Targeting bifurcated methyltransferases towards user-defined DNA sequences

There would be great value in targeting methylation toward user-defined DNA sequences. Directing methylation toward single CpG sites within a genome would provide a means to examine the effects of single epigenetic alterations on cellular phenotype. The spread, erasure, or maintenance of such modifications could be examined in different cellular contexts and at different genomic loci. Further, as aberrant methylation patterns cause or are implicated in many disease states, targeted methylation might be used as a therapeutic. Many groups have attempted to target methylation toward user-defined sites by fusing a methyltransferase enzyme to a sequence specific DNA binding domain. This strategy biases the methyltransferase toward specific DNA sequences, but the methyltransferase enzyme is active in the absence of the sequence specific DNA binding event. A better strategy would involve linking the DNA binding event of sequence specific proteins to the activity of the methyltransferase enzyme. The contents of this thesis describe work on an assisted protein assembly strategy for targeting methylation to single CpG sites within a genome. This strategy utilizes naturally or unnaturally bifurcated methyltransferases fused to zinc fingers to affect reassembly over a desired site. The bifurcated methyltransferases are engineered to have reduced affinity for each other and/or for DNA, preventing unassisted enzymatic reassembly at non-targeted CpG sites. Zinc finger binding to sequences flanking an internal CpG site increase the local concentration of these assembly-deficient, bifurcated methyltransferases, enabling enzymatic reassembly and methylation only over the targeted CpG site. In Chapter 2, we demonstrate the successful implementation of this strategy for two prokaryotic methyltransferases, M.HhaI and M.SssI. Further, we elucidate design parameters important for constructing active, targeted, bifurcated methyltransferases. In Chapter 3, we describe a novel directed-evolution strategy to quickly identify optimized zinc finger-fused bifurcated methyltransferases. Importantly, we also demonstrate that substitution of bifurcated methyltransferase fragments with new zinc fingers predictably targets methylation toward new zinc finger cognate sequences. Finally, in Chapter 4, we describe successful preliminary studies in human cell lines. We demonstrate the eukaryotic expression of both fragments, targeting specific sites in a mammalian expression vector and methyltransferase activity on chromosomal DNA. Advisor: Professor Marc Ostermeier Readers: Professor Sarah Woodson Professor Jeffrey J. Gra