Summary
The establishment of cell polarity de novo in the early mammalian embryo triggers the transition from totipotency to differentiation to generate embryonic and extra-embryonic lineages. However, the molecular mechanisms governing the timing of cell polarity establishment remain unknown. Here, we identify stage-dependent transcription of Tfap2c and Tead4 as well as Rho GTPase signaling as key for the onset of cell polarization. Importantly, advancing their activity can induce precocious cell polarization and ectopic lineage differentiation in a cell-autonomous manner. Moreover, we show that the asymmetric clustering of apical proteins, regulated by Tfap2c-Tead4, and not actomyosin flow, mediates apical protein polarization. These findings identify the long-sought mechanism for the onset of polarization and the first lineage segregation in the mouse embryo.
Introduction
The fertilized egg is totipotent and so can give rise to any embryonic or extra-embryonic tissue. In the mammalian embryo, totipotency becomes restricted as cells undertake the first cell fate decision and generate two distinct cell populations: the inner cell mass (ICM) and the outer extra-embryonic trophectoderm (TE). The ICM will form the epiblast (EPI), the future fetus, and the extra-embryonic primitive endoderm (PE), the future yolk sac. The TE will form the placenta. Formation of these three cell lineages by the time of implantation is a prerequisite for a successful pregnancy. Embryo polarization is key to the segregation of the ICM and TE lineages (Johnson and Ziomek, 1981a; Korotkevich et al., 2017) and in the mouse, this process happens at the late 8-cell stage(Fleming et al., 1986; Johnson and Ziomek, 1981a, b) when all cells acquire apical domains composed of the conserved Par protein complex, ERM proteins and enclosed by an actomyosin ring (Louvet et al., 1996; Plusa et al., 2005). In subsequent cell divisions, cells that retain this apical domain express the transcription factors Cdx2 and Gata3 to acquire TE fate, whereas the apolar cells maintain pluripotency to become ICM (Nishioka et al., 2009; Ralston et al., 2010).
Despite the importance of cell polarization for lineage segregation, the mechanisms establishing cell polarization and its timing in the mammalian embryo remain unknown. The apical domain present at the late 8-cell stage is unique as the ability of the cells to polarize is temporally restricted specifically to this stage. In addition, the apical domain can be established in a self-organized manner and in the absence of cell adhesion. However, the identity of the upstream genetic regulators and the mechanism driving cell polarity establishment remain unknown. Here, we use a combination of embryological, gene-editing and live-imaging methods to identify the crucial factors sufficient to trigger apical domain assembly, and the mechanisms by which these factors act.
Results
Transcription is required for apical domain assembly
It is well established that mouse embryo blastomeres polarize at the late 8-cell stage. We have recently found that apical domain formation requires actomyosin activation downstream of Rho GTPase signalling (Zhu et al., 2017). However, Rho signalling alone is insufficient to advance the timing of cell polarization (Zhu et al., 2017), suggesting the existence of additional factors parallel to actomyosin for apical domain formation. As there appeared to be a relationship between the developmental timing of zygotic genome activation and the onset of cell polarity establishment in different mammalian species(Brunet-Simon et al., 2001; Graf et al., 2014; Koyama et al., 1994; Nikas et al., 1996; Telford et al., 1990), we hypothesized that zygotic transcription might play a key regulatory role in this process. To address this hypothesis, we deployed two complementary approaches: first, to inhibit transcription prior to cell polarization and second, to increase the concentration of zygotic transcripts. In both cases, we examined the impact of these manipulations on the timing of apical domain formation.
To inhibit transcription, we treated embryos from the early 8-cell stage with two different transcription inhibitors, 5,6-Dichlorobenzimidazole 1-β-D-ribofuranoside DRB (Bensaude, 2011; Efrat and Kaempfer, 1984) and Triptolide (Bensaude, 2011; Vispe et al., 2009) and examined apical protein localization (Fig. 1A-D; Fig. S1A-G). Inhibition of transcription with either drug led to the failure of apical positioning of the polarity marker, Pard6 (Fig. 1B-D; Fig. S1C-E) although cells continued to undergo cytokinesis (Fig. S1A-B; F-G). Washing-out the reversible transcription inhibitor DRB allowed the resumption of apical domain formation after 9 h, suggesting that failure to polarize was a consequence of transcriptional inhibition (Fig. S1H-J). Thus, transcription is required from the early 8-cell stage for the mouse embryo to polarize.
To increase the concentration of zygotic transcripts, we wished to reduce the cytoplasmic volume as this has been shown to result in an increased concentration of newly synthesized mRNA (Bao et al., 2017; Padovan-Merhar et al., 2015). To this end, we resected 30-40% of cytoplasm from one of the two blastomeres, using a method that has been demonstrated not to compromise developmental potential (Zernicka-Goetz, 1998) (Fig. 1F; Fig. S2A). This resection resulted in a higher concentration of newly transcribed mRNAs in the cytoplasm as assessed by single-molecule fluorescence in situ hybridization (smFISH) (Wang et al., 2012) of a zygotically expressed house-keeping gene, Polr2a (Fig. S2B-E). To determine if blastomere resection affected the timing of cell polarization, we injected embryos with Ezrin-RFP mRNA, to visualize the apical domain in living cells, and then resected blastomeres at either the 2-cell stage or 4-cell stage (Fig. 1F-I; Fig.S2F-H). Both experimental and control embryos established all three lineages at the blastocyst stage, indicating that this procedure did not impair normal development (Fig. S2I-K). Importantly, blastomere resection advanced the timing of cell polarization by 2.1 hr (when carried out at the 2-cell stage; Fig. 1F-I; N=62 pairs; Movie S1) and by 3.3 hr (when carried out at the 4-cell stage; N=76 pairs; Fig. S3F-H; Movie S2). To determine whether the effect on cell polarization we observed was indeed due to transcription, we inhibited transcription in the resected cell, by subjecting it to a 3 hr pulse of DRB. In this case, both resected and control cells polarized simultaneously (Fig. 1K; Fig. S2I-K). Thus, although we cannot exclude the possibility that resection has several effects upon the cell, these results indicate that de novo transcription contributes significantly to apical domain formation.
Redundancy of Tfap2c and Tead4 activity in apical domain formation
We next considered whether zygotic genome activation might either directly activate expression of essential cytoskeletal regulators of cell polarization or indirectly through a specific transcriptional hierarchy. To identify candidate proteins for such regulatory roles, we interrogated previously published single-cell RNA-sequencing data (Goolam et al., 2016) and selected 118 polarity regulators whose transcript levels are upregulated between the 2-cell and 8-cell stages. We also identified 15 transcription factors most likely to be active at this time by analyzing ATAC-seq data (Wu et al., 2016) and of these, we selected 6 that become up-regulated by the 8-cell stage (Fig. S3; Table S1,2). We then downregulated the expression of each of these 124 genes by RNAi and scored the effect on the timing of apical domain formation by time-lapse imaging. We found that depletion of only two of these gene products, the transcription factors Tfap2c or Tead4, prevented cell polarization at the 8-cell stage (Fig. 2A-C, E; Fig. S4A-E). When Tfap2c and Tead4 were depleted individually, polarization was delayed to the 16-cell stage (Fig. 2A-C, E-F) but when both factors were depleted together, cell polarization was abolished (Fig. 2D-H). The depletion of Tfap2c and Tead4 also overcame the precocious cell polarization effect resulting from the reduction of cell volume by blastomere resection. This suggests that Tfap2c and Tead4 mediate premature cell polarization in the resection experiment (Fig. S4F-G).
To confirm the roles of Tfap2c and Tead4 in regulating cell polarization, we genetically depleted both genes by CRISPR-Cas9 mutagenesis. We designed three sgRNAs to target a single protein-coding exon of each gene (Fig. 2I) and injected them into the zygote together with Cas9 mRNA and Ezrin-RFP mRNA, to visualize apical domain formation in vivo. We then categorized the resultant blastomeres based on whether they had undetectable, moderate, or wild-type levels of Tfap2c or Tead4 proteins at the 8-16 cell stage (Fig. S5A-B). Through DNA sequencing we confirmed that the blastomeres with undetectable Tfap2c or Tead4 were homozygous mutants (subsequently termed Tfap2c-null or Tead4-null) (Fig. 2J; Fig. S5C-D). Simultaneous deletion of Tfap2c and Tead4 completely abolished cell polarization, whereas the effects of their individual deletions were less severe (Fig. 2J-l; Fig. S5E-F), in agreement with our RNAi results above (Fig. 2G-H). These findings lead us to conclude that zygotic expression of Tfap2c and Tead4 is required for cell polarization at the 8-cell stage.
Our findings that Tead4 was involved in the onset of cell polarization were unexpected as thus far, Tead4 has been only known to function downstream of cell polarization, following the nuclear re-localization of its transcriptional co-activator Yap to induce expression of TE transcription factors (Nishioka et al., 2009). To gain further insight into the earlier role of Tead4 that we had now uncovered, we examined the localization of Yap. Surprisingly, we found that Yap localizes to the nucleus together with Tead4 before cell polarization at the 8-cell stage (Fig. S6A-C)(Hirate et al., 2015). This nuclear localization of Yap was diminished by downregulation (Fig. S6D-E), and enhanced by upregulation, of Tead4 expression (Fig. S6F-G). These results suggest that at these earlier stages, Tead4 affects the localization of Yap indicating a polarity-independent Tead4 function.
Advancing expression of Tfap2c, Tead4 and Rho GTPase induces premature cell polarization
We next wished to determine whether advancing expression of Tfap2c and Tead4 was sufficient to advance the timing of cell polarization. To this end, we injected Tfap2c and Tead4 mRNAs into one blastomere at the 2-cell stage to elevate their expression by the 4-cell stage, together with Ezrin-RFP as an apical marker (Fig. S7A-C). Advancing the expression of Tead4 alone had no obvious effect on cell polarization (Fig. S7D-E,I-J). By contrast, advancing the expression of Tfap2c led to formation of cell protrusions that were enriched in apical polarity proteins, including Pard6 and Ezrin at the late 4-cell stage (Fig. S7F-I). Advancing the expression of Tfap2c and Tead4 together also induced premature formation of cell protrusions (Fig. S7D-G; Fig. 3B-C; Movie S3-4). These Tfap2c- or Tfap2c-Tead4-induced membrane protrusions were smaller than the natural apical domains formed at the late 8-cell stage and lacked the actomyosin-enclosed apical domain that normally forms (Fig. S7H). These results suggested that Tfap2c and Tead4 expression might be sufficient to lead to the polarization of apical proteins but not their expansion to form an actomyosin enclosed cap-like domain.
We have previously shown that apical domain formation requires activation of actomyosin by PKC-Rho GTPase signaling at the 8-cell stage although actomyosin activation alone is insufficient to trigger apical domain formation (Fig. 3B-C, Movie S5) (Zhu et al., 2017). We therefore hypothesized that activation of Rho GTPase might be required concomitantly with the Tfap2c and Tead4 to achieve complete cell polarization. To test this hypothesis, we expressed Tfap2c and Tead4 at the 2-cell stage (together with the Ezrin-RFP as a live apical marker) and constitutively active RhoA-Q63L at the 4-cell stage (Fig. 3A). Strikingly, when all three factors were expressed, complete cap-like apical domains became established at the 4-cell stage (Fig. 3B-C; Movie S6). These prematurely induced apical domains were enriched with Ezrin and Pard6 and thus strongly resembled the apical domains that form normally at the late 8-cell stage (Fig. 3B; Fig. S8B-C). To our knowledge, this is the first time that any premature cell polarization has been reported in the mouse embryo, where this process has been previously considered invariant.
To further confirm these results, we overexpressed all factors in half of the Ezrin-RFP labelled embryos, using the remaining cells as controls (Fig. 3D). In line with our findings from the expression of these factors in the entire embryo, we observed that individual blastomeres targeted with overexpression of the three factors polarized significantly earlier than control blastomeres from the same embryo (Fig. 3E-F; Movie S7-8). We did not observe any difference in the timing of cell division between blastomeres suggesting that polarization at the 4-cell stage by Tfap2c, Tead4 and active RhoA expression is not caused by a delay to cytokinesis (Fig. S8A).
Together our results indicate that the induction of the transcriptional program triggered by Tfap2c and Tead4 alongside the activation of actomyosin downstream of Rho GTPase signaling constitutes the timing mechanism that triggers apical domain formation at a specific stage of preimplantation development.
Advancing expression of Tfap2c, Tead4 and Rho GTPase advances morphogenesis and cell differentiation
During normal development, polarization at the 8-cell stage is followed by a zippering process in which adjacent apical domains expand and seal their boundaries at the late 16-cell stage, a process essential for blastocyst formation (Zenker et al., 2018). We therefore wished to determine whether the premature cell polarization induced by advancing the onset of Tfap2c/Tead4/RhoA-Q63L expression could also advance the zippering process. To this end, we again induced expression of Tfap2c/Tead4/RhoA-Q63L, to trigger the formation of apical domains at the 4-cell stage, and then followed subsequent development by time-lapse microscopy. The premature formation of the apical domains resulted in premature zippering at the 8-cell stage (Fig. S8B-D). These prematurely zippered sites were enriched with the tight junction protein ZO-1 just as in normal development at the late 16-cell stage (Fig. S8C). These results confirm that advancing the expression of Tfap2c/Tead4/RhoA-Q63L is sufficient to induce the formation of a precocious functional apical domain and also show that advancing cell polarization advances the subsequent step of morphogenesis.
Cell polarization in the mouse embryo is followed by cell fate specification, namely differentiation into the TE cells that inherit an apical domain. Thus, we determined whether changing the timing of cell polarization would affect the timing of TE formation. To this end, we induced apical cap formation at the 4-cell stage, by overexpression of Tfap2c, Tead4 and RhoA-Q63L, and examined the expression of cell differentiation markers. The induction of premature cell polarization induced premature expression of the key TE differentiation transcription factors, Cdx2 and Gata3 (Fig. S8E-H; Movie S9-10). These results suggest that the combined activities of Tfap2c, Tead4 and RhoA-Q63L are sufficient to advance the timing of not only cell polarization but also the cell differentiation program.
Apical protein centralization is achieved by apical protein clustering
We next wished to define the relative roles of actomyosin activation, Tfap2c, and Tead4 in driving apical domain assembly. The formation of the apical domain does not require canonical symmetry breaking cues but is highly responsive to the dynamics of actin cytoskeleton (Fleming et al., 1986; Johnson, 2009; Korotkevich et al., 2017). As it has been shown that disrupting the actin network abolishes apical domain formation (Sun et al., 2013; Zhu et al., 2017), we interrogated interactions between actin and the apical proteins examined by filming the development of LifeAct-GFP and Ezrin-RFP expressing embryos. Our time-lapse movies revealed that apical domain formation occurred in two steps (Fig. 4A). In the first, centralization step, apical proteins became concentrated around the center of the cell-contact free surface concomitant with local exclusion of actin (Fig. 4A-B). In the second, expansion step, apical proteins accumulated and then expanded to form an apical patch overlaying the actin meshwork before being concentrated in a surrounding ring-like structure (Fig. 4A,C; Movie S11). The initial apical protein centralization step failed to take place in Tfap2c and Tead4 depleted embryos, suggesting that these two transcription factors regulate this first step of apical domain formation (Fig. 2G; Fig. 4D).
To our surprise, we did not observe obvious actomyosin movements towards the center of cell-contact free surface that could drive apical protein centralization (Movie S11-12). To confirm this observation, we performed time-lapse imaging using high temporal resolution (4s per frame) and analyzed the movements of actin and Ezrin using embryos expressing LifeAct-GFP and Ezrin-RFP. We observed that the actin formed a circumferential, turbulent cortical flow, and that the movement of actin coordinated with the movement of Ezrin (Fig. 4E; Movie S13). However, the overall direction of these flows was not toward the center of the cell-contact free surface (Fig. 4F; Movie S14). Moreover, the inhibition of actomyosin contractility by blebbistatin treatment, impaired such cortical movements but failed to prevent apical protein centralization (Fig. S9A-B; Movie S15)(Maitre et al., 2015; Zhu et al., 2017). These findings suggest that the turbulent actin flow we observed on the cell surface is unlikely to cause apical protein centralization.
Besides the cortical actin flow, we observed that Ezrin forms clusters on the cell-contact free surface, and that the centralization of Ezrin is accompanied by an exponential growth of the size of Ezrin clusters in the center of the cell-contact free surface (Fig 5A-C; Movie S12). This result suggests the possibility that the centralization of apical proteins is achieved by the conjugation of apical protein clusters on the cell-contact free domain. We also observed a phase-dependent correlation of Ezrin and actin dynamics during the period of apical domain centralization. Initially, when the Ezrin cluster was small, actin and Ezrin localization exhibited a moderate positive correlation (Fig. 5D). Live imaging revealed that the conjugation of Ezrin clusters happened during the merging and splitting of actin clusters that was not prevented by blebbistatin treatment (Fig. 5E), suggesting that actin turnover and consequent actin cytoskeleton remodeling may trigger apical protein clustering and centralization. To test this idea, we treated mid 8-cell stage embryos with two actin inhibitors, firstly Jasp(Bubb et al., 1994), a drug that prevents actin de-polymerization; and secondly CK666(Sun et al., 2013), an inhibitor that prevents Arp2/3 mediated actin polymerization. In support of our idea, Jasp and CK666 treatment consistently blocked the growth of Ezrin clusters and therefore the centralization of Ezrin proteins (Fig. 5F-G; Fig. S9C-D). Interestingly, overexpression of RhoA-Q63L reduced the formation of clusters but increased membrane localization of Ezrin (Fig. S9E-F); on the contrary, depletion of RhoA resulted in ectopic clustering of actin and Ezrin on the cell membrane (Fig. S9G-H). These data suggest that RhoA signaling remodels the cortical property that negatively regulates cluster formation but prompts apical protein membrane localization. In addition, we also observed a correlation between cell curvature and the apical protein clustering site (Fig. S9I), which suggests that apical protein clustering preferentially happen at place with high curvature. To test this idea, we elongated early 8-cell stage cells to induce the asymmetry of cell curvature, using a method we have previously used (Methods)(Gray et al., 2004), followed by the tracking of the position of the apical domain using the time-lapse imaging. We found that the apical domain preferentially develops on the poles with high curvature (Fig. S9J). These results suggest that apical clustering responds to membrane geometry leading it to develop on the cell-contact free surface (Korotkevich et al., 2017).
Together, these observations indicate that apical protein centralization is mediated by the clustering of apical proteins that is regulated by both actin turnover and myosin activity.
Tfap2c and Tead4 control apical protein clustering
We next wished to determine the mechanisms by which Tfap2c and Tead4 regulate apical domain centralization. To this end, we examined how the level of Tfap2c-Tead4 activity would influence the configuration of actin and Ezrin on the apical surface. We found that depletion of Tfap2c and Tead4 resulted in smaller actin and Ezrin cluster sizes by the mid 8-cell stage. Conversely, overexpression of Tfap2c and Tead4 significantly increased the actin and Ezrin cluster sizes at the late 4-cell stage (Fig. 5H-M). These results suggest that Tfap2c and Tead4 regulate apical domain centralization by influencing expression of proteins that modulate apical protein clustering.
To gain further mechanistic insight into how Tfap2c-Tead4 regulate apical domain formation, we carried out RNA-sequencing of early 8-cell stage embryos depleted of Tfap2c and/or Tead4. For each group of embryos (control GFP RNAi, Tfap2c RNAi, Tead4 RNAi, Tfap2c/Tead4 co-RNAi), we collected two biological replicates with 10 embryos per sample, using embryos of two strains to eliminate any effect of the genetic background (Fig. 6A, Methods). The effect of different treatments on the global gene expression pattern was highly reproducible between biological replicates and between genetic backgrounds (Fig. S10A). Differential gene expression analysis (with a 2-fold cut-off) showed that depletion of Tfap2c led to the downregulation of 749 or 929 genes depending on the strain, whereas depletion of Tead4 led to downregulation of 242 or 314 genes (Fig. S10B-C). The co-depletion of Tfap2c and Tead4 led to an additional 135 or 95 genes being downregulated compared to single knockdown embryos, depending on the strain (Fig. S10B-C). Among the genes consistently downregulated in double-knockdown embryos of both strains, a significant proportion was associated with actin polymerization. These include the Arp2/3 complex, the tropomyosin complex, Marcks proteins, and the FREM family member Ebp4.1l5, and Cdc42 effector protein family members (Borg)(Vong et al., 2010) (Fig. 6B; Table S3). The levels of expression of these factors are upregulated between the 2- to 8-cell stages and correlate with the growth of Ezrin clusters over this period. Importantly, the depletion of Tfap2c and Tead4 eliminates their expression resulting in the reduced apical cluster size (Fig. 6B; Fig. 5H-M). To further validate whether the expression of these downstream targets are responsible for apical domain formation, we depleted two of the highly expressed candidates, Arpc1b and Marcksl1, and found that this led to the impaired apical domain formation (Fig. 6C-E). These results together suggest that Tfap2c and Tead4 regulate apical protein clustering by regulating the actin network, to centralize the apical proteins.
In summary, our results show that the activation of actomyosin by Rho GTPases concurrent with the transcriptional activity of Tfap2c and Tead4 triggers cell polarization and the segregation of the TE and ICM fates. Together, these three factors control the initiation of expression of transcription factors required for TE formation and the establishment of the apical domain to execute the first cell fate decision (Fig. 6F).
Discussion
The first bifurcation of cell fate in the mammalian embryo is a fundamental step that separates progenitors of the embryonic and extra-embryonic tissues. The establishment of the apical domain is the primary trigger for this process (Bedzhov et al., 2014; Chazaud and Yamanaka, 2016; Johnson and Ziomek, 1981a; Korotkevich et al., 2017). Although great progress has been made in understanding the connections between cell polarity and cell fate (Anani et al., 2014; Hirate et al., 2013; Samarage et al., 2015), little is known about how the establishment of the apical domain itself is triggered with such temporal specificity during development. Here, we provide evidence denoting the importance of zygotic genome activation in establishing the timing of cell polarization and demonstrate that the zygotic expression of the transcription factors Tfap2c and Tead4 is necessary for this process. We further identified a functional redundancy in their regulation of cell polarization, which can explain why this role was not discovered in previous studies where single knockout embryos were examined (Nishioka et al., 2008; Winger et al., 2006; Yagi et al., 2007). Our finding that the abundance of Tfap2c and Tead4 proteins increases after zygotic genome activation and correlates with the exposure of their DNA binding sites to open chromatin regions (Fig.S4) (Wu et al., 2016) indicates that their zygotic expression accounts for their functional activity. Accordingly, we show that inducing upregulation of Tfap2c and Tead4 in conjunction with Rho GTPase-mediated activation of actomyosin is sufficient to establish the apical domain and this allowed us, for the first time, to advance the timing of cell polarization. In turn, this advances the expression of downstream lineage-specific transcription factors that establish TE identity. This supports the idea of feedback loops between cellular events and the robust segregation of the first lineages (Cao et al., 2015; Yagi et al., 2007).
Apico-basal cell polarization at the 8-cell stage of the mouse embryo has often been viewed as a model for epithelial polarization. However, the mechanism establishing the apical domain is quite distinct from many other cell types as it can be formed in the absence of external cues such as those provided by the extra-cellular matrix or through cell adhesion (Korotkevich et al., 2017). Such spontaneous symmetry breaking properties have also been deployed by a broad array of tumor cells, but their underlying mechanisms remain elusive (Lorentzen et al., 2018). Here we show that the apical proteins form clusters at the cell-contact free surface and the conjugation of apical protein clusters, controlled by Tfap2c and Tead4 activity, leads to the apical protein polarization. Actomyosin flow has been reported to trigger cortical polarization in the C.elegans zygote. We show that although actomyosin flow helps to restore the apical domain after cytokinesis, at the 16-cell stage, we did not observe the actomyosin flow to contribute to the establishment of cell polarity at the 8-cell stage. Such mechanistic differences may correlate with the distinct time-scales of cell polarity establishment between different species, and the utilization of actomyosin flow may account for the more rapid establishment of a cortical domain in the worm embryo and in recovery of the domain in the 16-cell stage mouse embryo.
The clustering of apical proteins could be achieved through a change in the binding kinetics of apical proteins to lipids and actin. Indeed, we observed a strong positive correlation between Ezrin/Pard6 and PIP2 localization throughout the apical domain formation process (data not shown). The biophysical properties of the actin network, determined by actin turnover and myosin activity (mediated by Rho GTPases), may provide the driving force directing lipid-apical protein clustering. It has been reported that the PIP2 lipid prefers to cluster in areas with high cell curvature (Lin et al., 2018). Accordingly we also found a positive correlation between cell curvature and the position of the apical domain. These results may account for the centralized positioning of the apical domain on the cell contact-free surface. PIP2 regulated apical protein localization may provide a mechano-sensing mechanism independent of adherens junction to establish the apical domain, as observed in E-cadherin knockout cells (Korotkevich et al., 2017), it is of future interest to examine this possibility.
In summary, our results provide mechanistic insight into the timing and establishment of de novo cell polarization in the mouse embryo, the critical event for the transition from totipotency to pluripotency.
Author contributions
Conceptualization: M.Z and M.Z.G.; Investigation: M.Z., P.W., C.H; Writing: M.Z and M.Z.G.; Supervision: M.Z.G., N.J.
Declaration of Interests
The authors declare no competing interests.
Animals
This research has been carried out following regulations of the Animals (Scientific Procedures) Act 1986 - Amendment Regulations 2012 - reviewed by the University of Cambridge Animal Welfare and Ethical Review Body. Embryos were collected from F1 females (C57BI6xCBA) that had been super-ovulated by injection of 7.5 IU of pregnant mares’ serum gonadotropin followed by human chorionic gonadotropin (Intervet) 48 h later. F1 females were mated with F1 males.
Mouse embryo culture and inhibitor treatments
Embryos were recovered at the zygote or 2-cell stage in M2 medium and subsequently transferred to KSOM medium for long-term culture, as described previously (Zhu et al., 2017).
Inhibitor treatment: Puromycin (Invivogen, ant-pr-1) was diluted in KSOM to a working concentration of 10μg/ml. Cycloheximide (Sigma-Aldrich, C7698) was dissolved in DMSO and diluted in KSOM to a working concentration of 20μg/ml. 5,6-Dichlorobenzimidazole 1-b-D-ribofuranoside (DRB; Sigma-Aldrich, D1916) and Triptolide (Cayman Chemical, CAY11973) were dissolved in DMSO and diluted in KSOM to a working concentration of 50μM (DRB) or 5μM (Triptolide). C3-transferase was dissolved in distilled water and diluted in KSOM to 7μg/μl. For the control groups, the same dilutions of the vectors of different inhibitors were added to the medium.
Blastomere resection
The resection procedure was performed as previously described(Zernicka-Goetz, 1998). Briefly, the zona pellucida was removed for both 2-cell stage and 4-cell stage embryos. Embryos were transferred to a 1% agarose coated petri-dish covered by M2-medium containing 2µM Cytochalasin D (Sigma-Aldrich, C8273) prior to the resection procedures. For the 2-cell embryo, the embryo was first elongated by using a thin glass capillary with a flame-polished end. One of the blastomeres was resected using a thin glass needle, leaving approximately 30-40% of the cytoplasm (cytoplast) attached to its sister cell (Fig. 2B). Cell volume measurement was performed by applying a 3D polygon ROI around the periphery of the structure (indicated by Ezrin-RFP) throughout the Z-stack. The measurement was performed by using Icy software. For the 4-cell stage resection, the 4 blastomeres were first transferred to Calcium-Magnesium free M2 medium for 5 min and cells were dissociated by pipetting, as previously described (Graham et al., 2014). All four blastomeres were elongated using a thin glass capillary, and only two were resected (Fig. 2C). The small and control cells were transferred to M2 medium immediately after resection. The whole resection process for each embryo took up to 5 min with a survival rate greater than 80%. The small and control cells were transferred to KSOM medium for long-term culture.
Blastomere elongation
Elongation of 8-cell stage blastomeres were performed using the method described previously(Gray et al., 2004), the cells from an early 8-cell stage embryo (0-1hr post cell division) were disassociated as described in blastomere resection experiments. The single cell was placed in KSOM supplemented with containing sodium alginate (0.5%). The cell was then elongated by pipetting with a thin glass capillary. Immediate after the elongation procedure, a few drops of a 1.5% CaCl2 (0.3g CaCl2 dissolved in 20.0 ml 0.15M NaCl) were added, leading to the gelling of KSOM surrounding the cell and hence the maintenance of cell shape. The excessive CaCl2 was removed by replacing the medium with KSOM.
Microinjection
Microinjection was carried out as described previously (Zernicka-Goetz et al., 1997). In brief, embryos were placed in M2 medium on a glass slide with a depression and covered by a drop of mineral oil. Microinjection was performed with an Eppendorf Femtojet Microinjector. Negative capacitance was used to facilitate penetration through the membrane. dsRNA was injected at a concentration of 1μg/μl. Synthetic mRNAs were injected at the following concentration: Ezrin-Ruby (400ng/μl); Ezrin-Venus (400ng/μl); Tfap2c (15ng/μl); Tead4 (15ng/μl); RhoA-Q63L (3ng/μl); GFP-Myl12b (300ng/μl); Cas9 (100ng/μl). All sgRNAs were injected at 25ng/μl.
Preparation of DNA Constructs
pRN3P was used as the vector for all constructs as previously described (Zernicka-Goetz et al., 1997). To construct pRN3p-Tead4, Tead4 was amplified from mouse kidney cDNA and cloned into the pRN3p vector. To construct pRN3p-Tfap2c, Tfap2c cDNA was purchased from Origene (MR207174) and cloned into the pRN3p vector. pRN3p-Cas9 was a gift of J. Na (Tsinghua University, School of Medicine). Ezrin-Ruby, Ezrin-Venus, LifeAct-Ruby, GFP-Myl12b, RhoA-Q63L were as previously described (Zhu et al., 2017). All primers for constructs preparation are listed in Table S4.
mRNA, dsRNA, sgRNA preparation
For mRNA preparation, constructs for each mRNA were linearized using a restriction site downstream of the poly-A region. In vitro transcription was performed using the mMessage mMachine T3 kit (Thermo Fisher, AM1348) following the manufacturer’s instructions. mRNAs were purified using the lithium chloride precipitation method. For sgRNA preparation, the sequences of sgRNAs were designed using CRISPR design tool website (http://cirpsr.mit.edu). The DNA fragment containing T7 promoter, crRNA and sgRNA sequence were amplified using the Geneart gRNA kit (Thermo Fisher, A29377). sgRNAs were in vitro transcribed and purified using the gRNA Clean Up Kit (Thermo Fisher, A29377), following manufacture’s instructions.
All dsRNAs were designed using the E-RNAi website (Horn and Boutros, 2010) and were 350-500bp in length. The specific targeting regions for each dsRNA were amplified from a mixture of mouse kidney, lung, liver cDNAs. The in vitro transcription reactions were performed using the MEGAscript T7 transcription kit (Thermo Fisher, AM1334) following the manufacturer’s instructions. dsRNAs were purified by lithium chloride precipitation. All primers for dsRNA preparation are listed in Table S3.
Immunofluorescence
Embryos were fixed in 4% PFA at room temperature for 20min, and washed in PBST (0.1% Tween in PBS) three times. The embryos were then permeabilised in 0.5% Triton X-100 in PBS for 20 min at room temperature, washed in PBST three times, transferred to blocking solution (3% bovine serum albumin) for 2h and incubated with primary antibodies (diluted in blocking solution) at 4 °C overnight. After the incubation, embryos were washed in PBST and incubated with secondary antibodies (1:500 in blocking solution) for 1h at room temperature. Embryos were stained with DAPI (1:1000 dilution, in PBST, Life Technologies, D3571) for 15 min, followed by two washes in PBST. Primary antibodies: rabbit polyclonal anti-Pard6b (Santa Cruz, sc-67393, 1:200); mouse monoclonal anti-GFP (Nacalai Tesque Inc., 04404-84, 1:500). mouse monoclonal anti-Tfap2c (Santa Cruz, sc-12762, 1:200); goat monoclonal anti-Tfap2c (R&D Systems, AF5059-SP, 1:200); rabbit monoclonal anti-Tead4 (Abcam, ab97460, 1:200); mouse monoclonal anti-Tead4 (Abcam, ab58310, 1:100); goat monoclonal anti Sox17 (R&D Systems, af1924); mouse monoclonal anti Cdx2 (Launch Diagnostics, MU392-UC (Biogenex), 1:200); rabbit monoclonal anti Nanog (Abcam, ab80892, 1:200); mouse monoclonal anti-Tjp1 (Thermo Fisher Scientific, 33-9100, 1:200); rabbit monoclonal anti-phosphorylated-Yap (Cell Signaling Technologies, 4911S, 1:200); mouse monoclonal anti-Yap (Santa Cruz, sc-101199, 1:200); rabbit monoclonal anti-di-phosphorylated MRLC (Cell Signaling Technologies, 3674P, 1:100); goat polyclonal anti-Amot (Santa Cruz, sc-82491, 1:1000). Secondary antibodies: Alexa Fluor 568 Donkey anti-Goat (A-11057, ThermoFisher Scientific); Alexa Fluor 488 Donkey anti-Mouse, (A-21202, ThermoFisher Scientific); Alexa Fluor 568 Donkey anti-Mouse (A10037, ThermoFisher Scientific); Alexa Fluor 647 Donkey anti-Mouse (A31571, ThermoFisher Scientific); Alexa Fluor 568 Donkey anti-Rabbit (A10042, ThermoFisher Scientific); Alexa Fluor 647 Donkey anti-Rabbit (A-31573, ThermoFisher Scientific).
Real-time PCR
RNA was extracted from 8-cell stage embryos using the Arcturus PicoPure RNA isolation kit (Arcturus Bioscience). RT-PCR was performed using a StepOne Plus Real-time PCR machine (Applied Biosystem). The expression level was calculated using ddCT methods, normalised to a Gapdh PCR reaction and the endogenous control group. The primers used for RT-PCR are listed in Table S4.
Imaging and data processing
Imaging was carried out on a Leica-SP5 or a Leica-SP8 confocal using a Leica 1.4 NA 63X oil (HC PL APO) objective. Images were processed with Fiji software (Schindelin et al., 2012). For the analysis of nucleo-cytoplasmic signal intensity ratio, the region of the nucleus, and a cytoplasmic region of the same size, were cropped and the mean signal extracted using the Fiji ROI function. To normalise signals to the level of DAPI fluorescence, the Fiji ROI function was used to extract the nuclear region stained to reveal specific proteins and for the equivalent DAPI channel and normalised using the formula: I(protein of interest)/(DAPI). For Ezrin apical enrichment analysis, a freehand line of the width of 0.5μm was drawn along the cell-contact free surface (apical domain), or cell-contact (basal) area of the cell, signal intensity was obtained via ROI function of Fiji. The apical/basal signal intensity ratio is calculated as: I(apical)/I(basal). Cells on the same plane were subjected to this analysis. Compaction was assessed by measuring the inter-cellular blastomere angle in the mid-plane between adjacent cells (as described previously(Zhu et al., 2017)) by using the Fiji angle function.
For live-imaging, time-lapse recordings of embryos were carried out using a spinning disk or a Leica-SP5 scanning confocal. For the blastomere resection experiment, time-lapse frames were acquired every 1hr; for other live-imaging experiments, time-lapse frames were acquired every 20-30min. Images were acquired using a 3-4μm Z-step. Images were processed with Fiji software. Correlations were calculated using Prism software (http://www.graphpad.com).
Particle Image Velocimetry (PIV) analysis
PIV analysis was performed using PIVlab MATLAB algorithm (pivlab.blogspot.de). The sequential images with 3-4s/frame were used as input. 2-pass analysis with 80×40 pixel was used, and the images were analyzed with A-B,B-C,.. sequence. A mask demarcating the edge of the cell were applied to the image before processing the analysis.
Statistics
Statistical methods are indicated for every experiment in the corresponding figure legends. Qualitative data is presented as a contingency table and was analyzed using Fisher’s exact test. Normality of quantitative data was first analyzed using D’Agostino’s K-squared test. A one-sample t-test was used to test whether an observed distribution followed a hypothetical mean. If data showed a normal distribution, then for comparison of two or multiple samples, an unpaired two-tailed Student’s t test (two experimental groups) or a One-way ANOVA test (more than two experimental groups) was used to analyze statistical significance. Differences in variances were taken into account by performing a Welch’s correction. For data that did not present a normal distribution, a Mann–Whitney U-test (two experimental groups) or a Kruskal–Wallis test with a Dunn’s multiple comparison test (more than two experimental groups) was used to test statistical significance. To determine the influence of different groups in multiple variants, two-way ANOVA was performed. Statistical analyzes were performed using Prism software (http://www.graphpad.com).
RNA extraction and sequencing
For sample collection, 10 8-cell stage embryos injected with dsRNAs were treated with acidic tyrode solution to remove zona pellucida, washed in PBS (without Ca2+ and Mg2+) and transferred to transferred to hypotonic lysis buffer (Amresco, M334). mRNA was reverse transcribed by SuperScript II and pre-amplified using Smart-seq2 protocol as described previously (Picelli et al., 2014). Pre-amplified cDNA was fragmented by Tn5 enzyme, followed by library generation using TruePrep® DNA Library Prep Kit V2 for illumina Kit (Vazyme, TD501-503). Sequencing was performed on HiSeq X Ten platform.
RNA-sequencing data Processing
Raw reads with adaptors, low-complexity or low-quality were trimmed by trim_galore. Then clean data were mapped to mouse genome (mm10) by STAR. FPKM (Fragments Per Kilobase per Million mapped reads) of Refseq genes were calculated by Cufflinks. Htseq-count was used to count the mapped reads number. Subsequent reads number was used to perform differential gene expression analysis by using R packages “DESeq2” (Fold change >2, P value < 0.05). Heatmap and volcano plot were graphed using R (http://www.r-project.org/).
Acknowledgement
We are grateful to David Glover and Marta Shahbazi for valuable comments on the manuscript; Ed Munro for helpful discussion of the project; Stavros Malas for providing the Gata3-GFP transgenic line. This work was supported by Wellcome Trust (098287/Z/12/Z), ERC (669198) and Leverhulme Trust (RPG-2018-085) grants to M.Z.G.