Microglia are specialized tissue-resident macrophages of the central nervous system, which comprises the brain and spinal cord. They combine functions in innate immunity with continuous monitoring and maintenance of nervous tissue. Their branching processes contact nerve cells, blood vessels, and other supporting cells, while their molecular machinery detects infection, injury, altered cellular activity, and material requiring removal. Microglia are therefore not simply emergency responders: they also participate in the development, organization, and long-term maintenance of neural circuits.
Unlike most other glial cells, microglia originate predominantly from embryonic blood-cell precursors rather than the neural tissue that produces neurons, astrocytes, and oligodendrocytes. They establish a largely self-renewing population early in development. Their behavior depends strongly on age, anatomical location, and local conditions, making the traditional distinction between “resting” and “activated” microglia inadequate. Research on their origins, signaling pathways, and diverse states has linked them to normal neural development, responses to infection and injury, and the biology of several neurological diseases.
Identification, distribution, and historical discovery
Microglia are distributed throughout the brain and spinal cord, including gray matter, white matter, and specialized regions such as the retina. Estimates commonly place them at roughly 5–15% of brain cells, although the proportion varies with species, region, age, and counting method. They usually occupy partly distinct territories rather than forming a uniformly overlapping network. Their cell bodies are relatively small, but their branching processes extend through a much larger volume of surrounding tissue.
In healthy adult tissue, many microglia have a ramified appearance: a small soma gives rise to fine, repeatedly branching processes. During development or after injury, they may become more rounded, enlarge their cell bodies, thicken their processes, or acquire an amoeboid form associated with migration and engulfment. These appearances are useful descriptions, not definitive functional categories. A highly branched cell can be responding actively to local signals, and a rounded cell is not necessarily causing inflammation.
The Spanish neuroscientist Pío del Río-Hortega established microglia as a distinct cellular population in work published in 1919, using silver carbonate staining methods that revealed cells poorly distinguished by earlier techniques. His observations separated microglia from other non-neuronal elements and supported a mesodermal rather than neuroectodermal origin. This distinction was important because earlier classifications grouped several poorly characterized supporting cells together. Later electron microscopy, immunology, and developmental lineage tracing clarified microglial structure and ancestry.
Modern identification generally combines anatomy with molecular markers. IBA1, encoded by AIF1, is widely used to visualize microglial cell bodies and processes, but it also labels other macrophages. CD11b and CX3CR1 likewise identify overlapping myeloid populations rather than microglia alone. TMEM119 and P2RY12 are useful markers of many mature microglia, especially under relatively undisturbed conditions, but their abundance can change during development or disease. Consequently, no single marker identifies every microglial state while excluding every related cell.
An important boundary separates parenchymal microglia from macrophages associated with the meninges, perivascular spaces, and choroid plexus. These neighboring populations share immune functions and some markers, but differ in anatomical position, developmental history, and gene regulation. Blood-derived monocytes may also enter the nervous system during disease and become macrophages. Distinguishing resident microglia from these cells is essential when interpreting inflamed tissue, because similar appearance does not establish identical origin or function.
Embryonic origin and lifelong population maintenance
Most microglia arise from early embryonic progenitors associated with the yolk sac, an extraembryonic structure that supports early development and blood-cell production. In mice, lineage-tracing experiments place their ancestors in an early wave of hematopoiesis that precedes the definitive blood-forming system of the adult bone marrow. These precursors enter the developing nervous system during embryogenesis and expand as the tissue grows. Human developmental studies support a similarly early establishment of microglia, although the precise timing and cellular transitions cannot be investigated with the experimental lineage-tracing methods available in mice.
A major advance came in 2010, when genetic fate-mapping experiments demonstrated that adult mouse microglia descend from primitive embryonic myeloid progenitors. Such experiments permanently label a precursor population during a restricted developmental interval and then track its descendants. Subsequent studies using complementary approaches established that, under ordinary conditions, adult microglia are maintained mainly through local survival and cell division, not continual replacement by circulating monocytes.
Several signaling pathways sustain this population. The colony-stimulating factor 1 receptor, CSF1R, is particularly important for microglial development and survival. Its ligands include CSF1 and interleukin-34, whose distributions differ across nervous-system regions and developmental stages. Transcription factors including PU.1 and IRF8 help establish myeloid identity, while signals from the neural environment promote the more specialized microglial program. Population density therefore reflects both developmental history and continuing access to local survival signals.
Experiments that markedly deplete microglia have shown that surviving resident cells can proliferate and repopulate the tissue. This capacity does not mean that replacement is always complete or biologically neutral. The kinetics and properties of repopulating cells depend on the depletion method, duration, age of the animal, and condition of the nervous system. Under exceptional circumstances, including particular transplantation or conditioning procedures, blood-derived cells can establish long-lasting residence in the brain. Such engrafted macrophages may resemble microglia without fully acquiring their molecular identity.
Turnover also differs among species and experimental settings. In humans, retrospective birth-dating studies using atmospheric carbon-14 incorporated into cellular DNA have demonstrated that microglial renewal continues during adult life. However, renewal of a fraction of the population does not imply that all cells have the same lifespan. Local proliferation, cell loss, and clonal expansion can generate mixtures of younger and older cells. Microglia are thus a persistent resident population whose membership changes over time, rather than a fixed set of cells that remains untouched from embryogenesis onward.
Surveillance, motility, and communication with neural tissue
The older term “resting microglia” reflected their relatively inconspicuous appearance in fixed preparations, not physiological inactivity. In 2005, live imaging studies using two-photon microscopy showed that microglial processes continuously extend and retract in the living mouse cortex. Their cell bodies often remain comparatively stationary while the processes sample nearby tissue. This dynamic behavior permits local surveillance without requiring the entire cell to migrate through densely organized neural structures.
Extracellular nucleotides provide important signals. ATP and its breakdown products can appear outside cells during normal signaling, mechanical disturbance, or injury. The microglial receptor P2RY12 responds primarily to ADP and contributes to directed process extension toward sites of tissue damage. Other purinergic receptors have different properties; for example, P2X7 is an ATP-gated ion channel that can participate in inflammatory responses under appropriate conditions. These receptors allow microglia to distinguish and respond to changes in the chemical environment, although a nucleotide signal alone does not specify its cause.
Microglia also receive inhibitory and regulatory signals from surrounding cells. Neuronal CX3CL1, also called fractalkine, interacts with CX3CR1 on microglia. CD200 on neural cells can signal through microglial CD200 receptors. These interactions influence responses to the local environment rather than operating as simple universal off-switches. Their effects depend on developmental stage, brain region, and the type of challenge. Neurotransmission, extracellular ions, and local metabolic conditions further influence microglial behavior.
Contact with a neuron does not necessarily imply engulfment. Microglial processes can transiently appose synapses, dendrites, axons, and neuronal cell bodies. Such contacts have been associated with monitoring neuronal condition and responding to altered activity. Experiments in mice have identified specialized contacts at neuronal somata where microglial processes interact with sites related to neuronal metabolism. The functions and prevalence of individual contact types remain active research subjects, especially when extending observations from experimental animals to human tissue.
The cytoskeleton supplies the machinery for process movement, migration, and engulfment. Actin remodeling changes cell shape, while intracellular signaling coordinates direction and timing. These activities impose metabolic demands that vary with function: a cell surveying intact tissue has different requirements from one proliferating or digesting large amounts of damaged material. Microglial surveillance is therefore an integrated cellular activity linking receptors, energy supply, membrane trafficking, and physical movement, rather than a passive inspection of the extracellular space.
Molecular identity and context-dependent states
Microglial identity is maintained through interactions between inherited developmental programs and signals from the nervous-system environment. Transforming growth factor beta is an important component of this environment and supports expression of many characteristic microglial genes. The transcription factor SALL1 helps preserve a specialized microglial program and restrain inappropriate macrophage-like responses. These regulatory systems operate through networks of gene expression and chromatin organization, not through a single defining switch.
Many adult microglia express a recognizable set of genes, including P2RY12, TMEM119, SALL1, and CX3CR1. Nevertheless, the abundance of these transcripts differs among anatomical locations and can fall during disease. A microglial cell that loses a conventional homeostatic marker has not necessarily ceased to be microglial. Conversely, expression of a few such markers by another macrophage population does not prove complete equivalence. Identity, current state, developmental origin, and anatomical position are related but distinct features.
Modern profiling has revealed populations associated with proliferation, interferon responses, lipid handling, antigen presentation, and other activities. Single-cell RNA sequencing measures transcripts in individual cells and can identify groups with shared expression patterns. These groups are often called states or clusters. Their boundaries partly reflect analytical choices, and closely related cells may lie along a continuum rather than belong to discrete natural types. A cluster enriched in one experiment may also contain cells performing several different tasks.
The label “disease-associated microglia” became prominent after a 2017 study identified a characteristic transcriptional program in mouse models of neurodegeneration. This program included changes in genes involved in lipid metabolism, lysosomal activity, and immune signaling, with part of the transition depending on TREM2. Related signatures have since been identified in other settings, but they are not identical across diseases or species. Terms such as disease-associated, activated-response, or lipid-associated microglia describe experimentally observed patterns; none alone establishes whether the cells are protective or harmful.
The older M1/M2 classification, borrowed from simplified macrophage stimulation systems, does not adequately represent this diversity. Cells in living nervous tissue can simultaneously express genes assigned to both categories, and their responses change over time. A more informative description identifies the stimulus, location, timing, molecular features, and measured activity. Sex, age, genetic background, and tissue preparation also influence the observed profile. Molecular classification is most useful when connected to direct evidence of function, rather than treated as a substitute for it.
Development, cell clearance, and synaptic remodeling
During development, the nervous system produces cells and connections in excess of those retained in mature circuits. Microglia help remove cells undergoing apoptosis, a regulated form of cell death, and clear cellular fragments. This clearance limits the persistence of damaged material and can influence the local signaling environment. Engulfment is guided by molecular changes on dying cells, including exposure of phosphatidylserine on the outer surface of the cell membrane, together with receptors and bridging molecules that recognize those changes.
Microglia also interact with developing synapses. Synaptic pruning is the selective elimination or remodeling of connections during circuit refinement. In the developing mouse visual system, experiments have shown that microglia engulf presynaptic material and that the complement system contributes to this process. Complement proteins C1q and C3 can participate in marking neural elements, while microglial complement receptor 3 helps recognize complement-associated targets. Disruption of this pathway can alter the normal refinement of retinal inputs to the thalamus.
These findings do not establish a universal rule that complement labels every weak synapse for destruction. Pruning mechanisms differ among brain regions, developmental periods, and types of connection. Activity-dependent competition, neuronal signaling, astrocytic functions, and extracellular factors all contribute. Moreover, detection of synaptic proteins inside microglia does not by itself show whether a complete functional synapse was removed, whether only part of a terminal was taken up, or whether the material had already detached.
In some experimental settings, microglia remove small portions of axons or synaptic structures without engulfing the entire contacted element, a process described as partial phagocytosis or trogocytosis. Such observations broaden the range of possible interactions beyond whole-synapse elimination. Microglia may also influence the formation and stabilization of connections through secreted factors and direct contact. The effects can depend on the animal’s sensory experience and the maturity of the circuit, linking microglial activity to neuroplasticity without making microglia the sole regulators of plastic change.
Their developmental functions extend beyond synapses. Microglia can influence the survival and positioning of neural precursor cells, support aspects of oligodendrocyte development, and help clear material during the formation of white-matter tracts. These effects are strongly stage-dependent. Removing microglia early in development may alter several interacting processes at once, making later behavioral or anatomical changes difficult to attribute to a single mechanism. Developmental findings therefore require attention to timing, brain region, and the distinction between direct microglial actions and secondary consequences elsewhere in the tissue.
Immune recognition and inflammatory signaling
Microglia detect microorganisms and tissue damage through receptors shared with other innate immune cells. Toll-like receptors recognize selected microbial components and endogenous molecules associated with injury. Other sensors detect nucleic acids, altered membranes, or intracellular stress. Receptor engagement activates signaling networks that can change cell shape, migration, metabolism, and the production of immune mediators. The outcome depends on combinations of signals rather than recognition by one receptor in isolation.
Among the mediators produced are cytokines such as tumor necrosis factor, interleukin-1 beta, and interleukin-6, as well as chemokines that influence cell movement. Microglia can also mount interferon responses during viral infection. These substances act on microglia themselves, neurons, other glia, and vascular cells. Their effects are concentration- and time-dependent: signaling that helps contain infection or coordinate repair can disrupt neural function if excessive, prolonged, or spatially inappropriate.
Some inflammatory pathways involve assembly of an inflammasome, a multiprotein complex that promotes activation of inflammatory enzymes. The NLRP3 inflammasome is especially well studied in microglia. Under appropriate conditions, it activates caspase-1, which processes precursor forms of interleukin-1 beta and interleukin-18. Inflammasome activity can also connect to inflammatory cell-death mechanisms. Detection of inflammasome components, however, does not establish that the full pathway is active; evidence of assembly, enzymatic processing, or functional consequences is needed.
Microglia can increase expression of major histocompatibility complex molecules and other machinery involved in antigen processing. This gives them the capacity to interact with T cells under certain conditions. Their contribution to adaptive immunity is nevertheless distinct from that of professional antigen-presenting cells in lymphoid tissues. Border-associated macrophages, dendritic cells, and infiltrating immune cells may be more important for particular stages of an immune response. The relative contributions depend on the disease and anatomical compartment.
The term neuroinflammation encompasses these and other immune-related processes within nervous tissue, including changes in astrocytes, blood vessels, and recruited leukocytes. It is not synonymous with microglial activation, nor does it specify whether an overall response is beneficial or damaging. During infection, microglial containment and clearance functions may protect tissue, while associated inflammatory products can also injure nearby cells. Interpreting the response requires identifying the initiating event, the cells involved, and the sequence of changes over time.
Responses to injury, debris clearance, and repair
After focal tissue injury, nearby microglial processes can converge rapidly on the damaged site. This early response occurs before substantial recruitment of blood-derived immune cells in many experimental models. With more severe damage, microglia may migrate, proliferate, and transform their morphology. They accumulate near injured neurons, damaged axons, vascular lesions, and regions containing cellular debris. The response evolves over hours to weeks, so observations from a single time point capture only one stage.
A central function is phagocytosis, the engulfment of material into intracellular compartments for processing. Microglia take up dying cells, membrane fragments, protein aggregates, and damaged myelin. Engulfed material is delivered to lysosomes, where enzymes break down proteins, lipids, and other components. Receptors such as TREM2 and members of the TAM receptor family help coordinate recognition and uptake. Successful clearance can reduce the persistence of substances that would otherwise sustain inflammation or obstruct repair.
The consequences of engulfment depend on the target. Removing dead cells is different from eliminating a stressed but viable neuron. In experimental systems, inflammatory conditions can cause living cells to expose signals that favor their uptake, a process sometimes termed phagoptosis. Establishing this mechanism requires showing that preventing engulfment preserves cells that would otherwise survive, rather than merely delaying removal of cells already committed to death. This distinction is important when interpreting images of neuronal material inside microglia.
In white matter, myelin breakdown presents a substantial metabolic challenge because it releases large quantities of lipids, including cholesterol. Microglia must process, store, export, or redistribute these materials. Defects in lipid handling can lead to intracellular accumulation and altered inflammatory responses. Efficient debris removal can support remyelination by creating conditions in which oligodendrocyte precursor cells differentiate and produce new myelin. However, microglia do not themselves form the insulating myelin sheath.
Repair also depends on interactions with astrocytes, vascular cells, oligodendrocyte-lineage cells, and recruited macrophages. Microglia can influence the organization of tissue surrounding an injury and the signals available to surviving cells. In ischemic stroke and traumatic injury models, protective and damaging activities often occur concurrently or at different stages. Consequently, reducing a broad measure of microglial activation does not necessarily improve repair: it may suppress harmful inflammation while also impairing clearance, containment, or support of surviving tissue. Mechanistic interpretation requires separating these activities rather than treating the response as a single process.
Genetic evidence and neurological disease
Human genetics has provided strong evidence that microglial pathways contribute to neurological disease. In Alzheimer’s disease, several risk-associated loci involve genes expressed in microglia or related myeloid cells, including TREM2, CD33, INPP5D, and PLCG2. These associations connect disease susceptibility to immune signaling, lipid handling, and cellular responses to damage. They do not imply that every risk variant acts exclusively in microglia or that all affected pathways have the same effect throughout disease progression.
Rare coding variants in TREM2, including R47H, were linked to increased Alzheimer’s disease risk in studies published in 2013. TREM2 is a cell-surface receptor that signals through an adaptor containing TYROBP, also called DAP12. It contributes to microglial survival, responses to lipids, and behavior around amyloid deposits. In mouse models, disruption of TREM2 can impair the organization of microglia around plaques and alter plaque structure and associated neuronal injury. Effects on total amyloid burden vary with model and disease stage.
Microglia cluster around amyloid plaques and respond to tau-associated pathology, but their role is not simply the removal of toxic deposits. They can influence plaque compaction, inflammatory signaling, synaptic changes, and tissue responses to damaged neurons. Human disease unfolds over decades and includes interacting vascular, metabolic, and genetic factors. Results from animals engineered to develop selected pathological features therefore establish mechanisms under defined conditions rather than reproducing the full human disorder.
Other inherited conditions demonstrate the importance of particular pathways more directly. Pathogenic variants in CSF1R cause a hereditary leukoencephalopathy often termed CSF1R-related leukoencephalopathy or adult-onset leukoencephalopathy with axonal spheroids and pigmented glia. Its characteristic abnormalities involve white-matter degeneration and disturbances of the brain’s myeloid-cell system. Biallelic pathogenic variants affecting TREM2 or TYROBP can cause Nasu–Hakola disease, a rare disorder involving bone cysts and early-onset dementia. These conditions illustrate that impaired microglial signaling can be associated with severe neurological dysfunction.
In multiple sclerosis, microglia participate alongside infiltrating immune cells in inflammatory demyelination and in responses that may support myelin repair. Microglial changes also occur in Parkinson’s disease and amyotrophic lateral sclerosis, although their contributions vary across genetic forms, models, and disease stages. Across these disorders, the presence of reactive microglia is not by itself proof that they initiated the disease. Genetic evidence, longitudinal observations, and experiments that alter specific pathways are needed to distinguish initiating mechanisms from responses to ongoing tissue damage.
Aging, regional variation, and influences from outside the brain
Microglia change throughout life. The proliferative and phagocytic states common during early development differ from those typical of healthy adulthood. With aging, some populations show increased expression of inflammatory, lipid-processing, or stress-response genes, while others retain many features of younger cells. Morphological changes can include shortened or fragmented processes, but appearances vary across regions and individuals. Chronological age therefore does not define a single universal microglial state.
The term “priming” describes situations in which a prior condition makes microglia respond differently to a later challenge. In aging animal models, a peripheral inflammatory stimulus can produce a larger or more prolonged central response than in younger animals. This does not mean that every aged microglial cell is continuously inflamed. It indicates an altered response threshold or regulatory context. Long-lived changes may involve chromatin regulation, accumulated cellular material, altered metabolism, and continuing signals from aged neurons, glia, and blood vessels.
Regional differences are also substantial. White-matter microglia encounter myelin-rich environments, whereas cells in gray matter interact with dense synaptic networks. The cerebellum, hippocampus, cortex, and spinal cord differ in developmental history, local cell composition, and vulnerability to disease. These differences influence microglial density, turnover, and gene expression. A molecular pattern identified in one region should not automatically be treated as representative of the entire central nervous system.
The blood–brain barrier restricts and regulates exchanges between blood and nervous tissue, but does not isolate microglia from the rest of the body. Circulating inflammatory signals can act on vascular and barrier-associated cells, which then relay signals into the brain. Barrier disruption can allow additional plasma components and immune cells to enter. Peripheral infection, metabolic disturbances, and vascular injury can consequently alter microglial behavior through several routes without requiring direct entry of a microorganism into brain tissue.
The intestinal microbiota is another influence demonstrated particularly clearly in experimental animals. Germ-free mice show altered microglial maturation and responses, and microbial metabolites can contribute to communication between the intestine and nervous system. These findings do not establish a single microbiome composition that produces healthy human microglia. Diet, housing, microbial exposure, age, and species all affect results. Sex-related differences likewise depend on developmental stage, brain region, and experimental context; findings in one animal model cannot be assumed to describe every human population.
Experimental methods and therapeutic investigation
Microglial research depends on combining methods that answer different questions. Immunohistochemistry and conventional microscopy reveal location, morphology, and selected proteins in fixed tissue. Electron microscopy resolves intracellular organelles and membrane relationships, including engulfed material. Live two-photon imaging records process movement and interactions over time. Each method has limitations: fixed tissue cannot directly show a sequence of events, while live imaging often samples restricted regions under conditions influenced by surgery, anesthesia, or optical access.
Genetic tools permit more selective experiments in animals. Inducible lineage-tracing systems can label microglia and follow their descendants, while conditional gene deletion tests the contribution of particular pathways. Selectivity must be established for each tool because commonly used promoters can also target peripheral myeloid cells or border-associated macrophages. Drug-induced depletion through CSF1R inhibition is widely used to investigate population functions, but it can affect other cells and disrupt multiple microglial activities at once. A phenotype after depletion does not identify which lost function caused it.
Single-cell and single-nucleus sequencing, chromatin-accessibility assays, and spatial molecular methods have expanded the ability to map microglial diversity. Spatial approaches are especially useful because a cell near an amyloid plaque, damaged vessel, or demyelinated axon may differ from one elsewhere in the same specimen. Tissue dissociation can itself induce stress-related transcripts, while postmortem delay and RNA quality influence human data. Analyses therefore require careful attention to sampling and preparation before apparent differences are interpreted biologically.
Human induced pluripotent stem cells can be differentiated into microglia-like cells for experiments involving patient-associated variants or controlled genetic backgrounds. These cells can be studied alone, with neurons and other glia, in three-dimensional neural cultures, or after transplantation into animal brains. Each setting supplies a different environment. Cultured microglia often lose aspects of the identity found in intact tissue, whereas more complex models introduce additional variables. Comparisons with primary human cells help establish which features a model reproduces.
Molecular imaging provides indirect measures in living people. Positron emission tomography ligands targeting the translocator protein TSPO have been used to study changes associated with neurological disease. However, TSPO is not specific to microglia, and binding is influenced by factors including a common human genetic polymorphism. Therapeutic research is investigating TREM2 signaling, inflammatory pathways, cellular metabolism, and microglial survival or replacement. Such approaches must account for the coexistence of protective and damaging functions, as well as the possibility that the same pathway has different effects before injury, during active disease, and during repair.