Cell-type signatures of Alzheimer's disease shared across population groups

• 成果类型: Article
• 署名作者: Luquez, Tain; Algoo, Jonathan; Chiu, Rebecca; Mares, Jason A.; Yadav, Archana; Lam, Matti; Gaur, Pallavi; Lai, Xiaoying; Lee, Dylan I.; Paryani, Fahad; Batchelor, Rafe; Belli, Irla; Henry, Jordan; Hoter-Ishay, Bat; Mattison, Courteney; Starr, Lindsey; Lama, Tsering; Karaahmet, Berke; Cao, Wenqing; De Jager, Philip L.; Taga, Mariko; Barnes, Lisa L.; Marquez, David X.; Bennett, David A.; Zhang, Ya; Menon, Vilas
• 署名单位: Columbia University; Cornell University; Weill Cornell Medicine; NewYork-Presbyterian Hospital; Rush University; University of Illinois System; University of Illinois Chicago
• 刊物名称: NATURE
• ISSN/ISSBN: 0028-0836; 1476-4687
• DOI: 10.1038/s41586-026-10793-0
• 发表日期: 2026-09-03
• 关键词: cognitive impairment; association; pathology; variants; dementia; insights; Mutation; network

摘要

Genomic studies at single-cell resolution have identified several cell types associated with clinical and pathological traits in Alzheimer's disease1, 2, 3, 4, 5, 6, 7, 8-9, but have not examined associations that are shared across populations. To bridge this gap, here we use single-nucleus RNA sequencing and assay for transposase-accessible chromatin with sequencing to profile cortical and subcortical regions in post-mortem brain-tissue samples from Latin, white (excluding Latin) and African American (excluding Latin) individuals. Using discrete and continuous dissections of molecular programs, we identify cell-type-specific clusters associated with Alzheimer's disease in a region-specific manner across all three population groups, including microglial (GPNMB + and CD74 + subgroups), astrocytic (SERPINH1 +, CD44 + and WIF1 + subgroups) and neuronal (SST + GABAergic and superficial-layer glutamatergic) signatures. We also report continuous gene-expression factors in astrocytes and oligodendrocytes that are not captured by discrete cluster assignments, but which show strong associations with disease phenotypes; these factors are enriched for genes associated with annotated functions such as lipid processing and neurotransmitter reuptake. Finally, we find that molecular programs reveal six distinct subgroups of individuals with cognitive impairment that span all three populations, are not captured by neuropathology, and are instead distinguished by molecular signatures that are not universally present but are nonetheless associated with ante-mortem impairment. Overall, our study identifies key cell types and gene programs implicated in Alzheimer's disease that are shared across population groups, and underscores how representative sampling can capture both shared signatures and disease heterogeneity, thereby enabling better prioritization of key cell types for further investigation.