The resulting statistics of recurrent melanoma indicate a dearth of effective treatments for this disease. In order to identify targets for improved therapy and to better elucidate the mechanisms underlying the pathogenesis of this disease, we are performing gene expression profiling of human melanoma biopsies. Using a technique known as the serial analysis of gene expression (SAGE), which allows for the quantification of individual transcripts within a cell, as well as identifies novel transcripts, we can examine the gene expression profiles of melanomas at different stages of disease using minute quantities of tissue. Currently, three of these libraries are available on the public databases (http://cgap.nci.nih.gov), which we generated from three different melanoma tissue samples, with two representing vertical growth phase tumors and one representing a distant tumor metastasis. Studies currently in progress include the generation of libraries from compound nevi, additional metastases and normal melanocytes. We can examine these data using microarray analysis tools rarely utilized to analyze SAGE libraries. These types of multilevel analyses can reveal which genes underlie progression and which genes may be viable targets for future therapy. ? ? Thus far, data generated by microarray and SAGE analyses have highlighted the importance of G-protein mediated signaling, resulting in the activation of PKC and rises in intracellular calcium in melanoma progression. These effects can be mediated by the gene WNT5A, the over expression of which can lead to increases in melanoma cell motility and invasion. RNAi inhibition of this pathway, followed by microarray analysis, reveals that Wnt5a may make this contribution to invasion by silencing the expression of metastasis suppressers such as Kiss-1 and NME-1. Furthermore, we have also shown that Wnt5a can mediate melanoma metastasis via the initiation of the epithelial to mesenchymal cell transition, in a PKC-dependent fashion. In addition, Wnt5a can suppress the expression of melanoma antigens, such as MART1, implying it may play a role in escaping immune surveillance. Recent data suggests that this occurs via the activation of STAT3. In the light of recent data from other labs showing that targeting MART1 can cause tumor regression, these results are of significant importance and may provide a means to control MART antigen expression in tumors to facilitate immune responses against or the molecular targeting of such tumors.? ? Another gene implicated by the SAGE expression data is Claudin 1. We have found that claudin 1 is highly expressed in melanomas and that it's expression is regulated via PKC. Furthermore, increased expression of claudin 1 results in increases in invasion in melanoma cells, with concomittant increases in the matrix metalloproteinase enzymes MMP9 and MMP2. Staining of a tissue microarray indicates that this protein is upregulated in melanoma as compared to benign nevi. These data provide evidence of both an upstream regulator of PKC (Wnt5a) as well as a downstream target of this enzyme (claudin 1), both of which contribute to the increased invasiveness in melanoma. It is hoped that identifying important pathways such as these with high throughput gene expression profiling techniques will lead to identifying better molecular targets for the treatment of this disease.