UC Irvine

The recognition of higher-ordered structured nucleic acids is fundamental to immune defense, yet the molecular mechanisms governing substrate specificities remain incompletely understood. Inappropriate recognition of endogenous DNA or RNA can result in genome instability, chronic stress signaling, and the development of diseases including cancer. The overall goal of this dissertation was to define how secondary structures found in the human DNA and RNA genome regulate the activity of cytidine deaminases and double-stranded RNA binding proteins, that function as critical arms of innate and adaptive immunity, through the development, optimization, and application of quantitative biochemical and high-throughput sequencing technologies. First, I investigated the molecular basis underlying the distinct mutational signatures generated by the closely related cytosine deaminase enzymes APOBEC3A and APOBEC3B. By integrating biochemical assays, structural biology, high-throughput sequencing, and cancer genome analyses, I demonstrated that DNA secondary structure and local sequence context cooperatively determine substrate specificities of A3A and A3B. By connecting the distinct DNA hairpin architectures and sequence motifs that differentiate APOBEC3A and APOBEC3B deamination activities, we were able to assess cancer-associated mutagenesis driven by either A3A or A3B. To facilitate these studies, I utilized complementary in vitro techniques, including a quantitative fluorescence-based cytidine deaminase assay and helped to develop the novel Oligo-seq, a high-throughput sequencing platform capable of comprehensively mapping deaminase substrate preferences of A3A and A3B across complex DNA libraries. Importantly, we extended Oligo-seq application in a subsequent study and revealed that AID preferentially targets specific sequence contexts embedded within G-quadruplex and i-motif DNA structures found in the oncogenes BCL2 and MYC. These findings provided a mechanistic explanation for how DNA secondary structure contributes to off-target mutagenesis associated with B-cell lymphomagenesis. Finally, I investigated the mechanisms regulating endogenous activation of the dsRNA-sensing kinase PKR. By inducing selective RNA damage with UV irradiation and subsequent optimization of formaldehyde crosslinking immunoprecipitation sequencing (fCLIP-seq) strategy, I identified endogenous dsRNA substrates bound by PKR. Moreover, my data revealed that mechanistically, the dsRNA-binding protein PACT protects cells from chronic integrated stress response activation by competing with PKR for shared endogenous dsRNA substrates. These findings revealed that disruption of RNA homeostasis promotes sustained PKR activation, stress granule formation, and cell death through selective recognition of structured endogenous dsRNAs. Ultimately, this dissertation demonstrates that higher-order nucleic acid structure is a fundamental determinant of immune protein specificity across both DNA- and RNA-mediated pathways. Beyond providing mechanistic insight into cytidine deamination and dsRNA sensing, the biochemical and sequencing methodologies developed and optimized herein, establish versatile platforms for investigating nucleic acid recognition with unprecedented scale and resolution. My research provides a foundational basis to allow deeper interpretation of mutational processes in cancer and to advance therapeutic strategies that target enzyme-driven mediated genome instability and stress signaling.