Interleukin‐17 in rheumatoid arthritis: If T cells were to contribute to inflammation and destruction through synergy

The favorable clinical results obtained with tumor necrosis factor α (TNFα) and interleukin-1 (IL-1) inhibitors may suggest that the story of the contribution of cytokines to arthritis is almost over. However, additional cytokines that may contribute to joint inflammation have been described. Among these, IL-17 is of interest in this context, for two major reasons: first, similar to TNFα and IL-1, IL-17 has proinflammatory properties, and second, it is produced by T cells. Since the role of T cells has been a matter of debate, demonstration of the contribution of IL-17 to arthritis would also imply a role for the IL-17–producing T cells themselves. These issues are important, since they may lead to new therapeutic applications. The two initial reports describing the molecule later named IL-17 were published in 1995 and 1996. Researchers at Immunex (1) and Schering-Plough (2) defined a new cytokine for which no function was found until both groups demonstrated its effect on mesenchymal cells, wherein it induced the production of various cytokines associated with inflammation. IL-17 is a 17-kd protein that is secreted as a dimer. Mouse and human IL-17 share 25% of the amino acid sequence. Human IL-17 acts on mouse cells, whereas the reverse is not the case. Human IL-17 has 72% homology with HVS13, an open-reading frame from the T lymphotrophic herpes virus saimiri, and has 63% homology with murine CTLA8 (1). IL-17 was first described as a T cell product. At the present time, this description remains valid, with almost no exception. Eosinophils, which can be the source of almost any cytokine, were shown to be the source of IL-17 in bronchial secretions in asthma (3). The contribution of the eosinophil-derived IL-17 compared with that derived from T cells is unknown. More recently, sequences with 4 cysteine residues were found through a systematic genebank analysis. Since the presence of 4 cysteines is characteristic of IL-17, proteins with the same sequence were also named members of the IL-17 family (4). They are now listed as IL-17B–IL-17F and include the new IL-25 (5, 6). The nomenclature, however, remains confusing. The definition of their functions is still at an early stage. In particular, their role in inflammation and related diseases remains to be investigated in more detail. Accordingly, this report focuses primarily on the founding father of the IL-17 family. Similar to other cytokines, IL-17 acts through cell-membrane receptors. A mouse IL-17 receptor was isolated first and was then followed by the description of the human counterpart (7, 8). A comparison of the known sequences and protein structure indicated that the IL-17 receptor is a member of a new family (7). In contrast to the limited production of IL-17 on T cells, its receptor is widely expressed on almost any cell type. However, affinity measurements showed that the isolated IL-17 receptor interacts with IL-17 with low affinity, in contrast to the values usually obtained with other cytokine receptors which are often composed of more than one chain. This suggests that additional chains may be associated with the IL-17 receptor. However, as described below, a soluble form of this low-affinity chain is a potent inhibitor of IL-17 function. Issues involving a patent are probably the reason why additional research on IL-17 receptors is still missing. Intracellular targets of IL-17 have been identified, and their pharmacologic control is the target of active research. IL-17 shares transcriptional pathways with IL-1 and TNFα. In particular, p38 and nuclear factor κB are the key transcriptional factors for IL-17 function. These pathways have been identified in synoviocytes (9) and chondrocytes (10) and appear to be the same in osteoblasts and myoblasts. The use of common pathways is of importance when considering the combination of cytokines, which often act in synergy. IL-17 was described for its ability to induce the production of IL-6 and IL-8 by fibroblasts (2). Early studies used synoviocytes, which, from the beginning, suggests that IL-17 has been shown to play a role in joint inflammation. These properties were long ago also assigned to IL-1 and TNFα. When the 3 cytokines were compared for their effect on IL-6 production, clearly IL-1 was the most potent, since concentrations as low as a few pg/ml were as potent as 10 ng/ml of TNFα or IL-17. The most interesting part of these studies with disease application in mind was the demonstration of an additive and often synergistic effect between IL-1 or TNFα and IL-17 (11). This observation is important since, in any condition associated with inflammation, the combined production of TNFα, IL-1, and IL-17 is to be expected. For IL-17 production, the limitation is with the presence of T cells with IL-17–producing capacity, whereas the presence of monocytes producing IL-1 and TNFα is far less specific. In these conditions, IL-17 will have a regulatory role in which it is responsible for a fine tuning of the inflammatory response (12). This synergy was observed first with synoviocytes (11) and later with chondrocytes (13, 14), osteoblasts (15), and myoblasts. In addition, these cytokines interact with each other at the level of production. IL-17 was found to increase the production of IL-1 and TNFα by monocytes (16). Thus, low concentrations of IL-17 will increase the proinflammatory properties of these monocyte-derived cytokines. This feature imputes the role of IL-17 as a regulatory cytokine, in accordance with the role assigned to T cells. The mechanisms responsible for such synergy are not fully understood. The signaling pathways and transcription factors involved with IL-1, TNFα, and IL-17 are very similar. However, in vitro studies have used rather high concentrations of a single cytokine. When lower concentrations of cytokines, which reflect more directly the in vivo situation, are used in combination, a very potent effect on transcription-factor activation can be observed. Our preliminary observations with various mesenchymal cells, such as synoviocytes, osteoblasts, or myoblasts, indicate that cytokine combination results in a recruitment of additional transcription factors not activated when cytokines are used alone, even at optimal concentrations. It should be noted that the number of IL-17–producing cells in RA synovium is indeed very small (17). Although the level of secretion per cell could then be quite high, the biologic significance of such secretion appears to imply interactions between IL-17 and other monocyte/synoviocyte-derived factors. This could be the mechanism by which such a limited cell population might contribute to disease. Cytokines have an important effect on systemic response to stress. In addition, T cells are critical for cell-mediated immunity, suggesting a role for IL-17 in host defense. Indeed, IL-17 induces the maturation of CD34 hematopoietic lymphoid and myeloid progenitors (18); this includes, in particular, maturation of neutrophils in combination with stem cell factor and granulocyte colony-stimulating factor (G-CSF) (2). Circulating neutrophil counts were found to be regulated by a feedback loop involving IL-17 and G-CSF (19). In the same way, IL-17 induced neutrophil accumulation in infected lungs. In addition, IL-17 activated neutrophils to produce chemokines and hematopoietic growth factors (20). Mice deficient in IL-17 receptor showed an increased sensitivity to lung bacterial infection, resulting in higher mortality rates, which suggests a mechanism for the development of CD4 T cell defects associated with bacterial pneumonia (21). IL-17 contributes to other features of T cell–mediated immunity. In mice, IL-17 induced T cell–dependent, tumor-specific immunity (22). Inhibition of IL-17 could control experimental allogeneic graft rejection (23). Accordingly, detection of IL-17 expression may serve as a predictive parameter for subclinical renal allograft rejection (24). In rheumatoid arthritis (RA), it is possible that the immune defect associated with TNFα inhibition could be related to a role of IL-17 in the control of opportunistic infections such as tuberculosis. In this case, inhibition of TNFα modifies the synergy of the cytokines acting on, or produced by, T cells. Thus, in patients in whom systemic immune defenses are already altered, these effects of IL-17 on host defense have to be considered carefully when the inhibition of IL-17 may be used. The first results with regard to IL-17 in arthritis were obtained when an enzyme-linked immunosorbent assay was made available. Levels of IL-17 were found to be higher in RA synovial fluid when compared with osteoarthritis (OA) synovial fluid (25). It was critical to demonstrate that the measured IL-17 in synovial fluid was functional and produced by the synovium itself. The first step was to demonstrate the production of functional IL-17 by RA synovium (17). Quantification of IL-17 was performed using a specific biologic assay, which assessed the production of IL-6 by synoviocytes stimulated with supernatants first preincubated with a blocking anti–IL-17 antibody. Functional IL-17 was spontaneously produced at high levels by many RA samples, by a few OA samples at very low levels, and by none of the normal synovium-explant cultures. IL-17 messenger RNA expression was demonstrated by reverse transcription–polymerase chain reaction in RA synovium, but none was found in OA synovium samples. IL-15 produced by synoviocytes is also a potent inducer of IL-17 production (26). It is of interest to note that IL-17 was also found, by microarray analysis, to be highly expressed in the brain of patients with multiple sclerosis (27). In synovium, IL-17–producing cells represent a minor subset of lymphocytes found in lymphocytic infiltrates around vessels. These cells were found to express CD4 but had lost their CD3. On average, these cells are large, with a remote nucleus, and have the appearance of plasma cells (28, 29). The parallel can be drawn between B cells, which have lost their B cell membrane receptor when they start to produce soluble immunoglobulin as plasma cells. Addition of exogenous IL-17 to RA synovium resulted in an increase in the production of IL-6, whereas introduction of a blocking anti–IL-17 antibody reduced this effect. These first results indicate that IL-17 was present and could contribute to the active proinflammatory and destructive pattern that is characteristic of RA. In synovium, T cells are in contact with synoviocytes, monocytes, and dendritic cells. In bone, the same interactions occur, whereas chondrocytes, which are protected by cartilage extracellular matrix, are the target of soluble factors. These interactions are summarized in Figure 1 and the related functions of IL-17 are presented in Figure 2. Interactions between cells and cytokines associated with destruction in rheumatoid arthritis. IL = interleukin; TNF-α = tumor necrosis factor α. Effects of IL-17 on cell populations present in the rheumatoid joint. NO = nitric oxide; PGE2 = prostaglandin E2; RANKL = receptor activator of nuclear factor κB ligand (see Figure 1 for other definitions). On synoviocytes, IL-1β, TNFα, and IL-17 were shown to induce production of IL-6 and leukemia inhibitory factor (LIF), with IL-1 being much more potent than IL-17 or TNFα (11). Using low concentrations of IL-17 and IL-1β in combination, a synergistic effect on the production of IL-6 was observed, whereas an additive effect was observed on the production of LIF. Both IL-4 and IL-13 had a modest stimulatory effect on the IL-1– and IL-17–induced production of IL-6, but both inhibited production of LIF. These findings indicate that low levels of cytokines produced by monocytes (IL-1) and T cells (IL-17) can act together on synoviocytes. Thus, some RA synovium T cells producing IL-17 can activate mesenchymal cells, leading to an increased proinflammatory pattern that is sensitive, in part, to Th2-cytokine regulation. Increased production of chemokines and other cytokines was observed in the presence of IL-17. Synergy was again observed between IL-1, TNFα, and IL-17 for production of macrophage inflammatory protein 3α (MIP3α; now renamed CCL20), a chemokine involved in the migration of memory T cells and immature dendritic cells (30). On dendritic cells, IL-17 appears to induce differentiation of mouse progenitor dendritic cells when the cells are stimulated with G-CSF and IL-4 (31). This leads to graft rejection, which is controlled, in part, by inhibition of IL-17. In RA synovium, a defect in the maturation of dendritic cells was observed, with evidence of a relative accumulation of immature dendritic cells (29). IL-17 interacts with IL-1 and TNFα to influence the migration of dendritic cells, acting in synergy to induce the production of CCL20/MIP3α (30). Such production by synoviocytes was inhibited by the antiinflammatory cytokines IL-4 and IL-13. CCL20-producing cells were located in the lining layer and perivascular infiltrates in close association with CD1a immature dendritic cells. Addition of exogenous IL-17 to synovium explants increased CCL20 production. Conversely, specific soluble receptors for IL-1, IL-17, and TNFα inhibited CCL20 production to various degrees, but maximal inhibition was obtained only when the 3 receptors were combined. These findings indicate that interactions between monocyte and Th1 cell–derived cytokines contribute to the recruitment of T cells and dendritic cells by enhancing the production of CCL20 by synoviocytes. Proinflammatory cytokines produced by RA synovium have destructive patterns (Figure 1). In addition to the classic properties of IL-1 and TNFα, IL-17 has been shown to increase the spontaneous production of matrix metalloproteinase 1 (MMP-1) by synoviocytes, with IL-1 again being more potent. Addition of IL-4, IL-13, and IL-10 to synoviocyte cultures reduced the spontaneous production of MMP-1 and induced production of the tissue inhibitor of MMP-1 by synoviocytes stimulated with IL-17 and/or IL-1β. In the presence of anti–IL-17 blocking antibody, MMP-1 production and collagenase activity by RA synovium was reduced and associated with a reduction of type I collagen C-telopeptide (CTX) fragments released in the supernatants, demonstrating the direct contribution of IL-17 in destruction (32). From these results, IL-17 and its producing T cells appear to contribute to the RA inflammatory destructive process. On chondrocytes, IL-17 induced production of prostaglandin E2 (PGE2) and nitric oxide (NO) by cartilage explants in an IL-1–independent, but LIF-dependent manner (14, 33, 34). Similar to the effects of IL-1, IL-17 increased production of collagenase 3, mainly through activation of activator protein 1. Differences were observed in which IL-17 activated FosB and IL-1β activated cfos, suggesting differential pathways (35). Other members of the IL-17 family, such as IL-17E or IL-25, can also have direct cartilage-breakdown activity (34). On osteoblasts, IL-17 induced IL-6 production (36). As in chondrocytes, but only in combination with TNFα, IL-17 induced NO production in osteoblastic cells and fetal mouse metatarsal joints by a nuclear factor κB–dependent mechanism (37). During cell interactions between osteoblasts and osteoclast precursors, the presence of IL-17 induced osteoclastogenesis, an effect found to be mediated by PGE2 (25). IL-17 and other cytokines stimulating osteoclastogenesis, such as IL-1β and TNFα, increased the expression of receptor activator of nuclear factor κB ligand (RANKL), with a decrease in the expression of osteoprotegerin in osteoblasts/stromal cells (38). IL-17–producing T cells express the membrane form and secrete the soluble form of RANKL (39). These functions make IL-17 a new cytokine involved in bone resorption (15). In the context of RA, T cells in juxtaarticular bone are the source of IL-17, which then acts locally to influence joint destruction (28). The origin and role of IL-17 in cartilage and bone destruction during RA remain to be clarified. In human ex vivo models, addition of IL-17 enhanced IL-6 production and collagen destruction and inhibited collagen synthesis by RA synovium explants (28). In isolated mouse cartilage, IL-17 increased loss of cartilage proteoglycan and inhibited its synthesis (28, 40). On human RA bone explants, IL-17 also increased bone resorption and decreased bone formation (28). In these conditions, addition of IL-1 increased the effects of IL-17. Blocking of bone-derived endogenous IL-17 with specific inhibitors resulted in a protective inhibition of bone destruction. Conversely, intraarticular administration of IL-17 into a normal mouse joint induced cartilage degradation. These effects were independent of IL-1, since the IL-17 effect was still observed in IL-1–deficient mice (41). This was extended with the continuous administration of IL-17 by gene therapy (42). In conclusion, the contribution of IL-17 derived from synovium and bone marrow T cells to joint destruction suggests that the control of IL-17 could be used as a strategy for the treatment of RA. Because IL-17 is a T cell product, it was logical to search for the position of IL-17 in the now classic Th1/Th2 subsets (43). This classification is useful to understand cytokine patterns in relation to disease. RA is thus classified as a Th1 disease because of the high level of destruction and defective repair processes, whereas, conversely, scleroderma is characterized by abnormal repair activity leading to diffuse fibrosis. Accordingly, production of IL-4 is defective in RA but contributes to matrix formation in scleroderma. Defects have been identified in RA synovium T cells, which appear unable to switch to IL-4 production (44). Similarly, Th1 T cell clones in RA synovium are unable to switch to a Th2 phenotype in the presence of Th2-promoting conditions (45). Results with these T cell lines and clones derived from RA synovium allowed the classification of IL-17 as a Th1 cytokine (46). Indeed, IL-17 is always produced in association with and some clones have been found in association with a pattern with and low levels of This was extended when at function. When T cell clones were in contact with RA synoviocytes, IL-17–producing cells were the most potent of IL-6 production by synoviocytes they inhibited synoviocyte collagen results were obtained with Th2 These results indicate that IL-17–producing Th1 cells contribute to the defective repair activity in joint inflammation. of clones in Th1 conditions increased IL-6 production, whereas Th2 conditions inhibited IL-6 production and repair activity (45). Because rheumatoid synovium is by Th1 cells, cell interactions in a proinflammatory pattern with defective which can be at in part, by a Th2 clones from it was not possible to IL-17 with a cytokine Such an in reflect the contribution of the to the cytokine Indeed, of a switch from Th1 to Th2 is to with T cell clones from RA This may be to a in vivo to conditions Th1 cytokines are and Th2 cytokines are In IL-17E appears to be a Th2-promoting cytokine in of In addition, IL-17E mice showed both and high Th2 cytokine expression At specific IL-17 inhibitors have been in vitro and in They include blocking and a soluble IL-17 receptor. These are the of and for TNFα The interactions involved in and therapeutic are shown in Figure and therapeutic to control joint destruction and repair in rheumatoid arthritis. = interleukin-1 receptor = soluble = growth factor (see Figure 1 for other definitions). Although these inhibitors have not been in RA or other inflammatory the for such an application has been obtained from studies of and ex vivo human IL-17 receptor reduced the and joint of arthritis in the Using RA synovium and bone ex vivo models, addition of IL-17 reduced inflammation and destruction. Since IL-1, TNFα, and IL-17 have many additive and/or synergistic effects in it was of interest to their combined inhibition with the soluble IL-17 receptor would lead to an enhanced effect on ex vivo of synovium inflammation and bone destruction Indeed, therapy has shown in RA but some patients not a response and the treatment is RA synovium and bone explants were in the presence of the soluble receptor that is used in or in combination with type IL-1 and IL-17 soluble receptors. In the synovium, 3 soluble receptors used decreased production of IL-6 by and that of by therapy was more IL-6 production and collagen by to In bone, similar results on the effects of combination therapy on IL-6 production were For optimal inhibition was on the presence of the soluble IL-1 receptor. These results the of combination which may increase the of patients as as the of IL-17 function can also be inhibited with Th2 cytokines, which, at the same time, control IL-1 and TNFα. On T cell clones from RA synovium, addition of Th2-promoting conditions reduced the production of both and IL-17, conditions increased the production of these cytokines (45). These results demonstrate that RA Th2 clones Th1 and clones are more This was extended with the administration of IL-4 to mice with arthritis through gene therapy Although inflammation in mice, IL-4 bone and the formation of cells. RNA and protein levels of IL-17 in the synovial tissue were of RANKL was leading to reduced osteoclast activation bone destruction. in in vitro bone samples from RA patients of type I collagen by IL-4 also enhanced synthesis of type I suggesting that it tissue the effects associated with IL-17 blocking of IL-17 in may have by T cell–mediated immunity. This may the mechanisms of host the development of as as T cell–mediated and control of growth regard to the contribution of T cells to RA, results with IL-17 have some new on this highly This proinflammatory cytokine is produced by RA synovial more by Th1 When they in contact with synoviocytes, IL-17–producing T cells have an increased ability to influence destruction and to repair activity in synoviocytes. IL-17 produced by T lymphocytes from bone marrow contributes locally to juxtaarticular bone destruction. IL-17 acts in synergy with or to TNFα and IL-1, by their production and Accordingly, inhibition of IL-17 could be in human RA, as has been demonstrated in many of RA. for treatment are already but a is to combination therapy with the cytokine inhibitors may be of interest to increase the and of response arthritis