BackgroundDNA methylation and the Polycomb repression system are epigenetic mechanisms that play important roles in maintaining transcriptional repression. Recent evidence suggests that DNA methylation can attenuate the binding of Polycomb protein components to chromatin and thus plays a role in determining their genomic targeting. However, whether this role of DNA methylation is important in the context of transcriptional regulation is unclear.ResultsBy genome-wide mapping of the Polycomb Repressive Complex 2-signature histone mark, H3K27me3, in severely DNA hypomethylated mouse somatic cells, we show that hypomethylation leads to widespread H3K27me3 redistribution, in a manner that reflects the local DNA methylation status in wild-type cells. Unexpectedly, we observe striking loss of H3K27me3 and Polycomb Repressive Complex 2 from Polycomb target gene promoters in DNA hypomethylated cells, including Hox gene clusters. Importantly, we show that many of these genes become ectopically expressed in DNA hypomethylated cells, consistent with loss of Polycomb-mediated repression.ConclusionsAn intact DNA methylome is required for appropriate Polycomb-mediated gene repression by constraining Polycomb Repressive Complex 2 targeting. These observations identify a previously unappreciated role for DNA methylation in gene regulation and therefore influence our understanding of how this epigenetic mechanism contributes to normal development and disease.

BackgroundThe DNA methylation profiles of mammalian cell lines differ from those of the primary tissues from which they were derived, exhibiting increasing divergence from the in vivo methylation profile with extended time in culture. Few studies have directly examined the initial epigenetic and transcriptional consequences of adaptation of primary mammalian cells to culture, and the potential mechanisms through which this epigenetic dysregulation occurs is unknown.ResultsWe demonstrate that adaptation of mouse embryonic fibroblasts to cell culture results in a rapid reprogramming of epigenetic and transcriptional states. We observed global 5-hydroxymethylcytosine (5hmC) erasure within three days of culture initiation. Loss of genic 5hmC was independent of global 5-methylcytosine (5mC) levels and could be partially rescued by addition of vitamin C. Significantly, 5hmC loss was not linked to concomitant changes in transcription. Discrete promoter-specific gains of 5mC were also observed within seven days of culture initiation. Against this background of global 5hmC loss we identified a handful of developmentally important genes that maintained their 5hmC profile in culture, including the imprinted loci Gnas and H19. Similar outcomes were identified in the adaption of CD4+ T cells to culture.ConclusionsWe report a dramatic and novel consequence of adaptation of mammalian cells to culture in which global loss of 5hmC occurs, suggesting rapid concomitant loss of methylcytosine dioxygenase activity. The observed epigenetic and transcriptional re-programming occurs much earlier than previously assumed, and has significant implications for the use of cell lines as faithful mimics of in vivo epigenetic and physiological processes.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-014-0576-y) contains supplementary material, which is available to authorized users.

There is much interest in the mechanisms that regulate adult tissue homeostasis and their relationship to processes governing foetal development. Mice deleted for the Wilms' tumour gene, Wt1, lack kidneys, gonads, and spleen and die at mid-gestation due to defective coronary vasculature. Wt1 is vital for maintaining the mesenchymal–epithelial balance in these tissues and is required for the epithelial-to-mesenchyme transition (EMT) that generates coronary vascular progenitors. Although Wt1 is only expressed in rare cell populations in adults including glomerular podocytes, 1% of bone marrow cells, and mesothelium, we hypothesised that this might be important for homeostasis of adult tissues; hence, we deleted the gene ubiquitously in young and adult mice. Within just a few days, the mice suffered glomerulosclerosis, atrophy of the exocrine pancreas and spleen, severe reduction in bone and fat, and failure of erythropoiesis. FACS and culture experiments showed that Wt1 has an intrinsic role in both haematopoietic and mesenchymal stem cell lineages and suggest that defects within these contribute to the phenotypes we observe. We propose that glomerulosclerosis arises in part through down regulation of nephrin, a known Wt1 target gene. Protein profiling in mutant serum showed that there was no systemic inflammatory or nutritional response in the mutant mice. However, there was a dramatic reduction in circulating IGF-1 levels, which is likely to contribute to the bone and fat phenotypes. The reduction of IGF-1 did not result from a decrease in circulating GH, and there is no apparent pathology of the pituitary and adrenal glands. These findings 1) suggest that Wt1 is a major regulator of the homeostasis of some adult tissues, through both local and systemic actions; 2) highlight the differences between foetal and adult tissue regulation; 3) point to the importance of adult mesenchyme in tissue turnover.