Identification Of Prognostic Gene In Metastatic Melanoma Based On WGCNA

Background and aims: Skin cutaneous melanoma(SKCM)is the major cause of death for skin cancer patients,its high metastasis often leads to poor prognosis of patients with malignant melanoma.However,the molecular mechanisms underlying metastatic melanoma remain to be elucidated.It has been established that the metastatic ability of melanoma is regulated by an intricate gene interconnection network.In this study we aim to identify and validate prognostic biomarkers associated with metastatic melanoma.Methods: We first use the large-scale public gene expression profile from Gene Expression Omnibus(GEO)to construct a co-expression network,and then use weighted gene co-expression network analysis(WGCNA)to screen out clinically important modules and hub candidate genes.To construct a network and to identify hub genes,the gene expression profiles of GSE22153 dataset are downloaded from GEO database,which includes 57 metastatic melanoma cases associated with 4 molecular subtypes(high-immune response,pigmentation,normal-like and proliferative).In order to reveal the role of candidate hub genes in the pathogenesis of metastatic melanoma,Gene Ontology(GO)enrichment analysis and Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway analysis are performed.RNA-seq data and clinical information on metastatic melanoma from GSE22154 and The Cancer Genome Atlas(TCGA)are used for further screening for hub genes related to prognosis.Then,we analyze the overall survival rate of patients in the GEO dataset and TCGA,and screen out hub genes related to prognosis.Gene Set Enrichment Analysis(GSEA)further clarify the potential biological functions of hub gene in metastatic melanoma.In addition,we further verify the role of hub gene in metastatic melanoma through in vitro experiments.Results: We first construct a co-expression network using large-scale public gene expression profiles from GEO,from which candidate genes are screened out using WGCNA.A total of eight modules are established via the average linkage hierarchical clustering,module eigengenes of the blue,yellow,brown and turquoise modules are proved to have the highest correlation with the subtypes(high-immune response,pigmentation,normal-like and proliferative,respectively),which are accordingly selected as the clinically significant module for further analysis.Next,111 hub genes are identified from the clinically significant modules.In addition,through GO enrichment analysis and KEGG pathway analysis,we find that these 111 hub genes are mainly related to the growth and structure of the skin.Next,two other datasets from GEO and TCGA are used for further screening of biomarker genes related to prognosis of metastatic melanoma,and the genes with statistical significance to prognosis in both datasets are determined as the biomarkers related to survival.Via two independent survival analysis we find the low expression of the 11 genes AOAH,CD48,IL32,CORO1 A,GPR132,IL10 RA,ITGAL,LCK,LCP1,RCSD1 and TBC1D10 C are associated with poor prognosis of patients with metastatic melanoma.Besides,we find that IL10 RA has the highest correlation with clinically important modules among all identified biomarker genes.Further in vitro biochemical experiments,including CCK8 assays,wound-healing assays and transwell assays,have verified that IL10 RA can significantly inhibit the proliferation,migration and invasion of melanoma cells.Furthermore,GSEA shows that PI3K-AKT signaling pathway is significantly enriched in metastatic melanoma with highly expressed IL10 RA,indicating that IL10 RA mediates in metastatic melanoma via PI3K-AKT pathway.Conclusions: In conclusion,our WGCNA analysis identifies candidate prognostic biomarkers for further basic and advanced understanding of the molecular pathogenesis of metastatic melanoma.