Plasma MicroRNAs in Established Rheumatoid Arthritis Relate to Adiposity and Altered Plasma and Skeletal Muscle Cytokine and Metabolic Profiles

Background: MicroRNAs have been implicated in the pathogenesis of rheumatoid arthritis (RA), obesity, and altered metabolism. Although RA is associated with both obesity and altered metabolism, expression of RA-related microRNA in the setting of these cardiometabolic comorbidities is unclear. Our objective was to determine relationships between six RA-related microRNAs and RA disease activity, inflammation, body composition, and metabolic function. Methods: Expression of plasma miR-21, miR-23b, miR-27a, miR-143, miR-146a, and miR-223 was measured in 48 persons with seropositive and/or erosive RA (mean DAS-28-ESR 3.0, SD 1.4) and 23 age-, sex-, and BMI-matched healthy controls. Disease activity in RA was assessed by DAS-28-ESR. Plasma cytokine concentrations were determined by ELISA. Body composition was assessed using CT scan to determine central and muscle adipose and thigh muscle tissue size and tissue density. Plasma and skeletal muscle acylcarnitine, amino acid, and organic acid metabolites were measured via mass-spectroscopy. Plasma lipoproteins were measured via nuclear magnetic resonance (NMR) spectroscopy. Spearman correlations were used to assess relationships for microRNA with inflammation and cardiometabolic measures. RA and control associations were compared using Fisher transformations. Results: Among RA subjects, plasma miR-143 was associated with plasma IL-6 and IL-8. No other RA microRNA was positively associated with disease activity or inflammatory markers. In RA, microRNA expression was associated with adiposity, both visceral adiposity (miR-146a, miR-21, miR-23b, and miR-27a) and thigh intra-muscular adiposity (miR-146a and miR-223). RA miR-146a was associated with greater concentrations of cardiometabolic risk markers (plasma short-chain dicarboxyl/hydroxyl acylcarnitines, triglycerides, large VLDL particles, and small HDL particles) and lower concentrations of muscle energy substrates (long-chain acylcarnitines and pyruvate). Despite RA and controls having similar microRNA levels, RA, and controls differed in magnitude and direction for several associations with cytokines and plasma and skeletal muscle metabolic intermediates. Conclusion: Most microRNAs thought to be associated with RA disease activity and inflammation were more reflective of RA adiposity and impaired metabolism. These associations show that microRNAs in RA may serve as an epigenetic link between RA inflammation and cardiometabolic comorbidities.