| Metastatic melanoma has proven exceptionally difficult to treat, in part due to consistent defects in cell death pathways. The overexpression of antiapoptosis proteins from the Bcl-2 family is a hallmark of melanoma progression. Past strategies targeting these proteins have assumed equivalence between Bcl-2, Bcl-xL and Mcl-1, closely related anti-apoptotic factors. Recent evidence, however, has shown that Bcl-2 family members display different binding affinities and survival activities. To assess the specific function of Bcl-2, Bcl-xL and Mcl-1 in melanoma cell survival, we used a genetic approach based on targeted RNA interference. We found that Mcl-1 is the key determinant of melanoma resistance to multiple chemotherapeutic agents. Importantly, Mcl-1 is dispensable for normal melanocyte viability, offering a window for therapeutic intervention. We sought to utilize these findings to overcome the resistance of melanoma cells to proteasome inhibition. Proteasome inhibitors have shown great promise in treating certain cancers, but are ineffective as single agents against melanoma. Selective drug inhibition of Mcl-1, in combination with proteasome blockade, lead to the effective killing of melanoma cells, both in vitro and in vivo. Our studies suggest a functional hierarchy of Bcl-2 anti-apoptosis proteins that could be exploited to overcome melanoma resistance to proteasome inhibition.;More broadly, we sought to identify new mechanisms of melanoma chemoresistance. Electron microscopy of melanoma cells revealed the presence of autophagosomes in cells surviving after treatment with various chemotherapeutic agents. Multillamelar structures surrounding cellular organelles were particularly evident upon treatment with cyclopamine, a drug previously characterized as an inhibitor of Sonic Hedgehog signaling. These results were unexpected, as the Sonic Hedgehog pathway was not previously known to influence autophagy. The aberrant structures were sustained for extended periods, during which cells lacked proliferative capacity but remained viable. Notably, although early steps of autophagy were present in cyclopamine-treated cells, the later stage of autophagosome-lysosome fusion was absent. We hypothesized this autophagy blockade might provide a therapeutic opportunity. Indeed, forcing autophagy with hypoxia or low serum in the setting of cyclopamine treatment induced efficient melanoma cell killing. A better understanding of autophagy, a previously unexplored pathway in melanoma, may reveal new targets for therapy. |
