P4‐256: A SINGLE‐CELL‐BASED ATLAS OF HUMAN MICROGLIAL STATES SHOWS DIVERGENT ASSOCIATIONS WITH AGING‐RELATED NEUROPATHOLOGIES AND DEMENTIA

Recently it has been suggested that the potential microglia phenotypic states form a multidimensional space, rather than a linear spectrum, as was historically held. Novel single cell RNA sequencing studies performed on microglia isolated from mouse models of Alzheimer's disease and neurodegeneration identified novel functional states of murine microglia, which did not align with the historically held microglia phenotypes. These studies also pointed out that population level assessment of microglia phenotype is insufficient to resolve the contribution of this cell type to neurodegenerative diseases. To explore human microglia phenotype diversity in an unbiased way, we performed high-throughput single cell RNA sequencing of isolated microglia. We profiled 15,910 CD45+ cells isolated from the cerebral cortices of 7 aged (mean 92.5 years of age) and 8 middle-aged (mean 39.6 years of age) individuals. All aged donors had a non-zero burden of amyloid plaques and neurofibrillary tangles in their brain. We have identified 14 distinct subpopulations of human microglia. When analyzing the interrelatedness of the clusters we found that the largest cluster (cluster number 1 containing 42.2% of the cells) was centrally positioned with all the other clusters “originating” from this central, “homeostatic” cluster. Each microglia subset had a unique set of surface markers and expressed a unique combination of transcription factors. We have confirmed the existence of these subpopulations in situ with immunohistochemistry. The transcriptomic signatures specific to the identified microglia subsets showed divergent association with neurodegenerative and neuroinflammatory diseases, such as multiple sclerosis and Alzheimer's disease and had a unique relationship with the histopathological traits associated with Alzheimer's disease and aging. Marker genes of the disease associated microglia (DAM) phenotype, as well as other microglia signatures recently described in mouse, did not segregate clearly to any one human microglia subpopulation. Microglia in the human brain can acquire a multitude of phenotypic states, which are not well captured in mouse models. These unique human microglia subsets have distinct functional relevance in neurodegenerative diseases and aging. Accordingly our dataset offers novel insights into the involvement of microglia in these processes as well as provides new targets for biomarker and therapeutics development.