0919: Proteomic Profiling of Extracellular Vesicles from Synovial Fluid of RA and OA Patients Reveals Cell-type-specific subpopulations and Disease Related Patterns

Background/Purpose: Extracellular vesicles (EVs) are secreted by virtually all cells and are known to carry bioactive cargo including proteins, RNA, DNA and metabolites. Analysis of the cargo of EVs derived from the synovium may provide insight into disease-specific molecular pathways of arthritis. So far, most studies characterizing EVs in disease were based on bulk analysis, neglecting the complexity of EV populations present in biofluids of patients. To investigate and characterize distinct EV subpopulations in synovial fluid of patients with rheumatoid arthritis (RA) and osteoarthritis (OA) we have established an novel immunocapture separation method based on magnetic nano-sized beads targeting cell type specific markers. CD90/THY1 was selected as a synovial fibroblast marker due to the reported contribution of CD90+ to synovitis.

Methods: EVs were isolated from synovial fluid of patients with RA and OA via size exclusion chromatography (SEC) and subsequently characterized by Nano Tracking Analyzer (NTA), NanoFlow cytometry (NanoFCM) and Western Blot (WB). For separation of a CD90+ EV population prototype anti-CD90 MicroBeads (Miltenyi Biotec) were used. WB and mass-spectrometry (MS)-based proteomics were applied to analyse the EV subpopulations. The proteomic profile of separated EV subpopulations was compared with published reference single cell RNA-Seq Data from inflamed synovial tissue. Disease specific proteomic profiles of EVs from RA and OA were compared.

Results: Using our immunocapture approach we were able to separate CD90+ EVs from synovial fluid of arthritis patients. WB confirmed a CD90 enrichment, while proteomic analysis revealed a stromal cell signature in the CD90+ EV-population, in comparison to markers for infiltrated immune cells like T cells, B cells, monocytes and neutrophils enriched in the CD90- EVs. The comparison of EV populations derived from RA and OA synovial fluid showed CD90+ EVs in both diseases. Analysis of RA-derived EVs showed a prominent neutrophil and macrophage signature, while OA enriched proteins were of fibroblast, macrophage, endothelial and epithelial cell profile. Proteomic analysis revealed several proteins enriched in RA known for their involvement in inflammation (TREM1, CXCR1, CXCR2, RETN, PADI4, LOXL3, CTSZ, MPO) or in cartilage degradation (MMP8). In contrast, in OA proteins involved in cartilage degradation and bone destruction were significantly enriched (ADAM10, FMOD, GPRC5a, INPP4N, LRP1, TNXB, SPARCL1) as well as protein candidates associated with OA pathogenesis and severity (CAPG, CTSA, LTBP1, LUM, LRP1).

Conclusion: Using a novel immunocapture approach we were able to separate cell type specific subsets of EVs from synovial fluid of arthritis patients. Proteomic analysis confirmed a stromal cell pattern in CD90+ EVs and revealed enrichment of disease specific proteins in EVs derived from RA and OA synovial fluid respectively. The ability to investigate distinct cell-type specific EV subsets in synovial fluid provides insight into pathogenetic mechanisms in the synovium and may lead to the identification of biomarkers in a liquid biopsy approach.