role of proinflammatory cytokines in the pathophysiology of osteoarthritis

Transcription

1 FoCus on osteoarthritis role of proinflammatory cytokines in the pathophysiology of osteoarthritis Mohit Kapoor, Johanne Martel-Pelletier, Daniel Lajeunesse, Jean-Pierre Pelletier and Hassan Fahmi abstract Osteoarthritis (OA) is associated with cartilage destruction, subchondral bone remodeling and inflammation of the synovial membrane, although the etiology and pathogenesis underlying this debilitating disease are poorly understood. Secreted inflammatory molecules, such as proinflammatory cytokines, are among the critical mediators of the disturbed processes implicated in OA pathophysiology. Interleukin (IL)-1β and tumor necrosis factor (TNF), in particular, control the degeneration of articular cartilage matrix, which makes them prime targets for therapeutic strategies. Animal studies provide support for this approach, although only a few clinical studies have investigated the efficacy of blocking these proinflammatory cytokines in the treatment of OA. Apart from IL-1β and TNF, several other cytokines including IL-6, IL-15, IL-17, IL-18, IL-21, leukemia inhibitory factor and IL-8 (a chemokine) have also been shown to be implicated in OA and could possibly be targeted therapeutically. This Review discusses the current knowledge regarding the role of proinflammatory cytokines in the pathophysiology of OA and addresses the potential of anticytokine therapy in the treatment of this disease. Kapoor, M. et al. Nat. Rev. Rheumatol. 7, (2011); published online 30 November 2010; doi: /nrrheum Introduction osteoarthritis (oa), the most common form of arthritis, is a disease that tends to worsen over time as cartilage wears away. although articular cartilage breakdown is a major characteristic of oa, other joint tissues, including the synovial membrane and subchondral bone, actively participate in the progression of the disease. unfortunately, no curative therapeutics (that is, agents that can halt disease progression and reverse any damage) are available for oa. the main goals of current oa therapy are the control of pain and improvement of joint function. Commonly prescribed oa medications include nsaids, analgesic drugs, locally admini stered corticosteroids and viscosupplementation. For many patients, the lack of disease modifying oa drugs results in progressive cartilage damage that eventually ne cessitates surgical intervention. recent progress has considerably improved knowledge of both the factors involved in the development of oa and the mechanisms responsible for its progression. oa is now regarded not as a single disease, but rather as a common final stage of joint failure, the initial stages of which can be triggered and perpetuated by numerous causal factors. according to this view, the disease can be divided into subtypes, some of which might have a more biochemical, inflammatory or genetic signature than others, but all of which eventually converge into the common phenotype. nevertheless, the exact sequence of pathological events in oa remains unclear; the temporal relationship between subchondral bone damage, chronic Competing interests The authors declare no competing interests. inflammation of synovial tissue and cartilage erosion is still largely unknown. the development and progression of oa are now believed to involve inflammation even in the early stages of the disease. 1 epidemiological studies show a clear link between the progression of tibiofemoral cartilage damage and the presence of a reactive or inflammatory syno vium. 2,3 secreted inflammatory factors such as proinflammatory cytokines, therefore, are critical mediators of the disturbed metabolism and enhanced catabolism of joint tissue involved in oa. il 1β, tnf and il 6 seem to be the main proinflammatory cytokines involved in the pathophysiology of oa, although others, including il 15, il 17, il 18, il 21, leukemia inhibitory factor (lif) and chemokines have also been implicated (table 1 and Box 1). this review summarizes the involvement of proinflammatory cytokines in oa patho physiology and focuses mainly on il 1β, tnf and il 6 owing to the magnitude of their involvement in this disease. IL 1β and TNF among the proinflammatory cytokines involved in oa, il 1β and tnf are considered the major players; il 1β seems to be associated with cartilage destruction, and tnf with driving the inflammatory cascade (Figure 1). these two cytokines, which are produced by chondrocytes, mononuclear cells, osteoblasts and synovial tissues, induce the production of a number of inflammatory and catabolic factors. in patients with oa, levels of both il 1β and tnf are elevated in the synovial fluid, synovial membrane, subchondral bone and cartilage. 4 il 1β and tnf can act independently or in concert with other cytokines Osteoarthritis Research Unit, University of Montreal Hospital Research Centre (CRCHUM), Notre-Dame Hospital, 1560 Sherbrooke Street East, Montreal, QC H2L 4M1, Canada (m. Kapoor, J. martel- Pelletier, D. lajeunesse, J.-P. Pelletier, h. Fahmi). Correspondence to: J. Martel-Pelletier jm@martelpelletier.ca nature reviews rheumatology volume 7 JanuarY

2 reviews Key points Osteoarthritis (OA) is associated with cartilage destruction, subchondral bone remodeling and synovial membrane inflammation Proinflammatory cytokines are critical mediators in the disturbed metabolism and enhanced catabolism of tissue in the OA joint Interleukin (IL)-1β, tumor necrosis factor (TNF) and IL-6 seem to be the main proinflammatory cytokines involved in the pathophysiology of OA Data from cellular and animal studies have provided substantial evidence that blocking IL-1β and TNF production could counteract the degradative mechanisms associated with OA pathology Anticytokine therapies in OA clinical trials have so far yielded variable results A better understanding of the individual roles and functions of cytokines, such as IL-1β, TNF and IL-6, is of the utmost importance in order to develop adequate and specific anticytokine therapies Table 1 Cytokines involved in the pathophysiology of OA Cytokine TNF IL-1β activities in oa Elevated in OA synovial fluid, synovial membrane, subchondral bone and cartilage Suppresses the synthesis of proteoglycan, link protein and type II collagen in chondrocytes 18 Stimulates the release of MMP-1, MMP-3 and MMP-13 27,28 Induces the production of IL-6 33 and chemokines such as IL-8, 34 MCP-1 35 and CCL5 (also known as RANTES) 36 Elevated in OA synovial fluid, synovial membrane, subchondral bone and cartilage Suppresses type II collagen 20,21 and aggrecan expression 22 Stimulates the release of MMP-1, MMP-3 and MMP-13 27,28 Induces the production of IL-6 33 and chemokines such as IL-8, 34 MCP-1 35 and CCL5 36 IL-6 Elevated in synovial fluid and sera from patients with OA 81 Upregulates MMP-1 and MMP-13 expression in combination with IL-1β and oncostatin 82,83 Reduces expression of type II collagen 85 IL-15 IL-17 IL-18 IL-21 LIF Chemokines (CCL5, IL-8, GROα and MCP-1) Anti-inflammatory cytokines (IL-4, IL-10 and IL-13) Elevated in synovial fluid in early OA Suggested association with MMP-1 and MMP Induces IL-1β, TNF and IL-6 production Upregulates NO and MMPs 105 Downregulates proteoglycan levels 106 Elevated in human OA chondrocytes Structurally similar to the IL-1 protein family Acts through both IL-1-dependent and IL-1-independent pathways in OA 107 Elevated in synovial fluid in early OA Elevated in OA synovial membrane and fluid Produced in response to IL-1β and TNF Stimulates cartilage proteoglycan degradation, MMP synthesis and NO production 108 Elevated in OA tissues Induce inos, MMP-1 and IL-6 Stimulate proteoglycan depletion 36,108 Elevated in OA tissues Inhibit IL-1β, TNF and proteases Upregulate IL-1Ra and TIMP production 39 Abbreviations: CCL, CC-chemokine ligand; GRO, growth-related oncogene; IL, interleukin; IL-1Ra, IL-1 receptor antagonist; inos, inducible nitric oxide synthase; LIF, leukemia inhibitory factor; MCP, monocyte chemotactic protein; MMP, matrix metalloproteinase; NO, nitric oxide; OA, osteoarthritis; RANTES, regulated upon activation, normal T cell expressed and secreted; TIMP, tissue inhibitor of metalloproteinases; TNF, tumor necrosis factor. to initiate and propagate inflammation. simultaneous injection of il 1β and tnf into rabbit knee joints, for instance, causes more cartilage destruction than either cytokine alone. 5,6 although il 1β is widely accepted as a proinflammatory cytokine, deletion of Il1b accelerates the development of oa lesions in a mouse model of the disease, suggesting a complex role for this cytokine in maintaining cartilage homeostasis. 7 the biological activation of cells by il 1β is mediated through its specific cell surface receptor il 1 receptor type i (il 1ri). 8 expression of this receptor is increased on human oa chondrocytes and synovial fibroblasts in comparison with normal cells. 9,10 il 1β also binds to a second specific cell surface receptor, il 1rii, which is termed a decoy receptor owing to its inability to transduce a signal. 11 another natural inhibitor of il 1 activity, il 1 receptor antagonist (il 1ra), is produced by several cell types including chondrocytes and synovial fibroblasts and possesses anti inflammatory properties. 12,13 il 1ra functions as a competitive antagonist it binds to both il ri and il 1rii but does not transduce a signal. tnf binds to two specific receptors on the cell membrane, tnf receptor i (tnfri or p55) and tnfrii (or p75). tnfri expression is increased in oa chondrocytes and synovial fibroblasts compared with non oa cells. 14 tnfri appears to be the dominant receptor responsible for mediating tnf activity, at least in the articular tissue cells, but both receptor types are suggested to be actively involved in signal transduction and are linked to distinct intracellular protein cascades effects on anabolic and catabolic responses a number of studies demonstrated that il 1β and tnf downregulate the synthesis of major extracellular matrix (ecm) components by inhibiting anabolic activities of chondrocytes. 18,19 in cartilage, highly crosslinked triplehelical type ii collagen is the major component of ecm. in cell culture studies, treatment with il 1β reduces type ii collagen expression in chondrocytes isolated from several different species. 20,21 the expression of aggrecan, a major structural component of cartilage, is also downregulated in il 1β treated chondrocytes 22 and cartilage explants. 23 the suppressive effect of il 1β on proteoglycan synthesis in rat chondrocytes seems to be mediated through suppression of β 1,3 glucuronyltransferase i, the key enzyme in the biosynthesis of glycosaminoglycans linked to aggrecan core proteins. 24 similarly, tnf has been shown to suppress the synthesis of proteoglycan, link protein and type ii collagen in chondrocytes. 18,25 il 1β and tnf stimulate chondrocytes to release several proteolytic enzymes, among which are the matrix metalloproteinases (mmps): mmp 1 (interstitial collagenase), mmp 3 (stromelysin 1) and mmp 13 (collagenase 3) these three proteases are key regulators of cartilage destruction, making anticytokine therapy, especially that which targets il 1β and tnf, an attractive strategy to counteract oa. evidence is accumulating for the importance of the adamts (a disintegrin like and metalloproteinase with thrombospondin type 1 motifs) family of proteins in cartilage degradation in oa, especially adamts 4 and adamts 5. studies in adamts 4 29 and adamts 5 30 knockout mouse models of oa indicate that the latter 34 JANUARY 2011 volume 7

3 FoCus on osteoarthritis might be the predominant adamts involved in oa pathogenesis. a study using chondrocytes and cartilage explants from bovine and porcine species, however, demon strated that adamts 4 is induced upon stimulation with il 1β and tnf, whereas neither of these cytokines altered adamts 5 expression. 31 this finding is consistent with the results of a study of human oa syno vium, in which upregulation of adamts 4, but not adamts 5, was dependent on il 1β and tnf. 32 effects on inflammatory responses in articular cells, il 1β and tnf amplify and perpetuate the oa disease process by inducing the production of proinflammatory cytokines such as il 6 33 and chemo kines such as il 8, 34 monocyte chemoattractant protein 1 35 and CC chemokine ligand 5 (also known as rantes). 36 il 1β and tnf also stimulate the production of a number of other inflammatory mediators implicated in oa pathology. For instance, treatment of chondrocytes with these cytokines upregulates the expression of genes encoding inducible nitric oxide synthase (inos), soluble phospholipase a2, cyclooxygenase 2 (CoX 2) and microsomal prostaglandin e synthase 1, and also stimulates the release of nitric oxide (no) and prostaglandin e 2 (PGe 2 ). no and PGe 2 contribute to articular inflammation and destruction by enhancing the activation and production of mmps, inhibiting the synthesis of anabolic macromolecules such as collagen and proteoglycan, inhibiting the production of il 1ra, and promoting chondrocyte apoptosis. 4 il 1β and tnf also induce the production of reactive oxygen species (ros) mainly no and the superoxide anion that generate hydrogen peroxide, peroxynitrite and hydroxyl radicals, 37 which contribute to cartilage degradation. in addition, il 1β and tnf downregulate the expression of the antioxidant enzymes that scavenge ros, including superoxide dismutase, catalase and glutathione peroxidase, 38,39 thereby accelerating the damaging effects of ros on cartilage. mediators of il-1β and tnf activity the proinflammatory and catabolic effects of il 1β and tnf are mediated through the activation of several signal ing pathways including the c Jun n terminal kinase and p38 mitogen activated protein kinase (also known as extracellular signal regulated kinase) pathways and, most importantly, nuclear factor κb (nfκb) signaling. nfκb mediates the expression of several inflam matory genes such as those that encode inos, CoX 2 and chemokines, and also contributes to the induction of mmp 1, mmp 9, mmp 13 and adamts indeed, adenovirusmediated delivery of small interfering rna specific for the p65 subunit of nfκb reduced disease progression in a rat model of surgically induced oa. 41 the wnt β catenin signaling pathway has been shown to be among the major contributors to oa joint pathology. expression of wnt ligands and β catenin is upregulated in oa cartilage, 42 and activation of wnt β catenin signal ing enhances il 1β mediated matrix catabolism and stimulates mmp and adamts expression in articular cells. 43 elevated levels of the natural wnt antagonists Box 1 Evidence of a role for IL-17 in OA IL-17 is a proinflammatory cytokine thought to have a role in the pathophysiology of OA. IL-17 is secreted primarily by activated memory CD4 + T cells and binds to specific receptors that are expressed by virtually all cells and tissues. CD4 + T cells have been detected in the sublining layer of the synovium of patients with OA, 109 and T-cell reactivity against chondrocyte surface antigens has also been detected in patients with OA. 110 It is still uncertain as to whether T cells produce IL-17 in OA. Although the exact role of IL-17 in the pathophysiology of human OA is unclear, evidence suggests that IL-17 is involved in the early phase of the inflammatory process and induces the release of proinflammatory mediators in a wide range of cell types. In chondrocytes, IL-17 has been shown to induce IL-1β, TNF and IL-6 to upregulate the production of nitric oxide and matrix metalloproteinases and to reduce proteoglycan levels. 105,111,112 IL-17 has been shown to inhibit chondrocyte proteoglycan synthesis in intact murine articular cartilage, which is thought to be mediated by nitric oxide. 106 IL-17 is also known to stimulate the release of the angiogenic factor VEGF in synovial fibroblasts isolated from patients with OA. 113 Future studies need to be directed towards identifying whether IL-17 is actually involved in OA. Abbreviations: IL, interleukin; OA, osteoarthritis; TNF, tumor necrosis factor; VEGF, vascular endothelial growth factor. Inflammatory cells T H cells Intrinsic/external stimuli MMPs ADAMTS-4 Aggrecan Type II collagen Cartilage Chondrocytes Osteoclasts Cytokines (IL-1β, IL-6, IL-17, TNF) Cytokines Bone Cytokines Proteinases Cartilage destruction Osteoblasts such as dickkopf related protein 1 have been associated with reduced progression of hip oa in elderly women. 44 several endogenous molecular mechanisms are also involved in downregulating the effects of il 1β and tnf in oa. among them is the peroxisome proliferator activated receptor γ (PParγ), a ligand activated transcription factor and member of the nuclear receptor superfamily. activation of PParγ results in the downregulation of several inflammatory and catabolic responses mediated by il 1β and tnf in chondrocytes, 45 and a reduction in the progression of cartilage lesions in vivo in experimental oa models. 46,47 Synovium Synovial macrophages Synovial fibroblasts Subchondral bone Figure 1 The role of proinflammatory cytokines in the pathophysiology of OA. The levels of proinflammatory cytokines, including IL-1β, TNF and IL-6, are elevated in OA. These cytokines contribute to the pathogenesis of OA through several mechanisms including downregulation of anabolic events and upregulation of catabolic and inflammatory responses, effects that result in structural damage to the OA joint. Abbreviations: ADAMTS, a disintegrin-like and metalloproteinase with thrombospondin type 1 motifs; IL, interleukin; MMP, matrix metalloproteinase; OA, osteoarthritis; TNF, tumor necrosis factor. nature reviews rheumatology volume 7 JanuarY

4 reviews Table 2 Anticytokine therapies for OA tested in clinical trials or animal studies therapy mode of action main findings Anti-TNF Adalimumab Infliximab Anti-IL-1 Monoclonal anti-tnf antibody Monoclonal anti-tnf antibody In a case study of a single patient, subcutaneously administered adalimumab provided complete relief from nocturnal pain and decreased synovial effusion and synovitis; bone marrow edema was nearly abolished after 6 months of therapy 58 In a pilot clinical trial involving 12 patients, subcutaneous adalimumab was well tolerated but did not improve signs and symptoms of OA 60 In a pilot clinical trial involving 10 patients with erosive hand OA, intra-articular infliximab was well tolerated and reduced OA symptoms at 1 year of follow-up 59 Anakinra Recombinant IL-1Ra In preclinical studies, anakinra protected against development of experimentally-induced OA lesions in dogs 49 In clinical studies, anakinra administered by intra-articular injection was well tolerated in an open-label study, 54 although this treatment was not effective at relieving OA symptoms in a 12-week, randomized, placebo-controlled trial 55 Daily subcutaneous injections of anakinra for 3 months improved pain and global disability in a study of three patients with erosive OA of the hand 56 IL-1Ra gene therapy IL-1Ra gene delivered using retroviral or adenoviral vectors In preclinical studies, IL-1Ra gene therapy delivered by intra-articular injection reduced disease activity and progression in dog, 50 horse, 53 and rabbit 51,52 models of OA Abbreviations: IL, interleukin; IL-1Ra, interleukin-1 receptor antagonist; OA, osteoarthritis; TNF, tumor necrosis factor. anti-il-1β and anti-tnf therapies Data from in vitro human and animal oa explant studies and in vivo animal models of oa provide substantial evidence that blocking il 1β and tnf could be beneficial in counteracting the degradative mechanisms associated with oa pathology. in human oa cartilage, for instance, antagonists of il 1β and tnf (used alone or in combination) downregulate the expression of the major mmps associated with decreased aggrecan and type ii collagen. 48 several approaches might be taken to therapeutically target il 1β and tnf in oa (table 2). Targeting IL-1 directly anakinra, a recombinant form of native il 1ra that has an additional N terminal methionine and lacks glycosylation, inhibited the structural changes associated with oa in a canine model. 49 similarly, gene therapy using retroviral or adenoviral vectors carrying il 1ra was protective in rabbit, dog and horse models of oa in humans, an open label study demonstrated the safety of intra articular injections of anakinra in patients with painful oa. 54 However, a multicenter, randomized, placebo controlled study in patients with symptomatic oa of the knee demon strated that, despite providing slight pain relief 4 days after a single intra articular injection, ana kinra did not relieve symptoms at 12 weeks of follow up. 55 these data do not definitely prove the ineffectiveness of such a drug for oa, as several factors could have negatively influenced the evaluation of treatment response, such as the short half life of anakinra (approximately 4 hours, which prevents its accumulation in the joint) and the study s inclusion cri terion of a low level of pain at baseline. indeed, 3 months of daily sub cutaneous injections of il 1ra improved pain and disability in three patients with erosive oa of the hand. 56 Further studies using different strategies of administration are needed to better define the efficacy of il 1 inhibition in the treatment of oa. another anti il 1 strategy that could be applied to the treatment of oa is the therapeutic utilization of the decoy receptor il 1rii. attur and colleagues 57 have shown that adenovirus mediated overexpression of il 1rii prevents il 1 induced production of no, PGe 2, il 6 and il 8, and decreases proteoglycan synthesis. Targeting TNF directly Biological agents that inhibit the action of tnf, such as infliximab, etanercept and adalimumab, have shown excellent clinical efficacy in preventing the structural damage associated with rheumatoid arthritis (ra). Despite convincing preclinical evidence of the proinflammatory and cartilage destroying effects of tnf in oa, few studies have explored the effect of anti tnf therapy for oa in a clinical setting. in a case study of a 68 year old male with bilateral knee oa, treatment with adalimumab provided complete relief of nocturnal pain after two subcutaneous doses. 58 Furthermore, mri analysis of the patient s target knee after 6 months of therapy demonstrated that synovial effusion and synovitis had decreased and bone marrow edema was nearly abolished. 58 similarly, in an open label pilot trial in 10 patients with erosive oa of the hand, intra articular injections of infliximab reduced pain in all patients, with no reported local or systemic adverse reactions. 59 another small open label study in 12 patients with erosive oa demonstrated that treatment with adalimumab for 3 months did not improve signs and symptoms of erosive oa, although individual patients did experience some benefit. 60 most of the studies described above included a small number of patients and were of short duration (table 2); hence, a clinical trial of longer duration including more patients is required to accurately evaluate the effect and efficacy of tnf blockers in oa treatment. in this context, a double blind, placebo controlled, phase iii clinical trial of adalimumab 36 JANUARY 2011 volume 7

5 FoCus on osteoarthritis in patients (n = 84; 12 month duration) with severe and refractory hand oa is currently ongoing. 61 Targeting MMPs and ADAMTS since il 1β and tnf can increase the levels of several mmps involved in oa pathophysiology, an alternative anticytokine strategy is to inhibit mmp activity. in the past 3 decades, several broad spectrum mmp inhibitors have been tested with little success use of these agents was associated with a variety of adverse effects, such as musculoskeletal syndrome, mild anemia and elevated levels of liver enzymes, probably attributable to the low selectivity of the inhibition. Given the complexity of the pathophysiological processes mediated by various mmps, inhibitors of these proteins with greater selec tivity and specificity are in preclinical or early clinical stages of development. 66 most of these agents are being designed to specifically inhibit mmp 13, which is overexpressed in oa cartilage, while having minimal effects on mmp 1, the suppression of which is considered likely to cause adverse musculoskeletal effects. 67 adamts 4 could be another important target for inhibiting the destructive mechanisms of human oa. some inhibitors of adamts 4 are currently in the developmental phase. 68 Targeting signaling pathways as discussed earlier in this review, the nfκb signaling pathway is a potential target for oa therapy. 35,36 indeed, several pharmacological inhibitors of the nfκb pathway have been developed and have demonstrated protective properties in animal models of ra; 69 however, little is known about their effects in oa. the wnt β catenin system is very complex, but targeting this signaling pathway could be beneficial in counteracting the degrada tive mechanisms associated with oa. strategies to modulate downstream activities of il 1β and tnf via PParγ activation might also represent an approach to oa treatment. IL 6 il 6 signaling involves a multimeric receptor complex com prising a membrane bound il 6 receptor (il 6r), a soluble form of il 6r (sil 6r) and the receptor β subunit gp130. Following ligand induced inter action of this complex, stat1 and stat3 are phosphorylated, triggering their translocation to the nucleus and the subsequent transcription of il 6 target genes. 70,71 Chondrocytes produce low levels of il 6 under normal conditions. However, a number of cytokines and growth factors active in oa, such as il 1β and tgf β, directly stimulate il 6 production. 33,72 in human chondrocytes, PGe 2 also induces il 6 expression by activating nfκb via pathways involving cyclic amp, protein kinase a and phosphatidylinositol 3 kinase. 73 this situation is observed during fluid shear stress of human chondrocytes, which activates CoX 2 and results in the local production of PGe moreover, the direct infusion of high doses of il 6 increased C reactive protein (CrP) levels in healthy human volunteers 75 and in human peripheral blood mononuclear cells in vitro, 76 which has been linked to oa in humans. 77 a clinical trial in patients with oa showed that high baseline levels of il 6 and CrP were associated with an increased risk of cartilage loss (as assessed by mri). 78 Furthermore, high Bmi and increased levels of circulating il 6 have been linked to the risk of developing radiographic knee oa. 79,80 effects on catabolic and anabolic responses the exact role of il 6 in oa is controversial; a number of studies performed in chondrocytes failed to investigate or use sil 6r because it greatly increased the affinity of il 6 for its specific receptor. nevertheless, studies have shown that il 6, in the presence of sil 6r, has a role in inflammatory and degenerative joint diseases. levels of il 6 and sil 6r are increased in synovial fluid and sera from patients with ra and from those with oa. 81 in combination with il 1β and oncostatin, il 6 upregulates mmp 1 and mmp 13 expression in bovine and human cartilage explant cultures. 82,83 in addition, mechanical injury can potentiate the effects of tnf and il 6 sil 6r on proteoglycan degradation, 84 while treatment of chondro cytes with il 6 sil 6r reduces expression of type ii collagen. 85 the effects of il 6 in vivo in animal models are also uncertain. although il 6 deficiency in mice is associated with a decrease in the number of inflammatory cells in knee joints and a reduced response to collagen induced arthritis, 86 injection of il 6 into the joint cavity of il 6 deficient mice reduces cartilage destruction. 87 moreover, proteoglycan synthesis is decreased and the incidence of subchondral bone sclerosis is increased in male il 6 deficient mice compared with wild type animals. 88 interestingly, il 6 is thought to regulate acute phase proteins, which are anti inflammatory and immunosuppressive during the acute phase response; 89 this regulation could contribute to the observed dual effect of il 6 in vivo. effects on subchondral bone tissue Changes in the metabolism of subchondral bone tissue in oa are now considered to be an active component of the disease and not only a secondary manifestation. il 6 is a critical cytokine in this regard, as it triggers osteoclast differen tiation and bone resorption. 90 this effect of il 6 is attributable to an indirect interaction between osteoblasts and osteoclasts, whereby local production of and response to il 6 by osteoblasts stimulates the production of receptor activator of nfκb ligand (also known as osteoclast differentiation factor), il 1β, parathyroid hormone (PtH) related protein (PtHrP) and PGe 2, all of which activate osteoclasts. However, studies have shown that PGe 2 can trigger the production of il 6 by osteoblasts. 91,92 in addition, tnf, il 1β, PtH PtHrP and PGe 2 can all stimulate il 6, il 11 and lif producti on by osteoblasts. interestingly, in human oa, the level of il 6 production by subchondral bone osteoblasts can discriminate two subgroups of patients with oa. in one subgroup, characterized by low levels of il 6 and PGe 2 production, 92 oa is associated with increased bone resorption. 93 nature reviews rheumatology volume 7 JanuarY

6 reviews Table 3 Problems and progress in assessing disease-modifying effects of OA drugs area of concern Problems Progress Imaging Disease heterogeneity Biomarkers Most OA studies have focused on symptomatic patients (those with clinical manifestations such as pain and impaired joint function); structural changes to the articular tissues are already moderate-to-severe in these patients Plain radiography, the reference technique for assessing the severity of cartilage destruction, provides only indirect information about cartilage OA is not a single disease entity but represents different subtypes Variation between OA subtypes could have a bearing on efforts to assess the therapeutic efficacy of new treatments Individually, biomarkers do not provide much information about OA disease processes or about early disease Use of biomarkers in OA clinical trials has been tempered by, among other factors, circadian and diurnal variation in their levels, and by the fact that systemic levels of markers provide information on the overall metabolism of all joints and tissues in the body In advanced OA, little or no cartilage might remain, thus the interpretation of the levels of biomarkers can be difficult Advances in MRI enable quantitative assessment of the structure of articular tissues Reliable, sensitive and accurate MRI methods and systems have been developed to assess in a specific manner joint tissue alterations and intra-individual articular tissue changes associated with OA Identification of specific subsets of patients with varied ranges of cytokine patterns could be beneficial, as these subgroups could respond differently to treatment MRI also enables OA subgroup discrimination; however, whether this discrimination is related to the cytokine network is unknown Biomarkers in urine and serum (such as CTX-I, CTX-II, C2C and CPII) that reflect cartilage metabolic or inflammatory processes have been developed Clusters of markers, as opposed to individual markers, are suggested to correlate well with disease progression and the identification of early disease Abbreviations: CTX-I/II, C-terminal crosslinked telopeptide of type I/type II collagen; C2C, type II collagen neoepitope; CPII, type II procollagen carboxy-propeptide; OA, osteoarthritis. Conversely, in the other subgroup, high levels of il 6 and PGe 2 production 92 are associated with reduced bone resorptive properties. 93 expression of il 6 mrna and protein is increased approximately fivefold in trabecular bone osteoblasts from patients with oa compared with normal osteoblasts. 94 the role of il 6 in oa bone tissue might seem contradictory since il 6 should trigger osteoclast and bone resorption, whereas bone accretion and sclerosis of subchondral bone and osteophyte formation are observed late in the disease process. early oa might be linked with an initial phase of bone resorption that is then followed by one of increased bone formation. anti-il-6 therapy to our knowledge, no clinical studies have evaluated the effect of il 6 or il 6r inhibition in oa. However, data from studies in animal models of arthritis, as well as from clinical trials conducted in patients with ra, are encourag ing. in a mouse model of postmenopausal polyarthritis, reducing il 6 levels with raloxifene reduced the severity and frequency of arthritis while protecting the joints from destruction and the development of osteoporosis. 95 the humanized anti il 6r antibody tocilizumab improved disease parameters and quality of life, as well as inflammatory markers such as CrP, compared with placebo in phase iii trials of patients with ra. 96,97 Furthermore, tocilizumab in combination with methotrexate was shown to improve markers of bone and cartilage turnover in patients with ra. 98 although these results are encouraging, targeting il 6 for the treatment of oa needs further investigation in preclinical and clinical studies. as subchondral bone resorption appears to be involved in the pathophysiology of oa, the application of antiresorptive therapies could be considered. indeed, in rabbit and rat models of oa, antiresorptive therapy with risedronate, with or without concomitant nsaids, did not prevent inflammation or osteophyte formation, although this treatment did improve bone mineral density and joint laxity. 99,100 in human knee oa, oral risedronate failed to show any disease modifying effects, as assessed by radio graphy in a phase iii study, although this compound seems to reduce the level of the C terminal cross linking telopeptide of type ii collagen, a marker of cartilage degradation in humans. 101,102 there are several potential explanations for the failure of such trials. one reason might be that, for antiresorptive treatment to be effective in oa, the drug should be administered during the early stages of the disease. moreover, the imaging technology used in these investigations (radiography) might not be optimal for assessing a possible disease modifying effect (table 3). Conclusions over the years, it has become evident that the inflammatory cytokine network contributes substantially to the pathogenesis of oa. increased expression of proinflammatory cytokines in cartilage, synovial membrane and subchondral bone are believed to be linked to the develop ment and progression of structural changes in the oa joint. refinement of our understanding of the individual roles and functions of cytokines in particular il 1β, tnf and il 6 in oa pathophysiology is of the utmost importance. indeed, this knowledge will be instrumental in dissecting the underlying disease mechanisms in order to more effectively interfere with the dysregulated cytokine network. nonetheless, compounds that regulate cytokine synthesis or activity, or both, are favorable targets for oa therapy. Broader approaches 38 JANUARY 2011 volume 7

7 FoCus on osteoarthritis include activating or increasing the level of factors that inhibit these cytokines or their catabolic activities. the use of anti inflammatory drugs directed at cytokines to treat oa and prevent the progression of structural changes to the joint is based on a sound rationale. However, the results of clinical trials conducted to date, which have mostly studied the effect of anti il 1 therapy on oa symptoms, have been disappointing. Different strategies for drug delivery might be needed in order to maintain an optimum therapeutic level of the drug in the joint and, thereby, enable investigation of a potential disease modifying effect. moreover, some clinical trials using anticytokine drugs were stopped because of toxicity; however, it should be kept in mind that the duration of administration would be time limited if a drug effectively reduced or stopped the disease process. Clinical trials of anticytokine agents for the treatment of oa should investigate potential structure modifying effects with reliable and sensitive imaging technology such as mri. one such clinical trial using an anti inflammatory drug (licofelone) on patients with sympto matic knee oa has already been performed and showed that the treatment considerably reduced the cartilage volume loss, constituting a protective effect in these patients. 103 with ongoing advancements in the fields of imaging, drug delivery systems and molecular tools, the overall goal is to clearly dissect the role of cytokines in the patho physiology of oa and develop potent and specific anticytokine treatments in the near future. Review criteria This Review was prepared by a thorough search of the PubMed and ClinicalTrials databases for pertinent literature on the cytokines implicated in osteoarthritis. The following keywords were used alone or in combination: osteoarthritis, cytokines, proinflammatory cytokines, interleukin-1, tumor necrosis factor-alpha, inflammation, cartilage, chondrocytes, animal models of osteoarthritis and knockout mice. The search was restricted to English-language full-length articles only. The search criteria were not restricted by date. 1. Felson, D. T. Clinical practice. Osteoarthritis of the knee. N. Engl. J. Med. 354, (2006). 2. Ayral, X., Pickering, E. H., Woodworth, T. G., Mackillop, N. & Dougados, M. Synovitis: a potential predictive factor of structural progression of medial tibiofemoral knee osteoarthritis results of a 1 year longitudinal arthroscopic study in 422 patients. Osteoarthritis Cartilage 13, (2005). 3. Pozgan, U. et al. Expression and activity profiling of selected cysteine cathepsins and matrix metalloproteinases in synovial fluids from patients with rheumatoid arthritis and osteoarthritis. Biol. Chem. 391, (2010). 4. Martel-Pelletier, J., Lajeunesse, D. & Pelletier, J. P. in Arthritis and Allied Conditions. A Textbook of Rheumatology 15 th edn (eds Koopman, W. J. & Moreland, L. W.) (Lippincott Williams & Wilkins, Baltimore, USA, 2005). 5. Page Thomas, D. P., King, B., Stephens, T. & Dingle, J. T. In vivo studies of cartilage regeneration after damage induced by catabolin/interleukin-1. Ann. Rheum. Dis. 50, (1991). 6. Henderson, B. & Pettipher, E. R. Arthritogenic actions of recombinant IL-1 and tumour necrosis factor α in the rabbit: evidence for synergistic interactions between cytokines in vivo. Clin. Exp. Immunol. 75, (1989). 7. Clements, K. M. et al. Gene deletion of either interleukin-1β, interleukin-1β-converting enzyme, inducible nitric oxide synthase, or stromelysin 1 accelerates the development of knee osteoarthritis in mice after surgical transection of the medial collateral ligament and partial medial meniscectomy. Arthritis Rheum. 48, (2003). 8. Arend, W. P. Interleukin-1 receptor antagonist. Adv. Immunol. 54, (1993). 9. Martel-Pelletier, J. et al. The interleukin-1 receptor in normal and osteoarthritic human articular chondrocytes. Identification as the type I receptor and analysis of binding kinetics and biologic function. Arthritis Rheum. 35, (1992). 10. Sadouk, M. B. et al. Human synovial fibroblasts coexpress IL-1 receptor type I and type II mrna. The increased level of the IL-1 receptor in osteoarthritic cells is related to an increased level of the type I receptor. Lab. Invest. 73, (1995). 11. Colotta, F., Dower, S. K., Sims, J. E. & Mantovani, A. The type II decoy receptor: a novel regulatory pathway for interleukin 1. Immunol. Today 15, (1994). 12. Palmer, G. et al. Production of interleukin-1 receptor antagonist by human articular chondrocytes. Arthritis Res. 4, (2002). 13. Seitz, M. et al. Production of interleukin-1 receptor antagonist, inflammatory chemotactic proteins, and prostaglandin E by rheumatoid and osteoarthritic synoviocytes regulation by IFN-γ and IL-4. J. Immunol. 152, (1994). 14. Alaaeddine, N. et al. Osteoarthritic synovial fibroblasts possess an increased level of tumor necrosis factor-receptor 55 (TNF-R55) that mediates biological activation by TNFα. J. Rheumatol. 24, (1997). 15. Naume, B., Shalaby, R., Lesslauer, W. & Espevik, T. Involvement of the 55- and 75-kDa tumor necrosis factor receptors in the generation of lymphokine-activated killer cell activity and proliferation of natural killer cells. J. Immunol. 146, (1991). 16. Hohmann, H. P. et al. Expression of the types A and B tumor necrosis factor (TNF) receptors is independently regulated, and both receptors mediate activation of the transcription factor NF-κB. TNFα is not needed for induction of a biological effect via TNF receptors. J. Biol. Chem. 265, (1990). 17. Westacott, C. I., Atkins, R. M., Dieppe, P. A. & Elson, C. J. Tumor necrosis factor α receptor expression on chondrocytes isolated from human articular cartilage. J. Rheumatol. 21, (1994). 18. Saklatvala, J. Tumour necrosis factor α stimulates resorption and inhibits synthesis of proteoglycan in cartilage. Nature 322, (1986). 19. Goldring, M. B., Fukuo, K., Birkhead, J. R., Dudek, E. & Sandell, L. J. Transcriptional suppression by interleukin-1 and interferon-γ of type II collagen gene expression in human chondrocytes. J. Cell. Biochem. 54, (1994). 20. Chadjichristos, C. et al. Sp1 and Sp3 transcription factors mediate interleukin-1β down-regulation of human type II collagen gene expression in articular chondrocytes. J. Biol. Chem. 278, (2003). 21. Shakibaei, M., Schulze-Tanzil, G., John, T. & Mobasheri, A. Curcumin protects human chondrocytes from IL-l1β -induced inhibition of collagen type II and β1-integrin expression and activation of caspase-3: an immunomorphological study. Ann. Anat. 187, (2005). 22. Stöve, J., Huch, K., Günther, K. P. & Scharf, H. P. Interleukin-1β induces different gene expression of stromelysin, aggrecan and tumor-necrosis-factor-stimulated gene 6 in human osteoarthritic chondrocytes in vitro. Pathobiology 68, (2000). 23. Nietfeld, J. J., Wilbrink, B., Den Otter, W., Huber, J. & Huber-Bruning, O. The effect of human interleukin 1 on proteoglycan metabolism in human and porcine cartilage explants. J. Rheumatol. 17, (1990). 24. Gouze, J. N. et al. Interleukin-1β down-regulates the expression of glucuronosyltransferase I, a key enzyme priming glycosaminoglycan biosynthesis: influence of glucosamine on interleukin-1β-mediated effects in rat chondrocytes. Arthritis Rheum. 44, (2001). 25. Séguin, C. A. & Bernier, S. M. TNFα suppresses link protein and type II collagen expression in chondrocytes: Role of MEK1/2 and NF-κB signaling pathways. J. Cell. Physiol. 197, (2003). 26. Mengshol, J. A., Vincenti, M. P., Coon, C. I., Barchowsky, A. & Brinckerhoff, C. E. Interleukin-1 induction of collagenase 3 (matrix metalloproteinase 13) gene expression in chondrocytes requires p38, c-jun N-terminal kinase, and nuclear factor κb: differential regulation of collagenase 1 and collagenase 3. Arthritis Rheum. 43, (2000). 27. Lefebvre, V., Peeters-Joris, C. & Vaes, G. Modulation by interleukin 1 and tumor necrosis nature reviews rheumatology volume 7 JanuarY

8 reviews factor α of production of collagenase, tissue inhibitor of metalloproteinases and collagen types in differentiated and dedifferentiated articular chondrocytes. Biochim. Biophys. Acta 1052, (1990). 28. Reboul, P., Pelletier, J. P., Tardif, G., Cloutier, J. M. & Martel-Pelletier, J. The new collagenase, collagenase-3, is expressed and synthesized by human chondrocytes but not by synoviocytes. A role in osteoarthritis. J. Clin. Invest. 97, (1996). 29. Glasson, S. S. et al. Characterization of and osteoarthritis susceptibility in ADAMTS-4- knockout mice. Arthritis Rheum. 50, (2004). 30. Glasson, S. S. et al. Deletion of active ADAMTS5 prevents cartilage degradation in a murine model of osteoarthritis. Nature 434, (2005). 31. Tortorella, M. D., Malfait, A. M., Deccico, C. & Arner, E. The role of ADAM-TS4 (aggrecanase-1) and ADAM-TS5 (aggrecanase-2) in a model of cartilage degradation. Osteoarthritis Cartilage 9, (2001). 32. Bondeson, J., Wainwright, S. D., Lauder, S., Amos, N. & Hughes, C. E. The role of synovial macrophages and macrophage-produced cytokines in driving aggrecanases, matrix metalloproteinases, and other destructive and inflammatory responses in osteoarthritis. Arthritis Res. Ther. 8, R187 (2006). 33. Guerne, P. A., Carson, D. A. & Lotz, M. IL-6 production by human articular chondrocytes. Modulation of its synthesis by cytokines, growth factors, and hormones in vitro. J. Immunol. 144, (1990). 34. Lotz, M., Terkeltaub, R. & Villiger, P. M. Cartilage and joint inflammation. Regulation of IL-8 expression by human articular chondrocytes. J. Immunol. 148, (1992). 35. Villiger, P. M., Terkeltaub, R. & Lotz, M. Monocyte chemoattractant protein-1 (MCP-1) expression in human articular cartilage. Induction by peptide regulatory factors and differential effects of dexamethasone and retinoic acid. J. Clin. Invest. 90, (1992). 36. Alaaeddine, N., Olee, T., Hashimoto, S., Creighton-Achermann, L. & Lotz, M. Production of the chemokine RANTES by articular chondrocytes and role in cartilage degradation. Arthritis Rheum. 44, (2001). 37. Afonso, V., Champy, R., Mitrovic, D., Collin, P. & Lomri, A. Reactive oxygen species and superoxide dismutases: role in joint diseases. Joint Bone Spine 74, (2007). 38. Mathy-Hartert, M. et al. Interleukin-1β and interleukin-6 disturb the antioxidant enzyme system in bovine chondrocytes: a possible explanation for oxidative stress generation. Osteoarthritis Cartilage 16, (2008). 39. Scott, J. L. et al. Superoxide dismutase downregulation in osteoarthritis progression and end-stage disease. Ann. Rheum. Dis. 69, (2010). 40. Roman-Blas, J. A. & Jimenez, S. A. NF-κB as a potential therapeutic target in osteoarthritis and rheumatoid arthritis. Osteoarthritis Cartilage 14, (2006). 41. Chen, L. X. et al. Suppression of early experimental osteoarthritis by in vivo delivery of the adenoviral vector-mediated NF-κBp65- specific sirna. Osteoarthritis Cartilage 16, (2008). 42. Blom, A. B. et al. Involvement of the Wnt signaling pathway in experimental and human osteoarthritis: prominent role of Wnt-induced signaling protein 1. Arthritis Rheum. 60, (2009). 43. Yuasa, T., Otani, T., Koike, T., Iwamoto, M. & Enomoto-Iwamoto, M. Wnt/β-catenin signaling stimulates matrix catabolic genes and activity in articular chondrocytes: its possible role in joint degeneration. Lab. Invest. 88, (2008). 44. Lane, N. E., Nevitt, M. C., Lui, L. Y., de Leon, P. & Corr, M. Wnt signaling antagonists are potential prognostic biomarkers for the progression of radiographic hip osteoarthritis in elderly Caucasian women. Arthritis Rheum. 56, (2007). 45. Fahmi, H., Pelletier, J. P. & Martel-Pelletier, J. PPARg ligands as modulators of inflammatory and catabolic responses in arthritis. An overview. J. Rheumatol. 29, 3 14 (2002). 46. Kobayashi, T. et al. Pioglitazone, a peroxisome proliferator-activated receptor γ agonist, reduces the progression of experimental osteoarthritis in guinea pigs. Arthritis Rheum. 52, (2005). 47. Boileau, C. et al. The peroxisome proliferatoractivated receptor γ agonist pioglitazone reduces the development of cartilage lesions in an experimental dog model of osteoarthritis: in vivo protective effects mediated through the inhibition of key signaling and catabolic pathways. Arthritis Rheum. 56, (2007). 48. Kobayashi, M. et al. Role of interleukin-1 and tumor necrosis factor α in matrix degradation of human osteoarthritic cartilage. Arthritis Rheum. 52, (2005). 49. Caron, J. P. et al. Chondroprotective effect of intraarticular injections of interleukin-1 receptor antagonist in experimental osteoarthritis. Suppression of collagenase-1 expression. Arthritis Rheum. 39, (1996). 50. Pelletier, J. P. et al. In vivo suppression of early experimental osteoarthritis by interleukin-1 receptor antagonist using gene therapy. Arthritis Rheum. 40, (1997). 51. Fernandes, J. et al. In vivo transfer of interleukin-1 receptor antagonist gene in osteoarthritic rabbit knee joints: prevention of osteoarthritis progression. Am. J. Pathol. 154, (1999). 52. Zhang, X., Mao, Z. & Yu, C. Suppression of early experimental osteoarthritis by gene transfer of interleukin-1 receptor antagonist and interleukin-10. J. Orthop. Res. 22, (2004). 53. Frisbie, D. D., Ghivizzani, S. C., Robbins, P. D., Evans, C. H. & McIlwraith, C. W. Treatment of experimental equine osteoarthritis by in vivo delivery of the equine interleukin-1 receptor antagonist gene. Gene Ther. 9, (2002). 54. Chevalier, X. et al. Safety study of intraarticular injection of interleukin 1 receptor antagonist in patients with painful knee osteoarthritis: a multicenter study. J. Rheumatol. 32, (2005). 55. Chevalier, X. et al. Intraarticular injection of anakinra in osteoarthritis of the knee: a multicenter, randomized, double-blind, placebo-controlled study. Arthritis Rheum. 61, (2009). 56. Bacconnier, L., Jorgensen, C. & Fabre, S. Erosive osteoarthritis of the hand: clinical experience with anakinra. Ann. Rheum. Dis. 68, (2009). 57. Attur, M. G. et al. Functional genomic analysis of type II IL-1β decoy receptor: potential for gene therapy in human arthritis and inflammation. J. Immunol. 168, (2002). 58. Grunke, M. & Schulze-Koops, H. Successful treatment of inflammatory knee osteoarthritis with tumour necrosis factor blockade. Ann. Rheum. Dis. 65, (2006). 59. Fioravanti, A., Fabbroni, M., Cerase, A. & Galeazzi, M. Treatment of erosive osteoarthritis of the hands by intra-articular infliximab injections: a pilot study. Rheumatol. Int. 29, (2009). 60. Magnano, M. D. et al. A pilot study of tumor necrosis factor inhibition in erosive/ inflammatory osteoarthritis of the hands. J. Rheumatol. 34, (2007). 61. Efficacy study of an anti-tumor necrosis factor (TNF)-α agent in patients with hand osteoarthritis (DORA). [online], NCT (2010). 62. Skiles, J. W., Gonnella, N. C. & Jeng, A. Y. The design, structure, and clinical update of small molecular weight matrix metalloproteinase inhibitors. Curr. Med. Chem. 11, (2004). 63. Wojtowicz-Praga, S. et al. Phase I trial of Marimastat, a novel matrix metalloproteinase inhibitor, administered orally to patients with advanced lung cancer. J. Clin. Oncol. 16, (1998). 64. Murphy, G. & Nagase, H. Reappraising metalloproteinases in rheumatoid arthritis and osteoarthritis: destruction or repair? Nat. Clin. Pract. Rheumatol. 4, (2008). 65. Krzeski, P. et al. Development of musculoskeletal toxicity without clear benefit after administration of PG , a matrix metalloproteinase inhibitor, to patients with knee osteoarthritis: a randomized, 12-month, double-blind, placebo-controlled study. Arthritis Res. Ther. 9, R109 (2007). 66. Clutterbuck, A. L., Asplin, K. E., Harris, P., Allaway, D. & Mobasheri, A. Targeting matrix metalloproteinases in inflammatory conditions. Curr. Drug Targets 10, (2009). 67. Johnson, A. R. et al. Discovery and characterization of a novel inhibitor of matrix metalloprotease-13 that reduces cartilage damage in vivo without joint fibroplasia side effects. J. Biol. Chem. 282, (2007). 68. Tortorella, M. D., Malfait, F., Barve, R. A., Shieh, H. S. & Malfait, A. M. A review of the ADAMTS family, pharmaceutical targets of the future. Curr. Pharm. Des. 15, (2009). 69. Marcu, K. B., Otero, M., Olivotto, E., Borzi, R. M. & Goldring, M. B. NF-κB signaling: multiple angles to target OA. Curr. Drug Targets 11, (2010). 70. Wegenka, U. M., Buschmann, J., Lütticken, C., Heinrich, P. C. & Horn, F. Acute-phase response factor, a nuclear factor binding to acute-phase response elements, is rapidly activated by interleukin-6 at the posttranslational level. Mol. Cell. Biol. 13, (1993). 71. Zhong, Z., Wen, Z. & Darnell, J. E. Jr. Stat3: a STAT family member activated by tyrosine phosphorylation in response to epidermal growth factor and interleukin-6. Science 264, (1994). 72. Bender, S. et al. Interleukin-1β induces synthesis and secretion of interleukin-6 in human chondrocytes. FEBS Lett. 263, (1990). 73. Wang, P., Zhu, F. & Konstantopoulos, K. Prostaglandin E2 induces interleukin-6 expression in human chondrocytes via camp/ protein kinase A- and phosphatidylinositol 3-kinase-dependent NF-κB activation. Am. J. Physiol. Cell Physiol. 298, C1445 C1456 (2010). 40 JANUARY 2011 volume 7