Interplay of chromatin remodeling BAF complexes in mouse embryonic and epiblast stem cell conversion and maintenance

Abstract

Mouse embryonic stem cells (ESCs) and epiblast stem cells (EpiSCs) are pluripotent stem cells derived from preimplantation and postimplantation embryos, respectively. These cells are capable of interconversion through manipulation of key transcription factors and signaling pathways. While BRG1/BRM-associated factor (BAF) chromatin remodeling complexes are known to play crucial roles in ESC self-renewal and pluripotency, their roles in EpiSCs and their interconversion with ESCs remain unclear. This study demonstrates that the LIF/STAT3 and Wnt signaling pathways, in conjunction with canonical BAF (cBAF) and polycomb repressive complex two complexes, inhibit EpiSC gene expression, thereby preventing ESCs from converting to EpiSCs. Upon removal of LIF, the reduced LIF/STAT3 signaling lifts this inhibition, increasing TGF/nodal pathway activity. Subsequently, the cBAF complex facilitates ESC to EpiSC conversion by promoting EpiSC gene expression. Furthermore, unlike cBAF, inhibition of the ncBAF complex downregulates TGF-β signaling, thereby hindering both ESC to EpiSC conversion and EpiSC maintenance. Moreover, this study revealed the dual mechanisms, methylating histone or non-histone protein STAT3, by which polycomb repressive complex two components participate in the regulation of ESCs to EpiSCs. This research elucidates the interplay between distinct BAF complexes and specific signaling pathways in regulating the conversion and maintenance of ESCs and EpiSCs.

Keywords: BAF complex; ESCs; EpiSCs; LIF/STAT3; TGF-beta.

Figure 1

The cBAF complex collaborates with the LIF/STAT3 pathway to repress the transition ESCs to EpiSCs.A, heat map depicting the expression changes of naïve and primed genes in WT and Dpf2 KO ESCs. The color gradient represents the log2 FPKM (Fragments Per Kilobase Million) values of the genes. B, qPCR analysis of transcript levels for lineage marker genes in ESCs cultured for 24 h in ESC medium, with or without the addition of LIF. C, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3b, Dnmt3a, Lef1, and Pou3f1 genes in ESCs cultured in embryonic stem cell medium without LIF for the specified duration. D, genome browser view of ChIP-seq tracks for STAT3 in ESCs at the Lef1 and Otx2 loci. E, ChIP-qPCR analysis of STAT3 levels at the promoter regions of the Fgf5, Otx2, Dnmt3b, Dnmt3a, Lef1, and Pou3f1 genes in ESCs cultured in ESC medium without LIF for 24 h, followed by an additional 48 h without LIF, with and without 1 μM of 4-OHT treatment. F, quantification of luciferase activity in STAT3-ERT2 ESCs cultured in ESC medium without LIF, with and without 1 μM of 4-OHT, under the control of the basic promoter and the promoters of the Fgf5, Otx2, and Dnmt3a genes. The binding regions of p-STAT3 associated with the Fgf5, Otx2, and Dnmt3a genes were delineated in dark red color, cloned into pGL3-Basic vectors, and used in luciferase assay experiments. Quantification and statistical analysis. ∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001, ∗∗∗ indicates p < 0.001. Error bars represent the standard deviation. 4-OHT, 4-hydroxytamoxifen; cBAF, canonical BAF; ChIP-seq, chromatin immunoprecipitation followed by sequencing; EpiSC, epiblast stem cell; ESC, embryonic stem cell.

Figure 2

cBAF complex and LIF/STAT3 pathway collaboratively repress the transition of ESCs to EpiSCs.A, Western blot analysis depicting protein levels of EED, SUZ12, and EZH2 in ESCs cultured in ESC medium without LIF for the specified duration. GAPDH was used as a loading control. B, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, and Pou3f1 in ESCs cultured in ESC medium in the absence of LIF, with and without Ezh2 overexpression for 48 h. C, heatmap representation of normalized tag density profiles depicting reduced H3K27me3 in ESCs cultured in ESC medium with and without LIF, along with corresponding metaplots illustrating signal intensities. D, genome browser view of ChIP-seq tracks for H3K27me3 at the Fgf5, Otx2 and Dnmt3b loci in ESCs cultured with (Ctrl.) and without LIF (-LIF). E, ChIP-qPCR analysis of H3K27me3 levels at the promoter regions of the Fgf5, Otx2, Dnmt3a, Dnmt3b, and Lef1 genes in ESCs cultured in ESC medium with and without LIF for 48 h. F, Western blot analysis depicting protein levels of STAT3, and p-STAT3 in ESCs cultured in ESC medium without LIF, with and without overexpression of Ezh2 or Eed for 48 h. GAPDH was used as a loading control. G, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, Lef1 and Pou3f1 in ESCs cultured in ESC medium in the absence of LIF, with and without 10 μM of GSK126. H, heatmap illustrating downregulated genes in the absence of LIF that are upregulated upon 10 μM of GSK126 treatment, identified through RNA-seq analysis in ESCs cultured in standard ESC medium (Ctrl.), ESC medium without LIF (-LIF), and ESC medium without LIF with 10 μM of GSK126. I, GO analysis for biological processes associated with genes differentially expressed identified in (H). J, Western blot analysis depicting protein levels of STAT3, and p-STAT3 in ESCs cultured in ESC medium without LIF, with and without 10 μM of GSK126 for 48 h. GAPDH was used as a loading control. ∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001, ∗∗∗∗ indicates p < 0.001. Error bars represent the standard deviation. cBAF, canonical BAF; ChIP-seq, chromatin immunoprecipitation followed by sequencing; EpiSC, epiblast stem cell; ESC, embryonic stem cell; GO, Gene Ontology.

Figure 3

LIF/STAT3 pathway inhibits the expression of EpiSC genes via maintaining the activity of Wnt pathway.A, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, Pou3f1, and Lef1 in ESCs cultured in ESC medium in the absence of LIF, with and without 3 μM of CHIR99021 activator for 48 h. B, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, Pou3f1, and Lef1 in ESCs cultured in ESC medium in the absence of LIF, with and without overexpression of Wnt3a for 48 h. C, qPCR analysis of transcript levels for Wnt3a in ESCs cultured in ESC medium without LIF for the specified duration. D, Western blot analysis depicting protein levels of β-CATENIN in ESCs cultured in ESC medium without LIF for the specified duration. GAPDH was used as a loading control. E, genome browser view of ChIP-seq tracks for β-CATENIN and STAT3 at the Otx2 loci in ESCs. F, ChIP-qPCR analysis of β-CATENIN levels at the promoter regions of the Fgf5, Otx2, Dnmt3a, Dnmt3b, and Pou3f1 genes in ESCs cultured in ESC medium with and without LIF for 48 h. G, Venn diagram depicting the number of genes that are targeted by DPF2, β-CATENIN, and STAT3 from DPF2, β-CATENIN, and STAT3 ChIP-seq analyses. H, genome browser view of ChIP-seq tracks for BRG1, DPF2, β-CATENIN, and STAT3 at the Lef1, Otx2, and Dnmt3a loci in ESCs. I, confirmation of the interaction between FLAG-DPF2 and STAT3 in ESCs through Co-IP followed by Western blot analysis. J, confirmation of the interaction between BRG1 and β-CATENIN in ESCs through Co-IP followed by Western blot analysis. ∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001, ∗∗∗∗ indicates p < 0.001. Error bars represent the standard deviation. ChIP-seq, chromatin immunoprecipitation followed by sequencing; Co-IP, co-immunoprecipitation; EpiSC, epiblast stem cell; ESC, embryonic stem cell; qPCR, quantitative PCR.

Figure 4

The cBAF complex collaborates with the TGF-β pathway to maintain the expression of EpiSC genes.A, qPCR analysis of Fgf5, Otx2, Dnmt3a, Dnmt3b, Pou3f1, and Lef1 transcript levels in ESCs cultured in standard ESC medium (Ctrl.), ESC medium without LIF (-LIF), and Dpf2 KO ESCs cultured in ESC medium without LIF for 48 h. B, qPCR analysis of Fgf5, Otx2, Dnmt3a, and Dnmt3b transcript levels in WT and Dpf2 KO ESCs cultured in the indicated conditions. C, morphology of WT and Dpf2 KO EpiSCs cultured and in AF medium passaged twice. D, ChIP-qPCR analysis of BRG1 levels at the promoter regions of the Fgf5, Otx2, Dnmt3a, Dnmt3b, Lef1, and Pou3f1 genes in ESCs cultured in ESC medium with and without LIF for 48 h. E, qPCR analysis of Lefty1, Lefty2, Pitx2, Gdf1, and Nodal transcript levels in ESCs cultured without LIF for the specified duration. F, qPCR analysis of Fgf5, Otx2, Dnmt3b, Lef1, and Pou3f1 transcript levels in ESCs cultured in the absence of LIF for 48 h with and without 10 μM of SB431542. G, genome browser view of ChIP-seq tracks depicting BRG1, DPF2, β-CATENIN, and STAT3 binding in WT ESCs, SMAD2/3 binding in WT ESCs, and day 3 EBs induced with activin A (ACT) at the Otx2 locus. H, ChIP-qPCR analysis of SMAD2 levels at the promoter regions of the Fgf5, Otx2, Dnmt3b, Dnmt3a, Lef1, and Pou3f1 genes in ESCs cultured in ESC medium with and without LIF for 48 h. I, confirmation of the interaction between BRG1 and SMAD2 in ESCs through Co-IP followed by Western blot analysis. The Western blot for BRG1 was identical to that shown in Figure 3J, as both panels were obtained from the same Co-IP experiment. ∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001, ∗∗∗∗ indicates p < 0.001. Error bars represent the standard deviation. cBAF, canonical BAF; ChIP-seq, chromatin immunoprecipitation followed by sequencing; Co-IP, co-immunoprecipitation; EpiSC, epiblast stem cell; ESC, embryonic stem cell; qPCR, quantitative PCR; TGF, transforming growth factor.

Figure 5

ncBAF complex represses the transition of ESCs to EpiSCs.A, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, and Lef1 in ESCs cultured in ESC medium with and without the addition of 10 μM of I-BRD9. B, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, and Lef1 in ESCs cultured in 2i/LIF ESC medium, AF medium with and without the addition of 10 μM of I-BRD9. C, morphology of ESCs cultured in AF medium and passaged 1 to 3 times with 10 μM of I-BRD9 treatment. The scale bar represents 150 μm. D–I, qPCR analysis of transcript levels for Fgf5, Otx2, Dnmt3a, Dnmt3b, Lef1, and Pou3f1 in ESCs cultured in ESC medium, ESCs cultured and passaged 1 to 3 times in AF medium with and without the addition of 10 μM of I-BRD9. J, qPCR analysis of transcript levels for Lefty1, Lefty2, Pitx2, Gdf1, and Nodal in ESCs cultured in ESC medium with and without 10 μM of I-BRD9. ∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001, ∗∗∗∗ indicates p < 0.001. Error bars represent the standard deviation. ESC, embryonic stem cell; EpiSC, epiblast stem cell; ncBAF, noncanonical BAF; EpiSC, epiblast stem cell; qPCR, quantitative PCR.

Figure 6

Model for the interplay of distinct BAF complexes in embryonic and epiblast stem cell conversion and maintenance.A, the LIF/STAT3 and Wnt/β-Catenin signaling pathways, along with the cBAF and PRC2 complexes, cooperatively suppress the expression of EpiSCs genes. B, the TGF-β signaling pathway, in cooperation with the cBAF and ncBAF complexes, maintains the expression of EpiSC genes. C, the PRC2 complex regulates the expression of EpiSC genes through two mechanisms. cBAF, canonical BAF; EpiSC, epiblast stem cell; ncBAF, noncanonical BAF; PRC2, polycomb repressive complex 2; TGF, transforming growth factor.