Developmental Biology 317 (2008) 576–584

Contents lists available at ScienceDirect

Developmental Biology

j ourna l homepage: www.e lsev ie r.com/deve lopmenta lb io logy

Requirement of Oct3/4 function for germ cell specification

Daiji Okamura a,b,c, Yuko Tokitake a,b,c, Hitoshi Niwa d, Yasuhisa Matsui a,b,c,⁎a Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer, Tohoku University, 4-1 Seiryo-machi, Sendai 980-8575, Japanb Department of Molecular Embryology, Research Institute, Osaka Medical Center for Maternal and Child Health, 840 Murodo-cho, Izumi, Osaka 594-1101, Japanc CREST, Japan Science and Technology Agency, Saitama 332-0012, Japand Puluripotent Cell Studies, The Center for Developmental Biology, RIKEN Kobe Institute, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan

a r t i c l e i n f o

⁎ Corresponding author. Cell Resource Center for BioDevelopment, Aging and Cancer, Tohoku University, 48575, Japan.

E-mail address: ymatsui@idac.tohoku.ac.jp (Y. Matsu

0012-1606/$ – see front matter © 2008 Elsevier Inc. Aldoi:10.1016/j.ydbio.2008.03.002

a b s t r a c t

Article history:Received for publication 26 November 2007Revised 3 March 2008Accepted 3 March 2008Available online 14 March 2008

In mammalian embryos, PGCs (primordial germ cells) are specified from a pluripotent epiblast cellpopulation after implantation. In this study, we demonstrated an essential role for the germline-specifictranscription factor Oct3/4 in PGC specification. We generated chimeric embryos with ZHBTc4 ES cells lackingboth alleles of the Oct3/4 gene (pou5f1). Pluripotency was maintained by an Oct3/4 transgene, and itsexpression was suppressed by doxycycline (Dox). Transcription of the Oct3/4 transgene in the ES-derivedcells unexpectedly suffered constitutive suppression in chimeric embryos without Dox, and the ES-derivedcells contributed to PGC precursor-like cells, but failed to form PGCs. We then attempted to rescue Oct3/4expression in the ES-derived cells in the chimeric embryos by introducing an additional Oct3/4 transgene.The ES cell-derived cells indeed recovered Oct3/4 transcription in these chimeric embryos, and weresuccessfully specified to PGCs. We further confirmed the requirement of Oct3/4 by using another derivative ofZHBTc4 ES cells in which a Dex (dexamethasone)-dependent Oct3/4 transgene was introduced. In thepresence of Dox, Oct3/4 protein was absent in the nuclei of the ES-derived cells, which failed to form PGCs. Incontrast, the ES-derived cells could be specified to PGCs after activation of Oct3/4 function in the presence ofDex.

© 2008 Elsevier Inc. All rights reserved.

Keywords:Primordial germ cellEmbryonic stem (ES) cellOct3/4EpiblastChimeric embryoPrecursorSpecificationPluripotency

Introduction

The protein Oct3/4 is encoded by pou5f1 and represents agermline-specific transcription factor containing a POU-homeo boxdomain (Okamoto et al., 1990; Rosner et al., 1990; Scholer et al., 1990).Previous reports have indicated that Oct3/4 is essential for themaintenance of pluripotent cells. Mouse embryos deficient in pou5f1cannot develop beyond the blastocyst stage at pre-implantation andfail to establish a pluripotent inner cell mass (Nichols et al., 1998). Inaddition, two-fold up- or down-regulation of Oct3/4 in ES cells causesdifferentiation to endoderm/mesoderm or to troph ectoderm,respectively, indicating that a narrow range of the amount of Oct3/4 is critical for maintaining pluripotency (Niwa et al., 2000).Relatively small shifts in Oct3/4 expression thus lead to commitmentof stem cells to different cell lineages. As Oct3/4 is specificallyexpressed in germline cells, this molecule may be crucial throughoutthe development of germline cells (Pesce et al., 1998; Pesce andScholer, 2000) and has actually been reported to be necessary forsurvival of migrating PGCs in later stage embryos (Kehler et al., 2004).

medical Research, Institute of-1 Seiryo-machi, Sendai 980-

i).

l rights reserved.

However, clarification of the specific functions of Oct3/4 in germ cellspecification has remained elusive.

In mouse embryos, primordial germ cells (PGCs) are specified froma population of pluripotent epiblast cells which are also the origin ofall somatic cells after implantation (Ying et al., 2001; McLaren, 2003;Matsui and Okamura, 2005; McLaren and Lawson, 2005; Hayashi etal., 2007). The precursors of PGCs also give rise to extra-embryonicmesoderm including the allantoic mesoderm, and exist at theproximal rim of the epiblast in pre-gastrulating embryos (Lawsonand Hage,1994). Those cells may be induced and/or maintained by thefunction of BMP signaling from the adjacent extra-embryonicectoderm (Lawson et al., 1999; Ying et al., 2000; Yoshimizu et al.,2001; Chang and Matzuk, 2001; Ying et al., 2001; Tremblay et al.,2001; Hayashi et al., 2002; Chu et al., 2004; Chuva de Sousa Lopes etal., 2004; Okamura et al., 2005). Blimp1, known as a transcriptionalrepressor, is initially detected in lineage-restricted PGC precursorsin proximal epiblast and functions as a critical determinant of thegerm cell lineage by transcriptional repression of homeobox genesessential for somatic cell differentiation (Vincent et al., 2005;Ohinata et al., 2005). Before finally differentiating to PGCs, theprecursors cluster and interact with each other in extra-embryonicmesoderm, and this interaction causes the expression of PGC-specificgenes such as Stella/PGC7 and tissue non-specific alkaline phos-phatase (TNAP) (Sato et al., 2002; Saitou et al., 2002; Okamuraet al., 2003). In this study, we examined the function of Oct3/4 inPGC specification by generation of chimeric embryos with ES cells

577D. Okamura et al. / Developmental Biology 317 (2008) 576–584

showing distinct levels of Oct3/4 expression, and found that Oct3/4plays an essential role when the clustered PGC precursors are finallyspecified into PGCs.

Material and methods

ES cell culture and generation of chimera embryos

G3-3 ES cells were generated by transfecting a pCAG-EGFP-IP construct into ZHBTc4ES cells (Niwa et al., 2000, 2002). G4-2 ES cells were obtained by transfecting a pCAG-EGFP-IP construct into ES cells containing one allele of pou5f1. GOWT1 and 4EOGR1 EScells were generated by introducing the pCAG-EGFP-Oct3/4-IP and pCAG-EGFP-Oct3/4-GR-IP fusion constructions, respectively, into ZHBTc4 ES cells. EGFP-Oct3/4-GR wasgenerated by in-frame fusion of EGFP-Oct3/4 and the ligand binding domain of humanglucocorticoid receptor. These ES cells were maintained as described previously (Niwaet al., 2000). For generating chimeric embryos, 10 cells of each ES cell line were injectedinto C57BL/6 blastocysts, which were then transplanted into the uteri of pseudopreg-nant ICR females.

Animal care and experiments were approved by the Institutional Ethical Committeeof the Institute of Development, Aging and Cancer, Tohoku University, and the ResearchInstitute of Osaka Medical Center for Maternal and Child Health.

Immunological staining

Immunostaining of whole embryos and cultured explants was performed aspreviously described (Ciruna and Rossant, 2001). Primary antibodies were used at a1:300 dilution for rabbit anti-Oct3/4 (Shimazaki et al., 1993), at a 1:10000 dilution forrabbit anti-PGC7 (Sato et al., 2002), and at 1:1000 dilutions for rabbit anti-mil-2 andmouse anti-Integrin α7 (6A11) (MBL). For rabbit anti-phospho-Smad1/5/8 (CellSignaling TECHNOLOGY, #9511), whole-mount immunostaining was performed at a1:25 dilution of primary antibody as described (Chuva de Sousa Lopes et al., 2003).

Dissection of embryos and explant culture

Dissection of embryos and the explant culture of extra-embryonic mesoderm ofchimeric embryos were performed as described (Okamura et al., 2003). Staging ofembryos has been described previously (Downs and Davies, 1993). Explants of extra-embryonic mesoderm were cultured for 36 h before processing for immunologicalstaining with Stella/PGC7 antibody. Doxycycline and Dex amethasone were added at aconcentration of 1 μg/ml and 100 nM, respectively.

Single cell analysis

GFP-positive or -negative single cells were picked up using a micro-manipulatorfrom dissociated proximal epiblast at pre-streak stage (E6.25) and extra-embryonicmesoderm at the late streak (E7.0) of G3-3 chimeric embryos. cDNA was synthesizedfrom each single cell and then subjected to PCR amplification. As described previously(Dulac, 1998; Saitou et al., 2002), the amplified cDNAs were separated by agarose gelelectrophoresis, and the gels were then stained with ethidium bromide and blottedonto nylon membranes. The membranes were hybridized with the Oct3/4, Fragilis/mil-1 probes, or with the GAPDH probe as an internal control. In some cases, theamplified cDNA was subjected to RT-PCR using the following primer pairs: (Cdx2-5)5′-GTC TGG GCT TTC TTC TCC ACA A-3′, (Cdx2-3) 5′-CCC TTC ACC ATA CAA CTT CTCTAC C-3′, (Integrin α7-5) 5′-GGA GTA AGT TGC TAG CAT ACG TTG-3′, (Integrin α7-3)5′-CAG TGA CAC AGT CTC TTC CCT-3′, (Blimp1-5) 5′-CTT ACT GAC CTG TTG CTG TGACT-3′, (Blimp1-3) 5′-GTT CAT TGT GGG CTC AAC ACT CT-3′, (HPRT-5) 5′-GGA TTT GAAATT CCA GAC AAG-3′ and (HPRT-3) 5′-GCA TTT AAA AGG AAC TGT TGA C-3′. The PCRconditions were 40 cycles of 94 °C for 30 s, 1 cycle of 55 °C for 1 min and 1 cycle of72 °C for 1 min.

Results

To examine the possible roles of Oct3/4 in PGC specification, weattempted to observe the behavior of the PGC precursors, derivedfrom ES cell lines in which the Oct3/4 gene had been geneticallymanipulated, in chimeric embryos. We first used the ES cell line G3-3,a GFP-marked derivative of ZHBTc4 ES cells lacking both alleles of theendogenous Oct3/4 gene (Niwa et al., 2000). Their pluripotency wasmaintained by an exogenously introduced Oct3/4 transgene, the ex-pression of which could be suppressed by tetracycline or its derivative,doxycycline (Dox) via the Tet-Off system (Niwa et al., 2000). Weplaned to examine the possible effect of Dox on exogenous Oct3/4expression and PGC formation in the chimeric embryos at early- tomid-streak stages and in explant culture of the chimeric extra-embryonic mesoderm inwhich PGC precursors differentiate into PGCs(Okamura et al., 2003).

The expression of Oct3/4 protein in G3-3 ES-derived cells and theirbehavior in chimeric embryo

We first examined the behavior of the ES-derived cells in pre-and early- implantation embryos. Before implantation, the G3-3ES-derived GFP-expressing cells intermingled with the recipientinner cell mass, both of which showed comparable Oct3/4 pro-tein expression in blastocyst (Figs. 1A–D). However, we unexpect-edly found that exogenous Oct3/4 protein expression in the ES-derived cells was gradually repressed in most of the G3-3 derivedcells after implantation without Dox (Figs. 1E–P), and this re-pression occurred at the transcriptional level (Fig. 1Q). The sameresults were obtained using another cell line independently derivedfrom ZHBTc4 cells (data not shown). Although the precise mecha-nism of this repression is currently unknown, it is likely thattranscription of the exogenous Oct3/4 transgene in the ZHBTc4-derived cells was constitutively suppressed after differentiation intoepiblast cells due to a possible position effect of the integration siteof the transgene. In addition, most of the G3-3-derived cells thatlost Oct3/4 expression were preferentially localized at the proximalregion of the epiblast after implantation (Figs. 1E–P and Figs.2D–O).

G3-3 ES-derived cells lacking Oct3/4 expression exhibited the propertiesof proximal epiblast cells

As repression of Oct3/4 in ES cells induces loss of pluripotency anddifferentiation to trophectoderm (Niwa et al., 2000), the G3-3-derivedcells most likely differentiated to trophectoderm-derived extra-embryonic ectoderm, and might have been sorted toward extra-embryonic ectoderm in chimeric embryos. To test this possibility,expression of molecules specifically expressed either in proximalepiblast or extra-embryonic ectoderm was examined by immunos-taining (Figs. 2A–O). The results indicated that the G3-3-derived cellsshowed the signals of activated Smad1/5/8 whose activation isnormally induced in proximal epiblast cells by the BMPs signalsfrom the extra-embryonic ectoderm (Figs. 2A–G) (Hayashi et al., 2002;Chuva de Sousa Lopes et al., 2004). They also expressed mil-2 (Figs.2H–K) (Tanaka and Matsui, 2002), specifically expressed in PGCprecursors and PGCs (data not shown). On the other hand, theexpression of Integrin-α7 (Klaffky et al., 2001) and Cdx2 (Beck et al.,1995), specific markers of extra-embryonic ectoderm, was undetect-able in G3-3-derived GFP-expressing cells by immunostaining (Figs.2L–O) and by RT-PCR (Fig. 2P), indicating that these cells were notextra-embryonic ectoderm cells. Taken together, these results indicatethat the G3-3-derived cells lacking Oct3/4 expression and preferen-tially localizing at the proximal region of the epiblast had theproperties of proximal epiblast cells, including PGC precursors, butnot of extra-embryonic ectoderm.

It is possible that even a small amount of Oct3/4 protein below thesensitivity of our immunostaining procedure may still be active inepiblast cells. To examine this possibility, we cultured G3-3 ES cells inthe presence of Dox at different concentration, and found a nicecorrelation between the level of detectable Oct3/4 protein and itsfunctional activity (Fig. S1).

G3-3 ES-derived cells lacking Oct3/4 expression contributed to PGCprecursor-like cells, but failed to form PGCs

As indicated above, the G3-3-derived cells without detectableOct3/4 expression were able to contribute to proximal epiblast, andsome expressed a marker of PGC precursors (Fig. 2). We furtherobserved the developmental fates of the G3-3-derived cells in earlygastrulating chimeric embryos. Most of G3-3-derived cells movedinto extra-embryonic mesoderm, and some of them at the tip of theextra-embryonic mesoderm formed a cluster of cells together with

Fig. 1. The Oct3/4 expression in G3-3-derived cells was transcriptionally repressed after implantation, which was accompanied by localization of these cells at the proximal regionof the epiblast in chimeric embryos. (A–D) After culturing for 18 h, G3-3 ES cells (yellow arrowheads) injected into E3.5 blastocyst were intermingled with the recipient inner cellmass (white arrowheads), and both of these cells showed comparable Oct3/4 protein expression. (E–H) After implantation, G3-3-derived GFP-expressing cells (yellowarrowheads) showed weaker expression of Oct3/4 protein than recipient epiblast cells at E5.0, and its expression was undetectable in some cells (white arrowhead). (I–L) At E5.5,the ES-derived cells (yellow arrowheads) lost Oct3/4 protein and localized at the proximal region of the epiblast. (M–P) The ES-derived cells were still preferentially localized atthe most proximal region of the epiblast at the pre-gastrulation stages, E6.25. Among these cells, a small number of the G3-3-derived cells incidentally escaped the repression ofOct3/4 expression (* in panels M–P). TOTO-3 staining of nuclei (C, G, K, O) and merged images (D, H, L, P). Scale bar, 50 μm (A–P). (Q) Single-cell cDNA analysis by cDNA southernhybridization revealed that Oct3/4 transcript was undetectable or at quite low levels in G3-3-derived cells at the proximal region of the epiblast in chimeric embryos, while itsexpression was evident in the surrounding GFP-negative recipient epiblast cells.

578 D. Okamura et al. / Developmental Biology 317 (2008) 576–584

recipient cells, and they expressed mil-2 at E7.0 (Figs. 3A–D, arrow-heads). The clustered ES-derived cells also expressed Fragilis/mil-1(Fig. 3M) (Saitou et al., 2002, Tanaka and Matsui, 2002) and Blimp1(Fig. 3N) (Ohinata et al., 2005) at high frequency, indicating that theywere PGC precursors-like cells, even though Oct3/4 protein was stilldown-regulated in these cells (Fig. 3M). We next examined whetherthe clustered G3-3-derived PGC precursors-like cells could bespecified to PGCs at a subsequent stage. We stained the chimericembryos at E7.25–7.5 with an antibody to Stella/PGC7 to identify

nascent PGCs (Sato et al., 2002; Saitou et al., 2002). Despite theefficient contribution to PGC precursor-like cells, most of the ES-derived cells did not express Stella/PGC7 (Figs. 3I–L, blank arrow-heads), and the expression of Oct3/4 protein was still suppressed(Figs. 3E–H, blank arrowheads). In contrast, the recipient embryo-derived cells expressed Stella/PGC7 (Figs. 3I–L). Taken together, theseresults suggest that the G3-3-derived cells formed PGC precursors-like cells without detectable Oct3/4 expression, but failed to diffe-rentiate into PGCs.

Fig. 2. The G3-3 ES-derived cells exhibited the properties of proximal epiblast cells, but not of extra-embryonic ectoderm. (A–C) In normal embryos, the expression of phosphorylatedand activated Smad proteins at the proximal region of the epiblast was much higher than that at the distal region at E6.25, but its signal was not detected in extra-embryonicectoderm. (D–O) In the chimeric embryo, the G3-3 ES-derived cells in the proximal region of the epiblast showed signals for phosphorylated Smad proteins (D–G, yellow arrowheads)and mil-2 proteins (H–K), but they did not express Integrin-α7, a marker of extra-embryonic ectoderm (L–O). White blank arrowheads indicate the border between epiblast andextra-embryonic ectoderm (A–O). TOTO-3 staining of nuclei (B, F, J, N) andmerged images (C, G, K, O). Scale bar, 50 μm (A–G) and 25 μm (H–O). (P) Single-cell cDNA analysis by RT-PCRrevealed that transcripts for extra-embryonic ectodermmarkers, Cdx2 and Integrin-α7, were undetectable in proximal epiblast cells including the GFP-positive G3-3 ES-derived cellsand the GFP-negative recipient cells.

579D. Okamura et al. / Developmental Biology 317 (2008) 576–584

Rescued Oct3/4 expression in ES-derived cells restored the ability ofspecification to PGCs

To clarify the impact of loss of Oct3/4 expression in PGCspecification, we next attempted to rescue the Oct3/4 expression inchimeric embryos using GOWT1 ES cells, which are also a derivative ofthe ZHBTc4 cells but have an additional Oct3/4-GFP fusion transgeneunder the control of the CAG promoter. The GOWT1 ES-derived cellswere uniformly distributed within the epiblast (data not shown) andextra-embryonic mesoderm (Figs. 4A, E). In addition, most of the GFP-expressing GOWT1-derived cells showed comparable expression of

Oct3/4 protein to that in surrounding recipient cells (Figs. 4A–D), and aportion of GOWT1-derived cells within the extra-embryonic meso-derm successfully expressed Stella/PGC7 (Figs. 4E–H). These resultssuggest that the rescued Oct3/4 expression in the ES-derived cellsrestored the ability of specification to PGCs.

Suppression of Oct3/4 expression caused failure of PGC specification incultured extra-embryonic mesoderm

To further determine the requirement of Oct3/4 protein for PGCspecification, fragments of extra-embryonic mesoderm containing

Fig. 3. The G3-3 ES-derived cells continuously repressed for Oct3/4 expression were able to contribute to PGC precursor-like cells, but failed to form PGCs. (A–D) A cluster of PGCprecursors expressing mil-2, a specific marker of PGC precursors, was composed of G3-3 ES-derived GFP-expressing cells and GFP negative recipient cells in extra-embryonicmesoderm at E7.0 (arrowheads). (E–L) At E7.25, clustered GFP-expressing G3-3-derived cells, in which Oct3/4 protein was constitutively down-regulated (E-H, blank arrowheads),failed to be PGC7/Stella-positive PGCs (I–L, blank arrowheads). TOTO-3 staining of nuclei (C, G, K) and merged images (D, H, L). Scale bar, 50 μm. (M, N) Amplified single-cell cDNAobtained from extra-embryonic mesoderm in the G3-3-chimeric embryos at E7.0 were subjected to Southern blot analysis (M) and RT-PCR (N). Most of the clustered G3-3-derivedGFP-expressing cells showed transcriptional suppression of Oct3/4 (M), and expression of Fragilis/mil-1 (M) and Blimp1 (N), specific markers of PGC precursors.

580 D. Okamura et al. / Developmental Biology 317 (2008) 576–584

PGC precursors from the E6.75 chimeric embryos were cultured(Okamura et al., 2003) (Fig. 5). The extent of the contribution of EScells to extra-embryonic mesoderm varies among chimeric embryos,which may affect the efficiency of PGC specification from the ES-derived cells. Thus for culture we selected chimeric embryos showingover 50% chimerism in extra-embryonic mesoderm, which wasestimated by the area of GFP expression. After culture, the explantswere stained with the antibody for Stella/PGC7 and the number ofStella/PGC7-positive PGCs was counted. In cases where the explantsfrom the chimeric embryos of the ES cells were heterozygous for the

endogenous Oct3/4 alleles (G4-2) and of GOWT1 cells, over 60% ofStella/PGC7-positive cells co-expressed GFP (Figs. 5A–H and M, Table1), while only 12.6% of Stella/PGC7-expressing cells were GFP positivein explants from the G3-3 chimeric embryos (Figs. 5I–M, Table 1).Most of the recipient-derived Stella/PGC7-expressing PGCs andStella/PGC7-negative G3-3-derived cells were located within thesame cluster of cells (Figs. 5I–L), further supporting the idea that G3-3-derived cells preferentially contribute to a cluster of PGCprecursors, but fail to be specified into PGCs. Although the expressionof the Oct3/4 protein was mostly suppressed in the G3-3-derived

Fig. 4. Rescued Oct3/4 expression in the ES-derived cells restored the ability of specification to PGCs. (A–D) GOWT1-derived GFP-expressing cells, in which expression of Oct3/4 wasrescued by an additional Oct3/4 transgene, showed Oct3/4 protein expression at a comparable level to that of surrounding GFP-negative recipient cells in the extra-embryonicmesoderm at E7.25 (arrowheads). (E–H) PGC7/Stella-positive PGCs emerged from GOWT1-derived cells (arrowheads). TOTO-3 staining of nuclei (C, G) and merged images (D, H).Scale bar, 50 μm (A–H).

581D. Okamura et al. / Developmental Biology 317 (2008) 576–584

cells, its suppression may be leaky (* in Figs. 1M–P and 3E–H), whichmay explain why a few PGCs derived from G3-3 ES cells were found(Fig. 5M, Table 1). To examine this possibility, the explants of G3-3chimeric embryos were cultured in the presence of doxycycline (Dox)to completely suppress the Oct3/4 expression in the ES-derived cells.As expected, Dox markedly reduced the number of G3-3-derivedPGCs in the explants, while it had no effect on PGC emergence fromrecipient cells (Fig. 5M, Table 1 and data not shown). As wepreviously reported, cells showing Stella/PGC7 expression in thecultured explants were also positive for TNAP (tissue non-specific

Fig. 5. Suppression of Oct3/4 expression caused failure of PGC specification in cultured exchimeric embryos containing PGC precursors were cultured on a feeder layer of STO cells. Aftembryos of G4-2 ES cells, which were heterozygous for the endogenous Oct3/4 alleles (A–D)(arrowheads, * shows recipient-derived PGCs). (I–L) In explants from G3-3 chimeric embryosextra-embryonic mesodermwere stained using anti-Stella/PGC7 (B, F, J). TOTO-3 staining of nemergence from the ES-derived cells in cultured explants of extra-embryonic mesoderm. Thplotted with a red circle, and medians are depicted by a black stripe. Doxycycline (Dox) wa

alkaline phosphatase) activity or its gene expression, which is thealternative specific marker of PGCs (Okamura et al., 2003, 2005;Chuva de Sousa Lopes et al., 2004), which was also confirmed in thisstudy (data not shown).

Conditional Oct3/4 expression in ES-derived PGC precursors affected PGCspecification

We further confirmed the requirement of Oct3/4 for PGC spe-cification using another ZHBTc4-derived ES cell line, 4EOGR1, inwhich

tra-embryonic mesoderm. (A–L) Fragments of extra-embryonic mesoderm from E6.75er culture, explants were stained with anti-PGC7/Stella. (A–H) In explants from chimeric, and of GOWT1 ES cells (E–H), most of the Stella/PGC7-positive cells co-expressed GFP, few Stella/PGC7-expressing cells were GFP positive (arrowheads). Cultured explants ofuclei (C, G, K) and merged images (D, H, L). Scale bar, 50 μm (A–L). (M) Efficiency of PGCe ratio of the ES-derived cells in all of the PGC7/Stella-positive cells in each explant wass added to the culture medium at the indicated concentration.

Table 1The restoration of Oct3/4 protein and its activity in PGC precursors is necessary for PGCspecification

Type ofexplants

Doxycycline(1 μg/ml)

Dexamethasone(100 nM)

ES-derived cells/Stella/PGC7-positive cells

(%)a

G4-2 − − 123/199 61.8±22.6 n=7(explants)

GOWT1 − − 90/122 73.8±22 n=11G3-3 − − 27/213 12.6±8.7 n=8G3-3 + − 1/137 0.7±1 n=64EOGR1 + − 4/48 8.8±9 n=44EOGR1 + + 43/60 71.6±15.6 n=4

a Include standard deviations (±).

582 D. Okamura et al. / Developmental Biology 317 (2008) 576–584

a dexamethasone (Dex)-dependent GFP-Oct3/4-GR fusion trans-gene was introduced into ZHBTc4 cells. In culture of 4EOGR1 EScells, addition of doxycycline (Dox) completely suppressed Oct3/4

Fig. 6. Conditional Oct3/4 expression in 4EOGR1 ES-derived PGC precursors affected PGC speresponse to dexamethasone (Dex). One hour after addition of Dex (100 μM), most signals of th later, all signals had translocated to the nucleus (C, D). (E–L) Fragments of extra-embryopresence of doxycycline (Dox) (E–H), or of both Dox and Dex (I–L). Cultured explants werepositive cells were derived from the GFP-negative recipient cells (arrowheads). (I–L) In thecontributed to PGC7/Stella-positive cells (arrowheads). Cultured explants of extra-embryonandmerged images (H, L). Scale bar, 25 μm (A–D) and 50 μm (E–L). (M) Efficiency of PGC emerThe ratio of the ES-derived cells in all of the PGC7/Stella-positive cells in each explant was padded to the culture medium at the indicated concentration.

activity and resulted in differentiation of the cells, as in the parentZHBTc4 cells, while addition of Dex maintained their pluripotencyeven in the presence of Dox (Fig. S2). By administration of Dex, theGFP-Oct3/4-GR fusion protein in 4EOGR1 ES cells changes theirconformation and translocated from cytoplasm into nucleus within3 h (Figs. 6A–D and data not shown). Using the ES cells, weexpected that the Oct3/4 activity in the chimeric embryos could beconditionally manipulated by Dex. Fragments of extra-embryonicmesoderm of the 4EOGR1-chimeric embryos at E6.75 werecultured with Dox, or Dox and Dex. In the presence of Dox only,most of the ES-derived GFP-positive cells failed to express PGC7/Stella, as was the case for G3-3 cells (Figs. 6E–H and M, Table 1),while in the presence of both Dox and Dex, the ES-derived cellswere able to contribute to PGC7/Stella-positive cells (Figs. 6I–L andM, Table 1). Localization of the GFP-Oct3/4-GR fusion proteindetected by GFP fluorescence indicated Dex-dependent transloca-tion of the fusion protein into the nucleus (Figs. 6A–D and I–L),confirming the conditional activation of the Oct3/4 function in theexplants. On the whole, these results indicate that activity of Oct3/

cification. (A–D) The GFP-Oct3/4-GR fusion proteins in 4EOGR1 ES cells showed a quickhe GFP-Oct3/4-GR fusion proteinwere still dispersed in the cytoplasm (A, B). However, 6nic mesoderm from E6.75 chimeric embryos of 4EOGR1 ES cells were cultured in thestained with anti-PGC7/Stella. (E–H) In the presence of Dox, most of the PGC7/Stella-presence of Dex as well as Dox, the GFP-expressing 4EOGR1 ES-derived cells efficientlyic mesoderm were stained using anti-Stella/PGC7 (F, J). TOTO-3 staining of nuclei (G, K)gence from 4EOGR1 ES-derived cells in cultured explants of extra-embryonicmesoderm.lotted with a red circle, and medians are depicted by a black stripe. Dox and Dex were

Fig. 7. Oct3/4 plays critical roles in PGC specification. Oct3/4 expression in the clustered PGC precursors is essential for the final specification of PGCs. Oct3/4 may regulate expressionof genes conferring the properties of the nascent PGCs on the PGC precursors. After the PGC specification, expression of Oct3/4 protein is maintained only in PGC7/Stella-expressingnascent PGCs, while in surrounding somatic cells, it is rapidly down-regulated and finally disappears.

583D. Okamura et al. / Developmental Biology 317 (2008) 576–584

4 protein in PGC precursors is necessary for the precursor to bespecified to PGC.

Discussion

A recent study concerning functions of the transcriptional re-pressor Blimp1 in PGC development revealed that this molecule in thePGC precursors may repress the expression of genes involved insomatic cell differentiation and induce the precursors to PGCs(Vincent et al., 2005; Ohinata et al., 2005). However, independenttranscriptional regulation, including induction of genes required formaintenance and further specification to PGCs, in their precursorsmay also play essential roles. In the present study, we revealed thatOct3/4 activity is essential for PGC specification, but is dispensable forformation and maintenance of their precursors.

We initially designed experiments in which Oct3/4 transcriptionwas repressed in the presence of Dox using G3-3 ES cells, but wesubsequently found that the expression of Oct3/4 in the ES-derivedcells was gradually repressed in the chimeric embryos afterimplantation in a Dox-independent manner, probably due to integra-tion site-dependent repression of the Oct3/4 transgene after differ-entiation of ES cells (Figs. 1 and 3). On the other hand, the additionaltransgene in GOWT1 ES cells is likely to be integrated in a differentlocation in the genome, and therefore does not suffer the site-dependent repression (Fig. 4).

Even in the absence of Oct3/4 expression after implantation, wefound that the G3-3-derived cells successfully contributed to proximalepiblast cells and PGC precursor-like cells in the chimeric embryos(Figs. 1–3). None of these cells expressed the markers of extra-embryonic ectoderm, and some expressed activated Smads and mil-2protein, which are normally expressed in proximal epiblast as well asin PGC precursors (Fig. 2). Because Oct3/4 is expressed in all epiblastcells including the precursors of PGCs, the significance of precursorformation independent of Oct3/4 is currently unclear. However, arecent report indicated that the expression of Sox2, a heterodimericpartner of Oct3/4, is up-regulated at the time of PGC specification(Yabuta et al., 2006), suggesting that Oct3/4 becomes functional in thePGC precursors after expression of cooperative factors such as Sox2.

The present study also revealed that the Oct3/4 function isessential for the precursors to be specified to PGCs. We initially usedtwo different ES cell lines, G3-3 and GOWT1, both of which werederived from ZHBTc4 ES cells (Niwa et al., 2000). These ES-derivedcells showed differential expression of Oct3/4 in chimeric embryos,which most likely affected the efficiency of PGC formation from ES-

derived cells (Figs. 3–5, Table 1). However, it is also possible that theseES cell lines may have additional differences generated duringestablishment from ZHBTc4, which might affect their ability to formPGCs. To exclude this possibility, we used another ZHBTc4-derived EScell line, 4EOGR1, in which the Oct3/4 function could be conditionallyinduced by Dex, and the results further indicated that the Oct3/4activity was critical for PGC specification (Fig. 6, Table 1).

We previously found that interaction among the PGC precursorsmediated by a cell adhesion molecule, E-cadherin, plays a role in thefinal step of PGC specification (Okamura et al., 2003). E-cadherin maydirectly or indirectly transmit essential intracellular signals into thePGC precursors, which may activate transcription of essential genesrequired for final events of PGC specification, and Oct3/4 may colla-borate with the putative E-cadherin-dependent signaling (MatsuiandOkamura, 2005). A previous studyalso indicated the importance ofBlimp1 in the PGC precursors, and this transcription factor affectsgenes required for somatic cell differentiation (Vincent et al., 2005;Ohinata et al., 2005). In addition to the functions of Blimp1, it is mostlikely that distinct transcriptional cascades activate a group of genesthat are essential for the initial steps of PGC differentiation, and Oct3/4may play a key role in this process (Fig. 7). Identification of possibletarget genes of Oct3/4 in nascent PGCs in future studies is a desirableapproach to clarify the mechanisms of PGC specification.

Oct3/4 is necessary for survival of migrating PGCs in later stageembryos (Kehler et al., 2004). It is also expressed in spermatogonialstem cells as well as in growing and mature oocytes (Pesce et al.,1998). Although its functions in these cells are unknown, it is likelythat Oct3/4 may also play important roles in the maintenance ofspermatogonial stem cells and in the establishment of pluripotency inoocytes, and therefore investigating the possible functions of Oct3/4 inlater germ cells is an attractive subject for future studies.

Acknowledgments

We wish to thank Dr. H. Hamada for kindly providing the anti-Oct3/4 antibody and Dr. T. Nakano for kindly providing the anti-PGC7antibody. This work was supported in part by CREST of the JapanScience and Technology Agency (JST), and by grants-in-aid from theMinistry of Education, Science, Sports and Culture of Japan.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.ydbio.2008.03.002.

584 D. Okamura et al. / Developmental Biology 317 (2008) 576–584

References

Beck, F., Erler, T., Russell, A., James, R., 1995. Expression of Cdx-2 in the mouse embryoand placenta: possible role in patterning of the extra-embryonic membranes. Dev.Dyn. 204, 219–227.

Chang, H., Matzuk, M.M., 2001. Smad5 is required for mouse primordial germ celldevelopment. Mech. Dev. 104, 61–67.

Chu, G.C., Dunn, N.R., Anderson, D.C., Oxburgh, L., Robertson, E.J., 2004. Differentialrequirements for Smad4 in TGFbeta-dependent patterning of the early mouseembryo. Development 131, 3501–3512.

Chuva de Sousa Lopes, S.M., Carvalho, R.L., van den Driesche, S., Goumans, M.J., ten Dijke,P., Mummery, C.L., 2003. Distribution of phosphorylated Smad2 identifies targettissues of TGF beta ligands in mouse development. Gene Expr. Patterns 3, 355–360.

Chuva de Sousa Lopes, S.M., Roelen, B.A., Monteiro, R.M., Emmens, R., Lin, H.Y., Li, E.,Lawson, K.A., Mummery, C.L., 2004. BMP signaling mediated by ALK2 in the visceralendoderm is necessary for the generation of primordial germ cells in the mouseembryo. Genes Dev. 18, 1838–1849.

Ciruna, B., Rossant, J., 2001. FGF signaling regulates mesoderm cell fate specification andmorphogenetic movement at the primitive streak. Dev. Cell 1, 37–49.

Downs, K.M., Davies, T., 1993. Staging of gastrulating mouse embryos by morphologicallandmarks in the dissecting microscope. Development 118, 1255–1266.

Dulac, C., 1998. Cloning of genes from single neurons. Curr. Top. Dev. Biol. 36, 245–258.Hayashi, K., Kobayashi, T., Umino, T., Goitsuka, R., Matsui, Y., Kitamura, D., 2002. SMAD1

signaling is critical for initial commitment of germ cell lineage frommouse epiblast.Mech. Dev. 118, 99–109.

Hayashi, K., de Sousa Lopes, S.M., Surani, M.A., 2007. Germ cell specification in mice.Science 316, 394–396.

Kehler, J., Tolkunova, E., Koschorz, B., Pesce, M., Gentile, L., Boiani, M., Lomeli, H., Nagy,A., McLaughlin, K.J., Scholer, H.R., Tomilin, A., 2004. Oct4 is required for primordialgerm cell survival. EMBO Rep. 5, 1078–1083.

Klaffky, E., Williams, R., Yao, C.C., Ziober, B., Kramer, R., Sutherland, A., 2001.Trophoblast-specific expression and function of the integrin alpha 7 subunit inthe peri-implantation mouse embryo. Dev. Biol. 239, 161–175.

Lawson, K.A., Hage, W.J., 1994. Clonal analysis of the origin of primordial germ cells inthe mouse. Ciba Found. Symp. 182, 68–84 (discussion 84-91).

Lawson, K.A., Dunn, N.R., Roelen, B.A., Zeinstra, L.M., Davis, A.M., Wright, C.V., Korving,J.P., Hogan, B.L., 1999. Bmp4 is required for the generation of primordial germ cellsin the mouse embryo. Genes Dev. 13, 424–436.

Matsui, Y., Okamura, D., 2005. Mechanisms of germ-cell specification in mouseembryos. Bioessays 27, 136–143.

McLaren, A., 2003. Primordial germ cells in the mouse. Dev. Biol. 262, 1–15.McLaren, A., Lawson, K.A., 2005. How is the mouse germ-cell lineage established?

Differentiation 73, 435–437.Nichols, J., Zevnik, B., Anastassiadis, K., Niwa, H., Klewe-Nebenius, D., Chambers, I.,

Scholer, H., Smith, A., 1998. Formation of pluripotent stem cells in the mammalianembryo depends on the POU transcription factor Oct4. Cell 95, 379–391.

Niwa, H., Miyazaki, J., Smith, A.G., 2000. Quantitative expression of Oct-3/4 definesdifferentiation, dedifferentiation or self-renewal of ES cells. Nat. Genet. 24,372–376.

Niwa, H., Masui, S., Chambers, I., Smith, A.G., Miyazaki, J., 2002. Phenotypiccomplementation establishes requirements for specific POU domain and generic

transactivation function of Oct-3/4 in embryonic stem cells. Mol. Cell. Biol. 22,1526–1536.

Ohinata, Y., Payer, B., O'Carroll, D., Ancelin, K., Ono, Y., Sano, M., Barton, S.C.,Obukhanych, T., Nussenzweig, M., Tarakhovsky, A., Saitou, M., Surani, M.A., 2005.Blimp1 is a critical determinant of the germ cell lineage in mice. Nature 436,207–213.

Okamoto, K., Okazawa, H., Okuda, A., Sakai, M., Muramatsu, M., Hamada, H., 1990. Anovel octamer binding transcription factor is differentially expressed in mouseembryonic cells. Cell 60, 461–472.

Okamura, D., Kimura, T., Nakano, T., Matsui, Y., 2003. Cadherin-mediated cell interactionregulates germ cell determination in mice. Development 130, 6423–6430.

Okamura, D., Hayashi, K., Matsui, Y., 2005. Mouse epiblasts change responsiveness toBMP4 signal required for PGC formation through functions of extraembryonicectoderm. Mol. Reprod. Dev. 70, 20–29.

Pesce, M., Scholer, H.R., 2000. Oct-4: control of totipotency and germline determination.Mol. Reprod. Dev. 55, 452–457.

Pesce, M., Gross, M.K., Scholer, H.R., 1998. In line with our ancestors: Oct-4 and themammalian germ. Bioessays 20, 722–732.

Rosner, M.H., Vigano, M.A., Ozato, K., Timmons, P.M., Poirier, F., Rigby, P.W., Staudt, L.M.,1990. A POU-domain transcription factor in early stem cells and germ cells of themammalian embryo. Nature 345, 686–692.

Saitou, M., Barton, S.C., Surani, M.A., 2002. A molecular programme for the specificationof germ cell fate in mice. Nature 418, 293–300.

Sato, M., Kimura, T., Kurokawa, K., Fujita, Y., Abe, K., Masuhara, M., Yasunaga, T., Ryo, A.,Yamamoto, M., Nakano, T., 2002. Identification of PGC7, a new gene expressedspecifically in preimplantation embryos and germ cells. Mech. Dev. 113, 91–94.

Scholer, H.R., Ruppert, S., Suzuki, N., Chowdhury, K., Gruss, P., 1990. New type of POUdomain in germ line-specific protein Oct-4. Nature 344, 435–439.

Shimazaki, T., Okazawa, H., Fujii, H., Ikeda, M., Tamai, K., McKay, R.D., Muramatsu, M.,Hamada, H., 1993. Hybrid cell extinction and re-expression of Oct-3 functioncorrelates with differentiation potential. EMBO J. 12, 4489–4498.

Tanaka, S.S., Matsui, Y., 2002. Developmentally regulated expression of mil-1 and mil-2,mouse interferon-induced transmembrane protein like genes, during formationand differentiation of primordial germ cells. Mech. Dev. 119 (Suppl 1), S261–S267.

Tremblay, K.D., Dunn, N.R., Robertson, E.J., 2001. Mouse embryos lacking Smad1 signalsdisplay defects in extra-embryonic tissues and germ cell formation. Development128, 3609–3621.

Vincent, S.D., Dunn, N.R., Sciammas, R., Shapiro-Shalef, M., Davis, M.M., Calame, K.,Bikoff, E.K., Robertson, E.J., 2005. The zinc finger transcriptional repressor Blimp1/Prdm1 is dispensable for early axis formation but is required for specification ofprimordial germ cells in the mouse. Development 132, 1315–1325.

Yabuta, Y., Kurimoto, K., Ohinata, Y., Seki, Y., Saitou, M., 2006. Gene expression dynamicsduring germline specification in mice identified by quantitative single-cell geneexpression profiling. Biol. Reprod. 75, 705–716.

Ying, Y., Liu, X.M., Marble, A., Lawson, K.A., Zhao, G.Q., 2000. Requirement of Bmp8b forthegeneration of primordial germcells in themouse.Mol. Endocrinol.14,1053–1063.

Ying, Y., Qi, X., Zhao, G.Q., 2001. Induction of primordial germ cells from murineepiblasts by synergistic action of BMP4 and BMP8B signaling pathways. Proc. Natl.Acad. Sci. U. S. A. 98, 7858–7862.

Yoshimizu, T., Obinata, M., Matsui, Y., 2001. Stage-specific tissue and cell interactionsplay key roles in mouse germ cell specification. Development 128, 481–490.