Disulfidptosis-associated gene signature predicts prognosis and radioresistance in NSCLC.

BACKGROUND: NSCLC has an extremely high mortality rate, and limitations exist in the efficacy of current therapeutic modalities and the choice of treatment strategies. Cells overexpressing cystine/glutamate transporter proteins and experiencing a glucose shortage could lead to NADPH deficiency and excessive disulfide buildup, causing stress responses and cell death, which affects patient's prognosis. METHODS AND MATERIALS: Based on TCGA and GEO datasets, consensus clustering of NSCLC patients was performed to identify disulfidptosis-related genes and construct a prognostic model. The predictive ability of the model was validated in multiple datasets. Survival analysis, genomic mutations, and immune infiltration were used to characterize epigenetic alterations in different risk groups. Multiple radiotherapy-associated NSCLC datasets and radioresistant cell lines were employed to explore the relationship between disulfidptosis-related genes and radiotherapy. RESULTS: We identified nine disulfidptosis-associated prognostic genes (DAPGs) and calculated the disulfidptosis-relevant risk score (DRRS). The risk groups showed a difference in prognosis, genomic mutations, and immune infiltration. CAFs and epithelial-mesenchymal transition significantly enriched in the high DRRS group. The prognostic model also effectively predicted the prognosis of patients receiving radiotherapy, and expression of DAPGs, especially KIF14, is strongly associated with DNA damage and repair, and high expression of KIF14 was observed in radioresistant cells. CONCLUSIONS: Our findings imply that a prognostic model related to disulfidptosis could distinguish the prognostic differences among patients, allowing for more personalized treatment strategies. The mechanisms of KIF14 may provide a basis for combating radioresistance and a better understanding of the interaction between disulfidptosis and radiotherapy.