Integrated analysis of diagnostic, prognostic value and potential drug treatment of GSDME in head and neck squamous cell carcinoma.

BACKGROUND: Head and neck squamous cell carcinoma (HNSC) poses a global health challenge. Effective biomarkers for early detection are necessary to improve the survival rate of HNSC patient. The purpose of this study was using integrated bioinformatic analysis to investigate the potential biological roles of GSDME in HNSC. METHODS: The Gene Expression Omnibus (GEO) and Cancer Gnome Atlas (TCGA) databases were used to analyze the expression of GSDME in different cancer types. The correlation between GSDME expression and immune cell infiltration or immune checkpoint genes was examined by Spearman correlation analysis. DNA methylation analysis of the GSDME gene was conducted using the MethSurv database. Kaplan-Meier (K-M) survival curves, diagnostic receiver operating characteristic (ROC) curves, nomogram model, and Cox regression analysis were chosen to evaluate the diagnostic and prognostic predictive value of GSDME. Connectivity Map (Cmap) online platform, Protein Data Bank (PDB) database and Chem3D, AutoDock Tool and PyMol software were used to predict and visualize potential molecular drugs aimed for GSDME. RESULTS: GSDME expression level in HNSC was significantly higher than in the controls (p < 0.001). Differentially expressed genes (DEGs) correlation with GSDME were enriched in the GO pathways, such as protein activation cascade, complement activation and classical pathway (p < 0.05). According to GSEA, GSDME-associated differentially expressed genes were significantly enriched in KRAS signaling pathway and cytokine signaling molecule (p < 0.05). There is a significant relation between GSDME expression and immune cell infiltration in HNSC tissues, as well as immune checkpoint genes expression (p < 0.001). DNA methylation status of cg17790129 CpG islands of GSDME gene is correlated with HNSC prognosis (p < 0.05). Based on Cox regression analysis of HNSC patients, GSDME as a potential risk gene has high correlation with overall survival (OS) and disease specific survival (DSS) (p < 0.05). In a ROC curve analysis, HNSC tissues were differentiated from adjacent peritumoral tissues based on GSDME expression levels (AUC = 0.928). Totally six potential drugs targeted for GSDME were screened and the molecular docking tests between GSDME protein and candidate drugs were conducted. CONCLUSIONS: GSDME is a promising therapeutic target as well as a potential clinical biomarker in HNSC patients.