Alzheimer’s disease (AD) is a progressive neurodegenerative disorder with no effective treatments available. There is growing evidence that cuproptosis contributes to the pathogenesis of this disease. This study developed a novel molecular clustering based on cuproptosis-related genes and constructed a signature for AD patients.
The differentially expressed cuproptosis-related genes (DECRGs) were identified using the DESeq2 R package. The GSEA, PPI network, GO, KEGG, and correlation analysis were conducted to explore the biological functions of DECRGs. Molecular clusters were performed using unsupervised cluster analysis. Differences in biological processes between clusters were evaluated by GSVA and immune infiltration analysis. The optimal model was constructed by WGCNA and machine learning techniques. Decision curve analysis, calibration curves, receiver operating characteristic (ROC) curves, and two additional datasets were employed to confirm the prediction results. Finally, immunofluorescence (IF) staining in AD mice models was used to verify the expression levels of risk genes.
GSEA and CIBERSORT showed higher levels of resting NK cells, M2 macrophages, naïve CD4+ T cells, neutrophils, monocytes, and plasma cells in AD samples compared to controls. We classified 310 AD patients into two molecular clusters with distinct expression profiles and different immunological characteristics. The C1 subtype showed higher abundance of cuproptosis-related genes, with higher proportions of regulatory T cells, CD8+T cells, and resting dendritic cells. We subsequently constructed a diagnostic model which was confirmed by nomogram, calibration, and decision curve analysis. The values of area under the curves (AUC) were 0.738 and 0.931 for the external datasets, respectively. The expression levels of risk genes were further validated in mouse brain samples.
Our study provided potential targets for AD treatment, developed a promising gene signature, and offered novel insights for exploring the pathogenesis of AD.