Abstract 1119: Systems pharmacology approaches to study tumor drug mechanism of action

Abstract

Abstract Precision cancer medicine is predicated on the ability to identify drugs whose mechanism of action (MoA) matches one or more key tumor dependencies. Although the structure of many target proteins is available and their binding affinity to specific compounds can be experimentally assessed, drug efficacy and toxicity are infrequently determined based on single target affinity. Rather, the pharmacologic property of drugs is often the result of its complex polypharmacology, as mediated by both unknown or poorly characterized off-target effects (i.e., lower-affinity binding proteins) as well as by tissue-specific secondary effector proteins (i.e., non-binding) that are undetectable by traditional assays. MoA elucidation typically relies on in vitro affinity binding assays tailored to specific protein classes, such as kinases or metabolic enzymes, which cannot reveal critical high- and low-affinity binding targets among other protein classes, as well as context-specific downstream effectors. The goal of this study is to address critical unresolved questions in cancer pharmacology by characterizing the proteome-wide MoA of oncology drugs as a critical, yet highly elusive step necessary to fully understand their clinical efficacy and toxicity. To elucidate proteome-wide, drug-mediated changes in protein activity (drug MoA), we performed network-based analyses of genome-wide RNA-seq profiles representing the response of patient-matched cell lines to a comprehensive repertoire of clinically relevant drugs. More specifically, we generated genome-wide drug perturbation profiles from 23 cancer cell representing high-fidelity models of patients in clinical cohorts representing distinct tumor subtypes, using > 700 oncology drugs. This represents the largest resources of functionally annotated, clinically-relevant, genome-wide perturbational profiles for clinically relevant drugs, named PanACEA. VIPER-based analysis of this resource effectively elucidated the effect of each individual drug on the activity of ~6,500 regulatory and signaling proteins, in each of the 23 distinct represented cancer subtypes. Analyses of these data, using a graph theory approach, helped elucidate critical functional relationship between individual drugs and group drugs into functionally-distinct modules within each cancer context, thus providing critical insight into both drug MoA and polypharmacology. These analyses also effective drug inhibitors for critical cancer dependencies represented by transcription factors and co-factors, considered undruggable, that were experimentally validated. Finally, integrative analysis of PANCEA predictions and Cancer Dependency Map data significantly improved our understanding of cancer-related drug pharmacology, thus providing a unique resource for cancer related studies, which has been used in key clinical and pre-clinical trials, both published and in review. Citation Format: Lucas Zhongming Hu, Eugene Douglass, Ron Realubit, Karan Charles, Mariano Alvarez, Andrea Califano. Systems pharmacology approaches to study tumor drug mechanism of action [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1119.