Oxidative Stress-Related Genes in Uveal Melanoma: the Roles of CALM1 in Modulating Oxidative Stress, Apoptosis, and Prognostic Significance

Yue Wu^*Xiaoyan CaiMenghan HuRunyan CaoYong Wang^*

• First Affiliated Hospital of Anhui Medical University, Hefei, China

Abstract Background: Uveal melanoma (UVM) is a rare yet aggressive form of with a poor prognosis. This study aims to investigate the role of oxidative stress-related genes (OSGs) in UVM, focusing on their involvement in key signaling pathways, immune infiltration and their potential as prognostic biomarkers and therapeutic targets. Method: Differential gene expression analysis was conducted using 175 samples of normal retinal pigmented epithelium-choroid complex samples and 63 samples from UVM. Protein-protein interaction (PPI) networks were constructed to identify hub genes, and machine learning algorithms were utilized to screen for diagnostic genes, employing methods such as least absolute shrinkage and selection operator (LASSO) regression, random forest, support vector machine (SVM), gradient boosting machines (GBM), neural network algorithm (NNET) and eXtreme gradient boosting (XGBoost). A risk signature model was developed using data from the cancer genome atlas (TCGA) cohort and validated using the international cancer genome consortium (ICGC), GSE84976 dataset. Clinical samples were used to validate the diagnostic value. Experimental validation encompassed H2O2-induced oxidative stress assays and CALM1 overexpression analysis in UVM cells to evaluate its protective effects. Results: A total of 2,576 differentially expressed genes (DEGs) were identified, with 185 overlapping OSGs enriched in pathways such as HIF-1, FoxO, PI3K-Akt, and apoptosis pathways. Prognostic hub OSGs, including ACACA, CALM1, and DNM2, were associated with poor survival outcomes in training set and multiple validation data. Revalidation using clinically collected samples confirmed that CALM1 exhibits superior diagnostic value. The risk signature model demonstrated strong predictive accuracy for 5-year overall survival (AUC = 0.844). Immune infiltration analysis revealed an increased CD4+ memory activated T cells and mast resting cells in high-risk groups. Additionally, CALM1 overexpression attenuated H2O2-induced oxidative stress and apoptosis in UVM cells. CALM1 upregulation also mitigated the inhibitory effects of H2O2 on key cellular processes, including proliferation, migration, and invasion. Conclusion: This study underscores the critical role of OSGs in the progression of UVM and their potential as prognostic biomarkers and therapeutic targets. The identified risk signature model and the protective role of CALM1 offer valuable insights for developing targeted therapies and enhancing patient clinical outcomes in UVM.