Nucleobase, Nucleoside, And Neighboring Nucleotides: Intrinsic Preferences For Tet Enzyme-Mediated Oxidation Of 5-Methylcytosine

The Ten-eleven-translocation (TET) family of enzymes can oxidize the fifth base of DNA, 5-methylcytosine (mC) sequentially, to 5-hydroxymethylcytosine (hmC), 5-formylcytosine (fC), and 5-carboxycytosine (caC). The biochemical preference of TET enzymes for these substrates, in the canonical cytosine guanine dinucleotides (CpG), mimics the order in which they are generated and is reflected in levels of these oxidized modifications (oxmCs) detected in various genomes. Other than this exception, there is conflicting or limited data concerning intrinsic substrate preferences of TET, particularly with regards to different nucleic acid structures, sequence contexts, and extent to which TET mediates oxmCs in clustered proximity to one another. Thus, in this thesis, I present our efforts to determine intrinsic substrate preferences of TET enzymes, and in doing so expand upon our understanding of mechanisms driving these relative activities and the functional significance of observed levels of oxmCs in vivo. After a review of the field, in Chapter 2, I present our work comparing TET activity on different DNA and RNA structures in vitro. We found that TET is relatively promiscuous on a variety of DNA/RNA structures but prefers DNA, a specificity that is dictated by nucleic acid identity of the target base, as well helical conformation of the substrate. In Chapter 3, I newly expose the relative tolerance of TET activity on hmC with a non-G at the +1 base, although mCpG is still largely preferred. This tolerance for hmC oxidation by TET and fC and caC excision by TDG, regardless of the +1 base, supports a model explaining hmCpH depletion relative to mCpH and hmCpG in some genomes. In Chapter 4, I narrate our efforts to quantify clusters of oxmCs using modification-specific sequencing methods and observe that TET is intrinsically capable of clusters of at least fC and caC. Also, we explore the possibility that these clusters are mediated by either strand processivity of TET or underlying sequence context preferences. Finally, I propose two kinetics-based experiments to test our hypotheses regarding mechanisms driving these substrate preferences, along with ways to exploit our knowledge of TET enzymes for creation of more efficient and specific epigenetic editing tools.