Verification of a multivariate pharmacodynamic biomarker assay for HSP90 inhibitors

B16

This study is designed to identify and verify heat shock protein 90 (HSP90) pharmacodynamic RNA biomarkers and to develop a single multiplex panel to monitor them in a clinical setting. HSP90 is a ubiquitously expressed molecular chaperone which is involved in multiple cellular functions, such as signal transduction and protein folding. Together with other co-chaperones, HSP90 participates in many key processes in oncogenesis including growth factor independence, stabilization of mutant proteins, angiogenesis, and metastasis, by regulating the maturation and stability of its client proteins. Among HSP90 clients are Her2, v-Src, Bcr-Abl, p53 mutant, VEGF, NOS, MMP2, AKT, and steroid receptors. Multiple cancer drugs have targeted HSP90 to inhibit its activity in cancerous cells, including geldanamycin and its derivatives, and ridicicol. In pre-clinical and clinical development, HSP70, CDK4 and other proteins have been used as single biomarkers to monitor drug activity and response. We have developed a multivariate biomarker development system based on our proprietary multiplex PCR technology, which utilizes Beckman Coulter’s GeXP CEQ platform. Utilizing this process we have identified a set of RNA biomarkers whose in vitro expressionin a inhibitor sensitive cell line is significantly altered by the HSP90 inhibitors 17-N-Allylamino-17-demethoxygeldanamycin (17-AAG) and Ridicicol. Their expression changes in a dose and time dependent manner and the changes are not compound specific, suggesting that this panel may have broad utility for therapeutics that target the HSP90 pathway. We have produced a single multiplex assay that is capable of monitoring all of these targets in a single reaction. We have now expanded this study to include additional sensitive and insensitive cell lines for verification, as well as carried out IC50 assays to analyze for correlation to our gene expression results. Additionally, we have designed and carried out experiments to determine the correlation between data obtained in vitro and the drug response seen in vivo. We have successfully verified these biomarkers and correlated their activity to responses measured in IC50 assays. We now plan to begin validation of in vivo responses with human clinical samples.