Predicting prognosis of oral squamous cell carcinoma based on endoplasmic reticulum stress-associated gene signatures.

Endoplasmic Reticulum Stress (ERS), as a core mechanism of cellular response to protein homeostasis imbalance, plays a dual regulatory role in tumorigenesis and development. In this study, we aimed to analyze the regulatory network of endoplasmic reticulum stress-associated genes (ERSAGs) in oral squamous cell carcinoma (OSCC). By screening differentially expressed genes through the TCGA database, we explored the potential associations between ERS and OSCC across various aspects. We constructed an OSCC prognostic risk scoring model based on ERSRGs and validated the model's reliability using the GEO dataset as a validation set. In total, 43 differentially expressed ERSAGs were screened as well as 9 prognostic genes. Six genes (KLHL14, SLC25A4, STC2, TRIB3, ALG3, and CCNA2) were screened by Lasso regression to construct a prognostic risk score model. Further analysis suggested that KLHL14 may suppress OSCC progression by modulating tumor-infiltrating immune cells, specifically activated B cells and mast cells. Concurrently, experimental validation demonstrated that overexpression of CCNA2 significantly promotes the proliferation of OSCC. The results indicated that CCNA2 promotes the proliferation of OSCC cells cultured in vitro. This study is the first to construct a prognostic risk model for OSCC based on ERSAGs, which may assist in predicting prognosis of OSCC patients. The identified ERSAGs may contribute to the development of new therapeutic strategies, highlighting the potential clinical application value of ERS-related genes in predicting OSCC prognosis.