| Alterations in genes regulating growth and genetic stability are central to the development of all common forms of human cancer. It is now recognized that such alterations comprise epigenetic modifications as well as genetic lesions. For example, the inactivation of tumor supressor genes appears to result from epigenetic silencing associated with hypermethylation rather than intragenic mutations in many cases. The enzymes responsible for methylating specific tumor suppressor genes, as well as maintaining global methylation patterns, in human cells are unclear. In mouse cells, the prototypic DNA methyltransferase, Dnmt1, is responsible for most methylation, and it has been widely assumed that DNMT1 is likewise responsible for most methylation of the human genome, including the abnormal methylation found in cancers. To rigorously test this hypothesis, we disrupted the DNMT1 gene through homologous recombination in human colorectal carcinoma cells. These human cells lacking DNMT1 exhibited only a 20% decrease in overall genomic methylation. Though juxtacentromeric satellites became markedly demethylated, most loci analyzed, including the tumor suppressor gene p16INK4a, remained fully methylated and silenced. The human DNMT3b gene was likewise disrupted in the same carcinoma cell line. This disruption had little effect on the level of DNA methylation, reducing it by <3%. Surprisingly, however, concurrent genetic disruption of both DNMT1 and DNMT3b resulted in dramatic effects on DNA methylation, nearly eliminating methyltransferase activity and reducing genomic DNA methylation by >95%. These changes resulted in demethylation of repeated sequences, loss of IGF2 imprinting, abrogation of silencing of the tumor suppressor gene p16INK4a, and growth suppression. These results derived from our human somatic cell models of methyltransferase function demonstrate that two enzymes cooperatively maintain DNA methylation and epigenetic silencing in human cancer cells and provide compelling evidence that such methylation is essential for optimal neoplastic proliferation. |
