Epigenetic control of gene expression in human cancer

I performed my thesis work, which focused on loss of gene function associated with aberrant methylation of gene promoter CpG islands in cancer, in the laboratory of Dr. Stephen Baylin. My first efforts were directed at determining whether this promoter region abnormality might involve genes important to tumor invasion and metastasis.; I found that the TIMP-3 gene was silenced in association with aberrant promoter-region methylation in cell lines and primary tumors from human colon, lung, breast, kidney, and brain. I also observed that the methylation density of discrete regions within the TIMP-3 CpG island may be important for the silencing of this gene.; Densely methylated regions are associated with transcriptionally repressive heterochromatin characterized by the presence of deacetylated histones. In a colon cancer cell line, in which TIMP-3 is hypermethylated and silenced, inhibition of this histone deacetylation process could not reactivate TIMP-3 transcription. However, following minimal demethylation and slight gene reactivation in the presence of a low dose of a DNA methyltransferase inhibitor, addition of a histone deacetylase (HDAC) inhibitor resulted in robust re-expression of this gene suggesting that methylation is dominate over HDAC activity. This work has not only basic implications for cancer biology and gene regulation, but has also established a new potential strategy for cancer therapy.; The above observations led me to focus the remainder of my thesis work on understanding the mechanisms, which might link the process of DNA methylation with the formation of chromatin that would facilitate the transcriptional silencing of genes. I decided to study the enzymes that perform the function of DNA methylation, the DNA methyltransferases (DNMT1, 3a and 3b).; I found that (DNMT1, apart from its ability to catalyze DNA methylation, can function as a transcriptional repressor partially through binding to HDAC2 and a novel co-repressor protein, DMAP1 (D&barbelow;NM&barbelow;T A&barbelow;ssociated P&barbelow;rotein). DMAP1 is targeted to replication foci through interaction with DNMT1 throughout S-phase, while HDAC2 joins DNMT1 and DMAP1 only during late S-phase. This latter observation suggests a mechanism through which histones may become deacetylated in newly formed heterochromatin following DNA replication. I further demonstrated that Dnmt3a and 3b, like DNMT1, repress transcription in a methylation-independent manner, partially dependent on histone deacetylase activity. However, the domain involved in transcriptional repression by Dnmt3a and 3b is not found in DNMT1. Unlike DNMT1, which localizes to replication foci, Dnmt3b remains diffuse throughout the nucleus, while Dnmt3a localizes to large nuclear foci during and outside of S-phase. Dnmt3a foci have characteristics of pericentromeric heterochromatin, including the presence of HP1α and methyl-CpG binding proteins (MeCP2 and MBD1). DNMT1 is then positioned to re-establish transcriptionally repressive chromatin as cells replicate, while Dnmt3a may help to permanently maintain this repressive heterochromatin.