DNMTs and SETDB1 function as co-repressors in MAX-mediated repression of germ cell-related genes in mouse embryonic stem cells

Abstract

In embryonic stem cells (ESCs), the expression of development-related genes, including germ cell-related genes, is globally repressed. The transcription factor MAX represses germ cell-related gene expression in ESCs via PCGF6-polycomb repressive complex 1 (PRC1), which consists of several epigenetic factors. However, we predicted that MAX represses germ cell-related gene expression through several additional mechanisms because PCGF6-PRC1 regulates the expression of only a subset of genes repressed by MAX. Here, we report that MAX associated with DNA methyltransferases (DNMTs) and the histone methyltransferase SETDB1 cooperatively control germ cell-related gene expression in ESCs. Both DNA methylation and histone H3 lysine 9 tri-methylation of the promoter regions of several germ cell-related genes were not affected by knockout of the PRC1 components, indicating that the MAX-DNMT and MAX-SETDB1 pathways are independent of the PCGF6-PRC1 pathway. Our findings provide insights into our understanding of MAX-based repressive mechanisms of germ cell-related genes in ESCs.

Fig 1. Contribution of PCGF6-PRC1 to MAX-mediated gene repression.

(A) ChIP-qPCR analyses of Max-null ESCs (Dox+) and control ESCs (Dox−) using anti-RING1B antibody or control IgG in the promoter region of the late PGC markers and of hemoglobin β (Hbb-b1) as a negative control of MAX localization. Relative ratio of immunoprecipitated chromatin to input chromatin (% of input) was determined using real-time PCR. Data are presented as relative % of input normalized against % of input using anti-RING1B antibody in Hbb-b1 in control ESCs. The bar graph represents mean ± standard error of the mean (SEM) of two independent experiments. (B) Expression changes of 85 germ cell–related up-regulated genes (n = 3, > 2-fold change, one-way analysis of variance [ANOVA] P < 0.05) in Max-KD or L3mbtl2-KD ESCs compared with control ESCs based on our microarray data (GSE45181) [2]. Fold-change in the expression is shown by a box-whisker plot. The lines inside the boxes show the median. The whiskers indicate the minimum and maximum. ***P < 0.001 (Mann-Whitney U-test). (C) Fold-change in the expression of significantly up-regulated germ cell–related genes (top 22 of 85 germ cell-related upregulated genes in Fig 1B) in Max-KD and L3mbtl2-KD ESCs. The expression in control ESCs was set as 1.0 (dotted line). (D) Relative expression of the late PGC markers in Ring1a/b-DKO ESCs (4OHT+) determined by qRT-PCR. The expression in control ESCs (4OHT−) was set as 1.0. Values are plotted as mean ± SEM of 3 biological replicates. n.s.: not significant, *P < 0.05, **P < 0.01, ***P < 0.001 (Student’s t-test).

Fig 2. MAX-mediated repression of germ cell–related genes through DNA methylation.

(A) Samples immunoprecipitated using anti-MAX antibody or control IgG were analyzed by Western blotting using anti-DNMT antibodies. Principally, the same result was obtained in two independent experiments. The un-cropped data of these images are shown in S4A Fig. (B) Levels of CpG methylation of genes with a DMR in close proximity (±300 bp) for Max-KD ESCs and control ESCs from TMS. The ratios of methylated CpGs in the regions ±300 bp of the genes are shown. Values are plotted as mean ± SEM of 2 biological replicates. (C) GO analysis of 17 genes with DMRs hypomethylated in Max-KD ESCs compared with control ESCs. GO terms with corrected P value < 0.05 (top 7) are shown. (D) Motif analyses of 17 DMRs showed significant enrichment of E-box–like sequences. Motif sequences with the lowest q value (top 2) are shown. (E) DNA methylation status of the promoter regions of the late PGC markers in control and Max-KD ESCs, as determined by bisulfite sequencing. The filled and open circles indicate methylated- and un-methylated CpGs, respectively. The data shown were combined from two independent experiments. The percentage of methylated CpGs is indicated. (F) Box-whisker plots of the CpG methylation levels shown in Fig 2E. The lines inside the boxes show the median. The whiskers indicate the minimum and maximum. ***P < 0.001 (Mann-Whitney U-test).

Fig 3. MAX-mediated repression of germ cell–related genes through H3K9me3.

(A) Samples immunoprecipitated using anti-MAX antibody or control IgG were analyzed by Western blotting using anti-SETDB1 antibody. Principally, the same result was obtained in two independent experiments. The un-cropped data of this image is shown in S5B Fig. (B, C) ChIP-qPCR analyses of Max-null ESCs (Dox+) and control ESCs (Dox−) using anti-H3K9me3 antibody (B), anti-SETDB1 antibody (C) or control IgG. The data are displayed in the same way as in Fig 1A. (D) KD efficiency of Setdb1 in ESCs at day 2 post-siRNA treatment, as determined by RT-qPCR. (E) Relative expression of the late PGC markers in Setdb1-KD ESCs, as determined by qRT-PCR. The expression in control ESCs was set as 1.0. Values are plotted as the mean ± SEM of 3 biological replicates. *P < 0.05, **P < 0.01, ***P < 0.001 (Student’s t-test).

Fig 4. RING1A/B-independent DNA and H3K9 methylation, and a cooperative effect of Dnmts-TKO and Setdb1-KD on repression of the late PGC markers.

(A) DNA methylation status of the promoter regions of Stra8 and Sycp3 in control (4OHT−) and Ring1a/b-DKO ESCs (4OHT+) determined by bisulfite sequencing. The data are displayed in the same way as in Fig 2E. (B) ChIP-qPCR analyses of Ring1a/b-DKO ESCs (4OHT+) and control ESCs (4OHT−) using anti-H3K9me3 antibody or control IgG. The data are displayed in the same way as in Fig 1A. (C) KD efficiency of Max in Dnmts-TKO ESCs at day 3 post-siRNA treatment, as determined by RT-qPCR. (D) Relative expression of the late PGC markers in Max-KD/Dnmts-TKO ESCs, as determined by qRT-PCR. The expression in Dnmts-TKO ESCs with control siRNA treatment was set as 1.0. (E) KO and KD efficiency of Max (left panel) and Setdb1 (right panel) in Setdb1-KD/Max-null ESCs at day 3 post-siRNA treatment, as determined by RT-qPCR. KD efficiency of Setdb1 is normalized with the expression of Setdb1 in Max-null ESCs with control siRNA treatment. (F) Relative expression of the late PGC markers in Setdb1-KD/Max-KO ESCs, as determined by qRT-PCR. The expression in Setdb1-KD ESCs with Max expression (Dox-) was set as 1.0. (G) KD efficiency of Setdb1 in Dnmts-TKO ESCs at day 2 post-siRNA treatment, as determined by RT-qPCR. (H) Relative expression of the late PGC markers in Setdb1-KD/Dnmts-TKO ESCs, as determined by qRT-PCR. The expression in Dnmts-TKO ESCs with control siRNA treatment was set as 1.0. Values are plotted as the mean ± SEM of 3 biological replicates. *P < 0.05, **P < 0.01, ***P < 0.001 (Student’s t-test).

Fig 5. Fractionation of MAX-interacting complexes.

(A) Schematic illustration of the fractionation of ESC nuclear extracts by column chromatography. Nuclear extracts were prepared from VV3 ESCs and fractionated into fractions A-D on a heparin sepharose column. Fractions A-D were further fractionated into fractions I-IV on a Q sepharose column. (B to E) Samples immunoprecipitated using anti-MAX antibody or control IgG for fraction A-III (B), B-III (C), C-III (D), or D-III (E) were analyzed by Western blotting using anti-DNMT, anti-RING1B, or anti-MAX antibodies. Principally, the same result was obtained in two independent experiments. The un-cropped data of these images are shown in S6A–S6D Fig, respectively.