CDX2 downregulation in mouse mural trophectoderm during peri‐implantation is heteronomous, dependent on the YAP‐TEAD pathway and controlled by estrogen‐induced factors

Abstract Purpose To investigate the transition of CDX2 expression patterns in mouse trophectoderm (TE) and its regulatory mechanisms during implantation. Methods Mouse E3.5–4.5 blastocysts were used to immunostain CDX2, YAP, TEAD4, and ESRRB. Endogenous estrogen signaling was perturbed by administrating estrogen receptor antagonist ICI 182,780 or ovariectomy followed by administration of progesterone and β‐estradiol to elucidate the relationship between the transition of CDX2 expression patterns and ovarian estrogen‐dependent change in the uterine environment. Results CDX2 expression was gradually downregulated in the mural TE from E4.0 in vivo, whereas CDX2 downregulation was not observed in blastocysts cultured in KSOM. Fetal bovine serum (FBS) supplementation in KSOM induced CDX2 downregulation independently of blastocyst attachment to dishes. CDX2 downregulation in the mural TE was repressed by administration of ICI 182,780 or by ovariectomy, and administration of β‐estradiol into ovariectomized mice retriggered CDX2 downregulation. Furthermore, Cdx2 expression in the mural TE might be controlled by the YAP‐TEAD pathway. Conclusions CDX2 downregulation was induced heteronomously in the mural TE from E4.0 by uterus‐derived factors, the secretion of which was stimulated by ovarian estrogen.


| INTRODUC TI ON
Implantation is a specific and essential process in mammalian development. Importantly, pregnancy is established through successful implantation. In human, the conception rate is about 30% per menstrual cycle, and three-quarters of pregnancy losses are attributed to failure of implantation. 1 In addition, the low efficiency of implantation success is a major hurdle for infertile patients who use assisted reproductive technology (ART) to become pregnant. 2 Thus, although an understanding of implantation mechanisms is indispensable for enhancing the fertility of mammals, studies on implantation have been limited due to ethical considerations and to the difficulty of direct observation and manipulation.
In mammals, zygotes repeat cleavages and become compacted embryos, termed morulae, while moving in the oviduct. Morulae develop into blastocysts through blastocoel swelling after reaching the uterus. After hatching from the zona pellucida, blastocysts initiate implantation during a limited period called the implantation window. 3 This window is regulated by ovary-derived steroid hormones. 3,4 Especially, estradiol-17β induces uterine receptivity for blastocysts directly and blastocyst adhesion to the uterus indirectly via endometrial gland secretions such as leukemia inhibitory factor (LIF) and osteopontin (OPN). [5][6][7][8][9] In mice, estradiol-17β levels rise transiently during 3.5 days postcoitum (dpc), and the spike of ovarian estradiol-17β secretion strictly regulates the timing of implantation. 3,10 In human and rodents, the processes of implantation are subdivided into three steps: apposition, attachment, and invasion. Once ovarian estradiol-17β stimulates uterine receptivity and blastocyst activation, mouse blastocysts get close to and contact the uterine luminal epithelium (apposition) around E3.5, and start connecting to the uterine tissue (attachment) from E4.0. After the attachment process is completed by E4.5, the blastocysts start invading (invasion) the uterine tissue from E5.0. 10,11 Therefore, E3.5, 4.0, and 4.5 could be defined as apposition, attachment, and postattachment periods, respectively, in mice.
A blastocyst is constituted of inner cell mass (ICM) and trophectoderm (TE). In early-stage blastocysts, ICM includes progenitors of epiblast (EPI) and primitive endoderm (PrE), which are the origin of fetus and yolk sac, respectively, in a random "salt and pepper" pattern. Those progenitors are completely committed to EPI and PrE lineages, and PrE-committed cells localize along the blastocoel following blastocyst development for the implantation. [12][13][14][15] TE is subdivided into polar and mural parts, which contact and do not contact the ICM, respectively. In mice, polar TE differentiates into placental cells such as trophoblast giant cells (TGCs), spongiotrophoblasts, and labyrinthine trophoblasts. On the other hand, mural TE differentiates into only TGCs to form a yolk sac with PrE-derived parietal and visceral endoderm, which is indispensable for exchanging nutrients and endocrine signals between mother and fetus before placental formation. [16][17][18] In addition, mural TE acquires adhesiveness and invasiveness by transforming cell polarity, motility, and intercellular junctions through its differentiation into TGCs to promote implantation. [17][18][19][20] Thus, the differentiation of mural TE is essential for initiating and advancing implantation in mice. However, the details about mural TE differentiation remain mostly unknown.
CDX2, a member of the caudal-related homeobox transcription factor gene family, could be initially detected in morula-stage embryos and become restricted to outer cells (TE progenitors) by the blastocyst stage. 21 In these stages, CDX2 contributes to the repression of the pluripotent program and the acquisition of TE cell fate in the outside cells. [22][23][24] This CDX2 expression in preimplantation embryos is regulated by the Hippo signaling pathway. Hippo signaling is inactivated in the outside cells depending on cell polarity, which enables the downstream transcriptional cofactor YAP to access the nuclei, whereas activated Hippo signaling segregates YAP from nuclei via its phosphorylation in the inside cells, which do not possess polarity. Thus, in the outside cells, nuclear YAP binds to TEAD4 and the YAP/TEAD4 complexes induce Cdx2 expression. [25][26][27][28][29] On the other hand, CDX2 is also important for maintaining the stemness of mouse trophoblasts for correct placental formation, and this CDX2 expression is induced by epiblast-derived FGF4 via the MEK-ERK signaling pathway. [30][31][32][33] Especially, we previously reported that the abnormal expression of Cdx2 after differentiation disrupts the expression of some differentiation marker genes in mouse androgenetic embryoderived trophoblast stem cells (AG-TSCs). 34 Thus, CDX2 possesses dual functions (specification of TE and maintenance of trophoblast stemness) for the morphogenesis of extraembryonic tissues.
Here, we explored the transition of CDX2 expression patterns in mouse peri-implantation (E3.5, 4.0, and 4.5) blastocysts and its regulatory mechanisms to gain new insights into mural TE differentiation for implantation. The results of this study suggest that mural TE differentiation is heteronomously induced by uterus-derived factors secreted depending on estrogen signaling.

| Collection of mouse blastocysts
All mice were purchased from CLEA Japan and maintained in accordance with the Guidelines for the Care and Use of Laboratory

| Culture of mouse blastocysts
The zona pellucida (ZP) of each blastocyst was removed by incu- To inhibit the formation of YAP/TEAD4 complexes, the ZP-removed expanded blastocysts were cultured in KSOM supplemented with verteporfin (17334, Cayman) at 2.5 μM for 4 or 6 h.

| Whole-mount immunofluorescence of mouse embryos
Embryos were fixed in 4% paraformaldehyde for 30 min and per-

| Confocal microscopy and image analysis
Immunofluorescence images were obtained using a confocal laser scanning microscope (LSM710; Carl Zeiss, Oberkochen, Germany).

| Statistical analysis
Results are presented as mean ± S.D. of three or more independent experiments. Statistical analysis was performed with the Student's ttest. p values less than 0.01 were considered statistically significant.  Figure 1A). These results suggested that mural TE started to lose its epithelial property and acquire adhesiveness and invasiveness from E4.0. Actually, the epithelial marker gene E-cadherin was localized at the basolateral surface at E3.5 but was partially dissolved at E4.0 and E4.5 in the mural TE ( Figure 1B). Comparing the fluorescence intensity of CDX2 between polar and mural TE, we found that CDX2 expression in the mural TE gradually decreased from E4.0 and was completely repressed by E4.5, whereas the expression levels were equivalent between polar and mural TE at E3.5 ( Figure 1C,D). These results revealed that CDX2 expression was downregulated in the mural TE from E4.0 with implantation progression. We also clarified that mural TE lost the epithelial property in accordance with CDX2 downregulation.

| CDX2 expression in mural TE of blastocysts cultured in vitro
Next, we examined whether CDX2 is downregulated in the mural TE of blastocysts cultured in vitro. E3.5 blastocysts collected by uterine flushing were cultured in KSOM for 24 or 48 h and used for immunofluorescence of CDX2. To trace PrE specification, PrE marker gene GATA4 was also stained. We selected GATA4 as a PrE marker gene because GATA4 is detected more specifically in PrE-committed cells. 13,15,[36][37][38] Apparently, blastocysts cultured in KSOM were more expanded than E3.5 blastocysts, and their mural TE maintained a smooth epithelial surface. Immunofluorescence revealed that CDX2 expression was maintained in the mural TE of blastocysts cultured in KSOM (Figure 2A,B). On the other hand, whereas no or only a few cells faintly expressing GATA4 were detected in the ICM of E3.5 blastocysts, GATA4-positive cells were clearly detected in the ICM of blastocysts cultured in KSOM for 24 h (Figure 2A). These results indicated that CDX2 downregulation in the mural TE is not induced in culture in KSOM, while the differentiation of ICM into epiblast and PrE proceeded autonomously. It has been known that blastocysts start to attach to and spread on a dish (outgrowth) when they are cultured in medium supplemented with FBS, indicating that FBS stimulates the adhesiveness and invasiveness in TE. [39][40][41][42] In addition, the blastocyst outgrowth assay has been used as an in vitro implantation model. 43 We then cultured E3.5 blastocysts in KSOM with or without 10% FBS for 24, 48, or 72 h after removing ZP, and com-

| Effects of inhibiting ovarian estrogen signaling to CDX2 downregulation in mural TE
On the basis of the above results, we thought that the CDX2 downregulation in the mural TE was heteronomously induced by external factors. It has been known that ovarian estrogen signaling causes the secretion of uterus-derived factors into uterine fluid to induce blastocyst activation to initiate implantation. We thus hypothesized that ovarian estrogen-dependent uterine secretion is a factor in inducing CDX2 downregulation in the mural TE. Therefore, we exam-

| Expression of upstream factors for Cdx2 expression in mural TE during implantation
Recently, the question of when regulatory mechanisms upstream of CDX2 expression is switched to MEK-ERK signaling via epiblastderived FGF4 from Hippo-YAP-TEAD4 signaling has been discussed.
Intriguingly, we detected ESRRB expression only after E4.5 in polar TE, which is regulated by FGF-MEK-ERK signaling in mouse trophoblasts ( Figure 5A). This result is the same regardless of the mouse strain used (data not shown). This suggested that FGF4dependent MEK-ERK signaling is activated from E4.5 but not from E4.0, while CDX2 downregulation in the mural TE is initiated from E4.0. Therefore, we investigated whether CDX2 downregulation in the mural TE is dependent on Hippo signaling activation. It has been known that Hippo signaling inactivation contributes to the dephosphorylated form of YAP, resulting in its nuclear localization and the formation of YAP-TEAD4 complexes directly binding and activating the Cdx2 promoter in preimplantation blastocysts. We then performed immunofluorescence of YAP and TEAD4 and found that both nuclear YAP and TEAD4 in mural TE seemed to be gradually decreased from E3.5 to E4.5 ( Figure 5A). In addition, to assess the contribution of YAP/TEAD4 complexes to CDX2 expression in the blastocyst stage, expanded blastocysts prepared by IVF and in vitro culture were cultured in KSOM supplemented with verteporfin that inhibits the formation of YAP/TEAD4 complexes for 4 or 6 h.
Apparently, whereas control blastocysts remained in the expanded state, verteporfin-treated blastocysts tended to shrink. CDX2 immunofluorescence showed that CDX2 expression decreased significantly in the TE of verteporfin-treated blastocysts compared with control blastocysts (Figure 5B,C). These results suggested that the decline of YAP-TEAD signaling induced CDX2 downregulation in the mural TE and polar TE expressed CDX2 depending on YAP-TEAD signaling rather than FGF-MEK-ERK signaling before E4.5. Together, these results indicated that estrogen signaling-dependent CDX2 downregulation might be mediated by the decline of YAP-TEAD signaling.

| DISCUSS ION
Previous reports indicated that CDX2 is essential for not only cell fate specification toward TE but also maintaining the stemness of trophoblasts in mice. Thus, in this study, we focused on CDX2 to gain new insights into the differentiation of the mural TE to establish implantation. The present study showed that CDX2 was gradually downregulated in the mural TE during implantation in vivo, whereas CDX2 expression was maintained in the mural TE of blastocysts cultured in vitro. On the basis of these results, we propose that the differentiation of the mural TE is induced in vivo for implantation. In addition, our finding that E4.0 blastocysts already exhibited CDX2 downregulation in the mural TE indicated that mural TE differentiation is initiated at least from E4.0. At this stage, blastocysts attach to the uterine luminal epithelium. Therefore, it was thought that the attachment of blastocysts to the uterine luminal epithelium triggers CDX2 downregulation and mural TE differentiation. However, we also found that CDX2 downregulation was induced by FBS supplementation in KSOM without attachment to a dish in vitro. These results suggested that attachment would not be necessary for CDX2 downregulation.
Additionally, CDX2-downregulated mural TE necessarily showed a rough surface not only in vivo but also in vitro. Furthermore, we found basolateral localization of E-cadherin was disrupted in mural TE from E4.0. It has been known that Cdx2 homozygous mutant embryos fail to maintain epithelial integrity by disrupting tight and adherens junctions. 22,44 Therefore, the mural TE loses its epithelial property via CDX2 downregulation in preparation for implantation.
The initiation of implantation is regulated by ovary-derived hormones, especially estrogen. It has been known that mouse blastocysts could not initiate implantation and become dormant when estrogen signaling is pharmacologically or physically inactivated.
Moreover, a previous report indicated that OPN 41,[45][46][47][48] In addition, a previous report indicated that interaction between the RGD motif and integrins activates blastocyst adhesion competence by increasing adhesion complex assembly. 8 The present study revealed that FBS induces CDX2 downregulation in the mural TE, indicating that FBS includes factors that are the same as or analogous to the uterus-derived factors for not only the activation of blastocyst adhesion competence but also CDX2 downregulation in the mural TE. Further investigation of such factors would be beneficial to understanding the optimal environment for implantation.
Lastly, in this study, we tried to identify the upstream signaling for CDX2 downregulation. It has been known that Cdx2 expression is induced by Hippo-YAP-TEAD4 signaling in preimplantation TE and by FGF4-MEK-ERK signaling in postimplantation trophoblasts. Moreover, interestingly, it was revealed that enhancers for Cdx2 expression differ between blastocysts and TSCs. 49 50 Additionally, some reports revealed that ESRRB, whose expression is regulated by FGF4-MEK-ERK signaling in mouse TSCs, is detected in E4.75 polar TE. 52,53 In the present study, ESRRB was detected in E4.5 but not in E3.5 and E4.0 polar TE regardless of the mouse strains used, although CDX2 downregulation in the mural TE was started at least from E4.0. Therefore, we proposed here that the onset of CDX2 downregulation in the mural TE is not attributed to switching the regulatory mechanisms of Cdx2 expression from Hippo-YAP-TEAD to FGF-MEK-ERK pathway. We also immunostained YAP and TEAD4 and found that both YAP and TEAD4 were downregulated in the mural TE after E4.0, but those expression was maintained in the polar TE, as was the case with CDX2 expression. Moreover, the pharmacological inhibition of YAP-TEAD4 binding caused CDX2 downregulation in TE. These results suggested that Cdx2 expression is dependent on Hippo-YAP-TEAD signaling before at least E4.5 in TE, and the decline of YAP-TEAD signaling induces CDX2 downregulation in the mural TE, which contributes to mural TE differentiation.
In conclusion, our results indicated that the differentiation of the mural TE is initiated from peri-implantation stage E4.0. The results F I G U R E 5 Relationship between CDX2 downregulation and Hippo signaling activity in implantation-stage blastocysts. (A) Immunofluorescence analysis of YAP, TEAD4, and ESRRB in E3.5, 4.0, and 4.5 blastocysts. White arrowheads indicate the tips of mural TE. Yellow arrowheads indicate the ESRRB-positive TE. Scale bar: 50 μm. (B) Immunofluorescence analysis of CDX2 in DMSO-(control) or verteporfin (VP)-treated blastocysts. CDX2 expression in TE was found to be weaker in VP-treated blastocysts than in control blastocysts. Scale bar: 100 μm. (C) Comparison of CDX2 fluorescence intensity in the TE of control and VPtreated blastocysts. Statistical significance was determined by the Student's t-test (*p < 0.01). n indicates the number of nuclei analyzed. Bars indicate mean ± S.D.
also indicated the possibility that uterus-derived factors secreted depending on estrogen signaling triggers mural TE differentiation via the decline of YAP-TEAD signaling. These results suggest the significance of the interaction between embryo and mother for TE differentiation toward implantation. In future research, it will be necessary to identify the uterus-derived factors that induce TE differentiation, which contributes to efficient pregnancy success by improving the uterus environment. The authors thank Dr. Toshihiro Konno and Dr. Tamako Matsuhashi for providing helpful comments.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

H U M A N R I G HT S S TATE M E NT S A N D I N FO R M E D CO N S E NT
This article does not describe any experiments involving human participants.

A N I M A L S TU D I E S
This study was approved by the Ethical Committee for Animal Experiment of Tokyo University of Agriculture. All of the institutional and national guidelines for the care and use of laboratory animals were followed.