Metabolism as Medicine: A New Therapeutic Direction for Multiple Sclerosis

Disease-modifying therapies transformed relapsing-remitting multiple sclerosis (MS) over three decades, cutting relapse rates and slowing disability. They have not done the same for progressive disease. Once patients transition to secondary progressive or present with primary progressive MS, focal inflammatory attacks diminish and the clinical picture shifts to diffuse neuroinflammation compartmentalised behind a closed blood-brain barrier, together with a failure to remyelinate. Badawi and colleagues argue in this narrative review that part of the reason is a blind spot in how the field frames the problem. Research has concentrated on which immune cells are pathogenic and how to block or deplete them, and has paid far less attention to how those cells are bioenergetically programmed to sustain their behaviour. Their central claim is deliberately bounded: metabolic dysregulation acts as an amplifier and a modifiable hub inside MS pathogenesis, not as its primary cause.

Two Metabolic Programs, Two Master Regulators
The organising distinction is between glycolysis and oxidative phosphorylation. Resting naive T cells, regulatory T cells, memory T cells, regulatory B cells, M2-like microglia, and neurons run on oxidative phosphorylation and fatty acid oxidation under AMPK control, which maximises energy yield per glucose molecule and supports longevity. On activation, Th1 and Th17 cells, effector B cells, and M1-like microglia switch to aerobic glycolysis and glutaminolysis under mTOR and HIF-1α. That switch looks wasteful, since glycolysis yields two ATP per glucose against up to 36 from oxidative phosphorylation, but the advantage is speed and raw material. Glucose-6-phosphate diverts into the pentose phosphate pathway to make ribose for nucleotides, and other intermediates feed amino acid and lipid synthesis, which is what a cell needs to copy its contents for clonal expansion. HIF-1α is the part worth understanding in detail. Under normal oxygen it is normally degraded, but mTORC1 signalling stabilises it anyway, a state the review calls pseudo-hypoxia. Once stabilised it turns on nearly every glycolytic gene and also promotes degradation of FoxP3, the master transcription factor of regulatory T cells. One regulator therefore drives inflammation and dismantles regulation at the same time.

How the Lesion Environment Turns Regulators Into Attackers
What follows clinically is that regulatory cells do not simply get outnumbered in MS; they change program. Under the cytokine-rich, glycolytic-metabolite-rich conditions of an MS lesion, regulatory T cells shift away from fatty acid oxidation and towards glycolysis, lose suppressive capacity, and can convert into pro-inflammatory, IL-17-producing ex-Tregs. Simple resource competition compounds this, since rapidly proliferating glycolytic effector T cells consume large quantities of glucose and limit what remains for regulatory cells to maintain oxidative metabolism. Th17 cells add a further dependency: they need glutamine, and they run de novo fatty acid synthesis to build specific pathogenic lipids, a heightened metabolic activity that also leaves them open to ferroptosis, an iron-dependent cell death linked to lipid peroxidation. The B-cell compartment splits the same way, with effector B cells glycolytic and IL-10-producing regulatory B cells dependent on fatty acid oxidation, and the review reports evidence that the B-cell compartment in MS shifts towards the glycolytic, pro-inflammatory side.

Remyelination Failure Gets an Energetic Explanation
The most concrete section concerns repair. Oligodendrocyte precursor cell differentiation is bioenergetically expensive and requires a shift to oxidative phosphorylation with functioning mitochondria. The chronic MS lesion offers hypoxia-like conditions, oxidative stress, and competition for nutrients, so precursor cells proliferate and migrate but fail to differentiate, leaving axons demyelinated. Microglia in the chronically demyelinated cortex and normal-appearing white matter of progressive patients arrive in a state the authors describe as metabolically paralysed, with mitochondrial impairment and lipid droplets accumulated from myelin debris, unable to support repair and actively maintaining a hostile environment. Neurons run into a parallel problem: impaired electron transport chain function and depletion of tricarboxylic acid cycle enzymes in neuronal mitochondria produce a persisting axonal ATP deficiency that begins in acute lesions and continues into chronic disease. Read together, these three accounts turn remyelination failure into an energy-supply problem rather than a purely immunological one, which is a different kind of target.

Rereading Existing Drugs, With the Evidence Grades Attached
The review reinterprets nine classes of MS therapy through a metabolic lens, and its most useful feature is that it grades each reinterpretation by evidence certainty rather than presenting them as equally settled. Four are graded direct. Teriflunomide inhibits dihydroorotate dehydrogenase and blocks de novo pyrimidine synthesis, which starves rapidly dividing lymphocytes while resting lymphocytes fall back on salvage pathways. Dimethyl fumarate modifies Keap1 to stabilise Nrf2, driving glutathione and NADPH synthesis and shifting myeloid cells away from a glycolytic phenotype. Sphingosine-1-phosphate receptor modulators act on astrocytes, neurons, and oligodendrocytes well beyond lymph node retention. BTK inhibitors curtail B-cell mitochondrial respiration but not glycolysis, reducing the GM-CSF to IL-10 ratio without depleting B cells at all, an effect Phase 1 work in healthy volunteers has confirmed in vivo. Interferon-beta is graded indirect. The remaining four, glatiramer acetate, natalizumab, anti-CD20 antibodies, and cladribine, are graded as hypothesis, with the authors noting for natalizumab that the metabolic priming account of rebound disease is speculative and that alternative explanations remain credible. Four direct, one indirect, four hypothetical is the honest measure of how much of this reframing currently stands on evidence.

The Framework Qualifies Itself, and the Pipeline Is Early
Having built its argument on the Th17-versus-Treg and M1-versus-M2 dichotomies, the review then states that both are oversimplified. The M1/M2 scheme comes from in vitro work and does not survive contact with in vivo data: single-cell RNA sequencing of MS lesions identifies disease-associated microglia, lipid-associated macrophages, and interferon-responsive cells, each with its own metabolic signature, and myeloid cells in lesions often express classical and alternative markers simultaneously. Pathogenic Th17 cells in MS show a hybrid state using both glycolysis and oxidative phosphorylation. The practical consequence the authors draw is that any strategy aimed at locking cells into one metabolic state may be defeated by adaptive rewiring. The investigational pipeline reflects how early this is. Glutaminase inhibition with CB-839, borrowed from oncology, impairs Th17 differentiation in the animal model of MS. The pyruvate kinase M2 activator TEPP-46 works by disrupting the tetrameric structure the enzyme needs for glycolysis, pushing cells towards oxidative metabolism. ACC1 inhibitors such as ND-646 block the phospholipids Th17 cells require. Metformin crosses the blood-brain barrier and augments precursor cell differentiation, with trials underway. Cardiac glycosides inhibit HIF-1α effectively in the animal model but are not translatable because of toxicity, and 2-deoxyglucose is too blunt for systemic use. Almost all of this remains preclinical.

Disease Stage Reverses the Direction of Treatment
The hardest problem the review identifies is that the same axis demands opposite interventions depending on when you intervene. A glycolysis inhibitor that suppresses inflammation in relapsing-remitting disease may be inappropriate or harmful in progressive disease, where the goal is to enhance mitochondrial function in neurons and precursor cells rather than restrict it. Two further obstacles sit alongside it. Glycolysis and oxidative phosphorylation are required by healthy cells as well as pathogenic ones, so any usable drug needs a vulnerability specific to the pathogenic state, and the review nominates glutamine dependency in Th17 cells and particular lipid desaturation requirements in inflammatory microglia as candidates. Immune cells are also metabolically adaptable enough to reroute around a blocked pathway, which is a route to resistance. The authors are candid about what their own paper cannot settle: this is a narrative review with no formal quality assessment or quantitative synthesis, its selection reflects the authors' judgment, non-English articles and conference abstracts were excluded, much of the mechanism derives from an animal model that reproduces progressive human MS poorly, the therapeutic proposals are largely speculative, and the paper does not address the adverse effects of metabolic modulation at all. Their own translational priority list puts the sensible next step first: validate biomarkers for stratifying patients, including serum neurofilament light, serum GFAP, and FDG-PET, where pilot data suggest higher glycolysis in secondary progressive than in relapsing-remitting patients, before running proof-of-concept trials of repurposed metabolic combinations.

Disclaimer: This blog post is based on the cited review article and is intended for informational purposes only. It is not intended to provide medical advice. Please consult with a healthcare professional for any health concerns.

Reference:
Badawi, G. A., El-Sayed, R. M., & Fawzy, M. N. (2026). Immunometabolic reprogramming in multiple sclerosis: from pathogenic amplifier to therapeutic target in neuroinflammation and remyelination. Inflammopharmacology. https://doi.org/10.1007/s10787-026-02350-y