While the Protein Data Bank contains a vast collection of protein folds and complexes, many nucleic acid-protein assemblies remain undiscovered. Although genomic DNA predominantly adopts a B-form, multiple non-canonical forms play distinct regulatory roles, affecting transcriptional efficiency and occluding typical binders. Notably, Z-DNA is a left-handed form of DNA that can be found in stretches of purine-pyrimidine repeats. In Alzheimer’s disease, elevated levels of Z-DNA are observed in the hippocampus, and Z-DNA-forming sequences are present within the promoters of genes associated with Alzheimer’s disease. The ability to identify Z-form DNA within chromatin will enable direct investigation of the causal relationship between Z-DNA-prone promoter regions and pathology; however, most studies remain confined to sequence-level analyses, owing in part to the scarcity of high-resolution structural data on protein-Z-DNA complexes. De novo designed proteins can, in principle, be engineered to recognize such non-canonical DNA form by manipulating already existing motifs which could then serve as scaffolds for structural determination. Here, we outline a design strategy to generate a high-affinity, sequence-specific Z-DNA binder with backbone complementarity. We repurpose transcription-activator-like effector (TALE) proteins, whose modular, superhelical architecture, phosphate backbone and base pair recognition capability as our design scaffold. The nature of TALEs makes them well suited for targeting repetitive, Z-forming sequences while accommodating the distinct groove geometry. We reparameterize repeat modules to match Z-DNA’s altered phosphate spacing and groove topology that favor Z-specific electrostatics and hydration. This work aims to deliver a reliable Z-DNA sensor, enabling targeted interrogation of non-B DNA disease states or in vivo conditions.
Sebastian et al. (Sun,) studied this question.