Identification of Succinylation-Related Transcriptional Signatures in Alzheimer's Disease: Integrated Bioinformatics Analysis and Clinical Validation.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder involving mitochondrial dysfunction, oxidative stress, proteostasis imbalance, and immune-inflammatory alterations. Protein succinylation is closely linked to mitochondrial metabolic status, but peripheral blood transcriptomic signatures related to succinylation in AD remain insufficiently characterized. This study aimed to identify peripheral blood transcriptional candidate biomarkers associated with a predefined succinylation-related gene set in AD. The GSE63061 dataset was used as the discovery cohort, and GSE63060 served as an external validation cohort. Differential expression analysis, ssGSEA-derived succinylation scoring, and WGCNA were integrated with a predefined succinylation-related gene set. Candidate biomarkers were then refined through four feature selection approaches: LASSO, random forest, SVM-RFE, and Boruta, and their discriminatory performance was evaluated by ROC analysis, nomogram construction, calibration analysis, decision curve analysis, and SHAP analysis. Functional enrichment, GSVA/GSEA, GeneMANIA, immune-cell signature analysis, and single-cell analyses were used to explore biological relevance and cellular distribution. qRT-PCR was performed using peripheral blood-derived PBMC samples for experimental validation. Five genes, NDUFA1, NDUFS5, RPS27A, AIF1, and PRDX1, were identified as candidate biomarkers associated with a succinylation-related transcriptional signature in AD. These genes were mainly involved in mitochondrial complex I activity, oxidative phosphorylation, ribosomal function, protein homeostasis, redox balance, and immune inflammation. Immune-cell signatures and single-cell analyses further suggested that the signature was associated with T cells, B cells, NK cells, and monocytes/macrophages. qRT-PCR confirmed significant downregulation of all five genes in AD PBMC samples (P < 0.05). These findings suggest that peripheral blood transcriptional signatures associated with succinylation-related genes may provide candidate biomarkers and hypothesis-generating clues for AD.