Preprints.org logo
Preprint
Review

This version is not peer-reviewed.

Signal Integration and Cellular Decisions in Microglia

Submitted:

07 September 2026

Posted:

09 September 2026

You are already at the latest version

Abstract
Microglia encounter combinations of soluble, target-bound and physical signals whose functional meaning changes with spatial presentation, timing and the state of the receiving cell. Yet most mechanistic studies isolate one receptor or ligand at a time. This review defines direct functional interaction as experimentally demonstrated effect modification between specified inputs in the same functional setting, and distinguishes it from sequential or multicellular relays, context dependence, association and hypothesis. The literature is organized around five linked decisions: positioning and stable contact, target recognition, uptake or restraint, cargo degradation and recovery, and later cellular state or output. Noradrenergic opposition to ATP-directed process extension provides a direct interaction; neuron-microglia-astrocyte and injury-induced ATP-IL-1β circuits provide directional relays. Complement, phosphatidylserine, inhibitory checkpoints, trophic signals, mechanosensing, age, sex and anatomical niche instead show how individual pathways and context shape particular decisions without necessarily demonstrating cue-cue integration. Direct multi-input experiments remain uncommon. Human single-cell and spatial studies can identify candidate states and signalling niches, but observational data cannot by themselves establish ligand access, receptor function, direction or causal interaction. Claims are therefore described as direct interactions, relays, context-dependent effects or associations according to the experiment that supports them. Therapeutically, this matters because a pathway that contributes to pathology in one setting may support surveillance, clearance, vascular stability or repair in another.
Keywords: 
microglia
;  
signal integration
;  
purinergic signalling
;  
phagocytosis
;  
complement
;  
TREM2
;  
inflammasome
;  
cGAS-STING
;  
neurotransmitters
;  
neurovascular unit
Subject: 
Biology and Life Sciences  -   Neuroscience and Neurology

Review Approach and Scope

This is a critical narrative review rather than a systematic review. The literature set was assembled iteratively from primary studies and recent reviews and includes work available through August 2026. No protocol, exhaustive database search or quantitative screening process was used; omission from this review should therefore not be interpreted as evidence that a pathway or study is unimportant.
Primary studies were prioritized when they directly tested signal source, responding cell, event sequence or functional consequence. Particular weight was given to cell-restricted perturbation, temporal interruption or rescue, factorial manipulation of multiple inputs, and experiments that measured the relevant cellular decision rather than a broad state marker.
Throughout the review, the wording follows the experiment. Studies that manipulate specified inputs together can support functional interaction; source-and-receiver perturbations can support a relay; comparisons across states establish context dependence; and co-expression or proximity alone remains associative. These distinctions reflect what was measured rather than the prestige or complexity of the method.

Receptor Pathways and Cellular Decisions

Microglia continually survey neural tissue and contribute to injury sensing, circuit remodelling, target clearance, trophic support and coordination of multicellular responses. In vivo imaging established that apparently ramified microglia are highly dynamic, and developmental studies showed that microglia can alter synapse number and circuit maturation [1,2,3,4,5].
Much of the field is still organized around receptor-centred modules: P2Y12 for nucleotide-guided movement, complement receptors for target recognition, CSF1R for survival, cytokine pathways for inflammatory output, neurotransmitter receptors for neuronal control. These modules are experimentally convenient, but a microglial cell near a plaque, a capillary, a developing synapse or an injury border does not encounter one module at a time. It meets several chemical, target-bound and physical signals across different spatial and temporal scales [6,7,8,9,10].
The history and longevity of the cell may matter, but cell turnover and state persistence are separate questions. Adult mouse microglia are maintained mostly through local self-renewal; steady-state turnover behaves largely stochastically, whereas pathological challenge can drive local clonal expansion. Human cortical microglia renew more slowly, with a reported median rate of about 28% per year and some cells persisting for decades [11,12,13]. These species differences affect the opportunity for long-lived exposure, but they do not by themselves show that a molecular state is equally persistent: long-lived cells can change state rapidly, and population states can be continually re-imposed by the niche.
This review therefore asks a narrower question than which molecules regulate microglia. It asks what evidence shows that multiple inputs, cell state and tissue context combine to change a defined function: where a process moves, whether a contacted structure is retained or removed, whether internalized material is successfully degraded, and how an executed function changes later responsiveness [6,8,10,14,15,16].
This distinction matters because co-expression, proximity and transcriptional association do not establish ligand access, signalling direction or functional interaction. Global knockout, systemic pharmacology and rescue can establish organism-level relevance without locating the decisive pathway to microglia or demonstrating that two inputs interact [17,18,19,20,21].
Cell attribution is especially important at vascular and meningeal interfaces, where perivascular, meningeal and choroid-plexus macrophages overlap with parenchymal microglia in the myeloid markers most commonly used to identify them. Anatomical location, lineage and timing all need to be considered when interpreting vascular or CSF1R perturbations [6,22,23,24].
The sections that follow use the cellular decision sequence as their main spine, then examine state filters, sequential logic, multicellular relays and neuronal activity as cross-cutting cases. First, however, the evidence language used for a claimed interaction needs to be defined.

Evidence for Signal Interaction

Two receptors can be expressed by the same cell, or a ligand-producing cell can sit beside a receptor-expressing microglial cell, without either pathway modifying the same function [17,18,20]. A direct functional interaction requires specified inputs to be tested alone and together against a predefined endpoint. A statistical interaction demonstrates effect modification on the chosen response scale; additional pathway-specific perturbation is needed to establish mechanistic convergence. Noradrenergic opposition to ATP-directed process extension meets the functional standard [25,26]. Dopamine-receptor signalling and P2Y12 are both required during adolescent mesofrontal plasticity, but direct convergence has not been demonstrated [27].
Directional relays require evidence for source, receiver and order, ideally with interruption or rescue of the proposed intermediate. The neuron-microglia-astrocyte Wnt pathway after sensory perturbation and the reciprocal astrocyte-ATP-to-microglial-IL-1β circuit after focal injury satisfy that requirement in their tested models [28,29].
Context-dependent effects answer a different question: whether one pathway changes with dose, developmental stage, sex, region, disease state, receptor localization or physical presentation [30,31]. Single-cell, spatial and imaging studies can nominate candidate exposures and sender-receiver relationships, but they do not establish ligand access, receptor protein, direction or function without perturbation [17,18,20].

Spatial Temporal and Cellular Context

Spatial presentation can change a response without changing ligand identity. A focal ATP gradient carries directional information; widespread nucleotide elevation can erase it. Target-bound complement or exposed phosphatidylserine marks a local structure in a way that uniform soluble ligand cannot reproduce. Experiments also support a PIEZO1-dependent response to fibrillar or mechanically stiff amyloid environments, although native fibril stiffness has not been isolated fully from conformation, valency and ligand presentation [10,16,32,33,34].
Time matters just as much. Transient nucleotide signals guide processes; sustained hyperactivity disrupts ATP microgradients; neonatal 5-HT2B signalling can alter later inflammatory responsiveness; a brief ischaemic event can be followed by subacute microglial and circuit changes [34,35,36].
Receptor competence also depends on localization and cellular state. RNA expression alone does not establish a functional surface receptor: the 2025 α7 nicotinic-acetylcholine-receptor study found predominantly mitochondrial localization in isolated mouse microglia, no canonical choline-evoked calcium response, and altered bioenergetic output after agonism [37].
Identity pathways can change how the same intracellular machinery interprets a signal. TGF-β-SMAD signalling supports Sall1 expression, while SALL1 in turn redirects SMAD4 binding toward microglia-specific regulatory elements and away from alternative macrophage programmes [38,39].
Previous uptake changes immediate transcriptional and metabolic state, but whether it changes the response of the same cell to a later controlled signal remains largely untested. Distinct cargos induce cargo-associated programmes, and MerTK-dependent engulfment can initiate an autocrine TGF-β1 programme [14,15]. Purinergic signalling illustrates how location and tissue disturbance guide positioning while P2Y12-dependent behaviour shifts with contacted cell type, duration and neuromodulatory context [25,26,27,32,33,40].
Purinergic signalling explains how a process reaches a structure, but reaching a structure is not the same as deciding to remove it. Target selection introduces a different set of positive and inhibitory signals [16,32,33,41,42].
Preprints 232093 g001
Figure 1. Tissue context and cellular decisions in microglial signalling. Soluble neuronal signals, target-bound cues, trophic and vascular factors, and physical properties reach microglia at different spatial and temporal scales. Their effects depend on the receiving cell and can influence positioning, target recognition, uptake or restraint, degradation and later response. The diagram is a conceptual synthesis and does not imply that every displayed pathway interacts directly.
Figure 1. Tissue context and cellular decisions in microglial signalling. Soluble neuronal signals, target-bound cues, trophic and vascular factors, and physical properties reach microglia at different spatial and temporal scales. Their effects depend on the receiving cell and can influence positioning, target recognition, uptake or restraint, degradation and later response. The diagram is a conceptual synthesis and does not imply that every displayed pathway interacts directly.
Preprints 232093 g001

Purinergic Signals and Cellular Positioning

Spatial and Temporal Purinergic Signals

Focal tissue injury produces extracellular nucleotides that recruit microglial processes through P2Y12, and neuronal NMDA-receptor activation can evoke ATP-dependent process outgrowth in acute brain slices [32,33,43]. Together, these experiments established extracellular nucleotides as the link between local tissue events and directional microglial movement.
The extracellular nucleotide environment is dynamic, not a single ATP signal. ATP is converted through ADP and AMP to adenosine, and different purinergic receptors couple those metabolites to distinct motility, ionic and inflammatory responses; P2Y12, adenosine A2A and A3 receptors are the best-studied examples for process behaviour [40,44,45,46,47]. Release, diffusion and enzymatic conversion together determine which purinergic signal actually reaches a microglial process, at what distance, and when [43,48,49].
Source matters too. Nucleotides can come from damaged cells, active neurons, astrocytes, or vascular and blood-associated compartments, depending on the preparation [29,43,48,49]. That means temporal association between neuronal activity and microglial movement is not enough to infer direct glutamate sensing when ATP is the actual intermediary [43].

P2Y12 Beyond Chemotaxis

P2Y12 is usually introduced as the ADP-responsive Gi-coupled receptor that guides microglial processes toward extracellular nucleotides, but P2Y12-related signalling is associated with functions beyond focal injury [33,40]. In the developing retina and visual cortex, it is tied to apoptotic target availability and experience-dependent plasticity. At vascular interfaces, convergent microglial depletion and global P2ry12/Panx1 experiments support involvement in capillary regulation and focal blood-brain barrier repair, while receptor-specific microglial causality remains less completely isolated [50,51,52,53].
Pósfai et al. combined in vivo three-dimensional two-photon imaging with acute P2Y12 blockade and genetic deletion in mouse somatosensory cortex. Disrupting P2Y12 altered baseline surveillance, morphology and contacts with neuronal somata, smooth-muscle-bearing vessels and oligodendrocyte processes [40]. These findings support a role in normal surveillance and cell-cell contact, not only emergency chemotaxis [40].
The same study also examined human epilepsy tissue and found altered P2Y12 expression associated with disease severity and changes in microglia-neuron interactions [40]. Those human observations are correlational; the causal receptor perturbations were done in mice [40].
Transient and sustained nucleotide disturbances can lead to genuinely different outcomes. During neuronal hyperactivity, disruption of ATP microgradients comes with reduced microglial motility, changes in phagocytic-receptor expression, and impaired coupling between neuronal apoptosis and clearance [34]. Microglial conversion of ATP to adenosine can also feed back onto neurons and suppress neuronal activity through adenosine receptors [48].
A 2026 transient-ischaemic-attack model illustrates how a short event can precede a longer response. Five minutes of bilateral carotid occlusion without detected infarction was followed by ATP events, persistent microglial changes, altered microglia-synapse relationships, network disconnection and cognitive deficits [36]. Prophylactic intracisternal PSB-0739, given before ischaemia, improved several subacute outcomes. The result is consistent with P2Y12 involvement, but it does not establish microglia-specific genetic causality or post-TIA therapeutic efficacy; the ATP source and fate of internalized synaptic material also remain unresolved [36].

Neuromodulatory Control

Norepinephrine gives one of the clearest direct interactions in the whole field. In acute slices and cultured microglia, β2-adrenergic signalling promotes process retraction and directly opposes ATP-directed extension [25,26]. Even this interaction is not fixed, however: inflammatory stimulation changes adrenergic-receptor expression, so the effect depends on microglial state [25,26].
Dopamine works differently. During adolescent mesofrontal plasticity, rewarding experience or optogenetic activation of dopaminergic axons increases microglial surveillance and contacts with dopaminergic boutons, and both dopamine-receptor signalling and P2Y12 are required for the reported increase in surveillance and new bouton formation [27]. Older culture and injury studies show that dopaminergic receptor expression and dopamine’s effects on adhesion, motility and phagocytosis vary with microglial state [54,55,56]. What the adolescent study actually establishes is co-requirement in a defined circuit and developmental window. It does not identify a direct biochemical interaction between dopamine receptors and P2Y12 [27].
P2Y12-dependent microglial contacts can influence neuronal and network endpoints, but receptor-specific loss and whole-microglia depletion are not equivalent perturbations, and the two are often conflated. In mouse barrel cortex, germline P2Y12 loss and pharmacological microglial depletion both altered selected baseline firing, oscillatory and thalamocortical measures, while several stimulus-evoked responses were preserved [57].
Interpretation is still limited by developmental compensation in germline knockout, ecosystem effects of CSF1R-based depletion, and the haemodynamic component of functional-ultrasound measurements [24,52,57]. The safest conclusion is that microglial presence and P2Y12 function constrain selected network properties, not that one P2Y12 mechanism explains every circuit readout [57].

Target Recognition Uptake and Processing

A microglial process can contact a neuron, synapse, vessel or cellular fragment and then withdraw without engulfment. Target recognition, internalization and degradation are therefore distinct steps, not one event. Complement proteins and exposed membrane lipids can favour recognition, while inhibitory pathways such as CD47-SIRPα can restrain execution [16,41,42,58,59,60,61].

Complement and Phosphatidylserine

During development of the visual system, C1q and C3 take part in activity-dependent elimination of retinal inputs involving CR3-expressing microglia. Complement-dependent synapse loss has also been shown in amyloid models and in selected inflammatory and neurodegenerative settings [4,62,63,64].
Different complement components do different jobs. C1q can accumulate on selected structures, C3 cleavage products can supply opsonic signals, and CR3 takes part in recognition and downstream cytoskeletal responses, while C3aR and C5aR1 can influence migration and inflammatory signalling independently of opsonic uptake [65,66]. Moreover, the cellular source varies by component and context: astrocytes can produce C3 in defined amyloid-associated models, and microglia themselves contribute complement components including C1q in several CNS settings [67,68,69].
None of this makes complement a universal pruning mechanism. C3 deficiency protects synapses in some plaque-rich models, and CR3-dependent synapse elimination can contribute to pathology in a systemic-inflammation Parkinsonian model, but other circuits and cargos rely on entirely different recognition systems [64,70,71].
Postoperative neurocognitive disorder is another disease-specific context worth noting. Tibial-fracture surgery increased hippocampal C1q, classical-complement activity, synaptic loss and memory impairment, with C1q-associated excitatory and inhibitory synaptic material detected in microglial compartments. C1q neutralization and CA1 microglial C1q depletion preserved synaptic measures and improved postoperative memory, and NF-κB inhibition reduced C1q induction and improved cognition [69]. The study establishes a causal requirement for C1q in this model, though it does not fully resolve the upstream target-selection signal, or whether every internalized synaptic marker actually came from removal of an intact synapse [69].
Complement-dependent execution is shaped by intracellular state, too. During early postnatal hippocampal development, microglial INPP5D/SHIP1 loss increases complement abundance, synaptic material in microglial phagolysosomes, and synapse loss; neuronal CD55 can prevent the synaptic phenotype, and early deletion has consequences that adult deletion does not reproduce [72].
Phosphatidylserine is normally enriched on the inner plasma-membrane leaflet but becomes exposed during apoptosis, and it can also appear locally during cellular stress and membrane remodelling. In the CNS, microglia can recognize exposed phosphatidylserine through several receptor and bridging systems, including TAM receptors, MFG-E8-integrin pathways, GPR56 and TREM2 in defined settings [16,73,74,75,76,77,78,79,80].
Which receptor gets used depends on the target and the niche. GPR56 contributes to phosphatidylserine-dependent developmental synapse refinement, while MerTK is required to remove persistently phosphatidylserine-exposed inhibitory postsynaptic structures after neuronal Cdc50a disruption [74,79]. In an Aβ-oligomer co-culture and early or pre-plaque mouse contexts, phosphatidylserine-TREM2-dependent synaptic uptake was associated with reduced neuronal hyperactivity [80]. This does not establish that increasing synapse removal is generally beneficial in aged plaque-rich disease. The article later received corrections to statistical and source-data presentation; the authors stated that the conclusions were unchanged [80].

TREM2 and Inhibitory Checkpoints

TREM2 signals through TYROBP/DAP12 and influences survival, lipid handling, clustering around pathology, phagocytosis and state transitions, while ligand binding, surface delivery, proteolytic processing and soluble TREM2 can all vary independently of one another [15,31]. The TREM2-T96K variant makes the problem concrete: an assay-defined increase in ligand-dependent activity did not predict the in vivo phenotype in 5xFAD mice, where plaque-associated microglial responses, soluble TREM2 and disease-associated state transitions were altered, with the most prominent effects occurring in females [31].
The measured endpoint mattered as much as the variant itself. Aβ uptake was impaired in engineered cell assays, but several in vivo measures (plaque burden, neuronal or synaptic pathology, methoxy-X04 uptake) did not move in the same direction, and cognition was never tested [31]. A single receptor property, in other words, is not enough to predict tissue outcome [31].
CD47 on neuronal or synaptic membranes can engage microglial SIRPα and recruit inhibitory phosphatases that restrain cytoskeletal activation. During development, CD47 protects synapses from excessive elimination; loss of microglial SIRPα can increase engulfment and worsen selected synaptic or cognitive outcomes in neurodegenerative models [58,59,60,61].
It is plausible that positive signals such as complement or phosphatidylserine are balanced against inhibitory CD47-SIRPα signalling on the same target, but direct multi-input competition has not actually been demonstrated in every context where the idea gets invoked [16,58,59,60,61,62]. Retinal development further shows that neuronal SIRPα can alter access of neuronal CD47 to microglial SIRPα, and inflammatory models show that CD47 effects can shift with compartment and intervention timing [81,82,83].
Other checkpoints matter in specific states. CD22 increases on aged microglia, and its blockade improves selected homeostatic, phagocytic and cognitive measures in aged mice; low-affinity Fcγ receptors contribute to dopaminergic-neuron elimination in immunoglobulin-rich Parkinson-related models [84,85].

Uptake Degradation and Later State

Three-dimensional containment of a synaptic marker within a microglial volume is stronger evidence than two-dimensional overlap, but it still does not reconstruct how the material got there. Internalized signal can reflect whole-structure engulfment, trogocytosis, uptake of shed fragments, or clearance after degeneration [69,86].
In a developmental hippocampal preparation, Weinhard et al. observed presynaptic trogocytosis and microglia-induced spine-head filopodia rather than widespread wholesale postsynaptic engulfment [86]. The postoperative C1q study similarly supports internalization of synaptic material and a causal complement contribution, though fixed-tissue imaging on its own cannot establish that every event represents removal of an intact synapse [69].
Claims of whole-target elimination are strongest when lysosomal localization, live or correlative imaging, loss of the target structure, evidence of degradation, and receptor-specific perturbation all converge [16,69,86]. Terminology should follow the measurement: internalized synaptic material, trogocytosis, and elimination of an intact synapse are not the same observation, and should not be written as if they were [86].
Human iPSC-derived microglia exposed for 24 hours to synaptosomes, myelin, apoptotic neurons or amyloid fibrils adopt cargo-associated programmes involving APOE, GPNMB, lipid handling and lysosomes [15]. These experiments establish an immediate in vitro response to substrate exposure, not durable memory or an altered response to a later controlled challenge [15].
The MerTK pathway provides a mechanistic bridge from target recognition to later state. In optic-nerve injury and 5xFAD models, MerTK-dependent phospholipase-C signalling increases PU.1 and IRF8, induces microglial TGF-β1, and establishes an autocrine programme required for selected neurodegenerative responses [14].
Successful internalization does not guarantee successful digestion, either. After spinal injury, astrocyte-derived CCN1 acts through microglial SDC4 to support lipid-droplet buffering and processing of myelin-rich debris; astrocytic Ccn1 loss produces debris-laden microglial nodules with impaired recovery [87]. In a separate lysosomal relay, microglial β-hexosaminidase supports neuronal GM2 degradation, while neuronal GM2 accumulation in Hexb-deficient mice engages microglial MGL2 and promotes inflammatory output [88].
Cell state, then, is not only an endpoint of signalling. It becomes part of the starting condition for whatever comes next [14,15,87,88].

Cellular State and Model Context

Two microglial cells expressing the same receptor need not respond identically if their developmental history, trophic environment, metabolic state or local niche differs. CSF1R and TGF-β pathways show how the tissue environment sets up the state in which later signals get interpreted [38,39,89,90,91,92,93].

Trophic Signals and Microglial Identity

CSF1R signalling is essential for microglial development and maintenance, and its ligands CSF1 and IL-34 differ in cellular source, anatomical distribution and developmental timing despite acting through the same receptor [89,90,91,92]. Pharmacological CSF1R inhibition therefore changes more than cell number: other CNS macrophage populations can be affected too, and surviving or repopulating microglia encounter tissue that has already been altered by cell loss and accumulated signals [24,90].
IL-34 shows functional effects that go beyond survival. In the developing anterior cingulate cortex, excitatory-neuron-derived IL-34 supports microglial abundance and maturation; partial reduction increases inappropriate thalamocortical synapse engulfment, while overexpression suppresses physiologically appropriate engulfment [30]. Ligand abundance changes competence, not just population size [30].
In adult mice, microglial TGF-β1 is required to maintain homeostatic microglial features and normal cognitive function, and deletion from astrocytes or forebrain neurons does not reproduce the same phenotype [38]. During neurodegeneration, MerTK signalling can induce an additional microglial TGF-β1 autocrine programme [14].
TGF-β is secreted in a latent form and needs extracellular activation before receptor engagement, and the transcriptional response depends on intracellular regulatory state. SALL1 matters here in particular, because TGF-β-SMAD signalling supports Sall1 expression, and SALL1 redirects SMAD4 binding toward microglial regulatory regions [38,39]. Astrocyte-derived TGF-β can also suppress microglial inflammatory activation in defined IL-10-dependent settings [93].

Sex Age and Brain Region

Adult male and female mouse microglia differ in selected transcriptional and functional features, and age, sex and genotype jointly alter plaque-associated responses in Alzheimer models [94,95]. The TREM2-T96K phenotype reinforces this: several of its most prominent effects occurred specifically in female 5xFAD mice [31]. A genuine sex-specific claim, however, requires an explicit sex-by-treatment comparison: significance in one sex and non-significance in the other is not the same thing [19].
Age changes receptor and functional state in other ways too. CD22, for instance, rises on aged microglia and becomes experimentally targetable in aged mice [84]. Regional context matters just as much, since retinal development, cortical surveillance, white-matter repair and vascular interfaces all expose microglia to different neuronal, astrocytic, matrix and vascular environments [51,52,87].

Species and Human Models

Most causal experiments in this review come from mice, simply because longitudinal imaging, genetic manipulation and intact-circuit perturbation are far more accessible in animal models. Human and mouse microglia nevertheless differ transcriptionally, and removing human microglia from the CNS rapidly remodels part of their transcriptional and enhancer landscape [21].
Turnover adds another species difference. Mouse studies show ongoing local self-renewal with context-dependent expansion, whereas human cortical microglia renew slowly, at a reported median rate of about 28% per year [11,12,13]. Cell longevity defines the opportunity for accumulated exposure but should not be used as evidence that a molecular state persists for the same duration.
Human iPSC-derived microglia allow receptor-specific perturbation, and assembloid systems restore interactions with regionally patterned neural cell types. Even so, each model captures only part of the adult human environment [15,96].
Human midbrain-striatal assembloids give one causal example. Regionally patterned microglia show GABA_B-receptor-dependent calcium responses to circuit activity, and in SCN2A-mutant assembloids, increased microglial calcium activity and synaptic pruning are reduced by pharmacological GABA_B-receptor inhibition or microglial GABBR1 deletion [96].
Postmortem studies add a complementary kind of state resolution. Ex vivo profiling of microglia from 189 human brains identified Alzheimer-associated changes in genes, isoforms, co-expression and molecular subtypes, and 2026 spatial and single-nucleus profiling across very old brains identified microglial programmes associated with different stages of amyloid and tau pathology and with resilience [17,18]. These datasets identify states and niches that genuinely exist in human disease, but they cannot by themselves tell us which extracellular signals created them [17,18,20].

Sequential Innate Immune Signalling

Innate immune pathways illustrate sequential control, in which an earlier event changes whether a later input can execute a response. Canonical NLRP3 separates priming from activation in appropriate mouse-microglia models; cGAS-STING converts misplaced intracellular DNA into interferon and inflammatory output [97,98,99,100,101,102]. Figure 2 distinguishes demonstrated sequential dependency from state changes that still require a later rechallenge test.

NLRP3 Priming and Activation

Canonical NLRP3 responses are usually studied as a two-step process. Priming increases the availability of NLRP3 and pro-IL-1β, and a later activating input (extracellular ATP, for instance) can then promote inflammasome assembly, caspase-1 activation and mature IL-1β release in appropriately primed mouse microglia [101]. Increased NLRP3 expression or high extracellular ATP alone does not mean the full inflammasome pathway has executed [101].
Disease-model genetics establish biological consequence without replacing this two-step mechanism. In APP/PS1 mice, NLRP3 or caspase-1 deficiency alters amyloid-associated pathology and microglial responses; in tau models, NLRP3 loss reduces tau hyperphosphorylation and aggregation [97,98].
Inflammasome output can also create a new extracellular problem of its own. Microglia-derived ASC specks bind amyloid-β and promote aggregation and spreading in experimental systems: a feed-forward link between intracellular inflammasome execution and extracellular pathology [103].

cGAS STING in Ageing and Disease

Cyclic GMP-AMP synthase (cGAS) is activated by cytosolic DNA and produces the second messenger cGAMP, which binds and activates STING to drive IRF3-dependent type I interferon signalling [102]. This is how microbial DNA, or misplaced self-DNA, gets converted into a broader inflammatory response [102].
In microglia, mitochondrial damage can supply that self-DNA. In aged mice, mitochondrial DNA release engages cGAS-STING and contributes to reactive microglial states, neurodegeneration and cognitive decline [99]. In tauopathy, pathogenic tau activates microglial cGAS and type I interferon responses, partly through mitochondrial DNA leakage; Cgas ablation preserves synaptic and cognitive measures without simply reducing tau load [100].
In 5xFAD mice, inducible deletion of cGAS from adult CX3CR1-lineage CNS myeloid cells at the onset of amyloid pathology restricted plaque accumulation, altered plaque-associated microglial accumulation, reduced inflammasome-related readouts and improved selected behavioural measures [104]. Because Cx3cr1-CreERT2 can include long-lived CNS-associated macrophages and the spatial-memory probe outcome was unchanged, the findings should not be described as exclusively microglial or as uniform cognitive protection [104].
It also creates a mechanistic bridge to NLRP3: inflammasome activity can emerge downstream of a cellular state shaped by cGAS, rather than only from a simple extracellular ATP trigger [101,104]. That relationship is context-specific, though, and should not be generalized to every inflammasome response [101,104].
Several neurodegeneration studies support pathological consequences of excessive cGAS-STING activity. Pharmacological STING blockade in an amyloid/tau knock-in model reduced inflammatory, NLRP3-related, synaptic and pathological readouts, and microglial cGAS deletion was beneficial in 5xFAD mice [104,105].
Physiological ageing tells a different story. STING deficiency in aged mice worsened blood-brain barrier breakdown, microhaemorrhages and neuromotor deficits, and microglial STING expression was sufficient to protect against several age-associated changes [106]. So the same pathway can contribute to pathology in one setting while remaining protective in another [104,105,106].
A 2026 study identified S-nitrosylation of STING at C148 as a disease-associated regulatory event in Alzheimer-related models and human tissue; experimentally preventing this modification reduced STING-driven inflammatory and synaptic pathology in the reported systems [107]. This adds the post-translational state of STING to DNA availability and cGAS activity as another level at which the pathway can shift.
Human genetic context can alter it too. In tauopathy models carrying the APOE3 Christchurch R136S variant, tau-associated pathology and interferon responses were reduced together with suppression of microglial cGAS-STING signalling; pharmacological cGAS inhibition in conventional APOE3 tauopathy mice reproduced several of the same protective features [108]. The protective effect of the variant is not reducible to one pathway, but the study does give a direct link between genetic background, microglial innate immune signalling and tissue resilience [108].
NLRP3 and cGAS-STING, taken together, show that integration can cross the plasma membrane repeatedly: extracellular or pathological stress changes intracellular state, intracellular sensors generate inflammatory output, and that output goes on to change neighbouring cells and the extracellular environment [29,99,100,102,103,104,105,106,107,108].

Multicellular and Physical Niches

Once a microglial response changes neighbouring cells, microglia alone can no longer explain the mechanism. Astrocytes, neurons, vascular cells and other macrophage populations can provide the first signal, receive a microglial response, and then alter the tissue that instructed the microglial cell in the first place [6,28,29,109,110,111,112,113,114,115,116,117,118].

Astrocyte Microglia Relays

During development, astrocyte-derived IL-33 acts on microglia in the studied spinal-cord and thalamic settings, promotes synapse engulfment, and is required for normal circuit maturation [110]. A processed C-terminal fragment of astrocytic Hevin similarly engages microglial TLR4, promotes a TLR2-high phagolysosomal state, and supports developmental thalamocortical synapse refinement [111]. Both mechanisms are anatomically and developmentally restricted [110,111].
Communication can run the other way too. After sensory perturbation in postnatal barrel cortex, neuronal CX3CL1-microglial CX3CR1 signalling promotes microglial Wnt release; Wnt signalling in astrocytes reduces perisynaptic astrocytic contacts and is required for the microglial engulfment and synapse loss that follows [28]. The sequence runs neuron to microglia to astrocyte, and back to the synaptic environment [28].
Microglial IL-1α, TNF and C1q can jointly induce a neurotoxic astrocyte programme in defined culture and disease models, though the historical A1 label does not capture every reactive astrocyte state [112]. In experimental autoimmune encephalomyelitis, microglial TGF-α and VEGF-B have opposing effects on astrocyte pathogenicity, and microbial tryptophan metabolites act further upstream, through microglial aryl-hydrocarbon-receptor signalling [113].
Reciprocal communication can support repair, too. After spinal-cord injury, astrocytic CSF1 at the lesion border supports perilesional microglial proliferation and repair, and microglial IFN-β supports astrocyte survival and border formation; cell-specific loss and rescue support communication in both directions [109].
Following focal cortical injury, astrocytes generate Panx1-dependent ATP events whose magnitude scales with lesion severity. Microglia adjust IL-1β output in response, and microglial IL-1β acts through astrocytic IL-1R1 and calcium-calcineurin signalling to suppress further Panx1-dependent ATP release [29].
This pathway includes built-in negative feedback: injury drives astrocytic ATP, which recruits a microglial response that then helps restrain continued amplification of the astrocytic signal [29]. That does not make IL-1β generally protective. It shows that cytokine function depends on sender, receiver and timing, not on the molecule’s reputation [29].

Perivascular and Barrier Signals

In mouse Alzheimer models, SPP1 localizes predominantly to perivascular macrophages and, to a lesser extent, fibroblasts. Global Spp1 deficiency reduced the C1q-associated microglial programme, synaptic-material engulfment and synapse loss in the reported models [6]. The experiment establishes ligand requirement together with a plausible perivascular source, but source-specific necessity remains unresolved. A 2026 author correction updated portions of the text, methods and source-data files [6].
White matter offers a different version of niche control. After spinal injury, lesion-remote astrocytes produce CCN1, which acts through microglial SDC4 to support lipid-droplet buffering and digestion of myelin-rich debris; loss of astrocytic Ccn1 produces debris-laden but dysfunctional microglial nodules and impaired recovery [87].
Blood-brain barrier disruption introduces circulating proteins into a compartment from which they are normally excluded. Fibrin-derived signals can engage CD11b-containing integrins on myeloid cells, and selective interference with the fibrin-CD11b interaction suppresses inflammatory pathology in experimental CNS autoimmunity without globally blocking coagulation [114].
Vascular cells can signal actively as well. Endothelial MMP-3 after spinal injury contributes to microglial activation and oligodendrocyte damage, and VEGF has been linked to microglial chemotaxis toward amyloid-associated tissue and to altered scavenger-receptor expression in ischaemic settings [115,116,117]. Attribution near vessels is genuinely difficult, since parenchymal microglia, perivascular macrophages and infiltrating myeloid cells occupy closely related regions and share markers [24,118].
Longitudinal and lineage-resolved approaches matter here for exactly that reason: to work out whether vascular dysfunction recruits microglia, whether microglia stabilize an acute lesion, or whether a later inflammatory phase actually worsens tissue injury [24,52,53,118].

Mechanosensing

Chemical identity is only part of the information extracellular structures present. Microglial PIEZO1 contributes to responses to fibrillar or mechanically stiff amyloid environments, including calcium entry, plaque clustering, compaction and phagocytic readouts in an Alzheimer model [10]. Because soluble peptide and rigid fibrils also differ in conformation, aggregation, valency and ligand presentation, the contribution of native fibril stiffness cannot yet be separated completely from fibril chemistry [10].
What remains unresolved is how chemical and mechanical information get combined. Experiments that vary stiffness and molecular composition independently are needed before mechanosensing can be assigned a defined interaction with receptors such as TREM2 [10,31].

Neuronal Activity and Microglial Responses

Neuronal activity is treated here as a cross-cutting input rather than a separate receptor catalogue. Activity changes extracellular neurotransmitters, neuromodulators and nucleotides together, so a microglial response can arise through a transmitter receptor, indirectly through purinergic signalling, or through both routes [27,43,49,119,120]. Norepinephrine and dopamine were covered under Purinergic Signals and Cellular Positioning; the remaining systems illustrate direct receptor effects, indirect purinergic routes and context-dependent responses. Table 1 summarizes the best-supported decision and the main limitation for each system.

GABA and Glutamate

Favuzzi et al. identified GABA-receptive microglia during a postnatal cortical developmental window. These cells preferentially interacted with inhibitory synapses, and disrupting microglial GABA-receptor signalling altered the associated transcriptional programme, inhibitory connectivity and behavioural outcomes without a matching effect on excitatory synapses [125].
Human assembloids add complementary evidence. Regionally patterned human microglia show GABA_B-receptor-dependent calcium responses to circuit activity, and microglial GABBR1 deletion reduces abnormal pruning in SCN2A-mutant assembloids [96]. Together, these two studies link transmitter sensing to defined target classes and circuit effects, though both remain model- and stage-specific [96,125].
Neuronal NMDA-receptor activation can produce microglial process extension through ATP release: an indirect route from glutamatergic activity to a purinergic microglial response [43]. Direct ionotropic and metabotropic glutamate-receptor effects have also been reported in cultured or isolated microglia, but receptor profiles and functional outputs vary substantially across preparations [120,126,127,128,134].
For transmitter receptors generally, RNA expression is only a starting point. Receptor protein at the relevant cellular compartment, an immediate functional response, and receptor-specific perturbation are what actually provide stronger evidence of direct microglial sensing [37,43,120].

Acetylcholine and Histamine

Muscarinic M3 signalling has been linked to microglial recruitment, phagocytic responses and injury evolution after experimental stroke [129]. α7 nicotinic agonists have also been reported to reduce inflammatory outputs in fetal microglia and in cell-line or conditioned-medium systems [130,135,136].
A 2025 localization study complicates the standard surface-ion-channel interpretation, though. In isolated mouse microglia, α7 immunoreactivity turned out to be predominantly mitochondrial, canonical choline-evoked calcium responses were absent, and agonism increased ATP production instead [37]. The earlier pharmacological effects may well be real, but they probably do not share the same plasma-membrane mechanism assumed in neurons [37].
Histamine H3-receptor activation suppresses ATP-evoked calcium responses, chemotaxis, phagocytosis and cytokine secretion in primary mouse microglia [131]. This finding is notable because histamine and ATP were tested within the same functional response: direct evidence that one transmitter system can modify another extracellular input [131].
Other histamine receptors influence inflammatory and oxidative responses in culture, but systemic receptor manipulation is harder to attribute cleanly, since histamine receptors also regulate neuronal release and are expressed by multiple CNS cell types [137,138,139,140].

Serotonin and Model Limitations

Serotonergic signalling can increase injury-directed microglial motility while reducing phagocytosis in postnatal or cultured preparations: movement and uptake do not have to change in the same direction [132]. Microglia-specific 5-HT2B deletion during the neonatal period produces more persistent later inflammatory consequences than adult deletion, and 5-HT2B supports mononuclear-phagocyte integrity in an ALS model [35,133].
A separate post-stroke-depression study reported that HMGB1 associates with 5-HT7R and suppresses receptor-linked cAMP signalling. Interaction assays support direct receptor-level modulation, but in vivo attribution is less clean here, because 5-HT7R deletion was global, HMGB1 inhibition was systemic, and multiple CNS cell types changed receptor expression [141].
In a 6-hydroxydopamine Parkinsonian rat model, golexanolone treatment was associated with changes in IBA1-defined microglial morphology and glutaminase, tissue glutamate, astrocytic GABA-related measures, dopaminergic markers and behaviour [142]. The proposed microglia-glutamate-astrocyte GAT3/GABA-dopamine relay is plausible on its face, but microglial glutamate release was never measured directly, astrocytic GAT3 was never specifically interrupted, and the systemic drug could affect GABAergic signalling independently of microglia altogether [142].
The real contribution of a model like this is hypothesis generation. A causal test would manipulate microglial glutaminase, measure extracellular glutamate, interrupt astrocytic GAT3, and test downstream rescue within the same sequence [142].
Across transmitter systems, neuronal activity reaches microglia by several routes at once. Some transmitters directly modify microglial responses to other signals; some neuronal signals get converted into ATP before microglia ever respond; some transmitter receptors help define target-specific developmental programmes [25,26,27,43,96,125,131]. Receptor localization can change the mechanism further still, as the mitochondrial α7 receptor makes clear [37].
Selected microglial populations can respond to neurotransmitters. The mechanistic question is which transmitter reaches which receptor and cellular compartment, in which developmental or disease state, and which decision-specific function changes as a result [27,37,43,96,125,131].

Key Gaps and Experiments

The literature now supports many individual pathways, but far fewer experiments test how those pathways operate together, in the same microglial cell and the same tissue context. Complement can label a synapse, CD47 can inhibit engulfment, norepinephrine can change process behaviour, IL-34 can alter developmental state, and PIEZO1 can respond to amyloid stiffness, yet these mechanisms are almost always studied separately [10,25,30,58,62].
Norepinephrine changes microglial process behaviour and can directly oppose ATP-directed extension, and complement contributes to synaptic elimination in development and disease [4,25,26,63]. A direct test would manipulate noradrenergic tone and complement within the same preparation while separately measuring process movement, stable contact, complement deposition and target loss [4,25,26,63].
Complement and phosphatidylserine can favour target recognition, while CD47-SIRPα restrains engulfment [16,58,59,60,61,62,73,74,75,76,77,78,79,80]. The cleanest test of a common threshold model would present otherwise matched targets with controlled amounts of phosphatidylserine, complement and CD47, then follow uptake through degradation [16,58,62]. IL-34 could then be varied to ask whether trophic state changes how those target signals get interpreted [30].
Amyloid exposes microglia to chemical ligands and to altered physical properties at the same time. PIEZO1 senses fibril stiffness; TREM2 responds to lipid- and protein-associated ligands in the amyloid environment [10,31]. Varying stiffness and chemical composition independently, while perturbing PIEZO1 and TREM2, would tell us whether those two information streams converge on plaque clustering, compaction or uptake [10,31].
Cargo exposure changes lysosomal, lipid-handling and transcriptional programmes, but most experiments stop soon after uptake [14,15,87]. Longitudinal tracking of defined microglia after uptake of myelin, apoptotic material, synaptic components or protein aggregates could test whether later responses to P2Y12, MerTK, TREM2 or SIRPα are altered in those same cells [14,15,87].
Human single-cell and spatial datasets identify reproducible microglial states but do not identify the causal signal combination that produced them [17,18,20]. In amyloid models, SPP1 localization and global loss support a perivascular-cell-to-microglia hypothesis, but source-specific necessity remains untested [6]. A decisive experiment would combine perivascular-cell-restricted Spp1 loss and rescue with microglial receiver perturbation and functional measures of complement state, uptake and synapse preservation [6].
Three questions deserve priority. First, does a microglial cell combine positive and inhibitory labels on the same target? Second, does prior cargo processing change the response of that same cell to a later cue? Third, do chemical ligands and mechanical properties interact at plaques or damaged vessels? These experiments would move the field beyond pathway catalogues because each produces a falsifiable interaction or sequence test [6,10,14,15,16,25,30,58].

Therapeutic Implications

The same microglial pathway can damage tissue in one setting and support normal function in another. A treatment that simply suppresses microglial activation, then, is unlikely to be precise enough [36,40,52,53,80,99,100,104,105,106].
P2Y12 illustrates the problem. Prophylactic intracisternal antagonism reduced selected outcomes in a mouse transient-ischaemia model, whereas P2Y12-related signalling also contributes to normal surveillance, cell-cell contacts and vascular responses [36,40,52,53]. Noradrenergic β2-receptor stimulation can improve selected plaque-associated outcomes in 5xFAD mice even though β2 signalling acutely opposes ATP-directed process extension in other preparations [25,26,143]. Complement similarly participates in developmental refinement and pathological synapse loss [4,62,63,69].

Timing and Pathway Context

The astrocyte-microglia ATP-IL-1β circuit shows how timing and receiver identity can reverse a simple therapeutic expectation: early microglial IL-1β actually restrains further astrocytic ATP release after focal injury [29]. cGAS-STING offers another example. Excessive pathway activity contributes to pathology in several neurodegeneration models, yet microglial STING protects vascular and neurological function during physiological ageing [99,100,104,105,106].
The TREM2-T96K study shows that increasing an assay-defined receptor property does not guarantee improved processing or tissue outcome [31]. The same holds for phagocytosis more broadly: increasing uptake is not necessarily beneficial if the target is a viable synapse, and reducing engulfment can impair removal of genuinely damaged material [41,42,76,80]. Pioglitazone-induced reductions in synaptic C1q and engulfment-related readouts in APP/PS1 mice make a related point: changing the target environment can alter pathology without proving a direct, microglia-specific drug action [144]. Functional endpoints need to match the therapeutic objective, not just move in a favourable direction [31,41,80,144].
CSF1R inhibitors can deplete large fractions of microglia, but depletion also changes other CNS macrophage populations and the tissue environment that surviving or repopulating cells encounter afterward [24,89,90]. A disease driven by one harmful microglial function may not, in fact, require removing the entire population [24,89,90].

Combination Interventions

If one intervention changes target labelling and another changes execution, both agents, and their combination, should be tested in a design that can distinguish independent from interacting effects, and evaluated across the relevant treatment windows [19]. Otherwise, an apparent additive benefit can conceal the loss of a homeostatic function, or an effect arising in a non-microglial compartment entirely [19,24].
Selective disruption of a disease-associated interaction can preserve the other functions of the same pathway. The fibrin-CD11b example shows this well: interfering with the inflammatory fibrin-integrin interaction can suppress experimental CNS pathology without globally blocking coagulation [114]. The same logic motivates interest in pathological complement deposition, disease-associated STING modification, and other context-specific interactions, rather than indiscriminate pathway shutdown [63,69,107,114].

Human Translation

A mouse mechanism becomes more convincing for human translation when the relevant ligand and receptor protein are localized in human disease tissue, the proposed coupling is perturbed in human cellular systems, and the corresponding state or anatomical niche is observed across donors [15,17,18,21,96]. Human tissue establishes presence, localization and association in disease; it does not by itself establish a functional relationship.
The therapeutic goal that emerges from this review, then, is not global stimulation or suppression of microglia. It is reducing a defined pathological function while preserving surveillance, removal of genuinely damaged material, vascular support and tissue repair, wherever those functions are still needed [36,40,52,53,80,87,106].

Experimental Standards

The preceding section identifies mechanism-specific gaps and decisive experiments. The standards below explain how to interpret those experiments. The design should follow the biological claim and distinguish direct interaction, relay, context dependence and association [19,20,145,146].

Interaction and Sequential Designs

If signal A and signal B are suspected of affecting the same function, the minimum design is control, A alone, B alone and A plus B. The response scale and null model must be chosen in advance: a statistical interaction establishes functional effect modification on that scale, not necessarily receptor convergence or biochemical binding. Mechanistic interaction additionally requires a predefined pathway dependency or pathway-specific perturbation [19].
Dose, order and interval can all change the result. Sustained hyperactivity disrupts purinergic signalling, IL-34 shows dose-dependent developmental effects, and NLRP3 requires priming before activation [30,34,101]. Sequential designs should test A-to-B against B-to-A when the biology allows it, and should vary the interval when priming, adaptation or cargo history are plausible mechanisms [19,30,34,101].

Spatial and Functional Validation

Uniform bath application is a poor match for membrane-bound checkpoints, opsonized targets, focal nucleotide gradients or mechanical signals. Target-bound ligands, local uncaging, organotypic preparations and stiffness-controlled materials can preserve spatial or mechanical information that disappears the moment it hits a uniform solution [10,16,32,58].
The same principle applies to multicellular relays. Recombinant ligand can show that microglia are capable of responding; source-specific loss and rescue are the stronger test of whether a particular neighbouring cell actually supplies the relevant signal in vivo [6,28,29,109].
Broad state markers can show that microglia changed without ever revealing what the cell actually did. Positioning questions need process or cell movement; target-selection questions need contact and recognition; engulfment questions need internalization; clearance questions need degradation; vascular hypotheses need vessel measurements; circuit hypotheses need neuronal or network endpoints [8,19,36,52,57,86].
Transcriptional profiling is most useful when it explains a state tied to a measured function, not when it substitutes for the function itself [15,17,18].
Static microscopy should be read according to its actual resolution. Three-dimensional containment and lysosomal localization strengthen the case for uptake, but whole-target elimination is best supported when temporal imaging, target loss, cargo degradation and receptor-specific perturbation all converge [16,69,86]. Trogocytosis, scavenging of shed material, and elimination of an intact synapse should stay distinct terms, not synonyms [86].

Computation Replication and Model Choice

Single-cell and spatial datasets can identify candidate senders, receivers and niches, but ligand-receptor algorithms depend on the interaction resource and computational method used, and they do not establish ligand access, receptor protein, signalling direction or functional consequence [20]. Their most useful role is narrowing the experimental search space and nominating mechanisms for source-specific and receiver-specific perturbation [17,18,20].
Cells, synapses and image fields nested within the same animal or donor are not independent biological replicates, however many of them you count. Hierarchical models or appropriate aggregation are needed whenever the experimental unit is the animal or donor, and a large number of cells cannot compensate for too few independent subjects [145,146]. This matters especially for treatment-by-sex or genotype interactions [19,145,146].
Reductionist systems can test direct interactions; mouse models can establish intact-tissue and circuit consequences; human iPSC-derived systems allow human genetic perturbation; human tissue establishes whether the proposed state and niche exist in disease at all [15,17,18,21,96]. Mechanistic confidence grows when different models answer different parts of the same question, not when one model gets treated as a complete substitute for the human brain [15,17,18,21,96].
Some observations will remain associations, whereas others establish context-dependent responses, direct interactions or causal multicellular relays. The wording should state which inference the experiment supports and where that inference stops [17,18,19,20,28,29].

Conclusions

Microglia are commonly studied one receptor at a time, but receptor identity alone does not predict a cellular decision. Spatial presentation, duration, target identity, receptor and metabolic state, developmental stage and surrounding niche all alter the conditions under which a pathway operates [6,10,14,15,16,25,30,31,87]. Cargo exposure clearly changes immediate state; whether it changes a later response in the same cell remains an open experimental question.
Several component mechanisms and relays are mechanistically secure, but direct multi-input integration remains uncommon. Noradrenergic signalling functionally opposes ATP-directed extension, and defined astrocyte-microglia and neuron-microglia-astrocyte circuits establish directional relays. Complement, phosphatidylserine, inhibitory checkpoints, TREM2, PIEZO1 and cGAS-STING instead define individual pathways or context-dependent mechanisms whose combinations mostly remain untested [10,16,25,26,28,29,58,99,100,101,102,105].
Sequential control is established most clearly for canonical NLRP3 priming and activation in appropriate mouse-microglia models [101]. MerTK-dependent autocrine TGF-β1 signalling and cargo-associated transcriptional programmes show that an earlier event changes immediate cellular state [14,15], but they do not yet demonstrate that the same cell responds differently to a later controlled cue. That hypothesis requires A-then-B versus B-then-A designs, matched single-cue controls, defined washout intervals and same-cell functional tracking.
That reframing matters for therapy, because a pathway that damages tissue in one setting can support surveillance, vascular stability, clearance or repair in another [36,40,52,53,80,99,100,104,105,106]. The practical question a treatment has to answer is not “activate or suppress microglia,” but which function has become abnormal, what upstream history put the cell in that state, and whether the abnormal function can be corrected without erasing the responses the tissue still depends on [31,80,87,106,114].
The main opportunity is therefore to test known pathways in defined combinations and sequences. Without such experiments, many proposed examples of integration remain plausible juxtapositions rather than demonstrated functional interactions [6,10,19,25,28,29,30,58].

Author Contributions

M.T. is the sole author and is responsible for conceptualization, literature review, evidence appraisal, writing-original draft, and writing-review and editing.

Funding

This work received no external funding.

Institutional Review Board Statement

Not applicable. This work is a narrative review of previously published literature and contains no new human- or animal-subject research.

Data Availability Statement

No new datasets were generated or analysed for this narrative review. All cited sources are listed in the reference section.

Conflicts of Interest

The author declares no conflicts of interest.

Use of Generative AI

Generative AI tools, including ChatGPT (OpenAI), SciSpace, and LeapSpace, were used for preliminary organization, language drafting, figure design, and editorial review. The author reviewed the cited literature, verified the scientific claims and references, substantially revised the manuscript, and takes full responsibility for the final content, interpretations, and conclusions.

References

  1. Sierra A, et al. The “Big Bang” for modern glial biology: translation and comments on Pío del Río-Hortega’s 1919 series of papers on microglia. Glia. 2016;64(11):1801–1840.
  2. Nimmerjahn A, Kirchhoff F, Helmchen F. Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science. 2005;308:1314-1318. [CrossRef]
  3. Paolicelli RC, et al. Synaptic pruning by microglia is necessary for normal brain development. Science. 2011;333(6048):1456–1458.
  4. Schafer DP, et al. Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron. 2012;74(4):691–705.
  5. Wake H, et al. Resting microglia directly monitor the functional state of synapses in vivo and determine the fate of ischemic terminals. Journal of Neuroscience. 2009;29(13):3974–3980.
  6. De Schepper S, Ge JZ, Sierksma A, et al. Perivascular cells induce microglial phagocytic states and synaptic engulfment via SPP1 in mouse models of Alzheimer’s disease. Nature Neuroscience. 2023;26:406–415. Author Correction. Nature Neuroscience. 2026. doi:10.1038/s41593-025-02197-6. [CrossRef]
  7. Li Q, Barres BA. Microglia and macrophages in brain homeostasis and disease. Nature Reviews Immunology. 2018;18(4):225–242.
  8. Cserép C, et al. Microglia monitor and protect neuronal function through specialized somatic purinergic junctions. Science. 2020;367(6477):528–537.
  9. Whitelaw BS, Stoessel MB, Majewska AK. Movers and shakers: microglial dynamics and modulation of neural networks. Glia. 2023;71(7):1575–1591.
  10. Hu J, Chen Q, Zhu H, et al. Microglial Piezo1 senses Aβ fibril stiffness to restrict Alzheimer’s disease. Neuron. 2023;111(1):15–29.e8. [CrossRef]
  11. Askew K, Li K, Olmos-Alonso A, et al. Coupled proliferation and apoptosis maintain the rapid turnover of microglia in the adult brain. Cell Reports. 2017;18(2):391-405. [CrossRef]
  12. Tay TL, Mai D, Dautzenberg J, et al. A new fate mapping system reveals context-dependent random or clonal expansion of microglia. Nature Neuroscience. 2017;20(6):793-803. [CrossRef]
  13. Réu P, Khosravi A, Bernard S, et al. The lifespan and turnover of microglia in the human brain. Cell Reports. 2017;20(4):779-784. [CrossRef]
  14. Huang Y, Deng Z, Zhou Z, et al. MerTK-triggered TGF-β1 autocrine signal regulates microglial response to neurodegeneration. Nature Communications. 2026;17:2312. [CrossRef]
  15. Dolan MJ, et al. Exposure of iPSC-derived human microglia to brain substrates enables the generation and manipulation of diverse transcriptional states in vitro. Nature Immunology. 2023;24(8):1382–1390.
  16. Scott-Hewitt N, Perrucci F, Morini R, et al. Local externalization of phosphatidylserine mediates developmental synaptic pruning by microglia. EMBO Journal. 2020;39:e105380. [CrossRef]
  17. Kosoy R, Fullard JF, Bendl J, et al. Alzheimer’s disease transcriptional landscape in ex vivo human microglia. Nature Neuroscience. 2025;28:1830–1843. [CrossRef]
  18. Lu A, Chen W-T, Dalby M, et al. Human microglial transitions at the Aβ–tau inflection point associate with divergent pathways to dementia and resilience. Nature Medicine. 2026;32:2047–2059. [CrossRef]
  19. Nieuwenhuis S, Forstmann BU, Wagenmakers E-J. Erroneous analyses of interactions in neuroscience: a problem of significance. Nature Neuroscience. 2011;14(9):1105–1107. [CrossRef]
  20. Dimitrov D, Türei D, Garrido-Rodriguez M, et al. Comparison of methods and resources for cell-cell communication inference from single-cell RNA-Seq data. Nature Communications. 2022;13:3224. [CrossRef]
  21. Gosselin D, Skola D, Coufal NG, et al. An environment-dependent transcriptional network specifies human microglia identity. Science. 2017;356(6344):eaal3222. [CrossRef]
  22. Mondo E, et al. A developmental analysis of juxtavascular microglia dynamics and interactions with the vasculature. Journal of Neuroscience. 2020;40(34):6503–6521.
  23. Mayer MG, Fischer T. Microglia at the blood brain barrier in health and disease. Frontiers in Cellular Neuroscience. 2024;18:1360195.
  24. Bijnen M, et al. Brain macrophages in vascular health and dysfunction. Trends in Immunology. 2025;46(1):46–60.
  25. Gyoneva S, Traynelis SF. Norepinephrine modulates the motility of resting and activated microglia via different adrenergic receptors. Journal of Biological Chemistry. 2013;288(21):15291–15302.
  26. Schneble N, et al. Phosphoinositide 3-kinase γ ties chemoattractant- and adrenergic control of microglial motility. Molecular and Cellular Neuroscience. 2017;78:1–8. [CrossRef]
  27. Stowell R, Wang KH. Dopaminergic signaling regulates microglial surveillance and adolescent plasticity in the mouse frontal cortex. Nature Communications. 2025;16:7974. [CrossRef]
  28. Faust TE, Lee YH, O’Connor CD, et al. Microglia-astrocyte crosstalk regulates synapse remodeling via Wnt signaling. Cell. 2025;188(19):5212–5230.e21. [CrossRef]
  29. Luan P, Jia J, Chen Y, et al. A reciprocal glial circuit calibrates injury information and governs the tissue response balance. Neuron. 2026. Advance online publication. [CrossRef]
  30. Devlin BA, Nguyen DM, Ribeiro D, et al. Excitatory-neuron-derived interleukin-34 supports cortical developmental microglia function. Immunity. 2025;58:1948-1965.e6. [CrossRef]
  31. Pilat DJ, Le H, Prokopenko D, et al. The gain-of-function TREM2-T96K mutation increases risk for Alzheimer’s disease by impairing microglial function. Neuron. 2026;114(1):46–66.e13. [CrossRef]
  32. Davalos D, et al. ATP mediates rapid microglial response to local brain injury in vivo. Nature Neuroscience. 2005;8(6):752–758.
  33. Haynes SE, et al. The P2Y12 receptor regulates microglial activation by extracellular nucleotides. Nature Neuroscience. 2006;9(12):1512–1519.
  34. Abiega O, et al. Neuronal hyperactivity disturbs ATP microgradients, impairs microglial motility, and reduces phagocytic receptor expression triggering apoptosis/microglial phagocytosis uncoupling. PLOS Biology. 2016;14(5):e1002466.
  35. Béchade C, et al. The serotonin 2B receptor is required in neonatal microglia to limit neuroinflammation and sickness behavior in adulthood. Glia. 2021;69(3):638–654.
  36. Llovera G, Heindl S, Varga DP, et al. Blocking microglial reactivity via purinergic receptors prevents subacute cognitive deficits after TIA. EMBO Molecular Medicine. 2026;18:1150–1173. [CrossRef]
  37. Nakamura Y, Matsuda R, Kuribayashi S, Takemura M, Hisaoka-Nakashima K, Morioka N. Microglial α7-nicotinic acetylcholine receptors are expressed in mitochondria rather than on the plasma membrane: roles in mitochondrial function. Journal of Neurochemistry. 2025;169:e70139. [CrossRef]
  38. Bedolla A, Wegman E, Weed M, et al. Adult microglial TGF-β1 is required for microglia homeostasis via an autocrine mechanism to maintain cognitive function in mice. Nature Communications. 2024;15:5306. [CrossRef]
  39. Fixsen BR, Han CZ, Zhou Y, et al. SALL1 enforces microglia-specific DNA binding and function of SMADs to establish microglia identity. Nature Immunology. 2023;24(7):1188–1199. [CrossRef]
  40. Pósfai B, Szabadits E, Cserép C, et al. P2Y12 receptor function governs microglial surveillance and cell-cell interactions in the cerebral cortex. Glia. 2026;74:e70109. [CrossRef]
  41. Butler CA, et al. Microglial phagocytosis of neurons in neurodegeneration, and its regulation. Journal of Neurochemistry. 2021;158(3):621–639.
  42. Lemke G. How macrophages deal with death. Nature Reviews Immunology. 2019;19(9):539–549.
  43. Dissing-Olesen L, et al. Activation of neuronal NMDA receptors triggers transient ATP-mediated microglial process outgrowth. Journal of Neuroscience. 2014;34(32):10511–10527.
  44. Illes P, et al. Regulation of microglial functions by purinergic mechanisms in the healthy and diseased CNS. Cells. 2020;9(5):1108.
  45. Zarrinmayeh H, Territo PR. Purinergic receptors of the central nervous system: biology, PET ligands, and their applications. Molecular Imaging. 2020;19:1536012120927609.
  46. Orr AG, et al. Adenosine A2A receptor mediates microglial process retraction. Nature Neuroscience. 2009;12(7):872–878.
  47. Ohsawa K, et al. Adenosine A3 receptor is involved in ADP-induced microglial process extension and migration. Journal of Neurochemistry. 2012;121(2):217–227.
  48. Badimon A, et al. Negative feedback control of neuronal activity by microglia. Nature. 2020;586(7829):417–423.
  49. Fields RD, Burnstock G. Purinergic signalling in neuron–glia interactions. Nature Reviews Neuroscience. 2006;7(6):423–436.
  50. Sipe GO, Lowery RL, Tremblay M-È, Kelly EA, Lamantia CE, Majewska AK. Microglial P2Y12 is necessary for synaptic plasticity in mouse visual cortex. Nature Communications. 2016;7:10905. [CrossRef]
  51. Blume ZI, et al. Microglia in the developing retina couple phagocytosis with the progression of apoptosis via P2RY12 signaling. Developmental Dynamics. 2020;249(6):723–740.
  52. Bisht K, et al. Capillary-associated microglia regulate vascular structure and function through PANX1-P2RY12 coupling in mice. Nature Communications. 2021;12(1):5289.
  53. Lou N, et al. Purinergic receptor P2RY12-dependent microglial closure of the injured blood–brain barrier. Proceedings of the National Academy of Sciences. 2016;113(4):1074–1079.
  54. Färber K, Pannasch U, Kettenmann H. Dopamine and noradrenaline control distinct functions in rodent microglial cells. Molecular and Cellular Neuroscience. 2005;29(1):128–138.
  55. Fan Y, et al. Differential regulation of adhesion and phagocytosis of resting and activated microglia by dopamine. Frontiers in Cellular Neuroscience. 2018;12:309.
  56. Huck JH, et al. De novo expression of dopamine D2 receptors on microglia after stroke. Journal of Cerebral Blood Flow & Metabolism. 2015;35(11):1804–1811.
  57. Király B, Császár E, Balázsfi D, et al. Microglia modulate information processing in the mouse barrel cortex. Journal of Neuroscience. 2026;46(12):e0941252026. [CrossRef]
  58. Lehrman EK, et al. CD47 protects synapses from excess microglia-mediated pruning during development. Neuron. 2018;100(1):120–134.
  59. Ding X, et al. Loss of microglial SIRPα promotes synaptic pruning in preclinical models of neurodegeneration. Nature Communications. 2021;12(1):2030.
  60. Barclay AN, Van den Berg TK. The interaction between signal regulatory protein α (SIRPα) and CD47: structure, function, and therapeutic target. Annual Review of Immunology. 2014;32(1):25–50.
  61. Logtenberg ME, Scheeren FA, Schumacher TN. The CD47-SIRPα immune checkpoint. Immunity. 2020;52(5):742–752.
  62. Stevens B, et al. The classical complement cascade mediates CNS synapse elimination. Cell. 2007;131(6):1164–1178.
  63. Hong S, et al. Complement and microglia mediate early synapse loss in Alzheimer mouse models. Science. 2016;352(6286):712–716.
  64. Cai L, Zhang Y, Li J, et al. Complement receptor 3-dependent microglial synapse elimination drives Parkinson’s disease pathogenesis in systemic inflammation. Cell Death & Disease. 2026;17:319. [CrossRef]
  65. Coulthard LG, Hawksworth OA, Woodruff TM. Complement: the emerging architect of the developing brain. Trends in Neurosciences. 2018;41(6):373–384.
  66. Tenner AJ, Stevens B, Woodruff TM. New tricks for an ancient system: Physiological and pathological roles of complement in the CNS. Molecular Immunology. 2018;102:3–13.
  67. Lian H, et al. NFκB-activated astroglial release of complement C3 compromises neuronal morphology and function associated with Alzheimer’s disease. Neuron. 2015;85(1):101–115.
  68. Lian H, et al. Astrocyte-microglia cross talk through complement activation modulates amyloid pathology in mouse models of Alzheimer’s disease. Journal of Neuroscience. 2016;36(2):577–589.
  69. Lv J, Yao L, Tang L, et al. C1q-mediated synapse loss by microglial phagocytosis is associated with postoperative neurocognitive disorder in mice. British Journal of Anaesthesia. 2026;136(6):1855–1869. [CrossRef]
  70. Shi Q, et al. Complement C3 deficiency protects against neurodegeneration in aged plaque-rich APP/PS1 mice. Science Translational Medicine. 2017;9(392):eaaf6295.
  71. Schartz ND, Tenner AJ. The good, the bad, and the opportunities of the complement system in neurodegenerative disease. Journal of Neuroinflammation. 2020;17(1):354.
  72. Matera A, Compagnion A-C, Pedicone C, et al. Microglial lipid phosphatase SHIP1 limits complement-mediated synaptic pruning in the healthy developing hippocampus. Immunity. 2025;58(1):197–217.e13. [CrossRef]
  73. Fourgeaud L, et al. TAM receptors regulate multiple features of microglial physiology. Nature. 2016;532(7598):240–244.
  74. Li T, et al. A splicing isoform of GPR56 mediates microglial synaptic refinement via phosphatidylserine binding. EMBO Journal. 2020;39(16):e104136.
  75. Lemke G, Rothlin CV. Immunobiology of the TAM receptors. Nature Reviews Immunology. 2008;8(5):327–336.
  76. Fricker M, et al. MFG-E8 mediates primary phagocytosis of viable neurons during neuroinflammation. Journal of Neuroscience. 2012;32(8):2657–2666.
  77. Akakura S, et al. The opsonin MFG-E8 is a ligand for the αvβ5 integrin and triggers DOCK180-dependent Rac1 activation for the phagocytosis of apoptotic cells. Experimental Cell Research. 2004;292(2):403–416.
  78. Albert ML, Kim JI, Birge RB. αvβ5 integrin recruits the CrkII–Dock180–Rac1 complex for phagocytosis of apoptotic cells. Nature Cell Biology. 2000;2(12):899–905.
  79. Park J, et al. Microglial MERTK eliminates phosphatidylserine-displaying inhibitory post-synapses. EMBO Journal. 2021;40(15):e107121.
  80. Rueda-Carrasco J, Sokolova D, Lee S-E, et al. Microglia-synapse engulfment via PtdSer-TREM2 ameliorates neuronal hyperactivity in Alzheimer’s disease models. EMBO Journal. 2023;42:e113246. Author Correction. EMBO Journal. 2024. doi:10.1038/s44318-024-00159-5. [CrossRef]
  81. Jiang D, et al. Neuronal signal-regulatory protein α drives microglial phagocytosis by limiting microglial interaction with CD47 in the retina. Immunity. 2022;55(12):2318–2335.
  82. Han MH, et al. Janus-like opposing roles of CD47 in autoimmune brain inflammation in humans and mice. Journal of Experimental Medicine. 2012;209(7):1325–1334.
  83. Gheibihayat SM, et al. CD47 in the brain and neurodegeneration: an update on the role in neuroinflammatory pathways. Molecules. 2021;26(13):3943.
  84. Pluvinage JV, Haney MS, Smith BAH, et al. CD22 blockade restores homeostatic microglial phagocytosis in ageing brains. Nature. 2019;568(7751):187–192. [CrossRef]
  85. Casanova PV, Freitag-Berenguel I, Saavedra-Lopez E, et al. Microglial low-affinity FcγR mediates the phagocytic elimination of dopaminergic neurons in Parkinson’s disease degeneration. npj Parkinson’s Disease. 2026;12:35. [CrossRef]
  86. Weinhard L, di Bartolomei G, Bolasco G, et al. Microglia remodel synapses by presynaptic trogocytosis and spine head filopodia induction. Nature Communications. 2018;9:1228. [CrossRef]
  87. McCallum S, Suresh KB, Islam TS, et al. Lesion-remote astrocytes govern microglia-mediated white matter repair. Nature. 2026;649:959–970. [CrossRef]
  88. Frosch M, Shimizu T, Wogram E, et al. Microglia–neuron crosstalk through Hex–GM2–MGL2 maintains brain homeostasis. Nature. 2025;646(8086):913–924. [CrossRef]
  89. Lin H, et al. Discovery of a cytokine and its receptor by functional screening of the extracellular proteome. Science. 2008;320(5877):807–811.
  90. Bohlen CJ, et al. Diverse requirements for microglial survival, specification, and function revealed by defined-medium cultures. Neuron. 2017;94(4):759–773.
  91. Wei S, et al. Functional overlap but differential expression of CSF-1 and IL-34 in their CSF-1 receptor-mediated regulation of myeloid cells. Journal of Leukocyte Biology. 2010;88(3):495–505.
  92. Muñoz-Garcia J, et al. The twin cytokines interleukin-34 and CSF-1: masterful conductors of macrophage homeostasis. Theranostics. 2021;11(4):1568–1593. [CrossRef]
  93. Norden DM, et al. TGFβ produced by IL-10 redirected astrocytes attenuates microglial activation. Glia. 2014;62(6):881–895.
  94. Villa A, Gelosa P, Castiglioni L, et al. Sex-specific features of microglia from adult mice. Cell Reports. 2018;23(12):3501–3511. [CrossRef]
  95. Sala Frigerio C, Wolfs L, Fattorelli N, et al. The major risk factors for Alzheimer’s disease: age, sex, and genes modulate the microglia response to Aβ plaques. Cell Reports. 2019;27(4):1293–1306.e6. [CrossRef]
  96. Wu J, Chen X, Zhang J, et al. Human microglia in brain assembloids display region-specific diversity and respond to hyperexcitable neurons carrying SCN2A mutation. Science Advances. 2026;12(8):eady2977. [CrossRef]
  97. Heneka MT, Kummer MP, Stutz A, et al. NLRP3 is activated in Alzheimer’s disease and contributes to pathology in APP/PS1 mice. Nature. 2013;493(7434):674–678. [CrossRef]
  98. Ising C, Venegas C, Zhang S, et al. NLRP3 inflammasome activation drives tau pathology. Nature. 2019;575(7784):669–673. [CrossRef]
  99. Gulen MF, Samson N, Keller A, et al. cGAS–STING drives ageing-related inflammation and neurodegeneration. Nature. 2023;620:374–380. [CrossRef]
  100. Udeochu JC, Amin S, Huang Y, et al. Tau activation of microglial cGAS–IFN reduces MEF2C-mediated cognitive resilience. Nature Neuroscience. 2023;26(5):737–750. [CrossRef]
  101. Gustin A, Kirchmeyer M, Koncina E, et al. NLRP3 inflammasome is expressed and functional in mouse brain microglia but not in astrocytes. PLOS ONE. 2015;10(6):e0130624. [CrossRef]
  102. Sun L, Wu J, Du F, Chen X, Chen ZJ. Cyclic GMP-AMP synthase is a cytosolic DNA sensor that activates the type I interferon pathway. Science. 2013;339(6121):786-791. [CrossRef]
  103. Venegas C, Kumar S, Franklin BS, et al. Microglia-derived ASC specks cross-seed amyloid-β in Alzheimer’s disease. Nature. 2017;552(7685):355–361. [CrossRef]
  104. He S, Li X, Mittra N, et al. Microglial cGAS deletion preserves intercellular communication and alleviates amyloid-β-induced pathogenesis of Alzheimer’s disease. Advanced Science. 2025;12(12):e2410910. [CrossRef]
  105. Chung S, Jeong J-H, Park J-C, et al. Blockade of STING activation alleviates microglial dysfunction and a broad spectrum of Alzheimer’s disease pathologies. Experimental & Molecular Medicine. 2024;56:1936–1951. [CrossRef]
  106. Sulka KB, Carroll KA, Sawden M, et al. Microglial STING is a central safeguard against neurological decline with age. Cell Reports. 2025;44(6):115749. [CrossRef]
  107. Carnevale LN, Banerjee P, Zhang X, et al. Redox regulation of neuroinflammatory pathways contributes to damage in Alzheimer’s disease brain. Cell Chemical Biology. 2026;33(8):1104-1116.e8. [CrossRef]
  108. Naguib S, Lopez-Lee C, Torres ER, et al. The R136S mutation in the APOE3 gene confers resilience against tau pathology via inhibition of the cGAS-STING-IFN pathway. Immunity. 2025;58(8):1931-1947.e9. [CrossRef]
  109. Cao T, Hemati-Gourabi M, Liu Y, et al. Astrocyte-microglia crosstalk via CSF1 and IFN-β promotes central nervous system repair. Cell Reports. 2026;45:117418. [CrossRef]
  110. Vainchtein ID, Chin G, Cho FS, et al. Astrocyte-derived interleukin-33 promotes microglial synapse engulfment and neural circuit development. Science. 2018;359(6381):1269–1273. [CrossRef]
  111. Ramirez JJ, Hardin EJ, Sakers K, et al. Astrocyte-microglia crosstalk through Hevin and Toll-like receptor signaling controls developmental thalamocortical synapse refinement. Neuron. 2026;114(8):1433–1453.e11. [CrossRef]
  112. Liddelow SA, Guttenplan KA, Clarke LE, et al. Neurotoxic reactive astrocytes are induced by activated microglia. Nature. 2017;541(7638):481–487. [CrossRef]
  113. Rothhammer V, Borucki DM, Tjon EC, et al. Microglial control of astrocytes in response to microbial metabolites. Nature. 2018;557(7707):724–728. [CrossRef]
  114. Adams RA, et al. The fibrin-derived γ377-395 peptide inhibits microglia activation and suppresses relapsing paralysis in central nervous system autoimmune disease. Journal of Experimental Medicine. 2007;204(3):571–582.
  115. Lee JY, Choi HY, Yune TY. MMP-3 secreted from endothelial cells of blood vessels after spinal cord injury activates microglia, leading to oligodendrocyte cell death. Neurobiology of Disease. 2015;82:141–151.
  116. Ryu JK, Cho T, Choi HB, Wang YT, McLarnon JG. Microglial VEGF receptor response is an integral chemotactic component in Alzheimer’s disease pathology. Journal of Neuroscience. 2009;29(1):3–13. [CrossRef]
  117. Xu Z, et al. Vascular endothelial growth factor is neuroprotective against ischemic brain injury by inhibiting scavenger receptor A expression on microglia. Journal of Neurochemistry. 2017;142(5):700–709.
  118. Jolivel V, et al. Perivascular microglia promote blood vessel disintegration in the ischemic penumbra. Acta Neuropathologica. 2015;129:279–295.
  119. Albertini G, Etienne F, Roumier A. Regulation of microglia by neuromodulators: Modulations in major and minor modes. Neuroscience Letters. 2020;733:135000.
  120. Pocock JM, Kettenmann H. Neurotransmitter receptors on microglia. Trends in Neurosciences. 2007;30(10):527–535.
  121. Ma C, et al. Microglia regulate sleep through calcium-dependent modulation of norepinephrine transmission. Nature Neuroscience. 2024;27(2):249–258.
  122. Bharani KL, et al. A noradrenergic lesion aggravates the effects of systemic inflammation on the hippocampus of aged rats. PLOS ONE. 2017;12(12):e0189821.
  123. Yao N, et al. Lesion of the locus coeruleus aggravates dopaminergic neuron degeneration by modulating microglial function in mouse models of Parkinson’s disease. Brain Research. 2015;1625:255–274.
  124. Heneka MT, et al. Locus ceruleus controls Alzheimer’s disease pathology by modulating microglial functions through norepinephrine. Proceedings of the National Academy of Sciences. 2010;107(13):6058–6063.
  125. Favuzzi E, Huang S, Saldi G, et al. GABA-receptive microglia selectively sculpt developing inhibitory circuits. Cell. 2021;184(15):4048–4063.e32. [CrossRef]
  126. Venero JL, et al. DCG-IV but not other group-II metabotropic receptor agonists induces microglial BDNF mRNA expression in the rat striatum. Correlation with neuronal injury. Neuroscience. 2002;113(4):857–869.
  127. Geurts JJG, et al. Altered expression patterns of group I and II metabotropic glutamate receptors in multiple sclerosis. Brain. 2003;126(8):1755–1766.
  128. Zhang YN, et al. Metabotropic glutamate receptor 5 inhibits α-synuclein-induced microglia inflammation to protect from neurotoxicity in Parkinson’s disease. Journal of Neuroinflammation. 2021;18:23. [CrossRef]
  129. Costa A, et al. Deletion of muscarinic acetylcholine receptor 3 in microglia impacts brain ischemic injury. Brain, Behavior, and Immunity. 2021;91:89–104.
  130. Cortes M, et al. α7 nicotinic acetylcholine receptor signaling modulates the inflammatory phenotype of fetal brain microglia: first evidence of interference by iron homeostasis. Scientific Reports. 2017;7(1):10645.
  131. Iida T, et al. Histamine H3 receptor in primary mouse microglia inhibits chemotaxis, phagocytosis, and cytokine secretion. Glia. 2015;63(7):1213–1225.
  132. Krabbe G, et al. Activation of serotonin receptors promotes microglial injury-induced motility but attenuates phagocytic activity. Brain, Behavior, and Immunity. 2012;26(3):419–428.
  133. El Oussini H, et al. Serotonin 2B receptor slows disease progression and prevents degeneration of spinal cord mononuclear phagocytes in amyotrophic lateral sclerosis. Acta Neuropathologica. 2016;131:465–480.
  134. Haas LT, et al. Silent allosteric modulation of mGluR5 maintains glutamate signaling while rescuing Alzheimer’s mouse phenotypes. Cell Reports. 2017;20(1):76–88.
  135. Amaral CLD, et al. Activation of the α7 nicotinic acetylcholine receptor prevents against microglial-induced inflammation and insulin resistance in hypothalamic neuronal cells. Cells. 2022;11(14):2195.
  136. Mencel M, Nash M, Jacobson C. Neuregulin upregulates microglial α7 nicotinic acetylcholine receptor expression in immortalized cell lines: implications for regulating neuroinflammation. PLOS ONE. 2013;8(7):e70338.
  137. Dong H, et al. Histamine induces upregulated expression of histamine receptors and increases release of inflammatory mediators from microglia. Molecular Neurobiology. 2014;49:1487–1500.
  138. Pannell M, et al. The subpopulation of microglia sensitive to neurotransmitters/neurohormones is modulated by stimulation with LPS, interferon-γ, and IL-4. Glia. 2014;62(5):667–679.
  139. Rocha SM, et al. Histamine induces microglia activation and dopaminergic neuronal toxicity via H1 receptor activation. Journal of Neuroinflammation. 2016;13:137. [CrossRef]
  140. Ferreira R, et al. Histamine modulates microglia function. Journal of Neuroinflammation. 2012;9:90. [CrossRef]
  141. Du O, Wu C, Yang Y-X, et al. High mobility group box 1, a novel serotonin receptor-7 negative modulator, contributes to M2 microglial ferroptosis and neuroinflammation in post-stroke depression. Free Radical Biology and Medicine. 2025;237:666–683. [CrossRef]
  142. Pedrosa MA, Mincheva G, Martinez-Garcia M, et al. Golexanolone affords sustained improvement of Parkinson’s symptoms in rats by reducing microglia activation that restores the glutamate-GABA-dopamine pathway. Neuropharmacology. 2026;283:110759. [CrossRef]
  143. Le LHD, Feidler AM, Calcines Rodriguez L, et al. Noradrenergic signaling controls Alzheimer’s disease pathology via activation of microglial β2 adrenergic receptors. Brain, Behavior, and Immunity. 2025;128:307-322. [CrossRef]
  144. Zu J, Li C, Cui M, et al. Pioglitazone attenuates complement-mediated microglial synaptic engulfment in an Alzheimer’s disease model. Brain. 2026;149:668-679. [CrossRef]
  145. Lazic SE. The problem of pseudoreplication in neuroscientific studies: is it affecting your analysis? BMC Neuroscience. 2010;11:5. [CrossRef]
  146. Aarts E, Verhage M, Veenvliet JV, Dolan CV, van der Sluis S. A solution to dependency: using multilevel analysis to accommodate nested data. Nature Neuroscience. 2014;17(4):491–496. [CrossRef]
Preprints 232093 g002
Figure 2. Established sequential signalling and the experiment required to test durable state change. Canonical NLRP3 activation separates priming from a later activating input [101]. By contrast, an altered state after cargo exposure does not establish cellular memory unless the same cell shows a changed functional response to a controlled later cue. Reciprocal cue order, single-cue controls, washout verification and predefined same-cell endpoints are required [14,15].
Figure 2. Established sequential signalling and the experiment required to test durable state change. Canonical NLRP3 activation separates priming from a later activating input [101]. By contrast, an altered state after cargo exposure does not establish cellular memory unless the same cell shows a changed functional response to a controlled later cue. Reciprocal cue order, single-cue controls, washout verification and predefined same-cell endpoints are required [14,15].
Preprints 232093 g002
Table 1. Selected neurotransmitter and neuromodulator effects on microglia.
Table 1. Selected neurotransmitter and neuromodulator effects on microglia.
Signal Main decision Best-supported finding Important limitation
Norepinephrine Positioning/contact β2-adrenergic signalling opposes ATP-directed process extension in acute preparations [25,26]. Acute slice/culture interaction; locus-coeruleus manipulations also change neuronal activity, vascular tone, arousal and sleep [121,122,123,124].
Dopamine Positioning/circuit Dopamine-receptor signalling and P2Y12 are both required for increased surveillance and bouton formation in adolescent mesofrontal cortex [27]. Direct biochemical convergence and generalization across states remain unresolved [27,54,55,56].
GABA Target/circuit GABA-receptive microglia participate in defined developmental inhibitory-circuit programmes [96,125]. Evidence is model-, region- and stage-specific; it does not define a universal microglial GABA response.
Glutamate Positioning Neuronal NMDA-receptor activation can evoke ATP-dependent microglial process extension [43]. Direct microglial glutamate-receptor expression and function vary substantially by preparation [120,126,127,128].
Acetylcholine Positioning/metabolism M3 signalling affects injury responses; isolated microglia show predominantly mitochondrial α7 localization and bioenergetic responses [37,129,130]. Surface-channel mechanisms and cell attribution differ among culture, isolated-cell and in-vivo studies.
Histamine Positioning/uptake H3 activation suppresses ATP-evoked calcium, chemotaxis, phagocytosis and cytokine release in primary microglia [131]. The direct interaction is established in primary culture; in-vivo relevance and other receptor subtypes remain less resolved.
Serotonin Positioning/uptake/state Serotonergic effects on motility, phagocytosis and later inflammatory state depend on receptor subtype and developmental timing [35,132,133]. Acute pharmacology does not generalize across age, receptor subtype, disease or cellular compartment.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.