To explore the genetic characteristics of familial Parkinson’s disease (PD)-related genes and construct a model of the gene signatures to guide the prognosis and treatment of patients with skin cutaneous melanoma (CMM).

① By analyzing the TCGA and GTEx databases, the expression differences of familial PD-related genes were compared between melanoma and normal tissues. The GEPIA2 database was used to analyze the expression differences of related genes across different TNM stages and Clark levels. ② Based on the TCGA database, a signature model containing 7 key genes was established using prognostic forest plot, LASSO regression model, and risk factor analysis. Internal and external validation of this model was performed, and GO enrichment and KEGG pathway enrichment analyses were conducted for these genes. ③ Based on the TIMER database, the impacts of gene mutations in the prognostic model on immune cell infiltration in the tumor microenvironment (TME) were analyzed. ssGSEA heatmaps were employed to analyze differences in immune cell subsets between high-risk and low-risk groups. Through PD-1 immune checkpoint inhibitor treatment response analysis, external dataset validation, and half-maximal inhibitory concentration (IC50) values of different drugs, possible differences in drug treatment efficacy between high-risk and low-risk groups were investigated. ④ Western blotting and RT-qPCR were utilized to detect the protein expression and relative mRNA expression differences of the genes in this prognostic model between normal human melanocyte cell line PIG1 and human melanoma cell line A375.

① Analysis of the TCGA and GTEx databases revealed that, except for PARK14 (PLA2G6), all familial PD-related genes showed significant expression differences between melanoma tissues and normal tissues. Analysis of the HPA database found that the protein expression of familial PD-related genes was either up-regulated or down-regulated in melanoma samples compared to normal tissues. Analysis of the GEPIA2 database for the expression of these genes across different TNM stages and Clark levels also found differential expression in some genes. ② COX regression analysis based on TCGA database identified that expression levels of 13 genes could predict overall survival in melanoma patients. LASSO regression analysis of this data constructed a melanoma prognostic model containing 7 familial PD-related genes (LRRK2, UCHL1, SNCA, VPS13C, SPR, EIF4G1, FBXO7). The prognostic risk scores for all patients were calculated using a risk factor contingency plot, and then the patients were divided into high-risk and low-risk groups based on the median risk score (-0.145). The results showed that the high-risk group had a significantly worse survival rate, while the low-risk group exhibited a longer survival time. Internal validation of the model in the TCGA cohort: the calibration curves for the 3-year and 5-year nomograms showed that the predicted curves for melanoma patient prognosis were very close to the actual curves. External validation in GEO cohorts: 3 datasets from the GEO database containing complete clinical data and gene expression matrices (GSE19234, GSE22153, and GSE53118) were selected. Based on the prognostic model's risk score, the sensitivity and specificity on the ROC curve were better than TNM staging. GO and KEGG pathway enrichment analyses found that the 7 genes were mainly enriched in pathways related to tumorigenesis and tumor immunity, such as phosphorylation, programmed cell death, and cell cycle, suggesting the prognostic model is associated with tumorigenesis and immunity. ③ GSVA scoring revealed that these 7 genes were enriched in pathways related to tumorigenesis and anti-tumor immunity. CD8⁺ T cells, cytotoxic cells, and Th1 cells, which can inhibit tumor growth, were more infiltrated in the low-risk group. Meanwhile, the IC₅₀ values for nilotinib, BIBW2992, bleomycin, bortezomib, dasatinib, and MK-2206 were lower in tumor cells from low-risk group patients (P<0.05). ④ RT-qPCR and Western blotting analysis were used to verify the relative expression differences of these 7 genes at the transcriptional and translational levels in A375 and PIG1 cells. The experimental results showed that LRRK2 and UCHL1 were almost not expressed in both A375 and PIG1 cells. Meanwhile, compared to normal melanocytes, the mRNA expression of SNCA, VPS13C, SPR, eIF4G1, and FBXO7 genes was significantly up-regulated in melanoma cells (P<0.05). The protein expression of SPR, EIF4G1, and FBXO7 was significantly upregulated (P<0.05).

The 7-gene signature prognostic model developed in this study can effectively assess prognosis and treatment efficacy in CMM patients.