Evidence for a conserved role of retinoic acid in urodele amphibian meiosis onset

Authors

  • Angelina Wallacides,

    1. EA 3442 Aspects Cellulaires et Moléculaires de la Reproduction et du Développement, Nancy-Université, Nancy, France
    Search for more papers by this author
  • Amand Chesnel,

    1. EA 3442 Aspects Cellulaires et Moléculaires de la Reproduction et du Développement, Nancy-Université, Nancy, France
    Search for more papers by this author
  • Dominique Chardard,

    1. EA 3442 Aspects Cellulaires et Moléculaires de la Reproduction et du Développement, Nancy-Université, Nancy, France
    Search for more papers by this author
  • Stéphane Flament,

    1. EA 3442 Aspects Cellulaires et Moléculaires de la Reproduction et du Développement, Nancy-Université, Nancy, France
    Search for more papers by this author
  • Hélène Dumond

    Corresponding author
    1. EA 3442 Aspects Cellulaires et Moléculaires de la Reproduction et du Développement, Nancy-Université, Nancy, France
    • Nancy-Université, Faculté des Sciences, EA3442 Aspects Cellulaires et Moléculaires de la Reproduction et du Développement, Boulevard des Aiguillettes, BP 239, 54506 Vandoeuvre-lès-Nancy cedex, France
    Search for more papers by this author

Abstract

Pleurodeles waltl is a urodele amphibian displaying a ZZ/ZW genetic mode of sex determination. Gonad differentiation can later be modulated by hormone treatment. To investigate germ cell differentiation, we analyzed the expression of the meiosis marker PwDmc1 and show that germ cells enter meiosis in late larval life in females, and 2 months after metamorphosis in males. Organotypic cultures of gonad–mesonephros complexes demonstrated that retinoic acid triggers meiosis entry in P. waltl. In vivo analyses of both PwRaldh2 and PwCyp26b1 expressions, the enzymes required for RA synthesis and degradation respectively, indicate that meiosis onset depends on PwCyp26b1 repression in the gonad during normal or steroid-induced sex-reversed development. Taken together, our results show that RA-dependent meiosis entry could be a conserved mechanism of germ cell differentiation in vertebrates and provide evidence for crosstalk between steroid and RA signaling in the course of sex differentiation. Developmental Dynamics 238:1389–1398, 2009. © 2009 Wiley-Liss, Inc.

INTRODUCTION

The newt Pleurodeles waltl is a urodele amphibian that obeys female ZW heterogamety. Z and W sex chromosomes cannot be distinguished by conventional karyotypes but the genotype can be deduced at early larval stage from the analysis of a sex-linked marker, peptidase 1 (Ferrier et al.,1980). Of interest, ZW but not ZZ sex determination can be reversed by rearing the animals at 32°C instead of 20°C (reviewed in Dumond et al.,2008a).

When compared with other vertebrates, gonad differentiation is a slow process in amphibians and especially in Pleurodeles: it takes place mainly during larval life but goes on after metamorphosis and even in adulthood with differentiation of multiple testes (Dournon et al.,1990; Dumond et al.,2008a). Although the somatic compartment of the gonad and the role of steroid hormones have been studied extensively in this species (Kuntz et al.,2003a), little is known about germline differentiation due to the lack of molecular markers up to now. Using embryonic surgery experiments, Maufroid and Capuron (1985) showed that germ cells are induced by endodermal cells from the presumptive mesoderm. However, germ cells resulting from this induction cannot be identified until much later, just before hatching (stage 35) by using morphological features. From stage 35 to 41, primordial germ cells (PGCs) enter a resting period in both sexes while migrating to colonize the somatopleure, leading to the formation of genital ridges (Dournon et al.,1989,1990). Then, proliferation of both germ cells (stage 42 in ZZ larvae and stage 44 in ZW larvae) and somatic cells starts in the undifferentiated gonad. Beyond stage 53, ovarian differentiation in ZW larvae is characterized by the development of the cortical part, which includes large follicles while medulla regression leaves a cavity. In contrast, testis differentiation results from a strong development of the medulla. In Pleurodeles, stage 53 is the first stage at which gonadal sex may be ascertained by histological examination. Vitellogenic oocytes can be detected at metamorphosis and spermatogenesis begins during juvenile stage (see Fig. 1). Nevertheless, the precise time course of germ cell meiosis entry in both sexes and the factors triggering this process are still completely unknown.

Figure 1.

Timetable of Pleurodeles waltl development. Until stage 54, larvae are staged according to Gallien and Durocher (1957). Timing of development is expressed as “days postfertilization” (dpf). Sex differentiation process is first observed at stage 51 through induction of aromatase activity in females and is detected histologically at stage 53. Metamorphosis occurs at stage 56, approximately 90 days after the beginning of stage 54. During the thermosensitive period (from stage 42 to stage 54), females can be reversed to fertile males if reared at 32°C. During the hormone-sensitive period (stage 52 to stage 56), males treated with estradiol are reversed into females, while females treated with DHT are reversed into males.

As observed in other nonmammalian vertebrates, steroid hormones play a key role in Pleurodeles waltl sex differentiation wherein the P450-aromatase enzyme (P450arom) responsible for estrogen synthesis, is a major player (Kuntz et al.,2003b). Measurements of P450arom activity in gonad–mesonephros (GM) complexes reveal a higher activity in ZW larvae from stage 52 than in their ZZ counterparts (Chardard et al.,1995) and treatment of ZW larvae with an aromatase inhibitor induces sex reversal (Chardard and Dournon,1999). These results indicate that elevation of P450arom expression is the actual first sign of ovarian determination. The role of steroids in sex differentiation is also evidenced by the fact that, when added to the rearing water of ZZ larvae from stage 52 until metamorphosis, estradiol induces a male-to-female sex reversal whereas nonaromatizable androgens drive ZW larvae to differentiate as males (Chardard et al.,2003). Although these experimental data clearly demonstrate the effect of exogenous steroids on gonad differentiation, we have no knowledge about their potential effect on germline.

In mammals, meiosis entry followed by prophase I arrest is initiated during fetal development in females, whereas in males, spermatogenesis begins only after birth. Until recently, it was commonly believed that female germ cells have the intrinsic property to enter meiosis, whereas this process was thought to be delayed in males by an unknown factor. However, two studies from Koubova et al. (2006) and Bowles et al. (2006) demonstrated the involvement of retinoic acid (RA) in determining germ cell fate in mouse. They showed that RA is released from the mesonephros of both sexes following Aldh1A2 (also called Raldh2) activity from 10.5 days post-conception (dpc), while Cyp26b1, a P450-cytochrome enzyme that degrades RA, is expressed in the gonad. At 12.5 dpc, by the time of meiosis entry in females, Cyp26b1 is no more expressed in females, leading to high levels of RA in the developing ovary, and becomes strikingly male-specific. Moreover, inhibition of Cyp26b1 in the testis induces expression of specific premeiotic (Stra8) and meiotic (Dmc1, Scp3) markers, whereas blockade of RA signaling pathway in the ovary delays meiosis entry. Therefore, the authors conclude that Cyp26b1 is the “meiosis-inhibiting factor” in male mice.

More recently, the picture became more complex: Best et al. (2008) showed that changes in membrane trafficking in Sertoli cells of embryonic mouse testis induces male-to-female sex reversal and concluded to the existence of a secreted meiosis preventing substance (MPS). By using an organotypic culture system, Trautmann et al. (2008) further demonstrated that RA could in turn inhibit this MPS which is also necessary for mitotic arrest at 13.5 dpc and, therefore, induce germ cell death probably due to misunderstanding between somatic and germinal compartments in the embryonic testes. Finally, results from Suzuki and Saga (2008) indicated that Nanos2 is part of this germ cell sex-determining pathway because from 13.5 dpc, it is (1) required to maintain the suppression of meiosis by preventing Stra8 expression, (2) shown to force male differentiation when expressed in female germ cells, and (3) down-regulated by RA. Altogether, these data emphasize the complex relationship between soma and germ during gonad development as well as the key role of RA in this process but also strongly suggest the involvement of other factors. Finally, a recent molecular analysis of meiosis onset in avian embryo further suggested that a regulatory role of retinoic acid may be conserved in higher vertebrates (Smith et al.,2008).

In the present study, we analyzed meiosis onset in the amphibian Pleurodeles waltl. To determine at which developmental stage meiosis entry takes place in standard male (ZZ) and female (ZW) animals, we first analyzed the expression of the meiosis marker PwDmc1 (Dumond et al.,2008b). Then, we designed a protocol for organotypic cultures of GM complexes to evaluate the involvement of retinoid and steroid hormones in meiosis initiation. Altogether, our results demonstrate for the first time the conservation of RA-dependent mechanisms for germ cell differentiation in a urodele amphibian.

RESULTS

Meiosis Entry Occurs During Larval Life in Females and After Metamorphosis in Males

The first aim of the study was to follow germ cell differentiation during development in vivo. Therefore, partial cDNA of P. waltl Dmc1 ortholog (PwDmc1), which encodes a Rad51-like protein in the synaptonemal complex, was cloned by degenerated reverse transcriptase-polymerase chain reaction (RT-PCR) and rapid amplification of cDNA ends-PCR (RACE-PCR). Deduced amino acid sequence is 98% identical to the urodele amphibian Cynops, 92% to human and chicken. Expression of PwDmc1 mRNA was examined from stage 48 to adulthood as a marker of meiosis entry in male and female gonads (Fig. 2). At each stage, PwVasa transcript expression was measured to verify the presence of a detectable amount of germ cells. In females, the earliest expression of PwDmc1 mRNA was observed at stage 54 + 45 days (Fig. 2A) whereas in males, PwDmc1 mRNA was not detected until 8 weeks after metamorphosis (Fig. 2B). These results indicating an earlier meiosis entry in females were confirmed by a histological analysis (Fig. 3). Germ cells that have not yet entered meiosis are characterized by a very large polylobular nucleus whereas meiotic germ cell nuclei are circular and contain condensed chromatin. Once initiated, meiosis seems to be maintained all through the juvenile stage.

Figure 2.

Onset of meiosis during development of Pleurodeles waltl. Gonad–mesonephros (GM) complexes (stage 54) or isolated gonads (from stage 54 + 30 days to stage 56) were harvested from males or females. Before metamorphosis, analyses were performed on pools from 3 larvae whereas after metamorphosis, isolated gonads were sampled. Total RNA was reverse transcribed and amplification of the meiotic marker PwDmc1 was performed as described in the Experimental Procedures section. PwVasa amplification served as a control for relative germ cell amounts. Analyses were repeated twice on two RNA samples independently. The “54+xd” indicates stage 54 plus x days.

Figure 3.

Histology of ovary and testis at stage 54, 54 + 2 months, and metamorphosis + 2 months. FB, fat body; L, testis lobe; M, mesonephros; O, ovary; OC, ovarian cavity; Oo, ovocyte; Sc, Spermatocytes; T, testis. Asterisks indicate typical germ cells from each stage presented with higher magnification. At stage 54, neither female nor male gonads contain meiotic germ cells. Such cells are observed at stage 54 + 2 months in the ovary but only 2 months after metamorphosis in the testes. Scale bars = 20 μm.

Exogenous RA Induces Meiosis Entry In Vitro

To understand the physiological process initiating meiosis in Pleurodeles waltl and based on recent data highlighting the role of RA in mouse and chicken, we investigated the effect of all-trans RA on germ cells in vitro. GM complexes were carefully dissected at different stages from males or females and individually grown in presence or absence of 1 or 2 μM RA (see the Experimental Procedures section for further details). After 2 days of culture, 3 complexes were pooled for RNA extraction and PwDmc1 mRNA expression was checked by qualitative RT-PCR (Fig. 4A). Such analyses performed either in females at stage 50 or in males at stage 54 (that is before meiosis entry in vivo) revealed that 2 μM RA treatment for 2 days induces PwDmc1 transcription.

Figure 4.

Exogenous retinoic acid (RA) triggers premature meiosis entry in male and female gonad–mesonephros (GM) complexes cultured ex vivo. A: Three GM from females at stage 50 or males at stage 54 were exposed to 1 or 2 μM RA for 2 days. One microgram of total RNA was reverse transcribed and qualitative amplification of PwDmc1 was performed as previously described. Glyceraldehyde-3 phosphate dehydrogenase (GAPDH) amplification served as a control for relative RNA amounts. B: Real-time polymerase chain reaction (PCR) analyses of PwVasa and PwDmc1 expression were performed on cDNA samples obtained from male GM complexes cultured in 2 μM all-trans RA containing medium for 2 or 5 days. Results are expressed as the mean of C(t) ratios from triplicates. PwVasa amplification served as a control for relative gem cell amounts.

Furthermore, we repeated the culture experiment on stage 54 male GM complexes for 2 or 5 days in presence of 2 μM RA and measured PwVasa and PwDmc1 mRNA expressions by quantitative RT-PCR. Such analyses revealed that: (1) PwVasa expression is unaffected by 5 days organ culture; (2) PwDmc1 is transiently expressed; (3) PwDmc1 mRNA is no more detectable after 5 days in presence of RA (Fig. 4B).

These results suggest that, like in mouse and chicken, exogenous RA can trigger meiosis entry in Pleurodeles waltl. Nevertheless, in our hands, RA may be necessary but not sufficient to promote complete meiosis and/or endogenous factors rapidly degrade RA in the cultured gonads. Indeed, we cloned the RA degrading enzyme encoding gene PwCyp26b1 (GenBank accession no. EU213038) and observed a strong induction of its transcription under a 48h treatment in the presence of 2μM RA (data not shown).

Endogenous RA Induces Meiosis Entry In Vitro

To demonstrate the involvement of endogenous RA metabolism, we cloned PwRaldh2 (GenBank accession no. EU213037), the enzyme that synthesizes RA. RT-PCR analyses showed that this enzyme was expressed in organotypic cultures of GM complexes at all the stages studied. Then we tried to modulate the activity of either PwCyp26b1 or PwRaldh2. GM complexes from males at stage 54 were treated for 2 days with 2 μM R115866, an inhibitor of PwCYP26 enzyme. RT-PCR revealed a high level of PwDmc1 mRNA in response to R115866 (Fig. 5A), demonstrating that the amount of RA produced endogenously in male GM complex at stage 54 is sufficient to induce meiosis entry.

Figure 5.

PwCyp26b1 activity modulates meiosis entry in male and female gonad–mesonephros (GM) complexes cultured ex vivo. A: Three GM from males at stage 54 were exposed for 2 days to 2 μM of the Cyp26b1 inhibitor R11866 or to 1 nM of the nondegradable analogue of RA Am580. One microgram of total RNA was reverse transcribed and qualitative amplification of PwDmc1 and PwGapdh were performed as previously described. B: Histology of GM complexes cultured for 2 days in 1 nM Am580 containing medium. pGC, primordial germ cell; mGC, meiotic germ cell with condensed chromatin. C: Three GM from females at stage 54 + 90 days (after onset of meiosis) were exposed for 2 days to 60 μM of the PwRaldh2 inhibitor Citral alone or with 1 nM Am580. One microgram of total RNA was reverse transcribed and qualitative amplification of PwDmc1 and PwGapdh were performed as previously described. PwSox9 expression analysis serves as control for Citral treatment efficiency.

Moreover, this finding suggests that an in vivo PwCYP26-dependent degradation of RA could be responsible for delaying meiosis entry in male. The latter hypothesis was further supported by results from another organotypic culture experiment. We treated GM complexes from male at stage 54 with 1 nM Am580, a nondegradable analogue of RA that specifically binds to RAR-alpha receptors, and observed both the induction of PwDmc1 expression after 2 days (Fig. 5A) and the typical condensed chromatin of meiotic nuclei after 3 days (Fig. 5B). However, after 5 days, we were not able to perform RT-PCR analysis because cultured GM complexes displayed many dying germ cells with picnotic nuclei and altered structure of mesonephros. Finally, we treated female GM complexes, harvested at stage 54 + 90d (after natural meiosis initiation), with 60 μM Citral, an inhibitor of retinoic acid synthesizing aldehyde dehydrogenase (RALDH), and observed an inhibition of PwDmc1 expression, which can be reversed by a cotreatment with Am580 (Fig. 5C). In the latter experiment, endogenous PwSox9 expression was used as a control of RA signaling pathway inhibition since Afonja et al. (2002) described SOX9 up-regulation by RARalpha-selective (like Am580) but not RXR-selective agonists in breast cancer cell lines.

PwCyp26b1 Expression Is Repressed in Female Gonad at Meiotic Stage In Vivo

Because RA seems to play a key role in meiosis entry in organotypic cultures, we underwent in vivo experiments wherein GM complexes were dissected from males and females at stage 54 and stage 54 + 60d, id est before or after female meiosis onset, respectively. Western-blotting analyses of PwCYP26 and PwRALDH expressions were performed in both sexes and stages using specific polyclonal antibodies prepared and characterized in the laboratory (see the Experimental Procedures section for further details). The results presented in Figure 6A indicate that PwRALDH is expressed at the same level in males and females at both stages, whereas PwCYP26 expression is diminished in female gonads at the time of meiosis onset (54 + 60d). These results were further confirmed by quantification of the two enzyme expression levels in both sexes from stage 50 to adult stage (Fig. 6B). Before metamorphosis, protein expression was measured in GM complexes because dissection of individual gonads does not give sufficient material. Immunohistological analyses performed at stage 56 with the same antibodies confirmed a higher level of PwCYP26b1 expression in male mesonephros and gonad than in female ones, which correlates with the presence of meiotic germ cells in the ovaries but not in the testes. Such analysis also provided further information on PwRALDH and PwCYP26 localization in male or female GM complexes at the time of metamorphosis. Both enzymes are expressed in mesonephric ductules, with a higher level of CYP26b1 expression observed in male ones. RALDH2 expression is also observed into somatic and germ cells of the gonads, whereas CYP26b1 appears to be mainly expressed into testicular germ cells (Fig. 6C).

Figure 6.

Expression of the enzymes synthesizing (PwRaldh2) and degrading (PwCyp26b1) retinoic acid (RA) in larval gonad–mesonephros (GM) complexes of both sexes. A: Western blotting analysis of PwCYP26 and PwRALDH expression. A total of 30–50 μg of total protein were probed with 1:1,000 polyclonal-specific antibodies obtained and purified in our laboratory as described in the Experimental Procedures section. B: Male/female expression ratios of CYP26b1 and RALDH2 during Pleurodeles development. At stage 50, 54, 54 + 30, and 54 + 60, analyses were performed on pools of either 10 male or female gonad-mesonephros complexes; at juvenile and adult stages, expression level was determined on pools of 3 isolated gonads. Measurements were performed once at stage 50, on 2 pools at stage 54, 54 + 30, and 54 + 60 and juvenile and three times in adults. The graph shows the results of a representative experiment where all stages were analyzed on the same blot. C: Immunolocalization of the cells expressing both enzymes in male and female GM complexes at stage 56. Seven-micrometer sections of paraffin-embedded GM complexes were incubated with specific primary rabbit antibody, anti-PwCYP26 or anti-PwRALDH at the dilution 1:1,000 and then exposed to Alexa fluor 555 goat anti-rabbit IgG antibody diluted to 1:1,000 (Invitrogen Corp., Carlsbad, CA). DNA was stained with Hoechst 33342 to visualize the nuclei. Scale bars = 100 μm. GC, germ cell; MD, mesonephric duct; MuD, Müllerian duct; OC, ovarian cavity; Oo, oocyte; SC, somatic cell, WoD, Wolffian duct; asterisks indicate stained cells in testis.

After metamorphosis, Western blots were performed on gonads separated from adjacent mesonephroi. Quantification of enzyme expression levels clearly show that, before male meiosis onset occurs, PwCYP26b1 is still expressed at a higher level in male gonads than in female ones, whereas in the adult stage, this difference is no more observed. The same sex-dependent differential expression level was also obtained in kidney, but to a lesser extent (data not shown).

Hormonal Status of the Gonad May Alter Germ Cell Differentiation

Steroid hormones play a major role in sex differentiation in urodele amphibians and can even lead to complete functional sex reversal (Kuntz et al.,2003a). In Pleurodeles waltl, aromatase activity that leads to estradiol (E2) production increases in female gonads from stage 52 (Chardard et al.,1995; Kuntz et al.,2003a,b). This induction is also observed even if delayed until late stage 54, in males feminized with E2 treatment. Therefore, we investigated if the hormonal status of the gonad could modify endogenous production/degradation of RA and alter germ cell differentiation.

We analyzed PwDmc1 expression in ZZ larvae treated with E2 (100 μg/L from stage 50), or ZW larvae grown in presence of the nonaromatizable androgen DHT (400 μg/L from stage 50). No expression was detected until stage 54 + 60d (not shown). At the time of metamorphosis (stage 56), PwDmc1 expression was observed in E2-treated ZZ larvae but not in DHT-treated ZW ones (Fig. 7A). This result was correlated to the morphology of the gonad (Fig. 7B). Indeed, E2-treated ZZ animals were sex-reversed and possessed an ovary with an ovarian cavity and germ cells (meiotic or not) localized in the cortex, whereas the gonad from DHT-treated ZW larvae showed testes containing a few primordial germ cells with a polylobular nucleus in the medulla, surrounded by nonorganized somatic cells.

Figure 7.

Molecular and histological phenotype of sex-reversed animals after metamorphosis (stage 56). A: Expression of the meiotic marker PwDmc1 was assessed by semiquantitative reverse transcriptase-polymerase chain reaction (RT-PCR) on isolated gonads sampled from E2-treated males or DHT-treated females. Glyceraldehyde-3 phosphate dehydrogenase (GAPDH) amplification served as a control for relative RNA amounts. B: The global morphology of the gonads and the differentiation status of germ cells were observed after hematoxylin–eosin staining of gonads from animals of the same batch at the same stage. ED, efferent duct; FB, fat body; pGC, primordial germ cell; MGC, meiotic germ cell; OC, ovarian cavity; WD, wolffian duct. Scale bars = 20 μm. C: Transcription levels of enzymes synthesizing (PwRaldh2) or degrading (PwCyp26b1) retinoic acid were also measured in nontreated or treated animals by semiquantitative RT-PCR. One microgram total RNA was reverse transcribed, and amplifications of PwRaldh2 and PwCyp26b1 were performed as described in the Experimental Procedures section.

In an attempt to correlate germ cell phenotype and the presence of RA in hormone-treated or standard animals, we measured the mRNA expression levels of PwRaldh2 and PwCyp26b1 in vivo. At stages 54 + 90d and 56 (metamorphosis), high level of PwCyp26b1 expression was detected in gonads from standard males but not from standard females or E2-treated males. Similar levels of PwRaldh2 transcripts were observed in all samples (Fig. 7C). Therefore, the low level or absence of PwCyp26b1 expression seems to correlate with meiosis entry in sex-reversed animals.

DISCUSSION

In this study, we investigated the mechanisms that trigger germ cell differentiation in the urodele amphibian Pleurodeles waltl. We show that female germ cells enter meiosis at the end of larval life (54 + 60d), whereas in male germ cells, meiosis is delayed until juvenile stage (metamorphosis + 2 months). This result is correlated with morphological changes in germ cell nuclei. Using an organ culture system, we demonstrate for the first time in an amphibian model that RA determines germ cell fate in male gonad because addition of exogenous all-trans RA is sufficient to trigger premature meiosis entry. The same situation was described in mouse where exposure of embryonic testis to RA induces the expression of the meiotic markers Stra8, Dmc1, and Scp3 (Bowles et al.,2006; Koubova et al.,2006). Recently, a key role of RA in regulating the premeiotic marker Stra8 was also demonstrated in avian embryos through a comprehensive analysis of genes involved in RA metabolism during development (Smith et al.,2008).

To test if PwCyp26b1, the enzyme degrading RA in vivo, could be a meiosis inhibiting factor in male Pleurodeles, GM complexes were explanted before meiosis entry and grown in the presence of the specific Cyp26b1 inhibitor, R115866. This experiment demonstrated that inhibition of PwCyp26b1 (i.e., actual raise of endogenous RA level) was also able to induce PwDmc1 mRNA expression and chromatin condensation in larval male germ cells. However, prolonged exposure to RA through a 5-day treatment with Am580 is associated with nuclei picnosis. This may reflect a more general feature of Cyp26b1 activity, which was recently demonstrated in the mouse model to be not only necessary for germ cell differentiation in ovaries but also for their survival in testes (Maclean et al.,2007). Moreover, it was shown that mouse embryonic testes exposed to RA on 13.5 dpc in organotypic culture display an inhibition of mitotic arrest, which in turn, induces germ cell death with typical picnotic nuclei (Trautmann et al.,2008).

To identify the source of RA in vivo and the time course of its synthesis/degradation, we measured the expression of both PwCyp26b1 and PwRaldh2 at the protein level. In P. waltl, both enzymes are expressed in both gonad and mesonephros. Until stage 54, PwRaldh2 and PwCyp26b1 are expressed in GM complexes of both sexes at a similar level, as described in mouse and chicken before female meiosis onset (Bowles et al.,2006; Smith et al.,2008). At stage 54 + 60 days, by the time of female meiosis entry, we observed a reduction of PwCyp26b1 protein level in female complexes, suggesting that an elevation of RA level at least contributes to meiosis entry in vivo. In fact, when the same experiments are performed after metamorphosis, PwCyp26b1 transcript and protein levels also appear to be correlated to meiotic status because standard females as well as sex-reversed males display low level of PwCyp26b1 transcripts in their gonads which contain meiotic germ cells, whereas standard males or neomales (sex-reversed ZW larvae) display high levels of PwCyp26b1 transcripts but no meiotic cells. Therefore, Pleurodeles waltl germline may follow an RA-dependent differentiation process as in mouse and chicken.

In mammals, steroid hormones control genital ducts differentiation and secondary sex characteristics development, whereas in lower vertebrates like urodele amphibians, steroids act at primary steps of sex differentiation (Jost,1953; Hayes,1998). Indeed, treatments of P. waltl larvae with hormones, anti-hormones, or steroidogenesis enzyme inhibitors can lead to sex reversal (Kuntz et al.,2003a). For example, E2-treated males differentiate fertile ovaries, while treatments with androgens lead to various results: testosterone-treated males give rise to functional neofemales due to aromatization of the androgen by P450-aromatase while vestigial testes are observed in females treated with the nonaromatizable DHT (Chardard et al.,2003). Therefore, we finally addressed the question of steroid effects on meiosis entry and RA synthesis/degradation pathway in vitro and in vivo.

First, we observed that seasonal variation appeared in PwDmc1 expression in adults (from 18 months old), with male meiosis occurring all year long, whereas females seem to renew their oocyte stock only during summer (May to September) after the reproduction period (data not shown). These seasonal fluctuations of detectable number of germ cells in meiotic prophase in females are in agreement with an elevated ratio of [E2-17 beta]/[Androgen] which is stable (ratio = 0.2 to 0.3) during the major part of the year, except in July and August, when it reaches the value of 1.0 (Garnier,1985a). However, no relationships were found between the presence of meiotic germ cells that are observed along the reproductive season in males (September to May) and the two peaks of androgen plasmatic levels (October–November and March) (Garnier,1985b).

Second, we showed that meiosis entry occurs in males treated with E2 at the same time as in standard females, whereas it is delayed until after metamorphosis in DHT-treated females. Therefore, hormone-dependent sex reversal cannot only promote the reorganization of the gonad somatic compartment but also affect the germ cell differentiation process, at least in part, through the modulation of RA metabolism. Moreover, our results indicate that the transcription level of PwCyp26b1 is higher in male than in female and E2-treated neofemale, respectively, suggesting that a high level of E2 in gonad environment is associated with a high repression level of PwCyp26b1 expression.

Connections between RA signaling and E2 transduction pathways appear to be also relevant at the molecular level and seem to be a general feature of cells from mesodermic origin that undergo epithelium formation. From 13.5 dpc, mouse male germ cells stop mitotic proliferation and associate to somatic cells to form testicular cords. In fact, RA was demonstrated to block this process through PI3-kinase signaling (Trautmann et al.,2008). In addition, Giretti and Simoncini (2008) pointed out the regulatory actions of E2 in regulating, focal adhesion assembly involved in epithelial–mesenchyme transition (EMT), namely through the PI3-kinase pathway. Therefore, it would be interesting to address the potential effects of PI3K inhibitors on male gonad differentiation in the course of normal development or during E2-mediated sex reversal in Pleurodeles waltl.

Here, we currently bring the first data about the regulation of RA biosynthesis by steroids in the gonads of an amphibian species. Because it was demonstrated that the gonadic promoter of aromatase, which also drives gene expression in breast cancer, can be negatively regulated by retinoid receptors RAR and RXR (Chen et al.,2002; Rubin et al.,2002), the molecular basis of the crosstalk between RA and steroids during sex differentiation might be more complex than expected up to here and remain to be determined.

EXPERIMENTAL PROCEDURES

Animals

Embryos and larvae were staged (Fig. 1) according to the development timetable stages described by Gallien and Durocher (1957). During larval life, animals were analyzed from stage 48 to stage 56 at which metamorphosis takes place. Juveniles were also harvested every 2 weeks to study meiotic process. Before manipulation, animals were anesthetized with 0.03% benzocaine. The local Animal Care and Use Committee approved the experimental protocols, and guidelines for laboratory procedures were followed at all times.

Chemicals and Solutions

Leibovitz medium (L15) was diluted in sterile water (2:1) to respect the low osmotic pressure required for amphibian tissue culture and supplemented with 1% L-glutamine (Invitrogen, Cergy-Pontoise, France), 10% stripped fetal calf serum (Sigma-Aldrich, France), 100 U/ml penicillin G, 100 μg/ml streptomycin (Invitrogen). The concentrations of compounds in culture media were as follows: all-trans RA 2 μM, Am580 2 μM, Citral 60 μM (all three from Sigma-Aldrich), and CYP26 inhibitor R115866 2 μM (gift from Johnson and Johnson, New Brunswick, NJ). In vivo treatments were performed by addition of hormones in rearing water: Estradiol (Sigma-Aldrich) 100 μg/L, Dihydrotestosterone (Fluka, Sigma-Aldrich) 400μg/L.

Determination of Sexual Genotype and Phenotype

For sexual genotype analysis, at stage 48, a biopsy was obtained from the tail of the animals anesthetized with 0.03% benzocaine. The tissue was homogenized in 50 mM Tris-HCl, 2.5 mM MgCl2, 25 mM NaCl buffer. Samples were electrophoresed in a 0.7% agarose stacking – 10.5% starch running horizontal gel, in 25 mM Tris-Citrate pH 8.0 buffer. The patterns of peptidase-1 were revealed by specific hydrolysis of valyl-leucine substrate coupled with peroxidase colored reaction as previously described (Dournon et al.,1988).

RT-PCR

The detailed protocol for reverse transcription has been previously described (Kuntz et al.,2003b). Total RNA was extracted from pools of three GM complexes at stages 53, 54, or 54 + 60 days or from 20–50 mg of adult tissues using 200 μl of TRIzol reagent (Invitrogen Corp., Carlsbad, CA) and quantified. Total RNA (1 μg) was reverse transcribed using hexamer primers and 100 U Moloney murine leukemia virus reverse transcriptase (Invitrogen) in a total volume of 25 μl. A 2-μl aliquot of resultant cDNA was used for PCR.

The amplification was performed with 0.5 U of Taq DNA Polymerase (Invitrogen) in PCR buffer containing 25 mM of each deoxynucleotide triphosphate, 1.5 mM MgCl2, and 10 pmol of each primer in a total volume of 25 μl. Specific primers and conditions of amplification for PwGapdh, PwVasa, and PwDmc1 cDNAs have been described previously (Dumond et al.,2008b). Primers for PwCyp26b1 (dir 5′-GAC-AAG-AGC-TGC-AAG-CTG-CC-3′; rev 5′-GCT-GGA-TCT-TGG-GCA-GGT-AAC-3′), PwRaldh2 (dir 5′-CCC-ATT-GGA-GTG-TGT-GGA-CAG-ATC-3′; rev 5′-CTT –CTT-CCA-GCT-GCT-TCT-TG-3′), and PwSox9 (GenBank accession no. EU872027) (dir 5′-GAG-GGC-TCC-GAG-CAA-ACG-CAC-3′; rev 5′-GCT-CTG-CTC-GCT-GCC-CAG-TGT-3′) were used for a 35 cycles PCR (1 min 95°C; 1 min 55°C, 52°C, and 66°C, respectively; 1 min 72°C).

PCR products (12 μl) were run in a 1% agarose gel containing 0.5 μg/ml ethidium bromide.

Quantitative real-time PCR analysis was performed on an Opticon 2 System (BIO-RAD laboratories) by using SYBR-green labeling. Each cDNA sample was analyzed in triplicate with specific PwVasa or PwDmc1 primers (Dumond et al.,2008b). PwDmc1 expression was normalized to PwVasa transcript expression as a germ cell control gene.

Western Blotting

Pools of 3 gonad-mesonephros complexes from larvae were sonicated on ice in PY lysis buffer (Etique et al.,2006) for 5 sec (Branson sonifier 150, Branson Ultrasonics, Danbury, MA). After 10 min at 13,000 rpm centrifugation, supernatants were sampled and boiled for 5 min in Laemmli buffer. Total proteins (30–50 μg) were separated on a 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrotransferred to an Amersham Hybond enhanced chemiluminescence membrane (Amersham Bioscience Corp, Piscataway, NJ). The membrane was blocked for 1 hr at room temperature in Tris buffered saline containing 0.1% Tween and 5% bovine serum albumin. After blocking of nonspecific binding, membrane was incubated overnight at 4°C with a specific polyclonal anti-PwCYP26 (amino acids 2–80) or anti-PwRALDH (amino acids 6–98) rabbit serum, produced in the laboratory and diluted to 1:1,000 in the blocking buffer. The secondary anti-rabbit antibody was used following manufacturer's instructions (Amersham Bioscience Corp). Immunoreactive proteins were detected using enhanced chemiluminescence (ECL kit, Amersham Bioscience Corp).

Organ Culture

GM complexes were removed aseptically from the larvae anesthetized with 0.03% benzocaine (stage 48 to 54) under a binocular microscope. Whole explants were transferred to the in vitro culture system: the complexes were placed on a single Millipore filter (HA, pores size: 0.45 μm; Bedford, MA). The filters bearing the explants were floating on 1 ml of culture medium in tissue culture dishes and incubated at room temperature (20 ± 2°C) under noncontrolled gas composition. The complexes were maintained for 2–5 days with the medium being replaced once a day. Cultured explants were harvested for RNA extraction and RT-PCR analyses, or incubated for 48 hr in Bouin's fixative for histological analysis. No necrosis was ever observed in any part of the tissues.

Histology

GM complexes were fixed in Bouin's solution, embedded in paraffin, and sectioned at 7 μm. Sections were stained with hematoxylin–eosin–light green as described previously (Dumond et al.,2008b).

Immunohistochemistry Analysis

GM complexes were fixed overnight in 0.05 M Tris-HCl buffer pH 7.4 containing 4% paraformaldehyde, embedded in paraffin, and sectioned at 7 μm. PwCYP26 and PwRALDH immunolabeling were performed on sections deparaffinized in Histolemon (CarloErba Corp., Val de Reuil, France) and gradually rehydrated. Slices were incubated with BSA 4% for 10 min at room temperature to block nonspecific binding. Then these sections were incubated for 30 min with specific primary rabbit antibody, anti-PwCYP26 or anti-PwRALDH at the dilution 1:1,000. After two washes with Tris-HCl buffer, they were exposed for 30 min to Alexa fluor 555 goat anti-rabbit IgG antibody diluted to 1:1,000 (Invitrogen). Finally, the sections were rinsed with Tris-HCl buffer for 10 min and DNA was stained with Hoechst 33342 to visualize the nuclei. Fluorescence labeling was observed under a Nikon microscope.

Acknowledgements

We thank Eric Gelhaye for his kind help and abundant advice for antibody purification. We also thank Martine Chillet for her excellent work in histology, and Alexandra Kleinclauss and Elia Bouvry for technical assistance and animal rearing. A.W. was funded by a Syrian Government fellowship.

Ancillary