Using Blood-Based Biomarkers to Facilitate the Diagnosis of Alzheimer's Disease: Insights from a Novel Pyroptosis-Associated Molecular Signature Model.

Alzheimer's disease (AD) stands as the primary reason for aging-related dementia and is swiftly becoming recognized as one of the century's most expensive, fatal, and burdensome illnesses, which calls for more convenient diagnostic methods. Increasing studies indicate the involvement of pyroptosis in the advancement of AD. Yet, the precise role of molecules associated with pyroptosis in the diagnosis of Alzheimer's disease remains unexplored. We obtained the intersection of differentially expressed genes in GSE63061 as our training dataset (whole blood samples of AD patients and health controls) from the GEO database and further validated in GSE63060 as a validation set. The immune infiltration of selected pyroptosis-related genes was analyzed along with the ceRNAs co-expression network and transcription factor (TF) network construction using bioinformatics methods. A total of 7 differentiated expressed pyroptosis-related key genes (NLRP6, BAX, CASP4, DPP9, NLRP3, PYCARD, CASP6) were identified and constructed as a diagnostic model for AD. Through ROC analysis, the AUC values in both datasets were greater than 0.7 (AUC = 0.72, 95% confidence interval = 0.65-0.78), indicating that the diagnostic model we constructed has good diagnostic performance in distinguishing AD from the control. Furthermore, CIBERSORT algorithms were used to unravel the immune cell infiltration landscape related to pyroptosis-related key genes in AD and normal samples. Among them, PYCARD and Monocytes have the highest positive correlation (r = 0.54, p < 0.001). PYCARD and T cells CD4 memory-activated have the highest negative correlation (r = - 0.37, p < 0.001). Finally, a competing endogenous RNA (ceRNA) network and PPI network of the obtained pyroptosis-related key genes were subsequently constructed, and KEGG functional pathway analysis revealed the five most significant pathways, including fatty acid metabolism, fatty acid biosynthesis, ECM-receptor interaction, proteoglycans in cancer, and gap junctions. Besides, the upstream TF network of the 7 differentiated expressed pyroptosis-related key genes was constructed. Our team developed a diagnostic model for pyroptosis-related signatures in Alzheimer's disease, investigating ceRNA and TF regulatory networks, which may facilitate the diagnosis of AD with blood-based biomarkers.