单细胞与批量 RNA 数据的整合分析揭示了黑色素瘤相互作用组的复杂性与重要性

Background: Despite significant strides in anti-melanoma therapies, resistance and recurrence remain major challenges. A deeper understanding of the underlying biology of these challenges is necessary for developing more effective treatment paradigms. Methods: Melanoma single-cell data were retrieved from the Broad Single Cell Portal (SCP11). High-dimensional weighted gene co-expression network analysis (hdWGCNA), CellChat, and ligand-receptor relative crosstalk (RC) scoring were employed to evaluate intercellular and intracellular signaling.

The prognostic value of key regulatory genes was assessed via Kaplan-Meier (KM) survival analysis using the 'SKCM-TCGA' dataset. Results: Twenty-seven (27) gene co-expression modules were identified via hdWGCNA. Notable findings include NRAS Q61L melanomas being enriched for modules involving C19orf10 and ARF4, while BRAF V600E melanomas were enriched for modules involving ALAS1 and MYO1B .

Additionally, CellChat analysis highlighted several dominant signaling pathways, namely MHC-II, CD99, and Collagen-receptor signaling, with numerous significant ligand-receptor interactions from melanocytes, including CD99-CD99 communications with cancer-associated fibroblasts, endothelial cells, NK cells, and T-cells. KM analysis revealed that higher expression of SELL , BTLA , IL2RG , PDGFA , CLDN11 , ITGB3 , and SPN improved overall survival, while higher FGF5 expression correlated with worse survival.

Protein-protein interaction network analysis further indicated significant interconnectivity among the identified prognostic genes. Conclusions: Overall, these insights underscore critical immune interactions and potential therapeutic targets to combat melanoma resistance, paving the way for more personalized and effective treatment strategies.