Abstract

Microglia have critical roles not only in neural development and homeostasis, but also in neurodegenerative and neuroinflammatory diseases of the central nervous system1,2,3,4. These highly diverse and specialized functions may be executed by subsets of microglia that already exist in situ, or by specific subsets of microglia that develop from a homogeneous pool of cells on demand. However, little is known about the presence of spatially and temporally restricted subclasses of microglia in the central nervous system during development or disease. Here we combine massively parallel single-cell analysis, single-molecule fluorescence in situ hybridization, advanced immunohistochemistry and computational modelling to comprehensively characterize subclasses of microglia in multiple regions of the central nervous system during development and disease. Single-cell analysis of tissues of the central nervous system during homeostasis in mice revealed specific time- and region-dependent subtypes of microglia. Demyelinating and neurodegenerative diseases evoked context-dependent subtypes of microglia with distinct molecular hallmarks and diverse cellular kinetics. Corresponding clusters of microglia were also identified in healthy human brains, and the brains of patients with multiple sclerosis. Our data provide insights into the endogenous immune system of the central nervous system during development, homeostasis and disease, and may also provide new targets for the treatment of neurodegenerative and neuroinflammatory pathologies.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Data availability

Raw data for mouse and human single cell RNA-sequencing have been deposited in the Gene Expression Omnibus, and are available at the following accession numbers: GSE120629 (mouse), GSE120747 (mouse) and GSE124335 (human). All other data are available from the corresponding author on reasonable request.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    Prinz, M. & Priller, J. Microglia and brain macrophages in the molecular age: from origin to neuropsychiatric disease. Nat. Rev. Neurosci. 15, 300–312 (2014).

  2. 2.

    Colonna, M. & Butovsky, O. Microglia function in the central nervous system during health and neurodegeneration. Annu. Rev. Immunol. 35, 441–468 (2017).

  3. 3.

    Herz, J., Filiano, A. J., Smith, A., Yogev, N. & Kipnis, J. Myeloid cells in the central nervous system. Immunity 46, 943–956 (2017).

  4. 4.

    Prinz, M., Erny, D. & Hagemeyer, N. Ontogeny and homeostasis of CNS myeloid cells. Nat. Immunol. 18, 385–392 (2017).

  5. 5.

    Kierdorf, K. et al. Microglia emerge from erythromyeloid precursors via Pu.1- and Irf8-dependent pathways. Nat. Neurosci. 16, 273–280 (2013).

  6. 6.

    Hagemeyer, N. et al. Microglia contribute to normal myelinogenesis and to oligodendrocyte progenitor maintenance during adulthood. Acta Neuropathol. 134, 441–458 (2017).

  7. 7.

    Matcovitch-Natan, O. et al. Microglia development follows a stepwise program to regulate brain homeostasis. Science 353, aad8670 (2016).

  8. 8.

    Mathys, H. et al. Temporal tracking of microglia activation in neurodegeneration at single-cell resolution. Cell Reports 21, 366–380 (2017).

  9. 9.

    Jordao, M. J. et al. Single-cell profiling identifies myeloid cell subsets with distinct fates during neuroinflammation. Science 363, eaat7554 (2019).

  10. 10.

    Keren-Shaul, H. et al. A unique microglia type associated with restricting development of Alzheimer’s disease. Cell 169, 1276–1290 (2017).

  11. 11.

    Grabert, K. et al. Microglial brain region-dependent diversity and selective regional sensitivities to aging. Nat. Neurosci. 19, 504–516 (2016).

  12. 12.

    Herman, J. S., Sagar, & Grün, D. FateID infers cell fate bias in multipotent progenitors from single-cell RNA-seq data. Nat. Methods 15, 379–386 (2018).

  13. 13.

    Krasemann, S. et al. The TREM2–APOE pathway drives the transcriptional phenotype of dysfunctional microglia in neurodegenerative diseases. Immunity 47, 566–581.e9 (2017).

  14. 14.

    Tay, T. L. et al. A new fate mapping system reveals context-dependent random or clonal expansion of microglia. Nat. Neurosci. 20, 793–803 (2017).

  15. 15.

    Mildner, A. et al. Microglia in the adult brain arise from Ly-6ChiCCR2+ monocytes only under defined host conditions. Nat. Neurosci. 10, 1544–1553 (2007).

  16. 16.

    Benveniste, E. N. Role of macrophages/microglia in multiple sclerosis and experimental allergic encephalomyelitis. J. Mol. Med. (Berl.) 75, 165–173 (1997).

  17. 17.

    Brück, W. et al. Monocyte/macrophage differentiation in early multiple sclerosis lesions. Ann. Neurol. 38, 788–796 (1995).

  18. 18.

    Lawson, L. J., Perry, V. H. & Gordon, S. Turnover of resident microglia in the normal adult mouse brain. Neuroscience 48, 405–415 (1992).

  19. 19.

    Doorn, K. J. et al. Brain region-specific gene expression profiles in freshly isolated rat microglia. Front. Cell. Neurosci. 9, 84 (2015).

  20. 20.

    Wang, Y. et al. IL-34 is a tissue-restricted ligand of CSF1R required for the development of Langerhans cells and microglia. Nat. Immunol. 13, 753–760 (2012).

  21. 21.

    Greter, M. et al. Stroma-derived interleukin-34 controls the development and maintenance of Langerhans cells and the maintenance of microglia. Immunity 37, 1050–1060 (2012).

  22. 22.

    De Biase, L. M. et al. Local cues establish and maintain region-specific phenotypes of basal ganglia microglia. Neuron 95, 341–356.e6 (2017).

  23. 23.

    Ayata, P. et al. Epigenetic regulation of brain region-specific microglia clearance activity. Nat. Neurosci. 21, 1049–1060 (2018).

  24. 24.

    Hammond, T. R. et al. 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 (2019).

  25. 25.

    Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner Bioinformatics 29, 15–21 (2013).

  26. 26.

    Liao, Y., Smyth, G. K. & Shi, W. The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 41, e108 (2013).

  27. 27.

    Gu, Z., Eils, R. & Schlesner, M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32, 2847–2849 (2016).

  28. 28.

    Hashimshony, T. et al. CEL-Seq2: sensitive highly-multiplexed single-cell RNA-seq. Genome Biol. 17, 77 (2016).

  29. 29.

    Li, H. & Durbin, R. Fast and accurate long-read alignment with Burrows–Wheeler transform. Bioinformatics 26, 589–595 (2010).

  30. 30.

    Grün, D., Kester, L. & van Oudenaarden, A. Validation of noise models for single-cell transcriptomics. Nat. Methods 11, 637–640 (2014).

  31. 31.

    Grün, D. et al. Single-cell messenger RNA sequencing reveals rare intestinal types. Nature 525, 251–255 (2015).

  32. 32.

    Goldmann, T. et al. USP18 lack in microglia causes destructive interferonopathy of the mouse brain. EMBO J. 34, 1612–1629 (2015).

  33. 33.

    Raasch, J. et al. IκB kinase 2 determines oligodendrocyte loss by non-cell-autonomous activation of NF-κB in the central nervous system. Brain 134, 1184–1198 (2011).

  34. 34.

    Kuhn, A., Luthi-Carter, R. & Delorenzi, M. Cross-species and cross-platform gene expression studies with Bioconductor-compliant R package ‘annotationTools’. BMC Bioinformatics 9, 26 (2008).

  35. 35.

    Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).

  36. 36.

    Lucchinetti, C. et al. Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination. Ann. Neurol. 47, 707–717 (2000).

Download references

Acknowledgements

We thank M. Ditter, E. Barleon, T. el Gaz and J. Bodinek-Wersing for technical assistance; C. Gross for proofreading; N. Mossadegh for helpful discussions; and A. Zeisel and S. Linnarsson (Karolinska Institute, Sweden) for sharing data and support. The samples and data used for this project were kindly provided by the West German Biobank. T.M. was supported by the KANAE Foundation and the Japan Society for the Promotion of Science. M.P. is supported by the Sobek Foundation, the Ernst-Jung Foundation, the DFG (SFB 992, SFB1160, SFB/TRR167, Reinhart Koselleck Grant), the Ministry of Science, Research and Arts, Baden-Wuerttemberg (Sonderlinie ‘Neuroinflammation’) and the BMBF-funded competence network of multiple sclerosis (KKNMS). This study was supported by the DFG under Germany’s Excellence Strategy (CIBSS – EXC-2189 – Project ID390939984). J.P. is supported by the DFG (SFB/TRR167 B05 and B07), BIH CRG 2a, and the UK DRI Momentum Award. C. Stadelmann is supported by the DFG (STA1389/2-1), the Deutsche Multiple Sklerose Gesellschaft (DMSG), the National MS Society of the USA (NMSS) and the Hertie Foundation. C.B. is supported by the DFG (SFB/TRR167 B05).

Reviewer information

Nature thanks Arnold Kriegstein and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

Author notes

  1. These authors contributed equally: Takahiro Masuda, Roman Sankowski, Ori Staszewski

Affiliations

  1. Institute of Neuropathology, Medical Faculty, University of Freiburg, Freiburg, Germany

    • Takahiro Masuda
    • , Roman Sankowski
    • , Ori Staszewski
    • , Lukas Amann
    •  & Marco Prinz
  2. Berta-Ottenstein-Programme for Clinician Scientists, Faculty of Medicine, University of Freiburg, Freiburg, Germany

    • Roman Sankowski
  3. Department of Neuropsychiatry and Laboratory of Molecular Psychiatry, Charité – Universitätsmedizin Berlin, Berlin, Germany

    • Chotima Böttcher
    •  & Josef Priller
  4. Faculty of Biology, University of Freiburg, Freiburg, Germany

    • Lukas Amann
  5. Clinic for Neurosurgery, Faculty of Medicine, University of Freiburg, Freiburg, Germany

    • Christian Scheiwe
  6. Institute of Neuropathology, University Medical Center Göttingen, Göttingen, Germany

    • Stefan Nessler
    • , Patrik Kunz
    •  & Christine Stadelmann
  7. VIB Center for Inflammation Research, Ghent, Belgium

    • Geert van Loo
  8. Department of Biomedical Molecular Biology, Ghent University, Ghent, Belgium

    • Geert van Loo
  9. Department of Stereotactic and Functional Neurosurgery, Medical Faculty, University of Freiburg, Freiburg, Germany

    • Volker Arnd Coenen
    •  & Peter Christoph Reinacher
  10. Department of Neurosurgery, University Hospital Essen, Essen, Germany

    • Anna Michel
    •  & Ulrich Sure
  11. Department of Neurology, St. Josef Hospital, Ruhr University Bochum, Bochum, Germany

    • Ralf Gold
  12. DZNE, Berlin, Germany

    • Josef Priller
  13. UK DRI, University of Edinburgh, Edinburgh, UK

    • Josef Priller
  14. Signalling Research Centres BIOSS and CIBSS, University of Freiburg, Freiburg, Germany

    • Marco Prinz
  15. Center for NeuroModulation, Faculty of Medicine, University of Freiburg, Frieburg, Germany

    • Marco Prinz

Authors

  1. Search for Takahiro Masuda in:

  2. Search for Roman Sankowski in:

  3. Search for Ori Staszewski in:

  4. Search for Chotima Böttcher in:

  5. Search for Lukas Amann in:

  6. Search for Christian Scheiwe in:

  7. Search for Stefan Nessler in:

  8. Search for Patrik Kunz in:

  9. Search for Geert van Loo in:

  10. Search for Volker Arnd Coenen in:

  11. Search for Peter Christoph Reinacher in:

  12. Search for Anna Michel in:

  13. Search for Ulrich Sure in:

  14. Search for Ralf Gold in:

  15. Search for Josef Priller in:

  16. Search for Christine Stadelmann in:

  17. Search for Marco Prinz in:

Contributions

T.M., R.S., O.S., C.B., L.A., C. Scheiwe, S.N., P.K., G.v.L., V.A.C., P.C.R., A.M., U.S. and R.G. conducted experiments and analysed the data. M.P., C. Stadelmann and J.P. analysed the data, contributed to the in vivo studies and provided mice or reagents. T.M. and M.P. supervised the project and wrote the manuscript.

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to Marco Prinz.

Extended data figures and tables

  1. Extended Data Fig. 1 Comprehensive analysis of microglial diversity by scRNA-seq.

    a, Illustration depicting the workflow for the isolation of microglia from different regions of the CNS of embryonic (E16.5), juvenile (3 weeks of age) and adult (16 weeks of age) mice during homeostasis and during pathology (FNX or cuprizone-mediated demyelination for scRNA-seq). b, t-SNE plot showing 3,826 analysed microglial cells from different conditions tested in this study. Each dot represents a single cell. c, t-SNE plot depicting 13 clusters for all of the different conditions (3,826 cells). Colours represent each cluster. d, t-SNE plots depicting single microglia from the replicates from different regions of the CNS of individual embryos, juvenile and adult mice and diseased mice (3,826 cells).

  2. Extended Data Fig. 2 Subpopulations of microglia with distinct gene expression during development, in mice.

    a, Distribution of Malat1 gene expression in a t-SNE plot (2,966 microglia cells). Colour keys represent the respective expression levels. The right cloud represents the embryonic microglial population, whereas the left cloud combines both juvenile and adult microglia as shown in Fig. 1a. b, Left, single-molecule fluorescent in situ hybridization (smFISH) for Malat1 and Cx3cr1 shows the kinetics of Malat1+ microglia during development. Scale bar, 10 μm. Representative images out of two adult mice investigated are shown. Yellow and white arrowheads indicate Malat1+Cx3cr1+ microglia and Malat1+Cx3cr1 non-microglial cells, respectively. Right, frequency of Malat1+ microglia in the forebrain or cortex during development. Bar represents mean ± s.e.m. of 120 cells studied, from 3 animals per time point. c, t-SNE plot of Ctsd and Lamp1 gene expression; these genes were enriched in the C1 and C2 clusters, as shown in Fig. 1b. d, t-SNE plot of Eef1a1 and Rpl4 gene expression; these genes were enriched in the C6 cluster, as shown in Fig. 1b.

  3. Extended Data Fig. 3 CST3 is enriched in adult microglia.

    ac, Representative sections of the cortex from adult mice using immunofluoresence for CST3 (green), NeuN for neurons (red, a) and adenomatous polyposis coli (APC) for oligodendrocytes (red, b). The astrocyte marker ALDH1L1 (red, c), combined with CST3, was used on the hippocampal sections. Scale bars, 50 μm (overview), 20 μm (magnification). Representative images out of three mice investigated are shown. d, Quantification of CST3 immunoreactivity in the brain of the adult mouse. Bar represents mean ± s.e.m. of 3 animals (393 microglia, 1,817 neurons, 298 oligodendrocytes and 461 astrocytes).

  4. Extended Data Fig. 4 Diverse clusters of microglia during demyelination and neurodegeneration.

    a, Projection of 1,564 single microglia isolated from different regions of the CNS during homeostasis or FNX or cuprizone treatment, shown as a t-SNE plot. b, Heat map of top differentially regulated genes that were up- or downregulated in each cluster. The genes with the highest differentially expression are highlighted. c, t-SNE plot exhibiting 13 clusters for the 1,564 individual microglia isolated from different regions of the CNS during homeostasis or FNX or cuprizone treatment.

  5. Extended Data Fig. 5 Molecular characterization of subpopulations of microglia during demyelination and remyelination.

    ac, e, g, i, j, t-SNE plots showing expression of Axl, Igf1, Lyz2, Itgax, Gpnmb and Apoc1 (a); Fam20c, Ccl6, Psat1 and Ank (b); Fn1 (c); Spp1 (e); Cybb (g); H2-Aa and H2-Ab1 (i); and Ccl4 (j) transcripts after cuprizone challenge (3,826 microglial cells). Genes shown in a were upregulated in both demyelination and remyelination, whereas genes depicted in b, c and e or g and i were increased in demyelination-associated clusters C12, or in the remyelination- associated clusters C13, respectively. Colour keys represent the respective expression levels. d, Left and middle, smFISH for Fn1 and Cx3cr1 reveals subpopulations of microglia after a five-week cuprizone treatment in the corpus callosum. ‘1’ indicates Fn1+Cx3cr1+ microglia (yellow arrowheads, Fn1 mRNA); ‘2’ indicates Fn1Cx3cr1+ microglia. Scale bar, 10 μm (overview), 3 μm (inset). Representative images out of three mice investigated are shown. Non-filled arrowheads in the image of control mice indicate non-specific signals. Right, percentage of Fn1+ Cx3cr1+ microglia in the corpus callosum. Bar represents mean ± s.e.m. of 3 animals (168 cells investigated). f, Left and middle, smFISH for Spp1 and Cx3cr1 reveals subpopulations of microglia after a five-week cuprizone treatment in the corpus callosum. ‘1’ indicates Spp1+Cx3cr1+ microglia (yellow arrowheads, Spp1 mRNA); ‘2’ indicates Spp1Cx3cr1+ microglia. Scale bars, 10 μm (overview), 3 μm (insets). Representative images out of three mice investigated are shown. Right, percentage of Spp1+Cx3cr1+ microglia in the corpus callosum. Bar represents mean ± s.e.m. of 3 animals (165 investigated cells). h, Left and middle, smFISH for Cybb and Cx3cr1 reveals subpopulations of microglia after a five-week cuprizone treatment in the corpus callosum. ‘1’ indicates Cybb+Cx3cr1+ microglia (yellow arrowheads, Cybb mRNA); ‘2’ indicates CybbCx3cr1+ microglia. Scale bars, 10 μm (overview) and 3 μm (insets). Representative images out of three mice investigated are shown. Right, percentage of Cybb+Cx3cr1+ microglia in the corpus callosum. Bar represents mean ± s.e.m. of 3 animals (165 investigated cells). The colour key represents the expression levels.

  6. Extended Data Fig. 6 Microglial subtypes in healthy human brains.

    a, t-SNE plot of 1,180 human microglia showing the distribution of individual microglia from 5 patients. Each dot represents a single cell. Different colours indicate different patients. b, t-SNE plot of 1,180 individual human microglia isolated from 5 individual non-pathological brains depicts four major clusters (HHu-C1 to HHu-C4). Each dot represents a single cell. Colours correspond to each cluster. c, Heat map showing the distribution of the healthy human clusters in each individual patient. d, Bar graphs representing the relative abundance of microglial cells in the respective clusters from five individual non-pathological brains. Colours represent distinct clusters. e, Heat map of the top differentially regulated genes that were up- or downregulated in each cluster. f, t-SNE plots for CST3, P2RY13, CCL4 and EGR2 mRNA expression. CCL4 and EGR2 are enriched in the cluster HHu-C4.

  7. Extended Data Fig. 7 Detailed neuropathological characterization of lesions of multiple sclerosis.

    Histology of the brains of patients with multiple sclerosis (labelled as patients MS-1 to MS-5) using haematoxylin and eosin (HE), luxol fast blue (LFB-PAS), 2′,3′-cyclic-nucleotide 3′-phosphodiesterase (CNPase) and myelin basic protein (MBP) for myelin, human leukocyte antigen DR isotype (HLA-DR) and CD68 for myeloid cells, CD3 for T cells, CD20 for B cells and Bielschowsky (Biel) for axons. Scale bar, 50 μm. Lesions are typical early active multiple-sclerosis lesions, according to the standard classification system36.

  8. Extended Data Fig. 8 Molecular profile of subsets of microglia during multiple sclerosis.

    a, t-SNE plots representing the core signature genes for lymphocytes (TRAC, TRBC2, CD52 and IL32), myeloid cells (ITGAM, MS4A6A, TYROBP and CD14) and monocytes (CCR2, PLAC8, CLEC12A and FCN1) in the brains of patients with multiple sclerosis. Colour keys reflect the expression levels. bf, t-SNE plots of genes that are enriched in clusters Hu-C5 to Hu-C7 (b), Hu-C2 (c), Hu-C3 (d), Hu-C4 (e) and Hu-C8 (f) are shown (1,602 microglia cells). Colour codes represent expression levels. g, t-SNE plots depicting genes that are upregulated in the clusters Hu-C2, Hu-C3 and Hu-C8. Colour codes represent expression levels. h, t-SNE plots of genes that were upregulated in the disease-associated subsets of microglia in the mouse demyelination model, but not in the microglia in the brains of patients with multiple sclerosis. Colour codes represent expression levels. i, Immunofluorescence images for TMEM119 and IBA1 in healthy brains or the brains of patients with multiple sclerosis. Arrowheads indicate TMEM119+IBA1+ cells (filled) in the healthy brains, and TMEM119IBA1+ microglia (open) during multiple sclerosis. Representative images out of four mice investigated, per condition, are shown. Scale bar, 50 μm. j, Representative immunofluorescence images for IBA1+MRP14 (indicating microglia) and IBA1+MRP14+ cells (representing infiltrating early activated monocytes) in the healthy brain and brains of patients with multiple sclerosis. Insets show microglia (top row) and monocytes (bottom row) in the multiple-sclerosis lesion. Right, quantification. Bars represent means ± s.e.m. (n = 4 for each condition). Each symbol represents one patient. Scale bars, 50 μm (overview), 4 μm (inset).

  9. Extended Data Fig. 9 Canonical correlation analysis of scRNA-seq data from mouse and human microglia.

    a, Canonical correlation analysis (Seurat alignment procedure) visualizing shared correlation structures (that is, canonical correlation vectors (CCs)) between mouse and human datasets. Each dot represents single cell. b, CC plot of cells assigned as mouse clusters C7 to C13 and human clusters Hu-C1 to Hu-C10. Mouse demyelination-related clusters of microglia (C12 and C13) are transcriptionally close to human multiple sclerosis-associated clusters of microglia (Hu-C2, Hu-C3 and Hu-C8). Each dot represents single cell. c, Violin plots depicting a shared gene correlation structure that is conserved between mouse and human clusters. Values for CC1 and CC2 vectors for individual cells are shown.

  10. Extended Data Fig. 10

    Graphical abstract of experimental findings.

Supplementary information

  1. Supplementary Figure

    This file contains Supplementary Figure 1: FACS gating strategy for microglia isolation. CNS cells were gated for G1 and G2 (singlets), followed by being gated for living cell (G3, fixable viability dye), CD45intCD11b+ (G4), Ly6C-Ly6G- (G5), and CD206- (G6).

  2. Reporting Summary

  3. Supplementary Table

    This file contains Supplementary Table split between mouse and human data. The mouse data shows information on mice and cells used in each scRNA-seq analysis including genotype, strain, sex, condition, age, CNS region and cell numbers. The human data shows details on patients and cells used in each scRNAseq analysis including sex, condition, age, CNS region and cell numbers.

About this article

Publication history

Received

Accepted

Published

Issue Date

DOI

https://doi.org/10.1038/s41586-019-0924-x

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.