Parthenogenesis is a form of asexual reproduction where the offspring is derived entirely from an unfertilized female gamete. It is a normal process used by some reptiles and birds to reproduce. In mammals, parthenogenesis usually refers to embryonic development from an artificially activated oocyte without fertilization by a sperm. While using normal human embryos to derive ES cells is ethically disputable, using parthenogenetic embryos (called “parthenotes”), which are incapable of developing into full organisms, is less disputable (Sturm et al.,1994; Kono et al.,1996; Mognetti and Sakkas,1996). In fact, many recent studies have demonstrated that parthenogenesis is an efficient way to generate histocompatible human embryonic stem (ES) cells for transplantation-based stem cell therapies (Revazova et al.,2008; Hao et al.,2009; Lu et al.,2010). Histocompatible parthenote ES cells represent an important milestone in stem cell therapies, as they allow partial MHC matching to a substantial population of unrelated transplant recipients (Cheng,2008; Revazova et al.,2008).
In mammalian embryogenesis, the first lineage segregation event is the segregation between the trophectoderm and inner cell mass (ICM) lineages, which occurs between the morula and blastocyst stages, and the second lineage segregation event is the segregation between the primitive endoderm and epiblast lineages within the ICM (Reik et al.,2003; Morgan et al.,2005; Boyer et al.,2006; Dietrich and Hiiragi,2007). A previous study demonstrated that the ICM cultures from parthenogenetic embryos contained mostly differentiated parietal endoderm cells and only a few undifferentiated stem cells, suggesting lineage segregation defects in parthenotes (Newman-Smith and Werb,1995). In order to distinguish between the negative effects of in vitro cultures and that of parthenogenesis per se, our study compared gene expression in parthenotes with that in fertilized embryos that developed in vitro instead of in a maternal uterus.
In this study, we analyzed expression of many development-related genes in parthenote morulae and blastocysts, including the trophectoderm markers Cdx2, Elf5, and Tbr2(Eomes), the ICM marker Oct4, the epiblast markers Sox2 and Nanog, the primitive endoderm marker Gata4, and the signal transduction factors Fgf3, Fgf4, and their receptor Fgfr2. We found normal levels of Cdx2, Oct4, and Fgf4, but perturbed levels of Elf5, Tbr2, Sox2, Nanog, Gata4, and Fgf3, as well as increased Fgfr2 phosphorylation in parthenote embryos. Our study also reveals greatly increased Gata4 and Fgf3 but reduced Tbr2, Sox2, and Nanog expression in parthenotes compared to controls, indicating cell fate defects and impaired differentiation capacity of parthenote blastomeres. Hence we suggest that the increased Fgf3-Fgfr2 signaling, the downstream elevation of Gata4 expression and concomitant suppression of Nanog expression in parthenogenetic preimplantation embryos constitute the major molecular mechanism leading to lineage segregation defects of parthenotes.
Reduction of Cell Numbers at the Early and Late Blastocyst Stages
To exclude the possibility that the in vitro cultures instead of parthenogenesis per se cause delayed development of parthenotes, we used embryos developed from in vitro cultured fertilized oocytes as controls, instead of those directly from a maternal uterus. As shown in Figure S1A (which is available online), preimplantation embryos developed separately from a total of 78 parthenogenetic and fertilized oocytes were analyzed on consecutive days after oocyte collection (which was counted as E0.5). In order to compare the efficiencies of embryonic development between controls and parthenotes during in vitro culturing, we assessed the percentages of embryos reaching each developmental stage every day. We counted total cell numbers in parthenote and control embryos at both the morula and blastocyst stages. Morulae obtained at E3.0 usually had 12–20 cells and were grouped as “early morulae,” whereas morulae obtained at E3.5 usually had 25–33 cells and were grouped as “late morulae.” Blastocysts obtained at E3.5, which still expressed Oct4 and Nanog in the trophectoderm, were grouped as “early blastocysts” (Supp. Fig. S2 and data not shown). On the other hand, blastocysts obtained at E4.5, which have restricted Oct4 and Nanog expression in the ICM, were grouped as “late blastocysts” (Supp. Fig. S2).
As shown in Table S1, percentages of embryos reaching different preimplantation stages from E1.5 (the cleavage stage) to E5.5 (the hatched stage) were similar between controls and parthenotes. Representative charts of statistical analyses at E3.5, E4.5, and E5.5 are shown in Supp. Figure S1C. Taken together, these results indicate that the developmental timing of parthenote mouse embryos appears normal from E1.5 to E5.5. We also analyzed the total cell numbers in each embryo by enumerating the DAPI-stained nuclei on 3D projections constructed from serial optical sections of embryos. It was found that total numbers of cells in parthenote embryos were significantly decreased at the blastocyst stage (Fig. 1).
Normal Expression of Cdx2 and Reduced Elf5+ and Tbr2+ Cells in the Trophectoderm of Parthenotes
It has been reported that Cdx2, Elf5, and Tbr2(Eomes) are three of the earliest transcription factors initiating trophectoderm differentiation (Russ et al.,2000; Donnison et al.,2005; Niwa et al.,2005; Strumpf et al.,2005; Hemberger and Dean,2007; Jedrusik et al.,2008; Ng et al.,2008; Ralston and Rossant,2008), and Elf5 directly activates Cdx2 and Tbr2 expression (Choi and Sinha,2006; Ng et al.,2008). Hence we assessed the levels of Cdx2, Elf5, and Tbr2 in parthenote and control embryos by counting the nuclei stained with specific markers on 3D projections of each embryo (Fig. 2). It was found that normal Cdx2 expression in cells located on the periphery of the morulae (Fig. 2A) and in the trophectoderm of blastocysts (Fig. 1B, C). On the other hand, we failed to detect any specific Elf5 and Tbr2 signals at the morula stage in either controls or parthenotes (Fig. 2A, D). The lack of Elf5 and Tbr2 expression in normal morulae was also described previously (Kwon and Hadjantonakis,2007; Ng et al.,2008; Ralston and Rossant,2008; Hemberger et al.,2009).
The distribution of specific markers in whole embryos is shown in Supp. Table S2. With similar numbers of Cdx2+ cells in controls and parthenotes (Fig. 2A–C), percentages of Cdx2+ cells in whole embryos were also comparable between controls and parthenotes (Fig. 2G and Table S2). On the other hand, both Elf5+ and Tbr2+ cells were dramatically reduced in parthenotes compared to controls at the blastocyst stage (Fig. 2B, C, E, F, and Table S3). The percentages of Elf5+ cells are higher than those of Tbr2+ cells, which were lower than 50% of those in the control trophectoderm, both at the early and late blastocyst stages (Fig. 2B, C, E, F, and Table S2). We found that, while 90% of Elf5+ cells are also Tbr2-positive, the remaining 10% of Elf5+ cells do not express Tbr2. Taken together, our results indicate a normal percentage of Cdx2+ cells but reduced percentages of Elf5+ and Tbr2+ cells in the parthenote trophectoderm.
Decrease of Nanog+ and Sox2+ cells in Parthenotes
Since there was a great decrease in total cell numbers of parthenote blastocysts (Fig. 1), we were interested in assessing the numbers of ICM cells in parthenote blastocysts. We performed double immunostaining for Cdx2 and Oct4, and found normal levels of Oct4 expression but significantly decreased numbers of ICM cells in parthenotes compared to controls at both the early and late blastocyst stages (Supp. Fig. S2C). This reduction was proportional to the decrease in total cell numbers in parthenote blastocysts (Fig. 1). On the other hand, the percentages of ICM cells in parthenotes were comparable with those in controls (Supp. Fig. S2D), concurrent with the comparable percentages of trophectoderm cells (counted as Cdx2+ cells) between parthenotes and controls (Fig. 2G and Table S2). The percentages of Oct4+ cells in whole embryos were comparable between controls and parthenotes at both the morula and blastocyst stages (Supp. Fig. S2, Fig. 3, and Supp. Table S2). At the late blastocyst stage, when Oct4 expression was completely restricted to the ICM, Oct4+ cells constituted 100% of ICM cells in both control and parthenote embryos (n = 20) (Supp. Figs. S2B and S3D).
We also examined the expression of two other pluripotency genes, Sox2 and Nanog, in normal fertilized and parthenote embryos. Sox2 is expressed in the epiblast at the blastocyst stage and is required for peri-implantation lineage specification (Hayashi et al.,2002; Avilion et al.,2003; Boyer et al.,2006). Compared to control embryos, we found dramatically decreased numbers of Sox2+ cells in parthenote embryos at both the morula and blastocyst stages (Fig. 3). Percentages of Sox2+ cells in whole embryos are given in Table S2. As Sox2+ cells constitute the epiblast at the blastocyst stage (Hayashi et al.,2002; Avilion et al.,2003), our results demonstrate greatly diminished epiblast populations in parthenote ICMs.
Our analyses of the expression of another pluripotency gene and epiblast marker, Nanog (Chambers et al.,2003; Mitsui et al.,2003), further confirmed our results of Sox2 immunostaining. We observed colocalization of Nanog+ and Sox2+ cells in both control and parthenote embryos at both the morula and blastocyst stages (Supp. Fig. S3), indicating overlapping expression of Nanog and Sox2 in the same cell lineage during preimplantation development. Similar to Sox2+ cells, the numbers and percentages of Nanog+ cells in parthenotes were significantly lower than those in controls (Fig. 4 and Supp. Table S2).
Increase of Gata4+ Cells in Parthenotes
In addition to the epiblast, the other cell lineage in the ICM is the primitive endoderm (Rossant et al.,2003; Ralston and Rossant,2005; Chazaud et al.,2006; Yamanaka et al.,2006). We then assessed expression of the primitive endoderm marker, Gata4, in parthenotes. Our immunostaining results showed only scattered Gata4+ cells in control morulae at E3.0, whereas almost all cells in parthenote morulae expressed Gata4 (Fig. 4A, B, and Table S2). Between E4.0 and E4.5, we found a dramatically expanded Gata4 expression domain in parthenote blastocysts, covering not only the majority of ICM cells but also most trophectoderm cells (Fig. 4C). Therefore, almost 90% of total cells in parthenote blastocysts expressed Gata4, in contrast to restricted Gata4 expression in control primitive endoderm (Table S2). Percentages of Gata4+ cells in ICMs were also significantly higher in parthenotes than in controls (Fig. 4D). In conclusion, great decreases in Sox2+ and Nanog+ epiblast cells and progenitors in parthenote embryos were concomitant with dramatic increases of Gata4+ primitive endoderm cells and progenitors.
Increased Fgf3 Expression and Fgfr2 Phosphorylation in Parthenotes
The balance between the expression levels of Gata4 and Nanog determines the cell fate of becoming, respectively, primitive endoderm or epiblast (Mitsui et al.,2003; Rossant et al.,2003; Boyer et al.,2006; Chazaud et al.,2006; Yamanaka et al.,2006). Fibroblast growth factor 4 (Fgf4) signals through its receptor Fgf receptor 2 (Fgfr2) and the adaptor protein Grb2 to activate Gata4/6 expression and repress Nanog expression, thus promoting primitive endoderm formation (Rappolee et al.,1994; Cheng et al.,1998; Goldin and Papaioannou,2003; Chazaud et al.,2006; Nichols et al.,2009; Guo et al.,2010; Yamanaka et al.,2010; Frankenberg et al.,2011). Because we observed an increase in Gata4 and a decrease in Nanog levels in parthenote embryos, we examined the levels of Fgf4 and Fgfr2 proteins in parthenotes. At both the morula and blastocyst stages, it was found that immunostaining patterns and intensities of Fgf4 and Fgfr2 were comparable between controls and parthenotes (Supp. Fig. S4). It is noteworthy that Fgf4 signals were present in plasma membranes of all blastomeres, but were detected in the perinuclear regions of Cdx2− blastomeres only (white arrows in Supp. Fig. S4A, B). This indicates that Cdx2 and perinuclear Fgf4 expression is mutually exclusive among different cells, and the expression of Fgf4 may be generated only by cells destined to become the ICM.
Furthermore, in contrast to unchanged levels of total Fgfr2 in parthenotes, levels of phosphorylated Fgfr2 in parthenote morulae and blastocysts were significantly higher than those in controls, as revealed by the increased intensities of immunostaining for Tyr653/654-phosphorylated Fgfr2 (Fig. 5). As Fgfr2 receptors are activated upon phosphorylation, higher levels of Fgfr2 phosphorylation in parthenotes indicates increased Fgf signaling.
Because Fgf signaling is elevated in parthenotes without an increase of the Fgf4 protein level (Supp. Fig. S4A, B), we assessed other Fgf ligands to identify which one may cause the increased phosphorylation of Fgfr2 in parthenotes. It was shown that Fgf3 and Gata4/6 were co-expressed in the parietal endoderm of post-implantation embryos (Shimosato et al.,2007; Cai et al.,2008; Pilon et al.,2008). Hence we analyzed Fgf3 immunostaining and found that, in contrast to control embryos, which revealed only background levels of Fgf3, parthenotes showed strong Fgf3-positive signals throughout all plasma membranes of both morulae and blastocysts (Fig. 6). Unlike Fgf4, which was expressed in only Cdx2− blastomeres (Supp. Fig. S4A, B), Fgf3 was co-expressed with Cdx2 in some nuclei of parthenote blastomeres (Fig. 6A, B). Furthermore, Fgf3 was co-expressed with Gata4 in all nuclei of parthenote blastomeres, both at the morula and blastocyst stages (Fig. 6C, D). We also performed the in situ hybridization using an Fgf3 cDNA probe as described previously (Tannahill et al.,1992; Wahl et al.,2007). The levels of Fgf3 mRNA were dramatically increased in parthenote morulae and blastocysts compared with controls, consistent with our immunostaining data (Fig. 7). The dramatic up-regulation of Fgf3 level in parthenotes may contribute to the marked increase and ectopic expression of Gata4 (Figs. 4, 6).
Inhibition of Fgfr2 Phosphorylation Restores Normal Nanog and Gata4 Expression in Parthenotes But Does Not Increase Total Cell Numbers of Parthenote Embryos
To further verify the causal linkage between Fgfr2 phosphorylation and Gata4/Nanog levels, we treated E3.0 parthenogenetic morulae with the Fgfr2 inhibitor SU5402 and incubated for 20 hr, followed by immunostaining to analyze the expression levels of Gata4, Nanog, and Fgf3. SU5402 has been shown to be a potent inhibitor of Fgf signaling in the developing mouse embryos (Zuniga et al.,2004; Calmont et al.,2006; Miura et al.,2006; Di-Gregorio et al.,2007). These previous studies indicated that a 50% inhibition of Fgfr phosphorylation was achieved at 10–20 μM of SU5402 (Mohammadi et al.,1997). Other potent inhibitors of Fgf signaling include PD173074 and PD184352 (Nichols et al.,2009; Yamanaka et al.,2010). Nonetheless, because it was demonstrated that SU5402 has a mild effect on mouse embryonic development (Miura et al.,2006), we chose SU5402 for our study and tested the effects of SU5402 at a concentration between 10 and 20 μM.
After treatment with 10 μM of SU5402 for 20 hr, parthenote morulae displayed no significant change of Gata4 and Nanog expression, whereas parthenogenetic blastocysts showed significantly decreased Gata4+ cells and increased Nanog+ cells, despite that Gata4 was still expressed ectopically in many trophectoderm cells (compare Fig. 8A with C). At 15 μM of SU5402, Gata4 was detected in less than four trophectoderm cells of parthenote blastocysts, indicating a dramatic decrease in ectopic Gata4 expression (Fig. 8B); in addition, the numbers of Nanog+ cells in parthenotes treated with 15 μM of SU5402 were comparable with untreated control (zygotic) embryos (Fig. 8B, D). At 20 μM of SU5402, Gata4 expression was completely restricted to the ICM of parthenotes, and the percentages of Gata4+ and Nanog+ cells in whole embryos are comparable between parthenotes and controls (Fig. 8D). Thus, these results indicate that 20 μM of SU5402 completely restored the balance between Gata4 and Nanog levels.
It has been shown that treating normal (fertilized) embryos with 10 μM SU5402 significantly down-regulates most primitive endoderm-specific markers, including Gata4, and upregulates many epiblast-specific markers, including Nanog and Sox2 (Guo et al.,2010). Hence we also analyzed the effects of SU5402 on Gata4 and Nanog expression in control blastocysts. We found that SU5402 treatment decreased the total numbers of ICM cells, with the numbers of Gata4+ cells decreased to a much higher extent than the numbers of Nanog+ cells, leading to the reduction in the percentages of both Nanog+ and Gata4+ cells in whole blastocysts (Fig. 8A–C, E). Therefore, in spite of the decreased percentages of Nanog+ cells in whole embryos, the percentages of Nanog+ cells in control ICMs were significantly increased from 50.2 to 82.6% and 92.3% (1–2-fold increase) by 10 and 15 μM SU5402, respectively (Fig. 8E). On the other hand, the percentages of Gata4+ cells in control ICMs were significantly decreased from 49.8 to 17.4% and 7.8% (3–10-fold decrease) by 10 and 15 μM SU5402, respectively (Fig. 8E). Thus the Nanog/Gata4 gene expression levels in the ICM were inversely correlated after treatment with SU5402.
Although Gata4 and Nanog expression was rescued by SU5402 treatment, we found that the total cell numbers of parthenogenetic blastocysts were still significantly fewer than controls, even when treated with 20 μM of SU5402 (Fig. 8F). Thus SU5402 is capable of restoring normal Gata4 and Nanog levels in parthenotes, but fails to increase total cell numbers of parthenotes. The decrease in total cell numbers may be attributed to a decrease in cell proliferation or an increase in apoptosis (Hardy and Handyside,1996; Uranga and Aréchaga,1997). Because parthenotes displayed increased levels of Fgf3, especially in the nuclei (Fig. 6), and the nuclear isoform of Fgf3 was shown to inhibit cell proliferation (Kiefer and Dickson,1995; Antoine et al.,2005), it was of interest to assess whether Fgf3 expression in parthenotes was affected by SU5402. Surprisingly, even when ectopic Gata4 expression was completely suppressed by 20 μM of SU5402, the immunostaining intensity of Fgf3 signals was very strong in most nuclei of parthenotes (throughout the ICM and trophectoderm) (Supp. Fig. S5C). Supp. Figure S5D shows that the percentages of Gata4+ nuclei in parthenotes decreased as the concentration of SU5402 increased, whereas the percentages of Fgf3+ nuclei were not affected by the concentration of SU5402. Therefore, we conclude that the Fgfr2 antagonist SU5402 does not inhibit Fgf3 expression. Given that the nuclear isoform of Fgf3 inhibits cell proliferation (Kiefer and Dickson,1995; Antoine et al.,2005), the high level of Fgf3 in the nuclei of parthenogenetic embryos may suppress cell proliferation in parthenotes.
Our findings in this study are summarized in Figure 9. In this study, we analyzed expression of lineage-specific genes and found lineage segregation defects in parthenogenetic preimplantation embryos. We found that the number of Gata4+ primitive endoderm cells was dramatically increased while the number of Nanog+/Sox2+ epiblast cells was significantly decreased. Although with decreased numbers, these Nanog+/Sox2+ epiblast cells still exist in the parthenote ICM and can be used to derive patient-specific ES cells for stem cell therapies. On the other hand, there are other drawbacks of parthenogenetic embryos, which had been previously reported, including aberrant expression of imprinted and development-related genes that disrupt full development of organisms (Humpherys et al.,2001; Zvetkova et al.,2005; Bonk et al.,2007; Jiang et al.,2007; Mitalipov et al.,2007; Horii et al.,2008). Our study showed the abnormal gene expression in the preimplantation parthenogenetic embryo, which may affect the derivation and quality of parthenogenetic ES cells.
Consistent with previous studies (Hardy and Handyside,1996; Uranga and Ar&eacaute;chaga,1997), we found significantly reduced total cell numbers in parthenote embryos compared to control embryos from the early blastocyst stage, which was likely due to decreased cell proliferation in parthenotes between the morula and blastocyst stages (Hardy and Handyside,1996; Uranga and Aréchaga,1997). Impaired proliferation and differentiation were observed in both extraembryonic and embryonic lineages of parthenote embryos, especially in the extraembryonic and embryonic mesoderm as well as primitive endoderm lineages (Barton et al.,1985; Sturm et al.,1994; Mognetti and Sakkas,1996). Gata4 and Gata6 are two essential marker genes of the primitive endoderm (Boyer et al.,2006; Yamanaka et al.,2006; Cai et al.,2008; Kuijk et al.,2008), which forms yolk sac and is a major affected lineage in parthenote embryos (Sturm et al.,1994; Newman-Smith and Werb,1995; Chazaud et al.,2006; Yamanaka et al.,2006). Our observation of the overexpression of the parietal endoderm marker genes, Gata4 and Fgf3, in parthenote ICMs is consistent with the previously reported tendency of parthenote ICM outgrowths to differentiate into parietal endoderm cells, which are descendants of the primitive endoderm (Newman-Smith and Werb,1995; Boyer et al.,2006; Cai et al.,2008). Interestingly, we detected Gata4 expression as early as the morula stage, while previous studies failed to detect Gata4 expression till the mid-blastocyst stage (Plusa et al.,2008; Silva et al.,2009; Yamanaka et al.,2010). While the same anti-Gata4 antibody was used (from Santa Cruz Biotechnology, Santa Cruz, CA), differences in experimental procedures may contribute to the different results. For example, while Plusa et al. (2008) fixed embryos in 4% PFA overnight, we fixed for only 15 min after treatment with acid Tyrode's solution. The Gata4 staining in morulae should not result from unspecific signals, as our negative controls using hepatocyte cultures did not show any positive Gata4 signals (data not shown).
During lineage allocation and specification in the ICM, Sox2 and Nanog are two key transcription factors required for development of the epiblast, which forms the embryo body (Avilion et al.,2003; Rossant et al.,2003; Boyer et al.,2006; Chazaud et al.,2006; Yamanaka et al.,2006; Silva et al.,2009; Messerschmidt and Kemler,2010; Frankenberg et al.,2011). In addition, diminished Nanog expression, which was shown to cause excessive differentiation of ICM cells into parietal endoderm cells (Mitsui et al.,2003), was observed in parthenote embryos. Therefore, the misregulated balance between the expression levels of Gata4/6 and Nanog in parthenote ICMs may lead to excessive differentiation into the primitive endoderm lineage and impaired differentiation into the epiblast lineage. It is also noteworthy that elevated Gata4 and reduced Nanog expression in parthenote embryos was first observed at E3.0, suggesting that aberrant cell differentiation in parthenotes may occur as early as the morula stage. As Nanog was demonstrated to maintain self-renewal and the undifferentiated state of ICM and ES cells (Newman-Smith and Werb,1995; Mitsui et al.,2003), it is likely that dramatically reduced numbers of Nanog+ blastomeres in parthenote morulae indicate a significantly decreased amount of proliferating cells since E3.0, thus producing a significantly lower proliferation rate in parthenotes compared to controls between the morula and blastocyst stages.
It has been reported that Cdx2, Elf5, and Tbr2(Eomes) are three of the earliest genes required for specification and differentiation of the trophoblast lineage, with the functioning cascade being: Cdx2→Tbr2→Elf5 (Russ et al.,2000; Donnison et al.,2005; Niwa et al.,2005; Strumpf et al.,2005; Hemberger and Dean,2007; Jedrusik et al.,2008; Ng et al.,2008; Ralston and Rossant,2008). In spite of being downstream of Cdx2 and Tbr2, Elf5 is the key transcription factor creating a feedback loop reinforcing Cdx2 and Tbr2 expression in trophectoderm, which is indispensable for the differentiation of the trophoblast lineage (Donnison et al.,2005; Ng et al.,2008). As shown previously, hemizygous expression of Elf5 was sufficient to cause a marked decrease in Cdx2 protein and Cdx2 and Tbr2 mRNA levels (Ng et al.,2008).
Interestingly, it has been reported that parthenogenetic embryos show a remarkably higher global DNA methylation level compared to normal (fertilized) embryos, especially from the 4-cell to the blastocyst stage (Barton et al.,2001; Li et al.,2009). Previous studies have demonstrated that the early demethylation event in embryogenesis requires the presence of the paternal genome (Barton et al.,2001; Santos and Dean,2004; Morgan et al.,2005). It is likely that the lack of the paternal genome causes demethylation failure in early parthenogenetic embryos, leading to an elevated methylation level of the Elf5 promoter in many of the trophectoderm cells. Reduced Elf5 expression in these trophectoderm cells then leads to decreased Tbr2 expression. However, it is interesting that Cdx2 expression was not affected by reduced Elf5 expression in parthenotes. One possibility is that the remaining Elf5 expression in some cells is sufficient to maintain non-cell-autonomous Cdx2 expression in the neighboring cells. On the other hand, Tbr2 expression may be cell-autonomously regulated by Elf5; thus, Tbr2 is expressed only in Elf5-expressing cells in parthenote trophectoderm.
All of the aforementioned genes are pivotal factors regulating lineage segregation and cell fate determination (Russ et al.,2000; Strumpf et al.,2005; Shimosato et al.,2007; Cai et al.,2008). Similar to Tbr2(Eomes) homozygous mutants, most parthenote embryos demonstrated a lack of diploid dividing trophoblast cells, impaired or retarded development of the neuroectoderm and brain, and multiple mesodermal defects (Sturm et al.,1994; Bulfone et al.,1999; Russ et al.,2000; Kwon and Hadjantonakis,2007). Previous studies using chimeric mouse embryos have demonstrated that parthenogenetically derived cells do contribute to the trophectoderm at the blastocyst stage, but not at the post-implantation stages (Clarke et al.,1988; Thomson and Solter,1988, 1989). Therefore, there is an early post-implantation defect in the parthenogenetic trophoblast lineage. In spite of the trophoblast defect, parthenote embryos are capable of post-implantation development until the 25-somite stage (at around E10.0) (Sturm et al.,1994; Penkov et al.,1995; Kono et al.,1996; Mognetti and Sakkas,1996), and trophoblast giant cells were detected in parthenote placentae (Barton et al.,1985; Sturm et al.,1994; Mognetti and Sakkas,1996). On the other hand, Tbr2 homozygous mutants arrested at the peri-implantation stage (E6.0–E7.5) with differentiation defects of the trophectoderm and incompetence to form trophoblast stem cells (Russ et al.,2000). This apparent difference is probably due to the finding that Tbr2 expression is reduced but not completely absent in the parthenote trophectoderm, as observed in this study.
Our observation that Cdx2 expression is normal in the parthenote trophectoderm and both Elf5 and Tbr2 are expressed in over 40% of parthenote trophectoderm cells is consistent with the partially if not completely specified trophectoderm lineage in parthenotes (Sturm et al.,1994; Mognetti and Sakkas,1996). Interestingly, it has been shown that Fgf signaling is indispensible for trophectoderm development and maintenance of Cdx2 and Tbr2 expression (Tanaka et al.,1998; Nichols et al.,2009). On the other hand, ectopic Gata4 expression is correlated with increased Fgf signaling in the parthenote trophectoderm. Therefore, overexpression of Gata4 may be independent of reduced Tbr2 expression in parthenote embryos.
It has been shown that Fgf4-Fgfr2 signaling is indispensable for trophoblast proliferation and the maintenance of trophectoderm and ICM identities between E3.5 and E4.5 (Feldman et al.,1995; Chai et al.,1998; Tanaka et al.,1998; Haffner-Krausz et al.,1999; Goldin and Papaioannou,2003). In this study, we found elevated Fgfr2 signaling in parthenotes starting from the 16-cell morula stage, as indicated by increased Fgfr2 phosphorylation, in spite of unchanged Fgf4 expression in parthenotes. We found that Fgf3 expression was undetectable or expressed at a low-to-moderate level in normal mouse preimplantation embryos (Rappolee et al.,1988, 1994; Zhong et al.,2006), but was strongly expressed in parthenote embryos. Our finding that inhibition of Fgfr2 suppressed Gata4 but not Fgf3 expression in parthenotes (Supp. Fig. S4) indicates that Gata4 is not the only factor up-regulating Fgf3 expression, and that increased Fgf3 expression is precedent to increased Gata4 expression in parthenotes. Our observations indicated that Fgf3 did not up-regulate Gata4 expression when Fgfr2 signaling was inhibited. It remains to be studied which factor(s) induce(s) Fgf3 overexpression in parthenotes. Assessing the expression levels of Fgf3 in parthenogenetic embryos earlier than the morula stage will unravel the earliest stage at which Fgf3 up-regulation is first observed. The factor(s) inducing Fgf3 expression in parthenotes should be detectable at this stage.
Interestingly, Fgf3 was reported to have dual subcellular fates and play dual roles in regulating proliferation, with the secreted isoform promoting but the nuclear isoform inhibiting cell proliferation (Kiefer et al.,1994; Kiefer and Dickson,1995; Antoine et al.,2005). We have shown that Fgf3 immunostaining signals are apparently brighter in the nuclei than on the plasma membranes of parthenotes, indicating more nuclear than secreted Fgf3 proteins, thus leading to the inhibition of cell proliferation. The excessive nuclear Fgf3 proteins may contribute to the reduction of total cell numbers in parthenotes. Therefore, increased Fgf3 and ectopic Gata4 expression in parthenote trophectoderm together perturb the balance between differentiation and proliferation of trophoblast cells.
Interestingly, recent studies demonstrated that inhibition of the FGF/MAP kinase signal caused almost all ICM cells in normal (fertilized) embryos to become Nanog+ epiblast cells, and no or very few Gata4+/Gata6+ primitive endoderm cells were detected (Guo et al.,2010; Yamanaka et al.,2010). Consistent with previous findings of inverse correlation of Nanog/Gata4 expression levels (Guo et al.,2010; Yamanaka et al.,2010), we found SU5402 significantly increased the precentages of Nanog+ cells and decreased the percentages of Gata4+ cells in control ICMs. Interestingly, the extent of increase in the percentages of Nanog+ cells (less than 2-fold increase) was much lower than the extent of decrease in the percentages of Gata4+ cells (up to 6–7-fold decrease). This is in agreement with the previous study showing a 6-fold down-regulation of Gata4 expression and 2-fold up-regulation of Nanog expression in embryos treated with 10 μM SU5402 from the 16-cell stage for 24 hr (Guo et al.,2010).
Our results indicate that the decrease of total cell numbers in parthenotes is not associated with aberrant Nanog or Gata4 expression, as the restoration of normal Nanog and Gata4 expression is not concomitant with increased cell numbers. Instead, increased Fgf3 expression, which is not affected by SU5402 treatment, may be associated with decreased cell numbers in parthenotes, as the excessive nuclear Fgf3 proteins may suppress cell proliferation and thus reduce total cell numbers.
Parthenogenetic Activation and In Vitro Culture of Preimplantation Embryos
Animals used in this study were purchased from BioLASCO Taiwan (Taipei, Taiwan), and approval was received from the Academia Sinica Institutional Animal Care and Utilization Committee. B6DBA female mice at 10–14 weeks old were superovulated by an intraperitoneal (i.p.) injection of 5 IU pregnant mare serum gonadotropin (Merck, Darmstadt, Germany), followed by an i.p. injection of 5 IU human chorionic gonadotropin (hCG) (Sigma-Aldrich, St. Louis, MO) 48–50 hr later. In order to synchronize the time of fertilization with the time of parthenogenetic activation, 12.5 hr after hCG injection, the female mice were separated into two groups: the first group was individually paired with males of the same strain for 1.5 hr and then checked for copulation plugs, and the second group was sacrificed while the first group was mating. The unfertilized oocytes enclosed in cumulus masses were released from the ampullae, and cumulus cells were removed by pipetting using a mouth-controlled pipette (with an inner diameter of 200–300 μm) after 5 min of treatment with EmbryoMax® M2 medium containing 50–100 U/ml hyaluronidase (Millipore, Billerica, MA). The unfertilized oocytes were then washed and equilibrated in 35-μl droplets of potassium simplex optimized medium (EmbryoMax® KSOM, Millipore, Billerica, MA) at 37°C in a humidified atmosphere of 5% CO2 in air. After 1 hr of equilibration, parthenogenetic activation was conducted in CZBG medium containing 10 mM strontium chloride (SrCl2) and 5 μg/ml cytochalasin B for 4.5 hr at 37°C with 5% CO2 in air, as described before (Gao,2006). The formation of pronuclei was observed between 3 and 3.5 hr after incubation in the activation medium, i.e., between 16.5 and 17 hr after hCG injection. The timing of pronulcei formation in parthenotes was consistent with previous observations (Abramczuk and Sawicki,1975).
At the same time, for the oocytes of the first group of female mice, fertilization was reported to take place from 2.5 hr following pairing, and the first pronuclei were found to appear 2 hr after fertilization (Abramczuk and Sawicki,1975; Hardy and Handyside,1996). Thus the first pronuclei in fertilized oocytes appeared about 4.5 hr after pairing, i.e., approximately 17 hr after hCG injection. The mated female mice were sacrificed 4.5 hr after pairing, with the cumulus cells removed in the EmbryoMax® M2 medium containing 50–100 U/ml hyaluronidase (Millipore). The presence of pronuclei was confirmed under microscope. According to our observations, the time points of the first pronuclear formation are between 16.5 and 17 hr after hCG injection for both the fertilized control and parthenogenetic oocytes. All oocytes at the pronuclei stage were then transferred to KSOM medium and incubated till they reached the developmental stage at which they were analyzed. Embryos normally cleave to two-cells at E1.0–1.5, and form morulae at E2.5–3.0 (48–60 hr in culture). Early blastocysts are formed at E3.5 (70–72 hr in culture), blastocysts are formed at E4.0 (80–84 hr in culture), and late expanded blastocysts are formed at E4.5 (92–96 hr in culture). According to our observation, culturing fertilized oocytes in SrCl2-containing medium for the same time as unfertilized (parthenogenetic) oocytes did not affect immunostaining results in these control embryos, indicating that chemical activation by SrCl2 does not contribute to the differential gene expression between controls and parthenotes. These observations are consistent with many previous reports (Loren and Lacham-Kaplan,2006; Kyono et al.,2008; Chen et al.,2010).
For immunostaining, embryos were washed for 5–10 s in droplets of acidic Tyrode's solution (made by Sigma-Aldrich; purchased from Uni-onward, Taipei, Taiwan) to remove the zona pellucida, and then fixed in 4% paraformaldehyde (Sigma-Aldrich) in 1× phosphate-buffered saline (PBS) for 15 min at room temperature. Embryos were then permeabilized with 0.25% Triton X-100 for 15 min, followed by washing and blocking for 1 hr in blocking solution containing 0.05% Tween-20, 3% bovine serum albumin (BSA), and 5% normal goat serum in 1× PBS. After blocking, embryos were incubated at 4°C overnight with the following primary antibodies diluted in blocking solution: Cdx2 (mouse monoclonal; 1:100 dilution; BioGenex, San Ramon, CA), Elf5 (mouse monoclonal; 1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA), Fgf3 (rabbit polyclonal; 1:200 dilution; Abcam, Cambridge, MA), Fgf4 (rabbit polyclonal; 1:200 dilution; Abcam), Fgfr2 (rabbit polyclonal; 1:50 dilution; Abgent, San Diego, CA), phosphor-Fgfr2-pY653/654 (rabbit polyclonal; 1:50 dilution; Abgent), Gata4 (mouse monoclonal; 1:100 dilution; Santa Cruz Biotechnology), Nanog (rabbit polyclonal; 1:100 dilution; ReproCELL, Tokyo, Japan), Oct4 (rabbit polyclonal; 1:200 dilution; Santa Cruz Biotechnology), Sox2 (mouse monoclonal; 1:100 dilution; Millipore), and Tbr2 (rabbit polyclonal; 1:100 dilution; Millipore). On the second day, the embryos were washed and blocked for 1 hr in blocking solution, followed by incubation at room temperature for 1 hr with the following secondary antibodies conjugated with fluorophores: goat anti-mouse AlexaFluor 488 (green fluorescence) and goat anti-rabbit AlexaFluor 555 (red fluorescence) (Invitrogen Taiwan, Taipei, Taiwan). After incubation, the embryos were washed for 10 min in washing solution containing 0.2% Triton X-100 in 1× PBS, and counter-stained with 0.2 μg/ml DAPI in washing solution for 10 min, followed by mounting in VectaShield (Vector Laboratories, Burlingame, CA) on glass slides.
mRNA In Situ Hybridization
Fgf3 mRNA in situ hybridization was performed with DIG-AP Rembrandt® Universal RISH and Detection Kit (Invitrogen, Carlsbad, CA), following the manufacturer's instructions. The sequence of Fgf3 cDNA probe has been described previously (Tannahill et al.,1992; Wahl et al.,2007).
Series of confocal sections through 3D preserved embryo nuclei were collected using a Leica TCS SP5 confocal microscope (located on the 5th flour of Genomics Research Center, Academia Sinica) equipped with a Super Z galvanometer stage and Plan Apo 63×/1.4 NA oil immersion objectives. Fluorochromes were visualized using an argon laser with an excitation wavelength of 488 nm (for AlexaFluor 488), a DPSS laser with a laser line of 561 nm (for AlexaFluor 555), and a diode laser with a laser line of 405 nm (for DAPI). For each optical section, images were sequentially collected using the XYZ mode for two or three fluorochromes. The pinhole was set to 1–1.5 Airy units and the scan zoom was 1.5×. In order to compare the relative intensities of immunostaining between control and parthenote embryos, identical scanning parameters including the strength of laser emissions were maintained for controls and parthenotes stained with the same antibodies. Images of optical sections were then analyzed using Leica Application Suite, and 3D and maximum projections were constructed from serial stacks of sections for each embryo.
Cell Counting and Statistical Analyses
The image files of optical sections of each embryo were analyzed by the Count Nuclei/Cell Sorting Application Module for MetaMorph (MetaMorph Offline vers. 7.0; Universal Imaging Corporation™, Buckinghamshire, UK) to count cell numbers for the entire embryo and for each antibody immunostained sample. Statistical significances are represented by P values, which were calculated by Student's t-test. The lower the P value, the more significant the difference between control and parthenote embryos. The difference was regarded as non-significant when P ≥ 0.05, as significant when P < 0.05, and as highly significant when P < 0.01.
Inhibition of Fgfr2 Signaling
Culturing mouse embryos with SU5402 to inhibit Fgf signaling has been described in previous studies (Zuniga et al.,2004; Calmont et al.,2006; Miura et al.,2006; Di-Gregorio et al.,2007). SU5402 was dissolved in 100% dimethyl sulfoxide (DMSO) at 10 mM (stock solution) and stored at −20°C until use. Because a 50% inhibition of Fgfr phosphorylation was achieved at 10–20 μM of SU5402 (Mohammadi et al.,1997), we assessed the effects of SU5402 at 10, 15, and 20 μM, respectively, on the development of both control and parthenogenetic embryos. Embryos were cultured in the KSOM medium containing different concentrations of SU5402 for 20 hr since the early morula stage (E3.0) and then analyzed by immunostaining at the mid-blastocyst stage (E4.0). For experimental controls, we treated embryos for 20 hr with an equal concentration of DMSO in KSOM (0.1–0.2 % final).
We thank Chien-Hong Chen for providing the parthenogenetic activation protocol and Li-Wen Lo of the Core Facility of the Genomics Research Center for expert assistance with confocal microscopy. We also thank Drs. Hung-Chih Kuo and Cheng-Fu Kao in the Institute of Organismic and Cellular Biology for critical comments on this project and manuscript. This work was supported by grant 94M011-1 from the Genomics Research Center of Academia Sinica and NSC 98-3111-B-001-003 to John Yu.