Approximately half of human melanomas originate from precursor melanocytic proliferations termed melanocytic nevi. However, most melanocytic nevi do not undergo malignant transformation, but rather remain stable throughout life or disappear. To understand the determinants of melanocyte transformation, especially as they apply to the development of primary human melanoma from melanocytic nevus precursor lesions, we are planning a patient-based study with two goals. One is to define the clinical characteristics of early primary melanomas with greater precision, using skin epiluminescent microscopy, or dermoscopy, to visualize specific structures that can be correlated with lesional histopathology. The other is to sample and immortalize melanocytes from clinically benign acquired and congenital melanocytic nevi, in order to study signalling pathways and gene expression in these cells to gain more insight into a common first step in the process of melanocyte transformation. One potential benefit of this study is that the clinical knowledge gained about early primary melanomas, combined with the development of techniques to immortalize adult human melanocytes, may make it possible in a subsequent study to obtain and expand cells from these early malignant lesions to understand this stage of malignant transformation. This is especially important since most human melanoma cell lines used today are derived either from advanced primary tumors or metastases, and have little detailed clinical annotation associated with them. A clinical protocol detailing the procedures for this study has passed internal NCI scientific review and is being readied for submission to the NCI IRB. Furthermore, we have begun to culture normal human melanocytes from neonatal foreskins in the lab to provide material to test immortalization techniques using genes delivered from lentiviral vectors. We have obtained and prepared plasmids required to generate these lentiviral vectors. One of the genes we plan to use to immortalize these cells is a gene called Bmi-1, encoding a transcription factor of the polycomb group. Interestingly, a comparison of Bmi-1 expression on the protein level between normal human melanocytes and most melanoma cell lines showed a marked difference in expression between these two types of cells, with expression much higher in the melanoma cell lines than in their normal counterparts. This difference in expression is not apparent on the RNA level, however, with similar amounts of Bmi-1 mRNA detected in normal melanocytes and melanoma cells. We hypothesize that differences in Bmi-1 catabolism in these cell types, perhaps regulated by differential ubiquitination, or that differences in translational efficiency, perhaps regulated by microRNA(s), may account for the markedly different levels of protein expression observed in the setting of similar levels of message. We are planning experiments to test these hypotheses. At the same time, we are interested in whether the high level of expression of Bmi-1 observed in most human melanoma cell lines is functionally important for the maintenance of the malignant phenotype. We have begun preparations for experiments using RNAi to assess the importance of Bmi-1 expression in melanoma cells.
