Datasets:

datajuicer

/

the-pile-pubmed-central-refined-by-data-juicer

like 1

Follow

Data-Juicer 11

Background ========== Congenital fibrosis of the extraocular muscles (CFEOM) and Duane syndrome (DS) are complex strabismus disorders that present with congenital restrictive ophthalmoplegia with or without ptosis. These disorders were traditionally believed to reflect primary structural extraocular muscle (EOM) anomalies and have been referred to as \'congenital fibrosis syndromes\' \[[@B1]\]. Neuropathology studies of DS \[[@B2],[@B3]\] and one form of CFEOM (CFEOM1) \[[@B4]\], and the identification of *ARIX* as the gene mutated in a second form of CFEOM (CFEOM2) \[[@B5]\], however, support our hypothesis that CFEOM results from maldevelopment of the oculomotor (nIII) and/or trochlear (nIV) nuclei and DS results from maldevelopment of the abducens (nVI) nucleus. The continued definition of these phenotypes and identification of the underlying disease genes will assist clinical diagnostics and lead to a better understanding of the unique developmental features of the oculomotor lower motor neuron unit. Although several distinct CFEOM phenotypes have been defined \[[@B6]-[@B8]\], each likely resulting from maldevelopment of a unique combination of alpha motor neurons in nIII and/or nIV, most reports of CFEOM families describe a stereotypical clinical phenotype. The affected members of these pedigrees are born with bilateral ptosis and restrictive ophthalmoplegia. The primary vertical position of each eye is downward and cannot be elevated above the midline. On forced duction testing there is resistance to passive movement of the globe. Although the primary position of both eyes is infraducted, there is variability in the secondary position of each eye (i.e. exotropic, esotropic, or neutral), and the degree of residual horizontal movement within the lower quadrants (full to completely restricted). This CFEOM phenotype was first described in the medical literature in 1840 \[[@B9]\] and was recognized to occur as a familial trait in 1879 \[[@B10]\]. Subsequently, families segregating this phenotype have been published under myriad names \[[@B1],[@B4],[@B6],[@B11]-[@B21]\]. We now refer to individuals with this phenotype as having \"classic CFEOM\" and to families in which all affected members have this phenotype as \" CFEOM1 pedigrees\". We previously mapped a CFEOM1 locus, referred to as *FEOM1,* to a ≤ 3 cM region spanning the centromere of chromosome 12, flanked by *D12S1584 (AFM136xf6)* on the p-arm and *D12S1668 (AFMb320wd9)* on the q-arm \[[@B6],[@B15]\]. In addition to families with CFEOM1, we have identified several less common familial CFEOM phenotypes (CFEOM2 and CFEOM3). These phenotypes are classified as CFEOM based on the presence of affected members with congenital restrictive ophthalmoplegia affecting extraocular muscles in the nIII/nIV distribution. By definition, however, one or more affected family members do not have the classic CFEOM phenotype. In families with CFEOM2, the eyes of affected family members are fixed in an exotropic, or outward, position. Thus far, this phenotype segregates as an autosomal recessive trait and maps to the *FEOM2* locus on chromosome 11q13 \[[@B8]\], and affected individuals carry homozygous mutations in *ARIX*\[[@B5]\]. *ARIX* encodes a homeodomain transcription factor required for nIII and nIV development in mice and zebrafish \[[@B22],[@B23]\]. In families with CFEOM3, the CFEOM phenotype is variably expressed. Some affected members have classic CFEOM. Others, however, are unilaterally affected, the primary position of the eye is orthotropic rather than infraducted, and/or the eye can be raised into the upper quadrants. Thus far, this phenotype segregates as an autosomal dominant trait and maps to either *FEOM3* on 16qter \[[@B7]\] or to *FEOM1*\[[@B24]\]. In our attempt to understand phenotype-genotype correlations between specific CFEOM phenotypes and *FEOM* loci, we noted that the CFEOM1 phenotype in all pedigrees reported to date maps to the *FEOM1* locus. To determine if CFEOM1 is indeed genetically homogeneous, we identified all unpublished CFEOM1 pedigrees in our database, analyzed them for linkage to the *FEOM* loci, and found that most but not all were consistent with linkage to *FEOM1.* The two small pedigrees not linked to *FEOM1* were consistent with linkage to *FEOM3.* In addition, to further define the spectrum of human *ARIX* mutations, we identified all CFEOM1 families consistent with linkage to *FEOM2* or sporadic individuals with classic CFEOM and determined that none harbored mutations in the *ARIX* gene. Results ======= From our database, 33 pedigrees were of sufficient size and had sufficient clinical data and 14 sporadic individuals had sufficient clinical data to qualify for the study. Of these, 20 pedigrees met CFEOM1 and 5 sporadic individuals met classic CFEOM inclusion criteria. Although not an inclusion criterion for the study, the CFEOM1 phenotype in all 20 families was inherited as an autosomal dominant trait with full penetrance. The phenotype in 9 of the 20 pedigrees was previously demonstrated to map to the *FEOM1* locus \[[@B6],[@B15],[@B25],[@B26]\] (Table [1](#T1){ref-type="table"}). Therefore, the remaining 11 families were included in this study (Figs. [1](#F1){ref-type="fig"} &[2](#F2){ref-type="fig"}, Table [1](#T1){ref-type="table"}, and see additional files 1--3 \[[Additional File 1](#S1){ref-type="supplementary-material"}, [Additional File 2](#S2){ref-type="supplementary-material"}, [Additional File 3](#S3){ref-type="supplementary-material"}\]. {ref-type="fig"} were assigned as linkage studies were performed are not equivalent when compared between families. Black bars denote the potential disease-associated region. Diagonally hatched or white bars highlight the inheritance of the non-disease-associated haplotypes. References to specific individuals within the text refer to the generation number (Roman numeral) and position within generation (Arabic numeral). In all 9 pedigrees each family\'s disease-associated haplotype is inherited by all CFEOM1 individuals and by no asymptomatic individuals.](1471-2156-3-3-1){#F1} {ref-type="fig"}. In each family the CFEOM1 phenotype is co-inherited with *FEOM3* markers and not with *FEOM1* markers.](1471-2156-3-3-2){#F2} ::: {#T1 .table-wrap} ::: {.caption} ###### Summary of the genetic analysis of CFEOM1 pedigrees ::: **Pedigree** **CFEOM Phenotype** **Inheritance** **Forced Ductions** **Cytogenetic analysis** ***FEOM1*** ***FEOM3*** ***FEOM2*** ***ARIX* mutations** **Publication** -------------- --------------------- ----------------- --------------------- -------------------------- ----------------- ----------------- ------------------- ---------------------- ------------------------ A CFEOM1 AD \+ normal **LINKED** refs. \[[@B6],[@B15]\] B CFEOM1 AD \+ normal **LINKED** refs. \[[@B6],[@B15]\] C CFEOM1 AD \+ not done **LINKED** ref. \[[@B15]\] H CFEOM1 AD \+ normal **LINKED** ref. \[[@B15]\] AA CFEOM1 AD \+ normal **LINKED** ref. \[[@B15]\] AC CFEOM1 AD \+ normal **LINKED** ref. \[[@B15]\] AD CFEOM1 AD \+ normal **LINKED** ref. \[[@B15]\] CD CFEOM1 AD \+ not done **LINKED** ref. \[[@B25]\] CB CFEOM1 AD not done not done **c/w linkage** nc/w linkage NOT LINKED ref. \[[@B26]\] BJ CFEOM1 AD \+ normal **LINKED** current CZ CFEOM1 AD \+ normal **LINKED** current AG CFEOM1 AD \+ normal **c/w linkage** NOT LINKED NOT LINKED current AJ CFEOM1 AD \+ not done **c/w linkage** nc/w linkage nc/w linkage current AH CFEOM1 AD \+ normal **c/w linkage** nc/w linkage nc/w linkage current T CFEOM1 AD \+ not done **c/w linkage** nc/w linkage nc/w linkage current CT CFEOM1 AD \+ normal **c/w linkage** nc/w linkage nc/w linkage current BC CFEOM1 AD \+ not done **c/w linkage** nc/w linkage (c/w linkage) none current E CFEOM1 AD \+ normal **c/w linkage** **c/w linkage** NOT LINKED current K CFEOM1 AD \+ normal nc/w linkage **c/w linkage** **(c/w linkage)** none current BT CFEOM1 AD \+ normal NOT LINKED **c/w linkage** nc/w linkage current AD = autosomal dominant; + = positive forced duction testing for restriction; c/w linkage = consistent with linkage; nc/w li nkage = not consistent with linkage. ::: The 11 families are ethnically diverse, not consanguineous and, to the best of our knowledge, unrelated. Eight of the families reside in the US and are of mixed European ancestry (BJ, AG, AJ, AH, CT, E, K, BT), while families CZ, T, and BC are of Italian, Irish, and Japanese ancestry, respectively. Family history of CFEOM in several previous generations was documented in two pedigrees (AG, T), and a previous family history was recounted but could not be corroborated in five others (BJ, AJ, AH, E, K). In contrast, in pedigrees CZ, BC, CT and BT, neither the parents nor more distant relatives of the eldest affected family member were reportedly affected. Examination of individuals I--1 and I--2 of pedigrees BC and CT confirmed their unaffected status. Individuals I--1 and I--2 in pedigrees CZ and BT were deceased. These data suggest that the CFEOM1 mutation rate is not negligible. Cytogenetic analyses, when performed, did not reveal abnormalities (Table [1](#T1){ref-type="table"}). Nine of the 11 CFEOM1 pedigrees contain too few participants to establish linkage to a specific locus. Haplotype analysis of these families using multiple markers that span the critical region of a given locus can, however, eliminate linkage to the locus, determine genetic heterogeneity, and guide future mutation analyses. If the phenotype results from a mutation at a given locus, haplotype analysis at that locus will be consistent with linkage. If the phenotype does not result from a mutation at a given locus, however, haplotype data from a small family may be consistent or inconsistent with linkage. Thus, haplotype data in a small family that is consistent with linkage can result either from a disease mutation at that locus or by chance. Haplotype data in a small family that is inconsistent with linkage strongly suggests that the family\'s phenotype is not linked to the locus. Linkage to *FEOM1* ------------------ Genetic analyses of the two largest families (BJ, CZ) established linkage of their phenotype to the *FEOM1* locus (Fig. [1](#F1){ref-type="fig"}, Table [1](#T1){ref-type="table"} & see [Additional File 1](#S1){ref-type="supplementary-material"}). Maximum lod scores of 3.01 were obtained at a theta value of zero for the fully informative markers *D12S59* and *D12S1048* in family BJ, and the fully informative markers *D12S1648, D12S345,* and *D12S59* in family CZ. Linkage to *FEOM1* was ruled out in family BT (see [Additional file 1](#S1){ref-type="supplementary-material"}). A lod score of -2 was obtained at a theta value of 0.04 for the fully informative markers *D12S1621, D12S59* and *D12S1668.* The *FEOM1* critical region is ≤ 3 cM, and a theta of 0.04 corresponds to a genetic distance of approximately 4 cM, thus eliminating linkage of this family\'s CFEOM disease gene to the entire *FEOM1* critical region. Exclusion of linkage to the *FEOM1* locus is further supported by haplotype analysis of this family (Fig. [2c](#F2){ref-type="fig"}). The affected sister and brother inherit different *FEOM1* haplotypes from their affected mother, and the brother\'s affected daughter inherits her unaffected paternal grandfather\'s *FEOM1* haplotype, thus proving non-association between the *FEOM1* haplotype and the disease phenotype. The eight remaining CFEOM pedigrees were too small to produce statistically significant lod scores; however, seven of the eight families displayed haplotype and linkage data consistent with linkage to the *FEOM1* locus (Fig. [1](#F1){ref-type="fig"}, see [Additional File 1](#S1){ref-type="supplementary-material"}). Genetic and haplotype analysis of all nine families consistent with linkage to the *FEOM1* locus did not reveal any recombination events within the previously defined *FEOM1* critical region. The smallest family, K, revealed haplotype and linkage data that was inconsistent with linkage to the *FEOM1* locus (Fig. [2a](#F2){ref-type="fig"}, see [Additional File 1](#S1){ref-type="supplementary-material"}). A lod score of -2 was obtained at a theta value of 0.002 for the only two informative markers, *D12S59* and *D12S1090.* These theta values eliminate linkage to only 0.8 cM of the \< 3 cM *FEOM1* critical region and, therefore, this locus cannot be formally ruled out. Nevertheless, the minimum number of recombination events in this family occurs only if the affected son and daughter inherit different *FEOM1* haplotypes from their affected mother, thus strongly suggesting that the disease gene in this family does not map to the chromosome 12 locus. Genetic heterogeneity was tested taking into account the eleven new families. Admixture analysis of the two-point data with the HOMOG program showed evidence for linkage to *FEOM1* with heterogeneity for both markers tested (*D12S345* and *D12S1048*). Chi-squares of 34.601 and 40.377 were obtained for *D12S345* and *D12S1048* respectively which resulted in significant likelihood ratios of 3.26 × 10^7^ and 5.86 × 10^8^. Alpha (the proportion of linked families) was 0.90. Linkage to *FEOM3* ------------------ All families except BJ and CZ were analyzed for linkage to the 5.6 cM *FEOM3* locus flanked by *D16S486* and 16qter. Of the seven small families consistent with linkage to *FEOM1,* only the largest (AG) can be definitively excluded from linkage to *FEOM3* (see [Additional File 2](#S2){ref-type="supplementary-material"}). Five of the remaining families (AJ, AH, T, BC, CT) showed haplotype data inconsistent with linkage to *FEOM3,* but the theta values obtained at lod scores of -2 were insufficient to rule out the entire *FEOM3* critical region. The two families whose phenotype did not map to *FEOM1* (K, BT) had haplotype and linkage data consistent with linkage to *FEOM3* (Fig. [2b](#F2){ref-type="fig"} &[2d](#F2){ref-type="fig"}, see [Additional File 2](#S2){ref-type="supplementary-material"}). In addition, one of the small families consistent with linkage to the *FEOM1* locus (E) had haplotype and linkage data that was also consistent with linkage to the *FEOM3* locus (see [Additional File 2](#S2){ref-type="supplementary-material"}). Linkage to *FEOM2* ------------------ All families whose phenotype was not linked to *FEOM1* or *FEOM3* were tested for linkage to *FEOM2.* Assuming autosomal dominant inheritance with complete penetrance, two families (AG, E) are not linked and five families are inconsistent with linkage (AJ, AH, T, CT, BT) to the *FEOM2* locus (see [Additional File 3](#S3){ref-type="supplementary-material"}). Family BC is consistent with linkage to both the *FEOM2* and *FEOM1* loci (maximum lod score 0.3 at both loci), and family K is consistent with linkage to both the *FEOM2* and *FEOM3* loci (maximum lod score 0.3 at both loci). *ARIX* mutation analysis ------------------------ Genomic DNA samples from affected member of pedigree BC and K and 5 sporadic individuals with classic CFEOM were used as templates to sequence the three *ARIX* exons and flanking introns. No mutations were identified. Discussion ========== We have established clinical criteria for classic CFEOM and CFEOM1, and have identified 5 sporadic individuals with classic CFEOM and 20 pedigrees with CFEOM1. Of these 20 pedigrees, 18 are linked, or consistent with linkage, to the *FEOM1* locus. Two small CFEOM1 pedigrees are not consistent with linkage, however, establishing that CFEOM1 is genetically heterogeneous. Eleven of the 20 pedigrees are large enough to establish linkage to a specific locus; we previously reported that the CFEOM1 phenotype in 9 of these pedigrees maps to the *FEOM1* locus \[[@B6],[@B15],[@B25],[@B26]\] and we now demonstrate that the remaining two also map to *FEOM1.* Our analysis of the remaining 9 CFEOM1 pedigrees demonstrates that 6 most likely result from mutations in the *FEOM1* gene. Five of these 6 are consistent with linkage to *FEOM1* and are either not linked or not consistent with linkage to *FEOM2* and *FEOM3.* One is consistent with linkage to *FEOM1* and not *FEOM3* and, although consistent with linkage to *FEOM2, ARIX* mutations were not identified. Therefore, although not proved, the CFEOM1 phenotype in these 6 families seems likely to result from mutations in the *FEOM1* gene. The phenotype of a seventh family, E, is consistent with linkage to both *FEOM1* and *FEOM3* and will be screened for mutations at both these loci. In contrast to the 18 pedigrees whose CFEOM1 phenotype is consistent with linkage to *FEOM1,* pedigree BT is not linked to *FEOM1* and pedigree K is inconsistent with linkage to this locus. It is notable that haplotype analysis of both these small CFEOM1 families demonstrates co-inheritance with the *FEOM3* locus. In the reported family whose autosomal dominant CFEOM3 phenotype maps to the *FEOM3* locus, 9 of the 17 affected members had classic CFEOM. Our current data now suggests that, at least in small pedigrees, CFEOM1 can also map to the *FEOM3* locus. It will require the identification of additional large CFEOM1 families to determine if they too can map to this locus. *ARIX,* which encodes a transcription factor critical to nIII and nIV development in mice and zebrafish \[[@B22],[@B23]\], was recently identified as the *FEOM2* gene mutated in affected members of CFEOM2 families \[[@B5]\]. It was unknown, however, if classic CFEOM may also result from mutations in this gene. We now find that we are unable to identify *ARIX* mutations underlying classic CFEOM in either sporadic cases or in individuals from CFEOM1 families. This finding is consistent with our prediction that, compared to *ARIX,* the genes mutated in CFEOM1 may have a more restricted function in the development of nIII and that their expression may actually be regulated by *ARIX.* This prediction is based on the CFEOM1 phenotype and on the neuropathological study of an affected member of a CFEOM1 pedigree whose disease gene maps to the *FEOM1* locus. This study revealed absence of the superior division of the oculomotor nerve and the corresponding central caudal and medial nIII subnuclei, and marked abnormalities of the levator palpebrae superioris and superior rectus muscles \[[@B4]\]. These findings suggest that while *ARIX* is necessary for both nIII and nIV development, the CFEOM1 genes may be necessary for the development of only these two specific nIII subnuclei. Conclusions =========== The genetic analysis of the 11 CFEOM1 pedigrees in this report demonstrates that this disorder is genetically heterogeneous. While the phenotype of all large CFEOM1 pedigrees analyzed thus far map to the *FEOM1* locus, smaller CFEOM1 pedigrees may harbor mutations in the *FEOM3* gene. The CFEOM1 phenotype does not, however, appear to result from mutations in *ARIX.* The CFEOM1 families identified in this study contribute critical alleles toward the identification of the mutated *FEOM1* and *FEOM3* genes. Once identified, we anticipate that the study of the function of these genes will contribute to our understanding of midbrain motor neuron development. Materials and Methods ===================== Subjects -------- We established inclusion criteria for this study as follows. First, we established criteria for a \"classic CFEOM individual\" as an individual with congenital nonprogressive bilateral ophthalmoplegia and ptosis, an infraducted primary position of each eye with the inability to raise either eye above the midline, and forced duction testing positive for restriction, if testing was performed. We then established criteria for a \"CFEOM1 pedigree\" as a family in which every affected member met the criteria for \"classic CFEOM\". Second, we reviewed all participants enrolled in our ongoing CFEOM study and determined which individuals and which pedigrees met these two criteria. For a pedigree to be considered, we required clinical examination records and/or photographs/videos of the primary positions of gaze for all affected study participants. Third, from these CFEOM1 pedigrees we identified those with a family structure sufficient for linkage analysis. Pedigrees were required to have affected study participants in three or more generations, or affected study participants in two generations with at least two participating offspring of affected individuals, or two or more affected study participants within one generation. In this way we did not assume a mode of inheritance. Lastly, sporadic individuals with classic CFEOM were screened for *ARIX* mutations. The study was approved by the Children\'s Hospital institutional review board, and all study participants signed informed consent forms. Our methods adhered to the Declaration of Helsinki for research involving human subjects. Molecular studies ----------------- Blood samples were obtained from all participating family members, and lymphocyte DNA was extracted using the Puregene kit (Gentra, Research Triangle Park, NC). Chromosome analyses of GTG banded metaphase cells at a 400 band level minimum resolution were performed on one or more affected family members of each family whenever possible to rule out cytogenetic abnormalities. All families were analyzed for linkage to the *FEOM1* locus. Families not linked to the *FEOM1* locus were also analyzed for linkage to the *FEOM2* and *FEOM3* loci. Linkage studies were conducted using three or more locus specific polymorphic DNA micro satellite markers for each family. The *FEOM1* markers included *D12S1648, D12S61, D12S1584, D12S1621, D12S345, D12S59, D12S2080, D12S1048, D12S1668,* and *D12S1090*\[[@B6],[@B15]\]. The *FEOM2* markers included *D11S1337, D11S4162, D11S4196, D11S1314,* and *D11S1369*\[[@B8]\]. The *FEOM3* markers included *D16S539, D16S3077, D16S498, D16S486, D16S476, D16S3063, D16S689, D16S2621, D16S303,* and *D16S3407*\[[@B7]\]. The primer sequences for these polymorphisms are available from the Genome Database <>. Unlabeled primers were purchased from Genosys Biotechnologies, Inc. <>. Fluorescently labeled primers were purchased from Research Genetics, Inc. <>. Radioactive products were made by 30 cycles of PCR amplification of 10-μl reaction volumes containing 10--30 ng of genomic DNA, 40 ng of each primer, 200 μM each of dATP, dTTP, dGTP, and dCTP, 1 μCi α-^32^P-dCTP (3,000 Ci mmol^-1^) and 0.5 U Taq polymerase (Perkin Elmer). The radioactively labeled PCR products were separated on 6% denaturing polyacrylamide sequencing gels, and the alleles were visualized by autoradiography \[[@B6]\]. For the fluorescently labeled products, α-^32^P-dCTP was omitted, fluorescent primers were used, and the products were analyzed in an ABI PrismTM 377 DNA Sequencer (Perkin Elmer) following the manufacturer\'s specifications. Lod score calculations ---------------------- An individual was scored as affected based on clinical examination records and/or photographs of primary position of gaze. Lod scores were calculated using the Fastlink version 3.0 package of programs \[[@B27]\], assuming autosomal dominant inheritance with complete penetrance, and a disease incidence of 1 in 1,000,000 births, as described previously \[[@B7],[@B8]\]. Because of the absence of specific allele frequencies for each of the ethnic groups represented in the study, we assumed ten marker alleles of equal frequency. To assess nonallelic heterogeneity linkage data from two *FEOM1* markers spanning the *FEOM1* critical region (*D12S345* and *D12S1048*) were analyzed using a HOMOG version 3.35 program \[[@B28]\]. ARIX mutation detection ----------------------- *ARIX* was PCR amplified from genomic DNA using five primer sets and sequenced on an automated ABI 377 DNA seqeuncer (PE-Applied Biosystems) as described previously \[[@B5]\]. Supplementary Material ====================== ::: {.caption} ###### Additional file 1 Table 2: Two-point linkage data between CFEOM1 and the *FEOM1* locus ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 2 Table 3: Two-point linkage data between CFEOM1 and the *FEOM3* locus ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 3 Table 4: Two-point linkage data between CFEOM1 and the *FEOM2* locus ::: ::: {.caption} ###### Click here for file ::: Acknowledgements ================ This work was supported by National Eye Institute EY12498 and EY13583 and by the Children\'s Hospital Mental Retardation Research Center (P30 HD18655).

Background ========== Identification of species with molecular probes is likely to revolutionize taxonomy, at least for taxa with morphological characters that are difficult to determine otherwise. Among these are the single cell eucaryotes, such as Ciliates and Flagellates, but also many other kinds of small organisms, such as Nematodes, Rotifers, Crustaceans, mites, Annelids or Insect larvae. These organisms constitute the meiofauna in water and soil, which is of profound importance in the ecological network. Efficient ways for monitoring species identity and abundance in the meiofauna should significantly help to understand ecological processes. Molecular taxonomy with sequence specific oligo-nucleotide probes has been pioneered for bacteria \[[@B1],[@B2]\]. Probes that are specific to particular species or groups of related species can be used in fluorescent in situ hybridization assays to detect the species in complex mixtures or as symbionts of other organisms \[[@B3],[@B4]\]. Alternatively, the microarray technology is increasingly used for this purpose, allowing potentially the parallel screening of many different species. Most of the species-specific sequences that are used so far for this purpose are derived from ribosomal RNA sequences. However, any other sequence is also potentially suitable, as for example mitochondrial D-loop sequences in eucaryotes. The species-specific probes are usually derived from an alignment of the respective sequences, where conserved and non-conserved regions are directly visible. A program has been developed for ribosomal sequences that helps to build the relevant database, and supports the selection of suitable specific sequences (ARB \[[@B5]\]). In this, a correct alignment is crucial for finding the optimal probes, but alignments are problematical in poorly conserved regions. These, on the other hand, have the highest potential to yield specific probes. Moreover, the current implementation of probe finding calculates only the number of mismatching position to discriminate between the probes, but does not take into account the position of the mismatches within the stretches, which could influence the hybridization behavior. We have therefore devised here a new algorithm that allows working with datasets that need not to be carefully aligned and that takes the position of mismatches along the recognition sequence into account. The algorithm ------------- The algorithm includes three parts. The first one aims to provide a function that calculates the relative stability of matching oligos in dependence of the number and position of mismatches. The second one provides a strategy for probe finding that scans all possible sequence combinations, but works time efficient. The third part deals with matches caused by single nucleotide outloops of a given sequence. Stability function ------------------ Extensive studies exist for assessing the thermodynamic consequences of internal mismatches in short oligo-nucleotides (see fro example \[[@B6],[@B7]\]). These show that there are no simple rules and that the exact influence on the stability of a hybrid depends on the nature of the mismatch, as well as its flanking nucleotides. For example, mismatches including a G (i.e. G-G, G-T and G-A) tend to be less destabilizing than the other types of mismatches \[[@B7]\], although this can not directly be predicted from steric considerations. Comparable systematic studies on the relative influence of the position of the mismatch within the oligonucleotide do not exist yet, although it is clear that the influence is lower at the ends than in more central positions \[[@B7],[@B9]\]. Preliminary evidence with an oligo-dT stretch harboring A mismatches along the sequence suggests that the position dependence could be a continuous function \[[@B8]\]. We have therefore decided to use an *ad hoc* approach for the stability calculation that is mainly designed to discriminate against sequences with more central mismatch positions. We model the relative stability of mismatched oligos as follows. The position of the mismatch can be considered to be a \"weak point\". The location of the \"weak point\" is expressed as a probability function that takes into account the differential contribution of central versus terminal positions. The probability that the \"weak point\" is at position x is defined by p~1~. Under the experimental conditions of melting, the presence of the \"weak point\" is true, meaning that \[sum(p1) for all x\] = 1. We assume a Gauss distribution as the respective probability function, with the maximum in the middle of the duplex and the integral value along the duplex length set to 1 (Equation 1). Equation 1. \"Weak point\" location probability. L -- duplex length, σ -- distribution parameter, x -- duplex position. Note that the function in Equation 1 refers to discrete positions within the sequence, while the Gauss distribution is continuous and the integration from -∞ to +∞ is set to yield 1. The parameter σ is therefore chosen such that the discrete sum approaches 1 at any intended precision. In the program discussed below the accuracy of the sum value is 0.999. Although the preliminary experimental evidence \[[@B8]\] suggests that the destabilization function can be approximated with the Gauss distribution, the program implementation allows also to use a flat distribution, i.e. where a position-independent effect on the melting is assumed as an alternative, to compare the outputs of the two different assumptions. For assessing the relative amount of destabilization caused by a certain mismatch, we assume that the mismatch disturbs the surrounding base pairs from (y-n) to (y+n) positions, n can be called a border parameter that will need to be experimentally verified in the future. Because n can currently only be guessed, it is set as a program variable with a default value of 5. n might also depend on the nature of the mismatch, i.e. some types of mismatches might influence the surrounding bases less than the others. We therefore implemented further program variables that allow to define a different n depending on the nature of the mismatch (i.e. it is possible to set a particular n value for each possible type of mismatch). The overall relative stability of a given duplex is then expressed as a probability function. It is expressed as the sum of products of the individual position probabilities p~1~ (determined by the stability function) and p~2~ (determined by the border parameter). The value of p~2~ it the probability of \"melting\", conditioned that the \"weak point\" is disturbed. (Equation 2). Equation 2. L -- the length of the duplex, p~1~ -- the \"weak point\" location probability, p2 -- the \"melting\" probability due to the disturbance of the \"weak point\". p~2~ is a conditional probability of \"melting\" with p~2~ = 1 if the \"weak point\" is disturbed (in the region y ± n) and p~2~ = 0 at non-affected positions. This allows transforming Equation 2 into Equation 3. Equation 3. y -- the mismatch position, n -- the border parameter p~1~ can then be substituted by the function in Equation 1, to yield Equation 4. Equation 4: x -- the duplex position, y -- the mismatch position, n -- the border parameter, σ-distribution parameter In the case of several mismatches, the summing is done along all the respective mismatch regions. If the mismatches occur next to each other, their disturbed regions simply overlap and the summing is performed across the respective region. Probe finding ------------- The probe finding strategy is devised in a way (i) to avoid the need for exact alignments, (ii) to check probe specificity along the whole available sequence and (iii) to optimize performance. The workflow is depicted in Figure [1](#F1){ref-type="fig"}. It starts with a database in which each organism is represented by a single continuous sequence, such as a defined region of the 18S or 28S ribosomal genes. From this it takes first the sequences of the In-group organism(s) for which specific probes should be found and cuts these into short pieces of the specified oligo-nucleotide length (set as a program variable), following an approach proposed by Bavykin et al \[[@B11]\]. This is accomplished by a sliding window scheme with 1-nucleotide shifts across the whole length of the sequence(s). Two separate lists are created in this way. The first list is simply a straight list of all possible fragments from all In-group organisms. The second one consists of an array of lists for each of the In-group organisms (the two lists are identical if only one In-group organism is chosen). All duplicate oligos from the first list are then removed and each of the remaining oligos is checked whether it matches with each of the In-group organisms in the second list. A match is positive, when the relative melting probability is within the range of 0--25%, employing the function of Equation 4. Thus, this first calculation simply ensures that all candidate probes match with all In-group organisms. This calculation would be largely dispensable, if only a single In-group organism is chosen. {#F1} The next step is to subtract all oligos that match in any of the Out-group organisms. To avoid the comparison of all candidate oligos against all Out-group sequences, we identify first a group of sequences that is closely related to the In-group. For this one requires a rough alignment of all sequences, to calculate percentage similarity between them. Note that this serves only to identify a subgroup of sequences for speeding up the calculations, i.e. mistakes in the alignment are of no concern. The similarity calculator in the program extracts this related group of sequences by a simple percentage identity calculation across the given alignment. All sequences that are at least 90% similar to the In-group are used as Related-group. This percentage can be set as a program variable and should be set such that the Related-group does not become more than 5--10% of all sequences. The sequences of the Related-group are again converted into a fragment list as above, duplicates are removed and all candidate oligos are matched with this list. Now only those oligos are retained, which have a melting probability of at least 75% (the exact percentage values are program variables). The majority of oligos is removed in this step. The remaining candidate oligos are then matched against the remaining sequences in the Out-group with the same cut-off criterion. This stepwise selection scheme allows to significantly speed up the calculations even for very large datasets, but still ensures that all oligo-nucleotides of the desired length were directly or indirectly matched against all possible other oligos in the database. Single nucleotide loops ----------------------- Structure analysis with experimental oligo-nucleotides has shown that in a pair of hybridized oligos, one nucleotide can loop out, without interfering much with the stability of the hybridized pair \[[@B12]\]. This implies that one base of one strand of a duplex can loop out from the duplex and the rest of the strand can shift one position. This is depicted in Figure [2](#F2){ref-type="fig"}. A standard linear scanning algorithm would recognize the situation at the left as one with 11 mismatches, i.e. would suggest it as a specific probe. However, if the single nucleotide loop is taken into account, the match would be perfect and the probe would have to be considered as unspecific. Our scanning algorithm takes this problem into account by re-checking all candidate probes after the completion of the filtering steps. It does this by sequentially removing one nucleotide from the candidate probe and shifting the remainder by one position. The melting probability of the new oligo is then calculated and checked. The removed nucleotide is then reinserted and the cycle is repeated for the next position. The same procedure is done for the target sequence, so that outloops are considered to be possible on both strands of the duplex. Note that outloops of two nucleotides are considered to destabilize the helix too much to warrant a separate analogous calculation. {#F2} Parallel computation -------------------- A parallel program version allows probe finding to be done in parallel on several processors. Essentially the same algorithm is used in the parallel version of the program, whereby the parallelism is introduced in the matching steps. Each process takes its own part of the database and performs the matching as well as the stability calculations. The results are then gathered by the root process and superimposed. Program implementation ---------------------- The algorithm is implemented in a program called PROBE. The program consists of three modules that can be used independently. The first module finds the probes based on the given task (specificity group, length of probes, source database). The second one is the analytic module, which can be used if it is impossible to design a probe for a given organism group. This module depicts the situation with the given In-group and enables to find the closest group for which the task can be accomplished. The use of the analytic mode comes into play when PROBE fails to identify a set of probes for the given organism group. Such a failure can have two reasons -- either there is no probe, which identifies all organisms in the specificity group, or there is another organism outside the specificity group, which is also identified by all candidate probes suitable for the specificity group. For the first case, the specificity group must be broken down into several subgroups and the probes must be identified for these subgroups separately. For the second case, the organism that is very similar to the specificity group should be added to the specificity group and this may then have to be broken down into smaller subgroups. The analytic module creates a table with the organisms of the specificity group as well as the most related organisms. This table depicts then the matching or non-matching patterns for each of the possible probes, allowing a simple visual inspection of the best specificity groups. The output can be viewed and modified with spreadsheet programs such as Excel. The third module provides a report for the identified probe, including the mismatches in the duplexes within the specificity group, the best match out of the group and some other information. The program is written in standard C++ in a platform independent manner. Therefore, the program can be easily compiled for Linux and Windows without any modifications. The program binary files for Linux and Windows are available from the <> as freeware accompanied with all its source files, and a manual that describes further details. Results ======= As an example of the performance of the program we have used the full SSU database (RDP, release 8.1) \[[@B13]\] containing approximately 16.000 sequences to find a specific oligo-nucleotide probe with a length on 20 nt for *Thermotoga maritima.* The search was done on a Pentium III (800 MHz, 512 MB RAM) PC and took about 1.5 hours without outlooping and 16 hours with outlooping, indicating that the most time intensive step is the outlooping subroutine. The parallel version running on a cluster with 24 nodes (with the slowest node being a Pentium II -- 400 MHz with 256 MB RAM) took 2 hours for the same full task. Figure [3](#F3){ref-type="fig"} depicts the output from the check module, which allows comparing the oligos and their specificity that were found in this particular comparison. It shows that ARB suggests two oligos that are rejected by PROBE either because of mismatches occurring only at the ends, or under the outloop routine. Both programs find one oligo with acceptable high specificity. {#F3} Discussion ========== The algorithm presented here does not take into account the effect of relative GC content and stacking interactions of neighboring bases on the melting temperature of the oligo-nucleotides. Accordingly, the oligo-nucleotides suggested by the program can differ significantly in melting temperature. However, as this can easily be adjusted after the selection is made, we have not included a subroutine that takes GC content into account during the primary search, because this would slow down the calculations. Furthermore, we expect that GC content differences may be of less importance for the applications envisioned here, because they can be largely compensated by the choice of experimental conditions, such as buffers that compensate stability differences \[[@B13]\]. A more general problem is our way of calculating the relative stability factor. This does currently not take the nucleotide composition into account either. The reason is that there are too few experimental data as yet, that would allow to unequivocally include this in the calculations. The current experimental data sets focus on the types of mismatches in particular contexts, but not systematically on position specific effects \[[@B7],[@B15]\]. Moreover, they deal with relatively short model oligos only (up to 12 nt). However, the probes used for species identification are longer and the different effects can currently not be accurately assessed from experimental data for such longer probes. In our equation, it is mainly the border parameter n that would be affected by base composition and nearest neighbor interactions and we have therefore left this as a variable that can be set according to experimental results. In principle, it seems possible that n differs for different sequence compositions, i.e. GC-rich stretches have a smaller n than AT-rich ones. Thus, if one chooses a low n, one would risk that GC-rich oligos are suggested as specific probes that still show cross hybridization. However, it seems that these can easily be eliminated after the selection is made. Still, if experimental data indicate that this is a major problem, the program could easily accommodate such new insights. Finally, the stability function proposed in Equation 1 could possibly also have other shapes than Gaussian. Again this is a factor that needs further experiments. If it turns out that other functions are more appropriate, one can include this as additional options into the program. At the present we offer the extreme, namely a flat function, as an alternative option. Conclusion ========== We have designed a versatile algorithm for finding optimal species- and group-specific probes for molecular taxonomy that is sufficiently open to implement further experimental insights into the nature of the stability of mismatched oligo-nucleotides. Acknowledgements ================ We are grateful to Dr. Lysov from the Engelgardt Institute of Molecular Biology, Russian Academy of Sciences for the supporting A.P. in the initial phase of the project. We thank Prof. Speckenmeyer at the Institute of Informatics, University of Cologne for providing access to their LINUX cluster and Jens Rühmkorf for his help with installing the parallel version. This project was supported by a grant from the Ministerium für Schule Wissenschaft und Forschung des Landes Nordrhein-Westfalen.

Background ========== Nerve growth factor (NGF), a prototypical neurotrophic factor and a member of the neurotrophin family, promotes a wide range of responses in its target cells. These range from neuronal differentiation, maintenance of survival, and regulation of metabolic activities \[for review see \[[@B1]-[@B3]\]. Many of these actions include and require transcriptional regulation \[[@B4],[@B5]\]. However, the greater part of the changes in gene expression that underlie the NGF response remain to be elucidated. The PC12 line of rat pheochromocytoma cells \[[@B6]\] has proved to be a particularly favorable system for detecting NGF-responsive changes in gene expression \[[@B7]-[@B11]\]. These cells resemble sympathico-pheochromo-blasts and upon exposure to NGF cease proliferation and acquire, in a transcription dependent mechanism, many of the properties of post-mitotic sympathetic neurons including neurite outgrowth and electrical excitability. The robust nature of the response of PC12 cells to NGF coupled with their capacity to be examined both before and at various times after exposure to the factor has greatly facilitated study of the NGF mechanism of action, including gene regulation \[[@B12]\]. To obtain a comprehensive and quantitative over-view of NGF-promoted gene regulation, we \[[@B11]\] have used SAGE (Serial Analysis of Gene Expression). In this technique \[[@B13]-[@B15]\] cellular transcripts are converted to SAGE \"tags\" which are sequenced, quantified and, in many cases, matched with known genes. By comparing SAGE profiles for cells in different states (as for example before and after exposure to NGF), it is thus possible to obtain a comprehensive view of gene expression and regulation. Moreover, if a sufficient number of SAGE tags are analyzed, changes in expression levels of individual transcripts can be associated with a high level of statistical significance \[[@B13]\]. Initial analysis of approximately 157,000 SAGE tags from PC12 cells cultured without or with NGF for 9 days revealed nearly 800 transcripts (of a total of at least 21,000) that are regulated by ± 6-fold or greater in response to NGF \[[@B11]\]. Of these, approximately 150 were assignable to named genes of known functions that regulate cellular behaviors ranging from actin and microtubule cytoskeleton assembly/disassembly, gene transcription, RNA processing, neurotransmission, and energetics. A variety of criteria supported the reliability of the quantitative findings revealed by our SAGE analysis \[[@B11]\]. In the present study, we have extended our SAGE profiling of naïve and long-term NGF-treated PC12 cells to over 163,000 tags representing over 22,000 unique transcripts. Analysis of these reveals the presence of transcripts encoding 74 different ribosomal proteins (RPs). Surprisingly, we find that long-term NGF exposure leads to statistically significant changes in relative abundance of at least half of these transcripts by factors of up to nearly 5-fold. In addition, a time course for one of the RP transcripts (encoding RP L9) reveals that its relative abundance begins to change within 1 hr and is maximally regulated by 8 hr of NGF exposure. Results ======= SAGE libraries -------------- SAGE libraries were generated and analyzed as previously described \[[@B11]\] from matched sets of PC12 cells before and after 9 days of exposure to NGF. The present analysis is based on 76,280 15-mer tags from NGF-untreated PC12 cells and 87,004 tags from NGF-treated cells (after exclusion of duplicate ditags, mitochondrial transcripts and repetitive elements). Consideration of tags observed twice or more between the two libraries indicated the presence of approximately 22,000 unique transcripts. Of these, approximately 10% were regulated by more than 3-fold in response to NGF and approximately 4% by 6-fold or more. Detection and quantification of transcripts encoding ribosomal proteins ----------------------------------------------------------------------- Transcripts represented by SAGE tags were identified by direct matches of tags with the appropriate sequences of known rat genes (present in GenBank) or through matches of tags with appropriate sequences of rat ESTs (present in NCBI Unigene) that were in turn found to overlap with known rat genes. Positive identification required that the transcript or EST have a poly adenylation signal and poly A tail, and that the tag followed the most 3\' CATG of the transcript. In this way 74 tags were unambiguously assignable to known rat transcripts encoding proteins described as ribosomal components. Available sequences encoding rat ribosomal proteins L15 and S8 do not contain a CATG sequence and hence SAGE tags for these could not be identified. In addition, rat sequence data for transcripts encoding RPs L2, L25, L33 and S1 are not presently available in GenBank or Unigene. Table [1](#T1){ref-type="table"} lists the tags in our libraries corresponding to transcripts for ribosomal proteins along with their relative abundances. Considering that the eukaryotic ribosome contains approximately 82 proteins, our analysis includes transcripts encoding a major proportion of the known ribosomal proteins. ::: {#T1 .table-wrap} ::: {.caption} ###### The effect of NGF (9 days treatment) on expression levels of transcripts for various ribosomal proteins in PC12 cell cultures ::: TAG RP ACC\# \#TAGS- \#TAGS+ FOLD CHANGE +NGF *P* value ------------- ---------- ----------- --------- --------- ------------------ ----------- GGACCGCTCAA L3 X62166 44 27 ↓ 1.6 0.06 TTGAAGCTGAA L4/L1 X82180 49 39 ↓ 1.3 0.2 CTGCTATCCGA L5 X06148 40 28 ↓ 1.4 0.15 TACCCTCACAA L6 X87107 16 24 ↑ l.5 0.05 AGATCTATACA L7 Ml 7422 8 7 ↓ 1.1 0.5 CACCACTGTTG L7A X15013 63 47 ↓ 1.3 0.2 AATCCTGTGGA L8 P25120 119 58 ↓ **2.1** 0.0002 ATCAAGGGTGT L9 X51706 7 19 ↑ **2.7** 0.01 TTCAATAATAA L10 X87106 24 17 ↓ 1.4 0.2 GGCAAGCCCCA L10A X93352 26 30 ↑ 1.2 0.2 CGCTGGTTCCA L11 X62146 18 23 ↑ .3 0.1 ACATCATAGAT L12 R7RT12 59 15 ↓ **3.9** \< 0.0001 GCCCGAGCCAA L13 X78327 48 57 ↑ .2 0.1 AGGTCGGGTGG L13A X68282 87 118 ↑ **1.3** 0.002 AGGAGGCTACA L14 X94242 31 51 ↑ **1.6** 0.003 GCACGGGAATA L17 X58389 8 26 ↑ **3.2** 0.0007 GGTGTTGACAT L18 M20156 47 12 ↓ **3.9** \< 0.0001 AAGGTGGAAGA L18A JC4231 56 33 ↓ 1.7 0.03 GATCAGTCATT L19 X82202 65 63 ↓ 1.0 0.3 GCCTAATGTAT L21 M27905 43 32 1.3 0.2 TTTTGTATTAA L22 X78444 5 4 1.3 0.7 GTGATGGCCAC L23 X58200 72 63 ↓ 1.1 0.4 AAGGTCGAGCT L24 X78443 95 28 ↓ **3.4** \< 0.0001 CCCAGTTTTCA L26 X14671 31 19 ↓ 1.6 0.1 CCCACAAGGTA L27 X07424 16 30 ↑ 1.9 0.007 ATCCGAAAAAA L28 X52619 42 10 ↓ **4.2** 0.0001 GCCAAGGGTCG L29 X68283 97 40 ↓ **2.4** \< 0.0001 CCAGAACAGAC L30 X52619 36 9 ↓ **4.0** 0.0002 AAGGAGATGGG L31 X04809 64 111 ↑ **1.7** \< 0.0001 CTGCCTAGCGG L32 X06483 26 20 ↓ 1.3 0.2 TGCGCCAAGTG L34 X14401 27 30 ↑ 1.1 0.2 AAGAGAAGCTG L35 X51705 115 70 ↓ **1.6** 0.009 GTTCGTGCCAA L35A X03475 33 16 ↓ 2.1 0.04 CGGAAGGCGGC L36 X68284 38 77 ↑ **2.0** \< 0.0001 GATTCCGTGAA L37 X66369 58 79 ↑ **1.4** 0.01 AAAACAGTGGC L37A X14069 55 90 ↑ **1.6** 0.0003 TGACTATTAAA L38 X57007 25 21 ↓ 1.2 0.5 TCTTCTCACAA L39 X82551 15 27 ↑ 1.8 0.03 CAGATCTTCGT L40 X82636 89 32 ↓ **2.8** \< 0.0001 AGAGCGAAGTG L41 X82550 185 80 ↓ **2.3** \< 0.0001 CAAGGTGACAG S2 U92700 52 47 ↓ 1.1 0.4 CCTCAGCCAGT S3 X51536 21 35 ↑ 1.7 0.03 GTGAAGGCGGT S3A X75161 46 43 ↓ 1.1 0.4 ATGAAATCAAA S4 X14210 44 20 ↓ **2.2** 0.01 CCTTTGAGATC S5 X58465 72 34 ↓ **2.1** 0.004 GCAGAGTGCGC S6 NM_017160 63 27 ↓ **2.3** 0.0006 TTCAGCTCGAG S7 X53377 41 31 ↓ 1.3 0.3 CCCGTGTGCTC S9 X66370 54 32 ↓ 1.7 0.06 CAGTCTCTCAA S10 X13549 57 36 ↓ 1.6 0.02 TCTGTGCACCT S11 K03250 23 18 ↓ 1.3 0.4 TATGTCAAGCT S12 M18547 92 52 ↓ **1.8** 0.006 GTGTGGCACAG S13 X53378 65 65 1.0 0.2 TTGGCTGCCCA S14 X15040 99 50 ↓ **2.0** 0.0008 GTGGGTGTGTA S15 NM_017151 254 60 ↓ **4.2** \< 0.0001 AAGAGGCAAGA S15A X77953 35 48 ↑ 1.4 0.04 TGGCCCAAATT S16 X17665 155 48 ↓ **3.2** \< 0.0001 GGCCGCGTTCG S17 K02933 18 39 ↑ **2.1** 0.003 CAGAACCCACG S18 X57529 101 26 ↓ **3.9** \< 0.0001 ACCAAGATCTA S19 X51707 136 28 ↓ **4.8** \< 0.0001 CCTACCAAGAC S20 X51537 8 16 ↑ 2.0 0.07 GGTCTGGCTAG S21 X79059 0 3 ↑ 3.0 na CCGTGGGTGAT S23 X77398 52 68 ↑ 1.3 0.03 GCCTTTATGAG S24 X52445 310 87 ↓ **3.6** \< 0.0001 CCGCCCAAAGA S25 X62482 90 43 ↓ **2.1** 0.0003 GAAAAATAAAA S26 X02414 35 16 ↓ 2.2 0.03 CACAAACGGTA S27-1 AF184893 99 46 ↓ **2.2** 0.0004 GGTAGCCACTT S27A X81839 33 77 ↑ **2.3** \< 0.0001 GAATGACCTGC S28 X59277 40 37 ↓ 1.1 0.4 CTAGTCTTTGT S29 X59051 125 123 ↓ 1.0 0.3 GTTCTCTGGCT S30 X62671 29 33 ↑ 1.2 0.3 GGATTCGGTCT P0 Z29530 40 55 ↑ 1.4 0.04 TCCAATAAAGA P1 R5RT12 96 48 ↓ **2.0** 0.0002 GGATTTGGCCT P2 X15098 83 84 ↑ 1.0 0.3 GGAGGTTATGC 40 KD RP D25224 83 34 ↓ **2.4** \< 0.0001 Rat SAGE tags and the corresponding ribosomal proteins (and GenBank accession numbers) are given along with the number of times each tag was detected. For cells before and after NGF treatment, a total of 76,280 and 87,004 11 bp tags were analyzed, respectively. Tag numbers for non-treated cultures were normalized against those for NGF treated cultures. *P* values were calculated by Monte Carlo simulations using SAGE software. Fold changes in which *P* ≤ 0.01 are expressed in **bold.** ::: Relative expression of transcripts for ribosomal proteins --------------------------------------------------------- Among the information provided by analysis of SAGE data is the relative abundances of transcripts. The data given in Table [1](#T1){ref-type="table"} and Figure [1](#F1){ref-type="fig"} show the relative abundances of transcripts for PC12 cell ribosomal proteins with respect to one another as well as with respect to the total cell complement of transcripts. As noted above, our analysis has detected at least 22,000 unique transcripts in PC12 cells. The 74 ribosomal protein transcripts identified here thus account for no more than 0.33% of this total. By contrast, for NGF-untreated and -treated PC12 cells, tags corresponding to the 74 identified RP transcripts represent 5.2% and 3.5%, respectively, of total tags analyzed. As anticipated, this clearly places transcripts for ribosomal proteins as a whole in the high abundance category. {#F1} The most abundantly expressed ribosomal transcripts in NGF-untreated PC12 cells included those encoding ribosomal proteins S15, S24 and L41. These each accounted for approximately 0.2--0.35% of total cell transcripts. The high relative abundance of these ribosomal transcripts in NGF-untreated cells can be appreciated by the observation that of the 4 tags encountered more than 200 times in our analysis, 2 (S15 and S24) encoded ribosomal proteins; moreover, RP transcripts represented 6 of the 13 tags encountered from 100--200 times and 23 of the 36 tags encountered from 50--100 times. The data in Table [1](#T1){ref-type="table"} and Figure [1](#F1){ref-type="fig"} show that NGF treatment resulted in an overall decrease in relative abundance of transcripts encoding RPs (from 5.2% to 3.5% of total). This may reflect in part our observation that NGF treatment increases the numbers of transcripts representing low-abundance genes \[[@B11]\]. Nevertheless, transcripts for ribosomal proteins remained among the most highly expressed in the NGF-treated cells. For instance, tags for RPs L13A, L31, L37A, S24 and S29 had relative abundances of 0.1% or more. Of the 56 most abundantly expressed tags in NGF-treated cells, 21 represent ribosomal transcripts. NGF selectively regulates the expression of transcripts encoding ribosomal proteins ----------------------------------------------------------------------------------- In addition to yielding an overall decrease in the abundance of RP transcripts relative to total cellular transcripts, long-term NGF treatment also promoted selective changes in relative expression of transcripts encoding individual RPs (Table [1](#T1){ref-type="table"}, Figs. [2](#F2){ref-type="fig"},[3](#F3){ref-type="fig"}). Monte-Carlo simulation analysis of the SAGE data indicated that nearly half (a total of 35) of the 74 RP transcripts underwent changes that were significant at the *P* ≤ 0.01 level (see Table [1](#T1){ref-type="table"}). Of these 35 transcripts, 2/3 were down-regulated in response to NGF. {#F2} ![**Incremental changes in expression of RP transcripts evoked by long-term treatment with NGF.** NGF-promoted changes in expression of RP transcripts are given in incremental units. Up-regulated transcripts are given as \[(number of normalized tag numbers + NGF/number of normalized tag numbers - NGF) - 1\] and are thus expressed as positive values. Transcripts that are down-regulated by NGF are given as --\[(number of normalized tag numbers - NGF/number of normalized tag numbers + NGF) - 1\] and are thus expressed as negative values. For transcripts that do not change expression in response to NGF, the incremental change is zero. **A.** Transcripts for RP L3-L41 and P0-P2. **B.** Transcripts for RP S2-S30 and 40 KD RP.](1471-2202-3-3-3){#F3} Figure [2](#F2){ref-type="fig"} shows the distribution of fold-changes in transcript levels in response to NGF. 47 of the 74 RP transcripts were either invariant or showed changes of 2-fold or less; 10 individual transcripts decreased by 3--5 fold (L12, L18, L24, L28, L30, S15, S16, S18, S19 and S24) and one was elevated by more than 3-fold (L17). In a past study \[[@B11]\], the differences in NGF-promoted gene expression revealed by SAGE analysis of our libraries were found to be highly reliable based on 1) the absence of regulation shown by a number of anticipated \"housekeeping\" genes including β-actin, 2) the observation that the vast majority of transcripts were not responsive to NGF, 3) the detection of anticipated changes in expression of a number transcripts previously reported to be NGF responsive, and 4) the agreement in relative expression of 20 genes as revealed by SAGE and northern blot analyses. To further confirm the reliability of our SAGE findings, we used real time quantitative PCR to compare levels of 5 different RP transcripts in PC12 cells treated with or without NGF for 9--12 days. As shown in Figure [4](#F4){ref-type="fig"}, both techniques revealed similar changes in expression. {#F4} Rapid regulation of an RP transcript by NGF ------------------------------------------- We next determined the time course with which NGF regulates an RP transcript. For this purpose, we studied RP L9 which undergoes a 2.7-fold elevation in response to long-term NGF treatment. Because the changes in expression are relatively small, we used real time PCR for this end. As shown in Figure [5A](#F5){ref-type="fig"}, the relative abundance of transcripts encoding RP L9 were significantly upregulated by 1 hr of NGF treatment (1.8 ± 0.2-fold, n = 14) whereas there was no significant change in transcripts encoding RP S29, which does not undergo long-term NGF regulation. A time course (Figure [5B](#F5){ref-type="fig"}) revealed that elevation of L9 transcripts reaches maximal levels within 8 hrs of NGF exposure. ![Rapid regulation of RP L9 transcripts by NGF. **A.** Elevation of RP L9 transcripts after 1 hr of NGF exposure. PC12 cell cultures were treated with or without NGF for 1 hr and used for preparation of RNA and cDNA. Relative levels of L9 and S29 transcripts were determined by real-time PCR with normalization against levels of transcripts for GAPDH. Data are given as ratios for values of \[NGF treated/untreated\] and are expressed as means ± SE (n = 14 for L9 and n = 6 for S29). **B.** Time course for NGF response of transcripts encoding RP L9. PC12 cultures were treated with NGF for the indicated times and used for preparation of RNA and cDNA. Relative levels of L9 transcripts were determined by real-time PCR with normalization against levels of transcripts for GAPDH. Values are given as means ± SE (n = 3).](1471-2202-3-3-5){#F5} Discussion ========== NGF regulates expression of RP transcripts ------------------------------------------ In the present work, we used SAGE profiling to identify and quantify the relative numbers of transcripts for 74 ribosomal proteins in rat PC12 cells before and after long-term exposure to NGF. This has permitted us to detect selective changes in transcripts encoding specific RPs. A number of criteria support the reliability of our findings. The first regards the assignment of SAGE tags to specific RP transcripts. We required that all matching ESTs or transcripts have clearly definable poly A tails and poly adenylation signals. Due to the extensive data base for rat RP transcripts, it was possible to make all matches with rat sequences. In addition, we used 15 mer (CATG +11) base SAGE tags for our analysis; we have reported \[[@B11]\] that this leads to a significantly more reliable matching of SAGE tags to genes than with the often-used 14-mer tags. A second indicator of reliability stems from prior analysis of our SAGE libraries and data obtained from them \[[@B11]\]. For instance, we have shown that our SAGE libraries show little or no NGF-promoted changes in tag numbers for transcripts encoding a number of housekeeping proteins. Moreover, a number of transcripts previously found by alternative technologies to respond to NGF treatment, showed similar changes in our SAGE profiling study. Finally, Northern blot analysis verified SAGE-predicted responses of over 20 transcripts to NGF treatment. A third criterion for reliability was that Monte Carlo simulation indicated that many of the NGF-promoted responses of RP transcripts were at a probability of *P* \< 0.01. Because of the relatively high abundance of many RP transcripts, even comparably small changes in expression could be detected at this level of significance. A last criterion was that we used real time RT-PCR to verify effects of NGF on five RP transcripts. A previous study by Lee et al. \[[@B8]\], based on comparison of a total of approximately 7,000 random ESTs from naïve and NGF-treated PC12 cells, reported an NGF-promoted increase in expression of RPL7 transcripts and a decrease in RPL19 transcripts. This contrasts with the current data which revealed no significant change in expression of either of these transcripts. The reason for this discrepancy is unclear, but could in part originate from the relatively small number of transcripts that were surveyed in the former study. To our knowledge, there are currently no other reports regarding effects of NGF on levels of ribosomal transcripts. Relative levels of RP transcripts --------------------------------- Our observations indicate that there is a wide range in the numbers transcripts per cell that encode individual RPs. For instance, both before and after NGF treatment, there was over a 10-fold difference in the relative numbers of transcripts for RPs L7, L22, S21 as compared to those encoding RP S29. Because of the general scarcity of antisera prepared against mammalian RPs, we do not know whether this is reflected at the protein level. However, for a number of other NGF-regulated transcripts in PC12 cells, there is a good correlation between relative levels of message and protein \[[@B11],[@B16],[@B17]\]. Thus, it may be that although levels of individual RPs are assumed to be similar to one another, some may be present in limiting numbers. Alternatively, there may be a considerable disparity between relative abundances of RP transcripts and their corresponding proteins A recent SAGE study of targets for N-myc in a human neuroblastoma cell line \[[@B18]\], reported relative abundances for 66 RP transcripts. The existence of such data permit us to compare the relative levels of RP transcripts in two cell types (neuroblastoma and pheochromocytoma) of related origin (i.e., neural-crest-derived) that both have the potential for neuronal differentiation. The 74 RP transcripts detected here represented 5.2% of the total transcripts in NGF-untreated PC12 cells and 3.5% after NGF treatment. By contrast, the 66 RPs reported in the neuroblastoma study accounted for 4.1% of total transcripts for cells without N-myc over-expression and 12.6% for such cells transfected with N-myc. Thus, for non-N-myc transfected neuroblastoma cells and PC12 cells (± NGF) the contribution of total RP transcripts lies in a similar range and this parameter is greatly elevated in neuroblastoma cells by N-myc over-expression. With respect to transcripts for individual RP\'s, there are many similarities as well as several striking differences between the two cell types. When NGF-untreated PC12 cells are compared with non-transfected neuroblastoma cells, of the 55 RP transcripts detected in common for both systems, about half (a total of 28) have relative abundances within a factor of 2. Twenty transcripts are more than 3-fold higher in relative abundance in PC12 cells. In two of these cases (S9 and S17) no tags were detected in the non-transformed neuroblastoma cells; in several other cases (S12, S24, L24) the relative abundance in PC12 cells was over 20-fold that in the neuroblastoma cells. For most of these, expression of N-myc elevated the neuroblastoma levels to relative abundances within 2-fold of those in PC12 cells. However, for L24 and S24, the relative abundance in N-myc-expressing neuroblastoma cells was still 1/3 that in PC12 cells. Finally 7 RP transcripts (L7, L11, L21, L30, S7, S8 and P2) were of 3--5 times lower abundance in PC12 cells than in neuroblastoma cells and this difference was appreciably enhanced when the NB cells were transfected with N-myc. We have also compared our RP results with those for SAGE carried out with pooled human adult brain (tissue supplied by Gregory J. Riggins: <>). Comparison with data for NGF-treated cells reveals similarity in relative abundance within a factor of 2 for 39 of the 57 RP transcripts in common between the libraries and 48 of 57 to within a factor of 3. However, several large differences do occur. For example the relative abundances of transcripts encoding RPs L9, L39, S24, S13, and S17 are 5--10-fold higher in NGF-treated PC12 cells than in the pooled brain library and transcripts for RP L28 are 8-fold lower. These observations reinforce the notion that expression of individual RP transcripts can be significantly variable from cell type to cell type as well as sensitive to extrinsic signals. Potential significance of RP transcript regulation -------------------------------------------------- Although the ribosome has been considered as a \"molecular machine\" \[[@B19]\], it is of interest that the transcripts encoding individual proteins of this organelle are subject to regulation by NGF. Two related questions emerge regarding these findings: how do these changes compare with previous reports for regulation of RP transcripts and what might be the functional consequences of these changes? A number of studies have employed a variety of techniques to detect changes in gene expression associated with the oncogenic state and have reported selective elevation of specific RP transcripts in tumors. Examples include RPs L7a, L37 and S14 in prostate tumors and cell lines \[[@B20]\]; RPs L5, L7A, L18, S3, S6, S8, S12, S13, S28, P0 and P1 in colorectal cancers and cell lines \[[@B21]-[@B26]\]; L5 in astrocytomas \[[@B27]\]; L18a in squamous cell carcinoma \[[@B28]\]; L19 in breast tumors that over-express erbB-2 \[[@B29]\]; RPs S3A, S4, and S17 in lymphoid malignancies \[[@B30]\]; L38, S4, P0, and P1 in rhabdomyosarcoma cell lines \[[@B31]\]; and P0, P1, P2, L5, L9, L35, L39, S3A, S10, and S17 in liver tumors \[[@B32]\]. In addition, as noted above, a recent SAGE study identified targets of N-myc in a human neuroblastoma cell line \[[@B18]\]. Of 114 up-regulated genes detected, 66 encoded RPs with elevations ranging from 40% to 37-fold. Several of these genes were also up-regulated by c-myc. Taken together, these findings indicate that the transition to the transformed state is associated with elevation of various RP transcripts. A potential interpretation of such observations is that this up-regulation reflects the enhanced rate of cell division in tumor cells and the requirement for greater levels of protein synthesis. In agreement with this possibility, transcripts encoding RPs L6 and S7 are reported to be up-regulated in regenerating liver \[[@B33]\]. However, for the study involving responses of neuroblastoma cells to N-Myc, despite the massive up-regulation of RP transcripts, there was no over-all increase in the rate of protein synthesis \[[@B18]\]. Before considering potential functional consequences of the changes in RP expression described here as well as elsewhere, it must be conceded that changes in transcript abundance may not necessarily lead to changes in protein expression. The present lack of available antibodies/antiserum to most mammalian RPs hampers such a determination. However, even if cases occur in which altered RP transcript expression does not lead to changes in expression of the corresponding protein, our and others\' findings regarding selective regulation of RP transcripts would then raise the interesting issue as to why, and the mechanism by which, transcript and protein expression are uncoupled. If at least some of the NGF-promoted alterations in RP transcript abundance lead to changes in expression of the corresponding proteins, what might be the functional consequences? In the present system, NGF converts proliferating PC12 cells to a non-proliferating neuronally differentiated state. Our findings reveal that the relative overall abundance of transcripts for RP proteins fell by 1/3 in response to NGF. In addition, nearly half of the individual RP transcripts detected showed significant changes in expression and approximately 2/3 of these were decreases. This overall decrease in NGF-promoted RP transcript expression and the preponderance of decreases in expression of specific RP transcripts compared to increases would favor the interpretation that the changes observed here reflect, at least in part, the transition to the non-dividing phenotype. On the other hand, we observed that many of the RP transcripts did not show significant responses to NGF and that a number showed increases in expression. This suggests that the situation is likely to be more complex with individual RPs perhaps playing specific roles not simply associated with the state of cell proliferative capacity. In agreement with this, we found no significant change in expression of RP S29 transcripts even though these have been reported to be present at low levels in growth phase cells and elevated in quiescent cells \[[@B34]\]. In addition to leaving the cell cycle, NGF-treated PC12 cells undergo neuronal differentiation. This raises the possibility that some of the observed changes, as for many other NGF regulated genes, is related to acquisition of the neuronal phenotype. An analysis of gene regulation in human NTERA2 cells induced to leave the cell cycle and neuronally differentiate in response to retinoic acid revealed decreases in RP transcripts including L3, L7, L8, L10, L13, L39, S2, S6, S13, S16, S20, S19, S23, S27A and P0 \[[@B35]\]. Of these, In the present study only RP transcripts L8 and S6, S16, S19 showed significant down regulation in response to NGF whereas the others found in the NTERA2 study either did not show changes that were significant at the *P* ≤ .01 level or underwent a significant increase (RP S27A). Curcic et al. \[[@B36]\] reported that differentiation of BC3H1 myocytes is accompanied by a drop in L32 gene transcription; in contrast, we found no significant change in transcripts encoding this RP. In another study, RPs L35a and S5 were down regulated during murine erythryolukemia cell differentiation \[[@B37]\]; we also observed decreases in these transcripts. Mutation of RP S19 is associated with Blackfan\'s anemia and therefore appears to play a selective role in differentiation/proliferation of erythropoetic cells \[[@B39]\]. Here, NGF resulted in a nearly 5-fold drop in S19 transcripts. Thus, although changes in expression of specific RP genes may be associated with differentiation, the pattern of such changes may reflect the particular cell type and differentiation stimulus that is involved. Another condition associated with changes in RP expression is apoptotic death. RP L4/L1 transcripts, which were not significantly affected by NGF, are selectively upregulated in PC12 cells prior to 5 aza cytosine-induced death and over-expression of this gene in COS-7 cells induces apoptosis \[[@B39]\]. Over-expression of L7, L13A, S29 have been also reported to induce apoptosis \[[@B40],[@B41]\]. In the present study, NGF promoted little if any change in expression of these RP transcripts. Thus, although NGF is an effective anti-apoptotic factor, these actions do not appear to be mediated by down-regulation of potentially death-inducing RPs. At present, in contrast to studies on bacterial and archael ribosomes, relatively little is known about the functions of individual mammalian ribosomal proteins (\[[@B42]\]. However, assuming that changes in message abundance lead to alterations in protein levels, it is reasonable to anticipate that the NGF-promoted effects observed here may affect the protein synthetic capacity of the cell in some manner. Such changes are unlikely to be global; NGF does affect the cellular rate of protein synthesis, but this seems to be due at least in part to post-translational modification of the translational machinery \[[@B43]\]. The alternative is that the changes reported here may result in selective effects on translation of specific messages. A final, and important possibility to consider is that a number of RPs appear to possess extraribosomal functions \[[@B44]\] and consequently that the changes reported here may lead to responses that do not directly relate to ribosomal function. For example, RP L18, which was significantly down-regulated by NGF, has been reported to negatively regulate double-stranded RNA (dsRNA)-activated protein kinase (PKR) \[[@B45]\]. As an additional example RP S19, which is also down-regulated by NGF, has been implicated in erythropoiesis and has also been described as a monocyte chemotactic factor \[[@B46]\]. As one last example, rat RP L9 which we found to be up-regulated by NGF starting within a few hours and maximally by 8 hrs, has been shown to be the ortholog of the yeast gene *grc5* which appears to be involved in multiple cellular functions including growth control, cytoskeleton control and energy metabolism \[[@B47]\]. Identification of NGF-regulated RPs should now facilitate exploration of their potential roles in the responses of cells to neurotrophins. Conclusions =========== 1\. SAGE analysis provides a reliable, quantitative picture of ribosomal protein expression in PC12 cells before and after long term exposure to NGF. 2\. Transcripts for ribosomal proteins are among the most abundant transcripts in the cells; however there is a wide range between numbers of transcripts for individual RP transcripts. 3\. NGF promotes an overall decrease in relative RP transcript expression (for the 74 RP transcripts detected, from approximately 5% of total transcripts to 3.5%). This drop represents a decrease in relative expression of individual RP transcripts as well as an increase in message complexity in NGF-treated cells. 4\. Long term NGF treatment of PC12 cells promotes statistically significant changes in expression of over half of the transcripts encoding ribosomal proteins. Decreases in expression outnumbered increases by a ratio of approximately 2:1. The largest observed changes in expression are in the range of 3--5 fold. 5\. For at least one RP transcript (RPL9) the response to NGF is rapid; an elevation is detected within 1 hr of NGF exposure and reaches maximum regulation by 8 hrs. 6\. Examination of the literature reveals many other reports in which expression of individual or groups of RP transcripts are regulated in response to growth factors, differentiation agents or malignant transformation. Comparison with the present data indicate that changes in ribosomal protein transcripts is regulated in a cell and state dependent fashion with a large diversity in the particular RP transcripts that are subject to regulation. Thus, although the ribosome may be regarded as a \"machine\" there is a good deal of potential for plasticity with respect to the expression of its various components. This raises the possibilities that growth factors and other cell regulators may affect ribosomal function, and thereby the capacity of cells to transcribe specific transcripts. 7\. In the absence of information about the specific functions of most individual RPs in mammalian cells, one can only speculate at present on the physiologic significance of the reported changes. However, it seems highly plausible that the observed responses may play roles in NGF-promoted neuronal differentiation. This may be mediated in part by selective effects on translation of certain messages. In addition, past findings support the possibility that at least some of the regulated RPs have extra-ribosomal actions that may affect neuronal differentiation and function. Materials and Methods ===================== SAGE and matching SAGE tags with RP transcripts ----------------------------------------------- SAGE libraries were prepared from naïve and 9-day NGF-treated PC12 cells and sequenced as previously described \[[@B11]\]. To match SAGE tags with RP transcripts, tags were initially analyzed with the National Center for Biotechnology Information (NCBI) rat SAGE tag to gene mapping database <>, which matches possible 14-mer tags with known rat genes and expressed sequence tags (ESTs). With the use of sequences present in the NCBI UNIGENE rat database <> potential matches were further scrutinized to determine whether there was a match at the 15th base and to determine whether the matched sequence was at the most 3\' end of a known rat transcript or EST. We considered only cases in which a clear poly(A) tail and a polyadenylation signal were present at the 3\' end of the transcript or EST. Appropriate ESTs were further analyzed by an advanced BLAST search for matches with known rat RP genes. Monte-Carlo simulation analysis of the SAGE data was carried out using the SAGE 300 software package \[[@B13]\]. Real time quantitative PCR analysis ----------------------------------- PC12 cells were cultured as previously described \[[@B12]\] in complete medium (85% RPMI 1640 medium, 10% horse serum, 5% fetal bovine serum). Replicate cultures were treated with 50 ng/ml rhNGF (kindly donated by Genentech, Inc.) for the indicated times. Total cellular RNA was isolated as described previously \[[@B11]\] and 1.2 -- 5 μg was used for reverse transcription with 5\'-T~30~NN-3\' primer using Superscript II Rnase H-RT according to the manufacturer\'s specifications (Life Technologies, Grand Island, NY). Quantitative real time PCR reactions were performed as described by Troy et al. \[[@B48]\]. GAPDH was used to normalize input cDNA for samples originating from short-term NGF treatments and β-actin was used for this purpose for samples with long-term NGF exposure. Forward and reverse primers pairs used for quantitative PCR were: *S29,* 5\'GGTATCACAGGGTAGACAGT3\' and 5\'GGTATCACAGGGTAGACAGT3\'; *L9,* 5\'GAACTCTGAGGAGGGACTTC3\' and 5\'AAACTGTACTTGTTATCAGGAT3\'; *S27-1,* 5\'CGGCACGAGCGACCTCCCTA3\' and 5\'GTTCCCACTCATCTTGAATC3\'; *L24,* 5\'CAAGAAAGGACAGTCGGAAA3\' and 5\'TTCACAGGCTTCACAATCTT3\'; *S19,* 5\'TAACCAGCAGGAGTTCGTCA3\' and 5\'TTTGTTCTAATGCTTCTTGTT3\'; β-*Actin*, 5\'ATCCTGACCCTGAAGTACCC3\' and 5\'TACGACCAGAGGCATACAG3\'; and *GAPDH,* 5\'GAAACCTGCCAAGTATGATGA3\' and 5\'TCTCTCTTGCTCTCAGTATCC3\', respectively. Acknowledgements ================ We thank Claudine Bitel for her excellent technical assistance. Supported in part by grants from the National Institutes of Health.

Background ========== Over the last 50 years, the widespread usage of fluoridated water and fluoridated dentrifices have been cited as major reasons for a decline in caries since the early 1970s \[[@B1]\], and for the appearance of a significant association between oral hygiene and caries experience \[[@B2]-[@B4]\]. An inverse relationship exists between salivary fluoride concentration and caries experience in the deciduous and permanent dentition \[[@B5]\], but fluoride concentration is excluded from most caries prediction models \[[@B6],[@B7]\]. Acids in bacterial plaques cause caries in pits, fissures and interdental regions of teeth, but they also enhance the inhibitory effect of fluoride on demineralization, confounding the ability to predict caries from the salivary fluoride concentration \[[@B8],[@B9]\]. The greater the consumption of dietary sucrose, the greater the fall in pH and fraction of acidogenic, acid tolerant bacteria in tooth adherent plaques \[[@B10],[@B11]\]. The number of these bacteria (mostly mutans streptococci and lactobacilli), and the fluoride content, discriminate between severe and mild caries in 12--15 year-old children \[[@B12],[@B13]\]. Acid-tolerant bacteria require D-alanyl glycerol lipoteichoic acid (D-alanyl LTA) in their membranes and cell surfaces \[[@B14]\]. D-alanyl LTA is made by esterifying carboxyl-activated D-alanine to glycerol in membrane LTA by means of a D-alanyl-carrier enzyme, DCP \[[@B15]\]. Strains of *Streptococcus mutans* in which DCP is inactive do not initiate growth at below pH 6.5 and make glycerol LTA without D-alanine \[[@B14]\]. In the DCP active strains, soluble D-Alanyl LTA is extruded into culture fluid *in vitro*\[[@B16],[@B17]\] or plaque *in vivo*\[[@B18]\]. The D-alanyl esters are stable at pH 6.0 at 37°C, but hydrolyze to free D-alanine and LTA with a half-life of 3.9 h at pH 8.0 \[[@B19]\]. Healthy gingival sulci have a pH of 6.5 -- 7.5 and inflamed sulci a pH of 7.5--8.5 \[[@B20]\]. About 30% of young adults have serum IgG antibodies that precipitate with D-alanyl LTA, but not with D-alanine-free LTA \[[@B17],[@B21]\]. It is likely that plaques induce these IgG antibodies from gingival sulci that contain more acid-tolerant bacteria. An elevated IgG antibody response to D-alanyl LTA may therefore indicate the subjects in whom an inhibitory effect of fluoride on caries is enhanced. The fluoride concentrations of plaque and saliva are related to whether the drinking water is fluoridated \[[@B13]\] and to oral hygiene, which nearly always involves using a fluoridated dentrifice. The aim of this study was therefore to determine whether elevated antibody responders to D-alanyl LTA show a association of DMFT with fluoride exposure and gingival health not apparent in low responders. Methods ======= Subjects -------- Antibody was obtained from blood from four sources: 1) 105 dental students, 2) 147 patients seeking dental treatment, 3) 145 volunteer blood donors (volunteers), and 4) 37 siblings aged 5 through 25 from six Amish families. The dental students and patients were attending the University of Oklahoma Health Sciences Center between 1985 and 1988. The volunteers and Amish family members were attending centers elsewhere in the US at the same time. All subjects consented to provide blood for antibody analysis according to local Institutional Review Board procedures (see Acknowledgements). The student, patient and volunteer populations (397 subjects) were used to determine what IgG concentration constituted an elevated antibody response to D-alanyl LTA, to ensure that these antibodies were not unique to dental or Oklahoma populations and to examine whether the antibody concentration was sex or age-associated. The Amish family siblings were selected to determine the frequency of high antibody concentration in children and young adults. Each sibling had at least one parent high responder to increase the likelihood of exposure to an antibody-associated oral microbiota from birth. The clinical study participants consisted of 87 dental students (88.4% male) and 64 patients (31.3% male) who were medically healthy. All had 18 or more natural teeth and were aged \>22 and \<38 years. Of these participants, 67 dental students and 35 patients provided information that permitted an estimate of exposure to fluoridated water: residence(s) from birth through age 14 in the 1980 Fluoridation Census. Subjects not using the public water supply, or resident outside of the US for more than 18 months, were excluded. Exposure to water fluoridation scored 1 for each of five 3-year age cohorts: 0--2, 3--5, 6--8, 9--11 and 12--14 to give a fluoride exposure score (F Score) of 0 (no exposure) to 5 (complete exposure) described previously \[[@B22]\]. Most dental students had mild caries and gingivitis and most patients had moderate to severe caries and gingivitis. Exceptionally healthy or exceptionally diseased subjects were therefore increased compared to a similar number of subjects obtained as a random sample. This wide distribution of clinical measurements provided more stable estimates (narrower confidence intervals) of regression coefficients (β) than would be obtained from a similar number from a random survey of the general population. Regression lines are more robust when a greater range of measurements is used \[[@B23]\]. Clinical measurements --------------------- Dental caries experience was the number of Decayed, Missing and Filled Teeth (DMFT), excluding third molars and teeth reported missing for other reasons. Decayed teeth (DT) were also enumerated separately from missing and filled teeth (MFT). DT indicates a combination of delay in seeking therapy and faster development of new cavities. Fluoridated dentrifice use is related to oral hygiene but not toothbrushing frequency in adults aged as in the present clinical study \[[@B3]\] and young enough to have likely used fluoridated dentrifices from early childhood. Sensitive staining for plaque accumulation \[[@B24]\] was therefore used with measures of gingivitis and pocket depth at the mesio-buccal, buccal, disto-lingual and lingual surfaces of the six teeth employed for the simplified oral hygiene index \[[@B25]\], substituting adjacent teeth as necessary (24 sites sampled). Gingivitis was determined by whether a site bled within 30 sec of gentle probing, BOP \[[@B26],[@B27]\] and pocket depth by measuring the distance (mm) from the free gingival margin to the base of the sulcus or pocket. Finally, each subject was asked to suck an erythrosin tablet for 30 sec and the sites examined for stained plaque \[[@B24]\]. For each subject, the mean prevalences of plaque (PL) and bleeding on probing (BOP), and the mean pocket depth (PD), were calculated across all sampled sites. The clinical measurements were made by two experienced clinicians who were calibrated for this study. The clinicians agreed strongly with respect to all measurements (correlation coefficients, r \> 0.85; p \< 0.001) except gingival bleeding index, for which a weaker correlation was noted (r = 0.60, p \< 0.001). The data reported are the mean measurements from the clinical examiners. Antigen purification -------------------- D-alanyl LTA, but not D-alanine-free LTA, is present in culture filtrates of *Streptococcus mutans* GS5 \[[@B17]\]. Bacteria were grown at 37°C in trypticase soy broth to late stationary phase (96 h), when the maximal amount of antigen is extruded into the culture fluid \[[@B18]\]. After centrifugation to remove the bacteria, culture fluid (10 1) was concentrated 20-fold over a YM10 Diaflo Membrane filter (Amicon Corp., Beverley, MA). D-Alanyl LTA in the concentrate was detected by immunoelectrophoresis, using a standard human serum identified previously \[[@B16]\]. D-Alanine-free LTA does not react with this serum IgG \[[@B17],[@B18],[@B21]\]. The D-alanyl LTA was purified by passing the concentrated culture fluid over a 90 × 2.5 cm Sephacryl column in 0.4 M NaCl buffered with 0.05 M sodium acetate pH 5. Antigen in the fractions was collected. After equilibrating with 5 mM sodium acetate buffer pH 5.0, it bound to a short Sephacryl S-200 and eluted by adding 14 mM NaCl as described previously \[[@B16]\]. Measuring antibody content and determining high and low responders ------------------------------------------------------------------ IgG antibody content was measured by enzyme-immunoassay employing a Fast Assay Screening Test System at room temperature \[[@B28]\]. Pegs protruding from a lid were placed over a 96-well plate or trough containing 14 ml of 10 μg/ml D-alanyl LTA in acetate buffer pH 5 for 2 h (Becton Dickinson, Lincoln Park, NJ). The pegs were blocked with 14 ml of phosphate buffered saline (PBS) pH 7.0 in 1.0% Tween-20 and immersed in wells containing 0.1 ml serum. After overnight incubation, excess IgG antibody from the serum was washed away by thrice transferring the pegs to troughs containing 14 ml of PBS containing 0.05 % Tween 20 (PBS-Tween) for 5 min each time. The pegs were then immersed for 2 h in a trough containing 14 ml of anti-human IgG F(ab\'2) fragment conjugated to alkaline phosphatase in PBS-Tween and developed with nitrophenyl phosphate (Sigma Chemical Co. St Louis, MO). The concentration of antigen-specific IgG in standard serum was obtained by measuring the optimal amount of protein immunoprecipitated \[[@B16]\], and a standard curve of absorbance against concentration was obtained (Fig. [1](#F1){ref-type="fig"}). The greatest range of absorbance occurred when sera were measured at a dilution of 1:200 \[[@B28]\]. Replicate antibody assays were performed on each serum and the concentrations read off the standard curve. The antibody concentrations are shown ranked in Fig. [2](#F2){ref-type="fig"}. {#F1} {#F2} Data analyses ------------- The amount of IgG antibody to D-alanyl LTA varies with no obvious cutoff (Fig. [2](#F2){ref-type="fig"}). However, the sera containing precipitating antibody should tend to have high IgG antibody contents. The IgG antibody measurements were alternatively divided into clusters, using the unweighted pair group method with arithmetic averages \[[@B29]\] and NT-SYS, a package of multivariate statistical computer programs \[[@B30]\]. A low response supremum was obtained by taking the antilog of the mean IgG content of the non-precipitating sera plus one standard deviation, or the antilog of the highest IgG content of the cluster grouping containing the least antibody. The effect of age was determined after splitting the subjects into decile cohorts (Table [1](#T1){ref-type="table"}) and comparing the fraction of high antibody responders in each cohort. The youngest cohort was composed of Amish family siblings who were younger than any dental students, patients or volunteers. ::: {#T1 .table-wrap} ::: {.caption} ###### Age decile cohorts for determining changes in fraction of high antibody responders ::: Cohort No. Decile ^a^Number in Cohort High Responders ------------ -------- --------------------- ----------------- 1 5--15 ^b^19 ^c^21.05% 2 15--25 ^b^97 35.05% 3 25--35 176 36.36% 4 35--45 76 40.79% 5 45--55 31 38.71% 6 55--65 13 ^c^23.08% 7 65--72 6 ^c^16.67% ^a^381 subjects after excluding all Amish family members and 16 of the 397 dental student, patient and volunteer subjects whose age was not recorded. ^b^The Amish family siblings \<15 years comprised Cohort 1 and those \>15 years were included within Cohort 2. ^c^Comparison of cohorts 1, 6 and 7 with the remainder: X^2^ = 3.21, p = 0.073 (not significant). ::: A multiple linear regression procedure was utilized to examine the relationship of caries (DMFT) with age, F score, and measures of gingival health obtained in this study: PL, BOP, and PD. The regression on DMFT was used: 1) to estimate the partial regression coefficients (β coefficients) within the high and low antibody response group; 2) to test each β coefficient for significance after accounting for the effects of the other four variables; and 3) to examine for significant differences in β coefficients between the antibody response groups. A β coefficient is interpreted as the change in disease response (DMFT) per unit change in one of the independent variables after adjusting for all the other independent variables in the model. Within each antibody response group, the multiple regression coefficient (R^2^) provided an estimate of the proportion of variance in DMFT explained by the combination of variables tested. Stepwise regression then identified the best estimate of the variance in DMFT that was explained by multiple variables in the separate and combined high and low responder groups. These multiple regression analyses were repeated using gingivitis (BOP) as the dependent variable and OHPI, PD, DT and MFT and age as independent variables. All of the clinically examined subjects were included because F score was not an independent variable for BOP. To ensure that obtained relationships were robust, influential points (outliers) were identified using a statistic (DFFITS) which measured the change in coefficients caused by removing the data for each subject. If this change exceeded 2√(p/n), where p was the number of independent variables and n the number of samples \[[@B31]\], the point was influential. Repeating the regressions with all the influential points removed determined the degree to which these points had affected the results. Results ======= Definition of high antibody response ------------------------------------ IgG antibodies were initially detected in sera irrespective of whether D-alanyl LTA was immunoprecipitated. Excluding the Amish family group, there were 288 subjects whose sera failed to immunoprecipitate antigen (detected by immunoelectrophoresis). The mean IgG antibody concentration (log ng/ml) was 3.19 (0.64 standard deviation, s.d.) compared with 4.25 (0.57 s.d.) for 109 subjects whose sera did precipitate antigen (\'t\' test p \<10^-6^). High responders therefore had a log antibody content (ng/ml) that exceeded the mean plus standard deviation of non-precipitating serum (log ng IgG /ml \>3.83). However, low IgG concentrations formed a cluster whose supremum (log ng IgG antibody/ml) was 3.861, which corresponds to 7.26 μg/ml (left side of Fig. [2](#F2){ref-type="fig"}). Subjects were therefore classified as low responders if their log IgG antibody content exceeded 3.86 rather than 3.83. The odds ratio for a serum from a high antibody responder immunoprecipitating D-alanyl LTA was 14 times greater than for a low responder. High antibody response, age and gender -------------------------------------- Of the 397 students, patients and volunteers, 16 did not have their age recorded. High responders had a mean age of 32.7 years ± 9.4 s.d. (136 subjects) and low responders a mean age of 33.0 years ± 10.7 s.d. (245 subjects). Table [1](#T1){ref-type="table"} shows the fraction of high responders in different age cohorts. There was a high responder frequency of 35--40% from age 15 through 54. The reduced frequencies of high response in childhood and old age were not significant. However, within the Amish family offspring high IgG responders were significantly older. Table [2](#T2){ref-type="table"} shows that the high responder offspring had a mean age of 17.1 years compared with 13.3 years for low responders. This significant difference (\'t\' statistic = 2.42, degrees of freedom, d.f. = 35; p \< 0.03) was due to few high responder children and young teenagers and a slightly greater fraction of siblings aged 15--25 years who were high responders (50%) compared with the general population. ::: {#T2 .table-wrap} ::: {.caption} ###### Age of siblings from 6 families with at least one high responder parent ::: High responders Low responders ---------------------- ----------------- ------------------ F+M- 17 16,14.13,12,10,8 F+M+ 20,13 21,18,16,8,5 F+M+ 22,20,18 25,16,14 F?M+ None 12,11,10,9,8 F+M- 22,20,13,11 17 F+M\' 17,16,13 20,18,10,7 ^b^Mean age (s.d)^c^ 17.08(3.68) 13.25(5.00) ^a^F, Father; M, Mother; +, high responder; -, low responder, ? Not known. ^b^Mean age of the high and low responders ^c^s.d., standard deviation. ::: High antibody responders accounted for a similar fraction of subjects irrespective of whether they were in the clinical study, or dental students, or patients (Table [3](#T3){ref-type="table"}). Men were 49.1% of the 395 students patients and volunteers whose sex was recorded. Men had also a greater frequency of high response 40.21% vs 31.34% and a greater mean log ng/ml IgG antibody content, 3.52 ± 0.76 standard deviation (s.d.) vs 3.44 ± 0.80 s.d. Neither of these differences were significant (X^2^ = 3.00, p = 0.09; \'t\' statistic = 0.34, p = 0.73). In serum samples from 18 subjects aged between 22 and 38, the IgG antibody concentrations were essentially the same after 6 months as originally estimated (squared correlation coefficient, R^2^ \>0.95, p \< 0.01). The results indicate that, for subjects aged 15 -- 55, age and sex had little effect on the frequency of high D-alanyl LTA antibody response. ::: {#T3 .table-wrap} ::: {.caption} ###### Fraction of high antibody responders in or not in the clinical study. ::: Subject group Number \% high responders ----------------------------------------- -------- -------------------- Dental students in clinical study^a^ 87 37.9 Patients in clinical study^a^ 64 32.8 Same-age subjects not in clinical study 129 31.8 Other subjects not in clinical study^c^ 117 39.7 All subjects 397 33.8 ^a^See first section of Materials & Methods. ^b^Same-age subjects not in the clinical study were 12.4% dental students, 17.1% patients and 70.5% volunteers. ^c^Other subjects were a mixture of patients and volunteers: 70.9% older (ages \>38 and \<72 years), 15.4% younger (ages \>15 and \<22 years), and the remainder age unknown. ::: Association of DMFT with gingival health and fluoride in high and low responders -------------------------------------------------------------------------------- Table [4](#T4){ref-type="table"} lists the variables tested for association with DMFT and the observed β coefficients in high and low antibody responders. It was immediately apparent that the β coefficients from PL were similar in both groups, whereas those from BOP and F score were only significant in high responders and those from PD were only significant in low responders. Comparison of the differences in β coefficients between high and low responders, column 4 (column 2 -- column 3), indicated relationships of DMFT to pocket depth that were significantly different and relationships of DMFT to F score that were almost significantly different, p = 0.062 (Table [4](#T4){ref-type="table"}, column 4). ::: {#T4 .table-wrap} ::: {.caption} ###### Changes in the equality of the partial β coefficients for association of tested variables withcaries severity in high and/or low responders^a^. ::: Variable^a^ High responders ^b^n = 35 Low responders n = 67 Difference (Hi -- Low) ------------- --------------------------- ----------------------- ------------------------ F score ^c^-0.847 0.006 ^d^-0.853 PL ^d^0.259 ^d^0.234 0.025 BOP ^c^0.447 0.167 0.280 PD -1.903 ^c^4.249 ^c^-6.152 Age 0.099 ^d^0.261 -0.162 ^a^Variable names are defined under \"Clinical Measurements\" in the Methods ^b^n = number of subjects ^c^p \<0.05 for value of constant or variable (β) or for all values in indicated model. ^d^*p* \> 0.05 & \< 0.15. If no subscript, p \> 0.15 ^e^HLS: high response = 1, low response = 0. ::: Stepwise regression confirmed the similar associations of DMFT with plaque prevalence in both high and low IgG antibody responders, but significant associations with only F score and BOP prevalence in high responders and with only PD and age in low responders. In high responders, DMFT increased as plaque and BOP prevalences increased and fell as fluoride exposure increased. The equation obtained was: DMFT = 4.60 + 0.28 PL + 0.39BOP - 0.88FScore (R^2^ = 0.51, F statistic = 10.75, p \< 0.0001). By contrast, in low responders, DMFT increased with plaque prevalence, pocket depth and age. The equation was: DMFT = -13.37 + 4.58 PD + 0.27 Age + 0.30 PL (R^2^ = 0.26, F statistic = 7.41, p \< 0.0003). Within each equation, the constant and the respective β coefficients were significant (p \< 0.05), except for the β coefficient of age in low responders (p = 0.062). High responders receiving fluoridated drinking water for all 14 years of childhood (F score = 5), had a significantly lower DMFT than those never receiving fluoridated water (F score = O): DMFT = 7.50 ± 4.52 (s.d.) vs 11.60 ± 4.06 (s.d.); \'t\' test p \< 0.04. This was not true of low responders in whom the difference between F score 5 and F score 0 was not significant (DMFT = 9.33 ± 5.72 vs 12.07 ± 5.85; \'t\' test p = 0.14). When antibody was ignored in stepwise regression (control), DMFT increased with age, PL and BOP, and decreased with F score: DMFT = -2.11 + 0.17 Age + 0.25 PL + 0.27 BOP - 0.39 FScore (R^2^ = 0.29, F statistic = 6.62, p \< 0.0001). PL and BOP were individually significant. F Score and age were borderline, p = 0.09 and 0.14 respectively, and PD was not significant, being entirely replaced with age. Despite the subjects increasing to 102 (from 35 or 67) the strength of association was similar to that of low responders only. Association of DT with gingivitis in high and low responders ------------------------------------------------------------ Because caries experience associated with fluoride and gingival health in high responders, poor gingival health should increase the number of decayed teeth (DT) more than in low responders. When BOP was regressed against the variables in Table [5](#T5){ref-type="table"}, only DT (column 4) differed significantly between the high and low responders. Although DT alone significantly correlated with BOP in both response groups (p \< 0.01), it associated with BOP much more strongly in high responders (R^2^ = 0.57) than in low responders (R^2^ = 0.12). Fig. [3](#F3){ref-type="fig"} shows the respective correlations, and also the patient data (filled circles) skewed by few healthy subjects and the student data (unfilled circles) skewed by few moderate and no severely diseased subjects. Clearly, combining students and patients strengthened the respective associations (β coefficients). Stepwise regression indicated that, excluding DT, BOP associated with PL and PD similarly (BOP = -14.02 + 0.42 PL + 4.8 PD in high responders and BOP = -11.07 + 0.34 PL + 4.10 PD in low responders; R^2^ = 0.40 and 0.41 respectively; p \< 0.001). However, DT explained more variance (BOP = -1.46 + 0.24 PL + 0.95 DT, R^2^ = 0.62, p \< 0.001) in high responders, and less variance (BOP = -3.25 + 0.41 PL + 0.33 DT, R^2^ = 0.31, p \< 0.001) in low responders. {#F3} ::: {#T5 .table-wrap} ::: {.caption} ###### Changes in the equality of the partial β coefficients for association of tested variables withgingivitis (BOP) in high and/or low responders. ::: Variable^a^ High responders n = 54 Low responders n = 97 Difference (Hi -- Low) ------------- ------------------------ ----------------------- ------------------------ PL ^d^0.182 ^c^0.321 -0.140 PD ^c^2.529 ^c^3.588 -1.059 DT ^c^0.906 0.202 ^c^0.703 MFT 0.154 0.054 0.099 Age 0.022 -0.053 0.074 ^a^Variable names are defined under \"Clinical Measurements\" in the Methods ^b^n = number of subjects. ^c^p \< 0.05 for value of constant or variable (β). ^d^*p* \> 0.05 & \<0.15. If no subscript, p \> 0.15. ::: Influential points (outliers), whose presence might have affected the strength and significance of these complex regression analyses, were identified in five low responders and two high responders. When these subjects were omitted, the respective regression coefficients or their significance were little changed, indicating that the different, partial, linear regression coefficients in high or low responders were not artifacts of influential or outlying points. Discussion ========== This study has demonstrated that IgG antibodies to D-alanyl LTA are widespread in US adults. The fraction of high responders was essentially constant from early adulthood through middle age, but reduced in children (\<15 years) and old age (\>55 years). Within the adults (ages 15--55 years) a change from low to high response or *vice versa* was found unlikely from repeated measurements over 2--6 months. A similar lack of change in this IgG antibody concentration was reported 2--3 months after an additional 26 similarly aged patients had received oral hygiene therapy in another study \[[@B32]\]. Finally, the family studies established that a high antibody response was probably induced during the mid-teenage years. In order to apply the results of this study to children and young teenagers, longitudinal studies of the antibody response in relation to age and the clinical measurements in this study may need to be undertaken. Despite few investigations of caries risk in 22--38 year old subjects compared with a younger or older group \[[@B7]\], the association of DMFT with PL in this study agrees with that obtained from 35 year old Norwegians \[[@B3]\]. Plaque (simplified oral hygiene index measurement) accounted for 15% of the variance in number of decayed/filled teeth surfaces in that study, and for 19% of the DMFT variance within all 151 clinically examined subjects in this study (ignoring age, antibody and all other variables). Other studies have shown that the amount of fluoride applied from dentrifices is measured better from oral hygiene or plaque accumulation, as in this study, and not from the reported frequency of dentrifice use \[[@B2],[@B3]\]. Finally, because subjects aged more than 38 are unlikely to have used fluoridated toothpastes until later in life, they were omitted to avoid confounding the results. The rationale behind this study was that acidic plaque environments increase the amount of D-alanyl LTA and promote its immunogenicity. Accordingly, caries-protection by fluoride in the water supply and dentrifices was strong in high IgG antibody responders, accounting for just over 50% of the variance in DMFT. In addition, gingivitis (BOP prevalence) associated strongly and significantly with the number of decayed (untreated) teeth, suggesting that preventing gingivitis increased fluoride exposure from dentrifices and reduced the number of decayed teeth. Increased exposure to fluoridated water, and dentrifices associated with good gingival health, may result in fluoride inhibiting enamel remineralization at the acidic plaque pH likely present in high responders. In low responders, the fewer antibodies to D-alanyl LTA suggest less colonization by acid-tolerant bacteria and a weaker cariogenic attack. DMFT associated with age, as reported for other subjects whose sera did not precipitate D-alanyl LTA \[[@B22]\], and also with pocket depth. An increase in pocket depth is caused by periodontopathic bacteria that associate with an alkaline environment in the sulci over many years \[[@B20]\] and a microbiota that is neither acidic nor acid-tolerant \[[@B10]\]. The coefficient of DMFT association with PD in low responders therefore differed significantly from high responders within whom F score and gingival health were stronger covariates. Conclusions =========== An increased mutans streptococcal challenge accompanying low plaque pH (high antibody response to D-alanyl LTA) allows much of the variation in caries experience to be controlled by water fluoridation and by the use of fluoridated dentrifices associated with maintaining oral health. High IgG antibody responders are therefore better protected from caries in an optimally fluoridated environment. The concept that fluoride protects better from caries in a low pH environment \[[@B12]\] was recently used to explain why there is a poor association between caries experience and pH fall after a 10% sucrose rinse \[[@B33]\]. In low responders, increased fluoride exposure from dentriflce use to maintain oral health, or from water fluoridation, associate relatively poorly with caries experience. Although this study has indicated that the IgG antibodies to D-alanyl LTA do not become elevated until after age 17, when much caries may have already developed, they may be elevated to a lower level in children who eventually become high responders. The D-alanyl LTA antibody response is not detectable in saliva (unpublished studies), but it can be measured from only a thumb-prick of blood. Longitudinal studies of the D-alanyl LTA response in children could improve current efforts to predict caries susceptibility by relating it to fluoride or the fluoride ion product for fluoroapatite in saliva and the pH change after a sucrose rinse \[[@B5],[@B12],[@B13],[@B33]\]. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ This investigation was largely supported by USPHS Research Grant 1R01 DE-06740 from the National Institute of Dental Research, National Institutes of Health, Bethesda. We sincerely thank Dr. W. Bias, Immunogenetics Laboratory, Johns Hopkins University, Baltimore and Dr. F. Bach, Immunobiology Research Center, University of Minnesota Medical School, Minneapolis for donations of human serum from their patients; E. Carter and D. LeFlore for technical assistance; S. Pitts, J. Chowning and Dr. A. Cucchiara, Computing Center, University of Oklahoma Health Sciences Center, for computing assistance; Drs. R. Reynolds, M. Martin and L. Coggins Dept. of Oral Diagnosis for clinical assistance.

Background ========== In 1990 the National Institutes of Health Consensus Development Conference \[[@B1]\] concluded that breast conserving surgery (BCS) followed by radiation is an appropriate method of primary treatment for the majority of women with early stage breast cancer (AJCC stages \[[@B2]\] I, IIA, and IIB). Numerous clinical studies have shown that survival after BCS followed by radiation therapy is equivalent to survival following mastectomy for women in these stages \[[@B3]-[@B5]\]. Some absolute indicators for mastectomy remain \[[@B6],[@B7]\], in particular widespread malignant-type microcalcifications, previous radiotherapy, and a relation of tumor size to breast size that would not allow a cosmetically satisfactory result. Although there has been an increase in the use of BCS since the early 1990s, an apparent under-utilization of BCS among women for whom such treatment was not contraindicated has been documented \[[@B8]\]. Geographic location, type of hospital and health plan, and personal preferences have been investigated as possible factors to explain the continued use of mastectomy \[[@B9]-[@B12]\]. Comorbidity \[[@B13]\] has been found to be an important predictor among older women. The Surveillance, Epidemiology and End Results (SEER) cancer registries do not collect sufficient detail to examine some of these issues and medical records reviews are time consuming. Insurance claims data are a cost-effective alternative; they have already been collected and put into an electronic format by the insurance carriers. They cover large segments of the population, allow follow-up, use standardized codes \[[@B14],[@B15]\], and do not rely on subject recall. In cancer research, insurance claims data have been used to estimate the effectiveness of cervical cancer screening \[[@B16]\], to assess the role of screening practices in the incidence of prostate cancer \[[@B17]\], and to estimate mammography participation \[[@B18]\]. Medicare enrollees in Seattle and San Francisco who received care from a Health Maintenance Organization (HMO) were found to receive more BCS than women in Fee-For-Service (FFS) plans \[[@B19]\]. Whereas a Medicare/SEER registry linked data set has been used extensively to examine treatment and screening issues in individuals 65 years and older \[[@B20]\], linking of insurance claims data from private health plans in younger populations has been more difficult because of the large number of health plans in most geographic areas. However, Hawaii provides unique opportunities for insurance claims research because the majority of medical care is received within the state and more than 90% of the population \[[@B21]\] are covered by a limited number of health plans. As a pilot project, we were able to link data from the Hawaii Tumor Registry and from a health plan in Hawaii. The objectives of this analysis were to describe breast cancer treatment and comorbid conditions using the insurance claims data and to examine possible determinants of BCS *vs.* mastectomy for breast cancer patients with stage I and II disease. Methods ======= The study protocol was reviewed and approved by the University of Hawaii\'s Committee on Human Studies and by the Hawaii Tumor Commission that oversees the Hawaii Tumor Registry (HTR). Before linking of databases was initiated, a memorandum of agreement between all parties involved in this research project was signed. The resulting agreement safeguarded patients\' privacy at the highest possible level. All names and identifying information were deleted from the datasets and replaced with arbitrary numbers to be used for the data analyses. Data Sources ------------ The HTR has maintained a database of all cases of cancer diagnosed in the State of Hawaii since 1960 and became part of the SEER program in 1973. The HTR record contains demographic characteristics such as age, ethnicity, marital status, island of residence, as well as information on tumor size, extent of disease, lymph node involvement, and tumor grade. From medical records in hospitals and physicians\' offices, information is collected on the initial course (six months in the Hawaii Tumor Registry) of cancer-directed treatment following diagnosis. Quality control reviews have shown that case-ascertainment through HTR has been virtually complete \[[@B22]\]. Over 99% of cancer cases reported to the registry are histologically confirmed. The HTR also maintains a link with Hawaii Department of Health, which allows for death information to be captured in the HTR database. Linking ------- The insurance claims for this study were obtained from a local insurer who was a party to the memorandum of agreement. The linked data set contained cancer cases diagnosed from 1995 through 1998. During the linking process, a list of health plan members who had at least two cancer diagnostic codes in their claims history were matched against the HTR using a probabilistic method. For each matched record, the health plan furnished all claims data for that period. They also indicated whether the individual was enrolled in a Fee-For-Service (FFS) or a capitated (HMO) plan, a choice provided by the insurer. However, the majority of linked cancer cases belonged to the FFS plan and the HMO plan provided by this insurer differs considerably from a typical HMO. Individuals aged 65 years and older were included only if they were still working and had primary coverage through the health plan or if they were covered under Medicare but also had secondary coverage through the health plan. The dataset furnished by the health insurer contained all claims processed for cancer patients in this study. The data elements included: date of service, International Classification of Disease, version 9 (ICD-9) diagnosis codes \[[@B14]\], Current Procedural Terminology (CPT) codes \[[@B15]\], and the provider specialty code. The dataset contained claims for services from physicians, laboratories, freestanding facilities, as well as outpatient services provided by hospital facilities. Data set -------- Breast cancer cases accounted for 16.6% of all cancer cases recorded in the HTR during the period from 1995 to 1998. In the linked dataset, 27.7% of the cases were breast cancer cases. Using the ICD-9-CM code range \[[@B14]\] for breast cancer (174.0 to 174.9), we identified 1,377 female breast cancer cases. We then excluded 265 cases diagnosed after June 30, 1998 because we would not have a complete history of claims data covering at least six months of treatment. We also identified 30 women who had been recorded twice in the linked dataset because breast cancer was diagnosed in both breasts, as identified by the laterality codes. For 24 women, the diagnosis date was identical, in which case we considered the two diagnoses as one case. For the six cases with different diagnoses dates, we retained both records because there was a separate completed course of treatment for the cancer in each breast. We eliminated one case that was in the linked dataset and had inpatient data but no outpatient data, resulting in a dataset containing 1,088 breast cancer cases. Staging ------- We used the HTR information in the Extent of Disease (EOD-10) field \[[@B23]\] to determine the American Joint Committee on Cancer (AJCC) TNM stage \[[@B2]\], where T represents the primary tumor size, N refers to lymph node involvement, and M refers to presence of metastases. In cases where the tumor size was unknown, we assigned a stage only if the size was not a determinant in the TNM grouping. If the extension was unknown, we equated it to extension 10 (confinement to breast tissue and fat). For the small number of cases (N = 8) for whom lymph node involvement code was unknown, we considered this the same as no lymph node involvement because the tumor size was so small for these cases that a lymph node dissection was not considered necessary at the time of this study. The results of the TNM staging agreed well with the SEER summary stage codes. Cancer treatment ---------------- We used the CPT codes \[[@B15]\] in the claims dataset to identify the type of surgery performed. For mastectomies, we included codes for simple and subcutaneous mastectomies (19180 and 19182), radical mastectomies (19200 and 19220) and modified radical mastectomy (19240). BCS was identified by the CPT codes for partial mastectomies (19160, 19162). We also included codes for excision of breast cysts or lesions (19120 and 19125) as these codes meet the definition of a lumpectomy, although we recognize that some surgeons may have billed these codes for excisional biopsies as a diagnostic procedure. If BCS was initially performed, but followed by a subsequent mastectomy, the subject was classified in the mastectomy group. This situation would have occurred when the BCS code was actually used for a diagnostic biopsy and the subsequent surgical treatment was a mastectomy or in situations where a lumpectomy was first selected, but a mastectomy became necessary because pathological examination revealed cancerous cells in the margins of the excised tissue. CPT codes 77261 to 77799 were selected to determine whether radiation therapy was received. Several CPT codes and also a number of other codes used by the health plan for billing were considered evidence of chemotherapy treatment. Comorbidities ------------- Information on the existence of comorbidities was extracted from the claims data using the occurrence of ICD-9 codes associated with comorbid conditions included in the Charlson Index \[[@B24]\]. We also analyzed data for conditions that were not included in that index, but occurred at a high enough frequency to warrant examination as possible comorbidities, in particular hypertension and lipid disorders. We included the following twelve comorbid conditions into our index: diabetes with complications, diabetes w/o complications, hypertension, heart disease, cerebrovascular disease, chronic pulmonary disease, peripheral vascular disease, kidney disease, liver disease, rheumatological diseases, hypothyroidism, and lipid metabolism disorders. Comorbidities were based on physicians\' claims with the respective ICD-9 code \[[@B14]\] during the 12-months period preceding the month in which the cancer was diagnosed. The month of diagnosis was excluded to avoid identifying as comorbidities any complications or conditions directly resulting from cancer treatment. Laboratory, radiology, and other diagnostic services were excluded from the comorbidity identification process because tests may have been done to rule out the condition. Each subject was assigned a comorbidity score of 0 (no condition), 1 (one condition), or 2 (two or more conditions). Other variables --------------- In addition to stage at diagnosis and size of tumor, we examined several other variables that were possible determinants in the utilization of BCS *vs.* mastectomy in early stage breast cancer, including age at diagnosis, ethnicity, island of residence, and marital status. For ethnicity, we used the five major groups in Hawaii (Japanese, Caucasian, Hawaiian, Filipino, and Chinese). All other ethnicities were grouped into an \"Other\" category. The specific island of residence was coded in the HTR, but since close to 80% of the population resides on Oahu \[[@B25]\], we classified residence as either Oahu or non-Oahu. For marital status, we grouped all women who were identified as single, separate, divorced or widowed in the unmarried category. Women under age 50 were considered pre-menopausal and women 50 years and older were classified as postmenopausal. Tumor grade information from the HTR record was grouped into grade I (well-differentiated cells) *vs.* all other grades (II, III, IV and unknown). Statistical analysis -------------------- All analyses were performed with SAS version 8.00 for Windows (SAS Institute, Cary, NC). Simple Kappa statistics (κ) was calculated to validate the treatment information from the insurance claims and the HTR \[[@B26],[@B27]\]. BCS was used as reference group throughout the analysis. It was defined as the dependent (outcome) variable and coded as a dichotomous variable, with 1 indicating BCS received and 0 indicating mastectomy received. Age was entered in units of ten years and tumor size was grouped into units of ten millimeters (one centimeter). For analyses of ethnicity as a predictor of treatment selection, Caucasian was used as the control group and indicator variables for all other ethnic groups were created. Logistic regression \[[@B28]\] was used to explore the influence of each variable on the use of BCS *vs.* mastectomy. First, we considered each independent variable by itself in a model and then we entered all variables simultaneously in a logistic regression model. Odds Ratios (OR) \[[@B29]\] with corresponding 95% Confidence Intervals (CI) were calculated to measure the degree of influence of variables on the utilization of BCS *vs.* mastectomy. Results ======= Based on the TNM stages, we identified 722 women who had stage I, IIA, or IIB breast cancer (66.4% of the 1,088 breast cancer cases in the linked data set). These 722 cases represented 32.8% (722 of 2,203 cases) of all early stage breast cancer cases recorded in the HTR during the study period. Approximately two-thirds of the women in the study population (Table [1](#T1){ref-type="table"}) were diagnosed in stage I of the disease and only 24% and 12% in stage IIA and IIB, respectively. Approximately 30% of the study subjects were aged less than 50 years, close to half were 50 to 64 years old, and 21% were 65 years and older. Three out of four women resided on the island of Oahu. Overall, 52.8% of the women had received BCS. Of the cases diagnosed at stage I, 57% of the women had received BCS and 47% a mastectomy. This decreased to 50% for stage IIA and to 34% for stage IIB. Among women, 65 years and older, 56% had BCS, compared to 50% of those under 50 years of age and 53% of the 50--64 year old women. While 56% of women on Oahu underwent BCS, only 43% of the women residing on the outer islands received BCS. We observed no statistically significant differences between the BCS and mastectomy group in terms of ethnicity, comorbidity count, menopausal status, marital status, and insurance plan. We found very high agreement between HTR data and claims data in identifying BCS \[κ = 0.91 (95% CI 0.88, 0.94)\], only 32 women were misclassified. For the majority of women who received BCS (92.1%), the lumpectomy was followed by radiation therapy. Additional chemotherapy was given to 34.1% and 42.5% of the women who had a lumpectomy or a mastectomy, respectively. ::: {#T1 .table-wrap} ::: {.caption} ###### Determinants of Breast Conserving Surgery Among 722 Cases of Early Stage Breast Cancer ::: **Variable** **No. of patients** **Odds ratio (95% CI)** --------------------- ------------ --------------------- ------------------------- --------------------------- ----------------------- BCS Mastectomy Univariate analysis Combined analysis All 381 341 Ethnicity Caucasian 71 66 1 1 Japanese 170 147 1.08 (0.72--1.61) 0.85 (0.55--1.31) Hawaiian 48 40 1.12 (0.65--1.91) 1.24 (0.71--2.17) Filipino 35 47 0.69 (0.39--1.20) 0.65 (0.37--1.15) Chinese 32 25 1.19 (0.64--2.22) 0.97 (0.50--1.86) Other 24 16 1.45 (0.71--2.96) 1.33 (0.64--2.77) Residence Oahu 299 232 **1.71 (1.23--2.39)** **1.67 (1.17--2.38)** Other 82 109 1 1 Age at diagnosis \<50 yrs 110 109 Continuous (per 10 years) 50--64 yrs 186 166 **1.01 (1.001--1.028)** 1.01 (0.99--1.04) 65+ yrs 85 66 Menopausal status Pre Post 110 271 109 232 1 1.16 (0.84--1.59) 1 0.75 (0.46--1.23) Marital status Married 262 244 0.88 (0.64--1.21) 0.94 (0.68--1.32) Unmarried 119 97 1 1 Health plan FFS 330 300 0.89 (0.57--1.37) 0.87 (0.55--1.38) HMO 51 41 1 1 TNM stage I 266 200 1 1 IIA 86 85 0.76 (0.54--1.08) 0.99 (0.66--1.51) IIB 29 56 **0.39 (0.24--0.63)** 0.61 (0.32--1.19) Tumor size \<1 cm 113 83 Continuous (per 1 cm) 1--1.9 cm 177 139 **0.97 (0.96--0.99)** 0.98 (0.96--1.00) 2 + cm 91 119 Tumor grade Well diff. 77 49 **1.51 (1.02--2.24)** 1.25 (0.83--1.89) Other 304 292 1 1 Comorbid conditions 0 230 201 Continuous (per score) 1 109 102 0.97 (0.78--1.19) 0.91 (0.72--1.14) 2+ 42 38 ::: In the univariate models (Table [1](#T1){ref-type="table"}), we found that tumor size and grade, island of residence, age, and stage at diagnosis were predictors of breast cancer treatment. Women residing on Oahu were considerably more likely to have BCS than women living on all other islands in the state. For each one-centimeter increase in tumor size, there was a 3% lesser chance of undergoing BCS. Since size of tumor correlates to breast cancer stage, this also decreased the likelihood for women in stage IIA and IIB to receive BCS. Women with well-differentiated tumor grades were 50% more likely to undergo BCS as compared to women with all other grades. We found that for each ten-year increment in age, the chances of having BCS increased by 1%. We also observed that average tumor size was inversely related to age. Mean tumor sizes (in cm) with standard deviations were 1.82 ± 1.1, 1.59 ± 1.2, 1.39 ± 0.94 for women younger than 50 years, 50 to 64 years, and 65 years and older, respectively. Therefore, the smaller tumor sizes may account for the greater likelihood of BCS among older women. In a combined model with all independent variables in a logistic regression, island of residence remained the only significant predictor of BCS in this population. Women living on Oahu were 67% more likely to have BCS than women on the outer islands. Although all other variables lost their statistical significance, associations for age, TNM stage, tumor size, and tumor grade remained similar in magnitude as in the univariate models. Although none of the ethnicity variables was significant, it appeared that Filipino women were less likely to receive BCS than women from all other groups. Residence on outer islands did not explain this observation. Women with a TNM stage of IIB were still 40% less likely to receive BCS than women diagnosed at stage I, but the relation lost its statistical significance due to the small number of cases. Menopausal or marital status, type of health plan, or the number of comorbidities were not related to the type of surgery. Discussion ========== The place of residence at the time of diagnosis was the most important predictor for receiving BCS in this dataset of health plan members. Women with early stage breast cancer living on Oahu were 70% more likely to have BCS than women living on outer islands. Age at diagnosis and tumor size were also related to breast cancer treatment although they were not statistically significant in the combined model. Contrary to our expectation and a previous publication \[[@B13]\], the number of comorbid conditions did not effect treatment in this analysis. The importance of geographic location can be explained by the availability of radiation facilities on the different Hawaiian Islands. Whereas on Oahu treatment facilities can be reached within one hour, distances on the outer islands are much farther. Only the Island of Hawaii and Maui have a radiation facility, but they are hard to reach from many parts of the islands. Women residing on Kauai, Molokai, and Lanai have to fly to Honolulu daily to receive the course of treatment or they have to remain there temporarily. Because this may be not be economically feasible for some women, as well as physically and psychologically challenging, it is possible that many women choose to undergo a mastectomy instead. Alternatively, patients and physicians in rural areas may favor mastectomy because of differences in education or cultural attitudes toward medical advances. We do not have information to explain why a small proportion of women who underwent BCS (6.9%) did not receive radiation. However, compared to other reports \[[@B13],[@B30],[@B31]\], the proportion of women who received radiation is very high. An interview study with cancer patients \[[@B32]\] showed that some patients decide against their physicians\' recommendation for radiation because they fear the treatment or hold certain beliefs about cancer. For other patients, radiation may have been contraindicated due to comorbid conditions. In 1994, a national BCS rate of 42.6% was reported \[[@B33]\] with a rate of 46.7% for the Pacific region (Hawaii, California, Alaska, Washington, and Oregon). Our results for 1995 to 1998 indicate a slightly higher rate for Hawaii (52.8%). Geographic differences in BCS use have repeatedly been described in the literature. In western Washington \[[@B34]\], women residing outside the Seattle area were less likely to have BCS than women residing inside the county, in particular if radiation therapy facilities were not available in their county. Comparing rates across the country showed that BCS was much more common in the Northeast than in the South \[[@B33]\]. Low BCS rates were also found in rural areas, such as North and South Carolina \[[@B10]\], Minnesota \[[@B35]\], and the Southwest \[[@B36]\]. An analysis of 1991--1992 SEER data \[[@B37]\] demonstrated a strong association between BCS and distance to the next radiotherapy facility. The likelihood of undergoing BCS was 50% lower among women living more than 15 miles from the next radiotherapy facility. In contrast to our results, a number of studies have shown that the likelihood of BCS decreases with increasing age \[[@B11],[@B30],[@B33]\], possibly due to less concern about cosmetic results. Because of the relatively small proportion (21%) of women 65 years and older and the fact that the women 65 years and older in our study had a health plan other than Medicare, they were not directly comparable to the populations in the published literature. Therefore, our ability to observe an effect among older women was limited. In particular, we were not able to examine the treatment of women 80 years and older who have been described as receiving more BCS than women age 65 to 79 years \[[@B13]\]. Our findings of lower BCS in smaller and well differentiated tumors agree with previous studies \[[@B33]\], but the absence of an effect of comorbid conditions disagrees with a study in older women \[[@B13]\] that showed a higher rate of BCS in women with comorbidities. Type of health plan did not predict type of treatment in our study probably because the majority of women in this data set had a FFS plan. In previous analyses with a larger variety of health plans, HMO members were considerably more likely to receive BCS in San Francisco and in Seattle \[[@B19]\]. Medicare and Medicaid patients received more BCS in a North Carolina hospital \[[@B7]\]. However, a study using administrative data \[[@B38]\] described lower BCS use among Medicaid patients and treatment was similar by health plan \[[@B39]\] in Northern California. Our study had several limitations, including the possibility of incomplete claims histories if patients changed health plans during the course of treatment, but the high agreement with HTR information indicates that the validity of insurance claims information was high. Continuing efforts by the health plan through feedback loops has improved diagnostic coding \[[@B40]\] in Hawaii, but as documented for Medicare claims, problems remain \[[@B41]\]. As discussed above, our study sample of breast cancer patients does not represent the population of the state, excluding in particular Medicare and Medicaid recipients, members of a typical HMO, and the relatively small number of uninsured patients. The exclusion of Medicare claims lead to a younger population, whereas the lack of Medicaid claims biased the cases toward a higher socioeconomic status. On the other hand, the study population represented the population quite well in terms of ethnicity (Hawaii Department of Business 1998 1705 /id). Because our data set included 42.4% of all breast cancer cases under 65 years diagnosed during the study period, the results represents this population much better than the population of older women. Also, we were unable to measure some factors that have been found to be important in other reports, in particular, influence of physician age \[[@B11]\], physician specialty \[[@B42]\], type of hospital (teaching hospitals and hospitals with radiation facilities, private, county and public) \[[@B36],[@B38],[@B39],[@B43]\], socio-economic status and education \[[@B42]\], tumor-breast ratio and the expected cosmetic results \[[@B7],[@B9]\], physician/patient interactions before surgery \[[@B42]\], and psychological factors, such as fear of radiation or cancer \[[@B9]\]. The strengths of using of insurance claims data for this analysis were twofold. First, we had more detailed treatment information available than collected by the tumor registry. Second, the diagnostic codes in the claims data allowed us to estimate the number of comorbid conditions before the cancer diagnosis. The results of this pilot study combining tumor registry and insurance claims data raises an important issue for cancer practice in the State of Hawaii and elsewhere. Health care providers and insurance plans need to work with other agencies to develop viable solutions to facilitate access to radiation facilities for women residing in remote locations, such as islands with no radiation facility. Attitudes of patients and physicians living in rural areas may be important in the choice of treatment, but could not be investigated in this project. Future research that includes interviews with patients and physicians may investigate attitudes toward certain types of treatment and identify additional barriers to BCS, such as psychological problems. In order to understand how treatment decisions were made in practice, qualitative information from physicians and breast cancer patients should be collected and examined in detail. Further validation of treatment and comorbidity from insurance claims data would strengthen this type of research in the future. The inclusion of a larger proportion of breast cancer cases is needed to establish the validity of our findings for the population at-large. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We are very grateful to the employees of the health plan for their time and willingness to assist with this project. Special thanks go to Andrew White, PhD for his long-term support of our research efforts. The help of Marc Goodman, PhD and the staff of the Hawaii Tumor Registry is greatly appreciated. This research was funded by a special study grant from the National Cancer Institute, Surveillance, Epidemiology, and End Results program under contract number N01-PC67001.

Background ========== Health information systems (HIS) provide a scientific and technological framework to gather, manage, and interpret data to inform the public, policymakers, administrators, and health-care workers about the distribution and determinants of health conditions. Further, they can (and should) guide and measure the impact of interventions \[[@B1]\]. Public health surveillance -- a subset of HIS -- has been defined as *the ongoing, systematic collection, analysis, and interpretation of outcome-specific data for use in the planning, implementation, and evaluation of public health practice*\[[@B2]\]. Public health surveillance can be used to 1) assess the overall health status of a population, 2) describe the natural history of disease, 3) monitor disease trends, 4) detect epidemics, 5) evaluate the effect of prevention and control measures, 6) generate hypotheses, and 7) facilitate epidemiologic and laboratory research \[[@B3]\]. Before 1991, the Soviet Union centrally planned the infectious diseases surveillance systems (IDSS) of its 15 republics. Approximately 300 million persons were covered under the IDSS. Central monetary and technical support for the IDSS ended in 1991. As a result, the republics have struggled to maintain their respective IDSS. The former Soviet Union (FSU)-wide diphtheria outbreak in the 1990s \[[@B4]\] and the re-emergence of malaria in Tadjikistan in 1991 \[[@B5]\] and in Armenia \[[@B6]\] and Azerbaijan \[[@B7]\] in 1994 indicated that financial constraints resulting in the disruption of public health infrastructure and services had increased the risk of the re-emergence of infectious diseases \[[@B8]\]. After 1991, this transition to nationalism, privatization, and social reorganization posed new challenges to each republic of the FSU. The loss of centralized training, public health expertise, and resources especially impacted the surveillance systems in each republic. \[[@B9]\] Further, privatization of the FSU medical systems resulted in under-budgeted public health services and inadequately paid personnel \[[@B10]\]. Combined with these infrastructure problems, the increased population migration -- both within the FSU and internationally -- contributed to morbidity and mortality through population dislocation \[[@B8]\]. This has increased the risk of exposure to re-emerging microbial and environmental pathogens and limited access to health services and good nutrition \[[@B9],[@B10]\]. The most important of these changes was the financial crises resulting from the severance of economic ties among all republics of the FSU. Public health officials were challenged to transform the bulky, state-sponsored IDSS. With high inflation and unemployment, they also suffered from shortages of vaccines, hospital supplies, and essential drugs. Provision of basic public health services were compromised, including repairing antiquated water and sewer systems, resulting in increased risk for gastroenteritis and infections with hepatitis A \[[@B11]\]; the largest documented outbreak of typhoid in this century occurred in Tajikistan in February 1997 \[[@B12]\]. Practices such as reusing syringes during vaccination and poor sterilization procedures during dentistry have contributed to nosocomial outbreaks of HIV and a high prevalence of infections with hepatitis B \[[@B13]\]. The objective of this work was to assess the current status and functioning of various IDSS, so as to guide reform efforts. At the invitation of seven Ministries of Health (MoHs), we performed assessments in the Russian Federation and in the Republics of Kazakhstan, Tadjikistan, Uzbekistan, Turkmenistan, and Armenia, and in the Kyrgyz Republic. We found striking homogeneity in comparing the IDSS from one republic to another; and, for clarity, we present here representative findings by using the Armenian component of the IDSS (AIDSS) as a prototype. Discussion ========== We used the CDC guidelines to assess the seven IDSS \[[@B14]\]; these guidelines have recently been republished in revised form \[[@B15]\]. This strategy includes an assessment of public health importance, objectives and usefulness, operation of the system, cost, and the seven system attributes (i.e., simplicity, flexibility, acceptability, sensitivity, predictive value positive, representativeness, and timeliness). The assessment involves gathering both qualitative (i.e., simplicity, flexibility, and acceptability) and observations of the quantitative (i.e., sensitivity, predictive value positive, representativeness, and timeliness) attributes. In Armenia, we attempted to gather as much information as possible with respect to the construct and utility of the AIDSS from those most integral to its functioning and application. Therefore, we conducted face-to-face interviews and focus group discussions with approximately 50 epidemiologists at the Ministry of Health, the National Sanitary Epidemiologic Service, the Institute of Epidemiology, and two regional, three districts, and one city Sanitary Epidemiologic Service (SES) office. We also interviewed 23 health-care workers at village health centers, polyclinics, hospitals, and laboratories, which serve as the primary reporting units for the AIDSS. At the time of this assessment, the AIDSS was moving toward reform, and we chose not to use the limited resources to gather quantitative data (e.g., through chart reviews) to assess the quantitative attributes of the AIDSS (i.e., sensitivity, predictive value positive, representativeness, and timeliness). Rather, we relied on qualitative observations gleaned in the course of the interviews. We present here the qualitative observations made of the AIDSS and recommendations for reform. We believe these observations and recommendations reflect the status of the IDSS of the other FSU republics. Description of the AIDSS ------------------------ A republic of 3.3 million, Armenia gained its independence from the FSU on September 21, 1991 (Figure [1](#F1){ref-type="fig"}). The country is divided into 11 regions, which are subdivided into 37 districts (Figure [2](#F2){ref-type="fig"}). Responsibility for the AIDSS rests with the Department of Hygiene and Epidemiologic Surveillance -- a department of the Armenian MoH (AMoH) -- composed of 52 functional units known as the Sanitary Epidemiologic Service (SES). The units or stations of the SES parallel the geopolitical divisions of Armenia. There exists one SES station in each district. Each city has one SES station, except Yerevan, the capital. Yerevan has eight districts, each with one SES station. The national SES is also located in Yerevan. {#F1} {#F2} The SES, *per se,* developed from a model created in the late 1800s in russia and uniformly developed over many years in the republics of the FSU. Its principle functions are to collect and analyze public health surveillance data and to implement and enforce strategies for the prevention and control of infectious diseases. Traditionally, the SES had approximately 10% of the entire medical person-power and budget of the AMoH, and was separate from the curative medical care system \[[@B16],[@B17]\]. The SES was staffed by epidemiologists (physicians), microbiologists, sanitary hygienists, and other health workers (paramedics and physician assistants). The district SES was the basic public health unit that monitored infectious diseases, investigated outbreaks, attended to child and adolescent health, inspected the food-service industry, monitored water purity, and dealt with occupational and environmental health problems throughout Armenia \[[@B6]\]. City-, regional-, and national-level SES administrations were larger, with specialized staff. The SES collected infectious diseases data from all health-care facilities throughout the country. Before 1991, Armenia had a comprehensive and free health-care delivery system accessible to all citizens with health facilities and health-care workers employed under the auspices of the AMoH. Outpatient facilities (village health centers and polyclinics) and hospitals reported to the AIDSS. Each district had 12--45 village health centers, two polyclinics, and one hospital. The seven cities in Armenia had variable numbers of polyclinics and hospitals. Altogether, Armenia had 830 village health centers, 228 polyclinics, and 179 hospitals (Figure [3](#F3){ref-type="fig"}). Data from these health facilities were reported to the 52 districts and city SES and then to the national SES, which forwarded aggregated data to the Department of Hygiene and Epidemiologic Surveillance in the AMoH. {#F3} Objectives and Data Collection, Reporting, Analyses, and Response ----------------------------------------------------------------- The objectives of the AIDSS were to identify cases of infectious diseases, document outbreaks, and monitor trends in disease occurrence. It collected aggregated and case-based data on new cases of 64 infectious diseases. This list included some diseases with low pathogenicity (e.g., pediculosis and scabies) and some diseases with inadequate or non-existent preventive measures (e.g., infectious mononucleosis and parapertussis). There existed no tiered (e.g., confirmed, probable, and suspected) standardized case definitions. The reporting of confirmed or suspected cases of some infectious diseases required immediate reporting via telephone or in person within 12--24 hours. These included epidemic prone diseases (e.g., diphtheria, polio, plague). Data were provided monthly and yearly from the district to national SES. Reports from the national SES were sent monthly to the AMoH and the other 14 republics of the FSU and back to the district SES twice a year. The district and city SES reported immediately to the national SES. At all levels, epidemiologists used descriptive statistics for data analyses. These included the calculation of aggregated case numbers and incidence and prevalence rates based on estimates of the population size provided by the state agency charged with gathering census data; limited stratification by person, time, and place; and the assessment of trends. Epidemiologists did not use analytic methods to assess risk factors for diseases, even though they were collected. Cases and contacts of every disease were investigated by a district or city SES epidemiologist within 24 hours of the receipt of the case report, using a standardized form known as the *epid carta.* This form was not disease specific, yet lengthy (46 questions, many of which required subjective responses). Little or no feedback was provided to the original reporting sources and no routine or formal sharing of data and information occurred between the district SES and health-care facilities. Table [1](#T1){ref-type="table"} summarizes the public health practice activities in Armenia, stratified by health facility. ::: {#T1 .table-wrap} ::: {.caption} ###### Public health surveillance and action core and support activities, by health level, Republic of Armenia, 1996 ::: **Organization Level** **Public Health Surveillance and Action Core and Support Activity** --------------------------- --------------------------------------------------------------------- --- --- --- --- --- --- --- --- ----- --- Primary Health Facilities X X X X X X X X *X* X District SES X X X X X X X District Lab X X X X X X X X District Lab X X X X X X X Regional SES X X X X X X X Regional Lab X X X X X National SES X X X X National Lab X X X X ::: Qualitative Attributes ---------------------- ### Simplicity Referring to both its structure and ease of operation, the AIDSS was complex. Epidemiologists gathered voluminous information on each case; parallel vertical public health programs reported duplicative information; much time was required to collect, register, and report case information; and many staff maintained the system. ### Flexibility and Acceptability Addressing the extent to which the AIDSS could adapt to changing information needs or operating conditions and reflecting the willingness of participants to provide information and monitor the system, the AMoH, itself, determined both its flexibility and acceptability. Because the AMoH provided salaries for health-care personnel, it could enforce compulsory reporting, through monetary fines and professional demotions. Changes in reportable conditions or criteria were made rapidly because administrative SES employees carried out orders quickly and completely at all levels in the system. Quantitative Attributes ----------------------- While we did not quantitatively measure the attributes of sensitivity, predictive value positive, representativeness, and timeliness, we did make qualitative observations. ### Sensitivity Assessing the AIDSS\'s ability to account for all incidents of a disease (i.e., the proportion of cases detected, correctly diagnosed, and reported), we learned that detection of most reportable infectious conditions did come to medical attention. This was enhanced because all citizens received free health care and because primary care physicians were responsible for care (including community outreach) of persons in their assigned territories. However, due to constrained resources, laboratory confirmation of reportable infectious diseases was limited in practice. As such, reporting was based on clinical or epidemiologic, rather than laboratory information. ### Predictive Value Positive The likelihood that a disease report constituted a true case of that disease was diminished because of the lack of laboratory confirmation and standardized case definitions. ### Representativeness We felt that, ideally, the AIDSS accurately described the distribution of diseases in the population by person, time, and place because disease reporting was mandatory and failure to report was a punishable offense. And, because of the penalties, all official reporting to the national level did occur on a monthly basis. However, it was common practice for epidemiologists to conceal cases of infectious disease and willfully underreport epidemic morbidity, because outbreaks meant that the epidemiologists were not performing their duties of preventing and controlling infectious diseases. This paradox resulted in epidemiologists managing two sets of information: one officially reported to higher-ups and one unofficially kept (with more accurate numerators). ### Timeliness Information for action or for long-term planning was available because mandatory reporting to the SES within 12--24 hours of diagnosis for most conditions under surveillance allowed rapid implementation of control and prevention measures. ### Costs Because budgets were relatively non-existent in the FSU, historical data on the costs of operating the AIDSS were not available. In the FSU centrally planned economy, resources were obtained from cost centers (e.g., utilities were not metered, office supplies were requested by quantity and not cost, and salaries were provided from the central budget). However, numerous observations led us to conclude that the AIDSS was relatively inefficient and costly; in large part, because the expenditure was paid from the public sector. The AIDSS was labor intensive. Being paper-driven, reporting dieases for which no practical public health interventions exist misallocated scarce resources. Other common public health activities with high opportunity cost in the AIDSS were indiscriminate disinfection of homes and work sites. Environmental background monitoring practices by district and regional SES included routine collection of specimens and laboratory testing (e.g., air and water samples, food products, and items that children may come in contact with such as toys or eating utensils) in addition to evaluation of physical factors (noise, vibration, microclimate, electromagnetic fields, levels of lighting, and ionic radiation) at several sites (e.g., work places and day-care centers). When cases of hepatitis A, acute gastroenteritis, tuberculosis, diphtheria, or pediculosis were reported, disinfection of homes, schools, day-care centers and work places was conducted by public health workers who used chloramine application and steam cleaning of all hard surfaces and laundering of all clothing and bedding materials. The effectiveness of such disinfection practices or environmental background monitoring has not been documented, and is likely of doubtful public health utility. During disease outbreaks, it was common practice to investigate every case and culture all available materials, and decontamination efforts were instituted regardless of epidemiologic evidence. It was common practice to hospitalize children \< 1 year of age with pneumonia for 7--14 days, and children of all ages with acute gastroenteritis for 7--15 days. It was also common to hospitalize both adults and children with hepatitis A for 21 days, with syphilis for two weeks, with gonorrhea for three weeks, and with tuberculosis for one year. These isolation practices, meant to prevent disease transmission to the community, were consequences of central planning in which emphasis was placed on input indicators such as the occupancy rates of hospitals. Incentives (budget allocations) placed on input rather than output measures led to a level of medical infrastructure that has been difficult to maintain given current levels of funding available for the health sector. Because financial issues were a major driving factor, cost analyses of surveillance practices and control measures could identify areas for cost-savings. Analyses of the surveillance system in Ukraine (using 1996 budget figures) revealed that excessive culturing represented 47% of the cost *per capita* expenditure of the L\'viv Regional SES and disinfection procedures accounted for almost 30% of the entire Pustomity District SES\'s budget (V. Carande-Kulis, CDC, personal communication). Recommendations --------------- Based on this assessment, we developed recommendations with respect to the three main surveillance functions of data collection, analysis and interpretation, and retrospective and prospective responses. ### Data Collection • Eliminate punitive consequences to obtain accurate reporting; • Restrict the number of routinely reportable diseases based on measures of mortality, morbidity, severity, communicability, and preventability; • Categorize events under surveillance into a three-tiered surveillance system: ○ disease elimination (e.g., polio); ○ case-based (e.g., diphtheria); and ○ indicator-based (e.g., number of children immunized by two years of age); • Simplify reporting procedures and forms by ○ limiting urgent reporting of diseases to those that require prompt institution of control measures; ○ requiring only information necessary to direct control measures and perform basic analyses; and ○ developing disease-specific forms with diseases chosen for case-based surveillance; • Develop tiered (confirmed, probable, and suspected) standardized case definitions for all events under surveillance; and • Computerize demographic and risk-factor data for systematic and detailed analysis of reported diseases and rapid dissemination of information. ### Analysis and Interpretation • Provide ongoing capacity for training in analytic epidemiology; and • Base interventions on epidemiologic evidence. Use analytic epidemiology (case-control and cohort studies and presentation of data using 2 × 2 tables, odds ratios, relative risks, and tests of significance) for hypothesis generation, risk factor identification, outbreak investigations and intervention design and monitoring. ### Retrospective and Prospective Responses • Provide feedback to all reporting sources and share information across vertical program lines and with officials throughout the public health community in a timely fashion (e.g., via a monthly public health bulletin). A bulletin could include descriptions of important outbreak investigations, disease-specific analyses of surveillance data, graphic and tabular information on selected diseases, indicators of community health, and recommendations for public health concerns. Current HIS Reform Efforts in Armenia ------------------------------------- This assessment stimulated and guided reform efforts that were initiated in December 1992 through a cooperative project among the AMoH, United States Agency for International Development (USAID), and CDC. This project provided technical and material assistance toward reform of the Armenian HIS. The approach focused on training a cohort of public health officials and epidemiologists in the modern aspects of epidemiology, biostatistics, surveillance techniques, and scientific communications; developing Armenia-specific case definitions; facilitating HIS reform strategies through workshops and training sessions; and developing the capacity to publish an epidemiologic bulletin. Since 1996, this HIS reform activity has been self-sustained with no additional monetary support from USAID \[[@B6]\]. In 1996, the AMoH created a national HIS program for the development and reform of the HIS and the Armenian National Health Analytic-Information Centre \[[@B18]\]. The system has been transformed into a comprehensive HIS and includes chronic diseases, maternal and child health, and injury data. Diseases are now categorized by a three-tiered approach: disease elimination (e.g., polio), case-based (e.g., diphtheria) and indicator-based (e.g., number of children immunized by two years of age). Preparations are now being made for additional training and to assess and improve clinic case diagnoses, management, and recording, and clinic records. New regional centers equipped with computers and faxes have been organized for the collection, analyses, and reporting of health information. National and regional public health bulletins are being published monthly in three languages - Armenian, Russian and English -- and distributed to wide audiences. Tiered, standardized case definitions and essential health indicators for decision-making at the clinic and community level have been developed and disseminated. These include health status, performance, and resource indicators. Comprehensive HIS reform is critical throughout the FSU. Timely, accurate, and relevant health information are necessary to assess the burden of disease and disability; understand changing health patterns; measure the needs for and improve services; address inequities in health; provide information for policy formulation and planning; and provide a basis for intra- and international comparisons on health status and care utilization \[[@B1]\]. Timeliness, accuracy, and relevancy are augmented by efficiency. Integrating all sources of data into one comprehensive HIS prevents duplicate recording and reporting across services and programs, averts labor-wasting inefficiencies, and saves scarce resources. A comprehensive HIS includes the capacity to obtain data from vital registries, clinical, administrative, and other records; from provider and population-based surveys and sentinel systems for infectious and chronic diseases, and disabilities; and maternal and child health, nutrition, and program implementation indicators, including access, coverage, and service quality \[[@B1]\]. This reform activity should include the development of an indicator monitoring system based on selected essential, action-oriented indicators of health status, service performance, and resources that can be used for decision-making at the local level. Summary ======= We found the AIDSS to be a complex and sensitive, yet costly and inefficient surveillance system for infectious diseases. Despite the lack of standardized case definitions, feedback of information, and computer technology, it functioned fairly well before 1991. However, the functioning and continuation of the AIDSS has been affected both directly and indirectly by events of the past decade. Overall, the former AIDSS was useful because it detected cases of infectious diseases, estimated morbidity, monitored trends in disease occurrence, and documented outbreaks. The comprehensive no-cost health-care delivery system and compulsory reporting of diseases to the AMoH enhanced its flexibility, representativeness, and timeliness. The strengths of the AIDSS stemmed from the large numbers of health facilities and trained personnel and the separation of preventive from curative medicine, that secured an independent status (including separate budget) for preventive medicine and public health practitioners. The AIDSS also had weaknesses. Though the system meticulously tracked persons (from birth to death), very few of these data were computerized, analyzed, or used to develop, direct, or evaluate public health policy. In most cases, when used, data guided regulation and punishment rather than public health decision-making. Simply put, data were used to fix blame and punish rather than to find and implement effective interventions. Epidemiologists were motivated to perform actions that both pleased their supervisors and avoided the punishment of monetary loss or demotion. For example, it was common practice for epidemiologists to hide select cases of infectious disease and willfully underreport epidemic morbidity because outbreaks meant that the epidemiologists were not performing their duties of preventing and controlling infectious diseases. These disincentives to thoroughness and honesty resulted in surveillance data and reports that did not reflect true incidence and prevalence of disease, circumstances, needs, responses, or impacts \[[@B9]\]. Because the AIDSS was but a component of the systems designed for the entire FSU, the AIDSS did not address Armenia-specific needs. Further, because of the lack of tiered, standardized case definitions, there existed a potential for misclassification of diseases. Lack of feedback to reporting sources hampered improvements in clinical practice. Monitoring conditions for which there were no practical public health interventions, the multi-tiered and duplicative reporting processes, and the use of expensive and indirect monitoring and control measures such as excessive culturing, disinfection, and prolonged hospitalization led to waste of resources. As a result, the system did not guide control measures optimally nor use resources efficiently. As the republics of the FSU embrace various aspects of democratization, improvement of public health surveillance systems such as the IDSS should be a goal if decision makers are to use credible data for informed public health practice. Competing interests =================== We certify that we have participated sufficiently in the conception and design of this work, as well as its execution and the analyses of the data. Further, we have collaboratively written the manuscript and take public responsibility for it. We believe the manuscript represents valid work. We have reviewed the final version of the submitted manuscript and approve it for publication. Neither this manuscript nor one with substantially similar content under our authorship has been published or is being considered for publication elsewhere. If requested, we shall produce the data upon which the manuscript is based for examination by the editors. We certify that we have no affiliations with or involvement in any organization or entity with a direct financial interest in the subject matter or materials discussed in the manuscript. Drs. Wuhib, Chorba, MacKenzie, and McNabb were employees of the U.S. federal government when this work was performed and prepared for publication; therefore, it is not protected by the Copyright Act, and there is no copyright of which the ownership can be transferred. Dr. McNabb serves as corresponding author; his address is listed. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We acknowledge all SES epidemiologists in the respective Ministries of Health who shared their knowledge of the IDSS and contributed to the discussions on recommendations, and individuals from CDC including Drs. Daniel Bleed, Robin Ikeda, Lyle Conrad, Scott Wetterhall, Siiri Bennett, Gulbanu Altynbaeva and Scott Deitchman and Mr. Bruce Ross.

Background ========== The manner in which accuracy of clinical tests is mathematically summarised in the biomedical literature has important implications for clinicians. Appropriate accuracy measures would be expected to sensibly convey the meaning of the study results with scientifically robust statistics without exaggerating or underestimating the clinical significance of the findings. Lack of use of appropriate measures may lead authors of primary accuracy studies to draw biased conclusions.\[[@B1]\] In systematic reviews of test accuracy literature, there are many ways of synthesising results from several studies, not all of which are considered to be scientifically robust. For example, measures such as sensitivity and specificity commonly used in primary studies are not considered suitable for pooling separately in meta-analysis.\[[@B2]\] Variations in reporting of summary accuracy and use of inappropriate summary statistics may increase the risk of misinterpretation of clinical value of tests. A recent study evaluated a small sample of meta-analytical reviews of screening tests to demonstrate the variety of approaches used to quantitatively summarise accuracy results.\[[@B3]\] This study confined itself to a limited Medline search. It exclusively examined meta-analytical studies so reviews not using quantitative synthesis were excluded. It did not look at accuracy measures used to report results of primary studies separately from those used for meta-analyses. In order to address these issues, we undertook a comprehensive search to survey systematic reviews (with and without meta-analysis) of test accuracy literature to assess the measures used for reporting results of included primary studies as well as their quantitative synthesis. Methods ======= We manually searched for relevant reviews in the *Database of Abstracts of Reviews of Effectiveness* (DARE).\[[@B4]\] In order to limit the impact of human error inherent in manual searching, we complemented it with electronic searching. DARE was searched electronically with word variants of relevant terms (diagnostic, screening, test, likelihood ratio, sensitivity, specificity, positive and negative predictive value) combined using OR. From 1994 to 2000 DARE\[[@B4]\] has identified 1897 reviews of different types by regular electronic searching of several bibliographic databases, hand searching of key major medical journals, and by scanning grey literature (search strategy and selection criteria can be found at <>). The structured abstracts of these reviews were screened independently by the authors to identity systematic reviews of test accuracy. The full texts were obtained of those abstracts judged to be potentially relevant. Reviews addressing test development and diagnostic effectiveness or cost effectiveness were excluded. Any disagreements about review selection were resolved by consensus. Information from each of the selected reviews was extracted for the measures of test accuracy used to report the results of the primary studies included in the review. If a meta-analysis was conducted, information was also extracted for the summary accuracy measures. The various accuracy measures are shown in Table [1](#T1){ref-type="table"}. We sought the following in the primary studies: sensitivity or specificity, predictive values, likelihood ratios and diagnostic odds ratio. For meta-analysis, we sought the summary measures pooling the above results and summary receiver operating characteristics (ROC) plot or values. All extracted data were double-checked. We divided the reviews into two groups arbitrarily according to time of publication; one group covering the period 1994--97 (50 reviews) and another covering 1998--2000 (40 reviews). This allowed us to assess whether there were any significant differences in measures being used to report test accuracy results among reviews published earlier and those published later. As the approaches to summarising results are not mutually exclusive, we evaluated and reported the most commonly used measures and their most common combinations. We used chi-squared statistical test for comparison of differences between proportions. ::: {#T1 .table-wrap} ::: {.caption} ###### Measures of accuracy of dichotomous test results ::: ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ **Measures for primary studies** **Sensitivity (true positive rate)** The proportion of people with disease who are correctly identified as such. **Specificity (true negative rate)** The proportion of people with disease who are correctly identified as such. **Positive predictive value** The proportions of test positive people who truly have disease. **Negative predictive value** The proportions of test negative people who truly do not have disease. **Likelihood ratios (LR)** The ratio of the probability of a positive (or negative) test result in the patients with disease to the probability of the same test result in the patients without the disease. **Diagnostic odds ratio** The ratio of the odds of a positive test result in patients with disease compared to the odds of the same test result in patients without disease. **Measures for meta-analysis** **Summary sensitivity, Specificity, predictive values, likelihood ratios, and diagnostic odds ratio** Pooling of the above accuracy measures obtained from multiple primary studies (usually averaged and weighted according to size of individual studies). **Summary receiver operating characteristics curve (ROC)** A method of summarising the performance of a test as found in multiple primary studies, which takes into account the relationship between sensitivity and specificity. ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ ::: Results ======= Of the abstracts available in DARE, 150 were considered to be potentially relevant. Excluding reviews that addressed test development and diagnostic effectiveness or cost, 90 reviews of test accuracy were left for inclusion in our survey. There were 45 reviews of dichotomous test results, 42 reviews of continuous results dichotomised by the original authors, and 3 reviews that contained both result types. Meta-analysis was used in 60/90 (67 %) reviews, 50 in 1994--97 and 40 in 1998--2000. (See [Additional File](#S1){ref-type="supplementary-material"}: BMC_IncludedRefList_04032002 for a complete listing of the 90 reviews included in our study). As shown in Table [2](#T2){ref-type="table"}, sensitivity or specificity was used for reporting the results of primary studies in 65/90 (72%) reviews, predictive values in 26/90 (28%), and likelihood ratios in 20/90 (22%). For meta-analysis, independently pooled sensitivity or specificity was used in 35/60 (58%) reviews, pooled predictive values in 11/60 (18%), pooled likelihood ratios in 13/60 (22%), and pooled diagnostic odds ratio in 5/60 (8%). Summary ROC was used in 44/60 (73%) of the meta-analyses. There were no significant differences between reviews published earlier and those published later as shown in Table [2](#T2){ref-type="table"}. ::: {#T2 .table-wrap} ::: {.caption} ###### Measures of test accuracy reported in review of diagnostic literature (1994--2000) ::: ------------------------------------------------- ------------------ -------------------------------------- ------------ ------------------------------------- ------------------- **Time periods** **Measures of test accuracy** **1994--1997** **1998--2000** ***n*** ***% (95% confidence interval)*^+^** ***n*** ***% (95% confidence interval*^+^** **p-value^\*\*^** Included primary studies **50^\*^** **40^\*^** Sensitivity or specificity 35 70 (55--82) 30^a^ 75 (59--87) 0.77 Predictive values 13^b^ 26 (15--40) 13^c^ 33 (19--49) 0.66 Sensitivity, specificity and predictive values 12 24 (13--38) 12 30 (17--47) 0.69 Likelihood ratios 9 18(9--31) 11 28 (15--44) 0.41 Diagnostic odds ratios 0 0 (0--7) 0 (0--9) \- **Meta-analysis** **38^\*^** 76(62--87) **22^\*^** 55 (38--71) 0.16 Independently pooled sensitivity or specificity 22 58 (41--74) 13^d^ 62 (36--79) 0.86 Pooled predictive values 9^e^ 24(11--40) 2^f^ 10(1--29) 0.29 Pooled likelihood ratios 5 13 (4--28) 8 38 (17--59) 0.08 Pooled diagnostic odds ratios 5 13 (4--28) 0 0(0--15) 0.20 Summary ROC plot or values 23 61 (43--76) 11 52 (28--72) 0.60 ------------------------------------------------- ------------------ -------------------------------------- ------------ ------------------------------------- ------------------- ^\*^ numbers do not add up to totals because some reviews used more than one measures of accuracy; ^\*\*^ chi sq. test with Yates\' correction; ^+^ Exact (Clopper-Pearson) 95% confidence interval a includes study that only reported either sensitivity or specificity; b,c. includes study that only reported either positive or negative predictive value; d includes meta-analyses that only reported either pooled sensitivity or specificity; e,f. includes meta-analyses that only reported either pooled positive or negative predictive values ::: Discussion ========== Our study showed that sensitivity and specificity remain in frequent use, both for primary studies and for meta-analyses over the time period surveyed. Sensitivity and specificity are considered inappropriate for meta-analyses, as they do not behave independently when they are pooled from various primary studies to generate separate averages.\[[@B2]\] In our survey, separate pooling of sensitivities or specificity was used frequently in meta-analyses where summary ROC would have been more appropriate. \[[@B5]-[@B7]\]. Our findings about reporting of summary accuracy measures in meta-analyses are different to those reported previously.\[[@B3]\] We found a higher rate of use of summary ROC, though use of independent summaries of sensitivity, specificity and predictive values were similar. These differences may be due to differences in searching strategies (databases and time frames) and selection criteria. Our search was more recent and comprehensive, using DARE\[[@B4]\], which has covered seven different databases (Medline, CINAHL, BIOSIS, Allied and Alternative Medicine, ERIC, Current Contents clinical medicine and PsycLIT), and hand-searched 68 peer-reviewed journals and publications from 33 health technology assessment centres around the world since February 1994. Moreover, as we did not restrict our selection to meta-analytical reviews only, we were able to examine reviews summarising accuracy results of primary studies without quantitative synthesis, which constituted 33% (30/90) of our sample. Therefore, compared to the previous publication on this topic,\[[@B3]\] our survey provided a broader and more up-to-date overview of the state of reporting of accuracy measure in test accuracy reviews. Conclusions =========== The use of inappropriate accuracy measures has the potential to bias judgement about the value of tests. Of the various approaches to reporting accuracy of dichotomous test results, likelihood ratios are considered to be more clinically powerful than sensitivities or specificities.\[[@B8]\] Crucially, it has been empirically shown that authors of primary studies may overstate the value of tests in the absence of likelihood ratios.\[[@B1]\] There is also evidence that readers themselves may misinterpret test accuracy measures following publication.\[[@B9]\] It is conceivable that the problem of inconsistent usage of test accuracy measures in published reviews, as found in our survey, may contribute to misinterpretation by clinical readership. The reason for variation in reported accuracy measures may, in part, be attributed to a lack of consensus regarding the best ways to summarise test results. It is worth noting that despite authoritative publications about appropriate summary accuracy measures in the past,\[[@B5],[@B7],[@B10]\] (we have only quoted a few references) inconsistent and inappropriate use of summary measures has remained prevalent in the period 1994--2000. Our paper highlights the need for consensus to support change in this field of research. Competing interest ================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Supplementary Material ====================== ::: {.caption} ###### Additional file Reference list of the 90 reviews included in the survey ::: ::: {.caption} ###### BMC_IncludedRefList_04032002 ::: Acknowledgement =============== We wished to acknowledge J. Dinnes, J. Glanville and F. Song for their contribution in the searches and selection of the reviews for the survey. Funding ======= RCOG Wellbeing grant number K2/00

Background ========== Most current methods of cancer early detection, such as mammography or cervical cytology, are based on anatomic changes in tissues or morphologic changes in cells. Recently, various molecular markers, such as protein or genetic changes have been proposed for cancer early detection \[[@B1]-[@B4]\]. This has spurred many investigators with long-term cohort studies to serially collect and store blood or tissue specimens. The aim is to later perform a nested-case control study, where specimens from subjects with a particular type of cancer (cases) and specimens from a random sample of subjects without the cancer (controls) are tested for various molecular markers. Sometime this sort of study is called a retrospective longitudinal study \[[@B6]\] although retrospective longitudinal data could arise in other ways, as well. Unlike cross-sectional study designs, the markers are measured on specimens collected well before the onset of clinical disease in cases. This avoids the potential confounding effect of the target disease on the marker. For example, in the ATBC (alpha-tocopherol, beta-carotene) \[[@B7]\] and CARET \[[@B8]\] studies, subjects were randomized to placebo or drug to in a long-term study to determine the effect of the drug on lung cancer mortality. During the course of the trial serum was serially collected and stored in a biorepository. In a subsequent nested case-control study, stored serum samples from all cases of prostate cancer and a random sample of controls were tested for prostate-specific antigen (PSA). Importantly the nested case-control study of early detection biomarkers may be distinct from the original long-term study from which serum were collected. It is designed to answer a different question, it typically studies subjects with a different disease, and it often ignores the intervention in the original long-term study. Methods ======= We had three considerations in formulating appropriate guidelines. First we wanted to link the analysis to the goal of study, namely, to help decide on further study of the biomarker as a trigger for early intervention. Second we wanted to minimize possible biases in the selection of cases and the controls and in the investigation of many markers. Third, we wanted to extract as much information as possible relevant to the evaluation. Results ======= We offer the following guidelines for the design and analysis of nested case-control studies of early detection cancer biomarkers. 1\. For the clearest interpretation, statistics for binary markers should be based on true and false positive rates or predictive values based on the true prevalence -- not odds ratios, relative risks, or predictive values based on the prevalence in the study A promising marker should have a high degree of accuracy in discriminating between subjects who are likely to get cancer from those who are not. For a binary marker, which is either positive or negative, the basic measures of this type of accuracy are the true positive and false positive rates. Consider the hypothetical data in the Table [1](#T1){ref-type="table"} from a nested case-control study of early detection biomarkers. The true positive rate (TPR), or the test sensitivity, is the probability the marker is positive given cancer. The false positive rate (FPR), or 1 -- specificity, is the probability the marker is positive given no cancer. In Table [1](#T1){ref-type="table"}, TPR is estimated by 80/100 = .80, and FPR is estimated by 10/1000 = 01. For a perfect test, TPR = 1 and the FPR = 0. There is always a tradeoff between the TPR and FPR, so it is meaningless to assess one without assessing the other. For example, one could make the TPR equal to 1 simply by classifying every subject as positive for cancer. But this would be a poor classification rule because the FPR would also be 1. ::: {#T1 .table-wrap} ::: {.caption} ###### Hypothetical data for a binary marker from a nested case-control study ::: Marker - Marker + Total ----------- ---------- ---------- ------- No cancer 990 10 1000 Cancer 20 80 100 True positive rate (TPR) = 80/100 = .80 = sensitivity False positive rate (FPR) = 10/1000 = .01 = 1 -- specificity ::: Guidelines for FPR and TPR Because the underlying prevalence of cancer is so low in average risk populations, for acceptable cancer screening of asymptomatic people, the FPR should be very small \[[@B9]\]. As a starting point we recommend basing guidelines for FPR and TPR on the FPR and TPR for mammography, which, as discussed below, is FPR=.01 and TPR=.80. A mammogram is analogous to a biomarker test for cancer but there is extra information from various studies that makes it useful for setting guidelines. One reason for using FPR and TPR from mammography as a standard is that, unlike biomarker measurements in nested case-control studies, in mammography studies there is a biopsy at the time of a positive test. This biopsy is the gold standard for determining cancer status in subjects with positive mammograms and is used for computing TPR in a way not possible with biomarkers in nested case-control studies. The TPR for mammography is the probability of a positive biopsy as a direct result of mammography in women with cancer and is estimated via mathematical models or data collected after following subjects not biopsied. As discussed in Baker and Pinsky \[[@B10]\], estimates ranged from .74 to 1.00 with .80 a conservative value. Importantly the estimated TPR for mammography is not likely to be affected by overdiagnosis, which means that some screening-detected cancers would never have caused medical problems during the patient\'s life \[[@B11]\]. This would make the biomarker appear more promising than actually the case. Results from the HIP screening trial of mammography and clinical self-examination \[[@B12]\] suggest that if there were overdiagnosis with mammography, it would be relatively small. At the time of the last breast screening in the HIP trial, there were more cancers in the group randomized to screening than in the controls. But with further follow-up the number of cancers in the control group eventually equaled the number in the screened group, which would not have occurred if there were substantial overdiagnosis. A second reason for using FPR and TPR from mammography is that, based on various randomized trials with cancer mortality endpoint, mammography is generally considered an acceptable screening modality. The implication is that a similar FPR and TPR for a biomarker would lead to an acceptable screening modality. For a particular biomarker, these target values of FPR and TPR from mammography may need modification depending on various factors. One factor is the invasiveness of a follow-up procedure to investigate a positive test (e.g. needle biopsy of the prostate to investigate an abnormal PSA versus laparotomy to investigate an abnormal CA125). The more invasive the follow-up procedure, the lower the FPR must be to gain acceptance in practice. A second factor is additional work-up prior to a biopsy. If a positive biomarker is unlikely to trigger additional diagnostic work-ups prior to biopsy, a higher FPR might be acceptable. One caveat when using FPR from mammography is to be careful as to its definition. The restricted definition is the probability of a positive biopsy as a direct result of mammography in women without cancer. The less restricted definition is the probability of a suspicious mammogram warranting additional diagnostic follow-up of any type in women without cancer. Typically nested-case control studies of early detection biomarkers do not provide information on additional diagnostic follow-up. Therefore they cannot be used to estimate a less restricted FPR involving diagnostic follow-up. However, because nested case-control biomarker studies provide data on cancer diagnosis, they can be used to estimate a more restricted FPR based on unnecessary biopsies. Therefore the target FPR is based on the more restricted definition of FPR in mammography. For mammography the more restricted FPR is estimated by the fraction of women screened by mammography who received a biopsy in which no cancer was detected. As discussed in Baker and Pinsky \[[@B10]\], estimates of FPR from three studies ranged from .005 to .013 with a middle value of around .010. Inappropriateness of odds ratio and relative risk When evaluating binary early detection markers, it is inappropriate to report an odds ratio or relative risk, as is common in epidemiology or clinical trials. Because an odds ratio or relative risk is a single number, it does not capture the tradeoff between correctly classifying cancer and incorrectly classifying non-cancers. Also the odds ratio or relative risk can lead to an overoptimistic impression of the performance of an early detection test if the interpretation is based on experience in epidemiology or clinical trials. In the latter settings an odds ratio of 3 is often considered large. Much larger odds ratio are needed from early detection tests for useful application in the screening setting \[[@B13]\]. For example, for the target values of FPR = .01 and TPR = .80, the odds ratio equals (TPR × (1-FPR)/ ((1-TPR) × FPR) = 396, as in Table [1](#T1){ref-type="table"}. Appropriate computation of predictive values It is sometimes useful to use the FPR and TPR to compute the predictive value negative (PVN), the probability of no disease if the marker is negative, and the predictive value positive (PVP), the probability of disease if the marker is positive. For cancer screening, it is the PVP that is most important to the physician in clinical decision-making. Because the likelihood of any individual cancer type in an asymptomatic person is nearly always very low, a negative early detection test usually adds little information to the clinical impression. The computation of the PVN and PVP depend on the prevalence of cancer as well as on the FPR and the TPR, as shown below, PVP = (TPR × prevalence) / (TPR × prevalence + FPR (1-prevalence)) PVN = (1-FPR) × (1-prevalence) / ((1-TPR) × prevalence + (1-FPR) (1-prevalence)) When the prevalence is small, as with cancer screening, a small absolute change in FPR, but not TPR, can have a large impact on the PVP. It is erroneous to directly compute the prevalence from data in the nested case-control study, as in Table [1](#T1){ref-type="table"}, because the ratio of cases to controls in the study is not the same as in the population. For example suppose FPR = .01, TPR = .80, and prevalence = .003. We obtain, PVP = = (.8 × .003) /(.8 × .003 + .01 .997)= .19 PVN = (.99 × .997) / (.2 × .003 + .99 × .997) = .999 If we had incorrectly substituted the apparent prevalence in Table [1](#T1){ref-type="table"} of 100/1100 = .091, we would have incorrectly computed PVP = .89 and PVN = .980. For this reason, calculation and reporting of PVP and PVN using only data from a nested case-control study is not useful or appropriate. Extension to ordered categories via ROC curves Many markers for the early detection of cancer can be reported as ordered categories. Some markers, such as sputum cytology, inherently involve ordered categories, such as no evidence of cancer, slight atypia, moderate atypia, severe atypia, and frank cancer. Other markers, such as PSA, involve a continuous measure for which higher values indicate a greater probability of cancer. Dividing these continuous measures into ranges (either based on predetermined values or percentiles) gives ordered categories. With ordered categories, the statistics should still be based on FPR and TPR. Each ordered category is a possible cutpoint, where values equal to or higher than the cutpoint are called positive and values lower than the cutpoint are called negative. For each cutpoint, one can compute FPR and TPR (Table [2](#T2){ref-type="table"}) and generate a receiver-operating characteristic (ROC) curve, which is a plot of TPR versus FPR for the various pairs \[[@B14]\]. (See Figure [1](#F1){ref-type="fig"}). The higher and farther left the points on the ROC curve the better the test performance. As mentioned previously, acceptable cancer screening requires very small false positive rates. Therefore, for evaluating cancer biomarkers, we are only interested in the leftmost sliver of the ROC curve in Figure [1](#F1){ref-type="fig"}. {ref-type="table"}. Because acceptable cancer screening requires a very small false positive rate (FPR) and a reasonable true positive rate (TPR), only the four points farthest to the left are relevant for biomarker evaluation.](1471-2288-2-4-1){#F1} ::: {#T2 .table-wrap} ::: {.caption} ###### Hypothetical Data for an Ordered Marker From a Nested Case-Control Design ::: 1 2 3 4 5 ----------- ----- ---- ---- ---- ---- ------ No cancer 960 20 10 8 2 1000 Cancer 0 10 10 20 60 100 For cutpoint 4, the true positive rate is (20+60)/100 = .80 For cutpoint 4, the false positive rate is (8 + 2)/1000 = .01 ::: 2\. To avoid overdiagnosis bias, cases should be diagnosed as a result of symptoms rather than on screening For the TPR in the biomarker study to reflect the true TPR, cases should be diagnosed as a result of symptoms rather than on screening. For example in the study of PSA in the ATBC trial \[[@B7]\], cases were subjects diagnosed with prostate cancer as a result of symptoms. If the prostate cancer cases were detected as the result of screening, say with ultrasound, the TPR could be artificially elevated if there were overdiagnosis, as previously discussed. 3\. To minimized selection bias, the spectrum of control conditions should be the same in study and target screening populations For the FPR in the nested case-control study to reflect the true FPR in the target population, the spectrum of control conditions should be the same as in the target population. By control conditions, we mean characteristics of the population, such as the presence of other diseases or certain known risk factors that could elevate the false positive rate. The spectrum of conditions could differ considerably if the retrospective biomarker study were embedded in a randomized trial with strict eligibility requirements. For example, consider a biomarker for the early detection of lung cancer where the data comes from a biorepository arising from a randomized trial of healthy subjects. It would be inappropriate to apply the results to a population with a high prevalence of chronic obstructive lung disease, bronchitis, or viral pnuemonitis because these conditions could increase the number of positive readings in subjects without lung cancer. Because FPR is very small for screening to be acceptable, this spectrum bias could have important consequences in a clinical application. It would not always be possible to identify all relevant control conditions, but to the extent possible, the control conditions should be similar in both populations. 4\. To extract additional information, criteria for a positive test should be based on combinations of individual markers and changes in marker levels over time Data from multiple markers present an opportunity to extract potentially valuable information not available by analyzing markers separately. Consider the hypothetical data in Table [3](#T3){ref-type="table"}. The left side corresponds to subjects without cancer, and the right corresponds to subjects with cancer. In our idealized circumstance, markers A and B are independent for subjects without cancer and are perfectly correlated for subjects with cancer. The indicated region for A = 2 or A = 3 and B = 2 or B = 3 designates a positive test that has FPR = .04 and the TPR = .80. Suppose that marker A and marker B were evaluated separately. The indicated region corresponding to A only, namely A = 2 or A = 3 and the indicated region corresponding to B only, namely B = 2 and B = 3, each designates a positive test that has FPR=. 20 and TPR=. 80. Thus, in this particular example, the combination of markers leads to a much better test than the separate markers, with a smaller FPR (.04 versus .20) for the same TPR of .80. One could create a similar example with any number of marker levels. For an ideal test in which FPR = 0 and TPR = 1, the region would encompass all subjects with cancer but no subjects without cancer. ::: {#T3 .table-wrap} ::: {.caption} ###### Hypothetical data for two ordered markers, A and B ::: Subjects without cancer (N = 1000) Subjects with cancer (N = 100) ------------ ------------------------------------ -------------------------------- ------------ ------------ ---- ----------- ----------- ----------- B = 1 640 80 80 800 20 0 0 20 B = 2 80 **10^a^** **10^a^** **100^e^** 0 **20^b^** **0^b^** **20^f^** B = 3 80 **10^a^** **10^a^** **100^e^** 0 **0^b^** **60^b^** **60^f^** **A only** 800 **100^c^** **100^c^** 20 **20^d^** **60^d^** **For A and B combined;**^a^False positive rate (FPR) of indicated region (10+10+10+10)/1000=. 04; ^b^True positive rate (TPR) of indicated region = (20 +0+0+60)/100=. 80 **For A only**; ^c^False positive rate (FPR) of indicated region =(100+100)/1000=. 20; ^d^True positive rate (TPR) of indicated region = (20+60)/100=. 80 **For B only**; ^e^False positive rate (FPR) of indicated region =(100+100)/1000= .20; ^f^True positive rate (TPR) of indicated region = (20+60)/100=.80 ::: A real application of how marker combinations provide extra information comes from a recent study by Mok et al \[[@B15]\] on CA125 and Prostasin as markers for ovarian cancer. Although the data are not from a nested case-control study and have not been validated in subsequent studies, they are useful for illustration. Based on Figure 5 of Mok et al \[[@B15]\] with approximate values for the outpoints, we investigated regions with FPR=.02. The combination for a positive test of CA 125 greater than or equal to 40 U/ml and Prostasin greater than or equal to 5 μg/ml had FPR = .02 and TPR = .92. Evaluating CA 125 alone, the criterion for a positive test of CA 125 greater than 80 U/ml had FPR=.02 and TPR =.78. Evaluating Prostasin alone, the criterion for a positive test of Prostasin greater than 15 μg/ml had FPR=. 02 and TPR=. 32. Thus, in this real example, the combination of markers led to a better test than the separate markers, with a larger TPR (.92 versus .78 or .32) for the same FPR of .02. To more generally compare performance of a combination of markers with a single marker, we need to compare ROC curves. Creating an ROC curve from a combination of markers is different from creating an ROC curve from a single marker. With a combination of markers, the number of possible regions, as in Table [3](#T3){ref-type="table"} a, for calling a marker positive is extremely large. Some choices of regions correspond to AND rules, for example A\>1 and B\>1, as in Table [3](#T3){ref-type="table"} a. Some choices of regions correspond to OR rules, for example A\>2 or B\>2. Other choices are also possible, but for biological reasons, one would usually require all regions to be contiguous. A plot of the FPR and TPR for each region would lead to a cloud of points, rather than the smooth curve in Figure [1](#F1){ref-type="fig"}. To create the best ROC curve, one should select those points that are highest and farthest to the left, which is generally all that would need to be presented and only for small false positive rates. For complicated situations, Baker \[[@B16]\] proposed an algorithm to select the regions creating the best ROC curve without the need to enumerate all the regions. Mathematically, in any sample of data, the best ROC curve for a combination of markers must be as good or better than the ROC curve for any of the markers evaluated separately. The reason is that the set of possible regions for calling a combination of markers positive includes as a special case the regions for calling any single marker positive. Alternative approaches that do not directly optimize the ROC curve include linear logistic regression or linear discriminant analysis \[[@B17]\], which choose regions based on linear combinations of the markers, and neural networks \[[@B18]\], which choose regions in a very complicated nonlinear manner. Due to the potential for overfitting (to follow), it is not possible to make a blanket statement as to which approach for choosing regions is best. If one takes the set of regions for calling a combination positive that gives a good ROC curve in a random sample of data, it may give a poor ROC curve in another random sample of the data, simply due to selecting chance patterns in the first sample. As discussed in the section on overfitting, this motivates splitting the data into two random samples, training and test, and using the regions from the training sample to compute the definitive ROC curve based on data in the test sample. Changes in marker values over time also provide potentially valuable information not available when examining markers at a single time point. With marker measurements at two different times per subject (and approximately the same interval between times), a common summary measure is the slope. If investigators believe that both slope and baseline level predict cancer, the combination can be evaluated using the previously discussed methods for evaluating multiple markers, namely, treating baseline level and slope as two \"separate\" markers. With measurements at more than two times per subject, investigators may identify a more complicated feature, such as whether or not there is a sudden increase in marker levels \[[@B19]\]. When biomarker measurements occur at regular time intervals (and allowing different numbers of measurements for each subject), one can estimate TPR and FPR by using a first order Markov chain in reverse time, as described by Baker and Tockman \[[@B20]\] for the analysis of precancerous lesions for lung cancer. 5\. To avoid overfitting, the criteria for a positive marker combination developed in a training sample should be evaluated in a random test sample from the same study and, if possible, a validation sample from another study. With a single marker and a large number of subjects, there is usually no concern with overfitting. However with many combinations of markers, overfitting could invalidate results. Overfitting is often associated with step-wise regression models \[[@B21]\] but it can occur in other situations as well. Overfitting of a larger number of markers to a relatively small number of subjects produces a model that is overly sensitive to chance fluctuations in the data. As a simple example, overfitting occurs when a sports announcer reports that a baseball player had a very high batting average against left-handed pitchers in ballpark X over the past month. This average is not very reliable for future predictions because the particular set of factors, left-handed pitchers and ballpark X, were selected to give a high average. In reality the high average is more likely the result of chance factors that coincided with left handed pitchers at ballpark X during that particular month. One way to adjust for overfitting is to apply the factors in another sample not used for initial reporting. The first sample is known as the training sample, and the second sample is known as the test sample. For example, the prediction of the average against left handed pitchers in the ballpark X could be tested on data from a different month. Similarly, a standard statistical approach to adjust for overfitting is to randomly split the data into a training and test samples. This is called the split-sample approach. Promising marker combinations are identified in the training sample, but more reliable FPR and TPR measurements are made in the test sample because it involves different data. Baker \[[@B16]\] used the split-sample approach to evaluate the performance of four markers for prostate cancer. More sophisticated adjustments for overfitting include cross-validation and bootstrapping. Suppose the interest lies in a single statistic, such as the area under the part of the ROC curve corresponding to a small FPR. In a typical cross-validation analysis (i) the data are randomly split into ten samples each with 10% of the data, (ii) one of the 10% samples is deleted, (iii) the statistic is computed by applying the classification rule recalculated from the remaining 90% of the sample to the deleted 10% sample and (iv) an average is taken of the statistic over all ten 10% samples \[[@B22]\]. Although one could use a different fraction than 10%, ROC computation it is not possible with the most extreme version of cross-validation that involves leaving out one subject at a time. With bootstrapping, (i) the entire data set is resampled with replacement many times, (ii) on each resampling the statistics is computed, and (iii) the statistics from each resampling are combined in a special way \[[@B22]\]. In a recent study comparing these adjustments with other types of statistics in a different setting, Steyerberg et al \[[@B22]\] found that the split-sample approach tended to underestimate performance, cross-validation performed poorly on some statistics that were not normally distributed, and bootstrapping performed best overall. For our purposes of estimating an entire ROC curve rather than a summary statistic, more research is needed for cross-validation and bootstrapping, as it is not clear how best to combine ROC curves over different samples. Regardless of the method used to adjust for overfitting in forming a classification rule, to obtain the most reliable FPR and TPR measurements, the classification rule should ideally be evaluated in a *validation* sample from a *different* study, as in Baker \[[@B16]\]. 6\. To identify biomarkers with true and false positive rates similar to mammography, the training, test, and validation samples should each include at least 110 randomly selected subjects without cancer and 70 subjects with cancer (as based on FPR and TPR for mammography) The sample size is based on the need to determine if the biomarker is sufficiently promising for investigation as a trigger for early intervention in a future trial. As discussed previously, based on considerations from mammography, our target values are FPR = .01 and TPR= .80. In most situations, we think it would be of interest to specify a 95% confidence interval for TPR of (.70, .90). Using a normal approximation the target standard error is approximately .05. Setting the standard error of TPR, TPR × (1-TPR) / (square root of n), equal to .05 and solving for the sample size n, we obtain n= 64, which we round up to 70. In addition we think that in most situations the largest reasonable value of FPR would be .03 which is 3 times the number of false positives as with mammography screening. Because FPR is so small, we do not use a normal approximation. We specify a sample size of n = 110, so that under the binomial distribution with FPR=.01, the upper 2.5% bound equals .03 × 110. Strictly, these sample sizes apply only after a single criterion for a positive test has been identified. For a training sample, one might consider larger sample sizes. Conclusion ========== A major advantage of nested case-control studies for early detection biomarkers is that they can be done quickly if serum from a long-term study has been stored in a biorepository. Importantly the retrospective aspect does not compromise the validity. There are none of the usual problems with retrospective studies such as recall bias. Thus we anticipate that in the coming years, there will be many reports in literature from studies of this design. These guidelines should greatly help investigators design and analyze nested case-control studies for early detection biomarkers and help readers of the literature to interpret them. It bears emphasis that these studies do not prove clinical efficacy of the markers. Rather, they suggest which markers or marker combinations are the most promising candidates for further study as a trigger for early intervention in definitive trials with cancer-mortality endpoints. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgement =============== The authors thank David Ransohoff for his helpful comments.

Background ========== The introduction of small subunit ribosomal RNA as a tool in microbial taxonomy by Carl Woese and George Fox \[[@B1]\] led most microbiologists to assume that the concepts of animal and plant taxonomy could be extended to the realm of prokaryotes. In particular, it was assumed that a natural taxonomic system for microorganisms was feasible \[[@B2]\]. The goal of a natural taxonomic system is the formation of taxonomic groups that are defined by shared ancestry \[[@B3]\]. By definition, an ancestor that defines a monophyletic group can only give rise to members of this group. No organism outside this group has a lineage that traces back to the same ancestor (paraphyletic group); however, there might be earlier ancestors that define more inclusive monophyletic groups. The metaphor for organismal evolution that underlies a natural taxonomic system is a strictly bifurcating tree of species. A decade ago ribosomal RNA promised that one day it might be possible to place every extant organism on a universal tree of life, and the hope was that more genomic sequences would make this placement more accurate. However, the analyses of completely sequenced genomes initiated a reassessment of concepts in microbial evolution \[[@B4]\]. While some molecular markers were found to agree with one another e.g., \[[@B5]\], others do not \[[@B6]-[@B12]\]. Transfer of genetic information between divergent organisms has turned the tree of life into a net or web \[[@B13]\], and genomes into mosaics. Different parts of genomes have different histories, and representing the history of genome evolution as a single tree appears inconsistent with the data. Nevertheless, the assumption of a tree-like process still underlies many approaches. Genome content trees have been calculated based on the presence and absence of genes \[[@B14]-[@B16]\] or types of protein folds \[[@B17]\]. While there is limited agreement between genome and rRNA phylogeny, at present it remains unclear whether this similarity is based on shared ancestry of part of a less frequently exchanging genome core \[[@B18]\], or if the apparent congruence is itself the result of horizontal gene transfer \[[@B19]\]. Overall genome content is not best represented on a single tree. Fig. [1](#F1){ref-type="fig"} gives an example of an alternative depiction, where thickness of a line reflects percentage of genes shared between two genomes. The coherence among the three domains of life (Bacteria, Archaea, Eucarya \[[@B20]\]), is clearly reflected in genome content; i.e., Archaea share more genes with other Archaea than with Bacteria, but many features are incompatible with representing the relationships between different genomes as a tree. For example, the mesophilic euryarchaeon *Halobacterium* sp. has more genes in common with the mesophilic Bacteria than does the thermophilic crenarchaeote *Aeropyrum pernix*. However, the extremophilic euryarcheote *Archaeoglobus fulgidus* shares many more genes with the extremophilic bacteria, *Aquifex aeolicus* and *Thermotoga maritima* than does *Halobacterium*. While this example illustrates the web-like relationships among genomes, recent phylogenetic reconstructions from molecular data have explored only few alternatives to the tree-paradigm (e.g. \[[@B21],[@B22]\]). {#F1} One obvious drawback of the star-like representation in Fig. [1](#F1){ref-type="fig"} is that it utilizes BLAST search results only. Any phylogenetic information retained in the sequences is not utilized beyond the presence or absence decision based on a single expectation value cut-off. Because of recombination, individual genes themselves might be mosaic \[[@B23]\]; however, within-gene recombination of protein coding genes occurs mostly between closely related organisms. The redundancy of the genetic code greatly reduces recombination between divergent proteins. Even if a region is 100% conserved on the amino acid level, the encoding DNA can be so different as to allow the mismatch repair system to prevent recombination. For studies of single divergent orthologous protein encoding genes the assumption of a tree-like evolutionary history remains a reasonable expectation. In this manuscript we focus on methods that utilize the phylogenetic information that is retained in molecular sequence data, while not presuming that genomes as a whole evolved in a tree-like fashion. In an elegant approach Korbinian Strimmer and Arndt von Haeseler \[[@B24]\] utilized Bayesian posterior probabilities to assess the phylogenetic information contained in an alignment of four homologous sequences. With four sequences there are only three possible tree topologies, and thus the three posterior probabilities corresponding to these three trees must sum up to one. Utilizing a barycentric coordinate system, the resulting probability vector is represented as a point in an equilateral triangle (Fig. [2](#F2){ref-type="fig"}), where the distances of the point P to the three sides represent the three probabilities. Strimmer and von Haeseler applied this approach to depict the phylogenetic information content present in a multiple sequence alignment. They plot the results from the analyses of all possible quartets, where the four sequences are selected from a single multiple sequence alignment in the same coordinate system. If there is a lot of phylogenetic information in the alignment, then most probability vectors will fall close to one of the corners; conversely datasets containing little phylogenetic information will mainly result in vectors falling into the center of the triangle. Here we explore the application of this and similar approaches in comparative genome analyses. In particular, we compare different approaches to calculate Bayesian posterior probabilities, and we compare these probabilities to the more widely used bootstrap support values. We assess the reliability of the different probability mapping approaches through their application to mitochondrial genomes, and we illustrate their usefulness by mapping selected interphylum and interdomain relationships. ![Mapping of the probability vector onto an equilateral triangle. Each QuartOP is represented as a probability vector P inside an equilateral triangle. The position of P is determined by the barycentric coordinates (p~1~, p~2~, p~3~), which correspond to the posterior probabilities or bootstrap support values of the three possible tree topologies. The vertices of the triangle T~1~, T~2~ and T~3~ represent the three possible unrooted tree topologies. Geometrically, each of the coordinates (p~1~, p~2~, p~3~) equals the distance between P and the side of the triangle opposite the corresponding vertex. Points closer to a vertex T~i~ have a larger corresponding probability p~i~ and represent a more probable tree topology than the two alternatives. All the points are classified by their position in one of three zones: \"total\" zone, \"90%\" zone and \"99%\" zone, which are depicted schematically and not drawn to scale. In this diagram, point P corresponds to a dataset which has highest probability for the topology T~3~, but the probability is below 90%, so the point P is located in the \"total\" zone, but not in the 99% or 90% zone. Figure adapted from \[[@B24]\].](1471-2164-3-4-2){#F2} Results and Discussion ====================== Overview of data flow in probability mapping -------------------------------------------- An outline of our approach to genome probability mapping is given in Figure [3](#F3){ref-type="fig"}. Using SEALS \[[@B25]\] and MySQL we developed scripts that identify and retrieve **quart**ets of **o**rthologous **p**rotein-encoding open reading frames (QuartOPs) from four selected genomes. We use the term genome to denote the collection of all ORFs identified in a genome. (In the case of genomes that are not well annotated, it is feasible to use a very wide definition of ORF, e.g, all amino acid sequences encoded between two stop codons in any of the six possible reading frames. As long as one of the genomes included in the analyses is properly annotated, only those identified ORFs that are actually homologous to an identified ORF will become part of a quartet of orthologs.) We utilize an operational definition of an ortholog: two open reading frames are considered orthologous, if and only if they are each other\'s top scoring BLAST hit when one is used as a query to search the other genome. A QuartOP is formed when each of the open reading frames picks the other members of the quartet as the top scoring hit in searches of the respective genomes. QuartOPs are similar to the clusters of orthologous groups (COGs) maintained by the NCBI \[[@B26]-[@B28]\], but differ in that COGs require only unidirectional, circular best hit relationships for three of the reference genomes, whereas we require the reciprocal top hit relationship for the four genomes included in a quartet, and we do not limit our identification of QuartOPs to a number of reference genomes. Montague and Hutchison utilized a comparable approach in their definition of congruent COGs \[[@B29]\]. So far we have analyzed 68 genome quartets (see supplementary material). The number of QuartOPs identified per genome quartet ranges from 82 (for genome quartet \#6: *Deinococcus radiodurans, Treponema pallidum, Escherichia coli,* and *Halobacterium sp*.) to 1182 (for genome quartet \#63: *Agrobacterium tumifaciens, Sinorhizobium meliloti, Mezorhizobium loti* and *Caulobacter crescentus*). {#F3} Each of the aligned QuartOPs from a genome quartet was analyzed with respect to the posterior probability of the three possible tree topologies given the aligned QuartOP. Routinely we calculated these probabilities using Strimmer\'s and von Haeseler\'s approach \[[@B24]\]: Using each of the three topologies as a usertree, we calculated the maximum likelihood for each of the three topologies given the data. We then use the three maximum likelihoods to calculate the probability for topology *i* according to the formula: *P*~*i*~ = *L*~*i*~/(*L*~1~+*L*~2~+*L*~3~), where *L*~*i*~ is the likelihood for the best tree given topology *i*. Other types of reliability measures used to evaluate QuartOPs were bootstrap support values and Bayesian posterior probabilities estimated using MrBayes program (see below). An example for the comparison of four genomes from different phyla is given in Fig. [4A](#F4){ref-type="fig"}. Surprisingly, each of the three tree topologies is strongly supported by more than 40 QuartOPs, and most of the QuartOPs appear to strongly support one of the trees. None of the three possibilities has majority support. Figure [5](#F5){ref-type="fig"} lists the functional categories of those QuartOPs that strongly support the different tree topologies. None of the categories shows a preference for a particular tree topology. For each tree topology more than 50% of the strongly supporting QuartOPs belong to the category \"information storage and processing\", while this category contains only about 1/3 of the genes present in the genomes. While the genes in this category appear more conserved and phylogenetically informative, the strong support that the genes in this category provide is nearly evenly split between the three possibilities. {ref-type="fig"}. The three numbers associated with each tree topology indicate how many QuartOPs fall into each of the three zones: \"total\", 90% and 99% respectively. For definition of zones see figure [2](#F2){ref-type="fig"}. **A)** Probabilities are calculated according to Strimmer and von Haeseler \[[@B24]\]. There is no single topology that is supported by the majority of the QuartOPs and all three possible tree topologies are supported by roughly equal number of QuartOPs at the different probability levels. **B)** Probabilities are calculated with MrBayes program \[[@B31]\]. **C)** Bootstrap support values are plotted. For this case the zones are \"total\", 70% and 90% support, respectively. Bootstrapping appears to provide a more conservative reliability estimate than the posterior probabilities used in cases A and B. Nevertheless, each tree topology is still supported by a roughly equal number of bootstrapped datasets.](1471-2164-3-4-4){#F4} ![Distribution among different functional categories for those datasets that support one of the three topologies with better than 99% posterior probability. Tree topologies are indicated by column numbers 1, 2 and 3. Column 1 corresponds to topology ((1,4),2,3), columns 2 and 3 correspond to topologies ((1,3),2,4) and ((1,2),3,4) respectively. Divisions into functional categories are adopted from the COG database \[[@B27]\]. Functional categories are aggregated into four broad functional meta-categories. Distributions of datasets among the meta-categories are plotted as pie charts for each tree topology. In this case all three topologies are supported by roughly equal number of datasets from each meta-category.](1471-2164-3-4-5){#F5} Impact of model parameters and sequence conservation ---------------------------------------------------- To test if ill-aligned sequences might have had an impact on the analyses, we repeated the analysis of genome quartet \# 8 (see Figure [4](#F4){ref-type="fig"}) using only QuartOPs that contained very similar sequences. By default we only excluded top hits with an E-value larger than 10^-4^. We repeated the example given in Fig. [4A](#F4){ref-type="fig"} with a cut-off of 10^-20^, i.e., we not only required the sequences in a QuartOP to be each others top hit, but in addition we asked for a high similarity between the two sequences. As a result the support for the three topologies in genome quartet \#8 dropped to 54 (44, 38), 51 (45, 32), 39 (29, 28) (the numbers in parenthesis are the numbers of quartets that support the topology with posterior probability larger than 90% and 99%, respectively). To access the level of sequence conservation within the QuartOPs\' sequences, we calculated the average percentage of pairwise identity per QuartOP. It varied from 40.53 10.54% to 43.84 ± 9.7% when the E-value cut-off was varied between 10^-2^ and 10^-20^ (see supplementary material for the summary table). While pairwise sequence identity is not a universally dependable measure of phylogenetic information content, these values illustrates that the sequences within a QuartOP are neither identical to one another, nor so divergent as to be saturated with substitutions and of questionable homology \[[@B30]\]. Using only the most conserved QuartOPs does not change the qualitative result: each of the three possible tree topologies is supported by about an equal number of QuartOPs (see supplementary material). We recalculated the likelihoods for all QuartOPs in genome quartet \#8 using a model that incorporates **a**mong **s**ite **r**ate **v**ariation (ASRV). The posterior probabilities calculated according to Strimmer and von Haeseler did not change dramatically and each of the three tree topologies is still supported by roughly equal number of QuartOPs. The maps for this analysis are available in the supplementary material. Estimating Bayesian Posterior Probabilities ------------------------------------------- The formula used by Strimmer and von Haeseler \[[@B24]\] to calculate posterior probabilities (i.e. the probability that tree topology *T*~*i*~ is true given an aligned set of four sequences) considers only three trees (i.e. branch lengths and topology), each with the same prior probability. These three trees are those that have the highest likelihood for the three possible topologies. However, there are infinitely many other trees that differ from the three chosen ones only by differences in branch lengths. What is the effect on the calculated posterior probability of using only the single best tree as a representative of all the trees with the same topology? There is no *a priori* reason to exclude the other trees that have slightly lower likelihoods. A different approach that does not make these assumptions is the use of Markov Chain Monte Carlo methods to explore tree space. We used the program MrBayes written by Huelsenbeck and Ronquist \[[@B31]\]. Using a QuartOP with posterior probabilities of .76, .10 and .13 we explored different parameter choices for the biased random walk through tree-space. We chose two chains with 5,000 burn-in cycles, and 25,000 cycles with sampling after every cycle as a compromise between increased precision of the probability estimate and computation time (see Materials and Methods for more details). The result of calculating the posterior probabilities of all QuartOPs in genome quartet \#8 is given in figure [4B](#F4){ref-type="fig"}. Again all of the three tree topologies are strongly supported by some QuartOPs. When we repeated this analysis using the same settings, none of the probabilities changed by more than a few percent. The support for the three tree topologies changed from 67/47/37, 69/40/22 and 86/55/32 (the three numbers indicate total support and QuartOPs that supported a topology with more than .90 and .99 respectively) in the first run to 67/47/37, 70/41/22 and 85/55/32 in the second, indicating that the chosen parameters provided satisfactory reproducibility. Plots of both analyses are available in the supplementary material. Comparing figure [4B](#F4){ref-type="fig"} with [4A](#F4){ref-type="fig"} it is clear that in this case the Bayesian posterior probabilities estimated with MrBayes are more conservative assessments of reliability than the ones calculated according to \[[@B24]\]. The 99% support level calculated according to \[[@B24]\] approximately corresponds to the 90% support level calculated with MrBayes. Bootstrap support values versus posterior probabilities ------------------------------------------------------- To facilitate comparison of Bayesian posterior probabilities with a more widely used confidence measure, we generated 100 bootstrapped samples \[[@B32]\] from each QuartOP in case \#8. Each of the bootstrapped samples was analyzed using maximum likelihood with the same model of substitution as before. Each of the bootstrapped samples supports one of the three possible topologies, thus the sum of the bootstrap support values for the three topologies adds up to 100%, and the percentage of bootstrapped samples for each QuartOP that best supported each tree was again plotted in a barycentric coordinate system (Fig. [4C](#F4){ref-type="fig"}). Many more QuartOPs map into the central region of the triangle as compared to Figure [4A](#F4){ref-type="fig"} and [4B](#F4){ref-type="fig"}. Clearly, for this test bootstrap support values are more conservative measures of support than either of the posterior probabilities calculated above. Nevertheless, there are still several QuartOPs that strongly support each of the three tree topologies; however, there are 22 QuartOPs that support grouping *E. coli* with *Treponema pallidum* with better than 90% bootstrap support, whereas the alternatives are supported by only 12 and 13 QuartOPs, respectively. Comparing Figures [4A](#F4){ref-type="fig"} and [4C](#F4){ref-type="fig"} it appears that in analyzing quartets 70% bootstrap support is comparable to .99 posterior probability calculated according to \[[@B24]\]. Comparison of the different reliability assessment tools -------------------------------------------------------- ML-mapping according to \[[@B24]\] is the least conservative of the tools explored. For the test cases analyzed a posterior probability of .99 according to \[[@B24]\], corresponds to a Bayesian posterior probability of .90 calculating using a Markov chain exploration of tree-space using \[[@B31]\] and about 70% bootstrap support. We did not find a strong dependence of the results on the substitution models used in calculating likelihoods and separate runs indicated satisfactory precision of the calculated probabilities and bootstrap values. Given that we only analyzed about 300 QuartOPs using all three approaches it would be premature to generalize our findings; however, other analyses that utilized both bootstrapping and Bayesian posterior probabilities also found bootstrapping to be more conservative than posterior probabilities calculated using Bayesian methods with Markov chain Monte Carlo sampling (e.g., \[[@B33]-[@B35]\]). Mitochondrial genomes --------------------- While gene transfer into the mitochondrial genomes has been inferred \[[@B36]-[@B39]\], mitochondrial genomes are expected to have undergone many fewer legitimate and illegitimate recombination events than free-living prokaryotes. Clearly, if probability mapping is to be considered a reliable approach, we expect that when analyzing quartets of mitochondrial genomes, the different genes should all support the same tree topology. In most instances, this expectation is fulfilled (see Table [1](#T1){ref-type="table"}), even though we selected instances in which the splits could be expected to be ill resolved, e.g., echinoderm, mammal, insect, mollusk (m4), or protist, fungus, animal, plant (m7). The only exception was an ORF in quartet m7 that encodes the cytochrome oxidase subunit II. This ORF did not support grouping the animal with the fungal homolog as expected; rather it grouped the protist and animal homologs together (posterior probability according to \[[@B24]\] was " 0.99). Inspection of the aligned sequences (Fig. [6](#F6){ref-type="fig"}) revealed that there are more residues shared between the homologs from *Cafeteria roenbergensis* and *Homo sapiens* thanbetween the homologs from *Cafeteria roenbergensis* and *Arabidopsis thaliana*. No artifact that could be responsible for this unexpected grouping was detected. The same high support for this unexpected grouping is also recovered in bootstrap analysis and in posterior probabilities calculated with MrBayes \[[@B31]\]. {ref-type="table"}) that supports the unexpected ((Homo sapiens, Cafeteria), Saccharomyces, Arabidopsis) topology. The exact matches for each tree topology are colored in three different colors. Blue corresponds to the ((Homo sapiens, Cafeteria), Saccharomyces, Arabidopsis), yellow corresponds to the ((Homo sapiens, Arabidopsis), Saccharomyces, Cafeteria) and green corresponds to the ((Homo sapiens, Saccharomyces), Arabidopsis, Cafeteria) tree topology. As can be seen, the majority of the matches are in favor of ((Homo sapiens, Cafeteria), Saccharomyces, Arabidopsis) tree topology. There are nine parsimony informative positions favoring the latter topology, and only three for each of the other two topologies.](1471-2164-3-4-6){#F6} ::: {#T1 .table-wrap} ::: {.caption} ###### Results of analyses for the control mitochondrial genome quartets \#m1-m7. ::: **\#** **Genome 1** **Genome 2** **Genome 3** **Genome 4** **((1,2),3,4)** ** ** **((1,3),2,4)** ** ** **((1,4),2,3)** -------- ------------------- ---------------- -------------------- ---------------- ----------------- ------- ----------------- ------- ----------------- -------- -------- ------- --- --- --- **m1** **Drosophila** **Drosophila** Ceratitis capitata Apis mellifera **9** **8** **8** ** ** 0 0 0 ** ** 1 0 0 **melanogaster** **yakuba** ligustica **10** **8** **8** ** ** 0 0 0 ** ** 0 0 0 **9** **7** **5** ** ** 0 0 0 ** ** 1 1 0 ** ** ** ** ** ** ** ** ** ** **m2** **Alligator** Opossum **Stork** Donkey 0 0 0 ** ** **11** **11** **11** ** ** 0 0 0 0 0 0 ** ** **11** **11** **11** ** ** 0 0 0 0 0 0 ** ** **11** **11** **11** ** ** 0 0 0 ** ** ** ** ** ** ** ** ** ** **m3** **Turtle** Opossum **Stork** Donkey 0 0 0 ** ** **12** **12** **12** ** ** 0 0 0 0 0 0 ** ** **12** **12** **12** ** ** 0 0 0 0 0 0 ** ** **12** **12** **12** ** ** 0 0 0 ** ** ** ** ** ** ** ** ** ** **m4** **Starfish** **Donkey** Fruit Fly Doorsnail **8** **7** **7** ** ** 2 1 1 ** ** 0 0 0 **8** **7** **7** ** ** 1 0 0 ** ** 1 0 0 **8** **7** **5** ** ** 1 1 0 ** ** 1 0 0 ** ** ** ** ** ** ** ** ** ** **m6** **Reclinomonas** Saccharomyces **Arabidopsis** Homo sapiens 0 0 0 ** ** **4** **3** **3** ** ** 0 0 0 **americana** cerevisiae **thaliana** 1 0 0 ** ** **3** **3** **3** ** ** 0 0 0 0 0 0 ** ** **4** **3** **2** ** ** 0 0 0 ** ** ** ** ** ** ** ** ** ** **m7** **Cafeteria** Saccharomyces **Arabidopsis** Homo sapiens 0 0 0 ** ** **3** **2** **2** ** ** 1 1 1 **roenbergensis** cerevisiae **thaliana** 1 0 0 ** ** **2** **2** **1** ** ** 1 1 1 1 0 0 ** ** **2** **1** **0** ** ** 1 1 1 The groupings corresponding to the expected organismal phylogenies are given in bold. The three numbers in each table cell correspond to the three approaches used. The top number corresponds to results obtained using Strimmer and von Haeseler\'s approach \[24\], the middle number corresponds to results obtained using MrBayes program \[31\], and the bottom number corresponds to results of bootstrap support values calculation. Column \"Tot.\" lists the number of QuartOPs from \"total\" zone, column A lists the number of QuartOPs from \"90%\" zone (70% for bootstrap support), and column B lists the number of QuartOPs from \"99%\" zone (90% for bootstrap support). For definition of zones see Fig. 2. With the exception of the one dataset for quartet \#m7, the analyses proved to be consistent with organismal tree topologies. The alignment for the exceptional dataset is presented in Fig. 6. The common names for the organisms listed correspond to the following scientific names: alligator corresponds to *Alligator mississippiensis*, opossum to *Didelphis virginiana*, stork to *Ciconia ciconia*, donkey to *Equus asinus*, turtle to *Chelonia mydas*, starfish to *Asterina pectinifera*, doorsnail to *Albinaria caerulia*, fruit fly to *Drosophila melanogaster*. ::: The finding of a QuartOP in a mitochondrial genome quartet that supports a non-traditional grouping could either reflect a rare recombination event, selection pressures that led to convergent evolution in two lineages, or a chance event -- if one looks at enough samples one will find some that (considered by themselves) appear significant. At present it is not possible to decide between these three possible explanations. Our analysis of mitochondrial genomes shows that in most instances the calculated probabilities (ML-mapping, Bayesian posterior probabilities, or bootstrap values) support the expected tree topologies, albeit with surprisingly strong support values. Rarely, unexpected groupings can be recovered and support for these probably erroneous groupings can be high. In most instances the ML-mapping approach accurately revealed the expected relationships between the mitochondrial genomes. This confirms the suitability of this approach in genome analyses. Interphylum genome quartets --------------------------- Here we focus on examples that illustrate the utility of the probability mapping approach. Focusing on the relationships between the cyanobacteria with other bacterial phyla we calculated several genome quartets that include the *Synechocystis* sp. genome and three members each of other bacterial phyla (see Table [2](#T2){ref-type="table"}). In all cases that included both *Bacillus subtilis* as a representative of the low GC Gram positives, and *Synechocystis* sp., the majority of QuartOPs supported the topologies that grouped these two organisms together. The alternative topologies were significantly supported by some QuartOPs, but the number of strongly supporting QuartOPs was lower than for the *B. subtilis* -- *Synechocystis* grouping. This also was true when one of the other two genomes was from a high GC Gram positive (genome quartets \#53, \#54, \#55 and \#68). Only when two low GC Gram positives were included in the same quartet, was the intra-phylum grouping of low GC Gram positives supported by many more QuartOPs than the grouping of *B. subtilis* with *Synechocystis* sp. (genome quartets \#51 and \#52). ::: {#T2 .table-wrap} ::: {.caption} ###### Summary of the genome quartets that include *Synechocystis* sp., *Bacillus subtilis*, and two bacterial genomes from other phyla. ::: **\#** **Genome 1** **Genome 2** **Genome 3** **Genome 4** **((1,2),3,4)** ** ** **((1,3),2,4)** ** ** **((1,4),2,3)** -------- ----------------- ------------------ ------------------ ------------------- ----------------- ------- ----------------- ------- ----------------- ----- ------------ ------- ----- ----- --------- **9** *Synechocystis* *P. aeruginosa* *D. radiodurans* *B. subtilis* 94 76 63 ** ** 101 73 *57* ** ** 186 158 **126** **10** *Synechocystis* *P. aeruginosa* *T. pallidum* *B. subtilis* 69 54 50 ** ** 51 33 *28* ** ** 102 80 **67** **14** *Synechocystis* *R. sphaeroides* *B. subtilis* *E. coli* 65 53 44 ** ** 286 263 **248** ** ** 28 25 17 **15** *Synechocystis* *R. sphaeroides* *B. subtilis* *D. radiodurans* 95 72 60 ** ** 201 173 **149** ** ** 73 60 47 **16** *Synechocystis* *D. radiodurans* *B. subtilis* *T. pallidum* 63 50 *40* ** ** 94 74 **63** ** ** 60 46 *35* **17** *Synechocystis* *B. subtilis* *T. pallidum* *E. coli* 93 72 **66** ** ** 55 43 *34* ** ** 66 49 39 **18** *Synechocystis* *D. radiodurans* *T. pallidum* *E. coli* 86 68 54 ** ** 65 49 38 ** ** 75 54 47 **19** *Synechocystis* *D. radiodurans* *B. subtilis* *E. coli* 129 105 *86* ** ** 156 131 **104** ** ** 98 76 62 **50** *Synechocystis* *B. subtilis* *E. coli* *M. loti* 276 255 **228** ** ** 44 31 21 ** ** 54 43 38 **53** *Synechocystis* *B. subtilis* *E. coli* *M. leprae* 125 104 **82** ** ** 101 84 65 ** ** 119 19 66 **54** *Synechocystis* *B. subtilis* *E. coli* *M. tuberculosis* 141 114 **97** ** ** 101 82 69 ** ** 128 101 89 **55** *Synechocystis* *B. subtilis* *M. loti* *M. tuberculosis* 189 164 **139** ** ** 92 74 59 ** ** 80 58 44 **64** *Synechocystis* *C. trachomatis* *B. subtilis* *T. pallidum* 47 31 24 ** ** 108 96 **86** ** ** 36 25 18 **67** *Synechocystis* *C. trachomatis* *B. subtilis* *M. loti* 64 48 32 ** ** 116 104 **84** ** ** 72 51 44 **68** *Synechocystis* *C. trachomatis* *B. subtilis* *M. tuberculosis* 77 55 45 ** ** 94 80 **62** ** ** 68 52 38 **51** *Synechocystis* *B. subtilis* *E. coli* *S. aureus* 33 19 15 ** ** 361 349 [333]{.ul} ** ** 15 7 5 **52** *Synechocystis* *B. subtilis* *E. coli* *S. pyogenes* 34 22 18 ** ** 259 249 [227]{.ul} ** ** 24 17 9 The \# column refers to the unique number assigned to the genome quartets analyzed. Columns \"((1,2),3,4)\", \"((1,3),2,4)\" and \"((1,4),2,3)\" refer to the three possible tree topologies. Numbers in columns \"Tot\", \".90\" and \".99\" give the number of QuartOPs that support the indicated tree topology with a posterior probability higher than the other two posterior probabilities, or with 90% or 99% probability, respectively. The numbers in bold indicate the number of orthologs supporting the grouping of *Synechocystis* sp. and *Bacillus subtilis* in the absence of another low GC gram-positive in the genome quartets. Note that those numbers are the largest of the three numbers, a finding that supports the recent analyses by \[43\]. In the presence of another low GC Gram-positive in addition to *Bacillus subtilis*, the largest number of QuartOPs support grouping of low GC Gram-positives with each other (underlined). Other groupings that involve putative sister groups to the cyanobacteria (Deinococaceae and spirochetes) that had been suggested by others (e.g., \[40,41\]) are indicated in italics. ::: Previous analyses based on a limited number of proteins and signature insertions and deletions had suggested different bacterial groups as closest relatives to cyanobacteria. Among the suggested sister groups were the Deinococcales \[[@B40]\] and spirochetes \[[@B41],[@B42]\]. Our analyses do not support these earlier claims, but are in agreement with the recent analyses of genes involved in chlorophyll biosynthesis \[[@B43]\], which indicated that the low GC Gram positive heliobacteria are closest to the last common ancestor of all oxygenic photosynthetic lineages. The analyses summarized in Table [2](#T2){ref-type="table"} also illustrate that interphylum HGT, while turning genomes into mosaics, has not eroded all associations between bacterial phyla. In the case of cyanobacteria, a close association between low GC Gram positives and the cyanobacteria is supported by the majority of conserved genes. Similar observations of reproducible associations between phyla based on genome wide comparisons were recently published \[[@B44]-[@B47]\]. However, at present it cannot be decided to what extent these closer associations reflects shared ancestry or are due to preferred HGT \[[@B19]\]. Interdomain genome quartets --------------------------- In our search for the \"sister-phylum\" to the cyanobacteria we also analyzed a few quartets including Archaea. One noteworthy finding was that in the genome quartet including *Synechocystis* sp., *Halobacterium* sp., *Aquifex aeolicus* and *Thermotoga maritima* the grouping of *Halobacterium* sp. with *Synechocystis* sp. was recovered by many more QuartOPs (56 with p \> .99) than the grouping that would be expected following 16S rRNA phylogeny (12 QuartOPs with p \> .99; see Table [3](#T3){ref-type="table"}). To test if this association was specific for *Synechocystis* sp., we repeated the analyses replacing *Synechocystis* sp. with *Bacillus subtilis*. The result was qualitatively the same: at the p \> .99 level 53 QuartOPs supported grouping *Bacillus subtilis* with *Halobacterium* sp., and only 27 supported grouping *Aquifex aeolicus* with *Halobacterium* sp. (Fig. [7](#F7){ref-type="fig"}). {ref-type="table"}). For details on the figure notations see legend for Figure [4](#F4){ref-type="fig"}. **A.** Probabilities calculated according to Strimmer and von Haeseler \[[@B24]\]. **B.** Probabilities calculated with the MrBayes program \[[@B31]\].](1471-2164-3-4-7){#F7} ::: {#T3 .table-wrap} ::: {.caption} ###### Summary of the genome quartets that include the mesophilic archaeon *Halobacterium* sp. or the thermophilic archaeon *Archaeoglobus fulgidus*, deep-branching bacteria *Thermotoga maritima* and *Aquifex aeolicus*, and bacteria *Synechocystis* sp. or *Bacillus subtilis*. ::: -------- --------------------- ----------------------- -------------------- -------------------------- ------------------- --------- ------------------- ------- ------------------- --------- ---------- ------- ---------- --------- ---------- **\#** **Genome 1** **Genome 2** **Genome 3** **Genome 4** ***((1,2),3,4)*** ** ** ***((1,3),2,4)*** ** ** ***((1,4),2,3)*** **Tot.** **0.9** **0.99** ** ** **Tot.** **0.9** **0.99** ** ** **Tot.** **0.9** **0.99** **11** *Synechocystis sp.* *Thermotoga maritima* *Aquifex aeolicus* *Halobacterium sp.* 29 20 12 ** ** 45 34 27 ** ** 86 69 56 **13** *Synechocystis sp.* *Thermotoga maritima* *Aquifex aeolicus* *Archaeoglobus fulgidus* 47 36 30 ** ** 63 51 44 ** ** 50 34 25 **61** *Bacillus subtilis* *Thermotoga maritima* *Aquifex aeolicus* *Halobacterium sp.* 44 32 27 ** ** 27 19 11 ** ** 83 62 53 **62** *Bacillus subtilis* *Thermotoga maritima* *Aquifex aeolicus* *Archaeoglobus fulgidus* 64 50 40 ** ** 50 35 30 ** ** 41 27 23 -------- --------------------- ----------------------- -------------------- -------------------------- ------------------- --------- ------------------- ------- ------------------- --------- ---------- ------- ---------- --------- ---------- For table notations see legend for Table 2. Quartets \#11 and 61 indicate that the majority of the QuartOPs group *Halobacterium* sp. together with *Synechocystis* sp. and with *Bacillus subtilis* respectively, which is in disagreement with 16S rRNA topology. In two control quartets (\#13 and \#62) *Halobacterium* was substituted with *Archaeoglobus fulgidus*, and in these cases the majority of QuartOPs support the topology that is in agreement with SSU rRNA topology. ::: Clearly, there are many artifacts possible in analyzing divergent sequences. For many QuartOPs the ortholog from *Halobacterium* sp. is expected to be the longest branch. To test for the possibility that long branch attraction \[[@B48]\] might be the reason for the strong support of *Halobacterium* sp. grouping with *Synechocystis* sp., we repeated the analysis replacing the *Halobacterium* sp. genome with that from *Archaeoglobus fulgidus*, another archaeon. Gratifyingly, many more QuartOPs supported the grouping of the thermophilic archaeon *Archaeoglobus* with the thermophilic bacteria *Aquifex* and *Thermotoga*. The different interdomain genome quartets that include a meso- or thermophilic archaeon are summarized in Table [3](#T3){ref-type="table"}. An analysis of the putative functional assignments of the QuartOPs that grouped *Halobacterium* sp. with the mesophilic bacteria is given in Table [4](#T4){ref-type="table"}. To assess which of these categories have an increased percentage of QuartOPs as compared to distribution of ORFs within the genome, we also calculated the distributions of ORFs among functional categories in the *Halobacterium* sp. and *A. fulgidus* genomes. Open reading frames within a genome are distributed almost evenly among the four meta-categories (see columns labeled \"H\" and \"A\"). However, the QuartOPs that group the halobacterial orthologs with those from mesophilic Bacteria are distributed differentially among the meta-categories. Most of the QuartOPs are in the \"Metabolism\" and \"Information Storage and Processing\" meta-categories. These are also the categories in which *Halobacterium* sp. shows many more OuartOPs in support of topology 3 than *A. fulgidus*. ::: {#T4 .table-wrap} ::: {.caption} ###### Distribution of the datasets that strongly support (with 99% posterior probability) one of the three topologies among different functional categories. ::: **\#11** ** ** **\#13** ** ** **\#61** ** ** **\#62** ** ** **H** ** ** **A** ---------------------------------------------------------------- --- ---------- ---- --------- ---- ---------- ---- --------- ---- ---------- ---- --------- ---- ---------- ---- --------- ------- --------- ------- Information storage and processing 5 7 20 ** ** 7 7 10 ** ** 11 4 17 ** ** 12 4 0 ** ** 24 ** ** 17 **J** Translation, ribosomal structure and biogenesis 4 6 14 ** ** 5 6 9 ** ** 9 3 11 ** ** 10 3 0 ** ** 31 ** ** 44 **K** Transcription 1 1 0 ** ** 1 1 0 ** ** 1 1 0 ** ** 1 1 0 ** ** 30 ** ** 30 **L** DNA replication, recombination and repair 0 0 6 ** ** 1 0 1 ** ** 1 0 6 ** ** 1 0 0 ** ** 39 ** ** 26 Cellular processes 1 5 4 ** ** 5 7 1 ** ** 5 1 6 ** ** 6 3 1 ** ** 21 ** ** 16 **D** Cell division and chromosome partitioning 0 0 1 ** ** 2 0 1 ** ** 3 0 0 ** ** 3 0 0 ** ** 7 ** ** 5 **O** Posttranslational mod., protein turnover, chaperones 1 0 1 ** ** 1 1 0 ** ** 1 0 3 ** ** 2 0 0 ** ** 22 ** ** 18 **M** Cell envelope biogenesis, outer membrane 0 3 0 ** ** 1 1 0 ** ** 0 1 0 ** ** 1 0 0 ** ** 13 ** ** 14 **N** Cell motility and secretion 0 0 1 ** ** 1 1 0 ** ** 1 0 0 ** ** 0 0 0 ** ** 15 ** ** 10 **P** Inorganic ion transport and metabolism 0 2 1 ** ** 0 4 0 ** ** 0 0 3 ** ** 0 3 1 ** ** 30 ** ** 31 **T** Signal transduction mechanisms 0 0 0 ** ** 0 0 0 ** ** 0 0 0 ** ** 0 0 0 ** ** 14 ** ** 22 Metabolism 6 16 29 ** ** 18 30 12 ** ** 9 6 30 ** ** 22 28 14 ** ** 30 ** ** 37 **C** Energy production and conversion 2 3 1 ** ** 1 4 1 ** ** 0 0 0 ** ** 3 0 0 ** ** 23 ** ** 31 **G** Carbohydrate transport and metabolism 1 2 1 ** ** 2 2 0 ** ** 3 0 2 ** ** 2 1 1 ** ** 12 ** ** 8 **E** Amino acid transport and metabolism 1 5 16 ** ** 12 10 7 ** ** 2 2 14 ** ** 10 12 6 ** ** 27 ** ** 24 **F** Nucleotide transport and metabolism 2 4 10 ** ** 1 9 4 ** ** 2 2 11 ** ** 4 10 7 ** ** 10 ** ** 7 **H** Coenzyme metabolism 0 2 1 ** ** 2 5 0 ** ** 2 2 3 ** ** 3 5 0 ** ** 19 ** ** 16 **I** Lipid metabolism 0 0 0 ** ** 0 0 0 ** ** 0 0 0 ** ** 0 0 0 ** ** 9 ** ** 14 Poorly characterized 1 1 3 ** ** 1 1 0 ** ** 2 1 2 ** ** 1 1 1 ** ** 24 ** ** 30 **R** General function prediction only 1 1 3 ** ** 1 1 0 ** ** 2 1 2 ** ** 1 1 1 ** ** 64 ** ** 58 **S** Function unknown 0 0 0 ** ** 0 0 0 ** ** 0 0 0 ** ** 0 0 0 ** ** 36 ** ** 42 The distribution corresponds to the genome quartets listed in Table 3. Functional categories are as designated in Fig. 5. Columns 1, 2 and 3 correspond to the three possible unrooted topologies for each genome quartet (see Table 3). Column entries indicate the number of QuartOPs in each functional category. The last two columns represent the distribution of ORFs in *Halobacterium* sp. (H) and *Archaeoglobus fulgidus* (A) genomes among different functional categories. For these two columns, numbers in the rows corresponding to the meta-categories give the percentage of proteins in each meta category relative to the total number of classifiable proteins and numbers in the rows for each functional category indicate the percent distribution of the proteins within the corresponding meta-category. ::: The analyses described in this section reconfirm that genes have been transferred across domain boundaries \[[@B6]-[@B12]\]. Not surprisingly, these transfers appear to occur preferentially between organisms living in the same or similar environment. The genome of the mesophilic *Halobacterium* sp. contains many genes that group with the orthologs from mesophilic bacteria, whereas the majority of genes from the thermophilic archaeon *Archaeoglobus fulgidus* group with the orthologs from the extremely thermophilic bacteria. The majority of QuartOPs that group the halobacterial orthologs with the ortholog from the mesophilic bactria belong to two of four meta-categories: \"Information Storage and Processing\" and \"Metabolism\". QuartOPs in Information Storage and Processing meta-category that support the grouping of *Halobacterium* sp. with *Synechocystis/Bacillus* are listed in the Table [5](#T5){ref-type="table"}. A complete listing is available in the supplementary material. As expected, this list includes several tRNA synthetases, which were previously found to be frequently transferred \[[@B6]-[@B8]\], and enzymes involved in DNA repair (*cf*. \[[@B9]\]). More surprisingly, this list also includes translation initiation factors and several ribosomal proteins. The latter were assumed to be infrequently transferred, but recent analyses reported them to be horizontally transferred among bacterial lineages \[[@B10],[@B11]\]. The initiation factor IF-2 in *Halobacterium* sp. was previously shown to have strong similarity to the initiation factor IF-2 from Bacteria \[[@B49]\]. Most of the genes that group *Halobacterium* with the mesophilic bacteria encode functions that were postulated to be frequently exchanged \[[@B50]\]. While no meta-category appears exempt from HGT, some functions appear to be more often transferred than others (cf. Table [4](#T4){ref-type="table"}). ::: {#T5 .table-wrap} ::: {.caption} ###### List of genes putatively horizontally transferred between *Halobacterium* sp. (*H.* sp.) and the mesophilic Bacteria *Synechocystis* sp. and *Bacillus subtilis* (\"Information Storage and Processing\" meta-category only). ::: ***Protein Name*** ***H. sp. GI number*** ------------------------------------------------------------------------------ ---------------------------------------------------------- tRNA synthetases for serine, valine, methionine, cysteine, arginine, proline 10581491, 10581937,10579953, 10580644,10584349, 10580016 phenylalanyl-tRNA synthetase subunit alpha 10581896 Glu-tRNA amidotransferase subunits A, B 10580435, 10579969 tRNA-pseudouridine synthase 10581191 dimethyladenosine transferase 10580702 DNA gyrase subunits A, B \[10580453, 10580452\] DNA helicase 10580995 excision nuclease ABC chains A, B, C (involved in DNA repair) 10582016, 10581796, 10581790 endonuclease V (involved in DNA repair) 10579981 DNA mismatch repair protein 10579807 Putative translation factor SUA5 10581723 Translation initiation factor eIF-2B subunit alpha 10581299 Initiation factor IF2 10581429 ribosomal proteins L1, L11, L3, S4 \[10580652, 10580653\], 10581159, 10580672 GI numbers in brackets correspond to genes in operons. This list is derived analyses of genome quartets \#11 and \#61. A complete list of all GI numbers for each QuartOP as well as the four definition lines is available in the supplementary material. ::: Conclusions =========== Maximum likelihood mapping is a useful tool for analyzing and depicting the mosaic nature of genomes. ML-mapping is much less conservative than other approaches of estimating Bayesian posterior probabilities. If ML-mapping is used as the only probability mapping tool, the overestimation of supporting probabilities has to be taken into consideration. A posterior probability of .99 calculated with ML-mapping often corresponds to a posterior probability of only .90. Many relationships among prokaryotes cannot be depicted by a tree-like pattern reflecting a core of rarely transferred genes. Rather prokaryotic genomes are mosaics where different parts have different evolutionary histories. However, HGT between divergent organisms has not erased all patterns of interphylum relationship. For example, the majority of QuartOPs group the cyanobacteria with the low GC Gram positives as sister phyla. Due to horizontal gene transfer even organisms from different domains living in the same or similar environments share more genes with each other than organisms with a similar degree of divergence that live in different environments. These interdomain horizontal transfers mainly concern proteins involved in nucleotide, carbohydrate and amino acid transport and metabolism; however, proteins that are part of the translation machinery or are involved in DNA repair appear to be transferred across domain boundaries as well. Materials and Methods ===================== Genome Data ----------- Completed genomes were retrieved from the NCBI\'s FTP site <> in the form of amino acid sequences encoded by open reading frames (ORFs) as identified in the annotated genomes. Mitochondrial genomes were obtained from the Organelle Genomes Page at NCBI <>. The genomes were formatted using the *formatdb* program from the stand-alone BLAST package, initially of version 2.0.11 and later of versions 2.1.2 and 2.2.1 as they were released \[[@B51]\]. All analyses were performed locally. Data Flow in Quartet Analyses ----------------------------- For each set of four genomes, BLAST \[[@B51],[@B52]\] searches of every ORF in one genome against the other three genomes were performed using the *blastp* program. The E-value cutoff for the BLAST searches was set to 10^-4^ (in one test case an E-value cutoff of 10^-20^ was used). For every BLAST search the GI number of the top hit (if it was below the cutoff) was saved along with the GI number of the query sequence forming a GI pair. This resulted in twelve lists of GI pairs for each of the twelve possible pairwise genome comparisons. This information was further used to identify quartets of orthologous proteins (QuartOPs). Following Tatusov et al. \[[@B27]\] we defined QuartOPs as those sets of genes that mutually pick each other as the top scoring hit in the BLAST comparisons. The detection of the QuartOPs was performed using the MySQL database software <>. The lists of GI pairs were entered into twelve tables of a database. The tables were joined into one table under conditions that satisfy the definition of the QuartOPs (see above). This resulted in a table with four columns of GI numbers for QuartOPs. The amino acid sequences for each QuartOP were retrieved from GenBank at NCBI and were aligned using ClustalW 1.8 \[[@B53]\]. QuartOPs were analyzed using the ML-mapping approach according to Strimmer and von Haeseler, Bayesian probabilities mapping and bootstrap support values mapping techniques (see details below). Posterior probabilities according to Strimmer and von Haeseler -------------------------------------------------------------- For all three possible unrooted tree topologies maximum-likelihood values and posterior probabilities were calculated using in-house JAVA programs that were written utilizing classes from the Phylogenetic Analysis Library version 1.0 \[[@B54]\] and parts of Vanilla package version 1.0 \[[@B54]\]. If not indicated otherwise, likelihood values were estimated using the automatically selected suitable substitution model (chosen from BLOSUM62, CPREV, Dayhoff, JTT, MTREV24, VT and WAG) with no ASRV. The maximum-likelihood mapping approach was further used to visualize support for each tree topology \[[@B24]\], i.e. the posterior probability vector for each QuartOP was plotted into an equilateral triangle. Maximum-likelihood maps were generated using GNUPlot v. 3.7 <>. Posterior Probabilities calculated with MrBayes program ------------------------------------------------------- Posterior probabilities were also calculated with MrBayes version 2.01 \[[@B31]\]. Each QuartOP was analyzed with two simultaneous Markov chains for 25,000 cycles under the JTT substitution model \[[@B55]\] without ASRV. One chain was heated with the temperature set equal to the default value of 0.2. Samples were taken at each cycle. The \"burn in\" option was set to 5,000 cycles. The remaining 20,000 cycles were used to calculate posterior probabilities for each of the three tree topologies. The posterior probabilities were plotted to equilateral triangle as described above. For discussion of the choice of the parameters see below. Bootstrap Support Values ------------------------ As an alternative to posterior probability vectors, bootstrap support values were calculated and plotted. Each QuartOP was bootstrapped 100 times and the proportion of bootstrapped datasets supporting each tree topology was recorded as a bootstrap probability vector. The bootstrap probability vectors were plotted into an equilateral triangle with the zones changed to \"total\", \"70%\" and \"90%\" (see Fig. [2](#F2){ref-type="fig"}). Empirical Search for Optimal MrBayes Parameters ----------------------------------------------- To find parameters that will return consistent posterior probabilities within reasonable computation time, one QuartOP from mitochondrial genome quartet \#m1 was analyzed multiple times with different parameters. According to Strimmer and von Haeseler\'s approach \[[@B24]\] this QuartOP has posterior probabilities of 0.76, 0.10 and 0.13. In all runs samples were taken at each cycle; two chains and the JTT substitution model \[[@B55]\] without ASRV were used. First, we analyzed the dataset with 250,000 cycles. We tried different \"burn in\" options in the range of 1,000--20,000. The posterior probability values changed by less than 0.3% from case to case. We selected a \"burn in\" of 5,000 cycles in further analyses. Second, we tried different numbers of cycles to calculate posterior probabilities. The probabilities were calculated using 10,000--240,000 cycles with increment of 10,000 cycles. Again, the posterior probability values did not change significantly from case to case. Third, we raised the \"temperature\" parameter *T* to 2.0 for the second, heated chain. This did not result in changes of the estimated posterior probabilities. Fourth, we used 25,000 cycles and repeated the analysis 10 times, calculating average and standard deviation of all runs. For all three probabilities the standard deviation was less than 0.01. Based on these analyses we selected 25,000 cycles with a \"burn in\" of 5,000 as a compromise between precision of probability estimation and computational time spent. As a final test, we performed the analysis of the quartet \#8 twice with selected parameters. This did not result in significantly different maps. Graphs and tables depicting the results of these analyses are given in the supplementary material. Mapping taking ASRV into account -------------------------------- For the genome quartet \#8 we calculated posterior probabilities under the model which takes ASRV into account with Strimmer and von Haeseler\'s \[[@B24]\] approach and with the MrBayes program version 2.01 \[[@B31]\]. TREE-PUZZLE 5.0 \[[@B56]\] was used to calculate posterior probabilities according to Strimmer and von Haeseler \[[@B24]\]. A discrete approximation of the gamma distribution \[[@B57]\] was used to describe ASRV. Eight rate categories were used in TREE-PUZZLE \[[@B56]\], and four rate categories were used in MrBayes \[[@B31]\]. The maps are available in the supplementary material. Due to the amount of time required for calculations, the analyses were not performed for other genome quartets. Functional assignments using the COG database --------------------------------------------- Datasets for QuartOPs with strong preference for a particular tree topology (i.e. with posterior probability above 99% for that particular topology, or in other words the QuartOPs located in the very corners of the equilateral triangle) were extracted. For each of those QuartOPs the COG functional category \[[@B27]\] was identified. In order to detect the functional categories, the COG database was downloaded from NCBI\'s FTP site (initially the year 2000 release and later the year 2001 release). The COG database was formatted using the *formatdb* program of BLAST package. Every QuartOP was compared to the COG database using the *blastp* program. The category of the each sequence in the QuartOP was assigned according to the category of the top hit of each BLAST search. The numbers of QuartOPs in each functional category were calculated for each of the three tree topologies. Distribution of ORFs among COG categories for complete genomes of Halobacterium sp. and Archaeoglobus fulgidus -------------------------------------------------------------------------------------------------------------- Every predicted ORF in a genome was compared to the COG database (release of year 2001) using the *blastp* program with E-value cutoff 10^-4^. The category of each ORF was set to be equal to the category of the top hit of the corresponding BLAST search. Category Q was dropped from the results, because the corresponding genome quartets were analyzed with the previous release of the COG database (release of year 2000) that did not contain the Q category. Data Analysis Automation ------------------------ The repetitive tasks of analyses were automated using the SEALS package version 0.824 \[[@B25]\]. The tasks that were not available through SEALS package were programmed in PERL v. 5.005. The PERL scripts and JAVA programs are available upon request. Mitochondrial Genome Quartets Analyses -------------------------------------- Seven mitochondrial genome quartets were used as controls and were analyzed with the three approaches for genome quartet analysis described above. For calculation of posterior probabilities with MrBayes at least 25,000 cycles were used. List of Abbreviations ===================== **HGT** horizontal gene transfer **COG** cluster of orthologous groups **ML** maximum likelihood **rRNA** ribosomal ribonucleic acid **sp.** species **BLAST** Basic Local Alignment Search Tool **QuartOP** quartet of orthologous proteins **SEALS** System for Easy Analysis of Lots of Sequences **NCBI** National Center for Biotechnology Information **ASRV** Among Site Rate Variation Supplementary Material ====================== Supplementary material is located at the QuartOP web page <>. This web page includes the summary of all genome quartets analyzed (with maps), the results of control analyses, and a form to request the scripts described in this article. ML maps are available in postscript and PDF formats. An offline version of the QuartOP web page is available as a compressed archive named supp_material.zip and as a self-extracting archive supp_material.exe for Microsoft Windows users. The archive can be expanded using WinZip <> for Windows, StuffIt for Macintosh <>, or unzip utility for Unix. The uncompress utilities have to be run with the option to preserve the subdirectory structure inside the archive. To access the information in the archive, the file index.html has to be opened using an Internet browser. This index.html file is located in the root directory named \"offline_quartops\". All the files in the archive are hyperlinked and accessible through the index.html file. Supplementary Material ====================== ::: {.caption} ###### Additional file 1 ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 2 ::: ::: {.caption} ###### Click here for file ::: Acknowledgements ================ We thank Paul Lewis and Lorraine Olendzenski for many stimulating discussions and for critically reading the manuscript. The work was supported through the NASA Exobiology Program and through the NASA Astrobiology Institute at Arizona State University.

Background ========== Blood leukocytes are found in two sub-populations constituting the circulating and the marginating pools. The elements of these two sub-populations are interchangeable, i.e., marginated leukocytes return to bloodstream and vice-versa \[[@B1]-[@B3]\]. Therefore, a dynamical equilibrium situation occurs, and blood cell counts should reflect the rate constants of margination and return to circulation of those cells. Several studies addressed the relationships between the sub-populations of white cells within the blood pool. Among other features, the ratio between the sub-populations under normal conditions \[[@B4]\], under altered states \[[@B4],[@B5]\] and the stability of the equilibrium situation (i.e., to resume a previous value after perturbations, \[e.g., \[[@B6]-[@B8]\]) were approached. It is presently accepted that the blood sub-pools play an important role as a white cell reservoir when increased demands supervene. Increased demands arise in both pathological and non-pathological situations, as during exercise \[[@B9],[@B10]\], burns \[[@B11],[@B12]\], infectious diseases \[[@B13]\], inflammatory processes \[[@B5],[@B14]\], etc.. It is also currently recognized that some hormones acutely alter the dynamics of the blood sub-pools of white cells \[[@B6]-[@B8]\]. In this sense, epinephrine is known to possess a demarginating effect that lasts for less than 1 hour, and such an effect is thought to be the result of changes in the rate constants of margination and return \[[@B2],[@B6]\]. Therefore, the interplay between marginating and circulating leukocytes is a relevant issue that should be taken into account in the interpretation of many results. On the one hand, it is tacitly assumed that the rate constants of margination and return have higher values than other rate constants related to the dynamics of white blood cells (see below). On the other hand, there is no study addressing such an issue in a formal way in order to provide good estimation of these values. The aim of the present study is to provide the theoretical background to perform estimations of these rate constants, which may prove relevant for empirical studies on healthy normal situations as well as under altered states of the organisms. Constructing the extensive model -------------------------------- In this sub-section, we present a model based mainly on data from monocytes and neutrophils. The model describes the dynamics of three compartments of these cells and also contains the dynamics of specific growth factors. In the next sub-section we show a reduced model from this one, which will then be employed to obtain an improved estimation for the rate constants values. The extensive model is presented to assure consistence of the analysis. Bone marrow total rate production apparently has three components: a fixed production rate (P~1~), a production rate dependent on self-regulatory factors (P~2~) and a production rate dependent on inflammatory/infectious factors (P(I)) \[[@B3]\]. Therefore, the production rate is not a time independent function, and we denote it by the sum of the three terms above. Once the newly produced cell gains the bloodstream, it may marginate (rate constant M). A marginated cell may then return to circulation (rate constant R). Cells in the blood may leave to surrounding solid tissues, and they do that according to a time-independent rate constant S~B~\[[@B1]\]. Exit to tissues is also a function of rate constants dependent on inflammatory/infectious factors \[[@B3]\]. Thus, let us denote such a total exit rate constant from the circulation by S~B~ + S(t,I). Tissue leukocytes may proliferate locally (rate constant D), produce self-regulatory factors and, eventually, die (rate constant Z) \[[@B15]\]. This set of empirical data suggests the scheme presented in Figure [1](#F1){ref-type="fig"} and the following equation system for describing the dynamics of these cells: {#F1}  Where α are self-regulatory factors (e.g., CSF-1) and φ~i~ refers to cells that take part in the circulating (c), marginal (m) and solid-tissues (~T~) pools. G is a constant related to the production of self-regulatory factors, and K is the rate constant of their clearance. Notice that the exit to tissues comes both from the circulating pool, φ~c~, and from the marginated pool, φ~m~. Such an exit has the same rate constant, S~B~, independently of the sub-pool. This partition of the total exit constant rate to tissues in two components is the basic reasoning that leads to the inference about the existence of two distinct sub-pools within the blood white cells \[[@B1]\]. Therefore, a cell that touches, attaches and passes through the vessel wall cannot return to the circulating pool, even though during a certain time interval this cell was marginated (semantically but not functionally). In this sense, these cells exit to the surrounding tissues coming directly from the circulating pool, and their dynamics is contemplated by the product φ~c~S~B~ of the differential equation. Those marginated cells that can potentially return to circulation have their dynamics accounted for by the rate constant R. On the other hand, part of these marginated cells migrate to surrounding tissues and the product φ~m~S~B~ contemplates this rate. This dynamics is also consistent to the empirical evidence that margination and diapedesis are distinct features arising from different signals and receptors \[[@B16]\]. Reducing to the compact model ----------------------------- The injection of some substances, epinephrine in particular, changes the rate constants of margination and return to circulation (M and R) for very short periods of time (e.g., see \[[@B2],[@B6]-[@B8]\]). Under such a condition, the variable values change as the parameter values change. However, since the substance is rapidly metabolized, the parameters return to their previous value and then the variables would also return to some value maintained previously to the perturbation (in the parameters). Thereby, these substances act on the system described by equations 1a-d analogously to an impulsive function \[[@B17]\]. The returning phase is the relaxing period of the system, and it behaves as the perturbation had occurred in the variables instead of in the parameters. During the short time interval of the relaxing period, the production rate can be considered constant, P. Therefore, equations 1a and 1b maintain no loop-connection to the other two equations (1c and 1d), and a compact system can be written as:  Results ======= Analysis of the steady-state condition -------------------------------------- The first step in the analysis was to verify whether the extensive model allows the existence of a stable equilibrium point. An equilibrium point means a set of values of the variables (in this case, φ~c~, φ~m~, φ~T~ and α) that does not vary with time (if the system is left externally undisturbed). Stability is related to the behavior of the system in face of perturbations in the variables that displace them to some vicinity of an equilibrium point. The equilibrium point is said asymptotically stable if the system returns to the equilibrium point attained previously. Otherwise, the equilibrium point may be neutrally stable, if the system does not return to the equilibrium point but remains somewhere around it, or unstable, if the system leaves that vicinity away (e.g., \[[@B18]\]). We verified whether the extensive model posses one asymptotically stable equilibrium point as a first step because the biological system modeled seems to behave like this. Once disturbed, it returns to a previous value. The analysis revealed the existence of one equilibrium point, and such a point is asymptotically stable (see Methods). Let \* denote the value of the variable in the equilibrium point. Without loss of generality by considering the inflammatory/infectious factors absent and that φ~T~ and α are fixed (e.g., for a short time interval or changes in R or M, see Methods), the equilibrium values of the variables that we are interested in the present study are:  As stated before, blood cell counts should reflect the rate constants of margination and return. Let f^\*^ denote the ratio φ~m~^\*^/φ~c~^\*^ and thus:  Notice that one is not able to know, directly from the ratio φ~m~^\*^ /φ~c~^\*^, the values of R and M themselves. The demarginated state and its return to the equilibrium condition ------------------------------------------------------------------ In a very short time interval succeeding a bolus injection of some substances (e.g., epinephrine) the parameters R and M change and return to their previous values. Therefore, the circulating and marginal pools attain different values from φ~c~^\*^ and φ~m~^\*^, respectively. Given the stability of the equilibrium point (see above) these variables return to φ~c~^\*^ and φ~m~^\*^. This returning is governed by a second-order differential equation (see Methods). Let  be the value attained above (or below) the equilibrium value φ~c~^\*^ by the circulating cells after the perturbation (see Methods). In short,  (see equation 9). Therefore, taking the equilibrium value of the circulating pool as a reference value,  is the difference between this value and the value found at a time t after the beginning of the returning to the steady-state equilibrium condition. This leads to the following equation governing the fast-phase of resuming the steady-state condition:  where φ~c0~ is the peak (or nadir) value of the circulating pool attained after the perturbation, i.e., the initial value of this variable, and λ is the fast component of the process (λ~2~, see Methods equation 11b). Equation 5 can be linearized to simplify the estimation procedure. Finally, the λ of equation 5 is (see equation 11b): λ = -(R + S~B~)(f^\*^ + 1) This value of λ allows a good estimate of R during the fast phase of the returning to the equilibrium point (steady-state of the sub-pools). In the next section, we will work some examples of the application of the present model and its potential relevance. Worked examples --------------- The present study offers the means to compute the individual values of R and M. As stated before, the ratio f^\*^ is obtained by measuring the marginated and the circulating pools in steady-state conditions. This ratio is equal to M/(R+S~B~) (see equation 4). Therefore, its is important to note that the individual computation of R (and M) assumes that a series of other independent measurements were done: (a) φ~c~^\*^ (the steady-state value of the circulating pool); (b) φ~m~^\*^ (the steady-state value of the marginated pool); (c) S~B~ (the exit rate constant from the total blood pool to tissues). The knowledge of a general production rate value may also improve the picture, even though its is not a primary need. The first example intends to compute the R value for mice monocytes under normal conditions. However, the paucity of adequate data prevents a true computation. Therefore, we will compute an approximate value. The other two examples are completely imaginary. They show the usefulness of the model to address changes in the rate constants that would otherwise pass unnoticed. ### I. The normal condition The computations are based on data from van Furth and Sluiter \[[@B1]\]. The report give φ~c~^\*^ = 350 monocytes/mm^3^, φ~m~^\*^ = 500 monocytes/mm^3^ and S~B~ = 9.6 × 10^-4^ min^-1^. The f^\*^ value is 500/350 = 1.43. An intravenous bolus injection of epinephrine caused a two-fold increase in φ~c~ after 10 minutes of the injection. Therefore,  (or, φ~c~(0) = 700). After 60 minutes from the injection time (50 minutes after the monocytes peak), the variables resume their normal values. Notice the extremely limited temporal data. Let us suppose that we had a few more points, as illustrated in Figure [2A](#F2){ref-type="fig"}. From equation 5, the data could be linearized as: {#F2}  Where T is the discrete sampling time. Figure [2B](#F2){ref-type="fig"} shows the result of the linearization procedure. A linear regression would then result in λ = -0.113 min^-1^. Finally, from equation 6, we know that this value must be decomposed as -0.113 = -(R + S~B~)(1 + f^\*^) Thus, R = 0.0456 min^-1^, a value 47-fold greater than S~B~. In the data simulation, R = 0.045 min^-1^, therefore the procedure seems quite adequate. This example highlights the use of the present model as well as the difference between the eigenvalue (λ, obtained directly from the data) and the constant R. Also, as far as we know, this is the first attempt to estimate the value of R (and M) under a complete and formal model. ### II. Tumoral effects Consider, as another example, that mice carrying a certain tumor type have the exit constant rate equal to the normal value (S~B~ = 9.6 × 10^-4^ min^-1^). The f^\*^ value is 1.3 and these mice have φ~c~^\*^ = 490 monocytes/mm^3^, φ~m~^\*^ = 637 monocytes/mm^3^. Obviously, production rate have increased. Notice the potential confusing situation arising from these data. Does the tumor growth alter the dynamics between the blood pools? The epinephrine experiment (or some analogue) was conducted, and  (or, φ~c~(0) = 600). Figure [2C](#F2){ref-type="fig"} shows both the \"normal\" and the \"tumoral\" data. Linear regression of the log-transformed data computes λ = -0.065 min^-1^ (see Figure [2B](#F2){ref-type="fig"}) and thus R = 0.0273 min^-1^ since f^\*^ = 1.3. In data simulation, R = 0.027 min^-1^, again the results are in close agreement with the real parameter value. Therefore, in this constructed situation, one would be able to conclude that the rate constants of the transit between the blood pool were altered by the pathological condition. This could be an important result in tumor immunology, for example, allowing a better understanding of the pathology or devising new treatment strategies. ### III. Long Distance Running Effects Our final imaginary example is related to the effects of long sustained aerobic exercise. Consider that mice trained in this type of exercise are known to have increased production rate of monocytes. During the steady-state condition of the exercise, φ~c~^\*^ = 350 monocytes/mm^3^, φ~m~^\*^ = 700 monocytes/mm^3^, thus f^\*^ = 2. S~B~ increased to 8.5 × 10^-3^ min^-1^. An epinephrine experiment is performed during the exercise section and samples are taken during the decaying phase after the injection (the exercise section proceeds during the sampling period). The \"continuos\" count profile is illustrated in Figure [2C](#F2){ref-type="fig"}. Linear regression of the log-transformed data (see Figure [2B](#F2){ref-type="fig"}) computes λ = -0.041 min^-1^ and thus R = 0.0052 min^-1^ (in data simulation, R = 0.005 min^-1^). This is a 9-fold decrease in the R value in relationship to the normal condition. On the other hand, the rate constant M decreased less than 2.5 times in relationship to its value in normal conditions (see above). This result might be important in the understanding of many immune suppression/enhancement phenomena related to certain exercise protocols. Notice that the f^\*^ value by itself do not tell anything about the change in the individual value of each rate constant. Discussion ========== The compartmentalization of blood leukocytes in two sub-pools is an important feature of these cells. For example, the marginal pool may acutely function as an extra source of cells at increased demand conditions. In this sense, knowledge about the rate constants governing the transit between the sub-pools may prove relevant in studies that approach both physiological and pathological situations. In the present manuscript we provide the theoretical background to support empirical studies related to the calculation of these rate constants. From the general perspective of an extensive model, we first showed that this model corresponds to what is experimentally found in terms of existence and stability of an equilibrium point in the variables (the sub-pools). After that, we reduced the model to a compact one, concerning only the two blood sub-pools of interest here. Within the context of the compact model, it was shown how an impulsive function (e.g., an intravenous epinephrine bolus injection) perturbing the parameters (the rate constants of margination and return to circulation) can be translated to a perturbation in the variables. Then, the returning of the variables to their equilibrium values allows the estimation of the rate constants. This was done by the use of a second-order differential equation (see Methods). From that equation, the fast-decay component was identified and the corresponding eigenvalue contains the rate constants to be calculated. The present study demonstrates, for the first time, how to adequately and completely estimate the rate constants of margination and return to circulation of white blood cells. Notice that first-order differential equations (as equation 5) were employed before in gross estimates of these rate constants (e.g., \[[@B19]\]). However, the constant then computed comprised the whole eigenvalue, without discriminating its parts. In other words, its was not recognized that the computed value should be partitioned into three components: R, M and S~B~. The model also allows to establish the relationship between R and M. This relationship includes the S~B~ constant and the ratio between the marginal and the circulating sub-pools, f^\*^ (see equation 4). The need to take into account the f^\*^ value in the estimation of the constants is another important aspect for the first time presented. Therefore, the model leads to a more precise evaluation on how imposed disturbances interfere with the dynamics of the blood sub-compartments and thus increase our understanding of the physiology/pathophysiology of many conditions. Two imaginary situations of this kind were constructed and analyzed as examples of the use of the present model (see \"Worked Examples\" II and III). Conclusions =========== This study provides a complete model to approach the estimation of rate constants of margination and return to circulation of with blood cells. It shows, for the first time, how the value empirically found should be partitioned in order to adequately obtain the desired rate constants. In this sense, we were able to recognize that both an exit rate constant to tissues and the ratio between marginated and circulating cells should be taken into account in the computation procedure. Methods ======= 1. Equilibrium point and its stability -------------------------------------- The equilibrium point of the system described by equations 1a-d is found by setting all the derivatives to zero and computing the corresponding values of the variables that lead to such a condition. Without loss of generality, the inflammatory/infectious factors can be taken as zero. Therefore, the equilibrium point (denoted by \*) corresponds to the following values:  Conditions of existence: (1) Z \> D; and (2) K(Z-D) \> P~2~G. The first condition reflects that tissue leukocytes must die at a rate higher than their own local replication otherwise their population would increase forever. The second condition is similar to the first in the sense that it reflects that the self-stimulatory loop must be lower than the loss loop of the system. The stability of the equilibrium point is verified by the construction of the determinant of the matrix of the coefficients of the variables of the equations of the system \[[@B17],[@B18],[@B20]\]. In order to be asymptotically stable, all the eigenvalues (roots) of the characteristic polynomial of the matrix must have real parts lower than zero. Considering that the system has four equations, the characteristic polynomial is of the 4^th^-order with the following general formula: λ^4^ + a~1~λ^3^ + a~2~λ^2^ + a~3~λ + a~4~ = 0 where λ is an eigenvalue of the equation (note that the equation has 4 roots). The coefficients (a~i~) come from the parameters of the system. To have all the eigenvalues with real part lower than zero, the Routh-Hurwitz conditions must verify. In short: 1A. All a~i~ must be greater than zero. This condition is true for the system. 1B. a~1~a~2~ \> a~3~. This condition is true for the system. 1C. a~3~(a~1~a~2~ - a~3~) \> a~4~a~1~^2^. This condition is not easily verified analytically. Therefore, we performed a numerical analysis. Random values were assigned to the parameters (conditions of existence of the equilibrium point verified, see above) and the condition 1C checked (pseudo-random number generator normally distributed built-in function of MatLab 5.3, The MathWorks, Natick MA). This procedure was taken 150,000 times by a routine specifically written for it and the condition always verified true. Therefore, the equilibrium point seems asymptotically stable. The main point in this sub-section is to realize that both φ~T~^\*^ and α^\*^ are not affected by R and/or M. This is what assures the results concerning the compact model below. 2. The compact model and its second-order differential equation form -------------------------------------------------------------------- Considering that production rate and inflammatory/infectious factors would not change in a short time interval, the extensive model allows the blood pool to be treated as having no loop connection to the rest of the system (see equations 2a and 2b). The determinant of the Jacobian matrix of the compact model (i.e., the sub-system represented by equations 2a and 2b) is to set to zero:  Consider now a \"push\" in φ~c~^\*^ (and φ~m~^\*^) in such a way that:   is, therefore, the time course of the perturbation around the equilibrium point of φ~c~. Equation 8 is homogeneous for :  Taking into account that M = (R + S~B~)f^\*^ (see equation 4), the eigenvalues of equation 10 are:  Notice that: (a) both eigenvalues are pure real (this means that the equilibrium point is attained without oscillations in the variables); and (b) both eigenvalues are negative (the equilibrium point is asymptotically stable). Given that λ~2~ \< λ~1~, λ~2~ is the fast component of the process of resuming the equilibrium value φ~c~^\*^ (see sub-section \"the demarginated state and its return to the equilibrium condition\"). The value of λ~2~ is -(R + S~B~)(f^\*^ + 1). This value is presented in equation 6. List of abbreviations ===================== φ~c~: the number of cells (or concentration) in the circulating pool; φ~m~: the number of cells (or concentration) in the marginal pool; φ~T~: the number of cells (or concentration) in solid tissues; α: quantity (or concentration) of self-regulatory factors; R: rate constant of return to circulation; M: rate constant of margination; S~B~: rate constant of exit to tissues from the blood pool; P: bone marrow production rate of the cells; f^\*^: the ratio between the marginal and the circulating pools at steady-state condition. Acknowledgments =============== This study was supported by a FAPESP (State of São Paulo Science Foundation) undergraduate fellowship to K.I. (process number 00/01112-7) and by a FAPESP research grant to J.G.C.B. (process number 00/02287-5). This manuscript was significantly improved by the comments of two anonymous referees.

Background ========== Tetraethylammonium is known as a pore-occluding blocker of voltage-gated potassium channels \[[@B1]\], but it also blocks other cationic channels such as calcium-dependent K^+^ channels \[[@B2],[@B3]\] and the nicotinic acetylcholine receptor \[[@B4]\]. TEA at 0.1 to 10 mM also inhibits osmotic water flux through human AQP1 channels expressed in *Xenopus* oocytes, decreasing the net swelling rate in hypotonic saline by 30--40% as compared to AQP1-expressing oocytes not treated with TEA \[[@B5]\]. This blocking effect on osmotic water flux was demonstrated to involve AQP1 channels specifically by using site-directed mutagenesis (tyrosine 186 to phenylalanine) to generate a Y186F AQP1 channel that is insensitive to block by TEA, but retains sensitivity to block by mercury. The blocking effect of mercury in AQP1 channels is dependent on a neighboring residue, cysteine 189 \[[@B6]\]. TEA offers attractive advantages over mercury as a reversible and less toxic blocker for AQP1 channels in experimental analyses of water channel function; however, the relevance of TEA as a blocker for AQP1 channels outside the *Xenopus* expression system has not been examined previously. Differences in properties (including pharmacological sensitivities) have been observed for ion channels expressed in oocytes as compared to those in native tissues. For example, the *Torpedo* acetylcholine receptor and the *Shaker* K^+^ channel protein differ in the fractions of protein glycosylated, the composition of the oligosaccharide chains, and the degree of protein maturation when expressed in *Xenopus* oocytes as compared with channels in native tissue or other expression systems \[[@B7],[@B8]\]. Differences in glycosylation patterns can influence the binding of external blocking agents. The presence of tissue-specific targeting signals not recognized in *Xenopus* oocytes may lead to protein degradation \[[@B9]\]. These discrepancies raised the question of whether AQP1 channels in native tissues are sensitive to TEA, as they are in *Xenopus* oocytes. Data presented here show that TEA is effective as a blocker of AQP1 channels expressed in a mammalian renal cell line and in native renal epithelial membrane. AQP1 channels are complex solute conductors. They are constitutively permeable to water, and also function as regulated non-selective cation channels \[[@B10]\] when gated by intracellular cyclic GMP \[[@B11],[@B12]\]. Even though only a small proportion of the AQP1 channels that are present in the membrane appear to be available for cGMP-activation of the ionic conductance (1/56,000), model-based calculations nonetheless suggest that this contribution could be meaningful in a physiological setting \[[@B13]\]. Unlike transporters which move substantial amounts of substrate by design, ion channels function to change membrane potential by the net movement of relatively tiny amounts of charge (\~10 picomoles per cm^2^ membrane to generate a change of 100 mV) \[[@B14]\], hence the low proportion of active AQP1 ion channels is consistent with a functional goal of gated ionic signaling that is distinct from that of massive constitutive osmotic water permeability. The role of the AQP1 channel in mediating both water and ion fluxes raises interesting questions about whether the pathways are shared or structurally distinct. Expression of AQP1 channels in bilayers has shown that the water permeability but not the ionic conductance is blocked by p-chloromethylbenzenesulfonate, indicating that the two fluxes may involve distinct pathways \[[@B12]\]. Work presented here provides an additional line of evidence that the pathways for water and ions in AQP1 can be distinguished pharmacologically, using the blocker TEA as a probe. In combination with data from three-dimensional imaging analyses \[[@B15]\], these results support the suggestion that the AQP1 channel may contain separate parallel pathways for ion and water fluxes. These findings have potential significance in understanding the complex role for AQP1 in processes of salt and water movement across cell membranes. Results and Discussion ====================== TEA inhibition of water flux in AQP1-expressing MDCK cells ---------------------------------------------------------- TEA inhibited significantly the osmotically-driven net movement of water across monolayers of Madin-Darby Canine Kidney (MDCK) cells expressing human AQP1, as compared to AQP1-infected MDCK cells without TEA treatment (Figure [1](#F1){ref-type="fig"}). MDCK cells were induced to express AQP1 by infection with replication-deficient adenovirus (see Materials and Methods). These cells showed a high rate of water flux, ranging from 4 to 10 μl/cm^2^/hour, and a significant block by 1 mM TEA, ranging from 32 to 45% in 6 sets of replicate experiments. The TEA block of AQP1-mediated water flux for all data combined was 34.2%. In control MDCK cells infected with empty virus, the water flux was significantly lower (approximately 10-fold), ranging from 0.2 to 1.4 μl/cm^2^/hour. In these control cells, the net flux rates with and without treatment were not significantly different. Therefore, the enhanced water permeability and concomitant sensitivity to block by TEA were dependent on the expression of AQP1 channels. MDCK cells transfected with empty virus showed no augmentation of transepithelial water movement, confirming previous work that showed water does not flow across tight junctions in MDCK confluent layers \[[@B16]\]. Cells transfected with empty virus also showed no effect of TEA treatment. The magnitude of block of AQP1-mediated water permeability by TEA is consistent with that reported previously for AQP1 channels expressed in *Xenopus* oocytes \[[@B5]\]. {#F1} TEA inhibition of water flux in renal tubule -------------------------------------------- Preparations of pure descending thin limbs and pure ascending thin limbs from the inner medulla were segregated by morphological features. Descending thin limbs with their cells with nuclei protruding into the lumen differ markedly in appearance from ascending thin limbs with their large, round, flat nuclei when viewed with DIC optics \[[@B17],[@B18]\]. The differential expression of AQP1 in the two different types of segments (abundant in the descending thin limb and absent in the ascending thin limb), visualized by immunocytochemistry, confirmed the accuracy of the morphological identification (Figure [2](#F2){ref-type="fig"}). AQP1 provides the predominant pathway for water transport in the descending thin limb \[[@B19],[@B20]\]. In preparations of isolated perfused descending thin limbs, net osmotic water absorption rates (J~v~) and water permeability (P~f~) were significantly lower in the presence of 10 mM TEA than in the absence of TEA (Figure [3](#F3){ref-type="fig"}). This sensitivity to TEA correlates with the expression of AQP1 channels in this region. The overall effect of TEA in descending thin limb was to cause a 49% block of J~V~ and a 50% block of P~f~. The pairwise comparison of individual tubule preparations, before and after TEA, showed a significant percent change of (-) 52 ± 6 % (mean ± SEM; n = 10) in descending thin limb. In contrast, preparations of ascending thin limbs that lack AQP1 channels showed essentially no net water flux and no effect of TEA under osmotic gradient conditions. For data from ascending thin limbs, the pairwise comparison of individual tubule preparations before and after TEA showed a non-significant percent change of (+) 17 ± 35 % (n = 9). These data show that the blocking effect of TEA on water permeability originally observed in AQP1-expressing oocytes can be reproduced in a physiologically relevant condition. TEA blocks water permeability in a tissue that has abundant native expression of AQP1 channels, whereas it has no effect in tissue lacking AQP1 expression. {#F2} {#F3} Lack of TEA block of ion channel conductance -------------------------------------------- Heterologous expression of human AQP1 in *Xenopus* oocytes, a system that has been well characterized for electrophysiological assays, was used for two-electrode voltage clamp studies to evaluate the effects of TEA on ionic currents. External TEA(10 mM) failed to block the cGMP-dependent ionic current in AQP1-expressing oocytes (Figure [4](#F4){ref-type="fig"}). Current responses were assessed with voltage steps from +40 to -70 mV, from a holding potential of 0 mV (Fig. [4A](#F4){ref-type="fig"}). The initial response shows the current-voltage relationship recorded by two-electrode voltage clamp in an AQP1-expressing oocyte in isotonic 100 mM Na^+^ saline (see Methods). The activated response was measured in the same oocyte at 3--5 minutes after addition of 4 mM sodium nitroprusside (SNP), which elevates endogenous cGMP levels. The effect of 10 mM TEA (with 90 mM NaCl) was assessed by perfusing the recording chamber with TEA saline containing SNP. No appreciable change was observed. For the representative example shown in Figure [4](#F4){ref-type="fig"}, the conductance (measured as the linear slope of the current-voltage relationship between -50 and +30 mV; Fig. [4B](#F4){ref-type="fig"}) for the initial condition was 3.0 nA/mV. After channel activation the conductance was 58 nA/mV in Na^+^ saline and 57 nA/mV in 10 mM TEA saline. The reversal potential of the current (Fig. [4B](#F4){ref-type="fig"}) shifted from approximately -24 mV in the initial state to about -4 mV in the activated state, consistent with the activation of a non-selective cationic conductance permeable to both Na^+^ and K^+^. Similar results were observed for other SNP-activated oocytes (n = 5), as well as for AQP1-expressing oocytes activated by prior incubation in 10 mM 8Br-cGMP (as described in \[[@B11]\]). In AQP1-expressing oocytes, the conductance induced by 8Br-cGMP in control Na^+^ saline did not change significantly \[(+) 11.6 ± 6.2 % (mean ± SD; n = 7)\] after perfusion with 10 mM TEA saline, thus confirming the absence of a blocking effect of TEA on the ion conductance by AQP1 channels. The activated AQP1 ion channel is less permeant to TEA than to Na^+^ or K^+^\[[@B10]\]; however, because of the high extracellular Na^+^ concentration (90 mM) in the TEA bath saline in the present study, the measured reversal potential of the activated current was not appreciably different with and without 10 mM TEA. {#F4} Results presented here show that TEA serves as an effective, though partial, blocker of the osmotic water permeability mediated by human AQP1 channels expressed in a heterologous mammalian cell line and by native AQP1 channels in isolated rat renal descending thin limbs of Henle\'s loops. Before the characterization of TEA, mercurial compounds were among the only known blockers of AQP1, affecting both the constitutive water flux and the regulated ion conductance \[[@B6],[@B10]\]. For AQP1 channels expressed in *Xenopus* oocytes, the block of osmotic water flux is 5% in 0.01 mM TEA, 21% in 1 mM TEA, and 36% in 10 mM TEA, as referenced to the osmotic water flux in control AQP1-expressing oocytes \[[@B5]\]. Data presented here show that the level of block by TEA is comparable or greater for AQP1 channels expressed in mammalian cells. 1 mM TEA produces a 34% block of osmotic water flux in MDCK cells, and 10 mM TEA produces a 50% block in renal descending thin limbs. The effect of TEA on water permeability in cells and tissues that express high levels of AQP1 channels is likely to be significant, given the substantial amount of water that can be moved across AQP1-expressing epithelia (for example, see data on descending thin limbs in Fig. [3](#F3){ref-type="fig"}). The idea that TEA might be a candidate for a lead compound for the development of drugs with possible clinical applications \[[@B5]\] is supported by the present data that show the effectiveness of TEA in two different mammalian models of AQP1-mediated osmotic water permeability. While the usefulness of TEA *in vivo* as a research tool may be limited by toxicity and effects on other channels at higher concentrations, it offers a degree of reversibility and selectivity not seen with mercury, an agent which acts by covalent modification of all exposed cysteine residues \[[@B21],[@B22]\]. The regulated ion channel function of AQP1 is not blocked by external TEA. An intriguing question remains regarding the location of the permeation pathway for cations in AQP1 channels. Analysis of the crystal structure of GlpF suggests that potential ion binding sites may be located in the central pore, at the four-fold axis of symmetry in the tetrameric association of subunits that form the channel \[[@B15]\]. Interestingly, this arrangement of a tetramer of subunits to form a central ion pore is a hallmark of the family of voltage-gated ion channels \[[@B23]\], which rely on a tetrameric organization to create a central pore for ions that is lined by sequences contributed from all four subunits. In contrast, the four individual pores for water or glycerol, located within each subunit of AQP1 or GlpF, appear to be lined with hydrophobic residues expected to preclude ion permeability \[[@B15],[@B24],[@B25]\]. Importantly, TEA distinguishes between the two pathways in AQP1, supporting the idea that the ion and water pores are physically distinct although they co-exist in the same channel complex. The AQP1-mediated ion conductance may contribute measurably to the function of tissues in which AQP1 is abundantly expressed, such as in the mammalian renal proximal tubule \[[@B13]\]. Further work is needed to assess the additional levels of regulatory control that may constrain the availability of the AQP1 proteins to serve as ion channels gated by cGMP, within a background of constitutive water channel activity. The maintenance of an ionic pore, as well as a conserved cGMP-binding domain in the AQP1 carboxy tail for regulating activation \[[@B11]\], suggests that the ion channel function is likely to be essential for physiological regulatory processes that have yet to be fully appreciated. Materials and Methods ===================== MDCK cells ---------- ### Cell culture Madin-Darby Canine Kidney (MDCK) cells were obtained as a gift from Dr. R. Lynch, University of Arizona, and used between passages 13--14. Cells were grown and maintained in Dulbecco\'s Modified Eagle Medium (Life Technologies) that was supplemented with 10% fetal bovine serum (Hyclone), 170 mM glutamine and penicillin/streptomycin (100 μg/ml, Life Technologies) in humidified air containing 5% carbon dioxide at 37°C. ### Adenovirus vector The adenovirus (AV) backbone for the aquaporin-1 (AQP1) sense and antisense AV constructs was a replication-deficient \"first-generation\" AV with deletions of the E1 and E3 genes \[[@B26]\]. This \"empty\" AV contains the cytomegalovirus (CMV) promoter and bovine growth hormone polyadenylation (bHG) site separated by a polylinker that was used to clone AQP1 DNA as described previously \[[@B27]\]. A recombinant AQP1 virus was constructed using a plasmid containing the coding sequence for AQP1, pCHIPev \[[@B28]\]. pCHIPev was digested and the AQP1 insert was subcloned into the shuttle vector pSKAC, creating pSKAC/AQP1. pSKAC contains map units 0.0--1.3 of the AV which includes the left terminal repeat of AV, a CMV promoter, an AMV translation enhancer and a polylinker region. DNA fragments containing AQP1 DNA were liberated from pSKAC after restriction and ligated into adenovirus as described previously \[[@B27]\]. Human embryonic kidney cells (293 cells) were transfected with ligation mixture and individual viruses were isolated from cell lysates by two consecutive rounds of plaque purification using an agar overlay as described previously \[[@B27]\]. Individual viruses were amplified in 293 cells and purified over cesium step gradient. Individual AV DNA titers were determined by three different methods: 1) plaque titration on 293 cells, 2) immunofluorescence microscopy of AV protein expression (anti-penton group antigen, clone 143, Biodesign, Kennebunk, ME) in 293 cells infected with serial dilutions of AV and 3) absorbance at 260 nm (pfu/ml = A~260~ × dilution × 10^10^). ### Osmotic water flux assay Water permeability was measured as the net fluid movement driven by an osmotic gradient across intact monolayers of adenovirus-infected MDCK cells, by methods similar to those reported previously \[[@B27],[@B29]\]. Cells were seeded onto Transwell filters (Costar, 1 cm^2^, 0.4 μm pore size) at a density of 1.5 × 10^5^ cells per well. Cells were maintained in humidified air containing 5% CO~2~ for two weeks to allow for cell-cell junctions to mature. Cells were infected at the apical surface with adenovirus containing AQP1 cDNA or with empty adenovirus at a multiplicity of infection (MOI) of 10. Five days post-infection, all medium was removed carefully and completely from both upper (apical) and lower (basolateral) chambers. Exactly one milliliter of fresh prewarmed isotonic medium (\~300 mOsM) was added to the lower chamber and exactly 175 μl of hypertonic medium (\~450 mOsM) was added to the upper chamber at time zero. In experimental groups, TEA chloride was present at a 1 mM concentration in both the upper and lower chambers. After four hours of incubation at 37°C, all of the medium from each of the upper chambers was removed carefully and volume was measured using an analytical balance. ### Immunoblot analysis Sodium dodecyl sulfate (SDS) solubilized whole cell lysates containing 5% β-mercaptoethanol were electrophoresed into 12% polyacrylamide gels containing 0.1 % SDS. Fractionated proteins were blotted onto nitrocellulose using the Transblot system as per manufacturer instructions (Biorad, Hercules, CA). The blots were preincubated for 30 minutes at 22°C in Tris-buffered saline containing 5% nonfat powdered milk and 0.2% Tween-20 (TBS-T), and were then probed with affinity purified anti-AQP1 IgG (1:5000) for 2 hr at 22°C. The blots were washed (3 × 15 min) in TBS-T and were incubated for 2 hours with horseradish peroxidase-conjugated secondary antibodies (goat anti-rabbit, 1:5000, Pierce). The blots were washed (3 × 15 min) in TBS-T and specific labeling was visualized following enhanced chemiluminescence (Pierce, Rockford, IL) and exposure (10 seconds) to ECL-Hyperfilm (Amersham, Arlington Heights, IL). Immunoblots were digitized using the UVP gel documentation system and densitometry was performed using LabWorks software (Upland, CA). Isolated perfused rat inner medullary thin limbs ------------------------------------------------ ### Animals and dissection of tubules for immunocytochemistry and for in vitro fluid movement measurements Young male Munich-Wistar rats (Harlan Sprague-Dawley, Indianapolis, In), 90 g average body weight, were maintained on Teklad Mouse/Rat Diet No. 7001 with free access to water. Dissection of inner medullary thin limbs of Henle\'s loop from fresh rat renal tissue for immunocytochemistry and for *in vitro* microperfusion was performed as described in detail previously \[[@B17]\]. We used descending thin limb (DTL) and ascending thin limb (ATL) segments from the top 70% of the inner medulla above the pre-bend level. Descending and ascending thin limbs were identified by cell type using an inverted microscope with Nomarski differential interference contrast (DIC) optics at 400× magnification \[[@B17],[@B18]\]. ### Immunocytochemisty Rabbit antibody recognizing the carboxy tail domain of AQP1 (from W.D. Stamer and J.W. Regan, University of Arizona) was used for immunocytochemical confirmation of protein expression in isolated thin limbs of Henle as described previously \[[@B30]\]. Tubules were positioned onto a glass microscope slide covered with a layer of Cell-Tak^®^ adhesive (Becton Dickinson, Bedford, MA). Tubules were then fixed in 4% paraformaldehyde for 10 min and permeabilized with 100% methanol for 2 min at -20°C. Primary antibody or non-immune rabbit serum (diluted 1:500) was applied overnight at 4°C., followed by incubation with biotinylated goat anti-rabbit antibody (diluted 1:100, 60 min), and incubation with streptavidin conjugated to Cy5 (60 min). Digital fluorescent images were obtained with a Leica-TCS confocal microscope. ### Perfusion and bathing solution for in vitro microperfusions The Ringer solution used for perfusing and bathing the tubules was that initially used by Chou and Knepper \[[@B18]\] in perfusion of thin limbs from chinchilla kidneys and modified by us. It contained (in mM): 130 NaCl, 20 HEPES, 5 NaHCO~3~, 2.5 K~2~HPO~4~, 2 CaCl~2~, 1.2 MgSO~4~, 5.5 glucose, and 5 urea. The osmolality was about 290 mOsm/kg H~2~O, the pH was adjusted to 7.4 while the solution was gassed with 95 O~2~/5% CO~2~. ### Perfusion of tubules in vitro We perfused the dissected tubules *in vitro* by a technique essentially the same as that originally described by Burg et al. \[[@B31]\] and modified for use in our laboratory \[[@B17],[@B32],[@B33]\]. The perfusion rate was about 15--20 nl/min. The temperature of the bathing chamber was maintained at 37°C during perfusion and the bath was covered with a layer of light paraffin oil. To check for leaks and to measure net fluid movement across the epithelium with an imposed osmotic gradient, \[^14^C\]dextran (MW \~70,000) was added to the perfusate. ### Determination of net water movement and water permeability of thin limbs perfused in vitro There is no net transepithelial water movement with solutions of identical osmolalities in the bath and lumen. For this study, however, we determined the occurrence of net water absorption and water permeability when we imposed an osmotic gradient from bath to lumen (see below). Net water absorption, J~v~ (nl min^-1^ mm tubule length^-1^), was determined in each collection period with \[^14^C\]dextran in the perfusate using the following relationship: J~v~= (V~i~ - V~o~)/L. In this equation, V~i~(initial perfusion rate, nl/min) is calculated by dividing the cpm of ^14^C in the collected fluid by the cpm/nl of ^14^C in the initial perfusate and by the time of the collection period; V~o~ (collection rate, nl/min) is measured directly from the collection; and L (mm) is the length of the tubule perfused, measured with an ocular micrometer. For purposes of determining the water permeability of DTL and ATL, we imposed a 100 mOsmol/kg H~2~O osmotic gradient from bath to lumen with sucrose and determined J~v~ as just described. Water permeability (P~f~, μm s^-1^) was then determined from the equation P~f~ = J~v~/(A~s~ V~w~ δC~osmol~) where A~s~ = luminal surface area (π DL), V~w~ = partial molar volume of water (18 ml/M), δC~osmol~ = transepithelial osmolality gradient \[[@B18]\]. When the effect of TEA on water movement was examined, TEA bromide (10 mM) was added to the bathing medium and the NaCl concentration was reduced by 10 mM to maintain the osmolality constant. Collection periods were 5 min in length and each tubule served as its own control for measuring net water flux and water permeability in the presence and absence of TEA. \[^14^C\]dextran (sp act 1.08 mCi/g) was obtained from American Radiolabeled Chemicals, St. Louis, MO., USA. Results are summarized as means ± SE. The *n* value is the number of tubules. Each tubule came from a different animal. Xenopus oocytes --------------- ### Oocyte preparation Adult female *Xenopus laevis* were anesthetized with tricaine methane sulfonate (MS-222, Sigma Chemical Co). Stage V-VI oocytes were removed by partial laparotomy and defolliculated by collagenase treatment \[[@B11]\]. Cloned human Aquaporin1 DNA was provided by Dr. P. Agre \[[@B28]\], linearized with XbaI and used as a template for *in vitro* synthesis of cRNA with T3 RNA polymerase. Prepared oocytes were injected with 50 nl of sterile water (control oocytes) or 50 nl of sterile water containing AQP1 cRNA (\~0.2 to 0.5 ng/nl) and were incubated for 2--5 days at 18°C in culture medium (ND96: 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl~2~, 1 mM MgCl~2~, 5 mM Hepes, 2.5 mM pyruvic acid, 100 U/ml penicillin, and 100 μg/ml streptomycin, pH 7.6) to allow protein expression. ### Electrophysiology Two-electrode voltage clamp recordings were performed at room temperature with electrodes (0.5--3 MΩ) filled with 3 M KCl. Data were recorded with a GeneClamp 500 (Axon Instruments, Foster City, CA), filtered at 2 kHz, digitized at 50 to 2000 μs and analyzed with pClamp software (Axon Instruments). Control recording saline for two-electrode voltage clamp contained (in mM): 100 NaCl, 5 MgCl~2~, and 5 HEPES, pH 7.3. TEA saline was made with (in mM) 10 TEA.Cl, 90 mM NaCl, 5 MgCl~2~, and 5 HEPES, pH 7.3. Activation of the ionic current was induced by preincubation of AQP1-expressing oocytes for 10--30 minutes in 10 mM 8Br-cGMP (Sigma) in high K^+^ saline, containing (in mM): 30 KCl, 80 K gluconate, 5 MgCl~2~, 5 HEPES, pH 7.3, or by addition of 2--4 mM sodium nitroprusside (SNP), a stimulator of guanylate cyclase and cGMP production \[[@B34],[@B35]\]. Control oocytes treated in the same conditions showed no induction of an ionic current. Higher concentrations of SNP (\>10 mM) affected ionic permeability in some batches of control oocytes and were not used. Current voltage relationships were assessed in Na^+^ control saline and in TEA saline to assess potential blocking effects of TEA. ### Chemicals All chemicals were purchased from standard sources except as specified above and were of the highest purity available. List of abbreviations ===================== AQP1 Aquaporin-1 J~v~ Net osmotic water absorption rate MDCK Madin-Darby Canine Kidney cell P~f~ Osmotic water permeability SNP Sodium nitroprusside (nitric oxide donor) TEA Tetraethylammonium (chloride salt) Acknowledgements ================ We thank Dr. John Regan for antibodies to AQP1, Amy Marble and Dr. Kathryn Bolles for technical assistance, Dr. Heddwen Brooks for helpful discussions, Neil Atodaria and Eileen Ryan for technical assistance with permeability assays using MDCK cells, and Dr. Ron Lynch for the gift of MDCK cells (University of Arizona). Funding was provided by NIH R01 GM59986 (AJY), American Health Assistance Foundation \#G2001-026 (WDS), and NIH R01 DK16294 (WHD) and Research to Prevent Blindness Foundation.

Background ========== The extracellular signal regulated kinases ERK1 (p44^mapk^) and ERK2 (p42^mapk^) are activated in response to stimulation of receptor tyrosine kinases (RTKs) as well as heptahelical G protein coupled receptors (GPCR) and transmit signals which regulate cell differentiation and growth \[[@B1]-[@B3]\]. The molecular steps involved in signaling from GPCRs to ERK are incompletely understood. Data obtained in various cell systems have provided evidence in support of several signaling pathways including protein kinase C (PKC) \[[@B4]\], Ca^2+^-mediated mechanisms \[[@B5]-[@B12]\], and transactivation of receptor tyrosine kinases \[[@B13],[@B14]\]. In hepatocytes several hormones, including vasopressin, angiotensin II, norepinephrine, and PGF~2α~, that bind to GPCRs activate ERK \[[@B15]-[@B17]\]. The mechanisms mediating the ERK activation by GPCR agonists are not clarified; there is evidence that protein kinase C is involved \[[@B15],[@B18]\], but a role for Ca^2+^ also appears likely, since all the agents above activate phospholipase C and elevate intracellular Ca^2+^ in hepatocytes \[[@B19],[@B20]\]. Furthermore, agents that elevate intracellular Ca^2+^ through mechanisms bypassing receptors have been found to activate ERK \[[@B15],[@B21]\]. However, agonist-stimulated phospholipase C activity is rapidly down-regulated upon culturing of hepatocytes \[[@B22],[@B23]\], and we recently reported that norepinephrine and PGF~2α~ activate ERK under conditions where the level of inositol 1,4,5-trisphosphate (InsP~3~) was only slightly, and transiently elevated \[[@B17]\]. In the present study we have, therefore, examined more closely the role of Ca^2+^ in ERK activation induced by norepinephrine and PGF~2α~ and mechanisms downstream of elevated \[Ca^2+^\]~i~. Results ======= Agents that elevate \[Ca^2+^\]~i~ activate ERK ---------------------------------------------- In agreement with previous observations \[[@B15],[@B21]\] treatment of hepatocytes with thapsigargin, which inhibits Ca^2+^ reuptake to endoplasmatic reticulum \[[@B24]\], and A23187, which induces Ca^2+^ influx, stimulated ERK1/2 activity 2--2.5 fold (Fig. [1A](#F1){ref-type="fig"}). The elevation of intracellular Ca^2+^ resulting from stimulation with thapsigargin is shown in Fig. [1B](#F1){ref-type="fig"}. These observations are compatible with a role for Ca^2+^-elevating mechanisms in the events that trigger ERK1/2 activation in hepatocytes. ![ERK1/2 activation and Ca^2+^ response in hepatocytes. A: At 3 h after the time of seeding hepatocytes were preincubated with timolol (10 μM) for 30 min prior to stimulation with thapsigargin (1 μM), A23187 (10 μM) or norepinephrine (10 μM) for 5 min before they were harvested and ERK 1/2 activity assessed. Results represent mean ± S.E.M. of five different experiments. B: Single cell measurement of \[Ca^2+^\]~i~ as described in Materials and Methods. Results given as ratio (345/385 fluorescence) represent a typical single cell response after stimulation with thapsigargin (10 μM) in a fura-2 AM loaded hepatocyte.](1471-2121-3-5-1){#F1} Activation of ERK by norepinephrine and PGF~2α~ involves Ca^2+^ --------------------------------------------------------------- We then examined the role of Ca^2+^ in activation of ERK1/2 induced by stimulation of α~1~-adrenoceptors (with norepinephrine in the presence of timolol) and prostaglandin receptors (using PGF~2α~) \[[@B21],[@B25],[@B26]\]. The hepatocytes were pretreated with BAPTA-AM, which is activated intracellularly to bind Ca^2+^, EGTA, which binds extracellular Ca^2+^ and eventually may deplete intracellular Ca^2+^\[[@B27],[@B28]\], or gadolinium, a competitive inhibitor of Ca^2+^ influx \[[@B29]-[@B31]\]. BAPTA-AM completely attenuated the norepinephrine-induced rise of \[Ca^2+^\]~i~ (Fig. [2A](#F2){ref-type="fig"}), while the ERK1/2 activity in response to norepinephrine was partially decreased (Fig. [2B,2C](#F2){ref-type="fig"}). ERK1/2 activity induced by PGF~2α~ and the Ca^2+^ ionophore A23187 was also inhibited by BAPTA-AM, while the TPA response was unaffected (Fig. [2B,2C,2D](#F2){ref-type="fig"}). When the cells were pretreated with EGTA, the initial peak of the Ca^2+^ elevation was only slightly affected, while the prolonged phase of the Ca^2+^-response was abolished (Fig. [3A](#F3){ref-type="fig"}). The activation of ERK1/2 by norepinephrine or PGF~2α~ was partly decreased by EGTA (Fig. [3B,3C,3D](#F3){ref-type="fig"}). EGTA also markedly decreased the ERK1/2 response induced by A23187 and thapsigargin, while the TPA-induced ERK1/2 activation was unaffected (Fig. [3B,3C](#F3){ref-type="fig"}). Pretreatment with gadolinium decreased the adrenergic activation almost to the level obtained by EGTA (Fig. [4A](#F4){ref-type="fig"}). Gadolinium also decreased the A23187-induced activation of ERK1/2 (Fig. [4B](#F4){ref-type="fig"}). Taken together, the results suggest a role for Ca^2+^ in the activation of ERK by norepinephrine and PGF~2α~ and that this involves Ca^2+^ influx as well as release from internal pools. ![Effect of BAPTA-AM on \[Ca^2+^\]~i~ and ERK1/2 activation. A: Measurement of \[Ca^2+^\]~i~. Hepatocytes were preincubated with 0.55 % DMSO or BAPTA-AM (40 μM) during the last 25 minutes of the fura-2 AM loading. After 60 seconds of registration the cells were stimulated with norepinephrine (10 μM) in the presence of timolol (10 μM). Results show a typical single cell response. B-D: ERK1/2 responses. Hepatocytes cultured for 3 h were pretreated for 30 min with BAPTA-AM (40 μM) in the presence of timolol (10 μM) prior to stimulation with norepinephrine (10 μM), A23187 (10 μM), TPA (1 μM) or PGF~2α~ (10 μM) for 5 min before cells were harvested. All cultures contained DMSO at a concentration of 0.5 % during the preincubation and a final concentration of 1 % DMSO during incubation with agonist. B: Activity measurements of ERK1/2 representing mean ± S.E.M. of three experiments. C, D: Immunoblots using antibody against dually phosphorylated ERK1/2.](1471-2121-3-5-2){#F2} ![Effect of EGTA on \[Ca^2+^\]~i~ and ERK 1/2 activation. A: \[Ca^2+^\]~i~ measurements. Hepatocytes were preincubated in Krebs-Ringer-Hepes buffer with or without 5 mM EGTA for 15 min after fura-2 AM loading. After 60 seconds of registration the cells were stimulated with norepinephrine (10 μM) in the presence of timolol (10 μM). Results show a typical single cell response. B-D: ERK1/2 responses. Hepatocytes cultured for 3 h were pretreated with timolol (10 μM) for 30 min and EGTA (5 mM) for 15 min before stimulation with norepinephrine (10 μM), TPA (1 μM), thapsigargin (1 μM), A23187 (10 μM) or PGF~2α~ (10 μM) for 5 min (in the presence of 0.5 % DMSO). B: Activity measurements of ERK1/2 representing the mean ± S.E.M. of three experiments. C, D: Immunoblots using antibody against double phosphorylated, i.e. activated, forms of ERK1/2.](1471-2121-3-5-3){#F3} {#F4} Effect of antagonists of calmodulin and the multifunctional Ca^2+^/calmodulin-dependent protein kinase in ERK activation in hepatocytes --------------------------------------------------------------------------------------------------------------------------------------- A major mechanism for Ca^2+^-induced signaling is through formation of a complex with calmodulin \[[@B32],[@B33]\]. Calmodulin has been found to stimulate as well as inhibit ERK1/2 activity \[[@B12],[@B34],[@B35]\]. We therefore examined the role of calmodulin in these pathways. Pretreatment of hepatocytes with the calmodulin inhibitors trifluoperazine, J-8, and W-7 markedly inhibited the ERK1/2 activation after stimulation with norepinephrine and PGF~2α~ (Fig. [5](#F5){ref-type="fig"}). The results were confirmed with immunoblots (Fig. [6](#F6){ref-type="fig"}). Activation of ERK1/2 by A23187 was also markedly sensitive to pretreatment with W-7 (Fig. [5](#F5){ref-type="fig"}, [6](#F6){ref-type="fig"}). {#F5} {#F6} Calmodulin may act on several regulatory enzymes \[[@B32],[@B36]-[@B40]\], including the Ca^2+^/calmodulin-dependent protein kinases, which have been implicated in the activation of ERK1/2 \[[@B7]-[@B9]\]. We explored a possible role for the multifunctional Ca^2+^/calmodulin-dependent protein kinase in ERK1/2 activation in hepatocytes stimulated by norepinephrine and PGF~2α~. Pretreatment of the cells with KN-93, an inhibitor of the multifunctional Ca^2+^/calmodulin-dependent protein kinase \[[@B41]\], did not decrease the activation of ERK1/2 either by norepinephrine, PGF~2α~, or A23187 (Fig. [5](#F5){ref-type="fig"}, [6](#F6){ref-type="fig"}). In supplementary experiments we examined the effects of higher concentrations of KN-93 (up to 100 μM) or prolonged exposure times (up to 24 hours) which in none of the cases resulted in a decreased ERK1/2 activation (data not shown). These results suggest that calmodulin might be involved in hormone-induced activation of ERK1/2 in hepatocytes, however the data do not support a role for the multifunctional Ca^2+^/calmodulin-dependent protein kinase in Ca^2+^/calmodulin-mediated activation of ERK1/2 in hepatocytes Inhibitors of src kinases attenuate ERK activation in hepatocytes ----------------------------------------------------------------- Src kinases \[[@B42]\] have been implicated in the mechanisms resulting in ERK1/2 activation in response to stimulation of both G~i~- and G~q~- coupled heptahelical receptors \[[@B43]-[@B45]\], and several observations suggest that activation of Src in these pathways involves Ca^2+^\[[@B46],[@B47]\]. Data obtained in this study showed that the Src inhibitors PP1 and PP2, which are reported to primarily inhibit the Lck, Fyn, and Hck subtypes of Src kinases \[[@B48]\], markedly decreased the PGF~2α~-induced ERK1/2 activation and led to partial inhibition of the effect of norepinephrine, while the EGF induced ERK1/2 response was not reduced (Fig. [7](#F7){ref-type="fig"}). Furthermore, ERK1/2 activation induced by A23187 and thapsigargin was also decreased after Src inhibition (Fig. [7](#F7){ref-type="fig"}). The results suggest a role for Src kinases in the mechanisms leading to ERK1/2 activation both by PGF~2α~ and norepinephrine, and that this step at least in part may be located distal to increases in the intracellular level of Ca^2+^. {#F7} Discussion ========== The present findings confirm previous reports of a role for Ca^2+^ in ERK1/2 activation in hepatocytes \[[@B15],[@B21]\] and suggest that release of Ca^2+^ from intracellular stores as well as influx of extracellular Ca^2+^ is of importance for the hormone-induced activation of ERK1/2. Furthermore, the results suggest that calmodulin and Src kinases might be involved in the Ca^2+^-dependent activation of ERK1/ERK2. Evidence from several experimental models suggest that activation of ERK1/2 may occur through Ca^2+^-dependent as well as Ca^2+^-independent mechanisms \[[@B5],[@B28],[@B49]-[@B54]\]. The present data suggest that Ca^2+^ is involved in activation of ERK1/2 in hepatocytes in response to norepinephrine and PGF~2α~. The ERK1/2 response was decreased by chelation of intracellular and extracellular Ca^2+^ with BAPTA-AM and EGTA, respectively, as well as by gadolinium, which competitively inhibits Ca^2+^ influx. It may appear that extracellular and intracellular Ca^2+^ act in a concerted, possibly sequential manner in the mechanisms involved in activation of ERK1/2 by norepinephrine and PGF~2α~. An integration of Ca^2+^ signals from the extracellular and intracellular environment is presumably due to store-operated Ca^2+^ influx \[[@B31],[@B55],[@B56]\]. The mechanisms that initiate Ca^2+^ influx subsequent to depletion of intracellular stores are incompletely understood, but recent studies have suggested that direct interaction between InsP~3~ receptors and calcium channels in the plasma membrane may lead to activation of the calcium channels \[[@B57]\]. A diffusible Ca^2+^ influx factor may also be involved \[[@B58]\]. Previous studies have suggested that hormone-induced Ca^2+^ influx involves heterotrimeric G~i~ proteins in hepatocytes \[[@B59],[@B60]\]. It is notable that norepinephrine and PGF~2α~ activate ERK1/2 in the presence of a barely detectable increase in intracellular InsP~3~\[[@B17]\]. This may suggest either the occurrence of local elevations of InsP~3~ which do not affect global InsP~3~, or that Ca^2+^ pools are regulated by other mechanisms such as generation of sphingosine-1-phosphate \[[@B61]\]. A role for ryanodine-sensitive Ca^2+^ stores in the endoplasmatic reticulum has also been proposed in hepatocytes \[[@B62]\] and EGTA-sensitive pools that are located in plasma membrane micro villar structures have been described \[[@B63]\]. Our data further suggest that calmodulin, which has previously been implicated in growth regulation in liver \[[@B64]\], is involved in activation of ERK1/2. The ERK1/2 responses induced by norepinephrine and PGF~2α~ were markedly decreased after pretreatment with the calmodulin inhibitors trifluoperazine, J-8, or W-7. Besides calmodulin, it is conceivable that the effect of Ca^2+^ is mediated through other Ca^2+^-binding proteins \[[@B39],[@B65]\]. Calmodulin may also act in a Ca^2+^-independent manner \[[@B38],[@B66]\], which might account for the more pronounced inhibition of hormone-stimulated ERK activity by calmodulin antagonists than by agents inhibiting the Ca^2+^ signal. Alternatively, nonspecific effects produced by calmodulin antagonists in higher doses might explain their relatively stronger inhibition. Among the downstream targets of calmodulin, the Ca^2+^/calmodulin-dependent protein kinases have been implicated in ERK1/2 activation in smooth muscle cells \[[@B8],[@B9]\], but not in other cells \[[@B67],[@B68]\]. Furthermore, the multifunctional Ca^2+^/calmodulin-dependent protein kinase was located downstream of ERK1/2 activation by platelet-derived growth factor (PDGF) in vascular smooth muscle cells \[[@B69]\]. Pretreatment of hepatocytes with KN-93 did not decrease ERK1/2 activation induced by hormones or the Ca^2+^ ionophore A23187. Thus, while the multifunctional Ca^2+^ /calmodulin-dependent protein kinase exerts several effects in hepatocytes, including growth inhibition under certain conditions \[[@B70]-[@B72]\], it does not appear to be involved in ERK activation. It is of interest that the α~1~-adrenoceptor-induced c-*fos* expression in fibroblasts was also observed to involve calmodulin, but not the multifunctional Ca^2+^/calmodulin-dependent protein kinase \[[@B73]\]. Increasing evidence suggest a role of Src kinases downstream of Ca^2+^/calmodulin in ERK1/2 signaling \[[@B10],[@B14],[@B46]\]. The present results suggest that Src kinases may be involved in ERK1/2 activation induced by PGF~2α~ and norepinephrine, while the EGF induced ERK1/2 response appears to be independent of these Src kinases. Furthermore, the ERK1/2 activation induced by the Ca^2+^ ionophore A23187 or by thapsigargin was partially decreased by Src inhibition suggesting a role of Src distal to increases in intracellular Ca^2+^. Of the possible downstream targets for Ca^2+^/calmodulin in ERK signaling in hepatocytes our findings thus lend support to a role of Src kinases, although the results do not exclude the possibility that Src kinases and calmodulin act in parallel pathways leading to ERK activation. While the present results show a role for Ca^2+^ in ERK1/2 activation by norepinephrine and PGF~2α~, it is notable that even complete inhibition of Ca^2+^ signaling only partially inhibited ERK1/2 activity. Taken together with previous observations that inhibition of PKC almost completely inhibited ERK1/2 activation by norepinephrine, vasopressin, and angiotensin II \[[@B18]\], the results suggest that several mechanisms contribute to and may act in concert in the hormonal stimulation of ERK1/2 in hepatocytes. Conclusion ========== Our present data indicate that both extracellular and intracellular Ca^2+^ is involved in hormone-induced ERK1/2 activation in cultured hepatocytes, and suggest that calmodulin and Src kinases might play a role in these signaling pathways, while the multifunctional Ca^2+^/calmodulin-dependent protein kinase does not appear to be involved. Materials and Methods ===================== Materials --------- Dulbecco\'s modified Eagle\'s medium, Waymouth\'s medium MAB 87/3, penicillin and streptomycin were from Gibco, Grand Island, NY, U.S.A. Adenosine 5\'-triphosphate, collagen, collagenase, EGTA, phenylmethylsulfonyl fluoride, benzamidine, leupeptin, pepstatin A, myelin basic protein (MBP), norepinephrine, prostaglandin F~2α~, epidermal growth factor, insulin, timolol, gadolinium chloride (hexahydrate), sulfinpyrazone, and 2-mercaptoethanol were from Sigma, St. Louis, MO, USA. A23187, 12-O-tetradecanoyl phorbol-13-acetate, thapsigargin, N-(6-aminohexyl)-5-chloro-1-naphtalenesulfonamide (W-7), trifluoperazine dimaleate, 2-\[N-(2-hydroxyethyl)-N-(4-methoxybenzenesulfonyl)\]amino-N-(4-chlorocinnamyl)-N-methylbenzylamine (KN-93), and l,2-bis(o-aminophenoxy)ethane-N,N,N\',N\'-tetraacetic acid tetra(acetoxymethyl)ester (BAPTA-AM), PP1, PP2 were from Calbiochem, La Jolla, CA, USA. (N-8-Aminooctyl)-5-iodo-1-naphtalenesulfonamide (J-8) was from Alexis Biochemicals, Lausen, Switzerland. Fura-2 AM and Pluronic F-127 were from Molecular Probes, Eugene, OR, USA. Sodium(meta)vanadate was from Fluka Chemie AG, Buchs, Switzerland. Phenyl Sepharose CL-4B was from Pharmacia Biotech., Uppsala, Sweden. Dexamethasone was from Norwegian Medicinal Depot, Oslo, Norway, \[γ-^32^P\] Adenosine 5\'-triphosphate (3000 Ci/mol) was from Amersham International, Buckinghamshire, England. Isolation and culture of hepatocytes ------------------------------------ Male Wistar rats (170--220 g) fed ad libitum were used. Parenchymal liver cells were isolated by in vitro collagenase perfusion and low-speed centrifugation \[[@B74]\] with modifications as previously described \[[@B75]\]. Cell viability was at least 95 %, measured as the ability to exclude trypan blue. The cells were suspended in medium and plated in Costar wells at 20.000 cells/cm^2^, unless otherwise specified. The culture medium (0.2 ml/cm^2^) was a 1:1 mixture of Dulbecco\'s modified Eagle\'s medium and Waymouth\'s medium MAB 87/3 containing 16.8 mM glucose \[[@B76]\], supplemented with penicillin (100 U/ml), streptomycin (0.1 mg/ml), dexamethasone (25 nM) and insulin (100 nM). The cultures were gassed with 95 % air/5 % CO~2~ and kept at 37°C. Measurement of ERK activity --------------------------- The measurement of ERK1/2 activity was performed as previously described \[[@B17],[@B77]\]. In brief, the hepatocyte cultures were exposed to agonists for 5 minutes before rinsing and scraping the cells into a 10 % ethylene glycol buffer. The lysate was centrifugated (15,800 × g) for 10 min, and the supernatant was mixed with phenyl-Sepharose which was washed twice in a 10 %, twice in a 35 % ethylene glycol buffer, and finally ERK1/2 was eluted with a 60 % ethylene glycol buffer \[[@B78]\]. The eluate was assayed for ERK1/2 activity with MBP as substrate, thereafter spotted onto P81 paper (Whatman, Maidstone, UK), which was washed, dried and counted in a liquid scintillation counter. Protein content was determined with the BCA Protein Assay (Pierce, Rockford, IL, U.S.A.). Immunoblotting -------------- Aliquots with 20 μg cell protein (total cell lysate prepared in Laemmli buffer) were electrophoresed on 10 % polyacrylamide gels ( acrylamide 30:0.8) followed by protein electrotransfer to nitrocellulose membranes and immunoblotting with a polyclonal ERK1/2 antibody against the dually threonine- and tyrosine phosphorylated forms of ERK1 and ERK2 (Promega Corporation, Madison, WI). Assessment of the multifunctional Ca^2+^/calmodulin dependent protein kinase was performed by immunoblotting using an antibody against the phosphorylated from of the enzyme. Immunoreactive bands were visualised with ECL Western blotting detection reagents (Amersham International). Measurement of cytosolic Ca^2+^ in single hepatocytes ----------------------------------------------------- The calcium measurements were done as described previously \[[@B20],[@B79]\]. Freshly isolated hepatocytes (50.000/cm^2^) were plated onto glass coverslips coated with collagen and kept in the culture medium for 30 minutes in an atmosphere of 95 % air/5 % CO~2~ at 37°C. The cells were loaded in Krebs-Ringer-Hepes buffer (KRH) with 1 % albumin and 16.8 mM glucose, supplemented with 5 μM fura-2 AM, 0.25 % (v/v) DMSO, 0.025 % Pluronic F-127 and 250 μM sulfinpyrazone for 90 minutes at 37°C. After loading, the cells were washed once and incubated with 400 μl KRH buffer with 16.8 mM glucose. Cells were preincubated with timolol (10 μM) in the presence or absence of EGTA (5 mM) for 15 minutes after loading and the experiments were performed in the same buffer. Preincubation with timolol (10 μM) and DMSO or BAPTA-AM (40 μM) was performed within the last 25 minutes of the loading period. 100 μl norepinephrine was injected after 60 seconds of registration to a final concentration of 10 μM. The experiments were carried out at 37°C. Single cell Ca^2+^ measurements were based on the ratios of the fluorescence with excitation at 345 and 385 nm. Because fura-2 AM is partly compartmentalized in hepatocytes after loading, we did not calculate apparant values for \[Ca^2+^\]~i~ from the ratios. The equipment consisted of a PTI-Δ-scan excitation device, a Nikon inverted microscope, a Hamamatsu CCD video camera and a Sony video recorder. List of abbreviations ===================== BAPTA-AM, l,2-bis(o-aminophenoxy)ethane-N,N,N\',N\'-tetraacetic acid tetra (acetoxymethyl)ester; EGF, epidermal growth factor; EGTA, ethyleneglycol-bis(β-aminoethyl)-N,N,N\',N\'-tetraacetic acid; ERK1/2, extracellular signal-regulated kinase 1 and 2; Gd, gadolinium chloride (hexahydrate); GPCR, G protein-coupled receptor; J-8; (N-8-Aminooctyl)-5-iodo-1-naphtalenesulfonamide; InsP~3~, inositol (l,4,5)-trisphosphate; KN-93, 2-\[N-(2-hydroxyethyl)-N-(4-methoxybenzenesulfonyl)\] amino-N-(4-chlorocinnamyl)-N-methylbenzylamine; MBP, myelin basic protein; NE, norepinephrine; PGF~2α~, prostaglandin F~2α~ ; PP1, 4-amino-5-(4-methylphenyl)-7-(t-butyl)pyrazolo \[3,4-d\]pyrimidine; PP2, 4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo \[3,4-d\]pyrimidine; RTK, receptor tyrosine kinase. TFP, trifluoperazine dimaleate; Thaps, thapsigargin; TPA, 12-O-tetradecanoyl phorbol-13-acetate; W-7, N-(6-aminohexyl)-5-chloro-1-naphtalenesulfonamide. Acknowledgements ================ This work was supported by The Norwegian Cancer Society, The Research Council of Norway, and the Novo Nordisk Foundation. We thank Eva Østby and Ellen Johanne Johansen for technical help and John-Arne Røttingen for helpful discussions

Background ========== Non-participation in epidemiological studies has the potential to introduce bias into the results of such studies. Lack of participation can come about through noncontact, refusal or death of the respondent. This problem becomes accentuated in longitudinal studies in which non-random attrition can be expected at each wave. In studies of elderly samples, the problem of attrition is accentuated by the increased rate of deaths. Also, such samples may have reasons for refusal which are specific to older age groups. Therefore, it is important to determine the differences between those who take part in survey research and those who do not. Often, very little is known about those who do not participate in the initial phase of any study although a number of studies have been able to determine sociodemographic differences and even health differences using hospital records. Romans-Clarkson *et al.*\[[@B1]\] have reviewed the literature on initial non-responders prior to 1988. They found that non-responders are usually older, more often male, of lower socioeconomic status and have less education. They are more likely to smoke, live in urban areas, to have a negative attitude to health surveys and have higher mortality in subsequent years. Their own study, of a random sample of community residing women, found similar results and, by examining hospital records, found no difference in occurrence of physical or psychiatric illness. Similar results for initial refusals have been found in more recent studies \[[@B2],[@B3]\]. In longitudinal studies much more is known about those who do not participate in second and subsequent waves. Although non-response can be due to death of the respondent, no contact or refusal, not all studies examining non-response at follow-up have made this distinction. Lui & Anthony \[[@B4]\] found that non-response in an elderly sample one year after initial interview was associated with lower Mini Mental State Examination (MMSE) scores, being older and having fewer years of education. Schaie *et al*\[[@B5]\] found that non-response in a three wave study over 14 years was associated with lower scores on a number of IQ measures. In a study by Clark *et al.*\[[@B6]\] of depression in people aged 18 and above, non-response was not associated with depression after accounting for demographic variables. However, only one third of non-responders were refusals, most being no contacts. Looking specifically at refusals, Eaton *et al.*\[[@B7]\], in a follow-up of 18--64 year olds from the Epidemiologic Catchment Area (ECA) Program, found that refusers were older, more likely to be married and had lower educational levels than those interviewed. Psychopathology (assessed by either diagnoses or symptoms) was not significantly associated with refusal. Bucholz *et al*\[[@B8]\] examined refusers amongst an 18--49 year old community sample followed up after 11 years. Refusers were more likely to be male and to have a history of barbiturate abuse or dependence, but were no more likely to be in the problem-drinking, minimal alcoholic group than responders. Studies of elderly refusers have found that they are more likely to be cognitively impaired, to be in poor physical health and to have lower levels of education \[[@B9]-[@B14]\]. Of the possible reasons for non-response, refusal is of particular interest because it is amenable to change. The more we know about those who are alive and have been contacted, but refuse to take part, the more likely we are to be able to improve our approach and hopefully lower our refusal rates. The aim of this paper was to examine the relationship of non-response, particularly refusal, to the physical, psychological and cognitive state of elderly participants at subsequent waves of a longitudinal survey. Methods ======= The sample ---------- In 1990--91 a sample of 945 (of 1377 selected) persons drawn from the Electoral Rolls of Canberra and the nearby town of Queanbeyan and representative of the age distribution of this community were interviewed using the Canberra Interview for the Elderly (CIE) \[[@B15]\]. An additional sample of 100 (of 145 selected) residents of Nursing homes and sheltered accommodation were also interviewed. Interviews were sought from both the subject and informant, usually a close relative. The structured interview was administered to the participant by a trained lay interviewer. It was undertaken in the participant\'s home, taking 1.5 to 2 hours to complete. All interviewing and tests were completed in one visit. Wave 2 of the study was undertaken at a mean of 3.6 (range 3.3--4.2) years later and Wave 3, at a mean of 4.0 (range 3.7--4.4) years after Wave 2. Refusers at Waves 2 and 3 were divided into two types: those who refused all participation and those who were not interviewed themselves but an interview was achieved with a relative or friend. In most cases where only an informant interview was possible, the respondent was approached, initially, to ask if they would take part. However, in some cases, a spouse or child intercepted the interviewer, decided that the respondent was unable to carry out the interview but offered to be interviewed themselves. For the purposes of this study, those subjects for whom *only* an informant interview was obtained at Wave 1 were excluded from the analysis of refusals at Wave 2, while Informant Only interviews at Wave 2 were excluded from analysis of refusal at Wave 3. No attempt was made at Wave 3, to interview anyone who had refused outright at Wave 2. Measures -------- Physical health was assessed using 1) an Activities of Daily Living (ADL) scale \[[@B16]\] which asks the participant to rate their ability to perform a range of everyday activities, with answers ranging from \"no difficulty\" to \"unable to do\", 2) a measure of the number of chronic illnesses suffered by the respondent, and 3) self-reported measures of sight and hearing impairment. The sight impairment scale and the hearing impairment scale were each comprised of 5 questions, including a general question on how they rated their sight or hearing and 4 questions on sight or hearing problems in 4 different situations. Higher scores mean higher impairment. The Goldberg anxiety and depression scales \[[@B17]\] were used as indicators of mental health. This scale contains 18 questions, 9 on depression and 9 on anxiety with answer options being \"yes\" or \"no\". Cognitive impairment was assessed using the MMSE \[[@B18]\], a test covering a range of cognitive abilities including orientation for place and time, memory and spatial ability. A higher score for this test indicates better cognitive function. Premorbid IQ was measured by the National Adult Reading Test (NART) \[[@B19]\]. The NART is a measure of verbal IQ and is commonly used to estimate IQ in elderly subjects because it is highly resistant to the effects of ageing and dementia \[[@B19]\]. It involves the participant reading a list of words of increasing difficulty and being assessed on their pronunciation. Personality was assessed only at Wave 1 by extraversion and neuroticism scales from the short form of the Eysenck Personality Questionnaire-Revised \[[@B20]\], a measure consisting of 24 questions, 12 on extraversion and 12 on neuroticism. DSM-III-R dementia diagnosis was able to be determined from the interview. At the end of the Wave 1 interview only, participants were asked by the interviewer \"Was the questionnaire at all distressing -- did some of the questions upset you or make you feel anxious?\" with options being \"yes\", \"no\" or \"don\'t know\". Five percent answered \"don\'t know\" or the question had not been asked. This 5% was treated as missing data. Sociodemographic variables included age, sex, living alone (in community sample only), years of education and previous occupation status (white collar versus blue collar / manual). Statistical analysis -------------------- Univariate analyses on continuous variables were performed using one way analysis of variance followed by post hoc modified-Least Significant Difference (LSD) tests to compare the participants and refusers. Categorical variables were analysed using chi-square tests. Multivariate analyses were performed using logistic regression with simultaneous entry of predictor variables. Results ======= The response rate at Wave 1 was 69%. The only information available on non-responders was their age and sex. Those who refused participation were not significantly different from participants in age. However, significantly more women than men refused to take part. In the community sample the response rate for males was 76% and for females, 62%. In the nursing home sample, the response rates were 75% and 65% for males and females respectively. Table [1](#T1){ref-type="table"} gives the breakdown of response status at Wave 2 and Wave 3. Of the 77% of respondents from Wave 1 who were *contacted* at Wave 2, 85% participated, 5% allowed an informant to be interviewed and 10% refused any participation. At Wave 3, of those who were able to be contacted, 86% participated, 5% allowed an informant interview, and 9% refused any participation. ::: {#T1 .table-wrap} ::: {.caption} ###### Response rates at Wave 2 of those respondents interviewed at Wave 1 and response rates at Wave 3 of those interviewed at Wave 2. ::: **Wave 2** **Wave 3** ---------------------------------------- ------------ ------------ ----- ------ N \% N \% Respondent interviews at previous wave 981 100 638 100 Respondent interviews 638 65.0 379 59.6 Informant only interviews 36 3.7 22 3.6 Refusals 78 8.0 39 6.1 Died since previous wave 215 21.9 167 26.2 No contact 14 1.4 31 4.9 ::: Those who were not contacted were people who either could not be found or had moved too far away to be followed up. This group plus those who had died have been omitted from all further analysis. Refusal at Wave 2 ----------------- Table [2](#T2){ref-type="table"} compares three groups: those respondents who participated at Wave 2, those with an Informant Only interview and those who refused any participation. Significant differences were found for years of education, occupational status, NART, MMSE and ADL scores, sight impairment and distress at end of Wave 1 interview. Consideration of LSDs for continuous variables and adjusted residuals for categorical variables showed that both groups of refusers, but particularly outright refusers, were less likely to be white collar workers. However, only the outright refusers had significantly fewer years of education. Participants had significantly higher NART and MMSE scores than either those who refused or those with an Informant Only interview. The Informant Only group was significantly more physically impaired in terms of ADL, sight (but not hearing or chronic illness) and had lower MMSE scores. A dementia diagnosis at Wave 1 was significantly more likely in those who allowed an informant interview only at Wave 2, with those refusing being more likely to have had a dementia diagnosis than those who participated at Wave 2. Those who refused any participation were more likely to report distress at the end of the Wave 1 interview than those who participated at Wave 2. There was no difference between the groups on age, sex, anxiety or depression, extraversion or neuroticism. ::: {#T2 .table-wrap} ::: {.caption} ###### Comparison of participants, those respondents who refused but allowed an informant interview, and those who refused all participation at Wave 2: Means (and SDs) or percentages ::: Wave 1 variables Participants n = 560--638 Refused but informant interview n = 22--36 Refused all Participation N = 60--78 P-value -------------------------------------- --------------------------- -------------------------------------------- -------------------------------------- --------- Age at Wave 1 76.5 (4.9) 77.9 (5.4) 75.9 (4.0) .123 Sex (% male) 48.58 44.44 42.31 .532 Years of education 11.45 (2.64) 10.82 (1.95) 10.56 (2.28) .009 Occupational status (% white collar) 65.78 48.57 42.86 .000 Lives alone (%) ^\*^ 35.78 31.25 49.30 .066 NART 113.23 (9.05) 103.23 (11.51) 105.78 (10.89) .000 MMSE 27.69 (2.30) 23.10 (5.49) 26.14 (4.48) .000 DSM-III dementia (%) 4.7 34.6 13.5 .000 ADL 1.75 (2.51) 3.73 (4.56) 1.37 (1.56) .000 Chronic illness 2.94 (1.92) 2.94 (2.16) 2.67 (1.60) .498 Sight problems 0.88 (1.88) 2.03 (3.87) 0.92 (1.97) .005 Hearing problems 1.99 (2.45) 2.87 (3.11) 2.14 (2.66) .150 Anxiety 2.41 (2.23) 2.18 (2.58) 2.84 (2.63) .269 Depression 1.89 (1.88) 2.52 (2.38) 2.14 (2.15) .153 Extroversion 5.89 (3.23) 5.95 (3.14) 6.47 (3.11) .413 Neuroticism 3.04 (2.80) 4.05 (3.29) 3.28 (2.71) .223 Distressed by interview (%) 4.01 9.68 10.96 .016 ^\*^ Living alone or with others was only examined in those living in the community. ::: Logistic regression was used to compare 1) participants to refusers and 2) participants to the Informant Only group. Those variables that showed significant differences in Table [2](#T2){ref-type="table"}, were entered simultaneously into the regression. Dementia diagnosis was not included because it was strongly associated with the MMSE score. The results are shown in Table [3](#T3){ref-type="table"}. Premorbid IQ as measured by the NART was the only significant predictor of outright refusal at Wave 2, while both NART and MMSE scores were significant predictors of the Informant Only group at Wave 2. Stepwise entry of variables was also undertaken to further examine the role of predictors in the odds of refusal. The results were essentially the same as for simultaneous entry with the exception that in comparing participants with Informants Only, sight impairment was significantly worse in the latter group. Regression analyses were also undertaken using dichotomised variables: MMSE with a cut point of 23/24, sight impairment versus no impairment, ADL score indicating no need for assistance versus needing assistance. The results were substantively the same as when continuous variables were used. ::: {#T3 .table-wrap} ::: {.caption} ###### Odds ratios (ORs) (and 95% confidence intervals) from logistic regression where the OR of being 1) a refuser or 2) an informant-only respondent at wave 2 is associated with an increase in the predictor. ::: Wave 1 Variables Refusers compared to Participants n-716 Informant-only groups compared to Participants n-674 ------------------------------------------- ----------------------------------------- ------------------------------------------------------ Age 0.96 (0.91--1.03) 1.00 (0.92--1.10) Sex (males = 1 females = 2) 1.25 (0.72--2.18) 1.21 (0.47--3.08) Years of education ^\*^ 1.02 (0.90--1.16) 1.02 (0.82--1.27) Occupational status (white collar) 0.68 (0.38--1.23) 1.07 (0.42--2.72) NART^\*^ 0.95 (0.92 0.98) 0.94 (0.89--1.00) MMSE^\*^ 0.91 (0.88--1.02) 0.75 (0.63--0.89) Activities of daily living ^\*^ 0.90 (0.76--1.07) 1.04 (0.89--1.21) Sight impairment scale ^\*^ 1.03 (0.88--1.21) 1.20 (0.98--1.45) Distressed by interview (no = 1, yes = 2) 2.30 (0.86--6.15) 0.94 (0.16--5.37) ^\*^OR associated with an increase of one unit in the scale. ::: Refusal at Wave 3 ----------------- Refusal at Wave 3 was examined using the same analyses as for refusal at Wave 2. Significant differences between the three groups were found for age, years of education, NART, MMSE and ADL scores. Differences in years of education, NART and ADL scores were essentially the same as for the three Wave 2 groups, while MMSE scores were no different between participants and those refusing any participation but significantly lower in those for whom there was only an informant interview. This latter group was significantly older than the other two groups. Logistic regression comparing participants to outright refusers showed that the only significant predictor was years of education. No individual measure predicted refusal when participants were compared to the Informant Only group, although the block of variables did make a significant contribution. Change in physical and mental health and cognitive function between Waves 1 and 2 were examined as predictors of refusal at Wave 3. Changes in ADL showed the only significant association, with those providing an informant interview having a greater increase in ADL problems compared to respondents who were interviewed and those who refused any participation. Discussion ========== At the third wave of this longitudinal study of an elderly sample, respondent interviews were achieved for 39% of those originally interviewed. Another 39% had died since Wave 1. At Wave 2, refusal, for both those who allowed an informant to be interviewed and those who refused any participation, was significantly associated with lower occupational status, fewer years of education, lower verbal IQ scores and poorer cognitive function than participants. These results confirm those of other studies which have examined refusal in the elderly \[[@B4],[@B5],[@B7],[@B12],[@B13]\]. A number of these studies also report poorer physical health in refusers. However, in this study, only those with an Informant Only interview showed significantly higher disability (ADL) than participants. Also, the Informant Only group were more likely to have had a diagnosis of dementia at the Wave 1 interview. These findings suggest that including an informant interview in the study has enabled us to gather information on those who were too physically frail or cognitively impaired to participate themselves. These are groups that are important to include in any studies of health in the elderly. Unfortunately, those who refused were also more likely to have had a dementia diagnosis. It is interesting also to consider those characteristics that were not associated with refusal. It has been suggested that psychological state may contribute to non-participation in community studies \[[@B21]\]. While the data presented here can say nothing about initial refusal, at subsequent waves we found no association between depression and anxiety scores at previous waves and refusal. As the measures of current anxiety and depression were taken 4 years prior to refusal at subsequent waves, it may be reasonable to expect no association. However, the trait measure of neuroticism, which is relatively stable over time and highly correlated to depression and anxiety was also found to have no association. Similarly, Baton *et al*\[[@B7]\], Norris \[[@B22]\] and Clark *et al*\[[@B6]\] did not find any association between non-response and depression symptoms, while Launer \[[@B11]\] found that non-responders had reported more psychiatric symptoms. Refusers at Wave 2 were no different in age to those who participated. Some other studies \[[@B11]\] have found similar results, while refusers in the study by von Strauss *et al.*\[[@B13]\] were significantly older. This latter study, however, was interviewing people in their mid eighties. At Wave 3 of our study, when the average age of the sample was 84 years, those for whom only an informant interview was achieved were significantly older than those who participated. This present study also found no difference in the proportion of males and females refusing at Waves 2 and 3, although more women refused at the initial approach. Other studies examining refusal at initial approach have found that males are more likely to refuse\[[@B1],[@B8]\], however these tend to be samples covering the adult age range. Studies of older samples \[[@B10],[@B11]\] have found little difference in the gender of initial refusers. It is difficult to know why more women than men refused at the initial approach in this study. Reports of follow-up of older samples \[[@B4],[@B9],[@B13],[@B22]\] have found no differences in gender, as was found here. Multivariate analysis comparing those who refused any participation to those who participated showed that the verbal IQ score was the only significant predictor of refusal. Comparison of those who either refused or had an Informant Only interview found that both IQ and cognitive impairment contributed. However, the ADL score did not contribute significantly in this analysis. This was somewhat surprising as the ADL score, for those with an informant only interview, was more than double that of those who participated. Of the 44 people who said they felt some distress during the Wave 1 interview, 34 gave reasons; 17 (50%) gave the cognitive section as the reason for their distress, while a further five felt distressed by their inability to understand or answer the questions. A possible limitation of this measure is that some respondents may have been reluctant to admit to any distress to the interviewer and so the results may not be an adequate indicator of distress or discontent about the interview. Nevertheless, people who refused any participation at Wave 2 were more likely to have experienced distress. Von Strauss *et al.*\[[@B13]\], in a study of attitudes to a longitudinal study, found that those with impaired cognitive functioning and lower education showed the least positive attitude and that the first contact and the cognitive testing were the most stressful part of the study. Similarly, Levin *et al*. \[[@B14]\] examined refusal on follow-up in a sample of Parkinson\'s Disease patients and found that the only significant predictor was cognitive impairment. These authors commented that participants and their spouses were concerned about the development of dementia when contacted at follow-up. As with other studies, this one also found that those who refused had had lower MMSE scores and fewer years of education. One plausible interpretation is that those people who perform poorly on the cognitive tests are aware of this and become distressed by it. Non-response is a serious issue for survey research, in particular, longitudinal studies in which each wave results in further loss and the possibility of sample bias. In studies of the elderly, loss through death of respondents is inevitable. Kessler *et al*\[[@B23]\] and, more recently, Dunn \[[@B24]\], have examined statistical methods to adjust for non-response. These methods depend on the predictors of missingness and the purpose here has been to identify these predictors. These data can then be used for such methods as multiple imputation or Full Information Maximum Likelihood (FIML) estimation. An examination of the predictors of mortality in this sample \[[@B16]\] found physical ill health, cognitive functioning and being male to be significant predictors. This report has concentrated on those who are alive, have been contacted but refuse to participate. Conclusions =========== In conclusion, refusal was predicted by low scores on a verbal IQ test and cognitive impairment, possibly because of distress caused by cognitive testing. It is important that cognitive testing be presented in a non-threatening manner and that participants in the initial wave of such studies be reassured about their performance. However, bias may be reduced by including an informant interview in the study. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ This work was supported by grant 973302 from the National Health and Medical Research Council and by a grant from the Australian Rotary Health Research Fund. Help in various phases of the study was provided by Suzanne Dee, Ruth Scott, Susan Lindsay, Colleen Doyle, Karen Maxwell, Andrew Mackinnon and Bryan Rodgers.

Background ========== The study of cost-effectiveness comparisons between competing medical interventions has led to a variety of proposals for quantifying cost-effectiveness. Although the most widely used measure is still the incremental cost-effectiveness ratio (ICER), there is increasing preference for the cost-effectiveness acceptability curve (CEAC). Willan \[[@B1]\] has recently proposed an alternative that he calls the probability of cost-effectiveness. The differences between these various approaches can be subtle, and further complexity is introduced by some authors preferring a Bayesian formulation over more traditional frequentist analysis. One purpose of this article is to clarify some important issues, concerning (a) the perspectives of different decision makers and (b) the distinction between the true value of an unknown parameter and a statistical inference about that parameter. Discussion ========== We first review various approaches to measuring cost-effectiveness, including the ICER, the mean incremental net benefit, and the measure proposed by Willan \[[@B1]\]. We then contrast these measures and argue that Willan\'s proposal is of only secondary interest to a health care provider. All of the cost-effectiveness measures are in practice unknown parameters that must be estimated from data, and we next consider inference about these measures from both the frequentist and Bayesian approaches. Finally, a fundamental flaw in the estimator proposed by Willan \[[@B1]\] for his probability of cost-effectiveness is exposed. Measures of cost-effectiveness ------------------------------ We consider two competing treatments, drugs, or other health technologies, which we refer to as Treatment 1 and Treatment 2. Conventionally, Treatment 1 is often the standard treatment whereas Treatment 2 is a new or comparator treatment. In reality there will usually be far more than two competing treatments for any condition, but for the purpose of this article it is enough to consider, like Willan \[[@B1]\], just two treatments. A little notation is necessary. Let *C*~i~ be the cost associated with an individual patient when given Treatment *i*, and let *E*~i~ be the value of an appropriate effectiveness measure associated with that patient when given Treatment *i.* Now it is important to recognise variation between patients. One patient will incur different costs and experience different effectiveness from another. Therefore, *C*~i~ and *E*~i~ are random quantities, which we interpret as the cost and effectiveness under Treatment *i* for an individual patient *randomly* drawn from the population of all patients under consideration. The probability distributions of these random quantities describes how they vary over the population. In order to compare cost-effectiveness between the two treatments, we require a way to link costs to effectiveness, and this is done through a decision-maker\'s *willingness to pay* coefficient *K.* Formally, the decision-maker is prepared to pay *K* units of money to obtain one unit of effectiveness. Therefore, the *net benefit* of Treatment *i* for an individual (random) patient is *B*~i~(*K*) = *K E*~i~ - *C*~i~. This expresses net benefit on the monetary scale by converting the *E*~i~ units of effectiveness into *K E*~i~ units of money before subtracting the cost *C*~i~. (We could equally express net benefit on the effectiveness scale as *E*~i~ - *C*~i~/*K*, but the two approaches are clearly formally equivalent.) The notation also emphasises the dependence of the net benefit on the decision-maker\'s willingness to pay coefficient *K.* Treatment 2 would be clearly more cost-effective than Treatment 1 for an individual (random) patient if *B*~2~(*K*) \>*B*~1~(*K*). This can be expressed simply in terms of the *individual incremental net benefit* (individual INB) *D*~B~(*K*) = *B*~2~(*K*) - *B*~1~(*K*) = *K D*~E~ - *D*~C~, where *D*~E~ = *E*~2~ - *E*~1~ and *D*~C~ = *C*~2~ - *C*~1~ are the increments in effectiveness and cost, respectively. If all patients were the same, and experienced the same costs and effectiveness, then the individual INB would be the same for all patients, and could then be called *the* INB. Then the comparison of treatments would become trivial. The INB would quantify the gain (if positive) or loss (if negative) per patient that would result from switching from Treatment 1 to Treatment 2. Treatment 2 would clearly be more cost-effective than Treatment 1 if, and only if, the INB was positive. However, patients will vary, and the consequence of this is that individual INB will vary between patients, and there is no single value to represent the comparison between the two treatments. Across the population, there is a probability distribution of individual INB. The measures of cost-effectiveness that are in widespread use in health economics are based on the *mean* of this distribution. We denote the population *mean incremental net benefit* (mean INB) by Δ~B~(*K*). The standard notation in probability theory for a mean or *expected* value is , so the mean incremental net benefit is Δ~B~(*K*) = (*D*~B~(*K*)) = *K* Δ~E~ - Δ~C~, where Δ~E~ = (*D*~E~) and Δ~C~ = (*D*~C~) are the population mean increments in effectiveness and cost. Then Treatment 2 is defined to be more cost-effective than Treatment 1, in terms of the population mean, if Δ~B~(*K*) \> 0. The *incremental cost-effectiveness ratio* (ICER) can be expressed as ρ = Δ~C~/Δ~E~, and we can see that Δ~B~(*K*) \> 0, i.e. Treatment 2 is more cost-effective than Treatment 1, if ρ \<*K* and Δ~E~ \> 0, or if ρ \>*K* and Δ~E~ \< 0. The probability of cost-effectiveness as proposed by Willan \[[@B1]\] is the probability that an individual (random) patient will have a positive individual INB. We can denote this by θ(*K*) = *Pr*(*D*~B~(*K*) \> 0). It can also be seen as the proportion of all patients in the population who have positive individual INBs. Δ~B~(*K*) and θ(*K*) are just two summary measures of the distribution of net benefit in the population. If the distribution is symmetric about its mean, as shown for instance in Figure [1](#F1){ref-type="fig"}, then the two measures will be in agreement, in the sense that Δ~B~(*K*) will be positive if and only if θ(*K*) is greater than 0.5. {#F1} Thus, the distribution represented by the solid curve in Figure [1](#F1){ref-type="fig"} has mean Δ~B~(*K*) = 1.3 and θ(*K*) = 0.903, so that Treatment 2 is more cost-effective in terms of having a higher mean INB and the proportion of patients who will achieve a higher individual INB under Treatment 2 is 90.3%. Conversely, the distribution represented by the dashed curve has Δ~B~(*K*) = -0.7 and θ(*K*) = 0.242, so the mean INB under Treatment 2 is now less than under Treatment 1, and only 24.2% of patients will obtain a higher individual INB under Treatment 2. If, however, the distribution is not symmetric, then it is quite possible for the two measures to give apparently contradictory indications of relative cost-effectiveness. Figure [2](#F2){ref-type="fig"} shows another two possible distributions. In the distribution shown as a solid line, Δ~B~(*K*) = 0.2 and θ(*K*) = 0.414, so the mean INB is positive but only 41.4% of patients actually have a higher individual INB under Treatment 2. This is because those 41.4% include an appreciable proportion who obtain large positive individual INBs of 2 or more, whereas although the other 58.6% have negative individual INBs they never experience a value beyond -1. Conversely, in the distribution shown as a dashed line in Figure [2](#F2){ref-type="fig"}, Δ~B~(*K*) = -1 and θ(*K*) = 0.682, so that the mean INB is negative but 68.2% of patients have a positive individual INB. {#F2} Which measure is best? ---------------------- It is well-known in health economics that, from the perspective of a health care provider needing to decide which treatment to apply to the population of patients in their care, it is the mean cost and effectiveness over the whole population that matters \[[@B2]\]. This is because the decision is to apply to the whole population. The health care provider will have to pay a cost equal to the total of all the costs for individual patients under the chosen treatment, and when expressed on a per-patient basis this is the population mean cost. For a similar reason, the per-patient mean effectiveness under the chosen treatment measures the benefit that the health care provider obtains for that cost in terms of improved health for the patients in its care. If the health care provider\'s willingness to pay coefficient is *K*, then the appropriate measure of relative cost-effectiveness is the mean INB Δ~B~(*K*), and the correct decision is to fund Treatment 2 if Δ~B~(*K*) \> 0 or Treatment 1 if Δ~B~(*K*) \< 0 \[[@B3]\]. As discussed in the previous section, this can be expressed in terms of comparing the ICER ρ with *K*, but that approach is more complex, since the comparison depends on the sign of Δ~E~. From the perspective of a health care provider, then, needing to make a decision between two treatments, the decision rests on mean INB, and in fact only on its *sign.* There is no role for Willan\'s θ(*K*). As we have seen in Figure [2](#F2){ref-type="fig"}, the wrong decision could be made if it were based on θ(*K*). Willan \[[@B1]\] says, \"The use of θ(*K*) should be helpful to policy-makers\". We agree, in the sense that it does give extra information about the distribution of individual INBs in the population, but as such it is of secondary interest, only. It should not be used as the basis of the actual decision. Nevertheless, we believe that in general an understanding of the distribution of individual INBs in the population is useful ancillary information that may be helpful to a decision-maker in the subsequent implementation of the decision. The perspective of a health care provider is not necessarily the only one of interest. An individual clinician wishing to decide how to treat an individual patient may be willing to regard that patient as randomly drawn from a large population, and might be interested in θ(*K*). However, the situations shown in Figure [2](#F2){ref-type="fig"} argue for caution. Consider for instance the dashed curve. The patient is substantially more likely to have a positive individual INB than a negative one, and this may seem to suggest prescribing Treatment 2. There is, however, a risk of a large negative INB, corresponding to the patient having a very much worse outcome with Treatment 2 than with Treatment 1. In our opinion, the mean INB is as relevant to an individual decision as to the group decision of a health care provider. Inference about cost-effectiveness ---------------------------------- The measures of cost-effectiveness described in the preceding section are all unknown in practice because they depend on the unknown distribution of individual INBs for patients in the population. From the statistical point of view they are unknown parameters. In order to learn about them, we will need to obtain some relevant evidence. This might, for instance, as supposed in Willan \[[@B1]\], consist of observations of actual costs and effectiveness for a sample of patients in a clinical trial. We then need to construct appropriate methods of statistical inference for parameters of interest, based on the data. There is a substantial literature on this topic. Based on data from a clinical trial, various authors have presented estimators and confidence intervals for the ICER \[[@B4]-[@B12]\], and comparable inferences for the mean INB \[[@B3],[@B13]\]. All of these references employ the frequentist approach to statistical inference. Analyses under a Bayesian approach have also been given \[[@B14]-[@B18]\]. The fact that the ICER is a ratio, together with the way its interpretation changes as the sign of Δ~E~ changes, mean that inference about the mean INB is generally much more straightforward \[[@B17],[@B18]\]. Inference about the mean INB is generally presented by means of a *Cost-Effectiveness Acceptability Curve* (CEAC) \[[@B16]-[@B22]\]. As introduced by van Hout et al \[[@B19]\], the CEAC plots the probability that mean INB is positive against *K.* The value of such a graph lies partly in the difficulty of specifying *K* in practice. Decision-makers are generally reluctant to commit themselves to an explicit willingness to pay, and plotting against *K* allows them to assess the relative cost-effectiveness of the two treatments over a range of values of *K.* Strictly, the probability that Δ~B~(*K*) is positive can only be a Bayesian inference, since only in the Bayesian approach is it possible to assign probability distributions to unknown parameters. The frequentist analogue is to consider the P-value of a significance test of the null hypothesis that Δ~B~(*K*) \< 0 \[[@B17],[@B23]\]. It is important to remember, as always, that the interpretation of a P-value for a null hypothesis is much less direct and meaningful than the Bayesian probability that the hypothesis is true. In its more natural Bayesian form, the CEAC states, for given *K*, the probability, *based on the available evidence*, that the true value of the unknown parameter Δ~B~(*K*) is positive. It therefore states, for given *K*, the probability, *based on the available evidence*, that Treatment 2 is more cost-effective than Treatment 1, from the perspective of a health care provider needing to make a decision between the two treatments. In presenting the CEAC in practice, authors have tended to assert that the CEAC states, for given *K*, the probability that Treatment 2 is more cost-effective than Treatment 1, omitting to refer to the fact that this probability is based on available evidence, and omitting to state the decision context. Willan \[[@B1]\] objects to this presentation of the CEAC as giving \'the probability of cost-effectiveness\'. He writes: \"The interpretation that the acceptability curve is the probability that the intervention is cost-effective is not entirely accurate and could easily be misunderstood by policy makers. Consider the situation in which the observed INB for treatment is very small, but due to a very large sample size the acceptability curve at the value of λ \[our *K*\] of interest is 0.99. Attaching the label \"the probability that the intervention is cost-effective\" to this quantity could mislead policy makers into thinking that treatment is highly beneficial compared to the standard. What, in fact, is high is our confidence that the INB, however small, is not zero.\" We agree that to refer to the CEAC as simply \'the probability of cost-effectiveness\', or \'the probability that Treatment 2 is more cost-effective than Treatment 1\', is potentially misleading if its dependence on the available evidence and on the decision context is not clear. We advocate that the phrase \'based on available evidence\' should be used to emphasise the first point, or for a technical audience the Bayesian formulation of \'the posterior probability of cost-effectiveness\' would be appropriate. It might be helpful also to emphasise that we are judging cost-effectiveness from the perspective of a health care provider needing to decide between two treatments, although this context has been so pervasively adopted in health economics that we believe it can be taken as understood. Willan proposes that θ(*K*) should more properly be called \'the probability of cost-effectiveness\', but to use the phrase for θ(*K*) without further qualification would be *equally* misleading to policy makers. To parallel the above quotation from Willan \[[@B1]\], consider the situation in which the mean INB is positive but small, but due to there being very little between-patient variation we find θ(*K*) = 0.99. Attaching the label \'the probability that the intervention is cost-effective\' to this quantity could mislead policy makers into thinking that treatment is highly beneficial compared to the standard. What, in fact, is high is the proportion of patients in the population for whom the individual INB, however small, is positive. *Neither* measure asserts the degree to which one treatment is \'highly beneficial\' compared to the other. *Both* are concerned only with the sign of INB. The CEAC gives the probability, based on available evidence, that the mean INB is positive, while θ(*K*) gives the probability that an individual INB is positive. Willan \[[@B1]\] further objects to the fact that the CEAC changes as we get more information, because it is a statistical inference. As we get more evidence, our uncertainty about the sign of Δ~B~(*K*) for a given value of *K* will decrease until we become certain either that Δ~B~(*K*) is positive (whereupon the CEAC will tend to 1) or that it is negative (in which case the CEAC will tend to 0). This is entirely natural, and we do not understand this objection. Willan\'s θ(*K*) is a probability in a different sense, because it is a population parameter, not an inference about a population parameter. Inferences change as we get more data, while the true values of the underlying parameters remain fixed, but unknown. This does not make θ(*K*) in any sense a superior kind of probability. It happens that Willan is interested in inference about a parameter that can itself be considered as a probability (although we believe it would be more helpful to call it a proportion, i.e. the proportion of patients in the population with positive individual INBs). To make inference about it, he provides an estimator (although, as we shall see below, that estimator is logically flawed), but he could have considered calculating a P-value for the null hypothesis that θ(*K*) \> 0.5. That would be analogous to the CEAC, and would change with the available data in the same way. Willan\'s estimator ------------------- Willan proposes an estimator of θ(*K*) based on data comprising observed costs and effectiveness values for a sample of *n*~S~ patients given the standard, treatment 1, and another sample of *n*~T~ patients given the intervention, treatment 2. Now since these data do not include any observations in which the same patient is given both treatments, it is completely impossible to learn the true value of θ(*K*), no matter how large *n*~S~ and *n*~T~ might be. It is easy to demonstrate this impossibility with a simple example. Suppose that we have enormous samples such that we learn the true distribution in the population of costs and effects for treatment 1 and the true distribution of costs and effects for treatment 2. In particular, we will also learn the true distribution of net benefits *B*~i~(*K*) for each treatment. Suppose that the value of *K* is given and that the distribution of the net benefit *B*~1~(*K*) under treatment 1 is *N*(0,1) (i.e. normal with mean 0 and variance 1), while the distribution of net benefit *B*~2~(*K*) under treatment 2 is *N*(1,1). With all this information we know these distributions, and so we know exactly that Δ~B~(*K*) = 1. We therefore know with certainty that treatment 2 is more cost-effective than treatment 1 for a health care provider with the given value of *K.* Even with all this information we do not know θ(*K*), because this depends on how correlated *B*~1~(*K*) and *B*~2~(*K*) are in the population. At one extreme, they might be perfectly positively correlated, such that for every individual in the population it is true that *B*~2~(*K*) = *B*~1~(*K*) + 1. Then θ(*K*) = 1, because the individual INB is positive for *every* patient. At the other extreme we might have perfect negative correlation, so that for every individual in the population we have *B*~2~(*K*) = 1 - *B*~1~(*K*). Then treatment 2 is more cost-effective for all those individuals for whom *B*~1~(*K*) \< 0.5. The proportion of such individuals in the population is 69.15%, and so θ(*K*) = 0.6915. Willan\'s estimator effectively assumes that *B*~1~(*K*) and *B*~2~(*K*) are independent in the population, and for our example this implies that *D*~B~(*K*) is distributed as *N*(1,2), with the result that θ(*K*) = 0.7633. The assumption is arbitrary and completely unsupported. Indeed one might imagine that in practice there would be quite strong correlations, on the basis that a patient who responds well to one treatment might respond relatively well to the other, and similarly for costs. But we reiterate that there is absolutely no evidence about this correlation in the data which Willan supposes are available. Indeed for most kinds of intervention it is impractical to test two treatments on the same patient, and even when this is possible we must expect the picture to be complicated by cross-over effects. What Willan \[[@B1]\] actually estimates is the probability that a randomly chosen patient given treatment 1 will obtain a higher net benefit than *another* randomly chosen patient given treatment 2. This is an entirely different measure from θ(*K*) and we cannot imagine that it is of fundamental interest to any policy maker. Summary ======= 1\. From the perspective of a health care provider needing to decide which of two treatments to fund, it is the mean cost and mean effectiveness, over the whole population of patients within the provider\'s remit, that are of primary concern. This leads to the mean INB Δ~B~(*K*) as the appropriate measure of cost-effectiveness, and to the specific question of whether Δ~B~(*K*) is positive. 2\. Any measure of cost-effectiveness is a property of the population of patients under consideration, and is an unknown parameter. We make statistical inferences about parameters, based on available evidence. The true value of the parameter is fixed, independent of the available evidence, but unknown. Any statistical inference statement about the parameter is liable to change as the evidence changes. The CEAC plots the probability, based on available evidence, that Δ~B~(*K*) \> 0, and is the most relevant inference for a health care provider needing to decide between two treatments. Because it is an inference, the CEAC depends on the data. 3\. When reporting the CEAC in practice, its dependence on the data should be made clear by referring to it in such phrases as \'the probability of cost-effectiveness based on available evidence\' or \'the posterior probability of cost-effectiveness\'. It may also be useful to emphasise that cost-effectiveness is being judged from the perspective of a health care provider needing to decide which of two treatments to fund. 4\. Willan\'s probability of cost-effectiveness θ(*K*) may be useful to a decision maker in the same way as knowing other aspects of the distribution of individual INBs in the population would be useful, but it will generally be of secondary importance to the sign of Δ~B~(*K*). Since θ(*K*) is a parameter it does not depend on the data, but it is unknown, and any statistical inference about it will depend on the data. 5\. θ(*K*) should not be referred to simply as \'the probability of cost-effectiveness\' either, and we advise calling it, for example, \'the proportion of patients for whom the treatment is cost-effective\'. The fact that it is an unknown parameter should be emphasised, by a formulation such as \'based on available evidence, the proportion \... is estimated to be \...\' 6\. The proposed estimator of θ(*K*) given by Willan \[[@B1]\] is flawed. This parameter cannot be estimated consistently from the kind of data considered by Willan. His proposed estimate is in fact a probability concerning two randomly selected future patients, and is of doubtful interest to any decision maker. Conclusion ========== In conclusion, therefore, we reiterate the appropriateness of the CEAC as the primary comparator of relative cost-effectiveness between two treatments from the perspective of a health care provider. Willan\'s \'probability of cost-effectiveness\' would be of only secondary value in evidence presented to policy makers, and his proposed estimator of that probability is fatally flawed. However, we agree with Willan that assessments of cost-effectiveness should be more clearly stated, avoiding the unqualified phrase \'the probability of cost-effectiveness\'. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== For over ten years chemo-radiotherapeutic combinations have evolved as prevalent modalities in the cure of solid tumors. The rationale for these treatments relies mostly on the drugs\' ability to sterilize disseminated metastases (spatial cooperation) and on radiation-drug interaction for improved local control of the primary tumor. Supraadditive interaction, often resulting from impaired repair of radiation-induced sublethal damage, may lead to limiting toxicities. In contrast, pure additivity of the treatments may conceivably be turned to advantage because the dose-dependent response of cells to radiation includes a quadratic term. This is of concern, for example, to farnesyl transferase inhibitors (FTIs) of which six have already been tested in phase I clinical trials \[[@B1],[@B2]\] with most attention being paid to the non-peptidomimetic R115777 \[[@B3]-[@B5]\]. As one may expect a cytostatic rather than a cytotoxic effect from FTIs, assays of combined modality treatment with cytotoxic agents applied in close temporal proximity should be performed. This prompted us to investigate the cells\' response to RPR-130401 (Scheme [1](#F1){ref-type="fig"}), a non-peptidomimetic FTI issued from chemical optimization from series selected by random screening at Rhône-Poulenc-Rorer Co. \[[@B6]\], both alone or in concomitant association with ionizing radiation. A comprehensive approach to these problems requires an insight into the molecular targets of drug action. Together with mutation or deletion of the *p53* tumor suppressor gene, mutations in one of the four *Ras* genes (*H-, N-, K*~*A*~- and *K*~*B*~-*Ras*) represent the most frequent genetic alterations in human cancers \[[@B7]\] with predominance in pancreatic (90%), colorectal (50%) and lung (20%) tumors (for a review on ras mutations in relation to oncogenic transformation see \[[@B2]\]). Activation of Ras proteins requires prenylation of the CAAX sequence at their C-terminus, to ultimately permit their association with other proteins \[[@B8]\] or with the inner face of the plasma membrane \[[@B9]\]. A key step in this process is catalyzed by farnesyltransferase (FTase). The K-Ras protein can alternatively be prenylated by type I geranylgeranyltransferase (GGTase I), a closely related enzyme \[[@B10]\]. FTase has been a very attractive target for antitumor drug discovery because prenylation is required for oncogenic Ras to transform cells \[[@B11]-[@B13]\]. Numerous inhibitors have been developed, including FPP analogs, CAAX peptide analogs, bisubstrate analogs \[[@B14],[@B15]\] and more recently GGTase I inhibitors \[[@B16]\]. However, it is now known that a range of non-Ras proteins are targets for prenylation \[[@B17]-[@B19]\]. In spite of this, the quest for specific FTase-targeting drugs that would not affect GGTase is still pertinent, because geranylgeranylation of normal cellular proteins is five to ten times more prevalent than farnesylation \[[@B20],[@B21]\]. RPR-130401 acts as a competitive inhibitor of the FPP substrate with respect to the farnesylated proteins, with an IC~50~ of 28 nM for the inhibition of K-Ras prenylation *in vitro*\[[@B6]\], and has demonstrated *in vivo* antitumor activity in mice bearing human carcinoma xenografts \[[@B22]\]. RPR-130401 is highly selective (more than 300-fold) for FTase with regards to GGTase, an unexpected result since GGTase can accommodate and transfer both FPP and geranylgeranyl pyrophosphate to CAAX-motifs in proteins \[[@B23]\]. Consistently, RPR-130401 efficiently inhibits Ras farnesylation in cells but does not block geranylgeranyl transfer to Ras \[[@B24]\]. RPR-130401 is also very efficient in inhibiting lamin B farnesylation \[[@B6]\]. The lack of protein substrate specificity is of particular interest for the development of this series of FTIs, as non-Ras farnesylated proteins may participate in transformation by Ras-dependent or -independent pathways. The discovery of prenylation-dependent oncogenic protein tyrosine phosphatases harboring a CAAX-box \[[@B25]\] strengthens the interest for such FTIs as RPR-130401. For studies with RPR-130401 we chose three cell lines bearing a *K-Ras* allele with a point mutation, namely, the human colon adenocarcinoma HCT-116 with mutation at codon 13 \[[@B26]\], the human pancreatic carcinoma MiAPaCa-2 with mutation at codon 12 \[[@B27]\], and the human lung carcinoma A-549 with another mutation at codon 12 \[[@B28]\]. Cell growth or survival, cell cycle progression, ploidy, lamin B structure in nuclear lamina, cytoskeleton framework, and lamin B fragmentation were taken as an endpoint in single or combined treatment with ionizing radiation. The data show that RPR-130401 affects the integrity of the lamin B network, resulting in an aberrant onset of mitosis and cytokinesis and ensuing hyperploidy without significant changes in radiation susceptibility. Results ======= Growth inhibition by RPR-130401 ------------------------------- The response of HCT-116, A-549 and MiAPaCa-2 cells to RPR-130401 was investigated through growth inhibition assays with exponentially growing subcultures. Cells were exposed to the drug for up to 7-days at concentrations in the range 2.5--7,000 nM. RPR-130401 was found to inhibit cell growth with an IC~50~ of 50 nM for HCT-116, 120 nM for MiAPaCa-2 and 710 nM for A-549 cells. The effect correlated with a pronounced lengthening of the cells\' doubling time (Figure [2](#F2){ref-type="fig"}). {#F1} Altered cell cycle progression by RPR-130401 and radiation ---------------------------------------------------------- To investigate cell cycle redistribution by RPR-130401, HCT-116, MiAPaCa-2 and A-549 cells were exposed for 5-days to 1 × IC~50~ or 10 × IC~50~ FTI (relative to the cell line of interest) and submitted or not to γ-rays 24-h before harvest. BrdUrd was incorporated at the end of treatment for S-phase cell labeling. The three cell lines were not equally responsive to drug (Figure [3](#F3){ref-type="fig"}). RPR-130401 induced accumulation of HCT-116 cells in G2-M phase, with a concentration-dependent increase in the ploidy number and a decline in the normodiploid G1- and S-phase content. These observations correlate those made by other authors using Ras-mutated human pancreatic cells and L-744,832 \[[@B29]\] or human lung cancer cells and FTI-2153 \[[@B30]\]. MiAPaCa-2 cells showed a similar response at low (1 × IC~50~) drug concentration. However, with this cell line both the G1 phase depletion and the increase in the ploidy number, were substantially less pronounced at high (10 × IC~50~) than at low (1 × IC~50~) drug concentration. Finally, with the exception of the pre-mitotic arrest, the effect of RPR-130401 on cell cycle progression and ploidy change was comparatively minor in A-549 cells. Irradiation promoted drug-induced effects. Incidentally, flow cytometric analysis did not provide evidence of a significant amount of the sub-G1 DNA fragments expected from apoptosis. {#F2} Radiation survival in the presence of RPR-130401 ------------------------------------------------ Two methods were used to determine whether RPR-130401 resulted or not in an altered radiation survival. For lengths of drug exposure in excess of 12-h (1- to 5-days), cells were trypsinized and plated following treatment to preclude artifacts due to cell multiplicity. Short contact with drug (1-, 2-, 3- or 12-h) was performed in the most sensitive cell line only, i.e., in HCT-116 cells. In that case, cells were plated prior to treatment and the data corrected for changes in cell multiplicity (see Materials and Methods) where necessary. A total of 78 survival curves were generated and fitted to eq. (2) (see Materials and Methods) for data analysis. Irrespective of the method used, exposure of cells to RPR-130401 did not produce any significant modification of radiation sensitivity relative to controls, even with the largest drug concentration and the longest length of drug exposure (Figure [4](#F4){ref-type="fig"}). {#F3} Effects of RPR-130401 on lamin B distribution --------------------------------------------- Altered distribution of lamin B following exposure to the FTI was investigated in HCT-116, A-549 and MiaPaCa-2 cells. Cells were exposed for 5-days to 1 × IC~50~ or 10 × IC~50~ RPR-103401, submitted or not to γ-rays 24-h before harvest and lamin B was subsequently probed by immunofluorescence. Typical results are shown in Figure [5](#F5){ref-type="fig"}. ![Typical radiation survival curves for asynchronous growing cells without (•) or with exposure to 1 × IC~50~ (○) or 10 × IC~50~ (□) RPR-130401. Cells were plated, exposed to the drug for 3-days and irradiated. Each flask was trypsinized immediately after irradiation and 800 to 1,000 cells were replated in drug-free medium for colony formation assays. To account for the loss of plating efficiency du to drug exposure, survival values were normalized relative to the surviving fraction for drug alone. The curves were drawn for best fit to a linear quadratic equation (eq. (2), see Materials and Method) taking all data points into consideration. Found: α = 0.518 ± 0.036 Gy^-1^, β = 0.044 ± 0.016 Gy^-2^ for HCT-116 cells; α = 0.193 ± 0.036 Gy^-1^, β = 0.065 ± 0.014 Gy^-2^ for A-549 cells; α = 0.189 ± 0.058 Gy^-1^, β = 0.086 ± 0.026 Gy^-2^ for MiAPaCa-2 cells. Each data point came from an average over three flasks or more. *Bars, SD.* The figure at bottom right shows the surviving fraction at 2 Gy (SF~2~) of HCT-116 cells in the presence of RPR-130401. The conditions were the same as above. Data were corrected for the expression of potentially lethal damage \[82\].](1471-2210-2-2-4){#F4} Control cells (Figure [5](#F5){ref-type="fig"}A,[5](#F5){ref-type="fig"}D &[5](#F5){ref-type="fig"}G) showed a nuclear distribution of lamin B with localization to the nuclear lamina as indicated by a ring-like staining at the nuclear periphery. During mitosis the lamin was either cytoplasmic (Figure [5](#F5){ref-type="fig"}D) or poorly detectable (Figure [5](#F5){ref-type="fig"}G &[5](#F5){ref-type="fig"}H). The pattern of lamin B distribution was markedly altered following incubation with RPR-130401, depending on the cell line. For HCT-116 (Figure [5](#F5){ref-type="fig"}B,[5](#F5){ref-type="fig"}C) and A-549 cells (Figure [5](#F5){ref-type="fig"}E,[5](#F5){ref-type="fig"}F), the lamin B staining was still nuclear with the characteristic ring-like structure but, in addition to giant nuclei typical of treated HCT-116 cells (Figure [5](#F5){ref-type="fig"}C), cell doublets with no proper separation of the daugther nuclei after telophase were observed (Figure [5](#F5){ref-type="fig"}F). MiAPaCa-2 cells showed a different pattern in response to RPR-130401 (Figure [5](#F5){ref-type="fig"}H,[5](#F5){ref-type="fig"}H\'). With these cells, lamin B was not strictly confined in the nucleus as indicated by cytoplasmic patches (Figure [5](#F5){ref-type="fig"}H). Examination of DAPI-stained nuclei confirmed the absence of DNA in these patches (Figure [5](#F5){ref-type="fig"}). Lamin B figures were not appreciably modified by exposure to ionizing radiation; however, radiation induced the formation of micronuclei in relation to mitotic cell death (data not shown). Effects of RPR-130401 on actin and tubulin networks in HCT-116 cells -------------------------------------------------------------------- Some FTIs reportedly affect the actin cytoskeleton *via* the Rho (a Ras-related GTP-binding protein) pathway \[[@B31]\] and Rho-B, an endosomal Rho protein involved in receptor trafficking, has recently been suggested as an important mediator of the antineoplastic potential of FTIs \[[@B19],[@B32]\]. We therefore sought to investigate the effects of 50 nM or 500 nM RPR-130401 on the distribution of actin and tubulin networks in HCT-116-cells exposed to RPR-130401 and/or radiation. Pictures obtained using rhodamine-phalloidin and immunofluorescent labeling of isothiocyanate-conjugated tubulin in control and treated cells, are shown in Figure [6](#F6){ref-type="fig"}. {#F5} L-739,749, another FTI, was earlier shown to induce rapid morphological reversion of rat fibroblasts (Rat1-ras) transformed with H-ras \[[@B33]\]. Such phenotypic reversion was not observed using RPR-130401 in HCT-116 cells. Normal membrane ruffles, circumferential actin bundles and focal adhesions as seen in control HCT-116 cells (Figure [6](#F6){ref-type="fig"}A), were still present in cells incubated for as long as 5-days in the presence of RPR-130401 (Figure [6](#F6){ref-type="fig"}B). The size of the nuclei and the number of giant cells grew in parallel in the presence of the drug, up to threefold relative to controls. The stress fibers and the tubulin network were normal in these cells (Figure [6](#F6){ref-type="fig"}B &[6](#F6){ref-type="fig"}D). Examination of bisbenzimide-stained nuclei showed the absence of apoptotic bodies (Figure [6](#F6){ref-type="fig"}A\' &[6](#F6){ref-type="fig"}B\'), and no vesicles were formed in the cytoplasm of treated cells. RPR-130401-induced modification of the HCT-116 karyotype -------------------------------------------------------- Activation of Ha-ras gene expression in mouse NIH3T3 cells is known to result in genomic instability, identified from chromosome aberrations including acentric fragments, multicentric and double-minute chromosomes \[[@B34],[@B35]\]. We therefore investigated the effects of 50 nM or 500 nM RPR-130401 on the ploidy and chromosome integrity in HCT-116-cells exposed to RPR-130401 and/or radiation. Figure [7](#F7){ref-type="fig"} shows the distribution of the chromosome score within individual metaphases of colcemid-treated cells after 1-day of contact with RPR-130401. Control cells showed a near-diploid karyotype with a modal number of chromosomes of 45. Excess ploidy increased with the drug concentration, the radiation dose and the length of drug exposure. The mitotic index decreased during treatment (data not shown), thus confirming that the accumulation of HCT-116 cells was in the G2 phase. {ref-type="fig"}. The bar (C) represents 20 μm.](1471-2210-2-2-6){#F6} However, no chromosome aberration was seen at this stage using FISH analysis with all human centromeres probes. All chromosomes in the metaphases examined displayed one unique centromere, even in metaphases containing up to 200 chromosomes (data not shown). Western blot analysis of lamin B in treated cells ------------------------------------------------- Flow cytometry (Figure [3](#F3){ref-type="fig"}) and DAPI fluorescence of nuclei (Figure [5](#F5){ref-type="fig"}), did not provide evidence of a significant level of apoptosis in cells exposed to RPR-130401. In some instances, however, the pictures obtained from immunofluorescence analysis of the lamin B status in cells exposed for 5-days to RPR-130401, suggested breakdown of nuclear lamina, and it has been shown recently that inhibition of protein prenylation may cause apoptotic cell death in pancreatic cells \[[@B36]\]. As lamin B cleavage into 46 kDa fragments is a hallmark of apoptosis \[[@B37]\], we performed a Western blot determination of lamin B in extracts from HCT-116, A-549 and MiAPaCa-2 cells exposed for 5-days to 10 × IC~50~ RPR-130401. The results (Figure [8](#F8){ref-type="fig"}) unambiguously show complete absence of lamin B cleavage. {#F7} {#F8} Discussion ========== The main effects of RPR-130401 against the three K-Ras mutated cell lines used in this study were *(i)* a pronounced cytostatic effect, with an IC~50~ in the range 50--710 nM; *(ii)* accumulation of cells in G2 phase, with G1 and S phase depletion; *(iii)* induction of a large amount of hyperploid cells (HCT-116 and MiAPaCa-2) showing active S-phase DNA synthesis, together with some giant cells; *(iv)* profound alteration of the lamin B pattern, without cleavage of the protein; *(v)* no alteration of the microtubule and actin networks, even after prolonged incubation with a high drug concentration in the most sensitive cell line (500 nM, HCT-116 cells); *(vi)* strictly additive interaction with ionizing radiation with regard to cell kill, cell cycle redistribution and induction of hyperploidy. Two alternative pathways should be taken into consideration in order to explain these observations, namely, inhibition of the farnesylation of p21^Ras^ or related proteins from the Ras superfamily, or inhibition of the farnesylation of other proteins resulting in disruption of the mitotic apparatus or cytokinesis traverse. These hypotheses are discussed below. p21^Ras^ activation in normal and tumor cells is mandatory to the transition from cell quiescence to proliferation. However, it is generally agreed that inhibition of p21^Ras^ farnesylation does not suppress mitogenic functions in Ras-transformed cells \[[@B38]\]. Moreover, it has recently been shown that growth inhibition by FTIs does not necessarily proceed from inhibition of p21^Ras^ prenylation, and that various Ras-unrelated proteins may be important targets for FTI treatment \[[@B2],[@B19],[@B29],[@B39],[@B40]\]. Even though it results in an altered subcellular distribution of p21^Ras^\[[@B6]\], RPR-130401 is not able to block p21^Ras^ geranylgeranylation \[[@B24]\] in such a way that inhibition of p21^Ras^ farnesylation is not likely to account for the main effects of the drug. On the other hand, the peptidomimetic FTI L-739,749 has been reported to prevent Ras-induced transformation through inhibition of Rho-B prenylation \[[@B31]\] with a block in cell proliferation, increased apoptosis \[[@B41]\] and major changes in stress fiber formation, cell shape and mobility. None of these characteristic changes in cell morphology and architecture was observed upon prolonged exposure to RPR-130401 (Figure [6](#F6){ref-type="fig"}), which seems to rule out any member of the Rho family as a target for RPR-130401. Considering the efficient pre-mitotic block and hyperploidy induced by RPR-130401, as also found with other FTIs \[[@B16],[@B29],[@B30],[@B42]\], it is tempting to postulate that the target of the drug is a protein involved at some stage in mitosis and/or cytokinesis, and whose farnesylation is required for proper activation. Indeed, protein isoprenylation is important at the G2/M transition in *Schizosaccharomyces pombe*\[[@B43]\] and nuclear lamins are major targets for FTase \[[@B18],[@B44]\]. Lamin B farnesylation, which is a very sensitive target for RPR-130401 *in vitro*\[[@B6]\], appears to deserve consideration as a candidate to this role. As a matter of fact, among the three major type A-C lamins, lamin A and lamin B are initially translated with a CAAX motif. This motif is cleaved upon post-translational maturation of lamin A but persists in lamin B, unless the C-terminal aminoacid residue is mutated \[[@B23]\]. Studies with mutant lamin deleted of the CAAX sequence have shown that the CAAX motif is essential to anchoring lamin B to the nuclear envelope \[[@B45],[@B46]\]; moreover, farnesylation is required for the integration of newly synthesized lamins into the pre-existing nuclear lamina during interphase, and GGTase is unable to substitute for FTase at this stage \[[@B47]\]. These data altogether suggest that inhibition of lamin B farnesylation could be an important target of RPR-130401, resulting in an aberrant onset of mitosis and cytokinesis and ensuing hyperploid cell generation. Strong support to this scheme is given by immunofluorescence experiments showing that the intranuclear distribution of lamin B is profoundly altered in cells exposed to RPR-130401 (Figure [5](#F5){ref-type="fig"}). This occurred without lamin B breakdown (Figure [8](#F8){ref-type="fig"}), consistent with lack of a significant level of apoptosis in the three cell lines used. Interestingly SCH-66336, another FTI, reportedly allows accumulation of unfarnesylated prelamin A in patients under phase I investigation \[[@B1]\] and in several cell lines in culture as well \[[@B48]\]. It has also been shown that SCH-66336 induces diruption of microtubule binding to the centromeric proteins CENP-E and CENP-F \[[@B17]\], thus evoking again faulty mitosis as an important outcome of FTI exposure. Conclusions =========== Oncogenic *Ras* expression has been reported to have different outcomes in rodent and human cells with regard to radiation susceptibility. Studies in mouse \[[@B49]-[@B52]\] or rat cells \[[@B51],[@B53]-[@B59]\] relied on transfection with *N-Ras*\[[@B49],[@B50]\], *H-Ras*\[[@B51],[@B53]-[@B58]\] or co-transfection with *H-Ras* and *myc*\[[@B53]-[@B55]\], \[[@B58]-[@B60]\]. Except in one study \[[@B51]\], the authors have conclusively shown that transformation of rodent cells with the *Ras* oncogenes results in enhanced radiation resistance. Co-transfection with *H-Ras* and *myc* was found to act synergistically with regard to induction of the radioresistant phenotype \[[@B53],[@B54],[@B58]-[@B60]\], possibly in relation to inhibition of *myc*-dependent apoptosis \[[@B58],[@B61]\]. With two exceptions in support of increased radioresistance \[[@B62],[@B63]\], early studies on human cells did not show any clear-cut relationship between radiation sensitivity and expression of oncogenic *Ras*\[[@B64]-[@B67]\]. In addition, a study involving a cohort of patients with stage II and III carcinoma of the uterine cervix, demonstrated the absence of a significant correlation between rearrangements or structure of the *Ha-ras-1* and/or *c-myc* genes and tumor response to radiotherapy \[[@B68]\]. However, the contribution of *N- or K-Ras* activation to intrinsic radioresistance in human cells has recently been adressed directly, i.e., without the perturbations introduced by cell transformation \[[@B69]\]. The mechanisms involved in *Ras*-related radioresistance are still open to question, but it might reasonably be proposed that it proceeds from constitutive activation of pathways dowstream from *Ras* in signal transduction, such as raf-1 \[[@B70]\] or phosphatidylinositol 3-kinase \[[@B71]\]. Whether and how FTase inhibition may affect the response of cells to ionizing radiation has also been a long-standing problem. The outcome appears to depend very much on the cell line and the specificity of the FTI used. The FTase inhibitor FTI-277 reportedly acted as a radiosensitizer in *H-Ras*-transformed rat embryo fibroblasts \[[@B72]\] and in human tumor cells harboring *H-Ras* mutation \[[@B73]\]. On the other hand, the radiation susceptibility of human tumor cells expressing the wild-type form of *Ras* was not altered by FTIs \[[@B73]\]. However, enhanced radiation response by FTI-277 has been described in human cells expressing wild-type *Ras* together with the 24-kDa isoform (FGF2) of basic fibroblast growth factor, a factor involved in acquired radioresistance \[[@B74]\]. This was taken as an indication that farnesylated proteins other than Ras might be involved in the modulation of radiation response by FTI-277 \[[@B74]\]. Weak radiosensitization only was obtained with FTI-277 in K-Ras mutants, unless FTI-277 was combined with GGTI-298 to hinder geranylgeranyl transfer to K-Ras \[[@B73]\]. Last but not least FTI-induced radiosensitization, when it occurs, seems to be independent on the growth inhibitory effect of the FTI \[[@B73]\]. We show here that RPR-130401 does not significantly alter the radiation susceptibility of human cells bearing K-Ras mutation. This occurs in spite of significant hyperploidization, accumulation of cells in G2 and depletion of S phase. Finally, impaired repair of radiation damage by RPR-130401, appears to be very unlikely to occur. This seems valuable in the prospect of combining RPR-130401 as an adjuvant to conventional radiotherapy, as it should provide strictly additive interaction without the local and systemic adverse effects inherent in treatments with radiation sensitizers \[[@B75]\]. In this respect it is very exciting to note that two farnesyltransferase inhibitors, FTI-276 and L-744,832, have recently proven to act synergistically with ionizing radiation *in vivo* for growth inhibition of xenotransplanted *H-Ras-*mutated human tumor cells, without increase in acute or long-term radiation-induced damage to healthy tissues \[[@B76]\]. Materials and methods ===================== Reagents -------- RPR-130401 was synthesized as described \[[@B77]\] and stored at -20°C as a 20 mM stock solution in DMSO. Prior to use, the solution was diluted first into medium containing 20% DMSO, then to the final concentration in culture medium. DMSO was kept below 0.1% so as not to alter cell growth or radiation response. Colcemid and all products from cell culture were from Invitrogen (Cergy-Pontoise, France). BrdUrd, propidium iodide, bisbenzimide, DAPI, Giemsa stain, TRITC-phalloidin and mouse anti-α-tubulin monoclonal antibody (clone DM1A) came from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Mouse monoclonal antibodies directed against lamin B was from Oncogene Research Products (San Diego, California). FITC-conjugated sheep anti-mouse IgG was from Sanofi-Synthélabo (Paris, France). Alexa^®^ 488 goat anti-mouse IgG conjugate was from Molecular Probes (Eugene, Oregon). Rat monoclonal antibody directed against BrdUrd was purchased from Harlan Sera-Lab (Loughborough, United Kingdom), and fluorescein isothiocyanate-conjugated goat anti-rat IgG (heavy chain specific) was from Southern Biotechnology Associates (Birmingham, Alabama). Human centromere probes for FISH was obtained from QBiogene (Illkirch, France). Solvents and chemicals were of the highest purity available and came from Merck Eurolab (Paris, France). Cell culture ------------ HCT-116 human colon carcinoma (ATCC CCL-247), MiAPaCa-2 human pancreatic carcinoma (ATCC CRL-1420) and A-549 human lung carcinoma cells (ATCC CCL-185) were grown as monolayers in DMEM (HCT-116, MiAPaCa-2) or RPMI-1640 (A-549 cells) containing sodium pyruvate and Glutamax I, and supplemented with 10% foetal calf serum, 100 IU/ml penicillin, and 0.1 mg/ml streptomycin. Cells were routinely subcultured after trypsinization at a density of 6,000 cells/cm^2^ in culture flasks every 5-days and incubated at 37°C in a humidified atmosphere of 7% CO~2~ in air. Less than 9--10 passages of the same primary subculture were allowed. Prior to seeding, cells were rinsed twice with HBSS, harvested by trypsin-EDTA treatment and counted in an hemocytometer. Survival assays --------------- Radiation survival with or without combined drug exposure was determined by clonogenic assays. For determination of the response to radiation alone, cells from exponentially growing subcultures were seeded in 25 cm^2^ flasks at the suitable density, incubated for 3-h to allow cells to adhere and spread, and irradiated in culture medium at room temperature in an IBL-637 (^137^Cs) irradiator (CIS-Biointernational, Saclay, France) at a dose-rate of 0.92 Gy/min. All experiments were performed in triplicate or more. Two protocols were used in single or combined treatment assays. For short lengths of drug exposure (≤ 12-h), 1,000 cells were plated in triplicate in 25 cm^2^ flasks and incubated for 3-h. Cells were then exposed to drug for 1-, 2-, 3-, 4- or 12-h, irradiated and returned to the incubator for 1-h in the presence of RPR-130401, followed by two washes with HBSS and feeding with drug-free medium. Colonies were allowed to grow for 8--11 days, then fixed with methanol, stained with Giemsa and scored. Small colonies (\< 30--50 cells) were disregarded. Results were corrected for cell multiplicity (see below). Where the length of drug exposure was in excess of 12-h, cells were plated after treatments to avoid biases due to the formation of microcolonies. Typically, cells were plated in 25 cm^2^ flasks and exposed to drug for 1-, 2-, 3-, 4- or 5-days, and irradiated. Seeding was adjusted so as to obtain an equal number of cells at the time of irradiation, taking into consideration the cytostatic effect of the FTI and the length of drug exposure. Each flask was trypsinized immediately after irradiation, the cells were counted and 1,000 cells per flask were replated in drug-free medium for colony formation assays. Flow cytometry and cytogenetic analysis --------------------------------------- Cell cycle progression was monitored by dual parameter flow cytometry using a FACStar PLUS cytofluorometer (Becton Dickinson, Le Pont de Claix, France). Cells were grown for 3- to 6-days with or without RPR-130401. At that time, flasks were irradiated to 3.5 Gy (HCT-116), 5 Gy (A-549) or 5.5 Gy (MiAPaCa-2) and allowed to rest in the incubator for 24-h prior to fixation. 15-min before harvesting, cells were incubated with BrdUrd (10 μM) for pulse-labeling of S-phase cells, then collected, washed once with cold PBS and fixed in 70% ice-cold ethanol. Treatment of fixed cells for cytofluorimetric analysis, data acquisition and processing were done according to Demarcq *et al.*\[[@B78]\]. Cell cycle analysis was performed with ProCyt software (CEA-INSERM, Grenoble, France). Alternatively, HCT-116 treated cells were incubated with 0.2 μg/ml colcemid for 1-h, washed twice with HBSS and allowed to swell for 30-min in 75 mM KCl hypotonic solution. Cells were then fixed in 3:1 v/v acetic in ice and washed threefold in the same solvent. Fixed cells were dropped onto glass slides to spread chromosomes, and air dried. Slides were stained with Giemsa before scoring. Immunofluorescence ------------------ For immunofluorescence characterization, cells were grown on 20 × 20 mm coverslips in medium containing 10% foetal calf serum for 3- to 6-days with or without RPR-130401. At that time, the coverslips were irradiated to 3.5 Gy (HCT-116), 5 Gy (A-549) or 5.5 Gy (MiAPaCa-2) and allowed to rest in the incubator for 24-h prior to fixation. To visualize polymerized actin and tubulin, cells were processed according to Mies *et al.*\[[@B79]\]. Briefly, cells were fixed in 0.25% glutaraldehyde, 0.5% Triton X-100 for 1-min, followed by 10-min in freshly prepared 4% formaldehyde and extensive washing. The coverslips were then incubated (30-min, 37°C) with 0.2 μg/ml TRITC-labeled phalloidin, or with a 1:200 working dilution of monoclonal anti-α tubulin antibody followed by a 30-min incubation with a 1:1,000 dilution of FITC-conjugated sheep anti-mouse IgG at 37°C. To visualize lamin, cells were fixed in acetone/methanol (1:1, v/v) for 20-min at -20°C, allowed to air dry, then washed 3 times with PBS and treated with 0.5% Triton X-100 for 5-min at 4°C to permeabilize cells. The coverslips were incubated (30-min, 37°C) with 5 μg/ml of monoclonal anti-lamin B antibody followed by a 30-min incubation with 1:500 Alexa^®^ 488 goat anti-mouse IgG conjugate. To visualize nuclei, the cells were counter-stained with bisbenzimide (0.2 μg/ml) or DAPI (0.12 μg/ml). Cells were viewed on a Zeiss microscope and photographed with Ilford HP-5 film. Western blots ------------- For determination of lamin B cleavage, HCT-116, A-549 and MiAPaCa-2 cells were exposed for 5-days at 10 × IC~50~ each of RPR-130401. At the end of treatment, cells were harvested in PBS with 2 mM EDTA and lysed for 20-min on ice with gentle agitation in PHEM buffer (80 mM PIPES, 80 mM HEPES, 10 mM EDTA, 5 mM MgCl~2~, 1% Triton X-100, pH 7.4) supplemented with 1 mM PMSF, 10 μg/ml aprotinin and 1 μg/ml each of leupeptin and pepstatin. The insoluble fraction, containing lamin B, was recovered by centrifugation, solubilized in Laemmli buffer (3% lauryl sulfate, 10% glycerol, 5% 2-mercaptoethanol, 62.5 mM Tris-HCl, pH 6.8) and titrated. An aliquot (20 μg total protein) of each extract was loaded onto 7.5% polyacrylamide gel, electrophoresed, and blotted onto nitrocellulose (Schleicher & Schuell, Dassel, Germany). The membrane was saturated with bovine serum albumine and incubated with a mouse monoclonal antibody directed against lamin B. Protein detection was carried out using a peroxidase-conjugated goat anti-mouse secondary antibody and an ECL kit (Amersham Pharmacia Biotech, Orsay, France). Quantification was performed by densitometry with the aid of QuantityOne^®^ software (Bio-Rad, Hercules, California). Curve fittings and correction for cellular multiplicity ------------------------------------------------------- Care was taken to avoid contamination of seed cell suspensions by clusters, as their presence introduces large, systematic errors in viability measurements with clonogenic assays \[[@B80]\]. For this reason, the cellular multiplicity, *i.e*., the number of cells per potential colony-forming unit was carefully checked by microscope observation at the time of irradiation throughout the whole study. When the average multiplicity was less than 1.1 (for 1- and 2-h drug exposure), no correction was applied. Corrections were necessary for longer lengths of drug exposure (≥ 3-h), because viable colony-forming units contained a substantial amount of doublets and quadruplets at these times. In that case, the equation:  (where S~exp~ is the experimental cell survival determined from bulk colony scoring and a~n~ the fraction of colony-forming units with n viable cells) was used in order to calculate the single-cell surviving fraction (SCSF). Radiation survival curves with or without co-treatment with RPR-130401 were drawn for best fit to a linear-quadratic equation, as usual \[[@B81]\]:  where S~0~ is the clonogenic efficiency, S the residual survival, D the radiation dose, and α and β numerical parameters characterizing the radiosensitivity of the cell line. Equation (1) was solved with the aid of a home-made program. Least-squares regressions were performed on a Macintosh microcomputer using Kaleidagraph software (Synergy Software, Reading, Pennsylvania). List of abbreviations ===================== FTase, farnesyltransferase; FPP, farnesyl pyrophosphate; FTI, farnesyltransferase inhibitor; GGTase, geranylgeranyl transferase; HBSS, Hank\'s balanced salt solution; BrdUrd, 5-Bromo-2\'-deoxyuridine; PMSF, phenylmethylsulfonyl fluoride; FISH, Fluorescence In Situ Hybridation; IC~50~, amount of drug that reduces the growth of treated cells to 50% of that of controls. Acknowledgments =============== The authors wish to thank Drs. Patrick Mailliet and Jean-François Riou (Aventis Pharma) for the generous gift of RPR-130401. Grafetul thanks are also due to Hervé Tourbez for creating the program used for the correction of cell multiplicity and to Dany Rouillard for flow cytometric analysis. This work was supported by grants from the Association pour la Recherche sur le Cancer (ARC-9746) and the Institut Curie (Radiobiology and Genotoxicology program), and by financial aid from the Institut National de la Santé et de la Recherche Médicale.

Background ========== Clozapine is a widely used atypical neuroleptic with affinity for multiple receptors, including dopamine, serotonin, alpha adrenergic, muscarinic and histaminergic receptors \[[@B1],[@B2]\]. Although clozapine is effective in the treatment of schizophrenia refractory to traditional antipsychotic medication, it also has a number of significant side effects ranging from the potentially fatal but rare agranulocytosis to weight gain, constipation, seizures and urinary incontinence \[[@B3]\]. Urinary problems have been reported in the clinical literature with incontinence present in up to 44% of patients \[[@B4]\] and enuresis in 27% of patients \[[@B5]\]. Since clozapine has potent anti-muscarinic and anti-alpha adrenergic effects \[[@B2]\], it has been proposed that peripheral effects on the lower urinary tract might be responsible for the micturition disturbances \[[@B6]\]. Incontinence has also been reported following therapy with other atypical neuroleptics such as risperidone \[[@B7]-[@B9]\] and recently, olanzapine \[[@B10]\]. We have previously shown that intravenous administration of clozapine to anesthetized rats \[[@B11],[@B12]\] altered several micturition parameters including a decreased amount of volume voided per micturition and a concomitant increase in the residual volume. In addition, clozapine also inhibits the external urethral sphincter. Other, newer, atypical neuroleptics (e.g. olanzapine and risperidone; \[[@B13]\]) were also found to have similar effects but with differing potency compared to clozapine. Since risperidone has little or no anti-muscarinic activity, a primary anti-muscarinic peripheral effect may be ruled out as the cause of micturition disturbances following risperidone administration, and suggest a possible central role for atypical neuroleptics in regulating micturition. The purpose of the present study was to determine the effects of clozapine administered centrally at two different sites, spinal (L6-S1 spinal segments) vs. supraspinal (lateral ventricle), on micturition and the external urethral sphincter during cystometry in the anesthetized rat. By limiting the application site to a specific area of the central nervous system and comparing the results with our previous findings after systemic administration \[[@B12]\] it might be possible to determine whether the effects of clozapine are mediated at a central or peripheral level. Moreover, a comparison of the two central routes might indicate whether the central effect involves spinal or supraspinal structures. We report in this study that most of the effects of clozapine on micturition are mediated by its central effects. Furthermore, there are differences in spinal versus supraspinal effects. Results ======= Effects of central administration of clozapine on urodynamic parameters during single cystometry (Table [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}) ------------------------------------------------------------------------------------------------------------------------------------------------------------------ ::: {#T1 .table-wrap} ::: {.caption} ###### **1.** Effects of intrathecal (L6/S1 spinal level) administration increasing doses of clozapine on urodynamic and EMG parameters in anesthetized rats. Values are Mean ± S.E.M. ::: ***Urodynamic parameters*** ***EUS EMG parameters*** ------------- ----------------------------- -------------------------- ---------------------- ------------------------ --------------- --------------- -------------------------- ------------------------- ------------- --------------------- --------------- ------------------------- **Control** 0.29 (± 0.04) 0.18 (± 0.022) 34 (± 7.5) 3.4 (± 0.46) 16.9 (± 1.32) 13.1 (± 0.82) 2.4 (± 0.30) 1.5 (± 0.10) 100 100 100 100 **0.5** 0.30 (± 0.02) 0.17 (± 0.018) 43 (± 4.1) 3.3 (± 0.79) 17.0 (± 2.00) 13.3 (± 1.20) 1.9 (± 0.55) 1.3 (± 0.26) 98 (± 10.3) 89 (± 26.9) 116 (± 12.0) 84 (± 17.0) **5.0** 0.32 (± 0.06) **0.11^\*\*^ (± 0.031)** **69^\*\*^ (± 6.2)** 3.3 (± 0.49) 22.1 (± 1.49) 12.3 (± 0.66) **0.7^\*\*,a^ (± 0.46)** **0.4^\*\*,b^ (± .23)** 91 (± 20.0) **37^\*^ (± 24.0)** 94.7 (± 21.8) **32^\*\*,a^ (± 20.5)** **50** **0.47^\*\*^ (± 0.09)** **0.08^\*\*^ (± 0.019)** **79^\*\*^ (± 4.3)** **6.0^\*\*^ (± 0.86)** 21 (± 3.39) 13.6 (± 0.72) **0.9^\*^ (± 0.93)** **0.5^\*\*^ (± 0.19)** 93 (± 24) **40^\*^ (± 14.0)** 72 (± 14.8) 66 (± 22.2) ^\*^ p \< 0.05 compared to control; ^\*\*^ p \< 0.01 compared to control; ^a^ p \< 0.05 compared to same dose, different route; ^b^ p \< 0.01 compared to same dose, different route ::: ::: {#T2 .table-wrap} ::: {.caption} ###### **2.** Effects of intracerebroventricular (Lateral ventricle) admnistration of increasing doses of clozapine on urodynamic and EMG parameters in anesthetized rats. Values are Mean ± S.E.M. ::: ***Urodynamic parameters*** ***EUS EMG parameters*** ------------- ----------------------------- --------------------------- ---------------------- ------------------------ -------------- --------------- ------------------------ ------------------------ --------------------- ---------------------- --------------------- --------------------- **Control** 0.25 (± 0.027) 0.16 (± 0.018) 33 (± 7.4) 2.9 (± 0.29) 19.4 (± 1.5) 12.4 (± 0.46) 3.0 (± 0.28) 1.6 (± 0.11) 100 100 100 100 **0.5** 0.26 (± 0.026) **0.11^\*,b^ (± 0.02)** **54^\*^ (± 8.5)** 3.6 (± 0.30) 21.6 (± 3.0) 12.0 (± 0.90) 2.0 (± 0.33) 1.4 (± 0.19) 117 (± 9.4) **68^\*^ (± 13.9)** 95 (± 4.7) 101 (± 7.0) **5.0** 0.30 (± 0.022) **0.11^\*^ (± 0.019)** **62^\*\*^ (± 3.8)** 3.3 (± 0 44) 22.2 (± 25) 12.6 (± 0 73) **1.8^\*^ (± 0.35)** 1.3 (± 0.07) 128 (+15 1) **54^\*\*^ (± 8.4)** 101 (± 10.0) 94 (± 5.5) **50** **0.51^\*\*^ (± 0.11)** **0.084^\*\*^ (± 0.014)** **80^\*\*^ (± 5.2)** **7.2^\*\*^ (± 1.46)** 18.7 (± 1.9) 13.6 (± 0.34) **1.3^\*\*^ (± 0.39)** **0.6^\*\*^ (± 0.20)** **62^\*^ (± 14.9)** **26^\*\*^ (± 9.9)** **50^\*^ (± 14.9)** **65^\*^ (± 17.2)** ^\*^ p \< 0.05 compared to control; ^\*\*^ p \< 0.01 compared to control; ^a^ p \< 0.05 compared to same dose, different route; ^b^ p \< 0.01 compared to same dose, different route ::: Figures [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"} show representative examples of the effects of clozapine on single cystometry in anesthetized rats after intrathecal (i.t.; L6-S1 spinal segment) and intracerebroventricular (i.c.v.; lateral ventricle) administration, respectively. ![Representative traces of the effects of increasing dose of clozapine administered intrathecally (over the L6/S1 spinal segments) on the cystometrogram and external urethral sphincter of anesthetized rats (n = 6). Top panel shows bladder pressure during filling (0.11 ml/min) while bottom panel shows the integrated EMG recorded from the external urethral sphincter. Panels A and B also show some of the urodynamic parameters measured during cystometry: BC = amount of fluid infused into the bladder to elicit a contraction; PT = pressure at which contraction begins; PP = peak pressure during contraction; CT = contraction time; HFO = high frequency oscillations recorded during expulsion; ET = time between peak pressure and end of high frequency oscillations. The EMG activity was examined by dividing the contraction into 3 phases \[[@B19]\]: a contraction phase (Phase 1); an expulsion phase (Phase 2); a closing phase (Phase 3). A) Cystometrogram (CMG) following administration of vehicle (saline) intrathecally. (B) Expanded time scale shows the HFO and the concomitant bursting pattern in the EUS EMG. C & D) 0.5 nmoles of clozapine i.t. does not have an effect on the CMG or the EUS EMG. E & F) 5 nmoles of clozapine i.t. abolished the HFO (4/6 rats) and the bursting pattern in the EUS EMG. G & H) 50 nmoles of clozapine i.t. increased the bladder capacity and also abolished the HFO and the bursting pattern in the EUS EMG of some rats (2/6). Unstable, non-voiding contractions were occasionally seen following clozapine administration and preceding the final voiding contraction (1C; 1G). Calibration bar: A,C,E,G = 1 min; B,D,F,H = 1 sec.](1471-2210-2-6-1){#F1} {ref-type="fig"}). Non-voiding contractions are observed in this animal prior to the final voiding contraction (2G). Calibration bar: A, C,E,G = 1 min; B,D,F,H = 1 sec.](1471-2210-2-6-2){#F2} Bladder capacity was increased by clozapine given i.t. or i.c.v. only at the highest dose tested (Fig [1G](#F1){ref-type="fig"}; [2G](#F2){ref-type="fig"}). After 50 nmoles of clozapine i.t. the bladder capacity was 0.47 ± 0.09 ml, compared to 0.29 ± 0.027 ml after vehicle administration (Fig [3A](#F3){ref-type="fig"}). Similarly, after 50 nmoles of clozapine i.c.v the bladder capacity was 0.51 ± 0.11 ml, compared to 0.25 + 0.027 ml after vehicle administration (Fig [3A](#F3){ref-type="fig"}). ![Effects of increasing doses of clozapine administered intrathecally (i.t.) or intracerebroventricuarly (i.c.v.) on bladder capacity, micturition volume and residual volume. For comparison purposes, the effects of intravenous admnistration of clozapine (data from \[[@B12]\]) are also shown. Doses are presented in a log scale, with vehicle marked as V for the results of the present experiments, and (V) for vehicle in the i.v. doses. A) Bladder capacity increased only after 50 nmoles of clozapine either i.t. or i.c.v.; B) Micturition volume decreased after clozapine i.t. or i.c.v. However, 0.5 nmoles of clozapine i.c.v. significantly reduced the micturition volume and this effect was also significantly different from that obtained after the same dose i.t. C) Residual volume was increased by clozapine i.t. or i.c.v. All doses of clozapine i.c.v. produced significant increases in the residual volume, whereas only 5 and 50 nmoles i.t. produced significant results. ^\*^ = p \< 0.05; ^\*\*^ = p \< 0.01; \# = p \< 0.05 when compared to the effects of the same dose, different route.](1471-2210-2-6-3){#F3} Micturition volume was significantly decreased after 5.0 and 50 nmoles of clozapine i.t. (Fig [3B](#F3){ref-type="fig"}) to 0.11 ± 0.031 and 0.08 ± 0.019 ml, respectively compared to 0.18 ± 0.022 ml after vehicle. In the case of intracerebroventricular administration, 0.5 nmoles of clozapine decreased the micturition volume to 0.11 ± 0.02 ml compared to 0.16 ± 0.018 ml after vehicle (Fig [3B](#F3){ref-type="fig"}). The effect of 0.5 nmoles of clozapine i.c.v. on micturition volume was significantly different from the effect of the same dose given i.t. Further reductions in micturition volume were observed after 50 nmoles of clozapine i.c.v. Residual volume was significantly increased after 5.0 and 50 nmoles of clozapine i.t. to 69 ± 6.2% and 79 ± 4.3%, compared to 34 ± 7.5% after vehicle administration (Fig [3C](#F3){ref-type="fig"}). Clozapine i.c.v. resulted in a significant increase in the residual volume at all doses, including 0.5 nmoles. After 0.5 nmoles i.c.v. the residual volume was 54 ± 8.5% compared to 33 ± 7.4% after vehicle. This residual volume was not significantly different from the one observed after the same dose i.t. Larger doses of clozapine i.c.v. resulted in further increases in the residual volume to 62 ± 3.8 and 80 ± 5.2% after 5.0 and 50 nmoles, respectively. Clozapine i.t. or i.c.v caused an increase in the pressure threshold only at the highest doses tested (Fig [1G](#F1){ref-type="fig"}; [2G](#F2){ref-type="fig"}; [4A](#F4){ref-type="fig"}). After 50 nmoles i.t. the pressure threshold was 6.0 ± 0.86 compared to 3.4 ± 0.46 mm Hg after vehicle, whereas after i.c.v. administration the same dose increased the pressure threshold to 7.2 ± 1.46 mm Hg compared to 2.9 ± 0.29 mm Hg after vehicle. Clozapine i.t. or i.c.v had no effect on either peak pressure or contraction time in the range of doses tested (Fig [4B](#F4){ref-type="fig"},[5A](#F5){ref-type="fig"}) in this study. ![Effect of clozapine i.t. or i.c.v on pressure threshold and peak pressure. Data from intravenous administration \[[@B12]\] are shown for comparison. A) Pressure threshold increased only after the highest dose (50 nmoles) of clozapine either i.t. or i.c.v.; B) Clozapine i.t. or i.c.v. had no effect on the peak pressure. V= vehicle administration (saline) for i.t. or i.c.v. groups; (*V*) = vehicle administration for i.v. group. ^\*^ = p \< 0.05; ^\*\*^ = p \< 0.01](1471-2210-2-6-4){#F4} ![Effect of clozapine i.t. or i.c.v. on contraction time, expulsion time and amplitude of the high frequency oscillations (HFO). Data from intravenous administration of clozapine \[[@B12]\] are presented for comparison. A) Contraction time was not affected by clozapine i.t. or i.c.v.; B) Expulsion time was decreased by both clozapine i.t. or i.c.v. However, 5 nmoles of clozapine i.t. resulted in a greater reduction in expulsion time than the same dose given i.c.v. Only the highest dose (50 nmoles) resulted in a decrease in the expulsion time when given i.c.v. C) The amplitude of the HFO was significantly reduced by 5 and 50 nmoles ofclozapie i.t. but only by 50 nmoles of clozapine i.c.v. The effect of 5 nmoles of clozapine i.t. was significantly different from the same dose given i.c.v. V= vehicle administration (saline) for i.t. or i.c.v. groups; (*V*) = vehicle administration for i.v. group. ^\*^ = p \< 0.05; ^\*\*^ = p \< 0.01; \# = p \< 0.05 when compared to the effects of the same dose, different route.](1471-2210-2-6-5){#F5} Expulsion time was significantly decreased by clozapine at 5.0 and 50 nmoles either i.t. or i.c.v (Fig [5B](#F5){ref-type="fig"}). After vehicle i.t. the expulsion time was 2.4 ± 0.3 sec, which decreased to 0.7 ± 0.46 and 0.9 ± 0.93 sec after 5 and 50 nmoles i.t. Similarly, after vehicle i.c.v the expulsion time was 3.0 ± 0.28 sec and 5.0 and 50 nmoles of clozapine i.c.v decreased it to 1.8 ± 0.35 and 1.3 ± 0.39 sec, respectively. The effect of 5.0 nmoles of clozapine i.t. on the expulsion time was significantly different from the effect observed after i.c.v administration of this dose (Fig [5B](#F5){ref-type="fig"}). Clozapine also decreased the amplitude of (and in some animals even abolished) the high frequency oscillations (HFO) after i.t. administration (Fig [1](#F1){ref-type="fig"}; [5C](#F5){ref-type="fig"}). After 5.0 nmoles of clozapine the HFO were 0.4 ± 0.23 compared to 1.5 ± 0.1 mm Hg after administration of vehicle. In 4 of the 6 animals in this group, 5 nmoles of clozapine i.t. abolished the HFO. The reduction in amplitude observed at this dose after i.t. administration of significantly different from the effects observed after the same dose i.c.v. (Fig [5C](#F5){ref-type="fig"}). A greater dose of clozapine did not cause a larger effect although the amplitude of the HFO after 50 nmoles of clozapine i.t. was still significantly reduced (0.5 ± 0.19 mm Hg) when compared to vehicle administration. Clozapine i.c.v decreased the amplitude of the HFO only at 50 nmoles (0.6 ± 0.2 mm Hg compared to 1.6 ± 0.11 mm Hg after vehicle administration; Fig [2](#F2){ref-type="fig"}), and the HFO were abolished in only 1/6 of the animals in this group. Effects of central administration of clozapine on the activity of the external urethral sphincter during single cystometry (Table [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}) -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Clozapine i.t. did not have an effect on the activity of the EUS during Phase 1 (the rising phase during a contraction), and only the highest dose of clozapine i.c.v (50 nmoles) was found to decrease the activity of the EUS (Fig [6A](#F6){ref-type="fig"}) to 62 ± 14.9% of that seen during administration of vehicle. ![Effects of clozapine i.t. or i.c.v. on the EMG recorded from the EUS during cystometry in anesthetized rats. Data from i.v. administration of clozapine \[[@B12]\] are presented for comparison. A) Clozapine i.t. had no effect on the integrated EMG recorded during phase 1 of the contraction. Only the highest dose of clozapine (50 nmoles) i.c.v. decreased the EMG during this phase to 62%; B) Clozapine i.t. or i.c.v. decreased the integrated EMG during phase 2 of the contraction. All doses of clozapine i.c.v. significantly reduced the EMG whereas only 5 and 50 nmoles i.t. resulted in significant reductions. C) Only the highest dose of clozapine i.c.v (50 nmoles) decreased the EMG during this phase of the contraction. Clozapine i.t. had no effects. D) The amplitude of the individual bursts of EMG occurring during phase 2 of the contraction were reduced by clozapine i.t and i.c.v. However, 5 nmoles of clozapine i.t. produced a reduction of EMG that was significantly different from the effects of the same dose given i.c.v. Only the highest dose (50 nmoles) was effective in reducing the EMG when administered i.c.v. V= vehicle administration (saline) for i.t. or i.c.v. groups; (*V*) = vehicle administration for i.v. group. ^\*^ = p \< 0.05; ^\*\*^ = p \< 0.01; \# = p \< 0.05 when compared to the effects of the same dose, different route.](1471-2210-2-6-6){#F6} Phase 2 of the EUS EMG (occurring during the time of HFO) was decreased by clozapine either i.t or i.c.v. 5 and 50 nmoles of clozapine i.t. decreased the EMG to 37 ± 24 and 40 ± 14% of that observed during administration of vehicle, respectively (Fig [6B](#F6){ref-type="fig"}). In fact, in 4/6 animals, 5.0 nmoles of clozapine i.t. abolished the bursting pattern of the EMG observed during phase 2 (Fig [1F](#F1){ref-type="fig"}). Clozapine administered i.c.v at a dose of 0.5 nmoles decreased the EUS EMG to 68 ± 13.9% of the activity observed after vehicle administration (Fig [6B](#F6){ref-type="fig"}). Larger doses (5 and 50 nmoles) of clozapine i.c.v further reduced the EUS activity during this phase to 54 ± 8.4% and 26 ± 9.9%. Clozapine i.c.v, however, was not observed to abolish the bursting pattern during this phase (Fig [2](#F2){ref-type="fig"}), except in 1/6 animals. Phase 3 of the EUS EMG (recorded during the falling phase of a bladder contraction) was not affected by clozapine administered intrathecally (Fig [6C](#F6){ref-type="fig"}). However, the largest dose of clozapine given i.c.v (50 nmoles) decreased the activity to 50 ± 14.9% of the activity observed during vehicle administration. Finally, the amplitude of the bursts observed during phase 2 in the EUS EMG were decreased to 32 ± 20.5% by 5.0 nmoles of clozapine i.t. (Fig [6D](#F6){ref-type="fig"}). A similar dose of clozapine i.c.v had no effect. However, 50 nmoles of clozapine i.c.v. decreased the amplitude of the bursts to 65 ± 17.2% of that observed during vehicle administration. Discussion ========== In the present experiments, central application of clozapine (intrathecally over the L6/S1 spinal segments or intracerebroventricularly into the lateral ventricle) resulted in a number of changes in the urodynamic parameters of anesthetized rats. The major effect of clozapine was to decrease the voiding efficiency of the bladder by inhibiting expulsion parameters, such as micturition volume, residual volume and expulsion time. In addition, the activity of the EUS also decreased upon central application of clozapine. One problem in delivering substances centrally is the possibility that peripheral spread may confound the results. Since clozapine crosses the blood-brain barrier readily \[[@B14]\], the maximum dose selected in the present experiment was restricted to the first appearance of a decrease in blood pressure as an indication of possible peripheral alpha 1 antagonism due to peripheral leakage. We compared the effects obtained in the present study against those observed previously after intravenous administration \[[@B12]\] expecting that central effects would require significantly less application of clozapine than systemic administration. We considered a minimum difference of 10× magnitude in the central vs. peripheral dose that elicited a significant effect as an indication of possible central action. Our doses of 0.5, 5 and 50 nmoles of clozapine correspond to serum levels of 8, 80 and 800 ng/ml (assuming a blood volume of 8% of body weight) and are in the range of therapeutic levels (260--387 ng/ml \[[@B5],[@B15]\]). A second problem in administering a substance at a particular central location is that of redistribution to other parts of the central nervous system. Since application of a substance at the lumbosacral spinal level may travel to the brain after some time and vice versa, we established two criteria to help determine a possible central site of action: 1) the first dose to elicit a significant effect by either i.t. or i.c.v administration; 2) significant differences at the same dose but different routes (i.t. vs i.c.v). Figure [7](#F7){ref-type="fig"} is a summary of our findings when interpreted in light of the criteria stated above for determining (a) central vs peripheral effects and (b) spinal vs supraspinal site of action. The effects are presented as percent change over the control (vehicle) dose, for each of the central routes in the present experiment and our previously reported findings for systemic administration \[[@B12]\]. Theses changes are plotted at the dose (mg/kg) that first yielded significant results. The route that we consider most likely to be the site of action for clozapine is italicized. ![Summary of the effects of clozapine given i.t., i.c.v. or i.v. (data from \[[@B12]\]) on urodynamic parameters and the activity of the EUS during cystometry in anesthetized rats. The first dose to yield a significant effect has been plotted for all 3 routes. Doses have been converted to mg/kg for ease of comparison. Peak pressure data are not shown since no significant differences were found with clozapine after any route. A) Bladder capacity is increased to a similar extent by comparable doses of clozapine, regardless of the route. Therefore, a peripheral effect may be predominating. In the case of micturition volume (B) and residual volume (C), both clearly show a central effect since the first dose to yield a significant effect is 142× smaller than the i.v. dose. Both of these parameters show a possible supraspinal site of action. D) Pressure threshold increased only after the highest dose of clozapine i.c.v or i.t., however, that dose was still 14× smaller than the first dose i.v., suggesting a possible central effect. E) Expulsion time shows a 142× difference between the central dose and the peripheral dose. In addition, intrathecal administration shows a larger effect than after i.c.v., suggesting a possible spinal site of action. F) The effect of clozapine on the amplitude of the HFO appears to be mediated by a spinal site. G) Phase 1 of the EMG was decreased after equivalent doses of clozapine i.v. or i.c.v. Intrathecal administration had no effect. H) Phase 2 of the EMG (where the bursting occurs) clearly showed a central effect since a much smaller dose (142×) was required to produce an effect. In addition, the reduction after i.c.v. was greater than after i.t. therefore suggesting a possible supraspinal site of action. I) Phase 3 of the EMG was not affected by clozapine intrathecally. Only the highest dose of clozapine given i.c.v decreased the EMG during this phase. That dose was 14× smaller than the first dose to produce a significant effect after i.v. administration suggesting a possible supraspinal site of action. J) The amplitude of the individual bursts during phase 2 of the EMG was decreased by central effects of clozapine at smaller doses than those observed i.v. Also, the intrathecal administration was much more effective in reducing the amplitude (abolished in 4/6) animals, suggesting a possible spinal site of action.](1471-2210-2-6-7){#F7} Bladder capacity was increased to a similar degree by an equivalent dose of clozapine, regardless of route (Fig [7A](#F7){ref-type="fig"}; [3A](#F3){ref-type="fig"}), making a determination of most likely site of action difficult. Given that the peripheral effects at low doses were similar to the effects after the highest central dose of clozapine (by either route) it appears likely that this effect is due to peripheral actions of clozapine. Bladder capacity was reported to increase following i.c.v. administration of muscarinic antagonist (atropine \[[@B16]\]) or i.t. alpha2 antagonists (yohimbine \[[@B17]\]; atipemazole \[[@B18]\]). Given clozapine\'s strong affinity for muscarinic and alpha2 receptors \[[@B2]\] it is possible that central effects are also contributing to the increase in bladder capacity. Micturition volume, on the other hand, showed a reduction after smaller doses of clozapine i.c.v or i.t. when compared to the i.v. dose (Fig [7B](#F7){ref-type="fig"};[3B](#F3){ref-type="fig"}). Therefore, central effects of clozapine in controlling micturition volume are likely. Furthermore, when comparing the effects of i.t. vs. i.c.v administration, a 5 nmoles dose of clozapine i.c.v. elicited a reduction in micturition volume that was significantly greater from the effects observed after the same dose i.t. (Fig. [3B](#F3){ref-type="fig"}). Therefore, supraspinal effects of clozapine in regulating micturition volume are likely with spinal effects perhaps contributing. Clozapine i.t. also decreased the micturition volume, however the dose that first showed a significant effect was higher than the i.c.v. but smaller than the i.v. dose. Spinal antagonism of either alpha1 or alpha2 adrenergic receptors was reported to increase micturition volume \[[@B17]\] in anesthetized rats. However, intrathecal atipamezole (alpha2 antagonist) increased residual volume in awake rats \[[@B18]\]. Therefore, it is difficult to interpret clozapine\'s effects on micturition volume in terms of spinal alpha adrenoceptor antagonism and possibly the supraspinal effects predominate. Clozapine given i.t. or i.c.v. also increased the residual volume (Fig [7C](#F7){ref-type="fig"}; [3C](#F3){ref-type="fig"}). Since the effects were observed at doses that were lower than those after i.v. administration, a peripheral effect of clozapine to increase residual volume appears unlikely. The smallest dose of clozapine given i.c.v. (0.5 nmoles) increased residual volume to 160% (Fig [7C](#F7){ref-type="fig"}), suggesting that supraspinal effects predominate with possible contributing effects from spinal sites. Ishiura et al. \[[@B16]\] reported a decrease in voiding efficiency (comparable to an increase in residual volume) following atropine i.c.v. or i.t. suggesting that muscarinic receptors at supraspinal and spinal sites are involved and may account for our effects after clozapine i.t. or i.c.v. A decrease in residual volume has been reported following spinal antagonism of alpha2 receptors with yohimbine \[[@B17]\] but atipamezole produced an increase in the residual volume \[[@B18]\], similar to our findings with clozapine i.t. Clozapine increased pressure threshold after i.c.v. or i.t. administration only after 50 nmoles (Fig [7D](#F7){ref-type="fig"}; [4A](#F4){ref-type="fig"}). It should be noted that both central doses were at the range observed to result in cardiovascular changes and therefore the possibility of peripheral leakage of clozapine must be considered. Still the effective central doses are approximately 14× less than the first dose observed to produce significant results intravenously, and suggests a possible central site of action. Peak pressure was not changed by clozapine after either route in the present study after central administration, consistent with our findings afer intravenous administration \[[@B12]\]. Ishiura et al. \[[@B16]\] reported a decrease in maximal voiding pressure (equivalent to our peak pressure) after atropine i.v., i.c.v or i.t. in awake rats undergoing continuous cystometry. Given that clozapine has a relatively high affinity for muscarinic receptors \[[@B1],[@B2]\] it is surprising that we have not seen an effect on peak pressure. Clozapine, and also olanzapine \[[@B11],[@B13]\] were able to decrease the contraction amplitude after electrical stimulation of the pelvic nerve. Since the contraction pressures observed during cystometry were smaller than those observed after electrical stimulation of the pelvic nerve (and against a closed urethra) it is possible that the antimuscarinic effects of clozapine on bladder contraction pressure are not detected because during cystometry maximal bladder pressures are not necessary for emptying. Contraction time was not affected by clozapine i.c.v or i.t. (Table [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}; Fig [5A](#F5){ref-type="fig"}) however it was clearly decreased after intravenous administration \[[@B12]\]. Therefore, we suspect that solely peripheral effects may explain the effects of clozapine on contraction time. Antimuscarinic agents have been show to decrease contraction time (e.g. atropine \[[@B19]\]) and since clozapine has high affinity for muscarinic receptors it is likely that peripheral anti-muscarinic effects of clozapine are responsible for the reduction in contraction time after systemic administration. Expulsion time was reduced by clozapine after i.t. or i.c.v. administration at much smaller doses than were observed to produce similar results i.v. (Fig [7E](#F7){ref-type="fig"}; [5B](#F5){ref-type="fig"}). Therefore, central effects of clozapine are likely responsible for the reduction in expulsion time. Since the intrathecal effects were greater than those observed after i.c.v. administration it is possible that a spinal action predominates with possible contribution from supraspinal sites. The amplitude of the high frequency oscillations observed during the expulsion time in the rat micturition cycle \[[@B20],[@B21]\] was reduced by clozapine i.t. at a dose of 5 nmoles (Fig 8F). This dose was 10× smaller than the first i.c.v. dose observed to have a significant effect and approximately 142× smaller than the first i.v. dose to show a significant effect. Therefore, a spinal site mediating the effects of clozapine on the amplitude of high frequency oscillations appears likely. Previous results \[[@B12]\] suggest that D2 receptors modulate the amplitude of the HFO, since raclopride (selective D2 antagonist) decreased the amplitude of the HFO by 30%. In addition to clozapine\'s effects on urodynamic parameters, the EMG recorded from the external urethral sphincter also showed changes after clozapine i.t. or i.c.v. The EMG during phase 1 of the contraction only showed a reduction after the largest dose of clozapine i.c.v.. This dose is equivalent to our lowest dose i.v. and in fact all doses i.v. previously showed a significant reduction \[[@B11]\]. Intrathecal administration had no effect on the EMG at phase 1. The EMG during phase 2, corresponding to the period of high frequency oscillations was decreased by clozapine after i.t. or i.c.v administration at doses that were much lower than those necessary to produce an effect after i.v. Therefore, the effects of clozapine in reducing the EMG during this phase appear to be central in origin. Since the first significant effect was obtained after i.c.v. administration (Fig [7H](#F7){ref-type="fig"}) it is possible that supraspinal effects predominate with spinal sites contributing. The EMG during phase 3 (closing phase) of the contraction was decreased only after the highest dose of clozapine (50 nmoles) i.c.v. (Fig [6C](#F6){ref-type="fig"}). This dose was still 14× smaller than the first significant dose i.v. and therefore a central effect of clozapine is probable. Finally, the amplitude of the individual bursts of EMG recorded from the external urethral sphincter during phase 2 of the contraction also decreased after clozapine i.t. or i.c.v. (Fig [6D](#F6){ref-type="fig"}). Since the doses that yielded significant reductions were smaller than those observed after i.v. admnistration, a central effect is likely. Moreover, since 5 nmoles of clozapine i.t. produced a significant reduction when compared to the same dose i.c.v., a spinal effect of clozapine in mediating the reduction of the burst amplitude of the EMG is possible. In the anesthetized rat, the external urethral sphincter contracts and relaxes during the expulsion phase in a manner that is complimentary to the high frequency oscillations observed in the bladder pressure record \[[@B19]\] whereas in humans the external sphincter not active during voiding. Pharmacological blockade of the external urethral sphincter in the rat resulted in decreased micturition volume and increased residual volume \[[@B20]-[@B23]\] suggesting that the activity of the external urethral sphincter contributes to efficient voiding in the rat. Therefore, central administration of clozapine, by reducing the activity of the EUS, contributes to the decrease in voiding efficiency by reducing micturition volume and increasing residual volume. Alpha1 adrenergic antagonists have been shown to inhibit pudendal reflexes in anesthetized cats \[[@B24]-[@B26]\]. However, a systemic dose of prazosin (alpha1 antagonist) did not inhibit the EUS EMG activity during high frequency oscillations in the rat \[[@B27]\]. Thus, although alpha1 antagonism has been shown to modulate pudendal reflexes in the cat, their role in modulating the activity of the EUS during micturition in the rat appears unclear. In summary, our results in the present experiments suggest that most of the effects of clozapine on urodynamic parameters can be ascribed to central effects. Expulsion parameters, such as micturition volume, residual volume, expulsion time, and amplitude of the high-frequency oscillations, appear to be mediated by the central action (spinal or supraspinal) of clozapine. The activity of the EUS also appears to decrease after central application of clozapine. Therefore, central effects of clozapine appear to decrease the voiding efficiency of the bladder in the rat. Contraction time clearly showed a peripheral effect only, whereas changes in bladder capacity were difficult to explain from central effects and probably reflect peripheral effects of clozapine. Clozapine is metabolized mainly at the liver resulting in several metabolites \[[@B5],[@B14]\]. One of the major metabolites, N-desmethylclozapine has been shown to have pharmacological activity both in vitro \[[@B28]\] and in vivo in rats \[[@B29]\]. In addition, N-desmethylclozapine is found in large concentrations in the serum of schizophrenic patients \[[@B5],[@B15]\] and in rats \[[@B30]\]. The contribution of N-desmethylclozapine to clozapine\'s central effects has been questioned recently, since the levels of N-desmethylclozapine in the brain were much lower than those of clozapine \[[@B30]\] suggesting that N-desmethylclozapine does not cross the blood-brain barrier as readily as clozapine. Since we observed effects from central application of clozapine, we consider it unlikely that the effects of metabolites contributed significantly. However, whether any of the major clozapine metabolites also have a role in regulating micturition remains to be determined. Conclusions =========== Atypical neuroleptics are useful in treating patients that are refractory to \"traditional\" antipsychotic medication and produce fewer extrapyramidal side effects. However, other side effects still occur with varying severity and frequency \[[@B31]\] and continue to pose a challenge to effective treatment. Urinary disturbances as a result of clozapine therapy have been well documented, and include incontinence, enuresis and urgency \[[@B4],[@B5],[@B32]\]. Other atypical antipsychotics, such as risperidone \[[@B7]\] and olanzapine \[[@B10]\] have been reported to produce urinary incontinence. We have previously shown \[[@B11]-[@B13]\] that systemic administration of clozapine, olanzapine and risperidone to anesthetized rats reduced voiding efficiency. Risperidone had smaller maximal effects than olanzapine and clozapine and had no direct (peripheral) inhibitory effects on the amplitude of bladder contractions. In the present study we show that clozapine acts at supraspinal and spinal sites to inhibit certain urodynamic parameters and the external urethral sphincter of the rat resulting in decreased voiding efficiency. If these effects also occur in patients, they may contribute to the urinary disturbances reported following clozapine therapy. The exact receptor types (or combination of receptor types) responsible for clozapine\'s central effects on micturition were not investigated in the present study. However, isolating particular receptors that contribute to the effects of clozapine might be useful in designing neuroleptics that may avoid these side effects or in providing an adjunct therapy to relieve some of the side effects. Materials and Methods ===================== Surgical procedures ------------------- The experiments were conducted in compliance with the USDA Animal Welfare Act and amendments thereto and the revised Guide for the Care and use of Laboratory Animals DHEW (NIH) and were approved by the Animal Studies Subcommittee of the Bay Pines Veterans Administration Medical Center. Surgical procedures have been described in detail elsewhere \[[@B12]\]. Rats (female Sprague-Dawley; n = 16; 230--270 g; Harlan; IN) were anesthetized with halothane and placed on a heating pad. A catheter (PE-50) was introduced into the jugular vein to administer urethane (1.1 g/kg) over a period of 20 minutes while decreasing the level of halothane to prevent respiratory depression. Rats were instrumented with either an intrathecal cannula placed over the L6/S1 spinal segment or a cannula into the right lateral ventricle. An incision was made over the dorsal aspect of the neck and the overlying muscles were retracted to expose the atlanto-occipital membrane. An intrathecal catheter (PE 10) was introduced through a small slit in the atlanto-occipital membrane and positioned over the L6/S1 spinal cord segments \[[@B33]\]. Saline soaked gelfoam was placed around the catheter and the neck muscles and skin were sutured. Spinal segmental location of the catheter was verified post-mortem by performing a laminectomy. Following a small craniectomy, a stainless-steel cannula (27 ga) was placed in the lateral ventricle at the following coordinates: AP - 1.0, ML = 1.2; V= 3.2 mm \[[@B34]\]. The cannula was held in place with skull screws and dental acrylic. At the end of the experiment, 5 μl of fast-green (1%) was infused through the cannula while observing the CSF though a small slit in the atlanto-occipital membrane. Almost immediate visualization of the fast-green in the fourth ventricle was taken to indicate appropriate placement of the cannula into the lateral ventricle. After an abdominal incision, both ureters were tied distally and cut centrally and allowed to drain onto cotton gauzes that were directed outside the animal. A catheter (PE-90) was introduced into the bladder dome and tied in place with a purse string suture. A catheter (PE-50) was introduced into the right femoral artery for blood pressure recording. Stainless-steel wires (0.003 in.; A-M Systems; WA) insulated except at the tip were introduced into the external urethral sphincter for EMG recording (DAM 50; WPI; bandwidth= 3 to 3 kHz; gain: 1000--10,000). Urodynamic studies ------------------ In pilot animals, the dose and duration of effects was determined by administering clozapine (saline, 0.5,5,50,100 nmoles) at 10 min intervals either intrathecally (n = 2) or intracerebroventricularly (n = 2) during continuous cystometry (infusion rate = 0.11 ml/min) while recording bladder pressure, external urethral sphincter EMG, and blood pressure. Fifty (50) nmoles of clozapine administered by either route resulted in a decrease in arterial pressure (mean decrease in MAP was 20 ± 2 and 15 ± 3 mm Hg for intrathecal and intracerebroventricular administration, respectively. The onset time ranged from 1--1.5 min from the start of the infusion). Since clozapine has alpha1 antagonist effects \[[@B2]\] it is possible that the blood pressure decreases were due to spread of clozapine into the periphery following central administration. Doses smaller than 50 nmoles did not elicit a drop in arterial pressure. In addition, 50 nmoles represents a dose close to the smallest intravenous dose used previously (0.1 mg/kg \[[@B12]\]). Therefore, during single cystometry, the dose range was limited to 0.5 to 50 nmoles in order to reduce possible peripheral spread of clozapine. Single cystometry studies were conducted as follows. The bladder was emptied and allowed to equilibrate to air pressure for 5 minutes before beginning each cystometrogram. Room temperature saline was infused into the bladder (0.11 ml/min) while recording bladder pressure and the infusion was stopped when a contraction occurred. Volume expelled was determined by placing cotton gauze at the urinary meatus and weighing before and after micturition. External urethral sphincter EMG (EUS-EMG) was recorded throughout the cystometrogram and for some time after the filling had stopped. Increasing doses of clozapine (Sigma, vehicle, 0.5, 5, 50 nmoles in a volume of 5 μl; followed by a 7 μl saline wash) were administered through the intrathecal catheter (n = 6; Mean weight = 245 gm) or through the intracerebroventricular cannula (n = 6; Mean weight = 245 gm) at approximately 10-minute intervals. Clozapine was dissolved in a minimal amount of 0.1 N hydrochloric acid, and brought up to volume with saline (final pH = 6). Cystometrograms were started approximately 3 minutes after each drug administration. At the end of the experiment the rat was euthanized with an overdose of urethane (3.0 mg/kg; i.v.). Data analysis ------------- Bladder pressure, EUS-EMG and blood pressure during the single cystometrograms were displayed in an electronic chart recorder (RC Electronics; Goleta, CA) and analyzed off-line (Dataview, W.J. Heitler, U. St Andrews, Scotland). The following parameters were examined from the cystometrogram as described in detail earlier \[[@B12]\]: bladder capacity (amount of fluid infused to elicit a contraction); micturition volume (amount of fluid expelled); residual volume (\[bladder capacity-micturition volume\]/\[bladder capacity\] ×100); pressure threshold (pressure at which contraction begins); peak pressure (maximal pressure during contraction); contraction time; expulsion time (time between peak pressure and end of high frequency oscillations); amplitude of high frequency oscillations. The EMG activity was examined by dividing the bladder contraction into three phases in a modification of the technique of Chien et al. \[[@B19]\] a contraction phase (phase 1); an expulsion phase (phase 2) and a closing phase (phase 3). The raw EMG was rectified, integrated (0.5 second bin) and the area under curve of the EMG corresponding to each phase of the bladder contraction was measured (Sigma Scan/Image; Jandel Scientiflcs, San Rafael, Ca). Drug effects for the EMG were calculated as percent of control. Values are presented as Mean + S.E.M. Repeated measures ANOVA (GB Stat; Dynamic Microsystems; MD) were performed on all parameters and when statistical significance (p\<0.05) was obtained, comparisons between control and different drug dosages and between different routes (i.t. vs. i.c.v.) were made using Fisher\'s protected t-test \[[@B35]\] . List of Abbreviations ===================== CMG = cystometrogram EMG = electromyogram EUS = external urethral sphincter HFO = high frequency oscillations i.c.v. = intracerebroventricular i.t. = intrathecal i.v. = intravenous Acknowledgements ================ Gary A. Smith, Jr. provided excellent technical assistance. This work was supported by VA Merit Review Awards (PLV; IN) and by the Paralyzed Veterans Administration (AMS).

Background ========== In the era of aging populations and rising incidence of progressive diseases in industrialized countries, health politicians in these countries are increasingly faced with the decision what fraction of their budgets to spend on treatment and cure or on avoiding (further) health decline. Therefore, health economic analysis aimed at informing policy makers and supporting resource allocation decisions has to evaluate not only improvements in health but also prevented decline. The benefit of these \"upward-movements\" on a health scale compared to avoided \"downward-movements\" is often calculated as the potential gain in QALYs (\"quality-adjusted life years\"). Traditionally the standard gamble (SG) or the time trade-off (TTO) method is used to determine the quality of life weights needed to estimate the benefit of healthcare interventions in terms of QALYs. The use of these techniques in resource allocation decisions has been criticized as of two concerns, among others \[[@B1],[@B2]\]. First, because the benefit of an intervention is calculated as the difference in preferences for two health states initially valued *isolated* from each other \[[@B3],[@B4]\]. The second concern is that the use of SG and TTO in policy decisions assumes that preferences an individual has regarding his own (hypothetical) health equal the preferences this individual has for the health of others, therefore being unaffected by distributional concerns \[[@B5],[@B6]\]. Consequently, the person trade-off (PTO) method has been proposed because it forces respondents to take a social rather than a personal perspective in evaluating rationing frameworks \[[@B7]\]. As in the PTO *movements* on a health scale rather than distinct health *states* are presented, and it thus explicitly focuses on changes in health, the PTO seems to capture some of the factors that researchers have found to be important for the public in distributive decisions. Two of these are the severity of the initial health state and the potential for health \[[@B8]-[@B10]\]. The former describes the tendency to give priority to the worst off even if the associated health gain is smaller. The latter describes the aversion to discriminate against those with a limited potential for health as of chronic conditions or disabilities. This experimental study addresses a different factor influencing distributional concerns of the public, that is, the *direction* of a change in health. In health economic evaluative studies the benefit of avoiding a health state is traditionally calculated exactly as the inverse equivalent of illness, and therefore -- on a conceptual level -- as the same as cure or treatment. This may be of special relevance to the evaluation of prevention programs, where benefit is derived from reducing the likelihood of a specific health state thought to occur in the future. However, despite the effect of the direction of changes in health itself, i.e., avoided losses versus improvements, \'prevention\', as the term is commonly being used, also involves preferences towards time and uncertainty. This study focuses only on the equivalence of equally sized health effects achieved either by curative or preventive treatments, ignoring risk aversion and discounting of time effects. Imagine, for example, illness X with an associated health state x (which is worse than the top anchor \'perfect health\'). If there is a cure for X that will return diseased persons to perfect health, the benefit of cure is 1-x, assuming the conventional 0 to 1 utility scale and not taking the duration of illness into account. In case there is also the opportunity to avert X in people healthy at present, the benefit of avoiding health state x is also given by 1-x. The value people attached to a move from a health state worse than perfect health upwards is used interchangeably with the value of avoiding a move from perfect health downwards this health state. The underlying hypothesis is that the *distance* between two health states is the main carrier of value irrespectively of the direction of the movement and the mode by which it is achieved. This common method of benefit estimation has been applied to various prophylactic treatments and cost-per-QALY-estimates of avoiding health decline are compared not only to treatment of the disease under study but also to treatments of other diseases in cost-utility-ratio league tables \[[@B11]\]. However, there is very limited evidence on the value people attach to treatment and avoided decline when involved gains and losses are directly compared to each other in resource allocation contexts. If it was the case that the distance between two health states is perceived as being of equal value no matter whether this distance is prevented downwards or gained upwards, one would expect that people reveal answers close to \"100\" when faced with a PTO question like: \"In how many people a decline to health state X had to be prevented to make you indifferent to treating 100 persons in health state X and returning them to complete health\"? The available evidence and a number of rationales cast doubts on the hypothesis that curative and protective interventions would be valued as equivalents. Ubel et al. surveyed 289 prospective jurors in the US how to allocate funds on different healthcare projects for nursing home residents with varying levels of disability \[[@B12]\]. Healthcare activities were described as having the same magnitude of benefit either by improving the level of functioning or preventing further decline. While there was a general attitude towards prevention, when the strength of preference, measured on a five-point scale, was taken into account, there was no statistically significant preference in favor of prevention. In a survey conducted among the Swedish general public, participants were asked to trade-off the value of different programs saving varying numbers of lives either by acute or by preventive care \[[@B13]\]. Again, responders favored prevention over acute care with a mean of 1.2--1.4 lives saved by acute care being judged equivalent to 1 life saved by preventive care. Overall, there seems to be, if at all, a slight trend to prioritize towards preventive health care. Both surveys presented choices among certain outcomes close in time, thereby focusing mode of intervention and the direction, in which changes in health are achieved. There are a number of rationales for people expressing differing values for cure and avoidance of decline. As Johannesson and Johansson point out, giving priority to those in a state worse than perfect health already suffering from disease, may be explained by equity concerns \[[@B13]\]: Refusing treatment to individuals in a more severe initial health state and prioritizing those fortune to be healthy might be judged as unfair and people might therefore place more value on cure. Instead, one could argue that persons healthy at the moment experience a *loss* when they fall sick, while those presently suffering already have gone through that loss. The disutility of the process of loss itself might be considered more important than the process of gain itself, i.e., recurring to health. One could see parallels to the observation that patients often attach higher values to the health state they are experiencing than the general public \[[@B14]\]. If subjects in impersonal allocation tasks would simultaneously adopt both perspectives, this could explain that a greater value would be attached to avoiding decline as compared to treatment since healthy subjects obviously \"lose more\" than patients win. In addition, subjects saved from developing a disease or decline might be expected to live longer or in overall better quality of life, and interventions for avoidance of disease might be perceived as less harmful or more effective. There are also a number of psychological explanations why people would prioritize preventing decline compared to cure, for example, as of an impersonal equivalent to \"status quo\"- and \"endowment\"-effects \[[@B15]\]. Simply speaking, opting for protective treatments would keep the current situation unchanged for both the ill and the healthy. The main purpose of this paper is to investigate the value people place on equidistant health changes, either achieved by preventive or curative care, when directly compared in resource allocation decisions. A valuation method that captures this relationship best is to be identified. In addition to the common PTO technique, two new PTO frames will be applied which explicitly ask for the value of avoided health decline. A new idea for the PTO will also be introduced, which is intended to combine preferences for positive, and avoided negative health changes. This will be called the \"weight for avoided decline\" (WAD). As \"the coefficient of loss aversion\" in Prospect Theory, the WAD is interpreted as the \"currency\" by which effects in outcome are weighted and transformed when perceived as either a loss or a gain \[[@B16]\]. This conversion factor will be used to incorporate people\'s preferences in allocating resources among diseased patients awaiting treatment and healthy persons that could be saved from decline. Finally, the accuracy of the different valuation techniques will be assessed by comparing a prioritization ranking of health interventions directly obtained in a hypothetical allocation scenario with rankings predicted by the different valuation techniques. A second purpose of the study is to test whether utilities, elicited by the traditional standard gamble technique (SG) can be used to appropriately reflect social valuations of movements between health states and the direction in which these are experienced, as assumed in common health economic practice. Methods ======= Health states and scaling ------------------------- Six health states labeled A-F (with increasing severity) were prepared, which are described in table [3](#T3){ref-type="table"}. In a pre-test with 25 students from the University of Witten/Herdecke, detailed descriptions of health states covering various symptoms were used. Extreme variance in the subjective ranking of health states was observed, e.g. between sexes. Therefore, it was decided to use very simple and mostly additive health problems implying an obvious ranking of health states. The worst health state (F) had to be different from the conventional absorbing \"death\"-state. As values for the same distance between two health states had to be elicited either experienced as an improvement or as avoided decline, all health states had to be constructed so that hypothetical patients could both move into and out of each health state. Therefore, an irreversible health state could not be used. In addition to this methodological reason, there is a deeper rationale: Because life and death differ in their absolute meaning, preferences towards health care resource allocation involving both categories are likely to follow and mix distinctive ethical rules as if only one category was concerned. Though health state F might be perceived as worse than death by respondents, it differs from the traditional \"dead\"-state bottom anchor in its qualitative category, since \"worse than death\" describes a point on the quality-of-life continuum, different to the irreversible, \"unmodifiable\" and final character of death. For example, the rule of rescue in its narrower sense, namely the societal duty to save lifes in fatal situations, applies per se only in cases where the avoidable death of subjects is involved. ::: {#T1 .table-wrap} ::: {.caption} ###### Health states as described in the questionnaire ::: Health state Description -------------- ------------------------------------------------------------------------------------------------------------------------ A\) People in this health state are in complete health. B\) People in this health state have problems with the knee joints. C\) People in this health state have problems with the knee joints and suffer from Asthma. D\) People in this health state have problems with the knee joints, suffer from Asthma and are partially sighted. E\) People in this health state suffer from Asthma, are weak sighted and need a wheelchair. F\) People in this health state are completely dependent on others. They suffer extreme pain and are unconscious at times. ::: Readers should note this special scaling using a reversible lower end point because it makes the comparison between the PTO and SG scores reported herein with those found using the conventional scale (with death as the bottom anchor) inappropriate. Questionnaire instrument and valuation procedures ------------------------------------------------- An interactive survey was developed which was published in the World Wide Web. The survey was written in HTML and JavaScript. The questionnaire consisted of seven sections. They are described in some detail to make the reader familiar with the analytical steps and valuation procedures: 1\) The first part of the survey introduced all health states to respondents. Subjects were asked to rank health states according to their severity. This was intended to make them familiar with all health states. 2\) The conventional standard gamble method (SG) was presented for health states B to E after explaining the task. Instead of the usual definition of risk as death from treatment, risk in this study was defined as the probability of moving to health state F. The standard gamble was used as of two reasons: First, it was expected that the assessment of health states from a personal perspective would increase reflection and seriousness of participants. Second, results of the SG task were used to investigate the difference between personal (SG), and social (PTO) evaluation of health states. Following common health economic practice, SG utilities were also used to estimate the value of avoiding a health state by taking their inverse on the utility scale. For example, the benefit of avoiding a move from health state B downwards to health state C would be calculated as  where  is the utility of not moving from B to C and *U(B)* is the utility of being in health state B. Subsequently, the accordance of these values with those elicited by the other techniques was examined. 3\) After completion of the four SG questions, respondents were asked to imagine that they were health authorities whose task it is to decide which healthcare interventions to offer to the public. They were explicitly told that their budget was not sufficient to offer every treatment that people could benefit from and that they consequently have to compare and evaluate the benefit of the different options. It was stated that patients were equal in all other respects except the described health conditions. The section started with a PTO like choice that was only intended to make subjects familiar with the task and avoid start point bias \[[@B17]\]. These results were not used in data analysis. The introduction was followed by three PTO sections: The first contained four pair-wise comparisons between curative treatments. In the following, this is called PTO-UP, because choices had to be made between two treatments that offer *improvement*s in health. All patients were portrayed as suffering from the respective disease and their health status would not change unless they were provided the treatment, which would certainly return them to complete health (state A). The frame used was originally presented by Pinto Prades \[[@B18]\]: Starting with the PTO task applied to the worst health state compared to the next best health state, subsequent comparisons were made between a health state and the bordering, next best health state. As illustrated in Figure [1](#F1){ref-type="fig"}, first the improvement from health state F to A (arrow 1) was compared to the improvement from E to A (arrow 2), followed by the comparison between the gain from health state E to A and D to A (arrow 3) and so on. The advantage of this incremental approach is that it might be cognitively and morally easier to handle health states that are not too different from each other. However, within this approach, additivity of benefits is assumed. Again, to calculate the value of not moving from B to C, first the value of moving from B to A as compared to a move from C to A was assessed, which is given by {#F1} 100 *U*(*C* → *A*) = *x U*(*B* → *A*) (2.1) (The value of *U*(*C* → *A*) is known from the previous comparisons.) In the next step, the results of (2.1) were then used to calculate  4\) The next section of the survey continued with the same procedure except that now four PTO choices between protective interventions for people healthy at present were offered. Respondents were told that people would certainly fall ill in the near future, if they did not receive the preventive treatment. That is, two avoided downward-movements were presented. For example, first an avoided decline from health state A to health state F (Fig. [1](#F1){ref-type="fig"}, arrow 4) was compared to an avoided change from health state A to health state E (Fig. [1](#F1){ref-type="fig"}, arrow 5), and this was continued with adjacent health states. This method is called PTO-DOWN. The same eliciting procedure as in PTO-UP was used. The benefit of avoiding a move from B to C was calculated as the difference between an avoided move from A to C and A to B, namely  5\) Subsequently, five PTO choices between curative treatments for diseased persons and treatments saving persons from (further) health decline were presented. While the distance on the health scale was equal in both options, they differed in the direction of the potential health effect. For example, a choice between a treatment which would return 100 persons in health state F to complete health was compared with the opportunity to protect *X* persons in complete health (state A) from entering health state F (Fig. [1](#F1){ref-type="fig"}, arrows 6 and 7). The purpose of this procedure was to assess directly the rate at which health gains and avoided losses are traded against each other. This \"weight for avoided decline\" (WAD) for the distance between health states B and C is defined as:  If, for example, participants indicated by their answers, that they were indifferent between improving the health of 100 patients from health state C to B and avoiding decline from B to C in 80 persons, then *WAD*~B,C~ = 1.25. The size of the coefficient and whether it is constant, i.e., independent of the reference level, is the focus of interest. While a WAD higher than 1 indicates that people value avoiding decline over health improvements, a WAD lower than 1 implies that respondents prefer the latter to the former. With a WAD equal to 1, curative and protective treatments carry the same value, as indirectly assumed when taking the inverse value of health improvements as the value of avoided decline. While the size of WAD reveals whether individuals favor avoiding decline over health improving treatments or the opposite, it cannot be used directly for the comparison of interventions differing in both, direction *and* size of a health effect. Whereas an individual\'s WAD implies, e.g., preferences for prevention over cure, these preferences need also be related to health effects of different magnitude. That is, preferences for health improvements or avoided decline need to be traded against their relative size. Such information would be needed, for example, if a decision maker faces a decision on whether to fund curative care returning patients in HS C to complete health or to fund preventive care that avoids decline to HS E in patients currently in HS B. This makes it necessary to accommodate individuals\' preferences towards curative and preventive effects, measured in terms of WAD, as a special factor in the conventional measurement of value. One way to achieve this and make values for health improvements and avoided decline compatible, i.e. quantified on the same scale, is to adjust PTO-values for the direction in which health changes are experienced. Since the measurement of value attached to health improvements (PTO-UP) is more common than that of avoided health decline (PTO-Down) it was decided to use PTO-UP values as the \"baseline\", subject to adjustment. The individual-specific mean WAD was therefore applied as a weight to PTO-UP values. These weighted PTO-UP values may then be attached to *avoided health changes,* while the \"raw\" PTO-UP values would still be attached to health improvements. For example, the value of avoiding a move from health state A to C was calculated as the value of PTO-UP for the move from C to A, weighted by the individual-specific mean WAD. These combined values are indexed \"PTO-WAD\". To calibrate the resulting adjusted PTO-values to the 0--1 range, the PTO-Up values were weighted using a power transformation.  Thus, a WAD higher than 1 increases the social value of an avoided health change compared to the \"raw\" PTO value for curative treatment effects. If for example, subject\'s responses to the PTO-Up procedure reveal that the distance between health states C and A gained upwards is 0.4, and this individual\'s mean WAD is 1.5, which implies a preference towards preventive care, avoiding a decline from A to C would be calculated as 0.54. Whether the accuracy of the PTO-Up values in predicting preferences for prioritization among curative and preventive interventions can be improved by weighting them with the WAD, will be analyzed in the next step. 6\) The sixth section presented the direct allocation task to the subjects. From all 30 possible health changes on the scale (in both directions), five were randomly chosen from those *not presented in the previous three PTO tasks.* These were three curative movements (1) *C* → *B*, (2) *E* → *C*, (3) *F* → *B*, and two treatments saving from decline (4) , (5) . Subjects were told that they had a finite budget for the next year as health authorities, but that they did not yet know how sufficient it was and which of the presented interventions they could offer to the public. They were then asked to rank the interventions according to their preferences in giving priority to patient groups. Results were interpreted such that the intervention ranked first has the highest social value for respondents at an *ordinal* level. While the absolute size of the intervals between health states represented by the five scenarios cannot be derived from this ranking, several statements on their relative size can be made. For example, one would expect (3) *F* → *B* to be ranked higher than (1) *C* → *B* and (2) *E* → *C*. Consequently, if irrelevance of the direction of health change on preferences is assumed, higher priority would be attached to curative treatment (3) *F* → *B* as compared to avoiding decline in (5)  or (4) , while the improvement (1) *C* → *B* would rank lower than avoiding the decline in (4) , and (2) *E* → *C* would have lower priority than (5) . To analyze the assumption of equidistances, the ranking obtained from each individual was investigated for the number of preference reversals between curative and preventive treatments. Using the rules on relative size given above, a reversal was defined as either an avoided decline that was ranked higher than an improvement of a greater change (reversal in favor of avoiding decline) or as an improvement that was attached higher priority than an avoided decline of larger size (reversal in favor of curative treatment). The ordering obtained by the direct ranking task was then compared with the orderings implied by individuals\' responses to the four valuation techniques (PTO-UP, PTO-Down, PTO-WAD, SG). On the presumption that the direct ranking task reflects individuals\' preferences to prioritize among the five scenarios best, the number of \'errors\' found in each of the implied rankings was calculated. An \'erroneous\' rank was defined as a mismatch between the rank a scenario obtained in the direct prioritization task as compared to the rank the same scenario obtained predicted by the particular valuation technique. The number of potential errors ranges from 2, in case the ranks of two scenarios are reversed in the prediction, to a maximum of 5 in case all ranks are predicted incorrectly. However, not only the number of errors but also their degree is relevant. For example, a reversal between two adjacent ranks is less serious than a reversal between the first and the fourth rank. Therefore, the degree of incorrect prediction measured as the absolute sum of deviation in ranks between direct prioritization ranking and predicted ranking, averaged over all individuals, was also analyzed. The potential absolute sum of deviations in ranks ranges from 2, in case the ranks of two scenarios are adjacently reversed, to 12, reflecting the maximum disagreement in orderings. 7\) Demographic characteristics such as age, gender and employment status were asked in the last section of the questionnaire. In addition, subjects were asked to rate how difficult they found it to answer the survey\'s questions. They were also encouraged to give qualitative comments. A JavaScript function was programmed that calculated the time respondents needed to complete the survey starting. The recording of time started with the first standard gamble task (with the first \'click\' performed by responders) and stopped after the provision of demographic information (before final qualitative comments). The interactive nature of the survey allowed to program search algorithms in all PTO tasks that were different from the standard PTO design. Usually, people are asked to give a number *themselves* at which they are indifferent between two treatment options. As reported by others, the pre-test of this study showed that subjects had difficulties deciding on and matching precise numbers representing equal societal value to them \[[@B19]\]. Therefore, it was decided to *offer* numbers to respondents, similar to binary choices as in bidding game techniques. The elicitation procedure started with a scenario where both interventions could be offered to 100 patients each. Subjects then had to decide for which of the two treatments they would opt or whether they were indifferent. Had participants chosen one of the treatments, the response indicated the direction of inequality, i.e. which intervention was associated with higher value. Subsequently, a higher number of patients benefiting from the less favored treatment was presented. After each vote by which the participant indicated an unaffected direction of inequality, the number offered next was exponentially increased. The algorithm then identified the first choice where responders changed their preference for one of the alternatives, e.g. ticked the second treatment after they had chosen the first previously. The numbers in the range between this threshold and the preceding number (covering the point of indifference) were presented in a ping-pong style, alternating convergent numbers between high and low depending on the voting behavior. The task proceeded until subjects were indifferent, when the next pair-wise comparison was started. Examples of series of hypothetical choices and responses are given in table [3](#T3){ref-type="table"}. The whole elicitation procedure was facilitated by various visual aids. For example, in the SG section, a balance was shown that changed its appearance depending on the risk the respondent was willing to take (Figure [2](#F2){ref-type="fig"}). Health states were explained by written text and small animations representing the information. In the subsequent sections, these animations were used to represent health states and to avoid redundant text and information overload. Written information on health states and the tasks was always accessible by hyperlinks that opened pop-up windows restating the information. {#F2} ::: {#T2 .table-wrap} ::: {.caption} ###### Examples of search algorithm used in the PTO-tasks ::: Choice Vote 2\. Choice Vote 3\. Choice Vote 4\. Choice Vote PTO calculation -------- --------- ------------ ------ ------------ ------ ------------ ------ ----------------- ----- ----- --------- --------------- A B A B A B A B 100 100 A 100 110 A 100 130 A 100 170 Indiff. 100 A = 170 B 100 100 B 100 90 B 100 70 A 100 80 A 100 A = 85 B 100 500^\*^ A 100 700 B 100 600 B 100 550 Indiff. 100 A = 550 B Note: Each row shows a series of hypothetical choices between two interventions A and B that could be offered to the specified number of patients each, and responses to the choices. ^\*^ Arrived at after having consequently voted for A in the preceding choices ::: Subjects -------- A self-selected sample was recruited over the Internet by indexing the survey\'s web page in search engines, announcing it on survey meta-listings and by postings in German language Usenet groups, such as de.sci.\*, de.soc.\* and de.talk.\*. Subjects were invited to visit the survey\'s webpage. As an incentive, it was announced that those taking the questionnaire are offered the chance to win a coupon from an online bookstore worth \~ 10€. By separate transmission of responses and email addresses anonymity was guaranteed. The risk that the same person could provide multiple responses was minimized by placement of mutually dependent \'cookies\'. Cookies are small files that are written to the users\' hard disk by the Web server when accessing a web page and can be read and re-written alongside the visit by this server. In the study, information was written, checked, and tracked in multiple cookies, referring to each other, when initially accessing the survey, while proceeding the survey, before and after successful transmission of responses, as well as before and after submission of the email-address. The information was not stored in the usual temporary cookie folder on the users\' hard disk. This programming inhibited repetitive restarts of the survey by the same user as well as moving freely through the survey\'s pages. Multiple submissions of responses were also suppressed. Still, it cannot be ruled out completely that a person experienced in web programming with some energy and time might have been able to manipulate cookies to provide multiple responses. Data was analyzed with the statistical package STATA v7 \[[@B20]\]. Results ======= After 15 days, 129 individuals had completed the survey, of which 65% were male. The average age was 35 years, ranging between 19 and 63 years. 87% had tertiary education, predominantly in medicine (32%) or economics (19%), nursing sciences (11%) or psychology (8%). 24% were currently unemployed or retired and 9% were students. The average time needed to complete the survey was 21 minutes (median: 35). Data of two respondents had to be removed, as their answers to the PTO-Up choices were obviously not serious. Both subjects indicated that they were indifferent between a smaller health gain offered to a smaller number of patients and a larger benefit offered to more patients. 9% of respondents stated that they found it not difficult, little difficult (39%) or quite difficult (45%) to answer the questions, and 7% said they found it very difficult. Analyses of web statistics and access-logs show that only few individuals (7) started the questionnaire and dropped out during progress. All subjects provided the same, \"correct\" and consistent ranking of severity of the health states requested in the first task of the survey. As shown in figure [3](#F3){ref-type="fig"}, the differences in mean utility/social value of health states between the three elicitation techniques are considerable. Values obtained with the \"prevention\"-PTO method (PTO-Down) are generally higher than those assessed by the standard \"curative treatment\" procedure (PTO-Up). The difference between PTO techniques and SG is substantial and increases with severity of the health state. It was investigated whether the fixed format of the successive bids may have provided incentives to shorten the length of the survey by voting strategically and avoiding the ping-pong procedure. If this was true, one would expect that the fraction of responders taking the easiest way to do so, namely \"opting out\" themselves by choosing the \"indifference\"-option, would increase with successive tasks. However, this was not the case. The number of responders ending a PTO task by clicking the \"indifference\"-option ranges between 35 (28%) and 46 (36%) per choice, and these figures are not correlated with the sequential placement of the respective PTO-choice among all 13 PTO-choices (*Spearman r = -0.09, p = 0.8*). Also, the number of successive bids per PTO-figure does not decrease with the sequential placement of the respective PTO-choice among the 4 resp. 5 PTO-choices of each PTO-task. {#F3} The vast majority of responders (69%) preferred the allocation of resources to already diseased patients over avoidance of the same health state in *all* five comparisons. The opposite was true for 2 subjects (2%). 8 responders (6%) gave equal priority to improvements in health and avoiding decline in all comparisons. The remaining subjects (23%) favored cure or avoiding decline depending on the respective health state. There was no significant association between sex, employment-status, education or professional background and preferences towards cure or avoiding decline. However, respondents that voted in favor of cure or avoiding decline dependent on the compared health effects were more likely to find the survey difficult (χ^2^ = 21.79, *p* = 0.01). The mean WAD is nearly constant for all except health state B, for which it is moderately higher than the WAD\'s for the other health states (*p = 0.000 compared to WAD for health states C,D,E; p = 0.05 compared WAD for health state F*). The size of the WAD values implies that to be of equal value to responders, a health decline has to be avoided in twice as many persons than persons have to experience a health gain of the same magnitude. The mean intervals between health states calculated based on the values elicited by the different techniques and the substitution rate between avoided downward and gained upward movements (WAD) are presented in table [3](#T3){ref-type="table"}. All differences between SG and the two PTO techniques are statistically significant at the 5% level or better. Notably, differences in size of intervals between health states inferred from the two PTO-tasks increase with severity of health states. ::: {#T3 .table-wrap} ::: {.caption} ###### Mean intervals and WAD for movements between health states ::: Mean interval between health states --------------- ------------------------------------- --------------------------- --------------------------- --------------------------- Health states SG^§^ \[95% CI\] PTO-Down ° \[95% CI\] PTO-Up^\$^ \[95% CI\] WAD^\*^ \[95% CI\] A--B 0.0875 \[0.0637--0.1113\] 0.0480 \[0.0223--0.0736\] 0.0502 \[0.0304--0.0700\] 0.6443 \[0.5811--0.7075\] A--C 0.2056 \[0.1836--0.2276\] 0.0778 \[0.0488--0.1069\] 0.1032 \[0.0716--0.1348\] 0.5248 \[0.4616--0.5873\] A--D 0.4015 \[0.3738--0.4292\] 0.1484 \[0.1111--0.1857\] 0.1857 \[0.1389--0.2325\] 0.4748 \[0.4066--0.5431\] A--E 0.5658 \[0.5354--0.5961\] 0.2977 \[0.2487--0.3467\] 0.3694 \[0.3112--0.4277\] 0.4919 \[0.3951--0.5888\] A--F 1 \-- 1 \-- 1 \-- 0.5167 \[0.3542--0.6791\] *N* = 127 Results of paired t-tests: ^§^Significant different to PTO-Up at 1% or better; ° Significant different to SG at 1% or better; ^\$^Significant different to PTO-Down at 1% or better for A -- E, at 10% for A -- D and A -- C; ^\*^WAD values were calculated as described in eq. 4.1 (section 5) of the study design ::: The number of reversals observed in the direct ranking task is substantial. Calculated over all subjects, 151 reversals occurred in favor of curative treatments, i.e., an improvement was ranked higher than an avoided decline of greater size. In 71 rankings, the treatment that would improve health from C to B was ranked higher than avoiding decline from health state B to D and 80 individuals attached higher priority to the change from E to C compared to preventing the decline from C to F. Both reversals were prevalent in 68 rankings. 9 reversals occurred in favor of prevention. The improvement *F* → *B* was attached less priority than avoided decline from  and  in 5 and 4 cases respectively. Occurrence of both reversals was observed in 3 rankings. Only 3 reversals occurred among curative treatments. In all three cases, the improvement *E* → *C* ranked better than *F* → *B*. A comparison between the direct ranking task and the rankings predicted by the four valuation techniques reveals considerable deviation in the predicted rankings on the individual level. Table [4](#T4){ref-type="table"} shows the mean number of \"errors\" found in each predicted ranking compared to the direct ranking task. While the PTO-WAD ranking produces on average one reversal among the ranks of two scenarios, the unadjusted rankings implied by participants\' answers to the common PTO-UP procedure show on average one additional scenario ranked incorrectly. The mean absolute deviation in ranks between the direct prioritization ranking and the predicted rankings illustrates that the PTO-WAD not only produces the smallest number of errors, but also that these errors are less serious (table [4](#T4){ref-type="table"}). The average absolute deviation of two ranks suggests that the PTO-WAD produces one adjacent reversal of two scenarios compared to the direct ranking. The superiority of the PTO-WAD technique compared to the other valuation techniques on the aggregate level can be mainly attributed to the fact that it reproduces a considerable fraction of the reversals observed in the direct ranking task. 49% of the reversals between curative and preventive treatments reported about above were exactly replicated by the PTO-WAD technique. ::: {#T4 .table-wrap} ::: {.caption} ###### Mean and median number of errors and absolute rank deviation in the predicted rankings by elicitation technique ::: SG PTO-Down PTO-Up PTO-WAD ---------------------------------- ----- ---------- -------- --------- Mean number of errors 3.0 2.7 2.7 1.9 Median 4 4 4 2 Mean absolute deviation in ranks 3.9 3.0 3.0 2.1 Median 4 4 4 2 N = 127 ::: A considerable fraction of participants provided detailed and serious qualitative comments on the content level. These comments were discussing mainly (1) reasons underlying their choices in favor of curative or preventive healthcare, (2) issues of translating their choices into practice, which was mainly expressed by participants with a professional healthcare background, and (3) general issues of priority setting, health care policy and the German health care system. Surprisingly, several healthcare professionals stated, that they appreciate the presentation of the resource allocation problems and the opportunity to transmit their views anonymously as \"\...*these are the issues we have to face and we should aim to honestly and openly discuss them*\". Discussion ========== This study used an interactive web-based questionnaire to examine preferences towards improvements in health and avoidance of decline in a healthcare resource allocation context using a self-selected sample. All four preference elicitation techniques gave results of high internal consistency within and between subjects. This is especially surprising for the direct ranking and the PTO-tasks, since they were cognitively demanding. On the methodological level, this may be explained by the mode PTO choices were presented. Responders were confronted with numbers as in bidding games, which might be easier to handle than to match them \"freehand\". Another explanation is that the health scale used did not include the conventional \"death\"-state and therefore all interventions were equal with respect to the fact that ethical principles and preferences regarding life-saving treatments, such as the rule of rescue, were not involved. As the values obtained by the new PTO-Down method suggest, the same allocation preferences seem to underlie rationing choices regarding preventive interventions and curative treatments each evaluated among each other. Health effects of greater magnitude, gained upwards or avoided downwards, contributed to higher value. However, it is unclear for both types of health directions whether it is the size of health effect or the severity of initial, or avoided health state that was the main rationing criteria for responders. Unfortunately, the construction of the PTO choices does not allow decomposition of these factors. Contrary to the study by Ubel et al., which reports moderate preferences for preventive over curative treatments \[[@B12]\], responders to the present study strongly favored improving hypothetical patients\' health compared to avoiding decline. The prioritization of curative treatments was present both in the person trade-off procedure as well as in the direct ranking task (represented by the high number of reversals in favor of cure). This is surprising since the onset of disease was defined as a certain event in the near future. It is likely that the value people place on preventive as compared to curative treatments diminishes as the extent of uncertainty and time to disease onset increase. One explanation for the apparent discrepancy in results may be the different framing of questions. Ubel et al. presented the task as choices among different \"new projects\" of which a fixed number of nursing home residents could benefit from, i.e., on the program, rather than on the individual patient level. Contrary, in this study, participants were confronted with choices among treatments that could be offered to a varying number of patients. In addition, the SG tasks in which participants were asked to imagine themselves suffering from the health states preceded the rationing choices. The *presence* of suffering or need in terms of actual manifestation of the disease could have caused subjects to focus patients\' situation and to place more value on curative treatment, even if the distance between health states is considerably smaller. This interpretation is strengthened by the qualitative comments provided by participants of the present study. A reasonable fraction argued \"*\...it is cruel to deny treatment to those suffering with reference to the fact that resources are reserved to save others from entering this situation. What are patients left to expect from health care then?*\". Other qualitative comments suggest, that preferences for improving patients\' health rather than avoiding decline might in part be caused by fairness concerns. Several responders stated considerations such as: \"*It is some form of luxury to prevent a disease as long as there are people already suffering the same disease, and as long as it can be treated*\". Spending resources on already suffering persons first and accepting others getting seriously ill, which might then be prioritized against others again, equalizes each individuals\' chances to obtain appropriate treatment. Preferring avoidance of decline over improving patients\' health in resource allocation, on the other hand, would increase inequality between those remaining ill and those fortunate to profit from treatment before suffering. However, participants\' preferences towards curative treatments were not absolute. As both, results from the combined PTO-WAD technique as well as the direct prioritization ranking, show, a vast majority was willing to trade their preference towards cure against a larger number of patients that could be saved from decline. The basic idea of quantifying this relation between equally sized effects of different directions and applying this value as a weight for avoided health decline in comparisons between curative and preventive treatments proved feasible and has led to a substantial improvement in predicting individuals\' rationing decisions concerning curative and preventive treatments. However, using the individual-specific mean as a weight may be a too rough estimate since a reasonable fraction of responders made their choices for or against curative treatments dependent of the health states involved in the PTO-WAD comparisons. Other, more complex weighting functions may produce even better results. Optimally, one could use (traditional) PTO-Up values to evaluate the benefit of curative treatments, PTO-Down values for the assessment of interventions avoiding decline, and the weighting technique for choices involving both types of interventions. The inferior performance of the SG in predicting individuals\' rationing choices as compared to the PTO procedures, at least at the ordinal level, is not surprising since Pinto and others report similar results \[[@B18],[@B21],[@B22]\]. One could argue that it is not reasonable to expect that SG utilities would accurately predict societal rationing choices among preventive and curative treatments, since this is not what they are intended to measure. While this is of course true, it is exactly what is done in health economic evaluation and it is the purpose of the study to question this practice by presenting the disagreement between SG, PTO and the direct prioritization ranking. For example, recently Zaric and Brandeau developed a resource allocation model for HIV prevention programs \[[@B23]\]. The calculation of potential gain in QALYs of preventing HIV infection was based on utilities attached to HIV associated health states by HIV-infected patients. In addition to the problem of personal versus social valuation of health states, this approach assumes that society would attach the same value to avoiding and leaving HIV-associated health states. The study has several limitations and weaknesses. A main limitation is the small sample, which is highly biased towards the well-educated and those using the Internet. In addition to the traditional self-selection bias common to online-research, the current study is also characterized by a strong overrepresentation of participants with a medical background. This bias is likely to result in part from the chosen newsgroups, in which the survey was announced and might be reduced by a more wide-spread invitation strategy. However, the main aim of the study was to introduce a technique for the joint valuation of health effects experienced in different directions and to examine its internal consistency with a direct prioritization task, rather than to yield representative figures. The unconventional scaling of health states, the presentation of the PTO questions as bidding games and the heavy demand on cognitive capabilities during the ranking task highlight the experimental character of the study. In addition, the PTO tasks did not provide information on the preceding duration of illness of hypothetical patients. Some of the health states used might have been recognized as congenital disabilities, which is likely to introduce particular bias. Also, preferences towards curative treatments might in part result from uncertainty or discounting future outcomes. Though both curative and protective actions were described as having certain and immediate effects, subjects might have intuitively perceived the latter as less certain or subject to a time shift, or even questioned the credibility of this description, since preventive programs with \"certain effects in the near future\" are scarce. Contrary, curative treatments might have been perceived as more harmful or associated with side effects, which would have introduced a bias towards prioritization of avoiding deterioration. A serious concern is the additive definition of three health states. Whereas they may fit an additive scheme of cure of each condition, an additive scheme of prevention may have been seriously doubted by participants. Realism would have introduced a strong bias in favor of cure. One would then expect individuals with a healthcare professional background being most sensitive to this issue. However, no \"knowledge effect\", in terms of association between professional background and preferences towards cure or avoidance of decline was observed. Finally, the observed strong preferences towards cure might also be explained by the fact, that -- traditionally -- the healthcare budget is perceived to produce above all treatment and cure for diseased patients. Again, this potential bias might be even more serious taking the high fraction of medical professionals into account, that can be expected to be for the most part practicing and experiencing curative care. On the other hand however, targeting and moving resources to preventive care is often claimed by public opinion and the medical profession. Preventive programs often rank very high in priority setting lists \[[@B24]\]. As Ubel et al. point out, such attitudes might be based on the belief, that it only needs \"an ounce of prevention to bring large benefits\" \[[@B12]\]. As in the present study effects of avoiding decline and opportunity costs in terms of foregone curative effects were quantified and directly comparable, however, pro-preventive attitudes might have been altered. There are also a number of methodological concerns. The fact that all PTO-choices started with the compared treatments offered to 100 patients each may have introduced bias from anchoring effects (starting-point bias) \[[@B25],[@B26]\]. Though anchoring responses at 100/100 might have led to a systematic underestimation of participants\' \"true\" PTO-values, this should not have affected their relative size or their relation to the direct ranking task. More generally, the PTO may be susceptible to order effects which may in part explain the differences between the PTO-Up and PTO-Down results \[[@B19]\]. A major concern is that the PTO-WAD technique, which was used to directly evaluate the value of health improvements compared to avoided decline of the same magnitude might have introduced and mixed action, status-quo and omission biases \[[@B27]\]. To explain results, action bias must be strong enough to overcome status-quo and loss omission bias, both of which suggest that more participants opt for avoiding decline, as this keeps the current situation unchanged for all patients and the decision maker cannot be held responsible for health deterioration. Patt and Zeckhauser observed such strong action bias in some individuals in experimental studies involving environmental decisions though \[[@B28]\]. Finally, one could be concerned with the level of engagement and reflection of responders in the presented numerical exercises. Given that participants are not forced to deliberation, they might present quick and unreflective answers, not necessarily consistent with their deeper values. The high number of thoughtful and serious qualitative comments, the observed internal consistency in the PTO-tasks, and between PTO-WAD and direct prioritization ranking indicate that subjects were not responding haphazardly and made considered choices. Also the small number of reversals among curative treatments observed in the ranking task compared to those found among curative and preventive treatments indicate that these may reflect \"true values\" rather than resulting from ill judgment. Conversely, there is no objective and direct evidence on the involvement of participants and the rationales underlying their choices. Despite these limitations, the results of this study, however, suggest that the utility of being cured of a given health state might not be a good approximation for the societal value of avoiding this health state, especially in cases of competition between preventive and curative interventions. Conclusions =========== Participants strongly prioritized improving patients\' health rather than avoiding (further) decline. Weighting PTO values according to the direction in which changes in health are experienced improved their accuracy in predicting a direct prioritization ranking. This study also adds to the existing evidence that health state utilities obtained by the standard gamble method do not reflect social values in resource allocation contexts. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ I am particularly indebted to Thomas J. Strasmann for his support in graphical design and web programming. Special thanks are due to Erik Nord for his valuable comments on an earlier draft. I also thank Christian Koeck and Nicoletta von Laue for facilitating the survey. Any errors or omissions are the responsibility of the author alone.

Introduction ============ Following severe genotoxic damage, cells undergo either rapid or delayed death termed by Okada \[[@B1]\] as interphase and reproductive deaths, respectively. Sensitive cells undergo death within a few hours of genotoxic insult (if the damage is not so enormous as to cause immediate necrosis), whilst damage resistant cells execute death after a delay, precipitating either from the initial G2 arrest or after a number of cell divisions. In lymphoid and myeloid cell lines, all three types of cell death require new protein synthesis and exhibit the DNA ladder characteristic of apoptosis \[[@B2]\]. This delayed reproductive death, also known as \"mitotic death\", is characteristic of many p53 mutated tumours that are resistant to genotoxic insult \[[@B3],[@B4],[@B5]\]. The main features of mitotic death have been described by several workers \[[@B6],[@B7],[@B8]\]. They include (a) the absence/delay of the G1/S checkpoint and hence (b) the absence of interphase apoptosis coupled to this checkpoint; (c) delay in the G2 compartment followed by its adaptation and hence (d) a sequence of aberrant mitoses which end in mitotic death; (e) formation and disintegration of (multinuclear) giant cells (which is often also termed mitotic death or catastrophe); (f) delayed apoptosis; and (g) finally survival. Given this elaborate sequence of events associated with resistance to genotoxic insult and poor oncological prognosis \[[@B9]\], it is logical to suggest that these phenomena may contribute to cell survival. To assess this possibility, we must first explore the various facets of mitotic death. Uncoupling of apoptosis ======================= Terminal cell death ends any potential for repair or propagation of a cell population and so the earlier death occurs in the cell cycle following damage, the smaller the potential for possible cell survival. From studies of p53, the concept that there is a temporal element to the adaptive response of the cell to damage has emerged, and it has been suggested that there is a \"race between repair and cell death\" \[[@B10]\]. Therefore, it is clear that to ensure genomic integrity, apoptosis should be tightly coupled to cell cycle checkpoints and initiated rapidly \[[@B11],[@B12]\]. For this reason, apoptosis is mostly precipitated from the G1/S checkpoint in normal, healthy, p53 wild-type cells before segregation of genomes occurs, preventing any chance of potentiating or fixing the damage later in the cell cycle. As such, maintaining genome fidelity puts a tight limit on the proliferative and survival capacity of cells. Conversely, uncoupling cell death from cell cycle checkpoints and delaying the terminal effector stage of apoptosis should give cells a greater chance for repair, diversification of pathways and survival. This latter situation is observed in resistant tumours, where G1/S and S-phase checkpoints fail \[[@B13]\] and where mitotic death is initiated only at the last stage of the cell cycle, as the term suggests, around the final stages of mitosis. The tight negative correlation between delayed apoptosis and release from G2 arrest in p53 mutant tumours after irradiation (Cragg *et al.,* submitted), and other data \[[@B7],[@B14],[@B15]\] suggest that the main decision for death versus survival originates from the G2/M checkpoint. Thus, reaching the G2 compartment would appear crucial for cell survival and removing preceding checkpoints should increase resistance. Importantly however, these resistant cells have missed the chance of repair in G1 and S. Repair of DNA strand-breaks =========================== To repair damaged DNA, the damage must first be sensed. The precise mechanisms for this are not yet clear, although some current data suggests that p53 itself \[[@B16]\] and repair proteins of the Rad 9 and Ku group, also function as damage sensors \[[@B4],[@B17]\]. Recent evidence also indicates that in S-phase, sensing and repair of strand-breaks competes with continued DNA replication. Therefore, it may be that in resistant tumour cells, damage is not repaired because it not sensed \[[@B18]\]. In support of this notion, comparing S-phase delay for a panel of lymphoid cell lines after a single dose of irradiation, the most radioresistant cells delay least in S-phase and exhibit the most extensive chromosomal aberrations (Cragg *et al;* submitted). This situation is exacerbated by the fact that conversion of single-strand breaks (SSB) into pernicious double-strand breaks (DSB) and the error-prone trans-lesion DNA synthesis, are probably occurring in this environment \[[@B4],[@B19]\]. It is paradoxical then, that unlike p53 mutated cells, p53 wild-type cells which sense DNA damage and delay for repair in G1 and S, do not recover from large genotoxic insults \[[@B20]\]. Potentially lethal DSB are repaired by non-homologous end-joining (NHEJ) in G1/early S and by homologous recombination (HR), in late S/G2 \[[@B18],[@B21]\]. NHEJ is by definition error-prone and so error-free repair is only possible through HR. Reaching the G2 compartment provides the best chance for error-free DNA repair within the mitotic cycle. Delay in G2 -- a rescue compartment? ==================================== Cells with DSB arrest in G2 at the G2/M DNA damage checkpoint \[[@B22]\] and attempt HR. In fact, it was only recently discovered that HR acts in human cells \[[@B18],[@B23]\]. This HR repair capacity seems to be many-fold more active in p53-mutated tumours than in wild-type counterparts \[[@B24]\] and therefore, G2 arrest should be considered an important rescue compartment for damage-resistant tumours. In support of this, the ability of various tumour cell lines to accumulate after damage in G2 seems proportional to their clonogenicity (Cragg *et al.,* submitted) and resistance to multiple anticancer drugs \[[@B14]\], whilst abrogation of G2-arrest potentiates cell death \[[@B15]\]. This type of response, providing maximal accumulation of cells in G2, is due to loss of functional p53 \[[@B3],[@B13],[@B25]\] and perhaps explains why mutations in p53 are selected for during tumour progression. Although possibly elevated in p53-mutated tumours, it is clear that the repair capacity of the G2 compartment cannot be limitless. If the damage is too extensive and the cells have overcome (adapted) the G2/M checkpoint, the mitoses that ensue are aberrant due to the genome instability and chromosome breaks. In addition, the ability of chromosomes to stick at breakage sites is often displayed in sequential post-damage mitoses as \"bridge-break-fusion\" events and it is mostly this process which accounts for a large proportion of first aberrant and then lethal mitoses. All of these facets are due to secondary DNA damage events and represent the \"cost\" of reaching the G2 compartment. Several investigators have now shown that secondary chromosome lesions and mis-repair directly cause chromosome breaks and evoke \"mitotic catastrophe\" \[[@B2],[@B26],[@B27],[@B28],[@B29]\]. Mitotic catastrophe =================== There are very diverse descriptions of mitotic catastrophe. Initially, mitotic catastrophe was associated with incomplete DNA synthesis and premature chromosome condensation \[[@B30],[@B31]\], with features in common with apoptosis. Other authors define it as an aberrant form of mitosis associated with the formation of multinucleate giant cells that are temporarily viable but reproductively dead \[[@B11],[@B32]\]. Still others state that mitotic catastrophe is pre-determined in G2 and characterised by an abortive short cut into metaphase arrest \[[@B33],[@B34]\]. Mitotic failure often manifests with micronucleation \[[@B35],[@B36]\] and nuclear segmentation \[[@B37]\] and this in turn, is associated with mitosis restitution into interphase polyploid cells \[[@B38]\]. An apoptosis-like, TUNEL-positive death from metaphase-arrest can often be observed in tumours after various apoptogens \[[@B37]\]. It frequently accompanies mitosis restitution \[[@B39]\], and is commonly seen during the development of micronuclei and nuclear segments, which are themselves initially TUNEL-negative \[[@B40],[@B41]\]. Clearly, the features of mitotic catastrophe are eclectic \[[@B39]\]. In fact, although these events are usually viewed as deviations from the normal cell cycle, they involve phenomena that are not necessarily associated with mitosis, such as micronucleation, restitution, and polyploidy. Indeed, the last of these is clearly associated with the endocycle, and we would suggest that initiating the endocycle provides an alternative to mitotic catastrophe in resistant cells. Endopolyploid giant cells ========================= Polygenomic giant cells can arise from G2 arrested cells \[[@B7],[@B42],[@B43]\], by cell fusion or by mitotic restitution \[[@B38],[@B44],[@B45]\]. In all cases, tetraploidy appears to be a pre-requisite for initiation of endocycles. True endopolyploids contain several chromosome sets and are usually formed in a series of restitution cycles by (polyploidising) mitoses, which are aborted after segregation of sister chromatids. These cells are thought to be able to return to the diploid state either through mitosis omitting S-phases or by multipolar mitosis \[[@B46]\]. Interestingly, studies on human megakaryocytes in culture \[[@B47]\] showed that 8N-16N giant nuclei express cyclin B1 and that this co-precipitates with H1 histone kinase activity. Thus, these nuclei retain competency for mitosis and subsequently, multipolar mitoses with centrosomes co-localized with cyclin B follow. For the S-phase 16N-32N cells and for 32N and 64N cells, these activities are strongly reduced. These data are in line with numerous reports that 8N-ploidy represents a general limit, above which cells cannot execute mitosis. 16N-32N cells and cells of higher ploidy are usually unable to divide by mitosis, and therefore belong to the so-called hyperploid class of polyploids \[[@B44]\]. Hyperploidy, which can reach very high DNA content values, is widely observed during the development of transient organs in animals and plants \[[@B44],[@B46],[@B48]\]. Unlike in true endopolyploids, in hyperploid cells, sister chromatids usually do not become separated after replication rounds \[[@B44],[@B46],[@B48]\] and especially if slightly condensed, retain partial cohesion at highly reiterated sites. Most likely, such a polytenic kind of endopolyploidy arises initially from cells arrested in the G2/M or spindle checkpoint and renders these nuclei unable to segregate chromosomes by mitosis. Although mostly reproductively dead, these cells however, can also return to the diploid state, through a specific kind of cell division termed de-polyploidisation, which involves de-polytenization, somatic pairing of homologs and subsequent reduction divisions \[[@B44]\]. Such a process, termed also \"meiosis without karyogamy\", has been described by Grell in *Culex* \[[@B49]\] and in the asexual life-cycle of Radiolaria, reviewed in \[[@B50]\]. Although somatic reduction is likely to occur in higher plants, the possibility of a similar process occurring in mammalian cells remained until recently only speculation \[[@B42],[@B44]\]. In malignant tumours, there is often a population of cells with a DNA content above the diploid value of the so-called stem-line and in many cases a number of high-ploidy cells are evident \[[@B51]\]. The proliferative potential of this latter population remains controversial. Some authors claim that high ploidy cells have reduced proliferative potential \[[@B52]\] or even none at all \[[@B53],[@B54]\]. Others have found that although these cells are usually delayed in, and often deleted from, metaphase, they can proliferate either normally \[[@B55]\] or through multipolar mitoses \[[@B56]\], although these observations may refer only to oligoploid tumour cells. We have observed giant cells of 8N-128N and more, as a transient response of p53 mutated human lymphoma cell lines to genotoxic damage or spindle poisons. Within this response, some endopolyploid cells appear to undergo somatic reduction and/or multipolar mitoses and produce mitotic descendants \[\[[@B41],[@B57]\], also unpublished\]. Therefore, it is important to consider how endopolyploids originate from, and return to, the mitotic cycle, with respect to the different molecular environments required. Molecular machinery of the endocycle ==================================== The mitotic cycle is dependent upon the action of the mitosis-promoting factor (MPF), which is a complex of cyclin B and cdc2 kinase (p34 cdk1). The MPF phosphorylates various substrates including MAPs, lamins, and histone HI, which subsequently induce the spindle assembly, breakdown of nuclear envelope, and chromosome condensation, necessary for mitosis. In contrast, lack of MPF activity and inability to execute mitosis is considered a hallmark of the endocycle \[[@B8],[@B42],[@B58]\]. A recent study of reprogramming the cell cycle for endoreduplication in rodent trophoblast cells has shown that the arrest of the mitotic cycle at the onset of endoreduplication was associated with a failure to assemble the MPF \[[@B59]\], at the G2/M checkpoint. The same can be achieved by inhibiting cdc2 activity with the kinase inhibitor, staurosporine \[[@B45]\]. The main component of MPF inactivity appears to be lack of nuclear cyclin B1 localisation in the endocycle \[[@B8]\]. For endoreduplication, DNA synthesis must occur independently of mitosis and for this to happen several checkpoints must be overcome. These checkpoints are located in late S-phase (S/M \[[@B34]\], telophase \[[@B60]\], centrosome duplication \[[@B61]\] checkpoints), in G2-phase (G2/M checkpoint \[[@B22]\], and in M-phase (spindle and post-spindle checkpoints \[[@B62],[@B63]\]). In the endoreduplication cycle, cells shuttle only between G1 and S-phase, with a truncated late S-phase \[[@B8],[@B42],[@B44],[@B64]\] and therefore many of these checkpoints are naturally absent, allowing unchecked DNA synthesis. The MPF also negatively regulates the assembly of the replication complex \[[@B8]\]. Stable overexpression of cyclin E, which is part of the replication complex and the main driver of replication, is a characteristic feature of endocycling and tumour cells \[[@B8],[@B58],[@B60],[@B64],[@B65]\]. Therefore, it seems likely that when DNA synthesis becomes unscheduled, due to the inactivity of the anaphase promoting complex APC and the post-spindle replication checkpoint, this stable and abundant cyclin E down-regulates cdc2 kinase, and thus renders the MPF inactive, enabling decondensation of chromosomes and reset of interphase. Most of the checkpoints detailed above are p53 dependent, and so not surprisingly, the ability to initiate endocycles and form endopolyploid cells is characteristic of cells lacking wild-type p53 function \[[@B15],[@B43],[@B66]\]. Influence of DNA damage ======================= Genotoxic damage in particular enhances this transition from the mitotic cycle into the endocycle in p53 mutant cells \[[@B15],[@B66]\]. The arrest in the G2/M checkpoint is known to activate chk1, which induces the phosphorylation of cdc25. This results in cdc2 remaining phosphorylated on Tyr 15, causing its inactivation and through the subsequent molecular pathway, retaining the cytoplasmic localisation (nuclear export) of cyclin B1 \[[@B22]\]. Consequently, DNA damage, causing G2 arrest and preventing MPF assembly, may provide the necessary molecular environment to trigger entry into the endocycle instead of mitosis. Another molecular constraint to overcome, to allow endoreduplication, concerns the APC which couples mitosis to a new round of DNA replication. This function is performed by the targeted proteolysis of several substrates necessary for the completion of mitosis \[[@B62]\]. Arrest at the spindle checkpoint caused by post-damage secondary DNA breaks or by spindle damage, down-regulates APC activity, thus preventing exit from mitosis but allowing entry into the endocycle by the pathways detailed above \[[@B8]\] and in \[[@B67]\]. In summary, the characteristic features of p53 mutated cancer cells in conjunction with DNA damage, appears to facilitate the molecular environment necessary to provide a shift from mitosis into the endocycle. The main switch-points for these transitions are outlined on Figure [1](#F1){ref-type="fig"}, although it should be mentioned that apoptosis represents another, third option. {#F1} Our own studies (Cragg *et al.,* submitted) have shown that in p53-mutated cells, the radiation dose threshold for entering the endocycle from damage induced G2 arrest is higher than for entering mitosis, and in particular, in more resistant cells i.e. cells can still enter the endocycle after high doses of irradiation. One possible explanation for these data is that the endocycle provides a survival advantage. Can giant cells repair DNA? =========================== If giant cells do provide a survival advantage after genotoxic insult, then presumably they must be able to repair DNA. In support of this suggestion, chromatid exchanges have been found between sister chromatids in the endopolyploid nuclei of rodent fibroblasts \[[@B68],[@B69]\]. Importantly, crossing over at each site appears only to occur between two of the many aligned chromonemes on the multichromonemic chromosomes \[[@B70]\] potentially allowing the production of a fully repaired chromosome. Furthermore, some data suggest that homologous chromosomes, which are randomly distributed in the mitotic cycle of diploid cells \[[@B71]\], become relocated and paired in polyploids \[[@B70],[@B72]\]. This early data has acquired some recent support \[[@B73]\] as it has been shown that homologous chromosomes come together in generative polyploids long before meiosis begins. In endopolyploid tumour cells, for recombination and somatic reduction to happen, reorganisation of endopolyploid nuclei should occur to provide pairing of homologous chromosomes. In support of this suggestion, a complex somatic reduction process involving meiotic-like bouquets has been reported recently by our group for some Burkitt\'s lymphoma cell lines after high doses of irradiation \[[@B41],[@B59]\]. Although initially this process appears extremely exotic, it has serious molecular and evolutionary grounds. An attractive idea linking meiotic recombination with DNA repair in somatic cells was recently proposed by Kleckner \[[@B74]\] who suggested that crossing over and meiosis both originated during evolution from the necessity to repair DNA double strand breaks in the mitotic cycle. Furthermore, the molecular relationship between meiotic recombination and the mitotic DNA damage checkpoints has now been established \[[@B75]\]. In addition, it is worthwhile to note that the endocycle is probably more ancient than both meiosis and mitosis in evolution \[[@B76]\] and that somatic reduction introduces elements of meiosis into endopolyploid cells. This latter may be important, as recombination DNA repair by exchange between homologs in meiosis has been reported to be up to 10 times more effective than between sister chromatids \[[@B77]\]. Somatic reduction of endopolyploid cells, including pairing of homologs similar to that in meiosis, may therefore represent an effective means of both DNA repair and return to diploidy and the mitotic cycle. In fact, comparing the efficacy of HR in different cell cycle compartments reported in the literature, it can be inferred that HR efficiency increases from G1 to G2 \[[@B78]\] and then still further from G2 of the mitotic cycle to the endocycle. Possibly therefore, in the race between DNA repair and apoptosis, repair is elevated and apoptosis delayed further and further, allowing this extreme shift into the endocycle to occur in the most resistant tumours. Conclusion ========== In conclusion, after extensive DNA damage, p53 mutated cells miss DNA repair and apoptosis at the G1 and S phase checkpoints and accumulate at the G2 arrest compartment with point mutations and chromosome breaks. The latter exhibit at the spindle checkpoint with mitotic catastrophe. In the absence of p53, the various phenomena of mitotic death, including mitosis restitution, are induced in an attempt to execute an evolutionarily conserved survival program via transient endocycles. The molecular environment of G2 arrest and restituting mitosis down-regulates MPF activity allowing initiation of the endocycle. The resulting polyploid cells may harbour greater DNA repair capacity coupled with the potential to return to the mitotic cycle and may therefore provide a survival advantage.

Introduction ============ DNA methylation is involved in cellular development, differentiation and transformation \[[@B1]\]. In different types of tumours, aberrant methylation of CpG islands in the promoter region has been observed for many differentiation- and cancer-related genes resulting in the silencing of their expression \[[@B2]\]. Therefore, over the past decade, there has been increasing interest in the use of demethylating agents to induce the differentiation or the apoptosis of cancer cells \[[@B3],[@B4]\]. Treatment of the cells with the pyrimidine analogue 5-AzaCytidine (AzaC), which inhibits methylation of cytosine residues during replication in the newly synthesised DNA, has been demonstrated to reactivate the expression of many silenced genes, as well as the expression of the silenced retro viral genomes \[[@B5]\]. Silencing of one of the most important cell cycle regulatory proteins p16^INK4a^ by methylation of the CpG islands in the promoter region has been found to be a common event in tumours \[[@B6],[@B7]\]. Protein p16 suppresses S-phase entry by antagonising the cyclin-dependent kinases CDK4 and CDK6 \[[@B8]\]. Deciphering the molecular mechanisms underlying the phenotypic effects of the treatment with demethylating drugs is a crucial step in understanding what genes may be interesting targets for chemotherapy. The available data on the mechanism of action of these drugs strengthen the idea that it is different from that of agents that act primarily via their cytotoxic effects, such as Arc-C \[[@B9]\]. Several lines of evidence suggest that galectin-1 (Gal-1), a 14 kDa galactoside-binding protein distributed widely in immune cells, could be involved in these mechanisms. Several members of the galectin family have been found to modulate cell differentiation and cell survival \[[@B10]-[@B15]\]. Early studies demonstrated that the expression of Gal1 could be induced in cultured hepatoma-derived cells by treatment with AzaC \[[@B16]\]. Chiariotti and co-workers showed that reactivation of the silent Gal1 alleles is accompanied by a transition from a fully methylated to a fully unmethylated state of several CpG dinucleotides in the promoter region \[[@B17]\]. In addition, nonexpressing tissues exhibited highly heterogeneous methylation profiles \[[@B18]\]. Gal is considered to be a typical cytosolic protein, lacking a signal peptide for membrane translocation \[[@B19]\]. However, most of the functions assigned to galectins are confined to the cell surface or extracellular milieu \[[@B10],[@B20],[@B21]\], consistent with evidences for extracellular roles of Gal1 in regulation of cellular differentiation and proliferation. It is clear that Gal1 can be specifically secreted and targeted by an infrequent mechanism \[[@B22]-[@B24]\]. The constitutive expression as well as the secretion of Gal1 dramatically depend on cell types \[[@B25]\] and are responsive to developmental events \[[@B20],[@B22],[@B23]\]. An example is found during erythroid differentiation of the K562 human leukaemia cell line. During differentiation induced by erythropoietin and deprivation of granulocyte-macrophage colony-stimulating factor, the cells empty their cytoplasmic content of endogenous Gal1 into the external medium where it is bind to cell surface receptors \[[@B24]\]. The synthesis and secretion of Gal1 by leukocytes are of interest because lactosaminoglycans present at the leukocyte cell surface may be physiologically significant galectin receptors that could mediate autocrine or paracrine functions. Several lines of evidence indicate that Gal1 may function as an autocrine negative growth regulator or as a pro-apoptotic factor \[[@B26]-[@B28]\]. We have recently demonstrated that Gal1 binding to Burkitt lymphoma cells results in an intracellular signal, with inhibition of the tyrosine phosphatase activity of CD45 and therefore phosphorylation of Lyn kinase \[[@B29],[@B30]\]. In this work, we study the effect of AzaC treatment on the lymphoma cell line BL36. As p16^INK4a^ gene has been found to be downregulated by hypermethylation at high frequency in different types of tumours \[[@B6],[@B7]\], it is used as a control of AzaC effect. Then, the phenotypic effects of AzaC are compared to those obtained by the addition of exogenous Gal1. The findings that we report here lend further support to a potential role for Gal1 in the AzaC-induced pathway of differentiation in hematopoietic cells. Materials and Methods ===================== The BL36 B lymphoma cell line \[[@B31]\], a gift from Pr. Lenoir (CIRC, Lyon, France) to Pr. M. Raphael, was maintained in a complete medium of RPMI 1640, containing 10% heat-inactivated foetal calf serum, 2 mM L-Glutamine, 1 mM sodium pyruvate (complete medium). Treatment with AzaC was conducted as follows: 24 hours after seeding 5, 10 or 50 μM of AzaC were added to 5 ml of complete medium \[[@B16],[@B32]\]. The cells were feeded with the appropriate medium during the time of the experiments. Several independent methods were used to assess the proliferative versus death states of the cells. For direct determination of cell number, at the indicated time point the cells were harvested and counted using a Coulter Counter (Beckman Coulter France, Villepinte, France). Determination of the viable cell numbers used propidium iodide (PI) (Sigma, Saint-Quentin Fallavier, France). To determine the percent cell death including both apoptotic and necrotic cells, AzaC-treated cells were analyzed for AnnexinV binding and PI uptake as described \[[@B33]\] using the ApoDetect AnnexinV-FITC kit (Zymed). Briefly, after washing of cultured cells (3.10^5^) with PBS and resuspension in binding buffer, the cells were stained with 10 μl FITC-labelled AnnexinV and 10 μl PI (20 μg/ml). After 5 min of incubation at room temperature in the dark, again 200 μl of binding buffer was added and cells were analyzed by flow cytometry. The percent cell death was calculated by determining the percent of viable cells: The MTS assay (CellTiter 96™, Promega, Charbonnières, France) \[[@B34]\] was used to compare the percentage of metabolically active cells in treated cells vs. untreated cells, as previously described \[[@B14]\]. To study the cell cycle, 2.10^6^ BL36 cells were pelleted at 1000 g for 5 minutes and the pellet was incubated in a mixture of ethanol-PBS (70/30 v/v). The cells were pelleted another time and suspended in 800 μl PBS. One hundred microliters of 1 mg/ml RNase was added to the cell suspension. One hundred microliters of 400 μg/ml PI was added to the solution to stain the nuclear DNA. The DNA content of the cells was determined by a flow cytometer, FACScan (Becton Dickinson, Mountain View, CA), and the percentages of cells in G0+G1, S, and G2+M phases of the cell cycle were analyzed by a polynomial model (SFIT, Becton Dickinson). To determine the cell phenotypes, cells were suspended in PBS supplemented with 2% BSA, and incubated in suspension for 30 min with the fluorescently tagged primary antibody or negative control, or incubated with untagged antibody, washed, and incubated again for 30 min with FITC-conjugated second antibody. All incubations were performed at 4°C. Flow cytometry was performed using a Coulter Epics Elite ESP instrument. Human recombinant Gal1 (rGal1) was obtained as described elsewhere \[[@B14]\]. The protein was purified by affinity chromatography on a lactosyl-divinylsulfone-agarose column. Antibodies against human Gal1 were generated as described in \[[@B35]\] and \[[@B36]\]. The immunolocalization of Gal1 was carried out as described for K562 cells \[[@B24]\]. For the preparation of cell extracts, a total of 17.10^6^ cells were solubilized in 1 ml of extraction buffer: 50 mM Tris pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.1% SDS (w/v), 0.5% sodium deoxycholate (w/v), 0.5% Nonidet NP40 (v/v). Finally, 1 tablet of antiprotease cocktail (Roche, Meylan, France) was added to 10 ml of buffer. The cells were sonicated in ice three times. The lysat was centrifuged at 100, 000 g for 15 min at 4°C. The supernatant was collected and stored at -20°C until used. Proteins were resolved by discontinuous SDS-PAGE on 1.5 mm gel (T: 6--18%) according to the method of Laemmli. \[[@B37]\]. For Western blotting, proteins separated by SDS-PAGE were electrotransferred onto Immobilon-P membrane (Millipore). Blots were incubated with 1:50 anti-Gal1 mAb, or 1:500 anti-pl6 mAb (Pharmingen) for 2 h at room temperature. The blots were developed with anti-mouse Ig antibody-HRP diluted 1/10,000, followed by incubation in the Amplified Opti-4CN kit substrate (Bio-Rad). The image of the membranes was acquired from GS-700 Densitometer 4, and analyzed with Molecular Analyst Software (Bio-Rad). Results ======= During the four days of treatment with AzaC, cell proliferation appeared to slow (Fig. [1A](#F1){ref-type="fig"}). When BL36 cells were cultured with AzaC at the high concentration of 50 μM, cell proliferation was strongly inhibited. BL36 cells exposed to 5 μM AzaC also exhibited a significant reduced growth rate but these cells were \>95% viable by propidium iodide test even at day 4 (data not shown), and were found metabolically active by MTS assay as described below. According to these results, the following experiments addressing metabolic activity, death, and cell differentiation were performed. {#F1} The proportion of metabolically active cells was determined using the MTS test in cultures treated with AzaC in comparison with untreated cells (Fig. [1B](#F1){ref-type="fig"}). The cellular conversion of MTS to the ultraviolet-absorbing formazan product has been demonstrated to be directly proportional to cellular metabolism resulting in the formation of reducing equivalents such as NADH or NADPH \[[@B38]\]. A drop in the % of metabolically active cells of about 36% was observed over the 24 h period following the addition of 10 μM AzaC, and went up to 42% in cells treated with 50 μM AzaC. Cell death was confirmed by an increase in annexinV/PI staining (Fig. [2](#F2){ref-type="fig"}). Within 48 h of incubation with AzaC at 10 μM, the number of AnnexinV^+^ PI^-^ apoptotic cells and AnnexinV^+^ PI^+^ necrotic cells increased. For longer incubation times, induced cell death increased dramatically. {#F2} Studies investigating the mechanisms whereby B lymphoma cells are induced to undergo apoptosis demonstrated that an arrest in the cell cycle preceded apoptosis \[[@B39]\]. To determine whether AzaC modified the cell cycle distribution of BL36 cells, the DNA content of AzaC-treated cells was analyzed by PI staining. Cells were exposed to 10 μM AzaC for 2, 4 and 7 days. After 7 days, AzaC-treated cells exhibited a relative increase of cells in G0/G1 (77 %) in comparison with controls (56%). In the same time, the percentage of cells in the S phase compartment drops from 38 to 18%, suggesting an accumulation of cells into the G1 phase of the cell cycle (Fig. [3](#F3){ref-type="fig"}). However, there was no change in the relative cell cycle distribution at 2 or 4 days after AzaC treatment. By this time, a significant population of cells has undergone apoptosis. This suggests that, at least during the first days of treatment, inhibition of growth is uncoupled with an arrest in the cell cycle. {#F3} The effects of different concentrations of AzaC (5, 10, and 50 μM) on the expression of various surface antigen of BL36 cells were investigated. AzaC increased the cell population that expressed the cell-surface maturation marker CD23, CD30 present on activated B-lymphocytes \[[@B40]\] and CD138 (Syndecan-1). After the treatment with 50 μM AzaC some changes were observed that were not detected at lower concentrations: an increase of the cell population that expressed CD21, and a decrease of the cell population that expressed CD19 that is lost on maturation to plasma cells \[[@B41]\], and CD71 that is expressed on proliferating cells \[[@B42]\] (Table [1](#T1){ref-type="table"}). ::: {#T1 .table-wrap} ::: {.caption} ###### Modulation of urface Markers on BL36 Cells by AzaC ::: Surface marker Control Treatment AzaC = 5 μM AzaC- 10 μM AzaC = 50 μM ---------------- --------- ----------------------- ------------- -------------- CD 19 94.9 88.5 86.1 58.1 CD 21 \<20 \<20 \<20 20.3 CD 23 31.5 47.6 45.2 63.8 CD 30 \<20 \<20 48.4 21.5 CD 45 RA 76.7 74.4 71.2 64.6 CD 45 RO 21.4 35.1 33.3 38.5 CD 71 97.3 94.9 93.8 83.8 CD 138 20 23.9 23.8 24.8 BL36 cells were exposed with 5 μM, 10 μM and 50 μM AzaC for 4 days. Then cell surface antigens were analyzed by flow cytometry. Values below 20% were considered as negative. In the same conditions, CDs 3 and 10 remain negative, and no variability was observed for CDs 11 a, 18, 25, 38, 44, 48, 54, 58, 77, 80, and 95. ::: P16 is a biochemical marker of cell cycle progression \[[@B43]\] and is also commonly utilized as a marker of demethylation reaction \[[@B44]\]. Consequently, we examined the effects of AzaC treatment on expression of p16 in BL36 cells. As expected, when these cells were treated with AzaC for 48 h, we detected the presence of p16 by Western Blotting (Fig. [4A](#F4){ref-type="fig"}). Thus, taken together the data showed that AzaC treatment induced an inhibition of cell growth related to an arrest at G0/G1 phase of the cell cycle, and confirmed the efficiency of AzaC treatment for expressing the tumour suppressor gene p16. {#F4} To determine whether AzaC is able to induce the expression of Gal1 in B lymphoma cells, protein extracts prepared form cells treated with 5--10 μM AzaC and from untreated cells were separated by SDS-PAGE and analyzed by immunoblotting with anti-Gal1 antibody. Fig. [4B,C](#F4){ref-type="fig"} shows the rate of variation in Gal1 expression induced by the differentiating agent. The galectin was not detected either in untreated BL36 cells or BL36 cells treated with AzaC for 24 or 48 h. The galectin became detectable in AzaC-treated cells by 96 h. Then, to determine whether the newly synthesised galectin molecules in AzaC-treated BL36 cells are expressed on the cell surface, the binding of anti-Gal1 Ab to untreated and treated BL36 cells was studied by immunocytochemistry. To visualize only extracellular Gal1, cultures of BL36 cells were incubated with anti-Gal1 antibody without permeabilization of their membranes. Over the 4-day time-course, AzaC-induced Gal1 did progressively localise to the cell surface. However, Gal1 did not exclusively localise to the cell surface but was also distributed throughout the cytoplasm (Fig. [5](#F5){ref-type="fig"}). {#F5} Several of the changes in AzaC-treated BL cells might be attributed directly to Gal1 induction, as Gal1 has been implicated in differentiation and growth inhibition \[[@B11]-[@B13],[@B29],[@B45],[@B46]\]. A series of experiments were thus performed to address whether changes observed in BL36 were direct AzaC effects and/or secondary effects related to Gal1 expression induced by AzaC treatment. To this end, BL36 was treated with recombinant Gal1 (rGal1). Slower growth rates were confirmed for the rGal1-treated cultures. A drop in the % of metabolically active cells of about 28% was observed over the 24 h period following the addition of 700 nM rGal1. However the proportion of metabolically active cells in cultures treated with rGal1 vs. the proportion in untreated cultures was not modified after 4 days of treatment (Fig. [1C](#F1){ref-type="fig"}). Moreover the variation of cell death detected in treated-cells relative to the control cultures was discrete (\<10%) and limited to the first day after the treatment (Fig. [2](#F2){ref-type="fig"}). To determine whether the inhibition of proliferation could be correlated to a modified phenotype, the effects of 700 nM rGal1 on the expression of various surface antigens shown to be modified or not by AzaC were investigated. The populations of cells that expressed CD19, CD23, CD45RO and CD45RA were unchanged. On the other hand, 24 h rGal1 treatment increased the cell population expressing CD138 (Fig. [6](#F6){ref-type="fig"}), a marker for plasma cells, while it decreased the expression of CD71. These modifications were reinforced after 48 h rGal1 treatment (Table [2](#T2){ref-type="table"}). {#F6} ::: {#T2 .table-wrap} ::: {.caption} ###### Modulation of Surface Markers on BL36 Cells by rGal1 ::: Surface Marker Control 24 h Treatment (% positive) rGal1 -- 24 h Control 48 h Treatment (% positive) rGal1-48 h ---------------- -------------- -------------------------------------- -------------- ----------------------------------- CD71 98.00 94.90 93.00 45.00 CD138 \<20 20.20 25.00 70.00 BL36 cells were exposed with 700 nM rGal1 for 1 and 2 days. Then cell surface antigens were analyzed by flow cytometry. Values below 20% were considered as negative. ::: Discussion ========== In 1994, Chiariotti et al. \[[@B16]\] first reported experimental data showing that the expression of Gal1 can be induced in cultured hepatoma-derived cells by treatment with AzaC. Interestingly, Gal1 and AzaC individually have been shown to affect similar cell processes in cancer cells, including differentiation, and growth inhibition \[[@B12]-[@B15],[@B27],[@B29],[@B46]-[@B51]\]. Available data are consistent with the suggestion that the expression of Gal1, accompanied by its secretion and its binding to cell surface receptors, could be involved in the AzaC observed effects in hematopoietic cells where Gal1 modulates differentiation or apoptosis. However, the mechanisms of these effects on hematopoietic cells are unclear yet. In the present study, the effects of AzaC, on the cell phenotype, cell differentiation and cell death of BL36 cells were analyzed. The effects on cellular growth were time- and dose-dependent. Five μM AzaC caused significant inhibition on cellular growth, but the cell viability remained practically unchanged. AzaC increased the cell fraction in the G0/G1 phases, suggesting that AzaC inhibits cell division, which may be one of critical mechanisms of cell modulation by AzaC. AzaC-treated cells initially may become arrested at the G1 phase and then may either escape to the cycle arrest or die due to mechanisms leading to programmed cell death. We also show that incubation of BL36 with AzaC induces expression of Gal1. AzaC is thought to exert its effects as a competitive inhibitor of cytosine methylation, resulting in the expression of silenced genes. The gene for Gal1 is one whose expression is possibly enhanced in this manner. In the results reported here, we found that Gal 1 was detected in cytosol after 120 h of treatment by 10 μM AzaC and then Gal1 was externalized and bound to cell surface receptors 24 h later. A key to understanding the extracellular biological functions of Gal1 is how its secretion appears to be regulated and re-directed during development and differentiation. Gal1 is likely released form vesicles close to the plasma membrane. On the basis of the data, we propose that the released Gal1 be immediately recruited to modulate cell activity. Gal1 may do this by interacting with and modulating cell receptors via its carbohydrate recognition domains because the Gal1-receptor interaction is abrogated by thiodigalactoside \[[@B29]\]. Others and we have previously reported that Gal1 binds to T and B lymphoblastoid cells \[[@B29],[@B52],[@B53]\]. Other studies have demonstrated that galectins are immunosuppressive, in animal models of autoimmune diseases \[[@B54]-[@B56]\]. Whereas the full role of Gal1 in modulating immune function is not yet understood, the increase in Gal1 expression by AzaC in BL cells suggests that Gal1 may play a role in the behaviour of normal leukocytes and of tumour cells. What is the underlying mechanism? Although the regulatory machinery triggered by demethylating stimulus and resulting in phenotype modifications is not yet elucidated, it probably involves the stimulation of a signalling cascade that regulates cell proliferation and viability. A recently proposed model for such a cascade suggests the involvement of a cytoplasmic protein, AZ2 \[[@B57]\]. The amino-terminal part of the AZ2 protein is homologous to the previously reported TANK and I-TRAF, which participate in the signal transduction cascade from the TNF-receptor to the transcription factor NFkappaB. Demethylating stimulus may also modify a pathway activated by the membrane-anchored protein-tyro sine phosphatase CD45. Engagement of CD45 is known to regulate Src tyrosine kinases phosphorylation, phospholipase Cγ regulation, inositol phosphate production, diacylglycerol production, PKC activation, and calcium mobilisation \[[@B58]-[@B60]\]. Increased synthesis and secretion of Gal1 by the cell could account for part of the phenotypic alterations detected in AzaC treated cells. Gal1-induced dimerisation and/or segregation might inhibit the catalytic site in CD45, thereby blocking tyrosine phosphates activity. Because Gal1 binding to cell surface receptors results in tyrosine phosphorylation \[[@B29],[@B61]\], it may allow a kinase-dependent signal to be transduced. Several studies have linked Gal1 expression with growth inhibition \[[@B27]\] and cell death \[[@B26],[@B62]\]. However, the reports that some growth inhibitory agents did not induce Gal1 expression indicated that Gal1 expression is not dependent on the cell\'s growth state in general, through it may be involved in growth suppression \[[@B63]\]. Moreover, it is likely that Gal1 acts in a manner to regulate specific signal transduction processes that is determined by the cell type and by the state of cell differentiation. In this work, exogenous rGal1 added to BL cells inhibited cell growth. Moreover, Gal1 as well as AzaC induced an expression of the cell surface plasma cell antigen, CD138, a phenotypic marker that identify cells with plasmacytic differentiation \[[@B64]\]. This is consistent with the hypothesis that AzaC and Gal1 share similar signals for differentiation, however, since there was a significant difference in the expression of CD19 and CD23 after AzaC or Gal1 treatments it is likely that some pathways are specifically modified by AzaC. The mechanisms involved in these different pathways, important in clinical therapy, remain to be elucidated in the future. Ongoing studies are aimed at identifying as globally as possible the modifications resulting from AzaC treatment by using proteomics \[[@B65]\]. Acknowledgements ================ This work was supported, in part, by grants from the Ministère de l\'Education Nationale de la Recherche et de la Technologie (MENRT), and from the Ligue Française contre le Cancer (Comité de Seine Saint-Denis). FP was supported by ARC (Association pour la Recherche sur le Cancer).

Introduction ============ L-arginine (Arg), an essential amino acid, is required for normal growth of microbes, plants and animals. Deprivation of this amino acid from the culture medium or other sources of nutrition causes serious disturbances in cellular and organ function leading to total destruction. On the other hand, excessive doses of Arg also influence cell function, including cell death and cell proliferation. Substantial information has been obtained in the past decades on the role of Arg in tumour growth and in tumour therapy. Effect of Arg deprivation and supplementation on tumour cell proliferation ========================================================================== Arg, an essential amino acid, is required to maintain normal metabolism and proliferation of cells in culture \[[@B1]\]. Attempts to influence tumour cell proliferation by changes in amino acid balance were based on such observations. The role of the enzyme arginase, which decreases the amount of Arg, was thoroughly investigated in this respect and also used in the therapy of human tumours \[[@B2]\]. According to Umeda *et al.*\[[@B3]\], the proliferation of both HeLa cells in vitro and rat Novikoff hepatoma in vivo could be decreased by arginase, causing relative Arg deficiency. Otsuka \[[@B4]\] has shown that an enzyme, very similar to arginase inhibits DNA synthesis in normal rat liver. The proliferation promoting activity of L-Arg is also underscored by the fact, that Arg is converted by arginase to L-ornithine, which is the precursor of various polyamines essential for cell proliferation \[[@B5]\]. Tanaka *et al.*\[[@B6]\] have demonstrated the death of 3T3 cells after Arg deprivation. Wheatley *et al.*\[[@B7]-[@B10]\] analysed the effect of deprivation of eleven essential amino acids on several tumour cell lines and found that apoptotic-like cell death occurs as a consequence of this manipulation. The cell lines died considerably more quickly during Arg deprivation than in the absence of any other essential amino acids. Moreover, when co-cultures of normal and tumour cells were deprived of Arg the normal cells survived and the tumour cells died. According to these observations, Arg deprivation causes selective death of cultured malignant cells. Lamb and Wheatley \[[@B11]\] have also shown, that Arg deprivation most probably impairs the control of DNA synthesis at the G~1~ checkpoint, which normally inhibits its initiation of DNA synthesis under unfavourable conditions. Arg imbalance was also produced by excess of Arg supplementation in the diet. Brittenden *et al.*\[[@B13]\] Suggested a possible therapeutic effect of Arg-rich diet in malignant disease, in combination with anti-cancer chemotherapy. Ogilvie *et al.*\[[@B14]\] found that excess Arg combined with doxorubicin chemotherapy extended disease-free interval and survival time of dogs with lymphoma. According to the studies of Hester and Fee \[[@B15]\] on squamous cell carcinoma in the CH3/KM mouse the mechanism of action of high amounts of Arg may be the stimulation of host immune surveillance. However, Robinson *et al.*\[[@B16]\] found that Morris hepatoma-bearing rats fed with Arg-rich diet did not show any alteration in tumour growth or cytokine production. The role of Arg in carcinogenesis has been challenged by the experiments of Weinberger *et al.*\[[@B17]\] who found that high doses of Arg glutamate decreased the carcinogenic activity of various acetamine-derivatives in rats. Interesting data were reported on Arg-induced apoptosis of pancreatic acinar cells both in vitro and in vivo \[[@B18]\] providing a model of acute pancreatitis. The possible therapeutic use of Arg against pancreatic acinic cell carcinoma has not been examined yet. Arg-rich hexapeptides were identified from peptide libraries that inhibit the interaction of vascular endothelial growth factor to its receptor. These hexapeptides inhibit the proliferation of human umbilical vein endothelial cells and also block the angiogenesis induced by vascular endothelial growth factor in vivo, in the chick chorioallantoic membrane and in the rabbit cornea. One of the hexapeptides blocks the growth and formation of metastases of HM7 human colon carcinoma cells in nude mice \[[@B19]\]. These results may serve as leads for development of anticancer drugs. Arg imbalance was established in our early in vivo experiments \[[@B20]\]. High doses of L-Arg, D-Arg and DL-Arg (400--500--1000 mg/kg body weight intraperitoneally or orally) were administered to Wistar rats bearing subcutaneous Yoshida\'s sarcoma or to Swiss mice bearing subcutaneous Ehrlich carcinoma for 9--15 days (table [1](#T1){ref-type="table"}). D-Arg inhibited the growth of Yoshida\'s sarcoma significantly (50%, p \< 0.05), when applied in a daily dose of 500 mg/kg, orally. Intraperitoneal administration of the same dose to Ehrlich carcinoma bearing mice resulted in a 20%, statistically not significant, inhibition. ::: {#T1 .table-wrap} ::: {.caption} ###### Effect of D-Arg, L-Arg and DL-Arg on the growth of transplantable animal tumours ::: **Treatment** **Tumour** *Number of animals* **Daily dose (mg/kg)** **Duration of treatment (days)** **Inhibition %** *Remarks* --------------- -------------- --------------------- ------------------------ ---------------------------------- ------------------ ----------- ----------------------- *D-Arg*^1^ Yoshida 6 6 500 p.o. 12 50 p \< 0.05 D-Arg ^1^ Ehrlich s.c. 10 10 500 i.p. 10 20 p \< 0.1 L-Arg^3^ Yoshida 10 10 400 i.p. 10 -40 Enhancement p \< 0.05 L-Arg^3^ Yoshida 10 10 400 p.o. 9 -30 Enhancement p \< 0.1 L-Arg^2^ Ehrlich s.c. 10 10 500 i.p. 12 -40 Enhancement p \< 0.05 DL-Arg^3^ Ehrlich s.c. 10 10 500 i.p. 12 \- \- ^1^Serva (USA), ^2^Ajinomoto (Japan), ^3^Reanal (Hungary) Animals were sacrificed 24 hours after the last treatment and tumour weight was measured. Inhibition or enhancement is given as per cent of the control tumours. Abbreviations:i. p. -- intraperitoneal; p. o. -- peroral; s. c. -- subcutaneous; Arg -- arginine ::: L-Arg, however, enhanced the growth of Yoshida\'s sarcoma, when given intraperitoneally or orally in a dose of 400 mg/kg. The same tendency, namely significant (40%) enhancement was seen after intraperitoneal treatment (400 mg/kg) of Ehrlich carcinoma bearing mice. Intraperitoneal application of 500 mg/kg DL-Arg to mice, inoculated with Ehrlich carcinoma had neither inhibitory nor enhancing effect on tumour growth. Various pathways of metabolism of this amino acid may explain the mode of action of Arg imbalance. Among these methylation and hydroxymethylation appear to be of special importance. Effect of methylated, hydroxy and hydroxymethylated Arg on cell death and proliferation ======================================================================================= Arg is a highly reactive compound, both as a free amino acid and as a constituent of a protein. The structure of proteins can be altered by specific enzymatic modification of the side chains. One of these protein-modifying reactions is methylation, resulting in the addition of methyl groups to the guanidine residues of Arg. Methylated Arg also occur in free form, possibly resulting from enzymatic hydrolysis of methylated proteins in vivo (fig. 1). Mono-di- and trimethyl Arg, hydroxymethyl Arg, N-omega-hydroxy L-Arg, N-nitro-L-Arg methyl ester, Nitro-Arg were studied regarding the possible effect on cell death and cell proliferation of these compounds. Tyihák *et al.*\[[@B21]\] demonstrated that monomethyl Arg and dimethyl-Arg inhibit significantly the growth of tobacco tissue cultures in concentrations of 10--100 ppm in agar nutrient medium, NG-methylated Arg added to agar-medium also significantly inhibited the growth of the roots of lettuce seedlings. Further investigations along this line carried out by Szende *et al.*\[[@B22]\] have shown that NG-hydroxymethyl Arg inhibited dose-dependently and significantly the proliferation of HT-29 human colon carcinoma cells, P-388 mouse lymphoma cells and PC-3 human prostate carcinoma as well as K-562 human erythroleukaemia cells in culture. The cells of the treated cultures showed morphological signs of apoptosis in a high percentage. In our in vivo experiments \[[@B20]\] hydroxymethylated Arg was administered to Swiss mice inoculated with Ehrlich ascites tumour. The daily dose of NG-hydroxymethyl Arg was 400 mg/kg intraperitoneally, based on previous acute toxicity studies. After 10 days of NG-hydroxymethylated Arg treatment complete inhibition of the growth of Ehrlich ascites tumour was observed. C57B1 mice, inoculated with Lewis lung tumour intramuscularly and made tumour free by amputation of the tumorous leg 10 days after tumour transplantation, were treated for seven days with 400 mg/kg NG-hydroxymethyl Arg, daily, intraperitoneally. The treatment started 24 hours after amputation. The animals were sacrificed on the 8^th^ day after starting the treatment. Lung metastasis number and volume were determined. The average metastasis number in the treated animals was 27, in the controls 54. The average volume of the lung nodules was 34 mm^3^ in the treated and 50 in the control mice. The anti-proliferative and apoptosis-inducing effect of Arg-derivatives was confirmed by the studies of Singh *et al.*\[[@B5]\] who found that N-omega-hydroxy-L-Arg inhibited the proliferation of the high-arginase-expressing MDH-MB-468 cells and induced apoptosis after 48 hours. It has also been shown by Washo-Stultz *et al.*\[[@B23]\] that N-nitro-Arg methyl ester sensitised cells to apoptosis induced by sodium deoxycholate. L-Arg, nitro-Arg and methyl-Arg have been found to induce increase in cytosolic Ca concentration in cultured NIT-1 cells \[[@B24]\], leading to depolarisation of the plasma membrane potential, a phenomenon common during the process of apoptosis. NG-methyl-L-Arg \[[@B25]\] N-nitro-L-Arg methyl ester \[[@B23]\], N-hydroxy-L-Arg \[[@B5]\] and L-NG-methyl-Arg \[[@B26]\] all proved to be inhibitors of nitric oxide synthase. Nitric oxide synthase converts L-Arg to produce NO, which \"Janus-faced\" compound certainly may influence both cell proliferation and cell death. Role of formaldehyde in the mechanism of action of methylated arginines ======================================================================= Another possible mode of action of methylated and hydroxymethylated Arg can be deduced from the fact that these molecules are formaldehyde generators. It has been demonstrated by Hullán *et al.*\[[@B27]\] that NG-hydroxymethyl Arg as a biomolecule is one of the compounds that are responsible for the endogenous formaldehyde level. The guanidine group of L-Arg can bind one, two or three molecules of CH~2~O and in the reaction mono-, di- and trihydroxymethylated Arg derivatives are formed. This process is catalysed by the enzyme transmethylase \[[@B28]\]. The hydroxymethylated derivatives of Arg are relatively stable compounds. Arg is suitable to carry the endogenous CH~2~O in form of hydroxymethyl group in biological systems. The hydroxymethyl groups are attached to the guanidine group by reversible bindings \[[@B29]\]. Although little is known about the demethylation of NG-methylated Arg, NG-hydroxymethyl-L-Arg generates a direct CH~2~O-yielding activity, which may be responsible for its apoptotic effect \[[@B22],[@B30]\]. It has also been shown in our recent experiments \[[@B31]\] that the administration of analytically pure formaldehyde to cell cultures causes dose dependently apoptosis (1--10 μg/ml) or stimulation of DNA synthesis and cell proliferation (0.1--0.01 μg/ml). The calculated quantity of formaldehyde released by demethylation processes from hydroxymethyl Arg is in the above mentioned range and the formaldehyde-mediated biological action of this compound has to be taken into consideration. Lysine-arginine antagonism ========================== The lysine-arginine antagonism widely shown in nature is also represented in apoptosis resistant cell lines that contain A-to G alteration is the death domain, encoding L-arginine instead of L-lysine in codon 441 \[[@B32]\]. Arg 72-p53 genotype and cancer ============================== Over-representation of the homozygous Arg 72-p53 genotype in cervical carcinoma patients has been reported \[[@B33]\]. However, Tachezy *et al.*\[[@B34]\] did not find increased risk for human papilloma virus-associated cervical tumour development associated with Arg 72-p53. Conclusions =========== Although data in the literature are pointing to the anti proliferative effect of both Arg depletion and supplementation, Arg proved to be essential for tumour cell growth. This observation raises the question, whether decreasing the concentration of Arg in nutrients and consequently in blood serum or the administration of D-arginine may lead to retardation of tumour growth in humans, too. Arg, because of its strong basic guanidine group, plays an important role in molecular interactions in biological systems, such as interaction between Arg and formaldehyde, both of which are normal components of cells and biological fluids. As a result, hydroxymethyl derivatives of Arg are formed. These compounds may be the source of formaldehyde generation. Arg may be considered a formaldehyde capturer, carrier and generator molecule. These functions may also play role in the biological activity of Arg and its methylated and hydroxymethylated derivatives. An interesting therapeutic possibility worthy of further investigation may be the administration of methylated and hydroxymethylated Arg in order to induce tumour cell death or to prevent tumour cell proliferation.

Background ========== Arterial complications in Behcet\'s disease are rarely seen and usually involve great arteries like aorta and iliac arteries, and may lead to the perforation of the arterial walls and to the development of the aneurysms and their rupture \[[@B1]\]. Due to weakness and fragility of the aortic wall and arteritis open surgery has some complications like the development pseudoaneurysms. In this case report, an infrarenal huge saccular abdominal aneurysm extending to the left inguinal region in a 32 years old male Behcet patient that was first seen in spinal CT examination performed due to extensive back pain and consequently had a successful stent graft placement is presented. Case presentation ================= A 32 years old male patient with intense back pain admitted to the emergency room. From his history it was learnt that he had Behcet disease and being followed by another institution. Direct roentgenograms revealed a smooth contoured density increase on left side of vertebral column (Figure [1](#F1){ref-type="fig"}). Lumbar vertebral CT scan revealed significant destruction on the anteriolateral parts of vertebral corpuses between L2 and L4, more prominent on L3 and hypodense area noted. Then dynamic abdominal CT was performed which showed an infrarenal, 10 × 10 × 10 cm saccular aneurysm originating form a 3.5 × 1 cm defect on the left posterolateral aspect of abdominal aorta, with thrombus inside (Figure [2A, 2B](#F2){ref-type="fig"}). Due to the risk of pseudoaneurysm development digital subtraction angiography (DSA) was not performed and CT angiography (CTA) and 3D reconstructions were made (Figure [3A, 3B, 3C, 3D, 3E](#F3){ref-type="fig"}). CT examinations were performed with GE ProSpeed Helical scanner and for the reformat images and 3D reconstructions Advantage Windows version 2.0 Voxtool software on a separate workstation was used. The patient was not considered an ideal candidate for open surgery and endovascular stent graft placement was planned. All the measurements made by using abdominal CT and CTA images that revealed favorable lengths and diameters for endovascular treatment. A nitinol, polyster covered 14 × 70 mm stent graft (Stenford Grouppe Valendons fa Nanterre, France), beginning from 1.5 cm proximal and 1.5 cm distal to the aortic defect, ordered. Intervention took place in the operating room under spinal epidural anesthesia with the guidance of Schimadzu C-Arm fluoroscopy equipment that had no DSA function. Systemic iv 5000 U heparin was administered. Right common femoral artery was exposed and an arteriotomy was made. A 12 F (75 cm length) introducer was advanced and stent graft was deployed 1.5 cm below the renal arteries. During deployment systemic arterial pressure dropped to 50 mmHg. Intraprocedural control conventional angiography showed the patency of renal arteries and proper position of the deployed stent graft and entrapped contrast material inside the aneurysm sac (Figure [4A, 4B](#F4){ref-type="fig"}). Arteriotomy site was closed by primary sutures to prevent pseudoaneurysm development and wrapped by woven collagen covered Dacron graft (Hemashield, Meadox Medicals, Oakland, NJ). There was no intervention related intra or post procedural complications. Control abdominal CT examination and 3D reconstructions revealed complete sealing of the aneurysm sac by the stent graft, patency of the renal arteries and visceral arteries and no endoleak was noted (Figure [5A, 5B, 5C, 5D, 5E](#F5){ref-type="fig"}). No ischemic complications in the lower extremities were encountered. Patient is discharged with steroid therapy and is being followed up serial abdominal CT. After 18 months there were no complications (Figure [6A, 6B](#F6){ref-type="fig"}). {#F1} {#F2} {#F3} {#F4} {#F5} {#F6} Discussion ========== Behcet disease is first described by Hulusi Behcet, a Turkish dermatologist, in 1937 \[[@B2]\]. It is a multisystemic inflammatory disease that classically causing oral and genital ulcers, and ocular inflammation. It can also affect vascular system. Since it can affect both arteries and veins of all sizes, Behcet vasculopathy is different from other vasculitides \[[@B3]\]. There is no pathognomonic laboratory test or histologic finding specific to Behcet disease. Thus the diagnosis is based on clinical criteria. Various criteria were proposed (like Mason and Barnes, The Behcet\'s disease Research Committee of Japan, O\'Duffy and Goldstein, International Study Group (ISG) etc). The most commonly used criteria of ISG \[[@B4]\] require recurrent oral ulceration plus at least two of the following: recurrent genital ulcerations, eye lesions (like uveitis etc), skin lesions (erythema nodosum, folliculitis etc), positive pathergy test. ISG considered subcutaneous thrombophlebitis, deep vein thrombosis and arterial aneurysms as a criterion. Although these have high specifity for Behcet disease, they are not accepted as criteria due to their low sensitivity. The etiology of Behcet disease is still not certain but systemic vasculitis has an important role. Perivascular lymphocytic and plasma cell infiltration, endothelial cell proliferation and swelling, disruption of elastic lamina, degeneration of the tunica media and vasculitis of the small vessels in vasa vasorum are among the histologic features of venous and arterial lesions \[[@B5],[@B6]\] Behcet vasculopathy includes arterial and venous thrombosis formations, subcutaneous thrombophlebitis, venous collateralizations, pulmonary artery aneurysm, arterial aneurysm and occlusions \[[@B6]\] Arterial aneurysms are more common than arterial thrombosis. Although aorta is the most common site of involvement pulmonary, femoral, subclavian, popliteal and carotid arteries also could be affected. Histologic examinations of the arterial lesions show vasculitis in the vasa vasorum together with thickening in the media and separation of the elastic fibers, which in turn causes development of aneurysms. \[[@B6]-[@B8]\]. Endovascular grafts provide an important alternative to high-risk patients with more invasive surgical procedures. Patients who had endovascular stent graft placement had shorter operating room and hospital stay and less blood loss than open surgery. From the economical stand point the cost of stent graft placement is much higher than open surgery in our country. Although numerous improvements in catheter, stent and graft technology have been made, there are still some limitations like the size of delivery systems, endoleaks, position of the renal arteries etc. The success rate of endovascular stent grafting is high (90% in low risk patients, 80% in moderate to high risk patients). Short-term mortality (within the 30 days following the procedure) is reported to be 2.5% in low risk group, 8% in high-risk group. Although there is disagreement in reported mortality rates for open surgery \[[@B9]\], there are reports of mortality rates of between 1.6 % \[[@B10]\] to 7,6 % \[[@B11],[@B12]\]. As it is common for all new or minimally invasive techniques, there is a learning curve for endovascular stent grafting that affects both the results and costs. Thus for the best results efficiency and experience are needed. One of the crucial points of endovascular treatment of aneurysms is accurate and precise measurement of aneurysm. Although CT is highly effective, most of the vascular radiologists still would like to have preprocedural angiography. But in Behcet disease vessel wall vulnerability is extremely high and even minor lesions resulting from diagnostic procedures (e.g. intraarterial angiography etc.) could lead to the formation of aneurysms at the puncture site \[[@B13]\]. Due to the fragility of the aneurysm inherent in Behcet patients and the size of aneurysm, we did not perform diagnostic angiography. All measurements based on CT images and 3D reconstructions helped us to better understand and characterize the neck, dimensions and relations of the aneurysm to other arteries. Endovascular radiological intervention in the treatment of abdominal aortic aneurysm complications of young Behcet patients, in whom surgery is not suitable and intervention is necessary, could be an alternative treatment modality even without performing preprocedural conventional angiography or DSA using the advantages of helical CT and 3D reconstruction techniques. Competing interests =================== none declared Authors\' contributions ======================= Author 1, RK, conceived of the study, carried out the imaging work-up and the endovascular procedure, and drafted the manuscript. Author 2, OG, carried out the surgical part and participated in drafting the manuscript Author 3, AA, participated in the imaging work-up and en do vascular procedure. Author 4, RT, participated in the surgical part, and coordination. Author 5, TB, participated in the imaging work-up and coordination. All authors read and approved the final manuscript. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ Written consent was obtained from the patient for publication of the patient\'s details.. The manuscript has been presented in the TURK RAD, 17th National Radiology Congress (27--31 September 2000, Istanbul, Turkey) as a poster presentation.

Background ========== Diabetes mellitus constitutes a major Australian public health problem. It has been identified as one of our national health priority areas\[[@B1]\], and recent data suggests that up to 7.5% of the Australian adult population now suffers from this condition. After 20 years of diabetes most patients can be expected to develop diabetic retinopathy\[[@B3]\] which, if undetected, is likely to cause significant visual loss or blindness. A recent Australian survey recorded almost 1/3 of patients with self-reported diabetes having diabetic retinopathy.\[[@B4]\] However, if detected early by appropriate retinal examination, followed and treated by laser photocoagulation, most patients can be saved from major visual loss. \[[@B5]-[@B7]\] Our National Health and Medical Research Council (NHMRC) Guidelines on the management of diabetic retinopathy, recommend regular ocular review to detect, treat and minimize such visual morbidity. \[[@B8]\] Recent Australian studies, however, have documented only a relatively small proportion of Australians with diabetes receiving such sight-saving examinations. The Melbourne Visual Impairment Project, a large population-based epidemiological study of 4,744 Victorians with diabetes, found that only half the sample surveyed reported a retinal examination within the last 2 years. Worryingly, participants with diabetic retinopathy were no more likely to have had an examination than those without (52.9% v 49.3% p = 0.66)\[[@B3]\]. The National Divisions Diabetes Program, a survey of the treatment of diabetes in 4,359 patients across 7 Divisions of General Practice, recorded 49.6% of patients to have had an appropriate retinal examination over the 24 month reporting period. \[[@B9]\] A number of strategies have been implemented to increase the number of Australians with diabetes accessing regular eye checks. Remote clinics with annual \"fly-in\" ophthalmologists operate in a number of Australian states. Mobile retinal cameras, operated by state health departments or Divisions of General Practice, have recently improved community access to screening,\[[@B10]\] but are expensive to purchase and operate. Australian optometrists are offering their services in the area of diabetic retinopathy screening. However, little attention to date has focused around the use of appropriately-trained general practitioners in offering effective and accessible screening for diabetic retinopathy. We aimed to determine whether general practitioners could be trained to identify the important features of diabetic retinopathy in a short training time and with an acceptable degree of sensitivity and specificity. Method ====== Seventeen GPs from the Bayside Division of General Practice self-selected to participate in a Divisional Eye Upskilling Project, funded by the Department of Health and Human Services. As part of this program, the University of Queensland Departments of Ophthalmology and General Practice were asked to design a brief \'hands-on\' educational intervention to assist GP skill development in the ocular assessment and management of diabetes mellitus and primary open angle glaucoma. The complete program and its evaluation has been documented in a previous paper .\[[@B11]\] Details of the diabetic retinopathy training intervention are displayed in Table [1](#T1){ref-type="table"}. ::: {#T1 .table-wrap} ::: {.caption} ###### Diabetes segment of the GP Eye Upskilling Program ::: Workshop 1 Workshop 2 ------------ ----------------------------------- ------------ ----- -------------------------------------------------- --------- 1\. Overview diabetic eye disease 30 mins 1\. Patient fundoscopy clinic 30 mins 2\. Fundoscopy video 15 mins 2\. Question and answer session 15 mins 3\. Fundoscopy practice 30 mins 3\. Use of skills in the general practice time frame 15 mins 4\. Slide review session 15 mins 4\. Slide review 15 mins 5\. Patient fundoscopy clinic session 35 mins ::: All sessions were highly interactive and used patients for skill demonstration wherever possible. These sessions were conducted within the Professorial Eye Unit, Princess Alexandra Hospital, using patients with good clinical signs. Doctors were urged to practice and refine skills in their own practices between workshops, and to comment on barriers and supports to the use of the skills in their clinical practice. All GPs underwent a comprehensive practical assessment of their eye skills prior to undertaking the intervention, and completed an identical assessment at the conclusion (pre and post test). The pre and post evaluation relevant to the diabetes retinopathy skills component consisted of eight (8) two-minute patient fundal examinations conducted a month prior and a month following the skills program. Four (4) of the patients involved in this assessment suffered from diabetic retinopathy, and four (4) had normal fundi.. The gold standard used to determine the presence or absence of diabetic retinopathy was the agreed clinical assessment by two academic ophthalmologists. All four diabetic (4) patients had background diabetic retinopathy. Every attempt was made to make pre and post evaluation assessments for the diabetes patients identical. However, one of the four diabetic patients participating in the pre-test, was unable to return for the final evaluation session. He was replaced, as closely as possible, by a patient with similar fundal appearance. The four patients with normal fundi were recruited from the ward and outpatient departments on the day of assessment. One \"normal\" failed to attend the post-test. The post test was in all other respects conducted identically to the pre test. GPs were not informed of the retinopathy status of the patients used in the pre-evaluation testing and these patients were separated from clinical teaching sessions during the educational intervention. All patients were unknown to participating GPs. The University of Queensland ophthalmology staff supervised all training sessions and both assessments. Ethical approval to conduct the study was sought and received from the Princess Alexandra Hospital Ethics Committee. Statistical method ------------------ Participant performance on the patient fundal assessment, pre and post educational intervention, was used to determine sensitivity and specificity of screening for diabetic retinopathy. Sensitivity was calculated as the ratio of patient assessments correctly identified by the GPs as diabetic, to the total number of diabetic retinopathy positive cases (four subjects in total in both pre and post tests). Specificity was calculated as a ratio of the number of normal cases correctly identified over the total number of normals participating in both tests (four subjects in the pre-test, and three subjects in the post-test). \[[@B12]\] For each GP, sensitivity pre and post educational intervention was classified \'not satisfactory\' if sensitivity was less than 60% and classified \'satisfactory\' if sensitivity was equal to or greater than 60%. For each GP, specificity pre and post educational intervention was classified \'not satisfactory\' if specificity was less than 60% and classified \'satisfactory\' if specificity was equal to or greater than 60%. McNemar\'s Test was then used to determine if there was an association between Sensitivity Classification and educational intervention. This test was used because it accounts for the paired nature of the data. That is, the same GPs were evaluated Pre- and Post-workshop. Statistical significance is quoted at the conventional p \< 0.05 level. All hypothesis testing was based on two-tailed hypotheses. Due to small sample sizes exact significance levels were used. Results ======= Sensitivity ----------- Pre and post intervention sensitivity results are displayed as percentages in Table [3](#T3){ref-type="table"}. Ten GPs (59%) achieved a screening sensitivity of 25% or less in the pre test, three (18%) scored 50%, two (11.5%) scored 75%, and two (11.5%) recorded 100% sensitivity. In the post test, all seventeen GPs achieved between 50 and 100% sensitivity (1 GP achieved 50% sensitivity, 11 GPs (65%) achieved 75% sensitivity, and 5 GPs (29%) achieved 100% sensitivity). ::: {#T3 .table-wrap} ::: {.caption} ###### Sensitivity of GP screening for Diabetic Retinopathy (n = 17) ::: Pre-workshop Post-workshop ----- -------------- --------------- 0 4 0 25 6 0 50 3 1 75 2 11 100 2 5 ::: The McNemar test (Table [4](#T4){ref-type="table"}) was statistically significant and there was an association between sensitivity classification and educational invention (p = 0.001). There was a significant increase in the number GPs with \'satisfactory\' sensitivity from Pre- to Post-workshop. ::: {#T4 .table-wrap} ::: {.caption} ###### Specificity of GP screening for Diabetic Retinopathy (n = 17) ::: Pre-workshop Post-workshop -------------- --------------- ------ --- 0 3 0 1 25 6 50 4 33.3 3 75 4 66.6 6 100 0 100 7 ::: Specificity ----------- Specificity results, pre and post intervention, are displayed in Table [5](#T5){ref-type="table"}. As one of our \"normals\" failed to appear in the post test, the number of subjects used in the post assessment dropped from four to three, hence the specificities are expressed in increments of 25% for the pre test compared with increments of 33% for the post test. In the pre test, nine GPs (54%) recorded a screening specificity of less than 50%, four (23%) a specificity of 50% and four (23%) a specificity of 75%. In the post test, only four GPs (23%) recorded a specificity of less than 50%, six (35%) a specificity of 66%, and seven (41%) 100% specificity. No GP recorded 100% specificity in the pre test, whilst 7 (41%) recorded 100% screening specificity following the intervention. ::: {#T5 .table-wrap} ::: {.caption} ###### Sensitivity Classification of GP screening for Diabetic Retinopathy by Pre-and Post-workshop ::: Pre-workshop Total --------------- ------------------ -------------- ------- ---- Post-workshop not satisfactory 1 0 1 satisfactory 11 5 16 Total 12 5 17 ::: The McNemar test (Table [6](#T6){ref-type="table"}) was statistically significant and there was an association between specificity classification and educational invention (p = 0.001). There was a significant increase in the number GPs with \'satisfactory\' specificity from Pre- to Post-workshop. ::: {#T6 .table-wrap} ::: {.caption} ###### Specificity Classification of GP screening for Diabetic Retinopathy by Pre-and Post-workshop ::: Pre-workshop Total --------------- ------------------ -------------- ------- ---- Post-workshop not satisfactory 3 1 4 satisfactory 10 3 13 Total 13 4 17 ::: The NH&MRC publication \"Management of Diabetic Retinopathy: a guide for general practitioners\' \[[@B8]\] nominates that a \"60% detection rate of early Diabetic Retinopathy (DR) may be sufficient for a successful screening campaign, so long as diabetic patients deemed to not have DR are screened at least every two years.\" In this study, 24% of GPs met this criterion prior to the workshop, and 94% met it following the intervention. Discussion ========== The early identification of diabetic retinopathy is crucial, as effective therapy is now available for the maintenance of vision in most patients. However, the absence of a systematic approach to such identification for Australians with diabetes limits our ability to provide optimal visual outcomes. The limited number of ophthalmologists available in the community restricts their role to the confirmation of diabetic retinopathy, grading, and eventual treatment. The use of a mobile fundus camera and central reading of all photos remains a potential method of identification of retinopathy, although as yet this has not been widely embraced in any single large community (outside remote areas). GP screening is rarely suggested as an effective means of improving the early detection of diabetic retinopathy in Australia, yet it offers a number of clear benefits. General practitioners have almost universal access to Australian adults with diabetes, as GPs provide the bulk of adult diabetes diagnosis and care. As well as managing glucose control, medication, and lifestyle issues, GPs perform regular checks for peripheral neuropathy, nephropathy and macro vascular disease. An assessment for retinopathy could easily be included as part of the existing annual diabetes assessment (as nominated under the latest RACGP / DA Guidelines). This would negate the need to purchase and move expensive machinery, or arrange separate appointment times and assessments, and would be cost-effective for both the patient and community. A general practice screening model also offers the potential to opportunistically \"catch\" unscreened diabetics presenting to the general practitioner in a variety of other situations. General practitioners have a pre-established referral link with local ophthalmologists if early diabetic retinopathy is suspected. In most Australian states, this relationship has increasingly included an educational element via the RACGP / RACO National GP Eye Skills Program. A GP screening model thus has the potential to provide a widely-available, holistic and extremely cost effective screening service on a population basis -- providing general practitioner skills are adequate for the identification process. Previous work with general practitioners in this area suggested to us that GP reluctance to undertake retinal screening was due to concerns about the time taken for the examination; a long-standing and unfounded fear of the consequences of pupillary dilatation; and a lack of confidence in undertaking funduscopy with a direct ophthalmoscope. McCarty et al (13) identify lack of dilating drops in the practice, lack of confidence in detecting changes, concern re time taken and the fear of precipitation of angle-closure glaucoma as major barriers to GPs performing dilated ophthalmoscopy with their patients. Strategies to address all of these issues and perceptions were addressed as part of the \"\_How will this change my practice?\" module in our educational intervention (Table [2](#T2){ref-type="table"}). Follow up of a separate group of GPs taking this intervention as part of the national RACGP / RACO GP Eye Skills Workshop has demonstrated a sustained change in reported practice in this area \[[@B14]\]. We piloted a brief interactive training intervention to address these issues. Our GP intervention group reviewed and accepted recent data on the rarity of acute glaucoma following pupillary dilitation with short-acting mydriatics, and rapidly acquired effective skills in dilated diabetic fundal assessment. A focus of the training was the application of skills in each individual practitioner\'s work environment, and the practicing of skills between sessions in the general practice setting. This pilot aimed to assess the efficacy of this teaching program in raising the skills of those involved to the level of the current NHMRC guidelines for diabetic retinopathy screening. However, until patients are screened more regularly, the detection rate for diabetic retinopathy may need concomitantly to be higher than the 60% nominated in this guideline. ::: {#T2 .table-wrap} ::: {.caption} ###### Workshop Modules (National RACGP / RACO GP Eye Skills Workshop) ::: ----------- ----------------------------------------- Module 1 Vision Testing And Ophthalmic Screening Module 2 Concepts And Assessment For Glaucoma Module 3 The Slit Lamp Module 4 Opthalmoscopy Module 5 Incision Of Chalazion Module 6 Removal Of Corneal Foreign Bodies Module 7 Suture Repair Of Entropion Module 8 Ocular First Aid Module 9 The Six Point Eye Examination Module 10 How Will This Change My Practice? Module 11 Patient Assessment Clinic (Optional) ----------- ----------------------------------------- ::: Following training, 94 % of the general practice study group reached this level within a two minute patient examination (both eyes). In addition, the specificity of the identification was such that it would not provide a gross referral overload to ophthalmologists for more accurate assessment and treatment. The time and effort required could be easily integrated into an annual review for patients with diabetes. The Royal Australian Colleges of General Practice and Ophthalmology have implemented this workshop across Queensland in 2000. With appropriate training, motivation and support, GPs can make a major contribution in the area of effective screening for diabetic retinopathy. Clinicians, educators, bureaucrats and economists within our health system should be aware of this in framing future strategies in the area of diabetes. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ The authors would like to thank the project manager Manjula Reedy, and the Bayside GPs for their contributions.

Background ========== Collagen in the intervertebral disc, tendons, ligaments, heart valve leaflets, intestine and cornea displays a wavy collagen pattern termed \"crimping\". This crimp morphology contributes an important component to the mechanical responsiveness of these tissues \[[@B1]-[@B4]\]. When these tissues experience loading, the collagen crimp architecture is gradually straightened; this makes possible the slight elongation of collagen at the loaded site. The biomechanical behavior of the disc is closely linked to this specialized crimp morphology (Figure [1](#F1){ref-type="fig"}) \[[@B2],[@B3]\]. The collagen fibers which form the annulus possess the same planar crimped geometry as is seen in the tendon. In the disc, however, crimp parameters vary according to the radial distance through the annulus. The lamellar structure of the annulus consists of continuous layers of collagen fibers that encircle the nucleus pulposus. In successive lamellar layers, the fiber orientation changes with respect to the spinal column axis. From the outer annulus inward, the interlamellar angle decreases linearly. Within each lamellar bundle, the collagen fibers are aligned in parallel arrays, and the fibers display a planar crimped waveform with all fibers in register. A gradient of crimp angle also is present in the disc, with crimp angle increasing and the crimp period decreasing from the periphery of the disc inward. ![Diagrammatic illustration of the organization of the intervertebral disc with attention to lamellar structure in the annulus and crimping in the collagen fibers. (Reproduced with permission \[[@B3]\].](1471-2474-3-9-1){#F1} This complex and sophisticated architectural organization of lamellar disc structure serves to meet the unique biomechanical needs of the healthy disc. Although it is an important aspect of disc structure and function, this topic, which addresses the interface of microscopic anatomy and biomechanical function, has not been studied with specific reference to changes in the disc architecture during aging and degeneration. In the present report we suggest that the important stretch response of collagen made possible by collagen crimping is altered by morphologic changes in disc architecture during aging and degeneration. This alteration in the crimping behavior of collagen may thus contribute to the underlying tissue changes which culminate in fissures and annual tears in the degenerating disc. Methods ======= Control (normal) donor disc specimens were obtained from disc surgeries and the NCI Cooperative Human Tissue Network (CHTN). Studies were approved by the human subjects Institutional Review Board. Disc tissue studied here was from the annulus. Patient specimens were obtained from surgical procedures performed on individuals with herniated discs. Electron microscopy was used in the study of a total of 78 subjects (29 were from control donors \[aged newborn to 79 years\], and 49 were from surgical specimens from patients \[aged 16 to 77 years\] undergoing disc surgery). Light microscopic studies utilized the annulus from 40 control donors and 53 surgical specimens. The majority of the specimens studied here were from anterior sites of lumbar discs. Surgical specimens were transported to the lab within 30 minutes of surgical removal in sterile modified minimal essential medium with Earle\'s salts (MEM, Gibco, Grand Island, NY) with 1% (v/v) L-glutamine (Irvine Scientific, Santa Ana, CA), 1% (v/v) nonessential amino acids (Irvine Scientific), and 1% (v/v) penicillin-streptomycin (Irvine Scientific). CHTN specimens were shipped overnight in sterile MEM. Upon receipt into the lab, specimens from control donors were carefully dissected to obtain specimens from the outer annulus, inner annulus and nucleus. For electron microscopy, specimens were minced and fixed in Karnovsky\'s fixative, post-fixed with osmium tetroxide supplemented with 0.1% ruthenium red to enhance proteoglycan visualization in the extracellular matrix \[[@B5],[@B6]\], embedded in Spurr resin, thin sectioned and grid stained with uranyl acetate and lead citrate. In order to obtain representative specimens to examine with electron microscopy, 4--6 blocks of tissue were trimmed, thick-sectioned and examined with light microscopy and 1--2 blocks (in which the plane of sectioning was parallel to the long axis of the collagen fibrils) selected for thin sectioning. Sections were viewed on a Phillips CM10 transmission electron microscope. For light microscopy, specimens were fixed in 10% neutral buffered formalin, embedded in paraffin or glycol methacrylate, and stained with picrosirius red-alcian blue, Masson trichrome, or Goldner\'s stain. Three-dimensional intensity graphical presentation were obtained using BioScan^®^ OPTIMAS™ (BioScan, Edmonds, WA). Morphologic images were captured with a CoolCam 2000 video system and frame grabber (Cool Camera Company, Decatur, GA) and prints were prepared with a Sony Mavigraph color video printer (Sony Electronics, San Jose, CA). Results ======= Opportune sectioning planes through the annulus reveal collagen crimping at both the ultrastructural and light microscopic level. Figure [2A](#F2){ref-type="fig"} illustrates the normal annular morphology of a healthy young intervertebral disc. Disc cells are flattened and lie in a modest lacunar space within collagen lamellar bundles. Prominent wavy expanses of collagen (collagen crimping) are present. Aging and degenerating disc matrix, however, shows striking alterations in tissue architecture. Changes include the presence of layers of unusual extracellular matrix encircling disc cells and interterritorial matrix alterations which extend into the surrounding lamellar structure (Figure [2B](#F2){ref-type="fig"}). In these regions collagen fibers are fewer in number and fibers are separated from each other reducing collagen crimping areas. Sites of more advanced degeneration can contain areas where there is prominent focal loss of matrix (Figure [2C](#F2){ref-type="fig"}). Only sparse regions of collagen remain in such sites. {ref-type="fig"} and [2B](#F2){ref-type="fig"}: Photomicrographs of light microscopic features of the annulus in a disc from a three month old infant (A) and a control disc from a 68 year old female (B and C). Arrows in A mark regions of prominent collagen crimping. In B note the presence of abnormal matrix encircling disc cells and decreased matrix in intraterritorial regions. In Fig. [2C](#F2){ref-type="fig"} a region of prominent focal loss of matrix is present near encapsulated cells. (A, glycol methacrylate embedment, Goldner\'s stain; B and C, paraffin with Masson-trichrome; X 980).](1471-2474-3-9-2){#F2} At the ultrastructural level, wavy collagen bands characterize regions of good matrix (Figure [3A](#F3){ref-type="fig"}). In more degenerated areas, collagen fiber number is much less (Figure [3B](#F3){ref-type="fig"}). {ref-type="fig"}) which contrasts with sparser collagen in a region with more prominent matrix loss (Fig. [3B](#F3){ref-type="fig"}). (Surgical specimen from a 35 year old female. (Fig. [3A](#F3){ref-type="fig"}, X 5,900; Fig. [3B](#F3){ref-type="fig"}, X 13, 940).](1471-2474-3-9-3){#F3} When viewed with polarizing light microscopy, regions of the disc with collagen crimping stand out prominently, and visualization of crimping patterns can be enhanced by application of specialized image analysis techniques which provide a graphical representation of image intensities (Figure [4](#F4){ref-type="fig"}). In regions of good disc morphology, collagen crimping patterns produce image intensity graphs where crimp patterns show clear peaks and valleys in register (Figure [4A](#F4){ref-type="fig"} and [4B](#F4){ref-type="fig"}). In areas where degenerative changes in the matrix are prominent, however, collagen crimping is markedly disrupted (Figure [4D](#F4){ref-type="fig"}). Note that the area examined in Figure [4D](#F4){ref-type="fig"} shows some crimp morphology (arrows), thus confirming that the plane of sectioning was appropriate for the detection of collagen crimping. Figure [4C](#F4){ref-type="fig"} shows that rare collagen bundles in register with each other are still present in this area, thus verifying that the tissue section plane was appropriate for detection of collagen crimping. {ref-type="fig"}, surgical specimen from a 46 year old female) and from a region with marked matrix changes (Fig. [4D](#F4){ref-type="fig"}, surgical specimen from a 54 year old male). In Figures [4A](#F4){ref-type="fig"} and [4C](#F4){ref-type="fig"}, collagen crimping patterns throughout the same microscopic fields shown in Figs. [4B](#F4){ref-type="fig"} and [4D](#F4){ref-type="fig"} are converted to graphic images with three-dimensional intensity images using BioScan^®^ OPTIMAS™ using the following settings for both images: rotation: x, 25.000; y, 1.000; sampling, x: 45, y:45, and viewpoint Z: 3000.0 . White arrows in Fig. [4D](#F4){ref-type="fig"} mark small regions showing collagen crimping, but degenerative changes in the tissue disrupt crimping in the adjacent regions. (Figs. [4B](#F4){ref-type="fig"} and [4D](#F4){ref-type="fig"}, polarized light, picrosirius red/alcian blue, X 102).](1471-2474-3-9-4){#F4} Discussion ========== In order for the disc to successfully withstand mechanical forces, its matrix must have the appropriate structural features, and this matrix must be maintained by appropriate turnover of its molecules. Previous studies from our laboratory have identified similar matrix changes during disc aging and degeneration in disc tissue from both control subjects and patients undergoing disc surgery \[[@B7],[@B8]\]. Observations presented here focus upon secondary collagen changes involving crimping morphology which nonetheless have important consequences for tissue biomechanical function. During aging and degeneration of the intervertebral disc, disruption of annular fibers during stress can result in fissures which may extend out to the disc margin \[[@B9]\]. Such radial tears may be a source of the nonspecific early spinal symptoms in disc degeneration \[[@B9]\]. Although the importance of annular tears has long been recognized, the basis for these tears at the architectural level has not been investigated. In the present work, study of a collection of disc specimens from control and surgical patients has shown that the architectural features of collagen crimping in the annulus can be disrupted in three fashions. First, a common morphologic change during disc aging/degeneration is the accumulation of wide layers of unusual matrix which encircle disc cells and clusters of disc cells \[[@B7],[@B8]\]. These regions disrupt normal collagen fibrillar layers and thus reduce the functional stretch afforded by collagen crimping in regions of normal collagen. Secondly, in areas where the matrix has become sparse, only a few collagen fibers are present. Such regions would have diminished capability of affording appropriate stretch function to resist tearing under compression. Finally, marked extracellular matrix loss in focal areas altered normal lamellar collagen architecture. The presence of similar types of changes in both control and surgical disc tissue is not surprising in light of previous studies by Boden et al \[[@B10],[@B11]\] in which degenerative changes were identified with MRI in lumbar and cervical discs of older asymptomatic subjects. In other ultrastructural studies of disc tissue, we have reported considerable variation in cross-sectional diameters of collagen fibrils in both territorial and inter-territorial extracellular matrix \[[@B8]\]. Cross-sectional diameter variability was present in 34% of control and 59% of surgical specimens which were examined. Such variations may also have a possible effect on the ability of the disc to respond to biomechanical stress. A limitation of the present study is the qualitative nature of our study. Future studies would be greatly strengthened by quantitative analysis of variations of crimping periodicity in both surgical specimens and in representative sections of complete discs. Such quantitative studies are important since histologic/ultrastructural examination only focuses upon a small portion of tissue. Additional important information lacking from our study is determination of the extent to which the changes described here actually alter the biomechanical properties of the disc. We hope that our small study here will lead to more complete studies of this important topic since there is much to be learned about how disc degeneration relates to increased risk of herniation and pain. Little is understood about collagen turnover in either the healthy or diseased intervertebral disc. Studies by Antoniou et al have pointed out that it is very difficult to distinguish between synthesis of new matrix, degradation products, and resident matrix collagen \[[@B12]\]. It has been suggested that the turnover in the disc may be in excess of 100 years \[[@B9]\]. Newer methods used by Antoniou et al reveal that synthesis of Type I and II collagen and of aggrecan all decrease from age 2 to 15, but Type I collagen appears to decline more rapidly and to plateau during ages 15--40 years, whereas Type II collagen and aggrecan continue to decline during this period \[[@B12]\]. At older ages (60--80 years), there is increased denaturation of Type II collagen and a small increase in the synthesis of Type I procollagen. The maintenance of appropriate matrix composition in the disc is complicated by a marked decline in the number of disc cells with aging and an increase in cellular death by apoptosis \[[@B7]\]. Conclusions =========== In summary, use of our large disc tissue collection was an advantage in the present study since it provided examination of numerous disc specimens sampled through various planes of sectioning. Application of image intensity graphs provided objective tracings of collagen patterns. Although proteoglycan loss has long been acknowledged to be an important part of disc aging and degeneration, the work presented here suggests that important stretch responses of collagen made possible by collagen crimping may be markedly altered by morphologic changes in disc architecture during aging/degeneration; such changes may contribute to the early tissue changes involved in annular tears. This work points to the need to more fully understand the dynamic relationship between degenerative changes in disc architecture at the microscopic level and resultant alterations in biomechanical function. Authors\' contributions ======================= ENH contributed disc tissue from surgeries on patients, and participated in the study and reviewed findings with HEG. HEG conceived of the study and participated in its design, coordination, reviewed all light and electron microscopic material, and wrote the manuscript. All authors read and approved the final manuscript Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ The authors thank Jane Ingram and Tracie McClain for assistance with light microscopy, and Daisy Ridings, Pat McCoy and Winston Wiggins for assistance with electron microscopy.

Background ========== The distribution range of elephant populations in southern Africa where elephants are not restricted by humans largely overlaps with the distribution of mopane (*Colophospermum mopane*) woodlands. Mopane is a principal food item in the diet of elephants, not only in northern Botswana \[[@B1]\], but also elsewhere in southern Africa \[[@B4],[@B20],[@B28],[@B30]\]. High utilization rates of plants by prolific elephant population may result with the alteration of vegetation structure and even the decline of species diversity \[[@B9]\]. Some mopane woodland areas in northern Botswana appear to sustain higher elephant utilization rates than neighbouring mopane areas. The availability of water from surface seasonal water sources can only partially explain the distribution of elephant browsing \[[@B1],[@B31]\]. Two findings relating to factors that affect the feeding habits of other large herbivores suggest possible additional explanations for the patchy browsing by elephants on mopane. First, large herbivorous species are influenced by elevated nutrient concentrations in the forage \[[@B7],[@B24]\]. And second, the feeding patterns and the differential distribution range of large herbivores can be explained by localized differences in soil factors and nutrient levels in the forage \[[@B5],[@B22]\]. The elephant population in northern Botswana has been growing steadily over the past two decades \[[@B8],[@B21]\]. Already, some reports cautioned against the spread of excessive elephant damage to woodlands \[[@B15],[@B23],[@B27]\]. This risk gives practical urgency to identifying the factors that determine the pattern in which food -- in this case mopane -- is used by elephants. In particular, we seek to test here whether high mopane utilization by elephants is predicted by the higher nutrient qualities of either mopane itself or of the soil on which it grows. In tackling this question we assume that if elephants showed selective feeding in northern Botswana, then elephants would also display feeding preferences within a principal and abundant food source. Results ======= Soils, mopane woodlands and elephant utilization ------------------------------------------------ Analysis of variance for the independent parameters measured in each plot revealed high variations of variables related to soil properties and vegetation structure, but low variation in leaf protein contents (Table [1](#T1){ref-type="table"}). ::: {#T1 .table-wrap} ::: {.caption} ###### Mean values and standard deviations for variables describing soil properties, plant densities, and extent of elephant utilization in Mopane woodlands of northern Botswana. ::: Mean Std ------------------------------------------ -------- ------- pH 6.6481 0.23 Nitrogen (kg/acre) 27.18 19.26 Phosphorus (kg/acre) 40.27 23.15 Potassium (kg/acre) 133.51 36.28 Silt (%) 8.82 4.52 Clay (%) 3.67 0.91 Shrub (No/100 m^2^) 12.26 9.98 Tree (No/100 m^2^) 0.60 0.44 New elephant damage (No of utilized mopane shrubs /100 m^2^) 1.40 0.93 Old elephant damage (No of utilized mopane shrubs /100 m^2^) 14.71 6.83 Protein in leaves (%) 11.45 0.93 ::: Figure [2](#F2){ref-type="fig"} describes the influence of the geographical position of plots on the levels of protein in mopane leaves. Although the range of leaf protein was narrow across the plots (10.2--13.5%), it is possible to distinguish apart between 3 different clusters of points representing plots that correspond to separate sites selected for the analysis. Accordingly, Savuti had the lowest protein contents in leaves. South Gate dominated the medium range and plots of the Khwai River site had the highest protein contents in mopane leaves. Multiple comparisons indicated significant differences between the sites (T = 2.96, α = 0.05), except for Savuti and South Gate. {#F2} Further multiple comparisons indicated that the densities of shrubs did not differ significantly between the three sites. Similarly, the extent of new elephant damage to mopane plants did not differ significantly between the sites. Tree densities and the extent of old elephant damage however, differed significantly only between Savuti and South Gate. The match between the two observations can be explained in that mature trees showed signs of past elephant utilization. Levels of phosphorus in the soils did not differ significantly between the sites. Potassium levels differed significantly between the Savuti and Khwai sites. Soil nitrogen levels differed significantly between the three sites except for Savuti and South Gate (T = 2.96, α = 0.05). The matrix of variables incorporated for the multivariate analyses included 10 rows (each representing the 10 individual plots) and 8 columns (each representing one environmental parameter). The environmental parameters were tree densities, shrub densities, nitrogen (**N**), potassium (**K**) and phosphorus (**P**) contents of soils, protein contents of mopane leaves (**Pr**), densities of mopane shrubs and trees in plots, and new (**NE**) and old elephant damage (**OE**) to mopane plants. Figure [3](#F3){ref-type="fig"} provides the summary analysis of these parameters. The ordination of the first principal axes described 90.2% of the variation in the analysis. However, the test of Monte Carlo permutations indicated that the associations between variables representing plots and environmental parameters were not significant (P = 0.11). The geographical classification of sites related to clusters of neighbouring plots was retained to some extent in the analysis. Plots in the Savuti area however, showed a wide dispersal range along the principal axes due to high variability among the parameters examined. Although protein levels in mopane leaves differed between plots the differences did not add much to the overall variation in the analysis. The proximity of the point representing protein levels to the origin of the two principal axes indicates that there were poor associations between elevated protein levels in mopane leaves and high elephant damage to mopane plants or any other variable included in the analysis. {#F3} A breakdown of associations between variables to linear correlations between pairs of variables revealed only two significant correlations among the environmental variables measured. The associations described elevated soil nitrogen and potassium levels with increasing protein contents in mopane leaves (R^2^ = 0.68 and 0.64, P = 0.002 and 0.04 respectively). Discussion ========== A simple assumption would be that herbivores maximize their energy intake by selecting the most nutritious food items available within their home range. In reality however, this assumption is confounded by many ecological factors that reduce the energy return of potentially nutritious feed items. We already know that mopane woodlands are principal food items for elephants in northern Botswana and that the proximity of water sources also influences feeding patterns \[[@B1]\]. These two factors were imbedded a priori in the sampling design in an attempt to assess the magnitude of other factors that may influence the choice of food sources by elephants. The physiognomic diversity of mopane woodlands that vary from dense stands of shrubs to scattered tall canopied trees was not explained by either of soil factors examined. Nonetheless, soil factors seem to affect leaf chemical composition. High protein levels in mopane leaves correlated with elevated levels of nitrogen and phosphorus in the soils. Although nutrient intake by plants is generally independent of concentrations in the supporting substrate, this pattern suggests that mopane plants in some areas accumulate more nutrients than mopane plants elsewhere in northern Botswana. These differences were evident when comparing between mopane plots that were situated more than 30 kilometers apart. The question is whether the differential nutrient levels in mopane leaves influence the feeding patterns of elephants? Local soil properties, shrub and tree densities and protein levels in mopane leaves appear not influence elephants\' choice of feeding sites dominated by mopane. Thus, the seemly patchy distribution of elephant utilization of mopane shrubs in northern Botswana was not corroborated by the results. It seems that the differences in protein levels are too subtle as far as the elephants are concerned, or do not offset other factors, such as distance to permanent water sources and the proximity of other food sources. Furthermore, other chemical properties in leaves, such as carbohydrate content and secondary chemicals that inhibit digestion may influence elephant feeding habits than initially perceived. Management implications ----------------------- The importance of mopane plants as a food source for elephants cannot be underestimated in northern Botswana where woodlands grow on much of the dry land area. The resilience of mopane both in terms of survival rates of individual plants and recovery of biomass was demonstrated before \[[@B2],[@B3]\]. Yet, some mopane woodland areas in the region are utilized intensively by elephants and damage to these adds credence to support the notion that a prolific elephant population in some parts of the region may be so numerous as to become unsustainable by local woodland habitats. The results presented here indicated that elephants appear not to differentiate between areas with respect to the protein contents of mopane leaves. This to some extent diminishes the likelihood that elephants will concentrate feeding in a particular (high protein) area until they exhaust the available resources and destroy the vegetation in the process. Therefore, in contrast to early concerns, areas that are currently heavily utilized are unlikely to be further decimated by elephants. Insofar as this finding indicates a certain component of stability to the northern Botswana ecosystem \[[@B4]\]. Methods ======= The study area -------------- In this study, northern Botswana is defined as an area of more than 80,000 km^2^ between 18° and 21° south and 21° and 26° east, but excluding the permanent swamps of the Okavango Delta. Rainfall ranges from 400 mm in the south and 650 mm in the north-east. Rains occur during the summer from October to April and mainly in heavy storms. Temperatures vary between monthly mean maximum of 34° (October) and a mean minimum of 6°-7° (June) (Bhalotra 1987) \[[@B6]\]. The topography of the area is very flat and the altitude is around 950 m asl. The north easterly elongated Goha and Shinamba Hills, which rise to 150 m above the surrounding plain, form the only significant topographical feature. The area is covered with deep deposits of aeolian and alluvial sands and silts of the Kalahari beds. These soils are generally neutral or slightly acid, poor in nutrients, poorly structured and with low water holding capacity. More fertile vertisols and Molapo soils are mainly found on flood plains and in some depressions. The bedrock consists mainly of Karoo sediments, material of the Ganzi, Kwebe and Damara formations and granitoid gneisses of the old basement complex. Bedrock is exposed in the eastern and south-eastern parts of Botswana where the sand cover is thin. In the northern, central and western parts, solid rock is obscured by loosely consolidated deposits, known collectively as the Kalahari Beds, which attain thickness of 300 m or more \[[@B16]\]. Most of the area is covered by tree and shrub savanna, towards the north, changing into woodland and to dry deciduous forest with *Baikiaea plurijuga* dominating in the Chobe area while the south is dominated by shrub savanna, rich in *Acacia* species. Low lying areas and the edge of flood plains are often dominated by mopane woodlands. Soil and vegetation plots ------------------------- The relationship between soil factors, mopane plants and elephant utilization of plants was examined in 10 plots located in 3 sites within Chobe National Park and Moremi Game Reserve in northern Botswana, namely South Gate, the Khwai River and Savuti. To include a measure of the spatial relationships between plots, we chose one plot at the Khwai site to represent a benchmark point and measured the distances of the rest of the plots from that plot. The distance between plots situated within a site varied from 1 to 7 kilometers. Plots were allocated to cover as much as possible the existing regional range in soil properties, judged by the appearance of soil colour and texture. To override the influence of surface water as a source for the attraction of elephants, all plots were not situated further than 10 kilometers away from a permanent water source. Additional criteria for the allocation of plots included varying densities and physiognomic forms of mopane plants and evidence of elephant damage to plants. The position of the 3 sites followed a regional stratification to vegetation communities \[[@B10]\] and also the annual estimates of elephant densities \[[@B21],[@B10]\]. In each plot, squares of 10 × 10 m were allocated within two elongated rectangular areas that were parallel to the direction and on each side of a nearby track. The size of each plot varied with respect to the density of the mopane plants within. In all, the size of plots did not exceed 500 × 300 m. The number of squares per plot was determined after conducting an analysis of variance for plots where at least 150 squares were sampled. The allocation of squares within a plot was not contiguous to assure the independent status of squares \[[@B18]\]. Sampling took place during the middle of the dry season (August) when herbivores are expected to be more selective and also to offset variations in the concentrations of minerals in mopane leaves because of seasonal fluctuations \[[@B28]\]. In each square, the number of woody plants of the respective dominating species belonging to specific height classes was counted. Height classes ascribed for shrubs (\>1 and \<3 m) and trees (\>10 m) were represented in terms of the mean number of plants/100^2^ m (Table [1](#T1){ref-type="table"}). The category of shrubs included fallen trees, the individual branches of which were considered as shrubs if the size criteria were met. Multi-stemmed plants with stem diameters of less than 6 cm were treated as a single shrub unless a gap wider than the width of 10 cm could be distinguished between the stems. Trees were defined as plants with a stem diameter of 6 cm or more and higher than 3 m. Woody plants that corresponded to the definition of a tree and had several stems that were jointed close to the ground were considered as one tree. In all, 9,858 mopane plants were sampled. Elephant utilization -------------------- The frequency of mopane plants utilized by elephants in the assigned plots was recorded in addition to the estimation of vegetation structure. Damage to woody plants as a result of elephant activity was estimated for individual plants within each square as being as either new or old elephant damage, or both. New damage is that which had occurred since the recent rainy season. Damage becomes characteristically greyish in colour after rain soaks into the exposed inner plant parts. Damage in this context is not only excessive vegetation destruction, but is all easily noticeable utilization of woody material. Only damage that was positively attributed to elephants was recorded. Hence, damage was estimated in terms of the number of mopane shrubs utilized per square of 10 × 10 m \[[@B1]\]. Soil properties and leaf protein -------------------------------- Thirty soil samples were taken from the woodland plots in locations that were thought to represent the typical features of each plot, such as within tall tree stands, dense shrub formations, near plants that sustained high or low elephant utilization. Ten soil samples were taken within an area of 300 × 150 m to investigate the local variation in soil properties in one mopane woodland site. All samples were taken from 5 to 10 cm below the surface that represented the horizon at which plant root development is substantial \[[@B12],[@B20]\]. Each soil sample was subjected to five tests that included soil texture composition, pH and contents of nitrogen, phosphorus and potassium using a sampling kit \[[@B19]\]. The proportions of sand, silt and clay were derived by particle composition analysis using soil separation tubes and texture dispersing reagents \[[@B13]\]. Soil pH was determined through saturating the samples in a Tricon flocculating solution and indicator. Soil extracts were utilized for the determination of nitrogen, phosphorus and potassium contents \[[@B17]\]. Reagents for nitrogen included sodium bisulphate monohydrate sulfanilamide and ammonium sulphate. Stannous chloride, sodium molybdate and potassium chloride solutions indicated phosphorus levels. Sodium nitrite and sodium cobaltinitrite were the principal reagents for the extraction of potassium. Mopane leaves were sampled in each of the selected 10 plots. Crude protein was extracted from leaves of 40 to 50 mopane shrubs in each plot. To eliminate the seasonal effects on protein levels, the leaves were cut within the span of a few days during the dry season. The leaves were cut from shrubs between 1--3 meter height. The reasoning behind the selection of this particular height range is because elephants prefer to feed upon mopane shrubs and less from trees and seedlings \[[@B1]\]. Total nitrogen was determined after a wet digestion with sulphuric acid, by the Kjeldahl method \[[@B17]\]. The method does not separate between the organic and inorganic quantities of nitrogen (mostly in the form of nitrate) in leaves. However, it is unlikely that under conditions of low soil nitrogen and ample light in northern Botswana that plants accumulate large quantities of inorganic nitrogen and that these quantities vary significantly between the different areas. Data analysis ------------- Records of soil properties, leaf protein, plant densities and number of plants utilized by elephants were subjected to ANOVA, primarily to examine the extent of variation within a sampling plot \[[@B26]\]. A multivariate analysis provided a comparative measure for the variance between records of individual plots. The analysis produced an overview of the associations between plots and variables related to soil properties, vegetation structure, and elephant utilization. Canonical correspondence analysis \[[@B29]\], an eigenvector ordination technique for displaying variables in a two-way contingency table was used to generate ordination axes. These axes are linear combinations of a matrix in which plots were represented by columns of categories for soil properties, densities of shrubs and trees, and elephant damage. The matrix of variables consisted of eight rows representing each sampling plot. A Monte Carlo permutation test examined the significance of the first canonical ordination axis. Essentially, the test identified whether the associations between plots and environmental parameters were significant. The results of multivariate analyses were complemented by multiple comparisons and linear regression analyses \[[@B25]\]. The former compared between the mean values derived for plots and sites (T critical = 2.96, α = 0.05). Several multivariate analyses were performed on the same data set to examine the stability of associations between variables before deriving the final results. The analyses included Detrended Correspondence Analysis, Principal Component Analysis and Canonical Correlations \[[@B29]\]. Canonical Corrections with symmetric scaling of variables provided the optimal presentation of variables although the arrangement of variables along the principal axes remained consistent with the other analyses performed \[[@B29]\]. Authors\' contributions ======================= RBS conducted a 10 years study on the ecology of elephants in northern Botswana and drafted the manuscript. DWM conducted many studies on the behaviour and ecology of carnivores and large herbivores. {#F1} Acknowledgments =============== RBS thanks the Office of the President and the Department of Wildlife and National Parks for a permission to do this research. RBS thanks P. Gadilmang and M. Bathami for their enthusiastic assistance in the collection of data. We thank Dr. Steven Hill for constructive advice. This project was funded by the National Geographic Society.

Background ========== Delivering care to people with diabetes --------------------------------------- There is broad, international agreement over what constitutes high quality health care for people with diabetes \[[@B1],[@B2]\]. This will be enshrined in a National Service Framework for people with diabetes, due in summer 2002. However, in the face of poor current performance \[[@B3],[@B4]\] the most efficient method of delivering care remains unclear \[[@B4]\]. Following a 1994 systematic literature review \[[@B5]\] suggesting structured care improved patient care, an editorial in the British Medical Journal concluded that more evaluative research was needed before widespread adoption of any of the models could be recommended \[[@B6]\]. A subsequent systematic review of routine surveillance of patients with diabetes by Griffin and Kinmonth \[[@B7]\] concluded *\"Computerised central recall, with prompting for patients and their family doctors, can achieve standards of care as good or better than hospital outpatient care, at least in the short term. The evidence supports provision of regular prompted recall and review of people with diabetes by willing general practitioners and demonstrates that this can be achieved, if suitable organisation is in place\'.* However, the evidence base on which these conclusions are based is limited in several ways. Firstly there are only five randomised controlled trials (RCTs) involving 1058 patients. All of these studies are \'patient randomised\" trials, thus potentially under-estimating the effectiveness of the intervention (see Study Design). They were all evaluating more or less selected patients and general practices and none of them were explicitly evaluating a UK National Health Service (NHS) service area wide intervention. Only one of the four UK based studies evaluated patient based outcomes and included an economic assessment and this study only involved patients from three general practices \[[@B8]\]. Thus, the effectiveness of an area wide, patient focussed, structured recall and management system (in terms of process of care, patient outcome and economic impact) remains unknown. The current system ------------------ The current computerised diabetes management system runs in Hartlepool, Easington and Stockton, three Primary Care Group (PCG) areas, in the Northern and Yorkshire Region. It was introduced to all 36 general practices in Hartlepool and Easington Districts in mid-1995. Stockton (25 practices) agreed to join the system in 1999 and it was operational there by October 2000. There are three key components to the current system: 1\. A central register of patients with diabetes. 2\. A structured minimum dataset to be completed and returned to the central register. 3\. The provision of both patient specific and aggregated data to both patients and clinicians. The system (developed by Westman Medical Software) allows three methods of collection of data at each contact with a patient with diabetes who is registered on the database. Two methods use a standard form completed by clinicians to collect data concordant with the UK minimum data set \[[@B9]\]. Within secondary care, forms are completed at every new patient or annual review. In primary care, forms are completed by the practice nurse (usually) or general practitioner, either opportunistically or at practice diabetic clinics. In both cases, the completed data forms are sent to the Diabetes Register Facilitator for data entry. Thirdly, the hospital laboratory provides a monthly download of laboratory test details (e.g. HbA1c). A patient can be identified as having diabetes and added to the register by any permutation of one or more of these three routes. Feedback of individual patients\' data, including review status, is provided to general practices quarterly. This feedback is \'passive\' in that it does not explicitly prompt either patients or doctors as to required actions. Audit packages within the software can audit on every variable collected. District wide audit is provided on anonymised aggregated data; individual practice audits (with comparisons to other practices) are provided to participating practices at least annually. Feedback of the data to the patient (for hospital patients only) is by a patient information sheet and to the GP as a standardised letter. A Diabetes Register Facilitator co-ordinates and updates the register. A steering group composed of GH, the Diabetes Register Facilitator and representatives of the PCGs and patients, oversees the register and deals with issues such as confidentiality. ### Impact of the system to date Measures of the impact of the system to date relate only to Hartlepool, Easington and Stockton. The main impact on patient registration was in its first 12 to 18 months of operation: during 1995, 747 patients were registered on the system (0.4% prevalence) which had increased to 3867 (1.8% prevalence) by the end of 1996. The increase in registration has stabilised since then, reaching 4324 (2% prevalence) by 1999. During 1999, 70% of registered patients attended a clinic; 52% had their feet examined and 51% had their eyes examined. Seventy three per cent had an HbA1c result recorded and 69% a blood pressure measurement. These figures are similar to those reported by other centres using the same system \[[@B10]\]. ### The need for an extended system Recording of clinical measures increased during the first few years of operation of the system but began to plateau more recently (for example, 50% of patients had an HbA1c recorded during 1996, compared to 60% in 1997 and 63% in 1998). This plateauing of performance has been reported by others \[[@B10]\]. We believe that this is due to a lack of coordination (patients being lost to follow up) and lack of prompting of clinicians to deliver appropriate clinical interventions. Furthermore, given that most patients with diabetes are primarily seen in primary care the greatest potential impact is from optimising and extending the system in primary care. In order to address these shortcomings the additional key components, over and above those already in the system, will be: 1\. Locally adapted evidence based guidelines for the management and follow up of patients with diabetes. 2\. Automated prompts to patients and primary care clinicians that a review consultation is necessary. 3\. A structured management sheet (including patient specific management suggestions based on (1)). 4\. An enhanced monitoring system to follow up reasons for non-attendance from both patients and clinicians and to re-schedule appointments, based on nonreturn of a completed management sheet. 5\. Patient feedback for patients in primary care. There is some limited supportive trial evidence for these developments, although the existing studies involved small sample sizes and may not be generalisable to the NHS \[[@B11]-[@B13]\]. In evaluating the system with these extended features this study will also address the design shortcomings of previous studies of shared care in diabetes \[[@B6]\]. It will be tailored to each practice, PCG defined areas will be studied, rather than an unrepresentative sample of general practices; and the system will be transparent and replicable in other areas. Methods ======= Design of the study ------------------- The study design is a pragmatic two-arm cluster randomised controlled trial. The unit of randomisation will be the general practice. Simple patient randomised trials are rightly considered the most robust method of assessing most health care innovations \[[@B14]\]. This design, however, cannot be regarded as the gold standard for evaluating systematic approaches to chronic disease management, an essentially behavioural field of research \[[@B15],[@B16]\]. If both intervention and control patients were to be cared for within the same practice there is the risk that the management of control patients would be influenced by the practitioners knowledge of the care of intervention patients. This would result in an underestimation of the effect of the intervention \[[@B17]\]. Therefore, practices rather than patients are the appropriate unit of randomisation and analysis. As the current system has been in place for different lengths of time within the three participating PCGs, we will stratify the randomisation by PCG. Randomisation will be performed by a statistician independent of the research team using computer generated numbers to avoid allocation bias \[[@B18]\]. Study setting and recruitment of practices ------------------------------------------ The study will be based in the general practices of the three PCGs of Easington, Hartlepool and Stockton. Since the recent merger of Hartlepool and North Tees Acute Trusts all three PCGs are now exclusively served by one secondary care diabetes service (and thus the one diabetes register). GH is the lead clinician for diabetes services in the new Trust. The 61 general practices in the three PCGs constitute the target practices for the study and we will attempt to recruit all practices. The PCG diabetes leads or the PCG clinical governance leads in all three PCGs have provided letters confirming their support for the project. We do not envisage major difficulties with recruitment, given the need to agree local guidelines as part of the process involved in the Trust merger, the likely requirements in the forthcoming National Service Framework for diabetes, and the 100% practice coverage with the current diabetes system. We will (through the PCGs) write to all practices, giving information about the project to the senior partner or diabetes lead and practice manager of practices. Practices will be invited to opt out if they do not wish to be included in the study -- this is an approach we have used successfully before. The PCG diabetes lead, clinical governance lead and GH will be co-signatories of this letter. If practices do decline we will collect data on characteristics of non-participating practices to assess the impact on the generalisability of the trial\'s findings. Finally, if there are significant problems with recruitment, there are other practices which could be approached in a nearby PCG (South Tyneside) which uses the same software for its diabetes register. Details of the intervention --------------------------- ### Local guidelines and management prompts A guideline development group will be established to develop local guidelines for the management of diabetes, based upon available evidence based guidelines (Scottish Intercollegiate Guidelines Network \[[@B19]-[@B22]\](SIGN, 1996, 1997a, 1997b, 1997c), and Effective Care Bulletins \[[@B23],[@B24]\]. They will also use the forthcoming national diabetes guidelines as these become available. The group will be multidisciplinary and contain primary and secondary care doctors and nurses, patients and the Diabetes Register Facilitator \[[@B25]\]. The group will define review periods for specified patient groups (e.g. patients with diabetes satisfactorily controlled on diet alone should be reviewed every 12 months), referral criteria for patients moving from primary to secondary care and back and simple decision rules for the management prompts. These would be of two types. The first would prompt for actions to be performed and only require their performance to be documented (e.g. asking for a foot examination to be performed in a patient who does not have a recorded foot examination). The second would be more complex and suggest alterations to clinical management on the basis of data in the database (e.g. patients with persistently raised blood pressure should have their anti-hypertensive medication increased). These decision rules will be integrated into the recall and management system. ### Running the system The proposed enhancements to the system are designed to require the primary care team to perform no additional work over and above the current configuration. The current database has a patient identifier, a minimum dataset and retrieval systems to support the structured recall of patients. Westman Medical Software has agreed to amend the system as required. A \'circle of information exchange\' will be established between the participating general practices and the database. The local guidelines will be used to adapt the current centralised database, along with the practices\' preferred method of following up patients (for example, within consultations in routine surgeries or within special clinics). The central database system will identify when patients are due for review (based upon the local guidelines) and will generate a letter to the patient asking them to make an appointment for a review consultation. Patient information or educational materials could be included with the letter. At the same time, the central database will generate a letter to the practice stating that the patient should be making a review appointment in the near future. The letter to the practice will include a management sheet (to be held in the patient\'s record) to capture an agreed minimum data set to be collected during the consultation. This management sheet will also contain the relevant prompts (as described above). When the patient is seen in the practice, the primary care professional (currently this is usually done by the practice nurse) will complete the management sheet and return a copy for entry onto the central register within a designated period of time. This circle of information is broken if the patient does not visit the general practice as planned or the general practice does not return the management sheet to the central register. If this happens, the central register would alert the Diabetes Register Facilitator who will ascertain the reason for failure and take appropriate action, (e.g. send a reminder to the patient, prompt the practice to return the management sheet). A range of educational activities will be provided for intervention practices, as part of the usual local structures for contact with practices, with some additions, These will include: distribution of information about the trial in local newsletters; meetings with practice clinical governance leads; evening meetings for practice nurses (with small group discussion of the practical implications for intervention practices); and a telephone meeting with the practice diabetes lead (usually the practice nurse) in each intervention practice. Practices in the control arm will continue to receive the recall system as currently configured. ### Logistical considerations From the prevalence of patients with diabetes on the current register, there will be about 7500 patients on the system if 61 practices are recruited. Half of these will be in intervention practices. On current patterns of usage, we anticipate there being the need for 1.5 recalls per annum per patient on the register, resulting in about 6000 recalls per year for the intervention group. Assuming a 40 week working year, the system will need to dispatch, receive and process about 150 forms per group per week. ### Identification of patients Patients for the structured recall and management system are already identified on the Hartlepool and North Tees database. As some practices have children registered on the system, who are under the care of an exclusively secondary care adolescent service, an age limit of 18 years or over will be set for inclusion. Practices will be asked to check lists of their patients on the database regularly throughout the study. The central database will remove patients from the recall system who are known to have died or moved away. ### Patient consent Patients have already consented, or are being consented, to their data being held within the current diabetes register. The study will involve no extra \'routine\' data being collected, and this data will be anonymised before being sent for analysis; all data held for analysis will be held in accordance with the Data Protection Act. For the patient-based questionnaire study, we will seek additional patient consent to complete one survey. The three relevant Local Research Ethics Committees have approved the trial. Data collection --------------- The main study outcome measures will be rates of performance of process of care and the patient based measures of functional and psychosocial wellbeing. Data will be collected for 15 months after the start of the intervention. Fifteen months was chosen to allow for patients who are reviewed every 12 months but fail to attend on initial invitation. ### Process of care variables Process of care variables will be collected via the computerised database. The exact data to be collected will be determined by both the current content of the database and the guidelines but will include such data items as rates of attendance at clinics and annual reviews, conduct of eye and feet examinations, performance of investigations and prescribing. We will also collect data on clinical measures (e.g. HbA1c, and blood pressure levels). ### Outcome of care measures Outcome of care data will be collected, by postal questionnaire, 15 months after commencement of the study. A portfolio of validated \[[@B26]-[@B30]\] and responsive \[[@B30],[@B31]\] generic and disease specific instruments will be used to measure functional and psychosocial variables that will be potentially influenced by the intervention. These will include: i\) The SF36 health status profile which we will use to generate Mental (MCS) and Physical Component Summary Scales (PCS) \[[@B26],[@B27],[@B32]\]. ii\) The Newcastle Diabetes Symptoms Questionnaire \[[@B28]\]. iii\) The Bradley Treatment Satisfaction Questionnaire \[[@B29]\]. Patient costs questions will be developed by the study health economist. We have successfully used such packages of questionnaires within trials before and have achieved response rates in excess of 70% in similar surveys in this region. \[[@B33]\]. Sample size considerations -------------------------- On the basis of previous work we have made the following assumptions. The mean number of patients per practice for whom we will be able to collect process data will be 30 and the ICC (a measure of the lack of independence of responses from patients from the same practice) calculated from our local data is 0.14 for measures of process (whether a blood pressure measurement and whether an HbA1c measurement has been recorded in a 12 month period). Standard methods for determining the sample size requirements for a cluster randomised trial \[[@B34]\] indicate that we need 60 practices to detect a difference of 15% (42.5% v 57.5%) with 80% power assuming a significance level of 5%. Assessment of outcome of care will be based on health status scales such as the SF-36. Previous work has shown that this type of intervention is likely to produce an effect size of approximately 0.25 in such measures \[[@B32]\] and that the ICCs for such measures will be approximately 0.07. The most efficient study design (that minimises the number of patients required) is one that makes use of all the available practices. A sample of 27 patients from each of 61 practices will give us 85% power to detect an effect size of 0.25 assuming a significance level of 5%. With a predicted response rate of approximately 70% (based on our experience in the COGENT study \[[@B33]\]) after two reminders, our starting sample size will need to be 2379 patients (approximately 39 patients per practice). Principles of data analysis --------------------------- Analysis will be by intention to treat. Multilevel modelling (using the MlwiN package \[[@B35]\]) will be used to take into account the clustering of patients within practices \[[@B36]\]. Both binary variables (when a process was undertaken or not) and continuous variables (such as the physical health component of the SF-36) can be analysed using these techniques. For both types of variable, variation between practices will be fitted as a random effect and the difference between intervention and control practices will be fitted as a fixed effect. In the case of binary variables, a logit link function will be used. Economic evaluation ------------------- The economic impact of implementing the new structured recall and management system will be evaluated in terms of the marginal costs of adapting and running the system; the costs of developing and disseminating the guidelines; the educational activities for intervention practices; the implications for the use of health care services; and the costs to the patients and their carers. The benefits will be measured as described earlier on in the clinical study. The estimation of health service resource use will relate to diabetes-specific clinical visits, tests, investigations, and procedures. This data will be routinely collected as part of the management system implementation and subsequent costing, using health service pay and price data, will be undertaken using a mixed approach based on micro-costing and gross-costing methods \[[@B37]\]. Use of drugs, referrals to secondary care and the impact of the intervention on the change of use of patients\' and their carers\' time will also be monitored through postal questionnaires at the end of the follow-up period. A sensitivity analysis will be undertaken to test the robustness of the results to the uncertainty not related to sampling variations and to enhance the generalisability of the results \[[@B38]\]. We are aware that the costs of the system might be balanced only in the longer term against the cost savings related to averted complications \[[@B39]-[@B41]\]. However, the assessment of the benefits in terms of final outcomes (e.g lives saved, or QALYs) and long term costs is beyond the objective of the present study. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ The study is funded by Diabetes UK and the Northern and Yorkshire NHS Executive Research and Development Programme. We thank our collaborators at North Tees and Hartlepool NHS Trust, particularly Dr Carr, Dr MacLeod, Joanne Clayton and John Fitzsimmons; at Easington Primary Care Group and North Tees Primary Care Trust; and at South Tyneside Primary Care Group and South Tyneside Healthcare NHS Trust, particularly Drs Bradshaw and Parr, Clare Beard and Wynn Schembrie.

Background ========== The beta-thalassemia syndromes (thalassemia major and beta-thalassemia intermedia) originate from the absence of or from a reduction in the synthesis of structurally normal beta-globin chains. The beta-globin chain deficiency contributes to the pathogenesis of anemia by impairing the formation of functional Hb A and by causing ineffective erythropoiesis. The anemia stimulates the secretion of Epo and leads to erythroid hyperplasia of the bone marrow and to an accelerated erythropoiesis \[[@B1]\]. As a result, erythroblasts expressing the fetal globin gene are generated and fetal hemoglobin (Hb F) is synthesized in the postnatal life \[[@B2]\]. Two more factors, the amplification of the cellular clones producing F-cells and the slower turnover of these cells in circulating blood favour the Hb F presence in thalassemia \[[@B3]\]. In these affections the persistent synthesis of Hb F is very important because it restores the alpha/non-alpha globin ratio, thus reducing the bone marrow hyperplasia and ineffective erythropoiesis and therefore the severity of disease. Genetically, the beta-thalassemia syndromes are heterogeneous. The severity of these disorders is determined primarily by the type of the gene defect as well as the gene dosage. The molecular analysis has uncovered that different clinical variants of beta-thalassemia are associated with a wide spectrum of different mutations of the beta-globin gene; thalassemia major patients, characterized by a severe transfusion-dependent anemia and a marked increase in the synthesis of Hb F, are either homozygous carriers of mutations that abolish the beta-globin synthesis (beta° thai) or compound heterozygotes for beta° thal and severe beta^+^ thal defects which drastically reduce it. These mutations are referred to as severe beta-thal mutations. beta-thal int patients which have a less severe anemia and a modest increase in the synthesis of Hb F, are carriers mainly of beta^+^ thal mutations that reduce slightly beta-globin synthesis. These mutations are referred to as mild beta-thal mutations. Furthermore, deletions of the non-alpha globin gene cluster encompassing the beta- and delta-globin genes are associated with a genetic variant of beta-thalassemia (deltabeta thal), which is characterized by a marked increase in the synthesis of Hb F even in the heterozygous state. In these patients the gamma-globin genes remain intact and are expressed in the postnatal life, thus reducing the ineffective erythropoiesis and ameliorating the severity of the disease. It is worth noting that also genetic factors unlinked to the beta globin cluster, favour the Hb F synthesis in the post-fetal life \[[@B4]\]. Particularly interesting for its frequency is the polimorphic single-base substitution C→T at Ggamma -158, recognizable for the creation of a XmnI restriction site, that appears to cause an increase of Hb F expecially during erythropoietic stress. Recently, it has been suggested that the increased synthesis of Hb F associated with beta-thalassemia syndromes is directly or indirectly dependent from the Epo level \[[@B5],[@B6]\]. The aim of this study was to investigate the relationships between the genetic defects and the severity of the disease in patients affected by a thalassemic disease. In particular, the relationships between the genetic defect, the total Hb, and between the synthesis of Hb F and the Epo level have been examined in patients with beta thalassemia intermedia (beta thal int). In addition, the distribution and frequency of point mutations of the promoter region of gamma-globin genes were analysed and the relationship between these mutations and the content of Hb F was investigated. The analysis of the beta-globin gene and gamma-globin gene mutations was extended also to patients affected by Cooley\'s disease. Subjects -------- Seventy-four patients with clinical picture of beta thal int have been examined: 33 of them were affected by a mild or sub-silent form of the disease, and 41 by an evident clinical picture (Tab. [1](#T1){ref-type="table"}). In addition, 58 patients carried a beta thal defect and also a triplicated or quadruplicated alpha locus. Twenty-four patients were affected by thalassemia major, and 15 by Hb H disease. The control group included 53 healthy individuals and 42 iron-deficient, non thalassemic patients. ::: {#T1 .table-wrap} ::: {.caption} ###### Hematologic and hemoglobinic data and Epo levels in thalassemic patients and control subjects. ::: **Group of patients** **N. cases** **Hb g/dl** **Hb F %** **EPO mU/l** **Glob. synth. alpha/non-alpha** **Ret. ‰** -------------------------------------------------------------------- -------------- ------------- ------------------------------- -------------------------------- ---------------------------------- ------------ Sub-silent beta thal int 33 10.1 ± 1.0 20.5 ± 18.2 3.0 → .70.6 41.22 ± 32.00 11.66 → 136.6 2.10 31 Evident beta thal int 41 8.3 ± 1.16 33.40 ± 28.04 6.80 → 85.0 95.83 ± 64.33 17.84 → 218.97 2.20 60 beta thal int in double heterozygous beta thal + triplicated alpha 58 10.0 ± 1.38 5.15 ± 3.9 2.0 → 19.3 28.45 ± 17.11^\*^ 7.85 → 60.72 2.50 30 Hb H disease 15 9.4 ± 1.2 \< 1 32.88 ± 12.11^\*\*^ 0.52 28 Thal major 24 10.4 ± 1.0 7.7 ± 8.5^\*\*\*^ 1.60 → 40.9 46.83 ± 40.78 9.82 → 129.59 17 Healthy non-thal subjects 53 14.4 ± 1.1 \<1 11.90 ± 3.3 \~1.0 3 Iron-deficient non-thal subjects 42 8.7 ± 1.2 \<1 160.10 ± 312.6 2.7 → 1960.0 8 ^\*^ The measurement of Epo has been detected only in 17 patients. ^\*\*^ The measurement of Epo has been detected only in 10 patients. ^\*\*\*^ This value is not informative as patients were routinely transfused. ::: Patients with an homozygotic condition for a delta beta thal deletion were excluded from this study, since only Hb F is synthesized. The patients homozygotes for severe beta-globin defects, and carriers of an alpha thal defect, were also excluded. All of the individuals examined in this study were in the post-puberal age. Methods ======= Blood samples were collected in EDTA. Hematologic parameters were determined using a Technicon H1 Analyser (Bayer). Hb A~2~, Hb F and Hb H values were determined using an electrophoretic method \[[@B7]\]. Concentrations of Hb F higher than normal were confirmed by radial immunodiffusion (Quick Plate kit, Helena Laboratories, Beaumont, TX, USA). Serum iron and ferritin levels were measured by standard methods. The serum level of erythropoietin (Epo) was determined using the Epo-Trac kit (INSTAR Corporation, Stillwater, MN, USA). The average value of 53 healthy individuals was assumed as the normal Epo concentration. Reticulocyte counts were performed on fresh blood preparations stained with brilliant-cresyl-blue. The synthesis of globin chains in vitro was determined following the method of Clegg \[[@B8]\], using a high pressure liquid chromatograph. The Amplification Refractory Mutation System (ARMS) was used to detect mutations of the beta- and gamma-globin genes \[[@B9],[@B10]\] and a modified PCR method previously described in our laboratories \[[@B11]\], was used to detect alpha-globin genes defects. The patients affected by thalassemia major or intermedia were screened for all of the beta- and alpha-globin gene mutations present in the Italian population (Tab. [4](#T4){ref-type="table"}, ref. \[[@B4],[@B12]\]) and the mutations analysis was also extended to mutations of the promoter region of the gamma-globin genes which are until reported to occur in the Italian population (Ggamma-158 C→T; Agamma -196 C→T; Ggamma -175 T→C; Agamma -117 G→A; AgammaT and the deletion Agamma -225/-222 bp). The Ggamma -158 C→T and AgammaT substitutions were searched in all of the subjects reported in Table [1](#T1){ref-type="table"}. The remaining mutations of the gamma genes promoters were searched in 286 more chromosomes of beta-thalassemia patients or control subjects. ::: {#T4 .table-wrap} ::: {.caption} ###### Distribution of beta thal alleles in beta thal int and thal major patients. ::: **Type of thalassemic disease** **N. cases** **n. of identified alleles** **Alleles beta thal silent** **Alleles beta+ thal mild** **Alleles beta+ thal severe and beta ° thal** -------------------------------------------------------------------- -------------- ------------------------------ ------------------------------ ----------------------------- ----------------------------------------------- Sub-silent thal int 33 61 22 (37%) 8 (13%) 31 (50%) Evident thal int 41 81 9 (12%) 36 (44%) 36 (44%) beta thal int in double heterozygous beta thal + triplicated alpha 58 100 0 0 100 (100%) Thal major 24 47 1 (2.1%) 7 (15%) 39 (83%) The silent, mild, severe beta-thal defects present in Italy are: 1. Silent: -101 C→T; -92 C→T; IVS II-844 C→G; triplicated alpha^anti\ 3.7^; triplicated alpha^anti\ 4.2^: ; they have a hematologic-hemoglobinic picture completely normal or with some mild thal characters. 2. mild or sub-silent: -87 C→G or C→T; -86 C→G or C→A; IVS-1 6 T→C that have a constant but very mild thal hematologic picture and a mild increase of Hb A~2~. 3. severe: beta°39 C→T; IVS 1--110 G→A; IVS II-745 C→G; frameshift 6 (-A) that have all very marked hematologic and hemoglobinic characters. ::: The statistical significance was evaluated using the Student\'s t-test. Linear regressions were fitted by the least square deviations method and the correlation coefficient r was calculated according to a standard formula. A p value \< 0.05 was assumed to be statistically significant, as only few tests were performed. Results ======= The analysis of the hematologic data and the clinical symptoms of 74 patients affected by beta-thal int indicate that this disease can be classified into two clinical variants referred to as sub-silent and evident beta-thal int (Table [1](#T1){ref-type="table"}). The patients with sub-silent beta-thal int have a reduced total Hb content and show no or very mild clinical symptoms. The patients with evident beta-thal int show severe clinical symptoms, often associated with massive splenomegaly, and 41% of them have undergone splenectomy. The Hb content is significantly lower than that of the patient with sub-silent beta-thal int (t = 7.44; gl 72; p \< 0.001). The Hb F level is higher in with evident than in patients with sub-silent beta-thal int, and the difference is statistically significant (t = 2.28; gl 72; p \< 0.05). Thus, the severity of clinical symptoms of beta-thal int correlates with a decrease of the total Hb content and an increase of the Hb F content. Normally the Hb F content is positively correlated with the total Hb content, as illustrated in Fig [2](#F2){ref-type="fig"}. {#F2} In the 42 iron-deficient but not thalassemic subjects, the Epo level ranging from 2.7 to 1960 mU/1 with a mean value of 160.10 ± 312.60 mU/1, was negatively correlated (r = -0.74) to the Hb content (Fig. [1](#F1){ref-type="fig"}). The correlation is statistically significant (n = 42; p \< 0.001). ![Negative correlation between total Hb content and Epo level in iron-deficient patients. The logarithm of the Epo level plotted against the corresponding total Hb content of each patient is denoted by a blue square. The linear regression (solid line) fitting the data is described by the following equation: log \[Epo\] = 4.78--0.33 \[Hb\]. The correlation coefficient r is -0.74.](1471-2326-2-2-2){#F1} The mean Epo level of beta-thal int patients is apparently lower than that of iron-deficient patients (Tab. [1](#T1){ref-type="table"}). Regarding to total Hb, only between the groups of sub-silent beta-thal int patients and iron-deficient patients the difference is statistically significant (t = 2.17; gl 73; p \< 0.05). The mean values of the Epo level and the Hb F content in the two groups of beta-thal int seem to be positively related, but an accurate analysis of single cases indicates (Tab. [2](#T2){ref-type="table"}) that the Epo level and Hb F content are unrelated: subjects showing a low content of Hb F and an high level of Epo as well as subjects showing an high content of Hb F and a low level of Epo were frequently observed. This observation is reinforced by the analysis of groups of subjects in which the severity of anemia is similar. Within the group of evident beta tal int patients (the only adapt for this analysis) subjects showing an Hb F level higher or lower than 40% \[[@B5],[@B6]\] were selected. As illustrated in Fig. [2](#F2){ref-type="fig"} and Tab. [3](#T3){ref-type="table"}, subjects having a very similar Epo level and Hb content fall in both classes of Hb F content, indicating that Hb F content and Epo level are not related. ::: {#T2 .table-wrap} ::: {.caption} ###### Total Hb, Hb F, and Epo level in patients with evident beta thal int. ::: **Case** **Hb g/dl** **Hb F %** **EPO mU/l** ---------- ------------- ------------ -------------- 1 10.5 87 23 4 10.6 70 55 11 7.2 21 132 19 8.2 7 51 20 7.4 24 126 21 9.0 12 233 ::: ::: {#T3 .table-wrap} ::: {.caption} ###### Epo level and Hb F content in a group of evident beta thal int patients showing a similar degree of anemia and divided for Hb F \< or \> 40%. ::: **Patients\' group** **N. cases** **Hb F and EPO mean levels** ---------------------------------------- -------------- ---------------------------------------- ----------------------------------------- Evident beta thal int with Hb ≤ 9 g/dl 32 HbF = 13.6% EPO = 92.9 mU/l (25 cases) Hb F = 59.9% EPO = 118.8 mU/l (7 cases) ::: Different sets of mutations of the globin genes are associated with groups of patients characterized by a different severity of the disorder. The spectrum of the beta-globin gene mutations commonly found in the Italian population can be divided in three categories, the first of which includes the -101 C→T, -92 C→T, and IVS II 844 C→G single nucleotide substitutions; the second the -87 C→G, -86 C→G or C→A, and IVS I C→T single nucleotide substitutions, and the third the beta° 39 C →T, IVS I 110 G→A, IVS II 745 C→G single nucleotide substitutions and the frameshift 6 (-A) mutation. These three categories are referred to as silent beta-thal, mild beta-thal, and severe beta-thal mutations, respectively. In the group of 33 patients with sub-silent beta-thal int the frequencies of silent beta-thal, mild beta-thal, and severe beta-thal mutations were 37%, 13% and 50%, respectively (Tab. [4](#T4){ref-type="table"}). In the group of 42 patients with evident beta-thal int, the most common and equally frequent (44%) beta-globin gene defects were mild and severe beta-thal mutations. The frequency of silent beta-thal mutations was 12%. Conversely, severe beta-thal mutations were found at high frequency (83%) in thalassemia major patients. Silent beta-thal mutations were detected only in a small fraction (2.1%) of thalassemia major patients. In the 58 patients both heterozygous for a mutation of the beta-globin gene and heterozygous or rarely homozygous for the presence of three or four copies of one alpha-globin locus, the beta-globin gene defects were invariably beta° or severe beta^+^ thal mutations (Tab. [4](#T4){ref-type="table"} and ref. \[[@B13]\]); the alpha-globin gene mutation was in 54 out of 58 patients a triplicated alpha anti 3.7 I gene cluster; in 3 patients a quadruplicated alpha anti 4.2 gene cluster, and in one patient the homozygous condition for a triplicated alpha anti 3.7 I gene cluster. In these patients, the average Hb content was 10.0 ± 1.38 g/dl (Table [1](#T1){ref-type="table"}), the content of Hb F was slightly higher than normal, ranging from 2.0 to 19.3 with an mean value of 5.15 ± 3.9 %, and the average Epo level was 28.45 mU/1, ranging from 7.85 to 60.72 mU/l. This value is lower than that of regularly transfused thalassemia major patients. Although the Hb F content appears to be low in all these patients, their clinical picture is highly heterogeneous, 50% of the patients having a sub-silent beta thal int, 30% an evident beta thal int and 20% a severe anemia with 5 transfusion dependent patients. Ten out of 58 patients (17.2%) have undergone splenectomy. The patients with Hb H disease have a normal Hb F level, and a Epo level lower than that of patients with beta-thal int. In the beta-thalassemia patients and in the control subjects, screened for the six mutations of the promoter region of the gamma-globin gene observed in Italy, only the Ggamma -158 C→T and AgammaT single nucleotide substitutions were detected. In all cases except 7 the AgammaT mutation was to the Agamma -225--222 bp deletion. The frequencies of the Ggamma -158 C→T and AgammaT mutations are significantly different in the various groups of beta-thalassemia patients (Tab. [5](#T5){ref-type="table"}), with the lowest frequency of the Ggamma -158 C→T mutation (2.1%) in thalassemia major patients. ::: {#T5 .table-wrap} ::: {.caption} ###### Distribution of the Ggamma -158 C→T and AgammaT mutations in thalassemic patients and in control subjects. ::: **GROUP OF PATIENTS** **Allele frequencies** --------------------------------------------------------- ------------------------ ------ Sub-silent beta tal int 8.6 27.3 Evident beta tal int 18.3 26.6 beta tal int with genotype triplicated alpha + beta tal 16.6 17.6 Hb H disease 37.5 5.5 Tal major 2.1 35.4 Normal controls 25.0 18.0 ::: In the group of patients with evident beta-thal int the frequency of the Ggamma -158 C→T mutation was 6.2% in 7 patients having a total Hb content \> 9 g/dl and 1.6% in 32 patients having a total Hb content ≤ 9 g/dl. Discussion ========== It is well known that the severity of thalassemic diseases (thalassemia major and thal int) syndromes is primarily determined by the nature and type of mutations as well as gene dosage of the globin genes underlying these disorders. The results reported in this study extend this conclusion to two clinical variants of beta-thal int, referred to as sub-silent and evident beta-thal int. Distinct sets of mutations of the beta-globin gene have been found associated with these two variants of beta-thal int. Silent mutations of the beta-globin gene were detected at high frequency in patients affected by sub-silent beta-thal int, whereas mild and severe mutations of the beta-globin gene were the most common mutations amongst the patients affected by evident beta-thal int (Table [4](#T4){ref-type="table"}). This notion is reinforced by the high frequency (83%) of severe beta^+^ and beta° mutations in patients affected by thalassemia major. Several hematologic parameters of beta-thalassemia reflect the diversity of mutations at beta-globin locus. The total Hb content was found to be significantly different between patients affected by thalassemia major and patients with beta thal int, as well as between patients affected by sub-silent and patients with evident beta-thal int (Table [1](#T1){ref-type="table"}). These last two groups of patients showed also significant differences in the content of Hb F (t = 2.28; df = 72; p \< 0.05). The comparative analysis of the serum Epo level in groups of patients affected by thalassemia diseases or iron deficiency shows first of all that the synthesis of Epo, as predicted, is always directly related to the severity of anemia and, consequently, inversely related to the total Hb content. Furthermore, in the thalassemic diseases mean Epo level is consistently lower than that of iron deficient patients, probably as a result of the different pathophysiology of the anemia in the two conditions. No firm relationship could be established between the Epo and Hb F levels. The variations of these two parameters in each patient with evident beta thal int appear to be erratic (Tab. [2](#T2){ref-type="table"}); and in patients having a similar level of both total Hb and Epo, the amount of Hb F is indifferently higher or lower than 40% (Tab. [3](#T3){ref-type="table"}). Therefore, the data do not support the hypothesis that Hb F content is regulated by the serum level ofEpo. The relationship between total Hb content, Hb F content and Epo level were not studied in patients with Hb H disease, since Hb F is in these patients below the detection limit. In patients with a compound heterozygosity for a triplicated copy of a alpha globin gene cluster and a beta globin gene mutation, only a moderate increase of the Hb F level (mean 5.5 ± 3.9% within a range from 2.0 to 19.3%) was observed. This observation indicates that the excess of alpha globin chains, as assessed by the unbalanced alpha/non-alpha globin ratio (2.5) is not sufficient to increase the gamma-globin chains synthesis. Only the Ggamma -158 C→T and AgammaT single nucleotide substitutions out of 6 gamma gene promoter region mutations explored in this study were detected, and their frequencies varied in different groups of patients (Table [5](#T5){ref-type="table"}). The role of this mutation in determining the Hb F amount is not yet definitively cleared. The demonstration, already reported by other Authors \[[@B14]\], that the frequency of Ggamma -158 mutation is very low in beta thal major patients (Tab. [5](#T5){ref-type="table"}), and, inside the beta thal int patients, higher in those with a higher amount of total Hb and lower in those with a lower amount of total Hb, seems to indicate that this mutation is important in ameliorating the severity of the disease. Conversely, other obstervations \[[@B15]-[@B17]\] report that this mutation does not influence the Hb F content of subjects which are heterozygous for a beta thal mutation \[[@B16]\] as well as of heterozygous compounds for a beta thal mutation and a triplicated or quadruplicated copy of the alpha-globin cluster. At variance with these observations, in sickle cell disease patients carrying some beta^s^ haplotypes which show a high frequency (up to 100%) of the Ggamma -158 C→T mutation, show also an increase of the Hb F content that leads to a reduction of the severity of the disease \[[@B18]\]. The relationship between the content of Hb F and point mutations of the promoter region of the gamma-globin genes in thalassemic patients needs then to be further investigated. The analysis must be extended to those mutations of the gamma-globin gene promoter region that have not yet been detected in the population of Italy or other Mediterranean countries. The hypothesis \[[@B19]\] that the Ggamma -158 C→T mutation when linked in cis to the sequence motif (AT)~x~(T)~y~ mapping at position -530 in the 5\' untranslated region of the beta-globin gene could interact to generate a high expression of Hb F under the erythropoietic stress, has to be ulteriorly evaluated. Finally, the role played by loci unlinked to the beta-globin locus in the regulation of Hb F synthesis remains to be elucidated. Two candidate regulatory loci have been mapped at Xp22 \[FCP locus \[[@B20],[@B21]\]\] and 6q22.3-23.2 \[[@B22]\]. A study regarding all these factors, obviously necessary also for the various heterozygotic thalassemic conditions, is now in progress in our laboratories. Conclusions =========== The severity of beta-thalassemia syndromes including the increased synthesis of Hb F is primarily determined by the genotype of the beta-globin locus. The synthesis of Hb F does not appear to be regulated by the serum level of Epo. The Ggamma -158 C→T mutation is frequent among patients with beta-thal int, and very rare in patients with thalassemia major, but at variance with sickle cell disease, increased synthesis of Hb F in beta-thal int patients cannot be firmly related to the linkage in cis of this mutations. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ The Authors are grateful to Dr. Giacomo D\'Agostaro (ENEA, Rome) for his cooperation and helpful comments in preparing the manuscript; Dr. Paolo Cianciulli Director of the Day Hospital for Thalassemics of the Sant\'Eugenio Hospital of Rome, to have made possible the study of his thalassemic patients; to the Mrs. Brunella Sidorini and Mrs. Nadia D\'Arcangeli for their technical assistance; to Mr. Marco Ballante for his excellent contribution to the making of the manuscript.

Background ========== On the 25 January, 1999 an earthquake with an intensity of 6.2 points on the Richter\'s scale occurred in a focus located only 20 kilometers away from Armenia, Colombia. Although 28 towns or municipalities, including Armenia, were severely damaged by the earthquake, the worst effects occurred in Armenia itself and in its surrounding towns. The consequences of the catastrophe were devastating, mainly because of the loss of people\'s property and belongings: around 70% of Armenian houses were destroyed and in some neighboring towns the percentage was greater than 80%. In the rural area, the proportion of lost houses was also 80%. The death toll was estimated at 1184, and the number of people injured was more than 5,000 \[[@B1]\]. The region known as \"Departmento del Quindío\" includes, according to the political and administrative division of the Colombian territory, several small towns and its capital city, Armenia. The whole area of influence of the telluric movement is known as \"Eje Cafetero\" (Coffee Axis), because it is here that most of the world famous Colombian coffee is produced. This area is located within the country\'s central mountain range. Before the tragedy, Armenia had 296,330 inhabitants, with the whole Department of Quindío having approximately 500,000. The devastating effects on the houses of Armenia induced the organization of transitory housing facilities in tents and \"alojamientos\" (multi-family spaces), which caused difficulties in the provision of basic sanitary services, water supplies and food. Fears about epidemics were high, but no outbreaks of infectious diseases, including diarrhea, dengue or malaria occurred \[[@B2]\]. Temporary housing facilities remained until two years after the disaster, which raised concerns about public health conditions in this people. The University of Quindío undertook epidemiological studies in order to determine which health disorders were occurring. This included the effect on mental and physical health morbidity \[[@B2],[@B3]\]. One of the pathological aspects that was studied was the prevalence of intestinal parasites in this people. During this survey it was found that giardiasis was the most prevalent intestinal parasitic infection. Therefore, a study was designed in order to ascertain the epidemiological factors related to this high prevalence. In this paper we report the findings of the laboratory examination of stools and their association with data collected by questionnaire. Methods ======= Study design, characteristics of the camps and questionnaire ------------------------------------------------------------ A prospective study of the prevalence of *Giardia* in stool specimens of children living in temporary houses in Armenia was carried out. A total of 26,780 people lived in temporary housing camps built under the supervision of non-governmental organizations. This study was conducted in 18 camps (\"asentamientos\"), multi-family spaces which grouped temporary houses or tents, and were representative of a total of 86 camps which were established in Armenia after the January 25, 1999 earthquake. A census population of the families living in the 18 camps found 816 children aged between 3 and 13 years. A random sample of 193 targeted households was selected for a systematic household survey conducted between January 2000 and July 2001. Thus, 217 children aged between 3 and 13 years were sampled. The children were included only if their father or other adult with legal custody of the children agreed to give written informed consent. A household was defined as a unit of persons residing in one tent or temporary house. The median household size was made up of 5 persons. All households had municipal garbage disposal services. Electricity was available for all households. From each selected household one adult was interviewed using a standardized questionnaire which focused on both the demographic characteristics of children such as age and school years, as well as on the presence of clinical symptoms such as diarrhea (passing more than two liquid stools daily) or flu symptoms (rhinorrea without pharyngitis and with only minor or no fever at all) with a duration of less than eight days at the moment of interview. The weight of each child was recorded. Interviewers were selected from university students of biology who had previous experience in community work, were from the same city and knew the terms used by people to describe their typical symptoms and living conditions; they were previously trained by an epidemiologist. Stool samples of children were collected the day after the questionnaire was filled. The source of drinkable water and the type of toilet used were recorded for each child. Water supply varied between camps: in some camps water was provided by the municipal water ducts, in others there were individual tanks for each house. Municipal water (piped water) was the major source of drinking water and was used in 15 camps. The source of municipal water was surface water (\"Río Quindio\") provided through the aqueduct system which existed before the earthquake. Water chlorination concentrations were between 2,5 and 3 mg/liter. This is a closed water delivery system; however after the earthquake some of the ducts collapsed and leakages occurred. For this reason, in 3 camps individual tanks were used as a source of drinkable water. The tanks were made of plastic material and had a 2,000 liter storage capacity. Individual tanks were filled by a tank car. In addition, municipal health services distributed 1,217 kg of sodium isocianyde dichloride in all camps to be used in individual tanks. There were two types of hygienic room services. One was the communal toilet service that consisted of one latrine for each 9--10 families placed in the center of camp; other type of toilet was the individual sanitary service located in each temporary house and used only by the members of that household. The researcher also registered the \"clean status\" of the temporary house; a house was considered \"clean\" if no particular odors were detected, the food was stored in proper recipients, and food waste was disposed of in garbage cans. The place for food storage was also recorded. No questions about animal contact were done. Specimen collection and stool examination ----------------------------------------- Stool samples were collected early in the morning in sterile plastic containers supplied by the researchers and conserved on ice until laboratory examination. Samples with water or urine contamination were discarded. A second sample was requested from all children at day 5 after the first sample and a third sample one month later. Stool samples were stained with ferric hematoxilin and examined after centrifugation; this test has a sensitivity of 61% on the first sample and a cumulated sensitivity of 83% when three different serial samples are examined \[[@B4]\]. Statistical Analysis -------------------- Statistical significance tests for comparison of rates were performed when sample sizes were insufficient and asymptotic tests when sizes were sufficient. All tests were two-sided, and a p value of \<0.05 was considered as significant. Exact 95% confidence intervals (Cl) were calculated \[[@B5]\]. Results ======= In total, stools from 217 children were examined, 97 were girls (44.7%) and 120 boys (55.3%). In the first sample, prevalence of *Giardia* cysts was 60.4% (131 children) and *Giardia* trophozoite forms 4.6%. A second sample was obtained in only 53 of 217 children. Families were not informed about the result of the first sample before collecting the second one. Collection was requested on a voluntary basis. From these 53 children with a second sample, 40 were positive and 13 negative on the first sample. Thirty children were positive for *Giardia* cysts on the second sample (56%), only three of them being negative at the first sample. Only six children provided a third sample; three of them, who were all previously positive for *Giardia* cysts, were positive. Analysis of infections by other parasites was also performed (Table [1](#T1){ref-type="table"}). There were no statistically significant differences in the prevalence of giardiasis in children infected with other parasites. No cultures for bacteria were performed during this study. ::: {#T1 .table-wrap} ::: {.caption} ###### Parasitic infections in stool samples of children (n= 217) living in temporary housing after the 1999 earthquake disaster in Armenia. ::: **Organism** **Number of infected children** **Percent of sampled children** **Number of infected children also infected with giardiasis** ----------------------------------------------------------- --------------------------------- --------------------------------- --------------------------------------------------------------- *Ascaris lumbricoides* 36 16.6% 19 *Balantidium coli* 4 1.8% 3 *Endamoeba coli* 71 32.7% 40 *Endolimax nana* 125 57.6% 56 *lodamoeba buschlii* 31 14.3% 19 *Entamoeba histolytica/dispar* (determined by micrometry) 116 53.4% 75 ::: The analysis of epidemiological characteristics (Table [2](#T2){ref-type="table"}) showed that the presence of *Giardia* cysts was associated with the use of communal toilet versus use of individual toilet (OR 3.9). The water provision by municipal water services was associated with a higher risk for giardiasis, in contrast with people that used individual water tanks (OR 3.5). The finding of mucus in stool was higher in children with *Giardia* cysts (OR 2.3). A \"clean house\" was a protective factor for the presence of *Giardia* cysts (OR 0.4). ::: {#T2 .table-wrap} ::: {.caption} ###### Association between epidemiological and laboratory findings with the presence *Giardia* cysts in stool samples from children living in temporary housing after the 1999 earthquake disaster in Armenia. ::: **Risk factor** **With factor** **Without factor** **OR** **95% Confidence Intervals** **P** ------------------------------------------------------ ----------------- -------------------- -------- ------------------------------ ----------- Communal toilet (versus individual toilet) 21/25 110/192 3.9 1.2--16 **0.01** Municipal water service versus individual water tank 27/58 104/159 3.5 1.1--14 **0.02** Mucus in fecal sample 20/26 111/191 2.39 0.9--6.7 **0.049** Leukocytes in stool 16/22 115/195 1.85 0.70--5.35 0.15 Plastic stock of food versus carton stock of food 49/93 82/124 1.7 0.9--3.1 0.06 Yeast in stool 57/93 74/124 1.06 0.61--1.82 0.46 Diarrhea 23/42 108/175 0.75 0.4--1.4 0.84 Flu symptoms 79/136 52/81 0.7 0.4--1.4 0.45 Lipids in stool 120/201 11/16 0.67 0.18--2.2 0.65 Clean lodgment 62/118 69/99 0.48 0.27--0.84 **0.014** ::: In total 199 of the 217 children studied attended school. The source of water at school was municipal water and communal toilet was the type of sanitary service there. As data were collected concerning the putative factor (water source) and/or disease (*Giardia*), data were also analyzed on the presence or absence of the confounding variable (attendance at school). The data were then divided into strata for confounding variable (attendance at school) in a series of 2 × 2 tables, one for each level of the confounding variable (Figure [1](#F1){ref-type="fig"}). Using Mantel-Haenszel odds ratio method the OR was 15.1 (CI95% 8--36) with a p value of 0.00000000. This analysis indicate that drinking municipal water at camps was strongly associated with giardiasis even with the confounding variable \"school\" taken into account. {#F1} Presence of *Giardia* trophozoites was higher in patients with municipal water service and with lipids in stools (Table [3](#T3){ref-type="table"}); however due to the small number of children with trophozoites, no difference attained statistical significance. Other demographic factors including weight, age and the number of years at school were not associated with giardiasis (Table [4](#T4){ref-type="table"}). ::: {#T3 .table-wrap} ::: {.caption} ###### Association between epidemiological factors, laboratory findings and the presence of *Giardia* trophozoites in stool samples from children living in temporary housing after the 1999 earthquake disaster in Armenia. ::: **Risk factor** **With factor** **Without factor** **OR** **95% Confidence Intervals** **P** ------------------------------------------------------ ----------------- -------------------- -------- ------------------------------ ------- Municipal water service versus individual water tank 10/159 0/58 ? ?-1.1 0.06 Lipid drops in stools 2/16 8/201 3.41 0.45--16.5 0.16 Flu symptoms 6/81 4/136 2.62 0.69--10.8 0.11 Plastic stock of food versus carton stock of food 1/19 9/198 1.16 0.050--7.71 0.97 Leukocytes 1/22 9/195 0.98 0.04--6.44 0.66 Yeast in stools 4/93 6/124 0.88 0.21--3.0 0.69 Communal toilet (versus individual toilet) 1/25 9/192 0.84 0.03--5.50 0.71 Mucus in fecal sample 1/26 9/191 0.80 0.03--5.24 0.72 Diarrhea 1/42 9/175 0.45 0.01--2.86 0.88 Clean lodgment 3/118 7/99 0.34 0.07--1.34 0.97 ::: ::: {#T4 .table-wrap} ::: {.caption} ###### Mean age, school years and weight and presence of cysts of *Giardia* in stool samples from children living in temporary housing after the 1999 earthquake disaster in Armenia (Colombia). ::: ***Giardia* Cysts** **Mean ± SD** **P** -------------- --------------------- --------------- ------- Mean age No 7.3 ± 2.9 0.79 Yes 7.4 ± 3.1 Weight (Kg) No 23.1 ± 9.1 0.89 Yes 23.3 ± 9.9 School years No 1.6 ± 1.9 0.87 Yes 1.7 ± 2.2 ::: Discussion ========== Two previous studies carried out by the Colombian National Institute of Health found that *Giardia* was the most prevalent intestinal parasite. In 1965 a prevalence of 9.4% was found and in the 1980 survey the rate was 21.4%. These rates were the reported for children in the 5 to 14 year-age group living in the central region of Colombia, where the department of Quindío is located. In both studies (in 1965 and in 1980) diagnostic methods were the same as those used in the present study, microscopic examination and the rates were estimated from the examination of only one sample \[[@B6]\]. We found a higher prevalence. Our findings could indicate that giardiasis emerges during events which alter the existing sanitary conditions. We found that a higher prevalence was associated mainly with communal toilets and municipal duct provision at camp, which indicates that both factors favored dissemination of giardiasis in temporary camps after the disaster. As demonstrated by stratified analysis, drinking municipal water at camps was strongly associated with giardiasis even when attendance to school was taken in account as confounding variable. In future in similar conditions it will be important to make efforts addressed to guarantee individual toilet for each household and to control the water source by checking the presence of *Giardia* in water at temporary camps. In the present study, clinical symptoms were not associated with the presence of *Giardia* cysts or trophozoites in stools, but presence of cysts were significantly associated with mucus in stool. Mucus in stool can indicate inflammatory responses and our finding could indicate that chronic infection may induce inflammatory changes without major clinical symptoms. However, a longitudinal prospective study comparing infected and non-infected children with *Giardia* would be necessary in order to establish if chronic *Giardia* infection has an impact on growth or nutritional status in this particular population. In humans, the clinical effects of *Giardia* infection range from the asymptomatic carrier state to a severe malabsorption syndrome \[[@B7]\]. Factors possibly contributing to the variation in clinical manifestations include the virulence of the *Giardia* strain, the number of cysts ingested, the age of the host, and the state of the host immune system at the time of infection. Giardiasis was most prevalent in children who drank municipal water. *Giardia* survives in chlorinated water \[[@B8]\]. To protect against transmission all drinking water should receive chemical pretreatment, preferably with sedimentation, and filtration in addition to disinfection \[[@B9]\]. Municipal water should supply water with a concentration of less than 0.7--70 cysts per 100 liters \[[@B10]\]. There are two previous reports of outbreaks of giardiasis after natural disasters, both linked to flooding after heavy water run-off. One occurred in Montana (United States) where an outbreak of gastrointestinal illness affected 780 persons \[[@B11]\]. In Utah (United States) 1,230 people were affected with diarrhea \[[@B12]\]. Water systems providing unfiltered surface water and contaminated with flooding were identified as the origin of both epidemics. In Armenia the earthquake damaged the municipal system of water ducts \[[@B2]\]; this probably induced their contamination, explaining the high rate of giardiasis that we found in children. It is interesting to note that individual tanks were filled with municipal water. One explanation for the lower prevalence of giardiasis in children that used individual tanks is that in tanks there is additional sedimentation that can reduce the contamination of water with *Giardia*\[[@B10]\]. Health municipal services also distributed chlorine to be used in individual tanks. Although we could not establish the extent to which this chlorine was used, it is possible that additional chlorination contributed to reduce the probability of *Giardia* infection. There are no data available on the prevalence of giardiasis in general population in Armenia. It will be necessary to begin studies in order to establish the public health impact of this parasite infection in the city. Conclusions =========== Giardiasis was the most prevalent intestinal parasite found in children living in post-earthquake camps in Armenia, Colombia. Our results suggest that this is an emerging disease in post-disaster situations. However, data on prevalence of giardiasis before the earthquake and in population living away from camps is lacking. The use of communal toilets instead of individual household toilets was found to be an epidemiological risk factor associated to giardiasis. Adequate prevention measures which include efforts to make individual toilets available are also important to prevent dissemination of this parasite. The present study also found a relationship with the water provided by municipal closed ducts that were affected by the earthquake. This highlights the need for implementing adequate surveillance systems for the presence of *Giardia* in water sources in similar circumstances. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We thank Dr. Santiago Nichols from the \"Instituto Nacional de Salud\" at Bogota, who help us checking english spelling of this paper. This work was financed by research funds of the University of Quindio, a public Colombian University.

Background ========== One of the strategic objectives of the WHO is the eradication of poliomyelitis in all regions of the world within the next years. According to the WHO plans the transmission of wild poliovirus has to be interrupted worldwide by the end of 2002. The global certification will be held in 2005. Although the European region is actually polio -- free, it remains subject to the risk of importation from endemic regions. The most recent cases of poliomyelitis due to an importation of wild polio type 1 occurred in unvaccinated Bulgarian children \[[@B1]\]. To prevent such importations it is necessary to keep a high level of population immunity. The last indigenous case of poliomyelitis was diagnosed in Germany in 1990. The last imported wild viruses were detected in patients with travel history to India and Egypt in 1992. Since that period of time there were found only 0 -- 3 cases of vaccine-associated paralytic poliomyelitis (VAPP) per year. In order to prevent these VAPP -- cases the vaccination strategy has been changed in Germany. Since 1998 it is recommended to use only the inactivated vaccine (IPV). The neutralizing antibody test is the method of choice in conducting serological surveys to identify epidemiologically important immunity gaps in the population. It is considered essential to provide the precise procedure for carrying out polio neutralizing antibody tests. Use of standard reagents, the International Standard Serum (ISS), standard Sabin strains, and expression of the results in International Units (IU) is a feature aimed at obtaining consistent results. This will assure that comparison between studies can be made with confidence. An International Standard Serum does exist, but WHO recommendations do not indicate a protective level in IU \[[@B2],[@B3]\]. Materials and methods ===================== Serum samples ------------- The first German Health Survey 1997/98 was a representative study of the health status of the population in unified Germany. In this project which has been carried out by the Robert Koch Institute on behalf of the Ministry of Health about 7200 study participants aged between 18 and 79 years were going through a medical check-up and interviewed as to health -- relevant issues. Germans as well as foreigners, who lived in Germany, were included in the survey. As a result of various selection criteria (age, sex, and size of town) a representative collection of study participants was obtained \[[@B4]\]. A representative panel of 2564 serum samples from the total amount of sera from the Health Survey was investigated for the presence of antibodies to all three poliovirus serotypes. Additionally, 881 sera from children less than 18 years obtained from six large laboratories in Germany were included in this study. These laboratories performed wide-range diagnostics without any visible bias within the group of patients. The sera were collected in 1996/97 and stored at -20°C until testing. Neutralization test ------------------- The microneutralization test using 100 ID~50~ of challenge virus (Sabin strains) was performed according to the WHO-guidelines \[[@B5]\]\[[@B6]\]. An In-House Reference serum (IHR) of known neutralizing activity was included in each test to control reproducibility of results. Each test serum was investigated in duplicate. If the final serum dilutions of these duplicates differed more than one stage the test was rejected. The International Standard Serum (ISS, a preparation of pooled human sera) containing 25 IU, 50 IU, and 5 IU for Poliovirus 1, 2 and 3, respectively has been used to calibrate the potency of the IHR. ISS and IHR were titrated in parallel on ten separate occasions using eight replicates per serum dilution (starting from serum dilution of 1/4). Calculation of the antibody concentration in International Units (IU/ml) ------------------------------------------------------------------------ The calculation of the concentration with  resulted in C~IHR1~ = 3.125 IU/ml, C~IHR2~ = 6.25 IU/ml, C~IHR3~ = 0,625 IU/ml for Polio 1, 2, 3, respectively for the used IHR (example for Polio 1: a serum dilution Val~ISS~ = 1/256, and Val~IHR~ = 1/32 would result in C~IHR~ = 3.125 IU/ml). The calculation of the antibody concentration in a given test serum sample was done analogously.  In contrast to WHO -- recommendations not the geometric mean (GMT) of all measured IHR\'s was used but the concentration of the IHR in the given test only. Thus each test was calibrated in respect to the appropriate IHR. A given test was only accepted if the final serum dilution did not exceed a one-stage range of the GMT of all tested and accepted IHR\'s. For an accepted test the following condition holds:  The threshold values Co were calculated as follows:  with  *SDev(C)* = StandardDeviation{C, with serumdil. \< 1/4}. Results ======= In the past the cut-off value for \"sufficiently protected\" against poliomyelitis was set if the final serum dilution giving protection against 100 ID~50~ of challenge virus was greater or equal to 1/4 (titer 8). This condition should be substituted by expressing results in IU/ml. The cut-off values (Co) for the protective level were calculated as follows: for Polio 1: C~0~ = 0.080 IU/ml for Polio 2: C~0~ = 0.180 IU/ml for Polio 3: C~0~ = 0.075 IU/ml Neutralizing antibodies against poliovirus type 1, 2 and 3 were detected in 96.2% (95% confidence interval CI: 95.5%-96.9%), 96.8% (CI: 96.1%-97.5%) and 89.6% (CI: 88.4%-90.8%) of samples, respectively (Fig. [1](#F1){ref-type="fig"}). 85% of all tested persons had antibodies against all three virus types. Only 0.3% of tested persons had no detectable antibodies against all polioviruses. {#F1} Statistically significant regional differences in the immunity level between the adult populations in the East and West of Germany have been detected for Polio 3 (86,6% / 90,9%; p = 0,001). Within the given data set the decision \"sufficiently protected\" slightly changes according to Tab. [1](#T1){ref-type="table"} (expressing results in IU vs. serumdilution). Thus for Polio 3 about 3% (82/2564) of the samples were additionally classified as \"not protected sufficiently \" against disease. ::: {#T1 .table-wrap} ::: {.caption} ###### Differences in numbers of samples classified as \"sufficiently protected\" (serumdilution vs. concentration C in IU/ml), n = 2564 ::: **Polio 1** C\<C~0~ C\>=C~0~ ------------------ --------- ---------- Serum dil.\< 1/4 90 11 Serum dil.\>=1/4 12 2451 **Polio 2** Serum dil.\< 1/4 97 17 Serum dil.\>=1/4 1 2449 **Polio 3** Serum dil.\< 1/4 182 0 Serum dil.\>=1/4 82 2300 ::: Discussion / Conclusions ======================== The goal of eradication of poliomyelitis has been complicated by some events in 2000--2001. First, the discovery of circulation of a vaccine virus strain that mutated to regain its disease-causing ability in Hispaniola. This outbreak in the Dominican Republic and Haiti was the first in the Americas since 1991 and occurred in areas of very low vaccination coverage \[[@B7]\]. Second, the polio cases in Bulgaria. There had been no reported cases of polio in Europa since November 1998. In spring 2001 two gypsy unvaccinated children were paralyzed by a wild poliovirus in Bulgaria, which was imported from India \[[@B1]\]. These outbreaks send a warning to all countries to maintain their guard against polio, despite the rapid decline in its incidence. Sufficient surveillance systems and high levels of population immunity can prevent polio outbreaks. The level of immunity to poliomyelitis has been checked in Germany at regular intervals. For the first time in Germany the results were expressed in International Units. A representative serum panel from the German Health Survey was investigated. Additionally, sera from children less than 18 years old were included in this study. These sera were not representative for all selection criteria because they originated from diagnostic laboratories. A representative study of the health status of persons under an age of 18 years is prepared by the Robert Koch Institute. The pilot phase has started in 2001. Within the next years valid data will be obtained for the younger population of Germany. The given threshold values of 0.075 lU/ml, 0.180 IU/ml, and 0.080 IU/ml were calculated for Polio 1,2,3 respectively. The calculation was based on the assumption that the new criteria (threshold in IU/ml) should give comparable results with respect to the old one (threshold of serum dilution). Overall, neutralizing antibodies against poliovirus type 1, 2 and 3 were detected in 96.2%, 96.8% and 89.6% of samples, respectively. This seroprevalence indicates a very high level of immunity of the German population. Although the prevalence of antibodies against poliovirus was generally somewhat lower for type 3 than for types 1 and 2, it was still close to 90%. Data presented in this study are concordant to those obtained in other countries \[[@B8],[@B9]\]. A previous seroepidemiological study in Germany as well as this one revealed substantial regional variations of immunity levels to Polio 3 \[[@B10]\]. Compared to the results of this study other German studies have shown remarkable lower antibody prevalence to Polio 3 among blood donors in Berlin and in hospitalized patients \[[@B11],[@B12]\]. Limitations in some of these serological studies due to non-representative samples or methodological reasons resulted in different conclusions regarding immunity status. Standardization of the neutralization test as well as expressing results of serological studies in International Units will permit better comparability of polio immune status of populations in different studies, and in different countries. Immunity to poliomyelitis is largely dependent on humoral neutralizing antibodies, both after natural infection and after vaccination. It is unclear at present whether all those persons with a low level or no detectable antibodies are susceptible to infection; some, particularly the elderly, may be protected by memory immunity, an accelerated antibody response because the immune system has been primed previously. It has also been shown that enterovirus infections induce T-cell immunity \[[@B13]\]. The production of local secretory IgA antibodies in the gut mucosa may play a major role in protection \[[@B14],[@B15]\]. Although IPV offers excellent protection against disease, it is less effective in preventing poliovirus infection because of limited mucosal immunity. IPV recipients may contribute to the circulation of outbreak virus. Extensive circulation of Polio 3 was observed during 1984 outbreak in Finland. A study from the Netherlands has also shown that poliovirus circulation occurred during the early phase of the 1992--93 poliomyelitis outbreak. Most infected children had not been vaccinated \[[@B16]\]. A high level of immunity of the German population, including immigrants, must be monitored after the switch from attenuated to inactivated polio vaccine (OPV to IPV). Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ The authors are grateful to Helga Schickhoff for her excellent technical assistance.

Background ========== The most common localized tumor of the head and neck is the hemangioma, and more than 50% of all congenital hemangiomas are located on the head and neck \[[@B1]\]. The hemangioma may present in various forms, for example capillary, cavernous, or mixed, the most common being a capillary hemangioma. Most authors dispute whether this classification can be based on clinical appearance only. In fact, most of these tumors may actually have mixed capillary and cavernous hemangiomas, and are better known as Capillaro-Venous malformations. These capillaro-venous malformations may affect not only the cutaneous surfaces of head and neck but also the mucosal surfaces. Because these capillaro-venous malformations in the upper aero-digestive tract may cause difficulty in swallowing and/or breathing, and may even threaten the life if the patient aspirates while bleeding, they are best treated surgically \[[@B2]-[@B4]\]. As these malformations have a marked propensity for bleeding during excision, the value of preoperative embolization is well established. The objective of embolization is to devascularise these malformations by catheterization and embolization of each vessel supplying the tumor, which is not always achieved because most of them are fed, by many small vessels that may be too small for the catheter to pass. We present two cases where we have used a devascularization technique for these malformations by direct injection of *n-butyl cyanoacrylate* (NBCA) into the capillaro-venous bed. Case Presentation ================= Case 1 ------ A 24 years male had presented with slowly increasing bluish swelling over palate since birth that had occasionally bled. He had a capillaro-venous malformation over oral surface of the soft palate; the nasopharyngeal surface of the soft palate was not involved. The capillaro-venous malformation was punctured with a 22-guage needle and progression of the needle was monitored by 2-D fluoroscopy. After verifying the correct location of the needle by reflux of blood at the hub, the contrast agent was injected into the capillaro-venous malformation, revealing the local parenchymography draining to the regional veins without extravasation. NBCA and lipiodal then were mixed in equal proportions and injected slowly under fluoroscopy control with a 2 ml syringe to completely occupy the entire capillary bed (fig 1). Subsequently the capillaro-venous malformation was totally resected in the same sitting. A good plane of cleavage was obtained because of retraction of the malformation induced by devascularization. Peroperative blood loss was negligible (\<10 ml). The resected specimen was subjected to histopathological examination which confirmed the nature of capillaro-venous malformation and demonstrated that the embolization agent occupied almost whole of the capillaro-venous bed. Case 2 ------ A 55 years old male presented with repeated episodes of bleeding from mouth especially after taking meals. On detailed examination left tonsillar hemangioma (capillaro-venous malformation) was diagnosed to be the cause of this bleed. This hemangioma was also devascularised with direct puncture technique under general anesthesia (fig 2) as described above and subsequently left tonsillectomy along with excision of the hemangioma was done. Histological confirmation of the capillaro-venous malformation was done. Peroperative blood loss was negligible (\<10 ml) and a good plane of cleavage was obtained. Discussion ========== Preoperatively devascularization of the capillaro-venous malformation to decrease peroperative blood loss is desired by the surgeons. The conventional intravascular embolization technique has following limitations- (1) the devascularization is rarely complete as it is often not possible to embolize each and every feeding vessel, as some of the feeding vessels may be too small to negotiate the catheter, (2) a very experienced interventional radiologist team is required to carry out embolization, (3) time consuming, (4) expensive, (5) difficult to carry out surgical excision of the malformation in the same sitting. Whereas, the direct puncture technique has following advantages- (1) total devascularization is easily achieved, (2) does not require a team of intervention radiologists and can be carried out by the surgeon, (3) can be done immediately before surgery in the same sitting, (4) less time consuming, (5) relatively inexpensive. Although direct puncture technique is safe but sometimes anaphylactic shock \[[@B5]\] can occur. The embolizing agent getting transported into draining vessels causing complications e.g. pulmonary embolization \[[@B1]\] is reported. This technique can be safely carried out under fluoroscopic control by first doing local parenchymography and then keeping a close vigil on the slow progression of the embolization agent, suspending the injection when embolization agent penetrates into the draining vessels. This technique should be carried out under general anesthesia as direct injection of the embolization material into the capillaro-venous malformation in a conscious patient is extremely painful and also immediate surgery cannot be done. We chose NBCA as the embolising agent because it solidifies immediately on coming in contact with moisture, has antimicrobial properties \[[@B6]\] and has minimal tissue toxicity. NBCA is mixed with Lipiodal to make it radio-opaque. Other embolising agents may be used, for example ethibloc, ethanolamine oleate, sodium tetradecyl sulfate, polidocanol but they have various demerits in their use like Ethanolamine oleate has high viscosity that makes injection difficult, it also has a tendency of renal failure at high doses, and allergic reactions are also known. The main disadvantage of Ethibloc as an embolising agent is its long solidification time (3--5 minutes) making it susceptible to flow out into systemic circulation \[[@B7]\]. Sodium tetradecyl sulfate has a significant incidence of epidermal necrosis, causes hyper-pigmentation, and occasional anaphylaxis is known. Telangiectatic matting is seen following injection of Polidocanol, it also causes hyper pigmentation and anaphylaxis. Conclusion ========== We believe that this technique is relatively simple, less time consuming, cheap and permits surgical excision of the capillaro-venous malformation with lesser extent of bleeding. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== T wave alternans (TA) is a repolarisation phenomenon manifesting as a microvolt beat to beat change in amplitude of the T wave and ST segment. The mechanism responsible for TA is a dynamic (beat to beat) variation of the repolarisation sequence \[[@B1]\] due to either an alternation of action potential duration, frequently associated with the long QT syndrome \[[@B2]\] or an alternation of the time course of repolarisation \[[@B3]\] frequently associated with ischaemia \[[@B4]\]. The alternation of time course and multilevel distribution of electrical repolarisation activity during ischaemia \[[@B4]\] cause the electrocardiographic manifestation of TA in addition to alternating dispersion through islands of partially and fully refractory tissue which promote wavefront fractionation and vulnerability to reentrant malignant arrhythmias (MA)\[[@B5],[@B6]\]. The earliest description of TA was made in 1910 by Lewis \[[@B7]\]. Animal work by Hellerstein and Liebow showed a mechanistic link between TA and arrhythmogenesis \[[@B8]\] and recently the presence of TA has been recognised as being a harbinger of sudden death \[[@B9],[@B10]\]. Rosenbaum et al reported TA to be an independent prognosticator of vulnerability to arrhythmias predicted by electrophysiology testing\[[@B11]\] and these findings were repeated by Armoundas and co-workers showing TA to be superior to signal average electrocardiography as a predictor of malignant ventricular arrhythmias in high risk populations \[[@B12]\]. TA has demonstrated high sensitivities and specificities as a predictor of arrhythmic risk in unselected myocardial infarction populations \[[@B13]\] and idiopathic cardiomyopathy \[[@B14]\] but TA has not been used to differentiate risk in the ischaemic cardiomyopathy group. The purpose of this pilot study was to use indices of absolute micro voltage TA and the frequency content of alternating T waves to assess the ability to differentiate post infarction heart failure patients with significant substrates for developing re-entrant MA but without arrhythmic events, from those post infarction patients at most risk from developing MA. Materials and Methods ===================== Approval was obtained from the local Ethics Committee prior to performing the described research. The research was conducted in accordance the Helsinki Declaration, written informed consent was obtained from each patient prior to collecting data. The study population consisted of two groups of stable treated patients: Group 1 comprised 7 patients with extensive scarring secondary to myocardial infarction and compensated treated left sided heart failure with mean ejection fraction of 35%. The patient with the most recent infarct event in this group occurred over 20 months prior to the study. No patient within group 1 sustained a symptomatic arrhythmic event. Group 2 comprised 9 patients with a history of myocardial infarction with a mean ejection fraction of 38%, a distant episode of sustained ventricular tachycardia or aborted sudden death, positive induction of sustained ventricular tachycardia during an electrical stimulation study and a history of prophylactic implantation of an automatic defibrillator. 20 sequential electrocardiographic complexes were recorded from X,Y and Z orthogonal leads on each resting supine patient during a held half expiration breath without straining. The ECG recordings were made using and analog-to-digital converter (Powerlab), sampling at 1000 Hz. All the recorded signals were not filtered, because it was intended to make inter-group statistical comparisons in the frequency domain between the higher frequency band of 80--500 Hertz, where there may exist representations of possible late potential repolarisation. As all significant non-noisy signals have been found to occur within a 0 to 40 Hertz pass-band \[[@B15]\], the recorded signals underwent 40 Hertz low pass filtering prior to analysis in the time domain. For each lead on each patient, sequences of the alternate ST-T waves were ensemble averaged over a 400 millisecond duration using the positive or negative peak of the T wave as the fiducial central point of data collection over 400 milliseconds, see figure [1](#F1){ref-type="fig"}. In leads where the T wave was of low magnitude, the fiducial central point was taken to occur as the point at an average of the times from the peak QRS deflection to the fiducial T wave point in the other clearly demarcated orthogonal leads. The vectors of the squared magnitude of differences in ensemble averaged time aligned alternating ST-T sequences were calculated on each lead from each patient. For each patient, the resultant absolute difference vector (ATA) was calculated as the square root of the summed squared differences between alternate ensemble averaged T waves for each X, Y and Z vector. All calculations were performed using Mathcad Professional signal processing software on a Acer Pentium 2 PC. Comparisons were made between each mean group ATA as well as between the intra and inter-group means of vectors representing each third time interval of the respective mean ATAs. The spectral content was calculated for each group mean ATA by digital fast Fourier algorithm and cross spectral analysis was calculated between each group mean ATA to assess differences in the frequency components within each group. In order to assess group differences in the 80 to 500 Hertz frequency band, mean magnitude and standard deviation of the 80 to 500 Hz frequency band in the group with the smallest mean magnitude was considered noise and group differences between mean magnitudes in this band only considered significant if they exceeded three times the previously calculated standard deviation. {#F1} The following describes the method by which the digitised ensemble averaged alternate T wave vectors (X1, X2, Y1, Y2, Z1 and Z2) were used to calculate the mean ATAs for each group. Time of T wave Analysis= t = 400 milliseconds Subjects in groups 1 and 2 = 1 to N1 and 1 to N2 \[Vector **X1** - Vector **X2**\]^2^ = Vector (**DiffX**)^2^ \[Vector **Y1** - Vector **Y2**\]^2^ = Vector (**DiffY**)^2^ \[Vector **Z1** - Vector **Z2**\]^2^ = Vector (**DiffZ**)^2^ Resultant Absolute Difference Vector (ATA) = √ (**DiffX**^2^ + **DiffY**^2^ + **DiffZ**^2^)  Mean value of Group (1 or 2) mean ATA= \[Group (1 or 2) mean ATA\]/t Statistics ---------- Two tailed t test was used to assess group differences between mean ATAs and intra and inter-group differences between the first (0--133 milliseconds), second (134--237 milliseconds) and third mean ATA (238--400 milliseconds). The statistical group differences between the magnitude of components within the 80 to 500 Hz band have been described above. Results ======= An example of the time aligned alternate sequences of ST-T waves over 400 milliseconds duration is shown in figure [2](#F2){ref-type="fig"}. The mean value and standard deviation for each group member ATA are shown in Table [1](#T1){ref-type="table"}. There was no significant difference between the means for groups 1 and 2 mean ATAs. The mean value and standard deviation for group 1 *v* group 2 being 10.7 microvolts (standard deviation= 7.17 microvolts) *v* 11.74 microvolts (standard deviation= 8.48 microvolts) respectively. All the values for the difference between ensemble averaged T waves for each vector in each patient are shown in table [1](#T1){ref-type="table"}. There was no intra or inter-group significant difference in the mean values between the first, second or third time intervals for each groups mean ATA. The mean values for first, second and third intervals for Group 1 *v* Group 2 being 6.59 v 7.17, 6.37 *v* 7.96 and 6.06 v 6.18, respectively. The Power Spectrum for each groups resultant absolute difference vector between alternate sequences of ensemble averaged ST-T waves showed the largest values between 0 and 25 Hertz with maximum significant differences at the mean or DC frequency, see figures [3](#F3){ref-type="fig"} and [4](#F4){ref-type="fig"}. The cross spectral density function showed that both groups shared the same frequency components, see figure [5](#F5){ref-type="fig"}. There was no significant group differences between the mean magnitudes of frequency band 80--500 Hertz for Group 1 *v* Group 2. The standard deviation in this band for Group 1 (noise) = 0.562, the mean magnitudes for Group 1 v Group 2 in this band being 0.615 v 0.823. {#F2} {#F3} {#F4} {#F5} ::: {#T1 .table-wrap} ::: {.caption} ###### Values of the squared microvolt differences between ensemble averaged T waves meaned over time, in each patients X, Y and Z leads and values of each patients ATA. Patient numbers 1 to 7 represent Group 1 and patient numbers 8 to 16 represent Group 2. Values \< 0.015 rounded to 0. ::: X Y Z ATA ---- -------- -------- ------- ------- 1 304.85 0 72.76 19.43 2 0 0 0 0 3 144.96 0 94.67 15.48 4 24.6 0 0 4.96 5 38.8 55.65 0 9.72 6 4.6 44.62 7.07 7.5 7 192.93 32.26 91.58 17.8 8 143 278.9 0 20.05 9 2.16 2.16 17.5 4.67 10 40.57 0 6.66 6.87 11 186.6 0 0 13.66 12 57.5 31.92 33.9 11.1 13 0 0 0 0 14 0 17.3 0 4.16 15 225.6 291.7 9.36 22.9 16 19.45 320.41 133.9 21.8 ::: Discussion ========== The purpose of this pilot study was to assess the ability of the absolute measurements for the microvolt magnitude of TA to differentiate between one group with an established high risk of MA (Group2) versus a group of patients with multiple physical substrates for developing MA but without a history of a significant arrhythmic event (Group 1). Group 1 was chosen to be representative of those patients whom make up a large proportion of the inexorably increasing coronary heart disease population and on clinical criteria would theoretically be at highest risk from developing a MA and therefore most likely to benefit from accurate risk stratification and defibrillator implantation. It is important to scrutinise the limitations of any index proffered to help risk stratification and aid the focussing of limited resources of invasive electrophysiological investigation with a view to device therapy. Assuming a standard error of approximately 6% for the measurement of ejection fraction by acoustic quantification method, the two groups had similar ejections fractions and presumably similar degrees of post-infarction ventricular damage. The lower ejection fraction of Group 1 would be compatible with the intended selection bias and the greater degree of scarring and contractile dysfunction function in this group. Both groups were stable in terms of heart failure and the absence of dynamic symptomatic ischaemic syndromes, therefore the observed measurements of TA in time and frequency domains are expected to arise from mechanism associated with the post myocardial infarction scarred tissue and it\'s immediate periphery. Conventional TA indices either measure the alternans ratio \[[@B11]\] which is the extent to which the measured alternans exceeds the uncertainty (noise) of the measurement and conveys the statistical degree of confidence in the alternans measurement, or the cumulative alternans voltage in microvolts, which is derived from the frequency power spectrum and analogous, but not equivalent to resultant vector of absolute difference in alternate T wave sequences (ATA) used as the index in this study. The spectral method is an indirect method which measures the magnitude of TA as the amplitude of spectral power at the 0.5 cycles per beat frequency and then subtracts the mean background noise. This method will not compensate for the presence of superposed noise contamination of the 0.5 frequency component amplitude which is then squared to derive the power term. There is an assumption with the spectral measurement method, that the noise is all broad band but it may be non-linear or multiplicative and therefore not eliminated by the spectral method of subtracting the mean of the baseline noise. Perturbations to increase the heart rate have been used to improve the signal to noise ratio of the spectral method in order to improve the sensitivity to detect significant TA. This study describes a direct method which measures absolute TA which for reasons described in the next paragraph has a high signal to noise ratio. This may account for the higher mean values of ATA compared to the values of mean cumulative alternans voltage which is circa 1.2 microvolts. As the index ATA is an absolute measure of the alternating microvolt T wave amplitude, what it conveys is more in faith with the definition of TA. This pilot study demonstrates that the resultant vector of absolute differences in TA microvoltage measured as the index ATA, which was obtained from ensemble averaged alternate T wave sequences during breath holding, may demonstrate low specificity when applied to the risk stratification in those post-infarction patients with ischaemic cardiomyopathies. A recent study by Sakabe and colleagues have demonstrated a specificity of only 35% in a prospective study using TA to predict the development of ventricular tachycardia in non-ischaemic cardiomyopathy patients \[[@B16]\]. This paper supports our results and has significant implications for the specificity of TA when used as a discriminatory probe in the cardiomyopathy patient subset. Our method of measuring absolute TA minimises inaccuracies due to noise by ensemble averaging the data collected under optimal conditions of stability of the electrocardiographic tracing by suspending respiration in a supine rested patient. The noise remaining after ensemble averaging is then largely eliminated by subtracting the alternate ensemble averaged T wave signal sequences from each other, thus subtracting the equal amounts of background noise within each of the sequences from each other. It is known that TA can be produced by motion of the heart within the thoracic cavity, previous studies have not addressed the augmentation of such motion with respiration. It has been suggested that using a resultant vector of orthogonal leads is sufficient to mitigate against the effects of mechanical rotation of the heart when measuring TA \[[@B11]\] but theoretically the resultant vector will oscillate with cardiac and respiratory motion. Data from long, noisy, non-stationary ECG recordings, of the type found in Holter recordings or during exercise, would add to inaccuracies resulting from cardiac motion. We attempted to correct for any such inaccuracy arising from respiration by suspending the respiration in half held breath. The number of recordings of complexes was limited due to the time over which the patients could comfortably breath-hold but the accuracy of the measured absolute resultant vector of microvolt TA in a limited number of complexes was considered an advantage over a TA index compiled from a more numerous but more variable and inaccurate data source. For both groups the absolute difference in mean ATA was distributed equally in time throughout the ST-T wave, which corresponds to the finding that the maximum differences in frequency content for the ATAs in both groups occur at the mean or dc frequency component but there was also a significantly increased magnitude of ATA Within the 0--25 Hertz frequency band which would include the T wave peak frequency and subharmonics within the T wave. In the same way that spatial heterogeneities of depolarisation are responsible for the spectral component markers of late potentials within the QRS complexes, Fourier techniques were applied to the ST-T waves to determine the presence of any difference in the higher frequency bands between alternating ST-T sequences which may indicate the presence of alternating spatial heterogeneities of repolarisation. The finding of no statistically significant group differences in mean magnitude within the frequency band 80 to 500 Hertz is compatible with the findings of Steinbigler et al \[[@B15]\] and does not provide a probe to differentiate these two groups. This study only showed a significant difference in the mean frequency of alternating sequences, which suggests that the differences in measured alternating microvolt magnitudes of the ST-T wave were not due to a change in the morphology of the wave due to introduction of discrete frequency components. The ATAs in the time domain also showed homogenous distribution throughout the ST-T waves, which implies that the origin of TA is not an anatomically localised phenomena as it does not appear to arise from the repolarisation gradient generated between the epicardial, endocardial and M myocardial layers, which according to Antzelevitch and colleagues correspond to the epochs of the start of the peak T wave, the initial descending T wave and end of the descending T wave respectively \[[@B17]\]. There is controversy as to whether TA indices show significant specificity when identifying patients with idiopathic cardiomyopathy at high risk of ventricular arrhythmias \[[@B14],[@B16]\]. Alternatively, TA has been shown to be very useful in identifying arrhythmic risk in the non-specific post infarction population, where sensitivities and negative predictive values are 93% and 98% respectively \[[@B13]\]. However, the major arrhythmic risk lies in the large post infarction cardiomyopathy population but there has been little focus on the use of TA markers in this specific subgroup with coronary heart disease. On the basis of this pilot study, larger scale studies are indicated with electrophysiological ventricular stimulation as reference standard to further assess the sensitivity and specificity of absolute and spectral derived indices of TA to identify patients at high risk of MA within the population of post infarction ischaemic cardiomyopathic heart failure patients. Conclusions =========== This pilot study did not show any significant difference in the magnitude and distribution of absolute TA microvoltages or spectral components between Groups 1 and 2. Our results in ischaemic cardiomyopathy patients is supported by another study on cardiomyopathy subsets which has shown a low specificity for TA to identify arrhythmic risk. These results suggest larger scale studies are needed. Abbreviations ============= TA= T wave alternans ATA = Resultant absolute difference vectors MA= malignant arrhythmia Competing interests =================== None declared.

Background ========== Nutrient cycling between geographically distinctive ecosystems can produce zones of major productivity and biodiversity. It is generally recognized that downstream transport of terrestrial nutrients into marine estuaries produces one of the world\'s most productive habitats, but recent investigations suggest that the reverse flow, from marine to terrestrial habitats, may also be exceptionally important in structuring highly diverse coastal ecosystems \[[@B1]\]. Every year in the Pacific Northwest anadromous salmon (*Oncorhynchus* spp.) transport marine-derived nutrients from the North Pacific Ocean into coastal ecosystems. This salmon nutrient subsidy extends from aquatic habitats into riparian forests, and is thought to be ecologically equivalent to the migration of the wildebeest on the Serengeti \[[@B2]\]. Stable isotope studies in aquatic and terrestrial ecosystems reveal that salmon contribute highly to yearly protein intake for many vertebrates \[[@B1],[@B3]-[@B5]\] and invertebrates \[[@B6],[@B7]\], and provide substantial nutrient inputs to limnetic food webs \[[@B6],[@B8]-[@B10]\], and riparian vegetation \[[@B7],[@B11]-[@B13]\], emphasizing the ecological magnitude of this keystone resource for coastal communities. Transfer of salmon nutrients into terrestrial habitats occurs primarily through bear (*Ursus* spp.) mediated salmon carcass transfer \[[@B14]-[@B16]\] and urine deposition \[[@B12]\], but can also occur as a result of flooding events \[[@B11]\], hyporheic zone transfer \[[@B5]\], or the activities of other scavengers and predators \[[@B3],[@B5]\]. Since nitrogen is often limiting in coastal temperate rainforests of the Pacific Northwest \[[@B17]\], this salmon nutrient pulse to riparian forests can provide a significant proportion of plant total nitrogen \[[@B11]-[@B13]\], and is thought to increase riparian primary productivity, vegetation and litter quality, and soil nutrient capital \[[@B13]\]. Studies in forest ecosystems adjacent to salmon streams have so far been limited to vegetational use of salmon nutrients and have ignored other potential food web beneficiaries, particularly terrestrial invertebrates. Macro-invertebrates of coastal coniferous forests of the Pacific Northwest, including insects, arachnids, myriapods, annelid worms, isopods and gastropods, comprise the base of the myriad of nutrient and energy pathways from primary producers through to higher vertebrate consumers, and are highly important in many ecosystem processes including herbivory, litter decomposition, and nutrient cycling \[[@B18]-[@B20]\]. We use a dual isotope approach of δ^15^N and δ^13^C to assess: a) the extent of utilization of salmon-derived nitrogen and carbon by various trophic groups in a terrestrial invertebrate forest litter community and b) the mechanism of salmon nutrient utilization by invertebrates; either directly through salmon tissue consumption, or indirectly through utilization of salmon nitrogen sequestered into riparian vegetation or soil N pools. We compare the cycling of nutrients above and below waterfalls as a means of examining ecological discontinuities that may occur in litter-based macro-invertebrates between salmon and salmon-free forest sites, and speculate on possible implications to invertebrate community structure and higher vertebrate consumers. We also discuss components of invertebrate isotopic variability as it relates to microspatial variability in δ^15^N, invertebrate trophic structure, and invertebrate niche. Results ======= Invertebrate trophic groups varied predictably with respect to δ^15^N. The nested ANOVA analysis demonstrated that the majority of variance in δ^15^N was due to falls within watersheds (F = 9.191; p = 0.031; R^2^ = 0.819) and taxonomic group within all other factors (F = 13.71; p \< 0.001; R^2^ = 0.689). Variation in δ^15^N that occurred between watersheds or distance of collection from the stream contributed little to total variance and was insignificant in the model (See methods for violations). Invertebrates were enriched by 3--8‰ along salmon spawning reaches compared to similar groups collected above the falls, and showed a gradient of increasing values with increased trophic level at both salmon and non-salmon sites (Figure [1](#F1){ref-type="fig"}). There were highly significant differences in δ^15^N (t-tests: p \< 0.01) above and below waterfalls for all trophic groups at both watersheds. Multiple comparison tests (Tukey\'s post hoc) revealed distinct trophic separation in δ^15^N between at least two invertebrate groups depending on site of collection (Table [1](#T1){ref-type="table"}). Millipede detritivores had higher δ^15^N values than root feeding weevils on all sites but only on the Clatse above the falls was this trend significant. Carabid beetles demonstrated higher δ^15^N values than millipedes at all sites with significant differences on the Clatse River below and above the falls and on the Neekas River above the falls. Spider predators were significantly more enriched than carabid beetles on the Neekas River on both salmon and non-salmon sites, but demonstrated only marginally higher δ^15^N values than these beetles on the Clatse River. Carabid beetle omnivores and spider predators demonstrated significantly higher variance in δ^15^N below the falls than above on both watersheds (Carabidae Clatse: F~14,6~ = 14.61, p \< 0.005; Carabidae Neekas: F~21,6~= 21.94, p \< 0.001; Araneae Clatse: F~18,16~ = 5.41, p \< 0.002; Araneae Neekas: F~17,11~ = 4.94, p \< 0.02) (F-ratio tests). {#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### Tukey\'s multiple comparison post hoc tests for δ^15^N values in four invertebrate trophic groups collected above and below waterfalls on the Clatse and Neekas Rivers, British Columbia. N/A indicates post hoc tests not possible due to low sample sizes. ::: Clatse River ---------------------- ---------------------- ----------------------- -------------- ----------------------- -------------- Predators Root Feeders 5.93 \<0.001 8.31 \<0.001 Predators Detritivores 5.81 \<0.001 4.39 \<0.001 Predators Omnivores 1.28 0.615 1.30 0.123 Omnivores Root Feeders 4.65 0.002 7.01 \<0.001 Omnivores Detritivores 4.52 0.005 3.09 \<0.001 Detritivores Root Feeders 0.13 1.000 3.92 \<0.001 Neekas River Below falls Above falls Trophic grouping (1) Trophic grouping (J) Mean Difference (I-J) Significance Mean Difference (I-J) Significance Predators Root Feeders 8.55 \<0.001 5.33 N/A Predators Detritivores 6.14 \<0.001 5.32 \<0.001 Predators Omnivores 5.84 \<0.001 3.34 \<0.001 Omnivores Root Feeders 2.7 0.548 1.99 N/A Omnivores Detritivores 0.3 0.999 1.98 0.006 Detritivores Root Feeders 2.4 0.693 0.01 N/A ::: Invertebrate groups varied in δ^13^C but did not always vary predictably with trophic level or habitat (Figure [2](#F2){ref-type="fig"}). Nested ANOVA analysis using δ^13^C indicated significant variability only in taxonomic groupings (F = 11.801; p \< 0.001; R^2^ = 0.657), with all other levels insignificant. Relatively high δ^13^C values were observed in millipedes from both watersheds in salmon and non-salmon sites, most likely a reflection of inorganic carbon content. Multiple comparisons revealed trophic separation for spiders over carabid beetles in all sites (Table [2](#T2){ref-type="table"}). Spiders were enriched over root feeders on the Clatse River above the falls and on the Neekas below the falls. Carabids and root feeders did not differ in their δ^13^C values. Carabid beetles collected on the Neekas River were the only group to demonstrate isotopic enrichment below the falls (p = 0.042). Spiders on the Clatse River were found to be higher in δ^13^C above the falls than below (p= 0.016). {#F2} ::: {#T2 .table-wrap} ::: {.caption} ###### Tukey\'s multiple comparison post hoc tests for δ^13^C values in four invertebrate trophic groups collected above and below waterfall barriers to salmon migration on the Clatse and Neekas Rivers, British Columbia. N/A indicates post hoc tests not possible due to low sample sizes. ::: Clatse River ---------------------- ---------------------- ----------------------- -------------- ----------------------- -------------- Predators Root Feeders 0.77 0.555 1.17 0.004 Predators Omnivores 1.27 0.036 2.14 \<0.001 Omnivores Root Feeders -0.49 0.847 -0.92 0.058 Neekas River Below falls Above falls Trophic grouping (1) Trophic grouping (J) Mean Difference (I-J) Significance Mean Difference (I-J) Significance Predators Root Feeders 1.80 0.033 0.90 N/A Predators Omnivores 2.02 \<0.001 2.46 \<0.001 Omnivores Root Feeders -0.22 0.984 -2.08 N/A ::: We examined isotopic levels in relation to distance upstream from the ocean. At Clatse River, δ^15^N declined with increased distance upstream with the lowest levels occurring above the waterfalls. However, at Neekas River, δ^15^N levels were high but variable throughout the stream channel below the waterfall, above which there was a striking reduction in δ^15^N over short distance delineated by the geological barrier to salmon (Figure [3](#F3){ref-type="fig"}). {#F3} In order to assess niche differences within and among groups, we examined the relationships between δ^15^N and δ^13^C. Below the falls, there were significant positive correlations between δ^15^N and δ^13^C in spiders on the Clatse (R = 0.562; p = 0.012) and on the Neekas (R = 0.741; p = 0.001), and in carabid beetles on the Clatse (R = 0.682; p = 0.005) and on the Neekas (R = 0.538; p = 0.010) (Figure [4](#F4){ref-type="fig"}). None of the remaining correlations were significant in groups collected below the falls, and there were no significant correlations between δ^15^N and δ^13^C for any group collected above the falls. {#F4} We estimated contribution of marine-derived nitrogen to the total nitrogen content among invertebrate groups on both watersheds (Table [3](#T3){ref-type="table"}). At Clatse River, assuming no fractionation, values ranged from 19% in millipedes to 49% in weevils (with fractionation: 28% in millipedes to 71% in weevils). At Neekas River, assuming no fractionation, values ranged from 35% in ground beetles to 51% in spiders (with fractionation 47% in ground beetles to 70% in spiders). ::: {#T3 .table-wrap} ::: {.caption} ###### \% Marine-derived nitrogen (MDN) estimates in invertebrates collected above and below waterfall barriers to salmon migration on the Clatse and Neekas Rivers, British Columbia. Estimates of % MDN were made under conditions of no fractionation from the source and maximum fractionation of 4‰ from the source nitrogen by primary producers. ::: Clatse River --------------------------- ------------------------- -------------------------- Curculionidae 49.1% 71.5% Parajulidae 19.4% 28.2% Carabidae 30.6% 44.6% Agelenidae/Antrodiaetidae 30.4% 44.3% Neekas River Invertebrate family(ies) %MDN (No Fractionation) %MDN (Max Fractionation) Parajulidae 45.8% 62.3% Carabidae 34.8% 47.3% Agelenidae/Antrodiaetidae 51.2% 69.6% ::: Discussion ========== We demonstrate isotopic evidence for substantive incorporation of salmon-derived nitrogen into multiple trophic levels of terrestrial litter-based invertebrates from two salmon bearing watersheds. Enrichment in δ^15^N in terrestrial invertebrates occurs through two possible pathways: 1) direct consumption of salmon tissue and/or predation off of direct salmon consumers such as larval blowflies; or 2) indirect enrichment through δ^15^N enriched soil and vegetation N pools. Here, the use of the dual isotope method provides insight into the mechanism of salmon nitrogen utilization by terrestrial invertebrates. Direct consumption of salmon, with approximate δ^15^N and δ^13^C values of +11.2‰ \[[@B21]\] and -21‰ \[[@B9]\] respectively, would lead to enriched signatures of δ^15^N and δ^13^C in animal tissues. For example, consumption of salmon carcasses by larval blowflies (Calliphoridae) has been documented through the dual isotope method \[[@B7]\]. However, terrestrially derived carbon through C~3~ photosynthesis dominates δ^13^C pools in coniferous forest soils and salmon-derived carbon is assumed to contribute little to total carbon in litter and soil. The process of indirect utilization of salmon-derived nitrogen by animals has been observed previously in small mammals \[[@B11]\], whereby individuals were enriched in δ^15^N but not δ^13^C. Because we found little differences in δ^13^C in all trophic groups collected above versus below the waterfalls, this suggests that the primary mechanism of δ^15^N enrichment is by indirect processes through salmon-derived nitrogen subsidies to soil and vegetation N pools. δ^15^N / δ^14^N ratios of forest nitrogen pools are influenced by the isotopic values of nitrogen inputs and outputs and fractionation that occurs during nitrogen transformations within ecosystems \[[@B22]\]. Nitrogen inputs to typical Pacific coast forest ecosystems include atmospheric deposition and biological nitrogen fixation. In the case of forests adjacent to salmon streams there is substantial evidence that marine-derived nitrogen from salmon is transferred to forest ecosystems through predator activity \[[@B11],[@B12],[@B14]-[@B16]\], flooding events \[[@B11]\] and hyporheic zone transfer \[[@B5]\], and is incorporated into soil N pools through uptake by vegetation \[[@B6],[@B7],[@B11]-[@B13]\]. Vegetation δ^15^N values tend to parallel those in the soil and litter across multiple sites and are typically slightly depleted in δ^15^N relative to the soil source \[[@B22],[@B23]\]. Recent estimates for the contribution of marine-derived nitrogen from salmon in riparian ecosystems to total plant nitrogen have ranged from 15.5--24% \[[@B6],[@B12],[@B13]\]. These values may be conservative as they are based on the assumption of no plant fractionation from the original source nitrogen. In the case of high nitrogen inputs from salmon, vegetation may preferentially assimilate isotopically light nitrogen (even though it is also originally from salmon). However, in nutrient rich habitats fractionation from the source is potentially not as marked compared with nutrient poor soils \[[@B23],[@B24]\], making %MDN estimates challenging. %MDN estimates from hemlock (Mathewson & Reimchen unpublished data), possibly constituting a large percentage of litter biomass, vary from 23--34% on the Clatse River and 49--66% on the Neekas River depending on degree of fractionation from the source. These estimates are higher than previously reported, yet remain the baseline for comparison with %MDN estimates in our litter-based invertebrate community. Ponsard and Arditi \[[@B25]\] observed substantial site variation in litter and soil δ^15^N due to variations in soil processes and nitrogen sources across small scales (\< 1 km). Soil and litter δ^15^N and δ^13^C values are not yet available for our sites. However, δ^15^N values in litter-based terrestrial invertebrates are known to parallel the δ^15^N values in the litter and soil \[[@B25],[@B26]\]. We suspect that because vegetation and all invertebrates collected below the waterfall barrier to salmon migration are enriched in δ^15^N, that soil and litter δ^15^N are also enriched at these sites. Our data demonstrates that terrestrial invertebrates exhibit a substantial shift in δ^15^N over a sharp ecological discontinuity (ca. 250 m) in the source of nitrogen to the forest community, as a consequence of a distinct salmon-derived nitrogen subsidy to litter, soil and vegetation N pools. We estimate that %MDN to multiple trophic levels of litter-based invertebrates ranges from 19--71% on the Clatse River and 34--70% on the Neekas River depending on trophic grouping, and on the extent of fractionation from the original source nitrogen. These values are similar to %MDN estimates of hemlock and indicate that salmon-derived nitrogen is cycled from primary producers through multiple trophic levels of litter-based terrestrial invertebrates. Grouping all invertebrate samples over the entire 100 m riparian zone may have reduced the extent of statistical differences for δ^15^N in our comparisons above and below falls. This occurs because of a potential isotopic gradient of decreasing δ^15^N from salmon in terrestrial vegetation with increasing distance from the stream over a relatively small scale (\< 100 meters) \[[@B11]-[@B13]\]. Nevertheless, our %MDN estimates are higher than any other study investigating salmon nutrient transfer into terrestrial ecosystems and emphasizes the magnitude of the discontinuity that occurs across the waterfall barrier to salmon migration in these watersheds. These %MDN estimates assume salmon tissue δ^15^N as the marine end-member in the model. However, there are other factors that can influence these estimates. Vertebrate urine, particularly from bears (*Ursus* spp.) \[[@B12]\], faeces and guano deposition may contribute highly to nitrogen inputs during the salmon spawning season. Despite the fact that these inputs are ultimately from salmon tissue consumption, high fractionation during multiple transformation steps prior to nitrogen availability, such as ammonia volatilization \[[@B22]\], may lead to unknown shifts in the δ^15^N levels of the source nitrogen. This may increase the microspatial variability in δ^15^N in litter, soil, and vegetation, and subsequently invertebrates, along the salmon spawning channel. Variation in δ^15^N in carabid beetles and spiders collected below the waterfall barrier was substantially greater than above the falls. It was only marginally higher (non-significant) in root feeding weevils and millipede detritivores, possibly due to low sample sizes. This may indicate higher microspatial variability in δ^15^N in soil, litter and vegetation N pools, increased range of prey resources below the falls, and/ or invertebrate dispersal from other habitats into the zone of substantial salmon transfer. We detected variation in δ^15^N at different stream reaches, most likely as a function of abundance and species of spawning salmon. On the Clatse River, δ^15^N values decreased with increasing distance upstream. Potentially, this might result from a gradient in marine subsidies other than salmon as a function of distance from the estuary \[[@B27]\]. However, this trend was not observed on the Neekas River where δ^15^N values remain high, even at 2 km upstream. The difference between these two watersheds in the distribution of marine-derived nitrogen appears to be due to topography and the species and distribution spawning salmon. Clatse River is pink salmon dominated, with the majority of spawning, and subsequent predator activity, occurring in the lower 500 meters of the spawning channel \[[@B28]\] (personal observations). Above 600 meters the stream narrows and the riparian profile becomes increasingly steep on both sides. The Neekas River has high density chum spawning to the base of the falls with high salmon nutrient transfer and predator activity occurring in this region \[[@B28]\] (personal observations). Chum salmon contain twice the biomass of nitrogen than pink salmon, and this may partly explain the higher %MDN estimates obtained on the Neekas River compared to the Clatse. The distribution of δ^15^N in these terrestrial invertebrate groups thus appears to be directly correlated to salmon spawning density and biomass, and subsequent predator activity, a pattern that has been observed for δ^15^N in ground beetles (Carabidae) occurring between watersheds on Vancouver Island \[[@B7]\]. Differences in the variance of isotopic signatures within a population provide insight as to the range of diet available to the individual. For example, this has been found in stable isotope studies of marine mammals and chimpanzees \[[@B29],[@B30]\]. In the case of carabid beetles and spiders, high variability in δ^15^N along the salmon-spawning channel compared to above the falls, may indicate higher prey variability in this region. Variance in isotopic signatures can also indicate mobility between habitats \[[@B31],[@B32]\]. Carabid beetles, particularly on the Neekas River, exhibited high variance in signatures. The carabid beetle species collected, although brachypterous, can move freely between habitats \[[@B33]\], and captured individuals may not have obtained their nutrition along the salmon spawning channel for their entire life history. Correlations between δ^15^N and δ^13^C values provide further resolution into individual niche variability. We observed a significant positive correlation between δ^15^N and δ^13^C values in carabid beetles and spiders below waterfalls, with access to salmon nutrients, but not above falls. Both groups feed on a diverse array of prey including primary and secondary consumers, and in the case of the ground beetles, vegetative matter as well. Individuals within each group that fed at a higher average trophic level would be expected to exhibit more enrichment for δ^15^N and δ^13^C \[[@B34],[@B35]\]. Alternatively, individuals that fed on salmon directly or on prey that fed on salmon would also demonstrate isotopic enrichment in both isotopes \[[@B3]-[@B7]\]. Positive relationships in δ^15^N and δ^13^C below the falls and the absence of that relationship above the falls hints that direct consumption of salmon or salmon consumers below the falls may be a factor for some individuals of these species. However, increased range of food resources below the falls would also be consistent with this finding. Furthermore, smaller sample sizes above the falls may have reduced our ability to detect relationships. For the majority of the spiders and ground beetles, direct uptake of the marine isotopes most likely contributes only a minor component to yearly protein intake, as uptake of marine-derived nitrogen occurs by indirect means. The use of dual isotope model becomes most relevant when investigating terrestrial organisms that use salmon protein as a major contributor to diet. This is the case for several terrestrial necrophages including flies (Diptera: Calliphoridae, Scathophagidae, Anthomyiidae), and beetles (Coleoptera: Silphidae, Leiodidae, Staphylinidae) \[[@B7]\] (Hocking unpublished data). Animals are isotopically enriched in δ^15^N and δ^13^C relative to their dietary intake as a consequence of preferential excretion of the lighter isotope in metabolism \[[@B36]\], and this allows insight into relative trophic position within a community. Isotopic enrichment varies widely by body tissue, but there is an approximate stepwise enrichment of 3.4 ± 1.1‰ for δ^15^N \[[@B35],[@B37]\] and 0.4 ± 1.4‰ for δ^13^C \[[@B34],[@B38]\] for each sequential trophic level. Ponsard & Arditi \[[@B25]\] suggest that there are on average two trophic levels within litter-based invertebrate communities. We also find general evidence for two general trophic levels within the litter-based community at Clatse and Neekas Rivers usually consisting of: 1) root feeders and detritivores (weevils and millipedes) as primary consumers of plant material, and 2) predators (carabid beetles and spiders) that feed on these and other presumed plant feeders within the litter community. Our data, however, provides substantial evidence for a gradient in trophic level among our litter-based invertebrates rather than two distinct trophic groupings, a finding that coincides with that of Scheu & Faica \[[@B26]\]. Millipedes, for instance, were often found to be enriched in δ^15^N compared to root feeders, a finding that suggests that either weevils (Curculionidae) feed on roots that are somewhat depleted in δ^15^N compared to litter, or that millipede detritivores utilize some δ^15^N enriched protein food sources such as bacteria in their guts, or both \[[@B25]\]. Spiders were enriched in δ^15^N in all cases over those in carabid beetles, and below the falls on the Neekas this constituted a mean difference greater than a single trophic level. Evidence for omnivory is emerging in the carabid beetles \[[@B33],[@B39]-[@B42]\] and the observed discrepancy between spiders and carabid beetles is most likely a result of the purely predaceous versus omnivorous life histories of these groups. Spiders also demonstrated trophic enrichment in δ^13^C over carabid beetles at all sites. However, spiders were not consistently enriched over root feeders at each site and carabid beetles exhibited the lowest δ^13^C values. We conclude that, in general, carbon is a poor trophic level indicator \[[@B25]\]. Overall, this suggests that increased trophic and individual niche resolution in stable isotope studies will more likely extend from a detailed taxonomic separation rather than with guild analyses \[[@B26]\]. Implications ============ With the use of stable isotopes (δ^15^N and δ^13^C), spawning salmon have been shown to provide substantial nutrient inputs to limnetic food webs \[[@B6],[@B8]-[@B10]\], with implications for stream primary productivity and subsequent juvenile salmonid survivorship. Young salmon may in fact derive a large proportion of their required nitrogen and carbon from the death and decomposition of their parents, through food web utilization of salmon nutrients by algae and aquatic invertebrates. Other than inputs to terrestrial vegetation, salmon nutrient effects in forest food webs are poorly known. Input of salmon-derived nitrogen contributes to total available N in the soil and thereby increases forest primary productivity and vegetation and litter quality \[[@B11]-[@B13]\]. Nutrient subsidies (other than salmon) to terrestrial invertebrate communities can result in shifts in invertebrate community structure and abundance as a consequence of bottom-up ecosystem effects \[[@B27],[@B43],[@B44]\]. Soils in coniferous forests of low nutrient status are typically dominated by fungi as the primary decomposers of organic material, and thick humus layers quickly accumulate due to slow rates of nutrient turnover \[[@B45]\]. In nutrient-rich conditions, fungi are replaced by bacteria and invertebrates as the dominant decomposers, resulting in higher net rates of nitrogen mineralization and total available nitrogen \[[@B44],[@B45]\]. Shifts in invertebrate community structure and abundance due to a nutrient subsidy may have further implications for higher invertebrate and vertebrate consumers such as predaceous beetles, spiders, hymenopteran parasitoids, small mammals, amphibians and passerines. For example, in another form of marine subsidy, spider densities have been reported to be 4--5 times higher on islands with marine bird colonies than those without \[[@B46]\]. Furthermore, avian populations in boreal forests have been observed to respond to experimental nitrogen fertilization \[[@B47]\], a pattern that also may well be true in the case of nutrient inputs to forest communities along salmon streams \[[@B48]\]. Shifts in litter-based invertebrate community structure and abundance could have particular benefits for ground foraging birds such as the resident and migratory sparrows, thrushes and wrens. The widespread enrichment in salmon derived nitrogen among multiple trophic levels also hints at an ecosystem level effect that has further implications for shrub and canopy level invertebrate communities and their various vertebrate consumers \[[@B1],[@B5],[@B48]\]. Conclusions =========== The increasing evidence for the coast-wide decline in salmon abundance on the Pacific coast of North America \[[@B49]\] may have substantially more ecological implications to terrestrial forest food webs than previously recognized \[[@B5]\]. We present evidence for major uptake of salmon-derived nitrogen into a terrestrial invertebrate food web, with a sharp reduction in uptake across a waterfall barrier to salmon migration. These results supplement the conclusions of a diversity of recent contributions that have focused on the ecological consequences of the decline of salmon on the west coast of North America \[[@B1],[@B2],[@B5]-[@B13],[@B48]\]. Methods ======= Site Description ---------------- Two salmon bearing streams were investigated -- the Clatse (52° 20.6\'N; 127° 50.3\'W) and Neekas Rivers (52° 28.4\'N; 128° 8.0\'W), on the mid-coast of British Columbia, near Bella Bella, Canada. Both watersheds occur in the Coastal Western Hemlock Biogeoclimatic Zone along the boundary between the central very wet hyper-maritime (CWHvh2) and sub-montane very-wet maritime (CWHvm1) subzones \[[@B50]\]. Climate is considered cool and wet with mean annual temperature of approximately 8°C, and mean annual precipitation above 4000 mm (Environment Canada 2001). Dominant tree species include Western Hemlock (*Tsuga heterophylla*), Sitka spruce (*Picea sitchensis*), Amabilis fir (*Abies amabilis*), Western redcedar (*Thuja plicata*), and Red alder (*Alnus rubra*). Common understory species include Alaskan blueberry (*Yaccinium alaskaense*), red huckleberry (*V. parvifolium*), false azalea (*Menziesia feruginea*), deer fern (*Blechnum spicant*), bunchberry (*Cornus canadensis*), lanky moss (*Rhytidiadelphus loreus*), step moss (*Hylocomium splendens*), and common green sphagnum (*Sphagnum girgensohnii*) on zonal sites, and salmonberry (*Rubus spectabilis*), red elderberry (*Sambucus racemosa*), stink current (*Ribies bracteosum*), and spiny-wood fern (*Dryopteris expansa*) on nutrient rich sites. Deep acidic soils predominate with high organic matter content due to low rates of decomposition. Soil deposits are typically alluvial or glacial in origin, are heavily leached, and often contain iron deposits in the B layer. Mor humus types are most common with a thick layer of moss, but moder/mull humus forms occur in nutrient rich sites along the salmon spawning channel. Both the Clatse and Neekas watersheds are dominated by high-density returns of pink (*Oncorhynchus gorbuscha*) and chum (*O. keta*) salmon, with minor runs of coho (*O. kisutch*) and the occasional sockeye (*O. nerka*). In the last ten years, pink and chum salmon returns on the Clatse River average 17000 and 5000 individuals respectively. Chum salmon constitute the majority of spawning biomass on the Neekas (mean = 30000). Mean pink salmon returns on the Neekas River vary from an average of 33000 on even years to an average of 2700 on odd years (Department of Fisheries and Oceans Escapement data: 1990--1999). Suitable spawning habitat extends for 2.1 km on the Neekas River, roughly twice that of the Clatse (1 km), whereby both are interrupted by waterfalls that act as a barrier to salmon migration \[[@B28]\]. Invertebrate samples -------------------- In August of 2000 terrestrial macro-invertebrates were collected in each watershed through passive pitfall trapping and hand collection from the soil and course woody debris. Invertebrate sampling occurred above and below the waterfall barrier and up to 100 meters from the stream. On the Clatse River, main invertebrate sampling occurred from 200 to 800 meters upstream from the mouth, and again above the falls at 1200 and 1600 meters. The majority of invertebrate trapping on the Neekas occurred at 1 km, and again at 2 km, just below the falls. Control samples from the Neekas were collected just above the falls from 2250 to 2400 meters upstream from the mouth. Pitfall arrays were arranged in a three-way branching fashion. This included a central 10 cm diameter pitfall connected via three 24-inch by 6-inch aluminium drift fences (separated by 120°) to a perimeter pitfall at the end of each fence \[[@B7]\]. Pitfall arrays were cleared from four to five days after initial set-up, and to prevent rotting of invertebrate tissue 70% ethanol was used as a field preservative within each pitfall cup. Hand collection of invertebrates occurred more randomly as individuals were discovered in the riparian area. All specimens were stored in 70% ethanol prior to identification and isotopic analysis. Stable Isotope Analysis ----------------------- Whole invertebrate specimens were dried at 60°C for at least 48 hours and ground into a fine powder with a Wig-L-Bug grinder (Crescent Dental Co., Chicago, 111). Approximately 1 mg dry weight per ground specimen was then sub-sampled for continuous-flow isotope ratio mass spectrometry (CF-IRMS) analysis of nitrogen and carbon. Mass spectrometry analysis of δ^15^N and δ^13^C was conducted at the stable isotope facility, University of Saskatchewan, Saskatoon, Canada using a Europa Scientific ANCA NT gas/solid/liquid preparation module coupled to a Europa Scientific Tracer/ 20 mass spectrometer. Isotopic contents are expressed in \'δ\' (delta) notation representing the difference between the isotopic content of the sample and known isotopic standards (atmospheric N~2~ for nitrogen and PeeDee Belemnite (PDB) limestone for carbon). This is expressed in parts per thousand (‰) according to the formula (1): 1\) δ^15^N Or δ^13^C (‰) = (R~sample~ / R~standard~ - 1) \* 1000 where R is the ratio of the heavy isotope (^15^N or ^13^C) / light isotope (^14^N or ^12^C). Data Analysis ------------- Individual terrestrial macro-invertebrates processed for δ^15^N or δ^13^C were separated into four main groups based on taxonomic similarity and ranked according to degree of animal protein consumption, thus providing a proxy for relative trophic level within the litter based food chain: 1) Root feeders (Coleoptera: Curculionidae) \[[@B18],[@B51]\]; 2) Detritivores (Julida: Parajulidae) \[[@B25],[@B52]\]; 3) Omnivores (Coleoptera: Carabidae) \[[@B33],[@B39]-[@B42]\]; 4) Predators (Araneae: Agelenidae, Antrodiaetidae) \[[@B53],[@B54]\] (Table [4](#T4){ref-type="table"}). ::: {#T4 .table-wrap} ::: {.caption} ###### Family and species level designations by trophic grouping for invertebrates collected on the Clatse and Neekas Rivers, British Columbia, in August 2000. ::: Trophic grouping Family Species ------------------ --------------------------- --------------------------------------------------------------------------------------------- 1\) Root Feeders Curculionidae *Steremnius carinatus* Boh. *Steremnius tuberosus* Gyll. 2\) Detritivores Parajulidae Unknown 3\) Omnivores Carabidae *Pterostichus crenicollis* LeC. *Scaphinotus angusticollis* Mann. *Zacotus matthewsii* LeC. 4\) Predators Agelenidae Antrodiaetidae *Cybaeus reticulatus* Simon *Antrodiaetus pacificus* Simon ::: Curculionid beetles of the genus *Steremnius* feed as larvae and adults on the roots and slash of conifers and are assigned the lowest trophic rank, as there is no current evidence that these beetles utilize animal protein \[[@B18],[@B51]\]. Millipedes are detritivores, feeding primarily on dead plant material and fragments of organic matter. This potentially includes small amounts of animal protein from faeces, dead animals or microorganisms that occur on the litter material \[[@B25],[@B52]\]. The Parajulidae are indigenous to the forest ecosystems of the Pacific Northwest but are poorly known at the species level \[[@B55]\]. A priori, we assume here that the parajulid millipedes include minor contributions of organic matter derived from animal protein in diet. Carabid beetles of the genera *Pterostichus, Scaphinotus* and *Zacotus* are generalist forest floor predators on a variety of soil invertebrates including snails and slugs (Gastropoda), millipedes (Diplopoda), isopods (Isopoda), worms (Oligochaeta) and springtails (Collembola) \[[@B33],[@B40],[@B56]\]. However, documented observations of carabids feeding on plant material including seeds and fruit suggest that these beetles may be omnivorous rather than purely predaceous \[[@B39],[@B41],[@B42]\]. Arachnids of the genera *Cybaeus* and *Antrodiaetus* are known to be funnel-web \[[@B54]\] and trap-door spiders \[[@B53]\] respectively, feeding exclusively on animals including various insects, myriapods, isopods, other spiders and even small vertebrates \[[@B57]\]. Independent sample t-tests (two-tailed) were used to test for differences between invertebrate groups collected above and below the falls for δ^15^N and δ^13^C on each watershed (equal variances not assumed in all tests). All invertebrates collected within 100 meters of the stream were pooled for the analysis, and those collected less than 200 meters from the estuary were removed since these were assumed to possess ambiguous isotopic signatures where marine incursions other than salmon input may particularly obscure soil N pools \[[@B27]\]. F-ratio tests (two-tailed) were conducted for δ^15^N between invertebrate groups collected above versus below the falls under the null hypothesis of equal variances. We also performed separate Nested ANOVA\'s on δ^15^N and δ^13^C to examine the effects of trophic group, distance from the stream, above and below falls and watershed \[model: watershed, watershed(falls), watershed(falls(distance)), watershed(falls(distance (invertebrate group)))\]. However, assumptions of normality and homoscedasticity were not met and as such, we place more emphasis on the t-test comparisons. Tukey HSD multiple comparison post hoc tests were performed for δ^15^N and δ^13^C within sites under the null hypothesis that all invertebrate groups were isotopically indistinct. Since inorganic carbon in the form of CaCO~3~, present in the exoskeleton of our millipedes \[[@B52]\], is enriched in δ^13^C relative to organic forms \[[@B25]\], we removed millipedes from the post hoc analysis of δ^13^C among feeding groups. Pearson\'s Correlation Coefficients were used to examine the relationships between δ^15^N and δ^13^C within trophic groups at different sites to investigate the individual niche variability. Estimating % MDN ---------------- δ^15^N values in animals are influenced by the δ^15^N value of the principal N sources, and fractionation during nitrogen transformations within ecosystems. Principal N sources to riparian ecosystems include atmospheric N~2~ with a δ^15^N value of 0‰ \[[@B36]\], and salmon N with a δ^15^N value of approximately 11.2 ± 1.0‰ \[[@B21]\]. Variations in δ^15^N with trophic level appear to be relatively predictable such that biota are enriched by 3.4 ± 1.1‰ more than their food \[[@B37]\], a pattern that seems to hold true for soil macro-invertebrates \[[@B25],[@B26]\]. Estimates for % marine-derived nitrogen (MDN) in our litter-based macro-invertebrate food chain were obtained based on a combination of a limnetic trophic model proposed by Kline *et al.*\[[@B8]\] and a terrestrial vegetation model utilized by Helfield and Naiman \[[@B13]\] and is expressed mathematically by (2): 2\) %MDN = \[(Obs - TEM) / (MEM~TL~ - TEM) \] \* 100% where Obs is the observed δ^15^N value of a particular taxa below the waterfall barrier to salmon, TEM is the terrestrial end-member (the isotopic value obtained for the same taxa above the falls in absence of salmon input), MEM is the marine end-member (δ^15^N value of salmon of 11.2‰ \[[@B21]\] which should equal maximum vegetation δ^15^N values), and TL refers to the trophic level correction factor that applies to the marine end-member in the model. Since variability in utilization of MDN by the various invertebrate groups below the falls might obscure relative trophic level, we used invertebrate δ^15^N values above the falls on each watershed to provide an indication of relative trophic position. The trophic level correction factor was thus calculated by subtracting mean δ^15^N values in hemlock above the falls, (Mathewson & Reimchen unpublished data: Clatse mean δ^15^N = -1.55‰; Neekas mean δ^15^N = -3.93‰) from mean δ^15^N values in each invertebrate group above the falls on each watershed. This simplifies the above equation to (3): 3\) % MDN = \[(Obs - TEM) / (MEM - VEGabv) \] \* 100% where MEM equals salmon tissue \[[@B21]\] and VEGabv equals mean vegetation δ^15^N values above the falls. We also calculated %MDN for vegetation below the falls (Mathewson & Reimchen unpublished data: Clatse mean δ^15^N = +1.43‰; Neekas mean δ^15^N = +3.44‰) as a benchmark comparison to our invertebrate estimates. We were not able to assess the extent of fractionation occurring in the situation of 100% MDN at the level of primary producers (See assumptions in \[[@B13]\]). As such, we calculated two %MDN estimates based on no fractionation (MEM = 11.2‰) and maximum fractionation of 4‰ (MEM = 7.2‰), which is a typical maximum level of fractionation in vegetation from atmospheric N~2~ that is observed in the Clatse-Neekas non-salmon habitats (Mathewson & Reimchen unpublished data). This model assumes that invertebrate trophic level does not differ above and below the falls and that the marine end-member for vegetation δ^15^N values is represented by salmon tissue. Author\'s contributions ======================= MDH conducted the field research, sorted and processed the invertebrate samples, performed the statistical analyses, and drafted the manuscript. TER conceived of the study, participated in its design and coordination, and contributed to the manuscript preparation. All authors read and approved the final draft. Acknowledgements ================ We would like to thank Carsten Brinkmeier, Mike Windsor, Dan Windsor and Chester Starr for field assistance, Dr. Richard Ring for laboratory support and discussion, Myles Stocki for stable isotope analysis Dr. Bristol Foster for the zodiac, Larry Jorgenson for logistical assistance, and Rob Bennett for identification of the Arachnids. Thanks to the Heiltsuk First Nations and the Raincoast Conservation Society for local project support. This work was supported by funds from the David Suzuki Foundation and the Natural Sciences and Engineering Research operating grant (NRC2354) to TER, and from an NSERC -- Industrial Post-graduate Scholarship to MDH.

Introduction ============ Homologous DNA recombination is a fundamental process in the biochemistry of DNA repair and replication, which contributes to the generation of the genetic diversity critical for natural selection. An important step in the recombination process is the pairing of homologous double-stranded DNAs followed by the exchange of DNA strands between the paired molecules. Experimental studies have shown that members of the archetypal RecA family of recombinases are central to this reaction in all extant forms of life \[[@B1],[@B2]\]. Studies in *Escherichia coli* have shown that, although RecA is the principle protein involved in pairing and strand exchange, unrelated proteins, that have a much more restrictive phyletic distribution, can also promote similar reactions in a RecA dependent or RecA-independent manner \[[@B3]\]. These alternative or additional mediators of homologous recombination include the well-characterized prophage RecT, phage λ Redβ and phage P22 ERF proteins \[[@B4],[@B5]\]. Similarly, in yeast and vertebrates, the RAD52 protein is involved in the pairing and strand exchange reaction and can promote recombination in a RAD51 (the eukaryotic RecA homolog)-dependent or independent manner \[[@B6]\]. The RecT protein works in conjunction with the RecE-nuclease \[[@B7]\] and was initially described in genetic studies on the complemention of mutations in the RecBCD pathway of DNA repair \[[@B8]-[@B10]\]. Biochemically, RecT has been shown to bind single-stranded (ss) DNA 3\' overhang regions generated by the RecE nuclease, and promote strand exchange between homologous DNA partners by assisting the pairing of complementary single-stranded regions \[[@B4],[@B10]\]. The reaction catalyzed by the RecT/RecE system is similar to that described for the phage λ exonuclease (exo/Redβ) and the single-strand annealing protein Redβ. The similarity between these two systems is further extended by the observation that the RecT/E system can complement mutations in the λ exo/Redβ system \[[@B10],[@B12]\]. In eukaryotes, RAD52 protein has been shown to exhibit properties similar to those of RecT and Redβ proteins: it binds ssDNA and promotes strand exchange via the pairing of complementary single strands \[[@B6],[@B13]\]. *In vitro* studies on quaternary structures have shown that the single strand annealing proteins (SSAPs), RecT, Redβ, ERF and RAD52, form similar helical super-structures \[[@B14]-[@B17]\]. This has led to the proposal that RecT, Redβ, ERF and the eukaryotic RAD52 function in an analogous fashion, and even are \"structural homologs\" \[[@B14]\]. However, no sequence or secondary structural similarities have been noticed between different SSAPs and current understanding of their evolutionary history and phyletic range remains poor. Here, we describe the results of an in-depth sequence analysis of these proteins and delineate their evolutionary relationships and phyletic horizon in available genomes. We show that, in spite of the functional similarities, and the similar quaternary structures, there are three distinct superfamilies of SSAPs, namely the RecT/Redβ, RAD52 and ERF, that appear to be evolutionarily unrelated to each other. These superfamilies show a wide distribution in viral and cellular genomes, but appear to have originally evolved in large DNA bacteriophages. Through an analysis of the contextual information provided by the predicted operons, in which the SSAPs occur, we predict several previously undetected functional connections of these proteins, which might shed new light on the corresponding DNA repair/recombination pathways. Results ======= RecT and Redβ are evolutionarily related and define a widespread family of DNA recombination proteins ----------------------------------------------------------------------------------------------------- Several lines of evidence, including genetic analyses, and similarities in biochemistry and quaternary structures, suggest that the *E. coli* RecT and phage λ Redβ proteins are functionally equivalent as mediators of single-strand exchange in DNA recombination \[[@B10],[@B12]\]. However, no sequence similarity has been detected between these proteins leaving their actual evolutionary relationships unresolved. In order to gain a better understanding of their functions and origins, we undertook a detailed sequence analysis of these two proteins using iterative sequence profile searches with the PSI-BLAST program with a inclusion threshold of .01 iterated until convergence. Such searches, with Redβ proteins from different lambdoid bacteriophages as queries, retrieved not only other obvious Redβ homologs, but also the RecT protein family. For example, searches initiated with the Redβ homolog, PF161 protein (Genbank gi: 9836834 amino acids 1 to 188) from *Borrelia hermsii*\[[@B18]\], detected the *E. coli* RecT protein in the 5^th^ iteration with significant expectation (e) values (3 × 10^-3^). Subsequent iterations retrieved several more RecT-related proteins from diverse sources. Further, transitive searches with the proteins detected in the above searches resulted in the identification of more divergent homologs, such as a protein termed the \'enterohemolysin associated protein\' (EHAP1) from *E. coli*\[[@B19]\] and its orthologs in *Salmonella* (Fig. [1](#F1){ref-type="fig"}). An examination of the pairwise alignments generated by these searches showed that all these proteins shared a characteristic set of residues, including two highly conserved aromatic residues at the N- and C-termini, respectively, and two consecutive acidic residues near the C-terminus. These observations strongly suggested that RecT and Redβ, along with several other proteins, could be unified into a single protein superfamily with a core conserved domain of approximately 200 amino acid residues. {#F1} A multiple alignment of all members of the RecT/Redβ superfamily was generated using the T_coffee program followed by adjustments based on the PSI-BLAST search results. This alignment was used to predict their secondary structure using the JPRED and PHD methods; these predictions pointed to an α + β domain with a core of five β-strands and five α-helices (Fig. [1](#F1){ref-type="fig"}). Some of the strongest conservation is concentrated in the long helices, and the pattern includes some charged or polar residues, suggesting that they are probably exposed and participate in the protein-protein and protein-DNA interactions that are typical of this superfamily (helices 2,3, 4 in Fig. [1](#F1){ref-type="fig"}). The conserved, regularly spaced hydrophobic residues in the RecT/Redβ superfamily are predicted to be buried, allowing these domains to assume a globular structure. Experimental studies have shown that the strand transfer reaction mediated by RecT and its binding to dsDNA are sensitive to Mg^+2^ concentrations and it was proposed that the levels of free Mg^+2^ could regulate RecT activity \[[@B4]\]. Similarly, Redβ has been shown to promote single strand annealing in a Mg^+2^-dependent manner \[[@B20]\]. In this context, the conservation of the two C-terminal acidic residues in the majority of members of this superfamily suggests that these might be involved in the coordination of Mg^+2^ and implies that the metal ion-dependent conformational switching is likely to be a generic feature of this family. Phylogenetic analyses of the RecT/Redβ superfamily using the least squares and maximum likelihood methods distinguished three distinct groups, namely the RecT-like, the Redβ-like and the EHAP1-like families (Fig. [2](#F2){ref-type="fig"}). Previously, the RecT proteins have been known from very few bacteria and Redβ has only been detected in λ and closely related phages. However, we showed that the Redβ family is widespread in bacteria, such as *Borrelia hermsii, Xylella, Ureaplasma, Listeria, Streptococcus pyogenes, Mesorhizobium loti.* The RecT family is predominantly seen in the low-GC Gram-positive bacteria, such as *Bacillus, Streptococcus, Lactococcus and Listeria,* and their phages (Fig. [2](#F2){ref-type="fig"}). *E. coli* and *Legionella pneumophila* are the only two γ-proteobacteria that possess this protein, suggesting that they might have acquired RecT via a relatively recent horizontal transfer from Gram-positive bacteria. The sporadic distribution of the RecT and Redβ family proteins in bacterial genomes and their presence in phages suggest that these proteins ultimately are of phage origin and have been co-opted by the bacterial DNA recombination/repair systems. Consistent with this, practically all the bacterial members of these families appear to belong to prophages or their remnants, as they are mostly in the neighborhood of what appear to be clearly phage-derived genes. The EHAP1-like family is extremely divergent and represented thus far only in *E. coli* and the closely related *Salmonella.* Practically all members of the RecT/Redβ superfamily are single-domain proteins showing extended similarity to each other throughout their globular regions. The only exceptions are the *E. coli* EHAP1 and the PF161 protein encoded in the *Borrelia hermsii* circular plasmid, which are fused to C-terminal fragments of the ERF protein (see below). {ref-type="fig"}. The gene neighborhoods of the RecT/Redβ superfamily genes are shown in association with the corresponding branches whenever they they contained genes for proteins with plausible functional connections with SSAPs. The hatched boxes represent fragments of ERF (Indicated by E) and SSB (indicated by S) genes encoding C-terminal regions as described in the text. Gene abbreviations are as follows: GP46: GP46 of bacteriophage PSA, of bacteriophage PSA, of *Streptococcus thermophilus* bacteriophage 7201, ORF40: ORF40 of bacteriophage PVL of *Staphylococcus aureus* temperate phage φSLT, of *Staphylococcus aureus* temperate phage φSLT, ORF364: ORF364 of bacteriophage φ31.1, Ec2360: b2360 of *E. coli,* AcylTr: N-Acyltransferase, Bro: Bro-N domain fused to XF0704, Met: DNA Methyltransferase](1471-2164-3-8-2){#F2} ERF defines a superfamily of SSAPs that are evolutionarily distinct from the RecT/Redβ super family --------------------------------------------------------------------------------------------------- The ERF protein of phage P22 is involved in the circularization of the linear dsDNA phage genome upon entry into the host cell \[[@B21]-[@B23]\]. Experimental studies have shown that, mutations in ERF are complemented by Redβ and that *in vitro* ERF adopts quaternary structures analogous to those of Redβ and RecT \[[@B14],[@B17],[@B24],[@B25]\]. However, in the comprehensive analysis of the RecT/Redβ superfamily no statistically significant similarity could be detected between these proteins and the ERF proteins. To explore the evolutionary affinities of the ERF domains, we carried out a sequence profile analysis as described above for the RecT case using transitive PSI-BLAST analysis. As a result of these searches, homologs of ERF encoded in several bacterial and phage genomes from diverse taxa were identified. The alignments generated in these searches consistently pointed to a region of approximately 150 amino acids that is conserved in all these proteins, with a characteristic motif of the form: GuXXoYhp + YXhXXhh (where G is glycine, Y-tyrosine, u is a tiny residue, h-hydrophobic, p is a polar residue, o is an alcohol residue, + is a basic residue, and X is any residue; Fig. [3](#F3){ref-type="fig"}). This suggested that ERF was the prototype of a family of conserved bacterial domains. {ref-type="fig"}. Species abbreviations are as follows: Bbcp32-4: *Borrelia burgdorferi* cicular plasmid (cp) 32-4, Bbcp32-7: *B. burgdorferi* cp 32-7, Bbcp32-8: *B. burgdorferi* cp 32-8, Bbcp8.3: *B. burgdorferi* cp 8.3, Bbcp9: *B burgdorferi* cp 9, bIL67: bacteriophage IL67, c2: *Lactococcus* phage c2, D3: *Pseudomonas* phage D3, HK97: bacteriophage HK97, Lj771: *Lactobacillus johnsonii* prophage Lj771, Lm: *Listeria monocytogenes,* MM1: *Streptococcus pneumoniae* bacteriophage MM1, BPmv4: Bacteriophage mv4, P22: Bacteriophage P22, phiPV83: Bacteriophage φPV83, phiSLT: *Staphylococcus aureus* temperate phage φSLT, ST7201: *Streptococcus thermophilus* bacteriophage 7201, Unk: Unknown.](1471-2164-3-8-3){#F3} Secondary structure prediction based on the multiple alignment of the ERF domain suggests a globular α + β fold with five helices and three or four strands (Fig. [3](#F3){ref-type="fig"}). The above-mentioned motif that is typical of this family is associated with helix 4 of this domain; given the presence of conserved basic residues, it may be critical for DNA-binding and strand-transfer activity of the ERF-like proteins. Additionally, in the loop between helices 4 and 5 of the ERF domain there is a universally conserved acidic motif of the form DXD. Analogous to the RecT superfamily, this acidic dyad might coordinate a divalent cation and undergo a conformational change dependent on metal-binding. However, the average size of the core domains, the patterns of conserved residues, and the predicted secondary structures of the RecT/Redβ and ERF domains show no correspondence to each other, implying that there is no direct evolutionary link between these protein groups. ERF homologs are encoded by the genomes of several temperate phages of Gram-positive bacteria and γ-proteobacteria; additionally, we detected members of this superfamily in *Listeria* and in all the circular plasmids and one linear plasmid of *Borrelia burgdorferi* (Fig. [4](#F4){ref-type="fig"}). Thus, like the RecT/Redβ superfamily, the ERF family is likely to have emerged in the temperate phages, and was disseminated to the *Borrelia* circular plasmids and some bacterial genomes via prophages. {ref-type="fig"}. The internal branches with RELL bootstrap support \>70% are indicated by blue circles. The gene neighborhoods of the ERF proteins are shown whenever they contained gene coding for proteins with potential functional relevance. Gene abbreviations are as in Fig. [1B](#F1){ref-type="fig"}](1471-2164-3-8-4){#F4} Detection of bacterial homologs of RAD52 and identification of an aberrant HhH domain in these proteins ------------------------------------------------------------------------------------------------------- The baker\'s yeast protein RAD52 and its paralog RAD59 define a small family of proteins thus far represented in fungi, vertebrates and the early-branching ameboid eukaryote, *Entamoeba histolytica.* Rad52 functions in conjunction with the RecA ortholog, the RAD51 recombinase in double-strand break repair and meiotic recombination \[[@B6]\]. RAD52 binds ssDNA during recombination and also shows a quaternary organization similar to those of RecT/Redβ and ERF \[[@B16],[@B26]\]. However, RAD52-like proteins showed no detectable sequence similarity with either the ERF or the RecT/Redβ-like proteins. Sequence searches initiated with the conserved globular region of the eukaryotic RAD52 proteins readily detected their homologs from other eukaryotes and, at convergence, also retrieved from the database certain bacterial proteins, such as DR0423 from *Deinococcus* and CAC1936 from *Clostridium* respectively, with border-like statistical significance (e \~ .05). These bacterial proteins form a small family that is additionally represented in *Salmonella paratyphi* A, the temperate bacteriophage u136 of *Lactococcus lactis* (ORF252-encoded protein) and a Shiga toxin-converting phage from *E. coli*. Iterative profile searches initiated with CAC1936 from *Clostridium acetobutylicum* and its *S. paratyphi* A ortholog correspondingly retrieved *S. cerevisiae* RAD52 and its eukaryotic homologs, with borderline e-values at convergence (\~0.043). The alignment between these bacterial proteins and the eukaryotic Rad52 homologs was co-linear throughout the entire length of their shared globular region and the Gibbs sampling procedure detected two motifs of greater than 20 residues, with a probability of chance occurrence in these proteins less than 10^-18^ (Fig. [5](#F5){ref-type="fig"}). In addition to the similar conservation pattern, separate secondary structure predictions for both the eukaryotic RAD52 family and their potential bacterial homologs showed a complete concordance of the predicted structural elements between RAD52 and the bacterial proteins, strongly suggesting that they all belong to a single homologous superfamily (hereinafter the RAD52 superfamily). {ref-type="fig"}. Species abbreviations are as follows: BPul36: Bacteriophage ul36. Cab: *Clostridium acetobutylicum,* Dr: *Deinococcus radiodurans,* Hs: Homo sapiens, Kla: *Kluyveromyces lactis,* NC: *Neurospora crassa,* Sc: *Saccharomyces cerevisiae,* Sp: *Schizosaccharomyces pombe,* Sparatyphi: *Salmonella paratyphi* A, Ralbus: *Ruminococcus albus.* The Shiga toxin-converting phage RAD52-like protein (gi: 17977996) is nearly identical to the *Salmonella paratyphi* A RAD52 like protein. The RAD52-like proteins from Bacteriophage ul36 (gi: 8248159) and *Ruminococcus albus* are respectively adjacent to genes encoding the single-strand binding protein and the λ-type exonuclease.](1471-2164-3-8-5){#F5} The secondary structure predictions showed that the Rad52 superfamily proteins adopt a structure with interspersed α-helices and β-strands (Fig. [5](#F5){ref-type="fig"}). Additionally, fold predictions using 3DPSSM (E-value=.0085, corresponding to a 90% confidence in the prediction) and the hybrid fold method (Z-score = 19.5) predicted the presence of a potential Helix-hairpin-Helix (HhH) fold in members of the RAD52 superfamily. The HhH domain is a small nucleic acid-binding module comprised of two helices joined by a central loop (hairpin), which functions as the DNA-binding moiety of numerous repair and recombination proteins\[[@B27],[@B28]\]. Two HhH modules are predicted in the core conserved domain of the RAD52 family, the first one bounded by the predicted helices 2 and 3, and the second one bounded by helices 5 and 6 (Fig. [5](#F5){ref-type="fig"}). Although these predicted HhH modules are very divergent in sequence from the typical versions, the hairpin in both HhH modules of the RAD52 family proteins is bounded by small residues, typically glycine; this conforms to the signature motif characteristic of the classical HhH modules \[[@B28],[@B29]\]. However, in the case of the RAD52 superfamily the predicted HhH modules appear to have been welded into a large globular superstructure that maintained its evolutionary distinctness over time. The conservation pattern and predicted structural elements of the RAD52 superfamily are distinct from those predicted for the ERF and RecT/Redβ superfamilies (Fig [1](#F1){ref-type="fig"}, [3](#F3){ref-type="fig"}, [5](#F5){ref-type="fig"}), supporting the lack of a direct evolutionary relationship between these proteins. The RAD52 superfamily shows a sporadic phyletic distribution, and even in the crown-group eukaryotes, might have been secondarily lost in certain lineages, such as plants, nematodes and insects. The sporadic distribution of this family among phylogenetically distant bacteria, along with its presence in several prophages, suggests that, like the RecT/Redβ and ERF superfamilies, at least the bacterial RAD52-proteins might be of predominantly phage origin. The core of the eukaryotic recombination system appears to have been inherited from the system present in the common ancestor shared with the archaea \[[@B29]\]. However, RAD52 is thus far absent in all archaeal genomes and is restricted to a single orthologous group in the eukaryotes \[[@B29]\]. Thus, it appears plausible that eukaryotic RAD52 was ultimately derived through lateral transfer either from a bacterial genome or directly from a viral source, at a point at least predating the divergence of the crown group eukaryotes and *Entamoeba.* Contextual information from gene neighborhoods provides details regarding functional interactions of the SSAPs with DNA recombination pathways ---------------------------------------------------------------------------------------------------------------------------------------------- The clustering of functionally related genes in prokaryotic genomes into co-transcribed and co-regulated units, operons, often allows functional assignments through the principle of \'guilt by association\' \[[@B30]-[@B32]\]. Generally, genes whose products physically interact to form a complex or are involved in successive steps in a biochemical pathway form operons that are conserved over large evolutionary distances \[[@B30]\]. On previous occasions, we have used gene neighborhoods or operons to predict novel DNA repair complexes and their components \[[@B33]\]. Accordingly, a similar approach was applied to the three families of SSAPs (RecT/Redβ, ERF, Rad52), to shed light on their functional links. Notably, the genes encoding the three evolutionarily distinct SSAPs co-occurred with similar sets of DNA repair/recombination-related proteins (Figs. [2](#F2){ref-type="fig"},[4](#F4){ref-type="fig"}). In at least one case, each of them was found adjacent to the gene for the single-strand-binding protein (SSB), an OB-fold protein that binds ssDNA (Figs. [2](#F2){ref-type="fig"},[4](#F4){ref-type="fig"}). This association ties in with the function of the SSAPs in single-strand annealing, suggesting that they closely interact with SSB. It has been suggested in the case of RecT that it may compete with SSB for binding single strand overhangs and thereby make them available for the annealing process \[[@B3]\]. Similar interactions between other SSAPs and SSB, that probably coats the ssDNA generated by nucleases, appear likely. Genes for SSAPs from all the 3 distinct superfamilies may also occur adjacent to or in the vicinity of genes encoding nucleases or Holliday junction resolvases (HJRs). Genes for RecT/Redβ superfamily proteins are associated with genes encoding a λ-type exonuclease (LE) of the type II restriction enzyme fold, RecE, which also might be a divergent member of this fold, and a nuclease of the Endonuclease VII (EndoVII) fold \[[@B7],[@B34]\] (Fig. [2](#F2){ref-type="fig"}). The ERF superfamily genes are associated with a RusA superfamily nuclease/HJR and EndoVII fold nucleases (Fig. [4](#F4){ref-type="fig"}) \[[@B34]\]. Furthermore, the *Borrelia* plasmids that encode ERF, also almost always additionally encode a λ-type exonuclease, even if it is not the adjacent gene. In a single instance, in the Gram-positive bacterium *Ruminococcus albus,* the gene encoding a RAD52 superfamily protein occurs adjacent to a gene for a λ-type exonuclease. These nucleases probably contribute to the repair process, in which SSAPs are involved, by providing the initial break in the dsDNA and/or in digesting the nicked target to generate ssDNA. The RecT and Redβ family proteins often co-occur with the SbcC gene that encodes an ABC ATPase with a large coiled-coil segment. These proteins are known to cooperate with SbcD, nuclease of the calcineurin-like phosphoesterase superfamily and to degrade dsDNA in the 3\' → 5\' direction generating ssDNA \[[@B35],[@B36]\]. It seems likely that RecT/Redβ proteins, at least in certain cases, function in conjunction with the SbcCD-pathway, by utilizing the single-stranded regions generated by the SbcCD nuclease. Additionally, several genes, whose functions are less clear, tend to co-occur with the genes coding for the SSAPs. These include DNA methyltransferases and the primosomal protein DnaD from low-GC Gram-positive bacteria \[[@B37]\] that co-occur with both ERF and RecT superfamily members (Figs. [2](#F2){ref-type="fig"},[4](#F4){ref-type="fig"}). The poorly characterized phage-or prophage-specific genes that are frequently observed in these neighborhoods include ORF15 (*Streptococcus thermophilus* bacteriophage 7201), ORF86, ORF100a (*Staphylococcus aureus* temperate phage φSLT) and ORF364 (bacteriophage φ31.1) (Figs. [2](#F2){ref-type="fig"},[4](#F4){ref-type="fig"}). Secondary structure predictions indicate a high α-helical content for these proteins. It is likely that these α-helical proteins are phage innovations that could function as adaptors in the recombination pathway either as accessory protein-protein interacting domains or as DNA-binding domains. Evidence for convergent operon evolution and in situ non-orthologous displacement of genes in operons encoding SSAPs -------------------------------------------------------------------------------------------------------------------- A superposition of the gene neighborhood information upon the phylogenetic trees for the SSAP superfamilies provides insights into the evolutionary processes that led to the emergence of the operons that include the SSAP genes. As discussed above, the RecT/Redβ superfamily clearly splits into three distinct families (Fig. [2](#F2){ref-type="fig"}). The phylogenetic tree shows that SbcC co-occurs with the SSAP once within Redβ-family and once within the RecT-family. An examination of the tree and the respective gene neighborhoods suggests that independent juxtaposition of SbcC with Redβ-like and RecT-like genes on two separate occasions is the most parsimonious explanation. The alternative explanation, namely that the gene coding for the common ancestor of the Redβ and RecT already co-occurred with the *sbcC* gene is far less likely because it would require over 10 independent losses of this, apparently, functionally advantageous organization in different bacterial and bacteriophage lineages. Likewise, the observation that, in one or more cases, genes encoding each of the SSAPs co-occur in the same predicted operon with SSB or a λ-type exonuclease, suggests that similar operon structures may also emerge independently in evolution. Thus, the same or analogous operon organizations may emerge convergently on multiple occasions, probably due to the selective pressure arising from the strong interactions between the SSAPs and their functional partners such as SbcC, SSB and LE. The distribution of the gene neighborhoods, in which a member of the RecT superfamily occurs next to the RecE on the phylogenetic tree of the RecT/Redβ superfamily, indicates that the RecE-RecT combination was probably the ancestral state for at least the RecT and EHAP1 families (Fig. [2](#F2){ref-type="fig"}). This implies that, on at least two occasions, the gene for λ endonuclease displaced the functionally analogous *recE* gene and became the adjacent gene to RecT (Fig. [2](#F2){ref-type="fig"}). That this displacement might have occurred by *in situ* insertion of a non-orthologous gene is suggested by the detection, on three separate occasions, of unusual remnants of pre-existing genes. The RecT/Redβ superfamily members, namely EHAP1 from the enterobacteria and PF161 from *Borrelia hermsii,* contain a small, C-terminal fragment of the core conserved domain of the ERF superfamily, which is located C-terminal of their *bona fide* RecT/Redβ domains. These fragments of the ERF protein are closely related to other ERF domains from related organisms and are unlikely to fold into the native conformation characteristic of the full-length ERF domain. For example the ERF fragment fused to the EHAP1 RecT/Redβ domain is closely related to the P22 phage ERF domain. This suggests that in each of these cases a RecT superfamily gene was inserted in frame into a pre-existing ERF gene leaving behind only a non-functional fragment of it (Fig. [2](#F2){ref-type="fig"}). In a very similar case, the bacterial RAD52-like protein from a Shiga toxin encoding temperate phage is fused to an extreme C-terminal fragment that is nearly identical to the C-terminal most portion of the P22 ERF protein. In this case, it appears that the pre-existing ERF gene was displaced through the insertion of a bacterial RAD52-like gene. Interestingly, and in the same vein, the RecT proteins from *Bacillus* species contain a short C-terminal acidic module that is missing in other RecT proteins, but is highly similar to the C-terminal region of SSBs, particularly those from Gram-positive bacteria (data not shown). This suggests that, at some stage in their evolution, the *Bacillus recT* gene protein has recombined with the gene coding for SSB, which might even have resulted in a functional replacement of an SSB with an SSAP. Thus it appears likely that functionally equivalent genes may displace their analogs in operons via insertion into the same position. Conclusions =========== We show that functionally similar SSAPs belong to at least three evolutionarily distinct superfamilies. We unify the Redβ and RecT proteins and their homologs, which have not been reported as being related at the sequence level, into a single superfamily, supporting the notion that these proteins share a similar mechanism of action. The second superfamily typified by the ERF proteins is predominantly found in bacteriophages and is also present on all circular plasmids from *Borrelia,* suggesting a role in the recombination of these plasmids. The third superfamily, typified by the yeast RAD52 protein and previously detected only in eukaryotes, was shown to include bacterial and phage homologs and to contain a modified HhH domain. By comparing the gene neighborhoods of the SSAPs, we show that the predicted operons that include the SSAP genes evolve according to the \"LEGO\" principle. In these operons, the SSAP genes are linked to the genes for various DNAses and DNA repair related proteins, such as SSB and SbcC, which implies functional connections between the encoded proteins. Evidence is presented of convergent emergence of similar SSAP-encoding operons in different lineages and of *in situ* non-orthologous displacement of functionally similar genes in these operons. Materials and Methods ===================== Sequence searches of the non-redundant (NR) and the unfinished genomes databases, were done using the gapped BLAST and PSI-BLAST programs \[[@B38]\]. Iterative PSI-BLAST searches used for in-depth sequence analysis were done with the profile inclusion cutoff expectation value (E value) set at 0.1. Multiple sequence alignments were generated using the T_Coffee program \[[@B39]\] and the output was adjusted using PSI-BLAST search results and secondary structure predictions, which were conducted using the PHD \[[@B40],[@B41]\] and Jpred \[[@B42]\] programs. Fold predictions were done using the 3-D position specific score matrix (3DPSSM) \[[@B43]\] and the Hybrid fold method \[[@B44]\]. Phylogenetic analysis was carried out using the neighbor-joining algorithm, with subsequent local rearrangements using the maximum likelihood algorithm \[[@B45]\]. The robustness of tree topology was assessed with 10000 Resampling of Estimated Log Likelihoods (RELL) bootstrap replicates. The MOLPHY and Phylip software packages were used for the analyses \[[@B46],[@B47]\]. Authors\' contributions ======================= Author 1 (LMI) contributed to the discovery process, preparation of the manuscript and multiple sequence alignments. Author 2 (EVK) contributed to the analysis of the predicted operons encoding RecT and SSB proteins from Gram positive bacteria, Author 3 (LA) contributed to the discovery process, preparation of the manuscript and conceived the study.

Background ========== *Crocus sativus* L. (Iridaceae), commonly known as saffron is used in folk medicine for various purposes such as an aphrodisiac, antispasmodic and expectorant \[[@B1]\]. Modern pharmacological studies have demonstrated that saffron extracts have antitumour \[[@B2]-[@B4]\], radical scavenger, hypolipaemic \[[@B5]\], anticonvulsant effects \[[@B6]\] and improve activity on learning and memory \[[@B5],[@B7]\]. Chemical studies on *C. sativus* have shown the presence of constituents such as crocin, crocetin, safranal and picrocrocin \[[@B8]-[@B10]\]. Among the constituents of saffron extract, crocetin is mainly responsible for these pharmacological activities \[[@B5]\]. In traditional medicine, the stigma of this plant is used as an antiedematogenic remedy \[[@B1]\]. The aim of this study was to validate the anti-inflammatory and antinociceptive effect of the saffron stigma and petal extracts. Results ======= The maximum non-fatal doses of stigma aqueous and ethanolic extracts were 0.8 g/kg and 2 g/kg (i.p.), respectively. The maximum non-fatal doses of petal aqueous and ethanolic extracts were 3.6 g/kg and 8 g/kg (i.p.), respectively. LD~50~ values of the petal aqueous and ethanolic extracts were 6.67 g/kg, i.p. (4.95, 8.99) and 9.99 g/kg, i.p. (8.13,12.28), respectively. Phytochemical screening of the extracts indicated the presence of flavonoids, tannins and anthocyanins in the petal aqueous and ethanolic extracts. Alkaloids and saponins were found in the stigma aqueous and ethanolic extracts (Table [1](#T1){ref-type="table"}) ::: {#T1 .table-wrap} ::: {.caption} ###### Phytochemical screening of *Crocus sativus* petal and stigma aqueous and ethanolic extracts. ::: **Extract** **Alkaloids** **Flavonoids** **Tannins** **Saponins** **Anthocyanins** -------------- --------------- ---------------- ------------- -------------- ------------------ **Petal** *Aqueous* \- \+ \+ \- \+ *Ethanolic* \- \+ \+ \- \+ **Stigma** *Aqueous* \+ \- \- \+ \- *Ethanolic* \+ \- \- \+ \- ::: The aqueous and ethanolic extracts of *C. sativus* petal and stigma significantly reduce the number of mouse abdominal constrictions induced by a 0.7% acetic acid solution in a dose dependent manner (P\<0.001, except for the dose of 0.36 g/kg of aqueous stigma which was P\<0.01). Morphine and diclofenac induced a protection against abdominal constriction (P\<0.001). Naloxone, (2 mg/kg, s.c.) pretreatment after i.p. injection of the extracts practically did not inhibit the antinociceptive activity of both extracts (figures [1](#F1){ref-type="fig"}, [2](#F2){ref-type="fig"}, [3](#F3){ref-type="fig"}, and [4](#F4){ref-type="fig"}). Only, the antinociceptive effect of aqueous stigma extract, 0.32 g/kg, was partially blocked by naloxone (figure [3](#F3){ref-type="fig"}). Naloxone completely antagonized the antinociceptive activity of morphine. {#F1} {#F2} {#F3} {#F4} In hot plate test, aqueous (1.4, 2.5 and 3.6 g/kg, i.p.) and ethanolic (0.8, 3.2 and 5.6 g/kg, i.p.) petal extracts showed no significant antinociceptive activity (P\>0.05)(figures [5](#F5){ref-type="fig"} and [6](#F6){ref-type="fig"}). The aqueous (0.32, 0.56 and 0.8 g/kg, i.p.) and ethanolic (0.8, 1.4 and 2 g/kg, i.p.) stigma extracts also exerted no significant analgesic activity (P\>0.05)(figures [7](#F7){ref-type="fig"} and [8](#F8){ref-type="fig"}). Morphine, a positive reference, (10 mg/kg, i.p.) showed significant analgesic effect in the hot-plate test beginning 30 min after treatment (P\<0.001). In the xylene induced ear edema, the aqueous and ethanolic extracts of petal showed no significant anti-inflammatory activity but diclofenac and dexamethasone reduced the edema about 50% (Tables [2](#T2){ref-type="table"} and [3](#T3){ref-type="table"}). In higher doses, the aqueous (0.56 g/kg, P\<0.05; 0.8 g/kg, P\<0.01) and ethanolic (1.4 g/kg, P\<0.05; 2 g/kg, P\<0.05) extracts of stigma showed significant activity against the acute inflammation (Tables [4](#T4){ref-type="table"} and [5](#T5){ref-type="table"}). {#F5} {#F6} {#F7} {#F8} ::: {#T2 .table-wrap} ::: {.caption} ###### Effect of the intraperitoneal doses of *Crocus sativus* petal aqueous extract, diclofenac and dexamethasone on xylene-induced ear swelling in mice. ::: Treatment Dose Ear swelling (mg) Inhibition (%) --------------- ---------- ------------------- ---------------- Control 10 ml/kg 7.6 ± 0.6 -- Diclofenac 15 mg/kg 3.9 ± 0.4^\*\*\*^ 49.1 Dexamethasone 15 mg/kg 3.2 ± 0.3^\*\*\*^ 58.2 Extract 1.4 g/kg 6.8 ± 0.3 11.7 Extract 2.5 g/kg 6.4 ± 0.4 15.9 Extract 3.6 g/kg 6.3 ± 0.3 17.2 The increase in weight caused by the irritant (xylene) was measured by subtracting the weight of the untreated left ear section from that of the treated right ear sections. Values are the mean ± S.E.M. for 7 mice, ^\*\*\*^P\<0.001, compared to control (normal saline), Tukey-Kramer. ::: ::: {#T3 .table-wrap} ::: {.caption} ###### Effect of the intraperitoneal doses of *Crocus sativus* petal ethanolic extract, diclofenac and dexamethasone on xylene-induced ear swelling in mice. ::: Treatment Dose Ear swelling (mg) Inhibition (%) --------------- ---------- ------------------- ---------------- Control 10 ml/kg 6.8+0.5 -- Diclofenac 15 mg/kg 3.7 ± 0.6^\*\*^ 45.3 Dexamethasone 15 mg/kg 3.3 ± 0.5^\*\*^ 51.4 Extract 0.8 g/kg 5.5 ± 0.6 19.2 Extract 1.6 g/kg 5.5 ± 0.5 19.4 Extract 2.5 g/kg 5.1 ± 0.6 25.8 Extract 3.2 g/kg 4.8 ± 0.6 29.0 The increase in weight caused by the irritant (xylene) was measured by subtracting the weight of the untreated left ear section from that of the treated right ear sections. Values are the mean ± S.E.M. for 7 mice, ^\*\*^P\<0.01, compared to control (normal saline), Tukey-Kramer. ::: ::: {#T4 .table-wrap} ::: {.caption} ###### Effect of the intraperitoneal doses *of Crocus sativus* stigma aqueous extract, diclofenac and dexamethasone on xylene-induced ear swelling in mice. ::: Treatment Dose Ear swelling (mg) Inhibition (%) --------------- ----------- ------------------- ---------------- Control 10 ml/kg 7.6 ± 0.6 -- Diclofenac 15 mg/kg 3.9 ± 0.4^\*\*\*^ 53.6 Dexamethasone 15 mg/kg 32+0 3^\*\*\*^ 63.1 Extract 0.32 g/kg 5.0 ± 0.4 20.9 Extract 0.56 g/kg 4.3 ± 0.4^\*^ 32.6 Extract 0.8 g/kg 3.5 ± 0.5^\*\*^ 44.1 The increase in weight caused by the irritant (xylene) was measured by subtracting the weight of the untreated left ear section from that of the treated right ear sections. Values are the mean ± S.E.M. for 7 mice, ^\*^P\<0.05, ^\*\*^P\<0.01, ^\*\*\*^P\<0.001, compared to control (normal saline), Tukey-Kramer. ::: ::: {#T5 .table-wrap} ::: {.caption} ###### Effect of the intraperitoneal doses of Crocus sativus stigma ethanolic extract on xylene-induced ear swelling in mice. ::: Treatment Dose Ear swelling (mg) Inhibition (%) --------------- ---------- ------------------- ---------------- Control 10 ml/kg 6.2 ± 0.3 -- Diclofenac 15 mg/kg 2.6 ± 0.4^\*\*\*^ 58.3 Dexamethasone 15 mg/kg 3.1 ± 0.4^\*\*\*^ 49.5 Extract 0.2 g/kg 5.1 ± 0.3 18.2 Extract 0.8 g/kg 4.8 ± 0.3 22.4 Extract 1.4 g/kg 4.4 ± 0.4^\*^ 29.3 Extract 2.0 g/kg 4.5 ± 0.6^\*^ 27.8 The increase in weight caused by the irritant (xylene) was measured by subtracting the weight of the untreated left ear section from that of the treated right ear sections. Values are the mean ± S.E.M. for 7 mice, ^\*^P\<0.05, ^\*\*\*^P\<0.001, compared to control (normal saline), Tukey-Kramer. ::: In the chronic inflammation (formalin test), the aqueous petal extracts did not exhibit significant anti-inflammatory activity (P\>0.05) but the ethanolic petal extract showed significant activity (figure [9](#F9){ref-type="fig"}). Both extracts of the stigma exerted significant anti-inflammatory activity (figure [10](#F10){ref-type="fig"}). In diclofenac and the ethanolic stigma extract (1.4 g/kg) groups; the hind paw edema of rat disappeared after 6 days (P\<0.001) (figure [10](#F10){ref-type="fig"}). Diclofenac as well as the aqueous (0.8 g/kg) and ethanolic stigma extract (1.4 g/kg) showed anti-inflammatory effect on day one (P\<0.05) (figure [10](#F10){ref-type="fig"}). On days 4 and 5, both diclofenac and the extracts did not demonstrate anti-inflammatory activity (P\>0.05). {#F9} {#F10} Discussion ========== The present results indicate that aqueous and ethanolic extracts of *C. sativus* petal and stigma have antinociceptive activity in chemical pain tests. The stigma extracts showed activity against acute and chronic inflammation. The petal ethanolic extract showed anti-inflammatory activity in the chronic inflammatory test. In respect to LD~50~ values and maximum non-fatal doses, the stigma extracts were more toxic than the petal extracts. LD~50~ values of aqueous and ethanolic stigma extracts have been reported as 1.6 g/kg, i.p. (1.16, 2.22) and 3.38 g/kg, i.p. (2.55, 4.52) in mice \[[@B6]\]. According to a toxicity classification \[[@B11]\], stigma and petal extracts are \"relatively toxic\" and \"low toxic\", respectively. Treatment with the *C. sativus* extract also significantly prolonged the life span of cisplatin-treated mice almost three-fold \[[@B12]\]. Toxicity studies showed that the hematological and biochemical parameters were within normal range with saffron extract \[[@B2]\]. As preliminary phytochemical results indicated, it could be suggested that the antinociceptive and anti-inflammatory effects of the petal extracts may be due to their content of flavonoids, tannins and anthocyanins. Other studies have demonstrated that various flavonoids such as rutin, quercetin, luteolin, hesperidin and biflavonoids produced significant antinociceptive and/or anti-inflammatory activities \[[@B13]-[@B16]\]. There are few reports on the role of tannins in antinociceptive and anti-inflammatory activities \[[@B17]\]. Recently, it has been shown that crocins, *Crocus* glycosides, exhibited an anti-inflammatory effect in some models of inflammation \[[@B18]\]. In the present study, morphine, a centrally acting analgesic drug, produced an inhibitory effect on the nociceptive response in the hot plate test, a central antinociceptive test \[[@B19]\], while the aqueous and ethanolic extracts showed no antinociceptive activity in this test. Thus, the extracts may not act via central mechanisms, although agents that alter the motor performance of animals may increase the latency time on the hot plate test without acting on the central nervous system \[[@B19]\]. Antinociceptive activity of opioid agonists, opioid partial agonists, and non-steroidal anti-inflammatory agents can be determined by the writhing test \[[@B20]\]. It has also been shown that some plants such as *Hunteria zeylanica*\[[@B21]\] and *Ocotea suaveolens*\[[@B22]\] decrease stretching induced by acid acetic acid but have no effect on heat-induced pain. The association of both antinociceptive activity and moderate anti-inflammatory effect observed with the extracts has also been shown in non-steroidal anti-inflammatory drugs (NSAIDs). It is a well-established fact that NSAIDs exert their analgesic and anti-inflammatory activity by the inhibition of cyclo-oxygenase activity \[[@B23]\]. As the antinociceptive activity of most extracts in the writhing test was not inhibited by naloxone, the extracts may not act via opioid receptors and may exert their activity via a peripheral mechanism. The stigma extracts and the ethanolic petal extracts significantly diminished in a dose dependent way the induced paw edema in rats. The extracts were able to reduce the inflammation in acute and chronic phases. It is concluded that saffron stigma and petal aqueous and ethanolic maceration extracts have antinociceptive effects in chemical pain tests and have acute and/or chronic anti-inflammatory activity. The antinociceptive and anti-inflammatory effects of the extracts may be due to their content of flavonoids, tannins, anthocyanins, alkaloids and saponins. However, the chemical constituents and mechanism(s) responsible for the pharmacologoical activities remain to be investigated. Materials and Methods ===================== Animals ------- Male and female albino mice 25--30 g and Wistar rats weighing 150--210 g were obtained from a random bred colony in the animal house of Mashhad University of Medical Sciences. Animals were housed in colony room 12/12 hr light/dark cycle at 21 ± 2°C and had free access to water and food. Plant material -------------- Plants were collected from Torabat Hydarieh (in south of Khorassan province, I.R. of Iran) in October 1998 and dried in shadow and ground. The *C. sativus* L. was identified by Ferdowsi University (Ms. Molaei) and voucher samples were preserved for reference in the herbarium of School of Pharmacy, Mashhad, IR. Iran (143-0319-1). Preparation of extracts ----------------------- The powder of stigma or petal was extracted using maceration with ethanol or water. The powdered plant was macerated in water or ethanol (80 %, v/v) for 3 days and, subsequently, the mixture was filtered and concentrated under reduced pressure at 40°C. The yield (w/w) of aqueous and ethanolic extracts of stigma was 50.8% and 56.6%, respectively. The yield (w/w) of aqueous and ethanolic extracts of petal was 15.5% and 19%, respectively. The extracts were dissolved in normal saline. Phytochemical screening ----------------------- Phytochemical screening of the extract was performed using the following reagents and chemicals: Alkaloids with Dragendorffs reagent, flavonoids with the use of Mg and HCl; tannins with 1% gelatin and 10% NaCl solutions and saponins with ability to produce suds \[[@B24]\]. The maximum non-fatal dose (MNFD) and acute toxicity ---------------------------------------------------- Different doses of the extracts were injected to separate groups of five animals. After 48 h, the highest dose that did not induce any mortality was considered as the maximum non-fatal dose. The number of deaths was counted at 48 h after treatment. LD~50~ values and the corresponding confidence limits (CL, 95%) were determined by the Litchfield and Wilcoxon method (PHARM/PCS Version 4). Doses of 10, 40 (50), 70 and 100 % MNFD were chosen for most tests. Lower doses were found to be effective in the writhing test. Antinociceptive study --------------------- ### Hot-plate test The hot-plate test was assessed on groups of 8 mice. The temperature of a metal surface was maintained at 55 *±* 0.2°C. Latency to a discomfort reaction (licking paws or jumping) was determined before and after drug administration. The cut-off time was 20 s. The latency was recorded before and 30, 60, 120, 150 and 240 min following intraperitoneal administration of the agents. The prolongation of the latency times compared with the values of the control was used for statistical comparison. Control received normal saline (10 ml/kg, i.p.) and morphine (10 mg/kg, i.p.) was used as reference drug \[[@B25]\]. ### Writhing test Groups of 8 mice were used for controls and test mice. One hour after the administration of the extract, the mice were given an intraperitoneal injection of 0.7% v/v acetic acid solution (volume of injection 0.1 ml/10 g). The mice were placed individually into glass beakers and five min were allowed to elapse. The number of writhes produced in these animals was counted for 30 min. For scoring purposes, a writhe is indicated by stretching of the abdomen with simultaneous stretching of at least one hind limb. Control received normal saline (10 ml/kg, i.p.), diclofenac (10 mg/kg, i.p.) and morphine (10 mg/kg, i.p.) were used as reference drugs. Naloxone (2 mg/kg, s.c.) was administered 15 min prior to the extracts or morphine injections \[[@B25]\]. Anti-inflammatory study ----------------------- ### Xylene-induced ear edema Mice were divided into groups of seven. Thirty minutes after i.p. injection of the extract, diclofenac and dexamethasone, 0.03 ml of xylene was applied to the anterior and posterior surfaces of the right ear. The left ear was considered as control. Two hours after xylene application, mice were killed and both ears were removed. Circular sections were taken, using a cork borer with a diameter of 7 mm, and weighed. The increase in weight caused by the irritant was measured by subtracting the weight of the untreated left ear section from that of the treated right ear sections. The formula for computing percent inhibition was: average writhes in the control group (normal saline) minus writhes in the drug group divided by writhes in the control group times 100%. Control received normal saline (10 ml/kg, i.p.), diclofenac (10 mg/kg, i.p.) and dexamethasone (15 mg/kg, i.p.) were used as reference drugs \[[@B25]\]. Formalin induced inflammation ----------------------------- Rats were divided into groups of six. The inflammation was produced by subaponeurotic injection of 0.1 ml of 2% formaldehyde in the right hind paw of the rats on the first and third day. The animals were treated daily with the extracts or diclofenac intraperitoneally for 10 days. The daily changes in paw size were measured by wrapping a piece of cotton thread round the paw and measuring the circumference with a meter rule \[[@B26]\]. Materials ========= The following reagents were used: morphine and dexamethasone (Dam Pakhsh, I.R. Iran), naloxone hydrochloride (Tolid Daru, I.R. Iran), diclofenac (Zahrawi, I.R. Iran), acetic acid, xylene and formaldehyde (Merck). Statistical analysis -------------------- The data were expressed as mean values ± S.E.M. and tested with analysis of variance followed by the multiple comparison test of Tukey-Kramer. Acknowledgments =============== The authors are thankful to T.W. Stone .Professor of Pharmacology, Division of Neuroscience and Biomedical Systems, University of Glasgow, Scotland, for his linguistic help.

Background ========== Differentiation is commonly associated with intermingled changes in gene expression and cellular proliferation. In some differentiating cell types, changes in both gene transcription and proliferation are regulated by the same transcription factor \[[@B1]-[@B9]\]. CCAAT/Enhancer Binding Protein alpha (C/EBPα) is a transcription factor that is required for the differentiation of a number of tissues \[[@B10]-[@B18]\]. Mice homozygous for C/EBPα null alleles have severe defects in tissues involved in metabolic homeostasis \[[@B19]-[@B21]\]. Cellular proliferation is elevated in the liver of these knockout mice \[[@B22]\] suggesting that C/EBPα blocks proliferation *in vivo.* In cultured cells, C/EBPα expression leads to decreased colony formation upon antibiotic selection \[[@B23]-[@B25]\], decreased DNA synthesis \[[@B22],[@B24]-[@B27]\] and an enhanced proportion of cells in the G1 phase of the cell cycle \[[@B26]\]. Thus, C/EBPα regulates cellular proliferation, as well as gene transcription. C/EBPα contains a bZIP domain conserved at the carboxy terminus of a number of transcription factors \[[@B12],[@B28]\]. The bZIP domain consists of a basic region that binds directly to DNA, followed immediately by a leucine zipper. C/EBPα dimerizes via the leucine zipper. This dimerization is required for DNA binding \[[@B28],[@B29]\]. At least three transcription activation functions have been described in the more amino terminal regions of C/EBPα \[[@B30]-[@B32]\]. C/EBPα domains and activities associated with proliferation arrest also have been identified, but vary considerably between studies \[[@B23],[@B25],[@B33]-[@B36]\]. C/EBPα binds to and activates transcription of the gene promoter for the p21 inhibitor of cyclin-dependent kinase (CDK) \[[@B37]\]. This led to speculation that p21 gene activation may contribute to cell cycle arrest by C/EBPα \[[@B37]\]. Similarly, suppression of mitotic growth during adipocyte differentiation was associated with C/EBPα activation of the promoters of gadd45 (growth arrest and DNA damage-inducible gene 45), gas2 and gas3 (growth arrest-associated genes 2 and 3) \[[@B5],[@B6],[@B38],[@B39]\]. However, C/EBPα mutants defective in DNA binding still blocked proliferation \[[@B25],[@B33]\]. This suggested that direct promoter activation was unnecessary, or redundant, for C/EBPα proliferation arrest \[[@B18]\]. Possible mechanisms of transcription-independent proliferation arrest by C/EBPα have been suggested by a number of studies. Decreased proliferation was associated with C/EBPα stabilization of the p21 protein \[[@B22],[@B24]\]. p21 interacted directly with a large internal segment of C/EBPα that included transcription activation domain 3 \[[@B25]\] (see Fig. [1A](#F1){ref-type="fig"}). CDK2 and CDK4 also interacted with segments of C/EBPα close to, and within, transcription activation domain 3 \[[@B36]\]. CDK2 also interacted with the basic region of the C/EBPα \[[@B25]\]. p21 also has a second interaction site, within the leucine zipper of C/EBPα \[[@B25]\]. *in vitro,* C/EBPα enhanced p21 inhibition of CDK2 activity. C/EBPα inhibition of CDK2 activity correlated with p21 binding to C/EBPα transcription activation domain 3 \[[@B25]\]. However, proliferation arrest by C/EBPα still occurred in cell lines not containing p21 genes \[[@B40]\]. This indicated that proliferation arrest by C/EBPα did not rely solely upon C/EBPα enhancement of CDK inhibition by p21. ![**A,** Positions of the transcription activation (TA), DNA binding (basic) and dimerization (ZIP) domains along the linear sequence of C/EBPα. The numbers below the C/EBPα diagram indicate the amino acid positions at the boundaries of the domains. Also shown are C/EBPα sequences involved in known actions with the indicated cell-cycle proteins: p107 \[[@B34]\], p21 \[[@B25]\], CDK2 \[[@B25],[@B36]\], CDK4 \[[@B36]\] and E2F \[[@B33]\]. **B,** Schematic of the GFP fusions with full-length and mutant C/EBPα used in the studies reported here.](1471-2121-3-6-1){#F1} Another mechanism reported for proliferation blockage by C/EBPα involves the E2F-DP1 transcription complexes. E2F complexes activate genes required for entry into S phase. C/EBPα binds to and inhibits transcriptional activation by E2F \[[@B26],[@B33]\]. The complex of E2F with the retinoblastoma-related p107 protein is prevalent in cycling cells. Transcription activation domain 2 of C/EBPα was observed to interact specifically with p107 to disrupt p107/E2F complex formation \[[@B34],[@B35]\]. Disruption of E2F activity also was associated the basic domain of C/EBPα \[[@B33],[@B34]\]. Thus, a variety of interactions and mechanisms are potentially involved in proliferation arrest by C/EBPα. This variability may indicate divergent mechanisms for the block of proliferation by C/EBPα in different study conditions and/or cell types. In pituitary cells, C/EBPα is part of a multi-subunit complex \[[@B41],[@B42]\] that activates the growth hormone \[[@B42]\] and prolactin \[[@B43],[@B44]\] promoters. Full-length C/EBPα is absent from the pituitary progenitor GHFT1-5 cell line \[[@B42],[@B45]\] in which ectopically expressed C/EBPα activates co-transfected growth hormone and prolactin promoters \[[@B42],[@B44],[@B45]\]. We show here that ectopically expressed C/EBPα prolongs GHFT1-5 cells in the G1 and S phases of the cell cycle in a transcription-independent fashion. Different C/EBPα domains were required for G1- and S-phase arrest. C/EBPα domains previously described for the interaction with p21 were not required for arrest in GHFT1-5 cells. Rather, domains previously implicated in the regulation of the E2F and CDK2 pathways by C/EBPα were required for arrest of these progenitor cells. Results ======= Transcriptionally active and inactive GFP fusions with C/EBPα ------------------------------------------------------------- The domains of C/EBPα and their known interactions with cell-cycle proteins are summarized in figure [1A](#F1){ref-type="fig"}. We conducted studies to determine which, if any, of those activities affect the proliferation of GHFT1-5 pituitary progenitor cells. In prior studies, we noticed that fewer cells were present in GHFT1-5 cell cultures transfected with the C/EBPα expression vector than in cultures transfected with control vectors (X. Wang, F. S., unpublished data). The decreased cell number might be due to impaired proliferation of GHFT1-5 cells expressing C/EBPα. To examine if C/EBPα affected the progression of GHFT1-5 cells through the cell cycle, we needed to distinguish transfected cells that expressed C/EBPα from those that did not. We fused the cDNA for the green fluorescence protein (GFP) to either the 5\' (GFP-C/EBPα) or the 3\' (C/EBPα-GFP) end of the coding sequence of the C/EBPα cDNA. The cDNAs for the GFP-C/EBPα and C/EBPα-GFP fusion proteins were inserted into an expression vector and transfected into GHFT1-5 cells. Flow cytometry was used to specifically identify green fluorescent, C/EBPα-expressing cells. We initially characterized the abilities of the C/EBPα-GFP and GFP-C/EBPα fusion proteins to bind DNA and activate transcription in GHFT1-5 cells (Fig. [2](#F2){ref-type="fig"}). The C/EBPα-GFP or GFP-C/EBPα expression vectors were transfected into GHFT1-5 cells with a promoter consisting of a single C/EBPα binding site upstream of the growth hormone TATA box. This minimal promoter was specifically responsive to C/EBPα expression in GHFT1-5 cells \[[@B45],[@B46]\]. Cells transfected with the C/EBPα-GFP expression vector showed a statistically significant (p\<0.05, n = 5) 9.66 +/- 6.08-fold higher promoter activity than did cells sham-transfected with the same expression vector deleted of the C/EBPα cDNA (Fig. [2A](#F2){ref-type="fig"}). In contrast, activation by GFP-C/EBPα was a statistically insignificant 1.88 +/- 1.48-fold (Fig. [2A](#F2){ref-type="fig"}). Promoter activation by C/EBPα-GFP was reproducibly less than promoter activation by unfused C/EBPα. Western blots of nuclear extracts from the transfected cells showed that unfused C/EBPα was expressed at marginally higher levels (Fig. [2B](#F2){ref-type="fig"}). ![**A,** Chloramphenicol acetyltransferase activity expressed from a promoter transiently transfected into GHFT1-5 cells together with vectors expressing C/EBPα, fused or not, with GFP. The promoter contained a single C/EBPα binding site upstream of a TATA box \[[@B45]\]. CAT activities were normalized to the activity present in cells transfected with the expression vector not containing the C/EBPα cDNA (\"Sham\") and plotted as the mean +/- standard deviation from five independent experiments. **B,** Nuclear extracts from sham-transfected cells and cells expressing C/EBPα, C/EBPα-GFP and GFP-C/EBPα were separated by SDS-polyacrylamide gel electrophoresis. The separated proteins were transferred to a membrane and stained with an antibody directed against the FLAG epitope, which was appended to the amino terminus of C/EBPα in all the constructs. Arrow, expressed C/EBPα-GFP and GFP-C/EBPα. **C,** Whole cell extracts, from cells transfected with C/EBPα-GFP and GFP-C/EBPα or GFP-C/EBPαΔLZ expression vectors, were incubated with a radiolabeled oligonucleotide containing a high affinity consensus C/EBPα binding site. The observed complexes were competed with a 1, 10 and 100 fold molar excess of unlabeled oligonucleotide or were supershifted with an antibody directed against C/EBPα.](1471-2121-3-6-2){#F2} C/EBPα-GFP and GFP-C/EBPα both were expressed as intact, full-length proteins of the appropriate molecular weight \[[@B45]\](Fig. [2B](#F2){ref-type="fig"}, arrow). In addition, both C/EBPα-GFP and GFP-C/EBPα bound similarly to DNA (Fig. [2C](#F2){ref-type="fig"}). Extracts of cells expressing either C/EBPα-GFP or GFP-C/EBPα shifted the electrophoretic mobility of a radiolabeled oligonucleotide containing a C/EBPα binding site (Fig. [2C](#F2){ref-type="fig"}). No shift was observed with sham-transfected cells (not shown). The shifted bands were specifically competed by a 100-fold molar excess of the cold oligonucleotide and were supershifted with an antibody directed against C/EBPα. GFP-C/EBPα truncated of its leucine zipper (Fig. [1B](#F1){ref-type="fig"}) was ineffective at shifting the binding site (Fig. [2C](#F2){ref-type="fig"}, ΔLZ). Flow cytometry (discussed later) showed that comparable levels of green fluorescence were emitted from cells expressing C/EBPα-GFP, GFP-C/EBPα and ΔLZ. Thus, C/EBPα-GFP and GFP-C/EBPα were expressed and could bind to C/EBPα binding sites in DNA, but differed in their ability to activate gene transcription. Cell cycle distribution of GHFT1-5 cells ---------------------------------------- Transcriptionally active C/EBPα-GFP and transcriptionally inactive GFP-C/EBPα were compared for their effects on proliferation of GHFT1-5 cells. We first determined the distribution of untransfected GHFT1-5 cells in each phase of the cell cycle. GHFT1-5 cells were grown to subconfluence, collected and the DNA within the cells was stained with propidium iodide. The DNA content within each cell was quantified by flow cytometry as the amount of orange fluorescence from the propidium iodide-stained DNA (see Materials and Methods for details). The fluorescence intensity measured from each cell falls into one of three populations: cells centered around the lowest level of orange fluorescence, cells centered around the highest level of orange fluorescence which is twice that of the lowest, and cells with intermediate levels of orange fluorescence. This corresponds, respectively, to cells containing a 2n DNA complement prior to duplication of the genome (in G1 phase), cells containing a 4n DNA complement following genome duplication (in growth phase 2 or in mitosis, G2/M), and cells containing a partially replicated genome intermediate between 2n and 4n (in S phase). As averaged from nine independent experiments, the proportion of growing GHFT1-5 cells in G1, S and G2/M corresponded to 42%, 38% and 20%, respectively (Fig [3](#F3){ref-type="fig"}, white bars). {#F3} We next synchronized GHFT1-5 cells in particular stages of the cell cycle. GHFT1-5 cells were treated for 20 hours with nocodazole, which prevents microtubule polymerization/depolymerization and exit from M phase \[[@B47]\]. The delivery vehicle, DMSO, was added to parallel cell cultures. Cells also were treated with mimosine (in DMSO), which blocks cells in late G1 immediately before entry into S phase \[[@B48]\]. Mimosine also prolongs some cell types in the S phase itself \[[@B48]\]. GHFT1-5 cells grow with an average doubling time of approximately 18 to 24 hours. 20 hours of treatment allowed most cells to pass through one cell cycle and to accumulate as 4n cells upon nocodazole treatment (Fig. [3](#F3){ref-type="fig"}, light gray bars) or as 2n cells at the G1/S boundary upon mimosine treatment (Fig. [3](#F3){ref-type="fig"}, stippled gray bars). Thus, GHFT1-5 cells can be synchronized at particular stages of the cell cycle. Synchronization facilitated characterization of the cell cycle stages blocked or prolonged by C/EBPα expression. Transcriptionally active and inactive C/EBPα both retain GHFT1-5 cells in the G1 and S phases of the cell cycle --------------------------------------------------------------------------------------------------------------- We determined if either C/EBPα-GFP or GFP-C/EBPα altered the distribution of GHFT1-5 cells in the G1, S or G2/M phases of the cell cycle. GHFT1-5 cells were transfected with the expression vectors for C/EBPα-GFP or GFP-C/EBPα, or with the sham expression vector not containing the C/EBPα cDNA. The cells were incubated for one day to allow time for C/EBPα expression. Cells were then synchronized in M-phase or in G1/S by 20-hour incubations with nocodazole (Fig. [4A](#F4){ref-type="fig"}) or mimosine (Fig. [4B](#F4){ref-type="fig"}). Cells were collected and stained with propidium iodide. Flow cytometry was used to measure DNA content in 1) cells transfected with the sham expression vector (Fig. [4](#F4){ref-type="fig"}, Sham), 2) green fluorescent cells expressing GFP-tagged C/EBPα (Fig. [4](#F4){ref-type="fig"}, C/EBPα-GFP or GFP-C/EBPα) or 3) the subpopulation of cells from the C/EBPα-GFP or GFP-C/EBPα transfections that did not express measurable amounts of green fluorescent C/EBPα (Fig. [4](#F4){ref-type="fig"}, No C/EBP). The \"No C/EBP\" cells represent an internal control for cells not expressing C/EBPα collected simultaneously with cells expressing C/EBPα. The No C/EBP controls also indicated the extent to which green fluorescent, C/EBPα-expressing cells were distinguished from non-expressing cells by flow cytometry. For all our experiments, the distribution of cells in the G1, S and G2/M phases was similar for the \"No C/EBPα \" and sham-transfected cells. {#F4} The expression of C/EBPα-GFP or GFP-C/EBPα resulted in a statistically significant decrease in the amount of cells blocked in G2/M upon nocodazole treatment compared to sham-transfected cells (Fig. [4A](#F4){ref-type="fig"}) (p \<\< 0.01, n = 6). This indicated that C/EBPα expression prevented a significant proportion of GHFT1-5 cells from reaching G2/M to be blocked by nocodazole. When cells were blocked in G1 by mimosine treatment, C/EBPα expression caused no decrease in the proportion of cells in G2/M (Fig. [4B](#F4){ref-type="fig"}). This demonstrated that C/EBPα expression did not actively decrease the proportion of cells in G2/M. The C/EBPα-induced reduction in cells blocked in G2/M was associated with a highly statistically significant increase in the proportion of GHFT1-5 pituitary progenitor cells in G1 (p \< 0.001) and a statistically significant increase in the proportion of cells in S phase (p \< 0.03) (Fig. [4A](#F4){ref-type="fig"}). This retention of GHFT1-5 cells in the G1 and S phases of the cell cycle occurred for both the transcriptionally active C/EBPα-GFP and inactive GFP-C/EBPα fusions. The independence of C/EBPα-induced proliferation arrest from C/EBPα-regulated transcriptional activation also has been observed in other studies \[[@B25],[@B26]\]. Thus, simple transcription activation mechanisms do not account for the effect of C/EBPα on the cell cycle. Targeting of C/EBPα to the presumed centromeric G1/S checkpoint is insufficient for prolongation of G1 ------------------------------------------------------------------------------------------------------ Our prior fluorescence microscopy studies showed that C/EBPα-GFP, GFP-C/EBPα, and antibody-stained C/EBPα \[[@B45],[@B46]\], concentrated at specific intranuclear domains in GHFT1-5 cells (see Fig. [5A](#F5){ref-type="fig"}, GFP-C/EBPα). These domains coincide with regions that stain with the blue fluorescent DNA-binding dye Hoechst 33342 (Fig. [5A](#F5){ref-type="fig"}, Hoechst 33342). Hoechst 33342 stains AT-rich DNA that concentrates around the centromeres \[[@B45],[@B49]\]. Given the role of the centromere as a checkpoint for regulation of the mitotic phase of the cell cycle \[[@B50],[@B51]\], it was hypothesized that C/EBPα localization around the centromere could play a role in C/EBPα regulation of the cell cycle \[[@B49]\]. {ref-type="fig"}. The transfected cells were counter-stained with the blue fluorescent, DNA-binding dye Hoechst 33342 immediately before analysis by fluorescence microscopy. Green fluorescent (left panels) and blue fluorescent (right panels) images were collected from a single cell using filter sets that selectively discriminate GFP and Hoechst 33342 fluorescence (see Materials and Methods). The subnuclear position of GFP-C/EBPα and ΔLZ were compared to the subnuclear position of the peri-centromeric chromatin stained by Hoechst 33342. Representative images are shown.](1471-2121-3-6-5){#F5} To investigate if targeting of C/EBPα to the peri-centromeric chromatin is necessary or sufficient to cause G1 and S phase prolongation in GHFT1-5 cells, we first determined the domain of C/EBPα required to target C/EBPα to the peri-centromeric chromatin. It was previously suggested that the peri-centromeric targeting of C/EBPα was a function of DNA binding: C/EBPα binding sites are concentrated within the repeated DNA sequences that comprise the bulk of peri-centromeric chromatin \[[@B49]\]. Indeed, C/EBPα truncated to contain little more than the DNA binding, bZIP domain still concentrated at peri-centromeric chromatin \[[@B45]\]. By contrast, C/EBPα deleted of amino acids 310 to 358 (Fig. [1B](#F1){ref-type="fig"}, ΔLZ) no longer targeted to the peri-centromeric DNA (Fig. [5B](#F5){ref-type="fig"}). This deletion disrupts DNA binding (Fig. [2C](#F2){ref-type="fig"}) by eliminating the leucine zipper required for DNA binding by all members of the bZIP family of transcription factors \[[@B52]\]. Thus, as predicted \[[@B49]\], the bZIP domain is necessary and sufficient for C/EBPα targeting to peri-centromeric chromatin. When expressed in GHFT1-5 cells, ΔLZ was as effective as full-length GFP-C/EBPα in prolonging both G1 and S (Fig. [6A](#F6){ref-type="fig"}). Thus, leucine zipper dimerization, site-specific DNA binding and targeting of C/EBPα to peri-centromeric were not required for C/EBPα prolongation in G1 and S. In contrast, the isolated C/EBPα bZIP DNA binding domain (Fig. [1B](#F1){ref-type="fig"}, DBD), which still targeted to peri-centromeric chromatin, did not prolong G1 (Fig. [6B](#F6){ref-type="fig"}). S phase remained prolonged upon DBD expression. The different effects of the isolated DBD on G1 and S blockage indicated that C/EBPα regulation of G1 and S phase arrest was mechanistically distinct. Thus, C/EBPα regulates proliferation by two distinct pathways. Both pathways do not depend upon site-specific DNA binding, which is commonly considered a prerequisite for gene-specific transcription. {#F6} Prolongation in G1 requires the amino terminal domain of C/EBPα --------------------------------------------------------------- The C/EBPα DBD therefore was insufficient for prolongation of G1 (Fig. [6B](#F6){ref-type="fig"}). We next determined which additional domains of C/EBPα were required for G1 prolongation. C/EBPα amino acids 1--154 or 154--257 were appended to the DBD (see Fig. [1B](#F1){ref-type="fig"}). The addition of amino acids 1--154 (Fig. [7](#F7){ref-type="fig"}, DBD + 1--154) caused a statistically significant increase, relative to sham-transfected cells, in the proportion of cells in G1 (p \<\< 0.01, n = 6). The proportion of cells in G1 was not statistically different in cells expressing DBD + 1--154 from that in cells expressing full-length C/EBPα-GFP. Adding amino acids 1--125, instead of amino acids 1--154, caused a similar prolongation in G1 (Fig. [7](#F7){ref-type="fig"}, DBD + 1--125). Thus, an element in the amino terminal domain of C/EBPα was required for G1 prolongation. {ref-type="fig"}) and expressed in GHFT1-5 cells. The proportions of green fluorescent cells in G1, S and M phase in nocodazole-treated cells were determined and plotted as the mean +/- standard deviation from three independent experiments. The DBD + 1--154 mutant was repeated a total six times (not shown), for which the statistical significance remained p \< 0.01. Statistically significant increases in the proportion of G1 cells, compared to Sham cells, are indicated (One-way ANOVA: ^\*\*^, p \< 0.01; ^\*^p \< 0.05; -, no difference).](1471-2121-3-6-7){#F7} In contrast, addition of amino acids 154--257 to C/EBPα DBD did not rescue G1 prolongation (Fig. [7](#F7){ref-type="fig"}, DBD + 154--257). Again, appending amino acids 1--125 to this mutant restored G1 prolongation (A126--153), as did amino acids 1--97 (A 98--153). All of the mutants contained the DBD and prolonged the cell cycle in S phase (Fig. [7](#F7){ref-type="fig"}). Thus, the DNA binding, bZIP domain was sufficient to direct S phase regulation by C/EBPα whereas amino acids 1--97 contained a domain required for G1 regulation. Discussion ========== C/EBPα prolongs GHFT1-5 cells in G1 and S ----------------------------------------- The differentiative and anti-mitotic actions of C/EBPα have been observed in a variety of cell culture models \[[@B23],[@B24],[@B26],[@B40],[@B53]\]. To date, studies of the C/EBPα block of proliferation have relied on comparing the growth \[[@B23]-[@B25],[@B40]\] or DNA synthetic \[[@B22],[@B24]-[@B27]\] rates of C/EBPα-expressing and non-expressing cells. These studies suggest that multiple, possibly independent, mechanisms may contribute to C/EBPα arrest in a variety of cells \[[@B18],[@B23]-[@B26],[@B33]-[@B37],[@B40],[@B54]-[@B56]\]. Studies that distinguish the anti-proliferative actions of C/EBPα at different phases of the cell cycle, and in specific cell types, may clarify these discrepancies. Our studies showed that C/EBPα expression prolonged the proliferation of GHFT1-5 pituitary progenitor cells at two different phases of the cell cycle, G1 and S (Fig. [4](#F4){ref-type="fig"}). A G1 block has been reported previously in C/EBPα-expressing mouse L cells \[[@B26]\]. However, those studies did not define the activities within C/EBPα required for that blockage. Here, we observed that prolongation in G1 required the amino terminal 97 amino acids of C/EBPα (Fig. [7](#F7){ref-type="fig"}.). In contrast, amino acids 1 to 97 were not required for S phase arrest (Figs. [6](#F6){ref-type="fig"}, [7](#F7){ref-type="fig"}). This showed that C/EBPα prolongation of GHFT1-5 cells in the G1 and S phases occurred through molecularly distinct pathways. Such distinct effects on G1 and S may contribute to the wide variations in mechanisms reported, to date, to participate in proliferation arrest by C/EBPα. Cell Cycle Proteins and C/EBPα Inhibition of Proliferation ---------------------------------------------------------- Candidate mechanisms for the block of proliferation by C/EBPα include 1) activation of the gene promoters of the cyclin-dependent kinase inhibitor p21 \[[@B37]\], gadd45 \[[@B5],[@B6]\], gas2 and gas3 \[[@B38],[@B39]\], 2) post-transcriptional stabilization of p21 \[[@B22]\], or 3) interactions of C/EBPα with cell cycle proteins including p21 \[[@B22],[@B24],[@B25],[@B55]\], CDK2 \[[@B25],[@B36]\], CDK4 \[[@B36]\], retinoblastoma \[[@B54]\], the retinoblastoma-related protein p107 \[[@B34]\] and the G1/S-regulating E2F transcription complex \[[@B26],[@B33]\] (Fig. [1A](#F1){ref-type="fig"}). We found that transcriptional activation domain 3 and the leucine zipper were not required for prolongation of either G1 or S (Fig [7](#F7){ref-type="fig"}). These domains interact with p21, CDK2 and CDK4 \[[@B22],[@B24],[@B25],[@B36],[@B55]\]. Thus, C/EBPα interactions with at least p21 and CDK4 were not required for the anti-proliferative actions of C/EBPα in GHFT1-5 cells. CDK2, and other proteins, also interact with another domain of C/EBPα \[[@B25]\] required for proliferation arrest (discussed below). Arrest in S phase was associated with amino acids 259 to 310 (Fig. [6](#F6){ref-type="fig"}), which include the basic amino acid-rich, DNA binding region of C/EBPα. C/EBPα deleted of the leucine zipper dimerization domain, also required for DNA binding \[[@B25]\] (Fig. [2C](#F2){ref-type="fig"}), was as effective as full-length GFP-C/EBPα in prolonging both S and G1 (Fig. [6A](#F6){ref-type="fig"}). Thus, DNA binding *per se* was not required for proliferation arrest. The basic region of C/EBPα blocks the transcriptional activity of E2F by unknown mechanisms \[[@B33]\] and contains another known CDK2-interaction site \[[@B25]\]. Either of these activities may contribute to basic region-dependent, S phase arrest of GHFT1-5 cell proliferation. G1 block in GHFT1-5 cells required C/EBPα amino acids 1--97 (Fig. [7](#F7){ref-type="fig"}), comprising activation domains 1 and 2 \[[@B31]\]. Amino acids 1 to 97 have been implicated in C/EBPα growth arrest via the E2F/pl07 complexes \[[@B34],[@B35]\]. C/EBPα interaction with p107 is believed to prevent the formation of an S-phase promoting complex containing p107 and E2F \[[@B26],[@B34],[@B35]\]. Thus, G1 block in GHFT1-5 cells may be related to the ability of C/EBPα to interact with p107. However, the amino terminus of C/EBPα also functionally interacts with a number of other factors \[[@B31],[@B45],[@B46]\], any of which might contribute to proliferation arrest. Subnuclear Architecture and C/EBPα Inhibition of Proliferation -------------------------------------------------------------- Tang and Lane \[[@B49]\] showed that C/EBPα becomes associated with peri-centromeric chromatin upon expression during adipocyte differentiation. We observed that C/EBPα, expressed in mouse pituitary progenitor GHFT1-5 cells, also localized to peri-centromeric chromatin \[[@B45]\]. Here we show that targeting of C/EBPα to peri-centromeric chromatin in GHFT1-5 cells required the DNA binding domain of C/EBPα (Fig. [5](#F5){ref-type="fig"}). However, C/EBPα deleted of the DNA binding domain was as effective as full-length C/EBPα in prolonging GHFT1-5 cells in the G1 and S phases of the cell cycle (Fig. [6A](#F6){ref-type="fig"}). Thus, the specific association of C/EBPα with peri-centromeric chromatin, speculated to contribute to C/EBPα growth arrest \[[@B49]\], was not necessary for disruption of the cell cycle by C/EBPα. The independence of C/EBPα regulation of proliferation from its intranuclear localization may even be necessary since we have recently determined that pituitary cell differentiation is associated with the dispersal of C/EBPα from peri-centromeric heterochromatin by the pituitary-specific transcription factor Pit-1 (J. F. E., M. A. Kawecki, F. S., R. N. D., submitted). Transcription-Independent Inhibition of Proliferation ----------------------------------------------------- Transcription factors, like C/EBPα, may regulate cell proliferation by directly controlling the expression of proteins required for cell cycle progression. Indeed, C/EBPα activates transcription of the p21 \[[@B37]\], gadd45 \[[@B5],[@B6]\], gas2 and gas3 \[[@B38],[@B39]\] genes. However, similar levels of p21 mRNA in the livers of wild-type and homozygous C/EBPα knock-out mice \[[@B22]\] argue against a physiologically significant contribution of C/EBPα to p21 gene transcription in the liver. We also observed that transcriptionally inactive and DNA-binding-defective forms of C/EBPα still blocked GHFT1-5 cell proliferation in both the G1 and S phases of the cell cycle (Figs. [2](#F2){ref-type="fig"}, [4](#F4){ref-type="fig"},[6](#F6){ref-type="fig"}). DNA-binding-defective mutants of C/EBPα have also been observed to block proliferation in other cell lines \[[@B25],[@B33]\]. Together, this suggests that C/EBPα regulation of transcription, in general, might not contribute to the regulation of cell proliferation. It remains possible that C/EBPα activation of the transcription of some cell cycle regulatory genes does control proliferation in some cell types or under certain conditions \[[@B8],[@B37]-[@B39],[@B56]\]. The absence of a dependence of GHFT1-5 cell proliferation on transcription factor activities that are more classically associated with transcription regulation strongly suggests that C/EBPα, and perhaps other transcription factors, control proliferation and transcription through completely separable pathways. The effects of human papilloma virus E7 proteins on C/EBPα-induced differentiation and growth arrest also suggest that the effects of C/EBPα on proliferation are mechanistically divergent \[[@B40]\]. Having distinct mechanisms for regulation of gene transcription and proliferation allows a transcription factor to regulate these two critical processes completely independently of one another. This may be particularly important in differentiated, non-proliferating cells in which events that regulate a transcription factor\'s contribution to gene expression must be disconnected from that transcription factor\'s continuing blockage of proliferation. Conclusions =========== C/EBPα is a transcription factor that controls both proliferation and gene expression. C/EBPα uses molecularly distinct mechanisms to prolong the G1 and S phases of pituitary progenitor GHFT1-5 cells. The anti-proliferative effects of C/EBPα do not require activities in C/EBPα commonly required for gene-specific transcription. Proliferation arrest instead corresponds with C/EBPα domains previously implicated in post-transcriptional effects on E2F and/or CDK2 activity. Materials and Methods ===================== Plasmid construction -------------------- The GFP-C/EBPα vector was constructed by inserting the cDNA encoding the S65T derivative of GFP \[[@B57]\] at the start codon of the cDNA encoding rat C/EBPα. The fused cDNA was cloned into a previously described expression vector \[[@B58]\]. The GFP and C/EBPα portions of the fusion were separated by a 16 amino acid long linker including the epitope for the FLAG antibody. The C/EBPα-GFP vector was constructed by inserting the S65T GFP into the NcoI site present near the carboxy terminus of rat C/EBPα (also tagged at the amino terminus with the FLAG epitope) in our previously described C/EBPα expression vector \[[@B42]\]. This resulted in a deletion of the last four amino acids of C/EBPα. The DBD and DBD + 154--257 derivatives of C/EBPα were constructed by replacing, respectively, the amino terminal 257 amino acids of C/EBPα-GFP (to the SgrAI site of C/EBPα) and the amino terminal 153 amino acids of C/EBPα-GFP (to the NotI site of C/EBPα) with oligonucleotides containing a strong Kozak sequence. DBD + 1--154 was constructed by replacing amino acids 155 to 257 (from the NotI to SgrAI sites) with an oligonucleotide. The remaining constructs (Fig. [7](#F7){ref-type="fig"}) were generated by appending C/EBPα fragments to the NotI site of DBD + 1--154, or to the SgrAI site of DBD. ΔLZ was constructed by deleting amino acids from amino acid 310 (Tth111II site) at the junction of the basic region and leucine zipper of C/EBPα to the carboxy terminus of the GFP-C/EBPα fusion. The C/EBP-TATA reporter used to determine the transcriptional activity of the two GFP fusion proteins was previously described \[[@B45]\]. Transfection ------------ GHFT1-5 cells were propagated at 37°C and 5% CO~2~ in Dulbecco\'s modified Eagle (DME-H21) medium supplemented with 10% fetal calf serum. Approximately 1 × 10^7^ cells were transfected by electroporation, as previously described \[[@B42]\], with 5 μg of either the C/EBPα-GFP, GFP-C/EBPα, or control expression vectors. Following transfection, cells were propagated at 33°C. Incubation at 33°C results in significantly higher levels of green fluorescence (unpublished data), presumably because of better folding of the jellyfish GFP. Transfected cells were maintained in the dark throughout the experiment to minimize fluorescence activation of GFP prior to flow cytometry. DNA Binding and Promoter Activation ----------------------------------- Electrophoretic mobility shifts assays were done by mixing a radiolabeled oligonucleotide containing a consensus, high affinity C/EBP binding site (GATCGAGCCCCATTGCGCAATCATAGATC) together with extracts prepared from transfected cells as previously described \[[@B58],[@B59]\]. Competition with the same, unlabeled oligonucleotide in 1, 10 or 100 molar excess of the radiolabeled probed indicated specific binding. The protein DNA complex was supershifted with an antibody directed against C/EBPα (Santa Cruz Biotechnology, sc-61). For studies of the transcriptional activation properties of the fusion proteins, 1 μg of the C/EBP-TATA vector \[[@B45]\] was co-transfected and transcriptional activity was assessed by measuring the amount of chloramphenicol acetyltransferase reporter expressed with and without C/EBPα expression \[[@B45]\]. Fluorescence Microscopy ----------------------- Following transfection, most cells were plated into 14 cm dishes and allowed to grow for 24 hours. This ensured that the cells were not confluent by the time of collection for cell cycle analysis. A small portion of the transfected cells were cultivated on a 22 × 22 mm No. 1 borosilicate glass cover slip in a separate 6-well dish and treated with nocodazole, mimosine or vehicle alone as described below. The coverslips were removed for fluorescence microscopic observation immediately before collecting cells for flow cytometric analysis of DNA content and GFP expression. Cells were stained 10--20 minutes with 5 μg/ml of the cell-permeable DNA binding dye Hoechst 33342. The green fluorescence emitted from C/EBPα tagged with GFP and the blue fluorescence emitted from Hoechst 33342-stained DNA were readily distinguished by selectively exciting each fluorophore and capturing fluorescence emissions specific for each fluorophore using appropriate excitation and emission filter sets (Chroma Technology Corporation, Brattelboro, VT)\[[@B45]\]. The images shown in Fig. [5](#F5){ref-type="fig"} were acquired using an Olympus 40X PlanApochromat (0.95 numerical aperture) objective on an Olympus IX-70 microscope. Metamorph acquisition software (Universal Imaging Corporation, Downingtown, PA) on an Orca cooled interline camera (Hamamatsu, Bridgewater, NJ) were used to collect the images. Post-transfection Cell Synchronization -------------------------------------- The C/EBPα fusion proteins were expressed for one day following transfection, after which we added to the cell culture media 100 ng/ml nocodazole (Sigma) in dimethylsulphoxide (DMSO), 0.5 mM mimosine (Sigma) in DMSO, or DMSO vehicle. In some experiments, cells were also grown for 48 and 72 hours post-transfection before applying nocodazole (data not shown). The distributions of cells in G1, S and G2/M, in the presence or absence of C/EBPα expression, were similar for cells grown for 24 (reported here), 48 or 72 hours (data not shown) prior to nocodazole addition. Following nocodazole, mimosine or DMSO addition, the transfected GHFT1-5 cells continued to be cultivated at 33°C. Flow Cytometry -------------- 24 hours after drug or vehicle treatment, transfected cells were trypsinized, collected in DME-H21 media containing 10% fetal calf serum, and washed twice with PBS. The harvested cells (typically 5 × 10^6^ to 1 × 10^7^ cells) were resuspended in 300 μl of 8 μg/ml propidium iodide (Molecular Probes) in PBS containing 0.1% NP-40 and 10 μg/ml RNAse A, then incubated at room temperature in the dark for 30 minutes. The cell suspension was then analyzed on a FACScan flow cytometer (Becton Dickinson) at a flow rate of 12 μl/min. Fluorescence of the GFP and propidium iodide-bound DNA was excited with a 488 nm argon-ion laser. Green GFP fluorescence was collected using a 530/30-nm band pass filter, and orange emission from propidium iodide-bound DNA was detected using a 585/42-nm band pass filter. Photomultiplier tube voltage and spectral compensation were initially set using single-stained cells (cells expressing GFP fusions with C/EBPα but not stained with propidium iodide and cells transfected with the control vector containing no GFP cDNA but stained with propidium iodide). Electronic compensation was adjusted among the fluorescence channels to remove residual spectral overlap. The area and width of each event were measured to discriminate intact single cells from debris and from doublet or multiple cells stuck together. GFP fluorescence and propidium iodide-stained DNA fluorescence was thus collected from single cells. A minimum of 10,000 cellular events was collected for each sample. Data was analyzed using CELLQuest software (Becton Dickenson), and cell cycle subset analyses of DNA histograms were performed using ModFitLT™ software (Verity Software House, Topsham, ME). Statistical Analysis -------------------- One-way analysis of variance was used to compare data from multiple independent experiments. Statistically significant differences (\*\*, p \< 0.01; \*, p \< 0.05 compared to the Sham-transfected cells) and experiment number are indicated in the figures and legends. All significant differences were confirmed by paired, one-tailed t-tests. List of abbreviations ===================== C/EBPα CCAAT/Enhancer Binding Protein alpha G1 Growth phase 1 of the cell cycle S DNA Synthesis phase of the cell cycle G2 Growth phase 2 of the cell cycle M Mitsosis phase of the cell cycle CDK Cyclin-Dependent Kinase GFP Green Fluorescent Protein GFP-C/EBPα Fusion protein with GFP attached to the amino terminus of C/EBPα C/EBPα-GFP Fusion protein with GFP attached to the carboxy terminus of C/EBPα Δ Symbol signifying \"deletion\" when naming constructs LZ Leucine Zipper DBD DNA Binding Domain DMSO Dimethyl Sulfoxide Acknowledgements ================ We thank Dr. Paul Webb (U.C.S.F.) for critical reading of the manuscript. This work was supported by Public Health Service grant DK-54345 from the National Institutes of Health.

Background ========== Granuloma annulare is a granulomatous disease of unknown etiology. Disseminated granuloma annulare (DGA) is characterized by a chronic course of disease and frequent association with systemic disorders such as diabetes mellitus. Although spontaneous resolution can occur in some cases various therapies have been tried in DGA, including corticosteroids, several variants of psoralen plus ultraviolet-A radiation, ultraviolet- A1 radiation, systemic retinoids, and dapsone, with variable success \[[@B1]-[@B5]\]. We report a patient with recalcitrant DGA who was treated with fumaric acid esters (FAE). Case presentation ================= A 40-year old Caucasian woman presented with a 25-year history of DGA on both legs. Since one year, she also had lesions on the abdomen. Previous treatments with various therapeutic modalities (e.g., corticosteroids, dapsone, and bath psoralen plus ultraviolet-A radiation) were ineffective. On examination, she had erythematous annular plaques on the abdomen and on both legs (Fig. [1](#F1){ref-type="fig"}). Histopathologic examination of a punch biopsy specimen from the left leg revealed a normal epidermis. Below the epidermis there was mild collagen degeneration surrounded by palisading inflammatory cells. The infiltrates consisted of a mixture of monocytes, histiocytes, and occasional giant cells. These findings were consistent with the diagnosis of DGA. Complete work-up did not reveal evidence of malignancies, infections, and internal diseases such as diabetes mellitus. {#F1} Since the disease had been recalcitrant to various conventional therapies, we decided to start oral treatment with fumaric acid esters. The patient was treated with FAE in tablet form using two formulations differing in strength (low strength tablets: 30 mg dimethylfumarate, 67 mg monoethylfumarate Ca salt, 5 mg monoethylfumarate Mg salt, 3 mg monoethylfumarate Zn salt; high strength tablets: 120 mg dimethylfumarate, 87 mg monoethylfumarate Ca salt, 5 mg monoethylfumarate Mg salt, 3 mg monoethylfumarate Zn salt), supplied as Fumaderm^®^ initial and Fumaderm^®^ (Fumedica GmbH, Herne, Germany) \[[@B6]\]. Dosage of FAE was performed according to the standard therapy regimen for psoriasis patients displayed in Table [1](#T1){ref-type="table"}. After two months, a complete clearance of skin lesions on the abdomen was achieved. Long-standing lesions on the legs improved after three-month therapy (Fig. [2](#F2){ref-type="fig"}). No subjective side effects were observed during treatment. Regular laboratory investigations, including differentiate blood count and kidney function, did not revealed abnormal findings during therapy, with the exception of slight lymphocytopenia. After discontinuation of treatment with FAE the patient remained in remission during a six-month follow up period. {#F2} ::: {#T1 .table-wrap} ::: {.caption} ###### Dosage schedule of fumaric acid esters used for the presented patient with disseminated granuloma annulare ::: **Week** **Morning^\*^** **Noon^\*^** **Evening^\*^** **FAE formulation** ---------- ----------------- -------------- ----------------- --------------------- 1 1 \- \- Fumaderm^®^ initial 2 1 \- 1 Fumaderm^®^ initial 3 1 1 1 Fumaderm^®^ initial 4 1 \- \- Fumaderm^®^ 5 1 \- 1 Fumaderm^®^ 6--12 1 1 1 Fumaderm^®^ ^\*^ = number of tablets ::: Discussion ========== FAE has been shown to be an effective therapy option in patients with severe psoriasis vulgaris \[[@B6],[@B7]\]. During therapy with fumaric acid a persistent decrease in the lymphocyte count and stimulation of TH2 cytokine responses have been observed. Since psoriasis is regarded as a TH1-type inflammatory disorder, the immunomodulation away from the TH1 cytokine IFN-γ to the TH2 cytokine IL-10 may lead to improvement of the disease. Furthermore the anipsoriatic activity of fumaric acid may also be mediated by diminishing proinflammatory cytokine overexpression and the antigen-presenting capacity of monocytes and macrophages \[[@B8],[@B9]\]. It has been reported that FAE induce apoptosis in human dentritic cells as well as keratinocytes \[[@B10]\]. Histopathologically, localized granuloma as well as DGA are characterized by lymphohistiocytic and monocytic infiltrates that form palisading granulomas with central necrobiotic changes. In a recent study, numerous apoptotic macrophages have been observed within the necrobiotic areas \[[@B11]\]. A popular view concerning pathogenesis holds that granuloma annulare is based on a delayed-type hypersensitivity reaction to as yet undefined cutaneous antigens. Phototherapy (e.g., psoralen plus ultraviolet-A radiation, ultraviolet- A1 radiation) is effective in DGA and is known to suppress delayed hypersensitivity responses in the skin \[[@B3]-[@B5]\]. Previous findings suggest that a T cell-mediated immune response producing cytokines may be the dominant pathogenic factor in granuloma annulare \[[@B12]\]. Thus the efficacy of FAE in DGA may be mediate by similar immunomodulatory mechanisms that are observed in the treatment of psoriasis. Notably it has been observed that treatment with FAE was also effective in cutaneous sarcoidosis and necrobiosis lipoidica which are closely related to granuloma annulare U. Nowack, MD, and T. Gambichler, MD; unpublished data). Schulze-Dirks and Petzoldt \[[@B13]\] reported a female with a one-year history of DGA which resolved after six-week treatment of FAE. Since our patient had long-standing DGA, which was recalcitrant to potentially helpful therapeutic modalities, we do not consider that the therapeutic effect was due to spontaneous resolution. It has been demonstrated that FAE are well tolerated drugs suitable for long-term management (\> 6 months) in psoriasis. Subjective adverse effects such as flushing and gastrointestinal symptoms are frequently observed. Relative lymphocytopenia is the most frequent laboratory finding in long-term users. Therefore therapy with FAE should only be performed under controlled conditions \[[@B14]\]. FAE are a potentially beneficial therapeutic option for patients with recalcitrant DGA. Controlled trials are however necessary to fully explore the efficacy, optimal dosage, and safety of FAE in the treatment of DGA. List of abbreviations ===================== DGA: disseminated granuloma annulare FAE: fumaric acid esters Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== Liver sinusoidal endothelial cells (LSECs) differ from other endothelial cells. They possess open fenestrae that are grouped in sieve plates and lack a basal lamina \[[@B1]\]. Fenestrae measure about 150 nm and occupy 6--8% of the endothelial surface (porosity) \[[@B2],[@B3]\]. The endothelial filter characteristics determine the exchange between the blood and the hepatocytes, and are affecting the hepatic metabolism of lipoproteins including cholesterol and vitamin A \[[@B4]\]. Structural integrity of the fenestrated sinusoidal liver endothelium is believed to be essential for the maintenance of a normal exchange of fluids, solutes, particles and metabolites between the sinusoidal blood and hepatocytes. In the past, numerous publications appeared about the role of these dynamic structures under various physiological and pathological situations \[[@B5]\]. Their role and involvement in the regenerating liver after partial hepatectomy \[[@B6]\], shear stress \[[@B7]\], liposome-mediated transport \[[@B8]\], liver cancer \[[@B9]\], injury by free radicals \[[@B10]\] and chronic alcohol abuse \[[@B11]\], resulting in alcoholism-associated hyperlipoproteinemia \[[@B12]\] have been explored. In response to external signals, dynamic changes in fenestrae diameter and number occur and involve the dynamics of the actin cytoskeleton \[[@B13]\]. Discoveries in the past decade have revealed that an actomyosin-driven machinery contributes to the regulation of fenestrae diameter and this under the control of intracellular calcium levels \[[@B14]-[@B16]\]. Detergent-extracted whole mounts of LSECs showed that fenestrae are delineated by a fenestrae-associated cytoskeleton ring (FACR), which changes in diameter and thickness after different treatments. These rings therefore seem to act as a supporting lattice and \"muscle\" around fenestrae \[[@B17]\]. Furthermore, the actin cytoskeleton of LSECs has been shown to participate in cellular processes such as chemical- and cold-induced cell injury during liver transplantation \[[@B18],[@B19]\]. The recent availability of a battery of new actin binding drugs that affect the polymerization of actin by different mechanisms, provide a powerful tool to dissect the dynamics and functions of the actin cytoskeleton in various cell types \[[@B20]\]. Previously it has been demonstrated that the treatment of LSECs with cytochalasin B \[[@B21]\], latrunculin A \[[@B22]\], swinholide A, misakinolide or jasplakinolide \[[@B23]\], induce an increase in the number of fenestrae. Only after treating LSECs with misakinolide, it was possible to resolve the process of fenestrae formation and to identify a new structure involved in the process of fenestrae formation \[[@B23]\]. This illustrates the importance of the use of different anti-actin drugs to study the dynamic cellular processes that depend on the integrity and function of actin. The present study endeavours to elucidate the dynamic process of *de novo* formation of LSEC fenestrae by correlating specific changes in actin organization with fenestrae. We used two novel actin-binding compounds derived from marine organisms, halichondramide (HALI) and dihydrohalichondramide (di-h-HALI), that belong to a large group of structurally related marine macrolides \[[@B20]\]. Experiments on various cell lines show that HALI induces almost complete depletion of F-actin within minutes, illustrating the F-actin severing and monomer sequestering activities. Di-h-HALI, which differs from HALI only in having a single bond in the macrolide at positions 4--5 instead of a double bond, altered the filamentous F-actin distribution into large F-actin aggregates, and was found to possess strong barbed-end capping and weak severing activities \[[@B20]\]. Both agents possess biochemical similarities to the previous tested anti-actin drugs swinholide and misakinolide which have severing and barbed-end capping activities, respectively \[[@B20],[@B23]\]. We report here that HALI and di-h-HALI: (I) disrupts actin organization in LSECs in a distinctive manner; (II) significantly increase the number of fenestrae; (III) that di-h-HALI elicits fenestrae-forming centers (FFCs) from which nascent fenestrae are fanning out; and (IV) for reasons of comparison, we also examined the effect of HALI and di-h-HALI on cultured human umbilical vein endothelial cells (HUVECs) and bone marrow sinusoidal endothelial cells (BECs STR-4). It was reported earlier that various treatments could induce fenestrae in HUVECs \[[@B24],[@B25]\]; whereas sinusoidal endothelial cells derived from the bone marrow posses the capacity to form transendothelial openings in vivo (i.e., transendothelial channels and diaphragmed fenestrae) while these structures are lost or greatly reduced in number in vitro \[[@B26],[@B27]\]. Therefore, the effect of HALI and di-h-HALI on BECs STR-4 was investigated to address the question whether these microfilament-disrupting drugs could induce fenestrae in this cell line derived from another sinusoidal source. Results ======= Fluorescence microscopy ----------------------- In untreated LSECs, rhodamine-phalloidin staining reveals intense circular bundles lining the cell periphery and straight bundles traversing the cytoplasm. Whereas G-actin staining was most intense at the perinuclear region, and diffusely distributed in the cytoplasm (Fig. [1A](#F1){ref-type="fig"}). Exposure of cells to 100 nM HALI or di-h-HALI resulted in a loss of actin bundles and in the appearance of a drug-specific F-, and G-actin pattern (Fig. [1B, 1C](#F1){ref-type="fig"}). Maximal effects of HALI and di-h-HALI were obtained after 10 minutes treatment, and further incubation did not result in additional alteration of actin organization. With HALI, F-actin staining disappeared almost completely and only a weak speckled fluorescence could be observed in the entire cytoplasm, with only a few short and fine fibers remaining mainly at the cell margin. The rapid and massive F-actin depolymerization was reflected in a significant increase in G-actin, which was diffusely distributed in the cytoplasm (Fig. [1B](#F1){ref-type="fig"}). Di-h-HALI treatment caused complete dissolution of cytoplasmic F-actin bundles and appearance of numerous bright F-actin patches in the cytoplasm, whereas the peripheral F-actin bundles showed interruptions and became less dense. Furthermore, G-actin staining was not increased in intensity and was mainly localized around the nuclear area as in untreated cells (Fig. [1C](#F1){ref-type="fig"}). {#F1} Lower concentrations of both compounds (25 and 50 nM) resulted in partial loss of F-actin bundles that were present even after 120 minutes incubation. Exposure to higher concentrations of HALI or di-h-HALI (200 nM) resulted in a decreased viability as assessed by the percentage of cells stained by propidium iodide: from 97% (control) to 59% and 65% respectively. Scanning electron microscopy ---------------------------- To examine the effects of the two actin-binding agents on LSEC fenestration, purified cells were grown on collagen-coated cover slips and prepared for SEM. In these experiments, we treated cells with 100 nM of one of the compounds, because at this concentration the actin cytoskeleton was disassembled without further adverse effects on cell shape and viability. Untreated LSECs showed good preservation of their surface ultrastructural characteristics (Fig. [2A](#F2){ref-type="fig"}), revealing a central, bulging nucleus, surrounded by flat cytoplasmic extensions that contained fenestrae clustered in sieve plates. Within 30--60 minutes of di-h-HALI treatment, small cytoplasmic unfenestrated areas, surrounded by rows of very small fenestrae appeared, suggesting nascent fenestrae emerging from those areas (Fig. [2B](#F2){ref-type="fig"}). Curiously, HALI-treated LSECs did show such unfenestrated areas, and rows of very small fenestrae were absent even with different exposure times and concentrations. Because SEM gathers surface information with limited resolution, we prepared whole-mounts, and ultrathin sections for TEM to examine these peculiar areas in fine detail (vide infra). {#F2} While the maximum effect of the two drugs on actin organization was reached after 10 minutes, the maximum effect of the drugs on the number of fenestrae was reached at 120 minutes. Figure [2C](#F2){ref-type="fig"} depicts a di-h-HALI-treated LSEC after 120 minutes exposure to 100 nM di-h-HALI. The same SEM morphology was also seen in HALI-treated cells. Fenestrae were no longer clustered in sieve plates embedded in unfenestrated areas of cytoplasm (Fig. [2A](#F2){ref-type="fig"}), but treated cells contained abundant numbers of fenestrae interchanged with long and thin cytoplasmic arms, extending from the nucleus (Fig. [2C](#F2){ref-type="fig"}). Inside the fenestrated cytoplasm, the presence of small cytoplasmic unfenestrated areas could be observed. However, these areas were devoid of rows of very small fenestrae. Computer-assisted analysis of endothelial fenestration, using digitized SEM images, showed that fenestrae occur at a frequency of 3.0 ± 0.2 per micrometer squared in control LSEC. By comparing the effects of both microfilament-disrupting agents (at 100 nM) as a function of time, it becomes clear that HALI increases the number of fenestrae faster than di-h-HALI (Fig. [3](#F3){ref-type="fig"}). A significant difference in the number of fenestrae between control and HALI-treated LSECs could be discerned as early as 10 minutes after HALI application. In the case of di-h-HALI, a significant difference in the number of fenestrae was observed 20 minutes later than HALI (Fig. [3](#F3){ref-type="fig"}), corresponding with the observation of FFCs with connected fenestrae rows (Fig. [2B](#F2){ref-type="fig"}). After 120 minutes in the presence of HALI or di-h-HALI, the number of fenestrae per micrometer squared increased to 7.7 ± 0.5, and 6.0 ± 0.2 respectively. Lowering the concentrations of HALI or di-h-HALI to 25 or 50 nM gave insignificant changes in the number of fenestrae (data not shown). {#F3} We also measured the effect of the agents on fenestrae diameter at the end of treatment (Table [1](#T1){ref-type="table"} and Fig. [4](#F4){ref-type="fig"}). Fenestrae of HALI- or Di-h-HALI-treated LSECs have a mean diameter of 182 ± 75 nm and 165 ± 60 nm, respectively. A significant difference was found between the control and HALI- or Di-h-HALI-treated LSECs at the 0.0001 confidence level (Table [1](#T1){ref-type="table"}, Fig. [4](#F4){ref-type="fig"}). {ref-type="table"}**.** From this graph, we can conclude that treatment of LSECs with HALI or di-h-HALI results in smaller fenestrae diameters.](1471-2121-3-7-4){#F4} ::: {#T1 .table-wrap} ::: {.caption} ###### Fenestrae diameter (nm) after treatment with HALI or di-h-HALI obtained by scanning electron microscopy ::: Fenestrae diameter (nm) ± S.D. ± S.E.M. ----------- ------------------------- -------- ---------- Control 212 82 2.3 HALI 182^\*^ 75 1.7 di-h-HALI 165^\*^ 60 1.3 **Note.** Comparison of fenestrae diameter between control, HALI- and di-h-HALI-treated LSECs. Results are expressed as mean ± standard deviation (S.D.) and ± standard error of the mean (S.E.M). Data are obtained from three different experiments. ^\*^ Significant difference in diameter of fenestrae between the control and the LSECs treated with HALI (100 nM for 2 hours) and di-h-HALI (100 nM for 2 hours), (p \< 0.0001; Mann-Whitney U test -- two sided). ::: To address the question whether actin-disruption could induce fenestrae and reveal FFCs in large vessel and other capillary endothelial cells as well, the effect of 25, 50, 100 and 200 nM HALI or di-h-HALI for 10, 20, 30, 60 and 120 minutes was examined on cultured HUVECs and BECs STR-4 (Fig. [5](#F5){ref-type="fig"}). At all concentrations and incubation times, both cell types changed their morphology from well-spread to a more retracted appearance, but they never rounded up completely (Fig. [5A, 5C](#F5){ref-type="fig"}). Rhodamine-phalloidin staining revealed a complete F-actin dissolution as early as 30 minutes after treatment for both anti-actin agents when concentrations higher than 50 nM were used (data not shown). Detailed investigation of the topology of Di-h-HALI-treated HUVECs (Fig. [5B](#F5){ref-type="fig"}) and BECs STR-4 (Fig. [5D](#F5){ref-type="fig"}) revealed no sign of fenestrae and connected fenestrae rows, instead smooth cytoplasmic processes surrounding the bulging nucleus could be observed. {ref-type="fig"} for the difference with LSECs). Scale bar, 2.5 μm. **(C-D)** di-h-HALI treated BECs STR-4 show features similar to those seen in A and B (scale bars \[C\] 5 μm, \[D\] 2.5 μm).](1471-2121-3-7-5){#F5} Correlative fluorescence- and scanning electron microscopy ---------------------------------------------------------- Fenestrae in vitro have a diameter in the order of 200 nm (Table [1](#T1){ref-type="table"}), making it necessary to use electron microscopy for their study (Fig. [2](#F2){ref-type="fig"}). Therefore, to correlate fluorescently-labeled actin structures with SEM information, we combined one fluorescent and one SEM image of the same cell. This enables us to correlate actin and fenestrae-forming center topology. Examination of control LSECs stained for F-actin (Fig. [6A](#F6){ref-type="fig"}) and subsequently prepared for SEM (Fig. [6B](#F6){ref-type="fig"}), revealed features as shown in Fig. [1A](#F1){ref-type="fig"} and Fig. [2A](#F2){ref-type="fig"}, respectively. Projecting the green colored SEM micrograph on top of the corresponding red rhodamine-phalloidin stained LSEC, clearly illustrates a structural relation between F-actin and the surface topology of the cell (Fig. [6C](#F6){ref-type="fig"}). Dense peripheral actin filaments lined the cell margins, while straight actin fibers are mainly running in the peripheral cytoplasm. Close examination revealed negligible staining for F-actin in the nuclear area. Furthermore, F-actin was absent inside the fenestrated areas. However, actin filaments were running closely alongside the sieve plates (Fig. [6C](#F6){ref-type="fig"}). After 10 min of di-h-HALI treatment, the F-actin filaments are disrupted (Fig. [6D](#F6){ref-type="fig"}), sieve plates are clearly visible (Fig. [6E](#F6){ref-type="fig"}) and no sign of F-actin could be found inside these fenestrated areas (Fig. [6F](#F6){ref-type="fig"}). Moreover, the images show that the brightly stained F-actin patches as observed in Fig. [1C](#F1){ref-type="fig"} match with the fine globular topographic elevations present on the thin nonfenestrated arms which divide the cytoplasm in sieve plates (Fig. [6D, 6E, 6F](#F6){ref-type="fig"}). Thorough investigation of LSECs treated with di-h-HALI for 60 min, provided evidence that FFCs and the surrounding area are devoid of F-actin (Fig. [6G, 6H, 6I](#F6){ref-type="fig"}). Figure [6J](#F6){ref-type="fig"} illustrates that prolonged exposure to 100 nM di-h-HALI for 120 min did not result in additional alterations in actin organization (see also, Fig. [1C](#F1){ref-type="fig"}). At this time, the maximum effect of di-h-HALI on the fenestral number was reached (Fig. [6K](#F6){ref-type="fig"}), resulting in huge sieve plates which are lacking any sign of F-actin staining (Fig. [6L](#F6){ref-type="fig"}). {ref-type="fig"} and [2A](#F2){ref-type="fig"}: i.e., a well developed filamentous actin cytoskeleton (**A**) and sieve plates (**B**). The merge image (**C**) reveals that the fenestrated areas are clearly interspersed between the actin filaments. Scale bar, 2 μm. After 10 min di-h-HALI treatment (**D-F**), the images show that the brightly stained F-actin dots matches with the fine globular topographic elevations present on the thin nonfenestrated cytoplasmic arms. Within one hour of di-h-HALI treatment (**G-I**), the merge image of the fenestrated area reveals that the FFC and the area around is devoid of F-actin. Scale bar, 200 nm. (**J-L**) Images obtained after 120 min of di-h-HALI treatment. Note that the F-actin dots are localized in the thin nonfenestrated cytoplasmic arms; while the highly fenestrated cytoplasm lacks F-actin. Scale bar, 2 μm.](1471-2121-3-7-6){#F6} Transmission electron microscopy -------------------------------- One of the most straightforward ways to observe the complex architecture of the cytoskeleton is to examine whole mounts of cells by TEM \[[@B28]\]. This technique, allows the visualization of the cytoskeleton at the supramolecular level with a minimal disruption to the cells. Examination of control LSECs at low magnification showed the presence of an extensive network of cytoskeletal elements that fills the cytoplasm adjacent to the sieve plates (Fig. [7A](#F7){ref-type="fig"}). LSEC fenestration is characterized by the presence of a sieve plate-, and a fenestrae-associated cytoskeleton connected to a framework of microtubules and microfilaments \[[@B17]\]. Treatment with 100 nM HALI or di-h-HALI for 10 to 20 minutes resulted in the disappearance of microfilaments, and in the appearance of small cytoplasmic unfenestrated areas of intermediate electron density (gray centers), localized around the nucleus of all treated LSECs was apparent (Fig. [7B](#F7){ref-type="fig"}). Remarkably, between 30--60 minutes of di-h-HALI treatment, FFCs could be observed in all of these cells, consisting of rows of fenestrae, fanning out into the surrounding cytoplasm (Fig. [7C, 7D](#F7){ref-type="fig"}). Examination at higher magnification, reveals that these rows of fenestrae are clearly connected to the gray centers and that fenestrae with increasing diameter are emanating in the surrounding fenestrated cytoplasm (Fig. [7D](#F7){ref-type="fig"}). These structures are suggestive of de novo fenestrae formation and are similar to the recently described FFCs as revealed with the actin inhibitor misakinolide \[[@B23]\]. Even in static EM images these spiraling rows of fenestrae give the impression of a dynamic process, in which the newly formed fenestrae ebb away in the surrounding fenestrated cytoplasm as a tornado. This assumption was confirmed by quantifying the number of fenestrae rows connected to one FFC after different times of di-h-HALI treatment (Fig. [8](#F8){ref-type="fig"}), revealing that the onset of fenestrae formation starts after 10 min treatment and reaches its maximum after 60 min di-h-HALI treatment. At 120 min of treatment, when the burst of fenestrae formation has subsided (Fig. [7E](#F7){ref-type="fig"}), a significant shift towards a negligible low number of connected fenestrae rows could be registered (Fig. [8](#F8){ref-type="fig"}). At this time point, when the effect of di-h-HALI on the number of fenestrae reaches its maximum (Fig. [3](#F3){ref-type="fig"}), long cytoplasmic arms are extending from the nucleus into the cytoplasm and appear to divide the fenestrated cytoplasm into large sieve plates (Fig. [7E](#F7){ref-type="fig"}). At this stage, the burst of fenestrae formation has subsided and all fenestrae, including the newly formed ones, were delineated by the previously described FACR (Fig. [7F](#F7){ref-type="fig"}) with the same structure as in control LSECs \[[@B17]\]. Detailed and thorough investigation of HALI-treated LSECs at different times and concentrations only revealed the small unfenestrated areas (gray centers) that presumably represent inactive FFCs, but no sign of connected fenestrae rows. {#F7} {ref-type="fig"}**.** From these histograms it becomes clear that the onset of fenestrae formation starts after 10 min treatment and that the number of fenestrae rows connected to FFCs reaches its maximum at 60 min di-h-HALI treatment. At 120 min of treatment, when the burst of fenestrae formation has subsided (see also, Fig. [7E](#F7){ref-type="fig"}), a significant shift towards a low number of connected fenestrae rows could be registered.](1471-2121-3-7-8){#F8} In addition to the whole-mount TEM method, ultrathin sectioning was applied to cultured LSECs after treatment with 100 nM di-h-HALI at different time points. Examination of tangentially cut control LSECs shows two organizations of fenestrae as described previously \[[@B1],[@B29]\]; i.e., a somewhat complex sponge-like appearance around the nucleus (Fig. [9A](#F9){ref-type="fig"}), and grouped fenestrae in the peripheral cytoplasm (Fig. [9B](#F9){ref-type="fig"}). Within 30--60 minutes of di-h-HALI treatment, FFCs could be observed within the peripheral cytoplasm. Sectioned FFCs (Fig. [9C](#F9){ref-type="fig"}) are comparable to those seen with the whole-mount TEM method (Fig. [7C](#F7){ref-type="fig"}). Despite complete microfilament-disruption (Fig. [1C](#F1){ref-type="fig"}), the cytoplasm surrounding the FFCs show a few thin electron dense filaments fanning out in the surrounding cytoplasm (Fig. [9C](#F9){ref-type="fig"}). After 120 minutes of treatment, FFCs could be observed within the sieve plates, and they did not show connected rows of fenestrae at this stage. Moreover, on the edges of the sieve plates some short and thin filaments could be observed. At high magnification, the FFCs show a fine granular pattern of intermediate electron density (Fig. [9D](#F9){ref-type="fig"}). {#F9} Transversely cut sections revealed no additional structural information (data not shown). Discussion ========== Disassembly of filamentous actin in LSECs with the new microfilament-disrupting drug di-h-HALI enabled us to visualize FFCs (Figs. [2](#F2){ref-type="fig"}, [7](#F7){ref-type="fig"}, [9](#F9){ref-type="fig"}), most probably involved in the process of fenestrae formation, as demonstrated recently with the actin inhibitor misakinolide \[[@B20],[@B23]\]. Exposure of LSECs by other actin-perturbing agents including cytochalasin B \[[@B21]\], latrunculin A \[[@B22]\], swinholide A, jasplakinolide \[[@B23]\], and HALI (this paper) also produces a rapid increase in the number of fenestrae and the appearance of small unfenestrated areas which apparently represent inactive FFCs \[[@B23]\], indicating that actin disruption per se is sufficient to induce an increase in the number of fenestrae. However, the fact that only misakinolide and di-h-HALI resolved FFCs in the process of fenestrae formation indicate that specific alterations of the actin system are necessary to unmask active FFCs. Moreover, if our previous observations on FFCs as revealed by using misakinolide was an artefact of the drug, then it is most unlikely that di-h-HALI has the same side effect. The biochemical property that misakinolide and di-h-HALI have in common is their barbed end capping activity \[[@B20],[@B30]\], however misakinolide also forms actin dimers, whereas di-h-HALI possesses weak F-actin severing activity. In both cases an increased number of fenestrae (Fig. [3](#F3){ref-type="fig"}) and FFCs (Fig. [2](#F2){ref-type="fig"}, [7](#F7){ref-type="fig"}, [9](#F9){ref-type="fig"}) \[[@B23]\] could be observed, but misakinolide increases the number of fenestrae more rapidly than di-h-HALI (Fig. [3](#F3){ref-type="fig"}) \[[@B23]\], while di-h-HALI revealed approximately 40% more FFCs per squared micrometer (our unpublished data). While these differences may reflect distinct effects of these two agents on actin; it is possible that they also exert indirect effects on actin binding proteins as is the case with latrunculin \[[@B31]\]. In contrast to the other anti-actin drugs that we tested, the specific alterations that misakinolide and di-h-HALI induce in the state of actin organization either by their barbed end capping activity or by indirect effects on the actin cytoskeleton appear to slow down the process of fenestrae formation to such an extent that it becomes possible to visualize active FFCs. In addition, beside these actin-related effects, possible membrane-associated interactions of di-h-HALI or misakinolide \[[@B23]\] cannot be excluded and may promote or inhibit the fusion/fission process of the cell membrane during fenestrae formation and as a consequence retard this process in such a way that FFCs appear with connected fenestrae rows. Intrinstingly, it has been recently reported that cytochalasin D facilitates apical membrane invagination and promotes exocytosis in pancreatic acinar cells \[[@B32]\]; whereas cytochalasin B inhibits membrane invagination recovery in neurons \[[@B33]\]. Endothelial cells of large vessels, which normally do not have fenestrae, have the ability to form fenestrae within minutes \[[@B34],[@B35]\]. This indicates that the process of fenestrae formation probably does not involve de novo synthesis of proteins, but rather a reorganization of preexisting cellular components. Indeed, specialized structures involved in the formation of diaphragmed fenestrae in the capillaries of the exocrine pancreas \[[@B36]\], adrenal cortex \[[@B37]\], kidney glomerulus \[[@B38]\] and tumor micro vascular endothelium \[[@B39]\] have been reported. It is presumed that peristomal rings of cholesterol, knob-like structures, vesiculo-vacuolar organelles and the endothelial pocket may represent important contributors to the formation of diaphragmed fenestrae. Therefore, it is conceivable that fenestrae increase, whether diaphragmed or not, is depending on pre-existing structures which promote fenestrae formation. Although these specialized structures may have an important role in the formation of diaphragmed fenestrae, their role in LSECs is less certain, primarily because LSEC fenestrae lack a diaphragm, are exceptionally abundant, and differ structurally from fenestrae in other blood vessels \[[@B1]\]. The only fenestrae-related structure that LSECs and diaphragmed endothelial cells have in common in the complex process of fenestrae formation is the peristomal ring of sterols surrounding a fenestra \[[@B36]\]. Nonetheless, our observations on the effect of di-h-HALI on large vessel endothelial cells and bone marrow sinusoidal endothelial cells (Fig. [5](#F5){ref-type="fig"}), demonstrated once more that the increase in the number of fenestrae and the appearance of FFCs by actin-disruption is probably a unique process for the hepatic sinusoidal endothelium. It has been reported that phorbol myristate acetate (PMA) \[[@B24]\] and vascular endothelial growth factor (VEGF) \[[@B25],[@B40]\] could induce diaphraghmed fenestrae in HUVECs. However, De Zanger et al. \[[@B41]\] showed that LSECs are insensible to PMA with regard to fenestrae induction. In addition, Krause and collaborators \[[@B42]\] recently noticed that the increase in the number of fenestrae with time in five days old LSECs cultures is independent of the presence of VEGF. Taking together, these observations clearly illustrate that LSECs respond in a different way to inducers of diaphragmed fenestrae, indicating once more that the machinery for the formation of diaphragmed versus non-diaphragmed fenestrae probably differs. Recent data are accumulating to show that VEGF-induced diaphragmed fenestrae are derived from fused caveolae \[[@B39]\]. A recent study postulates that the same mechanism is used for the formation of LSEC-fenestrae \[[@B43],[@B44]\]. Evidence at the ultrastructural or molecular level about a possible relation between FFCs and interconnected caveolae is absent. Fusion of two opposing cell membranes to form fenestrae probably requires the presence of unique compositional membrane microdomains and a cell membrane-associated cytoskeletal structure. Several theories have been used to model the possible mechanisms of membrane fusion for LSECs and other cell types. In general, the process leading to membrane fusion and fission is subdivided into different steps, i.e.: adhesion-dehydration; disappearance of the hydration barrier; contact between phospholipid bilayers, and; molecular rearrangement, resulting in pore formation \[[@B45],[@B46]\]. As for LSECs, the first step corresponds to the formation of intramembrane protein-free zones \[[@B47]\], while the appearance of peristomal rings of sterols probably corresponds with the final step \[[@B36]\]. It seems reasonable to assume that these events take place in the rim of FFCs, and that these microdomains contain molecules to pull the bilayers of the cell membranes together at the edge of FFCs. Therefore, to define the FFC more precisely, we applied ultrathin sectioning and showed that filamentous structures of unknown nature are closely associated with these microdomains (Fig. [9C, 9D](#F9){ref-type="fig"}). In active FFCs, these filaments seem to serve as a guidance for the emerging nascent fenestrae. However, due to technical limitations it is impossible to get a nice plane overview picture of these filamentous structures on EM-sections. These limitations include the extremely small portion of the total cell volume that is included in a section, and the tendency of fine and thin structures cut obliquely or in cross section to disappear visually into the cytoplasmic ground substance. Together, these conditions make it almost impossible to correlate microscopical fluorescence data (Fig. [1B, 1C](#F1){ref-type="fig"}), with morphologically identified supramolecular structures in TEM (Fig. [9C, 9D](#F9){ref-type="fig"}). Therefore, in order to clarify the changes in actin organization that underlie the process of fenestrae formation correlative fluorescence- and SEM studies on the same cells was performed (vide infra). Surprisingly, examination of sectioned FFCs did not reveal additional structures regarding the architecture of this singular structure (Fig. [9C, 9D](#F9){ref-type="fig"}). Detailed investigation showed only a finely granular pattern of intermediate electron density. Although the molecular composition of FFCs remains unknown, F-actin was clearly found to be absent (Fig. [6I, 6L](#F6){ref-type="fig"}). Steffan et al. \[[@B21]\] postulated based on their in vitro and in situ studies with cytochalasin B that these pore-free microdomains may constitute an anchorage site for cytoskeletal elements. Our TEM sections (Fig. [9C, 9D](#F9){ref-type="fig"}) and correlative images (Fig. [6](#F6){ref-type="fig"}) support this statement but show that these cytoskeletal elements do not correspond with filamentous actin. In contrast, these F-actin patches clearly match with the fine globular topographic elevations present on the thin nonfenestrated cytoplasmic arms, and may represent anchor sites for linking F-actin with the plasma membrane \[[@B48]\]. Taking previous \[[@B20],[@B21],[@B23],[@B29],[@B47]\] and present findings together, we propose as an explanation for the events described that FFCs are anchored in the perinuclear area by the actin cytoskeleton where they cannot be resolved by electron microscopy due to their complex multi-fold organization. Disorganization of the filamentous actin cytoskeleton results in a centrifugally-like translocation of the FFCs towards the attenuated peripheral cytoplasm. Flattening of the FFCs occurs at the end of this movement and results in the appearance of FFCs with connected rows of fenestrae with increasing diameter. The spiraling rows may indicate that the translocated FFCs are rotating as they move into the peripheral region. The presence of a clear-cut FACR around every single fenestrae indicates that these centers already contain the necessary machinery and/or protein components for assembling the FACR around nascent fenestrae. However, caution is required when interpreting this hypothesis, because electron microscopy provides only static information. Nevertheless, our data on the number of fenestrae rows connected to one FFC at different time points (Fig. [8](#F8){ref-type="fig"}), illustrates in one way a dynamic driven process at the level of the FFC. Recent attempts to confirm our hypothesis on the translocation of pre-existing FFC in real-time with atomic force microscopy failed \[[@B49],[@B50]\]. The unique low elastic modulus of living LSECs, together with the damaging tip-sample interactions constitute a problem to acquire sequential images of the process of fenestrae formation in real-time under the influence of cytoskeletal-disturbing agents. Conclusion ========== A comparison of all anti-actin agents tested so far, revealed that the only activity that misakinolide and dihydrohalichondramide have in common is their barbed end capping activity; this activity seems to slow down the process of fenestrae formation to such extent that it becomes possible to resolve fenestrae-forming centers; fenestrae formation resulting from microfilament disruption is probably unique to LSECs. Materials and methods ===================== Cell cultures ------------- The method for the isolation of rat LSECs has been described in detail earlier \[[@B51]\]. Briefly, the liver of a male Wistar rat (Center for Laboratory Animals, Leuven, Belgium -- Rats received humane care in compliance with the institution\'s guidelines for the care and use of laboratory animals in research \[accreditation number 99.1-212-3\]) was perfused with collagenase A (Boehringer Mannheim, Catalogue Number 1088793, Belgium). After incubation of the fragmented tissue in the same solution, the resulting cell suspension was centrifuged at 100 × g for 5 minutes to remove the parenchymal cells. The supernatant, containing a mixture of sinusoidal liver cells, was then layered on top of a two step Percoll^®^ gradient (25--50%) and centrifuged for 20 minutes at 900 × g. The intermediate zone, located between the two density layers was enriched in LSECs. LSEC purity was further enhanced by selective adherence of Kupffer cells and spreading of the LSECs on collagen. For SEM, LSECs were cultivated in 24-multiwell plates on collagen-coated thermanox cover slips. For whole mount TEM, LSECs were cultivated on collagen-coated nickel grids (300 mesh) instead of cover slips \[[@B17]\]. Formvar (1%) supporting films on nickel grids (300 mesh) were used, coated with diluted collagen (10 μl of collagen-S stock solution \[Boehringer Mannheim, Catalogue Number 1098292, Belgium\], in 900 μl sterile water). Serum free LSEC culture medium consisted of RPMI-1640 with 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin and 10 ng/ml endothelial cell growth factor (Boehringer Mannheim, Catalogue Number 1074016, Germany). After 8 hours in culture, LSECs monolayers were extensively washed and subsequently used for experiments. The cultures were estimated to have greater than 95% purity, since less than 5% of the cells examined by EM were devoid of fenestrae. HUVECs obtained from Clonetics Corporation (BioWhittaker, Catalogue Number CC-0216, New York) were grown to confluency in 24-multiwell plates with or without thermanox cover slips, using endothelial cell growth medium -- 2 (BioWhittaker, Catalogue Number CC-3162, New York) supplemented with 2% foetal bovine serum (HyClon Labs, Logan, Utah). The cells were checked routinely for the presence of von Willebrand factor as described earlier \[[@B52]\]. The murine bone marrow sinusoidal endothelial cell line STR-4 (BEC STR-4) was established by transfecting primary endothelial cell cultures with SV40 \[[@B53]\]. BECs STR-4 were maintained in culture in RPMI-1640 supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, MEM (Gibco Life Sciences, Catalogue Number 11120, Belgium) and 10% bovine serum (fetal Clone I, Hyclone, UT, USA). Cultures were regularly checked for von Willebrand factor positivity and the capacity to uptake LDL \[[@B53],[@B54]\]. Treatment of cells with HALI and di-h-HALI ------------------------------------------ LSECs and HUVECs were treated with 25, 50, 100 and 200 nM HALI or di-h-HALI for 10, 20, 30, 60 and 120 minutes. The compounds were dissolved in dimethyl sulfoxide (DMSO) and the DMSO concentration in the assays were in all cases ≤ 0.05% and this concentration of DMSO had no effect on the ultrastructure and viability of cells as determined by EM and Hoechst 33342 / propidium iodide staining. Control media also contained DMSO in the same amount as the treated cells. All experimental variables tested, including control cells, were fixed at the same end point, i.e. all cells underwent a two hours treatment with 0.05% DMSO and the compounds were added 10, 20, 30, 60 or 120 minutes prior to fixation \[[@B21]-[@B23]\]. After incubation and fixation, cells were prepared for fluorescence microscopy, and EM as described below. Fluorescence microscopy ----------------------- In order to visualize simultaneously filamentous (F-actin) and globular (G-actin) actin, LSECs grown on glass cover slips were rinsed twice with phosphate-buffered saline (PBS) at pH 7.4, followed by fixation with freshly prepared 4% formaldehyde in PBS for 15 minutes at 21°C. After fixation, LSECs were submerged in absolute acetone for 3 minutes at -20°C. After this permeabilization, rhodamine-phalloidin (0.165 μM) and fluorescein-DNase I (0.3 μM) solution (Molecular Probes Inc., Eugene-Oregon, USA) was applied to LSECs for 20 minutes at 21°C. LSECs were subsequently washed 10 × with PBS and mounted on microscope slides in Vectashield containing DAPI (Vector Laboratories Inc., Burlingame, USA). As a control for the specificity of the staining reactions, LSECs were incubated first with 0.165 μM unlabeled phalloidin (Molecular Probes Inc., P-3457, Eugene-Oregon, USA) and 0.3 μM DNase I (Boehringer Mannheim S.A., Catalogue Number 104132, Germany) solution for 20 minutes at 21°C, before incubation with rhodamine-phalloidin and fluorescein-DNase I. No F- or G-actin staining was observed when control LSECs were incubated with unlabeled phalloidin or DNase I. Cells were viewed and recorded with a Leica DM-IRBE inverted microscope, equipped with a Leica WILD MPS 48/52 -- 35 mm camera. Elite Chrome 400 ASA slide film was used and slides were digitally scanned using a Minolta Dimâge Multi Scanner. The obtained images were transferred to Adobe Photoshop 5.5 software for color channel analysis and figure assembly. The magnification of the microscope was calibrated using fluoresbrite™ calibration grade microspheres, (Polylab BVBA, Ø 3.0 μm, Catalogue Number 18861, Belgium). Scanning electron microscopy ---------------------------- Cells were rinsed twice with PBS and fixed with 2% glutaraldehyde in Na-cacodylate buffer (0.1 M and 0.1 M sucrose) at pH 7.4 for 12 hours. They were subsequently treated with filtered 1% tannic acid in 0.15 M Na-cacodylate at pH 7.4 for 1 hour and postfixed with 1% osmium tetroxide in 0.1 M Na-cacodylate at pH 7.4 for 1 hour. SEM samples were dehydrated in a graded ethanol series, dried with hexamethyldisilazane, and sputter coated with 10 nm of gold. The samples were examined with a Philips SEM 505 (Philips Eindhoven, The Netherlands) at an accelerating voltage of 30 KV \[[@B55]\]. For automatic image analysis, the SEM was regularly calibrated at a magnification of × 20,000, using a 28.800 lines/inch grating stub with the specimen in eucentric position. 30 images at a magnification of × 20,000 were taken in randomly selected fields of each experimental variable, each image containing a minimum of 10 fenestrae. Digital images with a low-noise content were obtained by using a large spot size (20 nm) and were processed subsequently and stored on a Masscomp 5520S computer, running under the RTU UNIX operating system, as previously described \[[@B51]\]. Correlative fluorescence- and scanning electron microscopy ---------------------------------------------------------- Cells cultured on collagen-coated CELLocate-microgrid^®^ glass cover slips (Eppendorf, Catalogue Number 5245 962.004-00, Hamburg, Germany) were treated with HALI or di-h-HALI as described and subsequently stained with rhodamine-phalloidin to visualize F-actin by fluorescence microscopy (see \"Fluorescence microscopy\" section). Photographs were then taken with the fluorescence microscope and photographed cells were located simultaneously by using the alphanumerically marked grids on the cover slip. Cover slips were recovered and transferred to 2% glutaraldehyde in Na-cacodylate buffer for 12 hours and further processed for scanning electron microscopy \[[@B55]\]. Previously visualized cells in the fluorescence microscope were relocated with the aid of the alphanumerically marks on top of the cover slips and SEM-images of the corresponding cells were taken. Images of the same cells obtained from both microscopies were printed at an identical photographic end magnification \[[@B56]\], and were digitized using a Hewlett Packard ScanJet 3c scanner. The obtained images were transferred to Adobe Photoshop 5.5 software for color adjustment and figure assembly by using the replace color and duplicate layer/merge options. Transmission electron microscopy -------------------------------- In order to visualize the cytoskeleton as a whole-mount for TEM \[[@B17]\], cells cultured on collagen-coated nickel grids were rinsed twice with PBS and slightly fixed for 1 minute with freshly prepared 4% formaldehyde in PBS at 21°C. Cells were subsequently extracted in cytoskeleton buffer consisting of 1 mM ethylene glycol bis \[2-aminoethylether\]-N,N,N\',N\' tetra-acetic acid, 100 mM piperazine-N,N\'-bis \[2-ethanesulfonic acid\], 4% polyethylene glycol 6000 and 0.1% Triton X-100 in PBS at pH 6.9 for 1 minute at 21°C. After extraction, cells were processed as for SEM, but the tannin was omitted. Samples were further dehydrated and hexamethyldisilazane-dried. The specimens were examined in a Philips Tecnai 10 (Philips Eindhoven, The Netherlands) at an accelerating voltage of 100 kV. For sectioning, cells cultured on cover slips were fixed and dehydrated in the same way as for SEM. After dehydration, samples were embedded in Epon and after hardening of the embedding medium, the cover slips were removed using liquid nitrogen. Sections of 60 nm under various angles were cut with a diamond knife, stained first with uranyl acetate, subsequently with lead citrate, and examined in a Philips Tecnai 10 at 80 kV as described \[[@B51]\]. Statistical analysis -------------------- All experiments were repeated five times. Statistical analysis was performed with the Mann-Whitney U test. List of abbreviations used ========================== LSECs, liver sinusoidal endothelial cells; FACR, fenestrae-associated cytoskeleton ring; FFCs, fenestrae-forming centers; HALI, halichondramide; di-h-HALI dihydrohalichondramide; HUVECs, human umbilical vein endothelial cells; BECs STR-4, bone marrow sinusoidal endothelial cells STR-4; F-actin, filamentous actin; G-actin, globular actin; PBS, phosphate-buffered saline. Acknowledgements ================ We thank Mrs Marijke Baekeland and Danielle Blijweert for expert technical assistance. The help of Mrs Chris Derom with excellent photographic assistance is very appreciated. The authors are also thankful to Prof. Dr. Michael Goligorsky and Dr. Jun Chen (Department of Medicine and Physiology -- State University of New York at Stony Brook -- USA) for the assistance in performing the experiments with the HUVEC monolayers. The authors thank Dr. M. Kobayashi (Laboratory of Pathology -- Hokkaido University School of Medicine -- Japan) for the generous gift of the STR-4 bone marrow sinusoidal endothelial cell line. This research was financially supported by the \"Fund for Scientific Research-Flanders\" (grant N° G000599 & G038000) and by the \"Free University of Brussels\" (I. Vanderschueren Price 2000 -- Biomedicine \[F.B.\] & OZR230); and partially by the \"National Oceanic and Atmospheric Administration Award\" (NA46RG0090 NY Sea Grant Project R/XBP-5). E. Menu is an aspirant and F. Braet is a postdoctoral researcher of the \"Fund for Scientific Research-Flanders (FWO-Fl)\".

Background ========== Bladder cancer (BC) is the 5^th^ and 7^th^ most common malignancy among men and women, respectively, in Europe and the United States, with transitional cell carcinoma (TCC) dominating and accounting for more than 90% of all cases \[[@B1]\]. In the Middle East and African sub-Sahara, BC is the most common malignancy (25% of all cancers in men), with a clear dominance (87%) of the squamous cell carcinoma (SCC) sub-type \[[@B2]\]. In addition to the SCC differentiation and more pronounced male preponderance, a low mean age at diagnosis and rare involvement of the trigonal region characterize BC in Africa and the Middle East \[[@B3]\]. The geographic and clinical differences in BC behavior appear to be due mainly to etiologic differences: Whereas chemicals, including cigarette smoke and occupational exposures, cause TCC of the bladder in industrialized countries, a similarly strong association with urinary bilharziasis exists in Africa and the Middle East \[[@B4]\]. The mechanisms whereby urinary bilharziasis induces BC are not fully understood, but elevated urinary N-nitroso compounds \[[@B5],[@B6]\], elevated levels of B-glucuronidase\[[@B7]\], and chronic mechanical irritation of the urothelium by calcified eggs deposited in the bladder wall have all been implicated \[[@B6]-[@B8]\]. In contrast to the extensive cytogenetic and molecular genetic analyses that exist of bladder TCC in Western countries \[[@B9]\], little is known about the genetic alterations of post-bilharzial BC \[[@B10]\]. Cytogenetic investigations require in vitro culturing of tumor cells and therefore may be difficult to perform in areas with a high frequency of bilharziasis-associated BC (BAC). Comparative genomic hybridization (CGH), on the other hand, is a powerful molecular cytogenetic technique not dependent on the presence of vital cells. The technique utilizes differentially labeled tumor DNA and normal tissue DNA as competing probes and normal metaphases as templates to detect and localize gains and/or losses of genetic material across the entire tumor genome \[[@B11]\]. Although CGH cannot detect balanced chromosomal changes, its ability to identify genomic imbalances even in archival, paraffin-embedded tumor materials \[[@B12]\] makes it uniquely well suited when fresh samples are not available. Accordingly, we planned the present experiment to characterize the genetic defects underlying the development of BAC and to determine if its distinct morphologic and clinical characteristics evolve following genetic alterations different from those detected in BC in industrialized countries. Material and Methods ==================== Tumor material -------------- Fourteen formalin-fixed, paraffin-embedded blocks of post-bilharzial bladder carcinomas (8 SCC and 6 TCC) and 6 non-neoplastic bilharzia-associated bladder lesions from 20 Sudanese patients (12 males and 8 females) were obtained from the pathology archives of Ibn Sina Hospital and the National Health Laboratory at Khartoum, Sudan (Table [1](#T1){ref-type="table"}). All patients had a history of chronic urinary bilharziasis. No radiation therapy had been given before tumor sampling. The tumors were classified histologically according to the WHO (1973) grading system \[[@B13]\] and staged in accordance with the UICC tumor-node-metastasis (TNM) system \[[@B14]\]. ::: {#T1 .table-wrap} **Histopathology** **FISH** **CGH imbalances** --------------- ------ ------------------------------------- ------- ---------- -------------------- ------------------------------ ---------------------------------- -------------- Benign 747--98 M/65 Ch. inflammation 738--98 F/53 Ch. inflammation 162--99 M/70 Ch. Inflammation Moderate dysplasia 548--98 M/51 Granuloma CBO 352--99 F/55 Granuloma CBO Xc 3, AR 1 286--98 M/45 Granuloma with inflammatory polyp 9c 1, 17c 2 9p, 16p, 22q **Malignant** 453--99 F/49 SCC G1 ? CBO topo 1 2228--99 F/60 SCC G1 CBO 9c 1, 17c 3, Xc 3 1 cen-p31, 3p,9pter-q13 19p 229--99 M/70 SCC G2 T3 KCs Xc 2, 17c 2 210--99 M/70 SCC G3 T3 KCs 471--99 M/60 SCC G3 ? CBO+SM Xc 2, 17c 3 217--98 F/51 SCC G3 T3 CBO 9p 476--98 M/53 SCC G3 T3 CBO 9c 2, 17c 3 1p21-p22, 2q13-q21, 8p, 13q, 18q 19p 479--98 M/47 SCC G2--3 CBO 2q13-q21, 8p 19p 345--98 M/61 TCC G1 ? CBO+SM 240--98 F/55 TCC G2 ? Xc 1, 9c 2,17c 2 746--98 F/60 TCC G2 Xc 2--3, 9c2--3, 17c 2--6(4) 4540--98 M/62 TCC G2 ? 5605--97 M/50 TCC G3 T3 SM Xc 2, AR 3--5 8p21-pter 5653--98 F/50 TCC G3 T3 Xc 1--3, 9c 2, 17c 2 1 cen-p31 5p, 17q, 20q SCC, squamous cell carcinoma; TCC, transitional cell carcin oma; G1, well differentiated; G2, moderately differentiated; G3, poorly differ entiated; CBO, calcified bilharzial ova; I K:Cs, keratinzed cells; SM, squamous metapla ::: CGH analysis ------------ From each case, 20--30 paraffin sections (thickness 3--4 μm) were prepared for DNA extraction using the QIAamp Tissue Kit (QIAGEN GmbH, Germany). The yield of DNA was maximized with prolonged proteinase-K digestion according to previously published protocols \[[@B12]\]. CGH analysis was performed as previously described \[[@B11]\]. In brief, tumor DNA samples were labeled with FITC-dUTP (DuPont, Boston, MA, USA) by nick translation, whereas normal reference DNA was labeled with Texas Red (Vysis Inc., Downers Grove, IL, USA). In all cases, the tumor and reference DNA samples were sex-matched. Tumor and reference DNA were mixed with unlabeled Cot-1 DNA (Gibco, BRLTM), denatured, and applied onto slides with denatured metaphases of normal lymphocytes (Vysis Inc.). After hybridization at 37°C for 48 h, the slides were washed in 0.4 × SSC/0.3% NP-40 at 74°C for 2 min and in 2 × SSC/0.1% NP-40 at room temperature for 1 min. After air-drying, the slides were then counterstained with DAPI (Vysis Inc.). Two control hybridizations were also performed. In the first one, DNAs from a normal female and a normal male were labeled and hybridized to normal male metaphases. For the second experiment, DNA from a previously characterized breast cancer cell line (MPE 600, Vysis Inc.) and DNA from a normal female were labeled and hybridized to normal male metaphases. Six to 10 three-color digital images (DAPI, FITC, and Texas Red fluorescence) were collected from each hybridization using a Zeiss Axioplan 2 epifluorescence microscope (Carl Zeiss Jena GmbH, Jena, Germany) and a Sensys charge-coupled-device camera (Photometrics, Tucson, AZ, USA) interfaced to an IPLab Spectrum 10 workstation (Signal Analytics Corporation, Vienna, VA, USA). Interpretation of CGH results was carried out as previously described. Green-to-red ratios \< 0.80 were considered to signify losses of genetic material, ratios \> 1.2 gains, and ratios \>1.5 high-level amplifications. Heterochromatic regions, the short arms of the acrocentric chromosomes, and the sex chromosomes were not included in the evaluation. Fluorescence in situ hybridization (FISH) ----------------------------------------- Pericentromeric probes for chromosomes X (pBAMX7), 9 (pHuR98), and 17 (p17H8) were labeled with FITC (DuPont) and digoxigenin using nick translation. Prior to hybridization, xylene and absolute ethanol were used to remove the paraffin from the slides. After air-drying, the samples were pretreated with 0.01 M sodium citrate (pH7.3) in 92°C for 10 min, followed by proteinase Indigestion (10 mg in 2 × SSC) in 45°C for 30 min. The slides were then rinsed in 2 × SSC, dehydrated in an ethanol series, and air-dried. The probes with human placental DNA were denatured together with the slides on a hot plate at 75°C for 5 min, and hybridized at 37°C overnight. After washing and detection of digoxigenin with anti-digoxigenin rhodamine, the slides were counterstained with 0.1 μM 4,6-diamino-2-phenylindole (DAPI) in an antifade solution (Vectashield, Vector Laboratories, Burlingame, CA, USA). The analysis was performed using an Olympus epifluorescence microscope equipped with a CCD camera (Photometrics). Results ======= The clinical, histopathologic, CGH, and FISH findings are summarized in Table [1](#T1){ref-type="table"} and Figures [1](#F1){ref-type="fig"},[2](#F2){ref-type="fig"},[3](#F3){ref-type="fig"}. The pathologic examination showed calcified bilharzial ova in at least 8 of the cases as a direct indication of urinary bilharziasis (Figures [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}). The CGH results were generally characterized by a low number of genomic alterations. Of the 20 cases, 6 carcinomas (4 SCC and 2 TCC) and one benign lesion (bilharzial granuloma) showed chromosomal imbalances varying from 1 to 6 changes. In the remaining 13 samples, no DNA copy number changes could be detected. The most frequent changes (3 cases) were of chromosomes 1, 8, 9, and 19, with losses being observed from the short arms of chromosomes 1, 8, and 9 and gains of material from 19p. The FISH analyses performed on nuclei extracted from sectioned slides of 12 cases were consistent with the CGH findings (Table [1](#T1){ref-type="table"}). {#F1} {#F2} {#F3} Discussion ========== Schistosomiasis is endemic in 74 tropical developing countries. Some 600 million of the world\'s population run the risk of becoming infected, of which 200 million are already infected. In most of the cases, infestation with schistosomiasis leads to damage of the urothelium, which results in permanent loss of mature, superficial cells and an increased rate of cell division, eventually leading to the formation of a hyperplastic urothelium consisting of immature cells. The disturbed cellular differentiation pattern thus gives rise to a variety of non-cancerous phenotypes with a pronounced tendency to undergo malignant transformation. Compared with banding cytogenetics that has the capacity to provide detailed information on balanced as well as unbalanced chromosomal rearrangements, even in cases displaying intratumor clonal heterogeneity, CGH can only detect genomic imbalances present in major tumor clones \[[@B15]\]. On the other hand, CGH is superior in the sense that it can be applied to both fresh and frozen tumor samples as well as to formalin-fixed and paraffin-embedded materials \[[@B16]\], thereby circumventing the culturing process with the technical and selection problems this involves and which may cause considerable interpretation difficulties in conventional cytogenetic analysis. In 13 of the studied bladder lesions (4 SCC, 4 TCC, and 5 non-malignant lesions), no chromosomal imbalances were detected, indicating that the genetic alterations, if any existed, were qualitatively or quantitively beyond the detection limits of the CGH method. However, the positive results in the remaining 7 cases (4 SCC, 2 TCC, and 1 granuloma) demonstrate the capacity of CGH to detect chromosomal imbalances in tumor DNA obtained from archival materials. In the 11 cases in which FISH analysis with pericentromeric probes was performed, the results were consistent with those obtained by CGH analysis. In two cases (471--99 and 746--98), however, the FISH examination revealed chromosomal imbalances that had not been detected in the previous CGH analysis (Figure [3](#F3){ref-type="fig"}). The explanation for this discrepancy could be that too few neoplastic cells were present in the sample or that secondary intratumor genetic heterogeneity had arisen, again preventing the CGH results from reaching beyond the detection thresholds. Several preneoplastic urothelial lesions may precede the development of BAC such as inflammatory polyp, bilharzial granuloma, and bilharzial ulcer. To our knowledge, no data are available on the karyotypic characteristics of these lesions. Preneoplastic lesions in other organs have been shown to harbor nonrandom acquired genetic aberrations; in fibrocystic disease of the breast, for instance, cytogenetic studies have revealed clonal chromosomal abnormalities resembling the ones detected in carcinomas of the same organ \[[@B17]\]. Hence, another objective of the present study was to find out if benign bilharzial bladder lesions also carry chromosomal aberrations, and if they do, whether they resemble those seen in bladder carcinomas. Such changes, if present, would presumably constitute the earliest steps of a putative multistage cascade of post-bilharzial carcinogenesis \[[@B18]\]. Of the 6 non-neoplastic lesions examined, 5 were noninformative, in-as-much as no gains or losses were detected, and only in case 286--98 were chromosomal imbalances seen, one of them loss of 9p. The data are of course meager, but viewed in concert with what has already been reported \[[@B19]-[@B21]\], they come across as meaningful. Thus, loss of material from 9p in both malignant and non-malignant bilharzial lesions appears to be an initiating event in BAC. However, several have shown that loss of 9p (without loss of 9q) is a common genetic event in advanced, muscle invasive TCC, and since most (80%) BACs are advanced as far as grade and stage are concerned, loss of material from 9p may be seen more frequently in these carcinomas \[[@B22]-[@B25]\]. With the exception of the gain of chromosome material from 19p observed in three cases (2228--99, 476--98, and 479--98), almost all changes detected in our series (complete or partial loss of 1p, 8p, 9p, and 13q, and gain of 5p, 17q, and 20q) have been previously reported as recurrent genetic changes in chemically induced TCC of the bladder \[[@B9]\]. However, the gain of 19p should be interpreted with caution since this chromosome is prone to hybridization variability that may result in CGH artifacts. For the latter aberrations, the available data therefore indicate that they form part of a pathogenetic pathway followed by both BAC and chemical carcinogen-induced BC. A previous study reported homozygous deletions at 9p without any involvement of 9q in 92% of bilharzial carcinomas obtained from Egyptian patients compared with only 10% in bladder TCC obtained from Swedish patients \[[@B19]\]. In the same study, the type and position of *TP53* mutations also differed between the two tumor types, again suggesting molecular differences in the genetic mechanisms of bladder carcinogenesis depending on underlying etiology. Tsutsumi et al. \[[@B21]\] also suggested that loss of 9p21 heterozygosity was associated with early carcinogenesis of SCC of the bladder; they observed homozygous deletion of p16/p19 in 45% of squamous metaplasia from bladder cancer patients, demonstrating that this change occurred already in preneoplastic cells. Muscheck et al. \[[@B26]\] reached the same conclusion, and showed that different histologic subgroups of bladder tumors are characterized by distinct patterns of chromosomal alterations. The loss of 9p we observed in bilharzia-associated lesions is therefore in complete agreement with earlier findings \[[@B19]-[@B25]\] and could be an indication that bladder carcinogenesis with this etiology follows a more narrow pathogenetic pathway (loss of 9p) than is the case for BC in the industrialized world (the earliest genomic change in these tumors is loss of 9p, loss of 9q or loss of the entire chromosome). At the time of the first diagnosis, more than 80% of BAC present as bulky fungating nodular tumor masses with deep infiltration into the bladder wall. This necessitates radical cystectomy and urinary diversion. Unfortunately, at this late stage even these extensive therapeutic measures are rarely curative and long-term survival is obtained only in about 27--39% of the cases. It is clear that BAC is a potentially preventable disease, but until the means are found to achieve primary prevention through the control of schistosomiasis, one is dependent on earlier diagnosis, which means that a test for early detection of malignancy in high-risk groups, especially for the detection of early tumor stages, is badly needed. One can envisage the use of a molecular genetic test based on the consistent loss of 9p in cells detected in the urine for such screening purposes. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgments =============== This work was supported by grants from the Medical Faculty of Lund University, the Swedish Cancer Foundation, the Torsten and Ragnar Söderberg Foundation, the Swedish Match AB, the Swedish Society for Medical Research, the Norwegian Society for Cancer Research, the John and Augusta Perssons Fund for Medical Research, and the Finnish Cancer Institute

Introduction ============ Carcinoma of the ampulla of Vater is an infrequent neoplasia. Histologically the tumour is an adenocarcinoma arising from either pancreatic ducts, acinus or bile ducts and is often of a low grade malignancy. Both sexes are affected, the male to female ratio being 2:1 with most patients being between 50 and 69 years of age \[[@B1],[@B2]\]. Early diagnosis is vital because after a complete surgical resection the five-year survival rate can be expected to be approximately 40% \[[@B3]\]. Early symptoms (pain, cholestasis, intermittent jaundice) result from the extension of the tumour and either the obstruction or the compression of the papillary duct. Extension of the tumour into the duodenum causes occult anemia or intestinal obstruction \[[@B2],[@B4]\]. Endoscopic Ultrasonography (Endosonography, EUS) is a minimally invasive new imaging procedure. In this method a high frequency transducer (7.5 and 12 MHz) is placed into the duodenum, offering a detailed ultrasound image of the duodenal wall, at the convergence of the common bile and pancreatic duct. Even small papillary tumour masses can be detected with this method \[[@B5]\]. In this prospective study we evaluate the accuracy of EUS in detection and preoperative TNM staging (T and N classification) of ampullary carcinomas. Data on clinical presentation, endoscopic findings, endosonographic findings, US and CT scan findings, treatment and follow-up of 20 patients with carcinoma of the ampulla of Vater are reported. Patients and Methods ==================== Between July 1993 and February 2000 twenty patients having carcinoma of the ampulla of Vater were prospectively evaluated. The study was approved by the Ethics Committee of the Hospital. An Informed consent for diagnostic examinations was obtained from all patients. Seven women and 13 men, aged 40 to 87 years with a median age of 67 years were included in the study. ERCP and clinical findings of a subgroup of these patients were in part reported elsewhere \[[@B6]\]. The diagnosis was confirmed by histological examination of the specimens, taken at either endoscopy or surgery. All patients were subjected to clinical and laboratory tests. All patients were examined by Ultrasonography (US) and CT scan prior to endoscopic examination. Conventional CT scans were performed with a Philips Lx scanner with a slice thickness of a 10 mm, with oral and intravenous contrast in 13 patients (those diagnosed prior to 1997). Seven patients were subjected to spiral CT (after 1997). Detection of a tumour mass by means of US and CT scan with direct (visualization of the tumour) or indirect signs of tumour (bile or pancreatic duct dilatation, presence of lymph nodes), and distal metastasis were noted. Endosonographic examination was also performed in all patients. The Olympus Co GF-EUM3 system (frequency: 7.5 or 12 MHz) with a water filled balloon technique was used. Size and echogeneity of the tumour, local invasion, existence of suspected metastatic lymph nodes and anatomical relation of the tumour to the adjacent abdominal structures were assessed. Lymph nodes were assessed for malignancy on the basis of four well established criteria which were: larger than 10 mm, round, homogeneous echo pattern and sharp borders \[[@B7]-[@B10]\]. T and N classification, according to the TNM staging system \[[@B11]\] was also recorded. (Table [1](#T1){ref-type="table"}). ::: {#T1 .table-wrap} ::: {.caption} ###### T N M classification of carcinoma of the ampulla of Vater ::: **T (Primary tumour)** ------------------------------ ----------------------------------------------------------------------------------------------------------------------------------------- **Tx** Primary tumour cannot be assessed **T1** Tumour limited to the ampulla of Vater or aphincter of Oddi **T2** Tumour invades duodenal wall **T3** Tumor invades 2 cm or less into pancreas **T4** Tumor invades more than 2 cm into pancreas and/or into other adjacent organs **N (Regional lymph nodes)** **Nx** Regional lymph nodes cannot be assessed **NO** No regional lymph nodes metastasis **N1** Regional lymph nodes metastasis **M (Distant metastasis)** **Mx** Distant metastasis not assessable **M0** No distant metastasis **M1** Distant metastasis: Hepatic metastasis, peritoneal dissemination, lymph nodes metastasis along the splenic vein or at the splenic hilum ::: Duodenoscopy and ERCP were performed in all patients after US, CT and EUS assessment. Endoscopic biopsies were taken after sphincterotomy. Seventeen patients were operated upon. Patients were followed-up for a period of 2--75 months (median 27 months) until February 2000. One patient refused any therapeutic procedure and was lost to follow-up one month after diagnosis. Statistical analysis was done by the chi-square paired sample test for proportions with Yates continuity correction. Survival analysis in 17 patients that underwent surgery was performed using the Kaplan-Meier procedure. Results ======= Symptoms, clinical characteristics and laboratory tests of all patients are shown in table [2](#T2){ref-type="table"}. The most common symptoms were jaundice (70%), weight loss (55%), epigastric pain (50%), fever (45%), nausea and/or vomiting (35%). An elevated Alkaline Phosphatase (SAP) and γ-Glutamyl Transpeptidase (95%) were the most common laboratory abnormalities. Elevated ALT and AST (40%) and anemia (35%) were also common. In table [3](#T3){ref-type="table"} we summarize endoscopic, imaging, endosonographic findings as well as outcome of all patients. ::: {#T2 .table-wrap} ::: {.caption} ###### Clinical and Laboratory data of the patients on admission. Anemia is defined as a decrease of haemoglobin more than 10% of the normal value. Elevated levels of SAP and γ-Gt (Cholestatic enzymes) and ALT, AST are defined as more than double of the normal values. ::: Case Sex Age Jaundice Fever Pain Nausea Weight loss Anemia SAP, γ Gt ALT, AST ------ ----- ----- ---------- ------- ------ -------- ------------- -------- ----------- ---------- 1 M 64 \+ \- \+ \+ \+ \+ \+ \+ 2 M 43 \+ \+ \+ \- \+ \+ \+ \+ 3 M 70 \+ \- \- \- \+ \+ \+ \+ 4 M 80 \+ \- \- \- \+ \+ \+ \+ 5 M 67 \+ \- \- \- \- \- \+ \- 6 F 75 \+ \+ \+ \- \+ \- \+ \- 7 F 47 \+ \+ \+ \- \- \- \+ \- 8 M 62 \- \+ \+ \+ \- \- \+ \- 9 M 57 \+ \+ \- \- \+ \- \+ \- 10 F 78 \+ \+ \+ \+ \+ \+ \+ 11 M 40 \- \- \- \- \- \- \+ \- 12 F 64 \- \- \- \- \- \- \+ \- 13 M 77 \- \- \+ \+ \- \- \- \- 14 M 87 \+ \+ \+ \+ \+ \- \+ \- 15 F 68 \+ \+ \+ \+ \- \- \+ \- 16 M 73 \- \- \- \- \- \- \+ \- 17 F 63 \- \- \+ \- \- \- \+ \- 18 F 59 \+ \- \+ \+ \+ \- \+ \+ 19 M 69 \+ \+ \- \- \+ \+ \+ \+ 20 M 72 \+ \- \- \- \+ \+ \+ \+ ::: ::: {#T3 .table-wrap} ::: {.caption} ###### Endoscopic biopsy, US and CT scan findings (direct or indirect signs), EUS classification for T and N (TNM staging), treatment, follow-up and outcome of the 20 patients with cancer of the ampulla of Vater. ::: Case Endoscopic biopsy US CT EUS T and N Treatment Follow-up (Febr 00) Outcome (Febr 00) ------ ------------------- ---------- ---------- ------------- ---------------- --------------------- ------------------- 1 Ca Indirect Direct T3N1 Palliative 43 Died 2 Ca Indirect Indirect T2N1 Whipple 64 Died 3 \- Indirect Indirect T3N1 Local excision 2 Died 4 Ca Indirect Indirect T2N0 Stent 63 Died 5 \- Direct Indirect T3N1 Lost lost Lost 6 normal Indirect Indirect T3N1 Whipple 75 Died 7 Ca Indirect Indirect T4N1 Local excision 34 Alive 8 Ca Indirect Indirect T2N1 Whipple 21 Alive 9 \- Indirect Indirect T2N0 Stent 23 Alive 10 normal Indirect Indirect T3N1 Palliative 7 Died 11 Ca Direct Indirect T2N1 Whipple 30 Died 12 Normal Indirect Indirect T3N1 Local excision 26 Alive 13 Ca Indirect Indirect T3N1 Whipple 23 Alive 14 Ca Indirect Indirect T4N1 Whipple 2 Died 15 Ca Indirect Direct T3N1 Local excision 39 Died 16 Ca Indirect Direct T2N1 Whipple 22 Alive 17 Dysplasia Indirect Indirect T2N0 Whipple 15 Alive 18 Ca Indirect Direct T2N1 Whipple 12 Alive 19 Ca Direct Indirect T3N1 Whipple 8 Alive 20 Ca Indirect Indirect T3N1 Whipple 3 Alive ::: The endoscopic appearance of the papilla was abnormal (polypoid or exophytic tumour mass) in 17/20 patients (Fig. [1](#F1){ref-type="fig"}, Fig [3](#F3){ref-type="fig"}). In three cases the tumour was evident only after endoscopic sphincterotomy of a protruding papilla. In the rest of the cases a tumour mass was evident. On ERCP a dilatation of the peripheral part of the common bile duct (15/20) and of the main pancreatic duct (6/20) were found. Multiple endoscopic biopsies were taken from 17 patients. In one patient endoscopic biopsies showed only severe dysplasia, while in three patients the biopsies were normal. Carcinoma of the ampulla of Vater was diagnosed in the remaining 13 patients. {#F1} {#F3} Three patients had a surgical excision based on EUS findings despite an almost normal endoscopic appearance of the ampulla. Histology of the surgical specimens, comfirmed the presence of a carcinoma. On conventional US, the tumour was identified only in 3 cases (15%) as a hypoechogenic mass at the peripheral part of the common bile duct. In 14 patients indirect signs of malignancy (dilatation of the common bile duct: 12/20, main pancreatic duct dilatation: 2/20, enlarged lymph nodes: 9/20), were noted. In two patients enlarged lymph nodes along the splenic vein were detected (Ml according to the TNM classification). In the remaining three patients the examination was completely normal. On CT scan, findings were suggestive of a malignant tumour at the distal part of the common bile duct in 4 cases. 2/13 in conventional and 2/7 in spiral CT scan (30%) with an overall accuracy of 20%. Indirect signs of malignancy were evident in 16/20 (80%) cases (dilatation of the common bile duct: 14/20, main pancreatic duct dilatation: 5/20, enlarged lymph nodes: 12/20). In two patients enlarged lymph nodes at the splenic hilum and along the splenic vessels were detected (Ml). In one patient peritoneal dissemination of the mass was shown (Ml). On EUS tumour was identified in all patients. A heterogeneous hypoechoic mass was the common presentation (fig [2](#F2){ref-type="fig"}, fig [4](#F4){ref-type="fig"}). T and N classification was feasible in all cases. Invasion of the duodenal wall (T2) was found in 8 patients. Tumour invasion within 2 cm on the pancreatic parenchyma was found in 10 patients. These patients were classified as having a T3 carcinoma. Two patients had a tumour that invaded more than 2 cm into the pancreas and they were classified as T4. In 17 patients suspected metastatic lymph nodes were detected (N1). {ref-type="fig"}) patient. Hypoechoic mass invading the duodenal wall. Carcinoma of the ampulla of Vater:**T2N0**. (**DL:** duodenal lumen, **T:** tumour mass, **CBD:** common bile duct, **m:** muscularis propria, **nLN:** non metastatic lymph node, **P:** pancreas).](1471-2482-2-1-2){#F2} {ref-type="fig"}) patient. Hypoechoic mass invading the duodenum. Enlarged hypoechoic regional lymph nodes are also present. Carcinoma of the ampulla of Vater:**T3N1**. (**DL:** duodenal lumen, **T:** tumour mass, **LN:** metastatic lymph nodes).](1471-2482-2-1-4){#F4} Whipple resection (pancreatoduodenectomy) was performed in 11 patients. Four patients had a local resection of the tumour, 2 patients had a palliative anastomosis. In total 17 patients were operated upon. Two patients refused surgery and had a stent endoscopically placed. Table [4](#T4){ref-type="table"} shows the findings of US, CT and EUS concerning tumour and LN staging as well as indirect signs of tumour presence compared to the surgical findings. Preoperative T staging by means of EUS was confirmed in 14/17 patients (82%). From the remaining 3 patients, in 2 the T stage was overestimated (T3 and T4 instead of T2 and T3 respectively) and in one underestimated (T2 instead of T3). N classification was surgically confirmed in 12/17 (71%) of the cases. All of the remaining 5 patients were overestimated by means of EUS. No metastatic lymph nodes were found in surgical samples. ::: {#T4 .table-wrap} ::: {.caption} ###### Evaluation of ampullary cancer by means of US, CT scan and EUS and comparison to surgery findings in 17 patients who underwent surgery. Distribution of T stage, enlarged lymph nodes (for US and CT scan), N stage (for EUS and Surgery), Common bile duct (CBD) and main pancreatic duct (MPD) dilatation (for US and CT scan) and M stage (for US, CT scan and Surgery). ::: **US** **CT scan** **EUS** **Surgery** ----------------------------- -------------------------------- -------------------------------- ----------------------------- --------------------------- **T stage^\*^** Advanced 2/17 (12%) Advanced 4/17 (24%) Early: 6/17 Advanced: 11/17 Early: 8/17 **Enlarged LN or N stage** 8/17 (47%) 11/17 (65%) **M stage** 2/17 3/17 NA^\*\*^ 3/17 **CBD and MPD dialatation** CBD 12/17 (71%) MPD 2/17 (12%) CBD 13/17 (76%) MPD 5/17 (29%) NA NA ^\*^ Early stage: T1 and T2 ^\*^ Advanced stage: T3 and T4 ^\*\*^Nonapplicable ::: Overall accuracy of the EUS for detection of the tumour was 100%. EUS was significantly better than both CT scan (20%) and US (15%) in directly detecting ampullary tumour (p \< 0.001). The median survival time for the patients who underwent surgery was 64 months (95% CI: 9--119) The Kaplan-Meier survival curve of those operated on is shown in figure [5](#F5){ref-type="fig"}. Discussion ========== Cancer of the ampulla of Vater must be differentiated from cancer of the head of the pancreas. Prognosis and potential therapeutic intervention are different. 87% of patients with carcinoma of the ampulla and 47% of those with a malignancy of the duodenum have a potentially operable tumour compared with only 22% of patients with tumours arising from the head of the pancreas \[[@B2],[@B12]\]. Therefore a noninvasive accurate technique for detection and staging is mandatory for these patients. The endoscopic appearance of the papilla was abnormal in most of our patients. However the distinction between an impacted stone, a benign tumour or a malignancy is not always feasible. ERCP with endoscopic sphincterotomy was a helpful diagnostic procedure in cases of intraampullary tumors with a protruding papilla. In 3 cases of our patients a polypoid mass was evident only after sphincterotomy. However, ERCP is an invasive procedure with a considerable morbidity. Moreover, endoscopic biopsies do not always allow for reliable diagnosis and preoperative assessment of tumours of the papilla of Vater. Villous adenomas can be associated with an adenocarcinoma and a histological diagnosis of an adenoma does not exclude the presence of an adenocarcinoma elsewhere in the tumour. \[[@B13]-[@B16]\]. In our patients endoscopic biopsies were diagnostic for an adenocarcinoma in 13/17 cases (76%). In one patient a severe dysplasia was reported and in 3 cases histology was completely normal. In other series the positive histological yield was only 50% of the cases \[[@B17]\]. Our results are similar to the studies from Japan and Boston where approximately one-third of the patients were misdiagnosed \[[@B18],[@B19]\]. Staging of the disease is also very important since there is a close correlation between staging and survival rate \[[@B2],[@B20]-[@B22]\]. Conventional imaging techniques such as US and CT scan are not very efficient in both diagnosis and staging. CT scan cannot be used as the only diagnostic procedure for confirming or excluding the diagnosis of ampullary cancer. Staging of ampullary cancer by CT alone was found inaccurate for the evaluation of tumour resectability \[[@B23]\]. Moreover these methods can not detect small tumours. They can indicate the presence of a tumour by indirect signs such a dilatation of the common bile and main pancreatic duct, or the presence of malignant lymph nodes or distal metastases. ERCP can also demonstrate pancreatic and bile duct dilatation \[[@B2],[@B6]\]. In a recent study, CT scan was reported to detect dilatations of peripancreatic veins with nonvisualization of inferior peripancreatic veins as an indirect sign of tumour invasion of the peripancreatic tissue \[[@B24]\]. In our patients conventional US showed a tumour in only 15% of the cases. In 70% of the cases indirect signs of the disease and in 15% no abnormalities were identified. On CT scan, a tumour was directly visualized in only 20% of the patients. In most cases (16/20), indirect signs of the disease were evident but not diagnostic. Conventional CT scan was directly diagnostic in less cases (16%) than spiral CT (30%). This fact underlines the higher accuracy of spiral CT in detecting tumours of the pancreas region \[[@B25],[@B26]\]. EUS provides an accurate method for detecting and staging of ampullary tumours, especially those infiltrating into the pancreas. It also detects tumour infiltration of the proximal blood vessels. However, EUS is unable to detect distal metastasis. Finally, EUS is useful for planning an appropriate therapeutic approach and for preventing unnecessary extensive surgery. EUS findings have been found to correlate well with prognosis \[[@B27]-[@B32]\]. In our study survival for those patients selected by means of EUS for a Wipple\'s procedure was considerably better compared to those who had only a palliative procedure, even if this was not statistically significant due to the small number of patients. This finding agrees with a previous observation according to which EUS for selecting patients for local resection may be a cost-effective strategy \[[@B33]\]. In our patients EUS was highly accurate in directly detecting ampullary tumours making this modality the imaging method of choice. Moreover, in our study the EUS accuracy for tumour invasion (T staging), was 82%. Endosonography had also a relatively high accuracy (71%) in diagnosing regional lymph node metastases (N1 stage). In agreement with other studies comparing the accuracy of EUS in differentiating between early and advanced ampullary carcinomas (T1/T2 vs T3/T4) where a concordance between 78% and 84% was reported, we found similar accuracy rates \[[@B25],[@B27]-[@B29],[@B31],[@B33]\]. The reported accuracy of EUS in predicting regional lymph nodes metastasis in ampullary carcinoma varies between 59% and 68% \[[@B25],[@B27],[@B34]\]. These findings are comparable or slightly lower than ours (71%). The differences that occur between our results and some of the other studies, could be attributed to more advanced tumours in our patients compared to the mentioned studies. In fact, 60% of our patients had an advanced carcinoma. New modalities such as Intraductal US (IDUS) and spiral CT have been used for tumour evaluation. IDUS was significantly superior to EUS and CT scan. (100% vs 59.3% vs 29.6% respectively). Sensitivity and specificity rates for IDUS and EUS were 100% versus 62.5% and 75% versus 50%, respectively. Overall accuracy rate in tumour diagnosis for IDUS was reported significantly superior to EUS (89% vs 56%) \[[@B35]\]. However in our hands as well as in other studies \[[@B25],[@B27]-[@B29],[@B31],[@B33]\] overall diagnostic accuracy for tumour detection was higher reaching levels as high as 84%. The reason for the relatively low accuracy of EUS in this study compared to the rest is not evident. IDUS appears to be the most effective imaging method in visualizing and staging tumors of the papilla but this remains to be confirmed by other studies. Combining ERCP with a catheter probe sonography offers a new diagnostic modality that has some potential advantages for local staging of small tumours of the papilla \[[@B35]-[@B37]\]. The sensitivity and specificity of another diagnostic modality, spiral CT, for detecting involvement by the tumour of the superior mesenteric vein, portal vein and lymph nodes have been reported similar to that of EUS \[[@B26]\]. However this was not true in our hands. We think that this discrepancy was due to the small number of our patients undergoing spiral CT (7/20). In conclusion compared with classical imaging procedures like US and CT scan, EUS is an effective method for the evaluation of the extent of invasion of ampullary carcinoma as well as for the involvement of regional lymph nodes before operation. Spiral CT scan may be combined with EUS in order to evaluate distal metastases. Competing interests =================== None declared {#F5} Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== Microbial cholesterol oxidases (EC 1.1.3.6) (COX) catalyze the oxidation and isomerization of cholesterol to 4-cholesten-3-one. Interest in these enzymes mostly relies in their utility in the determination of cholesterol in biological samples such as serum and foods \[[@B1]\], and also in the bioconversion of a number of 3β-hydroxysteroids in organic solvents \[[@B2]\] and in reverse micelles \[[@B3]\] (for a recent review see \[[@B5]\]). Since earliest reports on crude preparations from *Mycobacterium sp.*\[[@B4]\], cholesterol oxidases have been described in a number of bacteria and fungi \[[@B5]\]. Enzymatic properties of cholesterol oxidase from *Rhodococcus* strains (some of which named formerly as *Nocardid*) are particularly suitable for use in the analytical determination of cholesterol, in which the hydrogen peroxide formed is used in a chromogenic reaction catalyzed by horseradish peroxidase \[[@B6]\]. The *Rhodococcus* enzyme has been usually reported to be membrane bound, extractable from whole cells by treatment with detergents or trypsin, although no phospholipids are detected in the enzyme extracts \[[@B7]\]. More recent reports have demonstrated the production of both extracellular and cell-bound cholesterol oxidase by strains of this genus such as *Rhodococcus* sp. GK1 \[[@B8]\], *R. erythropolis* ATCC 25544 \[[@B9]\] and the pathogenic specie *R. equi*\[[@B10],[@B11]\]. The kinetics of enzyme synthesis at both bench and large scale by *Nocardia rhodochrous* (renamed as *Rhodococcus rhodochrous*), a strain that produces only a cell-bound COX, has been studied and the growing conditions for bacterial enzyme synthesis in fermentor were defined \[[@B12]\]. Due to the high cell-bound to extracellular ratio of cholesterol oxidase produced by *Rhodococcus,* even in those strains that also produce extracellular enzyme, the use of detergents is compulsory in the cost-effective extraction of this enzyme. The properties of protein extraction and purification combine in polyoxyethylene type detergents whose cloud point is in the biocompatible range \[[@B13]\]. For instance, Triton X-114 is as effective as Triton X-100 to extract membrane proteins, but its cloud point in semidiluted solutions (temperature at which the detergent solubility decreases sharply and a liquid-liquid phase separation is produced) is 23°C as compared to 65°C of Triton X-100. Therefore, extracted proteins partition between a detergent-rich phase and a detergent-depleted phase thus occurring protein purification. Protein purification has been accomplished successfully from either animal cells and organelles, plant tissues and microbial cells \[[@B13]-[@B15]\]. Triton X-114 at 1%w/v in buffers has been utilized to study the partitioning behavior of commercial cholesterol oxidase from several bacterial sources, resulting in partitioning toward the detergent rich-phase in all cases \[[@B16]\]. The polyoxyethylene detergent C12EO5 added the to a non-clarified culture of *Nocardia rhodochrous* was used for direct solubilization and extraction of the cell-bound cholesterol oxidase followed by phase separation \[[@B17]\]. A four-fold preconcentration and five-fold purification were achieved in optimal conditions. Due to the high cost of C12E05 these authors tried the cheaper four narrow range ethylene oxide surfactant C12-C18E05 \[[@B18]\] which was found equally suitable for direct solubilization and extraction of cell-bound cholesterol oxidase, thus this system was expanded to pilot scale \[[@B19]\]. In a previous work \[[@B9]\] we described the cell-bound and extracellular cholesterol oxidase activities from *R. erythropolis* ATCC 25544, achieving in optimal conditions 55% cell-bound and 45% extracellular activity. Their enzymatic properties strongly supported the idea that the particulate and the extracellular cholesterol oxidases are two different forms of the same enzyme with an estimated molecular mass of 55 kDa. In this work we optimize the culture conditions in a 2-liter fermentor of this extracellular cholesterol oxidase producer strain and carry out the extraction, partial purification and concentration of both types of cholesterol oxidase by using Triton X-114 phase separation. The results obtained are very promising for the use of this strain and this technique in the industrial processing of bacteria to obtain cholesterol oxidase. Results and discussion ====================== Batch cultivation of *R. erythropolis* (ATCC 25544) --------------------------------------------------- The bacteria were grown on the GYS medium in a 2-liter scale fermentor in batch mode operation under pH and temperature controlled conditions. Under this conditions the cell yield was doubled (9.5 mg/ml vs. 4.8 mg/ml dry cell weight) and the cultivation time was reduced to one third (60 vs. 180 hours) as compared with shaken flasks. These results are in good agreement with the literature \[[@B12]\]. We found that addition of 2 g/l cholesterol to the culture broth \[[@B12]\], prepared as an aqueous emulsion with the aid of Tween 80 at a weight ratio 2:1 results in a high yield of COX production \[[@B9]\], but the preparation procedure of that emulsion had a marked influence in the final enzyme yield, although not on the cell weight, as seen in Table [1](#T1){ref-type="table"}. The spray-dry method resulted advantageous because the cholesterol :Tween 80 emulsion formed readily and COX production increased in overall by three times with respect to the preparation of the 80 mixture at the flame. Enzyme production improvement resulted larger as cell-linked (3.8-fold) than as extracellular (2.3-fold). This overall increase of COX production can be due to a better availability of cholesterol to the cell since particle size obtained by spray-dry is smaller. ::: {#T1 .table-wrap} ::: {.caption} ###### Effect of the cholesterol emuIsification method on the production of COX. ::: **COX activity (U/ml)^\*^** --------------------------------------- ----------------------------- ------------------- ------------------------ **Emulsification cholesterol method** **Cell-linked** **extracellular** **Dry weight (mg/ml)** Spray-dry 230 140 8.75 At the flame 60 60 9.05 Improvement 3.8 2.3 0.97 ^\*^Enzymatic activity figures correspond to 70 hours of fermentation. ::: Results from a typical batch fermentation are shown in Figure [1](#F1){ref-type="fig"}. Three stages can be differentiated during the fermentation process, (i) A first stage (0--16 h), in which 02 consumption increases continuously and HCl is consumed to keep the pH constant to 6.75. Buckland et al. \[[@B12]\] found that pH of the culture rose by 0.5 units in the first part of growth, and then fell. An exponential increase of cell mass is observed and low levels of COX activity appear linked to cells, (ii) The second stage (16--45 h) is characterized by a strong 02 consumption and a consumption of base. In this phase aerobic metabolism drives the cell growth but the growth rate is certainly limited by the O~2~ availability -- note pO~2~ is nearly zero under continuous stirring and air supply-. This stage is also characterized by a high rate of COX production of both types, cell-linked and extracellular, (iii) The third stage (45 h to the end of the process) a second phase of consumption of acid was recorded whereas pO~2~ increased again to reach saturating levels. The greatest increase of cell-linked COX production was observed in this stage whilst extracellular COX production stopped. {#F1} The profile of fermentation was very similar to that obtained by Buckland et al. \[[@B12]\] but differed in the accumulation of extracellular COX: the strain of Nocardia (NCIB 10554) used by these authors produced very low levels of extracellular enzyme while the strain tested in this work produces high levels. They also tested the effect of dissolved oxygen tension on the production of COX and found that in limiting conditions of oxygen supply the production of cell-linked COX was low. As seen in Figure [1](#F1){ref-type="fig"}, when oxygen supply is limiting (in the second stage) the rate of cell-linked COX production decreases, however is in these conditions when extracellular COX production takes place. Thus, there seem to be some relation between dissolved oxygen tension and extracellular COX production by the strain used in this work. The results obtained are coherent with those presented in a previous study in shaken flasks \[[@B9]\], where extracellular COX production is large and arises from the partial solubilization of the cell-linked enzyme \[[@B20],[@B21]\]. After 70 hours of fermentation the total enzyme activity obtained was ca. 360 U/ml, being 230 U/ml cell-linked and 130 U/ml extracellular, thus the cell-linked to extracellular ratio is 1.26. This ratio in shaken flasks ranged from 1.26, using the same amount of Tween 80 as in this work (0.1%), to 1.38, using 1% Tween 80, but in the latter the overall yields of COX production were 7-fold smaller \[[@B9]\]. The overall yield obtained in this work is comparable to that obtained by Buckland et al. \[[@B12]\] and by Minut et al. \[[@B17]\] but larger than that of Cheetham et al. \[[@B22]\]. Watanabe et al \[[@B24]\] compared the cell-linked and extracellular COX production of 31 strains of the genus *Rhodococcus* and *Nocardia* and found that among the best extracellular COX producers, the strains *Rhodococcus* sp. N° 31 and *R. equi* N° 24, displayed the highest cell-linked to extracellular ratio, 1.32 and 2.68 respectively. Use of Triton X-114 for extraction and purification of COX ---------------------------------------------------------- In most of the available wild producer strains, COX behaves as a cell-linked enzyme, which is particularly true in the genus *Rhodococcus*\[[@B7],[@B8],[@B23]\]. Significant levels of extracellular COX have been described to be produced by the pathogenic species *R. equi*\[[@B10],[@B11],[@B24],[@B25]\] and also by *R. erythropolis*\[[@B9]\] and *Rhodococcus* sp. \[[@B8],[@B24]\] in certain culture conditions. Several authors have investigated the ability of different detergents to disrupt lipid-protein associations and to release cell-linked COX in its native state. The use of Triton X-100 has been largely accepted \[[@B7],[@B9],[@B12],[@B26],[@B27]\] but other polyoxyethylene type non-ionic detergents whose cloud point is in the biocompatible range can be used for COX solubilization and purification \[[@B17]\]. Table [2](#T2){ref-type="table"} shows the amount and the percentage of cell-linked COX extracted by either Triton X-100 or Triton X-114 at several detergent concentrations. COX could be extracted from cells and solubilized by both detergents at 1% w/v. The % of recovery is of the same order as previously reported for these detergents \[[@B17]\]. The increase of the Triton X-114 concentration from 1 to 3% w/v resulted in concomitant increases of 1.5-fold the total activity, 2.6-fold the specific activity and a % of recovery above 90%. A Triton X-114 concentration as high as 6% w/v does not improve significantly the recovery nor the specific activity. Phase separation was further induced to gain COX purity in extracts and in the broth independently. ::: {#T2 .table-wrap} ::: {.caption} ###### Cell-linked COX extraction by Triton X-100 and Triton X-114 detergents. ::: **DETERGENT** **COX activity (UE/ml)** **% recovery** **COX activity (UE/mg Prot)** ----------------- -------------------------- ---------------- ------------------------------- TRITON X-1001% 234.2 59.3 1306.9 TRITON X-114 1% 243.4 61.6 566.9 TRITON X-114 2% 274.5 69.5 860.5 TRITON X-114 3% 363.8 92.1 1466.9 TRITON X-114 6% 365.0 92.4 1674.3 ::: The cell-free extract of Triton X-114 was subjected to phase separation as such. The culture broth was first supplemented with Triton X-114, well dissolved at 4°C and then warmed up to induce detergent phase separation. Figures [2a](#F2){ref-type="fig"} and [3a](#F3){ref-type="fig"} show the distribution of enzyme activity in each phase (detergent-depleted and -rich) for each of the COX sources (cells extract and culture broth) respectively. {#F2} {#F3} As Triton X-114 concentration is increased, COX partitions towards the detergent rich phase, increasing its specific activity (Figures [2b](#F2){ref-type="fig"} and [3b](#F3){ref-type="fig"}) thus resulting in enzyme purification and also in enzyme concentration since the volume of the detergent-rich phase is much lower than the initial volume. The 1% concentration of detergent was an exception to this rule since COX partitioned toward the depleted phase under our working conditions. Partitioning of commercial COX in buffers containing 1% Triton X-114 occurred toward the rich phase and was very influenced by the buffer concentration \[[@B16]\]. Therefore, it seems that the composition of phase separation media is extremely important to the partitioning of particular proteins. The purification was made evident by running samples of COX from cells and culture broth in SDS-PAGE gels. Figure [4](#F4){ref-type="fig"} shows that in both cases the detergent-rich phase was enriched in some proteins, including COX, whereas the depleted phase showed other different protein bands. {#F4} An exceptional result was obtained when performing COX purification from the culture broth supplemented with a 6% w/v Triton X-114. The total activity recovered after phase separation was ca. 3.5-fold that measured in the broth before phase separation. This result suggests that soluble COX produced by the culture is not fully active and that it can be activated by a treatment with 6% Triton X-114 but not with 4% or less. Further increase of Triton X-114 concentration results in no improvement with respect to 6% (results not shown). This phenomenon was not observed with COX extracted from cells, therefore the enzyme most likely exists in a fully active form in the cells. We have shown previously that cell-linked and soluble COX from the same strain are almost indistinguishable as judged by some enzymatic properties such as kinetic parameters, electrophoretic mobility of the native active enzyme and thermostability \[[@B9]\]. Now we show evidence of a differential characteristic of soluble COX as compared to cell-linked: the activation by 6% Triton X-114. The observed phenomenon accepts in principle several explanations: (i) all the soluble COX molecules become activated by 3.5-fold due to a detergent effect on the protein conformation, (ii) a fraction of soluble COX is active and a fraction 3.5-fold larger is fully inactive, but can be activated due to a detergent effect on the protein conformation, (iii) an inhibitor is removed as a consequence of detergent treatment. From the first hypothesis it could be expected some difference between both enzyme forms at least at kinetic level, which we did not observe in previous studies, although it cannot be discarded. The inhibitor hypothesis is perhaps less likely since the activation effect might have been observed at all concentrations of Triton X-114 and gradually. The second hypothesis may be the most likely according to our previous results since we characterize only active enzyme and not the enzyme protein. In that case it may be hypothesized that there is an active form of COX able to both interact with components of the cell membrane or the cell wall to remain cell-linked, and to stay soluble in the culture broth, and there is an inactive form which is soluble in the culture broth. Reversion of inactive to active is induced by a high detergent concentration, which may provide an environment resembling that of cell membranes or cell walls. The active soluble form secreted by bacterial cells might eventually and reversibly turn into inactive soluble COX. When detergent concentration of the 6% rich phase was lowered by dilution the specific activity did not change, therefore the conversion of active to inactive must be very slow, that is, the existing enzyme forms are not in equilibrium. In any case, the characterization of this activation phenomenon requires further studies that are now under progress in our laboratory. The enzymatic properties exhibited by COX have been shown to depend on the method of extraction from cells, either with Triton X-100, with buffer or with trypsin \[[@B7]\]. The extracted enzyme could be interconverted from one to another with appropriated treatments such as addition or removal of 0.5% Triton X-100. So they showed evidence for the existence of different isoforms of COX from *Nocardia rhodochrous* (renamed as *Rhodococcus rhodochrous*). These authors also pointed out that no phospholipids were co-extracted even when using Triton X-100 and suggested that cell-linked enzyme is anchored through a hydrophobic tail that interacts with naturally occurring surfactants of the cell wall. Taking that work into account and our own results it may be proposed that COX shed from cell walls aided by bacterial surfactant solubilization becomes the extracellular enzyme and so its level might be related with the production of bacterial surfactants. The effect of Triton X-114 phase separation step on the purification and concentration of COX is summarized in Table [3](#T3){ref-type="table"}. Taking together purification, concentration and % of recovery, the best results were obtained when using Triton X-114 at 3% w/v for the cell-linked and at 6% for the extracellular COX. Minuth et al. \[[@B17],[@B18]\] achieved 5-fold purification and 4-fold concentration adding the detergent pentaethyleneglycol mono *n*-dodecyl ether (C12EO5) to an non-clarified culture of *Nocardia rhodochrous.* Our results, 11.6-fold purification and 20.3-fold concentration indicate that Triton X-114 should be highly recommended for use with a clarified culture broth of an extracellular COX producer strain. ::: {#T3 .table-wrap} ::: {.caption} ###### Purification and concentration of COX during Triton X-114 phase separation. ::: **Cells extract of 3%Triton X-114 w/v** **Culture broth with 6% Triton X-114 w/v** -------------------------- ----------------------------------------- -------------------------------------------- **Purification fold** 1.67 11.6(2.52)^b^ **Concentration factor** 2.56 20.3 (4.38)^b^ **% recovery** 76 (70)^a^ 312 (65)^b^ ^a^Recovery in this step and, in parenthesis, with respect to cells. ^b^ Figures in parenthesis correspond to purification, concentration and recovery in this step as if no activation occurred, calculated on the basis of remaining activity in the upper detergent-depleted phase ::: The quantification of COX partitioning in the Triton X-114 two-phase system was accomplished by determining its partition coefficient on the basis of enzyme activity. To that, concentration of COX activity in each phase has been determined by measuring enzyme activity and phase volumes after phase separation. The partition coefficient is defined as activity concentration in the rich phase over activity concentration in the depleted phase: K= \[COX\]~rich~ / \[COX\]~depleted~ Results are shown in Figure [5](#F5){ref-type="fig"}. Concerning the enzyme extracted from cells, partition coefficient reaches an optimum at 3% w/v detergent, decreasing at higher detergent concentrations. We do not have direct evidence to explain this decrease but a reason for it could be a change in the composition of the rich phase that may contain more components of the bacterial cell wall extracted at high detergent concentration, thus interfering in the partitioning of COX. In the case of the extracellular COX the higher the detergent concentration the larger the partition coefficient. As seen in table [4](#T4){ref-type="table"}, the rich phase is quite more compact after phase separation in the culture broth than in the cell-extract and its relative volume is not proportional to the detergent concentration. Thus the rich phase is less hydrated, that is, more hydrophobic, with the culture broth than with the cell-extract and the more hydrophobic the higher the detergent concentration. As a result the partition coefficient for COX in the culture broth increases with the detergent concentration. Phase separation of Triton X-114 is affected by the presence of other surfactants such as Triton X-45 \[[@B28]\] and polyols such as glycerol \[[@B29]\], thus bacterial surfactants extracted during detergent treatment may affect phase separation as well. Natural surfactants may be also present in the broth but at a much lower concentration, therefore having less influence on the phase separation. {#F5} ::: {#T4 .table-wrap} ::: {.caption} ###### Partitioning of cell-linked and extracellular COX after phase-separation of Triton X-114 ::: **COX total units** **COX activity (U/ml)** **K** ------------------- ------------------------- --------------------- ------------------------- ------------ ---------------- ------- \% Triton X-114 \% volume of rich phase Rich phase Depleted phase Rich phase Depleted phase **Cell-linked** 1 10.8 8.5 234.8 78.7 263.2 0.30 2 15.9 137.9 136.5 867.3 162.3 5.34 3 29.6 276.5 87.1 934.1 123.7 7.55 4 37.5 278.0 86.5 741.3 138.4 5.35 6 41.8 280.0 85.0 669.8 146.0 4.58 **Extracellular** 1 6.4 5.0 113.7 78.2 121.5 0.64 2 9.3 22.8 99.7 245.7 109.7 2.24 3 10.7 58.8 69.4 550.5 77.7 7.08 4 12.8 70.5 55.1 552.3 63.1 8.75 6^a^ 15.3 84.0 41 547.9 48.4 11.32 6 15.3 390 41 2544 48.4 52.60 ^a^COX units in rich phase calculated as if no activation occurred: total units before partitioning (125) -- units in depleted phase (41) ::: The partitioning of commercial COX from *Nocardia erythropolis* and *Pseudomonas sp.* has been shown to depend on the detergent partitioning, since factors that affected the Triton X-114 distribution, such as temperature of partitioning, pH and phosphate concentration of the buffers, affected in the same way the distribution of enzyme \[[@B16]\]. Thus our results are in agreement with those findings. The reason for the different behavior of the detergent packing in cell extracts and in broth may lie in the fact that detergent partitioning is affected by physicochemical factors such as the presence of polyols, lipids, surfactants, etc \[[@B13]\], which can be present in cell extracts but not in the cell culture. Conclusions =========== From a technological point of view it is simpler to handle an extracellular enzyme than a cell-linked one. The amount of extracellular COX of the non-pathogenic strain *R. erythropolis* ATCC 25544 obtained in a 1.5 1 batch fermentation represents a 36% of the activity (130 out of 360 U/ml) when measured directly from the broth, but after the 6% Triton X-114 treatment and phase separation it represents a 65% of the total activity (442 out of 672 U/ml). In addition, active COX becomes 11.6 times purer and 20.3 times more concentrated. These results may make attractive and cost-effective the use of this bacterial strain and the Triton X-114 phase separation in a 6% w/v for the industrial production of COX used in serum and food cholesterol analysis. The purification step based on Triton X-114 phase separation should be followed by further steps, such as ion-exchange chromatography, which can combine non-ionic detergent removal and protein purification in one step, in order to obtain a preparation of COX suitable for analytical applications \[[@B16]\]. Investigations to improve the percentage of extracellular COX are currently under progress. Materials and methods ===================== Commercial cholesterol oxidase from *R. erythropolis* was purchased from Boehringer Mannheim (25 U/mg). Cholesterol from lanolin, cholesterol from human gallstones, Triton X-100 and Triton X-114 were purchased from Fluka. Triton X-114 was condensed three times in 5 mM sodium phosphate buffer pH 7.5 \[[@B30],[@B31]\]. The detergent phase of the third condensation had a concentration of 25% w/v TX-114 and was used as the detergent stock solution for all the experiments. The Triton X-114 concentration was determined from its absorption at 278 nm (A~278~ = 1.25 for 0.05% w/v) \[[@B29]\] Microorganism and culture conditions ------------------------------------ The strain used in this work was *R. erythropolis* ATCC 25544 which was routinely maintained in the laboratory by periodic subculturing in GMP medium \[[@B20]\] consisting of 0.1 g/l glucose, 0.02 g/l yeast extract, 0.04 g/l peptone, 0.04 g/l meat extract, 0.05 g/l NaCl, 0.0025 g/l MgSO4 and 0.25 g/l agar. The microbial production of cholesterol oxidase was assessed as previously described by us \[[@B9]\]. Cells were grown in GYS medium in a 2 1 reactor (BIOSTAT B from B. Braun Biotech Ltd.) with a working volume of 1.5 1. Air was supplied at 2.6 vol/vol/min; pH was set constant to 6.75 and temperature to 29°C. The GYS medium is a modification of the mineral medium described by Buckland \[[@B26]\] that consisted of 10 g/l glycerol, 20 g/l yeast extract, 2 g/l (NH~4~)~2~SO~4~, 2 g/l K~2~HPO~4~, 0.01 g/l CaCl~2~.2H~2~O, 0.01 g/l FeSO~4~.7H~2~O, 0.1 g/l MgSO~4~.7H~2~O. When this culture reached a dry weight of ca. 1.0 mg/ml, an aqueous suspension of Tween 80/cholesterol was added to a final concentration of 0.2% cholesterol and 0.1% Tween 80. An aqueous suspension of cholesterol was prepared in two ways; at the flame and by a spray-dry method. (i) In the first method, cholesterol and Tween 80 were mixed by heating at the flame until total dissolution of solids, then water was added to form an emulsion by vigorous shaking for 1 hour. (ii) In the second, cholesterol and Tween 80 were co-dissolved in diethyl ether; the solvent was then removed by spray drying and the solid material was recovered and used to readily prepare an stable aqueous suspension. Extraction and partial purification of cholesterol oxidase ---------------------------------------------------------- The extraction of cell-linked COX by using Triton X-100 was as described previously \[[@B9]\]. For the cell linked COX extraction and purification by the Triton X-114 method, the extract obtained in cold after removal of cells by centrifugation was submitted to temperature-induced phase separation. The coalescence of the detergent was facilitated by warming up to 37°C for 15 min that was followed by the sharp separation of the two resulting phases by spinning at 4000 g for 15 min at 25°C. Both phases, the lower detergent-rich and the upper detergent-depleted were assayed for both enzyme activity (see below) and protein \[[@B32]\]. Using Triton X-114 also purified the extracellular COX. The cold culture broth was supplemented with Triton X-114 to the desired final concentration and detergent was completely dissolved at 4°C. Phase separation was induced as above. Enzyme assay ------------ Cholesterol oxidase activity was assayed by a modification of the method of Allain et al. \[[@B1]\] as described previously \[[@B9]\]. One unit of activity was defined as the amount of enzyme that converts 1 μmole of cholesterol/min at 37°C. All samples were diluted before enzyme assay to a final Triton X-114 concentration of 0.1% to avoid detergent interference with the assay \[[@B16]\]. SDS-PAGE -------- The protein extracts were prepared as described previously \[[@B9]\]. SDS-PAGE electrophoresis \[[@B33]\] was carried out at 200 volts at 25° in a Mini Protean cell (Bio-Rad, Richmond, California). The gels were developed by using the silver staining technique. Acknowledgements ================ M.S. is a predoctoral fellow from the Spanish Ministry of Education (plan F.P.I.). This work was supported by grants BIO394-0451 and AGF99-0396 from C.I.C.Y.T. and PIB-95/03 from Comunidad Autónoma de Murcia.

Background ========== During the last 30 years, the incidence of allergic diseases, including pollen allergy and allergic asthma, has increased in industrialized countries \[[@B1]-[@B7]\]. Epidemiological studies suggest that the increase in the prevalence of allergic diseases may be associated with environmental exposures, lifestyle, temporal development of the immune system, and genetics \[[@B8]\]. Diesel exhaust, which is one of the major sources of environmental pollution in the 20th century, has been suspected to accelerate pollen allergy and allergic respiratory diseases \[[@B9]-[@B12]\], because diesel exhaust particles (DEP) may act as an adjuvant \[[@B13]-[@B22]\] or directly induce activated mast cells to degranulate \[[@B23]\]. Previous studies addressed the possibility that inhalation of diesel exhaust during pregnancy disrupted differentiation of the thymus in the fetus due to the accumulation of testosterone \[[@B24],[@B25]\]. Testosterone-binding cells have been found in the thymus of 18-day-old fetuses and testosterone can influence the function of specific thymic epithelial cells \[[@B26]-[@B28]\]. Such immunodeficiency of the thymus, especially of suppressor T cells, elevates serum immunoglobulin E \[[@B29]-[@B32]\]. These findings strongly suggest that the inhalation of diesel exhaust during the differentiation periods suppresses proliferation of the thymus cells and contributes to the development of allergies in postnatal life. This study was conducted to pursue three objectives. First, we sought to determine the impact that diesel exhaust inhalation during the differentiation periods of the immune system had upon the elevation of immunoglobulin E against Japanese cedar pollen. To show this, time of exposure to diesel exhaust was based on the developmental process of the thymus \[[@B33],[@B34]\]. The second objective was to determine which substances -- particulate matter or the gaseous phase present in diesel exhaust -- are responsible for the elevation of IgE. The third objective was to determine the mechanism by which the inhalation exerts its toxicity by measuring the serum levels of hormones. Materials and methods ===================== Generation of diesel exhaust ---------------------------- Diesel engine exhaust was generated by running a 309 cc engine (Model NFAD50, Yanmar Diesel Co., Osaka, Japan) at 2,400 rpm. Exhaust was diluted with clean air in a dilution tunnel and then drawn into the inhalation chamber (particulate matter = 1.73 mg/m^3^, nitrogen dioxide = 0.79 ppm). For the filtered group, most of the diesel soot particles in whole exhaust were removed by HEPA filtration (ATM 3QA, Nippon Muki Co., Tokyo, Japan). After filtration, 99.9998% of particles measuring 0.05 μm or more were eliminated. Ventilation was maintained by 15 air exchanges per hour. Concentrations of nitrogen dioxide and nitrogen monoxide were continuously monitored with a chemiluminescent analyzer (Model 8440, Monitor Labs Co., San Diego, CA). Gravimetric measurements of the particulate matter were conducted daily using an automatic beta-ray dust-mass monitor (Model BAM-102, Shibata Scientific Technology Co., Tokyo, Japan). Measurement of particle sizes with a particle fractionating sampler (Andersen Type low pressure impactor LP-20, Tokyo Dylec Co., Tokyo, Japan) confirmed that more than 90% of the particulate matter in the diesel exhaust measured less than 0.05 μm. Experiment 1 ------------ ### Animals and treatments Forty-three pregnant Fischer rats (F344/DuCrj) obtained from Charles River Japan (Kanagawa, Japan) were divided into seven groups on the sixth days of pregnancy. The experiment was conducted using only male pups from those mothers. Group 1 was exposed to clean air during all of the experimental period (*Control*). Group 2 was exposed to total diesel exhaust from the 7th day of gestation (day of impregnation = day 0) until 3 days after birth, then exposed to clean air until days 49, 82 and 96 after birth. (*Total-C-C*). Group 3 was exposed to filtered exhaust from the 7th day of gestation until 3 days after birth, then exposed to clean air until days 49, 82 and 96 after birth (*Filtered-C-C*). Group 4 was exposed to diesel exhaust from the 4th day after birth through the 22nd day after birth. The group was exposed to clean air during the other periods of the experiment (*C-Total-C*). Group 5 was exposed to filtered exhaust from the 4th day after birth through the 22nd day after birth. The group was exposed to clean air during the other periods of the experiment (*C-Filtered-C*). Group 6 was exposed to diesel exhaust from day 23 after birth until day 41. The group was exposed to clean air during the other periods of the experiment (*C-C-Total*). Group 7 was exposed to filtered exhaust from day 23 after birth until day 41. The group was exposed to clean air during the other periods of the experiment (*C-C-Filtered*). Each group had the same exposure period, 19 days, to diesel exhaust. The number of pregnant rats for groups 1, 2, 3, 4, 5, 6 and 7 were 7, 7, 7, 6, 6, 5 and 5, respectively. Litter sizes ranged from 6 to 11. The total number of pups (male: female) of each group from the mothers were 53 (28:25), 50 (24:26), 54 (25:29), 47 (22:25), 44 (23:21), 43 (22:21) and 45 (20:25), respectively. The experimental protocol is summarized in Figure [1](#F1){ref-type="fig"}. {#F1} Each group of animals was maintained in an inhalation chamber (1.6 m^3^) at 24 ± 2°C and 55 ± 5% humidity on a 12-hr light: 12-hr dark illumination schedule. The diet was standard rat chow containing 1.03% calcium, 0.70% phosphorus and 200 IU vitamin D3/100 g (MF, Oriental Yeast Co. Ltd, Tokyo, Japan). The pups were kept with their mothers until weaned at 22 days of age. At the same time, the young rats were divided into groups of males and females?\" the male rats were randomly assigned to groups of 6 to 8 housed in a single cage. All the animals were allowed free access to food and water. Exposure to diesel exhaust was 6 h daily from 10 a.m. to 4 p.m., for 19 days. Body weights, thymus and spleen weights were measured and blood samples were collected to measure the titer of immunoglogulin from the abdominal aorta under ether anesthesia on day 49 before immunization and five days after the 3rd or 4th immunization. ### Allergens and immunization procedure The crude Japanese cedar pollen allergens were kindly provided by Dr. Maejima through the Japan Automobile Research Institute (Tsukuba, Japan). The amount of protein in the crude Japanese cedar pollen allergens was 65 μg containing 17.4 μg of Cry j 1 and 1.73 μg of Cry j 2 per milliliter \[[@B35]\]. Intraperitoneal injection of 1 ml of 0.125 M ammonium bicarbonate containing 5 mg crude Japanese cedar pollen allergens mixed with 4 mg aluminum hydroxide gel was performed three or four times at 2-week intervals from the 49th day after birth. ### Measurement of anti-Japanese cedar pollen IgE antibody titers The titers of IgE antibodies were measured by the Prausnitz-Kustner reaction (P-K reaction) in 12-week-old male Fischer rats (F344/DuCrj) obtained from Charles River Japan (Kanagawa, Japan). One tenth milliliter of serial two-fold dilutions of the test sera was injected intracutaneously for the passive sensitization. Forty-eight hours later, the rats were challenged intracutaneously with 0.5 μg of Cry j 1 (Hayashibara Biochemical Lab. Inc. Japan) and 0.25 μg of Cry j 2 (Hayashibara Biochemical Lab. Inc. Japan) in 0.02 ml of PBS. At the same time, 1 ml of 0.5% Evans blue was injected intravenously. The test was carried out in duplicate, and the P-K titer was expressed as the geometric means of the reciprocal of the maximal dilutions that elicited a reaction of 5 mm in diameter. Enzyme-linked immunosorbent assay (ELISA) of anti-Japanese Cedar Pollen IgE antibodies was carried out by a modification of the method of Imaoka et al. \[[@B36]\]. Briefly, the wells of microtiter plates (Immulon 2, Dynatech) were coated with 100 μg/ml of a commercial anti-rat IgE mouse monoclonal antibody (Zymed, USA) diluted to 20 μg/mL with 0.1 M NaHCO~3~ (pH8.2) at 37°C for 3 h, and the plates were washed 5 times with phosphate buffered saline (PBS). The wells were then successively incubated at room temperature with the following reagents, washing 5 times between each pair of steps: 250 μL of Block Ace (Dainihon Seiyaku, Osaka, Japan) for blocking at 4°C overnight; 100 μL of serial 10-fold dilutions of the test and normal sera?\" 100 μL of biotinylated cedar pollen allergens (Biotin Labeled Cry j 1 and 2, Hayashibara Biochemical Lab. Inc. Japan) (1 μg/mL); 100 μL of streptavidin-peroxidase (Zymed, USA) diluted 1:1,000; 100 μL of substrate solution containing 3, 3\', 5, 5\'-tetramethylbenzidine (Morinaga, Japan). Absorbance was read at 450 nm with a microplate reader. IgE titers were expressed as the absorbance obtained by subtracting the absorbance of the serum from control group. Experiment 2 ------------ Eighteen pregnant Fischer rats were divided into three groups: control, total exposed and filtered exposed. The exposure period was from day 7 of pregnancy until 3 days after delivery, corresponding to *Control, Total-C-C* and *Filtered-C-C* in experiment 1. On day 4 after birth, blood samples of newborn rats were collected. The serum pooled from the male pups of the same group was used as a sample. Serum testosterone and estradiol levels were determined using Enzyme Immunoassay Kits (Cayman Chemical, USA). Exposure conditions were the same as aforementioned. Suckling rats with their mothers were divided into three groups on day 4 after birth: control, total exposed and filtered exposed, corresponding to *Control, C-Total-C* and *C-Filtered-C* in experiment 1. The number of mother rats was six. On day 23 after birth, blood samples of weaned males were collected and hormonal levels were assayed. Statistical analysis -------------------- All reported values are expressed as means ± standard deviations (SD) or means ± standard errors (SE) except for testosterone and estradiol of 4-day-old rats in experiment 2. The differences between the groups were analyzed using the Dunnett\'s test and Steel\'s test. The rank-order correlation between IgE titers was measured by ELISA and those of the P-K reaction were analyzed by Spearman\'s method. P-values \< 0.05 were considered significant. The treatment and care of the rats were carried out according to the protocol approved by the Animal Care and Use Committee of the Tokyo Metropolitan Research Laboratory of Public Health in a facility approved by the Japan Association for Accreditation of Laboratory Animal Care. Results ======= Experiment 1 ------------ The average body weight, spleen and thymus weight of male rats from 7 groups, *Control, Total-C-C, Filtered-C-C, C-Total-C, C-Filtered-C, C-C-Total and C-C-Filtered,* on days 49, 82 and 96 are shown in Table [1](#T1){ref-type="table"}. The average weight of the thymus of non-immunized rats on day 49 was significantly lower in *Total-C-C, Filtered-C-C* and *C-Filtered-C* than in the control group (p \< 0.01, p \< 0.01, p \< 0.05, respectively). There were no differences in thymus weight between the total-exhaust-exposed group and filtered-exhaust-exposed group. The average weight of the thymus of rats on day 82 after the third immunization was significantly lower in *Total-C-C* and *Filtered-C-C* than in the control group (p \< 0.01, p \< 0.05, respectively). The average weight of the thymus of rats on day 96 after the fourth immunization was significantly lower in *Filtered-C-C* and *C-Filtered-C* than in the control group (p \< 0.05 for each). There were no significant differences in thymus weight between the total-exhaust-exposed group and filtered-exhaust-exposed group. ::: {#T1 .table-wrap} ::: {.caption} ###### Average body, spleen and thymus weight of male rats from each group on days 49, 82 and 96 ::: non-immunized: rats on day 49 82-day-old rats immunized three times 96-day-old rats immunized four times -------------- ------------------------------- --------------------------------------- -------------------------------------- ------------------- --- -------------- -------------- ---------------------- --- --------------- --------------- ------------------- Control 8 143.5 ± 10.5 488.0 ± 20.5 323.1 ± 25.5 8 238.7 ± 5.7 601.8 ± 28.3 235.0 ± 19.8 8 252.9 ± 13.4 625.0 ± 42.0 204.8 ± 18.6 Total-C-C 8 143.5 ± 11.3 488.0 ± 38 .3 276.5 ± 18.9 ^\*^ 8 230.1 ± 9.3 601.0 ± 30.6 199.9 ± 25 .0 ^\*\*^ 8 257.6 ± 17. 8 631.8 ± 41.7 186.5 ± 20. 3 Filtcrcd-C-C 8 132.7 ± 11.9 464.1 ± 61.5 272.9 ± 49.3 ^\*^ 8 230.9 ± 11.5 588.2 ± 25.0 213.5 ± 26.7 8 249.9 ± 12.8 658.5 ± 30.4 179.6 ± 21.9 C-Total-C 8 136.1 ± 7.9 462.0 ± 30.9 308.2 ± 26.1 8 230.6 ± 8.5 575.1 ± 49.4 214.7 ± 36.6 6 256.8 ± 14.6 619.2 ± 44.3 210.5 ± 40.8 C-Filtcrcd-C 7 137.8 ± 7.1 475.0 ± 25.8 289.0 ± 23.5 7 232.5 ± 6.5 606.1 ± 32.5 214.5 ± 25.1 8 241.0 ± 17.0 608.5 ± 44.5 176.1 ± 13.7 ^\*^ C-C-Total 8 146.9 ± 6.4 488.7 ± 25.7 310.1 ± 17.8 7 239.7 ± 10.5 621.9 ± 14.7 243.1 ± 23.3 7 261.6 ± 11.5 648.8 ± 39.1 218.4 ± 24 .2 C-C-Filtcrcd 8 142.7 ± 11.7 490.2 ± 32.9 309.3 ± 19. 9 6 240.2 ± 9.5 620.0 ± 21.2 248.2 ± 19.5 6 260.3 ± 10.1 663.1 ± 69. 6 220.6 ± 16.0 Values are expressed as mean ± SD. ^\*^Different from *Control*, p \< 0.05; ^\*\*^ different from Control, p \< 0.01 ::: The individual anti-Japanese cedar pollen IgE antibody titer measured by the Prausnitz-Kustner reaction (P-K reaction) and the mean IgE titers in *Control, Total-C-C,* Filtered-C-C. C-Total-C, C-Filtered-C, C-C-Total and C-C-Filtered *after the third and fourth* immunization are shown in Figure [2](#F2){ref-type="fig"}. The geometrical means of IgE titers in *Control,* Total-C-C, Filtered-C-C, C-Total-C, C-Filtered-C, C-C-Total and C-C-Filtered were 32.0 ± 6.9, 181.0 ± 1.5, 166.0 ± 1.4, 49.4 ± 4.0, 43.1 ± 6.5, 4.0 ± 1.0 and 6.3 ± 2.0 after the third immunization, and 64.0 ± 2.7, 469.5 ± 1.6, 332.0 ± 1.7, 380.4 ± 1.7, 394.8 ± 1.7, 115.9 ± 1.3 and 57.0 ± 2.8 after the fourth immunization, respectively. Log~2~ P-K IgE titers of each group are shown in Table [3](#T3){ref-type="table"}. There were significant differences between *Total-C-C, Filtered-C-C, C-Total-C, C-Filtered-C* and *Control* after the fourth immunization (p \< 0.01, p \< 0.01, p \< 0.05, p \< 0.01, respectively). There were no differences between the total-exposure and filtered-exposure conditions. {#F2} ::: {#T2 .table-wrap} ::: {.caption} ###### Organ weight and serum hormone levels of 4- and 23-day-old male rats from each group. ::: No. body weight (g) spleen (mg) thymus (mg) testosterone (pg/ml) estrogen (pg/ml) ----------------- ----- ----------------- ------------------- ------------------- ---------------------- ------------------ 4-day-old rats *Control* 20 10.14 ± 1.11 51.4 ± O.1 24.1 ± 1.1 166.6 17.8 *Total-C-C* 24 9.90 ± 0.68 46.3 ± 1.0 ^\*^ 20.4 ± 1.4 ^\*^ 390.1 28.9 *Filtered-C-C* 25 9.55 ± 0.86 42.1 ± 4.3 ^\*\*^ 19.0 ± 1.5 ^\*\*^ 255.8 16.0 23-day-old rats *Control* 10 32.35 ± 2.33 134.1 ± 16.1 108.4 ± 17.4 361.1 ± 139.4 39.8 ± 19.8 *C-Total-C* 11 34.25 ± 3.61 144.5 ± 15.2 105.5 ± 15.1 497.4 ± 220.6 ^\*^ 40.4 ± 26.8 *C-Filtered-C* 8 32.84 ± 5.12 127.1 ± 20.1 93.2 ± 17.3 ^\*^ 659.5 ± 356.2 ^\*\*^ 52.5 ± 22.4 Values are expressed as mean ± SD. Testosterone and estrogen levels on day 4 were pooled serum from males. ^\*^ Different from Control, p \< 0.05; ^\*\*^ different from Control, p \< 0.01 ::: The average IgE titers measured by ELISA after the fourth immunization and expressed as optical density in *Control, Total-C-C, Filtered-C-C, C-Total-C, C-Filtered-C, C-C-Total and C-C-Filtered* were 0.47 ± 0.06, 0.79 ± 0.35, 0.86 ± 0.46, 0.80 ± 0.22, 0.56 ± 0.08, 0.46 ± 0.04 and 0.45 ± 0.03, respectively. The IgE titers in *Filtered-C-C* and *C-Total-C* were significantly higher than in *Control* (p \< 0.05, P \< 0.05, respectively). There was a significant rank-order correlation between the IgE titers measured by ELISA and those of the P-K reaction (p \< 0.01). Experiment 2 ------------ Table [2](#T2){ref-type="table"} shows the average body weight, spleen and thymus weight, as well as the serum levels of testosterone and estrogen of 4-day-old males exposed to clean air or diesel exhaust from the 7th day of gestation to 3 days after birth. The average spleen and thymus weights were significantly lower in *Total-C-C* (p \< 0.05, p \< 0.05) and *Filtered-C-C* (p \< 0.01, P \< 0.01) than in *Control.* The testosterone levels were higher in *Total-C-C* and *Filtered-C-C* than in *Control.* Table [2](#T2){ref-type="table"} also shows the average body weight, spleen and thymus weight and hormone levels of 23-day-old male rats exposed to clean air or diesel exhaust during the suckling period. The average thymus weight was significantly lower in *C-Filtered-C* than in *Control* (p \< 0.05). The testosterone levels were significantly higher in *C-Total-C* and *C-Filtered-C* than in *Control* (p \< 0.05, P \< 0.01, respectively). There were no significant differences between the total-exhaust-exposed group and filtered-exhaust-exposed group. ::: {#T3 .table-wrap} ::: {.caption} ###### IgE titers measured by P-K reaction against pollen in rats immunized three or four times ::: Immunized three times Immunized four times ---------------- ----------------------- ---------------------- --- ----------------- *Control* 8 5.0 ± 1.0 8 6.0 ± 0.5 *Total-C-C* 8 7.5 ± 0.2 9 8.9 ± 0.2^\*\*^ *Filtered-C-C* 8 7.4 ± 0.2 8 8.4 ± 0.3^\*\*^ *C-Total-C* 8 5.6 ± 0.7 6 8.6 ± 0.3^\*^ *C-Filtered-C* 7 5.4 ± 1.0 8 8.6 ± 0.3^\*\*^ *C-C-Total* 7 2.0 ± 0.0 7 6.9 ± 0.1 *C-C-Filtered* 6 2.7 ± 0.4 6 5.8 ± 0.6 Values are expressed as mean ± SD. Testosterone and estrogen levels on day 4 were pooled serum from males. ^\*^ Different from Control, p \< 0.05; ^\*\*^ different from Control, p \< 0.01 ::: Discussion ========== The present study provides the first evidence that inhalation of diesel exhaust during the differentiation periods of the immune system accelerated the elevation of immunoglobulin E (IgE) against Japanese cedar pollen. Elevation of IgE against Japanese pollen was detected in the rats exposed to total or filtered diesel exhaust from the 7th day of gestation until 3 days after birth (*Total-C-C, Filtered-C-C*). Previous studies addressed the possibility that inhalation of diesel exhaust during pregnancy disrupted the differentiation of the thymus in the fetus due to the accumulation of testosterone \[[@B24],[@B25]\]. In the present study, the testosterone levels, which were probably closely related to maternal hormone levels, were significantly higher in 4-day-old rats from *Total-C-C* and *Filtered-C-C* than the rats from *Control.* Testosterone-binding cells have been found in the thymus of 18-day-old fetuses. Testosterone can influence the function of specific thymic epithelial cells not only directly by acting on the thymus cells, but also indirectly by modulating the function of the thymus epithelial cells that bind testosterone \[[@B26]-[@B28],[@B37]\]. Other possibilities for the depletion of lymphoid cells are a reduction in the rate of bone marrow-derived prothymocyte entry into the thymus \[[@B12],[@B38]-[@B42]\], or a higher incidence of thymocyte apotosis \[[@B12],[@B43]-[@B45]\]. Since a certain number of lymphocytes in the spleen are from the thymus, decreased weight of the spleen on day 4 after birth in *T-C-C* and *F-C-C* might reflect disturbed or delayed development of the thymus. With regard to adjuvant effects of particles, \_particles deposited in the maternal organs might have acted as an adjuvant during the fetal and suckling periods and affected immune functions. There is a little possibility that diesel exhaust acted directly as an adjuvant for causing the elevation of IgE. The rats from the groups of *Total-C-C* and *Filtered-C-C* had been exposed to diesel exhaust through breathing for 3 days. Elevated IgE was also detected in the rats exposed to diesel exhaust directly during the suckling period, *C-Total-C* and *C-Filtered-C.* Those rats had been exposed to diesel exhaust in two ways, through inhaled air and mother\'s milk. It has been indicated that inhalation of nitrogen dioxide, a strong oxidative gas contained in diesel exhaust, reduces T lymphocyte subpopulations \[[@B46]-[@B48]\]. The other possibility for hypersensitivity to pollen is the elevated steroid hormone level detected in 23-day-old rats from *C-Total-C* and *C-Filtered-C.* The influences from deposited particles in the maternal organs could not be neglected. The rats exposed to exhaust directly during weaning period, *C-C-Total and C-C-Filtered,* did not show an elevation of IgE, though those rats had a shorter interval between exposure and the first immunization compared to the rats from the other groups. Holladay et al. suggested \[[@B34]\] that developmental exposure to immunotoxicants may play a role in inducing or exacerbating hypersensitivity, which is long-lasting or permanent. It is reasonable to assume that inhalation could have influenced the differentiation period of the thymus, such as the initiation of hematopoiesis, migration of stem cells, the expansion of progenitor cells and the colonization of bone marrow and maturation to immunocompetence and induced hypersensitivity in later life. The rats from *Total-C-C* and *Filtered-C-C* responded to a smaller dose of pollen than the rats from *C-Total-C* and *C-Filtered-C.* Judging from the decreased weight of the thymus and spleen on day 4 after birth, the impact of diesel exhaust inhalation on the development of the fetal immune system could be stronger than that on the neonatal immune system. Recently, it has been pointed out that the incidence of Japanese cedar pollinosis has increased not only in children but also in infants \[[@B49],[@B50]\]. There could be some connection between insufficient development of immune system during the fetal and neonatal periods because the mother has been exposed to polluted air and the increased prevalence of Japanese cedar pollinosis and atopic diseases in infants. Regarding which substances were more responsible for insufficient development of the immune system and elevation of IgE, filtered-exhaust that included the gaseous phase and ultra fine particles measuring less than 0.05 μm might be responsible. Although the two exposure conditions, total-exhaust-exposure and filtered-exhaust-exposure had almost the same affect, the latter had a more severe impact on pups. When comparing pups from fetal-period-filtered-exposed group (*Filtered-C-C*) to those of the fetal-period-total-exposed group (*Total-C-C*), the thymus weights were significantly lower in *Filtered-C-C* than in the *Control* on postnatal day 96. In pups from the suckling-period-exposed group (*C-Filtered-C*), the thymus weights on days 23, 49 and 96 were lower than those in the *Control.* There were no significant differences between the thymus weight of *Filtered-C-C* and *Total-C-C.* Filtered exhaust contained the remainder of diesel exhaust after filtration, such as gases and particles measuring less than 0.05 μm. The gaseous phase of diesel engine exhaust includes several agents that may affect immune functions, such as nitrogen dioxide and nitrogen oxide \[[@B46]-[@B48]\]. Recently the toxicity of ultrafine particles less than 0.1 μm has been reported, such as the direct invasion of ultrafine particles through the lung, the translocation of deposited ultrafine particles to extrapulmonary tissues, the impact of co-pollutants adsorbed to ultrafine particles \[[@B52]-[@B54]\] . It can not be negligible that ultrafine particles transferred from the mother to the fetus or inhaled act as immunotoxicants. Further studies are needed to show the mechanism that is responsible for susceptibility to allergies in rats exposed to diesel exhaust during the critical period of immune system development. The yearly average suspended particulate matter (SPM) concentration in a specified area at the motor vehicle exhaust monitoring station of motor vehicle NOx during 1999 was 0.045 mg/m^3^\[[@B51]\]. Assuming the proportion of diesel exhaust particles in SPM around the roads in large cities to be 40%, the concentration of the diesel exhaust particles used in this experiment is about 90 times the concentration occurring around the roads. The rats had been exposed to diesel exhaust for 16 days during their fetal life, which is more than two thirds of the gestation period. Taking the longer gestation period of human beings, a longer time of exposure per day and a larger volume of breathing into consideration, the inhalation dose used in the present study was not far from reality. The intranasal administration is a more reasonable simulation of the natural route of pollen exposure. It is also useful to estimate the amount of pollen grains being administered. In the preliminary studies, intranasal administration was not useful for elevating IgE in Fischer rats. The major purposes of the study were to show that environmental factors could change the sensibility to pollen besides genetic factors and to show the influence of diesel exhaust inhalation on the fetus and neonate. Antibody responses against Japanese cedar pollen differed among strains of rats, with Brown Norway (BN) rats being sensitive, Fischer rats moderately sensitive and Wistar rats not sensitive \[[@B36]\]. The average litter size in BN rats is 3 to 4, and that of Fischer rats is 8 to 10. For these reasons, intraperitoneal injection was administered in Fischer rats in the present study. Even if alternative routes of pollen grain administration were used, the predictability of such methods generally used in scientific research field is well acceptable to mimic and study human allergic reaction in a rat model. Competing interests =================== none declared Authors\' contributions ======================= NW designed the study, carried out the immunoassays and drafted the manuscript. MO carried out the regulation of exposure conditions in the inhalation chambers and participated in the immunoassays. All authors read and approved the final draft. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ This work was supported by CREST of JST (Japan Science and Technology Corporation).

Background ========== Vitamin A and its biologically active metabolites, the retinoids, are essential for the maintenance of normal epithelial differentiation \[[@B1]\]. Retinoid-deficient epithelial tissues acquire a premalignant phenotype, which is characterized by enhanced mitotic activity and loss of differentiation \[[@B2]\]. Retinoids have been used as chemopreventive and chemotherapeutic agents for various human cancers \[[@B3]\]. The antiproliferative and differentiative effects of retinoids are mediated through two classes of nuclear receptors, the RARs and RXRs. Both classes consist of three members, α, β and γ \[[@B4]\]. All these receptors are ligand-activated transcription factors. They activate transcription by binding to retinoic acid response elements (RAREs) located in the promoter regions of the target genes. One of the targets of retinoid receptors is the RAR-β gene itself. The RAR-β gene encodes for four transcripts and is expressed in normal epithelial tissue, where its expression is up-regulated by treatment with retinoic acid (RA) \[[@B5],[@B6]\]. Recent studies have demonstrated that in many malignant cell lines derived from various carcinomas the level of RAR-β2 mRNA (one of four transcripts of this gene) is decreased or undetectable \[[@B7]-[@B9]\]. A decrease of RAR-β2 mRNA level (compared to normal tissues) was also observed in carcinomas of cervix \[[@B10]\], and in carcinomas of head and neck \[[@B11]\]. These findings indicate that the specific loss of RAR-β2 mRNA expression may be an important event in tumorogenesis. This hypothesis has been supported by an observation that RAR-β2 can function as a tumor suppressor gene in two different carcinoma cell lines \[[@B12],[@B13]\]. Thus, a defect in RAR-β2 expression may be one of the events that lead the tumor cells to the escape from growth regulation. The etiology of cancer of the uterine cervix has been associated with several types of human papillomaviruses (\"high-risk\" HPVs) \[[@B14]\]. In addition to growth regulation, the product of RAR-β gene can inhibit transcription of the viral oncogenes (E6 and E7 genes) in high-risk HPV-immortalized cells \[[@B15]\]. The decrease of the expression of negative regulator of viral oncogene may be an additional important step on the way towards malignant progression of HPV-positive cells in this case. It has been shown that the retinoic acid can induce an increase of RAR-β2 expression in carcinoma cells \[[@B16]\] and in lesions of the oropharynx \[[@B17]\]. RA treatment has inhibited transformation of human keratinocytes by HPV-16 \[[@B18]\] and has enhanced regression of cervical moderate dysplasia \[[@B19]\]. Beneficial effects of RA treatment of advanced cervical squamous cell carcinomas (in combination with interferon α) have been demonstrated in clinical trials \[[@B20]\]. On the other hand, the existence of retinoic acid-insensitive cancer cell lines, including cervical carcinoma cell lines, and the lack of clinical effects of retinoids treatments in some cases are well established \[[@B21],[@B22]\]. The mechanism of the down-regulation and the loss of RA-inducible expression of RAR-β2 in cervical carcinomas and cell lines derived from them is not well understood yet. Recent findings have shown that hypermethylation of the promoter region is implicated in the transcriptional silencing of many tumor suppressor genes and several other genes functionally important in neoplastic process \[[@B23]\]. The promoter regions of many genes, including RAR-β2, contain CpG islands, that are protected from methylation in normal cells \[[@B24]\]. Aberrant methylation of CpG islands in tumors serves as an alternative pathway (in addition to intragenic mutations and loss of chromosomal loci) for complete inactivation of tumor suppressor genes. Hypermethylation of normally unmethylated CpG islands in the 5\' region of the RAR-β2 gene has been observed in breast and colon cancer cell lines, that do not express RAR-β2 mRNA \[[@B25]-[@B29]\]. In this study we examined whether the methylation could be responsible for the deficit of RAR-β2 gene expression in cervical squamous cell carcinomas. Materials and Methods ===================== Tissue specimens ---------------- Specimens of cervical tumor were collected in the Department of Radiosurgery of Russian Cancer Research Center: 20 specimens of invasive squamous cell carcinoma with low and moderate levels of differentiation and I, II and III FIGO stages and 20 specimens of morphologically normal tissues of cervix from the same patients. All tumors were HPV16- or HPV18-positive by PCR analysis. Cell lines ---------- The human cervical carcinoma cell HeLa, SiHa and CaSki (American Type Culture Collection, Rockville, MD) were maintained in DMEM, supplemented with 10% FCS and 2 mM glutamine. DNA and RNA preparations ------------------------ DNA and RNA were isolated from frozen tissues homogenized in the 4.5 M solution of guanidine thiocyanate by centrifugation through cesium chloride cushion, as described previously \[[@B30]\]. Northern blot analysis ---------------------- 10 μg of total RNA was fractionated in 0.8% formaldehyde agarose gel and transferred to \"Hybond C extra\" nitrocellulose membrane (Amersham, UK) according to Sambrook et al. \[[@B31]\]. The membranes were hybridized with ^32^P-labelled 650 bp and 610 bp EcoRI fragments of RAR-β2-specific DNA. In order to control the relative amounts of RNA the membranes were hybridized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) DNA probe. Hybridization was performed in 5 × SSC at 42°C in the presence of 50% formamide. The membranes were washed in 2 × SSC, 1% SDS at 65°C and were exposed to X-ray film at -70°C. Southern blot analysis ---------------------- 10 μg of DNA was digested overnight with MspI or HpaII restriction enzymes (10 units/μg of DNA, SibEnzyme, Novosibirsk, Russia). The restriction fragments were separated in 1% agarose gel, transferred to nylon membrane (Zeta-Probe, Bio-Rad, USA) and hybridized with ^32^P-labelled PCR DNA probe. PCR DNA probe was prepared by PCR amplification of a 680 bp fragment of RAR-β2 promoter and exon_3 with primers: forward, 5\'-ATT TGA AGG TTA GCA GCC CG and reverse 5\'-CAT TCG GTT TGG GTC AAT CC, in the following conditions: 94°C for 3 min, 30 amplification cycles (94°C, 30 sec; 56°C, 30 sec; 72°C, 1 min) and 72°C, 5 min . Hybridization and washing conditions were as described for RNA. Methylation-specific PCR (MSP) assay \[[@B32]\] ----------------------------------------------- Bisulfite conversion reaction was carried out according to S.J. clark et al. \[[@B33]\] with minor modifications. DNA (2 μg) was digested with RsaI restriction enzyme (SibEnzyme, Novosibirsk, Russia), denaturated by addition of NaOH to a final concentration 0.3 M, incubated for 15 minutes at 37°C and 3 minutes at 95°C, followed by cooling on ice. Sodium bisulfite and hydroquinone were added to the denatured DNA to final concentration 3.6 M and 0.5 mM, respectively. The sample was overlaid with mineral oil and incubated at 65°C for 16 hours. The DNA was purified using QIAEX II kit (Qiagen), desulfonated by addition of NaOH to a final concentration 0.3 M and incubation at 37° for 15 min, neutralized with 3 M ammonium acetate (pH 7.0), precipitated with 3 volumes of ethanol, resuspended in H~2~O and stored at -20°C. Primers that were specific for the upper strand of bisulfite-modified DNA and did not contain CpG dinucleotides in the original sequences were used in the first round of PCR: forward 5\'-AAG TGA GTT GTT TAG AGG TAG GAG GG and reverse 5\'-CCT ATA ATT AAT CCA AAT AAT CAT TTA CC \[[@B28]\]. 5-20 ng of DNA was amplified with these primers in the following conditions: 94°C, 5 min; 35 amplification cycles (94°C, 15 sec; 53°C, 15 sec and 72°C, 30 sec) and 72°C, 5 min. The obtained DNA fragments were amplified in the second round of PCR with two set of primers specific for the upper strand of bisulfite-modified DNA, which contains CpG dinucleotides: 1) forward 5\'-TTG AGA ATG TGA GTG ATT TGA and reverse 5\'-AAC CAA TCC AAC CAA AAC AA for the unmethylated sequence (U), 2) forward 5\'-TCG AGA ACG CGA GCG ATT CG and reverse 5\'-GAC CAA TCC AAC CGA AAC GA for the methylated sequence (M). The PCR was performed under the following conditions: 94°C, 2 min; 40 cycles (94°C, 30 s; 50°C, 30 s and 72 °C, 30 s for U-primers) and 35 cycles (94°C, 30 s; 58°C, 30 s and 72°C, 30 s for M-primers) and 72°C, 5 min. The PCR products were fractionated in the 3% agarose gel and photographed. Results ======= RAR-β2 mRNA expression ---------------------- The level of the expression of RAR-β2 mRNA was evaluated in 20 samples of invasive squamous cell carcinomas of uterine cervix at low and moderate levels of differentiation and in the morphologically normal tumor-adjacent tissues of cervix in order to select a group of tumors with deficient expression of RAR-β2 (Figure [1B](#F1){ref-type="fig"}, Table [1](#T1){ref-type="table"}). The level of RAR-β2 mRNA expression was significantly decreased in 8 out of 20 tumors constituting 5 -- 48% of the expression level of the corresponding morphologically normal tissue. An extremely low (undetectable) level of mRNA expression was observed in one tumor, as well as in the adjacent morphologically normal tissue (Fig. [1B](#F1){ref-type="fig"}, lanes 5,6; Table [1](#T1){ref-type="table"}, case 7). Thus, in 40% of cervical SCC the expression of the RAR-β2 mRNA was significantly decreased. An extremely low level of mRNA expression, wich was observed earlier in three cell lines HeLA, SiHa and CaSki \[[@B22]\], was confirmed in our experiment (data are not presented). {ref-type="table"}. The arrows show the positions of 28S and 18S RNA. **C. Southern blot analysis of DNA isolated from cervical carcinomas (T) and adjacent cervical normal tissues (N).** The blots were hybridized with 680 bp ^32^P labeled DNA probe for RARβ2. The bold arrow shows the positions of a minimal 1.1 kbp fragment, recognized by this probe. Lanes 1, 3, 5, 7, 9, 11, 13, 15 -- DNA digested by MspI; lanes 2, 4, 6, 8, 10, 12, 14, 16 -- DNA digested by HpaII; lanes 1, 2 -- 9N; lanes 3, 4 -- 9T; lanes 5,6 -- 4N; lanes 7, 8 -- 4T; lanes 9, 10 -- 7N; lanes 11,12 -- 7T; lanes 13,14 -- SiHa; lanes 15, 16 -- HeLa. Case numbers correspond to that indicated in table [1](#T1){ref-type="table"}. **D. MSP assay of DNA isolated from cervical carcinomas (T), adjacent cervical normal tissues (N) and leucocytes (L).** All DNA samples were treated with sodium bisulfite before amplification of a 146 bp fragment with methylation-specific primers (M) and unmethylation-specific primers (U). Sample numbers on the top of the gel correspond to that indicated in table [1](#T1){ref-type="table"}; mr -- molecular weight marker.](1471-2407-2-4-1){#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### Methylation Status of RAR β2 5\' region in cervical SCC and cervical cancer cell lines ::: Case number or type of cells RNA expression %^\*^ Southern blot analysis MSP assay ------------------------------ ---------------------- ------------------------ ----------- 1 5 nd M 2 19 M Nd 3 48 M M 4 22 M M 5 26,5 M M 6 18 M M 7 ^\*\*^ M M 8 19 nd M 9-20 95-103 U U HeLa ^\*\*^ M M SiHa ^\*\*^ M M CaSki ^\*\*^ nd M ^\*^ The expression levels of RAR β2 mRNA in carcinomas relative to morphologically normal tissues (%), determinations by Northern blot analysis; the RAR β2 mRNA levels have been normalized to GAPDH. ^\*\*^ Expression of RAR β2 mRNA was extremely low in tumor and morphologically normal tissue or cell lines. nd -- not done ::: Methylation status of the RAR-β2 5\' region ------------------------------------------- The methylation status of the 5\' region of the gene was studied in order to determine if this DNA modification, which inhibits transcription, contributes to the inhibition of the expression of the RAR-β2 mRNA. The study was carried out by two independent methods: by Southern blot analysis and by methylation-specific PCR assay. The localization of the recognition sites for the methylation-sensitive restriction enzyme Hpa II within the promoter region and the first transcribed exon of the RAR-β2 transcript (exon 3) is presented on the fig. [1A](#F1){ref-type="fig"}. The sizes of the fragments, which are produced when the restriction sites are methylated and can not be digested by Hpa II, are indicated in the lower part of the fig. [1A](#F1){ref-type="fig"}. The control digestion of DNA by Msp I (cuts at the identical to Hpa II site even if the cytosine residue is methylated) produced only one major band 1.1 kbp (Fig. [1C](#F1){ref-type="fig"}, lanes 1, 3, 5, 7, 9, 11, 13, 15). The absence of the 1.1 kbp band, as well as that of 1.15 -- 1.6 kbp, after the digestion of the DNA by Hpa II indicates that the H1-H5 sites are methylated (Fig. [1C](#F1){ref-type="fig"}, lanes 8, 10, 12, 14, 16). The presence of 1.1 kbp band and the absence of any additional bands indicates that these sites are unmethylated (Fig [1C](#F1){ref-type="fig"}, lanes 2, 4, 6). As demonstrated by Southern blot analysis, all analyzed sites were methylated in 6 out of 8 carcinomas with the deficit of the RAR-β2 mRNA (Table [1](#T1){ref-type="table"}). It was impossible to perform this analysis in two carcinomas due to lack of adequate amounts of DNA. Methylation of this region was found in the morphologically normal tissue in the case 7 (Fig. [1C](#F1){ref-type="fig"}, lane 10). This result correlated with the absence the mRNA expression in this samples (Fig. [1B](#F1){ref-type="fig"}, lane 5). Two cervical cancer cell lines, showed methylation of this region also (Fig [1C](#F1){ref-type="fig"}, lanes 14, 16). The same samples were subjected to the MSP assay. The MSP assay allows detection of methylated, as well as of unmethylated alleles after the modification of DNA by sodium bisulfite. The presence of nonmalignant cells (fibroblasts, blood cells and some normal epithelial cells) in tumor specimens does not interfere with the ability to detect methylation. The sensitivity of this assay to detect methylated alleles in the background of unmethylated alleles is \> 10^-4^\[[@B32]\]. The result of a typical experiment with a set of primers, that amplify a 146 bp product in the region of H4 and H5 sites of Hpa II (5\' untranslated region, 5\' UTR), is presented on the fig. [1D](#F1){ref-type="fig"}. The presence of an amplification product obtained in the reaction with methyl-specific primers (M) indicates for the methylation (Fig. [1D](#F1){ref-type="fig"}, samples IT, 3T, 7T). The absence of this product indicates for the absence of the methylated alleles (Fig. [1D](#F1){ref-type="fig"}, samples 13T, 9T, 13L and all N samples). Amplification with primers, which are specific for the unmethylated sequence (U), demonstrated that this region is mostly demethylated. It should be noted, that in all tumor samples, that bear the methylated alleles, we detected at least traces of the products of amplification with U primers. Apparently, the nonmalignant cells, that are present in tumor specimens serve as a source of this product. Both primer pairs (M and U) do not react with DNA, which was not treated with bisulfite (data not shown). Thus, the presence of an amplification product in the reaction with the U primers serves as a control of a correct DNA modification by bisulfite in the case of a negative reaction with the M primers. The results of the MSP assay confirmed the methylated status of DNA in the 5\' UTR region of RAR-β2 gene in tumors with decreased level of mRNA expression and in all cervical cancer cell lines examined. (Table [1](#T1){ref-type="table"}). In addition to one case, which was detected by Southern blot analysis, two patients were found to produce a poor amplification with the M primers in morphologically normal tissues (data not shown). Thus, methylation of CpG site in the promoter and the 5\' UTR region of RAR-β2 gene was detected in all 8 tumor samples with deficient expression of the RAR-β2 mRNA and all cervical cancer cell lines. Methylation of this region was not found in the tumors with normal expression levels of the RAR-β2 mRNA. Discussion ========== RAR-β2 is a putative tumor suppressor. Complete or partial inhibition of the expression of the gene was observed in many tumor cell lines and in primary human tumors \[[@B7]-[@B11],[@B22]\]. It was demonstrated in this study, that in 8 out of 20 SCC of cervix a deficient expression of the RAR-β2 mRNA was observed in comparison to the matched morphologically normal tissue of cervix. The mechanism of the inhibition of RAR-β2 expression remains obscure. It was demonstrated in the tumor cells lines derived from breast and colon carcinomas, that the expression of RAR-β2 mRNA is silenced by methylation of the promoter and the first exon, that is expressed in the RAR-β2 transcript. Treatment of DNA by demethylating agent 5-aza-2\'-deoxycytidine restored the inducibility of the gene by all-*trans*-retinoic acid (ATRA) in many, but not all tumor cell lines. For breast carcinomas a relationship between methylation status and decreased RAR-β2 expression was found only in invasive grade II tumors. In grade III tumors the RAR-β2 expression was never observed, regardless the methylation status \[[@B28]\]. In all 8 invasive SCC of cervix and in three cell lines, analyzed in our study, the deficit of the expression of the RAR-β2 mRNA was associated with methylation of the promoter and the region of the first exon, including the H3 site close to the RARE (Figure [1A](#F1){ref-type="fig"}). The methylation of this site plays an important role in the repression of the gene transcription in the breast cancer cells \[[@B29]\]. However, recently 5 cervical cancer cell lines were revealed in that the inhibition of the RAR-β2 mRNA expression was not associated with the methylation of the same region of the gene \[[@B34]\]. Apparently, some other events together with methylation of the promoter region are involved in gene inactivation in certain cancer cell lines and invasive carcinomas. In particular, overexpression of Nur 77 gene has been found to be associated with a loss of RAR-β2 inducibility and RA resistance in lung cancer cells \[[@B35]\]. In our study the methylated alleles of RAR-β2 were absent in the majority of morphologically normal tissues of cervix. However, methylated alleles of the gene were found in three samples of the normal tissue, which were adjacent to the tumors with methylated status of RAR-β2. It can be suggested, that methylation of the normal tissues of cervix is an early preneoplastic event, that favours the development of malignancy. Our data provide only indirect evidence that methylation and the loss of RAR-β2 expression are early steps of cervical carcinogenesis. Recently this suggestion was supported by the study, where the methylated alleles of the gene were found in 11% and 29% of low grade and of high grade cervical intraepithelial neoplasias respectively \[[@B34]\]. This authors found the methylation of RAR-β2 gene in 5 out of 17 invasive carcinomas also. Taken together these data indicate that methylation of RAR-β2 gene is frequent and early event during cervical cancerogenesis. Conclusions =========== These findings suggest that methylation of the 5\' region of RAR-β2 gene may contribute to gene silencing and that methylation of this region may be an important and early event in cervical carcinogenesis. Cervical carcinomas may serve as a potential target for the DNA methylation inhibitors due to a close correlation between the deficit of RAR-β2 mRNA expression and methylation of this gene. These inhibitors in combination with retinoids might be clinically useful in future as chemopreventive and chemotherapeutic agents for patients with cervical dysplasias and carcinomas. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We thank Dr. O.V. Sacharova for providing the clinical information and Dr. A.F. Kisseleva for technical assistance. This work was supported by grants 97-04-50140 and 01-04-49225 from the Russian Foundation of Basic Science.

Background ========== Interest in the peer review process and research aimed at determining the method of obtaining the best quality reviews has grown in recent years. \[[@B1]\] Checklists have been developed that aim to guide reviewers when assessing the quality of papers, but little evidence exists concerning the extent of agreement between two referees when evaluating the same paper. In addition, little is known about which dimensions of a checklist are likely to result in greater agreement between referees. There were two aims of this study: (1) to examine inter-rater agreement of the quality of abstracts submitted to a primary care research conference (Annual Meeting of the South West Association of University Departments of General Practice, Exeter 2000, UK), and (2) to compare the scores of abstracts accepted and rejected for the meeting. Materials and Methods ===================== Abstracts were anonymised and scored using a structured assessment comprising seven categories: (1) importance of the topic (2) originality (3) overall quality of the study design (4) appropriateness of the design used (5) achievement of aim (6) contribution to academic primary care (7) likelihood of provoking discussion. For comparison purposes, we have classified the assessment of categories 1, 2, 6 and 7 as more \'subjective\' in nature, and categories 3, 4 and 5 as more \'objective\'. Between one (poor) and four (excellent) marks were awarded for each category, giving a maximum possible score of 28 marks. Every abstract was assessed independently by two referees (AM and AG). Agreement between referees was assessed using intraclass correlation coefficients (ICC), a chance corrected measure of agreement. \[[@B2]\] The ICC indicates perfect agreement only if the two assessments are numerically equal and is preferable to the more usual (Pearson) correlation coefficient. The crude ICC is lowered by any systematic differences between referees\' scores. In terms of a plot of the two referees\' scores, a line with a non-zero intercept will further lower the ICC irrespective of any disagreement, represented by deviation of the slope of the line away from unity and scatter around the line. In a further analysis, this effect was investigated by subtracting the mean difference for each component from the higher of the two referees\' scores. The ICCs were then recalculated, giving estimates of agreement corrected for both systematic differences and chance. There are no universally applicable standard values for the ICC that represent adequate agreement, but the following convention is used here to aid interpretation: ICC \<0.20 \'slight agreement\'; 0.21--0.40 \'fair agreement\'; 0.41--0.60 \'moderate agreement\'; 0.61--0.80 \'substantial agreement\'; \>0.80 \'almost perfect agreement\'. Scores from referees from three different institutions were summed to give each abstract an overall score. Abstracts were ranked by this overall score and the top 45 were accepted for oral presentation at the meeting. Of the 52 abstracts refereed by AM and AG, mean total scores of those accepted and rejected for the meeting were compared using an unpaired t test. Results ======= Chance corrected agreement between the two referees\' scores measured using crude ICCs was greater for the three components relating to design and execution of the study (Table [1](#T1){ref-type="table"}: items 3 to 5) compared to those relating to more subjective elements of the abstract (Table [1](#T1){ref-type="table"}: items 1, 2, 6, 7). After adjustment for systematic differences in referees\' scores, ICCs for items 3 to 5 remained highest, demonstrating fair to moderate agreement. ::: {#T1 .table-wrap} ::: {.caption} ###### Inter rater agreement between two referees for 52 abstracts submitted for a primary care research conference ::: **Component** **Mean difference (AG minus AM)** **Crude ICC** **Adjusted ICC** ------------------------------------------- ----------------------------------- --------------- ------------------ 1\. importance of the topic 0.71 0 0.24 2\. originality 0.27 0 0.01 3\. overall quality of the study design 0.40 0.30 0.40 4\. appropriateness of the design used 0.17 0.40 0.41 5\. achievement of aim -0.12 0.44 0.45 6\. contribution to academic primary care 0.25 0.20 0.25 7\. likelihood of provoking discussion 0.15 0.22 0.24 Overall score 1.85 0.31 0.41 ICC = intraclass cluster coefficient ::: A total of 76 abstracts were submitted for the meeting. Of 52 received by the authors for assessment, 26 were accepted for oral presentation (Table [2](#T2){ref-type="table"}). Abstracts accepted for the meeting had a significantly higher mean score than those that were rejected (95% CI for difference 1.3 to 4.1, p = 0.0003) (Table [2](#T2){ref-type="table"}). ::: {#T2 .table-wrap} ::: {.caption} ###### Summary statistics of abstracts accepted and rejected for oral presentation at a primary care research conference ::: **N** **Mean score** **SD** **Range (possible 7 to 28)** ---------- ------- ---------------- -------- ------------------------------ Accepted 26 17.4 2.7 11 to 21.5 Rejected 26 14.6 2.3 11 to 19.5 ::: Discussion ========== This study has shown that when using a structured assessment form, two independent reviewers were more likely to agree on design or methodological components of a checklist than on subjective components of abstracts submitted for an annual research meeting. Abstracts accepted for the meeting had significantly higher total scores, but overlapped considerably with rejected abstracts. This was due to acceptance for the meeting being determined by an overall aggregate of scores awarded by referees from three institutions. While the subject of inter-reviewer agreement on different components of a checklist is relatively under-researched, some previous studies offer support for our finding that agreement is better when reviewers can be more objective in their assessments. Among a group of reviewers asked to rate a series of review articles, agreement on scientific quality of the papers was very high (60% of ICCs \> 0.7) both within and between groups with varying levels of research training and expertise. \[[@B3]\] All 10 dimensions of the checklist that reviewers rated could be regarded as objective. Divergent reviewers have been identified in a study comparing an overall rating score that indicated a recommendation to publish rather than individual dimensions of a review checklist. \[[@B4]\] This study does have limitations. Importantly, we assessed agreement between only two reviewers on a relatively small number of abstracts. This could be addressed by having more abstracts assessed by a greater number of reviewers. However the study was conducted pragmatically within the time and administrative constraints of a small annual scientific meeting rather than submissions to a journal over an extended period. Another limitation is that the reviewer checklist was constructed prior to conceiving the study. If future meetings are to be used to investigate the content of structured reviewer assessments, such checklists should be constructed with specific hypotheses in mind. Characteristics associated with good peer review are age under 40 years and training in epidemiology or statistics, \[[@B5]\] characteristics that applied to both reviewers in the present study. Structured assessment forms that ask the reviewer for their opinion of a paper\'s interest, originality or likelihood of provoking discussion may be more likely to result in scores that reflect the reviewer\'s own research interests. This is not necessarily a criticism -- it is perhaps only natural that individuals will differ in their opinions of how interesting they find, and think others will find, a particular paper. It is interesting that the two components with the lowest agreement, importance of the topic and originality of the study, both require more knowledge about a specific subject area than either of the other two subjective questions. Journal editors and meeting organisers should be aware that including subjective components in review checklists may result in greater disagreement between reviews. Conclusions =========== This study provides some evidence that inclusion of subjective components in a review checklist may result in greater disagreement between reviewers. An interesting area for further research would be to investigate the effects of attaching different weights to subjective and objective components of a checklist, or to exclude subjective components altogether from overall quality scores and simply use them a guide to acceptance or rejection. Competing interests =================== None declared. {#F1} Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We would like to thank Mrs Sylvia Smith for collation of all referees\' responses. During the period in which this work was conducted, AM was supported by a UK Medical Research Council Training Fellowship in Health Services Research. TF is supported by a UK NHS R&D Primary Care Career Scientist Award.

Background ========== During the first days after stroke, between one and two fifths of the patients develop fever or subfebrile temperatures. \[[@B1]-[@B4]\] Increased temperatures have been associated with relatively large infarct volumes, high case fatality, and poor functional outcome, even after adjustment for initial stroke severity. \[[@B2]-[@B6]\] The period in which hyperthermia is associated with poor outcome is probably limited to the first 12 or 24 hours from stroke onset. \[[@B6],[@B7]\] The harmful effects of an early rise in body temperature have been attributed to increased cerebral metabolic demands, \[[@B8]\] changes in the blood-brain barrier permeability, acidosis, and an increased release of excitatory amino acids. \[[@B9]\] In animal models of temporary focal cerebral ischemia, mild intra-ischaemic hyperthermia increased infarct volume, \[[@B10]\] whereas mild hypothermia reduced infarct size. \[[@B11]\] The above suggests that a pharmacological reduction of temperature in patients with acute ischaemic stroke may improve their functional outcome. Previously, we studied the effect on body temperature of high (6 g daily) and medium dose (3 g daily) paracetamol (acetaminophen) in a randomised placebo controlled trial of 3 × 25 patients (PAPAS: paracetamol (acetaminophen) in patients with acute stroke). \[[@B12]\] In the high dose paracetamol group, mean body temperature at 12 and 24 hours after start of treatment was 0.4°C (95% CI:0 -- 0.8°) lower than in the placebo group. Koennecke conducted a comparable study of 40 patients with acute stroke. He showed that treatment with paracetamol in a daily dose of 4000 mg resulted in a substantial reduction of the proportion of patients with body temperatures over 37.5°C, but he did not estimate the size of the temperature reduction. \[[@B13]\] Kasner et al. conducted a partly blinded trial of paracetamol against placebo in 39 patients with haemorrhagic or ischemic stroke, and oberved a difference of 0.2°C in body temperature. The patients who received paracetamol tended to be more often hypothermic (\<36.5°C) and less often hyperthermic (\>37.5°C), but these results were not statistically significant. \[[@B14]\] In this study, a lower dose (3.9 g daily) was used than in our study (6 g daily). Several studies in adults and children with fever suggest a larger antipyretic effect from ibuprofen than from paracetamol. \[[@B15]-[@B17]\] The effect of ibuprofen on body-core temperature in normothermic and subfebrile patients with stroke, who may have a disturbance of thermoregulation,\[[@B18]\] has not been studied. Pharmacological properties and toxicity of ibuprofen ---------------------------------------------------- Ibuprofen is a prostaglandin-synthetase inhibitor with analgesic, anti-inflammatory, and antipyretic features. It belongs to the group of propionic acid derivatives. It is usually well tolerated, has almost no side effects or drug interactions at therapeutic doses, and it affects the platelet aggregation in doses of more than 1 g by reversible inhibition of cyclo-oxygenase. After oral administration, ibuprofen is rapidly absorbed; the maximum concentration is reached after 1 to 2 hours. Rectal administration leads to a slightly slower absorption. Ibuprofen is bound for 90% to 99% to plasma proteins, but at usual concentrations only a small part of the binding sites is occupied. Ibuprofen is mainly distributed in the plasma compartment; the distribution volume is approximately 0.15 I/kg body weight. Elimination through the urine is almost complete within 24 hours. The metabolites do not have any detectable activity. More than 90% of a dose is found back as metabolites and their conjugates in the urine. Less than 1% is excreted as unchanged ibuprofen. The plasma half-life is 1.5 to 2.5 hours. Gastrointestinal side effects, including epigastric pain, nausea, heartburn, and sensations of fullness, are experienced by 5 to 15% of patients taking ibuprofen. The frequency of these side effects is lower than that of indomethacin or aspirin. A meta-analysis even showed that there is no significant difference in adverse effects in patients treated with ibuprofen in daily doses up to 1200 mg compared with patients given placebo. \[[@B19]\] Estimates of the incidence of gastro-intestinal haemorrhage in patients with acute stroke range from 0.1% to 3%. \[[@B20],[@B21]\] Whether treatment with non-steroidal anti-inflammatory drugs (NSAID) increases the risk of this complication in these patients is likely, but unknown. Other side effects have been reported incidentally. These include toxic amblyopia, fluid retention, edema, skin rashes, and thrombocytopenia. Several cases of aseptic meningitis, probably caused by ibuprofen, have been reported. Allopecia has been described in negroid women who also used relaxing hair lotions and steroids simultaneously. Teratogenic effects have not been reported. \[[@B22]\] Pharmacological properties and toxicity of paracetamol ------------------------------------------------------ Paracetamol (N-acetyl-p-aminophenol) is a para-aminophenol derivative and the major metabolite of phenacetin. It has a weak inhibitory effect on peripheral biosynthesis (which accounts for its weak anti-inflammatory activity), but it is a potent inhibitor of prostaglandin production within the central nervous system. This presumably accounts for its analgesic and antipyretic properties. It is rapidly absorbed through the gastrointestinal tract and uniformly distributed through most body fluids. After oral administration peak plasma levels are reached after 30 min to 1 hour. \[[@B23]\] Rectal administration leads to a slightly slower absorption, but effective plasma levels are reached after approximately one hour. \[[@B24],[@B25]\] Acetaminophen is mainly conjugated in the liver, and then excreted in the urine. Plasma half-life is 1.5 to 3 hours. Following a dose of 140 mg/kg bodyweight or more, the glucuronide pathway may become saturated, and (hepato-) toxicity may result. In patients with chronic liver failure, the threshold for toxicity is lower. \[[@B23]\] These patients will therefore be excluded from the present study. Methods and design ================== This study is a three-armed, randomised, double-blind, placebo-controlled trial, performed in the university hospitals of Rotterdam and Utrecht, and in the regional hospital \"Eemland\" in Amersfoort, The Netherlands. Aim of the study ---------------- To study the effect of high-dose ibuprofen and to confirm the previously observed reducing effect of high-dose paracetamol on body temperature, and to study their safety in normothermic and subfebrile patients with acute ischaemic stroke. Study medication ---------------- Patients will be treated for 5 days with paracetamol or ibuprofen in daily doses of 6 × 1000 mg or 6 × 400 mg, respectively, or with placebo. The study drug will be administered through identical capsules. The first gift of medication may be given as a suppository. The study medication will be provided in white paper boxes, numbered consecutively with a medication number. Each box contains 70 identical capsules containing paracetamol 500 mg, ibuprofen 200 mg, or placebo, and one suppository, containing paracetamol 1000 mg, ibuprofen 400 mg, or placebo. Every gift of medication consists of 2 capsules except the first gift, which is one suppository. The study-population will be stratified according to the time since onset of stroke (0 to 12 hours vs 12 to 24 hours). Randomisation and treatment schedule ------------------------------------ Treatment allocation will be random. The randomisation will be blocked in lots of six. A computer-generated list containing the medication number and information on the nature of the treatment (placebo, paracetamol, or ibuprofen) will reside with the independent safety committee and with the study pharmacist. Treatment allocation of any patient who is in the study will be disclosed before termination of the study if the safety committee deems this necessary. Local clinical investigators may apply for disclosure of the treatment allocation of an individual patient by the study pharmacist. This will be reported to the safety committee and to the steering committee by the study pharmacist. Pain should not be treated with NSAID. We advise codeine, 4 to 6 times 10 mg daily, or tramadol, 4 times 25 to 50 mg daily. When the treating physician judges that fever should be treated in a patient, the study medication should be stopped. The clinical investigators are advised to prescribe low-dose aspirin in all patients. Outcomes -------- The primary outcome will be body temperature at 24 hours from start of treatment. Secondary outcomes will be change in baseline temperature at 1 and 5 days from start of treatment, and time with elevated body temperature (\>37.0°C) (area under the curve) during the first 24 hours and the first five days. The tertiary outcome is functional outcome at 1 month, as determined by the scores on the modified Rankin Scale (mRS) and Barthel Index (BI). Setting ------- Ziekenhuis Eemland Amersfoort is a 600 bed regional hospital. Annually, 150 patients with acute stroke are admitted. The University Medical Centre Utrecht has 1043 beds, and admits a similar number of patients with acute ischaemic stroke. The University Hospital Rotterdam has 1229 beds. Annually, 250 patients with acute stroke are admitted. All three hospitals have an acute stroke unit and a dedicated stroke team. Study organization ------------------ The study personnel consists of an executive committee (RJM, EJB, DWJD, HMAvG, HBvdW), an advisory committee (PJK, UK), and an independent data-monitoring and safety committee. The co-ordinating centre for data management is located in Rotterdam. The investigators are responsible for patient selection and inclusion, and data acquisition. The investigators will notify the data management centre of newly included patients and of serious adverse events within 72 hours. The data-management centre is responsible for prompt data entry. Every week, a progress report entailing all serious adverse events (infection, gastro-intestinal haemorrhage, liver function disturbance, neurological deterioration, death, or other) will be provided by the executive committee. These data will be combined with the randomisation scheme by the secretary of the data-monitoring and safety committee, and summarised in a simple table. This table is forwarded to the other members of the data-monitoring and safety committee. The members will acknowledge receipt of the summary table. Upon receipt of the safety data, each member of the safety committee will notify the advisory committee of his approval of continuation of the study. If one or more members of the safety committee express a concern with regard to the safety of the study treatment, the chairman of the safety committee will summon the members of the safety committee and try to reach a consensus or majority advise. In light of the small number of patients in the study, no formal statistical stopping rule is given. Time path --------- In the previous study carried out in the same centres with similar inclusion criteria, it took 8 months to include 75 patients, i.e. 2 patients per week. It is therefore not unrealistic to take eight months for patient inclusion, one month to complete the follow-up and to close the database, and one month to report the study. The study has started in December 2000, and is expected to be completed at the end of the year 2001. Inclusion criteria ------------------ Patients will be included if they have an acute ischaemic anterior circulation stroke, a body temperature \>36.0°C or \<39.0°C, a CT scan that is compatible with acute ischaemic stroke, a focal deficit without rapid improvement, and a possibility to start treatment within 24 h after stroke onset. Exclusion criteria ------------------ Criteria for exclusion from this study are: 1) severe aphasia, defined as an aphasia score of 2 or 3 on the National Institutes of Health Stroke Scale (NIHSS); 2) treatment with an NSAID deemed necessary; 3) hypersensitivity to ibuprofen or paracetamol, (chronic) liver failure or cirrhosis; 4) (chronic) renal failure; 5) history of alcohol abuse; 6) active gastric ulcer disease or a history of peptic ulceration or gastro-intestinal hemorrhage in the preceding year; 7) colitis ulcerosa; 8) pregnancy; 9) use of corticosteroids; 10) a severe concomitant medical condition that could affect the assessment of the effect of the study medication on temperature; 11) residual neurological impairment resulting from a previous stroke that may hamper the assessment of functional outcome; 12) death appearing imminent, and 13) no informed consent given. Study activities ---------------- In the first 24 hours, tympanic temperature will be measured in both ears every 2 hours. During the remaining 6 days, temperature will be measured twice a day. On the first day, rectal temperature will be measured at start of treatment (t = 0) and at t = 24 hours. Day 0 is defined as the time period between onset of stroke and inclusion in the study. All baseline investigations are therefore carried out on day 0. Day 1 commences directly after the start of treatment with the study medication. All time-periods in the study are measured relative to the time of start of treatment. At baseline, the medical history, including previous strokes or transient ischaemic attacks, will be assessed and a general and neurological examination will be carried out. The NIHSS is used to assess stroke severity at onset. The mRS is used to determine the patients\' pre-stroke functional status. Laboratory investigations will include a full blood count, glucose, electrolytes, creatinin, AST, ALT, AF, LDH, γGT, total bilirubin, APTT and PTT, and CRP. A brain CT will be done before inclusion in the study. Body-temperature will be measured with a tympanic thermometer\[[@B26],[@B27]\] at 2-hourly intervals during the first 24 hrs after start of treatment, and thereafter every 12 hours until day 7. Temperatures will be taken from both ears, and averaged. The baseline temperature (at the start of treatment) and temperature at 24 hours after start of treatment will be determined by both rectal and tympanic thermometry. Compliance will be measured by counting leftover pills, and by a registry of the medication that was actually taken. Safety laboratory assessments will be carried out on day 6, including hemoglobin, AST, ALT, AF, LDH, γGT, total bilirubin and CRP. At the follow-up visit on day 28, mRS is used to assess functional outcome. \[[@B28]\] The score on the BI will also be assessed. Based on the neurological examination and the results of CT, duplex, and other studies, etiological stroke type will be determined according to the (modified) Trial of Orgaran in Acute Stroke (TOAST) classification. \[[@B29]\] Safety concerns and adverse events ---------------------------------- One should be careful not to overlook infections in patients who are treated according to the study protocol. Therefore, treating physicians are advised to lower their thresholds for clinical suspicion of infection to 38°C, to monitor the heart rate carefully, and start diagnostic studies and antibiotic treatment earlier than usual. The decision to start antipyretic mediciation in feverish patients is left to the discretion of the treating physician. On such an occasion, the study medication should be stopped. Serious adverse events are defined as any deterioration in health status that is potentially life threatening, and death, within the study-monitoring period (day 0 to day 7). This includes any life-threatening infection, such as pneumonia, sepsis, or urinary tract infection that needs antibiotic treatment according to the judgement of the treating physician, any liver function disturbance (AST, ALT, AF, or total bilirubin levels exceeding twice the local upper limit of normal), gastro-intestinal haemorrhage, and neurological deterioration (i.e. decrease in consciousness level of more than one point on the Glasgow coma scale, or a decrease of two or more points on the NIHSS). All serious adverse events have to be reported within 72 hours to the independent data-monitoring and safety committee. Ethical considerations ---------------------- Each patient will be given written information about the study. Informed consent will be obtained verbally and in writing before inclusion. If the patient is unable to write, written consent will be obtained from an appropriate source (spouse, child, or parent). The medical ethics committees of the three hospitals have approved this protocol. Analysis -------- The study results will be analysed on intent-to-treat basis, but an \"on-treatment\" analysis of primary and secondary endpoints will be carried out. There will be no formal interim analysis. The main results of the study will be presented as the mean difference in temperature between each treatment group and the placebo group. The precision of these estimates will be expressed with 95% confidence intervals, based on the *t* distribution. No adjustments for multiple comparisons will be made. Multiple linear regression will be used to adjust for possible confounding factors, such as age, stroke type, and stroke severity. The main analysis will be carried out according to the intent to treat principle, but an on treatment analysis will be provided. The analysis will be carried out with the STATA 7.0 statistical package. \[[@B30]\] Sample size ----------- Seventy-five (3 × 25) patients will be included in the study. In a recent cohort study of 162 patients admitted within 48 hours from onset, conducted at the University Hospital Rotterdam, mean rectal temperature was 36.6°C, with a standard deviation of 0.6°C. In order to detect a difference in core body temperature at 24 hours from onset of 0.5°C, with a significance level of α = 0.05, and power 1--β = 0.80, we need at least 23 patients in each treatment group; this is rounded to 25. The primary outcome occurs 24 hours after start of treatment. We expect more than 90% compliance in this period. We therefore need only 2 extra patients per treatment for lost patients in the on-treatment analysis. Publication of study results ---------------------------- All publications will be authored by the members of the executive committee and the advisory committee. Co-authors are supposed to have contributed to the design, the analysis and actual reporting of the study; their centres are expected to include patients roughly proportional to the number of acute stroke patients that are admitted annually. Discussion ========== We report the protocol of a phase II trial that aims to assess the feasibility of early temperature reduction in normothermic or subfebrile patients with acute stroke. This study intends to confirm the previously observed temperature-reducing effect of paracetamol (6 g daily) and to assess the effect of ibuprofen 2.4 g daily. The study is powered to find a difference of 0.5°C between active treatment and placebo. Half a degree Celsius does not seem much by itself, but body temperature has been shown to be a very strong prognostic factor in patients with acute stroke. For every degree Celsius of body temperature the risk of poor outcome increases by 2.2. The earlier the increase in temperature occurs, the stronger the relationship between brain damage and high temperature seems to be. \[[@B6],[@B7]\] Because of the consistently observed relation between hyperthermia and poor outcome after stroke, it is commonly advised to treat all stroke patients with a raised temperature with antipyretic drugs. \[[@B31],[@B32]\] However, a causal relationship between hyperthermia and poor outcome has never been demonstrated, and subfebrile temperatures and fever may just be epiphenomena of severe brain damage. In addition, there is no evidence that a pharmacological reduction of increased temperatures will lead to an improved outcome. Preventive administration of antipyretics may even mask infections and lead to delayed treatment with antibiotics. For these reasons, we think that the efficacy of antipyretic drugs to improve outcome after stroke has to be proven in a randomised controlled trial before treatment with these agents is accepted as a standard procedure. For the same reasons, we consider it ethical to withhold antipyretics from stroke patients with subfebrile temperatures in the placebo arm of such a trial. When the effect of paracetamol or ibuprofen on body temperature in acute stroke has been established, a large, pragmatic phase III study will be undertaken to assess the effect of these cheap and well known drugs on functional outcome after stroke. The PISA investigators are ========================== Executive committee: EJ van Breda, DWJ Dippel, RJ Meijer, HB van de Worp, HMA van Gemert. Data-management: EJ van Breda, H Hilkemeijer, L Kasijo. Statistical analysis: DWJ Dippel. Data monitoring and safety committee: A Algra, (chairman), J van Gijn, PA van Doom, A Koudstaal-Overduin (secretary). Advisory committee: PJ Koudstaal, LJ Kappelle. Writing committee: EJ van Breda, HB van der Worp, HMA van Gemert, RJ Meijer, LJ Kappelle, PJ Koudstaal, and DWJ Dippel. Participating centers: Ziekenhuis Eemland Amersfoort, The Netherlands: HMA van Gemert, A Hovestadt, JBS Boringa. University Hospital Rotterdam: DWJ Dippel, MPJ van Goor, RJ Meijer, D Siepman, E Maasland, SLM Bakker, F van Kooten, A Wijnhoud. University Medical Centre Utrecht: LJ Kappelle, HB van der Worp. Competing interests =================== None declared. Financial support ================= The study is funded by the Stichting Neurovasculair Onderzoek Rotterdam. List of abbreviations ===================== NSAID -- non-steroidal anti-inflammatory drug mRS -- modified Rankin Scale BI -- Barthel Index NIHSS -- National Institutes of Health Stroke Scale AST -- aspartate aminotransferase ALT -- alanine aminotransferase AF -- alkaline phosphatase LDH -- lactate dehydrogenase γGT -- gamma glutamyl transpeptidase APTT -- activated partial thromboplastin time PTT -- prothrombin time CRP -- C-reactive protein CT -- computed tomography TOAST -- Trial of Orgaran in Acute Stroke Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ Special thanks to DR Paul Janssen, pharmacologist at the University Hospital Rotterdam, who prepared the study medication and gave some useful advice on the study design, and to Dr Paul Mulder (Dept of Epidemiology and Biostatistics), who provided the randomisation codes.

Background ========== The weak opioid codeine is widely used in the management of pain. Various studies have demonstrated a synergistic analgesic effect of an opioid-NSAID combination \[[@B1]-[@B4]\], particularly if repeated doses are given. \[[@B5]-[@B7]\] The synergistic effect is thought to be caused by the known different pharmacodynamic mechanisms, opioids acting via opioid receptors in the central nervous system, NSAIDs affecting the synthesis of prostaglandins due to inhibition of the enzyme cyclooxygenase. NSAIDs have also been postulated to display additional antinociceptive effects in the central nervous system. \[[@B8]\] Moreover, a synergistic effect is also possible via a pharmacokinetic interaction between the two classes of drugs: NSAIDs may decrease the renal excretion of the pharmacologically active metabolite of morphine M-6-G (morphine-6-glucuronide).\[[@B9]\] We have previously demonstrated a marked inhibition of codeine glucuronidation in human liver tissue homogenate by diclofenac (K~i~ of 7.9 μM).\[[@B10]\] Others have shown \[[@B11]\] a more than 50% inhibition of the glucuronidation of the codeine structural analogue, dihydrocodeine, by 50 μM diclofenac in vitro. Codeine is predominantly metabolised by glucuronidation to C-6-G (codeine-6-glucuronide) (Figure [1](#F1){ref-type="fig"}). Minor metabolic pathways include N-demethylation to norcodeine and O-demethylation to morphine.\[[@B12],[@B13]\] The latter is catalysed by the polymorphically expressed CYP2D6. \[[@B14]-[@B16]\] There is increasing evidence, that the analgesic effect of codeine is mediated by its O-demethylated metabolite morphine \[[@B17],[@B18]\] and that the glucuronidated metabolite M-6-G possesses even greater analgesic potency than morphine itself.\[[@B19]\] In humans, the analgesic activity of C-6-G has not been reported; however, antinociceptive responses after intracerebroventricular administration have been reported in rats. \[[@B20]\] Since in vitro findings may not necessarily be of clinical relevance, we aimed to investigate whether diclofenac inhibits codeine glucuronidation *in vivo* in healthy volunteers in terms of pharmacokinetics, analgesic efficacy and side effects. {#F1} Methods ======= Subjects -------- Twelve healthy male volunteers, median age 31 years (range 26--42 years), median weight 77 kg (range 63--95 kg), median BMI 23.9 kg/m^2^ (range 19.9--28.4 kg/m^2^) participated in the study. One subject had to be replaced because of additional drug intake during the study. The study was approved by the ethics committee of the local medical board (Landesärztekammer Baden Württemberg) according to the Declaration of Helsinki (1996 Sommerset). All volunteers gave their written informed consent prior to study inclusion. The volunteers were healthy according to history, physical examination and laboratory tests, had no history of drug abuse and did not take any regular medication. The volunteers had previously been genotyped for CYP2D6 by allele-specific PCR \[[@B21],[@B22]\] and all were predicted to be extensive metabolisers (EM). In addition, the subjects were phenotyped with respect to sparteine oxidation polymorphism with a single dose of 100 mg sparteine sulphate and were classified as EMs according to the metabolic ratio (MR) of sparteine and its 2- and 5-dehydrometabolite.\[[@B23]\] In order to limit the effect of the variability of CYP2D6 on metabolic capacity, volunteers with a MR\<1 were generally included in the study, with one subject having a MR of 2.8 (intermediate metaboliser). Study design ------------ The study was designed as a randomised, placebo-controlled, double-blind, cross-over trial. Randomisation was performed using the computer program Sampsize 2.0 (Blackwell Science Ltd., Machin, Campbell, Fayers, Pinol). Each volunteer received on the study days 1 and 8 in random order either 100 mg codeine phosphate + 50 mg diclofenac sodium **or** 100 mg codeine phosphate + placebo. Codeine phosphate was provided as tablets (Codeinum Phosphoricum forte Compretten^®^, Glaxo Welcome GmbH/Cascan GmbH, Hamburg, Germany), diclofenac sodium and placebo were provided as capsules of identical appearance (manufactured by Contract Pharma GmbH & Co KG, Murr, Germany). Blood and urine sampling ------------------------ On study days 1 and 8 the following identical procedures were carried out: Blood samples were taken before drug administration as well as 0.25, 0.5, 0.45, 1, 1.5, 2, 2.5, 3, 4, 6, 8, 10, 12, 24, 35 and 36 hours after drug administration. Blood samples were centrifuged after 30 min. and serum was stored at -20°C until analysed. Urine was collected before drug administration and from 0--12, 12--24 and 24--36 hours after drug administration. The volume was measured and an aliquot of each collection was stored at -20°C until analysed. Determination of analgesic effect --------------------------------- The analgesic effects of codeine with and without simultaneous administration of diclofenac or placebo were assessed using the cold pressor test as previously described.\[[@B17]\] Briefly, the cold pressor test apparatus consisted of temperature-controlled water baths of 35 ± 0.5°C and 1.0 ± 0.5°C (ice-water bath). The nondominant forearm was placed into the warm-water bath for exactly 2 minutes. Fifteen seconds before transferring the forearm into the cold-water bath, a blood pressure cuff was inflated to 20 mmHg below the diastolic blood pressure and the eyes were covered with eye-patches. Subjects placed their forearm in a fixed position with the fingers wide apart into the cold-water bath (for a maximum time period of 2 min). They were instructed to clearly indicate the time of the first pain sensation as well as the time of intolerable pain, at which the forearm was removed from the cold-water bath. The time from the immersion of the forearm into the cold-water bath to the first pain sensation was measured, and is defined as the pain threshold. The pain tolerance is defined as the time from immersion of the forearm into the cold-water bath until the time of intolerable pain. A training session was carried out before the study to screen out volunteers with a pain tolerance of less than 15 sec or more than 120 sec and to familiarise the volunteers with the study procedure. The pain experiments were conducted on study days before drug administration (10--15 min before drug administration) and 1, 1.5, 2, 2.5, 3, 3.5, 4 and 6 hours after drug administration. Subsequently, the changes from baseline were determined and for each study day the area under the pain threshold change versus time curve and the area under the pain tolerance change versus time curve were calculated. Determination of adverse events ------------------------------- Adverse events were evaluated by the following method: The volunteers were required to list each symptom (fatigue, headache, dizziness, blurred vision, nausea, itching, exanthema and any other symptom) on a visual analogue scale (VAS) rated from 0 (not present) to 10 (most severe) at zero h (baseline, before drug intake), 2, 4 and 6 hours after drug intake. The values were corrected for the baseline measurement at t = 0 h. Analytical measurement of codeine and its metabolites ----------------------------------------------------- Codeine and its metabolites were determined using HPLC-electrospray mass spectrometry analogous to a previously published method\[[@B24]\] with minor modifications. Samples (1.0 ml of plasma, or 100 μl of urine diluted with 900 μl of water) were spiked with the deuterated internal standards (10 μl of standard mix with 50 pmol/μl codeine-G-glucuronide-d~3~, 5 pmol/μl codeine, 10 pmol/μl morphine-3-glucuronide-d~3~, 5 pmol/μl morphine-6-glucuronide-d~3~ and 1 pmol/μl morphine-d~3~) and extracted automatically using end-capped C~2~ solid-phase extraction columns. For the determination of the conjugates of norcodeine and normorphine, urine was hydrolysed\[[@B25]\] prior to extraction. The mobile phases used for HPLC were: (A) 1 % acetic acid in water and (B) 1 % acetic acid in acetonitrile. HPLC separation was achieved on a LiChrospher 100 RP-18 (Merck, Darmstadt, Germany) end-capped analytical column (125 × 3 mm I.D., 5 μm particle size) at a flow rate of 0.5 ml/min using a linear gradient from 8 % B to 40 % B in 8 min. The mass spectrometer (HP 1100 MSD, Hewlett-Packard, Waldbronn, Germany) was operated in the selected ion monitoring mode using the respective MH^+^ ions, *m/z* 476 for codeine-6-glucuronide, *m/z* 300 for codeine, *m/z* 303 for codeine-d~3~, *m/z* 286 for norcodeine, *m/z* 462 for morphine-3-glucuronide and morphine-6-glucuronide and *m/z* 465 for the deuterated morphine glucuronides, *m/z* 286 for morphine, *m/z* 289 for morphine-d~3~ and *m/z* 272 for normorphine. The limits of quantification achieved with this method were 0.5 pmol/ml for the morphine-glucuronides and morphine, 2 pmol/ml for codeine, normorphine and norcodeine and 5 pmol/ml for codeine-6-glucuronide with coefficients of variation less than 12 %. Analytical measurements of diclofenac ------------------------------------- Following the addition of 100 μl of a 10 μg/ml solution of 4\'-methoxydiclofenac in water for internal standardisation and 2 ml 10 mM phosphate buffer pH 6.5 to 1 ml of serum, samples were extracted by an automated solid phase extraction procedure using an ASPEC XL (Gilson, Villiers Ie Bel, France). Samples were applied on Bakerbond SPE™ C~18~ extraction columns (3 ml, 500 mg) which were equilibrated with 3 ml methanol and 3 ml of phosphate buffer. The columns were washed with 2 ml of phosphate buffer and 2 ml of phosphate buffer: methanol 6:4 (v/v) and eluted with phosphate buffer: methanol 2:8 (v/v). The solvents were evaporated and the residue was resuspended in 150 μl of the mobile phase. A 100 μl aliquot was injected on a HPLC-system consisting of an autosampler SIL 9A, a solvent delivery system LC 9A and an UV-detector (Shimadzu, Duisburg, Germany). Separation was performed on a Waters Spherisorb ODS1 5 μm 125 × 4.6 mm column equipped with a guard column filled with the same material (Bischoff, Leonberg, Germany) using 8 mM tetrabutylammonium bromide in 10 mM phosphate buffer pH 6.5: 65:30:5 (v/v/v) at a flow rate of 1 ml/min. Diclofenac and the internal standard 4\'-methoxydiclofenac were detected at 282 nm at a retention time of 11.0 and 12.4 min, respectively. Recovery was between 85 and 100.5%. Calibration curves were linear over a concentration range from 10 to 2000 ng/ml (r^2^ ranging from 0.9993 to 1). Inter-assay variability (n = 9) was 14.2, 8.3, and 7.8% for 30, 300, and 2000 ng/ml, respectively. Accuracy was within 10% at these concentrations. Pharmacokinetic evaluation -------------------------- Standard noncompartmental analysis for calculations using serum concentration-time data was performed using TOPFIT 2.0 (Gustav Fischer Verlag 1993). The following pharmacokinetic parameters were determined from serum concentration-time data and urine concentration data for codeine (cod), codeine-6-glucuronide (C-6-G), norcodeine (NC), morphine (morph), morphine-3-glucuronide (M-3-G), morphine-6-glucuronide (M-6-G) and normorphine (NM): C~max~- peak serum concentration \[pmol ml^-1^\], obtained from the visual inspection of the serum concentration-time curves t~max~- time to reach peak serum concentrations \[h\], obtained from the visual inspection of the serum concentration-time curves t~1/2~- terminal phase half-life \[h\] calculated according to ln(2)/λ~z~, the terminal elimination rate constant λ~z~ was determined from the slope of the regression line of In(concentration) vs time AUC~0-∞~- area under the serum concentration curve \[pmol ml^-1\*^h\], calculated by the trapezoidal rule, the segment to infinity was calculated from the last concentration measured by dividing the last concentration measured by the elimination rate constant (λ~z~) ratio AUC\' s- AUC~0-∞~ codeine/AUC~0-∞~ morphine Ae- cumulative amount of codeine, C-6-G, norcodeine, norcodeine-glucuronide, morphine, M-3-G, M-6-G, normorphine and normorphine-glucuronide \[% of dose\] excreted in urine CLo- apparent oral clearance of codeine calculated using dose/AUC~0-∞~ \[ml min^-1^\] CL~cod→c-6-G~- metabolic clearance of codeine to C-6-G calculated as Ae C-6-G/AUC~0-∞~ codeine \[ml min^-1^\] CL~cod→NC~- metabolic clearance of codeine to norcodeine calculated as (Ae norcodeine + Ae norcodeine-glucuronide)/AUC~0-∞~ codeine \[ml min^-1^\] CL~cod→morph~- metabolic clearance of codeine to morphine calculated as (Ae morphine + Ae M-3-G + Ae M-6-G + Ae normorphine + Ae normorphine-glucuronide)/AUC~0-∞~ codeine \[ml min^-1^\] CL~morph→M-3-G~- metabolic clearance of morphine to M-3-G calculated as Ae M-3-G/AUC~0-∞~ morphine \[ml min^-1^\] CL~morph→M-6-G~- metabolic clearance of morphine to M-6-G calculated as Ae M-6-G/AUC~0-∞~ morphine \[ml min^-1^\] CL~morph→NM~ metabolic clearance of morphine to normorphine calculated as (Ae normorphine + Ae normorphine-glucuronide)/AUC~0-∞~morphine \[ml min^-1^\] CL~ren\ cod~- renal clearance of codeine calculated as Ae codeine/AUC~0-∞~ codeine \[ml min^-1^\] CL~ren\ morph~- renal clearance of morphine calculated as Ae morphine/AUC~0-∞~ morphine \[ml min^-1^\] CL~ren\ C-6-G~- renal clearance of C-6-G calculated as Ae C-6-G/AUC~0-∞~ C-6-G \[ml min^-1^\] CL~ren\ NC~- renal clearance of norcodeine calculated as Ae norcodeine/AUC~0-∞~ norcodeine \[ml min^-1^\] CL~ren\ M-3-G~- renal clearance of M-3-G calculated as Ae M-3-G/AUC~0-∞~ M-3-G \[ml min^-1^\] CL~ren\ M-6-G~- renal clearance of M-6-G calculated as Ae M-6-G/AUC~0-∞~ M-6-G \[ml min^-1^\] CL~ren\ NM~- renal clearance of normorphine calculated as Ae normorphine/AUC~0-∞~ normorphine \[ml min^-1^\] Statistical analysis -------------------- In this cross over study, 12 subjects were required to ensure a 80% power of detecting a potentially clinically relevant difference in pharmacokinetic parameters of 30% with a variability of 25% at the 5% significance level for α. Pharmacokinetic data are presented as mean ± 95% CI. Pharmacokinetic data of codeine and its metabolites after administration of codeine + placebo vs. codeine + diclofenac are compared using the Wilcoxon matched pairs test. In addition, AUC and C~max~ of codeine and C-6-G were investigated in terms of bioequivalence: Point estimates (geometric means) and 90% CI were given for the ratios test/reference (test: pharmacokinetic parameters under consideration after codeine + diclofenac, reference: pharmacokinetic parameters under consideration after codeine + placebo). Test was considered bioequivalent to reference if 90% CI of log transformed AUC ratios were within 0.80--1.25 and C~max~ ratios were within 0.75--1.34. Adverse events are presented in two different manners: 1.) The number of subjects reporting *any* side effect after administration of codeine + placebo vs. codeine + diclofenac were compared using the chi square test. 2.) At different time points (2, 4 and 6 h after drug administration) VAS-rated mean sum scores of all reported side effects after administration of codeine + placebo vs. codeine + diclofenac were calculated; the mean sum scores are compared using the paired Wilcoxon matched pairs test at the different time points. The area under the pain threshold versus time curve and the area under the pain tolerance versus time curve (corrected to baseline at t = 0 h) are presented as mean ± SD and are compared using Wilcoxon matched pairs test. A p value of less than 0.05 was considered significant. Results ======= Pharmacokinetic data -------------------- Similar serum concentrations of codeine and its metabolite C-6-G were observed after treatment with codeine + placebo vs. codeine + diclofenac (Figure [2](#F2){ref-type="fig"}). Peak serum concentrations (C~max~) of codeine (764; 602--924 pmol ml^-1^ vs. 821; 663--980 pmol ml^-1^) and C-6-G (5263; 4738--5788 pmol ml^-1^ vs. 5621; 5167--6074 pmol ml^-1^), did not differ significantly between the two different treatments, nor did the time to attain peak serum concentrations (t~max~) and the terminal half-life t~1/2~ (Table [1](#T1){ref-type="table"}). The concentrations of codeine and its metabolites in the one intermediate metaboliser (MR = 2.8) were within the range of the concentrations of the other volunteers with MR\<1, however as expected the metabolites formed via O-demethylation (morphine, morphine glucuronides and normorphine) were the lowest or second lowest observed. {#F2} ::: {#T1 .table-wrap} ::: {.caption} ###### Time to attain peak serum concentrations (t~max~) and terminal half-life of codeine and its metabolites (n = 12). ::: codeine + placebo mean (95% CI) codeine + diclofenac mean (95% CI) p level ----------------------------- --------------------------------- ------------------------------------ --------- t~max~ codeine \[h\] 0.69 (0.51--0.87) 0.71 (0.51--0.91) n.s. C~max~ codeine \[pmol/ml\] 764 (602--924) 821 (663--980) n.s. t~1/2~ codeine \[h\] 3.37 (3.16--3.59) 3.43 (3.24--3.62) n.s. t~max~ C-6-G \[h\] 1.13 (0.89--1.35) 1.10 (0.87--1.34) n.s. C~max~ C-6-G \[pmol/ml\] 5263 (4738--5788) 5621 (5167--6074) n.s. t~1/2~ C-6-G \[h\] 3.54 (3.39--3.68) 3.54 (3.36--3.73) n.s. t~max~ morphine \[h\] 0.63 (0.45--0.80) 0.67 (0.53--0.81) n.s. C~max~ morphine \[pmol/ml\] 22.4 (12.3--32.5) 24.3 (13.7--35.0) n.s. t ~1/2~ morphine \[h\] 8.78 (5.97--11.59) 8.54 (6.74--10.61) n.s. t~max~ M-6-G \[h\] 1.5 (1.37--1.64) 1.35 (1.06--1.65) n.s. C~max~ M-6-G \[pmol/ml\] 63.4 (38.7--88.1) 68.8 (41.2--96.5) n.s. t~1/2~M-6-G \[h\] 7.77 (7.40--8.13) 8.10 (7.57--8.63) n.s. ::: The AUCs of codeine did not differ between the two treatments (Table [2](#T2){ref-type="table"}). In addition, the two treatment, based on AUCs and C~max~ of codeine and C-6-G, could be considered bioequivalent: Point estimates of the test/reference ratios and 90% CI were 1.007 (0.998--1.017) for AUC codeine, 1.004 (1.001--1.008) for AUC of C-6-G, 1.015 (0.990--1.042) for C~max~ of codeine and 1.008 (1.001--1.016) for C~max~ of C-6-G. However, a small but significant increase in the AUC of C-6-G (+ 5.5%) was observed in subjects after codeine + diclofenac (Table [2](#T2){ref-type="table"}). In terms of morphine formation, the ratios of AUC codeine/AUC morphine did not reveal a significant difference (53.4; 10.5--96.4 95% CI vs. 45.2; 16.4--74.0 95% CI) between the two treatments. ::: {#T2 .table-wrap} ::: {.caption} ###### Area under the serum concentration time curve (AUC~0-∞~) of codeine and its metabolites (n = 12). ::: \[pmol ml^-1\*^h\] codeine + placebo mean (95% CI) codeine + diclofenac mean (95% CI) p level -------------------- --------------------------------- ------------------------------------ --------- AUC codeine 2416 (2025--2807) 2515 (2176--2855) n.s. AUC C-6-G 29550 (26190--32910) 31120 (28150--34090) \< 0.05 AUC norcodeine 342 (257--428) 338 (228--448) n.s. AUC morphine 94.5 (52.9--136) 92.1 (53.8--131) n.s. AUC M-3-G 2137 (1298--2976) 2259 (1427--3091) n.s. AUC M-6-G 425 (254--596) 471 (298--644) \< 0.05 AUC normorphine 540 (396--684) 549 (387--710) n.s. ::: Peak serum concentrations (C~max~) of morphine (22.4 pmol ml^-1^; 12.3--32.5 95% CI vs. 24.3 pmol ml^-1^; 13.7--35.0 95% CI) and M-6-G (63.4 pmol ml^-1^; 38.7--88.1 95% CI vs. 68.8 pmol ml^-1^; 41.2--96.5 95% CI) did not differ significantly between the two different treatments, as well as the time to attain peak serum concentrations (t~max~) and the terminal half-life t~1/2~ (Table [1](#T1){ref-type="table"}). AUC of morphine did not differ between the two treatments (Table [2](#T2){ref-type="table"}). A small but significant increase of the AUC of M-6-G (+10.8%) was observed after codeine + diclofenac (Table [2](#T2){ref-type="table"}). Serum concentration-time curves of morphine, M-3-G and M-6-G are displayed in Figure [3](#F3){ref-type="fig"}. {#F3} No significant differences were observed in either the apparent oral clearance of codeine (CLo) (1805 ml min^-1^; 1498--2112 95% CI vs. 1700 ml min^-1^; 1470--1929 95% CI), metabolic clearance of codeine to C-6-G, norcodeine and morphine, of morphine to M-3-G, M-6-G and normorphine or renal clearance of codeine and its metabolites between the two treatments (Tables [3](#T3){ref-type="table"} and [4](#T4){ref-type="table"}). The total amount excreted of codeine and its metabolites excreted was 81.3% of dose (76.3--86.3 95% CI) vs. 84.9% of dose (81.8--88.0 95% CI) and did not differ between the two treatment groups. The urinary recoveries of codeine and its metabolites are listed in Table [5](#T5){ref-type="table"}. ::: {#T3 .table-wrap} ::: {.caption} ###### Partial metabolic clearances of codeine and morphine after administration of 100 mg codeine phosphate (n = 12). ::: Cl~met~ \[ml min^-1^\] codeine + placebo mean (95% CI) codeine + diclofenac mean (95% CI) p level ------------------------ --------------------------------- ------------------------------------ --------- codeine → C-6-G 1134 (882--1385) 1104 (928--1279) n.s. codeine → morphine 174 (118--230) 176 (126--226) n.s. codeine → norcodeine 102 (69.1--136) 97.1 (76.2--118) n.s. morphine → M-3-G 2631 (1227--4036) 2550 (1527--3573) n.s. morphine → M-6-G 702 (362--1042) 692 (446--938) n.s. morphine → normorphine 3090 (431--5749) 2600 (1456--3744) n.s. ::: ::: {#T4 .table-wrap} ::: {.caption} ###### Renal clearances of codeine and metabilites after administration of 100 mg codeine phosphate (n = 12). ::: Cl ~ren~ \[ml min^-1^\] codeine + placebo mean (95% CI) codeine + diclofenac mean (95% CI) p level ------------------------- --------------------------------- ------------------------------------ --------- codeine 66.5 (54.6--78.3) 64.2 (53.6--74.9) n.s. C-6-G 88.6 (76.4--101) 86.6 (78.6--94.5) n.s. norcodeine 220 (187--253) 233 (208--259) n.s. morphine 117 (72.4--162) 114 (83.6--144) n.s. M-3-G 88.2 (76.1--100) 86.2 (77.1--95.4) n.s. M-6-G 122 (108--136) 114 (101--127) n.s. normorphine 123 (106--141) 133 (115--152) n.s. ::: ::: {#T5 .table-wrap} ::: {.caption} ###### Urinary excretion of codeine and metabolites after administration of 100 mg codeine phosphate (n = 12). ::: Ae \[% of dose\] codeine + placebo mean (95% CI) codeine + diclofenac mean (95% CI) p level ------------------------- --------------------------------- ------------------------------------ --------- codeine 3.74 (3.33--4.15) 3.89 (3.10--4.68) n.s. C-6-G 62.2 (57.0--67.5) 64.7 (60.7--68.7) n.s. norcodeine 1.76 (1.35--2.18) 1.86 (1.36--2.36) n.s. norcodeine-glucuronide 3.87 (3.01--4.73) 4.01 (3.02--5.00) n.s. morphine 0.22 (0.13--0.32) 0.23 (0.14--0.33) n.s. M-3-G 4.40 (2.81--5.99) 4.66 (3.10--6.22) n.s. M-6-G 1.21 (0.77--1.64) 1.31 (0.83--1.78) n.s. normorphine 1.56 (1.91--1.93) 1.73 (1.34--2.11) n.s. normorphine-glucuronide 2.30 (1.80--2.79) 2.50 (1.99--3.01) n.s. ::: Diclofenac was detected in the serum of all subjects after codeine + diclofenac with mean peak serum concentration of 4.4 nmol/ml (3.77--5.06 95% CI) after 1.73 h (1.21--2.26 95% CI) (see Table [6](#T6){ref-type="table"}). ::: {#T6 .table-wrap} ::: {.caption} ###### Pharmacokinetic parameters of diclofenac after oral administration of 50 mg diclofenac sodium together with 100 mg codeine phosphate (n = 12). ::: mean (95% CI) -------------------- ------------------- C~max~ \[nmol/ml\] 4.42 (3.77--5.06) t~max~ \[h\] 1.73 (1.21--2.26) AUC \[h nmol/ml\] 5.63 (4.77--6.50) t~1/2~ \[h\] 0.97 (0.68--1.26) ::: Pharmacodynamic data and adverse events --------------------------------------- The area under the pain threshold-time as well as the area under the pain tolerance-time curves did not differ significantly after administration of codeine + placebo vs. codeine + diclofenac (8.68 ± 12.7 vs. 4.87 ± 9.82 s\*h and 23.3 ± 57.5 vs. 13.9 ± 35.6 s\*h, respectively). The time courses of pain threshold and pain tolerance are displayed in Figure [4](#F4){ref-type="fig"}. {#F4} Both codeine + placebo and codeine + diclofenac were well tolerated by the volunteers. Only minor side effects occurred and no rating higher than 5 was reported on a VAS scale. The number of subjects reporting any adverse event from VAS rated from 0 (not present) to 10 (most severe) at 2, 4 and 6 hours after drug intake is displayed in table [7](#T7){ref-type="table"} and did not differ significantly between the two treatments. Also, no significant differences could be observed concerning the mean sum scores of all reported side effects at the given time points (Table [8](#T8){ref-type="table"}). ::: {#T7 .table-wrap} ::: {.caption} ###### Number of subjects reporting any adverse event ::: codeine + placebo codeine + diclofenac ---------------- ------------------- ---------------------- fatigue 10 8 headache 4 1 dizziness 6 5 blurred vision 3 3 nausea 2 2 itching 0 0 exanthema 0 0 others 0 0 ::: ::: {#T8 .table-wrap} ::: {.caption} ###### Mean of sum scores of all adverse events ::: time after drug administration codeine + placebo mean ± SD codeine + diclofenac mean ± SD -------------------------------- ----------------------------- -------------------------------- 2 h 4.3 ± 4.4 2.3 ± 2.1 4 h 1.6 ± 1.5 1.3 ± 1.7 6 h 2.9 ± 2.8 1.7 ± 1.7 ::: Discussion ========== The possible pharmacokinetic interaction between codeine and diclofenac could have clinical implications: One could have speculated that by inhibiting codeine glucuronidation other metabolic pathways of codeine, especially O-demethylation to morphine may then be favoured in extensive metabolizers of CYP2D6, resulting in elevated morphine serum levels and a greater analgesic effect, and, possibly also increased adverse effects. A direct influence of diclofenac on O-demethylation of codeine to morphine via interaction with CYP2D6 has been excluded.\[[@B26]\] Considering our previous in vitro data\[[@B10]\], it was important to verify whether codeine glucuronidation was inhibited by diclofenac in vivo after administration of a commonly used dose. However, after single dose administration of codeine + diclofenac in healthy volunteers we did not observe major changes in the pharmacokinetics of codeine and its metabolites compared to codeine + placebo. In this study, codeine and diclofenac were given in a dose commonly used for clinical treatment of pain. In addition, no significant differences in the ratios AUC codeine/ AUC morphine were detected after the two different treatments, indicating that the formation of morphine from codeine remained unaffected. A slight increase in the AUCs of C-6-G (+5.5%) and M-6-G (+10.8%) after the diclofenac containing regimen was observed, reaching statistical significance (p \< 0.05). This might be a result of inhibition of renal excretion of both glucuronides by diclofenac or its glucuronide. In our opinion, these minor changes are not of clinical relevance after single dose administration. Nevertheless, the observed non significant decrease in renal clearance especially of M-6-G in the diclofenac containing regimen, might play a role during chronic treatment with opioids. In this study, we did not investigate a possible influence of codeine on the pharmacokinetics of diclofenac because it has been previously demonstrated that codeine does not influence the relative bioavailability of diclofenac in vivo.\[[@B27]\] The pharmacokinetic results obtained from this study do not confirm our previous in vitro data with human liver tissue homogenates, which revealed a marked non-competitive inhibition of codeine-6-glucuronidation caused by diclofenac (K~i~ of 7.9 μM in a concentration range of 100--10000 μM codeine and 0.5--100 μM diclofenac).\[[@B10]\] Although UGT (UDP-glucuronosyltransferase) 2B7, the UGT catalysing codeine-6-glucuronidation, morphine-6-glucuronidation and diclofenac glucuronidation \[[@B28]\], is known to be expressed in the intestine,\[[@B29]\] we did not favour a potential interaction of codeine-6-glucuronidation at the level of the small intestine because of its very low intrinsic clearance compared to liver tissue in vitro.\[[@B10]\] The pharmacokinetic data for codeine and its metabolites obtained in our study are supported by the data obtained by other investigators.\[[@B12],[@B13],[@B30]\] The codeine peak serum concentrations observed in our study ranged from 517 to 1481 pmol ml^-1^ and therefore did not reach the concentrations used in our previous in vitro investigation. Kirkwood et al. could demonstrate a more than 50% inhibition of the glucuronidation of the codeine structural analogue dihydrocodeine by 50 μM diclofenac. \[[@B11]\] In our study, mean peak serum concentrations of diclofenac were 4.4 nmol/ml (range 3.01--6.21 nmol/ml) and were in the same range as reported by Davies et al.\[[@B31]\] Mean t~max~ of diclofenac was 1.73 h, so peak serum concentrations of diclofenac were later than the t~max~ of C-6-G and M-6-G (table [1](#T1){ref-type="table"}). In addition, diclofenac is known to be highly bound to serum protein (\>99.7%).\[[@B31]\] Therefore it is likely that diclofenac, at least after administration of a common single dose, does not achieve serum levels fast enough and high enough to inhibit codeine glucuronidation in vivo. Furthermore, it is unlikely that a multiple dose regimen of diclofenac (e.g. 50 mg tid) would lead to the cumulation of diclofenac serum levels due to the short elimination half-life of diclofenac (mean t~1/2~ 0.97 h in our study). Therefore, in our opinion, steady-state administration of diclofenac will also not alter codeine glucuronidation in vivo. However, the data obtained in this study cannot exclude a pharmacokinetic / pharmacodynamic interaction at steady-state. Concerning codeine glucuronidation, our pharmacokinetic results demonstrate that in vitro conditions studied did not necessarily reflect the in vivo conditions at therapeutic doses. In terms of side effects, both codeine + placebo as well as codeine + diclofenac were well tolerated by the volunteers in our study. Only minor side effects occurred and no rating higher than 5 was reported on a VAS scale. There are several studies demonstrating the benefit of the combination of NSAIDs plus opioids in comparison to opioids alone in the treatment of postsurgical pain, pain induced by arthrosis and chronic pain in cancer patients. \[[@B1]-[@B7]\] In one study, after a single dose of 100 mg diclofenac rectally given to postsurgical patients receiving morphine PCA, a significant decrease in hourly morphine consumption and plasma morphine and morphine-6-glucuronide were observed after administration of diclofenac, but with a lag time of more than 5 hours.\[[@B32]\] In our study, we used the cold pressor test for analysing the analgesic effects of opioids because it has been proven to be sensitive to the effects of codeine and morphine in several studies.\[[@B17],[@B18]\] In contrast, this test is not appropriate to investigate the efficacy of NSAIDs alone.\[[@B33]\] In our study, the area under the pain threshold versus time curve and the area under the pain tolerance versus time curve did not differ significantly after the two treatments over the time intervals studied. In case of a pharmacokinetic interaction between codeine and diclofenac leading to higher serum levels of morphine and M-6-G, an increase in pain threshold and pain tolerance would have been postulated. A placebo part of the study was omitted because the primary objective was the assessment of a pharmacokinetic interaction. Because we could exclude a pharmacokinetic interaction between codeine and diclofenac, the synergistic analgesic effects in the above mentioned studies are likely to result solely from the different pharmacodynamic mode of actions. Most postsurgical patients and many cancer patients have an inflammatory component to their pain, which responds to cyclooxygenase inhibition. Furthermore, NSAIDs have been reported to have a specific effect in malignant bone pain. \[[@B34]\] Conclusions =========== A single dose of diclofenac does not alter the glucuronidation of codeine in healthy volunteers, which is in contrast to recent in vitro data. The formation of morphine from codeine was not affected. The combination of codeine and diclofenac was well tolerated. List of abbreviations used -------------------------- CI: confidence interval MR: metabolic ratio NSAIDs: non steroidal anti-inflammatory drugs C-6-G: codeine-6-glucuronide M-6-G: morphine-6-glucuronide UGT: UDP-glucuronosyltransferase Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We are grateful for the excellent technical assistance of Mrs. Anja Riebe, Mrs. Erika Schneider, Mrs. Sonja Seefried and Mrs. Monika Seiler. Funding: The study was fully supported by the Robert Bosch-Foundation, Stuttgart, Germany

Background ========== Traditional Chinese medicine has been practised by Chinese communities world-wide for many generations, and there is a wealth of literature information available related to the therapeutic use of this type of medicine. In recent years, there has been a global surge in the popularity of herbal/traditional medicine, and currently there is enormous interest in developing new pharmaceutical products from such resources. Over the last decade, bis-benzylisoquinoline (BBI) and especially protoberberine alkaloids (e.g., berberine and jatrorrhizine; Fig. [1](#F1){ref-type="fig"}) have attracted considerable attention in this respect; protoberberines represent a structural class of organic cations and have been found to be predominantly distributed in several genera of the families *Ranunculaceae* and *Berberidaceae* (e.g., *Berberis, Mahonia, Coptis*). Berberine, a major representative of the protoberberine alkaloids, displays diverse biochemical and pharmacological actions while being relatively non-toxic to man; its antimicrobial activity has been demonstrated against many bacterial and fungal species \[[@B1]-[@B4]\]. The drug was subsequently screened for anti-cancer activity following evidence of antineoplastic properties \[[@B5]-[@B7]\]. It has also been shown that berberine exhibits the ability to induce apoptosis in promyelocytic leukemia HL-60 and 3T3 fibroblast cells \[[@B5],[@B8]\]. In addition, some protoberberines are highly effective as cytotoxic agents against several carcinoma such as HeLa, SVKO(3), Hep-2, primary culture from mouse embryo and human fibroblast cells \[[@B9],[@B10]\]; berberine showed consistently the highest cytotoxicity among the alkaloids tested. Only recently it has been reported that berberine possesses a dual topoisomerase I and II poisoning activity \[[@B11]-[@B13]\] and binds to double helical DNA with a high affinity \[[@B14]\]. Furthermore, computer modeling studies on protoberberine-DNA complexes suggest that these alkaloids are able to bind to the host DNA by both intercalative (through rings C and D; Fig. [1](#F1){ref-type="fig"}) and minor groove (rings A and B) binding modes \[[@B12],[@B13]\]. {#F1} All the above findings raise the possibility that protoberberines may be effective in deactivation of carcinogens and tumour promoters. For in vitro screening of antimutagens/anticarcinogens a variety of prokaryotic and eukaryotic micro-organisms have been used. In the present study, a new eukaryotic model, viz. green unicellular flagellate *Euglena gracilis* was chosen \[[@B15]\]. This micro-organism possesses a multigenomic system with nuclear, mitochondrial and chloroplast DNAs. The detection ability of this model is based on the preferential and selective sensitivity of the chloroplast genetic apparatus to xenobiotics resulting in elimination of the functional chloroplast from the cells. Antichloroplastic activity of mutagens is macroscopically manifested by an irreversible loss of the capability of cells to form green colonies (bleaching effect). Compounds with antimutagenic properties prevent such a bleaching of *E. gracilis* cells by mutagens / carcinogens. Previous studies on several standard mutagens and antimutagens revealed a high sensitivity and reliability of the model \[[@B16]-[@B19]\]. Thus, the aim of the present work was to study the possible antimutagenic effect of berberine and jatrorrhizine as well as of the BBI alkaloids fraction and of the overall *M. aquifolium* crude extract, on acridine orange (AO) -- induced mutagenicity in the *E. gracilis* assay. Materials and methods ===================== Plant material -------------- The stem bark of *Mahonia aquifolium* (Pursh) Nutt. (*Berberidaceae*) was collected in October 1998 in the Arboretum Tesárske Mlyòany, Slovakia. Voucher specimens are deposited at the Herbarium of the Faculty of Pharmacy, Comenius University, Bratislava (No. Ma-108/8). Dried and powdered stem bark was macerated at room temperature with 62% aqueous EtOH (1:10 w/v) for 5 days and then filtered. The filtrate (crude *M. aquifolium* extract) was stored in a refrigerator until use in the biological assay. The BBI and the protoberberine alkaloid fractions were separated and purified using standard procedures as described previously \[[@B20]\]. Berberine \[CAS 2086-83-1\] and jatrorrhizine \[CAS 3621-38-3\], the main quaternary protoberberine alkaloids were isolated as iodides from the protoberberine fraction and identified by direct comparison of their physical and spectral properties with the literature data \[[@B20]\]. HPLC analysis of the BBI alkaloids was performed according to the procedure reported earlier \[[@B20]\]. Totally 6 tertiary BBI alkaloids (oxyacanthine, aromoline, baluchistine, berbamine, obamegine and aquifoline) were positively identified by referring to authentic compounds described earlier \[[@B20]\]. The identification was made on the basis of their UV absorption spectra and retention time. Microorganism ------------- *Euglena gracilis* (strain Z) was obtained from S.H. Hutner, Haskins Laboratory, Pace University, New York, NY, USA and maintained on Cramer-Myers (CM) medium \[[@B21]\] under static conditions at 27°C and with permanent lighting (16.4 W/m^2^). Chemicals --------- Acridine orange \[CAS 65-61-2\] was purchased from Merck, Darmstadt, Germany. Stock solutions of AO, berberine, jatrorrhizine, and BBI alkaloids were prepared by dissolving them in distilled water. Mutagenicity assay ------------------ *E. gracilis* cells diluted to concentration 8 × 10^5^ cells/ml CM medium were used in the experiments. Aliquots (0.2 ml) of the cell suspension were dispensed into test tubes, along with 2.3, 11.4 or 22.8 μM AO, indicated concentrations of berberine (0.03, 0.06, 0.15, 0.30, 0.45 μg/ml), jatrorrhizine (0.72, 1.44, 14.4, 28.8, 43.2 μg/ml), BBI alkaloids (0.72, 1.44, 14.4, 28.8,43.2 μg/ml) or the *M. aquifolium* extract (5.5, 11.0, 110,220,330 μg/ml), and at last the content of the incubation mixture was completed to a final volume (5 ml) by the addition of CM medium. Following a 24-h co-treatment, the cells were centrifuged at 3000 rpm for 20 min, the resultant pellet suspended in fresh CM medium and again centrifuged. After the pellet was resuspended in fresh CM medium, the cells were finally cultivated 14 days at 27°C under permanent illumination (16.4 W/m^2^). The experiments were repeated in three independent series. Just before the counting, the movement of *E. gracilis* was stopped by adding a drop of EtOH and the counting of green and white (mutant) colonies was carried out in a Bürker chamber under a microscope. The viability of the *Euglena* cells was estimated by counting the total white mutants in the presence of mutagen (positive control) compared to the number of the spontaneous white mutants in the absence of mutagen (negative control). We defined the relative decrease of the bleaching as the antimutagenic potency (AP) which was calculated by the formula  where B~0~ is the AO-induced *E. gracilis* bleaching (%) and B~r~ is the AO-induced and antimutagen-reduced *E. gracilis* bleaching (%). The statistical significance of all the calculated values were determined by paired Student\'s t-test. The values represent the means ± standard deviation (SD). Results and discussion ====================== The mutagenic effect of AO on *E. gracilis* was tested in three concentrations, 2.3, 11.4 and 22.8 μM, which induced, respectively, 43 ± 2 %, 58 ± 2 % and 65 ± 3 % of white mutant cells. The AO concentrations were chosen so as to ensure no significant change in the viability of cells compared to the negative controls (*Euglena* cells in the absence of the mutagen). Similarly, no spontaneous white mutants were found in any sets of positive controls. The dose-dependent inhibitory effect of the crude *Mahonia* extract, BBI alkaloids fraction, berberine and jatrorrhizine on the mutagenicity of AO applied at the above 3 concentrations is displayed in Figs. [2](#F2){ref-type="fig"},[3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"},[5](#F5){ref-type="fig"}. As shown in Figs. [2](#F2){ref-type="fig"} and [3](#F3){ref-type="fig"}, the plot of the percentage of bleached mutants vs. concentration for the *Mahonia* extract and the BBI fraction respectively displays typical curving even though over different concentration ranges. The lowest concentration that causes statistically significant (p \< 0.05) reduction in the percentage of white colonies (as compared to the positive control) is 0.7 and 11.0 μg/ml for the *Mahonia* extract and the BBI alkaloids, respectively. For both samples, the decrease of the percent proportion of mutant colonies gradually continued with further increasing the concentration; however, the antimutagenic efficiency was much higher for the BBI alkaloids sample, for which the bleaching percentage (corresponding to the AO 2.3 μM curve) reaches \~5% level for the highest concentration tested. {#F2} {ref-type="fig"}.](1472-6882-2-2-3){#F3} {ref-type="fig"}.](1472-6882-2-2-4){#F4} {ref-type="fig"}.](1472-6882-2-2-5){#F5} As to the protoberberines berberine and jatrorrhizine, which are of prime interest here, the concentration dependence of the bleaching effect of jatrorrhizine (Fig. [5](#F5){ref-type="fig"}) is markedly similar to those observed for the above two samples. In contrast, the behaviour of berberine is completely different; as shown in Fig. [4](#F4){ref-type="fig"}, berberine chloride suddenly reduced the number of AO-induced mutant cells (to ca 5% for AO 2.3 μM) at concentration 0.03 μg/ml but further increase of the concentration did not cause progressive reduction of the bleaching activity or even, in the case of the AO 22.8 μM series, the number of white colonies increased slightly but significantly with respect to that observed for the most effective concentration (p \< 0.05). For berberine, the antimutagenic potency (AP) was also calculated using the formula given at the end of the Material and methods section and the results are summarised in Table [1](#T1){ref-type="table"}. As can be seen, berberine exhibited significant antimutagenic potency (\>80%) almost over the whole concentration range tested when the concentration of the mutagen (AO) was 2.3 and 11.4 μM while for the increased AO concentration (22.8 μM) the antimutagenic potential of berberine was much less effective (AP \< 55%). ::: {#T1 .table-wrap} ::: {.caption} ###### Antimutagenic potency (AP) \[%\] of berberine in reducing the acridine orange (AO)-induced bleaching of *Euglena gracilis* cells. ::: mutagen concentration of berberine \[μg/ml\] --------- -------------------------------------- ---------- ---------- ---------- ---------- 2.3 89 ± 2.2 88 ± 5.3 90 ± 6.2 86 ± 1.7 78 ± 6.5 11.4 80 ± 3.0 86 ± 4.2 84 ± 1.6 83 ± 1.1 82 ± 3.6 22.8 54 ± 3.5 54 ± 1.8 49 ± 4.1 38 ± 2.3 36 ± 1.7 ::: Although the BBI and protoberberine alkaloids studied in this laboratory exhibit a broad spectrum of biological activities, for most of them the molecular mechanism of action is still unclear. In the case of the BBI alkaloids, one possible mechanism of their antimutagenic activity could be based on the reports that tetrandrine and other BBI alkaloids are excellent scavangers of reactive oxygen species (ROS) such as singlet oxygen and/or superoxide anion radical \[[@B22]\] since it is well documented that that ROS play a central role in multistage mutagenesis and carcinogenesis \[[@B23],[@B24]\]. However, the potential mechanism based on the ability of BBIs to prevent oxidative damage of DNA by ROS is unlikely since the ability of AO to produce ROS in a redox system has never been published. Instead, the antimutagenic potency of the BBI fraction is more likely linked to modulation of DNA at the transcriptional level. Recently, the role of eukaryotic transcription factors, NF-κB and AP-1, has been highlighted in mutagenesis and carcinogenesis \[[@B25],[@B26]\] and tetrandrine is known to inhibit the activation of NF-κB in the alveolar macrophage \[[@B22]\]. Nevertheless, the identification of the ultimate mechanism of the antimutagenic effect of the BBI alkaloids awaits further study. As described above, the significant antimutagenic potency against AO was observed for the two protoberberines studied herein. The potency of jatrorrhizine alone is comparable to that found for the overall BBI fraction and the potency of berberine is even by more than 2 orders of magnitude higher. To our knowledge, the ability of protoberberines to reduce the mutagenicity of environmental chemicals has not yet been reported so that the mechanism of their antigenotoxic action was not discussed. However, the cytotoxic activity of berberine and its analogues has been attributed to DNA topoisomerase I (TOPO-I) and II (TOPO-II) poisoning \[[@B11]-[@B13]\]; recently, a mechanistic model \[[@B12]\] for TOPO-I poisoning by protoberberines was developed according to which berberine binds to DNA and TOPO-I at the interface of the binary complex DNA -- TOPO-I in such a manner that rings C and D (Fig. [1](#F1){ref-type="fig"}) intercalate into the DNA helix, while rings A and B protrude out of the helix interior into the minor groove, where they are accessible to interact with specific functional groups on the enzyme. Thus, both drug-DNA and drug-enzyme interactions contribute to the overall affinity (potency) of the drug. In this light it is reasonable to suggest that the antimutagenicity of berberine against the genotoxic effect of AO on plastid DNA is exerted by a similar, if not identical, mechanism. A comparison of the antimutagenic potency of the two protoberberines tested here supports this notion. The only difference between berberine and jatrorrhizine (Fig. [1](#F1){ref-type="fig"}) is restricted to the substitution on the A ring (2,3-methylenedioxy in berberine vs. 2-methoxy-3-hydroxy substitution in jatrorrhizine), i.e. to the portion of the molecule which is involved in interactions with TOPO-I. Thus, insertion of the protoberberine molecule between the base pairs of DNA prevents the interaction of the AO mutagen at the same binding site as AO is known as a frameshift mutagen due to strong intercalative binding. A similar detrimental effect of the free hydroxy group on the TOPO-II poisoning activity has also been observed for berberrubine \[[@B13]\]. In this context it is important to note that camptothecin, a natural product currently in clinical use for the treatment of a variety of cancers \[[@B12],[@B25]\], has a pharmacological profile very similar to that of protoberberines; its antineoplastic and antimutagenic activity was demonstrated against many organic mutagens and also the DNA-TOPO-II complex was identified as its molecular target \[[@B26]\]. It is therefore reasonable to consider that the drug-induced poisoning of TOPO enzymes may be of general importance for anticancer and antimutagenic drugs. Conclusions =========== In summary, the data presented here showed that the BBI alkaloid fraction and especially the protoberberines, jatrorrhizine and berberine, inhibited the AO-induced mutagenicity in the *E. gracilis* assay. Berberine demonstrated significant antimutagenic activity against the chloroplast damaging effects of AO with a high potency. The protective effects of the latter compound against the chemical carcinogen suggest its potential chemopreventive ability, which is under investigation using other bacterial tests as well as animal tumour models. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ This research was supported by the Grant Agency of the Slovak Republic, Projects No. 1/7359/20, 1/8216/01, 1/8215/01 and 2005.

Background ========== The Nudix hydrolase family comprises enzymes that hydrolyse predominantly the diphosphate (pyrophosphate) linkage in a variety of nucleoside triphosphates, dinucleoside polyphosphates, nucleotide sugars and nucleotide cofactors having the general structure of a **nu**cleoside **di**phosphate linked to another moiety, **X**\[[@B1],[@B2]\]. They are found in archaea, eubacteria, animal, plants, and fungi and all possess the Nudix box sequence signature motif Gx~5~Ex~5~ \[UA\]xREx~2~EExGU (where U is an aliphatic hydrophobic amino acid) \[[@B1],[@B3]\]. The proposed functions of this family are to eliminate potentially toxic nucleotide metabolites from the cell and to regulate the concentrations of nucleotide cofactors and signalling molecules for optimal cell growth and survival. The number of genes encoding Nudix hydrolases varies widely, from zero in *Mycoplasma genitalium*\[[@B4]\] to 22 in *Deinococcus radiodurans*\[[@B5],[@B6]\]. This variation presumably reflects the growth or environmental adaptability, stress tolerance and metabolic capacity of the different organisms. The Nudix hydrolases thus offer an ideal system with which to study the evolution of a largely inessential protein family and its contribution to the individual biology of an organism. Understanding such variation requires a combination of detailed biochemical, genetic and cellular studies to reveal the individual functions of family members within the set in any given system. In the case of multicellular eukaryotes, the nematode *Caenorhabditis elegans* offers a genetically amenable model system with which to carry out such studies. There are 11 members of the Nudix hydrolase family in *C. elegans*. So far only two of these have been characterized -- a diadenosine tetraphosphate pyrophosphohydrolase (the orthologue of human NUDT2) \[[@B7]\] and an NADH diphosphatase \[[@B8]\]. Sequence comparisons would predict the existence of an ADP-sugar diphosphatase (NUDT5 orthologue) \[[@B9]\], an ADP-ribose diphosphatase (NUDT9 orthologue) \[[@B10]\], a diphosphoinositol polyphosphate pyrophosphohydrolase (NUDT3/4 orthologue) \[[@B11]\], two probable coenzyme A diphosphatases, one of which is highly similar to the mouse Nudt7 CoA diphosphatase \[[@B12]\], and 4 proteins of unknown function, including one with a strong similarity to the *Saccharomyces cerevisiae* PSU1/DCP2 protein \[[@B13],[@B14]\] and another similar to the developmentally-regulated mouse RP2 protein \[[@B15]\]. Recent characterization of the *S. cerevisiae* NADH and CoA diphosphatases \[[@B16],[@B17]\] and the mouse Nudt7 CoA diphosphatase \[[@B12]\] has revealed that they are located in peroxisomes. The function of these peroxisomal enzymes may be to regulate the concentration of these essential nucleotide cofactors for peroxisomal metabolism or, by analogy with the *E. coli* MutT 8-oxo-dGTPase, to eliminate toxic modified cofactor metabolites from the highly oxidizing peroxisomal environment. In order to investigate these possibilities in the *C. elegans* model system, we have cloned and characterised the putative *C. elegans* Y87G2A.14 CoA diphosphatase and shown that it displays the expected enzymatic activities and that it appears to be targetted to peroxisomes by a C-terminal PTS1 targeting signal. Results and discussion ====================== Cloning, expression and purification of Y87G2A.14 ------------------------------------------------- The *C. elegans* Y87G2A.14 gene encodes a 234 amino acid protein with an expected molecular weight of 26,601 Da. It was amplified by PCR from a *C. elegans* cDNA library. The PCR fragment was inserted into the pET-32b(+) expression vector and the nucleotide sequence of the insert was determined to be exactly the same as that submitted to GenBank under accession no. CAB54476. The recombinant plasmid pETY87G2A.14 was then used to transform *E. coli* BL21 (DE3) cells to generate a His-tagged thioredoxin fusion protein with an expected molecular mass of 43,731 Da. When the Trx-Y87G2A.14 fusion protein was expressed at 37°C, it was confined to inclusion bodies, so the induction temperature was decreased to 25°C to enhance protein solubility. As the expression level was low at this temperature, the induction time was increased to 8 h. These conditions markedly increased the solubility of Trx-Y87G2A.14 which was then purified from the soluble fraction (Fig [1](#F1){ref-type="fig"}, lane 2) to apparent homogeneity on NiCAM™-HC resin (Fig [1](#F1){ref-type="fig"}, lane 3). To determine the molecular weight of the Y87G2A.14 protein itself, the Trx-Y87G2A.14 fusion was cleaved with thrombin, which generated Y87G2A.14 with an apparent molecular weight of 27 kDa (expected molecular weight, 29,807 Da) and thioredoxin (15 kDa, Fig [1](#F1){ref-type="fig"}, lane 4). {#F1} Substrate specificity and product analysis ------------------------------------------ Purified Trx-Y87G2A.14 was inactive towards the following nucleotides when assayed at a fixed concentration of 0.5 mM: NADH, NAD^+^, NDP-sugars, 5\'-(d)NTPs, 5\'-NDPs, 5\'-NMPs and diadenosine polyphosphates. High activity was found with CoA and its derivatives. HPLC analysis of CoA hydrolysis by Trx-Y87G2A.14 showed that the enzyme was a CoA diphosphatase, cleaving the diphosphate linkage in CoA to yield adenosine 3\',5\'-bisphosphate (3\',5\'-ADP) and 4\'-phosphopantetheine (Fig [2](#F2){ref-type="fig"}). ![**Identification of reaction products of CoA hydrolysis.** Reaction mixtures containing 0.5 mM CoA were incubated at 37°C for 20 min with or without 0.1 μg Trx-Y87G2A.14 fusion protein and the products separated by HPLC as described in Materials and methods. Without enzyme (\-\-\-\-\--), with enzyme ([ ]{.ul}), gradient (--- --- --- --- ---). Positions of authentic standards are indicated.](1471-2091-3-5-2){#F2} Reaction requirements and kinetic parameters -------------------------------------------- Trx-Y87G2A.14 displayed optimal activity with 0.5 mM CoA as a substrate at pH 9.5. A divalent metal ion was absolutely required for activity, with optimal activity at 5 mM MgCl~2~. In common with all other Nudix hydrolases tested, fluoride was a strong inhibitor with a *K*~i~ value of approximately 3 μM (results not shown). *K*~m~, and *k*~cat~ values for CoA, CoA esters and oxidized CoA were calculated by non-linear regression from data obtained by HPLC analysis (Table [1](#T1){ref-type="table"}). A graphical example of the data for CoA in the form of a hyperbolic plot (Fig [3a](#F3){ref-type="fig"}) and double reciprocal plot (Fig [3b](#F3){ref-type="fig"}) show that the enzyme obeys simple Michaelis-Menten kinetics. The *k*~cat~*/ K*~m~ ratios show that the enzyme prefers reduced forms of CoA to oxidized CoA (Table [1](#T1){ref-type="table"}) with CoA itself the best substrate of those tested. {#F3} ::: {#T1 .table-wrap} ::: {.caption} ###### Kinetic parameters for the hydrolysis of CoA and CoA derivatives by Trx-Y87G2A.14 fusion protein ::: Substrate *K*~m~ (mM) *k*~cat~ (s^-1^) *k*~cat~ / *K*~m~ (10^3^M^-1^.s^-1^) ----------------------------- ------------- ------------------ -------------------------------------- CoA 0.22 13.8 64 Acetyl-CoA 0.23 11.8 51 Succinyl-CoA 0.24 12.7 53 3-Hydroxymethylglutaryl-CoA 0.20 10.5 54 CoASSCoA 0.32 2.9 9.1 Data were collected using the HPLC assay described in Materials and methods. ::: Subcellular localization ------------------------ Y87G2A.14 has the C-terminal tripeptide sequence SKI. This conforms to the pattern typical of PTS1 peroxisomal targeting signals found in many peroxisomal matrix proteins, suggesting that Y87G2A.14 may be targeted to these organelles \[[@B18],[@B19]\]. However, possession of a potential PTS1 sequence is not always sufficient on its own to result in peroxisomal targeting and other elements of the protein sequence may also be involved. Since targeting of animal peroxisomal proteins expressed in yeast has often been observed, yeast cells were transformed with expression plasmids encoding C-terminal or N-terminal fusions of Y87G2A.14 to yeast-enhanced green fluorescent protein (yEGFP) in order to determine the subcellular location of Y87G2A.14. The cells were then examined by confocal microscopy. Cells transformed with pY87G2A.14-yEGFP, in which the C-terminus of Y87G2A.14 is fused to the N-terminus of yEGFP, showed a diffuse, cytoplasmic fluorescence with no clear subcellular localization (Fig [4a](#F4){ref-type="fig"}). In contrast, cells transformed with pyEGFP-Y87G2A.14, in which the C-terminal tripeptide SKI is free to act as a targeting signal showed the clear punctate fluorescence that is indicative of yeast peroxisomes \[[@B16]\] (Fig [4b](#F4){ref-type="fig"}). The identity of SKI as the targeting signal was confirmed by transformation of cells with pyEGFP-Y87G2A.14Δ SKI, in which the C-terminal tripeptide was deleted during construction. This again showed a diffuse, cytoplasmic, fluorescence (Fig [4c](#F4){ref-type="fig"}). Together, these results strongly suggest that Y87G2A.14 is targeted to peroxisomes by its C-terminal tripeptide, SKI. {#F4} Conclusions =========== On the basis of its sequence, the *C. elegans* Y87G2A.14 gene product was predicted to be a peroxisomal coenzyme A diphosphatase. In addition to the Nudix motif, Y87G2A.14 possesses the PROSITE UPF0035 motif (Fig [5](#F5){ref-type="fig"}), which we have previously suggested confers a specificity for coenzyme A and its derivatives \[[@B12]\], and a C-terminal tripeptide, SKI, that conforms to the pattern typical of PTS1 peroxisomal targeting signals. The experiments described here confirm these predictions. Fig [5](#F5){ref-type="fig"} shows a multiple sequence alignment of the motif-containing region of Y87G2A.14 with related sequences from other organisms. Those marked with a tick have been experimentally shown to be coenzyme A diphosphatases \[[@B6],[@B12],[@B17]\]. In most cases, higher organisms possess two related sequences, e.g. mouse Nudt7 and Nudt8, one of which encodes a peroxisomal enzyme (e.g. Nudt7). However, *S. cerevisiae* has only one sequence containing the UPF0035 motif while *Arabidopsis thaliana* has three, and the second of the two *Drosophila melanogaster* sequences, RH61317, is currently only represented in GenBank by a single expressed sequence tag, so its status is still questionable. For the peroxisomal enzymes, either a putative C-terminal PTS1 or an N-terminal PTS2 targeting signal is present. Interestingly, in each case, the putative PTS2 signal is contained within or near a predicted mitochondrial targeting or chloroplast transit peptide sequence \[[@B20]-[@B22]\], suggesting a possible dual location for these proteins. Such a possibility has not yet been experimentally observed; however, mutation of a glutamate five residues to the C-terminal side of the PTS2 of rat peroxisomal 3-ketoacyl-CoA thiolase to a neutral or basic amino acid has been shown to result in partial mitochondrial targeting, suggesting that the negative charge on glutamate may normally block translocation to the mitochondria \[[@B23]\]. Whether or not a system exists *in vivo* to regulate dual targeting is clearly a topic requiring further investigation. The non-peroxisomal sequences provide no clear indication of possible subcellular location, hence they are likely to be cytoplasmic. Given the existence of mitochondrial, peroxisomal and cytoplasmic pools of CoA and CoA esters \[[@B24]\], it would not be surprising to find CoA diphosphatase activity in all these locations. However, the precise substrate specificities of the \"cytoplasmic\" activities remain to be determined. {#F5} Regarding the possible function of these enzymes in general, and the *C. elegans* peroxisomal enzyme in particular, a recent functional genomic analysis by RNA-mediated interference of *C. elegans* chromosome I, on which the Y87G2A.14 gene is located, revealed no phenotype in relation to growth, survival, fecundity or morphology when the expression of Y87G2A.14 was ablated \[[@B25]\]. This would indicate that, within the limitations of RNAi, the CoA diphosphatase activity of Y87G2A.14 is not essential. However, now that the biochemical properties of this protein have been established, a more detailed and targeted biochemical analysis can be undertaken that should reveal its cellular function and benefit to the organism. Nudix hydrolases are believed to regulate the concentrations of nucleotides for optimal cell performance and also to eliminate potentially toxic nucleotide metabolites from the cell. With regard to regulation, CoA diphosphatase activity is associated with the 400 kDa CoA synthesizing protein complex from *S. cerevisiae*, in which it forms part of an alternative pathway for CoA biosynthesis that differs from the principal route of 3\'-dephospho-CoA and CoA synthesis by this complex \[[@B26]\]. This CoA/4\'-phosphopantetheine cycle involves hydrolysis of CoA to 3\',5\'-ADP and 4\'-phosphopantetheine, which then reacts with ATP to give 3\'-dephospho-CoA then CoA. Whether such a pathway operates in peroxisomes and whether the *C. elegans* Y87G2A.14 protein is involved remain to be established. With regard to the elimination of toxic nucleotide metabolites, the 13-fold higher *k*~cat~ / *K*~m~ ratio for oxidized CoA (CoASSCoA) compared to CoA for the *S. cerevisiae PCD1* CoA diphosphatase previously suggested to us that this enzyme might preferentially remove non-functional and potentially toxic oxidized CoA and CoA esters from within the oxidizing environment of the peroxisomes \[[@B17]\]. However, neither the mouse Nudt7 \[[@B12]\] nor the *C. elegans* Y87G2A.14 proteins show this preference. Nevertheless, the potential production of adenine ring-oxidized derivatives of CoA by reactive oxygen species generated in the peroxisomes analogous to the 2-oxo-dATP and 8-oxo-dATP substrates of the mammalian MTH1 Nudix hydrolase \[[@B27]\] suggests that such species could be more relevant substrates for peroxisomal CoA diphosphatases *in vivo*. The amenability of *C. elegans* to studies of cellular and molecular stress will allow the question of the biological function of these enzymes to be addressed. Materials and methods ===================== *S. cerevisiae* strain BY4741 (*MAT a; his3D1; leu2D0; met15D0; ura3D0*) was from Research Genetics. Calf intestinal alkaline phosphatase, yeast inorganic pyrophosphatase, *Eco*R1 and *Bam*H1 were from Roche while *Bsp*H1 (*Pag1*) was from Helena Biosciences. *Pfu* DNA polymerase was from Stratagene. All other chemicals and nucleotides were from BDH or Sigma. The *E. coli* expression vector pET-32b(+) was from Novagen and the yeast-enhanced green fluorescent protein (yEGFP) fusion vectors pUG35 and pUG36 were a gift from J.H. Hegemann, Institute of Microbiology, University of Düsseldorf, Germany. The *C. elegans* cDNA library was prepared from adult nematodes by H. M. Abdelghany, School of Biological Sciences, University of Liverpool, U.K. Cloning of Y87G2A.14 from C.elegans ----------------------------------- A cDNA corresponding to the *C. elegans* Y87G2A.14 gene on chromosome 1 (GenBank accession no. CAB54476) was amplified from a cDNA library by PCR using as forward and reverse primers 5\' GCAAATCATGAAGTGTGTGGTTAGCCGAGCTG 3\' and 5\' TAAATGAATTCACTAAATTTTGGATTTCGGTTC 3\' respectively. These primers provided a *Bsp*H1 restriction site at the start of the amplified gene and an *Eco*R1 site at the end. After amplification with *Pfu* DNA polymerase, the DNA was recovered by phenol/chloroform extraction and digested with *Bsp*H1 and *Eco*R1. The digest was gel-purified and the restriction fragment ligated into the *Nco1* and *Eco*R1 restriction sites of pET-32b(+) as both *Bsp*H1 and *Nco1* form compatible ends with each other. The resulting construct, pETY87G2A.14, yielded Y87G2A.14 downstream of the 109-amino acid thioredoxin (Trx) fusion and His-tag and S-tag sequences under the control of an IPTG-inducible promoter. The structure of the insert was confirmed by sequencing. The construct was propagated by transformation of *E. coli* XL1-Blue cells. Expression of Y87G2A.14 in E. coli and protein purification ----------------------------------------------------------- *E. coli* strain BL21(DE3) was transformed with pETY87G2A.14. A single colony was picked from an LB agar plate containing 50 μg/ml ampicillin and inoculated into 10 ml LB medium containing 50 μg/ml ampicillin and incubated at 37°C. When the cells reached an A~600~ of 0.5, they were transferred to 1 litre of fresh LB medium containing 50 μg/ml ampicillin and grown to an A~600~ of 0.3 at 37°C, then transferred to an incubator at 25°C. Isopropyl-1-thio-β-D-galactopyranoside (IPTG) was added to 1 mM at an A~600~ of 0.8, and the cells induced for 8 h. The induced cells (4 g) were harvested, washed and resuspended in 20 ml breakage buffer (50 mM Tris-HCl, pH 8.0, 50 mM NaCl, 1 mM DTT). The cell suspension was sonicated and the cell lysate was then cleared by centrifugation at 10,000 × *g* at 4°C for 15 min. The supernatant was applied to a 15 × 50 mm column of NiCAM™-HC resin (Sigma) equilibrated with 50 mM Tris-HCl, pH 8.0, 0.5 M NaCl, 1 mM 2-mercaptoethanol at a flow rate of 0.5 ml/min. After eluting the unbound proteins, a linear gradient of 0--40 mM histidine in the same equilibration buffer was applied at flow rate of 1 ml/min and 1 ml fractions collected and analysed by SDS-PAGE. Those containing pure Trx-Y87G2A.14 fusion protein were collected, dialysed overnight at 4°C against 1 litre of 20 mM Tris-HCl, pH 8.0, 50 mM NaCl, ImM DTT and then concentrated by ultraflltration (Amicon) and stored at -20°C in 50% glycerol. Enzyme assays ------------- Potential substrates were screened by measuring the Pi released after nucleotide hydrolysis in presence of inorganic pyrophosphatase or alkaline phosphatase \[[@B9]\]. The standard assay (200 μl) for phosphodiester substrates was incubated at 37°C for 30 min and contained 50 mM l,3-bis \[tris(hydroxymethyl)-methylamino\]propane-HCl (BisTrisPropane-HCl), pH 8.0, 5 mM MgCl~2~, 1 mM DTT, 0.5 mM substrate, 0.1 μg of Trx-Y87G2A.14 fusion protein and 0.5 μg (1 unit) alkaline phosphatase. Assays with phosphomonoester substrates were as above, except 0.5 μg (100 mU) inorganic pyrophosphatase was used instead of alkaline phosphatase. The Pi released in each case was measured colorimetrically. Chromatographic analysis ------------------------ Kinetic parameters and reaction products generated from hydrolysis of CoA and its derivatives were measured by high performance anion-exchange chromatography. The reaction mixtures (100 μl) contained 50 mM BisTrisPropane-HCl, pH 9.5, 5 mM MgCl~2~, 1 mM DTT (in cases of substrates requiring reducing conditions), substrate in the range of 0.05--0.7 mM, (0.1--1 mM in the case of oxidized CoA) and were incubated at 37°C for up to 20 min (during which time the reaction rates remained linear) with 0.1 μg Trx-Y87G2A.14 fusion protein. A 90 μl sample of each reaction mixture was applied to a 1 ml Resource-Q column (Amersham Pharmacia Biotech) equilibrated with 0.045 M CH~3~COONH~4~ (pH 4.6, adjusted with H~3~PO~4~), and eluted with a linear gradient from 0 to 0.45 M NaH~2~PO~4~ (pH 2.7 adjusted with CH~3~COOH) for 10 min at a flow rate of 2 ml/min \[[@B28]\]. Elution was monitored at 259 nm and peaks identified with the aid of standards and quantified by area integration. GFP fusion constructs and subcellular localization -------------------------------------------------- Expression plasmids encoding C-terminal and N-terminal fusions of Y87G2A.14 to yeast-enhanced green fluorescent protein (yEGFP) \[[@B29]\] were constructed by amplification of the coding region of Y87G2A.14 from *C. elegans* cDNA by PCR using the same forward primer 5\' CGACGGATCCATGAAGTGTGT 3\' and one of the reverse primers 5\' TAAATGAATTCACTAAATTTTGGATTTCGGTTC 3\', 5\' CACTAAGAATTCTATTTCGGTTCAAATTTCCTACTTGC 3\', or 5\' GCTCGAATGAATTCAATTTTGGATTTCGGTTC 3\' to give PCR products \"C\", \"CΔSKI\" or \"N\" respectively. These primers provided a *Bam*H1 restriction site at the start of the amplified gene and *Eco*R1 sites at the end. PCR products \"C\" and \"CΔSKI\" were cloned as C-terminal fusion proteins to yEGFP, while PCR product \"N\" with a deletion of the Y87G2A.14 termination codon was cloned as an N-terminal fusion to yEGFP. After amplification with *Pfu* DNA polymerase, the DNA products were recovered by phenol/chloroform extraction and digested with *Bam*H1 and *Eco*R1. The digested PCR products \"C\", and \"CΔSKI\" were gel purified and the restriction fragments ligated between the *Bam*H1 and *Eco*R1 restriction sites of pUG36 (yEGFP-C-fusion) to give pyEGFP-Y87G2A.14 and pyEGFP-Y87G2A.14ΔSKI respectively. The digested PCR product \"N\" was gel purified and the restriction fragment ligated between the *Bam*H1 and *Eco*R1 restriction sites of pUG35 (yEGFP-N-fusion) to give pY87G2A.14-yEGFP. The structures of the inserts were confirmed by sequencing. The plasmids were propagated by transformation of *E. coli* XL 1-Blue cells. For microscopy, *S. cerevisiae* strain BY4741 was transformed with pyEGFP-Y87G2A.14, pyEGFP-Y87G2A.14ΔSKI or pY87G2A.14-yEGFP and grown on solid SC-Ura medium containing 2% glucose. Cells were viewed by conventional and confocal fluorescent microscopy on a Zeiss LSM510 confocal microscope with a 100 × 1.4 NA objective. Other methods ------------- Protein concentrations were estimated by the Coomassie blue binding dye-based colorimetric method using equal weights of bovine serum albumin, conalbumin, cytochrome c and myoglobin as standards \[[@B30]\]. Acknowledgements ================ S.R. AbdelRaheim is the recipient of a postgraduate scholarship from the Egyptian government. The authors are grateful to D.G. Spiller for assistance with the confocal microscopy. The authors thank J.H. Hegemann and H.M. Abdelghany for their kind gift of materials.

Background ========== *Lathyrus sativus*, belonging to the family of leguminoceae, is a popular drought resistant legume plant whose grains serve as food in the drought prone areas of Asia and Africa. The 2,3 α-β-Diamino propionic acid (DAP) present in *L. sativus* seeds when consumed causes Lathyrism in humans, affecting either the nervous system (Neurolathyrism) or bone formation (Osteolathyrism), \[[@B1]\]. Medical scientists evince keen interest in understanding the causes of neurolathyrism as a model for neurodegenerative diseases striking the more affluent sections of the human population. The scourge of neurolathyrism, known since ancient times, mainly affects the poor rural classes in developing countries. Though different methods such as boiling in water and development of DAP free cultivars \[[@B2]\], had been tried, as yet no effective method has been developed for safe consumption by the famine hit, poverty ridden, malnourished hungry millions in the drought prone areas in central India and Africa. Rajagopal et al. (1970) \[[@B3]\] have studied the metabolism of DAP in *Pseudomonas* and demonstrated that it is metabolized to pyruvic acid through Ammonia lyase reaction. Very scant information is available on the physicochemical and catalytic properties of diamino propionic acid ammonia lyase. In this communication we report the production of DAP ammonia lyase by *S. typhimurium* PU011 and describe a few of its properties. Results and Discussion ====================== Virulence Assay --------------- The results of virulence assay performed in mice employing *S. typhimurium* PU011 strain are shown in Table [1](#T1){ref-type="table"}. To be absolutely certain that enough number of virulent *S. typhimurium* and *S. typhimurium* PU011 are present in the mice after oral administration, to bring about mortality, each of the Salmonella strain at 2.5 × 10^6^ CFU per mouse, which is approximately twice the reported 50% Lethal Dose (LD~50~) \[[@B4],[@B5]\] was employed. In the light of observations reported by Leung et al (1991) that the administration of *S. typhimurium* in mice resulted in death within 28 days, the experiment was continued up to 28 days post inoculation. The mice administered with virulent *S. typhimurium* died on 8^th^ and 9^th^ day of post inoculation. However, the mice administered with *S. typhimurium* PU011 survived even after 28 days. ::: {#T1 .table-wrap} ::: {.caption} ###### Virulence of *S. typhimurium* in an orally infected typhoid mouse model ::: **Bacteria** **Percent Survival^\*^** **Clinical Symptoms of liver and spleen** **Conclusion** ------------------------------ -------------------------- ------------------------------------------- ---------------- *S. typhimurium* PU011 100% No symptoms Avirulent *S. typhimurium* 0% Enlarged, inflamed & edematous Virulent Control (Mice with PBS only) 0% No symptoms Avirulent ^\*^Average of six replicates ::: Further, the mice injected with virulent *S. typhimurium* showed the clinical symptom of diarrohea prior to death, which was conspicuously absent in mice that were inoculated with *S. typhimurium* PU011. The clinical examination of sacrificed mice injected with virulent *S. typhimurium* revealed typical symptoms of enlarged, inflamed and edematous liver and spleen, which were absent in mice, injected with *S. typhimurium* PU011 strain. Conclusively this is evident enough to show that *S. typhimurium* PU011 strain employed in our studies is a non-virulent strain. To the best of our knowledge this is the first report about the production of DAP ammonia lyase by a non-virulent *S. typhimurium* with wider ramifications. Biomass and Enzyme production ----------------------------- Vijayalakshmi (1975) \[[@B6]\] and Nagasawa (1988) \[[@B7]\] have identified *Pseudomonad* and *Salmonella typhimurium* respectively capable of producing DAP ammonia lyase that degrades the neurotoxin present in *L. sativus* seeds. We have accidentally stumbled upon a non-virulent bacterium *S. typhimurium* PU011 which was capable of producing DAP ammonia lyase. The kinetics of production of biomass and DAP ammonia lyase activity by this organism is shown in Fig [1](#F1){ref-type="fig"}. A continuous increase in biomass and specific activity of enzyme with increase in incubation period was observed, the maximum reaching at 16 h post inoculation. Interestingly, a direct correlation between biomass yield and enzyme activity up to 20 h after inoculation was observed. Nagasawa et al. (1988)\[[@B7]\] have also observed a similar pattern of growth and enzyme activity in *S. typhimurium*. {#F1} As to whether any carbon source other than DAP in the growth medium could support growth and enzyme production, it was observed that none other than DAP could substitute any of the carbon sources (data not shown). Since *S. typhimurium* PU011 employed in the present study was found to synthesize DAP ammonia lyase, it was thought fit to employ this enzyme for detoxification of DAP present in *L. sativus* seeds prior to consumption. Therefore, the characteristics of this enzyme were studied in detail. Kinetic properties ------------------ The Km value for the substrate, DL-DAP, was found to be 0.685 mM based on the Line weaver Burk plot (Fig [2](#F2){ref-type="fig"}). However, for the same enzyme from *Pseudomonas sp.* and *S. typhimurium*, Vijiyalakhsmi et al (1975) \[[@B6]\] and Nagasawa et al (1988) \[[@B7]\] have reported Km values at 0.267 mM and 0.12 mM respectively. The differences in Km values were attributed to D or L or DL forms of DAP used as substrate in the estimation of enzyme activity. {#F2} Effect of pH ------------ With regard to specific activity of DAP ammonia lyase in different buffer systems at various pH regimes, it was found that the enzyme had a sharp pH optimum of 8.0 in phosphate buffer and at any pH less than 6 or more than 11 a drastic reduction in activity was observed (Fig [3](#F3){ref-type="fig"}). Vijiyalakhsmi et al. (1975) \[[@B6]\] also have observed an optimum pH of 8.0 for maximum DAP ammonia lyase enzyme activity produced by *Pseudomonas*. However, in their studies Tris HCl buffer at pH 8.0 was found to inhibit 50% of enzyme activity. {#F3} Effect of temperature --------------------- The enzyme activity was measured using spectrophotometer at temperatures ranging between 30°C -- 60°C at increment of 5°C. Up to 45°C there was no decrease in enzyme activity; however, thereafter a decline in enzyme activity with increase in temperature was observed (Fig [4](#F4){ref-type="fig"}). The loss in DAP ammonia lyase enzyme activity with increase in temperature indicates the level of stability of the enzyme *per se* to different temperature regimes and not temperature activity relationship since the enzyme was not incubated along with the substrate. While 45°C was found to support maximum enzyme activity in *Salmonella sp*. (Nagasawa et al., 1988) \[[@B7]\], 40°C was found to support maximum activity in *Pseudomonas* (Vijiyalakshmi et al., 1975)\[[@B6]\]. The fact that maximum activity was scored at above ambient temperature is a good augury for the use of this enzyme in the field scale applications for detoxification of DAP in *L. sativus* in drought prone areas of the world where elevated temperatures and desiccation pose threats for inactivation of enzyme activity. {#F4} Purification of DAP ammonia lyase from *S. typhimurium* PU011 ------------------------------------------------------------- According to the procedure described in the text, the enzyme was purified 68 fold (Table [2](#T2){ref-type="table"}). When the purified enzyme was subjected to SDS-PAGE, a single band was observed corresponding to the molecular weight of 43 KDa (Fig [5](#F5){ref-type="fig"}). This is in conformity with the results of Vijiyalakshmi (1975) \[[@B6]\]. However, Nagasawa et al (1988) \[[@B7]\] reported that the relative molecular mass of DAP ammonia lyase produced by *Salmonella typhimurium*, estimated by ultra centrifugal equilibrium method, was 89,000 Da and the enzyme consisted of 2 subunits identical in Molecular mass. {#F5} ::: {#T2 .table-wrap} ::: {.caption} ###### Purification Table of Diamino propionic acid ammonia lyase ::: Treatments Total Protein (mg) Total Activity (U) Specific Activity (U/mg) --------------------------------- -------------------- -------------------- -------------------------- Crude Extract 214.0 40.02 0.19 Ammonium Sulphate precipitated 97.7 39.17 0.40 Passed through Sephadex -- G100 10.27 36.74 3.58 Passed through CM -- Sephadex 3.8 36.71 9.66 Passed through DEAE -- Sephacel 1.6 21.76 13.60 ^\*^Average of six replicates ::: The discovery of non-virulent *S. typhimurium* PU011 capable of producing DAP ammonia lyase has opened new vistas in the application of this enzyme in the tribal tracts of arid and semi-arid regions of Central India. For commercial exploitation of this enzyme, attempts are underway to clone and overexpress the gene coding for DAP ammonia lyase for applicative purposes in field. Conclusion ========== At 16 h post inoculation *S. typhimurium* PU011 produced maximum biomass yield and maximum enzyme activity in minimal medium containing DAP. Interestingly no carbon source other than DAP supported DAP ammonia lyase synthesis in the medium effectively. With respect to specific enzyme activity of DAP ammonia lyase at different pH regimes, the enzyme had a sharp activity at pH 8.0 in phosphate buffer and at any pH less than 6 or more than 11 a drastic decrease in enzyme activity was observed. Up to 45°C there was no decrease in enzyme activity; however, thereafter there was decline in enzyme activity with increase in temperature. Based on SDS-PAGE analysis of the DAP ammonia lyase enzyme, the molecular weight was determined to be 43 KDa having a Km value of 0.685 mM. Materials and Methods ===================== Culture and media ----------------- *Salmonella typhimurium* PU011 employed in the present study was isolated in our lab and further characterized by standard microbiological, serological and biochemical tests \[[@B8]\]. Cells were grown at 30°C in minimal medium (0.05% Yeast Extract, 500 μl Mineral salt solution and 2.0% Agar) supplemented with DL -- DAP (0.3%) for production of DAP ammonia lyase by the bacterium. Virulence Assay --------------- Virulent *S. typhimurium* (obtained from the Department of Microbiology, Jawaharlal Institute of Post Graduate Medical Education and Research (JIPMER), Pondicherry, India) and *S. typhimurium* PU011 suspensions were grown at 37C in LB broth overnight, with shaking at 200 rpm. The next day, bacteria were diluted 1:100 into fresh LB broth and incubated with shaking. After 4 h, bacteria were washed once with PBS and resuspended in PBS. BALB/c female mice, aged 6--10 weeks, were inoculated orally with bacterial suspensions at 2.5 × 10^6^ CFU per mouse, which is approximately twice the reported 50% Lethal Dose (LD~50~) \[[@B4]\], after being deprived of water for 4 h. Control mice were given PBS only. The liver and spleen of experimental mice that died on different days after administration of *S. typhimurium* were harvested, homogenized and the resultant slurry was spread on Mac Conkey plates to obtain bacterial counts. After sacrifice, the mice were critically examined for the presence of typical symptoms of enlarged, inflamed and edematous liver and spleen. Mice surviving after 28 days post inoculation were indicative of the non-virulence nature of *S. typhimurium* employed. Determination of Biomass and Enzyme activity -------------------------------------------- Cells were grown in minimal medium supplemented with 0.3% DAP at 30°C on a rotary shaker (180 rpm) as per the method outlined by Premkumar \[[@B9]\] and Rupesh \[[@B10]\]. For estimation of biomass, at various intervals of time 10 ml of culture was withdrawn, centrifuged at 8000 g for 10 min, washed once with distilled water, dried at 105°C and weighed. However, for estimation of enzyme activity 25 ml of culture was withdrawn, centrifuged at 8000 g for 10 min at 4°C, suspended in 0.5 ml of extraction buffer (10 mM Potassium phosphate buffer pH 8.0, 1 mM EDTA, 1 mM β -- Mercaptoethanol, 100 μM Pyridoxal-5-Phosphate and 5% Glycerol) and sonicated for 10 min using Branson Sonifier 250 (10 min bursts at 40% duty cycle with intermittent cooling every 2 min in order to maintain the suspension below 5°C). After sonication, the lysate was centrifuged at 12,000 g for 10 min at 4°C and the supernatant served as the crude extract for the estimation of enzyme activity. The DAP ammonia lyase activity was measured as per the method of Vijiyalakhsmi et al \[[@B6]\]. Protein estimation was done as per the method of Lowry et al \[[@B11]\]. Enzyme Assay (Vijayalakshmi et al., 1975) ----------------------------------------- DAP ammonia lyase was assayed spectrophotometrically, using a coupled assay system, utilizing the following reaction.  Sum: DAP + NADH + H~2~O + H^+^ → Lactate + NAD + 2NH~3~ The reduction in absorbance at 340 nm was measured in Hitachi U-2000 spectrophotometer. The assay system contained in a total volume of 3.0 ml in a 1 cm light path cuvette: 1 mmol Potassium phosphate buffer pH 8.0, 20 mmol DL-DAP, 0.3 μmol PLP, 15 Units of LDH, 1 μmol NADH and appropriate amount of enzyme. The reaction was studied at 28°C, and was started by the addition of DAP. The decrease in absorbance was recorded at 30 s intervals for a period of 2 min. One Unit of enzyme was defined as the amount of enzyme required for the conversion of 1 μmol of NADH to NAD^+^ in one minute. Specific activity is defined as Units per mg of protein.  Effect of pH ------------ Optimum pH at which a maximum specific activity of DAP ammonia lyase could be obtained was determined by employing different buffer systems at various pH as indicated in parenthesis. The buffer systems employed were: Glycine-HCl buffer (pH 3), Sodium acetate buffer (pH 5), Potassium phosphate buffer (pH 7), Potassium phosphate buffer (pH 8), Glycine-NaOH (pH 9), Glycine-NaOH (pH 11) and KCl-NaOH (pH 13). The crude enzyme extract was taken and incubated with the above buffers for 30 min. After incubation, the crude extract was precipitated with ammonium sulfate (80% saturation) and the precipitate was resuspended in minimal volume of 10 mM potassium phosphate buffer. The effect of pH was studied by measuring DAP ammonia lyase activity as per the procedure described earlier. Effect of Temperature --------------------- The crude enzyme extract in 10 mM Potassium phosphate buffer pH 8.0 was incubated at various temperatures ranging from 0°C -- 70°C. After 30 min of incubation, the crude extract was analyzed for DAP ammonia lyase activity. Purification of DAP ammonia lyase --------------------------------- ### Preparation of Cell free crude extract Cell free crude extract of *S. typhimurium* PU011 was prepared by sonicating the pellet in 0.5 ml extraction buffer for 10 min using Branson Sonifier 250. The sonicate was then centrifuged at 12,000 g for 10 min at 4°C and the supernatant served as the crude enzyme extract. Ammonium sulfate fractionation ------------------------------ The enzyme was precipitated by addition of appropriate amounts of ammonium sulfate (35--65% saturation) as per the method of Nagasawa et al, (1988) \[[@B7]\]. The precipitate was collected by centrifugation and dissolved in 1 ml of extraction buffer and dialyzed against the same buffer for 24 h. Chromatography on Sephadex G-100 -------------------------------- The crude enzyme preparation was subjected to gel filtration on sephadex G-100 (1.6 × 20 cm) column chromatography. The dialyzed enzyme solution was applied to sephadex G-100 column, which had been equilibrated with 10 mM phosphate buffer. The active fractions that showed absorption at 280 nm were collected separately and DAP ammonia lyase in each fraction was quantified. The fractions that showed maximum enzyme activity were pooled and passed through CM-Sephadex column. Chromatography on CM-Sephadex ----------------------------- Chromatography on CM-Sephadex column was performed on column (2 × 12.5 cm) of the resin equilibrated with 10 mM phosphate buffer pH 8.0, 1 mM β-mercaptoethanol and 1 mM pyridoxal-5-phosphate. The unbound fraction containing DAP ammonia lyase activity was reloaded onto DEAE-Sephacel column. Chromatography on DEAE-Sephacel ------------------------------- The active enzyme fraction was applied to a DEAE-Sephacel column (2 × 20 cm), which had been equilibrated with 10 mM phosphate buffer pH 8.0, 1 mM β-mercaptoethanol and 1 mM pyridoxal-5-phosphate. The enzyme fractions were eluted using 0 -- 1 M KCl gradient. All the column chromatography purifications were carried out in BIORAD Econo system. The active fractions were precipitated with 80% ammonium sulfate. The precipitate was resuspended in a minimum amount of 10 mM phosphate buffer and dialyzed against the same buffer for 24 h with three changes under continuous stirring at 4°C. The progress of purification of DAP ammonia lyase was monitored both by enzyme activity measurement and on 12% SDS-PAGE. SDS-Polyacrylamide Gel Electrophoresis (SDA-PAGE) (Laemmli, 1970) \[[@B12]\] ---------------------------------------------------------------------------- Routinely 0.75 mm thick gels were cast in a protein gel apparatus using a separating gel (12 %) and stacking gel (5%). The separating gel solution was first poured in between a sealed clean glass plate cassette leaving 3 to 4 cm space at the top. Water saturated n-butanol was layered gently on the top of the separating gel to make an even surface. After the separating gel got polymerized, the upper layer was removed completely and rinsed with water. Water was removed completely using a blotting paper. Stacking gel solution was poured onto the separating gel up to the top. A suitable comb was immediately inserted into the cassette and polymerization was allowed to take place. The slots were cleaned by rinsing with electrophoresis buffer. Protein samples were loaded into the slots and the slots were filled with 1 × electrophoresis buffer. The gel cassette was then fixed with the upper buffer tank and then fixed onto the lower buffer tank. Both the tanks were filled with 1 × electrophoresis buffer and electrophoresis was carried out at a constant current of 15 mA. After electrophoresis was over, the gel cassette was dismantled and the gel was placed in a tray containing Coomassiae Brilliant Blue staining solution (0.1 % Coomassie Brilliant Blue (R 250), 45 % Methanol, 10 % Glacial Acetic Acid and 45 % Distilled water) and the gel was agitated slowly for 4 h to overnight. The staining solution was poured out and replaced with destaining solution (45 % Methanol, 10 % Glacial Acetic Acid and 45 % Distilled water). Destaining was carried out for a minimum of 2 h with gentle agitation. Acknowledgements ================ The assistance rendered by M. K. Natarajan, JIPMER, Pondicherry, India in carrying out the microbiology work is gratefully acknowledged.

Background ========== Rapid and accurate detection, identification and susceptibility testing of mycobacteria remains important i) because the overall incidence of tuberculosis is increasing, also due to the HIV pandemic \[[@B1]\], ii) because of the increasing resistance to antituberculous agents \[[@B2]\] and iii) because an increasing number of mycobacterial species are being recognized as potentially pathogenic \[[@B3]\]. The importance of *M. avium* infection has increased in HIV patients \[[@B4]\], and clinical infections have been described with species like *M. heidelbergense*\[[@B5]\], *M. conspicuum*\[[@B6]\], *M. branderi* \[[@B7]\] and *M. interjectum*\[[@B8],[@B9]\], which have been recognized only recently. Current DNA amplification based diagnostic tests are expensive, have limited senstivity, are usually restricted to the detection of *M. tuberculosis* only and provide no or limited information on susceptibility (e.g. rifampicin only: RifTB LiPA, Innogenetics, Zwijnaarde, Belgium). Therefore, the need for culture has not been circumvented. The CDC decided to restrict the use of genotypic tests to confirmation of smear positive samples, so that they cannot be used to test the large number of specimens processed for mycobacterial detection every year in an average laboratory \[[@B10]\]. Only recently, the enhanced AMTD (Gen-Probe, San Diego, CA) gained FDA approval for direct detection of *M. tuberculosis* from smear-negative samples, but several problems are reported \[e.g. \[[@B11]\], and the high costs keep restricting its use. Here, we present our findings with the use of ARDRA \[[@B12]\] for the identification of *Mycobacterium* species. The method consists of amplification of the 16S rRNA gene (rDNA) and subsequent restriction digestion of the amplicon. The restriction patterns obtained with different restriction enzymes and combination of these patterns into a restriction profile was shown to enable identification of most clinically important mycobacteria by comparison of the obtained profiles with a library of ARDRA profiles obtained for reference strains of different species \[[@B12]\]. This PCR-RFLP analysis of the 16S rRNA gene, was published almost simultaneously with the more widely used technique (known as PRA), which is based on the amplification of the *hsp65* gene \[[@B13]-[@B15]\]. Results ======= The initial study describing the applicability of ARDRA for the identification of mycobacteria \[[@B12]\] used universal bacterial primers. However, this sometimes resulted in the false positive amplification from decontaminated samples of organisms other than mycobacteria. Therefore primers were developed, aimed at more specific amplification of mycobacteria. During the three year evaluation period, of which the results are reported here, amplification of nonmycobacterial organisms occurred in two cases. These organisms, namely *Corynebacterium glutamicum* and *Actinomyces odontolyticum,* stained acid fast on direct smear and are relatives of the mycobacteria. The restriction patterns obtained with the enzymes *Hha*I (isoschizomer of *CfoI*), *Mbo*I and *Rsa*I for the different species are numbered arbitrarily and are presented in Figures [1](#F1){ref-type="fig"}, [2](#F2){ref-type="fig"} and [3](#F3){ref-type="fig"} respectively. Figures [4](#F4){ref-type="fig"},[5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"} represent the restriction patterns obtained when digital restriction is carried out with the same enzymes on published GenBank sequences. The combination of these patterns is designed as ARDRA profiles and these are listed in Table [1](#T1){ref-type="table"}. For example, strains of the *M. tuberculosis* complex can be recognized by an ARDRA profile 1-1-1, while *M. grodonae* strains have ARDRA profile 8-4-2. For some species the ARDRA pattern obtained with enzyme is already characteristic, e.g. *Hha*I 1 is observed only for species of the *M. tuberculosis* complex. {#F1} {#F2} {#F3} {#F4} {#F5} {#F6} ::: {#T1 .table-wrap} ::: {.caption} ###### Library of ARDRA profiles (combination of restriction patterns) obtained for mycobacterial species. ::: **Species** **Genbank Number** ***Hha*I** ***Mbo*I** ***Rsa*I** ***Bst*UI** **Reference strains used^a^** ----------------------------- -------------------- ------------ ------------ ------------ ------------- ------------------------------- *M. tuberculosis* complex X52917 1 1 1 1 ITG 8021, IPB 92/0805 *M. conspicuum* X88922 1\' 1 2 2 *M. intracellulare* X52927 1\' 2 2 3 ITG 5913, ITG 5917 *M. gastri* X52919 1\' 4 1 1 *M. kansasii* M29575 1\' 4 1 1 ITG 8201 *M. bohemicum* AJ277283 1\' 4 1 3 *M. haemophilum* U06638 1\' 4 1 3 ITG 3065 *M. malmoense* AF152560 1\' 4 1 3 ITG 940611 *M. szulgai* X52926 1\' 4 2 1 ITG 4981 *M. scrofulaceum* X52924 1\' 4 2 3 ITG 4988 *M. xenopi* X52929 1\' 4 3 2 ITG 4986 *M. heckeshornense* AF174290 1\' 4 3 4 *M. marinum* AF251565 2 1 1 1 ITG 1728 *M. asiaticum* M29556 2 1 2 5 ITG 8182 *M. terrae* X52925 2 1 4 2 ITG 4922 *M. ulcerans* Z13990 2 1 9 1 ITG 724, ITG 1837 *M. avium* AF306455 2 2 2 3 ITG 4991, ITG 2666 *M. botniense* AJ012756 2 4 3 2 *M. terrae*-like MCRO6^b^ X93032 2 6 4 7 *M. nonchromogenicum* X52928 2 6 4 7 ITG 4980 *M. phlei* M29566 3 1 6 2\' *M. elephantis* AJ010747 3 1 8 3\' *M. chelonae* group IV^c^ AJ416940 3 3 5 2\' ITG 7701 *M. abscessus* group III^c^ AJ419970 3 3 6 2\' ITG 98-1296 *M. chelonae* group 1^c^ AJ419969 3 3 6 2\' ITG 95-0026 *M. immunogenum* AJ011771 3 3 6 2\' ITG 98-1289 *M. farcinogenes* AF055333 4 1 6 7 *M. fortuitum* X52933 4 1 6 7 SLZ A046 *M. senegalense* M29567 4 1 6 7 *M. septicum* AF111809 4 1 6 7 ITG 4166 *M. intermedium* X67847 5 4 6 1\' *M. interjectum* AJ272088 5 4 6 2\' ITG 96-116 *M. heidelbergense* X70960 5 4 8 2\' *M. simiae* X52931 5 7 6 2\' ITG 4485 *M. lentiflavum* X80769 5 7 6 2\' *M. peregrinum*^d^ X52921 7 1 6 9 ITG 99-2069, ATCC 14467 *M. gordonae* M29563 8 4 2 5 ITG 7838 *M. wolinskyi* Y12871 10 1 1 2\' *M. species*^a^ AF028712 10 1 1 2\' *M. goodii* Y12872 10 1 2 2\' *M. smegmatis* AJ131761 10 1 2 2\' ITG 4995 *M. kubicae* AF133902 10 1 6 1\' *M. tusciae* AF058299 10 1 6 2\' *M. flavescens* X52932 10 1 6 2\' VUB A016 *M. genavense* X60070 10 7 6 2\' ITG 97-102 *M. triplex* U57632 10 7 6 2\' *M. chelonae* group II^c^ AJ419968 13 3 6 6 ITG 96-0295 *M. branderi* X82234 14 1 1 9 *M. celatum*^e^ L08169 14 1 1 10 *M. triviale* X88924 15 1 10 10\' *M. aurum* X55595 16 1 6 8 a\. ITG: Institute for Tropical Medicine, Antwerp, Belgium; IPB: Institute Pasteur du Brabant, Brussels, Belgium; SLZ: Streeklaboratorium Zeeland, Goes, the Netherlands; VUB: Free University of Brussels, Brussels, Belgium. b. Strains designated MCRO6 have been studied by Turenne *et al.*\[[@B22]\] and Torkko *et al.*\[[@B33]\]. c. Sequences determined in this study. Roman numbering according to Portaels *et al.*\[[@B16]\], who distinguished four groups within the *M. abscessus/M. chelonae* complex, based on the 16S--23S rRNA spacer region. d. The GenBank sequence (AF058712) of strain ATCC 14467, submitted to GenBank as *M. peregrinum,* did not cluster within the *M. fortuitum* complex (Figure [8](#F8){ref-type="fig"}) and was highly similar to sequence Y12871 (*M. wolinskyi*). Moreover, the ARDRA profile calculated from sequence AF058712 (10-1-1-2\') did not correspond with the profile we obtained for strain ATCC 14467 (4-1-6-7). The sequence obtained in this study from strain ATCC 14467 (submitted as AJ422046) was identical to the *M. fortuitum* sequence X52933. Sequence AF058712 is indicated as *M. species* in the table. e. Received as *M. xenopi.* ::: Most mycobacterial species could be readily identified by comparison of the obtained ARDRA profile with the profiles from Table [1](#T1){ref-type="table"}. The species of the *M. tuberculosis* complex can not be differentiated on the basis of the 16S rDNA sequence, and therefore restriction digestion of this gene could not either. Most of the clinically relevant and the most abundant species were readily differentiated from each other. The following species could not be differentiated from each other after digestion with *Hha*I, *Mbo*I, *Rsa*I and *Bst*UI: *M. gastri* and *M. kansasii* (1\'-4-1-1), *M. bohemicum, M. haemophilum* and *M. malmoense* (1\'-4-1-3), MCRO6 and *M. nonchromogenicum* (2-6-4-7), *M. chelonae* group I, *M. abscessus* and *M. immunogenicum* (3-3-6-2\'), *M. farcinogenes, M. fortuitum, M. senegalense* and *M. septicum* (4-1-6-7), *M. simiae* and *M. lentiflavum* (5-7-6-2\'), *M. goodii* and *M. smegmatis* (10-1-2-2\'), *M. tusciae* and *M. flavescens* (10-1-6-2\'), and finally *M. genavense* and *M. triplex* (10-7-6-2\'). During 1998--2000, approximately 3500 samples were sent to the laboratory for direct smear examination and mycobacterial culture. Of these, 151 specimens, from 149 patients, were culture positive, and 20% of these were also smear positive. Table [2](#T2){ref-type="table"} summarizes the obtained identifications. For 148 of the 151 isolates, identification by ARDRA was straightforward and was obtained after an average of 36 hours after receipt of the cultured strain. Only three difficulties were encountered. One isolate presented with the ARDRA profile 5-4-6, not present in the ARDRA library at that time. Sequencing lead to an identification as *M. interjectum*\[[@B8]\], a species that was not yet covered by the library. A second isolate was first misidentified as *M. xenopi* (profile 1\'-4-3), but later on it was observed that the *Hha*I fingerprint differed clearly from the *M. xenopi Hha*I 1\' pattern by its low molecular size fragments (Figure [1](#F1){ref-type="fig"}). This *Hha*I pattern was designated *Hha*I 1\", resulting in the unique ARDRA profile 1\"-4-3. The isolate was identified as *M. heckeshornense* by sequencing of the 16S rRNA gene, and can now also be identified by ARDRA. The third problematic isolate had ARDRA profile 1-1-3, again a profile that had never been observed for any mycobacterial strain studied thus far. Sequencing of the 16S rRNA gave a 99.8 % similarity to the 16S rRNA sequence of the type strain of *M. tuberculosis.* The sequence revealed a mutation at *E. coli* position 646 from A to G, abolishing the *Rsa*I restriction site GTAC at that place. This mutation shifts the *Rsa*I pattern 1 to *Rsa*I pattern 3 because the two fragments of resp. 620 and 180 bp are replaced by a single fragment of 800 bp. Further morphological and biochemical tests revealed an identification as probably *M. africanum,* one of the species of the *M. tuberculosis* complex. It should be mentioned that *M. africanum* reference strains used in a previous study \[[@B12]\], were found to have the regular *M. tuberculosis* complex ARDRA profile 1-1-1. ::: {#T2 .table-wrap} ::: {.caption} ###### Distribution of the *Mycobacterium* species during 1998--2000 in the Ghent University Hospital, number of smear positives and time to positivity of culture. ::: *Mycobacterium* species Number of isolates (%) Number of samples positive on direct smear ^a^ (% of culture positive) Average time until positive culture in days (smear positives; negatives) --------------------------- ------------------------ ------------------------------------------------------------------------ -------------------------------------------------------------------------- *M. tuberculosis* complex 61 (40) 26 (46^b^) 16.7 (12.9; 20) *M. avium* 12 2 6.5 *M. heckeshornense* 1 1 32.5 *M. chelonae* 4 0 18 *M. fortuitum* 2 0 35.5 *M. gordonae* 41 (27) 1 22.3 *M. interjectum* 1 0 30.5 *M. intracellulare* 1 0 11.6 *M. kansasii/M. gastri* 2 0 No data *M. malmoense* 1 0 No data *M. szulgai* 3 0 20.3 *M. xenopi* 22 (15) 0 38.1 *Total* 151 40 (26) ^a^: No data for 6 of the 61 samples. All six of these samples were *M. tuberculosis* culture positive. ^b.^ Calculated as 26 positives on a total of those 56 culture positives for which data on smear positivity were available. ::: To construct artificial ARDRA patterns for the recently described species, we applied the programme RFLP on the published Genbank sequences. The resulting, theoretically to be expected, ARDRA profiles are presented in Table [1](#T1){ref-type="table"}. *M. tuberculosis* was found to be the most prevalent species, with 40% of the isolates, followed by *M. gordonae* and *M. xenopi,* species mostly isolated form non-pulmonar samples with low clinical relevance. Similarity calculation of published sequences and of sequences obtained in this study (Figure [8](#F8){ref-type="fig"}) indicates that the *M. abscessus/M. chelonae* complex (*M. chehnae, M. abscessus* and *M. immunogenum*) clusters separately from all other mycobacteria. Portaels *et al.*\[[@B16]\] described the presence of four groups within the *M. abscessus/M. chelonae* complex, based on the 16S--23S spacer sequences, as used in the INNO-LiPA Mycobacteria kit (Innogenetics, Zwijnaarde, Belgium). Groups II and IV consisted of genuine *M. chelonae* strains, which were usually from environmental sources. These two groups have characteristic ARDRA profiles, namely 13-3-6 and 3-3-5. Group III, which was found to represent genuine *M. abscessus* and *M. chelonae* group I, which most probably corresponds with *M. immunogenum,* a *M. abscessus*-like species that was recently described as being frequently involved in bronchoscope related pseudo-outbreaks \[[@B17]\], can not be distinguished from each other by ARDRA (profile 3-3-6). Strains of groups I and III are usually isolated from clinical samples, with only group III strains (genuine *M. abscessus*) being pathogenic \[[@B16]\]. {#F7} {ref-type="table"} lists the GenBank accession numbers of the sequences used to construct this tree. ARDRA patterns for *Hha*I, *Mbo*I, *Rsa*I and *Bst*UI are listed after the species name. a: GenBank AF028712. Erroneously listed in GenBank as *M. peregrinum* (see also legend of Table [1](#T1){ref-type="table"}). b: *M. gastri* clusters below 100% with *M. kansasii,* although it is generally agreed that the 16S rRNA gene sequences for *M. kansasii* and *M. gastri* are identical. This can be explained by the fact that the only available GenBank *M. gastri* sequence (X52919) contained several ambiguities. c. *M. lentiflavum,* initially not included in the manuscript is not presented in this tree. It clusters close to the branch including *M. heidelbergense, M. simiae, M. triplex* and *M. genavense.*](1471-2180-2-4-8){#F8} *M. tuberculosis* was the most frequently cultured species (40% of all culture positives), followed by *M. gordonae* (27%) and *M. xenopi* (15%) The latter two species were mostly isolated from non-pulmonar samples. 46% of the samples culture positive for *M. tuberculosis,* were auramine staining positive, and the average time of incubation until a positive liquid culture of *M. tuberculosis* was 12.9 days for the auramine positive samples and 20 days for auramine negative samples. The range of time to positivity for all *M. tuberculosis* positive samples was between 6 and 29 days. The average culture times for species other than *M. tuberculosis,* as far as data were available, are presented in Table [2](#T2){ref-type="table"}. Discussion ========== Restriction analysis of the amplified 16S rRNA gene, or amplified rDNA restriction analysis (ARDRA) was introduced into the Laboratory for Bacteriology of the Ghent University Hospital for the identification of cultured mycobacteria in 1993 \[[@B12]\]. Since then, several comparable approaches, based on restriction digestion of the amplified rRNA genes and spacer regions have been described \[[@B18]-[@B21]\]. During that period, this approach has been updated and refined. This was possible due to some technical changes, like increased quality control of gel electrophoresis and pattern interpretation and the use of primers specific for species of the order of the Actinomycetales instead of universal bacterial primers. Refinement was also possible because of the improvement of mycobacterial taxonomy and the possibility offered by PCR-RFLP techniques, like ARDRA, to easily adapt to this new information. Indeed, when new species are described, there is no need to develop new probes or primers. Instead, new ARDRA profiles can be easily added to the existing library. Also, ARDRA profiles for newly described species can be predicted by applying computer aided digestion of the available GenBank sequences, given the availability of sequences of sufficient quality \[[@B22]\]. ARDRA was found to be a useful tool for identification of mycobacterial isolates in a clinical routine laboratory, because of its speed -- compared to phenotypic identification, its reliability, practical applicability, flexibility and the possibility to identify most nontuberculous mycobacteria together with and at the same cost as *M. tuberculosis,* at an affordable price. Of the 151 isolates during the last three years, 148 could be identified without problems. The other three isolates, respectively *M. interjectum, M. heckeshornense* and an *M. africanum*-like strain, should be identifiable when met again in the future, since they presented with specific ARDRA profiles. Practical applicability of ARDRA -------------------------------- The theoretical turnaround time of ARDRA is 6 hours, and the average identification time in practice during this study was 36 hours. It should be emphasized that the technique was not fully implemented in the routine laboratory, but was carried out by the research laboratory technicians, which means that the practical turnaround time should be far less than 36 hours in a routine diagnostic laboratory. Technically, ARDRA is nondemanding, comprising only basic molecular biology techniques like simple DNA extraction, PCR, restriction digestion and submarine agarose gel electrophoresis. General considerations ---------------------- We have addressed previously the several limitations of molecular biology based detection in diagnostic bacteriology \[[@B23]\]. Others agree that the expectations that DNA amplification technologies would supplant microscopy, accurately predict culture results and provide an immediate definitive diagnosis were premature and that these claims have to be replaced with a more realistic view of the limitations and of the practical value of molecular diagnostics of tuberculosis \[[@B10],[@B24],[@B25]\]. Also the expectation that susceptibility would be carried out solely by means of DNA technology, had to be moderated \[[@B26]\]. Despite the fact that during the last ten years a tremendous effort, both in academic and commercial research, has been put into the applicability of nucleotide amplification techniques for the detection of mycobacteria directly from clinical samples, the CDC approved application of these techniques only for smear positive samples. This implicates that for 52% of the culture positive samples with *M. tuberculosis* encountered in this study, DNA technology would not have accelerated detection, since microscopy was negative. Recently, AMTD2 (GenProbe) gained FDA approval for testing smear negative sputa, but the cost of the technique keeps limiting its use to only those smear negative samples with strong clinical suspicion of tuberculosis. Moreover, since this kind of direct detection amplification technology is technically demanding or requires specialized equipment and kits, many laboratories carry out these tests only at well-set time intervals \[e.g. \[[@B11]\]\], delaying diagnosis with several days on average, and as such loosing some of the time gain offered by these direct detection methods. As a final remark, one should keep in mind that direct detection without direct antimicrobial susceptibility testing does not obviate the need for culture after all. ARDRA compared to other culture based genotypic identification techniques ------------------------------------------------------------------------- ARDRA and other gene restriction techniques \[[@B14],[@B15],[@B21],[@B27],[@B28]\] have been developed as a practical short cut to full sequence determination. From this study and others it is clear that the discriminatory power of these RFLP approaches for identification of mycobacteria is almost as high as that of sequencing. The discriminatory power and reliability of a commercially available rRNA-spacer based hybridisation assay (INNO-LiPA Mycobacteria) is high, but again this approach is somewhat more laborious and more expensive than in house PCR-RFLP-based techniques. Restriction digestion of a 439 bp stretch of the *hsp65* gene for identification of mycobacteria was described almost simultaneously with ARDRA and designated PCR-RFLP Analysis (PRA) \[[@B15]\]. Several laboratories have published their experience using this technique \[e.g. \[[@B14],[@B27],[@B28]\]\]. Possible drawbacks of *hsp65* gene restriction analysis are the smaller sized restriction fragments and the higher intraspecific variability, which may make interpretation more difficult. The small size differences have led to the use of polyacrylamide gel electrophoresis \[[@B27]\], which is less practical than agarose gel electrophoresis and the interpretation difficulties in general have led to reconsideration of the *hsp65* gene restriction profiles used thus far \[[@B28]\]. Comparable remarks can be made for PCR-RFLP analysis of the rRNA spacer region \[[@B21]\]. Future developments ------------------- At present, in an effort to have the best of both worlds, we are performing a double PCR, directly on smear positive, decontaminated samples, extracting DNA with the commerially available QiaAmp Tissue kit (Qiagen, Hilden, Germany). One PCR attempts to amplify the full length 16S rRNA gene (1500 bp), which can be used for ARDRA, at an annealing at 55°C, and one PCR amplifies a 123 bp region of the IS6110 at an optimal annealing of 68.6°C \[[@B29]\]. Both PCRs are carried out simultaneously in a T-gradient thermocycler (Biometra, Göttingen, Germany), programmed to have both annealing temperatures in the 96 well block. In case of the presence of the 123 bp fragment on an agarose gel, the identification of *M. tuberculosis* is completed, and can optionally be confirmed with ARDRA. Due to the higher sensitivity of the IS6110 PCR compared to the rDNA PCR, we could amplify the IS6110 fragment of *M. tuberculosis* directly from all smear positive, *M. tuberculosis* culture positive samples thus far. In case only the 16S rDNA fragment of 1500 bp is present, the absence of *M. tuberculosis* can be confirmed by ARDRA which also immediately provides with the identification of the *Mycobacterium* species other than tuberculosis. Thus far, this yielded in all cases an identification as a nontuberculous strain, confirming the absence of *M. tuberculosis.* Also in case both amplifications of the smear positive sample remain negative, this can be interpreted as the absence of *M. tuberculosis.* A positive culture is then awaited to identify the nontuberculous organism with ARDRA. Taxonomical considerations -------------------------- This study also confirmed the robustness of ARDRA based identifications. For example, for the first study \[[@B12]\], we received strains identified by reference laboratories as *M. avium* and *M. xenopi,* using phenotypic methods. During the first study already, some strains that had been sent as *M. avium* could be shown by ARDRA (and subsequent confirmation by another laboratory) to be *M. scrofulaceum.* For *M. xenopi,* ARDRA indicated the presence of several groups. In the meantime mycobacterial taxonomy has been refined and the groups we indicated as *M. xenopi* A and *M. xenopi* B \[[@B12]\] appear to correspond to genuine *M. xenopi,* resp. *M. celatum* (ITG 6147)\[[@B30]\], as becomes apparent when applying the programme RFLP on the 16S rRNA sequence data we obtained. Similarly, strains that we classified as *M. fortuitum* A and B \[[@B12]\], appear to correspond to *M. fortuitum* subsp. *fortuitum,* resp. *M. peregrinum.* Materials and Methods ===================== Strains ------- A collection of well characterized reference strains belonging to different mycobacterial species was used to create an ARDRA profile library. The strains used are listed in Table [1](#T1){ref-type="table"}. The 151 clinical strains used in this evaluation were collected during the period between January 1998 and December 2000 in the routine clinical laboratory of the Ghent University Hospital. Processing and culturing of the samples --------------------------------------- Decontamination of the samples was done by mixing 1 ml of sample with 1 ml of decontamination buffer (3% NaOH/N-acetyl L-cysteine (NALC)). After 15 min of incubation at room temperature the mixture was neutralized by adding 40 ml of 0.067 M phosphate buffer (pH 6.8), followed by centrifugation at 11600 g during 15 min. The supernatant was removed and part of the pellet was used for auramine staining and microscopy. The remaining pellet was suspended in 1 ml of phosphate buffer and used for inoculation of culture media. 100 μl was used for inoculation of a solid medium (Ogawa, Sanofi-Pasteur, Marnes la Coquette, France) and 500 μl for the inoculation of a liquid medium. During this study the automated liquid culture system was changed from the Bactec system (Becton Dickinson, Cockeysville, Md.) to the 3D BacT/Alert system (Organon Teknika, Boxtel, The Netherlands). DNA extraction -------------- Starting from liquid culture, 500 μl of a positive culture was transferred to a 1.5 ml screw cap Eppendorf tube. After centrifugation at 13.000 rpm for 15 minutes, the supernatant was removed and the resulting pellet was resuspended in 50 μl TE buffer (100 mM Tris-HCl -- 10 mM EDTA, pH 8.0). The mixture was heated for 30 minutes at 95°C, followed by a freezing step at -20°C for at least 30 minutes. Starting from solid culture, a loopful of a bacterial colony was suspended in 500 μl of TE buffer. The mixture was heated at 95°C for 30 minutes, followed by a freezing step at -20°C for at least 30 minutes. Prior to PCR, DNA extracts were thawed at 4°C and centrifuged shortly to pellet the debris. ***ARDRA*** for mycobacteria consists of the amplification of the 16S rRNA gene, followed by separate restriction digestion with *Hha*I, *Mbo*I and *Rsa*I The combination of the three obtained fingerprints is designated an ARDRA profile which can be compared with a library of ARDRA profiles, obtained from well-identified mycobacterial strains. In some cases, more discriminatory identification is possible by additional restriction with *Bst*UI. Amplification of the 16S rRNA gene ---------------------------------- The primers used to amplify the full length 16S rRNA gene (approximately 1500 bp) were MBUZ1 (GAC GAA CGC TGG CGG CGT GCT TAA C) and MBUZ2 (CGT CCC AAT CGC CGA TC). These primers are designated to amplify only the 16S rRNA gene for species of the order Actinomycetales. The PCR mixture consisted of 25 μl Qiagen Mastermix (Qiagen, Hilden, Germany), 0.2 μM of each primer, 5 μl of DNA extract, and was adjusted to 50 μl with distilled water. Thermal cycling consisted of an initial denaturation of 5 min at 94°C, followed by three cycles of 1 min at 94°C, 2 min at 55°C and 1 min at 72°C, followed by 30 cycles of 20 sec at 94°C, 1 min at 55°C and 1 min 72°C, with a final extension of 7 min at 72°C, and cooling at 10°C. ***Amplification of 123 bp of the IS6110 region*** was carried out as described \[[@B29]\] after DNA extraction from decontaminated sputum samples using the QiaAmp Tissue kit (QiaGen, Hilden, Germany). Restriction digestion --------------------- The restriction enzymes used were *Hha*I (isoschizomer of CfoI)(Amersham Pharmacia Biotech Benelux, Roosendaal, the Netherlands), *Mbo*I (Fermentas, Vilnius, Lithuania), *Rsa*I (Amersham Pharmacia). When necessary for further discrimination, digestion with *Bst*UI (New England Biolabs, Beverly, Ma.) was carried out. Each 16S rDNA amplicon was divided in three separate tubes in aliquots of 10 μl, to which 10 U of the respective restriction enzymes were added, with 2 μl of the corresponding enzyme buffer (10× concentrated, final concentration 2×) and each restriction digestion mixture was adjusted to 20 μl with distilled water and incubated during 2 hours at 37°C in a heater. Electrophoresis --------------- The DNA restriction fragments were electrophoresed in a 2.5% agarose electrophoresis gel, containing 2% Methaphor (FMC Bioproducts, Rockland, Me.) and 0.5% MP agarose (Roche) in the presence of ethidium bromide (50 ng/ml). The gels were photographed and the fingerprints were compared visually with the overview gels (Figures [1](#F1){ref-type="fig"}, [2](#F2){ref-type="fig"} and [3](#F3){ref-type="fig"}). 16S rDNA sequencing and comparative analysis -------------------------------------------- A fragment of the 16S rRNA gene (corresponding to positions 10-1507 in the *Escherichia coli* numbering system) was sequenced as described previously \[[@B31]\]. Sequencing primers were MB UZ1 (GACGAACGCTGGCGGCGTGCT TAAC, *E. coli* position 27-50), MB UZ2 (CGTCCCAATCGCCGATC, 1493 -- 1476), MBP1 (CCGGCCAACTACGTGCCAGC, 502 -- 522), MBP2 (CTGGAATTCCTGGTGTAGCGG, 673 -- 693), MBP3R (GCATGTCAAACCCAGGTAAGG, 1006 -- 986) and MBP4R (CCACCTTCCTCCGAGTTGACC, 1185 -- 1165) The 16S rDNA sequences obtained in this study are indicated in Table [1](#T1){ref-type="table"}. All steps of the comparative sequence analysis were performed by using the GeneBase software package (Applied Maths, St. Martens Latem, Belgium), as described \[[@B32]\]. First, pairwise alignment using UPGMA was carried out with a gap penalty of 100 %, a unit gap cost of 20 % and an ambiguity cost of 50 % of the mismatch cost. Subsequently, global alignment -- with *N. asteroides* ATCC 49872 (Genbank Z82229) used as the outgroup -- was carried out on the region corresponding to positions 67 through 1444 of the 16S rRNA gene of *E. coli,* with costs as above. Finally, a similarity matrix of the aligned sequences was constructed by global alignment homology calculation and a gap penalty of 20 %. The neighbour-joining method was used to construct the dendrogram based on this similarity matrix. Bootstrap values were calculated. ***Theoretical calculation of restriction patterns*** was done by means of RFLP (Applied Maths), which makes it possible to obtain restriction patterns using sequences in EMBL format, for every restriction enzyme. The programme GelCompar (Applied Maths) was then used to display the obtained fingerprints. ::: {#T3 .table-wrap} ::: {.caption} ###### Overview of the theoretical ARDRA profiles of the species that were newly described since the start of the evaluation period (1998), as obtained after computer aided digestion of the published sequences (see also Figures [4](#F4){ref-type="fig"},[5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"},[7](#F7){ref-type="fig"}). ::: **Species** **Source of isolation** **Reference** **GenBank Accession Numbers** ***Hha*l** ***Mbo*l** ***Rsa*l** ***Bst*UI** ----------------------------------- ----------------------------------- --------------- ------------------------------- ------------ ------------ ------------ ------------- *M. bohemicum* \[[@B34]\] AJ277284, U84502 1\' 4 1 3 *M. botniense* Water \[[@B35]\] AF012756 2 4 3 2 *M. elephantis* Elephant \[[@B36]\] AF010747 3 1 8 3\' *M. goodii* Human wounds \[[@B37]\] Y12872 10 1 \(1\) 2\' Y12872^a^ 10 1 2 2\' *M. heckeshornense* Lung disease \[[@B38]\] AF1174290 1 4 3 4 *M. kubicae* \[[@B39]\] AF133902 10 1 6 1\' *M. murale* Walls of childrens daycare centre \[[@B40]\] Unpublished *M. septicum* Catheter related bacteraemia \[[@B41]\] AF111809 4 1 6 7 *M. tuberculosis* subsp. *caprae* Goats in Spain \[[@B42]\] AJ131120 1 1 1 1 *M. tusciae* \[[@B43]\] AF058299 10 1 6 2\' *M. wolinskyi* Human wounds \[[@B37]\] Y12871, Y12873 10 1 1 2\' a\. After resolving the ambiguity (N) present at *E. coli* position 95 in the Genbank sequence Y12872. An identical base (G) is found at this position in all mycobacteria. ::: Acknowledgements ================ We thank Leen Rigouts for biochemical identification of some of the strains.

Background ========== The Hedgehog (Hh) proteins are evolutionarily conserved signaling molecules that control the normal growth and patterning of diverse animals including Drosophila and humans. In flies Hh is required for multiple developmental processes such as embryonic segment patterning, eye and appendage development (for reviews see \[[@B1],[@B2]\]) In vertebrates, three Hh homologues are expressed in a tissue specific manner and are responsible for the morphogenesis of various organs such as the neural tube and the limbs, and for cartilage and male germinal cell differentiation \[[@B3],[@B4]\]. In mammals, deregulation of the Hh pathway is responsible for cancers, especially basal cell carcinoma and medulloblastoma \[[@B5]-[@B7]\]. In all cases described so far, Hh initiates and/or maintains the transcription of target genes in responsive cells. Among the targets are *patched (ptc)*, which encodes a Hh receptor protein, and genes encoding signaling molecules. In Drosophila, *decapentaplegic (dpp)*, a signal of the TGFβ class, and *wingless (wg)*, a member of the Wnt family, are transcribed in response to Hh. Genes encoding related signaling molecules such as TGFβ, FGF, and Wnt are transcriptionally induced by Hh signals in vertebrates, as are genes for a variety of transcription factors. In Drosophila, Cubitus interruptus (Ci), a zinc finger transcription factor of the vertebrate Gli family, plays a central and complex role in the transcriptional regulation of Hh target genes. Ci acts as either transcriptional activator or repressor in a Hh-dependent manner (for reviews see \[[@B1],[@B8],[@B9]\]). In the absence of Hh signal, most of Ci is cleaved to generate a 75 kD nuclear protein (Ci75-R) consisting of the N-terminus and the zinc finger DNA binding domain of the protein. Ci75-R acts as a repressor on *hh* and *dpp* transcription \[[@B10]-[@B12]\]. Ci cleavage is proteasome dependent and requires Ci phosphorylation by the protein kinase A (PKA), the activity of the kinesin-related protein Costal2 (Cos2) and of Slimb (Slb), a F-box WD-40 protein (other members of this family are known to direct ubiquitin-mediated proteolysis of specific phospho-proteins) \[[@B13]-[@B21]\]. In the absence of Hh, Ci75-R is mainly localized into the nucleus while the uncleaved fraction of Ci (155kD, Ci-155) is retained in the cytoplasm by Cos2 and Suppressor of Fused (Su(fu)), a putative PEST motif-containing protein \[[@B14],[@B19],[@B22]-[@B25]\]. The exclusion of full-length Ci155 from the nucleus is also ensured by its constitutive export \[[@B14],[@B24]\]. Hh signaling inhibits Ci proteolysis, probably by reducing Ci phosphorylation level via the action of a phosphatase \[[@B10],[@B14],[@B18]\]. This results in the accumulation of full-length Ci-155. Hedgehog reception also relieves Ci-155 cytoplasmic retention and allows Ci-155 to be translocated into the nucleus via a basic nuclear localization sequence \[[@B14],[@B18],[@B19],[@B25]\]. Nevertheless, the persistence of its export leads to the accumulation of the vast majority of Ci-155 in the cytoplasm. Last, proper induction of Hh target genes also requires Hedgehog signaling to produce, by an unknown mechanism, an activated form of Ci called Ci-A \[[@B11],[@B23]-[@B26]\]. In the absence of Hh, Su (fu) appears to prevent Ci activation. Upon Hh reception, both the ser-thr protein kinase Fused (Fu) and Cos2 counteract Su(fu) to produce Ci-A \[[@B19],[@B23]-[@B27]\]. The molecular mechanism by which Hh signaling controls Ci remains poorly understood. Both full-length and truncated forms of Ci are detected in the cytoplasm as part of one or more large molecular weight protein complexes \[[@B10],[@B28]-[@B30]\]. The cytoplasmic complex that has been studied also includes Fu and Cos2 and Su(fu) \[[@B28]-[@B31]\]. In the absence of Hh signal, Fu-Cos2-Ci ternary complex binds microtubules, probably through Cos2 \[[@B28]-[@B30]\]. Cultured cell experiments showed that Hh signal triggers the release of the Fu-Cos2-Ci complex from the microtubules \[[@B28],[@B29]\]. Concomitantly Hh signal increases the level of phosphorylation of both Fu and Cos2 \[[@B28],[@B32]\]. Our working hypothesis is that Hh signal could control Ci fate (i.e. cleavage, subcellular localization, and/or activation) by inducing changes in the activity, composition, and/or subcellular localization of the transducing cytoplasmic complex. Since Cos2 is a putative motor protein with microtubule-binding activity, it could play a central role in this process by regulating the association of the complex with the microtubules and perhaps by directing its movement to specific locations within the cell. In order to better understand how Hh transducing complex may function, we focused on the precise relationships among its different members. The physical association of Cos2 with Fu and Ci has been previously demonstrated using gel filtration chromatography and co-immunoprecipitation from embryo and cultured cell extracts \[[@B28]-[@B30]\]. Here we have undertaken the identification and the mapping of the molecular interactions taking place between Cos2, Fu, Su(fu), and Ci using the yeast two-hybrid method and an *in vitro* biochemical assay. Our results show that (i) Cos2, as Su(fu), interacts with both the catalytic and regulatory domains of Fu and with the N-terminal part of Ci; (ii) Ci and Fu associate with Cos2 in the neck domain, located C-terminally to the motor domain. The precise identification of the interaction region of each protein *in vitro* provide new insights into the structure and possible mechanism of action of the complex during Hh signal transduction. Results ======= Cos2 -- Fu interaction ---------------------- *In vitro* interaction assays were done between a GST-Cos2 fusion protein expressed in bacteria and S-methionine-radiolabeled Su(fu) or Fu produced by *in vitro* translation. The Glutathion-S-tranferase was fused to the N-terminus of the full-length Cos2 protein. No interaction was detected between GST-Cos2 and Su(fu) (Figure [1A](#F1){ref-type="fig"}). The same negative result was obtained using GST-Su(fu) and radiolabelled Cos2 (data not shown). {#F1} In contrast, Fu binds specifically to the GST-Cos2 fusion protein but not to the GST protein alone (Figure [1A](#F1){ref-type="fig"}). We tested separately Fu1-305, the catalytic domain of Fu (called Fukin) and Fu306-805, the part of Fu that is not the kinase domain, referred to as the regulatory domain (called Fureg). Both domains have the capacity to specifically interact with the GST-Cos2 fusion protein (Figure [1A](#F1){ref-type="fig"}). Similar interactions were previously observed between Su(fu) and the two domains of Fu (\[[@B31]\] and data not shown). To confirm the Cos2-Fu interaction and to further map the interacting regions we employed the yeast two-hybrid system (Figure [1B](#F1){ref-type="fig"}). Neither full-length Cos2 nor full-length Fu and Fukin could be used in this assay due to interfering activities. Indeed, as we have shown previously, the presence (but not the activity) of the kinase domain of Fu prevents any protein interaction (\[[@B31]\] and data not shown). Using a transcription inhibition assay \[[@B33]\], we found that the full-length Cos2 fusion to the GAL4 DNA binding domain, and some partial Cos2 proteins (such as Cos1-751) that contain the motor domain, do not translocate efficiently to the nucleus (data not shown). The motor domain of Cos2 (AA 172--347) might tether the fusion protein in the cytoplasm, perhaps by binding to microtubules. We therefore analyzed partial Fu and Cos2 proteins. In the yeast two-hybrid assay, Fureg (Fu306-805) interacts specifically with the C-terminal part of Cos2 (Cos538-1201) (Figure [1B](#F1){ref-type="fig"}). Part of this region, Cos538-751, is sufficient to interact with Fureg. Cos538-751 encodes a region of Cos2 C-terminal to the putative motor domain (See also Figure [3](#F3){ref-type="fig"}) and which overlaps 14 of the 36 heptad repeats of a predicted coiled-coil motif of Cos2. In other kinesins the coiled coil repeats mediate homodimerization and belong to a flexible stalk domain involved in the movements of the motor domain along the microtubules \[[@B34]-[@B45]\]. {ref-type="fig"} and [2A](#F2){ref-type="fig"}.](1471-213X-2-4-3){#F2} We also mapped the region of Fu that interacts with Cos2 (Figures [1B](#F1){ref-type="fig"} and [3](#F3){ref-type="fig"}). The C-terminal-most region of Fu (AA 437 to 805) is sufficient to interact with both Cos538-1201 and Cos538-751. This region of Fureg is adjacent but non-overlapping to the part of Fu (AA 306 to 436) that interacts with Su(fu) \[[@B31]\]. No interaction between Fu and the different Cos2 constructs tested was detected with smaller pieces of Fureg (Fureg306-436, Fureg437-581, Fureg306-581, Fureg582-805; data not shown). In summary, Cos2 interacts with both the kinase and regulatory domains of Fu, just as Su(fu) does \[[@B31]\]. The regulatory domain of Fu interacts with a central region of Cos2 located C-terminally to the motor domain. Cos2 -- Ci interaction ---------------------- *In vitro* assays between the GST-Cos2 fusion protein and radiolabeled Ci revealed a direct specific interaction between these proteins (Figure [2A](#F2){ref-type="fig"}). Again, as with Fu, the yeast two-hybrid assay did not allow the detection of an interaction between full-length Cos2 or Cos1-751 and Ci. Nevertheless, interactions consistent with this GST \"pull down\" result were detected in yeast two-hybrid assays using full-length Ci and the region of Cos2 spanning residues 348 to 546 (Cos348-546) (Figure [2B](#F2){ref-type="fig"}). This part of Cos2 is adjacent to the Fureg binding site, in a region located immediately C-terminal to the motor domain at the base of the neck region. This region has been shown to regulate the directionality of the movement of kinesins along microtubules \[[@B34]-[@B45]\]. ![**Summary of the interaction domains** The interactions between Cos2, Ci, Fu, and Fureg described above and in former work \[[@B31]\] are summarized. The interaction between the region 941--1065 of Ci and Cos2 was previously shown by Wang et al. 2000 \[[@B23]\]. The domains of interactions are indicated in lines above or under each protein structure. No interaction could be detected between Cos2 and Sufu (data not shown). See also legends for Figures [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}. M: Cos2 motor domain, HR: Cos2 heptad repeats, DBD:Ci DNA binding domain; NLS: Ci nuclear localisation signal, AD: Ci activation domain.](1471-213X-2-4-2){#F3} To identify the region of Ci that interacts with Cos2, partial deletions of Ci were generated and tested for interaction with Cos2 using a combination of *in vitro* and two-hybrid assays. The first 430 amino acids of Ci are sufficient to interact *in vitro* with GST-Cos2 (Figure [2A](#F2){ref-type="fig"}). In contrast the first 346 residues of Ci, which have been previously shown to bind Su (fu) \[[@B31]\], do not interact in yeast with Cos538-1201 nor with Cos348-546 constructs (data not shown). These results suggest that Cos2 interacts with a part of Ci between amino acids 346 and 430 (Figure [2B](#F2){ref-type="fig"}). Indeed, Ci340-445 is sufficient to interact in yeast with Cos348-546 (Figure [2B](#F2){ref-type="fig"}). Thus, like Su(fu), Cos2 interacts within the N-terminal region of Ci, which is present in both Ci155 and Ci75-R. The region of Cos2 involved in this interaction is located at the base of its neck, which plays an important role in the activity of conventional kinesin. Discussion ========== Cos2, Fu and Su(fu) play multiple and complex roles in Ci control. Cos2 is central to this control both as a negative regulator in the absence of Hh and a positive regulator in the presence of Hh. In the absence of Hh signal, Cos2 action prevents Ci target activation by favoring Ci cleavage and by cytoplasmic retention of full-length Ci \[[@B14],[@B18],[@B23],[@B46]\]. In response to Hh, Cos2 is also required for complete Ci activation in the wing imaginal disc \[[@B19],[@B24]\]. Fu is also involved in positive and negative aspects of Ci regulation. Its kinase activity is required for Ci cytoplasmic release and activation in response to Hh signal. Its C-terminal regulatory domain, although also required for Fu kinase activity, cooperates with Cos2 and Su(fu) to negatively regulate Ci in the absence of Hh signal (see bellow) \[[@B19],[@B22],[@B25]-[@B27],[@B47]\]. Although *Su(fu)* mutants display only a very subtle phenotype, Su(fu) acts negatively in the Hh pathway. A decrease in *Su(fu)* (or cos2) dosage suppresses the effects of the loss of Fu kinase activity. *Su(fu)* loss of function aggravates cos2 phenotypes \[[@B47]\], so Su(fu) activity normally assists Cos2 in its negative regulation of Ci. Su(fu) seems to participate both in the retention of Ci in the cytoplasm and in preventing activation of full-length Ci \[[@B19],[@B26],[@B27]\]. Here, we show that Cos2 interacts directly with both Fu and Ci, as does Su(fu). This confirms and extends previous results, which showed these proteins to co-immunoprecipitate from embryo and cultured cell extracts and to be members of one or several large molecular weight cytosolic complex (es) \[[@B10],[@B28]-[@B30]\]. The mapping of the regions of each protein sufficient for interaction with its partner, *in vitro* and in yeast (Figure [3](#F3){ref-type="fig"}), sheds new light on the respective role of Fu, Su(fu), and Cos2 and on their relationships in the Hh pathway. Our results demonstrate that Cos2 can bind to the N-terminus of Ci, in a region (amino acids 340--445) located between a region interacting with Su(fu) (amino acids 240--346) and the zinc finger domain (amino acids 453--603). Wang *et al*. also used the two-hybrid method to look for interactions between Ci and Cos2. In contrast to our results, they found no interaction of Cos2 with the N terminus of Ci \[[@B23]\]. This discrepancy can have multiple causes as the fact that their constructs lack amino acids 441 to 445 of Ci or differences in the fusion proteins tested (i.e. differences in protein folding, stability, nuclear targeting etc) or in the reporter used\... Nevertheless, they also identified a second region of interaction with Cos2, located in the C-terminus of Ci (Ci 941--1065) \[[@B23]\]. Both the Cos2/Ci941-1065 interaction and the previously reported Su(fu)/Ci240-346 interaction have been shown to be retain Ci in the cytoplasm in the absence of Hh signal \[[@B19],[@B23]\]. Strikingly, neither Su(fu) nor Cos2 interaction regions of Ci map to the amino acids 703 to 850 (Ci-cyt) region identified by Aza-Blanc *et al*. as involved in Ci cytoplasmic tethering \[[@B10]\]. One hypothesis is that the region 703--850 of Ci contains an export signal \[[@B14],[@B24]\]. The interaction of Su(fu) and Cos2 with the N-terminal part of Ci is in agreement with the presence of both full-length Ci-155 and Ci75-R in the large complex(es). Since Fu, Su(fu) and Cos2 can also be found in complex(es) of similar size \[[@B10]\], Cos2 and Su(fu) may control the activity of Ci75-R *in vivo* by a so far unknown mechanism. Several roles can be proposed for the interaction between Cos2 and the Ci N-terminus. The association could reinforce the tethering of Ci in the cytoplasm by Cos2 and Su(fu). An attractive alternative is that the association could underlie the requirement for Cos2 to activate Ci155 in response to Hh. For example, Cos2 and Su(fu) could compete to bind Ci N-terminus and Hh signaling could favor the transient binding of microtubule-unbound Cos2, leading to Su(fu) release. We find that Cos2, like Su(fu), is able to interact with two different domains of Fu: the Fu kinase and regulatory domains. No phosporylation of Cos2 nor Su(fu) by Fu has been reported so far. In S2 cells, it has been reported that Hh induces Cos2 hyperphosphorylation in a Fu independent manner. Nevertheless, the interactions of respectively Su(fu) and Cos2 with the kinase domain of Fu suggest their direct phosphorylation by Fu. Some kinesins are regulated by phosphorylation, and Fu kinase activity could control Cos2 activity \[[@B48],[@B49]\]. The interaction between the regulatory domain of Fu and Cos2 is in agreement with the lack of Fu-Cos2 co-immunoprecipitation in *fu* mutants that are missing part or all of the regulatory domain (*fu*^*class\ II*^ mutant, see below) \[[@B28]\]. Possibly, the *in vivo* interaction of the kinase domain of Fu with Cos2 requires the binding of Cos2 to Fureg, or the interaction could be too transient to be detected in *fu*^*class\ II*^ mutants. The interaction of the regulatory domain of Fu with Su(fu) and Cos2 provides a support for the complex genetic relationships previously described to occur between *Fu*, *Su(fu)*, and *Cos2*. Mutations in the catalytic (class I) and the regulatory (class II) domains of Fu are genetically distinguishable from each other in a *Su(fu)* or *cos2* mutant background, suggesting distinct roles for the two domains of Fu \[[@B22],[@B47],[@B50]\]. Thus, *Su(fu)* ; *fu*^*class\ I*^ double mutants are almost wild type, while *fu*^*class\ II*^; *Su(fu)* double mutants have a cos2-like phenotype associated with the activation of Ci target genes. Similarly, *fu*^*class\ II*^ alleles strongly enhance the effect of a heterozygous *cos2* mutation, though alleles *fu*^*class\ I*^ do not. These observations strongly suggest that the regulatory domain of Fu also plays a negative role in Hh signal transduction, this effect being detected only in association with *Su(fu)* or *cos2* mutations. We propose that, in the absence of Hh signal, Fureg could act as a scaffold to anchor Su(fu) and Cos2 in the complex, favoring their negative effect on Ci. In *fu*^*class\ II*^ mutants, the negative effect of Su(fu) and Cos2 would be lessened due to their weaker anchoring in the complex. This would only be revealed in the presence of a decrease in Su(fu) or Cos2 dosage that would cause a further destabilization of the complex. In conventional kinesins, the C-terminal globular tail domain associates with vesicles. Our results show that the Cos2 C-terminus is not required to bind either to the Ci N-terminus (Ci340-445) or the Fureg domain. Instead, both interactions involve adjacent sequences located between the motor domain and the stalk homodimerisation domain. In kinesin itself, this region is called the \"neck\" domain and is important for efficient motor function. In at least some kinesin molecules the neck is involved in the determination of the direction of kinesin movement along the microtubule, in velocity, in processivity, and in mechanochemical coupling \[[@B34]-[@B45]\]. Thus, although no Cos2 motor activity has been described so far, Fu and/or Ci binding to its central region could induce structural changes and therefore modulate Cos2 motor activity. This hypothesis opens the possibility of a feedback control of Ci on Cos2 activity. Conclusion ========== The present data, together with published results, suggest that the existence of direct multiple interactions between Cos2, Su(fu), Fu, and Ci, each protein interacting with at least two other partners. We propose a model based upon a complex that includes Fu, Cos2, Su(fu), and Ci (Figure [4](#F4){ref-type="fig"}). Changes in the composition, activity and/or subcellular localization of this complex will control Ci fate in the absence and in the presence of Hh. In the absence of Hh signal, Su(fu) and Cos2 prevent Ci activation, Ci processing into the repressor form is favored and full-length Ci is exported from the nucleus. Upon Hh reception, Fu is activated, opposing Su(fu) to allow Ci to become an activator. Simultaneously the negative effect of Cos2 is alleviated, perhaps by release from the microtubules and/or a change in subcellular location. This lowers the rate of proteolysis of Ci into its repressor form and triggers nuclear import of full-length Ci. Furthermore, Cos2 and Fu promote the conversion of Ci 155 into an activator form, ultimately resulting in the transcription of Hh target genes. {#F4} So far, each protein of the complex can serve as a direct link between two other proteins, and could in this way form a scaffold for the complex. It is important to note that the interactions observed *in vitro* may not define a singular protein complex *in vivo*. A subset of the proteins could associate at any given time, or more than one type of complex may exist simultaneously. Additional interactions, may of course contribute to the stability, organization, or activity of the complex(es). Structural data and new functional tests will be required to learn more about the multiple interactions among the components, and their multiple roles in the control of Hh signaling. Materials and Methods ===================== Strains ------- RFY231: *Matα ura3-1 his3 trp1 Δ :: hisG 3lexAop-LEU2 : :leu2* was built by R. Finley (unpublished data). EGY189 *Matα ura3-1 his3 trpl-1 1lexAop-LEU2: :leu2* has a less sensitive *leu2* reporter gene than RFY231 due to a lower number of copies of *lexAop*\[[@B51]\]), RFY206: *Mata his3Δ200 leu2-3 lys2 Δ 201 ura3-52 trp1ΔhisG.* Yeast cultures and transformations were performed as in \[[@B31]\]. Plasmids -------- The two-hybrid pEG and pJG expression vectors were derived from pEG202 (translational fusion with the DNA binding domain of lexA) and pJG4-5 (translational fusion with the activation domain B42) \[[@B33],[@B52]\]. Two numbers added after the protein name indicate the first and last residues of truncated proteins. pJGCi (full-length), pJGCi1-346 and pJGFu306-805 (pJGFureg), were described in \[[@B31]\]. pEGCos2 (full-length), pEGCos1-751, pEGCos348-546, pEGCos538-1201, pEGCos538-751, pJGCi340-445, pJGFu437-805 were built by in frame cloning of restriction fragments or PCR products. pSH18-34 carries the lacZ reporter gene \[[@B52]\]. The GST-Cos2 translational fusion was made by introducing the entire Cos2 coding sequence in frame with the glutathione S-transferase (GST) gene in pGEX-4T3. All PCR products and junctions were sequenced. GST fusion protein binding assays --------------------------------- GST-Cos2 and GST protein production and purification, *in vitro* coupled transcription-translation, and *in vitro* interaction assays were performed as described in \[[@B31]\]. In all the figures, the amount of radiolabelled protein used in each \"input\" lane equals one-third the amount of protein used in the assay and equivalence in the amount of GST and GST-Cos2 was checked by Coomassie blue staining. All assays were done at least twice. Two hybrid assays ----------------- RFY231-pSH18-34 or EGY189-pSH18-34 were transformed with pEG (lexA fusion) constructs and subsequently mated with RFY206 transformants containing the pJG constructs (B42 fusion). Geranyl-geranyl transferase II β (GGTIIβ) and Rab3 are two proteins known to interact together \[[@B53]\]. They are respectively used as negative controls in tests of the pEGCos2 constructs and the pJGCi or pJGFu constructs. Diploids were selected on SCglu-UHW. For each diploid strain, 3 μl of cell suspension were spotted on SCgal-UHW (for β-galactosidase assays) or SCgal-UHWL solid media (for leucine auxotrophy assays). Growth on SCgal-UHWL was observed after 24 hours of incubation at 30°C. β-galactosidase assays were performed by the overlay method as described in \[[@B54]\]. All assays were done at least three times. Acknowledgements ================ We thank Russ Finley for plasmids and advice, M. Barre, F. Besse, F. Dussillol, S. Fouix, S. Szuplewski, D. Busson, H. Tricoire and the other members of both laboratories for their help in this work. We are deeply grateful to C. Lamour-Isnard for all her support and numerous stimulating discussions. This work was supported by grants from the ARC (N° 5523), by NSF-CNRS funds, and by the Howard Hughes Medical Institute.

Background ========== Acute Intermittent Porphyria (AIP) is an autosomal dominant inherited disorder because of a partial defect in Porphobilinogen Deaminase (PBGD), the third enzyme in the heme biosynthetic pathway. As a result, the porphyrins precursors, δ-aminolevulinic acid (ALA) and Porphobilinogen (PBG) can accumulate \[[@B1]\]. Furthermore, the free heme pool would decrease and adversely affect P-450 and several important antioxidant enzymes, leading to impaired hepatic detoxification reactions \[[@B2]\]. Clinically, the disease is characterized by occasional acute attacks of abdominal pain and various neuropsychiatric symptoms. Several therapeutic drugs, anaesthetics and infections are considered the most frequent precipitating factors of acute attacks in the Argentinean population \[[@B3]\]. Specific treatments for acute attacks include high glucose intake \[[@B4]\], hemin \[[@B5]\] and folic acid in combination with glucose \[[@B6]\]. In Northern Sweden, Lithner and Wetterberg \[[@B7]\] and Hardell *et al.*\[[@B8]\] have found a clear association between AIP and hepatocellular carcinoma, also confirmed by other authors \[[@B9]\]. Moreover, ALA has been demonstrated to produce reactive oxygen species through metal-catalyzed aerobic oxidation and to cause oxidative damage to proteins, lipids and subcellular structures. Exposure of plasmid pBR322 DNA to ALA produced DNA single-strand breaks \[[@B10]\]. In this study, we have examined the effect of ALA on human hepatocarcinoma HEP G2 and HEP 3B cell lines viability. Tumor cells were exposed to different ALA concentrations and ALA-S activity repressed conditions, produced by hemin or glucose addition to the medium. DNA fragmentation was also analyzed. Northern blot analysis for p53 and Western blot analysis for CDK2 and CDK4 were evaluated to determine cells sensitivity to ALA. Materials and methods ===================== Chemicals --------- Hemin, ALA, 3-\[4,5-dimethylthiazol-2-yl\]-2,5-diphenyltetrazolium bromide and L-glutamine were obtained from Sigma Chemical Co. (St. Louis, MO). D-Glucose, PBS, Dulbeco\'s Modified Eagles Medium (DMEM) and antibiotics were obtained from GIBCO BRL. Cell lines and cell culture --------------------------- HEP G2 and HEP 3B cell lines (American Type Culture Collection, Rockville, MD) were derived from a human hepatocellular carcinoma \[[@B11]-[@B13]\]. HEP G2 contained wild-type p53 \[[@B14]\] and HEP 3B cells contained deleted p53. Cell lines were grown in total darkness, at 37°C in 5% CO~2~ atmosphere in Dulbecco\'s Modified Eagle Medium supplemented with 10% fetal bovine serum, 10,000 U penicillin, 10,000 μg streptomycin and 2 mM L-glutamine. MTT assay --------- After exposure to drugs, the viability was determined by the MTT assay following published procedures \[[@B15]\]. Briefly, 40,000--50,000 cells in 500 μl of medium were plated onto each well of 24-well tissue culture plate. After 24 h of culture, ALA (0.5--5 mM), Hemin (8--24 μg/ml) or D-glucose (2--3 mg/ml) dissolved in PBS immediately prior to use, were added to the medium. Cells were cultured for 24 h. After exposure, the medium was aspirated and replaced with 300 μl of fresh drug-free medium containing 0.5 mg/ml (3-\[4,5-dimethylthiazol-2-yl\]-2,5-diphenyltetra-zolium bromide) MTT. The cells were incubated for an additional 4 h. This medium containing MTT was removed by inverting the plate and 200 μl of dimethylsulphoxide (DMSO) were then added to each well. After 10 minutes of shaking, absorbance in each well was read in a microplate reader at 570 nm. All experiments included untreated control cells, they were repeated three times and each concentration was tested in sextuplicate. Viability was expressed in terms of population growth, as percent of an untreated control population with standard error of the mean (SEM). Analysis of DNA fragmentation ----------------------------- Treated cells were grown in 75 cm^3^ flasks. Both attached and unattached cells were harvested, washed with PBS, resuspended in TNE solution (10 mM Tris-HCI, pH 7.6; 140 mM sodium chloride; 1 mM EDTA), and lysed at 37°C in 2 ml of extraction buffer (10 mM Tris-HCI, pH8.0; 100 mM EDTA, pH 8.0; 20 μg/ml pancreatic RNase; 0.5% SDS). After 2 h, proteinase K was added at a final concentration of 100 μg/ml and the mixture was incubated for another 3 h at 50°C. The DNA was extracted twice with equal volumes of phenol and once with chloroform. The DNA was then precipitated with 0.1 vol of sodium acetate (pH 4.8) and 2.5 vol of ethanol at -20°C overnight and pelleted at 13,000 g for 30 min. Samples were electrophoresed in 2% (w/v) agarose gel and DNA was visualized by ethidium bromide staining. Northern blot analysis ---------------------- Total RNA (40 μg), from exponentially growing cells treated with different concentrations of ALA (0.5--5 mM) during 24 h, was prepared by Chomsinsky and Sacchi \[[@B16]\] method, size-fractionated on 1% agarose/5% formaldehyde gels, transferred onto nylon membrane Hybond-N+ (Amersham Pharmacia Biotech), and ultraviolet cross-linked. The blots were then prehybridized and hybridized with pTRI-p53-Human or pTRI-GAPDH-Mouse (Ambion) to 65°C with Ultrahyb solution (Ambion) and washed as manufacturer described. Blots were exposed to AGFA films at -70°C and developed. cDNA probes were labeled with \[α-^32^P\]ATP (3,000 Ci/mmol) using Kleanow (Biolabs) and random primers, for 2 hours at 37°C. Intensity of bands were analysed with ImageMaster (Amersham Pharmacia Biotech). Western blot analysis --------------------- Exponentially growing cells (5 × 10^6^/25 ml) were treated with different concentrations of ALA (0.5--5 mM). After 24 h of incubation, cells were washed with PBS and lysed in 500 μl of RIPA buffer (50 mM Tris pH 7.4; 150 mM NaCI; 20 mM EDTA pH 8; 0.1% SDS). After sonication (3 × 15 s) and 30 min incubation with PMSF (10 mg/ml) at 4°C, cell lysates were centrifuged at 10,000 g for 10 min at 4°C. Protein content of supernatants was determined by Bradford protein assay. Cell lysates containing equal amounts of protein (50 μg) were resolved on 12 % SDS-polyacrylamide gel electrophoresis (PAGE). Proteins were transferred onto nitrocellulose membranes Hybond ECL (Amersham Pharmacia Biotech) at 300 mA for 1 hour at 4°C using Towbin buffer (25 mM Tris; 192 mM glycine; 3.5 mM SDS; 10% Methanol; pH 8.3). Membranes were blocked by incubation in TBS (20 mM Tris; 137 mM NaCl; 0.05 % Tween 20; pH 7.6) containing 5% dry low fat milk for 1 hour at room temperature. Blots were probed with either anti-CDK2 (1:300), anti-CDK4 (1:300) or anti-actin (1:100) goat polyclonal antibody (Santa Cruz Biotech, Santa Cruz, CA). Immune complexes were detected using donkey anti goat secondary antibody (1:1,500) (Santa Cruz Biotech, Santa Cruz, CA) and were visualized using ECL reagents (Amersham Pharmacia). Intensity of bands was analyzed with Image Master (Amersham Pharmacia Biotech). Results ======= Cytotoxic effects of ALA in Human Cancer Cells ---------------------------------------------- Human hepatocarcinoma cells were treated with various concentrations (0.5--5 mM) of ALA and the viability of cells was determined (Figure [1](#F1){ref-type="fig"}). ALA was toxic to both hepatocarcinoma cell lines. In HEP G2 cells, cytotoxicity occurred after exposure to 0.5 mM ALA for 24 h (growth inhibition: 30%) and the highest growth inhibition (70%) was observed with the highest ALA concentration (5 mM). HEP 3B cells were more resistant to ALA than HEP G2 cells, exhibiting only 12% growth inhibition after exposure to 0.5 mM ALA for 24 h and a 40% growth inhibition with 5 mM ALA. The strange shape of the viability curves obtained in both cases, with a restoration of control viability with 2 mM ALA, cannot be explained. {#F1} Hemin and D-Glucose effect -------------------------- Simultaneous treatment of the HEP G2 cells with ALA (0.5 mM) and non-toxic concentrations of hemin (8--24 μg/ml) for 24 h resulted in a significant decrease in sensitivity to ALA (Figure [2](#F2){ref-type="fig"}), reverting its cytotoxic effect. It is worth noting that treatment of HEP G2 cells with hemin alone stimulated cell survival (20% over control). However, the same treatment of HEP 3B (ALA and hemin) increased ALA toxicity. {#F2} Alike hemin, simultaneous treatment of the HEP G2 cells with ALA (0.5 mM) and D-glucose (2--3 mg/ml) for 24 h resulted in about two-fold decrease in sensitivity to ALA (Figure [3](#F3){ref-type="fig"}). And again the same treatment of HEP 3B (ALA and D-glucose) enhanced ALA cytotoxicity. {#F3} DNA fragmentation after ALA treatment ------------------------------------- As shown in Figure [4a](#F4){ref-type="fig"}, after 24 h exposure of HEP G2 cells to ALA (1--5 mM) both attached and unattached cells were harvested. In each case, nucleosomal DNA ladders, typical of apoptosis, were visible on agarose gel after staining with ethidium bromide. When HEP 3B cells were treated under the same conditions, nucleosomal DNA ladders were much less visualized (Figure [4b](#F4){ref-type="fig"}). Apoptosis was also corroborated morphologically by phase contrast microscopy in both cell lines. The decrease in viability was due to cytotoxicity as observed under phase contrast microscopy. ALA caused hepatoma cells first to shrink and then gradually to become dettached, starting about 14 hours after treatment as observed by phase contrast microscopy (data not shown). {#F4} Modulation of p53 and cell cycle in HEP G2 cells ------------------------------------------------ To elucidate the mechanisms of ALA causing apoptosis selectively, we have examined the effect of ALA on p53 mRNA levels. No any change in mRNA p53 levels for 24 h was observed in HEP G2 cells (Figure [5](#F5){ref-type="fig"}). Only at concentrations as high as 5 mM ALA, p53 mRNA expression was 25% diminished. {#F5} When CDK2 and CDK4 protein levels were analyzed in HEP G2 cells after exposure to ALA for 24 h (Figure [6](#F6){ref-type="fig"}), both kinases decreased with increased ALA concentrations, up to 2 mM. At concentrations above 2 mM protein content increased reaching basal levels at 5 mM (Figure [6](#F6){ref-type="fig"}). {#F6} Discussion ========== The heme precursor ALA accumulates under pathological conditions, such as AIP, tyrosinosis and lead poisoning, diseases often linked with increased liver cancer incidence \[[@B17]\]. The reactive oxygen species generated upon ALA metal-catalyzed oxidation, promote the formation of several radical-induced base degradation products in isolated and cellular DNA \[[@B17]\]. ALA is the precursor of Protoporphyrin IX (Proto IX). Exogenous ALA bypasses the feedback control and may induce Proto IX accumulation. Since heme-containing enzymes are essential for energy metabolism, every nucleated cell should have at least a minimal capacity to synthesize Proto IX and this is the rational for the use of Photodynamic Therapy \[[@B18]\]. ALA proved to be cytotoxic in neuroblastoma cells. This cytotoxic effect of ALA could be diminished by heme, presumably through modulation of ALA synthesis \[[@B19]\]. Here we have demonstrated the cytotoxic effect of ALA on two human hepatocellular carcinoma cell lines HEP G2 and HEP 3B, which was greater on the former. ALA effect would be attributed to its accumulation in HEP G2 cells, as it was shown by its reduction after simultaneous exposure with hemin or glucose, well-known repressors of ALA-S activity. There is evidence that following the exposure to external ALA, there is induction of endogenous ALA synthesis, and that ALA uptake is higher in more rapidly proliferating cells \[[@B20]\]. The strange shape of the viability curves is difficult to explain within the context of these experiments and could probably be due to different ALA and/or PPIX accumulation, which will be measured in future assays. In HEP G2 cells, protection by hemin would be due to induction of the HO-1 gene, and increased formation of bilirubine, through heme degradation with its known free radicals scavenger effect \[[@B21],[@B22]\]. Glucose protection could be attributed to overcoming reduced glucose uptake provoked by ALA. ALA was reported to affect cultured chick embryo neuronal and glial cells by decreasing glucose uptake \[[@B23]\]. In HEP 3B cells, treatment with ALA and hemin or glucose produced increased cytotoxicity, indicating that the mechanism of ALA action involved would certainly be different. Apoptosis is a physiological event occurring in response to multiple stimuli including growth factor withdrawal, radiation therapy and presence of chemotherapeutic agents. Several mechanisms regulate the apoptotic process, such as induction of a p53-dependent-pathway after DNA-damaging agents, modulation by the bcl2 family of proteins, and activation of some effectors including the interleukin-converting enzyme (ICE) family of proteases and endonucleases \[[@B24],[@B25]\]. Interestingly, ALA produced apoptosis visualized by DNA ladders, but its action did not appear to correlate with the p53 status; while HEP 3B p53 null cells are relatively resistant to ALA-induced apoptosis, HEP G2 p53 wild type cells are very sensitive to ALA-induced apoptosis. However, the treatment did not alter mRNA p53 levels as revealed by Northern blot analysis suggesting that apoptosis would be occurring through other mechanism, p53 independent. CDKs have been implicated as modulators of apoptosis \[[@B26]\]. Some important influences that may have an impact on susceptibility to apoptosis include altered generation of reactive oxygen or nitric oxide intermediates or propensity for mitochondrial damage or protease activation. The mammalian cell cycle is controlled by sequential activation and inactivation of the CDKs family, which are activated by cyclins and inhibited by CKIs. The regulators that operate in the different phases of the cell cycle are various. In our experiments, CDK2 and CDK4 protein levels decreased with ALA treatment preventing cells proliferation. Considering that both kinases are involved in cell cycle progression, diminution of their levels would induce G1 arrest and lead to the onset of the apoptotic process. When ALA was added to the medium at supraphysiological concentrations (higher than 3 mM), the rate of death cells were so high that the measured kinases levels would only correspond to a minimal proportion of viable cells. Perhaps, these cells represent a part of the population in S phase and therefore, with high levels of the kinases necessary to duplicate. These results prompted us to speculate that ALA toxic effect would be effective only in those cells which would be able to enter in arrest and would not exert any effect on those which are cycling. Further studies are in progress to further support our hypothesis. Disruption of the regulatory system controlling G1 phase progression is a common event in human hepatocarcinogenesis. Overcoming alterations that have occurred in the G1 phase regulatory machinery may provide a novel weapon to treat hepatocellular carcinoma \[\[[@B27]\] and references there in\]. In summary, ALA has a markedly cytotoxic effect on hepatocarcinoma cell lines at physiological levels, which could be attributed to an enhancement of apoptosis, independent of p53 but due to inhibition of cell cycle regulators CDKs. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== Mefloquine was developed by the US Army and introduced for the treatment of malaria in the late 1970s. \[[@B1]\] Mefloquine was first used for prophylaxis in 1985, and since then approximately 14.5 million people have been prescribed the drug for malaria prevention, versus 1.6 million for treatment. \[[@B2]\] In the first decade of mefloquine\'s use, the reported adverse effects were mainly gastrointestinal.\[[@B3]\] In the late 1980s it became clear that mefloquine could cause neuropsychiatric adverse effects.\[[@B4]\] The first randomised controlled trial of mefloquine prophylaxis in heterogeneous, non-immune Western travellers was published in 2001 and found that one-third of all mefloquine users reported neuropsychiatric adverse effects and 6% of all users reported at least one severe adverse event (defined as requiring medical advice).\[[@B5]\] On the evidence from a case series published in 1992 by the manufacturer, Hoffmann-La Roche, the World Health Organisation (WHO) recommends that travellers with a personal or family history of seizures or manic-depressive illness should not take mefloquine prophylaxis.\[[@B6],[@B7]\] However the Centers for Disease Control and their Canadian equivalent, CATMAT, do not recognise this as a valid contraindication to taking mefloquine.\[[@B8]\] A recent analysis of spontaneous reports held on the Dutch national pharmacovigilance database suggested that there is a mefloquine syndrome consisting of excessive sweating accompanied by malaise, nausea, diarrhoea, agitation, concentration problems and nightmares.\[[@B9]\] The aetiology of the adverse effects associated with mefloquine use remains obscure. Ten cohort studies in tourists found that women generally experienced worse side effects from mefloquine than men; \[[@B6],[@B10]-[@B18]\] an eleventh did not find this effect.\[[@B19]\] In randomised controlled trials, children have tolerated mefloquine well. \[[@B20],[@B21]\] Surprisingly, one cohort study found that some older adult travellers tolerate mefloquine better than younger adults. \[[@B22]\] It has also been reported that Asian patients tolerate mefloquine better than Caucasians and Africans. \[[@B23],[@B24]\] Despite early concerns, mefloquine appears to be safe in pregnancy.\[[@B25]\] Although the adverse effects of mefloquine are common, and often serious and long lasting, and the drug has been widely used for over 20 years, no real attempts have so far been made to investigate and explain these effects. A systematic review of mefloquine prophylaxis, performed within the Cochrane Collaboration, now tabulates 516 case reports published in 136 papers between 1976--2000, describing adverse effects from mefloquine at prophylactic and therapeutic dosages.\[[@B26]\] We retrieved and critically reviewed all the original papers listed in the Cochrane review to clarify the phenomenology of the adverse effects associated with mefloquine, and to look for clues to possibly safer use of the drug. Analysis of case reports ======================== Common features of the case reports ----------------------------------- Of the 516 published case reports of mefloquine adverse effects recorded in the mefloquine systematic review, 328 of the reports related to prophylactic mefloquine use, and 188 to treatment use. 324 of the 516 cases (63%) related to tourists or business travellers.\[[@B26]\] One-third of the reports were in languages other than English -- mainly French, German and Danish. The search strategy for finding the case reports is described in the review.\[[@B26]\] An annotated bibliography of the 516 published reports can be found at <>. 26% of the case reports recorded three or fewer individual patient parameters (for example, the patient\'s age, sex and mefloquine dose taken), and many of these reports were little more than a short description of an unexpected adverse event, set in the context of a larger study. 52% of the reports however contained some discussion of the causality of the symptoms attributed to the drug, and 11% proposed a mechanism by which these symptoms might be occurring. 56% of the 516 case reports we reviewed described one or more symptoms consistent with a *transient, anicteric chemical hepatitis* (eg, malaise, fever, anorexia, headache, abdominal pain, nausea, diarrhoea, concentration difficulties). Of the remaining reports, many were consistent with a *post-hepatic syndrome,* although in some cases the disorders described could have been due to conditions such as anxiety, depression, chronic fatigue or jet lag. 15% of all the case reports described symptomatology suggesting acutely *disturbed thyroid function* (eg, anorexia, fatigue, tremor, palpitations, nervousness, increased sweating, mood and/or sleep disturbance, memory and concentration disorders, emotional lability, altered bowel habit, depression). Table [1](#T1){ref-type="table"} categorises the 516 published case reports according to whether the clinical features were \'very likely\', \'plausibly\' or probably \'unrelated\' to liver or thyroid pathology. We have based this classification on standard lists of the common symptoms and signs of liver and thyroid disease. \[[@B27],[@B28]\] The direction of thyroid symptoms was mainly towards hyperthyroidism, though some patients exhibited signs of both an over- and an under-active thyroid (for example, tachycardia alternating with bradycardia). ::: {#T1 .table-wrap} ::: {.caption} ###### Clinical features of published case reports of adverse effects from mefloquine, and their possible relation to liver and thyroid pathology ::: **Prophylaxis case reports (N = 328)** **Treatment case reports (N = 188)** ----------------------------------------------------------------- ---------------------------------------- -------------------------------------- Number (%) of reports very likely **liver** related **75/328***(23)* **85/188***(45)* Number (%) of reports plausibly but debatably **liver** related **102/328***(31)* **29/188***(15)* Number (%) of reports plausibly **thyroid** related **51/328***(16)* **26/188***(14)* Probably unrelated to liver or thyroid **100/328***(30)* **48/188***(26)* Median dose (*range*) of mefloquine taken (mg) **750***(250--8750)* **1250***(150--5250)* Median duration (*range*) of adverse effects (days) **16***(1--550)* **4***(1--300)* ::: The table also shows the median duration of symptoms reported by \'prophylactic\' and \'treatment\' mefloquine users, and the median dose taken within each category. The median duration of adverse effects of patients who took mefloquine as treatment appears to have been shorter than that of those who took the drug as prophylaxis (4 days versus 16 days), even though the median dose of mefloquine taken was higher in the treatment group. We believe that one explanation for this unexpected finding may lie in the fact that the treatment users of mefloquine were more likely to have taken the drug as monotherapy, whereas the prophylaxis users more commonly took one or more co-medications, as well as alcohol. We discuss the possible significance of this later in this paper. Mefloquine and the liver ------------------------ Mefloquine is an aryl amino alcohol which accumulates in both the liver and the lungs, and is subject to enterohepatic circulation. \[[@B29]\] It has recently been found to cause acute hepatitis.\[[@B30]\] Mefloquine does not appear to cause florid signs of liver disease. However, transient subclinical disturbances of liver function are a common feature of many drugs metabolised in the liver, and this may explain the frequent finding of transaminase changes in safety studies of new drugs; these biochemical findings are usually dismissed as meaningless noise, but they may in fact be sensitive or oversensitive markers of vulnerability, of low specificity. That mefloquine induces liver enzymes is well documented. Jaspers et al reported significantly raised transaminases in Dutch marines who took mefloquine during 3 months in Cambodia, and who were not drinking alcohol at the time. \[[@B31]\] Takeshima found that of a cohort of healthy Japanese soldiers who took prophylactic mefloquine for 36 weeks without drinking alcohol, one-quarter developed symptoms compatible with liver pathology and four showed disturbed liver function. \[[@B32]\] Reisinger et al observed the same phenomenon in a cohort of short-stay European travellers to Africa, but it is not clear whether alcohol could have contributed to this effect.\[[@B33]\] One of the travellers, who was concurrently taking a liver-damaging agent, sulfadoxine,\[[@B34]\] showed gross morphological changes in his liver which were attributed to his use of prophylactic mefloquine. Liver biopsy showed intralobular cellular infiltrates consisting of macrophages and eosinophils as well as sporadic eosinophilic cell necroses; virology was negative. \[[@B33]\] Grieco et al described a 46-year-old woman who drank wine daily while taking mefloquine, and who became nervous and depressed, with nausea, vomiting and diarrhoea. She was dehydrated and in severe liver failure, with negative virology. Liver biopsy showed diffuse macrovesical hepatic steatosis. \[[@B35]\] \'Heavy sun exposure\' is noted in a case report of a 60-year-old Frenchman who reacted acutely to his second mefloquine tablet; it is likely that this sun exposure would have caused dehydration.\[[@B5]\] A 20-year old French traveller, concurrently taking an oral contraceptive, had epileptic seizures in her sixth week of mefloquine prophylaxis, directly after \'severe exertion\'.\[[@B5]\] It seems likely that in some mefloquine users dehydration will impose an added burden on the liver, and that this could contribute to a severe reaction to the drug. Many long-haul travellers using mefloquine are mildly dehydrated from in-flight alcohol and air conditioning, followed by hot and dry conditions, and more alcohol consumption, at their holiday or business destination. Of the 516 case reports we reviewed, eleven cited alcohol as possibly contributing to the adverse drug effects described. Wittes et al reported a remarkable challenge-rechallenge experiment where a healthy male geologist took both his third and his fourth weekly mefloquine tablet together with half a bottle of whisky, and on both occasions experienced acute paranoid delusions, depression and suicidal ideation; a fellow geologist who shared the same whisky bottle (and who was taking no antimalaria medication) experienced no such effects.\[[@B36]\] Vuurman et al, sponsored by Hoffmann-La Roche, tested in healthy volunteers whether or not alcohol might interact adversely with mefloquine.\[[@B37]\] They found psychomotor performance unimpaired, but their study design had important limitations. Only 20 participants took mefloquine and of these, two women dropped out due to adverse events (one with nausea, vomiting and dizziness, the other with malaise, fever and headache). The study protocol forbade \'strenuous physical activity\' and any prescribed medications. Alcohol was given under strict laboratory conditions 24 h after mefloquine ingestion, and then in small and interrupted doses, such that the blood alcohol concentration in any participant never exceeded 0.50 mg ml^-1^. The authors admit that their study did not address \'the question of what might happen should (mefloquine users) consume intoxicating amounts of alcohol. \[[@B37]\] Their findings can thus not be generalised to the broad population of tourists and business travellers. Approximately half of the case reports listed in the Cochrane review note some co-medication taken along with mefloquine.\[[@B26]\] Other quinoline derivatives (chiefly chloroquine and quinine) are the commonest co-medications mentioned in the case reports. After antimalaria drugs, an oral contraceptive (noted in 8 reports) is the next most commonly reported co-medication, followed by sodium valproate (7) and diazepam (4). All these drugs can cause liver damage.\[[@B34]\] Diazepam is also a thyroid hormone antagonist, and we discuss below the possible significance of this. Meszaros et al reported a male traveller who in addition to mefloquine took thioridazine, amitriptyline and fluphenazine (all capable of damaging the liver), and whose mefloquine-associated neuropsychiatric symptoms persisted for over a year. \[[@B38]\] Gullahorn et al reported a series of patients who experienced delirium on emerging from anaesthesia, possibly because in addition to mefloquine they had received isoflurane, an anaesthetic known to cause hepatocellular necrosis. \[[@B34],[@B39]\] One report describes an acute reaction in a man who took one mefloquine tablet each week together with two aspirin tablets. One hour after taking his fifth mefloquine tablet he experienced acute amnesia lasting approximately one hour.\[[@B40]\] Aspirin can cause hepatocellular necrosis,\[[@B34]\] and in addition can aggravate acute thyroid disturbance (discussed below) by competing with thyroid hormones for sites on binding proteins.\[[@B53]\] Mefloquine and the thyroid -------------------------- The preclinical studies of mefloquine by the US Army involved close monitoring in animal models and human volunteers of several organ systems, but not the thyroid. \[[@B1]\] The effect of mefloquine on thyroid function appears not to have been investigated in any phase III or phase IV study. Thyroid function has not been tested routinely in the diagnosis or management of patients suffering from mefloquine-related adverse effects. Thyroid disease, or some possible interference with thyroid activity, is reported in only three of the 516 case reports in the Cochrane review.\[[@B26]\] Bem et al described a 31-year old German woman with a \'thyroid condition\', who was also taking \'tranquillisers\' (unspecified) and alcohol, and who had an acute exacerbation of her schizophrenia after a single tablet of mefloquine; her symptoms persisted for 4 weeks.\[[@B5]\]Conget et al reported a 30-year old previously healthy woman who experienced abdominal pain, palpitations, instability, insomnia and a fine distal tremor in her second week of mefloquine prophylaxis. A thyroid function screen showed a raised serum thyroglobulin level (54 μg/ml, normal range 18.7 to 27.1); this returned to normal within a month of her stopping mefloquine.\[[@B41]\] Bauer et al reported acute psychotic reactions in a healthy US Peace Corps worker who took prophylactic mefloquine concurrently with diiodohydroxyquinoline for a presumed parasitic infection.\[[@B42]\] Presentation of the hypothesis ============================== The mefloquine syndrome ----------------------- The published literature describes a mefloquine syndrome that presents in a variety of ways including headache, gastrointestinal disturbances, nervousness, fatigue, disorders of sleep, mood, memory and concentration, and occasionally frank psychosis. Young western women are particularly vulnerable to mefloquine\'s adverse effects. Certain groups however (children, older adult travellers and Asian patients) tolerate mefloquine well. Hypothesis ---------- The phenomenology of mefloquine\'s adverse effect profile, together with incidental details in some of the published case reports (references to alcohol, and to known hepatotoxic co-medications such as the oral contraceptive pill) suggest that for many mefloquine users adverse drug effects may be the result of primary hepatocellular injury, caused by the drug in association with one or more concurrent liver insults. Further, it seems that in some of these symptomatic users of mefloquine a transient thyroid disturbance may appear as well, either as an endocrine disorder secondary to the primary liver damage, or as an independent pathological process. We therefore postulate that many of the adverse effects of mefloquine are a post-hepatic syndrome caused by primary liver damage. In some individuals we believe that symptomatic thyroid disturbance occurs, either as an independent process, or as a secondary consequence of the initial hepatocellular injury. Previous liver or thyroid disease, and concurrent insults to the liver (such as from alcohol, dehydration, an oral contraceptive pill, recreational drugs, and other drugs that can damage the liver) may be related to the development of many severe or prolonged adverse reactions to mefloquine. Co-medications that are thyroid hormone antagonists may also be risk factors. Relevant reports with other quinoline derivatives ------------------------------------------------- Mefloquine is a synthetic quinoline; other quinolines include primaquine, amodiaquine and chloroquine.\[[@B43]\] High doses of primaquine in rhesus monkeys have caused acute fatal liver damage.\[[@B44]\] Amodiaquine is still used to treat malaria, but was withdrawn from general use for malaria prophylaxis in 1986 after it was found to cause liver damage and hepatitis, mostly anicteric. \[[@B45],[@B50]\] Some of these reports mention co-medications known to damage liver cells (phenylbutazone, oral contraceptive, alcohol). \[[@B45],[@B46]\] Amodiaquine has not been linked to disturbed thyroid function. The possibility of a three-way interaction has already been suggested between a quinoline derivative (chloroquine) and the liver and the thyroid. Munera et al described a woman with hypothyroidism, well stabilised on thyroxine sodium 125 μg daily, who took prophylactic chloroquine and proguanil daily for 2 months for a vacation in Africa.\[[@B51]\] At four weeks her thyroid stimulating hormone (TSH) concentration was found to be very high (44.8 mU/1, normal range 0.35--6.0), but it returned to normal within a week of her stopping the drugs. Re-challenge with chloroquine and proguanil a year later again resulted in raised levels of TSH, a lowered concentration of free triiodothyronine (T~3~), and normal free thyroxine (T~4~) concentration. Liver function was not tested, but the authors speculated that \'Chloroquine\... seems to have enhanced the induction of liver enzymes. \[It\] probably increased the catabolism of thyroid hormones by enzymatic induction.\[[@B51]\] They also suggested that chloroquine might act centrally on the hypothalamus, through disruption of the feedback system by which thyrotropin releasing hormone stimulates the pituitary to release and later synthesise TSH.\[[@B52]\] A third mechanism by which chloroquine and chemically related drugs such as mefloquine might interfere with thyroid function is through structural homology to T~3~, resulting in thyroid hormone antagonism.\[[@B53]\] In the rat, chloroquine injections more than halve the T~3~ concentration, without changing the level of free T~4~.\[[@B54]\] Chloroquine has been reported to inhibit T~3~ uptake in mammalian cells by inhibiting receptor-mediated endocytosis. \[[@B53]\] Testing the hypothesis ====================== Our hypothesis needs to be tested through a large multicentre cohort study of mefloquine prophylaxis in tourists and business travellers, perhaps recruited in collaboration with one or more airlines. Small randomised controlled trials should test specific elements of the hypothesis, such as the postulated link between mefloquine and thyroid disturbance, and the presumed interaction between mefloquine and oral contraception. National pharmacovigilance databases should also be analysed systematically to see if the spontaneous reports of mefloquine\'s adverse effects tend to support our hypothesis or not. The multicentre cohort study of prophylactic mefloquine use should be questionnaire-based, and should enquire specifically into the major risk factors (alcohol intake during travel, hydration status, use of hormonal contraception and recreational drugs, other potentially hepatotoxic or thyrotoxic drugs, previous history of proven or suspected liver and/or thyroid abnormality) that we have proposed. The study design should include pre- and post-exposure testing of liver and thyroid function in at least a sample of the cohort. One or more case-control studies should be nested within the cohort study. \[[@B55],[@B56]\] These nested studies would allow for rigorous testing of the aetiological mechanisms which we have proposed for mefloquine\'s adverse effects. The studies should also resolve those prescribing issues on which experts\' opinions differ (eg, Is mefloquine safe in pregnancy? Is it safe for long-term prophylaxis? Should airline pilots be prescribed mefloquine? Should mefloquine be prescribed to people with a personal or family history of neuropsychiatric illness? Can mefloquine be given safely as a pre-travel loading dose, for rapid induction of chemoprophylaxis? \[[@B17],[@B57]\]). Mefloquine is a clinically important drug, commonly used by healthy people. A much higher standard of safety is therefore required for mefloquine prophylaxis than for drugs given to treat serious diseases.\[[@B58],[@B59]\] The study we propose is urgently needed. Implications of the hypothesis ============================== What our hypothesis explains ---------------------------- The hypothesis explains much of the complexity of the pattern of adverse effects of mefloquine, and also the fact that many healthy users of the drug suffer no adverse effects at all. It also explains some aspects of mefloquine\'s tolerability profile in travellers which until now have not been understood, notably that young women experience more adverse effects from mefloquine than men, and that children and older adult travellers (who do not use oral contraception, and who rarely misuse alcohol) seem to tolerate the drug well. In addition, our hypothesis may explain the ethnic and inter individual variations in the pharmacokinetics of this agent, which until now have not been understood. \[[@B23],[@B24]\] It has been known for some years that mefloquine users who take co-medications are about 1.5 times more likely to experience an adverse drug event than those users who take no co-medications, and twice as likely to experience severe adverse drug events.\[[@B14]\] The frequency of reported adverse drug events also increases when multiple co-medications are taken.\[[@B14]\] Our hypothesis plausibly explains these earlier findings. The use of marijuana and other psychoactive agents by chloroquine users has been associated with acute psychotic reactions, and it has been suggested that mefloquine users who take recreational drugs may likewise be predisposed to neuropsychiatric problems.\[[@B60],[@B61]\] We believe that this association is consistent with our hypothesis, since recreational drugs can cause hepatocellular damage and liver failure.\[[@B34]\] A puzzling observation is that although the adverse effects of mefloquine are usually reversible, in some patients these effects can persist for months or even years after the drug has been stopped.\[[@B61]\] We believe that the occasionally protracted time course of the adverse effects can be plausibly explained by supposing that in these patients mefloquine is just one of several concurrent insults to the liver, and that it is the continuance of the other insults (most commonly alcohol, and certain prescription drugs) that makes the mefloquine-induced syndrome persist. Some published evidence supports this view.\[[@B38],[@B62],[@B69]\] There is evidence from this study (Table [1](#T1){ref-type="table"}) that the adverse effects of mefloquine in those who have taken the drug prophylactically persist longer than they do in patients who have been treated with it, even though the latter group mostly receive larger doses of the drug; this paradox might be explained by the fact that although prophylactic users usually stop mefloquine as soon as they have experienced adverse effects, they often continue to assault their livers in other ways, and it is this which makes the effects persist. Most of the reported adverse effects of mefloquine fit into the model we propose. For example, mefloquine has been associated with a wide variety of dermatological adverse effects, and most of these can be linked with effects on the liver or thyroid.\[[@B41],[@B55]\] Convulsions and dizziness are other reported effects of mefloquine which can be related to liver or thyroid disturbance.\[[@B2],[@B41]\] Who should not take mefloquine? ------------------------------- We believe that people with a history of any proven or suspected liver or thyroid abnormality in the previous two years should avoid mefloquine. Travellers taking mefloquine should not drink alcohol, especially within 24 h of their weekly mefloquine dose. While taking mefloquine, travellers should be advised to maintain good hydration with water or carbonated drinks, especially on long plane journeys or during arduous work in hot conditions. Alcohol, tea or coffee should not be used to maintain hydration, since they all increase water loss. Travellers taking mefloquine should not take a hormonal contraceptive, nor any other drug known to injure liver cells.\[[@B34]\] They should not take any drug known to antagonise thyroid hormone. \[[@B53]\] We propose that drugs that are known to cause hepatocellular injury *and also* to be thyroid hormone antagonists should be absolutely contraindicated in mefloquine users; such drugs include amiodarone, benzodiazepines, calcium channel blockers and phenytoin.\[[@B34],[@B53]\] Because of the potential for additive toxicity, travellers taking mefloquine should avoid concurrent use of any other quinoline derivative (eg, amodiaquine, chloroquine, primaquine, quinidine, quinine, tafenoquine), whether for additional prophylaxis or for treatment. The administration of a different quinoline derivative at the same time as mefloquine may increase the risk of adverse effects.\[[@B70],[@B72]\] For the same reason, mefloquine users should avoid other quinine analogues, such as fluoroquinolone antibiotics. Fluoroquinolones, such as sparfloxacin, of loxacin and ciprofloxacin, are increasingly prescribed in severe cases of traveller\'s diarrhoea, and have been associated with severe reactions in mefloquine users. \[[@B73]\] Who should take mefloquine? --------------------------- Mefloquine is a safe and exceptionally useful drug for the mass prophylaxis and treatment of those resident populations in malaria-endemic areas which traditionally abstain from alcohol and hormonal contraception. It has been suggested that in such settings, mefloquine should be combined with artemisinin or a derivative to protect both drugs from resistance.\[[@B74]\] On the basis of our hypothesis, we believe that children, and also pregnant women, can safely be prescribed mefloquine (because neither group uses alcohol or oral contraceptives). Accompanied by explicit advice to avoid alcohol, maintain good hydration, and use co-medications with caution, we believe that prophylactic mefloquine could be recommended to certain occupational subsets of travellers who carry out safety-critical tasks and who until now have been denied the use of this drug. These occupational groups include airline pilots,\[[@B75]\] divers\[[@B76]\] and operators of heavy machinery.\[[@B77]\] Some authorities advise that travellers engaged in high-risk leisure pursuits, such as mountain climbing, should not use mefloquine. \[[@B78]\] We believe that this exclusion is unjustified, as long as there is no recent history of liver or thyroid disease, and provided the precautions we have proposed above (avoidance of alcohol, maintenance of hydration and non-use of hormonal contraception, recreational drugs and certain co-medications) are adhered to. Lobel et al consider that WHO\'s exclusion of people with a personal or family history of neuropsychiatric illness from taking mefloquine is based \'on limited evidence or theoretical concerns\', and we believe their scepticism is justified.\[[@B79]\] Neuropsychiatric illness may not contraindicate use of mefloquine, provided that the patient is not currently taking anything that can cause liver damage or thyroid disturbance. Competing interests =================== None declared. ::: {#T2 .table-wrap} ::: {.caption} ###### Proposed contraindications to the use of mefloquine ::: ----------------------------------- ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- **Mefloquine contraindications:** 1\. Known **hypersensitivity** to the drug. 2\. History of any proven or suspected **liver abnormality** within the previous 2 years. 3\. History of any proven or suspected **thyroid abnormality** within the previous 2 years, including any concurrent use of thyroid-stabilising medication. 4\. Concurrent use of drugs known to cause **hepatocellular injury.** Paracetamol and (especially) aspirin to be used with caution, since both can damage the liver and/or the thyroid.\[[@B34],[@B40]\] Mefloquine users should not take recreational drugs. 5\. Concurrent use of **oral contraceptive pill** or **HRT.** Healthcare advisers should recognise that some female travellers will need alternative contraception. 6\. Concurrent use of drugs known to be **thyroid hormone antagonists.**\[[@B53]\] 7\. Up to 2 units of **alcohol** per day may be taken by users of mefloquine prophylaxis until 24 h before their weekly dose, and from 24 h afterwards. 8\. Mefloquine users should avoid concurrent use of **any other quinoline derivative** (eg, amodiaquine, chloroquine, primaquine, quinidine, quinine, tafenoquine), whether for additional prophylaxis or for treatment. ----------------------------------- ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ::: Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We thank Professor Alan McGregor FRCP, of Guy\'s, King\'s and St Thomas\' School of Medicine, London, for his helpful comments on an earlier draft of this paper.

Background ========== Pleural effusion is a common clinical problem resulting from thoracic or systemic diseases. An effusion is termed as transudate when its formation is due to alterations in the mechanical factors such as increased pulmonary hydrostatic pressure or decreased plasma oncotic pressure \[[@B1],[@B2]\]. An effusion is classified as an exudate when the pleural fluid is formed because of loss of pleural surface integrity by inflammatory or infiltrative processes causing an increase in microvascular permeability \[[@B1],[@B3]\]. Based on this principle pleural fluid is termed an exudate if fluid to serum total protein ratio (TPR) is ≥ 0.5, or the pleural fluid absolute value of FLDH is ≥ 200 IU/L, or the fluid to serum ratio of LDH value (LDHR) is ≥ 0.6. An effusion is classified as a transudate if the TPR is \<0.5, FLDH is \<200 IU/L, and the LDHR is \<0.6 \[[@B4]\]. Roth and associates have documented that serum-effusion gradient of albumin is a better discriminator than Light\'s criteria in the diagnostic separation of transudates and exudates \[[@B5]\]. On the other hand, Burgess *et. al.* \[[@B6]\] using an albumin gradient of 12 g/L found the sensitivity and specificity to be 87% and 92% respectively and concluded that the criteria by Light *et. al* \[[@B4]\] remained the best method for distinguishing exudates from transudates. It is worth noting here that both studies used conventional analysis and not Receiver Operating Characteristics (ROC). Recently, we have established that FLDH and TPR perform better and that fluid to serum ratio of LDH (LDHR) had no role in the diagnostic separation of transudates and exudates using ROC analysis \[[@B7]\]. Therefore, this study was undertaken to compare the accuracy of serum-effusion albumin gradient (SEAG) and fluid to serum albumin ratio (ALBR) with the best available criteria i.e. FLDH from our previous report using ROC \[[@B7]\]. Materials & Methods =================== Patients referred for evaluation of pleural effusion to the respiratory unit at Rotherham Trust hospital, United Kingdom, from January 1989 to June 1991 for a prospective investigation of the dynamics of pleural effusion formation and removal were included in the study \[[@B8]\]. The cause of pleural effusion and their diagnostic separation was determined using clinical criteria set by Light and associates \[[@B4]\]. All patients were followed for at least 3 months or till the final cause of the pleural effusion was established. During the study, pleural effusion samples from two hundred and twelve (212) consecutive patients were collected. However, eight patients with uncertain diagnosis and four patients with possible multiple causes for the pleural effusion were excluded from the analysis. Effusions due to pulmonary embolism were classified as exudates according to Burgess *et al* \[[@B6]\]. Blood and pleural effusion samples collected and stored at -80°C were analyzed within a period of six months for albumin, total protein and LDH level. The pleural fluid and the corresponding serum samples were drawn simultaneously at the time of admission before any treatment was administered. LDH was measured with Boehringer Mannheim kit according to established methods and results expressed in IU/L \[[@B9]\]. The upper limit for the normal serum LDH value in our laboratory during the study was 200 IU/L. SEAG was calculated by subtracting the pleural fluid albumin value from serum albumin value. Statistical analysis -------------------- The median values for SEAG, ALBR and FLDH were compared in transudates and exudates using Mann-Whitney U test. A p-value of \<0.05 was considered significant. To compare the performance of SEAG, ALBR and FLDH in the diagnostic separation of transudates and exudates, ROC curves were generated for each of the criteria by plotting the true positive value (sensitivity) against the false positive value (1- specificity) for multiple test results. Area under the curve (AUC) with 95% confidence intervals was calculated using Wilcoxon estimate. The AUC for the three tests were compared using the method described by Hanley and McNeil \[[@B10]\]. Furthermore, to account for multiple comparisons, we performed Bonferoni\'s correction of the final p value. Optimum cut-off points for SEAG, ALBR and FLDH were computed by using the statistical software StatsDirect () taking the best possible values for sensitivity and 1-specificity on the ROC plot. Finally, the correlation between SEAG and ALBR was estimated using Pearson\'s correlation. Data on age is presented as mean ± standard deviation and other test variables are presented as median and interquartile range (IQR) as they are not distributed normally. Results ======= Of the 200 patients with pleural effusion, 116 were males and 84 were females. The mean age ± standard deviation was 62 ± 15. One hundred and fifty six (78%) effusions were exudates. The causes for the pleural effusion were: thoracic malignancy 90, thoracic infection including empyema 37, pulmonary embolism 11, pancreatitis 6, autoimmune disease 5, other exudates 7, cardiac failure 37, chronic liver disease 5, and nephrotic syndrome 2. The median and IQR for SEAG, ALBR and FLDH are given in table [1](#T1){ref-type="table"}. The median value for all three tests in transudates and exudates were significantly different. ::: {#T1 .table-wrap} ::: {.caption} ###### Median and Interquartile range (IQR) for the three tests ::: **Test** **Transudate (n = 44)** **IQR** **Exudate (n = 156)** **IQR** **P value ^\*^** ------------- ------------------------- ------------- ----------------------- ------------ ------------------ SEAG (g/L) 93.5 33.8--122.5 308.5 171--692 0.0001 ALBR 0.49 0.42--0.62 0.77 0.63--0.85 0.0001 FLDH (IU/L) 98.5 76.8--127.5 344 216--695 0.0001 ^\*^ Mann-Whitney U test ::: The ROC plots for SEAG, ALBR and FLDH are shown in figure-[1](#F1){ref-type="fig"}. At all test values, FLDH was more sensitive and specific than SEAG and ALBR. The optimum cut-off levels were 163IU/L for FLDH, 9 g/L for SEAG, and 0.68 for ALBR respectively. FLDH was the best of the three tests. The AUC and the corresponding 95% Cl for SEAG, ALBR and FLDH were 0.81 (0.75 to 0.87), 0.79 (0.72 to 0.86) and 0.9 (0.87 to 0.96) respectively. Although the AUC for SEAG and ALBR were of similar magnitude, the AUC for FLDH was significantly more than the AUC for ALBR (z = 2.75, p \< 0.01). Of the 156 exudates, FLDH correctly classified 132 (85%) 95% Cl 0.78--0.90, SEAG 112 (72%) 0.64--0.79, and ALBR 111 (71%) 0.63--0.78 respectively as exudates and all three tests classified 39 (89%) 0.75--0.96 of the 44 transudates correctly. Table [2](#T2){ref-type="table"} lists the positive likelihood ratios (LR +ve) with 95% Cl for the three tests at the optimum cut-off level. Figure-[2](#F2){ref-type="fig"} shows the correlation between SEAG and ALBR. There was a significant negative correlation between the two tests (r = -0.89, p \< 0.0001). {#F1} {#F2} ::: {#T2 .table-wrap} ::: {.caption} ###### Positive Likelihood Ratios (LR) at optimum cut-off points ::: **Test** **LR** **95% CI** ---------- -------- ------------ FLDH 7.3 3.5 to 17 SEAG 6.3 3 to 15 ALBR 6.2 3 to 14 ::: Discussion ========== Pleural effusion is a common clinical problem and it\'s diagnostic approach has been simplified by characterizing effusion into transudate and exudate \[[@B11]\]. Several biochemical markers have been used in the past for this purpose \[[@B12]-[@B14]\]. However, TPR and FLDH have remained the most widely used criteria so far. Therefore, this study was undertaken to assess the global performance of SEAG and ALBR with the existing best criteria i.e. FLDH \[[@B7]\]. The strength of our study is that we have utilized ROC analysis for comparing different tests. Furthermore, ROC analysis provides an opportunity to calculate likelihood ratios at various cutoff points of the diagnostic test. The upper limit of the normal LDH varies according to the methods used for estimating lactate dehydrogenase. In this study we have used the Boehringer Mannheim kit and the upper normal for this method is 200IU//L \[[@B9]\]. In our present study, the mean age and causes of pleural effusion were similar to that of Roth et al. \[[@B5]\]. However, the mean SEAG value of 14 g/L for transudate and 8 g/L for exudate in our series were different from that provided by Roth and associates about 20 g/L for transudates and 0.75 g/L for exudates. From ROC analysis it is clear that FLDH performed better compared to SEAG and ALBR as documented by the highest value for the AUC. ALBR has been reported to be inferior to SEAG in the diagnostic separation of transudates and exudates \[[@B5]\]. However, in our analysis the accuracy of ALBR and SEAG were similar as evidenced by similar AUC and positive LR. Furthermore, correlation analysis between SEAG and ALBR has shown a significant negative correlation (0.89). As the level of pleural fluid albumin increases depending on the degree of pleural microvacular injury in exudative processes, the ALBR increases correspondingly. Whereas, pleural fluid to SEAG is highest in transustive effusions as less albumin is filtered through the relatively normal pleural microvasculature. However, as pleural microvasculature is damaged in exudative processes, the level of albumin leaking into the pleural cavity would progressively increase depending on the degree of injury. Thus a very low SEAG can be expected in a trasudative processes with a correspondingly higher ALBR. This would partly explain the significant negative correlation observed between SEAG and ALBR thus indicating that these two tests likely to be operating through the same mechanism \[[@B15]\]. In our study using 9 g/L as the optimum cutoff level for SEAG, 51 of the 156 exudates were incorrectly labeled as transudates (false negative) and 5 of the 44 transudates were labeled as exudates (false positives). There were no trends for any specific diseases clustering in either of the groups. The pathophysiology of transudative or exudative pleural effusion in humans is not well studied. We have previously documented that the rate of formation and absorption are different in different disease states causing pleural effusion \[[@B8]\]. The decreased concentration of total protein and albumin in transudates can be explained by the selective sieving of these macromolecules by the normal intact microvasculature in transudates \[[@B3]\]. However, animal and human studies have shown that increase in hydrostatic pressure can increase the protein content in effusions \[[@B16],[@B17]\]. This is partly explained by the opening up of large pores in endothelium with elevated hydrostatic pressure \[[@B18],[@B19]\]. Thus in a patient with pleural effusion from congestive heart failure, it is possible that the amount of protein in the pleural fluid can increase significantly to exudative levels if the degree of pulmonary artery pressure causing the formation of effusion was very high. Such a linear correlation between the hydrostatic pressure and the protein content of ascites in patients with portal hypertension has been documented \[[@B17]\]. Therefore, this phenomenon may help explain the pathogenesis of exudative effusions documented in congestive heart failure when total protein in pleural effusion was used as a diagnostic criteria \[[@B5],[@B20],[@B21]\]. The dynamics of LDH in the pleural cavity is not well studied. Our recent investigation into the dynamics of pleural fluid LDH has documented that there was no correlation between serum and pleural fluid LDH concentration in transudates or exudates \[[@B7]\]. This compartmentalized elevation of LDH concentration in pleural cavity in exudative processes help explain the superior performance of FLDH as a diagnostic test in the separation of pleural effusion into transudate and exudate compared to SEAG, ALBR. We have reported earlier that LDHR has no role in the diagnostic separation of transudates and exudates. Furthermore, FLDH is the most accurate marker and combining TPR with FLDH improves the diagnostic utility marginally \[[@B7]\]. Conclusion ========== The discriminative value for SEAG and ALBR appears to be similar in the diagnostic separation of transudates and exudates. FLDH is a superior test compared to SEAG and ALBR. Based on these results, SEAG and ALBR appear to have limited diagnostic utility in clinical practice. Competing interests =================== None declared Author contributorship ====================== Author one JJ conceived the idea, designed the study, collected the data, took part in analysis of data and drafted the manuscript. Author two PB analyzed the data, interpreted the results and contributed in writing the manuscript. Author three GSB designed the study, supervised the data collection and contributed in writing the manuscript. Author four SS participated in the data analysis, interpreting results and contributed in writing and editing the manuscript. All authors read and approved the final manuscript. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgement =============== The authors would like to thank Mr. Steve Viney for carrying out the pleural fluid biochemical analysis.

Background ========== The genus *Hapalemur* of the Malagasy Prosimian family Lemuridae comprises three clearly different species: *Hapalemur griseus*\[[@B1]\], *H. simus*\[[@B2]\] and *H. aureus*\[[@B3]\]. Classification of different types within the *H. griseus* group has been controversial, particularly in the light of the variations existing in coat colour \[[@B4]\]. Systematic cytogenetic studies led to the description of four subspecies of *H. griseus* characterised by different karyotypes: *H. g. griseus* (HGG) (2N = 54) \[[@B5],[@B6]\], initially considered as the most widespread and supposed to live in the Eastern forest of Madagascar from the north to the south \[[@B7]\]; *H. g. occidentalis* (HGO) (2N = 58) \[[@B6]\] restricted in the north-west and the eastern part of Madagascar as east as Maroantsetra \[[@B4],[@B8]\]; *H. g. meridionalis* (HGM) (2N = 54) \[[@B9]\] found in the south, near Fort Dauphin, and differing from *H. g. griseus* in one metacentric chromosome ; *H. g. alaotrensis* (HGA) (2N = 54) \[[@B6]\], restricted to the bed of the Alaotra lake, with a karyotype similar to HGG containing more juxtacentromeric heterochromatin, and characterized by a larger body size \[[@B4]\]. In addition, a new variant of HG, *H. griseus ssp* (HGssp), has recently been described. This type is characterized by a chromosomal polymorphism (2N = 54, 55 and 56) and is living in the area of Ranomafana-Kianjavato \[[@B8]\]. The comparison of the karyotypes of the different *Hapalemur* led to a phylogenetic tree characterised by an earlier emergence *of H. simus* (HSI), followed by *H. aureus* (HAU) which represents the sister clade of the different groups of the *H. griseus* subspecies \[[@B8],[@B10]\]. As molecular studies were fruitfully used for taxonomic investigations of several taxa, including lemurs \[[@B11]-[@B15]\], and as no large morphological differences could be found between the cytogenetically close subspecies of *H. griseus,* mitochondrial cytochrome b and 12S DNA analyses were performed in order to determinate more accurately their species and/or subspecies status. Results ======= A total of 115 animals covering all the genus *Hapalemur* were captured in bamboo forests extending from the north to the south of Madagascar. The different capture areas of each species and subspecies as the number of animals are indicated in Fig. [1](#F1){ref-type="fig"}. {#F1} The PCR amplification of the cytochrome b resulted in a 357 bp fragment. Alignments using blast search demonstrated that this fragment corresponds to the beginning of the mitochondrial cytochrome b gene. All the sequences were aligned and the substitutions were scored. Animals sharing the same sequences were regrouped into a same haplotype. These different haplotypes are summarised in Fig. [1](#F1){ref-type="fig"} and table [2](#T2){ref-type="table"}. Consensus sequences constructed with the most common sequence in each species and subspecies (not shown) demonstrated that transition substitutions occur more frequently than transversions (an average of eight transitions for 1 transversion) and gaps are absent. As expected \[[@B16]\], most of the substitutions occur on the third position of the codon (54 out of 64 substitutions), while the first position changed nine times and the second once. The portion amplified comprises the four invariant codons considered necessary for cytochrome b function \[[@B17]\]. Changes were sometimes noticed in the third base of the codon in these positions, but the amino acid remains the same so that the functionality of the cytochrome b is not affected. ::: {#T1 .table-wrap} ::: {.caption} ###### Lower and higher values of Kimura two parameters distances (under the diagonal) and absolute distances (above the diagonal) between haplotypes of *Hapalemur* species and subspecies. For abbreviations see Figure [1](#F1){ref-type="fig"}. ::: HSI HAU HGM HGsspb HGA HGG HGsspa HGO -------- --------------------- --------------------- --------------------- --------------------- --------------------- --------------------- ---------------------- --------------------- HSI **0--7 0--0.01991** 36 42 38 46 38 42 30 35 30 36 30 39 31 40 HAU 0.11024 0.13002 **0--2 0--0.00563** 41 48 42 46 31 34 30 36 30 36 34 39 HGM 0.11797 0.14524 0.12893 0.15338 **0--9 0--0.02565** 0 10 23 28 21 29 21 34 25 32 HGsspb 0.11797 0.13085 0.13290 0.14524 0.00000 0.02855 **0--5 0--0.01414** 22 26 21 27 21 32 25 30 HGA 0.09098 0.11075 0.09431 0.10393 0.06558 0.08459 0.06558 0.07747 **0--1 0--0.00281** 1 3 2 9 4 10 HGG 0.09098 0.11075 0.09098 0.11075 0.06240 0.08789 0.06240 0.08092 0.00281 0.00847 **0--4 0--0.01133** 3 10 5 11 HGsspa 0.09098 0.12087 0.09098 0.11049 0.06240 0.10473 0.06240 0.09719 0.00563 0.02580 0.00847 0.02875 **0--10 0--0.02868** 5 13 HGO 0.09431 0.12469 0.10445 0.12117 0.07522 0.09767 0.07522 0.09055 0.01133 0.02883 0.01421 0.03180 0.01421 0.03781 **0--6 0--0.01705** ::: Kimura two parameter distances \[[@B18]\] and absolute distances calculated with the different haplotypes (Table [1](#T1){ref-type="table"}) as phylogenetic trees constructed with the computer program Pylogenetic Analysis Using Parsimony (PAUP) 4.0b.4a \[[@B19]\] (Figs [2](#F2){ref-type="fig"}, [3](#F3){ref-type="fig"}, [4](#F4){ref-type="fig"}, [5](#F5){ref-type="fig"}) were used to compare the different *Hapalemur* species and subspecies. Neighbor-joining analyses of cytochrome b haplotypes demonstrated that *H. simus* emerged first followed by *H. aureus* which is the sister clade of the *H. griseus* subspecies. Inside *H. griseus* subspecies, two subclades are present and the *H. griseus ssp* (HGssp) are distributed in both the subclades (Figs [2](#F2){ref-type="fig"}, [3](#F3){ref-type="fig"}). For this reason, the *H. griseus ssp* were separated in HGssp~a~ and Hgssp~b~. The first subclade (S1) contains HGA, HGG, HGssp~a~ and HGO. The second subclade (S2) is composed of HGM and HGssp~b~. Maximum parsimony analyses (data not shown) demonstrated the same topology, unless the bootstrap support for S1 dropped from 94 to 66. {ref-type="fig"}.](1471-2148-2-4-2){#F2} {ref-type="fig"}.](1471-2148-2-4-3){#F3} {ref-type="fig"}.](1471-2148-2-4-4){#F4} {ref-type="fig"}.](1471-2148-2-4-5){#F5} ::: {#T2 .table-wrap} ::: {.caption} ###### Haplotypes and GenBank accession numbers of Hapalemur cytochrome b sequences. ::: Species and subspecies/number of individuals Haplotype/number of individuals GenBank accession numbers ---------------------------------------------- --------------------------------- --------------------------- *Hapalemur simus* (HSI)/14 01/3 AJ428977 02/1 AJ428978 03/4 AJ428979 04/6 AJ428980 *Hapalemur aureus* (HAU) / 5 01/4 AJ428957 02/1 AJ428958 *Hapalemur griseus meridionalis* (HGM)/16 01/1 AJ428959 02/1 AJ428960 03/1 AJ428961 04/1 AJ428962 05/1 AJ428963 06/1 AJ428964 07/1 AJ428965 08/7 AJ428966 09/1 AJ428967 10/1 AJ428968 *Hapalemur griseus ssp* (HGssp) / 50 01/7 AJ429054 02/12 AJ429055 03/3 AJ429056 04/2 AJ429057 05/6 AJ429058 06/1 AJ429059 07/6 AJ429060 08/5 AJ429061 09/1 AJ429062 10/2 AJ429063 11/2 AJ429064 12/1 AJ429065 13/1 AJ429066 14/1 AJ429067 *Hapalemur griseus alaotrensis* (HGA) / 3 01/1 AJ428969 02/2 AJ428970 *Hapalemur griseus griseus* HGG)/18 01/1 AJ428971 02/12 AJ428972 03/1 AJ428973 04/1 AJ428974 05/1 AJ428975 06/2 AJ428976 *Hapalemur griseus occidentalis* (HGO) / 9 01/2 AJ428982 02/2 AJ428983 03/1 AJ428984 04/1 AJ428985 05/2 AJ428986 06/1 AJ428987 *Eulemur macaco flavifrons* (EMF)/11 01/11 AJ428981 ::: The use of the Kimura two parameter method (Table [1](#T1){ref-type="table"}) shows inside S1 a very close distance between HGA and HGG (ranging from 0.00281 to 0.00847), while a larger distance occurs between HGG and HGssp~a~ (0.00847 to 0.02875) as well as between HGG and HGO (0.01412 to 0.03180). In S2 (Table [1](#T1){ref-type="table"}), short distances are observed between HGM and HGssp~b~ (0--0.02855), (in this subclade, the sequence of HGM 02 and HGssp 01 are identical). Larger distances ranging from 0.10473 (HGssp~a~/HGM) to 0.06240 (HGM-HGssp~b~/HGG-HGssp~a~) separates S1 from S2 (Table [1](#T1){ref-type="table"}). In order to confirm the position of *H. griseus ssp* inside the *H. griseus* subclades, a portion of 12S mitochondrial DNA was amplified and sequenced. The resulted 438 bp fragments were aligned, and the sequences were analyzed with the program PAUP. The trees obtained with HGO, HGG, HGssp and HGM 12S sequences (Figs [4](#F4){ref-type="fig"}, [5](#F5){ref-type="fig"}) also demonstrated a separation of HGssp into two subclades. The first (HGssp~a~), which is composed of animals clustering with HGO and HGG when cytochrome b sequences are analyzed (Figs [2](#F2){ref-type="fig"}, [3](#F3){ref-type="fig"}), also clusters with HGO and HGG when 12S sequences are employed (Figs [4](#F4){ref-type="fig"}, [5](#F5){ref-type="fig"}). The second (HGssp~b~), constructed with animals clustering with HGM when cytochrome b sequences are analyzed (Figs [2](#F2){ref-type="fig"}, [3](#F3){ref-type="fig"}), also clusters with HGM when 12S sequences are employed (Figs [4](#F4){ref-type="fig"}, [5](#F5){ref-type="fig"}). Discussion ========== During the last decade, comparisons of mitochondrial DNA sequences have been very useful for the analyses of the phylogenetic species relationships including lemurs \[[@B11]-[@B15],[@B20]\]. In our study, cytochrome b and 12S mtDNA sequences were used, in order to clarify the position of each species and subspecies in the genus *Hapalemur.* Despite the fact that we analysed only short sequences, the genetic distances are considered to be relevant because of the large number of animals involved, since large sample size reduces errors in the estimation of evolutionary relatedness \[[@B11],[@B21]\]. Comparisons of cytochrome b sequences demonstrated that pairwise genetic distances between *H. simus* and *H. aureus* and the other *Hapalemur* (distances ranging from 0.09098 to 0.15338) are the highest. The corresponding phylogram also shows that these taxa are well separated. Our molecular data thus strongly support the species status for *H. simus, H. aureus* and *H. griseus.* Among the *H. griseus,* the genetic divergence observed between *H. g. meridionalis* and the HGG, HGA, HGssp~a~ and HGO (ranging from 0.06240 to 0.10473) is markedly higher than all other values of the same level of intra *H. griseus* comparison (ranging from 0.00281 to 0.03781). The phylogram also clearly separates the *H. g. meridionalis* from other *H. griseus.* This high genetic divergence in the ranging of those observed between species of lemurs, would support the classification *of H. g. meridionalis* in a separate species *\"H. meridionalis\"* despite the relatively small chromosomal differences observed \[[@B9]\]. Lower distances (ranging from 0.01133 to 0.03781) are observed between *H. griseus* and *H. g. occidentalis.* Moreover, these two forms cluster in the same clade. So, despite the cytogenetic differences existing between these two forms, which differ by two chromosomal rearrangements \[[@B10]\], we propose the maintenance of the subspecies status for *H. g. occidentalis.* On the basis of our analyses, *H. g. alaotrensis* is indistinguishable from *H. g. griseus.* The very short genetic distances found between *H. g. griseus* and *H. g. alaotrensis* (0.00281 to 0.00847) suggests a combination of these two groups into a single subspecies. However, the differences of the body sizes \[[@B4]\] and the differences in the content of heterochromatin found in both karyotypes \[[@B22]\] support the separation *of H. g. griseus* and *H. g. alaotrensis* in separate subspecies \[[@B23]\], but in no case in species apart as it has been previously proposed \[[@B24]\]. The systematic position of HGssp is more difficult to establish when cytogenetic and molecular data are compared. Cytogenetic data demonstrated the existence of an unique chromosomal polymorphic subspecies characterized by two karyotypes, 2N = 56 and 2N = 54, and their hybrids 2N = 55 \[[@B8]\]. In the view of molecular data, HGssp contains two groups, HGssp~a~ and HGssp~b~, separated by an important genetic distance (0.06240--0.09719) in the range of those observed between species. Each of these two groups contains both karyotypes, 2N = 56 and 2N = 54, as well as their hybrids. The HGssp~a~ are separated from HGG, HGA and HGO by a genetic distance in the range of those observed between subspecies (0.00563--0.03781), while HGssp~b~ appears similar to HGM. Moreover, the HGssp~b~ haplotypes are mixed with those of HGM. Taking into account these results, HGssp~a~ should be considered as belonging to the group of HGG, HGA and HGO, and HGssp~b~ as belonging to the group of HGM. The existence of two well-separated clades among the *Hapalemur griseus ssp* originating from the Ranomafana region could possibly resulted from either recent mitochondrial DNA introgression or ancestral polymorphism. As the boundaries between HGM and HGssp are still unknown, we could hypothesize that the limits could be close to Ranomafana, allowing hybridization between these two forms. A transfer of mitochrondrial DNA from HGM into the HGssp population could have occurred through a matrilinear process resulting in the HGssp~b~ haplotype. New investigations in areas located in the south of Ranomafana should thus allow the finding of HGM populations containing introgressed HGssp mitochondrial DNAs, unless this transfer occurs only from HGM to HGssp. A second hypothesis is that an ancestral polymorphic population containing both HGM and HGssp haplotypes and a chromosomal polymorphism have diverged in two separated populations. In the population of Ranomafana, the HGM and HGssp haplotypes as well as the chromosomal polymorphism were maintained. In the second population, only the HGM haplotype remained present, and a gain of a large block of heterochromatin gave rise to the karyotype characteristic of the HGM \[[@B9]\]. The comparisons of the phylogenetic trees based on mitochondrial DNA sequences with those previously obtained from cytogenetic data are only partially concordant. In both trees, *H. simus* emerges first, followed by *H. aureus* and then by the different *H. griseus* forms. The cytogenetic data allow to propose an evolutionary tree in which *H. g. occidentalis* emerges first, followed by *H. g. griseus,* and *H. g. meridionalis*\[[@B10]\], whereas on the tree based on cytochrome b and 12S sequences, *H. g. meridionalis* appears as a sister clade of the other *H. griseus.* This difference may be related to the short distances observed inside each clade which allowed no branching order. Conclusions =========== Our molecular studies of the *Hapalemur* genus raise the question of the classification of *H. g. meridionalis* in the species status *H. meridionalis.* They also confirm the subspecies status *of H. g. occidentalis* and the absence of arguments in favour of the classification of *H. g. alaotrensis* as a separate species. The sequencing of the *Hapalemur griseus ssp* originating from Ranomafana reveals animals clustering either with the *H. meridionalis* or with the group of *Hapalemur griseus griseus.* No monophyletic clade could be determined in this new cytotype, so that the taxonomic status of the *Hapalemur griseus ssp* remains undefined. As our molecular data did not match the branching sequence within the *H. griseus* group based on cytogenetic data, further investigations including nuclear DNA will be necessary in order to resolve this issue. Materials and methods ===================== Animal studied -------------- Six survey were organised from 1997 to 2001 allowing the capture of 115 animals in bamboo forests extending from the north to the south of Madagascar. Animals were captured using blowpipe projections and then sexed, weighted and measured. Skin samples were cut off under general anaesthesia with a 2 mg/kg injection of ketamine solution (Ketalar^®^ Parke-Davis) and conserved deep frozen in liquid nitrogen. The different capture areas and the number of animals are indicated in Figure [1](#F1){ref-type="fig"}. As the *H. griseus* taxonomy is essentially based on cytogenetic criteria, karyotypes were made in order to confirm the species and subspecies rank. From the north to the south, the following subspecies were caught: a total of nine *H. g. occidentalis* (four at Analamera (a), two at Ambato (b), two at Ambakoany (c), and one at Maroantsetra (d)). Three *H. g. alaotrensis* at the Alaotra lake (e); 18 *H. g. griseus* at Maromiza (f); 50 *H. griseus ssp.* (37 at Ranomafana (h), five at Ambolomavo (i), eight at Kianjavato (j)); 16 *H. g. meridionalis* (14 in Andohahela (k) and two in Mandena (l)). In addition, 12 *H. simus* (four from Ranomafana (h), and eight from Ambolomavo (i)), as well as two *H. simus* and five *H. aureus* from the Zoological Park of Tsimbazaza (g) were studied. Animals were released at their capture location, immediately after recovery from anaesthesia, except two *H. g. occidentalis* captured in Ambato as well as three *H. g. alaotrensis* which are kept in the Zoological Park of Mulhouse and one *H. g. occidentalis* from Maroan setra which is kept in the private Zoological Park of Mandraka. As outgroup, we used a sample *of Eulemur macaco flavifrons,* from one of the individuals captured in the Sahamalaza forest for an earlier study \[[@B25]\]. DNA extraction -------------- DNA samples were extracted from the skin biopsies using the standard proteinase K digestion followed by a phenol chloroform extraction as described by Sambrook *et al.*\[[@B26]\] with minor modifications. Small pieces of skin (\~9mm^2^) were suspended in 200μl of extraction buffer (Tris 0.2M pH 8.4; KCl 0.5M; proteinase K: 5 mg/ml), and incubated at 37°C overnight. The samples were than heated at 95°C and mixed with an equal volume (200μl) of phenol/chloroform (1/1). After centrifugation (6 mn at 8500 g) the aqueous part was precipitated with 2 volumes of absolute ethanol at -20°C in presence of 1/10 (V/V) of ammonium acetate 5M. After centrifugation (5 mn at 8500 g), pellets were rinsed with 200μl of 70% ethanol and dried at 37°C. Pellets were resuspended in sterile double-distilled water and the concentration of the DNAs were measured by absorption at 260 nm. DNA samples were then stored at -30°C. Amplification conditions ------------------------ ### Cytochrome b The polymerase chain reaction (PCR) was employed to generate a double-stranded fragment of 357 bp corresponding to a part of the mitochondrial cytochrome b gene. Each amplification was performed in the presence of 20 mM Tris-HCl pH 8.4, 50 mM KCl, 1.5mM MgCl~2~ (Gibco BRL), 200 μg BSA, 0.5 μM of each primer, 5U of Taq DNA polymerase (Perkin Elmer Cetus), 200 μM of each dNTP (Boehringer Mannheim) and 360ng of template DNA in a volume of 200 μl. The following primers derived from those described by Kocher *et al.*\[[@B27]\] were employed: Pr181 (5\'-CCATCCAACATGTCAGCATGATGAAA-3\') and Pr182 (5\'-CCCTCAGAATGATATTTGTCCTCA-3\'). Reactions were done in a Perkin Elmer Cetus DNA thermocycler 480 as follows: predenaturation (7 mn at 93 °C) and 35 cycles of denaturation (30 s at 94°C), annealing (45 s at 42°C), extension (1 mn 30 s at 72°C), followed by a final extension step (10 mn at 72°C). ### 12S mitochondrial DNA The same PCR conditions than those described for the cytochrome b amplifications were applied, with a concentration of 2 mM MgCl~2~ and the following set of primers: Pr179 (5\'-AAACTAGGATTAGATACCCTATTAT-3\') and Pr180 (5\'-AAGAGCGACGGGCGATGTGT-3\'). Amplifications were performed in a Perkin Elmer Cetus DNA thermocycler 480 as follows: predenaturation (10 mn at 94°C) and 40 cycles of denaturation (30 s at 94°C), annealing (45 s at 45°C), extension (1 mn 30 s at 72°C), followed by a final extension step (10 mn at 72°C). ### Sequences Amplification products were electrophoresed on 1.3% agarose gels in TBE buffer (Tris base 87 mM, boric acid 89 mM, EDTA 2 mM, pH 8.0) in the presence of 0.5 μg/ml of ethidium bromide. Electrophoresis were performed at 200 mA for 2 h in TBE. Gels were than examined and photographed under UV light with a Polaroid system. The major bands on the gels were cut and centrifuged (8500 g 10 mn) in order to recover the DNA fragments \[[@B28]\]. The fragments were then precipitated with 0.1 volumes of ammonium acetate 5 M and 2 volumes of absolute ethanol at -20°C. After centrifugation (8500 g 10-mn), the pellets were air dried, resuspended in 50 μl of sterile bidistilled water and sequenced on an automatic ABI PRISM sequencer with the Taq dye deoxy terminator cycle sequencing kit. Each sample was sequenced from 5\' to 3\' and 3\' to 5\'. Cytochrome b sequences were aligned and for each taxa, similar sequences were grouped under one haplotype. These haplotypes and the EMBL GenBank accession numbers are listed in table [2](#T2){ref-type="table"}. EMBL GenBank accession numbers for the 12S sequences are listed in table [3](#T3){ref-type="table"}. ::: {#T3 .table-wrap} ::: {.caption} ###### GenBank accession numbers of Hapalemur 12S sequences. ::: Species/subspecies N° GenBank accession numbers ---------------------------------------- ----------- --------------------------- *Hapalemur griseus meridionalis* (HGM) 25 AJ429205 29^\*^ AJ429206 *Hapalemur griseus ssp* (HGssp) 41 AJ429207 43 AJ429208 59 AJ429209 *Hapalemur griseus griseus* (HGG) 06 AJ429210 10 AJ429211 *Hapalemur griseus occidentalis* (HGO) 19 ^\*\*^ AJ429212 20 AJ429213 23 AJ429214 24 AJ429215 ^\*^ 12S sequence of HGM29 is similar to HGssp40, HGssp42 and HGssp73. ^\*\*^ 12S sequence of HGO19 is similar to HGO22. ::: The aligned sequences were analysed using neighbor-joining and maximum parsimony methods with the computer program Pylogenetic Analysis Using Parsimony (PAUP) 4.0b.4a \[[@B19]\]. Genetic distances (d) measured in neighbor-joining analyses were calculated using the Kimura two parameter method \[[@B18]\] with the following formula: d=-1/21n \[(1-2P-Q)Xsqr(1-2Q)\], with P=transitions/positions scored and Q=transversions/positions scored. For neighbor-joining and maximum parsimony analyses, bootstraps of 10000 replicates were performed to examine the relative support of each relationship in the resultant topologies. Maximum parsimony trees were calculated via fast stepwise addition with random addition sequence. Acknowledgements ================ This work was partially supported by the Association Européenne pour l\'Etude et la Conservation des Lémuriens (A.E.E.C.L.), the DREIF-MENESR (EA3428), and the Institut de Génétique et de Biologie Moléculaire et Cellulaire (Pr. P. Chambon). The authors thank the Commission Tripartite of the Malagasy Government and the Ministère pour la Production Animale et des Eaux et Forêts for their permission to capture the animals, to take samples and for delivery CITES. Thanks are also due to Dr J. Ganzhorn for the correction of the manuscript, Mrs M. Lavaux for her secretarial assistance, Dr B. Ravaoarimanana and M. Hauwy for their helpful comments, and to S. Vicaire (IGBMC, Strasbourg) for sequencing the cytochrome b and 12S fragments. Special thanks for insight comments on the cladograms are due to Dr. D. Montagnon.

Background ========== Fruit-set and growth of several horticultural plants are negatively affected by adverse environmental conditions. In general, sub and/or supra-optimal temperatures negatively affect reproductive processes and therefore curtail fruit production \[[@B1],[@B2]\]. Under greenhouse cultivation, low temperature, insufficient light intensity, and/or high humidity drastically reduce fruit productivity and quality in eggplant and other species. Moreover, environmental conditions often met in open field cultivation such as drought and high temperatures have a negative effect on fruit productivity and quality in eggplant and other species (e.g. tomato). Parthenocarpic fruit development (i.e. fruit-set and growth without fertilization) can significantly aid in the resolution of the aforementioned problems. Parthenocarpy can be triggered by exogenous factors, such as plant growth regulators, or it can be achieved by genetic factors. Genes causing parthenocarpic development have been identified in several plant species \[[@B3]-[@B5]\], and parthenocarpic eggplant varieties (e.g. Talina, Galine) have been introduced in the production process (e.g. protected cultivation). However, during winter cultivation of eggplant varieties in unheated greenhouses in the Mediterranean area, the negative effect of suboptimal environmental conditions on fruit production is usually counteracted by treating flower buds with plant growth regulators. Phytohormonal treatments make the production process more expensive due to the cost of both chemicals and labor. In principle, the genetic engineering of plants allows one to confer a trait of interest to different species and within a species to all the varieties of interest. To confer parthenocarpic fruit development, a chimeric gene has been constructed \[[@B6]\]. Specifically, the *DefH9-iaaM* gene contains the coding region of the *iaaM* gene from *Pseudomonas syringae* pv. *savastanoi* under the control of the placenta and ovule-specific promoter from the *DefH9* gene from *Anthirrinum majus*\[[@B6]\]. The *iaaM* gene codes for a tryptophan monoxygenase, which produces indolacetamide that in turn, is either chemically or enzymatically converted to the auxin indole-3-acetic acid. To date, the *DefH9-iaaM* chimeric gene has been shown to cause parthenocarpic fruit development in tobacco \[[@B6]\], eggplant \[[@B6]\], tomato \[[@B7],[@B8]\], strawberry and raspberry \[[@B9]\]. We have previously shown that *DefH9-iaaM* eggplants outperform control plants during protected winter cultivation by an average of 33% \[[@B10]\]. The present manuscript presents data on the agronomic performance of *DefH9-iaaM* eggplant hybrids during spring and summer cultivation. Seed-derived F1 plants perfectly reflects the real agronomical situation of eggplant production. In addition, they have allowed to demonstrate that the transgene is active after meiosis and give satisfactory results in the hemizygous state. Different types of eggplant hybrids have been evaluated during springtime in unheated greenhouses in two different locations. To evaluate GM parthenocarpic eggplants under optimal environmental conditions, a single trial has been carried out using two different genotypes under standard open field cultivation during summertime. Furthermore, we demonstrate that the *DefH9-iaaM* gene is expressed during late stages of fruit development. Results and discussion ====================== Greenhouse spring production ---------------------------- Spring production was evaluated in trials performed in two different locations: Monsampolo and Pontecagnano. Spring production was divided into early production, consisting of the first four harvests, and total production, consisting of sixteen harvests. Early spring production corresponds to the cultivation period from March to the first half of May, with temperatures somewhat low for fruit-set and growth. During this period, the average minimum and maximum temperatures ranged from 7° to 17°C in southern Italy (Pontecagnano) and from 14° to 17°C in central Italy (Monsampolo). The transgenic parthenocarpic hybrids gave a significantly higher yield, on the average a six-fold increase during early production, in comparison to their controls (Table [1](#T1){ref-type="table"}). The increment in the number of fruits per plant and the higher average weight of the fruits were the main causes of the increased early spring production of the parthenocarpic hybrids. The suboptimal/adverse environmental conditions did not affect the growth of GM parthenocarpic fruits and the average fruit weight was significantly higher in the GM eggplants in comparison to untransformed controls. The traditional parthenocarpic cultivar Talina produced fruits with an average weight similar to those of GM eggplant fruits (Table [1](#T1){ref-type="table"}). However, due to the higher number of fruits per plant, the productivity of the transgenic parthenocarpic hybrids was also increased, by an average of five-fold, when compared to cv Talina (Table [1](#T1){ref-type="table"}). During the whole harvesting period, the GM fruits were always seedless (Fig. [1](#F1){ref-type="fig"}) and were normal in both size and shape. {#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### Eggplant production of parthenocarpic hybrids and their respective controls at springtime. ::: Genotype Early production Total production ---------- ------------------ ------------------ --------- -------- ------- --------- P1 488 b 1.9 a 268.3 a 2241 a 8.4 a 253.9 a P2 695 a 2.6 a 290.8 a 2288 a 8.6 a 260.2 a C1 75 c 0.4 b 227.0 b 1547 b 7.8 a 187.5 b P5 522 b 2.0 a 281.8 a 2163 a 9.2 a 230.5 a C2 116c 0.7 b 170.6 b 1574 b 9.3 a 167.3 b Talina 114c 0.3 b 270.5 a 2360 a 9.4 a 229.9 a For each trait at least one common letter indicates no significant difference according to the Duncan test (P = 0.05). Mean values of yields per plant, number of fruits per plant, and fruit weight of three transgenic parthenocarpic hybrids (P1, P2 and P5), two controls (C1 and C2) and the commercial cultivar Talina. C1 hybrid plants represent the controls of P1 and P2 transgenic hybrid plants. C2 hybrid plants are the controls of the P5 transgenic hybrids. Data are the average of the Monsampolo and Pontecagnano locations. The experiments were carried out in greenhouse at springtime. ::: The increased productivity of GM hybrids characterised both the early spring production (i.e. the first four harvests) and the whole spring production cycle (i.e. sixteen harvests). During the whole spring production cycle, the hybrids P1 and P2 gave an average yield that was 46% higher with respect to the corresponding control C1 (Table [1](#T1){ref-type="table"}). The hybrid P5 gave a 37% higher yield with respect to its control C2. The average total number of fruits produced per plant in both locations was similar in all the hybrids (8--9 fruits/plant). However, the higher average weight of the GM fruits led to a higher total yield of transgenic hybrids with respect to their controls. When considering the whole spring cultivation cycle, the parthenocarpic cultivar Talina gave a total production that was not significantly different from either of the three GM hybrids (Table [1](#T1){ref-type="table"}). Open field (summer) production ------------------------------ Summer production was evaluated in an open field trial carried out during the optimal period of eggplant cultivation. Plants were transplanted on May 20^th^ and the last harvest took place on September 11^th^. The early production of the transgenic hybrids, consisting of the first three harvests, was significantly higher than that of the untransformed hybrids (Table [2](#T2){ref-type="table"}): P1 and P10 hybrids yielded, respectively, 36% and 76% more than their corresponding controls, C1 and C10. The difference in productivity between P1 and C1 hybrids, which have long-shaped fruits, was caused by the higher average weight of GM fruits. When comparing P10 and C10 hybrids, which have sub-oval fruits, the higher yield obtained with GM plants during the early harvesting period was due to the increased number of fruits per plant. ::: {#T2 .table-wrap} ::: {.caption} ###### Eggplant production of parthenocarpic hybrids and their respective controls at summertime. ::: Genotypes Early production Total production ----------- ------------------ ------------------ ------- --------- -------- ------- P1 1158a 2.7 a 433 b 3039 a 9.1 a 335 b C1 846 b 2.4 a 344 c 2211 b 7.7 ab 288 b P10 1287 a 2.4 a 553 a 2791 ab 6.8 bc 410 a C10 731 b 1.3 b 541 a 2386 ab 5.7 c 415 a For each trait, at least one common letter indicates no significant difference according to the Duncan test (P = 0.05). Mean value of yields per plant, number of fruits and fruit weight of two transgenic hybrids (P1 and P10) and their corresponding untransformed hybrids (C1 and C10), cultivated during summertime under open field conditions. ::: Consisting of ten harvests, the total production of P1 hybrids was 37% higher than control C1 eggplants (Table [2](#T2){ref-type="table"}). The difference in total yield between P1 and control C1 hybrids was statistically significant and due both to the higher number of fruits/plant and to the increased weight of GM fruits. It is noteworthy to point out that when considered individually, neither trait (number of fruits/plant or fruit weight) showed statistically significant differences between the GM and untransformed plants (Table [2](#T2){ref-type="table"}). Although higher in P10 hybrids in comparison to its control C10, the total yield (the number and average weight of the fruits) was not significantly different between the two. During the whole harvesting period, fruits from both P1 and P10 parthenocarpic hybrids were always seedless (Fig. [2](#F2){ref-type="fig"}), whilst control fruits always had seeds. Therefore, under open field cultivation, the *DefH9-iaaM* gene had a positive influence on fruit quality, as GM *DefH9-iaaM* fruits were always seedless. Fruit quality affects the economic value of eggplant production. {#F2} Although the environmental conditions were optimal and consequently did not affect negatively fruit-set, the *DefH9-iaaM* gene caused both faster development and growth of the fruits as indicated by the increased early-summer production (the first three harvests). In this regard, it is worthwhile to stress that expression of the *DefH9-iaaM* gene takes place in the placenta and ovules before pollen development. As a consequence, in GM parthenocarpic plants fruits are seedless and fruit development initiates well before non-GM controls \[[@B11]\]. In all trials we have never used homozygous lines because growers mostly employ F1 hybrids. The use of hemizygous primary transformants as pollinator plants allowed us to obtain in rather short time, by *in vivo* selection for kanamycin resistance, F1 plants transgenic for the *DefH9-iaaM* gene. Young, healthy and vigorously growing plants did not produce seeds. However, it is possible to obtain seeds from aged *DefH9-iaaM* transgenic plants both by selfing and crossing. By exploiting the delayed female fertility we have produced the homozygous plants needed as male parents for rapid seed multiplication of F1 eggplant hybrids. Therefore, the female sterility of young and mature plants is not an insuperable hindrance for mass propagation and commercial fruit production. Expression of the *DefH9-iaaM* gene takes place during both flower and fruit development in transgenic parthenocarpic eggplants ------------------------------------------------------------------------------------------------------------------------------- The *DefH9* gene is expressed specifically in the placenta and ovules during early phases of flower development \[[@B6]\]. To determine whether expression of the *DefH9-iaaM* gene also occurs during later stages of fruit growth and whether it is influenced by pollen fertilization, mRNA from transgenic flowers and fruits obtained either from emasculated or hand pollinated flowers was analyzed by RT-PCR at different stages of development, until the fruit attained a length of 28 cm. An amplicon of 161 bp corresponding to the spliced *DefH9-iaaM* mRNA was detected in all stages analyzed (Fig. [3](#F3){ref-type="fig"}, lanes 2--6). Thus, the presence of the mRNA of the *DefH9-iaaM* gene and consequently its action is most likely not limited to early stages of flower and fruit development. Pollination did not affect the steady state level of *DefH9-iaaM* mRNA (Fig. [3](#F3){ref-type="fig"}, compare lane 5 and lane 6). {#F3} Treatment with auxin often causes parthenocarpic development in several plant species \[for review, see: \[[@B12]\]\]. However, in some species and/or varieties, to efficiently sustain fruit growth, the hormonal treatment of the flowers must be repeated \[[@B13]\]. The finding that *DefH9-iaaM* mRNA is also present during later stages of fruit development is consistent with the interpretation that in *DefH9-iaaM* parthenocarpic plants, the placenta, the ovules, and the tissues derived therefrom are a source of auxin during the whole growth of the fruit. As a consequence, they efficiently sustain fruit growth. Conclusions =========== The data hereby presented show the positive influence of the *DefH9-iaaM* parthenocarpic gene on eggplant productivity under both greenhouse (spring) and open field (summer) cultivation. Taking into account the data previously obtained under winter greenhouse cultivation \[[@B10]\], we conservatively estimate that the overall increase in productivity is at least 30--35%. The increase in productivity of *DefH9-iaaM* eggplants is mainly due to a drastically improved fruit-set under sub-optimal temperatures and to an enhanced fruit growth and weight. Fruit quality is also improved because the fruits are seedless and do not show a placental cavity. The qualitative improvement of *DefH9-iaaM* eggplant fruits is interesting both for the fresh market and for the processing industry. During early spring greenhouse production, *DefH9-iaaM* parthenocarpic hybrids always gave fruits with an average weight suitable for fresh market commercialization, while untransformed hybrids, under sub-optimal conditions, rarely produced commercial fruits. Thus, the *DefH9-iaaM* gene quantitatively and qualitatively improves eggplant production under both greenhouse and open field cultivation. In all genotypes tested the *DefH9-iaaM* gene had a very positive effect on production and quality parameters. Such findings are of paramount importance as the hybrids tested have the same genetic background that the relative controls, except for the presence of the *DefH9-iaaM* gene. The *DefH9-iaaM* gene, already known to be expressed in the placenta and ovules during early phases of flower development, is expressed also in mature fruits, most likely in tissues derived from the ovules. From an economical standpoint, the main advantages conferred to eggplant by the *DefH9-iaaM* gene are: i) production of marketable fruits under environmental conditions adverse for fruit-set and growth; ii) reduction of cultivation costs (energy, phytohormones and labor) necessary for off-season and open field eggplant cultivation; and iii) enhancement of fruit quality. Last but not least, contrary to conventional wisdom, these advantages have been achieved without the use of either male or female sterility genes. Materials and methods ===================== Greenhouse spring cultivation ----------------------------- Trials were carried out in central Italy (Monsampolo del Tronto-AP) and in southern Italy (Pontecagnano-SA) (approval of the Italian Ministry of Health N° B/IT/97-29). The greenhouses were rather similar and made of galvanized steel and covered with plastic polyethylene (0.12 mm thick). An apparatus for drip-irrigation was used and the soil was completely mulched. A complete randomized block design with three replicate hybrid genotypes was adopted. Each experimental plot measured 3.12 m^2^ and contained eight plants in a double row. Transplanting was performed on March 3^rd^ in southern Italy and on March 27^th^ in central Italy. The P1, P2, P5, C1, C2 and the commercial Talina hybrids were employed. Transgenic parthenocarpic hybrids P1, P2 and P5 were obtained by crossing (as male parent) the primary transgenic plant DR2 *iaaM* \#34-1 with the line Tal 1/1 (P1), the primary transgenic plant DR2 *iaaM* \#28-1 with Tal 1/1 (P2) and the transgenic plant Tal 1/1 *iaaM* \#1-1 with the line Tina (P5). The hybrids P1 and P2 are homologous to C1 (DR2 × Tal1/1), except for the presence of the *DefH9-iaaM* gene integrated in their genome. The transgenic hybrid P5 is homologous to its untransformed control C2 (Tal1/1 × Tina). DR2 and Tina are parental lines obtained through classical breeding, Tal1/1 is a double haploid line derived from anther culture of the F1 commercial cultivar Talina. The segregation of the marker gene *nptII* was checked by spraying the plants with kanamycin \[[@B14]\] and allowed for the conclusion that the transgenes segregate as a single locus in the backcrossed progenies of the three independent events analyzed (Tal *iaaM* 1-1: χ^2^ = 0.01065, P = 0.917; DR2 *iaaM* 34-1: χ^2^ = 0.0496 P = 0.824; DR2 *iaaM* 28-1 χ^2^ = 0.06467 P = 0.799). Southern blot analysis showed that DR2 *iaaM* 28-1 and 34-1 had a single copy of the transgene, while Tal *iaaM* 1-1 had three copies of the transgene (Fig [4](#F4){ref-type="fig"}). Since the interaction genotype/location was not significant for the yield, the data were computed as average of the two locations and subjected to analysis of variance according to a randomized complete block design. Duncan\'s Multiple Range Test (P = 0.05) was used for means separations. {#F4} Open field (summer) cultivation ------------------------------- The open field trial was carried out under open field conditions at Monsampolo del Tronto (approval of the Italian Ministry of Health B/IT/99/21). Two transgenic parthenocarpic hybrids were tested: the hybrid P1 (Tal1/1 × DR2 *iaaM* \#34-1) with elongated fruits and the hybrid P10 (UGA × Tal1/1 *iaaM* \#1-1) with sub-oval fruits were compared to their homologous non-transgenic controls C1 (DR2 × Tal1/1) and C10 (UGA × Tal1/1), respectively. The UGA line has oval dark purplish fruits and it has been provided by Dr. S.C. Phatak. A complete randomized block design with the hybrids replicated four times was adopted. Each experimental plot measured 11.7 m^2^ and contained 30 plants in a double row. Transplanting was performed on May 10^th^. Early spring production consisted of the first four harvests (i.e. 4 out of 16 harvests during the whole production cycle), while early summer production, whose cultivation cycle consists of ten harvests, corresponds to the first three harvests. For all trials the number and weight of fruits were recorded. In addition, fruit sample for each harvest and replication was cut to check for the presence of seeds. Data were computed for the early harvesting time and for the whole harvesting season. Analysis of variance was performed according to a randomized complete block design. Duncan\'s Multiple Range Test (P = 0.05) was used for means separations. Plant DNA isolation and Southern blot analysis ---------------------------------------------- High-molecular-weight DNA was isolated from the young leaves of transgenic and untransformed eggplants by using Plant DNAzol (Invitrogen), according to the manufacturer\'s instructions. Ten μg of DNA was digested overnight with 80 units of *KpnI* (Promega) in a volume of 500 μl, separated on a 0.7% agarose gel and transferred to Hybond N (Amersham Pharmacia Biotech). A 1350 bp fragment of the *DefH9* gene was used as a radiolabeled probe. The membrane was hybridized overnight in 5X SSC/50% formamide (Sigma) at 42°C and washed two times for 15 min. in 2 × SSC/0.1% SDS, and two times for 15 min. in 0.1 × SSC/0.1% SDS at 42°C. Signals were detected using Kodak X-OMAT AR5 film (Sigma). RT-PCR analysis --------------- Semiquantitative (competitive) PCR analysis was carried out for 38 cycles (annealing temperature 63°C) using as template cDNA (8 ng) obtained by priming poly(A)^+^ mRNA with an *iaaM* specific primer (5\'-AATAGCTGCCTATGCCTCCCGTCAT-3\'). The mRNA was extracted from either young flower buds (5,8,11 mm) or eggplant fruits (placental tissue from fruits either 40 or 280 mm long). As an internal standard, 0.5 fg of a 600 bp long *DefH9* cDNA fragment was used in the PCR assay. To amplify the 161 bp long amplicon an *iaaM* specific primer (5\'-GGGTGAATTAAAATGGTCATACAT-3\') and a *DefH9* specific primer (5\'-CTTTGGAACTCGTGTTGAGCTCTCA-3\') were used. For the internal standard, a 3\' primer (5\'-TGAGCATTGATCTCCTGAGTGGTGT-3\') together with the *DefH9* specific primer were used to produce the 351 bp long amplicon. The PCR assays were performed with a thermostable DNA polymerase mixture (Expand High Fidelity PCR system, Roche) in presence of 3 μCi of ^32^P dCTP. The intensity of the bands was quantified by using an Instant Imager (Packard, Meriden, CT). Acknowledgements ================ This research was partially supported by the Consiglio Nazionale delle Ricerce (Progetto Finalizzato Biotecnologie II) and by the Ministero Politiche Agricole e Forestali Progetto \"Biotecnologie vegetali\". We thank Prof. Phatak, University of Georgia (USA) for the UGA line.

PROFESSOR JEAN DAUSSET\'S EDITORIAL IN ENGLISH ============================================== Why a new medical journal? Why especially this new journal? What are its ambitions? What audience can it reach? Specialists, generalists or both simultaneously? It is intended to be a link between the international scientific community, in particular within the fields of life science and its technological application. There is undoubtedly a niche to be occupied, an important role to be played to better link this community. I am delighted by the initiative taken by the promoters of this journal, who had the courage to enter the competitive field of scientific journals and to launch an exciting intellectual adventure. For my part, I would like to briefly broach some medical questions that are dear to me, and that are sufficiently general to be the subject of an editorial. Where indeed is medical practice going and what does the medicine of tomorrow hold for us? We have watched, amazed, the fantastic advances that are progressively entering current medical practice. The revolution we have witnessed has names, in particular: vaccination, developed by Pasteur, which still has new fields to conquer (until now there is no vaccine against malaria or against AIDS!). There is also antibiotic therapy, exploited to full advantage but beginning to run out of steam, showing its limitations because of the rapid rise of resistant microorganisms. It is only a few decades since developments in immunology have allowed the transplantation of organs or bone marrow. Transplantation---an old dream of humanity finally realized---whose successes themselves have led to many disappointments among the very many patients on the waiting lists, because of the shortage of the organs to be transplanted. It is nevertheless admirable that, in the whole world, 6 million individuals are ready to give their marrow to save patients, in particular those with leukemia. To this list of successes we should add the many victories in oncology due to improved monitoring of tumorigenic agents? And finally, we must not forget the methods of biological investigation and medical imaging, which have seen great improvements recently. This panorama would be quite incomplete if I did not add the spectacular rise of medical genetics, which makes it possible to locate and to clone the genes responsible for many monogenic diseases and, in the near future, the genes that act in concert to produce the many polygenic diseases that afflict our industrialized societies. Provided with this arsenal, the doctor, who is a specialist or general practitioner, can from now on offer to his patients increasingly effective therapies, often leading to the cure or amelioration of chronic diseases. Longevity continues to increase each year in the industrialized countries. It should be the same everywhere. I initiated, many years ago, the concept of predictive medicine, based on frequent associations between some diseases and the alleles of genes of the HLA system. For example ankylosing spondylitis, associated with HLA-B27, or insulin dependent diabetes, associated with HLA-DR3/DR4. This concept has now been extended to the entire human genome and the number of genes of susceptibility to (or protection from) many diseases can only increase. Apart from the HLA system, we know of susceptibility to familial breast cancer, given by genes such as BRCA1 and BRCA2. This idea can only grow in importance, but there is a danger. There can be no question of detection of susceptibility genes without simultaneous development of a preventive or therapeutic counterpart. We should not, under any circumstances, warn a healthy person of an uncertain future, because in this case---and I emphasize this point---we are dealing with a risk of uncertain proportions, without a corresponding method of prevention or cure. It would be unforgivable to cause useless anxiety. Whatever reservations we might have, the future direction is now clear. There are more and more specialties where tests of increased susceptibility to disease exist. These must be used for the benefit of the patient and not to his detriment. The physician of the 21st century, if he or she applies these new concepts with prudence, could become the adviser of healthy patients, who are no longer ill. Each person will be able to manage his health capital, as he manages his banking capital; with full knowledge, guided by his doctor, of the developments of modern genetics. Is it not interesting to note that medicine was, in its history, first of all curative, then preventive and finally predictive, whereas today the order is reversed: initially predictive, then preventive and finally, only in desperation, curative? Professor Jean Dausset Nobel Laureate

Background ========== Reused and poorly sterilised equipment causes infection in developing countries. For example, unsterile injections result in 8 to 16 million new infections of Hepatitis B every year, according to one recent estimate \[[@B1]\]. Preventing disease transmitted by medical equipment in developing countries is a priority for health care providers, and global policy agencies such as the World Health Organization (WHO) and the United Nations Children\'s Fund (UNICEF) seek the most appropriate strategies to achieve sterile injections and surgical procedures. For vaccinations, the WHO\'s and other agencies current strategy is to recommend auto-disable syringes for immunisation, and they aim for universal implementation of this by 2003 \[[@B2]\]. However, this strategy is not yet proven, the cost is high, delivery is not assured, and how they should be disposed of is not established. Some believe providers will continue to re-use injection equipment, and this is certainly likely where provision of auto-disable syringes are not assured. In addition, auto-disable technologies do not address the need for effective decontamination of equipment used in other procedures, such as general surgery, abortion and endoscopy. Therefore, policies for decontamination of medical equipment are still required and it is important that these are grounded in good evidence that they are effective. As a rule, manufacturer\'s instructions for decontamination of equipment should always be followed but there may be many cases where this information is not available. The World Health Organisation reviewed their current decontamination guidelines and reported their findings in a document, which has not yet been published (N Sang,. O Kasiolo: Presentation of inconsistencies found in WHO documents. *World Health Organisation (unpublished)* 1997). From eight sets of different recommendations, they identified inconsistencies. For example, guidelines gave a range of different temperatures and times for steam and dry heat sterilisation; and \'bleach\' was recommended in some guidelines, and not in others (N Sang,. O Kasiolo: Presentation of inconsistencies found in WHO documents. *World Health Organisation (unpublished)* 1997). In addition, some WHO publications recommended alcohol and povidone iodine for equipment \[[@B3]\] but these agents are not included in other published hospital guidelines \[[@B4]\]; and boiling (generally considered not to kill bacterial spores) is recommended for sterilisation of equipment \[[@B3],[@B5]\]. The evidence base for the WHO guidelines is not explicit, but this is not unusual in this field. There are many guidelines and recommendations for effective decontamination practice including reports \[[@B6]\], government documents \[[@B7]\] and field manuals \[[@B5]\], with no reference to the data on which the recommendations are based. We sought to synthesise current available evidence for the effectiveness of various procedures using explicit, pre-specified methods to help inform policies relevant to primary care settings in developing countries. As no explicit methods are recognised for this, we first developed the methods for synthesising this type of data. This paper describes the methods we used and illustrates how the synthesis works in practice, using the example of alcohol in relation to lipid viruses. We also summarise the findings from the analysis of bleach (sodium hypochlorite and similar chemicals), alcohol and povidone iodine. The full review of decontamination procedures can be downloaded from the Liverpool School of Tropical Medicine web site (Systematic Review of Medical Equipment Decontamination Practices -- <>). Methods ======= The Methods section is divided into three distinct sections: Inclusion Criteria, Search Strategy and Analysis. Studies were selected for inclusion in the review according to several explicit and pre-defined criteria. These criteria are described in detail here under headings of outcomes measured, types of study and the decontamination procedures under review. Following this, we describe the search strategy developed to identify all published studies that may meet these criteria. Finally, identified studies were assessed in detail to ascertain whether or not they actually met the inclusion criteria and the findings of each one summarised, compared and presented. We describe here the methods of analysis we used for this process and the criteria we used to compare studies. 1. Inclusion criteria --------------------- ### Outcomes The outcomes sought were those relating to equipment decontamination. This is not easy to quantify. From the UK Department of Health guidelines, decontamination is defined as: \'A process which removes or destroys contamination (defined as soiling or pollution with harmful, potentially infectious or other unwanted matter) and thereby prevents microorganisms or other contaminants reaching a susceptible site in sufficient quantities to initiate infection or any other harmful response. Three processes of decontamination are commonly used: sterilisation, disinfection and cleaning\' \[[@B7]\]. Decontamination procedures are defined by the types of microorganisms they kill or remove. Thus, a sterilising agent destroys all microbial life and a high-level disinfecting agent kills all microorganisms except large numbers of bacterial spores\[[@B4]\]. Microorganisms differ in their resistance to decontamination and we used indicator microorganisms from the groups shown (Table [1](#T1){ref-type="table"}) to represent the huge variety of potential contaminants. These indicators were chosen according to their particular resistance to decontamination (for example, picorna viruses) but some were also included on account of their particular medical significance in the area of equipment contamination (for example, HIV). The only decontamination outcome sought was the destruction of indicator microorganisms, demonstrated by a) negative microorganism culture *in vitro*, b) failure to demonstrate their viability in animal models or c) a negative result using sensitive molecular methods. ::: {#T1 .table-wrap} ::: {.caption} ###### Indicator microorganisms included as representative markers of the organism groups shown ::: **GROUP** **INDICATORS** --------------- ------------------------------------------------------------------------------------------------------------------------ **Bacterial** **Spores**: *Bacillus sp., Clostridium* **Tuberculous Mycobacteria**: (*M. bovis* &*M. tuberculosis*) **Non-Tuberculous Mycobacteria**: (other *Mycobacteria*) **Vegetative**: *Pseudomonas, Enterobacteria, Staphylococcus* **Viral** **Nonlipid (small)**: picorna viruses (e.g. Polio) **Lipid (medium)**: HIV, Hepatitis B, herpes viruses **Fungal** (e.g. *Trichophyton, Candida, Aspergillus*) **Other** CJD ::: It is clear that the ability to decontaminate equipment is dependant upon the initial level of contamination and we therefore needed a quantifiable benchmark against which to compare studies. The British standard protocol for testing disinfectants is the ability to show a 5log~10~ reduction in viable microorganisms (BS EN 1040: 1997 Chemical disinfectants and antiseptics. Basic bactericidal activity). We took this standard as a practical minimum requirement for studies to be included, considering it a reasonable level of kill while excluding as few relevant studies as possible. ### Types of study Observational studies with microbiological outcomes in the laboratory and field. Studies of in-use procedures, procedures experimentally applied to indicator microorganisms and applied to artificially contaminated equipment were included. Studies written in any language other than English and studies prior to 1976 (the date from which electronic abstracts are included in Medline) were excluded. Given the volume of studies retrieved from the database, we considered it impractical to include data from earlier studies and studies that would need detailed translation. ### Procedures included We included decontamination procedures likely to be available in resource-poor settings and effective for the decontamination of \'critical\' and \'semicritical\' equipment (Table [2](#T2){ref-type="table"}). Critical and semicritical are terms used to describe equipment\'s need for decontamination and are linked to the subsequent use of the equipment. Thus, equipment that is used in normally sterile areas of the body (e.g. needles that penetrate the skin) has a \'critical requirement\' for decontamination and is described as critical equipment. Semicritical equipment does not penetrate skin or other tissue and the requirement for decontamination is classed as less (e.g. specula). To be effective for critical equipment, a procedure must destroy all microorganisms, but for semicritical equipment, all microorganisms except large numbers of bacterial spores should be destroyed \[[@B8]\]. Procedures requiring specialised apparatus or conditions (such as irradiation, ethylene oxide, gaseous formaldehyde and peracetic acid washer/disinfectors) were not looked at. Low-level disinfectants and antiseptics (such as quaternary ammonium compounds and phenolics) were not included, as these are not effective for critical and semicritical equipment. ::: {#T2 .table-wrap} ::: {.caption} ###### Procedures used for the decontamination of medical equipment and included in the review ::: ------------------------------------------------------ **Heat-based** **Chemical** ----------------------- ------------------------------ Dry heat Bleach (sodium hypochlorite) Pressurised steam Formaldehyde Boiling and hot water 2% Glutaraldehyde\ hydrogen peroxide\ alcohol\ povidone iodine ------------------------------------------------------ ::: 2. Search strategy ------------------ A search strategy for Medline was developed with advice from colleagues in the Cochrane Infectious Diseases Group in Liverpool. The full search strategy is listed in Table [3](#T3){ref-type="table"}. We used database-defined subject headings for decontamination procedures and indicator microorganisms. The terms used for procedures were \'*sterilization\'* and \'*disinfectants\'*. We combined the exploded versions of these terms (incorporating all terms listed as sub-headings) with specific procedure sub-headings such as *\'alcohols\'*. Some areas of disinfectant use were not related to medical equipment and these were excluded from the search using the exploded headings \'*histocytological preparation techniques\'*, \'*handwashing\'* and \'*sanitary engineering\'*. ::: {#T3 .table-wrap} ::: {.caption} ###### Sequence of search expressions used in Medline (1976-April 2000) as the search strategy ::: **No.** **Search Expression** --------- ------------------------------------------------------------------ 1 exp disinfectants/ 2 exp hypochlorite/ or bleach.tw 3 exp formaldehyde/ 4 exp glutaraldehyde/ 5 exp hydrogen peroxide/ 6 exp alcohols/ 7 exp povidone iodine/ 8 or/2--7 9 exp sterilization/ 10 9 and (steam or heat or boil\$).tw 11 (1 and 8) or 9 or 10 12 exp histocytological preparation techniques/ 13 exp handwashing/ 14 exp sanitary engineering/ 15 11 not (12 or 13 or 14) 16 exp bacillaceae/ 17 exp mycobacterium/ or mycobacterium.tw 18 exp pseudomonas/ or pseudomonas.tw 19 exp staphylococcus/ or staphylococcus.tw 20 exp enterobacteriaceae/ 21 exp fungi/ 22 exp hiv/ or hiv.tw 23 exp hepatitis viruses/ or hepatitis.tw 24 exp picornaviridae/ 25 exp herpesviridae/ or herpes.tw 26 exp creutzfeldt-jakob syndrome/ or creutzfeldt-jakob syndrome.tw 27 exp cryptosporidium/ or cryptosporidium.tw 28 or/16--27 29 15 and 28 ::: The terms used for indicator microorganisms were those listed for each microorganism group. These were also exploded to include not only the relevant microorganisms but also any surrogates that may have been used experimentally. Text words for particular organism genera or diseases (e.g. *Herpes*) were also included in the search where relevant. 3. Analysis ----------- ### Assessment of studies for inclusion and quality We obtained and assessed the full text of all relevant papers from the search. A table of excluded studies and the reasons for exclusions are in the full report. The most common reason for exclusion was that the number of microorganisms exposed to decontamination was not measured, meaning no comparative analysis was possible. However, where the outcomes of these studies provide useful information about the procedures, we refer to them. The quality of included studies was assessed using criteria of a) quality of the procedure description, b) quality of the outcome description, c) quality of the study design and d) the number of observations for each recorded result. The form used is shown in Figure [1](#F1){ref-type="fig"}. Scores (out of a possible three) were given for each of these criteria and the total used to assign comparative reliability grades for each study. These grades were not definitive but allowed the comparison of studies within this review. {#F1} ### Data extraction We extracted the specific details of each procedure (such as time of exposure, concentration of chemical agent) and the level of kill it achieved for each indicator microorganism. In addition, we sought critical points of procedures, defined as procedural details that might explain variations in effectiveness of the same procedure when used in different studies. These we extracted systematically, and we particularly sought for details about: a\) The form of the microorganisms used in the study (for example, were they in solution or dried to equipment?) b\) the method of outcome assessment (for example, was it measured by culture or molecular means?). c\) the study design (for example, did the study include controls?). d\) Analysis of organic load (for example, was organic load present or its effect tested?). ### Data analysis An example of the analysis of extracted data is shown in Figure [2](#F2){ref-type="fig"}. The results were grouped within procedures by the organism tested for, separating procedures between critical and semi critical equipment. Thus, studies using bacterial spores or CJD as indicator microorganisms were grouped as evidence for critical equipment. Studies using any other microorganisms were grouped as evidence for semicritical equipment. Each procedure was treated separately and so where one publication tested a range of procedures in the same experiment, each test was counted as a separate study. {#F2} The studies identified used a wide range of different procedural specifications such as time of exposure to agent. To synthesise this information in a sensible way, we built up a representative set of specifications that were described as effective to destroy the particular indicator microorganisms. The specifications required to kill the most resistant microorganism were not always representative of the majority of studies. Therefore, the specifications effective for the majority of studies were chosen and each study classed as evidence for or against this standard. This process allowed flexibility to iteratively develop the standard specifications according to the findings of subsequent studies. Once we had classed each study, we summarised in an adjacent column the possible reasons for disagreement between studies showing a procedure was effective, and those showing it was not effective (\"possible causes of heterogeneity\" in Table A, shown in Figure [2](#F2){ref-type="fig"}). In this way, major factors influencing procedure efficacy became clearer and these often coincided with the critical points described above. Some included studies showed results that could not be described in these terms of \'evidence for\' or \'evidence against\' (for instance, during a time course of exposure to disinfectant, the period of measurement may stop short of the standard time specification iterated). These studies nevertheless contained valuable evidence or results that could be reasonably extrapolated and they are described and referred to when the effectiveness of a procedure is not clear from other studies. ### Presentation of results For each procedure and level of decontamination, minimum conditions for effective decontamination were presented. These conditions are a product of the iterated standard specifications and any secondary outcomes that were shown to have a critical impact on the effectiveness of the procedure (critical points). An assessment of the evidence and the reviewer\'s conclusions were also presented. Results ======= Out of 3396 records that were retrieved using the search protocol in the Medline database (from 1976-April 2000), 88 published papers met the inclusion criteria for the review. Within these papers there were 135 studies of the effectiveness of the chosen procedures against the indicator microorganisms (Table [4](#T4){ref-type="table"}). The full results can be downloaded from the Liverpool School of Tropical Medicine website <>. We present here summaries of the results for alcohol, bleach, and povidone iodine. ::: {#T4 .table-wrap} ::: {.caption} ###### Description of included studies, showing numbers describing the different decontamination procedures included in the review ::: **Procedure** **Total no. of studies** **Equipment level** ----------------------- -------------------------- --------------------- ---- **Dry heat** 2 2 0 **Pressurised steam** 8 6 2 **Boiling** 0 0 0 **Bleach** 33 10 23 **Formaldehyde** 8 3 5 **Glutaraldehyde** 39 12 27 **Hydrogen peroxide** 12 3 9 **70% alcohol** 20 0 20 **Povidone iodine** 13 0 13 ::: Alcohols -------- The effectiveness of alcohols was assessed in 23 of the included studies. None of these studies assessed alcohol\'s effectiveness for critical equipment. For semicritical equipment, the standard procedure derived from the included studies was exposure to 70--80% ethanol or isopropanol for 5 minutes. Alcohols were particularly effective for lipid viruses (such as HIV) \[[@B9]-[@B11]\] and bacteria \[[@B12]-[@B14]\] but non-lipid viruses (such as hepatitis A) and mycobacteria may be more resistant, especially when dried to equipment\[[@B15],[@B16]\]. There was not enough evidence to give a more definite assessment. The presence of organic load did not impair effective decontamination with alcohol except in cases where contaminating material (such as dried serum) limited its access to microorganisms\[[@B10]\]. Results for alcohols applied to lipid viruses are shown in Figure [2](#F2){ref-type="fig"}, to illustrate the application of the review methods described. Bleach ------ The effectiveness of bleach was assessed in 37 of the included studies; 10 effective for critical equipment and 27 for semicritical equipment. Comparison of these studies was made difficult by the different nomenclature used for \'bleach\'. As described, our definition of bleach includes sodium hypochlorite and similar chemicals. Bleach activity is regularly described in terms of concentration of available chlorine (% or mg per litre), parts per million (ppm) and ppm available chlorine. The most consistent terminology in the included studies was ppm available chlorine (used in 57% of papers identified) and the results are presented in these terms (for ppm, read ppm available chlorine). Where concentrations of available chlorine are quoted as percentages or mg per litre, these were converted to ppm for consistency. A formulation of 5,000 ppm available chlorine is equivalent to a 0.5% solution (of available chlorine). For critical equipment, a standard procedure of 5000 ppm sodium hypochlorite bleach for 5 minutes (pH = 7--8) was sufficient to kill bacterial spores. For semicritical equipment, the standard procedure of 1000 ppm for 10 minutes was found to be sufficient. From the critical points described, effective decontamination is compromised by the presence of various organic matter \[[@B17]-[@B19]\], a very large number of microorganisms \[[@B20]\] and situations where microorganisms are dried to equipment \[[@B15]\]. Mycobacteria and viruses are particularly resistant to the standard procedure for semicritical equipment when they are dried to a surface. These factors highlight the necessity of cleaning equipment before using hypochlorite and this is also a critical point that is described \[[@B21]\]. In addition, it should be remembered that bleach can be corrosive to metal equipment and this may be an important consideration for its use. Povidone iodine --------------- The effectiveness of povidone iodine was assessed in 15 of the included studies. As with alcohol, none of these studies assessed its effectiveness for critical equipment. For semicritical equipment, the standard procedure of a concentration giving 1% available iodine for 15 minutes was found to be sufficient. Bacteria and mycobacteria were particularly susceptible to this procedure but viruses may be more resistant \[[@B22],[@B23]\]. There were not enough studies using viruses to determine a definite effectiveness or non-effectiveness for povidone iodine. Discussion ========== The method ---------- Decontamination guidelines often give no reference to the source of their recommendations \[[@B7]\] and very rarely justify or qualify any evidence that is cited \[[@B4]\]. The evidence is difficult to assess, therefore, where inconsistencies occur between guidelines (N Sang,. O Kasiolo: Presentation of inconsistencies found in WHO documents. *World Health Organisation (unpublished)* 1997). The review described is an attempt to systematically gather the relevant evidence for effective decontamination and reproducibly assess it so that decisions may be made as to the most relevant procedure to use. We have developed methods to assess and analyse observational data where there is little standardisation in outcome measurement between studies. The novel means by which we have analysed the quality of this data has allowed the comparison of studies that lack the consistency and robustness of randomised controlled trials. The protocol we have used is also flexible enough to allow the review to be updated as more studies become available, much like a Cochrane Review. This approach may have applications in other areas of guideline appraisal and could be developed to provide a solid evidence base for other practices that may have been adopted. This report summarises evidence systematically gathered for bleach, alcohol and povidone iodine. There is good evidence for the effectiveness of bleach as a decontaminating agent for both critical and semicritical equipment and the standard procedural specifications derived using our methods support WHO guidelines. Similarly, the findings for alcohol also agree with WHO publications, though it is not recommended to the same extent as bleach. The exclusion of povidone iodine from the majority of decontamination guidelines (N Sang,. O Kasiolo: Presentation of inconsistencies found in WHO documents. *World Health Organisation (unpublished)* 1997) is supported by the evidence here, since the iodophor has uncertain activity against viruses. In order to manage the data and provide a valid analysis, certain stated bias was included. The first bias was using a cut-off time for inclusion of studies. As already described, 1966 to the present day is the limit of electronic databases and we limited this further to only include studies post-1976. This was mainly for practical reasons, to keep the volume of studies manageable and also to limit chasing irrelevant studies whose titles sound full of promise but for which no abstract is provided (abstracts are not included for pre-1976 records in Medline). The introduction of language bias was also for practical reasons. Many studies in languages other than English were potentially eligible for inclusion but they would have required very detailed translation in order to assess the precise de contamination methods used. Despite these limits, we consider that the number of studies retrieved were sufficient to develop representative specifications. One of the main areas for exclusion of relevant studies was their quantification or description of indicator microorganism destruction. In order to compare the potency of decontamination procedures between studies, we set the ability to show at least a 5log reduction in microorganisms as an inclusion limit. In reality, effective decontamination is achieved when there is no infection passed on and this will be influenced by the infective dose of the contaminating microorganism and the susceptibility of the host, as well as the initial level of contamination. These factors will be different for every situation and so decontamination, as such, cannot be quantified in a laboratory. Setting a minimum level of demonstrable destruction allows the comparison of data from diverse studies and is simply a tool to aid the analysis of data in this review. Setting such a limit obviously excluded studies that were otherwise relevant and, it could be argued, introduced misleading bias. Where there was heterogeneity in the evidence for a particular microorganism/procedure combination, however, these excluded studies were checked for data that may influence the outcome. In every case, such data did not challenge the standard specifications iterated from the included studies, suggesting this was a valid and useful cut-off point to incorporate. The explicit methods used to search for studies and assess them for inclusion enable this review to be updated. The methods we have used to analyse the procedural specification data also allow updating of the results by further iteration. By grouping evidence as for and against the standard set, reasons for disagreement between new studies and the standards developed in this review should be immediately apparent. Though the levels of evidence for each procedure were mixed, the results of this review agree with current decontamination guidelines, often reached by consensus and not systematically. This approach, therefore, provides a means by which observational and non-standardised data can be systematically summarised to give formal evidence for guidelines. The implications of the results for the WHO are to standardise decontamination guidelines across the various arms of the organisation. These methods of systematic review and analysis may enable evidence-based guidelines to be made in similar areas where data cannot be pooled or assessed using standard methods. Conclusions =========== Comparing the evidence with guidelines -------------------------------------- The standard procedures iterated for bleach differed from the WHO guidelines that included this agent (N Sang,. O Kasiolo: Presentation of inconsistencies found in WHO documents. *World Health Organisation (unpublished)* 1997) but did not contradict them. Guidelines only gave a recommendation for critical equipment and this was neat bleach (generally 52,000 ppm) for 30 seconds or 5000 ppm bleach for 10 minutes. Procedures for alcohol and povidone iodine are not described in the guidelines reviewed (N Sang,. O Kasiolo: Presentation of inconsistencies found in WHO documents. *World Health Organisation (unpublished)* 1997) but are mentioned elsewhere in WHO publications. The procedures derived here show some scope for their use as disinfectants but there is not enough available evidence to demonstrate their effectiveness against all the indicator microorganisms. In recommending alcohol, one WHO publication describes it as not effective against non-lipid viruses and this is supported by the evidence here\[[@B3]\]. The same publication recommends iodophors as effective for both non-lipid viruses and bacterial spores but the evidence for povidone iodine here disputes this. There are several considerations to be made when choosing a particular means of decontamination and this review focuses only on the microbiological effectiveness of agents. Other factors include cost and the compatability of decontamination agents with particular pieces of equipment. Manufacturer\'s guidelines for decontamination should always be followed where possible as these will take into account risks of equipment corrosion through inappropriate decontamination. Critical points of the procedures --------------------------------- The factors that impacted the effectiveness of procedures most frequently in the review were the presence of organic material and the drying of microorganisms to a surface before decontamination. These factors may either neutralise the potency of the decontamination agent being used or protect the target microorganisms from exposure to the agent. Thus, the critical point common to the majority of the procedures was the need to clean equipment before decontamination, and this is specifically mentioned or quantified in many of the studies. This demonstrates that, for the procedures described here, the application of a decontamination procedure is as important as the choice of procedure. Competing interests =================== None declared Authors\' contributions ======================= WS wrote the protocols, identified the studies, conducted the data extraction, and synthesised the data. PG identified the topic and advised on methods development, data extraction, interpretation and writing of the paper. TH helped identify the relevant questions, outcomes and indicators and advised on the literature search. All authors contributed to revisions and finalising of the paper and the report. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ A panel of specialists helped at the protocol and review stage by providing important technical advice and suggestions. We are very grateful for their input. They were: Barbara Stilwell (WHO), John Lloyd (PATH), Mary Catlin (PATH), Anthony Battersby (FBA systems), Craig Mackintosh (Clatterbridge Hospital). The panel are not responsible for the content, and the data presented and the views expressed are entirely those of the authors. This research was supported by the Department for International Development (UK) through their support to the Effective Health Care Alliance Programme based at the Liverpool School of Tropical Medicine.

Background ========== DNA of prokaryotic and eukaryotic cells and their viruses is often modified by methylation, carried out by S-adenosyl-L-methionine (AdoMet)-dependent DNA methyltransferases (MTases). Since a particular nucleotide sequence may exist in its methylated or unmethylated form, methylation can be regarded as an increase of the information content of DNA, which serves a wide variety of biological functions. In Eukaryota, DNA methylation plays a role in crucial regulatory processes, such as regulation of gene expression, embryonic development, genomic imprinting, and carcinogenesis (reviewed in ref. \[[@B1]\]). In Prokaryota, DNA methylation can be involved in DNA mismatch repair, regulation of gene expression, and control of timing of DNA replication (reviewed in ref. \[[@B2]\]). However, the majority of prokaryotic MTases are paired with a restriction endonuclease of cognate sequence specificity, together forming restriction-modification (RM) systems. RM systems are thought to serve as defense mechanisms that protect the cell against invasion of foreign genetic elements such as phages and plasmids \[[@B3]\]. It has been also suggested that RM systems are maintained in evolution because they participate in generating bacterial diversity by promoting homologous recombination \[[@B4]\] or because they act as as \"selfish\" genetic elements that undergo extensive horizontal transfer \[[@B5]\]. These three hypotheses are contrasting, but not mutually exclusive. MTases can be divided into three different groups on the basis of the chemical reactions they catalyze: generating N6-methyladenine (m^6^A), N4-methylcytosine (m^4^C), and C5-methylcytosine (m^5^C). It has been suggested that m^4^C and m^6^A MTases (collectively termed \"N-MTases\") may be more closely related to each other than to m^5^C MTases \[[@B6]\]. Nevertheless, subsequent analyses showed that the relationships between these groups of proteins are quite complicated and their evolution may have involved several independent conversions of the reaction specificity \[[@B7]-[@B9]\]. Amino acid sequence alignments of DNA MTases revealed several conserved motifs, of which I-VIII and X are common to most subfamilies, and a region of essentially higher variability \[[@B6],[@B10]\]. Based on the results of X-ray crystallography of members of all three groups and structure-based multiple sequence alignment, motifs IV-VIII were assigned to the active-site subdomain, motifs X and I-III to the AdoMet-binding subdomain, and the variable region was recognized as a separate domain, implicated in recognition of the target sequence (and termed TRD for target-recognition domain) (reviewed in refs. \[[@B11],[@B12]\]). Structural studies on m^5^C, m^4^C, and m^6^A MTases demonstrated that the TRDs of these proteins are structurally dissimilar and most likely were acquired in independent gene fusion events \[[@B11]\]. DNA MTases have been subdivided into 6 classes (α, β, γ, ζ, and the hypothetical δ and ε ; Figure [1a](#F1){ref-type="fig"}) according to the possible linear arrangements of three modules: the AdoMet-binding subdomain, the active site subdomain, and the variable TRD \[[@B6]\]. All of the enzymes preserve the same spatial arrangement of the motifs and α, δ, and ε form a circularly permuted set as does β, γ, and ζ (Figure [1b](#F1){ref-type="fig"}). The majority of DNA N-MTases fall into the α, β, and γ classes, with no *bona fide* γ-m^4^C MTases known. M.*Ngo*MXV \[[@B13]\] and its close homolog M.*Lmo*A118I \[[@B14]\] are the only experimentally characterized m^4^C MTases, whose architecture is very similar to γ-m^6^A MTases. Nevertheless, these remarkably small MTases (153 aa), as well as their uncharacterized homologs identified in sequence databases, lack both the classical TRD and the region corresponding to motif X and therefore can be regarded as \"minimal\" members of the family \[[@B13]\]. Another group of small (\~175 aa) MTases, which are similar to γ-m^6^A MTases but lack motif X and the TRD, are the atypical Dam (m^6^A) MTases encoded by several phages, including HP1, VT-2, and T1 \[[@B15]\]. ![Conserved fold and variable topology of the MTase domain, **a)** The linear organization of six classes of amino-MTases (α-ζ) postulated by Malone et al. \[[@B6]\], m^5^C MTases (the prevailing archetypal topology labeled as m^5^C, and the two minor classes as ζ and DRM2), and the minimal MTases lacking the discrete TRD. The AdoMet-binding region is shown as a blue arrow, the catalytic region is shown as a red arrow, conserved motifs are labeled accordingly **b).** The topology diagram: triangles represent β-strands, circles represent a- and 3~10~-helices, connecting lines represent loops; the thick lines correspond to the loops on the catalytic face of the protein, which harbor residues involved in cofactor binding and catalysis and most likely in DNA-binding. Elements forming the AdoMet-binding pocket are colored blue; elements forming the target base-binding/catalytic pocket are colored red. Circled Roman numerals represent nine motifs, the key motifs I and IV are shown in bold and underlined. Orange arrows show the topological breakpoints (N/C for generation of N- and C-termini) and sites of TRD insertion characteristic for the individual classes of MTases. The elements characteristic for the β-class MTases are shown in magenta, the corresponding elements found in M.*Mwo*I (δ-class) and TVN1413 (ζ-class) are shown in green and cyan, respectively.](1471-2148-2-3-1){#F1} Most m^5^C MTases resemble the N-MTases of the γ class, with the only difference being their motif X localized at the C-terminus instead of the N-terminus. However, a few m^5^C MTases have been described with unusual permutations (Figure [1](#F1){ref-type="fig"}). To my knowledge, M.*Bss*HII is the only DNA MTase, for which the ζ architecture has been convincingly confirmed by experiment \[[@B16]\] and homology modeling \[[@B17]\]. In addition, a novel class of permuted m^5^C MTases typified by DRM2 have been recently identified in plants; this prediction has also been supported by threading of the permuted sequence onto the common fold \[[@B18]\]. Some DNA MTases contain terminal extensions and various insertions \[[@B6]\], but it has never been demonstrated that they are related to known or predicted TRDs in other proteins. DNA MTases are only one of the numerous families of remotely related enzymes that exhibit a common fold \[[@B11],[@B19]\]. Other members of the superfamily methylate a variety of chemically diverse molecules, including various RNAs, proteins, lipids, small molecules, etc. While all members of the superfamily share the same structural core, only DNA MTases vary in the order of conserved motifs. The rest show the order X, I, II, III, IV, V, VI, VII, VIII and only protein-arginine MTases lack the last three motifs \[[@B19]\]. The question thus arises, why do DNA MTases, but not MTases in general, exhibit sequence permutations within the same structural framework? Two models have been proposed to explain sequence permutations arising during the evolution of N-MTases \[[@B9],[@B20]\]. Jeltsch argued that the process of domain permutation needs duplication and in-frame fusion of the MTase gene, producing one enzyme with two catalytic domains \[[@B20]\]. Subsequent introduction of a new start codon in the middle of the first gene copy and a stop codon at the equivalent position in the second gene copy would then result in a circularly permuted variant. For instance, the ζ- or β-like permutants could arise from a hypothetical tandem γγ-class MTase. This model corresponds to the widely accepted concept that a permuted protein may arise naturally from tandem repeats by extraction of the C-terminal portion of one repeat together with the N-terminal portion of the subsequent repeat, so long as the protein\'s N and C termini are in close spatial proximity (reviewed in ref. \[[@B21]\]). Although the idea itself offers a plausible explanation for the origin of permutants within many protein families, the only known duplicated m^6^A MTases are the type IIS enzymes of the αα-class, whose permutation would eventually produce enzymes of the δ or ε class that have not been identified to date. The TRDs of known MTases from β and γ classes are unrelated \[[@B22],[@B23]\], hence it is unlikely that simple conversions \"from γγ to β \" or \"from ββ to γ \" have occurred in nature. Furthermore, the N- and C-termini of M.*Taq*I, the only γ-m^6^A MTase whose 3D structure is known, are quite distant in space \[[@B22]\]. Still, this scenario may be valid for enzymes that have not been identified yet, or whose sequences have not been studied in enough detail. It has been hypothesized (ref. \[[@B9]\]) that the permuted DNA MTase variants were generated by intra- or intergenic rearrangements of gene fragments (i.e. \"module shuffling\"; reviewed in refs. \[[@B24],[@B25]\]) that left no evidence of duplication intermediates. However, only in one case (M.*Bss*HII) has it been possible to reconstruct a possible evolutionary history of shuffled fragments \[[@B17]\]. Moreover, no examples of N-MTases in different classes are known, whose TRDs are markedly similar. Hence, no convincing examples of permutations of an entire N-MTase molecule have been identified to date. The reported permutations in N-MTases concern only the segments within the catalytic domain, while the unrelated TRDs were acquired or evolved independently in distinct classes \[[@B20]\]. Identification of closely related DNA MTases with homologous TRDs that nevertheless lie in different classes might help decide between the two given hypotheses and shed light on how the different classes of DNA MTases arose. Results ======= Candidates for N-MTases in the midst of the process of permutation were sought amongst all DNA MTases, whose sequences were available from REBASE \[[@B26]\]. Those sequences exhibiting deviations from the typical spacing between the conserved motifs or unusual extensions at the termini were chosen for detailed analysis. Within some extensions, several known or predicted DNA-binding domains were identified (to be published elsewhere), however only two pairs of MTase sequences were found that fit the criteria of similarity between the AdoMet-binding subdomain, the active site subdomain, and the TRD, without conservation of their co-linearity. One MTase that emerged from the analysis was the M.*Mwo*I protein, which recognizes an interrupted palindrome GCNNNNNNNGC and methylates one of the cytosines in each strand, generating m^4^C \[[@B27]\]. M.*Mwo*I (GenBank record 2961238) was earlier classified as a member of the β-class, however it exhibits quite unusual length of 668 aa, which is approximately twice the length of a typical β-class MTase \[[@B6]\], and it lacks the variable insertion between motifs VIII and X, corresponding to the TRD in β-class MTases \[[@B9]\]. BLAST searches and threading analysis (see Methods) revealed that the N-terminal part of M.*Mwo*I (aa 1--270) aligns very well with the catalytic and the AdoMet-binding subdomains of β-class MTases. However, in β-class MTases these subdomains are separated by the TRD, which in M.*Mwo*I is replaced by only a few residues (Figures [2a](#F2){ref-type="fig"}, [3a](#F3){ref-type="fig"}). Comprehensive results of the threading analysis of the M.*Mwo*I sequence are available at the URL <>, the alignments of the N-terminal region are essentially identical to those reported previously \[[@B9]\]. Remarkably, The C-terminal region of M.*Mwo*I revealed no similarities to other sequences, with one prominent exception, namely the predicted TRD of the m^4^C MTase M.*Sfi*I (GenBank entry 2761010) with the BLAST expectation (e) value of 10^-3^. Further database searches using the sequence of M.*Sfi*I and the isolated fragments of putative TRDs confirmed that the three major subdomains of M.*Mwo*I and M.*Sfi*I exhibit significant sequence similarity, but the linear order of these elements differs between them. If this prediction is correct, M.*Mwo*I should be classified as the first member of the δ-class, rather than the β-class (Figures [1](#F1){ref-type="fig"}, [2a](#F2){ref-type="fig"}). It is noteworthy that using either PSI-BLAST or threading, no significant sequence or structural similarities of the TRD of M.*Mwo*I and M.*Sfi*I could be detected to TRDs of other MTases and generally to sequences of other proteins. The additional sequence region of M.*Mwo*I (aa 434--497), which may be regarded as a linker between the N-terminal catalytic domain and the newly identified C-terminal TRD, also showed no matches to any sequences in the database. Hence, the determination of which arrangement of subdomains, that of M.*Mwo*I or M.*Sfi*I, corresponds to the ancestral state must await discovery of their homologs. {#F2} ![Sequence alignment of β-MTases with the permuted MTases: **a)** M.*Mwo*I (δ-class) **b)** M.*Tvo*ORF1413P (ζ-class). Blue numbers indicate the sequence coordinates (residue numbers). Red numbers in parentheses indicate the size of terminal extensions or loops, which were omitted for clarity. Identical residues are highlighted in black, conservative substitutions are highlighted in gray. Conserved motifs are labeled according to the nomenclature described by Malone et al \[[@B6]\]. Secondary structure elements shown below the alignment were deduced from the M.*Pvu*II coordinates or predicted for the TRD of M.*Sfi*I/M.*Mwo*I.](1471-2148-2-3-3){#F3} M.*Sfi*I recognizes the sequence GGCCNNNNNGGCC, which belongs to a broader set of sequences recognized by the GCNNNNNNNGC-specific M.*Mwo*I. It is not known, how these enzymes recognize such a lengthy sequence with a non-specific spacer. Nonetheless, it can be imagined that the TRD of M.*Sfi*I evolved from the TRD of M.*Mwo*I by acquisition of new contacts to bases outside and inside the GC pair (N-\>G)GC(N-\>C)NNNNN(N-\>G)GC(N-\>C) or conversely, the stringent DNA-recognition specificity of the M.*Sfi*I-like TRD was relaxed to give rise to the less specific M.*Mwo*I. In the absence of protein-DNA co-crystal structures for the β-class of MTases and lack of suitable structural templates for modeling the TRD structure in M.*Sfi*I and M.*Mwo*I, prediction of the detailed protein-DNA contacts is unfeasible. However, I hope that the finding reported herein will prompt mutagenesis experiments -- it is tempting to speculate that swapping the predicted TRDs between M.*Mwo*I and M.*Sfi*I will result in an exchange of specificities. A second MTase that came from the initial screen was M.*Tvo*ORF1413P, interpreted as a member of the γ-class in ReBase \[[@B26]\]<>, but exhibiting an extension of over 150 aa located N-terminally to motif X instead of the C-terminal extension after motif VIII required by the structure of the γ-class. In a BLAST search initiated with the M.*Tvo*ORF1413P sequence, M.*Tha*I was reported as the best hit, with a highly scored alignment of the AdoMet-binding region (e-value 5\*10^-14^) and a quite poorly scored (0.049) alignment of the catalytic region. However, M.*Tha*I is a member of the β-class and these two regions of similarity are swapped in the primary sequences of the two MTases (Figures [2b](#F2){ref-type="fig"}, [3b](#F3){ref-type="fig"}). BLAST searches initiated with the N- and C-terminal parts of M.*Tvo*ORF1413P showed that its C-terminal region scores better (e-value 6\*10^-4^) when aligned with the catalytic domain of another CGCG-specific β-m^4^C MTase, M.*Tma*I (GenBank record 4980829), a close relative of M.*Tha*I (data not shown). Threading analysis of the M.*Tha*I sequence revealed its perfect compatibility with the M.*Rsr*I and M.*Pvu*II structures (results are available at <>, allowing homology modeling of the M.*Tha*I structure (Figures [3b](#F3){ref-type="fig"}, [4](#F4){ref-type="fig"}). These results reveal that the TRD of M.*Tha*I is shorter than TRDs of M.*Rsr*I and M.*Pvu*II, which may be an indication that in this enzyme some DNA-binding residues migrated to other loops \[[@B9]\]. On the other hand, threading of M.*Tvo*ORF1413P revealed that its C-terminus corresponds to the last β-strand of the common MTase core (motif VIII; Figure [1](#F1){ref-type="fig"}), strongly arguing against its assignment to the γ-class, which requires the presence of the TRD C-terminal to this element (threading results are available at <>, <> and <> with the three entries corresponding to the full length sequence, the N-terminal part and the C-terminal parts, respectively). Instead, all threading algorithms reported that the N-terminus of M.*Tvo*ORF1413P matches perfectly the additional β-strand (motif IX-N) in M.*Rsr*I and M.*Pvu*II and this region, along with the predicted TRD, aligns quite well with motif IX-N and the TRD of M.*Tha*I (Figure [3b](#F3){ref-type="fig"}). It is noteworthy, that for the core regions, the predicted secondary structure agreed very well both between M.*Tha*I and M.*Tvo*ORF1413P, and between these MTases and the experimentally determined structures of M.*Rsr*I and M.*Pvu*II (for details see the above mentioned links to MetaServer results). For the 42 N-terminal residues of M.*Tvo*ORF1413P no similarity to known sequences or structures could be demonstrated, and modeling based on N-terminally extended threading alignments resulted in misfolded structures; it is therefore possible that this region forms an elaboration of the common fold, which is unique to M.TvoORF1413P. {#F4} The MTase activity of M.*Tvo*ORF1413P remains to be demonstrated. However, its close homolog has been recently identified, which exhibits a genuine MTase activity (Drs. M.A. Abdurashitov and S.K. Degtyarev, personal communication). M.*Bst*F5I-4, whose sequence remains unpublished, is evidently homologous to M.*Tvo*ORF1413P over the region including the predicted N-terminal TRD, as well as motifs IX-N, X, and I-VIII (BLAST e-value 3\*10^-20^, 26% identical plus 23% conservatively substituted residues; with 60% identical residues in the predicted TRD, i.e. the 10 aa loop preceding motif IX-N; data not shown). It cannot be ruled out that the small TRD of M.*Tha*I, M.*Bst*F5I-4, and M.*Tvo*ORF1413P harbors only a fraction of specificity determinants and that other loops on the catalytic face of the protein contribute to specific DNA recognition. Nevertheless, according to the classical definition of the TRD (the variable region between motifs VIII and X \[[@B6],[@B28]\]), the presented results of sequence analysis and structure prediction suggest that the common ancestor of ζ-class MTases M.*Bst*F5I-4 and M.*Tvo*ORF1413P evolved from M.*Tha*I (β-class member) by sequence permutation. Discussion ========== Sequence analysis resulted in identification of two novel cases of sequence permutation in DNA MTases, and demonstration for the very first time, that MTases of different classes may exhibit significant sequence similarity not only in the catalytic domain, but also in the TRD. This finding suggests that the analyzed gene pairs diverged relatively recently, permitting a test of the hypothesis that the observed rearrangements occurred according to the \"permutation-by-duplication\" model \[[@B20]\] or to the alternative model, involving intragenic relocation of gene segments. If sequences resembling fragments of one of the DNA MTase analyzed herein were identified in its own neighborhood, this would provide strong evidence that gene duplication occurred. It would also suggest that this particular MTase is a permuted version of its homolog, whose neighborhood is free from duplicated fragments, rather than the opposite. Regrettably, the neighborhood of the *Sfi*I and *Mwo*I RM systems is unknown, however the context of *Tha*I RM system and M.*Tvo*ORF1413P can be analyzed using the complete genome sequences of *Thermoplasma acidophilum*\[[@B29]\] and *T. volcanium*\[[@B30]\], respectively. The genome sequences of both *Thermoplasma* species flanking M.*Tha*I and M.*Tvo*ORF1413P (10 000 base pairs in each direction) were compared using the BLAST-family programs at the level of DNA and putative translations in all open reading frames (with stop codons translated as missing characters, e.g. \"X\"). No evidence of sequences similar to the genes encoding these two MTases were found, except for another putative MTase M.*Tvo*ORF1416P located 2640 bp 5\' to M.*Tvo*ORF1413P. M.*Tvo*ORF1416P is a typical α-class member and exhibits significantly higher similarity to α-MTases such as M.*Psp*GI (BLAST e-value 2\*10^-73^) or M.*Mva*I (e-value 3\*10^-67^) than to M.*Tvo*ORF1413P (insignificant e-value 5.3). Therefore, the two MTases should be regarded as remote homologs of each other and as members of different phylogenetic lineages \[[@B9]\] that met rather accidentally in the *T. volcanium* chromosome rather than as products of recent duplication of one gene. The lack of evidence supporting the gene duplication mechanism in the case of MTases from *Thermoplasma* is not entirely convincing, especially since no direct evidence supporting the alternative \"cut-and-paste\" mechanism can be provided by sequence analysis. Hence, the events leading to M.*Sfi*I(β)-M.*Mwo*I(δ) and M.*Tha*I- (β) M.*Tvo*ORF1413P (ζ) permutations were reconstructed based on both mechanisms (Figures [5](#F5){ref-type="fig"}, [6](#F6){ref-type="fig"}, [7](#F7){ref-type="fig"}, and [8](#F8){ref-type="fig"}). For the sake of simplicity, it was assumed that the unique δ and ζ-class members were in these cases generated by permutation of common β-class members, however an analogous reconstruction could be carried out assuming the opposite directionality of rearrangements, leading to similar conclusions. {#F5} {ref-type="fig"}.](1471-2148-2-3-6){#F6} {ref-type="fig"}.](1471-2148-2-3-7){#F7} {ref-type="fig"} and [7](#F7){ref-type="fig"}.](1471-2148-2-3-8){#F8} Figures [5](#F5){ref-type="fig"} and [6](#F6){ref-type="fig"} shows possible histories leading to both permutations according to the gene duplication mechanism. The order of the sequence motifs remains conserved between M.*Sfi*I and M.*Mwo*I, but these two MTases differ in that a large segment bearing the TRD appears in the middle of the former but in the C-terminus of the latter, hence permutation of these two proteins is not circular (Figure [5](#F5){ref-type="fig"}). To produce M.*Mwo*I, duplication of M.*Sfi*I would have to be followed by both creation of a novel stop codon to eliminate the region encoding the \"new\" catalytic subdomain of the C-terminal repeat and deletion of the regions corresponding to the \"old\" TRD and the \"new\" AdoMet-binding subdomain. It is quite unlikely that all these changes occurred in a single event, and their occurrence in a series of steps would seem inevitably to produce a nonfunctional intermediate, which would require two steps to regain activity. If these changes did occur gradually, the product of gene duplication, in which one repeat retained the AdoMet-binding subdomain but lost the catalytic subdomain (or conversely), would expose the hydrophobic core of the remaining nonfunctional subdomain to the solvent. Folding and enzymatic function of such \"1 & ½\" mutant would be probably heavily compromised. The function of M.*Sfi*I and M.*Mwo*I is to protect the chromosome from being cleaved by the cognate ENase. Hence, it seems rather unlikely that the host cell would survive such a series of unlikely events passing through functionally compromised intermediates. Compared to M.*Sfi*I and M.*Mwo*I, evolution of M.*Tvo*ORF1413P from M.*Tha*I seems more likely (Figure [6](#F6){ref-type="fig"}), since in this \"classical\" case of circular permutation requires only removal of the terminal regions by formation of new start and stop codons. However, in this case deletion of terminal subdomains, or their large parts, would also have to be concurrent, otherwise nonfunctional intermediates could arise, leading to cell death due to insufficient protection against the cognate ENase. If only one of the repeats in the original tandem ββ fusion protein is damaged, deletion of the remaining nonfunctional part would most likely restore the highly active, single copy version of the parent MTase. It is noteworthy that the tandem duplication mechanism offers no stage, at which evolutionary pressure would result in optimization of a poorly active intermediate specifically towards the permuted version. Another problem with the scenario involving tandem ββ MTase fusion is that such fusions have never been reported to occur. In the X-ray structure of the β-class member M.*Rsr*I, the two identical subunits make quite extensive contacts (a loss of 1799.3 Ang\*\*2 of solvent accessible surface area per chain upon complex formation; see URL: <>, suggesting that the dimeric structure of this MTase is biologically relevant \[[@B23]\]. The TRD and the active site reside on opposite sides of the M.*Rsr*I monomer, but the unique dimeric configuration brings the TRD of one subunit near the active site of the other, indicating that dimerization may be required for recognition and methylation to occur. The N- and C-termini of M.*Rsr*I are located close to each other in the monomer (8.8 A), but the C-terminus of one monomer is located on the opposite side of the dimer in respect to the N-terminus of the other monomer, separated by a distance of 74 A in a straight line. If the configuration observed in the crystal structure of M.*Rsr*I is representative for other members of the β-class that use two cooperating MTase domains, covalent joining of the termini of these domains would require a very long linker peptide, looping around the dimer. Hence, tandem fusion seems disrupting for cooperation of two β-class MTases within the dimer. An alternative \"cut-and-paste\" mechanism (Figures [7](#F7){ref-type="fig"}, [8](#F8){ref-type="fig"}), inspired by the observed genomic rearrangements associated with the presence of restriction endonucleases \[[@B5],[@B31]\], involves generation of a functional gene from fragments. In one scenario (Figures [7](#F7){ref-type="fig"} and [8](#F8){ref-type="fig"}, left panels), the fragments may be generated due to combined action of various endo- and exonucleases that partially degrade the DNA fragment encoding the MTase gene, thereby producing recombingenic ends. If degradation occurs, at least two copies of the MTase gene must be present in the cell in order to reconstruct all important regions. Another scenario (Figures [7](#F7){ref-type="fig"} and [8](#F8){ref-type="fig"}, right panels) involves precise action of a sequence-specific endonuclease, which fortuitously cleaves the MTase gene in the regions corresponding to linkers between the TRD and the two subdomains of the catalytic domain. In this case, one or more copies of the MTase gene may be present in the cell. These two scenarios are not mutually exclusive, provided that the fragments resulting from any type of cleavage or degradation span all regions necessary for the MTase activity, and are able to recombine with each other or use \"sticky ends\" to ensure ligation. The \"cut-and-paste\" mechanism (Figures [7](#F7){ref-type="fig"}, [8](#F8){ref-type="fig"}) differs from the \"permutation-by-duplication\" mechanism (Figures [5](#F5){ref-type="fig"}, [6](#F6){ref-type="fig"}) in that it involves a momentary stage at which there are no intact, active MTase gene copies in the cell. The accompanying ENase gene might be fragmented as well, leading to elimination of the RM system from the cell. However, if the ENase remains active for a certain period of time, only those cells survive in which the MTase gene is restored from fragments. Such MTase may exhibit various deletions, duplications of certain regions and rearrangements, as long as these modifications allow the protein to provide protection against the ENase. The selection pressure will result in rapid optimization of the MTase function, most likely to the nearest maximum in the fitness landscape. With a certain probability, the permuted gene copy will arise, and under such \"all or nothing\" conditions, its sequence will be optimized towards the modification activity sufficient to protect the host\'s chromosome. It is worth mentioning that in the short term the ENase may remain active and provide selective pressure for restoring expression of the MTase even if its own gene has been destroyed. A hybrid mechanism can be envisaged in which the complete, fully functional MTase gene undergoes recombination with a fragment of a MTase gene (not shown). This mechanism has limitations similar to that of the \"permutation-by-duplication\" mechanism in that the newly fused fragment must not compromise the function of the original protein. However, there is no specific reason why a part of the original domain should be deleted with a higher frequency than the new fragment and why the latter scenario should be selected for, unless the alternative fragments are not identical and the new fragment encodes a function, which may increase the fitness of the protein. The scenarios shown in Fig. [5a](#F5){ref-type="fig"} require that the TRD can function autonomously, outside of its original structural environment. For some TRDs, at least, this is the case (reviews: \[[@B2],[@B12]\]). Not only is the movement of TRDs within a MTase plausible, but the exchange of TRDs between MTases can provide different specificities and thereby functional advantage \[[@B32]\]. Indeed, MTases that methylate more than one specific DNA target owing to the presence of several TRDs at various locations of the enzyme have been identified \[[@B33],[@B34]\], and shuffling of TRDs have been suggested to occur among both mono- and multispecific MTases \[[@B35]\]. However, while the unrelated TRDs of many structurally characterized MTases form structurally autonomous domains, the TRDs of two members of the β-class, M.*Rsr*I \[[@B23]\] and M.*Pvu*II \[[@B36]\], form only an amendment of the common fold, which is quite unlikely to behave as an independently folded and functionally autonomous unit (review: \[[@B12]\]). The putative TRD of M.*Tvo*ORF1413P and M.*Tha*I is most likely too small and too poorly structured to be regarded as an independent domain (Figures [3b](#F3){ref-type="fig"}, [4](#F4){ref-type="fig"}). However, the putative TRD of M.*Sfi*I and M.*Mwo*I is much longer and comprises at least four predicted helices (Figure [3a](#F3){ref-type="fig"}), therefore it cannot be excluded that it may form an independently folded, functional unit. It has been demonstrated that MTase M.*Aqu*I comprises two independent polypeptides corresponding to the catalytic domain and the TRD, which associate in solution to form a functional enzyme \[[@B37]\]. Correspondingly, one of the scenarios of evolution of the M.*Mwo*I enzyme involves temporary separation of the gene fragments encoding the catalytic domain and the TRD (Figure [7](#F7){ref-type="fig"}). According to this scenario, the linker region between the catalytic domain and the TRD originated from the initially non-coding sequence that initially separated the two functional units. It seems likely that covalent linkage of the two domains by the newly established linker increased the fitness of the rearranged MTase. The \"cut-and-paste\" scenario offers an explanation for the sequence permutations being observed in DNA MTases and not among ENases. This results from the asymmetry of selection for restoring methylation function and restriction function after the corresponding genes are fragmented. In other words, the newly permuted proteins are probably poor enzymes, but the ENase provides strong selective pressure for optimization of the MTase function, while in the second case selective pressure is relatively weak (functional ENases probably provide only a minor selective advantage and they are not required for protection of the host \"against\" the MTase). However, it cannot be excluded that ENases are simply not amenable to any sequence permutation for structural reasons, since these proteins exhibit different fold than MTases. Interestingly, sequence permutations have been observed only amongst DNA MTases \[[@B1]\], but not in other MTase families (enzymes acting on RNA, proteins, small molecules, etc.), despite their common structure. To date, no explanation has been offered for this peculiarity, even though it raised considerable interest in the field \[[@B11]\]. It is tempting to speculate that the rearrangements observed amongst DNA MTases but not other MTases are induced \[[@B5]\] by the increased exposure of their genes to the repertoire of various nucleases encoded by different hosts during horizontal transfer events common among the RM systems (J. Elhai, personal communication). High frequency of such events was inferred from sequence analyses \[[@B5],[@B38]\]. If this hypothesis is correct, interaction with the mechanisms of defense against alien genetic elements encoded by various Prokaryotes may be also responsible for permutations of entire domains within type I and type III RM systems and a plethora of combinations of various domains in many \"non-classical\" RM system subtypes (review: \[[@B12]\]). Summary ======= Analyses of the evolutionary scenarios presented herein favor the \"cut-and-paste\" mechanism or the hybrid mechanism (fusion of an intact MTase with the TRD) for the M.*Sfi*I(β)-M.*Mwo*I(δ) rearrangement and the original \"permutation-by-duplication\" mechanism or the \"cut-and-paste\" mechanism (rather than their hybrid) for the M.*Tha*I-(β) M.*Tvo*ORF1413P(ζ) rearrangement. Even though in these and probably other cases, certain scenarios may seem more likely than the other, none of them can be ruled out completely. The presented mechanisms are not mutually exclusive, and all have probably played significant roles in the generation of permuted MTases. Methods ======= The PSI-BLAST algorithm \[[@B39]\] was used to search the non-redundant version of current sequence databases (nr) and the publicly available complete and incomplete genome sequences at the NCBI website <>. All genuine and putative N-MTase sequences available from REBASE \[[@B26]\] were submitted as queries with default parameters. Protein structure prediction was carried out using the MetaServer available at <>, which combines several secondary structure prediction and threading methods (ref. \[[@B40]\] and references therein). These threading methods compare the query sequence (the target) with a library of structures (templates) and return 10 alignments that scored best according to the implemented criterion of compatibility. The results are evaluated by the Pcons server \[[@B41]\], which compares the models and the associated scores and produces a ranking of potentially best predictions (target-template alignments). Based on the results produced by the MetaServer, homology modeling was carried using the SWISS-MODEL/PROMOD II server \[[@B42]\]. Model evaluation was carried out using the PROSA II \[[@B43]\] program integrated with PROMOD II, suggesting that the stereochemistry and energetic parameters of the models were acceptable. Acknowledgements ================ I would like to thank Jeff Elhai, who as a non-anonymous reviewer provided extensive comments, and raised constructive criticism of the \"cut-and-paste\" mechanism, which lead to considerable improvement of the manuscript, Dr. Ichizo Kobayashi for critical reading and comments on the initial version of the manuscript and Drs. Murat Abdurashitov and Sergey Degtyarev for supplying unpublished data. This work was supported by KBN (grant 6P04 B00519).

Background ========== Respiratory syncytial virus (RSV) belongs to the Pneumovirinae subfamily of the Paramyxovirodae family of enveloped single-stranded negative sense RNA viruses. RSV infection of the lower respiratory tract cells results in cell death and sloughing into the lumen of the respiratory tree. Worldwide, RSV is the leading cause of infant mortality from respiratory infections and is so highly contagious that by age two nearly all children have been infected. RSV infection in infancy cause severe bronchiolitis and pneumonia and may predispose children to the subsequent development of asthma, the most common chronic illness of childhood \[[@B1]\]. Many studies have indicated that chemokines can play an important role in the onset and severity of asthma and it has been shown that RSV infection of lung epithelial cells increases chemokine production, although the mechanisms involved are largely unknown \[[@B2]-[@B5]\]. The chemotactic cytokines, or chemokines, compose a large superfamily of small structurally related polypeptides that play important roles in host defense by recruiting specific subsets of leukocytes to sites of inflammation and injury \[[@B6]\]. Chemokines have been associated with a number of inflammatory diseases and conditions, including asthma, sepsis, inflammatory bowel disease, and adult respiratory distress syndrome \[[@B7]-[@B9]\]. The chemokine superfamily can be divided into two major groups based on the position of the first two of four-conserved cysteine residues at the amino terminus, which are either adjacent (CC subfamily) or separated by one amino acid (CXC subfamily). The CXC chemokines such as IL-8 were originally identified as potent activators and chemoattractants for neutrophils, whereas the CC chemokines such as MCP-1 and RANTES mostly attract monocytes and eosinophils respectively \[[@B10]\]. Chemokines are secreted in a stimulus-and cell type-specific manner \[[@B11]-[@B17]\] and are regulated primarily at the level of gene transcription \[[@B18]-[@B24]\]. The transcriptional promoters of IL-8, RANTES and MCP-1 contain binding sites for the redox-responsive transcription factor NF-κB, which has been shown to be important for their regulation by viral infections and cytokines \[[@B18],[@B20],[@B23],[@B25]-[@B34]\]. We previously demonstrated that the chemokines IL-8, MCP-1 and RANTES are differentially regulated in A549 airway epithelial cells \[[@B35]-[@B38]\]. To further elucidate the mechanisms of chemokine expression in A549, we have compared the induction of IL-8, MCP-1 and RANTES by RSV infection with that of TNFα. Our findings suggest that RSV induction of chemokine gene expression involves a redox-sensitive NF-κB signaling mechanism that differs from that mediated by TNFα and involving predominantly the Rel A subunit of NF-κB. Materials and methods ===================== Materials --------- Dulbecco\'s Modified Eagle Medium (DMEM), fetal bovine serum (FBS), Dulbecco\'s phosphate buffered saline (DPBS), antibiotic/antimycotic, 1% trypsin/EDTA, Hanks Balanced Salt Solution (HBSS) and TRIZOL were purchased from Invitrogen Gibco Cell Culture (Carlsbad, CA). N-acetyl-L-cysteine, dexamethasone, glycerol and MTT tetrazolium salt were obtained from Sigma (St. Louis, MO). TNFα was obtained from R&D systems (Minneapolis, MN). ELISA kits were purchased from Pierce Endogen (Rockford, IL). Human CK5 RiboQuant ribonuclease protection assay kit was purchased from BD Pharmingen (San Diego, CA). \[α-^32^P\]UTP (250 μCi) was obtained from Perkin Elmer Life Sciences (Boston, MA). Gel shift assay system was purchased from Promega (Madison, WI). \[γ-^32^P\]ATP (500 μCi) was obtained from ICN (Costa Mesa, CA). Antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The A549 cell line and RSV Long strain were obtained from the American Type Culture Collection (Rockville, MD). Virus stock growth and maintenance ---------------------------------- RSV, Long strain, was grown on HEp-2, a human tracheal epithelial cell line. Cells were grown to 50% confluence in DMEM containing 7% FBS and 1% antibiotic/antimycotic. After two washes with 1X DPBS, a minimal volume of RSV, at a multiplicity of infection (MOI) of 1 or greater, containing less than 1% FBS was added. The virus and cells were incubated for 2 hours, after which DMEM was added to bring culture to normal growth volume and 7% FBS. Cultures were then incubated for 48 hours. All plates were scraped and contents were transferred to 50 ml conical tubes. The virus/cell cocktail was vortexed briefly and large debris was removed in a tabletop centrifuge. RSV was purified by ultracentrifugation through 30% glycerol in HBSS. RSV stocks were resuspended in HBSS containing 0.5% bovine serum albumin and 100 mM magnesium sulfate and frozen at -70° C. Cell culture, infection and treatments -------------------------------------- The A549 human type II lung carcinoma cell line was grown and maintained in DMEM containing 7% FBS and 1% antibiotic/antimycotic. For treatment, cells were grown to approximately 75% confluence, washed once with 1X DPBS and incubated overnight with DMEM containing 1% FBS. Cultures were then washed twice with 1X DPBS and pretreated with inhibitors DEX or NAC in serum-free DMEM to final concentrations as indicated in the figures. RSV and TNFα were added to the DMEM/inhibitor mix to final concentrations as indicated in the figures. One hour after agonist addition, FBS was added to 1% concentration for the duration of the experiment. Protein secretion ----------------- After addition of RSV and TNFα, cells were cultured for 24 h. Supernatants were collected and spun at high speed to remove cellular debris. The resulting suspension was frozen at -70°C until assayed following manufacturer\'s instructions by chemokine-specific ELISA (Endogen). RNase protection ---------------- RNase protection assays (RPA) were performed using an RPA kit purchased from BD Pharmingen. In brief, total RNA was isolated from stimulated A549 cells with TRIZOL. The hCK-5 multiprobe template set was used to synthesize RNA probes for the chemokines lymphotactin, RANTES, IP-10, MIP-1α, MIP-1β, IL-8, MCP-1, and I-309 as well as the house-keeping genes L32 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) labeled with \[α-^32^P\]UTP using T7 RNA polymerase. 3 × 10^5^ cpm of labeled probe were hybridized with 10 μg of total RNA for 16 h at 56°C. MRNA-probe hybrids were treated with RNase cocktail and phenol-chloroform extracted. Protected hybrids were resolved on a 6% denaturing polyacrylamide sequencing gels and exposed to a Molecular Dynamics detection screen overnight. Laser densitometry was performed using a Molecular Dynamics Storm scanner system (Molecular Dynamics, Inc., Sunny vale, CA). DNA binding studies ------------------- Electrophoretic mobility shift assays (EMSA) were performed essentially as described previously \[[@B39]\]. Briefly, nuclear protein extracts (7 μg protein) prepared form A549 cells by the method of Osborn *et al*\[[@B40]\] were incubated with 5 × 10^5^ cpm (\~0.1 ng) of ^32^P end-labeled consensus NF-κB oligonucleotide (5\'-AGTTGAGGGGACTTTCCCAGGC-3\') probe contained in the Promega gel shift assay system for 20--30 min at room temperature in 10 μl reaction volume using the kit provided reaction buffer. To demonstrate binding specificity, 100-fold molar excess (10 ng) of a specific or non-specific oligonucleotide (5\'-ATTCGATCGGGGCGGGGCGAGC-3\') was included in the binding reaction. Protein-DNA and protein-DNA-antibody complexes were resolved in 5% polyacrylamide gels preelectrophoresed for 30 min at room temperature in 0.25X TBE buffer (22.5 mM Tris-borate and 0.5 mM EDTA, pH 8.3). Gels were dried and exposed to radiographic film with an intensifying screen at -70°C. NF-kappaB subunit specific ELISA -------------------------------- NF-κB subunits Rel A (p65) and NF-κB1 (p50) were quantified using Trans-AM™ transcription factor assay kit from Active Motif (Carlsbad, CA). The assay is essentially an ELISA in which the consensus NF-κB binding site (5\'-GGGACTTTCC-3\') is immobilized onto the 96 well plate. Nuclear cell extracts (3 μg) were added to the wells and assayed for either Rel A or p50 binding as per the manufacturer\'s instructions. Optical density was determined on a spectophotometer at 450 nm. Results ======= RSV induction of chemokine expression differs from that induced by TNFalpha --------------------------------------------------------------------------- As shown in Fig. [1](#F1){ref-type="fig"}, RSV infection of A549 epithelial cells induces the production and secretion of the chemokines IL-8, MCP-1, and RANTES. In A549 epithelial cells substantial MCP-1 was constitutively expressed whereas little or no spontaneous IL-8 or RANTES was detected. However, upon RSV infection (MOI = 3) a significant increase in chemokine production and secretion was observed 24 hr post-infection for each of the chemokines. {#F1} To determine whether the chemokine induction was mediated by increased chemokine gene expression we assessed the kinetics of mRNA synthesis by chemokine specific RNase protection. A549 cells were infected with RSV (MOI = 1) or stimulated with TNFα (100 ng/ml) and total RNA isolated at various times post-treatment over a 24 hr time course. As shown in Fig. [2](#F2){ref-type="fig"}, RSV rapidly induced IL-8 and MCP-1 with steady-state mRNA readily detected as early as 1 hr and continued to increase over the 24 hr time course. In contrast, RANTES mRNA was not detected until 4 hrs post infection after which its expression rapidly increased also reaching a maximum at 24 hrs. The RSV induction kinetics differed dramatically from the induction kinetics for TNFα in the same experiment. Whereas RSV induction of the chemokines did not reach a maximum until 24 hrs, TNFα induction was maximal between 4 and 8 hrs and was either unchanged or lower at 24 hrs post-infection. {#F2} The magnitude of induction also differed between TNFα and RSV infection. In general TNFα was a stronger inducer of chemokine expression. This difference was also reflected in the amount of chemokine secreted from the cells (data not shown). Together, these data indicate that RSV infection of A549 epithelial cells induces chemokine expression with distinct kinetics suggesting the mechanism of RSV induction differs from that of proinflammatory cytokines. NF-kappaB inhibitors differentially effect RSV and TNFalpha induction of chemokine expression --------------------------------------------------------------------------------------------- IL-8, MCP-1 and RANTES gene expression are regulated by the redox responsive transcription factor NF-κB \[[@B18],[@B20],[@B23],[@B25]-[@B33],[@B41]-[@B44]\]. The NF-κB signaling pathways have been shown to be differentially affected by antioxidants and glucocorticoid steroids \[[@B45]\]. To determine whether RSV and TNFα induction of chemokine expression might be mediated by different NF-κB signaling pathways we pretreated A549 cells with the glutathione precursor N-acetyl-L-cysteine (NAC) or the synthetic glucocorticoid dexamethasone (DEX). Twenty-four hours post infection, total RNA was isolated and steady state mRNA expression analyzed by chemokine specific RNase protection. As shown in Fig. [3](#F3){ref-type="fig"}, the two inhibitors had very different effects on chemokine gene expression. Most strikingly, RSV induced chemokine expression was more sensitive to NAC, whereas TNFα induced chemokine expression was conversely more sensitive to DEX. Thus, NAC preferentially inhibited chemokine expression induced by RSV, while DEX predominantly inhibited chemokine expression induced by TNFα. {#F3} Differential inhibition of TNFalpha and RSV induced NF-kappaB ------------------------------------------------------------- NF-κB binds to the chemokine promoter as either homo- or heterodimers composed of Rel A (p65) and NF-κB1 (p50). Initially we detected TNFα and RSV induction of NF-κB by EMSA (KA Roebuck and LR Carpenter unpublished data) but because multiple gel shift complexes were induced it was not possible to distinguish the individual contributions of Rel A and p50 to the formation of these complexes. Therefore to assess the induction of the Rel A and p50 subunits individually in A549 nuclear cell extracts, we used a subunit specific NF-κB binding ELISA system that utilizes anti-Rel A and anti-p50 antibodies. As shown in Fig. [4](#F4){ref-type="fig"}, Rel A and p50 subunits of NF-κB are differentially expressed and induced in A549 epithelial cells. In untreated cells, no Rel A (solid bars) was detectable whereas in resting cells constitutive levels of p50 were observed (open bars), suggesting that unstimulated A549 epithelial cells contain p50 homodimers, which with regard to transcriptional activity have been shown to be functionally inert or inhibitory. In contrast, cells stimulated by TNFα (Fig. [4A](#F4){ref-type="fig"}) or infected by RSV (Fig. [4B](#F4){ref-type="fig"}) markedly induced Rel A, the transcriptionally active subunit of NF-κB. In addition, TNFα was also able to induce p50 by about 5-fold whereas RSV increased the constitutive p50 level by only about 50%. The induction of Rel A and p50 were specific since competition with either wild type or mutant consensus NF-κB oligonucleotide respectively inhibited or not the RSV and TNFα induced binding activity. These data demonstrate that TNFα is a potent inducer of both Rel A and p50 whereas RSV primarily induces the Rel A subunit of NF-κB. {#F4} To determine whether RSV and TNFα activate NF-κB differently, we examined the effects of NAC and DEX on Rel A and p50 binding activity. As shown in Fig. [4](#F4){ref-type="fig"}, Rel A and p50 subunits are differentially inhibited by NAC and DEX which have been shown to inhibit NF-κB by distinct mechanisms. Both NAC and DEX were unable to inhibit TNFα induction of either Rel A or p50. In contrast, in RSV infected cells both DEX and NAC inhibited Rel A and p50 binding activity. However, the inhibitory effects were most striking for Rel A suggesting the inhibitors primarily target the RSV induction of Rel A. The differential effects of NAC on TNFα and RSV induced NF-κB binding activity correlated with its effects on chemokine gene expression (Fig. [3](#F3){ref-type="fig"}) and protein secretion (data not shown) suggesting that RSV induces IL-8, MCP-1 and RANTES through redox-sensitive NF-κB binding complexes. DEX also showed differential effects on NF-κB binding activity with DEX preferentially inhibiting RSV induction of NF-κB. However, the effects of DEX on NF-κB binding induced by TNFα did not correlate with its effects on chemokine expression, suggesting in TNFα activation of chemokine gene expression DEX targets another mechanism of chemokine induction not involving NF-κB. Discussion ========== In this study, we examined the role of NF-κB in the activation of chemokine gene expression in response to TNFα and RSV infection. The inhibitor studies correlating chemokine gene expression and NF-κB binding activity indicate that RSV, in contrast to TNFα, induces IL-8, MCP-1 and RANTES expression in A549 epithelial cells through a redox-sensitive NF-κB signaling pathway that appears to involve predominately Rel A. IL-8, MCP-1 and RANTES are known redox regulated genes and our results are consistent with the findings that oxidant tone can regulate chemokine gene expression in RSV infected airway epithelial cells \[[@B46]-[@B48]\]. NF-κB activation is controlled by its inhibitory protein I-κBα which when phosphorylated on ser32 and ser36 marks it for degradation by the 26S proteosome complex. Thus I-κB is a potential regulatory target for the differential induction of NF-κB. Although we have not directly examined the role of I-κB, Fiedler *et al*\[[@B49],[@B50]\] have reported differential effects of RSV and TNFα on I-κB activity in A549 epithelial cells. In addition, Bitko and Barik \[[@B51]\] have demonstrated that another I-κB protein (I-κBβ) may also contribute via a redox-sensitive pathway to NF-κB activation by RSV. In addition to NF-κB dependent regulation there appears to also be NF-κB independent activation of chemokine expression since DEX inhibited TNFα induced chemokine expression but not NF-κB binding activity. Along these lines, it has been shown that RSV activation of chemokine gene promoters involves multiple inducible transcription factors including AP-1 \[[@B52],[@B53]\], which is also a redox sensitive transcription factor important in chemokine gene expression \[[@B28],[@B36],[@B54]\]. AP-1 can also be inhibited by DEX and has been shown to cooperate with NF-κB to mediate RSV induction of IL-8 \[[@B46],[@B55]\]. The differential effects of NAC and DEX on chemokine gene expression and NF-κB activation suggest that the mechanisms of RSV and TNFα activation of NF-κB and chemokine expression differ possibly involving distinct NF-κB signaling mechanisms. TNFα rapidly and potently activates Rel A and p50 subunits of NF-κB through a kinase-mediated phosphorylation cascade involving a high molecular mass signaling complex called the IKK complex \[[@B56]\]. TNFα activation of the multiprotein IKK complex results in the serine phorsphorylation of I-κB and the subsequent activation of NF-κB \[[@B57]\]. Consistent with the idea of RSV and TNFα activating distinct NF-κB signaling pathways, we found that the NF-κB inhibitor NAC had differential effects on chemokine gene expression. In contrast to TNFα, RSV induced chemokine expression was sensitive to NAC particularly the Rel A (p65) subunit. NAC is an antioxidant thiol that raises the intracellular pool of glutathione the major redox regulating mechanism in the A549 epithelial cell line \[[@B45]\]. Interestingly, it has been reported that NAC can have differential effects on NF-κB binding activity suggesting only a subset of NF-κB binding complexes are redox sensitive \[[@B58]\]. Moreover, it has been shown that NAC inhibits Rel A/p50 heterodimers but not p50 homodimers and can inhibit NF-κB binding to only certain variant κB elements which presumably interact with distinct redox sensitive NF-κB binding complexes \[[@B59]\]. These observations are consistent with our findings that indicate NAC predominantly inhibits Rel A homodimers since that was the major NF-κB subunit induced by RSV infection of A549 cells. Consistent with RSV and TNFα inducing functionally distinct NF-κB binding complexes, AP-1 was recently shown to be the preferential cooperative partner with NF-κB in RSV-induced IL-8 expression \[[@B46]\], whereas in TNFα-induced expression NF-κB cooperates preferentially with NF-IL-6 \[[@B18],[@B19],[@B22]\]. Apparently, the stimulus-specific gene regulation of chemokines is complex involving differential cooperativity of redox sensitive and redox resistant NF-κB complexes with other transcription factors to form unique higher order transcription complexes. Conclusions =========== Taken together, our data indicate that RSV, in contrast to the cytokine TNFα, activate NF-κB and chemokine gene expression through a redox sensitive NF-κB signaling pathway involving predominately Rel A. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgments =============== We thank Mark Peeples and Larry Thomas for reading the manuscript prior to submission. This work was in partial fulfillment of the doctoral degree of Laura R. Carpenter and was supported by a National Institutes of Health grant AR45835 and grants from the American Lung Association (Blowitz-Ridgeway Foundation).

Background ========== Osteoarthritis (OA) is the most common joint disease. Articular cartilage in OA has shown to lose its mechanical resistance, elasticity and smoothness, and is consequently worn out by the movements of the joint. This leads to reactive bone remodeling, forming osteophytes, microfractures, subchondral eburnation and pseudocysts, and exposure of the articular end of the bone \[[@B1]\]. The consequent roughness of the articular cartilage surfaces elicits secondary inflammatory reactions of the synovial membrane and bone. Unlike rheumatoid arthritis and other inflammatory joint diseases, the inflammatory component of OA is relatively mild \[[@B2],[@B3]\]. Clinical manifestations of OA of the knee are joint pain, stiffness in the morning or after rest, pain at night, limited joint motion and/or joint deformity. Joint pain in OA may originate from not only synovitis, but also stretching of the joint capsule or ligaments, periosteal irritation due to osteophyte formation, trabecular microfractures, intraosseous hypertension, or muscle spasm \[[@B4]-[@B7]\]. There are many treatment modalities for OA of the knee including nonpharmacologic (e.g. patient education, weight control, physical and occupational therapy, and aerobic exercise programs) and pharmacologic therapy (e.g. intraarticular steroid injections, paracetamol, topical analgesics, nonsteroidal anti-inflammatory drugs (NSAIDs) and opioid analgesics) \[[@B8]\]. Although NSAIDs are the most widely prescribed drugs to reduce joint pain and stiffness, the inflammatory component of OA is usually minimal, therefore, the need for the anti-inflammatory effect of NSAIDs in this condition is controversial \[[@B2],[@B9],[@B10]\]. Moreover, inhibition of prostaglandin biosynthesis is directly related to many common and occasionally severe side effects including gastrointestinal bleeding, hypertension, congestive heart failure, hyperkalemia, and renal insufficiency \[[@B11]-[@B14]\]. These disadvantages call for an evaluation of the risks and benefits of the therapy in comparison with a less toxic one for OA. Since the efficacy of NSAIDs in symptomatic treatment of OA of the knee depends on the analgesic rather than anti-inflammatory effect, paracetamol (analgesic drug) has been recently recommended as the first-line oral drug in the management of OA of the knee \[[@B8]\]. However, the long term use of paracetamol probably leads to hepatic and renal impairment \[[@B15],[@B16]\]. Thus, less toxic pain managing procedures, e.g. electroacupuncture (EA), may be considered as an alternative treatment of this disease. Even though EA is safe and effective in combating pain \[[@B17],[@B18]\], its role in OA of the knee is still controversial and comparative studies between EA and NSAIDs in this disease are rare \[[@B19]\]. Therefore, the aim of this study was to compare the efficacy of EA, diclofenac and their combination in short-term, symptomatic treatment of OA of the knee. Materials and methods ===================== Research design --------------- This study was a randomized, single-blind, placebo controlled trial. The treatment procedures consisted of placebo tablet plus placebo EA (placebo group), diclofenac tablet plus placebo EA (diclofenac group), placebo tablet plus true EA (EA group) and diclofenac tablet plus true EA (combined group). This study was approved by the Medical Ethics Committee of the Faculty of Medicine, Chiang Mai University and was in compliance with the Helsinki Declaration. Subjects -------- Two hundred out-patients of either sex, aged over 40 years, and who had been suffering from unilateral or bilateral OA of the knee according to the criteria of the American College of Rheumatology \[[@B8]\] for more than 3 months duration, were recruited. Lequesne\'s functional index, which was evaluated at the screening visit, had to be at least 6 points. Subjects had to be able to walk and give both verbal and written information regarding the study. Signed informed consent was obtained prior to entry. Exclusion criteria included an underlying inflammatory arthropathy, expectation of surgery in the future, recent injury in the area affected by OA of the knee, intraarticular corticosteroid injections or EA within the last 3 months, hypersensitivity to NSAIDs or paracetamol, abnormal liver or kidney function tests, evidence of leukopenia and coagulopathies screened by clinical laboratory, concomitantly receiving anticoagulants, history of peptic ulceration, anemia, uncontrolled hypertension, congestive heart failure, hyperkalemia, pregnancy, lactation and malignant tumors. Treatment procedures -------------------- ### 1. Drug administration During a run-in period of one week, the patients refrained from using any NSAIDs or analgesics except for a \"rescue analgesic\" (2 tablets of 500 mg paracetamol orally as needed, up to 4 times daily). The patients who had persistent pain and a Lequesne\'s functional index of at least 6 points at the end of the run-in period were randomized into the four groups mentioned above. Diclofenac sodium, 25 mg film-coated tablets were a gift from Novartis (Thailand) Limited. Placebo and diclofenac were prepared in identical appearance. Either the placebo or diclofenac was prescribed at 1 tablet, 3 times a day immediately after each meal for 4 weeks. In addition, 2 tablets of 500 mg paracetamol were still prescribed as a \"rescue analgesic\" during this study. ### 2. True and placebo EA The true EA treatment was standardized throughout the study. It was performed by the physician acupuncturist who received acupuncture training in the People\'s Republic of China. Four fine stainless steel needles were inserted into acupuncture points around the affected knee \[[@B20]\] as presented in Table [1](#T1){ref-type="table"} and Figure [1](#F1){ref-type="fig"}. All needles were used in order to conduct an electrical current through the points, and were inserted superficially (not more than 0.5 inch approximately in depth). Thus, an elicitation of needle sensation (so-called De Qi) during the insertion of the needles was not intended. The first pair of electrodes was connected to the Dubi and nearest adjacent point (medial Xiyan) and the second pair to the trigger point and Ququan. The electrical stimulation was applied slowly and simultaneously to each pair of needles until it reached the maximum toleration level of the patient. Biphasic pulses were used for the electrical stimulation at a frequency of 2 Hz, and it was administered for 20 minutes in each treatment. The patients were treated 3 times a week (Monday, Wednesday, and Friday) for 4 weeks (12 times). {#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### The selected acupuncture points used in this study. ::: Acupuncture points Location Needling manipulation ------------------------- ----------------------------------------------------------------------------------------------------------------- ------------------------------------------ Dubi (ST-35)^1^ In the depression of the lateral part of the patella ligament, when the knee is bent. Slightly towards the medial side^\*^. Medial Xiyan (Extra)^1^ In the depression of the medial part of the patella ligament, when the knee is bent. Slightly towards the lateral side^\*\*^. Trigger point^2^ At the level of the joint line, midpoint between the medial Xiyan and Ququan. Straight insertion^\*^. Ququan (Liv-8)^2^ At the medial end of the knee crease, in front of the semi-membranous muscle behind the lower end of the femur. Straight insertion^\*\*^. ^1^First pair of elctrodes ^2^Second pair of electrodes ^\*^Stimulated with positive polarity at visit \#1, 3, 5, 7,9,11 and negative polarity at visit \#2, 4, 6, 8, 10, 12. ^\*\*^Stimulated with negative polarity at visit \#1, 3, 5, 7,9,11 and positive polarity at visit \#2, 4, 6, 8, 10, 12. ::: The placebo EA was performed by attaching patch electrodes to the selected acupuncture points. Each electrode was connected to the sound producing dummy mode of the same apparatus, as in the true EA treatment. The duration and frequency of treatment were the same as those in the true EA treatment. Both true and placebo EA were performed by the same physician. Thus, the physician acupuncturist was the only person in the research team who knew which patients received the true or placebo EA. Compliance with treatment was assessed by counting the number of unused tablets (diclofenac or placebo) and the number of times acupuncture treatment was received. During the study period, all additional therapies (e.g. oral or topical NSAIDs, intraarticular corticosteroid injection, other analgesics, chondro-protective agents, surgical procedures on the knee joint etc.) were not allowed. However, all other treatment for concomitant disorders that did not interfere with the study could be continued, but it had to be documented. ### Assessments Clinical assessments were evaluated for base-line data at the end of the run-in period (week 0) and again at the end of the study (week 4). These assessments included the amount of paracetamol tablets taken/week, 50 feet-walk time, a patient\'s global pain as 100 mm visual analog scale (VAS) over the previous 3 days, the Western Ontario and McMaster Universities OA Index (WOMAC: score ranging from 0--96) \[[@B21]\], and Lequesne\'s functional index (score ranging from 0--24) \[[@B22]\]. At the end of this study, the orthopedist\'s and patient\'s opinion of change (much better, better, same, worse, much worse) were evaluated. The patients were considered to be responders if they met the following criteria. Firstly, the number of paracetamol tablets taken in week 4 was less than that at week 0, or less than 14 tablets/week. Secondly, at least 4 of the following 5 outcome parameters showed the following improvement: VAS, WOMAC, Lequesne\'s functional index decreased by at least 50%, and the orthopedist\'s or patient\'s overall opinion of change was better or much better. Clinical assessments in each patient were evaluated by the same physician who was blinded to the treatment. Complete physical examination and non-directive questioning for adverse events were also performed weekly for 4 weeks in order to acquire a safety assessment. Statistical analysis -------------------- One-way analysis of variance (ANOVA) was used to determine whether the four treatment groups differed in mean values of change from a base-line in paracetamol consumption, 50 feet-walk time, VAS, WOMAC and Lequesne\'s functional index. If there was any statistical significance between any of the four groups, the Scheffe method was used to demonstrate statistical significance between each of two groups. Differences among the treatment groups in the overall opinion of change, and number of patients considered to be responders, were evaluated by chi-square or Fisher\'s exact test. Results ======= Of the 200 subjects considered eligible for the study, 5 dropped out and 2 were withdrawn during the run-in period due to hypersensitivity to paracetamol and spontaneous pain relief. The remaining 193 subjects constituted the study population and were randomized into four parallel groups, which were not significantly different in base-line characteristics (Table [2](#T2){ref-type="table"}) and radiographic findings of affected knees (Table [3](#T3){ref-type="table"}). Only 1 patient in each placebo, EA and combined group as well as 2 patients in the diclofenac group received acupuncture treatment prior to this study. ::: {#T2 .table-wrap} ::: {.caption} ###### Base-line characteristics of patients evaluated at the end of the run-in period. ::: Treatment groups -------------------------------------------------- ------------------ ---------------- --------------- --------------- --------- n (M:F) 47 (12:35) 49 (11:38) 48 (10:38) 49 (10:39) 0.93^c^ Age (yr)^\*^ 62.70 (7.22) 62.14 (7.53) 65.10 (3.40) 61.84 (8.95) 0.16^a^ Body weight (kg)^\*^ 60.65 (10.24) 57.65 (10.64) 59.89 (9.74) 59.92 (9.66) 0.49^a^ Height (cm)^\*^ 153.94 (6.45) 151.94 (10.71) 152.19 (5.89) 153.32 (6.73) 0.54^a^ Duration of OA (yr)^\*^ 4.98 (3.32) 3.94 (2.83) 6.09 (4.96) 4.53 (3.86) 0.05^a^ Localization of OA 0.40^c^ Right/Left knee 2/3 2/6 1/4 3/5 Both knees 42 41 43 41 Number of paracetamol tablets taken per week^\*^ 22.06 (13.75) 18.94 (14.68) 21.40 (14.97) 19.04 (14.87) 0.63^a^ 50 ft-walk time (sec)^\*^ 22.04 (4.81) 22.36 (6.00) 24.54 (8.14) 22.77 (5.13) 0.21^a^ VAS^\*^ 63.49 (22.36) 64.79 (23.41) 66.87 (22.34) 57.63 (21.21) 0.21^a^ WOMAC^\*^ Pain index 10.19 (4.20) 11.02 (4.15) 10.25 (3.86) 10.50 (4.18) 0.75^a^ Stiffness index 4.51 (1.71) 4.08 (1.95) 4.35 (2.10) 4.27 (1.57) 0.72^a^ Disability index 37.04 (12.00) 35.65 (12.89) 38.00 (13.18) 37.94 (13.02) 0.79^a^ Total score 51.75 (15.96) 50.76 (17.98) 52.60 (18.13) 52.71 (17.65) 0.94^a^ Lequesne\'s functional index^\*^ 13.78 (3.78) 13.85 (3.22) 14.14 (2.98) 13.73 (2.92) 0.93^a^ ^\*^Data represent mean (SD) ^\*\*^Statistic analysis: ^a^ = one-way ANOVA, ^c^ = chi-square test. ::: ::: {#T3 .table-wrap} ::: {.caption} ###### The radiographic findings at entry into the study^\*^. ::: Treatment groups ------------------------------------------------ ------------------ ---- ---- ---- --------- Kellgren and Lawrence X-ray grade \[[@B32]\] Grade 1 4 1 3 5 0.40^f^ Grade 2 4 6 5 7 0.84^f^ Grade 3 19 15 25 23 0.31^c^ Grade 4 62 68 58 55 0.16^c^ Knee compartment with most severe change of OA Medial tibiofemoral 63 77 69 64 0.73^c^ Lateral tibiofemoral 12 5 13 11 0.24^c^ Patellofemoral 14 8 9 15 0.28^c^ Data represent number of patients. Statistic analysis: ^c^ = chi-square test, ^f^ = Fisher\'s exact test. ::: Of the 193 study patients, 186 (96.37%) completed the study. The remaining 7 patients were withdrawn from the trial due to flare of pain with joint swelling (2 in the placebo and 1 in the EA group), severe GI side effects (3 in the combined group), and flare of pain from an accidental fall not related to treatment (1 in the EA group). Since there were few patients withdrawn from the trial, the results were, therefore, not substantially affected, whether the analysis was performed by an intention to treat analysis or an analysis on available completers. Thus, the following data showed the findings in only 186 available completers. During 4 weeks of treatment, change in body weight compared to the base-line values did not significantly differ among the four groups (one-way ANOVA calculated weekly, data not shown). The rates of compliance with medications (placebo or diclofenac) and acupuncture (placebo or EA) in each group were more than 90%, and comparable among the four groups. At the end of the study, the improvement of symptoms, which were determined by the reduction in mean changes in most outcome parameters (except WOMAC pain index and WOMAC total score), was greatest in the EA group (Table [4](#T4){ref-type="table"}). The mean changes in VAS were significantly different between the EA and placebo group as well as the EA and diclofenac group (Table [4](#T4){ref-type="table"} and Figure [2](#F2){ref-type="fig"}). The mean changes in the Lequesne\'s functional index also differed significantly between the EA and placebo group (Table [4](#T4){ref-type="table"}, Figure [3](#F3){ref-type="fig"}). In addition, there were significant differences in mean changes in the subscale of standardized WOMAC (pain index) between the combined and placebo group (Table [4](#T4){ref-type="table"} and Figure [4](#F4){ref-type="fig"}). The proportion of patients with the orthopedist\'s and patient\'s overall opinion of \"much better\" was greatest in the EA group, followed by the diclofenac, combined and placebo group, respectively. In addition, the proportion of responders was also greatest in the EA group, followed by the combined, diclofenac and placebo group, respectively. However, statistical differences among the four groups were found only in the orthopedist\'s overall opinion and a number of responders (Table [5](#T5){ref-type="table"}). There were no differences between the four groups in the remaining outcome parameters. {#F2} {#F3} {#F4} ::: {#T4 .table-wrap} ::: {.caption} ###### Change in outcome parameters after 4 weeks of treatment^\*^. ::: Treatment groups ------------------------------------------------------ ------------------ --------------- --------------- --------------- ------------- Number of paracetamol taken (tablets taken per week) -5.16 (2.33) -4.43 (1.90) -7.89 (2.09) -5.13 (2.06) NS 50 feet-walk time (sec) -2.70 (0.52) -3.52 (0.46) -4.41 (0.70) -4.13 (0.54) NS VAS -22.86 (4.02) -32.99 (3.94) -48.24 (3.59) -35.59 (2.74) \<0.05^†^ WOMAC Pain index -3.31 (0.68) -4.90 (0.53) -5.65 (0.59) -6.28 (0.77) \<0.05^††^ Stiffness index -1.47 (0.31) -1.55 (0.27) -2.24 (0.31) -2.02 (0.28) NS Disability index -12.33 (1.88) -14.39 (1.77) -19.17 (2.05) -18.98 (1.92) NS Total score -17.11 (2.73) -20.84 (2.43) -27.07 (2.78) -27.28 (2.79) NS Lequesne\'s functional index -3.82 (0.51) -4.80 (0.61) -6.44 (0.59) -5.39 (0.52) \<0.05^†††^ ^\*^Compared values of week 4 versus those of week 0 (run-in peroid), data represent mean (SEM). Statistic analysis: ^a^ = one-way ANOVA and Scheffe method. NS= no statistical significance. ^†^ = EA vs placebo and diclofenac group. ^††^ = combined vs placebo group. ^†††^ = EA vs placebo group. ::: ::: {#T5 .table-wrap} ::: {.caption} ###### Overall opinions of change and number of responders evaluated at week 4. ::: ------------------------------------ ------------------ --------------------- ------------- ------------------- ----------- Treatment groups Parameter Placebo (n = 45) Diclofenac (n = 49) EA (n = 46) Combined (n = 46) *p*-value Orthopedist\'s overall opinion^\*^ 0.01^†^ Much better 6 18 21 16 Better 22 21 20 23 Same 16 10 5 7 Worse 1 0 0 0 Patient\'s overall opinion^\*^ 0.09^†^ Much better 19 25 31 22 Better 16 17 11 23 Same 9 7 4 1 Worse 1 0 0 0 Number of responders 13 18 27 24 0.02^††^ ------------------------------------ ------------------ --------------------- ------------- ------------------- ----------- ^\*^Data represent number of patients, ^†^ = chi-square test evaluated on the proportions of patients with the opinion of \"much better\", ^††^ = chi-square test. ::: The percentage of patients who experienced adverse effects (e.g. gastrointestinal and central nervous system symptoms, rash, edema, and hypertension) during this study did not differ between the four groups (data not shown), whereas, local contusions around the knee were common in the EA and combined group (approximately 45%). However, the contusions usually disappeared within 5--7 days. When the responders in each group were followed up for 2 months, the proportion of remaining responders was not significantly different between the four groups (Table [6](#T6){ref-type="table"}). ::: {#T6 .table-wrap} ::: {.caption} ###### Number of responders considered at the end of the study (week 4) and at 1 and 2 month(s) after treatment^\*^. ::: 1 month after treatment 2 months after treatment ----------------- ---- ------------------------- -------------------------- Placebo 13 11/12 9/12 Diclofenac 18 10/16 7/15 EA 27 21/25 19/24 Combined 24 14/24 14/24 *p* value^\*\*^ 0.12 0.19 ^\*^ Only the responders at the end of the study were followed up for 2 months. ^†^Some patients were unable to be evaluated due to loss of follow up or use of NSAIDs for other purposes during the follow-up period. ^\*\*^ Fisher test between the four groups. ::: Discussion ========== Although there are several lines of evidence from many controlled and uncontrolled studies for the short-term and long-term effectiveness of acupuncture in relieving clinical pain \[[@B23]-[@B28]\], the scientific data concerning the efficacy of acupuncture in OA are rare \[[@B19]\]. In addition, there are several systemic flaws \[[@B29]\] among these studies due to inadequate statistical power, inadequate sessions of acupuncture treatment, failure to control concomitant therapies, or no sham/placebo acupuncture controlled group. In this study, we minimized the methodological limitations of previous studies by using the randomized, single-blind, placebo controlled design with a larger sample size of OA patients coupled with standard outcome assessments. By using the percentage of the responders as the main efficacy criterion, the comparison between true and sham acupuncture needs at least 61 patients per group, whereas, only 35 patients per group are needed to compare between true and placebo acupuncture performed by not puncturing the skin \[[@B23],[@B30]\]. In order to increase the ability of differentiating true from placebo effects and minimize the sample size, we selected the procedure of attaching the acupuncture points with the patch electrodes as placebo EA, and at least 45 completers per group were treated in this trial. In this study, a double-blind design was considered inappropriate, since we used patch electrodes as placebo EA and our patients might have recognized the difference between true and placebo EA. A single blind was, therefore, a reasonable alternative. The acupuncture points used were selected because we intended to determine only the effects of local points around the affected knee. This applied especially to the medial aspect, which related to the knee compartment that was frequently involved in OA. Using these local points coupled with the needling technique (developed by Chawal Kanchanakul, acupuncturist of Dhammanamai Foundation, Chiang Mai, Thailand) demonstrated a rather simple, convenient, less painful, and more acceptable method for Thai patients, and it was effective in our pilot study. The points selected here were therefore different from other trials \[[@B24]-[@B26],[@B28]\], which also included the distal points at medial and lateral aspects of the leg. Low-frequency (2 Hz) EA was selected because it produces an analgesia of long duration, which outlasts the 20-min stimulation session by 30 min to many hours. In addition, its effects are cumulative after several sessions of treatment given either daily or less frequently (2--3 times a week) \[[@B30]\]. For these reasons, the low-frequency EA in this study was therefore given 3 times a week for 4 weeks, as commonly recommended in EA practice. However, to balance the acupuncture point stimulation by positive and negative polarities, each point was stimulated 6 times with positive and negative polarities in an alternate sequence during 12 sessions of treatment. In addition, each pair of electrodes was connected to each pair of adjacent points in order to obtain equal electrical sensation in each point during stimulation. In this study, the clinical responses observed in the placebo group might result from, 1) the placebo effect or natural fluctuations in the symptoms of OA that are unrelated to the analgesic effect of paracetamol, because some patients demonstrated reductions in these scores without or with minimal analgesic need, or 2) the direct effects of paracetamol as a rescue analgesic. The latter reason made the placebo group not absolutely inert because paracetamol is also the first-line drug in the treatment of OA of the knee. However, the use of a rescue analgesic could not be avoided due to ethical reasons. At week 4, the improvement of symptom scores (except change in WOMAC pain index and WOMAC total score) and the number of responders/patients with the opinion of \"much better\" were greatest in the EA group. These data indicate the great potential of EA in the symptomatic treatment of OA of the knee. This study also demonstrated that EA was significantly more effective than placebo with respect to changes in VAS and Lequesne\'s functional index, and significantly more effective than diclofenac with respect to the changes in VAS. This superiority of EA indicates its genuine efficacy \[[@B17],[@B23],[@B30]\], which was more effective than placebo (or diclofenac) in this study. A previous trial revealed 34% and 14% reductions in the mean values of WOMAC total score and Lequesne\'s index, respectively after 4 weeks of acupuncture treatment \[[@B28]\], whereas, our study demonstrated a 50% and 45% reduction, respectively. These discrepancies might be due to differences in acupuncture point selection, number of points and the electrical stimulation technique used, or electrical stimulation parameters. The clinical responses in the combined group were slightly superior to the diclofenac group, but not to the EA group (except for the tendency of being superior with respect to a change in the subscale of the standardized WOMAC: pain index). These responses indicate that EA may exert an adequate analgesic effect, a combination with diclofenac may not be of further benefit. Although a combination was not more efficacious than EA alone, it was significantly more so compared to placebo in terms of improvement in the WOMAC pain index, while EA was not. However, the nonsignificant change in clinical scores between the diclofenac and placebo group indirectly suggests that diclofenac may be as effective as paracetamol (as needed) in symptomatic treatment of OA of the knee. The nonsignificant change in paracetamol consumption among the four groups during treatment might result from the unnecessary use of paracetamol in a great majority of active treatment groups, despite a significant improvement in some parameters of the EA and combined group, and a tendency to improve parameters in the diclofenac group. Another possibility might be the large variation in paracetamol consumption in each group that contributed to a false negative result. The nonsignificant change in the 50 feet-walk time among the four groups suggests that this parameter may not be sensitive enough to demonstrate existing differences. Walk time determined during stair climbing should thus be a better alternative \[[@B31]\]. The failure to demonstrate the differences in other parameters (i.e. change in WOMAC stiffness, disability, and total scores) between the active treatment and placebo group may be due mainly to the rescue analgesic and partly to inadequacy of the sample size. This study demonstrated that EA treatment was safe and free from serious adverse effects. However, the indifference in adverse events between the diclofenac and placebo group might be due to the exclusion of patients who were at high risk to the adverse effects of NSAIDs during the screening visit or short-term trial. In this study, the rescue analgesic paracetamol contributed to several confounding effects. Therefore, the randomized placebo controlled trial without using rescue analgesic should be investigated further to confirm the effectiveness of this range of active treatment, especially in EA and its combination with NSAIDs. In summary, EA was significantly more effective than placebo regarding reductions in 100 mm VAS and Lequesne\'s functional index, but was significantly more effective than diclofenac in only the reduction of 100 mm VAS. The combination of EA and diclofenac treatment was more effective than placebo with respect to the reduction in the subscale of standardized WOMAC (pain index), but not more effective than EA treatment alone. Local contusions were minor adverse effects commonly found in the EA and combined group. The positive effects far outweigh the serious adverse ones of EA, which make this procedure an attractive alternative treatment for patients with OA of the knee. Competing Interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ This work was supported by the Faculty of Medicine, Chiang Mai University, Thailand.

Background ========== Plants in the family Cruciferae, such as cabbage (*Brassica olerecea*), mustard (*Sinapis alba*) and crambe (*Crambe abyssinica*) produce glucosinolates which are secondary metabolites that can help plants defend themselves against phytophagous insects, fungi and other pests. Glucosinolates are hydrolyzed by thioglucosidase (myrosinase) enzymes to yield an aglycone which undergoes non-enzymatic rearrangements to produce organic isothiocyanates, thiocyanates, nitriles and other products \[[@B1],[@B2]\], although non-enzymatic thermal degradation of glucosinolates may produce chemical species identical to products of the enzymatic hydrolysis \[[@B3],[@B4]\]. The chemical structures of the hydrolysis products (breakdown products) of glucosinolates vary depending on the precursors in different plant species, and different endogenous or exogenous factors such as pH and ferrous ions \[[@B2],[@B5],[@B6]\]. Glucosinolates and their breakdown products have been the focus of many studies because of the possibility of using them as natural pesticides \[[@B7],[@B8]\]. However, it is considered that glucosinolates themselves possess limited biological activity until they are hydrolyzed \[[@B9]\]. The breakdown products are generally relatively small molecules which make many of them volatile. Many of these volatiles have been shown to act as attractants for certain insects seeking food or egg-laying sites rather than direct insecticidal activity \[[@B10]-[@B14]\]. Allyl glucosinolate (sinigrin) breaks down in soils to allyl isothiocyanate and allyl cyanide, and these breakdown products and glucosinolate-containing plant tissue have been suggested for use in controlling soil-borne plant pests \[[@B9],[@B15]-[@B19]\]. Despite the many efforts that have been devoted to exploring glucosinolates and their breakdown products as naturally occurring pesticides, few products have been marketed commercially. 2-(2-Butoxyethoxy)ethyl thiocyanate (Lethane 384^®^) and isobornyl thiocyanoacetate (Thanite^®^) were marketed as insecticides, and methylisothiocyanate (Vorlex^®^) as a fumigant, although their discovery may or may not have been directly related to glucosinolates. The thiocyanates (Lethane 384 and Thanite) were discontinued by their manufacturers in 1984 \[[@B20]\] and methylisothiocyanate, the closest relative of the glucosinolate breakdown products, was also banned in 1991 \[[@B21]\]. Harvested crops including seeds, grains and cereals suffer a loss of at least 10% from insect pests during storage. Losses of 30% are common throughout large areas of the world \[[@B22]\]. Many chemical fumigants such as ethylene dibromide and methyl bromide that have been used to control these stored grain insects are being banned or facing phase-out. There are clearly needs for alternative fumigants. We have reported the insecticidal activity of the glucosinolates in crambe in our previous study, and found that allyl, 3-butenyl, 4-pentenyl, 4-hydroxybenzyl and 2-hydroxy-3-butenyl glucosinolate were the major glucosinolate components in crambe seed meal \[[@B8]\]. We also found that 2-hydroxy-3-butenyl glucosinolate was the predominant glucosinolate among the crambe glucosinolates, and its major breakdown product (\>95% of total glucosinolate) was 1-cyano-2-hydroxy-butene (CHB) \[[@B8]\]. CHB and other crambe glucosinolate breakdown products were found to have fumigation toxicity against insects in our early study \[[@B8]\]. Different fumigants have different insecticidal mode of actions, and some fumigants are known to affect the respiratory function of insects \[[@B20]\]. Although it has been reported that the CO~2~ expiration of *Tenebrio molitor* was reduced by feeding glucosinolate-containing rapeseed meal \[[@B24]\], there are no previous studies on the effect of the glucosinolate breakdown products on insect respiration. In this paper we report the insecticidal activity of crambe-related glucosinolate breakdown products as fumigants and the effect of these chemicals on the respiratory function of insects. The latter is achieved by measuring the CO~2~ emission of the insects that are exposed to the chemicals. The fumigation effects of the breakdown products were compared and discussed together with the known fumigants dichlorovos and chloropicrin. The effect of the glucosinolate breakdown products on CO~2~ emission was also compared with that of hydrogen cyanide and rotenone. Results and Discussion ====================== The house fly (*Musca domestica*) and the lesser grain borer (*Rhyzopertha dominica*) were chosen in the laboratory toxicity test, based on the availability and their significance as pests. In fact, glucosinolate-containing extracts from crambe were found to be toxic to the larvae of the house fly \[[@B8]\]. In the present study, we have found that some of the individual glucosinolate breakdown products from crambe were insecticidal toward *M. domestica* as potent fumigants. The respective LC~50~ of allyl thiocyanate (ATC), allyl isothiocyanate (AITC) and allyl isocyanate(AIC) were 0.1, 0.13 and 0.63 μg cm^-3^ (Table [1](#T1){ref-type="table"}). This is in the same potential range of the toxicity of chloropicrin (0.08 μg cm^-3^), a commercial fumigant. CHB, the major breakdown product of crambe glucosinolates was not as potent a fumigant as others. Its LC~50~ was 6.2 μg cm^-3^ (Table [1](#T1){ref-type="table"}). ::: {#T1 .table-wrap} ::: {.caption} ###### Fumigation toxicity of glucosinolate breakdown products against the house fly adults in comparison with selected fumigants (exposed for 24 h). ::: compound LC~50~ (μg cm^-3^) 95% FL ---------------------------- -------------------- -------------- dichlorovos \<0.02 -- chloropicrin 0.08 0.076--0.099 allyl thiocyanate 0.1 0.08--0.12 allyl isothiocyanate 0.13 0.10--0.16 allyl isocyanate 0.63 0.56--0.70 allyl cyanide 3.66 3.11--4.10 1-cyano-2-hydroxy-3-butene 6.2 4.91--9.73 control -- -- ::: Different insects have different susceptibilities toward chemicals. Although dichlorvos was almost 10 times as toxic as chloropicrin in house fly test (Table [1](#T1){ref-type="table"}), it was only 5 times more toxic in the lesser grain borer test (Table [2](#T2){ref-type="table"}). ATC, the most potent fumigant against the house fly was the most toxic against the lesser grain borer as well. Its LC~50~, 0.55 μg cm^-3^, against the latter insect was between that of dichlorvos (0.29 μg cm^-3^) and chloropicrin (1.3 μg cm^-3^). LC~50~\'s of allyl isothiocyanate (1.57 μg cm^-3^) was in the same range of toxicity as chloropicrin. Allyl isocyanate (2.2 μg cm^-3^) and allyl cyanide (2.8 μg cm^-3^) were slightly less toxic compared to chloropicrin, although they were all about 1/10 as toxic as dichlorvos (Table [2](#T2){ref-type="table"}) Allyl isocyanate was tested to compare with the glucosinolate breakdown products. ::: {#T2 .table-wrap} ::: {.caption} ###### Fumigation toxicity of glucosinolate breakdowns products against the lesser grain borer in comparison with selected fumigants (exposed for 24 h). ::: compound LC~50~ (μg cm^-3^) 95% FL CO~2~^\*^ ---------------------------- -------------------- ------------ ----------- dichlorovos 0.29 0.21--0.41 -- chloropicrin 1.3 1.20--1.42 -- allyl thiocyanate 0.55 0.50--0.60 5.1 allyl isothiocyanate 1.57 1.47--1.67 5.1 allyl cyanide 2.8 2.26--3.48 -- 1-cyano-2-hydroxy-3-butene \>19.60 -- -- allyl isocyanate 2.2 2.06--2.34 10.4 hydrogen cyanide -- -- 4.2^\*\*^ control -- -- 2.7 control (HCN) -- -- 4.6^\*\*^ ^\*^ CO~2~ in μg/mg body weight by *B. germanica.* Exposed for 1 h. ^\*\*^ Done in a separate experiment. ::: The high fumigant toxicity of the glucosinolate breakdown products against the lesser grain borer may have important economic impact on stored grain insects management. These naturally occurring chemicals were also insecticidal as fumigant on other stored grain insects such as the sawtoothed grain beetle (*Oryzeaphilus surinamensis* L.) and the red flour beetle (*Tribolium castaneum* Herbst) (Tsao, et al., unpublished data, 1997). Long-term fumigation tests against the lesser grain borer showed little improvement in the toxicity of the glucosinolate breakdown products. Also, the 2-day and 4-day tests resulted in higher mortality in the control (data not shown). Little is known about the mode of action underlying the high fumigation toxicity of the glucosinolate breakdown products. Dichlorvos exerts its insecticidal toxicity against the insect nervous system, by inhibiting acetylcholinesterase. Methyl bromide and chloropicrin\'s insecticidal mode of action has been shown as their reactions with the nucleophilic sites such as OH, SH or NH~2~ groups in vital enzymatic systems \[[@B25]\]. Many non-fumigant insecticides are known to affect the respiratory function of insects (e. g., rotenone, dinitrophenols) \[[@B20]\]. In our fumigation test technique, inhalation (intake through the spiracles) was the only way for the insects to receive the volatile chemicals, although certain degree of absorption could occur through the cuticle. Monitoring the CO~2~ production of the insects at sub-lethal doses in a closed chamber may provide valuable symptomatic information toward understanding how these chemical volatiles affect the respiration of insects and lead to the eventual death of the insects. Monitoring CO~2~ emission is one of the well established technique to study insect respiration and the effect of chemicals on the respiratory function, although it has sometimes been accompanied with simultaneous monitoring of oxygen consumption \[[@B24],[@B26]\]. The quantity of the chemicals used in our CO~2~ test was based on the toxicity tests of the chemicals (data not shown). At the sub-lethal levels, no mortality was observed within the test period. For AIC, ATC and AITC, the amount of chemical utilized was 200 μl of 10^4^ ppm corn oil solution. This is equivalent to ca. 10 μg cm^-3^ of the chemical in air. This concentration is lethal to the lesser grain borer and the house fly, but not to the cockroaches due to different physiological aspects and the shorter exposure time in the CO~2~ monitoring test. By exposing *B. germanica* to the fumes of the glucosinolate breakdown products AIC, ATC and AITC, we found that the CO~2~ production of this insect was significantly increased at each time interval (Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}, Fig. [2](#F2){ref-type="fig"}). AIC was the strongest CO~2~ emission (respiration) enhancer. *B. germanica* treated with AIC produced more than 3 times more CO~2~ than that in the control after being exposed for 1 h and 2 h. Even after 4 h, the CO~2~ emission was more than 75% higher for AIC (Fig. [2](#F2){ref-type="fig"}). Although ATC and AITC did not increase the CO~2~ emission as strongly as did AIC, these two chemicals still showed significant effects after 2 h. The effect was not as significant at 4 h for ATC and AITC (Fig. [2](#F2){ref-type="fig"}). {#F2} HCN and rotenone were tested for comparing the mode of action of glucosinolate breakdown products, because they are potential inhibitors of CO~2~ production \[[@B20]\]. AITC may have a different mode of action on the respiratory system than HCN and rotenone as the CO~2~ production of *P. americana* was significantly increased by AITC as compared to the control (Fig. [3](#F3){ref-type="fig"}). The CO~2~ emission of *P. americana* was not significantly affected by HCN or rotenone during the test period (Fig. [4](#F4){ref-type="fig"} and [5](#F5){ref-type="fig"}), indicating that these chemicals may not affect the respiratory system of the insect under the dose used in this experiment. The less effectiveness could be due in part to the low concentration of HCN and rotenone. Effort to use extended exposure time was also tried, however, high morality was observed when *P. americana* was exposed to HCN and rotenone at higher dose, or for longer time (\>60 min and \>35 min, respectively). There was no particular reason except availability for using *P. Americana* instead of *B. germanica.* {#F3} {#F4} {#F5} Toxicity of isothiocyanates has been attributed to inactivation of the thiol group of essential enzymes \[[@B27]\]. The mode of action of many isothiocyanate compounds has also been attributed to their capability for alkylating the nucleophilic groups of biopolymers such as DNA, thus having cytotoxic properties which can affect the formation of the spiracular epidermis and crochet on the prolegs of tobacco hornworm (*Manduca sexta* L.) caterpillars \[[@B28]-[@B30]\]. Our bioassay avoids totally any direct contact between the insects and the chemical, thus the insecticidal effect is very likely caused by inspiration of the chemical by the insects through their spiracles. The increased CO~2~ emission observed in this study indicated a different mode of action of these glucosinolate breakdown products. This new mode of action may be similar to that of the dinitrophenols, although further investigation is needed to prove this speculation. Compounds such as dinitrophenol are uncouplers or inhibitors of oxidative phosphorylation \[[@B31]\]. When the coupling between the respiratory chain and phosphorylation is broken, respiratory control is lost, and the electron transport along the chain occurs at full speed without producing ATP, the high energy molecule. Uncouplers kill insects not by inhibiting normal electron transport of the respiratory chain, but rather by enhancing the respiration \[[@B27],[@B31]\]. When the energy-transfer (production of ATP) is blocked, and at the same time, the respiration is accelerated, the insects have to use their stored energy sources and the exhaustion of those resources then lead directly to the death of the insects \[[@B31]\]. Insecticides such as rotenone and HCN are known to affect the respiratory function by inhibiting certain enzymes in the respiratory electron transport chain of insects, thus reducing the oxygen consumption and perhaps CO~2~ production as well \[[@B20]\]. This may explain why HCN did not increase the CO~2~ emission at low dose. As with many other conventional fumigants, we also found that some of the glucosinolate breakdown products and related chemicals were strong hatching inhibitors for the soybean cyst nematode (*Heterodera glycines*) (unpublished data, 1996) Conclusions =========== Using glucosinolate breakdown products as safer fumigants in pest control is a novel approach in integrated pest management, and at the same time, since these natural products are considered fully biodegradable and less toxic, considering many cruciferous vegetables are consumed daily, there should be fewer environmental concerns. The bioactivity of these compounds may be from their effect on the insect respiratory system indicated by the inhibited or accelerated CO~2~ production. Materials and Methods ===================== Chemicals --------- Dichlorovos and chloropicrin were purchased from ChemService (West Chester, PA). Allyl isothiocyanate, allyl isocyanate, allyl cyanide and rotenone were purchased from Aldrich Chemical Co. (Milwaukee, WI). Allyl thiocyanate was synthesized in this laboratory by reacting allyl chloride with potassium thiocyanate (KSCN) in dimethyl sulfoxide (DMSO). Briefly, KSCN (0.31 mol) and DMSO (200 ml) were added to a 1-L flask and heated on a water bath to 85°C followed by addition of allyl chloride (0.30 mol) to the flask. The mixture was allowed to cool to room temperature, and then the product ATC was extracted with ethyl ether (200 ml × 4). ATC was purified using column chromatography on silica gel with 2/10 ether/hexane, and the structure was confirmed by nuclear magnetic resonance (NMR) and infrared (IR) spectroscopy: ^1^H NMR δ~TMS~ (CDCl~3~) ppm: 3.58 (2H, d, -C[H]{.ul}~2~-SCN), 5.34--5.43 (2H, dd, C[H]{.ul}~2~-CH-), 5.86--6.02 (1H, m, CH~2~-C[H]{.ul}-); IR: 2930--2960, 2870, 2160, 1640,1400, 1230, 985, 930 cm^-1^. l-Cyano-2-hydroxy-3-butene (CHB) was extracted from the defatted crambe meal. One hundred g of the crambe meal (National Sun Inc., Enderland, ND) was extracted with CH~2~Cl~2~ at 45°C in a Soxhlet apparatus overnight, and isolated and purified by using column and thin layer chromatography. The structure of CHB was confirmed by NMR and mass spectrometry: ^1^H NMR δ~TMS~ (CDCl~3~) ppm: 2.60 (2H, dd, -C[H]{.ul}~2~-CN), 2.75 (H, brs, -CH(O[H]{.ul})-), 4.37--4.50 (1H, m, -C[H]{.ul}(OH)-), 5.25--5.45 (2H, m, C[H]{.ul}~2~-CH-), 5.85--5.98 (1H, m, CH~2~-C[H]{.ul}-); IR: 3480, 3000, 2930, 2270, 1640, 14251135, 1070, 1000, 945 cm^-1^. The above NMR was consistent with that of Das and Torssell \[[@B32]\] and the IR data is in accordance with Daxenbichler *et al.*\[[@B33]\]. Fumigation tests ---------------- ### Fumigation of the House Fly Approximately 1 g 1/1 dry sucrose/powdered milk was placed in the bottom of a 40-ml jar, along with a 2-cm piece of cotton roll moistened with distilled water. Ten to 15 house fly adults (*Musca domestica* L., laboratory reared Orlando regular strain, one week post-emergence) were anesthetized with carbon dioxide and placed in the 40-ml jars. The jars were each covered with nylon mesh secured by a rubber band. Three 40 ml jars were placed in the bottom of a 2745 ml amber jar. A filter paper (Whatman No. 1, 9-cm diameter) was folded in quarters and placed in the jar as well. About 30 min after the flies recovered from anesthesia, 200 μl of the appropriate corn oil (Flavorite™ pure corn oil, Preferred Products, Inc., Chaska, MN) solution of the chemicals was applied to the filter paper with an eppendorf pipette, the disposable tip of which was left in the jar. The control was treated under the same conditions with a corn oil blank. The above jar was tightly sealed and left for 24 h, when mortality was read. The insects were considered dead when they displayed no observable motion when prodded. LC~50~ was calculated based on 5 concentrations. ### Fumigation of the lesser grain borer Ten adult lesser grain borers (*Rhyzopertha dominica* Fabricius, laboratory reared Manhattan KS strain) were placed in a 1.5 × 5-cm glass tube with 2 g whole wheat kernels. The tubes were closed on both ends by a tight metal mesh secured with parafilm. Three tubes were tied together with a wire, and suspended from the top of a jar (490 ml). A filter paper (same as used in the fly test), was folded in quarters and placed at the bottom of the jar. One hundred μl of the appropriate corn oil dilution of the chemicals was applied to the filter paper with an eppendorf pipette, the disposable tip of which was left in the jar. Corn oil was used in the control. The jar was tightly capped and left for 24 h, at which time the mortality was recorded. An insect was considered dead when it displayed no observable motion upon prodding. LC~50~ was calculated based on 5 concentrations. ### Measuring CO~2~ emitted by insects We chose the German (*Blattella germanica* L.) and the American cockroach (*Periplaneta americana* L.) in this test because they are non-flying, thus producing a steady and measurable quantity of CO~2~. Two adult *B. germanica* (1 male, 1 female without ootheca) or 2 nymphs (average weight 0.55 g) of *P. americana* were weighed and put in a glass tube (5 × 1.5 cm) with both ends covered with mesh screen. A bigger size of tubing (5 × 2 cm) was used for the *P. americana.* The tube was then placed into a glass bottle (220 ml) together with a quarterly folded filter paper (same as used above). A rubber stopper with glass inlet and outlet was tightly capped on the bottle with both ends sealed with a rubber septum (Fig [1](#F1){ref-type="fig"}). An infrared carbon dioxide analyzer (LIRA^®^ Infrared gas analyzer, Model 3000, Mine Safety Appliances Company, Pittsburgh, PA) was used to measure the emitted CO~2~. The sealed bottle was flushed with CO~2~-free air (compressed air filtered through soda lime tubes and a NaOH trap) before introducing the chemical. A sub-lethal dose of the breakdown products (200 μl of 10000 ppm chemical in corn oil for the *B. germanica,* and 100 μl of 1000 ppm for *P. americana*) was immediately injected with a syringe through the septum onto the filter paper inside the system (Fig. [1](#F1){ref-type="fig"}). In this system, direct contact between the insects and the chemical is completely avoided. The CO~2~ emitted by the test insects was recorded at Range 1 (the most sensitive range) of the instrument at different exposure times. There were three replicates for each treatment unless stated otherwise, and the same volume of corn oil was used in the control. Tests on the effect of HCN were achieved by wetting the filter paper with 100 μl of 1 N HCl, and, after flushing out the CO~2~, injecting 20 μl of 1 N NaCN aqueous solution onto the paper. Rotenone (5 μl of 100 ppm in acetone) was injected to the thorax of the *P. americana* nymph in the CO~2~ test and same the amount of acetone was used in the control. Water or acetone was used in the control for HCN or rotenone. {#F1} The recorded peak meter readings (PMR) were converted to mg of CO~2~ according to the following formula \[[@B34]\]: CO~2~ (mg) = 0.0092+0.00795\*PMR+0.0000002\*(PMR)^2^ Data analysis ------------- Probit analysis was used to determine LC~50~, 95% FL, and slopes according to methods outlined by Finney \[[@B35]\]. Means were separated using least significant differences (LSD) (P = 0.05) \[[@B36]\]. Acknowledgements ================ The authors thank Dr. T. A. Anderson of the Texas Tech University, Texas, USA for his assistance in the use of the infrared carbon dioxide analyzer.

Background ========== Transforming growth factor beta (TGF-β) family proteins are multifunctional cytokines capable of regulating cell growth, extracellular matrix protein synthesis, and immune cell functions \[[@B1],[@B2]\]. Hyaluronidase is an extracellular matrix-degrading enzyme. Interestingly, malignant and metastatic breast and prostate cancer cells frequently overexpress both hyaluronidase and TGF-β proteins \[[@B3]-[@B7]\]. Previously we have shown that bovine testicular hyaluronidase, also known as PH-20, increases the susceptibility of various cancer cells to tumor necrosis factor (TNF or TNF-α) cytotoxicity \[\[[@B8]\], review\]. PH-20 induces the expression of proapoptotic p53 \[[@B9]\] and WOX1 (WW domain-containing oxidoreductase; also known as WWOX or FOR) \[[@B10]\], which contributes to the increased TNF cytotoxic effect. PH-20-induced downregulation of the extracellular matrix inter-α-inhibitor is also associated with the enhanced TNF cytotoxicity \[[@B9]\]. TGF-β family proteins, including β1,β2 and β3, induce TNF-resistance and suppress the PH-20 effect of increasing TNF cytotoxicity in murine L929 fibroblasts \[[@B8],[@B11]\]. TGF-β1 induces a novel extracellular matrix protein that prevents TNF-mediated cell death and blocks the activation of extracellular signal-regulated kinase (ERK; also known as p42/44 mitogen-activated protein kinase, p42/44 MAPK) in L929 cells \[[@B8],[@B12]\]. Additionally, TGF-β1 induces the expression of TIAF1 (TGF-β-induced antiapoptotic factor) \[[@B13]\] and TIF2 (TGF-β-induced factor 2) \[[@B14]\] that inhibit TNF cytotoxicity. Whether these TGF-β1-induced proteins restrict the ability of PH-20 to increase TNF cytotoxicity is unknown. PH-20 blocks TGF-β1-mediated growth suppression of epithelial cells \[[@B8],[@B15]\]. Additionally, PH-20 rapidly activates ERK in L929 cells and TGF-β1 reduces the PH-20-induced ERK activation \[[@B16]\]. In contrast, TGF-β1 synergistically increases PH-20-mediated inhibition of staurosporine apoptosis \[[@B16]\]. Thus, PH-20 and TGF-β1 are physiological antagonists for only certain cellular responses. Hyaluronidases Hyal-1 and Hyal-2 play a critical role in cancer invasion and metastasis \[[@B3]-[@B5]\], although inactivation of *HYAL1* gene has been shown in head and neck squamous cell carcinomas \[[@B17]\]. Hyal-1, also known as Luca-1, is a lysosomal enzyme that is secreted from cells. Hyal-1 is a candidate tumor suppressor \[[@B5],[@B18]\], although it enhances extravasation and metastasis of prostate cancer cells \[[@B3]\]. Also, mutations in Hyal-1 may contribute to the pathogenesis of a lysosomal disorder, mucopolysaccharidosis IX \[[@B19]\]. Hyal-2 is a lysosomal protein, which is active at low pH \[[@B20]\]. This protein is also known to be a glycosylphosphatidylinositol (GPI)-anchored cell-surface receptor for jaagsiekte sheep retrovirus \[[@B21]\]. Testicular hyaluronidase PH-20 counteracts TGF-β1-induced TNF-resistance in L929 cells \[[@B8]\]. The goal of the present study was to examine whether lysosomal Hyal-1 and Hyal-2 enhance TNF cytotoxic function and counteract TGF-β1-mediated TNF-resistance in L929 cells. It was determined that stable expression of murine Hyal-2 in L929 cells increased their sensitivity to TNF-mediated death, whereas Hyal-1 was somewhat less effective. TGF-β1 blocked this increased sensitivity. In contrast, Hyal-1 and Hyal-2 did not enhance L929 cell death by anticancer drugs such as daunorubicin, actinomycin D, staurosporine and camptothecin. Overexpressed Hyal-1 and Hyal-2 induced the expression of proapoptotic WOX1, but not p53, suggesting a role of WOX1 in the increased cellular susceptibility to TNF cytotoxicity. Additional mechanisms that are associated with the increased TNF cytotoxicity were also investigated. Results ======= As summarized in Figure [1](#F1){ref-type="fig"}, PH-20 induces the expression of proapoptotic p53 and WOX1 \[[@B9],[@B10]\] and downregulates antiapoptotic matrix inter-α-inhibitor \[[@B9]\], thereby enhancing TNF susceptibility in L929 cells. Also, PH-20 rapidly activates ERK and c-Jun *N*-terminal kinase (JNK1 and JNK2) \[[@B16]\]. Whether these signaling events lead to induction of p53 and WOX1 is unknown. ![Proposed basis for the antagonistic actions of PH-20 and TGF-β1 on L929 cell susceptibility to TNF cytotoxicity. Testicular hyaluronidase PH-20 enhances TNF cytotoxicity in L929 fibroblasts (see green), whereas TGF-β1 induces TNF-resistance in these cells (see red). Both upregulation of proapoptotic p53 and WOX1 and downregulation of antiapoptotic matrix inter-α-inhibitor are involved in the PH-20-increased TNF susceptibility in L929 cells \[[@B8]-[@B10]\]. Also, PH-20 rapidly activates ERK (or p42/44 MAPK) and c-Jun *N*-terminal kinases (JNK1 and JNK2) \[[@B16]\]. In contrast, TGF-β1 protects L929 cells from TNF cytotoxicity by upregulating antiapoptotic TIAF1, TIF2 and a novel matrix protein \[[@B11]-[@B14]\]. TGF-β1 counteracts PH-20-induced TNF-susceptibility in L929 cells \[[@B8]\]. TGF-β1 suppresses PH-20-mediated ERK activation \[[@B16]\], whereas ERK activation is not related with PH-20-increased TNF susceptibility in L929 cells \[[@B15]\]. Whether the proapoptotic p53 and WOX1 block the antiapoptotic function of TIAF1 and TIF2 is not known.](1471-2121-3-8-1){#F1} In contrast, TGF-β1 protects L929 cells from TNF cytotoxicity by upregulating the expression of antiapoptotic TIAF1, TIF2 and a novel matrix protein \[[@B11]-[@B14]\]. TGF-β1 blocks PH-20-increased TNF susceptibility \[[@B15]\]. Likewise, PH-20 inhibits TGF-β1-induced TNF-resistance \[[@B15]\]. TGF-β1 suppresses PH-20-mediated ERK activation \[[@B16]\]; however, the ERK activation is not related with PH-20-increased TNF susceptibility in L929 cells \[[@B15]\]. Whether the proapoptotic p53 and WOX1 block the function of antiapoptotic TIAF1 and TIF2 is not known. In this study, experiments were designed to determine whether lysosomal Hyal-1 and Hyal-2 enhance TNF cytotoxic function and counteract TGF-β1-mediated TNF-resistance in L929 cells. The full-length murine hyaluronidase *Hyal1* and *Hyal2* cDNAs were found in the EST database (GenBank accession AA688635 and BF139787, respectively) and the determined DNA sequences have been deposited in the GenBank (*Hyal1*, AF422176; *Hyal2*, AF422177). The *Hyal1* cDNA encodes a 463-amino-acid protein, which has 99.1% identity to a reported murine sequence (AF011567) \[[@B18]\]. Differences in the protein sequences are Gly-247, Gly-284 and Cys-450 in our Hyal-1 protein, compared to Arg-247, Glu-284 and Arg-450 in AF011567. Transitions in these amino acid residues are non-conservative. The murine *Hyal2* cDNA (AF422177) encodes a 473-amino-acid protein. The deduced protein sequence is identical to a reported murine sequence (AF302844), but has an Ile-355 to Val-355 transition compared to another clone (AF302843). Variation in the protein sequence is also found in a murine Hyal-2 homologue (NP_034619) \[[@B22]\], which possesses three unique 2-, 5- and 7-amino-acid segments, not identical to the above Hyal-2 proteins. L929 cells were engineered to stably express GFP-Hyal-1 or GFP-Hyal-2 fusion proteins. The presence of these proteins in these cells was demonstrated in Western blotting using anti-GFP antibody (data not shown). Exogenous bovine testicular hyaluronidase PH-20 stimulates rapid activation of c-Jun *N*-terminal kinases (JNK1 and JNK2) in L929 cells \[[@B16]\]. In contrast, the stably expressed Hyal-1 and Hyal-2 could not mediate constitutive activation of JNK1 and JNK2 in L929 cells (data not shown). Exposure of the Hyal-2-expressing L929 cells to TNF for 16--24 hr resulted in increased cell death by \~60--110%, compared to the control GFP-expressing cells (Fig. [2](#F2){ref-type="fig"}). These results are similar to the bovine testicular PH-20-increased TNF cytotoxicity of L929 cells \[[@B15]\]. In contrast, there were no significant differences in the extent of cell death when GFP-Hyal-2 and control GFP-expressing cells were exposed to staurosporine, actinomycin D, daunorubicin, or camptothecin for 16--24 hr (Fig. [2](#F2){ref-type="fig"}). ![Hyal-2 increases L929 cell sensitivity to TNF cytotoxicity. L929 cell lines that stably expressed GFP and GFP-Hyal-2 were established (see Materials and Methods) and exposed to TNF-α, staurosporine, actinomycin D (ActD), daunorubicin, or camptothecin for 24 hr. The extent of cell death were measured by staining the cells with crystal violet as previously described (mean ± standard deviation; n = 8) \[[@B10]\].](1471-2121-3-8-2){#F2} There was also an increased susceptibility to TNF cytotoxicity in the Hyal-1-expressing L929 cells (approximately 20--90% increase; also see Fig. [7](#F7){ref-type="fig"}), compared to the GFP control cells. However, there were no significant differences among these cells in the extent of death caused by staurosporine, actinomycin D, daunorubicin and camptothecin (data not shown). In similar experiments, the effect of transient expression of Hyal-1 and Hyal-2 on cellular sensitivity to anticancer drugs was examined. L929 cells were electroporated with the GFP-Hyal-1 or GFP-Hyal-2 construct, cultured for 48 hr and then exposed to actinomycin D, daunorubicin or camptothecin for 24 hr. There were no differences in cell death in these Hyal-1 and Hyal-2-expressing cells, compared to control cells (data not shown). These findings are consistent with the results for stable transfectants presented above (Fig. [2](#F2){ref-type="fig"}). Bovine testicular PH-20 enhances TNF sensitivity in L929 cells by increasing the expression of proapoptotic p53 and WOX1 \[[@B9],[@B10]\]. The ability of Hyal-1 and Hyal-2 to induce the expression of p53 and WOX1 in L929 cells was examined. L929 cells were electroporated with nothing (null), GFP, GFP-Hyal-1 or GFP-Hyal-2 constructs and cultured for 48 hr. Transient expression of GFP-Hyal-1 or GFP-Hyal-2 in L929 cells induced WOX1 expression, whereas no WOX1 expression was observed in non-transfected cells or control GFP-expressing cells (Fig. [3A](#F3){ref-type="fig"}). The 46-kDa band is full-length WOX1, while the 31-kDa band is a degraded WOX1 \[[@B10]\]. No detectable p53 expression was observed in these cells (Fig. [3A](#F3){ref-type="fig"}). ![Hyal-2 induces WOX1 expression in L929 cells. (**A**) L929 cells were electroporated with nothing (null), GFP, GFP-Hyal-1 or GFP-Hyal-2 construct (30 g DNA / 3 × 10^6^ cells) and cultured for 48 hr. Transient expression of GFP-Hyal-1 and GFP-Hyal-2 in L929 cells induced WOX1 expression (46 and 31 kDa), whereas no WOX1 expression was observed in non-transfected cells or control GFP-expressing cells. The 46-kDa band is full-length WOX1, and the 31-kDa band is a degraded WOX1 \[[@B10]\]. No detectable p53 expression was observed in these cells. α-Tubulin was examined to confirm equal protein loading onto each lane of the gel. (**B**) In comparison, the GFP-Hyal-2 stable transfectant constitutively expressed cytosolic WOX1 (46 kDa) and TNF (50 ng/ml) reduced the WOX1 expression. The reduction is probably due to TNF-mediated WOX1 nuclear translocation from the mitochondria \[[@B10]\]. In contrast, WOX1 was not expressed in the stable GFP-Hyal-1 cells, and TNF induced the expression of WOX1 in a time-related manner. TNF could not induce WOX1 expression in the control GFP-expressing cells. (**C**) L929 cells were transfected with GFP-Hyal-2 and/or WOX1 constructs by electroporation. GFP-Hyal-2 enhanced WOX1-mediated L929 cell death, as determined 48 hr post-transfection using crystal violet staining (two representative rows of cell stains shown at left and the analyzed data shown at right; mean ± standard deviation, n = 4). cell death. In controls, GFP-Hyal-1 or GFP-Hyal-2 alone failed to induce cell death.](1471-2121-3-8-3){#F3} In comparison, the cytosolic WOX1 (46 kDa) was constitutively expressed in the L929 cells stably transfected with the GFP-Hyal-2 construct, while TNF reduced WOX1 expression in a time-dependent manner (Fig. [3B](#F3){ref-type="fig"}). This reduction is probably due to TNF-mediated WOX1 nuclear translocation from the mitochondria \[[@B10]\]. In contrast, WOX1 was not expressed in L929 cells stably transfected with the GFP-Hyal-1 construct, and TNF induced WOX1 expression in 40 min (Fig. [3B](#F3){ref-type="fig"}). TNF did not induce WOX1 expression in the control GFP-expressing cells (Fig. [3B](#F3){ref-type="fig"}). Whether Hyal-2 enhances the apoptosis-inducing function of WOX1 was determined. In transient transfection experiments, L929 cells were electroporated with GFP-Hyal-2 and/or WOX1 constructs. GFP-Hyal-2 enhanced WOX1-mediated L929 cell death, as determined 48 hr post-transfection using crystal violet staining (Fig. [3C](#F3){ref-type="fig"}). In contrast, GFP-Hyal-1 did not enhance WOX1-mediated cell death (Fig. [3C](#F3){ref-type="fig"}). GFP-Hyal-1 and GFP-Hyal-2 alone did not mediate cell death (Fig. [3C](#F3){ref-type="fig"}). Next, the effect of TNF-mediated IκBα degradation and NF-κB activation was examined in both Hyal-1 and Hyal-2 cells. The GFP-Hyal-1 and GFP-Hyal-2 stable transfectants and control cells were exposed to TNF for various times. IκBα degradation and NF-κB activation (or NF-κB nuclear translocation) were observed at 20 to 60 min in Hyal-2 cells post TNF stimulation, but at the 20 min time point in both control and Hyal-1 cells (Fig. [4](#F4){ref-type="fig"}). {#F4} To further understand how Hyal-2 participates in the apoptosis pathway, the translocation of Hyal-2 to the mitochondria was examined during cell death. Transient expression of GFP-Hyal-2 in COS7 fibroblasts showed the localization of the expressed protein in the lysosome but not in the mitochondria, as determined by staining the cells using LysoTracker and Mitotracker Red (data not shown). These observations are consistent with other reports \[[@B5],[@B20]\]. COS7 cells are TNF-resistant. The reasons for using this cell type are its large size (which provides better resolution under fluorescence microscopy) and the ease of transfection and gene expression using liposome-based reagents such as FuGene 6 and GeneFector. Exposure of the GFP-Hyal-2-expressing COS7 cells to staurosporine (for inducing apoptosis) resulted in a time-related increase in co-localization of GFP-Hyal-2 with the mitochondria (Fig. [5](#F5){ref-type="fig"}), suggesting that a portion of GFP-Hyal-2 migrates to the mitochondria during apoptosis. {#F5} Next, whether TGF-β1 antagonizes the function of Hyal-1 and Hyal-2 in growth regulation was examined. TGF-β1 inhibits epithelial cell growth but promotes fibroblast growth \[[@B15]\]. TGF-β1 inhibition of epithelial cell growth is blocked by bovine testicular PH-20 \[[@B15]\]. In agreement with our previous observations \[[@B11]\], TGF-β1 promoted L929 cell growth by approximately 20% in 24--48 hr (Fig. [6](#F6){ref-type="fig"}). However, TGF-β1 did not promote the growth of L929 cells stably expressing Hyal-1 or Hyal-2 (Fig. [6](#F6){ref-type="fig"}). {#F6} Finally, the functional antagonism between TGF-β1 and Hyal-1 or Hyal-2 in regulating TNF susceptibility in L929 cells was examined. The Hyal-1 and Hyal-2 stable transfectants were pretreated with TGF-β1 for 1 hr and then treated with TNF overnight. Both Hyal-1- and Hyal-2-increased TNF cytotoxicity were suppressed by TGF-β1 (30--50%) (Fig. [7](#F7){ref-type="fig"}). However, TGF-β1 blocked TNF cytotoxicity of the GFP-expressing control cells to a greater extent (50--70%) (Fig. [7](#F7){ref-type="fig"}). {#F7} Discussion ========== In this study, the novel function of the lysosomal hyaluronidases Hyal-1 and Hyal-2 in increasing cellular sensitivity to TNF cytotoxicity is reported. TGF-β1 counteracted Hyal-1- and Hyal-2-increased TNF killing of L929 fibroblasts. TGF-β1 increased L929 growth, while Hyal-1 and Hyal-2 blocked this enhancement of cell growth. These observations are consistent with our previous finding that exogenous bovine testicular hyaluronidase PH-20 enhances TNF cytotoxicity in various cancer cells, and that TGF-β1 blocks the increased TNF cytotoxicity \[[@B8],[@B15]\]. As summarized in the Figure [8](#F8){ref-type="fig"}, the underlying mechanisms whereby Hyal-2 induces TNF cytotoxicity appear to be associated with 1) Hyal-2 induction of the proapoptotic WOX1, 2) Hyal-2 migration to the mitochondria during apoptosis, and 3) the prolonged period of IκBα degradation and NF-κB activation in TNF-treated Hyal-2-expressing cells. {#F8} Upregulation of p53 and WOX1 contributes, in part, to the PH-20-enhanced cytotoxic function of TNF \[[@B9],[@B10],[@B23]\]. WOX1 is a p53-binding protein. Ectopic expression of both p53 and WOX1 cDNA constructs results in colocalization of these proteins in the mitochondria \[[@B23]\]. Increased synthesis and complex formation between p53 and WOX1 in the cytoplasm are observed during apoptosis \[[@B23]\]. Both p53 and WOX1 mediate apoptosis synergistically. Overexpressed WOX1 alone mediates apoptosis independently of p53 \[[@B10]\]. However, p53 apoptosis requires the participation of WOX1. Blocking WOX1 expression with antisense mRNA abolishes p53-mediated cell death, suggesting that WOX1 is a potential partner of p53 in apoptosis \[[@B10],[@B23]\]. Both Hyal-1 and Hyal-2 induced WOX1 expression when transiently expressed in L929 cells. However, only Hyal-2 induced WOX1 expression following stable transfection of L929 cells. This correlates with the greater extent of Hyal-2 enhancement of TNF killing compared to Hyal-1. WOX1 expression was induced by TNF in the Hyal-1 stable transfectant. In transient transfection experiments, Hyal-2 enhanced WOX1-mediated L929 cell death, while Hyal-1 had no effect. Both Hyal-1 and Hyal-2 alone failed to induce cell death. Thus, induction of WOX1 expression contributes, in part, to Hyal-1- and Hyal-2-increased TNF sensitivity in L929 cells. TNF rapidly induces phosphorylation of IκBα, which normally peaks at 5--10 min, followed by proteasome-dependent degradation (10--20 min) \[[@B10],[@B24]\]. IκBα degradation and NF-κB activation occurred, as determined 20 min after exposure of control and Hyal-1 cells to TNF. However, there was an extended period of TNF-mediated IκBα degradation and NF-κB activation in Hyal-2 cells, ranging from 20--60 min. It is unclear whether this extend ed period of IκBα degradation and NF-κB activation directly contributes to the ability of Hyal-2 to enhance TNF-mediated apoptosis. IκBα is a physiological inhibitor of NF-κB activation; IκBα complexes with NF-κB and prevents its translocation to the nucleus \[[@B10],[@B24]\]. NF-κB activation is believed to induce antiapoptotic proteins to block cell death \[[@B8]\]. However, in several studies activated NF-κB cannot protect cells from apoptosis \[[@B8]\]. Also, NF-κB appears to be essential in the p53-mediated apoptosis in tested cell lines \[[@B25]\]. The mitochondrion is a reservoir of proapoptotic proteins and plays a key role in apoptosis \[\[[@B26],[@B27]\]; reviews\]. One of the pathways that leads to the release of proapoptotic proteins in the mitochondria has been determined. For example, the TNF/TNF receptor-signaling complex activates caspase 8, which leads to the activation and cleavage of Bid, a proapoptotic protein of the Bcl-2 family. Activated or truncated Bid (tBid) in turn activates Bak and Bax. Bak migrates to the mitochondrial outer membrane where it generates transmembrane pores; Bax induces the opening of the mitochondrial permeability transition pores. Cytochrome c is then released from the mitochondria into the cytosol, where it interacts with downstream proapoptotic proteins such as Apaf-1 and caspase 9. Activation of these proteins further activates nucleases, such as caspase-activated DNase (CAD), which cause DNA fragmentation. During staurosporine-mediated apoptosis, a portion of the lysosomal Hyal-2 migrates to the mitochondria. Whether Hyal-2 causes mitochondrial membrane damage is unknown. Presumably, Hyal-2 interacts with a specific protein target on the mitochondrial surface that changes the membrane permeability. This notion is supported by our recent screening of Hyal-1 and Hyal-2 interacting proteins in yeast two-hybrid interactions (Chang *et al*, unpublished) The functional property of Hyal-2 in degrading hyaluronic acid is controversial. Hyal-2 is found in the lysosome and is active under conditions of low pH \[[@B20]\]. Hyal-2 is also found on the cell surface as a glycosylphosphatidylinositol (GPI)-anchored receptor for jaagsiekte sheep retrovirus \[[@B21]\]. Nonetheless, the enzymatic activity of Hyal-2 has not been detected \[[@B21]\]. Whether the increased TNF cytotoxicity in L929 cells caused by Hyal-1 and Hyal-2 is associated with their enzymatic activities is being studied in this laboratory. The TGF-β1 inhibition of Hyal-1 and Hyal-2 effects could be abolished by cycloheximide or actinomycin D (data not shown). That is, inhibition of protein synthesis or gene transcription reduced the inhibitory effect of TGF-β1 in the prevention of Hyal-1- or Hyal-2-increased TNF cytotoxicity. Whether TGF-β1-induced antiapoptotic proteins such as TIAF1 and TIF2 inhibit the function of Hyal-1 and Hyal-2 remains to be determined. Conclusions =========== Like testicular PH-20, lysosomal Hyal-1 and Hyal-2 enhanced TNF susceptibility in L929 cells. TGF-β1 suppressed the induced TNF sensitivity. Furthermore, TGF-β1-mediated L929 cell proliferation was blocked by Hyal-1 and Hyal-2. These data indicate that there is a functional antagonism between TGF-β1 and Hyal-1 or Hyal-2. Hyal-2-increased TNF susceptibility in L929 cells appears to be associated with 1) Hyal-2 upregulation of the proapoptotic WOX1, 2) Hyal-2 migration to the mitochondria during apoptotic cell death, and 3) the prolonged IκBα degradation and NF-κB activation in Hyal-2-expressing cells stimulated with TNF. Materials and Methods ===================== cDNA constructs, transient expression and stable transfectants -------------------------------------------------------------- The full-length murine hyaluronidase *Hyal1* and *Hyal2* cDNAs were found in the EST database (Genbank accession AA688635 and BF139787, respectively) and purchased from Incyte Genomics (St. Louis, MO). The determined DNA sequences have been deposited in the GenBank (*Hyal1*, AF422176; *Hyal2*, AF422177). The coding regions were tagged with a *C*-terminal green fluorescent protein (GFP) sequence by subcloning into the *EcoR*1 site of the pEGFP-N1 vector (Clontech, Palo Alto, CA). Murine L929 cells were transfected with the constructed Hyal-1-pEGFP-N1 or Hyal-2-pEGFP-N1 vector, or a control pEGFP-N1 vector by electroporation (40 μg DNA/ 5 × 10^6^ cells; 250 volt and 960 Fd), as previously described \[[@B10]\]. The cells were cultured in the continuous presence of G418 (200 g/ml) for approximately 2 weeks, and the resulting total cell colonies were harvested. The presence of GFP-Hyal-1 and GFP-Hyal-2 proteins in these stable transfectants was examined by Western blotting using anti-GFP antibodies (Clontech). Where indicated, COS7 fibroblasts were transfected with the above constructs using a liposome-based reagent FuGene 6 (Roche, Indianapolis, IN) or GeneFector (Venn Nova, Pompano Beach, FL). Protein expression was visualized by fluorescence microscopy. Also, the cells were counterstained with LysoTracker or Mitotracker Red (Molecular Probes, Eugene, OR) to determine whether the expressed protein was located in the lysosome or mitochondria. Cell lines and chemicals ------------------------ Murine L929 fibroblasts and monkey kidney COS7 fibroblasts were cultured as previously described \[[@B10]\]. Staurosporine, daunorubicin, actinomycin D, camptothecin and DAPI were from Calbiochem (San Diego, CA), Sigma (St. Louis, MO) and Biomol (Plymouth Meeting, PA). Purified human platelet TGF-β1 was from R&D (Minneapolis, MN). Antisera against WOX1 were generated in rabbits against a synthetic peptide \[[@B10]\]. Antibodies against p53, IκBα, NF-κB p65 and α-tubulin were from Santa Cruz Biotechnology (Santa Cruz, CA) and Accurate Chemicals (Accurate Chemical, Westbury, NY). TNF cytotoxicity assay and Western blotting ------------------------------------------- The above established cell lines were treated with TNF-α (2.5--20 ng/ml; Pepro-Tech, Rocky Hill, NJ) for 16--24 hr. The extent of cell death was determined by crystal violet staining \[[@B10]\]. Where indicated, the cells were also exposed to staurosporine, actinomycin D, daunorubicin, and camptothecin for 16--24 hr, before determining the extent of cell death. To determine whether TGF-β1 blocked Hyal-1 and Hyal-2-increased TNF killing, the cells were pretreated with TGF-β1 for 2 hr and then exposed to TNF-α in the continued presence of TGF-β1 for 16--24 hr. The extent of cell death was then determined. Where indicated, GFP-Hyal-1, GFP-Hyal-2 and GFP stable transfectants were treated with TNF (50 ng/ml) for 0, 20, 40, 60 and 120 minutes. The cells were extracted using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Pierce, Rockford, IL) to separate the cytoplasmic and nuclear fractions. These preparations were subjected to SDS-PAGE and Western blotting using specific antibodies against IκBα, WOX1 and NF-κB. Gene transfection and fluorescent microscopy -------------------------------------------- COS7 cells were transfected with the GFP-Hyal-2 vector using the liposome-based FuGene 6 or GeneFector, cultured for 24--48 hr, and treated with staurosporine (1 M) for 0, 1, 2 and 4 hr. Mitochondria were stained with the membrane potential sensitive Mitotracker Red. Nuclei were stained by DAPI. The cells were then examined by fluorescence microscopy. Acknowledgment ============== Research was supported by the American Heart Association and the Guthrie Foundation for Education and Research. I thank Ms. Lori Schultz and Mr. John Heath for technical assistance, and Drs. Robert S. Aronstam, Margaret P. Quinlan and Victor Ruiz-Velasco for critically reviewing the manuscript.

Background ========== In *S. cerevisiae,* a key event accompanying exit from mitosis is the inactivation of Clb/Cdc28 protein kinase achieved through degradation of Clb and accumulation of the Clb/Cdc28 inhibitor Sic1 (reviewed in \[[@B1]\]). Anaphase-Promoting Complex/Cyclosome (APC/C) ubiquitin ligase and its substrate specific activator Hct1/Cdh1 are required for degradation of Clb2, the major mitotic cyclin in the budding yeast (reviewed in \[[@B2]\]). Whereas *hct1Δ* and *sic1Δ* cells appear relatively normal, *hct1Δ sic1Δ* cells are inviable, presumably due to their inability to extinguish Cdc28 activity in telophase \[[@B3]\]. The complexity in the regulation of mitotic exit is underscored by the existence of a set of genes essential for this process, referred to as the \"mitotic exit network (MEN)\". They encode the GTP-binding protein Tem1 and its putative guanine nucleotide releasing factor Lte1, the dual specificity protein phosphatase Cdc14, protein kinases Cdc5, Cdc15, and Dbf2/Dbf20, the Dbf2-binding protein Mob1 (reviewed in \[[@B4]\]) and the spindle-pole-body component Nud1 \[[@B5]\]. Conditional-lethal temperature sensitive (ts) mutations in any of the MEN genes cause cells to arrest in late mitosis with elevated mitotic Cdk activity when shifted to the restrictive temperature. Inactivation of Cdc28 by overexpression of Sic1 suppresses multiple MEN mutants and hyperactivation of Cdc28 by overexpression of Clb2 exacerbates their phenotypes, consistent with the idea that a key function of the MEN proteins is to eradicate Cdc28 activity during the transition from anaphase/telophase to G1 \[[@B6]-[@B9]\]. Recently, important advances have been made in our understanding of how the MEN is switched on during anaphase. Tem1 is thought to be kept inactive throughout early mitosis by the two-component GTPase-activating protein complex Bub2/Bfa1 which co-localizes with Tem1 at the spindle pole body \[[@B10],[@B11]\]. When the mitotic spindle penetrates the mother-daughter neck, the Tem1-bound spindle pole body is translocated into the daughter cell where a cortical pool of Lte1 is thought to activate Tem1 \[[@B12],[@B13]\]. In addition to this spatial regulation, the GTPase activity of Bfa1 is thought to be inactivated through its phosphorylation by Cdc5 \[[@B14]\], but it remains unknown how this process is coordinated with the presumed juxtaposition of Lte1 and Tem1 in the daughter cell. Despite this rapid progress in understanding the signals that regulate the MEN, the direct downstream targets of the MEN have yet to be apprehended. To address how the MEN components downstream of Cdc15 are organized and what their targets are, we sought *tab (telophase arrest bypassed)* mutants that alleviate the essential requirement for *CDC15.* In theory, *tab* mutants could act downstream of or parallel to *CDC15,* and could be either loss-of-function mutations in genes that inhibit exit from mitosis or gain-of-function mutations in genes that promote exit from mitosis. To our surprise, mutations in six linkage groups were uncovered, with three groups being recessive (*tab1-tab3*) and three groups being dominant (*TAB5-7*). We have shown in earlier studies that *tab2-1* is a reduction-of-function mutation in *NET1* (a.k.a. *CFI1*) \[[@B15]-[@B17]\] and that *TAB6* is a gain-of-function mutation in *CDC14*\[[@B18]\]. Net1 sequesters Cdc14 in the nucleolus and inactivates it in preanaphase cells. During exit from mitosis, a *TEM1/CDC15*-dependent signal frees Cdc14 from the nucleolus and allows it to promote inactivation of Clb/Cdc28 \[[@B15],[@B17],[@B19]\]. The fact that both Net1 and Cdc14 are directly involved in exit from mitosis validates the effectiveness of the tab screen and heightens the potential value of the uncharacterized *tab* mutants. Here, we present a detailed description of the *tab* mutant screen, and describe a third *TAB* gene, *TAB1/MTR10.* Results ======= Cdc15 promotes activation of Clb proteolysis by a post-translational mechanism ------------------------------------------------------------------------------ Transcription has been postulated to play an important role in triggering exit from mitosis \[[@B20]\]. For example, Dbf2, a protein kinase that promotes exit from mitosis, is associated with the CCR4 transcription regulatory complex \[[@B21]\]. In addition, the Swi5 transcription factor activates *SIC1* expression during late anaphase, and both *sic1Δ* and *swi5Δ* are synthetically lethal with *dbf2Δ*\[[@B22],[@B7]\]. To address if production of new proteins is essential for exit from mitosis, we examined the effects of the protein synthesis inhibitor cycloheximide on Clb2 degradation, a key aspect of this process. Mutant *cdc15-2* cells were uniformly arrested in late anaphase/telophase by incubation at the restrictive temperature 37°C. The cultures were then supplemented with 0, 10, or 100 μg/ml cycloheximide before being returned to 25°C to reverse the *cdc15-2* arrest. In the absence of cycloheximide, Sic1 accumulated and Clb2 disappeared as cells exited mitosis (Figure [1](#F1){ref-type="fig"}, lanes 1--4). In the presence of cycloheximide, Sic1 accumulation was abolished, but Clb2 degradation still occurred with apparently normal kinetics (Figure [1](#F1){ref-type="fig"}, lanes 5--8 and 9--12). Although this result does not exclude the likely possibility that synthesis of new proteins (e.g. Sic1) normally facilitates the exit from mitosis, it reveals that a key aspect of mitotic exit -- inactivation of Cdc28 protein kinase via degradation of the mitotic cyclin Clb2 -- can proceed beyond the Cdc15-dependent step without synthesis of new proteins. {#F1} A screen for *telophase arrest bypassed (tab)* mutants ------------------------------------------------------ To understand how the MEN mobilizes post-translational mechanisms to trigger mitotic exit, we conducted a genetic screen for \'telophase arrest bypassed\' (*tab*) mutants that bypassed the essential requirement for *CDC15* (Figure [2](#F2){ref-type="fig"}). Because the essential requirement for *CLNs 1, 2,* and *3* in entry into S phase can be bypassed by deletion of a gene (*SIC1*) that inhibits entry into S phase \[[@B23],[@B24]\], we reasoned by analogy that our screen might reveal key negative regulators of the anaphase/telophase -\> G1 transition. {#F2} Ninety independent cultures of *cdc15Δ \[pMET3-cdc15-2, URA3\]* were first grown in inducing medium (- methionine) at permissive temperature (25°C) to allow occurrence of spontaneous mutations. The cultures were subsequently plated on repressing medium (+ methionine) at a temperature semi-restrictive for the *cdc15-2* allele (30°C) to enrich for mutations that supported the growth of *CDC15-* deficient cells. To identify mutations that allowed complete bypass of *cdc15Δ,* representative colonies from each plate were tested for their ability to grow on 5-FOA-containing medium (5-FOA selectively blocks the growth of plasmid-bearing *URA3+* cells). The parental strain failed to grow due to its inability to proliferate without the \[*MET3-cdc15-2, URA3*\] plasmid. In contrast, twenty five mutants survived on 5-FOA. Furthermore, all of them had lost the plasmid as determined by PCR analysis (data not shown). Fifteen of these mutants were amenable to genetic manipulation and harbored single mutations. Fourteen mutants were assigned to three recessive and three semi-dominant/dominant linkage groups (Table [2](#T2){ref-type="table"}). We referred to these mutants as *tab* mutants for their telophase-arrest-bypass (Tab^+^) phenotype. In addition, we serendipitously found that *srp1-31* but not *srp1-49* could bypass *cdc15Δ* (Table [2](#T2){ref-type="table"}; data not shown). ::: {#T2 .table-wrap} ::: {.caption} ###### Characterization of *tab* mutants ::: Name \# alleles recessive/dominant^b^ identity growth at 37°C^d^ Nop1 pattern^e^ -------------- ------------ ---------------------------------- ------------ ------------------- ----------------- *tab1* 5 recessive *MTR10*^c^ inviable \- *tab2* 2 recessive *NET1* slow \- *tab3* 1 recessive slow \+ *TAB5* 2 semi-dominant normal \+ *TAB6* 1 dominant *CDC14* slow \+ *TAB7* 3 semi-dominant (-1) dominant (-2) normal \+ *srpl-31*^a^ recessive \- ^a^We have not determined if *srp1-31* is allelic to *tab3, TAB5,* or *TAB7.*^b^*All* alleles were tested for recessiveness/dominance of Tab^+^ except for *TAB7-3.*^c^See Methods. ^d^*All* alleles of a single linkage group displayed similar ts growth phenotype. ^e^All strains were grown to exponential phase at 25°C, and subjected to indirect immunofluorescence using anti-Nop1 antibodies. The wild type staining pattern was scored as \"+\", and partial delocalization or fragmented staining patterns were scored as \"-\". ::: Earlier studies identified *tab2* and *TAB6* as mutations in *NET1* and *CDC14,* respectively \[[@B15],[@B18]\]. Net1 detains Cdc14 in the nucleolus and inactivates it until a Tem1-dependent signal sets Cdc14 free to promote exit from mitosis \[[@B15]-[@B17]\]. The fact that both Net1 and Cdc14 are directly involved in exit from mitosis prompted us to further analyze *tab* mutants. Does bypass of *cdc15*Δ require Sic1 accumulation and Clb2 degradation? ----------------------------------------------------------------------- Either Sic1 accumulation or Clb2 degradation is proposed to be sufficient to sustain mitotic exit \[[@B3]\]. Although *cdc15-2* mutants are defective in both processes, ectopic activation of either one should restore exit from mitosis and growth. Thus, to gain insight into their molecular targets, we asked if bypass of *cdc15Δ* by *tab* mutations required *SIC1* or *HCT1.* Whereas all *tab* mutants tested were able to bypass *cdc15Δ* in the absence of *HCT1* and two (*net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^) were able to bypass *cdc15Δ* in the absence of either *SIC1* or *HCT1,* four *tab* mutants (*tab1-1, tab3-1, TAB5-1,* and *TAB7-1*) failed to bypass in the absence of *SIC1* (Table [3](#T3){ref-type="table"}). ::: {#T3 .table-wrap} ::: {.caption} ###### Does bypass require *SIC1* or *HCT1?* ::: Strain *sic1Δ*^a^ *hct1Δ*^b^ \+^c^ ---------- ------------ ------------ ------- *tab1-1* \- \+ \+ *tab2-1* +/- \+ \+ *tab3-1* \- \+ \+ *TAB5-1* \- \+ \+ *TAB6-1* +/- \+ \+ *TAB7-1* \- \+ \+ ^a^To assay if bypass required *SIC1, cdc15Δ::TRP1 tab \[pMET3-CDC15/URA3*\] cells were transformed with an *SIC1*-deletion cassette (*S. pombe his5*^+^ DNA fragment whose termini were engineered to be homologous to *SIC1* 5\' and 3\' untranslated regions). All transformants were tested for their ability to bypass *cdc15Δ* using the 5-FOA growth assay (sample size n = 4--17). They were also screened for the absence of *SIC1* as the result of homologous integration. In the case of *tab2* and *TAB6,* about half of *sic1Δ* transformants could bypass. This incomplete penetrance is not well understood. There are at least two sources of variability in this experiment. First, *sic1Δ* transformants of *tab cdc15Δ \[CDC15, URA3\]* were pregrown in rich media to accumulate cells that had lost the \[*CDC15, URA3*\] plasmid. Because plasmid-loss is a low-probability event and because cells without the plasmid are at a growth-disadvantage, the number of plasmid-free cells spotted on a 5-FOA plate probably was small and varied among different transformants. Second, mutant individuals can exhibit more variability than wild-type individuals, a phenomenon frequently observed in worms and flies. Therefore, it is possible that in only a fraction of plasmid-free cells, the amount of Cdc14 activity reached a level high enough to tolerate *sic1Δ*. Thus, one can imagine that for some transformants, the number of plasmid-free cells spotted on 5-FOA was small and that by chance they or their limited numbers of progeny could not exit mitosis, resulting in the absence of bypass colonies. In other transformants, the opposite could occur, giving rise to mixed results in the plate assay. ^b^Similar assay was conducted for *HCT1.* Sample size n = 1--7. ^c^Transformants from a and b that resulted from non-homologous integration (and hence no deletion of *SIC1* or *HCT1*) were used as a control. Sample size n = 12--86. ::: Efficient bypass of mitotic arrest by both *SIC1*-dependent and *SIC1*-independent *tab* mutants is accompanied by ectopic Sic1 accumulation and Clb2 degradation ----------------------------------------------------------------------------------------------------------------------------------------------------------------- We have previously shown that both *net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^ efficiently bypassed the mitotic arrest caused by Tem1-depletion \[[@B15],[@B18]\]. Furthermore, bypass was accompanied by ectopic Sic1 accumulation and Clb2 degradation. Because both *net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^ could bypass *cdc15Δ* in the absence of *SIC1,* we tested if the same bypass efficiency could be achieved by *tab3-1,* a *SIC1*-dependent bypass mutant. We used a *tem1Δ::GAL1-UPL-TEM1* strain that allowed for conditional depletion of Tem1 \[[@B15]\] to impose a conditional late mitotic block, and introduced the *tab3-1* mutation into this background. Wild type or *tab3-1* derivatives of *tem1Δ::GAL1-UPL-TEM1* were grown in galactose medium (*TEM1* induced), arrested in G1 with α factor, and released into glucose-containing medium to repress *TEM1* expression (time 0). As expected, the majority (\~80%) of *TAB*^+^ cells arrested with large buds. In the *tab3-1* strain, however, a significant fraction (\~36%) of cells had ≥ 3 cell bodies at 12 hours after release, indicating that they had undergone at least one extra round of budding (Figure [3A](#F3){ref-type="fig"}). Thirty-six percent represents a minimum estimate of *TEM1* bypass efficiency in *tab3-1* cultures, since cells with 1--2 cell bodies could have arisen from bypass events followed by successful cell separation. {#F3} Having established that *tab3-1* indeed bypassed Tem1 deficiency at a cellular level, we tested whether Sic1 accumulation and Clb2 degradation, two hallmark events that accompany exit from mitosis, also occurred. *tem1Δ::GAL1-UPL-TEM1* cells carrying *TAB3* or *tab3-1* alleles were arrested in G1 with α factor, and released into glucose medium (*TEM1* repressed, time 0). After cells had exited G1, α factor was added back to stop cells from cycling beyond the next G1 phase. As Tem1-depleted *TAB3* cells exited G1, Clb2 accumulated and remained at high levels, whereas Sic1 was degraded and remained at low levels (Figure [3B](#F3){ref-type="fig"}, left panels). In contrast, as Tem1-depleted *tab3-1* cells progressed through mitosis, Clb2 was almost completely degraded and Sic1 accumulated to high levels (Figure [3B](#F3){ref-type="fig"}, right panels). These data suggest that like *tab2-1* and *TAB6-1, tab3-1* efficiently bypassed *tem1Δ* by enabling Tem1-independent degradation of Clb2 and accumulation of Sic1. We conclude that the apparent *SIC1*-dependence of bypass by *tab3-1* and other *tab* mutants relates more to the strength of the particular mutation, rather than the mechanism of bypass. In support of this hypothesis, we recently observed that the *SIC1*-independent bypass mutant 15D2 is allelic to the *SIC1*-dependent *tab1-1* mutant (R. Azzam, unpublished data). *tab2-1* bypasses multiple deletion mutants of MEN -------------------------------------------------- All MEN genes are required for exit from mitosis, but it has not yet been tested whether this is the sole essential function of all MEN genes, or if some MEN genes play critical roles in other processes. This question is amplified by the observations that Cdc5 participates in both chromosome segregation and cytokinesis (\[[@B25],[@B26]\]; Park et al., submitted), Mob1 binds an essential protein kinase involved in spindle pole body duplication \[[@B27]\], and Tem1, Cdc15, Dbf2, and Mob1 have all been implicated in cytokinesis (\[[@B28]-[@B31]\]; reviewed in \[[@B1]\]. The *tab* mutants provide a unique opportunity to address this issue. Accordingly, we tested *if net1*^*tab*2-1^ could bypass deletion mutants of other MEN genes. *net1*^*tab*2-1^ bypassed *cdc15Δ, tem1Δ, dbf2Δ dbf20Δ, mob1Δ,* but not *cdc14Δ* (Figure [4A](#F4){ref-type="fig"}; \[[@B15]\]). In addition, it bypassed *cdc5Δ* with negligible (1000--10,000 fold lower) efficiency (Fig. [4A](#F4){ref-type="fig"}), suggesting that although the sole essential function of *CDC15, TEM1, DBF2,* and *MOB1* is to drive cells out of mitosis, *CDC5* has other critical cellular functions. {#F4} Some but not all *tab* mutants exhibit additional phenotypes including growth defects, perturbations in the nucleolus, and elevated rates of chromosome loss ------------------------------------------------------------------------------------------------------------------------------------------------------------ To further characterize the *tab* mutants, we tested whether they display other phenotypes in an otherwise wild type background. *tab* mutants grown to exponential phase at 25°C and shifted to 37°C for 3 or 6 hours did not arrest at a particular stage of the cell cycle (data not shown). When *tab* mutants were plated on YPD plates and grown at 37°C, *TAB5* and *TAB7* grew normally (Table [2](#T2){ref-type="table"}), suggesting that bypass of *CDC15* does not necessarily cause defective growth. However, *tab1* failed to form colonies, and *net1*^*tab*2-1^*, tab3,* and *CDC14*^*TAB*6-1^ grew slowly at 37°C, suggesting that at least some *tab* mutations reside in genes that have multiple functions. Indeed, *NET1* additionally modulates nucleolar structure and transcription \[[@B16],[@B18]\]. The release of Cdc14 from nucleolar Net1 promotes mitotic exit \[[@B15],[@B17],[@B18]\]. *net1*^*tab*2-1^ cells display a panel of nucleolar defects including reduction of nucleolar silencing, mislocalization of nucleolar antigens, alteration of rDNA morphology, and reduction of rRNA synthesis \[[@B16],[@B18]\]. To test whether perturbation of nucleolar structure is consistently associated with bypass of *cdc15Δ,* we immunolocalized the nucleolar antigen Nop1 in all *tab* mutants. Nop1 localization was also perturbed in *tab1-1* and *srp1-31,* but was normal in *tab3, TAB5, TAB6,* and *TAB7* cells (Table [2](#T2){ref-type="table"}; \[[@B18]\]), suggesting that a gross change in nucleolar structure is not required for bypass. Because the putative Tem1 antagonists and mitotic checkpoint proteins Bub2 and Bfa1 negatively regulate exit from mitosis and cytokinesis in response to both DNA damage and spindle misorientation (\[[@B32]\]; reviewed in \[[@B33]\]), we asked whether bypass of *TEM1*-dependent regulation can cause chromosome instability. All *tab* mutants are in the W303 background in which the *ade2-1* mutation makes colonies red after adenine in the growth medium has been depleted. A *URA3-marked* tester mini-chromosome carrying the *SUP11* suppressor transfer RNA gene \[[@B34]\]) was introduced into *net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^, the two *tab* mutants with well-characterized bypass mechanisms. Cells that retain the mini-chromosome should form white colonies due to suppression of *ade2-1* by *SUP11,* whereas loss of this chromosome should result in a red sector or sectors. The extent of chromosome loss (expressed as % colonies that were red or had red sectors) was low in wild-type and *CDC14*^*TAB*6-1^ cells (\< 5%), and high (94%) in *net1*^*tab*2-1^ cells (Figure [5](#F5){ref-type="fig"}). In comparison, in a strain lacking the anaphase-inhibitor Pds1, \~18% of the colonies suffered chromosome loss events (Figure [5](#F5){ref-type="fig"}). Since *net1*^*tab*2-1^but not *CDC14*^*TAB*6-1^ also perturbs nucleolar functions \[[@B18]\], *CDC14*^*TAB*6-1^ is a more specific and representative bypass mutant. Thus, *TEM1* regulation can be short-circuited without a major effect on chromosome loss. Consistent with this notion, cells lacking Bub2 (an inhibitor of Tem1) showed a similarly moderate extent of chromosome loss (5%) (Figure [5](#F5){ref-type="fig"}). We propose that additional functions of Net1 ensure faithful transmission of chromosomes. ![*net1*^*tab*2-1^ cells display strong chromosomal segregation defect. A *URA3*-marked tester mini-chromosome carrying the *SUP11* suppressor transfer RNA gene \[[@B34]\] was introduced into the indicated genetic background also harboring an *ade2* mutation. Cells that retained the mini-chromosome would form white colonies due to suppression of *ade2* by *SUP11;* the loss of this chromosome resulted in a red sector or sectors. When the loss event occurred at a very high frequency, a red colony could form. Cells were pre-grown in the uracil-lacking SD agar medium, and plated on YPD plates at room temperature. (A) After the initial signs of colony coloration appeared, the plates were kept at 4°C to enhance color development. White arrowheads indicate three sectoring events in a *net1*^*tab*2-1^(*net1-1*) colony. (B) Percentages of colonies that were red or had red sectors were calculated.](1471-2156-3-4-5){#F5} Molecular cloning of *TAB* genes and implication of nuclear transport in the regulation of mitotic exit ------------------------------------------------------------------------------------------------------- We attempted to clone by complementation genes corresponding to the recessive *tab* mutations by exploiting the fact that all of the recessive *tab* mutants are ts for growth (see below). By this approach, *TAB1* was revealed to be *MTR10,* a gene previously implicated in nuclear transport \[[@B35]-[@B37]\]. We were unable to isolate plasmids that complemented *tab3* in all libraries tested. Both *MTR10 (TAB1*) and *SRP1* belong to the nuclear transport receptor/adaptor family (reviewed by \[[@B38]\]). Mutant *srp1-31* and *mtr10*^*tab*1-1^ cells accumulate the ribosomal protein Rpb11b in the nucleus \[[@B39]\]. Thus, to evaluate a potential link between nuclear transport and the *tab* phenotype, we evaluated the localization of Rpl11b-GFP \[[@B39]\] in all *tab* mutants. Mutants expressing Rpl11b-GFP were grown to early-log phase at 25°C, shifted to 37°C to induce the ts mutant phenotype (which also reduces synthesis of ribosomal proteins including Rpl11b), and shifted back to 25°C to induce new ribosomal protein synthesis. If the rate of mutant phenotype reversal is slower than restoration of ribosome synthesis, then Rpl11b should accumulate in the nucleus. Whereas Rpl11b-GFP was distributed relatively evenly in wild-type cells, it accumulated in the nucleus in a significant fraction of *srp1-31* and *net1*^*tab*2-1^ cells at 25°C and after 25°C -\>37°C -\> 25°C shift (Figure [6](#F6){ref-type="fig"}; \[[@B39]\]). In addition, although *mtr10*^*tab*1-1^ cells were relatively normal at 25°C, 30% of them accumulated nuclear Rpl11b after 25°C -\>37°C -\> 25°C shift (Figure [6](#F6){ref-type="fig"}). Thus, multiple *tab* mutants accumulate Rpl11b in the nucleus, implying that nucleocytoplasmic trafficking is involved in the regulation of mitotic exit. Other *tab* mutants (such as *tab3, TAB5, CDC14*^*TAB*6^ and *TAB7*) showed no defect in this assay, suggesting that bypass *of cdc15Δ* can be achieved without perturbing nucleocytoplasmic transport. ![Some of the *tab* mutants show nuclear transport defects. Strains of the indicated genetic background were transformed with a CEN/ARS plasmid expressing Rpl11b-GFP, and grown at 25°C. A portion of the cultures were then shifted to 37°C for 1 hr and shifted back to 25°C for 35 min before the 37°C -\> 25°C samples were taken. (A) Rpl11 b-GFP was visualized as described previously \[[@B39]\]. (B) The percentage of cells (+/- 1%) with preferential nuclear Rpl11b-GFP staining in each sample was counted (n ≥ 100).](1471-2156-3-4-6){#F6} Discussion ========== MEN: A signalling network that can post-translationally trigger exit from mitosis --------------------------------------------------------------------------------- Both transcriptional and post-translational controls have been implicated in the operation of the Mitotic Exit Network and the regulation of the anaphase/telophase -\> G1 transition \[[@B40],[@B7],[@B19],[@B41]\]. To address whether the portion of the MEN that acts downstream *of CDC15* can trigger a key aspect of the anaphase/telophase -\> G1 transition (i.e. inactivation of Clb/Cdc28 via proteolysis of Clb2) in the absence of new gene expression, we evaluated the turnover of Clb2 in cells released from a *cdc15-2* arrest in the presence of cycloheximide. Interestingly, Clb2 was rapidly and efficiently degraded upon reversal of the *cdc15-2* block regardless of whether cycloheximide was present (Figure [1](#F1){ref-type="fig"}) although both cytokinesis and cell separation require protein synthesis \[[@B42]\]. This observation suggests that the exit from mitosis in budding yeast may be controlled by mechanisms fundamentally similar to those employed by animal cells, which proceed through mitosis with highly condensed, transcriptionally silent chromosomes. Interestingly, all of the MEN genes have putative homologues in the *C. elegans* genome <>. Isolation of *tab* mutants -------------------------- A deliberate search for bypass mutants was a key aspect of the screen reported here, since numerous genetic interactions have been observed among components of the MEN (consult \[[@B9]\] for a summary). Formally speaking, it is not possible to order gene functions based on suppressive interactions involving reduction-of-function (e.g. ts) alleles. For example, the *net1-1* allele suppressed ts mutations in both its downstream target *CDC14* and its upstream regulators *CDC5, CDC15,* and *DBF2* (Figure [4B](#F4){ref-type="fig"}). Since we sought mutants that bypassed *cdc15Δ,* the *tab* genes are predicted to function either downstream of or parallel to *CDC15,* but not upstream of *CDC15.* The tab screen yielded both dominant and recessive mutants which mapped to six linkage groups. The pattern of alleles (Table [2](#T2){ref-type="table"}) does not satisfy a Poisson distribution, suggesting that the screen has not reached saturation. The recessive *tab* mutants are likely to represent reduction-of-function mutations in genes that inhibit exit from mitosis, and the dominant *TAB* mutants are likely to result from gain-of-function mutations in genes that promote exit from mitosis. This prediction is supported by our prior characterization of the *tab2-1* and *TAB6-1* alleles of *NET1* and *CDC14,* respectively (see next section). The fact that we recovered recessive, semi-dominant, and dominant mutants in additional *tab* genes suggests that there are likely to be multiple dosage-sensitive positive and negative regulatory controls that act downstream of or parallel to *CDC15.* *tab* mutants can be classified into two groups: those that bypass *cdc15Δ* only in the presence of *SIC1* (e.g. *tab3-1*), and those that can bypass in the absence of *SIC1* (e.g. *net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^). It remains unclear whether mutants distribute into these classes based on bypass mechanism or allelic potency, but we favor the latter possibility, because some *tab1* alleles required *SIC1* to bypass *cdc15Δ,* whereas others did not (R. Azzam, unpublished data). Importantly, *tab* mutants from both classes appeared to bypass mitotic arrest in Tem1-depleted cells by enabling degradation of Clb2 and accumulation of Sic1 -- two hall-marks of mitotic exit (Figure [3](#F3){ref-type="fig"}; \[[@B15],[@B18]\]). A genetic pathway for anaphase -\> G1 transition ------------------------------------------------ Biochemical and cell biological experiments indicate that Net1 acts as part of a complex named RENT that tethers Cdc14 to the nucleolus and inhibits Cdc14 phosphatase activity, and that *CDC15* and *TEM1* are required for the release of Cdc14 from Net1 at the end of mitosis \[[@B15]-[@B17],[@B43]\]. Consistent with this notion, both *net1*^*tab*2-1^ (which encodes a mutant version of Net1 with presumably reduced affinity for Cdc14) and *CDC14*^*TAB*6-1^ (which encodes a mutant version of Cdc14 with reduced affinity for Tab2/Net1) bypass *cdc15Δ*\[[@B15],[@B18]\]. Both *net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^ impinge directly on mitotic exit, suggesting that other *TAB* genes may encode physiological regulators and effectors of the Mitotic Exit Network. The *net1*^*tab*2-1^ mutation efficiently bypassed *tem1Δ, cdc15Δ, dbf2Δdbf20Δ,* and *mob1Δ,* consistent with the notion that the sole essential function of all these genes is to inhibit Net1. *net1*^*tab*2-1^ failed to bypass *cdc14Δ,* because Cdc14 is the downstream target of Net1 \[[@B15],[@B17]\]. Cdc5 also regulates Net1 (Shou et al., manuscript in preparation), but *net1*^*tab*2-1^ bypassed *cdc5Δ* with very low efficiency, presumably because *CDC5* has other essential functions \[[@B44],[@B25],[@B26],[@B45]\]. The most parsimonious model that unifies all of these observations is as follows: Tem1 functions upstream of Cdc15, which in turn activates the Dbf2-Mob1 protein kinase complex \[[@B46]-[@B49]\]. Together with Cdc5, these proteins dislodge Cdc14 from nucleolar Net1. Cdc14 subsequently dephosphorylates (and thereby activates) proteins involved in Sic1 expression (Swi5) and C1b degradation (Hct1/Cdh1) \[[@B19],[@B41],[@B50]\], thereby eliminating mitotic Cdk activity and promoting exit from mitosis. The mechanisms by which Cdc5 and Dbf2/Mob1 promote the disassembly of RENT remain unknown, raising the possibility that other Tab proteins serve as intermediaries in this process. One attractive hypothesis based on our identification of *MTR10* as a *tab* gene is that MEN proteins regulate the nucleocytoplasmic distribution of a key regulator of the release process. Chromosome instability in *tab* mutants --------------------------------------- The Bub2/Bfa1 GTPase-activating protein (GAP) complex inhibits the MEN and cytokinesis in response to DNA damage and spindle misorientation (\[[@B32]\]; reviewed in \[[@B33]\]). *bub2Δ* and *CDC14*^*TAB*6-1^ cells show similar chromosome instability as wild-type cells. In contrast, *net1*^*tab*2-1^ cells have a dramatic chromosome instability phenotype, with some colonies showing multiple red sectors indicative of multiple chromosome-loss events, and many colonies completely red on the surface (Figure [5A](#F5){ref-type="fig"}). Net1 is unlikely to directly regulate transmission of the tester chromosome, because it is only localized to rDNA in the nucleolus \[[@B15]-[@B17]\]. Thus, the difference between *net1*^*tab*2-1^ and *CDC14*^*TAB*6-1^ presumably results from molecular defects unique to the former mutant, including defective nucleolar structure and function, or perhaps defective nuclear transport (\[[@B18]\]; Figure [6](#F6){ref-type="fig"}). Further study is required to reveal how the multifunctional Net1 protein helps to ensure chromosome transmission fidelity. Multiple *tab* mutations influence nucleocytoplasmic transport -------------------------------------------------------------- Nuclear transport-defective mutations in *MTR10* and *SRP1 (mtr10*^*tab*1-1^ and *srp1-31,* respectively) bypassed *cdc15Δ.* Mtr10 assembles with and functions as a nuclear import receptor for the mRNA-binding protein Npl3 \[[@B36],[@B37]\]. Interestingly, *srp1-31* also inhibits nuclear import of multiple proteins including Npl3 \[[@B51],[@B52]\]. Remarkably, reduction-of-function alleles of the karyopherin *KAP104* were also recovered in a screen for mutations that suppressed *cdc15-2*^*ts*^ (A. Toh-e, personal communication). *KAP104* is required for proper localization of mRNA binding proteins Nab2 and Hrp1 to the nucleus \[[@B53]\]. The only phenotype known to be shared by *srp1, mtr10,* and *kap104* mutants is accumulation of the 60S ribosomal protein Rpl11b in the nucleus \[[@B39]\]. *SRP1, MTR10,* and *KAP104* may also collaborate to maintain the proper nucleocytoplasmic partitioning of α factor (or factors) that regulates mitotic exit. An alternative possibility is that they sequester an inactive stockpile of α factor that, upon mutation or signalling by the MEN, is released from these transportins and enables exit from mitosis \[[@B54],[@B55]\]. Finally, it is possible that the transportin mutants enable bypass through an indirect perturbation of nucleolar architecture (Table [2](#T2){ref-type="table"}). Given that some of the MEN proteins reside in the cytoplasm whereas their target -- the RENT complex -- resides in the nucleolus, it is tempting to speculate that nuclear transport is a target of MEN action. Our failure to observe nuclear accumulation of Rpl11b in *tab3* and *TAB5-TAB7* mutants suggests that these *tab* genes bypassed *cdc15Δ* by a distinct mechanism. Regardless of the exact mechanisms of bypass, it is clear from these analyses that the release of Cdc14 from Net1 is more intricately connected to cell regulation than was previously imagined. Conclusions =========== Three main conclusions on how exit from mitosis is regulated in the budding yeast can be derived. First, the sole essential function of the Tem1, Cdc15, Dbf2/Dbf20, and Mob1 (but not Cdc5) proteins are to promote the release of Cdc14 from the nucleolus, as cells lacking these proteins can grow upon simultaneous reduction *of NET1* activity. Second, high chromosome loss rate was observed in *net1*^*tab*2-1^ but not *CDC14*^*TAB*6-1^, suggesting that although *CDC15*-independent mitotic exit does not dramatically affect chromosome transmission, additional functions of Net1 do. Third, both *mtr10*^*tab*1-1^ and *srp1-31* affect nuclear transport, suggesting that nuclear transport regulates mitotic exit. Materials and Methods ===================== Strains and plasmids -------------------- All yeast strains used in this study (listed in Table [1](#T1){ref-type="table"}) are isogenic to W303. Standard methods were employed for the culturing and manipulation of yeast \[[@B56]\]. All plasmid constructions were based on the pRS vector series \[[@B57]\]. To replace *CDC15* with *TRP1,* polymerase chain reaction (PCR) products containing *TRP1* flanked by 200-base pair (bp) homology to the 5\' and 3\' untranslated regions of the targeted gene were used to transform diploid strain RJD381. Correct integrants were verified by PCR using primers that amplified a DNA fragment that spanned the recombination junction. Deletions of *CDC5* and *TEM1* were carried out similarly. To construct \[*pMET3-CDC15, URA3*\] (pWS100) and \[*pMET3-cdc15-2, URA3*\] (pWS109), the corresponding genes (from start codon to 300 bps downstream of the stop codon) were amplified by PCR from genomic DNA of wild type (RJD381) or *cdc15-2* (RJD619) cells, and cloned into the SpeI and SacII sites of RDB620 (a derivative of pRS316 containing the *MET3* promoter inserted between the HindIII and EcoRI sites). ::: {#T1 .table-wrap} ::: {.caption} ###### *S. cerevisiae* strains ::: Strain Genotype^a^ -------- ----------------------------------------------------------------------- RJD381 *MAT**a**/MATα* RJD619 *cdc15-2 pep4Δ::TRP1 MAT**a*** WY4 *TAB7-2 MAT**a*** WY9 *cdc15Δ::TRP1 TAB5-1 \[pMET3-CDC15, URA3\] \[pRS315, LEU2\] MATα* WY10 *TAB5-1 \[pRS315, LEU2\] MATα* WY11 *cdc15Δ::TRP1 tab2-1(net1-1) \[pMET3-CDC15, URA3\] MATα* WY14 *tab3-1 MAT**a*** WY17 *cdc15Δ::TRP1 TAB6-1 \[pMET3-CDC15, URA3J \[pRS315, LEU2\] MAT**a*** WY18 *TAB6-1MATα* WY21 *cdc15Δ::TRP1 tab1-1 \[pMET3-CDC15, URA3\] \[pRS315, LEU2 \] MATα* WY34 *cdc15Δ::TRP1 TAB7-1 \[pMET3-CDC15, URA3\] \[pRS315, LEU2\] MAT**a*** WY38 *tab1-1 MAT**a*** WY39 *tab2-1(net1-1) MAT**a*** WY41 *cdc15Δ::TRP1 tab3-1 \[pMET3-CDC15, URA3\] MAT**a*** WY46 *tem1::GAL1-UPL-TEM1/TRP1 bar1::hisG MAT**a*** WY97 *tab3-1 tem1::GAL1-UPL-TEM1/TRP1 bar1:: LEU2 MAT**a*** WY214 *cdc5Δ::TRP1 \[pMET3-CDC5, URA3\] MATα* WY217 *cdc15Δ::TRP1 \[pMET3-CDC15, URA3\] MATα* WY218 *tem1Δ::TRP1 \[pGAL1-TEM1, URA3\] MATα* WY221 *cdc15Δ::TRP1 \[pMET3-cdc15-2, URA3\] MATα* WY243 *cdc14Δ::his5^+^ \[HA3-CDC14, URA3\] MATα* WY280 *dbf2Δ::LEU2 dbf20Δ::TRP1 \[DBF2, URA3\] MATα* WY288 *mob1Δ::HIS3 \[MOB1, URA3\] MATα* \"All strains are in the W303 background (*ade2-1 can1-100 his3-11-15 leu2-3-112 trp1-1 ura3-1*), which was provided by B. Fuller. *\[ \]* indicates *CEN/ARS* plasmid. WY280 and WY288 were derived after backcrossing the original strains into W303 three times. ::: Isolation of *tab* mutants -------------------------- The scheme is outlined in Figure [2](#F2){ref-type="fig"}. Ninety independent cultures of *cdc15Δ \[pMET3-cdc15-2, URA3*\] (WY221) were grown at 25°C in synthetic minimal medium + 2% glucose (SD) in the absence of methionine. For each culture, \~8 × 10^6^ cells were plated on SD+ methionine (2 mM) at 30°C (to simultaneously repress expression and partially inactivate the *cdc15-2* allele). One to two colonies were picked from each plate and tested for viability on SD+ 5-fluoroorotic acid (5-FOA) plates \[[@B58]\]. When more than one colony was picked from a single plate, care was taken to pick colonies of different size or morphology. Out of \~90 colonies tested, 25 survived on 5-FOA medium (Tab^+^). PCR analysis confirmed that all twenty five colonies had lost the \[*pMET3-cdc15-2, URA3*\] plasmid, indicating that they were true bypassers of *cdc15Δ.* Mutant characterization ----------------------- To determine if Tab^+^ phenotypes were due to single mutations, *cdc15Δ tab* strains were crossed to *cdc15Δ \[pMET3-cdc15-2, URA3\]* and the resulting diploids were sporulated and dissected. Approximately 50% of the viable spores should bypass *cdc15Δ* (as indicated by 5-FOA resistance) if the Tab^+^ phenotype was due to a single mutation. Out of twenty five mutants, fifteen satisfied this criterion, and the other ten were discarded because they either failed to sporulate, or yielded few viable spores after dissection, or did not segregate as single mutations. The same crosses were also used to determine if Tab^+^ was recessive or dominant: for every *tab* mutant, haploid *cdc15Δ tab \[pMET3-cdcl5-2, URA3\]* and diploid *cdc15Δ/cdc15Δ tab/+ \[pMET3-cdc15-2, URA3*\] cells were pre-grown on YPD (1% yeast extract/2% peptone + 2% glucose) at 25°C for 2--3 days, spotted in 5-fold serial dilutions onto SD+5-FOA plates, and the number of 5-FOA resistant colonies from the diploid was divided by that from the haploid to obtain ratio P. The mutant was defined to be dominant if P= 0.1-1, semi-dominant if P= 0.01-0.1, and recessive if P \< 0.01. To introduce *tab* mutations into the wild type background, *cdc15Δ tab* strains were crossed to *CDC15 TAB,* and *tab* spores were verified by crossing to *cdc15Δ \[pMET3-CDC15, URA3\],* and demonstrating that half of the resulting *cdc15Δ* segregants were 5-FOA resistant. All *cdc15Δ tab* mutants were also crossed to strains harboring either *\[pMET3-CDC15, URA3\] \[pRS313, HIS3\]* or *\[pMET3-CDC15, URA3\] \[pRS315, LEU2\]* to introduce these plasmids. The resulting strains grew better than the original *cdc15Δ tab* strains, and harbored selectable markers that facilitated subsequent crosses. To test whether *net1*^*tab*2-1^ can bypass *tem1Δ, dbf2Δ dbf20Δ, mob1Δ, cdc14Δ,* and *cdc5Δ,* we crossed *net1*^*tab*2-1^ (WY39) to WY218, WY280, WY288, WY243, and WY214, respectively. The diploids were sporulated, and if \~50% of the MEN-deleted spores were resistant to 5-FOA, then *net1-1* could bypass the MEN null mutation. To test temperature sensitivity (ts) of *tab* mutants, cells were spotted or streaked onto YPD plates, and incubated at 37°C. Their growth was scored after one to two days. Nop1 immunolocalization was performed as described earlier \[[@B15]\]. The chromosome loss assay was carried out using a strain \[[@B34]\] backcrossed into the W303 background by F. Uhlmann. When cells were pre-grown in YPD liquid (instead of SD-URA solid) media before being tested on YPD plates, the percentage of colonies that were red or had red sectors remained similar for wild-type, *CDC14*^*TAB*6-1^*,* and *bub2Δ* strains, but was \~25% for *net1*^*tab*2-1^. This number was lower than the 94% value observed in Figure [5](#F5){ref-type="fig"}, but still significantly higher than the wild-type value (\< 5%). Linkage analysis ---------------- To assign *tab* mutants to linkage groups, *cdc15Δ tab \[pMET3-CDC15, URA3*\] strains were crossed against each other, and the diploids were sporulated and dissected. If two *tab* mutations belonged to the same linkage group, then all spores were expected to survive on 5-FOA. Otherwise, approximately one quarter of the spores would fail to bypass *cdc15Δ.* The alleles of *TAB2, TAB3, TAB5, TAB6,* and *TAB7* (see Table [2](#T2){ref-type="table"}) were assigned this way. The recessive *tab* mutants exhibited a ts growth phenotype that co-segregated with the Tab^+^ phenotype. Complementation tests were used to assign two of the ts mutations as alleles of *tab 1.* After *TAB1* was shown to be *MTR10,* other *tab1* alleles were ascertained by their linkage to the *MTR10* locus. Molecular cloning of *TAB1* and *TAB2* -------------------------------------- To clone *TAB1* and *TAB2, tab1-1* and *tab2-1* mutant strains were transformed with plasmid libraries harboring yeast genomic DNA fragments, and incubated at 25°C for one day before being shifted to 37°C \[[@B59]\]. Libraries constructed by the laboratories of P. Heiter (ATCC\#77164) and R. Young \[[@B60]\] rescued *tab1-1,* and a 2μ library (ATCC\#37323, \[[@B61]\]) rescued *tab2-1.* Transformants (enough to cover multiple genome equivalents) were screened, and plasmids retrieved from colonies that grew at 37°C were re-transformed into the original mutant strain to verify their complementation activity. Candidate plasmids were sequenced, and the genomic regions containing the complementing activity were identified. *TAB1* resided on Chromosome XV, 631,500--637,950, and *TAB2,* on Chromosome X, 291,900--299,466. To confirm that these genomic fragments indeed carried the *TAB* genes, complementing fragments were cloned into the integrating vector \[pRS305, LEU2\], and linearized to target integration into the *tab* locus in a *cdc15Δ::TRP1 tab \[pMET3-CDC15, URA3*\] strain. The transformants lost their Tab^+^ phenotype, and when crossed to *cdc15Δ::TRP1 \[pMET3-CDC15, URA3\],* the diploid yielded no spores that could bypass *cdc15Δ* (n = 40--50). Since both genomic fragments contained multiple open reading frames (ORFs), each ORF was amplified by PCR from yeast genomic DNA and cloned into the integrating plasmid pRS305. The resulting plasmids were transformed into *cdc15Δ::TRP1 tab \[pMET3-CDC15, URA3*\] to identify the ORF that reversed the Tab^+^ phenotype. This analysis revealed *TAB1* to be *MTR10,* and *TAB2* to be *YJL076W (NET1).* Gene replacement ---------------- For gene replacement, the *Schizosaccharomyces. pombe his5*^+^ PCR amplification/ transformation method was used \[[@B62]\]. Correct integrants were verified by PCR using primers that amplified a DNA fragment that spanned the recombination junction. To assay if *SIC*1 or *HCT1* was required for the Tab^+^ phenotype of *tab* mutants, *SIC1* or *HCT1* was replaced by *his5*^+^ in *cdc15Δ::TRP1 tab \[pMET3-CDC15, URA3*\] strains. Each transformant was assayed for the occurrence of correct integration and its ability to survive on 5-FOA. Cell cycle synchronization -------------------------- Cells were grown in 1% yeast extract/ 2% peptone (YP) with 2% glucose (YPD), 2% galactose (YPG), or 2% raffinose (YPR). For the experiment described in Figure [1](#F1){ref-type="fig"}, *cdc15-2* (RJD 619) cells were grown to exponential phase in YPD at 25°C, and shifted to 37°C for three hours so that \>95% cells were arrested as large-budded cells. Cycloheximide (CHX) was supplemented to a final concentration of 0, 10, or 100 μg/ml to inhibit translation, and α factor (10 μg/ml) was added to trap cells in the subsequent Gl phase. After five minutes, the cultures were released to 25°C at time 0. Samples were taken at various time points after release. The experiment in Figure [3](#F3){ref-type="fig"} was carried out as previously described \[[@B15]\]. Cell extract preparation and protein detection ---------------------------------------------- Detection of proteins from crude yeast extracts was as described \[[@B15]\]. The following primary antibodies were used in immunoblotting: anti-Sic1 (1:8000), anti-Clb2 (1:2500), anti-Cdc28 (1:3000), and 12CA5 (directed against the haemagglutinin (HA) epitope; 1:1000, with 0.5 M NaCl). The first three antibodies were raised in rabbits and affinity purified, and the last one was from mouse. Acknowledgements ================ Special thanks go to Ramzi Azzam for mapping the putative *\"tab4\"* mutation to the *tab1* linkage group (Table [2](#T2){ref-type="table"}) and Kathleen Sakamoto for performing the Nop1 immunofluorescence experiment (Table [2](#T2){ref-type="table"}). We thank R. Azzam, K. Sakamoto, M. Budd, R. Feldman, B. Hay, P. Sternberg, and R. Verma for discussions; H. Mountain, P. Philippsen, A. Varshavsky, B. Futcher, and P. Silver for *pMET3, S. pombe his5*^+^, *UPL, HA3,* and *RPL11B-GFP* plasmids, respectively; F. Uhlmann and H. Rao for the *SUP11/URA3* sectoring strain; M. Nomura, L. Johnston, and F. Luca for *srp1, dbf2Δdbf20Δ \[pRS316-DBF2\], and mob1Δ \[pRS316-MOB1\] strains,* respectively; C. Thompson for yeast genomic library, and D. Kellogg, R. Feldman, and J. Aris for anti-Clb2, anti-Sic1, and anti-Nop1, respectively. W.S. was a Howard Hughes Medical Institute predoctoral fellow.

Current government policy in Canada is pressing for more and better care of the terminally ill in the community \[[@B1]-[@B3]\]. Several factors contribute to the re-emergence of home death, including: preference of the terminally ill and their families to remain in the home, recognition of the limited benefits found in medical care, improvements in home based medical technology and pharmacology, and hospital restructuring resulting in the closure of hospital beds \[[@B4],[@B5]\]. Related to this shift is an increasing elderly population which provides an impetus for the development of home based palliative care programs \[[@B6],[@B7]\]. Despite the above, fewer people are able to die at home than would wish to do so. While death at home is preferred by one half or more of terminally ill patients \[[@B8],[@B9]\], approximately 25% of people in Britain currently die at home \[[@B10]\]. Lower percentages have been reported in the United States and Australia \[[@B11]\]. The literature illustrates a number of factors needed to support a death at home. These include: the patients\' and caregivers\' desire for a home death, availability of skilled support 24 hours a day, health of the participating caregiver, and sufficient financial resources to allow caregiving to occur at home \[[@B12],[@B13]\] . The majority of published studies on the determinants of place of death focus on the relationships between sociodemographic and support network characteristics of the patient and place of death \[[@B14],[@B15]\]. Studies investigating place of death for patients who received palliative home care reveal that patients with a primary caregiver \[[@B16],[@B17]\] and those receiving additional home support \[[@B13]\] are more likely to die at home. Despite the volume of research in this area definite conclusions on the determinants of place of death remain elusive \[[@B15]\]. Identifying these determinants could improve supportive community care for the terminally ill. The purpose of this retrospective cohort study was to determine what patient and informal caregiver characteristics were associated with a home death among older adults receiving palliative care nursing services in the home. The results provide information which can assist in future planning and implementation of sustainable community-based programs tailored specifically to the needs of the terminally ill and their family caregivers. Method ====== Family caregivers were recruited into the study over a two year period \[1997--99\] through one of two community nursing agencies providing palliative care nursing services in homes located in an urban region in the province of Ontario, Canada. Patients were placed on the palliative case load of the participating agencies if they met one of the several criteria: identified end stage disease, oncology diagnosis indicated metastases, oncology patient was expected to undergo chemo or radiation therapy and would require symptom management, or presented complex psychosocial needs. The focus of services provided by the nursing agencies was physical symptom management and psychosocial and spiritual support. Palliative care nurses could also assess and facilitate referrals to other appropriate services. Inclusion criteria for study participants were that: the deceased was 50 years or older and the family caregiver could not only communicate well in english, but would provide informed consent. Caregivers were excluded from the study if the patient had dementia not due to malignancy and/or medication. Patients with dementia due to these two causes were excluded from the study as investigators felt that they represented a unique subpopulation presenting special needs which require separate considerations. Family caregivers were defined as relatives of the deceased who provided care to that individual on a regular basis, and did not receive remuneration for such services. Procedure --------- Family caregivers identified according to the above eligibility criteria were contacted by the nursing agency who explained the study to them. Once permission was obtained to release the names of interested caregivers to the study team a research assistant contacted participants to obtain consent and complete a telephone interview. Interviewers had undergraduate degrees in gerontology or the health sciences. Interviewer training included mock interviews and quality assurance assessments to ensure interview protocols were followed. Ethics approval was obtained from the local ethics board. Interview schedule ------------------ The interview procedure built upon that used by Addington-Hall and McCarthy \[[@B18]\] and covered the last year of life. Data collected covered two main areas: patient characteristics, and informal caregiver characteristics. Patient characteristics included: sex, age, marital status, diagnosis, dependence in activities of daily living \[ADL\] \[bathing, mobility, dressing, toileting and assistance at night\], length of time assistance needed in ADL, living arrangements, preferred place of death, and place of death. Informal caregiver characteristics included: sex, gender, education, employment status, self report of health status \[excellent, very good, good, fair, poor\], caregiving experience, and formal services used. To catalogue \"preferred place of death\" respondents were asked whether the care recipient expressed a preferred place to die and if so, where such was located. The caregiving experience was assessed by the Caregiver Reaction Assessment \[CRA\]. A 24-item scale, the CRA measures the reaction of family caregivers caring for older adults with physical and mental impairments \[[@B19]\]. Respondents used a five point likert scale ranging from strongly disagree to strongly agree as indicators of their agreement with 24 sentences. In completing the CRA, respondents were instructed to, \"Think back to your caregiver experience with \_\_\_\_\_. Try to remember how you felt most of the time when you were caring for him/her\". The scoring of positively worded sentences was reversed such that higher scores indicated more negative impacts of caregiving. The total possible score ranged from 24 \[no negative impact\] to 120 \[immense negative impact\]. Furthermore, the instrument included five subscales which measured: the extent to which caregiving impacts self-esteem \[range of 7--35\]; lack of family support \[[@B5]-[@B25]\]; impact on finances \[[@B3]-[@B15]\]; impact on schedule \[[@B5]-[@B25]\]; and, impact on health, which assessed the caregiver\'s physical capability and energy to provide care \[[@B4]-[@B20]\]. The instrument was multidimensional, factorially invariant across different patient groups, and factorially invariant across repeated measurement occasions \[[@B20]\] . In a multicenter study of partners of patients with cancer, reliability analysis of the CRA showed Cronbach\'s alphas between .62 and .83 for the separate subscales, indicating sufficient internal consistencies \[[@B21]\] . The study also found that the subscales represented conceptually different dimensions within the caregiving situation and supported the construct validity of the subscales. The CRA was preferred above other caregiver burden instruments for this study because other instruments were developed for caregivers of persons with mental impairment only or addressed only the negative consequences of caregiving \[[@B22]-[@B24]\]. Statistical analysis -------------------- Means, standard deviations and frequencies were used to describe patient and informal caregiver characteristics. Univariate odds ratios \[OR\] along with their 95% confidence intervals \[95% CI\] were used to determine which of the patient and informal caregiver variables were associated with a home death. The independent contribution of predictor variables was then assessed by a multivariate logistic regression. For the multivariate logistic regression, we used statistically significant variables according to univariate analyses and those considered by the investigators to be theoretically significant. Theoretical significant variables included the age and gender of the patient and caregiver, caregiver employment status and whether the patient received home making services. These variables were selected as previous research has revealed their significance in understanding patterns of caregiving \[[@B25]\]. For the purpose of logistic regression, a composite score of caregiver assistance provided to the patient on ADL was obtained by considering how long the assistance had been provided. The score could be one if the assistance was less than one month in duration, two for one to three months, three for three to six months, four for six to 12 months, and five for greater than 12 months. For example, if the patient had been receiving bathing assistance for more than six months but less than 12, a score of four would have been assigned for this type of need. The scores for all five activities were then added to form the composite score. The minimum possible score was zero, the maximum possible was 30. SPSS \[version 9.0 for Windows, SPSS Inc., Chicago, 1998\] was used for statistical computations and a probability level of 0.05 selected to determine statistical significance \[i.e., the 95% CI does not include 1.0\]. Results ======= Participants included 151 family caregivers \[consent rate of 51%\] of patients who had died approximately 9 months prior to the study interview. The average age of patients was 71 years \[SD = 9.34\] and 61 years \[SD = 13.56\] for the caregivers. Other characteristics of the study population are shown in Table [1](#T1){ref-type="table"}. ::: {#T1 .table-wrap} ::: {.caption} ###### Characteristics of study population (n = 151) ::: N (%) ------------------------------- ------------ **Patient Characteristics** Sex • women 63 (41.7) • men 88 (58.3) Diagnosis • cancer 124(82.1) • other 26(17.2) Setting of death • home 95 (62.9) • hospital 54 (35.7) • nursing home 1(.7) • other 1(.7) **Caregiver Characteristics** Sex • women 119(78.8) • men 32(21.2) Relationship to patient • spouse 99 (65.6) • child 46 (30.4) • extended family 6 (3.9) Lived with patient 130 (86) Not employed 102 (67.5) Education completed • elementary 45 (29.8) • high school 57 (37.7) • college 30(19.8) • university 19 (12.5) ::: Eighty-nine percent of caregivers reported the patient experienced some degree of functional restriction. The number of caregivers providing assistance in specific functional activities was: bathing, 133 \[88%\]; mobility, 123 \[81%\]; dressing and undressing, 114 \[76%\]; toileting, 101 \[67%\] and, assistance at night 97 \[64%\]. Since study participants were recruited through visiting nursing agencies, all were receiving home nursing visits. The three most frequently used services reported by caregivers included house calls by a family physician, 89 \[59%\]; home making services, 84 \[56%\]; and hospital palliative care services 37 \[25%\]. Unadjusted univariate OR and corresponding 95% confidence intervals are presented in Tables [2](#T2){ref-type="table"} and [3](#T3){ref-type="table"}. The reference category is indicated by an OR of 1.00 and a 95% CI of 1.00,1.00. The multivariate model revealed that the odds of dying at home were greater when the patient lived with the caregiver \[OR = 7.85; 95% CI = 2.35, 26.27\], the patient stated a preference to die at home \[OR = 6.51; 95% CI = 2.66,15.95\], and the family physician made home visits \[OR = 4.79; 95% CI = 1.97, 11.64\]. The odds were lower for patients who used hospital palliative care beds \[OR= 0.31; 95% CI = 0.12, 0.80\] or had caregivers with fair to poor health status \[OR = .22; 95% CI = 0 .07, 0.64\]. ::: {#T2 .table-wrap} ::: {.caption} ###### The association of patient variables and home death ::: Variable Home death Univariate Multivariate ---------------------- ------------ ------------ -------------- ------------- ------ ------------- Yes No OR 95% CI OR 95% CI Gender • female 40 (62.5) 23 (37.5) 1.04 0.53,2.04 NS • male 55 (63.5) 33 (36.5) 1.00 1.00, 1.00 Age • 50--64 25 (69.4) 11(30.6) 1.00 1.00, 1.00 NS • 65--74 35(61.4) 22 (38.6) 0.70 0.29, 1.70 • 75+ 35 (60.3) 23 (39.7) 0.67 0.28, 1.62 Marital status • spouse 64 (64.6) 35 (35.4) 1.24 0.62, 2.47 • other 31(59.6) 21 (40.4) 1.00 1.00, 1.00 Diagnosis • cancer 78 (62.9) 46(37.1) 0.90 0.37,2.18 • other 17 (65.4) 9 (34.6) 1.00 1.00, 1.00 Dependence in ADL • high 43(61.4) 27 (38.6) 0.87 0.44, 1.74 • low 44 (64.7) 24 (35.3) 1.00 1.00, 1.00 Lived with caregiver • yes 88 (67.7) 42 (32.3) 4.19 1.57, 11.15 7.85 2.35,26.27 • no 7 (33.3) 14 (66.7) 1.00 1.00, 1.00 Stated preference • yes 74 (79.6) 19 (20.4) 6.86 3.29, 14.32 6.51 2.66, 15.95 • no 21 (36.2) 37 (63.8) 1.00 1.00 ::: ::: {#T3 .table-wrap} ::: {.caption} ###### The association of caregiver variables and home death ::: Variable Home death Univariate Multivariate ----------------------- ------------ ------------ -------------- ------------ ------ ------------- Yes No OR 95% CI OR 95% CI Gender • female 75 (63.0) 44 (37.0) 1.02 0.46, 2.29 NS • male 20 (62.5) 12 (37.5) 1.00 1.00, 1.00 Age • \<50 20 (60.6) 13 (39.4) 1.00 1.00, 1.00 NS • 50--64 27 (57.4) 20 (42.6) 0.88 0.35,2.17 • 65--74 34 (69.4) 15 (30.6) 1.47 0.58,3.72 • 75+ 13(61.9) 8(38.1) 1.06 0.34,3.25 Education • \> high school 36 (67.9) 17(32.1) 1.40 0.69, 2.83 • other 59 (60.2) 39 (39.8) 1.00 1.00, 1.00 Employment • full-time 11 (47.8) 12 (52.2) 0.48 0.20, 1.18 NS • other 84 (65.6) 44 (34.4) 1.00 1.00, 1.00 Health status • excellent/very good 46 (74.2) 16 (25.8) 1.00 1.00, 1.00 1 1.00, 1.00 • good 33 (63.5) 19 (36.5) 0.60 0.27, 1.35 0.64 0.22, 1.86 • fair/poor 16 (43.2) 21 (56.8) 0.27 0.11,0.63 0.22 0.07, 0.65 Self-esteem • worst 21 (55.3) 17 (44.7) 0.63 0.29, 1.33 • better 71 (66.4) 36 (33.6) 1.00 1.00, 1.00 Scheduling • worst 19 (63.3) 11(36.9) 1.02 0.45,2.34 • better 76 (62.8) 45 (37.2) 1.00 1.00, 1.00 Health • worst 22 (66.7) 11 (33.3) 1.23 0.55,2.78 • better 73(61.9) 45(38.1) 1.00 1.00, 1.00 Family support • worst 19 (54.3) 16 (45.7) 0.62 0.29, 1.33 • better 75 (65.8) 39 (34.2) 1.00 1.00, 1.00 Finance • worst 14 (60.9) 9(39.1) 0.88 0.36, 2.20 • better 81 (63.8) 46 (36.2) 1.00 1.00, 1.00 Other caregiver(s) • yes 76 (65.5) 40 (34.5) 1.60 0.74, 3.45 • no 19 (54.3) 16 (45.7) 1.00 1.00, 1.00 Needed more help • yes 30 (52.6) 27 (47.4) 0.49 0.25, 0.98 NS • no 63 (69.2) 28 (30.8) 1.00 1.00 Physician home visits • yes 68 (76.4) 21 (23.6) 4.20 2.08, 8.46 4.79 1.97, 11.64 • no 27 (46.5) 35 (56.5) 1.00 1.00, 1.00 1.00 1.00, 1.00 Homemaking support • yes 54 (64.3) 30 (35.7) 1.14 0.59, 2.22 NS • no 41 (61.2) 26 (38.8) 1.00 1.00, 1.00 Palliative care bed • yes 16 (43.2) 21 (56.8) 0.34 0.16,0.72 0.31 0.12,0.80 • no 79 (69.3) 35 (30.7) 1.00 1.00, 1.00 1.00 1.00, 1.00 ::: Discussion ========== The findings of this study further our understanding of factors associated with home deaths among those who receive palliative nursing care in the home. Individuals who indicated a preference to die at home and resided with a healthy informal caregiver had better odds of dying at home. Home visits by the family physician were also associated with dying at home. The profile of caregivers and patients in this study is consistent with the research literature. Females are most likely to provide informal care \[[@B25]\]. The estimates range from 60% -- 75% of informal care in Canada being provided by women \[[@B25]\]. Caregivers are often a spouse or daughters, particularly an eldest daughter \[[@B26]\]. The main limitation of this study is the uncertain generalizability of the findings. The consent rate of 51% is consistent, if not better, than previously reported in bereavement studies \[[@B27]\]. Strobe & Strobe \[[@B27]\] conducted a review of 21 bereavement studies similar in design to the present study and reported that half of these had accrual rates of less than 50 %. However, the authors determined that depression was a factor in whether individuals participated. Specifically, men who were less depressed agreed to participate while women who were more depressed did so. The authors explained this sex difference in terms of sex roles in coping styles and norms for exhibiting emotions. Hence, non-participants may have different response patterns than participants. Further replication with prospective designs may resolve this issue and strengthen causal linkages on the present study findings. The strength of the statistical associations in this study establish clear service implications. As dying at home is dependent on caregivers\' capacity to manage domestic situations, recognition of the vital role families and other caregivers play in the community care system is essential. Family caregivers require support from the \"formal care\" system in their role of caring for a terminally ill family member. It should be stressed that caregivers in the present study represent an enhanced care group where they had the benefit of experiencing specialized palliative care nursing in the home. The availability of such service in Canada varies from community to community. For many Canadians access to such services becomes \"luck of the draw\" \[[@B1]\]. Without these supports however caregivers experience stress, burnout and ill health with the inevitable consequence of becoming unable to provide appropriate care to those in need \[[@B28]\]. It is also important to note that family physicians can enable home deaths. However, in these situations, family physicians are confronted by complex, and possibly unfamiliar, end-stage issues, therefore initiatives that enhance the palliative care skills and knowledge base could avert unnecessary institutionalizations \[[@B29],[@B30]\]. Finally, the importance of patient preference in predicting a home death heightens the belief that community care to the terminally ill needs to be client-focussed. Supporting patient and caregiver choices for service and treatment options are essential supportive measures to enable home deaths. Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== The development of practice guidelines -------------------------------------- In clinical medicine, variations exist that do not appear to be related to variations in the clinical presentation and severity of disease \[[@B1]-[@B3]\]. In response, practice guidelines have been developed in an attempt to reduce the wide practice variations and, through this process, to increase the appropriateness and quality of medical care and to reduce health care costs \[[@B4]-[@B8]\]. Despite the publication and dissemination of practice guidelines \[[@B9]\], there has been relatively little evaluation of the application and impact of clinical practice guidelines \[[@B9]-[@B14]\]. Some of the difficulty in the evaluation of these guidelines relates to the methods that were used to develop them \[[@B9]\]. Guidelines have often have been developed before adequate data have been available to assess the relationship between clinical practice patterns and desired clinical outcomes. Nevertheless, there have been some reviews of practice guideline evaluation \[[@B15],[@B16]\]. While epidemiological designs are commonly used to evaluate the effectiveness of health care interventions, never has this been discussed in the context of outcomes research. We propose the use of a methodological framework for outcomes research to evaluate practice guidelines. Methodological issues with the measurement of practice variations ----------------------------------------------------------------- In the debate about reasons to promote the development of practice guidelines, few have questioned whether the variations are real, or alternatively, whether they are simply a function of methodological flaws in the measurement of medical practices themselves, the result of variations in practice patterns across groups of patients with a similar diagnosis, or both. Furthermore, few studies have addressed whether practice variations, in fact, lead to outcome variations. Finally, little attention has been paid to the identification and measurement of initial conditions, that is, the potentially confounding factors and effect modifiers of the practice patterns outcomes relationship. Measurement of practice pattern variation ----------------------------------------- The measurement of medical practice patterns is susceptible to error. Measurement error may affect the validity of medical practice measurement in three major ways. First, it may lead to selection bias, in that subjects are selected to belong to a certain group based on an erroneous diagnosis. Secondly, it may lead to misclassification of exposure (information bias), in that patients treated with a specific practice pattern are classified in the wrong diagnostic group. Thirdly, it may lead to misclassification of outcomes, in that patients with a given outcome are classified in the wrong diagnostic group. Potential problems with the measurement of practice variations relate to the mechanisms that underlie the choice of groups that are compared in studies of practice variations. These mechanisms must be defined clearly to minimize selection bias. In many studies of practice variations, populations are arbitrarily divided according to hospitals, regions, counties, or countries. Little information is available about the factors that lead these groups to go to a particular hospital, live in a particular region, go to a particular doctor, etc. The population base from which each comparison group is derived should, in principle, be quite similar for all groups. Basically, if the groups are drawn from a similar population, unmeasurable and potentially confounding variables are more likely to be equally distributed between groups. In addition, the measurement of practice variations cannot be valid without information on relevant \"initial conditions\". Initial conditions are all confounding factors and effect modifiers, other than the treatment/practice patterns, that may cause or influence the clinical outcomes of interest. These factors may explain practice variations among groups that do not share similar initial conditions. To evaluate practice patterns-outcomes associations, potential confounders must be identified and controlled for in the analysis. Aside from clinical presentation and severity of illness; the initial conditions to be identified and characterized as completely as possible include physician, patient, and practice environment factors (Table [1](#T1){ref-type="table"}). Measurement of such factors is essential to minimize the chance of a systematic error following confounding biases and effect modification (Figure [1](#F1){ref-type="fig"}). {#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### Initial conditions to be taken into account when making inferences about practice patterns-outcomes associations ::: **Patient factors** **Physician factors** **Environmental factors** ------------------------------------- ------------------------------------------------------------- ------------------------------------------------------------- Demographics Severity of illness Physicians\' beliefs and interpretation of medical evidence Medical insurance status (fee-for-service vs. managed care) Heterogeneity of disease expression Medical uncertainty Availability of needed procedure Patient preference Specialist vs. generalist Physicians\' skills Proximity of medical centers with appropriate expertise Patient attitude towards disease Inappropriate use Nature of group ::: Identification and measurement of outcomes of interest ------------------------------------------------------ Limitations to the development and evaluation of practice guidelines also include the absence of a clear concept of the targeted outcomes and the paucity of outcomes data to support these guidelines \[[@B17]\]. There appears to be only a weak relationship between the purpose of guidelines and many of the outcomes usually measured in clinical research, that is, the source of evidence for guideline development (evidence-based). The initial goals of establishing practice guidelines -- to reduce costs and enhance the quality and appropriateness of treatment -- are, in fact, rarely the basis for guideline development, since little data is available for these outcomes. To some degree, the development of guidelines has been driven by the availability of data on clinical outcomes, such as morbidity and mortality, rather than those outcomes related to the primary goals of the guidelines. The evaluation of practice guidelines ------------------------------------- Throughout the development of practice guidelines, the major deficiency has been the lack of an evaluative method \[[@B18]-[@B27]\]. Thus, we suggest a methodological framework for outcomes research to be applied to evaluate practice guidelines. Outcomes research evaluates practice patterns as they occur in actual clinical settings. This type of research can describe practice patterns, evaluate their divergence from practice guidelines and determine the effect of practice variations on outcomes. Outcomes research is necessarily observational in nature and, although observational studies have been used to evaluate health care interventions, the proposed methodological framework has yet to be applied to outcomes research. Why should outcomes research be used to evaluate and validate practice guidelines? The primary goal of practice guidelines is the consistent adherence by physicians to practice patterns that achieve the \"best\" outcomes at the lowest cost. Outcomes research evaluates practice patterns as they occur in actual clinical settings, and is thus the logical method to evaluate practice guidelines. In fact, outcomes research and practice guidelines are connected through concepts that relate to efficacy and effectiveness research (Figure [2](#F2){ref-type="fig"}). Efficacy studies, which normally complement practice guideline development, are those performed in highly selected groups of patients to investigate if a particular intervention works under controlled conditions set by the study investigators. In contrast, outcomes research evaluates practice as it occurs in actual clinical settings \[[@B28]\]. Research in these settings is called effectiveness research because the investigators have limited control over the conditions that qualify the practice settings. The difference between efficacy and effectiveness research can be summarized as follows: does it work at all (efficacy) or does it work in the real world (effectiveness)? Thus, there exists a dynamic process in which evidence from both effectiveness and efficacy studies feeds into the development and evaluation of practice guidelines, as depicted in Figure [2](#F2){ref-type="fig"}. {#F2} Most practice guidelines are derived from efficacy studies rather than effectiveness studies. Therefore, it is not surprising that practice guidelines are not fully applicable in actual clinical practice. We suggest that effectiveness studies be used not only as a method to evaluate practice guidelines but also as a basis for their development. These could include both observational studies and effectiveness trials. Outcomes research better reflects practice in the real world and may make guidelines more likely to be applied. However, to date, little attention has been paid to the epidemiological underpinnings of the methods used to conduct outcomes research. Discussion ========== We will first propose a methodological framework for outcomes research. Then, we will show how it can be used to evaluate practice guidelines. Finally, we will address the limitations of the proposed methodological framework. Generic epidemiological issues in outcomes research --------------------------------------------------- In the proposed methodological framework, the generic issues related to outcomes research will be discussed in sequential order. In outcomes research, the first step is to identify the study population and the groups (hospitals, providers, regions, etc.) that will be compared. The next step is the measurement of practice patterns and outcomes. After groups are compared on the basis of the treatment they receive and outcomes of interest, associations are sought between practice patterns and the various measures of outcome. This step of the methodological framework raises issues of confounding bias because not all factors that can confound these associations are measured and controlled or even known. The presence or absence of confounding bias can be affected by the other sources of bias namely selection and information biases. Lastly, we discuss the issue of temporal trends. In the evaluation of practice guidelines, the measurement of practice patterns may not be contemporaneous with the publication of practice guidelines. This may explain and even lead to the frequently observed discrepancy between the actual practice and what the guidelines state that it should be. Finally, two particularities of outcomes research 1) the presence of ecological exposures in individual level studies and 2) the common use of large administrative databases are discussed. Specification of the model -------------------------- ### Definition of the elements of the proposed epidemiological model for outcomes research In the proposed model for outcomes research designed to evaluate practice guidelines, the outcome of interest can be a disease (Table [2](#T2){ref-type="table"}). For example, if the practice patterns that are being studied pertain to coronary revascularization, complications such as mortality and reinfarction after acute myocardial infarction may constitute the outcome of interest. Finally, the consequences of different practice patterns on medical resources (cost, quality and appropriateness) may be another possible outcome of interest. ::: {#T2 .table-wrap} ::: {.caption} ###### Epidemiological model for outcomes research to evaluate practice guidelines ::: Outcome = Exposure + Confounders + Error ----------------------------------------------------------------------- --------------------- ----------------------------------------------------------------------------------------------------- ------------------------------------------------------------------ • Clinical outcomes (for example, disease development, complications) • Practice patterns • Extraneous factors that can influence practice patterns and outcomes, that is, initial conditions • Misclassification of diagnosis, practice patterns and outcomes • Quality • Error in measurement of practice patterns and outcomes • Cost • Appropriateness • Unmeasured factors ::: In the studies of outcome research, practice patterns, (which constitute the exposure in the proposed model), range from the use of medication, diagnostic tests and therapeutic procedures to the length of hospital stay, transfer to other facilities and/or scheduled physicians visits. The primary goal of outcomes research is the evaluation of the effects of the selected practice patterns on the outcomes of interest. Consequently, any inference made about this association must be evaluated as a function of the potential selection, information (measurement error) and confounding biases. A limitation of outcomes research as it is most often performed is the lack of attention given to the measurement of each of the elements of the epidemiological model shown in Table 3. The basis of the proposed methodological framework will be the identification of generic sources of potential bias that relate to each element of the proposed model. Selection bias -------------- Since outcomes research is observational in nature, the choice of the study population and of the compared groups is highly susceptible to selection bias. As applied to outcomes research, selection bias is defined as a [distortion]{.ul} in the estimate of the practice patterns outcomes association due to the [way]{.ul} that subjects [are selected for inclusion]{.ul} in the study population and in the different groups to be compared \[[@B29]\]. A major consequence of selection bias is the potential confounding of inferences made about practice patterns-outcomes associations. This occurs when some characteristics of the subjects related to practice patterns or clinical outcomes influence the selection or exclusion of individual subjects, groups of subjects or practice environments. The selection process should be such that patients included in the study population come from the same target population \[[@B30]\]. Furthermore, patients or study members should have a similar probability of being selected and included in the actual population. Inclusion and exclusion criteria must be clearly defined in order to characterize the actual population as precisely as possible. Judging the internal validity of a study is more feasible when there is a detailed account of how the individuals were selected to become members of the actual population. Finally, the study population, also needs to be carefully characterized so that the inferences derived from the analysis of the study population can be evaluated for both internal validity (based on the data analyzed in the study) and external validity (the extent to which results obtained from the data analyzed in a particular study can be generalized to populations outside of the study). Any systematic differences between those actually studied and the source (target) population could result in biased estimates of the impact of a practice pattern on a clinical outcome. In many studies of outcomes research, groups exposed to different practice patterns are compared. The identification of such groups of patients is sought to assess the impact of different practice patterns on various outcomes in actual clinical settings and, as previously mentioned, can be used to assess practice guidelines. Because of such study design, it becomes unclear as to what the target population precisely is. Is it the group (the set of patients in a given environment) or is it the individuals receiving the various practice patterns within each group? For example, in a study of regional variations in the treatment of acute myocardial infarction in the U.S., the treatment of patients (practice patterns) was compared across different regions of the U.S. In this study, one wishes to generalize the findings about practice patterns-outcomes associations to all individuals with acute myocardial infarction (individual level). One also wishes to generalize the effect of the exposure, which is in this case practice patterns, to those prevalent in a given region (ecological level). The presence of these two levels, the individual and the ecological levels, introduces an added level of complexity in terms of the assessment of the effect of the exposure on outcome. When comparing practice patterns across regions using individual data, there is a certain degree of correlation brought about by the clustering of practice patterns that needs to be taken into account. Such a correlation is very difficult to quantify. In contrast, when assessing the effect of the exposure at the individual level, there are ecological factors (initial conditions particular to a given region) that need to be taken into account. The data originating from studies with mixed design, which are often the design of outcomes research studies, need to be analyzed with special attention to the degree of correlation between the individual covariates and to the presence of ecological exposure variables. Another potential source of selection bias is the choice of the groups to be compared, which depends on the criteria used to divide the groups. Individuals included in the groups to be compared should have the same probability of being included in these groups. Not infrequently in outcomes research, geographic criteria (such as country, regions, hospitals) are used because such criteria allow the identification of clinically comparable groups that receive very different treatments, whose resulting outcomes can then be assessed. However, such a process must be scrutinized for the possibility of selection bias other than the treatments that are being evaluated. Such selection bias would make groups not comparable as to clinical and other factors that could affect outcomes. The presence of a biased selection process could lead to confounding bias when practice patterns-outcomes associations are assessed. Such a situation may occur when the study groups are not comparable with regard to some characteristics of the subjects related to practice patterns or clinical outcomes that influenced the selection or exclusion of individual subjects, groups of subjects or practice environments. For example, in the same study of regional variations in the treatment of acute myocardial infarction, census regions of the U.S. were arbitrarily chosen as a basis for comparison. In this example, patients with similar risk of developing the outcome of interest, which is defined here as a complication after acute myocardial infarction, may not have had the same probability of being included in the different groups to be compared. Confounders may then bias the practice patterns/outcomes association if the selection of different risk groups is related to practice patterns. Selection bias can also affect the assessment of outcomes. Potential sources of this bias include loss to follow-up or missing data. Follow-up data is difficult to obtain in outcomes research studies, which often rely on administrative databases for data acquisition. Linkage, either of different databases or of the same database over time, is often performed \[[@B31]\]. A failure to link the databases for a number of individuals presents a problem equivalent to having data missing for these individuals. Information bias ---------------- The second step in outcomes research studies is the measurement of practice patterns and of the outcomes of interest. Here, issues of information bias must be considered. Information bias can be defined as a [distortion]{.ul} of the potential practice patterns outcomes association due to [misclassification]{.ul} of subjects with regard to practice patterns, outcome measures or both, or due to measurement error \[[@B29]\]. There are two major ways in which practice patterns can be misclassified. They relate to the sensitivity and specificity of the tests that are used for the diagnosis for which practice patterns are being evaluated and for the classification of the outcomes of interest. The measurement of the different practice patterns and their related outcomes largely depend on the identification of a group of patients who have a given diagnosis and require a given treatment. The characteristics that make a diagnosis more amenable to outcomes research are the following: 1) a precise diagnostic definition, 2) a diagnostic test with high sensitivity and specificity, 3) reproducibility among different individuals and locations, 4) easily coded, 5) related to a procedure, and 6) common and costly, so that it is likely to be collected in large, administrative databases frequently used in outcomes research. Because of such requirements, only a limited number of clinical conditions are amenable to outcomes research. Acute myocardial infarction is an example of a diagnosis that can be made with a high level of certainty because it has a precise diagnostic definition and well-defined diagnostic criteria, which, when taken together, have high sensitivity and specificity for the correct classification of patients. Therefore, it is easy to identify a study population that, in fact, has this disease and to describe their treatment. Thus, in order to minimize the misclassification of relevant practice patterns, the methods used to classify the disease and the outcomes that relate to the practice patterns under investigation must have high sensitivity and specificity \[[@B29],[@B31],[@B32]\]. Given the principles underlying the measurement of practice patterns and outcomes, how are the measurements generally made in outcomes research studies? The measurement of the exposure (practice patterns) in outcomes research is valid only if it corresponds to the \"true\" practice as performed in the clinical setting. Again, practice can only be \"true\" if the diagnosis is correct. The identification of both patients with the disease of interest and their treatment requires a source of information that has the features of a diagnostic test. In outcomes research, administrative databases are often used as an information source to identify a study population and to obtain data on exposure. The database coding of diagnoses and procedures can be used as a \"diagnostic test\" to identify the clinical condition for which practice patterns will be described and to classify the practice patterns themselves and the outcomes of interest. Such a \"diagnostic test\" will have higher sensitivity and specificity values for some diagnoses than for others. For example, administrative database coding will have higher sensitivity and specificity for procedure-related diagnoses (such as hip fracture) because the diagnostic code is related to a major operation and is likely to be recorded for administrative purposes. In contrast, a diagnostic criterion for osteoarthritis can be quite vague and administrative coding is likely to have very low sensitivity and specificity for this diagnosis. The use of databases as a diagnostic test must be validated in all outcomes research studies, especially those using administrative databases. Methods to validate these databases include chart reviews, *a priori* coding systems or both. These validation methods ensure that coding is as accurate and reproducible as possible, thus allowing the database to be used as a diagnostic test to identify the study population and the practice patterns and the outcomes in outcomes research. However, these validation methods are rarely used. Finally, appropriate measures of outcomes that will serve to evaluate practice guidelines must be identified. This presents a problem because most practice guidelines aim to reduce practice variations, which will, in turn, lead to improved appropriateness and quality of care. However, how appropriateness and quality of care are measured is controversial and will not be discussed here \[[@B33]-[@B97]\]. Nevertheless, defining the outcomes that will be used to evaluate practice guidelines is a crucial step in this process. Quality of life and functional status measures constitute another group of outcome measures that should be included for the evaluation of practice guidelines. These dimensions of outcomes have received more attention from health providers, while consumers have become more concerned about outcomes of care. However, these outcomes also are difficult to measure, because they rely heavily on patient interviews and questionnaires. They are likely to vary with patient expectations, culture, and climate and are thus potentially to be measured with error and be misclassified. A few reliable, valid instruments have been developed to assess health-related quality of life \[[@B91],[@B92]\], but such instruments are not easily used to collect this information from large databases. There is a need to develop instruments to measure these types of outcomes, whether they are conversion factors for existing databases (such using length of stay as a proxy for cost) or new measures that could easily be integrated in administrative databases. Such measures could include estimates of functional class or severity of illness. At present, many outcomes research studies measure mortality and disease-specific morbidity. The validity of the measurement of these outcomes is limited by the type of database that is used. For example, using death registries to obtain causes for death is a notoriously invalid source for this type of information. There are many examples of poor correlation between cause of death as established by death registries versus disease registries. Death certificates in New York City during 1992 were assessed to determine the accuracy and frequency of reporting tuberculosis as a cause of death. Of 310 persons who died with active tuberculosis in 1992 (based on a disease-specific registry), only 34% had tuberculosis listed on their death certificate. Thus, in this example, as in many others like it, using death certificates led to an inaccurate measure of disease burden \[[@B98]\]. Confounding bias ---------------- In outcomes research terms, confounding bias is present when the effect of the practice variations on the outcomes of interest is distorted because of the effects of [extraneous variables]{.ul} (variables that are causally associated with the practice variations and the outcomes of interest) \[[@B29]\]. This issue is crucial in outcomes research because, while outcomes research shares the purpose of a clinical trial (to evaluate different treatments), it primarily uses observational methods -- investigators conducting outcomes research have limited control over potentially confounding factors ([the initial conditions of individual groups of patients]{.ul}). Because outcomes research builds on existing practice variations and analyses the natural ongoing experiment, there is ample opportunity for confounding bias to invalidate any inference made about practice patterns outcome associations \[[@B99]\]. For example, variations in practice patterns could reflect variation not only in the use of a given procedure but also in the severity of disease. Assignment of patients to certain procedures on the basis of the severity of illness makes sense clinically, but in outcomes research, it is a common and important source of confounding if the procedure is either efficacious or particularly harmful in high-risk patients. Many indices have been developed to measure the severity of illness when using existing databases to correct for such confounding, but one can never be sure that this type of confounding has been entirely controlled \[[@B100],[@B101]\]. This presents an intrinsic limitation of outcomes research. Avoidance of confounding bias is limited by the source of data used to describe practice patterns, particularly when observational data, such as the large Medicare administrative databases, are used to compare outcomes among patients who receive different treatments. The potential for confounding bias arises because many factors other than the treatment under evaluation may affect patient outcomes. These factors include comorbid diseases, severity of illness, and patient, physician and environmental factors. Such factors are likely to influence treatment decisions but are difficult to capture fully in recorded data. Researchers cannot adjust for imbalances in prognostic factors that are unmeasured or poorly categorized and administrative data, in particular, may lack the precise and accurate coverage of clinical details needed to permit full and fair adjustments. Further data collection might solve this issue, but it is not always possible to collect additional information. Standard statistical modeling can attempt to adjust for the known differences between the groups, but this might not be sufficient for unmeasured differences. Several alternative methods have been suggested. One method is subgroup analysis \[[@B102]\] to adjust for unmeasured differences between groups of individuals who differ on known risk factors. Another method consists of the use of instrumental variables \[[@B103],[@B104]\]. Instrumental variables are observable factors that influence treatments but do not directly affect patient outcomes. This approach uses the so-called instrumental variables to mimic a randomization of patients to different likelihoods of receiving alternative treatments. McClellan *et al.*\[[@B103]\] applied this methodology to assess whether more aggressive use of invasive cardiac procedures improved outcomes in the elderly. In this study, the instrumental variable was the distance of the patient\'s residence from the nearest hospital with on-site angiography. The authors noted lower mortality among elderly individuals who received more aggressive treatment than among those treated more conservatively. Temporal trend bias ------------------- We propose a bias called a \"temporal trend bias\" that is particular to the use of outcomes research to evaluate practice guidelines. This bias results from the inability to control for secular trends. It reflects the fact that by the time practice guidelines are published and disseminated, new treatments and technology are being incorporated into clinical practice. Thus, it is difficult to identify a pure application of a practice guideline whose application is not undermined by recent advances in medicine and technology. For example, we evaluated the effect of a specific set of guidelines on return to work after acute myocardial infarction. The use of these guidelines had been successful in a university setting; this study assessed their use in a community setting. During the 5 years that elapsed between these two studies, practices changed. The use of guidelines was less successful in the community not only because they did not influence practice but also because usual care had grown closer to the proposed guidelines \[[@B105]\]. Ecological exposure in individual level studies ----------------------------------------------- A frequently encountered particularity of outcomes research study design is the presence of both ecological exposure and individual level covariates in the same analysis. Because the unit of analysis is a group, but inferences are made about the impact of a given practice pattern on individual outcomes, many outcomes research analyses have elements of both individual and ecological analyses \[[@B106]\]. In our study of regional variations in the treatment of acute myocardial infarction, measures describing practice patterns at the regional level, ecological exposure, (proportion of patients receiving angiography, angioplasty, and coronary artery bypass surgery) were linked to the outcome measures of mortality adjusting for individual level variables that measured severity of disease. Then, inferences were made about the use of these procedures at the patient level. Although the unit of analysis is the region, which would demand an ecological analysis, there are individual level covariates, which are likely to be correlated within each region, that need to be taken into account. When group measures are used that contain individual-level variability with some degree of correlatedness (within region) and aggregate-level variability (between regions), specific analytic tools must be used. It has been suggested that hierarchical logistic regression modeling be used to examine the interplay between sources of variation in the use of health-care services, that is, between ecological-level and individual-level sources. This type of modeling is designed to separate true variability across areas from observed variability. An application of this method is the work by Gatsonis *et al.*\[[@B107]\] who found that practice variations across regions of the U.S. in the use of angiography after acute myocardial infarction were largely explained by differences in patient characteristics and geographic region. However, states that had more on-site availability of angiography still tended to have higher angiography rates after accounting for between-region and within-region variability. After analysis for sources of variability, more reliable inferences about the associations between practice patterns and outcomes can be made. Sources of data --------------- The application of the proposed methodological framework for outcomes research largely depends on the sources of data that are used to evaluate the effect of the practice variations on outcomes \[[@B56]\]. Most commonly, the study design is a retrospective cohort analysis and the dataset that is used has been obtained either for administrative purposes (discharge databases) or for a randomized clinical trial that addressed a different question \[[@B108]\]. Less often, a prospective cohort study is designed to evaluate a particular set of practice guidelines \[[@B109]\]. Although a prospective design provides more control in data collection than a retrospective analysis, both designs are subject to selection, information and confounding biases. The ideal database to use for the evaluation of practice guidelines is one that allows the precise measurement of the practice patterns (exposure) and outcomes (disease) as well as the measurement of potential confounders (severity of illness, precision of diagnosis, socioeconomic characteristics). Unfortunately, such a database probably does not exist. The strength of administrative databases, such as that of Medicare is that they allow the observation of large numbers of patients for which practice patterns can be evaluated as they occur in actual clinical practice. Furthermore, administrative databases allow the observation of practice patterns outcomes associations in large numbers of unselected patients. However, the limitations of such databases include the missing information about potential confounding factors, such as severity of illness, and the limited ability to measure exposure and outcome accurately. Many databases that are not designed for clinical research either mismeasure patient outcomes or fail to capture outcomes that are important to both physicians and patients (such as quality of life and functional status). The control of these biases was the basis of the methodological framework for outcomes research proposed in this chapter. The application of outcomes research methods to practice guideline evaluation ----------------------------------------------------------------------------- The application of outcomes research methods to practice guideline evaluation can accomplish several goals. One important goal is the evaluation of practice guidelines, that is, to determine to what extent the guidelines accomplished their primary goals after their dissemination. We have suggested the model of chronic disease epidemiology as the methodological framework for outcomes research to evaluate practice guidelines. The steps to evaluate practice guidelines using outcomes research when the basic design is a retrospective cohort study are summarized in Figure [3](#F3){ref-type="fig"} Some limitations to the application of this model exist. The reasons for the inability of the proposed methodological framework to deal completely with the intrinsic biases in outcomes research are listed in Figure [4](#F4){ref-type="fig"}. They relate mostly to the databases usually used in studies of outcomes research. {#F3} {#F4} Summary ======= The proposed methodological framework for outcomes research to evaluate practice guidelines reflects the selection, information and confounding biases inherent in its observational nature which must be accounted for in both the design and the analysis phases of any outcomes research study. Indeed, a major limitation of outcomes research is the inability to account for unobserved heterogeneity that directly correlates with practice patterns and/or health outcomes. This may lend bias to any inferences made about practice variations and outcomes. \"Researchers cannot correct for the subtle reason doctors choose one treatment over another for a particular patient. That bias, in turn, can undermine the entire premise of outcomes research\" \[[@B110]\]. These are intrinsic properties of outcomes research that can be dealt with only in part, by applying the principles of chronic disease epidemiology. Thus, this proposed methodology can serve as a framework for the conduct of outcomes research in the evaluation of practice guidelines but its application will be limited. Competing interests =================== none declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== Macrophages, and their circulating form monocytes, are potent defenders of the integrity of our body by mediating crucial physiological and protective functions. Just to mention some : they are central actors in innate immunity and inflammatory reactions; they process and present foreign antigens either by themselves, or through their lineage descendents; they are also dead cells scavengers. Strikingly, their role as front line defense against myriad of potentially pathogenic infectious agents in the outside environment, is essential in many species from insects to humans. Macrophages use various sets of receptor proteins to react to these agents, a central role being plaid by the Toll family molecules \[[@B1],[@B2]\]. Such receptors recognize microbial lipids \[[@B3]\], lipoproteins \[[@B4],[@B5]\], microbial carbohydrates \[[@B6]\] and bacterial DNA specific patterns \[[@B7]\]. Whatever the specific type of Toll like receptors involved, their engagements induce macrophages activation. This results in the production and release of a broad variety of specialized molecules aimed at limiting the multiplication and propagation of the infectious agents (innate immunity). The front line effector molecules produced by activated macrophages are reactive oxygen species (ROS) and reactive nitrogen species (RNS), highly diffusible molecules which have strong cytotoxic activities, including against macrophages themselves \[[@B8]-[@B10]\]. Such toxic effects are potentially dangerous since extensive macrophage destructions in the body can lead to the development of the fatal complication known as septic shock. Most of the time however, auto-protective redox buffering mechanisms prevent extensive destruction of activated macrophages. The aim of the present work was to elucidate by which mechanisms macrophages stimulated by bacterial products, manage to avoid massive auto-destruction caused by RNS? Experiments were performed with cells of the murine macrophage cell line (RAW 264.7) stimulated *in vitro* with lipopolysaccharide (LPS) and/or interferon (IFN-γ), two stimuli which induce strong endogenous NO production \[[@B11]\]. Our results directly establish that macrophage resistance is tightly regulated by the expression of definite sets of auto-protective redox buffering molecules. Results ======= LPS and/or IFN-γ *in vitro*, stimulate cells of the murine RAW264.7 line along the differentiation pathway of secreting macrophages : induction of TNFα and NO production ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- We first made a detailed kinetics analysis of the response of RAW 264.4 cells to *in vitro* stimulation by LPS, or IFN, or both. We confirmed that these products stimulated the differentiation of the cells toward the phenotype of activated macrophages releasing NO and secreting TNFα. In our experiments, RAW 264.7 cells were exposed *in vitro* to either 50 units/ml of IFN-γ (IFN~50~), or 5 μg/ml LPS (LPS~5~), or a mixture of both stimulating products. Cell free culture supernatants were harvested at 0, 6, 24 and 48 H and assessed for the production of TNFα, nitrites/nitrates and citrullin. Titrating TNFα production by ELISA, the strongest stimulation was observed with either LPS or LPS + IFN, leading to abundant secretions of TNFα as early as 6 H after the beginning of the culture, with a subsequent plateau at 24 and 48 H. The quantities of TNFα released by 0.25 × 10^6^ cells initially seeded were in the range of 1400--1700 pg/ml at 24 and 48 H. Stimulation by IFN alone was less efficient inducing roughly only half of the above TNFα production (table [2](#T2){ref-type="table"}). In the same supernatants, the production of NO was evaluated by quantifying citrulline and the total nitrite/nitrate with the Griess reagent. Both products were titrated in quantity (15--30 μM range and 30--70 μM range respectively) in 24 and 48 H culture supernatants of RAW 264.7 cells stimulated by either LPS or LPS+IFN. Smaller quantities were titrated in supernatants of cells stimulated by IFN alone (Table [2](#T2){ref-type="table"}). Parallel experiments were performed to monitor NOs2 (iNOs) mRNA induction and NOs2 (iNOs) protein synthesis in RAW 264.7 cells exposed *in vitro* to either 50 units/ml of IFN-γ (IFN~50~), or 5 μg/ml LPS (LPS~5~), or a mixture of both stimulating products. Cells were pelleted at 0, 6, 24 H and 48 H. NOs2 (iNOs) mRNA was evaluated by semi-quantitative RT-PCR (Figure [1B](#F1){ref-type="fig"}), and NOs2 (iNOs) protein intracellular level by immunoblot (Figure [1C](#F1){ref-type="fig"}). NOs2 (iNOs) was undetectable both at the mRNA and protein levels in unstimulated RAW 264.7 cells. It was strongly induced at times 6, 24 and 48 H with the three stimulation protocols (Figure [1A](#F1){ref-type="fig"} &[1B](#F1){ref-type="fig"}). Only slight variations were noted depending of the stimulations used; IFN-γ alone being as previously, the weakest inducing stimulus compared to LPS and LPS + IFN-γ. {ref-type="table"}). HPRT was used as internal control for semi-quantitative estimation. **C/ NOS2 (iNOs) protein in RAW 264.7 cultured with LPS or IFN-ã or LPS+IFN-γ.** RAW 264.7 cells were treated with the different stimuli as indicated above for various times and cell pellets (1 × 10^6^ cells) were lysed with lysis buffer. Protein concentrations in samples were adjusted and electrophoresed on 15% SDS-polyacrylamide gel, then transferred to nitro-cellulose membrane and Western bloted using polyclonal rabbit anti-murine NOS2 (iNOs), as described in Materials and Methods.](1471-2172-3-3-1){#F1} ::: {#T2 .table-wrap} ::: {.caption} ###### Differentiation of RAW 264 7 cells toward secreting-activated macrophages following culture with LPS, IFN-γ or LPS+IFN-γ ::: ----------------- ---------------------------- ---------------------------- -------------------- ------------ ----------------------------- --------- ---------- ---------- --------- --------- ---------- ---------- **nitric oxide metabolism** **TNF production (pg/ml)** **nitrites/nitrates (μM)** **citruline (μM)** **Sample type** **0 h** **6 h** **24 h** **48 h** **0 h** **6 h** **24 h** **48 h** **0 h** **6 h** **24 h** **48 h** **medium** 156+/ 12 215+/ 18 332+/ 168 462 +/ 250 ND ND ND ND 20+/ 4 22+/ 5 25+/ 7 33+/ 8 **LPS** 156+/ 12 1480+/ 187 1646+/ 266 1733+/ 284 ND ND 26+/ 5 38+/ 18 20+/ 4 35+/ 12 47+/ 11 86+/ 31 **IFN** 156+/ 12 402+/ 28 681 +/ 42 819+/ 126 ND ND 4+/ 2 12+/ 3 20+/ 4 25+/ 5 34+/ 9 42+/ 8 **LPS/IFN** 156+/ 12 1392+/ 224 1502+/ 335 1754+/ 201 ND ND 13+/ 9 33+/ 15 20+/ 4 37+/ 12 43+/ 7 72+/ 22 ----------------- ---------------------------- ---------------------------- -------------------- ------------ ----------------------------- --------- ---------- ---------- --------- --------- ---------- ---------- ND: not detectable RAW 264.7 cells (250 000 cells/well) were cultured with medium or medium supplemented with either IFN-γ (50 U/ml), or LPS (5 μg/ml) or a mixture of IFN-γ (50 U/ml) + LPS (5 μg/ml). Cell-free culture supernatants were harvested at 0, 6, 24, and 48 hours. The concentrations of TNF-α (measured using an ELISA kit), nitrites/nitrates (measured using the Griess reagent) and citrulline (measured using colorimetric reaction), were evaluated in each culture supernatant as described in Methods. Data represent means +/- SEM of five independent experiments. ::: Death induction of RAW 264.7 cells stimulated by IFN-γ and/or LPS *in vitro*, with selection of a fraction of surviving activated RAW 264.7 ------------------------------------------------------------------------------------------------------------------------------------------- *In vitro* exposure of cells of the murine macrophage cell line RAW 264.7 to either 5 μg/ml LPS (LPS~5~), or 50 units/ml of IFN-γ (IFN~50~), or a mixture of both stimulating products, led to a substantial reduction of the growth rate and number of surviving cells after 24 H or 48 H in culture (Figure [1A](#F1){ref-type="fig"}). The strongest deleterious effects were observed culturing the cells with either LPS alone or both IFN and LPS, resulting in the recovery of only approximatively 20% viable cells as estimated by MTT assay (or trypan blue test) at 24 H, and 13% at 48 H. IFN alone induced less cytoxicity with 54% and 50% viable cells recovered at 24 H and 48 H respectively (Figure [1A](#F1){ref-type="fig"}). The surviving cells subsequently recovered an *in vitro* growth rate comparable to untreated parental RAW 264.7 cells in culture (not shown). Additional experiments indicated however that those surviving activated macrophages had acquired peculiar resistance capacity to exogenous oxydative and nitrosative stress. More experiments were performed to characterize this resistance phenotype and the differentiation induced cellular metabolic events responsible for this resistance. Nitrosative stress resistance of RAW 264.7 macrophages surviving LPS, or IFN, or LPS+ IFN, stimulation ------------------------------------------------------------------------------------------------------ We next established that the selected population of activated macrophages surviving *in vitro* stimulation by either LPS, or IFN, or LPS+ IFN, had acquired a nitrosative stress resistance phenotype. RAW 264.7 cells, recovered from 24 H stimulated cultures with LPS, or IFN, or LPS+ IFN, or from unstimulated cultures (control parental cells), were washed and seeded in fresh culture medium in the presence of 1 mM of the NO releasing compound diethylenetriamine nitric oxide (DETA-NO). This concentration had been established previously to reproducibly induce the apoptotic death of about 50% unstimulated parental RAW 264.7 cells after 48 H *in vitro* (Figure [2A](#F2){ref-type="fig"}). The susceptibility of LPS, or IFN, or LPS+ IFN activated macrophages to the NO releasing compound DETA-NO, was evaluated *in vitro* by measuring the number of metabolically active cells (reducing Formazan = MTT assay) or trypan blue negative cells, recovered at the indicated time. We established previously that the MTT test provides a reproducible quantitative estimate of cell viability giving comparable results with the trypan blue dye exclusion method. As can be seen in Figure [2B](#F2){ref-type="fig"}, unselected parental RAW 264.7 cells were susceptible to DETA-NO death induction *in vitro*, with only 55% cells surviving after 24 H, and 45% after 48 hours. By contrast, LPS, or IFN, or LPS+ IFN activated RAW 264.7 macrophages were less susceptible to DETA-NO death induction *in vitro*, with variations depending upon the stimulation protocol.(Fig [2B](#F2){ref-type="fig"}). Activated-macrophages recovered from cultures stimulated with LPS alone were almost completely resistant to the nitrosative stress. Activated-macrophages recovered from IFN, or IFN + LPS stimulated cultures, exhibited an intermediate NO resistance, with about 70% and 60% metabolically active cells, or trypan blue negative cells, recovered after 24 or 48 H in culture with DETA-NO. {#F2} Upregulation of redox protection/detoxification systems in LPS-differentiated NO resistant RAW 264.7 cells ---------------------------------------------------------------------------------------------------------- LPS (and/or IFN) induced NO resistance of differentiated RAW 264.7 macrophages could result from an up-regulation of their cellular redox protection/detoxification systems. We therefore evaluated the LPS (and/or IFN) induced modulation of a number of redox proteins, critical in the three major systems protecting cells against oxydative and nitrosative stress : the superoxide dismutase/catalase system (Cu/Zn SOD, Mn SOD and catalase); the glutathione system (glutathione (GSH), γ-glutamyl cystein synthase (γ-GCS), glutathione peroxidase (Gpx), gluthatione reductase (GR); and the thioredoxin system (thioredoxin (Trx), thioredoxin reductase (TR). The transcriptional modulation of several of these redox proteins was evaluated by semi-quantitative RT-PCR on 24 hours cell extracts, using the specific primers described in table [1](#T1){ref-type="table"}. As shown in Figure [3](#F3){ref-type="fig"}, Cu/Zn SOD, Mn SOD, and catalase, were significantly upregulated (130 to 200 % increased) in NO resistant activated RAW 264.7 macrophages obtained following 24 hours stimulation with LPS (and/or IFN). By contrast, γ-glutamyl cystein synthase (γ-GCS) and glutathion peroxidase (Gpx) were only marginally increased with IFN, while glutathione reductase (GR), thioredoxin (Trx) and thioredoxin reductase (TR) mRNA were not significantly affected. This gene transcription study was further documented by measuring cellular concentrations of glutathione (GSH) in lysates of activated RAW 264.7 macrophages, recovered 6, 24 or 48 hours after LPS (and/or IFN) stimulations. As can be seen in Figure [4](#F4){ref-type="fig"}, glutathione level was approximatively 3.5 folds increased at 6 hours in LPS stimulated RAW 264.7 cells, compared to basal level in unstimulated control RAW 264.7 cells. This LPS induced high glutathione level was still detected at 24 and 48 hours (2.6 and 5.2 fold increased respectively). IFN~50~ and IFN~50~/LPSs stimulations also increased intra cellular glutathione levels in RAW 264.7 cells, with the same kinetics than LPS alone, although with variable efficiency (Figure [4](#F4){ref-type="fig"}). Additional experiments were performed to quantify the cellular specific activity of several other redox protection proteins in lysates of RAW 264.7 macrophages, recovered 6, 24 or 48 hours after LPS (and/or IFN) stimulations. As can be seen in table [3](#T3){ref-type="table"}, catalase activity was approximately two folds increased at 6 and 24 hours after LPS stimulation of RAW 264.7 cells, compared to basal level in unstimulated cells, (but not after IFN~50~ or LPS~5~/IFN~50~ treatments), returning to basal level at 48 H. A sharp seven fold increased SOD activity was observed at 6 H in RAW 264.7 cells stimulated with IFN~50~ and IFN~50~/LPS~5~, which was not observed at any other time point, nor with the other stimulation protocols. No significant changes in Gpx activity was observed in stimulated cells. {#F3} {#F4} ::: {#T1 .table-wrap} ::: {.caption} ###### Pairs of synthetic primers used in RT-PCR to amplify the different molecular systems studied. ::: **Gene studied** **Sens sequence** **Anti-sens sequence** **size (pb)** ------------------ ------------------------------------------ ------------------------------------------- --------------- HPRT 5\'TGG AAT CCT GTG GCA TCC ATG AAA C 3\' 5\' TAA AAC GCA GCT CAG TAA CAG TCC G 3\' 348 catalase 5\' GCA GAT ACC TGT GAA CTG TC 3\' 5\' GTA GAA TGT CCG CAC CTG AG 3\' 229 GCS 5\' CCT TCT GGC ACA GCA CGT TG 3\' 5\' TAA GAC GGC ATC TCG CTC CT 3\' 346 Gpx 5\' CCT CAA GTA CGT CCG ACC TG 3\' 5\' CAA TGT CGT TGC GGC ACA CC 3\' 197 GR 5\' AGC CGC CTG AAC ACC ATC TA 3\' 5\' CCG TCT GAA TGC CCA CTT TA 3\' 601 NOS(2) 5\' ACG CTT CAC TTC CAA TGC AAC 3\' 5\' TGA GGG CTG ACA CAA GGC CTC 3\' 511 SOD(Cu/Zn) 5\' AAG GCC GTG TGC GTG CTG AA 3\' 5\' CAG GTC TCC AAC ATG CCT CT 3\' 246 SOD(Mn) 5\' GCA CAT TAA CGC GCA GAT CA 3\' 5\' AGC CTC CAG CAA CTC TCC TT 3\' 241 Trx 5\' CCC TTC TTC CAT TCC CTC TG 3\' 5\' AAC TCC CCC ACC TTT TGA CC 3\' 149 TR 5\' TCC TCT TTT TCT ACC CAC TG 3\' 5\' GTA TTC CTT GCT GTC ATC CA 3\' 464 ::: ::: {#T3 .table-wrap} ::: {.caption} ###### Activities of the redox enzymes catalase, Gpx and SOD, in differentiated NO resistant RAW 264 7 cells ::: -------------------- ------------ ------------ ---------- ----------- ---------- ---------- ------------ ----------- ----------- **time (hours)** **6** **24** **48** **6** **24** **48** **6** **24** **48** **control** 873+/-100 761+/-128 767+/-96 257+/-116 379+/-95 300+/-32 249+/-111 263+/-47 326+/-76 **LPS~5~** 1563+/-408 1036+/-670 738+/-90 335+/-206 249+/-55 323+/-52 454+/-174 205+/-136 231+/-169 **IFN~50~** 598+/81 555+/ 92 612+/76 420+/ 30 218+/46 248+/17 1637+/512 288+/ 85 66+/37 **LPS~5~/IFN~50~** 778+/67 698+/167 849+/ 33 280+/13 213+/61 268+/ 18 1833+/1601 284+/ 190 236+/115 -------------------- ------------ ------------ ---------- ----------- ---------- ---------- ------------ ----------- ----------- RAW 264.7 cells were cultured (10 × 10^6^ cells in 50 ml) with medium or medium supplemented with either IFN-γ (50 U/ml), or LPS (5 μg/ml) or a mixture of IFN-γ (50 U/ml) + LPS (5 μg/ml). Cells were harvested at 6, 24, and 48 hours. Activities of the enzymes : catalase (decomposition of H~2~O~2~), Gpx (oxidation of NADPH) and SOD (reaction with chromophore) were determined on cell lysates as described in Methods. Results are expressed as means +/- SEM of five independent experiments. ::: LPS induced NO resistance of differentiated RAW 264.7 macrophages is abrogated by chemical inhibitors affecting the intra-cellular redox protection/detoxification systems -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- The causal relationship between the acquisition of NO resistance and the upregulation of a selective set of redox protection/detoxification systems in LPS differentiated RAW 264.7 macrophages was established using chemical inhibitors targeting those redox systems. The compounds used were relatively specific inhibitors affecting the SOD/catalase system and the glutathione system, the two redox protection/detoxification biochemical sets significantly up-regulated in LPS-differentiated macrophages. DETC is an inhibitor of both Mn SOD and Cu/Zn SOD, ATZ is a catalase inhibitor, and BSO reduces the intra-cellular level of glutathione by inhibiting the γ-glutamyl cysteine synthase. In our experiments, RAW 264.7 cells were first differentiated *in vitro*, as previously, by stimulation with an optimal dose of LPS (5 μg/ml) for 48 h. Selected-surviving-LPS-differentiated RAW 264.7 macrophages (and control unstimulated RAW 264.7 cells) were washed and incubated for an additional 2 hours period in the presence of the redox inhibitors (or not). Subsequently, nitrosative stress resistance of the treated cells was evaluated by culturing them in the presence of 1 mM DETA-NO and assessing cell viability after 24 hours (number of metabolically active cells reducing formazan = MTT test). The results are presented in Figure [5](#F5){ref-type="fig"}. LPS differentiated macrophages were 90--100% resistant to the toxic effects of NO released by 1 mM DETA-NO *in vitro*, while unstimulated cells were susceptible to NO, with 45% cells only surviving to the same *in vitro* treatment. Noteworthy, incubation of LPS differentiated cells in the presence of the redox inhibitors abrogated their NO resistance and restored their susceptibility to the toxic effects of DETA-NO. The SOD inhibitor DETC was the most effective compound. It completely reversed the NO resistance status of LPS-differentiated macrophages : only 10% of DETC-treated LPS-differentiated macrophages were recovered at 48 hours following exposure to DETA-NO; compared to 90% recovery of LPS-differentiated macrophages cultured without DETC (Figure [5C](#F5){ref-type="fig"}). The two others redox inhibitors BSO (γ-GCS inhibitor) and ATZ (catalase inhibitor), also very significantly abrogated NO resistance in LPS-differentiated macrophages (with only 35% and 30% viable cells recovery following DETA-NO exposure of LPS-differentiated macrophages treated with these two compounds, respectively) (Figure [5A](#F5){ref-type="fig"} &[5B](#F5){ref-type="fig"}). In addition we noted that the SOD inhibitor DETC substantially sensitized unstimulated RAW 264.7 cells to the toxic effects of DETA-NO : only 10% DETC treated unstimulated RAW 264.7 cells survived NO exposure, compared to 40% control unstimulated RAW 264.7 not incubated with DETC (Figure [5C](#F5){ref-type="fig"}). {ref-type="fig"}). The % viability of cell exposed to exogenous NO was assessed after 48 hours as the percent of metabolically active cells reducing formazan (MTT test) compared to control (number of RAW 264.7 cells undifferentiated, not exposed to exogeneous NO, surviving at the end of the culture period = 100% cell viability).](1471-2172-3-3-5){#F5} Discussion ========== Macrophages mediate crucial functions protecting the organism against pathogenic infectious agents. Upon encounter with micro-organisms or stimulation by cytokines, they produce and release various sets of effector molecules aimed at destroying the foreign agents. Reactive oxygen species (ROS) and reactive nitrogen species (RNS) are their front line effector molecules \[[@B13],[@B14]\]. These highly diffusible products exert strong cytotoxic activities against micro-organisms and many cells, including against macrophages themselves. However, a fraction of activated macrophages exposed to redox stress manage to escape destruction. Our results establish that nitrosative resistance results from the adapatative tuning of the redox buffers in these cells. Reactive oxygen species and nitrogen monoxide are the earliest cytotoxic defense molecules produced by activated murine macrophages after contact with the infectious agents. Elevated oxygen production in macrophages (and polymorpho nuclear cells) relies upon the enzyme NaDPH oxydase \[[@B15]\]. Elevated and sustained production of nitric oxide (NO), a short-lived radical molecule generated from the guanido nitrogen group of L-arginine, is controled by an inducible NO synthase isoform (iNOS) \[[@B15]\]. This iNOS enzyme is tightly regulated in macrophages at variance with the two additional constitutive isoforms of NO synthase (cNOS) which are expressed in different tissues and function in basal physiological conditions. These later enzymes produce low levels of NO involved respectively in neurotransmission, and vascular relaxation \[[@B16]-[@B19]\]. Elevated production of NO in activated macrophages by the inducible NOS isoform following cytokines and/or bacterial lipopolysaccharide (LPS) challenge has been recognized as an essential anti-bacterial defense mechanism in rodent \[[@B14]\]. NO is not a strong oxidant by itself. However, NO can react with transition metals and reactive oxygen species (ROS) leading to the secondary generation of very toxic and reactive products like peroxynitrite (ONOO-) \[[@B20]\]. All these products, NO and its related nitrosocompounds, mediate toxicity in bacterial assailants. They also mediate growth arrest and apoptosis in normal mammalian cells \[[@B13]\] (including autocrine and paracrine toxicity for activated macrophages) and tumor cells \[[@B21]-[@B23]\]. Mammalian cells are equiped with intracellular protection systems, " redox buffer ", which protect them from endogen oxygen radicals produced by the respiration in mitochondria \[[@B24],[@B25]\]. Three biochemical sytems are the most important redox buffers. A predominant protection is by the Cu/Zn-dependent or Mn-dependent superoxide dismutase enzymes (SOD) which convert superoxide ions to hydrogen peroxide, then detoxified by catalase \[[@B26],[@B27]\]. A second protection is by gluthatione (GSH), an intracellular peptidic thiol molecule with both oxidant scavenger and redox regulating capacities \[[@B28]-[@B30]\]. Thiols on gluthatione combine with NO and form the less reactive product S-nitrosoglutathione (GSNO). A third protection is performed by thioredoxine (Trx) which can buffer ROS and RNS through oxidation of its intra chain disulfide bridge, reduced thioredoxine being then regenerated by the enzyme thioredoxine reductase (TR) \[[@B31]-[@B33]\]. The respective roles of these intracellular antioxidant autoprotective systems against ROS and/or RNS-mediated auto-injury in activated macrophages have not been defined in details. Using a murine macrophage cell line (RAW 264.7), abundantly producing NO upon stimulation with lipopolysaccharide (LPS) and/or interferon (IFN-γ), we examined changes in those intracellular redox homeostatic systems and, using relatively specific inhibitors, identified the most important redox protection systems. Our experiments indicate that macrophage stimulation induce an up-regulation of the two major intracellular redox buffering systems : the enzymatic SOD-catalase system and the RNS acceptor-neutralizer molecule glutathione. Such differentiated macrophages are resistant to exogenous NO This activation-induced new redox status is stable since the resistant phenotype of RAW 264.7 cells persisted after induction for at least one week in culture (data not shown). The regulation is both at the transcriptional and protein levels for SOD and catalase. The increased glutathione content in cells following stimulation is not explained by an increased transcription of the enzyme γ-glutamylcysteine synthase controling its synthesis; nor by an increased transcription of the regenerating enzyme glutathione reductase. It might result from the post translational stress-induced allosteric activation of the glutathione generating enzyme γ-glutamylcysteine synthase, or (and) from intra-cellular mobilization of an undefined inactive pool \[[@B34]\]. We did not observed any modulation of the third redox buffering system thioredoxine-thioredoxine reductase. This is at variance with our own previous results with THP1 monocytic human cells which heavily relied upon the thioredoxine system to maintain their redox homeostasis \[[@B33]\]. Such variation might represent inter species difference. The present data with RAW 264.7 cells are in good agreement with previous work from Brockaus & Brune group who reported that stably transfected RAW 264.7 cells overexpressing CuZn SOD were more resistant to endogenous or exogenous NO \[[@B35]\]. Using relatively specific inhibitors, we were able to establish a causal relationship between the upregulation of the intra cellular redox buffering systems mentioned above and the acquisition of their nitrosative stress resistant phenotype. The SOD inhibitor DETC almost completely reversed the NO resistance status of LPS-differentiated macrophages. It also substantially sensitized unstimulated RAW 264.7 cells to the toxic effects of DETA-NO. The two others inhibitors BSO (γ-GCS inhibitor) and ATZ (catalase inhibitor), also very significantly abrogated NO resistance in LPS-differentiated macrophages. Indeed the activation-induced resistance of macrophages has been studied by many groups (also designated as endotoxin-tolerized macrophages). Their results linked the resistance of macrophages to several, non exclusive, mechanisms such as : complex impaired expression and/or function of common signaling pathways \[[@B36]\] increased expression of Bcl-xL anti-apototic proteins \[[@B37]\], increased expression of Bfl-1 anti-apototic gene and simultaneous down regulation of caspase-8 mRNA \[[@B38]\], alteration of the Fas-Fas ligand transduction pathway \[[@B39]\], over-expression of CD14 receptor \[[@B40]\], autocrine cytokine regulatory network between IL-12 and IL-10 \[[@B41],[@B42]\]. The multiplicity and complexity of the macrophage responses was recently documented using the gene array technology which detected a broad spectrum of genes overexpressed in response to *S. typhimurium* and *S. typhimurium* LPS in RAW 264.7 cells \[[@B43]\]. We also observed in our experiments an increased expression of Bcl~2~ and Bcl-~XL~ and decreased expression of Bax and Bad in NO-resistant activated macrophages (not shown). Nevertheless, our results clearly demonstrated the crucial auto-protective importance of the up-regulation of the two major redox buffering systems SOD-catalase and glutathione, for nitrosative stress resistance of activated macrophage. A schematic representation of the many factors regulating survival of activated macrophages is presented in Figure [6](#F6){ref-type="fig"}. {#F6} Conclusion ========== Activated macrophages exposed to endogenous NO are submited to a redox nitrosative stress which acts as a strong selection process allowing the survival of the fraction of cells up-regulating key sets of autoprotective redox buffering molecules. What are the cascade of events and the mechanisms leading to the transcriptional activation of SOD genes and increased glutathione availability in activated macrophages, remain to be determined. Such findings might help define new strategies aimed at improving the induction of nitrosative stress resistant macrophages which could be the best efficient defenders of the organisms against infection and possibly cancer. Materials and Methods ===================== Materials --------- The monoclonal antibody (MoAb) against NOs2 (iNOs) was purchased from Pharmingen (San Diego, CA). The horseradish peroxydase-conjugated goat-anti-rabbit antibody was purchased from BIO-RAD (Bio-Rad laboratories, California). The NO donor compound 2,2\'(hydroxynitrosohydrazino)bis-ethanamine (DETA/NO) was purchased from Calbiochem. Purified LPS from *Escherichia Coli*, Leupeptin, pepstatin, phenylmethylsulfonyl fluoride (PMSF), EDTA, Hepes, CHAPS, 3-(4,5-dimethylthiazol-2-yl)-2-5-diphenyltetrazolium bromide (MTT), ATZ (aminotriazol), BSO (Buthionine sulfoximine), DETC (Diethylthiocarbamate) and DTT, were from Sigma. ECL chemoluminescence enhancer reagents were from Amersham. Cell cultures ------------- Cells of the mouse macrophage-RAW 264.7 line were obtained from the American Type Culture Collection \[[@B11]\] and cultured in RPMI 1640 medium supplemented with glutamax 1 (GIBCO/BRL/Life Technologies), 7% (v/v) heat-inactivated foetal calf serum (HyClone), 100 units/ml penicillin and 100 μg/ml streptomycin, at 37°C in a humidified air/CO~2~ (5%) atmosphere. Standart bulk cultures were with 10 × 10^6^ cells in 150 cm^2^/50 ml vials (Corning, polystyrene) in culture medium, or culture medium supplemented with either 5 μg/ml *Escherichia Coli* LPS (serotype O127:B8, from Sigma), or 50 units/ml recombinant mouse IFN-γ (Genzyme), or the combinaison of both (LPS 5 μg/ml plus IFN-γ 50 units/ml). Titration of macrophage-derived pro-inflammatory products --------------------------------------------------------- TNF-α were measured in undiluted supernatants using ELISA kit purchased from Medgenix Diagnostics (Rungis, France) according to manufacter\'s instructions, in triplicates for each sample. The total amount of nitrite/nitrate, the stable products of NO, was determined in supernatants using the Griess reagent. Briefly, 50 μl of culture supernatants collected at 6, 24 and 48 hours were mixed with 150 μl of Griess reagent (1% sulfanilamide/0,1% naphtylethylenediamine-dihydrochloride) at RT for 30 s. The absorbance at 543 nm was immediately determined on Dynex Revelation F 3.21 microplate reader. Nitrite concentrations in each sample were determined by extrapolation from a sodium nitrite standart curve. L-citrulline levels were determined by the colorimetric reaction of carbamido groups with diacetyl monoxime in acid solution \[[@B10]\]. Briefly, 30 μl of urease (25 U/ml) was added to 300 μl of supernatant, and incubated for 1 hour at 37°C. After addition of 37.5 μl trichloroacetic acid (TCA, 59% vol/vol), the precipitated proteins were pelleted by 5 minutes centrifugation at 12 000 g. 250 μl of supernatant was harvested and mixed with 300 μl of a 1:1 (vol/vol) mixture of 240 mmol diacetylmonoxime and a solution of phenazone (3 g in 104 ml H~2~O reacted with 12 mg FeSO~4~ and 21 ml H~2~SO~4~ 36N, and incubated 15 minutes at 90°C in dark. 200 μl of the reaction mixture was transferred to a microtitration plate for measurement of the optical density at 492 nm using Dynex Revelation F 3.21 microplate reader. Citrulline concentrations were determined by extrapolation from a citrulline standart curve. RNA isolation and RT-PCR analysis --------------------------------- Following *in vitro* stimulation for the indicated times with the indicated stimulus, total cellular RNAs were extracted from 2.5 × 10^6^ RAW 264.7 cells using TRI reagent (Euromedex, Souffelweyersheim, France) according to manufacturer\'s instructions. cDNAs were synthesised in 40 μl reaction mixtures containing 2 μg of RNA, 8 μl of 5X buffer (250 mM TRIS-HCL, 250 mM KCL, 50 mM MgCl~2~, 50 mM DTT and 2.5 mM spermidine), 2 μl of each dNTP 25 mM, 0.4 μl of 50 mM oligo-dT primers, 0.4 μl of RNAsin at 22 000 U/ml and 0.6 μl of Avian Myeloblastosis Virus-reverse transcriptase (Appligene, Strasbourg, France) at 42°C for 2 hours. PCR was performed using 1/20 of the cDNA reaction mixture in a 100 μl reaction volume containing 10 μl of 10X buffer (10 mM TRIS-HCL, 50 mM KCL, 1.5 mM MgCl~2~, 0.1% Triton X 100 and 0.2 mg/ml gelatine), 0.5 μl of each dNTP 25 mM, 0.4 μl of Taq Polymerase 5 000 U/ml (Thermus aquaticus, Appligene) and 1 μl of each primer 20 μM. The RT-PCR products were subjected to electrophoresis on agarose gel, stained with ethidium bromide, scanned and quantified by densitometry. The sequences of sense and anti-sense PCR-primers used for amplification and the predicted size of products are described on Table [1](#T1){ref-type="table"}. HPRT mRNA was used as internal standard. Modulation of mRNA expression in RAW 264.7 cells either undifferentiated, or differentiated, was standardized as follow. Bands were scanned and quantified by densitometry. The ratios : redox protein mRNA band intensity / HPRT mRNA band intensity were calculated for each stimulation conditions. The ratio values were then standardized taking the relevant mRNA band intensity ratio in undifferentiated RAW 264.7 cells as reference value 1. Immunoblot analysis ------------------- Pellets of 1 × 10^6^ cells were lysed in lysis buffer (25 mM HEPES, 0.5% Nonidet P40, 0.1% SDS, 0.5 M NaCl, 5 mM EDTA, 0.1 mM sodium deoxycholate, 1 mM PMSF, and 0.1 mg/ml leupeptin/pepstatin, pH 7.8) at 4°C for 30 min. Nuclei and membranes were removed by centrifugation at 12 000 g for 15 min. The amount of protein in each lysate was measured using the BSA microbiuret assay from Pierce. Loading buffer (42 mM tris-HCl, pH 6.8, 10% glycerol, 2.3% SDS, 5% 2-mercaptoethanol, and 0.002% bromophenol blue) was added to each lysate, which was subsequently boiled for 3 min and electrophoresed on an 15% SDS-polyacrylamide gel. Proteins were transferred to nitro-cellulose membrane. The membrane was saturated with TBS (150 mM NaCl, 10 mM tris HCl, pH 7.5) containing 10% skim milk overnight, washed twice 15 min with TTBS 0.1% (150 mM NaCl, 10 mM tris HCl, 0.1% Tween 20, pH 7.5), and incubated with a polyclonal rabbit-anti-murine NOS2 (iNOs) (1/1000 dilution), in TTBS 0.1% containing 3% skimmed milk for 1 h. The membrane was subsequently washed twice 15 min with TTBS 0.1%, and incubated for 45 min with the second antibody, peroxidase conjugated goat-anti-rabbit (1/5000). Bound antibodies were visualised by chemiluminescence using an ECL Western Immunoblotting Kit (Amersham). Susceptibility-resistance of RAW 264.7 cells to NO-mediated cell injury ----------------------------------------------------------------------- The susceptibility-resistance of RAW 264.7 cells exposed *in vitro* to exogenous NO was evaluated on cells distributed in 96 wells plates (4.10^4^ per well). Diethylene triamine-nitric oxide (DETA-NO) 1 mM was added to each well at t0 and incubation processed for the indicated times. This 1 mM dose induced substantial cytotoxicity of undifferentiated RAW 264.7 cells after 24 h (Figure [2A](#F2){ref-type="fig"}). After various incubation times, cell viability was assessed using 3 methods: trypan blue dye exclusion test, 3H thymidine incorporation and MTT assay (mitochondria-dependant reduction of 3-(4,5-dimethylthiazol-2-yl) 2,5-diphenyltetrazolium bromide (MTT) to formazan. The 3 tests give highly concordant results in every experiments, with or without NO-donor molecules. Results with the MTT assay only are presented. It was performed as follow : 20 μl of MTT (5 mg/ml in PBS 1X) was added in each wells and incubated at 37°C for 4 hours. Then, 150 μl of medium was removed from each well, and 100 μl of 0,5% HCl/propanol-1 was added to dissolve cristals. The net absorbance ratio (ratio A~550~-A~630~) in each well was immediately recorded using an ELISA microplate reader. The net absorbance in wells containing cells cultured in control medium was taken as the 100% viability value. The percent of viable cells exposed to DETA-NO was calculated by comparison. Determination of cellular content of glutathion ----------------------------------------------- The cellular concentration of glutathion was evaluated using the reagent kit Bioxytech GSH-400 (Bioproducts, Gagny, France) according to manufacturer instructions\', a glutathione colorimetric quantification test based upon specific evaluation of glutathione-SH groups. Activities of redox enzymes --------------------------- *In vitro* cultures of RAW 264.7 cells, in medium or medium supplemented with the differentiating agents, were interruped at the indicated times. Cells were pelleted, washed three times and lysed by three successive freezing-thawing cycles. The enzymatic activities were evaluated on cell lysates recovered after centrifugation at 12 000 g for 15 mn. The catalase activity was assayed by monitoring the decreased absorbance at 240 nm resulting from the decomposition of H~2~O~2~ for 1 min at 37°C. The assay mixture consisted of 2.98 ml of H~2~O~2~ solution from a stock solution of 0,1 ml of 30% (w/v) H~2~O~2~ diluted in 50 ml of sodium phosphate buffer, pH 7, and 20 μl of the incubation solution (0.025 mg of enzyme/ml of buffer). All assays were performed in triplicate. Activity is expressed in comparision with control activitis determined for each incubation time set as 100% value \[[@B12]\]. SOD and Gpx activities were determined using the reagent kits SOD-525 and Gpx-340 kits (from Alexi and Calbiochem respectively) following manufacturer\'s instructions. Statistical analysis -------------------- Results are reported as mean ± SD except where indicated, and differences between groups are calculated by means of the nonpaired Student\'s t-test. Experiments were performed in triplicate. Abbreviations ============= NO : nitric oxide GSH : reduced glutathione GSSG : oxidized glutathione LPS : lipopolysaccharide IFN-γ : interferon γ INOs : inducible isoform of nitric-oxide synthase GCS : γ-glutamylcysteine synthase GR : glutathione reductase GPx : glutathione peroxydase Trx : thioredoxin TR : thioredoxin reductase Cu/Zn SOD : Cu/Zn-dependent superoxide dismutase Mn SOD: Mn-dependent superoxide dismutase DETC: Diethylthiocarbamate ATZ: Aminotriazol BSO: Buthionine sulfoximine Acknowledgements ================ This work was supported by the Institut National de la Sante et de la Recherche Medicale, the Association pour la Recherche contre le Cancer (Grants n° 9974 and n° 5483), the Ministère de l\'Education Nationale de la Recherche et de la Technologie. PJF has received a research grant from the Association Française pour la Recherche Thérapeutique.

Background ========== Traditionally, the animal body cavity (coelom) has played a major role in interpretations of metazoan evolution, from groups (e.g., flatworms) lacking a coelom to those (e.g., nematodes) with a false coelom and finally to the bulk of animal phyla having a true coelom (Coelomata) \[[@B1],[@B2]\]. There has never been complete agreement on animal phylogeny and classification, but most researchers have divided living coelomate animals into deuterostomes (echinoderms, hemichordates, urochordates, cephalochordates, and vertebrates) and protostomes (arthropods, annelids, mollusks, and other phyla) based on differences in early embryonic development. An analysis of small subunit ribosomal RNA (18S rRNA) sequences challenged this arrangement by placing acoelomate and pseudocoelomate phyla in more derived positions among the protostomes, and in further defining a clade (Ecdysozoa) of molting animals that includes arthropods and nematodes \[[@B3]\]. This \"Ecdysozoa\" hypothesis has influenced diverse fields \[[@B4]\] and interpretations of developmental evolution in animals \[[@B5]-[@B7]\]. Since its publication, evidence has appeared both for and against this hypothesis \[[@B8]-[@B15]\]. Knowing the branching order of the major animal lineages, especially those three with fully sequenced genomes, is of importance to diverse fields such as medical genetics, physiology, neurobiology, paleontology, and astrobiology. With a genealogy of animals, it will be easier to determine the origins and inheritance of mutations, genes, gene functions, and structures. The three possible relationships of these animal phyla are: (I) arthropods + vertebrates, (II) arthropods + nematodes, and (III) nematodes + vertebrates. The first hypothesis corresponds to the traditional grouping Coelomata and the second corresponds to Ecdysozoa \[[@B3]\]. For convenience, we will use these names in reference to the two hypotheses while recognizing that this study, by necessity, involves only a subset of all animal phyla. The third hypothesis will be referred to as \"hypothesis III\" (Fig. [1](#F1){ref-type="fig"}). To test each hypothesis, sequence alignments of more than 100 nuclear proteins were assembled and subjected to a series of analyses designed to reveal biases that could result in an incorrect phylogeny. {#F1} Results and discussion ====================== Analyses of the individual protein datasets using neighbor-joining \[[@B16]\] show that most (62%) support Coelomata while 25% support Ecdysozoa and 13% support hypothesis III ([Supplemental Table 1](#S1){ref-type="supplementary-material"}, four-taxon analysis). Of the 25 proteins in which support for one of the three hypotheses is significant (≥ 95%), the results are 84% (21 proteins), 16% (4), and 0%, respectively; for those ten proteins with a highly significant (≥ 99%) topology, the results are 90% (9 proteins), 10% (1), and 0%, respectively. The four proteins showing significant support (\>95%) for a hypothesis other than Coelomata were reanalyzed using maximum parsimony and maximum likelihood; bootstrap values were not significant using other methods. Such divided results are typical of single-gene analyses because of limited information (\~400 amino acids), necessitating combined analysis. Coelomata was supported significantly (100% bootstrap confidence, posterior probability = 1.0) when the 100 four-taxon protein alignments were concatenated and analyzed using neighbor-joining, maximum parsimony, maximum likelihood, and Bayesian inference (Fig. [2A](#F2){ref-type="fig"}). Using the Shimodaira-Hasegawa (SH) test \[[@B17]\], this maximum likelihood topology was significantly different from the alternative hypotheses of Ecdysozoa (*P* \< 0.001) and Hypothesis III (*P* \< 0.001). These results agree with earlier studies involving 10--36 nuclear proteins \[[@B8],[@B9],[@B11]\]. {#F2} To test the stability of Coelomata to taxon sampling, we included new sequences of the planarian *Dugesia japonica* (Phylum Platyhelminthes) in 100 five-taxon protein alignments ([Supplemental Table 1](#S1){ref-type="supplementary-material"}, five-taxon analysis). The results, upon concatenation with this additional taxon, were unchanged (Fig. [2B](#F2){ref-type="fig"}): Coelomata continued to be significantly supported (≥ 98% bootstrap support, posterior probability = 1.0) and the alternative hypotheses were both rejected using the SH test (*P* \< 0.001), although the relationships among the basal phyla (Nematoda and Platyhelminthes) could not be resolved. Another potential bias is the specific taxa included in an analysis. For example, the original support for Ecdysozoa was obtained only with a particular genus of nematode, *Trichinella,* that had a short branch in the 18S rRNA tree \[[@B3]\]. To test this, phylogenies were constructed using different species of nematodes. The Coelomata hypothesis was significantly supported (≥ 97% bootstrap support, posterior probabilities = 1.0; Ecdysozoa rejected by SH test, *P* \< 0.006; Hypothesis III rejected by SH test, *P* \< 0.027) using either a genus in a different order, *Brugia* (18 proteins), or *Trichinella* (six proteins) (Fig. [2C, 2D](#F2){ref-type="fig"}). To further address the possibility that these results could be biased by taxon sampling, we included representatives from all available phyla for each protein. The results indicate that an increase in the number of taxa does not decrease single-protein support for Coelomata; in fact, the trend is the reverse (Fig. [2E](#F2){ref-type="fig"}). Simulation studies have shown that incomplete taxon sampling does not increase topological errors, and that most error is caused by limited sequence data \[[@B18]\]. In the initial study defining Ecdysozoa \[[@B3]\], rate variation was considered to be the major bias affecting the phylogenetic position of nematodes. In the 18S rRNA gene, nematodes typically have long branches indicating an increased rate of sequence change. Other nuclear genes also show this pattern, but to a lesser degree \[[@B8],[@B9]\]. Phylogenetic methods can accommodate moderate amounts of rate variation among lineages without producing an incorrect phylogeny \[[@B19]\]. However, if the rate of change is sufficiently large, longer branches in a phylogeny will sometimes attract one another \[[@B20]\]. If that happens, an ingroup species with a long branch may move to a more basal position in the tree. In analyses of the 18S rRNA gene, nematodes typically appear basal to arthropods + vertebrates. Because the use of a short branch nematode (*Trichinella*) resulted in a tree whereby nematodes clustered with arthropods, the basal position of nematodes in typical 18S analyses has been interpreted as long-branch attraction \[[@B3]\]. If nematodes cluster basally because of long-branch attraction, then the strongest support for Ecdysozoa should be obtained with the slowest evolving proteins. This was tested in an analysis of 36 nuclear proteins \[[@B8]\], but the results were equivocal. Therefore, we tested this suggestion with our four-taxon data set of 100 proteins, ordered by rate of evolution. Rate orders were determined in two ways: (i) nematode branch length and (ii) vertebrate-arthropod pairwise distance. The 100 proteins were grouped into concatenations of 10 proteins and 20 proteins to increase statistical resolution. The results show support for Coelomata at all rate orders, but the support is significant with the slowest evolving proteins, regardless of rate measure or number of proteins combined (Fig. [3](#F3){ref-type="fig"}). Concatenations of slow evolving proteins also show compositional homogeneity (pairwise disparity index test, *P* \< 0.05) \[[@B21]\], suggesting the basal position of the nematode results from true phylogenetic signal and not compositional bias. Support for Coelomata was weakest with the fastest evolving proteins (which also showed compositional heterogeneity), indicating that Ecdysozoa, not Coelomata, may be the result of a rate bias, compositional bias, or other artifact. {#F3} To ensure that this result was not affected by mutational saturation in our data set, the mean number of variants per variable site was determined for each protein and averaged for ten groups of proteins ordered by evolutionary rate (Fig [4A](#F4){ref-type="fig"}). As predicted, variable sites in the faster evolving proteins showed a higher number of variants than those in slow-evolving proteins. We also examined the minimum number of nucleotide changes required for sites where only the nematode sequence varied (Fig [4B](#F4){ref-type="fig"}). Slow-evolving proteins showed a smaller number of nucleotide changes required to alter amino acid identity, while faster evolving proteins required more changes. Thus, the nematode sequences in the slow-evolving proteins do not appear to be mutationally saturated. {#F4} Finally, the affect of lineage-specific rate variation on support for Coelomata was tested with the use of relative rate tests. Presumably, the selective elimination of genes with long branches will increase statistical support for the correct topology. Individual proteins from the four-taxon data set were each subjected to two different relative rate tests \[[@B22],[@B23]\]. Proteins determined to be rate-constant at the typically applied stringency level (5% significance), and at two greater stringency levels (10% and 40% significance) were concatenated and bootstrap support was determined using neighbor joining. The results of the two tests were similar. As stringency increased, 40--83% of proteins were rejected, and the relative nematode branch length (to the arthropod and vertebrate branches) dropped from 16% to 0%. However, in all cases, Coelomata remained highly significant (Fig. [5](#F5){ref-type="fig"}). Thus, the suggestion that a basal position of nematodes is the result of long-branch attraction \[[@B3]\] can be rejected. ![Effect of rate constancy on bootstrap support for Coelomata in four-taxon analysis. Graphs show results before application of tests (left, 0-level) followed by increasing stringency (5, 10, 40% significance) of the chi-square test \[[@B22]\] (circles) and Z-test \[[@B23]\] (triangles); the 5% level is normally used. (*A*) number of proteins passing rate constancy at each cutoff level. (*B*) relative nematode branch length upon concatenation of all rate constant proteins at each level. (*C*) bootstrap support for Coelomata for each rate-constant concatenation.](1471-2148-2-7-5){#F5} The importance of knowing the branching order of these species is illustrated by the immediate and wide acceptance of the Ecdysozoa hypothesis and its use in tracing patterns of developmental evolution \[[@B5]-[@B7],[@B10]\]. However, in the initial analysis of 18S rRNA sequences \[[@B3]\], Ecdysozoa was statistically significant only when a paralinear distance method was used; three other methods did not yield significant bootstrap support. In that study, Ecdysozoa also was not significant, using any method, when the flatworm sequence was included \[[@B3]\]. Subsequent analyses of the 18S rRNA gene have been interpreted differently \[[@B24],[@B25]\], but none has yielded statistically significant results supporting Ecdysozoa. Moreover, the molting cuticles of arthropods (chitin) and nematodes (collagen) are not homologous \[[@B4]\]. The significance of other morphological characteristics bearing on the position of nematodes continues to be debated \[[@B26]\]. Besides the 18S rRNA evidence, other genetic evidence for the grouping of nematodes and arthropods has come from qualitative interpretations of *Hox* gene \[[@B10]\] and β-thymosin \[[@B12]\] evolution. In the case of *Hox* genes, support comes from a single posterior gene sequence (Y75B8A.1) of the nematode *Caenorhabditis elegans* argued to have greater amino acid similarity with a posterior *Hox* genes of *Drosophila* and *Priapulus*\[[@B10]\]. Unfortunately, the *Hox* homeodomain is a short (60 amino acid) region with many sequence differences between these taxa. Definition of \"sequence signatures\" is qualitative and has not been tested statistically. In a subsequent study of nematode posterior *Hox* genes, other researchers were unable to determine if the simple nematode *Hox* cluster of six genes is an ancestral or a derived condition \[[@B13]\]. In the case of β-thymosin, a sequence signature also has been argued to support a grouping of *Drosophila* and *Caenorhabditis*\[[@B12]\]. However, it is a gene family known to have paralogs within animals, the position of introns differs between sequences from the two species, and only four other metazoan taxa were surveyed. In addition, knowing the presence or absence of a gene can be problematic without the complete genome sequence of an organism (in this case, genomes were known only in *Drosophila* and *Caenorhabditis*). Thus, although suggestive, it is too soon to judge the significance of this sequence signature. One difficulty with interpreting such qualitative evidence, including *Hox* gene orthology, is that almost any pattern can be found in nature if one looks. In other words, sequence signatures have not yet been surveyed systematically and objectively. In contrast, sequence evidence from randomly selected genes, analyzed phylogenetically, provides a more unbiased database amenable to statistical analysis. Conclusions =========== Although it is possible that a basal position of nematodes is the result of some unknown and widespread bias not yet identified, a simpler explanation is that the grouping of nematodes with arthropods is an artifact that arose from the analysis of a single gene, 18S rRNA. The results presented here suggest caution in revising animal phylogeny from analyses of one or a few genes or sequence signatures. Although many other aspects of animal phylogeny remain unresolved, our results indicate that insects (arthropods) are genetically and evolutionarily closer to humans (vertebrates) than to nematodes. Materials and methods ===================== DNA sequences from *Dugesia japonica*\[[@B27]\] were used to search the public protein database (Entrez) for orthologous counterparts in *Drosophila melanogaster, Caenorhabditis elegans, Homo sapiens, Arabidopsis thaliana, Saccharomyces cerevisiae,* and *Schizosaccharomyces pombe.* When available, sequences from other major animal phyla (e.g., Mollusca, Echinodermata, Annelida) also were obtained. In addition, the database was searched for all proteins from two other nematodes, *Brugia* and *Trichinella,* with orthologs in *Drosophila, Homo,* and *Arabidopsis, Saccharomyces,* or *Schizosaccharomyces. Arabidopsis* was used as the primary outgroup (95 out of 100 proteins in the four-taxon analysis, 94 out of 100 proteins for the five-taxon analysis, all proteins in *Brugia* and *Trichinella* analyses); yeast was used as the outgroup when *Arabidopsis* sequences were unavailable or paralogous. This was because many more genes were available for rooting with the plant than with the fungus. All three kingdoms are about equidistant from each other in terms of branch lengths \[[@B9]\] and therefore a plant serves about equally well for rooting an animal phylogeny as does a fungus. Orthology was assessed using reciprocal BLAST searches of the public protein database; those sequences receiving high scores in each search were also analyzed phylogenetically to ensure orthology. Short (\<100 amino acids) sequences were omitted. Sequences were aligned using Clustal X \[[@B28]\] and each alignment was visually inspected. Primary analyses of aligned protein data sets were conducted in MEGA2 \[[@B29]\]. Phylogenies were reconstructed using neighbor-joining \[[@B16]\] under a Poisson correction and a gamma distance (α = 2, or estimated from the data for combined analyses), with bootstrapping (2000 replications) for all analyses. Gamma parameters were estimated from the combined data using maximum likelihood under a Poisson correction \[[@B30]\] (4-taxon, α = 1.62; 5-taxon, α = 0.94, *Brugia,* α = 0.87; *Trichinella,* α = 0.66). In addition, phylogenetic analyses were conducted with maximum likelihood (JTT-F option) \[[@B31]\] and maximum parsimony (Max-Mini Branch & Bound option) \[[@B29]\] on combined data sets; in all cases they resulted in similar results (topology and significance) to the neighbor-joining analyses. Posterior probabilities of concatenated files were computed using Bayesian inference \[[@B32]\] (Jones model with gamma estimated from data; 10,000 generations; 4 chains with temp = 0.2). Shimodaira-Hasegawa tests \[[@B17]\] were performed in PAML \[[@B30]\] (JTT-F option, fixed gamma); p-values for each topology were recorded. Rate constancy was assessed using a chi-square test \[[@B22]\] under increasing stringency (5, 10, 40% significance levels); p-values were recorded for each protein. A Z-test \[[@B23]\] was also used under increasing stringency; z-values were recorded for each protein. Proteins determined to be rate constant at different significance levels were concatenated and analyzed in MEGA2 \[[@B29]\]. Nematode position and evolutionary distance were determined for each concatenation. New sequences, accession numbers of sequences, and sequence alignments may be found at the Evogenomics website <>. Authors\' contributions ======================= JEB carried out the bioinformatics research and drafted the manuscript. KI and TG supplied the unpublished EST sequences of the platyhelminth *Dugesia japonica.* SBH directed the research and assisted with drafting of the manuscript. Supplementary Material ====================== ::: {.caption} ###### Supplemental Table 1 Proteins analyzed and bootstrap support for each alternative hypothesis. ::: ::: {.caption} ###### Table-SuppInfo.doc ::: Acknowledgements ================ This work was supported by grants from the NASA Astrobiology Institute and National Science Foundation (to S.B.H.).

Background ========== There appears to be no one commonly accepted definition for quality of work life. In healthcare organizations, quality of work life (QWL) has been described as referring to the strengths and weaknesses in the total work environment \[[@B1]\]. Characteristics that describe the overall organization are viewed as part of the behaviour and reward system of the staff working in that setting. Organizational features such as policies and procedures, leadership style, operations, and general contextual factors of the setting, all have a profound effect on how staff view the quality of their work life. QWL is an umbrella term which includes many concepts. Therefore, concentrating on only one job characteristic, whether it is wages or management style, is an inadequate approach to assessing QWL. Because the perceptions held by employees play an important role in their decisions to enter, stay with or leave an organization, it is important that staff perceptions be included when assessing QWL. And although job satisfaction is not QWL, perception of QWL is often assessed using job satisfaction surveys. Previous studies have shown that low job satisfaction is a major cause of turnover among health care providers \[[@B2]-[@B4]\]. In addition, job satisfaction may affect the quality of service and organizational commitment \[[@B5]-[@B9]\] and may be a contributing factor associated with shortages of health care providers \[[@B10]\]. Such findings have recently increased interest in studying job satisfaction among health care providers \[[@B11]\]. The results of a 1993 meta-analysis of 48 studies looking at work satisfaction in over 15,000 nurses revealed that job satisfaction was associated strongly with reduced work stress, organizational commitment, communication with supervisors, autonomy, employee recognition, fairness, locus of control, years of experience, education, and professionalism. This study also found a strong relationship between job satisfaction and QWL for nurses \[[@B12]\]. After reviewing the literature on QWL and job satisfaction, and considering the wide variety of health care settings, situational contexts, and organizational structures (including management styles, reporting structures, staffing complements, and levels of training and experience) in which employees work, we hypothesized that the predictors of job satisfaction would vary depending on the organization. The purpose of this study was to identify organization specific predictors of job satisfaction within a health care system that consisted of six independent and distinct organizations located in five communities in Central West Ontario, Canada. Methods ======= Setting ------- The settings for this study included six independent and distinct health care organizations providing varying levels and types of care. All six organizations were affiliated with the St. Joseph\'s Health System (SJHS) located in five Central West Ontario communities. Collectively, the SJHS is one of the largest corporations in Canada devoted to health care. At the time of the study (2000), the SJHS employed 5,486 full, part and casual time (non physician) staff. Additional information about of each of the six organizations and their respective communities is provided in Table [1](#T1){ref-type="table"}. ::: {#T1 .table-wrap} ::: {.caption} ###### Characteristics of the Organizations within the St. Joseph\'s Health System. ::: Site Type of Organization No. of Staff No. of Beds or Visits/Yr Community Population^1^ -------- -------------------------------------------------- -------------- -------------------------- ------------------------- Site 1 Community Hospital 321 101 84,764 Site 2 Community Hospital / Long-Term Care Facility 649 186/124 95,821 Site 3 Visiting Nurse Organization 205 140,152 322,352 Site 4 Long-Term Care Facility 481 389 23,125 Site 5 Community Hospital 889 148 178,420 Site 6 Tertiary Care Hospital / Community Health Centre 2,941 459/88,837 322,352 ^1^1996 Census ::: Questionnaire development ------------------------- Items included in the \"Quality of Work Life Survey 2000\" were selected after a review of the literature and extensive consultation between research team members and the QWL Task Force (a management group consisting of representatives from each of the six SJHS organizations). The initial selection of items was influenced by a recently published Canadian study \[[@B13]\] and reports from two meta-analyses \[[@B1],[@B12]\]. The QWL Task Force then refined these items to consider, among other things, issues of accuracy, relevance, readability, grammar, potential for offensiveness, and appearance of cultural or gender bias. After several months of development, the instrument was pretested on a small group of staff at two of the participating organizations (Site 2 and Site 4 -- see Table [1](#T1){ref-type="table"}). This pretesting was done to ensure that individuals could follow the instructions associated with the format, to obtain estimates of the time required to complete the survey instrument, to identify items that were poorly written or ambiguous, and to identify an appropriate implementation strategy. The questionnaire and implementation strategies were revised accordingly. The final 65-item survey contained nine sections representing topic areas considered relevant to assessing QWL in the SJHS. Eight scale scores were developed from the individual items (see below and [Additional File](#S1){ref-type="supplementary-material"}: Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction). The ***Co-worker and supervisor support*** section included 10 closed-ended and 1 open-ended questions. A 3-item supervisor social support scale included questions about supervisor helpfulness, concern about the welfare of employees, and ability to facilitate effective interaction among employees. Co-worker support was measured by a 7-item scale reflecting the extent to which co-workers were seen as competent, understanding, and supportive of employees. Both scales where adapted from Woodward et al. (1999) \[[@B13]\]. The ***Teamwork and Communication*** section included 9 closed-ended and 1 open-ended questions. For determining teamwork, a 7-item scale was adapted from Taylor and Bowers (1972) to measure the extent to which one\'s work unit coordinates efforts, solves problems and works together effectively \[[@B14]\]. A 2-item scale developed for this project measured how communication was practiced within the organization. The ***Job Demands and Decision Authority*** section included 15 closed-ended and 1 open-ended questions. It included a 4-item scale adapted from Brosnan and Johnson (1980) to measure clarity regarding responsibilities, workloads and conflicting demands \[[@B15]\]. There was also a 9-item scale adapted from Karasek et al. (1998) to measure the extent to which respondents\' jobs gave them autonomy or decision-making latitude \[[@B16]\], and 2 questions which reflected the demands of one\'s work \[[@B17]\]. The ***Characteristics of Your Organization*** section included 6 closed-ended and 1 open-ended questions. This section was adapted from Woodward et al. (1999) and included a 4-item scale that inquired about the extent to which the organization encouraged the best efforts from staff, and how employees were treated \[[@B14]\]. Two additional questions examined the extent to which staff were kept informed, and organizational recognition of employee contributions. The ***Patient/Resident Care*** section included 5 closed-ended and 1 open-ended questions. The questions (developed for this project) were used to measure employees\' perceptions of the quality and timeliness of care provided for patients and residents at their respective organizations. The ***Compensation and Benefits*** section included 10 closed-ended and 1 open-ended questions. These questions were developed for this project to determine employee satisfaction concerning a number of employee benefits and level of pay. The ***Staff Training and Development*** section included 6 closed-ended and 1 open-ended questions. These questions (developed for this project) measured the extent to which each organization supports its staff in training, educational development and opportunities for advancement. The ***Overall Impressions of Your Organization*** section included 4 closed-ended and 4 open-ended questions. All of the questions (developed for this project) assessed staffs\' impressions of and overall satisfaction with their organization. The question \"Overall, how satisfied are you with your job?\" was used as the outcome variable in this study. The ***Staff Socio-Demographic Information*** section included 10 closed-ended questions (developed for this project) to collect information on gender, age, marital status, education, length of employment, supervisory status, time spent on job activities, job status and job classification. Within each of the first 8 sections, employees were asked to circle the response that best described their feelings using 5-point Likert scales. Employees were also asked for written comments pertaining to each of the sections and were provided space to comment on other issues they felt were important. Survey Procedure ---------------- Because of the diversity of organizations and staff within the SJHS, it was decided by the QWL Task Force, organization administrators and researchers that the implementation of the survey would be customized to best fit each of the organizations. It was felt that a varied approach would be more feasible for the organizations and that this would help maximize response rates. Although the procedures were not identical, all of the organizations provided as a minimum: advance notification (written or voice mail) of the survey to all staff (eligibility was based on whether the worker was active on the organization\'s pay roll at the time of the study and was not a physician); access to questionnaires for all staff (the QWL Task Force felt that each staff member in the SJHS should have the opportunity to complete a questionnaire); one or more reminder notices (e.g., letters, newsletters, voice mail, personal communication); and sealed drop off boxes for completed questionnaires. Pilot testing of the questionnaire revealed that employees felt that tracking individual employees for the purpose of follow-up (i.e., to increase response rate), violated the perception of anonymity and confidentiality. Therefore, to help ensure anonymity and confidentiality, follow-up attempts were limited to general reminder notices to all staff. Analysis -------- All closed-ended (or quantitative) responses were entered directly from the questionnaires into SPSS (version 10.0.5 for Windows, SPSS, Inc., Chicago, 1999). Prior to data analysis, most of the survey questions were re-coded. Questions which asked participants to select one response within a five point scale (never to always; very dissatisfied to very satisfied; very poor to very good; no, definitely not to yes, definitely) were collapsed into two categories. For example, for the response scale (1=very dissatisfied, 2=dissatisfied, 3=not sure, 4=satisfied, 5=very satisfied) those who indicated they were either satisfied or very satisfied were re-coded as \"satisfied\" while all others were re-coded \"not satisfied\" by default. In several instances, it was appropriate to combine two or more of the questions into a composite scale score. See \"Questionnaire Development\" section and [Additional File](#S1){ref-type="supplementary-material"}: Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction for additional details on how the composite scale scores were calculated. In total, there were eight scale scores (supervisor social support; co-worker support; teamwork; communication; role clarity; decision latitude; organization/staff relations; patient/resident care). Scale scores were generated by summing the participant responses (i.e. one to five) for all questions that made up the scale. In the rare situation where a participant failed to answer one or more of the questions that made up a scale score, missing values were replaced with mean values for that organization. Scale scores were categorized into meaningful dichotomous categories prior to analysis (e.g., satisfied or not satisfied). For the purpose of this study, QWL was operationally defined using the global question \"Overall, how satisfied are you with your job?\". Employees rated job satisfaction from very dissatisfied to very satisfied using a five point scale (very dissatisfied, dissatisfied, not sure, satisfied, very satisfied). For the analysis, however, those indicating they were either satisfied or very satisfied were considered to be \"satisfied\" with their jobs. All others were considered \"not satisfied\" with their jobs. Prior to analysis, study researchers reached a consensus on which survey questions to include as potential predictors of job satisfaction. In total, there were eight scale scores and 32 questions that were rationalized *a priori* as potential predictors of job satisfaction. Data from each of the organizations, as well as all of the organizations combined (representing the SJHS), were analyzed separately to identify predictors of job satisfaction. T-test, chi-square analyses and, when appropriate, Fisher exact tests were used to determine which of the variables were statistically associated with job satisfaction i.e., were potential predictors of job satisfaction. Descriptive information (numbers and percentages) for each of the variables was calculated by whether or not staff were satisfied with their jobs. In addition, p-values, odds ratios, and 95% confidence intervals for the odds ratios were calculated for each potential predictor of job satisfaction. Separate logistic regression analyses were used to identify the best predictors of job satisfaction for each organization and for all organizations combined (SJHS). Only variables which had a statistically significant association with job satisfaction were included in these analyses. Adjusted odds ratios and corresponding 95% confidence intervals are reported for each organization and the SJHS. The logistic regression analyses produces odds ratios which have been simultaneously adjusted for all other variables in their respective final models. The goodness of fit of the logistic regression models were assessed using the rho-squared statistic \[[@B18]\]. A rho-square value between 0.20 and 0.40 suggests a very good fit of the model. A probability level of \<0.05 was used to determine statistical significance. SPSS and Epi-Info (version 6.04a, Centers for Disease Control and Prevention, Atlanta, 1995) were used for statistical computations. Results ======= Table [1](#T1){ref-type="table"} provides additional information about each of the six health care organizations, including the type of organization, number of staff, number of beds or visits/year, and the size of the community where the organization was located. Respondent participation rate ----------------------------- Response rates are often used as an indicator of the representativeness of a sample of respondents. Of the combined 5,486 staff, 1,819 (33.2%) returned a completed questionnaire. Organization specific response rates varied from 25.3% to 55.3% (Table [2](#T2){ref-type="table"}). In an attempt to assess the representativeness of respondents, a comparison was made of available socio-demographic information between respondents and all staff within each of the organizations. Overall, female employees were more likely to respond than male employees (it should be noted, however, that the vast majority of staff (82% to 98%), were females within each of these organizations), as were full-time employees compared to part-time, casual or temporary employees. There were also some differences in respondents, across organizations, based on job classification. All organizations, however, had respondents within each job classification. A statistical estimating procedure was also used to assess how accurately respondents represent staff at each of the organizations \[[@B19]\]. This calculation suggests that the organization specific findings were accurate plus or minus 3.6% to 8.8%, 19 times out of 20 (Table [2](#T2){ref-type="table"}). ::: {#T2 .table-wrap} ::: {.caption} ###### Response rates and accuracy of responses by organization. ::: Site Number of Staff^1^ Number of Respondents Response Rate Accuracy (plus or minus 19 times out of 20)^2^ ---------- -------------------- ----------------------- --------------- ------------------------------------------------ Site 1 321 125 38.9% 8.8% Site 2 649 210 32.4% 6.8% Site 3 205 103 50.2% 6.8% Site 4 481 145 30.1% 8.1% Site 5 889 492 55.3% 4.4% Site 6 2,941 744 25.3% 3.6% **SJHS** **5,486** **1,819** **33.2%** **2.3%** ^1^Excludes physicians ^2^Standard error at the 95% confidence interval on a dichotomous variable with a 50/50 distribution. ::: Potential predictors of job satisfaction ---------------------------------------- Organization specific and combined SJHS (univariate) analyses (t-test, chi-square analyses and, when appropriate, Fisher exact tests) were used to determine which of the potential predictor variables were statistically associated with job satisfaction. Included in these analyses were the 40 potential predictor variables (8 scale scores and 32 individual questions). See [Additional File](#S1){ref-type="supplementary-material"}: Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction for a list of all variables. The number of statistically significant variables ranged from 15 to 30 depending on the organization and 32 for all organizations (SJHS) combined (see [Additional File](#S1){ref-type="supplementary-material"}: Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction). Best predictors of job satisfaction ----------------------------------- Separate logistic regression analyses were then used to identify the best predictors of job satisfaction for each organization and for all organizations combined (SJHS). All variables found to be statistically associated with job satisfaction from the univariate analyses were entered into these logistic regressions analyses. The best predictors of job satisfaction are presented in Table [3](#T3){ref-type="table"}. The ranking assigned to these variables relates to the order in which variables were added to the logistic regression models. For example, the rank \"1\" refers to the first variable that was added to the model i.e., the variable which best improved the fit of the model (or the most important variable). A more detailed description of the magnitude (as represented by the size of the odds ratios) and statistical significance (as represented by the 95% confidence intervals of the odds ratios) of the association between each of these predictors and job satisfaction is presented below for each organization and all organizations combined (SJHS). The best predictors of job satisfaction are again ranked according to their importance. All of the odds ratios presented below have been simultaneously adjusted for all other variables in their respective final logistic regression models. All logistic regression models achieved a rho-square between 0.20 and 0.40 suggesting they were very good (fitting) models for predicting job satisfaction. ::: {#T3 .table-wrap} ::: {.caption} ###### Best Predictors of Job Satisfaction^1^ Ranked by Organization^2^. ::: Predictors of Job Satisfaction Site 1 Rank Site 2 Rank Site 3 Rank Site 4 Rank Site 5 Rank Site 6 Rank **SJHS^3^ Rank** ---------------------------------------------------------- ------------- ------------- ------------- ------------- ------------- ------------- ------------------ Believes organization carries out its Mission Statement \- \- 3 1 1 1 **1** Good communication^4^ \- 1 \- \- \- 2 **2** Infrequently asked to do an excessive amount of work \- \- 1 . 3 \- **3** Good decision latitude^5^ \- \- \- 3 4 6 **4** Overall satisfaction with pay level \- 5 \- \- 5 \- **5** Satisfied organization recognizes employee contributions 1 \- \- \- \- 7 **6** Female employees \- \- \- \- 6 \- **7** Good role clarity^6^ 4 \- \- \- 7 \- **8** Satisfied the organization keeps employees informed \- 3 2 \- 2 \- **9** Good teamwork^7^ \- \- \- \- \- \- **10** Given enough time to get job done \- \- \- 4 \- 3 **11** Good organization/staff relations^8^ \- \- \- \- \- \- **12** Good decision authority^9^ 2 4 4 \- \- \- \- Satisfied with patient/resident care^10^ 3 \- \- \- \- \- \- Good supervisor social support^11^ \- 2 \- 2 \- \- \- Hours per week spent on job related activities \- \- \- 5 \- \- \- Job classification \- \- \- \- \- 4 \- Organization supports training and development 5 \- ^1^Separate logistic regression analyses were done to identify the best predictors of job satisfaction for each of the organizations and for the System as a whole. ^2^The rank reflects the order that variables were added to the logistic regression models. For example, the rank \"1\" refers to the first variable that was added to the model (i.e. the variable which best improved the fit of the model). ^3^In order to account for the organizational setting, site was forced into the final logistic regression analysis for the SJHS. ^4^Composite score from 2 (5-point) questions. A score of 8 to 10 indicates good communication. ^5^Composite score from 9 (5-point) questions. A score of 36 to 45 indicates good decision latitude. ^6^Composite score from 4 (5-point) questions. A score of 16 to 20 indicates good role clarity. ^7^Composite score from 7 (5-point) questions. A score of 28 to 35 indicates good teamwork. ^8^Composite score from 4 (5-point) questions. A score of 16 to 20 indicates good organization/staff relations. ^9^Decision authority is a subscale of decision latitude. It was found to be a better predictor of job satisfaction for this organization than decision latitude. It is a composite score from 3 (5-point) questions. A score of 12 to 15 indicates good decision authority. ^10^Composite score from 4 (5 point) questions. A score of 16 to 20 indicates satisfaction with patient/resident care. ^11^Composite score from 3 (5 point) questions. A score of 12 to 15 indicates good supervisor social support. ::: ### Site 1 (community hospital) The most important predictors of job satisfaction were: 1) being satisfied with the organization\'s recognition of employee contributions (OR 5.01, 95% CI 1.59 to 15.81), 2) good decision authority (OR 7.91, 95% CI 1.46 to 42.92), 3) being satisfied with patient resident care (OR 4.66, 95% CI 1.36 to 15.97), and 4) good role clarity (OR 4.24, 95% CI 1.16 to 15.49). The final model achieved a rho-square of 0.30. ### Site 2 (community hospital/long-term care facility) The most important predictors of job satisfaction were: 1) good open communication between staff (OR 2.55, 95% CI 1.03 to 6.35), 2) good supervisor social support (OR 6.27, 95% CI 1.36 to 29.00), 3) organization keeps staff informed (OR 3.73, 95% CI 1.51 to 9.20), 4) good decision authority (OR 3.49, 95% CI 1.25 to 9.73), and 5) being satisfied with pay level (OR 2.47, 95% CI 1.14 to 5.34). The final model achieved a rho-square of 0.24. ### Site 3 (visiting nurse organization) The most important predictors of job satisfaction were: 1) less frequently (never/seldom/sometimes) asked to do an excessive amount of work (OR 7.22, 95% CI 2.22 to 23.46), 2) being satisfied or very satisfied that the organization keeps employees informed (OR 4.52, 95% CI 1.43 to 14.32), 3) belief the organization carries out its Mission statement (OR 11.17, 95% CI 2.04 to 61.14, and 4) good decision authority (OR 5.29, 95% CI 1.32, to 21.22). The final model achieved a rho-square of 0.34. ### Site 4 (long-term care facility) The most important predictors of job satisfaction were: 1) belief the organization carries out its Mission statement (OR 4.63, 95% CI 1.77 to 12.51), 2) good supervisor social support (OR 3.32, 95% CI 1.22 to 9.04), 3) good decision latitude (OR 11.61, 95% CI 1.33 to 101.8), 4) often or always given enough time to get the job done (OR 3.05, 95% CI 1.00 to 9.35), and 5) spending 38 hours or more on the job or job related activities (OR 3.55, 95% CI 1.32 to 9.59). The final model achieved a rho-square of 0.34. ### Site 5 (community hospital) The most important predictors of job satisfaction were: 1) belief the organization carries out its Mission statement (OR 3.42, 95% CI 1.82 to 6.43), 2) satisfied that the organization keeps staff informed (OR 2.62, 95% CI 1.48 to 4.65), 3) not being asked frequently to do an excessive amount of work (OR 2.41, 95% CI 1.36 to 4.27), 4) good decision latitude (OR 5.65, 95% CI 2.09 to 15.25), 5) being satisfied with pay level (OR 2.41, 95% CI 1.37 to 4.23), 6) being female (OR 2.99, 95% CI 1.29 to 6.90), and 7) good role clarity (OR 2.45, 95% CI 1.02 to 5.86). The final model achieved a rho-square of 0.25. ### Site 6 (tertiary care hospital/community health centre) The most important predictors of job satisfaction were: 1) belief the organization carries out its Mission statement (OR 3.99, 95% CI 2.52 to 6.31), 2) good communication (OR 3.00, 95% CI 1.85 to 4.88), 3) being given enough time to get the job done (OR 2.63, 95% CI 1.58 to 4.40), 4) being a member of the nursing staff (OR 2.73, 95% CI 1.75 to 4.26), 5) good organization support for training and development (OR 3.51, 95% CI 1.59 to 7.76), 6) good decision latitude (OR 2.57, 95% CI 1.30 to 5.09) and 7) being satisfied with the organization\'s recognition of employee contributions (OR 2.05, 95% CI 1.07 to 3.91). The final model achieved a rho-square of 0.25. ### All sites combined (SJHS) The most important predictors of job satisfaction after adjusting for site were: 1) belief the organization carries out its Mission statement (OR 2.79, 95% CI 2.07 to 3.77), 2) good communication (OR 1.87, 95% CI 1.33 to 2.62), 3) less frequently being asked to do an excessive amount of work (OR, 1.80, 95% CI 1.33 to 2.43), 4) good decision latitude (OR 3.28, 95% CI 2.09 to 5.17), 5) being satisfied with pay level (OR 1.61, 95% CI 1.21 to 2.15), 6) being satisfied with the organization\'s recognition of employee contributions (OR 1.57, 95% CI 1.07 to 2.29), 7) being female (OR 2.83, 95% CI 1.81 to 4.42), 8) good role clarity (OR 1.73, 95% CI 1.17 to 2.56), 9) being satisfied that the organization keeps employees informed (OR 1.35, 95% CI 1.00 to 1.85), 10) good teamwork (OR 1.45, 95% CI 1.01 to 2.09), 11) being given enough time to get the job done (OR 1.57, 95% CI 1.10 to 2.23), and 12) good organization/staff relations (OR 2.02, 95% CI 1.13 to 3.62). The final model achieved a rho-square of 0.26. Discussion ========== The results of this survey were intended to assist decision-makers in identifying key workplace issues, as perceived by employees, in order to develop strategies to address and improve the quality of working conditions for staff within each of the individual health care organizations and the SJHS as a whole. This research represents the first step of an ongoing process to ensure better QWL for employees. In addition to the findings presented here, information from the survey\'s open-ended written comments have also been summarized for each of the six organizations (L Lohfeld, K Brazil, P Krueger, G Edward, D Lewis, E Tjam, E., personal communication, 2001) and the SJHS as a whole (St. Joseph\'s Health System Quality of Work Life Technical Reports 2000). This open-ended information provides additional and complementary information to that which is provided in this report. Together, these findings are currently being used by decision-makers at each of the organizations, and the SJHS, in an effort to improve employee QWL. It should be noted that at the time of this survey, all of the hospitals included in this study (as well all other hospitals within the Province of Ontario) were operating in an environment of restructuring and change. This was a time of anxiety for many health care professionals, hospital staff and the general public. In 1996, the Ontario government created a Health Services Restructuring Commission (HSRC) with a four year mandate to restructure Ontario\'s hospitals and health services system. The HSRC was given authority under the Public Hospitals Act and The Ministry of Health Act to direct public hospitals to change their roles, transfer services and programs, amalgamate or close. The HSRC completed its mandate, announced its decisions and was terminated in March 2000. The timing for this study was after the decisions of the HSRC were announced. All of the organizations included in this study were impacted to varying degrees either directly or indirectly the HSRC decisions. The most notable impacts occurred at Site 1 and Site 2. Site 1 (a community hospital) was ordered closed effective March 2001 with programs and services to be transferred to the other local community hospital while site 2 (a community hospital/long-term care facility) was ordered to transfer its acute care services to the other local hospital in its community thereby becoming a long-term care facility. During the time of the survey, a new building (adjacent to the current facility) for the new long-term care facility was under construction and was scheduled to open in 2002. These contextual issues could have influenced employee responses and therefore the predictors of job satisfaction for all of these organizations, particularly for site 1 and site 2. There are several positive attributes of this study. First, to our knowledge, it is the largest QWL investigation of health care workers in Canada with 1,819 completed interviews. Second, it is also unique in that we collected information from staff at six distinct and functionally diverse health care organizations. Third, because we could not find an \"off-the-shelf\" QWL instrument that suited our needs and collected all the information desired by key stakeholders, we developed (through a combination of modifying existing instruments and creating our own questions and scales) our own questionnaire. Finally, although the response rates were not as high as we would have hoped, the findings: appear to be consistent with what we expected *a priori* (the study\'s investigators had offices within 5 of the 6 organizations thus having inside knowledge about these organizations); appear consistent with the published literature; and were judged credible by management and staff at each of the sites. The statistical estimating procedure to assess how accurately respondents represent staff at each of the organizations also suggest that our findings were fairly representative of staff within these organizations, particularly the larger organizations. Conclusions =========== The results of this research show that job satisfaction is a multidimensional construct and is a product of the global evaluation of one\'s work place and context. This report provides valuable information about how employees in specific health care settings view their work environment. A number of organization specific predictors of job satisfaction were identified as a result of this study. The implications of these findings are currently being deliberated as they relate to improving QWL within each of the six health care organizations that make up the SJHS. These findings, may also be of relevance and value to employees, researchers, evaluators, human resource planners and administrators of similar health care organizations. The results of this survey can also be used as baseline measures against which the findings of future job satisfaction surveys can be compared. Such comparisons place this type of research within a continuous quality improvement framework. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Supplementary Material ====================== ::: {.caption} ###### Additional File Statistically Significant1 Organization Specific (Univariate) Predictors of Job Satisfaction. ::: ::: {.caption} ###### Click here for file ::: Acknowledgments =============== We would like to express our sincere appreciation to the QWL Task Force, to those involved in pretesting the survey, to Justin Garbedian, Project Assistant, and most importantly, to all staff who returned a completed questionnaire. This research was funded in part by the St. Joseph\'s Health System and the St. Joseph\'s Health System Research Network, Father Sean O\'Sullivan Research Centre, Hamilton, Ontario.

Background ========== *Helicobacter pylori (H. pylori)* has been etiologically linked with primary gastric lymphoma (PGL) and gastric carcinoma (GC) \[[@B1],[@B2]\]. Evidences in favor of cause and effect relationship between *H. pylori* and PGL include: higher frequency of isolation in PGL than in controls \[[@B3]\]; development of the disease during long-term follow up of infected subjects \[[@B4]\]; regression after eradication of the organism \[[@B5]\]; and rapid recurrence following re-infection \[[@B6]\]. Therefore, all components of Koch\'s postulates for establishing cause and effect relationship between *H. pylori* and PGL are fulfilled \[[@B7]\]. Evidences in favor of etiological relationship between *H. pylori* and GC are somewhat incomplete. In most studies from developed countries *H. pylori* infection has been correlated with GC \[[@B8]\]; however, the data from developing world where *H. pylori* infection is more frequent and is acquired at an early age is somewhat contradictory \[[@B9],[@B10]\]. There is no report of regression of GC after eradication of *H. pylori.* Recently, GC has been shown to develop in Mongolian gerbils after infection with *H. pylori*\[[@B11]\]. However, development of GC during follow up after fresh infection with *H. pylori* in humans has not been documented despite such report in PGL \[[@B4]\]. Accordingly we report a patient with PGL from India in whom surgical resection of the tumor combined with eradication of *H. pylori* was followed by regression of PGL. This patient developed metachronous GC associated with recrudescence / re-infection of *H. pylori.* Case presentation ================= A 32-year-old man presented in May 1998 with epigastric pain of six months duration. The pain was constant without any relation to meal or radiation to back. He also complained of anorexia, weight loss, vomiting of copious amount of stale food without any bile for four months. Melena occurred once three months ago. There was no history of fever. He denied history of gastric cancer in his family. Physical examination revealed emaciation and pallor. There was no lymphadenopathy, hepatosplenomegaly and abdominal lump. Examination of throat did not reveal any abnormality. Hemoglobin=85g/L, TLC=6.7 × 10^9^/L, ESR=78 mm/h. Chest radiograph was normal. Abdominal ultrasonogram did not reveal lymphadenopathy, hepatosplenomegaly or ascites. Upper gastrointestinal (UGI) endoscopy revealed hyperemic lower esophageal mucosa with multiple non-confluent erosions; there was excessive food residue inside the lumen and multiple hypertrophied gastric folds with erosions over these. Pyloric orifice was narrow and endoscope could not be negotiated into the duodenum. Six biopsies taken from the hypertrophied folds revealed only chronic inflammatory infiltrate and no specific pathology. *H. pylori* was not detected on histologic examination of gastric biopsies. In house rapid urease test (RUT) was positive within 15 minutes and scanning electron microscopy (SEM, Hitachi S530, Japan) from biopsies obtained from relatively normal looking areas revealed multiple round cells (might be lymphocytes, figure [1A](#F1){ref-type="fig"}) but no *H. pylori.* Serum titer for IgG antibody against *H. pylori* using ELISA based commercially available kit (Monobind Incorporation, California) was positive (21 IU/ml). Lansoprazole was started. With a diagnosis of gastric outlet obstruction, surgery was done after 2 weeks. After exploring the abdomen by midline incision thickening and nodularity of antrum were detected. There was no regional lymphadenopathy. Partial gastrectomy, removal of peri-gastric lymph nodes, closure of resected end of the duodenum in double layers and end-to-side ante-grade retro-colic gastrojejunostomy were done. Postoperative course was uneventful. Histological examination of resected specimen revealed transmural infiltration by mononuclear cells diffusely extending through muscularis mucosa into the submucosa without formation of lymphoid follicles, cellular atypia, abnormal mitotic figures (figure [2A](#F2){ref-type="fig"}) and lymphoepithelial lesions. On thorough histologic evaluation no evidence of synchronous GC was seen. Peri-gastric lymph nodes were not infiltrated by lymphoma. Bone marrow examination did not reveal any abnormality. On the basis of these morphological features a diagnosis of PGL (low-grade B cell lymphoma of MALT type by revised European-American classification) \[[@B12]\], stage I E: I~1~ (Musshoffs modification of Ann Arbor classification) \[[@B13]\] and *H. pylori* infection were made. Patient was treated with amoxycillin 1.5 g/d, omeprazole 40 mg/d and tinidazole 1 g/d for two weeks to eradicate the infection with *H. pylori.* UGI endoscopy repeated one month after completion of anti-*H. pylori* treatment revealed small sized stomach, no ulcer. RUT and Giemsa staining of biopsies obtained from proximal stomach were negative for *H. pylori.* However, a repeat UGI endoscopy done after six months revealed a benign-looking ulcer (1.5 cm in diameter) at gastrojejunostomy stoma. RUT and histology (Giemsa stain) on biopsies obtained from proximal stomach for *H. pylori* were still negative. Biopsy obtained by well technique from margin of the ulcer revealed chronic inflammatory infiltrate. Flow cytometry by standard technique \[[@B14]\] from these biopsies using anti-CD 19 (B cell marker) and anti-CD3 (T cell marker) monoclonal antibodies revealed polyclonal infiltrate with 6.6% CD 19 and 10.8% CD3 bearing cells (figure [3](#F3){ref-type="fig"}). With a possibility of benign stomal ulcer patient was treated with lansoprazole 30 mg/d for 3 months. After this period of treatment, healing of the ulcer and absence of *H. pylori* infection were documented by UGI endoscopy and negative RUT respectively (2 weeks after stopping lansoprazole). On follow up during the next six months, he remained well with improved appetite and weight gain (35 to 46 kg). He presented with recurrence of epigastric pain, anorexia and vomiting in August 1999. He had lost body weight (34 kg in August 1999). Examination revealed pallor, no peripheral lymphadenopathy and abdominal lump. UGI endoscopy: friable ulcerated growth involving most of stomach with reduced distensibility of the organ and patent gastrojejunostomy stoma. Since a possibility of recurrent PGL was considered, flow cytometry was done by standard technique \[[@B14]\] with biopsy from the tumor using anti-CD 19 (B cell marker) and anti-CD3 (T cell marker) antibodies; it revealed a polyclonal lymphocytic infiltrate consisting of 0.9% CD 19 and 3.9% CD3 bearing cells (figure [3](#F3){ref-type="fig"}). Biopsies obtained from relatively normal looking mucosa of stomach gave positive result to RUT. Histological examination revealed chronic gastritis but no intestinal metaplasia. SEM (Hitachi S530, Japan) with these biopsies revealed multiple bacilli suggestive of *H. pylori* (figure [1B](#F1){ref-type="fig"}). Histopathology of biopsy from margin of ulcers revealed multiple signet ring cells, cellular pleomorphism, mucin secretion, abnormal mitotic figures and gland formation at certain places (figure [2B](#F2){ref-type="fig"}). A diagnosis of GC and recrudescence / re-infection of *H. pylori* were made. Patient denied further treatment at this stage and left the hospital. {#F1} {#F2} {#F3} Discussion and conclusion ========================= High index of suspicion is required for early diagnosis of PGL in developing countries. Endoscopic features of PGL may be very non-specific. Endoscopic findings like slight increase in thickness of gastric folds, multiple benign looking gastric ulcers or erosions \[[@B15]\] are too non-specific to suspect the disease unless endoscopist specifically looks for it. Endoscopic mucosal biopsy may miss the diagnosis \[[@B16]\] and may show only benign looking lymphocytic infiltrate as in the present case. Molecular diagnostic methods essential for confirming the diagnosis are not widely available in developing countries including ours. Early diagnosis is important, as eradication of *H. pylori* has been commonly shown to cause regression in an early stage of the disease \[[@B17]\]. *H. pylori* has been found more often in patients with PGL \[[@B3]\], chronic *H. pylori* infection progresses from polyclonal lymphocytic proliferation to PGL \[[@B4]\], and its eradication leads to regression \[[@B5]\]. Recurrence of PGL following re-infection with *H. pylori* following eradication has been documented in the literature \[[@B6]\]. Presence of *H. pylori* infection in our patient at presentation was evidenced by positive RUT (sensitivity 92%, specificity 80%) \[[@B18]\] and serology; absence of organism in SEM might be due to sampling error as similar phenomenon has been reported earlier in presence of gastric atrophy and intestinal metaplasia, both being common in chronic gastritis and malignancy \[[@B19]\]. Though, biopsy was obtained from endoscopically normal-looking mucosa, SEM showed round cell infiltration suggesting involvement by lymphoma. Absence of *H. pylori* in that area is quite expected. Diagnosis of PGL in our patient could be further confirmed by molecular biological methods before eradication of *H. pylori.* However, typical histology picture is quite diagnostic of PGL \[[@B20]\]. We could demonstrate clinical and histologic regression of the disease following surgical resection followed by eradication of *H. pylori* (documented by negative RUT and histology obtained from unresected proximal stomach). Flow cytometry of gastric mucosal biopsy documented polyclonal lymphocytic infiltrate confirming absence of residual lymphoma. Molecular evidences of the neoplasm may persist after treatment despite histological regression \[[@B21]\]. However, eradication of *H. pylori* removes growth stimulus for PGL. Histological and molecular relapse in absence of re-infection with *H. pylori* are rare and transient \[[@B22]\]. Our patient developed GC on follow up after regression of PGL. This was temporally associated with re-infection with *H. pylori.* Positive RUT and demonstration of typical bacilli in SEM documented re-infection. Since we could not demonstrate a different genotype of *H. pylori* at the time of re-infection a possibility of recrudescence of earlier infection can not be entirely excluded though seems less likely. Development of intestinal metaplasia, a precursor of GC \[[@B23]\] and advanced cancer \[[@B11]\] has been reported after infection with *H. pylori* in Mongolian gerbils recently. This is perhaps the first report of development of GC following recrudescence / re-infection with *H. pylori* in human. We can not rule out a possibility of spontaneous development of GC. However, this seems less likely as in most cases of combined tumor reported earlier GC and PGL were synchronous in nature \[[@B24]-[@B29]\]. In a single series of 12 cases two developed metachronous tumor in both of whom GC preceded PGL \[[@B24]\]. In contrast, development of GC is expected to occur later in the course of *H. pylori* infection in patients with PGL \[[@B30]\]. The latter develops in the active phase of inflammation whereas the former in the pre-atrophic stage of gastritis \[[@B30]\]. In our patient resected specimen of PGL on thorough histologic examination did not reveal presence of synchronous GC. Therefore, as expected, GC in our patient developed later and was temporally associated with recrudescence / re-infection with *H. pylori.* Occurrence of metachronous GC after regression of PGL with eradication of *H. pylori* has been reported rarely \[[@B31],[@B32]\]. Four cases of metachronous GC after regression of PGL with *H. pylori* eradication have been reported till date \[[@B31],[@B32]\]. Development of GC in our patient was temporally associated with re-infection with *H. pylori* following regression of PGL. This is perhaps the first report of such an association. Considering several earlier studies \[[@B33],[@B34]\], which suggested reduction in risk of GC following *H. pylori* eradication, temporal association of development of GC with recrudescence / re-infection with *H. pylori* in our patient is not difficult to explain. However, development of GC within a short period of time after fresh infection with *H. pylori* is somewhat odd as GC takes time to develop after infection with *H. pylori*\[[@B35]\]. A possibility of genetic or immunological predisposition can not be excluded. Further, since the patient was already chronically infected with *H. pylori* he might have been predisposed. Some workers have suggested a possibility of immunological containment of dysplastic epithelial cells by lymphocytic infiltrate \[[@B36]\]. Whether immunological imbalance resulting from regression of lymphocytic infiltrate could contribute to rapid development of GC, which got further triggered by re-infection or recrudescence of pre-existing infection with *H. pylori* remains an open question. Development of GC in this patient might be related to changes induced by many years of *H. pylori* related gastritis rather than a few months of re-infection. However, since there was a temporal relation with re-infection, it could also be related to progression of existing changes induced over many years by re-infection. This remains a conjecture. In conclusion, we have shown regression of PGL after surgical resection combined with eradication of *H. pylori* and development of GC in association with reinfection with *H. pylori* in the present case. This may suggest the possibility of cause and effect relationship of both these malignancies to *H. pylori.* This case also highlights a unique problem in management of PGL in tropical countries where re-infection with *H. pylori* is supposed to be high \[[@B37]\]. List of abbreviations ===================== H. pylori: Helicobacter pylori; PGL: primary gastric lymphoma GC: gastric carcinoma TLC: total leukocyte count ESR: erythrocyte sedimentation rate UGI: upper gastrointestinal RUT: rapid urease test SEM: scanning electron microscopy IgG: immunoglobulin G ELISA: enzyme linked immunosorbent assay IU: international unit MALT: mucosa-associated lymphoid tissue CD: cluster of differentiation Competing interests =================== None declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We thank relatives of the patient for giving us written consent for publishing the patient\'s details.

Background ========== Mos belongs to a small family of cytoplasmic protein serine/threonine kinases having oncogenic activity \[[@B1],[@B2]\]. It is highly expressed in germ cells but barely detectable in a variety of somatic tissues \[[@B3]-[@B5]\]. Studies in Xenopus oocytes have established a role for c-mos in a) initiation of the maturation process and the meiosis I / meiosis II transition and b) in metaphase II arrest in mature oocytes \[[@B6]-[@B12]\]. In mouse c-Mos is apparently not required for initiation of maturation, however, like in Xenopus it is absolutely essential for the metaphase II arrest \[[@B13],[@B14]\]. The 124-v-mos oncogene represents one of several transforming gene isolates of the moloney murine sarcoma virus \[[@B15],[@B16]\] and shows unique constitutive protein kinase activity and enhanced transforming activity when compared to other v-Mos proteins or to c-Mos \[[@B2],[@B17]-[@B19]\]. The transforming mechanism of Mos involves signalling through the MAP kinase pathway as phosphorylation of MEK by c-Mos has been demonstrated \[[@B20]-[@B23]\] and mapping analyses have shown that Mos and Raf phosphorylate identical sites on MEK \[[@B16],[@B24]\]. The upstream events of the Mos/MEK/MAPK signalling cascade have not as yet been identified. In earlier studies we have shown that an activating mechanism of c-Mos is likely to involve a conformational change which is mimicked when a single amino acid is exchanged in the α-helix C loop of the kinase domain (Arg145-Gly) resulting in constitutive active c-Mos \[[@B19]\]. Recently Fisher and co-workers proposed an activating mechanism of c-Mos by sequential association with Hsp70 and Hsp90, in addition to phosphorylation \[[@B25],[@B26]\]. Presence of the activating Arg145-Gly amino acid substitution in 124-v-Mos does not change kinase specificity but is sufficient for constitutive kinase activity \[[@B19]\]. Hence the kinase activity of 124-v-Mos is independent of upstream effectors and we have used this oncogenic Mos derivative to identify substrates for the Mos protein kinase in vitro. Using the baculo virus expression system we have expressed active 124-v-Mos protein kinase, as demonstrated by its ability to auto-phosphorylate, predominantly on serine residues, and to phosphorylate vimentin in vitro. We have analysed a panel of acidic and basic substrates in immunocomplex protein kinase assays and identified two novel in vitro substrates for 124-v-Mos, the protein tyrosine phosphatase 1B and α/β-casein. Results ======= Three tryptic 124-v-Mos peptides include target sites for auto-phosphorylation ------------------------------------------------------------------------------ We have expressed 124-v-Mos with the baculovirus system in Sf9 insect cells and immunopurified 124-v-Mos using the anti-Mos N13 antiserum \[[@B19]\]. As a control, a Mos-unrelated protein, a synthetic kinase-inactive construct of PKC, PKCγ^K380R^\[[@B27]\], was expressed in Sf9 cells. Mos kinase assays, completed in the presence of \[γ-^32^P\]ATP, were resolved using SDS-PAGE and the Coomassie blue staining of the protein gel showed visible amounts of immunopurified 124-v-Mos (fig. [1B](#F1){ref-type="fig"}, arrowhead). The corresponding autoradiograph in figure [1A](#F1){ref-type="fig"} demonstrates that 124-v-Mos is expressed as a constitutive active protein kinase indicated by its ability to auto-phosphorylate in vitro. Further, a parallel kinase reaction was used for phosphoamino acid analyses which confirmed that 124-v-Mos auto-phosphorylation occurred predominantly on serine residues (fig. [1C](#F1){ref-type="fig"}) and a two-dimensional resolution of a tryptic digest of auto-phosphorylated 124-v-Mos showed that three tryptic peptides include auto-phosphorylation target sites (fig. [1D](#F1){ref-type="fig"}), demonstrating that auto-phosphorylation occurs on multiple sites of the Mos protein \[[@B28]\]. {#F1} 124-v-Mos phosphorylates vimentin but not tubulin in vitro ---------------------------------------------------------- Initially, we tested the kinase activity of 124-v-Mos using previously identified Mos substrates. It has been shown that 124-v-Mos, derived from mos-transformed fibroblasts, phosphorylates vimentin in vitro \[[@B29]\] and as presented here in figure [2C](#F2){ref-type="fig"}, in vitro kinase assays using immunopurified 124-v-mos from Sf9 insect cells showed strong vimentin phosphorylation. In contrast, tubulin which has been shown to be phosphorylated in vivo and in vitro by Xenopus c-Mos \[[@B30]\] was not a substrate for 124-v-Mos in vitro (fig. [2A](#F2){ref-type="fig"}). We have tested tubulin purified from various organs (mouse brain, testis and spleen) either polymerised, unpolymerised or pretreated with phosphatases but in none of these states found tubulin to be phosphorylated by 124-v-Mos (data not shown). {#F2} Demonstration of alpha and beta-casein phosphorylation by 124-v-Mos ------------------------------------------------------------------- In search of further substrates for the 124-v-Mos protein kinase we tested MBP; histone HI, H2AS, H3; protamine; protaminsulphate; purified PKC-α/-β II/γ and α- and β-casein. With the exception of α- and β-casein (fig. [3A](#F3){ref-type="fig"}) none of these substrates were phosphorylated by 124-v-Mos (data not shown). The possibility that factors other than 124-v-Mos in the immunoprecipitate might be responsible for the observed casein phosphorylation was eliminated by including a synthetic kinase-inactive construct of 124-v-Mos, 124-v-Mos^K121R^\[[@B19]\], as a control in addition to the Mos-unreleated protein, PKC\_^K380R^. A comparison of background phosphorylation on β-casein in the immunoprecipitates of both controls and 124-v-Mos specific phosphorylation showed that 124-v-Mos phosphorylates β-casein 7fold relative to background (fig. [3B](#F3){ref-type="fig"}). Critically, a tryptic digest of phosphorylated β-casein revealed that 124-v-Mos phosphorylates a specific tryptic peptide in β-casein which shows no background phosphorylation in either controls (fig. [3C](#F3){ref-type="fig"}, arrowhead) strongly supporting that 124-v-Mos is able to phosphorylate β-casein. Further, a two-dimensional phosphoamino acid analysis (fig. [3D](#F3){ref-type="fig"}) showed that 124-v-Mos phosphorylates α- and β-casein on serine and threonine residues at a ratio of 1:1. {#F3} The protein tyrosine phosphatase 1B is a novel in vitro substrate for 124-v-Mos ------------------------------------------------------------------------------- Protein tyrosine phosphatases constitute a diverse family of enzymes that can be divided into several subgroups, including receptor and non-receptor PTPs \[[@B31]\]. The non-transmembrane protein tyrosine phosphatase PTP-1B, a major intracellular PTP is widely expressed. PTP-1B has been demonstrated to be phosphorylated on multiple sites in a cell cycle specific manner whereby mitotic hyper-phosphorylation occurs, reflected by a protein mobility shift in SDS-PAGE analyses \[[@B32]\]. Using purified PTP-1B as a substrate, we show here that 124-v-Mos can phosphorylate PTP-1B in vitro (fig. [4A](#F4){ref-type="fig"}). We controlled this result by using immunoprecipitates from Sf9 cells expressing the synthetic kinase-inactive 124-v-Mos construct or purified PTP-1B alone in parallel kinase assays (fig. [4A](#F4){ref-type="fig"}). Other kinases such as PKC and CKII that phosphorylate PTP-1B in vitro are unable to induce a mobility shift of PTP-1B as observed in mitotic cells \[[@B32]\]. Likewise, as shown in figure [4B](#F4){ref-type="fig"}, a Mos-dependent phosphorylation did not result in a mobility shift of PTP-1B. ![**PTP-1B is a substrate for 124-v-Mos in vitro.** In vitro Mos kinase assays, using purified PTP-1B as a substrate, were resolved using 10% SDS-PAGE and the autoradiograph is shown in 4A. Immunoprecipitates of Sf9 cells expressing the kinase-inactive 124-v-Mos^K121R^ variant or PTP-1B alone were included as controls (A,B). A parallel kinase assay was blotted on nylon-membrane and PTP-1B was detected (B) using the PTP-1B-specific antiserum FG6 \[[@B29]\], arrowheads indicate the position of 124-v-Mos and PTP-1B.](1471-2091-3-6-4){#F4} Discussion ========== In this study we have expressed constitutive active 124-v-Mos using the baculovirus expression system and identified novel in vitro substrates for Mos by immunocomplex kinase assays. It has been shown that 124-v-Mos from mos-transformed mouse fibroblasts phosphorylates vimentin in vitro \[[@B29]\] and that v-Mos is physically associated with vimentin in transformed cells \[[@B33]\]. We have used vimentin as a positive control for 124-v-Mos kinase assays in vitro to demonstrate protein kinase activity of baculovirus expressed 124-v-Mos (fig. [2](#F2){ref-type="fig"}). It is known that the kinase activity of c-Mos is regulated by cellular factors and therefore we have chosen the oncogenic variant of c-Mos, 124-v-Mos, in our study since it is independent of activating mechanisms. Recently it has been shown that Hsp70 and Hsp90 physically interact with c-Mos in Xenopus oocytes and are required for c-Mos activation \[[@B25],[@B26]\]. Another factor controlling c-Mos kinase activity in Xenopus oocytes was identified by Chen and colleagues \[[@B34],[@B35]\] to be CKII, a tetrameric holoenzyme composed of two catalytic α-subunits and two regulatory β-subunits \[[@B36]\]. In Xenopus oocytes c-Mos kinase activity is inhibited by binding to the C-terminus of CKII β-subunit and by over-expression of the α-subunit of CKII this effect can be neutralized suggesting a binding competition between c-Mos and the α-subunit of CKII \[[@B34]\]. Another protein that interacts with c-Mos in Xenopus oocytes is tubulin. Tubulin not only co-precipitates with c-Mos but also serves as an in vivo and in vitro substrate \[[@B30]\]. In contrast, tubulin was not a substrate for 124-v-Mos in our immunocomplex kinase studies (fig. [2A](#F2){ref-type="fig"}). Possibly, this indicates that a cellular factor present in Xenopus oocytes and co-precipitating with c-Mos might be necessary for tubulin phosphorylation by the Mos protein kinase. This factor might not interact with the v-Mos protein, be absent in Sf9 insect cells or unable to interact with v-Mos. Interestingly, we have not detected any co-precipitation of the \_-subunit of CKII from Sf9 cells with 124-v-Mos in our immunoprecipitates (data not shown). However, as previously mentioned, Hsp70 is known to interact also with 124-v-Mos \[[@B26]\]. Having established that our recombinant 124-v-Mos protein is active in vitro, we tested a variety of molecules in immunocomplex kinase assays and identified α- and β-casein as very good substrates in vitro (fig. [3](#F3){ref-type="fig"}). This phosphorylation was specific to active 124-v-Mos as the overall phosphorylation on casein was significantly reduced using the synthetic kinase-inactive construct 124-v-Mos^K121R^ and more importantly, a tryptic peptide of casein was identified to be phosphorylated by 124-v-Mos only and not by either of the controls used in this study. As expected, casein phosphorylation occured on serine and threonine residues. The Mos-specific consensus phosphorylation site has not as yet been identified and only the mos-phosphorylation sites on MAP kinase kinase have been mapped revealing them to be identical to raf-phosphorylation sites \[[@B24]\]. Using the mos substrates identified in this study, it may be possible to determine the specific consensus phosphorylation site for the mos protein kinase as a basis for developing Mos-specific inhibitors. We have also identified protein tyrosine phosphatase 1B (PTP-1B) as a substrate for 124-v-mos in vitro (fig. [4A](#F4){ref-type="fig"}). PTP-1B is phosphorylated on multiple sites in vivo and during mitosis becomes hyper-phosphorylated resulting in a mobility shift in SDS-PAGE \[[@B32]\]. Protein kinase C and CKII phosphorylate PTP-1B in vitro but neither is responsible for the observed mitotic hyper-phosphorylation in vivo \[[@B32]\]. We show here that likewise PTP-1B phosphorylation by 124-v-mos is insufficient to effect a mobility shift (fig. [4B](#F4){ref-type="fig"}). PTP-1B phosphorylation occurs on serine 386, a phosphoacceptor site for Cdc2/cyclin B in vitro and serine 352, phosphorylated by an unknown kinase. The serine 352 phosphorylation site either might not be a target for Mos in vitro or PTP-1B may be sequentially phosphorylated by multiple kinases in vivo. Interestingly, it has been shown that PTP-1B hyper-phosphorylation does not occur uniquely in mitosis but also during osmotic shock and is induced by several other stress stimuli \[[@B37]\]. Given that activation of c-Mos is dependent on its interaction with the heatshock proteins, Hsp70 and Hsp90, it is tempting to speculate that the Mos kinase may phosphorylate PTP-1B also in vivo. Conclusions =========== The crucial biological functions of c-mos during meiosis have been analysed by antisense experiments in Xenopus lavis and by generating mos-deficient mice establishing mos as the main player in metaphase II arrest. In contrast, not much is known about activating mechanisms of mos and biochemical properties such as the mos-specific consensus phosphorylation site. In this study we immunopurified an oncogenic and constitutive active variant of mos, 124-v-Mos, and identified novel phosphorylation substrates, PTP1B and α- and β-casein. Our substrates represent a basis to determine the consensus mos-specific phosphorylation site and further, to analyze this phosphorylation ability functionally in vivo. Materials and Methods ===================== Protein expression and in vitro immunocomplex protein kinase assays ------------------------------------------------------------------- The construction and isolation of recombinant baculoviruses expressing active 124-v-Mos and the synthetic kinase-inactive variant of 124-v-Mos, 124-v-Mos^K121R^, is described in detail elsewhere \[[@B19]\]. According to the standard procedure published by Summers & Smith \[[@B38]\], recombinant proteins were expressed at 27°C in Sf9 cells for 48 hrs. and mos was immunopurified using the anti-Mos N13 antiserum as stated in \[[@B19]\]. Mos kinase assays were carried out in 50 \_l kinase reaction buffer (10 mM HEPES pH 7.3, 150 mM NaCl, 0.1% Triton X-100, 2 mM DTT, 15 mM MnCl~2~, 5 mM MgCl~2~, 2.5 mM β-glycerophosphate, 2.5 mM NaF, 20 μM ATP/ 10 μCi \[\_γ^32^P\]ATP), incubated for 20 min. at 25°C and stopped by the addition of Laemmli buffer. For in vitro substrate kinase assays, 2 μg of substrate was added to each kinase reaction. Phosphoproteins were resolved using 10% SDS-PAGE, Coomassie stained, dried and compared with the corresponding autoradiograph. Immunodetection of western blots were performed using the ECL system and protocol (Amersham). Substrates for in vitro immunocomplex kinase assays --------------------------------------------------- α- and β-casein (dephosphorylated, bovine origin) were purchased from Sigma and vimentin from Roche. Purified PTP-1B and the PTP-lB-specific antiserum FG6 were provided by N. Tonks, Cold Spring Harbor \[[@B32]\]. Tubulin was purified from either mouse brain, testis or spleen by F. Propst, Vienna. Two-dimensional phosphoamino acid analyses ------------------------------------------ Two-dimensional phosphoamino acid analyses were completed according to Boyle and colleagues \[[@B39]\]. Briefly, phosphoproteins were separated using SDS-PAGE, blotted on nylon-membrane and the desired protein bands were excised. The membrane strips were washed sequentially with 100% methanol and water and the phosphoproteins hydrolysed for 60 min. at 110°C in 5.7 N HCl. The hydrolysed samples were lyophilised, resuspended in 2.5% formic acid, 7.8% acetic acid and mixed at 15:1 with a non-radioactive amino acid standard (1 mg/ml of each phospho-serine, -threonine, -tyrosine; Sigma). Finally, samples were spotted on thin-layer chromatography plates and separated in two dimensions using the HTLE-7000 apparatus and manufacture\'s procedure (Two-Dimensional Peptide Mapping And Phosphoamino Acid Analysis, Featuring The Hunter Thin Layer Plate Electrophoresis System. B. Boyle & T. Hunter, C.B.S. Scientific Company, Del Mar, USA). First dimension: 20 min. electrophoresis at 0.8 bar, 1 kV in 2.5% formic acid, 7.8% acidic acid. Second dimension: 16 min. at 0.8 bar, 1.3 kV in 5% acidic acid, 0.5 % pyridine. The phosphoamino acids were fixed for 10 min. at 65°C and the standard non-radioactive amino acids visualised by spraying the chromatography plates with 0.25% ninhydrin followed by incubation for 15 min. at 65°C. The phosphoamino acids were located by comparing the autoradiograph with the stained standard amino acids. Tryptic digests and one- or two-dimensional separation of tryptic phosphopeptides --------------------------------------------------------------------------------- According to Boyle and colleagues \[[@B39]\] phosphorylated proteins were proteolytically digested with trypsin by incubating twice for 2 hrs. at 37°C, on each occasion with 10 μg trypsin (Promega, modified trypsin, sequencing grade) in 200 μl 50 mM NH~4~HCO~3~ and a two-dimensional separation of tryptic phosphopeptides was completed using the HTLE-7000 apparatus and manufacture\'s protocol: electrophoretic separation was performed on thin layer chromatography plates for 25 min. at 0.8 bar and 1 kV, followed by conventional chromatography in 39.25% n-butanol, 30.25% pyridine, 6.1% acetic acid. One-dimensional separation of tryptic phosphopeptides was achieved using 16% SDS-PAGE according to Schägger and von Jagow \[[@B40]\]. List of Abbreviations used ========================== Sf9, *Spodoptera frugiperda* cell line; MAPK, mitogen-activated protein kinase; MEK, MAP and erk kinase; Hsp, heat-shock protein; PTP, protein tyrosine phosphatase; MBP, myelin basic protein; PKC, protein kinase C; CKII, casein kinase II. Acknowledgements ================ We are most grateful to Friedrich Propst for purifying tubulin and generating the anti-Mos N13 antiserum.

Background ========== Most RNAs undergo a number of post-transcriptional modifications that increase the chemical diversity of the nucleotides. tRNAs are extensively modified, but rRNA, snRNA and mRNA also contain many modified nucleotides \[[@B1]\]. However, the functions of these nucleotides are poorly understood and few of the modifications are strongly conserved between species. Among all the modifications, only N-1 methylation of adenosine-58 in tRNA~i~^Met^ was shown to be truly essential \[[@B2]\] and in general, modified nucleosides are of minor importance for cell growth and/or survival (reviewed in ref. \[[@B3]\]). Among more than 100 naturally occurring nucleotide modifications, the simplest and most common are methylations of nucleotide bases or ribose moieties \[[@B1]\]. One example of this is the monomethylation of the exocyclic amine of guanine. The 16 S RNA of *E. coli* contains three such *N*^2^-methylguanine (m^2^G) residues, at positions 966, 1207, and 1516 \[[@B4]\]. m^2^G966 is located in the loop of a small stem-loop structure, which has been implicated in tRNA binding at the P site of the ribosome \[[@B5]\]. m^2^G1207 occurs in a region of the RNA that is involved in recognition of peptide chain termination codons \[[@B6]\], and m^2^G1516 is located in the 3\'-terminal stem loop of 16S rRNA, which contains also two *N*^6^, *N*^6^-dimethyladenosines (m^6^~2~A). In the ribosome structure, these m^2^G residues as well as other 7 modified residues come together forming a compact cage surrounding the location of the anticodon stem-loop structures of A and P site-bound tRNAs \[[@B7]\]. Nevertheless, these modifications are not essential for protein synthesis, since ribosomes constructed from totally unmodified 16S RNA are able to carry out all of the partial reactions of in vitro protein synthesis, albeit at ca. half-efficiency \[[@B8]\]. The enzymes responsible for the biosynthesis of m^2^G in 16S RNA are not completely characterized. The m^2^G MTases specific for G1207 and G966 (RsmC and RsmD, respectively, named for (r)ibosomal (s)mall subunit (m)ethyltransferase) have been purified \[[@B9],[@B10]\] and the gene encoding the RsmC protein was cloned and characterized \[[@B10]\]. Recently, sequence analysis revealed a family of related proteins in Gram-negative bacteria, encompassing subfamilies typified by RsmC, the YgjO open reading frame (ORF) believed to encode the RsmD protein, and the YbiN ORF suggested to encode the m^2^G1516 MTase \[[@B11]\]. Based on similarities between the highly conserved NPPF tetrapeptide in m^2^G MTases and the key motif of enzymes that methylate the exocyclic amino groups of adenine (m^6^A) and cytosine (m^4^C) in nucleic acids, it was proposed that these enzymes should be classified as a single family of N-MTases \[[@B11]\]. It has also been hypothesized that a long non-conserved sequence in the N-terminus of m^2^G MTases may be implicated in recognition of the target nucleic acid molecule, in analogy to other RNA and DNA MTases \[[@B11]-[@B13]\]. To learn more about the evolutionary origin and the sequence-structure-function relationships in the family of MTases, we attempted to infer their structural organization in a phylogenetic context. Results ======= Sequence analysis ----------------- The amino-acid sequences of *E. coli* RsmC and RsmD were used in PSI-BLAST database searches to identify orthologous proteins (hits reported in the 1^st^ iteration with e-values \<10^-40^ and sequence identity \> 40%; see Materials and Methods for details). The resulting multiple sequence alignment (Figure [1](#F1){ref-type="fig"}) was used to predict the secondary structure (separately for both protein families) and to precisely identify boundaries of the C-terminal catalytic domain. The Rsm sequences were divided into the N-terminal (variable) and C-terminal (conserved) domains (hereafter dubbed NTD and CTD) of approximately equal length. The sequences of NTD and CTD of RsmC and RsmD were used in additional PSI-BLAST searches and also submitted to the structure prediction MetaServer <>. The CTD exhibited similarity to a large class of genuine and predicted N-methyltransferases. Both the RsmC and RsmD CTD sequences showed highest similarity to each other (e-values in the range of 10^-20^--10^-19^ in the 1^st^ iteration), but also to a family of putative MTases from Gram-positive Bacteria and Archaea, typified by the Mj0882 protein from *M. jannaschii* (e-value 10^-19^ in the 2^nd^ iteration starting with RsmC, 10^-35^ in the 3^rd^ iteration starting with RsmD). Interestingly, the crystal structure of Mj0882 has been solved as a part of a structural genomics program (1dus; deposited in the Protein Data Bank in January 2000, cited as: LW Hung, L Huang, R Kim, SH Kim: Crystal structure and functional analysis of a hypothetical protein, Mj0882, from *Methanococcus jannaschii,* to be published). However, neither the biochemical characterization of this protein or the analysis of its structure has been published and its function remains a mystery. More distant homologs with shorter alignments reported in the CTD-initiated search included the HemK/YfcB family of protenr:*N*^5^-glutamine MTases \[[@B14],[@B15]\] and various distinct RNA and DNA MTase families (e-values \> 10^-16^ after the RsmD-initiated search converged in the 4^th^ iteration). Only Mj0882 and its close homologs could be aligned to the CTD of RsmC or RsmD over their entire length if the default PSI-BLAST parameters were used. Reciprocal database searches initiated with Mj0882 yielded its presumed orthologs from Archaea and Gram-positive Bacteria (detected in the 1^st^ PSI-BLAST iteration with e-values \< 10^-30^ and sequence identity to the query \> 40%) and CTD sequences of proteobacterial m^2^G MTases (RsmC detected in the 2^nd^ iteration, e-values ≥ 2\*10^-26^ and sequence identities ≤ 26%; RsmD detected in the 3^rd^ iteration with e values ≥ 2\*10^-27^ and sequence identities ≤ 24%). This result of reciprocal database searches strongly suggested an orthologous relationship \[[@B16]\] between the RsmC/RsmD and Mj0882 families. Further support for this prediction was obtained from the threading analysis. Among structurally characterized MTases, Mj0882 was reported as the definitely best structural template for the CTD of RsmC and RsmD (confident Pcons scores 5.78 and 6.09; see Table [1](#T1){ref-type="table"} and Table [2](#T2){ref-type="table"} for the summary of the threading analysis). ::: {#T1 .table-wrap} ::: {.caption} ###### Summary of the fold recognition analysis for RsmC-CTD. ::: Model number Pcons score Threading server reporting the model PDB ace. \# PDB entry description -------------- ------------- -------------------------------------- ------------- ----------------------- **1** **5.779** **FFAS** **1dus** **Mj0882** **2** **5.46** **INBGU** **1dus** **Mj0882** **3** **5.085** **3DPSSM** **1dus** **Mj0882** **4** **4.701** **SAM-T99** **1dus** **Mj0882** **5** **4.326** **mGENTHREADER** **1dus** **Mj0882** **6** **3.271** **FUGUE** **1dus** **Mj0882** 7 3.081 FFAS 1g6q Arg MTase Hmt1 8 3.002 SAM-T99 1g6q Arg MTase Hmt1 9 2.994 SAM-T99 1g6q Arg MTase Hmt1 10 2.99 SAM-T99 1f31 Arg MTase Pmt3 Hits to the Mj0882 structure are shown in bold. Comprehensive structure prediction results are available at the URL: <> ::: ::: {#T2 .table-wrap} ::: {.caption} ###### Summary of the fold recognition analysis for RsmD-CTD. ::: Model number Pcons score Threading server reporting the model PDB ace. \# PDB entry description -------------- ------------- -------------------------------------- ------------- ----------------------- **1** **6.092** **FFAS** **1dus** **Mj0882** **2** **5.856** **INBGU** **1dus** **Mj0882** **3** **5.43** **3DPSSM** **1dus** **Mj0882** **4** **5.132** **SAM-T99** **1dus** **Mj0882** **5** **4.893** **mGENTHREADER** **1dus** **Mj0882** **6** **4.171** **PDB-Blast** **1dus** **Mj0882** **7** **3.544** **FUGUE** **1dus** **Mj0882** 8 2.907 FFAS 2adm MTase 9 2.906 SAM-T99 1f31 Arg MTase Prmt3 10 2.874 SAM-T99 1g6q Arg MTase Hmt1 Hits to the Mj0882 structure are shown in bold. Comprehensive structure prediction results are available at the URL: <> ::: ::: {#T3 .table-wrap} ::: {.caption} ###### Summary of the fold recognition analysis for RsmC-NTD. ::: Model number Pcons score Threading server reporting the model PDB ace. \# PDB entry description -------------- ------------- -------------------------------------- ------------- ------------------------ **1** **3.615** **INBGU** **1DUS** **Mj0882** **2** **3.576** **FUGUE** **1DUS** **Mj0882** **3** **3.266** **FFAS** **1DUS** **Mj0882** **4** **3.237** **SAM-T99** **1DUS** **Mj0882** **5** **3.225** **3DPSSM** **1DUS** **Mj0882** **6** **2.624** **mGENTHREADER** **1DUS** **Mj0882** 7 2.091 3DPSSM 1EIZ O-MTase 8 1.812 SAM-T99 1BHJ glycine N-MTase 9 1.802 mGENTHREADER 1DL5 protein isoAsp-O-MTase 10 1.673 SAM-T99 1G6Q Arg MTase Hmt1 Hits to the Mj0882 structure are shown in bold. Comprehensive structure prediction results are available at the URL: <> ::: ::: {#T4 .table-wrap} ::: {.caption} ###### Summary of the fold recognition analysis for RsmD-NTD. ::: Model number Pcons score Threading server reporting the model PDB ace. \# PDB entry description -------------- ------------- -------------------------------------- ------------- ----------------------- **1** **3.805** **FUGUE** **1DUS** **Mj0882** **2** **3.573** **INBGU** **1DUS** **Mj0882** **3** **3.379** **FFAS** **1DUS** **Mj0882** **4** **3.242** **3DPSSM** **1DUS** **Mj0882** **5** **2.687** **mGENTHREADER** **1DUS** **Mj0882** **6** **2.262** **SAM-T99** **1DUS** **Mj0882** 7 1.864 SAM-T99 1 1F3L Arg MTase Pmt3 8 1.587 FFAS 1G6Q Arg MTase Hmt1 9 1.51 INBGU 1VID catechol O-MTase 10 1.426 FFAS 1D2H glycine N-MTase Hits to the Mj0882 structure are shown in bold. Comprehensive structure prediction results are available at the URL: <> ::: ::: {#T5 .table-wrap} ::: {.caption} ###### Summary of the fold recognition analysis for YbiN. ::: Model number Pcons score Threading server reporting the model PDB ace. \# PDB entry description -------------- ------------- -------------------------------------- ------------- ------------------------ **1** **3.406** **FFAS** **1DUS** **Mj0882** **2** **3.267** **3DPSSM** **1DUS** **Mj0882** **3** **3.083** **SAM-T99** **1DUS** **Mj0882** **4** **2.618** **mGENTHREADER** **1DUS** **Mj0882** **5** **2.552** **PDB-Blast** **1G6Q** **Arg MTase Hmt1** **6** **2.547** **INBGU** **1DUS** **Mj0882** 7 2.481 PDB-Blast 1DL5 protein isoAsp-O-MTase 8 2.467 PDB-Blast 1F3L Arg MTase Pmt3 9 2.462 FFAS 1G6Q Arg MTase Hmt1 10 2.4 PDB-Blast 1G6Q Arg MTase Hmt1 Hits to the Mj0882 structure are shown in bold. Comprehensive structure prediction results are available at the URL: <> ::: ![Multiple alignment of the Mj0882, RsmC, RsmD, and YbiN families. Sequences are denoted by the species\' name and the NCBI gene identification number; sequences lacking the number were obtained from the unfinished genome data. Numbers on the left side indicate the index of the first N-terminal residue shown, numbers in parentheses indicate how many residues in an insertion were omitted for clarity. Conserved motifs are labeled according to the nomenclature described for the AdoMet-dependent MTase superfamily \[[@B12]\]. Conserved residues are highlighted in black; the residues with invariant physicochemical character (hydrophobic, small etc.) are highlighted in gray. Conserved carboxylates predicted to bind AdoMet are shown in black on red background; the pair of residues proposed to make van der Waals contacts to guanine is shown in green on a yellow background; the Asn residue (His in Archaeal YbiN orthologs) proposed to catalyze the methyltransfer is shown in yellow on red background; the carboxylate proposed to bind the ribose of the target nucleoside is shown in yellow on magenta background. Secondary structure elements of Mj0882, extracted from the 1dus coordinates and virtually identical for the secondary structure independently predicted for other families (see Methods), are shown above the alignment.](1471-2105-3-10-1){#F1} In contrast, querying PSI-BLAST with the NTD sequences did not reveal any significant similarities. The NTDs could only be aligned to each other, displaying a number of conserved residues, if the full-length RsmC or RsmD sequences were used in the PSI-BLAST search. This suggested that the NTD could represent a novel type of a variable domain fused to the common catalytic domain or a strongly diverged version of some known nucleic acid binding domain. Quite unexpectedly, all threading algorithms (see Methods) reported that the sequences of the NTD of RsmC and RsmD are compatible with the MTase fold. Moreover, the majority of servers reported the Mj0882 structure as the best template with significant scores (Table [3](#T3){ref-type="table"} and Table [4](#T4){ref-type="table"}). In the rankings produced by the consensus server Pcons, the top 6 hits corresponded to the Mj0882 structure with the best score of 3.615 (RsmC NTD) and 3.805 (RsmD NTD). These results strongly suggested that the NTD and the CTD in bacterial m^2^G MTases are related to Mj0882 and to each other. Nevertheless, we found that the NTD lacked many residues that may be important for the enzymatic activity of Mj0882 and the CTD (see below). Hence, the NTDs could be catalytically inactive versions of the MTase domain. Therefore, the sequences of NTD were not included in subsequent studies of the m^2^G MTase AdoMet-binding site and the active site based on analysis of conserved residues. The YbiN family of predicted N-MTases has also been subjected to sequence comparisons and threading analyses. These results confirmed the previously observed similarity to sequences of RsmC and RsmD CTD \[[@B11]\] as well as to the Mj0882 family. PSI-BLAST reported similarity of YbiN to RsmC CTD in the 3^rd^ iteration with the e-value of 8 \* 10^-5^, whereas Pcons ranking of the top 4 hits corresponded to the Mj0882 structure with the best score of 3.41 (Table [5](#T5){ref-type="table"}). In the PSI-BLAST searches and in sequence profile comparisons using the FFAS algorithm \[[@B17]\], the YbiN family showed similarity to members of the HemK/YfcB family with scores comparable to that of the RsmC/RsmD/Mj0882 family (data not shown). It has been recently shown that the HemK family comprises MTases, suggesting that the (D/N/S)-P-P-(F/Y/W/H) motif is characteristic for a subset of MTases that methylate -NH~2~ groups of various substrates, rather than only bases in nucleic acids \[[@B14],[@B15]\]. Although the predicted key catalytic motif of the YbiN family (NPPFH) resembles more that of *bona fide* guanine-N2 MTases RsmC and RsmD than of any other N-MTase family, it cannot be excluded that these proteins methylate other substrates than guanine in RNA. We hope that this analysis will facilitate the choice of representative targets for experimental characterization, which would allow more precise inference of biological function (target specificity) and relationships between various families of N-MTases. Another piece of information that should shed light onto the broad picture of the N-MTase family is the previously unnoticed high similarity of the YbiN family to uncharacterized proteins from Eukaryota and Archaea (BLAST e-values 10^-27^, and 4\*^-04^, respectively). Their close similarity (confirmed in reciprocal searches with similar e-values) indicates an orthologous relationship to bacterial YbiN proteins. This was confirmed in reciprocal searches initiated with the eukaryotic and archaeal members of the family. Interestingly, while the predicted eukaryotic YbiN orthologs possess all sequence features typical for the bacterial MTases, their archaeal counterparts lack the key Asn residue in motif IV, suggesting that they may be inactive, unless the catalytic side chain \"migrated\" to another position in the primary sequence (see below). Regardless of the possible loss of functional residues, the presence of common motifs in the YbiN and Mj0882 families as well as in the NTD and CTD sequences of proteobacterial m^2^G MTases was confirmed using the Gibbs sampling procedure \[[@B18]\]. Motifs I, VI and VIII were detected with probability of occurring by chance \< 10^-13^. In addition, there is a good correlation of secondary structure elements independently predicted for all protein groups (Figure [1](#F1){ref-type="fig"}). The orthologous/paralogous relationships inferred from the results of PSI-BLAST searches were confirmed by phylogenetic analysis using the neighbor-joining method of Saitou and Nei \[[@B19]\]. The bacterial YbiN sequences and their archaeal and eukaryotic homologs grouped together with 100% bootstrap support and the RsmC/RsmD NTD, RsmC/RsmD CTD and Mj0882 families formed groups with bootstrap support \> 50% (data not shown). Taken together, these findings strongly suggest that the two domains of the proteobacterial m^2^G MTases RsmC and RsmD, and the predicted MTases YbiN and Mj0882 evolved from the common ancestor. An alternative scenario can be envisaged, in which the pseudodimeric structure of RsmC and RsmD results not from duplication but a fusion of the ancestral CTD with another distantly related MTase that has now degenerated. Nevertheless, fold-recognition results suggest that among various MTases of known structure, the NTD clearly shows highest similarity to Mj0882 (the archaeal counterpart of the CTD) (Table [3](#T3){ref-type="table"} and Table [4](#T4){ref-type="table"}). In addition, comparison of the sequence profile of the NTD sequences with profiles of other MTase families (unpublished data) using the FFAS algorithm \[[@B17]\] confirmed that the Mj0882/CTD family is the best match (data not shown), as reported in the course of fold recognition searches against the profiles comprising proteins from PDB. Moreover, the scenario involving intragenic duplication of the catalytic domain in the ancestor of the RsmC/RsmD lineage, which generated a functional pseudodimer, is supported by the dimeric structure of Mj0882 in the crystal (see below). Determination of the crystal structure of a member of the RsmC/RsmD lineage will provide insight into the relationship of the inactivated NTD to known MTase structures and the mutual orientation of the two domains. Structure-based interpretation of sequence alignment and prediction of the active site -------------------------------------------------------------------------------------- The Mj0882 structure was solved in the absence of any ligands. Prompted by the identification of close relationships between the family of m^2^G MTases and Mj0882, we sought to identify the cofactor- and guanine-binding sites in the 1.8 A crystal structure of the latter protein based on comparison with structures of known N-MTases. Searching the Protein Data Bank \[[@B20]\] using VAST \[[@B21]\]<> and DALI \[[@B22]\]<> revealed that the Mj0882 structure is more similar to various MTases alkylating hydroxyl groups, than to other nucleic acid N-MTases. This is not surprising, since previous evolutionary analysis of MTase structures and sequences demonstrated that some nucleic acid N-MTase families do not group together in the superfamily tree. The indication is that the common motif (N/D/S)-(I/P)-P-(F/Y/W/H) evolved independently several times on the common framework \[[@B23]\]. It will be interesting to determine, if m^2^G MTases share a relatively recent common ancestor with any of the m^6^A or m^4^C MTase families and if the similarities in their active site are synapomorphic or homoplastic. The issue of divergent or convergent evolution notwithstanding, the \"catalytic loop\", corresponding to motif IV, assumes a strikingly similar conformation in all N-MTase structures solved to date (Figure [2](#F2){ref-type="fig"}). This suggests that the interactions between the target amino group and the side chain of the first residue and the carbonyl oxygen of the second residue of the tetrapeptide are highly similar. The same conformation is retained in the Mj0882 structure, consistent with the prediction that this protein belongs to the N-MTase family. Nevertheless, in the superimposed structures there is a substantial variation in the conformation of the cofactor AdoMet or its analogs, and in their relative orientation with respect to the catalytic site (Figure [2](#F2){ref-type="fig"}). This variation was noted previously in the course of crystallographic analyses, and was attributed to subtle structural differences between cofactor-binding pockets of various MTases that impose distinct conformation of the same ligand \[[@B12],[@B24]\] and to the fact that different ligands (AdoMet, AdoHcy, and their analogs) make different interactions and do not necessarily retain the common conformation when bound to the same MTase \[[@B24],[@B25]\]. ![Comparison of the conformations of the \"catalytic loop\" (motif IV) of N-MTases in stereoview. MTases: *M.Taq*I (1g38 in PDB) \[[@B29]\] in yellow, *M.Rsr*I (leg2) \[[@B55]\] in red, *M.Dpn*M (2dpm) \[[@B56]\] in green; MTase ErmC\' (1qao) \[[@B24]\] in blue; MTase *M.Pvu*II (1boo) \[[@B57]\] in cyan; Mj0882 (1dus) in magenta. The ligand for each protein is shown: AdoMet for *M.Dpn*M and ErmC\', S-adenosylhomocysteine for *M.Pvu*II, 5\'-methylthioadenosine for *M.Rsr*I, and 5\'-\[2-(amino)ethylthio\]-5\'-deoxyadenosine (cofactor analog) and the target deoxyadenosine for *M.Taq*I. In case of *M.Rsr*I, ErmC\' and *M.Pvu*II, whose structures were solved in the absence of the target, the Tyr or Phe sidechain can be easily rotated to the M.TaqI-like orientation to make a face-to-face π-stacking interaction with the target base.](1471-2105-3-10-2){#F2} In order to dock the AdoMet molecule to the 1dus coordinates, we sought to identify those MTase structures, which display greatest similarity to the Mj0882 structure in the cofactor-binding region. VAST and DALI searches confirmed that the region spanning residues 55 to 117 in the Mj0882 sequence is the most similar to MTases that modify hydroxyl groups (data not shown), as well as the entire protein structure (see above). The catechol *O*-MTase structure (COMT) \[[@B26]\](1vid) was reported as the best hit by VAST (Score 7.1, RMSD 1.3). It was used as a structural template, because it was the only well-scored structure of a MTase complexed with AdoMet and not with its non-reactive analogs or the reaction product AdoHcy. Accordingly, the AdoMet moiety was copied from COMT to the Mj0882 structure based on superposition of the 1vid and 1dus coordinates. The obtained Mj0882-AdoMet complex showed no severe atomic overlaps and the cofactor seemed to fit the groove on the protein surface very well. According to AutoDock \[[@B27]\] the energy of the interactions between the 1dus structure and AdoMet in the docked complex is very favorable (-15.03 kcal/mol) even though it is lower from that calculated for the template COMT-AdoMet complex (-22.04 kcal/mol) in the 1vid structure \[[@B26]\]. From the docked model, in striking analogy to most MTase structures (reviewed in ref. \[[@B12]\]), the following three crucial contacts can be predicted: i) D^61^ from motif I coordinates the methionine amino group of AdoMet *via* an ordered water molecule. Even though the corresponding acidic residue is conserved in nearly all MTases analyzed to date, this contact has been identified only recently in the high-resolution structure of the RrmJ MTase \[[@B28]\]. ii) D^84^ from motif II coordinates the ribose hydroxyl groups, iii) D^113^ from motif III coordinates the amino group of the adenine moiety. Non-polar interactions between the side-chains of I^85^ and L^114^ and the adenine ring further contribute to the binding (Figure [3](#F3){ref-type="fig"}). {#F3} Docking of the target guanine to the Mj0882 structure (Figure [4](#F4){ref-type="fig"}) was guided by superposition of the \"motif IV loop\" of Mj0882 and *M.Taq*I, the only N-MTase co-crystallized with the nucleic acid substrate \[[@B29]\]. Under the assumption that both enzymes bind their targets in the same plane, the N2 group of guanine could be aligned with the N6 group of the adenine amino only if the purine ring was rotated by 120 degrees (*i.e.* with the atoms C2 and N3 of guanine superimposed onto atoms C6 and N1 of adenine). According to AutoDock \[[@B27]\] the energy of the interactions between the Mj0882-AdoMet complex and the docked guanine is quite favorable and comparable with that calculated for the target adenine in the *M.Taq*I structure (1g38) (-6.14 kcal/mol and -7.59 kcal/mol, respectively). These values are much lower than the energy of interactions between the protein and the cofactor (see above), because the AdoMet-binding groove is very deep and hydrophobic, while the base-binding site is relatively shallow. An alternative orientation of the target guanine in complex with 1dus could be obtained if its atoms C2 and N1 were superimposed onto atoms C6 and N1 of adenine. However, the latter model resulted in steric clashes between the ribose moiety and I^132^ (data not shown). Even if the flexible ligand docking option of AutoDock was used, which alleviated steric clashes, energies for the guanine in the \"alternative\" position (a cluster of similar conformations) were significantly worse (data not shown). In the first docked model the guanine moiety fits the groove on the protein\'s surface surprisingly well; a residue was identified, whose side chain might coordinate the ribose hydroxyl groups, namely D^41^ (Figure [3](#F3){ref-type="fig"}). Analysis of the multiple sequence alignment (Figure [1](#F1){ref-type="fig"}) revealed that this residue is highly conserved in the Mj0882, RsmD, RsmC, and YbiN families, supporting its functional importance. In the \"alternative\" docked model, D^41^ is too remote from the target base to make direct contacts to the target nucleoside (data not shown). However, a substantial conformational change upon substrate binding cannot be excluded, and direct participation of D^41^ in catalysis as well as other modes of guanine binding can be envisaged. Identification of this conserved carboxylate as a potentially important residue will hopefully prompt site-directed mutagenesis experiments and experimental determination of its role in binding and/or catalysis. {#F4} Although the presented docking model of Mj0882 complexed with AdoMet and guanine should be regarded as preliminary, it provides a useful platform for correlating the structure with sequence conservation in the Mj0882/RsmD/RsmC/YbiN family. The predicted AdoMet- and ribose-binding Asp residues (see above), as well as other invariant and highly conserved surface-exposed residues, map in the vicinity of the docked ligands, strongly supporting the model (Figure [5](#F5){ref-type="fig"}). The N2 group of the docked guanine is hydrogen bonded to the side chain of N^129^ and the carbonyl oxygen of P^130^. This is similar to the contacts made by the N6 group of the target adenine in the co-crystal structure of *M.Taq*I \[[@B29]\], which served as a docking template. As proposed for *M.Taq*I, these interactions could increase electron density of N2, change its hybridization from sp^2^ towards sp^3^, where the lone pair of N6 is no longer conjugated with the aromatic system, and thus contribute to its activation for nucleophilic attack on the activated methyl group of AdoMet. From the Mj0882 structure it is not obvious which residue may be responsible for eliminating the proton from the methylated N2 amine. The only conserved acidic residue in the vicinity of the active site corresponds to D^41^, and its localization is incompatible with the localization of the protons on N2 in the above-mentioned mechanism. If no massive conformational change of the m^2^G MTase occurs during RNA binding and catalysis, we propose that the proton is transferred directly to a bulk solvent molecule. In *M.Taq*I, the target base forms a face-to-face π-stacking interaction with Y^108^ from the NPPY tetrapeptide and a hydrophobic interaction with V^21^ from motif X. In Mj0882 these contacts are reversed (*i.e.* in motif IV Y^108^ is substituted by non-aromatic I and V^21^ is replaced by aromatic F^35^). Interestingly, all members of families RsmC, RsmD, and YbiN possess aromatic residues at both positions (Figure [1](#F1){ref-type="fig"}). It has been hypothesized that the methyltransfer reaction in N-MTases proceeds *via* a positively charged intermediate, which could be stabilized by cation-Ti interactions \[[@B30]\]. In m^2^G MTases, the electron-rich aromatic rings on both sides of the target could contribute to such stabilization. ![Molecular surface of Mj0882 colored by the conservation score. The score have been calculated using SWISS-PDB VIEWER \[[@B54]\] based on superposition of 1dus coordinates with homology models of YbiN and the CTD of RsmC and RsmD (all proteins from *E. coli*). The color scale varies from blue to yellow, with invariant residues in blue and most variable in yellow. The docked AdoMet and guanine moieties are shown in wireframe representation.](1471-2105-3-10-5){#F5} The predicted guanine-binding mode implies that the target nucleotide is \"flipped out\" of the rRNA into the catalytic site, similarly to the mechanism discovered for DNA: m^5^C MTases \[[@B31]\] and extended to DNA N-MTases \[[@B29],[@B32]\]. In bacterial 16S rRNA G966 and G1516 occur in extrahelical regions, which should facilitate their binding in the catalytic pocket of the enzyme, but G1207 is base-paired in the helix 34 stem \[[@B7]\]. The m^2^G1207 MTase RsmC reacts with the 30S ribosomal particles but barely at all with 16S RNA, while the m^2^G966 MTase RsmD requires only the S7 and S19 proteins to methylate its target in the 16S RNA \[[@B10]\]. It has been inferred that both these nucleotides are modified after the association of several ribosomal proteins with the 16S rRNA, which is likely to assume the native secondary structure, but before assembly of the 30S subunit is complete \[[@B10]\]. The positively charged surface of Mj0882 in the area next to the docked guanine suggests charge complementarity to the negatively charged phosphate groups in rRNA (Figure [6](#F6){ref-type="fig"}). It will be interesting to study the base-flipping capabilities of RsmC and RsmD, since they may shed the light on the evolution of presumed mechanistic differences between the nucleic acid MTases that flip their target out of the double helix and those modifying extrahelical bases. {#F6} Analysis of the apparently \"degenerated\" members of the family ---------------------------------------------------------------- Analysis of the multiple sequence alignment (Figure [1](#F1){ref-type="fig"}) suggested that the NTD of RsmC and RsmD lack several conserved residues. However, only in connection with the docking model of Mj0882 it can be concluded, which of these residues may be involved in binding or catalysis. Specifically, all NTD sequences lack the guanine N2-coordinating Asn and the aromatic or aliphatic residue in motif IV. The NTD of all RsmC, with the exception of *T. thermophilus,* lacks the N-terminal extension, which includes the aromatic residue predicted to make the face-to-face π-stacking interactions with the target guanine (F^35^ in Mj0882). This conserved aromatic residue is also missing from the longer N-terminus of RsmD and *T. thermophilus* RsmC. The NTD of RsmD retained the predicted ribose-binding acidic residue at the position corresponding to D^41^ in Mj0882, while in the NTD of RsmC this residue is often substituted, in most cases with Ser or Thr. All NTD sequences, with the possible exception of *C. crescentus* and *T. thermophilus* RsmC, lack the AdoMet-coordinating acidic residue in motif I, corresponding to D^61^ of Mj0882. The NTD of RsmD retained the Asp residue corresponding to D^84^ in Mj0882, however in nearly all RsmC sequences this residue is substituted with various polar amino acids. A region corresponding to motif III was so diverged in the NTD of both RsmC and RsmD that it could not be aligned with confidence and the position corresponding to D^113^ in Mj0882 could not be identified unambiguously. It is noteworthy that the residue corresponding to D^113^ (motif III) is weakly conserved in the MTase superfamily, but the acidic residue corresponding to D^84^ (motif II) is conserved in nearly all MTase families studied to date \[[@B12]\](JMB, unpublished data). In summary, the NTD of RsmC and RsmD lacks all residues, which were predicted to be a part of the active site based on the docking model and which are conserved between Mj0882 and the CTD. The NTD of RsmC also lacks most of the acidic residues predicted to coordinate AdoMet and the ribose moiety of the target nucleoside. Nevertheless, some of these residues are retained in the NTD of RsmD and in the NTD of *T. thermophilus* RsmC. Hence, it is inferred that the NTD of RsmC and RsmD lost the m^2^G MTase activity and are probably also deficient in the ability to bind the cofactor. There are known precedents of proteins with apparent severe defects in conserved MTase motifs. One example includes the yeast Kar4p protein required for expression of karyogamy-specific genes during mating \[[@B33]\]. Using fold recognition and homology modeling it was demonstrated that Kar4p is related to MTases and the \"permuted\" and m^4^C MTases, although it lost its active site and has the AdoMet-binding site blocked by an insertion. It is quite probable that Kar4p exerts its function due to the retained ability to bind to RNA or DNA (JM Bujnicki and coworkers, manuscript submitted). Another example is the eukaryotic heterodimeric MTase, in which one subunit binds AdoMet and presumably carries out the catalysis, while the other is indispensable for tRNA binding \[[@B2]\]. It has been demonstrated that the tRNA-binding subunit is homologous to the \"active\" subunit, but lost the cofactor-binding and catalytic side-chains. We have also identified a single \"active\" MTase gene in Bacteria and Archaea, suggesting that the ancestral prokaryotic form was a homodimer \[[@B34]\]. This prediction has been recently confirmed by the crystal structure of the Rv2118c protein from *Mycobacterium tuberculosis,* which forms a dimer of tight homodimers \[[@B35]\]. Interestingly, according to the EBI Macromolecular Structure Database, the suggested biologically relevant molecule for the Mj0882 structure is a homodimer <>. The Mj0882 homodimer exhibits 5 interchain salt bridges within the complex and shows a loss of 2045.5 Ang\*\*2 of solvent accessible surface area upon complex formation. However, in each subunit the active site is completely blocked by a loop from the other subunit. Nevertheless, it is tempting to speculate that the active form of Mj0882 is a homodimer that undergoes a conformational change to allow the target 16S rRNA to bind and that RsmC and RsmD MTases function as pseudodimers. If this is true, the relationship between Mj0882 and the RsmC/RsmD families would be similar to the relationships between prokaryotic and eukaryotic MTases. We also found that the archaeal members of the YbiN family lack the counterpart of the conserved N^129^ in Mj0882. However, analysis of the multiple sequence alignment revealed that they retained all other predicted binding residues, with the possible exception of the poorly conserved D^113^ (Figure [1](#F1){ref-type="fig"}). This prompted us to build a provisional homology model of the PH0266 protein, which revealed that the \"missing\" Asp sidechain in motif IV is spatially replaced by a His residue from motif X (data not shown). The alignment (Figure [1](#F1){ref-type="fig"}) reveals that the His residue is conserved at this position in all archaeal YbiN homologs, whereas bacterial and eukaryotic YbiN proteins have Pro or Ala and the Mj0882/RsmC/RsmD lineage have Gly or Ala (G^43^ in Mj0882). \"Migration\" of catalytic and cofactor binding residues have been documented in many cases (for a review see ref. \[[@B36]\]), including RNA MTases that generate m^5^C in rRNA \[[@B37]\] and m^7^G in mRNA \[[@B38]\]. Nonetheless, no active nucleic acid MTase have been reported bearing the His substitution for the catalytic residue in motif IV. We hope that this analysis will stimulate experimental characterization of archaeal YbiN homologs, which may eventually lead to validation of a novel version of the methyltransfer active site. Conclusions =========== Based on a comprehensive bioinformatic analysis of m^2^G MTases it was inferred that the prokaryotic RsmC and RsmD MTases are pseudodimers. The C-terminal catalytic domain is closely related to the structurally characterized Mj0882 protein, while the N-terminal domain lacks the cofactor-binding and catalytic side-chains. Based on analysis of sequence similarities within and between subfamilies, it is suggested that bacterial m^2^G MTases evolved from a duplicated Mj0882-like enzyme, although a fusion with another distantly related MTase cannot be definitely excluded. The supposed intragenic duplication or gene fusion followed by degeneration of the N-terminal domain was ancient and predated the gene duplication leading to radiation of RsmC and RsmD lineages. However, we were not able to determine the precise timing of the divergence, presumably due to the high noise-to-signal ratio in the multiple sequence alignment. This evolutionary scenario, along with the analysis of the crystal structure of Mj0882, suggests that m^2^G MTases function as dimers, with the one subunit carrying out catalysis and the other probably participating in recognition and binding of the target rRNA. An analogous case has been described for archaeal tRNA splicing enzymes, which exist in two forms. The *M. jannaschii* EndA nuclease is a homotetramer, in which two pairs of identical subunits have non-equivalent roles in tRNA binding and catalysis \[[@B39]\], whereas its *A. fulgidus* homolog is a homodimer of pseudodimers, comprising diverged domains of which only one retained the active site \[[@B40]\]. Analysis of the Mj0882 docking model in the light of the sequence conservation in the m^2^G MTase family allowed prediction of cofactor-binding and catalytic residues. In the YbiN family of predicted N-MTases (suggested to be MTases, but this specificity is uncertain) new eukaryotic and archaeal members have been identified. The latter exhibited an intriguing correlated loss of a conserved Asn residue in motif IV with acquisition of an atypical His residue in motif X. Both side chains are likely to occupy a similar spatial position, suggesting a novel version of the MTase active site. All these predictions can be verified experimentally. We hope that this analysis will significantly improve the insight into the molecular mechanism of m^2^G methylation, facilitate further structural and functional studies on m^2^G MTases and contribute to better understanding of relationships between them and other nucleic acid MTases. Methods ======= Sequence database searches -------------------------- The position-specific, iterative (PSI-) version of BLAST \[[@B41],[@B42]\] was used to search the non-redundant (nr) version of current sequence database and the publicly available complete and incomplete genome sequences available via the websites of the National Center for Biotechnology Information (Bethesda, USA; <> and the Bioinformatics Laboratory of International Institute of Molecular and Cell Biology (Warsaw, Poland; <>). All searches were initiated with stringent profile inclusion expectation (e) values (\<10^-20^) to avoid \"explosion\" of hits to multiple paralogous MTase families and the cutoff was relaxed in subsequent iterations as the profile become balanced (i.e. included more orthologs of the query, which reduced the probability of assigning a significant score to a non-ortholog that exhibited a fortuitous similarity to the query, but not to the orthologs of the query). The stringent criteria for provisional assignment of orthologous relationship in PSI-BLAST searches were: the alignment spanning more than 80% of both the query and the reported hit, e-value in the 1^st^ iteration \< 10^-30^, sequence identity \> 40%. In all searches initiated with RsmC, RsmD, and YbiN sequences, hits to the catalytic domain of HemK family members were reported before relaxation of the cutoff with e-values ranging from 10^-5^ to 10^-18^ (i.e. subsequently to the presumed members of the above-mentioned families). The choice of the relaxed cutoff was dictated by the assumption that the HemK family of protein MTases should be regarded as an outgroup and its members shall not be included in the profile. The multiple sequence alignment was retrieved from the PSI-BLAST, degapped, and used as a profile, to which all the full-length sequences were realigned using CLUSTALX \[[@B43]\]. Structure prediction -------------------- Protein structure predictions were carried via the MetaServer interface \[[@B44]\]<> using publicly available online services for fold recognition: FFAS \[[@B17]\], 3DPSSM \[[@B45]\], BIOINBGU \[[@B46]\], GenThreader \[[@B47]\], SAM-T99 \[[@B48]\], and FUGUE \[[@B49]\], and secondary structure prediction: JPRED2 \[[@B50]\] and PSI-PRED \[[@B51]\]. The results were processed by the Pcons consensus server \[[@B52]\], which produced a ranking of potentially best target-template alignments and evaluated the likelihood of the models to be correct. Protein structure analysis -------------------------- Comparison of atomic coordinates with the protein structures from the Protein Data Bank \[[@B20]\] were carried out using VAST \[[@B21]\] and DALI \[[@B53]\]. SWISS-PDB VIEWER \[[@B54]\] was used for all protein structure manipulation, calculation of conservation scores and electrostatic potential distribution, and to generate the figures. Provisional homology modeling was carried out using the SWISS-MODEL/PROMOD II server <>\[[@B54]\]. Authors\' contributions ======================= J.M.B. carried out the sequence and structure analysis and drafted the manuscript. L.R. provided the laboratory space and the access to the MetaServer, BIB-VIEW, local version of BLAST and sequence databases. Acknowledgments =============== We would like to thank all the genome sequencing groups that make their preliminary data publicly available, without which this work could not be done. Especially, the use of data generated at the Department of Energy Joint Genomic Institute, the Institute for Genomic Research, and the Stanford Genome Technology Center is gratefully acknowledged. We are indebted to Krzysztof Ginalski for help with AutoDock and to Richard Leach for critical reading of the manuscript. This work was supported by the Polish State Committee for Scientific Research (grants 8T11F01019 and 6P04 B00519)

Background ========== Enterohaemorrhagic *Escherichia coli* (EHEC) is considered to be common causes of haemorrhagic colitis (HC), thrombotic thrombocytopenic purpura, and haemolytic-uraemic syndrome (HUS) in humans \[[@B1],[@B2]\]. In a previous paper, we have demonstrated that EHEC are commonly found in the intestines of livestock, specially young cattle; and infections in humans are, in part, a consequence of consumption of undercooked meat or raw milk \[[@B3]-[@B6]\]. Several outbreaks have occurred mainly in USA, Canada, United Kingdom and Japan during the last decade \[[@B7]-[@B10]\]. Argentina has one of the highest records of HUS (300--400 cases/year; 22/100.000 children under 4 years of age). This rate can be compared with the anual incidence of 2.6/100.000 in Oregon, 3.0 in King County-Washington, 3 to 5 in Canada, United Kingdom, Chile and Uruguay \[[@B11]\]. In spite of EHEC strain O157:H7 has been reported as a cause of HUS in this country \[[@B12]\], this strain does not seem to be so common in this country as in the USA \[[@B11]\]. We have recently published an extensive list of non-O157:H7 EHEC serotypes isolated from cattle and foods in Argentina \[[@B13]\], several of them involved as cause of HC and HUS in somewhere. In Europe, eighty percent of the VTEC isolated from patients with diarrhoea corresponded to non-O157 serogroups such as O26, O91, O103, O111, O113, O128, O145 \[[@B14]\]. In 1984, Kudoh et al. \[[@B15]\] described the first outbreak of disease due to *E. coli* O145:H- in Japan, affecting to 100 children. Beutin et al. \[[@B16]\] have considered this serotype as a new emergen. Afterthere, several investigators notified the isolation of this serotype from cattle and humans in USA, Canada and Belgium \[[@B17]-[@B22]\]. Another serotype, , was found associated to HUS cases in U.K. \[[@B19]\], meanwhile O145:H8 and were isolated from healthy cattle in Canada \[[@B22]\] and Germany \[[@B23]\], respectively. Aside from known differences among the pathogenicity of VT1, VT2 and its variants, factors that may affect the virulence of EHEC are the ability to cause attaching and effacing lesions (*eae* gene) in the intestinal mucosa and possession of a 60 MDa megaplasmid (Mp) \[[@B1]\]. The aim of this work is to communicate the first isolation of O145:H- from cattle in this country and characterize the virulence cassette of several strains, providing useful information to evaluate the risk of foodborne transmission of this non-O157:H7 serotype. Results and discussion ====================== Nine Shiga toxin-producing *Escherichia coli* serogroups and eleven non-typable strains were identified from a feedlot in Argentina (data no shown). One of them (O145) corresponded to a serogroup never described before in this country. All O145 strains were characterized as H-. This non-O157:H7 serotype (O145:H-) had been previously involved in haemolytic uraemic syndrome outbreaks in Japan \[[@B15]\]. The genotype of those strains were characterized, corresponding to several virulence cassettes: VT2^+^*eaeA*^+^ Mp^+^ (n = 5), VT2^+^*eaeA*^+^ (n = 1), VT1^+^*eaeA*^+^ Mp^+^ (n = 2), and VT1^+^*eaeA*^+^ (n = 1). All these strains were able to ferment sorbitol within 18 h of incubation at 37°C. Strains isolated from the same animal were considered only when showed a different virulence pattern. O145:H- strains were distributed in two RAPD profiles, which corresponded to the presence of either, VT1^+^ or VT2^+^ genotype (RAPD profiles 2 and 1, respectively). No difference was detected by RAPD analysis between Mp^+^ or Mp^-^ strains (Fig. [1](#F1){ref-type="fig"}). This fact suggests that would be necessary to carry on RAPD studies by using new primers or to apply pulse field electrophoresis technique in the study of clonal relationships. {#F1} The presence of the virulence genes codifying for verotoxin (1 or 2), the ability to develop attaching and effacing phenomenon on enterocyte surface (*eaeA* gene) and the expression of an EHEC haemolysin (Hly~EHEC~), afford the potential of O145:H- as foodborne emergent human pathogen. This EHEC non-O157:H7 serotype, isolated from cattle in feedlot, was never found by us from grazing cattle in Argentina \[[@B13]\]. Our finding, and the high incidence rate of HUS in this country, put local health service authorities on guard. Conclusions =========== This was the first isolation of EHEC O145:H- serotype from Argentinian cattle which enlarges the list of non-O157:H7 serotypes isolated from bovine and ground beef in this country by us. All O145:H- strains carried several virulence factors which allow to predict their potential ability to develop haemolytic uraemic syndrome in humans. The strains, which had been isolated from different animal in the same feedlot, conformed two RAPD profiles based upon their verotoxin subtype, without clonal differences. Materials and methods ===================== Sample collection from cattle ----------------------------- Samples were collected from an Argentinian farm. following the method described previously \[[@B6]\]. Briefly, 59 beef cattle, belonging to a feedlot of Pampeana region, were sampled fortnightly during six months by a single rectal swab from each animal. The swabs were placed in transport medium and processed immediately at the laboratory. Bacterial colonies were grown on MacConkey agar incubating for 24 hs at 37°C. An aliquots of confluent growth, aproximately 100 colonies, was inoculated into Luria-Bertani broth for 3 h at 37°C and processed for DNA extraction \[[@B6],[@B24]\]. One ml from each culture was frozen at -70°C with the addition of glycerol for further isolation of individual VT^+^ colonies (50 to 100 colonies/sample) on a MacConkey agar plate. Microorganisms were confirmed as *E. coli* by means of biochemical test such as citrate, indole, urease and TSI profile. Detection of VT1, VT2, *eae* and 60 MDa plasmid by PCR ------------------------------------------------------ *Escherichia coli* growth and DNA preparation were performed as described previously \[[@B6],[@B24]\]. Primers *eae-1* and *eae-2* were designed to amplify a portion of the *eae* gene which is conserved between enteropathogenic and enterohaemorrhagic *E. coli* strains. Primer sequences for VT1, VT2 and *eae* were indicated in a previous paper \[[@B6]\]. Mp primer sequences were: MFS1F 5\' ACGATGTGGTTTATTCTGGA 3\' and MFS1R 5\' CTTCACGTCACCATACATAT 3\' (166 bp amplimer) \[[@B25]\]. T~m~ were calculated for each primer using Oligo 4.0 (Primer Analysis Software, National BioSciences). Amplification of bacterial DNA was performed in a total volume of 50 μl. A negative control (reagent blank) was included without addition of sample. Another control was designed by adding DNA from the strain lacking VT1, VT2, *eae* and Mp. Two reference strains O157:H7 were used as positive controls. The conditions for the PCR amplification of VT1, VT2, *eae* and Mp were as described previously \[[@B6],[@B24],[@B25]\]. Amplified products were analyzed by submarine gel electrophoresis and UV-transillumination (300 nm). Fermentation of sorbitol ------------------------ Sorbitol MacConkey agar was used to test the VT^+^ strains for this condition. Serotyping ---------- O and H antigens were determined by means of a microagglutination technique in tubes and plates described by Guinée *et al.*\[[@B26]\] and modified by Blanco *et. al.*\[[@B27]\] using all available O (O1 -- O175) plus six putative new O antigens (OX176-through OX181) \[[@B28]\] and H (H1 -- H56) antisera \[[@B29]\]. Non-specific agglutinins were removed by absorption with the corresponding cross-reacting antigens. All VTEC were processed for O serogroup determination. H serotyping was performed only on those strains which, having been isolated from the same sample, differed in either one virulence factor or the O serogroup. Random amplified polymorphic DNA (RAPD) analysis ------------------------------------------------ Bacteria were grown overnight at 37°C in LB broth with shaking. An aliquot of the culture was diluted 1/10 in water to determine the optical density at 600 nm \[39\]. For an optical density value of 0.5, a 500 μl aliquot of the stationary-phase culture was centrifuged (2 min at 12,000 × g) and suspended in 500 μl of bidestilled water. The suspension was then boiled for 10 min, centrifuged (2 min at 12,000 × g) and the supernatant was stored at -70°C. Five microliters were used as the template for PCR amplification. RAPD was performed in a final volume of 50 μl containing 20 mM (NH~4~)~2~SO~4~, 75 mM Tris-HCl pH 9.0, 0.1% (w/v) Tween 20, 2.5 mM MgCl~2~, 200 μM each dNTP, 1 μM primer, 2.5 U *Taq* DNA polymerase. Three different primers were evaluated: 1281 \[[@B31]\], 970--11 and M13 \[[@B32]\], but primers 1281 and 970--11 were not discriminatory. Amplification was done in a Genius thermal cycler \[Techne (Cambridge) Ltd.\] as follows: initial denaturation at 94°C for 5 min (heating rate 60°C/min), followed by 40 cycles of denaturation at 94°C for 1 min (heating rate 29°C/min), annealing at 50°C for 1 ½ min (cooling rate 26°C/min) and extension at 72°C for 1 ½ min (heating rate 29°C/min). Reaction products (10 μl) were analysed in a 1.8% agarose gel stained with ethidium bromide. ::: {#T1 .table-wrap} ::: {.caption} ###### Feno-genotypic characteristics of O145:H- strains isolated from cattle. ::: **Serotype** **Animal & Sample** **Virulence cassette** **Virulence profile** **Lane** -------------- --------------------- ------------------------ ----------------------- ---------- ---- --- --- O145:H- 52(1) \- \+ \+ \+ 1 1 O145:H- 18(1) \- \+ \+ \+ 1 2 O145:H- 27(3) \- \+ \+ \+ 1 3 O145:H- 46(3) \- \+ \+ \+ 1 4 O145:H- 38(4) \- \+ \+ \+ 1 5 O145:H- 27(3) \- \+ \+ \- 2 6 O145:H- 1(1) \+ \- \+ \+ 3 7 O145:H- 17(11) \+ \- \+ \+ 3 8 O145:H- 1(1) \+ \- \+ \- 4 9 **n** 3 6 9 7 **Lane** column indicates positions of RAPD study for each strain in Fig. [1](#F1){ref-type="fig"}. ::: Acknowledgements ================ Authors thank Nutrition and Metabolism Unit (INTA-Balcarce) for its collaboration and María R. Ortiz for her technical assistance. This work was supported by the Scientific Research Commission Prov. Buenos Aires (CIC), SECYT-UNCPBA, FONCYT and Univ. Santiago de Compostela. A.E. Parma and A. I. Etcheverría are members of CIC and G.H. Arroyo is member of CONICET.

Background ========== Since the initial description of four patients with coeliac disease (CD) and primary biliary cirrhosis (PBC) \[[@B1]\], there have been several subsequent reports demonstrating an association of these two diseases \[[@B2]-[@B7]\]. Recently, a number of epidemiological studies reported an increased, up to 7%, prevalence of CD in patients with PBC and an increased prevalence, up to 3%, of PBC in patients with CD \[[@B8],[@B9]\]. Screening for one disease when the other is identified has therefore been suggested. However, in other studies this association was found less pronounced or has not been confirmed \[[@B10]-[@B13]\]. Autoimmune cholangitis (AIC) was first reported as an overlap syndrome of autoimmune hepatitis and PBC, but is considered by most as a discrete entity of antimitochondrial antibody (AMA)-negative PBC with autoimmune characters \[[@B14]\]. Apart from a case report \[[@B15]\], no relation of adult CD to autoimmune cholangitis has so far been described. However, there have been no formal attempts to systematically evaluate patients with AIC for CD, presumably because AIC represents a rare form of chronic cholestatic liver disease. Although a definite diagnosis of CD still requires histopathologic examination of the small intestinal mucosa, the screening methods for large number of individuals are based on non-invasive serological tests, namely, anti-gliadin (AGA, IgG and IgA class), anti-reticulin (ARA R1-type, IgA class) and anti-endomysial (EMA, IgA class) antibodies \[[@B16]-[@B18]\]. The combined use of serum IgG-AGA (good sensitivity) and IgA-EMA (good specificity) has resulted in a reliable screening test for diagnosis of CD \[[@B19]\]. However, anti-gliadin antibodies (AGAs) are not suitable for CD screening in patients with chronic liver diseases including PBC, as an unacceptably high percentage of false-positives was reported in several studies \[[@B11],[@B20],[@B21]\]. In this respect, a previous study revealed that the serum IgA-EMA assay is the optimum test for predicting CD in patients with chronic liver disease \[[@B11]\]. Although IgA class antibodies to guinea pig liver-derived tissue transglutaminase (IgA-tTG) have been shown to be highly specific and sensitive for the diagnosis of CD \[[@B22]-[@B26]\], it was recently reported that in patients with chronic liver disease there was a high frequency of false-positives which disappeared when human tTG was used as the antigen in the ELISA system \[[@B27]\]. In the present study we evaluated AGAs, IgA-ARA, IgA-EMA and guinea pig liver-derived IgA-tTG in 62 patients with PBC and 17 patients with AIC. Intestinal biopsies have been subsequently performed in the majority of patients tested positive for at least one antibody class. Methods ======= Materials --------- Serum samples from 62 patients with PBC (53 women and nine men; mean age, 59 years; range, 32--85 years) and 17 patients with AIC (16 women and one man; mean age, 62 years; range, 52--77 years) were used in the study. The samples were collected prospectively over a 2 yr. period (1999--2000) and stored in aliquots at -20°C. Diagnosis of PBC was based on standard laboratory findings, compatible liver histology and the presence of AMA in titer \> 1:80. Class 2 AMA antibodies (M2) detected by ELISA were positive in all patients. According to Scheuer\'s classification \[[@B28]\], 34 patients (54%) had stage I-II and 28 (46%) stage III-IV disease. The mean follow-up period from the time of diagnosis was 5.2 years (range, 1--9.4 years). The diagnosis of AIC has been made on cholestatic patients with a liver biopsy compatible with PBC, without AMA or M2 positivity. ANA were positive (titers ≥ 1:160) in all 17 patients. Anti-smooth muscle antibodies were positive (titer ≥ 1:80) in 9 patients. According to Scheuer\'s classification, 10 patients (58%) had stage I-II and 7 (42%) stage III-IV disease. The mean follow-up after diagnosis was 6.6 years (range, 1--7.8 years). Ultrasound, computed tomography and endoscopic retrograde cholangiography ruled out the possibility of biliary obstruction, when the latter was suspected. No patient had a family history of CD, or IgA deficiency. One PBC patient was HbsAg positive and one PBC patient and two AIC patients were anti-HCV positive. One hundred Cretan adults selected at random from blood donors served as controls. Eighteen serum samples from symptomatic, biopsy-proven, AMA negative adults with CD were included in the study as methodological controls. Total IgA was measured in all CD patients and healthy controls to exclude IgA deficiency. Methods ------- Two operators unaware of the clinical parameters performed the immunological tests. All PBC, AIC, CD and control sera were screened for AGAs, IgA-ARA, IgA-EMA and IgA-tTG. IgA- and IgG-AGA were screened by a commercial ELISA at a dilution of 1:50 (Alphadia SA/NV, Belgium). Values greater than 50 U/mL were considered positive. Results were considered doubtful and repeated for values between 45 and 50 U/mL. For IgA-ARA R1-type, sera were tested by an indirect immunofluorescence (IIF) test at a 1:10 and 1:40 dilution, using rat liver, kidney and stomach as substrates (Sanofi Diagnostics Pasteur, France). Sera were screened for IgA-EMA by IIF, using both monkey oesophagus (Biosystems, Barcelona, Spain) and human umbilical cord (Eurospital SpA, Trieste, Italy) as substrates. Titers of ≥ 1:5 were considered positive. To avoid false-negative results, IgA-EMA were also evaluated at a 1:50 dilution. IgA-tTG were tested by a commercially available guinea-pig liver-derived tTG ELISA kit at a dilution of 1:101, according to manufacturers\' instructions (QUANTA Lite™ tTG ELISA, INOVA Diagnostics, San Diego, USA). The samples were classified as negative (\<20 Units), weakly positive (20--30 Units) and strongly positive (\>30 Units). Small intestinal biopsies were performed at endoscopy using endoscopic forceps. Four biopsy specimens from opposite sites of the distal duodenum were obtained and evaluated by one observer specialized in gastrointestinal tract pathology with regard to the villi/crypts ratio and the number of intraepithelial lymphocytes (IELs) per 100 surface epithelial cells. Small bowel biopsies from the patients with PBC and AIC were evaluated blindly with those obtained from the 18 patients with untreated CD. When biopsy specimen revealed features of CD \[[@B29]\], namely, a lowering of the villous height to crypt depth ratio (normal, 3--5:1) and/or an increase in IELs (normal, 10--30 per 100 epithelial cells) a more detailed histological classification was applied \[[@B30]\]. Test results were compared using a 2 × 2 contingency table and chi-squared statistics. When appropriate, Fisher\'s exact test was used. *P* values below 0.05 were considered significant. Results ======= IgA class anti-gliadin antibodies and IgA class antibodies to tissue transglutaminase were detected with a significantly increased frequency in patients with both PBC and AIC as compared to healthy blood donors (Table [1](#T1){ref-type="table"}). In contrast, none of the PBC and AIC patients was tested positive for anti-reticulin or anti-endomysial antibodies. Controls were tested negative for anti-reticulin, anti-endomysial and IgA class anti-tissue transglutaminase antibodies. ::: {#T1 .table-wrap} ::: {.caption} ###### Prevalence (%) of antibodies in patients and in controls ::: **PBC** n = 62 (%) **AIC** n = 17(%) **BD** n = 100 **CD** n = 18 ---------------------- -------------------- ------------------- ---------------- --------------- **IgA-AGA** 12(19)^\*^ 6 (35)^\*^ 1 18 **IgG-AGA** 3(5) \- 2 16 **Any AGA** 13 (21)^\*^ 6 (35)^\*^ 3 18 **Both AGAs** 2(3) \- 1 16 **IgA-ARA** \- \- \- 18 **IgA-EMA** \- \- \- 18 **IgA-tTG** 6(10)^\*^ 3(18)^\*^ \- 18 **IgA-tTG and AGAs** 2(3) 2 (12)^\*^ \- 18 n = number of patients; \* = where *p* less than 0.05, PBC = primary biliary cirrhosis; AIC = autoimmune cholangitis; CD = coeliac disease; BD = blood donors; IgA-AGA = anti-gliadin antibodies, IgA class; IgG-AGA = anti-gliadin antibodies, IgG class; AGAs = anti-gliadin antibodies; IgA-ARA = IgA class anti-reticulin antibodies, RI type; IgA-EMA = anti-endomysial antibodies, IgA class; IgA-tTG = antibodies to tissue transglutaminase, IgA class ::: PBC patients (Table [2](#T2){ref-type="table"}) ----------------------------------------------- ::: {#T2 .table-wrap} ::: {.caption} ###### Antibody titers and intestinal biopsies in patients with primary biliary cirrhosis ::: **Patient** **Disease stage^a^** **IgA-AGA** **IgG-AGA** **IgA-ARA** **IgA-EMA** **IgA-tTG** **Intestinal biopsy** ------------- ---------------------- ------------- ------------- ------------- ------------- ------------- ----------------------- 1\. I-II \- \- \- \- 49 Normal 2\. I-II 199 58,6 \- \- 50 Normal 3\. I-II \- \- \- \- 25 Normal 4\. I-II 55 \- \- \- \- ND 5\. I-II 166 \- \- \- \- Normal 6\. I-II 200 200 \- \- \- Normal 7\. I-II 70 \- \- \- \- ND 8\. III-IV 244 \- \- \- \- ND 9\. III-IV 144 \- \- \- \- ND 10\. III-IV \- 67,4 \- \- \- ND 11\. III-IV \- \- \- \- 21 ND 12\. III-IV 106 \- \- \- 32 Normal 13\. III-IV \- \- \- \- 28 Normal 14\. III-IV 97 \- \- \- \- Normal 15\. III-IV 53 \- \- \- \- Normal 16\. III-IV 57 \- \- \- \- Normal 17\. III-IV 253 \- \- \- \- ND ^a^According to Scheuer\'s classification; ND = not done ::: Both anti-gliadin antibodies were detected in 13 out of 62 patients with PBC (21% *vs.* 3% in controls, *p* \< 0.001). IgA-AGA were tested positive in 12 patients (19%, *p* \< 0.001), of whom 5 had stage I-II disease, and 7 had stage II-IV disease. IgA-AGA positivity was significantly more pronounced in patients with stage III-IV disease (25% vs 14%, *p* \< 0.05). IgA-tTG were detected in 6 patients (10%), 3 of whom had stage I-II disease and another 3 stage III-IV disease, while tested negative in all controls (*p* \< 0.001). Serology was negative in the two patients with a superimposed viral hepatitis infection. AIC patients (Table [3](#T3){ref-type="table"}) ----------------------------------------------- ::: {#T3 .table-wrap} ::: {.caption} ###### Antibody titers and intestinal biopsies in patients with autoimmune cholangitis ::: **Patient** **Disease stage^a^** **IgA-AGA** **IgG-AGA** **IgA-ARA** **IgA-EMA** **IgA-tTG** **Intestinal biopsy** ------------- ---------------------- ------------- ------------- ------------- ------------- ------------- ----------------------- 1\. I-II 157 \- \- \- 21,9 Normal 2\. I-II \- \- \- \- 26 Normal 3\. I-II 153 \- \- \- \- ND 4\. III-IV 55,9 \- \- \- \- ND 5\. III-IV 382 \- \- \- 32 Normal 6\. III-IV 65 \- \- \- \- Normal 7\. III-IV 60 \- \- \- \- Normal ^a^According to Scheuer\'s classification; ND = not done ::: IgA-AGA were tested positive in 6 patients with AIC (35% *vs.* 3% in controls, *p* \< 0.001), two with stage I-II disease and four with stage III-IV disease. IgA-AGA positivity was significantly more pronounced in patients with stage III-IV disease (57% vs 20%, *p* \< 0.05). IgA-tTG were detected in 3 patients with AIC (18%), two patients with stage I-II disease and one patient with stage III-IV disease, while tested negative in all controls (*p* \< 0.001). Serology was negative in two AIC patients with superimposed hepatitis C. CD patients ----------- All 18 patients with coeliac disease were tested positive for IgA-ARA, IgA-EMA and IgA-tTG, whereas 16 of them had also both IgA and IgG anti-gliadin antibodies. Healthy blood donors -------------------- None of the 100 tested sera were found to be IgA-EMA or IgA-tTG positive. However AGAs were present in 3 individuals (IgA-AGA in one and IgG-AGA in two). Duodenal biopsies ----------------- Overall, biopsies were performed in 10 out of 17 PBC patients (59%) and 5 out of 7 AIC patients (71%) tested IgA-AGA and/or IgG-AGA and/or IgA-tTG positive. Five patients died shortly after testing positive and four patients refused any further investigation. None of the biopsied PBC and AIC patients had features suggestive of CD. Small bowel biopsies of our 18 patients with CD revealed partial villous atrophy (Marsh IIIa) in three, subtotal villous atrophy (Marsh IIIb) in eleven and total villous atrophy (Marsh IIIc) in four. Discussion ========== In contrast to recent reports from Wales \[[@B8]\], Scandinavia \[[@B9]\], Northern Ireland \[[@B31]\], and Canada \[[@B32]\], the present study has not been able to demonstrate an increased risk of coeliac disease in this group of 62 Cretan patients with primary biliary cirrhosis. Our results are consistent with data from Sweden \[[@B11]\] and Italy \[[@B12],[@B13]\], suggesting that if there is no clinical suspicion of CD, screening with anti-endomysial antibodies should not be performed routinely in all patients with PBC. A possible explanation relevant to these controversies could be the methodological differences in detecting and defining CD, like the different serological tests used for screening (AGAs and/or IgA-EMA), the degree of intestinal damage, or the spectrum of clinical presentation of CD in patients studied. Patients with malabsorption, diarrhoea, weight loss, and total villous atrophy represent perhaps only 30--40% of the entire spectrum of gluten-sensitised individuals \[[@B19]\]. However, studies supporting the association between PBC and CD have included mainly symptomatic or complicated cases. In Kingham\'s et al study \[[@B8]\], CD in patients with PBC was sought by investigating features suggestive of malabsorption, a family history of CD, the presence of AGAs, or the finding of otherwise unexplained haematinic deficiency. In Sorensen\'s et al paper \[[@B9]\], it is not clear whether all patients with CD were systematically evaluated for PBC. PBC was found in patients hospitalised for CD, which presumably excluded patients with subclinical or silent disease. Furthermore, no information was available for patients treated as outpatients. The inconsistencies between studies may also be attributed to the ethnic variability of CD prevalence in the different populations. The prevalence of CD in Greece is currently unknown, because of the lack of official epidemiological data, but it is probably low. A crude estimation concerning of the prevalence of CD in Greece is provided from an endoscopic study from Athens, reporting one new case of unsuspected CD per 520 endoscopies \[[@B33]\]. When we retrospectively looked for, only 23 adult patients with symptomatic gluten-sensitive enteropathy were found over a 9 year period in a specialized University Unit, which is the main reference center for the total population of the island of Crete (approximately 500 000 people). We do recognize the fact that the size of the studied samples is not large enough to permit us to refute the presence of an association between CD and PBC safely. However, our inability to confirm this association is in agreement with recent data from large Italian series \[[@B13]\], consisted of 336 patients with CD and 65 cases with PBC. In this study, which can be assumed to be representative of the Mediterranean population, only one case of PBC among patients with CD was identified and no case of CD in the group of PBC. Patients with autoimmune cholangitis were not included in the studies aiming to estimate the risk of CD in patients with PBC. In our small series, no case of CD was found among the 17 patients with autoimmune cholangitis and thus an association could not be established. A possible limitation of our study may be the relatively low percentage of duodenal biopsies performed in patients with primary biliary cirrhosis tested positive for at least one antibody class. However, this was also the case in studies supporting the association. Our patients were screened for IgA-EMA by indirect immunofluorescence (IIF) test, using both monkey esophagus and human umbilical cord as substrates which have been shown to be highly sensitive and specific indicators of untreated coeliac disease \[[@B16],[@B17]\]. Thus missing a case of CD in our patients with PBC seems rather unlikely. It was previously reported that the development of IgA-EMA depends on the severity of intestinal histopathology of CD and a subgroup of asymptomatic coeliacs negative for AGAs or IgA-EMA will be under-diagnosed, especially those with minor tissue damage \[[@B30]\]. However, none of the studies supporting or failing to demonstrate the association have focused on this aspect. In our series, 21% of PBC and 35% of AIC patients were tested positive for AGAs. Furthermore, IgA-AGA positivity was significantly more pronounced in patients with Scheuer\'s stage III-IV disease. Floreani et al \[[@B20]\] reported that IgA-AGA were found in 3.4%, whereas both IgA and IgG-AGA in 6.8% of PBC patients. Likewise, Sjoberg et al \[[@B11]\] reported that the occurrence of false-positive AGAs in PBC patients was as high as 16%. We therefore believe, as previously suggested \[[@B11],[@B20],[@B21]\], that the presence of AGAs in PBC and AIC seems to be secondary to liver damage per se and represent a marker of non-specific immune reactivity. Finally, our results confirm the previously reported low specificity of guinea pig liver-derived tTG ELISA in patients with chronic liver disease \[[@B27]\]. We were unable to compare our results with those obtained by a human recombinant tTG ELISA as the latter is not currently commercially available in our country. Conclusions =========== We confirmed the previously reported high prevalence of false-positive anti-gliadin and guinea pig liver-derived anti-tissue transglutaminase antibodies in patients with chronic liver disease. We were unable to confirm observations suggesting an over-representation of coeliac disease in patients with primary biliary cirrhosis. Our findings support the idea that in areas of low prevalence of CD, in the absence of clinical suspicion, screening all patients with PBC for coeliac disease should not be performed routinely. Competing interests =================== none declared Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== Quality of life is an important clinical outcome in assessing the efficacy of health care. It is a concept that includes many subjective elements: physical, emotional and social function, attitudes to illness, patients\' daily lives-including family interactions, Spitzer, Dobson, et al. \[[@B1]\], Cella and Tulsky \[[@B2]\]. Treatments that are designed to control symptoms are suited for Quality of Life end points. When the goal of treatment is symptom palliation rather than prolongation, the use of Quality of Life as a primary end point of survival can be considered, Winer \[[@B3]\]. As a result of the increased interest in the impact of disease and treatment on the patient\'s life and functioning, several well-validated questionnaires that measure health-related quality of life (HRQoL) have been developed, some of which formed the foundation for the questionnaire that was used in this study \[[@B4]\]\[[@B5]\]\[[@B6]\]\[[@B7]\]. Cancer is not a disease of an individual but impacts the family system. Families must confront, and attempt to realize the meaning of cancer for the patient, for each family member individually and for the family system as a whole \[[@B8]\]. The diagnosis of cancer confronts both the patient and his/her caregiver(s) with a major life crisis, therefore family therapists now refer to the \'cancer family\' and not to the \'cancer personality\' \[[@B9]\]. Furthermore, as a life-threatening disease progresses, family members may assume new roles and responsibilities. They are usually supportive or concerned for their terminally ill patients, but they are frequently opposed to the patient\'s wishes or the recommendations of the treatment team. When the patient is incompetent, conflicts of interest may arise since decision-making is a great responsibility for the family \[[@B10]\]. A relative, who feels guilty or denies the hopelessness of the situation, may press the patient for cure rather than care, or to prolong or discontinue treatment \[[@B11]\]. It is worth noting, that every patient and family has a culture. Cultural aspects of values and behaviors are the key variables, along with life experiences, socio-economic status, and personality differences, that affect the meaning of cancer for both individuals and their families, as well as how they cope with the disease \[[@B10]\]\[[@B12]\]\[[@B13]\]. Thus, it is important to mention the close bonds found in a Greek family, especially towards severe problems such as cancer \[[@B14]\]. Furthermore, when relatives know the truth and try to keep it from their patients, they are likely to transmit their fears and anxiety through non-verbal ways. The cost of deception is high. For the patient who suspects but still hopes, the state of uncertainty is harder to bear than the certainty of the knowledge \[[@B15]\]\[[@B16]\]. This attitude of non-disclosure is not only associated with Greek culture but has also been observed in Japan \[[@B17]\]\[[@B18]\]\[[@B19]\]. Denying acceptance of the disease, family members even avoid discussing about cancer \[[@B20]\]\[[@B21]\]. The objectives of this study were to investigate the attitudes, perceptions and patterns of choice in the management of terminal stage cancer patients with respect to their families through the development of a relatives\' patient management questionnaire. Typically HRQoL questionnaires assessing the patients\' perceptions and choices are used to determine choice of treatment and palliation, however, with the realization that cancer families are also affected and involved by the disease and its management their perceptions of management choices are becoming important considerations in oncology. Methods ======= The study took place in the Pain Relief and Palliative Care Unit, of Areteion Hospital, in University of Athens. The study was carried out between March and September 2000, with hospital ethics committee approval. During this time 400 relatives of advanced cancer patients visited the Unit, from all over Greece. The questionnaire was randomly distributed to 146 Greek relatives who approached the Unit for pain relief and the control of other cancer related symptoms. It was drawn using stratified random sampling, based upon the level of first-degree relatedness, i.e. spouse, child, sibling, parent. Inclusion criteria were a) first-degree relative, b) mentally capable, c) informed consent. The relatives were asked to complete the questionnaire in the outpatient unit, and were given brief instructions on how to complete the questionnaire (particularly Question 3). The initial measurement consisted of 25-item questions, composed of 6 multi-item scales, and 7 single item scales. Twelve of the scales were presented into 3 optional statements to be scored \"yes\", \"sometimes\", \"no\", and the relatives were asked to mark the appropriate answer accordingly. Question number 3 consisted of 7 factors and relatives had to mark their choice in a hierarchical order. The questionnaire was designed to be self-assessment and was formulated after a thorough review of the relevant literature, incorporating previously validated questions from other scales and measures appropriate for use in palliative care settings for the assessment of relatives\' patient management perceptions and choices \[[@B4]\]\[[@B5]\]\[[@B6]\]\[[@B7]\]. The questionnaire contained 4 sections with different aspects of consultation: information disclosure (Qs, 1, 2, 4, 8); therapy choices (Q 3); support-communication and care (Qs, 5, 6, 7, 9 and 13) and hospitalization (Qs, 10, 11, 12). Patients\' medical records provided information about demographics and disease characteristics, specifically: type of cancer, presence of metastasis, and Karnofsky performance. Similar demographic data was obtained from the relatives: age and sex, with stratification based upon degree of relatedness, spouse, child, parent, sibling; educational status, primary, high school and university; residence, rural, village and town. Multivariate analysis was conducted on these stratifications. However, no analysis was conducted for the stratification residence, as only 10 relatives did not reside in the town. Relatives were also timed during the completion of the questionnaire. Quantitative variables are described by the mean ± standard deviation. Quantitative aspects were analyzed using *t*-test, one-way and two-way analysis of variance (ANOVA) and categorical variables by Pearson\'s chi-square, where appropriate. The reliability and validity was assessed by average inter-item and average item totals and Cronbach\'s alpha. Statistical significance was accepted with p ≤ 0.05. Results ======= The patients\' survival time was between 4 months to 5 years (median: 2 years) while they were off anti-cancer treatment and under palliative care for symptom control. Socio-demographic and clinical data for patients and relatives are presented in Table [1](#T1){ref-type="table"}. In all the questions, except for question number 3, the rate of answers given was 100%, the average time required to complete the questionnaire was 8--10 minutes. Regarding question number 3, 15% of responders did not complete this question. ::: {#T1 .table-wrap} ::: {.caption} ###### Socio-demographic characteristics ::: **Relatives\' Age (mean, ± s.d)** **Range** ---------------------- ----------------------------------- ----------- ------- **N** **%** **Sex** Male 39 27 Female 107 73 **Education** Primary 22 15 High School 73 50 University 51 35 **Patients\' Sex** Male 88 60 Female 58 40 **Patients\' Age (mean, ± s.d)** **Range** 67 ± 12 30--98 **N** **%** **Cancer Locations** Lung 38 27 Pancreatic 18 12 Rectum 15 10 Bladder 14 10 Other 61 42 ::: Descriptive statistics for the questionnaire are presented in Table [2](#T2){ref-type="table"}, for all questions except question 3. Some points of interest that derive from Table [2](#T2){ref-type="table"} include the following. First, a small percentage (23%) of relatives believe that diagnosis and prognosis should be revealed to the patients, while the majority (56% and 51% respectively) of the participants believe that only \"sometimes\" they would agree to do so. However, they are of the opinion that the patients should be given information on the possible treatment choices (71%) and the complications or side effects of the treatment (53%). A large percentage (81.5%) would like a health care team to be involved in caring for their patients, and 76% would like physicians from various specialties to co-operate, when this is judged necessary, and 78% of relatives would like the assistance of a social worker. However, only 44% would choose psychological support from specialists, even though 34% answered that they \"sometimes\" prefer the cooperation with psychologists. Several family members (67%), indicated that they would prefer to inform their patients of the palliative nature of the treatment, but only 44% would discuss with them their fears on death, with 28% \"sometimes\" choosing to discuss about it and 28% not. Only 30% chose the intensive care unit for a patient in the terminal stage, while 57% preferred the patients to stay at home with the provision of the appropriate care. It is worth mentioning that close communication and support are variables, which are now starting to be considered seriously from the vast majority of the relatives (96%). ::: {#T2 .table-wrap} ::: {.caption} ###### Relatives\' attitudes towards their patients\' quality of life ::: **Questionnaire items** **Percentage (%) of Relatives\'** --------------------------------------------------------------------------------------------------------------------------------------------------------------- ----------------------------------- --------- ------------------- --------- ------------ --------- **Responding \'yes\'** ***n*** **\'sometimes\'** ***n*** **\'no\'** ***n*** 1\) Do you believe that patient should fully be informed a) for the diagnosis? 23% 34 56% 82 20.5% 30 b) for the prognosis of the disease? 23% 34 51% 74 26% 38 2\) Do you agree the patient to be informed a) for the possible treatment choices? 71% 104 24% 35 5% 7 b) for the percentage effectiveness? 51% 74 37% 54 12% 18 c) for the complications or the side-effects of the treatment? 53% 77 33% 48 15% 21 4\) Do you believe that patient should be informed before amputative/disfiguring operations? 68.5% 100 21% 31 10% 15 5\) Would you like for the caring and treatment of cancer patients to be involved a) the health care team? 81.5% 119 16% 24 2.% 3 b) physicians from various specialties Cooperating as necessary? 76% 111 12% 17 12% 18 6\) Would you like to cooperate with psychological support specialists? 44% 64 34% 50 22% 32 7\) Would you like to cooperate with a social worker? 78% 114 19% 28 3% 4 8\) In the final stage of the patient would you prefer to inform him/her for the palliative nature of the treatment? 67% 98 24% 35 9% 13 9\) Would you like to discuss with the patient his/her fears on death? 44% 64 28% 41 28% 41 10\) In order to prolong the life of cancer patient for a few days in the final stage, would you choose the intensive care unit? 30% 44 1 30% 44 39% 58 11\) Do you believe that the advanced cancer patient when he/she receives the appropriate care should: a) be hospitalized 32% 47 30% 44 38% 55 b) remain at home 57% 83 23% 34 20% 29 12\) According to your opinion, the dying patient should: a) remain at hospital? 34% 50 34% 50 31.5^\*^ 46 b) die at home? 41% 60 29% 42 30% 44 13\) In order to prolong the life of cancer patient for a few days in the final stage, is communication and support from the relatives and friends important? 96% 140 4% 6 0 0 ::: The results of question number 3 \"choice of the medical treatment\" are presented in Table [3](#T3){ref-type="table"} (the distribution of factors influencing choice of treatment in a hierarchical order), and the percentage first choice for each of the items is depicted in Figure [1](#F1){ref-type="fig"}. According to the participants, the first choice on the hierarchy is the factor \" expectance of survival\" (28%), while only an 18.5% chose the factor of \"long term quality of life\" as first in the hierarchy. The most common second factor is \"expected effectiveness\" (28%). The \"long-term quality of life\", depicts third place (19%), while 18.5% chose the factor of \"expected effectiveness\". The patients\' \"short-term quality of life\" comes forth in the hierarchy (21%), while the factor of \"possible complications and side effects\" comes fifth (20.5%). The latter, also occupies the sixth place in the hierarchical ranking (23%). As the most common lowest ranked factor was, \"consequences in the patients\' sexuality\" (63%). {#F1} ::: {#T3 .table-wrap} ::: {.caption} ###### Ranking variables of treatment choices ::: **Expectance of survival** **Expected Effectiveness** **Complications Side-effects** **Consequences in sexuality** **Patient\'s Choice** **Short term QoL** **Long term QoL** ------------- ---------------------------- ---------------------------- -------------------------------- ------------------------------- ----------------------- -------------------- ------------------- **Choices** **1st** 41 (28%) 24(16%) 4 (3%) 0 18(12%) 8 (5.5%) 27(18.5%) **2nd** 20 (14%) 41 (28%) 5 (3%) 0 11(7.5%) 20(13.5%) 27(18.5%) **3rd** 20 (14%) 27(18.5%) 17(12%) 0 18(12%) 12 (8%) 28 (19%) **4th** 23 (16%) 12 (8%) 26(18%) 4 (3%) 17(11.5%) 31 (21%) 11(7.5%) **5th** 5 (3%) 17(11.5%) 30 (20.5%) 2 (1%) 24(16%) 28(19%) 17(11.5%) **6th** 12 (8%) 2 (1%) 34 (23%) 26(18%) 28(19%) 14 (9.5%) 7 (5%) **7th** 3 (2%) 0 6 (4%) 92 (63%) 8 (5.5%) 11 (7.5%) 4 (3%) **missing** 22 (15%) 22(15%) 22(15%) 22 (15%) 22(15%) 22(15%) 22(15%) ::: Reliability of the three different aspects of consultation: information disclosure (Qs, 1, 2, 4, 8), support-communication and care (Qs, 5, 6, 7, 9 and 13) and hospitalisation (Qs, 10, 11, 12) (scales) were assessed by average inter-item correlation, average item total correlation and Cronbach\'s alpha. This initial analysis prompted shortening of the questionnaire via the removal of questions 6 and 13 from the support-communication and care group, and question 12 from the hospitalisation group. This shortening of the questionnaire resulted in improving the average all item total and inter-item correlations to between 0.62--0.70. The internal-consistency reliability coefficients, Cronbach\'s alpha correlations, for the three scales were: information disclosure (Qs, 1, 2, 4, 8) 0.66, support-communication and care (Qs, 5, 6 and 7) 0.69 and hospitalisation (Qs, 10 and 11) 0.50. Using these shortened scales the construct validity revealed that the items were sensibly related to the domains that contained them (convergent validity correlations between 0.6--0.86) and had no association with items from different domains (discriminate validity). This reappraisal of the questionnaire has resulted in the generation of a 21-item scale consisting of 5 multi-item and 5 single-item scales, which demonstrates acceptable validity in assessing relatives\' perceptions and choices for the palliative management of terminal stage cancer patients. Multivariate analysis of the socio-demographic data was conducted for all remaining scales. For question 3, higher scores were placed on all factors except \"patient\'s long term quality of life\" ranked in the order University \> High School \> Primary (0.001 \> p \> 0.02, ANOVA). A similar trend, Child \> Spouse, was observed for relatedness (as only one relative was a parent and 9 were siblings theses were not included in the analysis) with significance reached in \"consequences in patients sexuality\", \"the patients choice\" and \"patient\'s short term quality of life\" (p= 0.007, 0.003 and 0.009, respectively; *t*-test). With respect to relatedness, statistical significantly more child than spouse favored \"yes\" to questions 1b) \"for the prognosis of the disease\", 2b) \"for the percentage of effectiveness\", 4) \"amputative/disfiguring operations\" and 5b) \"physicians from various specialties\" (p = 0.001, 0.036, 0.06 and 0.05 respectively, chi-square). Stratification in relation to the sex of the relative demonstrated significance towards men answering \"yes\" to Ib) \"for the prognosis of the disease\", 10) \"final stage\.... intensive care unit\" and 11b) \"remain at home\" (p = 0.012, 0.016 and 0.048 respectively). For women answering \"yes\" 2a) \"for possible treatment choices\" and 4) \"amputative/disfiguring operations\" (p = 0.037 and 0.010 respectively). As for educational status there was a clearly more pronounced firm \"yes\" to various questions with higher educational status (University \> High school \> Primary) for questions 2c) \" complications or side-effects of the treatment\", 4) \"amputative/disfiguring operations\", 5a) \"the health care team\", 5b) \"physicians from various specialties\", 7) \"cooperate with social worker\", 8) \"inform him/her for the palliative nature of the treatment\", 11b) \"remain at home\" (p = 0.028, 0.001, 0.017, 0.018, 0.024, 0.007 and 0.002 respectively); however the reverse rank order (University \< High school \< Primary) was observed for 10) \"final stage\....intensive care unit\" and 11a) \"be hospitalized\" (p = 0.001). Discussion ========== Traditionally, studies concerning cancer have focused on the effects of diagnosis, prognosis, treatment or the course of the disease on patients\' well-being which refers to psychological as well as social and physical well-being, components referred to as quality of life, Bloom \[[@B22]\], Lewis and Bloom \[[@B23]\], Meyerowitz \[[@B24]\], Penman, Bloom et al \[[@B25]\]. Most clinicians recognize that cancer is painful for patients and their families, but surprisingly there is little research to document this effect. Most of the studies in this area examine the quality of life according to patient assessments, Covinsky, Goldman et al \[[@B26]\]. In this article the assessment and validation of a questionnaire designed to assess terminal stage cancer patient family perceptions and choices in patient management (information disclosure, therapy choices, hospitalization, support-communication and care) including aspects regarding end-of-life and quality-of-life decisions are presented. The goals of palliative treatment are to improve patients\' functioning and their quality of life \[[@B15]\], input from relatives may offer additional measures in this respect. The questionnaire was well accepted by all the relatives that approached the unit. On average, it required 8--10 minutes to complete, and relatives appeared to be pleased that they could have the opportunity to participate in a study examining their patients\' quality of life. Following completion of the survey, the reliability and validity of the questionnaire was assessed by average inter-item and average item totals and Cronbach\'s alpha. Re-appraisal of both the support-communication and care and the hospitalization sections indicated that by shortening the initial 25-item scale to a 21-item scale generated a questionnaire that demonstrates acceptable validity and reliability. In assessing the benefits of the treatment of a fatal illness, health care professionals need to know about the quality of survival. Withholding the truth from a patient appears common in Greece \[[@B27]\], as reported in other nations Niimi \[[@B17]\], \[[@B18]\]\[[@B19]\]; with more information given in the U.S.A and U.K \[[@B28]\]\[[@B29]\]. Many studies suggest that most patients would like to be informed about the diagnosis and prognosis of their disease \[[@B30]\], and evidence suggests that there is little benefit in withholding such disclosure \[[@B17]\]\[[@B18]\]\[[@B31]\]. Frequently, it seems that the caregivers take all the responsibilities or even decide on the patient\'s behalf, and the suffering person remains in ignorance \[[@B32]\]. This is apparent in the present study, where the percentage of the relatives choosing the disclosure of diagnosis or prognosis to their patients is small (23%, and 23% respectively). However, a large percentage (56%) of the relatives believe that only sometimes the diagnosis should be revealed, while 51% indicated that only sometimes the prognosis should be announced. Multivariate analysis demonstrated that there was a significant trend (p = 0.001) for children of patients over spouse in disclosure of \"prognosis of disease\". A similar trend was observed with male relatives over females (p = 0.012). Furthermore, a trend towards a higher educational status also favored this response (however, this was not significant, p = 0.065), and that for patients\' with cancer devoid of metastasis there was also a significant trend towards disclose (p = 0.028). This was similarly matched by the response to question 3 \"patients long term quality of life\", where only in stratification towards patients with cancer (and not metastasis, p = 0.28) was there a significantly greater interest for hierachical importance. In this respect \"expectance of survival\" (p = 0.008), disclosure \"of the prognosis of disease\" (p = 0.028) and \"the involvement of physicians from various specialties cooperating as necessary\" (p = 0.016) all added support to relatives\' interest and involvement in the decision making process aimed at cure in non-metastatic patients. Despite the fact that they confront such a dilemma, they are still fairly certain as to their opinion with other aspects such as their attitudes about informing their patients for the possible treatment choices (71%) and for amputative/disfiguring operations (68.5%). Additionally, 53% of the relatives would choose the patient to be aware of the complications or the side effects of the treatment. Again with these latter two aspects there was a statistically significant trend towards relatives with a higher educational status indicating favorably (p = 0.001 and 0.028 respectively). Still, although in most European countries doctors tend to reveal the truth directly to the patient \[[@B27]\]\[[@B28]\]\[[@B29]\] without the family\'s consent, it is still not common practice yet in Greece, possibly as a result of the strong family bonds. The health care team-patient relationship is a triangle not a dyad, consisting of the health care professionals, the patient and the family. Each part supports the relationship between the other two, and each is affected by what else happens in the triangle \[[@B8]\]\[[@B33]\]. Hence, the involvement of health care teams is very important for the care and treatment of the patients. A high percentage (81.5%) of responders indicated \"yes\", especially the higher educated (p = 0.017), and a high percentage of relatives preferred to cooperate with a social worker (78%, again correlating with the more higher educated (p = 0.024)). The response to the involvement of a psychologist (44%), however, was somewhat contradictory, although only 22% were against it. Cancer affects the family in many ways and it is important for them to have effective co-operation with health care professionals in order to receive care. In the final stage, 67% of family members would prefer to inform patients about the palliative nature of the treatment (p = 0.07 for the higher educated). This is probably due to the fact that the relatives try to preserve the patients\' quality of life until their death. Ongoing open communication between cancer patients and family members is essential, both between each other and with health care professionals. Patients and family members need to feel supported in their expressions. The health care team can help by letting family members know that emotional stress and behavioral changes are not uncommon during illness, that these and other problems can be overcome with patience and support \[[@B22]\]. It is undoubtedly true, that social support can help people to cope with the psychological effects of cancer and help promote physical recovery; when this support is not available or is withheld the patients\' situation is aggravated \[[@B34]\]. The aspect that arises from the study is the fact that the vast majority of the caregivers give a special emphasis on communication between patients and the social environment (family, friends), and the support they receive with respect to disclosure of treatment choices, the involvement of health care teams, cooperation with social workers and physicians from various specialties, all of which demonstrated a stronger positive correlation with a higher educational status. Such communication provides support to the family -- between health care professionals, patients and relatives -- and the means to overcome the dilemma between disclosing information to the patient on prognosis, diagnosis and the treatment choice. Conclusions =========== The study argues that more attention needs to be paid to family members of cancer patients, as they are the primary caregivers in order to have more open-communication with their patients, and to be more empathic and realize the welfare of their patients\' quality of life. Good family interactions can help cancer patients adjust to diagnosis and treatment. Family relationships appear to be particularly important because the family is involved with the patient during the illness. The dynamics of family interactions can set the stage of social interactions with friends and health care providers. After all, cancer is a family illness and disrupts the relationship between them and the health care team. Competing interests =================== None declared Authors\' contributions ======================= All authors participated in the design of the study, conceived of the study, and participated in its design, coordination and statistical analysis. Finally, all authors read and approved the final manuscript. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== Nociceptors are specialized primary afferent neurons and the first cells in the series of neurons that lead to the sensation of pain \[[@B1]-[@B8]\]. The receptors in these cells can be activated by different noxious chemical or physical stimuli \[[@B9]-[@B11]\]. The essential functions of nociceptors include the transduction of noxious stimuli into depolarizations that trigger action potentials, conduction of action potentials from peripheral sensory sites to synapses in the central nervous system, and conversion of action potentials into neurotransmitter release at presynaptic terminals, all of which depend on ion channels \[[@B6],[@B12]-[@B16]\]. Recent expression cloning has led to the identification of the first pain sensory receptor. The cloned receptor is called VR1 (vanilloid receptor subtype 1) \[[@B9],[@B10]\]. The nucleotide sequence of VR1 predicts a protein of 838 amino acids with a molecular mass of 95 kDa. The predicted topological organization consists of six transmembrane domains with a hydrophobic loop between the fifth and sixth domain which lines the ion conducting pore \[[@B17]\]. VR1 has been expressed heterologously in several cell lines and has intrinsic sensitivity to thermal stimuli and to capsaicin (a pungent extract of the *Capsicum* pepper family) \[[@B18]\]. VR1 does not discriminate among monovalent cations \[[@B19]\]; however, it exhibits a notable preference for divalent cations with a permeability sequence of Ca^2+^ \> Mg^2+^ \> Na^+^ ≈ K^+^ ≈ Cs^+^\[[@B9]\]. Ca^2+^ is especially important to VR1 function, as extracellular Ca^2+^ mediates desensitization \[[@B20],[@B21]\], a process which enables a neuron to adapt to specific stimuli by diminishing its overall response to a particular chemical or physical signal. Although not activated by voltage alone, VR1 currents show outward rectification and a region of negative resistance in the current-voltage relation. The VR1 channel is a member of the superfamily of ion channels with six membrane-spanning domains, with highest homology to the *trp* family of ion channels. For those ion channels within this superfamily for which stoichiometry has been directly examined, all have been shown to be composed of four six-transmembrane domain subunits or pseudosubunits, with auxiliary subunits sometimes present as well \[[@B22]\]. An initial characterization of VR1 channels expressed in Cos and CHO cells has recently revealed that, under certain conditions, they run as multimers on pseudo-native (PFO) gels, with tetramers being one of the primary bands observed \[[@B23]\]. Thus, like other six membrane spanning domain channels, VR1 almost certainly forms as a tetramer; whether it combines with homologous subunits to form heteromeric channels remains to be determined. In this study we have examined the electrophysiological and biochemical properties of VR1 expressed in *Xenopus* oocytes. We found that its apparent affinity for the ligand capsaicin is comparable to that observed by others. When examined for size on denaturing gels, we found that the monomer appeared to be a doublet and that there was a band that corresponded to roughly twice the molecular weight of the monomer bands. Through site-directed mutagenesis, we determined that the doublet represented unglycosylated and glycosylated forms of the VR1 subunit monomer and identified the glycosylation site as N604. Next, using a VR1 subunit engineered to be of different size, we show that the larger band on the gel represented dimerized subunits. Several mechanisms underlying dimerization were examined and ruled out. Since VR1 likely forms as a tetramer, the strong interaction we observed between pairs of subunits raises the question of whether this subunit interaction is involved in VR1 function. Results ======= Electrophysiological and biochemical properties of VR1 ------------------------------------------------------ We expressed VR1 channels in *Xenopus* oocytes in order to characterize their electrophysiological and biochemical properties. Using outside-out patch-clamp recordings, we studied the responses of the channels to capsaicin. Figure [1A](#F1){ref-type="fig"} depicts a current family obtained using a saturating (4 μM) capsaicin concentration when the voltage was stepped from a holding potential of 0 mV to from -100 to +100 mV. A current-voltage relation measured from these currents (Figure [1B](#F1){ref-type="fig"}) shows the pronounced outward rectification exhibited by VR1. This rectification has been shown to be independent of external divalent cations \[[@B24]\], and is likely due to a combination of rectification in unitary conductance \[[@B9]\] and voltage-dependent gating \[[@B24]\]. Application of a lower concentration of capsaicin (0.5 μM) activated a smaller fraction of the channels in this patch, giving a smaller overall current (Figure [1C](#F1){ref-type="fig"}). By plotting the normalized currents in the presence of 4 μM and 0.5 μM capsaicin on the same graph (Figure [1D](#F1){ref-type="fig"}), we show that the voltage dependence remains unaltered under both experimental conditions. Figure [1E](#F1){ref-type="fig"} shows a dose-response relation for activation of VR1 channels by capsaicin for the same patch. Fits of dose-response relations with the Hill equation (see Experimental Procedures) yielded values of n = 1.8 ± 0.06 and a K~½~ = 614 ± 110 nM (mean of 5 patches). These values are similar to those previously reported \[[@B20],[@B24]\]. These data indicate that in our system VR1 RNA is expressed as a functional protein with characteristics similar to those reported by other groups. {#F1} To examine the size of VR1 with SDS/PAGE, we constructed a VR1 subunit with a FLAG epitope tag on its extreme C-terminal end (Figure [2A](#F2){ref-type="fig"}). By using both this C-terminal FLAG epitope and an N-terminal epitope for which a commercial antibody is available, we could determine whether the band(s) we observed corresponded to full-length VR1 subunits. When oocytes injected with VR1 RNA were examined with SDS/PAGE and Western blot, three bands were apparent (Figures [2B,2C,2D](#F2){ref-type="fig"}). Two of these bands were seen as a doublet at about 80 kDa and 84 kDa (best seen in Figure [2D](#F2){ref-type="fig"}). A third band of about 200 kDa was also observed. The size of these bands was the same for blots probed with the N-terminal antibody (Figure [2B](#F2){ref-type="fig"}) or the FLAG antibody (Figure [2C](#F2){ref-type="fig"}), indicating that they represent full-length VR1 subunits. For the remainder of this study we used the FLAG antibody due to the lower background it produced. {#F2} Identification of the glycosylation site in VR1 ----------------------------------------------- A doublet of monomers has been previously observed for VR1 expressed in CHO cells \[[@B23]\]. In that case, treatment of cells with peptide-*N*-glycosidase F eliminated the larger monomer band, suggesting that it represented a glycosylated form of VR1. Although the difference in apparent size in that study was 19 kDa and we see a difference of only 4 kDa (Figure [2C](#F2){ref-type="fig"}), we wondered whether glycosylation might explain our doublet as well. We therefore examined the predicted amino acid sequence of VR1 in order to identify potential glycosylation sites (see Experimental Procedures). Figure [2A](#F2){ref-type="fig"} depicts the proposed topology of the six transmembrane domains of VR1. A consensus sequence for N-linked glycosylation is located just distal to the fifth transmembrane domain at position 604. We introduced a point mutation to change the asparagine at position 604 into a serine (N604S). If indeed N604 is a glycosylation site, channels expressed from this N604S construct would be expected to lack glycosylation. Figure [2D](#F2){ref-type="fig"} shows that in N604S channels the upper band of the doublet has been eliminated. These data indicate that wild-type VR1 channels expressed in *Xenopus* oocytes are present in both unglycosylated and glycosylated forms and that the N-linked glycosylation occurs at position 604. Once we determined that the N604S mutation efficiently removed the glycosylated form of the monomer, we tested to be certain that N604S channels were present in the plasma membrane and had electrophysiological characteristics similar to those observed in wild-type channels. We performed patch-clamp recordings on oocytes injected with N604S RNA. The results of these experiments are shown in Figure [3](#F3){ref-type="fig"}. Like for wild-type VR1 channels, capsaicin activated N604S channels in a concentration-dependent manner (Figures [3A, 3C](#F3){ref-type="fig"}, and [3E](#F3){ref-type="fig"}). Current voltage relations at both a saturating (Figure [3B](#F3){ref-type="fig"}) and a subsaturating (Figure [3D](#F3){ref-type="fig"}) concentration of capsaicin show that the voltage dependence of N604S channels is similar to that of wild-type VR1 channels (in Figure [3](#F3){ref-type="fig"} compare data from N604S channels \[blue and red\] to data from wild-type VR1 channels \[green\]): the ratio of the current at +100 mV to that at -100 mV is not statistically different from wild-type VR1. Furthermore, the dose-response relation in Figure [3E](#F3){ref-type="fig"} shows that the apparent affinity of N604S channels for capsaicin is not significantly different from wild-type channels (t-test, p \> 0.05), with n = 2.3 ± 0.3 and K~½~ = 780 ± 71 nM (mean of 5 patches). These data indicate that eliminating the N-glycosylation site of the VR1 channel gives rise to a functional protein electrophysiologically similar to wild-type VR1. {#F3} Identification of the nature of the VR1 high molecular size band ---------------------------------------------------------------- At 200 kDa, the third band observed in Figure [2](#F2){ref-type="fig"} was approximately twice the molecular weight of the monomer. We therefore suspected that it might represent a dimer of VR1 subunits. However, it could also arise from a monomer in combination with another cellular factor. To distinguish between these possibilities, we engineered a VR1 mutant in which amino acids 2--52 in the N-terminal were deleted (termed Δ2--52). This construct gave rise to a high molecular weight band of 180 kDa, 20 kDa smaller than that observed for wild-type VR1 (Figure [4](#F4){ref-type="fig"}). When coexpressed with VR1, heteromeric channels composed of both types of subunits are expected to form. If the higher molecular weight band does represent a dimer of subunits, the heteromeric VR1/Δ2--52 ought to be intermediate in size between the homomeric dimers. When equal amounts of VR1 and Δ2--52 RNA\'s were coinjected into oocytes, three bands were observed: 200, 190 and 180 kDa (Figure [4](#F4){ref-type="fig"}). The 200 kDa band represents the homomeric VR1 dimer whereas the 180 kDa band represent the homomeric Δ2--52 dimer. The middle band, at 190 kDa, is indeed intermediate in size. Our interpretation is that this 190 kDa band is composed of one wild-type VR1 subunit and one Δ2--52 subunit, indicating that the high molecular weight band represents subunit dimers. Furthermore, quantification of the intensity of the bands on the gel revealed a 1:2:1 ratio of the three bands. This ratio of the three ways in which the two types of subunit can assemble is what would be expected if assembly between the different types of subunits were random. {#F4} Next, as with N604S, we wanted to be certain that Δ2--52 expressed as a functional channel in *Xenopus* oocytes. The results of electrophysiological experiments are shown in Figure [5](#F5){ref-type="fig"}. We found Δ2--52 to form functional, capsaicin-activated channels in *Xenopus* oocytes. Although an increase in rectification was consistently observed with these channels, compared to wild-type VR1 (Figure [5B](#F5){ref-type="fig"}), the ratio of the current at +100 mV to that at -100 mV is not statistically different from wild-type VR1. Fits of dose-response relations (Figure [5E](#F5){ref-type="fig"}) with the Hill equation yielded a mean n value of 2.1 ± 0.26, and a mean K~½~ value of 1.98 ± 0.8 μM; neither n nor K~½~ were significantly different from VR1 (t-test, p \> 0.05, for 5 patches). {#F5} Examination of putative cross-linking factors --------------------------------------------- A previous study has reported that the presence of capsaicin and chemical cross-linkers influence the formation of multimers in VR1 \[[@B23]\]. Moreover, this study reported that cross-linking could be a Ca^2+^ mediated process, through the activation of endogenous transglutaminases. In our expression system, the presence of a dimer was seen even in the absence of capsaicin. Thus, we set out to study the factors that could be involved in the formation of this complex. Our first approach was to determine whether this dimer could be due to the presence of a disulfide bond between subunits. Although we have β-mercaptoethanol in the sample buffer for all experiments, it is possible that a disulfide bond refractory to reduction by this reagent was present. We therefore used the stronger reducing agents DTT and TCEP in the biochemical assays, and used them in various steps of the purification (see Materials and methods). Figure [6A](#F6){ref-type="fig"} shows the results for this experiment. The control lane represents the results obtained from oocytes processed under control conditions (with β-mercaptoethanol as the only reducing agent). The next two lanes shown are those of oocytes exposed to the reducing agents DTT (100 mM) and TCEP (20 mM) both during processing and in the sample buffer. We quantified the ratio of intensity of the monomer band to the dimer band. Neither DTT nor TCEP treatment produced a difference in this ratio compared to the control condition (t-test, p \> 0.05 for 3 independent experiments). These data indicate that the presence of the VR1 dimer is likely not due to an intersubunit disulfide bond. {#F6} We next addressed whether Ca^2+^ could affect the ratio in our expression system. Oocytes were processed for biochemical assays under the various experimental conditions depicted in Figure [6B](#F6){ref-type="fig"}. In the first lane we show protein obtained from oocytes exposed to Ca^2+^ (1.8 mM in frog Ringer\'s solution) but not to capsaicin; both monomer and dimer bands can be observed. When capsaicin and Ca^2+^ were added together, the ratio of monomer: dimer remained unchanged in comparison to the previous experiment (p \> 0.05, for 3 independent experiments). We then tested whether eliminating Ca^2+^ from the oocyte media would affect this ratio. As seen in the last two lanes of this gel, the presence of EGTA (2 mM) did not alter the formation of the VR1 dimer, whether capsaicin was or was not present in the assay -- the ratio did not differ from control conditions (p \> 0.05, for 3 independent experiments under the same conditions). Our data indicate that under our experimental conditions Ca^2+^ does not play a pivotal role in VR1 dimerization. Finally, we tested whether the transglutaminase inhibitors cysteamine and monodansylcadaverine (MDC) could affect this process. As shown in Figure [6C](#F6){ref-type="fig"} (second and third lanes), the addition of these compounds for 1 hr to the solution bathing the oocytes and to the homogenization solution did not modify the ratio when compared to the control lane (p \> 0.05, for 3 independent experiments). The concentrations of cysteamine and MDC used here are identical to those previously shown to disrupt transglutaminase-induced cross-linking of VR1 channels in other cell types (Kedei *et al*., 2001). This result comes as no surprise since transglutaminases are known to be Ca^2+^ dependent \[[@B25]\], and our previous experiment demonstrates that Ca^2+^ has no effect on the dimerization we observed. Discussion ========== In this study we have examined the properties of VR1 in a *Xenopus* oocyte heterologous expression system. We report the following findings: (1) Full-length VR1 runs as a doublet of monomers on SDS gels, with apparent molecular weights of 80 and 84 kDa. The smaller band represents unglycosylated subunits and the larger band represents subunits glycosylated at N604. (2) Channels engineered to lack glycosylation (N604S mutants) are correctly folded and targeted to the plasma membrane, with functional properties similar to wild-type channels. (3) A 200 kDa band is also apparent on VR1 SDS gels. By coexpressing full-length and truncated subunits we show that this band represents a dimer of subunits. (4) The interaction between the pair of subunits in the dimer band was quite strong; the intensity of the dimer band relative to monomer was not affected by capsaicin or calcium, and remained unchanged after treatment with reducing agents and transglutaminase inhibitors. Unglycosylated and glycosylated forms of the VR1 monomer -------------------------------------------------------- We show that, in *Xenopus* oocytes, VR1 is expressed in both glycosylated and unglycosylated forms, with \~4 kDa difference in their apparent molecular weights. The complete disappearance of the upper band of the monomer doublet for the N604S mutation strongly suggests that the extracellular linker between the fifth transmembrane domain and the P-loop of VR1 is subject to N-linked glycosylation. Moreover, since channel function did not appear to be affected by the absence of glycosylation (Figure [3](#F3){ref-type="fig"}), it appears that glycosylation at this site is not essential for correct folding and targeting of the protein to the plasma membrane. The position of the glycosylation site in the structure of the channels is in a region of known importance in channel function. Lying between the fifth transmembrane domain and the P-loop, this region contributes to the extracellular vestibule of the ion-conducting pore \[[@B26]\]. The vestibules of ion channels are thought to influence the permeation of ions \[[@B27],[@B28]\] and in CNG channels it has been reported that elimination of the N-glycosylation site (which is in an analogous position between S5 and the P-loop) can induce changes in the apparent half-blocking constants for extracellular and intracellular Mg^2+^\[[@B29]\]. Whereas two monomer bands had been previously observed for VR1 expressed in CHO cells \[[@B23]\], a major difference between this study and ours is the difference in size between the two monomer bands (19 kDa -- about five times larger than we observe). On the other hand, Kedei *et al*. \[[@B23]\] show a doublet in channels purified from DRG cells, which express VR1 endogenously, that is similar to the one we observe. Further, unlike in CHO cells, no additional high molecular weight glycosylated bands were observed. Dimerization of VR1 ------------------- As a member of the six-transmembrane domain superfamily of ion channels, VR1 most likely assembles into tetrameric complexes. Evidence that VR1 is capable of forming multimers has been previously reported when studied under pseudo-native conditions \[[@B23]\]. In this previous study, tetramers were the major band observed, although larger and smaller bands were also seen. Interestingly, there is precedent for an ion channel to retain some intersubunit interactions even on denaturing gels like those used here. The bacterial K^+^ channel KcsA, whose X-ray crystal structure has been solved and is definitively a tetramer \[[@B30]\], runs as a tetramer on SDS gels \[[@B31]\]. Furthermore, mutations that disrupt a known intersubunit interface at the level of the pore disrupt this biochemical tetramerization. The structural interactions that underlie tetramerization of KcsA on gels are disrupted only by heating the sample and by pH 12 \[[@B31]\], treatments that had no effect on the dimerization of VR1 we observed (data not shown). Although it is tempting to conclude that the pH- and heat-resistent dimerization we observe with VR1 results from a covalent interaction, we cannot rule out other explanations such as strong hydrophobic interactions. What does a dimer observed on a gel mean given that the channels are almost certainly tetramers? Precedent for dimerization of limited domains of ion channels abounds. Cyclic nucleotide-gated channels, for example, appear to exhibit functional dimerization of their cyclic nucleotide-binding domains \[[@B32]\]. The \"RCK\" domain of BK channels has a dimerization domain even though BK channels, too, are tetrameric at the level of the pore \[[@B33]\]. Finally, the GluR2 ligand-binding core has a dimerization interface in the crystal structure \[[@B34]\]. Evidence, including the clustering of residues involved in receptor desensitization at this interface, suggests that the dimerization is not just crystallographic but functional. The dimerization observed in the above examples all involve ligand-binding sites. Given that VR1 is also a ligand-activated ion channel it is tempting to speculate that it may too contain such a dimer interface. Recent work identifying amino acid residues that likely comprise part of the capsaicin-binding site \[[@B35]\] suggest that capsaicin binds at the interface between transmembrane segment 3 and the cytoplasm. Could this be a point of intersubunit contact? Although the dimerization we observe may represent a native intersubunit interaction, other possibilities must be considered. For example, hydrophobic interactions can cause membrane proteins to aggregate during purification. Alternatively, a covalent interaction may link pairs of subunits. This possibility will be investigated in the future. Materials and methods ===================== Heterologous expression of channels in Xenopus oocytes ------------------------------------------------------ Segments of ovary were removed from anesthetized *Xenopus laevis.* After gross mechanical isolation, individual oocytes were defolliculated by incubation with collagenase 1A (1 mg/mL) in Ca^2+^-free OR2 medium (82.5 mM NaCl, 2.5 mM KCl, 1 mM MgCl~2~, 5 mM HEPES, pH 7.6) for 1.5--3 hours. The cells were then rinsed and stored in frog Ringer\'s solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl~2~, 1 mM MgCl~2~, 5 mM HEPES, pH 7.6) at 14°C. Oocytes were injected with 50 nL mRNA solution within two days of harvest. Electrophysiology ----------------- Electrophysiological recordings and/or biochemistry were performed 4--8 days after injection. After brief exposure to a hypertonic medium, the vitelline membrane was stripped from each oocyte with forceps. Outside-out patch-clamp recordings were made using symmetrical NaCl/HEPES/EDTA solutions consisting of 130 mM NaCl, 10 mM HEPES, 1 mM EDTA and 10 mM EGTA (pH 7.2). Capsaicin was prepared as a 4 mM stock in dry ethanol and was added to the extracellular solution only. The solution bathing the extracellular surface of the patch was changed using a RSC-200 rapid solution changer (Molecular Kinetics, Pullman, WA). Unless otherwise indicated, all chemicals were purchased from Sigma Chemical Co. (St. Louis, MO). Pipettes were polished to a resistance of 0.3--1 MΩ and immediately before use were dipped in a seal glue composed of one part light mineral oil, one part heavy mineral oil, and 10% parafilm beads to promote formation of high-resistance seals \[[@B36]\]. Currents were low pass-filtered at 2 kHz and sampled at a 10 kHz with an Axopatch 200B (Axon Instruments, Union City, CA). Data were acquired and analyzed with the PULSE data acquisition software (Instrutech, Elmont, NY) and were plotted and fit using Igor Pro (Wavemetrics Inc., Lake Oswego, OR). All currents shown are difference currents in which the current in the patch in the absence of capsaicin has been subtracted. All dose-response curves were measured at +100 mV at room temperature. Smooth curves shown in dose-response relations are fits with the Hill equation:  where I is the current at a given concentration of capsaicin, Imax is the maximal current, K~½~ is the concentration of half-maximal activation and n is the Hill coefficient. Current-voltage relations were plotted using the data obtained from voltage jumps from -100 to +100 mV for 100 ms from a holding potential of 0 mV. Data was normalized by dividing the values of the current at different voltages by the value of the current at -100 mV. When pooled data are discussed in the text, they represent the mean ± standard error of the mean (SEM) A Student\'s t-test (two-tailed) was performed on some data, as discussed in the text. The significance level was set at p \> 0.05. Mutagenesis ----------- A potential glycosylation site was identified by screening the predicted vanilloid receptor 1 (VR1) channel amino acid sequence \[[@B9]\]. A glycosylation consensus sequence of N-X-T/S was found at positions 604--606 (NNS). A point mutant was constructed as outlined below to replace the asparagine at position 604 with a serine. The mutant construct was designated as N604S. The point mutation and deletion mutation were constructed by a method involving oligonucleotides synthesized to contain a mutation in combination with wild-type oligonucleotides in PCR amplifications of fragments of the cDNA. The product of the PCR reaction was then cut with two different restriction enzymes to generate a cassette containing the mutation. The cassette was then ligated into the channel cDNA cut with the same two restriction enzymes. After transformation of bacteria with the ligation product, single isolates were selected, and the entire region of the amplified cassettes was sequenced to check for the mutation and insure against second-site mutations. mRNA was synthesized in vitro, using a standard reverse-transcription kit (mMessage mMachine, Ambion, Austin, TX). All constructs, including VR1, were made in a background of a VR1 subunit in which the FLAG epitope (DYKDDDDK) had been spliced on to the C-terminus. The presence of this epitope was found to have no detectable effects on the electrophysiological properties of the channels (data not shown). SDS/PAGE and Western blot ------------------------- Oocytes were prepared after the method of Rho *et al*. \[[@B29]\]. Typically 30 oocytes were lysed by trituration in 200 μL of a solution containing 100 mM Tris-HCl, 100 mM NaCl, 0.5% Triton X-100, 0.05 mg/mL pepstatin, 0.05 mg/mL leupeptin, and 0.05 mg/mL aprotinin (pH 8.0) and the homogenate was incubated for 15 minutes on ice. The homogenate was then centrifuged at 18,400 g for ten minutes at 4°C in a Jouan CR3i centrifuge. The soluble portion of the homogenate was then transferred to a new tube for an additional centrifugation. 10 μL of supernatant was removed, mixed with 20 μL of Laemmli sample buffer containing β-mercaptoethanol (19:1) and incubated at room temperature for five minutes. The samples were then subjected to SDS/PAGE using NuPage 3--8% or 7% Tris-Acetate precast gels (Invitrogen Corp., Carlsbad, CA). Proteins were then transferred to a PVDF membrane and Western blot analysis was performed. For all blots except that shown in Figure [2B](#F2){ref-type="fig"}, M2 anti-FLAG primary antibody was used, and for the blot shown in Figure [2B](#F2){ref-type="fig"} a polyclonal antibody raised against the N-terminal sequence of VR1 (amino acids 4--21: RASLDSEESESPPQENSC) was used (Neuromics Inc., Minneapolis, MN). Chemiluminescent detection was then carried out using the SuperSignal West Femto kit (Pierce, Rockford, IL). Chemiluminescent signals were captured with the Flourchem Imager (Alpha Innotech, San Leandro, CA), which has a linear range of 4 O.D. units. Densitometry of bands on Western blots was done with the Flourchem software. For comparison of the ratio of monomer to dimer, we included both monomer bands in the monomer category. Because of the large linear range of detection of our instrument, we could compare this ratio with accuracy even if the amount of total protein varied between gels. For SDS/PAGE experiments on disulfide bonds, the oocytes were treated as above, except that the lysis buffer contained either 100 mM DTT or 20 mM Tris (2-Carboxyethyl) Phosphine Hydrochloride (TCEP) (Pierce, Rockford, IL). To examine the effect of calcium and capsaicin on VR1 dimer formation, four different conditions were tested. For calcium-free condition, whole oocytes were rinsed three times with calcium-free frog Ringer\'s solution (96 mM NaCl, 2 mM KCl, 2 mM EGTA, 1 mM MgCl~2~, 5 mM HEPES, pH 7.6) prior to incubation with or without 10--20 μM capsaicin at room temperature for 30 minutes. For calcium- present condition, whole oocytes were bathed in frog Ringer\'s solution with or without 10--20 μM capsaicin for the same duration. The oocytes were triturated according to the method described above, except for the capsaicin-present condition, where 10--20 μM capsaicin was included in the lysis buffer. For experiments on transglutaminase inhibitors, the oocytes were bathed for one hour in frog Ringer\'s solution containing either 20 mM cysteamine or 250 μM MDC. These are expected to be saturating concentrations of the transglutaminase inhibitors. Oocytes were triturated according to the method above, except that either 20 mM cysteamine or 250 μM MDC were included in the lysis buffer. Acknowledgments =============== We would like to thank Dr. David Julius for his kind gift of the VR1 cDNA clone and Dr. Leon Islas for helpful discussions.

Background ========== Health care providers compete for managed care contracts based on cost-effectiveness and quality of care \[[@B1]-[@B4]\]. Information technology (IT) provides a cost-effective way to document productivity, performance measures, cost, and quality of care. Since IT has dropped in cost over time, physician practices are now turning to it to meet these needs. Information technology for this study is defined as computer software used to store, transport, or communicate information \[[@B2],[@B5]-[@B7]\]. The health care organizations that succeed in the 21^st^ century will be those that improve quality and reduce cost. These juxtaposed objectives most likely will be reached through improved handling of information \[[@B2],[@B8],[@B9]\]. The Committee on Quality of Health Care in America reported that most clinical information remains in paper form \[[@B9]\]. This committee made several recommendations for improving quality, including moving clinical information to an electronic format by the end of the decade. Information technology selection in health care has often been performed in a rather informal way, resulting in the purchase of \"white elephants\" \[[@B10]\]. The systems may not perform as planned and may cause additional work for medical staff. The systems are often purchased or developed in pieces without consideration to the overall business strategy \[[@B1]\]. To date, few publications have documented the selection process and the resulting impact of the IT on the health care organization. Most papers give anecdotal descriptions, often by vendors, but lack client perceptions of the information system\'s value \[[@B1],[@B2],[@B7],[@B11]-[@B14]\]. Even at the hospital level, only a few client perceptions of IT adoption have been reported \[[@B15]-[@B19]\]. The number of available papers that examine IT selections within physician practices is even smaller than those papers addressing hospital selections \[[@B3],[@B20]\]. However, many physicians are transitioning from paper to electronic formats for billing records, medical charts, etc. This study aims to understand the process for selecting IT for physicians\' practices and the perceptions of the IT after it is implemented. *The primary objective of this research was to identify the relationship (if any) between the IT selection process and the office staff\'s perceptions of the it\'s impact on practice activities.* Methods ======= To address the research objective, a literature review was completed; an expert panel was formed and consulted; a conceptual model was developed; a telephone interview survey was designed; an exploratory factor analysis was performed; and finally, a logistics regression analysis was performed. The conceptual model for this study was not based on one single overriding pre-established theory (Figure [1](#F1){ref-type="fig"}). Rather, it was drawn from a body of literature as well as from the observations of an expert panel regarding technology selection and how it facilitates or impedes practice activities \[[@B1]-[@B3],[@B11],[@B12],[@B16],[@B21]-[@B42]\]. The expert panel included physicians, health services researchers, informatics researchers, and health care industry consultants. {#F1} The telephone survey was conducted with 407 physician practices in Oregon \[[@B2]\]. The survey elements were based on the literature review and on the feedback from the expert panel. The survey addressed the following descriptive research questions: Q1: Who selects IT for a physician practice (e.g., administrators, clinicians, computer specialists)? Q2: What selection steps are used? Q3: What factors influence the purchase? Q4: Which IT features are selected? Q5: Who (within the practice) customizes the IT? Q6: Is time given to learn the IT? Q7: What are the clinical and office staff members\' perceptions of this IT\'s impact on several office activities (e.g., scheduling, communication, quality reporting)? The design of the telephone survey was reviewed by the Human Subjects Research Review Committee at Portland State University. Sample ------ Providence Health System in Portland, Oregon provided a database of practices (n = 933) for this study. These practices all served Providence Health System in some capacity -- e.g., as primary care physicians or specialists. Eligible practices had acquired software within the past five years but not within the past six months. Practices with software older than five years were disqualified because it was unlikely that the decision makers (if present) would recall the details of the selection process. Practices with software selected within the last six months were dropped because new software often requires a learning time period. The original sample of 933 contained 70 practices that had no computers and 35 that had software purchased only in past six months or more than five years ago. In total, 11.1% of the original sample were excluded. Of the remaining eligible practices (n = 828), 407 completed the telephone survey, representing a response rate of 49.2%. If a qualified respondent at a practice was not reached after at least three attempts (n = 269) or the respondents declined the interview (n = 152), the practice was counted as a nonrespondent. Qualified respondents were involved with software selection or software customization for the practice. Seven practices gave partial interviews and were also counted as nonrespondents. These respondents had to leave in the middle of the interview to address urgent clinic needs. Although these respondents were rescheduled, they were not reached to complete the interviews. Additionally, one respondent gave many \"don\'t know\" responses. The interviewer wrote in the comment section for this office that the respondent was not qualified for the study and should be dropped. Thus, in total, seven partial interviews, and one unqualified interview were dropped from the sample, reducing the total number of offices in the study to 399. The respondents and participating practices are summarized in Table [1](#T1){ref-type="table"}. ::: {#T1 .table-wrap} ::: {.caption} ###### Description of respondents and participating practices ::: **Frequency (n = 399)** ---------------------------------------------------------- ------------------------- *Role in practice* Administrator/office manager, finance manager, etc. 78.9% Billing or scheduling staff 9.0% Physician, physician\'s assistant or nurse practitioner 4.5% Other staff members 3.8% Information system managers 3.0% Nurses and medical assistants 0.8% *Type of practice* Various specialties 55.7% Primary care 32.6% Primary care and various specialties 11.7% *Practice size* Single practitioner 46.3% 2--10 practitioners 41.3% More than 10 practitioners 12.4% *Practice Ownership* Private 83.1% Health system owned 16.9% ::: Second interviews were gathered for 189 of the 407 responding practices. Since almost half of the responding offices represented single practitioners, many of these smaller offices had only one eligible participant. Telephone survey ---------------- The survey questions were developed based on the literature review and discussions with an expert panel. Since many of the respondents were not familiar with technical IT terms, care was taken to present the survey in a \"respondent friendly\" format. Thirteen college student interviewers and two supervisors conducted the interviews using a telephone interviewing software package, Computer Assisted Survey Execution System. A program was written to provide the interviewers with precise dialogue, questions, and precoded responses. As the interview progressed, the interviewer entered the responses into a personal computer. Since the study objective included capturing the perceived impacts of IT, we attempted to record perceptions from two representatives from each practice: the decision maker and a primary user (see [Additional File 1: \"Physician Practice Software Telephone Survey, Dialog and Questions\"](#S1){ref-type="supplementary-material"}). The initial interview that included questions related to the selection process and perceived impacts of the IT lasted approximately 15--25 minutes. The respondent was asked to describe a recent IT purchase (at least six months old). For each practice, the respondent indicated whether a person in a specific role -- e.g., an administrator -- was involved or not involved in selection, and involved or not involved in software customization. Customization in this study referred to providing input to the software vendor for writing software specific to the practice. During the interview we read the respondents a list of selection steps. For each step, the respondent answered \"yes\" or \"no\" as to whether it was performed. During the interview the respondents were read several potential factors that might have influenced the purchase. For each one they rated the statement on a 1-to-6 scale of importance, (ranging from \"no importance\" to \"very high importance\"). Finally, we asked the respondents to react to 12 statements describing potential impacts of the IT on selected practice activities. The statements were intentionally not grouped by any particular theme. The respondents rated each impact statement on a 1-to-5 scale of agreement (\"strongly disagree\", \"slightly disagree\", \"neither agree or disagree\", \"slightly agree\", \"strongly agree\") or selected \"not applicable.\" The second interview with a primary user of the software included mainly the perceived impact questions, and lasted 7--10 minutes. At the completion of the initial interview, each respondent was offered a summary of the results. Statistical evaluation ---------------------- The data from all interviews were first descriptively evaluated, primarily by computing frequencies of responses for each question. Factor analysis (principal components) revealed four latent factors related to the respondent\'s perceived impacts of the IT on four practice activities: scheduling, financial analysis, communication, and medical documentation \[[@B2]\]. Therefore, four subscales were created. The scheduling, financial analysis, communication subscales each included two items, and the medical documentation subscale included three items. Responses of \"not applicable\" were coded as missing. For each subscale the mean of the items was computed. Diagnostic plots of the four practice activity subscales suggested that an explanatory model might be best approached using logistic regression, which relaxes the assumption of normality. The four subscales were recoded to dichotomous variables corresponding to *agree* or *not agree.* If the mean score (of 2--3 impact statements) for a practice activity was greater than 3.0, the respondent was scored as \"1\" for *agree.* If the mean score for a practice activity was 3.0 (\"neither agree or disagree\") or less, the respondent was scored as \"0\" for *not agree.* Each of the four practice activity subscales became the dependent variable in a predictive model. The independent variables entered into the models included the demographic and selection variables. Multiple logistic regression ---------------------------- We attempted four predictive models, one for each of the newly created dichotomous subscales. Only respondents who found the impact statements relevant were included in the predictive models. Multiple logistic regression revealed relationships between the selection process and the perceptions related to the scheduling, financial analysis, and communication processes. Variables that achieved a significance level of p \< .05 were retained in the models. For the perceptions related to medical documentation, no significant selection variables survived the analysis. This was most likely due to the small number of practices with electronic medical records (n = 89) and aggregating all types of electronic medical record (EMRs) regardless of type and number of functions. It is also possible that the decision to purchase an EMR is often made outside the practice -- e.g., a large health system offers EMRs to the practices. For 11 of the 89 practices that had EMRs, the decision was made by a large health system. Data from these practices were not included in the predictive models, thus reducing the number of available practices with EMRs to 78. A summary of the models is presented in this paper. The complete analysis and models are available elsewhere \[[@B2]\]. The predictive models were built using a model building data set (299 randomly selected interviews). The models were then tested with a testing data set (the remaining 100 interviews). One-hundred interviews were needed to insure adequate statistical power. As a check for cross-validation, the accuracy with which the models predicted the perceived impact subscale values using the model building data set was compared to the accuracy achieved with the testing data set. Using the parameters established with the model building data set, agreement (or not agreement) to a perceived impact subscale was predicted for the testing data set. For cross-validation, the accuracy levels were compared using a z-test for proportions. As seen in Table [2](#T2){ref-type="table"}, the scheduling and financial analysis models had non-significant (p \> .05) drops in accuracy. This suggests that the models may be generalized to other physician offices with similar demographics. Since the accuracy level dropped dramatically for the communication model, this model did not \"cross-validate.\" The observations made in this study accurately describe the idiosyncrasies of this sample used to build the communication model, but may not accurately describe other samples of physician offices. ::: {#T2 .table-wrap} ::: {.caption} ###### Cross-Validation Summary ::: **Practice Activity** **Model Building Data Accuracy** **Testing Data Accuracy** **p** ----------------------- ---------------------------------- --------------------------- ------- Scheduling 73%(n = 136) 65%(n = 43) .437 Financial Analysis 86%(n = 166) 73%(n = 56) .059 Communication 90%(n = 89) 66%(n = 35) .003 Medical Documentation ::: Once the results were completed, the expert panel was reconvened to provide insight in interpreting the results. In the sections that follow, the descriptive results, comparison of the decision maker vs user, and each cross-validated model are summarized and discussed. Results and Discussion ====================== Most administrators were involved in the selection (68%) and customization (63%) processes (Table [3](#T3){ref-type="table"}). Clinical staff members were also very involved in selection (62%) but not as involved in customization (33%). ::: {#T3 .table-wrap} ::: {.caption} ###### Selection process ::: ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ **Who selects (Q1)? And Who customizes (Q5)? At least 1 of the following** **Selection Frequency (n = 399)** **Customization Frequency (n = 399)** --------------------------------------------------------------------------------------------------------------- ---------------------------------------------------------------- --------------------------------------- Administrators: office manager, financial manager, or medical director 68% 63% Clinical staff members: a physician, physician\'s assistant, nurse practitioner, nurse or medical technicians 62% 33% Computer consultant from outside the practice 48% 39% Office staff members: billing clerk, scheduler, receptionist, or secretary 42% 42% Representative from: health system, insurance company or patients 18% 18% Computer specialist within the practice 17% 13% **What selection steps are used (Q2)?** **Frequency (n = 399)** Performed cost comparisons 85% Viewed software demonstration 81% Issued a RFP (Request For Proposal) or RFI 79% Compared software options with the best in the field 78% Conducted prior user interviews 76% Performed a needs assessment 75% Developed selection criteria 73% Reviewed your long term business plan 60% Made a site visit 47% Developed a decision analysis 35% Formed a selection committee 21% **What factors influence the purchase (Q3)?**\ **Frequency (n = 399) Rated \"high or very high importance\"** **Influenzing Statements** The software appeared easy to use. 80% Software appeared to improve one or more of the business processes in the practice process. 79% The software provided the most value for cost. 73% The software would help the practice perform processes needed to reach our long term business strategy. 66% The vendor had many sites and was responsive to our needs during the selection process. 55% There were strong testimonies from prior users. 47% The software was already in use by other sites affiliated with this practice. 41% Software was compatible with existing practice systems in the practice. 36% ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ ::: Eighty percent or more of the practices performed cost comparisons and/or viewed software demonstrations. The frequencies for the steps the practices took in selecting software are depicted in descending order in Table [3](#T3){ref-type="table"}. Seventy percent or more of the practices stated that \"ease of use,\" \"improving a business process,\" and \"most value for cost\" were important factors influencing the purchase (Table [3](#T3){ref-type="table"}). The frequency of factors receiving either \"high\" or \"very high importance\" is also presented in descending order in Table [3](#T3){ref-type="table"}. The practices typically chose commercial packages that cost less than \$50,000. (Note: these data were collected during the fall, 1996). Information related to IT cost, customization level, and the number of users is presented in Table [4](#T4){ref-type="table"}. There were four basic software packages considered in this study. The type of package, associated computer activities, and frequencies are presented in Table [4](#T4){ref-type="table"}. The results indicate that more than 85% of the practices used the software for managed care or practice management activities. Fewer than half of the practices used the software for communication activities. Only 23% of the practices accessed a completed patient record with the software. ::: {#T4 .table-wrap} ::: {.caption} ###### Selected IT ::: **Details of IT** **Frequency (n = 399)** ---------------------------------------- ------------------------------------------------------------------------ ------------------------- *Software cost* Given to the practice (\$0) 6% Less than \$10,000 48% \$10--50,000 38% \$More than \$50,0 8% *Customization level* Commercial package (no customization) 49% Commercial package + customization 42% Completely custom package 9% *Number of users* Only 1 user 23% 2--5 users 44% More than five users 8% **Software Category** **Computer Activities** Which IT features are selected (Q4)? **Frequency (n = 399)** Electronic Medical Record Access and complete patient records using computerized patient records 23% Managed Care Track incoming and outgoing referrals 48% Track patient enrollment 44% Capitation accounting 32% Query database 38% Statistical reporting on utilization and outcomes 46% Follow clinical guidelines 24% **At least one managed care activity** **85%** Communication Email or telemedicine to external colleagues 15% Email within the practice 20% Remote link with other information systems 17% Access to internet 8% Electronic data interchange (EDI) 14% Online literature searches 8% **At least one communication activity** **45%** Practice Management Billing and collections 78% Appointment scheduling 50% Accounting spreadsheets 51% **At least one practice management activity** **92%** ::: Ninety percent of the respondents felt the software had impacted their billing process (Table [5](#T5){ref-type="table"}). The first column in Table [5](#T5){ref-type="table"} lists the theme of the impact statement. The middle column is the proportion of respondents who rated the software -- meaning the impact statement was *relevant* to their software. For those who found the impact statement relevant, the last column depicts the proportion who slightly or strongly agreed with the impact statement. For example, in Table [5](#T5){ref-type="table"}, 74% of the respondents felt the software affected the accuracy of their practice documents. Of those, 85% of the respondents agreed that practice documents were more accurate since the software was implemented. ::: {#T5 .table-wrap} ::: {.caption} ###### What are the clinical and office staffs members perceptions of this it\'s impact on office activities (Q7)? ::: **Impact Themes** **\"Relevant Proportion (n = 399)** **For \"relevant\" responders only, the \"agreed\" proportion** ---------------------------------------------------------- ------------------------------------- ----------------------------------------------------------------- Improved billing process 90% 89% More accurate documents 74% 85% Improved ability to analyze managed care costs 65% 85% Improved scheduling process 58% 76% Improved access to patient information at multiple sites 55% 83% Reduced malpractice costs 50% 53% Improved referral process 50% 68% Reduced time for recording patient information 47% 77% Improved communication 44% 76% Improved documented quality 38% 78% Quicker lab results 19% 60% Access to more journals 15% 38% ::: Comparison of decision-maker vs user ------------------------------------ The primary respondents agreed with users on their perceptions of the software\'s impact on scheduling and financial analysis activities (p \< .001). For the scheduling model, Phi was .359, with a maximal Phi of .778. For the financial analysis model, Phi was .418 with a maximal Phi of .920. Since the primary respondent was reasonably knowledgeable about the perceived impacts of the software, we did not include the user data in the remainder of the cross-validated models. The user provided only a few demographics and the perceived impact data, while the primary respondent provided the selection data as well as the perceived impact data. Predicting the impact of the software on scheduling activities -------------------------------------------------------------- For the scheduling model, five selection variables as a group predicted with 73% accuracy the subscale of whether the respondents on average would agree with the following two impact statements: \"The software has improved the scheduling of patients for routine, preventive and urgent appointments.\" \"The software has improved the referral process in sending and receiving referrals quickly.\" The statistically significant (p \< .05) predictors are presented in Table [6](#T6){ref-type="table"} along with the expected response by the respondent and the results of the multiple logistic regression analysis. The second column of the table contains the coefficient (or weighting value of B). The Wald statistic (Bj/standard error) gives a measure of significance of B for the predictor variable. ::: {#T6 .table-wrap} ::: {.caption} ###### Scheduling Model ::: **Predictor** **Coeff. (Bj)** **Wald Statistic** **p** **Odds Ratio** **Respondent\'s Reaction to Scheduling Statements** -------------------------------------------------------------------- ----------------- -------------------- -------- ---------------- ----------------------------------------------------- Software with electronic medical record features. 1.36 3.91 0.0481 3.89 More likely to agree The practice compared software options with the best in the field. 1.36 5.94 0.0148 3.88 More likely to agree Software with practice management features. 0.53 3.34 0.0676 1.70 More likely to agree Importance of prior user testimony 0.33 4.74 0.0295 1.39 More likely to agree The respondent personally selected the software. -1.61 8.22 0.0041 0.20 Less likely to agree ::: Looking at the odds ratios in Table [6](#T6){ref-type="table"}, the likelihood of agreement with the scheduling subscale is almost four times (odds ratio, OR = 3.89) as great when practices selected EMR packages than if they did not select EMR packages. At first this finding was surprising. Many EMRs, however, have automatic recall features when the patient should be called or sent a reminder for a health check. Similarly, the likelihood of agreement was almost four times (OR = 3.88) as great when the practice compared the software options with the best in the field than if it did not perform this step. The practices that selected practice management software were 1.70 times more likely to agree that the software had improved the scheduling and referring of patients than practices that selected other types of software. This finding was expected since these packages typically include a scheduling module. Additionally, practices that considered \"prior user testimony\" important in the selection process were 1.39 times more likely to agree with the scheduling subscale than those practices that did not consider prior user testimony as an important influence. Finally, a respondent who had personally selected the software was less likely to agree with the impact statements (OR = 0.20). The members of the expert panel felt this was a symptom of \"unmet expectations.\" The members of the selection team knew how the software was supposed to perform and were likely disappointed when it didn\'t live up to the vendor promises. These respondents had also probably seen the \"Cadillac\" performers and realized that their software had only achieved \"Chevrolet\" status. Another explanation is that these practices failed to fully implement the software or to adapt clinic workflows to fully utilize the software. In summary, practices that selected EMR or practice management software, that made software comparisons, or that considered prior user testimony as important were more likely to have perceived improvements in the scheduling process than were other practices. Predicting the impact of the software on financial analysis activities ---------------------------------------------------------------------- For the financial analysis model, five selection variables as a group predicted with 86% accuracy the subscale of whether the respondents on average would agree with the following two impact statements: \"The software has created a more accurate and timely billing process.\" \"The software has improved the practice\'s ability to track and analyze costs and revenues associated with managed care contracts.\" The most dramatic increase in odds of agreement (OR = 8.2) occurred when the practice reduced the workload to allow time to learn the software, Table [7](#T7){ref-type="table"}. However, only 36% of the 399 practices reported that reduced workloads were provided during the implementation phase. According to the survey conducted by Ambosa et al. \[[@B21]\], expecting medical staff to learn new software while caring for a full load of patients is a common reason for failure. ::: {#T7 .table-wrap} ::: {.caption} ###### Financial Analysis Model ::: **Financial Analysis Model Predictor** **Coeff. (Bj)** **Wald Statistic** **p** **Odds Ratio** **Respondent\'s Reaction to Financial Analysis Statements** --------------------------------------------------------- ----------------- -------------------- -------- ---------------- ------------------------------------------------------------- Time to learn (reduced workload to learn the software). 2.1 5.44 0.0197 8.20 More likely to agree Software with managed care features. 1.52 7.74 0.0054 4.59 More likely to agree Importance of \"value for cost\' purchase influence. 0.69 8.04 0.0046 2.00 More likely to agree Importance of compatibility purchase influence. -0.41 5.89 0.0152 0.66 Less likely to agree The cost of the software. -1.4 6.74 0.0094 0.25 Less likely to agree ::: The odds of agreement were increased by more than a factor of four (OR = 4.59) for each increase in managed care activities the software contained. Since most managed care software packages are marketed to assist the practice in documenting costs associated with managed care contracts, this finding was expected. Practices that considered value an important consideration were twice (OR = 2.0) as likely to agree with the financial analysis subscale. By contrast, practices that considered compatibility an important influence were less likely (OR = 0.66) to agree with financial analysis subscale. At first the compatibility result was surprising. However, 51% of these practices were first-time buyers, and usually buying billing software, so compatibility was not a critical consideration. Ninety-one percent of first-time buyers who rated compatibility as low-to-no importance agreed with the financial analysis subscale. It is also possible that practices with existing good financial analysis processes (and little room to improve) rated compatibility as important but disagreed that the new software had improved the existing good process. The finding that less expensive packages related to more satisfied buyers was interesting (OR = 0.25). There were many good financial packages available for less than \$10,000 in 1996. Practices that spent less than \$10,000 bought software packages with few, but very functional, features. Those practices that spent more than \$10,000 were purchasing complex systems, perhaps for multiple sites. Financial analysis may just have been a small module of these multi-purpose packages. In summary, practices that considered value important, that did not consider compatibility important, that selected managed care software, that spent less than \$10,000, or that provided learning time during implementation were more likely to perceive that the software had improved the financial analysis process than were other practices. Observations from both models ----------------------------- In looking over the predictors for the two cross-validated models (scheduling and financial analysis), some predictors naturally belong in one model or the other -- e.g., practice management software in the scheduling model and managed care software in the financial analysis model. The themes in the scheduling model center around software features (emr and practice management software, comparison of software options) and usability (prior user testimony and personal selection by respondent). The themes in the financial analysis model include cost (software cost, value), software features (managed care software and compatibility), and learning time. This might suggest that the respondents for the financial analysis model had differing roles in the practice than the respondents for the scheduling model. In both of these models, 79% of the respondents were administrators. Since all types of administrators (e.g., office managers, finance managers) were grouped together, it was impossible to identify the primary role of administrator who responded. The differences in the models also suggest that the predictors of success differ by the types of activities the software is intended to perform. It might appear odd that some predictors (e.g., learning time) did not carry through to both models. It is likely that the type and complexity of software package contributed to the learning demands on the office. Many of the respondents who agreed with the financial analysis subscale chose managed care software that bundled together many activities (tracking incoming and outgoing referrals, patient enrollment, capitation accounting, and/or utilization reporting). For practices learning this type of software, protected learning time was an important predictor of success. For practices implementing practice management software (scheduling, billing, and/or accounting spreadsheets), the learning demand was less. This naturally suggests that the decision to reduce the workload while learning a software package should consider the number and complexity of the tasks to be learned. Limitations and research opportunities -------------------------------------- The respondents for this study primarily represented practices that serve Providence Health System in Oregon. These practices served either as managed care providers or as fee-for-service providers. The only practices excluded were pure HMO providers -- e.g., Kaiser Permanente. The pure HMO practices were excluded because it was unclear whom to interview regarding software selections. Often these practices are given software directly from the organization. Eighty-seven percent of these practices in this study had 10 practitioners or less. Only 17% of these practices had in-house computer specialists assisting with software selection. The results of this study may not generalize to large practices that often have in-house computer specialists assisting with selection. A future study could include a nationwide survey of all types of physician practices, regardless of managed care status, ownership, specialty, or size. This study is retrospective in nature, requiring the respondents to recall a software purchase that occurred several months, perhaps more than a year, earlier. In an \"ideal study design,\" a questionnaire should be distributed to practices that have recently made selections. Another questionnaire addressing the impact on the practice could be sent at a pre-defined follow-up period -- e.g., six months after implementation. This \"ideal study design\" would be difficult to conduct without a sufficient list of practices that have recently purchased software. Perhaps software manufacturers and vendors could provide lists of recent clients (with permission) to interested researchers. The cross-sectional survey design of this study captured the technical aspects of the selection process (e.g., who was involved, what steps that were taken). Although the respondents were given a few \"open-ended\" questions, most provided little additional information. There could have been additional selection steps, influences, and impacts. It is also possible that the observed changes in impact were related to variables we didn\'t attempt to measure -- e.g., ability and desire of management to implement new technologies and to change existing practice activities. Focus groups might be more effective at capturing underlying management expertise. Another very time-invasive approach would be to conduct a series of case studies, documenting the decision-making process over time. This research would need support from practices for observers to remain on-site during the selection process. This format would also promote a more well-rounded, multiple perspectives evaluation. The current study relies on perceptive responses (primarily from office managers) to measure many variables, including impact variables. Their perceptions were related to business-related practice activities. Only 5.3% of the respondents were clinicians. It is likely that expanding this study to include more clinician responses would reveal perceptions related to other processes -- e.g., medical documentation or treatment processes. The subscales (related to practice activities) were formed from responses to only two to three original impact questions. A stronger design of these practice activities impacts would include several questions related to each activity. Given the exploratory nature of this current research, this limitation could not have been foreseen. However, the results of this study open doors for more confirmatory types of studies to design survey instruments that measure software impact with underlying practice activity constructs. This study does not attempt to demonstrate cause and effect. It would be important to have respondents rate existing practice activities (before purchasing software) to control for a \"ceiling effect\" -- practices with existing good processes have little room to improve. If such a trial were designed, it would also need to control for the type of IT and the needs of the buyer. To move toward a more direct measure of impact would require the practices to closely measure performance and behavior. For example, in this study, the respondent is asked if the practitioners have an improved ability to consult professional literature online. A direct measurement method would determine the number of online literature consultations before and after the software installation. Conclusions =========== The results of this research describe the software selection process as it occurs in physician practices. Using a telephone interview survey gave the researcher (and other interviewers) direct contact with the decision makers in each practice. The results of this study also describe how software is perceived to affect several practice activities. The objective of this study was to identify relationships (if any) between the IT selection process and the office staff\'s perceptions of the IT\'s impact on practice activities. The results of the multiple logistic regression models confirmed relationships between the selection process and the perceived impacts related to the scheduling and financial analysis activities. The results of this study demonstrated a relationship (not cause and effect) between the selection process and the user perception of software usefulness. Although many of the relationships were expected (e.g., performing software comparisons, interviewing prior users, and selecting certain software features improved perceptions about practice activities), perhaps one of the most important predictors of improvement was reducing the workload during implementation. Despite the importance of this predictor, only 36% of the practices performed this step in this study. If more practices had performed this step, it might have carried even more weight in the analysis. From a practical standpoint, many of the offices selected and implemented IT but expected the staff to learn the software while caring for a full load of patients. Investigators from a previous study by Ambroso et al. \[[@B21]\] cite this expectation as a common reason for IT failure. One of the secondary findings of this research is that the purchasers of the software (often office managers) had perceptions about the software\'s use similar to those of users (who were not involved in the selection process). This finding supports the use of a single-survey-response study design for understanding perceived impacts related to software\'s impacts on business-related practice activities. List of Abbreviations ===================== EMR: Electronic Medical Record IT: Information Technology OR: Odds Ratio Competing Interests =================== None Declared. Author comments on prior presentation of results ================================================ The results of this study were presented at the Portland International Conference on Management of Engineering and Technology, Portland Oregon, 1997 and 1999. The results were also presented at the Institute for Operations Research and Management Science, Philadelphia, Pennsylvania, 1999. The references for the conference proceedings are listed below. **Eden K,** Kocoaglu, D. Information Technology Selection Process and Perceived Impacts in Physician Practices. In *Technology and Innovation Management.* Portland State University, PICMET conference proceedings, 1999, pp. 562--568. Executive summary presented in proceedings, Portland International Conference on Management of Engineering and Technology, Portland, Oregon, 1999, pp. 392--394. **Eden K,** Kocaoglu, D. Selection of Information Technology in the Health Care Industry. Presented at Institute for Operations Research and the Management Sciences conference. Philadelphia, Pennsylvania, November, 1999. **Eden K,** Kocaoglu D. Selection and Implementation of Information Technology in the Health Care Industry. Preliminary results presented at the Portland International Conference on Management of Engineering and Technology, published in proceedings, Portland, Oregon, 1997, pp. 199--202. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Supplementary Material ====================== ::: {.caption} ###### Additional file 1 Scripted telephone survey, \"Physician Practice Software Telephone Survey, Dialog and Questions\", by K.B. Eden, The file contains the script, questions and pre-coded responses, variables names (in left margins, that appear as: \>xxxx\<), and several logical statements (e.g., goto, if, etc.) to lead the interviewer through the interview. ::: ::: {.caption} ###### Click here for file ::: Acknowledgments =============== I would like to thank Dundar Kocaoglu, PhD, Nancy Perrin, PhD, Mara Tableman, PhD, Wayne Wakeland, PhD, Laurie Skokan, PhD, Dr. Robert Eder, PhD and Bruce Bayley, PhD, for providing direction on this study. I would also like to thank Nancy Perrin, PhD, Mark Helfand, MD, MPH, William Hersh, MD, FACP, Joan Ash, PhD, John Beekman, PhD, Jane Beekman and the BioMed Central reviewers for critically reviewing earlier versions of this manuscript and providing thoughtful feedback. I would like to acknowledge Gary Miranda for his careful editing and suggestions to make this manuscript \"reader friendly.\" This project would not have been possible without the support and funding from Providence Health System and the use of the Portland State University Regional Research Institute. Finally, I would like to thank my husband, Kevin, and my children, Matthew and Erika, who were very supportive during this project.

Background ========== Coiled-coil domains are protein oligomerization motifs, which consist of two or more alpha helices that twist around one another to form a supercoil \[[@B1]\]. Peptides with the capacity to form coiled coils are characterized by a heptad repeat pattern in which residues in the first and fourth position are hydrophobic, and residues in the fifth and seventh position are predominantly charged or polar. This pattern can be used by computational methods to predict coiled-coil domains in amino acid sequences \[[@B2],[@B3]\]. Coiled-coil proteins can be grouped into two general classes. The first class is comprised of short coiled-coil domains of six or seven heptad repeats, also called leucine zippers. They are frequently found as homo- and hetero-dimerization domains in transcription factors (e.g. \[[@B4]\]). The second class is defined by long coiled-coil domains of several hundred amino acids, which are found in a variety of proteins involved in structuring cellular processes \[[@B1]\]. One of the three main classes of cytoskeletal proteins, the intermediate-filament proteins, represents a well-characterized group of coiled-coil proteins. A subgroup of intermediate-filament proteins are the nuclear lamins, which are involved in attaching chromatin to the nuclear envelope and have recently been implicated in inherited human diseases \[[@B5],[@B6]\]. Besides structural proteins of the cytoskeleton, the motor proteins that interact with them also contain coiled-coil motifs. Myosin, the actin motor protein, has an extended coiled-coil domain necessary for the assembly of the muscle thick filaments \[[@B7]\]. Kinesin and dynein, two microtubule motor proteins, also contain coiled-coil domains \[[@B8],[@B9]\]. The coiled-coil motif has also been identified in a variety of proteins associated with centromeres, centrosomes, the nuclear matrix, and chromatin \[[@B10]-[@B14]\]. TPR, a protein associated with the inner filaments of the nuclear pore complex is an extended coiled-coil protein \[[@B15]\]. The filamentous structures associated with both the outer and inner surface of the nuclear pore complex suggest that additional unknown coiled-coil proteins might be involved in its assembly. Several large coiled-coil proteins were also found associated with the cytoplasmic surface of the Golgi apparatus, and two of them have been shown to function in the docking of vesicles to the Golgi cisternae \[[@B16]-[@B18]\]. AKAPs are adapter proteins that attach protein kinase PKA pathways to cytoskeletal elements. At least one AKAP, AKAP 450 is a large coiled-coil protein attaching PKA to the centrosome \[[@B19]\]. Despite the multitude of cellular functions in which coiled-coil proteins participate, the general theme is the association of proteins with the solid-state components of the cell. In some cases, as in the attachment of PKA, this function appears to be crucial for the spatial regulation of signal transduction. Importantly, mutations in individual coiled-coil proteins lead to specific developmental or even behavioral phenotypes. The Drosophila mutations *Mushroom body defect* (a gene involved in the proliferation of neuronal precursor cells) and *quick-to-court* (a mutation that causes elevated sexual behavior) lie in two tissue-specifically expressed coiled-coil proteins of unknown function \[[@B20],[@B21]\]. In humans and mouse models, specific point mutations in nuclear lamins are the cause for autosomal Emery-Dreifuss muscular dystrophy \[[@B22],[@B23]\]. In contrast to animals and yeast, only a small number of long coiled-coil proteins have been identified from plants. MFP1 contains an extended coiled-coil domain and is located at the nuclear periphery of tobacco suspension culture cells \[[@B24],[@B25]\]. The carrot protein NMCP1 is also located at the nuclear rim and has been shown to migrate to the spindle poles in dividing carrot suspension culture cells \[[@B26],[@B27]\]. The cellular function of these plant proteins is not known. Several candidates for myosins and kinesins have been identified from plants and their function as motor proteins is under investigation \[[@B28],[@B29]\]. Besides these few examples, nothing is presently known about plant long coiled-coils and their potential function in anchoring and structuring of different cellular events. In the early nineties, several groups reported lamin-like proteins in different plant species \[[@B30]-[@B32]\]. A purification protocol adapted from animal lamins and applied to pea nuclei was used to purify four proteins between 49 and 66 kD \[[@B31]\]. These proteins were recognized by antibodies against mammalian intermediate filament proteins, and by antibodies against a peptide derived from lamin B. By immunofluorescence and immunogold labeling, these antigens were located mostly in the internal nuclear matrix, and not predominantly at the nuclear rim. Similar results were obtained by other investigators \[[@B32]\]. Although these early reports were promising, no lamin ortholog has been identified molecularly from plants. In a recent search of all publicly available plant sequences, including the full *Arabidopsis* genome, no homologs of lamins were found (\[[@B33]\] and Meier, unpublished results). Similarly, despite earlier reports of lamin-like proteins in yeast \[[@B34]\], the fully sequenced *S. cerevisiae* genome \[[@B35]\] also contains no lamin genes. It is therefore likely that non-animal eukaryotes have a distinct set of nuclear envelope proteins that functionally replace the lamins. Here, we have used a tomato protein containing long coiled-coil domains to search the Arabidopsis genome for related sequences. We have identified seven novel Arabidopsis proteins with extended coiled-coil domains, which form a protein family distinct from other Arabidopsis long coiled-coil proteins. They are characterized by four novel motifs, which are highly conserved in sequence and position, and which were not found in proteins outside of this family. In addition, a rice homolog was identified, indicating conservation between dicots and monocots. Results ======= LeFPP, a novel plant coiled-coil protein ---------------------------------------- A yeast two-hybrid screen was performed to identify interaction partners of the tomato nuclear envelope-associated protein LeMAF1 \[[@B36]\]. A prey plasmid (pAD-LeFPP) was identified that led to activation of the two reporter genes *HIS* and *LacZ* after retransformation with the pBD-LeMAF1 bait. We tested the specificity of this interaction by co-transforming pAD-LeFPP with two additional bait plasmids, pBD-AtMAF1 and pBD-AtRanGAP1. AtMAF1 is one of three Arabidopsis homologs of LeMAF1 (Patel and Meier, unpublished results), and has 45% amino acid identity with LeMAF1. AtRanGAP1 is one of the two Arabidopsis Ran GTPase activating proteins. Its N-terminus has 28% amino acid identity with LeMAF1 and 30% with AtMAF1 (Fig. [1C](#F1){ref-type="fig"}; \[[@B33],[@B37]\]). Fig. [1A](#F1){ref-type="fig"} shows that pAD-LeFPP leads to activation of the *HIS* gene in combination with pBD-LeMAF1 (sector a) and with pBD-AtMAF1 (sector b), but not with pBDAtRanGAP1 (sector c) or the binding domain vector pBD-GAL4 alone (sector d). This demonstrates that the observed interaction is specific for MAF1-like proteins. {#F1} Fig. [1B](#F1){ref-type="fig"} shows the quantification of β-Galactosidase activity in the pBD-LeMAF1/pAD-LeFPP strain. Lanes one and two are the positive and negative control provided with the kit, respectively, lane three shows the activity of the pBD-LeMAF1/pAD-LeFPP strain and lane 4 the activity of the pBD-GAL4/pAD-LeFFP strain. Although yeast two-hybrid interactions can not be easily quantified, the very high level of β-Galactosidase expressed by the pBD-LeMAF1/pAD-LeFPP stain suggests a comparably strong interaction in the yeast two-hybrid screen. The 1927 bp pAD-LeFPP cDNA was sequenced and was shown to contain an uninterrupted open reading frame in frame with the GAL4 activation domain. The conceptual translation of this sequence leads to a protein of 582 amino acids (aa) with a calculated molecular weight of 64.4 kDa and a pI of 4.7. The first methionine is at position 71, and does not have a plant start ATG consensus sequence \[[@B38]\], suggesting that the cDNA is not full length. A BLAST search with the pAD-LeFPP open reading frame indicated that it represents a novel protein with weak similarity to filament-like proteins from animals and yeast (see below). Secondary structure prediction algorithms showed that it is organized almost entirely in alpha-helical domains and contains extensive stretches of coiled-coil domains (Fig. [2A](#F2){ref-type="fig"} and Fig [3](#F3){ref-type="fig"}). The protein was named LeFPP for tomato filament-like plant protein. RNA blot analysis showed a single species of mRNA of ca. 2.4 kb. It is present in tomato leaves, fruits, flowers, light-grown seedlings, and dark-grown seedlings (Fig. [2B](#F2){ref-type="fig"}). {#F2} {#F3} A family of LeFPP-like proteins in Arabidopsis ---------------------------------------------- A BLAST search was performed with the LeFPP amino acid and DNA sequence. In a protein BLAST, sequences with highest scores represented a number of uncharacterized open reading frames in the Arabidopsis genome, followed by sequences for animal filament-like proteins such as myosin, kinesin, ankyrin, etc. While the e-values for the Arabidopsis ORFs were between 2e-56 and 1e-16, the best e-value for a non-plant sequence was 0.003 for an unknown human open reading frame. The Arabidopsis \"hits\" represented seven unique genes, some of which had repeated entries in GenBank (see Table [1](#T1){ref-type="table"}). ::: {#T1 .table-wrap} ::: {.caption} ###### Genomic organization and features of the Arabidopsis *FPP* gene family. ::: --------------------------------------------------------------------------------------------------------------------------------- Name GenBank Accession\ Alias^b^ Chr. \#\ Size ORF (aa)\ pI EST^c^ library (TAIR name^a^) (Map pos. (Mb) (Size protein (kDa) -------- -------------------- ----------- ---------------- --------------------- ----- ------------ ----------------------------- AtFPP1 AGG51666 T5M16.17 1 779 4.7 AV524363 Seedlings, aboveground (AT1G77580) (28.9) (87.8) org. APZL64f04F Mixed tissues AtFPP2 AAF16547 T26F17.2 1 629 4.9 AV524363 Seedlings, aboveground org. (AT1G21812) F8K7.24 (7.65) (71.1) AtFPP3 AAF27033 T12H1.24 3 603 5.3 AV547438 Roots (AT3G05270) T15D2.30 (1.5) (68.3) T22555 Mixed tissues AtFPP4 AAG12549 F6F9.12 1 1024 5.2 AV528396 Seedlings, aboveground org. (AT1G19835) F14P1.46 (6.8) (114.5) F6F9.17 AV556649 Green siliques AV546174 Roots AV547135 Roots AtFPP5 CAB81521 T19K4.250 4 981 6.3 RAFL16-63- unknown (AT4G36120) (16.3) (110.4) O05^d^ AtFPP6 AAG51782 T6B12.1 1 1054 4.8 AA042379 Hypocotyls (AT1G47900) (17.2) (118.5) N96153 Mixed tissues AtFPP7 AAC23780 F26B6.1 2 886 5.7 W43775 Hypocotyls (AT2G23360) T20D16.1 (10.24) (99.9) --------------------------------------------------------------------------------------------------------------------------------- ^a^The Arabidopsis Information Resource. ^b^Additional early annotation names of the same protein. ^c^GenBank accession numbers. ^d^RIKEN Arabidopsis Full-Length cDNA <> ::: A translated BLAST search identified one additional \"hit\" in the rice BAC H0212B02 of chromosome 4 (GenBank accession number AL442007). The \"hit\" lay between the annotated open reading frames H0212B02.17 and H0212B02.18 in a region of DNA not predicted to have coding capacity. However, running GenScan \[[@B39]\] on this part of the BAC, an open reading frame of 1074 amino acids was identified on the opposite strand that contained the \"hit\". The predicted coding region begins at position 114,922 of the H0212B02 DNA sequence, ends at position 107,317, and contains seven predicted introns. This rice open reading frame was named OsFPP and was included in the sequence and structural analysis shown in Fig. [3](#F3){ref-type="fig"} and Fig. [4](#F4){ref-type="fig"}. {#F4} Fig. [3](#F3){ref-type="fig"} shows a comparison of the secondary structure of LeFPP with the seven Arabidopsis ORFs and the rice ORF. All open reading frames are characterized by extended alpha-helical and coiled-coil domains. The extension of the coiled-coil domains shown in Fig. [3](#F3){ref-type="fig"} is derived from the \"coiled-coil\" analysis software in the DNASTAR sequence analysis software package, and was confirmed in all cases by the MultiCoil algorithm \[[@B2]\]. In addition, MultiCoil predicted that all coiled-coil regions have a high probability to form dimers and a low probability to form trimers. The seven Arabidopsis ORFs were named AtFPP1 -- AtFPP7 (Fig. [3](#F3){ref-type="fig"} and Table [1](#T1){ref-type="table"}). They fall into two groups according to size and extension of coiled-coil domains. AtFPP1, AtFPP2, and AtFPP3 are shorter (603 aa to 779 aa) and contain two extended coiled-coil domains connected by a short non-coiled-coil linker. AtFPP4 to AtFPP7 are longer (866 aa to 1054 aa) and contain a longer N-terminal and a shorter C-terminal coiled-coil domain, separated by a long non-coiled-coil linker domain (Fig. [3](#F3){ref-type="fig"}). The rice ORF (OsFPP) resembles structurally more the second group of Arabidopsis ORFs, with two shorter stretches of coiled-coil domains separated by a long linker. Novel sequence motifs define a subfamily of plant coiled-coil proteins ---------------------------------------------------------------------- Alignment of all nine amino acid sequences showed the presence of four novel sequence motifs, which are 100% conserved between the tomato and Arabidopsis proteins (Fig. [4A](#F4){ref-type="fig"}). Motifs II, III, and IV are also 100% conserved in OsFPP, while motif I is present, but slightly less conserved in the monocot sequence. Their location in each protein is indicated in Fig. [3](#F3){ref-type="fig"}. Their position with respect to the arrangement of the coiled-coil domains is strikingly conserved, despite differences in spacing between the four motifs due to the different length of the proteins and the coiled-coil domains. Motifs I and II always correlate with the N-terminal cluster of coiled-coil domains, with motif I being located at the beginning and motif II at the end of this cluster. Motif III is always located in the linker between the N-terminal and C-terminal coiled-coil domains. Motif IV is always located at the very end of the most C-terminal coiled-coil domain. This positioning is conserved even in the rice sequence, where the C-terminal coiled-coil domain is extremely short (Fig. [3](#F3){ref-type="fig"}). The amino acid identities between the eight dicot proteins are in the range of 20% to 50% (Fig. [4B](#F4){ref-type="fig"}), with the rice sequence showing between 13% and 26% identity with the dicot sequences. The Arabidopsis protein with the greatest similarity to LeFPP is AtFPP1 (40% identity). Table [1](#T1){ref-type="table"} shows an overview of the genomic organization of *AtFPP1* through *AtFPP7* and of the features of the predicted proteins. The genes are not clustered, and no two of them are arranged in tandem array. Except for chromosome 5, at least one member of this gene family is present on each chromosome. ESTs have been identified for all genes, indicating that they are expressed (Table [1](#T1){ref-type="table"}). Consistent with the ubiquitous expression pattern found for *LeFPP*, *AtFPP* expression covers a variety of tissues and developmental stages including shoots, hypocotyls, roots, and siliques. In order to identify more long coiled-coil proteins in the Arabidopsis genome, the complete list of annotated ORFs was searched with the keyword \"myosin\" to identify ORFs that had been annotated as myosin-like proteins. Because of the long stretches of conserved heptad repeats, BLAST searches with large coiled-coil proteins usually produce significant e-values with at least some of the large number of myosin heavy-chain sequences in GenBank, frequently leading to this annotation. The 40 ORFs identified this way were analyzed by MultiCoil and 37 were found to contain coiled-coil domains of at least 200 amino acids. 3 ORFs showed no coiled-coil motifs and were discarded. AtFPP2 and AtFPP5 had been annotated as \"myosin-like\" too and were therefore identified in this search. They were discarded to avoid duplication. None of the remaining 35 new coiled-coil proteins contained any of the conserved motifs identified in the FPP family. The 35 new sequences were aligned with the seven AtFPP sequences in a CLUSTAL analysis. Fig. [5](#F5){ref-type="fig"} shows the results represented as a phylogenetic tree. The AtFPP family clearly forms a family of sequences separate from the other coiled-coil proteins (boxed in Fig. [5](#F5){ref-type="fig"}). The three \"short\" AtFPPs, AtFPP1, AtFPP2, and AtFPP3 form a sub-family, as do the four \"long\" AtFPPs AtFPP4, AtFPP5, AtFPP6, and AtFPP7 (see Fig. [3](#F3){ref-type="fig"}). Interestingly, the two most closely related *AtFPP* genes, *AtFPP1* and *AtFPP2* are located in comparable positions on two duplicated segments of chromosome 1, indicating that they might be derived from a recent duplication event. {#F5} Discussion ========== A novel class of plant coiled-coil proteins has been identified that are conserved in dicots and monocots. Seven members of the FPP protein family are present in the Arabidopsis genome. They are characterized by the presence of two clusters of coiled-coil domains separated by a linker domain of variable size. Four highly conserved sequence motifs were identified, which are a signature feature of this protein family and which are also conserved in the homologs LeFPP and OsFPP from tomato and rice, respectively. The tomato homolog of the FPP family binds in a yeast two-hybrid assay to a nuclear envelope-associated protein, indicating that they might function in nuclear envelope biology. Besides the presence of extended coiled-coil domains, the FPP family members have no sequence similarity to the vertebrate lamins. They also do not show significant sequence similarity to other vertebrate nuclear envelope-associated proteins such as LAP1, LAP2, MAN1, and emerin (GenBank accession numbers A55649, S55255, Q9WU40, and XP_048410, respectively). None of these proteins appears to have a convincing ortholog in the Arabidopsis genome (data not shown), consistent with the hypothesis that the composition of nuclear envelope-associated proteins differs significantly between the plant and animal kingdom. Interestingly, a recent report about the protein composition of the nuclear envelope of the highly divergent eukaryote *Trypanosoma brucei* has identified a coiled-coil nuclear envelope protein (TbNUP-1) in this organism that also has no similarity with lamins \[[@B40]\]. TbNUP-1 is a 350 kDa protein with a striking repeat structure. Its full sequence is presently not known but a partial open reading frame of 268 amino acids has been constructed \[[@B40]\]. The near-perfect 144 amino acids repeats contain two coiled-coil domains each. Pairwise alignments of TbNUP-1 with AtFPP1 through ATFPP7 showed no significant sequence similarities beyond the typical low degree of similarity generally observed between coiled-coil proteins (data not shown). The TbNUP-1 open reading frame was also used for a BLAST search of the full Arabidopsis genome. Only comparably weak similarities with other \"myosin-like proteins\" were found. The four conserved sequence motifs identified in the FPP family appear to be unique to this group of coiled-coil proteins and are neither shared by other coiled-coil open reading frames present in the Arabidopsis genome nor by coiled-coil proteins such as lamins from other organisms. Their role in the structure or function of the FPP proteins is presently not known, but their high degree of conservation in the tomato and rice sequences imply that they are of relevance. Several functional small conserved motifs have been identified in other families of long coiled-coil proteins. The nuclear lamins contain short, conserved sequences flanking the lamin rod domain, which are phosphorylated by cdc2 kinase \[[@B41]\]. The 90 amino acid PACT domain, which immediately follows the coiled-coil domain of the related proteins pericentrin and AKAP450 and two uncharacterized coiled-coil open reading frames from Drosophila and *Schizosaccharomyces pombe*, confers centrosomal localization \[[@B19]\]. A conserved domain of ca. 50 amino acids was found in several Golgi-associated large coiled-coil proteins. It has been named GRIP domain \[[@B42]\] or Golgi localization domain (GLD) \[[@B18]\] and is sufficient to specify Golgi targeting in mammalian cells. Nearly all intermediate filament proteins exhibit a highly conserved amino acid motif (YRKLLEGEE) at the C-terminal end of their central alpha-helical rod domain which has been shown to be crucial for the formation of authentic tetrameric complexes and for the control of filament width \[[@B43]\]. No sequences with similarity to the consensus sequences of motifs II, III, and IV were found in the databases that might indicate their potential function. However, a sequence with some similarity to motif I was identified in the hemaglutinin-esterase fusion glycoprotein from influenza virus C (GenBank accession number S07412). Alignment of this sequence (amino acids 39 to 74) with the central conserved residues of motif I: A\--GWEK\-\--E\-\--LK\--L\-\-\-\-\-\-\-\-\-\-\--R\--H-D (\-- indicating non-conserved residues) shows a good match of identical and conserved amino acid residues: A\--GFEK\-\--D\-\--LK\--I\-\-\-\-\-\-\-\--R\--H-E. The fusion glycoproteins contain a coiled-coil domain and are involved in fusing the viral envelope with the host cell membrane. The conformational change of the protein necessary for membrane fusion is driven by ionic interactions between residues of the coiled-coil domain, and Glu74, the last amino acid aligning to motif I, is involved in this process \[[@B1]\]. In light of their putative interaction with the nuclear envelope-associated protein MAF1, it will be interesting to investigate if FPPs are associated with membrane systems of the plant cell. Conclusions =========== A family of novel long coiled-coil proteins has been identified from plants. They are characterized by two clusters of long coiled-coil domains separated by a non-coiled-coil linker and by four novel sequence motifs. Seven members of the *filament-like plant protein* (*FPP*) family are present in the Arabidopsis genome, and one homolog each has been identified from tomato and rice. Thus, this family of proteins appears to be conserved among higher plants, but we have not found convincing homologs from yeast or animals by sequence similarity searches. Tomato FPP was originally isolated in a yeast two hybrid screen with MAF1, a small plant-specific protein associated with the nuclear envelope \[[@B36]\]. Because the full Arabidopsis genome does not appear to contain genes for lamins, it is of particular interest to investigate which other long coiled-coil proteins might functionally replace lamins in the plant kingdom. The FPP family represents an exciting group of candidates for such proteins. Materials and methods ===================== Yeast two-hybrid screen ----------------------- Total RNA from young tomato leaves (*Lycopersicon esculentum*) was isolated as described by Wanner and Gruissem \[[@B44]\]. mRNA was purified using the mRNA purification kit (Pharmacia Biotech, Piscataway, NJ). To construct the yeast two-hybrid library the cDNA synthesis kit, the Gigapack III gold packaging extract and the HybriZAP two-hybrid predigested vector kit (all Stratagene, La Jolla, CA) were used according to the protocol of the manufacturer. The size of the primary library was determined to be 1.5 × 10^6^ plaque forming units. The complete cDNA of LeMAF1 (GenBank accession number AF118113) was cloned as an EcoRI/XhoI fragment between the EcoRI and SalI sites of the bait vector pBD-GAL4 (Statagene). Approximately one million primary transformants were screened for growth on histidine dropout plates. All steps of the library screening including the transformation, the isolation of clones, and the verification of the interactions, and the β-Galactosidase assays were performed as described by the manufacturer. pBD-AtRanGAP1 \[[@B33]\] and pBD-AtMAF1 were both constructed by cloning an EcoRI fragment containing the complete open reading frames into the EcoRI site of pBD-GAL4. The AtMAF1 open reading frame was isolated by PCR from Arabidopis genomic DNA using the 5\' primer 5\'-TCC ATG GCC GAA ACC GAA-3\' and the 3\' primer 5\'-CTA AGT TCA CTT CGA ACT GCT C-3\', which had been designed to match the Arabidopsis MAF1 homolog on the P1 clone MGG4 of chromosome 5 (GenBank accession number AB008267). RNA isolation and RNA blotting ------------------------------ Total RNA from tomato leaves (5--15 mm), fruits (3--8 mm in diameter), whole flowers, and light-grown and dark-grown tomato seedlings was isolated with the Trizol Reagent from Gibco BRL (Gaithersburg, MD). Total RNA (20 μg each) was separated on a formaldehyde gel, blotted to a nitrocellulose membrane, and hybridized with a radioactively labeled fragment corresponding to bp 1140 -- 1926 of the LeFPP cDNA, essentially as described by \[[@B45]\]. Database searches, secondary structure prediction, and sequence comparison -------------------------------------------------------------------------- Sequences with similarity to LeFPP were identified by conducting BLAST (Basic Local Alignment Search Tool; \[[@B46]\]) searches for similarity to sequences contained in GenBank. GenBank data were accessed through NCBI <>. Arabidopsis genome data were accessed through the TAIR database <>. Arabidopsis ESTs were accessed through the TIGR *Arabidopsis thaliana* database <>. Multiple sequence alignments were performed with the DNASTAR protein alignment protocol (DNASTAR, Madison, WI) using the CLUSTAL algorithm. Parameters for CLUSTAL alignments were: Pairwise alignment: Ktuple, 1; gap Penalty, 3; windows, 5; diagonals saved, 5. Multiple alignment: gap penalty, 10; gap length penalty, 10. The PAM250 weight table was used. Percent identities and phylogenetic relationships were derived from the CLUSTAL output using the MEGALIGN software package in DNASTAR. The secondary structures of the proteins, α-helical, and coiled-coil regions were analyzed using DNASTAR PROTEAN software and the coiled-coil regions were confirmed with the MultiCoil algorithm \[[@B2]\]. Coiled-coil proteins from Arabidopsis were identified by searching the lists of complete-genome ORF annotations <> for the keywords \"myosin\" (40 hits) and \"coiled-coil\" (0 hits). Arabidopsis ORFs for CLUSTAL analysis were downloaded either using their GenBank accession numbers or, if not yet available, by searching the TAIR database with the respective gene name.

Background ========== The potential adaptive significance of transposable elements (TEs) to the host genomes in which they reside is a topic that has been hotly debated by molecular evolutionists for more than two decades. While the biological importance of TEs seemed self-evident to those scientists involved in their initial discovery \[e.g., \[[@B1],[@B2]\]\], the subsequent realization that TEs could be maintained in populations even while imparting slight selective disadvantage to their hosts \[e.g., \[[@B3]-[@B5]\]\] drew into question the presumption of adaptive significance. However, even if TEs can be maintained in populations on a day-to-day basis without providing selective advantage, it does not preclude the possibility that the insertion of TEs in or near genes may, in some instances, be of adaptive advantage. If TE insertion variants have contributed to adaptive gene evolution, such variants might be expected to be in high frequency or fixed in populations and species. Initial surveys of natural populations of *Drosophila melanogaster* showing that TE insertion alleles are in uniformly low frequency seemed to negate the adaptive hypothesis \[[@B6]\]. However, the sporadic discovery of degenerate TEs or TE fragments as critical components of functional genes in both plants and animals was sufficient to keep the adaptive hypothesis alive throughout the pre-genomic era \[[@B7]-[@B11]\]. The current availability of the complete or nearly complete sequence of select genomes representing a variety of species is providing an unprecedented opportunity to examine the frequency and distribution of TEs in eukaryotic genomes. The results have been dramatic. TEs not only comprise a significant fraction of nearly all eukaryotic genomes thus far sequenced, they have been found to be components of the regulatory and/or coding regions of a surprisingly large number of genes \[e.g., \[[@B12]\]\]. For example, a recent genomic analysis of 13,799 human genes revealed that approximately 4% harbored retrotransposon sequences within protein-coding regions \[[@B13]\]. Similar results have been recently reported for the nematode *Caenorhabditis elegans*\[[@B14]\]. Here we analyze the polymorphism of two LTR retrotransposon / host gene associations across geographically widespread *D. melanogaster* populations and a representative population of the *D. melanogaster* sibling species, *Drosophila mauritiana*. Results ======= We have initiated a genomic analysis of LTR retrotransposons present in the *Drosophila melanogaster* genome \[e.g., \[[@B15]\]\]. Of particular interest is identification of genes harboring TEs and determining if these insertion alleles are in high frequency or fixed among natural populations as would be expected from the adaptive hypothesis. We report here the results of an analysis of two LTR retrotransposon-containing genes located on the second chromosome of the sequenced *D. melanogaster y; cn bw sp* strain. These two genes present an interesting contrast in that one of them, *Chitinase 3 (Cht3)*, is located within constitutive heterochromatin (Genbank accession: AE002743) while the other, *cathD*, is located in a euchromatic region of the chromosome (Genbank accession: AE003839). Our findings demonstrate that while the euchromatic *cathD* insertion variant was not detected in any of the natural populations examined, the insertion variant present in the heterochromatic *Cht3* gene was found to be apparently fixed throughout the species. These results are consistent with the view that the presence of TEs in constitutive heterochromatin may have relevance to the expression of heterochromatic genes \[e.g., \[[@B16],[@B17]\]\]. Genomic analysis of the sequenced y; *cn bw sp* strain of *Drosophila melanogaster* identified a full-length *Burdock* LTR retrotransposon located just 3\' to the *cathD* gene and a 359bp LTR fragment (complete LTR is 659 bp) of an *Antonia* LTR retrotransposon \[[@B15]\] located within an intron of the *Cht3* gene (Figure [1](#F1){ref-type="fig"}). A set of PCR primers were designed to amplify regions of both genes and retrotransposon sequences. Appropriate pairs of gene and element primers were used to detect the presence or absence of the respective retrotransposon inserts associated with each gene in strains representing 12 geographically dispersed populations of *D. melanogaster*. The results presented in Figure [2](#F2){ref-type="fig"} and Table [1](#T1){ref-type="table"} demonstrate that while the *Burdock* insertion located just 3\' to *cathD* gene is not present in any of 12 strains representing a geographically diverse sampling of natural populations, the *Antonia* LTR fragment located in the intron of the heterochromatic *Cht3* gene is fixed in all 12 strains tested. ![**Genomic structure of the *Cht3 and cathD genes* in the *Drosophila melanogaster* genome.(A)** Chromosome 2 illustrating location of *Cht3* and *cathD* genes (red lines) in reference to constitutive heterochromatin (in blue) \[[@B34]\]. Numbers above each red line refer to Flybase cytogenetic placement. (Chromosome not drawn to scale). **(B & C)** Green arrows represent Flybase-predicted gene regions with corresponding identification. Yellow blocks depict ESTs concordant to the predicted gene region. Blue boxes are predicted exon regions. Red boxes denote LTR position and internal arrows indicate orientation of retroelement. The black line and numbers represent position along the genomic clone sequence which is identified below the figure. Black arrows indicate direction and location of forward (f) or reverse(r) PCR primers. **(B)** An *Antonia* LTR fragment (359nt) is inserted in an intron of *Cht3* in 12 geographically distinct *Drosophila melanogaster* strains. **(C)** A full-length *Burdock* retroelement, only present in the sequenced y; *cn bw sp* strain, overlaps the predicted exon boundaries of the *cathD* gene by 6nt.](1471-2148-2-5-1){#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### Presence or absence of retroelement sequence associated with *cathD* and *Cht3* genes in strains representing 12 natural populations of *D. melanogaster*. ::: Georgaphic area Strain *cathD / Burdock* *Cht3 / Antonia* ----------------- --------------- ------------------- ------------------ Lab stock *y; cn bw sp* \+ \+ Americas Athens \- \+ California \- \+ Antilles \- \+ Europe Germany \- \+ Italy \- \+ Africa Capetown \- \+ Cotonake \- \+ Dimonika \- \+ Kenia \- \+ Niamey \- \+ Swaziland \- \+ Asia India \- \+ Males and females from each strain were tested. (+) indicates presence of retroelement sequence, (-) indicates absence of retroelement sequence. ::: It is formally possible that the presence of the *Antonia* LTR within the *Cht3* intron was the result of a chance fixation event prior to the expansion of *D. melanogaster* around the world. Thus, to further test the adaptive hypothesis we compared the level of sequence divergence within the LTR and its flanking intronic sequence between the two sibling species *Drosophila melanogaster* and *Drosophila mauritiana*. If the LTR-containing intron is under stabilizing selection, a lower than neutral rate of substitution would be expected. A total of 685 bp of the *Cht3* intron was sequenced. This region spans 264 bp of the 359 bp *Antonia* LTR fragment. The sequence of this region in a *D. melanogaster* (Dimonika, Africa) and *D. mauritiana* (Mauritius, Africa) strain was aligned with the homologous region in the sequenced *D. melanogaster y; cn bw sp* strain (Figure [3](#F3){ref-type="fig"}). The two *melanogaster* strains were 100% identical. The *melanogaster* sequences were found to be only 1.3% (9 substitutions/685 nucleotide sites) diverged from that of *D. mauritiana*. This value is significantly less than half of the expected 4.3 % (± 2.7) divergence based on the *Drosophila* neutral substitution rate of 0.016 (± 0.005) substitutions/site/million year \[[@B18]\] over the estimated 2.7 million years separating the two species \[[@B19]\]. To directly compare the substitution rate for the *Cht3* intron with that of another *Drosophila* gene intron, we randomly selected intron 1 of the *Drosophila* alcohol dehydrogenase (*Adh*) gene. *Adh* is a widely studied *Drosophila* gene and it has been sequenced in several *Drosophila* species including *D. melanogaster*, accession X60793 \[[@B20]\] and *D. mauritiana*, accession M19264 \[[@B21]\]. The sequence divergence between *D. melanogaster* and *D. mauritiana* in the *Adh* intron 1 (7.9%, Figure [4](#F4){ref-type="fig"}), is higher than that for the LTR containing *Cht3* intron (1.3%). These results strongly suggest that conservative selection has been operating on the LTR containing intron associated with the *Drosophila Cht3* gene over the past 2.7 million years. ![**Nucleotide alignment of the 659 bp intron 1 of the *Adh* gene in *Drosophila melanogaster* and *Drosophila mauritiana*.** Sequences obtained through GenBank for *D.melanogaster* (accession: X60793, \[[@B20]\] and *D.mauritiana* (accession: M19264, \[[@B21]\]). Sequences were aligned using MacVector (See Materials and Methods for details).](1471-2148-2-5-4){#F4} Discussion ========== For many years, constitutive heterochromatin was considered to be of little or no functional significance \[[@B22]\]. This view seemed to be supported by early molecular studies showing that heterochromatin consists almost exclusively of highly repeated and middle repetitive DNA \[e.g., \[[@B23],[@B24]\]\]. The middle repetitive fraction was viewed as the descendent of once active TEs that had the misfortune of inserting into transcriptionally inert heterochromatin at some point in their evolutionary history \[e.g., \[[@B6],[@B20]\]\]. The view of heterochromatin as a genetic wasteland gradually changed with the mapping of a number of functionally important *Drosophila* genes to constitutive heterochromatin \[e.g., \[[@B24]-[@B31]\]\]. Reexamination of *Drosophila* constitutive heterochromatin revealed that long stretches of highly repetitive DNA are interrupted by \"islands\" of retrotransposon sequences \[e.g., \[[@B32],[@B33]\]\]. *Drosophila* genes in heterochromatin are typically associated with these islands of retrotransposons \[[@B2],[@B31],[@B34]-[@B36]\]. It has been suggested that transposable elements inserted into heterochromatin may locally alter chromatin structure \[e.g., \[[@B16]\]\]. Our results suggest that in at least some instances, the association of heterochromatic genes with transposable element sequences may be of adaptive significance. Conclusions =========== The results presented here are consistent with the hypothesis that a 359 bp fragment of the *Antonia* retrotransposon located within the intron of the heterochromatic *Drosophila melanogaster Cht3* gene may be of adaptive evolutionary significance. Further genomic and molecular analyses will be required to assess the general importance of LTR retrotransposon sequences to the evolution of heterochromatic gene structure and function. Materials and Methods ===================== Gene Region Annotation ---------------------- BLASTS of sequenced DNA turned up several instances of genes proximal to an LTR retrotransposon. Sequence retrieval was initiated via BLASTN searches (default parameters- \[[@B37]\]) against the BDGP <> and GenBank <> databases using LTRs from previously identified *Drosophila* retroelements as queries \[[@B15]\]. Results with E-values \< e^-10^ were annotated on the corresponding clone, whereupon visual inspection of several annotations confirmed the presence of retroelements proximal to known genes. Selected genes were BLASTed against NCBI\'s EST database and mapped along with predicted transcript structures from Flybase <>. Chromosomal location of clones was also determined from Flybase. PCR --- *D. melanogaster* strains from Dimonika, Niamey, Swaziland, Kenia, Capetown, Cotonake, and India were obtained from Charles F. Aquadro, Cornell University. Germany, Italy, and Antilles strains were obtained from Nikolaj Junakovic, Universitá la Sapienza, Rome, Italy. California and Athens strains are from Daniel Promislow, University of Georgia. *D. melanogaster y; cn bw sp* strain was obtained from the Bloomington, IN, stock center. The *D. mauritiana* (241.0) strain was provided by the Bowling Green, OH, *Drosophila* stock center. PCR primers were designed with MacVector 7.0 <> and synthesized by Integrated DNA Technologies (Coralville, LA) (Table [2](#T2){ref-type="table"}). Three PCR reactions were performed per strain, per gene. For all PCR reactions, 1.0 μl of a single fly DNA prep \[[@B38]\] was used and amplification was performed in a Hot Top equipped Robocyler Gradient 96 (Stratagene, La Jolla, CA). 10 μl of product was separated on a 1% agarose gel in 0.5× TBE running buffer containing 0.25 μg mL^-1^ ethidium bromide. Gel images were visualized by UV transillumination. ::: {#T2 .table-wrap} ::: {.caption} ###### Primers used for PCR analysis. ::: Primer Sequence Position 1 ---------------------- ----------------------------------- --------------- *cht3(f)* 5\'-TGATGCCATACTCCTACTCCGTAAC-3\' 27910:27934 *cht3(f2)* 5\'-ATGAAAAACGGATGGACAGCC-3\' 18549:18569 *cht3(r)* 5\'-CATTCCTGTTTGCCAACCCC-3\' 28395:28376 *Antonia LTR(f)* 5\'-TTAAGCGAACGTCGGAGAC-3\' 21299:21317 *Antonia LTR(r)* 5\'-CCACTAGAAGGGTGAAAACTGC-3\' 21570:21549 *cathD(f)* 5\'-GGTGAAGCCGCCATTTTACG-3\' 220780:220799 *cathD(r)* 5\'-CGCCCAGCACAAACTTGATTAC-3\' 221240:221219 *Burdock LTR(f)* 5\'-TGACCGACGCTTCTAATCTTCC-3\' 221638:221659 *Burdock element(r)* 5\'-GGTTGGCAGTATGGGAACTTAGTG-3\' 221918:221895 ^1^ Position is relative to location of first nucleotide within clone, AE002743 for *cht3* or AE003839 for *cathD*. ::: ### Cht3 PCR The PCR products for primer set *cht3(f)* and *cht3(r)* and primer set *Antonia LTR(f)* and *Antonia LTR(r)* were amplified in a 25 μl reaction containing 3 mM MgCl~2~, 10X PCR buffer supplied by Pierce (Rockford, IL), 5% DMSO, 0.2 mM dNTPs, 0.5 μM of each primer, and 0.5 U of Taq DNA polymerase supplied by Pierce \[Rockford, IL\]. The program consisted of an initial incubation at 94°C for 3 min for 1 cycle, a 30 cycle extension at 94°C for 30 sec, 56°C for *cht3(f)/cht3(r)* primer set or 57°C for *Antonia LTR(f)/Antonia LTR(r)* primer set for 30 sec, 72°C for 1 min 30 sec, and a 1 cycle final extension of 72°C for 5 min. The PCR products for primer set *cht3(f2)* and *LTR(r)* were amplified in a 25 μl reaction containing Expand Long Template PCR System 10X PCR buffer \#1 supplied by Roche (Indianapolis, IN), 0.35 mM dNTPs, 0.32 μM of each primer, and 1.3 U of Expand Long Template PCR System DNA polymerase mix supplied by Roche (Indianapolis, IN). The program consisted of an initial incubation at 94°C for 3 min for 1 cycle, a 30 cycle extension at 94°C for 30 sec, 52°C for 30 sec, 68°C for 3 min, and a 1 cycle final extension at 68°C for 5 min. ### cathD PCR The reaction mix and program used for all sets of primers are the same as those described for primer set *cht3(f)* and *cht3(r)* and primer *set Antonia LTR(f)* and *Antonia LTR(r)* in the *Cht3* PCR (above). The annealing temperature for primer set *cathD(f)* and *cathD(r)* is 58°C, for primer set *Burdock LTR(f)* and *Burdock element(r)* is 59°C, and for primer set *cathDff*) and *Burdock element(r)* is 56°C. Sequencing ---------- PCR products of the *Cht3* intron were sequenced in the Molecular Genetics Instrumentation Facility at the University of Georgia. Sequences were aligned with Mac Vector 7.0 and compared to the published y; *cn bw sp* strain. Substitutions and insertion/deletion sites (indels) were summed for each sequence product and compared to the expected divergence based upon the neutral substitution rate. The expected number of polymorphisms between *D. melanogaster* and *D. mauritiana* was calculated based on the *Drosophila* neutral substitution rate of .016 (± 0.005) substitutions per site/million years \[[@B18]\] on 685 bp over a divergence time of 2.7 million years \[[@B19]\]. {#F2} {#F3} Note added in proof ------------------- The two Cht3 intron fragments descibed in Figure 3 have the following provisional accession numbers in GenBank: D. melanogaster, Africa - AY081055 D. mauritiana - AY081054 Acknowledgements ================ E.W.G supported through an NIH Genetics Training Grant. This work supported by a National Institutes of Health grant to J.F.M.

Background ========== Exposure of leaves to strong light sometimes causes the reduction of photosynthetic efficiency, a phenomenon referred to as photoinhibition \[[@B1]-[@B3]\]. The susceptibility of plants to photoinhibition depends on the species and growth light-environments \[[@B4]\]. In general, shade plants or low-light grown plants are more susceptible to photoinhibition than sun plants or high-light grown plants \[[@B4]\]. Since photoinhibition has a potential to lower productivity and plant growth, avoidance of photoinhibition is critical for the fitness and survival of plants in natural habitats \[[@B2],[@B5],[@B6]\]. It is now widely accepted that harmful Reactive Oxygen Species (ROS) produced upon illumination are involved in the mechanism of photoinhibition \[[@B7]\]. Singlet-excited oxygen (^1^O~2~) can be generated by the interaction of O~2~ with triplet-excited chlorophyll (^3^Chl) formed in the PSII reaction center \[[@B8]\]. Superoxide radical (O~2~^-^) is unavoidably produced by the Mehler reaction via electron transfer to O~2~ at photosystem I. Dismutation of O~2~^-^ results in the formation of hydrogen peroxide (H~2~O~2~) and the reaction between H~2~O~2~ and transition metal ions generates hydroxyl radical (•OH) which is the most reactive radical among ROS \[[@B7]\]. These ROS can oxidize molecules in chloroplasts including D1 protein in the PSII reaction center \[[@B9]\] and thiol enzymes in the Calvin-Benson cycle \[[@B10],[@B11]\] to inhibit partial reactions of photosynthesis eventually leading to photoinhibition. Recently, ROS have been shown to also bring about photoinhibition through the inhibition of *de novo* synthesis of D1 protein of PSII that is essential to recover from photoinhibition \[[@B12]\]. To protect photosynthetic machinery from ROS-mediated photoinhibition, chloroplasts contain a high amount of ascorbate in stroma at 20 to 300 mM \[[@B13]\]. In the stroma, ascorbate contributes to suppress the accumulation of photoproduced-H~2~O~2~ by acting as the electron donor for ascorbate peroxidase (APX) which detoxifies H~2~O~2~ to H~2~O \[[@B14]\]. Ascorbate can be also involved in the detoxification of ^1^O~2~, O~2~^-^ and •OH via nonenzymatic reduction \[[@B13]\]. In these reactions, ascorbate is univalently oxidized to monodehydroascorbate (MDA). To maintain ascorbate concentration in the chloroplasts, MDA should be promptly regenerated to ascorbate during illumination. MDA can be reduced to ascorbate by NAD(P)H in a reaction catalyzed by monodehydroascorbate reductase (MDAR) \[[@B15]\] or directly by reduced ferredoxin \[[@B16]\]. MDA, an organic radical, can undergo spontaneous dismutation to produce ascorbate and dehydroascorbate (DHA) in the absence of MDA reduction. DHA can be reduced by glutathione either through nonenzymatic reaction at a slow rate or enzymatic reaction at a much higher rate \[[@B17]\]. Dehydroascorbate reductase (DHAR, EC 1.8.5.1) is considered to be involved in enzymatic DHA reduction in the chloroplasts \[[@B17]\]. Until now, however, direct *in vivo* evidence showing the physiological significance of DHAR has been yet not available. We previously reported that leaves of *Ficus microcarpa* L. f. cv. Golden Leaves (GL), a tropical fig tree, lack heat-stable DHAR activity \[[@B18]\]. In GL, the canopy sun-leaves, which are always exposed to direct sunlight, show characteristic yellow whereas those in wild-type (WT, *Ficus microcarpa* L. f.) exhibit normal green even when exposed to direct sunlight. The mechanism for yellow leaves production in GL is unknown. In barley, it has been reported that leaves incubated with CO~2~-enriched air show yellow color due to photoinhibition of photosynthesis \[[@B19],[@B20]\], a phenomenon apparently similar to that observed in GL. We hypothesized that GL possess an incomplete machinery of photosynthesis which is susceptible to photoinhibition. The aim of this study was to directly examine the hypothesis. The side-by-side comparisons shown in this study demonstrate that GL is much more susceptible to photoinhibition of photosynthesis than WT. Results ======= Photoinhibition in the field ---------------------------- GL is a cultivar of the tropical fig tree *Ficus microcarpa* L. f. (WT) that is natively distributed in subtropical/tropical regions of Asia and Oceania. A significant characteristic of GL is the presence of yellow leaves that contain high amounts of flavonoids and negligible amounts of chlorophyll and carotenoids \[[@B21]\]. The reduced content of photosynthetic pigments in the yellow leaves can be increased by shading them from high-light. In shade conditions, therefore, GL exhibits normal green leaves which cannot be morphologically distinguished from WT leaves (Fig. [1](#F1){ref-type="fig"}). We have suggested that high susceptibility of GL to photoinhibition could explain the phenomenon \[[@B21]\]. However, there was no direct evidence available to show the susceptibility of GL to photoinhibition. {#F1} Figure [2](#F2){ref-type="fig"} shows diurnal change of maximum PSII efficiency (*F*~v~/*F*~m~) in field grown attached sun-leaves collected from the canopy. Chlorophyll content of sun-leaves in GL (yellow) was 0.05 g m^-2^ and that in WT (green) was 0.44 g m^-2^. There was no significant difference between GL and WT in *F*~v~/*F*~m~ values measured early in the morning (6 a.m.). The value of *F*~v~/*F*~m~ in GL decreased from 0.77 (6 a.m.) to 0.46 (2 p.m.) and subsequently recovered to 0.74 in the evening (6 p.m.). In contrast to GL, *F*~v~/*F*~m~ values in WT were almost constant throughout a day. Shade-leaves of GL(green), collected from the inside of tree, did not show any decrease in the *F*~v~/*F*~m~ value (data not shown). These data obtained in the field conditions suggest that GL is highly susceptible to high-light. {#F2} Photosynthetic performance -------------------------- We compared photosynthetic performance between GL and WT to confirm that the fundamental photosynthetic machinery of GL is normal. Shade-leaves of GL and WT, both of which are green, were used for this purpose. Figure [3](#F3){ref-type="fig"} shows the light response curves of CO~2~ assimilation rate (based on surface area) in detached shade-leaves. The chlorophyll content of shade-leaves in GL (green) was 0.58 g Chl m^-2^ and that in WT (green) was 0.70 g Chl m^-2^. GL showed a very similar light-response curve to WT and there was no substantial difference in the maximum activity of CO~2~ assimilation between them. These results clearly demonstrate that the photosynthetic capacity of GL is almost identical to that of WT, suggesting that photosynthetic machinery of GL is functionally not defective. {#F3} Photoinhibition in shade-leaves of Golden Leaves ------------------------------------------------ Figure [4](#F4){ref-type="fig"} shows effects of high-light on several indicators of photoinhibition: D1 protein content, *F*~v~/*F*~m~ value, O~2~ evolution rate and CO~2~ assimilation rate. Before high-light exposure, there was no substantial difference between WT and GL shade-leaves in D1 protein content on protein basis (Fig. [4A](#F4){ref-type="fig"}) and *F*~v~/*F*~m~ value (Fig. [4C](#F4){ref-type="fig"}). GL showed a significant decrease in the D1 protein content and *F*~v~/*F*~m~ value upon high-light illumination (Fig. [4A,4B,4C](#F4){ref-type="fig"}). Similar to the indicators specific for PSII activity, those for net photosynthetic activity (i.e. O~2~ evolution and CO~2~ assimilation rates) also decreased upon high-light illumination (Fig. [4D,4E](#F4){ref-type="fig"}). In contrast to these responses observed in GL, WT showed only a small decrease in D1 protein content and *F*~v~/*F*~m~ value (Fig. [4A,4B,4C](#F4){ref-type="fig"}). The O~2~ evolution and CO~2~ assimilation rates in WT did not change even under high-light condition (Fig. [4D,4E](#F4){ref-type="fig"}). {#F4} Reduced NPQ capacity in Golden Leaves ------------------------------------- When plants are exposed to high-light that exceeds the capacity of photosynthesis, excess absorbed light energy can be safely dissipated as heat \[[@B22]\]. This energy dissipation can be measured as the NPQ of PAM chlorophyll *a* fluorescence analysis. NPQ includes the quantum yield of PSII as an index of heat energy dissipation, photoinactivation of PSII and distribution of photon acceptance between PSII and PSI due to the state transition. It is known that the sate transition can be controlled by the redox state of Q~A~\[[@B23]\], which can be measured as 1-qP. Figure [5](#F5){ref-type="fig"} shows that there is no significant difference between GL and WT in yield and 1-qP values. Furthermore, measurements of the electron transport rate (ETR) showed essentially no difference in the sensitivity of PSII photoinactivation between GL and WT (data not shown). However, GL showed a much lower level of NPQ compared to WT (Fig. [5C,5F](#F5){ref-type="fig"}). These results suggest that low NPQ in GL can be attributed to dysfunction of the heat dissipation mechanism. {#F5} Discussion ========== This study has demonstrated that photosynthetic activities of GL are highly susceptible to photoinhibition under high-light conditions (Figs. [2](#F2){ref-type="fig"}, [4](#F4){ref-type="fig"}). There was essentially no difference in photosynthetic efficiency between GL and WT at least up to 700 μmol m^-2^ s^-1^ for a short time (Fig. [3](#F3){ref-type="fig"}), but NPQ in GL was lower than that in WT at 2300 μmol m^-2^ s^-1^ (Fig. [5C,5F](#F5){ref-type="fig"}). It has been reported that a *npq1* mutant of *Arabidopsis thaliana* shows increased susceptibility to photoinhibition, but is unaffected in short-term photosynthetic O~2~ evolution and CO~2~ assimilation \[[@B24],[@B25]\]. For the operation of NPQ, special xanthophyll pigments in the light-harvesting complexes of PSII are known to have critical roles \[[@B26]\]. The extent of NPQ in plants is strongly correlated with the levels of zeaxanthin and antheraxanthin that are formed from violaxanthin by the enzyme violaxanthin deepoxidase (VDE) located on the luminal side of thylakoid membranes \[[@B22]\]. In excessive light conditions, VDE is activated and converts violaxanthin to zeaxanthin via antheraxanthin when acidification in thylakoid lumen reaches a critical threshold \[[@B27]\]. The activity of VDE is also regulated by the concentration of ascorbate in the lumen because VDE can utilize only ascorbate as a reductant \[[@B28]\]. It has been shown that decreases in ascorbate availability severely affect VDE activity *in vivo*\[[@B28]\]. Because thylakoid membranes do not have any active transport mechanisms for ascorbate, the concentration of ascorbate in the stroma determines that in the lumen by a passive diffusion mechanism \[[@B29]\]. In *Arabidopsis,* it has been clearly demonstrated that NPQ performance *in vivo* is reduced by a mutation that causes ascorbate deficiency \[[@B13]\]. We previously reported that the ascorbate content of GL is severally decreased in daytime under field conditions \[[@B30]\]. Thus, it is reasonable to assume that decrease of ascorbate content would be a cause of dysfunction of NPQ in GL. Among antioxidant enzymes, we have shown that GL lacks heat-stable DHAR activity \[[@B18]\]. Plant cells possess several kinds of enzymes that exhibit DHAR activity: thioredoxin \[[@B31]\], glutaredoxin \[[@B32]\] and disulfide isomerase \[[@B33]\]. Recently, two distinct chloroplast proteins that exhibit DHAR activity have been isolated from spinach; one is *Kunitz*-type trypsin inhibitor \[[@B31]\] and the other is specific DHAR \[[@B34]\]. Because both chloroplastic DHARs show higher heat-stable activity than non-chloroplastic DHAR \[[@B31],[@B34]\], GL may lack chloroplastic DHAR(s) though molecular information is not available. In fact, heat-stable activity cannot be detected in nonphotosynthetic organs such as fruits, that do not include chloroplasts \[[@B18]\]. Although chloroplastic DHAR(s) have been presumed to function in the ascorbate regeneration system for maintaining the ascorbate concentration in the chloroplasts \[[@B7]\], the physiological significance of DHAR in chloroplasts is still controversial \[[@B35]-[@B38]\]. Since the activity of stromal MDAR is high, the spontaneous disproportionation from MDA to DHA and ascorbate may not occur under *in vivo* situations. Therefore, there is an argument that DHAR might be dispensable for the ascorbate-glutathione cycle to protect chloroplasts from high-light stress \[[@B36]\]. However, there is another location in chloroplasts where the stromal enzymes cannot directly regenerate ascorbate, namely, the lumen space. In the lumen, oxidation of ascorbate by xanthophylls, α-tochopherol and PSII result in the production of DHA and all of these reactions could be promoted under high-light conditions \[[@B13],[@B35]\]. Therefore, stromal DHAR would be essential to reduce DHA produced in the lumen side. In this context, chloroplastic DHAR must be physiologically indispensable for photoprotection mechanisms particularly when leaves are exposed to high-light stress conditions. This is consistent with previous reports that DHAR activity in leaves of *A. thaliana* is increased by high-light conditions \[[@B39]\]. Thus, we consider that ascorbate availability of GL is decreased by high-light due to a lack of DHAR activity which eventually results in lowered performance of NPQ (Fig. [5C,5F](#F5){ref-type="fig"}). Ascorbate is involved in photoprotection not only through NPQ but also primarily through ROS scavenging \[[@B7]\]. It should be noted that high susceptibility of GL to photoinhibition could be also explained by a low ROS scavenging capacity which may result in photodamage of target molecules in chloroplast \[[@B7]\] and the inhibition of D1 *de novo* synthesis \[[@B12]\]. The mechanism for yellow leaf production is still unknown. Sun-leaves of GL contain low amounts of chlorophyll and carotenoids but shade-leaves show comparable amount of those to WT, implying that exhibition of yellow leaves in GL is associated with a long-termed irradiation of high-light \[[@B21]\]. When leaves of GL were exposed to high-light, they showed significant symptoms of photoinhibition (Figs. [2](#F2){ref-type="fig"}, [4](#F4){ref-type="fig"}). It has been suggested that long-termed photoinhibition (or chronic photoinhibition) can lead photobleaching of photo synthetic pigments such as chlorophyll and carotenoids \[[@B3]\]. The *npq1 Arabidopsis* mutant which is partially defective in NPQ has been reported to show the photobleaching of chlorophyll in high-light conditions \[[@B40]\], a phenomenon apparently very similar to that observed in GL. It is plausible that photoinhibition is involved in an early stage of the yellow leaf-producing mechanism. Unlike herbaceous plants, the leaf-yellowing phenomenon does not cause senescence or cell death \[[@B21]\]. It has been shown that flavonoids accumulated in leaves could protect tissues from oxidative damage by complementing the ascorbate-glutathione cycle \[[@B15],[@B41],[@B42]\]. In addition to the ROS scavenging function, flavonoids have long been known to be effective sunscreen pigments particularly against UV radiation \[[@B43]\]. Increased flavonoid pigments in GL may play roles including ROS scavenging function, UV-protection and light-attenuation for shadeleaves. The mechanisms of photoinhibition susceptibility and, in particular of high-light tolerance in GL leaves remain to be fully clarified. Although it is difficult to apply molecular biology techniques to GL, we consider that the tropical tree could provide a unique opportunity for examining the important determinants of survival in high-light environments. Materials and Methods ===================== Plant materials --------------- Plant materials were *Ficus microcarpa* L. f. cv. Golden Leaves (GL) and *Ficus microcarpa* L. f. (WT) grown in the campus of the University of the Ryukyus on Okinawa island (26\'15\"N, 127\'45\"E) in Japan. The study period was from March to June in 1999--2000. The photo synthetic photon flux density (PPFD) on the plant canopy was c.a. 2500 μmol m^-2^ s^-1^ in full sunlight conditions during the period. There was no significant difference in growth conditions of PPFD between WT and GL. Green leaves, grown under light conditions of about 100 μmol m^-2^ s^-1^ as the maximum, were harvested and used for laboratory experiments as shade-leaves. To avoid the drought of leaves, leaf petioles were continuously kept in distilled water throughout the measurements. To activate the photosynthetic activity, detached leaves were exposed to an illumination of 1000 μmol m^-2^ s^-1^ for 30 min before experiments. Illumination of leaves ---------------------- Leaves were illuminated with white light from three halogen lumps (400 W). Leaf temperature was maintained at 30°C with an electric fan or a thermostatically controlled water bath during illumination. Various light intensities were obtained by placing wire screens in front of the light source. Measurements of chlorophyll fluorescence ---------------------------------------- Chlorophyll fluorescence was measured with a PAM-2000 chlorophyll fluorometer system under atmospheric conditions (Heinz Walz, Effeltrich, Germany). After a dark adaptation period of 15 min, minimum fluorescence (*F*~o~) was determined by a weak red light. Maximum fluorescence of dark-adapted leaf (*F*~m~) was measured during a subsequent saturating light pulse of white light (8000 μmol m^-2^ s^-1^ for 0.4 s). Maximum fluorescence (*F*~m~\') and steady-state fluorescence (Fs) of illuminated leaf were measured upon a subsequent saturating light pulse of white light (8000 μmol m^-2^ s^-1^, for 0.4 s) to determine NPQ (*F*~m~-*F*~m~\')/*F*~m~\'), yield of PSII \[(Fm\'-Fs)/Fm\] and qP (Fm\'-Fs)/Fm\'-Fo) \[[@B44],[@B45]\]. A 650 nm light-emitting-diode (LED) equipped with PAM-2000 was used for the illumination of leaf as actinic light. Measurements of gas exchange ---------------------------- CO~2~ assimilation was determined by the difference in CO~2~ concentration between inlet and outlet of the assimilation chamber with a CO~2~ gas analyzer (LI-6251; LiCor, Inc., Lincoln, Nebraska) in 0.036% CO~2~ and 21% O~2~\[[@B46]\]. Measurements of CO~2~ assimilation rates were carried out under an illumination of 2300 μmol m^-2^ s^-1^, except for light response curve measurements. Approximately 25 cm^2^ of leaf were employed for an experiment. The air in the chamber was stirred rapidly to maintain a high boundary layer conductance for CO~2~ diffusion. O~2~ evolution was determined using a gas-phase Walker-type oxygen electrode system (Model LD-1; Hansatech, Norfolk, U.K.) in 5% CO~2~ and 21% O~2~\[[@B47]\]. Measurements of O~2~ evolution were carried out on 4 cm^2^ of leaf disc under an illumination of 600 μmol m^-2^ s^-1^. Measurements of D1 protein content ---------------------------------- Leaf (0.5 g) was homogenized for 30 s at 0°C in a grinding medium containing 50 mM potassium phosphate (pH 7.0), 5 mM phenylmethylsulfonyl fluoride and 5% (w/v) polyvinylpolypyrrolidone. The homogenate (4 ml) was mixed with 4 ml of 125 mM Tris (pH 6.8) containing 5% (w/v) SDS, and then incubated at 80°C for 3 min to solubilize proteins. After cooling down the sample to room temperature, aggregates were removed by the centrifugation at 1,500 × g for 30 s. The supernatant (800 μl) was mixed with the same amount of a solution that contained 0.4 M sucrose, 8 M urea, 5 mM EDTA and 5% (w/v) 2-mercaptoethanol. SDS-PAGE was carried out according to a previously reported method \[[@B48]\]. Polyacrylamide gels containing 6.0 M urea were used for the stacking gel (4.5 %) and separation gel (13 %). Samples that contained 8 μg protein were loaded in each lane. Immunoblot analysis was performed using specific polyclonal antibodies raised against the D1 protein \[[@B48]\]. The intensities of the protein bands were determined with NIH image system version 1.61 (NIH, USA). List of abbreviations used -------------------------- APX, ascorbate peroxidase; ^1^Chl, singlet excited chlorophyll; ^3^Chl, triplet excited chlorophyll; DHA, dehydroascorbate; DHAR, dehydroascorbate reductase; D1, the reaction center-forming protein of photosystem II; *F*~v~/*F*~m~, ratio of variable to maximum chlorophyll *a* fluorescence; GL, Golden Leaves; MDA, monodehydroascorbate; MDAR, monodehydroascorbate reductase; NPQ, nonphotochemical quenching; •OH, hydroxyl radical; ^1^O~2~, singlet excited O~2~; O~2~^-^, superoxide radical; PPFD, photosynthetic photon flux density; ROS, reactive oxygen species; WT, wild-type. Acknowledgements ================ We are grateful to Dr. M. F. Cohen of University of the Ryukyus for critical reading of the manuscript. This work was supported by a Grant-in-Aid for Scientific Research (B) and (C) from Japan Society for the Promotion of Science to HY.

Introduction ============ Suspected acute appendicitis is the most frequent cause for emergency operations in visceral surgery worldwide. In the UK 37,289 patients had an emergency excision of the appendix in the year 2000 \[[@B1]\]. In approximately twenty percent of all cases however, the diagnosis is incorrect and patients undergo surgery without having acute appendicitis at all \[[@B2]-[@B5]\]. Operations of bland appendices may lead to morbidity in 4.6 percent \[[@B6]\] and to mortality in 0.14 percent \[[@B6]\] of cases. Despite the introduction of reports of highly accurate diagnostic procedures for the diagnosis of acute appendicitis a big retrospective cohort study \[[@B7]\] concluded that the rate of misdiagnosis (the false positive rate) has not changed over the last twenty years. One potential explanation of that finding might be, that studies reporting on test accuracy overestimate the true potential of correct classification due to inappropriate methodology and bias of reported results since primary research on evaluation of tests is generally poor in quality \[[@B8]-[@B10]\]. Online searches of the electronic databases revealed a number of broad reviews, commentaries and recommendations on tests for predicting acute appendicitis but there was a dearth of focused, rigorous diagnostic overviews of the available evidence. These publications showed that there are several prediction rules and tests or markers purported to be predictive of acute appendicitis. However, they offer only limited guidance for practice because traditional literature reviews evaluating tests for acute appendicitis have not applied the scientific strategies to assemble, appraise, and synthesize relevant evidence, which have been embodied in the criteria for high quality reviews. Given the variation that exists in both practice and research, the uncertainty regarding the quality of the underlying evidence, and the importance of early prediction of acute appendicitis in view of the available effective treatments, there is a clear need for a comprehensive, systematic and quantitative overview of the diagnostic value of the various tests purported to be predictive of acute appendicitis. At present there is a dearth of such reviews and in this commentary, we will describe how we are using such a systematic approach to collate and critically appraise the available literature in the diagnosis of acute appendicitis. Methods ======= Study identification -------------------- Non-comprehensive search strategies can lead to significant bias in the retrieval of relevant literature. This weakens the strength of inferences from systematic reviews and poses a particular problem in reviews of diagnostic tests \[[@B11],[@B12]\]. Therefore we will identify literature via general bibliographic databases including MEDLINE and EMBASE, specialist computer databases such as DARE and MEDION (a database of diagnostic test reviews set up by Dutch and Belgian researchers), the Cochrane Database of Systematic Reviews, relevant specialist registers of the Cochrane Collaboration, conference proceedings and BIOSIS without language restrictions. In addition we will contact individual experts and those with an interest in this field to uncover grey literature and we will contact the manufacturers of tests. Hand-searching of selected specialist journals, checking of reference lists and SCISEARCH to identify frequently cited articles will complete our searches. In cases of duplicate publication, the most recent and complete versions will be selected. A comprehensive database of relevant articles will be constructed -- a preliminary search has been carried out in order to estimate the size of the relevant literature. MEDLINE Searches located 800 potentially relevant citations. Expanding search to other databases, hand searching, reference list searching and or contact with authors might add another 100% citations, so the total is likely to be 1600. Letters will be sent to major centres and the first author of each shortlisted selected paper published in the last five years, asking them whether they know of any published or unpublished relevant studies not included on our list. The search strategy used to identify articles in MEDLINE is shown in: [appendix.doc](#S1){ref-type="supplementary-material"}. Study selection --------------- Studies will be selected for inclusion in the review in a two-stage process using the selection criteria based on those shown in Table [1](#T1){ref-type="table"}. First, a comprehensive database of the literature search will be constructed. The citations will be scrutinised by two reviewers to obtain copies of full manuscripts of all citations that are likely to meet the selection criteria. Two reviewers will then independently select the studies, which meet predefined, and explicit criteria regarding populations, tests, outcomes and study design. These criteria will be pilot tested using a sample of papers and agreement between reviewers will be measured. When disagreements occur the two reviewers will meet. Experience suggests that often the cause of the disagreement is a simple oversight on the part of one of the reviewers. When this is not the case the issue will be resolved by consensus involving a third reviewer. ::: {#T1 .table-wrap} ::: {.caption} ###### Study Selection Criteria. ::: ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ **Population:** Patients suspicious to have acute appendicitis **Diagnostic tests:** Prediction rules Inflammatory markers (C-reactive protein, leucocytes count) Transabdominal ultrasound Computer Tomography (CT) Magnetic Resonance Imaging (MRI) Scintigraphy Diagnostic laparoscopy clinical history and physical examination **Outcome measures:** Test accuracy, morbidity and mortality from misdiagnosis **Study design:** Diagnostic test studies will be selected. They consist of observational studies (prospective or cross-sectional) of defined non-randomised populations in which the results of the diagnostic test of interest are compared with the results of a reference standard allowing generation of a 2 × 2 table to compute indices of diagnostic accuracy. ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ ::: Study validation ---------------- Papers meeting the selection criteria will be appraised to rate their methodological quality. In addition to using ratings of study quality as possible explanations for differences in results, the extent to which primary research met methodological standards is important per se for assessing the strength of any conclusions that are reached. There is an ongoing debate over what constitutes the best quality assessment tool for diagnostic test studies. We will evaluate elements of study design, which are likely to have a direct relationship to bias in a diagnostic test study \[[@B10]\]\[[@B13]\]\[[@B14]\]\[[@B15]\]. The items shown in Table [2](#T2){ref-type="table"} will be used for methodological quality assessment. Agreement for the quality assessments will be calculated, and disagreement resolved, in the same fashion as for the assessment of study selection. We will evaluate the agreement between the two reviewers using percentage agreement and weighted kappa statistics \[[@B16]\]. ::: {#T2 .table-wrap} ::: {.caption} ###### Criteria for study validation. ::: --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- **Population:** Consecutive recruitment of an appropriate spectrum of eligible patients will be considered ideal. Convenience sampling, i.e. arbitrary recruitment or non-consecutive recruitment will be deemed inadequate. In the absence of any explicit information on the method of recruitment, the article will be categorised as unclear reported population enrolment. **Diagnostic test:** The description of the diagnostic test will be considered ideal if the methodology is reported together with the measurement parameter and the cut-off level for an abnormal result. In the absence of any of the above information in the manuscript, then the diagnostic intervention will be considered as unclear reported. **Outcome measures:** Blinding will be considered ideal if it is clearly reported that the results of the various tests were not divulged. Information on the number of patients recruited into the study and those whose outcome data were known will also be sought from the manuscripts. Withdrawal of patients from the study, missing data and lack of outcome data outwith the study hospital will be categorised as lost to follow-up. In particular we will look for evidence of verification bias where the rates follow-up and confirmation of outcome are different in patients with positive test results compared to those with negative test results. Any available randomised trials will be assessed for validity separately to the diagnostic accuracy studies considering factors associated with bias in such trials, e.g. concealment of randomisation, sequence generation, blinding and follow-up. --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ::: Data collection --------------- The extraction of study\'s findings will be conducted in duplicate using a pre-designed and piloted data extraction form to avoid any errors. Given the extent of insufficient reporting in the medical literature, we propose to obtain missing information from investigators whenever possible. It is otherwise impossible to distinguish between what was done but not reported and what was not done. A template of data extraction form is shown in: [appendix.doc](#S1){ref-type="supplementary-material"}. Analysis -------- By analysis we mean synthesis of results from individual studies (meta-analysis), and exploration of variation in results from study to study (heterogeneity) and generation of the most useful combination of tests. We will conduct meta-analyses to generate summary estimates of sensitivities, specificities, predictive values, likelihood ratios (LRs) and receiver operating characteristic (ROC) curves where appropriate \[[@B13],[@B14],[@B17]\]. If meta-analysis is considered to be inappropriate, we will describe the identified evidence in the context of appraised quality. If a meta-analysis is considered appropriate, we will examine the correlation between true positive rates and false positive rates in individual studies. If the correlation is poor, we will use LR as the main accuracy measure. If we find a correlation then we will generate a summary ROC curve \[[@B18]\] in addition to pooling of LRs. Many authorities considered this the preferred method of pooling test results from primary studies \[[@B13],[@B14],[@B17]\]. The summary ROC plot provides a way of summarising the performance of a test from the results of several studies over a range of test thresholds. However, our preference for LRs is based on the published recommendations that LRs are more clinically meaningful as measures of diagnostic accuracy \[[@B15]\]. Our experience has been that the true positive rates and false positive rates in individual studies are poorly correlated in which case it is not feasible to generate a summary ROC curve. Moreover, when the outcome of a test is of binary nature (positive or negative) LRs are more clinically meaningful than ROC curves. One disadvantage of analysis using LR is that it generates two measures for each test, one for a positive result and another for a negative result. A ratio of LRs will be used to generate a single measure called diagnostic odds ratio, which is more suitable for statistical analysis. For the purpose of meta-analysis, we will weight the logLR from each study in inverse proportion to its variance in order to combine the LRs from each study. To demonstrate the practical application of the summary LRs generated, we will calculate posttest probabilities for acute appendicitis using Bayes\' theorem. An estimate of the pretest probability will be obtained by calculating the prevalence of the outcome event in the population studied. The following algorithm of equations will be used for calculating post-test probability: pretest probability = prevalence of acute appendicitis pretest odds = pretest probability / (1 -- pretest probability) posttest odds = likelihood ratio × pretest odds posttest probability = posttest odds / (1 + posttest odds) In order to deal with the uncertainty of the estimate, we will generate 95% confidence intervals around the point estimate. Approximate variance for the posttest odds will be obtained by adding the variances of the combined LRs and pretest odds, enabling the calculation of its 95% confidence intervals. The 95% confidence intervals for the posttest probabilities will then be generated by converting the limits of the posttest odds to their respective probabilities. Heterogeneity of results between different studies will be formally assessed using the Breslow-Day test which compares for each study the ratio of the odds of having the outcome of interest when the test result is positive to the odds of having the same outcome when the test result is negative\[[@B19]\]. To explore causes of heterogeneity in the estimates of diagnostic accuracy of the tests for acute appendicitis, we will conduct a sensitivity analysis. This will be carried out by subgroup analyses to see whether variations in population, intervention, outcomes and study quality will affect the estimate of diagnostic accuracy. Results of pooled analyses will be provided within cogent patient groups. Discussion ========== In summary, systematic reviews of diagnostic literature to predict acute appendicitis allow us to assess the quality of the available evidence and to identify specific tests (including history, physical examination and tests) that have diagnostic value. These reviews should lead to formulation of recommendations for current practice and future research. Just as an evidence-based culture in delivery of health care has been supported by systematic reviews of literature on therapeutic interventions, we can expect to see an extension of this approach in the area of care involving use of diagnostic and screening tests. Competing interests =================== none declared Authors\' Contributions ======================= LMB and JS initiated the project and wrote the protocol. DBB, SAB, MGB and FMO screened the pilot searches, all authors commented on earlier drafts and approved the final manuscript. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Supplementary Material ====================== ::: {.caption} ###### Appendix.doc ::: ::: {.caption} ###### Click here for file ::: Acknowledgements ================ The authors would like to thank Gill Richie and Julie Glanville of the Centre for Reviews and Dissemination in York (UK) for searching the databases.

Background ========== Attention Deficit Hyperactivity Disorder (ADHD), known as Hyperkinetic Disorder under the ICD-10 classificatory system, is a common disorder that affects between 1.5% and 6% of children \[[@B1],[@B2]\]. The disorder is an early onset condition that is associated with educational failure, social difficulties and an increased risk of antisocial behaviour and problems in adult life including substance misuse and criminality \[[@B1]\]. ADHD now represents the commonest reason for follow up in child and adolescent psychiatry clinics \[[@B2]\] and an average general practitioner can expect to have between two and four children receiving treatment for it on their list. Nevertheless there is evidence that most children who fulfil diagnostic criteria remain undetected in the general population \[[@B3]\]. The treatment of ADHD has been the focus of extensive review and several authoritative reports and recommendations have been recently circulated \[[@B4]-[@B8]\]. These recommendations support the use of stimulant medication (such as methylphenidate (Ritalin, Equasym) and dexamphetamine (dexedrine)) as first line treatment (either alone or in combination with non-pharmacological interventions including behaviour therapy) for children with accurately diagnosed ADHD \[[@B4]-[@B9]\]. Moreover the high level of under treatment has also been highlighted in the recent National Institute of Clinical Excellence (NICE) guidance on the use of methylphenidate in ADHD \[[@B4]\]. Discussion ========== Current recommendations are that medication should be initiated by child and adolescent psychiatrists or paediatricians or learning disability specialists with particular expertise \[[@B4],[@B5]\]. Once the condition is stabilised children can then be followed up in primary care \[[@B4],[@B5]\] and NICE guidance recommends health authorities should draw up shared care protocols \[[@B4]\] . Is primary care in this country ready to follow these recommendations? Follow up of ADHD is often undertaken by primary care in the Unites States \[[@B8],[@B10]\]. Moreover adult ADHD is becoming increasingly recognised and some of the cohort of children and adolescents who are currently on stimulant medication will graduate to adult mental health services. These factors together with statements in the NICE guidance on the use of methylphenidate are likely to increase pressure on primary care to become more involved in the management of ADHD. However there is a great need for caution and careful planning before implementing shared care protocols for ADHD and sharing follow up with primary care. Primary care teams are already having to assume responsibility for follow up of an increasing number of chronic conditions and although for some (such as diabetes) there is evidence this is successful \[[@B11]\], the monitoring of children on treatment for ADHD may pose different problems. Most general practitioners and nurses will have undergone no or very little training in child and adolescent psychiatry. Many general practitioners do not feel confident about their skills in recognising and undertaking follow-up management of ADHD and feel that education is essential \[[@B12]\]. Attitudes amongst members of the primary care team members may also represent a problem. The media represents an important source of knowledge of ADHD for many general practitioners \[[@B12]\]. Given that recent media features have highlighted problems with over diagnosis and over treatment this is likely to affect attitudes within the primary care team to the use of stimulant medication and heighten anxieties about management. Although there is some evidence that most general practitioners are willing to take on physical monitoring of children on medication for ADHD (e.g. height, weight and blood tests if necessary) the only available data suggest general practitioners feel clinical monitoring of these children is a specialist responsibility \[[@B12]\]. Therefore shared care protocols for ADHD may have limited impact in reducing specialist workload and will need to incorporate effective mechanisms for close integration and good communication between services to ensure that patients and their families receive appropriate care. It is therefore important that a clear strategy for the follow up care of children with ADHD is devised. Planners need to ensure that roles and responsibilities of different healthcare sectors in monitoring children with ADHD on medication are clearly demarcated and acceptable to the professionals concerned. When these clear roles have been identified, it is then important to ensure that professionals possess the appropriate skills and knowledge to effectively undertake the assigned role. The resource implications of carrying out this monitoring also need to be addressed so that this can be done successfully (as has been achieved for monitoring of anticoagulant therapy). It is essential that patient needs are met in a coordinated and supported manner to avoid a situation of placing children and families into a \"no mans land\" of uncertain responsibility and inadequate skills, Summary ======= ADHD is likely to become increasingly important for primary care. To ensure that patient care is not compromised by the introduction of shared care protocols it is important that there is a clear dialogue between planners and healthcare professionals from both primary and secondary care. This is necessary to ensure mechanisms are in place for this care to be undertaken effectively and properly organised. Service delivery needs to be sensitive to patient need, acceptable to healthcare providers and tailored to their skills and finally must be adequately resourced. Authors\' contributions ======================= Both authors (AKT and AT) jointly initiated this paper, researched the topic and contributed to drafting and final preparation of this article. Declaration of competing interests ================================== AT (Anita Thapar) has organised academic meetings that have received sponsorship from Celltech (manufacturers of Equasym) and Janssen Cilag (manufacturers of Concerta). Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ We wish to thank Professor Eric Taylor for his helpful comments on this paper and Professors\' Richard Harrington and Martin Roland for reviewing the manuscript.

Background ========== Proteorhodopsin is a 249-amino acid membrane protein native to several uncultured species of γ-proteobacteria, which are a component of marine plankton \[[@B1]\]. Addition of retinal to *E. coli* expressing pR was shown to cause a reddish coloration of the bacteria with an absorption maximum near 520 nm. The pR contained in the bacterial membranes was shown to act as a light-activated proton pump, but only when retinal is present. Time-resolved UV/vis studies at pH 8 also revealed that the protein undergoes a photocycle, similar to that of wild type bacteriorhodopsin, but with a predominance of the O intermediate instead of M. The bR photocycle has been characterized by spectroscopic methods as having six principal photointermediates: bR, K, L, M, N and O. Each intermediate has a distinct absorbance maximum; the most studied are bR (570 nm), M (412 nm), and O (640 nm) since these are the ones that can be produced in the highest concentration at physiological pH values. Monitoring of the absorbance at individual wavelengths after photoexcitation is used to determine the relative concentrations and decay times of each of these photointermediates. The L → M transition in bR is characterized by the deprotonation of the Schiff base to Asp-85, producing the distinctive 412 nm absorbance maximum of M, and by so-called fast proton release, the ejection of a proton from a different (unknown) residue into the external medium on the \~10--100 μs time scale, depending on pH. Reprotonation of the Schiff base from Asp-96 occurs during the M → N transition with an absorbance maximum of 560 nm \[[@B2]\]. The N → O transition involves the reprotonation of the Asp-96 from the cytoplasmic space. Like bR, pR consists of seven transmembrane α-helices that include in the membrane interior all of the residues conserved among archaeal rhodopsin proton pumps. In particular, analogues of Asp-85, Asp96, Arg-82, and Lys-216 of bR are present in pR. Conspicuously absent are analogues for Glu-194 and Glu-204 of bR. The latter, as well as Arg-82, have been implicated in fast proton release. In particular, mutagenesis of Glu-194 or Glu-204 in bR results in loss of fast proton release \[[@B3],[@B4]\]. The absence of homologs for these residues in pR leaves open the question of whether it carries out fast H^+^ release. Experiments described here demonstrate that pR does indeed undergo fast H^+^ release, at least under elevated pH conditions that resemble somewhat those of the γ-proteobacteria\'s native open ocean environment. We also demonstrate that there is a post-translational modification of at least one of the three native cysteines when pR is expressed in *E. coli.* Both of these discoveries were made possible through purification methods for pR described herein. Results ======= Purification ------------ PR was obtained in 85% purity, assuming that values of ε~280~ and ε~546~ for pR are the same as for bR solubilized in DMPC/cholate/SDS mixtures at pH 8 (ε~280~= 7.85 × 10^4^ cm^-1^ M^-1^ and ε~551~ = 4.8 × 10^4^ cm^-1^ M^-1^) \[[@B5]\]. This assumption is actually expected to underestimate the purity of pR produced, by up to \~20%, since the pR we expressed has 10 tryptophan and 14 tyrosine residues, as compared to 8 tryptophans and 11 tyrosines in bR from H. salinarum. The absorbance of contaminant proteins was assumed to be 1.1 for a 1 mg/mL solution. By using these assumptions, the relative concentrations of pR and other proteins can be determined from the absorbance spectra of the various fractions (fig. [1](#F1){ref-type="fig"}). The resulting purity values correlate well with those Coomasie-stained SDS-PAGE gels (see below). The OG extract of cholate-washed membrane pellets starts out at a pR content of 7% total protein (w/w). The Phenylsepharose column increases the purity level to 24%, with approximately 5% loss. The final purification step by hydroxylapatite column chromatography produces pR with \~85% purity and a further loss of \~60%, i.e. the overall yield of the two column procedure was \~30%. {#F1} Polyacrylamide gel electrophoresis ---------------------------------- Relative to protein standards, the apparent molecular weight of bR is 25,000 while the apparent molecular weights of pR-wt and pR-TCM are 36,000 and 31,000, respectively (fig. [2](#F2){ref-type="fig"}, lanes E and C, respectively). SDS-PAGE (fig. [2](#F2){ref-type="fig"}) also confirms the estimates of purity level based on the assumed ε~280~/ε~546~ ratio identical with that of detergent solubilized bR. Interestingly, the pR appears to be a doublet band whose relative concentrations remain almost unchanged during purification. This doublet is also present in the less-purified sample of pR-TCM, with both bands shifted down by approximately the same amount (fig. [2](#F2){ref-type="fig"}, lane C). {ref-type="fig"}. Lane E contains the hydroxylapatite-purified pR wild type protein, corresponding to spectrum C in fig. [1](#F1){ref-type="fig"}.](1472-6793-2-5-2){#F2} Subsequent SDS-PAGE analysis of pR samples that had been stored for periods of time up to several months indicate that after sitting for several weeks in octylglucoside solution at 4°C, the largest post-translational modification on wild-type pR is eliminated -- presumably hydrolysed off of the cysteine(s) -- leaving only a 31,000-MW band indistinguishable from that seen for pR-TCM (data not shown). Furthermore, after boiling for several min in gel loading solution, this cleaved wild-type protein, as well as the TCM, both give an extra artifactual band near 36,000 dalton. The latter band, a singlet, is coincidentally at almost the same apparent MW as the doublet from the uncleaved post-translationally-modified wild type pR (fig. [2](#F2){ref-type="fig"}, lane E). These potential artifacts should be taken into consideration in any attempt to reproduce the results in Fig. [2](#F2){ref-type="fig"}. Photocycle kinetics and flash-induced proton concentration changes ------------------------------------------------------------------ Photocycle kinetics were measured at 400, 500. and 580 nm in the presence of the short-chain lipid DHPC. This lipid does not support the formation of closed bilayer vesicles, but rather forms micelles like a detergent. The time-resolved measurements showed no positive 400-nm absorbance signals at pH 8.0 or lower (Fig. [3](#F3){ref-type="fig"}). This is somewhat in disagreement with Béjà et al \[[@B1]\], who detected small 400-nm transient absorbance increases upon photolysis at pH 8.0. However, we observed a transient 400-nm absorbance increase at an elevated pH of 9.5 (fig. [3](#F3){ref-type="fig"}). {#F3} At pH 9.5 in the presence of DHPC, and observing transient changes at 500 nm (fig. [4](#F4){ref-type="fig"}), pR undergoes a 2-phase decay after the initial unresolved absorbance decrease. Multiexponential fits show that the first decay phase has a time constant of 4 μs, in good agreement with the 4-μs rise time of the 400 nm signal (Fig. [4](#F4){ref-type="fig"}). The amplitude of this decay represents about 80% of the initial absorbance depletion. The second phase of the 500 nm absorbance decay occurs with a substantially slower time constant of 0.5 s, returning the remaining 20% of initial absorbance change. The slowest decay components of the positive 400-nm signal and the negative 500-nm signal follow similar kinetics, although the amplitudes of these components differ by a factor of 3. At pH 9.5, the 580 nm trace has no significant positive values indicative of an O-like intermediate, although, in agreement with earlier measurements \[[@B1]\], at lower pH values a red-shifted transient is the predominant positive absorbance signal (data not shown). {ref-type="fig"}. The 500-nm trace shows the depletion signal of pR at the earliest times, and then the time course of the N intermediate as well as return of the pR resting state. The 580-nm trace is indicative of an O-like intermediate. The conditions are 1% DHPC, 100 mM NaCl, pH9.5 at 22°C. The laser excitation is as in fig. [3](#F3){ref-type="fig"}.](1472-6793-2-5-4){#F4} Figure [5](#F5){ref-type="fig"} shows a different type of time-resolved measurement, probing not the pR chromophore, but rather pH changes in the protein environment. Proton concentration changes in the aqueous bulk phase were measured with the pH sensitive dye cresol red, which has a pK~a~ of 8.2--8.5. The bottom trace in Figure [5](#F5){ref-type="fig"} shows the absorbance change of the indicator during the pR photocycle. The negative signal is indicative of a pH decrease, corresponding to transient H^+^ release from the protein into the solution. The best-fit time constant for the release phase is 6 μs. The positive 400 nm trace in fig. [5](#F5){ref-type="fig"} (reproduced from fig. [3](#F3){ref-type="fig"}) shows that the proton release and uptake follow kinetics very similar to the apparent formation and decay of M, as is typically seen in bR near neutral pH \[[@B6],[@B7],[@B21]\]. However, no proton release signal could be observed for pR at pH 6 or 8 (data not shown). {ref-type="fig"}.](1472-6793-2-5-5){#F5} Discussion ========== Purification of pR ------------------ Meaningful comparison of intrinsic physiological properties of pR and bR depends on the purification of pR. The *E. coli* expression system can easily be used to prepare pR at 85% purity with similar or less effort and time than required for bR (purple membrane) production from S9 *H. salinarum.* The initial purity level of the OG-solubilized cholate-extracted membrane is about 7% pR by weight. Phenylsepharose™ column chromatography separates proteins on the basis of hydrophobicity, and has been used previously in hR purification \[[@B8]-[@B10]\]. As an initial purification step, the Phenylsepharose™ column achieved a substantial increase in protein purity to 25%, along with removal of most of the lipid. The final purification step, utilizing a hydroxylapatite column, has been previously used with rhodopsin \[[@B11]\]. This column proved to be more efficient as a final step in the purification than as a preliminary one, because large amounts of contaminant protein tended to slow the flow rate drastically. This step of the purification yielded an increase in protein purity to \>85%. The overall yield of pR from membrane through the column purifications is \~30%. Most of this loss, \~65%, occurs during the hydroxylapatite column. Molecular weight differences ---------------------------- Béjà et al. reported a molecular weight for wild type pR of 27 kDa based on the predicted amino acid sequence of the protein \[[@B1]\]. Its 249 amino acids barely exceed the 248 of mature bR, which has a molecular weight of 26,000. However, we observed a significantly higher apparent molecular weight (\~36,000) for wild type pR on SDS-PAGE gels (Fig. [2](#F2){ref-type="fig"}, lanes D and E). Post-translational modification of pR must almost certainly account for some of the observed molecular weight difference between pR and bR. Lipids or sugars covalently bound to the protein surface would not be removed during the purification procedure and could cause a higher apparent molecular weight. Cysteines are frequently a site of lipid association with membrane proteins, (e.g. mammalian rhodopsin, which has two palmitoyl molecules attached to cysteine residues). Therefore we compared the SDS-PAGE gel mobility of wild-type pR to that of a cys-less mutant (pR-TCM). The elimination of the three possible sulfhydryl attachment sites lowers the apparent molecular weight of the pR-TCM by 5,000 (Fig. [2](#F2){ref-type="fig"}, lane C). From this, we conclude that at least one of the three cysteines in pR is probably modified post-translationally. However, this by itself does not account fully for the anomalous mobility of pR on SDS-PAGE gels, because pR-TCM is still approximately 5,000 higher in molecular weight than bR according to SDS-PAGE (fig. [2](#F2){ref-type="fig"}, lanes A & C). Only \~2,600 of this can be accounted for by the V5 eiptope and poly-histidine tail that are appended to the C-terminus of pR by the pBAD-TOPO expression vector that we used \[[@B1]\]. There is undoubtedly a further post-translational modification of unknown nature. Spectral comparisons -------------------- Béjà et al. reported an absorbance maximum for pR of 520 nm in *E. coli* membranes, using a difference bleaching technique to remove interfering absorbance bands from other membrane components in the impure pR sample \[[@B1]\]. We confirmed this result using crude *E. coli* membranes without detergent present (data not shown). However, we observed an absorbance maximum of 546 ± 5 nm for pR in OG at pH 7 at all stages of purification (Fig. [1](#F1){ref-type="fig"}). Small blue shifts were observed for pR samples in OG when measured at pH 8 and 9.5 (8 and 16 nm, respectively; data not shown). For the pR samples reconstituted in DHPC, which were used for flash photolysis experiments, the absorbance maxima were similar to those measured in OG (spectra not shown). A chromophore absorption maximum near 540 nm was also obtained by using difference spectroscopy of pR in crude *E. coli* membranes solubilized in OG (spectrum not shown). However, for pR measured directly in crude *E. coli* membranes, i.e. not solubilized in OG, we obtained the same value (520 nm) as reported previously \[[@B1]\]. Solubilization in detergent presumably leads to structural distortions of the native protein conformation, and therefore a change in the absorbance properties of the chromophore. However, the direction that λ~max~ for pR would have to shift upon solubilization in OG is inconsistent with the pattern for bR, whose λ~max~ decreases when it is solubilized in OG \[[@B12]\]. Furthermore, pR in OG showed resonance Raman spectra (D. Dunmire, R. A. Krebs and M. S. Braiman, unpublished data) indicative of a chromophore structure very close to native light-adapted (i.e. all-trans) bR in purple membrane. However, there is one major difference: pR in OG exhibits an upshifted, doublet C=N Schiff base band consisting of two components of nearly-equal intensity. These appear to correspond to the presence of two distinct subpopulations of pR, at least when expressed in *E. coli* and solubilized in OG micelles. The different values of λ~max~ for pR in membrane state \[[@B1]\] and OG solution might be related to the presence of these multiple subpopulations, but this connection remains unclear. Principal photointermediates of pR ---------------------------------- Of the six principal photointermediates present in bR, four can be discerned from the time-dependent visible absorbance traces from pR in fig. [4](#F4){ref-type="fig"}, along with previously published time traces at 600 nm \[[@B1]\]: the resting state (pR), M, N and O. The resting state, with an absorbance maximum of 546 nm (see above), provides the baseline spectrum for the difference time courses reported. The M intermediate of bR has an unprotonated Schiff base group, giving rise to a blue-shifted λ~max~ (400 nm). Likewise in pR, an increase in the 400 nm absorbance should indicate formation of a deprotonated Schiff base, and therefore the presence of an M-like intermediate. Interpretation of the 500 nm time course trace is more complicated. As in bR, it likely involves decay of M to N, as well as from N back to bR. The absorbance maximum of the N intermediate in bR is 560 nm, corresponding to a protonated Schiff base. This is not very different from λ~max~ for the resting state of pR (546 nm). The likely spectral overlap between pR and its own postulated N photoproduct complicates the determination of amounts of each that are present. The slow (\~500 ms) decay observed in both the 400 nm and 500 nm time courses indicates an equilibrium between the M and N intermediates that remains until pR returns to its initial resting state. Béjà et al. reported a strong positive 580-nm transient absorbance increase in suspensions of membranes prepared from *E. coli* expressing pR \[[@B1]\]. This positive absorbance difference is indicative of O intermediate formation. We looked for its presence in partially-purified samples reconstituted in DHPC in the pH range 6--11. (Data are shown only at pH 9.5; see Fig. [4](#F4){ref-type="fig"}). Our observations at pH 9.5 do not show any evidence of O intermediate formation (Fig. [4](#F4){ref-type="fig"}). Only at lower pH values was a clear positive 580-nm absorption change observed (data not shown). This is in agreement with results on bR, for which O formation is also enhanced at lower pH values, and becomes small or nearly unobservable in the alkaline range. Dependence of the M intermediate on pH -------------------------------------- Deprotonation of the Schiff base linkage of the retinal and Lys-216 is dependent on its pK~a~ which changes between photointermediate states. The Schiff base readily undergoes deprotonation in the M intermediate. However, no M intermediate formation occurs below a pH of \~9 (figure [3](#F3){ref-type="fig"}); instead the O intermediate predominates \[[@B1]\]. Predominance of O at lower pH values is also observed in bR. However, in bR the M intermediate is detectable at low and high pHs, but has a longer lifetime at higher pH due to a long-lived equilibrium between M and another intermediate, N \[[@B13]\]. It seems likely that an N intermediate of pR is similarly in equilibrium with its M intermediate, based on the fact that the transient positive (400 nm) absorbance increase is smaller than the negative 500 nm bleach, and the time course at 500 nm shows a partial return to baseline on a timescale of \~50 μs. The difference in pH dependence between the pR and bR photocycles can likely be attributed to differences in the microenvironment of the Schiff base, and is perhaps related to absence of Glu194/204 in pR. These adaptive differences presumably optimize the proteins to operate at maximum efficiency in the niches that their respective organisms occupy. In the case of the λ-proteobacteria, which in the open ocean (pH 7.8--8.0) occupy a signficantly more alkaline environment than halobacteria, perhaps the proton-release group is simply not under any evolutionary pressure to be capable of deprotonating in the M state at neutral pH. In this view, the principal role of E194 and E204 in bR may to modulate the pK~a~ of the H^+^-release group in the M state to a value lower than the pH of the organism\'s external environment. Fast proton release in pR ------------------------- Under the same conditions where M is observed (pH 9.5 and in 1% DHPC), pR undergoes fast proton release during its photocycle (Fig. [5](#F5){ref-type="fig"}). The pH indicator dye Cresol Red was used to detect pH changes in the bulk aqueous phase. These turn out to be similar to those observed for bR in the pH range 5.5--10. After photoexcitation, pR (like bR, presumably) ejects a H^+^ from a residue near its extracellular surface decreasing the pH of the solution. When the N → O transition takes place in bR, H^+^ is taken up from the medium, raising the pH once again. The H^+^ signals from pR measured with Cresol Red occur on a time scale similar to that assignable to M and N decay, returning to baseline about 1 s after photolysis. There is a clear kinetic correlation between M (and/or N) intermediate formation and fast H^+^ release in pR. The linkage between these two phenomena is further supported by the observation that neither a transient 400-nm absorbance increase, nor fast H^+^ release, is shown to occur at pH 8.0 and below. Nor is either observed in the absence of a reconstituting lipid (DHPC in these experiments). In bR, the ejected proton is thought to originate from a triad of amino acids, R82-E194-E204. However, in pR a homolog of only one of these three residues (the arginine) is present. This raises doubts about previous conclusions regarding the specific roles of these 3 residues in fast H^+^ proton release, in both pR and bR. In particular, the apparently obligatory roles of E204 and E194 in fast H^+^ release in bR are not matched in pR. Therefore, even in bR it is less likely that these groups themselves change protonation state between bR and M to provide the H^+^ released to the bulk medium. Instead, it now seems more likely that E204 and E194 merely help to lower the pK~a~ of the H^+^ release group from above 8, the apparent value for pR, into the vicinity of 6 for bR. It also seems very unlikely that the specific structural configuration of 2 carboxylic acid groups and arginine in bR could be conserved in pR, even if, as suggested previously \[[@B1]\], other surface carboxylic acids in pR could substitute in some ways for the roles of E194 and E204 in bR. The only way that the H^+^ release mechanism can be strongly conserved between bR and pR is if arginine itself serves as the principal donor group for fast H^+^ release in both, with nearby residues (such as E194 and E204) merely modulating the pK~a~ of the arginine in the M intermediate. However, it remains unclear how the pK~a~ of Arg-82 in bR could be made sufficiently low in its M intermediate to serve as the H^+^ release group at pH values down to 6.0. Alternatively, it is possible that the fast H^+^ release we observe from pR at pH 9.5 may differ from that in bR. One possibility is that in pR, the released proton could come directly from the chromophore counterion, Asp-97. This would be consistent with a proposed mechanism for fast H^+^ release that has been observed above pH 10 in the bR mutant E194Q \[[@B14]\]. In this mutant, Asp-85 was detected by low-temperature infrared difference spectroscopy to be deprotonated only in the N intermediate, and not in M \[[@B15]\]. It is not clear yet whether Asp-85 deprotonation in an N-like state could account for the proton-release kinetics of pR (Fig. [5](#F5){ref-type="fig"}), or whether the M intermediate itself might have a partially-deprotonated Asp-85. The reason for the requirement of DHPC in M intermediate formation and fast proton release is unclear. Delipidated bR in octylglucoside is fully capable of M formation and presumably proton release, although with altered kinetics \[[@B12],[@B16]\]. The requirement for pR to be in lipid to show fast H^+^ release and M formation stems either from a protein/lipid interaction needed to establish a stable, active tertiary structure, or from the need for the phosphate group in DHPC to act as a proton release group. The latter seems unlikely due to the DHPC molecule being zwitterionic at pH 9.5, with no proton on the trimethyl-modified nitrogen of the choline. Hence, the DHPC most likely interacts with the protein to effect minor structural changes needed to place the active site residues in their functional configuration. Conclusions =========== A comparison of the primary sequences of pR and bR at first glance seems as to preclude fast H^+^ release as part of the proton-pumping mechanism of pR due to the absence of residues analogous to Glu194 and Glu 204 of bR. However, fast H^+^ release is indeed observed in pR under conditions where an M intermediate is formed. Glu194 and Glu 204 in bR play a role in fast H^+^ release that is apparently not required for the mechanism of the bR family of proton transporters. It is therefore necessary to conclude that either the H^+^-release groups in pR and bR are non-homologous surface carboxylic acid residues (as suggested previously \[[@B1]\]), or else that a conserved non-carboxylic acid residue, i.e. Arg82/94 or Asp85/97, is the H^+^-release group in pR. The higher pH requirement for the M intermediate of pR presumably corresponds with adaptation to the more alkaline oceanic environment in which the γ-proteobacteria are found. The necessity of reconstituting pR with some lipid before it is capable of photocycling shows that the presence of lipids facilitates pR in assuming its fully active structure. *E.-coli*-expressed pR has post-translational modifications, including \~4000 daltons of substituents at one or more of its three cysteines. Such post-translational modifications might also play a role in explaining the different physiological properties of pR and bR. Materials & methods =================== Protein expression and detergent extraction ------------------------------------------- Proteorhodopsin was expressed by *E. coli* strain UT5600 containing an additional plasmid encoding for the first-reported pR gene (accession \#AF279106, obtained from the uncultured proteobacterium EBAC31A08 clone BAC \[[@B1]\]) with an Ara promoter and ampicillin resistance (kindly provided by O. Béjà). Single colonies were selected and grown overnight in LB/amp media (200 ml, 37°C, 300 rpm). This culture was then diluted 10× into several 500-ml cultures. After a further 2 h incubation in the shaker bath, a stock solution of 20% L-arabinose was added, to give a final concentration of 0.2% L-arabinose. This culture was then incubated for 4 h (37°C, 300 rpm). The cells (\~20 ml wet volume) were then collected by centrifugation (6000 rpm × 30 min) and washed 3× with 100 mM HEPES, pH 7.1 (buffer A). The cells were then resuspended in buffer A and incubated at 4°C with 50 μg all-trans-retinal (added as a concentrated ethanol solution) for 3 h. The cells were collected by centrifugation (6000 rpm × 30 min), then resuspended in 60 mL buffer A containing 0.3 mg/mL lysozyme, and stirred for 4 h at room temperature. The cells were again collected by centrifugation (6000 rpm × 30 min), then lysed with 20 ml of 20% sodium cholate, pH 7.1 (30 min., 4°C). The cells were centrifuged again (6000 rpm × 30 min), and the supernatants collected. After extracting 3× more with the same cholate solution, the pooled supernatants were diluted 10× with buffer A and centrifuged at 180,000 g for 45 min to collect the membrane pellet. This cholate-washed membrane pellet was then further extracted 3× with 3.0% β-octyl-D-glucoside (OG) in buffer A (30 min, with stirring, 4°C). The pooled supernatants, containing OG-solubilized pR (\~15 mg), were then diluted 6× with buffer A. Column purification ------------------- The diluted OG-solubilized membrane extract (10 mg in 300 mL total volume of 0.5% OG) was loaded on a 25 × 1 cm column containing Phenylsepharose™ (6 fast flow high sub; Amersham Pharmacia Biotech). The column was eluted with a 0.5%-2.0% OG gradient in buffer A (300 mL total volume, 0.5 mL flow rate). The pR eluted at an OG concentration of 1.5--2.0%. Fractions having an A~280~/A~546~ ratio of 4.0 or lower were pooled (9.5 mg pR recovered in all) and concentrated using Vivaspin™ 20 concentrators having a 5000 MW cutoff (Vivascience, Westford, MA). A portion of the Phenylsepharose™-purified pR (1.5 mg) was diluted to an OG concentration of 0.5% with 0.5 M KCl, 100 mM acetate. It was then loaded on a 10 cm × 1 cm hydroxylapatite (BioGel HTP, BioRad) column and eluted under pressure with a 0--600 mM phosphate gradient (200 mL total volume, flow rate 0.5 ml/min). Fractions with an A~280~/A~546~ ratio of 2.5 or lower were pooled and concentrated for subsequent experiments (0.5 mg). Mutagenesis ----------- Methodology for the site-directed mutagenesis of pR is discussed in detail elsewhere (R. Parthasarathy, T. Caterino, R.A. Krebs, M.S. Braiman, manuscript in preparation). The triple cysteine mutant (pR-TCM) has all three of its native cysteines (Cys-107, Cys-156, and Cys-175) replaced with serines, and was prepared using the same *E. coli* expression system and purification methods as the wild type. Polyacrylamide gel electrophoresis ---------------------------------- A 12% discontinuous SDS/polyacrylamide gel was used for molecular weight and purity analysis \[[@B17]\]. Flash photolysis ---------------- Time-resolved UV/vis spectroscopy methods were as described previously \[[@B18]\]. A Phenylsepharose™-purified pR sample was reconstituted into mixed micelles containing 1,2-diheptanoyl-SN-glycero-3-phosphocholine (DHPC), by adding a 1% solution of the short-chain lipid and then removing most of the detergent on a Sephadex G-25 column equilibrated with 1% DHPC in 100 mM NaCl. Proton release and uptake in the aqueous bulk medium were detected from the pR-containing micelles suspended in 1% DHPC, 100 mM NaCl, with 45 μM Cresol Red pH indicator dye. Flash-induced absorbance changes at 580 nm of samples with and without the Cresol Red were subtracted to determine the transient signals due to proton concentration changes. Photoexcitations were performed with 10-ns laser pulses of 3--6 mJ at 500 nm. The time courses in Figs. [3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"} are an average of 40 cycles with the exception of the Cresol Red experiments averaging 100 cycles (Fig. [5](#F5){ref-type="fig"}, bottom trace) \[[@B19]-[@B21]\]. Abbreviations ============= pR, proteorhodopsin; bR, bacteriorhodopsin; OG, β-octyl-D-glucoside; SDS, sodium dodecylsulfate; PAGE, polyacrylamide gel electrophoresis; pR-TCM, pR triple cysteine mutant \[C(107,156,175)S\]; DHPC, 1,2-diheptanoyl-SN-glycero-3-phosphocholine; DMPC, dimyristoylphosphatidylcholine; HEPES, N-2-hydroxyethylpiperazine-N\'-2-ethanesulfonic acid Acknowledgements ================ This work was supported by Syracuse University, and by a grant of the Deutsche Forschungsgemeinschaft (Sfb 449-TPA5 to U. Alexiev and M. P. Heyn). A.-M. DeVita was supported by an NSF-REU award to Syracuse University Chemistry Department.

Background ========== It is hypothesized that there are 1.5 million fungal species on earth, of which only about 70,000 have been described \[[@B1]\]. Thus nearly 1.43 million remain undescribed. It will be essential for the study of fungal evolution to determine the phylogenetic positions of undescribed fungi present in diverse environments. Many fungi have parasitic or symbiotic relationships to other organisms. This can make it difficult, if not impossible, to separate fungi from such organisms. On the other hand, PCR can be used to amplify the DNA fragments without isolation and cultivation. The genes used most widely in fungal phylogenetic studies are rRNA genes, elongation factor genes, tublin genes, and other universally conserved eukaryotic sequences. These genes are so conserved among all eukaryotes that they often result in artifacts when amplified by PCR. Therefore, we conclude that PCR primers that are unique to fungal genomes will be extremely useful for PCR-based phylogenetic study of fungi. In addition, although the comparison of rDNA and other genes among species is a powerful tool to show the phylogenetic relationships among fungi, no individual gene can answer all questions about fungal evolutionary relationships \[[@B2]-[@B4]\]. Many genes are not useful for quantitative analysis because they are multi-copy and copy-number is different among various fungi. Here we investigate the aminoadipate reductase gene. Most mycologists previously believed that lysine biosynthesis through the 2-aminoadipate pathway was a fungal-specific characteristic. However, some prokaryotes also synthesize lysine through the 2-aminoadipate pathway \[[@B5]-[@B7]\]. A comparison between the prokaryotic and fungal lysine biosynthetic pathways reveals that the synthesis of 2-aminoadipate from 2-oxoglutarate proceeds in the same way but the fungal process to synthesize lysine from 2-aminoadipate is different from the prokaryotic one \[[@B5],[@B8]\]. Therefore, prokaryotes have no aminoadipate reductase gene; on the other hand, the aminoadipate reductase is a key enzyme in the evolution of fungal lysine biosynthesis \[[@B8]\]. This enzyme consists of large and small subunits \[[@B9]\], and functions only in fungal cells. As far as we know, animals and plants have no homologue of this enzyme-coding gene. In the lysine biosynthetic pathway of the ascomycete yeast *Saccharomyces cerevisiae*, a complex of LYS2 (1392 aa) and LYS5 (272 aa) serves as aminoadipate reductase, which converts 2-aminoadipate into 2-aminoadipate 6-semialdehyde via an adenosylated derivative (Fig. [1](#F1){ref-type="fig"}). Before this report, seven coding regions of the large subunit of the aminoadipate reductase were deposited in the international DNA/protein database \[[@B10]-[@B15]\]. The *Saccharomyces cerevisiae lys2* gene is a single-copy gene. Comparisons of single-copy genes of large size can provide good phylogenetic resolution and offer advantages over multi-copy genes \[[@B16]\]. According to the seven ascomycetous aminoadipate reductases, fungal the coding regions of *lys2* genes are more than 4,000 nucleotides-long. {#F1} Results and Discussion ====================== First we performed a homology search using BLAST \[[@B17]\] with the given parameter values on the DNA data bank of Japan (DDBJ). We searched the homologous sequence of the *S. cerevisiae lys2* gene for EST database of *Arabidopsis thaliana, Caenorhabditis elegans, Dictyostelium discoideum, Drosophila melanogaster, Glycine max, Homo sapiens, Mus musculus, Oryza sativa, Xenopus laevis*, and *Zea mays*. As a result, we obtained no significant sequence. This result supports that the aminiadipate reductase gene is a fungal-specific gene. We determined the DNA sequence of 1,058-bp from *S. complicata* and 1,079-bp from *A. awamori*. These sequences have been deposited in the DDBJ under accession numbers AB076076 for *S. complicata* and AB076077 for *A. awamori*. An alignment of the nine ascomycetes was created using the program CLUSTAL W \[[@B18]\]. The alignment in this study is available on request using e-mail. We considered 325 amino acid sites excluding indels and PCR primers-sites. A maximum parsimonious phylogenetic tree was constructed using a Branch-and-Bound algorithm method, in MEGA version 2.1 \[[@B19]\], with 1,000 bootstrap replicates. We obtained equally two most parsimonious trees. The consensus parsimony tree (Fig. [2a](#F2){ref-type="fig"}) shows three major ascomycete lineages; the archiascomycete, the euascomycete, and the hemiascomycete. This is consistent with current ascomycete systematics based on other gene sequences. The bootstrap analysis indicated 96% and 90% support values for monophyletic lineages of euascomycetes and hemiascomycetes, respectively. However, it indicated only 57% support for the monophyly of the archiascomycetes. ![a) The bootstrap consensus tree of the two most parsimonious trees based on the amino acid sequences of the aminoadipate reductase. The most parsimonious trees using the Branch-and-Bound algorithm of MEGA version 2.1 \[[@B19]\] with 1,000 bootstrap analyses. b) The maximum likelihood phylogenetic relationships. This analysis was performed using PAML \[[@B20]\], version 3.1. The model of amino acid substitution by Whelan and Goldman \[[@B21]\] was used.](1471-2148-2-6-2){#F2} Maximum likelihood analysis was performed using PAML \[[@B20]\], version 3.1. In this analysis, we used the model of amino acid substitution by Whelan and Goldman \[[@B21]\]. The best tree (lnL: -4292.79) is shown in Fig. [2b](#F2){ref-type="fig"}. The ML tree is largely consistent with the most parsimonious tree, and shows the archiascomycetes to be monophyletic. The anamorphic yeast *Saitoella complicata* has unique morphological and chemotaxonomic characteristics, and its 18S rDNA shows affinity to those of archiascomycetes \[[@B3],[@B22]-[@B24]\]. As far as we know, this interesting and important yeast has not been analyzed in a molecular evolutionary study, aside from rDNA comparison. Here we determined the DNA fragment from the aminoadipate reductase gene and also show that the deduced amino acid sequence has an affinity for the archiascomycete *Schizosacchaormyces pombe* in phylogenetic analyses. The phylogenetic tree (Figs. [2a, 2b](#F2){ref-type="fig"}) clearly indicated that the budding yeast *S. complicata* was far from the other budding yeasts (the hemiascomycete lineage). The phylogenetic position of the black-koji mold *Aspergillus awamori* suggests a close relationship to *Penicillium chrysogenum* (99% bootstrap support). This is an expected result. The sequence similarity between *A. awamori* and *P. chrysogenum* is 86% in amino acid comparison and 76% in DNA comparison. This sequence-difference in the aminoadipate reductase comparison is greater than that in shown from rDNA comparison. Conclusions =========== The PCR primers designed in this study were shown to be effective for amplifying the aminoadipate reductase gene from divergent ascomycetes. In addition, this region of the PCR product is useful for clarifying the ascomycete phylogeny. We believe that this region would be a powerful tool for fungal ecological and evolutionary studies. Materials and Methods ===================== In this study we used *Aspergillus awamori* IAM 2112, *Saitoella complicata* IAM 12964. The genomic DNAs were isolated using a DNeasy Plant Mini Kit (QIAGEN, Valencia, CA). Multiple alignment was created among the seven known ascomycetous aminoadipate reductases (*Acremonium chrysogenum*, AJ261064; *Candida albicans*, U58133; *Neurospora crassa*, AL389890; *Penicillium chrysogenum*, Y13967; *Pichia sorbitophila*, AJ288950; *Saccharomyces cerevisiae*, M36287; *Schizosaccharomyces pombe*, AL353014) using the program CLUSTAL W \[[@B17]\]. According to the multiple alignment, we found only two conserved regions for the PCR-primers. Based on the conserved amino acid sequences, two primers were designed 5\'-GGNATHGCNCAYGAYCCNRTNCA-3\' and 5\'-GGYTTRTCNAYYTTNCCRTTNGGRTT-3\'. The amplification was carried out under the following conditions: denaturation at 94°C for 5 min, 30 cycles of (94°C for 1 min, 57°C for 1 min, 72°C for 1 min), and a final extension at 72°C for 10 min. The PCR products were cloned using a PCR Cloning Kit (QIAGEN). Direct sequencing for the PCR products and sequencing for the several cloned plasmids were performed using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, CA). Acknowledgements ================ We thank the two anonymous reviewers for their helpful comments. The Ministry of Education, Culture, Sports, Science, and Technology of Japan supported this work.

Background ========== Publication bias represents an Achilles\' heel in meta-analyses. A particular form of publication bias is the so-called language bias, by which randomised controlled trials (RCTs) with greater estimates of effect size tend to be published in English rather than in the original authors\' native languages \[[@B1]\]. This bias is compounded by Anglophone databases and journals considerably under representing the totality of RCTs. For example, more than half of trials published in a Hungarian journal, identified through a cover-to-cover hand search, were not reported in MEDLINE \[[@B2]\]. Electronic searches of non-English databases may yet not suffice: a mono-lingual English searcher would be able to identify only 73% of the trials in the Spanish/Portuguese LILACS \[[@B3]\] and 70% of the German PSYNDEX \[[@B4]\]. Attempts are underway to remedy this situation. Two notable examples are the Medical Editors Trial Amnesty \[[@B5]\] and the Cochrane Controlled Trials Register (CCTR). The latter now contains more than 300,000 citations. Despite these enormous efforts we still are unsure if any database provides comprehensive coverage, particularly for non-English trials. We have recently established a comprehensive register of Japanese trials of psychotropic drugs. We therefore examined how successful the Cochrane psychiatry groups have been in identifying trials conducted in Japan and reported in Japanese by comparing the yields of the groups\' specialized registers against this recently created \"gold standard.\" Methods ======= Three of the authors (TI, AI, SN) have established a comprehensive register of Japanese RCTs of psychotropic drugs. First, we requested all the pharmaceutical companies marketing a psychotropic drug in Japan (n = 20) to provide them with the references concerning their RCTs with Japanese participants. Second, we carefully searched the on-line bibliographic database, JMEDICINE (Jan 1981--June 2000). This contains reports of medical literature in Japanese. Thirdly, we also collected references to RCTs through the \"reference search\" of the identified reports. Since on-line JMEDICINE is not available before 1980, and even the best electronic search is likely to miss a considerable proportion of reports of trials, these authors made considerable efforts to hand search psychiatric and medical journals likely to contain reports of psychotropics trials (n = 35). Abstracts of academic meetings have not been included; neither have RCTs undertaken in Japan but published in the English language journals. We then asked two Cochrane groups in mental health, namely the Cochrane Schizophrenia Group and the Depression, Anxiety and Neurosis Group, to search their specialized registers for reports of trials possibly conducted in Japan. We employed two ways to identify such trials. The first strategy was to use the search term \"Japan\*\" (\*is a wild card). Because we were not sure if all candidates contained a word \"Japan^\*^,\" we employed the second strategy in which we searched for all the authors\' names appearing in the Japanese register. Results ======= The Japanese register contained 56 reports of RCTs of antidepressants for depressed people but the CCDAN register contained 18, with an overlap of only nine. It was possible to retrieve five more studies as the CCDAN register contained six English language duplicate publications of the five original Japanese RCTs. The remaining three studies found in the CCDAN register but not in the Japanese database were English reports of RCTs conducted in Japan but reported only in the English language. Searching the CCDAN register would therefore have identified a maximum of only 25% (=14/56, 95%CI: 15 to 38%) of all Japanese reports of relevant RCTs. The CSG register contained 36 reports of RCTs of neuroleptics given to Japanese people with schizophrenia. The Japanese register contained 61. Of these, only six had been registered in the CSG register; 13 more studies were possible to retrieve as the CSG register contained 18 English language duplicate publications of original Japanese RCTs. The remaining 12 reports found in the CSG register but not in the Japanese register were reports of RCTs conducted in Japan but published only in the English journals. Searching the CSG register only would have identified 19 or only 31% ( = 19/61, 95%CI: 21 to 44%) of all relevant reports in the Japanese language. Discussion and conclusion ========================= The comprehensiveness of the collaborative databases is now deemed all the more important because doubts have been expressed as to effectiveness of traditional statistical tools to examine publication bias such as funnel plots \[[@B6]\]. However, searching the Cochrane group registers would have missed 69% to 75% of all Japanese reports of RCTs of antidepressants or neuroleptics. These results clearly demonstrated the limitations of the current Cochrane group registers, despite their extensive searching (information available from the CCDAN and CSG modules in the Cochrane Library). Similar, or worse, yields may be expected with RCTs conducted in other East Asian countries (Korea, China), and in other fields of medicine. Many systematic reviews, at least in the field of mental health, are therefore likely to present results grossly underestimating research output from Asia. What evidence there is suggests that this will result in a systematic over estimate of effect \[[@B1]\]. Efforts must be expended for non-Anglophone countries to prepare their own register of RCTs for all health interventions in humans. The Japanese database of psychiatry trials has now been merged with the Cochrane group registers. Competing interests =================== None declared. Authors\' contributions ======================= TAF, TI and CEA participated in the design of the study and in its coordination. TI, CEA, HM, AI and SN participated in the collection of the datasets and in their analyses. TAF and CEA drafted the manuscript and all the authors participated in the rewriting. All authors read and approved the final manuscript. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== *Helicobacter pylori* is believed to be the major aetiological factor in the development of non-cardia gastric adenocarcinoma. Large-scale epidemiological studies have confirmed a strong association between *H. pylori* infection and both cancer \[[@B1]-[@B3]\] and the earlier histological stages, atrophy and intestinal metaplasia \[[@B4],[@B5]\]; both of which increase the risk of later neoplastic transformation. Animal models have also demonstrated the importance of *H. pylori* in gastric carcinogenesis \[[@B6],[@B7]\]. Increased rates of proliferation of the gastric mucosa are typical in *H. pylori* infection \[[@B8]-[@B11]\], and hyperproliferation within the gastrointestinal tract appears to be a marker for later malignant change \[[@B12]\]. The cause of the increased rate of proliferation is not clear, but the increased rates reduce to normal with clearance of the infection \[[@B8],[@B13]\]. Although hypeprproliferation is typical *in vivo,* studies testing the effects of *H. pylori* or its products *in vitro* have shown conflicting results, with both enhanced \[[@B14],[@B15]\] and diminished \[[@B16]-[@B18]\] proliferation reported. It is possible that other components of the inflammatory response typical of *H. pylori* infected mucosa could be at least partly responsible for driving the increased cell proliferation. The pluripotent pro-inflammatory cytokine interleukin-1β has a central role in the pathogenesis of *H. pylori*-induced mucosal inflammation. IL-1β gene expression and protein production are increased in *H. pylori* infection and reduce with successful eradication \[[@B19],[@B20]\]. The presence of the IL-1β genotype polymorphism associated with enhanced IL-1β-production has been associated with a significant increased risk of gastric cancer and pre-cancerous lesions \[[@B21],[@B22]\]. Interleukin-1β is a potent inhibitor of gastric acid secretion and it is hypothesized that the enhanced IL-1β response alters the topography of the gastric infection and thus promotes inflammation and subsequent atrophy of the gastric corpus \[[@B23],[@B24]\]. The possibility that IL-1β itself drives the increased proliferation of gastric epithelial cells has not been fully investigated. Alteration of gastric proliferation by IL-1β might contribute to the carcinogenic process, in addition to effects on acid secretion. Therefore the direct effects of IL-1β on gastric epithelial proliferation have been assessed. The mitogen-activated protein kinase (MAPK) cascades are well-characterised pathways transducing signals from the cell surface to the nucleus. The family includes distinct subgroups; extracellular signal-related kinases (ERKs), c-Jun NH~2~-terminal kinases (JNKs) and p38 MAPK \[[@B25]\]. The ERKs are activated by a variety of extracellular stimuli, and mediate the pro-proliferative effects of a number of hormones and growth factors \[[@B26],[@B27]\]. Activation by phosphorylation of a dual specificity protein kinase (MAP kinase kinase (MAPKK)), (also known as MEK), allows it in turn to activate a family of serine-threonine protein kinases, known as the ERKs. The ERKs in turn phosphorylate numerous cellular proteins including transcription factors and thus have a central role in propagation of mitogenic signals. Accordingly the role of the MAP-kinase pathway in mediating the responses to IL-1β has been assessed. Methods ======= Cell culture ------------ The human AGS gastric carcinoma cell line was purchased from the European Collection of Animal Cell Cultures (Porton Down, UK). Cells were grown in monolayer culture in RPMI 1640 medium supplemented with 100 μg/ml penicillin, 100 μg/ml streptomycin, 100 μg/ml gentamicin, 2.5 μg/ml amphoteracin B and 10 % foetal calf serum. Cells were grown in 75 cm^2^ tissue culture flasks at 37°C in an atmosphere of 5% CO~2~ and 95% air and passaged every 5--7 days. Proliferation studies --------------------- \[^3^H\]thymidine incorporation. Cells were grown in media containing 10% foetal calf serum, plated into 24-well plates at 10^5^ cells/well and allowed to attach overnight. After washing with serum-free media, cells were incubated in serum free medium containing 0.2 mM unlabelled thymidine for 24 hours in the presence of increasing concentrations of IL-1β, IL-8 or GM-CSF. DNA synthesis was estimated by measurement of \[^3^H\]thymidine incorporation into the trichloroacetic acid (TCA) precipitable material \[[@B28]\]. \[^3^H\]thymidine (0.1 μCi/ml, 10 Ci/mmol) was added 2 hours before the end of a 24 hour treatment period. Cells were washed twice with serum-free medium to remove unincorporated \[^3^H\]thymidine, and DNA was precipitated with 5% TCA at 4°C for 15 minutes. The precipitates were then washed twice with 95% ethanol, dissolved in 1 ml of NaOH, and analysed by liquid scintillation counting. Results are expressed as percent control unstimulated \[^3^H\]thymidine incorporation (mean ± SD) of 4--6 different experiments, each performed in triplicate. For the detection of growth inhibition, cells were incubated with either the specific MEK inhibitor PD 98059 (25 μM)\[[@B29]\], IL-1 receptor antagonist (500 ng/ml) \[[@B30]\] or the neutralising antibodies, anti-GM-CSF (5 μg/ml) or anti-IL-8 (10 μg/ml). Inhibitors or antibodies were added 30 minutes prior to cytokines. Cell growth ----------- Total viable cell numbers were assessed by a modified MTT (3-\[4,5-dimethylthiazol-2-yl\]-2.5 diphenyl tetrazolim bromide assay) \[[@B31]\]. Cells were plated into 24-well plates in medium containing 10% foetal calf serum. After attachment overnight, the medium was changed to 1% foetal calf serum-supplemented medium and increasing concentrations of IL-1β were added. Cells were cultured for 48 hours and then the medium was removed and fresh RPMI 1640 medium containing 0.5 ng/ml MTT was added. Cells were incubated at 37°C for 3 hours. The medium was then removed and 0.04 M HCl in isopropanol was added to extract the reduced formazan product. The resulting optical density at 550 nm was determined. Chemicals and reagents ---------------------- Recombinant human IL-1β and IL-8 were purchased from Sigma (Poole, UK), recombinant human GM-CSF and IL-1 receptor antagonist, anti-GM-CSF and anti-IL-8 were from R and D systems (Abingdon, Uk). PD 95059 was from Calbiochem (Nottingham, UK). RPMI 1640 was from Gibco BRL (Paisley, UK) and all other reagents were from Sigma. Concentrations of inhibitors used were taken from manufacturers\' data and published data. The ability of 500 ng/ml IL-1RA to abolish 10 ng/ml IL-1β-stimulation of IL-8 secretion in AGS cells was confirmed. The efficacy of the anti-GM-CSF at 5 μg/ml antibody to abolish GM-CSF (1 ng/ml) induced AGS cell proliferation was confirmed. Statistics ---------- Cytokine-stimulated results where compared with control unstimulated cells on the same 24-well plate. Data were compared by one-way analysis of variance and Student\'s t-test to determine statistical significance. Each experiment as performed in triplicate on 4--6 occasions. Results are expressed as mean ± standard deviation. Differences with P values of \< 0.05 were considered significant. Results ======= Effect of IL-1β on \[^3^H\]thymidine incorporation -------------------------------------------------- Interleukin-1β caused a dose-dependent increase in DNA synthesis as measured by thymidine incorporation. As shown in figure [1](#F1){ref-type="fig"}, significant stimulation was seen with 1--100 ng/ml IL-1β. The maximal stimulation of 52 ± 6 % above control was seen with 10 ng/ml. The higher dose of 100 mg/ml was slightly less effective in stimulating proliferation. ![Effect of IL-1β on \[^3^H\]thymidine incorporation into gastric epithelial AGS cells Cells were treated with increasing concentrations of IL-1β for 24 hours and DNA synthesis assessed by \[^3^H\]thymidine incorporation. Results expressed as means ± standard deviation. ^\*^*P* \< 0.01 vs. control](1471-230X-2-7-1){#F1} Effect of IL-1β on cell number ------------------------------ The increase in DNA synthesis by IL-1β was translated into an absolute increase in viable cell numbers. As shown in figure [2](#F2){ref-type="fig"}, IL-1β increased cell numbers in a dose-dependent manner similar to the effects on \[^3^H\]thymidine incorporation. The maximal stimulation was again seen at 10 ng/ml of IL-1β, which produced a 22 ± 5 % increase in total cell number. {#F2} Effects of cytokine inhibition or receptor antagonism on IL-1β-stimulation of proliferation ------------------------------------------------------------------------------------------- Pretreatment of the cells with interleukin-1 receptor antagonist abolished the stimulatory effects of IL-1β on \[^3^H\]thymidine incorporation (Figure [3](#F3){ref-type="fig"}). Previous studies have shown that IL-1β can activate gastric epithelial cells, including AGS cells, to secrete other cytokines, particularly interleukin-8 and granulocyte-macrophage colony stimulating factor (GM-CSF)\[[@B31],[@B32]\]. Therefore further studies were undertaken to assess if the stimulatory actions of IL-1β were mediated by either of these two cytokines. Neutralising antibodies to either IL-8 or GM-CSF had no effect on unstimulated \[^3^H\]thymidine incorporation. Neutralisation of IL-8 had no effect on IL-1β-stimuated growth but the anti-GM-CSF antibody reduced IL-1β-stimulated proliferation by 31 ± 4 % (*P* \< 0.01) (figure [3](#F3){ref-type="fig"}). ![Effect of cytokine inhibition on IL-1β-stimulation of \[^3^H\]thymidine into gastric epithelial cells AGS cells were treated with 10 ng/ml IL-1β for 24 hours plus either interleukin-1 receptor antagonist (IL-1RA 500 ng/ml), neutralising anti-IL-8 antibody (10 μg/ml) or neutralising anti-GM-CSF antibody (5 μg/ml). DNA synthesis was assessed by \[^3^H\]thymidine incorporation. Results expressed as means ± standard deviation. ^\*^*P* \< 0.01 vs. IL-1β stimulation in absence of inhibitor.](1471-230X-2-7-3){#F3} Effects of GM-CSF on proliferation ---------------------------------- In view of the results obtained with the anti-GM-SCF antibody, the direct growth-stimulatory actions of GM-CSF were examined. GM-CSF had a potent growth stimulatory action on AGS cells: significant enhancement of \[^3^H\]thymidine incorporation was seen at all concentrations (0.001--100 ng/ml) of GM-CSF. GM-CSF itself appeared to be a more potent stimulant than IL-1β ; maximal stimulation of 108 ± 17 % above control was seen with 100 ng/ml of GM-CSF (Figure [4](#F4){ref-type="fig"}). The inhibitory action of the anti-GM-CSF antibody was confirmed by abolition of the growth stimulatory action of 1 ng/ml GM-CSF (data not shown). Previous studies have shown that IL-1β-stimulated GM-CSF release under similar conditions in AGS cells to be approximately 10--20 pg/well/24 hours \[[@B32]\]. To confirm the results obtained with the anti-IL-8 antibody, \[^3^H\]thymidine incorporation was measured in response to IL-8. No enhancement of proliferation was seen at any concentration of IL-8 (0.001--100 ng/ml) (data not shown). Under similar conditions IL-1β-stimulated IL-8 release is approximately 3000 pg/well/24 hours \[[@B31]\]. ![Effect of GM-CSF on proliferation of gastric epithelial cells. AGS cells were treated with increasing concentrations of GM-CSF for 24 hours. Cell proliferation was assessed by \[^3^H\]thymidine incorporation. Results expressed as means ± standard deviation. ^\*^*P* \< 0.05, ^\*\*^*P* \< 0.01 vs control](1471-230X-2-7-4){#F4} Mechanism of IL-1β-stimulation of cell proliferation ---------------------------------------------------- The specific inhibitors genistein, which inhibits tyrosine kinases and PD 98059, which inhibits MAP kinase kinase (MEK), and thus inhibits the ERK-pathway, were used to assess the possible intracellular pathways mediating the effects of IL-1β. In order to examine the effects of IL-1β distinct from those of GM-CSF, these experiments were performed in the presence of the anti-GM-CSF neutralising antibody. As shown in figure [5](#F5){ref-type="fig"}, neither genistein nor PD 98059 altered unstimulated \[^3^H\]thymidine incorporation. Genistein completely abolished the IL-1β-stimulation of proliferation. Inhibition of MEK with PD 98059 reduced IL-1β-stimulated proliferation by 58 ± 5% (*P* \< 0.01) (figure [5](#F5){ref-type="fig"}) but did not completely abolish the growth stimulatory action of IL-1β. Further increases to supra-maximal concentrations of PD 98050 did not further inhibit IL-1β-stimulated proliferation (data not shown). ![Effect of inhibition of tyrosine kinase and MEK activity on IL-1β-stimulated gastric epithelial cell proliferation. AGS cells were treated with 10 ng/ml IL-1β for 24 hours in the presence of the tyrosine kinase inhibitor genistein (100 μM) or the MEK inhibitor PD 98059 (25 μM). Proliferation was assessed by \[^3^H\]thymidine incorporation. Studies were performed in the presence of anti-GM-CSF antibody (5 μg/ml). Results expressed as means ± standard deviation. ^\*^*P* \< 0.01 vs control](1471-230X-2-7-5){#F5} Discussion ========== This study demonstrated that IL-1β increased proliferation of AGS cells. This effect was reversed by the receptor antagonist, suggesting it was mediated via the interleukin-1 receptor. IL-1β stimulated both \[^3^H\]thymidine incorporation, as a measure of stimulation of DNA synthetic rate and also total cell numbers as measured by the MTT assay. This illustrates that the stimulation of DNA synthesis by IL-1β is translated into a real increase in cell numbers. A portion of the stimulatory action of IL-1β appears to be indirect. Neutralisation of GM-CSF in the media led to a significant reduction of IL-1β-stimulated proliferation. AGS cells are known to secrete GM-CSF in response to IL-1β \[[@B32]\]. GM-CSF itself was a potent stimulant of cell proliferation. Thus it seems likely that part of the growth stimulatory actions of IL-1β are due to an autocrine intermediary action of GM-CSF. There are limited previous data suggesting that GM-CSF stimulates the proliferation of non-haematopoietic cells; Dippold *et al* showed that exogenous GM-CSF stimulated the growth of two out of two cultures derived from gastric carcinomas and two out of nine pancreatic carcinoma cell lines \[[@B33],[@B34]\]. However, unlike the current study, autocrine production of GM-CSF was no detectable. Gastric epithelial cells also produce IL-8 in response to IL-1β \[[@B31]\]. However in this model system, IL-8 alone had no pro-proliferative action and neutralisation of IL-8 did not affect the stimulatory action of IL-1β. Therefore it is unlikely that IL-8 has an autocrine role in the growth-stimulatory action of IL-1β. It is possible that other cytokines produced by the gastric epithelial cells in response to IL-1β, *H. pylori or* other inflammatory insults could also play a role as autocrine or paracrine mediators of growth. It has recently been reported that another C-X-C chemokine, GRO/CINC-1, which is also upregulated in *H. pylori* infection stimulated proliferation in rat gastric epithelial cells \[[@B35]\]. The role of these other potential autocrine mediators deserves further study. The results of the inhibitor studies strongly suggest that tyrosine kinase activity is essential for the growth promoting action of IL-1β in AGS cells. The tyrosine kinase inhibitor genistein abolished the stimulatory action of IL-1β. IL-1β is known to activate a plethora of intracellular signalling pathways \[[@B31],[@B36]-[@B39]\], but in the current situation there appears to be an absolute requirement for signalling via a tyrosine kinase, subsequent on receptor activation. The mitogen-activated protein cascade is a well-characterised pathway mediating the cell growth-stimulatory actions of many growth factors and hormones. Inhibition of the ERK pathway, with the MEK inhibitor PD 98059, which prevents activation of ERKs by phosphorylation, had a significant inhibitory action against the stimulatory action of IL-1β. This suggests that activation of the ERK pathway is important in mediating the growth stimulatory actions of IL-1β. Activation of MAP kinase cascades, including the p42 and p44 ERK pathways and p46JNK and p55JNK c-Jun NH~2~-terminal kinases by IL-1β has been demonstrated in rat gastric epithelial cells \[[@B41],[@B42]\], but the functional importance of these pathways was not examined. Activation of ERKs and JNKs was inhibited by genistein \[[@B41]\], consistent with the inference of the current study that MAPKs lie downstream of tyrosine activity in IL-1β-induced signalling. However in the current study PD 98059 did not completely abolish the stimulatory action of IL-1β, suggesting that alternative pathways, activated subsequent on tyrosine kinase activity, also play a role in the signalling of the proliferative responses to IL-1β. Further studies are underway at present to examine these. There are conflicting data available concerning the direct effects of IL-1β on gastric epithelial proliferation. Although the current study and the study by Fan *et al* using stimulated-leukocyte conditioned media showed increased proliferation of the human AGS cell line \[[@B14]\], others have shown inhibition of serum, TGF-α and EGF-stimulated proliferation in RGM1 rat gastric epithelial cells by IL-1β \[[@B41],[@B42]\]. Tominaga *et al* showed that pre-treatment of RGM1 cells with IL-1β for 6 hours inhibited proliferation at 24 hours, but the inhibitory effect was lost at 48 hours, \[[@B41]\]. The reasons for these discrepancies are not clear. They may reflect intrinsic differences between the cell lines, differences in activation and involvement of paracrine growth-promoting pathways, species variability, an effect specific for certain growth factors, or underlying differences in the biology of cell lines derived from cancer (AGS) or normal tissue (RGM1). The studies demonstrating inhibition of proliferation by IL-1β were performed in the presence of powerful growth-promoting stimuli (high serum or specific growth factor concentrations in the culture media), whist the current studies were performed in serum-free or 1% serum media. It is possible that the multiple signalling pathways activated by IL-1β have differing effects on proliferation, the dominant effect depending on the complex inter-relationships of stimuli and signalling pathways under different circumstances. Pro-inflammatory cytokines such as IL-1β and TNF-α activate different signalling pathways with divergent results and different time-courses in gastric endocrine and parietal cells \[[@B37],[@B38],[@B43]-[@B46]\]. Further studies are underway examining the specific roles of the various signalling pathways in gastric epithelial cells under different conditions. There are now strong epidemiological data linking *H. pylori* with gastric carcinoma. The carcinogenic process appears to involve a number of steps: *H. pylori*-induced inflammation progresses to atrophy, intestinal metaplasia, dysplasia and eventually carcinoma \[[@B47]\]. Similarly *H. pylori* infection in Mongolian gerbils readily induces gastric atrophy and cancer \[[@B6]\]. Whilst gastric carcinogenesis is undoubtedly a multifactorial process, involving pathogenic bacterial and host factors, including HLA status, diet and antioxidant status \[[@B47],[@B48]\], it is clear that increased gastric epithelial proliferation, especially when relatively increased compared to apoptosis, is an important part of the pathway \[[@B49],[@B50]\]. Increased gastric epithelial proliferation is typical of *H. pylori* infection; it is demonstrable throughout all stages of the infection and eradication of the infection reduces proliferation. Increased proliferation is an important marker of increased risk of gastrointestinal adenocarcinoma \[[@B12]\]. The mechanisms of the increased proliferation are not completely understood. *In vitro* studies testing cultures of *H. pylori or* constituents have given conflicting results in different systems, with different cell types and bacterial strains. Direct stimulation of gastric epithelial proliferation by *H. pylori* has been reported by some authors \[[@B14],[@B15]\], whilst either neutral effects \[[@B15]\] or increased apoptosis and decreased proliferation have been reported by others \[[@B16]-[@B18]\]. Fan *et al* reported that conditioned media from either *H. pylori or* mitogen activated lymphocytes directly stimulated AGS cell proliferation \[[@B14]\], suggesting that the inflammatory response could be partly responsible for the enhancement of epithelial proliferation. IL-1β production is enhanced in *H. pylori* infection and this cytokine is regarded as being central to the regulation of the pro-inflammatory response in *H. pylori* infection. IL-1β is a profound inhibitor of gastric acid secretion *in vivo*\[[@B51]\] and in isolated parietal cells \[[@B37],[@B38]\]. Genetic polymorphisms of the IL-1β gene cluster causing increased transcriptional activity are associated with an increased risk of pre-cancerous and cancerous histological changes in *H. pylori* infection \[[@B21],[@B22]\]. The general hypothesis explaining this observation has been that enhanced IL-1β production consequent on *H. pylori* infection is responsible for greater suppression of acid secretion, which in turn allows greater colonisation of the acid-secreting body mucosa \[[@B24]\]. This greater colonisation provokes further inflammation, which ultimately leads to loss of specialised acid-secretory epithelium (atrophy) and further potentiates the vicious cycle of increased inflammation and diminished acid secretion \[[@B52]\]. Gastric atrophy significantly predisposes to cancer \[[@B3]\]. It is believed that atrophy in combination with mediators released in inflammation such as oxygen free radicals and nitric oxide, diet, anti-oxidant state and possibly bacterial overgrowth and generation of nitrosamines in the achlorhydric stomach \[[@B53]\] advance the histological changes, cause mutageneisis and drive progression to cancer. An alternative, but not mutually exclusive, hypothesis is that the enhanced cytokine response directly enhances epithelial cell proliferation and itself predisposes to cancer, in addition to the effects on acid secretion. The increased cell turnover of this hyperproliferative response would itself make the mucosa more vulnerable to the mutagenic effects of free radicals and other toxic products generated in the achlorhydric inflamed stomach. Elsewhere in the gastrointestinal tract, IL-1β has been shown to stimulate \[^3^H\]thymidine incorporation and increase cell number of cultured human colonic subepithelial myofibroblasts, which is thought to have importance in the remodelling of the mucosa in inflammation \[[@B54]\]. The pro-proliferative effects of IL-1β in the gastrointestinal tract deserve further study, given the importance of this cytokine in the regulation of the mucosal inflammatory response. Conclusions =========== The results of the current study suggest that IL-1β and GM-CSF can directly stimulate gastric epithelial proliferation. This might explain the proliferative responses to conditioned lymphocyte media demonstrated by Fan *et al*\[[@B14]\]. Enhanced epithelial proliferation due to IL-1β may contribute to the enhanced risk of gastric cancer and precancerous lesions in *H. pylori*-infected individuals with specific IL-1β alleles associated with higher levels of production. IL-1β stimulates proliferation via receptor-mediated activation of a tyrosine kinase pathway. Downstream signalling involves ERK-dependent and -independent pathways. Further studies will be necessary to clarify the mechanisms involved in IL-1β-stimulation of gastric epithelial proliferation, as well as data correlating IL-1β genotype, IL-1β protein production and epithelial proliferation *in vivo.* These will compliment the current study and further enhance our understanding of *H. pylori*-induced gastric carcinogenesis. List of abbreviations ===================== EGF -- epidermal growth factor; ERK -- extracellular signal related kinase; GM-CSF -- granulocyte-macrophage colony stimulating factor; IL -- interleukin; JNK -- c-Jun NH~2~-terminal kinase; MAP -- mitogen activated protein; MTT -- 3-\[4,5-dimethylthiazol-2-yl\]-2.5 diphenyl tetrazolim bromide; TCA -- trichloroacetic acid; TGF-α -- transforming growth factor alpha. Competing interests =================== None declared. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <>

Background ========== The National Institutes of Health have recommended that patients with chronic progressive renal insufficiency be referred to a multidisciplinary pre-dialysis team in order to minimize patient morbidity and ensure a smooth transition to dialysis therapy. \[[@B1]\] Referral to a pre-dialysis clinic is associated with many desirable outcomes including longer dialysis-free time intervals, \[[@B2]\] enhanced employment opportunities, \[[@B3]\] better patient education, \[[@B4],[@B5]\] and greater participation in the selection of dialysis modality. \[[@B4],[@B5]\] Patients referred to a multi-disciplinary pre-dialysis program also demonstrate better metabolic profiles, are less likely to require central venous catheter insertion, and require fewer in-hospital emergency dialysis starts and hospital admission days compared to patients who receive standard care. \[[@B6],[@B7]\] However, research exploring the benefits of pre-dialysis care from a patient-oriented perspective using self-reported quality of life measurements is limited. In a single Brazilian study, quality of life was compared between patients diagnosed with chronic renal failure less than one month before dialysis initiation (n = 53) and those patients diagnosed more than six months before starting dialysis (n = 60) using the Kidney Disease Questionnaire (KDQ). \[[@B8]\] Compared to patients diagnosed earlier, quality of life scores were significantly worse in the late-diagnosis group -- a finding that was particularly evident among elderly patients. However, statistical adjustment for covariates other that age and time of diagnosis was not performed. As well, only 28% of patients were over age 60, suggesting that the association between early diagnosis and improved quality of life may have been underestimated. Further study using a more representative sample reflecting the demographics of an aging Canadian population is therefore warranted. The present study hypothesized that attendance at a pre-dialysis clinic is an independent predictor of improved quality of life scores measured within the first six months of dialysis initiation in an incident cohort of chronic hemodialysis patients. Quality of life, measured using the SF 36-item Health Survey, was compared between patients who attended a pre-dialysis clinic and those who did not, adjusted for clinical and biochemical covariates. Methods ======= Patient population ------------------ Kingston General Hospital is a tertiary care university teaching centre in Southeastern Ontario, Canada, providing pre-dialysis, hemodialysis, peritoneal dialysis and transplantation programs. The hospital is the only centre offering nephrology consultation in this region, serving a population of about 450,000 people. Referral patterns are therefore very centralized. Patients receive hemodialysis at either Kingston General Hospital or one of several affiliated satellite units that offer both self-care and full nursing care services. Patients with renal failure are referred by primary care physicians to a nephrologist for initial assessment. The care of patients diagnosed with chronic and progressive renal failure who do not require imminent dialysis initiation is subsequently transferred by the nephrologist to the pre-dialysis clinic. Direct referral to the pre-dialysis clinic by a primary care physician is not permitted. The pre-dialysis clinic is scheduled weekly and staffed by nurse clinicians, dieticians, social workers, pharmacists, and nephrologists. The nephrologists at the hospital operate a group practice and staff the pre-dialysis clinic according to a rotating schedule. The medical records of all patients beginning chronic hemodialysis at Kingston General Hospital or affiliated satellite dialysis units were reviewed by the primary investigator for study eligibility. Study inclusion criteria were first hemodialysis initiation from January 1 1998 to January 1 2000, history of chronic irreversible renal failure, and completion of the QOL survey within six months of dialysis initiation. Exclusion criteria included receipt of dialysis for less than four weeks and previous kidney transplantation. Reasons for non-completion of the SF-36 questionnaire within six months of dialysis initiation included death (16 patients from the pre-dialysis group and nine patients from the non-pre-dialysis group) and transfer to peritoneal dialysis (two patients from the non-pre-dialysis group). Five patients were excluded due to a significant number of missing data points. An incident cohort of 120 adult hemodialysis patients satisfying the above inclusion and exclusion criteria was identified, of whom 74 patients had attended the pre-dialysis clinic on at least one occasion, while 46 patients began dialysis without previous pre-dialysis clinic attendance. Independent variables --------------------- Possible predictors of outcome included age (at dialysis initiation), sex, diabetes (yes/no), pre-dialysis clinic attendance (yes/no), length of pre-dialysis clinic attendance, date of QOL assessment, history of ischemic heart disease, congestive heart failure, peripheral vascular disease, stroke, malignancy, and chronic lung disease, Kt/V (measured within three months of QOL assessment), residual creatinine clearance (at dialysis initiation), albumin and hemoglobin levels (measured within one month of QOL assessment). Positive diabetic status included a history of either Type I or II diabetes. Patients were classified as having attended the pre-dialysis clinic if they presented on at least one occasion. Length of pre-dialysis clinic attendance was measured from the date of first attendance to date of dialysis initiation. The date at which the QOL assessment was performed was included as an independent variable (calculated from the date of dialysis initiation to QOL assessment) in order to account for any possible confounding effect due to the different times at which the Health Survey was administered. Kt/V was estimated based on a single-pool model using the urea reduction ratio method. Residual creatinine clearance at the time of dialysis initiation was estimated using the Gault-Cockroft method and the serum creatinine level at the start of dialysis initiation. History of co-morbid disease was extracted from medical records using local chart review and the Canadian Organ Replacement Registry (a national dialysis database). Angina was defined as any previous or present history of ischemic symptoms (possibly but not necessarily requiring anti-anginal therapy) or any past history of coronary artery angioplasty or bypass surgery. Congestive heart failure was defined as any previous or present history of dyspnea, orthopnea or paroxysmal nocturnal dyspnea due to cardiac etiologies and not solely attributable to renal failure and volume expansion alone. Echocardiographic documentation of systolic dysfunction was not required. Peripheral vascular disease was defined as either a typical history of exertional leg discomfort or previous history of vascular angioplasty or bypass surgery. Stroke was defined clinically and did not require radiographic confirmation. Patients were primarily of Caucasian race, reflecting the demographics of the southeastern Ontario region. Race was therefore not included as an independent variable. Dependent variables ------------------- QOL was assessed using the SF 36-Item Health Survey. \[[@B9],[@B10]\] This health survey is routinely administered every three months to all dialysis patients at Kingston General Hospital as standard care in order to monitor patient well-being. The survey is a well-validated self-report questionnaire that assesses QOL in eight domains: physical function (phy), physical role limitation (rlp), emotional role limitation (rle), social function (soc), body pain (pai), emotional well being (ewb), energy (eng), and general health (gh). The maximum score in each domain is 100 -- higher scores being desirable and indicative of better well-being or less pain. Changes in score of 5 units have been shown to be clinically relevant. \[[@B9]\] Additionally, the final multivariate model was used to explore the association between selected independent predictors of outcome and two composite scores that summarize physical and emotional QOL. Composite physical (praw) and emotional (eraw) scores were calculated based on an aggregation of the eight separate domain scores using standard published equations (where each of the domain scores is standardized using age-stratified Canadian national means and standard deviations). \[[@B11],[@B12]\] Statistical analysis -------------------- The distribution of independent and dependent variables was examined using univariate analyses. Comparison of continuous and categorical independent variables was performed between patients who attended the pre-dialysis clinic and those who did not using t-test, Mann-Whitney, and chi-square analyses as appropriate (Tables 1 and 2; see [Additional file Table 1](#S1){ref-type="supplementary-material"} and [Additional file Table 2](#S2){ref-type="supplementary-material"}). QOL scores were compared between patients who attended the pre-dialysis clinic and those who did not using t-test and Mann-Whitney analyses as appropriate (Table 3; see [Additional file Table 3](#S3){ref-type="supplementary-material"}). The association between independent variables and QOL scores in each of the eight health domains was examined using bivariate linear regression analysis (Table 4; see [Additional file Table 4](#S4){ref-type="supplementary-material"}). Based on the results of bivariate analyses, clinically relevant and statistically significant independent variables were selected for multivariate linear regression analysis (Table 5; see [Additional file Table 5](#S5){ref-type="supplementary-material"}). The final model was examined for adherence to linear model assumptions. In the situation where linear model assumptions were compromised due to the non-parametric distribution of outcome variables, non-parametric bivariate and multivariate analyses were also performed using Spearman correlation and logistic regression analyses respectively. As such, non-parametric analyses were used to examine physical and emotional role limitation scores due to the non-parametric distribution of these two dependent variables. The non-parametric distribution of these two domains is attributed to the relatively small number of questions devoted to these two domains in the SF 36-item Health Survey and is therefore not unexpected. Recognizing that earlier research using the health survey has traditionally reported results using parametric analyses despite the non-parametric distribution of the data, the present paper will present both parametric and non-parametric results in order to conform to previous methodologies and permit comparisons. Analyses were preformed using the SAS statistical analysis package for personal computers (version 6.12, Cary, NC). The study protocol was approved by the Queen\'s University Health Sciences and Affiliated Teaching Hospitals Research Ethics Board. Results ======= The distribution of categorical independent variables within the entire cohort and according to pre-dialysis clinic attendance is presented in Table 1 (see [Additional file Table 1](#S1){ref-type="supplementary-material"}). Overall, 55.8% of patients were male, while diabetes was identified in 43.6% of patients. Almost 40% of patients had a history of ischemic heart disease. The distribution of categorical independent variables was compared between the group of patients who attended the pre-dialysis clinic and those who did not. No statistically significant between-group differences were identified. Table 2 (see [Additional file Table 2](#S2){ref-type="supplementary-material"}) shows the distribution of continuous independent variables within the entire cohort and according to pre-dialysis clinic attendance. Mean age for the entire cohort was 62.7 years. Mean residual creatinine clearance clearance at dialysis initiation was 0.18 ml/s. A comparison of continuous independent variables according to pre-dialysis clinic attendance demonstrated no statistically significant differences between the two groups, with the exception of residual creatinine clearance. Residual creatinine clearance at dialysis initiation was significantly higher among the group of patients attending the pre-dialysis clinic (p = 0.05). Physical function, physical role limitation, emotional role limitation, social function, and general health scores were statistically higher among patients who attended the pre-dialysis clinic compared to patients who did not (Table 3; see [Additional file Table 3](#S3){ref-type="supplementary-material"}). Results of a series of bivariate linear regression analyses examining the association between each independent variable and each of the eight QOL domains are presented in Table 4 (see [Additional file Table 4](#S4){ref-type="supplementary-material"}). A positive parameter estimate denotes a positive correlation between independent variable and QOL score, while a negative parameter estimate indicates an inverse relationship. Younger age, male gender, pre-dialysis clinic attendance, longer attendance at the pre-dialysis clinic, higher residual creatinine clearance at dialysis initiation, absence of ischemic heart disease, stroke, and peripheral vascular disease, and higher albumin level were associated with a significantly higher physical function score. Higher scores in the physical role limitation domain were associated with younger age, pre-dialysis clinic attendance, and higher albumin level. Because physical role limitation scores did not follow a normal distribution, the bivariate analysis was re-examined using a nonparametric Spearman correlation analysis. Using a nonparametric analysis, pre-dialysis clinic attendance still predicted higher QOL score. Younger age (p = 0.02), pre-dialysis clinic attendance (p = 0.03), absence of ischemic heart disease (p = 0.04) and stroke (p = 0.03), and higher albumin level (p = 0.03) were associated with higher physical role limitation scores. Linear regression analysis suggested that higher emotional role limitation score was predicted by longer pre-dialysis clinic attendance. Due to the non-parametric distribution of the emotional role limitation score, a non-parametric analysis was also preformed, which demonstrated that pre-dialysis clinic attendance and longer pre-dialysis clinic attendance were the only predictors of better emotional role limitation score (p = 0.04 and p = 0.01 respectively). Significant predictors of better social function were pre-dialysis clinic attendance, absence of lung disease, and higher albumin level. Male gender, absence of ischemic heart disease, peripheral vascular disease and malignancy, and higher albumin level were significant predictors of higher pain scores (where higher score is indicative of less pain). Better emotional well-being was associated with male gender alone. Younger age, higher residual creatinine clearance at dialysis initiation, absence of peripheral vascular disease and lung disease, and higher albumin level predicted better energy level. Male gender, pre-dialysis clinic attendance, absence of ischemic heart disease, peripheral vascular disease and lung disease, and higher albumin level were significant predictors of higher general health scores. A total of nine independent variables were selected for inclusion in the final multivariate analysis: pre-dialysis clinic attendance, residual creatinine clearance, Kt/V, age, albumin, sex, ischemic heart disease, peripheral vascular disease, and diabetes (Table 5; see [Additional file Table 5](#S5){ref-type="supplementary-material"}). Age, sex, pre-dialysis clinic attendance, ischemic heart disease, peripheral vascular disease, and albumin were included as covariates because of their significant associations with QOL scores in bivariate analyses. Despite the absence of statistically significant associations with QOL scores, diabetes and Kt/V were included as covariates in the final multivariate model because of their perceived clinical relevance. Since residual renal function at dialysis initiation was significantly higher among patients attending the pre-dialysis clinic compared to those who did not, residual creatinine clearance was also included in the final model in order to permit statistical adjustment. The final multivariate model suggested that pre-dialysis clinic attendance was an independent predictor of four of eight health domains: higher physical function, emotional role limitation, social function, and general health scores. Physical function score was predicted by pre-dialysis clinic attendance, younger age, and absence of ischemic heart disease and peripheral vascular disease, while emotional role limitation and social function scores were predicted by pre-dialysis clinic attendance alone. Male gender and absence of peripheral vascular disease were independent predictors of higher pain score (reflecting less pain). Higher emotional well-being score was predicted by male gender. Pre-dialysis clinic attendance, male gender, and absence of peripheral vascular disease were independent predictors of higher general health score. Because of the non-parametric distribution of the physical and emotional role limitation scores, repeat analyses were performed for these two QOL domains using multivariate logistic regression analyses. The conclusions of logistic and linear regression analyses were identical. Multivariate analysis using the final model suggested that pre-dialysis clinic attendance, younger age, and absence of ischemic heart disease and peripheral vascular disease were independent predictors of the composite physical health score (Table 5; see [Additional file Table 5](#S5){ref-type="supplementary-material"}). Pre-dialysis clinic attendance was not an independent predictor of the mental health composite score. Adjusted r-squared values (Table 5; [Additional file Table 5](#S5){ref-type="supplementary-material"}) indicate the extent to which the variability of each health survey domain score can be explained by the final multivariate model. The final model accounts for 42% and 28% of the physical function score and composite physical score variability respectively, suggesting that the model is most effective in predicting physical function scores. Discussion ========== Although previous research has suggested that pre-dialysis clinic attendance is associated with better objective outcomes compared to standard care, research examining the benefits of pre-dialysis care with respect to patient-oriented outcomes such as QOL is limited. The results of the present study suggest that attendance at a predialysis clinic is an independent predictor of improved quality of life measured within the first six months of dialysis initiation in an incident cohort of chronic hemodialysis patients. The benefits of pre-dialysis clinic attendance are therefore not limited to objective measures of patient health such as decreased hospitalization and better biochemical profiles at the time of dialysis initiation. Rather, these results suggest that the benefits of pre-dialysis clinic attendance can be extended to include more subjective measures of patient health. The evaluation of medical therapies using patient-oriented outcomes has acquired a larger research profile in recent years. Nonetheless, to our knowledge, only one study has formally evaluated the benefits of pre-dialysis clinic attendance using patient-administered QOL questionnaires. \[[@B8]\] The present study confirms the conclusions of Sesso et al, who suggested that pre-dialysis clinic attendance improved QOL as measured by the Kidney Disease Questionnaire. However, only 28% of patients were over age 60 in the Brazilian study. Recognizing that the benefits of pre-dialysis clinic attendance were particularly pronounced among elderly patients, the value of pre-dialysis clinic attendance may have been underestimated by this earlier study. The mean age of 62.7 years in the present cohort is a closer representation of North American dialysis populations. The study is also benefited by a publicly funded and universally accessible Canadian healthcare system in which the importance of confounding variables such as socioeconomic status has been minimized. Furthermore, whereas the results of the Brazilian study were adjusted for age and time of diagnosis, differences in residual renal function and other covariates were not included. The present study has controlled for several covariates including age, sex, ischemic heart disease, peripheral vascular disease, diabetes, albumin, residual renal function, and Kt/V. Interestingly, pre-dialysis clinic attendance was associated with improved QOL scores in a wide range of domains that included both physical and emotional outcomes. Attendance was an independent predictor of better QOL score in four of eight domains: higher physical function, emotional role limitation, social function, and general health. Indeed, pre-dialysis clinic attendance was the only independent predictor of QOL score in the case of emotional role limitation and social function. Importantly, the difference in QOL scores between patients who did and patients who did not attend the pre-dialysis clinic is not only statistically significant but clinically significant as well. Previous research has suggested that a difference in SF-36 scores of five or more points is considered clinically relevant. \[[@B9],[@B11]\] The means by which a pre-dialysis clinic leads to better physical and emotional QOL is likely multi-factorial, resulting from opportunities for patient education, dietary counseling, modality selection, dialysis access creation, and management of co-morbid illnesses. The importance of early pre-dialysis correction of anemia using erythropoietin has also been emphasized in recent years. Several studies have suggested that earlier initiation of erythropoietin may improve patient morbidity and mortality by retarding or preventing the development of cardiomyopathy secondary to anemia. \[[@B13]\] The association between several covariates and QOL score seems intuitive in many cases. For example, in addition to pre-dialysis clinic attendance, younger age and absence of ischemic heart disease and peripheral vascular disease were independent predictors of better physical function score. Absence of peripheral vascular disease was also an independent predictor of better pain and general health scores. Although bivariate analysis suggested that higher serum albumin level predicted better QOL in many domains (including physical function, physical role limitation, social function, pain, energy, and general health), a statistically independent relationship between serum albumin and QOL was not demonstrated by multivariate analysis. Both pre-dialysis clinic attendance and increasing length of time attending the pre-dialysis clinic were predictors of better physical function and emotional role limitation scores in univariate analysis, but not multivariate analysis, thus suggesting statistical redundancy between these two covariates. The absence of an independent association between length of pre-dialysis clinic attendance (and presumably severity of chronic renal failure) and QOL outcome parallels the conclusions of Harris and colleagues, who found that, in a cohort of 360 non-dialysis patients with chronic renal failure, functional status was correlated with socioeconomic status and comorbid illness, but not severity of renal function. \[[@B14]\] Interestingly, multivariate analysis suggested that male gender was an independent predictor of better pain, emotional well-being, and general health scores. Although a significant association between male gender and QOL scores may seem doubtful, this finding has in fact been demonstrated previously in non-dialysis populations in which males consistently score slightly higher than females in all QOL domains. \[[@B12]\] QOL scores in the present study are comparable to earlier reports and underscore the fact that QOL is severely compromised among dialysis populations. \[[@B15],[@B16]\] In a prospective Dutch study involving 13 dialysis centres, QOL measured three months after dialysis initiation using the SF-36 questionnaire was significantly impaired compared to a general non-dialysis population, particularly with respect to physical role limitation and general health status. \[[@B15]\] Mean physical role limitation and general health scores for hemodialysis patients corresponded to the lowest 8% and 12% of the reference group respectively. Independent predictors of poor quality of life included a higher number of co-morbid illnesses, lower hemoglobin level, and lower residual renal function. Pre-dialysis attendance was not included as a covariate. Compared to more objective measures of patient well-being, the importance of QOL, not only as an outcome, but also as a predictor of patient morbidity and mortality, is probably under-estimated. For example, DeOreo has demonstrated that dialysis patients with a physical component summary score below the cohort median were twice as likely to die and 1.5 times more likely to be hospitalized compared to patients with scores above the median. These findings suggest that patient-reported QOL is as much a significant predictor of mortality as more well-established markers of patient care such as normalized protein catabolic rate and Kt/V. \[[@B17]\] Patient morbidity and mortality is significant even prior to dialysis initiation. \[[@B14],[@B18]\] Therefore, interventions that can potentially improve QOL should be pursued early and aggressively during the pre-dialysis phase of chronic renal insufficiency. The present study suggests that pre-dialysis care can improve QOL measured six months after dialysis initiation, while the conclusions of DeOreo argue that such improvement could lead to lower patient mortality. The inability of the present study to identify any predictors of the mental health composite score may be due to small sample size causing a Type II statistical error. However, other researchers have demonstrated that, while hemodialysis patients report low physical scores, mental and emotional scores are less affected, suggesting that the absence of any statistically significant predictors of poor mental health composite score in the present study may be valid. \[[@B15],[@B16]\] Surprisingly, diabetes did not predict QOL scores, even in bivariate analysis. This also could be due to effect of small sample size. Alternatively, failure to identify an association between diabetes and QOL scores may be due to short follow-up time in which the more long-term complications of diabetes could not be realized. Furthermore, the results of studies that have reported an association between diabetes and functional status are possibly limited by the use of cross-sectional study designs and prevalent cohort populations. \[[@B19]\] In a prospective cohort of dialysis patients over age 60, Kutner and colleagues showed that diabetic nephropathy was not an independent predictor of survival when controlled for functional status in multivariate analysis. \[[@B16]\] Kt/V was not a predictor of QOL outcomes in this study or previous investigations. \[[@B15]\] Some investigators have suggested that the absence of an association between Kt/V and QOL may be due to the effect of residual renal function and the possibility that clearance achieved by native kidneys is superior to that obtained by dialysis. \[[@B15]\] However, residual creatinine clearance was not a predictor of QOL outcomes in the present study. The absence of a statistically significant association between Kt/V and QOL scores could also be attributed to short follow-up time in which QOL assessments were performed soon after dialysis initiation. Alternatively, since Kt/V values were confined to a relatively narrow range in which the majority of patients achieved adequate dialysis dosing, a Type II statistical error simply due to lack of variability of the independent variable may account for the absence of any statistically significant association. The present study is limited by the retrospective design in which the possibility of selection bias cannot be definitively excluded. Indeed, previous research has demonstrated that referral of patients by primary care physicians to nephrology programs is influenced, either consciously or unconsciously, by patient-related factors including age, gender, diabetes, and other co-morbid illnesses. \[[@B20],[@B21]\] Consequently, referral bias is currently a feature of primary care practice and is likely unavoidable regardless of study design. Not surprisingly, the existence of referral bias in the present study is suggested by the lower residual renal function at the time of dialysis initiation seen in patients who did not attend the pre-dialysis clinic compared to those patients who did attend the clinic. Patient self-selection is also a consideration. For example, it is possible that non-compliant patients are less willing to attend the pre-dialysis clinic, and as such, the poorer QOL scores reported by patients who did not attend the pre-dialysis clinic may be related to issues of self-selection and compliance rather than pre-dialysis clinic attendance itself. However, other than residual renal function, no statistically significant demographic or clinical differences between those patients who did attend the pre-dialysis clinic and those who did not were identified in the present study. Notwithstanding the possibilities of referral and self-selection biases, the decision to examine an incident population restricted to new dialysis patients has avoided the other biases associated with earlier investigations using cross-sectional study designs and prevalent populations. \[[@B19]\] The retrospective study design also causes difficulty in isolating the independent influences of confounding covariates. For example, pre-dialysis clinic attendance and early dialysis initiation are separate yet potentially confounding variables, both of which are likely related to improved QOL scores. In the absence of randomized prospective studies, the effect of each covariate is difficult to examine independently since patients referred to a pre-dialysis clinic are probably less likely to require emergent dialysis initiation due to more regular follow-up. Consequently, it can be argued that the benefits of pre-dialysis clinic attendance may be the result of either selection biases (related to patient age and co-morbid illness) or confounding variables (such as timing of dialysis initiation and residual renal function) rather than pre-dialysis clinic interventions in and of themselves. However, a comparison of continuous and categorical independent variables between patients who did and did not attend the pre-dialysis clinic failed to demonstrate any statistically significant difference between the two groups apart from residual creatinine clearance at the time of dialysis initiation (which was therefore included as a covariate in multivariate analysis). The results of the present study support the NIH recommendation suggesting that patients with chronic progressive renal insufficiency should be referred to a pre-dialysis team in order to minimize patient morbidity and mortality. While previous research has demonstrated that pre-dialysis clinic attendance is associated with favourable objective outcomes (including better biochemical profiles and decreased hospitalization at the time of dialysis initiation), the results of the present study suggest that the benefits of a pre-dialysis program can be extended to include more subjective measures of health status. The importance of this finding is underscored by earlier research demonstrating that QOL scores are themselves independent predictors of patient mortality. However, it remains unclear to what extent the benefits of pre-dialysis clinic attendance are sustainable over time. Longitudinal studies evaluating the effects of pre-dialysis care with respect to prospective repeated-measure QOL assessments and objective outcomes are therefore required. Competing interests =================== None declared. Authors\' contributions ======================= Authors 1 (CW) and 2 (RP) designed the study and carried out data entry. Authors 3 (ML) and 4 (DH) performed the statistical analyses. Author 4 (DH) drafted the manuscript. All authors read and approved the final manuscript. Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Supplementary Material ====================== ::: {.caption} ###### Additional file 1 ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 2 ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 3 ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 4 ::: ::: {.caption} ###### Click here for file ::: ::: {.caption} ###### Additional file 5 ::: ::: {.caption} ###### Click here for file ::: Acknowledgements ================ The authors grateful acknowledge the assistance of the Canadian Organ Replacement Registry and medical record staff of Kingston General Hospital. We also appreciate the assistance of Ms. Wilma Hopman, Mckenzie Health services Research Group, Queen\'s University, Kingston, Ontario, Canada.

Background ========== The primary cause of senile cataract development is still unclear. To date the involvement of α-crystallin, a molecular chaperone for β- and γ-crystallin, has principally been considered in senile cataract development \[[@B1],[@B2]\], the decrease in the chaperone ability of α-crystallin with age being implicated. Xanthurenic acid is produced from a metabolite of tryptophan (3-hydroxy-kynurenine) \[[@B3]\] in the presence of 2,3-dioxygenase \[[@B4]-[@B7]\]. While 3-hydroxykynurenine \[[@B3]\] is photochemically inert \[[@B8]\] and serves as a protective UV filter of retina, xanthurenic acid is a photosensitizer \[[@B9],[@B10]\]. Xanthurenic acid accumulates with aging in mammalian lenses \[[@B11],[@B12]\] and is involved in the increased fluorescence of the lens with aging \[[@B13]\]. The glucoside of xanthurenic acid is also present in aged lenses \[[@B14]\]. Xanthurenic acid is an example of an endogenous ER stressor provoking an accumulation of unfolded proteins which in turn leads to an overexpression of Grp 94 and calreticulin in the lens epithelial cells of young mammals \[[@B15]\]. We have previously reported that porcine lens epithelial cells in culture respond to xanthurenic acid exposure by an overexpression of stress chaperone proteins, Grp 94 and calreticulin, in \[[@B15]\]. Here, we report that xanthurenic acid leads to human lens epithelial cells (HuLEC) death associated with caspase-3 activation, intracellular Ca^2+^ increase and calpain Lp82 induction. Previously, lens epithelial cell apoptosis was observed in an *in vitro* model of the cataract \[[@B16]-[@B18]\]. Materials and Methods ===================== Reagents -------- We used the following polyclonal antibodies from Santa Cruz Biotechnology Inc. CA, against cytochrome c, secondary antibodies IgG-fluoresceine (FITC)-conjugated. Primary antibody against active caspase-3 p17 was from Promega, Madison, USA. Secondary IgG-Texas Red-conjugated antibodies and Mitotracker CMXRos, DiOC~18~, Calcium Orange™, Hoechst 33342, propidium iodide were from Molecular Probes, Leiden, The Netherlands. Antibody against GPIP1 peptide of gelsolin was prepared as described previously \[[@B24]\]. Other reagents were from Sigma if not specified. Antibody against calpain Lp82 was from Dr. T. R. Shearer (University of Oregon, Oregon, USA). Preparation of human lens epithelial cells primary culture ---------------------------------------------------------- Lenses were obtained after transplantation of cornea from 58, 59, and 63 years old donors from Eye Bank University Hospital, Bern. The primary cells cultures were prepared separately from every donor using two lens capsules. The lens capsules from one donor were treated with 1.5 mg/ml of collagenase 1A and 4 overnight at 37°C. Thereafter, 1 ml of MEM medium with 10% FCS was added, and cells were centrifuged for 10 min at 300 g. The supernatant was discarded and the cells were re-suspended in 1.5 ml the growth medium, as described below, for 14 days in one well of 12-well plate. The cells were cultivated in the dark, in Minimal Essential Medium (MEM) with Earle\'s salts (Gibco BRL). Cells were grown under a humidified atmosphere of 5% CO~2~ in air at 37°C in MEM supplemented with 10% fetal bovine serum, penicillin (10 U/ml), streptomycin (10 μg/ml) and fungizone (250 ng/ml). When confluent, they were incubated in MEM or MEM supplemented with xanthurenic acid. A 20 mM stock solution of xanthurenic acid was prepared in 0.5 M NaHCO~3~, and diluted in 0.01 M PBS pH 7.4. Cytotoxicity and apoptosis assay -------------------------------- Cells were observed with differential interference contrast and phase contrast optics on a Zeiss Avionert 405 M inverted microscope, and images recorded with a Matsumoto 3-chip CCD cooled camera with images stored using Adobe Photoshop 4. Cell viability was determined by staining the cells with Hoechst 33342 and propidium iodide (PI) (Juro, Switzerland) using 50 μg/ml of each dye. Fragmented, apoptotic, nuclei were observed with excitation at 350 nm, and necrotic nuclei at 530 nm. Cell lysis and immunobloting ---------------------------- Cells were washed twice with cold 0.01 M PBS, pH 7.4. For Western blotting, cells were lysed in buffer containing 50 mM Tris (pH 8.0), 150 mM NaCl, 1 %Triton X-100, and the following protease inhibitors: 1 mM phenyl-methylsulfonyl fluoride, and leupeptin, aprotinin, and pepstatin, each at 1 μg/ml. The concentration of proteins was calculated from the absorption maximum at 280 nm, as described previously \[[@B15]\], and the concentration of xanthurenic acid from its absorption maximum at 342 nm (ε~M~ 6500). The lysate was centrifuged for 10 min at 14 000 g, and the supernatant was boiled in loading-buffer for 5 min. Proteins (50 μg per lane) were separated by SDS-PAGE containing 10 or 12.5% acrylamide. After transfer to Hybond ECL membrane (Amersham Pharmacia Biotech AB, Uppsala, Sweden) the proteins were probed with the appropriate antibodies. Chemilunimescence ECL system (Amersham Pharmacia Biotech AB, Uppsala, Sweden) was used for the detection of peroxidase-conjugated secondary antibody. Immunofluorescence studies -------------------------- Cells grown on glass coverslips were fixed for 10 min at room temperature in 4% paraformaldehyde in 0.1 M PIPES, pH 6.8, washed in PBS and permeabilized for 5 min in PIPES containing 0.05% saponin (65 μl per coverslip), washed in PBS, incubated for 10 min in cold aceton for additional fixing and permeabilisation, and again washed in PBS. Cells were incubated for 1.5 hour with the first antibody diluted in PBS containing 1% bovine serum albumine, and after washing incubated for 1.5 hour with the secondary antibody. The coverslips were then washed in PBS and incubated for 10 min with 65 μl of solution containing 1μl of Hoechst 33342 dye (1 mg/ml), washed in PBS, and incubated with Antifade Kits (Molecular Probes, Leiden, The Netherlands) according to the supplier\'s instruction. Staining of mitochondria was performed using Mitotracker CMXRos, as follows: confluent cells cultures were pre-incubated without or with xanthurenic acid in MEM medium for 72 hours. The medium was removed and replace with medium containing 100 nM Mitotracker CMXRos. After an incubation for 1 hour Mitotracker CMXRos was removed, coverslips were washed twice with PBS, and mounted on the slides using as antioxidative solution 9% w/v of Mowiol (Calbiochem) in 22% glycerol buffered with 0.2 mM Tris/HCl to pH containing 3.5% (w/v) of 1,4-diazabicyclo (2.2.2) octane (Sigma). Membranes were stained using overnight incubation in DiOC~18~ at concentration of 12 μM in MEM medium. Results ======= Xanthurenic acid activates caspase-3 and translocates gelsolin from the mitochondrial region to the cytoskeleton ---------------------------------------------------------------------------------------------------------------- We observed that 10 μM xanthurenic acid in the HuLEC cell culture medium activate caspase-3 (Fig. [1B](#F1){ref-type="fig"}). In the same cells the translocation of gelsolin from perinuclear region to cytoskeleton was observed (Fig. [1C, 1D](#F1){ref-type="fig"}). Previously it was reported that caspase-3 and -9 are associated with mitochondrial membranes \[[@B19]\]. Gelsolin is the cytoskeletal protein responsible for the disintegration of F actin during apoptosis induced by Fas \[[@B20]\]. It was also suggested that gelsolin keeps caspases in the inactive state \[[@B21]\]. We reported previously that gelsolin is abnormally cleaved during apoptosis induced by xanthurenic acid. This abnormal gelsolin cleavage could be a reason of absence of cytoskeleton breakdown in the presence of active caspase-3 (Fig. [1B, 1D](#F1){ref-type="fig"}). {#F1} Xanthurenic acid leads to mitochondrial migration, cytochrome c release, and destruction of mitochondrial structure ------------------------------------------------------------------------------------------------------------------- In the control cells mitochondria occupy the perinuclear region (Fig. [2A](#F2){ref-type="fig"}). In the presence of 10 μM xanthurenic acid mitochondrial migration was observed (Fig. [2B](#F2){ref-type="fig"}). However, at higher concentrations (20 and 40 μM) xanthurenic acid led to the destruction of mitochondria. An intrinsic apoptotic pathway is activated by cytochrome c release and apoptosome formation with APAF-1 and ATP \[[@B22]\]. The apoptosome leads to activation of caspase-9, which activates caspase-3. We observed that in the presence of 10 μM xanthurenic acid cytochrome c was release from mitochondria (Fig. [3](#F3){ref-type="fig"}). APAF-1 is present in the HuLEC and its level is independent from xanthurenic acid concentration (not shown). Thus, release of cytochrome c is responsible for the observed caspase-3 activation, and nucleus cleavage. {#F2} {#F3} Mitochondrial damage in the presence of xanthurenic acid is associated with nuclear cleavage -------------------------------------------------------------------------------------------- In control cells about 5 percent are apoptotic (Fig. [4A](#F4){ref-type="fig"}, Fig. [5](#F5){ref-type="fig"}). In the presence of 10 μM xanthurenic acid condensed and/or cleaved nuclei were observed, which a mostly stained only with Hoechst 33342, and not with propidium iodide. This indicated that the observed cell death was apoptotic and not necrotic (Fig. [4B, 4C](#F4){ref-type="fig"}). With concentrations of xanthurenic acid of 20 μM (Fig. [4D,4E,4F](#F4){ref-type="fig"}) and 40 μM (Fig. [4G,4H,4I](#F4){ref-type="fig"}) the destruction of the mitochondrial structure, was associated with nuclear cleavage. Cell death depended on the xanthurenic acid concentration (Fig. [5](#F5){ref-type="fig"}). In the presence of 10 μM xanthurenic acid about 40 percent of cells were dead, and an increase of xanthurenic acid to 20 μM provoked about 70 percent of cell death. {#F4} {#F5} Xanthurenic acid leads to damage of the cell membrane ----------------------------------------------------- The cell membranes were stained with DiOC~18~ and co-stained with Mito-Tracker Red CMXRos, and the nucleus was stained with Hoechst 33342 (Fig. [6](#F6){ref-type="fig"}). In the control cell mitochondria were in the perinuclear region and the cell membranes were uniformly stained with DiOC~18~ (Fig. [6A,6B,6C](#F6){ref-type="fig"}) In the presence of 10 μM xanthurenic acid the mitochondria migrated to the cell periphery and the cell membranes were not uniformly stained (Fig. [6D,6E,6F](#F6){ref-type="fig"}). In the presence of 20 μM of xanthurenic acid the mitochondrial structure was destroyed, nuclear DNA was degraded and membranes were not stained with DiOC~18~ (Fig. [6G,6H,6I](#F6){ref-type="fig"}). {#F6} Xanthurenic acid causes an increase of free intracellular Ca^2+^ and an induction of the lens calpain Lp82 ---------------------------------------------------------------------------------------------------------- Ca^2+^ increases are associated with cataract development. We investigated intracellular Ca^2+^ by loading the cells with acetometoxyl ester of Calcium Orange™. This dye becomes fluorescent when hydrolysed in the cell by esterases and conjugated with free Ca^2+^ Cells were incubated without xanthurenic acid or with xanthurenic acid at concentration of 0.125; 0.25, 0.5; 1; 2; 5, 10, and 20 μM. A presence of xanthurenic acid in the cell culture medium higher then 2 μM provokes an increase of intracellular Ca^2+^ in comparison with control. In the presence of xanthurenic acid at concentration of 10 μM and 20 μM an intensive staining with Calcium Orange™ was observed indicating an increase of free Ca^2+^ in the cell in a xanthurenic acid concentration-dependent manner (Fig. [7](#F7){ref-type="fig"}). {#F7} The lens specific calpain Lp82 \[[@B23]\] was not detectable using the Western blot analysis in the lens epithelial cell culture cultivated in the absence of xanthurenic acid (Fig. [8](#F8){ref-type="fig"}, lane 1). In the presence of xanthurenic acid the calpain Lp82 was induced (Fig. [8](#F8){ref-type="fig"}, lanes 2-4). Discussion ========== Our previous studies in cell culture showed that xanthurenic acid is a potent endogenous pathological substance in retinal pigment epithelium and smooth muscle cells \[[@B24]\]. In this study, we observed that xanthurenic acid leads to lens epithelial cells death associated with an induction of caspase-3 and calpain Lp82. Apoptosis was observed in models of cataract upon lens treatment with staurosporine, diamide, and ionophore \[[@B16]-[@B18]\]. In the selenite cataract model caspase-3 and calpain were induced \[[@B25]\]. In the normal apoptosis, such as observed with development, cells disappear because of caspase-3-dependent cleavage of DNA and the cytoskeleton \[[@B26],[@B27]\]. The caspase remodeling of the cytoskeleton was indicated as a possible mechanism leading to the aging of lenses \[[@B28],[@B29]\]. Apoptosis is considered as a common cellular basis for non-congenital cataract in mammals \[[@B30]\]. Recent data indicate that senile cataract is a result of proteolysis of crystalins by calpains \[[@B31]\]. Calpains are activated by Ca^2+^\[[@B32]\]. Previously, it was reported that ER Ca^2+^ homeostasis affects the cells\' sensitivity to apoptosis \[[@B33]\]. Lens epithelial cells overloaded by Ca^2+^ showed vimentin cleavage and opacification \[[@B34]\] Thapsigargin, a plant alkaloid, which depletes Ca^2+^ from the ER, was used to stop lens epithelial cell growth \[[@B35]\]. Here, we show that xanthurenic acid, an endogenous molecule, lead to induction of calpain Lp82 and caspase-3 activation. The simultaneus activation of caspase and calpain may lead to an abnormality of apoptosis because calpain cleaves caspases \[[@B36]\]. The calpain Lp82 is involved in cataract formation in connexin α-3 knockout mice \[[@B37]\]. The induction of calpain Lp82 leads to the cleavage of crystalins in the lenses and is involved in the senile cataract development \[[@B38]\]. Xanthurenic acid leads to formation of unfolded proteins \[[@B15]\]. An accumulation of unfolded protein can lead to Ca^2+^ release from intracellular stores as well to caspase induction \[[@B39],[@B40]\]. The observed death of the lens epithelial cells has apoptotic characteristics because release of cytochrome c and caspase-3 activation were observed. However, the apoptosis-like process does not lead to a collapse of the cytoskeleton driven by caspase-3 cleaved gelsolin in the presence of Fas-induced apoptosis \[[@B20]\]. In our study, in the presence of xanthurenic acid cells look normal when observed by the light microscopy. However, when visualized by fluorescence microscopy with Hoechst and propidium iodide a nuclear dysfunction is evident. We have previously reported that xanthurenic acid leads to an abnormal cleavage of gelsolin. The gelsolin stains cytoskeleton of the dead cells. Ca^2+^ and bis-phosphatidylinositol (PPI2) regulate, respectively, association and dissociation of gelsolin to actin \[[@B41]\]. Thus, also the increase in Ca^2+^ may play a role in the association of gelsolin to cytoskeleton. The mitochondrial damage observed in the presence of xanthurenic acid associated with cytochrome c release could lower energy necessary for the lens enzymes activation observed in senile cataract development. In this study we observed that xanthurenic acid accumulation can be an upstream event leading to an induction of the lens proteases: caspase and calpain. An accumulation of xanthurenic acid in the human lenses with aging can change the intracellular Ca^2+^ homeostasis. In summary, xanthurenic acid can induce the pathology of the lens epithelial cells without participation of light. Author contributions ==================== All authors contributed equally to realize this work Competing interests =================== None declared {#F8} Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ This work was supported by grant awarded to H. Z. M. by the Swiss National Foundation (32-59183.99) and MSE-Pharmaceutika GMBH, Bad Homburg, Germany. We thank Dr. T. R. Shearer for the kind gift of the antibody against calpain Lp82, Mr. R. Fischer and Mrs A. Marrero Nodarse for cell culture facilities, and Mrs D. Zuercher for lens epithelial cell culture preparation.

Background ========== Circadian clocks located within metazoan brains have been a focus of investigation for decades \[[@B1]-[@B4]\]. Such clocks are compelling in part because they influence timing across many levels of analysis; they are driven by molecular elements whose concerted output, in the form of neuronal signaling, controls physiology and behavior \[[@B5]\]. Recently, the presence of circadian clocks in peripheral tissues has been established in the fruitfly as well as in lower vertebrates and mammals \[[@B6]-[@B9]\]. For example, in *Drosophila* such clock functions have been demonstrated in antennae, wings, legs, a neuro-humoral gland, and internal excretory structures \[[@B6],[@B10],[@B11]\]. But it is unknown whether time-keeping mechanisms are identical wherever clocks are found within an animal. While it appears that the same molecules previously identified as clock components in the brain give rise to clock function in the periphery, it is not yet known whether clock genes and their encoded proteins act and interact in the same manner wherever they are expressed. Similarly, although circadian pacemaker cells in the *Drosophila* brain have been characterized anatomically as discrete clusters of cells called the Lateral Neurons \[[@B12],[@B13]\], the identity of clock cells in the tissues we studied is unknown. The same can be said about the function of these clocks: while circadian clock cells in the brain control behavioral rhythms, the biological functions of clocks in peripheral tissues are minimally understood, except perhaps in the case of the antenna \[[@B14]\]. Analysis of genetic and molecular mechanisms underlying circadian clocks in whole animals or isolated tissues is facilitated by the use of molecular reporters, such as firefly *luciferase* (*luc*). The use of firefly-*luciferase* (*luc*) as a real-time molecular reporter of gene expression has been illustrated in a variety of model systems including cyanobacteria, plants, *Drosophila*, and mammals \[[@B15]-[@B18]\]. This method has proved especially valuable in studies about molecular rhythms because real-time expression of clock genes and other chronobiological factors may be monitored over many daily cycles *in vivo* \[[@B17],[@B6],[@B19]-[@B21]\]. Applications of this technique are varied. For example, *luc* reporters have been employed to identify previously unappreciated genes that are involved in circadian timing \[[@B21],[@B22]\] as well as to demonstrate the presence of circadian clocks in isolated *Drosophila* tissues. We previously employed *luc* reporting to study the effects of a *cryptochrome* mutation (*cry*^*b*^), concentrating on clock gene expression in isolated antennal pairs from *Drosophila* \[[@B23]\]. Earlier studies established that *cry* encodes a \'deep-brain\' element of the photoreceptive pathway to the circadian clock \[[@B24]-[@B26]\]. But our study suggested a different role for *cry* in the periphery, because antennal rhythmicity in mutant specimens was dramatically reduced in constant darkness \[[@B23]\], an effect that is not consistent with the idea that the encoded protein (CRY) functions solely as a photoreceptor for the circadian system \[[@B24],[@B25]\]. While carrying out those studies, we became aware that there are interesting features of these peripheral rhythms, like the phase of peak expression, for example, that we were unable to compare rigorously amongst body parts. Such concerns motivated us to develop a set of tools for the analysis of circadian rhythms \[[@B27]\]. Here we apply two key features of this analytic package that were unavailable for our earlier report on the effects of *cry*^*b*^ in the periphery \[[@B23]\]. First is a method to eliminate nonlinear trends in the data, a potential source of artifact that could affect whether *luciferase* activity is scored as rhythmic or arrhythmic. Figure [1](#F1){ref-type="fig"} illustrates how rhythmicity generated by one antennal pair is extracted from rather weak output by the application of appropriate analytic tools. In addition, the normalization procedure facilitates comparisons between rhythms from different tissues by reducing the scale of the output to a mean of one for each case (see Results). Second, we were not previously able to examine effects of a given variable (genotype, body part) on circadian phase. The application of circular statistics to these data allows us to evaluate such effects. ![Effect of detrending and normalization on the analysis of *luciferase*-reported rhythms. This example is taken from an antennal pair isolated from a *BG-luc; cry*^+^ adult. *luc* activity was recorded for 8 days in constant darkness and at a constant temperature of 27°C. a) The upper panel in this column shows a plot of the raw data given in counts per second of bioluminescence plotted on the ordinate. Samples were recorded hourly as indicated on the abcissa. The dashed line illustrates a trend line defined by the application of a 72 hour low pass filter (see \[[@B27]\] for more detail about this use of filters). The lower panel shows the autocorrelation function for the raw data plotted immediately above. We use the autocorrelation function to detect rhythmicity\[[@B23],[@B27]\]. Correlation coefficients with range from -1 to 1 appear on the ordinate and the lag is plotted on the abscissa. The shaded area centered around 0 defines a 95% confidence interval. In this case, the shape of the function indicates one rhythmic cycle followed by a decrease, a pattern that suggests the signal is not rhythmic, b) The upper panel in this column shows the detrended and normalized signal obtained from the raw data shown in a). The signal was obtained by dividing each data value by the corresponding value on the trend line for a given time point. The values on the ordinate are given in arbitrary units and vary around a mean of 1 (as described in \[[@B27]\]). The lower panel shows the autocorrelation function for the detrended and normalized signal shown immediately above it. The application of these procedures reveals the presence of significant rhythmicity within the data set shown in a) above. The RI value is the value of the autocorrelation function at the third peak (marked by the asterisk); it indicates the strength of rhythmicity (see Table [1](#T1){ref-type="table"}, Materials and Methods and \[[@B27]\] for more detail). The RS value is obtained from the ratio of the RI value to the 95% confidence line. Thus when RS is ≥ 1, the rhythm is statistically significant (see Materials and Methods for more detail). The absence of an asterisk in the lower panel of a) indicates the absence of a third peak in the autocorrelation function for the raw data set.](1471-2202-3-5-1){#F1} ::: {#T1 .table-wrap} ::: {.caption} ###### Effects of the *cry*^*b*^ mutation on *BG-luc* and *tim-luc* reporter activity in isolated body parts under LD 12:12^1,2^ ::: Genotype^3^ N~a~^4^ Number Rhythmic^5^ (statistical significance) Period^6^ (Mean ± SEM) Rhythmicity Index^7^ (Mean ± SEM) Rhythmicity Statistic^8^ (Mean ± SEM) Amplitude^9^ (Mean ± SEM) Phase^10^ Activity Counts^11^ (Mean ± SEM) --------------------- --------- ----------------------------------------------- ------------------------ ----------------------------------- --------------------------------------- --------------------------- ----------- ---------------------------------- Heads *BG-luc; cry*^+^ 38 31 (14) 24.7 ± 0.4 0.32 ± 0.03 2.0 ± 0.2 0.10 ± 0.01 3.8,0.9 5759 ± 422 *BG-luc; cry*^*b*^ 35 4 (1) 22.5 ± 3.5 0.13 ± 0.02 0.8 ± 0.1 0.05 ± 0.01 10.6, 0.7 8382 ± 948 *tim-luc; cry*^+^ 22 14 (5) 24.7 ± 0.6 0.21 ± 0.03 1.3 ± 0.2 0.10 ± 0.01 3.1,0.8 32024 ± 3766 *tim-luc; cry*^*b*^ 5 0 \- \- \- \- \- \- Forelegs *BG-luc; cry*^+^ 37 33(11) 25.0 ± 0.3 0.45 ± 0.02 2.8 ± 0.2 0.11 ± 0.01 3.7,0.9 1089+93 *BG-luc; cry*^*b*^ 29 11(4) 25.2 ± 1.2 0.32 ± 0.07 2.0 ± 0.4 0.11 ± 0.01 5.0,0.8 902 ± 181 *tim-luc; cry*^+^ 24 24(5) 24.6 ± 0.2 0.32 ± 0.03 2.0 ± 0.2 0.12 ± 0.01 1.3,0.9 2605 ± 411 *tim-luc; cry*^*b*^ 17 4(0) 24.6 ± 3.2 0.02 ± 0.05 0.1 ± 0.3 0.04 ± 0.00 5.7, 0.9 1488 ± 240 Bodies *BG-luc; cry*^+^ 32 20(2) 24.2 ± 0.5 0.30 ± 0.03 1.9 ± 0.2 0.10 ± 0.01 5.1,0.9 4549 ± 458 *BG-luc; cry*^*b*^ 28 5 (0) 20.8 ± 0.4 0.02 ± 0.08 0.1 ± 0.5 0.21 ± 0.09 7.1,0.3 4245 ± 1087 *tim-luc; cry*^+^ 18 15(4) 24.4 ± 0.2 0.34 ± 0.04 2.1 ± 0.2 0.13 ± 0.01 2.9, 0.9 17678 ± 5053 *tim-luc; cry*^*b*^ 7 0 \- \- \- \- \- \- Wing *BG-luc; cry*^+^ 32 30(14) 25.1 ± 0.1 0.5 ± 0.02 3.1 ± 0.2 0.23 ± 0.01 2.2, 0.9 313+36 *BG-luc; cry*^*b*^ 25 20 (5) 27.0 ± 0.6 0.2 ± 0.05 1.3 ± 0.3 0.14 ± 0.02 7.5,0.5 333+31 *tim-luc; cry*^+^ 15 15(7) 25.2 ± 0.2 0.42 ± 0.04 2.6 ± 0.2 0.24 ± 0.02 -0.3, 1.0 497 ± 69 *tim-luc; cry*^*b*^ 5 5 (0) 26.4 ± 0.4 0.22 ± 0.05 1.4 ± 0.3 0.10 ± 0.01 7.4, 0.9 485+187 **1.** The experimental design is described in Krishnan et al \[[@B23]\] and in Materials and Methods. LD 12:12 refers to a light-dark cycle with 12 hours of light and 12 hours of darkness. **2.** Individual body parts were isolated by dissection and placed immediately in cell culture medium containing luciferin substrate for analysis under LD 12:12 (see Materials and Methods) **3.** Fly strains are described in Stanewsky et al \[[@B20],[@B21]\]. **4.** N~a~ is number of specimens analyzed. This analysis was applied to samples previously reported elsewhere \[[@B23]\]. The number of *cry*^+^ specimens was increased over the number previously reported as follows 10 additional heads, 9 additional forelegs and 15 additional bodies; the number of *cry*^*b*^ specimens is unchanged. **5.** Each sample is evaluated separately and considered rhythmic based on the correlogram and the requirement that rhythmicity falls between 18--40 hours according to spectral analysis (see Materials and Methods and also see \[[@B27],[@B30]\]). We asked whether *cry*^*b*^ affects rhythmicity in these body parts. Chi-squared tests showed significant effects of the mutation on rhythmicity with *BG-luc* or *tim-luc* for all body parts (p \< .02) except the wing (p \> .05 for both reporters). As discussed in the text, the finding of rhythmicity does not necessarily indicate statistical significance for the rhythm. The numbers in parentheses indicate how many of the specimens we called rhythmic also displayed statistically significant rhythmicity (the height of the peaks in the correlogram were above the 95% confidence line). The remainder of the rhythmic specimens were determined to be rhythmic because of the sinusoidal shape of the correlogram. See the text for further discussion about our criteria for rhythmicity (also see \[[@B23]\]). **6.** The estimate of circadian period is assessed by mesa \[[@B33]\] for each individual. Mean and standard error of the mean (SEM) are tabulated from the individual estimates. **7.** The rhythmicity index (RI) is a measure of the strength of the rhythm obtained from the autocorrelation function as described in Levine et al \[[@B27]\], see also \[[@B31]\]. Like the estimate of period, the RI is given as a mean with SEM based on the values obtained for each individual rhythmic sample. The *cry*^*b*^ mutation significantly reduces the RI value for each reporter in every body part (t-test, p \< .001). Note that these tests could not be performed for *tim-luc* specimens from isolated heads or bodies because there were no rhythmic samples to evaluate. **8.** The Rhythmicity Statistic (RS) is calculated as a ratio of the RI to the absolute value of the 95% confidence line for the correlogram obtained for each individual with means and SEM tabulated as above for RI (see Figure [2](#F2){ref-type="fig"} and Figure [3](#F3){ref-type="fig"}, for examples). The RS provides a quick indicator of whether the rhythm is statistically significant (RS ≥ 1) or not (see Materials and Methods). **9.** Amplitude is a measure of the distance from the peak (or trough) to the mean in the detrended and normalized rhythmic data (see Materials and Methods for more details) **10.** The two numbers given here represent the mean phase, or the direction in which the phase vector points and the correlation coefficient describing the distribution of phases among the specimens, or the length of the vector . Phase is determined for the group of rhythmic individuals using circular statistics \[[@B22],[@B30]\]. See Figure [7](#F7){ref-type="fig"} for example. **11.** Mean expression level is given as mean ± SEM for counts per second of bioluminescence/hour. ::: Our earlier conclusions \[[@B23]\] have been substantially buttressed by the new features of these time-series analyses: With respect to its effects on clock-gene cyclings, the *cry*^*b*^ mutation significantly reduces the number of rhythmic antennal pairs, heads, bodies and forelegs (but not wings) compared with *cry*^+^ controls. Moreover, we have intensified the analysis of the rhythmic specimens and thus observed that parameters of rhythmicity are altered in the subset of rhythmic mutant specimens. Among the new findings that have emerged from the current analyses, the most salient are that (1) *cry*^*b*^ has a significant effect on the phase of clock gene expression in rhythmic peripheral tissues; and (2) in isolated *cry*^+^ tissues, the relationship between the peak phase time of *luc*-reported expression from *per* and *tim* genes (respectively) varies such that the peaks are indistinguishable in antennae and they differ significantly in wings or forelegs, for example. The findings reported here strengthen the hypothesis that clock mechanisms in the periphery differ from each other as well as from those operating in the central nervous system of *Drosophila* \[[@B23]\]. Results ======= We chose *Drosophila* strains in which *luc* was either fused downstream from the 5\'-flanking sequence of the clock gene *timeless* (*tim-luc*), or downstream from a genomic sequence that spans the 5\'-flanking region of the *period* gene and extends through the N-terminal 2/3 of its coding sequence (*BG-luc*) \[[@B20],[@B21]\]. These *luc* fusion constructs were introduced into the *D. melanogaster* genome by germ-line transformation and shown to mediate daily cycles of LUC activity over the course of several days\' worth of monitoring live (whole) flies fed on luciferin-containing medium \[[@B20]\]. We were able to assess the effects of the *cryptochrome* mutation on clock-gene expression after carrying out genetic crosses that placed chromosomes bearing the reporter constructs into a homozygous *cry*^*b*^ background and comparing the effects of this mutation on LUC temporal patterns to those influenced by the normal *cry*^+^ allele \[[@B23]\]. In addition, these controls allowed comparisons of *tim-luc* to *BG-luc* expression. Isolated heads, forelegs, bodies, wings and antennal pairs were assayed, although data from the latter tissue-type were essentially the only ones reported \[[@B23]\]. Effects of *cry*^*b*^ on rhythms in isolated clock tissues under light-dark cycles ---------------------------------------------------------------------------------- We evaluated these various isolated tissues under a series of successive 12-hr-light12-hr-dark cycles (LD 12:12). These environmental conditions are the least likely to reveal deficits in the pattern of clock gene expression because the light-dark cycle provides a dominant stimulus to the clock (clock time is normally synchronized to this environmental cue) and is thought to increase the amplitude of clock gene cycling \[[@B20],[@B28],[@B29]\]. Rhythmicity was assessed for each specimen by autocorrelation analysis \[[@B30],[@B27]\]. This analysis provides a quantitative estimate of rhythmicity with statistical confidence; whereas the confidence interval is based solely on the number of data points used in the analysis without relying on the variability in the data set or any other feature of the measured values, the criterion for statistical confidence is the same for all individual or group analyses within an experiment (see \[[@B27]\] for more detail). In addition, the autocorrelation read-outs may be used in a less stringent (and more subjective) manner to evaluate rhythmicity, based on the shape of a given plot -- even when statistical significance is not achieved (see \[[@B27]\] for discussion of this feature of applying autocorrelation). This qualitative criterion is conservative from at least one perspective: If we imagine that *cry*^*b*^\'s effect might be an elimination of *per-luc* and *tim-luc* cyclings (given the mutant\'s isolation phenotype \[[@B21]\]), elementary scrutiny of correlogram plots increases the chance that data from a given specimen could be judged rhythmic, thereby decreasing the likelihood that we would go overboard in evaluating mutationally induced damage to the clock system. Figure [2](#F2){ref-type="fig"} provides an example of why we use a system of qualitative criteria to assess rhythmicity; while the correlogram of a pair of forelegs shown on the top row indicates a statistically significant rhythm, the foreleg pair shown on the bottom row is tabulated as rhythmic even though the correlogram does not achieve statistical significance. In an effort to include any specimen that could be judiciously apprehended as rhythmic, we applied this conservative criterion to obtain the data shown in Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}. We also show our analysis in graphic form for data averaged from each tissue of a given type to illustrate effects of *cry*^*b*^ in cases for which only the rhythmic specimens were included (Figures [3](#F3){ref-type="fig"} and [4](#F4){ref-type="fig"}; also see Materials and Methods). ![Comparison of *tim-luc; cry*^+^ foreleg specimens illustrates rhythmicity with or without statistical significance. From left to right: the first column shows raw data on the ordinate plotted over time on the abcissa, the second column shows detrended and normalized data plotted over time and the third column shows autocorrelation with p = 0.05 confidence interval depicted by a gray area centered at 0. The top row shows the analysis of a pair of forelegs with robust rhythmicity. The autocorrelation function is significantly rhythmic as indicated by the asterisk and the strength of rhythmicity (RI) is 0.52 (see \[[@B27]\]). The bottom row shows a signal with weaker rhythmicity. While the shape of the data plots and the autocorrelation function is consistent with a rhythmic signal, the height of the third peak (with asterisk) fails to achieve statistical confidence. At 0.05, the RI is an order of magnitude weaker than the signal shown in the top row. Nevertheless, this signal is tabulated as rhythmic (see text and \[[@B27]\] for further detail).](1471-2202-3-5-2){#F2} {ref-type="table"} or Table [2](#T2){ref-type="table"}, respectively. Counts-per-second of bioluminescence are plotted on the vertical axis with time in hours given on the horizontal axis. The shaded region around the mean activity line indicates standard error of the mean (SEM). The column labeled \'normalized data\' shows plots of the averaged detrended and normalized data as described under Figure [1](#F1){ref-type="fig"} (also see Materials and Methods and \[[@B27]\]). The shaded region represents the SEM. Units on the ordinate are arbitrary and the plot is centered around a mean of 1. Time (in hours) is represented on the abcissa. The column labeled \'autocorrelation\' shows correlograms for the normalized data. Correlation coefficients are plotted on the ordinate with a range of values from -1 to 1. The gray region centered around 0 describes a 95% confidence interval. The lag of the autocorrelation function is plotted on the abcissa. An asterisk is placed above the third peak of the autocorrelation function. The value at that point defines the Rhythmicity Index (RI), an estimate of the strength of rhythmicity \[[@B27],[@B31]\]. When the asterisk is not present, the autocorrelation function indicates a lack of rhythmicity. Values for the RI appear in the lower left corner of these plots along with a related number called the Rhythmicity Statistic (RS). The RS value is the ratio of the RI to the absolute value of the confidence line. This metric indicates that the rhythmicity described by the correlogram is statistically significant when the value is ≥ 1 (see Materials and Methods for more detail). The column labeled \'mesa\' shows a spectral analysis of the data that provides an estimate of the period \[[@B37],[@B27]\]. Spectral density is given in arbitrary units on the ordinate and the range of periods we assess is shown on the abcissa. Asterisks are placed over the highest peak shown in a range between 18--30 hours. Although we take this value as the estimate of circadian period, there may be other periodicities present within the horizontal range (the width) of the peak or elsewhere on the plot and these additional rhythmic components are also present in the data. Absence of an asterisk indicates either the absence of a peak or that any peak within the plot occurs outside the circadian range. Note that the autocorrelation plot is used to determine rhythmicity and mesa is used to provide an estimate of the period only when warranted by correlogram (for more details see Materials and Methods and \[[@B27]\]). a.) averaged data for *cry*^+^ or *cry*^*b*^ antennal pairs. These are assessed for rhythmicity on a specimen by specimen basis as tabulated in Table [2](#T2){ref-type="table"}. b.) same as a.) for isolated heads as tabulated in Table [1](#T1){ref-type="table"}. c.) same as b.) for isolated bodies. This analysis is continued in Figure 4.](1471-2202-3-5-3){#F3} {ref-type="fig"}. d) Averaged data for *cry*^+^ or *cry*^*b*^ wings. These are assessed for rhythmicity on a specimen by specimen basis as tabulated in Table [1](#T1){ref-type="table"}. e)same as d) for isolated forelegs.](1471-2202-3-5-4){#F4} ::: {#T2 .table-wrap} ::: {.caption} ###### Effects of the *cry*^*b*^ mutation on *BG-luc* and *tim-luc* reporter activity in antennal specimens under light-dark cycles (LD 12:12), constant darkness (DD and constant temperature (HH)^1,2^ ::: Genotype^3^ N~a~^4^ Number^5^ Rhythmic (statistical significance) Period (Mean ± SEM) Rhythmicity^6^ Index (Mean ± SEM) Rhythmicity Statistic (Mean ± SEM) Rhythm Amplitude (Mean ± SEM) Phase Activity Counts (MeanfSEM) --------------------- --------- ----------------------------------------------- --------------------- ----------------------------------- ------------------------------------ ------------------------------- ---------- ---------------------------- LD *BG-luc; cry*^+^ 36 33 (23) 24.2 ± 0.3 0.49 ± 0.03 3.0 ± 0.2 0.17 ± 0.01 2.8, 0.9 335 ± 33 *BG-luc; cry*^*b*^ 41 16 (3) 25.8 ± 1.0 0.12 ± 0.06 0.8 ± 0.4 0.09 ± 0.02 8.2, 0.9 299 ± 65 *tim-luc; cry*^+^ 20 16 (9) 25.0 ± 0.3 0.36 ± 0.05 2.3 ± 0.3 0.15 ± 0.02 2.1,0.9 309 ± 30 *tim-luc; cry*^*b*^ 36 11 (1) 26.5 ± 1.0 0.23 ± 0.05 1.4 ± 0.3 0.09 ± 0.01 9.2, 1.0 477 ± 105 DD *BG-luc; cry*^+^ 38 33(4) 25.7 ± 0.4 0.20 ± 0.03 1.2 ± 0.2 0.12 ± 0.01 5.5, 1.0 404 ± 27 *BG-luc; cry*^*b*^ 31 14(1) 26.8 ± 1.3 0.03 ± 0.04 0.2 ± 0.2 0.02 ± 0.00 10.5,0.3 350 ± 75 *tim-luc; cry*^+^ 23 22(0) 26.4 ± 0.4 0.12 ± 0.03 0.7 ± 0.2 0.08 ± 0.00 5.4, 0.8 269 ± 40 *tim-luc; cry*^*b*^ 37 8(0) 24.3 ± 2.2 -0.1 ± 0.09 -0.6 ± 0.5 0.04 ± 0.00 7.9, 0.5 625 ± 109 HH *BG-luc; cry*^+^ 112 94(5) 25.4 ± 0.3 0.12 ± 0.01 0.7 ± 0.1 0.08 ± 0.00 -0.1,0.6 487 ± 29 *BG-luc; cry*^*b*^ 107 28(2) 25.4 ± 0.5 0.14 ± 0.03 0.9 ± 0.2 0.06 ± 0.00 0.9, 0.6 624 ± 98 *tim-luc; cry*^+^ 112 74(0) 25.3 ± 0.3 0.14 ± 0.01 0.8 ± 0.1 0.09 ± 0.00 0.2, 0.7 902 ± 73 *tim-luc; cry*^*b*^ 80 37(4) 25.8 ± 0.2 0.11 ± 0.02 0.7 ± 0.1 0.08 ± 0.00 -1.4,0.5 674 ± 90 **1)** The experimental conditions are described in Krishnan et al \[[@B23]\] and in Materials and Methods. **2)** Ambient conditions are light-dark cycles with 12 hours of light and 12 hours of darkness (LD 12:12) or constant darkness at 25°C or constant temperature in which the specimens were first synchronized to temperature cycles in constant darkness consisting of 12 hours of 18°C followed by 12 hours of 27°C before they were evaluated under a constant high temperature of 27°C (HH). See Materials and Methods for further details. Also see Krishnan et al. \[[@B23]\]**3)** All headings in Table [2](#T2){ref-type="table"} are defined as in Table [1](#T2){ref-type="table"}. **4)** The analysis was applied to samples previously described in Krishnan et al \[[@B23]\]. The number of specimens analyzed (N~a~) in LD and DD is unchanged, while the number of specimens in HH has increased by 123 additional *cry*^+^ antennal pairs and 62 additional *cry*^*b*^ specimens (see Materials and Methods for more details) **5)** We asked whether *cry*^*b*^ affects rhythmicity in antennal pairs for each ambient condition. Chi-squared tests showed significant effects of the mutation on rhythmicity with either *BG-luc* or *tim-luc* in all cases (p \< .001). **6)** The *cry*^*b*^ mutation significantly reduced the RI value for each reporter in LD and DD (t-test, p \< .001) but not HH (p \> .05). ::: The *cry*^*b*^ mutation significantly reduced the number of rhythmic specimens for each of the body parts we assayed as measured by both reporters in LD 12:12, with the exception of the wing (see Table [1](#T1){ref-type="table"} and Table [2](#T2){ref-type="table"}). Moreover, whereas, on average, all of the *luc*-reported cycles were significantly rhythmic in a *cry*^+^ background, the minor fraction of *cry*^*b*^ specimens that appeared rhythmic (based on the shape of the correlogram) were not always significantly rhythmic on average. Mutant rhythmicity that passed muster in that manner was evident for both reporters only in the wing and antennal pairs. In addition, *BG-luc; cry*^*b*^ (but not *tim-luc; cry*^*b*^) cycles were significantly rhythmic in the forelegs. Otherwise, the mutation eliminated significant rhythmicity. We were interested in whether the mutant specimens that retained rhythmicity appeared normal. In this regard, a potentially more informative indication of the *cry*^*b*^ effect was provided by the Rhythmicity Index (RI), an estimate of a rhythm\'s strength that draws on the autocorrelation read-outs (see Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}; see Materials and Methods as well as \[[@B27],[@B31]\] for further discussion of this parameter). RI values indicated that the strengths of rhythmicity from the LD 12:12 monitorings were significantly lowered by *cry*^*b*^ for both reporters and in all of the body parts (Table [1](#T1){ref-type="table"} and Table [2](#T2){ref-type="table"}). Thus, the *cry*^*b*^ mutation reduces the strength of cycling in these body parts; in other words, these rhythms are less robust than in the *cry*^+^ controls. In Figures [3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"},[5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"}, plots of average counts for each tissue are presented, based on the rhythmic subset of specimens tabulated in Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"} for LD 12:12. These average signals indicate once again that *cry*^*b*^ impinges on clock-gene cyclings. For example, with the exception of the *BG-luc*-reported wing data shown in Figure [4d](#F4){ref-type="fig"}, simple inspection of the raw data column suggests that *cry*^*b*^ reduces or eliminates such cyclings under LD 12:12 (compare the raw data for *cry*^+^ v *cry*^*b*^ for example, in Figure [3a, 3b](#F3){ref-type="fig"}, or [3c](#F3){ref-type="fig"}). The appearance of rhythmicity in the *cry*^*b*^ specimens is mostly revealed after detrending and normalization (see Figures [3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"},[5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"} the second column labeled normalized). However, in the case of heads and bodies -- for which there were very few rhythmic *cry*^*b*^ samples (Table [1](#T1){ref-type="table"}) -- no rhythmicity is evident in the averaged plots. One explanation for this could be that the separate rhythms are not in phase with one another or are noisy; thus the average may not appear to be rhythmic. Nevertheless, applying autocorrelation to the averaged time-courses revealed daily cyclings to have occurred in all cases except for the average taken from the rhythmic subset of *BG-luc; cry*^*b*^ bodies. These results show that rhythmicity is evident in the mean signal as well as when it was initially tabulated on a specimen by specimen basis (see Materials and Methods). Nevertheless, there might not have been agreement between these two views of rhythmicity for any of the tissues (or between different approaches to the analysis of phase as described below). Although it is unlikely that arrhythmic specimens would give rise to a mean rhythm, it is conceivable that rhythmic specimens might give rise to an arrhythmic mean. {ref-type="fig"}. Experimental conditions are the same as in Figure [3](#F3){ref-type="fig"} with the exception that the reporter is *tim-luc.* The order of presentation is the same as in Figure [3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"}: a.) averaged data for *cry*^+^ or *cry*^*b*^ antennal pairs assessed as rhythmic on a specimen by specimen basis as tabulated in Table [2](#T2){ref-type="table"}. b.) and c.) show rhythmic specimens from averaged rhythmic *cry*^+^ isolated heads and bodies respectively. Note that there are no rhythmic *cry*^*b*^ heads or bodies to analyze because none were found as shown in Table [1](#T1){ref-type="table"}. d.) isolated wings as tabulated in Table [1](#T1){ref-type="table"}. This analysis is continued in Figure [6](#F6){ref-type="fig"}.](1471-2202-3-5-5){#F5} {ref-type="fig"}. e.) averaged data for *cry*^+^ or *cry*^5^ rhythmic isolated forelegs as tabulated in Table [1](#T1){ref-type="table"}.](1471-2202-3-5-6){#F6} Output from phase analysis -------------------------- We measured the time of peak luciferase activity for each of the five daily cycles included in the monitoring and analysis of a given specimen (see Materials and Methods). A mean peak time (per day) was thereby calculated for each specimen, with such a value taken as the representative phase for that specimen. One goal of quantifying the molecular phases was to compare the normal *tim-luc* time-course values to the *BG-luc* ones for each tissue in a *cry*^+^ genetic background. For this, we used circular statistics \[[@B27],[@B32]\]. Briefly, the phase of each specimen is plotted as a time point (Figures [7](#F7){ref-type="fig"},[8](#F8){ref-type="fig"}). An average vector is calculated based on the distribution of phase points around the unit circle. The angle of the vector corresponds to the mean phase for the group of points, and the magnitude of the vector represents the variability in the phase estimates \[[@B32]\]. The Watson-Williams-Stevens statistic was applied to evaluate whether two such vectors are significantly different from one another \[[@B32]\]. Significant differences were evident in the phase of these two clock genes\' cyclical expression under a light-dark cycle, for all tissues except isolated antennal pairs and the heads (see Figure [7](#F7){ref-type="fig"}) and [8](#F8){ref-type="fig"}. ![Circular phase analysis of luciferase expression in isolated body parts. The left column shows phase comparisons for *BG-luc; cry*^+^ (plotted with asterisk) vs *tim-luc; cry*^+^ (plotted with open circles). The right column shows *BG-luc; cry*^+^ (*plotted* as asterisk) vs. *BG-luc; cry*^*b*^ (plotted as open circles). All of the data shown here were collected in a light-dark cycle (LD 12:12). The time of lights on is indicated as 0 h on the circle. On these plots time moves forward in a counter-clockwise direction. Phase estimates for each rhythmic specimen are plotted just outside the unit circle and a mean vector summarizes the phase of the group with the mean direction indicating the time and the magnitude of the vector indicating the dispersion (variability) of the individual estimates (see Materials and Methods and \[[@B27],[@B32]\]). The Watson-Williams-Stevens test returns an F-statistic that is used to evaluate whether the vectors are significantly different from one another\[[@B32]\]. Rayleigh\'s test \[[@B27],[@B32]\] shows each vector is significantly different (p \< 0.05) from the null vector (random distribution) with the exception of the *BG-luc; cry*^*b*^ vectors shown on the right for heads and bodies in b and c respectively. Beneath each circular plot the difference in hours is given, followed by the mean time for each vector with the number of specimens, n, in brackets, and a p-value for the comparison, a.) comparisons for isolated antennal pairs, b.) isolated heads, c.) isolated bodies. Number of rhythmic samples for each group are given in Table [1](#T1){ref-type="table"} and Table [2](#T2){ref-type="table"}. This analysis is continued in Figure [8](#F8){ref-type="fig"}.](1471-2202-3-5-7){#F7} {ref-type="fig"}. d) isolated wings. e.) isolated forelegs. Number of rhythmic samples for each group are given in Table [1](#T1){ref-type="table"}.](1471-2202-3-5-8){#F8} From similar analyses performed on the *BG-luc* data in a *cry*^+^ vs a *cry*^*b*^ background, the mutation was revealed to affect the phase of *BG-luc*-reported cycling in all tissues except bodies (see Figure [7](#F7){ref-type="fig"}, and note that there were only 4 or 5 rhythmic *BG-luc; cry*^*b*^ samples for analysis of heads and bodies, respectively; and that neither one of the mean vectors representing these samples is significant by Rayleigh\'s test). Thus *cry*^*b*^ \'s effect on the phase of these molecular rhythms is as potent as the effect on cycling as such. The analysis of phase presented in Figure [7](#F7){ref-type="fig"} and Figure [8](#F8){ref-type="fig"} relies on an average estimate of phase for each specimen and it neglects any variability in the occurrence of peaktime associated with the daily rhythm in bioluminescence. As noted above, the mean vectors depicted within each panel of Figures [7](#F7){ref-type="fig"},[8](#F8){ref-type="fig"} represent a vector average of the individual phase estimates with the direction of the vector indicating the mean time and the length of the vector indicating the variability (dispersion) of the estimates. An alternative analysis, presented in Figures [9](#F9){ref-type="fig"},[10](#F10){ref-type="fig"}, preserves the intraspecimen variability. This analysis, called a bivariate analysis \[[@B32]\], represents an individual specimen as a vector in the x-y plane. The position of each point (plotted as an asterisk or an open circle) is determined by the variability of the occurrence of the peak for that individual record. Thus, the points that fall on the diameter of the circle are precisely consistent from day to day, while those falling closer to the origin indicate greater variability. The position of the point within the circle describes the mean peak time for each specimen. In this way, each point indicates the head of a vector that summarizes the phase for a particular specimen. The tail of the vector would connect the point to the origin but we do not plot the vectors this way because the figure would be difficult to interpret. A statistical comparison between the two groups (in this case *per-luc* v *tim-luc* or *per-luc; cry*^+^ v *per-luc; cry*^*b*^) is obtained by first calculating a mean vector that connects the origin to the center of the cloud of points defined by the respective group and then testing whether or not these representative vectors are different from one another. {ref-type="fig"}. The left column shows phase comparisons for *BG-luc; cry*^+^ (plotted with asterisk) vs *tim-luc; cry*^+^ (plotted with open circles). The right column shows *BG-luc; cry*^+^ (*plotted* as asterisk) vs. *BG-luc; cry*^*b*^ (plotted as open circles). The axes for each plot describe an x-y plane with the origin occurring at the center of the unit circle plotted within the plane for reference. The point (0,1) defines the beginning of the subjective day, or time 0. Time moves in a counter clock-wise direction on this circle. Each point denotes the head of a vector that summarizes the phase of an individual specimen. The tail of this vector would extend from the origin to the plotted point, with the direction indicating the mean peak time across cycles and the magnitude (distance from the origin) describing the variability of the peaks for each specimen. However, these tails are not plotted to simplify the appearance of the figure. A mean vector is calculated and fully plotted to show the mean phase time for each group of points as well as the mean variability (indicated by its length). Below each plot: the length of each mean vector is given by r; the mean time of each vector is given by phi and the p-value used to assess whether the phase differs between groups is obtained by Hotelling\'s two-sample test (for more detail see \[[@B32]\]). This analysis is continued in Figure [10](#F10){ref-type="fig"}.](1471-2202-3-5-9){#F9} {ref-type="fig"} legend for details.](1471-2202-3-5-10){#F10} The results shown in Figures [9](#F9){ref-type="fig"} and [10](#F10){ref-type="fig"} are consistent with those in Figure [7](#F7){ref-type="fig"} and [8](#F8){ref-type="fig"}, but the interpretation of the figures is different. For example, Figure [8e](#F8){ref-type="fig"} (on the right) shows that the *cry*^*b*^ mutation affects the mean peak phase significantly; this difference is evident in the distribution of the mean phase estimates estimates plotted outside the circle on the right in Figures [7](#F7){ref-type="fig"},[8](#F8){ref-type="fig"}. When these same *luc* records are plotted using the bivariate method as shown in Figures [9](#F9){ref-type="fig"},[10](#F10){ref-type="fig"}, both the temporal distribution and the intraspecimen variability determine the placement of individual points within the circle (on the right in Figure [10e](#F10){ref-type="fig"}). The plot in Figure [10e](#F10){ref-type="fig"} shows much more detail about the rhythmic output generated by each specimen; it is clear that the *cry*^+^ forelegs produced a stable rhythm in nearly all of the 33 cases as indicated by the presence of the asterisks on or near the edge of the circle. On the other hand, it is readily apparent from this plot that the 11 rhythmic *cry*^*b*^ specimens are more variable in the daily timing of peak phase because more than half of the points representing this group are plotted within the circle, relatively closer to the origin than the *cry*^+^ points. Nevertheless, the statistical test that corresponds to the treatment of data depicted in Figures [9](#F9){ref-type="fig"},[10](#F10){ref-type="fig"} only tells us that the phases are different; it is not possible to determine whether this difference comes from the temporal distribution or the variability or a combination of the two. Therefore, the interpretation of these two different phase analyses is that there is a significant effect of *cry*^*b*^ on the peak phase of the clock in the foreleg (Figures [7](#F7){ref-type="fig"},[8](#F8){ref-type="fig"}) and that there is also an effect of this variant on the mean occurrence of peak phase (Figures [9](#F9){ref-type="fig"} and [10](#F10){ref-type="fig"}). Effects of *cry*^*b*^ on antennal rhythms in constant darkness and constant temperature --------------------------------------------------------------------------------------- Temporally varying LUC activity was examined for antennal pairs in constant darkness (DD) to assess whether *cry*^*b*^ affects free-running rhythmicity. As noted above, these rhythms may appear to be weaker in constant conditions than under LD 12:12 (see Table [2](#T2){ref-type="table"}). Earlier studies have shown that the amplitude of rhythms in DD is reduced compared to LD \[[@B20]\]. These observations could reflect a weakening of synchrony between, for example, the two antennae within a well, or between multiple clock cells within a specimen. But the method for monitoring *luc* activity can also contribute to the appearance of damping within a record. Linear as well as nonlinear trends in the data come into play in experiments using luciferase reporter technology \[[@B19],[@B20]\]. Such artifacts were shown (in one substrate replenishment experiment) to be the result of inevitable depletion of luciferin from the medium surrounding the tissues over time \[[@B19]\]. Removal of these trends quantitatively clarifies the record of bioluminescent activity with regard to any rhythmicity that might be present \[[@B27]\]. The normalization procedure we use (see Figure [1](#F1){ref-type="fig"} and \[[@B27]\] for more detail) exposes the rhythm (if one is present) in a manner that is quantifiable without reference to the type of measurement that has been employed (for example, counts-per-second metric on the ordinate of the relevant plots becomes dimensionless). One salutary consequence is that comparisons between groups -- differing, for example, by body part-are based on the nature of the rhythmicity per se (independent of tissue-dependent levels of overall reporter expression). Therefore, the use of computational techniques to detrend and normalize the signals permitted us to extract and characterize rhythms for comparative analysis -- clearing out other features that were deemed relatively unimportant in terms of the manner by which tissue-type, background genotype, or transgene type would influence the appreciation of these clock-gene cyclings. As indicated in Table [2](#T2){ref-type="table"}, the effects of *cry*^*b*^ in DD are consistent with its effects in LD 12:12. In constant conditions, the mutation significantly reduced the number of rhythmic antennal pairs for both reporters. However, degrees of rhythmicity were also impinged on by the conditions alone, in that only the *BG-luc; cry*^+^ cycles were significantly rhythmic on average. From a more quantitative perspective, the mean RI value resulting from the DD monitorings was decreased by an order of magnitude in the *cry*^*b*^ mutation compared to *cry*^+^ for *tim-luc* as well as *BG-luc* activity (see Table [2](#T2){ref-type="table"}). This effect of *cry*^*b*^ is further underscored by an analogous effect on rhythm amplitude. The measures of mean amplitudes given in Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"} stem from computing the distances (as graphed) from the peak (or trough) to the \"mean line\" running through plots of the detrended and normalized signals, *cry*^*b*^ caused amplitude decrements for the time-courses mediated by expression of both *luc*-fusion transgene types (Table [2](#T2){ref-type="table"}). This *cryptochrome* mutation produced dramatic effects on the phase of both reporters in DD: The mean-phase values were estimated to occur more than 7 hours earlier for *cry*^*b*^ antennae compared to the *cry*^+^ appendages. This effect of *cry*^*b*^ indicates that the gene may contribute to regulation of the endogenous phase for the antennal clock. Alternatively, the extreme difference in phase could indicate the loss of synchrony between multiple clock cells within the antennae consistent with the large decrease seen in the RI value. These results suggest that *cry* may normally play a role in the determination of clock time or, alternatively, that it may act to synchronize clock time between individual clock cells. Previous studies of cryptochrome\'s chronobiological role in *Drosophila* emphasized that CRY may function as a deep brain photoreceptor involved in mediating resetting of the clock by light \[[@B26]\]. Support for this hypothesis came from the observation that locomotor-activity rhythms in *cry*^*b*^ flies are unaffected in constant darkness and also maintain rhythmicity under constant light, a condition that causes behavioral arrhythmicity in the wild type \[[@B33]\]. While antennal rhythms of gene expression and of odor sensitivity in DD are significantly attenuated in *cry*^*b*^ tissues (\[[@B23]\], unlike locomotor rhythmicity \[[@B21]\]), it remained possible that such effects are due to an altered role of light on the developing clock. To address this possibility, we reared experimental subjects and prepared antennal samples in a temperature cycle (27°C:18°C) in DD and then tested the antennal samples in constant darkness at a constant warm temperature 27° (here called HH -- for constant high temperature -- to distinguish this experiment from the standard LD rearing-\>DD tests described above). As shown in Table [2](#T2){ref-type="table"}, the effects of *cry*^*b*^ on molecular rhythmicity in HH are similar to what was observed in the LD and (standard) DD monitorings. The number of rhythmic specimens was significantly reduced by the mutation for both *BG-luc-* and *tim-luc-*reported rhythms. However, very few of the rhythmic specimens, among either *cry*^+^ or *cry*^*b*^ antennal pairs, was significantly periodic (see Table [2](#T2){ref-type="table"}). Further, unlike what was found in the other conditions, the RI values stemming from the HH test were similar across genotypes and reporters. Similiarly, there was no effect of the *cry*^*b*^ mutant on the amplitude of the normalized rhythmic output of our reporters in these experiments (Table [2](#T2){ref-type="table"}). A combined effect of the temperature regimen together with that of *cry*^*b*^ was detected on the phase of molecular cycling: A mutationally induced difference of 2.5 hours was observed in HH as opposed to 7 hours in DD. The direction of the *cry*^*b*^-induced phase modification was the same as that observed in DD at a lower temperature. More generally (and aside from the genotypic effect), entrainment, followed by subjecting these isolated appendages to a constant high temperature, caused a phase effect: Peak times of LUC-reported cyclings were shifted 4 or 6 hours earlier in the HH tests -- respectively, for *BG-luc; cry*^+^ and *tim-luc; cry*^+^ antennae -- compared with the results of LD rearing and DD monitoring. Discussion ========== We have developed a tool kit for the analysis of cyclical data \[[@B27]\] and here have applied it to molecular data involving circadian clocks in isolated tissues of *Drosophila.* Two features of this approach: the evaluation of rhythmicity and the evaluation of phase, have been emphasized. The first prominent feature is the application of filters for normalizing and detrending the raw time-course data. Such a preparation may indicate whether or not a rhythmic signal is present, as shown in Figure [1](#F1){ref-type="fig"}. Subsequent criteria are evaluated by means of the autocorrelation function to determine whether a rhythm is present in the data and, if so, whether it is statistically significant. However, we do not rely on statistical significance as an absolute test of rhythmicity. On the one hand, statistical confidence provides a quantitative standard that removes investigator bias; on the other hand, biological signals can fail this test while still displaying the features of rhythmicity. We have adopted a strategy for addressing this complexity based on several criteria as discussed in the companion paper \[[@B27]\]. The overall approach is somewhat subjective, yet it is systematic and does not rely on an arbitrary decision about periodicity in the data. We evaluate each record -- blind to treatment or genotype -- as raw data or normalized/detrended data. If the shape of the autocorrelation function displays periodicity (a sinusoidal pattern) in the circadian range, we accept the presence of rhythmicity and quantify the strength of the rhythm using the RI value. In addition, we keep track of which samples are significantly rhythmic and which samples are not with the RS metric (see Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"} for examples). Although we go into this matter more thoroughly in the companion paper, we reiterate here that this subjective use of the autocorrelation function to establish rhythmicity is not unprecedented; in fact, it has been recommended as a valid approach to situations like our experiments where the number of data points is relatively small \[[@B30]\]. In general, it must be noted that the use of both tabulated and averaged data also supports this approach. For example, the average of the RI values given for the rhythmic *per-luc* antennal samples in LD as indicated in Table [2](#T2){ref-type="table"} is lower than the RI values shown for the mean signal presented in Figure [3a](#F3){ref-type="fig"}. This difference occurs because the rhythmicity expressed by individual samples within the respective groups is similar and the averaging process (portrayed in the plotted data shown in Figure [3a](#F3){ref-type="fig"}) tends to strengthen the rhythms in question and to eliminate idiosyncracies within a given record. Thus, we have quantified how the *cry*^*b*^ mutation affects molecular rhythmicity in two ways. First, it reduces the number of rhythmic peripheral-tissue specimens in all cases, except -- and most intriguing -- for the wing. Second, even when only the rhythmic *cry*^*b*^ versus *cry*^+^ specimens are considered, the former genotype causes a significant reduction of RI values, with the exception of antennal pairs whose molecular time-courses were monitored at constant temperature. A variety of claims have been made about the phase of clock gene expression in various tissues and cell types. While immunohistochemical studies have suggested that the course of wild-type *per* cycling may differ between photoreceptors and neurons in the central brain \[[@B34]\], molecular studies have suggested that there is no difference between the phase of *per* mRNA cycling in photoreceptors vs. the rest of the head (35). The interpretation of such studies is complicated by the inability to assess phase over several cycles within a specimen. However, even when such measurements have been facilitated by the use of real-time monitoring of *luc*-reporters, there has been an assertion that phase is essentially the same amongst isolated tissues \[[@B19]\]. Nevertheless, given our observation that the effects of the *cry*^*b*^ variant may be tissue-specific, as well as the use of both *per-luc* and *tim-luc* reporters, we re-examined this question quantitatively. Thus, the second analytic feature we emphasize is the analysis of phase. We have applied two complementary approaches. The first approach uses circular statistics to evaluate the distribution of peak phase estimates. The second approach uses a bivariate analysis that reveals intraspecimen variability as well as mean peak time to test differences in phase. Specifically, we applied these methods to ask whether the two *luc* reporters (*BG-luc* and *tim-luc*) peak at the same time in a *cry*^+^ (and otherwise clock-normal genetic background) and also whether the *cry*^*b*^ mutation influences peak phase of the *BG-luc* reporter. The observation of tissue-specific phase relationships between *per-luc* and *tim-luc,* along with the apparent difference in the absolute timing of phase is not surprising. Such differences could be explained both on the basis of differences in cell types and differences in function. Whereas peripheral clocks are not necessarily neuronal \[[@B36]\], it is conceivable that tissue specific differences in the intracellular environment would force different peak times on the clock mechanism. Further, based on the tissue specific effects of *cry*^*b*^ it is possible that the configuration of a clock mechanism may vary between different cell types. For example, the full ensemble of clock factors present within these various tissues is not yet known. Conclusions =========== Following our previous development of a collection of analytical tools for the study of molecular and behavioral cycles \[[@B27]\], we now provide intensively performed examples of the analysis of molecular rhythms in isolated tissues from *Drosophila.* In the main, but by no means exclusively, the current analyses were applied to published data revolving round effects of the *cryptochrome* gene as it functions in *Drosophila\'s* circadian system \[[@B23]\]. The results of the present study show that procedures for detrending and normalizing the data clarify the presence versus absence of rhythmicity -- which tend strongly to be associated with *cry*^+^ functioning versus that putatively remaining in the *cry*^*b*^ mutant -- as well as permitting a more detailed understanding of the features accompanying a given molecular time-course. Our conclusions from extended and deepened analyses of these clock-gene cyclings are consistent with those in the earlier report \[[@B23]\], including the matter of peripheral-tissue rhythms being dependent in part on CRY-mediated functions. Moreover, current applications of our phase analysis uncovered new phenomena. These components of the tool-kit reveal that *cry*^*b*^ alters the phase of *period* and *timeless* gene expressions, suggesting further that -- at least in the periphery -- CRY plays a quasi-central role in molecular timekeeping. More broadly, in peripheral tissues whose genotype was clock-normal, phase differences that were evident among the different body parts strengthen the notion that overall features of the timekeeping mechanisms vary according to the local molecular milieu. This is perhaps as it should be -- because why would a structure such as the antenna necessarily wish to exhibit a maximum or mininim value for a given piece of clock-output that might be the same as those elaborated within the fly\'s legs? The development of analytical strategies for analyzing biological time-series will continue to be important as new real-time methods (potentially including images of tissues) provide greater access to molecular and physiological cycles. Such strategies add depth as well as detail to the study of biological time-keeping mechanisms. Materials and Methods ===================== Peripheral tissues were dissected from *D. melanogaster* adults as described in Krishnan et al \[[@B23]\]. The *per-luc (BG-luc)* and *tim-luc* transgenic strains were originally described elsewhere \[[@B20],[@B21]\]. Automated monitorings of luciferase (LUC) activity were the same previously described \[[@B20],[@B23]\]. All of the data from light-dark cycles (LD 12:12) and from constant darkness (DD) were from the LUC monitorings reported in Krishnan et al \[[@B23]\], although most of the pertinent data were not separated as to *per-luc* versus *tim-luc* expressions, and many of the results were included as supplementary material. Additional *cry*^+^ specimens in Table [1](#T1){ref-type="table"}, which did not appear in the earlier report, have been included here, substantially to augment the appreciation of normal clock-gene cycling in isolated tissues. We also include data from additional experiments collected under constant high temperature (27°C) \[[@B23]\]. These data were collected after the submission of the earlier report. In these new experiments flies were reared in DD and in temperature cycles (12-hr 27°C: 12-hr 18°C). One-to-three day-old ether-anesthetized males were collected and housed for 2 days in these same conditions before the antennal pairs were dissected. The antennal pairs were placed immediately in luciferin-containing medium in 96-well plates and transferred to the luminometer (as described in \[[@B23]\]), where they were exposed to 2 additional days of the temperature-cycle in DD before being maintained at a constant temperature of 27°C for the duration of the LUC-monitoring runs. The basic flow-chart for handling data from a given specimen and from an ensemble of like ones was as follows: Each record was evaluated for an adequate level of expression above \'no-specimen\' background to insure that the tissue had survived and was producing valid data \[cf. \[[@B23]\]\]. If expression levels were adequate, the specimen was assessed for rhythmicity. Temporally varying LUC activity was detrended and normalized as described in Figure [1](#F1){ref-type="fig"}. Next, autocorrelation analysis \[[@B30]\] was performed to look for a rhythmic pattern of activity. The correlogram is a plot that reports correlation-coefficient values on the vertical axis versus time-lags plotted on the horizontal axis. Zero lag always provides a value of 1, because the correlation is perfect between the activity record and a copy of itself at each point. However, as one copy of the activity record is shifted along the time axis (in either direction) by one point (in other words, a lag of 1 hour), the correlation falls off. This process is repeated, and a correlation coefficient is calculated for the activity record against itself for each lag shown in the correlogram (see bottom row of Figure [1](#F1){ref-type="fig"} for an example). If the correlation coefficients fall and rise in a periodic fashion, resembling a sinusoidal curve, the record of LUC activity is \"rhythmic.\" As to whether such periodicity is significant, a 95% confidence interval is calculated based on the number of observations in the experiment. Rhythmic activity is statistically significant if the peaks and troughs of the autocorrelation function cross the boundary set by the confidence interval centered at 0 on the correlogram (for further detail see \[[@B27]\] and references therein). As noted above in Results text, we do not require statistical significance to score a record as rhythmic. The appearance of rhythmicity in the correlogram is applied as a qualitative criterion so that we do not treat weakly rhythmic data as arrhythmic data. The justification for this strategy is given in Levine et al \[[@B27]\]. In addition, we keep track of the number of specimens that generated statistically significant rhythms as indicated in Table [1](#T1){ref-type="table"} and Table [2](#T2){ref-type="table"}. We quantify the strength of a rhythm with the aforementioned Rhythmicity Index. RI is the height of the third peak of the correlogram \[[@B27],[@B31]\]. Whereas the confidence interval for the correlogram is based solely on the number of observations, the RI value is based on the robustness of any regular fluctuations in the data. For example, we record LUC activity every hour. If we increased the sampling frequency from once every 60 minutes to once every 30 or 15 minutes, the confidence interval would become smaller while the shape and amplitude of the correlogram would presumably remain unchanged from the curve obtained by 60 minute sampling. The RI value, however, would not vary, while the estimates of statistical significance would vary according to the sampling frequency. It is nevertheless informative to monitor the statistical significance of periodicities; for this we devised a ratio that allows us to track whether or not a specimen is significantly rhythmic, the so-called Rhythmicity Statistic (RS). An RS value is a ratio of the RI value to the absolute value of the 95% confidence line. If RI is equal to or greater than the numerical height of the confidence line, then the rhythm is significant (by definition, the height of the peak is ≥ the height of the confidence interval used to determine statistical significance). Alternatively, if RI is less than the confidence line, the rhythm is not statistically significant. Thus, RS provides a numerical accounting of significance for an individual specimen or an average signal. We define the amplitude of a rhythmic LUC time-course as the (plotted) distance (in arbitrary units) from the mean peak or mean trough of the normalized activity record to 1 (the latter is always the mean value of detrended and normalized data, as exemplified in Figure [1](#F1){ref-type="fig"}). The complete range of the rhythm is twice the amplitude. Most of the analyses applied in the current study were described in detail in the companion paper \[[@B27]\]. We augmented such treatments of the data by developing two additional parameters, the Rhythmicity Statistic (RS) and a metric that allows specification of cycling amplitude. Analyses leading to the rhythmic parameters were applied to individual samples; their averages appear in Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}. In addition, averages of the LUC data computed across all records within a given group (which varied as to body part, *luc* reporter type, and *cry* allele) were obtained; and such averaged time-courses were formally analyzed, as shown in Figures [3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"} and Figures [5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"}. Consequently, there are differences between the Tables and Figures in values of parameters such as the Rhythmicity Index (RI, see below), because the tables specify the means determined from individual-specimen analyses, while the figures provide parameters determined after computing the average time-courses. Maximum Entropy Spectral Analysis (MESA) was employed to estimate the period of a rhythm, i.e., for a given time-course previously determined to be significantly periodic by correlogram. This method is presented in detail elsewhere \[[@B37],[@B27]\]. We have discussed the analysis of phase using circular statistics in the companion paper \[[@B27]\]. Briefly, an average estimate of peak phase, obtained for each specimen, is plotted as a point on a unit circle. A mean vector, extending from the center of the unit circle towards the diameter is calculated for each group of points; the direction of the vector indicates mean peak phase for the group and the length of the vector represents the variability or dispersion between the points (phase estimates for each specimen). The Watson-Williams-Stevens test returns an F-statistic for the comparison of vectors between two groups to determine whether they represent significantly different estimates of phase (see Figures [5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"} and \[[@B27],[@B32]\] for further detail). One disadvantage of this phase analysis is the loss of information about the variability of phase across cycles for each specimen. Therefore, we have also introduced another approach using bivariate statistics to represent the molecular cycles more completely \[[@B32]\]. In this approach, the record of each specimen is represented as the tip of a vector whose direction indicates the mean phase estimate (as above) and whose distance from the origin in the x-y plane represents intraspecimen variability (see Figures [7](#F7){ref-type="fig"},[8](#F8){ref-type="fig"}). Each group appears as a distribution of points. A mean vector extending from the origin to the center of the distribution is calculated to describe the overall mean peak phase and variability of the group of points (see Figures [7](#F7){ref-type="fig"},[8](#F8){ref-type="fig"}, for example). Finally, the two groups are compared using Hotelling\'s two-sample test to determine whether the respective vectors are different. If a significant difference between vectors is obtained, this method does not specify whether the difference is due to the variability, the peak phase estimate or some combination of the two. All of the analyses described above as well as the subsequent output (the figures) come from programs written for the current study and the just-previous one \[[@B27]\] using Matlab6 (Mathworks). This software is available on request.

Background ========== Prolonged or repeated agonist stimulation of G protein-coupled receptors frequently causes a reduction in the response to the agonist \[[@B1]\]. Short-term agonist exposure results in desensitization that is characterized by a rapid (minutes) and reversible uncoupling of the receptor from the G protein followed by sequestration and/or internalization of receptors from the cell surface. Down-regulation occurs after prolonged agonist treatment (hours to days), resulting in a loss of receptor binding sites \[[@B2]-[@B4]\]. The processes involved in homologous desensitization have been extensively investigated for the beta-2 adrenergic receptor, which is the prototypical G protein-coupled receptor. Initial uncoupling of the beta-2 receptor from the G protein after agonist binding is mediated by phosphorylation by G protein-coupled receptor kinase (GRK) of specific residues in the carboxyl tail of the receptor. The phosphorylated beta-2 receptor enhances the binding of beta-arrestin, which not only uncouples the receptor from the signal transduction process but also serves as an adapter protein that mediates entry into the internalization pathway \[[@B5]\], as well as serving as a platform for additional signaling pathways \[[@B6]\]. The mechanisms of beta-2 adrenergic receptor down-regulation appear to involve both an increase in the rate of degradation of the receptor as well as a decrease in the levels of beta receptor mRNA \[[@B7]\]. Down-regulation of the beta-2 receptor is, in part, due to trafficking of the beta-2 receptor to lysosomes via the clathrin-coated pit endosomal pathway \[[@B8],[@B9]\]. The alpha-2 adrenergic receptors also undergo agonist-induced desensitization and down-regulation. Desensitization occurs both in cell culture systems \[[@B10]-[@B12]\] and in the intact animal \[[@B13],[@B14]\]. In general the mechanisms of desensitization for the alpha-2 adrenergic receptors appear to be similar to those for the beta-2 receptor. There are three subtypes of the alpha-2 adrenergic receptor: alpha-2A; alpha-2B; and alpha-2C \[[@B15]\]. The human alpha-2A subtype undergoes phosphorylation and short-term desensitization \[[@B11],[@B12]\], whereas the human alpha-2C subtype is not phosphorylated \[[@B16]\] and does not desensitize \[[@B11]\]. The human alpha-2C subtype, however, does undergo arrestin-dependent internalization \[[@B17]\]. Many studies \[[@B11],[@B18]\], including those from our laboratory \[[@B10],[@B19],[@B20]\], have demonstrated agonist-induced down-regulation of alpha-2 receptors, although much less is known about the specific mechanisms for down-regulation as compared to the beta-2 receptor. In contrast to alpha-2A and alpha-2B adrenergic receptors from several species, the human alpha-2C receptor does not appear to down-regulate in response to agonist treatment in transfected cell culture systems \[[@B11]\]. By contrast, the opossum alpha-2C receptor in the OK cell line does down-regulate in response to norepinephrine \[[@B19],[@B20]\]. The reason for this difference is the focus of the current studies. The human alpha-2A adrenergic receptor has an acidic serine-rich motif (EESSSSD) in the third intracellular loop that has been shown to be important in agonist-induced desensitization but not in agonist-induced sequestration/internalization. Site-directed mutagenesis studies indicate that all four of these serines are phosphorylated by GRK, and all four are needed for desensitization \[[@B12]\]. The influence of this motif on down-regulation was not investigated, however. The opossum alpha-2C adrenergic receptor has a very similar acidic serine motif (EESSTSE) in the third intracellular loop (Figure [1](#F1){ref-type="fig"}; \[[@B21]\]), whereas the human alpha-2C receptor has a DESSAAAE sequence in the same region (Figure [2](#F2){ref-type="fig"}; \[[@B22]\]). Thus, we postulated that the lack of down-regulation in the human alpha-2C was related to having only two rather than four serines in the motif, and furthermore that the ability of the opossum alpha-2C to undergo down-regulation was dependent on having all four serine/threonine residues in the EESSTSE motif. To test this hypothesis, we mutated the opossum alpha-2C serines and threonine residues to alanines and valine, respectively, and tested whether the expressed receptors would down-regulate. Down-regulation was not observed in the mutated receptors following treatment with norepinephrine for 24 h. {#F1} {#F2} Results ======= Sequence of the opossum alpha-2C adrenergic receptor ---------------------------------------------------- In the process of confirming the mutations, we also completely re-sequenced the wild-type clone and found several differences from the originally published sequence (GenBank \#U04310 \[[@B21]\]). The corrected sequence differs from the original in 19 bases, which results in 11 amino acid residue changes as detailed in Figure [2](#F2){ref-type="fig"}. The sequence has been corrected in Genbank. Characterization of mutated receptors ------------------------------------- The results of \[^3^H\]rauwolscine saturation experiments using the wild-type and mutated receptors from two independent transfections of both mutations are given in Table [1](#T1){ref-type="table"}. The K~d~ values for the two mutants (0.071--0.118 nM) were similar to those of the wild-type (0.113 nM) and similar to the previously reported K~d~ of 0.056 nM for the wild-type alpha-2 receptor for \[^3^H\]rauwolscine in NaPO~4~ buffer \[[@B23]\]. Inhibition experiments with \[^3^H\]rauwolscine indicated that the affinity of norepinephrine for the mutated receptors was approximately two-fold higher (lower K~d~ value) than for wild-type receptor (Figure [3](#F3){ref-type="fig"}). ![Norepinephrine inhibition of \[^3^H\]rauwolscine binding. The data points are the means ± SEM of three independent inhibition experiments, each done in duplicate. The concentration of \[^3^H\]rauwolscine was 0.12 ± 0.01 pM. The IC~50~ values determined from these data are 380 nM, 170 nM and 160 nM for the wild-type (SSTS), SSVA and AAVA receptors, respectively. The 100% values were 1223, 755 and 285 cpm for the wild-type (SSTS), SSVA and AAVA mutants, respectively. The means of the K~i~ values determined in the three individual experiments are given in Table [2](#T2){ref-type="table"}.](1471-2210-2-9-3){#F3} ::: {#T1 .table-wrap} ::: {.caption} ###### Affinity (K~d~) and density (B~max~) of mutated alpha-2C adrenergic receptors expressed in CHO cells. ::: **Clone** **K~d~, nM** **B~max~, fmol/mg protein** **n** ------------------ --------------- ----------------------------- ------- Wild-type (SSTS) 0.113 ± 0.008 372 ± 81 7 SSVA clone \#1 0.105 ± 0.005 357 ± 103 3 SSVA clone \#2 0.088 ± 0.009 90 ± 31 4 AAVA clone \#1 0.118 ± 0.012 327 ± 105 3 AAVA clone \#2 0.071 ± 0.009 39 ± 18 4 The K~d~ and B~max~ values are geometric and arithmetic means respectively from nonlinear regression analysis of saturation experiments with over a 100-fold concentration range of \[^3^H\]rauwolscine. Clone \#1 and \#2 were from separate transfections ::: ::: {#T2 .table-wrap} ::: {.caption} ###### Affinity of norepinephrine for the mutated alpha-2C adrenergic receptors expressed in CHO cells. ::: **Clone** **K~i~, nM** **Slope factor** **n** ------------------ -------------- ------------------ ------- Wild-type (SSTS) 184 ± 9 0.79 3 SSVA 90 ± 4^\*^ 0.73 3 AAVA 71 ± 5^\*^ 0.65 3 K~i~ values were derived based on a single-site model and are the geometric means of three independent inhibition experiments, each done in duplicate. The slope factor was determined in a separate analysis based on a variable slope model. The data were not consistently better fit by a two-site model than a one-site model for any of the three clones. The concentration of \[^3^H\]rauwolscine was 0.12 ± 0.01 nM. ^\*^Significantly different (p \< 0.05) than wild-type based on a t-test of the pK~i~ values. ::: Down-regulation --------------- After a 24 h treatment with 0.3 μM norepinephrine, the wild-type receptor down-regulated to 65 ± 10 % of control (p \< 0.05; n = 7). By contrast, no down-regulation was observed with either the SSVA (114 ± 10 % of control) or the AAVA (110 ± 9 % of control) receptors (Figure [4](#F4){ref-type="fig"}). ![Effect of 24 h treatment with norepinephrine on binding of \[^3^H\]rauwolscine. Cell were incubated with 0.3 μM norepinephrine or vehicle for 24 h and membranes prepared. Specific binding to the alpha-2C adrenergic receptor was determined using 0.40 nM \[^3^H\]rauwolscine. The values are mean ± SEM for 7 experiments using two different clones. Each experiment consisted of three culture dishes with norepinephrine and three without norepinephrine for each clone and each dish was assayed in duplicate. The amount of specific binding in the wild-type following the norepinephrine pretreatment was significantly different from the binding in the control by a t-test (p \< 0.05). The binding levels of the two mutant clones were not different following norepinephrine pretreatment as compared to control.](1471-2210-2-9-4){#F4} Discussion ========== All three alpha-2 adrenergic receptor subtypes from nearly all species that have been tested undergo agonist-induced down-regulation \[[@B24],[@B25]\]. Interestingly, we have shown that the potency of norepinephrine to produce down-regulation differs among alpha-2 receptor subtypes \[[@B19],[@B24]\]. The concentration of norepinephrine required to produce half-maximal down-regulation is approximately 2 μM for the alpha-2A receptor. In contrast, norepinephrine is more potent at down-regulating the alpha-2B and alpha-2C subtypes, with a concentration of only 70 nM required to produce half-maximal down-regulation. This difference in potency appears to be a characteristic of the receptor protein itself, because the cellular environment in which a particular subtype is expressed does not affect the potency of norepinephrine to produce down-regulation. The human alpha-2C receptor is somewhat unique among the alpha-2 receptors in that it does not appear to undergo agonist-induced down-regulation \[[@B11]\]. The human alpha-2A adrenergic receptor, which does undergo down-regulation, has an acidic serine motif (EESSSSD) in the third intracellular loop. Sequentially mutating each of the four serines to alanine decreased the phosphorylation of the receptor by approximately 25% for each serine, and mutating either the first two or the last two serines to alanines prevented agonist-induced desensitization, but not agonist-induced sequestration/internalization \[[@B12]\]. The influence of mutations in this motif on down-regulation has not been investigated. The human alpha-2C receptor, by contrast, has a DESSAAAE sequence in the same region \[Figure [2](#F2){ref-type="fig"}; \[[@B22]\]\]. Thus, it is possible that the lack of desensitization and perhaps also down-regulation of the human alpha-2C receptor is due to the fact that there are only two hydroxyl residues in the DESSAAAE motif rather than the four hydroxyl residues in the alpha-2A EESSSSD motif. The opossum alpha-2C adrenergic receptor has a EESSTSE acidic serine/threonine motif in the third intracellular loop (Figures [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}), which could be a substrate for GRK based on the EESSSSD motif of the alpha-2A subtype. Thus, we postulated that the lack of down-regulation in the human alpha-2C subtype is related to having only two rather than four serines in the motif, and that mutating the threonine and serine residues at the end of the opossum EESSTSE motif to valine and alanine residues to give EESSVAE might prevent down-regulation of the opossum alpha-2C receptor. Consistent with this prediction, treatment with 0.3 μM norepinephrine for 24 h failed to down-regulate the receptor containing the motif mutated to EESSVAE, although the wild-type receptor with the EESSTSE motif was down-regulated by 35%. The inhibition data indicate that the affinity of norepinephrine for the receptors containing the mutations is slightly higher than for the wild-type. The affinity of norepinephrine for the mutated receptor lacking two of the hydroxyl residues was approximately twice as high as the wild-type. The elimination of the remaining two hydroxyl residues did not have any additional effect on the affinity of norepinephrine. These data eliminate the possibility that the lack of down-regulation in cells containing the mutated receptors is a consequence of the mutants having a lower affinity for norepinephrine than the wild-type. The elimination of the terminal two hydroxyl amino acid residues in the EESSTSE motif is sufficient to prevent down-regulation of the opossum alpha-2C adrenergic receptor. Similarly, mutation of only two serine residues in the human alpha-2A receptor resulted in a loss of desensitization \[[@B12]\]. Because the EESSSSE motif in the human alpha-2A receptor is a known GRK substrate, and because eliminating two potential phosphorylation sites prevents down-reguation, phosphorylation of the opossum alpha-2C receptor may be important in the mechanism of down-regulation of this receptor. By contrast, a deletion of twelve residues of the human alpha-2A receptor (293--304) which includes the EESSSSE motif did not appear to prevent agonist-induced down-regulation \[[@B26]\]. Thus a phosphorylation pathway may not be involved in agonist-induced down-regulation of the alpha-2A adrenergic receptor. In addition to the hydroxyl amino acid residues, acidic residues may also be an important sequence element for alpha-2 adrenergic receptor phosphorylation and regulation \[[@B27],[@B28]\]. The region from 300 to 313 in the opossum alpha-2C receptor, which encompasses the EESSTSE motif, is a high acidic region, with 50% (seven of fourteen) of the residues being acidic amino acids (Figure [2](#F2){ref-type="fig"}). In an effort to determine if four hydroxyl amino acid residues in an acidic environment might be a common motif for phosphorylation and receptor regulation, we searched the G Protein-Coupled Receptor Database for four consecutive hydroxyl residues (either serine or threonine) in a linear sequence that was at least 1/3 acidic with no basic residues which occurred in either the third intracellular loop or the carboxyl terminal tail. As can be seen from the results in Table [3](#T3){ref-type="table"}, this does not appear to be a common motif for G protein-coupled receptors, not even for alpha-2 adrenergic receptors. Indeed, this motif is not found in the alpha-2B subtype which does undergo agonist-induced down-regulation, although as noted above a highly acidic region appears to be important and could be related to the additional negative charge provided by phosphorylation. Thus, it is likely that there are multiple mechanisms that are involved in down-regulation of G protein-coupled receptors. ::: {#T3 .table-wrap} ::: {.caption} ###### G protein-coupled receptors with four hydroxyl amino acid residues in a acid environment. ::: **Receptor** **Species** **Sequence** **% acidic** **Location** -------------- --------------- -------------------------------------- -------------- -------------- Alpha-2C opossum ELEDIELEE**SSTS**E 50 3^rd^ Loop Alpha-2A multiple DLEE**SSSS**E 44 3^rd^ Loop Dopamine insect ** TTTST**PEEPDDE 42 3^rd^ Loop Rhodopsin pig DDEA**STTTS** 33 C terminus Rhodopsin arthopod ESDDVSESA**STTTS**AEE 33 C terminus Rhodopsin vertebrate EEDEA**STSS** 44 C terminus Rhodopsin vertebrate DEDEE**STSS** 56 C terminus Rhodopsin vertebrate DDEEA**SSSS** 44 C terminus cAMP class E Dictyostelium DEG**SSSSS**LSSSDEE 33 C terminus The G Protein-Coupled Receptor Database <> was searched for four consecutive hydroxyl amino acid residues (either serine or threonine) in a linear sequence that was at least 1/3 acidic (aspartic or glutamic) with no basic residues (lysine or arginine) and which occurred in either the third intracellular loop or the carboxyl-terminal tail. ::: Conclusions =========== A motif with four hydroxyl amino acid residues in an acidic environment appears to be important for down-regulation of the alpha-2C adrenergic receptor. The human alpha-2C receptor, which has a DESSAAAE motif with only two hydroxyl residues, does not undergo down-regulation. Modification of the EESSTSE to EESSVAE was sufficient to prevent down-regulation of the opossum alpha-2C adrenergic receptor. Because these are potential GRK phosphorylation sites, our data further suggest that GRK phosphorylation may be involved in down-regulation of the alpha-2C adrenergic receptor. Materials and Methods ===================== Materials --------- Restriction enzymes were purchased from New England Biolabs (Boston, MA). Norepinephrine and other chemicals were from Sigma (St. Louis, MO). Site-directed mutagenesis ------------------------- Mutations of the opossum alpha-2C adrenergic receptor cDNA in pRc/CMV vector described previously \[[@B21]\] was accomplished using the QuickChange™ Site Directed Mutagenesis Kit (Stratagene, LaJolla, CA). The SSTS sequence in the third cytoplasmic loop (amino acid residues 309--312; Figures [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}) was mutated to SSVA using forward and reverse primers with the following sequences (the bases introducing the mutation are underlined), which also introduced a new restriction site: GAACTGGAGGAGAGTAGC[GTGGCT]{.ul}GAGAACCGGCGCAGGAGGC and GCCTCCTGCGCCGGTTCTC[AGCCAC]{.ul}GCTACTCTCCTCCAGTTC, respectively. To screen for the presence of the mutation, the alpha-2 adrenergic receptor insert was cut from the vector using Hind III (895) and EcoR1 (1841) and cut with Ddel. In a similar manner a further AAVA mutation was prepared from SSVA using GACATCGAACTGGAGGAG[GCCGCA]{.ul}GTGGCTGAGAACCGGCGC and GCGCCGGTTCTCAGCCAC[TGCGGCC]{.ul}TCCTCCAGTTCGATGTC forward and reverse primers, respectively. The presence of the second mutation was determined using the AlwNi. To further confirm the sequence, the entire cDNA for the insert was sequenced for each mutation. Cell culture ------------ CHO-K1 cells (American Type Culture Collection) were maintained in Ham\'s F-12 medium (GIBCO) with 10% fetal bovine serum (GIBCO 26140-079) in 95 % air/5% CO~2~ at 37°C in a humid atmosphere. Wild-type and mutated receptors in the pRc/CMW vector were transfected into CHO-K1 cells using calcium phosphate precipitation technique as described previously \[[@B21]\]. Stable transfectants were selected with 500 μg/ml G-418 (Geneticin; Life Technologies) and screened by intact cell binding using \[^3^H\]rauwolscine as previously described \[[@B21]\]. Down-regulation assays ---------------------- The cells were plated at a density of 2 × 10^5^ cells/ml in 150-mm culture dishes with 20 ml of growth media and allowed to grow for four days to near confluence. The medium was replaced with fresh medium lacking G-418. Twenty-four hours later, either 2 μl of 3 mM norepinephrine in 5 mM HCl or just 2 μl of 5 mM HCl (vehicle control) was added to each of the plates. Cells were harvested 24 h after adding norepinephrine or vehicle. The medium was removed and the cells were rinsed with 10 ml of cold phosphate-buffered saline (PBS) containing 5 mM EGTA, pH 7.4. Another 10 ml of PBS was added to the plate and the cells were scraped off the plate using a rubber policeman. The plate was rinsed once with 10 ml of PBS. The scraped cells were centrifuged at 2000 × g for 5 min at 4°C, the supernatant was discarded, and the pellet was stored at -80°C until assayed. We used 300 nM norepinephrine in the down-regulation studies because we have previously shown that this concentration induces maximal down-regulation for the wild-type alpha-2C subtype \[[@B20],[@B24]\]. Radioligand Binding ------------------- Pellets were homogenized in 12 ml of ice-cold 50 mM Tris HCl buffer (pH 8.0) with a Tissumizer (Tekmar, Cincinnati, OH) for 20 sec at 75% of the maximum speed and the suspension was centrifuged at 47,000 × g for 10 min at 4°C. This step was repeated a second time. For saturation experiments, the resulting pellet (a crude particulate membrane fraction) was resuspended in 200 volumes of 25 mM NaPO~4~ buffer (pH 7.4) and re-homogenized for use in the radioligand binding assay. For down-regulation experiments the cells were re-homogenized in 5 ml of 25 mM NaPO~4~ buffer (pH 7.4). For saturation experiments, duplicate tubes were prepared containing 970 μl of tissue and 20 μl of \[^3^H\]rauwolscine at six concentrations, with final concentrations from approximately 6 pM to 0.6 nM. Nonspecific binding was determined using a final concentration of 200 μM (-)-norepinephrine. For competition experiments, duplicate tubes were prepared containing 970 μl of tissue, 10 μl of norepinephrine at various concentrations, and 20 μl of \[^3^H\]rauwolscine (81.5 Ci/mmol, NEN-Dupont, Boston, MA) at a single concentration of 0.12 nM, which is approximately the K~d~ value. Twelve appropriately spaced concentrations of norepinephrine were used in each competition experiment. The extent of down-regulation was determined using a single concentration of \[^3^H\]rauwolscine (0.40 nM; about 3 times the K~d~ value). We have previously shown that norepinephrine pretreatment does not alter the radioligand K~d~ of the opossum alpha-2C adrenergic receptor \[[@B20],[@B24]\], thus making it possible to estimate the receptor density with a single concentration of radioligand. Total binding was determined in triplicate and nonspecific binding in duplicate for each of the three dishes for each condition. All assays were incubated at room temperature for 60 min. Membrane suspensions were filtered through Whatman GF/B glass fiber filters with a 48-sample Brandel cell harvester (Biomedical Research and Development, Gaithersburg, MD). Filters and tubes were washed three times with ice-cold 50 mM Tris HCl buffer (pH 8.0). Radioactivity was determined by liquid scintillation photometry. Protein concentrations were determined using the Bio-Rad Protein Assay (Bio-Rad, CA) based on the Bradford method \[[@B29]\], with bovine serum albumin as the standard. Data analysis ------------- B~max~ (receptor density) and K~d~ (radioligand affinity) values were determined from the analyses of saturation experiments utilizing nonlinear regression with Prism 3 (GraphPad, San Diego, CA). B~max~ and K~d~ values are given as arithmetic and geometric means, respectively. Competition binding data were also analyzed with Prism 3 to determine the IC~50~ values based on the single-site model and slope based on the variable slope model. K~i~ values were calculated from the IC~50~ values \[[@B30]\], and are presented as geometric means. The statistical analysis of the down-regulation data was conducted using a two-tailed paired t-test of the mean of the duplicate assays for each dish of cells incubated in the presence of norepinephrine as compared to control cells.

Introduction ============ The major problem with the use of nonsteroidal anti-inflammatory drugs (NSAIDs) is the high prevalence of their gastrointestinal toxicity \[[@B1]\]. Accordingly 40% of patients have non-ulcer gastric lesions, 10--25% gastric ulcer and 65% NSAID enteropathy \[[@B2]\]. The development of safer NSAIDs is dependent on knowledge of the pathogenesis of damage. In general, two distinct phases (topical and systemic) are suggested to be involved in the pathogenesis of enteropathy \[[@B3]-[@B5]\]. The topical phase of damage may be due to the effect of NSAIDs to uncouple mitochondrial oxidative phosphorylation or inhibit electron transfer \[[@B4]-[@B6]\] either of which would result in a decrease in cellular ATP levels\[[@B7]\] and a disrupted intestinal barrier function. Consequently, it is suggested, increased intestinal permeability transforms the cellular damage into a tissue reaction where inflammation and ulcers occur because of mucosal exposure of luminal aggressive factors (such as acid, pepsin, bile, pancreatic secretions and bacteria) and the effects of decreased mucosal prostaglandins on the microvasculature. In respect of early pathogenic events in the development of NSAID damage it has been shown that uncoupling of oxidative phosphorylation and inhibition of electron transport occurs with all commonly available acidic-NSAIDs. Our earlier studies indicate \[[@B8],[@B9]\] that indomethacin, aspirin and naproxen uncouple mitochondrial oxidative phosphorylation at concentration between 30 and 500 μM and inhibit the electron transport chain at higher concentrations. The issue of inhibition of electron transport by NSAIDs in the pathogenesis of damage and its pathophysiological consequences is complicated by the fact that nitric oxide (NO•), which plays a pathogenic role in the damage, may bind to metalloproteins \[Fe-S\] of the respiratory chain and by itself increase intestinal permeability \[[@B10]\]. Thus NSAIDs may inhibit electron transport directly or indirectly by induction of nitric oxide (NO•) \[[@B11]-[@B16]\]. The aim of the current study was to assess and discriminate between the direct and indirect effects of NSAIDs on electron transport of the rat small intestine using electron paramagnetic resonance (EPR) spectroscopy. mitochondrial electron transfer system plays a significant role in the pathogenesis of NSAID enteropathy. In the mitochondria electron transfer occur in the iron sulphur proteins arranged sequentially. Loss (oxidation) and gain of electrons (reduction) in this iron sulphur proteins can be visualized as a spectrum in the EPR machine. It has been hypothesized that NSAID might have an affinity towards these iron sulphur proteins and alter the redox state. Materials and methods ===================== Experimental design ------------------- ### Electron paramagnetic studies Male Sprague-Dawley rats (200--250 g) were used throughout EPR studies. 12 rats received indomethacin 20 mg/kg, which is a commonly used dose to induce consistent small intestinal damage in the rat (characterized by mitochondrial damage on electron microscopy within an hour,\[[@B17]\] neutrophil infiltration of mucosa on light microscopy at 3--5 hours and macroscopic lesions (ulcers) at 12--18 hours \[[@B18]\]); . A separate group of 12 rats received the non-acidic pro-NSAID nabumetone (500 mg/kg). The maximum human doses of indomethacin in man is 150 mg/day and 2000 mg for nabumetone so that the equivalent doses given here are at least 8 and 17 fold greater than the recommended doses for indomethacin and nabumetone, respectively and many times higher than this required for anti-inflammatory activity. Nabumetone is unusual amongst NSAIDs in that it does not produce any small intestinal lesions in the rat, presumably because it is non-acidic (and hence not a proton translocator or an inhibitor of the electron transport chain) and its active component 6-methoxy-2-naphthylacetic acid (6-MNA) is not subjected to excretion in bile \[[@B9],[@B19]\]. Incorporation of nabumetone therefore allows assessment of the systemic effects of cyclooxygenase inhibition in isolation from its topical effects. In separate studies the inhibitor of inducible nitric oxide synthase, 2,4 diamino-6-hydroxy pyrimidine (DAHP; a potent inhibitor of tetrahydrobiopterin synthesis, which is a cofactor for the inducible nitric oxide synthase activity) 1.0 gm /kg b.w. was administered to 12 rats intra-peritonealy 1 hr before indomethacin (20 mg/kg) gavages. Preparation of drugs: Indomethacin and nabumetone were dissolved in 10% dimethyl sulfoxide and then diluted to the final concentration of 5% and administered orally by a gastric gavage without anaesthesia. Control rats were received the same volume and concentrations of dimethylsulfoxide. ### Intestinal permeability assessment Rats, 12 in each group, were fasted overnight and received solvent, 20 mg/kg indomethacin or nabumetone (500 mg/kg) by gavage. 12 rats received DAHP i.p. (1.0 g/kg) 1 hr before indomethacin (20 mg/kg). Rats were fed 2 hrs later and fasted the following night. Twenty hrs after NSAID administration these rats received the intestinal permeability marker 51Cr- EDTA (10 uCi / rat in 0.5 ml followed, by 0.5 ml water) by gavage and were placed in metabolic cages. Urine was collected for 5 hrs and the samples were counted in Wallace 1842-LKB gamma counter with appropriate standard. The percentage of the orally administered radioactive chromium excreted in urine was calculated. ### Morphological assessment 20 hours after administration of these drugs rat small intestinal mucosa (n = 12) was exposed by a cut through the anti- mesenteric side and laid out each group on a piece of cork for assessment of macroscopic damage.\[[@B24]\] The assessments were carried out by an independent person unaware of the drug treatments. An ulcer count was made to distinguish between pointed (\<5 mm) and longitudinal (\>5 mm) ulcers. ### Isolation of mitochondrial fractions Bovine heart mitochondria were isolated based on a modification of Schneider & Hogeboom method (1950) \[[@B20]\], diluted to 10 mg protein/ml and sonicated in 3 ml aliquots on ice with eight 15 sec bursts, interrupted by 15 sec intervals, at 18 amps using a micro tip probe. The sonicated suspension was centrifuged at 26,000 g for 10 min at 4\'C. The supernatant fraction was centrifuged at 130,000 g for 1 hr, and the sub mitochondrial particles pellet resuspended for EPR analysis after checking the purity of this fraction (by studying the electron transfer reactions using potassium ferricyanide as an electron acceptor and succinate as a substrate). This experiment was repeated for three times at various intervals. It has been attempted to isolate submitochondrial fraction of jejunal mitochondria from rats. In most of the cases we could not able to achieve a good yield of submitochodrial particles. Hence we used bovine heart miotochondria to understand the basic phenomena of NSAID binding sites. ### EPR spectroscopy 1 hr and 24 hr after administration of the drugs, the animals were anaesthetized (Hypnovil and Hypnorm) and underwent laparotomy. Segments of jejunum 10 cm distal to the ligament of Trietz were snap frozen with liquid nitrogen and stored for a maximum of 2 weeks. Samples were thawed and homogenized in 10% (w/v) of sucrose (74 mM), mannitol (225 mM), homogenising buffer (0.25 M) containing EDTA (1 mM), 5 mg/ml BSA solution in MOPS-NaOH buffer (10 mM), pH 7.4, by 15 strokes of a tight fitting teflon pestle. 100 ul samples (1 mg protein) were placed in quartz EPR tubes (3 mm-internal diameter) and immediately frozen in liquid nitrogen. X-band EPR spectra were recorded on a Bruker ESP 300 spectrometer fitted with an Oxford Instruments ESR900 liquid helium flow cryostat at 8 K and 20 K or 30 K. The effect of 10 ul of 250 mM indomethacin added to 90 ul (1 mg protein) of submitochondrial particles in a quartz EPR tube and read the EPR spectra at 8 K was assessed with appropriate controls. Spectra were an average of six scans and normalized for direct comparison. The experiments were repeated for three different times with separate group of animals. EPR conditions: Temperature 8 K or 20 K or 30 K: microwave power 20 mW, microwave frequency 9.35 GHz, modulation frequency 100 Khs, modulation amplitude 1 mT, receiver gain 1.25 × 10^5^, time constant 0.33 sec, sweep time 2.4 mT/sec. Spectra displayed are average of 6--8 scans. Statistics ---------- Data are presented as mean ±SEM. Statistical comparison between groups were made by Mann-Whitney U test. Results ======= Morphology ---------- No macroscopic changes were observed in controls or with indomethacin and nabumetone 1 hr after drug administration. At 24 hr the indomethacin-treated group exhibited severe intestinal ulceration (pointed ulcers: mean = 55 ± 10, longitudinal ulcers: mean = 15 ± 9). No ulcers were seen in controls, nabumetone or DAHP treated indomethacin gavaged rats. \[Table [1](#T1){ref-type="table"}\] ::: {#T1 .table-wrap} ::: {.caption} ###### The ulcerogenic and intestinal permeability (51Cr-EDTA excretion) effect of DMSO (Control), indomethacin, DAHP pretreated indomethacin and nabumetone gavaged rat small intestine. ::: Group Treatment Pointed ulcers Longitudinal ulcers \% Excretion of ^51^Cr-EDTA in Urine ------- ------------------------------ ---------------- --------------------- -------------------------------------- 1 Control(DMSO) 0 0 2.3 ± 1.2 2 Indomethacin 55 ± 10 15 ± 9 ^\*^ 10.5 ± 2.5 3 DAHP pretreated indomethacin 0 0 ^\*^ 11.5 ± 0.8 4 Nabumetone 0 0 2.7 ± 1.5 ^\*^ P \< 0.01 ::: Intestinal permeability ----------------------- Controls excreted 2.3 ± 1.2% of the orally administered 51Cr-EDTA in 5 h urine. Intestinal permeability was not significantly (p \> 0.5) increased following nabumetone (mean 2.7 ± 1.5%). Indomethacin treated rats and those pre-treated with DAHP (+indomethacin) had significantly increased (p \< 0.01) intestinal permeability to 51Cr-EDTA (10.5 ± 2.5% and 11.5 ± 1.8% respectively) \[Table [1](#T1){ref-type="table"}\] EPR spectroscopic study of NSAIDs on sub-mitochondrial particles ---------------------------------------------------------------- At 8 K, the EPR spectrum of bovine heart sub mitochondrial particles administered DMSO alone (Fig. [1](#F1){ref-type="fig"}) displayed a prominent resonance at g \~ 2.014. This signal was only observed in the oxidized state at temperatures below 25 K, and is characteristic of the \[3Fe-4S\]ox cluster of mitochondrial Complex II Centre S3. Extra features at g \~ 2.04, 1.99 are present, and are assigned to interactions between ubisemiquinone radicals \[[@B23]\]. {#F1} Indomethacin but not nabumetone or DAHP added to the sub mitchondrial particles significantly decreased the intensity of the latter featrures (Fig. [1](#F1){ref-type="fig"}) (g \~ 2.04 and 1.99) suggesting formation of radical species which is indicative of binding of indomethacin to and the inhibition of the respiratory chain direct to this site. EPR spectroscopic studies of NSAIDs on rat jejunal homogenates -------------------------------------------------------------- At 20 K, control rat intestinal homogenates also displayed a signal indicative of Centre S3, albeit weaker (Fig. [2a](#F2){ref-type="fig"}). A similar signal could be detected 1 hr and 24 hr after nabumetone administration. \[Fig. [2b](#F2){ref-type="fig"} and Fig [3b](#F3){ref-type="fig"})\] However, with indomethacin the \[3Fe-4S\]ox signal intensity was decreased at 1 hr, and by 24 hr the EPR spectrum of intestinal homogenates displayed a prominent new resonance centered at g \~ 2.0, with a peak at g \~ 2.07 and a triplet splitting centered at g \~ 2.01, indicative of nitrosyl-haem (haem-NO) formation. \[Fig. [2c](#F2){ref-type="fig"} and Fig [3c](#F3){ref-type="fig"}\]. DAHP pre treated indomethacin group significantly decreased the haem-NO intensity at 24 hr after administration \[Fig. [4](#F4){ref-type="fig"}\], thereby allowing detection of the Centre S3 signal at 8 K \[Fig. [5](#F5){ref-type="fig"}\]. DAHP pretreated indomethacin group displayed the similar signals at 1 hr, when compared to indomethacin alone treated group \[.i.e., decreased in the \[3Fe-4S\]ox signal at 1 hr, Data not shown\]. The haem-NO signal observed in Fig. [3c](#F3){ref-type="fig"} shows a splitting of A (14N) \~ 330 mT \[Fig. [4](#F4){ref-type="fig"} at 30 K\]. Similar signals have been observed for the NO-adducts of type II haem proteins such as cytochrome c oxidase and hemoglobin \[[@B21]\] Figure [3](#F3){ref-type="fig"} explains the unique features of nitrosyl signals (g \~ 2.04) at 20 K only in indomethacin treated group and not in the DMSO-control or nabumetone group. Figure [4](#F4){ref-type="fig"} displays the differential pattern of the nitrosyl signals (i)-(ii) at 30K after 24 hr of indomethacin (i) and DAHP treated indomethacin group(ii) when compare to the DMSO-control (iii) spectra. This prominent nitrosyl signals at 30 K is the characteristics of the free radical species and temperature sensitive when compare to the signals at 8 K, displayed in Figure [5](#F5){ref-type="fig"}. In indomethacin alone treated group (Figure [4](#F4){ref-type="fig"}) this broad signal displayed a peak at g \~ 2.04, a distinct trough at g \~ 1.98 and complex splitting (triplet) centered at g \~ 2.01 at temperature above 25 K, indicative of haem-nitrosyl formation.\[[@B21]\] {#F2} {#F3} {#F4} {#F5} Discussion ========== The present results show that indomethacin treated rats and not those pretreated with DAHP exhibited small intestinal ulcers at 20 hr. There was increased intestinal permeability and an early (1 hr) direct effect of indomethacin consistent with inhibition of electron transport in both groups but only the indomethacin treated group exhbited a haem-nitrosyl complex EPR signal at g \~ 2.07. Thus the present result suggests that the nitric oxide radical may play a role in the development of indomethacin-induced ulcers. Within the current pathogenic framework \[[@B5]\] where it is proposed that NSAIDs affect mitochondrial energy metabolism leading to increased intestinal permeability and where the micro vascular alterations, caused by changes in prostaglandin and nitric oxide metabolism are the driving force in the development of ulcers. The involvement of nitric oxide occurs temporarily later or at a different pathophysiological step than the intestinal permeability changes. The non-acidic Pro-NSAID nabumetone caused no intestinal lesion as previously described \[[@B22]\] and was not associated with changes in the EPR spectra, despite being effective inhibitors of cyclooxygenase and associated with decreased mucosal prostaglandins at much smaller doses\[[@B22]\]. These findings are consistent with previous in vitro and in vivo work showing that nabumetone itself does not alter mitochondrial energy metabolism whilst its active metabolite 6MNA does. However it would appear that as 6MNA is formed following absorption of nabumetone, over 99% of 6MNA is protein bound within the circulation and it is not excreted in bile, insufficient intestinal concentrations of 6MNA are achieved to affect enterocyte mitochondria. NSAIDs uncouple oxidative phosphorylation at 30--500 uM and inhibit mitochondrial electron transport at higher concentration in vitro \[[@B8]\] The relative importance and contribution of uncoupling and inhibition in the pathogenesis of the \"topical\" Phase of damage is uncertain, but either mechanism would lead to decreased cellular ATP and hence increased intestinal permeability. In this study we assessed the characteristics of NSAID modulation of electron transport in vivo in rats and in vitro in bovine heart submitochondrial particles. The EPR spectra observed at 1 hr would appear to be the result of a direct effect of indomethacin on electron transport unrelated to NO\', since pretreatment with DAHP did not alter the indomethacin spectra. The EPR spectra at 24 hr post indomethacin \[Fig. [2B (c)](#F2){ref-type="fig"}\] indicated formation of NO\' which is further substantiated by abolition of this signal when rats were pretreated with DAHP \[Fig. [3](#F3){ref-type="fig"}\]. The activity of iNOS is controlled mainly at the levels of transcription and translation and is dependent on the availability of co-factors including tetrahydrobiopterin (BH4). A recent study has demonstrated that inhibition of BH4 synthesis reduces NO• production in vivo in experimental endotoxic shock\[[@B12]\]. The same strategy of inhibition of BH4 synthesis has been applied in this present attempt to investigate ulcer formation in indomethacin-induced intestinal ulcer. We found that injection of DAHP an inhibitor of BH4 synthesis prevented ulcer formation, although it did not alter the increased intestinal permeability induced by indomethacin. This suggests that it does not modulate the direct biochemical actions of NSAIDs but interferes at a later stage in the pathogenesis of the damage. One possibility is that DAHP suppresses NO formation in neutrophils (iNOS), which infiltrate the mucosa in response to increased intestinal permeability. In order to study the specificity of the nitric oxide synthesis the experiments with N(G)-nitro-L-argenine methyl ester (L-NAME-10 mg /kg b wt s.c 1 hr prior to indomethacin) was carried out to prevent nitrosyl signals and ulcers induced by indomethacin. We did not find the inhibition of intestinal ulcers. This observation is further confirmed by the recent reports \[[@B25]\] stating the worsening effect of L-NAME (5--20 mg/kg) on these lesions was dose-dependently observed in association with further enhancement of the bacterial translocation and intestinal hypermotility following indomethacin. Thus it has been concluded that this effect may be due to the nonspecific inhibition of L-NAME and it is not a specific inhibitor of iNOS cofactor biopterin complex and hence excluded from EPR study. Conclusions =========== In summary, it is interesting that in spite of the increased intestinal permeability, there were no ulcers seen in the DAHP treated group indicating that inhibition of inducible nitric oxide synthase (in neutrophils and macrophages) could become a targeted therapeutic approach to reduce the intestinal toxicity of NSAIDs, and nabumetone is comparatively safer with neither inducing ulcer nor formation of nitrosyl complex in this study. Competing interests =================== None declared Authors\' contributions ======================= SS carried out the planning, animal experimentations, EPR measurements and drafted the manuscript. RS contributed for drafting the manuscript. SR carried out submitochondrial particle isolation and *In vitro* NSAIDs inhibition study. JKS helped for the computation of EPR spectra and contributed in the interpretation of the spectra. JW and IB conceived of the study and participated in its design, coordination, corrections and helping to draft this manuscript. Grant Support ============= This work was supported by an unconditional grant from Smith Kline Beecham, USA to Dr. S. Somasundaram. Abbreviations ============= EPR, electron paramagnetic resonance; NSAIDs, nonsteroidal anti-inflammatory drugs; DAHP, 2,4-diamino-6-hydroxy pyrimidine; DMSO, dimethyl sulfoxide Pre-publication history ======================= The pre-publication history for this paper can be accessed here: <> Acknowledgements ================ The authors thank Professor. R. Cammack for providing electron paramagnetic resonance spectroscope for the above study.

Background ========== Vacuolar-type H^+^-ATPases are enzymes responsible for the energization of membranes and the acidification of compartments within the eukaryotic cell via the establishment of proton and electrochemical gradients at the expense of ATP. The vacuolar type H^+^-ATPase in plants is a large multimeric enzyme complex whose function is to pump protons across a membrane via primary active transport. Vacuolar type ATPases are homologs of the F-type ATP synthases and probably convert the free energy of hydrolysis of the high-energy phosphate bond into rotational motion \[[@B1]-[@B9]\]. Vacuolar type H^+^-ATPases are critical for the maintenance of homeostasis in eukaryotic cells \[[@B10],[@B11]\]. In plant cells V-ATPases are responsible for the deacidification of the cytosol and the energization of secondary transport processes across the tonoplast, as well as in the endocytotic and secretory pathways \[[@B12]\]. In addition, the vacuolar type ATPase is thought to be primarily responsible for the acidification and expansion of the large central vacuole \[[@B12]-[@B15]\]. In plants a critical event in growth and development is the maturation and expansion of the central vacuole. Upon maturation, the vacuolar contents can comprise more than eighty percent of the total cell volume \[[@B16]\]. Primarily filled with water, the vacuole also is a repository for a wide variety of solutes fulfilling numerous important metabolic functions. The influx of water and metabolites into the vacuole is dependent in part on the generation of a proton motive force across the tonoplast. Vacuolar type H^+^-ATPases are large, multimeric enzyme complexes of 500--750 kDa consisting of 15 or more different proteins. Structurally, they can be divided into integral (V~0~) and peripheral (V~1~) membrane sectors \[[@B17]\]. The peripheral membrane sector is located on the cytoplasmic side of the membrane and is composed of at least five different subunits including the catalytic subunit, subunit A. The V~1~ peripheral sector is further divided into a head group and stalk region. The V~0~ integral sector is embedded in the membrane and consists of at least 4 different subunits. Subunit A of the vacuolar-type H^+^-ATPase is the catalytic subunit. The subunit A protein is approximately 70 kDa in most organisms studied and is a hydrophilic peptide located in the head group of the V~1~ peripheral sector in three copies per holoenzyme \[[@B18]\]. This subunit contains a nucleotide binding motif and functions to bind and hydrolyze ATP \[[@B19]\]. In addition, subunit A contains a highly conserved cysteine residue located within the enzymes\' catalytic center that may be involved in regulation of the holoenzyme \[[@B20]-[@B23]\]. Cloning, sequencing and characteristics of subunit A of the V-Type ATPase from *Arabidopsis thaliana* have been previously reported \[[@B24]\]. Small multigene families of V-ATPase subunit A are known to exist in flowering plants and algae. The catalytic subunit exists as two distinct genes with highly conserved exons and intron boundaries in twelve species of plants \[[@B25]\]. In the flowering plant *Daucus carota* (carrot) evidence exists for two distinct isoforms for the catalytic subunit, one of which is tonoplast specific while the other is localized to the Golgi \[[@B26]\]. In addition, two distinct mRNAs have been isolated from cotton (*Gossypium hirsutum*) \[[@B27]\]. Two subunit A mRNAs are expressed in the unicellular alga *Acetabularia acetabulum*\[[@B28],[@B29]\]. Isoforms of other V-ATPase subunits are also known from plants including subunit B \[[@B27],[@B30],[@B32]\], the proteolipid subunit \[[@B33]-[@B37]\] and subunits D and E \[[@B38]\]. What is lacking in our knowledge of vacuolar type ATPases is a sense of what purpose these multigene family proteins may serve in the cellular milieu and how they participate in the growth, development, and metabolism of higher plants. It has been proposed that multigene families might be encoding functionally distinct isoforms of V-ATPase subunits. These isoforms may be required to provide unique functions or regulatory constraints within the varied micro-compartments requiring V-ATPases \[[@B26],[@B27],[@B34],[@B39],[@B43]\]. In addition, different isoforms may be responsible for targeting to varied subcellular compartments \[[@B10],[@B26],[@B28],[@B36],[@B41],[@B44],[@B47]\]. Furthermore, it is possible that gene family members may behave in a differential manner with respect to environmental stress \[[@B36]\]. Three enzymes are primarily responsible for the maintenance of proton flux across the tonoplast in plants, the vacuolar H^+^-ATPase, the H^+^ pumping pyrophosphatase \[[@B48]\] and the Na^+^/H^+^ antiporter \[[@B49],[@B50]\]. All three enzymes have been implicated in allowing plants to cope with environmental stress conditions. Several studies have been conducted in an attempt to evaluate the role vacuolar type ATPases play in allowing plants to cope with environmental stress. These endeavors have primarily concentrated on salt and cold stress. Cold tolerant *Brassica napus,* a close relative of *Arabidopsis thaliana,* demonstrated an increase in both subunit A mRNA and protein in response to chilling at 2°C \[[@B50]\]. This chilling also resulted in a concomitant increase in cell sap osmotic pressure and endogenous ABA accumulation. These researchers indicated that *Arabidopsis* also was a cold tolerant plant which exhibited a similar response with respect to V-ATPase subunit A mRNA, however they provided no data. In rice (*Oryza sativa*) chilling plants at 10°C resulted in a two fold increase in vacuolar type ATPase activity evaluated by in vitro assay of isolated membranes \[[@B51]\]. This study did not analyze changes in protein level in the isolated membranes in response to chilling. Cold sensitive mung bean seedlings (*Vigna radiata*) subjected to chilling at 4°C for 48 hours contrastingly showed no increase in vacuolar type ATPase activity in isolated membranes \[[@B52]\]. In addition, these researchers found no change in the amount of V-ATPase protein levels in response to chilling. The facultative CAM plant *Mesembryanthemum crystallinum* displays a differential organ level response to salt stress while still engaged in C~3~ photosynthesis \[[@B53]\]. Subjecting plants to 400 mM NaCl resulted in an approximately two fold increase in the transcript levels of subunits A, B, and the proteolipid in roots prior to the shift to Crassulacean acid metabolism. Fully expanded leaves exhibited an increase only for the proteolipid mRNA, an increase of approximately three fold over non salt stressed plants \[[@B54]\]. Cell suspension cultures when subjected to salt stress exhibited increase in activity of both P-Type and V-Type ATPases without concomitant increase in protein levels \[[@B55]\]. Transcript levels of subunit E of the V-ATPase in *Mesembryanthemum crystallinum* were recently shown to increase in leaves but not roots of salt stressed plants with a concomitant increase in protein levels in leaves \[[@B56]\]. On the other hand, subunit E in *Hordeum vulgare* showed no effect on mRNA level in response to salt stress \[[@B57]\]. The effects of salt stress on V-ATPases has also been investigated in tobacco (*Nicotiana tabacum*) suspension cells. Subjecting NaCl adapted cells to salt stress resulted in an approximately four fold increase in both proton transport and ATP hydrolysis on isolated membranes compared to unadapted cells \[[@B58]\]. Remarkably this increase in overall V-ATPase activity coincided with an apparent reduction in the amount of enzyme on the tonoplast. This apparently contradictory finding may be the result of physical changes in the enzyme itself or changes in the lipid environment \[[@B58]\]. Especially confounding is the finding that mRNA levels for subunit A increase two to four fold in response to the same levels of salt stress in tobacco suspension cells \[[@B59]\]. Experiments with tomato (*Lycopersicon esculentum*) showed similar results with respect to induction of mRNA for subunit A with an approximately two to four fold increase seen with respect to salt stress \[[@B60]\]. Recent work in the resurrection plant *Tortula ruralis* indicated that transcript levels of the proteolipid (subunit c) increased in response to salt stress but that there was no corresponding increase in protein levels \[[@B61]\]. The limited work done to evaluate the role of V-ATPases in drought stress has provided conflicting results. Drought stress in winter *Brassica napus* resulted in a very similar response as cold stress, that is a dramatic increase in subunit A mRNA \[[@B50]\]. However, drought stress in *Lycopersicon esculentum* resulted in no apparent increase in mRNA levels for subunit A \[[@B60]\]. None of the previously mentioned efforts into evaluating the role of vacuolar type ATPases in environmental stress acclimation take into account the fact that subunits A, B, and the proteolipid are most certainly present as small gene families in the plants studied. Furthermore, it is possible that these gene family members may behave in a differential manner with respect to stress. The only work published previously on V-ATPase subunit isoforms and stress is concerned with the proteolipid gene family in *Arabidopsis thaliana.* Three proteolipid genes were evaluated for their response to salt stress as well as etiolation \[[@B36]\]. All three genes exhibited an approximately two fold increase in mRNA levels in response to salt stress (50 mM). However, with respect to etiolation two isoforms showed no response while the third demonstrated an increase in mRNA levels. A number of genes in plants have been found to contain multiple poly (A) attachment sites yielding different transcripts from a single gene. Several examples have recently emerged where differentially polyadenylated genes have transcripts that are post transcriptionally regulated based on sugar sensing. In rice (*Oryza sativa*), two transcripts generated off an α amylase gene exhibit differential stability based on sucrose availability \[[@B65]\]. A cell wall invertase gene (*Incw-1*) in Maize (*Zea mays*) is expressed as two transcripts differing in 3\' ends. The two transcripts are differentially expressed in response to sugars \[[@B66]\]. In this study we explore the possibility that differentially polyadenylated transcripts of the V-ATPase subunit A encoding gene show different responses with respect to etiolation, salt, or chilling stress. Results and discussion ====================== A vacuolar type H^+^-ATPase subunit A cDNA has been previously identified and sequenced (GenBank accession number U65638) \[[@B24]\]. The cDNA was identified as subunit A of the vacuolar type H^+^-ATPase via identity with other previously sequenced subunit A genes from other organisms. The primary nucleotide and amino acid sequence of the *A. thaliana* subunit A cDNA showed a high degree of identity with subunit A cDNA sequenced from other organisms. This cDNA encoded a putative open reading frame consistent with that of other subunit A genes and was shown to have all the motifs common to subunit A \[[@B67]\]. This high degree of conservation was evident across taxa as diverse as animals, plants, fungi, and protozoa \[[@B68]\]. Genome blot analysis was utilized to determine the number of genes in the *A. thaliana* genome corresponding to subunit A of the vacuolar type ATPase. Genomic DNA was isolated as described, subjected to restriction enzyme digestion, and immobilized on a positively charged nylon membrane. Ten single restriction enzyme digestions were performed. Hybridization was conducted at moderate stringency with a 300 base pair homologous *A. thaliana* cDNA probe corresponding to subunit A of the vacuolar type ATPase. Results indicated hybridization to only a single band for nine out of the ten enzymes (Fig. [1](#F1){ref-type="fig"}). The *Hae*III digest showed faint hybridization to a second band that was later shown to be the result of a *Hae*III site located in the region of the probe (results not shown). {#F1} All evidences acquired to date failed to indicate the presence of a second V-ATPase subunit A gene in *A. thaliana.* Whole genome blot analysis detected a single gene in the *A. thaliana* genome. Extensive screening of a cDNA library derived from all organs and multiple developmental stages of the plant revealed only a single subunit A cDNA. A polymerase chain reaction based screen failed to detect an intron length polymorphism (results not shown) between two subunit A genes found to be present in other plants \[[@B25]\]. Lastly, extensive database searches of GenBank, the *Arabidopsis* EST database, and the *Arabidopsis* Genome Initiative (TAIR, <>) report only a single subunit A gene in the current version of the *Arabidopsis* genome (chromosome 1 BAC F9K20). This situation of a single subunit A gene appears to be unusual in plants. Admittedly only a small subset of plants has been examined in detail. Of the species examined to date three have only a single subunit A gene detected, the alga *Coleochaete scutata,* sugar beet (*Beta vulgarise*\[[@B63],[@B64]\], and *A. thaliana*\[[@B25]\]. Flowering plants shown to have at least two subunit A genes include oat (*Avena sativa*), carrot (*Daucus carota*), tobacco (*Nicotiana tabacum*), tomato (*Lycopersicon esculentum*), cotton (*Gossypium hirsutum*), magnolia (*Magnolia virginana*), *Hydrastis canadensis, Chenopodium rubrum,* and clematis (*Clematis liguisticfolio*). In addition, the unicellular alga *Acetabularia acetabulum* has two V-ATPase subunit A genes. RNA blot analysis was performed to determine the organ level expression pattern of the subunit A gene. Vacuolar type ATPases are housekeeping enzymes critical for the maintenance of homeostasis. Therefore, it is expected that all living plant cells require V-ATPase function. Since *A. thaliana* seemed to possess only a single gene for subunit A of the V-ATPase it was expected that this gene should be expressed in all organs tested. Whole RNA was harvested from leaves, bolts, siliques, flowers, axenically grown seedlings, and axenically grown roots. RNA was denatured with glyoxal, fractionated via agarose gel electrophoresis, and immobilized on a positively charged nylon membrane. Hybridization was conducted at high stringency with a 1700 base pair homologous cDNA probe labeled with digoxigenin. In agreement with our expectations, results indicated high levels of expression in all organs tested (Fig. [2](#F2){ref-type="fig"}) with slightly higher levels of subunit A mRNA present in seedlings. This is most likely due to the fact that seedlings are the most rapidly growing and metabolizing stage of the plant tested. Seedlings are undergoing dramatic and rapid vacuole biogenesis to allow for rapid growth at this stage of development and therefore would be expected to require a higher level of subunit A expression than other more terminally differentiated organs. Roots, leaves, bolts, and flowers showed approximately equivalent amounts of signal with somewhat lower levels in siliques. {#F2} Most eukaryotic mRNAs are characterized by the presence of a poly (A) tail, a series of adenine residues post transcriptionally attached to the 3\' untranslated region. The primary function of the poly (A) tail is thought to be in message stability assisted by a series of poly (A) binding proteins \[[@B69]\]. In addition, the poly (A) tail in conjunction with the 5\' cap may be required for efficient translation of plant mRNAs \[[@B70]\]. Subunit A of the vacuolar type H^+^-ATPase in *A. thaliana* was shown to exhibit multi site polyadenylation. Sequencing the 3\' ends of seven independent clones isolated from the PRL-2 cDNA library and one EST obtained through the *Arabidopsis* EST database revealed four different length transcripts produced from the single subunit A gene based on different poly (A) tail attachment sites (Fig. [3](#F3){ref-type="fig"}). Multiple polyadenylation sites are fairly common in plant mRNAs but this phenomenon is rare in animal genes \[[@B71],[@B72]\]. Since the *A. thaliana* V-ATPase subunit A gene produced at least four different transcripts (all producing the same protein) and is one of two flowering plants observed to date not to have a multigene family for this subunit, experiments were designed to determine if muiti site polyadenylation might be a mechanism for gene regulation in plants. {#F3} V-ATPase proteolipid isoforms were shown to exhibit differential transcript expression levels in response to environmental stresses \[[@B36]\]. This fact implies a possible role for these isoforms in allowing the plant to deal with stress. Experiments were designed to determine if, in *A. thaliana,* differentially polyadenylated transcripts could be regulated in response to stress. First, it was critical to confirm that subunit A transcript levels were at all responsive to environmental stress conditions. Three environmental stresses were chosen including 100 mM NaCl, etiolation for four days, and chilling at 6 degrees for four days. In all three situations axenic, seven-day-old seedlings grown on agar were evaluated. Seven day old seedlings were chosen for their short growth time, limited space requirements, and ease in culturing. In addition, the facts that they are rapidly growing and metabolizing, were shown to produce ample quantities of subunit A transcript, and possess several organs including roots, hypocotyls, and cotyledons made them an attractive choice for analysis. Whole RNA blot analysis was used first to evaluate the overall response of subunit A to stress conditions. RNA was extracted from stress treated seedlings and non-treated control plants. Hybridization was performed as previously described. Results indicated a subunit A transcript response to stress. An approximate 2--4 fold increase in transcript level was detected in seedlings subjected to salt and chilling stress compared to control plants (Fig. [4](#F4){ref-type="fig"}). Etiolated seedlings showed essentially equivalent amounts of subunit A transcript compared to controls. Since etiolation results in rapid elongation of seedlings it was expected that there would be a concomitant increase in subunit A messenger RNA as seen with proteolipid subunit mRNA in *A. thaliana*\[[@B36]\] and subunit A message levels in rapidly expanding cotton (*Gossypium hirsutum*) fibers \[[@B73]\]. Equivalent loading of whole RNA was evaluated via spectrophotometry measuring absorption at 260 nm using a Perkin-Elmer Lambda-3 spectrophotometer and by visual confirmation using non denaturing ethidium bromide stained agarose gel electrophoresis (results not shown). {#F4} Having established that overall subunit A transcript levels were responsive to environmental stress conditions an assay was designed to examine the response of the four individual lengthed transcripts to the same stress conditions. The goal was to determine if any of the four were differentially expressed in response to stress. A PCR based approach was taken to evaluate transcript levels for the four different subunit A mRNAs. Primer pairs were designed (Table [1](#T1){ref-type="table"}) that would only amplify one of the four transcripts. This was accomplished by anchoring the downstream primer in the poly (A) tail. The upstream primer was located in the coding region of the subunit A cDNA. Anchoring the downstream primer in the poly (A) tail prevented that oligonucleotide from priming at any other site since 10 out of 18 nucleotides were T\'s at the 5\' end. The primer pairs were designed such that the four amplification products would differ in size by approximately 100 base pairs for ease of visualization on agarose gels and blots (Fig. [5](#F5){ref-type="fig"} and Fig. [6](#F6){ref-type="fig"}). Primer pair T-11 and T-12 corresponds to transcript-1 and produce an amplification product of 579 base pairs. Primer pair T-21 and T-22 corresponds to transcript-2 and amplify a band of 455 base pairs. Primers T-31 and T-32 correspond to transcript-3 and amplify a 362 base pair product. Lastly, primer pair T-41 and T-42 corresponds to transcript-4 and produce a 238 base pair amplification product. The transcript-4 amplification product, which is the smallest and nested within the other three, acted as probe to detect all four transcripts. This nesting was accomplished to avoid differences in signal during hybridization based on different probe lengths and labeling efficiency. {#F5} {#F6} ::: {#T1 .table-wrap} ::: {.caption} ###### Polymerase Chain Reaction primers. Primers SM-2, SM-3 and VA-1, VA-9 were utilized in the generation of digoxigenin labeled probes for hybridization. Primers T-11 through T-42 were utilized to amplify subunit A transcripts based on the position of the poly A tail. ::: **Primer** **5\' to 3\' sequence** **Amplification product** **5\' nucleotide position in cDNA** ------------ -------------------------- --------------------------------- ------------------------------------- SM-2 CCTGGTGCATTTGGTTGTGGAAA Subunit A cDNA (300 bp probe) 779 SM-3 CATCATACTAACATTGTAACCCAT Subunit A cDNA (300 bp probe) 1055 VA-1 TCTTCGAACACACAAGCCACTTT Subunit A cDNA (1700 bp probe) 274 VA-9 AGAGACCATGAATGAAACCCTCT Subunit A cDNA (1700 bp probe) 1959 T-11 TTTTTTTTTTACAAAATG Subunit A, 3\' end transcript-1 2264 T-12 TGAGAGAGCAGCTGGAAT Subunit A, 3\' end transcript-1 1702 T-21 TTTTTTTTTTATCCCCAA Subunit A, 3\' end transcript-2 2227 T-22 TCAGAAGTTCGAAGACCC Subunit A, 3\' end transcript-2 1789 T-31 TTTTTTTTTTGGAGATAA Subunit A, 3\' transcript-3 2210 T-32 CTGGATTCCGTGCTTTGG Subunit A, 3\' transcript-3 1866 T-41 TTTTTTTTTTCAAATCAA Subunit A, 3\' transcript-4 2173 T-42 GGAGAAAGAGGGTTTCAT Subunit A, 3\' transcript-4 1952 ::: A control experiment was performed prior to the actual stress analysis to ensure that the primer pairs amplified the correct transcript only and that the transcript-4 probe would detect all four amplification products. Whole RNA was extracted from seedlings and cDNA was generated using oligo dT as described in materials and methods. Subunit A transcripts were amplified from the generated cDNA with the appropriate primer pairs and transferred to solid support. Hybridization was performed at high stringency with the transcript-4 amplification product as probe. The results of this experiment indicated that all four primer pairs amplified only a single band of the appropriate and expected size and that the transcript-4 probe adequately hybridized to all four amplification products (Fig. [5](#F5){ref-type="fig"}). The differential expression of subunit A transcripts was evaluated in response to salt, cold stress, and etiolation. Complementary DNA\'s were generated from seven-day-old seedlings subjected to individual stress conditions. Transcripts 1--4 were evaluated via RT-PCR using transcript specific primers. Amplification products were immobilized on solid support and hybridized with the transcript-4 amplification product. The results of this experiment indicate that all four transcripts behaved identically with respect to individual stress conditions (Fig. [7](#F7){ref-type="fig"}). All four transcripts showed approximately 2--4 fold increases in response to 100 mM sodium chloride stress compared to untreated controls. Similarly, treatment of plants at six degrees for four days resulted in a similar increase in signal for all four transcripts. Etiolation resulted in slightly higher message levels for transcripts -1, -2, and -3 compared to controls. The lesser degree of increase for transcript-3 might be due to near saturation of the signal. The latter might be due to the higher abundance of the transcript, or due to more efficient amplification of the transcript-3 primer set. Transcript-4 seems to show slightly less signal than control plants but the autoradiograph is somewhat under developed. These results are in agreement with the whole RNA stress blot analysis. Apparent quantitative differences in upregulation between transcripts shown in (Fig. [7](#F7){ref-type="fig"}) might be due to differential amplification efficiency of the four different primer sets and do not necessarily represent true differential upregulation of the four transcripts. The data can only be evaluated quantitatively between samples amplified with the same primer set. In summation, no significant differential regulation of the four subunit A transcripts was detected in seedlings in response to the three stress conditions evaluated. {#F7} Conclusions =========== Evidence gathered to date indicates the presence of a single gene encoding subunit A of the vacuolar type H^+^-ATPase in *A. thaliana.* The search for additional subunit A genes was conducted using whole genome blot analysis, screening a lambda PRL-2 cDNA library, intron length polymorphism analysis, and extensive computer searches of the *Arabidopsis* Expressed Sequence Tag database, GenBank, and the *Arabidopsis* Genome Initiative. Most flowering plants evaluated to date have a small multigene family encoding subunit A of the vacuolar type H^+^-ATPase. The small size of the *A. thaliana* genome may be indicative of a trend toward streamlining of DNA in this plant, which may have eliminated a redundant subunit A gene over evolutionary time. Subunit A genes and subsequently proteins are highly conserved. Therefore, it is not likely that a subunit A gene would have escaped the screen by having a low degree of nucleotide identity. Vacuolar type ATPases are housekeeping enzymes probably needed in all living plant cells. The single *A. thaliana* subunit A gene was shown by whole RNA blot analysis to be transcribed in all organs of the plant tested. The organs analyzed included roots, leaves, bolts, siliques, and flowers. In addition, relatively high levels of transcript were found in whole seedlings. These results support the notion of a single subunit A gene in *A. thaliana* that is transcribed in all cells of the plant. Numerous examples are present in the literature \[see \[[@B14]\] and references therein\] of changes in transcript, protein, or activity levels of vacuolar type H^+^-ATPases in response to environmental stress. These results indicate a potential role for the vacuolar type H^+^-ATPase in maintenance of homeostasis during times of environmental stress. Whole RNA blot analysis performed on seedlings exposed to environmental stress conditions indicated that transcript levels of subunit A were responsive to salt and cold stress treatments. Transcript levels were shown to increase approximately 2--4 fold in seedlings treated with cold and salt stress. Etiolation did not produce as significant a change in transcript level as salt and cold stress. The single vacuolar type H^+^-ATPase subunit A gene detected in *A. thaliana* can produce at least four different length transcripts by using different polyadenylation sites. These transcripts differ only in their 3\' untranslated region and produce identical proteins. The presence of regulatory elements in the 3\'end of the subunit A gene could be a potential mechanism by which this protein is post transcriptionally regulated. However, the evidence gathered in this work indicates that the four different transcripts generated off the single subunit A gene are not significantly differentially regulated in response to salt, chilling, or etiolation in axenically grown seedlings. Multi site polyadenylation does not appear to be a mechanism for posttranscriptional control of gene expression for subunit A of the vacuolar type H^+^-ATPase in *A. thaliana* for the stress conditions and developmental stage tested. Materials and methods ===================== Plant materials --------------- *Arabidopsis thaliana* ecotype Columbia seeds were obtained from the *Arabidopsis* Biological Resource Center, Ohio State University, Columbus OH. Seeds were germinated in 4-inch pots according to standard protocols. Axenic seedlings, and soil grown plants were maintained in a growth chamber at 23°C under a 16-hour photoperiod until appropriate organs were mature. Axenic seedling cultures were prepared by placing surface sterilized seeds on 90 mm germination plates containing 1% agar and 4.4 g/L MS salts with minimal organics (Sigma Chemical Co. St. Louis, MO), pH 6.0. Seedlings were harvested when 7 days old. Axenic roots were obtained by plating surface sterilized seeds on Nunc square Bio-Assay dishes (Fisher Scientific, Pittsburgh, PA). Plates were allowed to dry under a laminar flow hood, sealed with paraflim and incubated vertically as above. Roots were harvested 7--14 days later. Growth effects of salt stress were evaluated by germinating surface sterilized seeds on agar containing various concentrations of sodium chloride. The optimal concentration chosen to evaluate the effects of salt stress on V-ATPase subunit A transcript levels was 100 mM. Higher concentrations resulted in failure to germinate or extremely poor growth. At 100 mM salt most seeds germinated but required more time to do so than controls. Plantlets did grow but were less robust than non-treated controls (results not shown). Seedlings were harvested after seven days of growth. In the case of etiolation stress, surface sterilized seeds were germinated on agar for three days under light and temperature conditions described above. After three days the plates were placed in a light sealed box at the same temperature conditions. Seedlings were harvested four days later. Plantlets exhibited the typical characteristics of etiolation, the hypocotyls were lengthened and the cotyledons reduced compared to light grown controls. Cold stress was evaluated by germinating surface sterilized seeds as described for three days under normal light and temperature. Agar plates containing seedlings were then transferred to a cold room at 6°C. Seedlings were maintained at this temperature for four days before being harvested. Isolation of nucleic acids -------------------------- Genomic DNA was extracted from mature leaves using a miniprep protocol \[[@B74]\] modified by the addition of a CTAB extraction \[[@B75]\]. *A. thaliana* leaves were harvested, hand washed and then ground to a fine powder in liquid nitrogen. Ground material was used immediately for DNA extraction or stored at -80°C. Post extraction samples were treated with RNAse A at 37°C for 15 minutes to remove contaminating RNA. Organ specific whole RNA extraction was performed as described \[[@B75]\]. Specific organs were harvested and immediately ground to a fine powder in liquid nitrogen. Special care was taken in the case of siliques and flowers to harvest the organs directly into liquid nitrogen to prevent degradation of the RNA. Standard precautions were taken to eliminate ribonuclease contamination in solutions and glassware including baking all glassware at 200°C for six hours, and treating all solutions (except those containing Tris) with 0.1% diethylpyrocarbonate for 12 hours followed by autoclaving. Nucleic acids were quantitated by spectrophotometry and via inspection of ethidium bromide stained agarose gels. Genome blot analysis -------------------- Genomic DNA (5--10 μg) was digested with restriction enzymes according to the manufacturers recommendations. Digested DNA was size fractionated via gel electrophoresis in 12 cm gels containing 0.8% agarose run in IX TAE or TBE at 25 volts for 12 hours. Fractionated DNA in gels was depurinated, denatured, and neutralized according to standard protocols \[[@B76]\]. DNA was transferred to positively charged nylon membranes (Hybond N^+^) (Amersham Pharmacia Biotech, Inc., Piscataway, NJ) by the wick transfer method using 20X SSC overnight. Membranes were baked at 80°C for 1.5 hours. Hybridization was performed according to a procedure from \[[@B77]\] modified as described below. Membranes were first washed for 5 minutes in 5X SSC at room temperature followed by 1 hour at 65°C in 1X SSC/0.1% SDS. Membranes were then pre-wet for 10 minutes in wash buffer (20 mM Na~2~HPO~4~ pH-7.2, 1 mM EDTA, 1% SDS). Prehybridization was performed at 67°C for two hours in a seal-a-meal bag (Dazey Corporation, Industrial Airport, KS) with hybridization solution (0.25 M Na~2~HPO~4~ pH-7.2, 1 mM EDTA, 20% SDS with 0.5% blocking reagent-Boehringer Mannheim, Biochemical Products, Indianapolis IN) at a volume of 125 μl/sq. cm. of membrane and with formamide added to a final concentration of 25% and 15 μg/ml boiled salmon sperm DNA. Hybridization was performed overnight at 67°C in a sealed bag with 125 μl/sq cm fresh hybridization solution containing formamide and salmon sperm DNA as above with the addition of 40 ng/ml digoxigenin labeled probe denatured by boiling for 5 minutes. The probe utilized was a 300 base pair digoxigenin labeled PCR product. Membranes were washed after hybridization four times at 65°C for 30 minutes each with vigorous agitation in wash buffer. Membranes were then soaked for 10 minutes in maleate buffer (0.1 M maleic acid, 3 M NaCl, 0.3% Tween 20, pH 8) at room temperature followed by blocking for two hours at room temperature in maleate buffer with 0.5% blocking reagent (Boehringer Mannheim, Biochemical Products, Indianapolis IN). Blocked membranes were sealed in a bag with 125 μl/sq. cm. of maleate buffer with 0.5% blocking reagent to which was added anti digoxigenin Fab fragment antibody (diluted 20,000:1) conjugated with alkaline phosphatase (Boehringer Mannheim, Biochemical Products, Indianapolis IN). After incubation for 30 minutes at room temperature membranes were washed in maleate buffer four times for 15 minutes each at room temperature. Next, membranes were soaked for 10 minutes in substrate buffer (100 mM TRIS, 100 mM NaCl, 50 mM MgCl~2~ pH 9.5) at room temperature. Incubation was then performed for ten minutes at room temperature with CDP-Star chemiluminescent substrate (Boehringer Mannheim, Biochemical Products, Indianapolis IN) diluted 1:100 in substrate buffer. Membranes were then exposed to X-ray film. Whole RNA blot analysis ----------------------- RNA was fractionated via denaturing agarose gel electrophoresis using glyoxal according to standard protocols \[[@B76]\]. Immediately before use the glyoxal was deionized by vortexing in the presence of mixed bed resin beads (Sigma Chemical Co. St. Louis, MO) until the pH was greater than 5. Determination of pH was made using Hydrion pH sensitive paper (Fisher Scientific, Pittsburgh, PA). Agarose gels (0.8%, 12 cm) cast in 10 mM sodium phosphate buffer were run for 2 hours at 100 volts with the buffer recirculated by a peristaltic pump. RNA gels were blotted using the same procedure as DNA gels omitting the depurination, denaturation, and neutralization steps. In addition, hybridization and chemiluminescent immunodetection were also performed using the same protocol for genome blot analysis except for the following. Just prior to prewetting in wash buffer for prehybridization RNA membranes were given two deglyoxylation washes in 20 mM Tris pH 8 for twenty minutes each at 65°C. The probe used for whole RNA blot analysis was a 1700 base pair PCR product. Generation of complementary DNA ------------------------------- Synthesis of cDNA was accomplished using MMLV reverse transcriptase (Gibco BRL Life Technologies, Bethesda, MA) and oligo dT according to the manufacturers instructions. Whole RNA was diluted to a concentration of 0.67 μg/μl in a total volume of 10--15 μl. The sample was heated at 90°C for 5 minutes and chilled on ice. The reverse transcription reaction mixture included the following components in a total volume of 20 μl. Nucleotides (dNTP\'s) (Amersham Pharmacia Biotech, Piscataway, NJ) were added to a final concentration of 0.5 mM. Forty units of a ribonuclease inhibitor (RNAsin) (Promega, Madison, WI) were added. Denatured, whole RNA (2 μg) was added in a volume of 3 μl. Oligo dT primer (1.1 μl) was added to a final concentration of 1.5 μM. Dithiothreitol was added to a final concentration of 10 mM. Dilution buffer (5X concentration, provided by the manufacturer) was added to a final concentration of 1X. After all the components were assembled, 200 units of MMLV reverse transcriptase (Gibco BRL Life Technologies, Bethesda, MA) were added. The sample was incubated at 42°C for 1 hour followed by heat denaturation of the enzyme at 95°C for 10 minutes. Samples were stored at -20°C. Polymerase Chain Reaction amplifications ---------------------------------------- In general, PCR reactions were performed in a volume of 50 μl with 50 picomoles of each primer, dNTP\'s (Amersham Pharmacia Biotech, Inc., Piscataway, NJ) at a final concentration of 0.5 mM, 1 μl of template, 1X PCR buffer (10X PCR buffer contained 700 mM Tris pH 8.8, 20 mM MgCl~2~, 1% Triton X-100, and 0.1% Tween 20) and Taq polymerase (Perkin-Elmer Inc., Wellesley MA) diluted to the manufacturers recommendations. The reaction components were assembled except Taq polymerase and heated at 95°C for 5 minutes. Polymerase was added and the PCR allowed to cycle for 1 minute at 94°C, 1 minute at 55°C and 1.5 minutes at 72°C for 30 cycles followed by 10 minutes at 72°C. All digoxigenin labeled probes for non-radioactive detection were generated using the Polymerase Chain Reaction. Probes were labeled using 10X PCR labeling mix (Boehringer Mannheim, Biochemical Products, Indianapolis IN) at a concentration of 1X replacing the normal dNTP mix. In each case 1 μl of template was added to the reaction mix. Template was created by amplifying a band off an Expressed Sequence Tag (EST G~5~B~3~T~7~) from the *Arabidopsis* expressed sequence tag database \[[@B78]\], fractionating the product on a low melting point agarose gel, and excising the band. The excised, diluted band was re-amplified to ensure fidelity of the PCR. After successful re-amplification the template was utilized for probe generation. A 300 base pair cDNA probe corresponding to nucleotide residues 779--1078 of the *A. thaliana* vacuolar type H^+^-ATPase subunit A gene was generated using primers SM-2 and SM-3 (Table [1](#T1){ref-type="table"}). This probe was used for whole genome blot analysis. An approximately 1700 base pair cDNA probe corresponding to nucleotide residues 274--1981 of subunit A was generated using primers VA-1 and VA-9. This probe was used for multi organ RNA blot analysis and stress response northerns. In both cases template was generated off EST G~5~B~3~T~7~. Cycling parameters for both probes were identical to those described above. Probes were precipitated by adding 1 μl of tRNA (10 mg/ml) or 1/50^th^ volume of glycogen (10 mg/ml in sterile distilled water) to the PCR reaction followed by 1/10^th^ volume of sterile 4 M lithium chloride. Next, 3 volumes of ice-cold absolute ethanol were added and the DNA precipitated for 30 minutes at -80°C. DNA was pelleted by centrifugation at 4°C for 15 minutes. Pellets were washed with ice cold 70% ethanol, air dried for 15 minutes and resuspended in 50 μl TE buffer/0.1 % SDS. Probes were quantitated by fractionation of 1/10^th^ volume in a 0.8% agarose gel compared to a known quantity of *Hind*III digested lambda DNA. Probes were stored at -20°C until use. Four different subunit A transcripts were amplified using RT-PCR. Primer pairs were designed with one upstream coding region primer and one primer anchored in the poly (A) tail of the transcript. In all four cases template was 1 μl of seedling cDNA. Amplification reaction mixtures were assembled as described, with PCR buffer modified to contain only 1 mM magnesium chloride. Transcript 1 was amplified using primers T-11 and T-12 (Table [1](#T1){ref-type="table"}). Transcript 2 was amplified using primers T-21 and T-22. Transcript 3 was amplified using primers T-31 and T-32. Transcript 4 was amplified using primers T-41 and T-42. The PCR was allowed to proceed in each case for 30 cycles of 1 minute at 94°C, 1 minute at the annealing temperature, and 1 minute at 72°C. Primers amplifying transcripts 1 and 2 were annealed at 35°C, primers amplifying transcripts 3 and 4 were annealed at 40°C. Following 30 cycles, primer extensions were completed at 72°C for 10 minutes. PCR products were then fractionated on 2.5% agarose gels. The transcript 4 amplification product was used as template for the generation of a digoxigenin labeled probe for use in the subunit A transcript stress response experiment. The transcript 4 amplification product was purified via low melting point agarose gel electrophoresis and used as template for probe generation. The digoxigenin labeled probe was generated as described using the transcript 4 cycling parameters and primers T-41 and T-42. DNA sequence of all PCR primers used in this research can be found in Table [1](#T1){ref-type="table"}. A diagrammatic representation of the PCR amplification products generated from transcript specific primers is shown in fig. [6](#F6){ref-type="fig"}. Abbreviations ============= H^+^-ATPase-proton pumping adenosine triphosphatase, V-ATPase-vacuolar type proton pumping adenosine triphosphatase, EST-expressed sequence tag, PCR-Polymerase Chain Reaction, RT-PCR-reverse transcriptase-Polymerase Chain Reaction. Acknowledgments =============== The authors wish to thank Shenell Antrobus and Alex Delcampo for their excellent technical assistance. This research was supported through NSF grant BSR 9020868 and by the University of Connecticut Research Foundation.

Background ========== Sequencing of Expressed Sequence Tags (ESTs) has resulted in the rapid identification of expressed genes \[[@B1]\]. ESTs are single-pass, partial sequences of cDNA clones from a large number of disease and normal tissue libraries. ESTs have been used extensively for gene discovery and for transcript mapping of genes from a wide number of organisms \[[@B2]-[@B4]\]. Even with the finished working draft of the human genome, the generation of a complete and non-redundant catalog of human genes is still a big challenge facing the genome research community. Full-length cDNA data are currently available for only 10,000 human genes \[[@B5]\], less than one-third of the total using the most conservative recent estimates of human gene numbers \[[@B6],[@B7]\]. Evidence of differential expression is one of the most important criteria in prioritizing the exploitation of genes in both academic and pharmaceutical research \[[@B8]-[@B10]\]. While identifying individual differentially expressed genes attracts most of the interest, a genome wide transcriptome map may not only provide a tool to identify candidate genes that are over-expressed or silenced in certain disease tissue, but may also help to understand the structure and organization of the genome. Genomes are the blueprints of life and they should not be considered as a simple collection of genes. In fact, the organization of genes into operons, complex regulons \[[@B11]\], or pathogenicity islands \[[@B12]\] suggests that related functions usually share physical proximity. Different types of transcriptome maps can help to identify different types of transcription domains. Those domains can now be analyzed as to how they relate to known nuclear substructures, such as nuclear speckles, PML bodies and coiled bodies \[[@B13]-[@B15]\]. Two strategies have been commonly used to evaluate large-scale gene expression: experimental and computational. The former is represented by DNA microarray technology \[[@B16]\]. Computational methods consist of generating a large number of random ESTs from non-normalized cDNA libraries. The variation in the relative frequency of those tags, stored in databases, are then used to point out the differential expression of the corresponding genes: this is the so called \"digital Northern\" comparison. Digital Northern data can be used to provide quantitative assessment of differential expression within a certain limit \[[@B17]\]. Velculescu et al. \[[@B18]\] introduced another digital method called serial analysis of gene expression (SAGE). The SAGE method requires only nine nucleotides, therefore allowing a larger throughput. In both protocols, the number of tags is reported to be proportional to the abundance of cognate transcripts in the tissue or cell type used to make the cDNA library. The recently announced first draft of the human genome \[[@B19],[@B20]\] holds in it an unprecedented wealth of information, available for public study and scrutiny. How are genes organized in the human genome? Is there any distribution pattern of tissue specific genes in terms of chromosomal location? In this study, we combined the concept of digital Northern and transcript mapping for all public and Incyte LifeSeq ESTs to evaluate the tissue specific transcriptome. The goal of this paper is not to evaluate the digital expression of individual genes; instead we are looking at the tissue enriched digital expression level for a given chromosomal region. Particularly, we looked at the distribution pattern of brain-enriched genes in the genome and how that pattern changes in brain tumor tissues. We are well aware of the fact that this method and associated approaches are quite primitive. However, the tissue specific transcriptome data strongly suggest that human genome organization is correlated to the tissue type and its dynamics. Results ======= Distribution of brain-enriched genes along the chromosomes ---------------------------------------------------------- With the unavailability of the complete annotated human gene catalog, it is not practical to document each individual gene that is expressed in brain within one chromosome region. Since the number of sequence tags is reported to be proportional to the abundance of cognate transcripts in the tissue or cell type used to make a given cDNA library, the number of ESTs within a chromosome region should reflect the abundance of the cognate transcripts in that region. Therefore comparison of the abundance of brain tissue derived transcripts relative to those from other tissues within the same chromosome region can highlight regions that have more brain-enriched gene expression. We performed digital expression analysis of brain-enriched genes across the human genome with a window size of 5 Mbp and an interval of 1 Mbp. The transcript density factor for normal (non-tumor) brain libraries (TDF~NB~) was calculated as described in Methods. Figure [1](#F1){ref-type="fig"} is an example of the distribution of TDF~NB~ over chromosome 1 using publicly available EST sequences from dbEST and reveals a number of \"peaks\" that represent transcripts that appear to be preferentially expressed in brain tissues. To check the validity of these peak regions and make sure that the difference is not due to random picking or partial sequencing of cDNA libraries (which is the common random fluctuation caused by digital Northern approach) \[[@B17]\], the analysis was repeated using ESTs and the associated library information from the Incyte Genomics LifeSeq database. The distribution pattern of TDF~NB~ shows an overall correlation coefficient of 0.658 for the whole genome between these two data sources. If we only analyze the region with Z-score \>= 2 (i.e. peak regions), the correlation coefficient is 0.935 which suggests that the peak regions resulting from the analysis of public data are most likely not artifacts. Figure [2](#F2){ref-type="fig"} shows the comparison of the distributions of TDF on chromosome 1 calculated from ESTs derived from brain tissue libraries vs. ESTs derived from breast tissue libraries. The overall pearson correlation coefficient for these two tissues is 0.113 which suggests that the peak regions observed in Figure [1](#F1){ref-type="fig"} are brain specific. {#F1} {#F2} There are 16 high TDF~NB~ regions (enriched with brain specific expression) with Z-score \>= 2 (Table [1](#T1){ref-type="table"}). To assess the validity of our finding using public data, similar analysis using Incyte LifeSeq data shows that the same peaks can be derived (data not shown). Table [2](#T2){ref-type="table"} summarizes all the low TDF~NB~ regions (lack of brain specific expression) over the whole genome with Z-score \>= 2. It\'s interesting to note that the majority of high TDF~NB~ or low TDF~NB~ regions have close to average gene density indicating that those regions are not biased toward extremely high or low gene density. ::: {#T1 .table-wrap} ::: {.caption} ###### Summary of chromosome regions with significant brain-enriched gene expression (high TDF~NB~ region) (window size 5 Mbp, Z-score \>= 2.0). All regions are confirmed by separate analysis using Incyte LifeSeq ESTs. Pearson Correlation Coefficient = 0.935. ::: Chromosome Region(Mbp) TDF~NB~ Gene Density Ratio ------------ ------------- --------- -------------------- 1 208--212 0.68 0.33 2 14--18 0.46 0.76 3 90--96 0.39 0.58 5 184--188 0.30 1.0 6 82--86 0.32 0.79 6 138--142 0.28 0.67 7 73--77 0.56 0.59 10 145--149 0.43 0.1 12 85--89 0.36 0.75 13 65--69 0.35 0.71 15 21--26 0.69 0.78 18 80--87 0.88 0.83 20 8--15 0.54 0.52 22 46--50 0.30 0.17 X 103--109 0.43 0.58 X 150--155 0.57 0.25 ::: ::: {#T2 .table-wrap} ::: {.caption} ###### Summary of chromosome regions lacking in brain-enriched gene expression (low TDF~NB~ regions) (window size 5 Mbp, Z-score \>= 2.0). All regions are confirmed by separate analysis using Incyte LifeSeq ESTs. Pearson Correlation Coefficient = 0.935. ::: Chromosome Region (Mbp) TDF~NB~ Gene Density Ratio ------------ -------------- --------- -------------------- 1 117--111 -0.80 1.11 1 218--222 -0.56 0.50 1 229--233 -0.62 0.79 2 92--99 -0.99 1.79 3 175--179 -0.76 1.0 3 217--211 -0.97 0.76 4 75--81 -1.31 2.78 7 152--159 -0.80 1.26 9 117--122 -0.76 0.99 10 57--61 -0.82 0.62 11 55--59 -1.24 0.25 11 111--117 -0.92 0.94 12 55--60 -0.71 4.15 14 111--116 -1.08 0.81 19 30--35 -0.60 0.41 19 53--57 -0.62 1.58 21 7--11 -1.83 0.03 22 10--14 -0.43 0.10 ::: Expression profile change of some chromosome regions with extreme TDF~NB~ in brain tumor ---------------------------------------------------------------------------------------- Our analysis strongly suggests that brain-enriched genes are distributed throughout the genome in a non-random fashion. Some regions are dense with brain-enriched genes or brain specific expression. It would be interesting to know if any of these patterns change during tumorgenesis. A similar analysis was performed using ESTs generated from brain tumor libraries and their digital expression profile relative to the pooled tissue was plotted against the genome. The chromosomal distributions of these putative brain tumor enriched transcripts and the normal brain enriched transcripts are quite different. Table [3](#T3){ref-type="table"} lists all the chromosome regions with high TDF in non-tumor brain libraries (TDF~NB~) which become low TDF or neutral TDF in brain tumor (TDF~TB~). Chr15, 21--25 Mbp, Chr12, 85--89 Mbp, and Chr18, 45--52 Mbp (Figure [3](#F3){ref-type="fig"}) are some of the examples. While most of the low TDF regions in normal brain remain low in brain tumor, a few regions did become high TDF regions in brain tumor tissues (Table [3](#T3){ref-type="table"}). Chr2, 93--99 Mbp and Chr19, 53--58 Mbp (Figure [4](#F4){ref-type="fig"}) are two examples. The digital expression profile in those regions was further confirmed by using data from Incyte LifeSeq (data not shown). {#F3} {#F4} ::: {#T3 .table-wrap} ::: {.caption} ###### High TDF~NB~ and low TDF~NB~ (Z-score \>= 2) regions which show significant differential expression between tumor brain tissues and non-tumor brain tissues. ::: Chromosome Region (Mbp) TDF~NB~ TDF~TB~ ------------ -------------- --------- --------- 2 14--18 0.45 0.06 2 91--99 -0.93 0.11 10 58--62 -0.74 0.92 12 85--89 0.36 -0.56 13 54--61 -0.78 0.19 15 21--25 0.74 0.21 18 45--52 1.0 -1.63 19 53--58 -0.64 1.8 20 10--15 0.64 -0.62 22 30--34 -0.48 0.15 X 103--109 0.43 0.09 X 24--28 -0.72 -0.11 ::: Discussion ========== A genome is not a simple collection of genes. It has been reported that significant correlation exists between the distribution of genes along the chromosome and the physical architecture of the cell in bacteria \[[@B21]\]. The human genome is much more complex and a complete understanding of its organization awaits completion of the finished sequence as well as a definitive annotation of the human gene catalog. Considerable evidence has already shown that related genes tend to exist as clusters in the genome. For example, 80% of the over 900 olfactory genes are found in clusters of 6--138 genes \[[@B22]\]. The 3.6 Mbp human major histocompatibility complex (MHC) on chromosome 6p21.3 is a critical repository for the immune response genes \[[@B23]\]. Extensive analysis of the genomic organization of the MHC region has revealed that at least 27 of its resident genes possess duplicated copies in at least one of the three other restricted chromosomal regions 1q21-q25, 9q33-q34 and 19q13.1-p13.4. For another example, ABC transporter gene family members are located on 6p21.3, 1q25 and 9q34 as clusters. The development of distinct tissue and cell types is a fundamental characteristic of growth in higher organisms. Tissue and cellular differentiation, in turn, is highly dependent on specific patterns of gene expression and transcript accumulation. Many studies have been successfully used to pinpoint genes exhibiting tissue or disease specific expression. This study suggests another approach that focuses on the tissue specific transcriptome map study and attempts to study the genomic proximity of tissue specific genes. We show here that some regions on specific chromosomes are enriched with brain-enriched gene expression. Given the chromosome window size (5 Mbp) used to calculate the TDF and the average size of the gene (\~30 Kbp), each 5 Mbp region should contain on average about 170 genes. In addition, we report here only regions that have a normal level of gene density. Therefore the very high TDF~NB~ regions are most likely contributed by the high level expression of brain-enriched gene clusters or the regions containing a high density of brain-enriched genes. We are aware of the importance of those brain-enriched genes that scatter across the genome whose brain specific or differential expression is diluted by the large volume of neighbor genes within the 5 Mbp window region. Another limitation of this approach is the inability to reveal some regions with equal up-down differential expression. The goal of this study is to pinpoint the chromosomal regions with the most significant brain-enriched gene expression and to elucidate the non-randomness of the tissue specific expression over the chromosome. Naturally, gene proximity within chromosomes is already known to be significant. Finding neighbors of a given gene can shed light on that gene, especially when the neighbors contain objects with similar features. Therefore, this tissue specific transcriptome map study may not only help us to understand the genome organization in the future, but may also provide means to gain leads to the functions of many genes for which this information is not currently available. Methods ======= Resources for databases and computer programs --------------------------------------------- GenBank release 120 was downloaded from <>. ESTs and their associated tissue library information and clone information were extracted and organized in a relational database (Sybase, SQL Server Release 11.0, CA, Sybase Inc.). The EST cDNA libraries were manually curated and catalogued into non-tumor brain libraries and brain tumor libraries. We obtained 208604 dbEST ESTs from 369 non-tumor brain cDNA libraries, 67351 ESTs from 148 brain tumor cDNA libraries. We also extracted 363473 ESTs from 100 Incyte LifeSeq non-tumor brain cDNA libraries and 126538 ESTs from their 23 brain tumor libraries. All non-commercial software used in this study was written in PERL 5.0. Transcript mapping ------------------ Transcript mapping was done based on the October 2000 Freeze of the University of California at Santa Cruz\'s working draft sequence <>, which presents a tentative assembly of the finished and draft human genomic sequence based on the Washington University-Saint Louis clone map <>. We mapped all the public ESTs (2.5 millions ESTs) from dbEST as well as the ESTs from Incyte Genomics\' LifeSeq database (5.1 millions ESTs) using a local alignment software package AAT. AAT is a local alignment software which extended the BLAST algorithm by assigning fixed penalty to long gaps \[[@B24]\]. To reduce the number of undesirable matches due to interspersed repeats, the DNA sequence is screened for interspersed repeats using the RepeatMasker program (Smit, AFA & Green, P et al <>). Only those ESTs that have over 95% identity to the genomic counterpart over half the length of ESTs\' length or 50 bp whichever is longer are included. Calculation of the transcript density factor (TDF) -------------------------------------------------- The TDF for normal brain-enriched gene expression (TDF~NB~) is calculated using 5 Mbp window moving along the chromosome with 1 Mbp interval and defined as: TDF~NB~ = ln(R~NB~/R)/(T~NB~/T) Where R~NB~ is the number of EST clones derived from non-tumor brain tissue within a window of 5 Mbp. R is the number of EST clones derived from all non-tumor tissue pooled libraries within a window of 5 Mbp. T~NB~ is the number of total mapped EST clones derived from non-tumor brain tissues. T is the number of total mapped EST clones derived from all non-tumor tissue pooled libraries. The distribution of TDF~NB~ of all the 5 Mbp regions should approximate a Gaussian distribution. Theoretically, TDF~NB~ should approach 0 if the expression level of genes in brain tissues has no difference from that in pooled tissues within a 5 Mbp chromosomal region. We define all the regions with high TDF~NB~ (Z-score \>= 2) as brain-enriched regions. Those chromosome regions that have high brain-enriched gene expression are referred as high TDF~NB~ region. Those chromosome regions that have low brain-enriched gene expression are referred as low TDF~NB~ region. The calculation of transcript density factor for brain tumor (TDF~TB~) is similar to the calculation of TDF~NB~. TDF~TB~ = ln(R~TB~/R)/(T~TB~/T) Where R~TB~ is the number of EST clones derived from tumor brain tissue within a window of 5 Mbp. R is the number of EST clones derived from all non-tumor tissue pooled libraries within a window of 5 Mbp. T~TB~ is the number of total mapped EST clones derived from tumor brain tissue libraries. T is the number of total mapped EST clones derived from all non-tumor tissue libraries. Correlation analysis of distribution pattern of TDF using public and Incyte data -------------------------------------------------------------------------------- The Pearson correlation coefficient (r) represents the degree of similarity (strength of correlation) between two sets of data. The correlation coefficient is calculated as follows:  where X~i~ is TDF values derived from public data and Y~i~ is the TDF values derived from the Incyte data. Values for the Pearson correlation coefficient range from -1 to 1 where zero indicates no correlation, -1 indicates a perfect negative correlation and 1 indicates a perfect positive correlation. The calculation of gene density ------------------------------- Using 5 Mbp window moving along the chromosome with 1 Mbp interval, the gene density ratio is defined as UniGene (Version 5.002) \[[@B4],[@B25]\] count divided by the average UniGene count in each 5 Mbp region. Average gene density ratio is equal to 1.0.

Introduction {#s1} ============ Stimulus discrimination and generalisation constitute two major abilities exhibited by most living animals. Discrimination allows treating different signals as distinct, while generalisation allows treating different but similar stimuli as equivalents \[[@pbio-0030060-b01],[@pbio-0030060-b02],[@pbio-0030060-b03]\]. Similarity along one or several perceptual dimensions determines the degree of generalisation between stimuli \[[@pbio-0030060-b02]\]. Determining such dimensions is fundamental for defining an animal\'s perceptual space. This objective remains, however, elusive in the case of the olfactory modality in which the dimensions along which odours are evaluated are not well known. Characteristics such as the functional chemical group or the carbon-chain length of a chemical substance may influence olfactory perception. It is known that at least some features of odorant molecules influence olfactory perception. For instance, some enantiomers can be discriminated by humans and nonhuman primates \[[@pbio-0030060-b04]\]. If and how chemical group and carbon-chain length are integrated as inner dimensions into an olfactory perceptual space remains unknown. Vertebrate and invertebrate nervous systems show important functional as well as anatomical similarities in the way in which olfactory signals are detected and processed in their brains, particularly at the level of their first olfactory centres, the olfactory bulb in the case of vertebrates and the antennal lobe (AL) in the case of insects \[[@pbio-0030060-b05],[@pbio-0030060-b06],[@pbio-0030060-b07]\]. Insects are useful models for studying olfaction, as their behaviour heavily relies on the use of olfactory cues. The honeybee Apis mellifera is one such model in which behavioural and neurobiological studies have been performed to unravel the basis of olfaction \[[@pbio-0030060-b08],[@pbio-0030060-b09],[@pbio-0030060-b10],[@pbio-0030060-b11]\]. Honeybee foragers are 'flower constant\' and learn and memorise a given floral species that they exploit at a time as long as it is profitable. Floral cues, among which odours play a prominent role, are then associated with nectar or pollen reward \[[@pbio-0030060-b12],[@pbio-0030060-b13]\]. However, under natural conditions, the blends of volatiles emitted by floral sources vary widely in quantity and quality both in time and in space \[[@pbio-0030060-b14],[@pbio-0030060-b15]\]. To cope with such changes in an efficient way, a 'flower constant\' forager should be able to generalise its choice to the same kind of floral sources despite fluctuations in their volatile emissions. In a pioneering investigation, von Frisch \[[@pbio-0030060-b16]\] trained freely flying bees to visit an artificial feeder presenting several essential oils (odour mixtures). Using a set of 32 odour mixtures, von Frisch observed that after learning that a blend was associated with sucrose solution, bees tended to prefer this odour blend, but they sometimes visited other blends that were similar (to the human nose) to the rewarded one. Olfactory generalisation in honeybees was mainly studied on restrained honeybees using the conditioning of the proboscis extension reflex (PER) \[[@pbio-0030060-b17],[@pbio-0030060-b18]\]. In this paradigm, harnessed honeybees are conditioned to odours associated with a sucrose reward. When the antennae of a hungry bee are touched with sucrose solution, the animal reflexively extends its proboscis to reach out towards and to lick the sucrose. Odours presented to the antennae do not usually release such a reflex in naive animals. If an odour is presented immediately before sucrose solution (forward pairing), an association is formed and the odour will subsequently trigger the PER in a subsequent unrewarded test. This effect is clearly associative and involves classical conditioning \[[@pbio-0030060-b18]\]. Thus, the odour can be viewed as the conditioned stimulus (CS), and sucrose solution as an appetitive unconditioned stimulus (US). Bees conditioned to individual odours or to olfactory mixtures can generalise PER to a wide range of different olfactory stimuli. Using the PER paradigm, Vareschi \[[@pbio-0030060-b19]\] showed that bees generalise most often between odours with similar carbon-chain lengths and between odours belonging to the same functional group. However, Vareschi conditioned odours in a differential way, with two rewarded and many unrewarded odours, so that several generalisation gradients (excitatory and inhibitory) may have interacted in an unknown way to determine the generalisation responses exhibited by the bees \[[@pbio-0030060-b19]\]. Using a similar approach and a restricted (6 × 6) set of odour combinations, Smith and Menzel \[[@pbio-0030060-b20]\] confirmed that bees generalise among odours with the same functional group, but their analysis did not detail the results obtained with individual odour combinations, thus rendering impossible the analysis of generalisation between odours with similar carbon-chain lengths. Free-flying bees trained in a differential way to a rewarded odour presented simultaneously with multiple unrewarded odours also generalise between odours with similar functional groups \[[@pbio-0030060-b21]\]. As for Vareschi\'s study \[[@pbio-0030060-b19]\], such an experimental design makes it difficult to interpret the generalisation responses due to unknown interactions between excitatory and inhibitory generalisation gradients. Recently, optical imaging studies facilitated our understanding of how olfactory stimuli are detected and processed in the bee brain \[[@pbio-0030060-b22],[@pbio-0030060-b23],[@pbio-0030060-b24],[@pbio-0030060-b25],[@pbio-0030060-b26]\]. The first relay of the bee\'s olfactory system involves the ALs, which receive sensory input from the olfactory receptor neurons of the antennae within a number of 160 functional units, the glomeruli \[[@pbio-0030060-b27],[@pbio-0030060-b28],[@pbio-0030060-b29]\]. Within each glomerulus, synaptic contacts are formed with local interneurons and projection neurons (PNs). PNs send processed information from the ALs to higher brain centres such as the mushroom bodies and the lateral protocerebrum \[[@pbio-0030060-b30]\]. Stimulation with an odour leads to a specific spatiotemporal pattern of activated glomeruli, as shown, using in vivo calcium imaging techniques that employ fluorescent dyes to measure intracellular calcium in active neurons \[[@pbio-0030060-b22],[@pbio-0030060-b24],[@pbio-0030060-b31]\]. The odour-evoked activity patterns are conserved between individuals and constitute therefore a code \[[@pbio-0030060-b23],[@pbio-0030060-b24]\]. Odours with similar chemical structures tend to present similar glomerular activity patterns \[[@pbio-0030060-b23]\]. Furthermore, it is believed that the neural code of odour-evoked glomerular patterns measured in the bee brain actually represent the perceptual code, although this idea was never tested directly. In the present work, we studied behavioural olfactory generalisation, using the PER conditioning paradigm, with 16 odorants varying in two chemical features, functional group and chain length. The odours belonged to four chemical categories: alcohols with the functional group on the first or second carbon of the carbon chain (henceforth primary and secondary alcohols, respectively), aldehydes, and ketones. They possessed therefore three functional groups (alcohol, aldehyde, ketone). Their chain length ranged from six to nine carbon atoms (C6, C7, C8, and C9). The pairwise combination of 16 odours defined a 16 × 16 matrix. These odours are well discriminated by free-flying bees \[[@pbio-0030060-b21]\] and give consistent odour-evoked signals in optical imaging studies \[[@pbio-0030060-b23]\]. Using a behavioural approach, we measured similarity between odours and calculated their perceptual distances in a putative olfactory space. These perceptual distances were correlated with physiological distances measured in optical imaging experiments \[[@pbio-0030060-b23]\]. The correlation between both datasets was highly significant, thus indicating that odours that are encoded as physiologically similar are also perceived as similar by honeybees. Although other studies have addressed the issue of perceptual correlates of neural representations \[[@pbio-0030060-b32],[@pbio-0030060-b33]\], we show for the first time that neural olfactory activity corresponds to olfactory perception defined on the basis of specific dimensions in a putative olfactory space, a finding that is of central importance in the study of the neurobiology of perception. Results {#s2} ======= We trained 2,048 honeybees along three trials in which one of the 16 odours used in our experiments was paired with a reward of sucrose solution (conditioned odour). Afterwards, each bee was tested with four odours that could include or not include the trained odour. Acquisition Phase {#s2a} ----------------- The level of PER in the first conditioning trial was very low (between 0% and 8.60%) for all odours ([Figure 1](#pbio-0030060-g001){ref-type="fig"}). All the 16 odours were learnt but not with the same efficiency. An overall (trial × odour) analysis of variance (ANOVA) showed a significant increase in responses along trials (*F* ~2,\ 4064~ = 2215.50, *p* \< 0.001) and a significant heterogeneity among odours (*F* ~15,\ 2032~ = 8.80, *p* \< 0.001). Responses to the CS in the last conditioning trial reached a level of approximately 70% for primary and secondary alcohols, 80% for aldehydes, and 61% for ketones. {#pbio-0030060-g001} In the case of aldehydes and primary and secondary alcohols, no significant chain-length effect within functional groups was found over the whole conditioning procedure (chain length × trial ANOVA; chain-length effect for primary alcohols: *F* ~3,\ 508~ = 0.18, *p* \> 0.05; secondary alcohols: *F* ~3,\ 508~ = 1.47, *p* \> 0.05; and aldehydes: *F* ~3,\ 508~ = 1.26, *p* \> 0.05). In contrast, bees conditioned to ketones showed a significant chain-length effect in the acquisition (chain length × trial ANOVA; chain-length effect: *F* ~3,\ 508~ = 20.00, *p* \< 0.005). Scheffé post hoc comparisons showed that acquisition was significantly better for nonanone (81.25% responses in the last conditioning trial) than for all other ketones. Octanone (68.75% responses in the last conditioning trial) was also better learned than hexanone and heptanone (45.31% and 48.44% responses in the last conditioning trial, respectively) ([Figure 1](#pbio-0030060-g001){ref-type="fig"}, bottom right). The effect over trials was significant in all cases (*p* \< 0.05) as bees learned all odours. The analysis of acquisition for each chain length separately revealed that it varied significantly depending on the functional group (functional group × trial ANOVA; C6: *F* ~3,\ 508~ = 18.89; *p* \< 0.005; C7: *F* ~3,\ 508~ =10.78; *p* \< 0.005; C8: *F* ~3,\ 508~ = 3.84; *p* \< 0.01; C9: *F* ~3,\ 508~ = 2.73, *p* \< 0.05). Scheffé post hoc comparisons generally showed that this effect was mainly due to ketones being less well learned than aldehydes and alcohols. Generally, the longer the carbon chain, the lower the heterogeneity in acquisition between functional groups. Thus, apart from short-chain ketones, all odours were learned similarly (reaching a level of acquisition between 60% and 80% in the last conditioning trial). Test Phase {#s2b} ---------- When the conditioned odour was presented in a test ([Figure 1](#pbio-0030060-g001){ref-type="fig"}, grey panels), the level of PER recorded corresponded mainly to that found in the last acquisition trial (McNemar tests \[2 × 2 Table\]: in all cases *p* \> 0.05). To compare generalisation after conditioning, and because acquisition levels were heterogeneous between odours, we built a generalisation matrix in which only bees responding to the CS at the end of training (3rd conditioning trial) were considered ([Figure 2](#pbio-0030060-g002){ref-type="fig"}). The number of individuals included in the statistical analysis varied within each 'training odour/test odour\' pair. The number of bees completing the tests varied between 17 and 28 for primary alcohols, between 13 and 29 for secondary alcohols, between 23 and 30 for aldehydes, and between 11 and 31 for ketones. The responses to the CS in the tests ranged between 70% and 100% in the generalisation matrix. All further analyses were carried out on this matrix. In the following sections, we will use the matrix data to analyse generalisation within and between functional groups, within and between chain lengths, and the asymmetries in olfactory generalisation. {#pbio-0030060-g002} Generalisation within Functional Groups {#s2c} --------------------------------------- [Figure 3](#pbio-0030060-g003){ref-type="fig"}A shows the percentage of PER to odours having different (white quadrants) or the same (grey quadrants) functional group as the conditioned odour. High levels of PER to odours different from the trained one correspond to high generalisation. In order to better visualise generalisation as depending on functional groups, we pooled all the observed responses within each quadrant of [Figure 3](#pbio-0030060-g003){ref-type="fig"}A (i.e., not considering chain length) and calculated the resulting percentage of PER ([Figure 3](#pbio-0030060-g003){ref-type="fig"}B). Grey bars correspond to generalisation to the same functional group; white bars correspond to generalisation to different functional groups. Generalisation mainly occurred within a given functional group (grey bars). This pattern was clearest for aldehydes ([Figure 3](#pbio-0030060-g003){ref-type="fig"}B, [3](#pbio-0030060-g003){ref-type="fig"}rd row) because bees conditioned to aldehydes responded with a high probability to other aldehydes but showed lower responses to any other odour (see also the clear aldehyde "response block" in [Figure 2](#pbio-0030060-g002){ref-type="fig"}). {ref-type="fig"}) represented as two-dimensional graphs for each conditioned odour. The right ordinate represents the CSs categorised in four functional groups, primary alcohols, secondary alcohols, aldehydes, and ketones (from top to bottom). The abscissa represents the test odours aligned in the same order as the conditioned odours (from left to right). The left ordinate represents the percentage of proboscis extensions to the test odours after being trained to a given odour. Each quadrant in the figure represents generalisation responses to one functional group after training for the same (grey quadrants) or to a different functional group (white quadrants).\ (B) Same data as in (A), but the observed responses within each quadrant were pooled and the resulting percentage of responses per quadrant was calculated. The abscissa and the right ordinate represent the four functional groups. The left ordinate represents the percentage of proboscis extensions to each of these groups after being trained to a given group. Grey bars correspond to grey quadrants in (A) and represent generalisation to the same functional group as the conditioned one. White bars correspond to white quadrants in (A) and represent generalisation to a functional group different from the conditioned one: 1-ol, 2-ol, al, and one mean primary alcohol, secondary alcohol, aldehyde, and ketone, respectively. Asterisks indicate significant differences along a row or a column (*p* \< 0.001)\ (C) Within-functional group generalisation, depending on chain length. The abscissa represents the functional groups tested. The ordinate represents the percentage of proboscis extensions to the functional groups tested after being trained to a given chain-length (lines). Thus, for instance, the first point to the left for C9 molecules (black circles) represents generalisation to 1-hexanol, 1-heptanol, and 1-octanol after conditioning to 1-nonanol. A significant heterogeneity was found in within-functional group generalisation for C8 and C9 but not for C6 and C7 molecules.\ (D) Generalisation within-functional groups. The figure shows results from pooling the data of (C) corresponding to each functional group. Each point shows the percentage of proboscis extensions to odours of the same functional group as the conditioned odour. Within-group generalisation was significantly heterogeneous (asterisks, *p* \< 0.001). Pairwise comparisons showed that generalisation within aldehydes was significantly higher than within primary alcohols or ketones and marginally higher than within secondary alcohols (different letters indicate significant differences).](pbio.0030060.g003){#pbio-0030060-g003} We analysed within-functional group generalisation as depending on chain length (see [Figure 3](#pbio-0030060-g003){ref-type="fig"}C). To this end we represented generalisation from C6, C7, C8, and C9 molecules having a given functional group to the other compounds having the same functional group (e.g., [Figure 3](#pbio-0030060-g003){ref-type="fig"}C, black circle curve, first data point: generalisation to 1-hexanol, 1-heptanol, and 1-octanol after conditioning to 1-nonanol). A significant heterogeneity appeared for C8 and C9 molecules (χ^2^ = 12.60 and 14.30, respectively, *p* \< 0.01 in both cases, *n* = 67--85) but not for C6 and C7 molecules (*p* \> 0.05). In the case of C8 and C9 molecules, generalisation was significantly higher within aldehydes (*p* \< 0.05). When comparing within-group generalisation over all four functional groups ([Figure 3](#pbio-0030060-g003){ref-type="fig"}D), a significant heterogeneity appeared (χ^2^ = 14.40, *df* = 3, *p* \< 0.01, *n* = 276--316). Pairwise comparisons (using a corrected threshold for multiple comparisons: α′ = 0.017) showed that generalisation within aldehydes was significantly higher than within primary alcohols (χ^2^ = 11.80, *df* = 1, *p* \< 0.0006) and ketones (χ^2^ = 9.90, *df* = 1, *p* \< 0.005) and close to significance in favour of aldehydes when compared to secondary alcohols (χ^2^ = 4.40, *df* = 1, 0.017 \< *p* \< 0.05). Generalisation within Chain Lengths {#s2d} ----------------------------------- [Figure 4](#pbio-0030060-g004){ref-type="fig"}A shows the generalisation responses of bees to odours having different (white quadrants) or the same (grey quadrants) chain length as the conditioned odour. In order to better visualise generalisation as depending on chain length, we pooled all the observed responses within each quadrant of [Figure 4](#pbio-0030060-g004){ref-type="fig"}A and calculated the resulting percentage of PER ([Figure 4](#pbio-0030060-g004){ref-type="fig"}B). Grey bars correspond to generalisation to the same chain length; white bars correspond to generalisation to different chain lengths. Generalisation was highest in the case of odours with the same or similar chain length. {ref-type="fig"}) represented as two-dimensional graphs for each conditioned odour. The right ordinate represents the CSs categorised in four chain lengths, C6, C7, C8, and C9 molecules (from top to bottom). The abscissa represents the test odours aligned in the same order as the conditioned odours (from left to right). The left ordinate represents the percentage of proboscis extensions to the test odours after being trained for a given odour. Each quadrant in the figure represents generalisation responses to one chain length after training for the same (grey quadrants) or to a different chain length (white quadrants).\ (B) Same data as in (A), but the observed responses within each quadrant were pooled and the resulting percentage of responses per quadrant was calculated. The abscissa and the right ordinate represent the four chain-length categories. The left ordinate represents the percentage of proboscis extensions to each of these categories after being trained for a given chain-length category. Grey bars correspond to grey quadrants in (A) and represent generalisation to the same chain length as the conditioned one. White bars correspond to white quadrants in (A) and represent generalisation to a chain length different from the conditioned one: C6, C7, C8, and C9 mean chain length of 6, 7, 8, and 9 carbons, respectively. Asterisks indicate significant differences along a row or a column (*p* \< 0.001).\ (C) Within chain-length generalisation as depending on functional group. The abscissa represents the chain lengths tested. The ordinate represents the percentage of proboscis extensions to the same chain length after being trained to a given functional group (lines). Thus, the first point to the left for ketones (red circles) represents generalisation to 1-hexanol, 2-hexanol, and hexanal after conditioning to 2-hexanone; the second point represents generalisation to 1-heptanol, 2-heptanol, and heptanal after conditioning to 2-heptanone. A significant heterogeneity was found in within-chain-length generalisation for aldehydes and ketones.\ (D) Generalisation within-chain lengths. The figure results from pooling the data of (C) corresponding to each chain length. Each point shows the percentage of proboscis extensions to odours of the same chain length as the conditioned odour. Within-chain-length generalisation was significantly heterogeneous (asterisks, *p* \< 0.001). Pairwise comparisons showed that generalisation within C9 molecules was significantly higher than within C7 and C6 molecules and marginally higher than within C8 molecules (different letters indicate significant differences).](pbio.0030060.g004){#pbio-0030060-g004} We analysed within-chain length generalisation as depending on functional group ([Figure 4](#pbio-0030060-g004){ref-type="fig"}C). To this end we represented generalisation from primary alcohols, secondary alcohols, aldehydes, or ketones of a given chain length to the other compounds having the same chain length (e.g., [Figure 4](#pbio-0030060-g004){ref-type="fig"}C, red circle curve, first data point: generalisation to 1-hexanol, 2-hexanol, and hexanal after conditioning to 2-hexanone). Generalisation within-chain length was generally higher for longer than for shorter chain lengths. This effect was significant for aldehydes (χ^2^ = 28.70, *df* = 3, *p* \< 0.01, *n* = 75--80) but not for primary and secondary alcohols (χ^2^ = 5.20 and 3.4, *df* = 3, *p* \> 0.05, *n* = 67--73 and *n* = 61--66, respectively). For ketones, a significant heterogeneity was found (χ^2^ = 10.00, *df* = 3, *p* \< 0.05, *n* = 40--79), but generalisation was more important between C8 than between C7 molecules. The generalisation corresponding to other chain lengths fell in between. When comparing within-chain length generalisation over all four chain-length groups ([Figure 4](#pbio-0030060-g004){ref-type="fig"}D, i.e., not considering functional group), a significant heterogeneity appeared χ^2^ = 23.2, *df* = 3, *p* \< 0.001, *n* = 247--293). Pairwise comparisons (using a corrected threshold for multiple comparisons: α′ = 0.017) showed that within-chain length generalisation was significantly higher within C9 than within C6 (χ^2^ = 18.50, *df* = 1, *p* \< 0.0001) and C7 molecules (χ^2^ = 15.00, *df* = 1, *p* \< 0.0001). Generalisation within C8 molecules was close to significance when compared to generalisation within C9 molecules (χ^2^ = 5.00, *df* = 1, 0.017 \< *p* \< 0.05), and it was significantly higher than generalisation within C6 molecules (χ^2^ = 4.3, *df* = 1, 0.017 \< *p* \< 0.05). Generalisation between Functional Groups {#s2e} ---------------------------------------- To analyse generalisation between groups, we took into account the responses to functional groups different from the conditioned one (see white bars in [Figure 3](#pbio-0030060-g003){ref-type="fig"}B). Bees showed heterogeneous patterns of generalisation (all vertical and horizontal comparisons in [Figure 3](#pbio-0030060-g003){ref-type="fig"}B were significant: χ^2^ \> 37.70, *df* = 3, *p* \< 0.001, in all eight cases). We found high between-group generalisation for primary and secondary alcohols: bees conditioned to secondary alcohols responded preferentially to primary alcohols, somewhat less to aldehydes, and even less to ketones (see [Figures 3](#pbio-0030060-g003){ref-type="fig"}A and [3](#pbio-0030060-g003){ref-type="fig"}B, second row). A similar but less obvious response gradation was found for bees conditioned to primary alcohols [Figures 3](#pbio-0030060-g003){ref-type="fig"}A and [3](#pbio-0030060-g003){ref-type="fig"}B, first row). In fact, the overall generalisation patterns were very similar for primary and secondary alcohols sharing the same chain length (see, for instance, the very close relationship between the two sets of blue \[primary alcohol\] and green curves \[secondary alcohols\] in [Figure 4](#pbio-0030060-g004){ref-type="fig"}A). As indicated before, bees conditioned to aldehydes generalised very little to odours belonging to other functional groups (see [Figure 3](#pbio-0030060-g003){ref-type="fig"}B, third row). Contrarily, bees conditioned to other functional groups highly generalised to aldehydes (see third column 'al\' in [Figure 3](#pbio-0030060-g003){ref-type="fig"}B). This shows that generalisation between aldehydes and odours belonging to other functional groups was asymmetrical. The topic of asymmetric generalisation will be considered below in more detail. Generalisation between Chain Lengths {#s2f} ------------------------------------ To analyse generalisation between chain lengths, we took into account the responses to chain lengths that differed from the conditioned one (see white bars in [Figure 4](#pbio-0030060-g004){ref-type="fig"}B). In general, responses to molecules with different chain lengths followed a clear decreasing gradient, depending on the difference in the number of carbon atoms between the molecules considered (see [Figure 4](#pbio-0030060-g004){ref-type="fig"}B; all horizontal and vertical comparisons were significant, χ^2^ \> 16.3, *df* = 3, *p* \< 0.001 in all eight cases). For instance, when conditioned to a C9 molecule (see [Figure 4](#pbio-0030060-g004){ref-type="fig"}B, fourth row), bees responded in 53%, 31%, and 23% of the cases to C8, C7, and C6 molecules, respectively, while they responded to C9 molecules in 67% of the cases. This gradient was also evident when generalisation took place between functional groups: for instance, after training with 2-nonanol (see [Figure 3](#pbio-0030060-g003){ref-type="fig"}A, second row), the response of bees to odours of different functional groups (solid lines in white boxes) always followed a similar decreasing tendency with the same (C9) or similar (C8) chain length on top. Asymmetry in Olfactory Generalisation {#s2g} ------------------------------------- As previously mentioned, some groups like aldehydes induced asymmetrical cross-generalisation (i.e., bees responded less to other functional groups after training for aldehydes than to aldehydes after training for other functional groups). We analysed this asymmetrical generalisation and built an asymmetry matrix ([Figure 5](#pbio-0030060-g005){ref-type="fig"}A). To this end, we calculated for each odour pair (A and B) the difference (in percentage) between generalisation from A to B and generalisation from B to A. Such differences were ranked in 10% categories from −55% to 55%. White boxes indicate no asymmetries. Blue shades in [Figure 5](#pbio-0030060-g005){ref-type="fig"}A indicate that cross-generalisation was biased towards odour A (i.e., conditioning to A resulted in lower generalisation to B while conditioning to B resulted in higher generalisation to A); red shades indicate that cross-generalisation was biased towards odour B (i.e., conditioning to A resulted in higher generalisation to B while conditioning to B resulted in lower generalisation to A). This representation showed that some odours induced generalisation while other odours diminished it. For instance, hexanal was well learnt but induced low generalisation to other odours, except to other aldehydes. On the other hand, bees conditioned to other odours very often generalised to hexanal. Thus, a clear blue row (or a red column) corresponds to hexanal in the asymmetry matrix. Conversely, 2-hexanone induced high generalisation to other odours but received few responses as a test odour. Thus a red row (or a blue column) corresponds to 2-hexanone in the asymmetry matrix. Most odours, however, showed little or no asymmetry. [Figure 5](#pbio-0030060-g005){ref-type="fig"}B presents the mean asymmetry found for each training odour. In six cases, the mean asymmetry deviated significantly from zero, which represents a theoretically perfect symmetry (*t*-test). Two odours (red bars) significantly induced generalisation (2-hexanone and 2-hexanol, *t*-test, *df* = 14, *p* \< 0.001 and *p* \< 0.01, respectively), while four odours (blue bars) diminished it significantly (hexanal, heptanal, and octanal, and 2-nonanone, *t*-test, *df* = 14, *p* \< 0.001 for the former and *p* \< 0.01 for the three latter odours). {#pbio-0030060-g005} Olfactory Space {#s2h} --------------- In order to define a putative olfactory space for the honeybee, we performed a principal component analysis (PCA) on our data to represent in a limited number of dimensions the relative relationships between odorants in a 16-dimension perceptual space ([Figure 6](#pbio-0030060-g006){ref-type="fig"}A). The first three factors represented 31%, 29%, and 15% of overall variance in the data (total of the first three factors: 75%). The analysis showed a clear organisation of odours depending on their chemical characteristics. First, chain length was very clearly represented by the first factor (see upper-right graph in [Figure 6](#pbio-0030060-g006){ref-type="fig"}A), from C6 to C9 molecules from the right to the left. On the other hand, the chemical group was mostly represented by factors 2 and 3. Whereas factor 2 separated mostly aldehydes from alcohols, with ketones falling between them, factor 3 segregated ketones from all other odours (lower-right graph, [Figure 6](#pbio-0030060-g006){ref-type="fig"}A). None of these factors separated primary and secondary alcohols. This analysis indicates that the chemical features of molecules (chain length and functional group), which are sometimes thought of as artificial perceptual (psychophysical) dimensions determined by experimenters \[[@pbio-0030060-b34]\] can be considered as true inner dimensions of the bees\' perceptual space. Cluster analyses performed on the data segregated odours mostly according to their chain length. In the first group ([Figure 6](#pbio-0030060-g006){ref-type="fig"}B, upper part), we found two subgroups, short-chain alcohols (C6 and C7, primary and secondary alcohols) and short-chain ketones (C6 to C8). On the other hand ([Figure 6](#pbio-0030060-g006){ref-type="fig"}B, lower part), three clear subgroups were formed: short-chain aldehydes (C6 and C7), long-chain alcohols (C8 and C9, primary and secondary alcohols), and a last group with long-chain aldehydes (C8 and C9) and 2-nonanone. Very similar results were obtained using Euclidian or city-block metrics. ![A Putative Honeybee Olfactory Space\ (A) Left: The olfactory space is defined on the basis of the three principal factors that accounted for 76% of overall data variance after a PCA performed to represent the relative relationships between odorants. Primary alcohols are indicated in blue, secondary alcohols in green, aldehydes in black, and ketones in red. Different chain-lengths are indicated as C6, C7, C8, and C9, which corresponds to their number of carbon atoms. For each functional group, arrows follow the increasing order of carbon-chain lengths. Right: Chain length was very clearly represented by factor 1. C6 to C9 molecules are ordered from right to left. The chemical group was mostly represented by factors 2 and 3. Whereas factor 2 separated mostly aldehydes from alcohols, with ketones falling between them, factor 3 separated ketones from all other odours. None of these three factors separated primary and secondary alcohols.\ (B) Euclidean cluster analysis. The analysis separated odours mostly according to their chain length. Linkage distance is correlated to odour distances in the whole 16-dimension space. The farther to the right two odours/odour groups are connected, the higher the perceptual distance between them (odour colour codes are the same as in \[A\]).](pbio.0030060.g006){#pbio-0030060-g006} Correlation between Optophysiological and Behavioural Measures of Odour Similarity {#s2i} ---------------------------------------------------------------------------------- We asked whether optophysiological measures of odour similarity, obtained using calcium imaging techniques at the level of the honeybee AL \[[@pbio-0030060-b22],[@pbio-0030060-b23],[@pbio-0030060-b24],[@pbio-0030060-b35]\], correspond to perceptual odour similarity measures as defined in our putative honeybee olfactory space. We thus calculated the Euclidian distance between odour representations in our 16-dimension "behavioural" space for all odour pairs (120 pairs). We then calculated distances between odours in optical imaging experiments, using the odour maps by Sachse et al. \[[@pbio-0030060-b23]\]. A correlation analysis was performed between both datasets. This analysis was possible because both the study by Sachse et al.\[[@pbio-0030060-b23]\] and our study used the same set of odours delivered under the same conditions. [Figure 7](#pbio-0030060-g007){ref-type="fig"}A presents the correlation obtained, including all 120 odour pairs. Both sets of data were highly significantly correlated (*r* = 0.54, *t* ~118~ = 7.43, *p* \< 2.10^--10^), a result that shows that odours, which were found to be physiologically similar in the optical imaging study, were also evaluated as similar in behavioural terms. Note, however, that data points cluster quite broadly around the main trend line, showing that many exceptions were found. In order to use a more exact measure of physiological odour similarity, we used the correlation results between primary and secondary alcohol maps provided by Sachse et al. \[[@pbio-0030060-b23]\]. By correlating this more exact value of physiological similarity with our behavioural data, we also found a highly significant relationship between physiological and behavioural data ([Figure 7](#pbio-0030060-g007){ref-type="fig"}B; *r* = 0.82, *t* ~26~ = 7.83, *p* \< 7.10^--8^). The correlation coefficient achieved with this second method was significantly higher than that achieved with the first method (*Z* = 2.52, *p* \< 0.05). A better fit between the two datasets was thus found, although outliers were still present in the data. These two analyses show that optophysiological and behavioural measures of odour similarity correlate well using the methods described here. Thus, in the case of the honeybee, olfactory neural activity corresponds to olfactory perception. ![Correspondence between Perceptual and Physiological Odour Similarity\ (A) Correlation between optophysiological measures of odour similarity (carried out using calcium imaging recordings \[[@pbio-0030060-b23]\]) and our behavioural measures of odour similarity. Euclidian distance between odour representations in our 16-dimension "behavioural" space for all odour pairs (120 pairs, x axes) and distances between odours in optical imaging experiments, using the odour category maps displayed by Sachse et al. \[[@pbio-0030060-b23]\] (also 120 pairs, y axes) were calculated. This correlation, including all 120 odour pairs, was highly significant (*r* = 0.54, *p* \< 0.001). Odours found to be similar in the optical imaging study were also similar in the behaviour. Data points cluster quite broadly around the main trend line, showing that many exceptions were found.\ (B) Correlation between measures of optophysiological similarity carried out using the optical imaging technique \[[@pbio-0030060-b23]\] and our behavioural measure of odour similarity. Using the exact data given for primary and secondary alcohols \[[@pbio-0030060-b23]\], a much better correlation between the two datasets was achieved than in (A) (*r* = 0.82, *p* \< 0.001), although outliers were still found in the data.](pbio.0030060.g007){#pbio-0030060-g007} Discussion {#s3} ========== In the present work, we have studied perceptual similarity among odorants in the honeybee, using an appetitive-conditioning paradigm, the olfactory conditioning of the PER \[[@pbio-0030060-b17],[@pbio-0030060-b18]\]. We showed that all odorants presented could be learned, although acquisition was lower for short-chain ketones. Generalisation varied, depending both on the functional group and on the carbon-chain length of odours trained. Generalisation was very high among primary and secondary alcohols, being high from ketones to alcohols and aldehydes and low from aldehydes to all other tested odours; thus, in some cases, cross-generalisation between odorants was asymmetric. Some odours, like short-chain ketones or aldehydes, induced more asymmetries than other odours. Higher generalisation was found between long-chain than between short-chain molecules. Functional group and carbon-chain length constitute orthogonal inner dimensions of a putative olfactory space of honeybees. Perceptual distances in such a space correlate well with physiological distances determined from optophysiological recordings performed at the level of the primary olfactory centre, the AL \[[@pbio-0030060-b23]\] such that olfactory neural activity corresponds to olfactory perception. Previous studies have attempted to describe olfactory generalisation in honeybees and to study structure--activity relationships \[[@pbio-0030060-b19],[@pbio-0030060-b20],[@pbio-0030060-b36],[@pbio-0030060-b37],[@pbio-0030060-b38]\]. These studies generally supported the view that generalisation mainly happens when odours belong to the same chemical group. Moreover, they also suggested that the rules underlying olfactory learning and perception of different chemical classes \[[@pbio-0030060-b20]\] or of particular odorants (e.g., citral \[[@pbio-0030060-b20],[@pbio-0030060-b37]\]) may vary. However, these studies used differential training, thus inducing several generalisation gradients (excitatory and inhibitory) that make the interpretation of generalisation responses difficult \[[@pbio-0030060-b21],[@pbio-0030060-b36]\]. Furthermore, these studies were carried out on a rather discrete number of odour pairs \[[@pbio-0030060-b37]\], did not detail the results obtained with individual odour combinations \[[@pbio-0030060-b20]\], or used a very reduced number of bees per conditioned odour (\[[@pbio-0030060-b21]\]; two bees per odorant).Thus, the present study is the first one to provide (i) generalisation data based on absolute conditioning (i.e., only one odour conditioned at a time), (ii) a systematical test of all odour combinations, (iii) robust sample sizes for each experimental situation, and (iv) important generalisation gradients. These are in our view crucial prerequisites to describe odour perception and similarity in a precise way. Chemical Group and Chain Length {#s3a} ------------------------------- Several studies in other species have shown the importance of functional group and carbon-chain length of the odour molecules for behavioural responses to odours. Differences in the response between molecules of diverse aliphatic and aromatic homologue odour classes (i.e., differing in functional group, chain length, and overall molecule form) were investigated in moths \[[@pbio-0030060-b39],[@pbio-0030060-b40]\], cockroaches \[[@pbio-0030060-b41]\], rats \[[@pbio-0030060-b42]\], squirrel monkeys \[[@pbio-0030060-b04],[@pbio-0030060-b43]\] and humans \[[@pbio-0030060-b38],[@pbio-0030060-b44],[@pbio-0030060-b45]\]. These studies show that both functional group and chain length affect the perceived quality of an odorant. Concerning chain length, the greater the difference in the number of carbons between odours, the easier the discrimination and the lower the generalisation (\[[@pbio-0030060-b21],[@pbio-0030060-b40],[@pbio-0030060-b42],[@pbio-0030060-b44]\] and present study). In our study, both chemical group and chain length of odour molecules determined the bees\' generalisation responses. Bees mostly generalised to other odours when these shared the same functional group. This effect was observed for all functional groups (see [Figure 3](#pbio-0030060-g003){ref-type="fig"}B) but was strongest for aldehydes. Other studies have found that aldehydes induced high within-group generalisation \[[@pbio-0030060-b20],[@pbio-0030060-b21],[@pbio-0030060-b36]\]. Thus, aldehydes may represent a behaviourally relevant chemical class for honeybees. Between-functional group generalisation depended on the functional group considered. It was high between primary and secondary alcohols, which appear therefore perceptually similar to the bees, and low between other chemical groups. Bees clearly generalised between odours that shared the same chain length. Increasing chain length promoted generalisation. Moreover, generalisation to other chain lengths decreased if the difference in the number of carbons between odours increased. This suggests a perceptual continuum between different chain lengths (but see below). Thus, the chemical structure of the odorants is critical for determining the amount of generalisation. A Putative Olfactory Space for the Honeybee {#s3b} ------------------------------------------- We found that the two controlled physical characteristics of odour molecules used in this study, functional group and chain length, correspond to internal dimensions in the bees\' olfactory perceptual space such as the three most important factors extracted in our PCA analysis, one mainly represented chain length and the other two were mostly influenced by functional group. Cluster analyses allowed separating odours in clusters according to their functional groups and their chain length. Interestingly, C6 and C7 molecules and C8 and C9 molecules were mainly grouped together, so that, for instance, all short-chain primary and secondary alcohols were grouped on one side, and all long-chain alcohols on the other side. The same happened for aldehydes, and in a different way for ketones (C9 separated from the rest). This discrepancy suggests that, although chain length appears mostly as a perceptual continuum in the PCA analysis, there may be a perceptual "jump" between short-chain and long-chain molecules. Neural Bases of Odour Perception {#s3c} -------------------------------- Both in vertebrates and in invertebrates, studies quantifying the neural responses to structurally similar odours in the first relay of the olfactory pathway have been performed (olfactory bulb: e.g., \[[@pbio-0030060-b46],[@pbio-0030060-b47],[@pbio-0030060-b48],[@pbio-0030060-b49]\]; AL: \[[@pbio-0030060-b23],[@pbio-0030060-b50]\]). These studies show that activity patterns are more similar when the difference in the number of carbons between molecules is small. It was hypothesised that such a physiological similarity is the basis for olfactory discrimination and generalisation as measured behaviourally. This has indeed been reported for mucosal activity in mice \[[@pbio-0030060-b51]\], electrical mitral cell activity \[[@pbio-0030060-b42]\], and/or radiolabelled 2-deoxyglucose uptake in the rat olfactory bulb \[[@pbio-0030060-b32]\]. Also, in *Manduca sexta,* qualitative similarities were observed between the degree of behavioural generalisation according to chain length \[[@pbio-0030060-b40]\] and the degree of overlap between electrophysiological temporal patterns of activity across AL neurons \[[@pbio-0030060-b50]\]. Several correspondences, but also discrepancies, can be found between our behavioural results and the physiological results obtained at the level of the bee AL \[[@pbio-0030060-b23]\]. First, within the regions of the AL accessible to optical imaging (about 25% of the glomeruli), patterns of glomerular activity for different odours are highly dependent on chain length, but much less so on chemical group. Thus, most active glomeruli respond to several functional groups as long as the chain length corresponds, but respond differentially to different chain lengths. Glomeruli T1--28 and T1--52 are specialised in short-chain molecules (respectively C5--C7 and C6--C7), whilst glomeruli T1--33 and T1--17 are specialised in long-chain molecules (respectively C7--C9 and C8--C9). These glomeruli also respond to most functional groups but in a graded way. For instance, glomerulus T1--17 responds more to alcohols in the intermediate range than to aldehydes or ketones, whereas T1--52 generally responds more to ketones in the short range, more to aldehydes in the long range, and overall little to alcohols. No individual glomerulus was found that responds specifically to a chemical group. However, it should be kept in mind that some regions of the ALs are not yet accessible to calcium imaging techniques (about 75% of the lobe; see below). Thus, a possible explanation is that glomeruli responding to specific chemical groups (or with responses more dependent on chemical groups than on chain length) were not imaged. Second, primary and secondary alcohols induce extremely similar activation patterns in the AL, but subtle differences could be found, so that for a given chain length, the representation of a secondary alcohol was between that of the primary alcohol of the same chain length and that with one less carbon atom (see [Figure 6](#pbio-0030060-g006){ref-type="fig"}B in Sachse et al. \[[@pbio-0030060-b23]\]). We found a similar arrangement of alcohol representations, with primary and secondary alcohols alternating on a common axis (see [Figure 6](#pbio-0030060-g006){ref-type="fig"}A). Third, optical imaging data showed that higher chain lengths support more similarity between patterns (see [Figure 6](#pbio-0030060-g006){ref-type="fig"}C in Sachse et al. \[[@pbio-0030060-b23]\]). Our finding that longer chain lengths induce more generalisation agrees with the imaging data. These last two points suggest that the general rules governing odour similarity at the neural and the behavioural level are similar. The Correspondence between Perceptual and Physiological Odour Similarity {#s3d} ------------------------------------------------------------------------ We aimed at comparing behavioural and physiological data in a more precise way, using correlation analyses between our behavioural similarity matrix, in which distances between two odour points represent psychological distances between stimuli, and a physiological similarity matrix obtained from optophysiological recordings of glomerular activation patterns \[[@pbio-0030060-b23]\]. Comparing distances between odours in these two matrixes resulted in a good correlation. This means that glomerular activity patterns recorded in the brain could predict behavioural responses and vice versa. The optophysiological dataset of Sachse et al. \[[@pbio-0030060-b23]\] has nevertheless some limitations with respect to the objectives of our work: (i) bath application measurements of AL activity using calcium green as a dye \[[@pbio-0030060-b23]\] record the combined activity of several neuronal populations of the AL, among which primary-afferent activity seems to have the most important contribution \[[@pbio-0030060-b52]\]; (ii) such measurements survey only the dorsal part of the AL, which constitutes 25% of the neuropile studied; and (iii) learning alters odour representations in the AL \[[@pbio-0030060-b35],[@pbio-0030060-b53],[@pbio-0030060-b54]\] such that there could be a mismatch between our data collected after olfactory conditioning and the dataset of Sachse et al. \[[@pbio-0030060-b23]\], which was obtained from naive bees. With respect to the first point, it could be argued that the AL circuitry transforms the primary-afferent representations of odours \[[@pbio-0030060-b25]\] such that recordings where primary-afferent receptor activity is predominant are not very useful for evaluating optophysiological similarity. However the very fact that we found a significant correlation between our behavioural data and the imaging data by Sachse et al. \[[@pbio-0030060-b23]\], strongly suggests that the perceptual quality of odorants mostly appears at the peripheral level. Clearly, this correlation was not perfect, and odour quality is most probably refined by further processing within the AL, and/or at higher stages of the olfactory pathway, such as in the mushroom bodies or the lateral protocerebrum. In honeybees, new methods have been developed, which allow recording selectively the activity of the efferent PNs \[[@pbio-0030060-b25]\]. However, the two studies published using this method \[[@pbio-0030060-b25],[@pbio-0030060-b26]\] do not provide an extensive odorant matrix as that provided by Sachse et al. \[[@pbio-0030060-b23]\]. In this sense the study on which we based our correlation analysis is certainly the only one of its kind published to date. However, in the future, a careful comparison of our behavioural data with both bath-applied imaging data emphasising receptor neuron input (as done here) and selective imaging of PNs would be extremely helpful in understanding to what extent AL processing shapes odour perceptual quality. With respect to the second point, calcium imaging recordings of AL activity are certainly limited to the dorsal part of the AL, which is the region accessible when the head capsule is opened in order to expose the brain for recordings. This is an inherent limitation of the method that the use of two-photon microscopy during calcium imaging measurements will soon allow us to overcome, as shown already by recordings obtained in the fruit fly Drosophila melanogaster \[[@pbio-0030060-b55]\]. Finally, with respect to the third point, it is known that learning alters odour representations in the AL, when bees are trained in a differential conditioning procedure, with one odour rewarded and another odour unrewarded \[[@pbio-0030060-b53]\]. This is not the conditioning procedure used in our work, which was absolute (only one odour rewarded at a time). In the bee, changes in the olfactory code due to absolute conditioning seem to be difficult to detect (C. G. Galizia, personal communication), such that this point may not be so critical for our correlation analysis. In any case, if there are changes in odour representations due to conditioning, recording glomerular activity patterns after conditioning would only improve our correlation analyses. Generalisation Asymmetries between Odours {#s3e} ----------------------------------------- We have found a number of asymmetries in olfactory cross-generalisation, with bees responding more to odour B after learning odour A than in the reverse situation. Previous studies have observed such a phenomenon, but it was mostly related to olfactory compounds with pheromonal value (aggregation pheromone citral \[[@pbio-0030060-b20],[@pbio-0030060-b37]\] and alarm pheromones 2-heptanone and isoamyl acetate \[[@pbio-0030060-b56]\]). In the present study, we found that six out of the 16 odours used induced significant generalisation asymmetries over the whole matrix; none of these six odours was related to any known pheromone (see [Table 1](#pbio-0030060-t001){ref-type="table"}). Generalisation asymmetries seem to be a general feature of honeybee olfaction. ::: {#pbio-0030060-t001 .table-wrap} ::: {.caption} ###### Chemical and Biological Characteristics of the Odours Used :::  The odours were listed by functional groups (primary alcohols, secondary alcohols, aldehydes, and ketones) and purity. Odour vapour pressure values (VP), pheromone characteristics and occurrence in floral scents (after Knudsen et al. \[[@pbio-0030060-b66]\]) are also given ^a^Notation: \*1, releases altering at hive entrance and stinging, repels clustering bees, inhibits scenting, repels foragers (sting chamber); \*2, releases altering at hive entrance, inhibits foraging activity, repels foragers (sting chamber); \*3, repels at hive entrance, releases stinging, encourages foraging activity (sting chamber); \*4, releases stinging, inhibits foraging activity, repels foragers (mandibular glands) ::: Odour concentration can affect stimulus salience. In our work, generalisation asymmetries could not be directly explained by differences in odour concentration (through differences in vapour pressure), because, for instance, the two odours with the highest vapour pressure in our sample (2-hexanone and hexanal) produced totally opposite results: 2-hexanone induced important generalisation, while hexanal strongly reduced generalisation. Also, although we used 16 different odours with a range of different vapour pressures, we found that acquisition was very similar for most odours, except for the short-chain ketones, which were less easily learned. This suggests that almost all odours used had a good salience for bees. Wright and Smith \[[@pbio-0030060-b57]\] studied the effect of odour concentration in generalisation in honeybees. They found that discrimination increased with concentration for structurally dissimilar odours but not for similar odours. Further experiments using odorants at different concentrations should be carried out to determine the effect of odour concentration on generalisation asymmetries. Generalisation asymmetries could be due to innate or experience-dependent differences in the salience of odours for honeybees, such that more salient odours would induce higher generalisation than less salient odours. This interpretation implies that most aldehydes (hexanal, heptanal, and octanal) are highly salient odours for honeybees, because aldehydes showed a clear "functional group" effect, which could reveal a certain bias of the olfactory system towards these odours. Ketones, on the other hand, showed a heterogeneous effect, as 2-hexanone seemed to have a low salience (it was not well learnt) and induced a high generalisation to other odours, while 2-nonanone consistently reduced generalisation to other odours. In the group of alcohols, only 2-hexanol induced generalisation to other odours. Therefore, only aldehydes showed a clear group effect on generalisation asymmetry. This effect could be due to innate odour preferences \[[@pbio-0030060-b58],[@pbio-0030060-b59]\] or to previous odour exposure within the hive \[[@pbio-0030060-b60],[@pbio-0030060-b61]\]. Innate odour preferences could be related to natural, floral odours that were more consistently associated with food resources \[[@pbio-0030060-b20],[@pbio-0030060-b62]\]. It is thus important to investigate whether or not such ecological trends exist in the natural flora associated with the honeybee and whether or not other bee species also present such clear biases, in particular towards aldehydes. Conversely, asymmetries could be the result of the conditioning procedure. This would be the case if conditioning modifies odour representation in an asymmetric way. Indeed, experience-induced modifications of odour representations have been found at the level of the honeybee AL. Thus, odour-evoked calcium signals in the AL can be modified by elemental \[[@pbio-0030060-b53]\] and nonelemental olfactory learning paradigms \[[@pbio-0030060-b35]\] such that the representations of odours that have to be discriminated become more distinct and uncorrelated as a result of learning. In the fruit fly D. melanogaster, new glomeruli become active after olfactory learning \[[@pbio-0030060-b54]\], while in the moth M. sexta new neuronal units in the AL are recruited after olfactory learning \[[@pbio-0030060-b63]\]. These elements suggest that modifications of odour representation after learning two different odours could indeed be asymmetrical: if, for instance, the neuronal representation of A after conditioning becomes A′, which is slightly farther away from B than A in the bee\'s olfactory space, and if the perceptual representation of B becomes B′ after conditioning, which is closer to A than B, then bees would show less generalisation in behavioural tests from A to B than from B to A. On the level of the AL network, glomeruli are connected via lateral inhibitory interneurons \[[@pbio-0030060-b25],[@pbio-0030060-b64],[@pbio-0030060-b65]\]. Due to this, glomerular activation by an odour A will transiently inactivate parts of the network and possibly parts encoding a subsequent odour B. Optical imaging experiments have shown that inhibition between glomeruli may be asymmetric \[[@pbio-0030060-b25]\]. In our case, glomeruli activated by odour A may inhibit glomeruli coding for odour B, while glomeruli coding for odour B may not inhibit those coding for odour A. In this hypothesis, asymmetric cross-generalisation could reflect a sensory phenomenon. Nevertheless, we believe that inhibitions at the level of the AL are rather short-lived such that a purely sensory priming effect seems improbable. If, however, the strength of lateral inhibitions between glomeruli can be modified by learning as proposed by Linster and Smith \[[@pbio-0030060-b65]\], then asymmetrical generalisation would come from the fact that inhibitory lateral connections are modified. In order to determine the physiological mechanisms underlying asymmetrical cross-generalisation and the possible role of AL networks in it, future work will aim at visualising the evolution of glomerular activity patterns during and after olfactory conditioning with odours that showed asymmetries in our study. Conclusion {#s3f} ---------- We have shown that the two odorant physical dimensions that varied in our study, functional group and chain-length, correspond to internal dimensions of the bees\' olfactory space. Generalisation was mainly due to these two characteristics with generalisation within functional group being more important. Such generalisation was particularly high for aldehydes, a fact that suggests that these odours may have an intrinsic value for bees. Generalisation between functional groups was mostly found between primary and secondary alcohols. Furthermore, a gradient in generalisation was found with respect to chain length. Asymmetric cross-generalisation was found in the case of certain odorants. Such asymmetries were neither strictly linked to chain length nor to functional group, but depended on particular odorants. The 16 odours used in our work represent a small part of the odorants that bees may encounter in nature (see Knudsen et al. \[[@pbio-0030060-b66]\]). For a complete description of the bees\' olfactory perceptual space, more odours having other molecular features have to be studied. New dimensions in the bees\' perceptual space could then be found. Finally, and most important, the perceptual distance between odours can be predicted on the basis of the differences in the patterns of glomerular activation in the first relay of the olfactory pathway: the AL, and vice versa. This emphasises the relevance of studying activity patterns in the brain in imaging studies and trying to relate them to perceptual tasks. Our work shows that this objective, which is at the core of cognitive neurosciences, can be achieved using an invertebrate model such as the honeybee. Materials and Methods {#s4} ===================== {#s4a} ### Insects {#s4a1} Every experimental day, honeybees were captured at the entrance of an outdoor hive and were cooled on ice for 5 min until they stopped moving. Then they were harnessed in small metal tubes in such a way that only the head protruded. The mouthparts and the antennae could move freely. Harnessed bees were left for 3 h in a resting room without disturbance. Fifteen minutes before starting the experiments, each subject was checked for intact PER by lightly touching one antenna with a toothpick imbibed with 50% (w/w) sucrose solution without subsequent feeding. Extension of the proboscis beyond the virtual line between the open mandibles was counted as PER. Animals that did not show the reflex were not used in the experiments. ### Stimulation apparatus {#s4a2} The odours were delivered by an odour cannon, which allowed the presentation of up to seven different odours, and a clean airstream \[[@pbio-0030060-b67]\]. Each odour was applied to a filter paper placed within a syringe (see below) that was connected to the cannon. An airstream was produced by an air pump (Rena Air 400, Annecy, France) and directed to the relevant syringes with electronic valves (Lee Company, Voisins-le-Bretonneux, France) controlled by the experimenter via a computer. In the absence of odour stimulation, the airstream passed through a syringe containing a clean filter paper piece (clean airstream). During odour stimulation, the airstream was directed to a syringe containing a filter paper loaded with odour. After a 4-s stimulation, the airstream was redirected to the odourless syringe until the next stimulation. ### Stimuli {#s4a3} Sixteen odours (Sigma Aldrich, Deisenhofen, Germany) were used in our work as CS and test stimuli (see [Table 1](#pbio-0030060-t001){ref-type="table"}). Racemic mixtures were used in the case of molecules that had chiral carbons. These odours are present in flowers and some in pheromones (see [Table 1](#pbio-0030060-t001){ref-type="table"}). Pure odorants (4 μl) were applied to 1-cm^2^ filter paper pieces, which were transferred to 1-ml syringes, cut to 0.7 ml to make them fit into the odour cannon. Fifty percent sugar solution was used throughout as US. ### Experimental design {#s4a4} Our work was designed to obtain a generalisation matrix with 16 different odours. Ideally, after conditioning each of the 16 odours as CS, the response to each odour (including the CS) should be measured (i.e., 16 × 16 = 256 cells). However, testing 16 odours implies presenting them without reward, a situation that may result in extinction of the learned response due to the repeated unrewarded odour presentations. Preliminary experiments were performed in which four groups of 180 bees were trained along three trials to 1-hexanol, 2-octanol, linalool, and limonene, respectively. Training was followed by tests with the four different odours, including the conditioned one. These experiments showed that after three conditioning trials, the response of the bees to the CS in the four tests remained at the same level, independently of the order of occurrence of the CS such that it was not influenced by extinction. We thus kept this protocol for the 16 × 16 matrix. Each of the 2,048 bees used in this study was thus subjected to three conditioning trials with their respective CS, and to four test trials, each with a different odour chosen among the 16 possible odours. Intertrial intervals of 10 min were used throughout. A randomisation schedule (detailed below) was developed for the test phase to reduce any possible day- and odour-combination effects. ### Conditioning trials {#s4a5} One bee at a time was placed into the conditioning setup. The total duration of each trial was 37.5 s After 15 s of familiarisation to the experimental context, the CS was presented to the bee for 4 s. Three sec after onset of the CS, the antennae were stimulated with the US, leading to a proboscis extension. The bee was allowed to feed for 3 s. Stimulus overlap was 1 s (interstimulus interval, 3 s). The bee was left in the conditioning place for 17.5 s and then removed. ### Test trials {#s4a6} The procedure was similar to that for conditioning trials but no US was given after odour delivery. After the four test trials, PER to the US was checked once again. Animals unable to show PER at this point were not considered for the analyses. Overall, less than 2% of the bees died during the experiment, and less than 1% of the survivors showed no US reaction at the end of the tests. ### Randomisation schedule {#s4a7} On each day, two to three experimenters worked in parallel, each training 16 bees at a time. In the training phase, the 16 bees were divided into four groups of four bees, and each group was trained to one of the 16 different odours. In the test phase, four out of 16 odours were presented to each of the 16 bees. The combination of four odours tested together changed in each experiment, so that any effect of having particular odours in the same test combination was suppressed. The whole experiment was planned in such a way that in any of our experimental groups, two given odours appeared at least once, but a maximum of three times together in a test sequence. This was possible by carefully picking out eight of the 16! (2.1 × 10^13^) possible experimental plans. Additionally, within each group, the testing order for the four test odours was determined randomly. ### Data analysis and statistics {#s4a8} During the experiments, we recorded the response to the presented odour, that is, whether bees extended their proboscis after the onset of the odour and before the presentation of the sucrose solution in the case of reinforced trials, such that the anticipatory response recorded was due to the odour and not to the US. Multiple responses during a CS were counted as a single PER. The percentages of PER recorded during acquisition were used to plot acquisition curves (see [Figure 1](#pbio-0030060-g001){ref-type="fig"}). To test whether bees learnt the different odours in a similar way, ANOVAs for repeated measurements were used both for between-group and for within-group comparisons. Monte Carlo studies have shown that it is permissible to use ANOVA on dichotomous data only under controlled conditions \[[@pbio-0030060-b68]\], which are met by the experiments reported in this study: equal cell frequencies and at least 40 *df* of the error term. The α level was set to 0.05 (two-tailed). To ensure that we analysed a true generalisation response in the tests, and hence built a true generalisation matrix, we kept only those bees which had actually learnt the CS (71% of the bees used in this work). We therefore performed new analyses that only included those bees that responded to the CS before the presentation of the US in the third conditioning trial. A lack of response to an odour in the tests could be due either to the fact that the bees had not made any association between CS and US or because their motivational level was low. For all odours tested, we observed that responses to the CS in the third conditioning trial were equivalent to responses to the CS in the tests (McNemar test; see Results). We represented the responses of the selected bees to the test odours (see [Figure 2](#pbio-0030060-g002){ref-type="fig"}). As the numbers of bees were now heterogeneous in the different groups, we could not use ANOVAs to analyse the responses in the tests (see above). We thus used χ^2^ tests for all further between-group comparisons. In the case of multiple two-by-two comparisons, the significance threshold was corrected using the Dunn--Sidak correction \[α′ = 1 − (1 − α)^1/k^ where *k* is the number of two-by-two comparisons in which each dataset is used\] in order to reduce the type I errors. Alpha values between α′ and 0.05 were considered as near significant. ### Olfactory space {#s4a9} To observe the relationships between odours in a reduced number of dimensions, we performed a PCA, which identified orthogonal axes (factors) of maximum variance in the data, and thus projected the data into a lower-dimensionality space formed of a subset of the highest-variance components. We calculated the three factors, which accounted for most of the observed variance. Calculating distances between odours in the resulting putative olfactory space allowed the evaluation of their perceptual similarity, not only based on direct generalisation between these odours (i.e., generalisation from odour A to odour B and vice versa), but also including responses to these odours after conditioning to other odours (e.g., C, D, E, etc.). We performed cluster analyses to group odours, according to their respective distance in the olfactory space, using both Euclidian and city-block metrics, with Ward\'s classification method. Both metrics gave very similar results, so we later used only Euclidian metrics. Euclidian (i.e., direct) distances in the 16-dimensional space are defined as with *i* and *j* indicating odours, *p* the number of dimensions---that is, conditioning groups---and *X~ik~* the response of bees to odour *i* after conditioning to odour *k.* These distances were used in correlation analyses with optical imaging data (see below). ### Correlation analysis between perceptual and optophysiological similarity measures {#s4a10} We studied whether or not physiological similarity between odours as determined by optical imaging studies of AL activity \[[@pbio-0030060-b22],[@pbio-0030060-b23],[@pbio-0030060-b35]\] actually reflects perceptual odour similarity for the bees. To this end, we performed correlation analyses between published optical imaging data that were obtained using the same set of odours as in our work \[[@pbio-0030060-b23]\] and our behavioural data. We used two sets of physiological data. First, to perform such a correlation on the whole dataset (including all 16 odours), we transcribed the activation maps presented by Sachse et al. \[[@pbio-0030060-b23]\] (see [Figure 7](#pbio-0030060-g007){ref-type="fig"}) into activation levels for each glomerulus from zero to three, according to the following signal scale: dark blue (0%--20%) and light blue (\>20%--40% activity), zero; green (\>40%--60% activity), one; yellow (\>60%--80% activity), two; and red (\>80% activity), three. As the activity under 40% was less accurately separated from noise, activation levels between 0% and 40% were ranked as 0. Scaling the physiological data in this way instead of using the original imaging activation data, gave a good overview of physiological similarity between odours for imaging data ([see Results](#s2){ref-type="sec"}). To provide a more precise correlation analysis between behavioural and imaging data, albeit on a more limited odour dataset (eight odours), we used exact correlation data (\[[@pbio-0030060-b23]\], [Table 1](#pbio-0030060-t001){ref-type="table"}). Each correlation value C, as calculated by Sachse et al. \[[@pbio-0030060-b23]\] between activity patterns for all pairs of primary and secondary alcohols, was converted into physiological distances by the operation 100 − C. All linear correlations were assessed by calculating Pearson\'s *r,* and using Student\'s *t*-test. Comparison between correlation coefficients obtained with the two methods was carried out statistically using a *Z* test as in \[[@pbio-0030060-b69]\]. We thank Nina Giotto for her help with the experiments and Giovanni Galizia, Monique Gauthier, Christiane Linster, and three anonymous referees for useful comments on previous versions of this work. M. Giurfa acknowledges the support of the Human Frontier Science Program (Young Investigator Award), the Fondation pour la Recherche Médicale, the Action Concertée Incitative Neurosciences Computationnelles (French Research Ministry), the Region Midi-Pyrenees, and the Institut Universitaire de France. J. C. Sandoz was supported by the Human Frontier Science Program, the Fyssen Foundation, and the Centre National de la Recherche Scientifique. F. Guerrieri was partially supported by the Programa "José Estenssoro," Fundación YPF (Argentina), and M. Schubert was supported by the Graduiertenkolleg 120 "Signal Cascades in Living Systems" (Germany). M. Schubert thanks Jochen Erber for support. **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** MG and JCS conceived and designed the experiments. FG and MS performed the experiments. FG, MS, and JCS analyzed the data. MG, FG, MS, and JCS wrote the paper. Citation: Guerrieri F, Schubert M, Sandoz J-C, Giurfa M (2005) Perceptual and neural olfactory similarity in honeybees. PLoS Biol 3(4): e60. AL : antennal lobe ANOVA : analysis of variance CS : conditioned stimulus PCA : principal component analysis PER : proboscis extension reflex PNs : projection neurons US : unconditioned stimulus

Introduction {#s1} ============ MicroRNAs (miRNAs) are small non-coding RNAs that serve as post-transcriptional regulators of gene expression in plants and animals. They act by binding to complementary sites on target mRNAs to induce cleavage or repression of productive translation (reviewed in \[[@pbio-0030085-b01],[@pbio-0030085-b02],[@pbio-0030085-b03],[@pbio-0030085-b04]\]). The importance of miRNAs for development is highlighted by the fact that they comprise approximately 1% of genes in animals, and are often highly conserved across a wide range of species (e.g., \[[@pbio-0030085-b05],[@pbio-0030085-b06],[@pbio-0030085-b07]\]). Further, mutations in proteins required for miRNA function or biogenesis impair animal development \[[@pbio-0030085-b08],[@pbio-0030085-b09],[@pbio-0030085-b10],[@pbio-0030085-b11],[@pbio-0030085-b12],[@pbio-0030085-b13],[@pbio-0030085-b14],[@pbio-0030085-b15]\]. To date, functions have been assigned to only a few of the hundreds of animal miRNA genes. Mutant phenotypes in nematodes and flies led to the discovery that the *lin-4* and *let-7* miRNAs control developmental timing \[[@pbio-0030085-b16],[@pbio-0030085-b17]\], that *lsy-6* miRNA regulates left--right asymmetry in the nervous system \[[@pbio-0030085-b18]\], that *bantam* miRNA controls tissue growth \[[@pbio-0030085-b19]\], and that *bantam* and *miR-14* control apoptosis \[[@pbio-0030085-b19],[@pbio-0030085-b20]\]. Mouse *miR-181* is preferentially expressed in bone marrow and was shown to be involved in hematopoietic differentiation \[[@pbio-0030085-b21]\]. Recently, mouse *miR-375* was found to be a pancreatic-islet-specific miRNA that regulates insulin secretion \[[@pbio-0030085-b22]\]. Prediction of miRNA targets provides an alternative approach to assign biological functions. This has been very effective in plants, where miRNA and target mRNA are often nearly perfectly complementary \[[@pbio-0030085-b23],[@pbio-0030085-b24],[@pbio-0030085-b25]\]. In animals, functional duplexes can be more variable in structure: they contain only short complementary sequence stretches, interrupted by gaps and mismatches. To date, specific rules for functional miRNA--target pairing that capture all known functional targets have not been devised. This has created problems for search strategies, which apply different assumptions about how to best identify functional sites. As a result, the number of predicted targets varies considerably with only limited overlap in the top-ranking targets, indicating that these approaches might only capture subsets of real targets and/or may include a high number of background matches (\[[@pbio-0030085-b19],[@pbio-0030085-b26],[@pbio-0030085-b27],[@pbio-0030085-b28],[@pbio-0030085-b29],[@pbio-0030085-b30]\]; reviewed by \[[@pbio-0030085-b31]\]). Nonetheless, a number of predicted targets have proven to be functional when subjected to experimental tests \[[@pbio-0030085-b19],[@pbio-0030085-b26],[@pbio-0030085-b27],[@pbio-0030085-b29]\]. A better understanding of the pairing requirements between miRNA and target would clearly improve predictions of miRNA targets in animals. It is known that defined *cis*-regulatory elements in *Drosophila* 3′ UTRs are complementary to the 5′ ends of certain miRNAs \[[@pbio-0030085-b32]\]. The importance of the miRNA 5′ end has also emerged from the pairing characteristics and evolutionary conservation of known target sites \[[@pbio-0030085-b26]\], and from the observation of a non-random statistical signal specific to the 5′ end in genome-wide target predictions \[[@pbio-0030085-b27]\]. Tissue culture experiments have also underscored the importance of 5′ pairing and have provided some specific insights into the general structural requirements \[[@pbio-0030085-b29],[@pbio-0030085-b33],[@pbio-0030085-b34]\], though different studies have conflicted to some degree with each other, and with known target sites (reviewed in \[[@pbio-0030085-b31]\]). To date, no specific role has been ascribed to the 3′ end of miRNAs, despite the fact that miRNAs tend to be conserved over their full length. Here, we systematically evaluate the minimal requirements for a functional miRNA--target duplex in vivo. These experiments have allowed us to identify two broad categories of miRNA target sites. Targets in the first category, "5′ dominant" sites, base-pair well to the 5′ end of the miRNA. Although there is a continuum of 3′ pairing quality within this class, it is useful to distinguish two subtypes: "canonical" sites, which pair well at both the 5′ and 3′ ends, and "seed" sites, which require little or no 3′ pairing support. Targets in the second category, "3′ compensatory" sites, have weak 5′ base-pairing and depend on strong compensatory pairing to the 3′ end of the miRNA. We present evidence that all of these site types are used to mediate regulation by miRNAs and show that the 3′ compensatory class of target sites is used to discriminate among individual members of miRNA families in vivo. A genome-wide statistical analysis allows us to estimate that an average miRNA has approximately 100 evolutionarily conserved target sites, indicating that miRNAs regulate a large fraction of protein-coding genes. Evaluation of 3′ pairing quality suggests that seed sites are the largest group. Sites of this type have been largely overlooked in previous target prediction methods. Results {#s2} ======= The Minimal miRNA Target Site {#s2a} ----------------------------- To improve our understanding of the minimal requirements for a functional miRNA target site, we made use of a simple in vivo assay in the *Drosophila* wing imaginal disc. We expressed a miRNA in a stripe of cells in the central region of the disc and assessed its ability to repress the expression of a ubiquitously transcribed enhanced green fluorescent protein (EGFP) transgene containing a single target site in its 3′ UTR. The degree of repression was evaluated by comparing EGFP levels in miRNA-expressing and adjacent non-expressing cells. Expression of the miRNA strongly reduced EGFP expression from transgenes containing a single functional target site ([Figure 1](#pbio-0030085-g001){ref-type="fig"}A). ![Complementarity to the miRNA 5′ End Is Important for Target Site Function In Vivo\ (A) In vivo assay for target site regulation in the wing imaginal disc. The EGFP reporter is expressed in all cells (green). Cells expressing the miRNA under ptcGal4 control are shown in red. Functional target sites allow strong GFP repression by the miRNA (middle). Non-functional target sites do not (right). Yellow boxes indicate the disc region shown in (B) and later figures.\ (B) Regulation of individual target sites by *miR-7.* Numbers in the upper left of each image indicate the mismatched nucleotide in the target site. Positions important for regulation are shown in red, dispensable positions in green. Regulation by the miRNA is completely abolished in only a few cases.\ (C) Summary of the magnitude of reporter gene repression for the series in (B) and for a second set involving *miR-278* and a target site resembling the *miR-9* site in *Lyra* \[[@pbio-0030085-b26]\]. Positions important for regulation are shown in red, dispensable positions in green. Error bars are based on measurements of 3--5 individual discs.](pbio.0030085.g001){#pbio-0030085-g001} In a first series of experiments we asked which part of the RNA duplex is most important for target regulation. A set of transgenic flies was prepared, each of which contained a different target site for *miR-7* in the 3′ UTR of the EGFP reporter construct. The starting site resembled the strongest *bantam* miRNA site in its biological target *hid* \[[@pbio-0030085-b19]\] and conferred strong regulation when present in a single copy in the 3′ UTR of the reporter gene ([Figure 1](#pbio-0030085-g001){ref-type="fig"}B). We tested the effects of introducing single nucleotide changes in the target site to produce mismatches at different positions in the duplex with the miRNA (note that the target site mismatches were the only variable in these experiments). The efficient repression mediated by the starting site was not affected by a mismatch at positions 1, 9, or 10, but any mismatch in positions 2 to 8 strongly reduced the magnitude of target regulation. Two simultaneous mismatches introduced into the 3′ region had only a small effect on target repression, increasing reporter activity from 10% to 30%. To exclude the possibility that these findings were specific for the tested miRNA sequence or duplex structure, we repeated the experiment with *miR-278* and a different duplex structure. The results were similar, except that pairing of position 8 was not important for regulation in this case ([Figure 1](#pbio-0030085-g001){ref-type="fig"}C). Moreover, some of the mismatches in positions 2--7 still allowed repression of EGFP expression up to 50%. Taken together, these observations support previous suggestions that extensive base-pairing to the 5′ end of the miRNA is important for target site function \[[@pbio-0030085-b26],[@pbio-0030085-b27],[@pbio-0030085-b29],[@pbio-0030085-b32],[@pbio-0030085-b34]\]. We next determined the minimal 5′ sequence complementarity necessary to confer target regulation. We refer to the core of 5′ sequence complementarity essential for target site recognition as the "seed" (Lewis et al. \[[@pbio-0030085-b27]\]). All possible 6mer, 5mer, and 4mer seeds complementary to the first eight nucleotides of the miRNA were tested in the context of a site that allowed strong base-pairing to the 3′ end of the miRNA ([Figure 2](#pbio-0030085-g002){ref-type="fig"}A). The seed was separated from a region of complete 3′ end pairing by a constant central bulge. 5mer and 6mer seeds beginning at positions 1 or 2 were functional. Surprisingly, as few as four base-pairs in positions 2--5 conferred efficient target regulation under these conditions, whereas bases 1--4 were completely ineffective. 4mer, 5mer, or 6mer seeds beginning at position 3 were less effective. These results suggest that a functional seed requires a continuous helix of at least 4 or 5 nucleotides and that there is some position dependence to the pairing, since sites that produce comparable pairing energies differ in their ability to function. For example, the first two duplexes in [Figure 2](#pbio-0030085-g002){ref-type="fig"}A (4mer, top row) have identical 5′ pairing energies (ΔG for the first 8 nt was −8.9 kcal/mol), but only one is functional. Similarly, the third 4mer duplex and fourth 5mer duplex (middle row) have the same energy (−8.7 kcal/mol), but only one is functional. We thus do not find a clear correlation between 5′ pairing energy and function, as reported in \[[@pbio-0030085-b34]\]. These experiments also indicate that extensive 3′ pairing of up to 17 nucleotides in the absence of the minimal 5′ element is not sufficient to confer regulation. Consequently, target searches based primarily on optimizing the extent of base-pairing or the total free energy of duplex formation will include many non-functional target sites \[[@pbio-0030085-b28],[@pbio-0030085-b30],[@pbio-0030085-b35]\], and ranking miRNA target sites according to overall complementarity or free energy of duplex formation might not reflect their biological activity \[[@pbio-0030085-b26],[@pbio-0030085-b27],[@pbio-0030085-b28],[@pbio-0030085-b30],[@pbio-0030085-b35]\]. {#pbio-0030085-g002} To determine the minimal lengths of 5′ seed matches that are sufficient to confer regulation alone, we tested single sites that pair with eight, seven, or six consecutive bases to the miRNA\'s 5′ end, but that do not pair to its 3′ end ([Figure 2](#pbio-0030085-g002){ref-type="fig"}B). Surprisingly, a single 8mer seed (miRNA positions 1--8) was sufficient to confer strong regulation by the miRNA. A single 7mer seed (positions 2--8) was also functional, although less effective. The magnitude of regulation for 8mer and 7mer seeds was strongly increased when two copies of the site were introduced in the UTR. In contrast, 6mer seeds showed no regulation, even when present in two copies. Comparable results were recently reported for two copies of an 8mer site with limited 3′ pairing capacity in a cell-based assay \[[@pbio-0030085-b34]\]. These results do not support a requirement for a central bulge, as suggested previously \[[@pbio-0030085-b29]\]. We took care in designing the miRNA 3′ ends to exclude any 3′ pairing to nearby sequence according to RNA secondary structure prediction. However, we cannot rule out the possibility that extensive looping of the UTR sequence might allow the 3′ end to pair to sequences further downstream in our reporter constructs. Note, however, that even if remote 3′ pairing was occurring and required for function of 8- and 7mer seeds, it is not sufficient for 5′ matches with less than seven complementary bases (all test sites are in the same sequence context; [Figure 2](#pbio-0030085-g002){ref-type="fig"}B). In addition, pairing at a random level will occur in any sequence if long enough loops are allowed. However, whether the ribonucleoprotein complexes involved in translational repression require 3′ pairing, and whether they are able to allow extensive looping to achieve this, remains an open question. Computationally, remote 3′ pairing cannot be distinguished from random matches if loops of any length are allowed. On this basis any site with a 7- or 8mer seed has to be taken seriously---especially when evolutionarily conserved. From these experiments we conclude that (1) complementarity of seven or more bases to the 5′ end miRNA is sufficient to confer regulation, even if the target 3′ UTR contains only a single site; (2) sites with weaker 5′ complementarity require compensatory pairing to the 3′ end of the miRNA in order to confer regulation; and (3) extensive pairing to the 3′ end of the miRNA is not sufficient to confer regulation on its own without a minimal element of 5′ complementarity. The Effect of G:U Base-Pairs and Bulges in the Seed {#s2b} --------------------------------------------------- Several confirmed miRNA target genes contain predicted binding sites with seeds that are interrupted by G:U base-pairs or single nucleotide bulges \[[@pbio-0030085-b17],[@pbio-0030085-b19],[@pbio-0030085-b26],[@pbio-0030085-b36],[@pbio-0030085-b37],[@pbio-0030085-b38],[@pbio-0030085-b39]\]. In most cases these mRNAs contain multiple predicted target sites and the contributions of individual sites have not been tested. In vitro tests have shown that sites containing G:U base-pairs can function \[[@pbio-0030085-b29],[@pbio-0030085-b34]\], but that G:U base-pairs contribute less to target site function than would be expected from their contribution to the predicted base-pairing energy \[[@pbio-0030085-b34]\]. We tested the ability of single sites with seeds containing G:U base-pairs and bulges to function in vivo. One, two, or three G:U base-pairs were introduced into single target sites with 8mer, 7mer, or 6mer seeds ([Figure 3](#pbio-0030085-g003){ref-type="fig"}A). A single G:U base-pair caused a clear reduction in the efficiency of regulation by an 8mer seed site and by a 7mer seed site. The site with a 6mer seed lost its activity almost completely. Having more than one G:U base-pair compromised the activity of all the sites. As the target sites were designed to allow optimal 3′ pairing, we conclude that G:U base-pairs in the seed region are always detrimental. {#pbio-0030085-g003} Single nucleotide bulges in the seed are found in the *let-7* target *lin-41* and in the *lin-4* target *lin-14* \[[@pbio-0030085-b17],[@pbio-0030085-b36],[@pbio-0030085-b37]\]. Recent tissue culture experiments have led to the proposal that such bulges are tolerated if positioned symmetrically in the seed region \[[@pbio-0030085-b29]\]. We tested a series of sites with single nucleotide bulges in the target or the miRNA ([Figure 3](#pbio-0030085-g003){ref-type="fig"}B). Only some of these sites conferred good regulation of the reporter gene. Our results do not support the idea that such sites depend on a symmetrical arrangement of base-pairs flanking the bulge. We also note that the identity of the bulged nucleotide seems to matter. While it is clear that some target sites with one nucleotide bulge or a single mismatch can be functional if supported by extensive complementarity to the miRNA 3′ end, it is not possible to generalize about their potential function. Functional Categories of Target Sites {#s2c} ------------------------------------- While recognizing that there is a continuum of base-pairing quality between miRNAs and target sites, the experiments presented above suggest that sites that depend critically on pairing to the miRNA 5′ end (5′ dominant sites) can be distinguished from those that cannot function without strong pairing to the miRNA 3′ end (3′ compensatory sites). The 3′ compensatory group includes seed matches of four to six base-pairs and seeds of seven or eight bases that contain G:U base-pairs, single nucleotide bulges, or mismatches. We consider it useful to distinguish two subgroups of 5′ dominant sites: those with good pairing to both 5′ and 3′ ends of the miRNA (canonical sites) and those with good 5′ pairing but with little or no 3′ pairing (seed sites). We consider seed sites to be those where there is no evidence for pairing of the miRNA 3′ end to nearby sequences that is better than would be expected at random. We cannot exclude the possibility that some sites that we identify as seed sites might be supported by additional long-range 3′ pairing. Computationally, this is always possible if long enough loops in the UTR sequence are allowed. Whether long loops are functional in vivo remains to be determined. Canonical sites have strong seed matches supported by strong base-pairing to the 3′ end of the miRNA. Canonical sites can thus be seen as an extension of the seed type (with enhanced 3′ pairing in addition to a sufficient 5′ seed) or as an extension of the 3′compensatory type (with improved 5′ seed quality in addition to sufficient 3′ pairing). Individually, canonical sites are likely to be more effective than other site types because of their higher pairing energy, and may function in one copy. Due to their lower pairing energies, seed sites are expected to be more effective when present in more than one copy. [Figure 4](#pbio-0030085-g004){ref-type="fig"} presents examples of the different site types in biologically relevant miRNA targets and illustrates their evolutionary conservation in multiple drosophilid genomes. {#pbio-0030085-g004} Most currently identified miRNA target sites are canonical. For example, the *hairy* 3′ UTR contains a single site for *miR-7,* with a 9mer seed and a stretch of 3′ complementarity. This site has been shown to be functional in vivo \[[@pbio-0030085-b26]\], and it is strikingly conserved in the seed match and in the extent of complementarity to the 3′ end of *miR-7* in all six orthologous 3′ UTRs. Although seed sites have not been previously identified as functional miRNA target sites, there is some evidence that they exist in vivo. For example, the *Bearded (Brd)* 3′ UTR contains three sequence elements, known as Brd boxes, that are complementary to the 5′ region of *miR-4* and *miR-79* \[[@pbio-0030085-b32],[@pbio-0030085-b40]\]. Brd boxes have been shown to repress expression of a reporter gene in vivo, presumably via miRNAs, as expression of a *Brd* 3′ UTR reporter is elevated in *dicer-1* mutant cells, which are unable to produce any miRNAs \[[@pbio-0030085-b14]\]. All three Brd box target sites consist of 7mer seeds with little or no base-pairing to the 3′ end of either *miR-4* or *miR-79* (see below). The alignment of *Brd* 3′ UTRs shows that there is little conservation in the *miR-4* or *miR-79* target sites outside the seed sequence, nor is there conservation of pairing to either miRNA 3′ end. This suggests that the sequences that could pair to the 3′ end of the miRNAs are not important for regulation as they do not appear to be under selective pressure. This makes it unlikely that a yet unidentified Brd box miRNA could form a canonical site complex. The 3′ UTR of the HOX gene *Sex combs reduced (Scr)* provides a good example of a 3′ compensatory site. *Scr* contains a single site for *miR-10* with a 5mer seed and a continuous 11-base-pair complementarity to the miRNA 3′ end \[[@pbio-0030085-b28]\]. The *miR-10* transcript is encoded within the same HOX cluster downstream of *Scr,* a situation that resembles the relationship between *miR-iab-5p* and *Ultrabithorax* in flies \[[@pbio-0030085-b26]\] and *miR-196*/*HoxB8* in mice \[[@pbio-0030085-b41]\]. The predicted pairing between *miR-10* and *Scr* is perfectly conserved in all six drosophilid genomes, with the only sequence differences occurring in the unpaired loop region. The site is also conserved in the 3′ UTR of the Scr genes in the mosquito, Anopheles gambiae, the flour beetle, Tribolium castaneum, and the silk moth,Bombyx mori. Conservation of such a high degree of 3′ complementarity over hundreds of millions of years of evolution suggests that this is likely to be a functional *miR-10* target site. Extensive 5′ and 3′ sequence conservation is also seen for other 3′ compensatory sites, e.g., the two *let-7* sites in *lin-41* or the *miR-2* sites in *grim* and *sickle* \[[@pbio-0030085-b17],[@pbio-0030085-b26],[@pbio-0030085-b36]\]. The miRNA 3′ End Determines Target Specificity within miRNA Families {#s2d} -------------------------------------------------------------------- Several families of miRNAs have been identified whose members have common 5′ sequences but differ in their 3′ ends. In view of the evidence that 5′ ends of miRNA are functionally important \[[@pbio-0030085-b26],[@pbio-0030085-b27],[@pbio-0030085-b29],[@pbio-0030085-b42]\], and in some cases sufficient (present study), it can be expected that members of miRNA families may have redundant or partially redundant functions. According to our model, 5′ dominant canonical and seed sites should respond to all members of a given miRNA family, whereas 3′ compensatory sites should differ in their sensitivity to different miRNA family members depending on the degree of 3′ complementarity. We tested this using the wing disc assay with 3′ UTR reporter transgenes and overexpression constructs for various miRNA family members. *miR-4* and *miR-79* share a common 5′ sequence that is complementary to a single 8mer seed site in the *bagpipe* 3′ UTR ([Figure 5](#pbio-0030085-g005){ref-type="fig"}A and [5](#pbio-0030085-g005){ref-type="fig"}B). The 3′ ends of the miRNAs differ. *miR-4* is predicted to have 3′ pairing at approximately 50% of the maximally possible level (−10.8 kcal/mol), whereas the level of 3′ pairing for *miR-79* is approximately 25% maximum (−6.1 kcal/mol), which is below the average level expected for random matches (see below). Both miRNAs repressed expression of the *bagpipe* 3′ UTR reporter, regardless of the 3′ complementarity ([Figure 5](#pbio-0030085-g005){ref-type="fig"}B). This indicates that both types of site are functional in vivo and suggests that *bagpipe* is a target for both miRNAs in this family. {#pbio-0030085-g005} To test whether miRNA family members can also have non-overlapping targets, we used 3′ UTR reporters of the pro-apoptotic genes *grim* and *sickle,* two recently identified miRNA targets \[[@pbio-0030085-b26]\]. Both genes contain K boxes in their 3′ UTRs that are complementary to the 5′ ends of the *miR-2, miR-6,* and *miR-11* miRNA family \[[@pbio-0030085-b26],[@pbio-0030085-b32]\]. These miRNAs share residues 2--8 but differ considerably in their 3′ regions ([Figure 5](#pbio-0030085-g005){ref-type="fig"}A). The site in the *grim* 3′ UTR is predicted to form a 6mer seed match with all three miRNAs ([Figure 5](#pbio-0030085-g005){ref-type="fig"}C, left), but only *miR-2* shows the extensive 3′ complementarity that we predict would be needed for a 3′ compensatory site with a 6mer seed to function (−19.1 kcal/mol, 63% maximum 3′ pairing, versus −10.9 kcal/mol, 46% maximum, for *miR-11* and −8.7 kcal/mol, 37% maximum, for *miR-6*). Indeed, only *miR-2* was able to regulate the *grim* 3′ UTR reporter, whereas *miR-6* and *miR-11* were non-functional. The *sickle* 3′ UTR contains two K boxes and provides an opportunity to test whether weak sites can function synergistically. The first site is similar to the *grim* 3′ UTR in that it contains a 6mer seed for all three miRNAs but extensive 3′ complementarity only to *miR-2*. The second site contains a 7mer seed for *miR-2* and *miR-6* but only a 6mer seed for *miR-11* ([Figure 5](#pbio-0030085-g005){ref-type="fig"}C, right). *miR-2* strongly downregulated the *sickle* reporter, *miR-6* had moderate activity (presumably via the 7mer seed site), and *miR-11* had nearly no activity, even though the miRNAs were overexpressed. The fact that a site is targeted by at least one miRNA argues that it is accessible (e.g., *miR-2* is able to regulate both UTR reporters), and that the absence of regulation for other family members is due to the duplex structure. These results are in line with what we would expect based on the predicted functionality of the individual sites, and indicate that our model of target site functionality can be extended to UTRs with multiple sites. Weak sites that do not function alone also do not function when they are combined. To show that endogenous miRNA levels regulate all three 3′ UTR reporters, we compared EGFP expression in wild-type cells and *dicer-1* mutant cells, which are unable to produce miRNAs \[[@pbio-0030085-b14]\]. *dicer-1* clones did not affect a control reporter lacking miRNA binding sites, but showed elevated expression of a reporter containing the 3′ UTR of the previously identified *bantam* miRNA target *hid* ([Figure 5](#pbio-0030085-g005){ref-type="fig"}D). Similarly, all 3′ UTR reporters above were upregulated in *dicer-1* mutant cells, indicating that *bagpipe, sickle,* and *grim* are subject to repression by miRNAs expressed in the wing disc. Taken together, these experiments indicate that transcripts with 5′ dominant canonical and seed sites are likely to be regulated by all members of a miRNA family. However, transcripts with 3′ compensatory sites can discriminate between miRNA family members. Genome-Wide Occurrence of Target Sites {#s2e} -------------------------------------- Experimental tests such as those presented above and the observed evolutionary conservation suggest that all three types of target sites are likely to be used in vivo. To gain additional evidence we examined the occurrence of each site type in all Drosophila melanogaster 3′ UTRs. We made use of the D. pseudoobscura genome, the second assembled drosophilid genome, to determine the degree of site conservation for the three different site classes in an alignment of orthologous 3′ UTRs. From the 78 known *Drosophila* miRNAs, we selected a set of 49 miRNAs with non-redundant 5′ sequences. We first investigated whether sequences complementary to the miRNA 5′ ends were better conserved than would be expected for random sequences. For each miRNA, we constructed a cohort of ten randomly shuffled variants. To avoid a bias for the number of possible target matches, the shuffled variants were required to produce a number of sequence matches comparable (±15%) to the original miRNAs for D. melanogaster 3′ UTRs. 7mer and 8mer seeds complementary to real miRNA 5′ ends were significantly better conserved than those complementary to the shuffled variants. This is consistent with the findings of Lewis et al. \[[@pbio-0030085-b27]\] but was obtained without the need to use a rank and energy cutoff applied to the full-length miRNA target duplex, as was the case for vertebrate miRNAs. Conserved 8mer seeds for real miRNAs occur on average 2.8 times as often as seeds complementary to the shuffled miRNAs ([Figure 6](#pbio-0030085-g006){ref-type="fig"}A). For 7mer seeds this signal was 2:1, whereas 6mer, 5mer, and 4mer seeds did not show better conservation than expected for random sequences. To assess the validity of these signals and to control for the random shuffling of miRNAs, we repeated this procedure with "mutant" miRNAs in which two residues in the 5′ region were changed. There was no difference between the mutant test miRNAs and their shuffled variants ([Figure 6](#pbio-0030085-g006){ref-type="fig"}A). This indicates that a substantial fraction of the conserved 7mer and 8mer seeds complementary to real miRNAs identify biologically relevant target sites. {#pbio-0030085-g006} 3′ compensatory and canonical sites depend on substantial pairing to the miRNA 3′ end. For these sites, we expect UTR sequences adjacent to miRNA 5′ seed matches to pair better to the miRNA 3′ end than to random sequences. However, unlike 5′ complementarity, 3′ base-pairing preference was not detected in previous studies looking at sequence complementarity and nucleotide conservation because UTR sequences complementary to the miRNA 3′ end were not better conserved than would be expected at random \[[@pbio-0030085-b27]\]. On this basis, we decided to treat the 5′ and 3′ ends of the miRNA separately. For the 5′ end, seed matches were required to be fully conserved in an alignment of orthologous D. melanogaster and D. pseudoobscura 3′ UTRs (we expected one-half to two-thirds of these matches to be real miRNA sites). We first investigated the overall conservation of UTR sequences adjacent to the conserved seed matches and found that overall the sequences are not better conserved than a random control with shuffled miRNAs ([Figure 6](#pbio-0030085-g006){ref-type="fig"}B). For both real and random matches, the number of sites increases with the degree of 3′ conservation (up to the 80% level), reflecting the increased probability that sequences adjacent to conserved seed matches will also lie in blocks of conserved sequence ([Figure 6](#pbio-0030085-g006){ref-type="fig"}B). For real 7mers and 8mers we found a slightly higher percentage of sites between 30% and 80% identity than we did for the shuffled controls. In contrast, the ratio of sites with over 80% sequence identity was smaller for real 7- or 8mers than for random ones, meaning that in highly conserved 3′ UTR blocks (\>80% identity) the ratio of random matches exceeds that of real miRNA target sites. This caused us to question whether the degree of conservation for sequences adjacent to seed matches correlates with miRNA 3′ pairing as would be expected if the conservation were due to a biologically relevant miRNA target site. Indeed, we found that the best conserved sites adjacent to seed matches (i.e., those with zero, one, or two mismatches in the 3′ UTR alignment) and the least conserved sites (i.e., those with only three, two, or one matching nucleotides) are not distinguishable in that both pair only randomly to the corresponding miRNA 3′ end (approximately 35% maximal 3′ pairing energy, data not shown). The observation that miRNA target sites do not seem to be fully conserved over their entire length is consistent with the examples shown in [Figure 4](#pbio-0030085-g004){ref-type="fig"} in which only the degree of 3′ pairing but not the nucleotide identity is conserved *(miR-7/hairy),* or at least the unpaired bulge is apparently not under evolutionary pressure *(miR-10/Scr).* Although this result obviously depends on the evolutionary distance of the species under consideration (see \[[@pbio-0030085-b43]\] for a comparison of mammalian sites), it shows that conclusions about the contribution of miRNA 3′ pairing to target site function cannot be drawn solely from the degree of sequence conservation. We therefore chose to evaluate the quality of 3′ pairing by the stability of the predicted RNA--RNA duplex. We assessed predicted pairing energy between the miRNA 3′ end and the adjacent UTR sequence for both *Drosophila* species and used the lower score. Use of the lower score measures conservation of the overall degree of pairing without requiring sequence identity. [Figure 6](#pbio-0030085-g006){ref-type="fig"}C shows the distribution of the 3′ pairing energies for all conserved 3′ compensatory *miR-7* sites identified by a 6mer seed match, compared to the distribution of 50 *miR-7* sequences shuffled only in the 3′ part, leaving the 5′ unchanged. This means that real and shuffled miRNAs identify the same 5′ seed matches in the 3′ UTRs, which allows us to compare the 3′ pairing characteristics of the adjacent sequences. We also required 3′ shuffled sequences to have similar pairing energies (±15%) to their complementary sequences and to 10,000 randomly selected sites to exclude generally altered pairing characteristics. The distributions for real and shuffled miRNAs were highly similar, with a mean of approximately 35% of maximal 3′ pairing energy and few sites above 55%. However, a small number of sites paired exceptionally well to *miR-7* at energies that were far above the shuffled averages and not reached by any of the 50 shuffled controls. This example illustrates that there is a significant difference between real and shuffled miRNAs for the sites with the highest 3′ complementarity, which are likely to be biologically relevant. Sites with weaker 3′ pairing might also be functional, but cannot be distinguished from random matches and can only be validated by experiments (see [Figure 5](#pbio-0030085-g005){ref-type="fig"}). To provide a global analysis of 3′ pairing comprising all miRNAs and to investigate how many miRNAs show significantly non-random 3′ pairing, we considered only the sites within the highest 1% of 3′ pairing energies. The average of the highest 1% of 3′ pairing energies of each of 58 3′ non-redundant miRNAs was divided by that of its 50 3′ shuffled controls. This ratio is one if the averages are the same, and increases if the real miRNA has better 3′ pairing than the shuffled miRNAs. To test whether a signal was specific for real miRNAs, we repeated the same protocol with a mutant version of each miRNA. The altered 5′ sequence in the mutant miRNA selects different seed matches than the real miRNA and permits a comparison of sequences that have not been under selection for complementarity to miRNA 3′ ends with those that may have been. [Figure 6](#pbio-0030085-g006){ref-type="fig"}D shows the distribution of the energy ratios for canonical (left) and 3′ compensatory sites (right) for all 58 real and mutated 3′ non-redundant miRNAs. Most real miRNAs had ratios close to one, comparable to the mutants. But several had ratios well above those observed for mutant miRNAs, indicating significant conserved 3′ pairing. A small fraction of sites show exceptionally good 3′ pairing. If we use 3′ pairing energy cutoffs to examine site quality for all miRNAs, we expect sites of this type to be distinguishable from random matches. The ratio of the number of sites above the cutoff for real versus 3′ shuffled miRNAs was plotted as a function of the 3′ pairing cutoff ([Figure 6](#pbio-0030085-g006){ref-type="fig"}E). For low cutoffs the ratio is one, as the number of sites corresponds to the number of seed matches (which is identical for real and 3′ shuffled miRNAs). For increasing cutoffs, the ratios increase once a certain threshold is reached, reflecting overrepresentation of sites that pair favorably to the real miRNA 3′ end but not the 3′ shuffled miRNAs. The maximal ratio obtained for mutated miRNAs never exceeded five, which we used as the threshold level to define where significant overrepresentation begins. For 8mer seed sites overrepresentation began at 55% maximal 3′ pairing; for 7mer seed sites, at 65%; for 6mer seed sites, at 68%; and for 5mer seed sites, at 78%. There was no statistical evidence for sites with 4mer seeds. We also tested whether sequences forming 7mer or 8mer seeds containing G:U base-pairs, mismatches, or bulges were better conserved if complementary to real miRNAs. We did not find any statistical evidence for these seed types. Analysis of 3′ pairing also failed to show any non-random signal for these sites. This suggests that such sites are few in number genome-wide and are not readily distinguished from random matches. Nonetheless, our experiments do show that sites of this type can function in vivo. The *let-7* sites in *lin-41* provide a natural example. Most Sites Lack Substantial 3′ Pairing {#s2f} -------------------------------------- The experimental and computational results presented above provide information about 5′ and 3′ pairing that allows us to estimate the number of target sites of each type in *Drosophila.* The number of 3′ compensatory sites cannot be estimated on the basis of 5′ pairing, because seed matches of four, five, or six bases cannot be distinguished from random matches, reflecting that a large number of randomly conserved and non-functional matches predominate ([Figure 6](#pbio-0030085-g006){ref-type="fig"}A). Significant 3′ pairing can be distinguished from random matches for 6mer sites above 68% maximal 3′ pairing energy, and above 78% for 5mers ([Figure 6](#pbio-0030085-g006){ref-type="fig"}E). Using these pairing levels gives an estimate of one 3′ compensatory site on average per miRNA. The experiments in [Figure 5](#pbio-0030085-g005){ref-type="fig"} provide an opportunity to assess the contribution of 3′ pairing to the ability of sites with 6mer seeds to function. The 6mer K box site in the *grim* 3′ UTR was regulated by *miR-2* (63% maximal 3′ pairing energy), but not by *miR-11,* which has a predicted 3′ pairing energy of 46%. Similarly, the 6mer seed sites for *miR-11* in the *sickle* 3′ UTR had 3′ pairing energies of approximately 35% and were non-functional. We can use the 63% and 46% levels to provide upper and lower estimates of one and 20 3′ compensatory 6mer sites on average per miRNA. For 5mer sites, the examples in [Figure 1](#pbio-0030085-g001){ref-type="fig"} show that sites with 76% and 83% maximal 3′ pairing do not function. At the 80% threshold level, we expect less than one additional site on average per miRNA, suggesting that 3′ compensatory sites with 5mer seeds are rare. The predicted *miR-10* site in *Scr* (see [Figure 4](#pbio-0030085-g004){ref-type="fig"}) is one of the few sites with a 5mer seed that reaches this threshold (100% maximum 3′ pairing energy; −20 kcal/mol). It is likely that other sites in this group will also prove to be functionally important. The overrepresentation of conserved 5′ seed matches (see [Figure 6](#pbio-0030085-g006){ref-type="fig"}A) suggests that approximately two-thirds of sites with 8mer seeds and approximately one-half of the sites with 7mer seeds are biologically relevant. This corresponds to an average of 28 8mers and 53 7mers, for a total of 81 sites per miRNA. We define canonical sites as those with meaningful contributions from both 5′ and 3′ pairing. Given that 7- and 8mer seed matches can function without significant 3′ pairing, it is difficult to assess at what level 3′ pairing contributes meaningfully to their function. The range of 3′ pairing energies that were minimally sufficient to support a weak seed match was between 46% and 63% of maximum pairing energy (see [Figure 5](#pbio-0030085-g005){ref-type="fig"}C). If we take the 46% level as the lower limit for meaningful 3′ pairing, over 95% of sites would be considered seed sites. This changes to 99% for pairing energies that can be statistically distinguished from noise (55% maximal; see [Figure 6](#pbio-0030085-g006){ref-type="fig"}E) and remains over 50% even for pairing energies at the average level achieved by random matches (30% maximal). It is clear from this analysis that the majority of miRNA target sites lack substantial pairing in the 3′ end in nearby sequences. Indeed the 3′ pairing level for the three seed sites for *miR-4* in *Brd* are all less than 25% (i.e., below the average for random matches) and *Brd* was thus not predicted as a *miR-4* target previously \[[@pbio-0030085-b26],[@pbio-0030085-b28],[@pbio-0030085-b35]\]. Again, we note the caveat that some of sites that we identify as seed could in principle be supported by 3′ pairing to more distant upstream sequences, but also that such sites would be difficult to distinguish from background computationally and that it is unclear whether large loops are functional. If there were statistical evidence for 3′ pairing that is lower than would be expected at random for some sites, this would be one line of argument for a discrete functional class that does not use 3′ pairing and would therefore suggest selection against 3′ pairing. Although the overall distribution of 3′ pairing energies for real miRNA 3′ ends adjacent to 8mer seed matches is very similar to the random control with 3′ shuffled sequences ([Figure 7](#pbio-0030085-g007){ref-type="fig"}; *R* ^2^ = 0.98), we observed a small but significant overrepresentation of real sites on both sides of the random distribution, which leads to a slightly wider distribution of real sites at the expense of the peak values around 30% pairing. Bearing in mind that one-third of 8mer seed matches are false positives (see [Figure 6](#pbio-0030085-g006){ref-type="fig"}A), we can account for the noise by subtracting one-third of the random distribution. We then see two peaks at around 20% and 35% maximum pairing energy, separated by a dip. Subtracting more (e.g., one-half or two-thirds) of the random distribution increases the separation of the two peaks, suggesting that the underlying distribution of 3′ pairing for real 8mer seed sites might indeed be bimodal. This effect is still present, though less pronounced, if 7mer seed matches are included. No such effect is seen for the combined 5- and 6mer seed matches. In addition, we see no difference between a random (noise) model that evaluates 3′ pairing of 3′ shuffled miRNAs to UTR sites identified by real miRNA seed matches and a random model that pairs the real (i.e., non-shuffled) miRNA 3′ end to randomly chosen UTR sequences, thus excluding bias due to shuffling. Overall, these results suggest that there might indeed be a bimodal distribution due to an enrichment of sites with both better and worse 3′ pairing than would be expected at random. We take this as evidence that seed sites are a biologically meaningful subgroup within the 5′ dominant site category. {#pbio-0030085-g007} Overall, these estimates suggest that there are over 80 5′ dominant sites and 20 or fewer 3′ compensatory sites per miRNA in the *Drosophila* genome. As estimates of the number of miRNAs in *Drosophila* range from 96 to 124 \[[@pbio-0030085-b44]\], this translates to 8,000--12,000 miRNA target sites genome-wide, which is close to the number of protein-coding genes. Even allowing for the fact that some genes have multiple miRNA target sites, these findings suggest that a large fraction of genes are regulated by miRNAs. Discussion {#s3} ========== We have provided experimental and computational evidence for different types of miRNA target sites. One key finding is that sites with as little as seven base-pairs of complementarity to the miRNA 5′ end are sufficient to confer regulation in vivo and are used in biologically relevant targets. Genome-wide, 5′ dominant sites occur 2- to 3-fold more often in conserved 3′ UTR sequences than would be expected at random. The majority of these sites have been overlooked by previous miRNA target prediction methods because their limited capacity to base-pair to the miRNA 3′ end cannot be distinguished from random noise. Such sites rank low in search methods designed to optimize overall pairing energy \[[@pbio-0030085-b16],[@pbio-0030085-b17],[@pbio-0030085-b26],[@pbio-0030085-b27],[@pbio-0030085-b28],[@pbio-0030085-b30],[@pbio-0030085-b35]\]. Indeed, we find that few seed sites scored high enough to be considered seriously in these earlier predictions, even when 5′ complementarity was given an additional weighting (e.g., \[[@pbio-0030085-b28],[@pbio-0030085-b43]\]. We thus suspect that methods with pairing cutoffs would exclude many, if not all, such sites. In a scenario in which protein-coding genes acquire miRNA target sites in the course of evolution \[[@pbio-0030085-b04]\], it is likely that seed sites with only seven or eight bases complementary to a miRNA would be the first functional sites to be acquired. Once present, a site would be retained if it conferred an advantage, and sites with extended complementarity could also be selected to confer stronger repression. In this scenario, the number of sites might grow over the course of evolution so that ancient miRNAs would tend to have more targets than those more recently evolved. Likewise, genes that should not be repressed by the miRNA milieu in a given cell type would tend to avoid seed matches to miRNA 5′ ends ("anti-targets" \[[@pbio-0030085-b04]\]). Although a 7- to 8mer seed is sufficient for a site to function, additional 3′ pairing increases miRNA functionality. The activity of a single 7mer canonical site is expected to be greater than an equivalent seed site. Likewise, the magnitude of miRNA-induced repression is reduced by introducing 3′ mismatches into a canonical site. Genome-wide, there are many sites that appear to show selection for conserved 3′ pairing and, interestingly, many sites that appear to show selection against 3′ pairing. In vivo, canonical sites might function at lower miRNA concentrations and might repress translation more effectively, particularly when multiple sites are present in one UTR (e.g., \[[@pbio-0030085-b42]\]). Efficient repression is likely to be necessary for genes whose expression would be detrimental, as illustrated by the genetically identified miRNAs, which produce clear mutant phenotypes when their targets are not normally repressed ("switch targets" \[[@pbio-0030085-b04]\]). Prolonged expression of the *lin-14* and *lin-41* genes in Caenorhabditis elegans mutant for *lin-4* or *let-7* causes developmental defects, and their regulation involves multiple sites \[[@pbio-0030085-b17],[@pbio-0030085-b36],[@pbio-0030085-b37]\]. Similarly, multiple target sites allow robust regulation of the pro-apoptotic gene *hid* by *bantam* miRNA in *Drosophila* \[[@pbio-0030085-b19]\]. More subtle modulation of expression levels could be accomplished by weaker sites, such as those lacking 3′ pairing. Sites that cannot function efficiently alone are in fact a prerequisite for combinatorial regulation by multiple miRNAs. Seed sites might thus be useful for situations in which the combined input of several miRNAs is used to regulate target expression. Depending on the nature of the target sites, any single miRNA might not have a strong effect on its own, while being required in the context of others. 3′ Complementarity Distinguishes miRNA Family Members {#s3a} ----------------------------------------------------- 3′ compensatory sites have weak 5′ pairing and need substantial 3′ pairing to function. We find genome-wide statistical support for 3′ compensatory sites with 5mer and 6mer seeds and show that they are used in vivo. Furthermore, these sites can be differentially regulated by different miRNA family members depending on the quality of their 3′ pairing (e.g., regulation of the pro-apoptotic genes *grim* and *sickle* by *miR-2, miR-6,* and *miR-11*). Thus, members of a miRNA family may have common targets as well as distinct targets. They may be functionally redundant in regulation of some targets but not others, and so we can expect some overlapping phenotypes as well as differences in their mutant phenotypes. Following this reasoning, it is likely that the *let-7* miRNA family members differentially regulate *lin-41* in C. elegans \[[@pbio-0030085-b17],[@pbio-0030085-b45]\]. The seed matches in *lin-41* to *let-7* and the related miRNAs *miR-48, miR-84,* and *miR-241* are weak, and only *let-7* has strong 3′ pairing. On this basis, it seems likely that *lin-41* is regulated only by *let-7*. In contrast, *hbl-1* has four sites with strong seed matches \[[@pbio-0030085-b38],[@pbio-0030085-b39]\], and we expect it to be regulated by all four *let-7* family members. As all four *let-7*-related miRNAs are expressed similarly during development \[[@pbio-0030085-b06]\], their role as regulators of *hbl-1* may be redundant. *let-7* must also have targets not shared by the other family members, as its function is essential. *lin-41* is likely to be one such target. The idea that the 3′ end of miRNAs serves as a specificity factor provides an attractive explanation for the observation that many miRNAs are conserved over their full length across species separated by several hundreds of millions of years of evolution. 3′ compensatory sites may have evolved from canonical sites by mutations that reduce the quality of the seed match. This could confer an advantage by allowing a site to become differentially regulated by miRNA family members. In addition, sites could retain specificity and overall pairing energy, but with reduced activity, perhaps permitting discrimination between high and low levels of miRNA expression. This might also allow a target gene to acquire a dependence on inputs from multiple miRNAs. These scenarios illustrate a few ways in which more complex regulatory roles for miRNAs might arise during evolution. A Large Fraction of the Genome Is Regulated by miRNAs {#s3b} ----------------------------------------------------- Another intriguing outcome of this study is evidence for a surprisingly large number of miRNA target sites genome-wide. Even our conservative estimate is far above the numbers of sites in recent predictions, e.g., seven or fewer per miRNA \[[@pbio-0030085-b27],[@pbio-0030085-b28],[@pbio-0030085-b29]\]. Our estimate of the total number of targets approaches the number of protein-coding genes, suggesting that regulation of gene expression by miRNAs plays a greater role in biology than previously anticipated. Indeed, Bartel and Chen \[[@pbio-0030085-b46]\] have suggested in a recent review that the earlier estimates were likely to be low, and a recent study by John et al. \[[@pbio-0030085-b43]\], published while this manuscript was under review, predicts that approximately 10% of human genes are regulated by miRNAs. We agree with these authors\' suggestion that this is likely an underestimate, because their method identifies an average of only 7.1 target genes per miRNA, with few that we would classify as seed sites lacking substantial 3′ pairing. A large number of target sites per miRNA is also consistent with combinatorial gene regulation by miRNAs, analogous to that by transcription factors, leading to cell-type-specific gene expression \[[@pbio-0030085-b47]\]. Sites for multiple miRNAs allow for the possibility of cell-type-specific miRNA combinations to confer robust and specific gene regulation. Our results provide an improved understanding of some of the important parameters that define how miRNAs bind to their target genes. We anticipate that these will be of use in understanding known miRNA--target relationships and in improving methods to predict miRNA targets. We have limited our evaluation to target sites in 3′ UTRs. miRNAs directed at other types of targets or with dramatically different functions (e.g., in regulation of chromatin structure) might well use different rules. Accordingly, there may prove to be more targets than we can currently estimate. Further, there may be additional features, such as overall UTR context, that either enhance or limit the accessibility of predicted sites and hence their ability to function. For example, the rules about target site structure cannot explain the apparent requirement for the linker sequence observed in the *let-7/lin-41* regulation \[[@pbio-0030085-b48]\]. Further efforts toward experimental target site validation and systematic examination of UTR features can be expected to provide new insight into the function of miRNA target sites. Materials and Methods {#s4} ===================== {#s4a} ### Fly strains {#s4a1} ptcGal4; EP miR278 was provided by Aurelio Teleman. The control, *hid, grim,* and *sickle* 3′ UTR reporter transgenes, and UAS-*miR-2b* are described in \[[@pbio-0030085-b19],[@pbio-0030085-b26]\]. For UAS constructs for miRNA overexpression, genomic fragments including *miR-4* (together with *miR-286* and *miR-5*) and *miR-11* were amplified by PCR and cloned into UAS-DSred as described for UAS-*miR-7* \[[@pbio-0030085-b26]\]. Details are available on request. UAS-*miR-79* (also contains *miR-9b* and *miR-9c*) and UAS-*miR-6* (*miR-6--1, miR-6--2,* and *miR-6--3*) were kindly provided by Eric Lai. *dcr-1* ^Q1147X^ is described in \[[@pbio-0030085-b14]\]. ### Clonal analysis {#s4a2} Clones mutant for *dcr-1* ^Q1147X^ were induced in HS-Flp;dcr-1 FRT82/armadillo-lacZ FRT82 larvae by heat shock for 1 h at 38 °C at 50--60 h of development. Wandering third-instar larvae were dissected and labeled with rabbit anti-GFP (Torrey Pines Biolabs, Houston, Texas, United States; 1:400) and anti-β-Gal (rat polyclonal, 1:500). ### Reporter constructs {#s4a3} The *bagpipe* 3′ UTR was PCR amplified from genomic DNA (using the following primers \[enzyme sites in lower case\]: AAtctaga AGGTTGGGAGTGACCATGTCTC and AActcgag TATTTAGCTCTCGGGTAGATACG) and cloned downstream of the tubulin promoter and EGFP (Clontech, Palo Alto, California, United States) in Casper4 as in \[[@pbio-0030085-b26]\]. ### Single target site constructs {#s4a4} Oligonucleotides containing the target site sequences shown in the figures were annealed and cloned downstream of tub\>EGFP and upstream of SV40polyA (XbaI/XhoI). Clones were verified by DNA sequencing. Details are available on request. ### EGFP intensity measurements {#s4a5} NIH image 1.63 was used to quantify intensity levels in miRNA-expressing and non-expressing cells from confocal images. Depending on the variation, between three and five individual discs were analyzed. ### 3′ UTR alignments {#s4a6} For each D. melanogaster gene, we identified the D. pseudoobscura ortholog using TBlastn as described in \[[@pbio-0030085-b26]\]. We then aligned the D. melanogaster 3′ UTR obtained from the Berkeley Drosophila Genome Project to the D. pseudoobscura 3′ adjacent sequence (Human Genome Sequencing Center at Baylor College of Medicine) using AVID \[[@pbio-0030085-b49]\]. For individual examples, we manually mapped the D. melanogaster coding region to genomic sequence traces (National Center for Biotechnology Information trace archive) of *D. ananassae, D. virilis, D. simulans,* and D. yakuba by TBlastn and extended the sequences by Blastn-walking. These 3′ UTR sequences were then aligned to the D. melanogaster and D. pseudoobscura 3′ UTRs using AVID. ### miRNA-sequences {#s4a7} Drosophila miRNA sequences were from \[[@pbio-0030085-b44],[@pbio-0030085-b50],[@pbio-0030085-b51]\] downloaded from Rfam (<>). The 5′ non-redundant set (49 miRNAs) comprised *bantam, let-7, miR-1, miR-10, miR-11, miR-100, miR-124, miR-125, miR-12, miR-133, miR-13a, miR-14, miR-184, miR-210, miR-219, miR-263b, miR-275, miR-276b, miR-277, miR-278, miR-279, miR-281, miR-283, miR-285, miR-287, miR-288, miR-303, miR-304, miR-305, miR-307, miR-309, miR-310, miR-314, miR-315, miR-316, miR-317, miR-31a, miR-33, miR-34, miR-3, miR-4, miR-5, miR-79, miR-7, miR-87, miR-8, miR-92a, miR-9a,* and *miR-iab-4--5p.* Additional miRNAs in the 3′ non-redundant set were *miR-2b, miR-286, miR-306, miR-308, miR-311, miR-312, miR-313, miR-318,* and *miR-6.* ### miRNA shuffles and mutants {#s4a8} For the completely shuffled miRNAs, we shuffled the miRNA sequence over the entire length and required all possible 8mer and 7mer seeds within the first nine bases to have an equal frequency (±15%) to the D. melanogaster 3′ UTRs (i.e., same single genome count). For the 3′ shuffled miRNAs, we shuffled the 3′ end starting at base 10 and required the shuffles to have equal (±15%) pairing energy to a perfect complement and to 10,000 randomly chosen sites. For each miRNA we created all possible 2-nt mutants (exchanging A to T or C, C to A or G, G to C or T, and T to A or G) within the seed (nucleotides 3--6) and chose the one with the closest alignment frequencies to the real miRNA in D. melanogaster 3′ UTRs and in the conserved sequences in D. melanogaster and D. pseudoobscura 3′ UTRs. ### Seed matching and site evaluation {#s4a9} For each miRNA and seed type we found the 5′ match in the D. melanogaster 3′ UTRs and required it to be 100% conserved in an alignment to the D. pseudoobscura ortholog allowing for positional alignment errors of ±2 nt. When searching 7mer to 4mer seeds we masked all longer seeds to avoid identifying the same site more than once. For each matching site we extracted the 3′ adjacent sequence for both genomes, aligned it to the miRNA 3′ end starting at nucleotide 10 using RNAhybrid \[[@pbio-0030085-b35]\], and took the worse energy. Supporting Information {#s5} ====================== Accession Numbers {#s5a} ----------------- The miRNA sequences discussed in this paper can be found in the miRNA Registry (<>). NCBI RefSeq (<>) accession numbers: *bagpipe* (NM_169958), *Brd* (NM_057541), *grim* (NM_079413), *hairy* (NM_079253), *hid* (NM_079412), *lin-14* (NM_077516), *lin-41* (NM_060087), and *Scr* (NM_206443). GenBank (<>) accession numbers: *sickle* (AF460844) and *D. simulans hairy* (AY055843). We thank Ann-Mari Voie for cheerfully producing the large number of transgenic strains used in this work. We are grateful to Marc Rehmsmeier for providing us with the RNAhybrid program prior to publication, to Eric Lai for providing unpublished fly strains, to Aurelio Teleman for comments on the manuscript, and to Lars Juhl Jensen for helpful discussions on the statistics. **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** JB, AS, and SMC conceived and designed the experiments. JB and AS performed the experiments and analyzed the data. JB, AS, RBR, and SMC wrote the paper. Citation: Brennecke J, Stark A, Russell RB, Cohen SM (2005) Principles of microRNA--target recognition. PLoS Biol 3(3): e85. *Brd* : *Bearded* EGFP : enhanced green fluorescent protein miRNA : microRNA *Scr* : *Sex combs reduced* ```{=html} <!-- --> ``` Academic Editor: James C. Carrington : Oregon State University

Compare the gene number of fruitfly (13,000) to human (20,000), and it\'s pretty clear that complexity emerges not just from gene number but from how those genes are regulated. In recent years, it\'s become increasingly clear that one class of molecules, called microRNAs (miRNAs), exert significant regulatory control over gene expression in most plant and animal species. A mere 22 nucleotides long, miRNAs control a cell\'s protein composition by preventing the translation of protein-coding messenger RNAs (mRNAs). When a miRNA pairs with an mRNA, through complementary base pairing between the molecules, the mRNA is either destroyed or is not translated. Hundreds of miRNAs have been found in animals, but functions for just a few have been identified, mostly through genetic studies. Many more functions could be assigned if miRNA targets could be predicted. This approach has worked in plants, because miRNAs and their targets pair through the near perfect complementarity of their base pairs. But the molecules follow different rules in animals---duplexes contain just short stretches of complementary sequence interrupted by gaps and mismatches---which makes predicting miRNA targets a challenge. In a new study, Stephen Cohen and his colleagues at the European Molecular Biological Laboratory in Germany establish basic ground rules for miRNA--mRNA pairing using a combination of genetics and computational analyses, and identify different classes of miRNA targets with distinct functional properties. Although the miRNA is only 22 nucleotides long, its 5′ and 3′ ends seem to have distinct roles in binding. Cohen and colleagues show that miRNA functional targets can be divided into two broad categories: those that depend primarily on pairing to the miRNA\'s 5′ end (called 5′ dominant sites), with varying degrees of 3′ pairing, and those that also need the miRNA\'s 3′ end (called 3′ compensatory sites). Surprisingly, miRNAs can regulate their targets simply by strong pairing with so-called seed sites that consist of just seven or eight bases complementary to the miRNA 5′ end. Target sites with weaker 5′ complementarity need supplemental pairing with the miRNA\'s 3′ end to function. The finding that so little sequence complementarity is needed means that there are many more target sites than had been previously recognized. The miRNA 3′ end, while not essential, is expected to confer some function, since it tends to be conserved in animals---miRNA 3′ ends provide an additional measure of regulatory control by permitting the function of target sites that have only limited complementarity to the miRNA 5′ end. The authors speculate that seed sites might be the first functional sites acquired by protein-coding genes that require repression, and that additional sites might be acquired to promote stronger repression. Based on their experimental results, Cohen and colleagues searched the Drosophila genome for biologically relevant targets, and estimate that the fly has about 100 sites for every miRNA in its genome. Since the fruitfly has anywhere from 96 to 124 miRNAs, that means it has 8,000 to 12,000 target sites (in the 11,000 genes sampled). This indicates that miRNAs regulate a large fraction of protein-coding genes. Of the known animal miRNAs, many regulate critical developmental processes. This new approach to predicting targets should help reveal just how much regulatory control actually flows from these tiny bits of RNA.

For Proust, a taste of cookie was enough to trigger vivid recollections of his childhood, the first of a long string of reveries that he fashioned into his famous memoir *Remembrance of Things Past*. For many animals, too, tastes and smells are evocative and play a crucial role in finding food, allowing them to build on past successes and to learn how to find their next meal. To locate blooming flowers, for example, honeybees rely heavily on scent. They can associate a whiff of an aldehyde, say, with a nectar-filled orchid. Then later they\'ll seek out the same or similar scents. To succeed in the wild, they must be able to distinguish relevant scents at varying concentrations, and within complex milieus of other scents. But to find food in varied conditions and adapt to new situations, they also have to generalize from past experience. Through both physiological and behavioral studies, scientists have investigated the response to smell in a wide range of organisms and have suggested that two key properties of scent-inducing chemicals are the functional class, such as alcohol or aldehyde, and the carbon-chain length. Bees trained to associate a particular chemical with a reward, for example, can then generalize to some extent to other chemicals with the same functional groups or similar carbon-chain lengths. In these situations, bees are surprisingly consistent in both in their behavior (extending their proboscis to an odor previously associated with food) and in their brains (brain activity in smell-processing centers). Each set of data, behavioral and neural, can be thought of as a "code" underlying the bee\'s response: present a scent, and a bee\'s brain and body will tend to react in a certain way.[](#pbio-0030122-g001){ref-type="fig"} {#pbio-0030122-g001} A new study of smell perception in honeybees (Apis mellifera) published in *PLoS Biology* gives a more comprehensive picture of how bees react to a suite of scents and also shows a remarkable correspondence between the codes for the insects\' behavior and brain activity. The researchers, led by Martin Giurfa, first trained bees to associate a specific chemical, such as the alcohol 1-nonanol, with a sucrose reward. Then the researchers tested the bees\' response to a set of other chemicals, varying in carbon-chain length from six to nine, and with four different functional groups: aldehydes, ketones, and primary and secondary alcohols. By watching how often the bees generalized---that is, how often they responded positively to a particular scent when they\'d been trained on another---the researchers could assign perceptual "distances" between pairs of chemicals. Drawing together all these distances, they created a preliminary map of the bees\' "perceptual space," similar to how surveyors measure distances between landmarks to map a landscape. From this comparison they found, for example, that the bees generalized more by functional group than by carbon-chain length. Previously, Giovanni Galizia\'s group, which works closely with Giurfa\'s group, had recorded bees\' brain responses to the same pairs of scents, assigning distances within centers of activity for each scent. Giurfa\'s team compared these two sets of data and found that the perceptual and neural distances correlated well, which suggests there\'s a species-specific code that ties together the insects\' brain and behavior. The brain recordings covered only a quarter of the bees\' main smell-processing center, the antennal lobe. Future studies with new methods of microscopy that visualize more of the brain and which focus on the olfactory message sent by the antennal lobe to higher-order brain centers should only improve our ability to investigate the correlations between brain and behavior, the authors say. Such studies would go even further toward cracking the codes underlying animals\' perception and memory.

Introduction {#s1} ============ Sex differences in recombination were discovered in the first linkage studies on *Drosophila* \[[@pbio-0030063-b01],[@pbio-0030063-b02]\] and *Bombyx* (Tanaka \[1914\] in \[[@pbio-0030063-b03]\]) almost one century ago. However, this observation remains today largely unexplained despite several attempts. Based on very limited observations (see [Table 1](#pbio-0030063-t001){ref-type="table"}), especially of *Bombyx,* in which the female is heterogametic, Haldane \[[@pbio-0030063-b03]\] suggested, as far as "these facts are anything more than a coincidence," that the lower autosomal recombination rate in the heterogametic sex may reflect a pleiotropic consequence of selection against recombination between the sex chromosomes. Later, Huxley \[[@pbio-0030063-b04]\] showed that *Gammarus* males also recombined less than females. He gave the same evolutionary explanation, although he restricted it to cases of a marked sex difference. ::: {#pbio-0030063-t001 .table-wrap} ::: {.caption} ###### Data on Which the Haldane-Huxley Rule is Based :::  Listed are the data available to Haldane \[[@pbio-0030063-b04]\] when he proposed the Haldane-Huxley rule ^a^ *r~m~* represents recombination in males ^b^ *r~f~* represents recombination in females ^c^ Plus and minus symbols indicate the direction of heterochiasmy, and zero indicates achiasmy ::: This conjecture has now been confirmed for achiasmate species (i.e., species in which only one sex recombines) and is referred to as Haldane-Huxley rule: Nei \[[@pbio-0030063-b05]\] showed theoretically that tight linkage should evolve on Y or W chromosomes, and Bell \[[@pbio-0030063-b06]\] compiled a large dataset showing that achiasmy evolved 29--34 times independently, each time with no recombination in the heterogametic sex. However, for heterochiasmate species, three problems with the Haldane-Huxley pleiotropy explanation were discovered \[[@pbio-0030063-b07],[@pbio-0030063-b08]\]. The first problem arose when substantial variation in male-female differences in recombination rate was found between pairs of autosomes within mice \[[@pbio-0030063-b08]\] and *Tribolium* \[[@pbio-0030063-b09],[@pbio-0030063-b10]\], and between genotypes for the same pair of autosomes \[[@pbio-0030063-b11]\]. The second problem was the discovery that hermaphrodite species (the platyhelminth *Dendrocoelum* \[[@pbio-0030063-b12]\] and the plant *Allium* \[[@pbio-0030063-b13]\]) may present strong heterochiasmy between male and female meiosis despite having no sex chromosomes or even sex-determining loci. The third problem was the discovery of species in which the heterogametic sex recombines more than the homogametic one (e.g., in some *Triturus* species) \[[@pbio-0030063-b14]\]. Because of these contradictory observations, variation in heterochiasmy has remained difficult to explain because of the absence of an alternative theory as well as the lack of a clear pattern in the data. In 1969, Nei \[[@pbio-0030063-b05]\] worked out the first "modifier" model to study the evolution of sex differences in recombination, and concluded for autosomes that "the evolutionary mechanism of these sex differences is not known at present." Surveying an updated dataset, Bell \[[@pbio-0030063-b06]\] concluded that "female gametes experience more crossing over among hermaphroditic plants (and perhaps animals), but this is not invariably the case among gonochoric animals (...) certainly (this) has never received any explanation." The idea that heterochiasmy may be explained by a sex rather than by a sex chromosome effect, which was ignored by Haldane because of *Bombyx,* was reconsidered. This led Trivers \[[@pbio-0030063-b15]\] to suggest that, because only males with very good gene combinations reproduce (relative to females, for whom reproduction success is often less variable), they should recombine less to keep intact these combinations. He accounted for exceptions by variation in the regime of sexual selection. The idea was criticized by Burt et al. \[[@pbio-0030063-b16]\], who also questioned the correlations---with an updated dataset---between heterochiasmy and either sex or heterogamety. These authors tried to correlate the level of heterochiasmy with the amount of "opportunity for sex-specific selection," but failed to find an effect. They were tempted to advocate neutrality, but were puzzled by the positive correlation between male and female recombination rate and by evidence showing compensation (e.g., female mice tend to recombine more on the X, as if they were compensating for no recombination in males; similarly, no species is known with achiasmy in both sexes \[[@pbio-0030063-b16]\]). In 1994, Korol et al. \[[@pbio-0030063-b17]\] insisted on a possible role for gametic selection but did not give evidence in favour of this claim. Recently, Lenormand \[[@pbio-0030063-b18]\], using Nei\'s modifier approach, showed that it is very difficult to explain heterochiasmy by sex-specific diploid selection. Rather, a sex difference in selection during haploid phase, or a sex difference in diploid selection on imprinted genes, is a more likely explanation. He predicted that, as far as haploid selection is concerned, the sex experiencing the more intense haploid selection should recombine less. Indeed, when allelic effects interact to determine fitness (i.e., when there is "epistasis," either negative or positive), recombining decreases mean fitness in the population of the next generation \[[@pbio-0030063-b19]\]. This effect occurs because recombination breaks up combinations of genes that have previously been built up by selection. For a given average recombination rate between sexes and for a given average epistasis between male and female haploids, it is always advantageous for the haploid population (male or female) with the greatest absolute value in epistasis to be produced with the lowest amount of recombination. In this way, the "recombination load" that the haploid population is exposed to is minimized. In this paper, we would like to come up with a more quantitative evaluation of the possible role of haploid selection in shaping heterochiasmy. For that purpose, we first updated the dataset of Burt et al. \[[@pbio-0030063-b16]\] on heterochiasmy, focusing on genetic maps that have become available over the last 15 years. We then determined how fast heterochiasmy evolves, in order to measure the amount of phylogenetic inertia on this trait. Finally, we determined whether variables such as gender, heterogamety, or the opportunity for selection in the haploid phase, could explain variation in heterochiasmy. If there is selection with substantial epistasis on some genes during the haploid phase, we expect the sex with the greater opportunity for haploid selection to show less recombination. Alternatively, if selection during the haploid phase is weak or without substantial epistasis, we do not expect it to produce a directional bias in the amount of recombination displayed by either sex. Results/Discussion {#s2} ================== Sex Chromosomes {#s2a} --------------- Heterochiasmy is a fast-evolving trait, and phylogenetic inertia does not satisfactorily explain its distribution. In contrast to achiasmy, we found that heterochiasmy is not influenced by the nature of the sex chromosomes. This is interesting, because it suggests that achiasmy and heterochiasmy are influenced by qualitatively different evolutionary forces, although they seem to differ only quantitatively. It would be useful to determine whether achiasmy evolved to reduce the average recombination rate or to change the relative amount of recombination between the sexes. The two situations may be discriminated by determining whether the homogametic sex in achiasmate species tends to recombine more than in closely related chiasmate species. Evidence for such compensation would indicate that achiasmy did not evolve to reduce the average recombination rate. In the absence of such compensation, however, achiasmy may simply reflect selection for tight linkage. In such a situation, we propose that Haldane-Huxley rule may be caused by the converse argument to the one previously considered: The presence of achiasmy only in the heterogametic sex may reflect selection to maintain nonzero recombination rate on X or Z chromosomes in the homogametic sex. In species in which the average autosomal recombination rate is selected against (i.e., towards a lower equilibrium value), loss-of-function (recombination) mutations with an effect restricted to one sex may spread only if they affect the heterogametic sex, because mutations suppressing recombination in the homogametic sex completely suppress recombination on the X or Z chromosome. The same argument applies to XO species and may explain why achiasmy is associated only with the heterogametic sex. In addition, this hypothesis does not require the existence of genes suppressing recombination between the sex chromosomes with autosomal pleiotropic effects. Under this hypothesis, there is no reason to find an effect of the presence of heteromorphic sex chromosome on the amount of heterochiasmy, as originally envisioned by Haldane and Huxley. Overall, this hypothesis would explain why heterochiasmy and achiasmy differ qualitatively and why we do not observe any effect of sex chromosomes on heterochiasmy. Heterochiasmy in Animals {#s2b} ------------------------ In animals, male-female dimorphism in haploid selection may also contribute to heterochiasmy. In general, there is no female haploid phase in animals, because meiosis is completed only at fertilisation. As far as at least some genes are expressed and under selection during the male haploid phase, this would tend to bias towards tighter linkage in males. Sets of genes responsible for male-specific meiotic drive systems would be good candidates and are often found in tight linkage. Measuring the opportunity for haploid selection in animals may be possible within some groups. Imprinting may, however, act as a confounding effect in many groups of animals while trying to measure the opportunity for "haploid" selection. Within-species comparisons of imprinted regions or of regions with sex-specific recombination using high-resolution maps \[[@pbio-0030063-b20]\] may be more fruitful to discriminate among potential causes of heterochiasmy in animals. In particular, there is evidence in humans that the reduction in crossing-over associated with imprinting is in the direction that theory predicts, even if this pattern is consistent with other explanations \[[@pbio-0030063-b21]\]. Finally, understanding exceptions within groups (e.g., male marsupials, contrarily to most mammals, recombine more than females of the species \[[@pbio-0030063-b22]\]) may also shed light on the different hypotheses. Heterochiasmy in Plants {#s2c} ----------------------- We found that plant heterochiasmy is correlated with the opportunity for male and female haploid selection. Female meiosis tends to exhibit lower recombination rates relative to male meiosis when selection is intense among female gametophytes (e.g., in Pinaceae) or mild among male gametophytes (e.g., in highly selfing species). This pattern is expected if heterochiasmy is determined by the relative magnitude of haploid selection in male and female individuals. Finding a pattern consistent with this general population genetic prediction is, of course, not firm evidence that male-female dimorphism in haploid selection is the evolutionary force generating heterochiasmy. Other correlates of selfing rates might have to be closely examined \[[@pbio-0030063-b23]\]. However, we consider this explanation the most parsimonious so far. Our finding provides, therefore, the first empirical evidence for a theory explaining male-female differences in the amount of recombination and contributes to our understanding of contradictory observations that have puzzled geneticists for almost a century. It also indicates that the amount of recombination may be shaped by indirect selection, and, therefore, corroborates theories based on selection and variation for the evolution of sexual reproduction. Materials and Methods {#s3} ===================== {#s3a} ### An extended dataset {#s3a1} We measured heterochiasmy as the log of the male/-to-female ratio (*ρ*) of autosomal recombination rate measured either with chiasma number or map length. We log-transformed the ratio to avoid bias due to measurement error in the denominator. Chiasma-count data for different species were compiled by Burt et al. \[[@pbio-0030063-b16]\], and we used their dataset, adding a few recent studies. We compiled genetic map data and linkage studies in animals and plants for which both a male and a female map were available. Only homologous fragments (i.e., between shared markers) in male and female maps were considered (especially in low-resolution maps). Heterochiasmy data were available for 107 species, with 46 sets of data based on genomic maps ([Table 2](#pbio-0030063-t201){ref-type="table"}). ::: {#pbio-0030063-t201 .table-wrap} ::: {.caption} ###### Dataset Pooled by Species with Levels of Phylogenetic Grouping Used in the Analysis :::  Note that references given in Burt et al. \[[@pbio-0030063-b17]\] were not repeated here ^a^ K, kingdom. Numeric indicators in this column are: 1, Animalia; 2, Plantae ^b^ P, phylum. Numeric indicators in this column are: 1, Arthropoda; 2, Chordata; 3, Embryophyta; 4, Platyhelminthes ^c^ C, class. Numeric indicators in this column are: 1, Actinopterygii; 2, Amphibia; 3, Magnoliopsidae (subclass asterids); 4, Aves; 5, Coniferopsida; 6, Insecta; 7, Liliopsida; 8, Mammalia; 9, Magnoliopsidae (subclass rosids); 10, Trematoda; 11, Turbellaria ^d^ Data refers to linkage map (LM) or chiasma count (CC) ^e^ Male and female indicate the value for the chiasma count or map length for each sex ^f^ Ratio refers to male/female recombination rate ^g^ *V~sc~* refers to the presence or absence of sex chromosome (see [Materials and Methods](#s3){ref-type="sec"}, "Sex chromosome effect") ^h^ Data were obtained from maps DBNordic2 and NIAIJapan (<>) \[[@pbio-0030063-b54],[@pbio-0030063-b55]\] ND, no data ::: ::: {#pbio-0030063-t202 .table-wrap} ::: {.caption} ###### Continued :::  ::: ### Phylogenetic inertia {#s3a2} Heterochiasmy may evolve so slowly that there is important phylogenetic inertia. Alternatively, it may be so fast-evolving that the amount of heterochiasmy takes on nearly independent values among related species. In the same way, heterochiasmy may be so variable between genotypes within a species that it may be difficult to measure and irrelevant to analyse species specific effects. In order to get a picture of phylogenetic inertia on heterochiasmy, we estimated the phylogenetic autocorrelation of *ρ* using Moran\'s *I* spatial autocorrelation statistic \[[@pbio-0030063-b24]\]. When standardized, values of Moran\'s *I* vary from −1 to 1. Positive values indicate that heterochiasmy is more similar than random within a taxonomic level, whereas negative values indicate that it is more different. Because a few species had multiple estimates of heterochiasmy, we also estimated the within-species correlation. The resulting correlogram is shown in [Figure 1](#pbio-0030063-g001){ref-type="fig"}. We found that heterochiasmy is a fast-evolving trait: Genotypes tend to be correlated within a species (I/I~max~ = 0.38, *p* = 7.9%), but this correlation is lower among species within genera (I/I~max~ = 0.18, P-value = 13%), and very low when comparing genera within families (I/I~max~ = 0.039, *p* = 63%). This pattern is very different from the one observed for highly autocorrelated traits using the same method (for instance, mammalian body size \[[@pbio-0030063-b25]\]). This analysis indicates that there is very little phylogenetic inertia overall on heterochiasmy, but that the species level is appropriate for our dataset. However, this low level of inertia may nevertheless inflate type-I error while testing the effect of independent variables on heterochiasmy. In order to avoid this problem, we tested the association between different variables and heterochiasmy using a generalized estimating equations linear model correcting for the full phylogeny (see below) \[[@pbio-0030063-b26]\]. {#pbio-0030063-g001} ### Sex chromosome effect {#s3a3} For each species, we reported the presence of sex chromosomes. We defined the variable *V~sc~* with the following values: −1 for XY/XX species, −1/2 for XO/XX or XY/XX without pseudoautosomal regions (marsupials), 0 for species without sex-chromosomes, and +1 for ZZ/ZW species. We distinguished the −1 and −1/2 cases to reflect the fact that, in the latter, recombination does not occur between sex chromosomes, so we expect a lower current selection pressure to suppress recombination. Under the Haldane-Huxley hypothesis, the presence of sex chromosomes is supposed to favour reduced recombination rate in the heterogametic sex. We therefore expect a positive effect of the variable *V~sc~* on *ρ*. We did not find such an effect in animals or plants (the linear effect of *V~sc~* on *ρ* is not significantly different from zero \[*p* = 0.75 in animals and *p* = 0.52 in plants\], assuming species were independent), and this result is unchanged if the −1 and −1/2 cases are not distinguished. Given this negative result, there was no need to do a phylogenetic correction. ### Gametic selection {#s3a4} In animals from our dataset, there is no female haploid phase because the completion of meiosis occurs only at fertilisation (sperm triggers the end of meiosis). In male gametes, very few genes are expressed, and sperm phenotype is determined mostly either by the diploid genotype of the paternal tissue or by its mitochondrial genome. Imprinted genes, which can also affect the evolution of heterochiasmy \[[@pbio-0030063-b18],[@pbio-0030063-b21]\], may be as numerous as haploid-expressed genes and act as a confounding factor while evaluating the "opportunity" for male or female gametic selection. As a consequence, we did not attempt to evaluate the opportunity for haploid selection in animals. Rather, we focused on plants, in which there is both a male (pollen) and female (ovule) haploid phase and during which many genes are expressed (e.g., as many as 60% of genes may be expressed in the male gametophyte \[[@pbio-0030063-b27],[@pbio-0030063-b28]\]). In order to evaluate the effect of the "opportunity for selection" for male haploid phase on *ρ,* we used selfing rate as an indirect variable estimating the degree of pollen competition. We assume that with high selfing rates, there is less genetic variation among competing pollen grains and, therefore, less scope for haploid selection. We defined *V~m~* (the degree of male gamete competition in plants) using three values depending on the amount of selfing: 0 for dioecious, self-incompatible or largely outcrossing (less than 5% selfing reported) species; 1 for species exhibiting low selfing rates (less than 30% reported); and 2 for other species. We used these three broad categories to reflect the fact that selfing rate is often variable within species and that it is often measured indirectly and with low precision. We therefore expect a positive effect of the variable *V~m~* on *ρ* if the opportunity for male gametic selection favours smaller *ρ* values, as predicted by the modifier model \[[@pbio-0030063-b18]\]. We tested this effect using the 57 species for which we were able to estimate *V~m~* ([Table 3](#pbio-0030063-t003){ref-type="table"}). We used a linear model in R \[[@pbio-0030063-b29]\] assuming that all species are either independent or phylogenetically related. In the latter case, we used a generalized estimating equations linear model \[[@pbio-0030063-b26]\] with a plant phylogenetic tree to the family level using data from Davies et al. \[[@pbio-0030063-b30]\], and several calibration points, including the *Picea*/*Pinus* divergence approximately 140 million years ago \[[@pbio-0030063-b31]\], that are not included in the Davies et al. dataset. We found an effect in the right direction with or without correcting for the phylogeny (linear effect of *ρ* on *V~m~*, *p* \< 0.0002 in both cases, [Figure 2](#pbio-0030063-g002){ref-type="fig"}). The fact that selfing plants exhibit higher recombination rates than their outcrossing relatives has been mentioned previously in the literature \[[@pbio-0030063-b32],[@pbio-0030063-b33]\]. However, in most cases, recombination was measured only in male meiosis. It would be valuable to reexamine this trend in the light of our results that recombination in male meiosis is typically greater than in female meiosis among selfers. {#pbio-0030063-g002} ::: {#pbio-0030063-t003 .table-wrap} ::: {.caption} ###### Plant Species Used to Test the Effect of Male and Female Opportunity for Selection :::  ^a^ Ratio refers to male-to-female recombination rate LM, linkage map; CC, chiasma count; *n,* haploid number of chromosomes; *V~m~*, measure of male opportunity for haploid selection; *V~f~*, measure of female opportunity for haploid selection ::: In order to evaluate the effect of the "opportunity for selection" during the female haploid phase on *ρ* in plants, we contrasted angiosperms with gymnosperms. In angiosperms, ovules do not compete much with each other on a mother plant, because resource accumulation starts after fertilisation (i.e., during fruit development in the diploid phase). In *Pinus* (three species in our dataset; see [Table 2](#pbio-0030063-t201){ref-type="table"}), male meiosis, female meiosis, and pollination occur in the year prior to fertilisation, but the pollen tube stops growing until the next spring, while the female gametophytes continue to accumulate resources and compete with each other over the course of the year. The same situation occurs in *Picea,* although the period between female meiosis and fertilisation is only 2--3 mo \[[@pbio-0030063-b34]\]. Perhaps more importantly, the endosperm (which is the organ managing resources for the zygote) is haploid in Pinaceae, in contrast to the double fertilisation that occurs in angiosperms to produce at least a diploid (typically triploid) endosperm \[[@pbio-0030063-b35],[@pbio-0030063-b36]\]. We therefore expect that *ρ* should be greater in Pinaceae, compared to angiosperms. We assigned *V~f~* (the degree of female gamete competition in plants) the values 1 for gymnosperms and −1 for angiosperms. We expected a positive effect of the variable *V~f~* on *ρ* according to the modifier model. An effect in the right direction was indeed detected (linear effect of *V~f~* on *ρ, p* = 0.011 and *p* = 0.0001, with and without correcting for the phylogeny as above, respectively; see [Figure 2](#pbio-0030063-g002){ref-type="fig"}). We thank G. Besnard, J. Britton-Davidian, J.-B. Ferdy, S. Glémin, L. D. Hurst, P. Jarne, O. Judson, M. Kirkpatrick, S. P. Otto, J.-M. Prosperi, and C. Vosa for helpful comments, information, and stimulating discussions. This study was supported by the Centre National de la Recherche Scientifique and French Ministry of Research. **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** TL and JD conceived and designed the experiments, performed the experiments, analyzed the data, contributed reagents/materials/analysis tools, and wrote the paper. Citation: Lenormand T, Dutheil J (2005) Recombination difference between sexes: A role for haploid selection. PLoS Biol 3(3): e63.

Introduction {#s1} ============ Engineering a system to emulate a particular behaviour can be an extremely informative approach to systems biology \[[@pbio-0030064-b01],[@pbio-0030064-b02],[@pbio-0030064-b03],[@pbio-0030064-b04],[@pbio-0030064-b05]\]. Even if the natural biochemical interactions are well characterized, it remains a considerable challenge to reconstruct a physical system with the appropriate behaviour. Step-by-step reconstruction allows theoretical assumptions and models to be refined. Not only is complexity reduced by removing the context of the whole organism, but by reconstitution of pattern formation from purified substances (in this case, RNA polymerases, ribonucleotides, and a translation extract) the sufficiency of a proposed mechanism of pattern formation can be demonstrated. In this work, our primary aim was the development of an in vitro system that allows the careful buildup of complex networks under controlled conditions. To demonstrate the usefulness of such a system, we decided to reconstruct a developmental pattern-formation program based on the formation of a gradient of a transcription activator---a "morphogen"---and to link it to a network of transcription repressors. In a sense, we set out to design a patterning system similar to chemical reaction-diffusion systems (see below). However, the use of components such as transcription activators and repressors in an in vitro transcription-translation system made the system significantly closer to a biological system, albeit highly simplified when compared to the complexity of eukaryotic transcription \[[@pbio-0030064-b06]\]. Patterning systems can be thought of as belonging to one of two principal types: First, there are systems with homogeneous initial conditions that self-organise after early random symmetry-breaking events (\[[@pbio-0030064-b07],[@pbio-0030064-b08]\]; for a biological example, *Fucus,* see \[[@pbio-0030064-b09]\]). Second, there are systems with initial localisation of the components, which can form concentration gradients of activities from their respective sources \[[@pbio-0030064-b10],[@pbio-0030064-b11]\]. The first class of patterning system, involving reaction-diffusion from initially homogeneous conditions, was first proposed by Turing in 1952 \[[@pbio-0030064-b07]\] and was further developed by Meinhardt and Gierer in the 1970s, in their model of patterning with short-range autoactivator and long-range lateral inhibitor \[[@pbio-0030064-b08]\]. Although there are many likely biological candidates for such activator-inhibitor systems (reviewed in \[[@pbio-0030064-b12]\]), none has been reengineered from first principles. In contrast, significant progress has been made in reconstituting purely chemical reactions that self-organise, such as the Belousov-Zhabotinski reactions \[[@pbio-0030064-b13],[@pbio-0030064-b14],[@pbio-0030064-b15]\] and Turing-type, Meinhardt-Gierer (M-G), and oscillatory reactions \[[@pbio-0030064-b16],[@pbio-0030064-b17],[@pbio-0030064-b18],[@pbio-0030064-b19],[@pbio-0030064-b20]\]. In a more biological context, similar spatiotemporal patterns, consisting of propagating concentration waves, have been modelled for a glycolytic enzyme oscillator in yeast \[[@pbio-0030064-b21]\]. In the second class of patterning system, where the initial conditions are nonhomogeneous, patterning begins with an asymmetry: a localised source of morphogen (shape-defining molecules that form a concentration gradient and function in a concentration-dependent manner so as to determine positional information in a patterning field \[[@pbio-0030064-b22]\]; reviewed in \[[@pbio-0030064-b23],[@pbio-0030064-b24],[@pbio-0030064-b25]\]). The question of how morphogen gradients are formed and maintained is still a matter of keen debate and study \[[@pbio-0030064-b26],[@pbio-0030064-b27],[@pbio-0030064-b28]\], with many proposed mechanisms (reviewed in \[[@pbio-0030064-b29],[@pbio-0030064-b30]\]). In the simplest case, a stable gradient could be formed by passive diffusion \[[@pbio-0030064-b31]\] and uniform degradation, although it has been suggested that enhanced morphogen degradation near the source leads to increased robustness against morphogen fluctuations during patterning \[[@pbio-0030064-b32]\]. Also, "ligand trapping" by the receptor for a morphogen can have significant effects on the shape of a morphogen gradient \[[@pbio-0030064-b33]\], as in the case of Torso diffusing in the extracellular space surrounding the *Drosophila* oocyte \[[@pbio-0030064-b34]\]. In most cases in metazoa, morphogens define patterns over fields of many cells (reviewed in \[[@pbio-0030064-b30],[@pbio-0030064-b35]\]), but there is one special case in embryonic development that has been particularly well studied, in which a morphogen gradient operates in a single-celled, multinuclear syncytium: the early *Drosophila* embryo \[[@pbio-0030064-b10],[@pbio-0030064-b11]\]. In this system, early patterning is mediated by maternal morphogen factors, which are thought to diffuse and form gradients to guide patterning within a large multinuclear cell (reviewed in \[[@pbio-0030064-b36],[@pbio-0030064-b37],[@pbio-0030064-b38]\]). After egg deposition, an embryo forms a segmentation pattern within 3 h, under the influence of a hierarchical sequence of gene expression interactions involving gap genes, pair-rule genes, and segment polarity genes \[[@pbio-0030064-b39]\]. Maternal elements---in particular, the morphogen Bicoid---guide this process, setting distinct initial conditions. Work by Driever and Nüsslein-Volhard \[[@pbio-0030064-b10],[@pbio-0030064-b11]\] demonstrated that Bicoid protein possessed three characteristics of a classic morphogen: (i) a localised source of cytoplasmic activity (through *bicoid* RNA transport to the anterior pole of the egg, involving microtubules and maternal genes \[[@pbio-0030064-b40],[@pbio-0030064-b41],[@pbio-0030064-b42],[@pbio-0030064-b43],[@pbio-0030064-b44]\]); (ii) formation of a concentration gradient from the source; and (iii) concentration-dependent activity that determines positional information within the gradient (reviewed in \[[@pbio-0030064-b38]\]). Bicoid has at least two functions that contribute to its function as a morphogen: transcription activation and translation inhibition. Acting as a transcription factor, Bicoid can activate a number of downstream gap genes, including *hunchback, knirps, giant,* and *Krüppel,* whose products cross-react in a complex and mainly repressive interaction network to modulate each other\'s expression (reviewed in \[[@pbio-0030064-b37]\]; modelled in \[[@pbio-0030064-b45],[@pbio-0030064-b46]\]). However, Bicoid does not simply function independently as a morphogen at the top of the gap gene hierarchy. Although Bicoid is responsible for anterior expression of zygotic Hunchback \[[@pbio-0030064-b47]\], it actually requires maternally expressed Hunchback as a cofactor to function anteriorly \[[@pbio-0030064-b48]\]. Meanwhile, in the posterior, the maternal *hunchback* mRNA is initially translationally repressed by the posterior determinant Nanos \[[@pbio-0030064-b49]\]. Moreover, the terminal gap genes *tailless* and *huckebein* are activated independently but serve to repress zygotic gap gene expression at the poles of the embryo, thus influencing patterning \[[@pbio-0030064-b50],[@pbio-0030064-b51],[@pbio-0030064-b52]\]. In its other role, as a translation inhibitor of *caudal* mRNA, Bicoid initially inhibits the uniformly expressed mRNA to form a concentration gradient of the protein Caudal \[[@pbio-0030064-b53],[@pbio-0030064-b54]\]. Thus, the Caudal gradient \[[@pbio-0030064-b55]\] is essentially the inverse of the Bicoid gradient, and Caudal functions as a transcription activator in the posterior of the egg, further influencing expression of the gap gene network. Some of the important interactions in the gap gene network are shown in [Figure 1](#pbio-0030064-g001){ref-type="fig"}A, although it should be noted that this overview is an oversimplification and does not consider differences in maternal and zygotic factor expression over time, nor does it consider all of the factors involved. ![Gene Circuits and Chambers\ (A) Principal interactions in the *Drosophila* gap gene network, modelled after \[[@pbio-0030064-b37]\]. Relative levels and distributions of Hunchback (Hb), Giant (Gt), Krüppel (Kr), Knirps (Kni), Bicoid (Bcd), and Caudal (Cad) shown from anterior (left) to posterior (right). Green arrows indicate activation, red T-bars repression.\ (B) Artificial gene network design, with transcription activators T7 and SP6 polymerases, and zinc finger repressors A, B, and C. Genes are immobilised on paramagnetic beads, and T7 forms a directional concentration gradient.\ (C) Principal interactions in a simple designed network.\ (D) Transcription-translation chamber. Genes for repressor A are localised at the "poles," whereas B and C are ubiquitous. Gel slabs 4--6 have been excised, exposing the magnets below, illustrating gel dissection for Western blot analysis.\ (E) Normalised Western data for four replicate chambers, showing mean levels of A, B, and C after 20 min (± One standard deviation).\ (F) Sample Western blot from the four-replicate experiment.](pbio.0030064.g001){#pbio-0030064-g001} From the outset, we chose to model our system around elements of the *Drosophila* gap gene network ([Figure 1](#pbio-0030064-g001){ref-type="fig"}A), because we could study many elements of morphogenesis, such as gradient formation and the sufficiency of cross-repression for setting pattern boundaries, without the need for considering multiple cells, membrane-bound receptors, and cell-to-cell interactions. Using an in vitro model, we wanted to address the following questions about patterning. First, how easy is it to generate an expression pattern in a gradient, using a diffusing activator from a localised source? Second, one of the outstanding issues in the field is to what extent correct positioning of the gap protein domain boundaries is specified by maternal morphogen gradients and by cross-repression between gap genes; a recent model suggests that repression is crucial for patterning and that threshold-dependent interpretation of the maternal morphogen concentration is not sufficient \[[@pbio-0030064-b56]\]. We therefore wanted to test the effect of transcription repression on pattern formation directly, by progressively adding more repression interactions in a designed gene network. Third, uniform degradation of a diffusing morphogen is often assumed to account for steady-state gradient formation, so we set out to test the effects of adding controlled degradation to an in vitro patterning system. Finally, we wanted to use our model to see how the scale and pattern of the system are affected by the relative rates of diffusion of individual components, and whether nonuniform diffusion of activators and inhibitors are required to form a pattern. Results {#s2} ======= Design of the Network and Development of the In Vitro Experimental Platform {#s2a} --------------------------------------------------------------------------- We began by designing a simplified gene network to emulate elements of the *Drosophila* gap gene system ([Figure 1](#pbio-0030064-g001){ref-type="fig"}). The aim was to develop a fully synthetic approach in which protein analogues completely unrelated to *Drosophila* would emulate some of the transcription activation and repression interactions thought to be important for patterning in the gap gene system. As activators, two sequence-specific polymerases were employed, T7 and SP6, that have been used successfully by others to engineer gene networks \[[@pbio-0030064-b57]\]. These two polymerases bind to their respective consensus DNA recognition sites to initiate transcription, and thus represent an extremely simplified mode of transcription when compared to the multifactor complexes required for eukaryotic transcription (reviewed in \[[@pbio-0030064-b06]\]). T7 polymerase was chosen to be the "master activator" of the system and, by crude analogy, was expected to carry out some of the functions of Bicoid, namely transcription activation of downstream members in the gap gene hierarchy, in a concentration-dependent manner, from a localised source \[[@pbio-0030064-b10],[@pbio-0030064-b11]\]. In *Drosophila*, the Bicoid morphogen gradient initially controls the shape of the Caudal protein gradient through translational repression of maternal mRNA \[[@pbio-0030064-b53],[@pbio-0030064-b54]\], although later Caudal expression is under zygotic transcriptional control. To simplify this level of complexity, we decided to model Caudal transcription activation by a second gradient of T7 polymerase, from the opposite pole to our primary "Bicoid" gradient (compare [Figure 1](#pbio-0030064-g001){ref-type="fig"}A and [1](#pbio-0030064-g001){ref-type="fig"}C). Residual activation between other members of the gap gene members (e.g., [Figure 1](#pbio-0030064-g001){ref-type="fig"}A, Hunchback activating Krüppel or Krüppel activating *knirps*) was modelled nonexplicitly by having a homogeneous distribution of a second sequence-specific transcription activator, SP6 polymerase ([Figure 1](#pbio-0030064-g001){ref-type="fig"}A and [1](#pbio-0030064-g001){ref-type="fig"}C). Repression interactions between gap gene members were modelled by constructing three site-specific repressors to represent the repressor activities of Hunchback, Giant, and Krüppel ([Figure 1](#pbio-0030064-g001){ref-type="fig"}B, repressors A, B, and C, respectively). The repressors were derived from artificial zinc finger DNA-binding domains that were engineered by phage display \[[@pbio-0030064-b58]\]. Variable gene-repression networks were therefore constructed by placing binding sites for the repressors in the appropriate gene expression constructs (see for comparison [Figure 2](#pbio-0030064-g002){ref-type="fig"}). Repressor sites were either overlapping with the polymerase initiation sites (demonstrated to be effective using triplex-forming oligonucleotides \[[@pbio-0030064-b59]\]), or immediately downstream of the initiation sites ([Figure 2](#pbio-0030064-g002){ref-type="fig"}). Therefore, by changing the identity of the repressor sites, the connectivity of the network could readily be modified to add or remove cross-repressive interactions. {#pbio-0030064-g002} Key to the strategy was the development of an experimental platform in which to model the volume of the *Drosophila* embryo and to carry out artificial gene network reactions. Plastic chambers were therefore developed, constructed over printed templates on petri dishes ([Figure 1](#pbio-0030064-g001){ref-type="fig"}D; see also [Materials and Methods](#s4){ref-type="sec"} and [Protocol S1](#sd001){ref-type="supplementary-material"}). The chambers were filled with a customised transcription-translation mixture, allowing gene network reactions to be carried out in situ. Additionally, small bar magnets were fixed under the chamber to create a spatially defined array, over which paramagnetic beads could be dispensed. By coating such streptavidin-linked beads with biotinylated PCR products, specific gene network constructs were tethered and sublocalised on the array. Furthermore, ultra-low melting point agarose was added to the transcription-translation mixture, both to increase viscosity and to allow the reaction to be "fixed" in a gelling step at 4 °C. Through fixing, gel slices could be excised and assayed by Western blotting against FLAG-epitope tags on the expressed proteins. This design therefore enabled quantification of each output species present in the network (genes A, B, and C) for any given chamber position and time point. The chambers were constructed such that the system components could be pipetted wherever desired, either homogeneously mixed with the transcription-translation mix or pipetted at defined loci, such as at the edges or "poles" of the chamber. As described above, these system components included the two soluble, purified transcription activators (T7 and SP6 polymerases) and three bead-tethered zinc finger transcription-repressor constructs, A, B, and C, which were themselves activated by the polymerases and could cross-repress each other (see [Figure 1](#pbio-0030064-g001){ref-type="fig"}). Positional information was therefore introduced into the artificial system in two ways. First, by injecting purified T7 polymerase at either pole of the chamber, the Bicoid activator distribution could be transiently modelled. Second, beads coated with different gene network constructs (genes A, B, and C), could be fixed at different positions on the magnetic array (see [Figure 1](#pbio-0030064-g001){ref-type="fig"}D). For example, repressor A genes were placed solely at the chamber edges ("poles") to model, loosely, the distribution of embryonic Hunchback activity. This part of the model is a significant oversimplification: Although Hunchback is eventually expressed in two domains, one anterior and one posterior \[[@pbio-0030064-b48]\], it is only expressed anteriorly in the early embryo. Furthermore, while maternal *hunchback* mRNA is evenly distributed, the anterior domain of Hunchback protein forms through zygotic translation and transcription activation (under the control of Bicoid), while maternal RNA is translationally repressed posteriorly, under the influence of Nanos \[[@pbio-0030064-b49],[@pbio-0030064-b60],[@pbio-0030064-b61]\]. In the later phase of *hunchback* regulation, the posterior Hunchback domain forms through a combination of factors, including activation by Tailless \[[@pbio-0030064-b62],[@pbio-0030064-b63]\] and *hunchback* autoactivation \[[@pbio-0030064-b64]\]. To complete the model, the genes for repressors B and C were distributed uniformly throughout the chamber on magnetic beads, to represent the ubiquitous distribution of genes in nuclei, throughout the embryo. Therefore, the spatial expression of genes B and C, who represent the downstream gap gene members *giant* and *Krüppel* (see [Figure 1](#pbio-0030064-g001){ref-type="fig"}A and [1](#pbio-0030064-g001){ref-type="fig"}C), was dependent on differential activation by the T7 polymerase gradient and crossregulation between gene network members. However, it should be noted that the initial expression of these genes in *Drosophila* may not be achieved by crossregulation, because localised mRNA is seen before any protein is detectable (Krüppel and Giant are only detected unambiguously in early cycle 13 \[[@pbio-0030064-b46]\]). Pattern Generation In Vitro from a Transcription Network {#s2b} -------------------------------------------------------- In our first experiments, we constructed a simple, minimal network with sequential transcription activation and repression (see [Figure 1](#pbio-0030064-g001){ref-type="fig"}B and [1](#pbio-0030064-g001){ref-type="fig"}C). Although this basic system is far less complex than the *Drosophila* gap gene system, it was indeed sufficient to generate a crude target behaviour (see [Figure 1](#pbio-0030064-g001){ref-type="fig"}E and [1](#pbio-0030064-g001){ref-type="fig"}F). Qualitatively, the pattern can be explained as follows, Gene A is activated by T7 polymerase from its source at either end of the chamber, and so is expressed most highly at these poles. Gene B is similarly activated, and so it is also less expressed in the middle of the cell. However, since gene B is repressed by protein A, its levels are also reduced at either pole. Finally, Gene C is activated by a ubiquitous SP6 polymerase, but is repressed by proteins A and B, and is consequently centrally distributed. Progressing from the minimal network, we explored systems with a variety of connectivities ([Figure 3](#pbio-0030064-g003){ref-type="fig"}), including a control network without repression interactions ([Figure 3](#pbio-0030064-g003){ref-type="fig"}A), and one with extensive mutual or feedback interactions ([Figure 3](#pbio-0030064-g003){ref-type="fig"}C). These were compared with the original network ([Figure 3](#pbio-0030064-g003){ref-type="fig"}B) in a series of time-course experiments. Generally, we observed that the more repression interactions in a system, the lower the overall protein production but the "sharper" the pattern. {ref-type="fig"}).\ (C) Mutual repression network with extensive negative interactions between species. Adding protease ("+ Degradation") creates weak but time-stable patterns for both the "Repressed" and "Mutual" networks (35 versus 90 min). Quantitated graphs for the above data are available in [Protocol S1](#sd001){ref-type="supplementary-material"}.](pbio.0030064.g003){#pbio-0030064-g003} All patterns degenerated to a significant degree by 60 min, indicating the transience of the system ([Figure 3](#pbio-0030064-g003){ref-type="fig"}, 60 min). However, by adding Factor Xa protease, we were able approximately to match levels of production and degradation. Thus, the outputs became sharper, weaker, and more dynamically stable, hardly varying between 35 and 90 min ([Figure 3](#pbio-0030064-g003){ref-type="fig"}B and [3](#pbio-0030064-g003){ref-type="fig"}C, "+ Degradation"). Computer Exploration of Parameter Space {#s2c} --------------------------------------- To study parameter sensitivity in our system more comprehensively, we constructed a computer model of the chamber and networks ([Protocol S1](#sd001){ref-type="supplementary-material"}). A series of coupled differential equations were simulated, yielding expression levels of gene products A, B, and C, for the three different levels of network connectivity coded by our gene network designs ([Figure 4](#pbio-0030064-g004){ref-type="fig"}). The modelling was carried out at two scales, that of our experimental system (18 mm long) and that of a *Drosophila* embryo (500 μm long). As in the experimental system ([Figures 3](#pbio-0030064-g003){ref-type="fig"} and [4](#pbio-0030064-g004){ref-type="fig"}A), the simulations revealed a large difference of pattern between the unrepressed and repressed systems. The patterns are more similar, however, between the simple and mutually repressed networks ([Figure 4](#pbio-0030064-g004){ref-type="fig"}B) but, as in our in vitro experiments (e.g., see [Figure 3](#pbio-0030064-g003){ref-type="fig"}B and [3](#pbio-0030064-g003){ref-type="fig"}C, 15 min), adding feedback repression makes the peaks better resolved. {ref-type="fig"}, showing outputs for proteins A (cyan), B (magenta) and C (dark blue).\ (A) Quantitated Western blot data from [Figure 3](#pbio-0030064-g003){ref-type="fig"}, after 25 min.\ (B) Simulation data plotted as percentage of total output protein against chamber length, at the chamber (18-mm) or *Drosophila* (0.5-mm) scale. The model is described in full in [Protocol S1](#sd001){ref-type="supplementary-material"}.](pbio.0030064.g004){#pbio-0030064-g004} Next, we explored the sensitivity of the simple repression network to diffusion parameters ([Figure 5](#pbio-0030064-g005){ref-type="fig"}). Generally, we found that the A- and C-peaks were least sensitive to parameter variation, as there are no antagonistic forces against their formation ([Figure 5](#pbio-0030064-g005){ref-type="fig"}A and [5](#pbio-0030064-g005){ref-type="fig"}C). By contrast, gene B is more sensitive: Twin-peak formation correlates with the relative diffusion ratios of activator (T7) and other mRNA/protein components ([Figure 5](#pbio-0030064-g005){ref-type="fig"}B). To generate "target behaviour," the activator must diffuse more rapidly than other species, within certain limits (approximately 5- to 50-fold faster for [Figure 5](#pbio-0030064-g005){ref-type="fig"}B). However, the absolute values (and ratios) for diffusion merely alter the timing of the transient B-peak formation in a system of a given scale. For simplicity, only the 0.5-mm system is illustrated in [Figure 5](#pbio-0030064-g005){ref-type="fig"}; similar conclusions were drawn from the 18-mm scale model. {#pbio-0030064-g005} As in our chamber experiments, the computer model output became more "time-stable" by adding a degradation element ([Figure 5](#pbio-0030064-g005){ref-type="fig"}D). *Drosophila* may exploit such mechanisms to some extent, since Bicoid protein degrades in vivo (t~1/2~ ≤ 1800 s \[[@pbio-0030064-b11]\]), although *bicoid* mRNA is unusually stable \[[@pbio-0030064-b65]\]. Discussion {#s3} ========== To develop a fully synthetic approach that will emulate elements of gap gene expression domain pattern formation, we created an in vitro transcription-translation system that allows flexible spatial gene network construction. The system is widely applicable, allowing control over factors such as localisation or diffusion, and the ability to add or remove components at will. Repressive Interactions and Pattern Formation {#s3a} --------------------------------------------- A basic aim of our system was to see whether we could engineer a gradient of protein expression, using a diffusing activator from a localised source. We found this task straightforward in the transcription-translation chambers, using injected T7 polymerase, and this led us to try more complex expression-repression interactions. We constructed three types of gene network---unrepressed, simple repressed, and mutually repressed (see [Figure 3](#pbio-0030064-g003){ref-type="fig"})---representing different levels of network connectivity. Generally, in both our in vitro and computer models, we found that adding more connections resulted in better-resolved patterning, although the absolute levels of gene expression were reduced. Our in vitro results are essentially qualitative at this stage, but appear to agree with the observations of others---that crossrepression is crucial for the control of patterning boundaries \[[@pbio-0030064-b56]\]. It will be interesting to learn whether more sophisticated elements can be engineered into the system to begin to emulate the more complex features of gap gene expression domain patterning. For example, dynamic anterior shifts are seen in domain expression over time because of asymmetric gap-gap crossrepression \[[@pbio-0030064-b56]\]. Asymmetric repression and other circuits could, in the future, be engineered into our chambers by altering the repressor binding sites in the appropriate constructs. Such a system would require component turnover to achieve steady-state patterning. We have begun to tackle this project through our experiments with controlled protease degradation, but a further requirement would be to have autocatalytic production of T7 polymerase from a localised source, rather than the injected pulse of purified polymerase in our current model. Interestingly, our experimental data showed a reproducible degree of patterning even in the unrepressed system ([Figure 3](#pbio-0030064-g003){ref-type="fig"}A, 15 min, C output). Because gene C is activated by a separate polymerase (SP6), this patterning cannot result from competition for activator. Therefore, competition for other resources (such as ribosomes, nucleotides, and tRNAs) may allow A and B to "inhibit" C. Indeed, supplying extra components (particularly wheat germ extract and SP6 polymerase together) increases protein production under these conditions, including that of C (unpublished data). If competition can generate patterns, albeit less well defined ones than repression-connected networks, this could perhaps represent an evolutionary "network precursor" state: Weak patterns could be generated by localisation and competition between factors, and these could later be consolidated by evolution of a "true" negative network connection. However, this hypothesis may not be relevant to the situation inside an insect egg, as this has probably evolved to deliver nutrients very efficiently to the embryo, even at a very primitive evolutionary stage. It therefore remains to be seen whether competition effects would be as significant in vivo. Since our models use minimal components to achieve spatial pattern formation, they demonstrate the ease with which very simple networks might evolve. Patterning may be achieved with only localisation, diffusion, and some kind of functional network connection, such as transcription activation, competition, or repression. The addition of extra layers of network properties, such as controlled degradation, could then fix and stabilize such patterns. In fact, since sublocalisation---followed by stepwise addition of network components---is sufficient to generate crude patterns, it might provide a plausible mechanism for early spatial network evolution inside a single cell. However, it should be noted that gap gene expression domain patterning probably evolved by a different mechanism, from an earlier multicellular state, where segments were added sequentially through polar growth. In fact, *bicoid* is absent in most other insects, and it has been proposed that *Drosophila* evolved *bicoid* by duplication of the homeodomain-encoding gene *zerknüllt,* found in lower Diptera \[[@pbio-0030064-b66]\]. Diffusion Rates and Patterning {#s3b} ------------------------------ We were intrigued by our observations, both in vitro and in silico, that patterning required the activator molecule to diffuse or propagate more rapidly than the inhibitors. This is interesting because it is the opposite of the M-G system (described in the Introduction and reviewed in \[[@pbio-0030064-b12]\]). Long-range activation is not unknown in chemical patterning systems \[[@pbio-0030064-b67]\], although many biological models appear to require the M-G criterion for long-range inhibition (e.g., \[[@pbio-0030064-b68]\]). The other obvious differences between our system and the M-G model are the initial localisation of components and the lack of autocatalysis of the activator. It will be interesting to determine whether such M-G patterning systems can be recreated in our chambers, once further factors are considered, such as the avoidance of "autocatalytic explosions" or "global inhibitions." The computer model that we developed allowed us to test a broad range of parameters, such as diffusion and degradation rates ([Figure 5](#pbio-0030064-g005){ref-type="fig"}), revealing differences between the requirements for patterning among the different species in the gene network. First, the more-connected member of the network (gene B) was much more sensitive to parameter variation than the less-connected members (genes A and C). This is perhaps to be expected, since protein B has two separate boundaries of expression (defined as a function of T7 distribution and both A and C expression), whereas proteins A and C have only single "edges" to be defined. Another important feature of the system emerged when scaling the parameters for the model patterns to *Drosophila* scale ([Figure 5](#pbio-0030064-g005){ref-type="fig"}, 0.5 mm "embryo," 2.5 h). We found that, assuming simple diffusion, B-peak formation was compatible only with unphysiologically slow diffusion values (diffusion constants for mRNA \[Dm\] and protein \[Dp\] = 0.02 μm^2^s^−1^; for T7 activator \[DX\] = 0.43 μm^2^s^−1^). Since cellular proteins are expected to diffuse more rapidly (approximately 1--100 μm^2^s^−1^), this could be an artefact, reflecting the simplicity of our model. Nonetheless, simple diffusion still appears too rapid to account for *Drosophila*-scale patterning. It should be noted that a potential barrier to free diffusion is the active nuclear import of Bicoid and Hunchback \[[@pbio-0030064-b69]\]. In a separate example, diffusion of pair-rule transcripts is overridden by microtubule transport \[[@pbio-0030064-b70]\]. Controlled sublocalisation may therefore be crucial to limit apparent diffusion in vivo, allowing more precise patterning. Perspectives {#s3c} ------------ The understanding of how precision of patterning is achieved in *Drosophila* is still far from complete. In a recent study, it was shown that the Bicoid profile is far more variable between embryos than that of Hunchback, but the mechanism by which this noise is filtered remains unknown \[[@pbio-0030064-b71]\]. As more and more detailed experimental data are collected \[[@pbio-0030064-b72]\], and new mechanisms are proposed to account for patterning, it will be important to test the sufficiency of these mechanisms through experimental reconstitution. For such purposes, the chambers described here may be easily adapted to test different hypotheses. In vitro systems are a useful first step towards testing the sufficiency of a network---which might then be reengineered in the original target organism. Combining simple reconstruction with theoretical modelling is a useful tool to discover and test general design principles in gene networks \[[@pbio-0030064-b73],[@pbio-0030064-b74],[@pbio-0030064-b75],[@pbio-0030064-b76]\]. Until now, however, the spatial component essential in many biological processes has been ignored in these approaches. We anticipate that other networks, such as signalling cascades or metabolic networks, might also be studied using our system and that the spatial element, introduced through the beads, might provide new insights into complex systems. Materials and Methods {#s4} ===================== {#s4a} ### Magnetic chamber construction {#s4a1} A detailed, step-by-step description of the construction of the chamber can be found in [Protocol S1](#sd001){ref-type="supplementary-material"}. Briefly, nine stirring-bar magnets (1.5 mm × 8 mm; VWR International, Vienna, Austria; \#4429025) were inserted vertically into a plasticine-filled standard petri dish, creating a magnetic array (see [Figure 1](#pbio-0030064-g001){ref-type="fig"}D). Construction was guided with a grid template, laser-printed on a transparent acetate sheet, and fixed over the magnets and plasticine. The template was a 3 mm × 18 rectangle with nine subdivisions ("slabs"). A sterile cell culture dish (Nalge Nunc, Rochester, New York, United States; \#150350) was fixed immediately above the magnetic array. Chamber borders (1 mm deep) were constructed on the base of this second dish, following the template, using strips cut from adhesive Hybriwell chambers (Sigma-Aldrich, St. Louis, Missouri, United States; \#H1159--100EA). ### Gene network constructs {#s4a2} Maps of the constructs are illustrated in [Figure 2](#pbio-0030064-g002){ref-type="fig"}. Repressors A, B, and C were derived from previously engineered zinc fingers \[[@pbio-0030064-b58]\]. Repressor A contained six zinc fingers, recognising the sequence 5′- AGGGAGGCGGACTGGGGA-3′, fused to the residues 11--55 of the Kox-1 repression domain \[[@pbio-0030064-b77]\] and a six-repeat FLAG epitope tag \[[@pbio-0030064-b78]\]. Repressor B contained six zinc fingers, recognising the sequence 5′- AGGGAGGCGGGAGCTTTC-3′ and fused to a three-repeat FLAG-tag. Repressor C contained three zinc fingers, recognising the sequence 5′- GGAGCTTTC-3′, fused to the Kox domain and a three-repeat FLAG-tag. The following polymerase consensus promoter regions were used: T7, 5′- TAATACGACTCACTATA**G** GGAG-3′; SP6, 5′- ATTTAGGTGACACTATA**G** AAGGG-3′. The gene network promoters were linked with neutral or repressor sites to the polymerase promoters. In the following nucleotide sequences, zinc finger binding sites are indicated in lowercase, initiation nucleotides in bold, and promoter overlaps underlined. Unrepressed T7, 5′- TAATACGACTCACTATA**G** GGAGAAACACCATAG-3′ (see [Figure 3](#pbio-0030064-g003){ref-type="fig"}A, constructs A and B, and [Figure 3](#pbio-0030064-g003){ref-type="fig"}B, construct A). Unrepressed SP6, 5′- ATTTAGGTGACACTATA**G** AAGGGAAACACCATAG-3′ (see [Figure 3](#pbio-0030064-g003){ref-type="fig"}A, construct C). T7 repressed by A (and weakly by B), 5′- TAATACGACTCACTATa**g**ggaggcggactgggga -3′ (see [Figure 3](#pbio-0030064-g003){ref-type="fig"}B, construct B). SP6 repressed by A (and weakly by B), 5′- ATTTAGGTGACACTATA**G**Aagggaggcggactgggga-3′ (see [Figure 3](#pbio-0030064-g003){ref-type="fig"}B and [3](#pbio-0030064-g003){ref-type="fig"}C, construct C). T7 repressed by A and C (and weakly by B), 5′- TAATACGACTCACTATa**g**ggaggcggactggggaTggagctttc-3′ (see [Figure 3](#pbio-0030064-g003){ref-type="fig"}C, construct B). T7 repressed by C (and weakly by B), 5′- TAATACGACTCACTATA**G**ggagctttc-3′ (see [Figure 3](#pbio-0030064-g003){ref-type="fig"}C, construct A). Constructs were cloned in pCaSpeR4, sequenced, and used to generate PCR DNA for in vitro transcription-translation. ### Gene network reactions {#s4a3} Paramagnetic beads were coated with PCR DNA (with one primer biotinylated) using a Dynabeads Kilobase Binder Kit (Dynal, Oslo, Norway; \#601.01). Typically, gene A was used at 800 fmol per 10 μl of beads, resuspended in 8 μl of water; 200 fmol of gene B and 140 fmol of gene C were combined with 20 μl of beads, and resuspended in 20 μl of water. Transcription-translation mixture was prepared that included 2.5 μl of water; 28 μl of ultra-low melting point agarose (Sigma; \#A2576) solution (prepared as 1.5% \[w/v\] in boiling water and cooled to 30 °C); and TNT Coupled Wheat Germ Extract System (Promega, Madison, Wisconsin, United States; \#L4130 and \#L4140), which comprised 20 μl of TNT wheat germ extract, 1.2 μl of TNT reaction buffer, 0.6 μl of amino acid mixture (1 mM), 1.2 μl of RNasin (not included in TNT kit), and 0.5 μl of SP6 polymerase. 54 μl of this mixture was dispensed per chamber. For degradation experiments, 2.25 units of Factor Xa (Amersham Biosciences, Little Chalfont, United Kingdom) were added per chamber. Coated Dynabeads were injected at appropriate positions over the magnetic array: typically, 100 fmol of gene A (1 μl), 5 fmol of gene B, and 3.5 fmol of gene C (0.5 μl). T7 polymerase (0.5 μl; from Promega TNT kit) was immediately injected at the chamber edges. After timed incubations at 25 °C, chambers were transferred to 4 °C for 35 min, to form a gel. Gel slices were cut with a razor blade (guided by the printed template) and aspirated with a P10 Gilson pipette. Samples were mixed with 10 μl of SDS-loading buffer and analysed by SDS-PAGE, Western blotting, and ECL, with anti-M2 FLAG antibody (Sigma; \#F3165). Further details on this step can be found in [Protocol S1](#sd001){ref-type="supplementary-material"}. ### Computer modelling {#s4a4} A Perl script was written to simulate the diffusion-coupled expression of genes A, B, and C, by T7 and SP6 phage polymerases, in a translation extract. The program parameters and script are fully described in [Protocols S1--S3](#sd001){ref-type="supplementary-material"}. 18 mm-scale chamber model: Parameters included separate diffusion (and degradation) rates for RNA and protein; a separate apparent diffusion for injected T7, modelled from experimental observations (rapid initial diffusion with exponential decay; Section 5 of [Protocol S1](#sd001){ref-type="supplementary-material"}); estimated binding constants for all interacting species (zinc finger dissociation constants were estimated from previous work on related three- and six-finger constructs \[[@pbio-0030064-b58],[@pbio-0030064-b79],[@pbio-0030064-b80]\]); and estimated transcription-translation rates. For adapting the model to the 0.5-mm *Drosophila* scale, chamber size was scaled down, and only simple diffusion was allowed for all components; for simplicity, transcription-translation rates were not varied (Section 3 of [Protocol S1](#sd001){ref-type="supplementary-material"})*.* Supporting Information {#s5} ====================== ::: {.caption} ###### Detailed Description of Model (1.2 MB PDF). ::: ::: {.caption} ###### Click here for additional data file. ::: ::: {.caption} ###### Parameter File for Simulations This file contains the default parameters for the computer model in a format that can be read by the Perl script. (5 KB DOC). ::: ::: {.caption} ###### Click here for additional data file. ::: ::: {.caption} ###### Computer Program Script for Simulations This text file is a Perl script to run the computer simulations described in the manuscript. (22 KB DOC). ::: ::: {.caption} ###### Click here for additional data file. ::: Accession Numbers {#s5a} ----------------- The Locuslink (<>, or GeneID ([, accession numbers of the genes and proteins discussed in this paper are Bicoid (40830), *caudal* (35341), *giant* (31227), *huckebein* (40549), *hunchback* (41032), *knirps* (40287), *Krüppel* (38012)*,* Nanos (42297), *tailless* (43656), Torso (35717), and *zerknüllt* (40828). We would like to thank K. Michalodimitrakis, C. Gonzalez, and B. Schönwetter for helpful discussions. MI is supported by an International Research Fellowship from the Wellcome Trust, United Kingdom. **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** MI and LS conceived and designed the experiments. MI performed the experiments. CL analyzed the data and contributed reagents/materials/analysis tools. MI, CL, and LS wrote the paper. Citation: Isalan M, Lemerle C, Serrano L (2005) Engineering gene networks to emulate *Drosophila* embryonic pattern formation. PLoS Biol 3(3): e64. Dm : diffusion constant for mRNA Dp : diffusion constant for protein DX : diffusion constant for T7 activator M-G : Meinhardt-Gierer

Introduction {#s1} ============ The Mlabri are an enigmatic group of about 300 people who nowadays range across the Nan, Phrae, and Phayao provinces of north and northeastern Thailand and the Sayaburi province of western Laos \[[@pbio-0030071-b01],[@pbio-0030071-b02]\]. Their traditional lifestyle is to move frequently through the dense forests of the high mountains, building temporary structures of bamboo sticks thatched with banana leaves, which they occupy for a few days, until the leaves turn yellow (thus accounting for their traditional Thai name, Phi Tong Luang, which means "spirit of the yellow leaves"). First contacted by Europeans in 1936 \[[@pbio-0030071-b03]\], they are unique among the hill tribes of northern Thailand in that, until recently, they subsisted by hunting and gathering combined with occasional barter trade with villagers. The origins of the Mlabri are controversial. Some investigators have assumed that there is a direct connection between the Mlabri and the ancient Hoabinhian hunting--gathering culture of Southeast Asia \[[@pbio-0030071-b01]\]. However, a limited investigation of blood group variation \[[@pbio-0030071-b04]\] raised the possibility that the Mlabri originated via a founder event from an agricultural group, and preliminary linguistic analyses support this idea. The Mlabri language seems lexically most closely related to Khmu and Tin, two languages of the Khmuic branch of the Mon-Khmer sub-family of Austro-Asiatic languages, both of which are spoken in agricultural highland villages \[[@pbio-0030071-b05]\]. The cluster of dialects jointly referred to as Tin, or Mal/Prai, \[[@pbio-0030071-b06]\] is spoken in the Thailand--Laos border region that the Mlabri also occupy, whereas Khmu is spoken over a much wider area \[[@pbio-0030071-b07]\]. The grammar of Mlabri additionally has features that deviate markedly from typical Mon-Khmer, suggesting that Mlabri developed as a result of contact between speakers of a Khmuic language and speakers of a quite different language of unknown affiliation \[[@pbio-0030071-b02],[@pbio-0030071-b08]\]. We report here the results of an investigation of genetic diversity in the Mlabri, to see whether patterns of genetic variation might provide further insights into the question of an agricultural versus hunting--gathering origin for the Mlabri. The rationale for using genetic analyses to investigate this question is that previous work has shown that hunter--gatherer groups typically differ from their agricultural neighbors in having reduced genetic diversity and high frequencies of unique mtDNA types \[[@pbio-0030071-b09],[@pbio-0030071-b10],[@pbio-0030071-b11],[@pbio-0030071-b12],[@pbio-0030071-b13],[@pbio-0030071-b14]\], so we might expect a similar pattern if the Mlabri have always been hunter--gatherers. The genetic results, combined with linguistic and cultural evidence, suggest that the most probable explanation for the origin of the Mlabri is an extreme founder event from an agricultural group, followed by adoption of a hunting--gathering lifestyle. Results/Discussion {#s2} ================== Genetic Analyses: mtDNA Diversity {#s2a} --------------------------------- We analyzed 360 bp of the first hypervariable segment (HV1) of the mtDNA control region in 58 Mlabri; surprisingly, all of the sequences were identical, with the following differences from the reference sequence \[[@pbio-0030071-b15]\]: 16140C, 16189C, and 16266A, as well as the common Asian 9-bp deletion in the intergenic region between the cytochrome oxidase subunit II and lysine tRNA genes \[[@pbio-0030071-b16]\]. No other human population has been found to lack mtDNA HV1 variation, and mtDNA HV1 variation in six other hill tribes (all agricultural groups) from the same region of Thailand was significantly higher ([Figure 1](#pbio-0030071-g001){ref-type="fig"}; [Table 1](#pbio-0030071-t001){ref-type="table"}). {#pbio-0030071-g001} ::: {#pbio-0030071-t001 .table-wrap} ::: {.caption} ###### Genetic Diversity Parameters Based on mtDNA HV1 Sequences, Y-STR Haplotypes, and Autosomal STR Genotypes for the Mlabri and the Six Other Hill Tribes :::  Diversity in the Mlabri is significantly lower than the average for the other groups for all three genetic systems, based on *t*-tests (not shown) ^a^ Probability of the observed heterozygosity excess under the stepwise mutation model, Wilcoxon one-tailed test ::: Y-Chromosome Diversity {#s2b} ---------------------- We analyzed nine short tandem repeat (STR) loci on the Y chromosome in 54 Mlabri, and again found significantly reduced variation in the Mlabri compared to the other six hill tribes ([Figure 1](#pbio-0030071-g001){ref-type="fig"}; [Table 1](#pbio-0030071-t001){ref-type="table"}). The Mlabri had just four Y-chromosome STR (Y-STR) haplotypes, two of which differed by a single repeat at a single locus from one each of the other two haplotypes ([Table 2](#pbio-0030071-t002){ref-type="table"}). The Y-STR haplotype diversity in the Mlabri is again lower than that reported for any other human population \[[@pbio-0030071-b17],[@pbio-0030071-b18]\]; the Akha, one of the six other hill tribes, also exhibited very low Y-STR diversity ([Table 1](#pbio-0030071-t001){ref-type="table"}). The average variance in the allele size distribution at the nine Y-STR loci shows an even greater contrast between the Mlabri and the other hill tribes: the average variance was 0.11 for the Mlabri, versus an average of 1.45 for the other six hill tribes. ::: {#pbio-0030071-t002 .table-wrap} ::: {.caption} ###### Y-STR Haplotypes in the Mlabri :::  The number of repeats for the allele at each locus in the four haplotypes is given ::: Autosomal DNA Diversity {#s2c} ----------------------- We analyzed nine autosomal STR loci in the Mlabri and the other six hill tribes, and again found significantly reduced variation in the Mlabri ([Figure 1](#pbio-0030071-g001){ref-type="fig"}; [Table 1](#pbio-0030071-t001){ref-type="table"}). The genotype frequencies did not deviate significantly from Hardy--Weinberg expectations for any locus in the Mlabri; however, even though these nine STR loci are on different chromosomes and hence unlinked, eight pairs of loci exhibited significant linkage disequilibrium (LD) (*p* \< 0.05; [Figure 2](#pbio-0030071-g002){ref-type="fig"}), as measured by a likelihood ratio test \[[@pbio-0030071-b19]\]. This is significantly more (*p* \< 0.01) than the 1.8 pairs expected by chance (out of 36 pairwise comparisons) to exhibit this level of LD. For each of the six agricultural hill tribes, the number of pairs of loci exhibiting significant LD was within expectations ([Figure 2](#pbio-0030071-g002){ref-type="fig"}). Moreover, the *p*-value of the likelihood ratio test is a measure of the strength of the association between two loci \[[@pbio-0030071-b19]\]; the average *p*-value was 0.20 for the Mlabri, versus 0.31--0.55 for the other six hill tribes, indicating that overall associations between these unlinked loci were stronger in the Mlabri than in the other hill tribes. However, the sample size for the Mlabri for the autosomal STR analyses was larger than the sample size for the other hill tribes (*n* = 35 for the Mlabri, versus *n* = 29--30 for the others), so it is possible that the lower average *p*-value for the Mlabri reflects more statistical power due to a larger sample size and not more LD. To test this, we sampled 30 Mlabri at random and redid the LD analysis; the conclusions did not change, indicating that the lower average *p-*value for the Mlabri does reflect more LD in the Mlabri. {#pbio-0030071-g002} One explanation for the reduced diversity at mtDNA, Y-STR loci, and autosomal STR loci, and the significant number of pairs of unlinked autosomal STR loci in LD, is a severe reduction in population size in the Mlabri. Following such an event, the number of alleles is reduced more than the heterozygosity, leading to an excess of observed heterozygosity compared to that expected for the observed number of alleles under mutation--drift equilibrium \[[@pbio-0030071-b20]\]. We therefore compared the observed and expected heterozygosity (at mutation--drift equilibrium, conditioned on the observed number of alleles) for the autosomal STR loci in the Mlabri and the six other hill tribes, under a stepwise mutation model. Only the Mlabri exhibited a significant excess of observed heterozygosity ([Table 1](#pbio-0030071-t001){ref-type="table"}). Although more complicated scenarios are possible, the simplest explanation is that the Mlabri (but not the other hill tribes) have undergone a severe reduction in population size, as also indicated by the mtDNA and Y-STR haplotype data, and as also suggested by a previous study of blood group variation \[[@pbio-0030071-b04]\]. Population Size Reduction in the Mlabri {#s2d} --------------------------------------- Assuming that there was a reduction in population size in the Mlabri that set the mtDNA and Y-chromosome diversity near or equal to zero, the coalescence times for the Mlabri mtDNA and Y-STR haplotypes provide an upper estimate as to when the population reduction occurred. We therefore applied Bayesian-based coalescence analysis \[[@pbio-0030071-b21]\] to the mtDNA sequences and the Y-STR haplotypes from the Mlabri and the other six hill tribes. For the six agricultural hill tribes, the resulting estimates of coalescence time are broadly distributed ([Figure 3](#pbio-0030071-g003){ref-type="fig"}), indicating little information in the data (except for the Akha, who do show a pronounced peak in the posterior probability distribution for the Y-STR data, in accordance with their lower Y-STR haplotype diversity). By contrast, the estimates of coalescence time for the Mlabri show a sharp peak ([Figure 3](#pbio-0030071-g003){ref-type="fig"}), with a median time of 770 y (approximate 95% credible interval 250--4,270 y) for the mtDNA sequences and 490 y (approximate 95% credible interval 170--1,290 y) for the Y-STR haplotypes. {#pbio-0030071-g003} Both the mtDNA and the Y-STR data therefore indicate that the Mlabri underwent a substantial reduction in population size about 500--800 y ago (and not more than about 1,300 y ago, if the mtDNA and Y-chromosome data reflect the same event). There are two possible scenarios: (1) a bottleneck, in which the Mlabri were reduced from a formerly large population to a much smaller population size, which then increased to the current level of about 300 individuals; or (2) a founder event, in which the Mlabri began as a very small number of individuals, became isolated, and then increased over time to their present size. Similar reductions in genetic diversity are predicted under either scenario, so the genetic data cannot distinguish between these. But some information can be obtained by considering the magnitude of the reduction in population size needed to completely eliminate mtDNA diversity in the Mlabri. The amount of population size reduction needed to completely eliminate mtDNA diversity in the Mlabri depends on how much mtDNA diversity was present prior to the size reduction. We assumed that the ancestral Mlabri population would have the same mtDNA diversity as one of the other hill tribes and then estimated the amount of population size reduction needed to completely eliminate mtDNA diversity by resampling with replacement various numbers of mtDNA types from the ancestral (pre-bottleneck) population. For example, we started with an ancestral population with the same distribution of mtDNA types as the Akha. We then sampled two mtDNA types (with replacement) from this ancestral population, repeated this procedure 1,000 times, and found that 243 out of the 1,000 resamples of size two had no mtDNA diversity; thus, the probability is 0.243 that a reduction to just two individuals would eliminate mtDNA diversity in an ancestral population that started with the same mtDNA diversity as the Akha. We then repeated this procedure, sampling three mtDNA types (with replacement), and obtained a probability of 0.007 that there would be no mtDNA diversity following a reduction to three individuals. Therefore, if the Mlabri were derived from a population with the same mtDNA diversity as the Akha, the population had to be reduced to not more than two unrelated females, in order to completely eliminate mtDNA diversity. This resampling analysis was carried out six times, with the putative ancestral mtDNA diversity corresponding to each of the six hill tribes. The results of this analysis were that for five of the ancestral populations, resampling three (or more) individuals gave a probability of no mtDNA diversity of less than 0.05; for the remaining ancestral population (which had the same starting mtDNA diversity as the Red Karen), resampling four (or more) individuals gave a probability of no mtDNA diversity of less than 0.05. We also carried out a similar analysis for the Y-STR types in the Mlabri. Here we again assumed an ancestral population with the same Y-STR haplotype diversity as one of the other hill tribes, then determined the maximum number of individuals that could be sampled at random that would have not more than two Y-STR types (since the four Y-STR types in the Mlabri consist of two pairs that differ by a single-step mutation at a single locus). The results of this analysis were that at most 3--6 individuals (depending on which hill tribe the ancestral population resembled most in terms of Y-STR diversity) could have been present after the size reduction, otherwise, with greater than 95% probability, more than two Y-STR types would have been retained. A critical assumption is the amount of genetic diversity present in the ancestral Mlabri population prior to the size reduction. The estimates used in the above analysis are based on agricultural populations, which in general have more mtDNA diversity than hunter--gatherer populations. We therefore also constructed putative ancestral populations with frequency distributions of mtDNA types identical to those found in the !Kung and in African Pygmies \[[@pbio-0030071-b22]\]; the results of the resampling analysis were the same. Another assumption of this analysis is that the event that led to the population size reduction completely eliminated the mtDNA diversity. Alternatively, some mtDNA diversity may have been present after the population size reduction, but was subsequently lost because of drift. Loss of mtDNA diversity due to subsequent drift is not likely if there was a single event reducing the Mlabri population size that was followed by population growth, since mtDNA diversity is retained in growing populations \[[@pbio-0030071-b23]\]. However, if the reduction in size occurred over several generations, then it may not have been as dramatic a bottleneck as the resampling analysis implies. To investigate this further, we employed a Bayesian approach, following the procedure previously used to estimate the number of founders for the Maoris \[[@pbio-0030071-b24]\] but allowing for new mutations, to estimate the number of founders for the Mlabri, assuming various time periods since the founding event. The results ([Figure 4](#pbio-0030071-g004){ref-type="fig"}) indicate that the most probable number of founders is one over all time periods; however, for longer time periods since the founding event, there is decreasing information on the number of founders from the observation of no mtDNA diversity in the Mlabri. As expected, the longer the time since the founding event (i.e., the slower the population growth rate), the greater the influence of drift in eliminating diversity that might have been present in the founding population. Nevertheless, given that the coalescent analyses indicate an upper date for the origin of the Mlabri of about 1,000 y ago, the lack of mtDNA diversity in the Mlabri is most consistent with a very small founding population size, perhaps even only one female lineage. {#pbio-0030071-g004} Origin of the Mlabri {#s2e} -------------------- The group that gave rise to the founder event that established the Mlabri could have been either a hunter--gatherer group, in which case the Mlabri maintained their hunting--gathering lifestyle from before, or an agricultural group, in which case the Mlabri subsequently adopted their current hunting--gathering lifestyle. While the genetic data cannot unequivocally distinguish between these two possibilities, they do suggest the latter. Other hunter--gatherer groups typically share few, if any, mtDNA types with neighboring agricultural groups, consistent with long-term isolation of the hunter--gatherer groups. For example, !Kung, African Pygmies, Andamanese Islanders, and south Indian hunter--gatherer groups can readily be distinguished from nearby agricultural groups on the basis of their mtDNA sequences \[[@pbio-0030071-b09],[@pbio-0030071-b13],[@pbio-0030071-b14],[@pbio-0030071-b25]\]. By contrast, the Mlabri mtDNA sequence has been reported in other, agricultural hill tribes \[[@pbio-0030071-b26],[@pbio-0030071-b27]\], and identical or closely related sequences have also been reported from Southeast Asia and China \[[@pbio-0030071-b09],[@pbio-0030071-b28],[@pbio-0030071-b29]\]. Similarly, the Mlabri Y-STR haplotypes are identical or closely related (differing by a single-step mutation at one locus) to Y-STR haplotypes found in Southeast Asia and Oceania \[[@pbio-0030071-b30],[@pbio-0030071-b31]\]. Also, the Mlabri do not exhibit any alleles at the nine autosomal STR loci that are not found in the agricultural hill tribes. The widespread sharing of mtDNA, Y-STR, and autosomal STR alleles between the Mlabri and agricultural groups in Southeast Asia are not expected if the Mlabri have always been hunter--gatherers. Instead, the genetic data suggest that the Mlabri are derived from an agricultural group. Moreover, the Mlabri vocabulary and folklore also give some evidence of ancient familiarity with agriculture coexisting with hunting and gathering (J. Rischel, personal communication). While preliminary in nature, the available linguistic evidence suggests that the present-day Mlabri language arose after some speakers of a Khmuic language, most likely Tin, became isolated and subsequently experienced intensive contact with speakers of some other, presently unknown language \[[@pbio-0030071-b02],[@pbio-0030071-b08]\]. Just how long ago the Mlabri and Tin languages diverged cannot be determined, but it has been suggested that Tin branched from Khmu about 600 y ago, and that Tin then branched into two varieties (Mal and Prai) some 200--300 y ago \[[@pbio-0030071-b06],[@pbio-0030071-b32]\]. These time estimates are based on a calibration of the chronology of sound changes in Tin against reasonably secure datings of sound changes in neighboring languages; the actual time depth may be underestimated, but most likely by not more than a few centuries. Thus, the linguistic evidence would date the origin of the Mlabri at less than 1,000 y ago, in excellent agreement with the genetic evidence. Other data that may shed light on the origins of the Mlabri, such as historical information, are scarce, since the Mlabri do not have a written language and the first recorded contact was only in 1936. However, the Tin Prai have an oral tradition concerning the origin of the Mlabri (J. Rischel, personal communication), in which several hundred years ago, Tin Prai villagers expelled two children and sent them downriver on a raft. They survived and escaped into the forest, turning to a foraging lifestyle and thus becoming the Mlabri. Although it is difficult to know how to evaluate such oral traditions, this story nevertheless intriguingly parallels the genetic and linguistic evidence concerning the origins of the Mlabri. In sum, genetic, linguistic, and cultural data all suggest a founding event in the Mlabri, involving a single maternal lineage and 1--4 paternal lineages some 500--1,000 y ago, from an ancestral agricultural population. The Mlabri then subsequently adopted their present hunting and gathering lifestyle, possibly because the group size at the time of founding was too small to support an agricultural lifestyle. Other examples of such cultural reversion are rare; probably the best known involves Polynesian hunter--gatherers on the Chatham Islands and the South Island of New Zealand \[[@pbio-0030071-b33]\], who abandoned agriculture and adopted a maritime-based foraging subsistence because of the rich marine resources and the inability of these islands to support cultivation of tropical crops. Other hypothesized examples of cultural reversion, such as the Punan of Borneo \[[@pbio-0030071-b34]\], the Guajá and other lowland Amazonian groups \[[@pbio-0030071-b35]\], and the Sirionó of Bolivia \[[@pbio-0030071-b36]\], are controversial, as it is not clear whether these groups are descended directly from earlier hunter--gatherer groups or whether they indeed have undergone cultural reversion. Detailed genetic analyses, as carried out here for the Mlabri, may shed further light in these cases. In any event, our conclusion that the Mlabri, a present-day group of hunters and gatherers, was founded recently and in all probability from an agricultural group further supports the contention that contemporary hunter--gatherer groups cannot be automatically assumed to represent the pre-agricultural lifestyle of human populations, descended unchanged from the Stone Age \[[@pbio-0030071-b37]\]. Indeed, even if they have not reverted from an agricultural lifestyle, most (if not all) contemporary hunter--gatherer groups interact with, and have evolved and changed along with, agricultural groups \[[@pbio-0030071-b38]\]. The Mlabri provide a unique opportunity to investigate the circumstances and consequences of a reversion from an agricultural to a hunting--gathering lifestyle that apparently was not dictated by purely ecological reasons (as in the case of Polynesian hunter--gatherers). Materials and Methods {#s3} ===================== {#s3a} ### Samples {#s3a1} There are three linguistically distinct subgroups of Mlabri \[[@pbio-0030071-b39]\], designated A, B, and C. Subgroup A (also known erroneously as "Mrabri") is the only group that has been studied in detail \[[@pbio-0030071-b01]\]; subgroup B (minor Mlabri) is practically extinct \[[@pbio-0030071-b02]\], and subgroup C (formerly "Yumbri") comprises less than 30 people \[[@pbio-0030071-b39]\]. Blood samples and genealogies of 91 Mlabri (all from subgroup A) were obtained with informed consent in 1999, and cell lines were prepared and DNA was extracted from the cell lines. The genealogical data were used to identify and exclude known relatives from the genetic analyses. Data on mtDNA and Y-STR variation from six agricultural hill tribes in the same geographic region (Akha, Lahu, White Karen, Red Karen, CR Lisu \[from near Chiang Rai\], and MHS Lisu \[from near Mae Hong Son\]), all of whom speak Sino-Tibetan languages, were published previously \[[@pbio-0030071-b26]\]. ### Genetic analyses {#s3a2} The first hypervariable segment (HV1) of the mtDNA control region (nucleotide positions 16,024--16,385) was amplified and sequenced directly, as described previously \[[@pbio-0030071-b29]\], from 58 Mlabri. PCR analysis of the intergenic region between the cytochrome oxidase subunit II and lysine tRNA genes, which harbors an informative 9-bp deletion, was carried out as described previously \[[@pbio-0030071-b28]\]. Nine Y-STR loci (DYS385a, DYS385b, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, and DYS394) were amplified and genotypes determined, using previously described methods \[[@pbio-0030071-b30]\], for 54 Mlabri. Nine autosomal STR loci (D3S1358, vWA, FGA, D8S1179, D21S11, D18S51, D5S818, D13S317, and D7S820) plus the amelogenin locus were amplified with the AmpFLSTR Profiler Plus PCR Amplification Kit (Applied Biosystems, Foster City, California, United States), using 2--4 ng of DNA in a 15-μl reaction volume. Genotypes were determined by fragment analysis on an ABI377 (Applied Biosystems) for 35 Mlabri, 29 Lahu, and 30 individuals from each of the other hill tribes. ### Statistical analyses {#s3a3} Genetic diversity, heterozygosity, and tests for goodness of fit to Hardy--Weinberg expectations were calculated with Arlequin 2.000 \[[@pbio-0030071-b40]\]. LD was estimated as the probability of the likelihood of the data assuming linkage equilibrium versus the likelihood of the data assuming association \[[@pbio-0030071-b19]\]; Arlequin 2.000 was used to obtain maximum-likelihood estimates of the haplotype frequencies for each pair of loci with the EM algorithm \[[@pbio-0030071-b41]\], and the null distribution of the *p*-value of the likelihood ratio test was generated by 10,000 random permutations. The program Bottleneck (<>) was used to compare the observed heterozygosity at each autosomal STR locus to that expected at mutation--drift equilibrium for the observed number of alleles, assuming a stepwise mutation model. Bayesian-based coalescence analyses of Y-STR haplotypes \[[@pbio-0030071-b42]\] were performed using the software Batwing (<>) and previously described prior distributions for the initial effective population size, population growth rate, and Y-STR mutation rates \[[@pbio-0030071-b30]\]. The coalescence time for mtDNA HV1 sequences was also estimated by a Bayesian procedure \[[@pbio-0030071-b21]\] as described previously for Xq13.3 sequences \[[@pbio-0030071-b43]\], using the same initial effective population size and population growth rate priors as for the Y-STR analysis, and a γ-distribution with parameters α = 14.74 and β = 0.0005 (corresponding mean = 0.00737) as a prior for the mutation rate \[[@pbio-0030071-b44]\]. Resampling of mtDNA and Y-STR types, in order to estimate the magnitude of population size reduction needed to eliminate mtDNA and reduce Y-STR diversity, was performed with the software Resample (<>). Bayesian analysis of the number of founders for the Mlabri was performed by pooling the mtDNA types in the other hill tribes to obtain a starting population, from which a certain number of founding mtDNA types were selected at random, assuming a uniform prior distribution between one and 20 founders. The sample was then allowed to grow from the number of founders to size 300 (the current size of the Mlabri population) over various time intervals, such that the shorter the time interval, the faster the growth rate. Simulations were performed both under the assumption of no new mutations, and with a mutation rate of one mutation/sequence/10,000 y. For each combination of parameters, 1,000,000 simulations were carried out. The simulation results were converted via Bayes\'s theorem into a posterior probability for the number of founding individuals, conditioned on the observation of no diversity in a random sample of size 58 (the sample size in this study). In practice, the posterior probability distributions were independent of the mutation rate (analyses not shown). We thank S. Brauer for technical assistance, M. Wood for assistance with the resampling analyses, and R. Cordaux, D. Gil, and M. Kayser for useful discussion. We are especially grateful to J. Rischel for permission to cite his unpublished observations, and for discussions of the linguistic and cultural evidence concerning the Mlabri. HO was supported by a fellowship from the Japan Society for the Promotion of Science (JSPS). Sample collection was supported by funds from the JSPS and the Ministry of Education, Science, and Culture, Japan; genetic research was supported by the Max Planck Society, Germany. We dedicate this paper to the memory of Surin Pookajorn. **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** MS and HO conceived and designed the experiments. HO and BP performed the experiments. MS, HO, BP, GW, and AvH analyzed the data. HO, SP, WSI, DT, and TI contributed reagents/materials/analysis tools. MS, HO, and BP wrote the paper. ¤Current address: Department of Genetics, Yale University School of Medicine, New Haven, Connecticut, United States of America **^‡^**Deceased 13 July 2004. Citation: Oota H, Pakendorf B, Weiss G, von Haeseler A, Pookajorn S, et al. (2005) Recent origin and cultural reversion of a hunter--gatherer group. PLoS Biol 3(3): e71. LD : linkage disequilibrium STR : short tandem repeat Y-STR : Y-chromosome short tandem repeat

Introduction {#s1} ============ Complex positive species interactions have been shown to expand the ecological niche and increase the persistence of the organisms involved in a variety of systems. In terrestrial systems, increased diversity of mycorrhizal symbionts is correlated with increased biodiversity of plant communities, resulting in greater stability and longer persistence at the community level \[[@pbio-0030077-b01]\]. In marine ecosystems, the coral Oculina arbuscula harbors a majid crab, *Mithrax forceps,* that prevents overgrowth of macroalgae and shading of the corals \[[@pbio-0030077-b02]\]. This allows O. arbuscula to maintain its facultative mutualism with photosynthetic zooxanthellae in well-lit habitats off the Atlantic coast of North Carolina, increasing the amount of energy available to the coral for growth and reproduction. At cold seeps in the Cascadia \[[@pbio-0030077-b03],[@pbio-0030077-b04]\] and Aleutian \[[@pbio-0030077-b05]\] subduction zones, bioirrigation through burrow formation and bioturbation by clams (*Calyptogena* spp.) has been shown to significantly affect the distribution of microbial anaerobic methane oxidation. Lamellibrachia luymesi inhabits areas associated with advection of hydrocarbons and other reduced chemicals to the seafloor (hydrocarbon or brine seeps) on the upper Louisiana slope (ULS) of the Gulf of Mexico from 400 to 1,000 m depth. L. luymesi does not posses a digestive system; rather, it acquires energy via internal sulfide-oxidizing bacterial symbionts \[[@pbio-0030077-b06]\]. L. luymesi differs from other vestimentiferan tubeworms by its ability to use a posterior extension of its body, the "root," to acquire sulfide from interstitial pools in sediments \[[@pbio-0030077-b07],[@pbio-0030077-b08]\]. Near the anterior plumes of tubeworms, sulfide concentrations typically decline below 0.1 μM as the tubeworms approach 1 m in length \[[@pbio-0030077-b09]\]. By using its roots, L. luymesi is able to delve into deeper sediment layers, providing access to more persistent sulfide sources. In the apparent absence of lethal predation \[[@pbio-0030077-b10],[@pbio-0030077-b11]\], the most significant hazard that this vestimentiferan tubeworm faces is sulfide limitation. Its high uptake rate of sulfide from hydrocarbon seep sediments, estimated at over 30 μmol · h^−1^ for a moderate-sized individual \[[@pbio-0030077-b12]\], suggests that sulfide flux may be limiting in L. luymesi\'s habitat. A diverse chemosynthetic community relies on the sulfide generated as a by-product of anaerobic degradative processes in the Gulf of Mexico \[[@pbio-0030077-b10],[@pbio-0030077-b11]\]. Reduction of seawater sulfate utilizing methane or other hydrocarbons as electron donors produces the majority of sulfide available at ULS seeps \[[@pbio-0030077-b13],[@pbio-0030077-b14]\]. Anaerobic methane oxidation is most commonly carried out by microbial consortia consisting of sulfate-reducing bacteria along with methanogenic archaea executing reverse methanogenesis \[[@pbio-0030077-b15],[@pbio-0030077-b16]\]. Methane oxidation linked to sulfate reduction and subsequent authigenic carbonate precipitation constrain ocean--atmosphere carbon fluxes \[[@pbio-0030077-b03],[@pbio-0030077-b04]\], accounting for up to 20% of the global methane flux to the atmosphere \[[@pbio-0030077-b17]\]. Oxidation of other hydrocarbons and organic material, carried out by sulfate-reducing bacteria in monoculture and in consortia with other microbes \[[@pbio-0030077-b18]\], may account for a larger proportion of sulfate depletion in ULS sediments \[[@pbio-0030077-b14]\]. These processes can result in a decoupling of sulfate reduction and methane oxidation rates \[[@pbio-0030077-b14]\], and form carbonates consisting mainly of non-methane-derived carbon \[[@pbio-0030077-b19]\]. L. luymesi may influence these anaerobic processes by utilizing its roots to release the sulfate generated by its symbionts during sulfide oxidation \[[@pbio-0030077-b07],[@pbio-0030077-b08],[@pbio-0030077-b12]\]. This hypothetical mechanism would provide sulfate for anaerobic methane oxidation and hydrocarbon degradation at sediment depths normally devoid of energetically favorable oxidants, thereby augmenting exogenous sulfide production. In this study, we address the question of whether known biogeochemical processes could supply sulfide at rates sufficient to match the requirements of long-lived L. luymesi aggregations. In the diagenetic model presented here, the hypothesized release of sulfate in sediments with sufficient electron donors results in sulfide generation at rates matching the sulfide uptake rate of L. luymesi aggregations for over 250 y. We speculate that the mutual benefits derived from the syntrophy among symbiotic tubeworms and microbial consortia implicit in the model would expand our current concept for the potential complexity of positive interspecific interactions and the benefits they confer. Results/Discussion {#s2} ================== L. luymesi Sulfate Release Allows Persistence of Aggregations {#s2a} ------------------------------------------------------------- The model predicts that inputs from known sources, including diffusion and advection of deep sulfide along with reduced seawater sulfate, will support a moderately-sized aggregation of 1,000 individuals for an average of 39 y (range, 22 to 78 y) ([Figure 1](#pbio-0030077-g001){ref-type="fig"}). A smaller aggregation of 200 individuals could be maintained with these sources for an average of 64.1 y (standard deviation, 10.6 y). In this model configuration, the duration of adequate sulfide flux is not congruent with the known sizes of aggregations and existing age estimates of L. luymesi individuals and aggregations. Adding sulfate release by tubeworm roots to the model results in sulfide generation and flux at rates that match the demands of large aggregations, allowing the tubeworms to survive for over 250 y ([Figure 1](#pbio-0030077-g001){ref-type="fig"}). This additional source of sulfate results in a two-orders-of-magnitude increase in sulfate flux in older (\>100 y) aggregations, accounting for over 90% of sulfate available after only 24 y. The sulfate released by the tubeworms would be used for anaerobic methane oxidation and hydrocarbon degradation. The nature of the relationship between symbiotic tubeworms and microbial consortia that we are proposing is a coupling of the sulfur cycle only, and not carbon. Light dissolved inorganic carbon (DIC) resulting from the oxidation of hydrocarbons is apparently not taken up by tubeworms as the carbon stable isotope signatures of L. luymesi are heavier than would be expected from a methane-derived DIC source \[[@pbio-0030077-b20],[@pbio-0030077-b21]\]. In addition, the well-studied hydrothermal vent tubeworm, *Riftia pachyptila,* obtains carbon in the form of CO~2~ across its plume \[[@pbio-0030077-b22]\]. However, this does not necessarily exclude the passive diffusion of DIC across the root surface, which could account for some of the variability observed in L. luymesi carbon stable isotope signatures \[[@pbio-0030077-b20],[@pbio-0030077-b21]\]. By augmenting the sulfate supply to microbial consortia for sulfate reduction, large aggregations of tubeworms may survive for hundreds of years in the model, mirroring the population sizes and individual lengths regularly observed and collected at seeps on the ULS \[[@pbio-0030077-b23]\]. {#pbio-0030077-g001} Model Results Are Robust to Parameter Variation {#s2b} ----------------------------------------------- An alternate hypothesis to explain the discordance between estimated sulfide supply and uptake rates is the presence of locally elevated seepage rates. Sensitivity analyses were carried out to determine the potential effects of uncertainty in seepage rate on supply estimated for aggregations without root sulfate release. A 10% increase in seepage rate resulted in a 5.6% increase in sulfide supply to aggregations 200 y old and older. This corresponds to only 16.4% of the sulfide required, which does not serve to extend aggregation survivorship (average, 39 y; range, 21 to 79 y) beyond that determined for lower flow rates. To supply the sulfide flux required by older aggregations, seepage rate would have to be at least 363 mm · y^−1^. This is over ten times greater than the rate used in the model (32 mm · y^−1^), which is the highest region-wide estimate for the Gulf of Mexico \[[@pbio-0030077-b24]\]. A rate of over 300 mm · y^−1^ approaches rates reported for active venting of fluids ([Table 1](#pbio-0030077-t001){ref-type="table"}). Active venting would result in the visual manifestation of seepage in the form of methane bubbles and oil droplets, which are generally restricted to mussel (Bathymodiolus childressi) beds at these sites \[[@pbio-0030077-b25]\]. In addition, larger, and therefore older \[[@pbio-0030077-b26]\], aggregations have lower epibenthic sulfide concentrations \[[@pbio-0030077-b08],[@pbio-0030077-b09],[@pbio-0030077-b25]\] suggesting that seepage becomes less vigorous over time and is not in the form of active venting in larger tubeworm aggregations. While difficult to obtain, in situ measures of advection rate of fluids at Gulf of Mexico seeps could be used to test these assumptions and may lend insight into the relationship between variability in tubeworm growth rate and sulfide availability. ::: {#pbio-0030077-t001 .table-wrap} ::: {.caption} ###### Reported Seepage Rates for Hydrocarbon and Methane Seeps :::  ::: The high degree of variability in growth rate and recruitment rate could also affect the ratio of supply and demand in the model. In an aggregation exhibiting anomalously low recruitment, the size of the rhizosphere would increase more rapidly than the biomass of the aggregation. This would lead to high rates of sulfide delivery and generation and low rates of sulfide uptake by tubeworm roots. When initial recruitment rate (*a* in equations [1](#pbio-0030077-e001){ref-type="disp-formula"} and [2](#pbio-0030077-e002){ref-type="disp-formula"}) is decreased by 10%, the length of time that supply exceeds demand increases by 3.7%. This effect appears to be linear, with a 20% decrease in initial recruitment rate resulting in a 7.4% increase in persistence. If growth rate is increased, thereby increasing the rate of rhizosphere growth in terms of surface area for diffusion and advection, there appears to be little effect of the ratio of supply to demand (20% increase in growth---0% change in persistence time). In fact, increasing growth to the upper limits of the error term ([equation 5](#pbio-0030077-e005){ref-type="disp-formula"}) lowers the amount of time that the aggregation can be supported since biomass and sulfide demand increase more rapidly than increases in supply resulting from additional surface area. By decreasing growth rate, aggregations may be supported for longer periods of time, with a 20% decrease leading to a 6.3% increase in persistence time and a decrease of 88% leading to persistence for over 250 y. While an 88% lower growth rate lies outside of the range of existing growth data, this could be accomplished by ceasing growth for extended periods of time in a quiescent stage. This possibility remains to be investigated in L. luymesi. By utilizing a variable recruitment rate in the model, both between realized aggregations and between years within a model run, along with a growth error term encompassing the full range of observed growth data, the model is capable of generating aggregations within the range of the 10%--20% variability tested in this analysis. Even these outlying aggregations (presented as maxima and minima in [Figure 1](#pbio-0030077-g001){ref-type="fig"}) support the qualitative conclusions drawn from model results. While the model was based on empirical data to the greatest degree possible, estimates of many of the parameters necessary to resolve the model were not available or are extremely difficult to measure in deep water with existing technology. Uptake rates were measured in the laboratory \[[@pbio-0030077-b08]\] for relatively small individuals (\<50 cm). While we attempted to approximate metabolic scaling by covarying uptake and growth rates, it is possible that large individuals require even lower sulfide flux. Model predictions are not overly sensitive to variability in this parameter. A reduction by 10% of the overall sulfide uptake rate results in a 5.2% increase in persistence time. To maintain an aggregation for over 250 y, mass-specific uptake rate would have to be reduced 6-fold. While this could also be accomplished by entering a period of quiescence as mentioned before, there is no existing evidence for this ability in vestimentiferans. The second version of the model is based on the assumption that L. luymesi is capable of releasing sulfate through its roots. It should be noted that in the model, sulfate release is constrained by the rate of sulfate generation by the tubeworm\'s sulfide-oxidizing symbionts, resulting in the near 1:1 ratio of supply and demand in [Figure 1](#pbio-0030077-g001){ref-type="fig"}. Though modeled sulfate flux across the roots into the rhizosphere may exceed 20 mmol · h^−1^ in older aggregations, the roots provide an ample respiratory surface such that rates of sulfate flux per unit root surface area do not exceed 0.4 μmol · h^−1^· cm^−2^ in the model. It remains possible that a proportion of the sulfate could be released through the plume of the tubeworms, though the energy required to pump sulfate against a concentration gradient (seawater \[SO~4~\] = 29 mM) \[[@pbio-0030077-b13]\] suggests that it would be more energetically favorable for the sulfate to passively diffuse out of the roots. It is also possible that sulfate flux could be increased by active bioirrigation delivering seawater to deeper sediment layers through the tubeworm tubes. This could allow the sulfide-oxidizing symbionts to store some of the oxidized sulfide as elemental sulfur rather than releasing it as sulfate, while maintaining sufficient sulfate flux to deeper sediment layers for sulfide generation. These mechanisms remain hypothetical and require further experimental investigations to evaluate their potential role in this system. Tubeworms Impact Seep Biogeochemistry {#s2c} ------------------------------------- Tubeworm sulfate release, in conjunction with high sulfide uptake rates, could contribute to the observation of declining advection rate in older aggregations. By increasing sulfate flux to deeper sediments, L. luymesi increases integrated rates of anaerobic methane oxidation and hydrocarbon degradation, which would enhance authigenic calcium carbonate precipitation within the rhizosphere. Under the conditions of root sulfate release in the model, calcium carbonate precipitation is rapid (0.109 to 0.316 μmol · l^−1^ · sec^−1^) in the first 53 y, with rates declining exponentially thereafter. By creating a barrier to fluid advection \[[@pbio-0030077-b04]\], this could result in the observed decrease in epibenthic sulfide concentration in older aggregations \[[@pbio-0030077-b08],[@pbio-0030077-b09]\] and the predicted cessation of tubeworm recruitment around this time \[[@pbio-0030077-b12],[@pbio-0030077-b23]\]. In order to prevent the precipitation of carbonate directly on the root surface, L. luymesi individuals may release hydrogen ions as well as sulfate through their roots. While hydrogen ion flux through the roots has not yet been empirically demonstrated, none of the nearly 5,000 tubeworms examined as part of this study were observed to have carbonate formed directly on their roots, suggesting that this form of precipitation is inhibited in some manner. In the model, diffusion of hydrogen ions across the root surface (the only form of release explicitly modeled) accounts for less than 40% of ion generation when carbonate precipitation is most vigorous. We speculate that L. luymesi may utilize the excess hydrogen ions generated by their sulfide-oxidizing symbionts to periodically raise the rate of hydrogen ion flux from their roots. This would not only supply additional hydrogen ions to sulfate-reducing bacteria, but could inhibit carbonate precipitation on the tubes and subsequent reduction of the root area utilizable as a respiratory surface. Further pH reduction could dissolve existing carbonate in sediments beneath the rhizosphere, thereby opening seepage pathways and allowing further root growth. This possibility is corroborated by the observation of young tubeworms that had apparently bored through bivalve shells in an experimental system (R. Carney, personal communication). Empirical measurements of hydrogen ion flux across the root tissue of L. luymesi are required to test these hypothetical mechanisms. The release of sulfate by tubeworm roots potentially explains the frequent observation of highly degraded hydrocarbons in the vicinity of large tubeworm aggregations \[[@pbio-0030077-b27]\]. The majority of sulfate supplied by tubeworm roots is utilized for microbial hydrocarbon degradation in the model ([Figure2](#pbio-0030077-g002){ref-type="fig"}). This process alone accounts for over 60% of the sulfide available to aggregations after approximately 80 y. In the absence of liquid and solid phase hydrocarbons, methane flux would have to be approximately four times the rate in the model in order to fuel sufficient sulfate reduction to support an aggregation for over 200 y. This could occur in sediments overlying rapidly sublimating gas hydrates, and hydrate abundance has been previously suggested as a potential factor influencing the distribution of chemosynthetic communities in the Gulf of Mexico \[[@pbio-0030077-b10]\]. However, model results indicate that large chain hydrocarbons are the most significant energy source for sulfate reduction in tubeworm-dominated sediments. Increased integrated rates of hydrocarbon degradation would lead to highly biologically altered hydrocarbon pools among the roots of tubeworm aggregations. Hydrocarbon oxidation has been implicated as one of the dominant processes in the carbon cycle at ULS seeps, accounting for over 90% of the carbon in carbonates collected in the vicinity of tubeworm aggregations \[[@pbio-0030077-b19]\]. Model analysis indicates that the minimum amount of organic carbon (including hydrocarbons as well as buried organic material) in sediments required to supply sulfide at rates matching aggregation demand (1:1 ratio) is 1.03% by weight, remarkably close to the lowest value found in any of the seep sediment core samples (1.2%) \[[@pbio-0030077-b13],[@pbio-0030077-b28]\], and greater than that found in ULS sediments away from seeps (0.71%) \[[@pbio-0030077-b29]\]. Determination of organic carbon concentration in sediments beneath tubeworm aggregations is necessary to test the prediction that elevated carbon content at seeps, primarily resulting from oil seepage, provides the energy source required to generate sufficient sulfide for tubeworm aggregations. {#pbio-0030077-g002} Additional sulfate flux from tubeworm roots could also explain the high apparent sulfate diffusion coefficients determined for tubeworm-impacted sediments \[[@pbio-0030077-b13]\]. Anomalous sulfate fluxes have been proposed to be a result of bioturbation and bioirrigation by macrofauna \[[@pbio-0030077-b03],[@pbio-0030077-b05]\], and recycling by microbial mats \[[@pbio-0030077-b13]\]. The results of the model presented here provide evidence for macrofaunal sulfur recycling, an additional component to be considered in future investigations of cold seep biogeochemistry. The hypothesized release of sulfate by tubeworm roots potentially explains numerous, apparently disparate observations, hinting at the great impact that L. luymesi aggregations may have on their abiotic environment. While the proposed interactions between symbiotic tubeworms and sulfate-reducing bacteria are essential for the persistence of L. luymesi aggregations in the model, we suggest that there are significant effects on the microbial community as well. This syntrophy will increase the abundance of sulfate-reducing bacteria and therefore increase the rates of anaerobic methane oxidation and hydrocarbon degradation carried out by microbial consortia that rely on sulfate as an oxidant. Tubeworm-generated sulfate supplies a more energetically favorable electron acceptor below the normal depth of sulfate penetration at seeps, relaxing the limitation on anaerobic oxidative processes at these sediment depths. Deeper sediment layers then become habitable to sulfate reducers, significantly altering the microbial community structure within the rhizosphere. Model configurations neglect the potential role of bioirrigation of seawater sulfate through L. luymesi tubes, which could further increase sulfate supply to deeper sediment layers. The possible role of tubeworm roots as substrata for the growth of microbial consortia, analogous to the habitat afforded mycorrhizal symbionts of higher plants, remains another possible benefit for the microbes. These predictions may be tested by determination of the relative abundance of microbial consortia at different depths of sediments both impacted by and isolated from tubeworms. Localization of the microbes on the root surface would provide evidence for a more intricate relationship. It is our hope that the results of this model may provide the impetus for future rigorous experimental tests of these ideas. Summary {#s2d} ------- The model results presented here are consistent with the hypothesis that L. luymesi releases sulfate into hydrocarbon-rich sediments to fuel sulfide generation, allowing for the persistence of the longest-lived animal known. The importance of this process to sulfide generation in the modeled rhizosphere implies a complex relationship between an animal with bacterial endosymbionts and external sulfate-reducing bacteria, often in consortia with methane-oxidizing or hydrocarbon-degrading microbes. This positive interspecific relationship, including members of all three domains, would benefit both the tubeworms and the microbial consortia involved. This expands our existing concept of the potential for complexity in mutualisms and the benefits they may confer. Further complex relationships are likely to be discovered through continued research into the role of positive species interactions at the individual and community levels. Materials and Methods {#s3} ===================== {#s3a} This study couples an individual-based population growth and sulfide uptake model \[[@pbio-0030077-b12]\] to a diagenetic diffusion/advection model to compare the relative magnitude of sulfide supply and uptake for long-lived tubeworm aggregations. A series of 1,000 iterations of the model under three different initial conditions (known sources of sulfate, known sources plus root sulfate supply, and known sources with elevated seepage rates) were carried out. The rhizosphere (volume of sediment encompassed by the root system of an aggregation) is modeled as an inverted dome beneath the sediment with a radius equal to the average root length of the population ([Figure 3](#pbio-0030077-g003){ref-type="fig"}). The rhizosphere was approximated by a series of two-dimensional discs at 2-cm intervals in order to reduce the complexity of a three-dimensional solution for a sphere of changing size. Sulfate (SO~4~ ^2−^), methane (CH~4~), sulfide (HS^−^), bicarbonate (HCO~3~ ^−^), and hydrogen ion (H^+^) fluxes across the rhizosphere boundary are determined. Sulfate reduction rates using methane, larger chain hydrocarbons, and buried organic matter as electron donors are modeled in order to estimate the sulfide available to tubeworm aggregations as they change in size over the course of 250 y. {#pbio-0030077-g003} ### Population growth model {#s3a1} The population growth model follows the methodology presented in \[[@pbio-0030077-b12]\] and includes population growth, mortality rate, individual growth rate, and sulfide uptake rate. The parameters underlying the population growth model were refined using growth data from an additional 615 individuals and population data from an additional 11 aggregations comprising 4,908 individuals. The model presented here includes data from a total of 23 tubeworm aggregations from three nearby sites (Green Canyon oil lease blocks 184, 232, and 234) collected over a period of 7 y on the ULS to arrive at generalized population growth parameters. L. luymesi individuals are dioecious, with males releasing sperm into the water column. Fertilization is believed to be external \[[@pbio-0030077-b30]\], though sperm has been found within the oviducts of females of the hydrothermal vent tubeworm R. pachyptila \[[@pbio-0030077-b31]\]. Eggs and embryos are positively buoyant and develop into a swimming trochophore-like larval stage within 3 d of fertilization \[[@pbio-0030077-b32]\]. Larvae are lecithotrophic and may remain in the water column for several weeks \[[@pbio-0030077-b32]\]. They require hard substrata for settlement, and acquire symbionts from their environment after metamorphosis \[[@pbio-0030077-b33],[@pbio-0030077-b34]\]. Settlement is initially rapid, and continues until the available substrate is occupied \[[@pbio-0030077-b12],[@pbio-0030077-b23],[@pbio-0030077-b35]\]. Population sizes of aggregations collected with existing sampling devices typically vary between 100 and 1,500 individuals (\[[@pbio-0030077-b12],[@pbio-0030077-b23]\]; this study), though far larger aggregations covering tens to hundreds of square meters are common at the sites sampled. Previous studies have shown that L. luymesi has an average longevity of 135 y \[[@pbio-0030077-b12]\], and requires an average of 210 y to reach 2 m in length \[[@pbio-0030077-b26]\], a size not uncommon among collected animals. Mortality events are exceedingly rare, dropping below 1% annual mortality probability for animals over 30 cm \[[@pbio-0030077-b12]\]. The expanded datasets of growth and mortality rates included here extend the longevity estimate for L. luymesi to an average of 176 y and the estimated age of a 2-m-long animal to 216 y. At the beginning of each iteration, population growth parameters are chosen for the following population growth model: where *N* is population size, *t* is time (in years), *K* is carrying capacity (set to 1,000 individuals for all simulations presented here), *a* describes the initial slope of the line, *b* defines the degree of density dependence, and *c* is a shape parameter. The first parameter *(a)* was generated using the following function: where *ɛ*\[N(0,1)\] is a normally distributed random deviate with an average of zero and a standard deviation of one. This allows the initial recruitment rate to vary within the range of all recruitment trajectories that have been observed \[[@pbio-0030077-b12]\]. The other parameters were not normally distributed; therefore, the log-transformed distributions were used to define the distribution of the random numbers generated. As the three parameters in the model were significantly correlated (ln(*a*) and ln(*b*), *r* = −0.853, *p* \< 0.001; ln(*a*) and ln (*c*), *r* = −0.461, *p* = 0.036), values of *b* and *c* were chosen from their relationship with *a*: The value of *a* was allowed to vary each year according to the pooled standard error associated with the estimates of *a* from the empirical data (standard error, 0.105). Once population size equaled or exceeded carrying capacity, recruitment was ceased, representing the lack of additional substrate or sulfide available in the water column. Once recruitment was determined for that year, the individual-based portion of the model began. Each individual was traced through time with respect to its length, root length, mass, mortality probability, mass-specific sulfide uptake rate, sulfate excretion rate, and hydrogen ion elimination rate. Growth rates of tubeworms were determined by staining tubes in situ ([Figure 3](#pbio-0030077-g003){ref-type="fig"}) and collection 12 to 14 mo later (\[[@pbio-0030077-b26]\]; this study). Individual growth rate was determined from the following function ([Figure 4](#pbio-0030077-g004){ref-type="fig"}): ![L. luymesi Growth Rate\ Size-specific growth of L. luymesi determined from stained tubeworms. Different colors indicate growth data from different aggregations. Blue points labeled "2000" are all from Bergquist et al. \[[@pbio-0030077-b26]\]. Other colored points refer to submersible dive numbers from 2003 when stained aggregations were collected.\ (A) Growth function and 95% confidence interval for size-specific growth.\ (B) Error function fitted to the residuals of the model.](pbio.0030077.g004){#pbio-0030077-g004} Length *(l)* is defined here as the distance from the anterior end of the tube to an outer tube diameter of 2 mm following the methodology of \[[@pbio-0030077-b26]\]. All growth rates were standardized to 365 d. The error term is an additional function fitted to the residuals of the first regression function ([Figure 4](#pbio-0030077-g004){ref-type="fig"}B), resulting in a variable growth rate. This error term was used rather than varying growth within the 95% confidence interval of the regression of length and growth rate because of the high degree of variability in growth among individuals. It should also be noted that there is a certain degree of variability in growth rate between aggregations ([Figure 4](#pbio-0030077-g004){ref-type="fig"}). This may be attributable to spatial or temporal variability in seepage rate or sulfide concentration between aggregations. Aggregations may be subject to persistently differing conditions on a small (meter) scale, or may encounter periodic fluctuations in habitat characteristics. Because we are uncertain whether this variation is persistent on the temporal scales that we are simulating, between-aggregation variability is not explicitly modeled, though by chance certain realized aggregations deviated from mean growth rate. The ratio of root length to tube length was determined from individual length using the following function: Annual mortality rate was approximated as the size-specific frequency of empty tubes in collected aggregations \[[@pbio-0030077-b12]\] with an overall annual mortality rate of 0.569%. This approximation is conservative and likely overestimates yearly mortality, as available data indicate that empty tubes should persist longer than 1 y \[[@pbio-0030077-b12],[@pbio-0030077-b36]\]. Mortality probability was determined for each 10-cm size class using the following function: where *m* is mortality probability and *l* is length. Individuals were considered dead if their probability of mortality exceeded a uniform random number between zero and one. By using generalized population growth parameters in the model presented here, we attempt to encompass the range of empirical data from sampled aggregations in our examination of sulfide uptake and supply rates. Taken together, the population growth model including recruitment, growth, and mortality provides a good qualitative if not quantitative fit for any individual aggregation, reflecting the size frequency of tubeworms within sampled aggregations \[[@pbio-0030077-b12]\]. It should be noted that the modeling of specific aggregations was not the aim of this study; rather, an attempt has been made to encompass the variability observed in the various populations of tubeworms that have been sampled. To examine the effect of uncertainty in the population growth parameters, sensitivity analyses were carried out. The initial slope of the recruitment rate (*a* in [equation 1](#pbio-0030077-e001){ref-type="disp-formula"}) was varied while individual size-specific growth rate was held constant (no error term in [equation 5](#pbio-0030077-e005){ref-type="disp-formula"}). Growth rate was then varied while holding the initial rate of population growth constant (no error term in [equation 2](#pbio-0030077-e002){ref-type="disp-formula"}). The effect of a 10% change in each parameter was determined, and then changes of greater magnitude were examined to determine the fastest rate of population or individual growth that could be supported by the sulfide available to the aggregation in the absence of sulfate release. Individual sulfide uptake was allowed to vary within the range of laboratory-determined sulfide uptake rates according to that individual\'s growth rate for that year: where *u* is uptake rate (in micromoles per gram per hour), *m* is mass (in grams), and *g* is growth rate (in centimeters per year). Growth rate was divided by the maximum growth rate (10 cm · y^−1^) such that highest growth rates resulted in highest uptake rates. By scaling uptake rate with growth, we approximate metabolic scaling, resulting in a decline in uptake rate by a factor of 3.7 over the range of tubeworm sizes in this study \[[@pbio-0030077-b12]\]. The amount of sulfate that could be excreted by each individual was determined from the amount generated by sulfide oxidation carried out by the internal chemoautotrophic symbionts assuming constant internal sulfate concentration, thereby accounting for changes in body volume. We do not account for the binding of sulfur by free amino acids, as this is believed to relatively minor compared the flux rates of sulfate and sulfide, and is reversible \[[@pbio-0030077-b37]\]. Hydrogen ions are also generated in the oxidation of sulfide by the tubeworm symbionts. Hydrogen ion elimination rate was determined in the model in the same fashion as sulfide uptake, with growth rate determining the variability in this metabolic flux according to laboratory-measured ion fluxes (mean, 10.96 μmol · g^−1^ · h^−1^; standard deviation, 1.88 μmol · g^−1^ · h^−1^) \[[@pbio-0030077-b38]\]. Simple diffusion of hydrogen ions across the root surface was included in the model, though the exact mode of proton flux has not yet been determined experimentally for L. luymesi \[[@pbio-0030077-b38]\]. As diffusion across the roots accounts for a relatively small proportion of total proton flux (less than 10% in large individuals), additional pathways are likely and require further investigation. ### Geochemical setting {#s3a2} Known sources of sulfide available to L. luymesi aggregations are sulfide transported with seeping fluids \[[@pbio-0030077-b10]\] and sulfide generated via reduction of seawater sulfate \[[@pbio-0030077-b39],[@pbio-0030077-b40]\]. The majority of the sulfide present at ULS sites is believed to be related to sulfate reduction coupled to anaerobic hydrocarbon oxidation \[[@pbio-0030077-b14],[@pbio-0030077-b39]\]. Other potential sources of sulfide associated with seepage include anaerobic oxidation of deeply buried organic material \[[@pbio-0030077-b10]\], "sour" hydrocarbons containing a proportion of sulfur \[[@pbio-0030077-b41]\], and hydrocarbon interactions with sulfur-bearing minerals such as gypsum and anhydrite found in the salt dome cap rocks of the ULS \[[@pbio-0030077-b08],[@pbio-0030077-b42],[@pbio-0030077-b43]\]. Concentrations of all chemical species in the sediments surrounding the rhizosphere were derived from the dataset included in Arvidson et al. \[[@pbio-0030077-b13]\] and Morse et al. \[[@pbio-0030077-b28]\]. Only those sediment cores taken around the "drip line" of tubeworm aggregations that contained detectable sulfide concentrations were used. Due to the vagaries of sampling with a submersible in sediments heavily impacted by carbonate and roots, those cores with detectable sulfide are believed to more accurately represent conditions around the periphery of the rhizosphere. Dissolved organic carbon (DOC) concentration was used as an estimate of methane concentration. While other forms of DOC make up this total concentration, methane accounts for 90%--95% of the hydrocarbon gasses dissolved in pore waters \[[@pbio-0030077-b28]\]. In seep sediments, the majority of DOC is likely to be in the form of hydrocarbon gasses. Because estimates of organic acid concentrations were not available, they could not be explicitly modeled. This would not affect the overall concentration of electron donors in the model, but could affect the sulfate reduction rate. Since sulfate reduction rate estimates for methane seeps in the Gulf of Mexico are among the highest recorded \[[@pbio-0030077-b14],[@pbio-0030077-b39]\], any differences in DOC composition (e.g., higher relative concentrations of dissolved organic acids) would serve to lower the overall sulfate reduction rate and sulfide availability. Sulfide supply estimates presented are likely overestimated most by the model without root sulfate release owing to the greater reliance on anaerobic methane oxidation in this form of the model. Simulations including sulfate release by tubeworms are affected to a lesser extent as the concentration of electron donors is not limiting in this model configuration. Solid and liquid phase organic carbon was separated into hydrocarbons and buried organic material according to their relative concentrations in hydrocarbon seep and surrounding Gulf of Mexico sediments. Background sediments on the ULS contain 0.71% organic carbon by weight \[[@pbio-0030077-b29]\]. At hydrocarbon seeps on the ULS, organic carbon accounts for 4.47% of total weight. This was assumed to be the sum of background organic input plus carbon in the form of C~6+~ hydrocarbons. It is possible that the higher biomass located at ULS seeps in the form of non-living macrofaunal and microbial materials may also contribute to the increased organic carbon concentration, but without empirical estimates, this could not be accounted for in the model. Hydrocarbons may consist of between 50% and 95% labile materials \[[@pbio-0030077-b44],[@pbio-0030077-b45],[@pbio-0030077-b46]\]. Based on existing data on degradation rates and residual hydrocarbons subjected to degradation \[[@pbio-0030077-b42],[@pbio-0030077-b47]\], a value of 50% labile material was used here. These assumptions of hydrocarbon concentration and degradation potential are therefore believed to be conservative. The following functions were fitted to the sulfide, sulfate, and methane concentration profiles ([Figure 5](#pbio-0030077-g005){ref-type="fig"}) to determine the boundary conditions at any given depth: {ref-type="sec"} and original data in \[[@pbio-0030077-b13],[@pbio-0030077-b28]\]). Best-fitted line based on least squares fit of [equation 9](#pbio-0030077-e009){ref-type="disp-formula"}.](pbio.0030077.g005){#pbio-0030077-g005} where *C* ~0~ is initial concentration, *C~∞~* is concentration at infinite distance, and *C~i~* is concentration at depth *d*. As there were no existing data for sediments below 30 cm, concentrations at infinite depth (*C~∞~*) were used (SO~4~ ^2−^ = 0 mmol · l^−1^, HS^−^ = 12 mmol · l^−1^, DOC = 11 mmol · l^−1^, DIC = 20 mmol · l^−1^, pH = 7.78). The first derivatives of the sulfide and methane profiles were used for the calculation of advective flux from depth. The first derivative of the sulfate profile was used for diffusive flux across the water--sediment interface of the rhizosphere, with advection rate subtracted from diffusive flux of sulfate across this surface. Advection (seepage) rate varied with time according to the following function: where *t* is simulation time in years and *sed* is sedimentation rate (6 cm · 1,000 y^−1^) \[[@pbio-0030077-b29]\]. Early seepage rate approximated the highest flux rates measured or estimated for methane seeps and declined with time in the model to the highest estimates for persistent, region-wide seepage in the Gulf of Mexico ([Table 1](#pbio-0030077-t001){ref-type="table"}). This follows a pattern of hydrocarbon seep development, with the highest seepage rates early in the evolution of the local seepage source followed by occlusion of fluid migration pathways by carbonate precipitation, hydrate formation, and possibly tubeworm root growth. By using the highest rate estimated (32 mm · y^−1^ = 0.000365 cm · h^−1^ in [equation 10](#pbio-0030077-e010){ref-type="disp-formula"}) as the basal seepage rate, we are testing the possibility that tubeworm aggregations could survive under the most favorable conditions possible in the absence of tubeworm sulfate supply. For sediments encompassed by the rhizosphere, sulfide, sulfate, methane, DOC, and hydrogen ion concentration profiles were determined iteratively prior to model implementation using a central difference scheme: where *C* ~*i*(*t*)~ is concentration in cell *i* at time *t, D* is the diffusion coefficient, *k* is the maximum reaction rate, and *K~s~* is the half-saturation constant for the reaction ([Table 2](#pbio-0030077-t002){ref-type="table"}). Reactions included anaerobic methane oxidation ([equation 17](#pbio-0030077-e017){ref-type="disp-formula"}), tubeworm sulfide uptake rate ([equation 8](#pbio-0030077-e008){ref-type="disp-formula"}), and carbonate precipitation rate ([equation 22](#pbio-0030077-e022){ref-type="disp-formula"}). The concentration in each 2 × 2 cm cell was calculated at 1 h time steps. At the end of each year, diffusion distance increased. The number of cells (total diffusion distance) was determined by the average root length of L. luymesi populations as realized in independent runs of the population growth model described above, and included here as model input only. A separate function was fitted to each of the concentration profiles: ::: {#pbio-0030077-t002 .table-wrap} ::: {.caption} ###### Parameters Involved in Diagenetic Model :::  ^a^Diffusion coefficients all corrected for temperature, pressure, and salinity according to Stumm and Morgan \[[@pbio-0030077-b51]\] and Pilson \[[@pbio-0030077-b52]\] ^b^All disassociation constants corrected for temperature, salinity, and pressure according to Stumm and Morgan \[[@pbio-0030077-b51]\] and Pilson \[[@pbio-0030077-b52]\] except: CaOH, no correction; CaHCO~3~, CaSO~4~, CaSO~4~H~2~O, MgHCO~3~, temperature only; H~2~CO~3~, temperature and salinity only; and HSO~4~, temperature and pressure only ::: where *d* is radial distance. The relationship between the parameter *a* and distance was used to generate concentration profiles for each disc comprising the rhizosphere. Because of the tight linear relationship between diffusion distance and the shape of the curve, concentration profiles could be generated for a disc of any size using the following function: where α is 1.7344 and β is 1.0104 for HS^−^, α is 0.2111 and β is 0.3363 for SO~4~ ^2−^, and α is 0.1626 and β is 0.2518 for CH~4~. Diffusional fluxes of sulfide, sulfate, and methane were calculated according to the first and second derivatives of the concentration profiles as determined by the diameter of each disc. ### Model implementation {#s3a3} The model estimates sulfide availability to the aggregation as a whole by summing the fluxes separately determined for each 2-cm disc composing the rhizosphere. Depth-dependant boundary conditions were set for each disc separately based on the sediment profiles ([Figure 5](#pbio-0030077-g005){ref-type="fig"}). Diffusional fluxes into each disc were calculated from the shape of the concentration profiles according to the following function \[[@pbio-0030077-b48]\]: where *C* is concentration, *r* is disc radius, and *D~s~* is the diffusion coefficient corrected for porosity by: where *D~o~* is the diffusion coefficient corrected for temperature and pressure, *n* is the chemical species-specific constant, and φ is porosity. The value of *n* was set to 2.75 as this was found to be a reasonable fit for all chemical species examined \[[@pbio-0030077-b49]\]. The ionic states of each species at the average pH value of tubeworm-dominated sediments (7.78) were used for the determination of diffusion coefficients. Porosity was determined from the following function: where φ~*z*~ is porosity at depth *z,* φ~0~ is porosity at the sediment--water interface, and φ~∞~ is porosity at infinite depth; φ~0~ was set at 0.841, φ~∞~ at 0.765, and *a* at 0.210, as determined from the best fit with the porosity data ([Figure 6](#pbio-0030077-g006){ref-type="fig"}) from Morse et al. \[[@pbio-0030077-b28]\]. {ref-type="sec"} and original data in \[[@pbio-0030077-b13],[@pbio-0030077-b28]\]). Best-fitted line based on least squares fit of [equation 9](#pbio-0030077-e009){ref-type="disp-formula"}.](pbio.0030077.g006){#pbio-0030077-g006} Diffusion across the sediment--water interface of the rhizosphere was also considered as an additional input of sulfate and hydrogen ions. This was included as one-dimensional diffusion across a circular surface (subtracting the area encompassed by the tubeworm tubes) with diffusion distance equal to rhizosphere diameter, and concentration differential from seawater concentration to the average concentration within the rhizosphere. Sulfate and hydrogen ion diffusion across the root surface was then added (if included in the set of model realizations) as simple Fickian diffusion. Concentration differential was the difference between internal concentration and average concentration for each disc of the rhizosphere assuming roots were evenly proportioned according to the volume encompassed by each disc. Internal sulfate concentration and pH ([Table 2](#pbio-0030077-t002){ref-type="table"}) represented an average of the values determined for R. pachyptila \[[@pbio-0030077-b22]\], a hydrothermal vent tubeworm. Internal sulfate concentrations and pH of L. luymesi have not been reported, but these values are generally consistent within taxa \[[@pbio-0030077-b50]\]. Uptake of sulfide and release of sulfate were summed across the entire tubeworm population, again assuming an even distribution of roots within the rhizosphere. The paucity of empirical data on the location of any individual tubeworm\'s roots within an aggregation precluded modeling space explicitly; therefore, it is assumed that each individual has equal access to the resources available within the rhizosphere. Within the rhizosphere, sulfide generation may be limited by sulfate supply, electron donor availability, or sulfate reduction rate. Sulfate supply was determined as the sum of flux across the series of discs approximating the rhizosphere dome, across the sediment--water interface, and from root sulfate (if available). Available sulfate is utilized for anaerobic methane oxidation first (the more energetically favorable process), then hydrocarbon and organic matter degradation. Electron donors included methane, complex hydrocarbons, and buried organic material. Solid and liquid phase hydrocarbons and organic material were assumed to be homogenous within the rhizosphere. Methane supply was determined as the sum of flux across each rhizosphere disc boundary. Hydrocarbon and organic material concentrations were determined as the amounts encompassed within the rhizosphere volume minus that oxidized in previous years. Sulfate reduction rate was determined from the relative amounts of the various electron donors with higher rates (0.71 μmol · ml^−1^ · h^−1^) for methane oxidation and lower rates (0.083 μmol · ml^−1^ · h^−1^) for organic matter or hydrocarbon degradation \[[@pbio-0030077-b39]\]. Microbes carrying out these processes are assumed to be evenly distributed within the rhizosphere. Total hydrogen sulfide availability to the aggregation was determined as the sum of sulfide diffusion and advection across each rhizosphere disc and sulfide generated within the rhizosphere from sulfate reduction according to the following reactions: Bicarbonate (HCO~3~ ^−^) is generated at a 1:1 stoichiometry during anaerobic methane oxidation and a 2:1 stoichiometry in the degradation of organic material. As hydrocarbons are degraded forming smaller chain hydrocarbons and organic acids, bicarbonate is generated at different stoichiometries. Because different-sized hydrocarbons and organic acids were not accounted for in the model, a rough average of these stoichiometries (1.47:1) based on toluene, ethylbenzene, xylene, and hexadecane degradation \[[@pbio-0030077-b18]\] was used to determine the amount of bicarbonate generated per mole of carbon. Hydrogen ions are also used up in a 1:1 stoichiometry with sulfate in the sulfate reduction half reaction as included in reaction 17. In order to account for carbonate precipitation, the model traced DIC concentration, calcium concentration, hydrogen ion concentration, buffer capacity, carbonate saturation, and carbonate precipitation rate. The buffer state of the rhizosphere was calculated to determine changes in pH resulting from hydrogen ion flux. Buffer capacity (*β*) was calculated using the following function \[[@pbio-0030077-b51]\]: where *A* and *B* represent the concentrations of the various acids and bases in the buffer system. In addition to hydrogen and hydroxyl ions, the buffer system included carbonate (CO~2~, H~2~CO~3~, HCO~3~ ^−^, and CO~3~ ^2−^), sulfide (H~2~S and HS^−^), sulfate (HSO~4~ ^−^ and SO~4~ ^2−^), and borate (B\[OH\]~4~ ^−^ and B\[OH\]~3~) speciation. Current pH was used to determine the ionic state of each species according to temperature-, pressure-, and salinity-corrected disassociation constants when available \[[@pbio-0030077-b51],[@pbio-0030077-b52]\] ([Table 2](#pbio-0030077-t002){ref-type="table"}). Change in pH was determined from hydrogen ion flux and buffer capacity as follows: Saturation state is highly dependent on the degree to which calcium and bicarbonate form complexes with other ions. The "free" calcium was determined as the proportion of calcium that is not associated with complexed bicarbonate (HCO~3~ ^−^), carbonate (CO~3~ ^2−^), hydroxyl (OH^−^), or sulfate (SO~4~ ^2−^) ions. Free carbonate was determined as the amount not forming complexes with calcium (Ca^+^) or magnesium (Mg^+^) ions in solution. Saturation state was then calculated from the product of the concentrations of free calcium and carbonate divided by the solubility product constant. If the saturation state was above one, then carbonate precipitation occurred at a rate determined by: where *k* ~1~ is 0.00597 l · mol^−1^ · sec^−1^ and *k~3~* = 0.456 l · mol^−1^ · sec^−1^ \[[@pbio-0030077-b51]\]. Because there is no empirical relationship between weight percent of carbonate and sediment porosity in tubeworm-dominated sediments \[[@pbio-0030077-b28]\], precipitation did not directly affect porosity. Precipitation was accounted for in the model by subtracting the volume of carbonate precipitate from the total volume encompassed by the rhizosphere. At the end of each annual time step, model output included average length of individuals, population size, sulfide uptake rate, sulfide supply rate, root sulfate flux (if included), root hydrogen ion flux, amount of sulfide supply accounted for by each process (sulfide seepage, anaerobic methane oxidation, organic matter degradation, and hydrocarbon degradation), number of individuals that could be supported by sulfide supply, carbonate precipitation rate, volume of carbonate precipitate, and pH. We would like to acknowledge K. Montooth, P. Hudson, and five anonymous reviewers for providing helpful comments on drafts of the manuscript. We are indebted to J. Freytag, S. Dattagupta, D. Bergquist, R. Carney, and R. Sassen for the many discussions and advice provided. EEC acknowledges funding from the Center for Environmental Chemistry and Geochemistry at Pennsylvania State University and the Nancy Foster Scholarship Program at the National Oceanographic and Atmospheric Administration (NOAA). This work was supported by the U.S. Minerals Management Service, the NOAA National Undersea Research Program, and the National Science Foundation. **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** EEC, MAA, and CRF conceived and designed the experiments. EEC performed the experiments and analyzed the data. KS and RSA contributed reagents/materials/analysis tools. EEC, MAA, KS, and CRF wrote the paper. Citation: Cordes EE, Arthur MA, Shea K, Arvidson RS, Fisher CR (2005) Modeling the mutualistic interactions between tubeworms and microbial consortia. PLoS Biol 3(3): e77. DIC : dissolved inorganic carbon DOC : dissolved organic carbon ULS

Introduction {#s1} ============ Topographic representation is a hallmark of cortical organization: primary somatosensory cortex contains a somatotopic map of the body surface, primary visual cortex contains a retinotopic map of visual (retinal) space, and primary auditory cortex contains a cochleotopic map of sound frequency. The necessity of auditory cortex for normal sound localization (which is disrupted by cortical lesions \[[@pbio-0030078-b01],[@pbio-0030078-b02],[@pbio-0030078-b03]\]) strongly implies a cortical representation of auditory space. That representation has been reasonably expected to consist of a spatiotopic map, based on the existence of such maps in other sensory systems and on the view, proposed by Jeffress \[[@pbio-0030078-b04]\], that spatial processing in the auditory brainstem and midbrain might involve a "local code" consisting of topographic maps of interaural spatial cues. A local code, or "place code," is one in which particular locations in space, or the spatial cues that correspond to those locations, are represented by neural activity at restricted locations in the brain. Evidence for local coding of auditory space has been demonstrated in mammalian superior colliculus \[[@pbio-0030078-b05],[@pbio-0030078-b06]\] and in avian inferior colliculus (IC) \[[@pbio-0030078-b07],[@pbio-0030078-b08]\] and optic tectum (homologous to mammalian superior colliculus) \[[@pbio-0030078-b09]\]. Nevertheless, local spatial coding has not thus far been demonstrated in the mammalian ascending auditory pathway. If the Jeffress model is correct and a local code for spatial cues exists subcortically, one might anticipate local coding to be maintained in the cortex, where the various cues might finally be integrated into a coherent map of auditory space. Numerous studies, however, have failed to provide evidence for such a map. The spatial tuning of neurons is often characterized using rate--azimuth functions (RAFs), which specify the average response rate (spikes per trial or per second) as a function of stimulus location in the horizontal dimension. Throughout the auditory cortex, such functions typically exhibit broad peaks (up to 180° wide) that cover the contralateral hemifield and broaden further with increasing sound level \[[@pbio-0030078-b10],[@pbio-0030078-b11],[@pbio-0030078-b12],[@pbio-0030078-b13],[@pbio-0030078-b14]\]. Similar functions have been reported for cortical sensitivity to interaural cues \[[@pbio-0030078-b15],[@pbio-0030078-b16]\], and for spatial and interaural sensitivity in the auditory brainstem and midbrain \[[@pbio-0030078-b17],[@pbio-0030078-b18],[@pbio-0030078-b19],[@pbio-0030078-b20],[@pbio-0030078-b21],[@pbio-0030078-b22],[@pbio-0030078-b23],[@pbio-0030078-b24]\], thus questioning Jeffress\'s view of binaural processing in mammals. The emerging alternative view replaces the local code with a "distributed code," in which sound-source locations are represented by patterns of activity across populations of broadly tuned neurons \[[@pbio-0030078-b12],[@pbio-0030078-b24],[@pbio-0030078-b25]\]. In the past, we argued for a distributed spatial code in the auditory cortex in part because the broad spatial tuning of cortical neurons would seem to preclude the existence of a local code and also because individual neurons are able to transmit spatial information throughout much, if not all, of auditory space \[[@pbio-0030078-b25],[@pbio-0030078-b26]\]. At least implicitly, we have advocated a uniform distributed code, assuming that uniform sampling of space by RAF peaks is required for maximally accurate spatial coding. Spatial centroids of neurons in the posterior auditory field (PAF), for example, sample space more uniformly than neurons in the primary auditory field (A1), and we have suggested that this feature partially underlies the increased ability of ensembles of PAF neurons to accurately signal sound-source locations \[[@pbio-0030078-b14]\]. A number of observations demonstrate, however, that the auditory cortex samples space nonuniformly. RAFs are plotted for a selection of neurons in the dorsal zone (DZ) of auditory cortex in [Figure 1](#pbio-0030078-g001){ref-type="fig"}, to illustrate a common observation of location-sensitive auditory cortical neurons: the majority favor contralateral stimulation, and typically exhibit either "hemifield" or "axial" tuning \[[@pbio-0030078-b11]\], responding to stimuli located throughout contralateral space or near the acoustic axis of the contralateral pinna, respectively. A smaller number of ipsilaterally tuned units are also observed, the majority of which exhibit hemifield or axial tuning characteristics similar to those of contralateral units. In A1 and DZ, ipsilateral- and/or midline-tuned neurons may be arranged in bands---parallel to the tonotopic axis---that interdigitate with bands of contralaterally tuned cells \[[@pbio-0030078-b25],[@pbio-0030078-b27],[@pbio-0030078-b28],[@pbio-0030078-b29]\]. The overall preponderance of contralateral tuning among cortical units seems to justify the view that each hemisphere represents the contralateral spatial hemifield, a view that is also supported by the contralateral sound-localization deficits that follow auditory cortical lesions \[[@pbio-0030078-b02],[@pbio-0030078-b30],[@pbio-0030078-b31]\]. Even within a single hemifield, however, no strong evidence for a topographic representation has been reported, and the observation that many units share similar hemifield RAFs demonstrates a profound inhomogeneity in the way cortical populations sample auditory space. {#pbio-0030078-g001} Additional evidence that the cortical representation of auditory space is inhomogeneous comes from studies of the ability of cortical responses to classify stimulus locations. Stecker et al. \[[@pbio-0030078-b14]\] found that the responses of most spatially sensitive units in cat cortical areas A1 and PAF could accurately discriminate the lateral hemifield (left versus right) of a stimulus, but often confused locations within the hemifield. This is shown for six PAF neurons represented by confusion matrices in [Figure 2](#pbio-0030078-g002){ref-type="fig"}. Similarly, Middlebrooks et al. \[[@pbio-0030078-b12]\] measured median localization errors---based on neural-network analyses of responses in the second auditory field (A2) and the field of the anterior ectosylvian sulcus (AES)---between 37.5° ± 8.9° and 43.7°*±* 10.2*°*, just under the theoretical limit of 45*°* attainable through perfect left/right discrimination and within-hemifield confusion. Taken together, these results suggest that auditory space is represented within the cortex by a population of broadly tuned neurons, each of which is able to indicate the lateral hemifield from which a sound originated, but generally little more. ![Classification Performance of Accurate PAF Units from \[[@pbio-0030078-b14]\]\ Neural spike patterns were classified according to the stimulus location most likely to have elicited them. In each panel, a confusion matrix plots the relative proportion of classifications of each target azimuth (x-axis) to each possible response azimuth (y-axis). Proportions are indicated by the area of a circle located at the intersection of target and response locations. Example units were selected from among those transmitting the most spatial information in their responses. In each case, discrimination of contralateral azimuths (negative values) from ipsilateral azimuths (positive values) is apparent, accompanied by significant within-hemifield confusion. As such, neural responses are sufficient for left/right discrimination only, and the spatial information transmitted by the most accurate units tends not to be much greater than one bit per stimulus.](pbio.0030078.g002){#pbio-0030078-g002} Results/Discussion {#s2} ================== Preferred Locations Oversample Lateral Regions of Contralateral Space---Steepest RAF Slopes Straddle the Midline {#s2a} ---------------------------------------------------------------------------------------------------------------- The idea that sound locations are signaled by the peaks of RAFs, which tend to be centered deep within the lateral hemifields, is at odds with localization behavior, which shows greatest resolution near the interaural midline \[[@pbio-0030078-b32],[@pbio-0030078-b33]\]. An alternative view, however, has emerged for the processing of interaural time and level differences by cortical and subcortical neurons. In that view, locations are coded by the slopes, rather than the peaks, of rate--interaural-time-difference or rate--interaural-level-difference functions \[[@pbio-0030078-b22],[@pbio-0030078-b24],[@pbio-0030078-b34]\]. Moreover, these slopes appear aligned with the interaural midline and provide maximum spatial information in that region \[[@pbio-0030078-b20]\]. If a similar arrangement can explain the inhomogeneity of spatial sampling in the auditory cortex, then we would expect to find cortical RAF slopes to be steepest near the interaural midline as well. In this report, we compare the responses of neurons in primary auditory cortex (A1) and two higher-order auditory cortical fields (PAF and DZ) in the cat. Compared to A1, areas PAF and DZ exhibit spectrotemporally complex responses that are significantly more sensitive to variations in sound-source location \[[@pbio-0030078-b14],[@pbio-0030078-b25]\]. Therefore, these areas are the most likely candidate regions of cat auditory cortex for spatial specialization. PAF, in particular, appears necessary for sound localization by behaving cats \[[@pbio-0030078-b30]\]. [Figure 3](#pbio-0030078-g003){ref-type="fig"} depicts the distribution of preferred locations ("azimuth centroids"; see \[[@pbio-0030078-b14]\]) along with locations of peak RAF slopes in all three fields. As we have reported previously, centroid distributions in [Figure 3](#pbio-0030078-g003){ref-type="fig"} reveal a preponderance of contralateral sensitivity regardless of cortical area or stimulus level \[[@pbio-0030078-b14],[@pbio-0030078-b25]\]. Distributions of peak-slope location, however, are tightly clustered around the frontal midline (median*±* standard error \[see [Materials and Methods](#s3){ref-type="sec"}\] in A1, +15° ± 2.5°; in DZ, +5° ± 2.1°; in PAF, −5° ± 2.6°). Values in A1 fall significantly farther into the ipsilateral field than do those in DZ (*p* \< 0.0004) or PAF (*p* \< 0.0002), consistent with both the broader spatial tuning and less extreme azimuth centroids of A1 compared to PAF or DZ units \[[@pbio-0030078-b14],[@pbio-0030078-b25]\]. Overall, the positioning of RAF slopes near the interaural midline suggests that auditory space is sampled inhomogeneously by the cortical population; the midline represents a transition region between locations eliciting responses from populations of contralateral- and ipsilateral-preferring units. {#pbio-0030078-g003} Neural Response Patterns Discriminate Best across Midline {#s2b} --------------------------------------------------------- Modulation of spike count is generally the most salient location-sensitive feature of neural responses, especially when data are averaged over many trials. However, temporal features of the neural response---such as first-spike latency, the temporal dispersion of spikes, and specific temporal features such as prototyped bursts of spikes or periods of inhibition---could also play an important role in stimulus coding by cortical neurons, and we have studied this role using pattern-recognition analyses applied to spike patterns \[[@pbio-0030078-b12],[@pbio-0030078-b14],[@pbio-0030078-b35]\]. Here, we assess the ability of neural spike patterns to subserve pairwise discrimination of stimulus locations by adapting the pattern-recognition approach of Stecker et al. \[[@pbio-0030078-b14]\] to a discrimination paradigm. This approach is similar to the receiver-operating-characteristic analysis used to estimate interaural and/or spatial thresholds from neural spike counts \[[@pbio-0030078-b36],[@pbio-0030078-b37],[@pbio-0030078-b38]\], with the addition of spike-timing information. Given a spike pattern---a smoothed, bootstrap-averaged peristimulus time histogram (2-ms bins) that approximates the instantaneous probability of spike firing over the course of 200 ms following stimulus onset---elicited by stimulation from an unknown location in space, the algorithm estimates the relative likelihood that the pattern was evoked by a sound from each of the 18 tested locations. From these relative likelihoods, we compute the index of discriminability, *d′* \[[@pbio-0030078-b39]\], for each pair of stimulus locations. In [Figure 4](#pbio-0030078-g004){ref-type="fig"} (right), pairwise *d′* is plotted as a function of the midpoint and separation between paired stimuli for a single PAF unit; the contour *d′* = 1 (dashed line) indicates the spatial discrimination threshold. Note that in this example suprathreshold discrimination is possible at much narrower stimulus separations when the stimuli span the interaural midline (left/right discrimination) than in cases of front/back discrimination spanning +/− 90*°*, about which point many features of the neural response (e.g., spike rate and latency) are symmetrical. As a result, the minimum discriminable angle (MDA, defined as the minimum separation along the *d′* = 1 contour) of 25*°* is found at a best azimuth (BA, the midpoint location of the most discriminable pair) of −5*°*, near the frontal midline. {#pbio-0030078-g004} In [Figure 5](#pbio-0030078-g005){ref-type="fig"}, MDA is plotted as a function of BA for the entire population of A1, PAF, and DZ units in which discrimination thresholds could be calculated. Overall, two main features of the results should be noted. First, despite the broad azimuth tuning of cortical neurons, the majority can discriminate between stimuli separated by less than 40*°*. A number of neurons successfully discriminate even smaller separations---especially in DZ, in which the median MDA (30.5*° ±* 2.5*°*) is significantly smaller than in A1 (40*° ±* 2.6°; *p* \< 0.007) or PAF (43*° ±* 3.8*°*; *p* \< 0.0002). Note that MDAs of even the most sensitive units exceed behavioral estimates of 5°--6*°* minimum audible angles in cats \[[@pbio-0030078-b40]\], but likely underestimate the true neuronal performance because loudspeaker separations were tested in minimum steps of 20*°*, and thus discrimination at smaller separations can only be assessed through extrapolation. Second, the distribution of BAs is tightly clustered around the interaural midline, with 50% of BA values falling within 18.5*°* (PAF), 25*°* (A1), or 26*°* (DZ) of the 0*°* or 180*°* azimuth. Note that this does not mean that units cannot discriminate off-midline locations. It does indicate, however, that the majority of units capable of discriminating between stimulus azimuths do so best for location pairs near the interaural midline. Very few units exhibit BAs located far within either lateral hemifield, although A1 units exhibit significantly more ipsilateral BA values (median, +14.5*° ±* 3.9*°*) than those in PAF (0*° ±* 2.9*°*; *p* \< 0.004) or DZ (5*° ±* 4.3*°*; *p* \< 0.05). As with the analysis of RAF slopes, the pairwise discrimination data reveal an inhomogeneous arrangement of spatial sampling by neurons in the cortical population. Accurate discrimination is found where RAF slopes are steepest (the midline), rather than where units respond most strongly (the lateral poles). {ref-type="sec"}). Symbols indicate the cortical area of each unit. Left and lower panels plot distributions of MDA and BA (in numbers of units per rectangular 20*°* bin), respectively.](pbio.0030078.g005){#pbio-0030078-g005} An Opponent-Channel Code for Auditory Space? {#s2c} -------------------------------------------- We have demonstrated quantitatively that the representation of auditory space in the cortex is inhomogeneous, consisting mainly of broadly tuned neurons whose responses change abruptly across the interaural midline. The population of auditory cortical neurons, then, appears to contain at least two subpopulations broadly responsive to contralateral and ipsilateral space. Neurons within each population exhibit similar spatial tuning and thus appear redundant with respect to spatial coding. The similarity of spatial tuning of units in these populations stands in contrast to their more profound differences in frequency tuning, for example. Each subpopulation, or "spatial channel," is capable of representing locations on the slopes of their response areas (i.e., across the interaural midline) by graded changes in response---a "rate code" for azimuth, generalized to incorporate spatially informative temporal features of the cortical response \[[@pbio-0030078-b14],[@pbio-0030078-b35],[@pbio-0030078-b41]\]. In other words, each spatial channel encodes space more or less panoramically, as we have argued previously for individual cortical neurons \[[@pbio-0030078-b12]\], although it now seems clear that some regions of space are represented with greater precision than others. In the past, we have argued that auditory space is encoded by patterns of activation across populations of such panoramic neurons. Here, we amend that view---which remains tenable---to reflect the observed inhomogeneity of spatial sampling in the cortex and account for differences in coding accuracy of midline and other locations. Following the proposals of von Békésy \[[@pbio-0030078-b42]\] and van Bergeijk \[[@pbio-0030078-b43]\] regarding interaural coding in the brainstem, we propose that auditory space is encoded specifically by differences in the activity of two broad spatial channels corresponding to subpopulations of contralateral and ipsilateral units within each hemisphere (i.e., by a left/right opponent process). We will refer to this proposal as the opponent-channel theory of spatial coding in the auditory cortex. An important consequence of the opponent-channel theory is that spatial coding may be robust in the face of changes in stimulus level. As is evident from past work, an important constraint on spatial coding in the cortex is the level dependence of many neurons\' tuning widths, such that sharp tuning is seen predominantly for low-level stimulation. For example, a number of narrowly tuned units in [Figure 3](#pbio-0030078-g003){ref-type="fig"} exhibit locations of peak slope that closely track their centroids at 20 dB above threshold, and one could argue that such units form the basis of a local (e.g., topographic) spatial code when stimulus levels are low. Such a code, however, would be significantly impaired by increases in stimulus level---predicting that sound localization should be most accurate at low levels. That prediction is not borne out in psychophysical tests \[[@pbio-0030078-b44],[@pbio-0030078-b45]\], and we have argued that spatial coding in the auditory cortex must employ relatively level-invariant features of the neural response \[[@pbio-0030078-b12]\]. Rather than relying on such features as they naturally occur, the opponent-channel mechanism constructs level-invariant features by comparing the activity of neurons that respond similarly to changes in level but differentially to changes in location, similarly to the coding of color by opponent-process cells in the visual system \[[@pbio-0030078-b46]\]. To illustrate the level invariance achieved by opponent-process coding, we analyzed the ability of cortical population responses to signal sound-source locations in the frontal hemifield under different stimulus-level conditions. The analysis (see [Materials and Methods](#s3){ref-type="sec"}) is simplified in a number of ways---for example, it utilizes a simple linear decision rule that weights contralateral and ipsilateral input equally, sums across multiple neurons within each subpopulation (ignoring any complexity of neural circuitry), combines data across different cortical areas known to exhibit different spatial sensitivities, and reduces each neural response pattern to a single overall spike count---but serves as a "proof of concept" that differences between the responses of neural subpopulations with quasi-independent spatial tuning can be used to estimate sound-source locations in an unbiased manner when stimulus levels vary, whereas the individual population responses cannot. Population responses (means of normalized spike rate across neurons in a population) to stimuli varying in location and level were computed separately for subpopulations composed of contralateral units or ipsilateral units in our sample of recordings in A1, PAF, and DZ. These subpopulations correspond to hypothetical "left" and "right" channels of a spatial coding mechanism. Classification of stimulus locations was based on either one of the subpopulation responses or the difference between the two, and involved linear matching to templates computed from a separate training set \[[@pbio-0030078-b14]\]. Population and difference RAFs are plotted in [Figure 6](#pbio-0030078-g006){ref-type="fig"} (left), along with confusion matrices (similar to [Figure 2](#pbio-0030078-g002){ref-type="fig"}) for sound-source classification based on each (right). Relatively accurate classification is exhibited by both subpopulation responses and by their difference when test and training sets reflect the same stimulus level. When training and test sets differ, however, responses are systematically biased. After training with 20-dB stimuli, localization of 40-dB stimuli by the contralateral subpopulation is biased toward the contralateral hemifield, because 40-dB ipsilateral stimuli and 20-dB contralateral stimuli elicit similar responses. Similarly, when trained with 40-dB stimuli, localization of 20-dB stimuli is biased toward the ipsilateral hemifield. This pattern is clear in the responses of both the contralateral-preferring and ipsilateral-preferring subpopulations. Classification based on their difference, however, is relatively unbiased. Significant undershoot (central responses for peripheral stimulus locations) results from compression of population RAFs by intense sound. While we know of no behavioral data relating to the effects of stimulus level on sound localization by cats, undershoot has been reported in numerous studies of their localization behavior \[[@pbio-0030078-b47],[@pbio-0030078-b48],[@pbio-0030078-b49]\]. Such undershoots, however, need not be assumed to reflect a limitation of the underlying neural representation of auditory space. {ref-type="sec"}) are plotted as a function of azimuth (x-axis) for stimuli presented 20 dB (red) and 40 dB (blue) above unit thresholds. Population responses were computed separately for subpopulations composed of contralateral units (top) or ipsilateral units (middle) corresponding to hypothetical "left" and "right" channels of an opponent-channel spatial coding mechanism. The difference (bottom) between responses of the two subpopulations is more consistent across stimulus level than is either subpopulation response alone. Error bars indicate the standard deviation of responses across 120 simulated trials.\ Right: stimulus--response matrices (confusion matrices; see [Figure 2](#pbio-0030078-g002){ref-type="fig"}) showing the proportion (area of black circle) of responses to a given (unknown) stimulus azimuth (x-axis) classified at each response azimuth (y-axis). Classification assigned each neural population response in the "test" set to the stimulus azimuth whose mean population response in an independently selected set of "training" trials was most similar. In some conditions, test and training trials were drawn from the same set of (matching level) trials: 20 dB (first column) or 40 dB (far right column). In others, test and training trials reflected different-level stimuli: 40-dB test stimuli classified based on a 20-dB training set (second column), or 20-dB test stimuli classified based on a 40-dB training set (third column). The contralateral and ipsilateral subpopulation responses (top and middle rows) accurately localize fixed-level stimuli, but are strongly biased when tested at non-trained stimulus levels. In contrast, the difference between responses (bottom row) remains relatively unbiased in all conditions, although responses to stimuli at untrained levels do exhibit compressed range and increased variability of classification.](pbio.0030078.g006){#pbio-0030078-g006} Based on its manner of level-invariant spatial coding, it seems clear that an opponent-channel mechanism should behave similarly in the presence of any stimulus change (e.g., in frequency, modulation, or bandwidth) that acts to increase or decrease the response of both channels. This suggests an efficient means for combining spatial information with information about other stimulus dimensions. This general principle of opponent-process coding should hold in any case where both channels exhibit similar sensitivity to the nuisance dimension (level, frequency, etc.) but dissimilar sensitivity to space, and illustrates one strength of the opponent-channel coding strategy: the ability to recover spatial information from the responses of neurons that are strongly modulated by other stimulus dimensions. As long as some of the cortical neurons involved in coding a particular acoustic feature are contralaterally driven and others are ipsilaterally driven, the spatial location of that feature can be computed without imposing additional distortion of its neural representation. Note that the opponent-channel theory as presented here involves contralateral and ipsilateral channels within each hemisphere. This feature is based on the observation of both types of neurons in a single hemisphere, and on the results of unilateral cortical lesions, which produce localization deficits mainly in contralesional space \[[@pbio-0030078-b02],[@pbio-0030078-b30],[@pbio-0030078-b31],[@pbio-0030078-b50]\]. The lesion data prevent us from considering opponent-channel mechanisms that place each channel in a separate hemisphere (e.g., left-hemisphere contralateral units versus right-hemisphere contralateral units) because in that case unilateral lesions should abolish localization throughout the entire acoustic field. As proposed here, however, the opponent-channel mechanism in either hemisphere should be capable of coding locations throughout space, not just in the contralateral hemifield. This would suggest that only bilateral lesions could produce localization deficits, which is also not the case. At this point, we can merely speculate that auditory cortical structures in each hemisphere provide input only to those multimodal spatial or sensorimotor structures that subserve localization behavior in contralateral space and, furthermore, that these inputs cannot be modified in adulthood following cortical lesions. General Discussion {#s2d} ------------------ In summary, the available data suggest that space is sampled nonuniformly in all fields of auditory cortex, with the majority of neurons responding broadly within one hemifield and modulating their responses abruptly across the interaural midline. Consistent with this view, we found cortical responses to be most sensitive to changes in stimulus azimuth at midline locations. Cortical neurons\' RAFs tend to be steepest near the midline even though their preferred locations are found distributed throughout the contralateral hemifield. Spatial discrimination by neural responses is also best at or near the interaural midline. Results of both analyses are compatible with the existence of a limited number of spatial channels in the cortex, and incompatible with either a uniform distributed representation or a local representation (e.g., a topographic map). The relative paucity of units with sharp tuning peaking near the midline strongly suggests that behavioral sound-localization acuity is mediated by the slopes and not the peaks of spatial receptive fields. In this report, we consider a model of spatial coding based on differences in the response rates of two broad spatial channels in the auditory cortex. It is similar to the mechanism proposed by Boehnke and Phillips \[[@pbio-0030078-b51]\] to account for differences in human gap detection when gaps are bounded by auditory stimuli occurring in the same or opposite hemifields. In each proposal, neural response rates are compared across channels, but each is also consistent with information encoded in the relative response timing of cortical neurons \[[@pbio-0030078-b25],[@pbio-0030078-b52]\]. Although the psychophysical and physiological data seem to agree on a two-channel mechanism, it is important to note that in this study, we treat units that respond more strongly to forward than rearward locations ("axial" units) as equivalent to units that respond equally to both quadrants ("hemifield" units). Similarly, we do not specifically examine the small number of units that respond best to midline locations. Distributed coding of interaural intensity by neural populations differing in binaural facilitation has been suggested previously \[[@pbio-0030078-b24]\]; similarly, populations of midline and/or axial units could be treated separately in a three-, four-, or five-channel opponent model of spatial coding. Such a model would follow the general principles of opponent-channel coding described here, but might differ in its ability to accurately code locations over wide regions of azimuth (see \[[@pbio-0030078-b24]\]). That the representation of space appears inhomogeneous in both primary and higher-order auditory cortical fields argues against the existence of a topographic "space map" within sensory cortex, pushing the emergence of any such map further into central structures than previously expected. The processing of interaural cues begins at the level of the superior olivary complex, but the integration of such cues into a complete topographic map of auditory space is presumed to begin with processing at the level of the IC or cortex. The suggestion that interaural cues are represented by a limited number of binaural channels in the IC \[[@pbio-0030078-b22]\] seems to imply that the space map must emerge at the level of auditory cortex or beyond, and the results of this study, along with others \[[@pbio-0030078-b15],[@pbio-0030078-b16]\], suggest that a "limited channel" code is maintained throughout primary and non-primary fields of the auditory cortex as well. PAF, in particular, appears to sit at the top of the auditory cortical processing hierarchy \[[@pbio-0030078-b53]\] but is similar to primary auditory cortex (A1) in this regard. We should note that spatial coding must subserve at least two distinct behavioral tasks, namely, the discrimination of sound-source locations and the localization of individual sources (e.g., orientation, or pointing). Much of the current discussion has focused on aspects of spatial coding relevant to discrimination, and on the observation that the RAF slopes of cortical neurons are better suited to the discrimination of nearby locations than are their broad RAF peaks. Nevertheless, we are interested in general mechanisms of spatial representation, and argue that cortical neurons\' broad spatial tuning suggests that neither aspect of sound localization is likely mediated by RAF peaks in cat cortex. This stands in contrast to the neural mechanism for sound localization in the IC of the owl, where sharp circumscribed spatial receptive fields form a place code for localization \[[@pbio-0030078-b07]\]. Owls\' behavioral discrimination of spatial locations, however, is sharper than these neural receptive fields, and appears---as in mammals---to be mediated by receptive-field slopes \[[@pbio-0030078-b38]\]. Thus, the owl makes use of place and rate codes for different behavioral tasks. The cat\'s auditory cortex, on the other hand, lacks the sharp spatial tuning necessary for map-based localization, so one coding strategy underlies both types of behaviors. It seems clear that these different coding strategies in owls and cats necessitate different mechanisms for generating motor responses and orienting to sound sources. The owl\'s space map exhibits a straightforward correspondence between restricted neural activity and locations in space, which might be ideal for computing audiovisual correspondence but requires further translation into motor coordinate systems before action can take place. It is possible that the opponent-channel code is transformed into a similar auditory space map within multisensory or sensorimotor areas, that is, not within auditory cortex itself. Alternatively, opponent-channel population codes in the auditory domain might be directly transformed into population codes in the motor domain without an intervening map-like representation. In either case, we could argue that the fundamental mode of spatial coding within the auditory system per se is non-topographic. In fact, it might be that auditory spatial topography is an emergent property of widespread neural populations and is evident only in perception and behavior, not in the physiology of single neurons. In considering the relative advantages of opponent-channel spatial coding within the cortex, one might wonder whether the formation of a spatiotopic map would be necessary or desirable. As described above, the opponent-channel mechanism could subserve behavior without an intervening map, and it provides an efficient means of combining information about space with information about other stimulus features. In this regard, at least, the opponent-channel mechanism solves---or simply avoids---the so-called binding problem \[[@pbio-0030078-b54]\] of how multiple stimulus features can be associated to create a unified neural representation. It does so without recourse to specialized mechanisms for binding \[[@pbio-0030078-b55]\] and without an explosion in the number of neurons necessary for a complete combinatorial code \[[@pbio-0030078-b56]\]. So long as feature maps (e.g., of frequency) contain neurons of each class (i.e., contralateral and ipsilateral), the spatial position of any particular feature can be reconstructed without the difficulty of binding activity in one feature map (frequency) with that in another (location). Finally, the three cortical fields studied in this report exhibited similar evidence for an opponent-channel mechanism, despite previously reported differences in their spatial sensitivity \[[@pbio-0030078-b14]\]. Although such differences appear modest when assessed physiologically, studies indicate that some fields are more critical for localization behavior than others \[[@pbio-0030078-b30]\]. An intriguing question for future research involves cortical fields---such as the anterior auditory field---that are not necessary for accurate localization. Are spatial channels maintained in such fields, or are they combined to produce space-invariant representations of other stimulus features? Materials and Methods {#s3} ===================== {#s3a} Data analyzed for this report were collected from extracellular recordings of 254, 411, and 298 units in areas A1, PAF, and DZ (respectively) of the cortex of chloralose-anesthetized cats \[[@pbio-0030078-b14],[@pbio-0030078-b25]\]. Methods of animal preparation, stimulus delivery, unit recording, and basic analysis have been described previously \[[@pbio-0030078-b14]\], and were approved by the University of Michigan Committee on Use and Care of Animals. Stimuli were delivered from loudspeakers placed in the free field, and consisted of 80-ms broadband noise bursts presented at levels 20--40 dB above unit threshold. Stimulus locations spanned 360*°* of azimuth in 20*°* steps, and are identified by angular distance from the frontal midline (0*°*). Positive azimuths increase to to the right (ipsilateral to the recording site), whereas negative values correspond to contralateral locations on the cat\'s left side. Unit activity was recorded extracellularly from the right cerebral hemisphere using 16-channel electrode arrays ("Michigan probes"), and spikes were sorted off-line based on principal-components analysis of their waveshapes. ### Locations of peak slope and centroids {#s3a1} Each unit\'s preferred location was characterized by the azimuth centroid of response (dark blue crosses in [Figure 3](#pbio-0030078-g003){ref-type="fig"}; see \[[@pbio-0030078-b14]\]); this is the spike-count-weighted average of contiguous stimulus locations eliciting a normalized response at or above 75% of maximum spike count per stimulus presentation. We additionally determined the locations of peak slope for each unit by smoothing its RAF (circular convolution with a 40*°* boxcar) and calculating the first spatial derivative of the result. Maximum and minimum values of the derivative indicate two peak-slope azimuths for each unit (black circles and endpoints of red horizontal lines in [Figure 3](#pbio-0030078-g003){ref-type="fig"}). ### Spatial discrimination by neural response patterns {#s3a2} Analyses of pairwise spatial discrimination (see [Figures 4](#pbio-0030078-g004){ref-type="fig"} and [5](#pbio-0030078-g005){ref-type="fig"}) employed a statistical pattern-recognition algorithm \[[@pbio-0030078-b14]\] to estimate the relative likelihood of each stimulus location, given the temporal pattern of neural response to a single (unknown) stimulus. We computed, for each pair of locations θ~1~ and θ~2~ in the loudspeaker array, the index of pairwise discriminability *d′* \[[@pbio-0030078-b39]\] based on the estimated relative likelihoods: where *z*(*P*) represents scaling to the standard normal distribution and the probability *P* of responding "1" is given by the (estimated) relative likelihood *l* of location θ~1~ (versus θ~2~), conditional on the actual stimulus location θ*~i~*. The analysis produces a map of *d′* between each pair of stimulus locations, plotted in coordinates of stimulus separation and overall location in [Figure 4](#pbio-0030078-g004){ref-type="fig"}. The map was interpolated to find a contour of *d′* = 1, which we define as threshold discrimination. The smallest stimulus separation along the threshold contour defines the MDA, and the overall location of that stimulus pair defines the unit\'s BA. Symbols in [Figures 4](#pbio-0030078-g004){ref-type="fig"} and [5](#pbio-0030078-g005){ref-type="fig"} indicate values of MDA and BA for individual units. ### Evaluation of a simple population code for space {#s3a3} To assess the level invariance of opponent-channel coding, we analyzed a simplified model of population spatial coding in the cortex. For each neural unit in a channel (e.g., a subpopulation of contralateral-preferring units), we accumulated a list of responses (spike counts normalized to the maximum response across all trials) on each trial with a given combination of stimulus azimuth and level. Azimuths were confined to the frontal hemifield (−80° to +80°) to avoid front--back confusions, which obscure but do not alter the appearance of bias in classification responses, and levels were either 20 or 40 dB above individual unit thresholds. We then computed population responses by randomly selecting one trial (with matching stimulus azimuth and level) from each unit and computing the mean of individual responses. We repeated the selection process 120 times for each combination of azimuth and level to simulate a set of 120 population "trials." The mean of these population responses for each stimulus is plotted on the left in [Figure 6](#pbio-0030078-g006){ref-type="fig"}. Separate "training" and "test" sets of population responses were computed by this method and used to assess the ability of subpopulations to classify stimulus locations. Individual population responses in the test set were classified to the azimuth with the most-similar mean population response across the training set. Confusion matrices in [Figure 6](#pbio-0030078-g006){ref-type="fig"} plot the proportion of test-set responses assigned to each stimulus azimuth. In some conditions, test and training sets were drawn from the same trials (matching level); in other conditions, training and test sets differed in stimulus level. We tested classification based on responses of a contralateral subpopulation, an ipsilateral subpopulation, and on the difference between subpopulation responses. Contralateral and ipsilateral subpopulations were composed of all units with centroids falling farther than 30° into the corresponding hemifield in our sample of A1, PAF, and DZ units. Differences were computed from the two subpopulation responses on a trial-by-trial basis, and classification was tested in the same manner as for the population responses themselves. ### Statistical procedures {#s3a4} Tests of statistical significance in this study were conducted using a 5,000-permutation bootstrap test (see \[[@pbio-0030078-b14]\] for details), reported to one significant digit. Standard error of the median, where reported, was obtained using a 2,000-permutation bootstrap, drawing *N* (the total number of data points) samples from the data with replacement on each permutation and recomputing the median. Distributions in [Figures 3](#pbio-0030078-g003){ref-type="fig"} and [5](#pbio-0030078-g005){ref-type="fig"} were computed by kernel density estimation (convolution) with a 20*°* rectangular window to obtain a continuous function of units per 20*°* bin. We thank Ewan Macpherson for assistance with data collection, Zekiye Onsan for technical and administrative support, and three anonymous reviewers for insightful comments. Funding was provided by the National Science Foundation (grant DBI-0107567) and the National Institute on Deafness and Other Communication Disorders (grants R01 DC00420, P30 DC05188, F32 DC006113, and T32 DC00011). Recording probes were provided by the University of Michigan Center for Neural Communication Technology (CNCT, NIBIB P41 EB002030). **Competing interests.** The authors have declared that no competing interests exist. **Author contributions.** GCS, IAH, and JCM conceived and designed the experiments. GCS and IAH performed the experiments and analyzed the data. GCS contributed reagents/materials/analysis tools. GCS wrote the paper. ¤Current address: Human Cognitive Neurophysiology Lab, Department of Veterans Affairs Research Service, VA Northern California Health Care System, Martinez, California, United States of America Citation: Stecker GC, Harrington IA, Middlebrooks JC (2005) Location coding by opponent neural populations in the auditory cortex. PLoS Biol 3(3): e78. A1 : primary auditory field BA : best azimuth DZ : dorsal zone IC : inferior colliculus MDA : minimum discriminable angle PAF : posterior auditory field RAF : rate--azimuth function