Identification of Cancer-specific Therapeutic Targets and Tumor Suppressor Genes in Glioblastoma Multiforme by Functional Genetics

Abstract

Glioblastoma Multiforme (GBM) is the most common and aggressive form of brain cancer. ~90% of adult GBM patients receiving standard of care therapies die within 2 years of diagnosis due to ineffective therapies. To identify new therapeutic targets for GBM, we performed lethal genome-wide short-hairpin RNA (shRNA) and RNA-guided clustered regularly interspaced short palindromic repeats-Cas9 (CRISPR-Cas9) knockout screens in patient-derived GBM stem-like cells (GSCs) and also non-transformed human neural stem cells (NSCs), non-neoplastic tissue of origin controls. Previous experiments with RNA interference (RNAi) identified multiple molecular vulnerabilities specific to GSCs in processes that include kinetochore regulation and 3' pre-messenger RNA splice site recognition. It is likely that additional genes contributing to these or other biological processes could yield similar or superior cancer-lethal effects. From an unbiased RNAi viability screen to putative transcription factors in GSCs and NSCs, we identified a new kinetochore protein, BuGZ/ZNF207, that is differentially required for expansion in the GSCs, but not in NSCs. Inhibition of BuGZ results in loss of both BUB3 and BUB1 from kinetochores, reduction of BUB1-dependent phosphorylation of centromeric histone H2A, attenuation of kinetochore-based Aurora B kinase activity, and chromosome congression defects in cancer cells. From genome-wide CRISPR-Cas9 knockout screens, we identified multiple GSC-sensitive genes, including the WEE1-like kinase protein PKMYT1, which is essential to most GBM isolates, but not NSCs. Molecular and mechanistic studies revealed that PKMYT1 acts redundantly with WEE1 to inhibitory phosphorylate CDK1-Y15 and to promote timely completion of mitosis in NSCs, but that this redundancy is lost in most GBM isolates and in NSCs harboring activated alleles of EGFR and AKT1. PKMYT1 depletion in GSCs and genetically altered NSCs requiring PKMYT1 lead to cytokinesis failure and cell death during mitosis. In addition, CRISPR-Cas9 knockout screens revealed multiple genes required for in vitro expansion of NSCs, including: ARID1A, ARID1B, CREBBP, EP300, NF2, PDCD10, PTPN14, TAOK1, TGFBR2, and TP53. Knockout of these genes caused shortened cell cycle transit times and drastic growth advantages in NSCs, and in the case of CREBBP knockout, caused precocious entry into S-phase and deregulation of cell cycle gene expression. Together, these functional genetic studies identify novel cancer therapeutic targets and growth suppressor genes in human GBM isolates and NSCs that will direct the development of therapeutics to these cancer-specific cellular targets and complexes for cancer patients.