Multi-omic Integration of Serum Proteins and Metabolites Reveals Distinct Molecular Signatures Associated With Clinical, Patient-reported, and Imaging Measures in Multiple Sclerosis
Fatemeh Siavoshi1, Matthew Smith1, blake Dewey1, Keenan Walker2, Elias Sotirchos1, Shiv Saidha1, Kathryn Fitzgerald1, Ellen Mowry1, Peter Calabresi1, Pavan Bhargava1
1Johns Hopkins University, 2National Institutes of Health
Objective:
To identify shared blood-based molecular signatures linking serum metabolomics and proteomics with clinical disability, patient-reported outcomes, and neuroimaging markers in multiple sclerosis (MS), providing mechanistic insight into pathways underlying disease severity.
Background:
Identifying molecular factors that drive MS severity remains a major challenge. Although multi-omic profiling offers a powerful framework to study disease mechanisms, large-scale integrative analyses in MS remain limited. Leveraging comprehensive serum proteomic and metabolomic datasets enables unbiased identification of molecular signatures associated with disease severity and provides insight into processes contributing to MS progression.
Design/Methods:
In this cross-sectional study, serum samples from 187 people with MS were profiled using the SomaScan 7k proteomic and Metabolon global untargeted metabolomic platforms (7,000 proteins and 1639 metabolites). Assessments included clinical disability metrics (Expanded Disability Status Scale, Nine-Hole Peg Test, Timed 25-Foot Walk, and Processing Speed Test), patient-reported outcomes derived from Neuro-QoL domains, and imaging measures from magnetic resonance imaging volumetrics and optical coherence tomography. After preprocessing, omic layers were grouped using two complementary integrative approaches, Multi-Omics Factor Analysis (MOFA) and Hierarchical All-against-All Association (HAllA). Associations between molecular groups and outcome measures were evaluated using multivariable linear regression, adjusting for age, sex, and race.
Results:
Two distinct molecular groups were identified. The first group, associated with imaging outcomes, comprised metabolites related to mitochondrial and membrane metabolism (DHEA-S, phosphoethanolamine, taurine, sphingomyelins) and proteins involved in immune signaling, receptor activity, and axonal guidance, including TNFRSF1A, FZD2, FZD7, EPHA2, EFNA5, CD46, TIMP1, and SPON2. The second group, associated with clinical and patient-reported outcomes, was characterized by carnitine and lipid derivatives (palmitoleoylcarnitine [C16:1], myristoylcarnitine [C14], sphingomyelins) together with proteins linked to ubiquitination, proteostasis, oxidative stress, and cytoskeletal regulation, including HSPB1, HSPA5, PARK7, BAG3, UBE2L6, ATXN3, GLO1, and PRDX1.
Conclusions:
Using multi-omic analysis, we identified distinct molecular groups and pathways associated with multiple MS severity metrics in a large cohort.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.