Microglial
spatial heterogeneity remains a crucial yet poorly studied question in light of potential
cell-directed therapies for Alzheimer`s
disease (AD). Little is known about the dynamics of spatially distinct
microglia states, which are either adjacent or non-associated with the plaque site, and their selective contributions to
neurodegeneration in vivo. So far, research has essentially focused on
pathology-associated
microglia. Here, we combined novel multicolor
fluorescence fate mapping, single-
cell transcriptional analysis, epigenetic profiling, advanced
immunohistochemistry and computational modelling to comprehensively characterize the relation of plaque-associated and non-plaque- associated
microglia during
neurodegeneration. This approach enabled us to identify and characterize non-plaque-associated microglia as a unique and highly dynamic microglial state in a mouse model of AD. Non-plaque-associated microglia modulate
network expansion, quickly adapt to environmental cues and their transition to plaque-associated microglia can be specifically modulated during
disease, contrary to their reputation as a passive bystander subpopulation. This description of the dynamics of spatially segregated microglial states and their distinct molecular features may therefore open promising new avenues for state-specific therapeutic interventions during
neurodegeneration.