Abstract / Summary
Abstract Neuroinflammation is a hallmark of Alzheimer’s disease (AD); however, which neuroinflammatory effectors are consistently altered across independent human cohorts, brain regions, and microglial states—and can be engaged by available compounds—remains incompletely defined. Here, we integrated bulk transcriptomic data from 462 human AD brain samples across four independent cohorts, comprising 61,466 nuclei profiled by single-nucleus RNA-seq, to demonstrate that the NLRP3/ASC inflammasome represents a convergent therapeutic target supported by multiple molecular modalities. Neuroinflammatory pathways were consistently enriched across cohorts. Meta-analysis revealed significant upregulation of inflammasome-related genes, including GSDMD (log₂FC = 0.72; pooled p = 5.4 × 10 -5 ), PYCARD , and CASP1 ; GSDMD expression was highest at Braak stage VI, consistent with progressive engagement of the pyroptotic effector arm in advanced AD. Cell state–resolved single-nucleus analysis further showed that the proportion of inflammasome-high nuclei was higher in disease-associated microglia (DAM) from AD donors, although this difference was not detectable when all microglia were analyzed together. To evaluate the therapeutic relevance of this pathway, we treated 5xFAD mice with dimethyl itaconate (DI), a membrane-permeable inhibitor of NLRP3 signaling. DI treatment was associated with trends toward improved spatial working memory and reduced fibrillar amyloid deposition, and it broadly reversed neuroinflammatory and DAM-associated transcriptional signatures, including those of Pycard and Gsdmd , in bulk brain RNA-seq. Together, these findings support the NLRP3/ASC inflammasome pathway as a convergent neuroinflammatory target in AD and show that DI attenuates the corresponding neuroinflammatory transcriptional program in vivo, providing a translational rationale for immunometabolic therapeutic strategies in AD.