Spatial proteomics reveals human microglial states shaped by anatomy and neuropathology

Abstract

Microglia are implicated in aging, neurodegeneration, and Alzheimer's disease (AD). Traditional, low-plex, imaging methods fall short of capturing in situ cellular states and interactions in the human brain. We utilized Multiplexed Ion Beam Imaging (MIBI) and data-driven analysis to spatially map proteomic cellular states and niches in healthy human brain, identifying a spectrum of microglial profiles, called the microglial state continuum (MSC). The MSC ranged from senescent-like to active proteomic states that were skewed across large brain regions and compartmentalized locally according to their immediate microenvironment. While more active microglial states were proximal to amyloid plaques, globally, microglia significantly shifted towards a, presumably, dysfunctional low MSC in the AD hippocampus, as confirmed in an independent cohort (n=26). This provides an in situ single cell framework for mapping human microglial states along a continuous, shifting existence that is differentially enriched between healthy brain regions and disease, reinforcing differential microglial functions overall.

Figure 1.. Multiplexed ion beam imaging (MIBI) maps the macro- and microenvironment of the human brain through spatial proteomics.

(A) The MIBI workflow including human brain anatomical region and field of view (FOV) selection, staining with a 38-marker antibody panel and the computational pipeline spanning single-cell and niche analysis. (B) The human hippocampus imaged with 106 stitched FOVs (700 μm × 700 μm each) showing VGLUT1 (green), MAP2 (red), MCNpase and GFAP (blue), DNA + RNA (yellow, through free Indium 1153+ staining) and CD31 + CD105 (magenta). (C) An enlarged view of FOV1, the CA2 and partial dentate gyrus, shown in (B), highlighting fine cellular features including nuclei, pyramidal neuronal soma and dendrites, synaptic densities, myelin, astrocytes and vasculature. (D) Pixel correlation of pan-brain proteins used to identify multi-cellular niches in the brain, highlighting strong positively and negatively correlated protein programs. (E) Images of proteins depicted in (D) demonstrating pixel signal overlap and exclusion for (i) nuclear proteins, (ii) astrocytes proteins, (iii) vascular proteins, (iv) axonal and dendrite proteins in neurons, and (v) endogenous iron (Fe). (F) Sub-regional organization of the human hippocampus from expert neuropathological annotation (left) and the protein abundance in each sub-region (right) calculated as average pixel value per mm², with the inner cornu ammonis, subiculum and dentate gyrus areas highlighted.

Figure 2.. Microglial segmentation and phenotyping within the context of large anatomical brain regions.

(A) Iba1⁺ (green) microglia depicted in a single FOV from the human hippocampus in a < 1 mm depth of field, with nuclear staining by histone H3 (HH3, blue) and vascular staining by CD31 + CD105 (red). (B) An enlarged view of FOV1 in (A) highlighting a microglia cell (Cells 1) positive for Iba1 (green) and CD45 (red) with its HH3⁺ (blue) nucleus in the plane. (C) Microglia were segmented with eZsegmenter by creating a hybrid mask (green) based on Iba1 (red) and CD45 (blue) expression, normalized for cell size. (D) Inclusion of microglial nuclei (HH3, blue) from the segmentation mask and exclusion of other glial markers like GFAP (red) and S100β (blue). (E) Microglial cellular density in local expert-annotated sub-regions across the hippocampus and cerebellum, depicted as color overlay on dots (microglia) across each stitched image. (F) Quantification of local microglial density in each acquired FOV for each large anatomical brain regions in grey and deep white matter areas, including the hippocampus (HIP), substantia nigra (SN), middle frontal gyrus (MFG), caudate and cerebellum (n = 1 for each brain region). Box and whisker plots represent the minimum, first quantile, median, third quantile and maximum, with individual points representing FOVs. (G) Microglial phenotyping protein expression within segmented cellular masks across five sets of single cells showing differential intracellular localization of each protein. (H) Endogenous elemental iron (Fe) found within and outside of segmented microglia and its colocalization with Ferritin-L. (I) Pearson correlation of microglial phenotyping proteins at a cellular level, with specific positively and negatively correlated protein programs highlighted.

Figure 3.. Classification of microglial cells and cell fragments using multiplexed morphological features.

(A) Multiplexed morphological features extracted with eZsegmenter including metrics of cell size. (B) Aggregated data spread of microglial morphological features across all brain regions, including both grey and deep white matter. P/A ratio: perimeter to area ratio. Each dot represent a FOV taken from one brain region. (C) Single-cell representative images on microglial masks of each morphological metric spanning low, middle and high levels. (D) The correlation of nuclear signal as measured by single-cell HH3 expression and circularity across the entire microglial data set is weak and reveals heterogeneous populations of differential nuclear inclusion and circularity levels. (E) Representative images of differential protein localization within circular cell bodies (CB), larger cytoplasmic cell pieces and small round processes (PR) within the segmented microglial mask (magenta fill and dashed magenta line). (F) Four FlowSOM morpho clusters calculated from the entire microglial data set using select morphological features uncovers nucleate and anucleate cells and cell parts with varying degrees of branching and size. (G) Clustered overlays of the four morpho clusters within a single hippocampal FOV, highlighting an enlarged area (FOV1) and the presence or absence of a nucleus in the plane (dashed white line). (H). Differential protein expression of select microglial phenotyping proteins in each of four morpho clusters. (I) Distribution of the four morpho clusters in each individual FOV acquired for each large anatomical brain region across grey and deep white matter. MFG: middle frontal gyrus.

Figure 4.. High-dimensional single-cell proteomic phenotyping of microglia through trajectory analysis across healthy human brain regions.

(A) Dimensionality reduction of all extracted microglia from all brain regions (from one healthy donor) with UMAP and progression along the microglial cell state continuum (MSC) calculated by SCORPIUS pseudotime using only protein features. (B) Differential relative protein expression for each phenotypic protein along the MSC from low, medium and high MSC. (C) Representative images of single cells and their expression of protein markers from low, medium to high MSC within the segmented mask (green fill, red dashed line). (D) Progression of relative expression of morphological features across the MSC calculated with protein features with representative images of cells. (E) Distribution of microglial cellular density in each large brain region across the MSC from one healthy donor. (F) Total number of statistically significant differences between feature comparisons (including both protein and morphology) between each brain region pairwise comparison in binned low, middle and high MSC. (G) Expression of the five most highly statistically significant differentially expressed features between brain regions along the MSC (top), with their statistical significance represented as their Pairwise Comparison of Trajectories by Binned Permutations (PCTBP) score (significant above a threshold of 1) and area between the curves for each brain region pairwise comparisons (bottom, shown as a single open circle for each pair). (H) Summary of the individual features (protein and morphology) and their total number of significant pairwise comparisons for each brain region pair. (I) Features that are significantly differentially expressed along the MSC in grey (black line) vs. deep white (grey line) matter areas depicted in volcano plot with individual proteins and their relative expression along the MSC, as well as the microglial cellular density of grey and deep white matter along the MSC.

Figure 5.. Microglial states are enriched in local microenvironments as defined by pixel clustering and neighborhood analysis.

(A) Local brain textures identified through pixel clustering of all FOVs from all healthy human brain regions from a healthy donor using pan-brain protein markers, with expression of all panel protein makers for each of twenty pixel clusters. (B) Distribution of pixel clusters for each large brain region, quantified as average frequency per FOV for grey and deep white matter. (C) Spatial compartmentalization of overlaid pixel clustered brain textures in the hippocampal local sub-regions with enlarged FOVs (right): FOV1 depicting grey matter with high synaptic density (pixel clusters 1 and 2), nuclei (pixel cluster 4), pyramidal neurons (pixel cluster 10), dendrites (pixel cluster 20), areas of grey and white matter mixtures (pixel cluster 5); FOV2: astrocyte endfeet alone vasculature (pixel cluster 6), intravascular immune cells (pixel cluster 12); FOV3: white matter axons (pixel cluster 15) and myelin (pixel cluster 9). (D) Brain texture pixel clusters present in FOV1 shown as single layer images with enlarged individual features for clarity. (E) Single microglia depicted spatially within the hippocampus and colored by their MSC value, with expert annotated subregions (dashed white line). (F) Quantification of the proximal niche around each microglia calculated as the pixel cluster frequency within a 20 mm radius from the centroid of each microglial cell, represented in FOV4, and plotted for the entire grey matter hippocampus arranged along the binned MSC bin (180 bins) from low, middle to high MSC. (G) The pixel clusters of the deep white matter of the hippocampus overlay across the acquire FOVs (top) with microglia depicted spatially and colored by their MSC. Pixel cluster frequencies around microglia along the MSC for the deep white matter in the hippocampus (bottom). (H) Correlation (positive and negative) of pixel clusters progression with MSC progression for each healthy brain region in the grey (left) and deep white (right) matter.

Figure 6.. Spatial analysis of microglia in Alzheimer’s disease alters their cellular state according to proximity to amyloid plaques.

(A) Segmentation of amyloid plaques (left) and tau tangles (right) from one human AD hippocampus by eZsegmenter through masking (blue) Pan-Aβ (magenta) and PHF-tau (yellow) expression. Human donor details: 82-year-old male with AD dementia, Braak score of V, MMSE score of 19, APOE ε3/ε3, and a post-mortem interval of 2.95 hours. (B) Phenotyping of plaque types with FlowSOM clustering (left, top) results in four plaque types with varying levels of Pan-Aβ, Aβ₄₂ and Aβ₄₀. Density of each plaque type quantified as number of plaques per mm² in the grey and deep white matter of the human AD hippocampus (left, bottom). Relative levels of each plaque phenotyping protein for each plaque type (right). (C) Spatial enrichment analysis of plaques and tangles proximal to single microglia (left) through masking each feature and quantifying the presence of a plaque or tangle within a specified radius (r = major axis length of each cell + 10 pixel buffer) around each microglial mask. Total number of microglia with a plaque or tangle of each type within their direct vicinity (right). (D) Quantification of differentially expressed proteins in disease associated microglia (DAMs) as defined by their proximity to each plaque type, with representative plaque types 3 and 4 shown with microglia overlaid as dashed magenta lines. (E) Microglial cellular density along the MSC between the healthy (green line) and AD (navy blue line) human hippocampus (top) with the total number of statistically significant and not significant pairwise differences within protein and morphology features. (F) Statistically significant feature differences (only proteins) between the healthy and AD human hippocampus as volcano plot, with their PCTBP scores and area between the curves. (G) All feature comparisons (protein and morphology) shown along the MSC for the healthy and AD human hippocampi. (H) Validation of MIBI findings through an independent cohort of 13 healthy and 13 AD hippocampi through low-plex immunohistochemistry: (top) staining the tissue microarray cores of the CA1 region of the hippocampus with Iba1 (blue) and HLA-DR (brown) as a dual stain and in silico separation of each protein signal; (bottom) segmentation of microglia through masking both Iba1 and HLA-DR signals together in eZsegmenter and quantification of signal intensity at a single-cell level in each FOV acquired in healthy and AD hippocampus groups. A proxy MSC was calculated with Iba1 and HLA-DR expression from extracted cells (bottom, middle) and the average microglial cellular density plotted for the healthy (green line) and AD (navy blue line) hippocampus (with individual donors in dashed lines) and Iba1 (light blue) and HLA-DR (peach) expression along the MSC. Statistical significance (p-value) was calculated along binned MSCs (bottom, right) with PCTBP scores > 1 interpreted as statistically significant.