Identification Of Core And Context-Specific Fitness Genes In Glioblastoma Stem Cells Via Genome-Wide Crispr-Cas9 Screens

Abstract Advances in the characterization of the genetic, epigenetic and transcriptomic landscapes of glioblastoma (GBM) tumor samples and primary cell lines have yielded considerable insight into tumor initiation and progression. A major finding from these studies is the high degree of inter- and intra-tumoral variability displayed in GBM, which has increased interest in the identification of clinically relevant disease subtypes for precision targeting. While multiple groups have identified subtypes of GBM tumors and GBM stem cell lines, efforts to identify specific actionable targets have yet to bear fruit. To address this gap, we performed genome-wide TKO CRISPR-Cas9 fitness screens to functionally interrogate the genomes of a panel of 16 genetically/epigenetically characterized GBM stem cell lines. Our screens identify over 1,500 fitness genes per GBM stem cell line which collectively provide a functional landscape of genetic vulnerabilities in GBM. The size and heterogeneity of our screen panel provided us with the capability to identify both core (i.e. SOX9, Protein Ufmylation, JUN) and context specific fitness genes. Specifically, we have identified genes preferentially required for the proliferation of GBM stem cells based on transcriptionally defined subtypes (i.e. SOX2, OLIG2 in Proneural GSCs), genetic alterations (i.e. CDK6 in CDKN2A/B deleted GBM) and disease stage (i.e. ATP6AP2 in recurrent GBM). Experimental validation of key subtype-specific fitness genes is currently underway using genetic knockout and drug/inhibitor treatment. Our data provides biological insight and novel mechanistic understanding of heterogeneity in GBM and points to opportunities for precision targeting of defined GBM subtypes.