Phagocytosis and neuroinflammation: orchestrating CNS homeostasis, repair, and the resolution of inflammation

Abstract

The ingestion of foreign material – phagocytosis – is a fundamental feature shared across phyla, from single-celled amoeba to more complex mammalian cells. Phagocytosis is a crucial process for maintaining homeostasis in body systems, including in the central nervous system (CNS). In this review, we first explore how phagocytic cells maintain CNS homeostasis by removing excess synapses, apoptotic cells, and debris. Next, we discuss the dual role of phagocytosis in CNS pathologies, including multiple sclerosis, aging and Alzheimer’s disease, traumatic CNS injury, and ischemic stroke. During CNS pathology, phagocytosis aids debris removal and tissue repair, yet also contributes to damage through engulfment of viable synapses and cells and via release of cytotoxic mediators. Finally, we highlight current clinically-relevant approaches and consider future directions for leveraging phagocytosis to enhance CNS repair and improve neurologic outcomes.

Phagocytosis: A fundamental process for CNS homeostasis and repair

Phagocytosis – the act of ingesting extracellular material – is a fundamental cellular activity, as suggested by its evolutionary conservation across phyla. Indeed, studies from the 18th century hinted at the ability of single-celled organisms to ingest small objects, and, in the 19th century, studies in sea star larvae showed that a rose thorn puncture elicited massive accumulation of cells that “devoured” the foreign material [1]. These pioneering studies helped unveil the importance of phagocytosis in immune defense, fueling further research into how phagocytic cells aid healthy tissue dynamics, immune defense, tissue repair, and inflammatory resolution.

In the central nervous system (CNS), phagocytosis has crucial roles in shaping development and maintaining health. The main professional phagocytes of the CNS, microglia and astrocytes, phagocytose excess synapses to ensure neural circuit formation and maintenance, and remove cells undergoing apoptosis (i.e., programmed cell death) to prevent accumulation of potentially toxic cellular debris and molecular danger signals.

After CNS damage or disorder, phagocytosis has complex roles in regulating repair. Successful engulfment of apoptotic cells (efferocytosis) initiates anti-inflammatory signaling [2], whereas failed phagocytosis drives secondary necrosis and release of damage-associated molecular patterns (DAMPs) that amplify inflammation [3]. Further, phagocytes that process different types of cargo develop distinct inflammatory profiles [2]. Thus, enhancing phagocytic uptake of harmful debris, while minimizing damage to viable cells, could ameliorate CNS pathology.

In this review, we aim to provide an integrative framework on how phagocytic pathways coordinate debris clearance and inflammatory regulation across CNS disorders. We will discuss the roles of phagocytosis in the rodent and human CNS. First, we will introduce the main CNS phagocytes – microglia and astrocytes – which orchestrate CNS debris clearance. Next, we will examine the proactive role of phagocytes in maintaining healthy CNS homeostasis and baseline neuroimmunity. We will discuss how phagocytosis helps and hinders repair in various neurologic disorders. Finally, we will explore modulating phagocytosis as an emerging translational target for ameliorating CNS disease and disorders.

Phagocytic cells of the CNS: Microglia, astrocytes, and other cells

The CNS is a highly heterogenous environment with region-specific demands that vary over the course of the lifespan; therefore, CNS cells must adapt rapidly to changing demands and injury.

While neurons are the principal cells of the CNS, numerous other cell types support the homeostatic maintenance of the CNS. One major activity of these supportive glial cells – and infiltrating peripheral immune cells during pathology – is refinement and cleanup of the CNS environment through the process of phagocytosis (Figure 1, Key Figure).

Figure 1. Phagocytosis and immunity in the central nervous system.

The figure depicts areas in the healthy CNS (left half of brain) and pathological CNS (right half of brain). The ‘lesion or plaque’ area represents pathology due to CNS injury, stroke, multiple sclerosis, or age-related neurodegeneration (e.g., Alzheimer’s disease). a. Phagocytosis and immunity in the healthy CNS. Microglia are the major resident CNS immune cell, while astrocytes and perivascular macrophages also contribute to neuroimmunity. These immune-competent cells maintain CNS homeostasis by pruning excess synapses; removing cells undergoing programmed cell death (apoptosis); surveying for and responding to damage or infection; and maintaining a healthy low baseline inflammatory tone. b. Roles of phagocytosis and immunity in the lesion or plaque core. Phagocytosis effector cells in the pathologic CNS – including microglia, astrocytes, and/or infiltrating immune cells – respond by engulfing apoptotic and myelin debris, thereby limiting necrosis and secondary damage. During aging and Alzheimer’s disease, microglia remove amyloid-β plaques. Cells in the lesion/plaque epicenter develop morphological and functional features of reactivity, and peripheral immune cells are recruited to major sites of damage and inflammation. Neuroimmune cells secrete inflammatory mediators that can have reparative and cytotoxic effects on spared tissue. c. In peri-lesion, penumbral, or peri-plaque tissue, phagocytosis enables removal of apoptotic debris, but also removes injured-but-viable cells and synapses. Neuroimmune cells in the peri-pathology region develop reactive phenotypes and release immunomodulatory mediators. Overall, targeted activation of phagocytosis to remove damaged cells, while preserving viable cells and tissue and promoting secretion of reparative mediators, could reduce spreading progression of pathology and enhance neurologic recovery. Figure created in part with BioRender.com.

Phagocytic efficiency in the CNS depends on cell type-specific factors. First, receptor repertoire and expression determine which debris can be recognized [4,5]. Second, distinct motility between cell types defines debris contact – microglia exhibit highly motile processes, whereas astrocytes have less dynamic processes but extensive arbors [6]. Third, anatomical location shapes phagocytic roles: microglia are distributed throughout the parenchyma, astrocytes contact parenchymal and perivascular spaces, and infiltrating monocytes/macrophages localize to lesions and perivascular areas [7]. Together, these factors determine each cell type’s contribution to debris clearance.

Microglia, the primary innate immune cell of the CNS, constantly survey their environments and adopt localized responses [8]. Microglial transcriptional state and functions vary across the lifespan and in response to local physiological demand [9–11] (see Box 1). Microglia engage in myriad activities, including synapse remodeling, neurogenesis, myelination, BBB integrity, vasculogenesis, tissue repair and regeneration, and immune regulation. In early life, microglia contribute to multiple stages of neural circuit formation, including neurogenesis and axon guidance [12,13]. They refine neural circuitry into adolescence and adulthood via their participation in developmental apoptotic cell removal [14] and synaptic pruning [15,16]. As animals enter adulthood, baseline microglia functions skew more towards activities that promote CNS repair and regeneration [17]. Microglia have various receptors that facilitate cell-cell communication and responses to diffusible signals, allowing microglia to undergo rapid morphological and functional shifts in response to CNS perturbation.

Box 1. Transcriptional regulation of microglial cell state.

Microglia, the primary phagocytes of the CNS, exhibit striking transcriptional plasticity in response to pathology. Other CNS phagocytes, particularly astrocytes, exhibit distinct transcriptional responses to pathology that differ from microglial DAM states (see [19,20,26]). This box provides an overview of microglial cell signatures and how these phenotypes relate to phagocytic capacity.

Microglia during early development display phenotypic heterogeneity, including a subset that exhibits proliferative signatures and a progressively more prominent subset of adult-like microglia expressing cell type-specific homeostatic markers [151]. In the healthy adult mouse, microglia cell signatures are relatively consistent across the CNS [79,93] – most adult microglia express canonical markers including Sall1, Cx3cr1, P2yr12, and Selplg. Although the core homeostatic microglial phenotype is conserved in microglia throughout the CNS, there is some regional variability in microglial transcriptional and proteomic signatures in mice and humans [152–154]. Overall, adult homeostatic microglia exhibit a consistent and stable phenotype with expression of a suite of prototypical markers.

During pathological conditions, microglia develop a “disease-associated microglia” (DAM) cell signature (see [155]). The DAM phenotype includes downregulation of homeostatic microglial genes and de novo upregulation of key immune response genes, including Spp1, Itgax, Cd9, Csf1, and Apoe [79,93]. Although the DAM signature shares common gene expression changes across CNS disorders, microglia also develop distinct, disorder-specific phenotypes [155]. Additional microglial subpopulations commonly initiated by pathology include antigen-presenting response, cytokine response, and interferon response microglia [156,157].

These transcriptional changes directly regulate phagocytic capacity and efficiency: DAM microglia upregulate phagocytic machinery and enhance their ability to process engulfed material [33,150]. Indeed, DAM-like microglia upregulate genes related to phagocytic uptake (e.g., Axl, MerTK), lysosomal function (e.g., Cd63, Atp6v1a) and lipid metabolism (e.g., Plin3, Acly) [33,150], which mediate effective debris clearance.

Progress in human microglia research reveals phenotypic transitions similar to mouse microglial cell state during pathology [158]. Human microglia show dynamic regulation from early to late-stage AD pathology [159,160]. In addition, innovative use of iPSCs exposed to disease-relevant brain substrates effectively phenocopies in vivo microglial cell states [161]. Thus, future research could integrate multimodal data from mouse models, human iPSCs, and human tissue to provide insights into mechanisms and treatment of CNS pathologies.

A key unresolved question is whether transcriptional shifts drive enhanced phagocytosis, or whether phagocytic activity itself initiates these gene expression changes. Future research using iPSC-derived microglia and single-cell approaches will help dissect these causal relationships [162]. Ultimately, synthesizing complex transcriptomic and proteomic data will help identify promising phagocytic pathways in microglia that can be targeted therapeutically.

Following detection of an inflammatory signal, microglia are directed to sites of inflammatory activity with chemotactic signals, where they upregulate production of inflammatory mediators including opsonins and cytokines and engage in phagocytic activity. Importantly, the phagocytic activity of microglia was recognized in their initial characterization, receiving the description of “voracious monsters” by del Rio-Hortega [18]. However, controversy about whether microglia phagocytic activity is ultimately beneficial or damaging has persisted for years, and likely relates to the context and CNS demands. For example, in the healthy adult, a microglia state involving surveying with little inflammatory priming may be considered “homeostatic”; in contrast with aging, microglia with a similarly dampened inflammatory state may not meet CNS pathological demands, thereby worsening toxic debris accumulation.

In addition to microglia, astrocytes and peripheral immune cells perform phagocytic functions in brain. Astrocytes are dynamic glial cells that coordinate numerous activities including uptake of ions and neurotransmitters, metabolic support, release of trophic factors, and structural support of the blood brain barrier. Astrocytes also modulate synaptic transmission and plasticity, contributing to formation and elimination of synapses. During pathology, astrocytes develop a reactive cell state: reactive astrocytes newly secrete inflammatory mediators, form the glial scar to constrain damaging debris, and engage in phagocytosis [19–23]. Indeed, astrocytes can recapitulate or compensate for microglia activities when they are absent or dysfunctional: for example, in the microglia FIRE mouse which lacks microglia, the somewhat limited pathology is thought to relate to compensatory activity from astrocytes [6,24–26]. Many of the phagocytic targets of astrocytes can also be phagocytosed by microglia, leading to questions about how these distinct cell types coordinate their activities [27,28]. Additionally, infiltrating peripheral cells, such as macrophages, engage in phagocytic activity in the CNS. Peripheral immune cells tend to be more inflammatory and aggressive than CNS-resident cells, which can lead to pathology when they infiltrate the CNS milieu [7,29].

Receptors involved in CNS phagocytosis

Phagocytosis in the CNS is orchestrated by a diverse array of receptors that recognize “eat-me” signals on apoptotic cells and debris, while avoiding viable cells marked by “don’t eat-me” signals. These receptors are expressed on microglia, astrocytes, and infiltrating macrophages; their activation mediates engulfment and shapes inflammatory and reparative responses [30,31].

Phagocytic receptor expression shows notable cell-type enrichment. TREM2 and P2RY12 are microglia-enriched [32,33], whereas TAM receptors (MerTK, MEGF10) are expressed on both microglia and astrocytes [26,34]. Scavenger receptors also show broad expression: CD36 is prominent on microglia/macrophages [35], while CD163 and CD68 are commonly expressed across myeloid phagocytes [36,37]. Complement receptors are predominantly microglial [15], although astrocytes express complement components [19]. Astrocytes use the cholesterol transporter ABCA1 to facilitate lipid processing during debris clearance [20]. These cell type-specific patterns affect which cells clear specific debris.

Although comprehensive coverage of CNS phagocytic receptors is beyond the scope of this review, many receptors and signaling pathways have been implicated in phagocytosis (reviewed in [38,39]). Below, we highlight phagocytic receptor systems with particularly strong genetic and functional links to CNS pathology and repair (Figure 2).

Figure 2. Phagocytic receptor systems implicated in removal of apoptotic debris and Amyloid-β plaques in the CNS.

a. A summary of key phagocytic receptor systems implicated in removal of apoptotic cells and debris. Distinct molecules are increased on the surface of apoptotic debris and are identified by phagocytic receptors, including TREM2, TAM receptors (Tyro3, Axl, MerTK), and integrins (e.g., α_vβ₂, aka CR3). “Don’t eat-me” signals, such as CD47, inhibit uptake by phagocytes. Scavenger receptors clear debris, pathogens, and damage-associated molecules. CD33 is a Siglec family member that binds sialic acids and limits phagocytosis. b. Key receptors that directly bind and regulate removal of Aβ plaques. TREM2 and α_vβ₂ facilitate Aβ engulfment. In contrast, CD33 interferes with Aβ uptake via binding sialic acid residues within the plaques and likely by inhibiting the TREM2 pathway. iC3b: protein fragment in the complement system, oxLDL: oxidized low-density lipoprotein, DAMPs: damage-associated molecular patterns; Siglec: sialic-acid binding immunoglobulin-like lectin; CR3: complement receptor 3; TAM: Tyro3, Axl, Mer receptor family. Figure created in part with BioRender.com.

TREM2 and TYROBP

Triggering receptor expressed on myeloid cells 2 (TREM2) is a transmembrane receptor expressed predominantly on microglia. TREM2 forms a signaling complex with the adaptor protein TYROBP (also known as DAP12), initiating intracellular cascades that promote phagocytosis and modulate inflammation. In addition, extracellular proteases can cleave the ectodomain of TREM2, creating soluble TREM2 (sTREM2) that accumulates in the cerebrospinal fluid (CSF) when TREM2 is highly expressed or under inflammatory conditions [40].

TREM2 is essential for microglial uptake of apoptotic neurons and amyloid-β (Aβ) plaques, and its deficiency impairs debris clearance and worsens outcomes in models of stroke and neurodegeneration [41]. TREM2 also binds to nucleic acids and lipids released from damaged cells, suggesting it acts as a broad sensor of CNS injury.

TAM receptors: Tyro3, Axl, and MerTK

The TAM receptor family—Tyro3, Axl, and MerTK—are receptor tyrosine kinases that recognize an “eat-me” signal, externalized phosphatidylserine, on apoptotic cells via bridging ligands such as Gas6 and Protein S (reviewed in [42]). TAM receptors are expressed on microglia and astrocytes and are critical for phagocytosis and resolution of inflammation; TAM receptor deletion impairs debris clearance and exacerbates pathology across multiple CNS disorders [42,43].

Integrins and complement receptors

The integrin α_Mβ₂ (a.k.a. Cd11b/CD18, Mac-1, and complement receptor 3 [CR3]) is expressed predominantly on microglia and macrophages in the CNS. α_Mβ₂ links innate immune responses to complement signaling – α_Mβ₂ binds to the complement component C3b, which evokes synaptic pruning during development [15,44]. In Alzheimer’s disease (AD) models, α_Mβ₂ facilitates uptake of Aβ but also contributes to synapse loss and inflammation [45].

CD33 and Siglec family receptors

CD33 is a sialic acid-binding immunoglobulin-like lectin (Siglec-3) that negatively regulates phagocytosis. Expressed on microglia, CD33 binds sialylated glycans on cell surfaces and Aβ plaques. CD33 inhibits TREM2-mediated activation of SYK kinase [46], ultimately limiting debris uptake and promoting plaque persistence. CD33 polymorphisms are associated with AD risk, and its inhibition enhances microglial phagocytosis in vitro and in vivo. In mice, the homologous receptor Siglec-E similarly suppresses phagocytosis and oxidative burst, suggesting a conserved role for Siglecs in limiting neuroinflammation [47].

CD47 and SIRPα: “Don’t eat-me” signals

CD47 is a transmembrane protein that interacts with signal regulatory protein alpha (SIRPα) on phagocytes to inhibit engulfment. This “don’t eat-me” signal is critical for preserving viable cells during inflammation. In the CNS, CD47 expression is upregulated in stressed but salvageable neurons and glia, and its blockade enhances clearance of apoptotic cells and improves outcomes in models of stroke and hemorrhage [48]. However, indiscriminate inhibition of CD47SIRPα signaling may risk removal of viable cells, underscoring the need for spatial and temporal precision in therapeutic targeting.

Scavenger receptors: CD36

Scavenger receptors recognize a broad range of ligands, including oxidized lipids, apoptotic cells, and Aβ. CD36, a class B scavenger receptor, forms a receptor complex with integrins and CD47 to mediate uptake of debris including Aβ and myelin [35].

CNS disorders – role of phagocytosis in damage and repair

Under pathological conditions, phagocytic cells develop new roles and elevated importance in shaping disease trajectory. Whereas phagocytosis in the healthy CNS involves controlled, localized uptake of debris tagged for removal, phagocytic cells during pathology confront damaged cells in an inflammatory landscape. This inflammatory milieu alters phagocyte behavior: cytokines and DAMPs drive reactive states that enhance debris clearance capacity yet also increase cytotoxic mediator production and maladaptive engulfment of viable tissue [7,19,38].

Common themes across CNS pathologies include initial beneficial debris clearance that becomes overwhelmed, and the critical distinction between core pathological areas requiring aggressive clearance versus penumbral regions where phagocytosis may remove viable cells. A key example is the “foamy macrophage/microglia” phenotype: phagocytes overwhelmed by lipid-rich debris accumulate lipid droplets and oxidized lipids that can elicit inflammatory responses, yet also contain debris and prevent toxic lipid spread [49,50]. This highlights how challenges in intracellular processing elicit context-dependent cellular responses (see Box 2).

Box 2. Intracellular processing of phagocytic material.

Phagocytosis involves the initial engulfment of extracellular debris into a membrane-bound vacuole called a phagosome. This is only the first step of debris removal. For effective clearance to occur, the phagosome must mature and fuse with a lysosome, forming an acidic, enzyme-rich phagolysosome [163]. This compartment is the cell’s primary hub for digestion. Phagolysosomal function is regulated by coordinated molecular pathways: Rab GTPases orchestrate phagosome maturation and fusion [163]; PI3K-mTOR signaling links phagocytosis to cellular metabolism [164]; and NOX2-generated reactive oxygen species degrade cargo but can create toxic oxidized byproducts that impair lysosomal function when excessive [165].

Within the phagolysosome, cargo is degraded by lysosomal acid hydrolases, which break down myelin and apoptotic debris, recycle nutrients such as lipids and amino acids, and help resolve inflammation to maintain CNS homeostasis [166]. Successful phagolysosomal degradation is anti-inflammatory: lysosomal enzymes neutralize pro-inflammatory DAMPs, and the lysosome membrane sequesters inflammatory lipids from cytosolic sensors [166]. In addition, efferocytosis initiates production of anti-inflammatory cytokines and specialized pro-resolving mediators [2].

Dysfunctional intracellular processing is a hallmark of neurologic disorders, and often exacerbates pathology [167,168]. Phagocytic uptake may progress initially, but insufficient cargo processing within microglia and other immune cells leads to the accumulation of toxic products. This dysregulation is underscored by genetic mutations (e.g., in Trem2 or Grn) that impair lysosomal function, which exacerbates pathological accumulation of undigested material [169,170].

Severe CNS pathologies such as MS, CNS injury, and AD generate excessive amounts of lipid debris. Although phagocytes initially engulf this debris, they have difficulty with its degradation; this leads to the formation of large, lipid-filled cytoplasmic vacuoles in cells called foamy macrophages (or microglia) [49,149]. This processing failure creates a bottleneck: phagocytes become saturated with undigested material and lose capacity for additional uptake, allowing toxic debris to accumulate extracellularly. Further, these foamy macrophages/microglia can develop a persistent pro-inflammatory state that amplifies cytotoxic conditions. Conversely, mouse and human foamy macrophage/microglial formation can be decoupled from proinflammatory cell states: deletion of the negative lipid regulators BHLHE40/41 enhances cholesterol clearance and non-toxic lipid storage [171]. This implies that foamy cell formation constitutes a conserved, adaptive strategy for responding to pathology, and that the inflammatory state of these cells can be modified by cell-intrinsic and environmental cues [172,173]. Overall, efficient intracellular processing of engulfed material is essential for coordinating CNS repair.

Emerging therapeutic approaches aim to target intracellular processing to aid phagocytosis, immune resolution, and neurologic repair. Strategies include pharmacologically enhancing lysosome biogenesis and function, and amplifying lipid metabolism by targeting lipid transporters and metabolic pathways [149]. Accelerating the intracellular breakdown and clearance of lipid-rich debris could delay the formation of foamy macrophages and also boost upstream phagocytic engulfment, thereby enhancing the clearance of toxic debris and dampening pro-inflammatory CNS conditions.

In the following discussions, we consider how endogenous phagocytic responses modulate CNS damage and repair in multiple sclerosis (MS), aging and Alzheimer’s disease (AD), traumatic CNS injury, and ischemic stroke. For reviews on the roles of phagocytosis in additional neurologic disorders (e.g., Parkinson’s disease and viral infection) see [38,51].

Multiple sclerosis

MS is a chronic autoimmune disease characterized by multifocal demyelination of CNS axons, leading to disrupted axonal communication and neurologic function. The initiation and progression of MS are associated with phagocytic activities involving microglia and macrophages. These innate immune cells localize around sites of demyelination, and can also provoke remyelination during remission in relapsing-remitting MS. Here, we highlight in MS the roles of specific phagocytic pathways, including TREM2, MerTK, and scavenger receptors.

The TREM2 pathway activates reparative microglial cell states and enhances remyelination in rodent models of MS. TREM2 is a transmembrane protein primarily expressed on microglia that complexes with TYROBP and binds to various ligands to engage in phagocytic activity. During chronic demyelination, Trem2 knockout (KO) mouse microglia retain some capacity to engulf myelin debris but are unable to process intracellular myelin cholesterol [52]. This leads to pathological cholesterol (lipid) accumulation in Trem2 KO microglia, which is associated with impaired transition to reparative transcriptomic cell states, stress of the endoplasmic reticulum, and worsened neuronal damage [52]. Indeed, Trem2 KO microglia from male mice exhibit impaired cholesterol esterification and lipid droplet biogenesis [53], highlighting the adaptive nature of these responses in addressing demyelination. This cholesterol accumulation transforms microglia and macrophages into foamy cells, which contain many indigestible lipid droplets and can persist indefinitely [54] (see Box 2). During experimental autoimmune encephalomyelitis (EAE), female Trem2 KO mice exhibit similarly shifted microglial and monocyte inflammatory cell states and worsened clinical course [55]. Using the cuprizone model of demyelination, blocking TREM2 function in male mice through both the demyelination and remyelination phases prevents formation of a MAFB+ microglia cell population [56]. MAFB+ microglia are associated with remyelination and repair and are found in active MS lesions in humans, suggesting that TREM2 is required for microglia to develop a reparative, promyelinating phenotype. The aforementioned results in rodent models likely have translational relevance, given that TREM2 is increased in the CSF and on CSF monocytes in patients with MS [57], and genetic TREM2 deficiency in humans drives robust white matter demyelination and early-onset dementia [58]. Overall, TREM2 appears to have complex, context-dependent roles in MS and demyelination.

TAM receptor pathways are required for beneficial microglial inflammatory responses, phagocytosis, and repair in models of MS. Treatment of human microglia with TGF-β induces MerTK expression and enhanced myelin uptake, and MerTK antagonism inhibits phagocytic capacity and shifts microglia and macrophages towards a pro-inflammatory cell state [59]. Mechanistically, MerTK signaling has dual roles: upon binding phosphatidylserine, MerTK activates PI3K-AKT for engulfment while inducing SOCS1/3 to suppress inflammatory pathways [60]. After cuprizone-induced demyelination, MerTK KO female and male mice have deficient clearance of myelin debris and remyelination; in parallel, MerTK KO brains exhibit increased interferon signaling that likely worsens microglial ability to repair focal demyelination [61]. Similarly, deletion of the TAM receptor Axl or the bridging ligand Gas6 exacerbate lesions, demyelination, and deficits during EAE [62,63]. Administering Gas6 to female and male mice improves the clinical course of EAE [62] and boosts remyelination in males after cuprizone-induced demyelination [64]. Overall, these data suggest that activating TAM receptor pathways is a promising approach for ameliorating MS pathology.

Several scavenger receptors show promise for promoting repair in models of MS. Overexpressing CD36 boosts myelin uptake by cultured mouse or human microglia, and increasing CD36 via niacin treatment after lysolecithin-induced demyelination in mice improves myelin removal, remyelination, and oligodendrocyte precursor cell proliferation [65]. Future studies should complement biomarker data with gain- and loss-of-function experiments to define in greater detail the role of notable scavenger receptors during MS [36,37].

Aging, dementia, and Alzheimer’s disease

Dysregulated phagocytosis is a prominent feature of both healthy aging and aging in pathological contexts (e.g., neurodegenerative diseases such as AD). The aging brain exhibits increased deposition and insufficient phagocytic clearance of potentially harmful immunomodulators, such as cytokines, misfolded proteins, and damage-associated molecular patterns [66–68]. A hallmark of AD is pathological accumulation of amyloid-β (Aβ) plaques, with inadequate phagocytic removal primarily by microglia [69]. Accordingly, causative genes for early onset AD are related to Aβ formation and processing (e.g., APP and presenilins). Notably, the most robust genetic risk factor for AD is the APOE4 allele (vs. the most common allele APOE3, or the most protective allele, APOE2) [70]. APOE regulates microglial uptake and clearance of Aβ and apoptotic debris. In human induced pluripotent stem cell (iPSC)-derived astrocytes, APOE primes the astrocytes to support microglial phagocytosis [69,71]. Aging-related shifts in microglia cell state [66,72–75] and ineffective phagocytic clearance are partially due to cellular senescence, but may also represent a struggle to compensate and repair an aging and degrading CNS environment. Although microglia are the primary phagocytes clearing Aβ, astrocytes contribute through ABCA1-mediated pathways facilitating lipid processing and debris removal [20]. Numerous AD risk genes are enriched in the phagocytic pathway (see [39,76]); here, we discuss three pathways with particularly strong genetic and functional links to AD pathogenesis and aging-related microglial dysfunction: TREM2, CD33, and integrins.

TREM2 binds various CNS ligands that increase with age, including Aβ, apolipoproteins, and apoptotic debris. During typical aging, TREM2 has mixed roles in regulating CNS inflammation and pathology [19]. Microglia lacking Trem2 persist in a homeostatic phenotype that is neither cytotoxic nor reparative – i.e., Trem2-deficient microglia are unable to develop disease-associated microglia (DAM) cell states [77] (see Box 1). Similarly, human iPSC-derived microglia lacking TREM2 fail to develop DAM states, exhibit impaired phagocytosis and CXCR4-dependent migration, and have reduced accumulation around Aβ plaques [78]. Interestingly, Trem2-deficient mice exhibit impaired microglial responses and worsened myelin degradation, but paradoxically show reduced age-related synapse loss and cognitive decline [58,79–81], highlighting TREM2’s complex roles in aging. TREM2 activation dampens inflammation through DAP12-dependent signaling [82], so Trem2 loss creates a dual deficit: impaired phagocytosis and unchecked inflammatory responses.

TREM2’s role has also been extensively investigated in the context of AD. Microglia and astrocytes from individuals with dementia or clinical AD internalize excess synapses, which likely contributes to neurodegeneration [83–85]. In mouse models of AD, TREM2 on microglia can limit synapse removal by binding and sequestering the complement protein C1q, thereby quenching microglial-mediated synaptic pruning driven by the complement cascade [86]. Permanent TREM2 deletion or loss-of-function risk variant R47H reduces microglial accumulation at Aβ plaques in female mice, leading to impaired Aβ uptake and worsened pathology [87,88]. In mice that overexpress the R47H-TREM2 variant, inhibition of the downstream signaling molecule AKT limits R47H-driven microglial reactivity, pro-inflammatory cell signatures, and synapse loss (more robust in females than males) [88]. Similarly, human microglia from R47H AD patients exhibit exaggerated expression of inflammatory and AKT pathways and enrichment of additional pathways in sex-specific patterns [88]. This implies that early pathological R47H-TREM2 signaling via AKT in mice and humans drives hyperreactivity of microglia in the presence of tau to amplify AD-related damage and deficits, and that wildtype TREM2 is protective at this early stage. TREM2 also regulates progression of late-stage AD. Later stages of AD-like pathology in mice are characterized by tau pathology; in models of tau-driven degeneration, TREM2 exacerbates microglial reactive cell states and neuropathology [89–91]. Further, female and male mice permanently lacking Trem2 show persistent decreases in accumulation of myeloid cells around plaques, and at later times this is associated with increased plaque size and total area, and impaired Aβ uptake by microglia and macrophages [87]. In contrast, transient late-stage reduction of CNS Trem2 using antisense oligonucleotides in male APP/PS1 mice enhances phagocytosis by microglia and reduces plaque deposition [92].

With regards to expression, TREM2 is upregulated in multiple microglial subsets in aged mice and humans, including “TREM2-dependent activated microglia” enriched for phagocytic pathways, and “senescent microglia” more prevalent in aged and 5XAD brains [93,94]. In human CSF, sTREM2 increases in early-stage AD as a biomarker before symptom onset [95–97]. sTREM2 decreases at later stages of AD; AD individuals with higher sTREM2 exhibit decelerated cognitive decline [98]. Thus, TREM2 enables inflammatory cell state transitions and engages in crosstalk with other phagocytic pathways.

CD33 is expressed by microglia and macrophages, and regulates phagocytosis by inhibiting TREM2 signaling. Cd33 deletion in 5xFAD mice protects against memory deficits and reduces Aβ burden – effects abolished by additional KO of Trem2 [99]. However, CD33 deletion in human microglia enhances Aβ phagocytosis but also increases pro-inflammatory responses [100]. In mouse microglia and human U937 monocytic leukemia cells, the protective CD33 splice variant CD33m enhances Aβ clearance via intracellular signaling without extracellular binding [38,101], suggesting that fine-tuning CD33 localization or isoform balance may optimize therapeutic benefit.

Integrins promote excess synaptic pruning in AD models. The microglial integrin α_Mβ₂ (CR3) binds complement C3 to mediate synaptic pruning. In mouse models of AD, deletion of α_M or C3 ameliorates synapse loss and cognitive deficits [102,103], and α_Mβ₂ binding to Aβ plaques provokes both phagocytosis and cytotoxic mediator release [104,105]. Integrin activation generally exacerbates Aβ pathology [106].

Traumatic brain injury and spinal cord injury

CNS injury elicits abrupt inflammation at the site of damage, driving reactivity of resident neuroimmune cells and uncontrolled infiltration of circulating white blood cells (leukocytes) (see [7,29]). Acute inflammation after CNS injury involves immune cells developing pro-inflammatory cell states; the release of cytotoxic factors and cytokines; and spreading cell death. Ultimately, the CNS injury-induced immune response expands the primary lesion, a process called “secondary damage.” Inflammation in the lesion core persists for months to years in the injured rodent and human CNS.

Neuroinflammation after CNS injury is dampened by early phagocytic responses – particularly by microglia and macrophages. In the first 3 days after spinal cord injury (SCI) in mice, microglia are the predominant phagocytic cell and are relatively efficient at engulfing debris. At 7 days and later, macrophages take on a more prominent role in phagocytosis – but these cells are more susceptible to apoptosis and are less efficient at myelin uptake and processing compared to microglia [5]. Ultimately, these cells become foamy macrophages (Box 2) [49,50]. Repair after CNS injury is regulated by various phagocytic pathways, including the complement cascade and TREM2.

CNS injury activates the complement system: complement components are released from the blood, and CNS-localized cells express complement-related machinery (for reviews, see [107–109]). Complement proteins opsonize myelin and cell debris, tagging it for removal after CNS injury; however, these complement proteins also amplify immune cell recruitment and activation, and exacerbate bystander damage in adjacent intact tissue [107,110–113]. Complement inhibition strategies consistently improve outcomes after CNS injury by reducing inflammatory cascades while preserving debris clearance capacity [110–115]. This likely has functional implications for humans; complement deposition is detected after SCI and is positively correlated with requirement for pain medications [116]. Thus, although the complement system helps remove myelin debris after CNS injury, it has substantial detrimental roles in exacerbating pro-inflammatory cascades, destroying spared cells and synapses, and disrupting membrane continuity.

TREM2 enables microglial inflammatory activation, which can have disparate effects on recovery after CNS injury. sTREM2 is highly expressed in the CSF of humans with SCI [55]. Trem2 KO mice display increased locomotor recovery after SCI; female Trem2 KO mouse microglia maintain more of a homeostatic-like phenotype and the KO epicenter has fewer infiltrating macrophages [55,117]. Similarly, after traumatic brain injury (TBI), Trem2 KO microglia exhibit impaired transition from homeostatic state to a potentially reparative DAM cell state; however, in contrast with the SCI locomotor results, male Trem2 KO mice with TBI exhibit worsened motor function, learning, and memory; and larger lesions [118]. Differences in functional outcomes in Trem2 KO mice after SCI vs. TBI may be due to tissue-specific injury dynamics; divergent methods between studies; or sex differences in neuroimmune dynamics (for discussion on sex differences in phagocytosis, see Box 3).

Box 3. Sex differences in CNS phagocytosis: divergence in mechanisms and pathology.

Biological sex is a fundamental biological variable that shapes CNS immune function across the lifespan, and influences susceptibility to neurological disorders [174,175]. For instance, females are more predisposed to autoimmune conditions like multiple sclerosis (MS) [176,177], whereas males have more adverse outcomes following CNS trauma [7,75]. Despite this robust sexual dimorphism in neuroimmune responses, most research so far, particularly when it comes to animal model studies but also in human research, relies on single-sex cohorts or fails to report sex-disaggregated results. Thus, there is a critical need to further study sex-specific CNS phagocytic responses.

Mechanisms of microglial metabolism and reactivity exhibit sex differences. In a mouse model of AD, microglia from females have aberrant glycolytic metabolic preference and exaggerated inflammatory state compared to those from males [178]. This deleterious metabolic state in female microglia results in reduced capacity for Aβ phagocytosis, which correlates with increased Aβ plaque number and area [178]. In contrast, amoeboid microglia observed in male AD patients may indicate better preservation of phagocytic function [178].

The functional outcome of key phagocytic pathways is also sexually dimorphic and disorder-dependent. The AD-linked R47H-TREM2 variant shows stronger detrimental effects on synapse loss in female AD mice [88]. TREM2 function in microglia is complex: whereas female Trem2 KO mice exhibit worsened EAE pathology, they maintain a more homeostatic-like phenotype after SCI [55,117]. This variability in the effects of modulating TREM2 highlights how genetic factors interact with both biological sex and specific pathology to determine phagocytic outcomes.

CNS immune programming, priming, and reactivity is influenced by sex steroid hormones – estrogens generally promote anti-inflammatory responses and are neuroprotective in the CNS, modulating microglial function via Estrogen Receptor beta signaling [179].

Given the robust influence of biological sex on immune regulation and disease pathology, future research should prospectively integrate both sexes [74,180,181]. This approach may help reveal sex-specific signaling pathways and define optimal therapeutic windows, ultimately driving development of tailored interventions that fine-tune phagocytosis and improve clinical outcomes.

Intracellular signaling cascades determine whether phagocytosis helps or harms repair. The HV1-NOX2-ROS pathway exemplifies this context-dependency: the voltage-gated proton channel HV1 enables sustained NOX2-mediated reactive oxygen species (ROS) generation for debris degradation, yet after CNS injury, prolonged activation drives excessive ROS that oxidizes perilesional tissue and exacerbates secondary damage [119]. Removal of HV1 in mice limits oxidative injury and improves outcomes after TBI, SCI, and stroke [120–122]. Thus, dysregulated phagocytic machinery represents a potential therapeutic target after CNS injury.

Ischemic brain injury: stroke

Ischemic stroke causes an acute-to-chronic neuroinflammatory response, and post-stroke pathology is shaped by phagocytic processes [123]. As with TBI and SCI, neuroinflammation after ischemic brain injury is regulated by the complement pathway and TREM2-TYROBP signaling.

Complement activation acutely after ischemic injury causes phagocytosis of surviving neurons and synapses in the stroke penumbra; inhibiting complement activation in mice via complement inhibitor B4Crry, or via a C3a receptor antagonist, spares penumbral tissue to enhance functional recovery [124]. Interestingly, intranasal C3a delivery starting 7 days after stroke in male mice enhances peri-infarct cortical connectivity and reduced astrocytic reactivity [125], suggesting that divergent complement-related approaches can be timed optimally to enhance recovery.

After stroke, removal of TREM2 or its transmembrane signaling adapter TYROBP (DAP12) is generally detrimental to recovery – Trem2 deficiency in male mice impairs phagocytosis of dying cells in the lesion site [41]. TREM2 likely has a beneficial role in microglia and macrophages by promoting these cells to develop a more anti-inflammatory and reparative cell state – e.g., Trem2 is a key upregulated gene after ischemia in repair-associated macrophages, which aid in lipid recycling, vascular integrity, and CNS repair [126].

Phagocytic receptors such as MerTK and MEGF10 may exacerbate stroke pathology. MEGF10 binds eat-me signals C1q and phosphatidylserine, and acts with the ABCA1 transporter to engulf debris after ischemia [20]. After ischemic stroke, microglial or astrocyte deletion of either receptor reduces phagocytosis and benefits penumbral sparing and recovery in male mice; after hemorrhagic stroke, deletion in microglia/macrophages (but not astrocytes) is protective [26]. This highlights how identical receptors can have distinct roles depending on cellular and pathological context. In contrast, activating TAM receptors via intranasal delivery of the TAM bridging ligand Gas6 in male mice is beneficial after middle cerebral artery occlusion [127].

Finally, balancing “eat-me” and “don’t eat-me” signals is crucial after stroke. Blocking eat-me signals (phosphatidylserine or C1q) in male mice reduces excess synaptic pruning to improve outcomes after photothrombotic stroke [128]. Conversely, inhibiting CD47 “don’t eat-me” signaling in mice after stroke enhances uptake of persistent apoptotic debris and improves outcomes [129–131]. This highlights the therapeutic potential of modulating “eat-me” and “don’t eat-me” signals to benefit stroke outcomes.

Therapeutic modulation of phagocytosis in CNS disorders: selected examples and translational challenges

As discussed above, phagocytic pathways have broad implications for tissue pathology and repair in various neurologic disorders. Promising phagocytosis-targeting therapeutic strategies are under development or in clinical trials, including approaches that enhance efferocytosis, block “don’t eat-me” signals, modulate eat-me signals, target specific phagocytic receptors, and enhance intracellular processing (reviewed in [2,38,132]). Here, we highlight selected examples that illustrate key translational principles and challenges, focusing on TREM2, TAM receptors, and eat-me signal modulators that have advanced furthest in CNS applications (Figure 3).

Figure 3. Emerging approaches, strategies, and outcomes for optimizing CNS phagocytic efficiency.

a. Agonists for phagocytic receptors, such as a TREM2 agonist antibody, drive intracellular signaling that can enhance phagocytosis and inflammatory pathways within CNS phagocytes. b. Adding engineered fusion or hybrid proteins can amplify uptake of pathological CNS debris and amyloid-β (Aβ) plaques. For example, Aβ removal can be augmented by a hybrid protein pairing an Aβ-specific antibody moiety with Gas6 [142], thereby crosslinking Aβ with TAM receptors like MerTK. c. Another goal of phagocytic modulation is preserving lesion- or plaque-adjacent CNS tissue; this can be achieved by masking “eat-me” signals exposed on injured-but-viable cells. In the illustrated hypothetical approach, externalized phosphatidylserine (PS) is masked by annexin V from its cognate receptors. d. Focal activation of phagocytosis aims to enhance uptake of apoptotic debris and plaques specifically in the pathology epicenter. Emerging strategies include local injection of phagocytosis-driving particles (e.g., nanoparticles, AAVs); local transient breakdown of the blood-brain barrier (BBB) to enable drug delivery; or other innovative strategies like local activation using nanomaterials, chemogenetics, or optogenetics. Future research should further test, optimize, and combine these emerging strategies to boost phagocytosis selectively in pathological areas of the CNS. Figure created in part with BioRender.com.

TREM2 activation is one potential approach for ameliorating pathology in several CNS disorders [133]. In Trem2^+/− mice with cuprizone-induced demyelination, systemic delivery of the mouse TREM2-activating antibody AL002a enhances microglial phagocytic activity and myelin uptake, improving remyelination and axon integrity [134]. Similarly, TREM2 agonist antibodies reduce Aβ plaques and enhance microglial surveillance in AD models [135–137].

However, conflicting results highlight mechanistic gaps in our understanding. Several TREM2 agonist antibodies failed to benefit pathology in mouse AD and MS models, with some worsening remyelination [138]. These disparate results likely relate to differences in CNS penetrance, agonist efficacy, and disease context. Two clinical trials^i,ii using TREM2 agonist antibodies for AD have been discontinued due to safety signals [139], underscoring the need to optimize timing, biodelivery, and molecular design.

TAM receptor targeting shows promise across multiple disorders. Gas6-TYRO3 and ProS1-AXL signaling correlates with neuroprotection in human MS brains [140]. Intrathecal Gas6 treatment in mice improves EAE clinical scores [62], while intranasal Gas6 delivered to male rats reduces infarct size after stroke [141]. Innovative approaches include fusion proteins that combine anti-Aβ antibodies with Gas6 domains to enhance plaque clearance [142].

Eat-me signal modulation can balance debris removal with tissue preservation. Blocking complement component C1q or externalized phosphatidylserine in male mice reduces excessive synaptic pruning and improves outcomes after stroke or mild TBI [128,143]. Don’t eat-me signal inhibition is another promising translational strategy: CD47 blockade has efficacy in multiple cancer types by enabling macrophage-mediated tumor clearance, revealing promise in animal models that has led to ongoing clinical trials for treating cancer [144]. Similar approaches in CNS disorders show therapeutic potential (as discussed above in the ‘Ischemic brain injury: stroke’ section); however, poor CNS penetrance limits therapeutic efficacy of many phagocytic therapies, requiring improved drug delivery strategies [88,89].

Aberrant phagocytosis and impaired function of the phagocytic cell can worsen neuropathology [145,146]. After cerebral ischemia, male mice lacking MerTK or MGF-E8 have improved neuroprotection and locomotor recovery – likely due to microglial MerTK-mediated uptake of stressed-but-viable neurons [147]. MCAO-induced stroke in mice or monkeys is associated with impaired phagocytosis by microglia: microglia in the pathological environment experience energy depletion and induce self-preserving autophagy, thereby limiting process motility, lysosomal function, and engulfment [148].

Additional immunomodulatory strategies show promise. Cancer immunotherapy has successfully targeted CD47-SIRPα signaling to enhance tumor clearance [144]; similar approaches are being explored for CNS disorders. Complementary strategies targeting eat-me signals or intracellular lipid processing may overcome limitations of single-pathway receptor agonism.

The approaches highlighted here represent a subset of emerging phagocytic therapeutics. Additional strategies include enhancing intracellular lipid processing; targeting signaling cascades; modulating phagocyte metabolism; and leveraging combinatorial approaches [2,144,149,150]. The diversity of these approaches reflects the complexity of phagocytic regulation and the need for context-specific therapeutic strategies.

Overall, future therapeutic development should optimize phagocytic pathway engagement by integrating temporal dynamics and cellular heterogeneity to enhance debris removal while preserving viable tissue.

Concluding remarks and future perspectives

Phagocytic pathways are promising yet complex targets for manipulating repair during neuroinflammatory conditions. Phagocytosis is performed by various CNS cells, including microglia, astrocytes, and macrophages. Distinct phagocytic receptor pathways have their own signaling mechanisms and parallel immunomodulatory effects, and several of these pathways engage in crosstalk. Phagocytosis is a vital process that shapes CNS development and homeostasis by pruning excess synapses and removing apoptotic cells; further, phagocytosis enables repair during neuroimmune perturbation. The role of phagocytic pathways depends on various features of the neurologic condition. CNS injury, MS, and AD have disorder-specific features: they have distinct etiologies, timecourses, and inflammatory intensities. From a therapeutic perspective, additional key challenges must be addressed, including the heterogeneity of each disorder in humans and the influences of genetics and the environment (both as independent factors and combined, expanding variability in disease course). Thus, these disparate neurologic conditions likely have disorder-specific and individualized phagocytic needs and optimal phagocytosis-based manipulations.

On the surface, removing apoptotic cells and debris during CNS pathology via phagocytosis sounds productive – but this is often paired with potentially damaging events, including engulfment of still-viable cells; phenotypic transition of overwhelmed phagocytes towards proinflammatory states; and release of cytotoxic mediators. Thus, several unresolved issues remain to be addressed in future work (see Outstanding questions). First, future studies should reveal which phagocytic pathways help maximize benefits of phagocytic removal of debris or intracellular processing of lipids to enhance phagocytic capacity. Alternatively, studies could aim to amplify pathways that reduce damaging effects of phagocytic-related activities (e.g., phagocytosis of lesion-adjacent live cells; excessive release of pro-inflammatory mediators). Furthermore, cell type-specific phagocytic mechanisms must guide therapeutic development, as targeting ubiquitously expressed receptors may have off-target effects across CNS cell populations. A critical challenge is understanding how phagocytic receptors coordinate debris clearance with inflammatory regulation, and future research should disentangle whether anti-inflammatory effects result from receptor signaling, debris removal, or metabolic reprogramming. In parallel, given sex differences in innate and adaptive immune function, future studies on phagocytosis should consider both sexes, and include sex as a variable in openly-shared raw data for future meta- or mega-analyses.

Outstanding questions.

• Phagocytic receptors gate debris processing and inflammation: phagocytosis can drive removal of damaging debris, while simultaneously initiating pro-inflammatory activities of the phagocyte that can harm healthy tissue. How can one amplify beneficial effects and minimize potentially damaging aspects of phagocytic pathways?

• Individual receptor pathways involved in phagocytosis and neurologic repair can have divergent effects even within a single disorder. For example, activation of TREM2 can ameliorate or worsen AD pathology, depending on modulation strategy, duration, and timing. What are the mechanisms underlying these divergent effects of single phagocytic pathways?

• Phagocytic pathways can have distinct effects on pathology in different CNS disorders. This may relate to the initiating event (sudden damage vs. chronic dysregulation), intensity of initial inflammation, or a combination of factors. How do phagocytic manipulations interact in distinct manners with different CNS disorders? What are common features that evoke protection versus exacerbation of injury?

• Phagocytic cells responding to CNS pathology can elicit bystander damage of healthy cells, by phagocytosing injured-but-viable cells and by producing cytotoxic mediators. For each CNS disorder or disease: how can phagocytic manipulations be optimized in space and time?

• Sex differences exist in the balance between innate and adaptive immunity, and numerous CNS disorders have sex differences in risk, development, or presentation of symptoms. How does sex affect CNS phagocytosis in these neurologic disorders? Would it be useful to develop sex-specific strategies for manipulating phagocytosis in the CNS?

Another emerging approach to optimize phagocytic efficacy is customizing drug delivery in space and time (Figure 3). For CNS injury and stroke, this could be achieved by enhancing phagocytic activity specifically in the lesion, while minimizing bystander damage – e.g., via combinatorial strategies to modulate phagocytosis using innovative ultrasound, chemogenetic, or nanomaterial technologies [148]. For MS or AD, targeted phagocytosis at lesions or plaques could be putatively implemented by activating a phagocytic pathway in the presence of a disorder-specific molecular signature. CNS injury and stroke often have a defined initiating event and stereotypic neuroinflammatory evolution, so phagocytic manipulations can be delivered in amenable therapeutic windows. Future studies in animal models could use transgenic cell fate-mapping strategies and single-cell sequencing in various CNS disorders to delineate optimal timing of phagocytic manipulations. Finally, the use of novel hybrid phagocytosis-related receptors, ligands, antibodies, or cytokines could enhance phagocytic uptake and boost reparative signaling.

Overall, phagocytosis represents a promising frontier for orchestrating CNS repair. From early observations of cells “devouring” foreign material in sea star larvae to current prospects of targeted approaches for treating neurodegeneration, the field has evolved from descriptive studies to mechanistic understanding with therapeutic potential. Future studies are expected to leverage spatiotemporal targeting, eat-me/don’t-eat-me signal modulation, and hybrid therapeutic designs to enhance beneficial debris clearance while preserving viable tissue – ultimately unveiling phagocytic strategies that ameliorate CNS pathology and improve clinical outcomes.

Highlights.

• Phagocytes in the central nervous system (CNS), including astrocytes, microglia, and macrophages, shape development and homeostasis by pruning synapses and removing apoptotic debris.

• Phagocytosis is mediated by various ligand-receptor dyads and signaling pathways, enabling CNS phagocytes to respond to neuroimmune shifts across the lifespan and during pathology.

• Phagocytosis pathways regulate recovery in various models of CNS pathology, including multiple sclerosis, CNS injury, ischemic stroke, and age-associated neurodegeneration.

• Phagocytosis pathways are intimately integrated with inflammatory cell state and remove viable cells in pathology-adjacent tissue, highlighting the complexity of targeting these systems.

• To maximize benefit and minimize off-target damage, new phagocytic-based approaches should optimize drug delivery timing and location, tailored for each CNS pathology.