Multi-omic Evaluation of Metabolic Alterations in Multiple Sclerosis Identifies Shifts in Aromatic Amino Acid Metabolism

ABSTRACT

The circulating metabolome is a product of interactions between the genome, epigenome, exposome and microbiome. The metabolome may be altered in people with multiple sclerosis (MS); however, existing metabolomics studies were relatively small or characterized a limited number of metabolites. Herein, we performed a multi-site study profiling the circulating metabolome to obtain relative abundances for 269 metabolites in a large cohort of MS patients and healthy controls. After adjusting for batch effects and extensive quality control, we created an overall metabolic dysfunction score, defined apriori sets of metabolites using known metabolic pathways, and derived novel networks of correlated metabolites using a weighted correlation network analysis (WGCNA). We assessed whether metabolic dysfunction, individual metabolites, metabolic pathways or WGCNA-identified module scores differed between people with MS versus healthy controls (HC) after adjusting for age, sex and race using generalized estimating equations (participants could provide multiple samples). In a subset of patients, information on disability status was also available. Similar models assessed the association between metabolites and metabolite sets with measures of disability. In people with MS, we identified striking abnormalities in a WGCNA-defined module enriched in aromatic amino acid (AAA) metabolites (FDR-adjusted p-value=2.77E-18) that are also strongly associated with disability (FDR-adjusted p-value for AAA module=1.01E-4). Consistent results were obtained using apriori-defined metabolite sets or in analyses of individual metabolites. The identified abnormalities likely relate to imbalances in gut microbial metabolism of AAAs resulting in reduced production of immunomodulatory metabolites and increased production of metabotoxins (indole acetate, phenylacetylglutamine, p-cresol sulfate, p-cresol glucuronide). Single cell RNA sequencing data analysis demonstrated altered AAA metabolism in CSF and blood derived monocyte cell populations, while treatment of human peripheral blood mononuclear cells with AAA-derived metabotoxins resulted in increased production of tumor necrosis factor-α. We identify novel metabolic alterations in people with MS potentially contributing to disease pathophysiology.