Fusion gene products such as PML-RARalpha and BCR-ABL generated by leukemia-specific chromosomal translocations have been identified as target molecules for the treatment of leukemia. Here we describe one possibility for extending the frontier of mechanism-based medicine for acute myeloid leukemia (AML). FLT3, a receptor tyrosine kinase (RTK) preferentially expressed in hematopoietic progenitor cells, frequently has a gain-of-function mutation in AML. To search for FLT3-targeted compounds, we screened the growth-inhibitory effects of several tyrosine kinase inhibitors (TKIs) on mutant FLT3-transformed 32D cells. Herbimycin A at a concentration of 0.1 microM markedly inhibited the growth of the transfectants but at that concentration was ineffective in parental 32D cells. It suppressed the constitutive tyrosine phosphorylation of the mutant FLT3, but not the phosphorylation of the ligand-stimulated wild-type FLT3. In mice transplanted with transformed 32D cells, the administration of herbimycin A completely prevented leukemia progression. Recent studies have indicated that herbimycin A binds directly with HSP90, a molecular chaperone, and destabilizes HSP90-associated proteins. Another HSP90 inhibitor, radicicol, also induced apoptosis selectively in transformed 32D cells. HSP90 is a promising target for the treatment of AML with mutant FLT3.