Osteoclast <scp>microRNA</scp> Profiling in Rheumatoid Arthritis to Capture the Erosive Factor

In rheumatoid arthritis (RA), only a subset of patients develop irreversible bone destruction. Our aim was to identify microRNA (miR)-based osteoclast-related signature predictive erosiveness in RA. Seventy-six adults with erosive (E) or nonerosive (NE) seropositive RA and 43 sex- age-matched healthy controls were recruited. Twenty-five miRs from peripheral blood mononuclear cell (PBMC)-derived osteoclasts selected RNA-Seq (discovery cohort) assessed by qPCR (replication cohort), as 33 target genes (direct targets associated regulated pathways). The top five found differentially expressed either decreased (hsa-miR-34a-3p, 365b-3p, 374a-3p, 511-3p [E versus NE]) increased (hsa-miR-193b-3p controls]). vitro, inhibition miR-34a-3p had an impact on osteoclast resorption. An integrative network analysis their highlighted correlations between mRNA miR expression, both negative (CD38, CD80, SIRT1) positive (MITF), differential gene expression NE E (GXYLT1, MITF) KLF4). Machine-learning models used evaluate the value genes, combination clinical data, predict erosion. One model, including set (predominantly 365b-3p) combined factor titer, provided 70% accuracy (area under curve [AUC] 0.66). Adding directly targeted belonging related pathways improved power model for phenotype (78% accuracy, AUC 0.85). This proof-of-concept study indicates that identification subjects at risk erosions may be studying PBMC-derived osteoclasts, suggesting novel approaches toward personalized treatment. © 2023 Authors. JBMR Plus published Wiley Periodicals LLC behalf American Society Bone Mineral Research.