Engineering Caspase-Activated DNase for Nucleosome-Resolution Chromosome Conformation Capture
Chromosome conformation capture methods such as Hi-C and Micro-C have transformed our understanding of genome organisation, revealing architectural features from megabase-scale compartments to kilobase-scale loops. Yet the choice of fragmentation enzyme imposes constraints that shape which features are most readily observed. Restriction enzyme workflows rely on sequence specific cut sites and often use detergent conditions that can perturb chromatin structure, while micrococcal nuclease (MNase) exhibits pronounced sequence specificity and generates termini that require end processing before efficient ligation. These properties add time, introduce variability, and may bias fragment recovery. This thesis develops CAD-C, a chromosome conformation capture method that exploits the apoptotic nuclease CAD (Caspase-Activated DNase) to address these limitations. CAD is a double strand-specific endonuclease that cleaves preferentially in linker DNA between nucleosomes and generates predominantly blunt or near-blunt double-stranded breaks with 5′-phosphate and 3′ hydroxyl termini, the end chemistry preferred by T4 DNA ligase. This native ligation competence can eliminate enzymatic end-repair steps in Direct CAD-C and supports efficient proximity ligation under chromatin-preserving conditions. I first describe the design, expression, and purification of a recombinant CAD construct optimised for chromatin fragmentation. Because CAD requires its inhibitor-chaperone ICAD for proper folding, I developed a dual-cassette bacterial expression system that co-expresses CAD with a TEV cleavable ICAD variant, enabling on-demand activation of the nuclease. I also identify purification conditions that reduce contaminating bacterial nuclease activity and improve preparation quality for controlled chromatin digestion. I next characterise CAD cleavage patterns on crosslinked chromatin and introduce the CHOMP principle (CAD Hindered Overdigestion at Margins of Proteins), which describes how CAD’s steric bulk can bias cleavage toward defined offsets from chromatin-bound proteins, leaving accessible termini available for ligation. This framework provides a conceptual basis for interpreting CAD-C contact maps and the locations where ligation-competent ends form. I then present the CAD-C protocol in three workflow variants: Classic, Express, and Direct. Benchmarking against Micro-C and Hi-C in GM12878, hTERT-RPE1, and K562 cells shows that CAD-C recovers major architectural features, including compartments, domains, and loops, with performance comparable to Micro-C in these datasets. At the same time, CAD-C exhibits a distinctive short-range contact signature and preferentially recovers a subset of active regulatory contacts, including promoter-associated loops. I develop a ligation-gating model to explain how CAD’s native ligation competence can increase near-diagonal contact density while sharpening loop contrast relative to MNase-based methods. Finally, I introduce CADwalks, a per-molecule analytical framework for Direct CAD-C con catemers. Because CAD generates nucleosome-resolution fragments whose boundaries encode protein-DNA footprint information, and because its ligation-ready termini support efficient concate mer formation, CADwalks preserve the relative order, orientation, and co-occurrence of fragments captured on the same concatemer molecule, information that pairwise decomposition discards. At CTCF boundaries, CADwalks reveal a nucleosome-scale directional asymmetry consistent with oriented loop extrusion. This thesis is primarily a tool-development and validation effort. The experimental work em phasizes proof-of-principle demonstrations rather than exhaustive perturbation biology. Together, CAD-C and CADwalks show that changing the chromatin fragmentation enzyme changes not only assay resolution and workflow efficiency, but also the biochemical gate through which contacts are captured. They establish a nucleosome-resolution, per-molecule framework for chromosome con formation capture and provide a foundation for future studies of chromatin architecture, regulatory looping, and gene control.