The genomics of economically important marine bivalves is studied to provide better understanding of the molecular mechanisms underlying their different reproductive strategies. The recently available gonad transcriptome of the black-lip pearl oyster Pinctada margaritifera is a novel and powerful resource to study these mechanisms in marine mollusks displaying hermaphroditic features. In this study, RNAseq quantification data of the P. margaritifera gonad transcriptome were analyzed to identify candidate genes in histologically-characterized gonad samples to provide molecular signatures of the female and male sexual pathway in this pearl oyster. Based on the RNAseq data set, stringent expression analysis identified 1,937 contigs that were differentially expressed between the gonad histological categories. From the hierarchical clustering analysis, a new reproduction model is proposed, based on a dual histo-molecular analytical approach. Nine candidate genes were identified as markers of the sexual pathway: 7 for the female pathway and 2 for the male one. Their mRNA levels were assayed by real-time PCR on a new set of gonadic samples. A clustering method revealed four principal expression patterns based on the relative gene expression ratio. A multivariate regression tree realized on these new samples and validated on the previously analyzed RNAseq samples showed that the sexual pathway of P. margaritifera can be predicted by a 3-gene-pair expression ratio model of 4 different genes: pmarg-43476, pmarg-foxl2, pmarg-54338 and pmarg-fem1-like. This 3-gene-pair expression ratio model strongly suggests only the implication of pmarg-foxl2 and pmarg-fem1-like in the sex inversion of P. margaritifera. This work provides the first histo-molecular model of P. margaritifera reproduction and a gene expression signature of its sexual pathway discriminating the male and female pathways. These represent useful tools for understanding and studying sex inversion, sex differentiation and sex determinism in this species and other related species for aquaculture purposes such as genetic selection programs.
Citation: Teaniniuraitemoana V, Huvet A, Levy P, Gaertner-Mazouni N, Gueguen Y, Le Moullac G (2015) Molecular Signatures Discriminating the Male and the Female Sexual Pathways in the Pearl Oyster Pinctada margaritifera. PLoS ONE 10(3): e0122819. https://doi.org/10.1371/journal.pone.0122819
Academic Editor: Hanping Wang, The Ohio State University, UNITED STATES
Received: October 9, 2014; Accepted: February 24, 2015; Published: March 27, 2015
Copyright: © 2015 Teaniniuraitemoana et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This study was financially supported by ANR funding (POLYPERL project, reference ANR-11-AGRO-006) and by the French Government through the Contract Project BiodiPerl 2011-2013. VT is a Ph.D. student financially supported by Ifremer and holds a grant from la Délégation à la Recherche de Polynésie française. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The black-lip pearl oyster Pinctada margaritifera (Linnaeus, 1758) var. cumingi (Jameson, 1901) is a benthic bivalve occurring in subtropical and tropical coral reefs. It is widely distributed in the Indo-Pacific region: throughout the Equatorial zone, down to Northern Australia and extending up to the south of Japan .
For French Polynesia, the black pearl industry, based on the cultured of P. margaritifera, is the second most important economic resource. Over the last three decades, research projects have been conducted to support sustainable development of pearl farming and to improve the quality of the pearls produced , using genetic methods. Genetic selection programs have been encouraged because it was demonstrated that the pearl quality is dependent on the genetics of the graft donor oyster . The aim of the P. margaritifera selection program, is to breed families of graft donor oysters selected for their capacity to produce high quality pearls and/or particular colors and/or rapid growth [4–6].
The successful hatching of selected spat depends on the production of gametes and embryos from synchronous selected breeders raised in laboratory conditions. Reproduction of pearl oysters under controlled conditions mainly depends on the availability of females, which are unfortunately not abundant among farmed oysters due to the protandrous hermaphroditism of this species . As with other pearl oysters, P. margaritifera shows a consecutive switch in sex: individuals may change gender (from male to female) from the end of their second year onwards . However, Thielley (1993)  showed that changes from female to male could occur when conditions, either natural (temperature or food) or unnatural (handling or cleaning), were stressful. These observations would better correspond to alternate hermaphroditism. In all cases, simultaneous hermaphrodites and animals with undetermined status were uncommon. Knowledge of the underlying physiological mechanisms and factors controlling sex ratio [8,10,11] especially for the female pathway, are essential to establish a genetic improvement program.
Sex ratio is the combined result of sex determination and sex differentiation. In the animal kingdom, sex determination, the process by which the sex (gender, male or female) of an individual is established, can be genetic (genetic sex determination, GSD), environmental (environmental sex determination, ESD), or result from an interaction of both of these factors . These genetic and/or environmental sex determining switches lead to sex differentiation, the process by which specific molecular cascades transform an undifferentiated gonad into a testis or an ovary . Although these mechanisms are well known in vertebrates, data on sexual determinism and sex differentiation in marine mollusks are still poorly understood. In bivalve mollusks, sex determination systems appear complex and dynamic, from diocy to hermaphroditism . In the pacific oyster Crassostrea gigas, an alternative protandrous hermaphrodite bivalve, genetic models suggest that sex determination would be controlled by a major gene, which may be under environmental influence [14,15]. To date, several genes homologous to sex-determining pathway genes in model species have been identified in C. gigas such as Cg-DMl and Cg-SoxE homologous to male sex-determining pathway genes, and Cg-β-catenin, Cg-foxl2 and its natural antisense transcript Cg-foxl2os homologous to female sex-determining pathway genes. Their expression profiles in this species strongly suggest their implication in male and female gonad differentiation but no evidence was found of a role in sex determination. However their expression profiles may imply the existence of a sex determination time window at the end of the gametogenetic cycle [16–20]. In most bivalves, the main environmental factors affecting reproduction, and probably gender, are temperature, food availability [21,22] and, to a lesser degree, photoperiod . In the Pacific oyster Crassostrea gigas, Enríquez-Díaz et al. (2009)  showed that while intensity was influenced by the quantity of available food, the timing and speed of gametogenesis were regulated by temperature.
During recent years, genomics of marine bivalves of major economic interest has been studying in order to provide better understanding of the molecular mechanisms underlying reproduction, including reproductive cycle, sex differentiation and sex determination [25–28]. Recently, Teaniniuraitemoana et al. (2014)  performed a gonad transcriptome analysis of P. margaritifera, using Illumina-based RNAseq, unraveling some molecular mechanisms involved in sex determination, sex differentiation and gametogenesis. In this study, the authors revealed the importance of pmarg-dmrt, pmarg-fem1-like and pmarg-foxl2 as starting points for further functional studies on the sex-determining cascade of the pearl oyster.
In the present study, using the previously published RNAseq dataset , we applied a more stringent method of statistical selection on the differential expression levels in order to draw up a short list of candidate genes of the male and the female sexual pathway. These candidates were next assayed by real-time PCR in different sexes and stages of pearl oyster gonad. Through these analyses, we obtained new insights into the reproduction of P. margaritifera based on a dual histo-molecular analytical approach, allowing a model to be built based on expression ratio of three gene pairs, and predictions of sexual pathway to be made for individuals of this hermaphrodite species. This study offers molecular signatures of the female and male sexual pathways that will be useful for future research on reproduction in P. margaritifera, especially research on sex differentiation and sex determination.
Materials and Methods
The authorization (No. 4226) of the pearl oyster translocation from Takaroa atoll (14°26’59.12”S, 144°58’19.91”W, Tuamotu Archipelago, French Polynesia) to the Vairao lagoon (Ifremer marine concession No. 8120/MLD: 17°48’26.0”S, 149°18’14.4”W, Tahiti, French Polynesia) has been issued by the Ministry of Marine Resources on September 27, 2012 after the favorable notice given by the town councils of Takaroa and Vairao, respectively. At the departure of Takaroa atoll, pearl oysters were packed in ice-box for shipment by air. Upon their arrival at Vairao, pearl oysters were immersed for 30 min in a hyper-salted water bath (120 g of salt L-1 water) following the recommendations with the authorization for transfer in order to eliminate anemones and sea squirts. Then, the pearl oysters are stored in the lagoon of Vairao for 3.5 months to recover physiologically of air transport and hyper-salted treatment. In any case, this study involves protected or endangered species.
Animal and tissue sampling
Three-year-old adult P. margaritifera of an average size of 118 mm ± 8.67 from Takaroa atoll, acclimated in the Vairao lagoon for three and a half months, were brought to the Ifremer laboratory in Tahiti, French Polynesia, in April 2013. For each oyster, gonad tissues were sampled, making a transversal section of the gonadic area for histological examination. A piece of the rest of the gonad was conserved in RNAlater (Qiagen) (50 mg mL-1) and stored at -80°C until RNA extraction.
Gonad tissues of 44 oysters were sampled for histological examination using methods described by Fournier et al. (2012) . After dissection, each visceral masse was fixed in formalin solution 10% diluted in seawater for 48 h before being transferred in ethanol at 70% for 48 h for preservation. Then the visceral masses were cut according to their sagittal plane with a microtome blade and were dehydrated through a graded ethanol series, embedded in paraffin, sectioned at 3–4 μm on a rotary microtome, stained with Giemsa dye and, finally, mounted on microscope slides. Gonad differentiation stage and sex were determined under a light microscope and samples were classified according to the different categories of gonadic tissues previously described in Teaniniuraitemoana et al. (2014) : male and female at “Early” stage (gonad in early gametogenesis; Male: n = 4; Female: n = 3), “Intermediate” stage (gonad developing; Male: n = 4; Female: n = 4), at “Mature” stage (oyster ready to spawn; Male: n = 4; Female: n = 2), “Regressed” stage (gonad has stopped generating gametes; Male: n = 9; Female: n = 8); and “Undetermined” (gonad contains no gametes at all; n = 6). No gonads in “Inversion” (where the gonad presents both male and female gametes) were observed in this sample.
Candidate marker genes selected from the analysis of RNAseq dataset were assayed by real-time PCR on the 44 gonad samples. After removing RNAlater by pipetting and absorption, total RNA of each individual was extracted using TRIzol Reagent (Invitrogen) and treated with DNAse I using a DNA-free Kit (Ambion) following manufacturer’s instructions. RNA concentrations were measured on an ND-100 spectrophotometer (Nanodrop Technologies) at 260 nm, using the conversion factor 1 OD = 40 μg mL-1 RNA. For each sample, 0.5 μg of total RNA were reverse-transcribed using Transcriptor First Strand cDNA Kit (Roche) and amplified by real-time PCR on a Stratagene MX3000P. The amplification reaction contained 12.5 μL 2X SYBR green qPCR Master Mix (Stratagene), 10 μL cDNA template, and 2.5 μL of primers (4 μM) in a final volume of 25 μL. Each run included a positive cDNA control and a blank control (water) for each primer pair. Relative gene expression was calculated using two reference genes, ef1a and gapdh1, by the 2-ΔCt method , as follows: Relative expression(target gene, sample x) = 2^-(Cttarget gene, sample x—Ctreference gene, sample x). The relative stability of the ef1a and gapdh1 combination, considering the sex and reproductive stage of each gonad sample, was confirmed using NormFinder  (Stability value for best combination = 0.288). PCR efficiency (E) was estimated for each primer pair by determining the slopes of standard curves obtained from serial dilution analysis of a cDNA to ensure that E ranged from 90 to 110%. The primers used for amplification are listed in S1 Table.
All the statistical analyses were performed using R v2.15.2, an environment and a language for statistical computing .
Differential expression (DE) analysis.
RNAseq quantification data of P. margaritifera gonads were taken from the study of Teaniniuraitemoana et al. (2014)  (Bioproject PRJNA229186—Genbank accession number SRP033217) who analyzed the gonad transcriptome based on the Illumina sequencing of 36 cDNA libraries, each corresponding to gonad samples of different reproductive stages and sexes.
The differential expression level of contigs between different gonadic categories was tested using the DESeq package v1.11.3  as described in Teaniniuraitemoana et al. (2014) . In the present study, the differential expression of contigs was considered as statistically significant when the value of FoldChange was > 10 or < 0.1 i.e., when there was a tenfold change in expression under one of the conditions; and a padj < 0.001, p-value adjusted with a false-discovery rate (FDR) correction for multiple testing by the Benjamini-Hochberg method .
Relative expression analysis.
Normality and homoscedasticity of gene relative expression data were checked using Shapiro-Wilk and Bartlett’s tests. One-way ANOVA was performed followed by Tukey’s multiple comparison tests to determine expression differences of candidate marker genes between the gonadic categories considered.
Hierarchical clustering analysis.
Hierarchical clustering was applied firstly to RNAseq samples and secondly to real-time PCR samples in order to cluster them according to the similarity of expression pattern of the statistically significant differentially expressed contigs and expression pattern of the selected candidate marker genes, respectively.
We adapted and performed a multivariate regression tree (MRT), widely used in the domain of ecology for modeling species-environment relationships , to evaluate the hierarchical importance of the effect of relative gene expression on the oyster sexual pathway. Divisions in the MRT were determined by cross-validation. We performed these analysis with the mvpart package v1.6–1  using the default parameters of the rpart function.
Differential expression analysis reveals new insight into P. margaritifera reproduction
To provide an overall view of the transcriptional changes in the gonad of P. margaritifera, a stringent differential expression analysis was performed between the ten different gonadic categories on the RNAseq dataset already available . In accordance with the value of FoldChange > 10 or < 0.1 and padj < 0.001, the DESeq method identified 1,937 contigs differentially expressed between gonadic categories.
A hierarchical clustering method was applied on the RNAseq samples to cluster them according to the similarity of expression pattern among the 1,937 differentially expressed contigs (Fig. 1). Analysis of the dendrogram generated identified two clusters: one male, gathering the samples of male sex, and one female, gathering the samples of female sex. Moreover, the hierarchical clustering put one of the two gonad samples of the “undetermined” category in each cluster and did the same with the two gonad samples in sexual inversion. The one male gonad sample in regression was classified with female gonads. Thus, Fig. 1 illustrates how the samples “Und1”, “Inv1” and “MR2” showed a female gene expression pattern whereas the samples “Und2” and “Inv2” showed a male gene expression pattern.
Samples are divided into two clusters based on their contig expression pattern, discriminating male (blue) and female gonads (red). (M/F)E: male/female at early stage; (M/F)I: male/female at intermediate stage; (M/F)M: male/female at mature stage; (M/F)R: male/female at regression stage; purple or Inv: gonads in sexual inversion; grey or Und: gonad sex is undetermined.
From the clustering analysis results coupled with histological analyses, we modified the previous classification of P. margaritifera gonads proposed in Teaniniuraitemoana et al. (2014)  to establish a new one, no longer based on a histological approach alone but on a dual histo-molecular analytical approach considering the gene expression pattern of gonads. This new classification consists of 16 gonadic categories and can be divided into two main groups: gonads in the male and the female pathway characterized by a male and a female gene expression pattern, respectively. These new gonadic categories are described in Tables 1 and 2. Fig. 2 is used to describe the two types of gonad in sex inversion, i.e. InvF and InvM.
Two types of sexual inversion are distinguished: sexual inversion from male to female (InvF, Table 1) (A), and sexual inversion from female to male (InvM, Table 2) (B). Og: ovogonia; O: oocytes; Dg: degenerating oocytes; Spg: spermatogonia; T: connective tissue; Spc: spermatocytes; Spz: spermatozoids.
Selection of nine candidate genes of the sexual pathway in P. margaritifera
We performed a second differential expression analysis considering the gonadic categories from the new proposed histo-molecular classification. Thus, this second DE analysis was performed using the DESeq package between the M category and each of the categories of the female sexual pathway as follows: M vs F, M vs FR, M vs MRF, M vs IndF and M vs InvF. We then compared all the DE results in order to obtain the contigs commonly differentially expressed in all the comparisons described above. Nine genes were thus selected (Table 3). Three genes among these nine, pmarg-foxl2 coding for the Forkhead box protein L2 (Contig_43072) and two unknown genes pmarg-c43476 and pmarg-c45042 (Contig_43476 and Contig_45042), were expressed at least tenfold more in MRF, UndF, InvF, F and FR oyster gonads than in male gonads. They were thus considered as markers of the entire female pathway (FP). Two other unknown genes, pmarg-c19309 and pmarg-c54338 (Contig_19309 and Contig_54338), were expressed at least tenfold more in MRF and UndF oyster gonads (i.e., oysters presented female gene expression pattern at the molecular level but no female germ cells are visible at the cellular level) than all other categories, and therefore represent markers of the “early” female pathway (eFP). Two genes coding for the GATA-type zinc finger protein 1 (pmarg-zglp1, Contig_25360) and vitellogenin-6 (pmarg-vit6, Contig_15150), showed expression at least tenfold higher in female oyster gonads in gametogenesis than in other categories, and thus were classed as markers of the female category (F). Finally, the two last genes, pmarg-fem1-like and pmarg-dmrt were expressed at least tenfold more in categories of the male pathway (UndM, InvM, M and MR), and consequently were considered as markers of the male pathway (MP). RNAseq expressions of these nine genes, classed into the new gonadic categories, are presented in Fig. 3.
Heat map of the RNAseq expression in the new gonad categories of the nine selected genes. Color represents the normalized expression after variance-stabilizing transformation (DESeq). Expression levels are depicted with a color scale, in which shades of red represent higher expression and shades of green represent lower expression. FP: the overall female pathway, eFP: early female pathway, F: female in active gametogenesis, MP: the overall male pathway. M or F: male or female in gametogenesis; (M/F)R: male/female at regression stage; Und(M/F): Undifferentiated gonad at the cellular level beginning a male or a female determination/differentiation at the molecular level; Inv(M/F): oyster in sexual inversion female to male or male to female; MRF: male at regression stage at the cellular level beginning a female differentiation at the molecular level.
To validate at the biological level these nine candidates as markers of the corresponding pathway, gene mRNA levels were then assessed using real-time PCR in a new set of oyster gonads (Fig. 4A-I). These samples were classified into five gonadic categories only: M, MR, Und, F and FR, because it was not possible to determine the stages MRF, UndF or UndM, and FRM using histology. Concerning FP candidate genes, results of relative expression showed a significantly higher expression in female gonads overall, especially for pmarg-c43476 (F: 0.0099 ±0.0077; and FR: 0.0104 ±0.0107) (Fig. 4A, 4B and 4C). In the same way, relative expression of F and MP candidate genes showed levels equivalent to their expression found in the RNAseq dataset. Indeed, pmarg-vit6 (0.39 ±0.53) and pmarg-zglp1 (0.005 ±0.006) were significantly more expressed in F gonads (Fig. 4F and 4G), and pmarg-dmrt and pmarg-fem1-like were significantly more expressed in M gonads (0.005 ±0.004 and 0.04 ±0.02 respectively) (Fig. 4H and 4I). There was no significant difference between gonadic categories in the relative expression of early female pathway candidate genes (eFP), pmarg-c19309 and pmarg-c54338 (Fig. 4D and 4E).
Relative expression profiles (real-time PCR) of pmarg-foxl2 (A), pmarg-c43476 (B), pmarg-c45042 (C), pmarg-c19309 (D), pmarg-c54338 (E), pmarg-vit-6 (F), pmarg-zglp1 (G), pmarg-dmrt (H), and pmarg-fem1-like (I) in gonad categories (new set of samples). Different letters indicate statistically significant differences determined by ANOVA. M or F: male or female in gametogenesis; (M/F)R: male/female at stage of regression; Und: Undetermined sex.
A 3-gene-pair expression ratio signature of the sexual pathway in P. margaritifera
To cluster samples according to their similarity of expression patterns by the nine candidate genes, we performed a hierarchical clustering using spearman correlation, based not on relative expressions (9 variables) but on their ratio (36 variables), as in Ma et al. (2004) . Following this method, samples were grouped into two main clusters discriminating samples in the male and female pathways (Fig. 5). Moreover, each cluster was separated into two sub-clusters which divided oyster gonads on the early male or early female pathways (eMP or eFP), including undifferentiated and regressed gonads, from gonads in the course of active male or female gametogenesis (MP or FP).
Samples are divided into two main clusters based on their gene expression ratio pattern (36 variables), discriminating male and female pathways. Each cluster is divided into two sub-clusters discriminating oysters in early male or early female pathway (eMP or eFP) and oysters in the course of male or female pathway (MP or FP). M or F: male or female in gametogenesis; (M/F)R: male/female at regression stage; Und: gonad sex undetermined.
Next, to determine the major gene expression ratios explaining this clustering and thus having an effect on the sexual pathway, we performed a multivariate regression tree. A four-leaf MRT (where “leaf” refers to each terminal node), shown in Fig. 6, best described sexual pathways, explaining 100% of the total variance (Error = 0, Cross-Validation error = 0.167, Standard error = 0.0702). From the MRT model obtained, it appeared that, among the 36 input variables, the most important factors/predictors were three relative gene expression ratios implicating four candidate marker genes: pmarg-c43476, pmarg-foxl2, pmarg-c54338 and pmarg-fem1-like. The first regression tree split, which explains 43% of the variance, is based on the relative gene expression ratio of pmarg-c43476 to pmarg-fem1-like. If the expression ratio of these two genes is lower than 0.02 then, following the regression tree to the right, this means that the oyster sample is on the male pathway. When this ratio is higher than 0.02, following the MRT to the left, a second split then occurs based on the relative gene expression ratio of pmarg-foxl2 to pmarg-c54338, which explains 30% of the variance. Samples showing pmar-foxl2/pmarg-c54338 expression ratios greater or equal to 12.2 are on the female pathway, whereas samples showing a ratio lower than this critical value are on the early sexual pathway. Finally, the third and the last split of the regression tree (27% of the variance), is based on the pmarg-foxl2/pmarg-fem1-like expression ratio, and differentiates samples on the early female pathway (eFP) from those on the early male pathway (eMP) by a critical value of 0.1. Thus, following this MRT model of the sexual pathway of P. margaritifera, we can associate oysters with the particular pathways by analyzing the expression ratios of these four genes (i.e., pmarg-c43476/pmarg-fem1-like, pmarg-foxl2/pmarg-c54338 and pmarg-foxl2/pmarg-fem1-like).
Candidate predictor/explanatory variables are the relative expression ratio of the nine potential marker genes (36 variables), and the response variables are the four sub-clusters of gonad samples (n = 44) determined by hierarchical clustering analysis (eMP or eFP: oyster on the early male or female pathway; MP or FP: oyster in the course of male or female pathway). This tree explained 100% of the total variance. The bar graphs at the end of each leaf represent the number of samples of the corresponding group. Statistics of the MRT are presenting at the left of the figure. Error: Residual error; CV Error: Cross-validation error; and SE: Standard error.
To test the predictive ability of this MRT model of the sexual pathway of P. margaritifera on other datasets, we predicted the sexual pathways on which were engaged the RNAseq samples. To do this, we used the RNAseq normalized expression (estimatesizefactor, DESeq) of the samples. The regression tree prediction matched with the previous hierarchical clustering (Fig. 1). Samples of the male cluster were predicted to be on the male pathway and samples in the female cluster were predicted to be on the female pathway as shown in Table 2. The undetermined oyster gonads (“Und1” and “Und2), oyster gonads in sexual inversion (“Inv1” and “Inv2”) and the male regressed gonad sample “MR2”, were classified as in the earlier analysis (Table 4, Fig. 1).
As marine bivalves are organisms of major economic interest, genomic research has recently been conducted to provide greater insight into the molecular mechanisms governing various complex physiological traits [39–44]. Among those, several genomic studies have been employed successfully to better understand cellular and molecular mechanisms involved in reproductive behavior, including gametogenesis, sex determinism and sex differentiation [25–27,45,46]. Recently, in a previous study, a gonad transcriptome analysis of P. margaritifera was performed using Illumina-based RNAseq, which allowed to elucidate some molecular mechanisms involved in sex determination, sex differentiation and gametogenesis in this protandrous hermaphrodite species .
Here, we re-analyzed the quantification data of the de novo gonad transcriptome of P. margaritifera to identify marker genes of the sexual pathway and therefore acquire further insights into the complex and unusual reproduction of the black-lip pearl oyster.
A dual approach to modeling P. margaritifera reproduction
Although hierarchical clustering, based on the expression pattern of the 1,937 contigs determined as differentially expressed using a stringent statistical analysis, did not discriminate gametogenic stages, a clear grouping of the different sexes was observed. This marked and specific gonadic gene expression had already been shown for P. margaritifera  as well as for other marine bivalves with different reproductive functioning, including the Pacific oyster Crassostrea gigas (an irregular and alternative protandrous hermaphrodite) , Pacific lion-paw scallop Nodipecten subnodosus (a functional hermaphrodite) , and European clam Ruditapes decussatus (a gonochoric species) .
Interestingly, it appeared that individuals undergoing sexual inversion were distributed in both male and female clusters. Although these individuals undergoing sex inversion and presented gametes of both sexes, the main gene expression pattern resulted from the sex that an oyster would become. This molecular aspect is in accordance with the cellular aspects found by observation of histological sections of gonads in the course of sexual inversion. Indeed, the new germ line, consisting of several oocytes or spermatocytes, appears at the periphery of the acini, whereas the former sex phenotype is represented only by a few residual spermatozoids or degenerated oocytes . Moreover, in the hierarchical clustering, one of the two undifferentiated individuals was grouped on the same branch as the females, and the other on the same branch as the males. Misclassification has already been observed in two other marine bivalve species, C. gigas and R. decussatus, where individuals at undetermined sex stages were found to express male- or female-specific genes, suggesting that sex differentiation had already taken place, even if it was not visible at cellular level using histology [25,27]. Here, the most unexpected result was the finding of an individual histologically determined as a regressed male grouped with female oysters. This result, leads us to hypothesize that although this oyster could be determined as male due to the presence of residual spermatozoids, it had already initiated female sex differentiation according to its gene expression pattern. Following this result, regression stage appears to constitute the sex-switching time window in P. margaritifera, as already proposed in C. gigas. For this latter irregular alternative hermaphrodite species, the sex-determining time window may occur in the period around the end of a reproductive cycle and the beginning of the next .
All these results reveal the importance of considering both histological and molecular, approaches in the description and the understanding of the atypical reproduction of P. margaritifera. Indeed, these previous results allow us to propose a kinetic of the gonadic sex and stages found in the male and the female sexual pathway of this species, modeled in Fig. 7.
M or F: male or female in gametogenesis; (M/F)E: male/female at early stage; (M/F)I: male/female at intermediate stage; (M/F)M: male/female at mature stage; (M/F)R: male/female at regression stage; Und(M/F): gonad undifferentiated at the cellular level, beginning a male or a female differentiation at the molecular level; Inv(M/F): oyster in sexual inversion female to male or male to female; MRF: male in regression stage at the cellular level beginning a female differentiation at the molecular level; and FRM: female in regression stage at cellular level beginning a male differentiation at the molecular level.
This is the first histo-molecular reproduction model proposed for a marine bivalve species, taking into account molecular signatures coupled with cellular descriptions. Thus, at the end of a gametogenetic cycle, female or male gonads (M or F) would pass through a regression phase (MR or FR) before starting a new cycle. This gonad regression occur after a spawning event as other marine bivalve species [47–49] but might also occur when food are not abundant. Indeed, P. margaritifera gametogenesis is known to be modulated by food level and it was showed recently that the under-feeding decreases or stops the germinal activity preventing any resumption of gametogenesis . From this regression stage, gonads might also completely regressed and become undetermined (UndM or UndF) before starting a new gametogenetic cycle. Here, oysters are certainly in sexual rest. Gonads are undifferentiated at cellular level but shows early molecular signs of differentiation allowing us to know in which sexual pathway the oysters are. Moreover, as mentioned above, regression stage appears the sex-switching time window of P. margaritifera. Indeed, the transition from the male to the female pathway seems to occur from the gonad regression of male oysters and the inverse for the transition from the female to the male pathway. Indeed, male or female oysters pass through a regression stage (MRF/FRM) before reaching an inversion stage (InvF/InvM) to become females or males at gametogenesis. The factors involve in this sex change are not well known in P. margaritifera. In C. gigas, interaction between environmental, as temperature, and genetic factors could be involved in sex determination and thus in sex inversion [14,15,19]. In our bivalve species, Thielley (1993)  showed that changes from female to male could occur when environmental conditions (temperature or food) were stressful, but no study has yet shown environmental or genetic factors inducing changes from male to female. This proposed model of the reproduction of P. margaritifera will help to elucidating these mechanisms.
The nine identified marker genes reveal a 3-gene-pair expression ratio model of the sexual pathway in P. margaritifera
Mechanisms of sex determination, which are critical for development and reproduction, are among the least well-known processes in biology, since they vary a great deal among species, in both primary signal and downstream genetic pathways [50,51]. Sex determination is a crucial question that still needs to be elucidated in marine bivalves, especially pearl oyster, because of important implications for aquaculture. Providing a repertoire of molecular markers of the sexual pathway is essential for further studies on sex determination for this protandrous hermaphrodite bivalve.
Here, we selected nine genes as potential markers of the sexual pathway based on their differential RNAseq expression, considering the new gonadic categories of the histo-molecular classification. There were four unknown genes whose role cannot yet be interpreted, and five genes functionally assigned by orthology: pmarg-foxl2, pmarg-vit6, pmarg-zglp1, pmarg-dmrt and pmarg-fem1-like. These latter genes have already been shown to play a role in sex determination or sex differentiation in other organisms. pmarg-foxl2, coding for the Forkhead box protein L2, has previously characterized and discussed in the pearl oyster . In mammals, foxl2 is expressed in follicular cells and is the earliest known marker of ovarian differentiation [52,53]. The two genes pmarg-vit6 and pmarg-zglp1 coding for vitellogenin-6 and the GATA-type zinc finger protein 1, respectively, were identified here as putative markers of the female pathway. Indeed, vitellogenin, a female specific glycoprotein constituting the yolk protein, was identified as being necessary for building oocytes [25,54]. The GATA-type zinc finger protein 1, also named GATA-like protein-1 (GLP-1), was identified to be required for normal fertility in female mice, as its deficiency leads to the absence of oocytes at birth . Finally, the two genes pmarg-dmrt and pmarg-fem1-like, orthologs of family genes known to be involved in sex differentiation and sex determination respectively in model organisms and previously discussed in Teaniniuraitemoana et al. (2014) , are underlined here for the male pathway.
To go beyond the limits of histological analysis mentioned above, we considered the 44 new gonadic samples based on their relative marker gene expression ratios. Hierarchical clustering revealed that oyster gonad samples can be clustered into four groups differentiating, on a first level, oysters on the main male and female pathways, and on a second level, oysters in the early stages of the male or female pathways. Herein, we successfully identified three gonad-expressed gene ratios correlated with the oyster sexual pathway. Among the 36 gene ratios, three were enough to explain 100% of the variance and thus can be considered as major sexual determinants. This 3-gene-pair ratio sexual pathway signature is pmarg-c43476/pmarg-fem1-like, pmarg-foxl2/pmarg-c54338 and pmarg-foxl2/pmarg-fem1-like. It implicates two genes defined as potential markers of the female pathway, pmarg-c43476 and pmarg-foxl2; one unknown gene, pmarg-c54338, detected as a marker of the early female pathway; and pmarg-fem1-like, a marker of the male pathway. Interestingly, the level (0.1) of the gene expression ratio between pmarg-foxl2 and pmarg-fem1-like appears crucial in the sex-switching and sex differentiation of P. margaritifera distinguishing the early female and early male pathways. The P. margaritifera fem1-like gene was characterized as a derived member of the Fem1 family, which includes the fem-1 gene required for normal masculinization of somatic and germline tissue in the worm Caenorhabditis elegans . The high expression of pmarg-fem1-like might inhibit the expression of the pmarg-foxl2 and thus cause the male differentiation of the oyster. Such a pattern can be observed in C. elegans: fem genes inhibit tra-1, which would otherwise produce the protein TRA-1A whose sequence-specific DNA-binding would otherwise bring about female sex differentiation by regulating the transcription of target genes [57,58]. These previous results strongly suggest also that the P. margaritifera foxl2 gene might not only be involved in early ovary differentiation but also in ovary maintenance. Indeed, its large over-expression throughout the time course of the female pathway could suggest that pmarg-foxl2 might restrain the genetic program for testis differentiation by active repressing of pmarg-fem1-like gene. This maintenance mechanism involving foxl2 was already demonstrated in mammals by Uhlenhaut et al. (2009) . The authors showed that the mammalian ovarian phenotype has to be maintained throughout adulthood, mainly by a down regulation of the Sertoli cell-promoting gene Sox9 expression which Foxl2, coupled with estrogen receptors, is the critical factor responsible . Although our results are insufficient to conclude that pmarg-fem1-like and pmarg-foxl2 are major sex-determining genes in P. margaritifera, they would certainly play a key role in the molecular cascade of sex differentiation at a downstream level. Furthermore, unknown genes detected with relevant expression would be of great interest for further studies, as shown in other aquatic organisms .
The most significant outcomes of our study are the two models we constructed allowing us to improve our understanding of the specific reproduction of the marine bivalve P. margaritifera. The first is a model based on two approaches of the reproduction in this species defining the gonadic stage occurring in the sexual pathway by considering two levels of analysis: histological and molecular. In particular it reveals that the sex-inversion time window of P. margaritifera occurs during a regression phase of the gonad.
The second is a 3-gene-pair expression ratio based model, which makes it possible to predict the sexual pathway in this hermaphrodite species and can thus be used as a tool for future studies on sex inversion, sex differentiation and sex determination in P. margaritifera. Such work will aim to control reproduction for selection programs, thus supporting the sustainable development of pearl farming in French Polynesia. This latter model also revealed the importance of the expression of pmarg-foxl2 and pmarg-fem1-like for sex inversion and sex differentiation and the need to elucidate their role in P. margaritifera based on further functional research, such as the use of RNA interference now available in marine bivalves .
The authors are indebted to J.Y. Daniel, V. Quillien, V. Vonau, J. Fievet, C. Cabrau, L. Quintric, F. Marquer and all the staff of the GeT-PlaGe (http://www.genomique.genotoul.fr/) and Sigenae (http://www.sigenae.org) core facilities for their helpful assistance. We thank H. McCombie for English language editing.
Conceived and designed the experiments: VT AH GLM YG. Performed the experiments: VT GLM. Analyzed the data: VT. Contributed reagents/materials/analysis tools: PL. Wrote the paper: VT. Critically revised the manuscript: AH GLM YG NGM.
- 1. Le Pennec M, Anastas M, Bichet H, Buestel D, Cochard J, Cochennec-Laureau N, et al. Huître perlière et perle de Tahiti. Faaa, French Polynesia: HQ Imaging; 2010. p. 204.
- 2. Gueguen Y, Montagnani C, Joubert C, Marie B, Belliard C, Tayalé A, et al. Characterization of molecular processes involved in the pearl formation in Pinctada margaritifera for the sustainable development of pearl farming industry in French Polynesia. In: Watabe S, Maeyama K, Nagasawa H, editors. Recent Advances in Pearl Research. Tokyo: TERRAPUB; 2013. pp. 183–193.
- 3. Arnaud-Haond S, Goyard E, Vonau V, Herbaut C, Prou J, Saulnier D. Pearl formation: persistence of the graft during the entire process of biomineralization. Mar Biotechnol. 2007;9: 113–16. pmid:17139560
- 4. Ky C-L, Blay C, Sham-Koua M, Vanaa V, Lo C, Cabral P. Family effect on cultured pearl quality in black-lipped pearl oyster Pinctada margaritifera and insights for genetic improvement. Aquat Living Resour. 2013;26: 133–145.
- 5. Blay C, Sham-Koua M, Vonau V, Tetumu R, Cabral P, Ky C-L. Influence of nacre deposition rate on cultured pearl grade and colour in the black-lipped pearl oyster Pinctada margaritifera using farmed donor families. Aquac Int. 2013;22: 937–953.
- 6. Ky C-L, Blay C, Sham-Koua M, Lo C, Cabral P. Indirect improvement of pearl grade and shape in farmed Pinctada margaritifera by donor “oyster” selection for green pearls. Aquaculture. Elsevier B.V.; 2014;432: 154–162.
- 7. Tranter D. Reproduction in australian pearl oysters (Lamellibranchia). IV. Pinctada margaritifera (Linnaeus). Mar Freshw Res. CSIRO PUBLISHING; 1958;9: 509–525.
- 8. Chávez-Villalba J, Soyez C, Huvet A, Gueguen Y, Lo C, Moullac G Le. Determination of gender in the pearl oyster Pinctada margaritifera. J Shellfish Res. 2011;30: 231–240.
- 9. Thielley M. Etude cytologique de la gamétogenèse, de la sex-ratio et du cycle de reproduction chez l’huître perlière Pinctada margaritifera (L) var. cumingii (Jameson), (mollusque, bivalves). Comparaison avec le cycle de Pinctada maculata (Gould). Université française du pacifique. 1993. p. 233.
- 10. Vahirua-Lechat I, Laure F, LeCoz JR, Bianchini JP, Bellais M, Le Moullac G. Changes in fatty acid and sterol composition during oogenesis in the pearl oyster Pinctada margaritifera. Aquac Res. 2008;39: 1739–1746.
- 11. Le Moullac G, Soyez C, Sham-Koua M, Levy P, Moriceau J, Vonau V, et al. Feeding the pearl oyster Pinctada margaritifera during reproductive conditioning. Aquac Res. 2013;44: 404–411.
- 12. Valenzuela DM, Murphy AJ, Frendewey D, Gale NW, Economides AN, Auerbach W, et al. High-throughput engineering of the mouse genome coupled with high-resolution expression analysis. Nat Biotechnol. Nature Publishing Group; 2003;21: 652–9. pmid:12730667
- 13. Piferrer F, Guiguen Y. Fish gonadogenesis. Part II: Molecular biology and genomics of sex differentiation. Rev Fish Sci. 2008;16: 35–55.
- 14. Guo X, Hedgecock D, Hershberger WK, Cooper K, Allen K. Genetic determinants of protandric sex in the Pacific oyster, Crassostrea gigas Thunberg. Evolution (N Y). 1998;52: 394–402.
- 15. Hedrick PW, Hedgecock D. Sex determination: genetic models for oysters. J Hered. 2010;101: 602–611. pmid:20525766
- 16. Naimi A, Martinez A-S, Specq M-L, Mrac A, Diss B, Mathieu M, et al. Identification and expression of a factor of the DM family in the oyster Crassostrea gigas. Comp Biochem Physiol A Mol Integr Physiol. Elsevier Inc.; 2009;152: 189–196. pmid:18854223
- 17. Naimi A, Martinez A-S, Specq M-L, Diss B, Mathieu M, Sourdaine P. Molecular cloning and gene expression of Cg-Foxl2 during the development and the adult gametogenetic cycle in the oyster Crassostrea gigas. Comp Biochem Physiol B Biochem Mol Biol. Elsevier Inc.; 2009;154: 134–142. pmid:19481171
- 18. Santerre C, Sourdaine P, Martinez A-S. Expression of a natural antisense transcript of Cg-Foxl2 during the gonadic differentiation of the oyster Crassostrea gigas: first demonstration in the gonads of a lophotrochozoa species. Sex Dev. 2012;6: 210–21. pmid:22572405
- 19. Santerre C, Sourdaine P, Marc N, Mingant C, Robert R, Martinez A-S. Oyster sex determination is influenced by temperature—first clues in spat during first gonadic differentiation and gametogenesis. Comp Biochem Physiol A Mol Integr Physiol. Elsevier Inc.; 2013;165: 61–69. pmid:23416889
- 20. Santerre C, Sourdaine P, Adeline B, Martinez A-S. Cg-SoxE and Cg-β-catenin, two new potential actors of the sex-determining pathway in a hermaphrodite lophotrochozoan, the Pacific oyster Crassostrea gigas. Comp Biochem Physiol A Mol Integr Physiol. Elsevier Inc.; 2014;167: 68–76. pmid:24120488
- 21. Chávez-Villalba J, Cochard J-C, Le Pennec M, Barret J, Enríquez-Díaz M, Cáceres-Martínez C. Effect of temperature and feeding regimes on gametogenesis and larval production in the oyster Crassotrea gigas. J Shellfish Res. 2003;22: 721–731.
- 22. Dutertre M, Beninger PG, Barillé L, Papin M, Rosa P, Barillé A-L, et al. Temperature and seston quantity and quality effects on field reproduction of farmed oysters, Crassostrea gigas, in Bourgneuf Bay, France. Aquat Living Resour. 2009;22: 319–329.
- 23. Saout C, Quéré C, Donval A, Paulet Y-M, Samain J-F. An experimental study of the combined effects of temperature and photoperiod on reproductive physiology of Pecten maximus from the Bay of Brest (France). Aquaculture. 1999;172: 301–314. pmid:9930771
- 24. Enríquez-Díaz M, Pouvreau S, Chávez-Villalba J, Le Pennec M. Gametogenesis, reproductive investment, and spawning behavior of the Pacific giant oyster Crassostrea gigas: evidence of an environment-dependent strategy. Aquac Int. 2009;17: 491–506.
- 25. Dheilly NM, Lelong C, Huvet A, Kellner K, Dubos M- P, Riviere G, et al. Gametogenesis in the Pacific oyster Crassostrea gigas: a microarrays-based analysis identifies sex and stage specific genes. PLoS One. 2012;7: e36353. pmid:22590533
- 26. Llera-Herrera R, García-Gasca A, Abreu-Goodger C, Huvet A, Ibarra AM. Identification of male gametogenesis expressed genes from the scallop Nodipecten subnodosus by suppressive subtraction hybridization and pyrosequencing. PLoS One. 2013;8: e73176. pmid:24066034
- 27. Sousa JT, Milan M, Bargelloni L, Pauletto M, Matias D, Joaquim S, et al. A microarray-based analysis of gametogenesis in two Portuguese populations of the European clam Ruditapes decussatus. PLoS One. 2014;9: e92202. pmid:24643002
- 28. Matsumoto T, Masaoka T, Fujiwara A, Nakamura Y, Satoh N, Awaji M. Reproduction-related genes in the pearl oyster genome. Zoolog Sci. 2013;30: 826–850. pmid:24125647
- 29. Teaniniuraitemoana V, Huvet A, Levy P, Klopp C, Lhuillier E, Gaertner-Mazouni N, et al. Gonad transcriptome analysis of pearl oyster Pinctada margaritifera: identification of potential sex differentiation and sex determining genes. BMC Genomics. 2014;15: 491. pmid:24942841
- 30. Fournier J, Levesque E, Pouvreau S, Le Pennec M, Le Moullac G. Influence of plankton concentration on gametogenesis and spawning of the black lip pearl oyster Pinctada margaritifera in Ahe atoll lagoon (Tuamotu Archipelago, French polynesia). Mar Pollut Bull. Elsevier Ltd; 2012;65: 463–470. pmid:22560741
- 31. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25: 402–8. pmid:11846609
- 32. Andersen CL, Jensen JL, Ørntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Reseach. 2004;64: 5245–5250. pmid:15289330
- 33. R Development Core Team R. R: A Language and Environment for Statistical Computing [Internet]. R Foundation for Statistical Computing. 2011. p. 409.
- 34. Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. BioMed Central Ltd; 2010;11: R106. pmid:20979621
- 35. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B. 1995;57: 289–300.
- 36. De’ath G. Multivariate regression trees: a new technique for modeling species-environnement relationships. Ecology. 2002;83: 1105–1117.
- 37. Therneau TM, Atkinson B, Ripley B, Oksanen J, De’ath G. mvpart: Multivariate partitioning. R package version 1.6–1. 2013
- 38. Ma X-J, Wang Z, Ryan PD, Isakoff SJ, Barmettler A, Fuller A, et al. A two-gene expression ratio predicts clinical outcome in breast cancer patients treated with tamoxifen. Cancer Cell. 2004;5: 607–616. pmid:15193263
- 39. Meng X, Liu M, Jiang K, Wang B, Tian X, Sun S, et al. De novo characterization of Japanese scallop Mizuhopecten yessoensis transcriptome and analysis of its gene expression following cadmium exposure. PLoS One. 2013;8: e64485. pmid:23741332
- 40. Li Q, Zhao X, Kong L, Yu H. Transcriptomic response to stress in marine bivalves. Invertebr Surviv J. 2013;10: 84–93.
- 41. Gavery MR, Roberts SB. Predominant intragenic methylation is associated with gene expression characteristics in a bivalve mollusc. PeerJ. 2013;1: e215. pmid:24282674
- 42. Shi Y, He M. Differential gene expression identified by RNA-Seq and qPCR in two sizes of pearl oyster (Pinctada fucata). Gene. Elsevier B.V.; 2014;538: 313–322. pmid:24440293
- 43. Rosa RD, de Lorgeril J, Tailliez P, Bruno R, Piquemal D, Bachère E. A hemocyte gene expression signature correlated with predictive capacity of oysters to survive Vibrio infections. BMC Genomics.???; 2012;13: 252. pmid:22708697
- 44. Joubert C, Piquemal D, Marie B, Manchon L, Pierrat F, Zanella-Cléon I, et al. Transcriptome and proteome analysis of Pinctada margaritifera calcifying mantle and shell: focus on biomineralization. BMC Genomics. BioMed Central Ltd; 2010;11: 613. pmid:21040589
- 45. Banni M, Negri A, Mignone F, Boussetta H, Viarengo A, Dondero F. Gene expression rhythms in the mussel Mytilus galloprovincialis (Lam.) across an annual cycle. PLoS One. 2011;6: e18904. pmid:21573210
- 46. Ghiselli F, Milani L, Chang PL, Hedgecock D, Davis JP, Nuzhdin S V, et al. De Novo assembly of the Manila clam Ruditapes philippinarum transcriptome provides new insights into expression bias, mitochondrial doubly uniparental inheritance and sex determination. Mol Biol Evol. 2012;29: 771–786. pmid:21976711
- 47. Shafee MS, Daoudi M. Gametogenesis and spawning in the carpet-shell clam, Ruditapes decussatus (L.) (Mollusca: Bivalvia), from the Atlantic coast of Morocco. Aquac Fish Manag. 1991;22: 203–216.
- 48. Steele S, Mulcahy MF. Gametogenesis of the oyster Crassostrea gigas in southern Ireland. J Mar Biol Assoc UK. 1999;79: 673–686.
- 49. Aragón-Noriega EA, Chávez-Villalba J, Gribben PE, Alcántara-Razo E, Maeda-Martínez AN, Arambula-Pujol EM, et al. Morphometric Relationships, Gametogenic Development and Spawning of the Geoduck Clam Panopea Globosa (Bivalvia: Hiatellidae) in the Central Gulf of California. J Shellfish Res. 2007;26: 423–431.
- 50. Haag ES, Doty A V. Sex determination across evolution: connecting the dots. PLoS Biol. 2005;3: e21. pmid:15660158
- 51. Yusa Y. Causes of variation in sex ratio and modes of sex determination in the Mollusca—an overview. Am Malacol Bull. 2007;23: 89–98.
- 52. Cocquet J, Pailhoux E, Jaubert F, Servel N, Xia X, Pannetier M, et al. Evolution and expression of FOXL2. J Med Genet. 2002;39: 916–921. pmid:12471206
- 53. Schmidt D, Ovitt CE, Anlag K, Fehsenfeld S, Gredsted L, Treier A-C, et al. The murine winged-helix transcription factor Foxl2 is required for granulosa cell differentiation and ovary maintenance. Development. 2004;131: 933–942. pmid:14736745
- 54. Matsumoto T, Nakamura AM, Mori K, Kayano T. Molecular characterization of a cDNA encoding putative Vitellogenin from the Pacific oyster Crassostrea gigas. Zoolog Sci. 2003;20: 37–42. pmid:12560599
- 55. Strauss TJ, Castrillon DH, Hammes SR. GATA-like protein-1 (GLP-1) is required for normal germ cell development during embryonic oogenesis. Reproduction. 2011;141: 173–181. pmid:21123517
- 56. Hodgkin J. Sex determination in the nematode C. elegans: analysis of TRA-3 suppressors and characterization of FEM genes. Genetics. 1986;114: 15–52. pmid:3770465
- 57. Zarkower D, Hodgkin J. Molecular analysis of the C. elegans sex-determining gene tra-1: a gene encoding two zinc finger proteins. Cell. 1992;70: 237–249. pmid:1339311
- 58. Zarkower D, Hodgkin J. Zinc fingers in sex determination: only one of the two C. elegans Tra-1 proteins binds DNA in vitro. Nucleic Acids Res. 1993;21: 3691–3698. pmid:8367286
- 59. Uhlenhaut NH, Jakob S, Anlag K, Eisenberger T, Sekido R, Kress J, et al. Somatic sex reprogramming of adult ovaries to testes by FOXL2 ablation. Cell. 2009;139: 1130–1142. pmid:20005806
- 60. Yano A, Guyomard R, Nicol B, Jouanno E, Quillet E, Klopp C, et al. An immune-related gene evolved into the master sex-determining gene in rainbow trout, Oncorhynchus mykiss. Curr Biol. 2012;22: 1423–1428. pmid:22727696
- 61. Fabioux C, Corporeau C, Quillien V, Favrel P, Huvet A. In vivo RNA interference in oyster—vasa silencing inhibits germ cell development. FEBS J. 2009;276: 2566–2573. pmid:19476495