Limits...
A Cbx8-containing polycomb complex facilitates the transition to gene activation during ES cell differentiation.

Creppe C, Palau A, Malinverni R, Valero V, Buschbeck M - PLoS Genet. (2014)

Bottom Line: Prolonged gene activation results in eviction of PRC1 despite persisting H3K27me3 and H2A ubiquitination.We further demonstrate that the exchange of Cbx7 for Cbx8 is required for the effective activation of differentiation genes.Taken together, our results establish a function for a Cbx8-containing complex in facilitating the transition from a Polycomb-repressed chromatin state to an active state.

View Article: PubMed Central - PubMed

Affiliation: Institute of Predictive and Personalized Medicine of Cancer (IMPPC), Badalona, Barcelona, Spain.

ABSTRACT
Polycomb proteins play an essential role in maintaining the repression of developmental genes in self-renewing embryonic stem cells. The exact mechanism allowing the derepression of polycomb target genes during cell differentiation remains unclear. Our project aimed to identify Cbx8 binding sites in differentiating mouse embryonic stem cells. Therefore, we used a genome-wide chromatin immunoprecipitation of endogenous Cbx8 coupled to direct massive parallel sequencing (ChIP-Seq). Our analysis identified 171 high confidence peaks. By crossing our data with previously published microarray analysis, we show that several differentiation genes transiently recruit Cbx8 during their early activation. Depletion of Cbx8 partially impairs the transcriptional activation of these genes. Both interaction analysis, as well as chromatin immunoprecipitation experiments support the idea that activating Cbx8 acts in the context of an intact PRC1 complex. Prolonged gene activation results in eviction of PRC1 despite persisting H3K27me3 and H2A ubiquitination. The composition of PRC1 is highly modular and changes when embryonic stem cells commit to differentiation. We further demonstrate that the exchange of Cbx7 for Cbx8 is required for the effective activation of differentiation genes. Taken together, our results establish a function for a Cbx8-containing complex in facilitating the transition from a Polycomb-repressed chromatin state to an active state. As this affects several key regulatory differentiation genes this mechanism is likely to contribute to the robust execution of differentiation programs.

Show MeSH
Cbx8 binding coincides with gene activation.(A) The occupancy of Cbx8 on genes was analyzed by ChIP in untreated ES cells and differentiating cells treated with RA for 3 days. Data is plotted as percentage of ChIP-enriched DNA in respect to DNA from input material. Error bars denote s.d., and n = 3. (B) qRT-PCR analysis of mRNA levels in the same cells as A. Levels in untreated cells have been set to one. Error bars denote s.d., and n = 3. (C) ChIP and ReChIP analyses performed with differentiating ES cells treated with RA for three days are shown. Anti-H3K36me3 and control IgG antibody were used in a primary ChIP (left panel) or in a secondary ChIP performed on material enriched in a primary ChIP with anti-Cbx8 antibody (right panel). ReChIP data is plotted relative to IgG. Error bars denote s.d., n = 3.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4263398&req=5

pgen-1004851-g002: Cbx8 binding coincides with gene activation.(A) The occupancy of Cbx8 on genes was analyzed by ChIP in untreated ES cells and differentiating cells treated with RA for 3 days. Data is plotted as percentage of ChIP-enriched DNA in respect to DNA from input material. Error bars denote s.d., and n = 3. (B) qRT-PCR analysis of mRNA levels in the same cells as A. Levels in untreated cells have been set to one. Error bars denote s.d., and n = 3. (C) ChIP and ReChIP analyses performed with differentiating ES cells treated with RA for three days are shown. Anti-H3K36me3 and control IgG antibody were used in a primary ChIP (left panel) or in a secondary ChIP performed on material enriched in a primary ChIP with anti-Cbx8 antibody (right panel). ReChIP data is plotted relative to IgG. Error bars denote s.d., n = 3.

Mentions: We used retinoic acid (RA) to induce mouse E14 ES cells to start differentiating towards the neuronal lineage. We confirmed previous results [16], [17] showing that Cbx8 was virtually absent in self-renewing ES cells but potently induced on protein and RNA levels after three days of RA-induced differentiation (Fig. 1A and S1A Figure). To assess the genome wide distribution of Cbx8 in differentiating ES cells, we enriched Cbx8-bound chromatin by chromatin immunoprecipitation (ChIP) using an antibody generated against the unique part of the protein (S1B-G Figure) and analyzed the co-precipitated DNA by direct massive parallel sequencing (ChIP-seq). Taking advantage of the fact that Cbx8 is virtually absent in untreated, self-renewing ES cells (Fig. 1A), we decided to use both IgG as well as Cbx8 ChIPs from untreated cells as negative controls. We were able to uniquely map 9–16 million reads per sample (S2A Figure). For our further analysis we used a set of high confidence binding sites that were identified by the overlap of peaks that were called by MACS comparing Cbx8 ChIP from RA-treated ES cells to IgG and those called comparing Cbx8 ChIP from RA-treated ES cells to the antibody-specific background ChIPed from untreated cells (S2B Figure). By this method we were able to identify a subset of 171 peaks corresponding to Cbx8 binding sites of high confidence (S1 Table). Peaks were annotated to the nearest gene if the center of the peak was inside a window flanking the transcribed region by 3 kb. Plotting the average read coverage on genes with annotated Cbx8 indicated that Cbx8 tends to accumulate on gene bodies but also to spread into upstream and downstream regions (Fig. 1C). Using this annotation we found that the large majority (141/171) of identified peaks are associated to annotated genes (Fig. 1D). We performed a microarray analysis comparing self renewing untreated ES cells and differentiating cells after 3 days of RA treatment and crossed the data with our ChIP-seq to identify possible transcriptional changes on Cbx8 target genes. Taking into consideration the established function of Polycomb proteins in gene repression, we expected target genes to either not change because they are maintained in a repressed state or to be downregulated. We could extract data for 121 of the 141 gene-associated peaks. We found that about one third of these Cbx8 binding sites (53 peaks) annotated to genes that displayed a more than 1.5-fold change in gene expression. To our surprise the large majority of these genes (44/53) was not down- but upregulated (Fig. 1E). Most of these upregulated genes were repressed in untreated cells as indicated by very low average probe intensities (Fig. 1F). Though a previous report has already shown that Cbx8 can be found on a handful of activated genes in differentiating cells [19], our data suggested that this could actually be true for a substantial fraction of Cbx8 target genes. In order to confirm this, we selected a panel of target and control genes and simultaneously analyzed Cbx8 recruitment and the corresponding mRNA levels. We were able to confirm differentiation-induced recruitment of Cbx8 on all target genes tested while control genes were negative (Fig. 2A). Target genes included many important key differentiation genes such as Sox9, Gata6 and Nkx6-1 whose expression was potently induced (Fig. 2B). In order to exclude the possibility that Cbx8 binding and active transcription might occur on different and exclusive alleles within the cell population, we analyzed the co-occurrence of Cbx8 and H3K36me3, which is a mark of active transcription [20], by coupled ChIP in differentiating ES cells treated for three days with RA. First, we have analyzed five Cbx8 target genes and four non-target genes that included Oct4, Nanog, Gapdh and Rpo. Despite the fact that Nanog is downregulated after three days of RA treatment (Fig. 2B) it still retained some H3K36me3 that was in a similar range as on the activated gene Sox9 or the constitutively active gene Gapdh (Fig. 2C, left panel) suggesting that removal of the active mark H3K36me3 follows a slower dynamic than the actual gene repression. Taking advantage of the fact that with Gapdh, Nanog, and Sox9 we had identified target and non-target genes of Cbx8 with comparable H3K36me3 levels, we have used anti-Cbx8 antibody-enriched chromatin as input material for a secondary ChIP with IgG and anti-H3K36me3 antibody. As shown in the right panel of Fig. 2C, we found a clear enrichment of H3K36me3 over IgG on Sox9 and on the other Cbx8 target genes but not on non-target genes such as Gapdh or Nanog. As chromatin ranging in size between 300–500 bps has been used for these experiments H3K36me3 and binding of Cbx8 could be occurring on different H3 tails on the same or even neighboring nucleosomes. The observation that Cbx8 can be simultaneously detected on the same locus provides further support to the idea that Cbx8 is recruited to genes that become actively transcribed.


A Cbx8-containing polycomb complex facilitates the transition to gene activation during ES cell differentiation.

Creppe C, Palau A, Malinverni R, Valero V, Buschbeck M - PLoS Genet. (2014)

Cbx8 binding coincides with gene activation.(A) The occupancy of Cbx8 on genes was analyzed by ChIP in untreated ES cells and differentiating cells treated with RA for 3 days. Data is plotted as percentage of ChIP-enriched DNA in respect to DNA from input material. Error bars denote s.d., and n = 3. (B) qRT-PCR analysis of mRNA levels in the same cells as A. Levels in untreated cells have been set to one. Error bars denote s.d., and n = 3. (C) ChIP and ReChIP analyses performed with differentiating ES cells treated with RA for three days are shown. Anti-H3K36me3 and control IgG antibody were used in a primary ChIP (left panel) or in a secondary ChIP performed on material enriched in a primary ChIP with anti-Cbx8 antibody (right panel). ReChIP data is plotted relative to IgG. Error bars denote s.d., n = 3.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4263398&req=5

pgen-1004851-g002: Cbx8 binding coincides with gene activation.(A) The occupancy of Cbx8 on genes was analyzed by ChIP in untreated ES cells and differentiating cells treated with RA for 3 days. Data is plotted as percentage of ChIP-enriched DNA in respect to DNA from input material. Error bars denote s.d., and n = 3. (B) qRT-PCR analysis of mRNA levels in the same cells as A. Levels in untreated cells have been set to one. Error bars denote s.d., and n = 3. (C) ChIP and ReChIP analyses performed with differentiating ES cells treated with RA for three days are shown. Anti-H3K36me3 and control IgG antibody were used in a primary ChIP (left panel) or in a secondary ChIP performed on material enriched in a primary ChIP with anti-Cbx8 antibody (right panel). ReChIP data is plotted relative to IgG. Error bars denote s.d., n = 3.
Mentions: We used retinoic acid (RA) to induce mouse E14 ES cells to start differentiating towards the neuronal lineage. We confirmed previous results [16], [17] showing that Cbx8 was virtually absent in self-renewing ES cells but potently induced on protein and RNA levels after three days of RA-induced differentiation (Fig. 1A and S1A Figure). To assess the genome wide distribution of Cbx8 in differentiating ES cells, we enriched Cbx8-bound chromatin by chromatin immunoprecipitation (ChIP) using an antibody generated against the unique part of the protein (S1B-G Figure) and analyzed the co-precipitated DNA by direct massive parallel sequencing (ChIP-seq). Taking advantage of the fact that Cbx8 is virtually absent in untreated, self-renewing ES cells (Fig. 1A), we decided to use both IgG as well as Cbx8 ChIPs from untreated cells as negative controls. We were able to uniquely map 9–16 million reads per sample (S2A Figure). For our further analysis we used a set of high confidence binding sites that were identified by the overlap of peaks that were called by MACS comparing Cbx8 ChIP from RA-treated ES cells to IgG and those called comparing Cbx8 ChIP from RA-treated ES cells to the antibody-specific background ChIPed from untreated cells (S2B Figure). By this method we were able to identify a subset of 171 peaks corresponding to Cbx8 binding sites of high confidence (S1 Table). Peaks were annotated to the nearest gene if the center of the peak was inside a window flanking the transcribed region by 3 kb. Plotting the average read coverage on genes with annotated Cbx8 indicated that Cbx8 tends to accumulate on gene bodies but also to spread into upstream and downstream regions (Fig. 1C). Using this annotation we found that the large majority (141/171) of identified peaks are associated to annotated genes (Fig. 1D). We performed a microarray analysis comparing self renewing untreated ES cells and differentiating cells after 3 days of RA treatment and crossed the data with our ChIP-seq to identify possible transcriptional changes on Cbx8 target genes. Taking into consideration the established function of Polycomb proteins in gene repression, we expected target genes to either not change because they are maintained in a repressed state or to be downregulated. We could extract data for 121 of the 141 gene-associated peaks. We found that about one third of these Cbx8 binding sites (53 peaks) annotated to genes that displayed a more than 1.5-fold change in gene expression. To our surprise the large majority of these genes (44/53) was not down- but upregulated (Fig. 1E). Most of these upregulated genes were repressed in untreated cells as indicated by very low average probe intensities (Fig. 1F). Though a previous report has already shown that Cbx8 can be found on a handful of activated genes in differentiating cells [19], our data suggested that this could actually be true for a substantial fraction of Cbx8 target genes. In order to confirm this, we selected a panel of target and control genes and simultaneously analyzed Cbx8 recruitment and the corresponding mRNA levels. We were able to confirm differentiation-induced recruitment of Cbx8 on all target genes tested while control genes were negative (Fig. 2A). Target genes included many important key differentiation genes such as Sox9, Gata6 and Nkx6-1 whose expression was potently induced (Fig. 2B). In order to exclude the possibility that Cbx8 binding and active transcription might occur on different and exclusive alleles within the cell population, we analyzed the co-occurrence of Cbx8 and H3K36me3, which is a mark of active transcription [20], by coupled ChIP in differentiating ES cells treated for three days with RA. First, we have analyzed five Cbx8 target genes and four non-target genes that included Oct4, Nanog, Gapdh and Rpo. Despite the fact that Nanog is downregulated after three days of RA treatment (Fig. 2B) it still retained some H3K36me3 that was in a similar range as on the activated gene Sox9 or the constitutively active gene Gapdh (Fig. 2C, left panel) suggesting that removal of the active mark H3K36me3 follows a slower dynamic than the actual gene repression. Taking advantage of the fact that with Gapdh, Nanog, and Sox9 we had identified target and non-target genes of Cbx8 with comparable H3K36me3 levels, we have used anti-Cbx8 antibody-enriched chromatin as input material for a secondary ChIP with IgG and anti-H3K36me3 antibody. As shown in the right panel of Fig. 2C, we found a clear enrichment of H3K36me3 over IgG on Sox9 and on the other Cbx8 target genes but not on non-target genes such as Gapdh or Nanog. As chromatin ranging in size between 300–500 bps has been used for these experiments H3K36me3 and binding of Cbx8 could be occurring on different H3 tails on the same or even neighboring nucleosomes. The observation that Cbx8 can be simultaneously detected on the same locus provides further support to the idea that Cbx8 is recruited to genes that become actively transcribed.

Bottom Line: Prolonged gene activation results in eviction of PRC1 despite persisting H3K27me3 and H2A ubiquitination.We further demonstrate that the exchange of Cbx7 for Cbx8 is required for the effective activation of differentiation genes.Taken together, our results establish a function for a Cbx8-containing complex in facilitating the transition from a Polycomb-repressed chromatin state to an active state.

View Article: PubMed Central - PubMed

Affiliation: Institute of Predictive and Personalized Medicine of Cancer (IMPPC), Badalona, Barcelona, Spain.

ABSTRACT
Polycomb proteins play an essential role in maintaining the repression of developmental genes in self-renewing embryonic stem cells. The exact mechanism allowing the derepression of polycomb target genes during cell differentiation remains unclear. Our project aimed to identify Cbx8 binding sites in differentiating mouse embryonic stem cells. Therefore, we used a genome-wide chromatin immunoprecipitation of endogenous Cbx8 coupled to direct massive parallel sequencing (ChIP-Seq). Our analysis identified 171 high confidence peaks. By crossing our data with previously published microarray analysis, we show that several differentiation genes transiently recruit Cbx8 during their early activation. Depletion of Cbx8 partially impairs the transcriptional activation of these genes. Both interaction analysis, as well as chromatin immunoprecipitation experiments support the idea that activating Cbx8 acts in the context of an intact PRC1 complex. Prolonged gene activation results in eviction of PRC1 despite persisting H3K27me3 and H2A ubiquitination. The composition of PRC1 is highly modular and changes when embryonic stem cells commit to differentiation. We further demonstrate that the exchange of Cbx7 for Cbx8 is required for the effective activation of differentiation genes. Taken together, our results establish a function for a Cbx8-containing complex in facilitating the transition from a Polycomb-repressed chromatin state to an active state. As this affects several key regulatory differentiation genes this mechanism is likely to contribute to the robust execution of differentiation programs.

Show MeSH