Allele-specific sensitization of pancreatic ductal adenocarcinoma to PARP inhibition via adenine base editing

Synthetic lethality represents a promising therapeutic paradigm in oncology, particularly for pancreatic ductal adenocarcinoma (PDAC) patients with BRCA1/2 mutations, who demonstrate increased sensitivity to PARP inhibitors (PARPi) like olaparib. However, the broader application of this strategy is constrained by the limited incidence of homologous recombination deficiency (HRD) in PDAC that parallels BRCA loss. To address this, we employed SpRY-mediated adenine base editing (ABE) to screen for allele-specific mutations in HR-related genes across normal pancreatic epithelial (HPNE) and PDAC cell lines, including both KRAS-G12D mutant (PANC-1) and wild-type (BxPC-3) models. These edited cells were subjected to olaparib treatment at three escalating doses. A subsequent slope-based analysis of viability reduction identified several mutations that specifically sensitized PDAC cells to PARPi. Strikingly, the NBN-X mutation conferred robust olaparib hypersensitivity in KRAS-wild-type BxPC-3 cells, while PALB2-Y and RAD50-Z mutations markedly enhanced olaparib efficacy in KRAS-G12D mutant PANC-1 cells compared to normal HPNE controls. Mechanistically, the NBN-X variant sensitizes non-KRAS mutant PDAC by disrupting MYC-driven homologous recombination (HR) repair. In contrast, the RAD50-Z variant synergizes with KRAS-driven suppression of HR to potentiate PARPi efficacy. These allele-specific effects were functionally validated, demonstrating up to a 4.1-fold decrease in olaparib IC50 and a 2.5-fold increase in apoptosis. Critically, this potent sensitization was highly tumor-specific, yielding a significantly improved safety profile over conventional BRCA-mutant models both in vitro and in vivo. Leveraging these findings, we are using virtual screening and AI-driven drug design to develop small molecules that mimic the PARPi-hypersensitivity induced by our identified mutations. These lead compounds will be advanced for evaluation in sophisticated PDAC models, including patient-derived xenografts (PDX) and genetically engineered mice. Our results highlight a novel strategy for personalizing PARPi therapy in PDAC and suggest a substantially expanded therapeutic window that could improve outcomes for patients with this aggressive disease.