Acquired resistance to HSP90 inhibitor 17-allylamino-17-demethoxygeldanamycin (17-AAG) in adult and pediatric glioblastoma cell lines

2461

Introduction. The molecular chaperone heat shock protein 90 (HSP90) is an exciting anticancer drug target due to its role in maintaining the stability and function of oncogenic client proteins. The HSP90 inhibitor 17-allylamino-17-demethoxygeldanamycin (17-AAG), in phase II clinical trials, binds to the N-terminal of HSP90 and inhibits its ATPase activity. 17-AAG inhibits the growth of glioblastoma (GB) cell lines. However, drug resistance account for poor GB prognosis. Objective. To explore acquired resistance to HSP90 inhibitors in GB cells. Methods and results. By continuous exposure to increasing 17-AAG concentrations, acquired resistance to 17-AAG was generated in both adult (SF268, U87MG) and pediatric (SF188, KNS42) GB cell lines. High levels of resistance with resistance indices (RI=IC50 of resistant line/IC50 of parental line; sulforhodamine B assay) of 20-137 were obtained rapidly (2-8 weeks). After cessation of 17-AAG, RI were decreased (6-25), except for the pediatric line KNS42. After 7-26 weeks, the RI remained stable in all cell lines. Cross-resistance was found with other 17-AAG related ansamycin benzoquinones (17-DMAG, 17-AG) but not with structurally unrelated HSP90 inhibitors, radicicol and our synthetic potent resorcinylic pyrazole/isoxazole amide compounds (VER-49009, VER-50589). An inverse correlation between NAD(P)H:quinone oxidoreductase 1 (NQO1) expression (western blot) / activity (enzymatic assay) and 17-AAG IC50 was observed with the resistant lines. The NQO1 inhibitor ES936 abrogated the differential effects on 17-AAG sensitivity between the parental and resistant lines. These results suggested that NQO1 plays a major role in 17-AAG resistance in GB cells. However, after 17-AAG cessation in the pediatric resistant line SF188-RA6, NQO1 protein/activity returned to parental level although 17-AAG resistance was retained, suggesting an additional mechanism of resistance, albeit restricted to ansamycin benzoquinones. NQO1 mRNA levels (quantatitive RT-PCR) inversely correlated with 17-AAG IC50 in the adult resistant lines, but levels remained stable in the pediatric resistant lines, suggesting different mechanisms of NQO1 down-regulation between adult and pediatric lines. NQO1 was also the mechanism of resistance in an acquired 17-AAG resistant melanoma cell line. Interestingly, we were unable to generate acquired resistant lines to the resorcinylic pyrazole/isoxazole amide agents. Conclusion. Low NQO1 activity is a likely mechanism of acquired resistance to 17-AAG in GB and in other cell types (e.g. melanoma), underlining the problematic metabolism of 17-AAG. New series of HSP90 inhibitors, such as the resorcinylic pyrazole/isoxazole amide analogues, are able to circumvent this mechanism of resistance, providing additional support for their clinical development.