DDRKOL: A Focused CRISPR Library for Systematic Identification of DNA Damage Response Dependencies in Glioblastoma

Abstract

Background: DNA damage response (DDR) pathways are central regulators of genome maintenance and major determinants of cancer cell survival. The extensive genomic instability and high replicative stress that characterize glioblastoma render tumor cells highly dependent on DDR pathways to preserve genome integrity and sustain proliferation. This reliance creates potential therapeutic vulnerabilities, making the systematic identification of essential DDR genes a promising strategy for uncovering novel therapeutic targets. Methods: We developed DNA Damage Response KnockOut Library (DDRKOL), a custom CRISPR/Cas9 sgRNA library targeting 819 DDR genes with approximately 10 sgRNAs per gene, together with positive (essential), negative (non-essential) and non-targeting controls. Parallel depletion screens were performed in Cas9-expressing U87-MG and A172 cells cultured for 15 population doublings. Hits were prioritized utilizing TCGA and DepMap databases and validated by viability, clonogenic, apoptosis and GFP competition assays. Clinically relevant patient-derived glioblastoma spheroids and an orthotopic xenograft model was employed to characterize the effects of hit genes. Results: Sequencing confirmed near-complete recovery of the designed sgRNAs from the plasmid pool, with uniform representation across the library and complexity preserved through transduction and selection. Essential-gene controls depleted strongly while non-targeting controls remained neutral, confirming screen performance in both cell lines. The screens identified DDR dependencies in each line and defined a shared core composed of 20 genes belonging to homologous recombination, nucleotide excision repair and ATM/DSB signaling pathways. This shared dependency landscape highlighted four high-confidence candidate genes (TOP2A, CDK1, XRCC6, and RAD21), which were successfully validated across multiple orthogonal assays. These genes displayed grade-associated expression and their expressions were positively correlated with proliferation markers in TCGA. Individual knockouts reduced viability, colony formation and competitive fitness, induced apoptosis, and impaired growth of patient-derived glioblastoma spheroids. Both genetic depletion and pharmacological inhibition of TOP2A induced S/G2-M cell cycle arrest. In orthotopic xenografts, TOP2A depletion prevented tumor progression, and led to significantly prolonged survival. Conclusion: DDRKOL represents a robust and versatile focused CRISPR platform for systematic functional interrogation of the DDR associated genes. Using glioblastoma, we demonstrate that the library reliably identifies biologically significant and clinically relevant genetic dependencies through multiple orthogonal validation approaches. As a reusable platform rather than a disease-specific tool, DDRKOL can be broadly applied across diverse biological contexts to discover context-dependent DDR vulnerabilities, therapeutic targets, and mechanisms of treatment resistance.