Integrative Single-Cell and Bulk Transcriptomic Analysis Defines a Molecular Stratification Strategy Based on...

Acral melanoma (AM) is a rare subtype of melanoma with high malignancy. Although the transcriptional landscape of the tumor microenvironment (TME) in AM has been characterized, the heterogeneity of malignant cells and underlying molecular features remain to be fully delineated. We analyzed 23 AM samples across three single cell RNA sequencing (scRNA-seq) datasets and quantified the intra-tumoral melanoma cell heterogeneity (ITMH) based on the degree of deviation for individual cells in principal component space. We compared the molecular characteristics of melanoma cells and TME between high-intra-tumoral-heterogeneity (ITMHhi) and low-intra-tumoral-heterogeneity (ITMHlo) groups. Differentially expressed genes (DEGs) in melanoma cells were extracted and a single sample Gene Set Enrichment Analysis (ssGSEA)-based differential scoring method was applied to stratify patients into ITMHhi and ITMHlo subtypes in bulk RNA-seq cohorts. Weighted Gene Co-expression Network Analysis (WGCNA) was employed to explore the gene modules related to tumor heterogeneity. ScRNA-seq analysis revealed diverse transcriptional heterogeneity among melanoma cells within and between patients. ITMHhi and ITMHlo samples distributed separately in the transcriptional space. We observed that the heterogeneity of melanoma cells was correlated with genomic, immunological, and clinical characteristics. The ITMHhi group exhibited distinct molecular programs and high genome instability in tumor cells, as well as high progenitor exhausted and inflammatory microenvironment. We also defined ITMH-related AM subtypes with different gene expression patterns and clinical outcomes in bulk RNA-seq data. Hub genes associated with high heterogeneity were identified by DEG and WGCNA analysis, and a prognostic risk model was developed to predict patients' survival based on the hub genes. Our study provided insight into understanding the intra-tumoral heterogeneity of AM and its potential impact on patient stratification, TME remodeling and prognosis prediction.