Abstract PR09: A systems biology approach to elucidate novel drug targets in KRAS mutant tumors

Abstract

Abstract Mutations of RAS GTPase family oncogenes affect more than 20% of human cancers. Of the three major RAS isoforms (HRAS, KRAS, and NRAS), KRAS is the most frequently mutated, affecting 57% of pancreatic adenocarcinomas, 33% of colorectal cancers, 17% of lung adenocarcinomas, and many other tumor subtypes. Most RAS mutations abrogate its GTPase activity, thus activating aberrant downstream signals associated with worst prognosis and tumor resistance to both chemotoxic and targeted therapy. Therapies against RAS have failed, in part, because of our limited understanding of RAS-mediated signal transduction pathways. Therefore, we hypothesize that the systematic and complete dissection of RAS relevant regulatory interactions can help elucidate druggable mechanisms to abrogate its aberrant activity or elicit synthetic lethality in RAS mutated tumors. Here, we have employed computational systems biology approaches to begin to dissect KRAS regulatory pathways. Specifically, we reverse-engineered a cell context specific KRAS regulatory network from TCGA transcriptome-wide expression data. The KRAS regulatory network was used to predict KRAS activation levels across all colon and lung adenocarcinoma samples available in TCGA in order to generate a patient-by-patient estimate of KRAS activity. Genome-wide somatic mutations and copy-number variations were then tested for significant association with the predicted KRAS activity. Consistent with known KRAS biology, samples harboring KRAS or BRAF activating mutations were significantly associated with increased KRAS activity. However, we also observed that mutations in the DEAD box helicase DDX6 and the Glycolipid Transfer-domain protein GLTPD2 to be significantly associated, suggesting that these proteins may be involved in KRAS pathways. We dissected the effects of KRAS mutations on the activity of all transcription factors, identifying those more closely activated or repressed in a KRAS-specific context. KRAS is activated by its own mutations, along with known KRAS interactors, such as SOS1 and RAPGEF6. Interestingly, we found that these activating KRAS mutations significantly correlate with increased activities of the Rho-associated kinases ROCK1 and ROCK2. Moreover, the protein-protein interaction database PrePPI predicts KRAS to homodimerize, and surprisingly, to interact with ROCK1/2 through a new PPI interface. We are currently confirming these results through biochemical and biophysical analyses. This abstract is also presented as Poster B34. Citation Format: David A. Wah, Federico M. Giorgi, Donald Petrey, Mariano J. Alvarez, Tonya Silkov, Josh Broyde, Gabrielle E. Rieckhof, Barry Honig, Andrea Califano. A systems biology approach to elucidate novel drug targets in KRAS mutant tumors. [abstract]. In: Proceedings of the AACR Special Conference on RAS Oncogenes: From Biology to Therapy; Feb 24-27, 2014; Lake Buena Vista, FL. Philadelphia (PA): AACR; Mol Cancer Res 2014;12(12 Suppl):Abstract nr PR09. doi: 10.1158/1557-3125.RASONC14-PR09