Disulfidptosis-related genes signature predicts prognosis and immune microenvironment in colon cancer.

BACKGROUND: Colon cancer (CC), characterized by high incidence and mortality, ranks among the most prevalent digestive malignancies, and reliable molecular tools to predict prognosis and immunotherapy response are needed. Disulfidptosis is a recently identified form of programmed cell death induced by disulfide stress and represents a potential therapeutic target for CC. However, the prognostic and immunological implications of disulfidptosis-related genes (DRGs) in CC remain underexplored. METHODS: RNA sequencing and clinical data from TCGA and GEO databases (GSE39582, GSE17536) were analyzed. A prognostic model including five DRGs (RAB7A, SLC7A11, INF2, FLNA, OXSM) was established using WGCNA, univariate Cox, and LASSO-Cox regression. Patients were stratified into high- and low-risk groups based on risk scores. The model was cross-validated by Kaplan-Meier, time-dependent ROC, nomogra, and multivariable Cox analyses. Immune infiltration (ssGSEA), tumor-mutation burden, miRNA-DRG networks, and drug-sensitivity correlations (CCLE/GDSC/CellMiner) were assessed. Protein expression of RAB7A and OXSM was further examined in tissue microarrays (TMAs) of 97 CCs with matched normal mucosae using immunohistochemistry. RESULTS: A 5-DRGs prognostic model was established which serves as an independent signal of overall survival and immunotherapy efficacy in CC. Low-risk group patients exhibited an immune-favorable microenvironment and were more abundant with activated CD56 bright NK, γδ-T, and Th17 cells, higher tumor mutational burden (TMB), and elevated CTLA-4 expression, which were likely associated with improved response to immune checkpoint inhibitors (ICIs). As a risk factor, RAB7A relating to poor prognosis was identified in our study, and was a potential target for therapeutic agents such as PLX4720 and PD-0325901, while OXSM was the opposite. Therefore, this model serves as a translatable framework for the precision management of CC, enabling rational risk stratification and facilitating personalized treatment decisions. CONCLUSION: The 5-DRGs prognostic model can be used to assess the prognosis of patients with CC, and reflect the characteristics of their immune microenvironment. With RAB7A and OXSM as key determinants, this model provides a clinically applicable framework for risk stratification and personalized treatment guidance.