Oct4 (also known as Pou5f1) is a class V POU domain transcription factor expressed in pluripotent embryonic and germ line cells. In mammals, Oct4 is critical to maintain pluripotency of the inner cell mass (ICM) and survival of germ cells (Rosner et al., 1990; Schöler et al., 1990; Nichols et al., 1998; Pesce et al., 1998). In embryonic stem cells (ESC), Oct4 in conjunction with transcription factors Nanog and Sox2 orchestrates pluripotency and proliferation characteristics of ESC (Boyer et al., 2005; Loh et al., 2006; Wang et al., 2006). Understanding the function of these genes during embryonic development will provide a basis for proper manipulation of ESCs.
In mice, Mm-Oct4 is expressed in blastomeres, early pluripotent embryo cells, and the germ cell lineage (Rosner et al., 1990; Schöler et al., 1990; Yeom et al., 1996; Pesce et al., 1998). Mm-Oct4 mutant embryos are lethal during preimplantation stages before egg cylinder formation and suggest that Mm-Oct4 is necessary for the establishment of pluripotency of the ICM (Nichols et al., 1998). In addition, Mm-Oct4 may regulate cellular differentiation competence by controlling chromatin structure at different loci since Mm-Oct4 is necessary to maintain chromatin structure in the Nanog locus (Levasseur et al., 2008). Studies in cultured stem cell lines have confirmed the role of Mm-Oct4 in maintaining pluripotency and regulating differentiation through changes in its expression levels (Niwa et al., 2000). Additionally, Mm-Oct4 is also expressed in the primordial germ cells (PGCs) as they proliferate and migrate, and also in the maturing and adult gonads (Rosner et al., 1990; Schöler et al., 1990; Pesce et al., 1998). In humans (Hs)-Oct4 expression has also been detected in the gonads (Anderson et al., 2007). Conditional inactivation of Mm-Oct4 in the PGCs induced apoptosis of PGCs rather than differentiation, suggesting that Mm-Oct4 is necessary for PGC survival (Kehler et al., 2004). The in vivo study of pluripotency begs the introduction of additional models to mice that can complement past experiments and shed light on the evolution and molecular nature of pluripotency. Teleost fish are excellent genetic models to understand the evolution and molecular mechanisms of different developmental pathways. Fish are easy to manipulate genetically and allow for direct live observation of their transparent embryo development.
In zebrafish, spg/pou2, which is the Oct4 ortholog, is necessary for pre-gastrula morphogenesis, mid-hindbrain boundary (MHB) formation (Belting et al., 2001; Burgess et al., 2002), specification of the endoderm (Lunde et al., 2004; Reim et al., 2004), and participates in establishing the dorso-ventral axis (Reim and Brand, 2006). Although some of these functions have been conserved in mice, others, which have not been documented in mammals, may represent new functions acquired during fish evolution or function losses during tetrapod evolution. For example, the lack of Mm-Oct4 expression in the MHB argues against a specific role for Mm-Oct4 during MHB formation. In fact, conditional inactivation of Mm-Oct4 in all neural progenitor cells produced no apparent behavioral abnormalities and brain morphology was normal (Lengner et al., 2007). Moreover, spg/pou2 expression has not been reported in the developing PGCs or the adult gonads and, thus, its mRNA expression in the mature oocytes may only reflect mRNA accumulation for use during early stages of embryonic development before mid-blastula transition (Marlow and Mullins, 2008).
Thus, the study of the pluripotency network in fish has been hampered by the absence of a Nanog ortholog in fish as well as the different roles of spg/pou2 in zebrafish compared to mammals, which suggested that the mammalian pluripotency gene networks were either not present or only partially present in zebrafish. However, we have recently characterized the Nanog gene in medaka and showed that Ol-Nanog regulates proliferation of early embryonic cells (Camp et al., 2009), as human (Hs)-Nanog does in human ESC cultures (Zhang et al., 2009). To characterize a second component of the pluripotency network, we assessed Ol-Oct4 mRNA and protein pattern of expression in medaka embryos. In this study, we present the mRNA and protein expression patterns of Ol-Oct4 during embryonic development. Ol-Oct4 expression pattern is different from that of zebrafish spg/pou2 as Ol-Oct4 is expressed in adult gonads and no detectable expression is observed in the brain region after st. 19. Our results in medaka show that Ol-Oct4 has a similar expression pattern to that of Mm-Oct4 suggesting that Ol-Oct4 may have similar roles to its mammalian homologs. These results further validate the suitability of the medaka fish as a model for the study of pluripotency mechanisms and evolution.
RESULTS AND DISCUSSION
Ol-Oct4 Is the Homolog of Mammalian Oct4
There is currently a predicted medaka Oct4 sequence present in the database (GenBank AY639946). However, in order to identify putative duplications of the medaka Oct4 homologue, we performed different genome searches in the medaka genome using the described Ol-Oct4 sequence and members of the PouV family of transcription factors from zebrafish, Xenopus, mouse, and human. We did not find additional copies of the gene or closely related homologs in medaka. Alignment of the Ol-Oct4 predicted protein with other Oct4 proteins from humans, mice, Xenopus (Xl-Oct25), and zebrafish (Pou2/Spg) shows that conservation is highest at the POU domain as compared to the full-length protein (Fig. 1A, C). It is interesting to note the conservation outside the POU domains of several Proline residues among all species compared at the medaka positions 117, 188, 433, and 442 (Fig. 1A). Additionally, synteny analysis of Ol-Oct4 shows that it is located near the Clic1 and Fut7/Q5FP2_ORYLA genes as zebrafish pou2/spg (Fig. 1B). The syntenic relationship between Zebrafish pou2/spg and Xenopus, chicken, and mouse Oct4 genes has been described elsewhere (Burgess et al., 2002; Morrison and Brickman, 2006; Lavial et al., 2007). These data suggests that the described Ol-Oct4 gene is the ortholog of the mammalian Oct4 gene.
Ol-Oct4 Is Expressed During Early Embryo Development, and in PGCs and Adult Gonads
Ol-Oct4 expression in the developing embryo
To examine the expression pattern of Ol-Oct4, we first identified the different stages at which it is expressed using RT-PCR on mRNA isolated from embryos at different stages. We detected Ol-Oct4 expression in all stages from the unfertilized egg to the fry stage (Fig. 2A), as opposed to another pluripotency regulatory gene Ol-Nanog (Camp et al., 2009), which is not detected from st. 20 of embryonic development until fry stages. We cloned the whole predicted cDNA sequence for Ol-Oct4 and performed whole mount in situ hybridization (WMISH) expression analysis at different stages of embryonic development. Ol-Oct4 was detected in all cells of the early developing embryo (Fig. 2B–E) until blastula stages (st. 10). This expression is similar to that of zebrafish spg/pou2 (Belting et al., 2001) and Mm-Oct4 (Yeom et al., 1991). At st. 15 (mid-gastrula), Ol-Oct4 expression was limited to a central group of cells around the embryonic streak (Fig. 2G). At st. 17 (early neurula), expression was detected in the middle of the embryonic body and at st. 18 (late neurula, optic bud formation) a second domain of Ol-Oct4 expression appeared at the posterior tip of the embryonic body, similar to the expression of spg/pou2 in the posterior part of the embryo (Fig. 2H, I). However, as previously observed in mice, we did not detect Ol-Oct4 expression during brain regionalization, suggesting that spg/pou2 expression in the zebrafish brain is not conserved among teleost fish. This observation indicates that some Oct4 functions could be different between medaka and zebrafish, whose lineages branched out 320 Myrs ago, very early during teleost evolution (Kasahara et al., 2007). At st. 19 (2-somite stage), expression in the central region is lost and the posterior domain of expression is stronger and remains strong at least until st. 23 (Fig. 2J–L; 12 somites, tubular heart).
To further characterize this Ol-Oct4 dynamic embryonic expression pattern, we generated polyclonal antibodies against a specific 20-mer peptide of the Ol-Oct4 protein. At st. 7 (32-cell stage), the Ol-Oct4 protein could be clearly detected in the nucleus of the blastomeres, with some cytoplasmic background staining similar to that obtained with the same serum after incubating it with the immunoreactive peptide or in a secondary antibody-only staining (Fig. 3A–C and data not shown). To test whether the antibody of Ol-Oct4 was specific for the Ol-Oct4 protein, we injected a specific Ol-Oct4 morpholino (MO-Oct4) that inhibits mRNA translation. Injection of a control morpholino (MO-C) did not affect Ol-Oct4 nuclear staining (Fig. 4A–C), whereas analysis of Ol-Oct4 protein levels in vivo after MO-Oct4 injection showed loss of Ol-Oct4 staining in the nucleus (Fig. 4D–F). This result confirms that the generated antibody specifically detects Ol-Oct4. At st. 10 (early blastula), nuclear staining was maintained except in cells undergoing mitosis, as detected using DAPI staining (Fig. 3D–I). Intriguingly, cells presented cytoplasmic punctuated staining for Ol-Oct4 (Fig. 3D–L), which may suggest the existence of a putative translocation regulatory mechanism for this transcription factor. At st. 25 (18 somites, onset of blood circulation), Ol-Oct4 was not detected in the posterior part of the embryo where the mRNA was detected. This may suggest further complexity in the regulation of mRNA translation. On the other hand, Ol-Oct4 expression was observed to co-localize with GFP from the Nanos-3′UTR-GFP construct (Fig. 5A–E, J–M) known to only be expressed in PGC (Saito et al., 2006). However, only a subset of PGC co-expressed Ol-Oct4 (Fig. 5A–I, J–M). This result was further confirmed in transgenic Olvas-GFP embryos, which express GFP under the control of the medaka Vasa gene promoter (Tanaka et al., 2001). Ol-Oct4 expression was detected only in a few GFP-positive cells of the Olvas-GFP transgenic embryos (Fig. 5N–Q). Our expression results suggest that Ol-Oct4 may play a role in PGC biology in medaka, which may be similar to that in mice, where Mm-Oct4 is necessary for PGC survival (Kehler et al., 2004).
Ol-Oct4 expression in the adult gonads
Several pluripotency genes are expressed in the adult gonads where they may play a role in gonadal stem cell maintenance and/or gamete differentiation. Such is the case of Oct4, which has been detected in mouse and human gonads (Pesce et al., 1998; Anderson et al., 2007). We analyzed whether Ol-Oct4 was also expressed in the adult gonads of medaka. Analysis using RT-PCR and WMISH showed that Ol-Oct4 was expressed in the ovary and weakly in the testis (Fig. 6A–C). Further analysis of protein expression in the testis showed that Ol-Oct4 expression was confined to the periphery where undifferentiated spermatogonia are located (Fig. 6D–K), which constitutes the germ stem cell population of the testis. This expression pattern in the medaka testis is similar to the adult mouse testis where Mm-Oct4 is expressed in the most undifferentiated type A spermatogonia population (Pesce et al., 1998) and is necessary for spermatogonia stem cell self-renewal (Dann et al., 2008). In the ovaries, Ol-Oct4 expression was detected in the nuclei of peripheral cells and small oocytes. Additionally, during oocyte maturation Ol-Oct4 expression was detected in the chromosomic material and in a punctuated pattern around this nuclear staining (Fig. 6L–O). Moreover, Ol-Oct4 expression was also detected in the germ plasm (Fig. 6L–S) as identified by actin expression (Marlow and Mullins, 2008). Staining with only the secondary antibodies showed that Ol-Oct4 and actin expression patterns were specific (Fig. 6T–W). The protein expression results in testis and ovaries suggest that Ol-Oct4 plays a role during gamete maturation beyond the accumulation of maternal mRNAs and proteins. In fact, in mice, conditional inactivation of Mm-Oct4 in the PGCs provoked a sterility phenotype in the adult mice, with females containing few primordial or growing follicles and males showing complete or partial depletion of spermatogonia (Kehler et al., 2004). Also, Mm-Oct4 regulates the developmental competence of mouse oocytes, indicating that Mm-Oct4 has a functional role in gamete maturation (Zuccotti et al., 2008).
In conclusion, our results suggest that Ol-Oct4 is the homolog of mammalian Oct4 and has an expression pattern more similar to that of mammalian Oct4 during embryonic development and in the adult gonads. On the other hand, Ol-Oct4 does not share some expression domains with zebrafish spg/pou2; mainly, Ol-Oct4 is not detected during regionalization of the brain and it is, however, expressed in the PGCs and adult gonads. These expression results suggest that Ol-Oct4 may play similar roles to those of mammalian Oct4. Along with the recent functional characterization of Ol-Nanog (Camp et al., 2009), our expression results suggest that the pluripotency mechanisms controlled by the Oct4-Nanog-Sox2 triad may be conserved between mammals and medaka fish. This suggests that Medaka could serve as a model for studying pluripotency in vivo.
Adult medaka (Oryzias latipes) CAB strain animals (kindly provided by Paola Bovolenta, CSIC, Madrid, Spain) and Olvas-GFP transgenic animals (kindly provided by Minoru Tanaka, NIBB, Okazaki, Japan) were kept in recirculating water aquaria at 28°C on a 14-hr light/10-hr dark daily cycle. Embryos were collected by natural spawning in Yamamoto solution (Yamamoto, 1975) and staged as previously described (Iwamatsu, 2004). Embryos were raised at 25°C.
Total RNA was extracted from groups of 40 embryos at the desired stage or from different adult dissected tissues using Trizol reagent (Sigma, St. Louis, MO). cDNAs were synthesized from 1 μg of total RNA using Random primers hexamers (Roche, Nutley, NJ) and Superscript III reverse transcriptase (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Equal amounts of first-strand cDNAs were then amplified using primer pairs for the studied genes. The primer pair sequences were as follow: Oct4 F: 5′ACACGGATGAGGAG TATTC 3′; Oct4 R: 5′GTTTACGCCAA AGAAAGC 3′; Actin F: 5′ACTGGAAC AATGAAGGAG 3′; Actin R: 5′ATACA AGCAATACTACAACC 3′. The PCR reactions were performed using 35 cycles with a denaturation step at 94°C for 35 sec, annealing at 54.7°C for 50 sec, and extension at 72°C for 30 sec. After amplification, the reaction product was analyzed in a 2% agarose gel containing ethidium bromide (1 μg/mL) and images of the bands were captured on a Versadoc Image analyzer (BioRad, Hercules, CA). A negative control to confirm the absence of gDNA was performed using Actin primers on RNA without Reverse Transcriptase. RT-PCR experiments for each gene were performed three times.
WMISH and Protein Expression Analysis
Ol-Oct4 cDNA was cloned in pCS2+ using appropriate primers for amplification of the predicted sequence and subsequent sequencing (GenBank AY639946). Ol-Oct4 probe synthesis was produced by digesting plasmids with XhoI and transcribing with T7 polymerase. WMISH was performed as described (Camp et al., 2009) on embryos fixed at the desired stages. Sense probes were used as controls (Fig. 2F).
For immunohistochemistry, the medaka-specific anti-Ol-Oct4 antibody was raised in rabbits (Abnova, Taipei City, Taiwan) against the following peptide sequence QTREQIKMPEI KIEKDTDEE. Anti-Ol-Oct4, anti-PCNA (Santa Cruz Biotechnology, Santa Cruz, CA; Ref. Sc-56) and anti-Actin (Abcam, Cambridge, MA; Ref. ab6276) were used at 1:200 in embryos fixed at desired stages in 4% paraformaldehyde (PFA)/1×PBSTw or gonadal cryosections. For anti-Ol-Oct4 staining, embryos were first pretreated with 10 μg/mL proteinase K for 1 min for embryos until st. 10, for 20 min for embryos at st. 25, for 5 min for testis cryosections, and for 7 min for ovary cryosections. Embryos and gonadal cryosections were incubated with the appropriated secondary antibodies: Alexa488-conjugated anti-rabbit (Invitrogen, Cat. No. A11008), Alexa488-conjugated anti-mouse (Invitrogen, Cat. No. A11001), Cy3-conjugated anti-rabbit (Jackson ImmunoResearch, West Grove, PA; Cat. No. 111-166-045), or Cy3 conjugated anti-mouse (Jackson ImmunoResearch, Cat. No. 115-166-062) at 1:500 and DAPI. Control immunofluorescence experiments were performed using either secondary antibodies alone, pre-immunization serum (pre-serum), or pre-adsorbed Oct4 antibody with the immunogenic peptide. Pre-adsorption of the Oct4 antibody with the immunogenic peptide was performed at an antibody-to-peptide ratio of 1 mol:20 mol in 1% BSA and PBS, and incubating overnight at 4°C with rotation. The treated solution was then used as primary antibody in immunohistochemistry experiments (Fig. 3M–R; Fig. 6T–W). Finally, all embryos or cryosections were mounted using fluoromount G medium.
mRNA and Morpholino Injection
pEGFP-nanos 3′UTR (EGFP ORF fused to the 3′UTR of nanos; kindly provided by Minoru Tanaka, Okazaki, Japan) was used for mRNA synthesis using the SP6 Ambion mMessage mMachine Kit (Ambion, Inc., Austin, TX). The morpholino (MO; Gene Tools LLC) sequences are: MO-Oct45′-CTGAAAATGGCGCAGCTTTTTAGT G-3′ and control MO (MO-C) 5′-CCTCTTACCTCAGTTACAATTTATA-3′. MOs in a concentration of 0.5 mM or synthesized mRNA in a concentration of 200 ng/μL were injected in CAB embryos at st. 2 using a pressure Narishige IM300 microinjector.
We thank M. Tanaka and P. Bovolenta for reagents and Deborah Burks for encouragement and helpful comments. This work was supported by grants from Spanish MEC (BFU2005-09186/BFI and BFU2008- 04624/BFI to J.L.M), the Valencian Health Conselleria and Spanish ISCIII: Regenerative Program of the Valencian Community (to J.L.M), and FIS 02/3003 and PI070789 (to A.G.-E.).