Morphogenesis of extra-embryonic tissues directs the remodelling of the mouse embryo at implantation

Abstract

Mammalian embryos change shape dramatically upon implantation. The cellular and molecular mechanism underlying this transition are largely unknown. Here, we show that this transition is directed by cross talk between the embryonic epiblast and the first extra-embryonic tissue, the trophectoderm. Specifically, we show via visualisation of a Cdx2-GFP reporter line and pharmacologically mediated loss and gain of function experiments that the epiblast provides FGF signal that results in differential fate acquisition in the multipotent trophectoderm leading to the formation of a tissue boundary within this tissue. The trophectoderm boundary becomes essential for expansion of the tissue into a multi-layered epithelium. Folding of this multi-layered trophectoderm induces spreading of the second extra-embryonic tissue, the primitive endoderm. Together, these events remodel the pre-implantation embryo into its post-implantation cylindrical shape. Our findings uncover how communication between embryonic and extra-embryonic tissues provides positional cues to drive shape changes in mammalian development during implantation.

Fig. 1

Polar trophectoderm transformation. a Schematic for blastocyst to egg cylinder transformation. b Embryos at late pre-implantation blastocyst (E4.5) and early egg cylinder (E5.0–E5.25) stages stained for markers for primitive endoderm (PE)/visceral endoderm (VE) (Gata6) and trophectoderm (TE)/extra-embryonic ectoderm (ExE) (Cdx2). Representative of 20 embryos for each stage. c Embryo at peri-implantation stage (E4.75–E5.0). Polar trophectoderm (pTE) expansion to form the ExE (red outline) precedes PE spreading (white outlines) during blastocyst to egg cylinder transition. Representative of 20 embryos. d Stills from a time lapse movie (Supplementary Movie 1) of a Cdx2-GFP E3.5 blastocyst showing flow of cells from polar to mural TE. Cyan dots: tracking of individual cell nuclei. e Colour coded cell tracks according to number of generations as extracted after single cell tracking from Supplementary Movie 1. Cyan dots: polar TE cells. Grey dots: mural TE cells. f Displacement map extracted after single cell tracking from Supplementary Movie 1. g Polar TE displacement map extracted after single cell tracking from Supplementary Movie 1. h Stills from a time lapse movie of Cdx2-GFP E4.5 implanting blastocyst combined with the displacement map (magenta: polar TE; white: mural TE) showing stop of cell flow from polar to mural TE upon implantation. i Stills from a time lapse movie of Cdx2-GFP E4.5 implanting blastocyst combined with single cell tracking (magenta: polar TE; cyan: mural TE) showing stop of cell flow from polar to mural TE upon implantation. j Model of polar TE cell flow in pre- and post implantation blastocysts. Scale bars = 20 µm

Fig. 2

Establishment of tissue boundary within the trophectoderm. a Maximum intensity projection (MIP) image of an implanted E4.75 blastocyst. Arrowheads indicate tissue boundary between polar and mural TE. Representative example of 20 embryos. b MIP image of an implanted E4.75 blastocyst. Arrowheads indicate actomyosin enrichment at the boundary between polar and mural TE. Representative example of eight embryos. c Stills from a time lapse movie of Cdx2/E-Cad GFP E4.5-implanting blastocyst. Red line follows the boundary between high-Cdx2 cells (polar TE) and low-Cdx2 cells (mural TE). As development progresses, polar and mural TE cells segregate and a linear tissue boundary formed (arrowheads). d Quantification of cell–cell junction angle at the polar/mural TE boundary and within polar TE. Two-sided unpaired student’s t test; ****P < 0.0001; mean ± SEM. n = 56 polar/mural TE and 63 polar TE cell junctions. Source data are provided as a Source Data file. Scale bars = 20 µm

Fig. 3

Cell shape changes and polarized cell divisions drive polar TE expansion. a Stills from a time lapse movie of Lifeact-GFP blastocyst. Red arrowheads indicate the polar/mural TE boundary. Boundary formation is followed by polar TE expansion. b Quantification of polar TE cell aspect ratio at different developmental stages. Two-sided unpaired student’s t test; ****P < 0.0001; mean ± SEM. Pre-implantation blastocyst: n = 216 cells; Implanted blastocyst: n = 152 cells; Transitioning blastocyst; n = 75 cells. c Quantification of polar TE apical cell surface are. Two-sided unpaired student’s t test; ****P < 0.0001; mean ± SEM. Blastocyst before: n = 186 cells; Transitioning blastocyst; n = 163 cells. d Stills from a time lapse movie of Lifeact-GFP implanted blastocyst. Polar TE expansion is evident at 2.5 h upon polar TE cell shape changes and polarised cell division. e Representative examples of polar TE cells dividing parallel to Embryonic(Em)/Abembryonic(Ab) axis during polar TE expansion. f Rose diagram for quantification of cell division orientation in blastocysts before (green) and during (purple) polar TE expansion. Kolmogorov–Smirnov test: ****p < 0.0001. Blastocyst before polar TE expansion: n = 24, Bastocyst during polar TE expansion: n = 36. Source data are provided as a Source Data file. Scale bars = 20 µm

Fig. 4

Tissue boundary is necessary for polar trophectoderm expansion. a Implanted blastocyst (E4.75) stained for phosphorylated ERK (pERK). pERK is detected in the primitive endoderm (PE) and the trophectoderm (TE). pERK is restricted to the polar region of the TE. b Implanted blastocyst (E4.75) stained for phosphorylated AKT (pAKT). pAKT is restricted to the polar region of the TE. For a and b, representative example of 10 embryos. c Schematic representation of the experimental design to examine the contribution of actomyosin contractility and FGF signalling in the formation of polar/mural TE tissue boundary formation. d Representative examples of control, Y27632 (ROCK inhibitor) and SU5402 (Fgfr inhibitor) treated blastocysts cultured as described in c and analysed for the presence of polar/mural TE boundary (cyan dotted line). Polar/mural tissue boundary formation is defective in the absence of actomyosin contractility and FGF signalling. e Quantification of polar/mural TE boundary formation efficiency in control, Y27632 and SU5402-treated blastocysts. χ² test; ****P < 0.0001, mean ± SEM. For d and e FGF n = 29 control, 15 Y27632 treated and 15 SU5402-treated embryos. f Quantification of cell–cell junction angles at the polar/mural TE interface. The angles of junctions after abrogation of actomyosin contractility (Y27632) and signalling (SU5402) are narrowed in agreement with the defective formation of polar/mural TE tissue boundary in these conditions. Two-sided unpaired student’s t test; ****P < 0.0001; mean ± SEM. Control: n = 57; Y27632: n = 38; SU5402: n = 45. g Representative examples of control, Y27632 (ROCK inhibitor) and SU5402 (FGFR inhibitor)-treated blastocyst cultured as described in b and analysed for polar TE (white outline) expansion. h Quantification polar TE expansion efficiency in control, Y27632, and SU5402-treated blastocysts. χ² test; ****P < 0.0001. For f and g n = 29 control, 15 Y27632 treated and 15 SU5402-treated embryos. i Model for the molecular and cellular events necessary for polar TE expansion during peri-implantation morphogenesis. Source data are provided as a Source Data file. Scale bars = 20 µm

Fig. 5

Primitive endoderm behaviour during the blastocyst to egg cylinder transition. a Examples of embryos used for analysis of primitive endoderm (PE) cell shape at different stages of peri-implantation development. Representative of 20 embryos for each stage. b Quantification of PE cell aspect ratio at different developmental stages. The red lines between different columns correspond to events taking place within the trophectoderm (TE). Two-sided unpaired student’s t test; ****P < 0.0001; mean ± SEM. Pre-implantation blastocyst: n = 117 cells; Implanted blastocyst: n = 219 cells; Transitioning blastocyst: n = 117 cells; early egg cylinder: n = 184 cells. c Stills from a time lapse movie of Lifeact-GFP E4.5-implanting blastocyst. PE (cyan) acquires a columnar morphology at implantation stages and this is followed by tissue spreading through cell-shape changes upon polar TE (red) expansion as quantified in b. Red double-headed arrows: polar TE expansion. Cyan arrows: PE spreading. Stills without coloured overlay are presented in Supplementary Fig. 7b. d Representative examples of implanting blastocyst (E4.5–E4.75). Parietal endoderm migration is completed before the expansion of the TE (double-headed arrows in bottom row) and before PE (red outline in 2D images) spreading over the lateral sides of the epiblast (bottom row). Red arrows indicate the front of the migrating parietal endoderm. e Representative examples of transitioning blastocysts (E5.0) displaying expanded polar trophectoderm and early egg cylinders. Primitive endoderm cells are found enclosed between the embryonic and Reichert’s basement membrane during blastocyst to egg cylinder transformation. arrowheads: Reichert’s membrane. Arrows: embryonic basement membrane. Dots: proximal cells. Representative images of 20 embryos per stage. Source data are provided as a Source Data file. Scale bars = 20 µm

Fig. 6

Trophectoderm folding orchestrates the final step the blastocyst to egg cylinder transformation. a Stills from a time lapse movie of Lifeact-GFP blastocyst. ExE tissue (red) folding (white arrows) through apical constriction is followed by spreading (yellow arrows) of the visceral endoderm (VE) (yellow outline) and formation of egg cylinder. Stills without coloured overlay are presented in Supplementary Fig. 10a. b Quantification of ExE apical cell surface area, VE apical cell surface area and VE length through time during blastocyst to egg cylinder transformation from three individual embryos shown in Supplementary Movie 8. Cyan boxes indicate the developmental period, during which ExE cell’s apical constriction precedes VE spreading. VE tissue spreading during egg cylinder formation is owing to cell-shape changes as quantified in Fig. 5b. c Quantification of polar TE/ExE apical cell surface area relative to PE/VE position. d Quantification of PE/VE cell aspect ratio relative to PE/VE position. e Quantification of PE/VE cell aspect ratio relative to polar TE/ExE apical cell surface area. f Quantification of polar TE/ExE cell aspect ratio relative to PE/VE cell aspect ratio. For c–f, Pearson correlation test was used, ****p < 0.0001. g Rose diagram for quantification of long cell axis orientation in distal and proximal VE at the early egg cylinder stage. Kolmogorov–Smirnov test: ****p < 0.0001. Distal VE: n = 153 cells, Proximal VE: n = 75 cells. h Representative examples of control and Y27632 (ROCK inhibitor)-treated embryos. Embryos were recovered at E4.75 and cultured ex vivo for 12 hours. In the absence of actomyosin contractility (Y27632) egg cylinder formation fails due to defective ExE folding (Supplementary Movie 10). i Quantification of blastocyst to egg cylinder transformation efficiency in control, and Y27632 treated embryos. χ² test; ****P < 0.0001. For h and i n = 16 control and 29 Y27632-treated embryos. j Model of blastocyst to egg cylinder transformation as a result of TE morphogenesis. Source data are provided as a Source Data file. Scale bars = 20 µm