Constructing and deconstructing cancer using CRISPR-Cas9

Abstract

Cancer is a genetic disease that arises through the sequential acquisition of genetic and epigenetic alterations in oncogenes and tumor suppressor genes. Large-scale efforts to re-sequence protein-coding genes from human cancer cell lines and tumor biopsies have begun to catalog the spectrum of mutations existing in human cancers. One major limitation of these studies has been the inability to rapidly and systematically determine which of these mutations are causally related to tumorigenesis, particularly in the context of in vivo models of the disease. Although existing genetically engineered mouse models (GEMMs) have led to critical insights into the initiation and progression of human cancer, their use for rapid functional characterization of new cancer genes has been historically limited, partly due to the cost and time required to generate appropriate murine models. In the first part of this thesis, I describe a novel CRISPR-Cas9-based approach for rapid functional investigation of candidate genes in vivo using well established autochthonous mouse models of cancer. By employing this platform in a GEMM of lung adenocarcinoma in vivo, I have functionally validated both known and novel tumor suppressor genes - all of which promote one or more aspects of lung cancer. These findings underscore the power and versatility of this platform for the rapid and functional interrogation of the cancer genome. In the second part of this thesis, I describe the development of a novel CRISPR-Cas9-based genetic screening approach for systematically uncovering genotype-specific vulnerabilities that could be exploited for the therapeutic benefit of specific lung cancer patient subpopulations. I demonstrate the successful application of this system for the discovery of novel Keap1 mutant-specific genetic dependencies, many of which could potentially be pursued for the clinical benefit of patients whose tumors harbor loss of function mutations in KEAP1 (~17% of lung adenocarcinoma patients). These results demonstrate the power of CRISPR-based genetic screens for uncovering novel genetic dependencies in the context of clinically relevant cancer-associated genotypes.