Mitochondrial and metabolic dysfunction of peripheral immune cells in multiple sclerosis

The autoimmune response is considered a key pathogenic mechanisms underlying MS, wherein an abnormal immune response targets self-antigens within the CNS [34]. This response is triggered by the activation of autoreactive T cells and B cells, which recognize and attack myelin and other components of the CNS [2, 35]. Activated myelin-specific CD4⁺ T cells that react to myelin antigens are primarily responsible for this reaction [36]. Myelin antigens, together with HLA class II molecules and accessory molecules on the surface of APCs, reactivate CD4⁺ T cells in their native environment. The presence of autoantibodies against myelin proteins in the serum and CSF of MS patients indicates that the immune system perceives these proteins as foreign entities and initiates an immune response against them. [37, 38]. Additionally, the presence of inflammatory cells, such as T cells and B cells within MS lesions demonstrates an ongoing immune response within the CNS [39]. Reactivation of MS triggers the release of inflammatory cytokines and soluble mediators, which disrupts the blood–brain barrier (BBB), stimulates chemotaxis, and leads to a larger-scale influx of inflammatory cells into the CNS [40] (Fig. 1A). The experimental autoimmune encephalomyelitis (EAE) model, induced by immunizing animals with myelin-derived proteins (or polypeptides), such as myelin oligodendrocyte glycoprotein (MOG), proteolipid protein (PLP), and myelin basic protein (MBP), serves as a classic animal model of MS. EAE mice model is characterized by the myelin-reactive CD4⁺ T cells and B cells [35, 41].

The chronic inflammatory response in MS is characterized by the activation and proliferation of T cells, B cells, and other immune cells, resulting in the release of pro-inflammatory cytokines, chemokines, ROS and nitric oxide (NO) [16] (Fig. 1B). This chronic inflammation leads to the destruction of myelin, axons, and oligodendrocytes, ultimately causing neurological dysfunction and disability [42]. Recent studies have implicated the innate immune system in the pathogenesis of MS. Activation of toll-like receptors (TLRs) activation leads to the release of pro-inflammatory cytokines and chemokines, contributing to the chronic inflammatory response in MS [43]. Pro-inflammatory Th1 and Th17 have been associated with the pathology of MS [44]. On the contrary, regulatory T cell (Treg) mediates the dysregulation of T cell response in MS by reducing their number and activity [45, 46]. T cells and B cells can be detected in damaged white and gray matter [47]. In MS, B cells differentiate into memory cells or plasma cells, which subsequently secrete autoantibodies involved in antibody-dependent cellular cytotoxicity and complement damage. B cells also have the ability to secrete pro-inflammatory cytokines, including lymphotoxin (LT) and tumor necrosis factor (TNF)-α, significantly contributing to T cell activation [48]. Mucosal-associated invariant T (MAIT) cells, a subset of innate-like CD8⁺ T cells, express semi-invariant T cell receptors as well as accumulate in the MS brain and produce IL-17 [49]. Innate immunity involves mast cells, macrophages (Mϕ) and natural killer (NK) cells. Mast cells in MS plaques release histamine and trypsin simultaneously to facilitate BBB opening [16]. Mϕ produce cytokines such as interleukin (IL)-12, IL-23, and are also involved in the clearance of demyelinating debris [16]. Some subgroups of microglia and Mϕ produce TNF-α, ROS and NO in the activated state, exerting direct neurotoxic effects. CD56^bright NK cells promote the production of granzyme B, which is involved in cytotoxicity and autologous CD4⁺ T cells proliferation [50].

Peripheral immune cells are activated and enter the CNS where they contribute to the destruction of the myelin sheath [51] (Fig. 1C). Myelin-specific T cells exacerbate demyelinating lesions in CNS autoimmunity by releasing cytokines such as IL-17, interferon (IFN)-γ, and granulocyte macrophage-colony stimulating factor (GM-CSF) [52, 53]. Besides, autoreactive T cells mistakenly target and attack myelin, which is the protective covering of nerve fibers in the CNS. These activated T cells release inflammatory cytokines and chemokines that attract other immune cells to the site of inflammation, leading to the further damage to the myelin sheath [54, 55]. In progressive MS, cortical demyelination is associated both with B cell-rich structures in the meninges and plasma cell accumulation in experimental CNS inflammation [56‐58], which may be involved in secretory products independent of antibodies, and multiple cytokines produced by B cells in progressive MS patients are cytotoxic to oligodendrocytes and neurons [59, 60]. On the other hand, autoreactive B cells can differentiate into plasma cells, producing antibodies that bind to the myelin sheath and oligodendrocyte proteins. These bound antibodies result in the induction and activation of the complement proteins on tissue surfaces, directly damaging the myelin sheath and exacerbating demyelination [61, 62]. Inflamed monocytes derived from individuals with MS exhibit a diminished ability to engulf and remove myelin debris, a process crucial for prompt and effective remyelination [63]. Mϕ release a substantial quantity of cytokines and NO, penetrating the CNS, which is a vital element in initiating the demyelination response in MS [64]. Distinct subgroups of Mϕ exert contrasting influences on myelination, with inflammatory Mϕ facilitating demyelination [65] and immune-modulatory Mϕ facilitating remyelination [66]. Dendritic cells (DCs) culture stimulated with low-level IFN-γ-released exosomes can increase myelination and reduce oxidative stress both in vitro and in vivo in EAE mice [67].

The modulation of T lymphocyte homeostasis in MS is influenced by mitochondrial involvement. Several mitochondrial proteins, including B cell lymphoma 2 (Bcl2), ocular atrophy 1 (OPA1), prohibitin 2 (PHB2), Sirtuin-3 (SIRT3), mitochondrial metalloendopeptidase OMA1, and autophagy related 5 (ATG5), are closely associated with mitochondria-mediated death in MS T cells, probably caused by oxidative stress [29, 68‐72]. Some studies have shown that T cells from MS patients exhibit altered mitochondrial structure, reduced glycolysis, downregulated expression and activity of OXPHOS subunits, and decreased mitochondrial membrane potential (MMP) [73‐77]. Treatment with IFN beta-1α has the potential to restore the diminished glycolysis and mitochondrial respiration activity observed in T cells derived from patients with RRMS. This restorative effect is linked to elevated levels of aldolase, hexokinase 1 (HK-1), enolase 1 (ENO1), glucose transporter 1 (GLUT1), dihydrolipoamide-S-acetyl transferase (DLAT), and dihydrolipoamide-S-succinyl transferase (DLST) production [77]. Another study found that during relapse, RRMS patients exhibited increased extracellular acidification rate (ECAR) and oxygen consumption rate (OCR), which were not observed during the remission when compared to the healthy controls. Teriflunomide treatment could prevent the proliferation of metabolically active high-affinity T cells involved in the progression of RRMS patients during relapse. This effect is associated with a functional downregulation of OCR and ECAR, along with complex III activity of the ETC in activated T cells. However, it has no impact on mitochondrial content or structure [78].

B cells can be found in CNS lesions throughout all stages of MS, primarily localized to the perivascular cuffs. B cells serve as effective APCs expressing costimulatory molecules such as CD40, CD80, and CD86 as well as MHC class II [61, 106]. Compared to myeloid APCs, they exhibit significantly higher efficiency, approximately 10,000 times, in capturing soluble and membrane-tethered antigens and presenting them to T cells [107]. In SPMS patients, mitochondrial damage has been observed around CNS periventricular CD20⁺ B cells [108].

In CSF memory B cells, single-cell RNA sequencing studies have shown an increase in cholesterol production [109]. Obeticholic acid (OCA) is a unique medication that regulates various metabolic processes, including cholesterol production, glucose metabolism, inflammation, and apoptosis [110]. It has been proven the effectiveness of OCA in treating EAE, reducing high-density lipoprotein (HDL) levels, cholesterol, and the number of B cells, while OCA alleviates B cell exhaustion by reducing the expression of programmed cell death protein 1 (PD1) and programmed death ligand 1 (PD-L1) [111]. Evidence suggests that increased levels of glycolytic enzymes, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and triosephosphate isomerase (TPI), in the CSF may contribute to B cell clonal growth in the CNS of MS patients [112].

In vitro studies have demonstrated that inhibition of mitochondrial respiration decreases B cell activation (but not glycolysis). On the other hand, both mitochondrial respiration and glycolysis are involved in B cell proliferation, with a notably higher dependence on mitochondrial respiration. Additionally, the expression of costimulatory molecules CD80 and CD86 is reliant on mitochondrial respiration rather than glycolysis. The dysregulated activation and upregulation of CD80 and CD86 on B lymphocytes from untreated individuals with MS can be effectively counteracted by inhibiting Bruton’s tyrosine kinase (BTKi) as well as modifying metabolic processes. Furthermore, BTKi therapy reduces OCR in circulating B cells, attenuates their activation and antigen presenting potential in vivo, partially through the PI3K/AKT/mTOR pathway [113]. Taken together, these findings indicate that glycolysis, cholesterol metabolism, and mitochondrial respiration play crucial roles in regulating B cell functions associated with the pathophysiology of MS. Table 2 provides a summary of the effects of mitochondrial and metabolic dysfunction pathways on B cell activities.

Monocytes, a type of immune cell, have been shown to influence mitochondrial function in patients with MS. They play a significant role in the demyelination process observed both in MS and EAE and they exhibit high adaptability. Once they infiltrate the CNS during disease progression, monocytes can differentiate into either inflammatory or immune-modulatory macrophages, depending on local conditions. The ratio of these macrophage types ultimately determines the course of MS [114]. Functional enrichment analysis using GWAS datasets revealed a strong association among unique risk genes in monocytes and mitochondria and lipid metabolism [11].

The small molecule inhibitor 6877002 targeting the CD40-TRAF6 pathway alters human inflammatory monocytes, resulting in reduced production of ROS, TNF, and IL-6, while increasing the production of IL-10. This treatment also enhances the ability of monocytes to reduce trans-endothelial migration, which is significant when monocytes infiltrate the CNS during neuroinflammation in EAE [115]. Lactate dehydrogenase (LDH), consisting of lactate dehydrogenase A (LDHA) and LDHB subunits, is a crucial enzyme in the anaerobic glycolysis metabolic pathway implicated in MS pathology [116]. LDHB favors the conversion of lactate to pyruvate, while LDHA, as a rate-limiting enzyme, has a higher affinity for pyruvate, converting it into lactate [117]. Remarkably higher levels of LDHA on CD11b⁺ monocytes in blood from EAE mice have been observed compared to the control condition [118].

Monocytes derived from EAE mice treated with 2-DG display decreased glucose uptake and reduced lactate production in the surrounding medium. Additionally, these monocytes exhibited decreased ECAR, which correlated with a decrease in the expression of key glycolytic enzymes including Glut1, HK-2, TPI, pyruvate kinase M (PKM), LDHA, and monocarboxylate transporter (MCT) 1. Monocytes treated with 2-DG adopted an immune-modulatory phenotype, and once them transplanted into EAE mice, resulted in a significant improvement in disease severity [114]. These findings indicate that glycolysis is highly upregulated in activated monocytes in MS.

Comparison between the MS group and the control group revealed that monocytes treated with MS-CSF produced more pyruvate and glutamine. Extracellular levels of glutamine, lactate, and pyruvate were significantly increased, while levels of glutamic acid were significantly decreased [119]. Quantification of monocyte numbers and assessment of monocytic ROS levels revealed an upregulated trend following DMF administration, allowing for the distinction between responders and non-responders. Mechanistically, a single nucleotide polymorphism (SNP) in NOX3 (rs6919626 A) is associated with ROS production and responsiveness to DMF treatment [120]. Mitochondrial dysfunction pathways were upregulated in monocytes of individuals who did not respond to IFN-beta treatment. Monocytic ROS production slightly decreased, and several mitochondrial ETC-related genes showed significant alterations in monocytes that responded to IFN-beta treatment compared to non-responders [121, 122]. These results suggest that mitochondrial translation, oxidative stress, and intracellular glycolysis are crucial for monocyte function and pathology, particularly in the context of mitochondrial therapy drugs (Fig. 3). Table 3 provides a summary of the pathways involving mitochondrial and metabolic disorders that impact monocyte activities.

Mϕ actively participate in the immune response to autoimmune diseases and have been implicated in the destruction of myelin and axons in MS lesions [123]. Immunoregulatory Mϕ (M2) drive the remyelination of damaged axons, produce neurotrophic factors, and facilitate oligodendrocyte function [64]. DMF efficiently inhibits antigen-presenting function in Mϕ, suppressing pro-inflammatory mediators such as ROS, inducible NO synthase (iNOS), TNF-α, IL-1β, and IL-6. DMF achieves this by decreasing extracellular signal-regulated kinase (ERK) phosphorylation, promoting M2-like Mϕ in an EAE mouse model [124]. Furthermore, it has been demonstrated that DMF significantly blocks glycolysis by inhibiting GAPDH in murine Mϕ [125]. This may help explain how DMF promotes M2-like Mϕ, considering that M1 Mϕ primarily utilize glycolysis while M2 Mϕ exhibit greater OXPHOS activity [125, 126]. Moreover, FhHDM-1, a 68-mer peptide secreted by the helminth parasite Fasciola hepatica, inhibits the progression of EAE/MS by upregulating OXPHOS and decreasing glycolysis in Mϕ. This leads to the inhibition of Mϕ activation and production of pro-inflammatory cytokines, such as TNF and IL-6 [127, 128].

ROS are rapidly produced in the CNS of MS patients, predominantly by activated Mϕ responsible for demyelination and disruption of axons [129, 130]. These oxidative bursts in Mϕ contribute to focal axonal degeneration, which is further exacerbated by mitochondrial damage and iron release from MS lesions [131, 132]. The p47phox subunit of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, primarily localized within the zone of initial damage or lesions in patients with acute or early RRMS, accumulates in infiltrating Mϕ and mediates ROS generation [133, 134]. RNA sequencing analysis reveals that fibrin induces the core oxidative stress signature (including neutrophil cytosolic factor (Ncf) 2, Sod2, and Nox2) and the p47phox subunit of NADPH oxidase expressed by ROS⁺ Mϕ [135, 136]. In an EAE mouse model, the transcriptomic signature of fibrin indicates a decrease in oxidative stress, demyelination, axonal harm, and prevention of paralysis when mice are exposed to the fibrin-targeting antibody 5B8. This evidence suggests that fibrin plays a significant role in triggering gene programs associated with oxidative stress during the activation of peripheral Mϕ in EAE [135]. Taken together, these findings suggest that the inflammation-associated oxidative burst in Mϕ contributes to demyelination and tissue injury mediated by free radicals in the pathogenesis of MS.

MCTs play a crucial role in exporting lactate in glycolytic cells [137]. MCT-4 and LDHA are upregulated in inflammatory Mϕ of postcapillary venules in MS patients. Knockdown of LDHA or MCT-4 significantly reduces lactate levels in macrophage supernatant after lipopolysaccharide (LPS) treatment. Additionally, the MCT-4 inhibitor α-cyano-4-hydroxy-cinnamic acid (CHCA) effectively decreases lactate levels and prevents inflammatory activity in Mϕ. The LDHA inhibitor 3-dihydroxy-6-methyl-7-(phenylmethyl)-4-propylnaphthalene-1-carboxylic acid (FX11) also exhibits lactate-reducing properties in Mϕ, inhibiting their inflammatory function [118]. Another study shows that enhanced glycolysis in infiltrating MHC-II⁺ Mϕ leads to increased expression of glucose transporters (GLUTs) such as GLUT1, GLUT3, GLUT4, and MCT1, actively contributing to demyelination in MS lesion tissue [138]. These findings indicate that elevated glycolysis is partially associated with increased expression of glycolytic kinase proteins, which are integral to inflammatory macrophage activation and contribute to demyelination in MS lesions (Fig. 4). The impact of mitochondrial and metabolic dysfunction mechanisms on Mϕ functions is summarized in Table 4.

Neutrophils, a crucial population of granulocytes, play a key role in initiating inflammatory responses [139]. Neutrophils in MS display an activated phenotype characterized by increased surface expression of TLR-2 and N-formyl-methionyl-leucyl-phenylalanine receptors (fMLPR). Additionally, their adhesion and migration energies are increased [140]. Once neutrophils enter the target site and become activated, they exhibit various effector functions aiming at neutralizing the entry of pathogens. These effector activities include phagocytosis, degranulation, and the secretion of ROS.

The abundance and function of mitochondria in neutrophils have been discussed for a long time. It is believed that instead of involved in ATP synthesis primarily, neutrophil mitochondria appear to have diverse functions including mtDNA copy number and mitochondrial potential [141]. Moreover, a study found no significant differences in intracellular ROS generation between RRMS patients and healthy controls [142]. However, selective deletion of CXCR2 in neutrophils is crucial for downregulating Ncf1. This downregulation leads to the inhibition of ROS and IL-1β production in neutrophils, resulting in attenuated CNS neuronal damage in EAE disease [143]. In case of lacking Socs3, neutrophils demonstrated heightened activation of granulocyte colony-stimulating factor (G-CSF)/STAT3 signaling, resulting in atypical EAE characterized by intensified neutrophil activation and elevated generation of ROS [144]. The impact of mitochondrial and metabolic dysfunction pathways on neutrophil activities is summarized in Table 5.

DCs, a type of APCs, are divided into conventional DCs (cDCs) and plasmacytoid DCs (pDCs) [145]. pDCs produce a large amount of type I IFN, which plays a role in antiviral innate immunity and exists in the blood as immature cells [146]. The number of cDCs and pDCs is significantly upregulated in the CSF of MS patients [147, 148]. In EAE, pDCs promote Tregs expansion and inhibit the development of CD4⁺ T cells into Th1 and Th17 cells [149, 150]. Interferon beta-treated DCs induce the gene expression of IL-12p35 and IL-27p28, which suppress the differentiation of Th17 cells and induce IL-10 secretion through the activation of STAT1 and STAT3 in MS [151].

Research highlights the importance of sirtuin 6 (SIRT6) in regulating several processes such as DNA repair, inflammation, immunology, and energy metabolism, including glucose and lipid metabolism [152]. Inhibition of SIRT6 attenuates the onset of EAE by reducing DC migration and activation, potentially mediated via metabolic pathways [153]. NADPH oxidase 2 (NOX2) plays a crucial role in regulating the endocytosis of MOG antigen processing in DCs and supports the presentation of MOG antigen to CD4⁺ T cells through LC3-associated phagocytosis (LAP). Deletion of NOX2 in DCs inhibits the myelin peptide presentation of DCs, preventing the recruitment of CD4⁺ cells to the CNS in EAE mice [154]. Activation of HIF-1α by lactate induces NDUFA4L2 and limits the pro-inflammatory activities and activation of DCs. The inhibition of transcription factor X-box binding protein 1 (XBP1) by NDUFA4L2 in DCs plays a crucial role in controlling pathogenic autoimmune T cells in EAE mice. This regulatory mechanism is achieved through the downregulation of oxidative phosphorylation reprogramming and ROS production [155].

Loss of responsiveness to TGF-β receptor II (TGF-βRII) in monocyte-derived DCs (moDCs) results in increased secretion of IL-12 by moDCs, inducing polarization toward an interferon-gamma-producing Th1 phenotype and exacerbating EAE disease. Mechanistically, upregulated IFN-γ increases Nox-2 and ROS production in moDCs during chronic EAE [156]. DMF induces type II DCs by reducing intracellular GSH through the formation of conjugates, leading to upregulation of heme oxygenase-1 (HO-1) levels [157]. GSH acts as an effective scavenger of ROS, while HO-1 is a sensitive heat shock protein that may result in higher ROS production [158, 159]. Additionally, DMF prevents DCs from expressing pro-inflammatory cytokines and costimulatory molecules [160, 161]. In part, the modulation of the GSH-HO-1 pathway by DMF treatment regulates the oxidative stress during DC maturation and antigen-presenting capacity. These findings suggest that targeting mitochondrial activities of DCs holds promise for the treatment of MS. Table 6 provides a comprehensive overview of the impact of mitochondrial and metabolic dysfunction pathways on DC activities.