IL-17 Biology in RA

In the past few years, there has been increasing evidence that cytokines of the IL-17 family (and, in particular, IL-17A produced by the T helper (Th17)-cell axis) play an important role in RA pathogenesis. Initially, the cDNA of human IL-17 was cloned from a library of CD4 + T cells and stimulation of peripheral blood T cells was shown to induce IL-17 . IL-17 stimulates epithelial, endothelial and fibroblast type cells to produce cytokines such as IL-6 and IL-8, together with matrix metalloproteinase (MMP)-type moieties . The cellular sources of IL-17 are now broadening – originally, it was considered a T helper cell product but several studies now reveal expression across a range of cells including macrophages, neutrophils and, potentially, mast cells. Interestingly, the differentiation of Th17 cells is different in mouse and human – this has considerable importance in the interpretation of murine models of arthritis in which IL-17 is dominant. Whereas Th17 induction in mice relies on IL-6, transforming growth factor (TGF)-beta and IL-23 (the latter for maintenance of Th17 cells) , for human T cells, IL-21 and TGF-β seem to be important . However, this is still in debate and reflects the complexity of the T-cell axis. In 1999, IL-17 was described in RA synovium and synovial fluid. Despite the relatively low-level expression in fluid and comparatively few positive cells detected in tissues, interest in IL-17 increased based mainly on mouse models with a focus on the Th17 axis . Collagen-induced arthritis (CIA) in a mouse lacking IL-17 demonstrated suppression of induction and amelioration of disease progression . This was supported by antibody studies blocking IL-17 in arthritis mouse models and using mice deficient in the IL-17 receptor . A different chronic relapsing arthritis model showed that the combined blockade of IL-17 with granulocyte/macrophage colony stimulating factor (GM-CSF) also works in a TNF-independent model . More interestingly, TNF blockade stimulates an increase in circulating Th17 and Th1 cells, through preventing their accumulation in the joint . Finally, IL-17 promotes erosions by inducing pro-erosive cytokines and by up-regulation of receptor activator of NFκB ligand (RANKL) on osteoblasts and stimulating an osteoclast differentiating environment . Due to the broad pro-inflammatory effects of IL-17 in RA, blockade of this cytokine seems reasonable. Successful treatment of psoriasis with anti-IL-17 antibodies supports the achievability in IL-17-driven auto-immune diseases . Two phase I studies have been reported to show benefit compared with placebo in RA – in one the rate of response was low although differentiated from placebo. In a second trial using a discrete antibody against IL-17A, the ACR20 response rate was high but confounded by rather high placebo rates – ongoing phase II programmes will clarify the utility of IL-17 in due course as a target.

IL-23/IL-12p40 Biology in RA

The cytokine IL-23 is a heterodimeric cytokine composed of a p19 and a p40 subunit; the latter is shared with IL-12 (p35/p40), which drives naïve T cells into Th1 cells. IL-23 is produced by activated dendritic cells and macrophages and seems to be important in induction and maintenance of Th17 cells. Expression in RA synovial fluid is controversial with groups reporting low and others very high levels. Interestingly, in synovial tissue, bioactive IL-23 seems to be present only in low levels despite high expression of IL-23p19 . Regarding bone homeostasis, IL-23 up-regulates RANK on myeloid precursors and induces RANKL expression on CD4 + T cells . This direct effect, combined with an indirect effect via IL-17, strongly supports osteoclast differentiation and further erosive disease. Recent genetic studies in psoriasis patients provide strong support for an association of IL-23 related loci with the involvement of single nucleotide polymorphism for IL-23A, IL-12B and IL-23R . So far, these variants have not been observed in RA.

Due to the shared subunits, interpretation of models conducted in mice deficient in p40 is difficult as it lacks both IL-12 and IL-23. To test the effect of IL-23 in auto-immune disease, anti-IL-23 antibodies were administered to rats with CIA, which demonstrated less severe disease . Another study in mice developing spontaneous arthritis (IL1RA-/- mice) showed more severe disease after application of IL-23 – this model is particularly Th17-cell dependent, reflecting the role of IL-1 in amplifying Th17 cells. To better understand the relationship between IL-23p19 and IL-12p40, Murphy and colleagues used mice deficient for these specific subunits . Loss of IL-23p19 protected mice against arthritis, whereas IL-12p40-deficient mice experienced exacerbated disease. This not only demonstrates how potent single cytokine subunits in the right environment may act but also how the counter-subunit totally changes the profile from ‘good’ to ‘bad’ behaviour.

The importance of IL-12/IL-23 in human pathology is now confirmed in phase III clinical trials of ustekinumab, a human monoclonal antibody against IL-12p40 targeting IL-23 and IL-12. In psoriasis, this agent is now licenced for use with substantial beneficial effects observed in >75% of patients . This does not however distinguish between these two cytokines – targeting Th1 and Th17 cell differentiation may be effective but might have future side effects due to inhibition of two T cell subsets. Thus, in the future, a more rational approach might be blockade of IL-23p19 alone; trials are underway to test this possibility.

IL-23/IL-12p40 Biology in RA

The cytokine IL-23 is a heterodimeric cytokine composed of a p19 and a p40 subunit; the latter is shared with IL-12 (p35/p40), which drives naïve T cells into Th1 cells. IL-23 is produced by activated dendritic cells and macrophages and seems to be important in induction and maintenance of Th17 cells. Expression in RA synovial fluid is controversial with groups reporting low and others very high levels. Interestingly, in synovial tissue, bioactive IL-23 seems to be present only in low levels despite high expression of IL-23p19 . Regarding bone homeostasis, IL-23 up-regulates RANK on myeloid precursors and induces RANKL expression on CD4 + T cells . This direct effect, combined with an indirect effect via IL-17, strongly supports osteoclast differentiation and further erosive disease. Recent genetic studies in psoriasis patients provide strong support for an association of IL-23 related loci with the involvement of single nucleotide polymorphism for IL-23A, IL-12B and IL-23R . So far, these variants have not been observed in RA.

Due to the shared subunits, interpretation of models conducted in mice deficient in p40 is difficult as it lacks both IL-12 and IL-23. To test the effect of IL-23 in auto-immune disease, anti-IL-23 antibodies were administered to rats with CIA, which demonstrated less severe disease . Another study in mice developing spontaneous arthritis (IL1RA-/- mice) showed more severe disease after application of IL-23 – this model is particularly Th17-cell dependent, reflecting the role of IL-1 in amplifying Th17 cells. To better understand the relationship between IL-23p19 and IL-12p40, Murphy and colleagues used mice deficient for these specific subunits . Loss of IL-23p19 protected mice against arthritis, whereas IL-12p40-deficient mice experienced exacerbated disease. This not only demonstrates how potent single cytokine subunits in the right environment may act but also how the counter-subunit totally changes the profile from ‘good’ to ‘bad’ behaviour.

The importance of IL-12/IL-23 in human pathology is now confirmed in phase III clinical trials of ustekinumab, a human monoclonal antibody against IL-12p40 targeting IL-23 and IL-12. In psoriasis, this agent is now licenced for use with substantial beneficial effects observed in >75% of patients . This does not however distinguish between these two cytokines – targeting Th1 and Th17 cell differentiation may be effective but might have future side effects due to inhibition of two T cell subsets. Thus, in the future, a more rational approach might be blockade of IL-23p19 alone; trials are underway to test this possibility.

TWEAK

TWEAK (TNF-like weak inducer of apoptosis) is a member of the TNF superfamily of cytokines and shares a high level of homology with other TNF cytokine members but signals through its own receptor TWEAKR (TNFRSF12A). It was first described in1997 and was shown to be expressed in various tissues, including brain, heart and spleen but was undetectable in unstimulated peripheral blood mononuclear cells (PBMCs) . Stimulation of human monocytes with interferon (IFN)γ induces secretion of TWEAK ,and upon binding to its receptor, Fn14, promotes the release of the pro-inflammatory cytokines and chemokines, IL-6, IL-8 and MMP-1, from dermal fibroblasts and synoviocytes . In addition, it has also been shown to up-regulate the expression of the adhesion molecule intra-cellular adhesion molecule (ICAM)-1 in synovial fibroblasts . It has also been shown to exhibit cytotoxic effects in adenocarcinoma cell lines and induce proliferation of endothelial cells driving angiogenesis . TWEAK has now been implicated in multiple pathologies including multiple sclerosis , ankylosing spondylitis (AS) and RA . TWEAK and its receptor are both expressed in the synovial membrane of RA and psoriatic arthritis patients . The concentration of TWEAK was shown to be significantly elevated in the serum of 40 age-matched AS and RA patients compared with healthy controls and levels correlated strongly with the concentration of serum TNFα and DAS28 used as a measure of disease severity . Interestingly, patients treated with anti-TNF had lower serum concentrations of TWEAK. Given the potential pathological role of TWEAK, there is now a great deal of interest in developing biologics to inhibit its function. Inhibition of TWEAK with a neutralising antibody ameliorated the clinical and histological disease severity, reduced synoviocytes proliferation and angiogenesis in a murine CIA model and reduced serum concentrations of monocyte chemoattractant protein (MCP)-1/2 . However, T-cell proliferation and anti-collagen antibody titres were not affected. Given the success of biologics against TWEAK, there is current interest in developing new biologics with better efficacy for translation into the clinic for the treatment of multiple auto-immune diseases.

GM-CSF

GM-CSF signals through a heterodimer receptor complex comprising of an α-ligand specific sub-unit and a βc sub-unit (shared with IL-3), which allows high affinity binding of GM-CSF. GM-CSF has been implicated in the regulation of neutrophil, macrophage and dendritic cell production and differentiation from bone-marrow precursors . GM-CSF is has been detected within the synovium of patients with RA and the concentration of GM-CSF is elevated in the plasma of patients with RA and systemic lupus erythematosus . As such, GM-CSF has been implicated in the pathology of arthritis. GM-CSF is now known to have a pro-inflammatory function and has been shown to modulate T cell function in which it promotes differentiation of the newly described Th17 cells in the murine CIA model through enhanced production of IL-6 and IL-23 from dendritic cells . In addition, GM-CSF been shown to be secreted by rheumatoid synovial fibroblasts-stimulated TNFα and IL-17 , creating a positive feedback loop between GM-CSF-producing fibroblasts and IL-17-producing T cells. Although regarded as being involved in arthritis pathology, GM-CSF was originally reported to inhibit bone resorption through the process of osteoclastogenesis . However, it has recently been shown that osteoclasts differentiated in the presence of GM-CSF have stronger osteoclastic activity when stimulated with TNFα implying GM-CSF may drive bone erosion in arthritis by a previously unrecognised mechanism. Furthermore, it has been shown that mice deficient for GM-CSF failed to develop CIA, confirming the pro-inflammatory effect of GM-CSF whilst administration of exogenous GM-CSF exacerbated the severity of arthritis . More recently, administration of anti-GM-CSF antibodies has been shown to ameliorate clinical and histological disease resulting in a decreased concentration of synovial TNFα and IL-1 and anti-GMCSF treatment inhibited proliferation of human RA fibroblast synoviocytes . As yet, clinical trials using antibodies against GM-CSFR have shown some promise but remain at an early stage; further larger analyses are awaited with interest.

BLyS (BAFF) and APRIL

B-cell depletion by the anti-CD20 targeted monoclonal antibody, rituximab, has been successful in ameliorating disease in a proportion of RA patients. As such, B-cell depletion is now a favourable strategy for the treatment of inflammatory arthropathies and other auto-immune conditions alike. BLyS (B-lymphocyte stimulator) and APRIL (A Proliferation Induced Ligand) are both members of the TNF superfamily of cytokines. They are potent regulators of B-cell survival and differentiation, which are expressed by T cells, macrophages, dendritic cells and osteoclasts . BLyS promotes B-cell proliferation and transgenic mice overexpressing BLyS exhibit increased numbers of peripheral blood B-cells, splenomegaly, increased size of germinal centres and a higher level of major histocompatibility complex (MHC)II expression suggesting an increased capacity to present antigen and B-cell activation . In addition, silencing of BLyS ameliorates arthritis characterised by a suppressed Th17 differentiation . The capacity of APRIL-deficient mice to develop arthritis is also significantly reduced and is associated with decreased serum concentrations of collagen-specific immunoglobulin G (IgG)2a, IL-17 and increased Th2 cytokines . In addition, APRIL over-expression correlates with increased serum levels of the auto-antibody immunoglobulin (Ig)M and induces the secretion of pro-inflammatory cytokines from RA synovial fibroblasts . The concentrations of APRIL within the synovium are increased in RA and correlate strongly with the number of infiltrating plasma cells . Furthermore, heterodimeric complexes of APRIL and BLyS are found at elevated levels in the serum of RA patients compared with healthy controls and their concentrations within synovial fluid independently correlate with an increase in synovial lymphoid architecture/organisation in which dendritic cells appear to be the primary cellular source and disease severity by DAS28 . These observations strongly support a pathophysiological role for APRIL and BLyS in driving the aggravation in inflammatory arthropathies. Thus, much effort has been expended in developing biologics. Belimumab; an anti-BLyS human monoclonal antibody, is now in clinical trials for the treatment of systemic lupus erythematosus and similar studies are likely to be extended to other auto-immune diseases such as arthritis and multiple sclerosis. Similarly, atacicept (which inhibits both BLyS and APRIL) has been tested in a variety of human inflammatory conditions and clinical trial data are awaited with interest.

Adipocytokines

Adipocytokines, or adipokines, are bioactive proteins mainly produced by white adipose tissue that are known to regulate glucose and lipid metabolism, energy homeostasis and cardiovascular function. In addition, they can regulate immune and inflammatory responses . Adipokine biology may be of particular interest in understanding the accelerated cardiovascular disease found in RA patients . Furthermore, local fat pads within or surrounding joints have been demonstrated to secrete cyokines and therefore may be implicated in disease pathogenesis . Here, we present the current evidence for the expression and complex roles of adipokines, in particular, leptin, adiponectin, resistin and visfatin, in the inflammatory and degenerative processes underlying RA. These cytokines have not yet reached clinical intervention but we consider them of significant clinical importance going forward; hence their inclusion herein.

Leptin

Leptin is a peptide hormone mainly produced by adipocytes, encoded by the gene obese (ob) in mice and LEP in humans, which regulates body weight by inhibiting food intake. It belongs to the class I cytokine superfamily and can modulate many inflammatory and immune responses, including inducing T-cell activation, driving a Th1 response and multiple effects on monocytes/macrophages, neutrophils, basophils, eosinophils, natural killer (NK) and dendritic cells (reviewed in Matarese et al .). Circulating leptin concentrations in RA patients have been measured by several groups. In some studies, patients with RA showed considerably higher plasma levels of leptin than healthy controls, and significantly higher levels in women than in men . A positive correlation was also found between serum leptin concentration and the value of DAS28, ESR and the number of tender joints . Furthermore, in fasted RA patients, a fall in circulating leptin was associated with CD4+ lymphocyte hyporeactivity and increased IL-4 secretion . Leptin concentrations have also been assessed in matched blood and synovial fluid samples where leptin levels were significantly higher in plasma than in synovial fluid samples obtained simultaneously and higher than in control samples . However, other studies show that while circulating leptin levels did correlate with body mass index (BMI) and percentage body fat, they did not differ from levels in healthy controls , and, in some cases, actually were inversely correlated with inflammatory markers such as C-reactive protein (CRP) and IL-6 .

Data from animal models have also shown conflicting results making it difficult to draw conclusions about the potential role of leptin in arthritis. Experimental antigen-induced arthritis is less severe in leptin-deficient ob/ob mice or leptin receptor-deficient db/db mice, with lower cytokine levels, antigen-specific antibodies and T cell proliferative responses . By contrast, in a zymosan-induced arthritis (ZIA) model, which is independent of an adaptive immune response, joint swelling was similar; however, the resolution of acute inflammation was delayed in ob/ob or db/db mice compared with wild-type mice, suggesting that chronic leptin deficiency interferes with adequate control of the inflammatory response in ZIA . Furthermore, in a murine Staphylococcus aureus -induced septic arthritis model, leptin production was decreased and treatment with recombinant leptin significantly decreased the severity of disease and the inflammatory response .

Adiponectin

Adiponectin is protein released by adipocytes with anti-diabetic and anti-inflammatory properties. It can exist as a trimer, hexamer or larger multimer with 12–18 subunits and circulates at very high plasma concentrations. Patients with RA showed considerably higher plasma concentrations of adiponectin than osteoarthritis (OA) patients or healthy controls , and these levels: (1) correlate with the severity of RA evaluated by extent of joint destruction , (2) negatively correlate with CRP levels and (3) were higher in women . In addition, several studies have shown an increase in plasma levels of adiponectin after administration of anti-TNF agents in patients with RA . Synovial fluid concentrations of adiponectin were significantly higher in patients with RA than in those with OA , and immunohistology showed that numerous cells in the synovial biopsies of RA expressed adiponectin and its two receptors (adipoR1 and adipoR2) . The addition of adiponectin to synovial fibroblasts in culture stimulates IL-6, IL-8 and MCP-1 production via the adipoR1 receptor and the NF-κB pathway. Human and murine chondrocytes also express functional adipoRs and adiponectin treatment-induced NOS2 and increased IL-6, MMP-3, MMP-9 and MCP-1 secretion by murine-cultured chondrocytes, thus indicating a role in cartilage homeostasis . In murine CIA models, intra-articular injection of adiponectin , or systemic adenoviral delivery of adiponectin prevent inflammation and joint destruction.