Single-cell technologies have revealed important cell heterogeneity across the human brain. In the context of neurological diseases, the relationship between different cell subpopulations and pathological features can be crucial for their understanding. However, the combination of spatial, single-cell, and high-plex protein information together with analytical tools for the accurate segmentation of ramified cells remains a challenge to address in neuroscience.
Here, we present CODEX-CNS, a modification of CO-Detection by indEXing (CODEX) technology for its use in human brain tissues. This technology consists of multiplexed fluorescent imaging that allows the detection of up to 100 proteins using DNA-barcoded antibodies. In this paper we bring technical advancements to the CODEX protocol and an improved data analysis pipeline for accurate segmentation of complex cell morphologies. As proof-of-principle, we were able to detect the different parenchymal brain cells and their cytoarchitecture, as well as blood-brain barrier and meningeal components with a 32-plex antibody panel. More specifically, we used CODEX-CNS in human brain samples of healthy and Alzheimer’s disease donors to study microglial phenotypes in relationship to their spatial context. Applying our customized cell segmentation algorithm and clustering analysis, we identified diverse microglial subpopulations differentially distributed between brain areas and according to their distance to amyloid-β plaques.
These data provide a new approach for the neuroscience community that allows the characterization of microglial subpopulations at the protein level with both single-cell and spatial resolution.