Mouse models have been an essential tool for cancer research for many decades, and have seen application to ever more diverse and complex questions with the advent of genetic engineering technologies. However, little has been done to characterize the genome-wide profiles of mutations in tumors generated from some of the most commonly used mouse models of cancer. Here we used whole-exome sequencing to compare the landscape of somatic mutations in carcinogen-induced and genetically engineered mouse models of Kras-driven lung cancer. We show that these models adopt distinct routes to tumor development, with carcinogen-induced tumors having a much higher mutational burden and better recapitulating the mutational landscape seen in human lung cancer. In addition, the exome-wide mutation spectra in these tumors display highly consistent and distinct signatures of the initiating carcinogen, and demonstrate the potential utility of mouse sequencing studies to discern complex mutation signatures of suspected carcinogens.
The wild type (WT) allele of Kras has been shown to act as a tumor suppressor in mouse models of lung cancer, and we demonstrate that urethane-induced tumors from normal mice carry mostly Q61R Kras mutations, while those from Kras heterozygous mice carry mostly Q61L mutations, indicating a major role of germline Kras status in mutation selection during initiation. In order to gain a more comprehensive view of the genetic underpinnings of Kras mutation selection, we characterized Kras mutations in carcinogen-induced lung tumors from a large population of Mus musculus x Mus spretus backcrossed mice. Genome-wide quantitative trait locus (QTL) mapping identified a novel locus on chromosome 10, in addition to the Kras locus, that modulates Kras mutation selection. We also investigated the relationship between the tumor suppressive effect of WT Kras and its role in modulating mutation selection, and found that both isoforms of Kras (Kras4A and Kras4B) contribute to these phenomena, while canonical tumor suppressive pathways do not seem to be involved. These approaches provide a framework for understanding the interactions between germline and somatic events that control Kras mutation specificity in cancer.