Final Report Summary - PRIME (Polycomb repressor interactions in relation to the mammalian epigenome.)
Polycomb repressor complexes are highly conserved chromatin modifying machines that play a central role in regulating gene expression through development. The targeting of these complexes to specific genes has generally been attributed to the activity of sequence specific DNA binding factors and/or non-coding RNAs. However, recent evidence has suggested that targeting may also involve Polycomb complexes reading pre-existing chromatin states, leading to either stimulation or inhibition of their binding/activity. PRIME (Polycomb Repressor Interactions with the Mammalian Epigenome) set out to explore these interactions in relation to the major Polycomb complexes, PRC1 and PRC2.
When the PRIME project was initiated we, together with collaborators in Oxford and Japan, had made key breakthroughs, demonstrating that unmethylated CpG dinucleotide is a critical determinant of both PRC1 and PRC2 occupancy. Additionally we made the unexpected discovery that H2A ubiquitylation (H2AK119ub1), catalysed by PRC1, can be recognised by PRC2. The latter finding, the converse of the known interaction of PRC1 with PRC2 catalysed H3 lysine 27 methylation (H3K27me3), provided a rational explanation for Polycomb occupancy at CpG islands of target genes, based on initial targeting of PRC1 complexes via the subunit KDM2B, which recognises unmethylated CpG. Whether or not PRC1 activity precedes PRC2 recruitment in normal physiology and how PRC2 recognises H2AK119ub1 however were unknown. Our work within the PRIME project has gone a considerable way towards addressing these key questions. We determined unequivocally that PRC1 recruitment precedes that of PRC2 in a widely studied model system, the inactive X chromosome, in which Polycomb recruitment to an entire chromosome is mediated by a non-coding RNA, Xist (Almeida et al, Science 356, p1081-1084, 2017). Moreover, we were able to demonstrate how positive feedback mechanisms enhance Polycomb occupancy through recognition of H2AK119ub1 by both PRC1 and PRC2 complexes. To understand how PRC2 recognises H2AK11ub1 we analysed a candidate PRC2 sub-complex which includes the cofactors Aebp2 and Jarid2. We found that Aebp2 functions as an anti-Polycomb factor (Grijzenhout et al, Development 143, p2716, 2016), and as such is unlikely to mediate H2AK119ub1 recognition. However, in a separate study we demonstrated that Jarid2 directly mediates recognition of H2AK119ub1 by PRC2 (Cooper et al, Nature Communications, doi: 10.1038/ncomms13661 2016). Building on these important advances we went on to define the mechanism for PRC1 recruitment by Xist RNA, involving the RNA binding protein hnRNPK (Pintacuda et al, Molecular Cell 68, p955-969). In a related study we defined a novel protein complex that is recruited by the histone modification H3K36me3 which is found in the body of active genes and is known to antagonise activity of PRC2 complexes (Zhang et al, Nature Communications, doi: 10.1038/s41467-018-06235-9 2018). We also further characterised the reciprocal relationship between CpG methylation and Polycomb activity by performing high-resolution allelic bisulfite (CpG methylation) and ChIP-sequencing (H3K27me3) analysis of the inactive X chromosome (Gdula et al, Nature Communications, doi: 10.1038/s41467-018-07907-2 2019). Finally we performed detailed analysis to determine the relative contribution of both PRC1 and PRC2 to gene repression in X chromosome inactivation (Nesterova et al, PREPRINT bioRxiv 477232; doi: https://doi.org/10.1101/477232 2019). Together our findings have significantly advanced our understanding of how the Polycomb system interplays with other epigenetic pathways in gene regulation during cellular differentiation and development in mammals.
When the PRIME project was initiated we, together with collaborators in Oxford and Japan, had made key breakthroughs, demonstrating that unmethylated CpG dinucleotide is a critical determinant of both PRC1 and PRC2 occupancy. Additionally we made the unexpected discovery that H2A ubiquitylation (H2AK119ub1), catalysed by PRC1, can be recognised by PRC2. The latter finding, the converse of the known interaction of PRC1 with PRC2 catalysed H3 lysine 27 methylation (H3K27me3), provided a rational explanation for Polycomb occupancy at CpG islands of target genes, based on initial targeting of PRC1 complexes via the subunit KDM2B, which recognises unmethylated CpG. Whether or not PRC1 activity precedes PRC2 recruitment in normal physiology and how PRC2 recognises H2AK119ub1 however were unknown. Our work within the PRIME project has gone a considerable way towards addressing these key questions. We determined unequivocally that PRC1 recruitment precedes that of PRC2 in a widely studied model system, the inactive X chromosome, in which Polycomb recruitment to an entire chromosome is mediated by a non-coding RNA, Xist (Almeida et al, Science 356, p1081-1084, 2017). Moreover, we were able to demonstrate how positive feedback mechanisms enhance Polycomb occupancy through recognition of H2AK119ub1 by both PRC1 and PRC2 complexes. To understand how PRC2 recognises H2AK11ub1 we analysed a candidate PRC2 sub-complex which includes the cofactors Aebp2 and Jarid2. We found that Aebp2 functions as an anti-Polycomb factor (Grijzenhout et al, Development 143, p2716, 2016), and as such is unlikely to mediate H2AK119ub1 recognition. However, in a separate study we demonstrated that Jarid2 directly mediates recognition of H2AK119ub1 by PRC2 (Cooper et al, Nature Communications, doi: 10.1038/ncomms13661 2016). Building on these important advances we went on to define the mechanism for PRC1 recruitment by Xist RNA, involving the RNA binding protein hnRNPK (Pintacuda et al, Molecular Cell 68, p955-969). In a related study we defined a novel protein complex that is recruited by the histone modification H3K36me3 which is found in the body of active genes and is known to antagonise activity of PRC2 complexes (Zhang et al, Nature Communications, doi: 10.1038/s41467-018-06235-9 2018). We also further characterised the reciprocal relationship between CpG methylation and Polycomb activity by performing high-resolution allelic bisulfite (CpG methylation) and ChIP-sequencing (H3K27me3) analysis of the inactive X chromosome (Gdula et al, Nature Communications, doi: 10.1038/s41467-018-07907-2 2019). Finally we performed detailed analysis to determine the relative contribution of both PRC1 and PRC2 to gene repression in X chromosome inactivation (Nesterova et al, PREPRINT bioRxiv 477232; doi: https://doi.org/10.1101/477232 2019). Together our findings have significantly advanced our understanding of how the Polycomb system interplays with other epigenetic pathways in gene regulation during cellular differentiation and development in mammals.