Abstract

The distinction between benign and malignant melanocytic neoplasms can be challenging on the basis of histomorphology. While cytogenetic and molecular assays help address this challenge, they are less readily available and have greater turn-around-time compared to immunohistochemistry. PRAME has emerged as a useful marker to support a diagnosis of melanoma, but there are limitations to its sensitivity and specificity. To address these limitations we used single cell RNA sequencing (scRNA-seq) data from 19 malignant melanomas (GSE72056) and identified 247 genes expressed in PRAME-negative melanoma cells, with functions in MYC signaling, protein metabolism, and cell metabolism and energetics (OR >= 3.2, p < 0.05). In a separate cohort of 53 melanomas and 22 benign nevi (GSE112509), 48 of these 247 genes were upregulated in melanomas compared to benign nevi (adj. p < 0.01), and were enriched in PI3K/AKT/mTOR signaling and tumor invasiveness genes (OR >= 13.9 , adj. p < 0.05). Using data from the Human Protein Atlas, we subsequently filtered out genes with high protein expression in benign skin. We identified 32 candidate genes expressed in PRAME-negative melanoma cells and upregulated in melanomas compared to benign nevi, including genes previously implicated in melanoma pathogenesis. This study forms the basis for additional work to find biomarkers useful for the distinction of nevi from melanoma.

Financial Disclosure: No current or relevant financial relationships exist.