Transient silencing made permanent

Abstract

All cancers have genes that are permanently silenced by DNA methylation. Yeshayahu Schlesinger, Howard Cedar (The Hebrew University, Israel), and colleagues, and Martin Widschwendter (UCL, London, UK), Peter Laird (USC, Los Angeles, CA), and colleagues inspect these irreversibly silenced genes. In normal tissues, they find, these genes carry transient repression signals, which are inappropriately made permanent in cancer. Large parts of the genome get methylated and thus permanently silenced in the early embryo, but genes associated with CpG islands are spared. Some are instead transiently repressed by a complex called polycomb. This prevents inappropriate expression during development, but allows repression to be reversed when needed. Irreversible silencing by DNA methylation does occur at many CpG island genes during cancer progression. Schlesinger et al. now show that >60% of genes that are methylated in colon cancer are marked by polycomb in normal tissues. In an accompanying paper, Widschwendter et al. calculated from previously published data that polycomb target genes are 12-fold more likely than nontargets to be methylated in cancer. Different sets of CpG island genes get methylated in different cancers, but both groups found that, regardless of cancer type, the correlation between methylation and polycomb tagging is consistent. No causal link between transient repression and permanent silencing marks, or between methylation of genes and cancer, has been established. Cedar speculates, however, that methylation of polycomb-tagged CpG genes could be an early, even causative, event in cancer. “Polycomb target genes are required for differentiation,” he explains. “Therefore, cells have a mechanism for getting rid of polycomb. If, prior to that, these genes get abnormally methylated, the cell gets stuck in a state of proliferation, unable to differentiate.” The model suggests that cancer might originate from adult stem cells rather than from cell dedifferentiation. Perhaps an abnormally active DNA methyltransferase in certain stem cells incorrectly targets transiently repressed loci.