Site-directed nucleases have recently emerged as potent tools for gene disruptions in various model organisms, including disruptions of tumor suppressor genes. However, whole organism tumor suppressor gene disruption is time-consuming and often causes deleterious effects such as lethality. To mitigate these problems, we have developed somatic cell gene-targeting approaches. Our strategy is to target candidate tumor suppressor genes specifically in melanocytes. Using the miniCoopR system, in which transgene-expressing melanocytes are reconstituted in a mitfa(lf) background, we have expressed TALENS in melanocytes and validated targeted knockdown of p53 in melanomas. In order to achieve greater targeting efficiency, we are using CRISPR-Cas system. With our improved genome-editing system we aim to target candidate tumor suppressors that are recurrently deleted in human and zebrafish melanomas. Additionally, we are assessing the function of these candidate tumor suppressors in human melanoma cell lines using down-regulation and over-expression studies and will present our progress. Combining the throughput afforded by genetics in zebrafish with functional in vitro studies has enabled our identification of novel melanoma oncogenes. We now extend that approach to tumor suppressor genes.