Modeling the Genesis and Progression of Human Melanoma in the Mouse

The major goal of the Cancer Modeling Section of the Laboratory of Cancer Biology and Genetics was to elucidate the complex molecular and genetic programs governing tumor initiation and metastatic progression through the development and analysis of better, more relevant genetically engineered mouse models of human cancer. Our aim was to employ such mouse models, and the tissues and cells derived from them, to uncover the molecular mechanisms underlying these processes, and identify candidate factors or pathways that may ultimately serve as therapeutic targets in patients. Our research efforts in this regard were focused predominantly on cutaneous malignant melanoma for several reasons. Melanoma is an extremely aggressive, often fatal disease that has proven to be largely resistant to currently available therapeutic approaches. Compounding this health hazard, the incidence of melanoma has steadily increased over the last 40 years, and continues to rise at a time when the incidence of many other cancers is falling. Melanoma is particularly well suited to our major research interests. Its etiological initiating factor, ultraviolet (UV) radiation, is long known but its role in melanoma genesis remains poorly understood. Moreover, its propensity to metastasize, even years after removal of the original primary tumor, makes this cancer especially deadly.

We created a mouse model, based on deregulated signaling of the receptor tyrosine kinase MET, that when provoked by UV radiation develops melanocytic lesions in stages that are highly reminiscent of human melanoma with respect to both histopathologic architecture and molecular wiring. This mouse is being used to experimentally address questions that have long perplexed the melanoma research community: is childhood sunburn a significant melanoma risk factor; is UV-B or UV-A the more influential melanoma agent; does sunblock prevent melanoma genesis? We used this mouse model to determine what UV radiation actually does to melanocytes and how it induces melanoma genesis. For example, by designing mice in which melanocytes express green fluorescent protein in a tetracycline-inducible fashion, we can use fluorescence-activated cell sorting to isolate melanocytes from mice following UV irradiation and subject them to detailed molecular analyses, including array-based expression profiling and deep sequencing to search for mutations. Other mice we generated in which Cre recombinase is conditionally targeted to melanocytes allowed us to determine which tumor suppressor genes prevent progression from each stage of melanoma.

We were particularly interested in mechanisms associated with the spread and survival of metastatic cells. All data derived from mouse melanomas were compared with their human counterparts to determine the relevance of our findings. Ultimately, we planned to use our melanoma mouse model to perform preclinical studies, aimed at an examination of the role of cancer stem cells in drug-resistant residual metastatic disease.

Our collaborators included Ron DePinho, Dana-Farber Cancer Institute; Frances Noonan and Ed DeFabo, George Washington University; Paul Meltzer, NCI; Giorgio Trinchieri, NCI.