Exploring micro-rna content of circulating extracellular vesicles as biomarkers in early rheumatoid arthritis

Abstract Background/Aims Early diagnosis and prompt and effective intervention for people with immune-mediated inflammatory diseases (IMIDs) such as rheumatoid arthritis (RA) improves outcomes. Non-invasive biomarkers that reflect unique aspects of underlying biology could facilitate this. Extracellular vesicles (EVs) released by many cell types are abundantly present in body fluids where they contribute to intercellular signalling by transferring cargo including microRNAs (miRs) - themselves implicated in disease pathogenesis. We investigated whether serum EV miR expression (i) discriminates patients with newly presenting, untreated RA from those with an alternative IMID or none, (ii) predicts therapeutic response to first-line methotrexate (MTX) treatment amongst RA patients, or, in the same individuals (iii) exhibit dynamically altered expression in MTX-responders versus non-responders during treatment. Methods Using a validated protocol, serum EVs were isolated from 43 early RA patients. Identically isolated EV were available from sera of 23 newly presenting, untreated polymyalgia rheumatica (PMR, “IMID control”) patients and 12 “non-inflammatory control” patients (NICs) in whom significant inflammatory disease had been excluded. miRs were isolated and subject to NanoString sequencing (798 miRs). Following quality control, DESeq2 was first used to identify miRs uniquely deregulated in RA EVs at baseline. Within the early RA cohort, a baseline comparison between individuals who subsequently achieved clinical remission after 6 months treatment with MTX (DAS28-CRP<2.4 without having required systemic steroids in preceding 2 months) and those that did not, was also undertaken by partial least squared-discriminant analysis (PLS-DA) with 5-fold cross validation. Linear modelling was used to assess differences in dynamic EV miR expression between responders and non-responders. Multiple test correction was applied (false discovery rate; FDR). Results A signature of 28 miRs uniquely differentially expressed between early RA patients and NICs (i.e. not also between PMR and NICs) was identified, which was robust to MTC after correction for age and sex. Amongst those upregulated in RA were miR-337-3p, whose predicted target proteins include Jak2 and STAT3, and miR-144-3p, recently shown to mediate the expression of pro-inflammatory cytokines and chemokines by macrophages. Whilst EV miRs differentially expressed at baseline between RA patients who subsequently achieved early MTX-induced remission and those who did not were not robust to MTC, PLS-DA indicated their predictive potential. Finally, several miRs whose change in expression over the course of treatment differed significantly between responders and non-responders; namely miR-548ar-3p, miR-212-3p, miR-188-5p, miR-410-3p, and miR-338-5p amongst others, may offer insight into the mechanism of MTX’s therapeutic action. Conclusion Serum EV miRs are a possible source of clinically valuable biomarkers that may shed light on pathophysiology and/or mechanisms of MTX efficacy. Our findings highlight their potential as diagnostic and prognostic tools in early RA. Larger studies and validation in independent cohorts are warranted to pursue this potential, alongside mechanistic evaluation. Disclosure D. Maunder: None. P.M. Brown: None. B. Barron-Millar: None. D. Lendrem: None. N. Naamane: None. J. Macdonald: None. X. Wang: None. A.E. Anderson: None. A.W. Morgan: None. S.L. Mackie: Consultancies; Roche/Chugai, Sanofi, AbbVie and AstraZeneca. Honoraria; Pfizer, Vifor and UCB. Grants/research support; Vifor. R. Crossland: None. A.G. Pratt: Consultancies; Inflection Biosciences.