Cytokines exert a wide variety of immunologic actions and are key mediators in the pathogenesis of human autoimmune disease. Perhaps the clearest evidence of the central role of these molecules in disease is the success of tumor necrosis factor α (TNFα) blocking therapies in the treatment of inflammatory arthritis (1). Studies of these agents have not only advanced clinical management but have elucidated, for the first time, definitive critical checkpoints in proinflammatory cascades. The significant proportion of patients in whom the disease does not respond to TNFα blockade, or achieves only modest improvement, indicates that TNFα-independent pathways also operate in inflammatory synovitis. Identification of such pathways, together with elucidation of factors that drive TNFα production itself, is now of paramount importance in developing new therapies. The identification of new cytokine activities in this context therefore deserves close attention. Interleukin-18 (IL-18) is a novel cytokine of the IL-1 family that has been identified in a number of autoimmune diseases, including rheumatoid arthritis (RA), inflammatory bowel disease, and psoriasis (2). High levels of IL-18 expression in adult-onset Still's disease (AOSD) are described in the report by Kawaguchi et al in this issue of Arthritis & Rheumatism (3).
IL-18 is an 18-kd glycoprotein derived by cleavage of a 23-kd precursor, pro–IL-18, by caspase 1. Pro–IL-18 is expressed in macrophages, dendritic cells, Kupffer cells, keratinocytes, chondrocytes, synovial fibroblasts, and osteoblasts, while IL-18 receptor α/β (IL-18Rα/β) is present on naive T lymphocyte subsets, mature Th1 cells, natural killer cells (NK cells), macrophages, neutrophils, and chondrocytes. IL-18R signals through the IL-1 pathway, via IL-1R–associated kinase/MyD88/TNF receptor–associated factor 6, to activate nuclear factor κB (4).
IL-18 is an important regulator of innate and acquired immune responses. It induces proliferation, cytotoxicity, and cytokine production by Th1 and NK cells, primarily in synergy with IL-12 (5). During early T cell differentiation, however, IL-18 can promote Th1 or Th2 responses independent of IL-4 or IL-12, suggesting a broader role in T cell maturation (6). Activation of macrophages and dendritic cells may operate through direct, interferon-γ (IFNγ)–independent pathways (via constitutive IL-18Rα), or indirectly through T cell–derived cytokine production (7). Consistent with the notion that IL-18 plays an early role in the induction of immune responses, IL-18 messenger RNA is constitutively expressed in macrophages, facilitating rapid generation and regulation of cytokine. Regulation of IL-18 is also mediated via splice variants of IL-18 binding protein (IL-18BP) that may be present in the extracellular milieu, where they bind IL-18 with high affinity and neutralize effector function (8). The results of in vivo studies using neutralizing antibodies and IL-18–deficient mice provide evidence of a regulatory role for IL-18 in host responses to infection (9) and in models of autoimmune diseases, including diabetes and experimental autoimmune encephalomyelitis (for review, see refs. 2, 4, and 5).
Is there a role for IL-18 in rheumatic disease? We recently defined the expression and potential proinflammatory activities of IL-18 in RA synovial tissues (10). IL-18 is present in RA synovial membrane, where it is localized in CD68+ cells of dendritic morphology in aggregates and lining layer and in fibroblast-like synoviocytes (FLS) (10). Importantly, both pro- and mature IL-18 can be detected in synovial biopsy specimens by Western blot. IL-18Rα is present on synovial lymphocytes and macrophages. In synovial cultures, IL-18 induces production of proinflammatory cytokines, especially TNFα, in synergy with at least IL-15 and IL-12. The mechanisms by which IL-18 induces TNFα release are currently being investigated. IL-18 can directly induce cytokine release from CD14+ synovial macrophages that express IL-18R. However, it may also amplify TNFα release by enhancing cell–cell interactions between synovial T cells and macrophages. Importantly, dose-response studies reveal that only very low concentrations of IL-18 are needed to promote significant levels of TNFα production in vitro, especially in synergy with IL-15 (2).
IL-18 promotes other activities relevant to disease pathogenesis, including angiogenesis and synovial neutrophil activation (11, 12). Moreover, activated chondrocytes express IL-18Rα, and IL-1β induces expression of mature IL-18. Independent of IL-1β, IL-18 reduces chondrocyte proliferation, up-regulates inducible nitric oxide (NO) synthase, stromelysin, and cyclooxygenase 2 (COX-2) expression, and enhances glycosaminoglycan release (13). Factors regulating IL-18 in synovium are as yet unclear. IL-18 expression is up-regulated in vitro in FLS by IL-1β/TNFα. This indicates that positive feedback loops regulating proinflammatory cytokine production and Th1 cell activation may arise not only from invading immune cell activity, but also following activation of resident tissue cells (e.g., FLS) in synovial inflammatory responses. IL-18 induces NO release by RA synovium (2), which provides for negative feedback of IL-18 synthesis since NO inhibits caspase 1 activity. It is probable that primary regulatory activity resides in IL-18BP expression. Our knowledge about the presence and isoform specificity of IL-18BP in synovial membrane is currently unclear, but this could be of importance since preferential expression of low-binding IL-18BPs could promote IL-18–mediated synovial inflammation.
Not all activities of IL-18 described thus far are of a proinflammatory nature. In particular, IL-18 inhibits osteoclast maturation through granulocyte–macrophage colony-stimulating factor and can suppress COX-2 expression through IFNγ synthesis (14). Available in vivo data suggest, however, that the net effect of IL-18 is proinflammatory, at least in the context of articular inflammation. IL-18 administration promotes development of inflammatory, erosive arthritis in partially primed DBA/1 mice, predominantly through driving TNFα production (15). IL-18–deficient DBA/1 mice exhibit significantly delayed onset and milder severity of collagen-induced arthritis (CIA), characterized by reduced TNFα expression and by lower collagen-specific Th1 responses in vitro (16). Finally, recent reports indicate that antibody-mediated IL-18 neutralization suppresses streptococcal cell wall–induced arthritis through an IFNγ-independent mechanism and that IL-18BP–Fc fusion retards established CIA (17, 18).
High serum levels of IL-18 detected in patients with AOSD, as described by Kawaguchi et al (3), are therefore intriguing. Concentrations reported far exceed those of other inflammatory cytokines, such as IL-6, TNFα, and IFNγ, that have been previously detected in AOSD sera, perhaps indicating a specific relationship between IL-18 production, AOSD pathogenesis, and the systemic nature of the clinical presentation. Certainly, levels are equivalent to those detected in neoplastic and hemophagocytic syndromes, in which cytokine dysregulation has been related to systemic clinical features, such as fever and lymphadenopathy, and to disease activity (19, 20). In AOSD, serum IL-18 levels correlated with clinical outcome defined by response to immunosuppressive treatment. In a subgroup of patients, IL-18 also behaved as a marker of disease activity over time, analogous to ferritin (3). Similar data correlating IL-18 levels, ferritin levels, and disease activity in AOSD have recently been reported (21). Confirmation of these pilot data in larger, prospective clinical studies will be essential.
Elucidating the pathogenic role of IL-18 in AOSD may prove difficult in the absence of a reliable model system, however, although clues should be obtained from pathways in RA in which information is emerging, as described above. Efforts to detect ongoing Th1 responses and NK cell activation would therefore be of interest. IL-18–promoted Fas-mediated apoptosis in the liver of mice has been reported (22). Direct effects of IL-18 in skin and in liver may also be anticipated in AOSD, providing a potential mechanism for the characteristic rash and hepatic disturbance in this disorder. Other questions about the role of IL-18 arise. It will be important to determine the proportion of serum IL-18 in active form and free of IL-18BP interactions, since recent data suggest that IL-18BP activity is increased in AOSD (21). The cellular source of IL-18 synthesis is unclear, although circulating or tissue monocyte/macrophages seem the most likely candidates. Increased levels of TNFα in circulating monocytes have been detected in patients with AOSD (23), and similar studies with regard to IL-18 may prove helpful. In addition, it will be important to determine factors regulating IL-18 release, its precise relationship with acute-phase reactants and other proinflammatory cytokines, and indeed, the primary site of effector function.
IL-18 represents an exciting, novel mediator of inflammation that is clearly up-regulated in a number of autoimmune rheumatic diseases. Several specific biologic agents capable of targeting IL-18 are already in preparation. Although IL-18 exhibits a broad functional profile (Figure 1 and Table 1), most data elucidated thus far indicate that proinflammatory effects predominate, particularly in inflammatory arthritis. Thus, IL-18 likely represents an attractive novel therapeutic target.
|Lymphocyte: Th1 and Th2 maturation and activation|
|Macrophage: cytokine release, NO production, cognate interactions|
|Chondrocyte: GAG release, reduced proliferation, NO production, MMP expression|
|NK cell: cytokine release, cytotoxicity|
|Endothelial cell: promotes angiogenesis|
|Neutrophil: cytokine/chemokine release, respiratory burst, granule release, adhesion molecule up-regulation|
|Osteoclast maturation (via GM-CSF)|