Microglial activation and tau propagate jointly across Braak stages

Abstract

AbstractBackgroundCompelling experimental evidence suggests that microglial activation is involved in the spread of tau neurofibrillary tangles over the neocortex in Alzheimer’s disease (AD). Here, we tested the hypothesis that the spatial propagation of microglial activation and tau accumulation colocalize in a Braak‐like pattern in the human brain.MethodWe studied 130 subjects (86 cognitively unimpaired and 44 cognitively impairment elderly) with microglia activation [11C]PBR28 PET, Amyloid [18F]NAV4694 PET, cognitive tests, magnetic resonance imaging, and cerebrospinal fluid (CSF) biomarkers. Participants also had baseline and follow‐up tau [18F]MK6240 PET scans. We used linear regressions and partial correlations to estimate the association between topographical distributions of microglial activation and tau accumulation.Result(A)Microglial activation (MA) connectivity analysis between regions comprising Braak histopathological stages (Braak I‐VI) across our population demonstrated that MA regions correlated hierarchically with each other, where a Braak‐like stage region only correlates with the previous and subsequent stage. (B)Tau tangles connectivity analysis between regions that comprise Braak stages (Braak I‐VI) across our population demonstrated that tau regions correlated hierarchically with each other, where a Braak stage only correlates with the previous/subsequent stage (C)The scatterplot shows that the elements of the MA‐MA matrix and tau‐tau matrix were highly positively correlated with each other, reinforcing the notion of a joint topographic propagation of MA and tau tangles through Braak stages. (D) Rates of tau propagation over 1‐year hierarchically correlated in Braak stages. (E)Significant Pearson correlation between longitudinal tau propagation and baseline tau network. (F)Significant Pearson correlation between longitudinal tau propagation and baseline microglia network. (G)Partial residuals plots show the results of regression between network of tau propagation over 1‐year and both baseline topographic circuits, MA‐MA and tau‐tau networks. The model suggests that the pathways of longitudinal tau propagation depended on underlying microglia, rather than tau circuits. (H)The figure shows thatmicroglial activation levels in the transentorhinal cortex predicted longitudinal tau spread over the neocortex.ConclusionOur results support a model in which an interaction between amyloid load and activated microglia set the pace for tau spread across Braak stages, while the simultaneous cerebral presence of these pathologies potentiates the development of dementia.