Abstract
A subset of non-small cell lung cancers (NSCLC) are dependent upon oncogenic driver mutations, including the most frequently observed driver mutant KRAS, which is associated with a poor prognosis. As direct RAS targeting in the clinic has been unsuccessful to date, use of Hsp90 inhibitors appeared to be a promising therapy for KRAS-mutant NSCLC; however, limited clinical efficacy was observed due to rapid resistance. Furthermore, the combination of the Hsp90 inhibitor (Hsp90i), ganetespib, and docetaxel was tested in a phase III clinical trial and failed to demonstrate benefit. Here, we investigated the mechanism(s) of resistance to ganetespib and explored why the combination with docetaxel failed in the clinic. We have not only identified a critical role for the bypass of the G2-M cell-cycle checkpoint as a mechanism of ganetespib resistance (GR) but have also found that GR leads to cross-resistance to docetaxel. Reactivation of p90RSK and its downstream target, CDC25C, was critical for GR and mediated the bypass of a G2-M arrest. Overexpression of either p90RSK or CDC25C lead to bypass of G2-M arrest and induced ganetespib resistance in vitro and in vivo Moreover, resistance was dependent on p90RSK/CDC25C signaling, as synthetic lethality to ERK1/2, p90RSK, or CDC25C inhibitors was observed. Importantly, the combination of ganetespib and p90RSK or CDC25C inhibitors was highly efficacious in parental cells. These studies provide a way forward for Hsp90 inhibitors through the development of novel rationally designed Hsp90 inhibitor combinations that may prevent or overcome resistance to Hsp90i. Mol Cancer Ther; 16(8); 1658-68. ©2017 AACR.