Cellular network perturbations point to a new microglia‐astrocyte community accelerating Alzheimer’s disease progression

Abstract

AbstractBackgroundAlzheimer’s Disease (AD) is a fatal neurodegenerative disease where abnormal accumulation of amyloid‐β and tau aggregates drive neurodegeneration and leads to cognitive impairment, dementia and eventually death. Combining results from recent studies, each finding a novel AD‐associated cell state of different non‐neuronal cell types, suggests a system level change of the cellular environment of the AD brain and its potential role in disease outcome.MethodTo better understand the dynamics of the cellular environment in aging human brains and AD, we applied single‐nucleus RNA‐sequencing to profile 424 aging brains of individuals across different clinicopathological characteristics. We characterized the cellular diversity within the aging human prefrontal cortex, resulting in a detailed atlas of ∼1.7 million cells classified into 95 cellular subsets of different cell types, including rare populations of microglia and astrocytes.ResultsQuantifying variations across individuals, we uncovered multiple cell subsets associated with disease pathologies and cognitive impairment. Of these, we prioritize an uncommon Microglial subset 13 (Mic.13, average 3.2% of microglia) and Astrocyte subset 10 (Ast.10, average 3.8% of astrocytes). Mic.13 expresses AD risk gene TREM2 and microglial markers linked to AD (e.g. APOE, SPP1), and its proportion is strongly associated with AD pathologies, amyloid‐β plaques and neurofibrillary tangles, and cognitive decline. Applying a mediation model, we showed that Mic.13 partially mediates the effect of amyloid‐β on tau (proportion mediated [p.m.] = 18%). While the impact of Mic.13 on cognitive decline is largely mediated by tau (p.m. = 43%), an additional subset highlighted by our analysis Ast.10, associated with both tau and cognitive decline, partially mediates the impact of tau and Mic.13 on cognitive decline (p.m. = 7.2% and p.m. = 12% respectively). Integration of all cellular states revealed an intricate network of multi‐cellular communities, each composed of multiple states of different cell types. Investigating the dynamics of the cellular communities in the aging brain and along disease progression, uncovered unique trajectories spanning from a healthy to diseased brains.ConclusionOur charted atlas of the cellular environment and modelling of the cellular dynamics supports a system level change underlying AD, with an active role for multiple cell types in disease progression, and highlights novel research directions and potential therapeutic targets.