Multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD), both immune-mediated demyelinating disorders of the central nervous system (CNS), have been acknowledged as being some of the most debilitating neurological conditions. MS is a chronic immune-mediated demyelinating disorder that primarily affects Europeans, and is uncommon in Asians, Africans, and indigenous Americans [1]. It has been suggested that the etiology of MS is centrally demyelination induced as a consequence of inflammatory cells invading the CNS through blood vessels in the blood-brain barrier (BBB) [2]. These inflammatory cells, in addition to activated microglia and astrocytes, have been shown to contribute to oligodendrocyte injury [1, 3, 4]. The relapsing-remitting phenotype of MS, which accounts for 85-90% of all MS cases, is caused by peripheral inflammation as a result of periodic invasion [4, 5]. In the advanced progressive stage of MS, the repetitive destruction of oligodendrocytes causes persistent oxidative stress, mitochondrial dysfunction, brain iron accumulation, hypoxia, defective remyelination, and glutamate hyperexcitability, eventually leading to neurodegeneration [1, 6]. The clinical manifestations of MS vary based on the area affected, frequent locations being the optic nerve, the cerebral cortex, white matter, brainstem, and spinal cord [1]. Disease-modifying therapies (DMTs), which primarily target inflammatory cascades and have pleiotropic effects, have been the standard therapies for MS [3]. The exact pathophysiology of MS, however, remains unknown, and no specific antibody has been identified and investigated facilitating the detection of MS. As a result, the diagnosis of MS still utilizes the 2017 McDonald criteria which indicate space and time dissemination [7].

Until 2004 NMOSD was considered an MS variant but then a specific antibody was discovered [8]. That antibody is immunoglobulin G (IgG) directed against aquaporin 4 (AQP4), in combination with complement system activation [8, 9]. That antibody causes invasion of the CNS by peripheral granulocytes through the CNS, resulting in astrocytic endfeet damage, severe inflammation, secondary oligodendrocytic injury, and demyelination [9, 10]. The six essential clinical characteristics that must be present in addition to the presence of AQP4 IgG for a diagnosis of NMOSD are longitudinally extensive transverse myelitis (LETM), optic neuritis (ON), area postrema syndrome (APS), brainstem encephalitis, diencephalitis, and telencephalitis [9]. Autoantibody elimination and inflammatory cascade suppression have been the main goals of NMOSD therapy [9, 11, 12].

Extracellular vesicles (EVs) are nano-sized, non-nucleated structures that are hypothesized to facilitate intercellular communication and contribute to autoimmune processes. EVs have been connected to the development and deterioration of a range of pathological conditions, including cancer, cerebrovascular disease, Alzheimer’s disease, and central demyelinating diseases (MS and NMOSD) [13-17]. Many studies have described the potential involvement of EVs as possible biomarkers and as a novel diagnostic approach known as a liquid biopsy for a variety of disorders [18-20]. Furthermore, one of the therapeutic uses of EVs which has the greatest potential is as carrier for the drug delivery system [21]. As a result of those previous studies, the aim of this review was to summarize and discuss the evidence available from in vitro, in vivo, and clinical studies on the role of EVs in central demyelinating diseases, primarily as regards MS and NMOSD, as well as EV therapy strategies for treatment of central demyelinating diseases. Any contradictory findings have been included and considered in this review.

The PubMed database was searched using the keywords: “demyelinating”, “multiple sclerosis”, “neuromyelitis optica”, “exosome”, “extracellular vesicles”, “microvesicles” and “central nervous system” with a demarcation from 2011 to 2024. The search was limited to research articles published in the English language.

EVs are double-layered and non-nucleated vesicles that contain a diversity of cytoplasmic biological substances. Different cell types, in both physiological and pathological stages, secrete distinctive EVs in order to maintain cellular homeostasis and communicate with adjacent cells [22]. Considered as a carrier for biochemical compounds, enzymes, signaling molecules, biogenesis factors, chaperones, cytoskeletal factors, and nucleic acids especially microRNA are contained within EVs. Based on their biogenesis, release mechanisms, size, composition, and function, the three primary subtypes of extracellular vesicles (EVs) are exosomes, microvesicles, and apoptotic bodies [13, 23-26]. Exosomes are the smallest EVs, ranging from 30~100 nm in diameter, and are formed by the inward budding of the endosomal compartment, subsequently being released to the periphery by the exocytotic process. Exosomes usually contain lipids and nucleic acids [24]. Microvesicles, however, are approximately 100~1,000 nm in diameter and since they arise directly from the cell membrane, they typically express transmembrane molecules and glycoproteins depending on their parental cells [25]. Apoptotic bodies are approximately 1,000 to 5,000 nm in diameter and mostly develop in the late stage of apoptotic cells. Phosphatidylserine is often transferred to the outer membrane of apoptotic bodies, which then stimulates phagocytes to remove the dead cells [26]. As previously stated, EVs are involved in cell-to-cell communication; hence, dysfunctional EVs lead to pathogenic processes and promote autoimmunity [27].

Extracted EVs from MS and NMOSD patients were investigated using various methods. However, the interpretation and comparison of the findings across studies must be considered carefully. Different isolation and characterization techniques used in the different studies caused different EV findings [28]. Focusing on the isolation technique, ultracentrifugation (UC) is the most widely used isolation method and is considered the gold standard for EV extraction and separation. UC does not need EV labeling and can avoid cross-contamination. However, this technique is time-consuming and expensive [22]. Polymer precipitation is another approach that uses polyethylene glycol as a medium, the technique involves the incubation of EVs under centrifugation conditions by decreasing the EV solubility. Contrarily, this approach has a high false-positive rate, and the polymer has an adhesive feature that makes it difficult to remove for future EV investigation. Size-based isolation by means of size exclusion chromatography is an alternative technology with a high enrichment efficiency and low cost. The main disadvantage of this technique is its impurity [22, 28]. The different characterization techniques also can have impact on different EV characteristics. Characterization techniques are mainly divided into external characterization (primarily morphology and particle size detection) and internal characterization (membrane protein and lipid raft). The electron microscope is used to detect EV morphology, but the technique is complicated and unsuitable for the rapid testing of many EV samples. Nanoparticle tracking analysis (NTA) is a real-time detection method but carries a high risk of impurity [28]. Camera levels and detection thresholds affect the quality of NTA. Western blotting is a conventional technique that can qualitatively and quantitatively analyze marker proteins. However, the detection of marker proteins varies depending on the type of parental cell, and it is not suitable for the detection of EV marker proteins in biological fluids [28].

Since the quantity of EVs in blood plasma or serum is believed to be between 10⁷ and 10¹² EV/ml, serum EVs have garnered a great deal of attention [29]. However, many investigators have demonstrated the EVs in various samples beyond serum. Different sample types can result in different EV profiles. In contrast with serum, cerebrospinal fluid (CSF) is a type of biological fluid secreted from choroid plexus ependymal cells. CSF surrounds the brain and spinal cord in order to protect the CNS, support homeostasis, and remove waste products. Approximately 80% of CSF EVs originate from the blood, whereas the remaining are produced by nerve cells [30]. However, due to the high integrity of the BBB, the protein level of CSF becomes 100 to 300 times lower than protein found in serum [30, 31]. Therefore, a high amount of CSF is required in order to obtain sufficient EVs for analysis. Albumin and immunoglobulin, the two most abundant proteins in the CSF, may hinder the identification of low-abundance proteins and result in a substantial loss of EV information. As a result, direct comparisons between methods may be difficult according to the advantages and disadvantages of individual techniques. Size, function, sample type, and parental source of EV make isolating and characterization of EV problematic [28, 32].

Extracellular vesicles and pathological correlations in multiple sclerosis

The central paradigm of central demyelinating diseases, especially in the case of MS, is an increase in the inflammatory process and a downregulation of immune tolerance. EVs derived from various origins carry different characteristics and functions.

Extracellular vesicles classified by their cellular origins

Endothelial cell derived extracellular vesicles

The pathophysiology of MS is thought to involve endothelial dysfunction as a crucial factor. The expression of certain markers on the endothelial surface, such as CD31 (PECAM-1), CD51 (integrin), CD54 (ICAM-1), CD62E (ELAM-1 or E-selectin), CD63E, and CD105 (endoglin), increases the adhesion of inflammatory leukocytes to endothelial cells and promotes their transendothelial migration into the CNS [33-37].

It is believed that radiological exacerbation, which refers to an increase in the severity of lesions seen on imaging without the presentation of any clinical symptoms, occurs prior to or simultaneously with clinical exacerbation. Notably, CD31 and CD54, which are localized at endothelial cell intercellular junctions, show high levels of expression during radiological relapse even in the absence of clinical symptoms [34, 36]. Once the exacerbation reaches the clinical exacerbation threshold, endothelial cell damage occurs, leading to the disruption of the BBB. CD31 and CD62E are among the earliest markers to be released, indicating endothelial injury. CD51, along with CD54, which acts as a transmembrane receptor, facilitates leukocyte recruitment and adhesion to endothelial cells via Th1 cell activation [34, 36, 38]. CD51 also plays a role in signaling between the extracellular matrix and cells and has been found to indirectly contribute to nerve regeneration. Following an exacerbation, demyelinated neurons attempt to repair and promote remyelination. The expression of CD51 increases after an attack but remains elevated even after treatment, suggesting that remyelination is a long-term process [39]. However, not all brain regions exhibit the same levels of these endothelial EV markers. Based on the process of neurovascular injury, the subcortical gray matter periventricular region, which is highly vascularized, is the most affected area of the brain [35]. The association between different brain regions may provide additional insights from imaging data for predicting the clinical course and early diagnosis of relapse.

Platelet-derived extracellular vesicles

Platelets have been classified as inflammatory cells due to their multifaceted functions beyond hemostasis. Within platelet-alpha granules reside a wide range of EVs responsible for inflammatory process [40]. Following BBB disruption from impaired endothelial cells and astrocyte foot processes, peripheral inflammatory cells are recruited into the CNS. Platelet-derived EVs have been proposed to play a significant role in leukocyte recruitment and to facilitate transendothelial migration into the CNS, leading to inflammation and demyelination. In patients with MS, CD31+ marker of platelet EV is observed only in the inflammation-predominant stage, RRMS [33, 35]. This phenomenon is not evident in the very early stage, CIS, nor in the advanced stage, SPMS. The postulated explanation is that there is limited degree of inflammation demonstrated in CIS, whereas in SPMS, neurodegeneration rather than neuroinflammation is the primary pathophysiology [41]. However, some level of inflammation persists across all stages of MS evidenced by the heightened expression of CD61 (integrin beta 3), CD62 (P-selectin), and surface IgM on platelets [33, 42, 43]. The alteration of platelet-derived EVs could potentially serve as a marker for identifying MS stages and predicting MS prognosis [33, 35].

Leukocyte-derived extracellular vesicles

Inflammation arises from an imbalance between the increase in proinflammatory leukocytes and the decrease of regulatory T cells (Treg). A recent study has found that specific markers of T cells in CSF, such as CCR3 and CCR5, exhibit high expression during radiological relapse, but not during clinical relapse [44]. It is suggested that activated CD4+ and CD8+ T lymphocytes play a significant role in triggering inflammation. An in vitro study has demonstrated that co-culturing human endothelial cells with T cell EVs leads to a decrease in transendothelial electrical resistance, indicating BBB disruption and astrocyte damage [45, 46]. Furthermore, both in vitro and clinical studies have shown that B cells impact the reduced function of oligodendrocytes [47, 48]. In conjunction with Treg dysfunction, the apoptotic rate of inflammatory leukocytes decreases, resulting in the proliferation of the inflammatory leukocyte population. The excessive accumulation of leukocytes subsequently leads to inflammatory damage within the CNS. However, both T cells and B cells predominantly contribute to the inflammatory stage seen in RRMS. In SPMS, there is a significant decrease in CD14 and CD45 expression, suggesting that neurodegeneration can surpass the extent of neuroinflammation [49].

Microglia/macrophage-derived extracellular vesicles

Microglia, which are considered to be resident macrophages in the CNS, play a crucial role in maintaining self-tolerance and preventing autoimmunity. It has been hypothesized that once leukocytes enter the CNS compartment, microglia may contribute to the promotion of inflammation. In patients with MS, microglial EVs have been found to be significantly upregulated, even in the CIS [50]. During exacerbations, markers associated with microglia/macrophages, such as IB4, showed a significant increase, indicating inflammation mediated by microglia [51]. In an in vitro model, EVs derived from activated microglia/macrophages were shown to impair synaptic function, as evidenced by reduced expression of markers like Syt1 and N1g1 [52]. Additionally, microglia/macrophage EVs might hinder remyelination by inhibiting the accumulation of oligodendrocyte progenitor cells and reducing levels of myelin basic protein (MBP) [53].

Oligodendrocyte-derived extracellular vesicles

Following demyelination, oligodendrocytes play a role in promoting remyelination. Myelin oligodendrocyte glycoprotein (MOG) is secreted by oligodendrocytes in response to demyelination. An elevated level of MOG expression is indicative of increased inflammation. The available evidence suggests that MOG expression in the remitting stage of RRMS is similar to that in healthy individuals but becomes highly expressed during relapsing and progressive stages [54]. Another protein, MBP, also serves as a marker of demyelination. A significant increase in MBP levels in the serum is observed across all stages, whereas healthy individuals do not exhibit such elevation [55]. These findings suggest a potential association between EVs released by oligodendrocytes and the demyelinating response.

Astrocyte-derived extracellular vesicles

Although astrocytes are the primary target of AQP4-IgG in NMOSD, it has also been observed that EVs derived from astrocytes are present in the experimental autoimmune encephalomyelitis (EAE) model of MS. The EVs obtained from EAE mice showed a significantly higher level of glial fibrillary acidic protein (GFAP). Since GFAP is essential for maintaining the cytoskeleton of astrocytes, the increased expression of GFAP in CSF suggests the presence of astrocyte damage and inflammation [56]. However, it is important to note that astrocyte dysfunction may not be the primary mechanism underlying the pathogenesis of MS, even though it is widely accepted as a central process in NMOSD. The subsequent section will provide a more detailed of NMOSD.

Extracellular vesicles categorized by surface membrane expression

Extracellular vesicles with fibrinogen-laden membranes

Fibrinogen-laden EVs are recognized for their significant involvement in inflammation, extending beyond their traditional role in coagulation processes [57]. Within the CN, various neuroglial cells, including oligodendrocytes, astrocytes, and microglia, express receptors for fibrinogen, facilitating their interaction with fibrinogen-laden EVs and subsequently leading to the activation of encephalitogenic CD8+ T cell responses. In the EAE model, mice injected with fibrinogen-enriched EVs demonstrated the development of encephalitogenic T cell responses, while this phenomenon was not observed in the fibrinogen-deficient group. Laboratory confirmation of inflammatory markers such as CXCL10 and CCL2 provided evidence supporting the proposed involvement of the fibrinogen alpha chain in stimulating CD8+ T cells [58]. Furthermore, fibrinogen-laden EVs were observed to directly transfer antigens and pathogenic peptides to antigen-presenting cells (APCs) via major histocompatibility complex (MHC) class I complexes on their surface membrane [59]. In summary, it is hypothesized that fibrinogen-laden EVs may contribute to human leukocyte antigen (HLA) class I disequilibrium and susceptibility to autoimmunity, ultimately playing a role in the development of MS.

Extracellular vesicles with glycolipid membranes

Sulfatide belongs to the sulfoglycolipid group and is associated with myelin formation. Beginning in the endoplasmic reticulum and finishing up in the Golgi apparatus, ceramide is transformed to galactocerebroside which then is sulfated to produce sulfatide. In addition to being a membrane constituent, sulfatide is also associated with protein transportation, intercellular communication, and glial-axon interactions [60]. Recently, a study found that sulfatide mediated immunomodulatory mechanisms in certain conditions such as autoimmunity, infections, and cancers [61]. In MS, a specific sulfatide fraction known as C16:0 exhibited distinct differences compared to those without MS and showed inverse correlation with the expanded disability status scale (EDSS) [62]. These pathways of immune regulation are primarily triggered by the activation of type II natural killer T (NKT) cells which identify sulfatides via CD1d molecules [63]. C16:0 sulfatide exerts an anti-inflammatory role by inhibiting cytokine production. Additionally, the over-functioning of the acid sphingomyelinase enzyme (ASMase), which is usually seen in MS patients, enhances the hydrolysis of other healthy sphingomyelin into ceramide, resulting in a reduction of sphingomyelin and subsequent inflammation [64]. Changes in the quantities, distribution, and metabolism of glycolipids ultimately exacerbate MS.

Extracellular vesicles as cargo of biological substances

Protein, peptides, and enzymes

At least 50 CSF proteins showed significant increases compared to healthy individuals [65]. ASMase, responsible for sphingolipid metabolism, was highly expressed in CSF. ASMase is also associated with ASMase-dependent ceramide signaling, which coordinates intracellular signal cascades and governs the activation and proliferation of CD4+ T-cells. ASMase can also induce neuronal mitochondrial dysfunction and axonal damage [64]. Another protein, Apolipoprotein E4 (ApoE4), was found to be enriched in the CSF of MS individuals [65]. ApoE4 is associated with neurodegeneration due to decreased clearance of pathogenic amyloid-β, as commonly seen in many neurological diseases, especially Alzheimer’s disease [66]. This finding may explain the subsequent neurodegeneration observed in SPMS. Other proteins such as KLKB1, DKK3, C6, and S100A9 were also analyzed and found to be responsible for scar healing, homeostasis, and clot formation, as commonly observed as sequelae of demyelination [65]. However, limited data were gathered due to the very small amount of EV extracted from the CSF. Additionally, CSF is not routinely taken in clinical practice.

MicroRNAs

MicroRNA consists of an 18 to 22-nucleotide single-stranded non-coding molecule. MicroRNA genes are generally transcribed by RNA polymerase II within the nucleus and are subsequently processed in the cytoplasm. The mature microRNA is finally transported into the extracellular space in order to control cell-to-cell communication. Circulating microRNAs have been found to have a correlation with various conditions and diseases, even in the case of MS. Many studies have demonstrated the role of microRNA as an inflammatory trigger in serum, and it also confers a degree of immune protection. A study found suppressed function of anti-inflammatory T cells called Foxhead box P3 (FOXP3+) Treg is observed in MS patients. Let-7i is markedly increased in the serum of MS patients and suppresses the function of insulin-like growth factor 1 receptor (IGF1R) and transforming growth factor beta receptor 1 (TGFβR1). IGF1 and TGFβR1 are important for Treg differentiation and clonal expansion, and impairment of both functions subsequently promotes inflammation. Increments of miR-15b-5p, miR-30b-5p, miR-342-3p, and miR-451a were observed in RRMS patients [67]. MiR-15b and miR-23a were found to be targeted at fibroblast growth factor 2 (FGF-2), which is linked to inflammation, demyelination, and remyelination. MiR-23a was also found to be involved with the level of oligodendrocyte differentiation. A higher level of miR-23a meant there was substantial oligodendrocyte damage. MiR-30b and miR-342 had previously been recognized as potential biomarkers for Alzheimer’s disease and Parkinson’s disease [68-71]. The additional association with MS pathology of both miRNAs apparently supported their contribution to neuro-axonal damage in CNS [72]. In contrast, miR-122-5p, miR-196b-5p, miR-301a-3p, and miR-532-5p, which are considered immune protection EVs, are downregulated in MS patients [72]. MiR-122-5p inhibits inflammatory cell proliferation and migration. MiR-196b-5p is secreted from hematopoietic cells in both normal conditions and malignancy. MiR-301a-3p is a critical endogenous regulator of Th17 cell proliferation and maturation. MiR-532-5p can suppress the inflammatory pathways and has previously been reported as a tumor suppressor in ovarian cancer [72].

Not only serum EVs, but also CSF-derived EVs have been investigated in central demyelinating diseases. Upregulation of miR-18a-5p, miR-21, miR-30a-5p, miR-145-5p, miR-150, miR-328, miR-342-3p, miR-374a-5p, miR-645, and Let-7g-5p, together with the down-regulation of miR-106a, miR-132-5p, miR-146a, miR-191, miR-199a-3p, miR-320a-5p, and miR-365 were associated with oligoclonal band production in CSF and inflammation [73, 74]. Increased miR-181c and miR-633 with a reduction in miR-922, were all associated with inflammatory suppression and may be associated with the reactive response to inflammation [75]. Even though there are several studies on MS and miRNAs in the CSF, they do not directly claim the miRNAs found in these studies were EV-associated. Different extraction techniques, study design, and validation procedures contributed to the variability of results in EV studies [76-79].

Several studies have tried to demonstrate the differences of EVs in serum and CSF to explain the exact pathogenesis of MS [54, 80, 81]. Major myelin proteins, including MPB, proteolipid protein, and MOG, were substantially elevated in the serum and CSF in MS patients. However, peripheral blood mononuclear cell (PBMC)-derived EVs, considered as a potential peripheral EV source, failed to cause expression of MOG or any other myelin proteins. In addition, evidence showed that myelin-associated EVs could transmigrate from the CNS into the periphery through the BBB. This result suggested that MS pathology may originate in the CNS and then spread to the periphery [56]. Another study focusing on miRNAs in MS patients found that serum and CSF miRNAs differed from healthy controls and varied across different disease stages. At least twenty-one miRNAs were significantly expressed, mainly upregulated during an MS relapse, in comparison with the clinical remission stage. However, the level of some miRNAs, such as miR-21-5p, miR-142-3p, miR-223-3p, miR-342-3p, miR-423-5p, and let-7f-5p, were also highly expressed in other CNS inflammatory and infectious diseases [82]. The expression of miRNA was associated with pleocytosis and the existence of oligoclonal bands, which were not only unique to MS but also associated with intrathecally immune-mediated processes such as CNS infection and inflammation. However, due to a lack of specificity, using a single miRNA alteration as a surrogate marker for MS may not be applicable, and a combination of miRNA panels may be suitable for this purpose. All data from clinical studies are shown in Table 1 and 2. Table 3 and Table 4 show the data from in vivo and in vitro studies, respectively.

Extracellular vesicles and pathological correlations in neuromyelitis optica spectrum disorder (NMOSD)

Given the underlying issue of astrocyte dysfunction in NMOSD, individuals with this condition typically show increased expression of astrocyte-associated markers. A recent study has revealed that EVs obtained from CSF, such as GFAP, tetraspanin, haptoglobin-related protein, and C4b binding protein-alpha, were significantly upregulated in NMOSD but not in MS [81]. This discovery suggests that the mechanisms underlying BBB dysfunction differ between MS and NMOSD. Experimental models and human NMOSD lesions have demonstrated a transient disruption of the BBB caused by infiltration of immune cells including macrophages and polymorphonuclear cells (PMNs) with only a few T and B cells, a phenomenon absent in MS [83]. The migration of PMNs into the brain could additionally exacerbate NMOSD lesion formation by allowing additional AQP-4 antibodies and complement to enter the brain. The activation of the complement cascade during astrocyte destruction, in turn, results in the release of the anaphylatoxin C5a, a potent PMN chemoattractant and activator leading to a vicious inflammatory cycle [84].

Alteration of circulating miRNAs may serve as possible biomarkers for NMOSD. A study demonstrated miRNA dysregulation, as shown by an increase in miRNA-122-3p and miR-200a-5p, which occurs in NMOSD, but not in MS [85]. The miR-122-3p targets the MAPK signaling pathway associated with both innate and adaptive autoimmunity [86]. The miR-200a-5p targets the Ras and Wnt signaling pathways, modulating cell proliferation and the inflammatory response. Additionally, the elevation of miR-122-3p is correlated with the increasing EDSS, indicating that miR-122-3p might be a potential biomarker for the prediction of NMOSD relapse. Comparable to other well-defined autoimmune diseases, EV alteration in systemic lupus erythematosus (SLE) is associated with the inflammatory pathway and disease progression. Serum EV miR-21 and miR-155 of SLE patients were upregulated, whereas the expression of miR-146a was down-regulated. Even though the specific mechanism of EVs is unclear, these EVs might activate inflammatory cells via the TLR7 pathway, allowing them to constantly release IFN-α and proinflammatory cytokines, which may contribute to the pathogenesis of SLE. Additionally, one study reported that the alteration of the Ras signaling pathway enhanced inflammatory T cell function and triggered autoimmunity [87]. Hence, the overlap of Ras signaling pathway dysfunction across two diseases suggests that EVs may promote self-intolerance and autoimmunity.

Contrarily, certain microRNAs probably play a protective effect on NMOSD patients. An in vitro study demonstrated that co-culturing miR-129-2-3p with oligodendrocytes and rat optic nerve resulted in a reduction of demyelination [88]. These results affirm the dual role of microRNAs in NMOSD, where they can act both as anti-inflammatory and pro-inflammatory agents.

In the antibody-negative NMOSD group the majority had evidence of myelin oligodendrocyte glycoprotein antibody-associated disorders (MOGAD). However, until now, there has been no published data that demonstrates any association between EVs and MOGAD. Further studies are still required. The roles of EVs in MS and NMOSD were summarized in Fig. 1.

Since the launch of an investigation into the first disease-modifying therapy (DMT) as a therapeutic target for central demyelinating diseases. Recent studies have shown an association between DMTs and the alterations in EV profiles. Focusing on Interferon-β or IFNβ (including IFNβ-1a and IFNβ-1b), IFNβ could lessen the expression of endothelial surface EVs. The increases in platelet and leukocyte EVs confirmed the reduction of transendothelial migration from the periphery into the CNS [42, 89-91]. Fingolimod inhibited S1P receptor activity, causing lymphocyte sequestration in peripheral lymph nodes and the prevention of lymphocyte migration into the CNS. An increase in platelet EVs, leukocyte EVs, and monocyte EVs, was demonstrated in the plasma of fingolimod-treated individuals confirming that fingolimod can hinder immune cells from peripheral transmigration into the CNS [92]. Fingolimod also modified the expression of EVs as shown by a decrease in the expression of endothelium-derived EVs such as CD105, which inhibited the adhesion of inflammatory cells to the BBB [93]. Natalizumab, a humanized monoclonal antibody, modified platelets and lymphocyte EVs, resulting in immune cell suppression and 4-integrin adhesion on activated endothelial surfaces [92]. DMTs which can limit the peripheral inflammatory cells, are preferable treatments since most inflammatory cells originate from outside the CNS rather than within it. Considering that DMTs were responsible for changes in EV levels, the measurement of blood EVs may be useful surrogate-markers for the assessment of therapeutic efficacy in MS. However, more studies are required to add weight to these findings and assess this possibility further. All of these findings are summarized in Table 5.

In contrast, the mainstay treatment of NMOSD is immunosuppressants. In acute exacerbation, intravenous methylprednisolone or plasma exchange is the current major therapy [9]. In the maintenance phase, immunosuppressive drugs including azathioprine, mycophenolate mofetil, methotrexate and rituximab are prescribed to NMOSD patients [9]. However, the data on NMOSD treatment and EVs is scant. Further study in the field is still required.

Currently, Investigators have been focusing on the role of EVs as a drug delivery tool. EVs are naturally produced and seldom induce intolerance, therefore they have been considered as next-generation possible therapeutic strategies. Mesenchymal EVs are one of the possible groups of EVs being explored in therapeutic approaches for MS [94-96]. Several researchers have hypothesized that placental mesenchymal cells could have a protective function in MS, since previous studies showed decreased exacerbating rates of MS in pregnant women, especially in the third trimester, and a rapid emergence in exacerbation rates after delivery [97]. At various points throughout pregnancy, the maternal circulation becomes flushed with a wide range of EVs representing the maternal-fetal interaction. Pregnant women secrete numerous EVs to suppress the immune response caused by the presence of fetal antigens. Approximately 20% of circulating exosomes in the pregnant woman are of placental origin, and their concentration increases during pregnancy, reaching their peak in the third trimester. Placental EVs originated from various kinds of cells. In a normal physiologic state, after the blastocyte adheres to the placenta, fetal trophoblastic cells differentiate into cytotrophoblasts and syncytiotrophoblasts. EVs generated by cytotrophoblasts and syncytiotrophoblasts enhance embryonic and placental development. Martire et al. demonstrated an association between placental EVs and MS patients by analyzing EVs secreted from placental tissue culture. Major EV groups, specifically embryonic, mesenchymal, and hematopoietic stem cells, showed high expressive capacity. EVs from these origins have been reported to inhibit the inflammatory cascade and enhance immune tolerance. Contrarily, the EVs derived from antigen-presenting cells (such as HLA-A/B/C molecules, CD86, and CD1c), T cells (CD2, CD3, CD25), activated immune cells with cytotoxic properties (CD56), and B cells (CD19, CD20) were significantly decreased in pregnant women [98]. Tumor necrosis factor (TNF)-related EVs, specifically Fas ligand (FasL) and TNF-related apoptosis-inducing ligand (TRAIL), suppressed inflammation by triggering apoptosis in peripheral mononuclear cells. Trophoblast-related EVs are also related to Treg proliferation and NF-κB activation inhibition [99]. Steroid hormone concentrations, which include estrogen and progesterone, surge in pregnant women, particularly in the third trimester. Progesterone at high concentrations causes lymphocytes to release progesterone-induced binding factor (PIBF). The findings demonstrated that PIBF continuously rises during pregnancy and declines dramatically after delivery. PIBF is associated with the secretion of anti-inflammatory cytokines from Th2 cells, including IL-4, IL-5, and IL-10 [98, 100]. In addition, EAE mice injected with placental EVs showed a decrease in T cell proliferation, increase Treg cell, and a subsequent decrease in inflammation adding weight to these findings [101, 102]. The same effects were observed in EVs produced from bone marrow, periodontal ligament cells, adipose tissue, and umbilical cord mesenchyme, which raised the levels of protective cytokines, such as IL10, transforming growth factor-beta (TGF-β), and stromal cell-derived factor 1 alpha (SDF-1α), but reduced pro-inflammatory cytokines including tumor necrosis factor-alpha (TNF-α) and NALP-3 [94, 96, 103-109].

The effectiveness of mesenchymal EVs has not only been observed in human-derived mesenchymal EVs, but also in EAE mice treated with Rhesus monkey mesenchymal stromal cell EVs. This treatment showed evidence of stimulation of oligodendrocytes and prevention of the polarization of M2 to M1 microglia, resulting in remyelination [95]. EVs from doxycycline-inducible let-7g transgenic (Let-7Tg) mice also resulted in measurable benefits with regard to inflammatory suppression. Let-7 inhibits CD4+ T cell proliferation, which may contribute to the lower number of CNS-infiltrated Let-7Tg CD4+ T cells. Th17 differentiation and IL1 receptor 1 (IL1R1) activity were suppressed by the Let-7g transgene, resulting in a pro-regenerative environment [110].

Since the goal of improving MS demyelination is to promote remyelination, many EVs were examined to investigate their efficacy with regard to possible effects [111-113]. Intranasally administered miR-219a-5p extracts boosted OPC proliferation in EAE mice resulting in remyelination [114]. IFN-stimulated dendritic cell EV aided remyelination, as indicated by an increase in MBP. One study found that the exposure of aging animals to EV-containing serum from youth animals also increases oligodendrocyte proliferation and promotes remyelination [115]. The experimental model discovered that environmentally enriched EVs primarily containing miR-219 can promote remyelination, but miR-219 deficient EVs cause remyelination failure and inhibit oligodendrocyte progenitor cell proliferation [116, 117]. An in vitro study showed that IL4 promotes remyelination by causing microglia to transform into the pro-regenerative form [53]. All of these findings suggest that personalized medicine is possible through the development of EVs as carrier for numerous therapeutic compounds for MS. All data is summarized in Table 6. Potential interventions are also illustrated in Fig. 2.

EVs have been shown to be linked to the pathogenesis of immune-mediated central demyelinating disorders, however the actions and presence of individual EVs have been found to vary between different diseases. Platelet and endothelial EVs, as well as activated microglia-derived exosomes, were found to be associated with synaptic and BBB dysfunctions in MS. In vivo data provided evidence of an association between fibrinogen and astrocyte-derived exosomes with neuroinflammation and demyelination. Although robust evidence from serum and CSF of MS individuals highlighted the role of various EVs, including those from leukocytes, microglia, and microRNAs, concerning central inflammatory pathways, there was still no definitive evidence regarding the implementation of EVs as a surrogate biomarker in diagnosis, prognosis management or therapeutic approaches. NMOSDs are a new spectrum of demyelinating diseases that to date lack data concerning the role of EVs. There is limited clinical evidence to indicate the difference in the action of exosomes between NMOSD and MS, however some miRNA dysregulation has been shown to be related to clinical relapses of those diseases. Even though MS and NMOSD are considered CDD and share overlapping clinical features the data pertinent to EVs and CDD mostly focus on MS, whereas the data on NMOSD remain limited. Increased research and data into the clinical use of EVs as a potential therapeutic strategy is still required. However, as they have been considered carriers for diverse biological molecules, there is a very real possibility that EVs could be exploited as a next-generation targeted therapy as a precision medicine approach for MS and NMOSD.

This work was supported by the Distinguished Research Professor grant from the National Research Council of Thailand (N42A660301 to SCC); Research Chair Grant from the National Research Council of Thailand (NC) and a Chiang Mai University Center of Excellence Award (NC).

The authors have no competing interests to declare that are relevant to the content of this article.

Extracellular vesicle marker in central demyelinating diseases related to inflammatory pathways from plasma or serum: reports from clinical studies

Extracellular vesicle markers in central demyelinating diseases related to inflammatory pathways from csf: reports from clinical studies

Targeting extracellular vesicle markers in central demyelinating diseases related to inflammatory pathways: data from in vivo studies

Targeting extracellular vesicle markers in central demyelinating diseases related to inflammatory processes: data from in vitro studies

Extracellular vesicles markers in central demyelinating diseases and therapeutic strategies: reports from clinical studies

Extracellular vesicles markers in central demyelinating diseases and therapeutic strategy: reports from in vivo and in vitro studies