VGLL1 cooperates with TEAD4 to control human trophectoderm lineage specification

Abstract

In contrast to rodents, the mechanisms underlying human trophectoderm and early placenta specification are understudied due to ethical barriers and the scarcity of embryos. Recent reports have shown that human pluripotent stem cells (PSCs) can differentiate into trophectoderm (TE)-like cells (TELCs) and trophoblast stem cells (TSCs), offering a valuable in vitro model to study early placenta specification. Here, we demonstrate that the VGLL1 (vestigial-like family member 1), which is highly expressed during human and non-human primate TE specification in vivo but is negligibly expressed in mouse, is a critical regulator of cell fate determination and self-renewal in human TELCs and TSCs derived from naïve PSCs. Mechanistically, VGLL1 partners with the transcription factor TEAD4 (TEA domain transcription factor 4) to regulate chromatin accessibility at target gene loci through histone acetylation and acts in cooperation with GATA3 and TFAP2C. Our work is relevant to understand primate early embryogenesis and how it differs from other mammalian species.

Fig. 1. VGLL1 is highly expressed during human TELC induction from naive PSCs.

a Schematic depicting the generation of TELC-D5 from 4CL H9 ESCs with TE medium and the generation of TSCs from TELC-D5 cells with TSC medium. VGLL1/TEAD4 knockdown (KD) and multiple types of sequencing were performed as indicated. b Phase contrast images of 4CL H9 ESCs and TELCs at the indicated time points. Scale bar, 40 µm. Representative of three independent experiments. c Immunostaining images for pluripotency (NANOG and KLF17), naïve/TE shared (TFAP2C and TEAD4), and TE-enriched (GATA3 and VGLL1) genes in 4CL H9 ESCs (upper panels; scale bar, 50 µm) and TELC-D5 cells (lower panels; scale bar, 100 µm). Nuclei were counterstained with DAPI (blue). Representative of three independent experiments. d Heatmap showing the expression of pluripotency and TE genes in bulk RNA-seq for the indicated conditions. Example genes for each cluster are shown in the boxes. n = 2 biological replicates. e UMAP comparing the human embryonic day 4 (E4) to E14 stages with 4CL H9 ESCs and TELC-D5 cells. All reference datasets used in this study are summarized in Methods. f Bubble plot showing the frequency of expression and scaled average expression of representative genes in, 4CL H9 ESCs, TELC-D5 cells and human embryo E4-E14 datasets. TE, trophectoderm; CTB, cytotrophoblast; EVT, extravillous trophoblast; SCT, syncytiotrophoblast. g Volcano plot showing DEGs between 4CL H9 naïve ESCs and TELC-D5 cells. DEGs higher in TELC-D5 cells (log₂(fold change) > 1) are shown in red. P value was calculated using the Wald test and adjusted for multiple testing using the Benjamini-Hochberg correction. h RT-qPCR showing the expression of VGLL1 in 4CL H9 ESCs and TELC-D5 cells. Data are presented as the mean ± standard error of the mean (SEM) of the fold-change compared to naive ESCs. n = 3 biological replicates. P value was calculated using a two-tailed unpaired Student’s t-test, ***P < 0.001. i Western blotting analysis for the indicated proteins in 4CL H9 ESCs, TELC-D5 cells, and TSCs derived from TELC-D5 cells. Representative of three independent experiments.

Fig. 2. Expression kinetics of VGLL1 in mammalian preimplantation embryos.

a UMAP visualization showing the different cell types in the human blastocyst at E5. Dots are colored by cell type. Cell types were annotated by cluster-specific gene expression patterns. b Violin plot showing the log-normalized expression of TE genes in different cell types of the human blastocyst at E5. c Left panel: the trajectory of the different cell types in the human blastocyst at E5 reveals two branches: pre-lineage to the TE branch and pre-lineage to the ICM branch. Right panel: colors from dark blue to light blue indicate progression through the pseudotime. d Expression patterns of TE genes along the pseudotime trajectory of cells in the human blastocyst at E5. e UMAP visualization showing the different cell types in the monkey blastocyst at E6-E9. Dots are colored by cell type. Cell types were annotated by cluster-specific gene expression patterns. f Violin plot showing the log-normalized expression of TE genes in different cell types of the monkey blastocyst at E6-E9. g Left panel: the trajectory of the different cell types in the monkey blastocyst (E6-E9) reveals two branches: pre-lineage to the TE branch and pre-lineage to the EPI branch. Right panel: colors from dark blue to light blue indicate progression through the pseudotime. h Expression patterns of TE genes along the pseudotime trajectory of cells in the monkey blastocyst at E6-E9. i UMAP visualization showing the different cell types at the mouse 16-cell stage. Dots are colored by cell type. Cell types were annotated by cluster-specific gene expression patterns. j Violin plot showing the log-normalized expression of TE genes in different cell types at the mouse 16-cell stage. k Left panel: the trajectory of the different cell types at the mouse 16-cell stage reveals two branches: pre-lineage to the TE branch and pre-lineage to the ICM branch. Right panel: colors from dark blue to light blue indicate progression through the pseudotime. l Expression patterns of TE genes along the pseudotime trajectory of cells at the mouse 16-cell stage.

Fig. 3. VGLL1 is required for human TELC induction from naive PSCs.

a RT-qPCR showing the VGLL1 knockdown efficiency for 4CL H9 ESCs transduced with shVGLL1−2 or shVGLL1−4 compared to the shLuc control at day 5 of TELC differentiation. Data are presented as the mean ± SEM, n = 3 biological replicates. P value was calculated using a two-tailed unpaired Student’s t-test, ***P < 0.001. b Western blotting analysis of the indicated proteins for 4CL H9 ESCs transduced with shVGLL1−2 or shVGLL1−4 compared to the shLuc control at day 5 of TELC differentiation. Representative of three independent experiments. c Representative phase contrast images of 4CL H9 ESCs transduced with shLuc (control) or shVGLL1 (2 and 4) at day 5 of TELC differentiation. Scale bar, 100 µm. Representative of three independent experiments. d Analysis of cell numbers for 4CL H9 ESCs transduced with shLuc, shVGLL1−2 or shVGLL1−4 throughout the TELC induction time course. Data are presented as the mean ± SEM. n = 3 biological replicates. P value was calculated using a two-tailed unpaired Student’s t-test, ***P < 0.001, **P < 0.01. e Heatmap showing the expression of pluripotency, cell cycle and TE genes in 4CL H9 ESC-derived cells from the indicated conditions. Example genes are shown for each cluster in the boxes. n = 2 biological replicates. f RT-qPCR showing the expression of TE-related genes for 4CL H9 ESCs transduced with shVGLL1−2 and shVGLL1−4 compared to the shLuc control at day 5 of TELC differentiation. Data are presented as the mean ± SEM. n = 3 biological replicates. P value was calculated using a two-tailed unpaired Student’s t-test, ***P < 0.001, **P < 0.01, *P < 0.05. g Enriched GO terms for downregulated genes in 4CL H9 ESCs transduced with shVGLL1 compared to shLuc control at day 5 of TELC differentiation. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. h Enriched GO terms for upregulated genes in 4CL H9 ESCs transduced with shVGLL1 compared to shLuc control at day 5 of TELC differentiation. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. i TA cloning followed by Sanger sequencing results showing homozygous deletion for VGLL1-knockout clones [clone 45 (C45) and clone 68 (C68)]. WT: wild-type. j Western blotting analysis of VGLL1 expression in H9 WT and VGLL1-knockout clones at day 5 of TELC induction. Representative of three independent experiments. k Representative phase contrast images of H9 WT and VGLL1-knockout clones at day 5 of TELC induction. Scale bar, 100 µm. Representative of three independent experiments. l Analysis of cell numbers for H9 WT and VGLL1-knockout clones at day 5 of TELC induction. Data are presented as the mean ± SEM. n = 3 biological replicates. P value was calculated using a two-tailed unpaired Student’s t-test, ***P < 0.001. m Heatmap showing the expression of pluripotency, cell cycle and TE genes in bulk RNA-seq of H9 VGLL1-knockout clones compared to WT at day 5 of TELC induction. Example genes are shown for each cluster in the boxes. n = 2 biological replicates. n Enriched GO terms for downregulated genes in H9 VGLL1-knockout clones compared to WT at day 5 of TELC induction. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction.

Fig. 4. VGLL1 interacts with TEAD4 to regulate human TELC induction from naive PSCs.

a Pie chart showing the distribution of VGLL1-binding sites throughout the annotated genomic locations in CUT&Tag of TELC-D5 cells differentiated from 4CL H9 ESCs. b Venn diagram showing the overlap between downregulated genes in shVGLL1 compared to shLuc control and genes bound by VGLL1 at promoter sites in TELC-D5 differentiated from 4CL H9 ESCs. c Enriched GO terms for genes bound by VGLL1 in TELC-D5 cells differentiated from 4CL H9 ESCs. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. d Motif analysis using HOMER showing the significantly enriched DNA-binding motifs at VGLL1-bound sites in TELC-D5 cells differentiated from 4CL H9 ESCs. P value was calculated using a hypergeometric test (one-sided). e Immunoprecipitation using lysates from TELC-D5 cells differentiated from 4CL H9 ESCs with anti-VGLL1 (left panel) or anti-TEAD4 (right panel) and subsequent Western blotting analysis with anti-TEAD4, anti-VGLL1, anti-YAP and anti-WWTR1. Representative of three independent experiments. f Pie chart showing the distribution of TEAD4-binding sites throughout the annotated genomic locations in TELC-D5 cells differentiated from 4CL H9 ESCs. g Pie chart showing the distribution of YAP-binding sites throughout the annotated genomic locations in TELC-D5 cells differentiated from 4CL H9 ESCs. h Venn diagram showing the overlap between VGLL1-, TEAD4- and YAP-bound sites in CUT&Tag of TELC-D5 cells differentiated from 4CL H9 ESCs. Representative cell cycle and TE genes corresponding to the overlapping regions are shown. i Co-occupancy analysis by signal density pileups of VGLL1, TEAD4 and YAP peaks in CUT&Tag of TELC-D5 differentiated from 4CL H9 ESCs. GO analysis for each group is shown. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction.

Fig. 5. VGLL1 regulates histone acetylation levels and chromatin accessibility at TE target loci.

a Western blotting analysis for the indicated histone marks in 4CL H9 ESCs transduced with shLuc, shVGLL1−2 or shVGLL1−4 at day 5 of TELC differentiation. Representative of three independent experiments. b Venn diagram showing the overlap between VGLL1- and TEAD4-bound sites and H3K27ac-enriched sites in CUT&Tag of TELC-D5 cells differentiated from 4CL H9 ESCs. Representative cell cycle and TE genes corresponding to the overlapping regions are shown in the boxes. c Enriched GO terms for VGLL1, TEAD4 and H3K27ac co-occupied genes in CUT&Tag of TELC-D5 differentiated from 4CL H9 ESCs. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. d Co-occupancy analysis by signal density pileups of TEAD4 CUT&Tag peaks in 4CL H9 ESCs and TEAD4 CUT&Tag peaks in TELC-D5 cells differentiated from them. Peaks were divided into three categories: naïve-specific, shared and TE-specific. Example genes are shown for each category. e Signal intensity of H3K27ac in 4CL H9 ESCs (Naïve-H3K27ac) and TELC-D5 cells (TELC-D5-H3K27ac), and VGLL1 in TELC-D5 cells (TELC-D5-VGLL1) from the three categories defined in d. f Venn diagram showing the overlap between VGLL1- and TEAD4-binding at H3K27ac-enriched sites in CUT&Tag with chromatin accessibility peaks (ATAC-seq) in TELC-D5 cells differentiated from H9 4CL naïve ESCs. Representative cell cycle and TE genes corresponding to the overlapping regions are shown in the boxes. g Co-occupancy analysis by signal density pileups of VGLL1, TEAD4, H3K27ac CUT&Tag peaks and chromatin accessibility (ATAC-seq) peaks in TELC-D5 cells differentiated from 4CL H9 ESCs. h Enriched GO terms for open chromatin genes (ATAC-seq) co-occupied by VGLL1, TEAD4 and H3K27ac (CUT&Tag) of TELC-D5 cells differentiated from 4CL H9 ESCs. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. i Genome browser tracks showing VGLL1, TEAD4, H3K27ac genomic enrichment peaks (CUT&Tag) and chromatin accessibility peaks (ATAC-seq) for representative TE gene loci of TELC-D5 cells differentiated from 4CL H9 ESCs. j Signal intensity of TEAD4 at VGLL1-bound sites in WT and VGLL1-KO clone 68 (VGLL1^-/--C68) at day 5 of TELC induction from 4CL H9 ESCs. k Signal intensity of H3K27ac at VGLL1-binding sites in WT and VGLL1-KO clone 68 (VGLL1^-/--C68) at day 5 of TELC induction from 4CL H9 ESCs. l Signal intensity of chromatin openness at VGLL1-binding sites in WT and VGLL1-KO clone 68 (VGLL1^-/--C68) at day 5 of TELC induction from 4CL H9 ESCs.

Fig. 6. VGLL1 is necessary for sustaining the human TSC identity.

a Phase contrast images of 4CL H9 ESC-derived TSCs transduced with shLuc, shVGLL1-2 or shVGLL1-4. Scale bar, 100 µm. Representative of three independent experiments. b Analysis of cell numbers for 4CL H9 ESC-derived TSCs transduced with shLuc, shVGLL1-2 or shVGLL1-4. Data are presented as the mean ± SEM, n = 3 biological replicates. P value was calculated using a two-tailed unpaired Student’s t-test, ***P < 0.001, **P < 0.01, *P < 0.05. c Heatmap showing the expression of pluripotency, cell cycle and TSC genes for the indicated conditions. Example genes are shown for each cluster in the boxes. n = 2 biological replicates. d Enriched GO terms for downregulated genes in shVGLL1 compared to shLuc control in 4CL H9 ESC-derived TSCs. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. e Immunoprecipitation using 4CL H9 ESC-derived TSC lysates with anti-VGLL1 (left panel) or anti-TEAD4 (right panel) and subsequent Western blotting with anti-TEAD4, anti-VGLL1, anti-YAP and anti-WWTR1. Representative of three independent experiments. f Western blotting for the indicated histone marks for shVGLL1-2 and shVGLL1-4 compared to the shLuc control in 4CL H9 ESC-derived TSCs. Representative of three independent experiments. g Genome browser tracks showing VGLL1, TEAD4, H3K27ac genomic enrichment peaks (CUT&Tag) and chromatin accessibility peaks (ATAC-seq) for representative TSC genes loci in 4CL H9 ESC-derived TSCs. h Co-occupancy analysis by signal density pileups of VGLL1, TEAD4 binding in 4CL H9 ESC-derived TELC-D5 cells and TSCs. The three major groups are: group 1 (common between TELC-D5 and TSC), group 2 (TELC-specific) and group 3 (TSC-specific). For clarity, other groups are not labelled. i Enriched GO terms for groups 1, 2 and 3 of panel h. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. j Genome browser tracks showing VGLL1, TEAD4 and H3K27ac genomic enrichment peaks (CUT&Tag) in 4CL H9 ESC-derived TELC-D5 cells and TSCs for representative gene loci from group 2 (upper panel) and group 3 (lower panel).

Fig. 7. VGLL1/TEAD4 cooperate with GATA3 and TFAP2C to regulate TE identity.

a Venn diagram showing the pairwise overlap between VGLL1-, TEAD4-, GATA3- and TFAP2C-bound sites in 4CL H9 ESC-derived TELC-D5 cells. b Four-way venn diagram showing the overlap between VGLL1-, TEAD4-, GATA3-, and TFAP2C- bound sites in 4CL H9 ESC-derived TELC-D5 cells. c Enriched GO terms for genes in Group 1 of panel b. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. d Enriched GO terms for genes in group 2 of panel b. P value was calculated using a hypergeometric test (one-sided) and adjusted for multiple testing using the Benjamini-Hochberg correction. e Motif analysis using HOMER showing the significantly enriched DNA-binding motifs at GATA3-binding sites in 4CL H9 ESC-derived TELC-D5 cells. P value was calculated using a hypergeometric test (one-sided). f Motif analysis using HOMER showing the significantly enriched DNA-binding motifs at TFAP2C-binding sites in 4CL H9 ESC-derived TELC-D5 cells. P value was calculated using a hypergeometric test (one-sided). g Genome browser tracks showing VGLL1, TEAD4, GATA3 and TFAP2C binding in 4CL H9 ESC-derived TELC-D5 cells for representative TE loci.