Molecular Rheumatology Department, Trinity College Dublin Dublin, Dublin, Ireland
Disclosure information not submitted.
Órla Tynan1, Mary Canavan2, Achilleas Floudas3, Conor Smith4, Aoife O' Rourke4, Dumitru Anton5, Carl Orr6, Douglas Veale7 and Ursula Fearon8, 1Molecular Rheumatology Department, Trinity Biomedical Sciences Institute, Trinity College Dublin, Ireland, EULAR Centre for Arthritis and Rheumatic Diseases, St Vincent University Hospital, University College Dublin, Dublin, Ireland, 2Molecular Rheumatology Department, Trinity Biomedical Sciences Institute, Trinity College Dublin, EULAR Centre for Arthritis and Rheumatic Diseases, St Vincent University Hospital, University College Dublin, School of Biochemistry & Immunology, Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin, Ireland, 3Dublin City University, Dublin, Ireland, 4Translational Immunology, Trinity Biomedical Sciences Institute, Dublin, Ireland, 5Molecular Rheumatology Department, Trinity Biomedical Sciences Institute, Trinity College Dublin, EULAR Centre for Arthritis and Rheumatic Diseases, St Vincent University Hospital, University College Dublin, Dublin, Ireland, 6EULAR Centre for Arthritis and Rheumatic Diseases, St Vincent University Hospital, University College Dublin, Dublin, Ireland, 7St.Vincent's University Hosp, Dublin, Ireland, 8Trinity College Dublin, Dublin, Ireland
Background/Purpose: Synovial fibroblasts (FLS) are key contributors to joint inflammation and damage in patients with Rheumatoid (RA) and Psoriatic Arthritis (PsA). Recent studies have identified FLS subsets with distinct pro-inflammatory roles in RA. However, there is a scarcity of data regarding the contribution of FLS in PsA pathogenesis and overall lack of unifying nomenclature. Therefore, the aim is to identify the phenotypic and functional characteristics that define distinct FLS populations and function in RA vs PsA, with implication for disease pathogenesis and therapeutic response.
Methods: Single cell (Sc) RNAseq was performed on 88,953 RA and PsA FLS from intact synovial biopsies and FLS populations were defined by advanced bioinformatic analysis. Subsequently, multiparametric flow cytometric analysis (22 markers) was performed on RA and PsA patient synovial biopsies to examine FLS phenotype and function. Further characterization of differences in the invasive activity of RA and PsA FLS was conducted ex vivo through quantification of matrix metalloproteinases using MSD multiplex-assays/RT-PCR, whilst metabolism was assessed by Seahorse-XFe-technology. Additionally, flow cytometric analysis of key FLS activation/functional markers was performed on RA and PsA FLS across passages 0-3 (P0-3) to define phenotypic alterations once removed from the joint microenvironment.
Results: ScRNAseq analysis demonstrated 11 distinct FLS populations in RA and PsA, with differential frequency of clusters observed with THY1+ FLS dominant in RA vs THY1- FLS dominant in PsA. Flow analysis of PDPN+ FLS demonstrated significant increases in HLADR+, YAP+, Cad11+, and pS6+ FLS in RA (all p< 0.05), whilst PsA FLS demonstrated a significant increase in CD55 expression (p=0.0079). Further flow analysis identified 6 FLS populations that could be matched to 6 main populations in the scRNAseq. When compared directly between diseases, patients with RA displayed significant enrichment in THY1+CD34-CD55-FAP+ and THY1+CD34+CD55-FAP+ FLS (p=0.0093), while patients with PsA displayed enrichment in THY1+CD34-CD55+FAP+(p=0.02) and THY1-CD34-CD55+FAP+ FLS (p=0.0013). HLADR and Cad11 were significantly higher in RA subpopulations, compared to increased metabolic markers in PsA subpopulations by flow cytometry. In parallel, single-cell analysis of these populations demonstrated immune/inflammatory responses in RA dominant populations in contrast to matrix degrading and metabolic markers in PsA populations. Expanded RA and PsA FLS (P0) confirmed these differences in matrix degrading/metabolic pathways, matching the single cell/flow cytometric analysis. While P0 FLS maintained similar phenotypic profiles to ex vivo FLS, once expanded to P3, FLS started to lose specific phenotypic characteristics. Specifically, the expression of HLADR and CD55 reduced across passages 0-3, while CD34 and CD54 increased, supporting their transient nature.
Conclusion: Distinct FLS populations with unique functional properties were identified in RA and PsA. However, once removed from the joint microenvironment, FLS subset stability appears transient with convergence towards common phenotypes.
Ó. Tynan: None; M. Canavan: None; A. Floudas: None; C. Smith: None; A. O' Rourke: None; D. Anton: None; C. Orr: None; D. Veale: None; U. Fearon: None.