| Description |
Complex cancer phenotypes are defined by their aggressive nature and lack of known or accessible therapeutic targets. My dissertation focuses on the use of a personalized medicine approach for the identification of novel therapies against two complex cancer phenotypes: Basal-like/Claudin-low breast cancer and RAS-active nonsmall cell lung cancer. RAS-active cancer is characterized by the activation of the complex signaling network of RAS, which lacks effective therapeutics capable of inhibiting the RAS protein itself or the overall pathway. Further complicating treatment is the ability of the RAS pathway to be activated independent of the presence of an activating mutation in the RAS protein. To broadly characterize pathway activation independent of RAS protein mutation, I used a gene-expression-based biomarker for RAS network activity in nonsmall cell lung cancer (NSCLC) cell lines, and identified RAS activation in both RAS-mutant and wild-type lines. I then screened for drugs whose efficacy significantly correlated to RAS network activity and showed that EGFR and MEK co-inhibition is an effective treatment personalized against RAS-active NSCLC. Finally, I demonstrated that EGFR and MEK co-inhibition induced apoptosis and blocked both EGFR-RAS-RAF-MEK-ERK and EGFR-PI3K-AKT-RPS6 nodes simultaneously in RAS-active, but not RAS-inactive NSCLC. Secondly, I identified a novel compound effective against Basal-like and Claudin-low breast cancer (BL-CL). BL-CL is a molecular subtype of breast cancer characterized by an aggressive, recurrent and nonluminal nature, epitomized by the lack of known therapeutic targets and poor patient prognosis. Using high- iv throughput screening of a marine invertebrate compound library and sequential purification of crude fractions, I identified a previously uncharacterized sulfated sterol, Topsentinol L Trisulfate (TLT), purified from a marine sponge, and showed that it inhibits AMPK and CHK1 but activates p38. Furthermore, I indentified the potential use of known AMPK and CHK1 inhibitors, alone or in combination, as an effective therapy against BL-CL. Lastly, sensitivity to TLT was projected against various human tumors by generating a gene-expression signature that predicted breast and bladder cancer as the cancer types most receptive to TLT therapy. This work describes the identification of novel treatments personalized against BL-CL and RAS-active NSCLC, providing a framework for future pre-clinical studies. |
