Microglia were first characterized by del Rio Hortega about 100 years ago but our understanding of these cells has only gained traction in the last 20 years. We now recognize that microglia are involved in a plethora of activities including circuitry refinement, neuronal and glial trophic support, cell number modulation, angiogenesis and immune surveillance. Specific to immune surveillance, microglia detect threats which then drive their transformation from ramified to amoeboid cells. This morphological transition is accompanied by changes in cytokine and chemokine expression, which are far less conserved than morphology. To simplify discussion of these expression changes, nomenclature ascribed to states of macrophage activation, known as Macrophage 1 (“M1”; classic) and Macrophage 2 (“M2”; alternative), have been assigned to microglia. However, such a classification for microglia is an oversimplification that fails to accurately represent the array of cellular phenotypes. Additionally, multiple subclasses of microglia have now been described that do not belong to the “M1/M2” classification. Here, we provide a brief review outlining the prominent subclasses of microglia that have been described recently. Additionally, we present novel NanoString data demonstrating distinct microglial phenotypes from three commonly used central nervous system inflammation murine models to study microglial response and conclude with an introduction of recent RNA sequencing studies. In turn, this may not only facilitate a more appropriate naming scheme for these enigmatic cells, but more importantly, provide a framework for generating microglial expression “fingerprints” that may assist in the development of novel therapies by targeting disease-specific microglial subtypes.
Microglia, the resident immune cells of the central nervous system (CNS), were first characterized 100 years ago by Pio del Rio Hortega (reviewed in[
At present, it is unclear how microglia are capable of mediating a wide range of activities that, in some cases, are seemingly in contrast to each other. For example, during development, microglia regulate neuronal numbers by both driving cell death[
In addition to providing a brief review of several parameters and subclasses that define microglial heterogeneity, we also present novel RNA expression profile data that are consistent with the development of distinct microglial phenotypes as a consequence of distinct inflammatory environments. As presented in more detail below, we isolated cortical microglia from mice in three commonly used models to study various aspects of multiple sclerosis - cuprizone, lipopolysaccharide (LPS) and experimental autoimmune encephalomyelitis (EAE). Orally administered cuprizone results in CNS demyelination secondary to oligodendrocyte death. Intraperitoneal injection of the endotoxin LPS mediates a peripheral immune response that results in widespread CNS neuroinflammation. Similarly, EAE is induced by a peripheral injection of a bacterial exotoxin that is accompanied by Complete Freund’s Adjuvant and a myelin antigen resulting in breakdown of the blood brain barrier. Although microglia from all three models presented pro-inflammatory profiles, the microglia from each expressed a unique set of factors suggesting environmental-specific responses. Although these observations are consistent with environmental cues driving heterogeneity, it remains possible, and perhaps likely, that microglia also represent intrinsically distinct populations.
Currently, a prevailing thought is that microglia, which derive from the embryonic yolk sac, develop initially as a single-cell type lineage[
Some of the earliest evidence of heterogeneity within the microglial population was presented by Lawson
Variations in cell density and transcription profiles are not limited to regional differences as similar distinctions have also been reported between microglia from male and female mice. Male mice present with more microglia in the cortex, hippocampus, dentate gyrus, and amygdala in early postnatal brains. As the mice mature, these densities flip with female mice presenting with a greater cell density in these regions[
The mechanisms responsible for these sex differences are not known and transcriptomic studies comparing male and female microglia reveal expression differences in both the healthy and perturbed states[
Perhaps the most recognized heterogenic aspect of microglia is their morphology. Two main classes have been identified - amoeboid-like, with few processes; and ramified, with numerous thin, highly-branched processes. Following initial colonization of the embryonic CNS, the majority of microglia present with an amoeboid-like morphology[
Recently, a new class of microglia was identified based on morphology. These microglia are “dark” based on their electron dense cytoplasm and are observed in non-homeostatic conditions[
Baalman
Upon activation following a controlled cortical impact (CCI)-induced traumatic brain injury, the association between CNS microglia and the AIS is lost, consistent with the regulation of microglial function and response by the local environment. Interestingly, our laboratory has also reported contact between microglia and the AIS[
Frequency of microglial-AIS contact is not altered in LPS-induced neuroinflammation. Female c57black6 mice were given a single intraperitoneal injection of LPS (5 mg/kg) or vehicle (0.9% saline, 10 mL/kg). Confocal z-stacks spanning an optical thickness of 25 μm, using a pinhole of 1 Airy disc unit and Nyquist sampling (optical slice thickness, 0.48 μm), were collected from neocortical layer V for each of six sections (spanning 1.1 mm anterior to the bregma to 2.5 mm posterior to the bregma) per mouse, resulting in 12 images per animal (
Herein, we have reviewed several subclasses of microglia that have been defined based on morphology; however, it is unclear if these subclasses are truly distinct, or if they are merely the consequence of artificial classifications based on techniques used for identification, and loose criteria for defining subtypes (reviewed[
In an effort to more conclusively characterize microglia and to elucidate their functions, morphologic characterizations have been complemented by molecular classification studies. Initial attempts were based on presumed states of activity based on limited expression profiles. Simply, microglia were classified as either “activated” or “resting” but both terms are misleading. Microglia are never “resting” as we now recognize that they are constantly extending and retracting their processes to survey their surroundings[
Reactive microglia have been further divided into “M1” and “M2” states, referring to the classical (pro-inflammatory) and alternative (resolving/anti-inflammatory) phenotypes based on expression profiles. The “M1” and “M2” nomenclature is a naming scheme originally derived from the T cell literature and applied to macrophages based on their state of activation
Another subclass of reactive microglia that is specific to non-homeostatic conditions is known as Disease-Associated Microglia (DAM). First identified in Alzheimer’s disease and amyotrophic lateral sclerosis models[
Although unique to non-homeostatic conditions, the function of DAM is not known. It has been hypothesized that these cells respond to a CNS stress signaling system that is akin to the peripheral immune system’s pathogen- and damage-associated stress signals (PAMPs and DAMPs)[
Following injury or disease, reactive microglia are rapidly recruited to sites of damage where they phagocytose debris and dying cells, consistent with the described functions of DAM. Likewise, AXIS microglia may also be recruited to sites of damage following injury or disease[
In addition to regulating neuronal function through secreted factors, microglia also regulate neurons through physical contact[
Recently, we analyzed CNS pathology in three models of neuroinflammation. In all three models, microglia presented with reactive phenotypes and these cells maintained, or even increased, contact with the AIS. However, in two of the models, the AISs were disrupted and in one, the AISs were preserved. Since AIS integrity temporally correlated with the presence of reactive microglia and contact was at least maintained in all three models, we proposed that differential AIS integrity was consequential to the heterogeneity among the reactive microglia from all three models.
For our studies, we exploited the immune-mediated inflammatory models of EAE[
We utilized these three models to further investigate microglial heterogeneity. AIS disruption only occurred in the LPS and EAE models, while microglial-AIS contact was abundant in all three models. Thus, while microglial reactivity and contact increased prior to and was coincident with disruption in EAE, contact alone did not disrupt AIS integrity[
Microglia with reactive morphologies predominate in the cortex of all three models[
Microglia differentially express inflammation-associated genes in three neuroinflammatory models that demonstrate robust microglial reactivity. A: representative images of surveying microglia from the cortex of naïve mice and reactive microglia from LPS 24 h, 3 week Cuprizone, and EAE Early 3 & 4 mice; B-D: analysis of NanoString data of 248 differentially expressed inflammation-associated genes in CD11b+ cells. Background subtraction was performed using the maximum value across samples of the negative controls and data normalization was performed using the geometric mean expression of six internal reference genes (
Recently, several single cell RNA sequencing studies have begun to more clearly define subsets of microglia in the developing,mature and healthy, and pathologic CNS[
Using fluorescent assisted cell sorting gated by CD11b, CD45 and Cx3Cr1, Hammond
Although described 100 years ago, we are only just beginning to put together the various pieces of the microglial puzzle. We now recognize their involvement in establishing and maintaining a homeostatic CNS environment through trophic support and pruning of both neuronal and glial populations, modulating CNS wiring and circuitry, and facilitating axonal organization and outgrowth, myelin formation, and immunosurveillance in the healthy brain. Moreover, we are also beginning to appreciate their critical roles in disease, potentially both as CNS protectors by recognizing and removing infected, dying and dead cells, and also as CNS villains secondary to hyperactivation or dysregulation. We are also beginning to recognize that microglia may present as functionally distinct subclasses, which provides an explanation as to how a single lineage cell type can manifest into a plethora of diverse roles. However, it remains to be determined if distinct subclasses of microglia truly exist, or if microglia exist on a spectrum where they have the capacity to take on a multitude of identities depending on their environment. To address this issue, consistent approaches in cell isolation and analysis should be established and implemented. Additionally, as presented by other authors[
Made substantial contributions to experimental conception and design and manuscript preparation: Dupree JL
Made substantial contributions to experimental conception and design, in technical support, mRNA data analysis and interpretation and manuscript preparation: Benusa SD
Made substantial contributions to microglia-AIS contact analysis and interpretation: George NM
NanoString raw data files are provided in Supplementary Material.
This work was supported by grants from the National Institute of Health [“Microglial neurofascin: a novel mediator of microglial/axon initial segment interactions?” R21NS1016515; (JLD)] and the Veterans Affairs [“Attenuating microglial-dependent axonal pathology in EAE” (No. BX002565 (JLD)]. Microscopy was performed at the VCU Massey Cancer Center Microscopy Core Facility and supported, in part, with funding from NIH-NCI Cancer Center Support Grant P30 CA016059.
All authors declared that there are no conflicts of interest.
Not applicable.
Not applicable.
© The Author(s) 2020.