OBJECTIVE: Disulfidptosis is a recently discovered mechanism of cell death caused by disulfide stress. Arachidonic acid metabolism (AAM) is one of the metabolic mechanisms of polyunsaturated fatty acids. However, few studies have explored the relationship between disulfidptosis and AAM and how together they affect breast cancer prognosis. The aim of this study was to establish a prognostic model of disulfidptosis and arachidonic acid metabolism in breast cancer and to investigate the potential mechanisms of disulfidptosis and arachidonic acid in breast cancer. METHOD: A comprehensive approach including single-cell sequencing analysis (scRNA-seq), weighted gene coexpression network analysis (WGNA), and transcriptome differential expression analysis was used to investigate this relationship. Cox regression and Lasso regression analysis were used to screen for arachidonic acid metabolism and disulfidptosis-related genes (AAMDRGs) in breast cancer. This study also comprehensively analyzed the risk score of AAMDRGs, as well as clinical features, tumor microenvironment, somatic mutations, immunotherapy, drug sensitivity, and molecular docking data. RESULT: Seven genes associated with disulfidptosis and AAM (ATP5F1B, RPL4, PRDX1, TCP1, CLDN7, GSTK1, PYCARD) were identified in this study. These genes were used to establish a prognostic model related to disulfidptosis and arachidonic acid in breast cancer. The model had excellent performance. Multi-indicator ROC analysis, DAC, and C-index all suggested that the model had high prediction accuracy. This study found that RPL4 is a key candidate gene that can serve as a biomarker for BC, with low expression associated with a poorer prognosis in cancer patients. In addition, this study found that axitinib has a predicted potential pharmacological binding relationship with RPL4 based on molecular docking analysis, which provides a basis for future preclinical and clinical validation studies. CONCLUSION: This machine learning-driven disulfidptosis-related and arachidonic acid metabolism-related model provides clinically actionable prognostic stratification, outperforming conventional gene signatures in precision oncology applications. RPL4 is a potential candidate biomarker associated with disulfidptosis and AAM-related pathways in breast cancer, and its potential as a therapeutic target requires further experimental validation.
Yan et al. (Thu,) studied this question.