Glia Connect Inflammation and Neurodegeneration in Multiple Sclerosis

Abstract

Multiple sclerosis (MS) is regarded as a chronic inflammatory disease that leads to demyelination and eventually to neurodegeneration. Activation of innate immune cells and other inflammatory cells in the brain and spinal cord of people with MS has been well described. However, with the innovation of technology in glial cell research, we have a deep understanding of the mechanisms of glial cells connecting inflammation and neurodegeneration in MS. In this review, we focus on the role of glial cells, including microglia, astrocytes, and oligodendrocytes, in the pathogenesis of MS. We mainly focus on the connection between glial cells and immune cells in the process of axonal damage and demyelinating neuron loss.

Introduction

Multiple sclerosis (MS) is a chronic, inflammatory, demyelinating disease of the central nervous system (CNS) with progressive neuroaxonal degeneration [1]. The diagnosis and monitoring of MS have benefited from the advances in magnetic resonance imaging (MRI) and positron emission computed tomography (PET) methods that enable the visualization of features related to chronic inflammation in MS [2].

Chronic inflammation in MS arises from the activation of innate and adaptive immune responses in the CNS. The immune system and particular immune pathways play a central role in the pathogenesis of MS. In the context of MS neuroinflammation, the cerebral cortex, cerebrospinal fluid (CSF), the meninges, and choroid plexus are filled with innate and adaptive immune cells. With the development of single-cell sequencing [3], mass cytometry, and fate-mapping techniques, researchers have discovered various immune cell populations in the CNS, largely residing in the interfaces between the brain and the periphery [4].

The blood-brain barrier (BBB) is a diffusion barrier, which impedes the influx of most compounds from blood to the brain. Three cellular elements of the brain microvasculature compose the BBB—endothelial cells, astrocyte end-feet, and pericytes (PCs) [5]. Endothelial cells at the BBB are tightly packed and have low pinocytotic activity, enabling them to block inflammatory cells and molecules in the peripheral blood from entering the CNS [6].

During the development of MS, the integrity of the BBB is chronically impaired [7], peripheral immune cells are activated and then cross to the CNS via the damaged BBB, ultimately leading to immune-mediated tissue injury. Finally, fibrinogen appears perivascularly at the edges of MS lesions and is histopathologically associated with developing demyelination and inflammation, suggesting that blood proteins leak into the CNS and maintain chronic inflammation [8].

Furthermore, at sites of CNS injury and disease, the recruitment of peripheral immune cells modulates tissue-resident glial cells. Peripheral immune cells crosstalk with glial cells by secreting diverse cytokines. They influence each other bidirectionally. On the other hand, neuroinflammation including peripheral immune cells and glia affects synaptic transmission at different levels in MS. Glia forms a bridge connecting inflammation and neurodegeneration.

Immunopathogenesis in MS

Adaptive Immunity

The role of adaptive immune responses in the pathogenesis of MS has been extensively studied. Peripheral immune cells are first activated outside the brain; T cells are activated by molecular mimicry between foreign peptides and self-peptides presented by dendritic cells (DCs), and then cross to the CNS via the damaged BBB, ultimately leading to immune-mediated tissue injury. This is the most accepted hypothesis for the development of MS. However, a study reported that the pathogenesis of some cases of MS begins within the CNS - oligodendroglial and axonal injury are independent of peripheral immune cell infiltration - then continues with peripheral immune activation [9, 10].

In the neuroinflammation context, the main way in which T cells enter the CNS involves direct interactions between the T cells and the endothelium of post-capillary blood vessels. The central lymphatic system associated with the dural venous sinuses allows DCs and T cells to migrate into the deep cervical lymph nodes for antigen presentation. In the same way, B cells shuttle between cervical lymph nodes and the CNS parenchyma as a major functional consequence of chronic inflammation in the CNS [11, 12]. Reflux in the paravascular lymphatic system facilitates the transport of brain antigen proteins into the blood [13, 14].

B cells that can differentiate into plasmablasts and plasma cells are also found in the brain parenchyma of MS patients [15]. B cells and plasma cells found in the brain parenchyma, meninges, and CSF of MS patients produce anti-myelin antibodies that are detectable by the presence of oligoclonal bands [16, 17].

Recent studies have confirmed that the complement system is activated in MS. Functional immunoglobulins secreted by B cells and plasma cells lead to myelin destruction by activating the classical complement system [18]. Recent anti-CD20 antibody treatment has revealed the importance of B cells in the immunopathogenesis of MS [19]. The antibody-independent function of B cells is to attract and activate T cells and myeloid cells for recruitment to the CNS, thereby enhancing cellular immune responses.

Innate Immunity

The innate immune system in the CNS contains innate immune cells and soluble inflammatory mediators. These innate immune cells mainly include macrophages, microglia, and astrocytes. All these cells are involved in MS immunopathogenesis. The innate immune response begins with pattern recognition receptors on myeloid cells recognizing potential external threats such as pathogen-associated molecular patterns (PAMPs) or internal threats such as damage-associated molecular patterns [9].

Traditional clinical MRI imaging technology has low sensitivity for the detection of this chronic inflammation. Over the past 20 years, clinical quantitative dynamic MRI scanning with contrast agents has increased the sensitivity to inflammatory signatures, including a reduction in BBB integrity associated with inflammation. Quantitative lesion volume measures and magnetic resonance susceptibility imaging are sensitive to the activity of macrophages in the rims of white matter lesions. PET and magnetic resonance spectroscopy methods can also be used to detect activated innate immune systems in the brain and spinal cord. These high-level imaging methods for the visualization of chronic inflammation make a great contribution to clinical applications [2].

The Correlation Between Inflammation and Neurodegeneration in MS

MS is a chronic inflammatory disease of the CNS, characterized by axon demyelination and neuroaxonal degeneration which becomes more dominant as the disease progresses [1]. Patients are gradually deprived of their memory, motor abilities, and ability to handle everyday tasks [20]. In most MS patients, MS begins with a relapsing-remitting course (RRMS) process, and after 15-25 years, RRMS transforms into progressive neurodegeneration, called secondary progressive MS (SPMS) [21]. At the same time, 10–15% of MS patients directly enter the neurodegenerative stage at the onset of the disease, which is called primary progressive MS (PPMS) [1, 22]. In RRMS and the progressive forms of MS, chronic central inflammation plays a major role in causing neurodegenerative processes [23, 24]. The pathogenesis of RRMS is the transient infiltration into the CNS by peripherally adaptive or innate immune cells, whereas the mechanism of progressive MS is mainly driven by immune cells residing in the CNS or its rim.

The first lesions in the brain of MS patients are usually focal areas of demyelination in the white matter, caused by the invasion of peripheral immune cells through the BBB. Further sustained activation of innate immune cells such as bone marrow-derived macrophages and infiltration of peripheral immune cells, including B cells and CD8⁺ T cells, eventually lead to myelin degeneration, oligodendrocyte loss, and axonal damage [1, 20, 25–27].

In the healthy CNS, lipid-rich myelin sheaths encase axons and aid in the conduction of electrical impulses. CNS myelin has two functions: it provides metabolic support to the axon and allows rapid transmission of action potentials along the axon [28–30]. In the healthy CNS, oligodendrocytes are mature, terminally differentiated oligodendrocyte progenitor cells that form myelin sheaths around axons. Myelin sheath formation from mature myelin oligodendrocytes is divided into 3 steps. First, the mature myelin oligodendrocytes express myelin proteins such as myelin basic protein (MBP). Then, myelin proteins build up an elaboration of the myelin membrane to initially wrap the axon. Finally, the continued elaboration of the membrane further wraps the axon to form the multilayered and compacted sheath [31].

In MS, neurodegeneration is dependent on inflammation, which is highly correlated with axonal damage [32]. The autoimmune attack (including the pathological activation of T cells/B cells and macrophages) drives inflammatory demyelination, the myelin sheath is lost, but the underlying axon remains intact [31]. Existing oligodendrocytes whose sheaths have been damaged do not contribute to the regenerative process and are cleared away by activated microglia and astrocytes [33]. The remyelination response generates new sheaths from newly formed oligodendrocytes. When remyelination is hindered, energy-efficient conduction cannot be restored due to the absence of remyelination; this leads to the accumulation of mitochondria at the node and ultimately axonal degeneration [34]. This degeneration triggers a secondary inflammatory response, characterized by the presence of activated macrophages around the degenerating axon.

In MS, the activity of lesions reflects the degree of degeneration and the disease process. Based on the number of inflammatory cells in the autopsy brains of patients with different MS pathogenesis, it has been found that inflammation in the brain not only exists in RRMS but also in progressive MS; peripheral plasma cell infiltration is the most evident in MS. Meningeal B cells form ectopic lymphoid-like aggregates and play a major role in inflammatory demyelination and neurodegeneration of the cerebral cortex [35]. A marker of meningeal B cell aggregation is CXCL13 in serum and CSF [36], which is elevated in patients with PPMS or SPMS [37]. The presence of ectopic leptomeningeal lymphoid tissue in the brains of SPMS patients is associated with extensive cortical demyelination and neurodegeneration [38]. In addition to the complete demyelination of the affected area, the loss of neurons and synapses in adjacent areas can also be substantial [39].

Neuropathologically, white matter lesions in MS are classified according to the characteristics and extent of microglia-macrophage-mediated inflammation and demyelination [26]. Lesions with persistent demyelination or the presence of microglia and macrophages throughout the volume are called active lesions. Lesions with central demyelination or thin myelin sheaths associated with remyelination, and with microglia-macrophage lesions at their borders are classified as mixed active–inactive lesions, and these lesions become dominant as MS progresses [26, 40, 41]. During the pathogenesis of MS, chronic inflammation persists in white matter lesions. Although patients with progressive MS have fewer active lesions in the white matter region than RRMS, mixed active–inactive is seen in the postmortem brains of almost all people with MS [42]. Mixed active–inactive plaques are most abundant in the brains of people whose disease duration is >10 years and who are >50 years of age at death [43].

Inflammation of the CNS can lead to a series of molecular changes in immune cells. In both active lesions and mixed active-inactive lesions, bone marrow-derived macrophages change from a state in which they mainly express the non-inflammatory molecule CD206 [44] to a pro-inflammatory phenotype, with the expression of markers such as CD68, CYBA, major histocompatibility complex class II antigens, inducible nitric oxide synthase, and ferritin [32, 45]. Also, disease-associated microglia reduce the expression of the homeostatic markers and up-regulate the expression of pro-inflammatory markers [46]. During active demyelinating lesions, the myelin breakdown product iron accumulates in macrophages [47].

In addition, several growth factors and immune signaling pathways establish a link between inflammation and neurodegeneration in MS, such as the growth factor TGF-β, the complement cascade, and the extracellular receptor TREM2 [48]. Immune cells secrete neurotoxic products, including reactive oxygen species (ROS), glutamate, cytokines, and chemokines. They further elicit an immune response, altering the cellular metabolism of neurons and their axons. In the short term, these immune cell products are essential for tissue defense and stress responses, but in the long term, they cause intrinsic stress in the CNS, disrupt homeostasis, and ultimately lead to neurodegenerative diseases [49].

Besides the detrimental aspect of neuroinflammation in MS, glial cells play a positive role in remyelination after demyelination [50]. In the area of mixed active-inactive lesions, activated microglia and astrocytes mainly exhibit an anti-inflammatory phenotype and promote remyelination [51]. Activated glial cells drive remyelination by facilitating OPC recruitment and differentiation into oligodendrocytes in three ways. First, they clear the damaged myelin debris via CX3CR1, triggering receptor expressed on myeloid cells 2 (TREM2) [52] and colony-stimulating factor 1 receptor (CSF1R) signaling [53], and they clear cholesterol via apolipoprotein E, LXRα, ABCA1, and ABCG1 [54]. Second, they secrete pro-regenerative factors, such as activin A [55], insulin-like growth factor 1 (IGF1) [56], galectin 3 [57], tumor necrosis factor (TNF) [58], and IL-1β [59] to recruit OPCs to demyelination zones and support oligodendrocyte lineage cell responses. Third, they facilitate OPC responses to the extracellular matrix during remyelination via the secretion of matrix metalloproteinases (MMPs) and TGM2 [60].

CNS Glial Cells Connect Inflammation and Neurodegeneration in MS

Glial Cells at Sites of MS Pathology

Microglia and astrocytes are widely distributed in normal cerebral white matter, and microglia in non-lesion white matter areas of MS patients mainly exert the function of removing damaged neurons and cell debris. Activation of microglia in the white matter of the brain becomes more representative of MS progress [41]. The number and features of microglia in MS lesions can represent the progression of MS disease. Active lesions are characterized by macrophages and microglia residing throughout the area of the lesions, while mixed active–inactive lesions have a hypocellular lesion center, and a small number of macrophages and microglia are limited to the edge of the lesion. In inactive lesions, macrophages and microglia display an almost non-inflammatory phenotype [44]. Astrogliosis is rare in normal white matter and is evident in areas surrounding MS lesions. Microglial activation and astrogliosis in normal white matter can predict the onset of MS lesions [61].

In the acute lesions of SPMS, iron accumulation is evident in activated microglia and macrophages [62]. Intracellular iron accumulation occurs after activated microglia and macrophages engulf myelin debris, dead oligodendrocytes, and leaking red blood cells [63]. The use of high-field MRI has demonstrated that iron rim lesions appear in MS foci with iron accumulation in microglia-enriched areas [64]. These iron rim lesions are more destructive than non-iron rim lesions [65], and iron rim lesions are almost absent around the remyelinated shadow plaques [66], indicating that iron accumulation in glia has negative effects on the recovery of MS. In progressive MS, abnormal iron accumulation in microglia promotes the establishment of a proinflammatory phenotype in microglia [67]. For example, exhaustive transcriptome analysis of microglia in human tissues found that iron accumulation in microglia drives glycolysis in cells, making them more efficient in producing ATP, promoting morphological changes, and secreting cytokines, thereby a pro-inflammatory cell phenotype is established [68].

Microglia Connect Inflammation and Neurodegeneration in MS

Microglia are ubiquitous resident phagocytes among the innate immune cells in the CNS, where they can self-replenish. Unlike macrophages, microglia originate from the yolk sac [69]. In neural homeostasis, microglia have important functions such as pruning synapses, helping remyelination, removing debris, producing cytokines, tissue surveillance, and supporting other brain cells [70–72].

Microglia are the first responders to neuroinflammation or neural injury, and they rapidly adjust their phenotype in response to the brain environment [73]. Microglia are not just quiescent sentinels ready for neuroinflammation, they are also highly reactive to brain dynamics, actively producing signaling molecules to maintain homeostasis or trigger diseases. Microglia play an important role in all MS lesions and during advanced neurodegeneration. Activated microglia are found in lesions and white matter areas in MS patients. Microglia with a pro-inflammatory phenotype are usually scattered in small clusters in white matter areas. These disease-associated microglia reduce the expression of the homeostatic marker molecule P2ry12 and up-regulate the expression of the pro-inflammatory markers TMEM119 and CD68, antigen presentation MHCII antigen, and ROS markers [46]. The microglia with a pro-inflammatory phenotype also produce ROS, nitrogen [74], pro-inflammatory cytokines [75], and crosstalk with adaptive immune cells [76], activating the complement system and affecting neurotransmitter transmission [77], ultimately causing axonal degeneration [78], oligodendrocyte and myelin damage, and contributing to MS progression (Figs 1 and 2) [46].

Fig. 1.

Pathophysiological roles of glial cells and peripheral non-neuronal cells in MS. In MS, peripheral non-neuronal cells, including T cells, B cells, and cells of the monocyte-macrophage lineage, are recruited to the CNS in response to neuronal damage. Crosstalk between different cell types by several immune cytokines and chemokines exacerbates this neurotoxicity. Peripheral immune cells promote the activation of resident glia. Glial cells are thought to acquire neurotoxic properties, comprising both gain-of-function toxicity and loss of supportive functions. Glial cells in turn regulate peripheral non-neuronal cells, followed by synaptic loss and demyelination. Chemotactic factors, such as C–C motif chemokine 2 (CCL2). NO, nitric oxide; ROS, reactive oxygen species; TGFβ1, transforming growth factor β1; TNF, tumor necrosis factor.

Fig. 2.

Inflammatory synaptopathy in MS and EAE. Proinflammatory cytokines (including TNF and IL-1β) released by activated lymphocytes, microglia, and astrocyte induce glutamate accumulation and GABA reduction in the synaptic cleft in MS and EAE. Furthermore, inflammation can induce structural alterations, comprising synaptic loss characterized by degeneration of the presynaptic and/or postsynaptic site.

Microglia have multiple cellular phenotypes and different transcriptional phenotypes depending on the stage of MS (relapsing/remitting or progressive) and lesion type (active or chronic). In the late stage of MS, microglia swing between homeostatic and pro-inflammatory phenotypes and exert beneficial effects including phagocytosis, brain-derived neurotrophic factor secretion, and remyelination support. Microglia contribute to remyelination by secreting IGF-1 and fibroblast growth factor and inducing the proliferation of OPCs.

Traditional studies have identified different states of microglia according to their morphological phenotypes. The classification of quiescent and reactive microglia is usually distinguished by the number of cell branches. However, the cell morphology of microglia is highly dynamic; further studies of gene transcripts to distinguish different states of microglia through new technologies have the potential to allow a better description of dynamic microglial states [79].

Astrocytes Connect Inflammation and Neurodegeneration in MS

Astrocytes account for almost 50% of all glial cells in the CNS, with many stellate protrusions per cell. This special cellular structure helps astrocytes to make close contact with synapses, blood vessels, and other glial cells. Astrocytes are central resident cells necessary for synapse formation and function. They directly affect the function of local synapses through rapid neurotransmitter uptake and extracellular potassium balance (Fig. 2). Astrocytes also support neurons through sophisticated regulation of metabolism, neuronal excitability, and plasticity. Astrocytes are also neurovascular units, an important component of the BBB, maintaining cerebral blood flow. During CNS injury or disease, the BBB, including astrocytes, restricts the entry of peripheral immune cells into the CNS parenchyma. In the context of neuroinflammation, crosstalk between astrocytes and microglia promotes the occurrence of inflammation.

In the nineteenth century, Jean-Marie Charcot identified astrocytes as key components of MS lesions; this was the first evidence of the importance of astrocytes in MS pathology. The subsequent experimental autoimmune encephalomyelitis (EAE) mouse model further elucidated the role of astrocytes in neuroinflammation. In the acute EAE B6 mouse model, depletion of reactive astrocytes increases the disease severity and CNS inflammation [80]; while in the EAE non-obese diabetic mouse model, depletion of reactive astrocytes ameliorates the pathogenesis of disease and decreases the recruitment of active microglia [81], suggesting the complex and diverse functions of astrocytes. We found that astrocytes with different phenotypes are enriched in MS, such as a subset of neurotoxic reactive astrocytes abundant in MS and characterized by complement component 3 expression. In addition, a subset of astrocytes driven by transcription factor (MAFG) and GM-CSF signaling promotes inflammation and neurodegeneration in EAE and MS. There are also some anti-inflammatory phenotypes of astrocytes, such as a subset of anti-inflammatory LAMP1⁺TRAIL⁺ astrocytes, which limit EAE pathogenesis by inducing T cell apoptosis [82].

CNS inflammation in MS is typically triggered by BBB dysfunction and peripheral immune cell invasion, followed by glial including astrocyte activation, which respond to pro-inflammatory cytokines and PAMPs. In MS patients, astrocytes in active lesion areas are hypertrophic and strongly express GFAP, secreting pro-inflammatory cytokines, chemokines, and demyelination signaling molecules [83]. Among these, the chemokines CCL2, CXCL10, and CXCL12 produced by reactive astrocytes attract leukocytes including T cells, B cells, and plasma cells to the perivascular space and CNS parenchyma, further activate the complement system, and cause axonal damage and demyelination. Conversely, astrocyte responses are regulated by T cells, and IL-10 produced by regulatory T cells limits astrocyte activation and reduces neuroinflammation [84]. GM-CSF signaling also limits astrocyte anti-inflammatory functions by downregulating TRAIL expression [85] (Fig. 1). In MS lesions, astrocytes can also phagocytose myelin debris, followed by intracellular iron accumulation, but their phagocytic capacity is not as good as that of microglia and macrophages [86, 87]. In the chronic inflammation of progressive MS, iron accumulation disrupts astrocyte antioxidant pathways, resulting in its damage [88].

In the healthy CNS, astrocytes contribute to the maintenance of neuronal function through a variety of mechanisms involving ion channels, neurotransmitters, and metabolites [89]. Astrocytes trap glutamate and prevent it from accumulating in the synaptic cleft, thereby protecting the synapse from glutamate toxicity. Furthermore, astrocytes act as modulators of synaptic inhibition. In MS, astrocytes influence the glutamine neurotransmission across the synaptic cleft by increasing the release and/or decreasing the reuptake of glutamate into astrocytes, causing excessive stimulation of glutamate receptors (GluRs) and, consequently, excitotoxic damage of neurons and oligodendrocytes [90].

Crosstalk between CNS glia is a regulatory mechanism in MS pathogenesis. Recently described microglia–astrocyte interactions mediated by the semaphorin4D–plexinB1/2 and ephrinB3–EPHB3 axis, promote CNS inflammation [82]. Neurotoxic astrocytes similar to those induced by microglia in the EAE mouse model have been observed in the postmortem brain tissue of MS patients [91]. In EAE mice, microglia-derived TGFα reduces disease severity by directly regulating astrocytes, whereas microglia-derived VEGFβ aggravates disease by regulating astrocytes [92].

Oligodendrocytes Connect Inflammation and Neurodegeneration in MS

Myelinating oligodendrocytes are derived from OPCs in the early stages of neural development. During embryogenesis, neural stem cells generate highly proliferative OPCs. The differentiation of OPCs into mature and myelinating cells is a multistep process [93], tightly controlled by spatiotemporal activation and repression of specific growth and transcription factors [94, 95]. Mature oligodendrocytes spread throughout the gray and white matter of the CNS, wrap around neuronal axons, and are responsible for the production of myelin, maximizing axonal conduction velocity [96]. Myelin is an extension of the oligodendrocyte cell membrane with a highly complex structure.

An undifferentiated pool of OPCs is maintained in the adult CNS and contributes to the production of myelinating oligodendrocytes throughout adulthood. Following nerve injury or inflammation, axonal damage requires newly generated and differentiated mature oligodendrocytes to accelerate recovery.

Components of the myelin sheath [MBP-, myelin oligodendrocyte glycoprotein, or proteolipid protein] are the main targets of autoimmune attack, mainly including the antigen-specific CD4⁺ T cells in the CNS [97].

Also, inflammation plays an important role in OPC/oligodendrocyte survival, proliferation, differentiation, and remyelination. The blockade of OPC differentiation contributes to failed remyelination in MS owing to the inflammatory microenvironment and the presence of an array of inhibitory molecules like Lingo-1 and PSA-NCAM [98]. The inflammatory microenvironment is associated with the presence of ROS, which may in turn affect the fate of OPCs [99]. The inflammatory microenvironment is also associated with the lack of certain factors such as IGF1, TGF-β1, GGF2, or integrins, contributing to limited remyelination [58, 59, 100]. Moreover, in experimental inflammatory demyelination, OPC differentiation is inhibited by effector T cells and IFNγ. This study found that induction of immunoproteasomes in OPCs occurs in human demyelinated MS brain lesions [101]. The crucial immunoproteasome subunit PSMB8 (also known as LMP7) increases MHC class I expression on OPCs, rendering OPCs a more prominent target for the cytotoxic CD8⁺ T cells that are abundant in MS lesions [102].

In addition, recent evidence suggests that oligodendrocytes also have immune-related functions. Falcão et al. reported that disease-specific oligodendroglia are also present in human MS brains. OPCs express MHC-II in response to IFNγ, exhibit phagocytic capacity, and regulate T cell survival and proliferation, activating memory and effector CD4-positive T cells. They claimed that oligodendrocytes and OPCs are not only passive targets but instead active immunomodulators in MS [103]. Niu et al. reported that interaction between oligodendroglia and the vasculature in MS disrupts the BBB, triggering CNS inflammation [104]. Also, OLCs regulate neuronal network activity through direct connections to the glial syncytium [105].

Crosstalk Between Peripheral Immune Cells and Glia

There is little peripheral lymphocyte infiltration in the normal brain, but lymphocytes are recruited to the CNS as part of a neuroinflammatory response. The migration of B cells into the CNS is regulated by T_FH cells, based on studies in EAE and evidence in the CSF of patients with MS [106]. B cells with a neurotoxic phenotype have been observed in patients with MS and shown to secrete soluble factors that directly kill oligodendrocytes and neurons in vitro [97, 107].

The progression of MS is mainly the result of interactions between peripheral immune cells and glial cells in the CNS, especially microglia. At sites of CNS injury and disease, the recruitment of peripheral immune cells modulates tissue-resident glial cells [108]. IFNγ and IL-4 are known T cell-derived cytokines driving the pro-inflammatory and anti-inflammatory phenotypes of microglia in culture, respectively [109], suggesting T cells directly modulate resident microglia during injury and disease. Microglia can also present antigens to lymphocytes by expressing MHC class II molecules; their role as antigen-presenting cells is seemingly dispensable for the induction of EAE [110].

T cells, similar to microglia, produce multiple molecules, eventually altering the activation state of astrocytes. Activated astrocytes in turn recruit peripheral immune cells into the injured BBB by secreting CCR2, activating inflammasomes, and further causing axonal damage. In addition, activated astrocytes produce fewer neurotrophic factors, thereby failing to further support neuronal integrity, eventually accumulating neurodegeneration [111].

After CNS injury, microglia transform into an inflammatory state and further recruit mononuclear macrophages into the CNS; the macrophages in turn regulate microglial function and state by promoting microglia to express a range of inflammatory mediators in vitro [112]. Peripheral monocyte-derived macrophages stimulated by interferon-γ facilitate microglia to induce the expression of TNF and inducible nitric oxide synthase. Microglia and macrophages interact to enhance each other's phagocytic capability, which is detrimental to neuronal survival, eventually causing CNS damage and disease [113]. Monocyte-macrophages can regulate not only the activity and function of microglia but also the function of astrocytes. Soluble factors secreted by inflammatory macrophages enhance the release of CCL2 and MMP9 by astrocytes, which further recruit peripheral mononuclear macrophages into the CNS. The crosstalk between astrocytes and macrophages further accelerates CNS inflammation and damage [114].

Neutrophil clearance is one of the key events in the resolution of neuroinflammation [115]. The clearance of neutrophils is mediated by specific factors derived from macrophages and glia, which ultimately resolve inflammation and promote recovery after CNS injury [116]. Activated microglia and peripheral infiltrating neutrophils are in contact with each other during CNS inflammation, as shown by microscopy in vivo. Microglia accelerate the clearance of neutrophils by secreting specific inflammatory factors. The clearance of neutrophils by macrophages and microglia in turn affects the function and phenotype of phagocytes. Also, complement components released by neutrophils promote the differentiation, maturation, and migration of astrocytes from neural stem cells in vitro [117]. In addition, neutrophils produce a series of cytokines and proteolytic enzymes affecting microglia, astrocytes, and oligodendrocytes [118].

The inflammasome, for example, as a mediator connecting glial and immune cells, is an intracellular multiprotein complex that mediates innate immune responses against PAMPs. NLRP3, mainly localized to microglia, was the first inflammasome to be studied in the brain. Activation of the NLRP3 inflammasome has also been found in other CNS cells, such as astrocytes, neurons, endothelial cells, and pericytes. Also, immune cells (T cells and myeloid cells) express components of the NLRP3 inflammasome complex in MS. One of the hallmarks of NLRP3 inflammasome activation is the production and secretion of the proinflammatory cytokines IL-1β and IL-18 [119]. Activation of the NLRP3 inflammasome in glial cells amplifies neuroinflammation and promotes EAE progression. Similar to the EAE mouse model, the IL-1β mRNA level is increased in the peripheral blood mononuclear cells of MS patients. The mRNA and protein levels of NLRP3 inflammasome components and pro-inflammatory cytokines are upregulated in MS patients. IFN therapy effectively inhibits the level of IL-1β in serum and further improves the clinical manifestations of MS by reducing the expression of inflammasome and the secretion of IL-1β [120].

Glia Connect Inflammation and Neurotransmitters

Brain-infiltrating autoreactive T cells and glial cells in the CNS are significantly activated during MS, and these activated immune cells broadly modulate synaptic transmission. Infiltrating T cells and glial cells are important sensors of the local microenvironment of the CNS and secrete many cytokines, growth factors, or neurotransmitters. Activated microglia affect synaptic structure and function by shedding microvesicles [121] or by stripping synapses after inflammatory injury [122]. Therefore, these glial cells connect inflammation and neurotransmitters (Fig. 2).

Recent studies have found that incubation of CD3⁺ T cells isolated from the spleen of EAE mice with brain slices from normal mice alters normal brain glutamatergic and GABAergic synaptic transmission [123, 124], indicating that immune cells can change the conduction of neurotransmitters and affect the function of synapses. Similarly, co-incubation of activated microglia from EAE mouse brains with brain slices from normal mice results in synaptic defects similar to those in EAE mice [123]. Interestingly, the effects of CD3⁺ T cells and activated microglia on synaptic function are completely reversed by specific cytokine antagonists, suggesting that cytokines secreted by immune cells contribute to synaptic defects [123].

Pro-inflammatory and anti-inflammatory cytokines secreted by glial cells play an important role in regulating bidirectional communication between glial and neurons, as well as in regulating synaptic transmission [125]. Therefore, in the pathological state of MS, synaptic transmission and neuronal survival depend on intrasynaptic cytokine concentrations, the balance between pro-inflammatory and anti-inflammatory cytokines, and the expression of neuronal surface receptors [126].

Neuroinflammation affects synaptic transmission at different levels. On the one hand, inflammation alters the frequency of presynaptic neurotransmitter release, and the amplitude and duration of the postsynaptic current [124]. On the other hand, neuroinflammation affects spontaneous excitatory and inhibitory postsynaptic currents (sEPSCs and sIPSCs) of neurons, ultimately altering synaptic excitability and disrupting neuronal function [125]. In MS, the effects of neuroinflammation on synaptic function are commonly studied using two approaches: (1) correlation of cortical excitability and plasticity—as measured by transcranial magnetic stimulation—with levels of cytokines in the CSF [127]; and (2) isolation of CSF from MS patients and co-incubation with healthy mouse brain slices to explore the effects of specific cytokines in the CSF of MS patients on synaptic transmission [128].

Dysregulation of Neurotransmitters in MS

The concentrations of glutamate, GABA, and other neurotransmitters at the synaptic cleft depend on the regulation of their synthesis, release, degradation, and reuptake. In MS, these regulatory processes are disturbed, and the receptors involved in neurotransmission are abnormally expressed, ultimately leading to disturbances in neurotransmitter transmission [90].

The Glutamatergic System in MS

There is increasing evidence that free glutamate in the synaptic cleft is increased in MS, and abnormally high levels of glutamate have been detected in the CSF, white matter, and gray matter of MS patients [129]. Increased glutamate concentration is involved in many processes depending on the interaction between immune cells and nervous systems. Neurons, T cells, macrophages, astroglia, and activated microglia are all able to synthesize and release glutamate [130], guaranteeing a continuous local supply of glutamate neurotransmitters. The abnormal accumulation of glutamate in the synaptic cleft is due to increased glutamate release in MS. At the same time, in oligodendrocytes from MS patients, the expression of the metabolizing enzyme glutamate dehydrogenase is down-regulated, while the expression of the glutaminase enzyme is increased [131], the uptake of glutamate by astrocytes is insufficient, and blockade of the uptake of glutamate in the synaptic cleft also leads to disruption of glutamate homeostasis. Abnormal accumulation of glutamate in the synaptic cleft leads to overstimulation of GluRs, which eventually leads to excitotoxic damage to neurons and oligodendrocytes. In MS, glutamate excitotoxicity is an important bridge linking neuroinflammation and neurodegeneration [132, 133].

Glial cells in the nervous system play an important role in the disturbance of the glutamate neurotransmitter system in MS. In MS, activated microglia inhibit the expression of neuronal cell surface glutamate transporters (GluTs: EAAT1, EAAT2, and EAAT3) involved in glutamate uptake [131, 132, 134]. Microglia also secrete chemokines to recruit peripheral immune cells to infiltrate and enhance glutamatergic transmission, including monocytes, macrophages, and lymphocytes. Activated microglia promote TNF-α release and further trigger the excitation of post-synaptic currents [135].

In an ex vivo MS-CSF incubation study on striatal slices, TNF (expressed at higher levels in the CSF of patients with progressive MS than in patients with RRMS) promoted glutamatergic transmission by increasing the duration of sEPSCs, leading to neuronal swelling [136]. This suggests that TNF acts as a major neurotoxic molecule during MS progression (Fig. 2). In MS, increased pro-inflammatory cytokines and free glutamate in the synaptic cleft upregulate neuronal GluR expression and exacerbate synaptic dysfunction, thereby enhancing local glutamate excitotoxicity [137]. This is the main mechanism of neurodegeneration caused by glutamate excitotoxicity in MS, so drugs that modulate GluR expression and function may have a protective effect on neuronal excitotoxicity. GluR antagonists have been shown to ameliorate motor deficits and neuropathology in the EAE model [138]. In addition, drugs that inhibit the release of glutamate from nerve cells and immune cells have been reported to have beneficial effects in EAE [139], and inhibition of glutamate release from immune cells has also shown promising therapeutic prospects in MS patients [129]. Therefore, modulating extracellular glutamate levels appears to be one of the promising therapeutic strategies for preventing neurodegeneration in MS.

The GABAergic System in MS

In contrast to the abnormal accumulation of glutamate neurotransmitters in the synaptic cleft in MS patients, lower levels of GABA have been detected in their CSF and blood [140]. CSF from MS patients with acute brain injury impairs GABAergic transmission in normal mouse striatal slices and the growth factor IL-1β may be responsible for this effect. Exogenous IL-1β mimics the effect of MS CSF on synapses and the phenotype is blocked by incubation with IL-1R antibody. More recent transcriptomic and proteomics analyses have shown that molecular proteins regulating GABA concentrations in the synaptic cleft are abnormally expressed in MS patients [141, 142]. The GABA concentration in the synaptic cleft is known to be regulated by both NMDA and GABA_A receptors. Ex vivo experiments have demonstrated that the reduced GABA_A-mediated neuronal inhibition depends on IL-1β (Fig. 2).

In addition to its neurotransmitter function, GABA is an important immunomodulator that affects a variety of immune cell functions, including cytokine release, cell proliferation, and the phagocytic activity of immune cells [143, 144]. The role of GABA in regulating immune system function enriches our understanding of its role as a neuroprotective and anti-excitotoxic molecule.

Conclusions and Perspectives

During the pathogenesis of MS, the roles of adaptive and innate immune cells depend on the disease process and the spatial region of the brain. We need to better understand the immune cell populations and pathways that regulate glial interactions with neurons, oligodendrocytes, and myelin by using advanced technologies, such as massive single-cell sequencing datasets of glia in health and disease, cerebral organoids, clustered regularly interspaced short palindromic repeats (CRISPR) screening, transgenic knockout of specific cells in the mouse model, high content imaging, and constantly improving sequencing technologies. The diverse and dynamic properties of the microglia, astrocytes and other glial cells present in disease pathology are potential therapeutic targets. The ability to treat MS in the future could rely on a strategy that combines immune intervention with early neuroprotective, remyelinating drugs.