Analysis of Methylation Levels and Prognostic Correlations in Oral Squamous Cell Carcinoma Based on TCGA/GTEx Data

Hangzhou Normal University Affiliated Hospital (School of Clinical Medicine, School of Stomatology), Hangzhou, China

^* Corresponding author: c18357757158@126.com

Abstract

Objective: To investigate the relationship between DNA methylation and the prognosis of oral squamous cell carcinoma (OSCC), and to introduce normal tissue data from the GTEx database as a control. Methods: DNA methylation data and expression data of OSCC were obtained from the TCGA database, and methylation and expression data of normal tissues were obtained from the GTEx database as controls. Differential methylation site analysis of DNA methylation chips was performed using the CHAMP package in R software. Methylation annotation was conducted using GENCODE 28V probes, and prognostic-related genes were identified through Cox regression analysis. Kaplan-Meier survival analysis, univariate and multivariate analyses, and receiver operating characteristic (ROC) curves were used to evaluate factors affecting prognosis. Results: A total of 38 differentially expressed genes were identified, including 29 upregulated genes (VMP1, SPNS1, EIF4EBP1, NLRC4, KLHL24, HIF1A, TP63, ITGB4, BID, VEGFA, DRAM1, GAA, BNIP3, ITGA3, BIRC5, SPHK1, CXCR4, NRG1, ITGA6, FADD, DDIT3, SERPINA1, IFNG, IL24) and 9 downregulated genes (NRG2, TP53INP2, FOS, HSPB8, PTK6, NRG3, PRKN, CCL2, ULK3). A total of 5213 differentially methylated sites were detected in the transcription start site (TSS) and 14233 in the gene body. After removing duplicate data, 3696 independent sites were obtained. Comparison with normal tissue data from the GTEx database yielded 1522 differentially methylated sites, of which 1521 showed consistent trends in both databases. Univariate Cox regression analysis identified 53 genes with prognostic significance. Multivariate Cox regression analysis based on these 53 genes confirmed that 4 genes (VMP1, HIF1A, VEGFA, IL24) were statistically significant (P<0.05). Conclusions: The methylation levels of VMP1, HIF1A, VEGFA, and IL24 are associated with the prognosis of oral squamous cell carcinoma, and the normal tissue data from the GTEx database provide important control support for the study. Hypermethylation of these genes was associated with downregulation of their expression levels, which may impact key functional pathways such as angiogenesis (VEGFA), hypoxia response (HIF1A), and immune regulation (IL24). These findings provide mechanistic insights into the role of DNA methylation in the development and prognosis of OSCC.