Metabolic Signatures of Multiple Sclerosis: What Multi-Omics Reveals About Disease Biology
Multiple sclerosis (MS) is a chronic inflammatory and neurodegenerative disorder of the central nervous system characterized by immune-mediated damage, demyelination, failure of effective remyelination, and progressive neurological impairment. One persistent clinical challenge is that MS diagnosis depends heavily on neurological manifestations together with radiological evidence of lesions disseminated in space or time. Because many early symptoms are nonspecific, diagnosis and initiation of disease-modifying therapy may be delayed, despite evidence that earlier treatment can reduce long-term disability. Andersen and colleagues therefore investigated whether measurable changes in circulating metabolites could provide biological information relevant to MS diagnosis and pathogenesis. Their study, Metabolome-based signature of disease pathology in MS, specifically examined serum rather than cerebrospinal fluid, because blood collection is substantially less invasive and more suitable for repeated clinical assessment. The authors sought not merely to identify isolated metabolic abnormalities, but to combine metabolomics with gene-expression and genetic data to obtain a broader molecular view of MS pathology.
A Comprehensive Metabolomic and Multi-Omic Strategy
The investigation involved 25 participants: 12 men with MS and 13 controls. All participants were non-Hispanic white, non-smoking males, and the controls were frequency matched to the MS cases according to age and body mass index. Importantly, the individuals with MS had received no disease-modifying therapy for at least three months before biospecimen collection, reducing the likelihood that treatment effects would dominate the observed metabolic profile. Serum metabolites were examined using both untargeted two-dimensional gas chromatography coupled with time-of-flight mass spectrometry and a targeted platform measuring amino acids, acylcarnitines, hexoses, phospholipids, and sphingolipids. Of approximately 400 initially captured metabolic variables, 325 passed quality-control criteria. The researchers then used random-forest machine learning to identify metabolites informative for MS classification, followed by logistic regression and receiver operating characteristic analysis to estimate their discriminatory performance. Gene-expression measurements and MS-associated genetic variants were subsequently integrated with these metabolic findings.
Six Metabolites Emerge as the Strongest Candidates
Random-forest analysis initially identified 12 metabolites as informative for distinguishing participants with MS from controls. Eight were significantly associated with MS in logistic regression, while six achieved an area under the receiver operating characteristic curve greater than 80%, the threshold established by the investigators for their leading candidates. These six metabolites were pyroglutamate, laurate, tetradecenoyl-L-carnitine or acylcarnitine C14:1, N-methylmaleimide, and two phosphatidylcholines designated PC ae C40:5 and PC ae C42:5. Their individual AUC values ranged from approximately 0.81 to 0.86, and all six were present at higher levels in MS cases than in controls. The box plots presented in Figure 1 of the article visually demonstrate these group differences, although some overlap between cases and controls remains evident. The metabolic classes represented are biologically diverse, involving glutathione metabolism, fatty-acid metabolism and oxidation, membrane lipid composition, and electrophilic or transient-receptor-potential-related signaling. These findings therefore suggest that MS-associated metabolic disturbance may involve several interconnected cellular systems rather than a single biochemical pathway.
Oxidative Stress, Fatty Acids, and Membrane Biology
The individual metabolites provide important clues regarding potential mechanisms of disease. Pyroglutamate, the highest-ranked metabolite in the random-forest analysis, is an intermediate related to glutathione metabolism and may therefore reflect altered antioxidant regulation and oxidative stress. The authors note that elevated pyroglutamate has previously been observed in metabolomic studies of MS and could potentially be both a marker and contributor to oxidative injury. Laurate, a medium-chain saturated fatty acid, is particularly noteworthy because experimental studies cited by the authors suggest that it can promote differentiation of pro-inflammatory Th1 and Th17 cells while reducing regulatory T-cell differentiation. The two phosphatidylcholines are also biologically relevant because phosphatidylcholines are fundamental constituents of cellular membranes and myelin. Their metabolism can generate inflammatory lipid mediators and lysophosphatidylcholines capable of damaging myelin. Thus, elevations in these compounds should not be interpreted simply as diagnostic signals; they may reflect altered immune activation, membrane turnover, lipid metabolism, or tissue injury occurring during MS. This distinction is essential because the study examined established disease rather than individuals before MS onset.
Mitochondrial Dysfunction and Immune Metabolism
Acylcarnitine C14:1 provides an especially interesting connection between cellular energetics and immune function. Acylcarnitines are intermediates of mitochondrial fatty-acid oxidation, making changes in their concentrations potentially informative about energy metabolism and mitochondrial efficiency. In this study, gene-expression patterns associated with C14:1 were strongly enriched for antigen-presentation pathways and included numerous class II HLA genes. The authors propose that these observations may reflect metabolic reprogramming during immune activation: activated CD4+ T cells can shift their energy production away from fatty-acid oxidation toward glycolysis. Pathway analyses across the six candidate metabolites further revealed recurring enrichment of mitochondrial dysfunction, oxidative phosphorylation, sirtuin signaling, and apoptosis-related pathways. PC ae C40:5 and N-methylmaleimide were particularly associated with signatures of mitochondrial dysfunction, while pyroglutamate-linked genes were enriched in iron homeostasis and sphingolipid-related processes. Overall, approximately one third of metabolite-associated enriched pathways overlapped with pathways connected to recognized MS genetic risk, strengthening the possibility that the metabolic abnormalities reflect biologically meaningful disease processes rather than isolated biochemical variation.
Connecting Metabolites with Genetic Susceptibility
A major strength of the investigation is its attempt to relate circulating metabolites to inherited MS susceptibility. The researchers examined 175 putative non-MHC MS risk variants after genetic quality control, as well as the HLA-DRB115:01 and HLA-A02 alleles. HLA-DRB115:01, one of the most important established genetic susceptibility factors for MS, was significantly associated with acylcarnitine C14:1, whereas no corresponding association was detected for HLA-A02. Moreover, multiple non-MHC variants showed associations with the six candidate metabolites. Particularly notable were variants related to ETS1, IL2RA, and AFF1 that were associated with both phosphatidylcholines. The C14:1 finding was accompanied by associations with expression of several class II HLA genes involved in antigen presentation, creating a potential molecular bridge between inherited immune susceptibility and altered energy metabolism. These relationships remain exploratory rather than proof of causality, but they illustrate the value of integrating genomic, transcriptomic, and metabolomic information when studying a multifactorial disease such as MS.
Scientific Significance, Limitations, and Future Directions
The study presents a compelling proof of concept for metabolomics-based investigation of MS, but its findings require cautious interpretation before any clinical application. The sample comprised only 25 participants and lacked an independent replication cohort, meaning that the reported discriminatory performance may not generalize to larger or more heterogeneous populations. The deliberate restriction to non-Hispanic white, non-smoking men helped reduce confounding but simultaneously limits conclusions concerning women, other racial and ethnic groups, smokers, different body compositions, and diverse MS subtypes. Furthermore, because the researchers studied prevalent rather than newly diagnosed or preclinical MS, the identified metabolites could represent consequences of ongoing inflammation, demyelination, or neurodegeneration rather than factors that precede disease development. The authors therefore call for larger, diverse studies, evaluation of associations with disease progression, and comparison with other inflammatory and neurological diseases to establish specificity. The principal contribution of this work is consequently not the establishment of a ready-to-use blood test, but the demonstration that integrated metabolomics can reveal a coherent biological signature linking oxidative stress, lipid metabolism, mitochondrial dysfunction, apoptosis, immune activation, and genetic susceptibility in MS.
Disclaimer: This blog post is based on the provided research 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.
References:
Andersen, S. L., Briggs, F. B. S., Winnike, J. H., Natanzon, Y., Maichle, S., Knagge, K. J., ... & Gregory, S. G. (2019). Metabolome-based signature of disease pathology in MS. Multiple sclerosis and related disorders, 31, 12-21.
