Mechanisms of DNMT3B-mediated DNA Methylation Maintenance

PROJECT SUMMARY DNA methylation (5mC) is catalyzed by three DNA methyltransferases (DNMTs): DNMT1, DNMT3A, and DNMT3B. DNMT3A and 3B are often referred to as de novo DNMTs, functioning to establish 5mC patterns during early embryonic development. DNMT1 is referred to as the maintenance methyltransferase, functioning

to copy established 5mC patterns on newly replicated DNA in differentiated and self-renewing stem cells. Despite this simplified division of labor, DNMT3A and 3B also contribute to 5mC maintenance, though the regulatory mechanisms involved are incompletely understood. A key 5mC regulatory mechanism is the

interaction between DNMTs and histone post-translational modifications (PTMs). For example, lysine methylation on H3K4 and H3K36 has been reported to repel or recruit DNMT3B binding, respectively. The functional outcomes of these interactions aid in the regulation of gene expression and transcription fidelity.

Specifically, DNMT3B-mediated methylation of actively transcribed gene bodies was shown to protect against spurious transcription, and this function is dependent on tri-methylation of H3K36 (H3K36me3). However, my preliminary data, as well as data from prior literature, show that H3K36me3 is dispensable for gene body 5mC

maintenance. These results challenge the established model and suggest that there are alternative mechanisms which regulate DNMT3B-mediated 5mC at intragenic regions. The overarching goal of this fellowship proposal is to define mechanisms of DNMT3B recruitment and activity in gene bodies of embryonic

and differentiated cells. In addition to histone H3, DNMT3B was reported to directly interact with the transcription elongation-associated hPAF1C complex in embryonic development. Further, hPAF1C components are required for deposition of H3K36me3 across gene bodies, and hPAF1C mutation results in aberrant expression of genes silenced by DNA hypermethylation. Taken together, these data lead me to

hypothesize that DNMT3B is recruited to chromatin through multivalent engagement of the N-terminus of histone H3 and hPAF1C to establish and maintain gene body 5mC. This hypothesis will be tested with two Specific Aims. In Aim 1, I will utilize oncohistone mutations, CRISPR/Cas9 genome editing, and genetic

complementation experiments to rigorously define the contribution of H3K4 and H3K36 methylation in the regulation of DNMT3B-mediated 5mC maintenance. Additionally, I will use publicly available cancer genome data to understand the clinical relevance of DNMT3B- mediated gene body 5mC. In Aim 2, I will use chemical

inhibitors of transcription elongation, protein-protein interaction analysis, and deep RNA-sequencing to investigate the interaction between DNMT3B and hPAF1C and its role in co-transcriptional deposition of 5mC in gene bodies. Collectively, these studies will lead to a deeper understanding of mechanisms of epigenetic

and transcriptional crosstalk that contribute to the propagation of aberrant 5mC patterning in human cancers.