The PI3K-pathway in tumorigenesis and cancer drug resistance

The PI3K-pathway is frequently deregulated in cancer; specifically, its catalytic subunit, p110alpha, is mutant in 26% of breast cancers. The pharmaceutical industry is seeking specific inhibitors targeting PI3K, some of which have already entered clinical trials. Despite their great promise, a justified concern is the emergence of drug resistance.;At least two distinct genetic mechanisms of drug resistance have been documented, involving either mutations in the drug target, or deregulation of parallel or downstream pathways. Consequently, resistance to PI3K-inhibition may be conferred by one of these mechanisms. To select for resistance to the dual PI3K/mTOR inhibitor BEZ235, we performed a selective pressure screen using growth in the presence of the drug. We used immortalized human mammary epithelial cells (HMECs), an engineered cell-system highly dependent on PI3K. Upon isolating a population of cells resistant to the inhibitor, we found that a dominant contributor to the drug resistance is genomic amplification of MYC. We could demonstrate that MYC is both sufficient and necessary to drive resistance to PI3K and/or mTOR inhibition.;Notably, a direct downstream effector of PI3K is PDK1, a kinase amplified in a fraction of breast cancers, whose causative role in cancer has not been firmly established. As a key downstream effector, PDK1 may be an attractive drug target in tumors with hyperactivated PI3K-pathway. We investigated if PDK1 can contribute to breast tumorigenesis and serve as a necessary downstream effector for transformation by upstream lesions. We could demonstrate that PDK1 can transform HMECs with efficiency similar to oncogenic PI3K, and in addition, that it is necessary for transformation by oncogenic alleles of PI3Kalpha and RAC1. Furthermore, we identified a truncated splice variant of the AGC-kinase MSK1 as a novel downstream effector of PDK1.;Collectively, this work suggests that PDK1 could act as a causative transforming oncogene in human disease and that amplification of MYC may be a driver of resistance to PI3K inhibition. These findings may aid in selecting patients who would benefit from PI3K-targeted therapy.