148 POSTER Comparison of the cellular and biochemical properties of ansamycin and non-ansamycin based Hsp90 inhibitors

Abstract

Background: Heat shock protein 90 (Hsp90) has emerged as an important target for the treatment of cancer due to its essential role in several key oncogenic signaling pathways. Several classes of Hsp90 inhibitors have recently advanced into clinical trials including ansamycin derivatives that are semi-synthetic derivatives of the natural product geldanamycin (e.g. 17-AAG, IPI-504, 17-DMAG) or synthetic small molecules (e.g. purine derivatives, isoxazoles, pyrazoles). Ansamycin derivatives are potent Hsp90 inhibitors that demonstrate selective cell growth inhibition toward cancer cells as compared to normal cells. We have determined the biochemical and cellular properties of a group of published Hsp90 inhibitors, including both natural product derived and synthetic compounds. Materials and methods: The biochemical affinity of inhibitors to Hsp90 was determined using a competition binding assay using radioactively labeled competitor compound and Hsp90 purified from Hela cells. The growth inhibition induced by Hsp90 inhibitors was evaluated in human normal and cancer cells by measuring cell growth in the presence of varying concentrations of compounds using the Alamar Blue assay after 72h of compound addition. Results: The biochemical affinities to purified Hsp90 for the inhibitors tested range from 0.1 to 500 nM. There is an approximate correlation between biochemical affinity and cell growth inhibition of cancer cells with a relative activity ranking of isoxazoles > ansamycins > purines. While ansamycins demonstrate selective growth inhibition of cancer cells compared to normal cells as previously described in the literature, the more potent isoxazole derivative also potently inhibits the growth of some normal cell types. Three of the most potent compounds on cancer cells were also tested in vivo for their ability to induce the degradation of a Hsp90 client protein (mutEGFR) in a xenograft model of human NSCLC. In this model, we observe similar suppression of mutant EGFR and similar induction of cleaved caspase 3 for all three compounds, independent of their biochemical affinity for purified Hsp90. One possible explanation for this observation is provided by our measurements of off-rates of inhibitors from Hsp90. Conclusion: The experiments presented above raise the question of whether synthetic Hsp90 inhibitors with high affinity for Hsp90 have lost some of the in vitro therapeutic window between cancer and normal cells that makes Hsp90 inhibitors such attractive candidates for cancer therapeutics. Inhibitor Ki (nM)