Numerous cytokines are associated with the pathogenesis of rheumatoid arthritis (RA), including tumor necrosis factor (TNF), interferon-γ (IFNγ), interleukin-1 (IL-1), IL-6, IL-15, and IL-17. IL-17 is a proinflammatory cytokine that is secreted by T lymphocytes (1). Ligation of the IL-17 receptor, which is expressed on several cell types, including epithelial cells, endothelial cells, and fibroblasts, induces the secretion of IL-6, IL-8, granulocyte colony-stimulating factor (G-CSF), monocyte chemotactic protein 1, prostaglandin E2 (PGE2), TNFα, and IL-1β (2–5), as well as neutrophil chemotaxis and granulopoiesis (6, 7). IL-17 induces the expression of matrix metalloproteinase 1 (MMP-1) and MMP-13 in RA synovial cells and osteoblasts (8, 9). IL-17 is present in RA synovium (10) and induces the expression of RANKL, which contributes to bone resorption (11, 12). Local overexpression of IL-17 increases the severity of murine arthritis, and neutralizing anti–IL-17 antibody reduces the severity of arthritis (13, 14). IL-17–deficient mice have a reduced incidence and severity of collagen-induced arthritis (CIA) (15). Taken together, the available data suggest that IL-17 is important in both CIA and RA.
IL-23, a dimeric cytokine composed of the IL-12 p40 subunit and a unique p19 subunit (16), is secreted by dendritic cells (DCs) and induces the production of IL-17 by T cells (17, 18). IL-23 p19–/– mice are resistant to experimental autoimmune encephalomyelitis and CIA, suggesting that this cytokine is important in autoimmune diseases (19, 20). There is some evidence that costimulatory molecules may also be involved in the regulation of IL-17, since IL-17 production is reduced in ICOS–/– mice (21, 22).
DCs are professional antigen-presenting cells that are capable of activating and regulating T lymphocytes (23). Previous studies in our laboratory have shown that a single injection of bone marrow–derived DCs genetically engineered to express IL-4 reduces the incidence and severity of CIA and alters primary immune responses in vivo (24, 25). Such effects were not produced by T cells or by fibroblasts overexpressing IL-4, which suggests that the migratory properties of DCs are required. Although genetically modified DCs were only detectable in the spleen for a few days following injection, the effects of IL-4 DCs were long-lasting. Soluble IL-4 has a short half-life and has to be administered by continuous infusion or by repeated injections, whereas IL-4 DCs can be administered as a single injection and are effective in CIA, making this an attractive therapeutic approach. The molecular mechanisms by which IL-4 DCs alter an ongoing immune response in vivo are not yet understood in detail, although previous studies have shown that the polyclonal response of splenic T cells shifts to Th2 after administration of IL-4 DCs. However, there is no evidence for the suppression of IL-12 expression in IL-4 DCs despite a down-regulation of the expression of the IL-23 p19 subunit (25).
RA has been viewed as a Th1 disease, mediated in part by the expression and biologic effects of IFNγ. IFNγ, but not IL-4, is detected in the RA synovium (26). However, deficiency of either IFNγ or the IFNγ receptor causes an acceleration of CIA in mice (27). Furthermore, treatment of CIA with IL-4 DCs did not appear to deplete primed collagen II–specific Th1 cells in vivo (24). The existence of a third Th population, a distinct subset of IL-17–producing T cells (Th17), was recently proposed, which may be the pathogenic population in some autoimmune diseases (15, 17, 28, 29). The present studies were performed to further our understanding of the mechanisms underlying the response to IL-4 DCs in CIA. We evaluated T lymphocyte responses to IL-4 DCs during different phases of arthritis, and we found that IL-4 DCs selectively regulate the production of IL-17.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
The Th1/Th2 paradigm has recently been modified to include a third type of differentiated T helper cell, called Th17, which secretes IL-17. Th17 cells have a distinct lineage and do not share a developmental pathway with Th1 or Th2 cells (28, 29). IL-17 receptor is ubiquitously expressed (1). IL-17 stimulates the production of IL-1β, TNF, IL-6, IL-8, G-CSF, and PGE2 from macrophages and synoviocytes (2, 5, 31).
IL-23 augments IL-17 production by memory Th17 cells, but is itself not sufficient to generate Th17 cells. Data from transgenic mice expressing a T cell receptor for ovalbumin have shown that neutralization of IFNγ and IL-4, in addition to the presence of IL-23, enhances the generation of Th17 cells from naive CD4 T cells in vitro. Once the Th17 phenotype is stabilized, neutralization of IFNγ or IL-4 does not alter the IL-17 response (28). Th17 cells develop independently of STAT-1, STAT-4, and STAT-6 signaling and the transcription factors T-bet and c-Maf (28, 29). Experiments done with CD28- or ICOS-knockout mice have shown that these costimulatory molecules also play a role in the generation of Th17 cells (29). Recent studies have shown that IL-6, in combination with transforming growth factor β (TGFβ), inhibits the generation of FoxP3-expressing T regulatory cells and induces the generation of Th17 cells (32–34). TGFβ is not only critical to the development of Th17 cells, but it also induces the expression of IL-23 receptor, which facilitates Th17 survival and expansion (34). These studies provide evidence that Th1, Th2, and Th17 cells develop from naive T cells and that the generation of T regulatory cells and Th17 cells occurs via alternative pathways, which are selected according to the presence or absence of IL-6.
IL-17 has been detected and has been proposed to be pathogenic in systemic sclerosis (35), systemic lupus erythematosus (36), and RA (10). Anti–IL-17 antibody was shown to reduce the incidence and severity of experimental autoimmune encephalomyelitis, even though the splenic and lymph node cells obtained from the mice retained the capacity to produce IL-17. There was a lack of recruitment of Th17 cells into the brain secondary to a reduced expression of CCL2, CCL17, and CXCL1, suggesting that IL-17 can increase the expression of several lymphotactic chemokines. Transgenic mice that overexpressed IL-17 in the lung epithelium also had increased levels of several chemokines (29).
Neutralization of IL-17 has been shown to suppress the onset, and decrease the severity, of arthritis in animal models of RA (13, 14, 37). IL-17–/– mice have reduced collagen-specific T cell responses during the initiation phase of arthritis and a reduced incidence and intensity of arthritis (15). Overexpression of IL-4 has been shown to down-regulate IL-17 and prevent cartilage destruction and bone erosion (38, 39).
Investigators in our laboratory have previously shown that when bone marrow–derived DCs that overexpress IL-4 are injected into mice during the initiation phase of CIA, the incidence and severity of arthritis are reduced and that IL-4 DCs differentially regulate the production of IL-12/IL-23 after stimulation through their CD40 receptors. Since IL-17–knockout mice have a reduced incidence of arthritis and since local overexpression of IL-4 down-regulates IL-17 expression in the joints, we hypothesized that IL-4 DCs may be reducing the incidence of arthritis by decreasing the production of IL-17 by collagen-specific T cells. To study this, we isolated splenic T lymphocytes from DBA mice 15 days after collagen immunization and cultured them in the presence of untransduced or IL-4–transduced DCs in a collagen rechallenge response. IL-17 was measured in the culture supernatants by ELISA or ELISpot. Our data showed that T cells cultured in the presence of IL-4 DCs produced less IL-17 and similar quantities of IFNγ in a collagen-specific response. This suggests that IL-17 is a cytokine of paramount importance in the early response to collagen and that IL-4 DCs reduce the incidence and severity of arthritis by decreasing the IL-17 response in T cells, even when the IFNγ response is fixed. It is likely that this muted IL-17 response during the early stages of arthritis translates into a reduced incidence and severity of arthritis. A decrease in both IL-17 and IFNγ was only seen when the polyclonal stimulus anti-CD3 was used. These effects were not limited to splenic T cells, but were also seen in lymphocytes from draining lymph nodes.
IL-4 DCs produce excessive IL-4, and we evaluated the effect of neutralizing IL-4 in cultures of IL-4 DCs or adding IL-4 to cultures with untransduced DCs. Our results showed that IL-4 can decrease the production of IL-17 by T cells cultured with DCs and collagen. Recent studies by Harrington et al (28) and Park et al (29), which were published when the experiments presented herein were near completion, have also shown that IL-4 can influence the production of IL-17, in that neutralizing IL-4 can lead to increased levels of IL-17. In our studies, neutralizing IL-4 in T cells cultured with IL-4 DCs resulted in reconstitution of the abrogated IL-17 response to collagen.
IL-23 has been implicated as a primary driver of IL-17 production by memory Th17 cells (17, 18). We hypothesized that IL-4 may be reducing the production of IL-17 through an IL-23–dependent mechanism, possibly by down-regulating the production of IL-23 by DCs. If this were the case, then adding recombinant IL-23 would increase the production of IL-17. However, we found that IL-23 at 10 ng/ml did not alter IL-17 levels and that IL-23 at 100 ng/ml only modestly increased IL-17 production in the presence of IL-4 DCs. This amount of IL-23 is several hundred times higher than the amount of IL-4 present in the cultures. When the same amount of IL-23 was added to T cells cultured with untransduced DCs and collagen, we observed a much higher production of IL-17. This suggests that the IL-4–mediated suppression of IL-17 is not due to suppression of IL-23 production by IL-4 DCs. The data also indicate that IL-4 renders the T cells resistant to IL-23, at least with respect to IL-17 production, and that this resistance is not overcome by excess concentrations of IL-23. Primed T cells cultured with untransduced DCs and collagen showed high levels of IL-17 production, which was suppressed by IL-4 and was not reversed by the addition of IL-23. These results provide further proof of IL-23 resistance in the presence of IL-4 and suggest that down-regulation of IL-17 by IL-4–transduced DCs is mediated by IL-4 and not by the retroviral transduction of DCs.
Furthermore, the increase in IL-17 production after neutralization of IL-4 was only modestly suppressed by the addition of anti-p19 antibody, thus suggesting that T cell production of IL-17 is only partly mediated by IL-23 in T cells cultured with IL-4 DCs or with untransduced DCs. It is likely that cognate costimulatory molecules play a role in the production of IL-17 by primed T cells obtained during the early stage of CIA.
IL-12, which shares a common subunit with IL-23, drives IFNγ production in T cells (40). Neutralizing IFNγ has been shown to increase the generation of Th17 cells (28, 29). IL-4 DCs produce increased amounts of IL-12 and IFNγ after stimulation with lipopolysaccharide or CD40L (25). We also evaluated the effect of IL-12 and IFNγ in our system. We found that IL-12 suppressed IL-17 production, but the effect of IL-4 DCs on IL-17 production was independent of IL-12, since the IL-17 response was not restored by blocking IL-12. Neutralizing IFNγ did not reconstitute the IL-17 response, which provides further evidence that the IL-4 DC–mediated suppression of IL-17 is robust and dominant over IFNγ and is not secondary to excess IFNγ that is produced by IL-4 DCs or by T cells influenced by IL-12.
The mechanism underlying the IL-4–mediated suppression of IL-17 is not known. However, it is robust and dominant over IL-23 and IFNγ. It is possible that IL-4 down-regulates the expression of IL-23 receptor or induces T regulatory cells, which then suppress IL-17 production. It is also possible that IL-4 blocks the costimulatory molecule–mediated production of IL-17 by down-regulating receptor/ligand expression or function. IL-4 binding to its receptor on T cells could block intracellular signaling events that would otherwise have led to the production of IL-17.
T cells from mice with end-stage arthritis failed to decrease their production of IL-17 when cultured in the presence of IL-4 DCs. This suggests that while IL-4 DCs modulate the early initiation phase of arthritis by decreasing IL-17 production by T cells, they are unable to elicit such a response during the late phase of arthritis. It has been shown that IL-4 DCs can suppress CIA early after its clinical onset (41). Our data seem not to be in complete agreement with those obtained in previous studies by Lubberts et al (39), who reported that local injection of a viral vector expressing IL-4 resulted in the amelioration of established inflammation and the reduction of IL-17 in the injected joint. It is possible, however, that the systemic IL-17 response in mice with established arthritis is regulated differently by IL-4 than is the local expression of IL-17 in arthritic joints.
The cytokine imbalance in CIA is difficult to explain by the prevailing Th1/Th2 paradigm and, instead, is likely to involve other cytokines, especially IL-17. The data currently available suggest that the IL-17–producing cells remain sensitive to IL-4 DCs for a limited interval following the appearance of joint inflammation. It is possible that, while the IL-17 response during the early initiation phase of CIA is amenable to regulation by IL-4, the IL-17 response during end-stage arthritis is mediated by a distinct set of costimulatory signals that cannot be overcome by IL-4. Alternatively, the IL-17–producing T cells in end-stage arthritis are at a point in their differentiation at which IL-4 receptor signaling is defective.
Distinct windows appear to exist in the IL-17 response from T cells in arthritis, during which it can be modulated by changing the cytokine environment. Once the IL-17 response is stabilized, as in end-stage arthritis, then such changes are not possible or, at least, are more difficult to accomplish. This paradigm helps to explain the early events in immune-mediated arthritis and possible ways to modulate key pathogenic responses. Such an understanding will help in the development of novel therapies targeted to IL-17 in RA. Further studies evaluating the role of costimulatory molecules in the generation and regulation of IL-17–producing T cells, as well as the in vivo IL-17 response in CIA and its modulation by administration of IL-4 DCs, are ongoing.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Dr. Fox had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Dr. Sarkar, Ms Tesmer, and Dr. Fox.
Acquisition of data. Dr. Sarkar, Ms Tesmer, Dr. Hindnavis, and Ms Endres.
Analysis and interpretation of data. Dr. Sarkar, Ms Tesmer, Dr. Hindnavis, Ms Endres, and Dr. Fox.
Manuscript preparation. Dr. Sarkar, Ms Tesmer, and Dr. Fox.
Statistical analysis. Dr. Sarkar.