Abstract
The 90-Kda heat shock protein (Hsp90), together with a number of other chaperones are involved in normal cellular differentiation and cancerogenesis. Hsp90 is implicated in the conformational maturation of a large variety of protein kinases. We have shown its overexpression in a series of 116 acute myeloid leukemias (AML). Geldanamycin and its analogue 17-AAG selectively block the activities of Hsp90, but do not interact with other members of the Hsp family. The objective of the study was to test the effect of 17-AAG on cell lines (Jurkat, U-937, HL60, HL60R : resistant to daunorubicin (DNR), and KG1a) and on 24 AML samples tested positive for Hsp90 (14 strongly positive and 10 moderately positive). CD34+ cells from bone marrow donors were used as controls. Cells were maintained in liquid culture in the presence or absence of 17-AAG. Moreover hematopoietic and leukemic progenitor assays were performed in semi-solid media containing or not 17-AAG. CD34+ cells were non affected by 17-AAG. Jurkat, U937, HL60 were partially sensitive to 17-AAG. Addition of low dose DNR to 17-AAG induced an apoptotic death in 48 hours in all samples, including HL60R. KG1a (exhibiting high HSP90 levels) was more sensitive to 17-AAG alone with a death rate of 100% at higher doses. A similar effect was observed in AML samples : 17-AAG induced cell death in all 14 cases highly positive for HSP90, but a partial death in 5 cases, and no death in 5 cases mildly positive for HSP90. DNR increased apoptosis in only 5/10 of these cases. The other 5 patients were also positive for bcl-2. When they were first incubated with 17-AAG and then treated with bcl-2 antisens oligodeoxynucleotides (bcl-2AS), a complete apoptosis was obtained. So, in the 10 samples with moderate HSP90 expression a synergystic effect was found when a combination of 17-AAG with either chemotherapy or bcl -2AS was used. While normal progenitors were not affected by 17-AAG, leukemic progenitors were completed inhibited in the 14 patients who responded to 17-AAG in liquid culture. This apoptosis was caspase -3 dependant. Finally, we investigated the implication of the ERK signal pathway in Jurkat line and in 14/24 17-AAG responder samples using an ERK inhibitor (PD98059). When this inhibitor was added prior to 17-AAD and DNR, the apoptotic effect was completely inhibited. So, ERK signal pathway may interfere with the Hsp90 function. In conclusion, these results raise the possibility that Hsp90 antagonists represent a novel antileukemic strategy in sequential addition to conventional chemotherapy
Author notes
Corresponding author