Smad4 is essential for epiblast scaling and morphogenesis after implantation, but nonessential before implantation

Abstract

Bone morphogenic protein (BMP) signaling plays an essential and highly conserved role in embryo axial patterning in animal species. However, in mammalian embryos, which develop inside the mother, early development includes a preimplantation stage, which does not occur in externally developing embryos. During preimplantation, the epiblast is segregated from extra-embryonic lineages that enable implantation and development in utero. Yet, the requirement for BMP signaling is imprecisely defined in mouse early embryos. Here, we show that, in contrast to previous reports, BMP signaling (SMAD1/5/9 phosphorylation) is not detectable until implantation when it is detected in the primitive endoderm - an extra-embryonic lineage. Moreover, preimplantation development appears to be normal following deletion of maternal and zygotic Smad4, an essential effector of canonical BMP signaling. In fact, mice lacking maternal Smad4 are viable. Finally, we uncover a new requirement for zygotic Smad4 in epiblast scaling and cavitation immediately after implantation, via a mechanism involving FGFR/ERK attenuation. Altogether, our results demonstrate no role for BMP4/SMAD4 in the first lineage decisions during mouse development. Rather, multi-pathway signaling among embryonic and extra-embryonic cell types drives epiblast morphogenesis postimplantation.

Keywords: Epiblast; Extra-embryonic; Maternal and zygotic gene deletion; Morphogenesis; Mouse; PSMAD1/5/9.

Fig. 1.

BMP signaling becomes active in the primitive endoderm lineage at implantation. (A) SMAD1/5/9 phosphorylation (pSMAD1/5/9) in wild-type CD-1 embryos at E3.75, E4.5, E4.75 and E5.5. In all cases, positive pSMAD1/5/9 signal co-localizes with GATA6 as a marker of PrE and VE. (Asterisks indicate maternal uterine tissue, not part of the embryo shown.) (B) Quantification of total number of pSMAD1/5/9-positive cells in wild-type embryos in A and Fig. S1A. (C) Quantification of the percentage of embryos from A and Fig. S1A that display any pSMAD1/5/9-positive cells versus no pSMAD1/5/9-positive cells. (D) Heat map of the mean normalized expression of BMP pathway genes from scRNA-seq data from Nowotschin et al. (2019). (E) pSMAD1/5/9 in wild-type embryos collected at E2.75 and cultured for 36 h in media containing 300 ng/ml exogenous BMP4. (F) Quantification of the total number of pSMAD1/5/9-positive cells in embryos from E revealed significantly more pSMAD1/5/9-positive cells in BMP4-treated embryos. (G) pSMAD1/5/9 staining is absent in Bmp4 z null embryos at E5.5. (H) Quantification of total number of pSMAD1/5/9-positive cells in wild-type and Bmp4 null embryos at E5.5 revealed significantly fewer pSMAD1/5/9-positive cells in Bmp4 null embryos. White arrowheads indicate positive pSMAD1/5/9 signal. Red arrowhead indicates a GATA6-positive cell, which does not express pSMAD1/5/9. All pairwise comparisons were assessed by one-way analysis of variance (ANOVA) with Tukey's post-hoc test. **P<0.01. ns, not significant. Data are mean±s.d. Scale bars: 10 μm.

Fig. 2.

Maternal and zygotic Smad4 and Bmp4 are dispensable for blastocyst formation and preimplantation cell fate specification. (A) Immunofluorescence for SOX17 and NANOG as respective markers of primitive endoderm (PrE) and epiblast (EPI) in flushed E3.75 wild-type CD-1 embryos and embryos lacking maternal and zygotic Bmp4 (mz null). Quantification did not reveal any significant difference in cell number or cell fate between Bmp4 mz null embryos and controls. (B) Immunofluorescence for SOX17 and NANOG in flushed E4.25 embryos lacking maternal Bmp4 only (m null) and Bmp4 mz null embryos. Quantification did not reveal any significant difference in cell number or cell fate between Bmp4 mz null embryos and controls. (C) Immunofluorescence for SOX17 and NANOG in flushed E3.75 Smad4 m null and Smad4 mz null embryos. Quantification did not reveal any significant difference in cell number or cell fate between Smad4 mz null embryos and controls. (D) Immunofluorescence for SOX17 and NANOG in flushed E4.25 Smad4 m null and Smad4 mz null embryos. Quantification did not reveal any significant difference in cell number or cell fate between Smad4 mz null embryos and controls. ‘Mixed’ indicates co-expression of SOX17 and NANOG. All pairwise comparisons were assessed by unpaired two-tailed Student's t-test. Error bars represent s.d. Scale bars: 10 μm.

Fig. 3.

BMP-independent function of Smad4 is required for post-implantation epiblast organization and maintenance. (A) E4.75 Smad4 mz null and m null embryos stained by immunofluorescence for SOX17 and NANOG. (B) Quantification of epiblast (EPI), primitive endoderm (PrE) and trophectoderm (TE) cell numbers from embryos in A revealed a significant decrease in EPI cells in Smad4 mz null embryos when compared with controls. (C) Quantification of the EPI, PrE and TE cells as a percentage of total cell number from embryos in A revealed a significant decrease in EPI percentage in Smad4 mz null embryos. (D) E5.5 Smad4^−/− embryos stained by immunofluorescence for OCT4 and GATA6 as markers of EPI and VE, respectively. Smad4^−/− refers to combined Smad4 z null and Smad4 mz null embryos. (E) Quantification of the number of OCT4-positive cells in wild-type, Smad4^+/− and Smad4^−/− embryos. (F) Quantification of EPI and PrE cell numbers from Smad4^+/− and Smad4^−/− embryos at E5.5 revealed a specific, significant decrease in EPI cell number in Smad4 mz null embryos when compared with controls (P<0.05, unpaired two-tailed Student's t-test). The difference in VE cell numbers was not significant (P>0.05). (G) Quantification of the proximal-distal length of wild-type, Smad4^+/− and Smad4^−/− embryos at E5.5. (H) Quantification of the proximal-distal length of the EPI of wild-type, Smad4^+/− and Smad4^−/− embryos at E5.5. (I) Quantification of the proximal-distal length of the EPI as a percentage of total length of wild-type, Smad4^+/− and Smad4^−/− embryos at E5.5. (J) Quantification of the proximal-distal length of the EXE of wild-type, Smad4^+/− and Smad4^−/− embryos at E5.5. (K) Quantification of the proximal-distal length of the EXE as a percentage of total length of wild-type, Smad4^+/− and Smad4^−/− embryos at E5.5. (L) Quantification of the proportion of Smad4^+/− and Smad4^−/− embryos with a proamniotic cavity at E5.5. (M) Quantification of the two-dimensional area of the proamniotic cavity of wild-type, Smad4^+/− and Smad4^−/− embryos at E5.5. Cavity area was measured within the plane exhibiting the largest cavity area for each embryo. Comparisons in B, C and F were assessed by unpaired two-tailed Student's t-test. Comparisons in E, G-K and M were assessed by analysis of variance (ANOVA) with Tukey's post-hoc test. **P<0.01, ***P<0.001, ****P<0.0001. ns, not significant. Data are mean±s.d. Scale bars: 10 μm.

Fig. 4.

Inhibition of FGF signaling partially rescues epiblast cavitation in E5.5 Smad4 null embryos. (A) Wild-type (CD-1) and Smad4 z null embryos collected at E5.5 and cultured for 6 h after dissection in control media or media containing FGFR/MEK inhibitors (see Materials and Methods), then stained by immunofluorescence for OCT4 and phosphorylated ERK (pERK). Dashed line in enlargement denotes the proamniotic cavity. (B) Quantification of the proportion of treated and untreated Smad4^−/− embryos with a proamniotic cavity at E5.5. (C) Quantification of the two-dimensional area of the proamniotic cavity of treated and untreated Smad4^−/− embryos at E5.5. Cavity area was assessed on the z-plane with the largest cavity space for each embryo. (D) Quantification of proximal-distal length of the epiblast (EPI) in treated and untreated E5.5 Smad4^−/− embryos. (E) Quantification of proximal-distal length of the EPI as a proportion of total length in treated and untreated E5.5 Smad4^−/− embryos. (F) Quantification of proximal-distal length in treated and untreated E5.5 Smad4^−/− embryos. (G) Quantification of OCT4-positive cell number in treated and untreated E5.5 Smad4^−/− embryos. ***P<0.001. Comparisons in C-G were assessed using unpaired two-tailed Student's t-test.. Data are mean±s.d. Scale bars: 10 μm.

Fig. 5.

Working model: FGF inhibition rescues rosette formation but not embryo growth in Smad4 null embryos. (A) In wild-type embryos, SMAD4 activity inhibits ERK phosphorylation in the extra-embryonic ectoderm (EXE), which allows for rosette formation and cavitation in the epiblast (EPI). BMP4 from the EXE activates SMAD1/5/9 phosphorylation and SMAD4 activity in the visceral endoderm (VE), but this activity is not required for EPI cavitation. (B) In Smad4 null embryos, pERK is upregulated, causing an increase in pERK in the EXE and preventing EPI cavitation. (C) Treatment with FGF inhibitors prevents ectopic upregulation of pERK in Smad4 null embryos, resulting in a small proamniotic cavity. (D) Proposed mechanism for regulation of EPI cavitation by SMAD4.