Rheumatoid synovial fluid interleukin-17-producing CD4 T cells have abundant tumor necrosis factor-alpha co-expression, but little interleukin-22 and interleukin-23R expression

Rheumatoid arthritis (RA) is a systemic chronic inflammatory disorder associated with persistent and destructive synovitis leading to cartilage and bone erosion. The underlying cause of RA is unknown; however, the pathogenesis of RA is thought to be the result of complex cell to cell interactions between amongst others, T cells, macrophages and fibroblasts. In established disease, the preponderance of IFNγ-expressing and paucity of IL-4-expressing T cells, in situ and ex vivo, had until recently led to the description of RA as an immune mediated inflammatory disease associated with a predominantly T helper type-1 (Th1)-like cytokine profile [1‐3].

More recently, effector T cells (Th17 cells), that produce interleukin-17A (IL-17) [4, 5] and that are functionally distinct from Th1 and Th2 helper T cells, have been identified in mice and subsequently in humans [6]. Th17 cells have an important role in the clearance of extracellular bacteria and fungi, but also appear to play a pathogenic role in several inflammatory and autoimmune diseases. In experimental animal models, IL-17-producing T cells are involved in the pathogenesis of experimental autoimmune encephalomyelitis (EAE), collagen-induced arthritis (CIA), colitis and psoriasis [7‐9].

In mice, the development of Th17 cells is driven by the transcription factor retinoic acid-related orphan receptor γt (RORγt). Differentiation from naïve T cells requires TGFβ, IL-1, and IL-6 [10, 11]. In humans, the origin of Th17 cells and the factors that regulate their development remain controversial, but like murine Th17 cells, IL-1 and IL-6 are essential and it is likely that TGF-β also plays a role. Both murine and human Th17 cells require IL-23 for their expansion and survival. Th17 differentiation is not only regulated by cytokines but also by environmental and dietary factors, such as aryl hydrocarbons [12, 13] and vitamin D₃ [14, 15]. In addition to IL-17, Th17 cells have been shown to produce IL-21, IL-22, TNFα and IFNγ [16].

In RA, IL-17 has been detected in synovial fluid (SF) and synovium [17‐21]. Its expression is associated with inflammation and joint destruction, as well as with production of IL-1β and TNFα. In addition to stimulating the production of these proinflammatory cytokines, IL-17 acts synergistically by amplifying their effects [22, 23]. We have previously identified IL-17-producing T cells within SF and synovial tissue, and demonstrated that RA synovial fibroblasts treated with IL-17 and TNFα promote the survival and functional lifespan of neutrophils, contributing to the increased number of neutrophils observed in the rheumatoid synovial microenvironment [23]. Based upon the combined evidence for a role of IL-17 in inflammation, targeting of IL-17 is now being tested as a new therapeutic strategy for the treatment of RA [24].

However, relatively little is known about the phenotype, cytokine profile and frequency of Th17 cells in the synovial environment and how they relate to RA disease activity. Two reports have shown that the frequency of Th17 cells was increased in the blood of RA patients compared with healthy donors [15, 25], whilst Shahrara et al. demonstrated that the percentage of Th17 cells was higher in RA SF compared with normal and RA peripheral blood [15, 25, 26]. In addition, IL-17-producing T cells were shown to be enriched in the joints of children with juvenile idiopathic arthritis [27]. In contrast, Yamada et al. reported fewer Th17 cells in the joints of RA patients compared to peripheral blood [28]. In light of these conflicting data, we investigated the phenotype and frequency of IL-17-producing T cells in the blood, SF and synovial tissue of patients with RA, examining the cytokine profile of this population and correlation with disease activity.

There is considerable evidence, from both animal and human studies, that IL-17 plays a role in inflammatory arthritis [17‐21, 34]. Mechanistically, IL-17 promotes osteoclastogenesis, partly through upregulation of RANKL expression [35], and also synergises with other inflammatory cytokines, such as TNFα and IL-1, to amplify their effects [22, 23]. These observations have led to the development of new therapeutic strategies aimed at targeting IL-17 or suppressing Th17 cells [24]. However, recent studies of the frequency of Th17 cells in RA patients have yielded conflicting data. Furthermore, the profile of cytokines produced by these cells ex vivo and their relationship with disease activity has not been well defined. Some studies have only examined this population in the blood of RA patients [25], whilst only two studies have examined these cells in SF or synovium [26, 28].

We found no significant difference in the frequency of IL-17-producing CD4 T cells in the blood of RA patients and healthy controls. This contrasts with data from the study of Shen et al. (which looked at blood from 12 RA patients) but is consistent with data from Yamada et al. (which looked at blood from 123 RA patients and mononuclear cells isolated from the joints of 12 patients) [25, 28]. The study by Shen et al., whilst primarily focussed on Th17 cells in ankylosing spondylitis, demonstrated an increased frequency of IL-17-producing CD4 T cells in RA patients blood compared to healthy controls which correlated with both CRP and swollen joint counts (SJC). We would suggest that the larger sample sizes in our study (blood examined from 44 RA patients) and the study of Yamada et al. may provide a broader representation of what is recognised to be a very heterogeneous disease.

Our observation of a small but significant increase in the frequency of IL-17-producing CD4 T cells in SF compared to peripheral blood of RA patients, and the same for the frequency of memory CD4 T cells (CD45RO-positive) producing IL-17, are consistent with data from Shahrara et al. [26] but differ from the findings of Yamada et al. [28]. It should be noted however that Yamada et al. studied cells from the joints of only 12 RA patients and data from synovial fluid (n = 8) and synovial tissue (n = 4) were grouped together for analysis; this may be relevant in the context of our observation that the percentage of Th17 cells was lower in synovial tissue than in the blood of RA patients whereas the reverse was seen when synovial fluid was compared with blood.

Our data also demonstrate that there is marked variability in the frequency of IL-17-positive CD4 T cells in SF between RA patients. The underlying cause for this remains unclear but drug therapy may be a contributing factor. As discussed, Yamada et al. reported that patients taking prednisolone or MTX had a higher frequency of IL-17-producing CD4 T cells, but could not attribute this directly to treatment [28]. In our study we did not observe this phenomenon. Most patients in this study were receiving treatment with a non-steroidal anti-inflammatory drug, a disease modifying anti-rheumatic drug, a biologic agent or a combination of these (Table 1). We assessed the relationships between therapy and frequencies of IL-17-positive CD4 T cells but found no significant relationships. As it is not yet known how current therapeutic regimes affect the frequency of IL-17-producing CD4 T cells, it is possible that medication, in addition to other factors, such as disease duration and seropositivity, may contribute to the heterogeneity in frequency of IL-17-producing CD4 T cells amongst patients studied here. In the context of medication, appropriately designed longitudinal studies would be needed to address the impact of drug therapy on IL-17 producing cell populations.

IL-17-producing T cells were originally misclassified under the Th1 umbrella. Since the characterisation of Th17 cells, it has been identified that some IL-17-producing T cells coexpress IFNγ. This IFNγ-positive IL-17-positive subset is particularly enriched in the gut of patients with Crohn's disease [36]. Here we found that the majority of IL-17-producing CD4 T cells in the blood and SF of patients with RA also coexpress IFNγ. However, in the SF we observe a greater frequency of IFNγ-positive IL-17-negative CD4 T cells compared with blood, suggesting that Th1 cells and not Th17 cells predominate in the established RA joint. The relative paucity of IL-17-positive CD4 T cells in either the SF or synovium is at odds with the high levels of IL-17 detected in RA SF [17‐21]. One explanation may lie with the use of a non-physiological stimulus in vitro, such as PMA/ionomycin; stimuli in the RA joint, such as antigen or cytokines, may drive a different cytokine response in vivo. Ziolkowska et al. demonstrated that RA PBMC produced significantly more IL-17 than SFMC when challenged with PMA/ionomycin in vitro, however, stimulation with IL-15 induced greater IL-17 production from SFMC than PBMC [21]. It is thus possible that the inflammatory environment of the rheumatoid synovium drives Th17 cells to produce IL-17 in a cytokine-dependent manner. In addition, the concept that CD4 T cells may not be the only source of IL-17 in the joint is being increasingly recognised. Mast cells have recently been identified as a source of IL-17 in RA synovium and are potent producers of IL-17 upon stimulation with TNFα, immune complexes and LPS [37].

IL-22 has been characterised as one of the effector cytokines secreted by Th17 cells [9, 30, 31], but it is also produced by Th1 cells [38]. Although this novel cytokine belongs to the IL-10 family, it is proinflammatory in function [39]. In animal models such as CIA, IL-22 is proinflammatory, whilst IL-22(-/-) mice are less susceptible to CIA than wild-type mice [32]. In human studies, IL-22 has been proposed to play a role in the pathogenesis of autoimmune inflammatory diseases, such as psoriasis [40] and Crohn's disease [41, 42]. In RA, expression of IL-22 was found to be upregulated in synovium and capable of inducing synovial fibroblast proliferation and chemokine production [33]. Here, we found that IL-17-producing CD4 T cells in the SF coexpressed very little IL-22. Furthermore, we identified an IL-22-producing CD4 T cell population distinct from IL-17-producing CD4 T cells, but this population was not elevated in the blood of RA patients compared to healthy controls nor was it enriched in RA SF. This CD4 T cell subset may be similar to that recently reported to secrete IL-22 but not IL-17 and to be involved in the skin pathophysiology of psoriasis [43, 44]. The paucity of IL-22-producing CD4 T cells in SF lends support to the notion that the primary source of IL-22 in the joint is not T cells but rather synovial fibroblasts and/or macrophages as reported by Ikeuchi et al. [33].

IL-22 production by Th17 cells has been shown to be dependent upon IL-23 [38, 45, 46], thus our observation, which corroborates that of Shen et al. [25], that very few IL-17-producing CD4 T cells express IL-23R provides a potential explanation for the paucity of IL-22-positive IL-17-positive CD4 T cells we observed. Considering the importance of IL-23 to the development and maintenance of Th17 cells, it is striking that over 90% and 85% of IL-17-producing CD4 T cells in the SF and blood respectively did not express IL-23R. However, our data are consistent with recent evidence showing that bioactive IL-23 (p19/p40) was barely detectable in joints of patients with RA [47].

In contrast to the pattern of IL-22 expression, the majority of IL-17-producing CD4 T cells in SF coexpressed TNFα; these double positive cells were significantly elevated in SF compared to blood. Although the percentage of IL-17-producing CD4 T cells coexpressing TNFα is very small compared to that of TNFα-producing CD4 T cells found in the SF, this may be relevant in terms of therapy. One of the mechanisms by which monoclonal antibodies against TNFα mediate their therapeutic effects, in addition to TNFα blockade, may be through induction of apoptosis in monocytes and T cells by outside to inside signalling through transmembrane TNFα [48‐50]. Membrane expression of TNFα on IL-17-producing CD4 T cells would make these cells a potential target of monoclonal antibodies and may help to explain the beneficial effects of these therapies.

Upon identification of the IL-17/IL-23 axis, Th17 cells were viewed as a driving force in the pathogenesis of several autoimmune diseases previously associated with a Th1 predominance. In the case of multiple sclerosis evidence has supported this shift in paradigm [51, 52]. However in RA, evidence is less clear. There is no doubt that cytokines associated with the Th17 lineage, such as IL-17, IL-6, IL-1β, IL-22, can be found within the RA joint [20, 33], and as demonstrated here, there is evidence for a small but significant enrichment of Th17 cells within RA SF. In our study, however, we have only examined established RA and many patients had longstanding severe disease; thus 60% (6 of 10) of patients from whom synovial tissue was obtained and 29% (4 of 14) from whom synovial fluid was obtained were on anti-TNF agents. Recent data from Colin et al. and previous data from our group suggest that the situation may be different in the early phase of RA [15, 20]. Further studies are required to define the role of Th17 cells in the early phase of disease. The absence of an enrichment of Th17 cells in the synovial tissue of established RA patients is paradoxical to the role this cytokine is reported to play in the pathogenesis of RA [17‐21]. This leaves the possibilities that Th17 cells located outside the synovium, such as the juxta-articular bone marrow, may be an additional source of IL-17, that other cells in the synovium, including mast cells [37] may also produce IL-17 or that the experimental techniques used to reveal IL-17 production ex vivo do not reveal the true potential of T cells in the rheumatoid joint to produce IL-17. Future work will need to address these issues.

Our findings demonstrate a modest enrichment of IL-17-producing CD4 T cells in RA synovial fluid compared to peripheral blood. Th17 cells in synovial fluid produce more TNF than their peripheral blood counterparts, but are not a significant source of IL-22 and do not express IL-23R. However, the percentage of CD4 T cells which produce IL-17 in the rheumatoid joint is low, suggesting that other cells may be alternative sources of IL-17 within the joints of RA patients.