CBIO-17. IDENTIFYING A NETWORK OF ESSENTIAL & TUMORIGENIC GENES IN GLIOBLASTOMA USING WHOLE-GENOME CRISPR Cas9 SCREENING

Abstract

Abstract GBM is the most common primary malignant brain tumor in adults, with a 100% recurrence rate and a median survival of 21 months. As such, advances in therapy are desperately needed. Genomic approaches have shown that GBM has high intra-tumoral heterogeneity. However, a comprehensive understanding of determinants of growth is required to identify new therapeutic targets. CRISPR-Cas9 screening technology has enabled whole-genome screens that allow objective identification of genes governing specific phenotypes. Here, we performed a genome-wide CRISPR knockout screen in H4 human glioma cells to identify genes that drive proliferation. Our screen identified ~150 novel essential growth genes. From this list, we identified 5 genes that were previously unstudied, show significant elevations in expression at the RNA and protein levels (p< 0.05), and show significant survival benefit in patient datasets (p< 0.05) – PSMB3, CHCHD4, THOC1, SPDYE5, HSPA1. Our validation experiments showed that knocking out these genes resulted in cell death in multiple GBM patient-derived xenograft (PDX) lines. In addition, animals with knockout cells implanted demonstrated extended survival (p< 0.01). Furthermore, overexpression of these genes in a normal neural stem cell line resulted in transformation to a cancer phenotype, as evidenced by sphere formation in a soft agar assay (p< 0.01). Further investigation of one of these genes, PSMB3, which is a subunit of the proteasome, showed that ubiquitinated proteins were significantly increased after PSMB3 knockdown, suggesting that disruption of the proteasome system was the likely cause of cell death. We performed a ubiquitin immunoprecipitation (IP) to identify which genes were being uniquely ubiquitinated in GBM. Results showed pathways involving retinoblastoma genes, DNA repair genes, and DNA replication. Together, this data suggests that both our CRISPR screens and our ubiquitin IP have yielded promising and novel therapeutic targets for GBM, a disease desperately in need of new strategies.