Rheumatoid arthritis (RA) is a systemic disease that causes progressive joint damage and disability. Rheumatoid synovium is histologically characterized by prominent infiltration of inflammatory mononuclear cells, such as T cells and macrophages, and the proliferation of synovial fibroblasts. Inflammatory cytokines, including tumor necrosis factor α (TNFα), interleukin-1β (IL-1β), and IL-6, which are mainly produced by macrophages, play a central role in the development of synovitis (1, 2). For example, TNFα is shown to directly induce synovial fibroblast proliferation, which leads to pannus formation (3). TNFα is also critical for the expression of inflammatory chemokines and adhesion molecules, which, in combination, facilitate further recruitment of inflammatory leukocytes and perpetuation of inflammatory responses (2).
Neutralization of these inflammatory cytokines has been proven effective as a therapeutic strategy in human RA (4). However, many patients do not respond to such anticytokine therapy, indicating heterogeneity of the disease. Different cytokines may play a dominant role in these patients. In this regard, efforts should be continued to search for any novel cytokines that are critically involved in the pathogenesis of RA.
In contrast to proinflammatory cytokines such as TNFα, IL-10 is regarded as an immunomodulatory cytokine with a broad spectrum of biologic activities, including antiinflammatory and immunosuppressive effects. Collagen-induced arthritis can be ameliorated by administration of IL-10 (5). In the rheumatoid synovium, large amounts of IL-10 are produced, and neutralization with anti–IL-10 antibodies increases the production of the proinflammatory cytokines, including TNFα and IL-1β (6). These observations have established the antiinflammatory role of IL-10 in RA, although the results of clinical trials in which IL-10 was administered to RA patients have not been encouraging.
Recently, a novel cytokine of the IL-10 family, IL-22, was identified. By use of the complementary DNA (cDNA) subtraction method, IL-22 was cloned as a molecule induced by IL-9 in murine T cells. IL-22 encoded 180 amino acids, showing 22% amino acid identity with IL-10 (7). Subsequently, a human homolog was cloned and found to have 25% identity with human IL-10. Initial studies have shown that IL-22 is expressed by thymic lymphocytes, mast cells, T cells activated with anti-CD3, or concanvalin A (8). Another study has shown that in peripheral blood mononuclear cells (PBMCs), expression of IL-22 was exclusively detected in T cells, especially upon Th1 polarization, and in natural killer cells (9). In vivo, IL-22 was shown to be expressed in the thymus and the brain (8).
A receptor for IL-22, IL-22 receptor 1 (IL-22R1), was cloned and the functional IL-22 receptor was then identified as a complex of IL-22R1 and IL-10R2 (the second chain of the IL-10 receptor complex) (10). IL-22R1 is expressed in tissues such as those from the liver, kidney, pancreas, and skin, but not in PBMCs (11). Upon lipopolysaccharide (LPS) stimulation, IL-22R1 messenger RNA (mRNA) expression is highly up-regulated in the liver, in contrast to IL-10R2, which is expressed constitutively in a variety of tissues (11). A soluble receptor that binds to IL-22 was also cloned as a way of identifying the class II cytokine receptor family, and this was designated as IL-22 binding protein (IL-22BP). IL-22BP has 34% amino acid identity with the extracellular domain of the IL-22R1 and has been shown to neutralize the effects of IL-22 (12, 13).
Although the pathophysiologic function of IL-22 is largely unknown, the following data suggest that it has proinflammatory properties, in contrast to the characteristics of IL-10. IL-22 induced up-regulation of serum amyloid A (SAA) in Hep-G2 cells. In vivo, IL-22 also increased SAA expression in the liver, and in turn, LPS injection induced up-regulation of IL-22 in various tissues, including those of the kidney and the liver (14). In contrast to IL-10, IL-22 neither inhibited production of proinflammatory cytokines by macrophages nor induced immunoglobulin production by activated human B cells (15). IL-22 activated STAT-3 and, to a lesser extent, STAT-5 in MES13 cells, and STAT-1 and STAT-3 in Hep-G2 cells. Of interest, a recent study has shown that IL-22 in a rat hepatoma cell line, H4IIE, activated not only STAT, but also kinases such as ERK, JNK, and p38 MAPK (16). Since p38 MAPK is regarded to be an important kinase for the inflammatory responses, these data support the proinflammatory property of IL-22.
In this study, we examined whether IL-22 plays a role in the pathogenesis of RA. To this aim, we analyzed the expression of IL-22 at both the mRNA and the protein level, using samples derived from RA patients. We further examined the effects of recombinant IL-22 (rIL-22) on cultured synovial fibroblasts derived from RA patients (RASF). Our results demonstrate that high levels of IL-22 are expressed in RA synovial tissues, and that rIL-22 induces synovial fibroblast proliferation and production of chemokines by RASF.
- Top of page
- PATIENTS AND METHODS
IL-22 is a recently identified cytokine of the IL-10 family, but its pathophysiologic function is not well known and neither its expression nor its role in human inflammatory diseases have been explored. In this study, we found that IL-22 was expressed by the synovial fibroblasts and macrophages of the rheumatoid synovium, and that IL-22R1 was expressed by rheumatoid synovial fibroblasts. In vitro, cultured RASF expressed both IL-22 and IL-22R1. Recombinant IL-22 induced proliferation and expression of MCP-1 by RASF. Based on these findings, we propose that IL-22, produced by synovial fibroblasts and macrophages, may promote inflammatory responses in RA synovial tissues by inducing synovial fibroblast proliferation and production of chemokines. RASF expressed both IL-22 and IL-22R1, suggesting that IL-22 works in an autocrine as well as a paracrine manner.
IL-22 was originally identified as a cytokine that is produced by T cells stimulated with IL-9 (6). Therefore, we first expected that activated T cells in the rheumatoid synovium might produce IL-22; however, macrophages and synovial fibroblasts, but not T cells, expressed high levels of IL-22. Immunofluorescence studies showed that both vimentin-positive synovial fibroblasts and CD68-positive macrophages were positive for IL-22. Cultured RASF also expressed IL-22, as confirmed by RT-PCR. Previous data have suggested that IL-22 is produced by hematopoietic cells, which in turn affect nonhematopoietic mesenchymal cells (9). In this context, our findings are unique in that IL-22 could be produced not only by hematopoietic cells, but also by mesenchymal cells.
In contrast to IL-22, IL-22R1 expression in the RA synovium, detected by immunohistochemical analysis, was limited to vimentin-positive synovial fibroblasts. Macrophage-lineage cells did not express IL-22R1. In vitro, cultured RASF expressed IL-22R1 both at the mRNA and the protein level. The functional IL-22 receptor complex consists of IL-22R1 and IL-10R2 (10). Since previous studies have shown that RASF express IL-10R2, it is conceivable that RASF could respond to IL-22 in the presence of IL-22R1 (21). Indeed, we found that rIL-22 induced RASF proliferation and production of MCP-1 by RASF. Because synovial fibroblast proliferation and macrophage infiltration by chemokines are key events in the development of synovitis in RA, these are important findings in that they presume the promotion and progression of rheumatoid inflammation by IL-22. Induction of cell growth by IL-22 is shown in other cell types. For example, IL-22 induces proliferation of IL-22R1–transfected Baf3 cells (10). Aggarwal et al showed that IL-22R1 is expressed by the pancreas, and that IL-22 stimulates isolated primary pancreatic acinar cells and the acinar cell line 266-6 to up-regulate mRNA of osteopontin, a chemotactic factor for macrophages (22). Collectively, these previous findings support our hypothesis that IL-22 can induce cell growth and production of molecules with chemotactic activity.
IL-22 has 25% amino acid homology with IL-10 (8). Whereas IL-10 is regarded as an antiinflammatory and immunosuppressive cytokine, previous studies as well as our own investigations have suggested that IL-22 works as a proinflammatory cytokine. Molecular mechanisms explaining the difference of the effect between IL-10 and IL-22 have been partially elucidated. IL-10 exerts its function by acting on IL-10 receptors, which consist of IL-10R1 and IL-10R2 (23). Upon IL-10 binding, IL-10R–associated tyrosine kinases (JAK1 and Tyk2) are activated, followed by phosphorylation of STAT-1 and STAT-3 (23). Phosphorylated STAT-1 and STAT-3 translocate regulatory molecules such as suppressor of cytokine signaling, which results in the antiinflammatory properties of IL-10. Although IL-22 exerts its effect mainly through activation of STAT-1, STAT-3, and STAT-5, a recent study has shown that IL-22 can also activate other important kinases such as ERK-1/2 and p38 MAPK (16). Activation of ERK-1/2 and p38 MAPK are key events leading to proliferation or inflammatory responses, including chemokine production. It is plausible that in our system, IL-22 induced proliferation of synovial fibroblasts and MCP-1 production through the activation of ERK and p38 MAPK, respectively. Consistent with this hypothesis, we found that IL-22 induced activation of both ERK-1/2 and p38 MAPK (Figure 8).
Given that IL-22 is a potent inflammatory cytokine in RA, an important question is whether IL-22 can be a therapeutic target. In RA, a variety of cytokines, including TNFα, IL-1β, and IL-6, are thought to contribute to tissue injury. This has been proven by the fact that blocking these cytokines is beneficial in patients with RA (2). The relative contribution of these inflammatory cytokines to RA can differ. In this context, the possibility that blocking IL-22 impedes the progression of arthritis should be explored in animal models of RA. The naturally occurring soluble molecule IL-22RA, which has blocking activity, has been described (12, 24) and would be an interesting tool for therapeutic application.
In conclusion, our results suggest that IL-22, produced by synovial fibroblasts and macrophages, promotes inflammatory responses in RA by inducing the proliferation of synovial fibroblasts and production of chemokines by synovial fibroblasts. Further studies are necessary to establish the pathophysiologic role of IL-22 in RA.