Background Deep profiling of synovial tissue samples from rheumatoid arthritis (RA) patients may reveal the molecular underpinnings of phases of RA progression and provide new therapeutic targets to intervene earlier in disease pathogenesis.
Objectives We sought to identify the molecular pathways expressed in different stages of disease (from seropositive subjects without clinically apparent synovitis to those with established disease) in synovial tissue compared to non-RA controls.
Methods Transcriptomics profiling was performed on RNA isolated from synovial tissue biopsies. Normal synovium was collected from subjects with knee pain and without diagnosis of OA or RA (n=28). Arthralgia tissue was collected from ACPA-positive subjects without synovitis (n=10). Early RA tissue was collected from patients recently diagnosed (<1 year) with RA (n=57). Established RA tissue was collected from ACPA-positive subjects with >1 year of disease duration (n=95). Protein expression was confirmed on infiltrating immune cells from synovial biopsy cell suspensions by flow cytometry in separate RA subjects.
Results Several pathways previously identified as important for RA pathogenesis (e.g., lymphocyte activation, osteoclast differentiation, NF-kappa B signaling) were enriched in differentially expressed genes in disease synovial biopsies compared to normal tissue samples. Interestingly, several genes known to function in T cell activation as signal 2 co-stimulatory or co-inhibitory molecules were differentially expressed, even in arthralgia and early RA subjects. 66 of 81 known co-stimulatory or co-inhibitory genes profiled were differentially expressed (FDR <5% and absolute fold-change >2) in disease samples from at least one cohort. The genes encoding co-stimulatory proteins that were increased compared to normal included CD28, CD40LG, CD40 and ICOS. Interestingly, some of the genes encoding co-inhibitory proteins were increased (PDCD1/PD-1, CD274/PD-L1, HAVCR2/TIM3, TIGIT, BTLA), whereas others showed decreased expression (C10orf54/VISTA and LAG3) compared to normal controls. We focused on CD28 expression, which is elevated in “pre-RA” arthralgia samples, proposing that anti-CD28 therapeutics could be candidates for RA disease prevention. By flow cytometry we demonstrated that a majority of CD4+ (>90%) and CD8+ (>60%) T cells from RA synovial biopsy cell suspensions (n=4) showed surface expression of CD28.
Conclusions We have generated a unique dataset from different phases of RA progression. Our results provide guidance for the selection of co-signaling molecules as therapeutic targets as well as for preventing the progression to RA. In particular, CD28 expression was elevated in synovial tissue biopsies before the development of RA and it was also expressed on the majority of synovial tissue infiltrating CD4+ and CD8+ T cells from RA patients. These observations provide rationale to target CD28 for both RA treatment and disease interception.
Disclosure of Interest A. Walsh Shareholder of: Johnson & Johnson, Employee of: Johnson & Johnson, M. Canavan: None declared, Y. Guo Shareholder of: Johnson & Johnson, Employee of: Johnson & Johnson, T. McGarry: None declared, X. Yin Shareholder of: Johnson & Johnson, Employee of: Johnson & Johnson, M. Wechalekar Grant/research support from: Johnson & Johnson, M. Smith Consultant for: Johnson & Johnson, S. Proudman: None declared, C. Orr: None declared, S. Kelly: None declared, C. Pitzalis Grant/research support from: Johnson & Johnson, Consultant for: Abbott, Astellas, MedImmune, BMS, Celgene, Grunenthal, GSK, MSD, Pfizer, Sanofi, Roche, UCB, D. Veale: None declared, U. Fearon: None declared, S. Nagpal Shareholder of: Johnson & Johnson, Employee of: Johnson & Johnson