DDDR-70. Unlocking the Genetic Code of Chemotherapy Response in Glioblastoma

Abstract Glioblastoma (GBM) remains one of the deadliest brain tumors, with minimal and short-lived response to temozolomide (TMZ), the frontline chemotherapeutic. To uncover the genetic circuitry driving this resistance, we conducted a large-scale in vivoCRISPR-Cas9 loss-of-function screen targeting 2,500 genes involved in transcriptional regulation, RNA splicing, and epigenetic modification. Using 8 sgRNAs per gene across multiple collaborative laboratories, our unbiased screen revealed a striking spectrum of genes whose disruption either conferred resistance or heightened sensitivity to TMZ. Among resistance-associated hits were RNF10, WHSC1, SMAD2, GFI1, ZNF551, TRIM33, MNDA, HNF1A, POGZ, EID2B, and ARID1A—with ARID1A emerging as a top candidate regulating epigenetic plasticity and therapy evasion. Conversely, genes such as ASCL1, SNAI2, MED12, ATF7, VRK1, ZNF26, HOXB8, SEBOX, and ZSCAN4 sensitized tumors to TMZ. For functional validation, sgRNAs from GeCKO and Brunello libraries were transfected into U251 GBM cells, with knockout efficiency confirmed by Western blot. TMZ response was quantified using CellTiter-Glo viability assays across drug gradients. Strikingly, loss of resistance-driving genes sensitized GBM cells to TMZ, while knockout of sensitizer genes reversed this effect—confirming their functional roles in chemoresistance. To dissect spatial dynamics, we employed high-resolution spatial transcriptomics, mapping the expression patterns of resistance- and sensitivity-associated genes across the tumor microenvironment. This integration revealed spatially defined gene programs contributing to therapy failure or response. Together, our work decodes key regulators of chemoresistance in GBM and nominates a set of high-priority, druggable targets—such as ARID1A—for therapeutic development. These insights open new avenues for overcoming treatment resistance and improving outcomes for GBM patients.