In silico identification and molecular characterization of genes predominantly expressed in the fish oocyte
© Bobe et al. 2008
Received: 16 July 2008
Accepted: 23 October 2008
Published: 23 October 2008
Skip to main content
© Bobe et al. 2008
Received: 16 July 2008
Accepted: 23 October 2008
Published: 23 October 2008
In fish, molecular mechanisms that control follicle-enclosed oocyte progression throughout oogenesis and oocyte developmental competence acquisition remain poorly understood. Existing data in mammals have indicated that the so called "oocyte-specific" genes play an important role in oogenesis, fertilization, and early embryo development. In teleost species, very little is known about "oocyte-specific" genes. The present study therefore aimed at identifying and characterizing oocyte-specific genes in fish.
Using digital differential display PCR, mouse ESTs exhibiting an oocyte-predominant expression were identified. Those murine ESTs were subsequently used to identify cognate rainbow trout (Oncorhynchus mykiss) ESTs using a reciprocal Blast search strategy. In the present study we report the identification of five previously uncharacterized rainbow trout cDNAs exhibiting a oocyte-specific, oocyte-predominant, or gonad-specific expression: zygote arrest 1 (zar1), v-mos Moloney murine sarcoma viral oncogene-like protein (mos), B-cell translocation gene (btg3), growth differentiation factor 9 (gdf9), and mutS homolog 4 (msh4). The orthology relationship of each of these genes with vertebrate counterparts was verified by phylogenetic analysis. Among those five genes, three had never been characterized in any fish species. In addition, we report the oocyte-predominant expression of btg3 for the first time in any vertebrate species. Finally, those five genes are present in unfertilized eggs as maternally-inherited mRNAs thus suggesting that they could participate in ovarian folliculogenesis as well as early embryonic development.
The expression patterns of zar1, mos, btg3, gdf9 and msh4 in rainbow trout and the functions of their orthologs in higher vertebrates strongly suggest that they might play an important role in follicle-enclosed oocyte development, meiosis control and early embryonic development in fish. Future investigations are however required to unravel the participation of those strong candidates in the molecular processes that control folliculogenesis and/or oocyte developmental competence in fish.
Oocyte developmental competence can be defined as the oocyte ability to be fertilized and to subsequently develop into a normal embryo. In fish, molecular mechanisms that control oocyte developmental competence remain poorly understood. In the past few years, transcriptomic investigations have been initiated to tentatively link oocyte transcriptome and oocyte developmental potential in order to identify key genes involved in the control of oocyte developmental competence . While these types of approaches have been successful, information on the specific molecular mechanisms that make a good oocyte are still limited. One alternative way to fully understand the molecular mechanisms controlling oocyte quality is to study genes that are specifically or predominantly expressed in the oocyte. In mammals it has been shown that the so called "oocyte-specific" genes can affect folliculogenesis, fertilization and early development [2–4]. These genes have been extensively studied in mammals. Yet, very little information is available about those genes in fish despite the recent identification of ovarian-predominant genes in zebrafish . The purpose of the present study was therefore to identify and characterize genes exhibiting a predominant oocyte expression in fish. Taking advantage of the numerous murine tissue-specific libraries available in public databases, we used an in silico approach to identify genes exhibiting an oocyte-predominant expression in rainbow trout (Oncorhynchus mykiss). Our study led to the identification and characterization of five previously uncharacterized rainbow trout cDNAs exhibiting an oocyte-specific, oocyte-predominant, or gonad-specific expression: zygote arrest 1 (zar1), v-mos Moloney murine sarcoma viral oncogene-like protein (mos), B-cell translocation gene (btg3), growth differentiation factor 9 (gdf9), and mutS homolog 4 (msh4).
Zygote arrest 1 (Zar 1) is a maternal-effect gene critical for the oocyte-to-embryo transition first identified in the mouse . In this species, zar1-/- mice are infertile as most of their embryos stop developing at one-cell stage. Since its discovery, Zar1 was characterized and its expression studied in several vertebrates species including mammals [9–12], chicken  and Xenopus . By contrast, available data in fish are scarce. Zebrafish (Danio rerio) and pufferfish (Fugu rubipres) zar1 sequences have been reported but no information is available on tissue or cellular expression of zar1 in any fish species. Similarly to what has been reported in all studied vertebrate species , the rainbow trout zar1 sequence exhibits an atypical PHD motif (C-X2-C-X13-C-X2-C-X4-C-X1-C-X17-C-X2-C) (Figure 1). The phylogenic analysis confirmed that the rainbow trout zar1 sequence was orthologous to the previously characterized vertebrate ZAR1 sequences including mouse Zar1 (Figure 2). We also show, for the first time in any fish species, that rainbow trout zar1 is strongly expressed in the ovary whereas a limited expression is observed in metaphase II oocytes (unfertilized eggs). Within the ovary, the expression was limited to the ooplasm as demonstrated by in situ hybridization. A very low signal was also observed in testis whereas no detectable expression was seen in any other tissue. In agreement with the results reported here, Zar1 was shown to be expressed exclusively in the oocyte in chicken and mouse [8, 12, 13]. In contrast, expression in other tissues such as testis [8, 11], muscle , lung , and brain  was also observed in various vertebrates species. In bovine, pig and human, the mRNA expression observed in the testis results from an alternative splicing of the ZAR1 gene . Together, rainbow trout zar1 sequence and tissue expression are consistent with a role in oocyte/embryo development in fish that would be similar to what has been shown in the mouse. Further studies are needed to thoroughly explore any relationship between zar1 expression in the oocyte and the acquisition of oocyte developmental competence.
In the mouse oocyte, Mos encodes for a serine-threonine kinase involved in the maintenance of the meiotic arrest at metaphase II [14–16]. A disruption of Mos results in spontaneous parthenogenetic activation of oocytes [14, 16]. In Xenopus, mos has also long been implicated in the maintenance of the meiotic arrest . In contrast, data on mos function and expression are extremely limited in fish. In the goldfish (Carassius auratus), mos is also involved in the metaphase II arrest but does not participates in oocyte maturation . In the present study, we observed that rainbow trout mos mRNA is specifically expressed in the oocyte and not detected in any other tissue. Interestingly, mos mRNA is present in the unfertilized egg and is therefore maternally inherited. To the best of our knowledge, no information is available on the tissue distribution of the mos mRNA in fish. In several mammalian species, Mos mRNA was only found in embryos, ovary and testis [19, 20]. In addition, mos mRNA was found to be expressed in the shark testis . While it is unknown if rainbow trout mos is expressed in the testis at other stages, its expression in the oocyte is consistent with existing data in higher vertebrates. Interestingly, the expression of mos mRNA in the unfertilized egg suggests that mos could participate in early development in addition to its well documented role in meiotic arrest.
BTG3 also named ANA and TOB5 belongs to a family of proteins, the BTG family; know for their anti-poliferative activity. In this family, 6 different proteins have been characterized in vertebrates . The phylogenetic analysis carried out in the present study clearly shows that we have identified the rainbow trout btg3 cDNA. The Btg3 gene was originally cloned in the mouse  and reported to be expressed in several cell lines and in a wide variety of murine and human adult tissues [23, 24]. Similarly, porcine Btg3 mRNA was detected in most tissues assayed . In the present study, btg3 mRNA could be detected in many tissues at very low levels. However, a strong and predominant expression was monitored in the oocyte. Together, our observations are consistent with existing data in mammals. However the oocyte-predominant expression of BTG3 was never reported in any vertebrate species and a thorough expression analysis will be necessary in other vertebrate species. Interestingly, several studies have shown that BTG4, another BTG family member, was preferentially expressed in the chicken oocyte  and in bovine reproductive tissues . In fish, a recent transcriptomic study also revealed that btg4 was predominantly expressed in zebrafish ovarian tissue .
In the mouse, the 30 kDa protein encoded by the Btg3 gene was cell cycle-dependent and peaked at the end of the G1 phase . Overexpression of the human cognate protein resulted in an impaired serum-induced cell cycle progression from the G0/G1 to S phase in NIH3T3 cells . More recently, in an attempt to study DNA damaged-induced genes, BTG3 was identified as a p53 target exhibiting an antiproliferative activity. Together, the predominant oocyte-expression of rainbow trout btg3 and the antiproliferative activity of the cognate protein in mammals suggest that btg3 could play an important role in oocyte development in fish. In addition, the presence of btg3 mRNA in the trout female gamete suggests a role for btg3 during early embryonic development, possibly in response to UV-induced DNA damage.
Gdf9 is an oocyte-specific gene of the TGF beta superfamily involved in folliculogenesis. It participates in the successful transition from primary to secondary follicles and it was previously shown that Gdf9-null mice are sterile [27, 28]. In fish, gdf9 was very recently characterized in zebrafish  and sea bass (Dicentrarchus labrax) . In the present study, the phylogenic analysis clearly showed that rainbow trout gdf9 was orthologous to those previously characterized gdf9 proteins in teleosts [6, 7] despite the difficulty to construct a reliable phylogenetic tree among vertebrate species. Northern blot analysis showed an ovarian-specific expression of gdf9 in sea bass . Similarly, semi quantitative PCR showed a gdf9 expression in zebrafish oocyte and testis, and possibly a weak signal in follicular cells . In the present study, we clearly showed using real-time PCR and in situ hybridization that rainbow trout gdf9 is exclusively expressed in the oocyte. Interestingly, significant levels of gdf9 mRNA were detected in the unfertilized egg, thus demonstrating that gdf9 is maternally inherited in rainbow trout. This observation is supported by semi-quantitative PCR data in zebrafish showing strong mRNA levels at early blastula stage and sharp decrease during gastrulation . While data on gdf9 in fish are scarce, the observed expression patterns are consistent with existing data in mammals. However, the functions of gdf9 in fish, including a possible role during early development, remain currently unknown.
mutS homolog 4 (MSH4) is a meiosis-specific gene belonging to the DNA mismatch repair (MMR) system. In yeast (Saccharomyces cerevisiae) MSH4 is required for reciprocal recombination and proper segregation of homologous chromosomes during meiosis I . In humans, MSH4 protein is only found in testis and ovary . In mice, Msh4 plays an essential role in the control of meiotic recombination and a disruption of this gene leads to male and female sterility due to meiotic failure . In fish, very little is known about msh4. To date, msh4 cDNA and protein sequences were never characterized from any fish species and only sequences automatically predicted from zebrafish and tetraodon genomes are available. In rainbow trout, in agreement with existing data in mammals, the tissue distribution study shows a gonad-specific expression pattern and a strong testicular expression. In addition, high expression levels were found in the late vitellogenic ovary, immediately prior to meiosis resumption. Based on existing data in yeast and mammals, it can be speculated that msh4 plays an important role in meiosis in fish. However, the msh4 mRNA is also detected in metaphase II oocytes at low levels. This indicates that msh4 mRNA is maternally inherited and could thus participate in early development, possibly through DNA mismatch repair functions. Further investigations are needed to study the expression of msh4 in fish and characterize its participation in oocyte and embryo development.
Using an in silico analysis, we have successfully identified 5 previously uncharacterized rainbow trout cDNAs exhibiting an oocyte-specific, gonad-specific, or oocyte-predominant expression. Among those 5 genes, 3 had never been characterized in any fish species. In addition, we report the oocyte-predominant expression of btg3 for the first time in any vertebrate species. Finally, expression patterns of those 5 genes in fish and the functions of their orthologs in higher vertebrates strongly suggest that they might play an important role in fish oocyte development, meiotic arrest and early embryonic development.
A differential digital display (DDD) analysis was previously performed with mouse ESTs providing a list of murine oocyte-specific genes . Cognate rainbow trout expressed sequence tags (ESTs) were subsequently identified using a reciprocal blast search strategy. A tblastX search was performed against all rainbow trout expressed sequence tags (ESTs) available in dbEST  using oocyte-specific mouse sequences identified in silico. The corresponding clones were obtained from INRA-Agenae program resource center (Jouy-en-Josas, France)  and fully sequenced in both directions using the dye-termination method (ABI PRISM 310, PE Biosystems). The deduced amino acid sequence was used for sequence alignment and phylogenetic analysis. Alternatively, the amino acid sequence was deduced from rainbow trout ESTs belonging to the same UniGene cluster.
Phylogenetic analysis was performed using the phylogenomic analysis pipeline available in the FIGENIX platform http://www.up.univ-mrs.fr/evol/figenix/. FIGENIX retrieved sequences, provided multiple sequence alignments, performed phylogenetic reconstruction, and deduced orthology and paralogy relationships (for a detailed description of pipelines and models used, see ). For each studied gene, the protein sequence was entered in the phylogenomic inference task, which was run with the default parameters and with Ensembl database (release 49) . We chose the NJ (neighbor joining) topology for the graphical representation. The trees (npl) are the fusion of three phylogenetic trees built based on neighbor joining , maximum parsimony, and maximum likelihood . The Dayhoff PAM matrix provided the distance matrix for the NJ method. The evolutionary distance separating sequences is defined as the number of mutational events per site underlying the evolutionary history separating sequences. Thus, evolutionary relations among sequences are represented by the tree structure, where branch length represents the evolutionary distance [38, 39]. Thus, evolutionary relations among sequences are represented by the tree structure, where branch length represents the evolutionary distance [38, 39]. In phylogenetic tree, bootstrap values are reported on each node for each npl method. Bootstrapping was carried out with 1000 replications.
Investigations were conducted according to the guiding principles for the use and care of laboratory animals and in compliance with French and European regulations on animal welfare. Rainbow trout (Oncorhynchus mykiss) in their first reproductive season were obtained from an experimental fish farm (PEIMA, Sizun, France). Fish were deeply anaesthetized in 2-phenoxyethanol (10 mg/ml of water), killed by a blow on the head and bled by gill arch section. Tissues were sampled from 3 ovulated females. Testis samples were obtained from 3 different males at stage II of spermatogenesis . For RNA extraction, tissues were homogenized in Tri-reagent (Molecular Research Center, Cincinnati, OH) at a ratio of 100 mg per ml of reagent and total RNA was extracted according to manufacturer's instruction. For in situ hybridization, ovarian tissue was sampled from an ovulated female, fixed in Dietrick's fixative (10% Formaldehyde, 28. 5% ethanol, 2% glacial acetic acid) at 4°C overnight, rinsed in tap water for 1 hour and held in 50% ethanol until further processing.
Real-time PCR primers
Dehydration (increasing ethanol: 15 min in 50% ethanol, twice 15 min in 70% ethanol, 15 min in 80% ethanol, 30 min in 96% ethanol, and 30 min in 96% ethanol/butanol vol/vol), clearing (butanol once for 30 min, and twice for 3 h each), and paraffin infiltration (once for 1 h and twice for 2 h, at 60°C) were performed in a Citadel 1000 tissue processor (Shandon, Pittsburgh, PA). Dehydrated tissues were embedded in plastic molds in paraffin using a HistoEmbedder (TBS88, Medite, Germany).
Digoxigenin-labeled anti-sense RNA probes were produced using the Promega T3/T7 RNA polymerase Riboprobe Combination System as recommended by the manufacturer, using as DNA template a PCR product obtained following amplification of the plasmid inserts with M13 reverse and M13 forward primers. Digoxigenin-labeled riboprobes were then purified by precipitation in ammonium acetate 7.5 M/ethanol for 2 hours at -20°C, and RNA concentrations were measured using a NanoDrop® spectrophotometer. Serial cross-sections of 5 μm were deparaffinized, re-hydrated in TBS (50 mM Tris, pH 7.4, 150 mM NaCl) and post-fixed in 4% PFA for 20 min. ISH was performed using the "In situ Pro, Intavis AG robotic station". Incubation volumes for all ISH steps were set to 250 μl. Digestion was carried out for 20 min with 3 μg/ml of proteinase K. Pre-hybridization (2 h, 60°C) and hybridization (12 h, 60°C) were carried out in 50% formamide, 2 × SSC, 1 × Denhardt, 10% dextran sulfate, and 250 μg/ml tRNA. For hybridization the digoxigenin-labeled anti-sense RNA probes were diluted in hybridization buffer at a final concentration of 3 ng/μl. Washing steps (2 × SSC, 60 min) were performed at 60°C followed by an RNAse treatment at 37°C. The digoxigenin signal was then revealed with an anti-digoxigenin antibody conjugated with alkaline phosphatase (Roche Diagnostics Corp.) and a NBT/BCIP revelation system (Roche Diagnostics Corp.) as recommended by the manufacturer. Slides were mounted with mowiol 4–88 (Calbiochem).
Supported by an INRA Animal Physiology and Livestock Systems department grant to PM and JB.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.