Chemically inducible split base editors for precise and controllable in vivo genome editing

Project Summary Hypertrophic cardiomyopathy (HCM) affects 1 in 500 adults, with treatment limited to managing symptoms rather than addressing the root cause. Around 40% of familial HCM patients have mutations in the gene for myosin- binding protein C (MYBPC3), causing the disease. Theoretically, a single dose of a genomic editor correcting MYBPC3 mutations in the heart would cure the disease for these patients. While the base editors (BE), fusing a nickase-Cas9 (nCas9) to a cytosine deaminase (CBE) or adenine deaminase (ABE), can achieve high-efficiency single nucleotide substitutions, their propensity for off-target editing due to a lack of temporal control over the editing activity poses significant safety concerns. We recently developed split ABE (sABE) that utilizes chemically induced dimerization (CID) to control the TadA deaminase activity. Our sABE retains a high on- target editing activity similar to the canonical intact ABE but displays significantly reduced off-target editing with a narrower activity window and improved precision. When delivered via dual adeno-associated virus (AAV) vectors, sABE has achieved efficient single base conversion on the PCSK9 gene in mouse liver. This gene is a prominent drug target for atherosclerosis and regulating blood cholesterol levels. This achievement marks the first in vivo CID-controlled gene knockdown. We propose addressing the current weaknesses of the sABE system and applying it to HCM treatment. In Aim 1, We will expand CID systems and optimize the splitting strategy in sABE. Our current sABE employs a rapamycin-dependent CID system. Rapamycin has immunosuppressive and autophagic effects in vivo, which can limit its applicability to certain diseases. In Aim 2, we will create sBE variants to achieve versatile base editing with low off-target activity. sABE can only achieve A-to-G editing, while disease-causing mutations can result from other mutation types. We will translate our inducible design to an expanded set of BE systems using sTadA deaminase variants, achieving C-to-T and C- to-G editing, etc. We will then apply sBE variants to knock down PCSK9 in vivo to demonstrate their functionality. In summary, achieving tight spatiotemporal control of the base editing enzyme activity with alternative small molecule inducers will yield versatile split base editors, significantly reducing genomic and transcriptomic off- target editing while maintaining similar levels of on-target editing activity. Correcting MYBPC3-associated HCM and decreasing PCSK9 expression and blood cholesterol levels will validate these sBE systems in vivo, laying the groundwork for their future adoption in treating a broad range of diseases.