A Gene, an Exposure, and an Intervention in MS Metabolism

Metabolism in multiple sclerosis (MS) gets studied in three separate ways that rarely cite each other. One line asks what an inherited gene product does to the cells that drive the disease. A second asks what an environmental exposure does to the metabolites circulating in patients. A third asks whether a behaviour can move those metabolites back. Four recent papers, one from each of those lines plus a review that tries to put them on common ground, show how far apart the approaches are in method and how close two of them turn out to be in subject matter.

A Gene Product That Sets How Glia Activate
Shen and colleagues work on microproteins, short polypeptides that standard annotation tends to miss. In a screen for microproteins that modify inflammatory outcomes they had already found MOCCI, the product of C15orf48 and its mouse counterpart AA467197, as a regulator of gut and lung inflammation. Whether it did anything in the brain was unknown. It does. MOCCI is upregulated in microglia and astrocytes in both mouse and human brain when inflammation occurs, and it is required for what the authors call proper, complete and beneficial activation of those cells. Inducing it pushes glia into a neuroprotective state and promotes resolution of inflammation, while cells deficient in it show reduced migration, reduced phagocytosis and reduced cytokine secretion. In the cuprizone mouse model of MS, MOCCI has a role in both demyelination and remyelination. The authors describe it as a molecular brake on neuroinflammation, which puts a mitochondrial gene product in charge of whether glial activation ends well or badly.

An Exposure Acting Through Host and Microbiome Together
Smokers with MS have worse disease, and Briggs and colleagues went looking for the chemistry in between. They measured serum bile acids, tryptophan metabolites and tobacco-related metabolites across four independent MS cohorts totalling 266 people, in a design that required each finding to appear in discovery and then again in a separate set before it counted. Mixed-effects regression tested associations with current smoking and with nicotine exposure, and mediation analysis then asked whether any surviving metabolite sits on the path from smoking to severity. Current smokers and nicotine-exposed patients showed reductions in both bile acids and tryptophan metabolites, with indolepropionate standing out, a neuroprotective and anti-inflammatory metabolite that gut microbes produce. Lower indolepropionate statistically mediated roughly 20% of smoking's adverse effect on MS severity. Metagenomic analysis, which the authors label hypothesis-generating rather than confirmatory, identified smoking-enriched taxa with MS links and found indolepropionate co-occurring broadly with microbial networks.

An Intervention Moving the Same Chemistry the Other Way
Kupjetz and colleagues pooled four randomized controlled trials, 273 participants at a mean age of 44.5 years and 67% female, to ask whether exercise shifts the kynurenine pathway. Their earlier work had reduced serum kynurenines to two latent factors, one reflecting a neurotoxic pattern and one a neuroprotective pattern, and the first thing they checked here was whether those factors hold up. Repeating the exploratory factor analysis in each subcohort gave similar loadings, so the constructs are stable across samples rather than artifacts of one dataset. Then the interventions. A single session of combined strength-focused exercise produced an acute rise in the neuroprotective pattern of 0.421 (95% CI 0.216 to 0.626, p < 0.001). Twenty-four weeks of endurance exercise produced a chronic fall in the neurotoxic pattern of −0.432 (95% CI −0.786 to −0.080, p = 0.019). The two effects came from different trials and different exercise modalities, and the authors are careful that the determinants and clinical relevance of these changes still need work.

Two of the Three Land on the Same Pathway
The smoking study and the exercise study were designed independently, in different countries, for different questions, and both ended up measuring tryptophan metabolism. One found a harmful exposure depleting tryptophan metabolites in a way that accounts for a fifth of its effect on severity. The other found a modifiable behaviour pushing the kynurenine arm of the same pathway toward its neuroprotective pattern. Neither paper claims the other's result, and the metabolites they measure are not identical, since indolepropionate comes from the microbial indole branch while kynurenines come from the host enzymatic branch. But both sit downstream of the same amino acid, and they move in opposite directions for opposite reasons. Whether that is one mechanism seen twice or two unrelated results in the same chemistry is not something either paper can answer.

What the Review Proposes as Common Ground
Mosaffa Jahromi and colleagues set out the framework these results would have to fit, in a review covering MS alongside rheumatoid arthritis and type 1 diabetes. Their argument is that these are not only disorders of immune regulation but of metabolic reprogramming in lymphocytes, where the balance between oxidative phosphorylation and aerobic glycolysis, increased fatty acid synthesis, and disturbed glutamine metabolism shape how T and B cells activate, differentiate and become pathogenic. They note that the same signature of raised glycolysis, impaired mitochondrial function and mTOR signalling appears in systemic lupus erythematosus too, which positions immunometabolism as a shared mechanism across autoimmunity rather than a disease-specific one. They also name the obstacle that keeps this from becoming therapy, which is targeting the relevant cells without systemic toxicity, and single out epigenetic-metabolic crosstalk as the part that needs understanding before any of it translates.

What Each Design Can and Cannot Settle
The four papers fail at four different points, and knowing where is most of the value in reading them together. A microprotein knockout establishes necessity in a mouse, but cuprizone is a demyelination model rather than an autoimmune one, so MOCCI's role in human MS is inferred from expression rather than demonstrated. Mediation analysis is arithmetic performed on observational data; the 20% figure describes how much of an association travels through a measured variable, not how much of a causal effect would disappear if indolepropionate were restored. Pooled randomized trials deliver the strongest causal claim of the four, since the exercise was assigned rather than chosen, but two significant results drawn from four trials with different modalities and durations is a thin basis for a dose. And a review organizes what is known without adding evidence, which is why its most useful contribution here is naming the gap rather than closing it.

Disclaimer: This blog post is based on the cited studies 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.

References:
Shen, K., Guo, K., Leong, S. H., Lim, R., Chew, W. L., Lee, C. Q. E., & Ho, L. (2026). Mitochondrial microprotein MOCCI controls neuroinflammation by altering glial activation states. Frontiers in Immunology, 17, 1798798. https://doi.org/10.3389/fimmu.2026.1798798
Briggs, F. B. S., Litwiler, J., Montini, F., et al., & Tobin, W. O. (2026). Tobacco smoking disrupts bile acid and tryptophan metabolism in multiple sclerosis. Multiple Sclerosis Journal, 32(9), 972–984. https://doi.org/10.1177/13524585261454207
Kupjetz, M., Langeskov-Christensen, M., Riemenschneider, M., et al., & Zimmer, P. (2026). Exercise-induced modulation of kynurenine pathway metabolite patterns in multiple sclerosis: a targeted metabolomics analysis across four randomized controlled trials. Brain, Behavior, and Immunity, 138, 106930. https://doi.org/10.1016/j.bbi.2026.106930
Mosaffa Jahromi, A., Mirzaei, F., Mirzazadeh, S., Meri, S., Kabelitz, D., & Kalantar, K. (2026). Metabolic switch: lymphocytes and the onset of T-cell-mediated autoimmunity. Cellular & Molecular Immunology, 23(10), 1171–1193. https://doi.org/10.1038/s41423-026-01460-5