Deciphering the prognostic and therapeutic effects of ion channel genes in the occurrence and progression in SKCM

Abstract BACKGROUND: Skin cutaneous melanoma (SKCM) is a cutaneous malignancy with a poor prognosis, and it is highly malignant and aggressive, making it the skin cancer with the highest mortality rate. OBJECTIVE: The present study was designed to explore the effect of prognostic modeling on SKCM occurrence and prognosis through ion channel genes. METHODS: Gene expression data and related clinical information of patients with SKCM were obtained from the TCGA and GEO databases, respectively, and the TCGA data were included as a training group. The related ion channel genes were detected from the ion channel database.A one-way cox survival analysis of ion channel modification-related genes was undertaken to screen for prognostically relevant ion channel genes using the 'survival' software package. Ion channel-related features were built with the LASSO-Cox regression model and validated against external datasets for accuracy and reproducibility in predicting prognosis in SKCM patients. Univariate and multivariate Cox analyses were undertaken in the training set to observe if the feature was independent of traditional clinical variables. The independence and robustness of signature were further validated using stratified Kaplan-Meier analysis of clinical variables. We identified gene modules with strong positive correlations in the group of low risk and performed GO/KEGG analysis of gene with strong positive correlations in the modules. Butterfly plots were then applied to vividly show the correlations between risk scores and TIP scores, eight immunotherapy prediction scores and different tumour signalling pathways. Differences in four different groups of immune-related genes (immune checkpoints, chemokines, immune cell markers, and HLA) in the groups of low- and high-risks were also compared. We identified differences in immune infiltration between the groups of low- and high-risks with 7 software, as well as genes significantly co-expressed in the group of high risk following the WGCNA algorithm, and performed functional enrichment analyses of the genomes to probe potential mechanisms of SKCM occurrence and prognosis. Finally, we assessed the sensitivity of chemotherapy versus immunotherapy in the groups of low- and high-risks. RESULTS: Through univariate Cox analysis, we secured 330 genes, of which 30 genes were statistically significantly correlated with survival. Prognostic models of 14 genes were constructed through Lasso-Cox analysis. Signature, which consists of 14 ion-channel-related genes, had good predictive effect on SKCM 1-, 2-, and 3-year DSS. Signature is an independent prognostic factor for SKCM and its predictive effect is superior to that of traditional clinical variables. dca suggests that using our model to predict survival in SKCM patients could benefit patients. The gene enrichment in the blue module of adaptive immune-related pathways, biological processes, molecular functions, and cellular components was found to be negatively related to risk scores by immunoenrichment, and significantly positively related to the low-risk group in this study. TIP scores and 8 immunotherapy scores were negatively related to risk scores, and therefore better predicted the response to immunotherapy. The EGFR and VEGF pathways were positively related to risk scores. Therefore, there was significant heterogeneity between the groups of low- and high-risks. Correlations between genes and immune cell content in a large number of models suggests that ion channel-related genes possibly affect the pathogenesis and prognosis of SKCM by modulating the immune microenvironment. These findings revealed that the signature was a significant marker of immune rejection. Using this signature, patients in the group of low risk could be predicted to have increased immune reserve. We also concluded that the group with lower risk was a better candidate for chemotherapy by chemotherapy-related analysis. CONCLUSIONS: The results of this study indicated that we had uncovered a prognostic model of prognostic value for SKCM patients. It also revealed that ion channel-related genes were highly correlated with the immune status and prognostic survival of patients with SKCM, and they might modulate SKCM pathogenesis in multiple ways. The results of this study would help to clarify how to develop and expand new targets for the effective management and treatment of SKCM.