good Job in Subsidy for Working Capital Loans on Pig Breeding Farms and Large-scale Pig Farms.” This policy provides support for short-term loan discounts to large pig farms. The Chinese government vigorously supports the development of large-scale pig-farming enterprises in terms of policies. In addition, to reduce the risk of African swine fever, downstream slaughter companies also prioritize purchasing slaughter pigs from large pig farms, and large-scale pig farms have become the industry’s mainstream. In addition, African swine fever has also accelerated the integration of the upstream and downstream of China’s pig farming industry chain.In China, the growth in livestock before the pandemic was mainly the result of increased consumption of pork, beef, and chicken [6,7,8]. From 2000 to 2018, China’s pork production increased from 39.66 million tons to 54.04 million tons, and the proportion of pork in total meat production fell from 65.90% to 62.70%. In 2019, due to the impact of the African swine fever epidemic, the scale of pig farming in China dropped sharply, and the proportion of pork in 2019 dropped sharply to 55.6%. Since the beginning of the new century, China’s total output of cattle, sheep, and poultry meat has increased. From 2000 to 2019, China’s poultry meat production increased from 12.71 million tons to 22.39 million tons, and its proportion rose from 21.1% to 29.3%. During the same period, China’s beef production rose from 5.13 million tons to 6.87 million tons, and its proportion increased from 8.5% to 8.7%. China’s mutton production rose from 2.64 million tons to 4.88 million tons, and its proportion increased from 4.4% to 6.4%. The COVID-19 outbreak has had a significant impact on many sectors at the global, regional, and national levels, including the livestock sector [9,10,11]. Furthermore, the meat, poultry, and animal product processing plants have been the sectors most affected by the COVID-19 pandemic [12,13,14].Since the discovery of African swine fever in China in 2018, panic has spread throughout the country, which has caused a large number of pig deaths, a shortage of pork products, and a sharp increase in the prices of pork and other meat products [15]. In the COVID-19 pandemic, many countries have adopted measures such as lockdowns and traffic restrictions. The Chinese government has also taken a series of practical actions. For example, to prevent the spread of the virus, farmers’ markets and feed mills were closed [16], slaughterhouses were delayed to resume work, and transportation was restricted [17]. These factors led to a shortage of feed supplies, unsalable livestock products, and increased production and transportation costs [18]. During the pandemic, China implemented strict traffic controls to control the movement of people and vehicles [17], which had a severe impact on the supply of livestock production materials and the sale of products [15]. It is difficult for livestock enterprises and farmers to deliver livestock products to the market [19], and it even leads to the interruption of the entire livestock industry supply chain [15].Many government relief policies are not enough to make up for the loss of livestock product sales caused by the COVID-19 pandemic [20]. Understanding the specific impact of the pandemic on livestock product sales and the revenue of livestock enterprises, and how to minimize the impact has become a top priority. At present, the quantitative research on the impact of the COVID-19 pandemic on the sales of livestock products and the revenue of livestock enterprises is relatively limited. Based on the perspective of financial statements, we use the revenue function of listed Chinese livestock companies from 2015 to 2020 to quantitatively estimate the impact of the pandemic. The research results show that the COVID-19 pandemic has reduced the revenue of the livestock industry, but the decline in revenue of large livestock enterprises is lower than that of small and medium-sized livestock enterprises. Based on the above research results, we provide further suggestions for measures that governments and livestock companies can consider to reduce the impact of the pandemic, thereby increasing the level of sales of livestock products. The remaining part is as follows: Section 2 provides background and hypothesis development. In Section 3, we introduce the research methods and models. Section 4 and Section 5 are the results and conclusions, respectively.2. Background and Hypothesis DevelopmentWhen an infectious disease breaks out, the hungry population will increase [21,22], and workers will decrease accordingly. Since the virus can spread from person to person through the air [23], and workers in meat and poultry processing plants need to work closely with each other, this increases the risk of virus transmission [11,23,24,25]. COVID-19 has caused meat and poultry processing plant employees to become sick, putting livestock companies at risk of closing down, which reduces the supply of livestock products [20]. As processing plants are blocked [26,27,28,29,30,31], many poultry producers have had to discard animals because they cannot supply them to meat processing plants [31]. In addition, restrictions on social distancing and transportation have caused sales difficulties, and many dairy farmers have had to dump milk [26]. The closure of restaurants, hotels, and schools has also affected the sales of livestock products [32]. Due to sales difficulties and the high cost of raising animals, this has led to the unfortunate problem of euthanasia of animals [26,33,34].Due to the COVID-19 pandemic, the labor force of the livestock industry shrank [23,35,36], and many processing plants closed [27,28,29,30,31] or reduced processing capacity [23,30]; livestock companies faced difficulties in slaughtering and processing livestock [20]. The increase in consumption of livestock products has always been a driving force for the development of the livestock industry. However, the COVID-19 outbreak has caused significant losses to all economic sectors including the livestock industry [37], and livestock companies and farmers have been losing their regular customers [38,39,40,41]. The blockade of hotels, schools, and restaurants (the main consumers of animal products) and the tourism slump have led to a sharp decline in market demand for animal products, which further makes it difficult for livestock companies to sell their products [37].After the outbreak of COVID-19, the sales of animal products on online sales platforms in China showed a significant increase [9], but the sales in the physical market have dropped significantly. In addition, there were rumors that livestock can spread COVID-19 and people should stop buying and eating animal products [42,43,44]. These misunderstandings that led to the belief that livestock or animal products are the hosts or carriers of the virus have left people with the impression that humans may be infected with COVID-19 through the consumption of animal products, which has further exacerbated the decline in sales of meat and other animal products [9]. The interruption of international trade routes also limits the development of livestock enterprises [9,45,46], which may eventually affect the sales of meat products and dairy products export suppliers, and further reduce the income of livestock companies and farmers [9].The closure of live animal markets in many countries has caused livestock producers to face difficulties in selling animal products, which has greatly reduced the opportunities for livestock products to enter the market for consumers to purchase. In addition, due to traffic restrictions and the fear of being infected by the virus, people have drastically reduced the frequency of going to supermarkets and markets to purchase animal products from on-site stores, which has further led to a decline in sales of animal products. However, even if people reduce the overall frequency of going to the market, in a limited time, Chinese people still tend to buy livestock products produced by large livestock companies. Therefore, we propose the following hypothesis:Hypothesis 1 (H1). The COVID-19 pandemic has reduced the revenue of the livestock industry, but the decline in revenue of large livestock enterprises is lower than that of small- and medium-sized livestock enterprises. 3. Method3.1. Theoretical ModelHigh-quality employees are one of the prerequisites for the stable operation of enterprises [47,48,49,50,51,52]. We use the following equation to represent the livestock industry production function (The definitions of x1,x2,x3,f, d and b are shown in Table 1):(1)y=f(x1,x2,x3,f, d, b)In Equation (1), y is the sales, the revenue function of livestock enterprises is as follows:(2)r(p;x1,x2,x3,f, d, b)=max pysubject to y=f(x1,x2,x3,f, d, b) r is revenue, p represents the prices of livestock products. The revenue function of livestock enterprises is converted into the following equation:(3)lnr=α0+δlnp+∑i=13αilnxi+β1lnf+δ1ln d+ϵ1ln bThis study normalized by setting p = 1 [53]. we can further simplify Equation (3) as follows:(4)lnr=α0+∑i=13αi ln xi+β1ln f+δ1ln d+ϵ1ln bScholars have studied the use of translog revenue function in many industries [54]. The translog revenue function in this study is as follows:(5)ln r=α0+∑i=13αi ln xi+β1ln f+δ1ln d+ϵ1ln b+12∑i=13∑i=13αil ln xi ln xl+12β11(lnf)2+12δ11(lnd)2+12ϵ11(lnb)2+∑i=13γi1lnxiln f+∑i=13εi1lnxiln d+∑i=13μi1lnxiln b+θ11ln f ln d+ρ11ln f ln b+σ11ln d ln b+ρ11ln f ln b+σ11ln d ln b3.2. Data and Variables3.2.1. Data Source and Sample PeriodWe obtained the data for this study from the quarterly financial statements data provided by the CSMAR database, including the livestock industry data from 2015 to 2020. We have excluded unreasonable observations, such as the number of employees or zero net fixed assets. Finally, the livestock industry obtained 255 valid observations.3.2.2. Variable DefinitionsCOVID-19 (COVID) and the top two livestock companies (BIG) are the dummy variables of this study’s livestock industry revenue model. This study uses the revenue to define large livestock companies. From 2015 to 2020, the revenue of China’s top two large livestock enterprises accounted for a relatively high proportion of the industry’s revenue. This study takes the top two livestock companies (BIG) as one of the dummy variables of the model. Under the pandemic, this study can use BIG to explore the impact of COVID on large and non-large livestock companies. We bring together the definitions of the above variables and other variables in Table 1.3.3. Estimation ModelThe livestock industry estimation model of this study is as follows:(6)lnREVENUE=α0+α1lnMSTAFF+α2lnRSTAFF+α3lnOSTAFF+β1lnFIXED+δ1lnDEVELOP+ϵ1lnBIOLOGY+12 α11 (ln MSTAFF)2+12 α22 (ln RSTAFF)2+12 α33 (ln OSTAFF)2+12 β11 (ln FIXED)2+12 δ11 (ln DEVELOP)2+12 ϵ11 (ln BIOLOGY)2+α12lnMSTAFFlnRSTAFF+α13lnMSTAFFlnOSTAFF+α23 ln RSTAFF ln OSTAFF+γ11lnMSTAFFlnFIXED+γ21 ln RSTAFF ln FIXED+γ31 ln OSTAFF ln FIXED+ε11lnMSTAFFlnDEVELOP+ε21ln RSTAFF ln DEVELOP+ε31 ln OSTAFF ln DEVELOP+μ11lnMSTAFFlnBIOLOGY+μ21lnRSTAFFlnBIOLOGY+μ31 ln OSTAFFlnBIOLOGY+θ11 ln FIXED ln DEVELOP+ρ11 ln FIXED ln BIOLOGY+σ11 ln DEVELOP ln BIOLOGY+φ1 BIG+φ2 COVID+φ3 BIG COVIDThe following is the APE derivation process of variables:The APE of MSTAFF on REVENUE:(7)∂ ln REVENUE^/∂ ln MSTAFF=α^1+α^11 ln MSTAFF¯+α^12 ln RSTAFF¯+α^13 ln OSTAFF¯+γ^11 ln FIXED¯+ε^11 ln DEVELOP¯+μ^11 ln BIOLOGY¯ The APE of RSTAFF on REVENUE:(8)∂ ln REVENUE^/∂ ln RSTAFF=α^2+α^22 ln RSTAFF¯+α^12 ln MSTAFF¯+α^23 lnOSTAFF¯+α^21 ln FIXED¯+ε^21 ln DEVELOP¯+μ^21 ln BIOLOGY¯The APE of OSTAFF on REVENUE:(9)∂ ln REVENUE^/∂ ln OSTAFF=α^3+α^33 ln OSTAFF¯+α^13 ln MSTAFF¯+α^23 ln RSTAFF¯+γ^31 ln FIXED¯+ε^31 ln DEVELOP¯+μ^31 ln BIOLOGY¯The APE of FIXED on REVENUE:(10)∂ ln REVENUE^/∂ ln FIXED=β^1+β^11 ln FIXED¯+γ^11 ln MSTAFF¯+γ^21 ln RSTAFF¯+γ^31 ln OSTAFF¯+θ^11ln DEVELOP¯+ρ^11 ln BIOLOGY¯The APE of DEVELOP on REVENUE:(11)∂ ln REVENUE^/∂ ln DEVELOP=δ^1+δ^11 ln DEVELOP¯+ε^11 ln MSTAFF¯+ε^21 ln RSTAFF¯+ε^31 ln OSTAFF¯+θ^11 ln FIXED¯+σ^11 ln BIOLOGY¯The APE of BIOLOGY on REVENUE:(12)∂ ln REVENUE^/∂ ln BIOLOGY=ϵ^1+ϵ^11 ln BIOLOGY¯+μ^11 ln MSTAFF¯+μ^21 ln RSTAFF¯+μ^31 ln OSTAFF¯+ρ^11 ln FIXED¯+σ^11 ln DEVELOP¯The APE of BIG on REVENUE:(13)∂ ln REVENUE^/∂ BIG=φ1+φ3 COVIDThe APE of COVID on REVENUE:(14)∂ ln REVENUE^/∂ COVID=φ2+φ3 BIG4. Results4.1. Descriptive Statistics and Correlation MatrixTable 2 shows the descriptive statistics. It can be seen that during the study period, the medians of revenue, R&D personnel, ordinary personnel, net fixed assets, R&D investment, and net productive biological assets are all less than the average. From the perspective of the average revenue of the livestock industry every year, the revenue of livestock products in 2015–2019 shows an upward trend year by year, showing the increase in consumer demand for livestock products. However, under the impact of the 2020 pandemic, the consumption of livestock products has declined, and the average revenue of Chinese livestock companies has also declined significantly. In 2020, livestock companies continue to expand investment in fixed assets, productive biological assets, and human resources. This shows that China’s livestock industry generally believes that the decline in livestock products sales caused by the COVID-19 pandemic is temporary.Table 3 shows the Spearman and Pearson correlation coefficients (On the right is Pearson, and on the left is Spearman). The values in Table 3 indicate a positive correlation between COVID and REVENUE, but the correlation is drawn without controlling for other variables. After we control other variables, we can analyze the impact of COVID on REVENUE through APE, and we will have a more accurate conclusion. In addition, COVID is significantly positively correlated with OSTAFF.4.2. The Pandemic and the Revenue of Livestock IndustryThis study analyzes the impact of COVID-19 on the revenue of the livestock industry. Taking the time of COVID-19 occurrence in 2020 as the boundary, the Chow Test is used to explore the relationship between the pandemic and the revenue of the livestock industry, the F statistics were 2.16. The rejection of the null hypothesis indicates that the COVID-19 would impact the revenue of the livestock industry. Meanwhile, this study puts BIG as a dummy variable into the model.4.3. Estimation Results4.3.1. The Revenue FunctionTable 4 shows the estimated results of Chinese livestock enterprises’ revenue functions. This study uses the more flexible translog format instead of using the simple Cobb-Douglas format (Log-linear). We needed to check whether the translog format can provide a correct representation of Chinese livestock enterprises’ revenue function. Therefore, this study tests whether these conditions in Equation (6) is met:(15)αil=β11=δ11=ϵ11=γi1=εi1=μi1=θ11=ρ11=σ11=0 for all i=1,2,3.Table 4 shows the result of the F statistical value is 3.71, which significantly rejects the null hypothesis of the log-linear specification, indicating that the translog format is suitable for analyzing Chinese livestock enterprises’ revenue functions.4.3.2. Livestock Product SalesTable 5 shows the average partial effect (APE). The APE of COVID to the REVENUE of Chinese livestock enterprises is negative and significant. Meanwhile, Table 4 shows that the estimated value of the COVID coefficient is significantly negative in the revenue function equation of Chinese livestock enterprises. These figures show that the COVID-19 pandemic reduces livestock product sales.We analyzed the possible reason is that the inconvenience of transportation caused by the pandemic makes people spend less time shopping to buy livestock products. Another reason that cannot be ignored is the decrease in income brought about by the pandemic, which has caused consumers to reduce their consumption of livestock products and instead save their income.The APE of COVID on BIG and non-BIG are −0.254 and −0.387, respectively. These results indicate that under the pandemic, the decline in revenue of large livestock enterprises is lower than that of small- and medium-sized livestock enterprises. Hypothesis 1 of this study has been verified by the above contents. In China, large-scale livestock companies have further improved their Internet marketing channels during the pandemic, and the livestock products produced by them are better packaged for preservation. Therefore, in the predicament of the pandemic, although the overall consumption of livestock products has dropped significantly, the consumption of livestock products produced by large enterprises has decreased by a smaller amount. This result may provide inspiration for managers of small and medium livestock companies. It is possible that small- and medium-sized livestock companies can significantly reduce the reduction in livestock product sales caused by the pandemic by establishing a comprehensive door-to-door service and online sales system similar to large-scale livestock companies.5. ConclusionsDue to the decline in China’s livestock production during the pandemic, China has increased its imports of foreign livestock products. According to data released by the General Administration of Customs of China, in 2020, China’s total meat imports totaled 9.91 million tons, an increase of 60.4% year on year. Among them, pork imports were 4,392,200 tons, a year-on-year increase of 108.34%; chicken imports were 1,433,000 tons, a year-on-year increase of 98.28%; beef imports were 2,118,300 tons, a year-on-year increase of 27.65%. According to Euromonitor data, the market size of China’s imported meat products in 2020 is 15 billion US dollars, and Brazil is China’s largest source of meat imports, accounting for up to 21%. Brazil’s meat exports to China have proliferated, from USD 800 million in 2015 to USD 3.1 billion in 2020, with an average annual growth rate of 31%. During the pandemic, the consumption demand for meat products in Brazil’s domestic market decreased, and prices were far below the international market level. At the same time, during the pandemic, the output of Chinese livestock products has dropped significantly. The Chinese market can form a benign interaction with the export of Brazilian livestock companies.Due to human mobility restrictions and rising unemployment, the COVID-19 crisis has impacted livestock production, negatively impacting many livestock companies and farmers. We need to take measures to promote the livestock industry’s development better. The following options are available for policymakers and competent authorities to consider: (1) in terms of transportation, providing animal product transport drivers with special transportation permits, and promoting the easier import and export of animal products under the premise of quarantine safety; (2) under the premise of ensuring epidemic prevention requirements, allowing more markets and supermarkets (especially fresh markets and fresh supermarkets) to resume opening; (3) improving the level of aseptic packaging and refrigerated transportation of animal products; (4) formulating policies that require livestock workers to abide by certain epidemic prevention codes of conduct [55]; (5) providing children with high-quality livestock products at low prices can increase the revenue of livestock enterprises and farmers and further protect the health and nutritional level of children under the pandemic; (6) Support livestock enterprises through special subsidies from dedicated financial institutions; (7) helping producers improve online sales channels to promote the sales of animal products and the growth of enterprise revenue.This study is limited to analysis from the perspective of the financial statements of listed companies in China’s livestock industry. Future study may consider conducting cross-regional comparative studies on the livestock industry in different countries, which may obtain the different impacts of COVID-19 on the livestock industry in different economies.

animals : an open access journal from mdpi

10.3390/ani11040939

PMC8066009

As biosecurity is generally low in backyard chicken flocks, infections with various pathogens are common. This puts other poultry nearby, including commercial flocks, at risk. Some chicken pathogens can also infect humans and cause disease. In this study, backyard poultry flocks were tested for parasites. Eighty-four fecal samples, 82 from chickens and two from turkeys, from 64 backyard flocks throughout the state of Alabama were collected in the summers of 2017 and 2018. The most frequently observed parasites were coccidia, unicellular parasites capable of causing diarrhea. Eggs of various roundworms were observed in 20.3–26.6% of the flocks. These parasites were usually present in low numbers only. Other detected parasites were the flagellates Histomonas meleagridis and Tetratrichomonas gallinarum in 4.7% and 18.8% of flocks. Both can cause severe disease in poultry. Detected parasites that can cause disease in humans were Cryptosporidium spp. in 18.8% of the flocks and Blastocystis spp. in 87.5% of the flocks. The results will help to provide information that can be used to design outreach programs to improve the health and wellbeing of birds in backyard flocks.

Keeping chickens as backyard pets has become increasingly popular in the United States in recent years. However, biosecurity is generally low in backyard flocks. As a consequence, they can serve as reservoirs for various pathogens that pose a risk for commercial poultry or human health. Eighty-four fecal samples, 82 from chickens and two from turkeys, from 64 backyard flocks throughout the state of Alabama were collected in the summers of 2017 and 2018. Coccidia oocysts were seen in 64.1% of flocks with oocyst counts in most samples below 10,000 oocysts per gram. Eggs of Ascaridia spp. or Heterakis gallinarum were observed in 20.3% of the flocks, and eggs of Capillaria spp. in 26.6% of the flocks. Egg counts were low, rarely exceeding 1000 eggs per gram. DNA extracted directly from fecal samples was investigated by PCR for other relevant parasites. The results showed that 4.7% of flocks were positive for Histomonas meleagridis, 18.8% of flocks for Tetratrichomonas gallinarum, 18.8% of flocks for Cryptosporidium spp. and 87.5% of flocks for Blastocystis spp. The results will help to provide information that can be used to design outreach programs to improve health and wellbeing of birds in backyard flocks.

1. IntroductionOver the past two decades, there has been an apparent increase in backyard flocks in the United States [1,2,3,4,5,6]. “Backyard flock” is a term that generally refers to a privately owned flock of poultry, more often chickens than turkeys, that are kept at a residence. The most common reasons for backyard flock ownership in the United States are to keep the chickens as pets, a learning tool for children or as a source of eggs [3]. In other parts of the world, small non-commercial chicken flocks are referred to as village chickens and contribute to the subsistence of their owners [7].Regardless of the location, many of these small flock owners tend to lack knowledge of proper biosecurity measures, e.g., wearing designated clothes/shoes, not allowing guests to interact with the chickens. They are not aware of the risks associated with exposing their flock to wild birds and rodents [3,8,9]. Zoonotic avian diseases such as salmonellosis are a risk for small flock owners, either by direct contact with backyard poultry flocks or by consumption of contaminated meat or eggs [10,11]. Low biosecurity in backyard flocks may also be an issue for commercial poultry flocks as backyard flocks can become a reservoir for pathogens [12]. This is especially relevant in a state like Alabama, which ranks second in broiler production in the United States [13].Eimeria spp. are considered ubiquitous in chicken flocks [14]. However, their prevalence in village chickens can be between less than 5% [15,16] and up to more than 60% [17]. Roundworms, mostly Ascaridia galli and Heterakis gallinarum, have been detected in 15–25% of chickens by coproscopy [15,18] and up to 80% by visual inspection of intestines [19]. In 25% of dead village chickens, helminths were regarded as causative factor for the loss [20]. Currently, limited information is available about parasites found in backyard flocks in the United States. In birds submitted from backyard flocks to eight veterinary diagnostic laboratories across the United States, internal parasites were regarded as the primary cause of mortality in 2.6% of the birds. However, parasitic infections were the most common secondary finding, being observed in 25.5% of the birds [1].The aim of the present study was to determine the population of relevant parasitic organisms found in backyard poultry flocks without ongoing disease.2. Materials and Methods2.1. Sample CollectionEighty-four fecal samples from 64 different, non-commercial backyard flocks with less than 50 chickens throughout the state of Alabama were included in the study. The flocks were selected opportunistically, and pooled fecal samples of 10 to 50 g were collected in Ziploc bags and submitted by the owners. Forty-seven samples from 41 flocks were submitted in the summer of 2017 and 37 samples from 23 flocks in the summer of 2018. Two of the fecal samples were from turkeys kept on the same premises with sampled chickens. Each sample was stored at 4 °C upon arrival for microscopy and at −20 °C for DNA extraction. Four owners submitted samples of their flocks in both 2017 and 2018; however, in the present study they are considered different flocks.2.2. Oocysts and Nematode Egg DetectionEach fecal sample was mixed thoroughly, and 1 g was suspended in 29 mL saturated NaCl solution. Debris was filtered out through a sieve. A McMaster chamber was filled with the fecal mixture and placed on a microscope where Eimeria spp. oocysts and nematode eggs were counted. The total number of oocysts and eggs in the chambers were multiplied by 100 to obtain the oocysts per gram (opg) and eggs per gram (epg) [21,22]. Eimeria oocysts were between 10 and 30 µm long and between 10 and 20 µm wide with a thick, double layered, smooth oocyst wall. Some oocysts were sporulated while most were not. Ascaridia spp. and H. gallinarum eggs were between 75 and 80 µm long and between 45 and 50 µm wide with a thick, smooth shell. Ascaridia spp. and H. gallinarum eggs were not differentiated due to similar egg morphology [23,24]. Capillaria eggs were about 70 µm long and 30 µm wide, with a thick smooth shell and two polar plugs.2.3. Oocyst Purification and qPCR to Detect EimeriaOocysts were purified and concentrated from 4 g feces of 47 samples with Eimeria oocysts as described by Hafeez et al. [25]. Three positive samples were not further processed due to lack of material. DNA was extracted from the purified oocysts using the QIAGEN QiaAmp DNA mini kit (QIAGEN, Valencia, CA, USA) according to the manufacturer’s protocol, and Eimeria DNA was quantified by qPCR with 45 cycles detecting Eimeria 5S rDNA as described [26,27]. DNA load was expressed as the number of the cycles of the qPCR minus the quantification cycle (Cq). Its correlation with the parasite load in opg was assessed by calculating Spearman’s rho using R 3.6.0 [28].2.4. Stool DNA Extraction and PCR for Other ParasitesDNA was extracted from one fecal sample per flock using the QIAGEN QIAamp Stool Mini Kit according to the manufacturer’s instructions (QIAGEN, Valencia, CA, USA). Histomonas meleagridis, Tetratrichomonas gallinarum, Blastocystis spp., and Cryptosporidium spp. and were detected by PCR using established protocols. Primers and references are listed in Table 1. Positive and negative controls were included in all PCR runs, and a negative control was included in all DNA extractions.3. Results3.1. Eimeria and Nematode PrevalenceEimeria were detected in 41 flocks (64.1%) and 50 samples (59.5%). Median parasite load was 800 opg; however, several samples had greater than 10,000 opg. In the two samples from turkeys, no coccidia were observed. Ascaridia spp. or H. gallinarum eggs were detected in 13 flocks (20.3%) and 16 samples (19.0%), while Capillaria spp. were present in 17 flocks (26.6%) and 22 samples (26.2%). Median epg for all nematodes was less than 500 (Table 2, Figure 1). In one of the two samples from turkeys, 200 epg Ascaridia spp. or H. gallinarum eggs were observed. Of the four flocks that submitted samples in both 2017 and 2018, one flock had a change in status for coccidia from negative to positive and two had changes in Ascaridia/Heterakis egg status from negative to positive. There were no changes in status for Capillaria spp. eggs from year 2017 and 2018.3.2. Quantification of Eimeria Oocysts by qPCRThe 5S rDNA qPCR failed to detect Eimeria DNA in four samples with 100 opg (two samples), 4000 opg, and 30,800 opg. A spearman’s rho of 0.31 showed only a weak correlation between the parasite load seen in the feces and the DNA load detected by qPCR (Figure 2).3.3. Prevalence of Other ParasitesOf the 64 DNA samples, one from each flock, tested by PCR, 4.7% were positive for H. meleagridis, 18.8% for T. gallinarum, 18.8% for Cryptosporidium spp., and 87.5% for Blastocystis spp. (Table 2). Of the four flocks whose owners submitted samples in both 2017 and 2018, only one flock had a change in status for both T. gallinarum and Blastocystis spp. with the flock being positive in 2017 but negative in 2018. There were no changes in any of the other species of parasites in the flocks that submitted samples in both years.4. DiscussionBackyard flocks may be a concern to public health and the commercial industry as they could potentially be a reservoir for pathogens. This is due to the fact that many of these flocks have poor biosecurity and have frequent access to the outdoors, which allows them to come in contact with wild birds and other animals, such as rodents, that can transmit disease [6,32].In the present study, Eimeria spp. oocysts were detected in 59.5% of the samples and 64.1% of the flocks and counts in most samples were low. This reflected the equilibrium between infection and immunity present in older chickens, as well as the lower stocking density in extensively kept, often free ranging backyard flocks [3], which decreases the infection pressure. Compared to the prevalence of Eimeria spp. reported in village chickens, which ranges from less than 5% [15,16] up to more than 60% [17], this was comparatively high. One of the most important factors influencing the prevalence of Eimeria spp. in small flocks is the season [16,33,34]. In India, prevalence varied between 61% during monsoon season, i.e., warm and humid conditions, and 22% during the preceding cooler months [35]. The prevalence in the samples of this study taken during summer in Alabama was similar to the former number.Four samples in which coccidia oocysts were observed, tested negative by qPCR. One likely reason for the discrepancy is a lack of sensitivity of the qPCR: two of the samples in question contained only 100 opg. On the other hand, two samples with considerably higher oocyst counts tested negative as well. The most likely reason is that the observed oocysts were not Eimeria infecting chickens but other coccidia, including Eimeria spp. infecting other hosts; the primers of this qPCR were designed based on sequences of Eimeria infecting chickens [26] and might not amplify other Eimeria spp. or coccidia. In fact, Eimeria from other hosts such as squirrels and mice were detected in some of the samples when amplified by pan-Eimeria PCR primers [27], potentially the result of coprophagy by chickens or contamination of the samples with feces from other hosts.There was only a weak correlation between the parasite load seen in the feces and the DNA load detected by qPCR. The reasons probably include the presence of Eimeria from other hosts in addition to varying losses of oocysts during the purification of the oocysts and age of the samples. Testing samples from commercial poultry by the same methods showed a better correlation and no sample in which Eimeria oocysts had been seen tested negative by qPCR (results not shown).A prevalence of 20% for eggs of A. galli or H. gallinarum and 26% for eggs of Capillaria spp. in the present study were similar to the prevalence of these parasites in village chickens in Africa when fecal samples were investigated [15,18]. However, the prevalence was lower than in organic layer chickens in Europe, where flock prevalence of the two parasites was between 49.3% and 100% [24]. However, the mean of 576 epg was similar to the results presented here [24].Since coccidia and roundworms were the most encountered parasites in similar studies, they were our primary target and consequently flotation was used for detection. However, flotation might not be the most suitable method for tapeworm eggs, and we might have missed infections with those.In European commercial pullet and layer flocks, antibodies against H. meleagridis were detected in up to 37.3% of the tested birds and 89.3% of the tested flocks [36,37]. In contrast, in the present study, the prevalence as detected by PCR was extremely low. This compares to findings by Cadmus et al. [1] who diagnosed histomoniasis based on lesions only in very few chickens. No nematode eggs were detected in the samples that tested positive for H. meleagridis. Unfortunately, these samples were very dry, which decreased the likelihood to detect nematode eggs.In commercial poultry in Germany, T. gallinarum DNA was detected in 17.7% of flocks in which lesions resembling histomoniasis were observed, which is similar to the flock prevalence found here [38]. To our knowledge, there are no previous studies investigating its prevalence in commercial or backyard poultry without reported disease. In the present study, a single flock had a concurrent infection with both T. gallinarum and H. meleagridis.Zoonotic parasites that were investigated included Cryptosproidium spp. and Blastocystis spp. There are several species of Cryptosporidium that are known zoonotic agents. Cryptosporidium spp. can cause intestinal disease in humans [39]. Cryptosporidium meleagridis, an avian pathogen, has been shown to be increasingly important as a human pathogen as it makes up 10–20% of human cryptosporidiosis cases in Peru and Thailand [40]. Due to the low host specificity of C. meleagridis and other Cryptosporidium spp., it is important for backyard flock owners to be aware of this and improve biosecurity as they could potentially become ill.Blastocystis spp. are very common in chickens and seem to have a low host specificity. Blastocystis infections in humans may result in clinical symptoms such as diarrhea, abdominal cramps, and nausea. However, it is unclear if Blastocystis spp. infecting chickens can cause disease in humans [41,42]. In this study, there was a high prevalence of Blastocystis spp. with 87.5% of backyard flocks being infected. This is unsurprising as another study found a Blastocystis prevalence of 95% in commercial chickens [43]. Overall, only one flock was free of the parasites investigated and 11 flocks were only infected with a single parasite. Fifty-two flocks had concurrent infection with two or more of the selected parasites. However, prevalence of the investigated parasites in backyard flocks was lower than expected as chickens with access to the outdoors generally have higher rates of parasites [44,45].The comparatively low prevalence of parasites is likely underestimated. Shedding patterns of the parasites may have an effect on the results. Eimeria are shed in variable amounts based on days post-infection [46]. Compared to our and similar results using coproscopy, detecting worms macroscopically during necropsy resulted in significantly higher prevalence of more than 65% [19,47]. A study that looked into diurnal fluctuations of nematode egg excretion found that egg shedding was higher during the day, early morning to noon, than in the afternoon and night [48], and H. meleagridis is shed only intermittently from chronically infected chickens [49]. In addition, sample quality was not always optimal due to flock owners collecting samples and not properly storing them; and for some parasites the McMaster flotation method using 1g of feces, especially of pooled samples, may be too insensitive to show all present dispersive forms of parasites.5. ConclusionsWe detected a variety of poultry parasites in the investigated flocks, posing a risk to commercial poultry and their owners. The results of this study will help to provide information to owners of backyard chicken flocks that can be used to design timely and appropriate extension/outreach material. Informing them of the types of internal parasites typically observed and what steps can be implemented to improve the health and wellbeing of birds will improve the overall health of backyard flocks. In addition, alerting flock owners of potential zoonotic parasites that are present in their backyard flock may lead to improvements in their biosecurity measures.

animals : an open access journal from mdpi

10.3390/ani13111773

PMC10252101

In this review, the nutritional value and functional properties of spray-dried animal plasma (SDAP) for use in pet food are discussed. Although SDAP is a co-product of meat production for human consumption, this ingredient has a high nutritional standard, technological properties for wet and dry pet food production, and its bioactive components have high potential for use with a focus on intestinal health, immunity, neuroprotection, and control of “inflammageing” in animals. These properties are discussed in this literature review.

Plasma is a co-product from pork and beef obtained during the processing of animals for human consumption. The spray-drying process maintains the solubility of spray-dried animal plasma (SDAP) and its nutritional and functional properties, making this ingredient multifunctional in human and animal nutrition. In pet food, SDAP has been used in the production of wet foods (pates and chunks in gravy) as an emulsifying and binding agent, with the potential to replace hydrocolloids partially or totally, which have some negative implications for digestibility, fecal quality, and intestinal inflammation. From a nutritional point of view, SDAP has high digestibility and an amino acid profile compatible with high-quality ingredients, such as powdered eggs. Studies in companion animals, especially in cats, have shown that SDAP is an ingredient with high palatability. Despite the immunomodulatory, anti-inflammatory, prebiotic, and neuroprotective properties demonstrated in some animal models, there are still few publications demonstrating these effects in dogs and cats, which limits its use as a functional ingredient for these species. In this review, the potential use of SDAP in pet food, aspects related to the sustainability of this ingredient, and opportunities for studies in companion animals are discussed.

1. IntroductionIn the past decade, increased demand for high-quality raw materials, supply chain disruptions, and demand for ingredients with a low environmental impact have created many challenges in the animal feed and pet food industry [1]. The humanizing desire for high-quality food for dogs and cats along with the demand for highly digestible, palatable, and biologically valuable ingredients has created challenges and opportunities. There is an increased interest in protein ingredients that have functional properties, meet nutritional requirements, optimize health, and reduce the environmental impact. Some functional animal and plant-origin ingredients have the potential to meet these demands and spray-dried animal plasma (SDAP) is one of them.Plasma is a co-product from pork and beef obtained during the processing of animals for human consumption. Anticoagulants are added to the blood during collection. Then, the blood is centrifuged to separate the plasma and cellular fractions (red blood cells and platelets), and these fractions are subsequently spray-dried into a powder form (Figure 1). This processing system maintains the solubility and functionality of the ingredient when added to dry or wet food for dogs and cats.SDAP is used in both human and animal food. In pet food, it can be used for several purposes. For wet food with a high moisture and fat content that has the possibility of particle segregation, the main technological application of SDAP is as an emulsifying and binding agent (“binder”) to improve water retention, texture, juiciness, and homogenization of the final product [2,3,4]. In dry and extrusion-processed pet food, the technological properties of SDAP have not yet been widely explored, although emulsifying agents and binders are commonly used to improve the kibble characteristics.The inherent solubility and functional properties of plasma are maintained after manufacturing. Plasma as a fraction of animal blood is a very nutritionally rich ingredient with high concentrations of essential amino acids and bioactive components that confer its biological functionality [5,6]. The high amino acid concentration, immunoglobulins, bioactive peptides, growth factors, enzymes, and metalloproteins provide immunomodulatory [6,7,8], prebiotic [5,9,10], anti-inflammatory [9,11] and neuroprotective properties [5,12,13]. Figure 2 shows a summary of the main applications of this ingredient in dog and cat food.Considering the pet food processing, functional properties, and nutritional advantages of SDAP, this review aims to discuss the results of available studies with this ingredient, its main applications currently used by the industry, and perspectives for innovative uses in dog and cat nutrition.The following areas will be covered: sustainability of SDAP, wet pet food processing, nutritional value for dogs and cats, immunological effects, the use in aging animals, limitations on the use in pet diets, future options applying SDAP in pet food, and conclusions and perspectives.2. SDAP as a Sustainable IngredientHistorically, animal blood from the meat industry was considered a waste product with the potential to cause environmental pollution. Therefore, it was processed for use in animal feed, and this destination was considered the most viable from an economic and environmental point of view [14]. Higher-value animal blood co-products obtained from industrial processes to produce beef or pork for human consumption, such as SDAP, have since been developed. The process of using waste from the production of food for humans and its use in other stages of the production chain is known as “Industrial Ecology” [15] and, in the case of slaughterhouse waste, “Rendering or Animal Recycling” [14]. Animal recycling contributes significantly to the reduction in the emission of greenhouse gases if the destination of these co-products were different, such as landfills or composting [14].Life Cycle Assessment (LCA) is a way to quantify the environmental impacts of production and use of products and has been applied as a tool in choosing ingredients in animal formulations, aiming to minimize the impacts of formulations on animal production. For this reason, industrial co-products of animal and vegetable origin that are not used for human food benefit the circular economy, which promotes the continued use of parts of a primary resource, successively minimizing the waste generated while at the same time giving purpose to the co-products generated in the different stages of the production chain. Especially, meat co-products minimize environmental impacts and have been the focus of research to enhance their nutritional value [16,17].Pet food uses many co-products of plant and animal origin that are not consumed by humans. Although the volume of pet food produced globally represents less than 10% of the global production of animal feed, this market uses more than 30% of all co-products generated in the process of animal recycling and therefore, has an important role in the current use of industrial co-products. In the case of co-products from the rendering process, LCA recommends that the allocation of their impacts be performed economically [18,19] since it would not be correct to attribute similar impacts to the products generated from rendering, which precisely has the role of not disposing meat production waste to the environment.Ingredients such as beef tallow, chicken fat, and animal-based meal, including SDAP, have an important role in minimizing the environmental impacts caused by the production of food for human consumption in the circular economy.3. Wet Pet FoodThe protein content ranges from 70% to 80% of the final spray-dried plasma ingredient, depending on the membranes used for plasma concentration before drying. The drying stage uses reverse osmosis to remove water for the product with 70% protein or ultra or nanofiltration that, in addition to water, removes some salts for the product with 80% protein. SDAP is mainly composed of albumin, immunoglobulin G (IgG), and coagulation proteins [20]. Because the spray-drying process preserves the quality of plasma components and maintains their solubility [21], SDAP has important functional properties in wet food processing to produce pate or chunks in gravy and can be used as an emulsifying and binding agent, having a role like hydrocolloids [20].Hydrocolloids represent a diverse group of readily dispersible, fully, or partially water-soluble, long-chain polymers that increase in volume in water. They change the physical properties of the environment by forming gels, thickening, emulsifying, recoating, and stabilizing food components [22]. Although plasma is not included in the hydrocolloid group, which is essentially composed of polysaccharides and collagen, SDAP has very similar properties and is commonly used in wet foods for this purpose. For this review, a market investigation by one of the largest e-commerce retailers in Brazil (www.petlove.com, accessed on 3 February 2023) was performed; among 25 brands of wet pet foods, 44% of these wet pet foods declare SDAP in the composition, probably as an emulsifying or binding agent. Other typical hydrocolloid agents were declared in pet food that did not include SDAP, such as xanthan gum, guar gum, carrageenan gum, collagen, and modified starch. Other agents commonly used for this purpose are wheat gluten, soy protein, and whey protein.Wet pet foods are mainly formulated with co-products from slaughterhouses that have high protein, lipid, and moisture content (70–85%). Gelling agents and emulsifiers are used to avoid particle segregation and improve the texture and homogenization [23].In a study comparing the technological properties of binders commonly used in wet food to form chunks in gravy or loaves, the inclusion of 1.5% or 2.5% SDAP increased the hardness and reduced water loss when compared to wheat gluten [4]. Similarly, in another study comparing the inclusion of 2% SDAP with 2%, 4%, or 6% wheat gluten, there was a 2.5-fold increase in chunk hardness for the 2% SDAP, and juiciness was improved by approximately 20% because of the increased absorption of water from the gravy in contact with the chunk [24]. Juiciness was defined in that study as the amount of gravy that was absorbed by the chunks after cooking and corresponding to the change in weight of chunks after cooking and storage for two weeks. Therefore, juiciness can be described as the ability of meat to release juice (defined as the quantity of water preserved after cooking) when a pet chews it. The sensation of juiciness happens in two ways: the first corresponds to the quantity of juice that is released into the mouth when the meat is first chewed, and the secondary perception of juiciness is due to salivary flow stimulated by the presence of fat in the mouth. Juiciness or succulence is well correlated with improved palatability and enjoyable eating for humans and our pets.Figure 3 demonstrates the emulsification properties of SDAP in pate products.SDAP confers toughness to the pate due to its high-water retention capacity when included at 20% in the recipe as compared to 20% wheat gluten (WG). This is shown in Figure 4.No studies were found in the literature that directly compared the gelling properties of SDAP and hydrocolloids. Hydrocolloids are considered additives, while SDAP is a high-protein ingredient for pet food that provides technological and nutritional properties, and it is used mainly because of these characteristics. Among the typical hydrocolloids, carrageenan gum has a high gelling and emulsifying capacity [23] and together with other hydrocolloids, such as xanthan gum, guar gum, or locust bean gum, are the most used in pet food.Although various gums have important technological actions, there are some undesirable effects associated with their use [25]. Dogs that were fed diets containing 0.4% guar gum had reduced protein digestibility and stool quality [26]. Similar effects were also reported with a mixture of guar gum and carrageenan, although the authors also reported intestinal fermentative benefits with these additives [27]. Another undesirable effect of the use of hydrocolloids is their ability to induce intestinal inflammation and gastric ulcers as described in rats, mice, rabbits, and guinea pigs after ingesting carrageenan gum or carboxymethyl cellulose [28,29].On the other hand, faster healing of gastric ulcers was observed in pigs after ingesting plasma proteins in their drinking water when compared to the control group that did not receive the addition of these proteins in the water [30]. In addition, performance horses that consumed a supplement with plasma-based proteins prevented the development of stress-induced gastric ulcers [31]. However, no studies have evaluated the effects of plasma proteins on ulcers in dogs and cats. SDAP is a highly digestible ingredient with immunomodulatory and anti-inflammatory properties that could be used at higher levels to partially reduce the inclusion level of hydrocolloids to avoid undesirable effects while maintaining the desirable technological characteristics in the pet food product.4. Nutritional Value for Dogs and CatsThe main nutritional attribute of SDAP is its high protein concentration with an appropriate composition of essential amino acids compared to other ingredients used in pet food. However, due to the differences in protein concentrations between ingredients, the amino acid score (AAS) calculation represents an adequate comparative measure between protein sources [32]. The AAS is calculated by the amino acid concentration within the food source, divided by the recommended intake of that amino acid from reference tables for the species. In the case of dogs and cats, currently, the most used tables are from the European Pet Food Industry Federation (FEDIAF) and the Association of American Food Control Officials (AAFCO).Compared to other protein ingredients, SDAP has the highest total AAS for lysine, tryptophan, and threonine, which, together with methionine, are the main limiting amino acids in diets for dogs and cats [33]. The superior amino acid profile of SDAP is also accompanied by the higher digestibility of this ingredient in dry [34] and wet [35] foods, which improves the biological value of the protein compared to conventional sources. Table 1 shows a comparison between SDAP and the main protein sources used in pet food [18,36].The AAS of SDAP is adequate and superior to other ingredients commonly used in pet food but is relatively low in methionine (Figure 5).Plasma protein is composed mainly of albumin, globulins, and other components normally found in the blood. Because of this composition, the main nutritionally focused application is to meet biological needs for amino acids. The nutritional value of a protein source is determined primarily by its digestibility and its amino acid composition concerning the species’ requirement (AAS), so the latter directly influences the biological value [38].As shown in Table 1, SDAP has an amino acid profile comparable to high-quality protein sources such as egg powder, except for the relatively lower levels of methionine in SDAP. There have been a few studies evaluating SDAP as a protein ingredient in diets for dogs and cats. In three studies, dogs that were fed diets with up to 3% SDAP had increased apparent dry matter (DM) digestibility in all experiments and had increased crude protein digestibility in two of the studies, with no changes in stool quality [34]. Similar results were found using cat diets containing 3% SDAP or 3% wheat gluten as binders [35]. Cats that were fed diets with SDAP had higher DM digestibility without any significant changes in protein digestibility, and there were significant reductions in the volume of feces excreted by the animals.While few studies have evaluated the digestibility coefficients of SDAP in companion animals, this ingredient has been studied more extensively in pigs and poultry to minimize the use of antibiotics and improve immunity indicators, due to its potential benefits on the digestive system. The digestibility of SDAP by poultry and swine is equivalent to other highly digestible ingredients such as egg powder [39], brewer’s yeast extract [29,40], whey protein concentrate [41], and hydrolyzed fish meal [41]. Despite the similarity with other highly digestible ingredients, SDAP has consistently shown additional benefits in weight gain, feed conversion, and improved disease resistance when included in diets for weaned pigs [42].The biological value of ingredients is important because protein sources with high biological value can be included in lower concentrations and can reduce formulation costs. A low biological value for wheat gluten was observed, resulting in weight loss in puppies that were fed low-protein diets when compared to other sources such as hydrolyzed casein and soy protein isolate [38].A widely used measure of protein quality in humans is the Protein Digestibility Corrected Amino Acid Score (PDCAAS). This measure is obtained by multiplying the amino acid score (AAS) by the apparent protein digestibility of a given source [32]. Using the data in Table 1 and the reported digestibility of 90% for both SDAP and WG [38,39], the PDCAAS for lysine, methionine, and tryptophan are very low for WG, which contributes to the low biological value of this ingredient, while SDAP presents lower values only for methionine. Among the protein sources compared in Figure 6, the PDCAAS of SDAP is higher for the main amino acids than the other ingredients, including egg powder, with the exception that methionine and cystine are lower in SDAP.5. Immunological Effects of SDAPDietary SDAP also has immunological properties in the digestive system, but these effects have not been determined for dogs and cats. In swine and rodent models, some mechanisms of action of SDAP in intestinal protection are proposed in Table 2.Among the benefits to the digestive tract with the use of SDAP, its prebiotic and modulating effects on local gut immunity have been studied. In a study comparing the effects of 0% or 8% SDAP in diets fed to mice on intestinal microbiome populations [9], mice that were fed SDAP had a significant increase in the count of microorganisms of the Phylum Firmicutes that is generally associated with the production of short-chain fatty acids (SCFA), which regulate the growth of pathogenic microorganisms and improve gut immunity. Among the genera that showed the highest increase were Lactobacillus spp. and Blautia spp. The changes in the intestinal microbiota for mice that were fed SDAP were associated with an increased expression of mucosa interleukin 10 (IL-10), transforming growth factor beta (TGF-β), mucin 2 (MUC-2), and trefoil factor 3 (TFF3), all of which increase immune tolerance of the intestinal mucosa to minimize the risk of inflammation. Similar results of SDAP modulating the intestinal microbiota have been observed in pigs and even in fish [45,46].Part of the intestinal protector effects associated with the consumption of SDAP are related to its high concentration of immunoglobulins, especially IgG, which confers a direct effect against pathogenic microorganisms and prevents lesions in the intestinal mucosa. Mice genetically predisposed to inflammatory bowel disease were fed diets containing 2% bovine serum immunoglobulin isolate (BSI) [10] or 8% SDAP [5]. Both studies showed that BSI and SDAP prevented the increase of pro-inflammatory mucosal cytokines and chemokines including IL-2, IL-6, IL-17, MIP-1β, and MCP-1. Furthermore, BSI increased the expression of anti-inflammatory cytokines TGF-1β, while SDAP increased IL-10. Thus, immunomodulatory components in SDAP and BSI play an important role in protecting and improving intestinal immune tolerance.The immune modulatory effects related to SDAP supplementation are not limited to the local intestinal mucosa. It was already observed that dietary SDAP reduced acute pulmonary inflammation induced by lipopolysaccharide inhalation in mice [47] and improved pregnancy rates and favorably altered the uterine mucosa cytokine profile in transport-stressed mice [48].An important aspect to consider when using bioactive compounds in animal nutrition is their viability along the digestive tract. Studies have verified in dogs and cats that 7.6% and 4.9%, respectively, of the porcine IgG ingredients consumed remain viable in feces, demonstrating that IgG can have activity throughout the digestive tract [49]. However, the porcine IgG ingredients used in these studies were applied after the kibbles were extruded because it was unknown if the extrusion process could inactivate their functions.Plasma proteins in immunologically challenged animals show positive effects on the growth rate, feed intake, and general health conditions when compared to animals not receiving these proteins. Weaning stress can reduce feed consumption, resulting in a compromised intestinal barrier and predisposing pigs to diarrhea and infections [50]. A recent meta-analysis [6] reported that increasing the concentrations of SDAP by up to 10% in the diet for weaned pigs significantly increased feed intake, showing that SDAP was palatable and minimized the negative impacts of postweaning stress on pig performance. The mechanisms of action and the effective concentration of SDAP in the diet of various animal species still need to be better defined (Table 3).However, the SDAP effects on animal performance improvement are not only attributed to the increased feed intake but also to the presence of immunoglobulins, growth factors, and other proteins present in the plasma, which have a positive effect against important intestinal pathogens such as Salmonella typhimurium, Salmonella enteritidis, Escherichia coli, Staphylococcus spp. and their enterotoxins [47,85,86]. The intestinal and immune-related benefits of SDAP seem to be mostly related to the activity of immunoglobulins and/or bioactive peptides produced during the digestion of SDAP in the gastrointestinal tract. Other authors have suggested a prebiotic effect of SDAP on promoting the growth of beneficial intestinal microbiota. This may help the control of pathogenic microbial populations in the gut and therefore reduce antigen stimulation and the subsequent production of pro-inflammatory cytokines and chemokines such as IFN-γ, TNF-α, IL-1, IL-6, and LTB-4. These compounds are associated with the activation of gut-associated lymphoid tissue (GALT), which promote mucosal lesions and predispose the animals to microbial translocation [9,58].The effects of SDAP on increasing intestinal immune tolerance seem to be primarily related to an increased production of inflammation-regulating cytokines and chemokines, such as IL-10, TGF-β, β-defensin, iNOS, integrins, and other adhesion molecules [86]. However, microbiota interactions can also modulate gut tolerance (Figure 7).Inflammatory bowel disease (IBD) in dogs is the most common cause of intestinal disease and may be associated with parasitosis, food allergies, idiopathic inflammation, or a previously existing disease. The primary changes in dogs with IBD are increased intestinal permeability, increased dysbiosis index (Figure 8), increased cellular infiltrates of mononuclear cells, and increased pro-inflammatory cytokines (IL-2, IL-6, TNF-α, IFN-γ, and IL-1β) and C-reactive protein [88]. Although SDAP supplementation has not been studied in dogs with IBD, many of its attributes could help alleviate clinical conditions. Supplementation with BSI reduced irritable bowel syndrome in humans [89,90].According to a systematic literature review [91] that included the results of 11 publications about animals that were immunologically challenged, a consistent effect of plasma protein-derived ingredient supplementation on daily weight gain, feed intake, and feed conversion was reported, along with a significant reduction in pro-inflammatory cytokines such as IL-6, TNF-α, and IL-1β. The authors suggested studies using humans with enteropathies were needed given the consistent findings in challenged animals. Similar results of dietary SDAP on pig performance were described in a systematic review [92]. However, in this review, only results from experiments without intentional challenges to the animals were used but improved physiological conditions in animals supplemented with SDAP were reported.Inflammation associated with IBD may be caused by an alteration in the balance between Treg and proinflammatory-activated Th cells [71,88]. SDAP supplementation reduced the ratio between activated Th lymphocytes and Treg lymphocytes, indicating that SDAP restores the balance between these lymphocyte populations [11]. A similar response observed in mice genetically predisposed to IBD has been observed in a Staphylococcal enterotoxin B model of mild intestinal inflammation [51] and under acute lung inflammation induced by LPS [47].6. SDAP in Aging AnimalsThe aging process in dogs and cats is very similar to humans [93]. General changes in body function occur because of reduced adaptive capacity, which in part results from a process known as “inflammageing”. This is characterized by mild to moderate immunological stimulation, which culminates in oxidative, inflammatory, and degenerative modifications in some organ systems [94]. Among all of them, the changes in the nervous system seem to be more evident because they are more easily perceived, since the animals show deficits in locomotion, sense organs, learning ability, changes in the sleep–wake cycle, and even in learned functions, such as the habit of defecating and urinating in inappropriate places. Such changes are part of a syndrome known as Cognitive Dysfunction, which, from a pathophysiological standpoint, is very similar to Alzheimer’s disease in humans [95]. Currently, the decline in cognitive functions has been partially associated with modifications in the gut microbiota [96], and for this reason, studies on the brain–intestine interaction (“gut-brain axis”) have gained relevance in the search for nutritional strategies to minimize the effects of aging in animals.Based on this principle, the association between the low-level inflammation and the decline of cognitive functions in mice predisposed to early aging (SAMP8) and how supplementation with 8% SDAP was able to minimize the impacts of aging were studied [12]. The authors showed a significant improvement in the cognitive functions of animals supplemented with SDAP compared to the non-supplemented ones (negative control group) using short- and long-term memory tests. These effects were partially attributed to the reduction of pro-inflammatory (IL-6 and NF-κβ) and oxidative (hydrogen peroxide concentration) markers, increased expression of inflammation-modulating cytokine (IL-10), and increased adhesion molecules at the blood–brain barrier (e-cadherin and ZO-1), which reduce capillary permeability in the central nervous system, protecting neurons from potentially toxic compounds that can induce neuronal degeneration.In addition, the neuronal protective effects of SDAP were attributed to its effects on the gut in two other publications [5,83] by this same research group, which demonstrated that aging in mice induces mild-grade inflammatory changes and that senescence promotes an increase in populations of potentially pathogenic microorganisms, with a reduction in microorganisms with probiotic potential (Lactobacillus hayakitensis and Blautia hansenii) that promote protection against infections by Clostridium spp. In this study [83], supplementation of SDAP reversed such deleterious changes in the aging animals, promoting an effective response to an intestinal challenge.Two of the main neuronal changes identified histopathologically in Alzheimer’s disease are the presence of the amyloid precursor protein (APP) and neurofibrillary tangles. In SAMP8 mice, the SDAP supplementation reduced the expression of proteins related to these two Alzheimer’s disease-related structures [13].Despite the lack of studies using aging dogs and cats supplemented with SDAP, the results obtained with aging mice suggest important benefits for improving quality of life by minimizing the effects of inflammageing and preserving cognitive functions.7. Limitations on the Use of SDAP in Pet Food DietsThere are some factors that can limit the use of SDAP in pet food. First, SDAP comes from an undervalued raw material (blood), but the process to separate the plasma from the red cells and the dehydration by spray-drying technology to maintain the high solubility of this ingredient increase its cost. So, pet food producers need to account for it in the cost of their final recipe, which could impact the inclusion levels in the product. Second, when formulating SDAP into diets or supplements, the nutrient profile including amino acid and minerals should be accounted for during formulation to meet desired nutritional requirements for complete food and supplements. Third, researchers still need to investigate whether SDAP should be applied before or after extrusion for optimal results [34]. When applying externally, methods of application such as tumbling or vacuum coating can impact effectiveness of application and limit the level applied. When blended internally before extrusion, further research is needed to understand the impact on SDAP biological activity, because only one digestibility study has evaluated the internal versus external application of SDAP [34].8. Future Options Applying SDAP in Pet FoodConsidering the multiple functions of SDAP presented in this review, it is important to further investigate its effects and concentrations in dogs and cats, due to its high potential for applications, especially in dry pet food, treats, and supplements. The use of SDAP in dry pet food opens new opportunities for pet food producers to develop a diversity of new concepts, such as the use of SDAP to improve cognitive functions and mobility in diets oriented to senior pets. Additionally, there is an opportunity to use SDAP in veterinary diets for pets with gastrointestinal disorders, especially for companion animals presenting IBD or other inflammatory disorders. Furthermore, due to the prebiotic effect of SDAP, its use in puppies and kittens can help develop a beneficial microbiota in early life stages to help pets develop a strong, healthy gastrointestinal tract. In addition, the binding properties of SDAP may help the development of dry pet food recipes with high fresh meat inclusion, while maintaining the physical properties of dry kibbles.9. Conclusions and PerspectivesThe nutritional composition of SDAP along with its bioactive components make this animal co-product an ingredient with high digestibility and biological value that also has prebiotic, immunomodulatory, and neuroprotective properties, with multiple potential new applications for its use in wet and dry pet food products. The technological benefits of SDAP as an emulsifier and binder in wet food are well demonstrated, and its use by the industry is widely recognized.In addition to the above-mentioned properties which make SDAP a multifaceted ingredient, it also contributes to a reduced environmental impact when compared to other ingredients, even those of plant origin, since it is produced from industrial waste from human food production and is a sustainable ingredient.Although all the benefits of SDAP have been demonstrated for dogs and cats, it is necessary to conduct more studies with a dose–response for SDAP, because publications about wet and dry pet foods with a technological and nutritional focus were evaluated with formula inclusions of up to 2%, while studies focused on its functional properties used 8%. Thus, there is a gap between the formula levels used that needs further study to determine the optimum dietary concentration of SDAP in various applications for pet food products.

animals : an open access journal from mdpi

10.3390/ani12010094

PMC8749871

Litter size is an important economic trait in pigs. Improving the number born alive is an important breeding goal of the pig husbandry. A shorter farrowing duration is welcome for facilitating the management and sows’ health. Therefore, the aim of this study was to explore the effect of litter size and parity on farrowing duration, to determine whether a shorter length of farrowing duration could be considered as a breeding parameter in pig breeding. Our results showed the total number born had no significant relation with farrowing duration, but number of stillbirths increased with the prolongation of farrowing duration and decrease of live litter size if farrowing duration was longer than 240–300 min. Different parities sows had little difference in the same farrowing duration interval except for gilts. A shorter farrowing duration within 300 min might be considered in pig breeding without worrying about the decreasing of live litter size or the negative effect of parity.

Litter size has increased and farrowing duration has also prolonged in recent years. The aim of this study was to analyze the effect of litter size and parity on farrowing duration (FAR) to estimate the possibility of selecting a short farrowing duration. We recorded 32,200 parturitions of 8420 Landrace × Yorkshire sows, determined farrowing duration, litter size, parity, gestation length. Results showed that total number of born (TNB) and parity obeyed a cubic (p = 0.0004, p = 0.004) relationship while number born alive (NBA) and number born dead (NBD) obeyed a linear (p = 0.0239, p = 0.0035) relationship with FAR. Gestation length obeyed a linear (p = 0.02) relationship with FAR. FAR of sows with stillbirth was longer than that of sows without stillbirth. Stillbirth rate increased rapidly from about 2% to 4%, especially when FAR was over 240 min. FAR gradually prolonged with the parities. FAR of 7th parity sows was longer than that of 1st~6th parity sows (p < 0.05), but different parity sows had little difference in the same FAR interval except for gilts. Results indicated it was possible and necessary to consider FAR into pig breeding without worrying about decreasing of live litter size or negative effect of parity if FAR was shorter than 300 min.

1. IntroductionReproductive traits are the most important concepts in determining the output of a pig farms. Litter size is an important reproductive trait, and it is one of the decisive factors affecting the economic benefits of the entire pig industry. Over tens of reproduction traits are considered in calculating the breeding index to rank and select breed pigs. However, the average stillbirth rate of piglets was still 3–8%, of which approximately 75% happened during parturition [1], the increase of stillbirth rate would inevitably reduce the profits of pork production [2,3]. Larger litter size may lead to a longer farrowing duration (FAR) and is related to the lower birth weight of piglets [4]. It has been suggested that large litter size can also lead to the increase of stillbirth rate, which is likely to bring significant negative impact on animal welfare [5]. Farrowing is known to be a stressful process of both piglets and sows, and a longer duration of farrowing process often requires extra care from nursing staff and sometimes impairs the uterus health of sows, which would bring certain difficulties in management and even a fertility reduction of sows during the next parturition. Many stillbirths occur during parturition due to dystocia and prolong farrowing.Duration of farrowing (FAR) was reported to prolong with the number of stillbirths [6]. A prolonged FAR is said to impair fertility in pigs [7], as sows with a longer FAR tended to have a higher risk of post parturient disorders [8]. However, at the same time, the prolongation of FAR was also in relation to the increase of litter size [9,10]. In addition, genetic, breed, age, length of gestation, feeding management, environment and nutritional factors could also affect FAR [11]. There have been some studies focused on the environment conditions control in the farrowing house in order to ensure the smooth production of piglets [12,13,14].The aim of this study was to explore the exact effect of litter size including total number born, number born alive and number dead on FAR. We also wanted to evaluate the impact of parity on FAR and litter size in order to evaluate the possibility of short FAR as a breeding indicator for pigs.2. Material and Methods2.1. AnimalsThe observational study was conducted in 2017–2018 on a commercial pig farm in Central China. Parturitions (n = 32,231) of 8420 randomly selected healthy Landrace × Yorkshire hybrid sows between 1st and 10th parity were included in this study. The sows farrowed in 26 batches. Pregnant sows were fed 1.8–2.0 kg/day for the first 7 days; 2.0–2.5 kg/day from days 7 to 30; 2.5–3.0 kg/day from days 30 to 70; 2.5 kg/day from days 70 to 90; and 3.5 kg/day from days 90 to 110. The diets were formulated according to their body condition. During 3 days before farrowing, sows were fed 2.5 kg/day. On the day of farrowing, sows were fed 0.5–1.5 kg/day. For the first 5 days after parturition, sows were fed 2.0, 2.5, 3.0, 3.5, and 4.0 kg/day, respectively.All pregnant sows were fed in groups and transferred to farrowing houses 1 week before farrowing and fed in single crates in farrowing houses with free access to drinking water. The farrowing pens were 2.50 × 2.30 m2 in size, the farrowing crates were 2.10 × 0.70 m2 in size. The average room temperature was 25 °C. The environment of this pig farm was controlled by Big Dutchman ventilation system and cooling pad in summer and warm air heating system in winter. All sows were raised under regular disinfection and vaccination procedures (Table 1).2.2. MeasurementsDuration of farrowing (FAR) was defined as the time interval between the first birth to the complete expulsion of placenta. We also determined the following sow traits: farrowing batch, farrowing barn, farrowing house, farrowing pen, the length of pregnancy, parity, feeders and nursing staffs. The length of pregnancy was defined as the time interval between fertilization and farrowing. Total number born piglets (TNB), number of born alive piglets (NBA) and number of born dead piglets (NBD) in each litter were investigated by nursing staffs. TNB consisted of NBA, NBD and mummies. Piglets that had no breathing and heartbeat at birth, or had heartbeat but no breathing and unable to breathe after rescue, were defined as NBD.FAR and delivery mode (normal delivery and assisted delivery) were recorded by the nursing staff who works day and night shifts in farrowing houses. Nursing staffs’ job is for delivery and nursing. When the sow’s amniotic fluid is out, they need to wipe the sow’s udder and hind body with warm water and disinfect with 0.1% potassium permanganate solution and wait for the sow to farrow. They need to keep an eye on each sow in farrowing and keep an eye on each sow as she gives birth, in order to deal with the dystocia of sows, the suspended death of newborn piglets, and to cut umbilical cord, to assist the piglets to keep warm and access the colostrum as soon as possible.The assisted delivery was adopted if sows had strong parturition symptoms such as giving an expulsion effort but no piglet was delivered after 1–2 h of amniotic fluid outflow, or a fetus was present in canal or uterine cervix but sows had not enough labor force or the litter interval was more than 1 h. The assisted delivery consists of an injection of 0.5–1.0 mL oxytocin or and an artificial assisted delivery.2.3. Statistical AnalysisAll the data was analyzed by SPSS Statistics 26 (IBM SPSS Statistics, IBM Corp, Armonk, NY, USA) and presented with R software. Variables including FAR, TNB, NBA, NBD, parity (P) and gestation length (G) were checked for normal distribution. An initial univariable screening was performed to identify potential influencing factors of FAR and litter size. The correlations of FAR and TNB, NBA, NBD, P and G were tested and estimated. The regression model that had the highest fitting degree and achieved significance level was selected. Based on the significance levels of those tests, a multivariable linear model for FAR was created. Farrowing house, farrowing pen, nursing staff, feeder, and batch were included as random factors and delivery mode as fixed effects. The correlations of parity (P) and gestation length (G) on litter size were tested and estimated. Based on the significance levels of those tests, a multivariable linear model for TNB and NBA and a lineal model for NBD were created. Farrowing house, farrowing pen, nursing staff, feeder, and batch were included as random factors and delivery mode as a fixed effect. Results were presented as Mean ± SD and considered significant at p ≤ 0.05.3. ResultsParturition records of FAR longer than 10 h (n = 225), live litter size less than 5 (n = 1172) and more than 19 (n = 31) were excluded. Data of parity ≥ 7 were classified as parity 7. A total of 32,200 parturitions of 8420 Landrace × Yorkshire hybrid sows were included in the following analysis: 3451 assistant delivery parturitions (10.7%). Distribution of farrowing duration, litter size, parity and gestation length were presented in Figure S1.3.1. Effects of Litter Size on Farrowing DurationTotal number of born (TNB) (5–20), number born alive (NBA) (5~18) and number born dead (NBD) (0–9) obeyed a cubic (R2 = 0.77, p = 0.0004; R2 = 0.54, p = 0.0442; R2 = 0.79, p = 0.0043) relationship with farrowing duration (FAR) (10.2–600 min) (Figure 1). In the final multivariable model for FAR, the interact of TNB and NBA (TNB × NBA) were significant (p = 0.012). The average FAR was 250.05 ± 67.63 min, TNB was 11.61 ± 2.45, NBA was 11.29 ± 2.42, and NBD was 0.26 ± 0.58, respectively. An overview of effect of litter size on FAR was listed in Figure 1 and Table 2, Table 3 and Table 4.From Table 2, farrowing duration indicated no significant difference between different TNB, but significant difference between different NBA (p < 0.05). Sows of different TNB and NBA had almost the same FAR variations, about 27%. The length of FAR first decreased and then increased with the increase of NBA. When litter size was less than 11, NBA took a little longer FAR than TNB. When litter size was over 12, NBA generally took a shorter time to farrow than TNB. Sows with 5 and 16 NBA had the longest FAR of 257.49 ± 76.8 min and 256.62 ± 67.61 min, longer than sows with 13, 17 and 18 NBA. Sows with 18 NBA had the shortest FAR of 239.98 ± 65.88 min, which was shorter than that of all the other sows, followed by sows with 17 and 13 NBA of 246.21 ± 65.05 min and 247.19 ± 65.89 min.From Table 3, 78.69% parturitions (n = 25,339) had no NBD. 17.95% parturitions (n = 5779) had 1 NBD. 2.56% parturitions (n = 824) had 2 NBD. 0.50% parturitions (n = 162) had 3 NBD. 0.14 % parturitions (n = 44) had 4 NBD. 0.09% parturitions (n = 30) had 5 NBD. 0.04% parturitions (n = 14) had 6 NBD. 0.01% parturitions (n = 3) had 7 NBD. 0.01% parturitions (n = 4) had 8 NBD. 1 parturition had 9 NBD. Farrowing duration indicated significant difference between sows with different NBD (p < 0.05). Length of FAR of sows with 0, 1 and 5 NBD had about 27% variability. Length of FAR of sows with 6, 7 and 8 NBD had over 38% variability. FAR prolonged with NBD from 0 to 4. FAR of sows with 0 NBD were 62.64 min longer than that of sows with 4 NBD (247.18 ± 66.02 min vs. 309.82 ± 99.09 min) (p < 0.05). Sows with 6 NBD had the longest FAR, 72 min and 63 min longer than sows with 0 NBD and 1 NBD (319.24 ± 153.66 min vs. 247.18 ± 66.02 min, 256.60 ± 68.44 min) (p < 0.05).From Table 4, after dividing FAR into 5 groups: 0–180 min, >180–240 min, >240–300 min, >300–360 min and >360–600 min, number born dead and stillbirths rate increased with the length of FAR, the longer the FAR, the more the NBD (from 0.21 to 0.45) and stillbirth rate (from 1.83% to 0.45%), especially when FAR was longer than 240 min. When the length of FAR was over 240 min, there would be significantly more NBD (p < 0.05).3.2. Effects of Parity on Farrowing Duration and Litter SizeParity obeyed a cubic relationship with FAR (R2 = 0.98, p = 0.0043), TNB (R2 = 0.997, p = 0.0003), NBA (R2 = 0.997, p = 0.0002) and NBD (R2 = 0.979, p = 0.0052) (Figure 2). An overview was shown in Figure 2 and Table 5 and Table 6. From Table 5, FAR gradually prolonged with parity, from 212.65 ± 61.55 min to 260.42 ± 68.02 min. 1st parity sows (n = 89) had a FAR of 212.65 ± 61.90 min, 2nd parity sows (n = 6915) of 243.01 ± 67.68 min, 3rd parity sows (n = 7477) of 247.23 ± 67.23, 4th parity sows (n = 5078) of 250.23 ± 65.84, 5th parity sows (n = 3559) of 251.10 ± 67.28 min, 6th parity sows (n = 3065) of 252.01 ± 68.75 min, 7th parity sows (n = 6018) of 260.40 ± 68.02 min, respectively. Sows of parity 1 and 2 had a significantly shorter FAR, while sows of parity 6 and 7 had a significantly longer FAR (p < 0.05).TNB and NBA increased from 10.83 ± 2.35 and 10.59 ± 2.34 to 11.99 ± 2.44 and 11.66 ± 2.39 with the increase of the first four parities, and decreased to 11.03 ± 2.38 and 10.65 ± 2.34 with the increase of the last four parities. NBD always increased as the increase of parities, from 0.16 ± 0.40 to 0.32 ± 0.65, by 0.16 ± 0.25. Sows of parity 4 had the most TNB (11.99 ± 2.44) and NBA (11.66 ± 2.39). There was no significant difference among TNB and NBA of sows in parity 3 to 5, or among NBD of sows in parity 2 to 6. Sows of parity 3, 4 and 5 had significant more TNB and NBA than sows of parity 1, 2, 6 and 7 (p < 0.05). Sows of parity 7 had significant more NBD than sows of parity 1, 2 and 3 (p < 0.05).Then, we looked into if there was significant variation among seven parities within each of the 5 FAR groups: 0–180 min, >180–240 min, >240–300 min, >300–360 min and >360–600 min (Table 6). Significant differences were detected when FAR was of 0–180 min, >240–300 min, and >300–360 min. When sows had FAR of 0~180 min and of >300–360 min, gilts had a shorter FAR than sows of all the other 6 parities (p < 0.05). When sows had FAR of >240–300 min, gilts had a shorter FAR than sows of parities 2 to 6 (p < 0.05).3.3. Effect Gestation Length on Farrowing Duration and Litter SizeGestation length (109–118 days) obeyed a lineal (R2 = 0.51) relationship with FAR, TNB (R2 = 0.946, p = 0.0004) and NBA (R2 = 0.915, p = 0002) (Figure 3). No significant relationship was found between gestation length and NBD. The average gestation length (G) was 114.28 ± 0.98 days. FAR was positively associated with length of gestation (β = 3.67; p = 0.02). An overview was shown in Table 7. There were only seven and three parturitions of gestation length of 109 days and 118 days. Additionally, significant difference was only found among sows of gestation length of 109 days and 118 days for FAR, TNB and NBA. After removing these 10 parturitions records, there was no difference in FAR between 110–117 days of gestation length. TNB and NBA of sows with 112 and 113 days of gestation were significantly higher than those in other gestations.4. DiscussionA longer duration of farrowing (FAR) is thought to be against the health of sows [15] and their next delivery performance which would reduce the working efficiency for feeders and nursing staffs in farrowing houses. Some studies also suggested a positive correlation between FAR and litter size, the longer FAR, the more litter size. Therefore, in this study, we investigated the influence of litter size including TNB, NBA and NBD on FAR and the impact of parity on both FAR and litter size using a large scale of farrowing data in order to look into whether it was reasonable to consider a shorter FAR as one of the breeding parameters in pigs in the further.By analyzing 32,200 parturition records from 8420 Landrace × Yorkshire sows, we found that the average length of FAR was about 250 min, or about 4.2 h, which was another piece of evidence that the FAR is getting longer and longer. In 2004, FAR was reported to last 133 min [9], and increased to 166 min [6] and 268 min [12], though the common use of oxytocin-like compound in pig production can shorten the length of FAR [16]. This might be partly because of the rapid increase of litter size in pigs since 2004. The average length of FAR in this study was significantly shorter than the 6~8 h of a batch of European superior sows [17]. Average TNB in this study was 11 compared to as high as 20 in the most hyper prolific sows in Europe.In this study, we found that the litter size and parity had an impact on FAR by using the multivariable linear model. The interaction between TNB and NBA were detected. In the study of Bjorkman et al., they used the multinomial logistic regression model to explore the relationship between litter size, parity and farrowing duration. They found a positive correlation between number born alive and FAR, and number of stillbirth was significantly correlated with FAR [18]. The coefficient of variation of TNB and NBA were larger, about 27%. Significant difference was not found in FAR between different TNB, but found between different NBA because of the interact between FAR and stillbirth [19]. Shortening the length of FAR will not reduce the effective NBA. FAR of >240 to 300 min might be the ideal cut-off point in this population (with 5 to 22 litter size). This was in coincidence with the general believing that FAR that exceeded 300 min was likely to cause dystocia and new-born piglet complications, and piglets that experienced a longer FAR are more likely to die at birth [17,20].Besides, NBD increased gradually with FAR, especially when FAR was longer than 240 min, the NBD increased rapidly and linearly. Studies have shown that a longer FAR could affect perinatal mortality or subsequent piglet growth. FAR and the birth order of piglets were reported to be the main factors determining the risk of stillbirth [20]. Some studies also pointed that there was a positive correlation between the number of born alive and FAR, and the number of stillbirths was significantly negatively correlated with FAR [18,21].In addition, we found that parity had a significant impact on FAR. As the parity increased, FAR gradually became longer. Studies show that parity has a slight effect on the number of colostrum metabolites, which might affect the reproduction performance of sows, including litter weight at birth and piglet mortality [22]. Parity and FAR could affect the incidence of postpartum disease in sows; if FAR was longer than 4 h, the sows would be at greater risk for fever 1 day after parturition [23], which affected the reproductive performance of sows and sped up the sows’ elimination. Besides, the effect of parity on FAR might be due to the aging of sows’ uteri after multiple parturitions, which weakens their muscles’ ability to contract during parturitions, resulting in prolonged FAR.Some studies have shown that FAR was not only affected by parity, but also by the sow’s physical condition, feeding and environment. For example, a higher energy intake in sows during late gestation could improve the farrowing duration traits [24], adding the Bacillus during the perinatal period of sows could shorten FAR and the weaning-estrous interval [25]. FAR was positively correlated with the gestation, and FAR with 118 day of pregnancy was the longest. However, there was no significant difference between gestation (110–117 days) and FAR after removing the data of pregnancy was 109 and 118. A study reported that the length of gestation (112–119 days) was negatively correlated with FAR, but the effect of litter size on FAR was not considered [6].In general, we found that litter size had an impact on FAR, and there was a positive correlation between parity and FAR. The total number born and the number born alive increased gradually in 1–4th parity. All of these indicated that we can select sows with a shorter FAR and of the 1–4th parity during pig breeding.5. ConclusionsIn summary, we found that number of stillbirths increased with the increase of farrowing duration and decrease live litter size if farrowing duration was longer than 240–300 min. It might be possible to choose a shorter farrowing duration without compromising decreasing litter size in pig breeding. In order to further confirm the effect of litter size and parity on farrowing duration, it is necessary to further explore it from the perspective of genome.

animals : an open access journal from mdpi

10.3390/ani11051475

PMC8161186

In recent decades, the use of phytogenic feed additives as natural growth promoters has been substantially increased in the poultry industry. Fermented pine (Pinus densiflora) needle drink is used as a functional beverage in Korea due to its antioxidant effects. Therefore, the aim of the current study is to determine the effects of fermented pine needle extract on laying performance, egg quality, lipid parameters, and lipid oxidation of eggs in laying hens. Supplementation of fermented pine needle extract in laying hens’ diet improved productive performance and egg quality parameters.

This study aimed to investigate the supplemental effects of fermented pine (Pinus densiflora) needle extract (FPNE) in laying hen diets on productive performance, egg quality, and serum lipid metabolites. A total of 108 40-week-old Hy-line brown laying hens were randomly assigned to one of the three dietary treatment groups: (1) basal diet + 0 mL FPNE/kg diet (CON), (2) basal diet + 2.5 mL FPNE/kg diet (T1), or (3) basal diet + 5 mL FPNE/kg diet (T2) for 6 weeks. Each group consisted of four replicates of nine hens each. Feed and water provided ad libitum. Results showed that dietary supplementation of FPNE increased egg production percentage (linear, p < 0.01 and quadratic, p < 0.05), egg mass (linear, p < 0.05), and feed intake (linear, p < 0.05) during the entire experimental period. In addition, dietary inclusion of FPNE significantly increased the eggshell color (linear, p < 0.01), egg yolk color (quadratic, p < 0.01), and eggshell breaking strength (linear, p < 0.05 and quadratic, p < 0.05) while the Haugh unit decreased (quadratic, p < 0.05). However, serum lipid profile did not differ among the dietary treatments (p > 0.05). Notably, antioxidant activity of egg yolk was improved by significantly decreasing the malondialdehyde content in egg yolks after 6 weeks of storage (linear, p < 0.001 and quadratic, p < 0.05). In summary, dietary inclusion of FPNE could improve laying performance and the antioxidant capacity of eggs.

1. IntroductionIn the past two decades, the increasing consumer demand for poultry products from hens raised without antibiotics has intensified the research on the development of antibiotic alternatives to maintain or improve poultry health and performance [1,2]. In this context, phytogenic feed additives (PFA), plant-based extracts containing various bioactive compounds such as polyphenols, organic acids, essential oils, terpenoids, and aldehydes, have gained significant attention, and are reported to have beneficial effects on animal production and health [3,4]. The antiviral, antimicrobial, antioxidant, and anti-inflammatory properties of PFA are thought to be their primary modes of action [5]. For example, Pinus densiflora of the family Pinaceae, commonly known as Korean Red Pine, has been a potential source of bioactive components. P. densiflora has the wide natural distribution area in Japan, Korea, North-Eastern China, and the extreme South-East of Russia. Pine needles contain several bioactive components such as α-pinene, caryophyllene, beta-pinene and bisbenzene, camphene, borneol, phellandrene, quercetin, kaempferol, and terpene, which have been reported to have antimicrobial, antimutagenic, and antioxidant effects [6,7,8]. Pine needles also contain calcium (28 mg/100 g) and various amino acids such as glutamic acid, phenylalanine, leucine, and lysine, as well as vitamins such as niacin, riboflavin, beta carotene, and thiamine [9,10]. Despite a rich source of bioactive compounds, some researchers indicated that high content of condensed tannins in pine needles might affect nutrient absorption by reducing protein digestibility in animals [11,12]. In this context, fermentation has been reported as an effective process to remove tannins and thereby improving the nutritional quality of pine needles [13].Fermented pine needle drink has been used as a functional beverage due to its various biological properties in Asia–Pacific regions especially in Taiwan, China, and Korea [13]. Self-fermented pine (P. densiflora) needle extract has been reported to contain phenolics, flavonoids, essential oils, and terpenoids, which might act as nourishing agents [13,14]. In an in vivo study, when self-fermented pine (P. densiflora) needle extract aged for 7 years (0.5 mL/day) was orally gavage in cholesterol-fed rats for 4 weeks, it reduced the blood plasma cholesterol and triglycerides [15]. In addition, the same extract (200 µg/mL) also inhibited the frequency and amplitude of pacemaker currents in interstitial cells of Cajal of the murine small intestine via ATP-sensitive potassium channels, which suggests the regulatory role of self-fermented pine needle extract in gastrointestinal motility [15]. Gastrointestinal motility is crucial for ensuring the proper transportation of ingested food and absorption of nutrients along the gut [16]. Chen et al. [17] demonstrated that fermented pine (P. morrisonicola) needle preparations (vinegar and alcohol) have better antioxidant activity than non-fermented products. Recently, Chiu et al. [13] demonstrated antioxidant and anti-inflammatory activities of ethyl acetate extract of fermented pine (P. morrisonicola) needles in lipopolysaccharide-treated RAW 264.7 macrophage cells. These effects were mediated through the modulation of NF-κB signaling pathway. The same authors also observed higher contents of phenolics and flavonoids in the fermented extract as compared with the non-fermented one, which might be due to various enzymatic reactions of microorganisms. Only one study in broilers has demonstrated that the fermented pine needle powder significantly improved the antioxidant status of birds [12]. However, literature investigating the effect of fermented pine needle extract (FPNE) in laying hens’ diet is very scarce. Therefore, the aim of this study is to evaluate the effect of dietary supplementation of FPNE on laying productivity, egg quality, serum lipid parameters, and antioxidant capacity of eggs.2. Materials and Methods2.1. Pine Needle FermentationPine needles used in this experiment were collected from Pinus densiflora, a native to Bonghwa, Gyeongbuk, Korea. The collected needles were washed, air-dried, and stored at 4 °C until fermentation. The pine needles were analyzed for their nutrient composition using AOAC procedures [18] (Table 1). For fermentation, the pine needles were chopped to 1 cm in size and mixed with water and sugar in a ratio of 1:1:0.6 on weight basis and then spontaneously fermented in an earthen pot (120 L) with the covered lid at room temperature for 6 months. The pine needles were then removed, and the remaining solution was aged in the same pot at room temperature for another 6 months. After aging, this solution was termed as a fermented pine needle extract (FPNE). The FPNE was then transferred to a stainless-steel beverage tank and stored at 4–8 °C until further use. 2.2. Characterization of FPNEThe pH of the FPNE was directly measured using a digital pH meter (iSTEC 735P, Daejeon, Korea). The total polyphenols, total flavonoids, and antioxidant activity of FPNE was measured. To measure the total phenol content, 0.1 mL of the extract was mixed with 2 mL 2% (w/v) Na2CO3 and vortexed for 3 min. Thereafter, 0.1 mL of 1N Folin–Ciocalteu reagent was added to the mixture and absorbance was measured at 700 nm after 30 min of incubation against a water blank. The total flavonoid content was measured using the aluminum chloride colorimetric method. Briefly, 0.5 mL of the extract was mixed with 1.5 mL 95% ethanol, followed by the addition of 0.1 mL of 10% (w/v) aluminum chloride, 0.1 mL of 1 M potassium acetate, and 2.8 mL of distilled water. After 30 min, the absorbance was measured at 415 nm against a water blank. The total polyphenol and flavonoid content were expressed as microgram of quercetin equivalent (QE) per ml of the extract. The antioxidant activity of FPNE was determined by measuring the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging ability. Briefly, 32 µL of the extract, 128 µL of distilled water, and 640 µL of 0.15 mM DPPH solution (Sigma Co., St. Louis, MO, USA) were added and allowed to react in the dark for 15 min at room temperature. The absorbance was measured at 515 nm against a water blank and the results were expressed as percent inhibition using the Equation (1):Radical scavenging activity (%) = (Ac − As)/Ac × 100,(1)Ac: Absorbance of control and As: Absorbance of sample.2.3. Experimental Animal and DesignA total of 108 Hy-Line brown laying hens (40 week of age) randomly allotted to three experimental groups with four replicates of nine birds each. Hens were housed nine per cage in a double-tier metal-cage system (90 cm wide, 90 cm high and 735 cm2 area) at a controlled temperature of 22 ± 2 °C. A corn-soybean meal basal diet was formulated to meet the nutritional requirement of laying hens [19] (Table 2).The amount of FPNE added to the basal diet was 0 mL/kg (CON), 2.5 mL/kg (T1), and 5 mL/kg (T2). The appropriate amount of FPNE was added to the basal diet and mixed for 5 min using a feed mixer (DKM350SU, Daekwang Co., Ltd., Hwaseong, Korea). The experimental period lasted for 6 weeks after 2 weekadaptation on the basal diet. Feed in mash form and water were provided ad libitum. The photoperiod was set at 16 h of light and 8 h of darkness (16L:8D) throughout the study.2.4. Laying PerformanceEggs from individual cages were weighed and recorded daily. Egg production was calculated as the rate of egg production (including normal and broken eggs) per hen per day. The average weight of the eggs was determined by measuring the weight of all the eggs laid daily. Egg mass was calculated by multiplying the egg production rate by average egg weight. Average daily feed intake was recorded weekly as the difference between the feed offered and the refusals during a week. Feed conversion ratio was calculated by dividing feed intake by egg mass.2.5. Egg Quality AnalysisEvery week, 3 eggs per replicate from each treatment group were randomly selected for analyses. Eggs were weighed individually, and eggshell strength was measured using an eggshell strength tester (Fugihira Industry Co., Ltd., Tokyo, Japan). Then the egg was cracked, and egg contents were placed on a glass plate, and the height of albumen was measured to calculate the Haugh unit (HU) as described by Haugh [20], using the formula: HU = 100 × log (H + 7.57 − 1.7 × W0.37), where H is the albumen height, and W is the egg weight. Color of fresh yolk was measured using the Roche yolk color fan (F. Hoffmann-La Roche Ltd., Basel, Switzerland). Eggshell color was measured by comparing with eggshell color fan (Eggshell color fan, Samyang, Korea). For eggshell thickness, the thickness of the eggshell was measured using a digital micrometer (Digimatic Micrometer, Series 293–330, Mitutoyo, Japan).2.6. Serum Lipid Parameters At the end of experiment, after fasting for 6 h, blood samples of eight laying hens per treatment was collected from extrinsic vein using a 5 mL sterilized syringe. Serum was collected after centrifugation and stored at −20 °C until analyses. The concentration of total cholesterol (TC) and triglycerides (TG) were measured by colorimetry using a biochemical analyzer (HITACHI 7600, Tokyo, Japan). High density lipoprotein-cholesterol (HDL-C) was measured using an HDL diagnostic kit (HDL-cholesterol kit, Youngdong Medical Corporation, Seoul, Korea). The sum of low density lipoprotein- and very-low density lipoprotein-cholesterol (LDL- + VLDL-cholesterol) was calculated by subtracting the HDL-C content from the TC.2.7. Egg Quality during StorageFour eggs were randomly selected from each replicate to observe the changes in HU during storage at 18 °C for 4 weeks. The thiobarbituric acid reactive substances (TBARS) were measured in egg yolk after 4 weeks of storage using the lipid peroxidation assay in egg yolk as described by Botsoglou et al. [21] with some modifications. In addition, five intact eggs from each treatment were stored at room temperature for 4 weeks and then egg yolk was removed and stored for 2 more weeks to study the oxidative changes in eggs. Of the collected yolk, 1.5 g was mixed with 5% aqueous trichloroacetic acid solution containing 0.8% butylated hydroxytoluene and homogenized at 4000 rpm for 5 min. After homogenization, it was reacted with thiobarbituric acid reagent and the mixture was incubated at 70 °C for 30 min. After cooling to room temperature, the absorbance of the mixture was measured at 532 nm against a blank reaction mixture. The concentration of TBARS was calculated using malondialdehyde (MDA) as a reference standard.2.8. Statistical AnalysisData were expressed as mean and the standard error of the mean (SEM). Individual cage was treated as the experimental unit for productivity- and egg quality-related parameter analyses. The experimental unit for blood parameters was an individual bird. The data of lipid oxidation in egg yolk during storage was analyzed by considering egg as an experimental unit. Orthogonal polynomial contrasts were employed to measure the linear and quadratic effects of additive dosage of FPNE using the SAS software (SAS Institute Inc., Cary, NC, USA). Differences were considered significant at p < 0.05. 3. Results3.1. Characterization of FPNEThe pH of FPNE was measured as 3.15 ± 0.02. Table 3 present the antioxidant activity and antioxidant compounds of FPNE. The contents of polyphenols and flavonoids in FPNE were determined as 50.1 ± 0.005 µg of QE/mL and 16.1 ± 0.010 µg QE/mL, respectively. The antioxidant capacity as determined by DPPH radical scavenging assay was found to be 66.9 ± 1.21%.3.2. Egg ProductivityTable 4 illustrates the effect of dietary FPNE supplementation on performance of the laying hens over 6 week period. Egg production rate was significantly increased at 43 week (linear, p < 0.05), 45 week (linear, p < 0.01), and during the overall period (43–48 week) (linear, p < 0.01 and quadratic, p < 0.05) with the increasing levels of FPNE in laying hens’ diet. Dietary FPNE significantly improved the feed intake linearly during 44–45 week (linear, p < 0.05) and the entire period (43–48 week) (linear, p < 0.05). In addition, egg mass was significantly increased in FPNE supplemented groups at 45 week (linear, p < 0.01) and for the overall period (linear, p < 0.05). However, supplementing the diet with FPNE did not affect the egg weight and FCR at any time of the experimental period (p > 0.05).3.3. Egg QualityTable 5 shows the effect of supplemental FPNE on egg quality traits in hens from 43 to 48 week. The egg quality data showed high variability at each week of experiment. For instance, HU was significantly decreased at 46 week (linear, p < 0.001 and quadratic, p < 0.01) by the dietary supplementation of FPNE. However, no significant effect was observed at 43, 44, 46, 47, and 48 week (p > 0.05). Eggshell color was significantly increased in FPNE supplemented groups at 44 week (quadratic, p < 0.01) and 45 week (linear, p < 0.01). However, no significant effects were observed in egg yolk color scores among the dietary treatments at each week of study (p > 0.05). The eggshell breaking strength was significantly improved in the FPNE fed groups as compared with the CON group at 45 week (quadratic, p < 0.05), 47 week (linear, p < 0.01), and 48 week (quadratic, p < 0.05). Moreover, eggshell thickness was significantly decreased at 45 week (linear, p < 0.05) while significantly increased at 48 week (quadratic, p < 0.05) by dietary FPNE. Considering the overall effect of FPNE supplementation for 6 week, eggshell color (linear, p < 0.01), egg yolk color (quadratic, p < 0.01), and eggshell breaking strength (linear, p < 0.05 and quadratic, p < 0.05) were significantly increased in FPNE-supplemented groups. However, there was no significant difference in eggshell thickness among the dietary groups (p > 0.05) during 43–48 week. Furthermore, dietary FPNE negatively affected the HU (quadratic, p < 0.05) during the entire experimental period.3.4. Blood Lipid ProfileData of serum lipid parameters at the end of the experiment (48 week of age) are presented in Table 6. The addition of FPNE to laying hens’ diet had no significant effect on serum lipid parameters (p > 0.05).3.5. Effect of FPNE Supplementation on Haugh Unit and Lipid Oxidation of Eggs during StorageThe effect of dietary FPNE supplementation on the egg quality during storage is presented in the Table 7. The HU decreased with the increasing storage period at 18 °C among all the treatments. Dietary FPNE did not affect the HU of eggs during storage when compared with the non-supplemented group. However, supplementation of laying hens’ diet with FPNE significantly decreased egg yolk MDA concentrations (linear, p < 0.001 and quadratic, p < 0.05) after 6 weeks of storage, as compared with the control group.4. DiscussionIn this study, we aimed to evaluate the potential effects of supplementing FPNE to layers’ diet on productivity performance, egg quality, serum lipid parameters, and antioxidant capacity of eggs. The addition of FPNE to laying hens’ diet increased the overall egg production and egg mass as compared with the control group. It could be possible that FPNE might have stimulated the hepatic secretion of egg yolk precursors through protecting hepatocytes from oxidative damage resulting in the enhancement of yolk formation and ovulation [22,23]. The increased egg mass in this study was related to the improved egg production rate in FPNE-fed hens. In contrast, recently, Moon et al. [24] indicated that FPNE (0.4 and 1.2 mL/kg of diet) as a component of PFA mixture had no effects on egg production and egg mass when averaged for 6 weeks. This variability in result might be related to the lower dose of FPNE or multiple components present in the feed additive mixture. In line with the present study, many studies of the inclusion of phytogenic preparations in layers’ diets indicated the dose-dependent increase in the feed intake in hens [24,25]. It is generally considered that PFA could improve the flavor and palatability of poultry feed, subsequently increasing the total feed intake. The observed improvements in egg production, egg mass, and feed intake with FPNE supplementation might be due to the presence of essential oils, terpenoids, and polyphenols which are reported to improve digestion, absorption, and utilization of nutrients in the digestive tract [3,5]. A dietary supplementation level of 0.5 mL/kg FPNE in the layers’ diet resulted in the best improvement in productive performance in terms of egg laying rate, egg mass, and feed intake.The current results indicate that FPNE can be used to enhance the intensity of the egg yolk and eggshell color as well as the breaking strength of the eggshell. In accordance with results from the present study, increased yolk pigmentation was found by several authors when supplementing PFA to laying hens’ diet [5,23]. The effect of FPNE supplementation in laying hen diets on egg yolk and eggshell color, as well as eggshell breaking strength, was not reported previously. The higher egg yolk color score in FPNE supplemented groups could be attributed to the antioxidant components of FPNE which might have reduced the lipid peroxidation in egg yolk. Moreno and Osorno [26] reported that antioxidant components are responsible for eggshell color; however, the effect of dietary additives was not considered by the authors. The higher eggshell breaking strength in the eggs from FPNE-fed hens could be explained by the study of Radwan Nadia et al. [27], who reported that better eggshell breaking strength in the phytogenic supplemented group might be due to the fact that natural antioxidant compounds promote uterine health and increase calcium absorption as well as improve digestibility of nutrients. However, the exact mechanism is not completely understood. HU is the measure of internal egg quality. Generally, eggs are graded according to their HU values: AA, ≥72; A, 71–60; and B, <60 [24]. Herein, HU in all the dietary treatments were above 80, suggesting that the eggs produced in this study are of good quality. The internal quality of eggs deteriorated as the storage time increased, especially at room temperature [28]. The dietary inclusion of FPNE reduced egg yolk MDA concentrations during storage for 6 weeks, suggesting its antioxidant action through reducing lipid peroxidation. A dietary supplementation level of 0.25 mL/kg FPNE in the layers’ diet quadratically improved some of the egg quality parameters such as egg yolk color, eggshell breaking strength, and egg yolk antioxidant capacity. Since different concentrations of FPNE resulted in the higher values for productivity and egg quality parameters, it is imperative to conduct additional studies for the optimization of dietary supplementation levels of FPNE in laying hens.Pine needles have been shown to promote lipid metabolism when supplemented in poultry diets [29,30]. Cholesterol and triglycerides are the important markers of lipid metabolism. Guo et al. [29] demonstrated the reduction of serum cholesterol and triglycerides in broilers by pine needle powder supplementation (10 and 50 g/kg of diet) via improving the antioxidant functions in birds. The cholesterol-reducing effects of FPNE have also been previously reported in rats fed high cholesterol diet [15]. However, in the present study, no effect of dietary FPNE on serum lipid metabolites was observed, which may be due to its lower dose or short-term supplementation.5. ConclusionsThe dietary supplementation of FPNE significantly improved the egg production, egg mass, and feed intake in a dose dependent manner. The egg quality parameters like egg yolk color, eggshell color, eggshell breaking strength, and egg yolk antioxidant capacity were also increased by FPNE. Further long-term studies are needed to ascertain the effect of dietary FPNE on antioxidant capacity in laying hens.

animals : an open access journal from mdpi

10.3390/ani11092581

PMC8468771

The presence of gastrointestinal parasites such as coccidia (protozoa), gastrointestinal nematodes, flukes, and tapeworms are a considerable problem in goat keeping. Parasitic infections cause a deterioration of animal health, delay in growth rate, weight loss, reduced milk production, and miscarriages. The aim of the study was to compare the prevalence and intensity of parasitic infections observed in the digestive tracts of goats kept on organic and conventional farms. Our findings indicate that conventional goat herds demonstrate a similar prevalence of parasitic diseases as organic herds. Nevertheless, the prophylactic programs used to combat parasitic infections in both types of farms appear ineffective and require improvement. There is a need for goat herds to be covered by ongoing parasitological monitoring, including parasitological testing before and after the pasture season, to detect carriers and shedders of parasite eggs, oocysts, and cysts. It is also recommended that keepers employ rotational or intensive rotational grazing methods and take care to ensure the hygiene of animal quarters and livestock rooms. Furthermore, accurate diagnosis of parasitic infections, as well as effective monitoring and prophylaxis, are essential for keeping goat herds free from parasitic infections.

The aim of the study was to compare the prevalence and intensity of gastrointestinal parasitic infections in goats kept on organic (n = 76) and conventional farms (n = 82). In general, a higher prevalence of some gastrointestinal parasitic infections was found in the conventional farms compared to the organic farms: the mean prevalence of Eimeria spp. was 85.4% in conventional farms and 77.6% in organic farms, that of Fasciola hepatica was 6.10% in conventional farms and 2.63% in organic farms, and that of Moniezia expansa was 31.7% and 17.1%, in conventional and organic farms, respectively. Both farm types demonstrated a similar mean prevalence of nematodes (80.3 vs. 84.2%). Conventional farms demonstrated a significantly higher intensity of infection with E. arloingi, Haemonchus spp., Nematodirus spp. and Moniezia expansa compared to organic farms. They also demonstrated a higher intensity of infection with Eimeria spp. than organic farms. The prophylactic programs used to combat parasitic infections in both types of farms appear ineffective and require improvement. There is a need for goat herds to be covered by ongoing parasitological monitoring. It is also recommended that keepers employ rotational or intensive rotational grazing methods and take care to ensure the hygiene of animal quarters and livestock rooms.

1. IntroductionRecent years have seen a growth in the number of organic farms, particularly since some countries have favourable conditions for their development. Many provide various economic incentives for organic production, such as financial support from various national and EU programs, growing demand and ease of marketability, lower production costs and higher product prices. Organic methods are also widely regarded as being more environmentally friendly and better for animal welfare. Food produced based on natural methods is seen as healthy and is appreciated by consumers [1,2].However, animals kept under organic conditions are more exposed to the risk of contracting infectious diseases, including parasitoses, with one reason being the limited use of antiparasitic drugs compared to conventional farms. In organic animal husbandry, disease prevention or control relies mainly on non-chemical methods. Although synthetic antiparasitic drugs can be used in organic farms (under the conditions regulated by the legal EU and national regulations [3]), parasites are usually controlled by authorised drugs based on phytotherapy [4].Data on the parasitosis of animals kept in organic farms is scarce, and those that are available show a high proportion of infected animals, often by several parasites [5,6]. The most common parasitoses noted in goats from organic farms are those of the gastrointestinal nematodes and Eimeria spp., while Moniezia spp. and Fasciola hepatica are recorded less frequently, and they are also major preoccupations for organic farmers [4,7].These groups of parasites include numerous pathogenic species capable of causing disease, e.g., Eimeria arloingi, Haemonchus contortus [8,9] or Trichostrongylus colubriformis, as well as less pathogenic species, including E. hirci, E. punctata, Nematodirus spp. [10]. The presence of these parasites is a considerable problem in goat keeping because they cause a deterioration of animal health, delay in growth rate, weight loss and miscarriages, as well as increased costs associated with production and veterinary treatment, and even death [11,12,13,14]. In addition, there are costs related to the decrease in animal productivity (weaker weight gain in meat breeds, a decrease in milk production in dairy breeds). Hoste and Chartier [15] noted that subclinical infection caused by H. contortus and T. colubriformis induced a decrease in body condition score and persistent decrease in milk yield, ranging from 2.5–10% in goats with the lowest performance and 12–25% in goats with the highest milk production. A significant decrease in the fat content of milk was also observed. In turn, Kyriánová et al. [5] reported that high intensity of strongylid infection contributed to a significant decrease (p < 0.01) in milk protein. These changes in the composition of milk may be reflected in its price and limit the technological suitability of goat milk [16]. According to Charlier et al. [17], the annual estimated costs of helminth infections in dairy goats in Europe was 67–107 million €.Unfortunately, parasitic diseases are often underestimated by farmers, and deworming is not always preceded by a faecal examination for parasites. In addition, vets often assume a goat to respond the same way as a sheep or a cow and use incorrect drug dosing. Goats have a more rapid metabolism of anthelmintics than sheep. Therefore, applying the same doses to goats as to sheep may promote more rapid selection for resistance in parasites. However, care must be taken to avoid poisoning when increasing the dose rate of anthelmintics not registered for goats [18,19,20,21]. Therefore, as the main site of infection for goats is the pasture, an important element in the protection of small ruminants against parasitic diseases in organic farms is correct management of grazing. Maintaining so-called safe pastures, i.e., those free from invasive parasites, can reduce or eliminate parasitoses in the herd [4].The aim of the study was to compare the prevalence and intensity of parasitic infections observed in the digestive tracts of goats kept on organic and conventional farms.2. Materials and MethodsThe faeces from goats from 2 organic farms and 4 conventional farms were subjected to coproscopy examination in September 2019. In total, 76 goats were studied from the organic farms and 82 goats from conventional farms, located in the West Pomerania region of NW Poland. Faecal samples (≈10 g) were collected from the rectum of animals and placed in labelled plastic bags. The samples were transported and refrigerated at 4 °C. Laboratory analyses were performed within 48 h.Both the goats on organic and conventional farms were kept on deep bedding. In the conventional farms, deworming was carried out twice a year: once in spring before pasture, and again in autumn after pasture (levamisole hydrochloride, 8%). In the organic farms, deworming was carried out when needed (after examination) using drugs with the same active substance as in conventional farms. The information on farms obtained from the herd owners is given in Table 1.Infection with gastrointestinal nematodes, Eimeria spp. and Moniezia expansa tapeworms was determined on the basis of coproscopic examinations using the Willis–Schlaaf flotation method [21]. The intensity of infection was determined quantitatively by the McMaster method [21]. Larval cultures were obtained from the isolated eggs, and these were used to identify the gastrointestinal nematodes to genus level [22,23]. Liver fluke eggs were detected by decantation [24].The species composition of coccidia in the studied goats was determined based on oocyst morphology (shape, colour, form index), the time of sporulation, as well as the presence or absence of a micropyle and its caps, residual bodies, polar bodies and Stied bodies. The oocysts were cultured in a moist chamber at a temperature of 24–26 °C, with a 2.5% aqueous solution of potassium dichromate (K2Cr2O7). Species composition was established based on keys provided by Chartier and Paraud [9].The results were statistically analysed with Statistica 13.3 (TIBCO Software Inc., Palo Alto, CA, USA). The χ2 test was used to compare the prevalence between particular species of parasites on organic and conventional farms, while the intensity of infection was examined using the Mann–Whitney U-test. Differences were determined to be statistically significant at p < 0.05. The confidence interval of the proportion was calculated by the modified Wald method, as recommended by Agresti and Coull [25].3. ResultsIn general, the mean prevalence of gastrointestinal tract parasitic infections was comparable (p = 0.658) between conventional farms and organic farms (Table 2).The mean prevalence of Eimeria spp. was 85.4% in conventional farms and 77.6% in organic farms (Table 3). Conventional farms demonstrated a significantly higher prevalence of infection with E. arloingi (χ2 = 5.16; p = 0.023), compared to organic farms. For other species, the difference was not statistically significant. Conventional farms also demonstrated a significantly higher intensity of infection (Z = −2.15; p = 0.031) compared to organic farms. However, statistically significant differences were noted only for 1 species—E. jolchijevi (Z = −2.88; p = 0.004).Many animals were infected with several coccidia species. Single and multi-species Eimeria infections in goats on organic farms are presented in Table 4. Mixed infections were found in 73.7% of studied goats. The most common mixed infections were those comprising three (19.77%) and four (27.68%) species of Eimeria. Among the three-species co-infections, the presence of E. caprina (n = 8) and E. chrisienseni (n = 8) was most frequently observed. In the case of four-species co-infections, E. alijevi was most common (n = 11). Single-species infections were found in 13 goats, and the most widespread was E. arloingi (n = 5).Single and multi-species infections in goats on conventional farms are presented in Table 5. Mixed infections were found in 81.7% of studied goats. The most common mixed infections were those comprising three (29.27%) and four (29.76%) species of Eimeria. Among the three-species co-infections, the presence of E. arloingi was most frequently observed (n = 18). In the case of four-species co-infections, E. ninakohlyakimovae was most common (n = 13). Six- species co-infections have not been recorded.The mean prevalence of infection with nematodes was comparable (χ2 = 0.41, p = 0.523) in both types of farms, but intensity of infection was significantly higher in conventional farms than organic farms (Z = −5.71; p < 0.001) (Table 6). Significantly higher prevalence in conventional farms was observed only in the case of Nematodirus spp. (χ2 = 8.64, p = 0.003) and Haemonchus spp. (χ2 = 10.32, p = 0.001).Compared to those on the organic farms, the goats on conventional farms were also found to display a significantly higher intensity of infection with Trichostrongylus spp. (Z = −4.15; p < 0.001), Oesophagostomum spp. (Z = −4.04; p < 0.001), Cooperia spp. (Z = −4.22; p < 0.001), Haemonchus spp. (Z = −2.48; p = 0.013), Nematodirus spp. (Z = −2.74; p = 0.006).Conventional farms demonstrated a significantly higher prevalence of infection with Moniezia expansa (χ2 = 4.52; p = 0.033) compared to organic farms. In the case of Fasciola hepatica, higher prevalence was observed in conventional farms than in organic farms (6.10% vs. 2.60%), but the difference was not significant (p = 0.290).The occurrence of particular species of helminths in single and multi-species infections in goats on organic and conventional farms are presented in Table 7. Of these, five-species co-infections dominated in organic farms, and four-species co-infections in conventional farms.4. DiscussionThe level of infection in goats is influenced by a range of environmental factors that favour the development of the parasitic stages outside the host or limit their survival. On farms, these factors largely depend on the management strategies intended to prevent and control endoparasitic diseases in goat farming systems [26]. Rahmann and Seip [26] reported a greater spread of parasites when goats were kept on deep bedding and were maintained in an alcove-pasture rearing system. In contrast, a lower prevalence of infection was observed among goats maintained in an alcove system, with daily manure removal, as well as during winter, which may be due to the reduced chance of contact between the host and the parasites. The prevalence and intensity of infection were also influenced by the size of the herd, stocking density and choice of prophylaxis programs. Our present findings indicating high prevalence of gastrointestinal parasite infection are consistent with those of previous studies [27,28,29]. In goat breeding, a serious problem is presented by parasitosis caused by gastrointestinal nematodes and Eimeria protozoa, as well as by flukes such as Fasciola hepatica and Moniezia tapeworms [30,31,32,33,34,35,36].Although Eimeria spp. generally occur worldwide in goats, no geographical predisposition has been observed for any particular species. Any observed diversity in the prevalence and distribution of coccidiosis is influenced more by the hygiene and temperature in the farm, as well as microclimate, host resistance and susceptibility of the breed to coccidia [37]. Our present findings indicate that Eimeria species are widespread in the studied herds. The mean prevalence of Eimeria spp. in conventional and organic farms was 85.4% and 77.6%, respectively. These results are similar to those noted in Iran (83.4%) [31], India (79.2%) [32] and in Turkey (73.6%) [38]. A higher prevalence has been reported in Portugal (98.61%) [39], Florida, USA (97%) [40], China (92.9%) [41] and Spain (96.1%) [42]. However, all the cited works concerned research conducted on conventional farms. Unfortunately, only a small number of papers are available on the differences in the prevalence of coccidiosis between organic and conventional animal husbandry methods. However, similar studies have been carried out in herds of cows in West Pomerania (Poland) [6], as in the present studies, the results indicate no significant differences in the prevalence of Eimeria spp. between both herd types. In the present study, the presence of a relatively high prevalence of Eimeria spp. in both farm types probably results from the contamination of the local habitat with Eimeria oocysts. Höglund et al. [43] recommend that gastrointestinal parasites can be controlled by good management, such as the use of parasite-safe pastures. It is possible that the studied farms were not employing enough effective grazing management strategies. Additionally, due to the lack of clinical symptoms in goats, no coccidiostats were used in any of the farms. This could result in a number of oocyst seeders being present in the herd, which would act as a source of infection for healthy animals.For most Eimeria species, no significant difference in prevalence was observed between the organic and conventional farms. The exception was E. arloingi, which demonstrated a significantly (p = 0.023) higher prevalence in conventional farms than organic farms. This difference is difficult to explain. E. arloingi is one of the most pathogenic Eimeria species to goats [9]. It may be the case that the animals possessed greater immune resistance or that the studied conventional farms may have had a previous history of coccidiosis. As indicated by Silva et al. [39], protective immune responses against Eimeria infections are inhibited, among others, by stress, which can be caused by numerous factors, such as herd size, diet changes, weather conditions or nutritional status. Thus, the lower standard of welfare of goats on conventional farms could have made them more susceptible to infection than those on the organic farms.In goats, the most pathogenic species of Eimeria are believed to be E. arloingi, E. ninakohlyokimovae, E. caprovina, E. christenseni, E. faurei and E. gilruthi [9,37,44,45,46]. It is worth noting that of these, the most pathogenic species, viz. E. ninakohlyakimovae, E. arloingi and E. christenseni, were quite commonly observed in the studied herds (Table 3). In general, seven species were identified in faecal samples from both organic and conventional farms: E. arloingi, E. ninakohlyakimovae, E. caprina, E. alijevi, E. jolchijevi, E. hirci and E. chrisienseni. Değer et al. [38] reported nine different species of Eimeria in goats from Turkey, these being E. arloingi, E. christensini, E. alijevi, E. hirci, E. ninakohliyakimovae and E. jolchijevi, in addition to E. pallida E. apsheronica and E. punctata. Nine species of Eimeria were also identified by Kahan and Greiner [40] in goats from Florida, USA. This group included both E. punctata and E. caprovina. In turn goats from Egypt were found to be infected with E. ninakohlyakimovae, E. hirci, E. caprina, E. christenseni, E. jolchijevi, E. apsheronica and E. arloingi [47].The most prevalent Eimeria species on the organic farms were found to be E. chrisienseni (40.8%), E. alijevi (40.8%) and E. arloingi (36.8%), while the most common on the conventional farms were E. arloingi (54.9%), E. ninakohlyakimovae (48.8%) and E. chrisienseni (41.5%). In previous studies conducted in this area, E. arloingi, E. alijevi, E. ninakohliyakimovae and E. chrisienseni were the most common species in goats [48]. Our results are also similar to those of Değer et al. [38]. In turn, Mohamaden et al. [47] noted that the most prevalent species in goats from Egypt were E. arloingi (37.04%), E.ninakohlyakimovae (30.86%) and E. hirci (24.69%).Epizootiological studies have found mixed coccidial infections played a significant role in goat health around the world [9,31,32,49]. Indeed, in our study, the most common mixed infections were those comprising three and four species of coccidia. Similarly, Değer et al. [38] reported that multiple infections with three (19.4%) or four species (17.4%) were not rare. In addition, Değer et al. [38] also reported the presence of mixed infections in 66.9% of examined goats. This is a lower figure than in the present study, where 73.7% goats from the organic farms demonstrated mixed infection and 81.7% conventional farms.As reported by Mohamaden et al. [47] 700 oocysts/g indicated a subclinical infection of goats. The number of oocysts ranged from 1000 to 1 × 106 oocysts/g faeces in the faeces of asymptomatic animals and from 100 to 10 × 106 oocysts/g faeces in symptomatic animals [50]. In our study, the mean intensity of infection with Eimeria spp. was higher than 1000 opg, but no clinical signs of coccidiosis were observed in either the organic or the conventional farms. This may be due to the fact that adult animals are protected by the cellular immune responses induced by primary Eimeria infections [51]. This can lead to the development of enzootic stability between host and parasite and non-clinical status in goats.A significantly higher intensity of infection with Eimeria spp. was observed in goats from conventional farms than those on organic farms (p = 0.031). Chartier and Paraud [9] indicated that breeding intensification, high stocking rates, poor hygiene and physiological and nutritional stress all represented risk factors for high excretion. These factors were not observed to a high degree in the present farms. A greater intensity of infection was noted in farms characterised by higher breeding and stocking rates. In addition, Ruiz et al. [42] found oocyst shedding intensity to be related to herd size.The dominant endoparasites in goats are gastrointestinal nematodes, mainly Haemonchus contortus, Teladorsagia circumcincta, Trichonstrongylus spp., all of which also demonstrated high pathogenicity [34]. A range of gastrointestinal nematodes were also identified in the herds in the present study, including Chabertia ovina, Trichostrongylus spp., Oesophagostomum spp., Cooperia spp., Haemonchus spp., Nematodirus spp. and Strongyloides spp. (Table 6). In organic farms from the Czech Republic, the most prevalent nematodes were H. contortus (42%), Trichostrongylus spp. (23%), Oesophagostomum columbianum (13%), and Teladorsagia circumcincta (11%) [5].In conventional herds, nematode infections can be controlled by anthelmintic treatments, but this is prohibited, or at least limited, in organic production. However, excessive use of drugs from a single chemical group can result in the parasites developing resistance to the active substances and the consequent failure of treatment [22,52]. In addition, the use of inadequate doses (e.g., for sheep) may also result in the emergence of drug resistance in the nematode population due to its faster breakdown and elimination [53]. This is probably the cause of the significantly (p < 0.05) higher prevalence (Haemonchus spp., Nematodisrus spp.) and intensity of infection (Trichostrongylus spp., Oesophagostomum spp., Cooperia spp., Haemonchus spp, Nematodirus spp.) of some of the studied nematodes in the goats from conventional farms. The conventional farm owners reported that goats were treated with levamisole twice a year without prior testing. In contrast, the goats on organic farms were only treated in case of severe infection. The available literature indicates that levamisole has low efficacy against nematode in goats, ranging from 43.4 to 52.6% in the period 10–60 days after administration [52]. Generally, this could be one of the reasons for the high prevalence of gastrointestinal nematodes infections, which was above 80% in both types of farms. This prevalence was higher than that reported by Dey et al. [54] in Bangladesh (62.1%), Zvinorova et al. [55] (31.0–40.0%) in Zimbabwe and Jegede et al. [56] in Nigeria (37.5%). In contrast, a very low prevalence of nematodes has been reported in organic goat herds in Greece (7.4%) [57]. Authors [57] reported the presence of strongyle-type eggs in 3.4% samples, Nematodirus spp. eggs in 1.1% samples and Trichuris spp. eggs in 2.9% samples.In the present study, the dominant nematode was found to be Haemonchus spp. in the conventional farms (59.8%) and Trichostrongylus spp. in the organic farms (56.6%). While no significant differences in prevalence were observed between the two types of farms, in the case of Trichostrongylus spp., significant differences (p = 0.001) were observed for Haemonchus spp. This situation can be attributed to some goat breeds possessing a genetic resistance to nematode infections. Comparative studies found that Boer goats, the breed kept on organic farms, demonstrated the highest expression of the DRB1∗1101 gene after exposure to Haemonchus contortus [58].Animals are typically infected with several species of nematodes at the same time. The prevalence of mixed infection in this study was 73.7% on organic forms and 81.7% on conventional ones. Elsewhere, 46% of goats were found to have mixed infections in Argentina [59], while only 6.25% of goats from Ethiopia demonstrated mixed infection by gastrointestinal nematodes [60].In organic farms, much attention was paid to grazing management [27], the purpose of which is to prevent and control endoparasite diseases. The farms included in the present study employed strategies based on shortening the grazing period and sowing plants with antiparasitic properties: A number of plants, such as garlic (Allium sativum), mugwort (Artemisia absinthium), black walnut (Juglans nigra), mugwort (Artemisia vulgaris) and common thyme (Thymus vulgaris), are known to demonstrate such properties [61,62,63]. In addition, a restricted living environment, such as a farm, is associated with a higher risk of parasitic infections [64]. Therefore, it is possible that the goats kept on organic farms demonstrated a lower intensity of infection with gastrointestinal nematodes due to the fact that lower numbers of animals were kept in a given area and that they may have been grazing on plants exhibiting antiparasitic properties. In addition, Saanen goats were kept on the organic farm. This breed is known to carry three single-nucleotide polymorphisms (SNPs) pertaining to four candidate genes of the cytokine family (IL2, IL4, IL13, and IFNG), which may be associated with its greater resistance to gastrointestinal endoparasitic infections [58].In the present study, the only tapeworm found on both the organic and conventional farms was Moniezia expansa. Similar results in goats were obtained by Górski et al. [35]. However, they report a significantly lower prevalence (2–2.6%) compared to our present findings (17.1% and 31.7%, Table 6). It is possible that the higher prevalence of Moniezia spp. observed in the present study was due to the samples being collected in autumn (September): the prevalence of Moniezia spp. is conditioned by the periods of activity of oribatid mites, their intermediate hosts, which is greater in summer or autumn [65,66].Moniezia spp. was noted in 31% of goats from Nigeria [67], 12% in Ethiopia [68], and 2.64% in Slovakia [69]. Fagbemi and Dipeolu [70] reported that the prevalence of Moniezia spp. was generally low in animals kept on an extensive system of management with low stocking rates. This could also explain the significantly (p = 0.033) lower prevalence of this tapeworm in goats from organic farms.In the present study, significant differences in the prevalence of Moniezia spp. were found between the two types of farms, but both demonstrated similar intensity of infection (150 and 165 EPG). A significantly higher intensity of infection was noted in Slovakia (320 EPG) [69].Fagbemi and Dipeolu [70] note that low pathogenicity of Moniezia spp. was usually associated with low-grade infections. However, this pathogenicity can increase if an animal is coinfected with other parasites. For example, mixed infection of Moniezia and Trichuris can result in severe malnutrition, leading to pulmonary oedema and the death of the animal [71]. In the organic farms, Moniezia spp. was present in mixed infections of three to eight species, while in the conventional farms, it was observed as both single infections and mixed infections containing from two to eight species of the parasite (Table 7). However, no clinical signs were observed in animals, possibly due to the low intensity of infection.The liver fluke Fasciola hepatica, the causative agent of fasciolosis, is associated with wetlands or periodically flooded areas. Such areas are a habitat for intermediate hosts such as the snail Galba truncatula. The course and symptoms of the disease depend on the age of the hosts, their nutritional status and the intensity of infection [72,73]. Our findings indicate that the occurrence of F. hepatica in goats is rare in Poland, as also indicated by Górski et al. [35]. This fluke was also found sporadically in Greece. According to Kantzoura et al. [57], the prevalence was 0.5% in sheep from organic farms and from 0 to 2.5% in those from conventional farms, depending on the region.Fluke infection is usually diagnosed by coproscopic examination based on decantation, mainly due to its simplicity. However, this method is only effective three to four months after infection, when sexually mature flukes start producing eggs. In addition, as egg production by flukes is not constant, the method has relatively low sensitivity [74]. As a result, the true prevalence of F. hepatica may be underestimated. Even so, it must be remembered that the appearance of liver flukes in a herd is a significant problem for farmers. Fluke infection may increase the susceptibility to infection with other parasites. For example, Cuervo et al. [59] noted significant (p < 0.05) positive associations between F. hepatica and Strongyle eggs. The authors suggest that infection by F. hepatica may act as a contributing factor for Strongyle infection, increasing the chance almost twofold (1.96 and 1.83, respectively), probably due to the host immunosuppression caused by the trematode, or possibly due to the weakened state of the host.In this study, no significant differences in prevalence of F. hepatica were observed between two types of farms. However, this may be due to the small number of infected animals: two in the organic farms and five in the conventional farms. In contrast, a higher prevalence of F. hepatica was observed in cows from organic farms than those from conventional farms in Denmark [75]. However, the authors [75] note that the conventional farms implemented F. hepatica control strategies based on grazing management, resulting in a decreased level of infection. Olsen et al. [76] suggest that the higher prevalence of F. hepatica in organic animals may arise from greater access to pasture or to lower treatment levels in organic herds.5. ConclusionsIn general, our findings indicate that goats in conventional herds demonstrate a higher prevalence of parasitic diseases than organic herds. Only E. arloingi, Haemonchus spp. Nematodirus spp., Moniezia spp. and Fasciola hepatica demonstrated a greater prevalence in conventional farms. The prophylactic programs used to combat parasitic infections in both types of farms appear ineffective and require improvement. There is a need for goat herds to be covered by ongoing parasitological monitoring, including parasitological testing before and after the pasture season, to detect carriers and shedders of parasite eggs. It is also recommended that keepers employ rotational or intensive rotational grazing methods and take care to ensure the hygiene of animal quarters and livestock rooms. Furthermore, accurate diagnosis of parasitic infections, as well as effective monitoring and prophylaxis, are essential for keeping goat herds free from parasites.

animals : an open access journal from mdpi

10.3390/ani11082245

PMC8388394

In anaesthetized horses, blood oxygenation impairment often occurs. This systematic review compared the effects of low and high inspired oxygen fractions on the arterial oxygen tension and other pulmonary gas exchange parameters in horses during inhalation anaesthesia. Five studies, four experimental and one clinical, were deemed suitable for inclusion. A meta-analysis was performed on the four experimental studies. The oxygen partial pressure was significantly lower with a lower inspired oxygen fraction. However, indices of pulmonary gas exchange were significantly worsened. It is concluded that, while only a limited number of studies are available, the use of a higher inspired oxygen fraction in horses during inhalation anaesthesia will result in higher levels of oxygen in the blood; it will also worsen the lung gas exchange status. Further studies are needed to increase the level of evidence on this subject.

In anaesthetized horses, pronounced ventilation/perfusion mismatching often occurs. Several authors have investigated the effect of lower inspired oxygen fractions (FiO2) to reduce formation of absorption atelectasis. This systematic review compared the effects of low (<0.6) and high (>0.8) FiO2 on the arterial oxygen tension (PaO2), the alveolar-to-arterial oxygen tension difference (P(A-a)O2), and the PaO2/FiO2 ratio in horses during inhalation anaesthesia. Using the Systematic Review Protocol for Animal Intervention Studies, four experimental and one clinical investigations were deemed suitable for inclusion. A meta-analysis was performed on the four experimental studies. The PaO2 was significantly lower (p = 0.0007, mean difference −23.54 kPa, 95% CI −37.18, −9.90) with a lower FiO2. However, the P(A-a)O2 was also significantly lower (p < 0.00001, mean difference −20.80 kPa, 95% CI −26.28, −15.32) when using a low FiO2. For the PaO2/FiO2 ratio, only one study fitted the inclusion criteria, so no meta-analysis was performed. It is concluded that, while only a limited number of studies are available, the use of a higher FiO2 in horses during inhalation anaesthesia will result in higher levels of PaO2, but also a larger P(A-a)O2 difference. Further studies are needed to increase the level of evidence on this subject.

1. IntroductionSince almost sixty years ago, studies in humans have shown that during halothane anaesthesia with spontaneous breathing, ventilation/perfusion relationship (V/Q) abnormalities and intrapulmonary shunt may develop, resulting in a reduction in the arterial partial pressure of oxygen (PaO2). One of the main causes of impaired oxygenation of the blood seems to be the development of atelectasis during anaesthesia, which reduces lung compliance and PaO2. There is supporting evidence that this condition may develop in humans, horses, and other animal species [1]. This was confirmed in computed tomography (CT) studies in humans, which revealed atelectasis of the most dependent parts of the lungs in 90% of the anaesthetised patients [2,3]. Development of atelectasis is considered to happen immediately after induction of anaesthesia [4,5]. Lung compression, gas absorption, and surfactant impairment are the major causative factors for atelectasis development [2].Because of these reasons, it has been assumed that an alveolar partial pressure of oxygen (PaO2) of at least 26.6 kPa (200 mmHg) is needed in order to preserve normal PaO2 values [6,7,8]. However, higher levels of the fraction of inspired oxygen (FiO2) will increase the rate of gas absorption from partially or completely occluded alveoli and produce atelectasis [2,6]. In fact, the use of FiO2 1.0 may even be the major causative factor for atelectasis development [4,9,10,11,12], since the composition of the inspired gas is directly related to the rate of the alveolar collapse of a completely closed lung unit.In animals, lower PaO2 values have also been recorded in patients intraoperatively than in conscious subjects breathing the same FiO2. As in humans, a major factor for this is V/Q alterations [1]. In horses, pronounced V/Q abnormalities are commonly found, mainly caused by atelectasis formation in the dependent lung regions [13], a condition first published in an original and seminal paper on the subject, wherein it was shown that under halothane anaesthesia, a severe reduction in pulmonary ventilation may develop in horses [14]. However, the use of FiO2 1.0 to compensate for the atelectatic areas [15] may itself result in severe pulmonary atelectasis, creating a controversy among clinicians regarding the optimal FiO2.Evidence from studies in animals (dog [16], cat [17], sheep [18], and horse [19,20,21]) indicate that the use of a low FiO2 may be beneficial in reducing lung atelectasis. On the other hand, there is evidence that FiO2 0.3 [22] or 0.5 [23] does not improve arterial oxygenation or gas exchange compared to FiO2 above 0.9. Some studies evaluate aeration based on the CT images of the lungs for atelectasis formation, while others investigate the oxygenation status of the animals.Since there is lack of supporting evidence for the best FiO2 values intraoperatively in horses, we conducted this systematic review. The objective of our review was to systematically identify, appraise, and synthesise the evidence in relation to different FiO2 levels (high or low) in horses anaesthetised with inhalant anaesthetics. Specifically, our PICO question was: “Does a reduced FiO2 (below 0.6) compared to FiO2 above 0.8 improve blood oxygenation in horses during anaesthesia?”.2. MethodsA study protocol was established using the Systematic Review Protocol for Animal Intervention Studies (SYRCLE) [24].2.1. Type of StudiesWe included controlled studies on either experimental or client-owned animals, which compared at least two different FiO2 values during inhalant anaesthesia in horses. Reviews were excluded. Only publications in the English language were evaluated.2.2. Population/Species StudiedThe target species was the horse, of all ages. Only normocapnic patients were included.2.3. InterventionsFor the purpose of this review, a standard (control) treatment was defined as an FiO2 more than 0.8 and the intervention/exposure treatment as an FiO2 below 0.6. The mixture of inspired gas should contain medical air or nitrogen, but not other gases.2.4. Outcome MeasuresArterial partial pressure of oxygen (PaO2). All values were transformed to kPa.Alveolar–arterial difference in the partial pressure of oxygen (P(a-a)O2). All values were transformed to kPa.Arterial partial pressure of the oxygen to fraction of inspired oxygen ratio (PaO2/FiO2).2.5. Search MethodFour electronic databases were searched:MEDLINE via PubMed;Web of Science/CAB Abstracts;SCOPUS.The search string was: (“oxygen”) AND (“oxygenation” OR “atelectasis” OR “gas exchange” OR “oxygen tension” OR “pressure of oxygen” OR “oxygen partial”) AND (equine* OR horse*) AND (anaest* OR anest*) This string was adapted according to the search rules/code of the database used. All dates of publication were searched until end of May 2021.2.6. Selection of StudiesTwo groups, with two persons each (I.S. and K.P., and C.B. and S.S.), screened the results of the search output. Discrepancies were resolved with collaboration and critical discussion between the two groups. The first selection phase consisted of the evaluation of the title and abstract of the studies. The studies selected in this phase passed onto the second phase—the critical reading of the full paper. Whenever the authors of this review were also authors of an eligible study or had been a reviewer thereof, they were excluded from the evaluation.2.7. Data Extraction and ManagementDetails of the eligible studies were independently extracted by the two groups of reviewers. Data extracted were:Authors, title, year of publication, and journal;Number of animals in intervention and control groups;Horses, age, weight, status ASA, inhalant agent, and spontaneous/mechanical ventilation;Outcome measures;Presence of any other outcome measures;Excluded animals (dropouts).2.8. Assessment of Risk of Bias in the Included StudiesThe two groups of reviewers assessed the included studies using the SYRCLE’s Risk of Bias tool [25]. The following details were agreed on: When the study was randomised, but there was no mention of the randomisation method, we judged the study to have an unclear risk of bias. Random housing of the experimental animals was judged as low or unclear, as well as the animal assessors and animal selection blindness, because we assumed that these were mostly irrelevant to our review.2.9. Data AnalysisData were introduced into a specific software (Review Manager/RevMan Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014), where they were stored, analysed, and synthesised for the production of the meta-analysis. All three outcomes were continuous variables and were analysed with the inverse variance method with a random effects model. No subgroup analysis was performed (e.g., inhalant agent, and recumbency), because of the small number of the included studies. Effect measures are presented as the mean differences. Heterogeneity and overall effects were calculated. Statistical significance was set to α = 0.05.3. ResultsA total of 448 papers were retrieved. A PubMed search returned 135 results, Scopus 158, and Web of Science/CAB Abstracts 155. After removing the duplicates, 302 papers remained. The first selection phase revealed 19 papers eligible for further evaluation. The second selection phase revealed 5 papers, which were included in this review, and 14 papers were excluded (Figure 1).3.1. Characteristics of the Included StudiesFour studies were experimental, and one study was a clinical trial. In the experimental studies, isoflurane was used to maintain anaesthesia in three studies (two with mechanical ventilation [23,26] and one with spontaneous ventilation [27]), and halothane in one (spontaneous ventilation [19]). In the clinical trial, isoflurane was used for the maintenance of anaesthesia with mechanical ventilation [28] (Table 1).All the experimental studies had a crossover design, with 5–8 horses included in each one. Three were adequately randomised, but one had an unclear risk of randomisation bias [19]. The clinical study was a prospective randomised one. Outcome measures were serially (over time of anaesthesia) recorded in all studies. We collected and analysed data for the meta-analysis at the 90 min timepoint from all the experimental studies, which was the commonest timepoint among the studies. In the clinical study [28], data had been recorded at three timepoints during anaesthesia, and the pooled data are presented in the paper; therefore, the data were extracted for the qualitative analysis, but were not used for the meta-analysis.In all studies, the standard treatment was an FiO2 above 0.85. Intervention treatment was an FiO2 of 0.5 in two experimental studies. In the other three studies, the intervention groups received an FiO2 of 0.21 (experimental study) and 0.3 (one experimental and one clinical).In only one study [28] an a priori power analysis was performed, however the post hoc power of the study was poor.3.2. Risk of Bias of the Included StudiesThe risk of bias was found to be unclear or low in most of the studies. The risk of bias tables are shown in Figure 2 and Figure 3.3.3. Characteristics of the Excluded StudiesFourteen studies were excluded after critically evaluating them: one review [29]; five due to missing control/intervention allocation of the animals [30,31,32,33,34]; two studies [20,35] because injectable agents were used for the maintenance of anaesthesia; one with an FIO2 of the interventional group above our pre-defined threshold of 0.6 [36]; and one retrospective study with an average FIO2 of the control group below our threshold of 0.8 [37]. Furthermore, three more studies were excluded as other gases than nitrogen were used to decrease the FIO2: in two studies [21,38] the intervention group received a mixture of oxygen with helium, and in one study [39] the inspired mixture consisted of oxygen and nitrous oxide. Finally, one study [40] had an unorthodox study design: 24 animals were used to compare the influence of a delivered oxygen fraction (FdO2) of 1.0 and 0.6 during isoflurane anaesthesia. Sixteen horses underwent an arthroscopy in dorsal recumbency and were equally and randomly allocated over the two treatments, while the remaining eight horses received both treatments in a randomized crossover study in lateral recumbency for a wound healing study. The main reason for exclusion of this study is that the authors targeted a fixed FdO2, which may have resulted in some variability of the FiO2 among individual horses, although a similar influence would be expected with both treatments.3.4. The Effect of Low FiO2 on PaO2Data for the FiO2 were extracted from all five studies; in all of them, a low FiO2 statistically significantly reduced the PaO2 of the horses. Four studies were included in the meta-analysis (Figure 4). Data from a total of 24 animals were analysed. The heterogeneity of the studies was statistically significant (p < 0.00001, I2 = 99%) and the overall effect was statistically significant (p = 0.0007, mean difference = −23.54, 95% CI −37.18, −9.90), in favour of the high FiO2.Because of the high heterogeneity, a sensitivity analysis was also performed, by removing each study from the model. By removing the studies of Portier et al. (2009), Hubbell et al. (2011), or Crumley et al. (2013), there was a minor change in the I2 and the p-values. However, when removing the Cuvelliez et al. (1990) study, the I2 was reduced to 78% (still significant, p = 0.01), with an overall effect again statistically significant (p < 0.00001, mean difference = −18.58, 95% CI −24.26, −12.91), in favour of the high FiO2 (19 animals in the model). It seems that the Cuvelliez et al. (1990) study is the major source of heterogeneity; as can be seen, it has the largest mean difference between the two groups.3.5. The Effect of Low FiO2 on P(a-a)O2Data of P(a-a)O2 measurements were found in four studies. In one study [28], there was a statistically non-significant difference between the control and the intervention groups, regarding the P(a-a)O2, but these data were not included in the meta-analysis, because they were pooled out of several timepoints. In the remaining three studies, the low FiO2 statistically significantly reduced the P(a-a)O2 (Figure 5). Data from a total of 18 animals were analysed. The heterogeneity of the studies was statistically non-significant (p < 0.1, I2 = 57%) and the overall effect was statistically significant (p < 0.00001, mean difference = −20.80, 95% CI (−26.28, −15.32)), in favour of the low FiO2.3.6. The Effect of Low FiO2 on PaO2/FiO2Data on PaO2/FiO2 measurements were found in two studies. In one study [28], there was a non-significant difference in PaO2/FiO2; however, these data were not used for the meta-analysis, because they were pooled out of several timepoints. In the other study [23], the low FIO2 statistically significantly reduced the PaO2/FIO2. No meta-analysis was produced for this outcome.4. DiscussionAccording to our knowledge, this is the first systematic review with a meta-analysis to evaluate the effect of the inspired oxygen fraction on the oxygenation of the blood in anaesthetised horses. This review combined the results of five studies, four experimental and one clinical (not used in the meta-analysis), including a total of 64 horses (24 in the meta-analysis). It is interesting that very few clinical studies were found, although we believe that the extent of the use of low FiO2 mixtures in clinical practice is higher than depicted in the published literature; an indication is given in a retrospective study [37]. In an attempt to have more meaningful results, we excluded studies with injectable agents, because the effect of the different injectable anaesthetic drugs on the respiratory system and the pulmonary function is diverse.Moreover, we excluded studies where the inspired gas mixture consisted of gas other than oxygen with medical air or nitrogen. For instance, we excluded two studies with helium in the inspired mixture [21,38], because of the specific physical properties of helium. Interestingly, in both studies, PaO2 and P(a-a)O2 were higher (P(a-a)O2 (non-significantly in [38]) with a high FiO2, whereas the PaO2/FiO2 ratio was lower. Moreover, we excluded another study [39] with N2O in the mixture, an agent with special anaesthetic and analgesic properties, which may have affected pulmonary function. In this study, PaO2 was also higher in high FiO2. Although we have not included these three studies in our review, we assume that this has not substantially affected our results.A variety of factors, e.g., breed, gender, age, type of ventilation, application of positive end-expiratory pressure, tidal volume used, etc. [1], as well as body position, body weight, and thoracic conformation [41,42] may affect the oxygenation of the blood. Because of the diversity of the equine population recruited in these studies, we assumed that the heterogeneity of the studies was high, and it was treated as such in the meta-analysis (random effect analysis was used). That was also the reason why a subgroup analysis was not performed.As an appraisal method to estimate bias, we used the SYRCLE’s risk of bias tool for animal studies, which is an adapted version of the Cochrane risk of bias tool. Although it is not fully validated, it takes into account the specific aspects of the experimental design of animal studies compared to clinical studies [25]. In our review, we have included experimental as well as randomised clinical animal studies, so we believe that the choice of the SYRCLE’s risk of bias tool was the best available option.From this review, it is clear that the arterial partial pressure of the oxygen is higher when a high oxygen fraction is inhaled, which is an expected result, especially when mechanical ventilation is applied. However, this is not the only index of oxygen exchange. The horse may not be hypoxic, although severe intrapulmonary V/Q mismatch may develop intraoperatively, and lead to increased mortality post-operatively [15]. Furthermore, in practice, as long as the PaO2 is in the range to fully saturate haemoglobin, differences in PaO2 have limited relevance. It may, e.g., be more useful to study the influence of the FiO2 on the incidence of hypoxaemia. Other indices are also used to assess pulmonary gas exchange. P(a-a)O2 is an index of intrapulmonary shunt or V/Q scatter, although it can be affected by PaO2, cardiac output, body temperature, pH and base excess of the blood, haemoglobin concentration, and alveolar ventilation [7]. This review revealed that the alveolar–arterial difference is higher in high fractions of inspired oxygen, an indication of compromised pulmonary function. Unfortunately, there is very limited information of the effect of high inspired oxygen on another index, the PaO2/FiO2 ratio. It would be interesting to have data of this outcome measure, since it has been shown in humans that the PaO2/FiO2 ratio, as well as the P(a-a)O2 difference, depends on FiO2 [43], whereas in sheep, the P(a-a)O2 difference correlation to shunt seems to be weaker than that of the PaO2/FiO2 ratio [44].Another technique to detect the matching between alveolar ventilation and pulmonary blood perfusion is the multiple inert gas elimination technique (MIGET), which uses six (usually) inert gases [45,46]. One study [20] using this technique in horses was found during our literature search. This study shows an increased intrapulmonary shunt when high FiO2 is administered, despite the high PaO2 measured. This shunt persisted into recovery. These results support the findings of this review, although unfortunately we could not include this study in the systematic review and meta-analysis, because this experimental study used dissociative anaesthesia, and also did not calculate the indices we were looking for.It is known that the alveolar O2 concentration affects the development of absorption atelectasis; however, in horses, compression of the lungs because of the shape and the position of the diaphragmatic dome, in combination with a high pressure exerted by the abdominal contents (especially in dorsal recumbency), may promote compression atelectasis. Thus, lowering the FiO2 may not be the only intervention to improve oxygenation, and other strategies, e.g., alveolar recruiting strategies, may be more effective in improving oxygenation.Our review possesses some limitations. The first one is the small number of included studies and animal population. It seems that very few studies have been performed on that topic, and the ones available often differ substantially in their methodology and described outcome, making it questionable to combine their results. Thus, this systematic review is accompanied by a meta-analysis with a fairly limited number of studies. While limited data as well as high heterogeneity are reasons to avoid a meta-analysis, there is no agreement on a cut off regarding these factors [47]. In our opinion, the meta-analysis serves as an additional aid to evaluate the presented information in a concise fashion. Certainly, there is a need for more research in this area, because equine intraoperative hypoxia is a serious problem in clinical practice. The second limitation is that the subgroup analysis was not possible for any comparison. Future well-designed experimental studies and clinical trials using targeted evaluating tools are necessary to increase the level of evidence, for better decision making. The third limitation (although not an inherent one of our review, rather than a limitation of the included studies) is the lack of power analysis in the included studies, which may have compromised the level of evidence. Having in mind these limitations, and despite a clear trend towards specific results, the findings of this review should be interpreted cautiously.5. ConclusionsConsidering a low to medium level of evidence, the reduction of FiO2 in horses under anaesthesia may improve some oxygenation indices, e.g., shunt, but will decrease blood oxygenation.

animals : an open access journal from mdpi

10.3390/ani11061539

PMC8227126

The distribution and expression of aquaporins (AQPs) in the testes and spermatozoa of several animal species play important roles in spermatogenesis and spermatozoon transit in this region. The aim of this study was to evaluate AQP7, AQP8, and AQP9 localization and expression in the efferent ductules and epididymal regions (the caput, corpus, and cauda) of normal and cryptorchid dogs. The results from immunohistochemistry, Western blotting, and real-time reverse transcription polymerase chain reaction (RT-PCR) show regional tissue distributions, particularly at the level of the epithelium of efferent ductules and both the regions caput and cauda of the canine cryptorchid epididymis. These findings support the hypothesis that these channel proteins respond differently to multiple stimuli that cause cryptorchidism (hormones, heat, osmolarity, etc.) and participate in the mechanisms of cell “resilience” or apoptosis taking place in the epididymis.

The efferent ductules and the epididymis are parts of the male reproductive system where spermatozoa mature. Specialized epithelial cells in these ducts contribute to the transport of fluids produced by spermatozoa’s metabolic activity. Aquaporins (AQPs) have been demonstrated to be expressed in the spermatozoan membrane and testis epithelial cells, where they contribute to regulating spermatozoan volume and transit through environments of differing osmolality. Due to the lack of detailed literature regarding AQP expression in the canine male genital tract, the aim of this study was to investigate both the distribution and expression of AQP7, AQP8, and AQP9 in the efferent ductules and epididymal regions (caput, corpus, and cauda) from normal and cryptorchid dogs by using immunohistochemistry, Western blotting, and real-time reverse transcription polymerase chain reaction (RT-PCR). Our results show different patterns for the distribution and expression of the examined AQPs, with particular evidence of their upregulation in the caput and downregulation in the cauda region of the canine cryptorchid epididymis. These findings are associated with a modulation of Hsp70 and caspase-3 expression, suggesting the participation of AQPs in the luminal microenvironment modifications that are peculiar characteristics of this pathophysiological condition.

1. IntroductionThe epididymis is a tract in the male reproductive system that plays a pivotal role in the complex processes of sperm maturation, transit, and storage. It is a convolute tubule divided into three anatomical parts—the caput, corpus, and cauda—representing physiologically unique compartments that create the luminal microenvironment essential for sperm survival. Of note, specialized epithelial cells in the epididymal segments contribute to the metabolic activity of spermatozoa during its transit along the epididymis. In detail, the proximal regions of the epididymis are devoted to sperm maturation [1], while sperm storage takes place in the distal region [2].In humans and larger mammals, but not in rodents, a major portion of the proximal region of the epididymis is characterized exclusively by efferent ductules, a complex histological structure localized particularly at the junction of the caput epididymis. Efferent ductules are channels involved in sperm transit from the rete testis to the epididymis. These ducts have been described as compact and small if compared with the epididymal tubules, which appear long and convoluted [3]. In rodents, the epididymis is divided into four regions: initial segment, caput, corpus, and cauda, whereas in large animals and humans, the initial segment is absent [3]. As described in humans, and also in dogs, the proximal region of epididymis is occupied by blind-ending efferent ductules [3,4,5]. A crucial role of these aforementioned tubules is to increase sperm concentration by a process of water reabsorption (up to 96% of the testicular fluid) prior to reaching the epididymis [6]. This event is also characterized by active transport of ions, followed by passive movements of water under estrogen regulation [3,7,8].Studies have demonstrated that the epididymis is a metabolically active tissue comprising morphologically and functionally distinct cells that regulate important processes, including fluid reabsorption, secretion of proteins, and the production of antioxidants [9]. The luminal water reabsorption by the epididymal epithelium can have the following three main consequences [10]: (1) an increase in the sperm concentration in the deferent ducts of the testes, (2) the modulation of protein concentrations [11,12], and (3) an increase in osmolality that facilitates the regulation of sperm volume [13].To realize these processes, the epididymal epithelial cells of these ducts are specialized, as demonstrated by a wide repertoire of protein and gene expression, transporters, and receptors [14,15] that allow them to respond uniquely to different stimuli, such as hormonal and other regulatory factors. Temperature is another major factor influencing epididymal physiology [2]. Some studies have demonstrated that changes in body temperature modify the ionic and protein composition of the cauda fluid by altering the cauda epithelium and, thus, impair its normal function of storing and prolonging the life of spermatozoa [16,17]. Conditions including cryptorchidism, characterized by modifications of testicular temperature caused by the failure of one or both testes to descend into the scrotum, and a varicocele, which is an enlargement of venous blood vessels within the scrotum, have been demonstrated to be detrimental to sperm production and viability [18,19]. Moreover, heat stress during cryptorchidism induces male germ cell death by apoptosis, which is considered to be the predominant mechanism involved in this condition [20]. It has been well demonstrated that heat stress, altering the normal cellular functions, induces the synthesis of several heat shock proteins, including Hsp70. This protein has a critical role as a marker protein of cell thermotolerance [21], and its expression has been found dramatically increased in rat epididymis as a result of the exposition of animals to acute heat stress (39 °C for 0.5, 1, and 3 h) [22]. More recently, similar results were also observed in male cavy exposed for 60 days to 39 °C, demonstrating decreased sperm mobility, sperm count, and testicular antioxidant enzymes accompanied by an increased serum level of Hsp40 [23].In the last decade, studies focusing on the mechanisms involved in spermatozoon maturation and the luminal microenvironment in contact with spermatozoa have demonstrated the presence of specific water channel proteins named aquaporins (AQPs), expressed either in the spermatozoan membrane or in the testis epithelial cells. The results have suggested that these water channel proteins participate in processes involving water, ions, and molecules, such as lactate in sperm maturation [24,25,26]. AQPs belong to a family of 13 integral membrane proteins normally expressed in the tissues of all living organisms that play a fundamental and pivotal role in water and small-solute transport across cell membranes. AQPs are generally divided into two main groups based on their transfer specificity: classical water-transporting AQPs (AQP0, 1, 2, 4, 5, 6, and 8) and solute-transporting aquaglyceroporins (AQP3, 7, 9, and 10). In addition, a third aquaporin group, comprising AQP11 and AQP12, the most distantly related paralogs, are defined as superaquaporins [27] or subcellular aquaporins [28], according to their cellular localization.Characterization studies on AQPs in the male reproductive tracts of several domestic animals (buffalo, hamsters, horses, pigs, rams, and sheep) [29,30,31,32,33,34,35] have demonstrated the presence of these channel proteins in different tracts, also showing different functions according to the examined regions, cell types, membrane domains, species, breeding seasons, and age of the subjects [36,37,38,39].Regarding domestic animal species, studies on dogs have often been conducted regarding either testicular organization and function [40,41,42] or clinical problems related to the specific dysfunction of this tissue [43]. From the limited studies reported, data on the expression and possible role of AQPs in the canine male reproductive tract are limited to the testicular area [38,39,44]; the presence of AQP1 and AQP9 in different cell types in the testis [38,39,44]; and the presence of AQP3, 4, 7, and 9 at the level of the gubernaculum testis (GT) [45], suggesting the need for the further investigation of their roles using disease models of the reproductive system.The available data, mostly obtained from laboratory animal studies, demonstrate that systematic analyses of the expression of AQPs in epididymal epithelial cells have been lacking [46] and that there have been few studies focused on AQP8 [37,47]. Noteworthily, as indicated by Kirchhoff (2002) [48], the canine epididymis represents a useful model with high relevance to humans, especially at the level of tissue-specific gene expression. Comparable results at the molecular level have been reported from analyzing epididymis from a wide variety of dog breeds and mongrels [49], thus emphasizing the importance of studies on this topic.The aim of this study was to investigate both the distribution and expression of AQP7, 8, and 9 in the caput, corpus, and cauda of the epididymis of normal and cryptorchid dogs by using real-time reverse transcription polymerase chain reaction (RT-PCR), Western blotting, and immunohistochemistry. In the second part of the study, we aimed at analyzing, in the same tracts of normal and cryptorchid canine epididymis, Hsp70 and caspase-3 expression by Western blotting, considering their possible relation to AQPs since channel proteins also regulate cell homeostasis in relation to a physiological process leading to cell death, apoptosis, which is a common characteristic of cryptorchidism. 2. Materials and Methods2.1. Animals and Tissue Collection A total of 10 dogs (5 normal and 5 cryptorchid with unilateral cryptorchidism), medium sized, and aged between two and eight years old, were enrolled in this study. The dogs came to the University of Federico II veterinary clinic, Naples, Italy, for orchiectomy between April and November 2018. All the experimental procedures were conducted with the approval of the University of Naples Federico II and local Ethics Committee (approval number 0-050-377), in accordance with relevant national and international guidelines. Epididymis samples were obtained immediately after the removal of testicles and transported on ice (+4 °C) within 40 min for further processing. The epididymal tissues were divided into two groups: normal epididymis (epididymis from normal dogs) and cryptic epididymis (retained epididymis from cryptorchid dogs). From each group, the caput, corpus, and cauda of the epididymis were sectioned and fixed in Bouin’s fluid for immunohistochemistry. For Western blotting and real-time RT-PCR, the samples were frozen at −80 °C until analysis.2.2. ImmunohistochemistryFixed samples were embedded in paraffin and sectioned at a thickness of 6 μm. For immunohistological study, the sections were deparaffinized and hydrated through xylene in a graded ethanol series. They were also pretreated with citrate buffer (pH 6) for 2 min (two times) in a microwave (700 watts) for antigen retrieval. After cooling for 30 min, the samples were incubated in 3% H2O2 for 20 min, in a humid chamber, to quench endogenous peroxidase. Then, they were washed in phosphate-buffered saline (PBS) for 5 min (three times) and blocked in diluted normal goat serum (Vector laboratories, Inc., Burlingame, CA, USA) for 30 min. Next, they were incubated with rabbit polyclonal antibodies against AQP7 (orb13253), AQP8 (orb101163), and AQP9 (orb10127) (dil., 1:500; Biorbyt LLC, San Francisco, California, USA), at 4 °C overnight.The following day, the sections were washed with PBS and incubated with a secondary antibody, Ultra-Polymer Goat anti-Rabbit/Mouse IgG (ImmunoReagents, Raleigh, NC, USA) conjugated with a peroxidase polymer backbone (1:4), for 30 min in a humid chamber. After washes with PBS, they were treated with diaminobenzidine 3,3′ tetra hydrochloride (DAB) (Vector laboratories, Burlingame, CA, USA) until the desired stain intensity developed. Finally, the sections were dehydrated through an ascending alcohol series mounted with Eukitt® (Sigma-Aldrich, Taufkirchen, Germany), observed under a Nikon Eclipse E 600 light microscope, and photographed using a Nikon Coolpix 8400 digital camera. The negative controls were reactions performed without the primary antibody incubation step, as described elsewhere [50], and a sample from a rat kidney was used as the positive control.2.3. Western BlottingThe epididymal tissues were processed for Western blotting according to the previous adopted method [51]. In brief, the tissues were homogenized in RIPA buffer and then centrifuged at 14,000 rpm for 15 min, at 4 °C, to remove the nuclei and cell debris. The resultant supernatant was collected, and the protein concentration was determined using the Bradford assay (Bio-Rad Laboratories Inc., Hercules, CA, USA). Lysates containing equal protein contents (30 μg) were resuspended in Laemmli buffer and loaded on a 4–20% Mini-PROTEAN, TGX Stain-Free precast electrophoresis gel (Bio-Rad Laboratories, Inc., Hercules, CA, USA).After electrophoresis, the proteins were transferred to a nitrocellulose membrane using a Mini Trans-Blot apparatus (Bio-Rad Laboratories Inc., Hercules, CA, USA), and the transfer of the protein onto the membrane was checked using a ChemiDoc molecular imager (Bio-Rad Laboratories, Inc., Hercules, CA, USA). For AQP8 and 9 immunoblotting, the nitrocellulose membrane was stripped using stripping buffer (Hi Media laboratories, Mumbai, India) and re-probed. The Western blot analysis of caspase-3 and Hsp70 was performed using the remaining protein lysates. Following gel transfer, the membrane was blocked with 5% non-fat milk diluted in TBS-T buffer (1.5 M NaCl, 200 mM TRIS-HCL, and 0.1% Tween 20, pH 7.2) at room temperature under constant motion for an hour.After washes in TBS-T, the blot was incubated overnight at 4 °C with primary rabbit polyclonal antibodies directed against AQP7 (dil., 1:500, orb13253), AQP8 (dil., 1:500; orb101163), AQP9 (dil., 1:500; orb10127), caspase-3 (dil., 1:500), and Hsp70 (dil., 1:500; orb415817) (Biorbyt LLC, San Francisco, California, USA).One day after incubation, the membrane was washed three times with TBS-T for 10 min and incubated with a Goat anti-Rabbit secondary antibody conjugated with horseradish peroxidase (HRP) (ImmunoReagents, Raleigh, NC, USA; dil., 1:1000) for 1 h at room temperature. After washing the membrane 3 times with TBS-T, ECL (Bio-Rad Laboratories, Inc., Hercules, CA, USA) was used to visualize the proteins, and an image was taken using the ChemiDoc molecular imager (Bio-Rad Laboratories, Inc., Hercules, CA, USA). For a positive control, canine testicular tissue extract was used, while for a negative control, the membrane filter was treated only with TBS-T used for antibody dilution but omitting the primary antibody. The standard molecular weight marker used was Precision Plus Protein™ All Blue Prestained Protein Standards (1–250 KdA, #1610373, Bio-Rad Laboratories, Inc., Hercules, CA, USA). Obtained bands for the examined proteins were visualized with Image Lab 6.0 program and were normalized to total proteins using the stain-free technology gels. The results are expressed as the intensity relative to that for normal tract.2.4. Real-Time RT-PCR and Data ProcessingTotal RNA extraction, cDNA synthesis, RT-PCR, and sequencing were performed as previously described [38]. Briefly, total RNA was extracted using an Ultra-Turrax homogenizer from tissues in ice-cold TRIzol reagent (Life Technologies, Carlsbad, CA, USA). After chloroform extraction and isopropyl alcohol precipitation, the RNA was dissolved in RNAase-free diethyl dicarbonate (DEPC) water. Then, the RNA content was quantified using an Eppendorf BioPhotometer (Eppendorf AG, Basel, Switzerland). For the conventional and real-time RT-PCR reactions, primers specific for the selected canine transcripts were designed using Primer Express.These specific primers, which amplify 200 (AQP7 and 8) and 150 (AQP9) base pairs, were designed based on the published GenBank gene sequences for the Canis lupus familiaris AQP7, 8, and 9 mRNAs.The GenBank accession numbers and sequences of the primers are listed in Table 1. To prepare cDNA, 1 μg of total RNA was retrotranscribed using High Capacity cDNA Reverse Transcription Kits (Applied Biosystems, Carlsbad, CA, USA), according to the manufacturer’s instructions, using random hexamers as the primers. The PCR cycle conditions were as follows: 94 °C (30 s), 60 °C (30 s), and 72 °C (1 min) for 35 cycles and then 72 °C. The PCR products for canine AQP7, 8, and 9 were purified using a GFX PCR DNA and Gel Purification Kit (28-9034-70, GE Healthcare, Little Chalfont, Buckinghamshire, UK) and sequenced. Quantitative RT-PCR was used to study the mRNA transcript profiles for these genes. The real-time PCR reactions contained 1 μL of cDNA (20 ng/well) and 24 μL of SYBR Green Master Mix (Applied Biosystems, Carlsbad, CA, USA) containing specific primers. The PCR conditions were as follows: 50 °C for 2 min and 94 °C for 10 min, followed by 40 cycles of 94 °C for 15 s and 60 °C for 1 min. The GAPDH gene was also amplified in separate tubes under the same conditions to serve as an active endogenous reference for normalizing the quantification of the mRNA target.Real-time detection was performed on an ABIPRISM 7300 Sequence Detection System (Applied Biosystem, Foster City, California, CA, USA), and data for the SYBR Green I PCR amplicons were assessed with the ABI 7300 System SDS Software. The relative expression for all the epididymal segments was quantified using the delta-delta Ct method (2−ΔΔCt), as described previously [52]. GAPDH expression was used for the normalization of the AQP7, 8, and 9 expression levels between the different samples.2.5. Statistical AnalysisThe results of the densitometric analysis and real-time RT-PCR are expressed as the mean ± standard deviation (SD) for all the values calculated. Variance analysis (ANOVA) for unpaired data and Tukey’s HDS test for independent samples were used to analyze the significance of differences in the relative abundances of the AQP7, 8, and 9 mRNA between the different epididymal segments in normal and cryptorchid dogs. In addition, the statistical significance of differences in the AQP7, 8, and 9 mRNA levels between the calibrator (normal epididymal segments) and cryptic counterparts were also determined using Student’s t-tests. All the experiments were performed in triplicate. A value of p < 0.05 was considered to indicate a statistically significant difference.3. Results3.1. Immunohistochemistry of AQP7, AQP8, and AQP9 in Normal and Cryptorchid Canine EpididymisThe results of the immunohistochemistry analyses of AQP7, 8, and 9 in the different epididymal segments of normal and cryptorchid dogs are shown in Figure 1, Figure 2 and Figure 3, respectively. Histologically, the normal epididymis is lined by pseudostratified epithelium consisting of three main regions: caput, corpus, and cauda. Principal cells represent the major cell type throughout the entire epididymis, followed by clear, narrow, apical, basal, and halo cells and the number, appearance, and function of these cell types vary in each segment of the epididymis. The height of the principal cells changes from tall columnar in the caput of the epididymis to low columnar in the corpus and cauda of the epididymal segments. The cryptorchidism results in reduced diameter of the duct and reduction in the length of the epidydimal epithelium. Fine histological alterations were observed between normal and cryptorchid epididymis, which consisted of the narrowing of the epididymal duct, flattering of the epithelial cells, disorganization of the tubular arrangement, and the presence of abundant interstitial tissue throughout the entire epididymis with the absence of spermatozoa in the lumen of the duct itself.In the normal dogs, intensely stained AQP7-containing granules lined the entire cytoplasmic profile and the basal portion of the efferent ductules epithelium (Figure 1A, arrows). Positive material, weakly stained, was found in the apical (Figure 1B,D, arrowheads) and basal (Figure 1D, pounds) portions of the principal cells in the caput and cauda of the normal epididymis. Several basal cells positive for AQP7 were localized in the corpus (Figure 1C, double arrows). Some immunoreactive narrow cells were also detected in the caput of the normal epididymis (Figure 1B, line arrow), whereas in the cryptorchid epididymal segments, immunoreactive basal cells were observed in the caput and cauda epididymis (Figure 1E,F, double arrows). Dispersed positive granules immunoreactive to AQP7 were found in the cytoplasm of the principal cells of the cryptorchid cauda (Figure 1F), while the corpus did not exhibit any labeling more than the background (data not shown). For AQP7, in the normal epididymis, the density of immunoreactivity (IR) was higher in the cauda and corpus than the caput epididymis, and the labeling was confined to cytoplasmic domains.In the normal subjects, AQP8 IR was observed in the cytoplasm of the efferent ductules epithelium (Figure 2A, arrows), particularly in the apical and basal portions. Immunoreactive material was also detected in the apical cytoplasm (Figure 2B–D, arrowheads) of the principal cells of all the epididymal segments. AQP8 IR was found in the basal portion (Figure 2C,D, pound) of the principal cells of the corpus and cauda epididymis. Some narrow cells reactive to AQP8 were intensely stained and localized in the caput epididymis (Figure 2B, line arrow). In the cryptorchid epididymal segments, however, basal cells (Figure 2E, double arrows) exhibited immunoreactivity in the caput. Immunopositivity was also observed in the apical portion of the principal cells of the cryptorchid cauda, with low intensity (Figure 2F, arrowheads). To better identify the positive cytotypes, inserts were added in Figure 1A,B and Figure 2D,E.As described for the other peptides, AQP9 IR was found with the same immunolocalization in the efferent ductules (Figure 3A, arrows). A faint positivity was detected in the apical portion of the principal cells of the normal caput and cauda epididymis (Figure 3B,D, arrowheads). A few basal cells were positive (data not shown). In contrast to that for the other AQPs, intranuclear immunoreactivity was also observed for AQP9 throughout the segments of the normal epididymis (Figure 3B–D). However, there was a very faint immunoreactivity in the apical portion of the principal cells of the caput and cauda segments of the cryptorchid epididymis (Figure 3E,F, arrowheads).Results of AQPs expression and their cellular distribution in the efferent ductules and epididymal segments of normal and cryptorchid dogs are summarized in the Table 2 and Table 3, respectively.3.2. Analysis of Expression of AQP7, AQP8, and AQP9 in Normal and Cryptorchid Canine Epididymis by Western BlottingThe results of the Western blot analysis for AQP7, AQP8, and AQP9 are reported in Figure 4A.For AQP7, immunostaining was clearly observable at 25–42 kDa and about 60 kDa for all the examined tracts of the epididymis of the normal and cryptorchid dogs, although differences in band intensity were apparent for both molecular weights between the normal and cryptorchid dogs. In particular, the data from the densitometric analysis (Figure 4B) demonstrated similar band intensities for AQP7 in all the tracts of the epididymis for the normal dogs, while a slight increase was observed only for the caput epididymis in the cryptorchid dogs relative to that in the normal dogs. A different pattern of AQP7 expression was observed in the corpus and cauda regions of the epididymis of the cryptorchid dogs, where the levels of the protein were relatively low.The Western blotting image in Figure 4A shows, for AQP8, two bands corresponding to 35–48 kDa and ~60 kDa for all the tracts, although differences in intensity were particularly evident for the tracts of the cryptorchid dogs. The relative intensity of AQP8 (Figure 4B) was similar for all the epididymal tracts of the normal dogs. By contrast, while moderately high levels were observed for the caput and corpus of the epididymis of the cryptorchid dogs relative to those for the normal dogs, a lower intensity was observed for the cauda region of the epididymis.The Western blotting image shows (Figure 4A), for AQP9, an intense band corresponding to ~60 kDa and two weak bands corresponding to 35–48 kDa for all the tracts of the normal and cryptorchid epididymis, albeit slight differences are observed between different tracts and normal vs. cryptorchid conditions. Densitometric analysis (Figure 4B) reveals a lower intensity for the corpus and cauda tracts in the epididymis of the cryptorchid dogs than those of the normal dogs, while only a slight increase in intensity is observed for the caput epidydimal tract of the cryptorchid dogs.3.3. mRNA Expression of AQP7, AQP8, and AQP9 in Normal and Cryptorchid Canine Epididymis According to Real-Time RT-PCRReal-time RT-PCR analysis was performed to determine the expression of these AQPs in different segments of the epididymis from the normal and cryptorchid dogs. As shown in Figure 5A, in the normal dogs, all three AQP mRNAs were expressed in all the segments of the normal epididymis, with similar expression levels except for AQP7, which showed the highest value in the cauda. The observations were different for the cryptorchid dogs; AQP7 and AQP8 showed low mRNA levels in the corpus and cauda and high levels in the caput of the epididymis (Figure 5B); by contrast, AQP9 mRNA was high in all three segments of the epididymis, with a slight increase in expression in the caput (Figure 5B). In addition, this analysis showed a similar trend for the AQP7 and AQP8 transcripts in the segments of the epididymis of the cryptorchid dogs as compared with the normal dog samples. Increases in the mRNAs were observed for the caput and corpus, while a decrease was shown in the cauda (Figure 5C). The increase in the AQP7 and AQP8 mRNAs relative to the normal sample levels was particularly high only in the caput. However, AQP9 mRNA was slightly less abundant in all the epididymal segments of the cryptorchid dogs compared to in the normal dogs (Figure 5C).3.4. Western Blot Analysis of Hsp70 and Caspase-3 Expression in Normal and Cryptorchid Canine EpididymisTo better investigate the possible role of AQPs in the epididymis and their relationships with factors such as heat and oxidative stress, which are both involved in cryptorchidism, Hsp70 and caspase-3 were evaluated (Figure 6A). The densitometry showed an increase in Hsp70 protein expression in the cryptorchid condition compared to the normal dog samples (control) in all the segments of the epididymis, with the highest and most significant value in the caput followed by the corpus and cauda regions (Figure 6B).Caspase-3 expression showed a reduced expression in both the epidydimal segments of the caput and corpus of cryptorchid dogs compared to the control, with a significant decrease for the first epididymal segment. In the cauda, the protein showed its highest expression level with a value of relative intensity > 2.5 compared to the control. (Figure 6B).4. DiscussionIn the present study, we examined the localization and expression of the AQP7, 8, and 9 proteins and their mRNAs in the efferent ductules and epididymis of normal and cryptorchid dogs. To clarify the possible roles played by the examined AQPs in the epididymal tissue, we analyzed the expression of Hsp70 and caspase-3, which are well known to be involved, among numerous other factors, in the heat stress and oxidative mechanisms related to cryptorchidism [53,54,55].Our results obtained by immunohistochemistry demonstrate the presence of the examined AQPs in different epithelial cells of the efferent ductules and of all the tracts of the epididymis in both normal and cryptorchid dogs, although a less dense distribution of cellular immunoreactivity to AQPs was shown in the cryptorchid dogs than in the normal dogs. Moreover, the analyses of the localization and relative protein and mRNA expression of the AQPs showed significant differences among the examined segments of the epididymis.In normal subjects, the results of AQP8 and AQP9 IR demonstrated their presence in the basal and apical portions of the cells of the efferent ductules epithelium. Similar cellular distribution for AQP9 was well described by Domeniconi et al. (2007) [39], who reported its expression in the apical brush border of non-ciliated cells. The possible role played by AQP9 could be to allow the movement of water across the epithelia to improve the passage of glycerol, considering that this molecule serves as a metabolic substrate for sperm to produce CO2 [44,56].Conversely, our results relative to the positivity of epithelial cells to AQP7 disagree with that of Domeniconi et al. (2008) [44], leaving us to attribute this discrepancy to different factors, including a different sensitivity of the used antibody and the methods of tissue preparation and immunocytochemical procedures utilized in each case. Of note, the only immunohistochemical study in rats showed both AQP7 and AQP9 in the efferent ductules, with a moderate expression of AQP7 either over the microvilli of epithelial cells or along the basolateral plasma membranes [26,44]. The co-expression of different AQPs along the epithelial cells of this reproductive tract of dogs suggests their possible role in regulating luminal fluid composition to promote spermiogenesis [57]. Moreover, considering that AQPs are influenced by hormones, the absence of AQP IR demonstrated in the cryptorchid dogs could be due to a reduced production of androgens because of the cryptorchid condition, as previously observed in rats [58]. Additional experimental data using mice deficient in α-estrogen receptors (αESRKO mice) demonstrated a decreased AQP1 and AQP9 expression in the efferent ducts, confirming the important hormonal control on the efferent ductules function in retention of water [59].In normal epididymal segments, our results demonstrated slight differences in the cellular distributions of the aquaglyceroporins AQP7 and AQP9 along the epididymal segments; AQP7 was localized in the cytoplasm of the principal cells of the caput and cauda segments and basal cells only in the corpus, whereas AQP9 was distributed only in the principal cells of the entire epididymal tract, albeit the immunoreactivity was exclusively intranuclear. The AQP7 expression was also confirmed by Western blot demonstrating bands at ~30 kDa (faint band) and ~42–60 kDa, respectively. This result confirms previous reports on dogs [44] showing only the first band, while protein extracts from bull spermatozoa only displayed a band corresponding to 45 kDa [60]. The presence of AQP7 in the epididymal basal cells was interpreted as being part of the intricate cooperative participation in the removal of water and/or small uncharged molecules from the epididymal lumen throughout the epididymis [44]. The basal cells are numerous in the corpus as previously demonstrated [61,62,63], and their particular morphological characteristic gives them huge potential for the movement of water, small uncharged molecules, or both in the epididymis morphology [26].The distribution of AQP9 in all the segments of the normal canine epididymis as well as its expression as a band corresponding to 30 kDa confirms previous data by Domeniconi et al. (2007) [39]. Other previous findings show this protein in the epididymis of rats [64] and mice, respectively [65], albeit with a localization in the apical stereocilia of principal cells. AQP9 could participate in the trafficking of water and/or solute permeability of these epithelia, contributing to the processes taking place in this tissue. Furthermore, the intranuclear localization of AQP9 observed in the epithelial cells of the epididymis in our study has also been reported in sheep and post-pubertal pigs [37], which suggests a putative role in purine and pyrimidine transport [66].The co-expression of AQP7 and AQP9, having in common the ability to allow the cellular membrane trafficking of water, small solutes, and glycerol, albeit distributed in different epididymal cells, could sustain the effective transport of glycerol and glycerylphosphorylcholine from the epithelium to the lumen, where these molecules play a pivotal role in the process of sperm maturation [67].AQP8 IR was shown in the apical portion of the principal, basal, and narrow cells of the normal epididymal segments in a region-specific manner. The presence of AQP8 was also observed, although with slight immunoreactivity, in the rat epididymal basal cells [68], and its role could be associated with that assumed for these cells at the level of the conductive airways. At this level, AQP8 could be involved in the differentiation of epithelial cells during maturation or after recovery from injury to the airways [68]. Notoriously, basal cells present in all mammalian species are supposed to sustain a stem function in the renewal of the epithelium [69,70,71]. The presence of these AQPs at the level of the epididymis, as well as the suggestion of their involvement in different processes, are all coordinated by hormonal control that could include: (1) the water absorption from the testicular fluid; (2) the creation of a luminal microenvironment in physical and nutrient parameters essential to spermatozoa growth; and (3) the transfer of solutes and glycerol to sustain spermatozoa along the cauda epidydimal segment.In the cryptorchid condition, significant and intense differences were observed in terms of AQP localization and expression in all the segments of the epididymis. AQP7 and AQP8 showed a moderate shift in their cellular localization compared to control passing from principal, narrow, and basal cells (control) to the principal and basal cells, albeit with a faint IR. AQP9 IR showed a cellular localization along the epididymal segments similar to that observed for the control. The presence of AQP7 and AQP8 in the basal cells in the course of this disease could hypothesize a different and new role exerted by these proteins. Basal cells have been largely investigated for their multifaceted roles, facilitating the cell–cell crosstalk [72] and functioning as luminal sensors of principal cells [73] throughout the prostaglandin production [74]. Particularly noteworthy is the participation of basal cells in immune functions and their possession of genes that encode proteins involved in cell adhesion, cytoskeletal arrangement, ion transport, cellular signaling, and inflammatory responses [75]. Moreover, Pinel et al. (2019) [76] have reported a new and intriguing involvement of these cells as responsible for the equilibrium between apoptosis and cell renewal in the epididymal epithelium. The role of these cells and the associated AQP expression could be associated with the pathway of Hsp70 and caspase expressions. In particular, in the caput where AQP IR was quite exclusively associated with these cells (only a faint AQP9 IR is associated with epididymal principal cells) and a high level of Hsp70 expression was shown, a mechanism of homeostasis regulation at the level of the lumen microenvironment to counteract the oxidative stress responsible for sperm cell death could be hypothesized.This hypothesis is corroborated by evidence that the cellular localization of Hsp70 is selective for basal cells in the human epididymal epithelium [53] and the fact that we demonstrated, by Western blotting, high AQP7 expression in the caput of the cryptorchid canine epididymis. In addition, Lu et al. (2013) [77] proposed that other mechanisms, such as necrosis or inflammation, could be preferred over apoptosis at this level. Prieto-Martinez et al. (2017) [78] evaluated AQP expression in boar sperm in relation to sperm cryotolerance and demonstrated that AQP7 was expressed in ejaculates exhibiting good freezability (GFE), suggesting its possible role as a freezability marker.Similarly, for AQP9, the higher expression observed in the caput of the cryptorchid dogs, associated with high levels of Hsp70, could suggest its involvement in the protection from stress. Of note, previous in vitro studies demonstrated an upregulation of AQP9 in rat intestinal epithelial exposed to hypertonic stress [79].The localization of AQPs in the cauda region of the epididymis either along the basal or principal cells could suggest different roles of these proteins in this epididymal segment. In particular, the principal cells where AQPs are localized could be involved in the wave of apoptosis determined by androgen loss. AQP9 could play a role in this mechanism since its downregulation in mouse Sertoli cells exposed to excess of 17β-Estradiol (E2) has been demonstrated [80]. Similarly, AQP9 downregulation has also been shown to be associated with apoptosis and caspase-3 increase in retinal ganglion cells (RGCs) exposed to stress, thus demonstrating the role of this protein in transporting lactate as an energy substrate and as a source of ROS scavengers. Indirectly, these findings suggest that the cryptorchid condition could influence AQP9 preventing its role in transporting glycerol in this district and thus could induce cell death [81].Thus, we could hypothesize that cryptorchidism in dogs could impair a general dysregulation of the epithelial cells along the efferent ductules and epididymal segments, influencing the expression of AQPs. The particular AQP cellular distribution and the modulation of Hsp70 and caspase-3 (Figure 7) could suggest their different responses to cellular alterations are probably associated with mechanisms ranging from inflammation or necrosis to male germ cell death, which are often observed in the cryptorchid condition or in the experimental condition of the efferent duct ligation [82,83]. Further studies are needed to characterize possible segment-specific regulation of epididymal cell apoptosis as well as other possible mechanisms involved in the canine cryptorchid condition.5. ConclusionsThe different patterns of the expression and distribution of AQPs in the efferent ductules and in the epididymal segments of cryptorchid dogs suggest their involvement in luminal microenvironment modifications that are peculiar characteristics of this pathophysiological condition. The cryptorchid condition characterized by multiple associated processes (androgen decrease, heat stress, etc.) could influence the sensitivity of AQPs at the cellular level, modulating their activity. Further investigations using in vitro models are needed to clarify the exact role played by each of the examined AQPs, and a simultaneous examination of the different immunological, oxidative, and inflammatory states of the epididymis would improve the knowledge on the pathogenesis of canine cryptorchidism.

animals : an open access journal from mdpi

10.3390/ani11072078

PMC8300236

Pasture-borne worm infections impact cattle health and productivity worldwide. The present study assessed exposure of dairy cattle herds to the three most important pastural parasites, i.e., gastrointestinal worms, liver flukes and lungworms, in three parts of Germany by measuring antibodies in bulk tank milk samples. The results show a high level of exposure to gastrointestinal worms, while antibodies against liver flukes were less frequently detected and lungworm-positive herds were rare. Regional and breed differences regarding parasite exposure were detected. In addition, the presence of antibodies was associated with access to fresh grass, access to hay, silage quality and deworming frequency. Furthermore, parasite exposure was significantly associated with a poor body condition across all regions. Parasite-exposed cows of high-performance breeds also produced on average less milk per year than dual-purpose breeds.

Pasture-borne parasites adversely affect bovine health and productivity worldwide. In Europe, gastrointestinal nematodes, especially Ostertagia ostertagi, the liver fluke Fasciola hepatica and the lungworm Dictyocaulus viviparus represent the most important parasites of dairy cattle. The present study assessed exposure towards these parasites among 646 cattle herds in three parts of Germany during 2017–2019 via antibody detection in bulk tank milk (BTM). Overall, O. ostertagi levels indicative of production losses were detected in 41.2% (266/646; 95% confidence interval (CI): 37.4–45.1%) of BTM samples, while F. hepatica seroprevalence amounted to 14.9% (96/646; 95% CI: 12.2–17.9%). Only 2.3% (15/646; 95% CI: 1.4–3.9%) of samples were D. viviparus antibody-positive. Significantly lower O. ostertagi as well as F. hepatica seroprevalence was detected in dual-purpose breeds compared to high-performance breeds from the same region. Management factors related to parasite exposure included access to fresh grass and hay, silage quality and anthelmintic treatment. Furthermore, F. hepatica and O. ostertagi seropositivity was significantly associated with suboptimal herd-level body condition. Interestingly, the relationship between seropositivity and productivity differed between breed types. Negative impacts on milk yield were detected only in high-performance breeds, while O. ostertagi seropositivity was associated with a lower milk fat content in dual-purpose herds.

1. IntroductionPasture-borne parasites represent a major global problem for bovine health and productivity. Gastrointestinal nematodes (GIN) of the family Trichostrongylidae, causing parasitic gastroenteritis, are relevant for grazing cattle worldwide [1]. In temperate regions of Europe, including Germany, Ostertagia ostertagi is the most prevalent species [2,3,4]. Although infections are often subclinical in dairy cows, studies in several European countries have shown a significant negative correlation between O. ostertagi antibody levels and cow productivity, especially in terms of milk yield [5,6]. In addition, a lower milk protein content has been observed in animals with patent GIN infections [7]. The liver fluke Fasciola hepatica and the lungworm Dictyocaulus viviparus are less prevalent than GIN, but nevertheless represent significant economic burdens [1]. Liver fluke infections mostly cause chronic disease in cattle, resulting in reduced milk yield, impaired fertility, and condemnation of affected livers [8,9,10]. Furthermore, a correlation between elevated β-hydroxybutyrate levels in milk, indicating a negative energy balance and a state of ketosis, and F. hepatica antibody titers has been demonstrated [11]. Parasitic bronchitis due to D. viviparus may result in severe clinical signs, thus affecting animal welfare and leading to costs for treatment, or even animal mortality [12]. In addition, patent D. viviparus infections are associated with a lower average daily milk yield [13]. On herd level, negative effects on milk fat and milk protein content have also been observed [14,15]. Screening of bulk tank milk (BTM) samples for the presence of antibodies via ELISA constitutes a reliable and easy method to assess herd parasite exposure [16]. As O. ostertagi exposure is generally considered high, available studies usually report mean ELISA optical density ratios (ODRs) rather than prevalence. In Germany, previous studies on dairy cattle using BTM samples indicated mean ODRs of 0.45–0.66, which were regarded intermediate compared to other European countries [17,18]. In this context, BTM ODRs ≥ 0.5 are considered as indicative of a reduction in milk yield [17]. In addition, Fanke et al. [19] reported a seroprevalence of 28.2% using an ODR of 0.6 as cut-off, with 46.5% of herds displaying ODRs between 0.3 and 0.6, and no regional differences. Regarding F. hepatica, on average 23.6% of German dairy herds were seropositive in 2008, with considerable regional differences and the highest prevalence rates in the northern and north-western parts of the country [20]. High levels of F. hepatica exposure in north-western Germany were also confirmed by recent studies [11,19]. Similar to F. hepatica, the lungworm D. viviparus shows an unequal distribution in Germany, with regional seroprevalence rates ranging from 0.0% in the south-western federal state of Saarland up to 31.2% in central and northern parts of the country, as determined in the year 2008 [21]. Despite the negative consequences of parasite exposure, pasture access for dairy cattle is desirable from an animal-welfare perspective [22] and is increasingly demanded by consumers [23,24]. At the same time, rising levels of anthelmintic resistance and changes in the global climate as well as changes in management practices, e.g., an increase in organic farming systems with a restricted use of anthelmintics, may lead to altered patterns of parasite prevalence in farmed cattle [25]. For example, an increase of F. hepatica prevalence among cattle and/or an increase in the geographical spread of this parasite have been observed in some European countries during recent years [26]. Furthermore, a climate-related decrease in F. hepatica, but an increase in O. ostertagi prevalence was noted over an eight-year study period in Belgium [27]. Similarly, an increase in diagnosed cases of parasitic gastroenteritis as well as dictyocaulosis has been noted from 1975–2014 in the United Kingdom, with more lungworm outbreaks occurring during the winter months [28].To provide an up-to-date estimate of dairy herd exposure towards GIN (O. ostertagi), F. hepatica and D. viviparus, the current study assessed 646 BTM samples from three different parts of Germany, collected in the period 2017–2019. The dairy industry in these three parts is characterized by distinct structural differences, with medium-sized, numerous farms in northern Germany and numerous, small farms in southern Germany, while fewer, larger farms exist in the eastern part of the country, the former German Democratic Republic [29]. Furthermore, high-performance dairy breeds, such as the Holstein-Friesian, dominate in the northern, western and eastern parts, whereas dual-purpose breeds, such as the German Simmental, are common in southern Germany. Therefore, seroprevalence patterns were analyzed with regard to the different sampling regions and years. Furthermore, regression models were used to assess the impact of management factors on seropositivity and associations with body condition and herd productivity.2. Materials and Methods2.1. Farm Selection, Questionnaire and Farm VisitsIn the frame of the “PraeRi” project, a government-funded research project on animal health, biosecurity and housing environment on German dairy cattle farms [30], a total of 8944 farms were initially contacted, of which 765 agreed to participate in the study. Farms were located in three parts of Germany, namely in the North (i.e., the federal states Schleswig-Holstein and Lower Saxony), the East (i.e., the federal states Mecklenburg-Western Pomerania, Brandenburg, Thuringia, Saxony-Anhalt) and in the South (i.e., the federal state of Bavaria). Each farm was visited by a team of veterinarians once during the study period. During this visit, the body condition scores (BCS) of cows (up to a maximum of 292 cows/farm) were assessed and categorized according to Edmonson et al. [31] and Metzner et al. [32]. Because body condition changes in a breed-specific manner during lactation, cows were categorized as being below, within and above the optimal BCS range depending on their breed and stage of lactation, as described by Oehm et al. [33]. Additionally, the cow-comfort-quotient according to Nelson [34], skin lesions in different body parts [35,36,37], and lameness [38,39] were assessed. To evaluate the quality of hay and grass silages used by the farmers, samples were taken and analyzed by the service laboratory of the Lower Saxony Chamber of Agriculture (LUFA Nord-West). Using the classification into different LUFA quality scores (QS) ranging from 1 to 4, the farms were categorized as follows: QS < 3: silages with normal to slightly reduced quality (category 0); QS = 3: at least one silage with highly reduced quality (category 1); QS = 4: at least one spoiled silage (category 2). Regarding fully slatted floors, farms were categorized as 1 if at least one compartment for cows had fully slatted floors and 0 if no compartment had fully slatted floors.During the visit, farmers were interviewed based on an extensive questionnaire with regard to the operational type of the farm, management practices, e.g., access to pasture, feeding regimen and anthelmintic treatment, as well as occurrence of health problems. Based on the interview, the usage of hay in the feeding rations was categorized as follows: low percentage or no hay in the ration (category 0), hay dried differently than on the floor (category 1) and floor-dried hay (category 2). With regard to anthelmintic treatments, different age classes were considered separately: regarding calves (pre-weaning or <6 months of age) and young cattle (weaning/6 months of age to first calving), categories corresponded to no anthelmintic treatment (category 0), treatment of young cattle as well as calves (category 1), treatment of calves only (category 2) and treatment of young cattle only (category 3). The information for lactating and dry cows was categorized accordingly, with category 0 for no anthelmintic treatment, category 1 for treatment of lactating and dry cows, category 2 for anthelmintic treatment of lactating cows only, and category 3 representing treatment of dry cows only.2.2. Breed Information and Milk Production ParametersAnimal-level breed information was obtained from the national cattle registration database (HI-Tier, Bavarian State Ministry for Agriculture and Forestry). Breeds were grouped into “high-performance dairy breeds” (HD), including Holstein-Friesian, Brown Swiss, Angler, Jersey and their crosses, and “dual-purpose breeds” (DP), including German Simmental, Pinzgauer, Deutsches Schwarzbuntes Niederungsrind (the founder of the modern Holstein breed) and their crosses. Farms were assigned to the category HD or DP if ≥80% of animals were of the respective breed type; otherwise, they were categorized as mixed-breed farms.Data on milk production were provided by the milk recording systems of the different federal states (Landeskontrollverbände). Monthly test-day data were available from the year prior to the farm visit (i.e., based on approximately 11 test-day records). Based on these recordings, herd averages for annual milk yield per cow (kg), milk protein content (%), milk fat content (%), somatic cell count (SCC), lactation number and number of lactating animals were calculated, excluding farms with less than 10 lactating cows.2.3. BTM SamplingFollowing the farm visit, one BTM sample was collected per farm. Sampling was conducted primarily from August–October of 2017, 2018 and 2019, i.e., towards the end of the grazing season, as D. viviparus as well as O. ostertagi BTM antibody titers reach the highest levels during these months [40,41]. Furthermore, a few samples taken in November were also included. BTM samples were collected in tubes containing a lyophilized bacteriostat (Exactobac-L, nerbe plus GmbH & Co. KG, Winsen/Luhe, Germany) and sent to the Institute for Parasitology, University of Veterinary Medicine Hannover, where they were centrifuged at 2000× g for 15 min. The superficial fat layer was removed, and the skimmed milk was stored at −20 °C until analysis.2.4. ELISA AnalysesSamples were run in duplicates in all ELISA analyses. For determination of anti-O. ostertagi antibodies, a commercial ELISA kit based on crude adult worm extract was used according to the manufacturer’s instructions (SVANOVIR® O. ostertagi-Ab, Boehringer Ingelheim Svanova, Uppsala, Sweden). Results from this test are expressed as optical density ratios (ODRs), with ODRs ≥ 0.5 identifying herds likely to suffer from a negative impact on herd milk yield (infection category +) and ODRs ≥ 0.8 identifying herds likely to suffer substantial production loss due to ostertagiosis (infection category ++) [17,42]. Antibodies against the F. hepatica f2 antigen were measured using a commercial ELISA test kit as described by the manufacturer (IDEXX Fasciolosis Verification Test, IDEXX GmbH, Kornwestheim, Germany). In this test, results are obtained by comparing the net extinction (NE) of the sample with the NE of the positive control, yielding a sample/positive control ratio (S/P). According to the manufacturer, test results correlate with in-herd prevalence as follows: S/P ≤ 30%: no or very low infection; 30% < S/P ≤ 80%: low infection (proportion of <20% within the herd infected) (infection category +); 80% < S/P < 150%: medium infection (20–50% infected) (infection category ++); S/P ≥ 150%: strong infection (>50% infected) (infection category +++).An in-house ELISA based on recombinant D. viviparus major sperm protein was used to assess antibodies against D. viviparus as previously published [40,43,44]. An ODR ≥ 0.41 was considered positive (infection category +) as validated for BTM by Schunn et al. [40].2.5. Data AnalysesAll collected data were entered into a central SQL-data base and variables of interest were extracted as Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) spreadsheets. Statistical analyses were conducted using SAS software (Version 9.4 for Windows, SAS Institute Inc., Cary, NC, USA) and R. v. 4.0.2 [45]. The distribution of infection categories was compared between the different regions and study years using χ2-tests (O. ostertagi, F. hepatica) or Fisher’s Exact tests (D. viviparus), respectively. In addition, Fisher’s Exact tests were used to assess breed differences in the distribution of infection categories in the data subset from the southern region. Furthermore, the influence of different predictor variables on O. ostertagi and F. hepatica seroprevalence was assessed using logistic regression, whereas seroprevalence of D. viviparus was too low to conduct a meaningful analysis. In this context, ODRs < 0.5 for O. ostertagi were interpreted as negative, and ODRs ≥ 0.5 as positive. Further, S/P-values ≤ 30% for F. hepatica were categorized as negative, and S/P-values > 30% as positive. Predictor variables were chosen based on their impact on animal health. The goal was to assess if O. ostertagi and F. hepatica seroprevalences are lower in herds with improved overall animal health and housing conditions. Before building the logistic models, normal distribution for all predictor variables was checked using the Shapiro–Wilk test. Associations between predictor variables were reviewed using Spearman’s correlation coefficient, the Kruskal–Wallis test and Cramér‘s V. The univariate models included “access to fresh grass”, “silage quality”, “anthelmintic treatment of calves and young cattle”, “anthelmintic treatment of lactating and dry cows”, “presence of fully slatted floors” (as an indicator of suboptimal housing), health indicators such as mastitis [%], metritis [%], pneumonia [%], skin lesions [%], BCS [%], the annual average milk yield per cow [kg], the cow-comfort-quotient and lameness [%] as predictor variables. Additionally, access to and drying method of hay was included in the models for F. hepatica, since metacercaria can survive the drying process as described previously [46]. The dependent variable “BCS” was defined as the proportion of animals below their optimal BCS range on the farm. Likewise, “mastitis”, “metritis”, “pneumonia”, “skin lesions” and “lameness” were defined as the proportion of animals with the condition. The variables for the multivariate models were selected using a 5% level of significance in the univariate analyses. The multivariate models were built by backward stepwise selection using the -2 Log-Likelihood (-2LL) and a significance level of 20% to estimate the goodness of fit. To assess the effect of predictors on prevalence, the odds ratio (OR) and its 95% confidence interval (CI) were primarily used, considering CIs that did not include 1 as significant, since p-values may be unreliable [47]. Because of their presumably large impact on parasite infections, “access to fresh grass”, “anthelmintic treatment of calves and young cattle” and “anthelmintic treatment of lactating and dry cows” were included in the final models regardless of their p-value.Linear regression models (LMs) were used to assess the association between BTM ELISA results and BCS as well as milk production parameters (annual average milk yield per cow [kg], milk protein content [%] and milk fat content [%]) in each study region. Models included the BTM ELISA result category for the three different parasitoses as predictor variables. For BCS, the models contained breed type (HD, DP or mixed) and number of scored animals as further variables. For milk production parameters, breed type, average lactation number, average SCC, average number of lactating animals, and farm type (“conventional”, “organic”, “in transition from conventional to organic”) were included as potential confounders. Due to the expected non-linear relationship, a quadratic term was included for lactation number. As a significant correlation of milk yield with milk fat and protein content was observed, milk yield was included as a potential confounder in these models. Two-way interactions were considered and retained in models if statistically significant (p ≤ 0.05). For LM validation, the distribution and homogeneity of model residuals was checked graphically. Cook’s distance and residuals-vs.-leverage plots were used to identify potentially influential observations; these were dropped if their exclusion affected model estimates. Final models were compared to null models containing only an intercept term (R function “anova”).3. Results3.1. Regional and Annual Patterns of Seroprevalence Of the 765 participating farms, 646 contributed BTM samples, including 201 from the North, 204 from the East and 241 from the South. Overall, O. ostertagi antibody levels indicative of production losses (ODRs ≥ 0.5) were detected in 41.2% (266/646; 95% CI: 37.4–45.1%) of farms, with ODRs between 0.5 and 0.8 in 29.7% (192/646; 95% CI: 26.3–33.4%) and ODRs ≥ 0.8 in 11.5% (74/646; 95% CI: 9.2–14.2%) of farms. Across all samples, the average O. ostertagi ODR value was 0.48 (standard deviation [SD]: 0.24). Significant regional differences in the distribution of infection categories were observed (χ2-test, χ2 = 15.6, Df = 4, p = 0.004), with the highest proportion of farms with ODRs ≥ 0.5 in the northern region (49.3%, 99/201, 95% CI: 42.2–56.4%), followed by the South with 39.4% (95/241, 95% CI: 33.3–45.9%) and the East with 35.3% (72/294, 95% CI: 28.8–42.3%) (Figure 1). In contrast, no significant differences were observed between the study years (χ2-test, χ2 = 5.4, Df = 4, p = 0.245; Table 1).Regarding F. hepatica, the overall seroprevalence in BTM samples (S/P ≥ 30%) amounted to 14.9% (96/646; 95% CI: 12.2–17.9%), with approximately equal proportions of farms displaying titers indicative of a low (4.6%; 30/646; 95% CI: 3.2–6.6%), medium (5.0%; 32/646; 95% CI: 3.5–7.0%) and high (5.3%; 34/646; 95% CI: 3.7–7.4%) in-herd prevalence. Significant regional differences were observed (χ2-test, χ2 = 51.8, Df = 6, p < 0.001), with the highest seroprevalence recorded in the South (24.9%, 60/241, 95% CI: 19.7–30.9%), followed by the North (16.9%, 34/201, 95% CI: 12.2–23.0%), while seroprevalence in the East was low (1.0%, 2/204, 95% CI: 0.2–3.9%, Figure 1). Comparison between the different sampling years revealed no significant differences (χ2-test, χ2 = 5.9, Df = 6, p = 0.434; Table 1). Overall D. viviparus BTM seroprevalence amounted to 2.3% (15/646; 95% CI: 1.4–3.9%), with significant regional differences (Fisher’s Exact test, p < 0.001). The highest seroprevalence was observed in the North (4.5%, 9/201, 95% CI: 2.2–8.6%), followed by the East (2.5%, 5/204, 95% CI: 0.9–5.9%), while seroprevalence in the South was low (0.4%, 1/241, 95% CI: 0.02–2.7%). Similar to the other parasitoses, no significant annual differences were detected (Fisher’s Exact test, p = 0.347; Table 1). Regarding co-exposure rates, all D. viviparus-positive samples also displayed an O. ostertagi ODR ≥ 0.5. Furthermore, 85 samples (13.2%, 95% CI: 10.7–16.1%) showed both an O. ostertagi ODR of ≥ 0.5 and a positive F. hepatica result, with five farms (0.8%, 95% CI: 0.3–1.9%) being additionally D. viviparus-positive. Only the co-exposure rate of O. ostertagi/F. hepatica differed significantly from the numerically expected rate of 6.1% (χ2 = 17.1, df = 1, p < 0.001), indicating a positive association, while no significant differences between the observed and expected values were determined for the other combinations. Patterns of co-exposure in the different sampling regions are shown in Table 2. 3.2. Breed DifferencesHigh-performance dairy cattle breeds dominated in 64.2% (415/646) of farms, whereas 29.1% (188/646) of farms had predominantly DP cattle. A further 5.9% (38/646) of farms had mixed herds, whereas breed information was insufficient for the remaining 0.8% (5/646). In the northern and eastern parts of Germany, 97.5% (195/200) and 92.6% (187/202) of farms were assigned to the HD category, respectively. In contrast, 76.6% (183/239) of farms in the South had DP herds, but only 13.8% (33/239) had HD and 9.6% (23/239) had mixed herds. In the South, significant differences in antibody level category distribution were observed regarding O. ostertagi and F. hepatica, with lower prevalence in DP than HD herds (Fisher-Exact tests, p = 0.025 and p < 0.001, respectively; Figure 2). Differences were not assessed for D. viviparus due to the low prevalence in this region. 3.3. Association of Seroprevalence with Management FactorsSubstantial regional differences in herd management were observed for the factors of access to fresh grass, access to hay and anthelmintic treatment (Supplementary Table S1). While 47.8% of herds had access to fresh grass in the North, this applied to only 24.9% of herds in the East and 32.0% in the South. In contrast, less than 10% of farms included hay in rations in the North and East, whereas more than 20% of farms fed hay in the South. Spoiled silages were found in 50.8% of farms in the North, 29.8% of farms in the East and 38.6% of farms in the South. Regarding anthelmintic treatment, 73.0% and 50.8% of farms in the North regularly treated calves/young cattle and dry/lactating cows, respectively, as compared to only 38.5% and 23.1% in the East and 30.5% and 19.5% in the South. The following variables were included in the final multivariate logistic regression models for O. ostertagi after exploratory univariate analyses: access to fresh grass, silage quality (North only), anthelmintic treatment, lameness (South only) and average annual milk yield per cow (Table 3). Access to fresh grass significantly increased the odds of a BTM ELISA result ≥0.5 in all three regions, with ORs of 5.5 (95% CI: 2.5–12.1) in the North, 3.7 (95% CI: 1.8–7.9) in the East and 7.0 (95% CI: 3.1–15.4) in the South. Furthermore, the use of spoiled silages increased the odds for a positive result by 3.0 (95% CI: 1.4–6.5) in the North. Anthelmintic treatment of young cattle was positively associated with O. ostertagi BTM seroprevalence in the North and in the South, while treatment of both calves and young cattle was positively associated with O. ostertagi status in the South only (Table 3). The association of seropositivity with average annual milk yield per cow was significant in all three regions. Therefore, the impact of O. ostertagi exposure on milk production was further investigated in detail using linear regression (see below).With regard to F. hepatica, access to fresh grass, access to hay (South only), anthelmintic treatment, lameness (South only) and annual average milk yield were included in the final multivariate models. Risk factors for F. hepatica could not be identified in the North and were not investigated in the East due to low prevalence. In the South, however, access to fresh grass, access to hay and the treatment of calves and young cattle with anthelmintics were positively associated with F. hepatica seropositivity, whereas a negative relationship was detected regarding the average annual milk yield per cow as well as lameness (Table 4). 3.4. Association of Seropositivity Categories with BCS and Herd Productivity Parameters In all three regions, the proportion of cows displaying a suboptimal BCS was significantly associated with BTM ELISA results (Table 5, Figure 3). In the North and the South, this proportion was significantly increased on farms with both positive O. ostertagi (category ++) and F. hepatica antibody levels (category ++) compared to seronegative farms, respectively. In addition, the proportion of cows displaying a suboptimal BCS was also significantly increased in herds with an O. ostertagi category ++ and a F. hepatica category + antibody level in the South. In the East, where almost no farms were F. hepatica-seropositive, a significantly higher proportion of BCS-suboptimal cows was found on farms with an O. ostertagi ODR ≥ 0.8 (category ++). However, the effect of infection on BCS was not consistent across all antibody level categories, as a significantly lower proportion of suboptimal cows was present on farms with a F. hepatica (category +) status in the South and O. ostertagi (+)/F. hepatica (+++) status in the North, as compared to seronegative farms. Additionally, breed type was significantly associated with BCS, with a lower proportion of suboptimal cows in DP herds (Table 5). However, no indication of a breed difference regarding the effect of parasite exposure on BCS was found (data not shown). Finally, the number of scored cows had a small but significant negative effect regarding the proportion of suboptimal cows.Herd productivity parameters were available for 596 farms (192 from the North, 196 from the East, 208 from the South). A significant negative impact of F. hepatica seropositivity on milk yield was detected in the northern region, with an estimated annual reduction of 1129.2–1335.3 kg milk/cow in seropositive vs. seronegative herds, depending on seropositivity category (Table 6). In the East, where F. hepatica seroprevalence was low, a significant negative effect of O. ostertagi seropositivity was apparent, with an estimated reduction in annual milk yield of 832.7 kg/cow in herds with an ODR ≥ 0.8 (category ++). As the model for the southern part of Germany indicated a significant interaction of breed type and F. hepatica BTM result (Table 6), separate models were additionally calculated for HD and DP herds. A significant negative impact of F. hepatica seropositivity (infection category +++) on milk yield was only detected in HD, but not in DP herds, which instead displayed a significantly higher milk yield in the F. hepatica category +++ (Supplementary Table S2). Further factors significantly associated with milk yield were breed type, lactation number, SCC, number of lactating animals and farm type (Table 6, Supplementary Table S2). Regarding D. viviparus, no significant impact on milk yield was detected in northern and eastern Germany. The dataset from the South, for which milk production data were available, contained no D. viviparus-positive herds.Regarding average milk protein content (%), the number of producing animals had no significant effect on milk protein and removing this variable significantly improved model fit. Likewise, removing the variable “milk yield” from the model for the North led to an improved model. Only farm type had a significant effect in all three datasets, while significant differences between breed types were noted in the East and in the South (Table 7). A significant association of BTM results with milk protein was only detected regarding F. hepatica category ++ in the North, with a positive effect (Table 7). For the southern region, separate models were again calculated for HD and DP herds, and only the HD model suggested a negative association of F. hepatica infection category +++ with protein content (Supplementary Table S2), however, the model was not significantly different from a null model and thus needs to be considered of limited reliability. Average milk fat content (%) was significantly negatively associated with milk yield in all three regions (Table 8). In contrast, a significant association with O. ostertagi BTM ELISA results was only noted in the South, where herds with an ODR ≥ 0.8 had significantly lower average fat content (Table 8). This was primarily driven by DP herds, as the dataset did not contain any HD herds in this antibody level category (Supplementary Table S1).4. DiscussionThe aim of the current study was to provide an up-to-date estimate of dairy cow exposure to O. ostertagi, F. hepatica and D. viviparus in three parts of Germany, which show considerable structural differences in dairy farming, and to assess relationships with management factors as well as production parameters. The overall mean O. ostertagi ODR of 0.48 is similar to previous investigations and can be considered intermediate in comparison to other European countries, where mean ODRs as low as 0.31 and as high as 0.83 have been measured [17,18]. In contrast to Fanke et al. [19], who reported no significant regional differences, the current study demonstrated a higher level of O. ostertagi exposure in the northern as compared to the eastern and the southern study areas. This can be explained by the high proportion of farms which provided access to fresh grass in this area. In all three datasets, access to fresh grass was significantly associated with O. ostertagi exposure, in line with the results of previous studies [17,48,49]. Furthermore, the presence of spoiled silage on the farm increased the odds for O. ostertagi BTM ODRs ≥ 0.5 in the North. Possibly, spoiled silage may negatively affect the overall health status of cows and their susceptibility towards GIN infections. In addition, anthelmintic treatment of young cattle was positively associated with O. ostertagi seropositivity in the North and the South, probably because farms with known GIN problems increase their use of anthelmintics. Frequent use of anthelmintics in young animals, however, impairs the development of an efficacious immune response, rendering them more susceptible to O. ostertagi as adults [50].The overall F. hepatica seroprevalence of 14.9% determined in the current study is lower than previously determined values for German dairy herds of 23.6% [20] and 22.9% [19]. This is mainly driven by a lower F. hepatica seroprevalence (16.9%) in northern Germany as compared to previous investigations. In samples collected in 2008, Kuerpick et al. [20] determined F. hepatica BTM seroprevalences of 29.4–38.4% in those federal states, which were included in the northern region in the present study. Indeed, a spatial model of fasciolosis in dairy cattle, including meteorological factors, identified the northwestern part of Germany as a high-risk area, but indicated a low risk in eastern and southern parts of Germany [51]. However, 2018 and 2019 were exceptionally dry years in central Europe [52], which may have led to unfavorable conditions for the snail intermediate host of F. hepatica, which depends on moist habitats. Thus, the decline in F. hepatica seroprevalence may be related to climate change. In contrast, May et al. [11] observed high F. hepatica seroprevalences of 33.1–37.0% during 2017–2018 in East Frisia, a coastal area of Lower Saxony, employing the same serological test as the current study. However, the landscape of East Frisia is characterized by coastal marshlands traversed by drainage channels and was probably less affected by the dry weather. In addition, farmers received feedback regarding the BTM ELISA results of the previous studies, so the decline in seroprevalence observed in the present study may be due to effective F. hepatica control measures [11]. Furthermore, the geographical distribution of the farms within the northern region may be responsible for the observed discrepancies, or the study design may have created a bias since it was based on voluntary participation by farmers. In cases of known problems with fasciolosis on the farm, farmers may have chosen not to participate. Furthermore, it is possible that the increased use of anthelmintics in calves and young cattle as well as in lactating and dry cows in the northern region in the current study led to a lower prevalence.In contrast, the low F. hepatica seroprevalence of 1% in the East is in accordance with previous observations [20]. For southern Germany, a heterogeneous distribution of F. hepatica has been described, with high levels of exposure in Alpine regions, but low levels in the Bavarian plains [53]. The overall seroprevalence of 24.9% is similar to the value of 32.2% determined by Koch in 2005 [53]; however, the data are not completely comparable due to different cut-offs used to define seropositivity. Interestingly, a significantly lower seroprevalence of F. hepatica as well as O. ostertagi was determined in DP as compared to HP dairy herds in the South. This might be driven by differences in animal husbandry. Alternatively, DP breeds might have higher liver fluke resistance than HP dairy breeds, which have often been selected for indoor production systems. May et al. [24] determined a higher level of resistance against GINs in two Holstein-Friesian lines selected for grazing systems as compared to lines selected for indoor production. However, similar studies comparing Holstein-Friesian to German DP breeds with regard to liver flukes are not available so far. Previous research in temperate climatic zones has shown that management factors, such as treatment with flukicides, may have a larger impact on herd infection status with F. hepatica than climatic variables [42,43]. In the present study, anthelmintic treatment in calves and young cattle was positively related to F. hepatica exposure in Bavaria, probably because farms with known fasciolosis problems practice frequent treatment. Furthermore, access to fresh grass increased the odds of a positive F. hepatica BTM ELISA result, in line with previous studies [49]. This demonstrates that pasture and feeding management remain key factors in decreasing parasite exposure. Since metacercaria of F. hepatica can survive in hay for up to six months, depending on the degree of desiccation [46], access to hay was included in the analysis. A significant impact of this variable was only found in the dataset from the South. In this part of Germany, the proportion of farms feeding hay was larger than in the other parts, whereas less farms provided access to fresh grass. Finally, the percentage of cows with lameness was negatively associated with herd seropositivity in the South. Further studies are needed to shed light on the underlying reasons and possible confounding factors, especially since the occurrence of lameness was distributed evenly in all three regions (data not shown). Seroprevalence values for D. viviparus among German dairy herds determined in the year 2008, using the same serologic test as the current study, ranged from 0.0–31.2% in different regions, with an overall value of 17.1% [21]. In comparison to the dataset from 2008, both the overall D. viviparus seroprevalence of 2.3%, as well as the regional seroprevalence values determined in the current study were surprisingly low, as rates of 10.4–31.2% were determined by Schunn et al. [21] in the regions relevant for the current study. As mentioned above, with regard to F. hepatica, farmers were informed about the results of these studies, and may have changed their management practices, resulting in the observed prevalence decline. Unfortunately, D. viviparus herd seroprevalence in Germany has not been assessed between 2008 and 2017, hampering the interpretation of our data. Alternatively, the decline in seroprevalence may reflect endemic stability, as the magnitude and duration of the antibody response are reduced in repeatedly infected cows, resulting in lower BTM ODRs [54]. Therefore, further studies are necessary to determine whether D. viviparus may be on the decline in Germany, e.g., due to increasingly dry and warm summers as observed especially in the years 2018 and 2019 [52], or whether other factors are responsible for the low seroprevalence in the current study. As mentioned regarding F. hepatica, the study design may have caused a bias, as farms with a dictyocaulosis history may have been less likely to participate. Given the low seroprevalence of D. viviparus, risk factor analysis was not possible in the present study. In previous studies, herd size, mowing of pastures and length of the grazing period were associated with D. viviparus antibody status [15,55].The present study determined a significant association of parasite exposure with health indicators and cow productivity. O. ostertagi and F. hepatica exposure were significantly associated with a low herd-level body condition in all three datasets. In the North as well as the South, herds with evidence of co-exposure to both parasites had a significantly higher proportion of cows in suboptimal condition. In the East, where F. hepatica seroprevalence was very low, a significantly higher proportion of thin cows was found on farms with an O. ostertagi BTM ODR ≥ 0.8. Similarly, Ostertagia spp. and F. hepatica infected beef cattle displayed lower carcass weights than helminth-free animals in the United Kingdom [3]. Additionally, liver fluke infections in dairy cows have been associated with higher levels of beta-hydroxybutyrate in milk, indicating a negative energy balance [11]. Therefore, the current study contributes to the evidence that helminth infections do not only lead to subclinical impacts on dairy cow productivity, but may also contribute to a low body condition, which entails further health risks [56]. In line with previous studies, a negative association between O. ostertagi and F. hepatica exposure and milk yield was demonstrated in the current analysis, although not to the same extent in all regions. Similar to the pattern regarding BCS, a significant association of O. ostertagi exposure and reduced milk yield was only detected in the East, where a loss of 832.74 kg/milk per cow and year was estimated in herds with a BTM ODR ≥ 0.8. This value is similar to an estimated loss of 975 litres of milk per cow and year in O. ostertagi-seropositive herds in a previous German study [19]. In the other study areas, F. hepatica rather than O. ostertagi exposure was significantly associated with a reduced milk yield, with estimated losses of more than 1000 kg milk/cow and year. Interestingly, breed differences seem to play an important role regarding regional differences. In the South, where DP breeds (mainly German Simmental) are common, a negative association between milk yield and F. hepatica exposure was only detected in herds consisting of HD breeds. Therefore, DP breeds not only had lower parasite seroprevalences in the current study, but also seem to be more resilient towards negative effects of parasite infection on milk yield. Similar differences in resilience exist among sheep and goat breeds with a moderately heritable genetic basis, and have been suggested for cattle [57,58]. These breeds may thus be especially suited for organic farming, where “robust” breeds are preferred due to restricted drug use, including anthelmintics [57]. However, DP herds in the South showed a significantly lower milk fat content associated with O. ostertagi seropositivity, while this was not observed in HD breeds in the current study, nor in previous investigations [5,7]. Therefore, resilience may be limited to milk yield, whereas GIN infection may adversely affect milk fat content in DP breeds, which should be confirmed in further studies, preferably on an individual animal basis. A drop in milk fat content due to F. hepatica infections as observed in previous studies [9,11,59] was not noted in the present dataset. Furthermore, no significant association of parasite seroprevalence with milk protein content was found, except for a 0.1% higher average protein content in F. hepatica category ++ vs. seronegative herds in the North. Most previous studies found either no [8,59] or a significant negative association between F. hepatica infections and milk protein content [9,11]. Therefore, the estimated increase in milk protein in this group of cattle may be due to an unknown confounding factor. 5. ConclusionsThe current study indicated that O. ostertagi and F. hepatica exposure among dairy herds seems to be rather stable in Germany, while further studies are needed to assess whether D. viviparus is indeed on the decline. Management factors related to parasite exposure included access to fresh grass and hay, silage quality and anthelmintic treatment, with some regional differences related to variation in animal husbandry. The significant differences between cattle breed types with regard to parasite seroprevalence as well as impacts on production parameters represent a novel finding. Lower prevalence of O. ostertagi and F. hepatica in dual-purpose herds and their increased resilience in terms of unaffected milk yield indicate that this breed type may be especially suited for pasture-based or organic settings, if these findings are confirmed in further studies.

animals : an open access journal from mdpi

10.3390/ani11030843

PMC8002547

Twinning in dairy cattle is caused by many different factors, both genetic (i.e., inherited) and non-genetic (i.e., animal management). In dairy operations, twinning is an undesirable trait associated with other reproductive and metabolic diseases, higher operational costs, and higher rates of culling on farm. The animal welfare and economic impacts have resulted in the development of a genomic prediction for twinning (i.e., TWIN) by Zoetis such that producers can make informed breeding decisions for breeding Holstein females that are less likely to become pregnant with twins in a given lactation. This prediction is included in a holistic breeding tool (i.e., selection index) for producers so that they can improve multiple health, fertility, and production traits in parallel with reducing twinning when making breeding decisions for future generations. The objectives of the present study were (1) to describe how the twinning prediction was developed (and included in a selection index), (2) show that the prediction works effectively using real life farm data, and (3) propose how this genetic tool can be used in collaboration with management practices to proactively reduce twin pregnancies on farm. The results of this study provide evidence that twinning can be proactively managed on dairy farms using genetically powered tools.

Twinning is a multifactorial trait influenced by both genetic and environmental factors that can negatively impact animal welfare and economic sustainability on commercial dairy operations. To date, using genetic selection as a tool for reducing twinning rates on commercial dairies has been proposed, but not yet implemented. In response to this market need, Zoetis (Kalamazoo, MI, USA) has developed a genomic prediction for twin pregnancies, and included it in a comprehensive multitrait selection index. The objectives of this study were to (1) describe a genetic evaluation for twinning in Holstein cattle, (2) demonstrate the efficacy of the predictions, (3) propose strategies to reduce twin pregnancies using this information. Data were retrieved from commercial dairies and provided directly by producers upon obtaining their permission. The twin pregnancies trait (TWIN) was defined as a pregnancy resulting in birth or abortion of twin calves, classified as a binary (0,1) event, and analysed using a threshold animal model. Predictions for a subset of cows were compared to their on-farm twin records. The heritability for twin pregnancies was 0.088, and genomic predicted transmitting abilities ((g)PTAs) ranged from −7.45–20.79. Genetic correlations between TWIN and other traits were low, meaning that improvement for TWIN will not negatively impact improvement for other traits. TWIN was effectively demonstrated to identify cows most and least likely to experience a twin pregnancy in a given lactation, regardless of reproductive protocol used. Effective inclusion of the prediction in a multitrait selection index offers producers a comprehensive tool to inform selection and management decisions. When combined with sound management practices, this presents a compelling opportunity for dairy producers to proactively reduce the incidence of twin pregnancies on commercial dairy operations.

1. IntroductionTwinning in cattle is a complex trait that can be influenced by a multitude of factors such as milk production, season, breed, parity and ovulation rate, previous twin calving events, pharmaceutical use, and genetics [1,2,3,4,5,6,7,8]. Although generally perceived to be a positive reproductive outcome in the beef industry (due to increased operational efficiencies in terms of increased weaned calf weight per dam [9,10]), twinning is a negative reproductive outcome for dairy operations, as it adversely affects cow and calf health and welfare, and ultimately reduces herd profitability [11,12]. Despite cattle being a monotocous species, twinning rates in cattle range from <1% in beef populations up to 5.8% in dairy populations, with appreciably high twinning rates observed in the Holstein breed, relative to other cattle breeds [12,13,14,15,16]. The deleterious health effects of twinning are numerous and include increased incidences of mortality, stillbirths and abortions, and reduced birth weights for calves [2,17,18,19]. For cows, higher incidences of periparturient diseases such as retained placenta, metritis, displaced abomasum, and ketosis are associated with twin-bearing cows; this is all in addition to a reduced overall productive lifespan for these cows [2,20,21,22,23]. Moreover, twinning also negatively impacts reproductive performance of cows and calves alike, resulting in increased days open, increased services per conception, cystic ovarian disease and dystocia for cows, and freemartinism in most heifer calves twinned to bull calves [1,2,4,21,24,25]. Taking this myriad of detrimental effects into account, twinning increases veterinary, culling and replacement heifer costs on farm, thus eroding profit margins and impacting operational sustainability. Previous estimations of the economic ramifications of twinning range from $97–$225 per twin pregnancy on farm, with an overall estimated annual impact on US dairy profitability ranging between $22.5–$112.5 million [16,26,27,28,29]. These health, welfare and economic challenges imposed by twinning on the industry are projected to rise, as selecting for higher milk output has previously been associated with higher rates of twinning, and the aim of dairy operations is to maximize milk production output per cow [1,2].Once diagnosed, current intervention and management practices for twin pregnancies include (1) GnRH treatment to maintain the gestation, (2) induced embryo reduction through aspiration of the allanto-amniotic fluid or manual rupture, (3) induced abortion using PGF2α, and (4) culling. Limitations of these methods include an increased likelihood of pregnancy loss in the embryo that was not selected for manual reduction, and in the case of pregnancy maintenance, the aforementioned periparturient diseases and abortion, stillbirth, calf mortality, and dystocia [11,12,27,30,31]. Opinions differ on what is the optimal way to manage females carrying twin calves, with some advocating manual embryo reduction [27], and others suggesting selective culling and selective maintenance of pregnancy, depending on the value of the cow and calf(ves) as the preferred alternative [12]. Since prior twin pregnancy diagnosis/calving events is a risk factor for future twinning events, these methods are somewhat short-sighted and do not help reduce the prevailing trend in twin pregnancies in the industry in a proactive manner. Importantly, regardless of which management decision is ultimately chosen, all of these methods are reactive in nature. Alternative, proactive solutions, such as those that use genetics and genomics to predict the likelihood of females to be diagnosed with twins in a given lactation are warranted, as the trend toward increased twinning in the dairy cattle population is ever increasing, given the industry selects for higher-output cattle [12].There is an established historical and ongoing effort by research groups across the globe to quantify the genetic component of twinning in cattle. Notably, since the 1980′s the USDA Meat Animal Research Center (USDA-MARC) has quantified the genetic parameters of, and identified areas in the bovine genome associated with twinning in cattle, in addition to demonstrating the effectiveness of selective breeding for twinning in cattle based on their observed twinning rates [7,8,21,32]. Twinning rates in this research herd have exceeded 50%, and a proportion of the foundation herd was reported to be comprised of Holstein genetics (with just under one third of the genetic makeup of the population tracing back to the Holstein) [7,33]. Heritability estimates reported globally for twinning in dairy cattle range from 0.004–0.29, and vary depending on the extent of variability (residual and genetic) in the populations sampled, sample sizes, and model used (i.e., linear, threshold, animal, sire) [7,8,13,16,29,32,34,35,36,37,38,39,40,41]. These non-zero estimates, alongside the reported genetic variation in twinning highlights that there is ample opportunity for selection and breeding decisions to reduce the frequency of twinning events in the dairy population. Quantitative trait loci (QTLs, i.e., regions in the genome) previously identified as being associated with twinning in cattle further substantiates the existence of a genetic and genomic component to twinning, with QTLs identified on Bos taurus autosome 1, 5, 6, 7, 8, 10, 14, 15, 19, 23, and 24 [39,41,42,43,44,45,46,47,48,49,50,51].Recent advancements in implementing hormonal synchronization protocols which proactively reduce the incidence of double ovulation in dairy cattle has shown to be an effective non-genetic (i.e., environmental) tool in this space [30]. However, in spite of the relatively large body of work investigating the genetics(omics) underpinning twinning, Fricke [12] stated that “at present, dairy herders and their consultants are ill prepared to make sound management decisions to mitigate the negative effects of twinning on their operations because of a lack of basic and applied scientific data on twinning in dairy cattle”. This statement has largely held true from a genomics perspective, due to the relative lack of industry-available predictive tools for producers to proactively reduce twinning rates on their dairy operations globally. To date, there have been a number of studies that conducted genetic(omic) evaluations for twinning rate in cattle [16,35,37,39,40,52], however these evaluations have not been translated in to predictions for twinning that producers in the industry can use to proactively select for cows less likely to experience a twin pregnancy in a given lactation. Using genetic selection as a tool to reduce on-farm twinning rates has been researched, yet, so far, has not been implemented. Genomic predictions for twinning events can provide the foundation needed to make proactive and informed breeding and management decisions to reduce their frequency, provided they are included in a comprehensive, economically driven multitrait selection index tool.Harnessing such genetic- and genomically powered tools, in tandem with practical and strategic breeding and management strategies can holistically offer producers the ability to proactively reduce the incidence of twinning in their herds, helping improve animal wellbeing, economic outcomes [53,54], and peace of mind. Selection indices are tools that condense information on the genetic merit of animals across multiple traits in to one value, allowing a producer to easily rank animals and subsequently create breeding and management strategies to maximize genetic gain in the herd toward their selection goal(s) [55,56]. Historically, the goal of selection indices was to focus solely on improving production traits [57,58], to the detriment of cow fertility and health traits arising from the antagonistic relationships that exist between these traits [54,59]. The appreciation of these antagonistic relationships, along with the shift to not only focusing on traits that increase on-farm profit, but those that decrease costs of production has resulted in the inclusion of these non-production traits in multitrait selection indices globally [56,60].In response to industry needs for genetic improvement of dairy wellness traits, Zoetis Genetics developed an industry-available genetic and genomic evaluation for wellness traits in Holstein cattle in 2016, known as CLARIFIDE® Plus (Zoetis Genetics, Kalamazoo, MI, USA). CLARIFIDE® Plus provides producers with a comprehensive suite of genomic predictions, plus a multitrait selection index tool, the Dairy Wellness ProfitTM (DWP$) index to facilitate sound management and selection decisions based on the genetic predisposition of their heifer calves for these traits that will be expressed later in life. DWP$ is an economic multitrait selection index that was formulated to estimate the potential lifetime profitability an animal would generate under US dairy economic conditions and includes cow and calf wellness, production, fertility, functional type, longevity, livability, calving ability, and milk quality traits, as well as polled test results [56,61]. In 2018 and 2020, CLARIFIDE® Plus and DWP$ were updated to include additional traits shown to impact the lifetime profitability of a dairy animal; one of those traits included in the 2020 update was twinning (i.e., TWIN). The objective of the present study was to (1) describe a genetic and genomic evaluation for twinning in Holstein cattle (including how this information is incorporated into a comprehensive and easy to interpret multitrait selection index), (2) demonstrate the ability of the evaluation’s predictions to accurately predict twinning incidences (regardless of on-farm reproductive protocol used), and (3) propose practical ways to holistically apply this information to offer a proactive solution that producers can implement to reduce the incidence of twinning in Holstein dairy herds.2. Materials and Methods2.1. Data Sources for Genetic EvaluationData was available from 276 herds located in 26 states in the US and collected approximately from 1990 to date. Each herd supplied, on average 22,998 phenotypic records. Data were obtained directly from producers upon obtaining their signed permissions; the herds were not routinely monitored or compensated for event recording by Zoetis. Production, reproduction, and pedigree information was extracted from on-farm herd management software backup files using proprietary scripts. Genotypes were obtained from the Zoetis Genotyping Lab (Zoetis Genetics, Kalamazoo, MI, USA). Animals were genotyped with a variety of low-density SNP chips with a number of SNPs ranging from about 3000 to over 35,000, and several types of medium-density chips with 50,000–80,000 SNPs. All animals genotyped on chips with <40,000 SNPs were imputed using the program FImpute [62] to 45,245 SNPs used in the genomic evaluation. 2.2. Data Editing and Genetic Trait Definition for Genetic EvaluationInitial data editing included checking animal identification for accuracy and consistency across data files, using similar criteria as described in Norman et al. (1994) [63]. Each animal was required to have a lactation record with a valid calving date and a lactation number as well as a calving interval between 250 and 999 days. Phenotype records were checked against the pedigree and all animals found to be male in the pedigree file or having a calving date preceding their birth date were removed.The genetic trait of twin pregnancies (i.e., TWIN) was defined as a pregnancy resulting in birth or abortion of twin calves (alive or dead). TWIN was treated as a binary event-having a value of one if the cow was recorded as carrying or giving birth to twins and zero otherwise. Records of the same cow without TWIN recorded, as well as records of all herdmates without recorded twins, were added as ‘singleton pregnancy’ records (i.e., ‘0’). Twinning information was extracted from a combination of the events and remarks recorded on the herd management software. Terminology used to record TWIN events varied across the herds, which were standardized and collapsed into the binary TWIN outcome as shown Table 1.Further, each herd by year and season of calving (HYS) group was required to have a minimum of 20 records and at least one recorded twinning event. HYS groups that did not meet these criteria were omitted from analysis, as it was assumed that the herd did not record twinning at all, or did not record it during that time period.2.3. Statistical Models for Genetic EvaluationThe analyses for the present study were conducted largely as described by Gonzalez-Peña et al. (2020) [64]; the following threshold animal model with repeated observations was used to conduct the analysis:λ = Xβ + Zhh + Zaa + Zpp + e(1) where λ represents a vector of unobserved liabilities to twinning; β is the vector of fixed parity effects, with the corresponding incidence matrix X; parities 1, 2, 3, 4, and ≥5 were considered; h is the random herd-year-season effect, where h~N(0,Iσh2), with the variance σh2. Four seasons were defined within each calving year: Winter (December–February), Spring (March–May), Summer (June–August), and Fall (September–November); a is the random animal effect with a~N(0,Hσa2), where σa2 is the additive genetic variance and H is the pedigree relationship matrix augmented using genotypes; p is the random effect of permanent environment with p~N(0,Iσp2), with σp2 is the permanent environment variance and e is the residual, where e~N(0,I).Variance components were estimated using the same model, but without genotypes using the program THRGIBBS1F90 version 2.108 from the BLUPF90 family [65]. Genetic evaluation was performed using the programs from the BLUPF90 family [66]. A univariate threshold model based on single step genomic BLUP methodology (ssGBLUP) was applied. The inverse of the traditional pedigree relationship matrix, 𝐴−1, was replaced by the inverse of H matrix that combines pedigree and genomic relationships [67,68]:(2)H−1=A−1 + 000G−1−A22−1 where 𝐴−1 is an inverse of the pedigree relationship matrix, 𝐺−1 is an inverse of the genomic relationship matrix and 𝐴22−1 is an inverse of the pedigree-relationship matrix for genotyped animals only. The ‘algorithm for proven and young animals’ (APY) developed by Ignacy Misztal’s group from the University of Georgia in Athens (UGA, Georgia) was applied. The program CBLUP90IOD2 version 3.39 was used to obtain genomic breeding values using preconditioned conjugate gradient (PCG) with the number of rounds set to 200. The core consisted of 25,000 randomly selected animals. The genomic matrix conditioning parameters tau and omega were set to 1.0. Inbreeding was considered when constructing the pedigree relationship matrix. The reliabilities of estimated breeding values (EBVs) were obtained with the program ACCF90GS version 2.54, which approximates reliabilities using contribution from genotypes, phenotypes, and pedigree. To reduce computational requirements, the contribution from genotypes is replaced by the value of the diagonal of the G matrix, gii. Reliabilities for genotyped animals were approximated as per Gonzales-Peña et al. (2020) using the guidance of Daniela Lourenco, University of Georgia, Athens, personal communication, 2016 [69].The solutions from the CBLUP90IOD2 program (raw EBVs) were transformed into probabilities of exceeding the threshold value, with threshold values being calculated from the current data. For each animal solution, we calculated probability that a standard normal variable with the mean equal to this solution and the variance of one exceeds the threshold. The probabilities were then multiplied by 100 (to represent percent), divided by 2 (to obtain predicted transmitting abilities (PTAs)), and expressed as the deviation from the average PTA of all animals (genotyped or not (i.e., having a gPTA or a PTA)) born in 2015 with a phenotypic record for TWIN. The resulting (g)PTAs may be interpreted as differences of the individual animal’s risk of having twins from the average (base) risk. Higher values of (g)PTAs indicate a higher chance of twinning. For example, an animal with a TWIN (g)PTA of −2.0 has a 2% lower risk, whereas an animal with a TWIN (g)PTA of 2.0 has a 2% higher risk of having twins in a given lactation than the base animal [64]. For ease of interpretation, (g)PTAs were transformed into standardized transmitting abilities (STAs), with a mean of 100, a standard deviation of 5, and the reversed sign (so that higher values represent lower risk of TWIN) as per McNeel et al. (2017) [54,64]. Correlations between TWIN STAs and other traits (Zoetis Wellness trait STAs and trait (g)PTAs in the US genetic evaluation (Council of Dairy Cattle Breeding)) were estimated using product-moment (Pearson) correlations, similar to Gonzalez-Peña et al. (2020) [64].2.4. Inclusion of Twinning Prediction in a Multitrait Selection IndexAn in-depth description of the development of DWP$ is detailed in Fessenden et al. (2020) [56]. Briefly for the purposes of this study, (1) the TWIN STA, its phenotypic correlations, and genetic relationships with other traits were estimated, (2) the economic value of the TWIN trait as it relates to its contributions to a dairy animal’s lifetime profitability was estimated (by determining the economic value of all incomes and losses for a 12-unit increase in the TWIN trait), and (3) the STA for TWIN is multiplied by its corresponding economic weight, alongside all other traits in the DWP$ index, which are then summed together to determine an animal’s overall selection index value [56,70].2.5. Demonstration of Evaluation EfficacyTo demonstrate the efficacy of the TWIN predictions generated by the evaluation, a subset of 8,219 females from 5 US herds that used commercial genomic tests were selected, and the relationship between the TWIN phenotype and genetic(omic) merit for TWIN investigated for each herd. Criteria for inclusion in these demonstration cohorts included (1) the animals/herds did not contribute phenotypes to the genetic evaluation so as to not bias results, (2) the herds demonstrated adequate recording of TWIN events on a per-animal basis, and (3) a single reproductive protocol was required to be used consistently on farm (n = 3, defined as: no synchronization protocol, presynch ovsynch protocol, or Double Ovsynch protocol). Table 7 outlines the overview 5 herds used in the demonstration cohort.Once selected, TWIN STAs for these animals were adjusted to reflect what their STA values would have been in 2013. This measure was included to reduce bias in the TWIN estimate predictions such that the predictions closely reflect what the producer would have received before the animals entered the milking string. This measure reduces the bias that could arise from having 2014–2020 records (pedigree and performance) feeding into the evaluation. Females were then ranked based on TWIN STA predictions within each herd, and allocated to one of three genetic groups (i.e., tertiles): worst 33%, 34–66%, and best 33%, similar to what has been reported by others [54,56,71]. Data analysis was generated using SAS software (version 9.4, SAS Institute Inc., Cary, NC, USA). For this analysis, differences between genetic groups were considered to be statistically significant when p-value < 0.05. The binary TWIN events (0,1) for each herd were analyzed using PROC GLIMMIX with a binomial distribution and a logit link function using the following model:Y = Xβ + Zμ + e(3) where Y represents the vector of the TWIN phenotype; β represents the fixed effects of TWIN genetic group (worst 33%, 34–66%, best 33%) and lactation (2nd–4th); μ represents the random effects of animal nested within lactation to account for repeated measures; and e represents the residual, with X and Z design matrices relating observations Y to β and μ, respectively. Marginal means (i.e., twinning incidence), the standard error of the mean, and the p-values are reported. Finally, cost per cow per genetic group for TWIN was calculated as per McNeel et al. (2017) [54], using the average of the most and least conservative estimates ($97 and $225, respectively) developed in a comprehensive economic analysis by Mur-Novales et al. (2018) [27]. A total of 6280 females were analysed.Secondarily, the association of individual animal TWIN STAs, lactation and TWIN outcomes was estimated as supporting data, with the model’s β term updated to represent the fixed effects of TWIN STA (85–115 inclusive, i.e., ±3SD) and lactation (1st–4th), and μ updated to represent the random effect of animal. Plots of the association between the individual animal STAs and TWIN outcome are reported.Tertiarily, the association between TWIN STA, season of conception (spring, summer, autumn, winter), and peak lactation output (from the previous lactation) with TWIN outcome was estimated as supporting data. This was completed given previous reports of associations between season of conception and peak lactation with TWIN [1,2,3,5]. In this analysis, the model was updated such that the β term represented the fixed effects of TWIN STA, season of conception, and peak lactation output of the previous lactation. This tertiary analysis was completed using 3rd lactation records for the 5 herds. This lactation was chosen given TWIN incidence was appreciable, peak lactation data from the previous lactation was available, and culling bias was relatively low. Type III Tests of Fixed Effects for TWIN STA, season of conception and peak lactation output of the previous lactation are reported. A total of 4,162 females were analysed.3. Results3.1. Data Characteristics for Genetic EvaluationThe total number of pedigree records, phenotypic records, animals with phenotypes, animals with genotypes, animals with both genotypes and phenotypes, and the incidence of twinning as of August 2020 are given in Table 2.3.2. Variance Components and Summary Statistics for Genetic EvaluationTable 3 shows the estimated variance components for twinning in Holsteins. The estimated heritability of twinning was 0.088, with a repeatability of 0.18.The means, standard deviations, minimum and maximum of (g)PTAs, STAs, and reliabilities for TWIN are presented in Table 4. Figure 1 shows the associated distribution of (g)PTAs for twinning.3.3. Correlations of Twinning with Other TraitsTable 5 and Table 6 show the product-moment (Pearson) correlations of genomic predictions for TWIN with predictions for the DWP$ index and other wellness traits offered in CLARIFIDE® Plus, as well as the correlations of TWIN with economically important traits in the genetic evaluation produced by the Council on Dairy Cattle Breeding (CDCB). As expected, the strongest correlations exist between TWIN and RETP, ABORT and METR. The weak correlation between TWIN and production traits communicate that making genetic progress for TWIN would not negatively impact genetic progress for those production traits, and vice versa.3.4. Demonstration of Evaluation EfficacyAn overview of the herd data is presented in Table 7, showing an incidence for first and third lactation per herd. First lactation incidences range from 0.56–1.48%, and third lactation incidences range from 6.95–14.00%.Differences in twinning incidence (marginal means) were statistically significant between the genetic groups (p < 0.0001). As shown in Table 8, the differences in twinning incidence between the top and bottom tertiles was 9%, 5%, 6%, 6%, and 7% for herd 1, 2, 3, 4, and 5 respectively. Previously published disease economic costs demonstrate that the differences in marginal means by genetic groups (disease incidence) translate into appreciable differences in expected economic costs on a US dairy operation (Table 8). Cows in the lowest genetic risk group (Best 33%) for Twinning (i.e., highest TWIN STAs) had 9%, 5%, 6%, 6%, and 7% percentage points lower frequency of twinning for herd 1, 2, 3, 4, and 5 respectively, which represents a 47%, 56%, 50%, 67%, and 78% difference in incidence of twinning when compared to the highest risk group (Worst 33%). This translates into decreased losses per cow of $15, $8, $9, $9 and $11 for herd 1, 2, 3, 4, and 5 respectively.Table 9 reports the significance for each of the fixed effects included in the tertiary model, with p-values ranging from 0.0033 (TWIN STA 2013 for herd 5) to 0.9129 (season of conception for herd 1). TWIN STA was a significant predictor of TWIN for 4 out of the 5 herds, with the exception of herd 2 (p-values ranging from 0.0033–0.03 for herds 1, 3, 4 and 5, and 0.1147 for herd 2). Season of conception was not a significant predictor for any herds (p-values ranging from 0.2779–0.9129). Peak lactation (previous lactation) was a significant predictor for TWIN for 1 of the 5 herds, with a p-value of 0.0495 for herd 3 (and ranging from 0.0650–0.8849 for the other herds).4. DiscussionThe incidence of twinning records in the genetic evaluation of 3.25% reported here is similar to incidences reported in previous studies, which range from <1–5.8% [12,13,14,15,16,37]. Reported estimates in this study for heritability, repeatability, genetic correlations and (g)PTAs are consistent with those previously published [7,8,13,16,29,32,34,35,36,37,38,39,40,41]. To the best of our knowledge, this is the first time a genetic and genomic evaluation for TWIN in female dairy cattle has been described and demonstrated to predict actual twinning events using producer-recorded data, and subsequently included in a holistic multitrait selection index. The observed efficacy of using producer-recorded data for conducting successful genetic evaluations of dairy cow wellness traits corroborates the results of previous studies [53,61,64,72]. Given the demonstrated variation in, and efficacy of TWIN STA predictions observed in this study, in addition to the economic and animal health-related impacts of TWIN, the justification for including a genomic prediction for twin pregnancies in a balanced multitrait selection index is strong. That being said, truly efficacious protocols for reducing the incidence of TWIN on farm will arise from combining genetic (omic) tools and best practices previously documented [11,12,27,30], examples of which are outlined in Section 4.4 below. 4.1. Variance Components and Genetic (omic) CorrelationsPrevious studies that estimated genetic parameters for twinning in cattle suggest that heritability estimates for the trait range from 0.004–0.29, with repeatabilities (where reported) ranging from 0.04–0.55 [7,8,13,16,29,32,34,35,36,37,38,39,40,41]. The heritability for TWIN reported in this study is 0.0882, with a repeatability of 0.1767 (Table 3); both fall into the ranges previously reported. Considerations to be taken into account when interpreting these varying estimates in the literature include impacts relating to the extent of variability (residual and genetic) in the populations sampled, the heritability calculation used, the trait definition, the prediction model used (animal/sire, linear/threshold), and study population used. Regardless of differences in approach, the results of this study corroborate previous reports, and the non-zero heritability estimate, alongside the evident genetic variation for TWIN in cattle populations signifies that inclusion of the trait in well-structured breeding strategies that will apply appropriate selection pressure will result in genetic progress toward lower risk of TWIN pregnancies in a given lactation.There exists both qualitative and quantitative evidence of an association between twinning and other traits of (economic and wellness) importance in dairy production. Previous phenotypic associations reported implicate correlations between twinning and mortality, stillbirths and abortions, and reduced birth weights for calves [2,17,18,19]. For cows, higher incidences of periparturient diseases such as retained placenta, metritis, displaced abomasum, and ketosis, in addition to reduced productive lifespan for twin-bearing cows have been reported [2,20,21,22,23]. Additionally, associations have been reported between twinning and reproductive performance of cows and heifers alike, including increased days open, higher ovulation rates, increased services per conception, cystic ovarian disease and dystocia for cows, and freemartinism in most heifer calves twinned to bull calves [1,2,4,21,24,25].In the present study, we reported genetic correlations between genomic predictions for TWIN and genomic predictions for an index and traits of importance for dairy wellness and profitability (Table 5 and Table 6). As expected, TWIN is positively correlated with other wellness traits pertaining to reproductive wellness: abortions (r = 0.25), retained placenta (r = 0.24) and metritis (r = 0.13), indicating that twinning is associated with other reproductive problems in dairy females. The highest correlation (r = 0.25) was observed with abortions, which was expected, as twin pregnancies often end in abortions. The correlations between TWIN and KETO and TWIN and DA are positive yet weak (0.05 and 0.02, respectively), suggesting that previous observations in the literature regarding such correlations between these health events could be influenced to a larger extent by environmental factors. Nevertheless, the non-zero correlation suggests there is a genetic association between these traits, albeit weak. As these traits are all expressed in STAs, positive correlations are interpreted as improvement in one will result in improvement of another. This indicates that applying selection pressure on these traits will result in genetic progress in the same direction for all (with the exception of LAME); in a similar vein, this also implies that cows that are more genetically predisposed to TWIN are more likely to experience abortion, retained placenta and metritis events. These correlations make sense physiologically, as the transition period for cattle is a stressful period, and is potentially exacerbated for twin-bearing cows. These correlations may also imply that these reproductive wellness traits are controlled by the overall reproductive system of the cow. The correlations of TWIN STAs with the CDCB trait PTAs are interpreted in the opposite manner—A positive correlation infers that improvement in a trait (g)PTA will result in a corresponding disimprovement in TWIN STA. All correlations with the traits in the US CDCB genetic evaluation were weak in nature. The strongest of these is the negative correlation between TWIN and PL (r = −0.055), which is not unexpected, given twinning has been previously associated with reduced productive lifespan in dairy cattle [23]. Thus, selecting for TWIN will result in indirect selection for a more favorable productive life. As with Ron et al. (1990) [35], the present study did not find an appreciable correlation between genomic predictions for TWIN and production traits (magnitudes ranging from 0.011 to 0.038 for milk, fat and protein; Table 6). Thus, selecting for reduced risk of TWIN will have little or negligible impact on making genetic progress for production traits. It is not unreasonable to surmise that previous reports implicating milk production and twinning being correlated could be underpinned by the many environmental and physiologic factors that influence twinning risk i.e., parity, season, ovulation rate, previous twin calving events, pharmaceutical use, hormonal fluctuations in high-feed intake, high-producing cows, plane of nutrition etc. [1,2,3,4,5,6,7,8,30]. In contrast to phenotypic associations in the literature, there is a weak to negligible correlation between genomic predictions for TWIN and cystic ovary, productive life, livability and the fertility traits (DPR, HCR and CCR) with magnitudes ranging from 0.011–0.031 (Table 6). The differences in magnitudes of correlations could also be due to differing models, populations sampled, data editing procedures and, simply, because the multifactorial nature of these quantitative traits results in environmental/physiological influences having a large impact on the expression of these traits. This offers the opportunity to use holistic and complementary management strategies (e.g., nutritional, pharmaceutical and breeding strategies) to reduce the relative risk in cattle experiencing these events. Furthermore, these negligible correlations showed that predictions for TWIN provides new information about an animal’s genetic potential for TWIN, further demonstrating the importance for its inclusion in a multitrait selection index that focuses on dairy wellness and profitability.Most importantly, the correlation between TWIN and DWP$ was positive at 0.115, inferring selecting on DWP$ will result in genetic progress for TWIN, in parallel with all other trait categories in the index (cow and calf wellness, production, fertility, longevity, functional type, milk quality and calving). TWIN’s inclusion in the 2020 update of the DWP$ index was predicated on the fact that it was identified as a trait that impacts the lifetime profitability of a dairy female, and which occurs at an appreciable incidence in the industry. Subsequently, to quantify how selecting on DWP$ would alter genetic progress of its (aforementioned) underlying traits, the expected response to selection per standard deviation of genetic improvement of the index has been estimated previously (Zoetis data on file, Appendix A, Supplementary Materials S1). Inclusion of TWIN at a 1% weighting, coupled with a 1 standard deviation of genetic improvement on DWP$ index STA results in an expected response to selection of 0.81 STA (i.e., a lower relative risk of TWIN in a given lactation), alongside 218lbs of milk, 1.44 months of PL, 0.27% DPR, $9.47 Calving Ability, 0.80 STA for retained placenta and 0.55 STA for abortion. For further information on this response to selection, please refer to the Appendix A and the Supplementary Materials section.4.2. (g)PTAs and (g)STAs of TwinningTo the best of our knowledge, this study is the first to report genomically enhanced PTAs ((g)PTAs) and STAs using a threshold animal model and ssGBLUP methodology for TWIN in Holstein dairy females. The means, standard deviations, and extreme values of gPTAs for TWIN are comparable to those previously obtained for other wellness traits in Holstein [61] and Jersey cows [64]. TWIN (g)PTAs reported in the present study range from −7.449–20.792%, with means ranging from −0.096–0.147% and SD ranging from 2.55–2.61%, depending on cohort evaluated (Table 4). As seen from Figure 1 (normal distribution curve), the numerical variation in (g)PTAs is considerable, and we have reported one of the widest ranges in (g)PTAs seen for TWIN. The most extreme positive (g)PTA value for twinning was obtained for a bull with 106 phenotyped daughters. Out of a total of 1,145,323 animals, 1,940 had a (g)PTA greater than 10. Generally, a broader range of (g)PTAs is preferable, because it enables better segregation of animals of different genetic merit for the trait. The average genomic reliability of about 40% for young genotyped animals without their own phenotypes or progeny is also comparable with genomic reliabilites for other similarly heritable traits. A total of 6 animals in the analysis had a reliability of zero, which may not be expected in genotyped animals. Zero or very low reliabilities are the artifact of the formula used for approximation of reliabilities, which relies on the diagonal values of the genomic relationship matrix (GRM). Animals that are not very well related to the majority of the population (e.g., crossbred animals), or animals genetically isolated from US Holstein genetics may have high values of the GRM diagonals, resulting in underestimation of reliabilities.These present estimates are based on a large, longitudinal dataset which has undergone rigorous data editing procedures, and demonstrates wide-ranging variability in genomic merit within the study population. The inclusion of producer-recorded on-farm records shows that even with the potential intra-herd variation in phenotype recording (i.e., some herds potentially do not record, or record inconsistently given the obvious nature of the phenotype), the data editing and modeling approach employed yielded heritability and (g)PTA/STA estimates demonstrating exploitable variation, whilst reflecting actual on-farm situations. Moreover, the advantages of using an animal model lies in the fact that they are more suited to calculating variance components given that they account for all relationships amongst animals by taking in to account selection and assortative mating (i.e., accounts for both maternal and paternal influences) [73]. Furthermore, using cutting-edge ssGBLUP methodology for genomic evaluations removes risk of double counting and reduces pre-selection bias, resulting in improved accuracy of selection for low heritability traits and traits with incomplete information [61,74,75]. This prediction supports the much-needed shift in the industry from an historically reactive treatment to a proactive preventative to combat the rise in twinning incidences in dairy herds across the globe.We converted the TWIN (g)PTAs to STAs for ease of interpretation and thus to facilitate ease of selection decisions. As McNeel et al. (2017) [54] outlined, a value of 100 represents the average expected TWIN risk, with the standard deviation being 5. STAs greater than 100 represent lower expected average disease risk relative to the base population in this study. Thus, higher STAs are more desirable, and selecting for a higher TWIN STA value will apply selection pressure for reduced genetic risk of TWIN. Genetic progress and thus adaptability in the long-term relies on ample exploitable genetic variation in the population, which has been demonstrated in the present study [76].4.3. Demonstration of Evaluation EfficacyIn the current study, we hypothesized that cows with the highest genetic risk (lowest STAs) for TWIN would have a higher phenotypic incidence of TWIN than cows with the lowest genetic risk (highest STAs). As outlined by McNeel et al. (2017) [54], an accepted best practice for any genetic evaluation is to evaluate the association of its genetic predictions with the observed phenotypes of the evaluated animals in an external population. To our knowledge, this is the first time such a demonstration has been completed for genetic(omic) predictions for twin pregnancies. The results of the current study demonstrate the ability of the TWIN prediction to accurately predict TWIN incidence in a given lactation. Given the negligible incidence of twinning in the 1st lactation for all 5 herds (Table 7), the analysis was restricted to 2nd–4th lactation. This trend has been demonstrated and outlined previously [30]. The association between individual TWIN STAs and twinning incidence illustrated in Figure 2 demonstrate that as expected, when STA values increase, twinning incidence decreases across all herds, all reproductive protocols, and all lactations. Unsurprisingly, the herd that used no synchronization protocol (herd 1) had the highest twinning incidence of all herds regardless of lactation, and those that used a synchronization protocol had a relatively lower incidence, which corroborates previous reports that Double Ovsynch protocols reduce twinning incidence on farm [30] and provides evidence that using genetics and management practices in tandem can help reduce on farm twinning incidences. Although some studies have previously reported seasonal effects on the incidence of twinning [2,3], inclusion of a season of conception effect in the model in the present study did not contribute to explaining variation in twinning incidences/events, given both TWIN STA and peak lactation output for the previous lactation were included in the model. This lack of association reported here mirrors results reported previously [1]. Peak lactation was significant for one of the five farms, which corroborates previous studies [5,77]. However, TWIN STA is the most consistently significant predictor of twinning incidence on farm, as demonstrated in the present study. This demonstration shows that producers can use the TWIN prediction to effectively reduce twinning incidence on-farm, with associated economic benefits (as reported in Table 8—up to a 78% reduction in frequency of twinning, with up to $15 in reduced cost losses per cow can be achieved). In future, when sufficient record numbers from herds representing a global cohort of females are accumulated, it could be beneficial to conduct another demonstration study on a superset of consolidated data with reproductive protocol included as a fixed effect in the model. In any case, as outlined by Van Vleck et al. (1991 (b)) [52], genetic evaluations are useful for early selection as producers to not need to have parturition records to make decisions based on these predictions. In the present study, we have corroborated this observation through the demonstration presented here. Thus, direct selection on TWIN can play an important role in reducing twinning incidences in Holstein populations, again, provided it is part of a holistic breeding and management strategy that selectively targets appropriate replacement heifers according to the producer’s breeding objective(s).4.4. Inclusion in Index & Breeding Strategies to Complement Current Management StrategiesAs outlined by Fessenden et al. (2020) [56], selection indices are a crucial component of many breeding programs, and are designed to facilitate selection for balanced genetic improvement across multiple traits of economic and production importance to producers [56,57]. Thus, inclusion of TWIN in a selection index is warranted, and in the present study TWIN was included in the DWP$ index. Defining practical ways to incorporate this information into a holistic breeding and management strategy will empower producers to extract the maximum value from the body of knowledge that exists today on twinning in dairy cattle. Any consideration for breeding for higher TWIN STAs (and thus lower risk of TWIN) should be undertaken within the framework of a multitrait breeding objective to account for relationships between traits and avoid unintended consequences [78]. Selection indices facilitate multitrait breeding objectives to be realised (once appropriate selection pressure is applied to the population) by providing a tool for producers to rank and select animals based on a single value that incorporates genetic merit information from many traits [55].To our knowledge, this is the first time a prediction for TWIN has been included in a multitrait selection index for the dairy industry, which constitutes a novel step forward when it comes to harnessing genetic(omic) technologies in a practical way for producers to proactively reduce TWIN in Holstein populations. Previous genetic(omic) merit estimates generated for TWIN served other purposes such as (1) demonstrating variation in PTAs in cattle populations, (2) identifying QTLs associated with twinning rates and (3) inclusion as an input for genome-wide association studies [16,32,35,37,39,40]. The objective of this study was to include it in a selection index to empower producers in making sound breeding and management decisions to proactively reduce twinning risk.This empirical theoretical and practical evidence demonstrating the exploitable genetic component to TWIN in Holsteins offers a solution for reducing TWIN in dairy cattle populations. One major advantage of this genetic component is that genetic progress made upon selection is cumulative, and can be permanent provided it is consistently selected for by producers. Environmental influences that can impact TWIN risks are numerous and have been outlined by Fricke (2001), Fricke (2015), López-Gatius et al. (2017) and Mur-Novales et al. (2018) [11,12,27,30]. Opinions differ on what is the optimal way to manage females carrying twin calves, with some advocating manual embryo reduction [27], and others suggesting selective culling and selective maintenance of pregnancy depending on the value of the cow and calf(ves) as the preferred alternative [11,12]. Another important strategy for producers to consider employing on-farm is hormonal manipulation of dairy females using a Double Ovsynch protocol before artificial insemination to reduce the likelihood of double ovulation [30]. The physiological basis for this strategy is that the protocol increases progesterone during follicular development which results in a reduced incidence of double ovulation and subsequent dizygotic twinning. Importantly, this strategy was the first non-genomic (i.e., environmental) management strategy proposed that proactively mitigates twinning risk in dairy females, and thus should be included in a holistic breeding and management approach to mitigate twinning risks on-farm.All methods previously proposed have merit and their various combinations should be employed depending on the specific circumstances of each herd e.g., resource availability and allocation (e.g., for purchasing semen from high index and TWIN STA bulls, purchasing therapeutics for Double Ovsynch protocols etc.). An important consideration is that the positive effects of environmental management decisions to reduce twinning events are short-lived considering they impact the animal today and cannot be transmitted to the next generation, whereas the genetic(omic) merit of a cow is a function of all selection decisions that were made throughout her ancestral generations [78]. TWIN is a multifactorial trait, and should be treated as such when designing breeding and management strategies to reduce in on herd. Hence, a truly effective strategy to reduce TWIN in dairy populations depends on the complementary combination of both environmental and genetic management strategies to balance mitigation of TWIN risk in both a short-term and long-term fashion. Given previous limitations regarding lack of selection tools available to producers, the results described in the present study provide additional options for producers to best manage their females in terms of selecting females for reduced TWIN risk, and also informing management decisions for twin-bearing cows. Previously, Fricke (2015) [30] has proposed a two-pronged approach including (1) using a Double Ovsynch protocol before artificial insemination, and (2) selective reduction of unilateral twins/selective maintenance of cows with bilateral twin pregnancies. Lopez-Gatius (2020) [31] outlined using gonadotrophin-releasing hormone treatment for pregnancy maintenance and embryo reduction preferentially using PGF2 depending upon multiple factors such as genetic merit and stage of lactation. We propose a hybrid approach in which genetic merit information is utilized to differentially inform management decisions for cows, using a selection index value as the first consideration, followed by the TWIN trait merit, and subsequent management protocols. The first step of a genomically enhanced management strategy involves genomically testing all females in the herd, with the aim to select on indexes that include TWIN predictions (such as DWP$). Applying an appropriate level of selection pressure based on the index merit values (e.g., retain the top 80% of females every year, provided potential changes to the herd’s parity structure is economically viable) should result in genomic progress for TWIN along with production and wellness traits thus breeding a robust cow. Sire selection should consider the genomic merit for DWP$ and TWIN (high STAs) in order to minimize TWIN pregnancies. Coupled with Double Ovsynch protocols, this strategy should further reduce the incidence on TWIN on-farm. Optionally, the top 80% females retained annually on DWP$ could then be secondarily ranked on TWIN, and Double Ovsynch strategies employed for those with the lowest TWIN STAs, and/or those with the lowest STAs are only bred to bulls of a high TWIN STA merit (followed by sufficient early detection of TWIN via transrectal ultrasonography to inform future management decisions). In line with recommendations suggested by Lopez-Gatius (2020) [31], cows with a high index merit that are carrying twin calves may be candidates for pregnancy maintenance. However, given that previous twinning is a risk factor for future twinning events, such cows should ideally be bred to high index and TWIN merit bulls, after a DoubleOvsynch protocol implementation. In essence, a ‘coevolution’ of breeding and management strategies is required to establish a short-term and long-term sustainable solution to twinning in dairy herds. The TWIN and DWP$ predictions are a useful tool for dairy producers interested in using genetic(omic)ally enhanced strategies to improve their overall herd twin pregnancy incidence and profitability.Our evaluation results are based on over one million producer-recorded twinning records, and over one million genotypes. Continuous data collection and inclusion of both new data from existing herds and historical data from new herds will increase the sampling depth in the Holstein population, simultaneously allowing for the variation in (g)PTA estimates observed to converge toward the true population variation, and reliability of records to increase given sufficient time. A concurrent and complementary approach when combining genetic(omic) tools and reproductive protocols ushers in the new era of precision management of dairy cattle, which should extract the maximum value of genetic(omic) predictions in relation to improving dairy wellness and herd sustainability.5. ConclusionsThis study showed that on-farm, producer-recorded twinning data can be successfully used in routine genetic and genomic evaluations for Holstein cattle. Moreover, the twinning predictions outputted by the evaluation have been demonstrated to effectively identify cows that are more genetically predisposed to experiencing twin pregnancies (regardless of reproduction protocol) when they enter the milking herd. Genetic correlations with Zoetis dairy reproduction wellness traits were low yet positive and indicate that they are controlled by the overall reproductive system of the cow. Successful inclusion of this trait in to a multitrait selection index whose aim is to improve the overall herd wellness and profitability offers a new selection tool for dairy reproduction and wellness improvement which was not available to producers until now. Using such a tool to drive on-farm genetic selection for improved reproduction and wellness traits will confer cumulative improvement of a herd’s reproductive and health status. In the shorter-term temporary relief offered by other management interventions (therapeutic or otherwise) currently available today can be used in tandem. Moreover, concurrent management interventions can complement genetic selection strategies to further reduce the risk/incidence of twinning in the herd concomitantly. The inclusion of novel reproduction traits such as twinning in a selection index offers producers a more comprehensive tool for selecting potentially more robust, sustainable and profitable animals.

animals : an open access journal from mdpi

10.3390/ani11113198

PMC8614429

The Can de Palleiro or Galician shepherd is a canine breed that was in danger of extinction but is currently growing rapidly in popularity. In this study, different behavioural traits of the breed were evaluated in order to assess breeds, select the best breeding animals and identify behaviour problems. This is the first study carried out in the Can de Palleiro breed using different scientifically validated tests. Questionnaires filled by the owners (C_BARQ) were collected, and a behavioural test (SAB) was conducted to evaluate the response of the dogs to a specific stimulus at a certain time and in a certain environment. In addition, the results from the Can de Palleiro breed were compared with those obtained from the general canine population of Galicia. Thereby, the Can de Palleiro breed showed less owner-directed aggression, dog-directed fear, excitability, non-social fear and separation-related problems and better trainability.

The Can de Palleiro (CP) is an autochthonous canine breed from Galicia (NW Spain). Interestingly, no previous research has been published about the behaviour of this breed. Thus, the aim of the present study was to obtain a deeper understanding of CP behavioural and temperamental traits and detect any potentially problematic behaviour by using the Canine Behavioural Assessment and Research Questionnaire (C-BARQ) and the Socially Acceptable Behaviour (SAB) test. Behavioural information was obtained from 377 dogs—177 CPs and 200 general population (GP) dogs—using the C-BARQ. Additionally, 32 dogs were enrolled to perform the SAB test (19 CPs and 13 GP dogs) in order to directly evaluate their temperament. Our results indicated that CP dogs had a lower tendency to show aggressiveness towards their owners (0.18 times lower, p = 0.033) and less fear of other dogs (by 0.43 times, p = 0.001), as well as higher trainability levels (2.56 times higher, p < 0.001) when compared to GP dogs. CP dogs also had increased odds of showing chasing behaviour (3.81 times higher, p < 0.001). Conversely, CPs had reduced odds of non-social fear, separation-related problems and excitability (by 0.42, 0.35 and 0.48 times, respectively; p < 0.001, p < 0.001 and p = 0.002). The current research represents a starting point for the study of the behaviour of CPs, which appear to be a working breed, with guarding and, especially, herding characteristics.

1. IntroductionCan de Palleiro (CP) is an autochthonous canine breed from Galicia (NW Spain), also known as the Galician Shepherd. The CP is of Indo-European descent, which is evident from its rustic characteristics, with its ancestors likely being native to Galicia, having been imported by Galicians during the Palaeolithic era while expanding from the British Isles and European continent. The breed shares some of its origin story with German Shepherds, Belgian Shepherds, Dutch Shepherds and the Portuguese Castro Laboreiro.The CP is a rustic and lupine-type dog with a straight profile which is eumetric and mesodolichomorphic; it is of medium size, with a height of about 60–62 cm at the withers, harmonic proportions, a strong constitution, and fairly wide (‘thick-set’) bones. Females are somewhat shorter and lighter in appearance. In general, CPs are considered to have a strong and reticent character with strangers but usually show great loyalty to their owners with whom they are docile and calm [1,2,3].The breed was traditionally employed in the rural areas of Galicia as herding dogs and as guard dogs at farmers’ houses where they would sleep in the hayloft, or palleiro. However, since the middle of the 20th century, a significant reduction in the rural population of CPs and the introduction of foreign breeds has put CPs in danger of becoming extinct [1,4].At the very end of the 1990s, the regional government in Galicia decided to study the possibility of reviving the breed and started by searching for surviving specimens. Thus, an official breed standard was described, which allowed a recovery project aimed at increasing the number of CP individuals. In addition, the CP club was created, which has been guardian of the breed since 2002 by overseeing all the official offspring and by organising adoptions to help revive the Galician Shepherd [3,4]. Nonetheless, CP remains a particularly vulnerable dog, with only 1775 specimens currently registered in the herd book (unpublished data). Moreover, in addition to its characteristic shepherd function, its employment in activities related to police or rescue services units is now being contemplated.Although the breed standard refers to the temperament and behaviour of CPs, to date, no studies have used validated tools to evaluate these traits [3]. Gathering scientific knowledge on the behavioural characteristics of dog breeds has both scientific and applied purposes [5]. Behaviour is driven by a complex interaction between endocrine and neuroendocrine factors [6]. In turn, these factors are influenced by both genetics and the environment. In fact, it was shown that several behavioural problems such as fear or aggression have a genetic component [7,8,9]. Thus, it is important to carefully choose breeding individuals to try to avoid undesirable behaviours in future generations. In this sense, understanding the behaviour of a breed is likely the first step for promoting more responsible breeding.Several standardised tests to evaluate dog behaviour and temperament have already been developed and validated. In general, a sample of dogs is subjected to the same or similar stimuli as part of these tests, while human observers attempt to measure several behaviours [10,11,12,13,14]. Although the Socially Acceptable Behaviour (SAB) test was originally developed as a test of aggression toward unfamiliar people, this tool has also been validated to assess aggressive and fearful behaviours in dogs by analysing their response to 16 standardised subtests [10,14,15,16,17].Indeed, it has also been shown that the SAB test can help to reduce unwanted behaviours in dog populations when used to direct breeding policies [18]. Nonetheless, although this test is comprehensive, it is often difficult to conduct, and the possibility that the experimental setting itself could result in the emergence of novel behaviours cannot be excluded [19].Other methods focus on the assessment of day-to-day behaviour using a questionnaire for dog owners. They are often used as a means of validating behavioural tests and typically provide detailed information about a given dog’s tendency to display different behaviours because owners can observe their animals in a variety of situations over an extended period. In this sense, a widely used questionnaire is the canine behavioural assessment and research questionnaire (C-BARQ). The C-BARQ contains questions regarding aggression, fear and anxiety, trainability, excitability, separation-related behaviour, attachment, attention seeking, and chasing [20,21,22]. It has been used with different purposes, including the comparison of the behavioural profile of different breeds [19,22], to study civilian working dogs [23], military working dogs [24], drug detection dogs [25] and guide and service dogs [26].Based on the above, the aim of this present article was to use the C-BARQ and SAB tests to assess the behavioural traits of CP and to compare the results with those obtained for the general canine population.2. Material and Methods2.1. Surveyed Animals This study was carried out at the Rof Codina Veterinary Teaching Hospital (VTHRC) that is part of the Veterinary Faculty of Lugo and is the referral centre for veterinary clinics in Northwest Spain. We included 377 dogs, 177 (46.9%) of which were CPs. The CP dogs were recruited through the Can de Palleiro Club with which the VTHRC has a collaboration agreement. The remaining 200 general population (GP) dogs (53.1%) were recruited through social media networks such as Facebook, Instagram, and LinkedIn and by sharing an e-mail chain, starting with our personal contacts. The GP dogs were dogs of different breeds (except CP) and were used as a representative sample of the general Galician canine population (Table 1).2.2. Behaviour Assessment and Data CollectionBehavioural information about the 377 dogs was obtained using the C-BARQ, a validated questionnaire developed by Hsu and Serpell (2003) [18]. Additionally, a randomly selected sample of 32 dogs from the overall population (19 CPs and 13 GPs, including 1 miniature Pinscher, 1 Dachshund, 1 Poodle, 1 German Shepherd, 1 Podenco and 8 mixed-breed dogs of different morphotypes) completed the SAB test [10] to directly evaluate the temperament of these animals. Descriptive statistics summarising the characteristics of the 377 dogs, as well as the results from the 32 dogs that undertook the SAB test, are presented in Table 2.The C-BARQ was filled in by the owners via an online platform. The owners were able to contact one of the veterinarians responsible for the ethology service at the HCVRC to discuss and solve any possible doubt or misunderstanding they had regarding the C-BARQ. This questionnaire comprises 100 questions which describe different ways in which dogs typically respond to common events, situations, and stimuli in their environment. The responses are grouped into 14 behavioural traits as follows: stranger-directed aggression, owner-directed aggression, dog-directed aggression, dog-directed fear, familiar dog aggression, trainability, chasing, stranger-directed fear, non-social fear, separation-related problems, touch sensitivity, attachment/attention seeking, excitability and energy [20,21,22] (Appendix A Table A1). All the traits are expressed on a 0 to 4 scale, in which 0 indicates no sign of the behaviour in question, and 4 indicates the presence of a severe form of the behaviour.The SAB test was set up by following the procedure described by Planta and De Meester (2007) [10] and was performed at the HCVRC by one of the authors (S.M.M.). The test comprises 16 subtests which analyse posture and behavioural responses to different stimuli which are displayed in a fixed order [10,15,16,17] (Appendix A Table A2). Thus, posture and behavioural strategy scores (Appendix A Table A3 and Table A4) were recorded according to De Meester et al. (2011) [15]. The neutral position was defined as the posture adopted in an active but relaxed mood according to the breed standard.2.3. Statistical AnalysisAll the statistical tests were performed using SPSS 15.0 software (SPSS Inc., Chicago, IL, USA). The effect of breed (GP vs. CP) on the C-BARQ scores for the different behavioural traits was examined by applying the Kruskal–Wallis test. The effect of breed on the responses observed through the SAB test was assessed using the Fisher exact test. In this test, to assess the agreement between observers in the positions and strategies shown by the dogs for each subtest, the kappa (κ) index was used.Afterwards, ordinal regression models were fitted to assess the effect of breed (CP vs. GP) on the C-BARQ score for the different traits (given that data from all 377 dogs were available for this test). Likewise, the following explanatory variables were considered in the models fitted as control variables: gender, age, age at the time of acquisition, neutered vs. unneutered, dog activity pattern (if the dog displayed any exceptional activities or not, i.e., bike, walking, herding) and whether the owner had previously owned other dogs.One model was tested for each of the following C-BARQ traits: stranger-directed aggression, owner-directed aggression, dog-directed aggression, dog-directed fear, familiar dog aggression, trainability, chasing, stranger-directed fear, non-social fear, separation-related problems, touch sensitivity, attachment/attention seeking, excitability and energy.The scores obtained for each of the traits were divided into five categories [27]:-0: C-BARQ score = 0.-1: C-BARQ score >0 to 1.-2: C-BARQ score >1 to 2.-3: C-BARQ score >2 to 3.-4: C-BARQ score >3 to 4.Therefore, the following odds were modelled [27]:-C-BARQ score 0, 1, 2, 3 vs. 4-C-BARQ score 0, 1, 2 vs. 3, 4-C-BARQ score 0, 1 vs. 2, 3, 4-C-BARQ score 0 vs. 1, 2, 3, 4The ordinal regression model provided the odds ratios for higher levels of the C-BARQ score (relative to being in or below a given score). When a variable changed the effect of the remaining coefficients by 10% or more, it was considered a confounder, and we retained it in the model, regardless of its level of significance [27].The parallel line test was used to assess the hypothesis of proportionality. Ordered logistic regression assumes that the coefficients that describe the relationship between the lowest versus the highest response variable categories are the same as those that describe the relationship between the next lowest category and all the higher categories [28].3. Results3.1. Canine Behavioural Assessment and Research Questionnaire ResultsThe mean and median C-BARQ scores (along with quartiles) obtained for the CP and their counterparts for the GP are provided in Table 3. Differences between groups (CP vs. GP dogs) were significant except for familiar dog aggression and attachment/attention seeking. Compared to GP dogs, CPs showed significantly higher scores for stranger-directed aggression, dog-directed aggression, trainability, chasing and energy, as well as significantly lower owner-directed aggression, dog-directed fear, stranger-directed fear, non-social fear, separation-related problems, excitability and touch sensitivity.3.2. Socially Acceptable Behaviour ResultsTwo CP out of the 32 CP dogs for which the test was performed presented extreme fear during the test, and so the SAB was interrupted and cancelled for animal welfare reasons.The kappa values to assess the agreement between the observers for the different postures and strategies displayed by the dogs for each subtest showed values >0.85 except for posture 7 in subtest 3 (κ = 0.74) and posture 3 in subtest 10 (κ = 0.78).The following results from the SAB test were statistically significant. When the dogs were exposed to an unfamiliar sound, 69.2% of the GP dogs showed strategy 3, while only 26.3% of the CPs showed this behaviour. Additionally, 23.0% of the GP dogs showed postures compatible with extreme fear (score = 6) when they were approached by unfamiliar people who were staring at the dog, while none of the CPs (0%) showed this behaviour. Finally, when the owner tried to pet the dog with a doll, 100% of the GP dogs showed slight signs of stress (strategy 1), while only 70.6% of the CPs did.3.3. Regression ModelsAccording to the regression models, no relationship was found between the groups (CP vs. GP) for stranger-directed aggression, dog-directed aggression, familiar-dog aggression, stranger-directed fear, attachment/attention seeking, touch sensitivity or energy after correcting for the control variables.Regarding owner-directed aggression, the ordinal model indicated that CPs had a C-BARQ score 0.18 times lower than the GP dogs (p = 0.033) for this factor. The factors of age and having previously owned other dogs were retained in the model because their exclusion modified the coefficients by more than 10%. The remaining control variables (i.e., gender, neutered vs. unneutered, dog activity pattern and age at the time of acquisition), being non-significant, were excluded from the model, since their elimination did not modify the remaining coefficients by more than 10%. The same procedure was followed in the subsequent models (Table 4).Ordinal regression also indicated that CP dogs had a 3.0-fold lower score for dog-directed fear (p = 0.001) and a 2.56 times higher score for trainability (p < 0.001) than GP dogs. Trainability was also influenced by the neutering status, having previously owned other dogs, and the dog activity pattern (Table 4).CP dogs had an increased odds ratio of a higher chasing score by a factor of 3.81 (p < 0.001). Conversely, CPs had reduced odds of non-social fear (0.42-fold, p < 0.001), separation-related problems (0.35-fold, p < 0.001) and excitability (0.48-fold, p = 0.002). Age had also a significant effect on non-social fear. The parallel line test, non-significant in the models fitted, confirmed the suitability of the ordinal regression models.4. DiscussionThe CP dog breed is rapidly growing in popularity as a breed of working and companion animals. However, scientific evidence of the behavioural characteristics of CPs is still very limited; therefore, to the best of our knowledge, this current study is the first investigative research on the behavioural profile of CPs.To assess dog behaviour and tackle possible behavioural problems, it can be effective to obtain information about individual dogs from their owners, who usually best understand the typical behaviours of their dogs. Serpell and Hsu (2001) developed the C-BARQ instrument to measure behavioural traits in pet dogs [29]. The C-BARQ is a useful and validated resource for investigating dogs’ behaviour, and several studies have used it to examine and compare breed differences in behavioural traits, with previous findings [16,21,22,30].In this research, C-BARQ was distributed between CP and GP owners. GP dogs included breeds of all the Federation Cynologique Internationale groups, as well as mixed-breed dogs of different morphotypes.The Kruskal–Wallis test showed significant differences between the CP and the GP groups for stranger-directed aggression, dog-directed aggression, trainability, chasing, energy, owner-directed aggression, dog-directed fear, stranger-directed fear, non-social fear, separation-related problems, excitability and touch sensitivity. Nevertheless, when the breed was assessed in a multivariate model with other control variables that could also be risk factors for these behavioural problems (i.e., age, age at acquisition, sex, neutering, having previously owned other dogs and activity pattern [31,32,33,34,35], no differences were found for stranger-directed aggression, dog-directed aggression, familiar-dog aggression, stranger-directed fear, attachment/attention seeking, touch sensitivity or energy. However, one possible limitation of this study is that, given its limited sample size, multivariate models were not applied in the case of the SAB test.According to the ordinal model, CP dogs had a significantly lower C-BARQ owner-directed aggression factor score than GP dogs. Similar results have also been found for other shepherd breeds [19,36] with which CPs share a common origin. Indeed, the breed standard [3] also reports that CPs are loyal and docile dogs with their owners. Like other shepherd and guarding breeds (i.e., Australian Shepherds or Rottweilers) [36], CPs tend to show less dog-directed fear and higher levels of trainability.Interestingly, lower scores for excitability, non-social fear and separation-related problems were found for CP dogs, with these lower scores corresponding to a steady temperament as defined in the breed standard [3].The C-BARQ chasing factor refers to the tendency of some dogs to display predatory chasing of cats, squirrels, birds and/or other small animals when given the opportunity. In this work, CPs obtained higher chasing scores than GP dogs. Nevertheless, most herding breeds strongly express predatory motor patterns such as stalking, while more advanced aspects of the canine hunting sequence (grabbing) are differentially developed among herding dogs. For instance, herding breeds such as the Australian cattle dog (which is used to working with typically stubborn cattle) strongly express grab-biting behaviours [37,38].Even though only a few animals from the overall cohort performed the SAB test in this work, its results coincided with some of the C-BARQ findings. Thus, the GP group but not the CP group showed stronger avoidance behaviour towards unfamiliar sounds, which also matches with the differences in non-social fear found with the C-BARQ. Of note, CP dogs showed less fear when their owner tried to pet their dog with a doll than GP dogs; correspondingly, in the C-BARQ, CPs showed less aggression toward family members than GP dogs. However, the combination of an unfamiliar stimulus (the doll) with the owner, who is most likely associated with a positive emotional state in the dog, might have decreased the novelty effect of the doll, thereby increasing the dog’s capacity to maintain emotional homeostasis [39,40].However, the SAB test showed significant differences between the groups for extreme fear when the dogs were approached by unfamiliar people staring at them, although no differences between the groups were found in the ordinal model for social fear using the C-BARQ. A possible explanation could be that this subtest was performed in the absence of the owner. The behaviour of both confident and fearful dogs can change when their owner is not present; both confident and fearful dogs can experience an increase in the posture score (lower postures), but either group may also react in the same way as if the owner were present. Therefore, it is impossible to predict a given dog’s behaviour in the absence or presence of its owner [15].Control variables included in the ordinal models are considered risk factors for behaviour problems in dogs. Thereby, some research revealed that males have a higher risk for behavioural problems than females [20,31,32,41,42]. Nevertheless, in the present study, only excitability was influenced by sex. Controversial results were found in previous studies. In fact, Takeuchi et al. [43] and Bradshaw et al. [44] showed that males are more excitable than females, whereas other surveys found that females tend to be more excitable [45,46]. The role of gonadectomy on behavior is complex; indeed, it was used to treat some behaviour problems, such as urine marking, mounting, roaming and intrasexual aggression in male dogs [47,48,49]. However, similar to our results, some studies revealed neutering as a risk factor for fear, anxiety, aggression and even some cognitive alterations [35,50,51,52,53]. It has been suggested that these behaviour problems in neutered dogs are related to the continuous elevation of luteinizing hormone at supraphysiologic concentrations occurring in gonadectomized animals [54].One limitation of this study is that the CP dogs in the SAB test were heavily female-skewed, and overall the dogs subjected to the SAB were skewed towards not neutered dogs; this could have influenced the results. More studies are necessary, with a higher and more homogeneous sample of dogs taking the SAB, to obtain more accurate results.5. ConclusionsThis current research represents a starting point for the study of CP behaviour. According to the differences in behavioural traits between CPs and the GP dogs we measured in this study, CPs seem to be a working breed with guarding and, especially, herding characteristics. As such, CPs exhibited lower scores for owner-directed aggression, dog-directed fear, excitability, non-social fear and separation-related anxiety and higher scores for chasing and trainability.

animals : an open access journal from mdpi

10.3390/ani12030311

PMC8833336

Deoxynivalenol (DON)-contaminated feed may cause anorexia, vomiting, immunosuppression, and intestinal dysfunction in pigs, which would lead to growth retardation and great losses in the pig industry. In this study, the effects of resveratrol (RES) on growth performance, the intestinal barrier, antioxidant capacity, and mitochondrial function in weaned pigs fed with DON-contaminated diets were investigated. Dietary supplementation with resveratrol increased the average daily feed intake of piglets. Diets supplemented with resveratrol increased the villus height and the ratio of the jejunum villus height to crypt depth, increased the activities of superoxide dismutase (SOD), and increased the total antioxidant capacity in the jejunum mucosa. After being supplemented with RES, the level of reactive oxygen species (ROS) in mitochondria was decreased, while the mitochondrial membrane potential in the jejunum was increased. In conclusion, these results suggested that resveratrol effectively relieved DON-induced oxidative stress in weaned piglets, improved intestinal barrier function, enhanced mitochondrial function, and improved the growth performance of piglets.

This study aimed to investigate the potential effects of resveratrol (RES) on intestinal function and oxidative stress in deoxynivalenol (DON)-challenged piglets. Twenty-four healthy Duroc × Yorkshire × Landrace weaned piglets at the age of 28 ± 1 days were randomly divided into four groups with six repetitions per group. The four groups were as follows: the control group (CON), fed with a basic diet; the RES group, fed with a basal diet + 300 mg/kg RES; the DON group, fed with a basal diet containing 2.65 mg/kg DON; and the DON + RES group, fed with a basal diet containing 2.65 mg/kg DON + 300 mg/kg RES. The results showed that the growth performance and intestinal function of DON-challenged piglets were significantly decreased (p < 0.05). Compared with the DON group, the average daily feed intake of piglets in the DON + RES group was significantly increased (p < 0.05). Additionally, dietary RES ameliorated DON-induced intestinal morphology impairment, as indicated by the increased (p < 0.05) jejunal villi height and the ratio of the jejunal villi height/crypt depth. Furthermore, after the addition of RES, the activities of superoxide dismutase (SOD) and total antioxidant capacity (T-AOC) in the jejunum mucosa were significantly increased, and the content of malondialdehyde (MDA) was significantly declined (p < 0.05). In addition, the level of reactive oxygen species (ROS) in the mitochondria was significantly reduced by RES, while the mitochondrial membrane potential in jejunum was significantly increased by RES (p < 0.05). However, there was no obvious difference between DON + RES and DON groups on average daily gain and the ratio of feed togain, except for the significant inhibition of average daily feed intake (p < 0.05). In conclusion, RES could effectively alleviate the DON-induced oxidative stress on weaned piglets, and reduce the damage to mitochondria and intestinal morphology, so as to improve the growth performance of piglets.

1. IntroductionDeoxynivalenol (DON), with a stable structure and bioactivity, is a toxic metabolite produced by Fusarium graminearum. It is widely found in cereal crops, such as wheat and corn, as well as their by-products [1]. Piglets are often at high risk of exposure to DON, because corn and wheat are common ingredients in their formulated feed. The main toxic effects of DON involve anorexia, vomiting, decreased growth performance, immunosuppression, intestinal dysfunction, and increased susceptibility to intestinal infectious diseases, resulting in large economic losses to the animal industry [2,3].Physical, chemical, and biological methods are commonly used for the detoxification of DON. Presently, the addition of physical adsorbents into feed is the most popular method for detoxification. However, most of the adsorbents have low adsorption capabilities for DON due to their weak polarities and poor electrophilicities [4,5]. Because oxidative stress is regarded as an important mechanism of DON toxicity [6], extensive attention has been focused on substances that effectively inhibit oxidative stress and improve immune function [7,8]. Resveratrol (RES) is a naturally occurring polyphenol found in peanuts, red wine, grapes, pistachios, mulberries, and chocolate [9,10,11]. It has been shown that RES possesses efficient antioxidant activities both in vitro [12] and in vivo [13]. Supplementation with RES could significantly reverse the decline in the growth performance of piglets, and improve the feed efficiency of intrauterine growth-retarded suckling piglets [14].Intestinal functions, such as the morphology of the villus and crypt, play a vital role in the growth performance of piglets [15]. Meanwhile, as a valid indicator of intestinal permeability in mammals, diamine oxidase (DAO) shows a high level of content and activity in the small intestinal villi, but low levels of content and activity in other tissues in normal conditions. However, when the intestinal mucosa is damaged, DAO will pass through into the blood, causing an increase in DAO activity in the plasma [16]. Lactic acid is a metabolic product of bacterial fermentation, which is rarely absorbed under normal circumstances, and cannot be degraded rapidly in mammals. As intestinal permeability increases, the D-lactic acid produced by intestinal bacteria will also enter the bloodstream through the damaged mucosa [17]. Hence, the levels of DAO and D-lactic acid in plasma usually reflect the degree of damage to the intestinal barrier, also called intestinal permeability. Moreover, it has been suggested that the addition of RES could protect the intestinal mucosal–epithelial barrier in an ischemia/reperfusion rat model because of its antioxidant capacity [13,18]. Nevertheless, information regarding the effect of RES on intestinal injuries induced by DON in piglets is unavailable. In the present study, 300 mg/kg RES was added to a diet containing 2.65 mg/kg DON to determine the role of RES on growth performance, intestinal barrier, and anti-oxidation in piglets challenged by DON.2. Materials and Methods2.1. Laboratory Animals and DietsThe Institutional Animal Care and Use Committee of Zhejiang University (Hangzhou, China) has approved all procedures for the experiment, with the permit number for conducting animal experiments of ZJU2018-118-44. Twenty-four weaned piglets (28 ± 1-day-old, Duroc × Landrace × Yorkshire, average body weight, 9.45 ± 0.4 kg) were divided into four groups with six replicates per group and one piglet per replicate according to similar weight and parity, with a ratio of males to females of 1:1. Each piglet was kept in an individual enclosure with 1.0 m2 of area per piglet. During the experiment, feed and water were freely provided, and disinfection and immunization were conducted in accordance with the routine procedures of the pig farm. The trial period was 28 days, involving a 7-day pretrial period and a 21-day formal period.A basal diet of “corn–wheat–soybean” was used according to the nutritional requirements for weaned piglets recommended by the NRC (2012). The nutritional formula of the basal diet is shown in Table 1. The experimental groups and diets were as follows: (1) CON, fed with a normal diet; (2) RES, fed with a normal diet + 300 mg/kg RES; (3) DON, fed with a wheat diet containing 2.65 mg/kg DON/deoxynivalenol; and (4) DON + RES, fed with a wheat diet containing DON + 300 mg/kg RES.2.2. Preparation of DON-Contaminated Wheat and Measurement of Mycotoxin ContentDON-contaminated wheat and dose selection were prepared as previously described [19,20]. The contents of DON in wheat and feed were determined by immunoaffinity column purification-high performance liquid chromatography according to a previous study [21]. The contents of aflatoxin B1, zearalenone, and T-2 toxin were determined through liquid chromatography–tandem mass spectrometry [22]. The mycotoxins in mildewed wheat were as follows: DON, 8260.14 μg/kg; aflatoxins B1, 2.78 μg/kg; zearalenone, 375.56 μg/kg; and T-2 toxin, 212.64 μg/kg. The content of mycotoxin in feed and the hygienic standard of feed (GB13078-2017) are shown in Table 2.2.3. Experimental Design and Sample CollectionEach piglet was weighed at the beginning and end of the feeding and slaughtering trials, and feed consumption was recorded daily during the trial period. The average daily gain, average daily feed intake, and the feed-to-weight ratio of each piglet were calculated at the end of the experiment. At the end of the experiment, all the pigs were slaughtered and dissected. The heart, liver, spleen, and kidney were cleaned with 4 °C saline solution, and surface moisture was drained for weighing. The organ index was calculated as organ weight (g)/live weight (kg).2.4. Morphological Analysis of the Intestinal TractThe morphological analysis was conducted according to Wang et al. A segment of proximal jejunum approximately 1 cm was selected and rinsed with saline solution [23]. After the surface liquid was drained, the sample was fixed in 10% formalin at 4 °C. After dehydration with a concentration gradient of ethanol, the samples were made transparent via xylene treatment and embedded in paraffin, then sliced using an RM2135 rotary microtome (Leica, Wetzlar, Germany) followed by hematoxylin–eosin staining (HE) and neutral gum sealing. Quantitative analyses of villi and crypts were performed using a Qwin image analyzer (Leica). The vertical distance from the top of the villi to the crypt opening was regarded as the height of villi, while the vertical distance from the crypt opening to the crypt base was regarded as the depth of the crypt. Three HE-stained sections were selected from each sample. Three fields were randomly selected for each section. At least 20 measurements were taken for each crypt and villus measurement per pig. An average value was calculated.2.5. Measurement D-Lactic Acid and Diamine Oxidase (DAO) in the PlasmaThe activity of DAO and the concentration of D-lactic acid were determined using kits from Nanjing Jiancheng Bioengineering Institute (Nanjing, China) according to the kit instructions.2.6. Measurement of the Redox Status of Intestinal MucosaThe activities of superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and the content of malondialdehyde (MDA) in the jejunal mucosa were determined using ELISA kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the previous studies [24,25].2.7. Extraction of Mitochondria from Intestinal MucosaThe extraction of mitochondria was conducted according to the previous study [26]. Approximately 0.5 g of jejunal mucosa tissue was homogenized in precooled MSH buffer (10 mmol/L HEPES containing 200 mmol/L mannitol, 70 mmol/L sucrose, 1.0 mmol/L EGTA, and 2.0 mg/mL serum albumin) and centrifuged at 1000× g at 4 °C for 10 min. The supernatant was then collected at 3500× g at 4 °C and left still for 10 min to obtain mitochondrial deposition. The protein concentration was determined using the BCA method [27].2.8. ROS Levels in Mitochondria of the INTESTINAL MucosaThe intestinal mucosal mitochondria were treated with 2 μmol/L 2′,7′-dichlorofluorescein diacetate, which could pass through the outer mitochondrial membrane. They were incubated at room temperature for 20 min, and then the fluorescence intensity was determined using a fluorescence microplate reader [28].2.9. Measurement of the Intestinal Mitochondrial Membrane Potential (ΔΨm)The ΔΨm was determined using JC-1 ΔΨm detection kits (Beyotime Institute of Biotechnology, Haimen, China) according to the manufacturer’s instructions [29]. The isolated mitochondria were suspended in 0.5 mL medium containing 5 mmol/L JC-1. After mixing, the fluorescence was immediately measured using an FLx800 fluorescence microplate reader (BioTek, Winooski, VT, USA). When the mitochondrial membrane potential was high, JC-1 aggregates in the matrix of the mitochondria formed a polymer, which produced a red fluorescence; when the mitochondrial membrane potential was low, JC-1 could not aggregate in the matrix of mitochondria, forming a monomer, which produced a green fluorescence. The ΔΨm of intestinal mitochondria could therefore be determined by the fluorescence ratio of aggregates to monomers.2.10. Data Processing and AnalysisThe data were processed and analyzed as a 2 × 2 factorial arrangement by ANOVA using SPSS statistical software, version 26.0 (SPSS, Chicago, IL, USA). The statistical model includes the effects of DON (challenged or not challenged), resveratrol (supplemented or not supplemented), and their interaction. Data are shown as means ± SD. The differences between means were analyzed using Duncan’s multiple-range tests. Significance was considered as p < 0.05.3. Results3.1. The Effect of Dietary RES on the Growth Performance of Piglets after DON ChallengeAs shown in Table 3., compared with the CON group, the final body weight, average daily gain, and average daily feed intake were significantly decreased (p < 0.05) by DON. Meanwhile, the dietary addition of RES in the DON group significantly increased average daily feed intake (p < 0.05), and there was an increasing trend (0.05 < p < 0.10) in the final body weight in the DON + RES group relative to the DON group. There was no significant resveratrol−DON interaction on the growth performance of piglets.3.2. The Effect of Dietary RES on the Organ Index of Piglets after DON ChallengeTable 4 shows that the organ indices of the liver and kidney of piglets in the DON group were significantly greater than those in the CON group (p < 0.05), indicating that DON treatment caused enlargements of the liver and kidney. Meanwhile, no significant difference was found between the DON group and the RES group on the organ index of piglets. There was no significant resveratrol−DON interaction on the organ index of piglets.3.3. The Effect of Dietary RES on the Jejunal Morphology of Piglets after DON ChallengeAs shown in Table 5., in comparison with the CON piglets, the DON-contaminated diet significantly (p < 0.001) decreased the height of the jejunal villus and the ratio of villus height/crypt depth, and increased (p = 0.034) the crypt depth. Meanwhile, dietary supplementation with RES in the DON group exerted a protective effect on jejunal morphology by significantly increasing (p = 0.005, p = 0.006) the villus height and the ratio of villus height/crypt depth in the jejunum of piglets. There was a significant resveratrol−DON interaction (p = 0.009) on the villus height/crypt depth of piglets.3.4. The Effects of RES on the Jejunal Permeability of Piglets after DON ChallengeTable 6 shows that the levels of DAO and D-lactic acid in the plasma were significantly elevated (p < 0.05) in the DON group when compared with the CON group. Meanwhile, no significant difference (p > 0.05) was found between the DON group and the DON + RES group on the levels of DAO and D-lactic acid. There was no significant resveratrol−DON interaction on the jejunal permeability of piglets.3.5. The Effects of RES on Antioxidant Activities in the Jejunum of Piglets after DON ChallengeWhen compared with the control group, DON significantly decreased the activities of SOD and T-AOC in the jejunal mucosa of piglets (Table 7; p < 0.05), while the content of malondialdehyde (MDA) was raised remarkably (p < 0.05) in the DON group. Nevertheless, the addition of RES could significantly alleviate the negative effect of DON on jejunal SOD, T-AOC, and MDA (p < 0.05). There was no difference in GSH-Px activity among the groups (p > 0.05). Regarding the T-AOC activity and MDA content, there were interactions between RES and DON treatments in the jejunum of piglets (p < 0.05).3.6. The Effects of RES on Mitochondrial ROS Levels and Membrane Potential in the Jejunum of Piglets after DON ChallengeAs shown in Table 8, the ROS level of the jejunal mitochondria was evidently induced in the DON group (p < 0.05), while the mitochondrial membrane potential was significantly reduced (p < 0.05). Meanwhile, the addition of RES in the DON group inhibited the level of ROS (p < 0.05) and enhanced the membrane potential (p < 0.05). However, the ROS level between the DON + RES and CON groups still exhibited a significant difference (p < 0.05), suggesting that the addition of RES in the DON group could not completely eliminate the adverse effects of DON. Moreover, there was an interaction between RES addition and DON treatment on the ROS level of piglets (p < 0.05).4. DiscussionA DON-contaminated diet usually affects growth performance by reducing the feed intake of piglets [30,31]. RES is a natural polyphenol belonging to the phytoalexin family. It has been shown to possess the ability to improve the antioxidant capacity, immune function, intestinal morphology, growth and reproductive performances, and meat quality of pigs [32,33,34]. In this study, the average daily gain and average daily feed intake were significantly decreased by the addition of DON, while supplementation with RES in the DON-contaminated group significantly increased the average daily feed intake, suggesting that RES alleviated the adverse effects of DON on weaned piglets. However, the increase in the average daily gain and the reduction of feed/gain by RES were not significant, which might be related to the experimental period of the study. Furthermore, in contrast with a previous study [32], the addition of RES to the basic diet had no significant effect on growth performance, which might be related to feed composition, as well as the age or weaning weight of piglets [35].The organ index is essential for judging the functional status of organs. DON may cause damage to the liver, spleen, thymus, kidney, and intestinal epithelial cells [36,37]. Supplementation with RES in the diet significantly improved the bursa and spleen indices of black bone chickens under heat stress [38], and protected the liver from injury under oxidative stress in mice [39,40]. Similarly, our results showed that DON caused no significant damage to the heart and spleen, but led to the swelling of the liver and kidney. Meanwhile, the addition of RES to the DON group had no significant effect on the organ index, which could result from the different animal models used in experiments.The integrity of the structure and morphology of the intestinal mucosa is the basis for normal digestion and the absorption of nutrients. DON exposure destroyed the integrity of the intestinal barrier, as well as the protein structural organization of intestinal epithelial cells, leading to oxidative stress, inflammation, and apoptosis. The degree of injury from DON treatment to the structure and morphology of intestinal villus epithelial cells in pigs was influenced by the dose and feeding period [41,42]. It has been shown that RES increased the height of the jejunum villi, decreased the crypt depth, and decreased the number of apoptotic cells, improving the intestinal function of piglets [43,44]. We verified that DON significantly decreased the height of the jejunal villi and the ratio of the jejunal villi height/crypt depth, which is consistent with a previous study [45]. Moreover, the dietary addition of RES in the DON group could ameliorate DON-induced intestinal morphology disorder by increasing the villus height and the ratio of the villus height/crypt depth of piglets when compared with the DON group, indicating that supplementing with RES in the DON-contaminated diet could remarkably alleviate the adverse effects of DON and improve the index of intestinal morphology. Unlike previous results [44], the addition of RES in the basal diet had no positive effect on jejunum morphology, which might result from the diverse feed composition and the day age of the piglets. It is therefore possible that RES may only play key roles under stress conditions in piglets.External stress can damage the integrity of the intestinal epithelial barrier, resulting in the metabolic dysfunction of nutrients. The levels of DAO and D-lactic acid in plasma usually reflect the degree of damage to the intestinal barrier, also called intestinal permeability. DON could cause damage to the intestinal mucosa, changing the permeability of the intestinal mucosa. The activities of DAO and the content of D-lactic acid in piglets are increased after DON challenge [46]. Cao et al. also demonstrated that RES supplementation in the diet of piglets reversed the decline of occludin, claudin-1, and zonula occludens-1 protein levels in the jejunum induced by diquat, thus affecting intestinal mucosal permeability and improving intestinal barrier function [29]. Our data showed that the levels of DAO and D-lactic acid in plasma were significantly increased in a diet containing DON, which is consistent with the previous results [47]. However, the activities of DAO and the concentration of D-lactic acid in the plasma of DON-contaminated diets plus RES were not significantly different from those in the DON group. The possible reasons for the different results with Cao et al. could result from the differences in the initial body weight of the piglets and the intestinal injury model. Therefore, further research, such as the assessment of genes and the protein expression of the tight junction, would help reveal the underlying mechanism of RES on the intestinal barrier of piglets.DON could disrupt the normal function of mitochondria and cause the release of free radicals, which induce lipid peroxidation and affect the integrity of cell membranes and the signaling of the redox cycle, resulting in increases in ROS, MDA, and TBARS, as well as decreases in the activities of antioxidant enzymes, such as glutathione and SOD [9,48]. DON can also induce mitochondrial injury by reducing mitochondrial membrane potential and induce apoptosis by up-regulating the expressions of apoptosis-related factors, such as caspase-3, caspase-8, and caspase-9 [9,49,50]. It was reported that diets supplemented with RES in piglets alleviated the oxidative stress induced by diquat through an increase in T-AOC and a decrease in H2O2 and MDA in jejunum mucosa [29]. It was also shown that RES significantly increased the activity of glutathione and SOD in the liver, and decreased the level of MDA in the serum of weaned piglets [44]. In the present study, the levels of SOD and T-AOC in the jejunum mucosa of piglets was significantly declined due to DON in the feed, while the level of MDA significantly increased. Furthermore, the ROS level in mitochondria in the DON group was significantly increased, while the mitochondrial membrane potential in jejunum was decreased when compared with the control group. Compared with DON group, RES increased the activities of SOD and T-AOC, and the mitochondrial membrane potential, and decreased the levels of ROS and MDA of jejunum. Nevertheless, although RES significantly reduced the level of mitochondrial ROS production in the DON group, it still maintained a relatively high level of ROS when compared with the CON group, suggesting that the addition of RES merely increased the antioxidant ability of weaned piglets to a certain degree. Therefore, if the DON content in the diet was too high, the mitochondrial function would still be damaged.5. ConclusionsIn conclusion, the present study demonstrated that supplemental resveratrol attenuated oxidative stress, and improved mitochondrial function and intestinal morphology in DON-challenged piglets. This study showed that resveratrol might serve as an effective additive to treat intestinal disorders involved in DON-induced growth-retardation in piglets.

animals : an open access journal from mdpi

10.3390/ani11102876

PMC8532638

Mutations of 17β-hydroxysteroid dehydrogenase type3 (HSD17B3) gene cause disorder of sex differentiation (DSD). In this study, the open reading frame sequence of ovine HSD17B3 was revealed, and the effects of amino acid substitution on ovine and human HSD17B3 enzymatic activities were evaluated. Although ovine HSD17B3 has a conserved amino acid sequence, it possesses two amino acid substitutions that are consistent with the reported variants of human HSD17B3. Substitution of these amino acids in ovine HSD17B3 for those in human did not affect the enzymatic activities. Similarly, substitution of these amino acids of human HSD17B3 for those in ovine also did not affect the enzymatic activities. However, enzymatic activities declined in the missense mutations of the HSD17B3 gene associated with DSD, which occurred in the conserved amino acids between both species.

17β-hydroxysteroid dehydrogenase type 3 (HSD17B3) converts androstenedione (A4) into testosterone (T), which regulates sex steroid production. Because various mutations of the HSD17B3 gene cause disorder of sex differentiation (DSD) in multiple mammalian species, it is very important to reveal the molecular characteristics of this gene in various species. Here, we revealed the open reading frame of the ovine HSD17B3 gene. Enzymatic activities of ovine HSD17B3 and HSD17B1 for converting A4 to T were detected using ovine androgen receptor-mediated transactivation in reporter assays. Although HSD17B3 also converted estrone to estradiol, this activity was much weaker than those of HSD17B1. Although ovine HSD17B3 has an amino acid sequence that is conserved compared with other mammalian species, it possesses two amino acid substitutions that are consistent with the reported variants of human HSD17B3. Substitutions of these amino acids in ovine HSD17B3 for those in human did not affect the enzymatic activities. However, enzymatic activities declined upon missense mutations of the HSD17B3 gene associated with 46,XY DSD, affecting amino acids that are conserved between these two species. The present study provides basic information and tools to investigate the molecular mechanisms behind DSD not only in ovine, but also in various mammalian species.

1. IntroductionAndrogens are sex steroid hormones that play various roles both in male and female physiological processes via pathways involving the androgen receptor (AR). Testosterone (T) is the most important androgen, and is produced from cholesterol in the gonads and adrenals in a series of steps by cytochrome P450 hydroxylases and hydroxysteroid dehydrogenases (HSDs) [1,2]. First, 17-HSDs catalyze the final step, the conversion of androstenedione (A4) into T. Further, 17-HSDs have been identified as the enzymes that interconvert 17-keto and 17-hydroxy sex steroids, which include at least 14 family members [3]. Moreover, 17-HSDs catalyze the reactions of not only steroid hormone metabolism, but also other metabolic pathways. Among the family members, HSD17B1 and HSD17B3 are involved in gonadal T production in mammals. HSD17B1 is mainly expressed in ovarian granulosa cells to regulate the production of active sex steroids [4,5]. It is also detectable in some other tissues, such as placenta, adipose tissue, and testis, in a species-specific fashion. HSD17B1 efficiently catalyzes the conversion of the weak estrogen estrone (E1) into active estradiol (E2). It can also convert A4 to T as active androgen or a precursor for E2 production catalyzed by CYP19A1.HSD17B3 is expressed almost exclusively in testicular Leydig cells to regulate the production of active androgens [6,7]. It contributes to the development of male sexual characteristics by converting A4 into T during fetal life and puberty. Mutations of HSD17B3 genes cause a disorder of sexual development (DSD) and undermasculinizaion as a result of low T production and an abnormal A4/T ratio in multiple species, such as human, dog, and mouse [6,8,9,10,11]. In livestock, DSD represents one of the major causes of reduced productivity [12,13,14,15]. Therefore, an understanding of sex differentiation-related genes, including HSD17B3, is necessary to explore new mutations/variants and to develop diagnostic tools to identify them. It is also important to study the molecular mechanisms of sex differentiation. To date, mutations of DSD-related genes, including HSD17B3, have never been reported in sheep. However, as other animal species, the existence of 54,XY DSD sheep, despite the presence of a sex-determining region Y gene (SRY-positive), has been reported [12]. Furthermore, various studies have used prenatal sheep to investigate the effects of T on gonadal development and sexual dimorphism in the brain [16,17,18,19,20,21,22,23]. Nevertheless, the molecular characteristics of the HSD17B3 gene in sheep have remained unclear. In this study, we identified the open reading frame (ORF) of this gene and compared the enzymatic activities of HSD17B3 with those of ovine HSD17B1. We also investigated the effects of amino acid substitution of HSD17B3 genes on the associated enzymatic activities.2. Materials and Methods2.1. Animals and Tissue CollectionThis study was conducted in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals following protocols approved by the Obihiro University of Agriculture and Veterinary Medicine Committee on Animal Care (Permission number: 20–25). Ovine fresh placenta was collected from a sheep farm and sheep ovarian tissue was obtained from the slaughterhouse in Obihiro. Ovine liver, intestine, and ovaries were collected from immature sheep at a local abattoir (Asahikawa, Hokkaido, Japan).2.2. Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) and SequencingTotal RNA from each tissue was extracted using TRIzol (Thermo Fisher Scientific, Waltham, MA, USA) or TRIsure reagent (Bioline, Luckenwalde, Germany). Total RNA of ovine testis and adrenal was purchased from Zyagnen (San Diego, CA, USA). RT-PCR was performed as described [24,25,26]. The cDNA was synthesized from total RNA of each tissue using random hexamers and SuperScript III Reverse Transcriptase (Thermo Fisher Scientific, Waltham, MA, USA). PCR was performed using Ex Taq (Takara Bio Inc., Shiga, Japan), according to the manufacturer’s instructions. The reaction products of the RT-PCR assay were subjected to electrophoresis in a 1.25% agar gel, and the resulting bands were visualized by staining with ethidium bromide. The primers used for PCR are described in Table S1. To identify the ORF of ovine HSD17B3, the PCR product obtained using testis cDNA as template was cloned into the pGEM-T Easy Vector (Promega Cooperation, Madison, Woods Hollow Road Madison, WI, USA) and sequenced on an ABI PRISM 3500 Genetic Analyzer with a Big-Dye Sequencing Kit version 3.1 (Thermo Fisher Scientific, Waltham, MA, USA), using universal primers (SP6 and T7).2.3. Sequence Alignment and Phylogenetic AnalysisThe alignment analysis of HSD17B3 sequences was performed using Clustal W. The neighbor-joining phylogenetic tree was constructed using MEGA version X. Analyzed proteins and their accession numbers are as follows: human HSD17B3 (CAG46692.1), monkey HSD17B3 (NP_001253433.1), goat HSD17B3 (XP_005684205.2), bovine HSD17B3 (AAI09701.1), porcine HSD17B3 (NP_001231719.1), rabbit HSD17B3 (XP_002708310.1), horse HSD17B3 (XP_023482861.1), donkey (XP_014712426.1), dog HSD17B3 (XP_003638918.1), cat HSD17B3 (XP_003995509.1), mouse Hsd17b3 (NP_032317.2), rat Hsd17b3 (NP_446459.1), and chicken Hsd17b3 (XP_425046.4).2.4. Cell Culture, Transfection, and Luciferase AssayHuman embryonic kidney 293 (HEK293) and African green monkey kidney fibroblast-derived CV-1 cells (ATCC, Manassas, VA, USA) were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and penicillin (100 IU/mL) /streptomycin (100 μg/mL). They were transfected using HilyMax (Dojindo Laboratories, Kumamoto, Japan). One day before transfection, cells were seeded on 48-well plates and cultured with DMEM supplemented with 10% Hyclone Charcoal/Dextran treated FBS (GE Healthcare UK Ltd., Buckinghamshire, England). HEK293 cells were transfected with expression vectors of GFP and HSD17Bs. At 2 days post-transfection, they were treated with vehicle or steroids for 2 (A4) or 3 h (E1). Then, supernatants of culture media were collected for the incubation of CV-1 cells or E2 measurements. CV-1 cells were transfected with reporter and AR-expression vectors. At 24 h post-transfection, they were treated with vehicle (EtOH), A4 (1 nM), T (1 nM), or supernatant of HEK293 cell culture media for 24 h. Transfections were performed in triplicate within a single experiment. Luciferase assays were performed as described previously [27]. Each data point represents the mean of at least four independent experiments.2.5. PlasmidsThe pQCXIP expressing ovine HSD17B3 and HSD17B1 conjugating FLAG fusion tag in their C-terminus and N-terminus, respectively, were generated by amplifying the open reading frame of each gene, adding a fusion tag to each end, and cloning them into a pQCXIP vector (Invitrogen, Carlsbad, CA, USA). Ovine AR expression vector was generated by amplifying its open reading frame and cloning it into a pQCXIP vector. pQCXIP-green fluorescent protein (GFP), Slp-ARU/Luc reporter, and pQCXIP-human AR were prepared as described [7,28]. Constructs having mutations in HSD17B3 genes were prepared by the QuikChange Site-Directed Mutagenesis Kit (Agilent Technologies Inc., Santa Clara, CA, USA). The nucleotide sequences of the constructs were confirmed by DNA sequencing as described above.2.6. Western Blotting AnalysesExtraction of total proteins from cultured cells and subsequent quantification were conducted as described previously [29]. Equal amounts of protein (50 or 100 μg) were resolved using 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes. Western blot analyses of FLAG, HSD17B3, and GAPDH were performed with antibodies directed against FLAG (1:1000; M2, Sigma-Aldrich Co. LLC, Saint Louis, MO, USA), HSD17B3 (1:1000; PAF173Hu01, Cloud-Clone Corp., Katy, TX, USA), and GAPDH (1:1000; 14C10; Cell Signaling Technology, Inc., Danvers, MA, USA), respectively. Immunoreactive proteins were detected using horseradish peroxidase-labeled secondary antibodies (1:10,000; Jackson ImmunoResearch Labs, West Grove, PA, USA) with Clarity Western ECL substrate (Bio-Rad Laboratories Inc., Hercules, CA, USA). Signal intensity was calculated using Image J software (Figure S1, [30]).2.7. Measurements by Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS)Quantification of E2 in culture media by LC-MS/MS is based on the methods as described ([27,28,31], ASKA Pharma Medical Corporation, Kanagawa, Japan). As internal standards, 17β-E2-13C-4 was added to a medium, which was diluted with distilled water. The steroids were extracted with methyl tert-butyl ether (MTBE). After the MTBE layer was evaporated to dryness, the extract was dissolved in 0.5 mL of methanol and diluted with 1 mL of distilled water. The sample was applied to an OASIS MAX cartridge, which had been successively conditioned with 3 mL of methanol and 3 mL of distilled water. After the cartridge was washed with 1 mL of distilled water, 1 mL of methanol/distilled water/acetic acid (45:55:1, v/v/v), and 1 mL of 1% pyridine solution, the steroids were eluted with 1 mL of methanol/pyridine (100:1, v/v). After evaporation, the residue was reacted with 50 μL of mixed solution (80 mg of 2-methyl-6-nitrobenzoic anhydride, 20 mg of 4-dimethylaminopyridine, 40 mg of picolinic acid, and 10 µl of triethylamine in 1 mL of acetonitrile) for 30 min at room temperature. After the reaction, the sample was dissolved in 0.5 mL of ethyl acetate/hexane/acetic acid (15:35:1, v/v/v) and the mixture was applied to an InertSep SI cartridge, which had been successively conditioned with 3 mL of acetone and 3 mL of hexane. The cartridge was washed with 1 mL of hexane and 2 mL of ethyl acetate/hexane (3:7, v/v). E2 was eluted with 2.5 mL of acetone/hexane (7:3, v/v). After evaporation, the residue was dissolved in 0.1 mL of acetonitrile/distilled water (2:3, v/v) and the solution was subjected to an LC-MS/MS.2.8. Statistical AnalysesData are presented as the mean ± SEM or the mean ± SD. Differences between groups were assessed by the one-way ANOVA followed by Tukey’s multiple comparison test using EZR (Easy R, Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria) as described [31]. p-Values less than 0.05 were considered significant.3. Results3.1. Conservation of Ovine HSD17B3 GeneAlthough predicted ovine HSD17B3 is registered in a public database, there are three isoforms that have different deduced amino acid sequences. To identify the true open reading frame (ORF), ovine HSD17B3 was amplified from testis cDNA as a template by RT-PCR using primers constructed from the conserved sequence in the 5′ and 3′ non-coding regions of Bovidae HSD17B3. Ovine HSD17B3 cDNA (DDBJ accession number: LC631120) comprises an open reading frame (ORF) of 933 bp that encodes 310 amino acids (Figure S2). This length is consistent with data of XM_042243038, despite the presence of some different amino acids. It is also consistent with the length of HSD17B3 from the other mammalian species, such as bovine, goat, porcine, and human (Figure 1A). Alignment of the deduced amino acid sequences showed the presence of a well-conserved cofactor-binding motif (GXXXGXG), an active site (YXXXK), and androgen-binding amino acids.Ovine HSD17B3 showed the highest identity with that of goat in terms of both nucleotide (98.82%) and amino acid (98.39%) sequences (Table S2). Bovine homologs also showed high identity (nucleotide: 96.14%, amino acid: 93.61%). Ovine HSD17B3 showed moderate identity with the homologs in other mammalian species (nucleotide: 77.72–89.14%, amino acid: 71.12–83.81%), whereas its identity with chicken homologs was low (nucleotide: 66.49%, amino acid: 57.86%). A phylogram based on amino acid sequences placed ovine in a clade of Cetartiodactyla with high bootstrap values (Figure 1B). As in the case of other mammals, ovine HSD17B3 was strongly expressed in testis (Figure 1C).3.2. Evaluation of Enzymatic Activities of Ovine HSD17B3Next, the enzymatic activities of ovine HSD17B3 to produce T were evaluated using our system established for human HSD17B3, which quantifies the conversion from A4 to T based on AR-mediated transactivation [7]. Although the ovine AR gene is registered in NCBI database (https://www.ncbi.nlm.nih.gov/nuccore/NM_001308584.1, accessed on 31 August 2021), to the best of our knowledge, its responsiveness to androgens has never been investigated. To construct an ovine-based system, a potential for ovine AR-mediated transactivation between T and A4 at various concentrations was analyzed in CV-1 cells. The ovine AR-mediated transactivation was increased by T from a concentration of 10−10 M (Figure 2A). In contrast, A4 hardly increased AR-mediated transactivation at this concentration. Although T had higher potentials than A4 at all concentrations, the ratio of T-induced activity to A4-induced activity was extremely high at concentrations of 10−10 and 10−9 M (Figure 2B). These results suggest that, based on the efficiency of transactivation, ovine AR can also discriminate between T and A4 at these concentrations. Next, HEK293 cells were transfected with the expression vectors of GFP and ovine HSD17B3 (Figure 2C). At 2 days post-transfection, the cells were incubated with media containing A4 at 10−9 M until 2 h. Supernatants of culture media were collected at each time point, and added to CV-1 cells transfected with ARE-Luc and AR expression vectors. AR-mediated transactivation in CV-1 cells increased linearly by each culture medium in a manner dependently on the culture period until 1 h after A4 addition in HSD17B3-expressing HEK293 cells. This reflected high enzymatic activities for converting A4 to T in HSD17B3-expressing cells. In contrast, culture media from GFP-expressing cells never activated AR-mediated transactivation, even at 2 h after A4 addition (Figure 2C,D). Similarly, culture media from both GFP- and ovine HSD17B3-expressing HEK293 cells supplemented with A4 at a concentration of 10−10 M hardly induced AR-mediated transactivation (Figure 2D).Using this system, we compared the enzymatic activities of ovine HSD17B3 and HSD17B1 (Figure 3). Ovine HSD17B1 also showed high enzymatic activities, although these were slightly lower (but not significant) than HSD17B3 (Figure 3B). To compare the function to produce active estrogen, we also investigated the enzymatic activities of HSD17B3 and HSD17B1 for converting E1 into E2 (Figure 3C). Although HSD17B3-expressing cells significantly converted E1 into E2 compared with GFP-expressing cells, the levels of conversion were much lower than that of HSD17B1-expressing cells.3.3. Enzymatic Activities Associated with Various Missense Mutations in Ovine and Human HSD17B3 GenesAlthough various missense mutations with or without the manifestation of 46,XY, DSD have been reported in the human HSD17B3 gene, the enzymatic activities of mutant proteins have often not been defined [6,32,33,34]. Ovine HSD17B3 possesses two amino acid substitutions (V31I and G289S) that are consistent with the reported variants of human HSD17B3 (Figure 4A,B). We expressed these amino acid-substituted enzymes in HEK293 cells, and compared the enzymatic activities with that of wild-type protein using the above-mentioned system. As shown in Figure 4C, the levels of ovine AR-mediated transactivation by culture media of I31V- and S289G-expressing cells were comparable to that of ovine wild-type protein-expressing cells. Similarly, substitutions of amino acids at the same positions in human HSD17B3 (V31I and G289S) hardly affected human AR-mediated transactivation by culture media (Figure 4D). In contrast, mutations in amino acids conserved between these two species (L128S and P193H) decreased AR-mediated transactivation by culture media.4. DiscussionMammalian HSD17B3 genes possess highly conserved sequences to exert common functions to convert A4 into T in testicular Leydig cells [6]. Although various previous studies investigated the effects of T on the gonadal development and sexual dimorphism of the brain in prenatal sheep [16,17,18,19,20,21,22,23], the molecular characteristics of this gene have not been revealed. In this study, we identified the ORF of ovine HSD17B3 and evaluated its enzymatic activity. The amino acid sequence of ovine HSD17B3 showed high similarity to that of Bovidae homologs, whereas it showed moderate similarity to other mammalian homologs, including human HSD17B3. However, important amino acids, such as those forming the cofactor-binding motif, active site, and androgen-binding region, were almost completely identical in all species. This conservation is likely to have led to the finding that A4 added to culture media at 10−9 M was mostly converted into T within 2 h in cells expressing HSD17B3 derived from various species [7,31]. Using this characteristic, we established a system for measuring the enzymatic activities of HSD17B3 via ovine AR-mediated transactivation, based on our previous study using human AR [7]. Although this system is an indirect method to quantify T production, there are some advantages. It is very sensitive and rapidly detects the conversion of substrates, due to a low background activity of C3 group nuclear receptors and a good response to T by ectopic expression of AR in CV-1 cells. In addition, it can be achieved using only common cultured cells (HEK293 cells and CV-1 cells). Ovine AR-mediated reporter assay is possibly useful not only for evaluating enzymatic activities of HSD17B3 mutants, but also for evaluating transcriptional activities of AR mutants. Mutations of the AR gene are also the cause of DSD in other animal species [35], although gene mutations causing DSD, including HSD17B3 and AR, have never been reported in sheep.Mutation of human HSD17B3 causes 46,XY DSD as a result of low T concentration. Newborn males with 17β-HSD3 deficiency have complete or predominantly female external genitalia with a blind vaginal pouch [6,8,9]. To date, more than 50 pathogenic and benign mutations in the HSD17B3 gene have been reported. In addition to humans, it was recently reported in DSD dogs that truncation of the encoded protein was caused by 2-bp deletion of the HSD17B3 gene [10]. Furthermore, male Hsd17b3KO mice were reported to show a delay in puberty, owing to an abnormal A4/T ratio [11]. These findings indicate that the HSD17B3 gene plays important roles in sex differentiation in various mammalian species. Therefore, it is conceivable that mutations of HSD17B3 could be the cause of DSD in sheep. Because XY DSD was also reported in sheep [12], it is expected that the sequence information and measurement of enzymatic activity of HSD17B3 in this study can contribute to revealing the mechanisms of sex differentiation and the causes of DSD in sheep.Consistent with a previous report of human HSD17B3, ovine HSD17B3 weakly converts E1 into E2 [36]. Although the physiological significance of this activity is unclear, HSD17B3 may contribute to testicular E2 production. In support of this hypothesis, testicular E2 levels in Hsd17b3 KO mice were found to be decreased, despite E1 levels being increased [11]. Estrogen-estrogen receptor (ER) pathways are essential for spermatogenesis [37,38,39]. Male Cyp19a1 and ERα KO mice are infertile as a result of the defective spermatogenesis [40,41]. In sheep, it has often been reported that active spermatogenesis is concomitant with the elevation of plasma E2 levels [42,43]. Therefore, it would be interesting to investigate the function of HSD17B3 in testicular E2 production and spermatogenesis.Consistent with the findings in other species, ovine HSD17B1 is strongly expressed in the ovary (our unpublished data). In addition to the conversion of E1 into E2, ovine HSD17B1 can strongly convert A4 to T. Previous studies showed that murine Hsd17b1 equally converts E1 and A4 into E2 and T, whereas the catalytic efficiency of human HSD17B1 for the A4 to T reaction is very weak [44]. However, using the AR-mediated transactivation system, we suggested that these results could be caused by substrate inhibition of human HSD17B1 through unphysiological A4 concentrations [7]. Human HSD17B1 produces T from A4 at a physiological concentration (10−9 M), a level similar to those of mouse Hsd17b1 and porcine HSD17B1 [7]. Therefore, it is conceivable that mammalian HS17B1 can efficiently convert not only E1 to E2, but also A4 to T.Although the amino acid sequences of HSD17B3 proteins are relatively conserved in mammals, some amino acid substitutions that are consistent with the reported variants of human homolog have occurred in other animal species. Ovine HSD17B3 possesses two such amino acid substitutions (V31I and G289S, Figure 4B). Because substitutions of amino acids at these positions in human and ovine HSD17B3 did not affect the enzymatic activities, it is reasonable that V31I and G289S substitutions are benign variants. In addition to DSD, it was reported that G289S is associated with the risks of prostate cancer and hypospadias [33,45]. However, consistent with our results, the enzymatic activity of these mutant proteins associated with this mutation was reported to be similar to that of wild-type protein in multiple previous studies [7,46,47]. Furthermore, the amino acid residue at this position is serine in almost all mammalian species except for primates (Figure 1A), although no studies have shown the associated susceptibility to prostate cancer and hypospadias in particular animal species. Hence, it is conceivable that G289S is not involved in these diseases. In contrast, missense mutations of the HSD17B3 gene in 46XY, DSD patients (L128S and P193H) result in a loss of enzymatic activities (Figure 4D). These mutations affect amino acids that are conserved between ovine and human. Because HSD17B3 genes play the conserved roles by converting A4 into T at least in male mammals, more multi-species sequence comparisons should provide valuable information for analyzing the relationship of gene mutations and variants with DSD and other diseases in various species.5. ConclusionsWe identified the correct ORF of the ovine HSD17B3 gene. Although ovine HSD17B3 has an amino acid sequence is conserved relative to those of other mammalian species, it possesses two amino acid substitutions that are consistent with the reported variants of human HSD17B3. Substitutions of these amino acids in ovine HSD17B3 to those in the human one did not affect the enzymatic activities. However, the enzymatic activities decreased in the missense mutations of the HSD17B3 gene associated with 46,XY DSD, affecting amino acids that are conserved between both species. Taking these findings together, the present study provides valuable insights for investigating the molecular basis of DSD not only in ovine, but also in other species.

animals : an open access journal from mdpi

10.3390/ani11061668

PMC8227273

The EU is one of the main markets for marine ornamental species. Their entrance into the EU, as well as their circulation between member states, is supposed to be highly regulated. Surprisingly, it is currently impossible to answer simple questions, such as how many Nemos and Dorys are imported each year into the EU, or where do they come from? This lack of knowledge is difficult to understand, as all these organisms enter the EU by air-shipping and must be controlled at customs offices in international airports. This scenario favors “business as usual” and does not allow to verify the claims on sustainability commonly made by the marine aquarium industry. However, the EU already operates a platform that may allow to collect such information in a reliable way and shed light on this blurry industry, the Trade Control and Expert System (TRACES). This platform can start by surveying marine ornamental fishes, so the EU can finally know how many Nemos and Dorys are being imported and where they are sourced from. If this approach works, marine ornamental invertebrates can also be monitored, and reliable databases can finally be assembled to document the marine aquarium trade in the EU.

The EU is one of the main importers of marine ornamental species sourced from tropical coral reefs around the world. While the entrance of live organisms into the EU, along with their intra-EU circulation, is framed within stringent control mechanisms, to date, no reliable figures exist concerning which marine ornamental species are imported, in what numbers, and where they are sourced from. This lack of reliable data in the EU on the trade of marine ornamental species is puzzling if one considers that all these imported specimens must be controlled at customs offices located in international airports. Such data deficiency favors the prevalence of blurry supply chains and a “business as usual” mindset that hampers any serious effort to promote sustainability in the marine aquarium industry. To safeguard the collection of findable, accessible, interoperable, and reusable (FAIR) data, we suggest that the EU platform Trade Control and Expert System (TRACES) refines its surveillance on the trade of marine ornamental species. The detailed survey of marine ornamental fishes alone can be used as a proof of concept to validate the use of TRACES for this purpose and, if successful, it can later be expanded to all other taxonomic groups of marine ornamental species.

1. IntroductionThe marine aquarium trade is a global-scale industry that supplies enthusiastic hobbyists, along with public and private aquariums, with a multitude of vertebrate and invertebrate organisms that mostly originate from coral reefs in southeast Asia [1]. Commonly being referred as a multi-million dollar activity, the fact is that figures on the value of this industry are commonly outdated [1,2]. Indeed, accurate and reliable sources of information on this important activity are either missing or unavailable. Unfortunately, it is not only the value pertaining to this industry that is lacking reliable and up to date figures. The volume, taxonomy, and geographic origin of the live specimens being traded are also largely unknown [2,3]. While some works have tried to estimate these figures for importing markets, such as the USA [4,5], Australia [6], and Europe [2,7,8,9], they all acknowledge limitations (e.g., taxonomy and numbers of specimens being traded are largely unknown) and consider that the figures presented are most likely underestimates.With the EU often being reported as one of the main importing markets of marine ornamental species and championing stringent control mechanisms that monitor the entrance and circulation between member states and associated states of live organisms [8,10], it is puzzling that no reliable figures exist for such a well-established economic activity [2,7,8]. This fact is even more surprising if one acknowledges that most marine ornamental species, namely specimens being collected from the wild, enter the EU via air shipping, thus having as their entrance points highly monitored and regulated facilities, namely customs offices at international airports. At present, the EU largely ignores the overall value associated with the trade of marine ornamental species, the list of species being imported, their numbers, and their country of origin. The EU is therefore unable to find Nemo and Dory because no reliable traceability protocol is in place to monitor marine ornamental species once they enter the European market.2. Blurry Supply Chains Allowing “Business as Usual” to Go OnThe EU acknowledges the advantages of tracing all movements of live organisms in arriving to, moving between, and exiting its member states. This survey reduces the potential impacts of disease outbreaks and may allow authorities to trigger faster mitigation actions along trade chains. To do so, trade chains must be well-defined and one must be able to simultaneously trace the previous and subsequent link of any player in that chain. That is certainly not the case of the marine aquarium trade [3]. Trade chains associated with the importation of marine ornamental species have long been recognized as being diffuse and blurry, making it virtually impossible to pinpoint the number and accurate geographic origin of most specimens being traded [2,11,12]. Several specimens are collected in remote regions and pass over multiple middlemen until they arrive to an international wholesaler in the main cities of exporting countries served by international air flights (e.g., Philippines and Indonesia). These are the trade chain players that prepare marine ornamentals for international air-shipping to importing countries. For many specimens, the first time they will likely be recorded in a database is at the wholesale facility. However, such databases are not framed under FAIR data principles [13], as these data on marine ornamental species will not be readily findable (F), will likely only be accessible (A) to some corporate staff and authorities, will not be interoperable (I) (as these databases are often isolated data islands), and will not be reusable or re-used (R).Not knowing the accurate place of origin of marine ornamentals being collected impairs any reliable assessment on the effect of the fishing effort targeting donor populations in the wild. How can claims on sustainability be performed by the marine aquarium industry, or any managing authority, under this scenario? Even if only employing sustainable collection practices, operations per se may not be sustainable.The blurry nature of marine ornamentals supply chains is not only an issue when these species are being traded at domestic level before exiting their source country. At times, the country of origin being declared when marine ornamentals are entering the EU may not be the country where those specimens were collected. While some large sized wholesalers operating in the EU import marine ornamentals directly from source countries, other smaller players import them from third countries that act as logistics hubs (e.g., Singapore) that facilitate imports.3. TRACES, the Right Tool to Monitor Marine Ornamental Species Trade in the EUThe most puzzling issue concerning why the EU continues to largely ignore what marine ornamental species are being imported, in what numbers, and where they originate from is that all member states already employ one of the most advanced tracking systems in the world that targets the import, export, and intra-EU trade of live animals, animal products, and plants—the Trade Control and Expert System (TRACES). TRACES is an EU multilingual online platform that was put forward to safeguard sanitary certification related with the import, export, or trade at intra-EU level of animals and plants [9]. EU law specifically refers that live animals, animal products, feed, and plants must be accompanied by official certificates that attest to their compliance with existing EU regulations. When live animals and plants are imported into the EU, as well as when they are traded within the EU single market, TRACES records all official controls, as well as the pathway of those organisms from their origin to their destination. This online platform is already used by 85 countries with over 40,000 users in the world, including several non-EU countries (e.g., Iceland, Norway, Switzerland…), is available in 34 different languages, being accessible 24 h a day, seven days a week, and is free of charge after registration [9]. Authorities in EU points of entrance, or at destination, are pre-notified and can plan controls in due time, including on animal welfare related issues. These features alone already make TRACES a very appealing platform to monitor the import of marine ornamentals. Additionally, most countries exporting these species to the EU also export other goods that are also monitored under TRACES (e.g., animal products and food) and, as such, are already familiar with the platform. According to the European Commission, TRACES aims “to streamline the certification process and all linked entry procedures and to offer a fully digitized and paperless workflow”. With a built-in statistical tool that allow users to extract data referring to imports into the EU, exports from the EU, and intra-EU trade, TRACES is the right tool to allow one of the largest markets of marine ornamental species in the world to gain an unprecedented insight on this illusive economic activity. This approach has already been previously suggested by Biondo [8,14], but is yet to be put into practice. Although some technical adjustments may have to be put into place to allow the collection of data relevant for monitoring the trade of marine ornamental species, these will allow the EU to perceive for the first time the sustainability of this economic activity using a science-based approach.4. The Way ForwardTRACES was not specifically developed to monitor the trade of wildlife in the EU, although all wildlife legally entering the EU will somehow be recorded in this platform. While TRACES has distinguished freshwater from marine ornamentals since 2014, the diversity of traded species and number of imported specimens remains not fully reported. Moreover, the multitude of marine ornamental fishes and invertebrates traded for marine aquariums may be a phenomenal obstacle to tackle and likely impair any feasible solution to be implemented in the short-term to better understand what is imported, from where, and in what numbers. As such, to test the suitability of the solution here advocated, it is suggested that only marine ornamental fishes are covered in detail. This initiative will act as a proof of concept that, once validated, may latter be extended to some groups of marine ornamental invertebrates and ultimately cover the whole biodiversity of organisms supplying the marine aquarium industry.Making the detailed reporting of marine ornamental fish species name, number, and origin mandatory by exporters to veterinary authorities at customs in international airports would allow TRACES to be the first platform able to generate FAIR data for one of the biggest importing markets of marine ornamentals in the world. This goal will only be achieved if the origin countries from which marine ornamental species are sourced from are only allowed to export these organisms to the EU if the species and number of specimens being traded, where have they been collected, and how are clearly indicated. Marine ornamental fishes originating from aquaculture should also be clearly labelled as such, to avoid a biased perception of the fishing pressure targeting the populations of those species in the wild. However, it must be safeguarded that data are reported in a way that serves the purposes of sustainability and conservation. Reporting species lists at higher taxonomic levels (including family or genus) would hamper the use of data to flag species being more prone to overfishing. The same caveat would also arise if the volume of traded species is reported in any other unit than number of specimens. Reporting as place of origin any other country than the one where fishes were originally captured from the wild would also impair the reliable use of such data.With the marine aquarium trade in the EU being so speciose when compared to other economic activities that are also surveilled by TRACES, it is certainly important that the nomenclature employed is up to date or at least allows synonyms and alternative names so data recording can continue to be performed in a reliable way [8]. The reporting of species names should be made from dropdown menus where fish species being listed cannot be restricted solely to species already being traded. Such a procedure would overlook new species being recruited into the trade, and therefore the list provided should be comprehensive and cover all known fish species from tropical coral reefs. As already referred to above, it is paramount that the use of synonyms is allowed to accommodate taxonomic updates that are regularly performed by the scientific community.5. ConclusionsWithout a reliable way to compile sound data on species composition, numbers, and origin of marine ornamental species imported to the EU, any claims by the marine aquarium industry on its sustainability are based on assumption and not science-based facts. The EU may lead the first approach ever delivering FAIR data on marine ornamental species. EU member states have pledged their commitment to ocean conservation and coral reefs regeneration (see Mission Starfish 2030 [15]). Moreover, at the last Conference of the Parties to CITES (the Convention on International Trade in Endangered Species of Fauna and Flora) in 2019, designed to ensure sustainable international trade in threatened animals and plants species, the EU co-proposed to start scrutinizing the trade of marine ornamental fishes, with the proposal being supported by the 183 member states [16,17]. If TRACES is used to its full potential, not only will the EU be able to rapidly find out how many Nemos and Dorys are imported every year by member states, but also from where they originate. This active survey on the imports of marine ornamental species, starting with fishes and expanding to cover marine invertebrates, is most of all an effort to safeguard that Nemo and Dory may continue to occur in their natural habitats—coral reefs.

animals : an open access journal from mdpi

10.3390/ani11102937

PMC8532636

Decomposition is a complex process that involves several factors, such as temperature, pH, humidity, and microbes. Microbes play a significant role in the carcass decomposition process. Microcosm burial set-ups were prepared and poultry carcasses were decomposed for 60 days in unsterilized and sterilized soil, incubated under aerobic or anaerobic conditions. The moisture content, pH, alpha and beta diversity were affected by the soil microbial community and oxygen availability during the decomposition of poultry carcasses. The bacterial taxa composition was also altered during the poultry carcass decomposition. These changes suggested that the soil with an intact microbial community and oxygen availability influenced the bacterial community structure during the decomposition of poultry carcasses. The results of this study provided information on the different bacterial species which might be associated with the decomposition of poultry carcasses.

The impact of soil with an intact microbial community and oxygen availability on moisture content, soil pH, and bacterial communities during decomposition of poultry carcasses was investigated. Poultry carcasses were decomposed in soil with or without a microbial community, under aerobic or anaerobic conditions. The samples collected in each microcosm burial set-up were analyzed by targeted 16S rRNA amplicon sequencing and Amplicon sequence variants (ASV) method. Our results showed that moisture was high in the burial set-ups under anaerobic conditions and pH was high in the burial set-ups under aerobic conditions. Meanwhile, the Chao1 and Shannon index significantly differed between the different burial set-ups and across different time points. In addition, bacterial taxa composition during the early period of decomposition differed from that of the late period. A total of 23 phyla, 901 genera, and 1992 species were identified. Firmicutes was the most dominant phyla in all burial set-ups throughout the decomposition. At day 60, Pseudogracilibacillus was dominant in the burial set-ups under aerobic conditions, while Lentibacillus dominated in the burial set-ups under anaerobic conditions. This study demonstrated that the soil microbial community and availability of oxygen significantly affected the changes in moisture content, pH, and bacterial composition during the decomposition process.

1. IntroductionThe outbreak of highly contagious diseases, such as avian influenza has been of great concern in many countries [1,2,3,4,5]. The disposal of the infected and potentially infected animals is necessary, specifically for disease containment [6]. Disposal options for such animals include burial, incineration, composting, rendering, lactic acid fermentation, alkaline hydrolysis, and anaerobic digestion [7,8]. Among these disposal methods, burial, composting, rendering and incineration are the commonly utilized methods [9]. Decomposition causes significant and sequential changes in the bacterial communities within the soil, and which is often correlated with changes in the stage of decomposition. The decomposition process is affected by several environmental factors, such as temperature, humidity, pH and the cadaver itself [10,11,12,13,14]. Decomposition can also be affected by soil microorganisms and oxygen availability. Decomposition of buried carcasses mainly relies on the activity of microorganisms producing extracellular proteolytic enzymes which break the polymers of organic matter into oligomeric and monomeric molecules [15]. The decomposing carcass contributes nutrients to the soil through nutrient leaching, which affects the microbial communities in the soil near the carcass and surrounding environment [10,16]. Microbial activity in, on, and around carcasses is recognized as a crucial factor that can affect the decomposition rate [17,18,19]. Soil microorganisms play an important role in the decomposition process as they synthesize proteases in response to an animal organic matter supply and can degrade prion proteins in vitro [20,21]. Microbial communities are important in maintaining soil quality due to their involvement in organic matter dynamics, nutrient cycling, pathogenic spread and decomposition [16,17,22,23,24]. Several studies have demonstrated, metabolites from decomposing carcasses can affect the land and surrounding environment [16,25]. Heat is known to inactivate enzymes secreted by soil microorganisms. The removal of soil microorganisms by autoclaving typically results in a decrease in microbial biomass and enzyme activity [26,27]. Meanwhile, the access or restriction of oxygen content on the body is an important factor in the decomposition rate [28].Studies have been conducted to investigate whether the presence of an endogenous soil microbial community will influence carcass decomposition. However, the composition of microbial communities between soil with, or without the microbial communities, and decomposed aerobically, or anaerobically has not been clearly elucidated. Therefore, an investigation on changes to bacterial communities in animal burial soil is necessary to identify the possible pathogenic microorganisms which can cause contamination of the environment and a risk to public health. Thus, the present study investigated the changes in the moisture content, pH, and bacterial communities of decomposing poultry carcasses buried in either (i) soil with an intact microbial community (unsterilized soil) or (ii) soil that was sterilized and incubated under aerobic or anaerobic conditions for 60 days.2. Materials and Methods2.1. Ethical Statement All experimental protocol of this study was approved by the Animal Care and Use Committee (Approval number: SCNU IACUC-2019-7) of the Sunchon National University (Suncheon, Jeollanam-do, Korea). All experiments were performed as per the guidelines and regulations set by this governing body. 2.2. Carcass and Soil PreparationTwo Hy-Line Brown hens, weighing 2.50 ± 0.1 kg, were purchased commercially and used in the study. The Hy-Line Brown hens were euthanized and handled as per the guidelines of the Institutional Animal Care and Use Committee of the Sunchon National University. The carcasses were separated from the bones, and the chicken remains, including feathers, internal organs, muscular, epithelial, and visceral tissues, were crushed and homogenized. Approximately 10 kg of soil was obtained from an agricultural field in the experimental farm of the Sunchon National University. The soil was sieved through a 2 mm sieve to remove crop roots and stones and divided into two parts. One part of the collected soil was sterilized (S), while the other part was kept unsterilized (soil with an intact microbial community, U). To prepare the sterilized soil, the sample was placed in a polypropylene bag and autoclaved at 121 °C, 15 psi thrice in four days to destroy microbes, fungi and their spores [11].2.3. In Vitro Set-Up for Carcass Decomposition A 2 × 2 microcosm burial set-up was made using two types of soil (i. soil with an intact microbial community; ii. sterilized soil) under two incubation conditions (a. with oxygen access; b. without oxygen access) (Figure 1). The four microcosm burial set-ups were as follows: UA (unsterilized soil–aerobic condition), SA (sterilized soil–aerobic condition), UAn (unsterilized soil–anaerobic condition), and SAn (sterilized soil–anaerobic condition). Three replicates for each burial set-up incubated at different periods (0, 5, 10, 30, and 60 days) were prepared. A total of sixty sterilized containers (450 mL each) were used, with each containing 112.5 g of soil and 37.5 g of meat. The mixture of soil and meat was decomposed under aerobic and anaerobic conditions. Anaerobic conditions were created by sealing all parts of the container and placing it in an incubator with 5% flowing CO2 gas. For simulating aerobic conditions, the lid of the reactor was punctured so that air could pass through the hole before placing it in an incubator. All experimental set-ups were incubated at 25 °C for a total period of 60 days.2.4. Sample CollectionSamples were collected on days: 0, 5, 10, 30 and 60. Day 0 was the initial placement day followed by subsequent sample collections on days 5, 10, 30 and 60. Individual samples (approx. 30 g) were taken from each replicate container for each period and distributed into two sterile 50 mL Falcon® tubes. One tube was stored at 4 °C until physicochemical analysis, and the other tube was stored at −80 °C for further molecular analysis.2.5. Moisture Content and pH during Carcass DecompositionThe pH was measured by suspending 1 g of soil in 5 mL of sterilized distilled water followed by vortexing for 1 min. The large particles from the mixture were allowed to settle for 5 min and the supernatant was collected and its pH was measured with a pH meter (SevenCompact™ pH/Ion meter S220, Mettler Toledo, Switzerland). The average readings of the three samples were used to estimate the pH for each soil sample [11]. The moisture content of the soil and carcass mixture was estimated according to the standard method AS 1289 B1.1. One gram of burial soil was weighed, placed in an aluminum plate, and oven-dried overnight at 105 °C.2.6. DNA Extraction, PCR Amplification and 16S rRNA Amplicon SequencingThe DNA was extracted from 0.25 g of soil samples obtained from each burial set-up using the DNeasy® PowerSoil® Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The extracted DNA was stored at −20 °C until further processing. The quality and quantity of the extracted DNA were checked using Quant-IT PicoGreen (Invitrogen, Grand Island, NY, USA). DNA sequencing libraries targeting the V3–V4 hypervariable region of the 16S rRNA gene were performed according to the Illumina 16S metagenomic sequencing library preparation method [29]. This consists of two PCR steps. In the first amplification, specific primers were used, while in the second, index information for sample identification was added. The DNA was amplified by primary PCR using universal primer pair with Illumina adapter overhang sequences, S-D-Bact-0341-b-S-1 (5′–TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG CCT ACG GGN GGC WGC A–3′) and S-D-Bact-0785-a-A-21 (5′–GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GGA CTA CHV GGG TAT CTA ATC C–3′) [15]. The PCR was performed with 2.5 μL of DNA sample (5 ng/μL), 5 μL each of the universal forward and reverse primer, and 12.5 μL of 2× KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Wilmington, MA, USA) in a total volume of 25 μL. The cycle condition comprised an initial denaturation at 95 °C for 3 min, followed by 25 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, and a final extension at 72 °C for 5 min. The PCR products were purified with AMPure XP beads (Agencourt Bioscience, Beverly, MA, USA) to remove free primers and primer-dimer species. Following purification, 2 uL of the primary PCR product was amplified for secondary PCR for library construction. Primer sequences used for the secondary PCR are as follows: a Nextera XT Index primer pair (Illumina®, USA), Primer 1 (N7xx): 5′–AAT GAT ACG GCG ACC ACC GAG ATC TAC AC–[i5]–TCG TCG GCA GCG TC–3′ and Primer 2 (S5xx): 5′–CAA GCA GAA GAC GGC ATA CGA GAT–[i7]–GTC TCG TGG GCT CGG–3′. The PCR consisted of 5 μL of sample DNA, 5 μL each of Nextera XT Index primers 1 and 2, 25 μL of 2× KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Wilmington, MA, USA), and 10 μL of PCR Grade Water. The cycle conditions comprised an initial denaturation at 95 °C for 3 min, followed by 8 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, and a final extension at 72 °C for 5 min. The final PCR product (final library) was cleaned up before quantification using AMPure XP beads (Agencourt Bioscience, Beverly, MA). Finally, the PCR products were quantified using qPCR according to the qPCR Quantification Protocol Guide (KAPA Library Quantification Kit for Illumina Sequencing platforms) and library quality was assessed using the TapeStation D1000 ScreenTape (Agilent Technologies, Waldbronn, Germany). Equimolar amounts of the barcoded V3–V4 amplicons were pooled and paired-end sequenced (2 × 300 bp) on an Illumina MiSeq® platform (Illumina Inc., San Diego, CA, USA) using v3 reagents, according to the manufacturer’s instructions at the Macrogen Inc. (Seoul, Korea).2.7. Sequence Data Processing and Metataxonomic AnalysisAfter sequencing was completed, Illumina MiSeq raw data was classified by sample using an index sequence, and a paired-end FASTQ file was generated for each sample. Sequences were demultiplexed, and barcodes and adaptors were removed using the Cutadapt v3.2 program [30].Denoising strategies were applied to obtain amplicon sequence variants (ASVs) using the Divisive Amplicon Denoising Algorithm 2 (DADA2) v1.18.0 [31] in the R program v4.0.3. For paired-end read, forward reads were truncated at 250 bp, reverse reads were truncated at 200 bp, and sequences with expected errors of 2 or more were excluded. Then, the error model for each batch was established to remove the noise for each sample. After assembling the paired-end sequence corrected for sequencing error into one sequence, the Chimera sequence was removed using the DADA2 Consensus method to form ASVs. In addition, for comparative analysis of the microbial community, the QIIME v1.9 [32] program was used to normalize by applying subsampling based on the number of reads of the sample with the minimum number of reads among all samples.Taxonomy was assigned to ASVs using the BLAST+ v2.9.0 [33] against the Reference Database (NCBI 16S Microbial DB). The taxonomy information for the organism of the subject with the highest similarity was assigned. At this time, if the query coverage of the best hit matching the database is less than 85% or the identity of the matched area is less than 85%, taxonomy information is not allocated.Using QIIME with the above ASVs abundance and taxonomy information, a comparative analysis of various microbial communities was performed. The ASV abundance was normalized by rarefying each sample such that all the samples had the same number of total counts (25,610 reads). Alpha diversity was assessed by Chao1 and Shannon index. Based on the weighted and unweighted UniFrac distances, beta diversity between samples (information on microbial community diversity among samples in the comparison group) was obtained, and the relationship between samples was visualized through PCoA.2.8. Statistical AnalysisData were analyzed using the general linear model (GLM) procedure of Statistical Analysis Systems (SAS) version 9.4 (SAS Institute Inc., Cary, NC, USA). All variables were tested for normality using the Shapiro–Wilk test. We used analysis of variance (ANOVA) to test for effects of different burial set-ups (UA, SA, UAn, and SAn), decomposition day (0, 5, 10, 30, and 60), and their interaction. The model used was:Yijk = µ + αi + βj + (αβ)ij + ϵijk where, Yijk is the response variable; µ is the overall mean; αi is the main effect of different burial set-ups (B); βj is the main effect of decomposition day (D); (αβ)ij is the interaction between the different burial set-up and decomposition day (B × D) and ϵijk is the random error of the kth observation from the (i, j, k)th treatment.Differences of moisture, pH, and alpha diversity index between burial set-ups on day 0, 5, 10, 30, and 60 were assessed using Tukey’s honestly significant difference (HSD) post hoc test. A p-value < 0.05 was considered statistically significant. Beta diversity was calculated based on the unweighted and weighted UniFrac distance matrix and the ordination plot was visualized using principal coordinate analysis (PCoA). We used permutational analysis of variance (PERMANOVA) to determine significant differences in beta diversity. A Venn diagram of the membership-based representation of unique, shared, and core bacterial community were generated using jvenn [34]. The bacterial abundance profiles were calculated at phylum, genus, and species levels and were plotted as a bar graph. Linear discriminant analysis Effect Size (LEfSe) was performed to determine the bacterial taxa that most likely explained differences between burial set-ups. For LEfSe analysis, only taxa (species level) with an LDA score of >2.0 and a p-value of 0.05, as determined by the Kruskal–Wallis rank-sum test, are shown.3. Results3.1. Effect of Oxygen Availability and Sterilized Soil in Moisture Content and pH during Carcass DecompositionThe moisture content and pH demonstrated significant differences (p < 0.05) between burial set-ups, days of decomposition, as well as the interaction of burial set-ups and decomposition day (Figure 2; Table S1). Moisture content ranged between 36.05% to 61.16% and became relatively constant from day 30 (Figure 2a). The moisture significantly varied between burial groups (p < 0.05) and across different time points (p < 0.05). Burial set-ups under anaerobic conditions (UAn and SAn) had a higher moisture content compared to those in aerobic conditions. The burial set-ups (UAn and SAn) in anaerobic conditions have high moisture at days 5 to 10, decreased from day 30 and became relatively stable until day 60. Meanwhile, the moisture in the burial set-ups (UA and SA) under aerobic conditions tended to increase until day 30 and become stable until day 60. On day 60 of decomposition, higher moisture was observed in the sterilized soil-anaerobic burial set-up.The pH significantly differs between burial groups (p < 0.05) and across different sampling days (p < 0.05) (Figure 2b; Table S1). The pH of burial set-ups UA and SA under aerobic conditions was higher (p < 0.05) than those in anaerobic conditions. The pH in UA and SA decreased on day 10 and then increased on day 30 and become relatively stable until day 60. Meanwhile, the pH in UAn and SAn increased until day 30 and then become stable until day 60. In addition, the pH of the samples increased as the decomposition process progressed and remained relatively constant from days 30 to 60. At the end of day 60 of decomposition, a higher pH (p < 0.05) was observed in both the sterilized and unsterilized soil-aerobic condition burial set-ups (SA and UA) compared to other burial set-ups.3.2. Species Richness and Diversity of Bacterial Community in Different Burial Set-UpsA total of 6578 ASVs were obtained from all samples through the Illumina sequencing analysis. Rarefaction curves revealed that all samples were sequenced to sufficient depth to achieve asymptote, indicating a good representation of the microbial community (Figure S1). Burial set-ups, days of decomposition, and their interaction affect significantly both Chao1 and Shannon indexes (Figure 3; Table S2).Chao1 differed significantly between burial set-ups at all time points (p < 0.05) (Figure 3a). During the initial day (day 0), Chao1 was significantly higher in UA and UAn burial set-ups compared to SA and SAn burial set-ups (p < 0.05). On day 5, it was significantly higher in UA compared to SA, UAn, and SAn burial set-ups (p < 0.05). On day 10, Chao1 was significantly higher in UA and UAn burial set-ups compared to SA and SAn burial set-ups (p < 0.05). Meanwhile, Chao1 varied significantly between the burial set-ups on day 30 (p < 0.05), with UAn burial set-up having the highest Chao1. On the other hand, Chao1 was significantly higher in UAn compared to other burial set-ups on day 60 (p < 0.05). We observed that the Chao1 in UA burial set-up significantly decreased as decomposition progressed, while UAn, SA, and SAn tended to decrease from days 5 to 30 and increased in day 60 (p < 0.05).The Shannon index differed significantly between burial set-up, day of decomposition, and their interaction (p < 0.05) (Table S2). The Shannon index significantly varied between burial set-ups at days 0, 5, 10, and 60 (p-value = 0.0022, 0.0092, 0.0209, and 0.0297, respectively) (Figure 3b). Burial set-ups UA and UAn had a significantly higher Shannon index compared to SA and SAn burial set-ups at days 0, 5, and 10 (p < 0.05). Meanwhile, no significant difference was found among the burial set-ups on day 30, however, the Shannon index decreased on this day. On day 60, the Shannon index increased, with UAn burial set-up having the highest Shannon index among the other burial set-ups (p = 0.0297).The Venn diagram showed the similarities and differences between the communities in the different burial set-ups (Figure 4; Tables S3 and S4). The shared taxa by all samples in each burial set-up were deemed to be core bacterial communities. From 1992 ASVs, 328 bacterial ASVs were shared among all burial set-ups. Meanwhile, a total of 824 ASVs were identical in two or three groups, while 338, 135, 280, and 87 ASVs were unique to UA, SA, UAn, and SAn burial set-ups, respectively.Based on the unweighted PCoA, the principal component 1 (PC1) and PC2 analysis showed 23.33% and 8.74% of variance explained, respectively (Figure 5a). The bacterial communities were clustered based on the unsterilized and sterilized burial group set-up and by decomposition period. The unsterilized soil group (UA and UAn) at the initial day of decomposition formed a distinct cluster, separate from the other samples. Similarly, the samples in the sterilized soil group (SA and SAn) formed a cluster together and separated from the other samples. This indicates that the bacterial composition from these clusters was different from those in days 5 to 60. Meanwhile, the bacterial community at days 5, 10, 30, and 60 of UA and UAn were clustered together. Likewise, SA and SAn burial groups were clustered together at days 5, 10, 30, and 60. On the other hand, weighted PCoA revealed 30.30% and 22.91% variation for PC1 and PC2, respectively (Figure 5b). The plot showed that the samples from SA and SAn formed a distinct cluster. Similarly, UA and UAn also formed a cluster together. Meanwhile, the bacterial communities of different burial set-ups were similar at day 30. This indicates that the bacterial composition from those clusters was likely composed of similar microorganisms.3.3. Bacterial Community Composition during Poultry Carcass DecompositionA total of 23 phyla, 61 classes, 133 orders, 293 families, 901 genera, and 1992 species were identified in the samples. The bacterial community was mainly comprised of Firmicutes, Proteobacteria, and Actinobacteria, with a relative abundance of 77.53%, 11.53% and 6.76% of the total ASVs in all samples, respectively.On the initial day (day 0), Firmicutes dominated the SA and SAn burial set-ups, with the relative abundance of 81.07% and 80.79%, respectively (Figure 6). Meanwhile, UA and UAn were mostly composed of Firmicutes, Proteobacteria, Actinobacteria, Acidobacteria, Chloroflexi, and Bacteroidetes. At the early stage of decomposition (day 5), the abundance of Firmicutes increased in the UAn, and SAn burial set-ups but decreased in SA and UA burial set-ups. In contrast, Proteobacteria increased in UA, but decreased in UAn, SA, and SAn. In the following days (10–60), an increase of Firmicutes and a decrease of Proteobacteria could be observed. Meanwhile, Actinobacteria is more abundant in burial set-ups with unsterilized soil than with sterilized soil. In addition, the abundance of Actinobacteria tended to decrease from day 5 in all burial set-ups.At the genus level, Bacillus, Clostridium, Pseudescherichia, Limosilactobacillus, Neobacillus, Rummeliibacillus, Lentibacillus, Lactobacillus, Pseudogracilibacillus, and Anaerosalibacter were the most common genera in the decomposing carcass (Figure 7). On the initial day, Limosilactobacillus and Lactobacillus were the most abundant genera in all burial set-ups. Meanwhile, Gaiella was abundant in burial set-ups with unsterilized soil (UA and UAn) on the initial day. However, the relative abundance of Gaiella in those set-ups decreased as the decomposition progressed.In the UA burial set-up, Pseudescherichia and Neobacillus predominated the bacterial community on days 5 and 10; Bacillus (64.18%) on day 30 and Pseudogracilibacillus (27.81%) and Bacillus (15.54%) on day 60 of decomposition. In the UAn burial set-up, Clostridium (23.36%), Neobacillus (20.58%), Limosilactobacillus (14.08%), and Lactobacillus (11.02%) were dominant on day 5; Clostridium (27.23%), Rummeliibacillus (15.07%), and Limosilactobacillus (10.41%) on day 10; Bacillus (58.27%) and Clostridium (12.51%) on day 30 and Lentibacillus and Clostridium (18.32%) on day 60 of decomposition. In the SA burial set-up, Pseudescherichia (58.62%) and Neobacillus (10.22%) was dominant on day 5; Pseudescherichia (40.07%) and Clostridium (14.02%) on day 10; Bacillus (59.91%) on day 30 and Pseudogracilibacillus (39.08%) on day 60 of decomposition. In the SAn burial set-up, Clostridium (26.93%), Limosilactobacillus (26.04%), and Neobacillus (15.01%), and Lactobacillus (11.52%) were dominant on day 5; Rummeliibacillus (40.94%), Clostridium (29.27%), and Limosilactobacillus (12.94%) on day 10; Bacillus (67.34%) and Clostridium (10.05 %) on day 30 and Lentibacillus (27.27%), Tissierella (26.97%), Clostridium (20.85%), and Anaerosalibacter (14.01%) on day 60 of decomposition. The relative abundance of Bacillus significantly increased on day 30 in all burial set-ups and decreased on day 60 of decomposition. On day 60, Pseudogracilibacillus was abundant in UA (27.81%) and SA (39.08%), while Lentibacillus was abundant in UAn (32.26%) and SAn (27.29%).At the ASV (species) level, the results showed that the bacterial community composition varied in different burial set-ups throughout the decomposition process (Figure 8). During the initial day, we identified that Limosilactobacillus reuteri was predominant in all burial set-ups. On day 60, the relative abundance of Pseudogracilibacillus endophyticus under aerobic conditions (UA and SA) was high compared to other species. On the other hand, Lentibacillus lacisalci dominated the community on day 60 in the burial set-up under anaerobic conditions (UAn and SAn). LEfSe analysis revealed that Bacillus cereus and Gaiella occulta were the species that showed significant increase in abundance in UA and UAn burial set-ups, respectively (p < 0.05) (Figure S2). In terms of day of decomposition, Megamonas hypermegale, Blautia glucerasea, and Cuneatibacter caecimuris were significantly higher on day 0 both in SA and SAn burial set-ups (p < 0.05). Neobacillus jeddahesis significantly increased at day 5 in UAn and SAn; Bacillus kyonggiensis at day 10 in SA burial set-up, while Schnuerera ultunensis and Urmitella timonensis at day 30 in both UA, SA, and UAn burial set-ups. In addition, Tissierella praeacuta and Anaerosalibacter bizertensis were significantly abundant on day 60 in SAn burial set-up; Sporanaerobacter acetigenes and Clostridium acetireducens in UAn; Cerasibacillus quisquiliarum in UA and Lentibacillus lacisalci were significantly high in abundance on day 60 of decomposition in UAn and SAn burial set-ups (p < 0.05).4. DiscussionIn this study, we investigated if the moisture content and pH were affected during the decomposition of poultry carcasses in vitro. The study also employed 16S rRNA Illumina MiSeq high-throughput sequencing using the ASV approach method to determine the changes in the bacterial communities in different burial set-ups of decomposing poultry carcasses.The results of the study revealed that moisture content was higher in the burial set-ups under anaerobic conditions than those in aerobic conditions. This suggests that during the decomposition under anaerobic conditions, the water from the decomposing material was extracted, and moisture was evaporated and accumulated within the container. Several studies reported that the moisture content during decomposition increases as the temperature increases [35,36]. Since there was no oxygen supply in the anaerobic burial set-ups, the heat increased and accumulated inside the container, hence, the bound water in the decomposing body was released and moisture content increased [35]. At this period, the microbial activity would increase, as well as the production of enzymes by the microorganisms which aids in the decomposition of the carcasses.In terms of pH, the pH varied in each burial set-up and across time. We observed a higher pH in burial set-ups under aerobic conditions compared to those in anaerobic conditions. Several studies reported that an increase in pH is due to the breakdown of nitrogen-containing organic matter, which leads to the accumulation of NH3 that dissolves in moisture to form alkaline NH4+ [37,38]. Thus, the increase in pH may be related to the moisture content and is important during carcass decomposition. In an actual livestock burial site, there is a lack of oxygen during the decomposition process and as a result the moisture content will increase and produce a large amount of leachate which may slow down the decomposition process. Meanwhile, a decrease in pH is attributed to the reduced production and increased release of NH3 and the subsequent release of ions [35,37].The microbial diversity can be determined by several factors including physical, chemical, and biological characteristics of an ecosystem [39,40]. Carcass decomposition is directly associated with the alteration of bacterial community due to the breakdown of tissue, which releases an ammonia-rich, high-nutrient fluid that alters both the pH and nutrient content in the soil [41,42,43]. Microorganisms are key players during the decomposition of carcasses [11,44,45,46]. The Chao1 and Shannon index varied between different burial set-ups and across different sampling days during decomposition. Our findings demonstrated that the Chao1 and Shannon index decreased starting from day 5 of decomposition. This could be due to the different burial conditions and effect of the soil moisture and pH during decomposition [10,11,46,47]. In addition, we found that more bacterial taxa were found in the burial set-ups with unsterilized soil (UA and UAn) compared to burial set-ups with sterilized soil (SA and SAn). This indicates that the removal of intact microbes in soil by sterilization treatment significantly affected the diversity during the decomposition of poultry carcasses [11]. However, the burial set-up with unsterilized soil decomposed under anaerobic conditions (UAn) had a higher Chao1 and Shannon index compared to other burial set-ups at day 60. We speculate that as the decomposition progressed, the environment became anaerobic leading to an increase in anaerobic microorganisms. The breakdown of tissue is mainly dominated by anaerobic autolysis and later by microbe and insect infiltration [48], which could explain the dominance of anaerobic microorganisms. Furthermore, we observed that as pH increased during decomposition, the bacterial species present in the burial set-ups was decreased. The trend in pH was similar to the changes seen in bacterial communities, indicating that the bacterial communities were influenced by the pH. According to Rousk et al. [49], the relative abundance and diversity of bacteria were positively associated due to the narrow pH ranges for optimal growth of bacteria. Moreover, the survival of microbes is dependent on the pH, thus, when pH increased, the microbes which can survive at a higher pH become dominant [50]. Several studies reported that the pH of the soil is an important abiotic predictor associated with bacterial diversity [50,51,52,53]. Bacterial communities are more diverse in soils with near-neutral pH than in acidic or alkaline soils [50,51]. We observed that the alpha diversity was highest in all burial set-ups in the initial day than those found in other periods. In addition, these samples had near-neutral pH (6.5 ± 0.5) than the other samples. Likewise, the sterilization treatment of soil had a role in the diversity of microorganisms during the decomposing carcasses. Our findings were in accordance with the study of Lauber et al. [11], wherein a decrease in the alpha diversity and shift in the taxonomic composition was reported in the burial set-up with sterilized soil. These results suggested that the soil bacterial community was affected when the soil microbes were removed by sterilization treatment. The presence of these soil microbes can influence the decomposition process. These microbes can be characterized and utilized in order to promote the decomposition of livestock carcasses in an actual burial site. In addition, when comparing the burial set-ups by oxygen condition, the Chao1 in the unsterilized soil-anaerobic condition burial set-up (UAn) was higher compared to other set-ups. Microbial studies suggest that anaerobic bacteria flourish at most times during decomposition. According to Janaway [54], aerobic bacteria flourish during the early stages of the decomposition process as there is oxygen present within the body. Furthermore, as the microbial population increases, the accumulation of gases during the decomposition process makes the environment anaerobic which prompts the microbial community to shift [54].Based on the PCoA, the comparison of bacterial communities between different burial set-ups at each time point showed significant variation. The bacterial species during early decomposition differs in the late decomposition. During the initial day, the bacterial community in the unsterilized soil burial setups (UA and UAn) and the sterilized soil burial setup (SA and SAn) were different from each other. This is probably due to the sterilization treatment which eradicated the intact microbial community in the soil. Similarly, a decomposition study by Lauber et al. [11] reported that bacterial communities of carcasses buried in soil with an intact microbial community and soil without an intact microbial community vary from each other. Moreover, several studies reported that bacterial communities vary significantly at different stages of decomposition [55,56]. However, as the decomposition progressed, we observed changes in the bacterial species present at each period across all burial set-ups. Based on these results, it showed that the presence of an intact microbial community in soil and oxygen availability during carcass decomposition influenced the bacterial community in the burial set-ups. Meanwhile, samples from days 5 to 60 showed a close association and were clustered by unsterilized or sterilized soil. This indicates that during this period, the burial set-ups that were clustered together had a similar bacterial composition. We speculate that these might be related to the moisture content and pH during the decomposition process, wherein it showed a stable pattern from days 30 to 60. The changes in the moisture content and pH during this period influenced the bacterial composition in the burial set-ups. Since the moisture content and pH were relatively stable at this point, we hypothesize that the bacterial community towards the end of decomposition also showed similar taxonomic composition.Firmicutes, Actinobacteria, and Proteobacteria were the predominant phyla during carcass decomposition in all burial set-ups. Similar to other studies, these phyla were reported to be associated with the decomposition of carcasses [43,46,52,57,58]. Variation in the taxonomic composition was observed during the decomposition process. The changes in bacterial communities suggest that various bacterial species play a role during the decomposition period. During early decomposition, Proteobacteria and Actinobacteria were abundant; however, Firmicutes dominated the community as the decomposition process progressed. A decomposition study by Pechal et al. [46] showed that Proteobacteria declined in abundance, while Firmicutes increased in abundance. In this study, we observed that the abundance of Proteobacteria decreased on day 10, while the abundance of Firmicutes increased as the decomposition progressed. The taxa within Proteobacteria are commonly associated with the spoiling of meat and have been found on the hides of slaughtered animals [46]. Proteobacteria are common in soil and play an important role in the decomposition of fats and carbohydrates [57]. Meanwhile, Firmicutes are associated with the gut microbiome [58] and soil communities [52]. Firmicutes are involved in reducing large macromolecules, such as proteins, complex fats and polycarbohydrates to their building blocks [58]. Thus, they are more prominent during active decomposition [43,59]. Furthermore, during the initial day of decomposition, we observed that Firmicutes became more abundant in the burial set-ups with sterilized soil than in burial set-ups with unsterilized soil. Similar to other studies [43,57,58,59], Firmicutes gradually increase during the early process of decomposition (day 5), whereas Proteobacteria decrease as decomposition progressed. On the other hand, Actinobacteria are widely distributed in both terrestrial and aquatic ecosystems, especially in soil, where they play an important role in the recycling of refractory biomaterials through the decomposition of complex mixtures of polymers in dead plants, animals and fungal materials, such as chitin, keratin, and lignocelluloses [60,61].The bacterial community at the genus level revealed that Bacillus, Clostridium, Pseudescherichia, Lactobacillus, Rummeliibacillus, Limosilactobacillus, Neobacillus, Lentibacillus, Pseudogracilibacillus, Anaerosalibacter, Gaiella, and Sporanaerobacter were among the common genera found in the burial set-ups. These microbes are potentially pathogenic in soil samples during carcass decomposition. In this study, the abundance of Lactobacillus was high during the early days and decreased as decomposition progressed. Similarly, Wang et al. [62] reported a sharp decrease in the abundance of Lactobacillus after 16 days of composting poultry carcasses. Members of Lactobacillus are associated with the decomposition of lipids and complex carbohydrates associated with animal carcasses [62]. Meanwhile, Clostridium was detected in the early period of decomposition in burial set-ups in anaerobic conditions. Clostridium species are ubiquitous anaerobic putrefactive bacteria that can be found in various environments and are potentially pathogenic [54,63,64]. Several decomposition studies have reported that the increase in abundance of Clostridium is due to the anaerobic layer that develops around a carcass during decomposition [23,65]. Clostridium species play an important role in biomass digestion as they synthesize a wide variety of extracellular enzymes which aids in the degradation of various compounds, such as carbohydrates, amino acids, alcohols, amino acids, or purines [23,65,66,67,68]. In the present study, Gaiella occulta was found abundant in the burial set-ups with unsterilized soil. Gaiella sp. which belongs to Actinobacteria were commonly found in soil [69,70]. This suggests that Gaiella occulta was intact in the unsterilized soil since this species was not found in the burial set-ups with sterilized soil. Meanwhile, Megamonas hypermegale, Blautia glucerasea, and Cuneatibacter caecimuris were found in SA and SAn burial set-ups during the initial day. These microorganisms are reported to be associated with the poultry intestinal microbiota [71,72,73,74]. Pseudescherichia vulneris is an environmental microorganism that can colonize humans and animals [75,76]. P. vulneris was reported as one of the contaminants in slaughtered poultry carcasses [77]. Rummeliibacillus sp. and Neobacillus sp. were also detected in the decomposing poultry carcasses in the different burial set-ups, however, very little is known about these microorganisms and their involvement in the carcass decomposition. Rummeliibacillus suwonensis is a Gram-positive, facultatively aerobic, rod-shaped, non-motile, terminal spore-forming bacterium which was first isolated in soil from Suwon, Korea [78]. Neobacillus jeddahensis is an aerobic, Gram-positive, rod-shaped, mesophilic bacterium that was isolated from the feces of a man from Jeddah, Saudi Arabia [79]. Limosilactobacillus reuteri was found in all burial set-ups, which indicates that this species originated from the animal. L. reuteri is a bacterium that is found in a variety of natural environments including different body sites, such as the gastrointestinal tract, urinary tract, and skin in a human and a large number of mammals [80]. Sporanaerobacter acetigenes is a strictly anaerobic, moderately thermophilic, and sporulating bacteria. Sporonaerobacter sp. are commonly observed in anaerobic utilization of municipal wastes, activated sludge, and decomposition of entombed pigs [81]. They are important members of the bacterial community for the destruction of the protein fraction of complex substrates in the form of volatile fatty acids (VFA). Anaerosalibacter bizertens is an anaerobic, spore-forming, thermophilic bacterium that was isolated from sludge and is reported to be a reducing species during the decomposition process [64,82]. We detected a significantly high abundance of this bacterium in the burial set-ups under anaerobic conditions (UAn and SAn) at day 60 of decomposition.At the end of the decomposition period, the prevalent genus for burial set-ups under aerobic conditions is Psedogracillibacillus, while the most prevalent in the burial set-up under anaerobic conditions is Lentibacillus. Pseudogracilibacillus is an aerobic, spore-forming, gram-positive bacterium [83]. We detected Pseudogracilibacillus endophyticus in the burial set-ups under aerobic conditions (UA and SA), this suggests that this bacterium may be positively associated with aerobic conditions during decomposition. Lentibacillus is an aerobic or facultatively anaerobic, gram-variable, endospore-forming, moderately halophilic bacterium [81,84]. Lentibacillus was among the dominant bacteria detected in a composting system along with other anaerobic, thermophilic, endospore-forming, and/or halophilic gram-positive bacteria, such as Pelotomaculum, Clostridium, and Caldicoprobacter [45]. In this study, we detected a high abundance of Lentibacillus lacisalsi in the burial set-up under anaerobic conditions (UAn and SAn). This could indicate that this bacterium may be positively associated to an anaerobic condition during poultry carcass decomposition. P. endophyticus and L. lacisalsi both belong to Firmicutes; however, it is unclear at this time on their specific involvement in the carcass decomposition. These species of bacteria may be associated in the soil and have a role in decomposition with the metabolites they produce. The changes in the abundance at genus and species level could be attributed to the availability of oxygen rather than the soil condition as time goes by in each burial set-up.5. ConclusionsIdentification of microorganisms and changes in the community structure is important in order to determine the species involved and their possible role in the community during carcass decomposition. Our findings suggested that the presence of microbes in soil and oxygen availability significantly influenced the changes in moisture content, pH, bacterial abundance, and community composition during the process of decomposition of carcasses in vitro. This study provided plausible information on the possible bacterial species involved in the decomposition of poultry carcasses. Also, this study could be used to utilize potential microbes to increase the decomposition rate of animal carcasses and antagonistic action against contagious animal pathogens, such as avian influenza under actual burial conditions.

animals : an open access journal from mdpi

10.3390/ani12070846

PMC8996930

We tested two computer-vision-based indexes to analyze the rearing-environment enrichment on broiler movement as a function of comfort temperature and heat stress. The results indicated that the simultaneous application of cluster and unrest indexes could monitor the movement of the group of broilers under different environmental conditions. Future monitoring and alert systems based on computer vision should consider the complexity of the environment for detecting heat stress in broiler production.

Computer-vision systems for herd detection and monitoring are increasingly present in precision livestock. This technology provides insights into how environmental variations affect the group’s movement pattern. We hypothesize that the cluster and unrest indexes based on computer vision (CV) can simultaneously assess the movement variation of reared broilers under different environmental conditions. The present study is a proof of principle and was carried out with twenty broilers (commercial strain Cobb®), housed in a controlled-environment chamber. The birds were divided into two groups, one housed in an enriched environment and the control. Both groups were subjected to thermal comfort conditions and heat stress. Image analysis of individual or group behavior is the basis for generating animal-monitoring indexes, capable of creating real-time alert systems, predicting welfare, health, environment, and production status. The results obtained in the experiment in a controlled environment allowed the validation of the simultaneous application of cluster and unrest indexes by monitoring the movement of the group of broilers under different environmental conditions. Observational results also suggest that research in more significant proportions should be carried out to evaluate the potential positive impact of environmental enrichment in poultry production. The complexity of the environment is a factor to be considered in creating alert systems for detecting heat stress in broiler production. In large groups, birds’ movement and grouping patterns may differ; therefore, the CV system and indices will need to be recalibrated.

1. IntroductionProspects of future scenarios indicate that the world population will grow to 9.3 billion people in 2050, which requires a significant increase in food demand. Globally, chicken meat is expected to represent 41% of all animal-protein sources by 2030 [1]. In order to meet this strong demand for market growth, intensive production of broilers has prevailed. In most of these systems, the rearing environment restricts opportunities for species-specific behaviors, which are essential for good welfare [2]. Broilers are housed at a high density and with selected genetic characteristics for rapid growth [3]. However, the bone structure of the chicken did not follow this process of high development of the upper body part (breast), which triggers leg disorders and the consequent loss of mobility with increasing body weight [4]. The automatic detection of the activity level of groups of broilers makes it possible to identify deviations outside the expected patterns and generate real-time notification alerts to the producer, which allows a faster readjustment with benefits for the welfare of the animals [5]. Stimulating physical activity in birds prevents the occurrence of locomotor problems that impair wellbeing [6,7]. Previous studies indicate that enriched environments have the positive potential to stimulate and increase the activity level of broilers [4,8,9,10,11,12]. Physical activity strengthens the locomotor system, especially at the beginning of the growth phase [13]. In summary, environmental enrichment introduces improvements in existing production systems and considers which artifacts stimulate the behavioral activities inherent to the species, promoting improvements in biological function [14,15]. Computer vision (CV) applies mathematics and computer science to provide image-based automated process control [16]. CV allows continuous and real-time measurements during the flock production cycle in a fully automated, noninvasive way [17]. The data images were collected past steps to preprocessing, segmentation (region of interest), features extraction, and classification or regression [16]. Thus, the producer can monitor various biological processes and bioresponses related to animal welfare, health, feeding and drinking behaviors, and flock productivity [18,19,20]. We hypothesize that computer vision associated with movement indexes can monitor locomotor-health problems and prevalence in broiler flocks [5,21,22,23,24]. Studies based on proof of concept are present in animal production and evaluate the technical, practical, and financial feasibility of an idea or hypothesis [25,26]. Several studies have been developed to monitor locomotor-health [22,27,28,29] body-mass estimation [30,31]. The effect of environmental enrichment on broiler activity levels, gait assessment, locomotor problems, zootechnical performance, and behavior and wellbeing have been previously studied using video-image-processing techniques [32,33,34,35]. The computer-vision technology was also validated in a laboratory scale for automatic monitoring and gait-score classification [36]. It was also used to identify abnormal deviations in the activity level of commercial birds [5] and for evaluating the occupancy rate of laying hens in compartments with different levels of ammonia concentration [37]. Under heat-stress conditions, there is a significant decrease in growth rate, increased mortality, a compromised immune system, loss of meat quality, behavioral changes, and a decreased level of wellbeing [38,39]. The rapid diagnosis of animals in thermal discomfort is crucial to prevent the stress from being prolonged, preserving broiler performance, health, and welfare. The continuous analysis of image processing obtained by video cameras allows for the generation of activity indexes that monitor the thermal state of broilers [40,41,42]. Most studies in the current literature involving heat stress were conducted from 21 days of age [43]. Animals change their behavioral pattern as a function of the rearing temperature, being close to each other when subjected to cold or spread out in the environment in the heat [42]. Livestock workers have routinely used these postural patterns to assess thermal comfort and adjust to environmental management settings [44]. Thus, observation of behavioral parameters is a noninvasive way of detecting heat stress [45]. Previously CV has been used for the generation of cluster and unrest indices, developed respectively by [23] and [24], and have been applied as indicators of thermal comfort in commercial poultry production. The results obtained indicated that the unrest index could detect the agitation of poultry under different thermal conditions, with a significant decrease in the movement of birds under heat stress [24]. On the other hand, the cluster index revealed a significant difference in the clustering behavior of birds under conditions of comfort and heat stress. In addition, it identified behavioral differences between the heavy rearing breeds [23]. The unrest index was used to measure the walking ability of broilers with different gait scores [46]. Both indices have the potential to develop a remote-monitoring system to accurately detect differences in the behavior of birds raised in floor pens bedded with wood shavings. The application of these indices has not yet been explored in the rearing of broilers in enriched environments.This study is a proof of concept to assess the use of the cluster and unrest indexes simultaneously to test the sensitivity and viability in order to evaluate the movement of broilers under different conditions such as heat stress and pen enrichment.2. Materials and Methods2.1. Description of the Controlled-Environment ChamberThe controlled-temperature room has three compartments (C1, C2 and C3), measuring 1.6 × 1.4 × 3.0 m3. Only two compartments were used for the present study, and they were randomly selected. In each compartment, a manual tube-type feeder (Zatti® Model number 181,528, Zatti Industry and Commerce, Coronel Freitas, Santa Catarina, Brazil), an automatic pendulum-type drinker (CASP®, Model pendular drinking automatic 2003, CASP, Amparo, São Paulo, Brazil), and a temperature and humidity sensor were installed close to the animals’ level, at a distance of 0.40 m from the floor. Each compartment had an air conditioner, two dehumidifiers, two heaters, a dimmable LED lamp to control light intensity (lx), and a video camera. Each compartment was accessed independently through a door (0.7 m wide × 2 m high). The computers responsible for managing the experimental units were installed in the support room (climate control and video recording). Figure 1 [47] shows the schematic with the respective positioning of all the equipment used in the controlled-temperature room and the technical-support room during the experimental period.The environmental-control center manages each compartment of the controlled-environment room using software developed in the Delphi programming language (version 6.0, Borland Software Co., Austin, TX, USA). The software allows measuring, processing, controlling, and recording continuously collected data. This system allows the user to check temperature, humidity, light intensity, and air renewal rates in real time. According to the established temperature and humidity values, the equipment is automatically activated (on and off). Relative humidity was programmed to remain at 60% continuously. Only the air temperature varied during the experimental period, with operating values recommended by the [48] of 23 °C for thermal comfort and 31 °C for heat-stress treatment. The particularities of the environmental-control system, equipment, operating limits, stability, and validation with broilers are presented in [47]. The steps of inputs, data collection and processing, experimental treatments, and outputs are represented in Figure 2 and explained in the sequence.2.2. Image AcquisitionFor the animal-behavior monitoring data and further analysis, surveillance cameras (Intelbras® VMD 3120 IR, Intelbras Corporation, São José, Santa Catarina, Brazil) with a resolution of 976 × 496 (H × V) and automatic activation of the infrared device in cases of low light were installed on the ceiling of the geometric center of each compartment. The two validation tests used recorded video images between noon and 18:00 h. Video recordings were automatically stored on an NVR video recorder (Intelbras® Multi HD Serie 1000, 1080p, Intelbras Corporation, São José, Santa Catarina, Brazil). Figure 3 shows the areas observed for the unenriched (a) and enriched (b) compartments.2.3. Birds and HusbandryA total of thirty-day-old mixed-sex chicks of the Cobb® strain were obtained from a commercial farm. Twenty chicks with similar weights and the same distribution of males and females were selected in two treatments, each containing ten birds (without environmental enrichment and with environmental enrichment). The compartments and animals were randomly assigned to assign the treatments on the first day of housing in the controlled-environment chamber.Both compartments were kept without environmental enrichment during the first three days of adaptation in the climatic chamber. The compartment selected as “enriched” was provided with colored plastic rings suspended by a string, a plastic box containing fine sand, and a wooden perch. According to previous literature, the enrichment was selected to positively affect the birds’ natural behaviors (perching, pecking, and dust bathing) [13,49,50]. A bell drinker and feeder were placed in each compartment. Water and commercial feed, based on corn and soybean meal, were provided ad libitum throughout the rearing period. The supply of commercial feed and water to the birds was ad libitum throughout the experimental period and followed the nutritional recommendations of the breeder’s manual [48]. Once a day, the offered ration was weighed and manually inserted into the tube feeder. The automatic water-supply system used a tubular drinking fountain with height adjustment. The floor was covered with shavings bedding (0.05 m). We adopted the breeding-company-recommended period of light (24 h of light until the seventh day and increasing 1 h of darkness every two days). On the 14th day of growth, the birds remained exposed to 20 h of light and 4 h of darkness from 21:00–01:00 until the end of the experiment (42nd day of growth). We also adopted the breeding manual, so the broilers were kept in thermoneutrality conditions during the first to the twentieth day of growth [48].The acquisition of video images was performed automatically for seven consecutive hours from noon to 18:00 h for the two consecutive days of analysis. According to a previous study [5], broilers’ activity patterns were similar for three weeks throughout the day. Such assumption allowed us to validate the analysis of broiler movement through the cluster and unrest indexes in two days for experimental conditions in the controlled-environment room, characterizing the present study as a proof of concept. The age of 21 days is when the heat stress starts to impair productive performance (decrease in feed consumption and weight gain) and negatively challenge animal metabolism and immunity [43,51,52]. Our experimental tests were precisely at the age of 21 and 22 days under thermoneutrality and heat-stress conditions, respectively, for both treatments. Heat stress was tested for one day, with the chickens at 22 days of age, with the birds being kept in thermal comfort during the previous housing days. Figure 4 illustrates the diagram of activities used in data collection for proof-of-concept validation.The relative humidity remained within what was recommended in the breeders’ manual [48] during the experiment. A previous commissioning study of the controlled-environment-room operation [47] allowed one hour to reach the heat-stress temperature condition (31 °C) and maintain the system. For this reason, the control setup started at noon.2.4. Video AnalysisThe efficiency of two comfort indexes based on group behavior was verified. The calculation of these indexes is based on information extracted from images recorded through image-analysis techniques. This proof of concept evaluated the extraction of information and the calculation of indexes by a computer-vision system.Videos were analyzed at the frequency of one frame per second (fps). Considering that there was no effect of the compartments, we used a completely randomized design in a split-plot scheme in time, in which we tested two factors: (1) temperature (comfort or heat stress), as the main factor; and (2) environmental enrichment (present or absent), as the secondary factor. The seven hours of recording analyzed were divided into 14 blocks of time (30 min each block, which corresponds to the analysis of 25,200 frames per condition (temperature vs. environment), totaling 100,800 frames in the experimental period. The images were processed frame by frame, initially using low-pass filters to smooth out image noises such as feathers on the bedding and wood shavings on the birds. After segmentation, mathematical morphology techniques were applied to fill holes and exclude the remaining noise. The group behavior of chickens was measured using the cluster index [23] and the unrest index [24], described in Equations (1) and (2). (1)Cluster Indexi=2×A¯×h2+w2P¯×D¯×nA−1 where Cluster Index(i) is the cluster index of the birds observed in the ith frame of the video; A¯ and P ¯ are the average area and perimeter (in pixels) of the shapes observed in the frame, respectively; D¯ is the average distance between the centers of mass of the shapes in the scene; nA are the number of clusters; and h and w correspond to the height and width (in pixels) of the cropped image. (2)Unrest Index(i,i−1)=k.max{dH(F(i),F(i−1)), dH(F(i−1),F(i))} where Unrest Index(i, i−1) is the unrest index (cm) of the birds between two frames recorded with 1 (one) second difference; i is the position of the frame in the video; F(i) is the current frame; F(i−1) is the previous frame; dH is the Hausdorff distance [53] between birds from one frame to another; and k is the proportionality factor calculated by Equation (3). (3)k=2Htan(α/2)w where k is the proportionality factor; H is the height (cm) of the installed camera concerning the floor; α is the opening angle of the camera lens; and w is the length (pixels) of the CCD sensor, which corresponds to the length of the largest measurement of the frame captured by the camera. The video capture rate was 30 fps, but a frequency of 1 fps was adopted as the most adequate for image analysis, considering the birds’ movement speed.Cluster and unrest indexes were calculated frame by frame. In this way, the values obtained for each plot correspond to an average referring to 1800 images. The data were explored by graphs of the indexes calculated in the simultaneous application observation time, verifying possible interactions and differences in the crowding and unrest behaviors between the enrichment and temperature treatments evaluated. We applied the ANOVA with repeated measures, followed by the Tukey mean test to confirm the differences in the birds’ crowding and movement behavior between the evaluated environmental treatments.3. Results and DiscussionFigure 5 corresponds to the results obtained for the cluster index (crowding) from noon to 18:00 in the group of broilers housed in an enriched and nonenriched environment, subjected to thermal conditions of neutrality and heat stress. Each point on the graph represents a repetition of a subdivision plot in time (total 14). Each plot corresponds to the analysis average of 1800 frames/video for generating the cluster index for the treatments.It can be seen from Figure 5 that the cluster index detected in the enriched environment is similar between the conditions of thermoneutrality and heat stress. However, the highest peaks occurred in thermoneutrality and the lowest in heat stress. This observation of trend analysis means that the ambient temperature above the comfort limit was not enough to change the crowding pattern of broilers aged 21 days reared in enriched environments. Therefore, the isolated analysis of this index cannot be considered an indicator of heat stress for broilers raised in enriched environments. We observed significantly lower crowding rates in the comfort temperature and nonenriched environment than the enriched-compartment results.Figure 6 illustrates the unrest index for the treatments from noon to 18:00 h. Each point of the graph represents a repetition of the subdivision plot in time (total 14). Each plot corresponds to the average of 1800 frames/video analysis for the generation of the unrest index of the treatments.During the evaluated period, a tendency for the unrest index to be higher under thermoneutrality conditions is observed, regardless of the environmental enrichment. Furthermore, heat stress reduces the movement of animals, with more significant losses for the nonenriched treatment. Table 1 shows the differences observed in the behavior of gathering and movement of the birds, where it is observed that the environmental enrichment promoted more movement of the birds both under conditions of comfort (52.74 vs. 48.71) and thermal stress (35.38 vs. 32.81), which confirm previous graphical analyses. However, heat stress is a limiting factor for the movement of animals, reducing the positive potential of the presence of environmental enrichment. Environmental enrichment provided higher crowding rates both in comfort (8.42 vs. 4.66) and in heat stress (7.91 vs. 7.11), and in an enriched environment, birds under heat stress crowded more than birds raised in nonenriched environments (4.66 vs. 7.11). Environmental enrichment increased broilers’ unrest (movement) index from fast and slow-growing strains under thermoneutrality conditions, confirming our findings [32]. Exposure to high temperatures above the thermal comfort zone is challenging for birds housed in complex environments. Broilers raised in enriched environments from 1-day-old exposed to stress conditions at 22 days of age (heat, noise, and containment in a crate) showed heat stress as the worst adverse condition [54].Environmental enrichment benefits, especially perches and litter boxes, have been extensively studied [2,4,7,9,10,11,50]. In the present study, we noted behavioral changes in the group compatible with these benefits, suggesting that the proposed computer vision based on cluster and unrest indexes can be safely used for these assessments.Figure 5 shows the crowding behavior of broilers housed in enriched and nonenriched environments at the different temperatures tested. Results of the present study indicated that complex environments favored the crowding of the group of broilers under heat stress. Visually reviewing the videos, we observed that birds under heat stress conditions clustered around enrichment objects, indicating that environmental enrichment can minimize the negative effect of heat stress on birds. Figure 7 illustrates the frames for comfort (a and c) and heat stress (b and d) conditions in enriched (a and b) and nonenriched (c and d) environments.When analyzing the positioning of the animals in the video images, it was observed that broilers have different distribution patterns depending on the complexity of the environment (enriched versus nonenriched). Note that the birds are better distributed throughout the compartment for the environment without environmental enrichment in the comfort treatment (Figure 7a). However, in the heat-stress condition (Figure 7b), birds crowded near the drinker to benefit from the microclimate close to the water, which resulted in higher crowding rates [23,55]. The behaviors of remaining seated birds—increased water consumption, spreading wings, increased respiratory rate, and panting—are favored to dissipate excess heat [55,56,57,58].In our study and previous work, bird distribution inside the pen appears to be highly related to the location of food and water [59]. Environmental enrichment also alters the distribution pattern of birds, with a higher prevalence of agglomeration close to enrichment objects, both in small-group experiments [32,60,61] and at a commercial scale [2,8,62]. Cornetto and Estevez [60] observed that the birds were forced to occupy the central region earlier in groups of larger sizes than for smaller group sizes. Slow-growing broilers used environmental-enrichment objects more frequently when compared to fast-growing broilers [32]. These findings reinforce the differences among housed flocks and the importance of recalibrating the CV system to each housing situation (group size, strains, age, and housing conditions).The CV could simultaneously apply the Cluster and Unrest indexes to monitor the movement of the group of broilers under different environmental conditions, indicating the possible differences in the environmental conditions. The authors suggest that more research should be conducted to evaluate the potential positive impact of environmental enrichment in poultry production. The complexity of the environment is a factor to be considered in creating alert systems for detecting heat stress in broiler production.4. ConclusionsThe cluster and unrest indexes calculated from videos analyzed by computer-vision techniques allowed us to simultaneously evaluate the movement of broilers raised in different environments and detect variations that allowed us to estimate the level of wellbeing. We recommend that the indexes be used to evaluate the movement and agglomeration of broiler flocks in environments with different enrichment levels to evaluate the improvement of wellbeing. In large groups, birds’ movement and grouping patterns may differ; therefore, the CV system and indices will need to be recalibrated. The use of CV to assist with monitoring can assist caregivers during the rearing of broiler chickens.

animals : an open access journal from mdpi