Abstract
Recent studies indicate that the development of drug resistance and increased invasiveness in melanoma is largely driven by transcriptional plasticity rather than canonical coding mutations. Understanding the mechanisms behind cell identity shifts in oncogenic transformation and cancer progression is crucial for advancing our understanding of melanoma and other aggressive cancers. While distinct melanoma phenotypic states have been well characterized, the processes and transcriptional controls that enable cells to shift between these states remain largely unknown. In this study, we initially leverage the well-established zebrafish melanoma model as a high-throughput system to dissect and analyze transcriptional control elements that are hijacked by melanoma. We identify key characteristics of these elements, making them translatable to human enhancer identification despite the lack of direct sequence conservation. Building on our identification of a zebrafish sox10 enhancer necessary for melanoma initiation, we extend these findings to human melanoma, identifying two human upstream enhancer elements that are critical for full SOX10 expression. Stable biallelic deletion of these enhancers using CRISPR-Cas9 induces a distinct phenotype shift across multiple human melanoma cell lines from a melanocytic phenotype towards an undifferentiated phenotype and is also characterized by an increase in drug resistance that mirrors clinical data including an upregulation of NTRK1, a tyrosine kinase, and potential therapeutic target. These results provide new insights into the transcriptional regulation of SOX10 in human melanoma and underscore the role of individual enhancer elements and potentially NTRK1 in driving melanoma phenotype plasticity and drug resistance. Our work lays the groundwork for future gene-based and combination kinase-inhibitor therapies targeting SOX10 regulation and NTRK1 as a potential avenue for enhancing the efficacy of current melanoma treatments.