Acute phase proteins can be positive (that is, their concentrations increase in response to inflammation - for example, C-reactive protein, serum amyloid A) or negative (that is, their concentrations decrease in response to inflammation - for example, apolipoproteins that protect from inflammation by inhibiting the contact between activated lymphocytes and monocytes for the production of IL-1 and TNF) (Figure ​(Figure1)1) [4]. Describing autoimmune diseases as a hyperactivity of the immune system, and using IL-1 receptor agonist (IL-1Ra) and apolipoprotein A₁ as examples, Professor Dayer went on to consider how cytokines interact with one another in the body and the role that they might play in the development of autoimmune diseases. Discussing the delicate balance that exists between IL-1 and IL-1Ra, Professor Dayer noted that, during inflammation, leptin produced by adipocytes can stimulate the production of IL-1 by the hypothalamus (Figure ​(Figure2)2) [5]. Acting at this level, IL-1 becomes a cachexin resulting in the loss of both adipose tissue and lean body mass. The adipocytes also produce IL-1Ra, however, which is able to block the cachectic action of IL-1 and increase appetite [6].

ATAC was first cloned in both mice [7] and humans [8] more than 15 years ago, and was initially believed to act as a chemoattractant for lymphocytes [7]. Subsequent studies in humans indicated that ATAC/lymphotactin was primarily produced in the synovium of RA patients and so, given its role as a chemoattractant, might be a key modulator for T-cell trafficking in the pathogenesis of RA [9]. Studies using murine models suggest that the receptor for ATAC/lymphotactin is only present on CD8⁺ dendritic cells, such as those found in the spleen, which, given the role of CD8 cells in the development of self-tolerance by the immune system, may be implicated in the development of autoimmunity.

Cartilage destruction and bone erosion are major problems in RA, and studies have shown that these processes may be mediated by cytokines. Murine arthritis models have demonstrated the therapeutic potential of anti-TNFα and anti-IL-1 antibodies [10]. During this session on cytokines and arthritis, Professor Wim van den Berg (Nijmegen, The Netherlands) postulated that different cytokines may dominate at different stages of the inflammatory process. For example, in early-stage collagen-induced arthritis both anti-TNFα and anti-IL-1 treatments have been shown to be effective [11]. Moreover, IL-17 - a T-cell cytokine expressed in the synovium and synovial fluid of patients with RA - has been shown to be a potent inducer of TNFα and IL-1, and is involved in both the initiation and progression of murine arthritis models [12]. IL-17 not only synergizes with TNFα, but also enhances inflammation and destruction independent of IL-1 and TNFα, making it an additional potential target for the treatment of RA. As such, tailor-made treatment is required for the different patient groups.

Professor Stefan Rose-John (Kiel, Germany) discussed the inflammatory properties of IL-6 and the complexity of IL-6 signalling, together with the consequences of and various techniques employed in IL-6 blockade. During his presentation, Professor Rose-John noted that all IL-6 signalling is mediated via binding of the IL-6 receptor to the ubiquitously expressed glycoprotein 130. The IL-6 receptor is normally membrane bound and expressed only on hepatocytes and some leukocytes; however, this receptor can be cleaved, or shed, from the cell via the actions of a metalloprotease (ADAM17), producing a soluble IL-6 receptor that can then bind to cells which do not normally express this receptor. This phenomenon, so-called trans-signalling, enables IL-6 to exert its effects upon a much wider range of cell types, including smooth muscle cells, endothelial cells and neural cells [13]. Binding of IL-6 to soluble IL-6 receptor has been shown to be proinflammatory, and has thus been implicated in the pathogenesis of RA - making this pathway a target for therapeutic interventions [13]. Recent studies have revealed that selective blockade of this alternative IL-6 signalling pathway using an engineered variant of soluble glycoprotein 130 (sgp130Fc) led to substantial clinical improvement in a preclinical arthritis model [14].

Although TNFα has been implicated in the pathogenesis of RA, it also plays an important role in host defence, with complete TNFα blockade associated with an increased risk of mycobacterial infection [15]. Professor Sergei Nedospasov (Berlin, Germany and Moscow, Russia) described in detail the development of a novel humanized murine model for the study of TNFα in collagen-induced arthritis [16]. In this heterozygous model, both alleles are active - producing human and murine TNFα, and thus allowing detailed comparison of human versus mouse regulation of TNFα expression. In the homozygous model, approximately 40 to 50% of humanized mice will develop arthritis (mediated by human TNFα), which can be treated clinically using TNF blockers. Using this, and several other murine models that they have developed, Professor Nedospasov and his team aim to determine the source of TNFα that protects against infection and to develop TNFα inhibitors that target specific cell types or compartments in the body where TNFα is overproduced.

Session 2 continued the theme of cytokines in arthritis and began with an overview of the biological role of IL-1 and its novel homologue IL-33 in inflammatory responses. The original members of the IL-1 superfamily were IL-1α, IL-1β and IL-1Ra; however, several additional molecules with structural homology have been added to this family in recent years, namely IL-1F5, IL-1F6, IL-1F7, IL-1F8, IL-1F9, IL-1F10, IL-18 and IL-33 (Figure ​(Figure3).3). Professor Cem Gabay (Geneva, Switzerland) discussed how the balance between IL-1and IL-1Ra influences the development and severity of arthritis. Using conditional knockout mice in which the expression of IL-1Ra has been selectively targeted in myeloid cells, Professor Gabay's group showed that these mice had a more rapid onset and severe form of collagen-induced arthritis with lower levels of IL-1Ra in lymph nodes and increased Th1 and Th17 responses [17]. IL-33 - which binds to a member of the IL-1 receptor family, inducing similar intracellular signals to IL-1 - is also expressed in human synovial fibroblasts with increased expression in arthritic joints of murine models, suggesting a potential role in the pathogenesis of arthritis [18].

Interferons are a family of naturally secreted proteins with immunomodulatory functions. IFNβ has anti-inflammatory properties and plays a role in bone homeostasis. IFNβ treatment has been shown to reduce the severity of collagen-induced arthritis in mice [19,20] and rhesus monkeys [21]. Conversely, IFNβ deficiency resulted in the development of severe collagen-induced arthritis in mice as a result of increased activation of stromal cells and osteoclasts [22]. After reviewing the lack of efficacy of subcutaneous injections of IFNβ protein three times weekly for the treatment of RA [23], Dr Margriet Vervoordeldonk (Amsterdam, The Nether-lands) went on to discuss the potential of intra-articular IFNβ gene therapy for the treatment of RA. In a set of proof-of-principle studies using an adenoviral vector and recombinant adeno-associated virus type 5 for local delivery of the rat IFNβ gene [24,25], a beneficial effect has been shown on arthritis development in two different rat models of arthritis.

Local delivery of adenoviral vector or recombinant adeno-associated virus type 5 vectors expressing rat IFNβ after the onset of disease reduced paw swelling impressively in both injected and uninjected joints. Strikingly, IFNβ treatment protected against bone and cartilage erosions. Together, the results provide a rationale for IFNβ as a therapeutic target for intra-articular gene therapy for arthritis.

Techniques that may be employed to manipulate the cytokine environment using tolerogenic dendritic cells were presented by Professor John Isaacs (Newcastle, UK). Dendritic cells are antigen-presenting cells that initiate and orchestrate immune responses. By generating tolerogenic dendritic cells, it is possible to downregulate immune responses. Tolerogenic dendritic cells have been generated by culturing monocytes in the presence of IL-4 and GM-CSF and then exposing them to immunosuppressive agents such as dexamethasone [33]. Tolerogenic dendritic cells have been shown to be highly stable and refractory to further stimulation by either lipopolysaccharide or peptidoglycan. They also maintain their tolerogenic phenotype in a proinflammatory cytokine environment. Professor Isaacs described preclinical studies where intravenous administration of tolerogenic dendritic cells has been shown to decrease the severity of inflammatory arthritis and to inhibit disease progression. Studies in humans are planned in the near future, and it is hoped that this approach may provide a route for immune reprogramming in autoimmunity.

An update on IL-17 and the growing IL-17 family of cytokines was provided by Professor Pierre Miossec (Lyon, France). Studies have shown production of functional IL-17 by RA synovium explants and have shown that overexpression of IL-17 in the knee joint of mice induces bone destruction and cartilage damage [34]. A whole family of IL-17 molecules, now known as IL-17A to IL-17F, has been identified together with a range of IL-17 receptors. The highest degree of homology (50%) is observed between IL-17A and IL-17F, both of which are associated with the pathogenesis of RA [35]. In contrast, IL-17E displays the lowest level of homology with IL-17A (17%) and has not been shown to play any role in RA [36-38]. IL-17 has been shown to induce the expression of IL-1, IL-6, IL-8, GM-CSF and TNFα, all of which may contribute to the pathogenesis of RA. In concluding, Professor Miossec noted that the biology of IL-17A indicates a role in the development of inflammation and joint destruction, and that the contribution of other IL-17 family members to this process must be considered. Different novel therapeutic approaches targeted at these pathways are now under investigation with interesting preliminary results.

The role of B cells in autoimmune diseases was further discussed during a presentation by Dr Simon Fillatreau (Berlin, Germany). Murine models of multiple sclerosis have been employed in an attempt to elucidate the interactions that may occur between cytokines, B cells and T cells. Experimental autoimmune encephalomyelitis, induced by immunizing mice with myelin proteins, is mediated by Th1 and Th17 CD4⁺ T cells and is clinically manifested by paralysis. Experiments have shown that recovery from experimental autoimmune encephalomyelitis requires the production of IL-10 by B cells [47]. Myeloid differentiation primary response gene 88 is a universal adapter protein that signals via Toll-like receptors to activate the transcription factor NF-κB and trigger the regulatory function of B cells, and hence control the inflammatory T-cell response. In addition, B-cell activation via the B-cell receptor and CD40 has also been implicated in recovery from experimental autoimmune encephalomyelitis. A two-step process whereby B cells limit inflammation and aid recovery from experimental autoimmune encephalomyelitis thus operates within this murine model. Firstly, myeloid differentiation primary response gene 88 signalling in B cells induces potent anti-inflammatory cascades capable of suppressing chronic immune responses. Secondly, the B-cell receptor and CD40 then act to amplify these B-cell-mediated cascades. These findings may have implications for the study of multiple sclerosis in humans.

Professor Peggy Crow (New York City, USA) discussed how microarray analyses have enabled the identification of genetic signatures in the peripheral blood of individuals with autoimmune diseases. In addition, increased IFNα pathway activation has been associated with increased autoimmunity together with increased inflammation and tissue damage.

Continuing on the same theme, Professor Lars Rönnblom (Uppsala, Sweden) concurred that administration of IFNα can cause autoimmune disease and that many patients with autoimmune conditions have an ongoing production of IFNα. Moreover, anti-IFNα antibodies applied in preclinical and early clinical trials have been shown to attenuate autoimmunity.

Professor Joachim Sieper (Berlin, Germany) provided an overview of clinical data that clearly demonstrated the benefits of IL-12 and IL-23 blockade in individuals with psoriasis [58,59]. This approach has proved less beneficial in individuals with either Crohn's disease or psoriatic arthritis. Moreover, further investigations are required in order to establish whether joint IL-12 and IL-23 blockade is required, or whether blockade of IL-23 alone would be sufficient for the treatment of autoimmunity.

A comprehensive review of the properties of IL-21 and the IL-21 receptor was provided by Professor Peter Lipsky (Bethesda, MD, USA). Professor Lipsky outlined the role played by IL-21 in B-cell differentiation and T cell-B cell interactions, and commented that IL-21 is a new target for the downregulation of B-cell responsiveness (Figure ​(Figure5).5). After noting that IL-21 is elevated in individuals with systemic lupus [60], Professor Lipsky speculated on the possible benefit to these individuals of effective IL-21 blockade.