Collections > Electronic Theses and Dissertations > An ex vivo Familial Genetic Strategy for Determining Mechanism of Action

One of the greatest challenges in anticancer drug development is the discovery of molecular targets and biochemical interactions required for drug action. Lapses in drug efficacy and unanticipated toxicity, the two biggest causes of drug failure in clinical trials, are often attributed to our limited understanding of drug mechanism and cost the pharmaceutical industry millions. Genomics is rapidly emerging as tool for mechanism elucidation. Our approach is one of the latest to link drugs to the genes which influence their activity. This ex vivo familial genetics strategy uses a collection of extensively genotyped, normal, healthy, human cell lines from multigenerational families. Cell lines are phenotyped for cytotoxic response to anticancer agents, heritability analysis gives a measure of the degree to which genetic influences response, and linkage analysis suggests regions of the genome which are associated with the observed variation in response. To evaluate this strategy as method for mechanism elucidation, we first asked whether the system could produce pharmacological and genomic profiles related to a shared mechanism for a class of structurally related compounds. The in vitro sensitivity of CEPH cell lines the camptothecin, Topoisomerase 1 inhibitors (Top1), was studied. Heritability analysis estimates that genetics accounts for as much 20% of the observed variation in cytotoxic response to these drugs. Linkage analysis revealed a pattern of six quantitative trait loci (QTLs) that were shared by all of the camptothecins and independently replicated with a second of camptothecin analogues. The pattern of QTLs observed with the camptothecins was compared to those of the indenisoquinolones, a structurally distinct class of Top1 inhibitors. The objective was to identify which if any QTLs are related to the general mechanism of Top1 inhibition or should be considered class-specific. Finally, the model was assessed for its ability to stratify compounds by mechanism based on their biological and genomic profiles. Cell lines were phenotyped for response to approximately 30 drugs belonging to 8 mechanistic classes. Intraclass biological and genomic profiles were more similar to each other than to compounds belonging to distinct mechanistic classes. This work could have a significant impact on drug discovery and development as it provides a strategy for not only making predictions about mechanism of action for novel therapies, but for identifying genes involved in variable response to chemotherapeutic agents as well.