Interleukin-17 and its _target genes: mechanisms of interleukin-17 function in disease

Rheumatoid arthritis

Many, if not most, autoimmune diseases are now connected in some manner to IL-17 or the Th17 pathway. In particular, RA has been intensively studied, starting even before the recognition of the Th17 subset. The key features of inflamed arthritic joints are proliferating synovial fibroblasts, joint and cartilage erosion, infiltrating CD4⁺ T cells and autoantibody-producing plasma cells. In addition, increased numbers of innate immune cells (DC, granulocytes and macrophages), in some cases ectopic germinal centres (GC) are found within joints. Early studies showed that high levels of IL-17 were found in the rheumatoid synovium of patients with RA but not of controls or of patients with osteoarthritis. Consistently, adding IL-17 to an in vitro culture system stimulated bone resorption and collagen destruction.²³ Furthermore, neutralizing IL-17 or its receptor in collagen-induced arthritis mouse models resolved RA symptoms, IL-17A-deficient mice are protected from collagen-induced arthritis, and adding IL-17 ectopically by gene therapy exacerbated disease.^24–26 Consequently, IL-17 appears to promote both inflammation and bone destruction in RA.

These findings pose the question of how IL-17 mediates its pathogenic activities. Interleukin-17 induces pro-inflammatory cytokines such as TNF-α, IL-1β and IL-6 from cartilage, synoviocytes, macrophages and bone cells.²⁷ Collectively, these pro-inflammatory cytokines contribute to RA flare-ups and also establish a chronic inflammatory state by a self-reinforcing positive feedback loop wherein IL-17-induced IL-6 maintains the Th17 T-cell population.²⁸ The IL-17 also stimulates the production of multiple chemokines, including IL-8/CXCL8, CXCL1 (KC/Groα), CXCL2 (MIP2α/Groβ), CCL20 (MIP-3α), CCL2 (MCP1) and CCL7 (MCP3).^3,5,27,29 These serve to recruit neutrophils, macrophages and lymphocytes to the synovium, thereby enhancing inflammation. Secondary lymphoid GC formations are found ectopically in RA synovial tissue.^30,31 Although it has been reported that synovial lymphoid neogenesis is not a major determinant of RA-specific autoantibody responses, this phenomenon is indicative of ongoing inflammation.³²

Irreversible deformities in joints are a key feature of advanced RA, caused by extensive cartilage and bone erosion (Table 1). Interleukin-17 contributes to this process by inducing expression of matrix metalloproteinases (MMP) 1, 2, 3, 9 and 13, which drive degradation of extracellular matrix within the joint.^33–39 It also induces prostaglandin E₂ via cyclooxgenase-2, which enhances inflammation by many mechanisms including vasodilatation.⁴⁰ Furthermore, IL-17 induces expression of receptor activator of nuclear factor-κB ligand (RANKL) in osteoblasts;²³ RANKL is a membrane-bound receptor of the TNF superfamily that promotes differentiation of osteoclasts. Strikingly, Th17 cells also express elevated levels of RANKL, suggesting that they may be particularly adept at promoting bone turnover.⁴¹ Accordingly, IL-17 not only enhances inflammation, but stimulates osteoclast differentiation leading to subsequent bone and cartilage damage.⁴²

Although IL-17 alone has the capacity to induce pro-inflammatory factors, its activities are vastly increased when combined with other cytokines, particularly TNF-α (Table 2). This is probably the situation in inflamed joints, where multiple inflammatory cytokines are over-expressed.⁴³ Interleukin-17 synergizes with TNF-α to promote induction of nearly all its _target genes, and in many cases synergy has also been observed with IFN-γ and IL-1β (reviewed in ref. 27). As a consequence, IL-17 alone or together with other inflammatory cytokines in the inflamed joint mediates adverse events in RA.

Systemic lupus erythematosus

Systemic lupus erythematosus (SLE) is a multi-organ systemic autoimmune disease characterized by autoantibody production. Although traditionally considered mainly a B-cell disease, recent reports indicate that there is likely to be a role for IL-17 in lupus (reviewed in ref.44). Interleukin-17 is elevated in the serum of many SLE patients.^45,46 The BXD2 mouse strain develops SLE-like features and spontaneous erosive arthritis with age.⁴⁵ These mice produce pathogenic autoantibodies because of increased somatic hypermutation and enhanced class-switch recombination, mediated by over-expressed activation-induced cytidine deaminase in GC.⁴⁷ In studies to elucidate the underlying mechanism of this phenotype, Hsu et al.⁴⁵ unexpectedly discovered IL-17-dependent signalling in BXD2 B cells that led to enhanced autoantibody production and increased numbers of GCs. CD4 T cells from BXD2 mice were also more prone to differentiate into Th17 cells. The IL-17 functions in this setting by inducing genes encoding Regulator of G-protein signalling 13 and 16 (Rgs13 and Rgs16), which inhibit G-coupled protein receptors such as CXCR4. Hence, IL-17 disrupts trafficking of B cells within lymph nodes, which appears to promote spontaneous generation of autoreactive GC and hence elevated autoantibody production.⁴⁵

Another animal model with lupus-like features is the Ets-1^−/− mouse, characterized by increased high-titre autoantibodies that deposit in kidney.⁴⁸ In these mice, more naive T cells differentiate into a Th17 phenotype.⁴⁹ Ets-1 is transcription factor expressed in various immune cells and it skews towards a Th17 phenotype by interfering with the ability of IL-2 to inhibit Th17 differentiation. The Ets1^−/− mice have low IL-2 production, but T cells from these mice also appear to be refractory to the inhibitory effects of IL-2 on Th17 development. Interestingly RORγt expression and IL-2-induced signal transducer and activator of transcription 5 (STAT5) were comparable with wild-type and Ets-1^−/− T cells, and so the specific pathway by which Ets-1 exerts its effects is down-stream or independent of STAT5.

Interestingly, IL-17 has also been recently shown to synergize with another TNF superfamily member, B-cell activating factor (BAFF), to protect B cells from apoptosis, thereby increasing the number of autoantibody-producing cells.⁴⁶ Increased BAFF expression is found in ∼ 22–25% of SLE serum samples, and BAFF transgenic mice have a lupus-like phenotype.⁵⁰ Both IL-17R and BAFFR use a common adaptor molecule Act1; however, Act1 is a positive regulator of IL-17R signalling, whereas it is a negative regulator of BAFFR down-stream pathways.⁵¹ Following IL-17 and BAFF stimulation of B cells, Act1 is preferentially recruited to IL-17RA, providing an intriguing mechanism for synergistic signalling between these two systems. Subsequently, IL-17 and BAFF together enhance expression of the transcription factor Twist1, which initiates a cascade of gene expression leading to induced expression of the anti-apoptotic genes Twist-2 and Bfl-1, ultimately promoting B-cell differentiation to autoantibody-producing plasma cells⁴⁶ (Tables 1 and 2). Hence, the impact of IL-17 on B cells may explain its role in contributing to SLE pathogenesis.

Inflammatory bowel disease

Inflammatory bowel disease, encompassing both Crohn’s disease and ulcerative colitis, is a chronic relapsing inflammatory disorder in the gastrointestinal tract, caused in part by an unregulated immune response to intestinal bacteria. The gut immune system is composed of heterogeneous cell populations, which differ along the gastrointestinal tract (reviewed in ref.52). A notable feature of the gut mucosa is specialized epithelial cells called microfold (M) cells, which sample the gut lumen and transport bacteria and transport antigens to APC on the basolateral surface. In healthy individuals, intestinal DC typically induce T-cell unresponsiveness, which is needed to maintain tolerance to commensal organisms and food antigens.⁵² Recognition of gut microbes is mediated by Toll-like receptors (TLR) and also intracellular pattern recognition receptors of the nucleotide oligomerization domain (NOD) family. Strikingly, genome-wide association studies (GWAS) identified NOD2 (CARD15) as an IBD-associated gene,^53,54 leading to the current paradigm that IBD results from unregulated immune responses against commensal bacteria. In addition to NOD2, additional genes identified in GWAS studies have implicated the Th17 pathway, most notably the IL-23R,⁵⁵ which is expressed primarily on IL-17-producing cells. Other Th17-associated genes implicated in GWAS studies of IBD include JAK2, IL12B, STAT3, CCR6 and TYK2,^56–58 all of which point to a major role for this pathway in IBD pathogenesis.

Animal models support a role for the IL-23/IL-17 axis in IBD. For example, the IL-10^−/− mouse develops spontaneous colitis, which is prevented when the mice are crossed to IL-23p19^−/− animals (lacking Th17 cells) but not when crossed to IL-12p35^−/− animals (lacking Th1 cells).⁵⁹ Similarly, in chemically induced IBD such as trinitrobenzene sulphonic acid (TNBS) -induced IBD, disease severity is higher in mice deficient in IL-12p40^−/− (lacking both IL-23 and IL-12) compared with IL-12p35^−/− (lacking just IL-12).⁶⁰ In the TNBS model, IL-17 _target genes such as IL-6 and CXCL2 were reduced in the colon, and IL-17RA^−/− mice were protected from severe disease.⁶¹ In contrast, results differ in another chemically induced model, dextran sulphate sodium (DSS) -induced colitis. The DSS disrupts the epithelial cell barrier, causing mucosal microflora to activate mucosal macrophages. In this setting, neutralizing IL-17 worsened symptoms, which was associated with increased expression of TNF-α, IFN-γ, IL-6 and CCL5/RANTES.⁶² A recent study compared cytokine profiles in acute (day 7) and chronic (day 35–75) TNBS- or DSS-induced IBD. Data suggested that various cytokines and effector cells are working in different time-points.⁶³ For example, IL-6, TNF-α, IL-17 and CXCL1 were elevated in the acute DSS model (7 days), which later shifted to a more Th2/Th17-like profile (IL-4, IL-10 as well as IL-17). In the acute TNBS model, IFN-γ, IL-12, IL-17 and CCL2 were elevated, reminiscent of a mixed Th1/Th17 profile. Although detailed cellular mechanisms are still not well defined, the timing of disease and initiation of disease may dictate which Th cell type is most instrumental in establishing disease.

RAG1^−/− mice adoptively transferred with CD45RB^hi CD25⁻ CD4⁺ T cells develop aggressive colitis and wasting disease, which is strongly IL-23-dependent but also IFN-γ-dependent. Interestingly, however, IL-17 is protective in this setting, because IL-17^−/− CD45RB^hi T cells induce a more aggressive disease. Surprisingly, IL-17RA^−/− CD45RB^hi T cells were also more aggressive, indicating that these cells respond to IL-17.⁶⁴ Naive Th0 cells do not express IL-17RA, but start to do so upon differentiation to Th1. Interleukin-17 suppresses IFN-γ secretion by repressing T-bet, the master regulator of Th1 cells, which in turn leads to suppression of IFN-γ, IL-12Rβ1 and osteopontin expression (Table 1). Therefore, in this transfer model, IL-17 protects against IBD by limiting Th1 cell activity.⁶⁵

Psoriasis

Psoriasis is a chronic inflammatory skin disorder characterized by dermal hyperplasia. The key histological features of psoriatic skin are epidermal keratinocyte hyperproliferation, vascular proliferation and infiltration of DCs, macrophages, neutrophils and T cells.⁶⁶ The critical roles of IL-23/IL-17 were highlighted in a GWAS study that linked IL-23R polymorphisms to psoriasis, similar to IBD.⁶⁷ Based on this finding, a model of psoriasis was developed using intradermal injection of IL-23. In this model, anti-IL-17 treatment decreased granulocyte colony-stimulating factor (G-CSF) and MMP-13, although it had no effect on erythema, induration and parakeratosis.⁶⁸ Several IL-10-family cytokines have been found in psoriatic skin, including IL-19, IL-20, IL-22 and IL-24. However, IL-19^−/− and IL-24^−/− mice still developed skin thickening following IL-23 injection, although mice lacking the common receptor IL-20R2 were resistant.⁶⁸ In contrast, Zheng et al.⁶⁹ observed no elevation of IL-19, IL-20 or IL-24 in the IL-23 injection model, whereas IL-22^−/− mice exhibited significant reduction in skin thickness. A role for IL-22 is consistent with findings in human microarray studies, which identify IL-22 as well as many typical Th17 genes.^70–72 As a consequence, IL-22 from Th17 cells and an IL-20R2-using cytokine have roles in developing psoriasis-like symptoms caused by ectopic IL-23.

A mechanism of IL-17 activity in psoriasis likely stems from its co-operative gene regulation IL-22 and other stimuli. Together with IL-17, IL-22 synergistically increases expression of skin antimicrobial peptides, including β-defensin-2 (BD-2), S100A7 (psoriasin) and S100A8/9 (calprotectin).⁷³ Supporting this, S100A7–9 are elevated in psoriasis, correlating with disease onset.⁷⁴ Interestingly, psoriasis patients are more resistant to skin infections than people without psoriasis, perhaps as the result of elevated antimicrobial peptide production. Another antimicrobial peptide, cathelicidin (LL37), is synergistically increased by treatment with IL-17 in combination with 1,25-dihydroxyvitamin D3.⁷⁵ LL37-bound self-DNA fragments trigger TLR9 in DC, which induces a potent adaptive immune response, possibly one of the mechanisms by which self-tolerance is broken.⁷⁶

Experimental autoimmune encephalomyelitis

Experimental autoimmune encephalomyelitis (EAE) is a model of multiple sclerosis, a T-cell-mediated autoimmune disease of the central nervous system. It is elicited by immunization of neuroantigens, such as myelin basic protein and proteolipid protein. As with many other autoimmune conditions, Th1 cells were long thought to be responsible for EAE pathology, despite the fact that IFN^−/−, IFNγR^−/− and IL-12p35^−/− mice were susceptible.^77–79 Landmark studies comparing the IL-12p35^−/− and IL-23p19^−/− mice showed clearly that the Th17 pathway was responsible for pathology.⁸⁰ Further evidence for the role of Th17 cells in driving EAE was shown in STAT6^−/−/T-bet^−/− doubly deficient mice, lacking Th1 and Th2 cells.⁸¹ Furthermore, EAE in these mice could be ameliorated by treatment with anti-IL-17 antibodies. An elegant study noted that regulation of IL-17 pathogenic activity can be controlled by IL-10. Specifically, Th17 cells derived ex vivo following treatment with TGF-β and IL-6 do not cause EAE upon adoptive transfer, whereas Th17 cells derived ex vivo with TGF-β, IL-6 and IL-23 develop EAE.⁸² Interleukin-23 suppresses IL-10, correlating with elevated expression of IL-17 _target genes such as CXCL10 (IP10), CCL2, CXCL2 and CCL20.⁸² Both IL-17 and IL-22 also disrupt the tight junctions that form the blood–brain barrier, enabling Th17 cells to migrate into the central nervous system and cause neuronal damage.⁸³

In summary, increasing numbers of autoimmune conditions implicate the Th17 pathway. Although the specific genes and mechanisms exhibit some variation depending on the location, chronicity and type of disease, there are many common threads in terms of IL-17-mediated gene expression. These findings have made blocking the IL-17 pathway an attractive _target for anti-cytokine therapy.²² On the flip side, the same pathways that promote disease in autoimmunity are beneficial in many infection settings, which will be considered in the following sections.

Extracellular bacterial infections