High-parameter cytometry unmasks microglial cell spatio-temporal response kinetics in severe neuroinflammatory disease

Background Differentiating infiltrating myeloid cells from resident microglia in neuroinflammatory disease is challenging, because bone marrow-derived inflammatory monocytes infiltrating the inflamed brain adopt a 'microglia-like' phenotype. This precludes the accurate identification of either cell type without genetic manipulation, which is important to understand their temporal contribution to disease and inform effective intervention in its pathogenesis. During West Nile virus (WNV) encephalitis, widespread neuronal infection drives substantial CNS infiltration of inflammatory monocytes, causing severe immunopathology and/or death, but the role of microglia in this remains unclear. Methods Using high-parameter cytometry and dimensionality-reduction, we devised a simple, novel gating strategy to identify microglia and infiltrating myeloid cells during WNV-infection. Validating our strategy, we (1) blocked the entry of infiltrating myeloid populations from peripheral blood using monoclonal blocking antibodies, (2) adoptively transferred BM-derived monocytes and tracked their phenotypic changes after infiltration and (3) labelled peripheral leukocytes that infiltrate into the brain with an intravenous dye. We demonstrated that myeloid immigrants populated only the identified macrophage gates, while PLX5622 depletion reduced all 4 subsets defined by the microglial gates. Results Using this gating approach, we identified four consistent microglia subsets in the homeostatic and WNV-infected brain. These were P2RY12(hi) CD86(-), P2RY12(hi) CD86(+) and P2RY12(lo) CD86(-) P2RY12(lo) CD86(+). During infection, 2 further populations were identified as 'inflammatory' and 'microglia-like' macrophages, recruited from the bone marrow. Detailed kinetic analysis showed significant increases in the proportions of both P2RY12(lo) microglia subsets in all anatomical areas, largely at the expense of the P2RY12(hi) CD86(-) subset, with the latter undergoing compensatory proliferation, suggesting replenishment of, and differentiation from this subset in response to infection. Microglia altered their morphology early in infection, with all cells adopting temporal and regional disease-specific phenotypes. Late in disease, microglia produced IL-12, downregulated CX3CR1, F4/80 and TMEM119 and underwent apoptosis. Infiltrating macrophages expressed both TMEM119 and P2RY12 de novo, with the microglia-like subset notably exhibiting the highest proportional myeloid population death. Conclusions Our approach enables detailed kinetic analysis of resident vs infiltrating myeloid cells in a wide range of neuroinflammatory models without non-physiological manipulation. This will more clearly inform potential therapeutic approaches that specifically modulate these cells.