Synthetic lethality between PBRM1 deficiency and PARP inhibitors: exploiting G2/M checkpoint arrest in colorectal cancer.

BACKGROUND

Polybromo 1 (PBRM1), encoding the BAF180 subunit of the polybromo-associated BAF (PBAF) chromatin-remodeling complex, is commonly lost or mutated across malignancies. Despite its prevalence, tailored therapeutics for PBRM1-defective cancers remain limited. PBRM1 loss is associated with elevated replication stress and DNA damage responses, implying dependence on compensatory repair pathways. Given the lack of genotype-matched drugs, exploiting DNA-repair dependencies may provide a precision option for PBRM1-deficient disease. We pursued a synthetic-lethality strategy in colorectal cancer to test whether clinically used PARP inhibitors selectively suppress PBRM1-deficient cells and to define the linked cell-cycle and stress-response mechanisms. We also compared PARP inhibition with broader chromatin-remodeler targeting.

METHODS

Isogenic PBRM1-/- HCT116 colorectal carcinoma cells were generated by CRISPR/Cas9 lentiviral editing using sgRNAs cloned into lenti-CRISPR-V2. Knockout was confirmed by Western blotting, Sanger sequencing, and RT-qPCR. A focused compound screen compared four agents PARP inhibitors olaparib and rucaparib, the multi-target chromatin remodeler inhibitor AU-24,118, and the SMARCA2/4-targeting degrader AU-1530 using dose-response CCK-8 viability assays and selectivity indices. We then validated our results by 12-day colony-formation assays. Mechanistic analyses measured drug-induced G2/M accumulation by propidium iodide staining and flow cytometry and quantified apoptosis by Annexin V/PI dual staining, with significance assessed by t-test or two-way ANOVA.

CONCLUSIONS

PBRM1 loss confers selective hypersensitivity to PARP inhibitors, which intensify DNA-damage signaling, promote G2/M checkpoint arrest, trigger apoptosis, and induce stress-response genes such as CSRNP3. Although this effect appears context-dependent and was not observed uniformly across all PBRM1-/- models tested. These results support further evaluation of PBRM1 as a potential predictive biomarker in defined molecular contexts rather than as a universal marker of PARP inhibitor sensitivity.