Inflammatory arthritis ‐ an exciting confluence of human and animal research

Abstract

The years 2002–2011 in the United States have been declared the Bone and Joint Decade. The mission of the United States Bone and Joint Decade (USBJD) is ‘to promote and facilitate collaboration among the public, patients, and organizations to improve bone and joint health through education, research, and advocacy.’ At least 30% of Americans require treatment for a musculoskeletal disease, and the estimated US cost for treatment of all persons with a musculoskeletal disease diagnosis and indirect lost wages was $849 billion annually for the years 2002 to 2004, 7.7% of the gross domestic product (1). Arthritis includes both non-inflammatory osteoarthritis and inflammatory diseases such as rheumatoid arthritis and spondyloarthropathies. Inflammatory arthritides are characterized pathologically by changes in synovium, cartilage, and bone induced by activation of the immune system. These histologic changes are the consequence of disease states as varied as chronic autoimmune diseases like rheumatoid arthritis (RA) and ankylosing spondylitis, autoinflammatory crystal-induced diseases like gout, and infectious arthritis. In this volume of Immunological Reviews, we focus on the most recent research into the pathophysiology of autoimmune and crystal-induced arthritis. There has been an explosion in our understanding of the etiology and pathogenesis of inflammatory arthritis in recent years, and, although our treatments are much improved, we have no ability to prevent these diseases. Explorations of the genetic, cellular, and molecular underpinnings of pathology have been fruitful avenues of investigation. Finally, research into the mechanisms of the inflammatory arthritides presents a wonderful paradigm for the convergence of clinical and basic studies in both humans and mice. RA is the paradigmatic autoimmune arthritis and, as with most autoimmune diseases, is induced by a combination of genetic, environmental, and stochastic mechanisms. Dysregulation of the immune response involves innate immune cells including neutrophils, macrophages, and antigen-presenting cells, as well as loss of tolerance by both T cells and B cells. The marked response to blockade of tumor necrosis factor (TNF) in patients pushed researchers to focus on the innate immune response and the function of macrophages in RA. However, similar to most autoimmune diseases, the major genetic susceptibility alleles for RA are the human leukocyte antigen (HLA) class II genes. Additionally, the most sensitive diagnostic test for RA is the presence in the serum of autoantibodies directed against citrullinated proteins (2, 3). Taken together, these findings should encourage researchers to re-examine models of RA in which the adaptive immune response and specifically the function of CD4+ T cells is central to pathogenesis. Significant progress has been made in the last several years in the identification of the genes that contribute to RA susceptibility (4). The importance of genes such as CD40, TRAF1 (TNF receptor associated factor 1), and TNFAIP3 (TNFα-induced protein 3) suggest a model of RA centered on activation of nuclear factor-κB (NF-κB)-dependent inflammatory pathways. Nevertheless, researchers are still trying to understand the mechanism for the major correlation between susceptibility HLA class II genes alleles and RA, an association established more than 20 years ago. Susceptibility to RA is increased in HLA-DR1 and -DR4-positive patients. Provocative recent data show that the initial identification of the HLA-DR locus was incomplete. But more detailed analyses suggest that disease is also affected by HLA-DQ genes that are in linkage disequilibrium with the DR locus. Chella David and his colleagues (5) have developed transgenic mice expressing human rather than mice MHC class II molecules that model the class II-dependent loss of T-cell tolerance that may initiate disease in RA. Are there other murine models in which disrupted T-cell tolerance leads to inflammatory arthritis? David and his colleagues provide data that susceptible MHC alleles alter the T-cell repertoire, either altering effector cells or by failing to select a robust repertoire of regulatory T cells. Pernis and her group (6) describe the spontaneous development of inflammatory arthritis that develops in mice when the T-cell receptor threshold for selection in the thymus is disrupted, leading to altered effector cells. Additionally, previous observations have suggested that regulatory T cells are dysfunctional in untreated RA patients (7). Yet Caton and his colleagues (8) describe a model of spontaneous inflammatory arthritis that develops despite the presence of normal, functional regulatory T cells. Future research will need to dissect the seesaw relationship between effector T cells and regulatory T cells in RA, an intriguing therapeutic axis. Separately, the presence of autoantibodies—both rheumatoid factors and antibodies to citrullinated proteins—demonstrate the loss of B-cell tolerance present in RA. Duskin and Eisenberg (9) survey the data hypothesizing pathologic function for autoantibodies in mouse and human inflammatory arthritis. Two additional articles in the current issue (2, 3) examine the development of anti-CCP (anti-cyclic citrullinated peptide) autoantibodies. The predisposition to citrullination explains some of the genetic and environmental susceptibility to RA. DR4-positivity does associate with anti-CCP-positive disease, and smoking—perhaps through increased inflammation—increases the rate of citrullination. Both reviews (2, 3) link the etiology of the inflammatory arthritis in RA to subclinical bacterial infections. Thus, RA may be the first disease in which researchers concretely draw the long sought link between infection and autoimmunity. The development of autoantibodies to a broad array of citrullinated proteins may result from intermolecular epitope spreading of reactivity to post-translationally modified extracellular matrix proteins? Utz and his colleagues (10) have modeled the autoimmune response to spliceosome proteins that develop in mixed connective tissue disease and lupus as a model system for understanding the development of autoantibodies. Enhancing the function of regulatory T cells provides an attractive target for treating inflammatory arthritis. Similarly, Fillatreau and his colleagues (11) propose that tolerance is additionally maintained by interleukin-10 (IL-10)-secreting regulatory B cells. This subpopulation should provide an additional therapeutic target. The susceptibility to RA mediated by MHC class II genes presumably occurs through changes in the repertoire or function of CD4+ T cells. This finding is in contrast with recent research into the pathogenesis of ankylosing spondylitis. Ankylosing spondylitis was the first disease linked to an HLA gene—the MHC class I molecule HLA-B27. Yet, our understanding of the mechanism has strayed away from presentation of arthritogenic peptides to MHC class I-restricted CD8+ T cells. The genetics of ankylosing spondylitis is strikingly similar to inflammatory bowel disease and highlights a role for the innate immune system (12). How might an innate immune response cause disease associated with an MHC class I molecule? Robert Colbert et al. (13) have suggested that ‘immunologic recognition of HLA-B27 might not be necessary to cause disease if intracellular effects are paramount.’ Instead, human, rodent, and in vitro studies suggest that the cellular and molecular biology of the HLA-B27 heavy chain polypeptide might differ from other class I molecules and initiate an unfolded protein response in the endoplasmic reticulum. In this model, ankylosing spondylitis might be an autoinflammatory disease operating through pathways similar to those induced by uric acid crystals. Uric acid activates the innate immune system and, at excess concentration, causes gout. Gout is an autoinflammatory disease; spontaneous episodes of inflammation occur with no autoantibodies and no antigen-specific T cells. In the normal setting, uric acid may be a major signal to the immune system for the presence of dying cells (14). Uric acid utilizes multiple pathways to activate macrophages. Shi (15) has shown that uric acid crystals signal by altering membrane lipid structures. Martinon (16) discusses the activation of the NALP3 inflammasome and IL-1 mobilization. Thus, similar to autoinflammatory diseases like Muckle–Wells syndrome (17–19), gout should respond to IL-1 blockade (20). Gout then is a prime example of rapid translation of basic observations in the laboratory into therapeutic advances. The study of the effector cells in the joint should lead to new targets to prevent secondary effects of inflammatory arthritis. The clinical presentation of inflammatory arthritis requires inappropriate activation of lymphocytes, neutrophils, and macrophages in the joint. Although RA and ankylosing spondylitis are systemic autoimmune diseases, with extra-articular manifestations, much of the pathology is confined to the synovium and adjacent bone and cartilage. Indeed, the long-term joint destruction and disability in the inflammatory arthritides result from synovial and bone remodeling. Thus, the final set of articles in this issue focus on the joint and bone. Normal synovium, consisting of fibroblast-like synoviocytes and macrophages, is a layer one to three cells thick. Over the course of RA, the synovium is infiltrated by immune cells and becomes hyperplastic pannus that invades underlying cartilage and bone. Pannus develops in response to cytokines secreted by infiltrating immune cells. However, fibroblast-like Type B synoviocytes also contribute to pannus formation by producing pro-inflammatory cytokines and by adopting an aggressive growth phenotype reminiscent of transformed cells (21). Are there synovium-specific targets for pharmacologic intervention? Brenner and his group (22) previously showed that the integral membrane glycoprotein, cadherin-11, regulates cell–cell adhesion between fibroblast-like synoviocytes during development and pannus formation. Indeed, antibody-mediated inflammatory arthritis is attenuated in mice treated with anti-cadherin-11 antibodies (23). Targeting the synovium provides an interesting alternative to current treatment modalities directed towards immune activation. Changes in the synovium and bone are mediated by lymphocyte activation. However, Pitzalis and colleagues (24) discuss evidence that there is two-way communication between the adaptive immune system and the joint. They describe the mechanisms that regulate expansion of the lymph nodes that drain joints involved in RA and the development of tertiary lymphoid structures in the joints themselves. Thus, T and B cells and macrophages induce synovial proliferation and osteoclast activation. In turn, the inflamed joint nurtures the lymphoid environment. Osteoclast activation and focal bone loss characterize RA. The development and activation of both osteoclasts and osteoblasts is regulated by the nuclear factor of activated T cells (NFAT) family of transcription factors (25). Walsh and Gravallese (26) ask why the cellular and cytokine networks in RA lead to osteoclast activation, while osteoblasts are overactive in spondyloarthropathies in the setting of osteoclast repression. Taken together, this series of reviews highlights the genetic pathways, environmental events, and immunologic and target organ cellular regulation of at least three different forms of inflammatory arthritis. We are hopeful that better understanding of these pathways will lead to therapeutic targets regulating the immune system as well as effector cells in the target organs.