Decoding Multiple Sclerosis Through the Metabolome

Multiple sclerosis (MS) is a immune-mediated disorder characterized by inflammation within the central nervous system, demyelination, impaired remyelination, axonal injury, and progressive neurodegeneration. Although disease-modifying therapies can reduce inflammatory activity and may limit long-term disability, treatment initiation frequently depends on a diagnostic process that combines clinical manifestations with magnetic resonance imaging and, in selected cases, cerebrospinal fluid analysis. This process may be prolonged because early neurological symptoms are often nonspecific and radiological evidence must demonstrate lesion dissemination in time or space. Consequently, there is considerable interest in developing minimally invasive blood-based biomarkers that could complement existing diagnostic procedures. In the study “Metabolome-based signature of disease pathology in MS,” Andersen and colleagues investigated whether serum metabolite profiles could distinguish individuals with MS from unaffected controls. Their approach was particularly notable because it combined broad metabolomic profiling with transcriptomic and genetic analyses, thereby moving beyond the identification of isolated biomarkers toward a systems-level interpretation of MS pathophysiology.

A Multi-Omic Experimental Strategy
The study included 12 male patients with MS and 13 male controls recruited through the MURDOCK Study. All participants were non-Hispanic White, nonsmokers, and frequency-matched for age and body mass index. Patients had received no disease-modifying treatment for at least three months before biospecimen collection, reducing the likelihood that the observed metabolic differences were direct pharmacological effects. Serum samples were analysed using two complementary platforms. Untargeted two-dimensional gas chromatography coupled with time-of-flight mass spectrometry provided broad detection of chemically diverse compounds, whereas the targeted Biocrates AbsoluteIDQ p150 platform quantified predefined amino acids, acylcarnitines, hexoses, phosphatidylcholines, and sphingolipids. Of 400 initially measured metabolic variables, 325 met quality-control criteria. The investigators additionally generated whole-blood gene-expression data for 24 participants and assessed established MS-associated genetic variants, including the major HLA risk allele HLA-DRB115:01 and 175 non-major-histocompatibility-complex single-nucleotide polymorphisms. This design enabled the researchers to examine not only whether metabolites differed between groups, but also how those differences might relate to immune regulation, mitochondrial biology, membrane metabolism, and inherited susceptibility.

Machine Learning Identifies Six Candidate Metabolites
A random forest classifier containing 5,000 trees was used to rank metabolites according to their importance for distinguishing MS cases from controls. Twelve metabolites emerged as informative, and subsequent logistic-regression and receiver-operating-characteristic analyses identified six candidates with areas under the curve greater than 0.80: pyroglutamate, laurate, tetradecenoyl-L-carnitine or acylcarnitine C14:1, N-methylmaleimide, phosphatidylcholine PC ae C40:5, and phosphatidylcholine PC ae C42:5. Each was present at a higher level in the MS group. The box plots presented on page 5 of the article visually demonstrate the separation between cases and controls, although individual values overlap and several distributions contain outliers. Laurate produced the highest reported area under the curve, approximately 0.86, followed closely by pyroglutamate at 0.85; the remaining four metabolites produced values of approximately 0.81–0.82. These results indicate promising discriminatory potential within the study population. However, they should not be interpreted as evidence of clinical diagnostic performance because the same small dataset was used for biomarker discovery and evaluation, and no independent replication cohort was included.

Oxidative Stress, Fatty Acids, and Immune Activation
Pyroglutamate and laurate provide biologically plausible links between systemic metabolism and MS pathology. Pyroglutamate, also known as 5-oxoproline, is an intermediate of the γ-glutamyl cycle and is closely connected to glutathione metabolism. Elevated pyroglutamate may therefore indicate disruption of antioxidant homeostasis and increased oxidative stress, both of which are relevant to mitochondrial injury, oligodendrocyte damage, and neurodegeneration in MS. Gene-expression pathways associated with pyroglutamate included glutathione biosynthesis, iron-homeostasis signalling, ceramide degradation, and sphingosine metabolism. These pathways connect redox imbalance with lipid-mediated apoptosis and myelin biology. Laurate, or lauric acid, is a saturated medium-chain fatty acid with potential immunomodulatory activity. Experimental studies cited by the authors suggest that laurate can promote pro-inflammatory T-helper 1 and T-helper 17 differentiation while suppressing regulatory T-cell development. Consistent with this interpretation, laurate-associated expression patterns were enriched for inflammasome, SAPK/JNK, ephrin, interleukin-8, and lymphotoxin-related signalling. The finding does not establish that dietary laurate causes MS; serum concentrations may reflect endogenous metabolism, diet, intestinal processes, or altered immune-cell energetics. Nevertheless, it supports the broader hypothesis that lipid availability can influence inflammatory polarization.

Membrane Lipids and Mitochondrial Energy Metabolism
Two ether-linked phosphatidylcholines, PC ae C40:5 and PC ae C42:5, were elevated in patients with MS. Phosphatidylcholines are major constituents of cellular membranes and myelin, and changes in their circulating abundance may reflect membrane remodelling, altered lipoprotein transport, phospholipase activity, inflammatory signalling, or tissue injury. The authors identified associations between both phosphatidylcholines and the expression of PLA2G4C, which encodes a phospholipase A2 family enzyme capable of generating bioactive lipid products. Genetic variants near ETS1, IL2RA, and AFF1 were also associated with both metabolites, suggesting possible connections between immune-regulatory loci and lipid metabolism. PC ae C40:5 was especially associated with oxidative phosphorylation, mitochondrial dysfunction, sirtuin signalling, death-receptor signalling, and calcium-induced T-cell apoptosis. Acylcarnitine C14:1 provided a complementary signal. Acylcarnitines are intermediates in mitochondrial fatty-acid transport and β-oxidation; their accumulation can indicate incomplete fatty-acid oxidation or altered mitochondrial substrate use. Expression patterns associated with C14:1 were strongly enriched for antigen-presentation pathways and multiple class II HLA genes. Furthermore, carriage of HLA-DRB115:01 was significantly associated with C14:1 levels, linking the principal genetic risk locus for MS to a measurable feature of energy metabolism.

N-Methylmaleimide and Convergent Pathways of Cellular Injury
N-methylmaleimide was the least biologically characterized of the six metabolites, yet its associations generated several mechanistically relevant hypotheses. The compound is an electrophilic thiol-reactive molecule and can act as an agonist of the transient receptor potential ankyrin 1 channel, or TRPA1. TRPA1 is expressed in sensory neurons, astrocytes, and other cell types and participates in calcium signalling, nociception, neurogenic inflammation, and responses to oxidative or electrophilic stress. Experimental demyelination models cited in the article suggest that reduced TRPA1 activity may protect oligodendrocytes from apoptosis. In the present study, N-methylmaleimide-associated genes were enriched for oxidative phosphorylation, mitochondrial dysfunction, sirtuin signalling, apoptosis, and the mevalonate pathway. When the pathway results for all six metabolites were considered together, several recurrent biological themes emerged: altered mitochondrial function, disrupted energy metabolism, oxidative stress, death-receptor signalling, and immune-cell apoptosis. Table 2 on page 6 illustrates these convergent associations, including oxidative-phosphorylation genes linked to PC ae C40:5 and N-methylmaleimide and antigen-presentation genes linked to C14:1. These findings imply that the serum signature may capture interconnected processes rather than six independent abnormalities.

Scientific Significance, Limitations, and Future Directions
The principal strength of this study is its integrative framework. By combining untargeted and targeted metabolomics with gene-expression profiling and genotyping, the investigators produced a multidimensional model of metabolic disruption in MS. The careful restriction of the cohort by sex, ancestry, smoking status, treatment exposure, age, and body mass index reduced several sources of confounding. However, the same homogeneity substantially limits generalizability. The sample consisted of only 25 participants, all of whom were male and non-Hispanic White, while MS disproportionately affects women and occurs across diverse populations. The cross-sectional design also prevents determination of whether the metabolites contribute to disease development, arise as consequences of established pathology, or reflect unmeasured dietary and environmental factors. Statistical associations with genetic variants and gene-expression pathways were exploratory, and the large number of comparisons relative to the available sample creates a substantial risk of false-positive findings. Accordingly, the six metabolites should be regarded as candidate markers rather than validated diagnostic biomarkers. Future studies should evaluate them in larger, independent, prospective cohorts that include women, diverse ancestries, different MS phenotypes, clinically isolated syndrome, neurological disease controls, and longitudinal measurements. Validation should also determine whether a combined multimetabolite panel provides information beyond magnetic resonance imaging, cerebrospinal fluid markers, neurofilament light chain, and conventional clinical variables.

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.