A Holistic Approach Towards Cancer Drug Discovery: Targeting Synthetic Lethal Interactions And Chemoresistance Pathways

Abstract

Personalized cancer medicine focuses on the predictive value of genetic profiling, allowing for the classification and specialized treatment of individuals in subpopulations who share similar genetic alterations observed in tumors. Colorectal cancer is the second leading cause of cancer deaths in the United States and defects in DNA mismatch repair due to mutations in mismatch repair genes MSH2 and MLH1 play a key role towards the development of ~20% of hereditary and sporadic colorectal cancers. Inactivation of mismatch repair genes results in genomic instability and the subsequent accumulation of somatic mutations. Furthermore, since a defective mismatch repair system confers a “mutator phenotype” on the cell, the resulting cancers are less responsive to standard chemotherapeutic treatments and more likely to become chemoresistant.

This dissertation describes a novel method for determining the genetic basis of anticancer drug resistance. Since the high mutation rate of mismatch repair defective cells typically results in the generation of a highly heterogeneous population of cells, the addition of a compound to the heterogeneous culture will result in the selection and growth of the cells bearing the mutation(s) required to grow in the presence of the compound. For this reason, the mismatch repair defective mutator strain, msh2Δ, was used to identify the genetic basis of drug resistance to clinically relevant anticancer agents. Whole genome sequencing was used to uncover the genes responsible for conferring resistance. This work also describes a high throughput approach for developing therapeutic strategies for targeting mismatch repair defective tumors. This dissertation provides a basis for the use of yeast as an experimental model system in personalized cancer drug discovery.