Abstract 5214: Reconstructing targetable pathways in KRAS dependent lung cancer by integrating transcriptome, proteome and phosphoproteome.

Abstract

Abstract Activating mutations in the Ras oncogenes characterize 20-40% of all non-small cell lung cancer (NSCLC), the leading cause of cancer mortality in the United States, which establishes Ras genes as the most commonly mutated oncogenes in this malignancy. Mutations in the KRAS oncogene characterize more than 20% of all NSCLC malignancies and recent studies suggest that those tumors can be divided into KRAS dependent (KRAS-Dep) and KRAS independent (KRAS-Ind), according to their dependency on KRAS mutant for cell survival and proliferation. Although, the KRAS signaling pathway has been considerably studied using high throughput transcriptomic or proteomic technologies in isolation, the specific network of effector genes and proteins that drives carcinogenesis in NSCLC is still far from being understood. Integrative profiling with different omics technologies harbors the potential for characterizing the specific network of effector proteins associated with specific oncogenes. We developed a bioinformatics approach to reconstruct active and targetable networks associated with KRAS dependency in non-small cell lung cancer (NSCLC). By integrating matched global transcriptomics, proteomics and phosphoproteomics datasets, we stratified NSCLCs into those dependent on, or independent of, KRAS signaling by defining the S score that combines transcript, protein and phosphoprotein abundance in order to identify differential expressed proteins. We demonstrate that the S score outperforms naïve integration methods and improves the functional specificity of the proteins identified as differential expressed. Next, we reconstruct targetable network modules associated with KRAS dependency by employing the Prize Collecting Steiner Tree (PCST) algorithm to formulate the module reconstruction. This reconstruction identified three conspicuous modules centered on KRAS and MET, LCK and PAK1, and B-Catenin. We validated activation of those proteins in KRAS dependent cell lines and performed functional studies defining LCK as a critical gene to the cell proliferation in KRAS-dependent, but not KRAS-independent, NSCLCs. We further define a functional module of LCK signaling in NSCLC, in which LCK activates downstream proteins such as PAK1 in order to regulate cell death. Finally, we find evidence of LCK signaling in human NSCLCs tissues. These results are the first evidence that suggest LCK as potential targetable protein in the context of KRAS dependency. In summary, our integrative analysis establishes a novel mechanistic basis for KRAS dependency in NSCLC and nominates KRAS-LCK-PAK1 network module as a potentially druggable pathway in KRAS-dependent lung cancers. Citation Format: Alejandro O. Balbin, John Prensner, Anirban Sahu, Anastasia Yocum, Sunita Shankar, Mohan Dhanasekaran, Xuhong Cao, Alexey Nesvizhskii, Arul Chinnaiyan. Reconstructing targetable pathways in KRAS dependent lung cancer by integrating transcriptome, proteome and phosphoproteome. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 5214. doi:10.1158/1538-7445.AM2013-5214