Background: The loss of self-tolerance plays a major role in the induction of rheumatoid arthritis (RA) leading to autoimmune-mediated joint inflammation and subsequent joint damage. Regulatory T cells (Tregs) maintain self-tolerance and limit excessive immune responses. However, Tregs show high levels of plasticity and heterogeneity, influencing functional activities. Identification and characterization of Treg subpopulations can elucidate their role in synovial inflammation.

Objectives: To identify Treg subpopulations in patients with RA and to examine their function in synovial inflammation.

Methods: CITE-seq was applied on PBMCs from four treatment naïve RA patients and four healthy controls, measuring the expression of 20 surface markers and gene expression on a single cell level. Unsupervised clustering was performed to identify Treg subpopulations. Further verification of these subpopulations was performed in PBMC from 150 RA patients and 30 healthy controls using flow cytometry. A suppression assay was performed to evaluate the suppressive capacity of the sorted Treg subpopulations. Finally, fibroblast interaction and cytokine-secreting potential of sorted Treg subpopulations were investigated using ELISA.

Results: CITE-seq data-driven analysis of T cells from treatment naïve early RA patients identified Treg subpopulations based on the presence of HLA-DR and/or CCR6. We verified the surface expression of HLA-DR and CCR6 on Tregs in PBMC of patients with established RA, revealing a significant decrease in HLA-DR ^- CCR6 ⁺ Treg cells in RA patients compared to healthy individuals. A significant decrease of these cells was found in RA patients with moderate and severe disease activity. It prompted us to investigate the functional differences between these Treg subpopulations.

First, we showed that all Treg subpopulations, including HLA-DR ^- CCR6 ⁺ Treg, are able to suppress CD4 ⁺ responder cells in the Treg suppression assay. Next, co-cultures with RA synovial fibroblasts were performed to mimic the synovial interaction between these cells. The HLA-DR ^- CCR6 ⁺ Treg subpopulation showed inflammatory interaction, indicated by the secretion of IL-6, IL-8 and MMP-3 after co-culture. Finally, we found that activated HLA-DR ^- CCR6 ⁺ Treg can produce IL-17A, which was increased upon co-culturing with RA synovial fibroblasts.

Conclusion: We identified four Treg subpopulations based on the surface expression of HLA-DR and CCR6 in early and established RA, of which the HLA-DR ^- CCR6 ⁺ Treg subpopulation was significantly decreased.

All subpopulations have a suppressive Treg phenotype, however, the HLA-DR ^- CCR6 ⁺ Treg subpopulation shows inflammatory interaction with RA synovial fibroblasts, resulting in increased secretion of inflammatory mediators. Together, these data show that the HLA-DR ^- CCR6 ⁺ Treg is a double-edged sword, having anti-inflammatory and pro-inflammatory characteristics.

REFERENCES: NIL.

Acknowledgements: NIL.

Disclosure of Interests: None declared.