Spatial proteomics of Alzheimer's disease-specific human microglial states.

Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Nat Immunol . Author manuscript; available in PMC: 2026 Apr 14. Published in final edited form as: Nat Immunol. 2025 Jul 22;26(8):1397–1410. doi: 10.1038/s41590-025-02203-w Search in PMC Search in PubMed View in NLM Catalog Add to search Spatial Proteomics of Alzheimer’s Disease–Specific Human Microglial States Dunja Mrdjen Dunja Mrdjen 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Dunja Mrdjen 1, # , Bryan J Cannon Bryan J Cannon 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Bryan J Cannon 1, # , Meelad Amouzgar Meelad Amouzgar 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Meelad Amouzgar 1 , YeEun Kim YeEun Kim 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by YeEun Kim 1 , Candace Liu Candace Liu 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Candace Liu 1 , Kausalia Vijayaragavan Kausalia Vijayaragavan 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Kausalia Vijayaragavan 1 , Christine Camacho Christine Camacho 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Christine Camacho 1 , Angie Spence Angie Spence 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Angie Spence 1 , Erin F McCaffrey Erin F McCaffrey 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Erin F McCaffrey 1 , Anusha Bharadwaj Anusha Bharadwaj 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Anusha Bharadwaj 1 , Dmitry Tebaykin Dmitry Tebaykin 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Dmitry Tebaykin 1 , Syed Bukhari Syed Bukhari 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Syed Bukhari 1 , Marc Bosse Marc Bosse 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Marc Bosse 1 , Felix J Hartmann Felix J Hartmann 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA 3 Systems Immunology and Single-Cell Biology, German Cancer Research Center (DKFZ), 69120, Heidelberg, Germany Find articles by Felix J Hartmann 1, 3 , Adam Kagel Adam Kagel 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Adam Kagel 1 , John Paul Oliveria John Paul Oliveria 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by John Paul Oliveria 1 , Koya Yakabi Koya Yakabi 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Koya Yakabi 1 , Geidy E Serrano Geidy E Serrano 4 Banner Sun Health Research Institute, Sun City, 85351, AZ, USA Find articles by Geidy E Serrano 4 , Maria M Corrada Maria M Corrada 2 Department of Neurology, University of California, Irvine, 92697, CA, USA Find articles by Maria M Corrada 2 , Claudia H Kawas Claudia H Kawas 2 Department of Neurology, University of California, Irvine, 92697, CA, USA Find articles by Claudia H Kawas 2 , Robert Tibshirani Robert Tibshirani 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Robert Tibshirani 1 , Thomas G Beach Thomas G Beach 4 Banner Sun Health Research Institute, Sun City, 85351, AZ, USA Find articles by Thomas G Beach 4 , M Ryan Corces M Ryan Corces 5 Gladstone Institute of Neurological Disease, Gladstone Institutes, San Francisco, 94158, CA, USA Find articles by M Ryan Corces 5 , Will Greenleaf Will Greenleaf 6 Department of Genetics, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Will Greenleaf 6 , R Michael Angelo R Michael Angelo 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by R Michael Angelo 1 , Thomas Montine Thomas Montine 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Thomas Montine 1, * , Sean C Bendall Sean C Bendall 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA Find articles by Sean C Bendall 1, * Author information Article notes Copyright and License information 1 Department of Pathology, Stanford University, School of Medicine, Palo Alto 94304, CA, USA 2 Department of Neurology, University of California, Irvine, 92697, CA, USA 3 Systems Immunology and Single-Cell Biology, German Cancer Research Center (DKFZ), 69120, Heidelberg, Germany 4 Banner Sun Health Research Institute, Sun City, 85351, AZ, USA 5 Gladstone Institute of Neurological Disease, Gladstone Institutes, San Francisco, 94158, CA, USA 6 Department of Genetics, Stanford University, School of Medicine, Palo Alto 94304, CA, USA # These authors contributed to this manuscript equally. Authors Contributions Statement: DM, BJC, TM and SCB conceptualized, designed, and executed the project. DM, BJC, CC, AS, KV, MB, JPO, FJH, AB, KY validated antibodies and protocols on IHC and MIBI for human brain tissue. DM, BJC, acquired all MIBI images. RMA, MB, BJC, CL maintained and troubleshot MIBI instrument. AK designed MIBI TIFF processing tools for noise and background removal. MA created the trajectory inference applications for mapping microglial cell states. BJC and MA derived and applied statistical metrics used to identify differences between regions and disease states in consultation with RT. BJC designed the segmentation tools that captured microglia and proteopathy objects, as well as the tools for quantifying local proteopathy-microglial interactions. DT created the scripts for analyzing local microglial proteomic neighborhoods. CL designed and ran pixel clustering on the case study samples. EFM provided additional analysis of microglia pixel cluster neighborhoods. SB, TGB, GES, MMC, CHK, TM provided the human brain tissue for study. TM, SCB provided review and editorial assistance of the manuscript. MRC provided the ATAC-Seq data, while YK, WG, and BJC provided ATAC-Seq analysis. DM, BJC, were the primary writers of the manuscript. KV assisted with figure designs and edits. All authors read and approved the final manuscript. * Correspondence – [email protected] ; [email protected] Issue date 2025 Aug. PMC Copyright notice PMCID: PMC13075585  NIHMSID: NIHMS2145041  PMID: 40696045 The publisher's version of this article is available at Nat Immunol Previous version available: This article is based on a previously available preprint posted on Research Square on June 2, 2023: " Spatial proteomics reveals human microglial states shaped by anatomy and neuropathology ". Abstract Microglia are implicated in aging, neurodegeneration, and Alzheimer’s disease (AD). Low-plex protein imaging fails to capture cellular states and interactions in the human brain, which differs from rodent models. With Multiplexed Ion Beam Imaging (MIBI) we spatially mapped cellular states and niches in cognitively normal (CN) human brains, identifying a spectrum of proteomic microglial profiles. Defined by immune activation states that were skewed across brain regions and compartmentalized according to microenvironments, this spectrum enabled the identification of proteomic trends across the microglia of ten cognitively normal individuals and orthogonally with single nuclei epigenetic analysis (snATAC-seq), revealing associated molecular functions. Notably, AD tissues exhibit significant regulatory shifts in the immunologically active cells at the end of the proteomic spectrum, including enrichment of CD33 and CD44 and decreases in HLA-DR, P2RY12, and ApoE expression. These findings establish an in situ, single-cell spatial proteomic framework for Alzheimer’s disease-specific microglial states. Introduction Microglia are dynamic, tissue-resident immune cells in the central nervous system that perform diverse functions throughout brain development, adulthood, and aging, as well as in neurodegenerative and neuroinflammatory disorders 1 . Early definitions of microglia focused on morphology or expression of broad immune markers (often categorized as “M1” or “M2” states) 2 . Single-cell proteomic 3 – 5 and transcriptomic phenotypes have identified multiple phenotypes in both health and disease 6 – 12 , and in development and aging 13 – 15 . The field faces challenges in unifying these findings, especially for human microglia, where data can be limited to a few brain regions or rely on nuclear rather than cytoplasmic markers. 16 – 23 . Standard clustering algorithms used to categorize microglia can artificially separate states that overlap in vivo 3 , 24 . These limitations are magnified by the fact that most studies use rodent models, whose microglial diversity may not translate fully to humans 25 . Additionally, existing spatial transcriptomics and immunohistochemistry approaches often lack the resolution and multiplexing necessary to capture microglia’s complexity in situ. 26 – 32 . Thus, a high-dimensional, spatially resolved analysis is needed to compare microglial states across human brain regions and disease contexts. In previous work, multiplexed ion beam imaging (MIBI) 33 – 40 has been used to perform high-dimensional, spatial proteomic analyses of post-mortem human brain and other various tissues. Here, we built a 38-plex spatial proteomic analysis with MIBI to capture major brain cell types, AD hallmark proteins and 17 microglial phenotypes, then mapped 93 679 microglia across five regions of a cognitively normal donor. Instead of discrete subsets, cells fell along a continuous “microglial state continuum” (MSC) of immune activation that varied by anatomy and local niche. Applying MSC to nine more healthy brains (paired CA1 and caudate) reproduced a low-to-high activation gradient in 17,455 microglia and, when compared to microglia open chromatin regions read by snATAC-seq, displaying a proteomic signature that fit into an expanded epigenetic axis—from synapse-supporting to immune-effector states. Finally, contrasting 24,266 microglia from matched brain regions of 12 AD cases showed a skewed high-MSC profile characterized by diminished microglial homeostasis, displaying both increased inflammatory signatures as well as broadly dysfunctional levels of HLA-DR and engulfed synaptic protein, especially in hippocampal CA1. Altogether, our study offers a quantitative spatial proteomic framework for understanding how human microglia vary by region and in disease, highlighting potential new avenues for therapeutic intervention. Results Multiplex spatial proteomics can organize human brain niches We developed a 38-plex MIBI panel for human brain ( Supplementary Table 1 – 2 ) and analyzed five brain regions from a >85-year-old male (APOE ε3/ε3, post-mortem interval 3.4 hours) who was cognitively and neurologically normal ( Figure 1A ). Regions included hippocampus (HIP), cerebellum, SN, caudate nucleus, and MFG. Each section was stained for markers of neuronal structures (e.g., MAP2, VGLUT1), astrocytes (GFAP, S100β, CD44), myelin (MCNPase), vasculature (CD31, CD105), metabolic pathways (VDAC1, CPT2), proteolysis (K48-linked polyubiquitin), oxidative stress (8-OH-Guanosine), iron, hallmark AD pathology (amyloid-β [Aβ], tau), and 17 microglial proteins (e.g., Iba1, CD45, Tmem119, P2RY12, TREM2). Figure 1. MIBI maps protein distribution of macro- and micro-environments of human brain. Open in a new tab 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 x 700 μm each) showing VGLUT1 (green), MAP2 (red), MCNpase and GFAP (blue), DNA + RNA (yellow, through free Indium 115 3+ 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 mm 2 , with the inner cornu ammonis, subiculum and dentate gyrus areas highlighted. In the HIP, we tiled ~6.5 mm x 5.5 mm at ~470 nm pixel resolution ( Figure 1B ). Composite overlays showed characteristic laminar patterns ( Figure 1C ). Pixel-level correlation analyses validated expected relationships among markers—for instance, nuclear markers (HH3, Indium115 3+ , 8-OH-Guanosine) 41 correlated with each other ( Figure 1D , 1E.i ), astrocyte markers (GFAP, S100β, CD44) partially overlapped ( Figure 1E.ii ), ApoE was widely dispersed ( Figure 1E.iii ) 44 , and axonal proteins NFH and NFL co-localized in myelinated tracts ( Figure 1E.iv ),. Iron colocalized with Ferritin-L in microglia ( Figure 1E.v ) as expected 42 , 43 . HIP sub-regions were annotated by neuropathology (subiculum, CA1–4, dentate gyrus [DG], alveus, etc.) ( Figure 1F ). Each displayed distinct protein densities: CA2 was enriched for CPT2, the DG had high levels of nuclear markers, and the alveus showed dense myelin (MCNPase). Microglial markers also differed among subregions, reflecting localized functional demands (e.g., P2RY12, Tmem119, TREM2 in subiculum; Iba1, CD68, GPNMB in stratum lacunosum moleculare; CD74 and TSPO in alveus). Similar patterns were observed in other regions ( Figure S1A – D ). Overall, these multiplexed images revealed regional and sub-regional complexity in microglial marker expression, alongside other cell types and structures. Segmenting microglia can quantify proteomes and cell geometry in a spatial context To investigate microglial diversity, we segmented single cells using Iba1 and CD45 ( Figure 2A , B ). Across the five regions, we identified 203,714 putative myeloid cells, which were filtered down to 93,679 bona fide microglia by excluding vessel-associated macrophages and monocytes ( Figure 2C , S2A , B ). Morphological segmentation showed minimal overlap with astrocytes ( Figure 2D ). Microglial density (cells/mm^2) varied substantially across regions, with the HIP containing the highest density ( Figure 2E , F ). Within the HIP, microglia were enriched in CA2 and CA4, while in the cerebellum they were more abundant in white matter ( Figure S2B ). These distributions align with previous histological studies of human microglia density 45 – 49 . Figure 2. MIBI enables microglial segmentation and phenotyping across anatomical regions. Open in a new tab 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 gray and deep white matter areas, including the hippocampus (HIP), substantia nigra (SN), middle frontal gyrus (MFG), caudate and cerebellum. Box-plots summarise microglial densities from n = 420 imaging fields of view (FOVs) collected across 5 brain regions in a single cognitively-normal donor (biological replicate = 1; technical replicate for each FOV = 1). Centre line, median; box, 25th–75th percentiles; whiskers, 1.5× interquartile range; points, individual FOVs.Centre line, median; box limits, 25th and 75th percentiles; whiskers, 1.5× the interquartile range; points beyond whiskers, individual outliers. G. Microglial phenotyping protein expression within segmented cellular masks across five sets of single cells showing differential intracellular localization of each protein. H. Pearson correlation of microglial phenotyping proteins at a cellular level, with specific positively and negatively correlated protein programs highlighted. I. Clustered overlays of the four morphology 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). Examining subcellular distributions, some markers filled the cytoplasm (Iba1, CD45), while others formed punctate clusters (MRP14, GPNMB, CD74, and MerTK – which also stains perivascular macrophages). Tmem119 and P2RY12, often considered “homeostatic” microglia markers in mice 50 , showed distinct expression intensities ( Figure 2G ). Correlation analyses grouped many phagocytic and immune markers together (HLA-DR, CD68, MRP14, GPNMB) while Fe and Ferritin-L showed moderate correlation ( Figure 2H ), consistent with it being a marker of iron storage in microglia 51 . We also extracted morphological features (cell size, elongation, symmetry, etc.) (Extended Table S3). FlowSOM 52 clustering of morphology alone yielded four meta-clusters: (1) small, rounded cells lacking nuclei (likely fragments), (2) branched cells with some concavity, (3) rounded cells with nuclei, and (4) large, branched cells ( Figure 2I , S2D , E ). To focus on intact cells, we excluded the small, nucleate-lacking fragments in cluster 1 from subsequent analyses. Microglia span a proteomic state continuum across regions When we clustered microglia based on protein expression (FlowSOM), the clusters formed a near-continuous gradient rather than discrete populations ( Figure S3A – E ). To better visualize this, we applied SCORPIUS 53 trajectory analysis and aligned cells along a microglial state continuum (MSC) ( Figure 3A ). The MSC reflected the observed cluster gradient, with one extreme having overall low expression of immune proteins and the other high ( Figure 3B , C , S3J ). Markers like HLA-DR, MerTK, CD11c, MRP14, and TREM2 rose steadily along the continuum, while others (e.g., P2RY12) showed a parabolic pattern. Figure 3. Microglial trajectory analysis reveals a single-cell proteomic continuum in CN brains. Open in a new tab A. Dimensionality reduction of all extracted microglia 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. F. 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). We tested whether the continuum was biased by the large number of cells from the HIP. Calculating region-specific spectrums and comparing them to the combined dataset showed robust consistency ( Figure S3F – H ). This indicates that the continuum reflects a general axis of microglial activation across all brain regions. We also attempted a morphology-based continuum (MSC-morph), which did not correlate well with proteomic changes ( Figure 3D , S3I – K ), reinforcing that morphology alone is insufficient to define microglial states. Next, we examined how microglia from each region distributed along the MSC ( Figure 3E ). The MFG and caudate skewed toward the low end, suggesting microglia in these regions typically express fewer of these specific immune markers in this patient. The cerebellum was enriched in the mid-MSC, while the HIP and SN showed higher average MSC values, indicative of more robust immune phenotypes associated with our markers of interest. To quantify protein differences among regions at different continuum intervals (low, mid, high), we developed a binned permutation test (PCTBP) ( Figure S3L – M ). This analysis revealed that VDAC1, ApoE, CD44, Iba1, and P2RY12 contributed most to regional variation ( Figure 3F , S3N ). Microglial proteomic states segregate within local niches We next assessed whether microglia along different continuum values localize to specific micron-scale niches. Given the difficulty in segmenting non-microglial cell shapes (e.g. astrocytic projections, neuronal axons and dendrites) we instead performed pixel-level clustering across all images, identifying 20 unique pixel clusters that captured distinct tissue textures, including synaptic fields (VGLUT1+CD56+), white matter (MCNPase+), mitochondria/iron-dense areas, neuronal soma, astrocyte bodies, and vasculature ( Figure 4A , Supplementary Table 2 ). The composition of these clusters varied by region ( Figure 4B ). Figure 4. Local pixel-defined niches associate with specific microglial states. Open in a new tab A. Local brain textures identified through pixel clustering of all FOVs from all cognitively normal human brain regions 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 gray 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 gray 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 gray 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 um radius from the centroid of each microglial cell, represented in FOV4, and plotted for the entire gray matter hippocampus arranged along the binned MSC bin (180 bins) from low, middle to high MSC. In the HIP, pixel clusters delineated layers of dense synapses, dendrites, and myelin-rich alveus ( Figure 4C , D ). We assigned each microglial cell a local environment by tallying the nearest pixel clusters within a 20 μm radius ( Figure 4F , S4A – C ). Low continuum microglia were enriched in synapse-dense regions, while higher MSC cells were closer to myelin tracts or vasculature ( Figure 4E , F ). This suggests that microglial states may be partly driven by local signals in distinct structural or functional compartments. Microglial proteomic states segregate across cognitively normal human brains To test whether the MSC generalizes to broader populations, we imaged paired hippocampus CA1 and caudate from nine cognitively normal (CN) older adults ( Supplementary Table 4 – 6 ). Using a 40-plex MIBI panel, we segmented 17,455 microglia ( Figure 5A ). We recalculated an MSC using overlapping microglial markers (Iba1, CD45, P2RY12, Tmem119, HLA-DR, TREM2, ApoE, CD68, C1q, CD44, K48-linked polyubiquitin), obtaining similar low-to-high immune gradients ( Figure S5A – C ). Figure 5. Proteomic immune signature in microglia connects to broader epigenetic spectrum. Open in a new tab A. Summary of total cognitively normal brain tissue imaged by MIBI. Microglia were segmented from images using the same CD45+, Iba1+ strategy as before. CA1 region represents the Cornu Ammonis Region 1 of the hippocampus. Representative tiled images of a patient’s CA1 and Caudate microglia captured by MIBI. Composite images represent the merged CD45 and Iba1 signal, while inset images show individual CD45 and Iba1 signal with microglia segmentation masks in white. B. Volcano plot showing statistically significant feature differences between the cognitively normal CA1 and cognitively normal caudate MSC trajectories, with each features’ pairwise comparison of trajectories (PCT) significance score (y axis) and area between curves, here termed as effect estimate score (x axis). C. Representative images of High MSC microglia with differential phenotypes with respect to CA1 vs Caudate brain region. CA1 microglia show higher HLA-DR, P2RY12, and CD44, while Caudate microglia show higher ApoE, TREM2, and CD68. Patient label in brackets. D. Top: UMAP of original single-cell ATAC sequencing cell types from Corces et.al 2020. Represents 70,631 individual cells in total across cortical, striatum, hippocampus, and substantia nigra brain areas across 8 different patients. Microglia used for further analysis are highlighted in black circle. Bottom: UMAP of microglia alone, colored by epi-MSC cluster: low, medium, high. E. Heatmap of hierarchy clustered genes between Low, Med(ium), and Hi(gh) epi-MSC clusters. MSC-associated genes are those representing the combined proteins used to construct the MSC trajectories from both the original tissue source and the larger cohort. Genes are z-score normalized. F. Heatmap of hierarchically clustered microglia-associated gene ontology (GO) gene sets derived from Thrupp et.al 2020. between the three epi-MSC clusters. GO sets have been grouped into 10 metaclusters and named according to similar individual GO sets. A full list of GO sets can be found in Figure S6E . Despite this shared continuum, inter-individual variation was high ( Figure S5D ). However, region-specific differences in expression persisted when comparing CA1 and caudate across low, mid, and high continuum bins. For example, CA1 microglia showed significantly higher P2RY12, HLA-DR, and CD44 across all continuum bins, while caudate microglia consistently expressed more ApoE ( Figure 5B , 5C , Figure S5C ). In the high continuum bin, TREM2 and CD68 were also elevated in the caudate. Functional markers like Cleaved Caspase 3 were more associated with CA1 microglia, while K63-linked polyubiquitin was more associated with caudate microglia. These patterns suggest that anatomical differences in microglial immune activation exist even among cognitively normal older individuals. We compared our proteomic data to publicly available single-cell ATAC-seq of cognitively normal human microglia ( Figure 5D , top). Clustering these cells (epi-MSC) again revealed three states aligning with low, mid, and high immune activation ( Figure 5D , bottom). Genes in the high cluster were linked to proteins involved in antigen presentation and immune regulation ( Figure 5E , F , S5E ). Conversely, the low cluster showed open chromatin regions associated with genes encoding proteins associated with synaptic regulation (e.g., GPR37, FGF10, KCNN2, MEOX2, CLEC14A, NKX2) 56 – 69 , hinting that these microglia might be specialized for synapse maintenance or remodeling rather than classical immune surveillance. AD hippocampal microglia profiles shift away from antigen specific, homeostatic signature Given the organization seen in cognitively normal tissue, we then applied our MSC approach to individuals with Alzheimer’s disease. First, we analyzed hippocampal subregions (CA1, CA4, DG) from an 82-year-old man with AD (APOE ε3/ε3, post-mortem interval 2.9 hours). We segmented plaques (Aβ) and tangles (PHF-tau) ( Figure S6A – C ). Although microglia around plaques or tangles expressed elevated HLA-DR, ApoE, CD68, and CD74 ( Figure S6D - G )—consistent with “disease-associated microglia” (DAM) 7 —these cells were relatively rare (<1% of the total microglial population). Most microglia in the AD hippocampus were not in immediate proximity to hallmark pathologies. To examine broader AD effects, we imaged CA1 and caudate from 12 individuals with AD dementia ( Figure 6 A , B , Figure 7A ). After segmentation (24,266 microglia), we mapped each cell onto the cognitively normal MSC. While DAMs were again visually identified ( Figure 6 A , B ), especially at the extreme end of the continuum, the many microglia did not directly interact with classic proteopathies. Collectively, the proteomic spectrum of AD microglia displayed a distinctive shift at the high-activation end ( Figure S7A ). Compared to cognitively normal brain aging, AD high MSC microglia showed significantly higher CD44 and CD33 and significantly lower immune engulfment and presentations proteins HLA-DR, P2RY12, ApoE, and engulfed post-synaptic protein PSD95 ( Figure 7B , C , S7A ). This effect was restricted to CA1 relative to Caudate tissue except for ApoE ( Figure S7B ). Like the heterogeneity seen between cognitively normal patients, AD microglia density distributions varied among patients and brain region ( Figure S7C ). Overall this suggests an immune proteomic profile that lacks key antigen presentation or homeostatic functions, possibly indicative of an “exhausted” or dysfunctional phenotype. Figure 6. Microglial protein continuum in AD hippocampus extends past plaque/tangle niches. Open in a new tab Representative tiled images of a (A) cognitively normal (CN) and (B) an Alzheimer’s Disease (AD) patient’s CA1 (top) and Caudate (bottom) microglia captured by MIBI. Composite images combine segmented microglia overlaid on either amyloid beta 40 (Aβ40) or PHF-Tau, while inset images show individual microglia either interacting or independent of individual plaques via Aβ40 staining (insets, i – CA1, iii – Caudate) or tangles via PHF-Tau staining (insets ii – CA1, iv – Caudate). Microglial are colored based on their mapped value to the microglial state continuum built in Figure 5 . Figure 7. AD skews intrinsic and extrinsic proteomic profiles of immune-activated microglia. Open in a new tab A. Top: Summary of total cognitively normal brain tissue imaged by MIBI. Microglia were segmented from images using the same CD45+, Iba1+ strategy as before. CA1 region represents the Cornu Ammonis Region 1 of the hippocampus. Bottom: Quantification of local microglial density in each acquired FOV for each patient’ CA1 and Caudate tissue areas, comparing cognitively normal and AD microglia density. Centre line, median; box limits, 25th and 75th percentiles; whiskers, 1.5× the interquartile range; points beyond whiskers, individual outliers. B. Volcano plot showing statistically significant feature differences between the cognitively normal CA1 and AD CA1 MSC trajectories, with each features’ PCT significance score (y axis) and area between curves, here termed as effect estimate score (x axis). C. Representative images of high MSC microglia with differential immune, intrinsic phenotypes with respect to cognitively normal CA1 vs AD CA1 brain region. Cognitively normal CA1 microglia show higher HLA-DR, P2RY12, and ApoE, while AD CA1 microglia show higher CD33, and CD44. Patient label in brackets. D. Representative images of high MSC microglia with differential proteopathy related, intrinsic phenotypes with respect to cognitively normal CA1 vs AD CA1 brain region. High AD CA1 microglia show higher enrichment of proteopathic amyloid beta 40 and PHF-tau. Patient label in brackets. E. Significantly different proximal mean protein abundance between the cognitively normal CA1 and AD CA1 MSC trajectories, with each features’ binned pseudotime trajectory difference significance score (y axis) and area between curves, here termed as effect estimate score (x axis). F. Representative images of proximal mean protein abundance in high MSC microglia with differential extrinsic phenotypes, with respect to cognitively normal CA1 vs AD CA1 brain region. Cognitively normal CA1 microglia show enrichment in areas of excitatory-related synaptic protein PSD95 and VGlut1, as well as astrocytic protective protein Glutamine Synthetase (GlutSyn), while AD CA1 microglia show higher enrichment in areas of inhibitory-related protein Parvalbumin, as well as gliosis-related astrocytic protein GFAP. Patient label in brackets. Subsequent immunohistochemical staining for Iba1 and HLA-DR in the SN, MFG, HIP, and caudate of the same cohort confirmed that the CA1 region alone showed a significant decrease in HLA-DR–expressing microglia in AD relative to cognitively normal aging ( Figure S7D ), consistent with its known susceptibility in AD. Finally, we examined disease-associated proteins in microglia at different MSC states. High continuum, CA1 AD microglia had significantly more Aβ40 and PHF-tau ( Figure 7E , S7A ). Spatially, these cells were also significantly enriched near parvalbumin (inhibitory synapses) rather than PSD95 (excitatory synapses) ( Figure 7D , S7E , F ), suggesting a possible link to hippocampal hyperexcitability in AD 70 , 71 due to a potential microglial-mediated loss of inhibitory synaptic connections. Neighborhood analyses further showed glutamine synthetase (a sign of healthier astrocytic metabolism) 72 was significantly higher near cognitively normal high continuum microglia, whereas GFAP (reactive astrogliosis) 73 was significantly higher near AD high continuum microglia ( Figure 7E , F , S7F , G ). These data support the notion that AD microglia adopt a broader, regionally pervasive phenotype beyond immediate plaque or tangle niches, including aberrant immune profiles and AD-related hyperexcitability and excitotoxcitity 70 , 74 , 75 . Discussion By combining multiplexed ion beam imaging, single-cell segmentation, and trajectory modeling, we created a spatial proteomic framework to examine microglia in both cognitively normal and diseased human brain regions ( Figures 1 – 7 , Supplementary Table 1 – 2 , 4 – 6 ). Our data reveal that microglia proteins primarily occupy a continuous spectrum of immune activation—the microglial state continuum (MSC)—rather than forming discrete, mutually exclusive subtypes. This continuum is evident across multiple individuals and brain regions, as shown in Figures 3 – 5 , and is shaped by local tissue environments ( Figures 4 ) and sub-regional niches. Single-cell transcriptomic and proteomic studies have defined various microglial “subtypes”, but these states often overlap. Our results ( Figures 2 – 3 ) support the idea that microglia dynamically transition between low and high immune marker expression, with intermediate states in between. The parabolic expression of P2RY12 ( Figures 3B , S3J ) underscores that certain markers do not simply rise or fall monotonically but may peak at intermediate activation levels. This fluidity aligns with microglia’s capacity to rapidly adapt to changes in the CNS environment. Brain regions significantly influence the distribution of microglia along the MSC in a single individual ( Figures 3 – 4 ). The hippocampus and SN harbor higher MSC cells, while MFG and caudate lean toward lower ends of the spectrum. Pixel clustering and neighborhood analyses ( Figure 4 , Supplementary Table 2 ) reveal that microglia in myelin-rich or astrocyte-dense areas adopt higher immune-related protein phenotypes, whereas those in synapse-dense niches show lower expression of activation markers. These patterns may reflect localized cues such as neurotransmitters, trophic factors, or damaged myelin, driving region-specific microglial states. In a larger cohort of cognitively normal older adults ( Figure 5 , Supplementary Table 4 – 6 ), we again observed the MSC, but with considerable inter-individual variability. Notably, CA1 vs. caudate differences in microglial proteins persisted across low, mid, and high continuum bins ( Figures 5B – C , S5C ). Furthermore, integrating single-cell ATAC-seq data ( Figure 5D – F , S5E ) demonstrated that a wide spectrum (epi-MSC) of microglial signatures from synaptic maintenance to antigen presentation and immune regulation exists in cognitively normal human brains. Our immune–heavy spatial proteomics MSC likely captures more nuanced functional aspects in the immune active portion of the epi-MSC. These findings reinforce that microglia shift along a broad molecular axis in cognitively normal aging, modulated by both individual-level and regional factors, and that multi-modality study of these cells is required to fully understand their functional roles. We imaged a large collection of AD CA1 and Caudate brain area, identifying some microglia associated with typical proteopathy (i.e., DAMs), and many that were not ( Figures 6 , S6 ). In AD, microglia protein expression exhibits a skewed high continuum phenotype characterized by elevated CD44 and CD33 and reduced HLA-DR, P2RY12, and ApoE, especially in CA1 ( Figures 7B – C ). Conventional, two-color immunohistochemistry across multiple regions confirms a selective decrease in HLA-DR–expressing microglia in AD hippocampus ( Figure S7D ). While a small fraction of MSC very high microglia near plaques or tangles (i.e., DAMs) do upregulate antigen presentation and broad reactivity ( Figure S6 ) the majority of microglia at the higher end of our continuum ( Figure 7D , E ), display a more diffuse, possibly “exhausted” or “dysfunctional” immune proteomic phenotype, with significantly lower immune regulatory proteomic features 76 – 87 . These changes extend beyond immediate plaque or tangle boundaries, implicating widespread alterations in the AD hippocampus, compared to cognitively normal individuals. Our neighborhood analyses ( Figure 7E - G , S7F ) reveal that AD high continuum microglia are associated with parvalbumin-rich inhibitory interneurons and reactive astrocytes (GFAP-high, glutamine synthetase–low). Immune aberrant microglial microenvironments, via inhibitory impairment, could contribute to hippocampal hyperexcitability and synaptic dysfunction commonly observed in AD. Future work is needed to determine whether shifting microglia away from this dysregulated high MSC phenotype towards the antigen presentation heavy phenotype seen in cognitively normal patients would improve AD conditions. Additional shifts up the continuum towards a DAM phenotype or away from it entirely to non-immunologically active profiles at the lower end of the spectrum may also restore more balanced immune and synaptic support. The MSC framework highlights how microglia in humans are not simply “resting” or “activated” but span a continuous range, influenced by anatomical location and pathological context. Therapeutic strategies might aim to push microglia toward more beneficial phenotypes—e.g., robust phagocytosis without excessive inflammation—or restore antigen presentation and surveillance capacities lost in AD ( Figures 5 – 7 ). Additionally, the significant inter-individual heterogeneity underscores that patient-specific factors likely modulate microglial phenotypes 88 , 89 , suggesting the need for personalized diagnoses and interventions. While this study focuses on older adults (both cognitively normal and AD), future analyses could extend to younger individuals and other neurodegenerative or inflammatory conditions. Moreover, combining high-dimensional proteomics with transcriptomics (including single-cell RNA-seq) and epigenetics (single-cell ATAC-seq) may reveal deeper regulatory mechanisms guiding microglial transitions in AD that protein-based approaches alone struggle to capture. Our approach also emphasizes the importance of retaining spatial context, as local tissue architecture ( Figures 4 , S4 ) and disease hallmarks ( Figures 6 – 7 ) shape microglial states in ways that dissociated single-cell data alone cannot fully capture 90 . In conclusion, we present a multiplexed, quantitative spatial proteomic map of human microglia across multiple brain regions, showing that these cells occupy a continuum of immune activation states. This targeted protein-based immune spectrum state likely reflects one portion of the wide diversity of microglial roles as shown through open chromatin accessibility mapping. This proteomic continuum is heterogenous at both the regional and individual levels, influenced by local microenvironments, and altered in AD toward a potentially dysfunctional phenotype. By integrating single-cell imaging, molecular trajectory modeling, and epigenetic data, we provide a framework for future research and therapeutic targeting of microglia in aging and neurodegenerative diseases. Methods Human brain tissue Human FFPE brain samples that were imaged by MIBI were acquired from the Arizona Study of Aging and Neurodegenerative Disorders and Brain and Body Donation Program at Banner Sun Health Research Institute ( brainandbodydonationprogram.org ) 91 . Brain regions imaged included HIP, cerebellum, SN, caudate nucleus and MFG were from a cognitively normal donor who was a >85-year-old male with no dementia, an MMSE score of 27, Braak stage I, APOE genotype ε3/ε3, and a post-mortem interval of 3.4 hours [ Figures 1 – 4 ]. The AD dementia case imaged by MIBI was HIP from an 82-year-old male with MMSE score of 19, Braak stage V, APOE genotype ε3/ε3, and a post-mortem interval of 2.9 hours [ Figure S6 ]. The tissue microarray of cognitively normal and AD samples used to expand MIBI findings comprised 21 hippocampal human FFPE brain cores selected from patients of The 90+ Study cohort ( https://doi.org/10.1002/alz.12981 ) whose pathologic evaluation was performed at Stanford Pathology Department [ Figures 5 – 7 , S5 , S7 ]. Donors were age-matched and selected based on their clinical diagnosis and neuropathological scores evaluated by NIA-AA guidelines 91 . The 90+ Study at University of California at Irvine; Institutional Review Boards of both institutions approved this research. The donor details are listed in Supplementary Table 4 Sample-size rationale and blinding No statistical methods were used to pre-determine sample sizes; the cohort of 23 post-mortem brains and the multiple regions profiled per donor were dictated by tissue availability but are comparable to — or larger than — sample sets in recent high-plex or single-cell studies of human tissue 5 , 8 , 29 , 39 . On this basis we expect statistical power similar to these published benchmarks while offering broader neuro-anatomical coverage. Investigators were blinded to the individual sample clinical labels during data generation and low level processing. Antibody preparation Antibodies were first screened by IHC on FFPE brain sections to select for the best clones. These were then conjugated to isotopic metal reporters as described previously 36 , 40 . Following conjugation antibodies were diluted in Candor PBS Antibody Stabilization solution (Candor Bioscience). Antibodies were either stored at 4°C or lyophilized in 100 mM D-(+)-Trehalose dehydrate (Sigma-Aldrich) with ultrapure distilled H 2 O for storage at −20°C. Before staining, lyophilized antibodies were reconstituted in a buffer of Tris (Thermo Fisher Scientific), sodium azide (Sigma-Aldrich), ultrapure water (Thermo Fisher Scientific) and antibody stabilizer (Candor Bioscience) to a concentration of 0.05 mg ml −1 . The antibodies, metal reporters and staining concentrations are listed in Supplementary Table 1 for Figures 1 – 4 , S6 , and Supplementary Table 5 for Figures 5 – 7 , S7 . For detailed metal-antibody protocol MIBItag see dx.doi.org/10.17504/protocols.io.bhyej7te . Tissue staining FFPE brain tissues were sectioned (5μm section thickness) from tissue blocks on gold and tantalum-sputtered microscope slides. Slides were baked at 70°C overnight, followed by deparaffinization and rehydration with washes in xylene (3×), 100% ethanol (2×), 95% ethanol (2×), 80% ethanol (1×), 70% ethanol (1×) and ddH2O with a Leica ST4020 Linear Stainer (Leica Biosystems). Slides next underwent antigen retrieval by submerging sides in 3-in-1 Target Retrieval Solution (pH 9, DAKO Agilent) and incubating at 97°C for 40 min in a Lab Vision PT Module (Thermo Fisher Scientific). After cooling to room temperature for 1 h, slides were washed in wash buffer (1× PBS IHC Washer Buffer with Tween 20 (Cell Marque) with 0.1% (w/v) bovine serum albumin (Thermo Fisher)). Next, all slides underwent two rounds of blocking, the first to block endogenous biotin and avidin with an Avidin/Biotin Blocking Kit (BioLegend). Slides were then washed with wash buffer and blocked for 1 h at room temperature with 1× TBS IHC Wash Buffer with Tween 20 with 3% (v/v) normal donkey serum (Sigma-Aldrich), 0.1% (v/v) cold fish skin gelatin (Sigma-Aldrich), 0.1% (v/v) Triton X-100, and 0.05% (v/v) Sodium Azide. The first round of staining was done with free indium 115 3+ (8 mM diluted in PBS, Fluidigm) in staining buffer (1x TBS IHC Wash Buffer with Tween 20 with 3% (v/v) normal donkey serum) and incubated overnight at 4°C in a humidity chamber. The following day, slides were washed twice for 5 min on a shaker in wash buffer. The second round of staining was done using the cocktail of metal conjugated antibodies prepared in staining buffer at their respective concentrations and filtered through a 0.1 μm centrifugal filter (Millipore) prior to incubation with tissue overnight at 4°C in a humidity chamber. Following the overnight incubation with the antibody cocktail, slides were washed twice for 5 min in wash buffer. On the third day, anti-biotin 152 Eu was prepared as described and incubated with the tissues for 1 h at 4°C in a humidity chamber. Following staining, slides were washed twice for 5 min in wash buffer and fixed in a solution of 2% glutaraldehyde (Electron Microscopy Sciences) solution in low-barium PBS for 5 min. Slides were then washed in PBS (1×), 0.1 M Tris at pH 8.5 (3×) and ddH2O (2×) and then dehydrated by washing in 70% ethanol (1×), 80% ethanol (1×), 95% ethanol (2×) and 100% ethanol (2×). Slides were dried under vacuum overnight prior to imaging. For the detailed staining protocol see https://www.protocols.io/view/mibi-staining-dm6gprk2dvzp/v5 Immunohistochemistry For IHC screening of panel antibodies, FFPE human hippocampal tissue was sectioned onto standard glass slide at 5 μm thickness. Slides containing tissue were baked at 70 °C overnight. Tissue sections were then processed and stained using the sequenza method with single primary antibody. The IHC protocol mirrors the MIBI protocol, with the addition of blocking endogenous peroxidase activity with 3% (v/v) H 2 O 2 (Sigma-Aldrich) in ddH2O after epitope retrieval. On the second day of staining, instead of proceeding with the MIBI protocol, tissues were washed twice for 5 min in wash buffer and stained using ImmPRESS universal (Anti-Mouse/Anti-Rabbit horse radish peroxidase) kit (Vector Laboratories). For detailed IHC staining techniques see dx.doi.org/10.17504/protocols.io.bf6ajrae and dx.doi.org/10.17504/protocols.io.bmc6k2ze . For the double antibody staining of the larger IHC cohort, sections were baked at 70°C overnight followed by deparaffination and antigen retrieval (Dako; S2367). Endogenous horseradish peroxidase (HRP) and alkaline phosphatase (AP) were blocked by applying BLOXALL (Vector Laboratories; SP-6000–100) for 30 minutes followed by blocking buffer solution as for sequenza for 1 h at room temperature. Antibodies Iba1 (clone: EPR16588 ; cam; ab220815) and HLA-DR (clone: CD3/43; Abcam; ab7856), both diluted to 1 μg/mL in 3% (v/v) normal horse serum, were added together to each section and incubated at 4°C overnight. After washing three times in wash buffer (95% 1xTBS IHC Wash Buffer with Tween 20), secondary antibodies (ImmPRESS Duet Reagent: HRP Anti-Mouse IgG and AP Anti-Rabbit IgG; Vector Laboratories) were added for 10 minutes at room temperature followed by three washes with wash buffer. Antibodies were revealed with Vector Blue AP substrate (Vector Laboratories; SK-5300) for 5 minutes followed by DAB HRP substrate (Vector Laboratories; SK-4105) for 40 seconds. For detailed protocols see dx.doi.org/10.17504/protocols.io.81wgbyoryvpk/v1 . Image acquisition on MIBI Prior to imaging, slides were sputter coated with 10 nm of gold over the entire stained tissue section on each slide in order to ensure no charging effects of the tissue which impact on field uniformity during imaging. Imaging was performed using a MIBI instrument with a Hyperion ion source by sequential rastering: pre-rastering at half the ion dose to remove the gold coating on FOVs of interest, followed by final rastering and image collection at full ion dose. Xe + primary ions were used to sequentially sputter pixels for a given FOV. The following imaging parameters were used: Aperture setting, 300 μm; acquisition setting of 100 kHz; FOV size, 700 μm x 700 μm at 1,024 ×1,024 pixels per FOV; sample bias, 20 V (pre-raster, ions not funneled into the TOF chamber to preserve the detector) and 50 V (final raster, ions funneled into the TOF chamber); dwell time, 0.5 ms (pre-raster) and 1 ms (final raster for image acquisition); median gun current on tissue, 13.6 nA; an ion dose of ~15 nAmp.ms.nm −2 (pre-raster) and ~30 nAmp.ms.nm −2 (final raster). Acquisition time was 8.7 minutes per FOV (pre-raster) and 17.33 minutes per FOV (final raster). FOVs were tiled across each brain region with an overlap of 50 μm in the x and 20 μm in the y direction. A total of 420 FOVs were acquired across all brain regions, including gray and deep white matter [ Figures 1 – 4 , S6 ]. Using the preset Ionpath MIBI imaging settings: Pre-raster settings: Modified Coarse acquisition settings (40 nA, FOV size = 400 μm x 400 μm at 500 × 500 pixels per FOV), Image Acquisition Coarse acquisition setting; FOV size = 400 μm x 400 μm at 1,024 ×1,024 pixels per FOV; dwell time of 0.5 ms [ Figures 5 – 7 , S5 , S7 ]. Low-level image processing Multiplexed image sets were extracted, slide background-subtracted, denoised and aggregate filtered as previously described 36 , 40 , using Matlab scripts and in-house GUIs for MIBI image processing. Additionally, non-specific binding to charged neurons (due to over fixation of tissue in formalin prior to embedding in paraffin blocks) was subtracted by masking neurons using the indium 115 3+ signal and removing signal from channels that contained non-specific binding. For visualization purposes, individual FOVs were stitched together to reconstruct large areas of each brain region using the Fiji/ImageJ image processing environment and existing plugins [ Figures 1 – 4 , S6 ]. The larger cohort of cognitively normal and AD human brain MIBI images were extracted, denoised, and normalized using the MIBI-processing toffy toolkit https://github.com/angelolab/toffy [ Figures 5 – 7 , S5 , S7 ]. Microglial and pathology segmentation To segment microglia, amyloid plaques and PHF tau tangles we used ezSegmenter, a MATLAB regionprops thresholding-based segmentation GUI developed in-house and available as a part of the MIBI image processing toolkit here: https://github.com/angelolab/MAUI , as previously described 40 . Briefly, multiplexed TIF images from multiple FOVs are loaded into the GUI. Microglial masks were created using the combined Iba1 and CD45 protein expression, amyloid plaque masks were created using pan-Aβ protein expression and PHF tau tangle masks were created using PHF tau protein expression. Parameters for masking are adjusted for each mask separately (e.g. Gaussian Blur, minimum and maximum object pixel size) and fixed across all FOVs. Masks are then used to extract pixel-level signal intensities across each channel and then cell or object size normalized before import into an output cell table csv file. Single-cell or object data was then imported into R for arcsin h transformation and normalization. We also removed sparse monocytes found in blood vessels that would otherwise contaminate our analysis. We gated out intravascular monocytes (i.e., MRP14high, round, nucleate cells found close to CD31+ & CD105+ blood vessels) ( Figure S3B ). For our larger cohort, where MRP14 was not included, we again filtered out small processes and cells that were close to CD31+ & CD105+ blood vessels [ Figures 1 – 4 , S6 ]. An updated version of ezSegmenter, written in Python and available at https://github.com/angelolab/ark-analysis was used to segment microglia for downstream single-cell analysis using R or Python scripting. Similar parameters to the original segmentation and filtering of intravascular monocytes were used [ Figures 5 – 7 , S5 , S7 ]. Microglial Clustering Protein and morphology features of segmented microglial cells were mean-centered and scaled, and unsupervised clustering was performed on protein-only features using FlowSOM 52 v2.2.0 in R version 4.1.3. Computation of the microglial state continuum The microglial state continuum (MSC) was generated using the SCORPIUS 53 algorithm with a lowess smoother and 1000 iterations. Trajectory inference was performed using either protein-only, or morphology-only features. Microglia from the HIP, Caudate, Cerebellum, SN, and MFG brain regions were included (gray and white matter), and features were mean-centered and scaled prior to trajectory inference. Trajectory inference was also performed on microglia from individual brain regions and a spearman rank correlation was used to compare the pseudotime estimate of each individual region, to the all-region MSC, which had strong agreement with each other [ Figures 1 – 4 ]. The MSC on the larger cohort of human brain samples was constructed similarly to above, using all 9 cognitively normal patient CA1 and Caudate samples to construct the trajectory [ Figures 5 – 7 , S5 , S7 ]. Pairwise Comparison of Trajectories by Binned Permutations (PCTBP) To compare protein and morphology expression between microglia from different brain regions along the MSC in a pairwise manner, we developed a non-parametric statistical method using binned permutations along the trajectory with 3 major steps. Shown in Figure S3L : Calculate the optimal bin width using Freedman-Diaconis rule 93 . Perform permutation analysis: Randomly scramble, without replacement, the pair of labels to compare. Within each bin, calculate the mean difference between the scrambled labels. Compare the absolute permutation mean difference from the scrambled labels to the absolute true difference (test statistic) between the true labels in the same bin. Repeat this step 1000x times. The permutation score is the proportion of samples that have a test statistic at least 1.5x greater than the permutation mean difference. A minimum 1.5x difference is required to improve robustness of the method by incorporating a fold change threshold. The permutation score of each bin along the trajectory is corrected for multiple comparisons and smoothed using a locally estimated scatterplot smoothing and the final PCTBP score is the average smooth permutation score along the trajectory. The bin-based permutation approach to generating the PCTBP score allows for flexible analysis that can capture both local and global differences along the trajectory. The PCTBP score is generated using permutation scores along the entire binned continuum, unless otherwise noted. When comparing low, middle, and high parts of the continuum, smoothing is performed prior to splitting the low, middle, and high PCTBP scores [ Figures 3 , 4 ]. For the larger cohort, a simpler pairwise comparison of trajectories (PCT) was calculated without permutation and using pseudobulks to infer differences at the population level instead of the single-cell level [ Figures 5 – 7 , S5 , S7 ]. Data normality assumptions All trajectory and enrichment comparisons were performed with non-parametric methods (Spearman correlations, permutation-based PCT/PCTBP scores or Wilcoxon tests), which do not assume normality or equal variances. For the few parametric summaries shown (e.g., Pearson correlation in Supplementary ), data distribution was assumed to be normal but this was not formally tested. No additional variance-equality tests were applied. Multiple comparisons were controlled with Benjamini–Hochberg false-discovery-rate adjustment where indicated, and individual data points or density contours are presented throughout to visualise underlying distributions. Statistics and reproducibility Representative multiplexed ion-beam images and their quantitative overlays originate from one cognitively normal donor that was imaged in five brain regions. Each section was acquired once; no additional biological replicates were collected for these specific illustrative panels [ Fig. 1e , Fig. 2a and Supplementary Fig. 4a (single-brain “case-study” images)]. These panels incorporate data from n = 21 independent human brains (9 cognitively normal and 12 Alzheimer’s disease). Every donor contributed microglia populating the low , medium and high bins of the protein-defined microglial state continuum (MSC), confirming that the full spectrum is reproducibly present across individuals and diagnoses [ Fig. 5a , c , 6a , b and Fig. 7c , d , f (multi-donor cohort images)]. Each tissue section was imaged once; no technical replicates were generated. Estimating microglia state in the AD hippocampus using machine learning To estimate the MSC of microglia in the AD hippocampus, we used kernel support vector machine (KSVM) with epsilon regression and a radial basis function Gaussian kernel. KSVM was done using the ksvm() function in the R package kernlab 94 v0.9.30 with parameters type = ‘eps-svr’ and kernel = ‘rbfdot’. To validate predictive accuracy of the model, we used 5-fold cross-validation. Microglia data from all regions except AD hippocampus was randomly split into 5 folds and SCORPIUS was computed on the training set for each fold. The KSVM model was trained with protein features and pseudotime as the response variable, and then prediction was performed on the hold-out set. We compared the predicted pseudotime to the pseudotime computed separately on the hold-out set. Model performance was assessed using correlations and residuals. Predicted and true pseudotime estimates were significantly correlated with values between 0.99 and 1.0, residuals were randomly distributed along the fitted line, and absolute mean and standard deviation of residuals across all folds was less than 1.8% and 0.02%, respectively [ Figure S6 ]. To estimate the MSC of microglia in the AD hippocampus of the larger cohort of patients, we used elastic principal graphs 87 to establish landmarks to fit the AD patient microglia trajectories to [ Figure 5 – 7 , S5 , S7 ]. Pixel clustering and spatial enrichment analysis Pixel clusters were identified using the Pixie pipeline 95 . Briefly, single-pixel expression profiles were extracted from pre-processed MIBI images from all brain regions combined. A Gaussian blur was applied using a standard deviation of 2 for the Gaussian kernel. Pixels were normalized by their total expression, such that the total expression of each pixel was equal to 1. A 99.9% normalization was applied for each marker. Pixels were clustered into 100 clusters using a self-organizing map (SOM) based on the expression of 17 markers: HH3, CD56, Indium, APOE, CD45, CD105, NEFL, NEFH, VDAC1, MCNPase, VGLUT1, Iba1, MAP2, S100β, GFAP, pan-Aβ and PHFtau. The average expression of each of the 100 SOM clusters was found and the z-score for each marker across the 100 SOM clusters was computed. All z-scores were capped at 3, such that the maximum z-score was 3. Using these z-scored expression values, the 100 SOM clusters were hierarchically clustered using Euclidean distance into metaclusters. These metaclusters were manually adjusted and mapped back to the original images such that each pixel was assigned to one pixel cluster. To characterize the microenvironment in direct vicinity of each microglial cell, we defined a radius of 40 pixels (20 μm) from the centroid of each microglial mask and calculated the proportion of pixel clusters within. Microglia were binned into 180 bins along the MSC for the hippocampus, and 80 bins for the other brain regions. Average pixel cluster proportions were calculated for the cells within each bin and plotted in ascending order along the MSC [ Figure 4 , S4 ]. Close neighborhood pixel expression was derived using expanded microglial segmentation masks by a set number of pixels (pixel_expansion = 20, ~8 um), equally and in all directions, followed by the removal of the original microglial segmentation masks. This process created donut-shaped masks of microglia-adjacent areas. In the cases of masks being too close to the 1024×1024 image border, we stopped expanding the neighborhood area at the image border. Then, we calculated mean marker expression values for each microglia neighborhood area by applying the donut-shaped masks to the data TIFs. We kept track of microglia cell IDs and linked the neighborhood area expression values to their microglial cells, which enabled us to proceed with data analysis of this dataset [ Figures 7 , S7 ]. Obtaining microglia single-cell ATAC-seq data from publicly available dataset Microglia cells were parsed out from the brain single-cell ATAC-seq data from a published paper 55 . We downloaded raw and meta data that is available online ( GSE147672 ) and also contacted the authors for the fragment files and the ArchRProject file 96 , which the authors generously shared with us, to facilitate consistent analysis. We generated a new ArchRProject from the fragment files and selected 4,564 cells that were labeled as microglia from the original authors’ annotation. Dimensional reduction and batch removal processes were done with addIterativeLSI() and addHarmony() functions from ArchR package ( www.ArchRProject.com . v1.0.2). Batch-removed reduced dimensions were used for all downstream analyses. scATAC-seq microglia clustering Unsupervised clustering of microglia was performed via addClusters() function from ArchR package. Default parameters were used. Among four clusters identified, an outlier cluster that was separated in reduced dimensions had no specific marker genes identified. Thus, this cluster was considered promiscuous cells from the original dataset and was disregarded. A finalized 4,511 cells were clustered into 3 clusters. Calculation of epi-MSC and annotation of clusters To calculate the corresponding metric of MSC in chromatin accessibility data, we generated epi-MSC using the addModuleScore() function in ArchR package. First, gene activity scores were calculated from the chromatin accessibility data by using addGeneScoreMatrix(). Then the gene activity scores of the protein markers used in generating MSC (CD14, TSPO, P2RY12a, MERTK, GPNMB, FTL, TMEM119, CD44, CD68, CD74, APOE, VDAC1, FCGR3A, FCGR3B, PTPRC, AIF1, ITGAX, HLA-DRA, HLA-DRB5, HLA-DRB1, S100A9) were given as features for addModuleScore(). As three clusters exhibit apparent epi-MSC module score distributions, they were named Low, Medium, High epi-MSC for the rest of the analyses. Marker genes identification from gene activity score Marker genes for each cluster were identified using getMarkerFeatures() function from ArchR package in R, which use pairwise wilcoxon test to identify features that are definitional of each cluster. Cutoff of FDR <= 0.05 and Log2FC >= 1 were used. Total 392 genes were called as marker genes (epi-MSC Low: 295 genes, Med: 10 genes, Hi: 87 genes). The mean of gene activity scores of marker genes and MSC-associated genes per each epi-MSC cluster are plotted as heatmap by plotMarkerHeatmap() function in ArchR. Transcription factors (TFs) list was obtained from public database for human TFs. ( https://humantfs.ccbr.utoronto.ca/download.php ). Cell surface protein list was obtained using biomaRt package (v2.46.3) in R, using ensembl human dataset and filtering for “gene_biotype” = “protein_coding” and “name_1006” = “external side of plasma membrane” or “cell surface”. Any TFs or cell surface proteins that appear in marker genes were labeled in Figure 6G . Calculation of Gene Ontology (GO) module scores and heatmap generation To calculate GO module scores, GO terms and associated genes were obtained from Thrupp et.al 2020. Each gene set per GO term was given as a feature list for addModuleScore() function in ArchR package to add GO module score for each microglia. Among 198 GO modules, 194 modules that showed statistically differential distribution between epi-MSC Low and HIgh were chosen (one way ANOVA, p <0.01). The mean value of GO module scores for each epi-MSC cluster were plotted as a heatmap using ComplexHeatmap (v2.6.2) package in R. The rows (GO modules) were k-means clustered (k=10) for 100 times and a final consensus k-means clustering was used. The number of meta-clusters (k) was chosen manually to best represent the GO module score patterns. The number of MSC-associated genes per each GO divided by the total number of MSC-associated genes is annotated as epi-MSC overlap. Pathology and DAM analysis Plaques were clustered using FlowSOM based on pan-Aβ, Aβ 1–40 , Aβ 1–42 and PHF-tau expression within them. To identify microglia that localized directly around plaques and tangles (i.e. DAMs), we bucketed microglia containing masked plaques or tangles within a set radius from each cell centroid. The radius was determined by taking the major axis length of each cell and adding a 10-pixel buffer to ensure only those cells in relatively close proximity to pathology would be considered DAMs [ Figure S6 ]. Supplementary Material Supplemental Figures NIHMS2145041-supplement-Supplemental_Figures.pdf (6.1MB, pdf) Supplementary Tables NIHMS2145041-supplement-Supplementary_Tables.xlsx (73.4KB, xlsx) Acknowledgements: This work was supported by NIH grants R01AG056287, R01AG057915, R01AG068279, U19 AG065156, U24CA224309, P30AG066515, U54HL165445, R01AG078702, R01AG088656, R01NS121404, P01AG036695, R01CA251858, R01CA240638, R01AG078702, R01AG088656, R01NS121404, P01AG036695, R01CA251858, R01CA240638. A gift from Christy and Bill Neidig. Canadian Institute of Health Research Postdoctoral Fellowship to J.P.O. T32 AI007290 to B.J.C and E.F.M. D.M. was supported by Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung (P2ZHP3_181563), Novartis Foundation for medical-biological Research (#18B067), and the Glenn Foundation for Medical Research Postdoctoral Fellowship in Aging Research. The following grants supported the 90+ Study tissue resource -R01AG021055, the UCI ADRC-P30AG066519, and NIA P30 AG19610. Figure 1 cartoons made using BioRender.com Footnotes Competing Interests Statement: SCB and RMA are consultants for and shareholders in Ionpath Inc. that commercializes MIBI technology. SCB and RMA are inventors on and receive royalties for patents relating to MIBI technology licensed to Ionpath Inc by Stanford. All other authors declare no competing interests. Data Availability: All imaging data and cell tables are made available in a public repository here https://doi.org/10.25740/sq409qv3664 97 . Single-nuclear ATAC data previously processed can be found here in GSE147672 Code Availability: Software for running the MIBI equipment was developed by SAI (MiniSIMS 2 Data Systems). The code for the analysis can be downloaded at https://github.com/BendallLab/publications/2024-Cannon_Mrdjen_etal_MSC . R version used: 4.3.3. All the information required for cell and object segmentation are available at Ark-Analysis https://github.com/angelolab/ark-analysis . References: 1. Prinz M, Jung S, and Priller J. (2019). Microglia Biology: One Century of Evolving Concepts. Cell179, 292–311. 10.1016/j.cell.2019.08.053. [ DOI ] [ PubMed ] [ Google Scholar ] 2. Ransohoff RM (2016). A polarizing question: do M1 and M2 microglia exist? Nat. Neurosci 19, 987–991. 10.1038/nn.4338. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Mrdjen D, Pavlovic A, Hartmann FJ, Schreiner B, Utz SG, Leung BP, Lelios I, Heppner FL, Kipnis J, Merkler D, et al. (2018). High-Dimensional Single-Cell Mapping of Central Nervous System Immune Cells Reveals Distinct Myeloid Subsets in Health, Aging, and Disease. Immunity 0, 380–395.e6. 10.1016/j.immuni.2018.01.011. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Ajami B, Samusik N, Wieghofer P, Ho PP, Crotti A, Bjornson Z, Prinz M, Fantl WJ, Nolan GP, and Steinman L. (2018). Single-cell mass cytometry reveals distinct populations of brain myeloid cells in mouse neuroinflammation and neurodegeneration models. Nat. Neurosci 21, 541–551. 10.1038/s41593-018-0100-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Böttcher C, Schlickeiser S, Sneeboer MAM, Kunkel D, Knop A, Paza E, Fidzinski P, Kraus L, Snijders GJL, Kahn RS, et al. (2019). Human microglia regional heterogeneity and phenotypes determined by multiplexed single-cell mass cytometry. Nat. Neurosci 22, 78–90. 10.1038/s41593-018-0290-2. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Keren-Shaul H, Spinrad A, Weiner A, Matcovitch-Natan O, Dvir-Szternfeld R, Ulland TK, David E, Baruch K, Lara-Astaiso D, Toth B, et al. (2017). A Unique Microglia Type Associated with Restricting Development of Alzheimer’s Disease. Cell 169, 1276–1290.e17. 10.1016/j.cell.2017.05.018. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Krasemann S, Madore C, Cialic R, Baufeld C, Calcagno N, El Fatimy R, Beckers L, O’Loughlin E, Xu Y, Fanek Z, et al. (2017). The TREM2-APOE Pathway Drives the Transcriptional Phenotype of Dysfunctional Microglia in Neurodegenerative Diseases. Immunity 47, 566–581.e9. 10.1016/j.immuni.2017.08.008. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Sankowski R, Böttcher C, Masuda T, Geirsdottir L, Sagar, Sindram E, Seredenina T, Muhs A, Scheiwe C, Shah MJ, et al. (2019). Mapping microglia states in the human brain through the integration of high-dimensional techniques. Nat. Neurosci 10.1038/s41593-019-0532-y. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Frigerio CS, Wolfs L, Fattorelli N, Thrupp N, Voytyuk I, Schmidt I, Mancuso R, Chen W-T, Woodbury ME, Srivastava G, et al. (2019). The Major Risk Factors for Alzheimer’s Disease: Age, Sex, and Genes Modulate the Microglia Response to Aβ Plaques. Cell Rep 27, 1293–1306.e6. 10.1016/j.celrep.2019.03.099. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Srinivasan K, Friedman BA, Etxeberria A, Huntley MA, van der Brug MP, Foreman O, Paw JS, Modrusan Z, Beach TG, Serrano GE, et al. (2020). Alzheimer’s Patient Microglia Exhibit Enhanced Aging and Unique Transcriptional Activation. Cell Rep. 31, 107843. 10.1016/j.celrep.2020.107843. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Absinta M, Maric D, Gharagozloo M, Garton T, Smith MD, Jin J, Fitzgerald KC, Song A, Liu P, Lin J-P, et al. (2021). A lymphocyte–microglia–astrocyte axis in chronic active multiple sclerosis. Nature 597, 709–714. 10.1038/s41586-021-03892-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Marschallinger J, Iram T, Zardeneta M, Lee SE, Lehallier B, Haney MS, Pluvinage JV, Mathur V, Hahn O, Morgens DW, et al. (2019). Lipid droplet accumulating microglia represent a dysfunctional and pro-inflammatory state in the aging brain. Nat Neurosci. 2020 Feb;23(2):194–208. doi: 10.1038/s41593-019-0566-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Safaiyan S, Besson-Girard S, Kaya T, Cantuti-Castelvetri L, Liu L, Ji H, Schifferer M, Gouna G, Usifo F, Kannaiyan N, et al. (2021). White matter aging drives microglial diversity. Neuron 109, 1100–1117.e10. 10.1016/j.neuron.2021.01.027. [ DOI ] [ PubMed ] [ Google Scholar ] 14. Hammond TR, Dufort C, Dissing-Olesen L, Giera S, Young A, Wysoker A, Walker AJ, Gergits F, Segel M, Nemesh J, et al. (2019). Single-Cell RNA Sequencing of Microglia throughout the Mouse Lifespan and in the Injured Brain Reveals Complex Cell-State Changes. Immunity 50, 253–271.e6. 10.1016/j.immuni.2018.11.004. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Li Q, Cheng Z, Zhou L, Darmanis S, Neff NF, Okamoto J, Gulati G, Bennett ML, Sun LO, Clarke LE, et al. (2019). Developmental Heterogeneity of Microglia and Brain Myeloid Cells Revealed by Deep Single-Cell RNA Sequencing. Neuron 101, 207–223.e10. 10.1016/j.neuron.2018.12.006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. de Haas AH, Boddeke HWGM, and Biber K. (2008). Region-specific expression of immunoregulatory proteins on microglia in the healthy CNS. Glia 56, 888–894. 10.1002/glia.20663. [ DOI ] [ PubMed ] [ Google Scholar ] 17. De Biase LM, Schuebel KE, Fusfeld ZH, Jair K, Hawes IA, Cimbro R, Zhang H-Y, Liu Q-R, Shen H, Xi Z-X, et al. (2017). Local Cues Establish and Maintain Region-Specific Phenotypes of Basal Ganglia Microglia. Neuron 95, 341–356.e6. 10.1016/j.neuron.2017.06.020. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Grabert K, Michoel T, Karavolos MH, Clohisey S, Baillie JK, Stevens MP, Freeman TC, Summers KM, and McColl BW (2016). Microglial brain region−dependent diversity and selective regional sensitivities to aging. Nat. Neurosci 19, 504–516. 10.1038/nn.4222. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Hart AD, Wyttenbach A, Hugh Perry V, and Teeling JL (2012). Age related changes in microglial phenotype vary between CNS regions: gray versus white matter differences. Brain. Behav. Immun 26, 754–765. 10.1016/j.bbi.2011.11.006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. O’Koren EG, Yu C, Klingeborn M, Wong AYW, Prigge CL, Mathew R, Kalnitsky J, Msallam RA, Silvin A, Kay JN, et al. (2019). Microglial Function Is Distinct in Different Anatomical Locations during Retinal Homeostasis and Degeneration. Immunity 50, 723–737.e7. 10.1016/j.immuni.2019.02.007. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Tan Y-L, Yuan Y, and Tian L. (2020). Microglial regional heterogeneity and its role in the brain. Mol. Psychiatry 25, 351–367. 10.1038/s41380-019-0609-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Stogsdill JA, Kim K, Binan L, Farhi SL, Levin JZ, and Arlotta P. (2022). Pyramidal neuron subtype diversity governs microglia states in the neocortex. Nature 608, 750–756. 10.1038/s41586-022-05056-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Paolicelli RC, Sierra A, Stevens B, Tremblay M-E, Aguzzi A, Ajami B, Amit I, Audinat E, Bechmann I, Bennett M, et al. (2022). Microglia states and nomenclature: A field at its crossroads. Neuron 110, 3458–3483. 10.1016/j.neuron.2022.10.020. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Mair F, Hartmann FJ, Mrdjen D, Tosevski V, Krieg C, and Becher B. (2016). The end of gating? An introduction to automated analysis of high dimensional cytometry data. Eur. J. Immunol 46, 34–43. 10.1002/eji.201545774. [ DOI ] [ PubMed ] [ Google Scholar ] 25. Geirsdottir L, David E, Keren-Shaul H, Weiner A, Bohlen SC, Neuber J, Balic A, Giladi A, Sheban F, Dutertre CA, et al. (2019). Cross-Species Single-Cell Analysis Reveals Divergence of the Primate Microglia Program. Cell 179, 1609–1622.e16. 10.1016/j.cell.2019.11.010. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Mathys H, Davila-Velderrain J, Peng Z, Gao F, Mohammadi S, Young JZ, Menon M, He L, Abdurrob F, Jiang X, et al. (2019). Single-cell transcriptomic analysis of Alzheimer’s disease. Nature 570, 332–337. 10.1038/s41586-019-1195-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Olah M, Menon V, Habib N, Taga MF, Ma Y, Yung CJ, Cimpean M, Khairallah A, Coronas-Samano G, Sankowski R, et al. (2020). Single cell RNA sequencing of human microglia uncovers a subset associated with Alzheimer’s disease. Nat. Commun 11, 6129. 10.1038/s41467-020-19737-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Olah M, Patrick E, Villani AC, Xu J, White CC, Ryan KJ, Piehowski P, Kapasi A, Nejad P, Cimpean M, et al. (2018). A transcriptomic atlas of aged human microglia. Nat. Commun 9, 1–8. 10.1038/s41467-018-02926-5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Chen W-T, Lu A, Craessaerts K, Pavie B, Frigerio CS, Corthout N, Qian X, Laláková J, Kühnemund M, Voytyuk I, et al. (2020). Spatial Transcriptomics and In Situ Sequencing to Study Alzheimer’s Disease. Cell 182, 976–991.e19. 10.1016/j.cell.2020.06.038. [ DOI ] [ PubMed ] [ Google Scholar ] 30. Swanson MEV, Murray HC, Ryan B, Faull RLM, Dragunow M, and Curtis MA (2020). Quantitative immunohistochemical analysis of myeloid cell marker expression in human cortex captures microglia heterogeneity with anatomical context. Sci. Rep 10, 11693. 10.1038/s41598-020-68086-z. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Swanson MEV, Scotter EL, Smyth LCD, Murray HC, Ryan B, Turner C, Faull RLM, Dragunow M, and Curtis MA (2020). Identification of a dysfunctional microglial population in human Alzheimer’s disease cortex using novel single-cell histology image analysis. Acta Neuropathol. Commun 8, 1–16. 10.1186/s40478-020-01047-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Masuda T, Sankowski R, Staszewski O, Böttcher C, Amann L, Scheiwe C, Nessler S, Kunz P, van Loo G, Coenen VA, et al. (2019). Spatial and temporal heterogeneity of mouse and human microglia at single-cell resolution. Nature 566, 388–392. 10.1038/s41586-019-0924-x. [ DOI ] [ PubMed ] [ Google Scholar ] 33. Keren L, Bosse M, Thompson S, Risom T, Vijayaragavan K, McCaffrey E, Marquez D, Angoshtari R, Greenwald NF, Fienberg H, et al. (2019). MIBI-TOF: A multiplexed imaging platform relates cellular phenotypes and tissue structure. Sci. Adv 5, eaax5851. 10.1126/sciadv.aax5851. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Liu CC, McCaffrey EF, Greenwald NF, Soon E, Risom T, Vijayaragavan K, Oliveria J-P, Mrdjen D, Bosse M, Tebaykin D, et al. (2022). Multiplexed Ion Beam Imaging: Insights into Pathobiology. Annu. Rev. Pathol 17, 403–423. 10.1146/annurev-pathmechdis-030321-091459. [ DOI ] [ PubMed ] [ Google Scholar ] 35. Risom T, Glass DR, Averbukh I, Liu CC, Baranski A, Kagel A, McCaffrey EF, Greenwald NF, Rivero-Gutiérrez B, Strand SH, et al. (2022). Transition to invasive breast cancer is associated with progressive changes in the structure and composition of tumor stroma. Cell 185, 299–310.e18. 10.1016/j.cell.2021.12.023. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Keren L, Bosse M, Marquez D, Angoshtari R, Jain S, Varma S, Yang S, Kurian A, Van Valen D, West R, et al. (2018). A Structured Tumor-Immune Microenvironment in Triple Negative Breast Cancer Revealed by Multiplexed Ion Beam Imaging. Cell 174, 1373–1387.e19. 10.1016/j.cell.2018.08.039. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Angelo M, Bendall SC, Finck R, Hale MB, Hitzman C, Borowsky AD, Levenson RM, Lowe JB, Liu SD, Zhao S, et al. (2014). Multiplexed ion beam imaging of human breast tumors. Nat. Med 20, 436–442. 10.1038/nm.3488. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Greenbaum S, Averbukh I, Soon E, Rizzuto G, Baranski A, Greenwald N, Bosse M, Jaswa EG, Khair Z, Kwok S, et al. (2021). Spatio-temporal coordination at the maternal-fetal interface promotes trophoblast invasion and vascular remodeling in the first half of human pregnancy. 2021.09.08.459490. 10.1101/2021.09.08.459490. [ DOI ] [ Google Scholar ] 39. McCaffrey EF, Donato M, Keren L, Chen Z, Delmastro A, Fitzpatrick MB, Gupta S, Greenwald NF, Baranski A, Graf W, et al. (2022). The immunoregulatory landscape of human tuberculosis granulomas. Nat. Immunol 23, 318–329. 10.1038/s41590-021-01121-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Vijayaragavan K. & Cannon BJ, Tebaykin D, Bossé M, Baranski A, Oliveria JP, Bukhari SA, Mrdjen D, Corces MR, McCaffrey EF, et al. (2022). Single-cell spatial proteomic imaging for human neuropathology. Acta Neuropathologica Commun 10, 158. 10.1186/s40478-022-01465-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Watson ML, and Aldrige WG (1961). Methods for the use of indium as an electron stain for nucleic acids. J. Biophys. Biochem. Cytol 11, 257–272. 10.1083/jcb.11.2.257. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Rouault TA (2013). Iron metabolism in the CNS: implications for neurodegenerative diseases. Nat. Rev. Neurosci 14, 551–564. 10.1038/nrn3453. [ DOI ] [ PubMed ] [ Google Scholar ] 43. Ryan SK, Zelic M, Han Y, Teeple E, Chen L, Sadeghi M, Shankara S, Guo L, Li C, Pontarelli F, et al. (2023). Microglia ferroptosis is regulated by SEC24B and contributes to neurodegeneration. Nat. Neurosci 26, 12–26. 10.1038/s41593-022-01221-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Huang Y, and Mahley RW (2014). Apolipoprotein E: Structure and function in lipid metabolism, neurobiology, and Alzheimer’s diseases. Neurobiol. Dis 72, 3–12. 10.1016/j.nbd.2014.08.025. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Wu Y, Dissing-Olesen L, MacVicar B. a., and Stevens B. (2015). Microglia: Dynamic Mediators of Synapse Development and Plasticity. Trends Immunol. 36, 605–613. 10.1016/j.it.2015.08.008. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Hughes AN (2021). Glial Cells Promote Myelin Formation and Elimination. Front. Cell Dev. Biol 9, 661486. 10.3389/fcell.2021.661486 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. McNamara NB, Munro DAD, Bestard-Cuche N, Uyeda A, Bogie JFJ, Hoffmann A, Holloway RK, Molina-Gonzalez I, Askew KE, Mitchell S, et al. (2023). Microglia regulate central nervous system myelin growth and integrity. Nature 613, 120–129. 10.1038/s41586-022-05534-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Mittelbronn M, Dietz K, Schluesener HJ, and Meyermann R. (2001). Local distribution of microglia in the normal adult human central nervous system differs by up to one order of magnitude. Acta Neuropathol. 101, 249–255. 10.1007/s004010000284. [ DOI ] [ PubMed ] [ Google Scholar ] 49. Menassa DA, Muntslag TAO, Martin-Estebané M, Barry-Carroll L, Chapman MA, Adorjan I, Tyler T, Turnbull B, Rose-Zerilli MJJ, Nicoll JAR, et al. (2022). The spatiotemporal dynamics of microglia across the human lifespan. Dev. Cell 57, 2127–2139.e6. 10.1016/j.devcel.2022.07.015. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Bennett ML, Bennett FC, Liddelow SA, Ajami B, Zamanian JL, Fernhoff NB, Mulinyawe SB, Bohlen CJ, Adil A, Tucker A, et al. (2016). New tools for studying microglia in the mouse and human CNS. Proc. Natl. Acad. Sci 113, E1738–E1746. 10.1073/pnas.1525528113. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Kenkhuis B, van Eekeren M, Parfitt DA, Ariyurek Y, Banerjee P, Priller J, van der Weerd L, and van Roon-Mom WMC (2022). Iron accumulation induces oxidative stress, while depressing inflammatory polarization in human iPSC-derived microglia. Stem Cell Rep. 17, 1351–1365. 10.1016/j.stemcr.2022.04.006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 52. Van Gassen S, Callebaut B, Van Helden MJ, Lambrecht BN, Demeester P, Dhaene T, and Saeys Y. (2015). FlowSOM: Using self-organizing maps for visualization and interpretation of cytometry data. Cytometry A 87, 636–645. 10.1002/cyto.a.22625. [ DOI ] [ PubMed ] [ Google Scholar ] 53. Cannoodt R, Saelens W, Sichien D, Tavernier S, Janssens S, Guilliams M, Lambrecht B, Preter KD, and Saeys Y. (2016). SCORPIUS improves trajectory inference and identifies novel modules in dendritic cell development. 079509. 10.1101/079509. [ DOI ] [ Google Scholar ] 54. Hopperton KE, Mohammad D, Trépanier MO, Giuliano V, Bazinet RP. (2018) Markers of microglia in post-mortem brain samples from patients with Alzheimer’s disease: a systematic review. Mol Psychiatry. 23(2),177–198. 10.1038/mp.2017.246. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 55. Corces MR, Shcherbina A, Kundu S, Gloudemans MJ, Frésard L, Granja JM, Louie BH, Eulalio T, Shams S, Bagdatli ST, et al. (2020). Single-cell epigenomic analyses implicate candidate causal variants at inherited risk loci for Alzheimer’s and Parkinson’s diseases. Nat Genet 52, 1158–1168. 10.1038/s41588-020-00721-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 56. Bang S, Donnelly CR, Luo X, Toro-Moreno M, Tao X, Wang Z, Chandra S, Bortsov AV, Derbyshire ER, and Ji R-R (2021). Activation of GPR37 in macrophages confers protection against infection-induced sepsis and pain-like behaviour in mice. Nat. Commun 12, 1704. 10.1038/s41467-021-21940-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 57. McCrary MR, Jiang MQ, Giddens MM, Zhang JY, Owino S, Wei ZZ, Zhong W, Gu X, Xin H, Hall RA, et al. (2019). Protective effects of GPR37 via regulation of inflammation and multiple cell death pathways after ischemic stroke in mice. FASEB J. 33, 10680–10691. 10.1096/fj.201900070r. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 58. Bang S, Xie Y-K, Zhang Z-J, Wang Z, Xu Z-Z, and Ji R-R (2018). GPR37 regulates macrophage phagocytosis and resolution of inflammatory pain. J. Clin. Investig 128, 3568–3582. 10.1172/jci99888. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 59. Zhang R, Shen Y, Zhang Q, Feng X, Liu X, Huo X, Sun J, and Hao J. (2023). TRIM21-mediated Sohlh2 ubiquitination suppresses M2 macrophage polarization and progression of triple-negative breast cancer. Cell Death Dis. 14, 850. 10.1038/s41419-023-06383-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 60. Chen J, Wang Z, Zheng Z, Chen Y, Khor S, Shi K, He Z, Wang Q, Zhao Y, Zhang H, et al. (2017). Neuron and microglia/macrophage-derived FGF10 activate neuronal FGFR2/PI3K/Akt signaling and inhibit microglia/macrophages TLR4/NF-κB-dependent neuroinflammation to improve functional recovery after spinal cord injury. Cell Death Dis. 8, e3090–e3090. 10.1038/cddis.2017.490. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 61. Dolga AM, Letsche T, Gold M, Doti N, Bacher M, Chiamvimonvat N, Dodel R, and Culmsee C. (2012). Activation of KCNN3/SK3/KCa2.3 channels attenuates enhanced calcium influx and inflammatory cytokine production in activated microglia. Glia 60, 2050–2064. 10.1002/glia.22419. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 62. Buchanan RA, Foley KE, Pepper KW, Reagan AM, Keezer KJ, Hewes AA, Diemler CA, Preuss C, Soto I, John SWM, et al. (2019). Meox2 Haploinsufficiency Accelerates Axonal Degeneration in DBA/2J Glaucoma. Investig. Ophthalmol. Vis. Sci 60, 3283–3296. 10.1167/iovs.18-26126. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 63. Chung J, Wang X, Maruyama T, Ma Y, Zhang X, Mez J, Sherva R, Takeyama H, Initiative TADN, Lunetta KL, et al. (2018). Genome-wide association study of Alzheimer’s disease endophenotypes at prediagnosis stages. Alzheimer’s Dement. 14, 623–633. 10.1016/j.jalz.2017.11.006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 64. Ding L, Chu M, Jiao Y, Hao Q, Xiao P, Li H, Guo Q, and Wang Y. (2018). TFDP3 regulates the apoptosis and autophagy in breast cancer cell line MDA-MB-231. PLoS ONE 13, e0203833. 10.1371/journal.pone.0203833. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Kim Y, Lee S, Zhang H, Lee S, Kim H, Kim Y, Won M-H, Kim Y-M, and Kwon Y-G (2020). CLEC14A deficiency exacerbates neuronal loss by increasing blood-brain barrier permeability and inflammation. J. Neuroinflammation 17, 48. 10.1186/s12974-020-1727-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 66. Su Z, Li Y, Lv H, Cui X, Liu M, Wang Z, Zhang Y, Zhen J, Tang W, Wang X, et al. (2021). CLEC14A protects against podocyte injury in mice with adriamycin nephropathy. FASEB J. 35, e21711. 10.1096/fj.202100283r. [ DOI ] [ PubMed ] [ Google Scholar ] 67. Minocha S, Valloton D, Arsenijevic Y, Cardinaux J-R, Guidi R, Hornung J-P, and Lebrand C. (2017). Nkx2.1 regulates the generation of telencephalic astrocytes during embryonic development. Sci. Rep 7, 43093. 10.1038/srep43093. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 68. Valero M, Viney TJ, Machold R, Mederos S, Zutshi I, Schuman B, Senzai Y, Rudy B, and Buzsáki G. (2021). Sleep down state-active ID2/Nkx2.1 interneurons in the neocortex. Nat. Neurosci 24, 401–411. 10.1038/s41593-021-00797-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 69. Thrupp N, Frigerio CS, Wolfs L, Skene NG, Fattorelli N, Poovathingal S, Fourne Y, Matthews PM, Theys T, Mancuso R, et al. (2020). Single-Nucleus RNA-Seq Is Not Suitable for Detection of Microglial Activation Genes in Humans. Cell Reports 32, 108189. 10.1016/j.celrep.2020.108189. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 70. Gimeno-Ferrer F, Eitner A, Noora N, Bauer R, Schmidt-Hieber C, Schaible H-G, and Richter F. (2024). Neuropathological changes in the TASTPM mouse model of Alzheimer’s disease and their relation to hyperexcitability and cortical spreading depolarization. Sci. Rep 14, 7224. 10.1038/s41598-024-57868-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 71. Prange O, Wong TP, Gerrow K, Wang YT, and El-Husseini A. (2004). A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. Proc. Natl. Acad. Sci 101, 13915–13920. 10.1073/pnas.0405939101. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 72. Andersen JV, Christensen SK, Westi EW, Diaz-delCastillo M, Tanila H, Schousboe A, Aldana BI, and Waagepetersen HS (2021). Deficient astrocyte metabolism impairs glutamine synthesis and neurotransmitter homeostasis in a mouse model of Alzheimer’s disease. Neurobiol. Dis 148, 105198. 10.1016/j.nbd.2020.105198. [ DOI ] [ PubMed ] [ Google Scholar ] 73. Yang Z, and Wang KKW (2015). Glial fibrillary acidic protein: from intermediate filament assembly and gliosis to neurobiomarker. Trends Neurosci. 38, 364–374. 10.1016/j.tins.2015.04.003. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 74. Aloi MS, Prater KE, Sánchez REA, Beck A, Pathan JL, Davidson S, Wilson A, Keene CD, Iglesia H. de la, Jayadev, S., et al. (2023). Microglia specific deletion of miR-155 in Alzheimer’s disease mouse models reduces amyloid-β pathology but causes hyperexcitability and seizures. J. Neuroinflammation 20, 60. 10.1186/s12974-023-02745-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. Das M, Mao W, Voskobiynyk Y, Necula D, Lew I, Petersen C, Zahn A, Yu G-Q, Yu X, Smith N, et al. (2023). Alzheimer risk-increasing TREM2 variant causes aberrant cortical synapse density and promotes network hyperexcitability in mouse models. Neurobiol. Dis 186, 106263. 10.1016/j.nbd.2023.106263. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 76. Griciuc A, Serrano-Pozo A, Parrado AR, Lesinski AN, Asselin CN, Mullin K, Hooli B, Choi SH, Hyman BT, and Tanzi RE (2013). Alzheimer’s Disease Risk Gene CD33 Inhibits Microglial Uptake of Amyloid Beta. Neuron 78, 631–643. 10.1016/j.neuron.2013.04.014. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 77. Walker DG, and Lue LF (2015). Immune phenotypes of microglia in human neurodegenerative disease: Challenges to detecting microglial polarization in human brains. Alzheimer’s Res Ther 7, 1–9. 10.1186/s13195-015-0139-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 78. Rangaraju S, Dammer EB, Raza SA, Rathakrishnan P, Xiao H, Gao T, Duong DM, Pennington MW, Lah JJ, Seyfried NT, et al. (2018). Identification and therapeutic modulation of a pro-inflammatory subset of disease-associated-microglia in Alzheimer’s disease. Mol Neurodegener 13, 24. 10.1186/s13024-018-0254-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 79. Boche D, and Gordon MN (2022). Diversity of transcriptomic microglial phenotypes in aging and Alzheimer’s disease. Alzheimer’s Dement. 18, 360–376. 10.1002/alz.12389. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 80. Sepulveda J, Kim JY, Binder J, Vicini S, and Rebeck GW (2024). APOE4 genotype and aging impair injury-induced microglial behavior in brain slices, including toward Aβ, through P2RY12. Mol. Neurodegener 19, 24. 10.1186/s13024-024-00714-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 81. Kenkhuis B, Somarakis A, Kleindouwel LRT, van Roon-Mom WMC, Höllt T, and van der Weerd L (2022). Co-expression patterns of microglia markers Iba1, TMEM119 and P2RY12 in Alzheimer’s disease. Neurobiol. Dis 167, 105684. 10.1016/j.nbd.2022.105684. [ DOI ] [ PubMed ] [ Google Scholar ] 82. Lau S-F, Wu W, Wong HY, Ouyang L, Qiao Y, Xu J, Lau JH-Y, Wong C, Jiang Y, Holtzman DM, et al. (2023). The VCAM1–ApoE pathway directs microglial chemotaxis and alleviates Alzheimer’s disease pathology. Nat. Aging 3, 1219–1236. 10.1038/s43587-023-00491-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 83. Dawe RJ, Yu L, Arfanakis K, Schneider JA, Bennett DA, and Boyle PA (2020). Late-life cognitive decline is associated with hippocampal volume, above and beyond its associations with traditional neuropathologic indices. Alzheimer’s Dement. 16, 209–218. 10.1002/alz.12009. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 84. Persson K, Bohbot VD, Bogdanovic N, Selbæk G, Brækhus A, and Engedal K. (2018). Finding of increased caudate nucleus in patients with Alzheimer’s disease. Acta Neurol. Scand 137, 224–232. 10.1111/ane.12800. [ DOI ] [ PubMed ] [ Google Scholar ] 85. Mattiace LA, Davies P, and Dickson DW (1990). Detection of HLA-DR on microglia in the human brain is a function of both clinical and technical factors. Am. J. Pathol 136, 1101–1114. [ PMC free article ] [ PubMed ] [ Google Scholar ] 86. Hendrickx DAE, van Eden CG, Schuurman KG, Hamann J, and Huitinga I. (2017). Staining of HLA-DR, Iba1 and CD68 in human microglia reveals partially overlapping expression depending on cellular morphology and pathology. J. Neuroimmunol 309, 12–22. 10.1016/j.jneuroim.2017.04.007. [ DOI ] [ PubMed ] [ Google Scholar ] 87. Albergante L, Mirkes E, Bac J, Chen H, Martin A, Faure L, Barillot E, Pinello L, Gorban A, and Zinovyev A. (2020). Robust and Scalable Learning of Complex Intrinsic Dataset Geometry via ElPiGraph. Entropy 22, 296. 10.3390/e22030296. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 88. Zeng H, Huang J, Zhou H, Meilandt WJ, Dejanovic B, Zhou Y, Bohlen CJ, Lee S-H, Ren J, Liu A, et al. (2023). Integrative in situ mapping of single-cell transcriptional states and tissue histopathology in a mouse model of Alzheimer’s disease. Nat. Neurosci 26, 430–446. 10.1038/s41593-022-01251-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 89. Mancuso R, Fattorelli N, Martinez-Muriana A, Davis E, Wolfs L, Daele JVD, Geric I, Preman P, Serneels L, Poovathingal S, et al. (2022). A multi-pronged human microglia response to Alzheimer’s disease Aβ pathology. 2022.07.07.499139. 10.1101/2022.07.07.499139. [ DOI ] [ Google Scholar ] 90. Zapata CF, Giacomello G, Spruth EJ, Middeldorp J, Gallaccio G, Dehlinger A, Dames C, Leman JKH, Dijk R.E. van, Meisel A, et al. (2022). Differential compartmentalization of myeloid cell phenotypes and responses towards the CNS in Alzheimer’s disease. Nat. Commun 13, 7210. 10.1038/s41467-022-34719-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Methods-only References 91. Beach TG et al. (2015) Arizona Study of Aging and Neurodegenerative Disorders and Brain and Body Donation Program. Neuropathology. 35(4):354–89. 10.1111/neup.12189. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 92. Montine TJ, Phelps CH, Beach TG, Bigio EH, Cairns NJ, Dickson DW, Duyckaerts C, Frosch MP, Masliah E, Mirra SS, et al. (2012). National Institute on Aging-Alzheimer’s Association guidelines for the neuropathologic assessment of Alzheimer’s disease: a practical approach. Acta Neuropathol. (Berl.) 123, 1–11. 10.1007/s00401-011-0910-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 93. Freedman D, and Diaconis P. (1981). On the histogram as a density estimator:L2 theory. Z. Für Wahrscheinlichkeitstheorie Verwandte Geb. 57, 453–476. 10.1007/BF01025868. [ DOI ] [ Google Scholar ] 94. Karatzoglou A, Smola A, Hornik K, and Zeileis A. (2004). kernlab - An S4 Package for Kernel Methods in R. J. Stat. Softw 11, 1–20. 10.18637/jss.v011.i09. [ DOI ] [ Google Scholar ] 95. Liu CC, Greenwald NF, Kong A, McCaffrey EF, Leow KX, Mrdjen D, Cannon BJ, Rumberger JL, Varra SR, Angelo M. Robust phenotyping of highly multiplexed tissue imaging data using pixel-level clustering. Nat Commun. 2023. Aug 1;14(1):4618. doi: 10.1038/s41467-023-40068-5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 96. Granja JM, Corces MR, Pierce SE, Bagdatli ST, Choudhry H, Chang HY, Greenleaf WJ. (2021) ArchR is a scalable software package for integrative single-cell chromatin accessibility analysis. Nat Genet. Mar;53(3):403–411. 10.1038/s41588-021-00790-6 . Erratum in: Nat Genet. 2021 Jun;53(6):935. 10.1038/s41588-021-00850-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 97. Cannon B. and Bendall S. (2024). Spatial proteomics reveals human microglial states shaped by anatomy and neuropathology - DATA. Stanford Digital Repository. 10.25740/sq409qv3664. [ DOI ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplemental Figures NIHMS2145041-supplement-Supplemental_Figures.pdf (6.1MB, pdf) Supplementary Tables NIHMS2145041-supplement-Supplementary_Tables.xlsx (73.4KB, xlsx) Data Availability Statement All imaging data and cell tables are made available in a public repository here https://doi.org/10.25740/sq409qv3664 97 . Single-nuclear ATAC data previously processed can be found here in GSE147672 ACTIONS View on publisher site PDF (6.0 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top