Identification of mitochondrial oxidative stress-related genes for estimating prognosis, immune infiltration landscape and drug candidates in pancreatic adenocarcinoma.

BACKGROUND: Pancreatic adenocarcinoma (PAAD) represents a highly lethal form of cancer. Mitochondrial stress responses play a critical role in cancer development and therapeutic responses. METHODS: Transcriptomic and clinical survival data for PAAD and normal pancreatic tissues were sourced from TCGA, GEO, and GTEx databases. Random survival forest and Cox regression analyses were used to identify MOS prognostic genes and construct a prognostic risk model. Subsequently, receiver operating characteristic (ROC) curve analysis, calibration curve analysis, and decision curve analysis were performed to evaluate the model’s performance in both the training and validation sets. Finally, enrichment analysis, protein-protein interaction network construction, unsupervised cluster analysis, tumor immune microenvironment assessment, and chemotherapeutic drug sensitivity analyses were conducted. Finally, we also performed RT-qPCR for validation. RESULTS: Five MOS-related genes were identified to construct the risk signature: ABCG1, CENPT, GMNN, NOG, and TOP2A. The predictive performance of the risk signature was validated in both the training set (TCGA) and the test sets (GSE62452 and GSE78229), with areas under the ROC curve for 3-year survival being 0.77, 0.74, and 0.75, respectively. Mechanistically, functional enrichment analysis revealed that pathways associated with cancer cell proliferation and mitochondrial stress responses were upregulated in the high-risk population. Furthermore, specific small molecules may hold therapeutic potential for treating PAAD patients. RT-qPCR results validated the changes in relative RNA expression levels. CONCLUSION: This study demonstrates that mitochondrial oxidative stress is correlated with the prognosis of PAAD and may serve as a novel biomarker and therapeutic target for clinical diagnostics.