Just like a bossy older sister, DNA methylation tells the Polycomb complex where it can and cannot bind. That’s the gist of a recent paper showing that removing DNA methylation actually frees up the Polycomb machinery to bind to new places, changing gene expression.
DNA methylation is linked to Polycomb repression, but the question of whether this had anything to do with transcription was still up in the air.
So the research team in the U.K., U.S., and Australia figured they would learn a thing or two about this if they just got rid of DNA methylation altogether. Would Polycomb just replace DNA methylation with its tell-tale H3K27me3 marks in the exact same places? Or would something else happen?
They used severely hypomethylated mouse somatic cells (Dnmt1-/-) and genome-wide mapping to help answer those questions. Here’s some of what they found:
- Once DNA methylation was gone, H3K27me3 didn’t simply appear at those same formerly methylated promoters. Almost 1500 promoters had less H3K27me3 in Dnmt1-/- cells than in control cells. Only 81 promoters had more H3K27me3.
- H3K27me3 winds up spreading out into regions outside of gene promoters and its location generally correlated with areas that used to have DNA methylation.
- Some genes normally repressed by PRC2 were expressed in hypomethylated cells.
Why didn’t H3K27me3 go up more at promoters in hypomethylated cells? The researchers say that PRC2 is probably diluted from its normal targets in those cells.
“The global DNA methylation terrain essentially shrinks the ‘epigenetic space’ available to the Polycomb machinery by masking potential binding sites,” says Richard Meehan of the MRC Human Genetics Unit, University of Edinburgh and Western General Hospital. “Remove this restriction, and Polycomb-mediated repression dilutes from its normal address points and can migrate to new landing pads, introducing a silencing state that was paradoxically prevented by local DNA methylation.”
Read all the details at Genome Biology, March 2013.