Abstract
Disulfidptosis is a recently identified type of programmed cell death. It is characterized by aberrant accumulation of intracellular disulfides. The clinical implications of disulfidptosis in clear cell renal cell carcinoma (ccRCC) remain unclear. A series of bioinformatics approaches were employed to analyze ten disulfidptosis-related molecules. Firstly, the expression patterns of the disulfidptosis-related molecules were different between normal and ccRCC tissues. A comprehensive cohort of patients with ccRCC was then assembled from three public databases and subjected to cluster analysis based on disulfidptosis-related molecules. Consensus cluster analysis revealed three distinct disulfidptosis clusters. We then conducted weighted gene co-expression network analysis (WGCNA) to identify highly correlated genes. 267 hub genes were screened out through WGCNA, and three gene clusters were then determined. Finally, we identified 87 genes with prognostic value and then used them to develop a disulfidptosis scoring (DSscore) system, which was proven to independently predict survival in ccRCC. Patients in the high-DSscore group exhibited a significant survival advantage and better immunotherapeutic responses compared with those in the low-DSscore group. However, the patients in the low-DSscore group exhibited a greater degree of chemotherapeutic response. In addition, the expression of disulfidptosis-related molecules was validated by qRT-PCR, and the potential of disulfidptosis-related molecules to indicate distinct cell subtypes were validated by single-cell RNA-sequencing. In conclusion, DSscore is a promising index for predicting the prognosis and efficacy of immunotherapy in patients with ccRCC and may provide a basis for novel strategies for future studies.