Rationale: In our search for radiation modifiers, molecular targets which affect cellular proliferation, cell cycle progression, and pro-survival signaling pathways are often the most efficacious. Indomethacin and other non-steroidal anti-inflammatory drugs (NSAIDs) affect these processes, and our preliminary results indicate that it may occur through a heat shock protein (HSP)-related mechanism. These proteins are often involved in conformation and maintaining distribution of client proteins throughout the cell following a stressful challenge. Particularly, HSP90 appears important, as it coordinates the function of over 100 proteins, many of them pro-survival factors, and is constitutively overexpressed in tumor cells. To abrogate HSP90 function, therefore, may also equate to enhancement of tumor cell response to therapies. We began studies with the ansimycin antibiotics geldanamycin (GA) and its demethoxy analogue (17-AAG) because of their specificity for the ATP-binding domain of HSP90, where client protein interaction occurs, and because each are clinically effective at doses of a magnitude less than other agents, including NSAIDs. Research Synopsis: We explored the utility of the HSP90 molecular chaperone in its client protein shuttling capacity both independently and in conjunction with exogenous stress. Clonogenic and apoptotic assays show that GA and 17-AAG are cytotoxic agents unto themselves at modest doses and behave as potent radiosensitizers in vitro and in vivo. These effects likely occur through abrogation, inactivation, and/or selective degradation of client proteins necessary for tumor cells to effectively respond to stressful stimuli. Furthermore, surviving cells following HSP90 inhibition indicate a lack of proliferation capability and a late G2 or M phase cell cycle arrest. This also affects the radiation response and, from our results, appears to be rooted in a similar abrogation of cell cycle-related proteins as observed in signal transduction machinery (Bradbury et al, in preparation). Interestingly, this effect appears to be through a p53-dependent mechanism. Cells made to overexpress HPV16-E7, a characteristic of most cervical carcinomas, are functionally p53-null and respond more completely to GA and 17-AAG than do benignly transformed cells. In this regard, GA and 17-AAG may specifically target tumor cells for destruction, leaving normal cells unharmed. 17-AAG is presently in Stage II clinical trials for other treatment modalities; expansion of these trials to include therapeutic radiation is presently underway.
