Most human melanomas of the superficial spreading and nodular clinical subtypes contain activating mutations in either the NRAS small GTPase or the BRAF protein kinase. The same mutations occur at approximately the same frequency in melanocytic nevi, which can be precursors to melanoma. To understand further the role of activated BRAF in human melanoma, we are pursuing an observation we made that BRAF in human melanoma cells interacts with the molecular chaperone heat shock protein (HSP)-90. HSP-90 has previously been demonstrated to interact with another form of RAF, CRAF. Disruption of the interaction between CRAF and HSP-90 with the natural product geldanamycin, an inhibitor of HSP90 ATPase activity, or its semi-synthetic analog, 17-AAG, destabilizes CRAF and inhibits its ability to direct activation of the downstream MAP kinase signaling cascade. We have analyzed the effect of 17-AAG upon BRAF stability and downstream MAP kinase signaling in human melanoma cell lines. 17-AAG at concentrations up to 1 micromolar appeared to destabilize BRAF in 2/6 human melanoma cell lines examined but destabilized CRAF in 4/6 of these cell lines. Despite the persistence of BRAF in the majority of cell lines treated with 17-AAG, downstream activation of the MAP kinase cascade was inhibited, shown by absence of MEK and ERK phosphorylation, in each of the cell lines exhibiting and in 2 additional human melanoma cell lines examined. We are interested in the mechanism by which 17-AAG inhibits BRAF-dependent activation of MEK and ERK without inducing BRAF degradation and are testing hypotheses about this mechanism primarily using a biochemical approach. To analyze the effects of expressing melanoma-specific mutations in NRAS and BRAF in melanocytes, we are developing an tetracycline-inducible transgenic mouse system using the dopachrome tautomerase (Dct) promoter to permit the controlled induction of gene expression in vivo in murine melanocytes. Using the Dct promoter may render it possible to induce expression of these genes early in embryogenesis as well as in the putative melanocyte stem cell that has been described in the bulge region of the murine hair follicle. Some of this work is being done in collaboration with Dr. Glenn Merlino of NCI. Initial characterization experiments suggest that expression of transgenes is induced in the expected location of the murine follicle. More extensive characterization of transgenic lines will be required to describe the transgene expression pattern and inducibility of expression. Inducible lines should be quite useful for studying the effects of expression of activated mutants of BRAF and NRAS in isolated murine melanocytes, and establishing the propensity of these mutations to cause melanoma formation in appropriate genetic backgrounds.
