M2-associated microglial states in oligodendrocyte regulation and white matter repair in vascular dementia

Abstract

Vascular dementia (VaD), the second most common form of dementia, lacks approved disease-modifying therapies. White matter injury and demyelination are major pathological features, and oligodendrocyte-lineage damage directly limits myelin repair. M2-associated microglial responses regulate inflammation, clear cellular and myelin debris, and support oligodendrocyte differentiation and maturation. Single-cell studies, however, have exposed the limitations of the classical M1/M2 dichotomy in brain disease. In this review, M2a-, M2b-, and M2c-like states are interpreted based on experimentally specified inducing conditions and concordant molecular or functional evidence. We examine how these programs relate to oligodendrocyte injury, myelin loss, and white matter repair in VaD. with particular attention to their potentially stage-dependent contributions. Integrating these findings may inform the timing and functional focus of future interventions aimed at preserving oligodendrocyte function and promoting white matter repair in VaD.

Search strategy

Studies cited in this review were published between 1987 and 2026. Relevant literature was identified primarily through searches of the PubMed database, with particular emphasis on studies published during the most recent decade, from January 2016 to July 2026. PubMed was searched using combinations of Medical Subject Headings (MeSH) and title/abstract terms. The principal MeSH terms included “Dementia Vascular”, “Brain Ischemia”, “Microglia”, “Macrophage Activation”, “Oligodendroglia”, “Oligodendrocyte Precursor Cells”, “White Matter”, “Myelin Sheath”, “Remyelination”, and “Phagocytosis”. These were supplemented with title/abstract terms including “vascular dementia”, “chronic cerebral hypoperfusion”, “white matter injury”, “M2 microglia”, “M2-like microglia”, “M2a”, “M2b”, “M2c”, “alternative activation”, “oligodendrocyte differentiation”, “myelin repair”, and “myelin debris”. Searches were performed using different combinations of disease-, cell-, and function-related terms. Earlier studies were included when they provided foundational evidence for oligodendrocyte development, myelination, microglial/macrophage activation, or the molecular mechanisms discussed in this review. The final search was conducted on July 30, 2026.

Introduction

Vascular dementia (VaD) is a syndrome of cognitive decline attributable to cerebrovascular disease or vascular risk factors. It is considered the second most common form of dementia after Alzheimer’s disease (AD), and its incidence is expected to rise as populations age [1]. According to a 2026 report from the World Stroke Organization, approximately 8.5 million people worldwide have pure VaD, whereas nearly 9.1 million have mixed dementia. By 2050, the combined population with pure or mixed VaD is projected to reach 42.7 million [2]. This disease burden carries substantial economic costs, with annual spending on VaD diagnosis, treatment, and follow-up estimated in the trillions of dollars, or approximately 0.421% of global GDP [3].

The pathogenesis of VaD involves cerebrovascular injury, cerebral ischemia and hypoxia, neuronal damage, and subsequent cognitive decline [4]. Clinical manifestations include memory loss, impaired attention, and language or visuospatial deficits, often accompanied by neurological sequelae such as hemiplegia, sensory loss, dysarthria, and dysphagia. White matter injury caused by ischemia and hypoxia is a major contributor to cognitive dysfunction [5], with demyelination and axonal loss among its principal pathological features. Oligodendrocyte injury is a direct cellular contributor to demyelination [6].

Oligodendrocytes are a major glial cell type in the central nervous system (CNS) and are closely coupled to myelin structure and function. Mature oligodendrocytes extend flattened membrane processes around neuronal axons to form the dense, multilayered myelin sheath [7]. In addition, oligodendrocytes provide metabolic support and contribute to myelin maintenance. After myelin injury, surviving oligodendrocytes can support local repair through metabolic adaptation [8]. Severe myelin damage is accompanied by oligodendrocyte loss and the recruitment of oligodendrocyte precursor cells (OPCs), which differentiate into mature oligodendrocytes and contribute to remyelination [9]. Thus, preserving oligodendrocyte-lineage function and supporting OPC-mediated repair are central to restoring myelin integrity in VaD.

Currently, the clinical management of VaD remains focused primarily on controlling vascular risk factors, preventing cerebrovascular events, rehabilitation, and symptomatic treatment, while no disease-modifying therapy has yet been approved specifically for VaD [4, 10, 11]. Although these measures may reduce the risk of subsequent vascular events and alleviate some clinical symptoms, they do not directly reverse established white matter injury or myelin loss. Promoting white matter repair therefore remains an important unmet need in VaD treatment.

M2-associated microglial functions include inflammatory regulation, cellular and myelin-debris clearance, and trophic support. Together, these processes can shape oligodendrocyte survival, differentiation, and remyelination [12–14]. To date, however, most studies have examined M2-like microglia as a broad functional category. The relative contributions of M2a-, M2b-, and M2c-like states to oligodendrocyte regulation and myelin repair in VaD remain incompletely defined. Accordingly, this review summarizes current evidence on the roles of M2a-, M2b-, and M2c-like microglial states in VaD and considers how these findings may provide new directions for therapeutic development and clinical translation.

Differentiation of oligodendrocytes and myelin formation

Origin of oligodendrocytes: from neural stem cells to OPCs

In mammals, cells of the CNS originate from the embryonic ectoderm. Inductive signals specify the neural plate, which folds and closes to form the neural tube. Neuroepithelial cells in the neural tube wall retain self-renewal and multilineage differentiation capacity. Oligodendrocyte-lineage cells arise primarily from progenitors generated in the ventricular zone and subventricular zone (SVZ) [15, 16].

Transcription factors including Olig1, Olig2, Sox10, Nkx2.2, and Zfp488 [17–19] regulate the specification of neural stem cells (NSCs) into OPCs. Among these factors, Olig2 serves as a key transcriptional regulator during oligodendrocyte lineage specification. It determines the commitment of NSCs toward the oligodendrocyte lineage [20] and cooperates with Sox10 to promote subsequent OPC differentiation and maturation [18]. Loss of Olig2 directly impairs OPC generation. Sox10 acts as a downstream regulator of OPC specification and maturation and promotes OPC proliferation and migration by regulating genes such as platelet-derived growth factor receptor α (PDGFRα) [18, 21]. Nkx2.2, Zfp488, and Sox4 cooperate with Olig2 and Sox10 to promote oligodendrocyte-lineage specification. Within this lineage, OPCs are the principal proliferative and migratory cell population. Representative markers include Olig2, PDGFRα, and NG2 [22]. The transition from NSCs to OPCs marks the initial stage of oligodendrocyte development. OPCs subsequently migrate from ventricular regions into white and gray matter under the guidance of local cues, providing a cellular basis for later myelination (Fig. 1).

Fig. 1.

Origin and early differentiation of the oligodendrocyte lineage

Cells of the mammalian CNS originate from the embryonic ectoderm and develop through the neural plate and neural tube. Oligodendrocyte-lineage cells primarily derive from NSCs in the SVZ of the neural tube. Within the core transcriptional network, Olig2 promotes lineage commitment, whereas Sox10 supports OPC specification and maturation. Nkx2.2, Zfp488, and Sox4 cooperate with these factors to drive OPC generation and maintenance.

Migration and distribution of OPCs

During OPC migration, platelet-derived growth factor A (PDGF-A) released by neurons and astrocytes binds PDGFRα on OPCs, activating downstream pathways that promote migration and proliferation [23, 24]. After leaving ventricular regions, OPCs can associate with the abluminal surface of adjacent blood vessels. Wnt-dependent C-X-C chemokine receptor type 4 (CXCR4) expression enables OPCs to respond to endothelial C-X-C motif chemokine ligand 12 (CXCL12) and migrate directionally along cerebral vessels [25].

The extracellular matrix (ECM) also contributes to OPC migration. Basement membrane glycoproteins, including fibronectin and laminin, provide adhesion sites for migrating OPCs [26, 27] and cooperate with growth factors such as PDGF-A to enhance chemotactic signaling. Netrin-1 regulates migration direction through its concentration-dependent deleted in colorectal cancer (DCC)/Unc5 receptor system [28]. N-cadherin-mediated interactions regulate OPC adhesion and migration on adjacent glial surfaces [29]. In parallel, contact-dependent self-repulsion enables neighboring OPCs to maintain largely non-overlapping territories and restore local cell density after cell loss [30]. Together, these mechanisms promote the relatively uniform distribution of OPCs throughout the brain parenchyma, thereby maintaining an adequate progenitor pool for subsequent myelination (Fig. 2).

Fig. 2.

Mechanisms underlying the migration and distribution of OPCs

Multiple signaling pathways coordinate OPC migration and distribution. PDGF-A released by neurons and astrocytes binds PDGFRα and promotes OPC proliferation and migration [23, 24]; the Wnt-CXCR4–CXCL12 axis guides OPC migration along cerebral blood vessels. Fibronectin and laminin provide adhesive support [25–27]; netrin-1 regulates migration direction through DCC and Unc5 receptors; and n-cadherin-mediated adhesion may influence oligodendroglial migration on adjacent glial surfaces, whereas contact-dependent self-repulsion helps maintain the relatively even spatial distribution of OPCs throughout the CNS parenchyma [29, 30].

Maturation of oligodendrocytes

After reaching their target regions, OPCs respond to local signals, withdraw from the cell cycle, and begin terminal differentiation. OPC maturation proceeds through three commonly described stages, each with characteristic morphological, functional, and molecular features. Transcriptional programs, signaling pathways, and local environmental cues coordinate this progression toward myelinating oligodendrocytes.

Pre-oligodendrocyte stage (pre-OL)

Signals such as triiodothyronine (T3) and neurotrophin 3 (NT-3) initiate OPC maturation and entry into the Pre-OL stage [31]. At this stage, the expression of progenitor markers such as PDGFRα and NG2 becomes progressively attenuated, whereas O4 becomes more prominent [32, 33]. Cells also shift from a bipolar or tripolar morphology toward a more complex multipolar form. In parallel, inhibitory pathways that maintain OPC proliferation and an undifferentiated state, including Notch and Wnt/β-catenin signaling, are downregulated [34–38]. At the epigenetic level, DNA methyltransferase 3A (DNMT3A) and histone deacetylases 1 and 2 (HDAC1/2) participate in chromatin remodeling and reduce Wnt/β-catenin-mediated inhibition of differentiation [39]. Histone deacetylation also reshapes chromatin accessibility and facilitates activation of myelin-related loci such as myelin basic protein (MBP), proteolipid protein 1 (PLP1), and myelin-associated glycoprotein (MAG), establishing an epigenetic basis for terminal differentiation and myelination [40–42]. At this stage, however, pre-OLs have not yet acquired the capacity to form compact myelin and must undergo further differentiation.

Immature oligodendrocyte (immature OL) stage

As cells enter the immature oligodendrocyte stage, they exit the cell cycle and begin preparing for myelin formation. As pre-OLs progress toward a more mature stage, they undergo marked morphological remodeling: cell volume increases, and their processes elongate, branch extensively, and establish contact with axons, preparing the cells for subsequent myelin formation [43, 44]. Olig1/2 and Sox10 remain expressed, whereas PDGFRα and NG2 are further downregulated. O4 expression is retained, whereas O1/ Galactocerebroside (GalC) expression becomes more prominent as cells enter the immature oligodendrocyte stage [43, 45]. Myelin protein genes activated during the Pre-OL stage are expressed at low levels that remain insufficient for formation of a complete myelin sheath [45]. Cells at this stage can establish initial contact with axons but have not completed maturation [46]; their subsequent maturation can be supported by neurotrophic factors, including those released by M2-like microglia.

Mature myelinating oligodendrocyte stage (mature OL)

Signals from neurons and the local microenvironment drive immature oligodendrocytes into the final stage of differentiation [43, 47]. Their processes undergo extensive remodeling and flatten to form myelin membranes. Olig1/2 continue to regulate this transition. Under the coordinated control of Olig1/2, Sox10, and Myrf, mature oligodendrocytes express high levels of myelin-associated proteins, including MBP, proteolipid protein (PLP), MAG, and myelin oligodendrocyte glycoprotein (MOG), thereby supporting myelin formation [48–51]. Neuron-derived leucine-rich glioma-inactivated 1 (LGI1) binds oligodendrocytic myelin paranodal and inner loop protein (OPALIN) on oligodendrocytes and regulates pathways including mechanistic target of rapamycin (mTOR), thereby supporting differentiation and myelination [47]. Mature oligodendrocytes recognize and ensheath axons. They also comprise transcriptionally distinct populations, including MOL1 to MOL6, that contribute to myelin maintenance and function [52].

Formation of myelin

Myelin formation involves axon selection by mature oligodendrocytes, cytoskeletal reorganization, and membrane assembly. Oligodendrocytes preferentially myelinate axons with a diameter greater than 0.4 μm [53]. Their processes express adhesion molecules such as L1 cell adhesion molecule (L1CAM) and N-cadherin, which facilitate stable contact with suitable axons [54]. This process depends on reciprocal signaling between oligodendrocytes and axons [55]. In demyelinating lesions, soluble factors released by M2-like microglia can also support oligodendrocyte differentiation and remyelination [14].

After stable axonal contact is established, oligodendrocyte processes undergo extensive remodeling and begin to envelop the axon. The processes extend along the axonal surface through “growth cone”-like structures, while the cell membrane spreads across the axon surface, forming inner and outer membranes [56]. The edges of these processes expand and overlap as the membrane spirals around the axon in one or both directions [57]. A mature oligodendrocyte can support 20–60 myelin segments, enabling efficient myelination of multiple axons [58]. The newly formed sheath then requires compaction to become structurally stable and functional.

During myelin densification, the cytoplasm within the cellular processes is gradually compressed and expelled, causing the multilayered lipid membranes to stack tightly [56]. Throughout this process, the myelin proteins MBP, PLP, and MAG are expressed at high levels and participate in sheath assembly. MBP promotes membrane compaction, PLP supports sheath integrity, and MAG stabilizes axon–myelin contact, together producing a compact mature sheath [57, 59–61].

During remyelination after white matter injury, M2-associated microglial programs provide complementary forms of support. M2a-like trophic programs promote oligodendrocyte differentiation and myelin-protein expression [14, 62]. M2c-associated phagocytic programs clear inhibitory myelin debris and create a microenvironment more permissive for oligodendrocyte differentiation and remyelination [63–65]. M2b-like immunoregulatory programs indirectly protect oligodendrocyte-lineage cells and myelin integrity by limiting excessive inflammatory responses [66–68]. Across experimental models, these programs provide complementary and overlapping support for myelin repair by contributing to an immune microenvironment permissive for oligodendrocyte differentiation and remyelination [69] (Fig. 3).

Fig. 3.

Terminal differentiation of oligodendrocytes and myelination

After reaching their target regions, OPCs progress through the pre-oligodendrocyte (Pre-OL), immature oligodendrocyte (immature OL), and mature myelinating oligodendrocyte (mature OL) stages. T3 and NT-3 promote entry into the Pre-OL stage, during which PDGFRα and NG2 expression begins to decline and O4 becomes more prominent. As cells progress toward the immature OL stage, O4 expression is retained, whereas O1/GalC becomes increasingly evident. Immature oligodendrocytes exit the cell cycle, extend more highly branched processes, establish initial contact with axons, and begin expressing myelin-related genes. In parallel, downregulation of Notch and Wnt/β-catenin signaling, together with chromatin remodeling mediated by DNMT3A and HDAC1/2, facilitates activation of the differentiation program. Terminal differentiation is coordinated by Olig1/2, Sox10, Myrf, and the LGI1–OPALIN–mTOR axis and is accompanied by increased expression of MBP, PLP, MAG, and MOG. Mature oligodendrocytes extend and wrap their processes around selected axons, expel cytoplasm, and compact their membranes to form structurally stable myelin sheaths.

Changes in the VaD microenvironment and their effects on oligodendrocyte-lineage cells and myelin

Vascular injury and stage-dependent microglial responses

Cerebrovascular injury is central to the pathogenesis of VaD. Local ischemia and hypoxia disrupt blood–brain barrier (BBB) integrity, increase oxidative stress, and provoke inflammatory responses. Together, these changes create a microenvironment that damages myelin and impairs OPC differentiation and maturation [70, 71]. BBB disruption can further facilitate the entry of circulating inflammatory mediators and immune cells into the brain parenchyma, thereby reinforcing white matter inflammation and limiting remyelination [71].

Within this altered microenvironment, microglia can engage pro-inflammatory programs and release mediators such as IL-1β and TNF-α [72–74]. These mediators can act directly on OPCs and mature oligodendrocytes, suppressing OPC proliferation and differentiation and promoting oligodendrocyte apoptosis, thereby compromising myelin integrity [68, 75, 76]. As myelin breaks down, debris accumulates within lesions when clearance is insufficient. This debris inhibits OPC differentiation and remyelination [77] and can sustain pro-inflammatory microglial responses [78]. Inflammatory activation, however, is not uniformly detrimental. Related studies have indicated that an early, transient pro-inflammatory phagocytic response can facilitate myelin-debris clearance, inflammation resolution, and the generation of new oligodendrocytes [79]. As injury evolves, this early response may be accompanied or followed by repair-associated microglial functions that support debris processing, inflammatory regulation, and oligodendrocyte differentiation [14, 79]. By contrast, prolonged or excessive inflammatory activation can amplify tissue injury and promote inappropriate phagocytosis of stressed but potentially viable myelin [80]. These findings support a temporally coordinated model of microglial responses rather than a strategy based on uniformly suppressing M1-like activity or globally promoting an M2 state [81].

Contributions of astrocytes and the neurovascular unit to white matter repair

Microglial responses during the progression of vascular dementia do not occur in isolation. Astrocytes and other components of the neurovascular unit also participate in remodeling the local environment and may act in concert with microglia to support myelin repair. Astrocytes can promote OPC differentiation and maturation through the release of brain-derived neurotrophic factor (BDNF). Selective reduction of BDNF expression in glial fibrillary acidic protein (GFAP)-positive astrocytes has been shown to decrease the generation of new oligodendrocytes and exacerbate white matter injury [82]. Astrocytes also synthesize and export cholesterol, providing an important lipid source for newly generated oligodendrocytes and supporting regenerative myelination [83].

Astrocytes may additionally influence microglial functional states. In vitro co-culture studies have shown that CD109/Emp1-positive reactive astrocytes can promote microglial proliferation and increase the expression of markers associated with an M2-like response [84]. The reciprocal effect, however, appears less consistent. In another co-culture system, conditioned medium derived from M2-like microglia did not substantially alter astrocytic growth-factor expression or enhance the ability of astrocytes to promote OPC differentiation [85]. These observations suggest that communication between astrocytes and microglia may be bidirectional but not necessarily equivalent in strength or biological consequence. Repair-associated microglial functions, including debris clearance and inflammatory regulation, may complement the trophic and metabolic support provided by astrocytes, although the extent of this cooperation in VaD-related white matter injury requires direct in vivo validation.

Vascular cells also contribute to the microenvironment in which oligodendrocyte regeneration occurs. Cerebral endothelial cells release soluble trophic factors, including BDNF and fibroblast growth factor (FGF), which support OPC survival and proliferation, whereas oxidative stress can weaken this protective paracrine effect [86]. Pericyte regulation appears to be both signal- and context-dependent. Pericyte-derived laminin subunit α2 (LAMA2) can promote OPC differentiation [87]. Under chronic cerebral hypoperfusion, however, pericyte-derived bone morphogenetic protein 4 (BMP4) may redirect OPCs toward an astrocytic fate, thereby restricting oligodendrocyte-lineage proliferation and differentiation [88]. The vascular microenvironment may therefore support or inhibit myelin repair depending on the prevailing cellular state and signaling context.

Taken together, myelin repair is best understood as a multicellular process rather than the consequence of a single microglial state. Its outcome depends on coordinated, and sometimes opposing, signals from microglia, astrocytes, endothelial cells, pericytes, and oligodendrocyte-lineage cells. A local environment that combines effective debris clearance and inflammatory control with sufficient trophic, metabolic, and vascular support is likely to be more conducive to white matter repair (Fig. 4).

Fig. 4.

Multicellular regulation of the white matter microenvironment in vascular dementia. Cerebrovascular injury in VaD alters the local white matter microenvironment and elicits coordinated responses from microglia, astrocytes, endothelial cells, and pericytes. (A) An early, transient inflammatory–phagocytic microglial response may facilitate the initial clearance of myelin debris. When injury persists or inflammation fails to resolve, sustained production of pro-inflammatory mediators, including IL-1β and TNF-α, can aggravate oligodendrocyte and myelin injury and inhibit OPC differentiation. By contrast, repair-associated microglial responses combine debris clearance, inflammatory regulation, and trophic support, thereby creating conditions more permissive for oligodendrocyte-lineage repair. (B) Astrocytes contribute through trophic support, cholesterol-dependent metabolic support, and modulation of repair-associated microglial responses. (C) The vascular niche provides additional regulatory cues. Endothelial cell-derived BDNF and FGF support OPC survival and proliferation, whereas pericyte-derived LAMA2 promotes OPC differentiation. Together, these multicellular interactions shape oligodendrocyte maturation, remyelination, and white matter repair

Regulation of M2-like microglial functional states in VaD

Microglia are one component of this broader repair network, but their contribution is not uniform. In different contexts, they may support oligodendrocyte survival and differentiation, limit damaging inflammation, or remove myelin debris. These functions are commonly discussed in relation to M2a-, M2b-, and M2c-like states.

The M2a, M2b, and M2c framework extends the traditional M1/M2 dichotomy and provides a practical basis for comparing repair-associated microglial functions. These categories were derived largely from macrophage responses induced under defined experimental conditions and were later applied to related functional states in microglia [89, 90]. Their translation to the injured brain, however, is not straightforward. Markers and functional properties assigned to these states frequently overlap in vivo, and their relative expression changes with the type, severity, and duration of injury and with the local microenvironment [90, 91]. Recent single-cell and single-nucleus transcriptomic studies further show that microglia occupy interconnected and dynamically changing states in which phagocytic, metabolic, inflammatory, proliferative, and tissue-repair programs may coexist to varying degrees [91–93]. Despite these limitations, the M2a/M2b/M2c framework remains useful for comparing functional programs dominated by trophic support, immune regulation, and phagocytosis. By integrating defined experimental induction conditions with their corresponding molecular profiles, including relevant signaling pathways and multiple markers, microglial states along a continuum can be described in functional terms. Accordingly, the labels “M2a-like,” “M2b-like,” and “M2c-like” denote resemblance to particular functional programs rather than stable, mutually exclusive microglial subtypes.

M2a-like microglial state

M2a-associated responses were among the earliest alternatively activated states defined under experimental conditions. Compared with the pro-inflammatory profile traditionally assigned to classically activated M1 microglia, M2a-associated responses are linked mainly to tissue repair and control of excessive inflammation. They are induced primarily by the T helper 2 cytokines interleukin-4 (IL-4) and interleukin-13 (IL-13) [94]. Under experimentally defined conditions, IL-4/IL-13 exposure, signal transducer and activator of transcription 6 (STAT6)-associated signaling, and a concordant molecular profile support classification as an M2a-like state. CD206 and CD163 are commonly associated with this response [95].

In inflammatory settings, IL-4 and IL-13 availability can increase. IL-4 binds a receptor complex containing IL-4 Rα and the common γ chain (γc), whereas IL-13 signals through an IL-13 Rα1–IL-4 Rα heterodimer [96, 97]. Binding of IL-4 or IL-13 activates the Janus kinase (JAK)–STAT6 pathway [98–100] and engages several related signaling modules, including insulin receptor substrate (IRS)–phosphoinositide 3-kinase (PI3K)–protein kinase B (AKT) [101, 102], mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) [102], and peroxisome proliferator-activated receptor γ (PPARγ) [103], thereby promoting the acquisition of M2a-associated features.

Following induction through these pathways, M2a-associated microglia support oligodendrocyte differentiation through trophic and microenvironmental mechanisms. IL-4/IL-13-associated responses are linked to Olig2 and Sox10 expression and favor the progression of OPCs toward immature and mature oligodendrocytes [104]. In vivo studies likewise associate IL-4-induced, CD206-positive microglial responses with increased oligodendrocyte differentiation [105]. These findings support an M2a-associated repair program.

In addition, M2a-associated microglia can secrete neurotrophic factors, including BDNF and insulin-like growth factor 1 (IGF-1) [106]. BDNF binds to tropomyosin receptor kinase B (TrkB) expressed on OPCs and oligodendrocytes, thereby activating downstream MAPK/ERK signaling. This pathway increases the expression of myelin proteins such as MBP, PLP, and MAG and has been implicated in oligodendrocyte differentiation and subsequent myelination, while BDNF also supports neuronal survival [104, 107, 108]. Consistent with these findings, conditional deletion of TrkB in the oligodendrocyte lineage reduces myelin protein expression and results in thinner myelin sheaths [109], indicating that BDNF not only provides trophic support but also contributes directly to myelin formation. IGF-1 acts through the insulin-like growth factor 1 receptor (IGF1R) on oligodendrocyte-lineage cells to activate PI3K–AKT and MAPK/ERK signaling. PI3K–AKT signaling promotes OPC survival, limits apoptosis, and can further engage the mTOR pathway to support myelin protein synthesis [110]. In parallel, MAPK/ERK signaling promotes cAMP response element-binding protein (CREB) phosphorylation and contributes to oligodendrocyte development and differentiation [111, 112]. Similarly, conditional deletion of IGF1R in the oligodendrocyte lineage reduces the numbers of OPCs and mature oligodendrocytes and impairs normal myelination [113]. Together, the BDNF–TrkB and IGF-1–IGF1R axes provide a mechanistic basis for M2a-associated support of oligodendrocyte survival, differentiation, and myelination.

Further evidence comes from in vitro studies in which IL-13-induced M2a-like microglia increased the expression of myelin proteins, including MBP and MOG, and enhanced the differentiation of OPCs into mature oligodendrocytes [14]. Similarly, Li et al. found that extracellular vesicles (EVs) derived from IL-4-induced M2-like microglia promoted OPC proliferation and differentiation. These vesicles were enriched in miR-23a-5p, which directly targeted Olig3 in luciferase reporter assays. Because Olig3 can inhibit oligodendrocyte differentiation under specific conditions, its suppression may contribute to the observed effects on oligodendrogenesis, remyelination, and white matter repair [62].

In addition to supporting oligodendrocyte-lineage development, M2a-associated responses create conditions favorable to myelin repair by limiting excessive neuroinflammation. M2a-like microglia release mediators such as IL-10 and transforming growth factor-β (TGF-β), restrain strongly pro-inflammatory responses, and reduce IL-1β and tumor necrosis factor-α (TNF-α) production [67, 95]. These effects reduce inflammatory stress on OPCs and mature oligodendrocytes and make the local environment more permissive for differentiation and myelination.

Studies also show that M2-like microglia recognize and engulf myelin debris through receptors such as CD36 and CD206 [114]. This process may interrupt the cycle in which uncleared myelin debris sustains inflammation and inhibits OPC differentiation. Whether this function is specific to M2a-like microglia remains uncertain. CD206-positive phagocytes cannot be classified as M2a on this basis alone, particularly because debris clearance is shared by several reparative myeloid states.

From a functional perspective, the trophic and differentiation-supporting programs associated with M2a-like responses are particularly relevant during the initiation of remyelination. At this stage, OPCs require a permissive microenvironment and appropriate cues to progress toward mature oligodendrocytes. Experimental demyelination studies link repair-associated microglial and macrophage responses to enhanced oligodendrocyte differentiation [14]. However, these findings do not establish a discrete activation window for M2a-associated microglia in vivo, and whether a comparable temporal pattern occurs during VaD-related white matter injury remains unclear.

In summary, current evidence indicates that IL-4 or IL-13 can induce a trophic microglial response that supports OPC differentiation and oligodendrocyte maturation. This effect is shown most directly in microglial cultures exposed to defined stimuli. In vivo changes in CD206 or CD163 expression may be compatible with an M2a-associated reparative response, but do not by themselves clarify the functional state of the responding cells. However, whether these IL-4/IL-13-associated mechanisms are similarly engaged during VaD-related white matter injury requires further direct investigation (Fig. 5) (Table 1).

Fig. 5.

Mechanisms through which M2a-like microglia support oligodendrogenesis and myelin repair. IL-4 and IL-13 promote an M2a-like microglial response primarily through JAK–STAT6 signaling, with additional contributions from the IRS–PI3K–AKT, MAPK/ERK, and PPARγ pathways. (A) BDNF and IGF-1 released by M2a-like microglia act on TrkB and IGF1R, respectively, in oligodendrocyte-lineage cells. BDNF–TrkB signaling primarily activates MAPK/ERK, whereas IGF-1–IGF1R signaling engages the PI3K–AKT–mTOR and MAPK/ERK–CREB pathways. These pathways support OPC survival and oligodendrocyte differentiation and increase the expression of myelin-associated proteins, including MBP, PLP, and MAG. (B) Extracellular vesicles derived from M2-like microglia may transfer miR-23a-5p to OPCs, suppressing Olig3 and reducing its inhibitory effect on OPC differentiation. (C) IL-10 and TGF-β released by M2a-like microglia restrain pro-inflammatory microglial responses and reduce the production of IL-1β and TNF-α, thereby limiting inflammation-associated myelin injury. (D) M2a-like microglia may also participate in myelin-debris clearance through phagocytic receptors such as CD36 and CD206, although this function is not specific to the M2a-like state. Dashed arrows indicate proposed or indirectly supported relationships

Table 1.

Comparison of M2-like microglial functional programs

M2-like	Typical inducing stimuli	Associated markers*	Predominant functional emphasis	Overlapping functions	Temporal relevance	Strength of evidence
M2a	IL-4 and IL-13	CD206/MRC1, Arg1, Ym1/Chil3, CCL17, and CCL22	Trophic support and promotion of OPC differentiation and oligodendrocyte maturation	Also contributes to inflammatory control, maintenance of a reparative microenvironment, and, to some extent, clearing damaged material	Its trophic and differentiation-supporting functions may be particularly relevant during the initiation of remyelination [14]	Moderate evidence: direct support from defined microglial cultures and selected in vivo models; several in vivo studies remain marker-based, and direct VaD evidence is limited [14, 94, 105]
M2b	Immune complexes combined with TLR or IL-1 R ligands	CCL1; often accompanied by the expression of IL-10, IL-6, and IL-1β	Immunoregulation and remodeling of inflammatory responses	IL-10-associated immunoregulation exerts inflammation-limiting and tissue-protective effects	In an intracerebral hemorrhage model, CCL1 increased later than M2a- and M2c-associated markers, suggesting a relatively delayed response [115]	Limited evidence: derived mainly from experimentally induced macrophage or microglial responses and mechanistic extrapolation; direct validation in VaD is lacking [66, 116, 117]
M2c	IL-10, TGF-β, and glucocorticoids	CD163, CD206/MRC1, MerTK, TREM2, and Axl	Clearance of myelin debris, lipid handling, inflammation resolution, and tissue remodeling	Also provides trophic support and facilitates oligodendrocyte repair by improving the local microenvironment	Its debris-clearing and lipid-handling functions may remain relevant during the subacute and chronic phases of white matter remodeling [63, 64]	Relatively strong evidence: supported by both in vitro studies and multiple in vivo models of white matter injury, although subtype-specific evidence in VaD remains limited [63, 65, 118, 119]

Assignment of an M2-like program requires concordance among inducing conditions, signaling pathways, molecular profiles, and functional readouts. The listed markers are supportive rather than diagnostic and cannot independently establish cellular origin

M2c-like microglial state

M2c-like microglia are commonly associated with exposure to immunoregulatory stimuli, including IL-10, TGF-β, and glucocorticoids. IL-10 activates the JAK1/TYK2–STAT3 pathway, TGF-β signals through Smad2/3, and glucocorticoids act through the glucocorticoid receptor, thereby promoting an M2c-like state [120, 121]. This classification is reinforced by a concordant molecular signature encompassing CD163 and CD206, phagocytic receptors including MER receptor tyrosine kinase (MerTK), triggering receptor expressed on myeloid cells 2 (TREM2) and Axl [95, 122].

The induction of M2c-associated features involves signaling pathways such as JAK1–STAT3 and Smad2/3. After binding to IL-10 R on the surface of microglia, IL-10 activates JAK1 and TYK2, leading to STAT3 phosphorylation, dimerization, and nuclear translocation [120]. TGF-β activates Smad2/3 signaling, which can cooperate with STAT3-dependent transcription and reinforce an inflammation-resolving program [121]. During white matter injury, accumulating myelin debris engages phagocytic receptors and reinforces a damage-clearance response [123]. This creates a functional cycle in which tissue damage promotes phagocytic activity and debris removal.

Compared with the predominantly trophic profile attributed to M2a-associated responses, M2c-associated microglia generally display stronger phagocytic and debris-clearing properties. This activity involves receptors such as MerTK, TREM2, and Axl [118, 123, 124]. Among these, TREM2 participates in the recognition and processing of lipid-rich myelin debris and in the regulation of microglial responses to white matter injury. Following ligand recognition, TREM2 signals through DNAX-activating protein of 12 kDa (DAP12) and spleen tyrosine kinase (Syk), thereby enhancing microglial phagocytic activity [63, 125]. Studies show that TREM2 deficiency impairs myelin-debris clearance, increases debris accumulation, and aggravates myelin degeneration [126].

The importance of TREM2 in myelin remodeling extends beyond debris internalization. Myelin is rich in cholesterol and sphingolipids, and extensive phagocytosis places a substantial metabolic burden on microglia. Inadequate processing of internalized lipids may lead to intracellular lipid accumulation, cellular foaming, chronic inflammatory activation, and subsequent microglial dysfunction. TREM2-related signaling, including the phospholipase C γ2 (PLCγ2)–protein kinase C (PKC) axis, may facilitate lipid processing and limit the adverse effects of metabolic stress [127]. TREM2-mediated phagocytosis may also cooperate with MerTK and Axl. In an AD mouse model, Savage et al. found that PPARγ and retinoid X receptor (RXR) signaling increased TREM2, MerTK, and Axl expression and enhanced myeloid-cell phagocytosis [128]. Although the study used an AD rather than a VaD model, the findings suggest that lipid sensing and nuclear-receptor signaling may regulate TREM2, MerTK, and Axl during chronic injury. Changes in the expression of these receptors may influence phagocytosis and intracellular lipid processing.

M2c-associated microglia also influence oligodendrocyte generation and myelin repair. In a Theiler’s murine encephalomyelitis virus (TMEV)-induced demyelination model, conditioned medium from in vitro-polarized M2c-like microglia increased oligodendrocyte-lineage generation from NSCs and promoted nuclear β-catenin accumulation. These findings implicate canonical Wnt/β-catenin signaling [119]. Mechanistically, Wnt7a binds to Frizzled receptors and the low-density lipoprotein receptor-related proteins 5 and 6 (LRP5/6) co-receptors on NSCs, leading to the recruitment and activation of Dishevelled and subsequent inhibition of the β-catenin destruction complex composed of Axin, adenomatous polyposis coli (APC), glycogen synthase kinase 3β (GSK3β), and casein kinase 1 (CK1) [129, 130]. The resulting reduction in β-catenin phosphorylation and proteasomal degradation allows β-catenin to accumulate in the cytoplasm and translocate to the nucleus, where it associates with T-cell factor/lymphoid enhancer factor (TCF/LEF) transcription factors to regulate gene programs involved in oligodendroglial lineage commitment and progression [119, 131]. Similar to M2a-associated microglia, M2c-like cells also produce BDNF and IGF-1, which support oligodendrocyte differentiation, maturation, and remyelination [122, 132]. Because this experiment used conditioned medium from microglial cultures exposed to defined M2c-inducing conditions, it provides relatively direct evidence that an experimentally induced microglial program can regulate oligodendroglial differentiation.

During inflammation resolution and tissue remodeling, IL-10- and TGF-β-associated responses suppress persistent inflammatory signaling and restrain strongly pro-inflammatory features [133]. These responses limit pro-inflammatory microglial activation within the CNS. Their immunoregulatory activity also reduces residual inflammatory cues and establishes a more quiescent microenvironment that favors oligodendrocyte differentiation and myelin repair.

In summary, M2c-like microglial responses are defined by stimulation with IL-10, TGF-β, or glucocorticoids, accompanied by STAT3- or Smad-dependent signaling, a corresponding molecular profile involving CD163, CD206, MerTK, TREM2, and Axl, and functional evidence of phagocytosis or inflammation resolution. Their clearance- and lipid-handling functions may be especially relevant during the subacute and chronic phases of white matter remodeling, when prolonged processing of cellular and myelin-derived debris is required [63, 64, 127]. However, these findings should not be interpreted as indicating that M2c-related responses are initiated only during late stages of injury. Rather, they highlight the potential importance of clearance and lipid-processing functions during tissue remodeling. The specific contribution of these pathways to M2c-like responses in VaD remains to be further clarified (Fig. 6) (Table 1).

Fig. 6.

Mechanisms through which M2c-like microglia support oligodendrogenesis and myelin repair. IL-10 and TGF-β induce M2c-like responses through JAK1/TYK2–STAT3 and Smad2/3 signaling, respectively. (A) M2c-like microglia secrete Wnt7a, which activates canonical wnt signaling in NSCs. Signaling through Frizzled/LRP5/6 and dishevelled inhibits the axin–APC–GSK3β–CK1 destruction complex, allowing β-catenin stabilization, nuclear accumulation, and TCF/LEF-dependent transcription. In NSC cultures, Wnt7a favored oligodendroglial lineage specification. (B) BDNF and IGF-1 provide trophic signals that support OPC survival and differentiation and facilitate progression toward myelinating oligodendrocytes. (C) IL-10 and TGF-β restrain persistent pro-inflammatory microglial responses, thereby reducing inflammation-associated myelin injury. (D) TREM2, MerTK, and Axl contribute to the recognition and removal of myelin debris. TREM2–DAP12–syk signaling promotes phagocytic uptake, whereas TREM2-associated PLCγ2–PKC signaling supports intracellular lipid processing after debris internalization. PPARγ/RXR signaling can further increase the expression of phagocytic receptors and reinforce this clearance program

M2b-like microglial state

Within the classical M1/M2 framework, M1 microglia are generally viewed as pro-inflammatory, whereas M2 microglia are commonly described as anti-inflammatory. This binary classification does not adequately capture the functional diversity of activated microglia. M2a-associated responses are mainly linked to trophic support and tissue repair, M2c-associated responses to phagocytic clearance and inflammation resolution, and M2b-associated responses to a mixed immunoregulatory profile. M2b-associated responses therefore cannot be classified simply as an “anti-inflammatory phenotype” [134].

An experimentally defined M2b-like state is typically induced by combined stimulation with immune complexes and Toll-like receptor (TLR) or IL-1 receptor ligands, rather than by a single cytokine signal. This induction depends on cooperation between Fc gamma receptor (FcγR)-mediated and TLR/IL-1 R-associated pathways, with FcγR–Syk–PI3K/AKT and TLR/IL-1 R–myeloid differentiation primary response 88 (MyD88)–nuclear factor κB (NF-κB)/MAPK representing two commonly implicated signaling axes [115, 135]. Consistent with this dual-input mechanism, M2b-associated microglia exhibit a mixed cytokine profile. They release IL-10, which may restrain excessive inflammatory activation, while retaining the capacity to produce mediators such as IL-1β and IL-6 [66]. M2b-associated responses are therefore not purely anti-inflammatory but are more appropriately viewed as immunoregulatory, combining inflammation-limiting activity with preservation of a controlled immune response. Accordingly, assignment of an M2b-like state should ideally require evidence of the defined immune-complex–TLR/IL-1 R induction context, activation of FcγR–Syk- and MyD88-associated signaling, and a corresponding mixed cytokine profile.

Studies examining the temporal dynamics of different M2-associated microglial functional states remain limited. In an intracerebral hemorrhage model, Ohnishi et al. found that several M2a- and M2c-associated markers were elevated as early as 1 day after injury, whereas a clear increase in the M2b-associated marker CCL1 was not observed until day 3 [115]. In this acute injury model, M2b-associated responses appeared later than the measured M2a- and M2c-associated responses. However, the substantial overlap and limited specificity of currently available markers make it difficult to distinguish these functional states reliably over time. Consequently, longitudinal studies comparing their temporal dynamics remain scarce; therefore, no fixed sequence of transitions among M2-associated states has yet been established.

Direct evidence that M2b-associated microglia regulate oligodendrocyte differentiation remains scarce. Given their distinctive dual pro- and anti-inflammatory properties, M2b-like responses are best viewed as modulators of the repair microenvironment rather than as purely anti-inflammatory or reparative states. Two mechanisms are particularly relevant.

First, IL-10 reduces inflammatory stress on oligodendrocyte-lineage cells. IL-10 suppresses the production of pro-inflammatory mediators and limits strongly inflammatory microglial responses [116, 117]. Laffer et al. showed in IL-10-deficient mice that loss of IL-10 shifted microglia toward an M1-like phenotype and exacerbated inflammation in the CNS [67]. Another study showed that activation of the IL-10/IL-10R1 signaling pathway favors an M2-like microglial state while suppressing the acquisition of M1-like features [136]. Through these effects, an IL-10-rich immunoregulatory environment can protect OPCs and mature oligodendrocytes and create conditions favorable to myelin repair. Second, the ability of M2b-like microglia to secrete IL-6 provides another potential route through which they may influence oligodendrocyte-lineage cells. IL-6 signals either through membrane-bound IL-6 R or through soluble IL-6 R, which activates glycoprotein 130 (gp130)-expressing cells. OPCs express gp130 and can therefore respond to IL-6/sIL-6 R complexes [137]. Valerio et al. found that an IL-6/sIL-6 R fusion protein enhanced the survival and differentiation of purified rat oligodendroglial-lineage cells through gp130-associated signaling [138]. Zhang et al. further reported that OPCs pretreated with IL-6/sIL-6 R generated more MBP-positive myelin structures after transplantation into an experimental demyelination model [139]. These studies show that IL-6 trans-signaling can regulate oligodendrocyte-lineage cells. However, they do not establish a direct link between M2b-like microglia and these IL-6-mediated effects. The proposed M2b-derived IL-6–OPC axis in VaD therefore remains unproven and requires direct testing using microglia–OPC co-culture systems and cell-specific tracing.

In summary, available evidence suggests that M2b-like responses may influence myelin repair mainly by regulating the inflammatory environment. Experimental studies support an IL-10-mediated immunoregulatory role, whereas the IL-6–OPC pathway has not yet been demonstrated. However, current understanding of M2b-like responses is largely based on macrophage systems or non-VaD inflammatory models. Whether similar responses occur in VaD-associated white matter injury and how they affect oligodendrocyte-lineage cells remain important questions for future investigation (Fig. 7) (Table 1).

Fig. 7.

Immunoregulatory mechanisms linking M2b-like microglia to oligodendrocyte and myelin repair. M2b-like microglia are induced by combined stimulation through immune complex–FcγR and TLR/IL-1 R pathways. FcγR engagement activates syk–PI3K/AKT signaling, whereas TLR/IL-1 R stimulation recruits MyD88 and activates NF-κB/MAPK pathways. The resulting state has a mixed immunoregulatory profile, with the production of both pro-inflammatory mediators, including IL-1β, IL-6, and TNF-α, and the anti-inflammatory cytokine IL-10. (A) IL-10 may restrain pro-inflammatory microglial responses, reduce inflammation-associated oligodendrocyte injury, and help establish a microenvironment permissive for repair. (B) a proposed mechanism involves IL-6 trans-signaling. IL-6 binds soluble IL-6 R to form an IL-6/sIL-6 R complex, which may activate gp130 on OPCs and thereby support OPC survival and differentiation, as well as myelin protein expression. This pathway remains hypothetical in the context of M2b-like microglia and has not been directly validated in vascular dementia models. Dashed arrows indicate proposed or indirectly supported relationships

Therapeutic strategies directed at M2-associated microglial functions

M2a-, M2b-, and M2c-like states represent overlapping functional programs involving trophic support, immunoregulation, and phagocytic activity. These observations suggest that future interventions may need to consider the functional characteristics and disease context of microglial responses. Current studies have mainly examined broad M2-associated responses, and evidence for stage-specific modulation of these programs remains limited. This review summarizes compounds reported to influence M2-associated microglial functions and discusses their potential relevance to white matter repair in VaD. Although many of these strategies have not yet been directly evaluated in VaD models, they may provide potential directions for developing future therapeutic approaches for VaD.

Enhancing M2c-associated phagocytosis and clearance

M2c-associated programs emphasize myelin-debris recognition, phagocytosis, and lipid handling through receptors including MerTK and TREM2. Enhancing these functions can reduce debris-mediated inhibition of OPC differentiation and create a more permissive environment for remyelination.

Pentoxifylline (PTX) is a nonspecific phosphodiesterase inhibitor and a synthetic derivative of theobromine [140, 141] and was included for clinical use in the Chinese Pharmacopoeia in 2010 [64]. Zheng et al. used bilateral common carotid artery stenosis (BCAS) to establish a mouse model of chronic cerebral hypoperfusion and administered PTX intraperitoneally. PTX attenuated white matter injury in the corpus callosum and improved cognitive function, potentially by activating PPARγ, upregulating MerTK expression, and enhancing microglial clearance of myelin debris [64].

TREM2 agonists offer another approach to enhancing M2c-associated clearance functions. TREM2 activation promotes microglial recognition and phagocytosis of myelin debris. In a 5XFAD AD model, Wang et al. found that acute systemic administration of the human TREM2 agonist antibody AL002c increased microglial proliferation-associated signaling, whereas long-term administration reduced neuroinflammatory adverse effects and improved behavioral performance [142]. In a cuprizone-induced demyelination model, the murine TREM2 agonist AL002a enhanced microglial uptake and degradation of myelin debris. Immunofluorescence showed increased degraded MBP (dMBP) within lysosomal structures in CD68⁺ phagocytes. Pretreatment with AL002a also enhanced phagocytic clearance by bone marrow-derived macrophages from Trem2⁺ /⁻ mice [65]. These findings support TREM2 agonism as a means of enhancing myelin-debris clearance. Its efficacy and safety in VaD remain to be established.

Enhancing M2a-associated differentiation and repair

IL-4 and IL-13 are canonical inducers of M2a-associated microglial responses [94]. They bind receptor complexes containing IL-4 Rα or IL-13 Rα1 [96] and activate downstream pathways, including JAK/STAT6 [96, 98] and PI3K/AKT [101]. These pathways promote the production of trophic factors such as BDNF and IGF-1 [106], thereby supporting OPC proliferation, differentiation, and white matter repair.

Engineered IL-4 and IL-13 agonists can improve ligand stability, receptor affinity, or signaling potency. For example, Junttila et al. engineered a type I IL-4 receptor-selective superkine with approximately 3,700-fold higher affinity for the γc chain and three- to ten-fold greater cellular potency than wild-type IL-4 [143]. Using structural grafting and directed evolution, Yang et al. developed the de novo IL-4 mimetic Neo-4. Neo-4 reproduced key IL-4 signaling properties while showing greater stability and potentially improving its manufacturability and translational utility [144]. Oshima et al. generated the high-affinity IL-13 variant IL-13 R112D by substituting aspartic acid for arginine at position 112. Compared with wild-type IL-13, IL-13 R112D exhibited a five- to ten-fold increase in receptor-binding affinity and approximately ten-fold greater biological potency in cell-based assays [145]. These engineered ligands provide tools for testing sustained IL-4/IL-13 signaling. Their capacity to induce beneficial M2a-like microglial programs and repair ischemic white matter, however, requires direct validation in relevant demyelination models.

M2-like microglia-derived EVs (M2-EVs) represent a distinct delivery strategy. In a cerebral ischemia model, Li et al. found that M2-EV-enriched miR-23a-5p promoted the proliferation and differentiation of OPCs and accelerated myelin regeneration and white matter repair [62]. Qin et al. further found that miR-23a-3p in human umbilical cord mesenchymal stem cell-derived exosomes promoted oligodendrocyte differentiation and myelin-protein expression through PI3K/AKT and Tbr1/Wnt signaling [146]. These studies provide preclinical support for extracellular-vesicle-mediated repair. However, their cell-state specificity, biodistribution, manufacturing consistency, and efficacy in VaD remain unresolved.

Modulating M2b-associated immunoregulation: IL-10 inducers

Standardized in vitro models of M2b-like microglia remain difficult to establish because induction depends on combined immune-complex and innate-receptor signaling. Their mixed inflammatory and immunoregulatory profile also complicates functional targeting. Accordingly, therapies directed specifically at M2b-like responses remain largely unexplored. In AD and stroke models, Sun et al. found that AD110 increased IL-10 expression and reduced brain injury. The compound also increased CD206, a marker associated with M2a- and M2c-like responses [147]. AD110 therefore cannot be considered a specific M2b-directed therapy. Its effects are more consistent with broad modulation of M2-associated programs, although its IL-10-inducing activity remains relevant to immunoregulatory strategies. Direct validation in VaD and cell-specific analysis are required before this mechanism can be assigned to M2b-like microglia (Table 2).

Table 2.

Candidate interventions affecting M2-associated microglial functions relevant to VaD

Drug	Associated microglial program	Mechanism of Action	Relevance to VaD	References
PTX	Phagocytic clearance	Activates PPARγ, upregulates MerTK expression, and enhances the phagocytic clearance of myelin fragments by microglia	Validated in a chronic cerebral hypoperfusion model, with reduced white matter injury and improved cognition	[64, 140, 141]
TREM2 agonists (AL002c, AL002a)	Phagocytic clearance	Enhances TREM2 receptor clustering, activates the DAP12-Syk signaling axis, and improves microglial recognition and phagocytosis of myelin fragments	Supported in demyelination and AD models; not directly validated in VaD	[65, 142]
IL-4/IL-13 Superagonists	Trophic signaling	Engineered ligands with increased affinity, stability, or signaling potency; activate IL-4/IL-13 receptor pathways, including STAT6	Not validated in VaD; effects on microglial states and white matter repair require direct testing	[143–145]
M2-EVs	Trophic delivery	Enriched in miR-23a-5p, which specifically inhibits Olig3, lifts the suppression of OPC differentiation, and promotes myelin regeneration	Preclinical support in cerebral ischemia models; no direct validation in VaD	[62]
AD110	Broad immunoregulation	Selectively promotes IL-10 release while upregulating CD206 (M2a/M2c marker); no significant effect on pro-inflammatory factors	Not validated in VaD; supports IL-10 induction but not M2b-specific targeting	[147]

Discussion and outlook

VaD, the second most common form of dementia after AD, remains a major public health challenge. Its onset and progression are closely associated with chronic cerebral hypoperfusion, cerebral small vessel disease, BBB disruption, and white matter injury [1, 4]. Myelin repair after white matter injury is not mediated by a single cell type or signaling pathway; rather, it emerges from coordinated interactions among microglia, astrocytes, endothelial cells, pericytes, and oligodendrocyte-lineage cells within the local white matter microenvironment [83, 148]. Within this network, microglia participate in inflammatory regulation, myelin-debris clearance, and the support of oligodendrocyte differentiation. The effects of microglia on white matter repair are therefore likely to change with the local environment and the stage of injury.

Earlier studies commonly used the M1/M2 dichotomy to describe the functional states of microglia. Although this framework provided a useful basis for understanding microglial function, advances in single-cell sequencing and spatial omics have increasingly revealed its limitations in capturing the complexity of microglial responses in neurological disease. M2a, M2b, and M2c were initially defined as functional phenotypes under specific experimental conditions. These states may be associated with trophic support, immune regulation, and phagocytic clearance, respectively, but the boundaries between them are not absolute [149, 150]. These functions are not mutually exclusive. They may coexist within the same lesion, share molecular features, and vary in prominence as white matter injury progresses.

IL-4/IL-13–STAT6-associated M2a-like responses are mainly linked to trophic support, regulation of the inflammatory milieu, OPC differentiation, and oligodendrocyte maturation [94, 95, 105]. IL-10/TGF-β–STAT3/Smad-associated M2c-like responses are more closely associated with myelin-debris clearance, intracellular lipid processing, resolution of persistent inflammation, and support for oligodendroglial differentiation [120, 121]. M2b-like responses, typically induced by combined immune-complex and TLR or IL-1 receptor stimulation, appear to modulate inflammation through the coordinated production of IL-10 and selected pro-inflammatory mediators [66, 115, 135]. Their influence on oligodendrocyte-lineage cells appears to be less direct and has been studied less extensively. These responses do not necessarily occur in a fixed sequence, and their relative contributions may change as tissue pathology develops from active injury and debris accumulation to oligodendrocyte differentiation, remyelination, and longer-term remodeling.

Reliable assignment of an M2a-, M2b-, or M2c-like state requires more than the detection of one or two markers. Ideally, the inducing conditions, activated signaling pathways, molecular profile, and functional output should point in the same direction [150, 151]. State-associated molecules are most informative when considered in relation to their biological roles. CD206 and CD163 may support the presence of a repair-associated response, but neither marker is sufficient to identify a specific functional state [150, 152, 153]. CCL1, together with a mixed IL-10, IL-6, and IL-1β profile, is more compatible with an M2b-like immunoregulatory program [66, 154]. MerTK, TREM2, and Axl are more directly related to phagocytosis, efferocytosis, and lipid handling [64, 118, 127, 128]. By contrast, CCR2 is primarily informative about recruitment from the circulating monocyte pool [155]. Marker expression is therefore most useful when interpreted together with the induction context, signaling pathways, and observed function.

A separate but related issue is cellular origin. Although resident microglia remain the primary parenchymal myeloid population considered in this review, CNS border-associated macrophages [152, 153, 156] and recruited monocyte-derived macrophages [157] can express overlapping repair-associated markers and engage similar clearance or immunoregulatory functions at vascular and lesion interfaces. Accordingly, purified microglial cultures can establish the cell-intrinsic capacity to perform a given function, whereas lineage tracing and appropriately validated microglia-restricted genetic approaches provide stronger in vivo attribution than marker-based analyses of mixed tissue [150, 156]. Functional state and cellular origin should therefore be evaluated separately.

Together, these issues help explain why direct evidence separating M2a-, M2b-, and M2c-associated functions in white matter repair remains limited. Many studies continue to use “M2-like microglia” as a broad category, often on the basis of CD206, IL-10, or a small panel of related molecules. Such evidence may indicate engagement of a repair-associated program, but it does not establish the underlying signaling context or dominant biological function. Moreover, much of the available experimental evidence has not been derived directly from VaD models, but instead from models of AD, experimental autoimmune encephalomyelitis (EAE), and ischemic stroke (IS). Although these conditions may all involve microglial activation, myelin damage, and white matter alterations, their pathological processes do not fully correspond to those of VaD. EAE is commonly used to model chronic inflammation and immune-mediated demyelination [158]; IS results from sudden focal cerebral ischemia [159]; and AD is characterized primarily by amyloid deposition and tau-related pathology [160]. By contrast, VaD often develops in the setting of prolonged cerebral hypoperfusion and is influenced by aging, hypertension, diabetes, and other vascular risk factors [4, 161]. These differences affect lesion distribution, BBB integrity, the duration of injury, and the composition of the myeloid response.

These differences are also likely to influence the timing and persistence of microglial responses. In chronic cerebral hypoperfusion models, white matter injury and glial responses generally develop over several days to weeks after surgery [162], whereas focal lesions form more rapidly in IS models and may be accompanied by early recruitment of peripheral immune cells [157]. These temporal differences may influence how microglial states emerge and how oligodendrocyte-lineage cells respond to local signals. Other disease models therefore provide mechanistic clues but cannot define the corresponding sequence in VaD. Therefore, conclusions regarding specific microglial programs should be considered provisional until they are validated in VaD-relevant models. VaD-related studies support roles for inflammatory regulation, myelin-debris clearance, and oligodendrocyte support after white matter injury. Direct evidence for the timing and relative contributions of M2a-, M2b-, and M2c-associated programs remains limited. Future work should prioritize models that better reproduce chronic hypoperfusion and VaD-related white matter pathology. Single-cell sequencing and spatial transcriptomics can resolve microglial states and their distribution within white matter lesions [163], but transcriptional identity alone does not establish function. Combining these approaches with lineage tracing, cell-specific genetic manipulation, and microglia–OPC co-culture may help determine how defined pathways affect oligodendrocyte survival, differentiation, and remyelination.

From a translational perspective, several issues surrounding therapeutic strategies directed at M2-associated microglial functions remain unresolved. Drugs administered peripherally must first cross the BBB to reach the CNS, and different therapeutic modalities vary considerably in their ability to do so. Although BBB disruption is common in VaD and related cerebral small vessel disease, its severity and spatial distribution vary across patients and disease subtypes [164, 165]. The BBB therefore remains a substantial obstacle to effective CNS drug delivery. Small molecules and biologic modalities, including antibodies, cytokine-based agents, and EVs, differ substantially in systemic stability, BBB penetration, tissue distribution, and brain exposure [166]. Some agents may also be taken up or cleared by peripheral tissues before reaching the CNS. For TREM2-activating antibodies and cytokine-based agents in particular, it will be important to determine their actual concentrations within the brain, their distribution across affected regions, and the duration of their biological effects. Future studies should therefore incorporate pharmacokinetic and biodistribution analyses to establish whether candidate agents reach the CNS and achieve sufficient exposure within the intended target regions.

Therapeutic timing is likely to shape treatment outcomes. VaD-related white matter injury evolves across stages rather than as a single event. Across experimental models, the relative contributions of inflammatory, phagocytic, and regenerative microglial programs appear to vary with the stage of demyelination [167]. Broad suppression of inflammatory activity during the early phase may interfere with myelin-debris clearance and the initiation of oligodendrocyte repair [79]. Conversely, once myelin debris has accumulated and OPC differentiation is impaired, anti-inflammatory signaling alone may be insufficient to restore repair unless debris clearance is also improved [168]. Interventions targeting clearance, immunoregulation, or trophic support should therefore be matched to disease stage. Their optimal timing, sequence, and duration remain to be established.

In addition, IL-4, IL-13, IL-10, TREM2, MerTK, and PPARγ are not specific to microglia. Systemic administration may also affect peripheral myeloid cells and other immune-cell populations, potentially leading to immune imbalance or other adverse effects [118, 124, 150]. Future preclinical and clinical studies should therefore evaluate not only white matter repair and cognitive outcomes, but also changes in peripheral immunity and long-term safety.

In summary, inflammatory regulation, trophic support, and myelin-debris clearance coupled with intracellular lipid processing are the principal microglial functions shaping oligodendrocyte responses and myelin repair across experimental models. Their relative contributions are likely to vary across disease stages. This function- and stage-oriented framework provides a more defensible basis for therapeutic design than attempts to induce a nominal M2 subtype. Current understanding remains limited by differences among disease models, uncertain boundaries between microglial states, and the scarcity of VaD-specific evidence. Before these mechanisms can be advanced toward clinical application, future studies should integrate VaD-relevant models, single-cell and spatial analyses, cell-specific functional validation, and pharmacokinetic assessment.

Abbreviations

AD

Alzheimer’s disease

AKT

Protein kinase B

APC

Adenomatous polyposis coli

BBB

Blood–brain barrier

BCAS

Bilateral common carotid artery stenosis

BDNF

Brain-derived neurotrophic factor

BMP4

Bone morphogenetic protein 4

CNS

Central nervous system

CREB

cAMP response element-binding protein

CK1

Casein kinase 1

CXCL12

C-X-C motif chemokine ligand 12

CXCR4

C-X-C chemokine receptor type 4

DAP12

DNAX-activating protein of 12 kDa

DCC

Deleted in colorectal cancer

DNMT3A

DNA methyltransferase 3A

EAE

Experimental autoimmune encephalomyelitis

ECM

Extracellular matrix

ERK

Extracellular signal-regulated kinase

EVs

Extracellular vesicles

FcγR

Fc gamma receptor

GalC

Galactocerebroside

GFAP

Glial fibrillary acidic protein

GSK3β

Glycogen synthase kinase 3β

HDAC1/2

Histone deacetylases 1 and 2

IGF-1

Insulin-like growth factor 1

IGF1R

Insulin-like growth factor 1 receptor

IRS

Insulin receptor substrate

IS

Ischemic stroke

JAK

Janus kinase

CD

Cluster of differentiation

L1CAM

L1 cell adhesion molecule

LAMA2

Laminin subunit α2

LGI1

Leucine-rich glioma-inactivated 1

LRP5/6

Low-density lipoprotein receptor-related proteins 5 and 6

MAG

Myelin-associated glycoprotein

gp130

Glycoprotein 130

TGF-β

Transforming growth factor-β

MAPK

Mitogen-activated protein kinase

M2-EVs

M2-like microglia-derived extracellular vesicles

MBP

Myelin basic protein

MerTK

MER receptor tyrosine kinase

MOG

Myelin oligodendrocyte glycoprotein

mTOR

Mechanistic target of rapamycin

MyD88

Myeloid differentiation primary response 88

NF-κB

Nuclear factor κB

NSCs

Neural stem cells

NT-3

Neurotrophin 3

OL

Oligodendrocyte

OPALIN

Oligodendrocytic myelin paranodal and inner loop protein

OPCs

Oligodendrocyte precursor cells

PDGF-A

Platelet-derived growth factor A

PDGFRα

Platelet-derived growth factor receptor α

PI3K

Phosphoinositide 3-kinase

PKC

Protein kinase C

PLCγ2

Phospholipase C γ2

PLP

Proteolipid protein

PLP1

Proteolipid protein 1

PPARγ

Peroxisome proliferator-activated receptor γ

Pre-OL

Pre-oligodendrocyte

PTX

Pentoxifylline

RXR

Retinoid X receptor

STAT3

Signal transducer and activator of transcription 3

STAT6

Signal transducer and activator of Transcription 6

SVZ

Subventricular zone

Syk

Spleen tyrosine kinase

TCF/LEF

T-cell factor/lymphoid enhancer factor

T3

Triiodothyronine

TLR

Toll-like receptor

TMEV

Theiler’s murine encephalomyelitis virus

TrkB

Tropomyosin receptor kinase B

TREM2

Triggering receptor expressed on myeloid cells 2

TYK2

Tyrosine kinase 2

VaD

Vascular dementia

IL

Interleukin

TNF-α

Tumor necrosis factor-α

FGF

Fibroblast growth factor

Author contributions

Zhang Qinyuan, Zhang Bomin, and Wang Yixin: writing—original draft. Cheng Fafeng, Li Changxiang, Wang Xueqian, and Wang Qingguo: writing—review and editing. Wu Yiping, Liu Yanting, Feng Sicheng, and Zhang Chuxin: visualization and writing—original draft. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. U21A20400); Beijing Natural Science Foundation (No. 7232279); the Fundamental Research Funds for the Central Universities (2024-JYB-JBZD-043); CACMS Innovation Fund (CI2023C016LH); the Fundamental Research Funds for the Central Universities (2026-JYB-JBZD-001).

Data availability

Not applicable.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.