Abstract
KRAS-activating mutations drive human non-small cell lung cancer and initiate lung tumorigenesis in genetically engineered mouse (GEM) models. However, in a GEM model of KRAS(G12D)-induced lung cancer, tumors arise stochastically following a latency period, suggesting that additional events are required to promote early-stage tumorigenic expansion of KRAS(G12D)-mutated cells. PI3Kα (PIK3CA) is a direct effector of KRAS, but additional activation of PI3'-lipid signaling may be required to potentiate KRAS-driven lung tumorigenesis. Using GEM models, we tested whether PI3'-lipid signaling was limiting for the promotion of KRAS(G12D)-driven lung tumors by inducing the expression of KRAS(G12D) in the absence and presence of the activating PIK3CA(H1047R) mutation. PIK3CA(H1047R) expression alone failed to promote tumor formation, but dramatically enhanced tumorigenesis initiated by KRAS(G12D). We further observed that oncogenic cooperation between KRAS(G12D) and PIK3CA(H1047R) was accompanied by PI3Kα-mediated regulation of c-MYC, GSK3β, p27(KIP1), survivin, and components of the RB pathway, resulting in accelerated cell division of human or mouse lung cancer-derived cell lines. These data suggest that, although KRAS(G12D) may activate PI3Kα by direct biochemical mechanisms, PI3'-lipid signaling remains rate-limiting for the cell-cycle progression and expansion of early-stage KRAS(G12D)-initiated lung cells. Therefore, we provide a potential mechanistic rationale for the selection of KRAS and PIK3CA coactivating mutations in a number of human malignancies, with implications for the clinical deployment of PI3' kinase-targeted therapies.