The link between gut microbiota and multiple sclerosis from the perspective of barrier function

In fact, early evidence suggesting the involvement of the GM in autoimmune diseases can be traced back to studies on EAE. Germ-free (GF) mice, which are bred and maintained in isolators to prevent exposure to and colonization by microbiota, exhibit immunological immaturity. The mice show a marked reduction in proinflammatory Th17 cells and a skewing toward Th2 responses (22). Furthermore, GF mice demonstrate reduced severity of EAE, corresponding with lower levels of proinflammatory cytokines in the intestine and spinal cord, alongside an increase in regulatory T cells (Tregs) (23, 24). Perhaps most importantly, GM from patients with MS (PwMS), when transplanted into GF transgenic mice expressing myelin-reactive T cells, leads to an increased incidence of spontaneous EAE, which underscores the sufficiency of GM changes in MS for driving CNS autoimmunity (25).

Numerous studies have documented that PwMS experience gut dysbiosis (26 –31) as summarized in Table 1. Specifically, PwMS exhibit an enrichment of bacterial genera such as Ruminococcus, Blautia, Dorea, Bifidobacterium, Bilophila, Sutterella, Pedobacteria, Flavobacterium, Pseudomonas, Acinetobacter, Eggerthella, and Akkermansia. Conversely, there is a reduced abundance of genera including Clostridium, Faecalibacterium, Eubacterium, Parabacteroides, Haemophilus, Adlercreutzia, Ruminococcus, Butyricimonas, Bacteroides, Coprobacillus, Lactobacillus, and Prevotella in PwMS. Furthermore, PwMS have been observed to possess a distinct mycobiome (fungus) (29, 32).

Studies also indicate the correlation between gut microbiota species richness and MS phenotypes. Patients who were clinically non-active exhibited an increased abundance of Faecalibacterium prausnitzii, Gordonibacter urolithinfaciens, Anaerostipes hadrus, Gemmiger formicilis, and Roseburia inulinivorans (28). Specific microbial taxa were also found to be linked with a reduced risk of MS relapse, such as Butyricicoccus desmolans, Odoribacter splanchnicus, Lachnospiraceae NK4A136, and Ruminococcaceae, whereas Blautia, Lachnoclostridium, Lachnospiraceae_UCG-004, and Coriobacteriales were associated with an increased risk (33). In both relapsing-remitting MS (RRMS) and progressive MS, Clostridium bolteae, Ruthenibacterium lactatiformans, and Akkermansia was observed, while a decrease in Blautia wexlerae, Dorea formicigenerans, and Erysipelotrichaceae CCM was noted. The administration of disease-modifying therapies (DMTs) also impacts the microbiota composition. For instance, reduced levels of Prevotella and Sutterella have been observed in patients with untreated MS (34).

The altered gut microbiome observed in MS has fueled intense research interest in elucidating the factors that shape this microbial community and the mechanisms through which these microbes may influence MS pathogenesis. Microbiota dysbiosis can initiate a cascade of events, including the proliferation of pathogenic bacteria and the release of harmful toxins, resulting in a proinflammatory environment and a compromised gut barrier (35, 36). The leaky gut syndrome (LGS) is further characterized by increased intestinal permeability, which facilitates bacterial translocation and the growth and colonization of pathobionts, thereby triggering systemic inflammation. Numerous studies have demonstrated that EAE mice exhibit increased gut permeability (35, 37). Thus, LGS is linked to MS and can play an important role. We will discuss the specific relationship among between LGS, MS, and gut microbiota dysbiosis in Section 3.1.3.

Furthermore, certain bacteria can directly modulate the immune system in MS, influencing the development and behavior of immune cells, such as CD4 T cells, B cells, DCs, and macrophages, which are the main culprits in the pathophysiology of MS (38). MS is predominantly mediated by myelin-specific CD4+ T helper cells, with the Th17 cell lineage being particularly implicated. Th17 cells are known to produce the pro-inflammatory cytokine IL-17 and migrate to the CNS during active disease phase (39, 40). Concurrently, PwMS display dysregulation in CD4+ Tregs, characterized by a marked reduction in their suppressive ability (41, 42). GF mice demonstrate resistance to EAE and a lack of Th17 cells (24). However, mono-colonization with segmented filamentous bacteria (SFB) is sufficient to restore susceptibility to EAE disease and to induce the expansion of Th17 cells in the CNS (24). These findings underscore the necessity of gut bacteria in the EAE pathology. Interestingly, certain bacterial species, such as Akkermansia muciniphila and Acinetobacter calcoaceticus, are also capable of activating intestinal Th17 cells and promoting inflammation in the spinal cords of EAE mice (7, 43 –45). These bacteria are found in increased abundance in the small intestine of PwMS, potentially enhancing the pathogenicity of CNS-autoreactive T cells within the intestine. The microbiota can also influence the expansion and maintenance of Tregs (32). Kasper et al. showed that the Bacteroides fragilis can enhance Treg numbers in cervical lymph nodes, which leads to amelioration of EAE (46). Moreover, alterations in gut microbiota composition may indirectly influence the capacity of Tregs to control autoimmunity by inducing of Th1 and Th17 cells or by modulating the T cell microenvironment (47). Besides well-established T cell, recent research has elucidated a gut microbiota-dependent, anti-inflammatory function of B cells in MS. Rojas et al. found a marked reduction in immunoglobulin A (IgA)+ plasma cells within the gut during EAE (48, 49). A subsequent study revealed that IgA+ B cells migrate across the BBB during active MS and exhibit specificity for MS-associated immunostimulatory bacterial strains.

Microorganisms also influence the innate immune response. Toll-like receptors (TLRs), widely distributed on immune cells and nonimmune cells including intestinal epithelial cells, neurons, and glial cells, are pattern recognition receptors (PRRs) capable of detecting exogenous and endogenous pathogenic molecules (50 –52). They can be activated by microbe-related antigens like peptidoglycan, lipoteichoic acid, and LPS in the intestine, thereby initiating downstream reactions by recruiting signaling molecules through the myeloid differentiation factor 88 (MyD88) pathway or MyD88-independent signal transduction (53). MyD88 signaling results in the activation of transcription factors including the nuclear factor kB (NF-κB) and activating protein-1 (AP-1), which activate the expression of a variety of genes encoding proinflammatory cytokines and chemokines, as well as molecules important in antigen presentation. This could facilitate the reactivation of myelin-reactive T cells in the target tissue in EAE and MS (54, 55). MyD88 knockout mice are resistant to the development of active EAE, further supporting for a role of MyD88-dependent signaling in disease development (56). The composition of the intestinal flora also impacts TLRs expression, thereby influencing intestinal barrier integrity and immune homeosis (57). Microbiota is also a source of signaling molecules, immune mediators, and gut hormones, which have been shown to be involved in TLR signaling (51, 58, 59).

GMs co-evolved with humans to provide essential enzymes for digesting complex fibers, transforming humans into holobionts dependent on gut bacteria for functions such as vitamin production, nutrient digestion, and immune regulation (60 –62). Dysbiosis can result in alterations in the metabolites, contributing to a proinflammatory environment (63). Recent advancements in metabolomics technology have revealed significant changes in microbial metabolites like short-chain fatty acid (SCFAs), tryptophan metabolites, bile acids, and phytoestrogens in MS. Furthermore, various classes of bacterial compounds like LPS, have been shown to penetrate systemic circulation, even reaching the CNS. We will discuss these topics in Section 3.3.

The BBB refers to the barrier between plasma and brain cells, formed by brain endothelial cells (BECs), the perivascular foot processes of astrocytes, a basement membrane (BM), and pericytes (PCs). In July 2001, the National Institute of Neurological Disorders and Stroke introduced the neurovascular unit (NVU) concept to highlight the dynamic interactions between the BBB, neurons, extracellular matrix, and microglia, which collectively regulate BBB structure and function (64, 65).

BECs exhibit low penetrance to intravascular materials due to a thick luminal glycocalyx layer, specialized tight junction (TJ) structures, lack of fenestration, and selective transporters, which underpin the trans-endothelial permeability and guarantee metabolic and immunological homeostasis for normal brain functions (66). BECs also actively recruit inflammatory cells into the CNS to suppress local inflammation at BBB (67). NVU astrocytes have versatile roles, supporting vascular endothelium, responding to immune stimuli, forming endfeet and the glia limitans, and regulating intracerebral fluid flow. Astrocyte endfeet, together with secreted basal material, form the glia limitans, an immune barrier preventing T cells entry into the parenchyma (68). Connexin 43 in astrocytes helps maintain the BBB, and its loss causes continuous immune cell recruitment (69).Pericytes ensure endothelial integrity, regulating astrocytic endfeet, leukocyte trafficking, and vascular immune homeostasis and vasomotor (69, 70). Pericytes can regulate astrocytic endfeet and BBB endothelium formation. They also limit lymphocyte and monocyte transmigration into the brain (71, 72).

The BM of the BBB is a multilayered extracellular matrix composed of laminin, collagen IV, nidogen, and proteoglycan, formed by the interplay between astrocytes, BECs, and pericytes (66, 73). It allows fluid and soluble molecule passage while blocking leukocyte infiltration and binding growth factors (73). BM laminins affect T lymphocyte extravasation and migration into the brain. The perivascular space (PVS), situated between the BM secreted by BECs and astrocytes, is a key component of the highly organized glymphatic system, which include meningeal lymphatic vessels (MLVs) (69, 74). This system, characterized by astrocyte endfeet expressing polarized aquaporin-4 (AQP4) water channels, shares key functions with the peripheral lymphatic vessels and aids in CSF-interstitial fluid (ISF) exchange, waste removal, and immune cells trafficking (75, 76). The CSF exchanges with ISF through the PVS of the penetrating arteries and is ultimately drained by arachnoid granulations and MLVs, aiding nutrient delivery and metabolic waste clearance within the brain parenchyma (77, 78). Macromolecules in the subarachnoid space are transported to deep cervical lymph nodes (dCLNs) and superficial cervical lymph nodes via MLVs (79). MLVs also work with CNS immune cells, including microglial, to regulate immuno-lymphatic interface and enhance neurotrophic signaling (79, 80). Studies have shown that AQP4 serves as a crucial regulator of fluid dynamics within the brain (81).

Unlike the BECs, the choroid plexus vascular barrier (PVB) is fenestrated, allowing small molecules, water, and solutes to pass, which is crucial for CSF production (82). Animal models have demonstrated that the PVB remains permissive under normal physiological conditions but can close in response to intestinal and systemic inflammation (82, 83).

Oligodendrocytes and microglia are also crucial for BBB integrity, although they do not directly form the BBB. Seo et al. revealed oligodendrocytes can enhance TJs via TGF-β signaling (84). Studies indicate that BBB leakage initially causes astrocyte damage, followed by alterations in oligodendrocytes. Multiple mechanisms, such as imbalances in protein synthesis and degradation (85, 86), the impact of aquaporin-1 and AQP4 (87), and ionic equilibrium (88), have been suggested to explain this phenomenon.

To function as an exquisite machine with highly regulatable dynamics, the BBB or NVU is crucial for brain health. Maintaining brain homeostasis requires the NVU coupling to work synergistically among BECs, pericytes, astrocytes, microglia, neurons as well as cerebral lymphatic system.

BECs, with a thick glycocalyx layer on the luminal surface as mentioned above, block macromolecule leakage and leukocyte adhesion. Hypoxia, inflammation, and TNF-α can disrupt the glycocalyx. Therefore, attenuated glycocalyx coats are involved in the early pathogenesis of neuroinflammation and brain aging. Microglia, central players in neuroinflammation, can transform into a phagocytic state and remodel neuronal connectivity, damaging the BBB by engulfing astrocyte endfeet AQP4 when peripheral inflammation breaches microvessels (89). Activated microglia also stimulate astrocytes to release TNF and glutamate (90), interacting with BECs and neurons to produce chemokines to recruit leukocytes into the CNS (91). Importantly, they also communicate with infiltrating lymphocytes and other immune cells, potentially worsening CNS inflammation (75). The cerebral lymphatic system affects MS progression by influencing immune cell movement, inflammatory responses, and oligodendrocytes function. In acute MS lesions, glial cells retraction and astrocyte damage occur, leading to reduced diffusivity along the PVS, which correlates with increased disability and longer disease duration in MS (92, 93). Impaired lymphatic fluid flow results in the accumulation of inflammatory cells and neurotoxic elements, impairing the clearance of toxic molecules and metabolites from the ventricles and deep gray matter, as well as the clearance of inflammatory microglia, thereby exacerbating cortical demyelination and gray matter pathology (94 –96). MLVs facilitate meningeal T cells migration to dCLNs, and their ablation attenuates CD4⁺ T cell infiltration and spinal cord demyelination, improving EAE prognosis (97, 98). However, MLVs may also exert neuroprotective effects in MS by modulating the function of oligodendrocytes and astrocytes (99).

Chemokines, cytokines, and immune cells, may also influence BBB during systemic or local inflammation. Pro-inflammatory cytokines such as IL-1, TNF-α, and IL-6 have been linked to neuroinflammation in the CNS and peripheral nervous diseases including MS, Parkinson's disease (PD), Alzheimer's disease (AD) and diabetic neuropathy (100, 101). They activate signaling pathways like NF-κB and JAK/STAT, leading to neuroinflammation and BBB disruption (102 –104). This disruption allows more immune cells and cytokines into the inflammation lesions, worsening inflammation in MS and other conditions (105, 106). Persistent cytokines activity leads to neuronal loss, demyelination, and chronic activation of microglia, astrocytes, and peripheral immune cells. These cytokines initiate and sustain a neuroinflammatory feedback loop, culminating in BBB breakdown, increased oxidative stress, synaptic dysfunction, and neuronal death. The persistence of this inflammatory environment is a significant element in the genesis and progression of neurological disorders, including MS (107 –109). Many other cytokines like IL-22, and IFN-γ also damage the BBB by modulating TJs and increasing the expression of transmigratory molecules expression on BECs (110 –112). Circulating cytokines also enhance inflammasome activation like NLRP3, which could downregulate TJ proteins and increase BBB permeability (113).

Chemokines are crucial for BBB integrity, lymphocyte chemotaxis, CNS immunosurveillance, and neural regulation (114). Studies have found the chemokines levels, such as CXCL13, CXCL9 and CCL2, are significantly increased in MS (115, 116), potentially activating the p38 mitogen-activated protein kinase (MAPK) pathway and compromising the BBB (117). The NVU controls peripheral leukocytes entry into the CNS, a process that can be disrupted by cytokines. IL-17, in particular, impairs NVU function by attracting circulating neutrophils and downregulating TJs like occludin and ZO-1 (118). Human BECs express low levels of IL-17R normally but increase expression near active MS lesions. It also facilitates CD4+T cell transmigration and enhances the ICAM-1-dependent monocyte adhesion through the BBB (119).

MS has long been seen as a T-cell-mediated disease, especially involving CD4 myelin-reactive T cells including Th1 cells and Th17 cells. Th1 cells primarily secret IFN-γ and TNF-α, which play a crucial role in activating local glial cells and antigen-presenting cells (APCs). Th17 cells can secrete matrix metalloproteinases 3 (MMP-3) and MMP-9, which can degrade the BM and facilitate peripheral leukocyte migration through the BBB (111, 120). Furthermore, Th17 lymphocytes highly express granzyme B, which subsequently kills neurons and recruits more CD4+ lymphocytes (119). Activated Th cells interact with various autoantigens like CNS resident cells, leading to a rapid clonal expansion and an amplified immune response. This ultimately triggers a cascade of inflammatory, demyelinating, and neurodegenerative events, thereby further affecting BBB permeability in MS (121). Recent studies have demonstrated that plasma cells originating from the gut, which secrete IgA, play a role in mitigating neuroinflammation in the CNS through the production of IL-10 (49). In contrast, the accumulation of IgA-producing cells that are reactive to gut bacterial strains associated with MS has been correlated with acute inflammatory episodes in MS (122).

NVU coupling relies heavily on the canonical Wnt/β-catenin, Sonic Hedgehog (SHH), PDGF-β, and TGF-β signaling pathways (123, 124). Additionally, inflammatory mediators, including mitochondrial reactive oxygen species (ROS), can stimulate proinflammatory signaling pathways (Jak-STAT, NF-κB, and NLRs) in BECs, pericytes, and astrocytes, potentially damaging the BBB and interfering with the morphogen signaling (Wnt/β-catenin and SHH) as well as transcriptional program of the BECs, leading to NVU breakage.

The BBB disruption will finally leads to transcytosis, cerebral ion metabolism imbalance, brain perfusion abnormalities, and influx of erythrocytes, cytotoxic iron, as well as fibrinogen, thrombin, and immunoglobulins, which might further drive pathology of MS. The plasminogen cascade activation is associated with MMP activity and BBB disruption in acute MS lesions (125, 126). In progressive MS, postmortem brain tissue shows increased fibrin and fibrinogen deposition in the motor cortex (127). Plasma-derived extracellular vesicles in RRMS are also enriched in fibrinogen (128). Studies reported that fibrin and fibrinogen deposition likely follow BBB breakdown, with blood-derived thrombin mediating further BBB breakdown, eliciting a Ca2+ influx, nitric oxide and ROS production, stress fibers formation, and TJ disruption in MS (129). In the NVU, BECs regulate ion transport through modulating TJ, receptors, and ion channel expression (130). In MS, the BBB disruption might impair selective ion exchange and lead to the neurotoxicity. Iron buildup has been associated with increased ROS, lipid peroxidation, decreased antioxidants, and neurodegeneration in these patients (126). Lastly, modern approaches to BBB disruption in MS also focus on the vascular changes at the NVU, where BBB function and cerebral perfusion are closely interconnected. The NVU harmoniously couples cerebral blood flow with neural activity in different regions of the brain through vascular activity, which has been reported to play a central role in MS pathology (125, 130, 131). Under MS pathology, increased NVU permeability is secondary to BECs dysfunction. Additionally, pericytes contract and undergo apoptosis, leading to capillary constriction and increased BBB damage. Global hypoperfusion in both the white and gray matter is associated with active MS with cognitive dysfunction (132). Therefore, the involvement of the NVU in MS highlights the importance of cerebral hypoperfusion in MS pathology and could represent a potential treatment target.

Studies increasingly show that gut microbes significantly impact BBB integrity. In 2014, Braniste et al. found that GF mice exhibited higher BBB permeability in various brain regions compared to pathogen-free (PF) mice, which was linked to reduced occludin and claudin-5 expression. After fecal transplantation from PF mice or administration of SCFA-producing bacteria to GF mice restored BBB integrity by increasing TJ expression. Current research suggests that the gut microbiota regulate the BBB through a variety of pathways, including the vagus and sympathetic nerves (133), the immune system (134), the endocrine systems (135), and microbial metabolites such as SCFAs, microbial structural components such as LPS and peptidoglycans (8), and microbial membrane vesicles (136). We will discuss this part in the following section.

The intestinal barrier consists of mucus layer, epithelial barrier, and gut vascular barrier, behaving as a coordinated and multilayered network that protects host physiology from external insults and regulates several gut functions. Intestinal immune compartments are divided into inductive sites, like mesenteric lymph nodes and gut-associated lymphoid tissue (GALT), where adaptive immune cells are primed and differentiated, and effector sites, such as the intestinal lamina propria and epithelium, where these cells localize to support barrier integrity and immunity (162).

The gut epithelial barrier, made up of columnar epithelial cells (enterocytes) and specialized secretory cells (Paneth and goblet cells), is underpinned by intestinal stem cells in mucosal crypts (163, 164). The mucus layer, secreted by goblet cells, shields gut epithelial cells from harmful substances and provides a habitat for microbiota, facilitating beneficial interactions while preventing pathogen entry (165, 166). Components such as TJs, antimicrobial peptides (AMPs), secretory IgA, and glycosylated proteins contribute to this protective mechanism (167 –169). Mucosal surfaces also serve as an immune barrier acting as part of the innate immune response against microbial pathogens (170). Beneath the mucus layer, the gut epithelial lining acts as a semipermeable barrier, maintaining a balance between microbial-host interactions. Enterocytes are connected by junctional complexes that regulate paracellular transport and maintain intestinal permeability (166, 171). The gut barrier also features ATP-binding cassette transporters that prevent toxin accumulation and inflammation, influenced by gut microbiota (172, 173). Enteric glial cells, similar to astrocytes in regulating BBB, can also impact gut epithelial barrier function (130, 174, 175).

The immunological layer of the intestinal barrier, following the mucus and the epithelial lining includes innate lymphoid cells and intraepithelial lymphocytes that protect against pathogens and modulate immune responses (176 –178). The GALT consists of multifollicular lymphoid tissues, including Peyer's patches, isolated lymphoid follicles, the appendix, cecal and colonic patches and rectal lymphoid tissues, with all dependent on the microbiota (179, 180). These tissues house diverse immune cells including CD4+ Th cells, Tregs, CD8+ T cytotoxic cells, DCs, macrophages, and innate lymphoid cells (ILCs) that initiate and propagate immune responses, with the IgA+ Marginal Zone B cells (MBCs) being particularly prominent. IgA+ B cells from GALT are vital for the gut-meningeal immune axis and protect the CNS from gut-derived infections (179, 181). GALT that drains mucosal surfaces will constantly encounters foreign structures from commensal microbiota, infectious pathogens and antigens, which also establish tolerance to autoantigens with the changes in autoreactive T cells phenotypes, including peptides from the CNS (182). The gut vascular barrier, with fenestrated endothelium and TJs, prevents microbial entry into circulation and controls the access of dietary compounds (183). This barrier is crucial for gut-brain axis communication, with disruptions linked to the closure of PVB in mice (83). This finding suggests a functional linkage between barriers along the BGM axis, potentially underlying the frequent comorbidity of neurological and gastrointestinal symptoms (184).

The enteric nervous system (ENS), an autonomous division of the autonomic nervous system (ANS), autonomously regulates gastrointestinal functions through its submucosal and myenteric plexuses (185). It contains neurons, glia, and immune cells, forming intrinsic circuits for gastrointestinal motility, secretion, immunity, and tissue repair (186). ENS neurons can release many neurotransmitters and are closely related to vagus efferent input, forming intrinsic sensorimotor circuits (187, 188). Enteric glial cells interact with neurons, EECs, immune cells and epithelial cells, thereby modulating barrier function (175, 189, 190). Studies of human Peyer's patches also reveal peptidergic innervation including Substance P, Vasoactive Intestinal Peptide (VIP) and Calcitonin gene-related peptide (CGRP) immunoreactivity in cells within the GALT (191).Microbiota and enterochromaffin cell (ECC)-derived 5-HT further influence glial homeostasis. Interstitial cells, including interstitial cells of Cajal (ICCs) and platelet-derived growth factor receptor alpha-positive (PDGFRα+) cells, can facilitate gut motility via electrical coupling with smooth muscle (192, 193). Macrophages, the most abundant immune cells in the GI tract, can modulate barrier homeostasis, with their activation influencing ENS integrity and being linked to microbiota dysbiosis (194 –197).

The intricate interplay between gut microorganisms and the immune system is regulated at the gut barriers through multifaceted mechanisms. Intestinal microorganisms and their metabolites impact both immune system and intestinal epithelial barriers (IEBs), while intestinal layers reciprocally shape microbial composition and immune activity. This microbiome-mediated barrier homeostasis is pivotal in regulating of neuroinflammation.

The mucus layer critically governs gut microbial community and immune interactions. Adhesion to host epithelial cells and mucus is a key property for gut bacteria colonization, which can be modulated by host-specific mucin glycosylation (198 –200). Dysregulated mucin glycosylation is correlated with increased inflammation and microbial translocation by regulating mucin degradation (201 –203). The transmembrane mucins also enhance the intestinal immune functions (204). The IEB impairment is a crucial mechanism for several inflammatory and immune-mediated disorders (205). As result of gut barrier imbalances, some microorganisms, bacterial products, and toxins may translocate across the epithelium uncontrollably, leading to both the local and systemic inflammation (34, 206 –209). Paracellular translocation, often linked to the impairment of TJs, has been associated with direct damage to enterocytes and their supporting structures, along with significant changes in intestinal TJs gene expression and downregulation of both ZO-1 and occludin (210, 211).

The interplay between intestinal epithelial cells (IECs) and mucosal immune components, such as intraepithelial lymphocytes and lamina propria immune cells, sustains intestinal immune homeostasis. IECs detect antigens, secrete antimicrobials, and modulate immune responses, while immune cells regulate IEC-derived cytokines (212, 213). Interactions between DCs and IECs maintain anti-inflammatory environments under steady-state conditions (214). Commensals reinforce IEB integrity via TJ regulation and IECs proliferation (215). IECs-associated inflammasomes like NLRP6 are critical for mucosal homeostasis and infection defense (216). Inflammasome-deficient mice exhibit microbiota dysbiosis, amplifying inflammatory responses and IBD susceptibility (217, 218). NLRP6 deficiency also disrupts goblet cell mucus secretion and the production of epithelial IL-18 and AMPs, impairing bacterial control (219, 220), and leading to AMPs imbalance, dysbiosis, and autoimmunity.

Additionally, as previously mentioned, gut microbiota can regulate the expression and phenotype of inflammatory cells both locally and at distant sites. Studies indicate that the microbiota in the GALT remotely influences T cell development in the thymus via soluble factors (221). In the EAE model, MOG-specific T cells proliferate substantially in the GALT under SPF conditions, but less so in germ-free environments, suggesting microbiota-induced T cells stimulation (221). Recent findings emphasize the crucial role of IgA antibody-secreting cells (ASCs) in the CNS, acting as a "brain firewall" to protect the BBB and maintain intestinal homeostasis (222). In both mice and humans, meninges contain gut-derived, commensal-specific IgA ASCs, which help prevent pathogens from entering the CNS (223). Gut bacteria stimulate secretory IgA production, which compartmentalizes commensal bacteria away from the host epithelium and modulates chemotaxis and TLRs signaling (224, 225). In mice, non-invasive bacteria residing gut are coated with IgA, promoting the production of diverse, species-specific IgA in mice (226). In an adaptation to this specific microenvironment, intestinal plasma cells (PCs) might have a distinct metabolic profile. IgA ASCs can utilize diet- and gut microbiota-derived SCFAs as one carbon source to maintain metabolism (227). Inflammatory responses induced by environmental factors or intestinal dysbiosis might dramatically change oxygenation and the metabolic profile of the PC niches in the gut.

Recently identified ILCs are key regulators of intestinal immune responses and have also been implicated in CNS autoimmunity. Among ILCs, ILC3s are notable for their similarities to Th17 cells, which are crucial in CNS inflammation and can be modulated by many cues from the gut microbiota (228, 229). ILC3s are critical for the generation of the organized lymphoid tissue in the intestinal wall and regulating microbiota content and the integrity of the intestinal barrier (230, 231). Found in different GALT compartments, ILC3 interact with immune cells including Th1 cells, Th17 cells, and Tregs, efficiently controlling effector T cells and promoting a Treg balance (232 –234). ILC3s produce IL-17 to attract neutrophils to the intestine during bacterial and fungal infections (235, 236), which can also induce AMPs and TJs production (237). They are also a key source of IL-22, crucial for maintaining the intestinal barrier (238). IL-22 production is stimulated by a glial-derived neurotrophic factor from enteric glial cells in response to TLR ligands (239), and is also enhanced by SCFAs that act through AhR and FFAR, respectively (240 –243).

An altered microbiome also affects bacteria-associated products that influence neuroimmune responses. Besides microbial metabolites, structural components derived from bacterial cell walls and membrane vesicles also significantly impact host physiology and gut permeability, see Section 3.3.

Overall, barrier permeability is dynamic and must be carefully orchestrated and constantly adapted to maintain homeostasis, with gut microorganisms playing a major part in achieving this goal.

The topic of intestinal permeability (IP) in neuroinflammation is actively being studied, with several lines of investigation exploring the plausible relationships between gut barrier disruption and MS, as well as on translational implications based on IP.

In a study of 12 jejunal biopsies from MS patients, Lange and Shiner observed subtle histological changes, including villous atrophy and intestinal inflammatory cell infiltration (244). The latest study used the lactulose/mannitol test to evaluate intestinal permeability in MS patients and found that 73% of cases presented with abnormal permeability (245). Elevated serum zonulin levels in both RRMS and SPMS further confirm diminished intestinal barrier function in MS, as zonulin can rapidly increase both intestinal and BBB permeability in vitro. Similar findings were also described in the EAE model, with increased intestinal permeability, reduced submucosal thickness, and altered TJ expression in IECs, which have been associated with a mucosal imbalance between Th1/Th17 and Treg cell subsets in intestinal lamina propria, Peyer's patches, and mesenteric lymph nodes (41, 246). They also found that treatment with probiotic Escherichia coli strain Nissle 1917 preserved TJs and decreased intestinal permeability, leading to reduced EAE severity and decreased pro-inflammatory cytokines (41).

The above studies indicate that PwMS indeed experience an alteration in the intestinal barrier due to an altered intestinal immune response and microbial dysbiosis (247). The leaky gut may be involved in the pathophysiological process of MS via the following mechanisms. Firstly, intestinal barrier dysfunction has been associated with susceptibility to systemic infections, which are common complications in MS patients (247, 248). Furthermore, the interaction between intestinal barrier and commensal microbiota could modulate the immune response pathologically, shaping the development of immune cells such as CD4+ T cells, B cells, DCs and macrophages. Additionally, changes in intestinal permeability could exacerbate neuroimmune dysregulation by allowing transmucosal passage of injurious or immunogenic antigens. Interestingly, recent work suggests a connection between the IP changes (IPC) and MS risk factors. For example, Vitamin D deficiency may reduce intestinal calcium absorption, causing gut stasis and subsequent IPC, which would allow gut microbiota to transfer more endotoxins into the blood and trigger inflammatory cytokines production within the CNS (249).

Alterations in the gut homeostasis in MS could increase translocation of bacterial and their toxic products through an impaired intestinal barrier. A recent study found higher plasma levels of endotoxin LPS in MS, linked to in vivo IL-6 production and in vitro Th17-like responses (195). In another study, investigators also found increased LPS-binding protein levels in the serum of MS patients (196). Besides LPS, MAMPs such as bacterial lipoproteins and double-stranded RNA can enter the bloodstream and modulate the immune system through TLRs, which are present in microglia and to modulate the initiation and severity of EAE models (197). Dysbiosis may alter gut bacteria metabolites, reducing health-promoting ones like SCFAs and dietary tryptophan, which may further contribute to increased gut barrier permeability and pro-inflammatory setting. Additionally, microbiota dysbiosis disrupts IgA synthesis and AMPs production, which act as anti-inflammatory mediators beyond the gut.

Increased intestinal permeability, alterations in TJs functioning, and modifications in intestinal morphology occurred along with the changes in the immune cells including T cells, IgA ASCs and ILCs, as well as gut microbiota dysbiosis in GALT, thus indicating that disruption of intestinal homeostasis was dependent on the immune response at the initiation of EAE. Thus, the combination of LPS- and MAMPs-induced inflammation, leaky gut, metabolic imbalance, and immune activation creates a perfect storm for dysregulated immune activation that can fuel chronic disease in MS.

Produced by microbiota fermenting dietary fiber and resistant starch in the intestines, SCFAs (acetate, propionate, and butyrate) provide energy for both the host and the gut microbiota, and can enter host circulation and cross the BBB, enabling a role in maintaining barrier integrity (60, 267 –269). Lower levels of SCFAs have been observed in MS patients (270 –273). Furthermore, diminished SCFAs have been correlated with increased intestinal permeability and worsening EDSS in MS (270, 271, 273). Moreover, known SCFAs-producing gut microbiota are reduced in MS, including Butyricimonas, Bacteroides, Lachnospira, and Eubacterium (272, 273). GF mice, naturally lacking SCFAs, show a compromised BBB, while introducing butyrate or butyrate-producing bacteria like Clostridium tyrobutyricum can improve BBB dysfunction in these mice (141). The mechanism by which SCFAs influence barrier function is not fully understood. SCFAs bind G protein-coupled receptors (GPCRs) (141) and the free fatty acid receptors (FFAR2 or FFAR3) on intestinal epithelial cells and brain ECs, protecting the barrier from oxidative stress (274 –278). Knox et al. also found that butyrate and propionate promote remodeling of actin cytoskeleton and TJs in an BBB model (279). In GF mice, SCFAs can improve barrier function and TJs expression at the choroid plexus in antibiotic-treated mice (280). SCFAs are also known to support mitochondrial function (281, 282), as they protect against mitochondrial disruption in brain endothelial cell treated with LPS (279). For intestinal homeostasis, SCFAs can mediate sodium transport, energize intestinal epithelial cells, and influence gene transcription that supports colon homeostasis by inhibiting histone deacetylase activity (283 –285). SCFAs also reduce T cell proliferation and cytokine production in the gut, partly by inhibiting the activation of NF-κB pathway in immune cells and intestinal epithelial cells (286 –288).

Bile acids (BAs), derived from cholesterol metabolites in the liver and modified in the gall bladder, become primary bile acids conjugated with glycine or taurine. Primary Bas, cholic acid and chenodeoxycholic acid (CDCA), can be further metabolized by gut microorganisms into secondary bile acids (2BAs), such as Deoxycholic acid (DCA), chenodeoxycholic acid (CDCA) and lithocholic acid (LCA) (289), which can enter systemic circulation and affect the CNS (290, 291). DCA and CDCA have been shown to have disruptive effects on the gut barrier (292, 293), whereas LCA seems to have a protective role (294). CDCA and DCA have also shown disruptive effects on the BBB in animal models, which may suggest common mechanisms of disruption across barriers (295). BAs can interact with many receptors such as Farnesoid X receptor (FXR), the VDR, PXR and Takeda G protein-coupled receptor 5 (TGR5), to exert various functions (296 –298). Without these receptors, the intestinal barrier weakens, allowing the translocation of bacteria (299). Moreover, FXR modulates gut immune responses driven by microbes during inflammation, potentially linking them to BA metabolism dysregulation (300). Gut microbes can activate TGR5, affecting the expression of EECs involved in immune regulation (301). This, in turn, directly influences macrophage polarization and the subsequent inflammatory response. Once TGR5 is activated, BAs may suppress the production of inflammatory cytokines such as IL-1, IL-6, and TNF-α (302).

Tryptophan is acquired through digestion of dietary protein in the small intestine (303 –305). This essential amino acid is crucial for protein synthesis and the production of serotonin (5-HT) and kynurenine (155, 306, 307). Studies have noted reduced levels of tryptophan and its metabolites in PwMS, also correlating with EDSS scores (308 –310). Dietary tryptophan restriction in EAE models can abolish BBB disruption, leukocyte infiltration, and CNS demyelination, likely by inhibiting Th1/Th17 skewing and impairing migratory capacity (311). This effect is partially lost in GF mice, suggesting a microbiota-dependent mechanism. However, tryptophan and its metabolites can also exert protective effects, which are partially mediated by binding to the aryl hydrocarbon receptor (AhR). AhR regulates astrocyte and microglial crosstalk in the CNS, which controls inflammation and neurodegeneration (312, 313). Furthermore, tryptamine-mediated EAE suppression relies on AhR and modifies the gut microbiome composition to increase butyrate-producing microbiota (310).

Microbial fermentation can also produce compounds like methylamines, indoleacetate, phenylacetate, and phenolic compounds (314), as well as branched-chain amino acids (BCAAs) such as 2-methylbutyrate, isovalerate, and isobutyrate (314). BCAAs may play a role in autism spectrum disorder pathophysiology and barrier modulation (315). Gut microbes convert dietary methylamines dylcholine into trimethylamine (TMA), which is subsequently rapidly converted into TMA N-oxide (TMAO) in the liver and circulates systemically (314). TMAO can enhance BBB function through annexin A1 signaling (307, 311). Bacterial fermentation of dietary tyrosine and phenylalanine into p-cresol (314), whose metabolite, p-cresol glucuronide, protect human BECs line hCMEC/D3 upon LPS challenge (316). Kynurenine has been shown to protect barrier function in a colitis mouse model (317) and, along with tryptophan, crosses the BBB via the amino acid transporter SLC7A5 or L-type amino acid transporter 1, affecting neurotransmitter production (307).

In summary, the gut microbiome significantly influences how diet-related metabolites affect health and disease. A better understanding of how these diet- related metabolites alter the composition and function of gut bacteria could pave the way for improved treatments for PwMS.

Recognizing the significance of structural components derived from bacterial cell walls and bacterial membrane vesicles is also crucial due to their effects on host physiology. Microbial structures, such as LPS and bacterial membrane vesicles, have previously been discussed as regulators of gut barrier function as well as BBB through various signals at the micro-gut-brain axis (193, 289, 318).

LPS, a component of Gram-negative bacteria cell wall, is recognized for its association with compromised gut barrier function and activation of immune system (194, 279, 280). Gut microbiota disorders can increase LPS release, leading to higher intestinal permeability and activation of gastrointestinal immune cells to release inflammatory cytokines (319, 320). In MS, elevated levels of LPS have been detected in the bloodstream (321). The same study also reported increased levels of LPS in the brain, spinal cord, and blood of EAE model (321). LPS activates TLR4 on microglia, leading to the release of inflammatory cytokines and chemokines (322), and promotes neuronal apoptosis and endothelial cells damage (323, 324). And Singh et al. showed that LPS also interacted with lipoteichoic acid on the cell wall of G+ bacteria, reduced mRNA levels of ZO-1, occludin, and JAMs, while increasing levels of TNF-α and IL-1β at the border of NVU. Additionally, LPS also affects adhesion proteins, membrane transporters, the basal lamina, and the extracellular matrix in the BBB (324). Therefore, LPS affects the integrity of BBB and NVU through a variety of mechanisms, providing a potential target for the treatment of related diseases.

Peptidoglycans, found in the cell walls of G+ and, to a lesser degree, G- bacteria, play key roles in host physiology. Bacterial membrane vesicles are lipid bilayer capsules released from the outer membranes of both Gram-negative and Gram-positive bacteria. They can transport and protect various cargoes, including proteins, DNA, RNA, metabolites, enzymes, peptidoglycans, polysaccharides, and toxins (325, 326). Gut microbial membrane vesicles can traverse the intestinal barrier, enter the bloodstream, and cross the BBB, constituting a key component of the BGM axis (200, 327). Notably, these vesicles influence gut barrier function by modulating mucosal innate immune cells such as macrophages and DCs (328). Thus, LPS and other MAMPs could constitute another pathway through which compromised barrier function impacts neuroimmune responses in MS.