Specification of primordial germ cells in medaka (Oryzias latipes)
© Herpin et al; licensee BioMed Central Ltd. 2007
Received: 29 August 2006
Accepted: 11 January 2007
Published: 11 January 2007
Skip to main content
© Herpin et al; licensee BioMed Central Ltd. 2007
Received: 29 August 2006
Accepted: 11 January 2007
Published: 11 January 2007
Primordial germ cells (PGCs) give rise to gametes that are responsible for the development of a new organism in the next generation. Two modes of germ line specification have been described: the inheritance of asymmetrically-localized maternally provided cytoplasmic determinants and the induction of the PGC fate by other cell types.
PGCs specification in zebrafish appears to depend on inheritance of germ plasm in which several RNA molecules such as vasa and nanos reside. Whether the specification mode of PGCs found in zebrafish is general for other fish species was brought into question upon analysis of olvas expression – the vasa homologue in another teleost, medaka (Oryzias latipes). Here, in contrast to the findings in zebrafish, the PGCs are found in a predictable position relative to a somatic structure, the embryonic shield. This finding, coupled with the fact that vasa mRNA, which is localized to the germ plasm of zebrafish but does not label a similar structure in medaka opened the possibility of fundamentally different mechanisms governing PGC specification in these two fish species.
In this study we addressed the question concerning the mode of PGC specification in medaka using embryological experiments, analysis of RNA stability in the PGCs and electron microscopy observations. Dramatic alterations in the somatic environment, i.e. induction of a secondary axis or mesoderm formation alteration, did not affect the PGC number. Furthermore, the PGCs of medaka are capable of protecting specific RNA molecules from degradation and could therefore exhibit a specific mRNA expression pattern controlled by posttrancriptional mechanisms. Subsequent analysis of 4-cell stage medaka embryos using electron microscopy revealed germ plasm-like structures located at a region corresponding to that of zebrafish germ plasm.
Taken together, these results are consistent with the idea that in medaka the inheritance of maternally provided asymmetrically-localized cytoplasmic determinants directs cells to assume the germ line fate similar to zebrafish PGCs.
Two basic strategies for the specification of primordial germ cells (PGCs) have been described. In Drosophila, C. elegans and Xenopus the inheritance of germ plasm with associated germ cell determinants appears to be required for PGC formation [1–3]. In these organisms, maternally provided asymmetrically-localized germ plasm, or germ plasm-like structures that are visible in electron micrographs (so called "nuage") are found in positions where the PGCs are formed. In Drosophila it has been shown that the germ plasm has an instructive role in directing nuclei to the germline  while zebrafish embryos from which the germ plasm was removed lacked germ cells . In mouse (and by extension other mammals) and urodelean amphibians the mechanisms of germline specification are different. Here, morphologically distinct germ plasm has not been identified in early embryos but rather, as revealed in transplantation experiments, cellular interactions are thought to be responsible for PGC specification [6, 7]. Consistently, the formation of the founding population of PGCs in the mouse was shown to depend on extracellular factors of the bone morphogenetic protein (BMP) family [8–11]. Whereas the actual specification and formation of PGCs in mammals and in urodeles appear to be independent of germ plasm, material resembling nuage, a germ plasm organelle, is found in germ cells of these organisms at later developmental stages [12, 13].
Germ-cell specification in fish was studied mainly in bony fish (teleosts) where the work relied on observations using light and electron microscopy. The germ cells were recognized by their large size and by the typical electron dense nuage-like structures. Using these criteria, the earliest time point at which germ cells could be recognized in teleosts was around the onset of somitogenesis [14–16]. Some investigators deduced that the position where the PGCs were first identified was their site of origin (For references see ). Transplantation experiments of the embryonic shield, the fish organizer, in the teleost Fundulus, to the extraembryonic yolk sac of hosts resulted in, among other tissues, the development of germ cells and gonads with the highest frequency of germ cells found when the posterior quarter of the Fundulus shield was grafted [18, 19]. One could not exclude the possibility however, that a small population of PGCs existed already before the end of gastrulation and migrated to the position where they were identified either morphologically or by embryological transplantation experiments. Thus, the question of the time and place of germ-cell origin in fish remained open.
The findings obtained in zebrafish were extended to suggest that generally in fish, PGCs are specified by determinants found in the germ plasm and that the position of the germ plasm could be inferred from the position of RNA molecules such as vasa and nanos which mark the early germ plasm. Surprisingly, the cloning of the vasa and nanos genes from another teleost, medaka (Oryzias latipes) and analysis of their mRNA expression patterns did not provide direct support for the notion that germ plasm is responsible for PGC specification in all fish species [31, 32]. In contrast to the enrichment of vasa transcripts in the position of the germ plasm observed in zebrafish, the transcripts of olvas, the medaka vasa homologue, were uniformly distributed until stage 16 at late gastrulation (, Fig. 1B and 1D). At this stage, PGCs, recognized by higher levels of olvas expression, are found on either side of the shield, the fish organizer . Similarly, medaka nanos transcripts did not aggregate at the two first cleavage planes (Fig. 1E to 1H). As mentioned above, in zebrafish the germ plasm contains vasa and nanos mRNAs and is found at the distal parts of the first two cleavage planes. These four locations are randomly oriented relative to the dorsal aspect of the embryo leading to specification of PGCs in random dorsoventral positions . In contrast, at these stages, olvas and nanos mRNAs were found to be uniformly distributed in the corresponding stages and when the PGCs are first observed using olvas as a molecular marker, they are found in a defined position relative to somatic structures. If the early expression site of vasa and its homologues indeed mark the site and stage of PGC origin, then the expression pattern of olvas would imply a different mechanism for specifying PGC in medaka.
In this work we investigate the question of the mechanisms responsible for PGC specification in medaka. We determine the effect of perturbations in somatic development on PGC specification and describe the properties of medaka PGCs regarding the metabolism of specific RNA molecules. The results presented here are consistent with the notion that despite the differences in the expression pattern of vasa during early embryogenesis in these two teleosts that represent totally different branches of the phylogenetic tree, it is likely that similar basic mechanisms are responsible for PGC specification.
Several lines of evidence strongly indicate that in mouse the PGC fate is induced by signals that originate in adjacent tissues. In this organism, germ cells arise just before or during early gastrulation, around the proximal part of the epiblast adjacent to extra-embryonic tissues [6–11]. The finding that medaka PGCs are first detected at a specific position relative to the embryonic shield ( and Fig. 1D) could be interpreted as evidence for induction of PGCs by the adjacent somatic cells. Consistent with this suggestion, axial mesodermal tissues in fish are known to be capable of inducing cell fate in adjacent tissues (e.g. ).
The first manipulation was carried out by injection of 600 pg β-catenin RNA into one marginal blastomere at the 8- or 16-cell stage (based on the protocol for zebrafish by ). This resulted in formation of a complete secondary axis in about 3% of the injected embryos. Control embryos were injected with the same amount of GFP-globin RNA. The injected embryos were fixed at stages 17 and 23 and whole-mount in situ hybridization was performed using an olvas antisense RNA probe to visualize the PGCs (for stage 23 see Fig. 2A and 2B). The position of PGCs in embryos with an induced secondary embryonic axis was similar to that observed in control embryos of the same developmental stage. If PGC fates are indeed induced by somatic tissues of the organizer, one would expect an increase in PGC number when two shields are present in the embryo. However, the average number of PGCs in embryos with duplicated axes was found to be similar to the number of PGCs in control embryos (Fig. 2C).
Mesoderm patterning during early gastrulation was modulated by either overexpressing BMP2 (Fig. 2H to 2K) or expressing a dominant negative form of BMP2/4 receptor I (DN(BMP2/4RI)) (Fig. 2L to 2O), which results in ventralization and dorsalization of mesoderm, respectively. Again, no alteration in the total number of PGCs was observed, although their localization was severely impaired (Fig. 2D to 2P). Thus, the number of PGCs is independent of mesoderm patterning.
These findings do not support the idea that PGC fate is induced by neighboring cells at the site where they are first detected. Rather, these results are more consistent with the model according to which PGCs in medaka are specified by inheritance of pre-localized maternal components whose amount is the limiting factor determining the number of PGCs. As BMP-induced mesoderm formation occurs during gastrulation and the induction of the secondary axis by β-catenin arises at a very early stage of development well before gastrulation initiates, we consider it unlikely that the PGCs are specified by an earlier inductive process.
As mentioned above, in Drosophila and C. elegans the PGCs are specified by inheritance of maternally provided asymmetrically-localized cytoplasmic determinants. In these organisms, maternally provided transcripts that are expressed in the PGCs are selectively stabilized within these cells while being degraded in somatic cells [35, 36]. Similarly, in zebrafish the mRNA of the vasa gene , as well as nanos1 (nos1) mRNA [29, 38] are degraded in somatic cells but are stabilized and translated in cells that contain germ plasm.
If specification of PGCs in medaka occurs by inheritance of germ plasm similar to zebrafish, then the phenomena of differential stability of RNAs between somatic and primordial germ cells is expected to take place in this species, as well.
To determine whether this mechanism is relevant for the observed olvas mRNA and protein distributions, a reciprocal experiment was performed in which the original full-length olvas RNA was injected in zebrafish embryos. The analysis of its fate with an antisense olvas probe revealed no specific stabilization of olvas RNA in the PGCs of zebrafish (data not shown). Thus, neither medaka olvas 3'UTR nor olvas open reading frame (ORF) mediate germ cell specific RNA stability in zebrafish. Interestingly, in medaka, olvas-GFP protein expression was rapidly restricted to PGCs (Fig. 3E to 3I), although the olvas-GFP fusion RNA itself was not efficiently differentially stabilized in the PGCs of medaka (data not shown). Therefore, the preferential localization of the olvas-GFP protein to the PGCs (Fig. 3E to 3I) could be attributed to complementary specific posttranscriptional mechanisms: (i) in agreement with Kurokawa et al., a PGC-specific enhanced translation due to the olvas 3'UTR . (ii) Somatic inhibition of translation due to the open reading frame itself. (iii) Soma-specific destabilization of the protein .
In addition, like olvas 3'UTR in medaka, zebrafish vasa 3'UTR has been shown to be essential for proper protein PGC localization in zebrafish. Hence, when injected in medaka, GFP-zfvasa 3'UTR failure of selective GFP expression (Fig. 3J to 3N compared to Fig. 3E to 3I and 3O to 3S) points out possible specie-variability of either targeted cis-acting RNA sequence elements or RNA-interacting protein sequence, but likely using a similar conserved mechanism.
Thus, while we provide evidence for the existence of conserved mechanisms governing PGCs specification in zebrafish and medaka, the similarities apparently can not be extended to the control of PGC-specific expression of every molecule, as shown here for the zf-vasa and olvas RNAs. Hence, the mechanisms directing PGC-specific expression of olvas could depend on combined machineries associating enhanced/repressed specific soma/germline translation, protein stabilization and possibly, to a lower extend, RNA stability.
Our findings in medaka are consistent with the suggestion that the mode of PGC specification is conserved with zebrafish and possibly other teleosts. First, we show that dramatic alterations in somatic development (axis duplication and mesoderm formation alteration) do not alter the PGC number in medaka, which appears inconsistent with an inductive mode of PGC specification that depend on signals from somatic cells. Second, the process of germ-line specific RNA stabilization shown here for medaka has been described for those organisms whose PGCs are specified by inheritance of maternally provided cytoplasmic determinants. Consistently, material resembling germ plasm was identified in 4-cell stage medaka embryos. The apparent paradox that the vasa-homologue of medaka differs from its zebrafish counterpart with respect to its early spatial distribution and its ability to be subjected to the "degradation-protection" control mechanism may be reconciled by the finding that vasa mRNA appears to be dispensable for early PGC development in zebrafish .
β-catenin sense RNA was transcribed from pT7Ts , which contains the full-length zebrafish β-catenin cDNA. The injection of the zebrafish 600 pg β-catenin RNA into one of the marginal blastomeres of 8- and 16-cell stage medaka embryos induced complete secondary axis in 3% of the embryos. As a control, embryos were injected with 600 pg of GFP-globin RNA.
Zebrafish BMP2b and dominant negative truncated version of BMP2/4 receptor I (DN(BMP2/4RI)) were transcribed from pSP64TS and pSP64T respectively. Capped synthetic mRNA were injected (1 nL) into the cytoplasm of a single-cell Medaka embryo at different concentrations (2 to 50 ng/μL and 40 to 200 ng/μL for BMP2b and DN(BMP2/4RI) respectively).
The GFP-nos1 3'UTR construct includes the mmGFP5 ORF cloned upstream of the 3'UTR of the zebrafish nanos1 gene . The GFP-zfvasa 3'UTR construct was made by inserting the PCR amplified zebrafish vasa 3'UTR after GFP (XhoI crated site).
The olvas-GFP construct was prepared by inserting the mmGFP5 ORF into the SacII site in the full-length olvas cDNA (obtained from the cab-wild-type strain using the forward oligo AAAAAGCTTTCAGTTTGAAGCTAACAGCAGCAC and reverse oligo AAATCTAGATTTGKTGAAAACTTTTAATTATYAGGAGAA.
GFP-globin: mmGFP5 was cloned in between the globin 5'UTR and globin 3'UTR of pSP64T .
All RNA molecules were synthesized using the mMessage mMachinekit (Ambion).
For transmission electron microscopy individual eggs were fixed by immersion in a solution containing 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate buffer at pH 7.4 for 4 h. The chorion was subsequently removed and the eggs were fixed further overnight at 4°C using 2% glutaraldeahyde in 0.1 M phosphate buffer at pH 7.4. After an additional fixation with 1% OsO4 and pre-embedding staining with 1% uranyl acetate, eggs were dehydrated and embedded in EMbed 812 resin (Plano).
The sectioning was performed using a Leica Ultracut ultramicrotome. 600 nm thick sections were stained with tolouidine blue and visualized in a Zeiss light microscope to select the area of interest. Thereafter, 60 nm thick sections were collected and counterstained with 1% uranyl acetate and lead citrate and examined using a Philips CM 120 BioTwin (Philips Inc.Eindhoven, The Netherlands) transmission electron microscope.
We thank S. Schories for Betta splendens embryos, and J. Wittbrodt and F. Loosli and M. Tanaka for the medaka strains and medaka nanos probe, help and encouragement. We thank M. Köprunner and M. Reichman for helpful comments on the manuscript. This work was supported by a grant of the Rudolf-Virchow-Zentrum for Experimental Medicine (DFG Forschungszentrum) to M.S.
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.