The Genomic Landscape Of Kras Mutant Genetically Engineered Mouse Models Of Human Cancers

Genetically engineered mouse models (GEMM) of cancer are invaluable tools for oncology research. They recapitulate critical aspects of tumor etiology and allow for unique interrogation of tumor evolution and maintenance. In addition, these models are poised for assessing therapeutic drug efficacy and resistance mechanisms. Although these models are widely used, the mutational landscape and transcriptional profile of GEMMs have yet to be thoroughly characterized. In this study we sought to investigate the precise genomic landscape of three well-validated GEMMs representing two types of Kras mutant cancers, non-small cell lung cancer (NSCLC) and pancreatic ductal adenocarcinoma (PDAC). Using next generation sequencing techniques, we profiled an adenovirally-induced Kras mutant NSCLC model (KrasLSL.G12D/+;p53frt/frt) and two Kras mutant models of spontaneous PDAC: KPP (KrasLSL.G12D/+;p16/p19fl/fl;Pdx1.Cre) and KPR (KrasLSL.G12D/+;p16/p19fl/+;p53LSL.R270H/+;Pdx1.Cre). All three GEMMs incorporate genomic aberrations that are documented to occur in the human patient population. Additionally these models have been reported to approximate histological and phenotypic aspects of their human correlate. We carried out whole exome sequencing for each model (NSCLC n = 115; PDAC n = 73 and n = 40, respectively) to comprehensively characterize the genomes. Interestingly, both intra and inter-tumoral genomic heterogeneity was evident both within and amongst Kras mutant tumor models. The models most significantly diverge with respect to copy number alterations. The NSCLC model demonstrated whole chromosome 6 gain and focal deletion on chromosome 9, in addition to other genomic alterations. However, the KPP PDAC model tumors exhibited significant focal amplification of mutant Kras in 75% (55/73) of cases. In combination with RNA-seq data, we can show that these genomic alterations correlate with divergent transcriptional signatures despite the shared Kras mutation, especially in MAPK pathway signatures. This data suggests tumors within these models have unique signaling pathway dependencies. Finally, we compared the tumors resulting in the murine models to that of Kras altered human tumors from TCGA at the genomic level. The genomic landscapes of Kras mutant GEMMs and corresponding Kras mutant patients share significant similarities. We identified mutations in oncogenes and tumor suppressors, in addition to alterations in gene copy number. Taken together, the results provide a clearer understanding of the genomic landscape of these GEMMs and their relationship with human patients. More importantly, these analyses will allow us to better interpret results from previous and future experiments using these Kras mutant GEMMs of cancer. Citation Format: Wei-Jen Chung, Anneleen Daemen, Jason Long, Jason Cheng, Chris Tran, Zora Modrusan, Oded Foreman, Melissa Junttila. The genomic landscape of Kras mutant genetically engineered mouse models of human cancers. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2686.