Genetic studies of late-onset Alzheimer’s point to variations in the function of the innate immune system—and its CNS cadre, the microglia. Why would immune genes determine whether a person will get the disease? Perhaps because the innate immune system forms a first line of defense against toxic amyloid and tau proteins. But how to move from GWAS hits to understanding this biology? One approach that is paying dividends is to stop mashing up brain tissue and instead analyze single cells, and a pair of recent papers are suddenly propelling the field forward. In addition, two studies delve into the details of the immune response to amyloid and tau, using transcriptomics in animal models, to begin to connect the dots between genes and disease.

  • Genetic studies of late-onset AD find genes expressed in microglia.
  • Single-cell analysis reveals microglia are highly diverse, with specific subsets appearing in old age and after brain injury.
  • AD risk genes affect microglia responses to amyloid, and tau pathologies.

In one foundational study, the groups of Beth Stevens of Boston Children’s Hospital and Steven McCarroll of Harvard Medical School, also in Boston, jointly characterized gene expression in 79,000 mouse microglia, one cell at a time, sampled across the animals’ life spans. In the November 21 Immunity, they expose immense diversity among microglia, and glimpses of how these populations change in aging and brain injury. In another foundational study, Pandurangan Vijayanand and colleagues at La Jolla Institute for Allergy and Immunology, California, catalogued gene expression in 13 different human immune cell types as a function of genetic polymorphisms. Their work debuted in Cell on November 15, and comes with an open database for others to use.

How will such information generate new insight? By providing a resource for studies of human microglial response genes. In a preprint currently on bioRχiv, John Hardy and collaborators at University College London used transcriptomics to identify new candidate AD genes that influence the microglial response to amyloid. Which variants of these genes a person has inherited likely influences how deftly, and safely, their microglia dispose of age-related amyloid. On the tau side, in the November 3 Neuron, Hui Zheng and colleagues at Baylor College of Medicine in Houston identify an immune network that gets activated in response to pathogenic tau. Hello again, microglia: Central to this network is the complement protein C3 and its microglial receptor, both of which figure in synapse loss and neurodegeneration. Variants in these sets of genes can now be explored for their effects on expression in microglial subsets, and in the human immune cell expression database.

Microglia perform multiple roles in the brain throughout life. In development, they prune synapses and assist myelination; in adulthood, they patrol the brain and attend to injuries. When microglia go berserk, they contribute to neurodegeneration, as they do in Alzheimer’s. The old, simplistic classification of microglia as resting, inflammatory, or anti-inflammatory has gone by the wayside as single-cell RNA sequencing studies have laid bare more diversity, including specialized subsets associated with amyloid deposition or neurodegeneration (Jul 2018 conference news; Keren-Shaul et al., 2017; Mathys et al., 2017). These cells can be a force for good or ill in the brain—helping to mop up toxic amyloid, or unleashing deadly neuroinflammation. Which road microglia take in a given person is turning out to be determined by genetic variation tied to AD risk (for review, see Hansen et al., 2018). 

Stevens’ new work adds critically needed context for those studies, and more to come. “We realized that before we can really understand what’s going on in AD, we needed a deep dive into the diversity of normal microglia, from embryonic development to aging. We need to understand the different populations and how they change over time,” she told Alzforum.

To do that, first author Timothy Hammond led a team that isolated and sequenced 76,149 individual microglia from 41 mice, male and female, prenatal to aged, and after brain injury. Because microglia respond rapidly to perturbations in their environment, Hammond prepared cells entirely in the cold to minimize activation. In all, the investigators cataloged nine transcriptionally distinct microglia subtypes. The prenatal and newborn brains harbored the most diversity of microglia. The six main subtypes in the young brains were marked by genes that were strongly upregulated or unique to each subset. Based on their gene-expression profiles and regional distribution in the brain, the subsets appeared to subserve different functions.

Microglia Map.

Clustering analysis of transcriptome data from mouse microglia turned up nine distinct subtypes, some with their own divisions. Cluster 7 predominates in adult mice, cluster 8 appears in aged mice, cluster 9 after brain injury. [Courtesy of Hammond et al., Immunity 2018.]

In the healthy adult brains, microglia became more uniform, with one major subtype. Alas, with advancing age, they diversified again. Old mice had up to fourfold more microglia expressing the chemokine Ccl4 and other inflammation markers scattered throughout the brain. Stevens suspects that while small, this population may be a culprit in age-related neuroinflammation.

After injury, a new sort of microglia arose. When the investigators induced a multiple sclerosis-like, focal white-matter lesion with an injection of lysolecithin, they detected microglia that had downregulated canonical genes such as Fcrls, P2ry12, Cx3cr1, Trem2, and C1qa, and upregulated APOE, interferon-responsive and cytokine genes associated with microglial activation. The injury-induced microglia were not all the same: Some selectively increased expression of Ccl4, others the interferon-responsive gene Ifi2712a, or proliferative genes. In brain tissue from the mice, the injury-induced microglia clustered around the lesion, while microglia in distant areas kept an undisturbed phenotype. The results might hold in human disease: the investigators found Ccl4-positive microglia lurking around MS lesions in human brain samples.

Some injury-associated microglia upped expression of genes that marked developing microglia, supporting a role for reawakened development pathways in disease. One notable example: Members of the Ms4a family. Ms4a genes encode transmembrane lipid sensors that function in the microglial response to membrane damage, and variants of Ms4a are linked to Alzheimer’s risk (Jul 2018 news). 

Stevens and colleagues compared injury-response microglia to recently described disease-associated microglia from AD mice (DAMS, Jun 2017 news). Both shared a transcriptional signature of 12 core genes, including Spp1, APOE, and Lpl. However, the scientists report that lysolecithin injury-induced microglia express additional, unique genes. The authors speculate that microglia may mount some common responses to brain injury or disease, but also lesion-specific responses. With these initial results in hand, Stevens said analysis of microglia in other models and diseases is ongoing now.

Stevens was surprised to find no sex differences in the constellation of microglia in young, healthy animals. Perhaps age or challenge bring out such differences, and the authors have not yet compared aged or injured mice by sex, Stevens told Alzforum.

The investigators have created an open, searchable resource where users can probe gene-expression patterns in this data set.

The data is but a start, Stevens said. “The next step will be doing this with human brain. Most importantly, we want to understand how transcriptional signatures in individual cells fit back to what the GWAS and genetics are telling us,” she said. Work on transcriptional profiling of human microglia has started in multiple labs, including that of Philip de Jager at Columbia University, New York (Jul 2018 conference news). 

For a roadmap on how to relate an immune cell’s gene expression to its genetics, researchers can look to the work of Vijayanand and colleagues. They have assembled a compendium of how gene variants affect gene expression in human immune cells. This is necessary because generally in GWAS, most single-nucleotide polymorphisms (SNPs) are not in protein coding but in regulatory regions that dial up or down the expression of neighboring genes. How to define which genes are targets of a risk SNP, and in which cell type? The La Jolla researchers sequenced the transcriptomes of 1,544 purified cell populations isolated from the blood of 91 healthy people, covering the gamut of immune cell types. Then they identified common variants associated with expression of 12,254 unique genes in the cells. Unlike the lack of effect sex had on microglial genes in mice, it did have dramatic effects on gene expression in human immune cells. A central database lets scientists enter in their “favorite” SNPs and see how that variant affects gene expression in specific immune cells.

“This is a stunning achievement, combining genome-wide SNP analysis with RNA sequencing in 13 human immune cell types, primarily lymphocytes,” wrote Richard Ransohoff, Third Rock Ventures, Boston, to Alzforum. “The present resource will be of enormous value to the community of researchers interested in how the immune system affects disease, and additionally provides a useful game plan for extending this form of study to other tissues of interest,” he wrote.

For human microglia, research groups are already collaborating to amass sufficient numbers of genotyped samples for a similar analysis, Ransohoff told Alzforum. He recommends comparing microglia with other brain cells as well as with other tissue macrophages (full comment below).

A new study from John Hardy’s lab at University College London illustrates how knowing more about microglial phenotypes at the transcriptional level reveals new insight into how genes influence AD risk. GWAS for late-onset AD uncovered a roster of largely microglial genes that harbor risk alleles. Earlier, Hardy’s lab found that several of these genes get coordinately upregulated in response to amyloid in mouse models of AD (Jan 2015 news). In that analysis, the group had measured gene expression genes using microarrays, but this limited them to the set of genes that are on the arrays.

In the new study, first author Dervis Salih aimed to expand the network by sequencing microglial mRNA. He identified147 mouse genes whose expression rises in unison in microglia that respond to amyloid. The network centers on the microglial receptor TREM2, includes mouse orthologs of six human GWAS hits, and resembles a human network identified separately from RNA sequencing of pathology samples. The new work ties in five additional genes with SNPs associated with AD—OAS1, CXCL10, LAPTM5, ITGAM, and LILRB4—that did not attain genome-wide significance in previous GWAS. Importantly, like the six already known genes—HLA-DOB, TREM2, SPI1, MS4A6A, ABI3, and CD33—the five new ones respond to the beginning stages of amyloid deposition, increasing their expression in young APP/PSEN1 mice. The network containing all 11 genes is different from a network associated with tau pathology in P301L mice, suggesting it reflects a relatively amyloid-specific response, the authors write.

“This is a clever approach, and a good way to use the data from GWAS to focus in on genes of interest,” said Jean-Charles Lambert, Pasteur Institute, Lille, France. “However, we have to keep in mind these genes have not crossed the statistical threshold to genome-wide significance. They are not yet ‘genes associated with AD risk.’”

Even so, human genetic variation mapping to this expanded network once again singles out innate immunity as a critical factor in AD. To dig deeper into how genetic polymorphisms affect immune function, many groups, including his, are now using cells derived from fibroblasts of genetically defined people, often using polygenic risk scores to choose their subjects, Hardy wrote to Alzforum.

Compared with amyloid, much less is known about the immune response to tau, or the role microglia play in it. The new work from Zheng’s lab implicates the complement pathway in tau’s neurotoxicity. First author Alexandra Litvinchuk documents that expression of the complement protein C3 and its receptor C3aR1 accumulates in the hippocampi of people with AD, and expression there correlates with cognitive decline and the Braak stage of tau pathology. She also sees higher cortical C3 and C3aR1 expression in the tauopathies corticobasal degeneration, Pick’s disease, and progressive supranuclear palsy than in healthy controls.

The Texas investigators see the same trend in mice. In P301S tau transgenic mice, C3, and C3aR1 mRNA and protein rise with age, and C3 is expressed in neurotoxic astrocytes, as previously shown in humans (Liddelow et al., 2017). Surprisingly, deleting the microglial C3aR1 receptor abolished astrocyte C3 expression, lessened tau pathology, and normalized hippocampal-dependent memory behavior. Without this complement receptor, hippocampal synapses were structurally and functionally more intact, and neurodegeneration was milder than in the mice’s parental P301S strain.

Gobbling Synapses. Microglia from PS19 mutant tau mice (green three-dimensional reconstruction) engulf synapses, as indicated by synaptophysin (Syp). Knocking out complement 3a receptor (C3aR) suppresses their appetite. [Courtesy of Litvinchuk et al., Neuron 2018.]

What was happening at the genetic level? Expression profiling of unfractionated hippocampal tissue identified nearly 2,000 genes that were differentially expressed in old P301S mice compared with wild-type. In particular, many genes that function in innate immune response and inflammatory pathways were up, but went back to normal in the P301S/C3aR knockout. Moreover, upregulation of some of these genes, including AD risk genes TREM2, SPI1 and MS4A6A, was associated with C3aR1 expression in human AD expression data sets, suggesting this mouse pathway recapitulates an aspect of human disease. In total, 15 C3aR1-coexpressed genes have been linked to late-onset AD in GWAS studies.

The investigators also examined cell-type-specific gene expression in purified hippocampal astrocytes and microglia. They found elevated expression of markers for the DAM microglia phenotype, and neurotoxic A1 astrocytes in these mice. Deleting C3aR1 diminished both. “That highlights an important feedback loop between disease-responsive microglia and astrocytes that may spiral out of control in the early stages of disease,” said Shane Liddelow, New York University. The work validates earlier findings from Liddelow and others on C3-expressing, neurotoxic A1 astrocytes in AD and in rodent tau models (Jan 2017 news; Sep 2017 news), he wrote to Alzforum (full comment below).

The expression data focused Zheng’s attention on the STAT3 transcription factor as an effector of C3aR receptor signaling. That was an unexpected finding, she said, because STAT3 is normally activated by cytokines. However, treating the P301S mice with a brain-penetrant STAT3 inhibitor three times a week for two months reduced tau pathology, reactive astrocytes, microgliosis and brain inflammation, similar to the effect of genetic C3aR knockout.

Zheng’s results dovetail with recent work from Morgan Sheng’s lab at Genentech, South San Francisco, which implicates complement 1q in synapse loss triggered by tau. In this work, presented at Keystone and published online November 1 in Neuron, unbiased proteomics of synaptic proteins turned up an abundance of complement C1q in synapses of tau-P301S transgenic mice, and also in AD patients (Jul 2018 conference news; Dejanovic et al., 2018). The presence of complement triggered subsequent destruction of synapses by microglia, and an C1q-blocking antibody prevented synapse loss in the mice.

For her part, Zheng also caught microglia gobbling synaptic proteins in their mouse model, which was eased by C3aR knockout. “Both papers together point out a robust and prominent role of complement in driving tau pathology,” she told Alzforum. In that vein, at least one company has started human testing of anti-C1q antibodies for neurodegeneration. Annexon Biosciences, South San Francisco, recently completed a Phase 1 in healthy volunteers.—Pat McCaffrey


  1. This is a very exciting piece of research. I’ve eagerly awaited these data sets. A few points that make this work so novel, and will ensure that it is cited widely in the coming years:

    1. Single-cell microglia analysis—obviously interesting and follows nicely from the DAM paper from Keren-Shaul et al., 2017.
    2. Age-related changes in ssMicro transcripts—this highlights different populations of microglia that, one can now hypothesize, play specific and different roles at different stages of development. The addition of the senescent age (over 500 days) pairs nicely with the aging astrocyte transcriptomes published earlier in the year by Clarke et al., 2018, and Boisvert et al., 2018
    3. Sex-specific ssMicro transcripts. This is going to be extremely helpful for investigations into microglia functions going forward. We obviously know there are differences in female/male responses and disease progression, not just for AD, but for other chronic neurodegenerative diseases too, so having these baseline “control” data is key to making new hypotheses moving forward.
    4. Disease-response ssMicro transcriptomes. The capstone to this work was the response of the microglia to the mouse demyelinating EAE model. The most exciting piece for me! This really put to bed one of the most intriguing pieces of data in our original Nature paper (Liddelow et al., 2017) that Ben Barres and I were unable to make sense of. That is, when we used the colony-stimulating factor 1 receptor inhibitor drug PLX3397 to ablate microglia, we were able to remove 99 percent of microglia as per FACS, but the astrocyte response to inflammatory insult (modeled with LPS injection) was maintained. We thought that perhaps the PLX drug did not kill microglia but might instead simply have downregulated microglia-specific markers we were using for FACS/IF, or that the small subset of microglia that remained were the ones that would have mounted a response to the inflammation anyway. These new data show that the latter is likely the cause, i.e., if you look for TNFα, Il1α, and C1q, there is only a small population of microglia in the disease state that expresses all three.

    Note that the authors provide an easily searchable online portal. This means I can check genes easily on the fly while at meetings, or while out running and thinking of an idea. The portal also couples nicely with the earlier datasets published by the McCarrol and Macosko labs, e.g.,


    . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

    . Normal aging induces A1-like astrocyte reactivity. Proc Natl Acad Sci U S A. 2018 Feb 20;115(8):E1896-E1905. Epub 2018 Feb 7 PubMed.

    . The Aging Astrocyte Transcriptome from Multiple Regions of the Mouse Brain. Cell Rep. 2018 Jan 2;22(1):269-285. PubMed.

    . Neurotoxic reactive astrocytes are induced by activated microglia. Nature. 2017 Jan 26;541(7638):481-487. Epub 2017 Jan 18 PubMed.

    View all comments by Shane Liddelow
  2. The Schmiedel et al. paper offers similar highlights to those found in Hammond et al.

    1. Sex differences.
    2. Disease differences, though this time done by comparing cell-type specific transcriptome data sets with already-available GWAS data sets.
    3. Done in human cells. It would have been nice to see a comparison with the “same” cell types from rodents, to see how well individual immune cell transcriptomes are preserved across species so that we can start to make some decisions about which cells are appropriate models for human.

    The website accompanying Schmiedel et al. is just beautiful. The sheer amount of data on it is a dream, as is the easy-to-manipulate visualization, e.g., changing FC cutoff to determine above/below expression of a gene in multiple cell types. Paired with the Hammond microglia paper, this manuscript provides a key pair of searchable data that the field has been lacking. (I would also add the manuscript by Tristan Qingyun Li/Ben Barres with additional single cell transcriptomics. It’s available on bioRχiv but accepted and to be out soon).


    . Single-Cell RNA Sequencing of Microglia throughout the Mouse Lifespan and in the Injured Brain Reveals Complex Cell-State Changes. Immunity. 2019 Jan 15;50(1):253-271.e6. Epub 2018 Nov 21 PubMed.

    . Developmental Heterogeneity of Microglia and Brain Myeloid Cells Revealed by Deep Single-Cell RNA Sequencing. Neuron. 2019 Jan 16;101(2):207-223.e10. Epub 2018 Dec 31 PubMed.

    View all comments by Shane Liddelow
  3. I have an academic soft spot for this study, as it perfectly validates some of our original findings on neurotoxic A1 reactive astrocytes in human AD and rodent tau models (Liddelow et al., 2018; Shi et al., 2018). In extensive work, Hui Zheng’s lab highlights that astrocyte C3 reactivity is paired with, and increasing with, Braak staging in patients. They also beautifully show that A1 C3 reactivity in GFAP+ astrocytes is present in PS19 mice, and that this reactivity is removed following ablation of the C3aR on microglia. This underscores an important feedback loop between disease-responsive microglia and astrocytes, which may be setting up a cascade in early disease stages.

    Interestingly, PS19 mice showed polarization to an A1-specific phenotype, not just upregulation of GFAP or broad upregulation of both A1 and A2 transcripts. Moreover, ablation of C3aR reverted DAM-activated microglia, which in response blocked only the A1 response but did not drive alternative activation states. This suggests a highly specific response to this genetic approach.  

    Key points in Litvinchuk et al.:

    1. Braak pairs with astrocyte activation in humans: higher Braak stage means more A1 neurotoxic reactive astrocytes.
    2. PS19 mice adequately recapitulate the DAM microglia phenomenon, and recapitulate the A1 neurotoxic reactive astrocyte phenotype seen in human patients.
    3. Ablating C3aR in microglia stops astrocyte activation, likely blocking a feedback loop that spirals out of control during disease progression.

    Moving forward, it will be interesting to see if this crosstalk initiates AD pathology/degeneration, or is simply a response to it. With the data sets provided by Hammond et al. and Schmiedel et al., we now have good baselines to start investigating this in both rodents and humans.


    . Neurotoxic reactive astrocytes are induced by activated microglia. Nature. 2017 Jan 26;541(7638):481-487. Epub 2017 Jan 18 PubMed.

    . ApoE4 markedly exacerbates tau-mediated neurodegeneration in a mouse model of tauopathy. Nature. 2017 Sep 28;549(7673):523-527. Epub 2017 Sep 20 PubMed.

    View all comments by Shane Liddelow
  4. Microglial scRNA-Seq was performed across the mouse life span from E14.5 through P540. After cluster analysis, care was taken to specify at what age, and where, in the brain varied clusters were detected. It was observed that maximal microglial heterogeneity occurred during development, significantly resolved during healthy young adulthood, and then selectively and differentially re-emerged during either aging or focal injury. Particularly compelling observations included a select microglial phenotype associated with presumptive white-matter tracts before myelination, a finding of special interest given recent reports from Marco Prinz and Xinhoa Piao’s labs that microglia provide signals to promote myelination (Hagemeyer et al., 2017; Giera et al., 2018). 

    Another fascinating cluster was seen only at E14.5 and exhibited a phenotype consistent with that of highly proliferative cells, including elevated expression of fatty-acid-binding proteins and glycolytic machinery genes. The investigators' dedication in spatiotemporal localization of cells expressing transcriptomic phenotypes will provide both information and a strategic template for conducting such studies. In this analysis, the absence of sexual dimorphism was notable, consistent with previous reports that microglial sexual dimorphism is commonly elicited by stress conditions such as germ-free state. 

    Data from this paper will constitute a fascinating comparison with those coming from Olah, Phil DeJager et al., currently on bioRχiv. They isolated and sequenced more than 15,000 microglia from human samples, both autopsy and biopsy, from young and aged individuals. Many overall technical conclusions were shared between the two groups. In particular, scRNA-Seq represents an essential step for clarifying the population structures and transcriptomic phenotypes of microglia. DeJager compared their data with those they'd previously obtained by co-expression module analysis of cortical tissue from autopsy samples. They concluded that the scRNA-Seq data set was essential to developing robust insights into microglial phenotypes and heterogeneity in human brain. Of particular interest, Olah/DeJager evaluated four transcriptomic clusters found only in samples from brain donors more than 90 years of age. These transcriptomes bore resemblance to that described by Mathys et al., 2018, in a mouse model of neurodegeneration.

    Neither Hammond et al. nor Olah et al. yielded substantial support for the existence of a “DAM” transcriptome (Keren-Shaul et al., 2017). In particular, Hammond found genes diagnostic of the DAM transcriptome distributed among a number of distinct transcriptomic phenotypes (characterized by varying subsets of non-DAM genes). Olah did not observe clustering of transcriptomic states to be associated with a coherent co-expression pattern of DAM genes. Selection of scRNA-Seq platform and mouse disease model, as well as human versus mouse species variation and preferred analytic scheme, doubtless contributed to these differing observations among the reports.


    . Microglia contribute to normal myelinogenesis and to oligodendrocyte progenitor maintenance during adulthood. Acta Neuropathol. 2017 Sep;134(3):441-458. Epub 2017 Jul 6 PubMed.

    . Microglial transglutaminase-2 drives myelination and myelin repair via GPR56/ADGRG1 in oligodendrocyte precursor cells. Elife. 2018 May 29;7 PubMed.

    . Temporal Tracking of Microglia Activation in Neurodegeneration at Single-Cell Resolution. Cell Rep. 2017 Oct 10;21(2):366-380. PubMed.

    . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

    View all comments by Richard Ransohoff
  5. Stunning achievement, combining genome-wide SNP analysis with RNA-Seq—protein-coding; pseudogenes; lncRNAs; sncRNAs—in 13 human immune cell types, primarily lymphocytes. Once cis eQTL, i.e., SNPs that affected expression of a gene on the same chromosome, were identified, abundant salient, non-trivial, unexpected observations were made. Of all genes expressed across this panel of immune cells, more than half exhibited eQTLs, suggesting a strong effect of germline genetics on human adaptive immune responses. Of genes showing cis eQTL, 40 percent demonstrated effect in only one cell type, and the great majority in only one to three cell types. Importantly, most of these genes were expressed in multiple cell types so cis eQTL occurred due to an epigenomic landscape present in one, two or a few cell types. Further, it was common for cis eQTL to affect expression in a cell that expressed the regulated gene at lower levels than a cell not showing eQTL. Some eQTL were only apparent after cell activation, in this case polyclonal activation of naive CD4+ or CD8+ T cells.

    SNPs that appeared in GWAS analysis of the genetic architecture of human disease, primarily cancer and autoimmune conditions, were evaluated and shown to be nearly 20 percent of all SNPs associated with eQTL. Main emphasis was on T cells and their subtypes. Informative examples were provided: The LACC gene is known to harbor a risk allele for inflammatory bowel disease (IBD), and its strong expression in myeloid cells led to the interpretation that myeloid cell LACC mediates genetic risk for IBD. However, the eQTL analysis identified the effect only in T cells, which may refocus research on this gene-disease connection. For Crohn's disease, a substantial fraction of eQTL affected only one cell type, raising the question whether genetic subtypes of CD could be identified. 

    The sine qua non for cell-specific eQTL identification is sufficient numbers of genotyped samples. Efforts are underway to obtain such data resources, which require collaboration among groups with experience obtaining and analyzing purified human brain cell populations. Implications for brain eQTL in microglia are several: Given that genes harboring disease SNPs are expressed in multiple cells types, e.g. BIN1 in both neurons and microglia, or GRP56 in microglia, astrocytes, and oligodendrocyte lineage, microglia should be compared with other brain cells, as well as with other tissue macrophages.

    Effects of biological sex on expression of cell-type-specific genes in immune cells were impressive. Of all expressed genes in immune cells, 1,875 showed sex bias. More than 90 percent were on autosomes; they were enriched in protein-coding genes and nearly always showed sex bias in one or two cell types, despite expression across multiple cell types.

    The present resource will be of enormous value to the community of researchers interested in how the immune system affects disease. It also provides a useful game plan for extending this form of study to other tissues of interest. One would like to see experiments in which monocytes (classical and non-classical) are subjected to stimulation to elicit additional activation-dependent eQTL. Cells studied here were positively selected, which can induce activation. Some data from this study could also be replicated in negatively selected cells as well.

    View all comments by Richard Ransohoff

Make a Comment

To make a comment you must login or register.


News Citations

  1. TREM2: Diehard Microglial Supporter, Consequences Be DAMed
  2. MS4A Alzheimer’s Risk Gene Linked to TREM2 Signaling
  3. Hot DAM: Specific Microglia Engulf Plaques
  4. A Delicate Frontier: Human Microglia Focus of Attention at Keystone
  5. Network Analysis Points to Distinct Effects of Amyloid, Tau
  6. Microglia Give Astrocytes License to Kill
  7. ApoE4 Makes All Things Tau Worse, From Beginning to End
  8. Synaptic Tau Clangs the Dinner Bell for Hungry Microglia

Research Models Citations

  1. Tau P301S (Line PS19)

Paper Citations

  1. . Genetic variability in response to amyloid beta deposition influences Alzheimer's disease risk. Brain Commun. 2019;1(1):fcz022. Epub 2019 Oct 10 PubMed.
  2. . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.
  3. . Temporal Tracking of Microglia Activation in Neurodegeneration at Single-Cell Resolution. Cell Rep. 2017 Oct 10;21(2):366-380. PubMed.
  4. . Microglia in Alzheimer's disease. J Cell Biol. 2018 Feb 5;217(2):459-472. Epub 2017 Dec 1 PubMed.
  5. . Neurotoxic reactive astrocytes are induced by activated microglia. Nature. 2017 Jan 26;541(7638):481-487. Epub 2017 Jan 18 PubMed.
  6. . Changes in the Synaptic Proteome in Tauopathy and Rescue of Tau-Induced Synapse Loss by C1q Antibodies. Neuron. 2018 Dec 19;100(6):1322-1336.e7. Epub 2018 Nov 1 PubMed.

External Citations

  1. open, searchable resource
  2. central database
  3. Phase 1

Further Reading

Primary Papers

  1. . Single-Cell RNA Sequencing of Microglia throughout the Mouse Lifespan and in the Injured Brain Reveals Complex Cell-State Changes. Immunity. 2019 Jan 15;50(1):253-271.e6. Epub 2018 Nov 21 PubMed.
  2. . Impact of Genetic Polymorphisms on Human Immune Cell Gene Expression. Cell. 2018 Nov 29;175(6):1701-1715.e16. Epub 2018 Nov 15 PubMed.
  3. . Genetic variability in response to amyloid beta deposition influences Alzheimer's disease risk. Brain Commun. 2019;1(1):fcz022. Epub 2019 Oct 10 PubMed.
  4. . Complement C3aR Inactivation Attenuates Tau Pathology and Reverses an Immune Network Deregulated in Tauopathy Models and Alzheimer's Disease. Neuron. 2018 Dec 19;100(6):1337-1353.e5. Epub 2018 Nov 8 PubMed.