Systematic Interrogation of Cancer Driver Gene Function

Abstract

A comprehensive catalog of the genetic alterations found in human cancer is emerging through large-scale tumor sequencing efforts. These studies have confirmed that many tumors harbor mutations or copy-number alterations in well-characterized tumor suppressors and oncogenes and have also identified hundreds of other alterations that occur at lower frequencies. However, in many cases, the roles these alleles play in tumorigenesis remain incompletely understood. New genetic perturbation tools have been developed that enable systematic high-throughput assessment of gene function in experimental models of cancer. These functional approaches provide the means to uncover the roles of cancer-associated alleles in oncogenesis, and to query the entire genome to identify tumor suppressors, oncogenes, non-oncogene dependencies, and drivers of therapeutic resistance. In the present work, we applied unbiased functional genetic approaches to study the tumor suppressor TP53 and the oncogene ERBB2. TP53 (p53) is the most frequently somatically mutated gene in cancer, and more than 1,500 different mutations have been identified. We systematically interrogated the TP53 mutational spectrum in cancer by (1) silencing wild-type or mutant TP53 in hundreds of cancer cell lines, (2) testing every possible TP53 missense and nonsense mutation in isogenic p53WT and p53NULL cell lines, and (3) estimating the baseline probability of acquiring each TP53 mutation using gene-agnostic mutational signatures. Our findings indicate that cancers select for loss-of-function and dominant-negative TP53 mutations that are generated by specific mutational processes. ERBB2 (HER2) is amplified in a subset of breast cancers, and although patients with this genetic alteration often benefit from drugs that target HER2, resistance is common. We systematically interrogated the genetic underpinnings of the response to anti-HER2 therapy by performing genome-scale gain- and loss-of-function screens. We identified dozens of resistance genes, most of which encode components of the RTK, RAS, PI3K, or SRC signaling pathways. Importantly, we found that pharmacological inhibition of these pathways could restore sensitivity to HER2 inhibition in the resistant setting. In summary, our work illustrates the power of using functional genetic screens to interrogate biological systems and indicates that mutational processes and phenotypic selection play important roles in shaping cancer genomes.