PMC

RYBP represses endogenous retroviruses. (A) Upregulation of MuERV mRNAs among those encoded by LTR-retroelements measured by RT-qPCR of ES cells after 4 days of 4′-OHT treatment. (B and C) Schematic representation of MuERV retroelements and expression. RT-qPCR expression analysis of indicated regions. LTR, long terminal repeat; Gag, Pro-pol, and dUTPase, gene region encoding a Gag-Pro dUTPase polyprotein; PBS, primer binding site; PPT, polypurine tract. Means and SDs of three experiments are shown. (D) Lsd1 levels in wild-type and RYBP-mutant ES cells.

RYBP binding to chromatin in Dnmt1-deficient ES cells and effects of RYBP depletion on DNA methylation. (A) Enrichment profiles of RYBP bound to methylated promoters (Mael, Dazl, Mov10l1, and Ddx4, upregulated in RYBP-KO ES cells) and unmethylated promoters (Lef1, Bmp6, and Hmx3, not affected by RYBP deletion). Red sectors indicate CGI locations, and TSSs are depicted by arrows. RYBP enrichment in Dnmt1-mutant cells colocalized closely with CGI regions regardless of promoter class. (B) Methylation patterns of CGI promoters upregulated in RYBP-mutant ES cells. Cartoons delineate locations of CGIs relative to TSSs (arrows). CpG nucleotides are shown by black (methylated) or white (unmethylated) circles. Methylation extent is given as percentage of methylated CpG and shows little or no variation between wild-type (wt) and RYBP-deficient ES cells.

RYBP binds genomic regions in an H3K27me3-independent manner but does not contribute to PRC1 recruiting. (A) Distribution of hybridization intensity signals of coimmunoprecipitated DNA in wild-type (wt) and Eed-deficient ES cells compared to those of control antibodies (dotted lines). RYBP-bound plots showed no differences between wild-type and Eed-KO ES cells (left), whereas Ring1B enrichment in mutant cells dramatically decreased. The results indicate distinct responses to histone modification so that RYBP binding to promoters, but not Ring1B, is essentially independent of H3K27me3 marks. (B) PRC1 recruiting to targets is independent of RYBP. ChIP-qPCR analysis of Ring1B and Mel18 (PRC1 subunits) localization to the indicated PcG targets in wild-type and RYBP-deficient ES cells, showing little or no variation with changes in RYBP levels. RYBP binding was used as a control of RYBP depletion in mutant cells. Means and SDs of two experiments shown.

RYBP-deficient ES cells. (A and B) Western blot of cell extracts and phase-contrast images from cultures after 4 days of ethanol (EtOH) or hydroxytamoxifen (4′-OHT) treatment. (C) BrdU incorporation during a 20-min pulse in cells grown for different times after treatments (4 or 12 days [+d4 or +d12, respectively]). (D) Expression of stem cell markers in ethanol-treated (wild-type) or 4′-OHT-treated (RYBP-KO) ES cells identified by immunofluorescence (Oct4 and SSEA1) or immunohistochemistry (alkaline phosphatase [AP]). RYBP immunofluorescence was included as a control of gene inactivation. (E) Deregulated differentiation of RYBP-mutant ES cells. Time course (days in x axis) gene expression analysis by RT-qPCR of mRNAs encoding cell lineage markers; error bars are SDs of relative mRNA levels indicated in the y axis.

RYBP association with chromatin. (A and B) Distribution of RYBP-bound promoters according to histone marks and CGI length. In panel B, the fraction of total CGI promoters in each category (defined by CGI length) is indicated. The data showed preferential association with those marked with H3K27me3 (K4⁻ K27⁺, P = 4.4 × 10⁻⁴⁸; K4⁺ K27⁺, P = 8.1 × 10⁻¹⁵) and long CGIs. Note the enrichment in promoters bearing H2AK119Ub1 marks (H2AUb1⁺, P = 1.7 × 10⁻⁴³); this subset of K4⁺ K27⁺ promoters corresponds to core PcG targets. (C) Distinctive Ring1B and RYBP enrichment around transcription start sites (TSSs) (denoted by arrows) depending on their response to RYBP inactivation, showing a relatively higher RYBP density on promoters upregulated in RYBP-deficient ES cells. (D) ChIP-qPCR assay of RYBP and Ring1B enrichment including promoters of preimplantation-specific genes repressed in an RYBP-dependent manner that are not included in the microarray. (E) RYBP association with genomic regions of LTR-retrotransposons using ChIP-qPCR. IgG was used as a control for nonspecific binding. Means and SDs of two or three experiments indicated.

Gene expression alterations in RYBP-deficient ES cells. (A) Microarray analysis of preimplantation- and germ line-specific genes upregulated in RYBP-KO cells shortly (4 days) after 4′-OHT treatment. Polycomb targets include a random set of core targets (Ring1B bound and H3K27me3 and H2AK119Ub1 marked). Mean values, deduced from fluorescence intensities, and SDs of duplicate experiments with each of two independent ES cell lines. (B) RT-qPCR values of gene expression changes validating results from microarray analysis indicated as means and SDs of four to six experiments. Bars filled in black, dark gray, or light gray correspond to preimplantation- and germ line-specific genes and Polycomb targets, respectively. (C) Comparison of genes upregulated in RYBP-deficient and Ring1A- and Ring1B-deficient ES cells 4 days after 4′-OHT treatment. Absolute number of genes (within bars) and relative number (in parentheses) of upregulated genes corresponding to chromatin categories defined by H3 and H2A marks and Ring1B binding, as indicated. Note the underrepresentation of core PcG targets in RYBP-mutant ES cells. Numbers in italics correspond to values when gene expression changes in +d12 RYBP-KO ES cells are considered.