Drs. Kwok and Cho contributed equally to this work.
Interleukin-21 promotes osteoclastogenesis in humans with rheumatoid arthritis and in mice with collagen-induced arthritis
Article first published online: 28 FEB 2012
Copyright © 2012 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 64, Issue 3, pages 740–751, March 2012
How to Cite
Kwok, S.-K., Cho, M.-L., Park, M.-K., Oh, H.-J., Park, J.-S., Her, Y.-M., Lee, S.-Y., Youn, J., Ju, J. H., Park, K. S., Kim, S.-I., Kim, H.-Y. and Park, S.-H. (2012), Interleukin-21 promotes osteoclastogenesis in humans with rheumatoid arthritis and in mice with collagen-induced arthritis. Arthritis & Rheumatism, 64: 740–751. doi: 10.1002/art.33390
- Issue published online: 28 FEB 2012
- Article first published online: 28 FEB 2012
- Accepted manuscript online: 3 OCT 2011 09:30AM EST
- Manuscript Accepted: 11 SEP 2011
- Manuscript Received: 27 SEP 2010
- Basic Science Research Program through the National Research Foundation of Korea
- Ministry of Education, Science and Technology. Grant Number: 20008-2005645
- Korea Health Technology R&D Project
- Ministry for Health, Welfare & Family Affairs, Republic of Korea. Grant Number: A092258
Bone destruction is a critical pathology involved in the functional disability caused by rheumatoid arthritis (RA). Osteoclasts, which are specialized bone-resorbing cells regulated by cytokines such as RANKL, are implicated in bone destruction in RA. The aim of this study was to determine whether interleukin-21 (IL-21), a potent immunomodulatory 4–α-helical bundle type 1 cytokine, has osteoclastogenic activity in patients with RA and in mice with collagen-induced arthritis (CIA).
The expression of IL-21 in synovial tissue was examined using immunohistochemistry. The concentrations of IL-21 in serum and synovial fluid were determined by enzyme-linked immunosorbent assay. The levels of RANKL and osteoclastogenic markers were measured using real-time polymerase chain reaction. CD14+ monocytes from patients with RA or mouse bone marrow cells were cocultured with fibroblast-like synoviocytes (FLS) from patients with RA or CD4+ T cells from mice with CIA in the presence of IL-21 and subsequently stained for tartrate-resistant acid phosphatase activity to determine osteoclast formation.
IL-21 was up-regulated in the synovium, synovial fluid, and serum of patients with RA and in the synovium and serum of mice with CIA. IL-21 induced RANKL expression in mixed joint cells and CD4+ T cells from mice with CIA and in CD4+ T cells and FLS from patients with RA. Moreover, IL-21 enhanced in vitro osteoclastogenesis without the presence of RANKL-providing cells and by inducing RANKL expression in CD4+ T cells and FLS.
Our data suggest that IL-21 promotes osteoclastogenesis in RA. We believe that therapeutic strategies targeting IL-21 might be effective for the treatment of patients with RA, especially in preventing bone destruction.
Rheumatoid arthritis (RA) is a multisystem autoimmune disease of unknown etiology that is characterized by a hyperplastic synovial membrane, also called pannus, that is capable of destroying adjacent articular cartilage and bone (1, 2). The pathology in the synovial membrane of patients with RA includes infiltration of inflammatory leukocytes, proliferation of synoviocytes, and extensive angiogenesis, which are collectively referred to as rheumatoid pannus (2–4). Among the various pathologic events occurring in affected joints, bone destruction is extremely important clinically, because it is related to functional impairment and the progression of joint damage in patients with RA in prolonged remission (5). Osteoclasts, which are specialized bone-resorbing cells regulated by RANKL and macrophage colony-stimulating factor (M-CSF), are primarily implicated in bony erosion in RA (6, 7). In the microenvironment of inflamed RA joints, local production of proinflammatory cytokines (interleukin-1 [IL-1], tumor necrosis factor α [TNFα], IL-6, and IL-17) as well as RANKL provided by activated CD4+ T cells and synoviocytes leads to stimulation of osteoclastogenesis and bone destruction (8, 9).
IL-21, a potent immunomodulatory 4–α-helical bundle type 1 cytokine, is produced by CD4+ T cells and natural killer (NK) T cells and has pleiotropic effects on both innate and adaptive immune responses (10). IL-21 is also produced by Th17 cells and plays a critical role in their regulation (11, 12). Th17 cells comprise a proinflammatory helper T cell subset implicated in osteoclastogenesis and bone destruction (13) as well as in the development of RA (14, 15). Therefore, it is conceivable that IL-21 might have some role in RA-related osteoclastogenesis. However, the precise impact of IL-21 on osteoclastogenesis in RA remains to be determined.
To determine the effect of IL-21 on osteoclastogenesis in RA, we investigated the expression of IL-21 in the synovium of patients with RA and in mice with collagen-induced arthritis (CIA). We also examined whether IL-21 would induce the expression of RANKL, the leading player in osteoclastogenesis (16), in cultured mixed joint cells or CD4+ T cells from mice with CIA and in fibroblast-like synoviocytes (FLS) or CD4+ T cells from patients with RA. We studied the osteoclastogenic potential of IL-21 on monocytes from patients with RA and on bone marrow cells from mice with CIA, using tartrate-resistant acid phosphatase (TRAP) staining for osteoclasts and quantitative assays to measure osteoclast markers. Our results demonstrated that IL-21 promotes osteoclastogenesis in the absence of RANKL-providing cells as well as indirectly by possibly inducing RANKL expression in CD4+ T cells from mice with CIA and in FLS from patients with RA. This suggests that a strategy targeting IL-21 might be effective in treating patients with RA, especially in terms of preventing bone destruction.
MATERIALS AND METHODS
Four- to six-week-old male DBA/1J mice were purchased from SLC Inc. The mice were maintained under specific pathogen–free conditions at the Institute of Medical Science of the Catholic University of Korea and were fed standard mouse chow (Ralston Purina) and water ad libitum. All experimental procedures were approved by the Animal Research Ethics Committee of the Catholic University of Korea, which conforms to all US National Institutes of Health guidelines.
Induction of CIA.
To induce CIA in mice, 0.1 ml of an emulsion containing 100 μg of bovine type II collagen and Freund's complete adjuvant (Arthrogen-CIA) was injected intradermally into the base of the tail as a primary immunization. Two weeks later, 100 μg of type II collagen dissolved and emulsified 1:1 with Freund's incomplete adjuvant (Difco) was administered to the hind leg as a booster injection.
Preparation of mixed arthritic joint cells.
The ankle joints of 6- to 8-week-old mice with CIA were skinned and digested with Liberase (0.1 mg/ml) and DNase I (1 mg/ml) (Roche Diagnostics) for 60 minutes at 37°C. A 70-μm nylon cell strainer (BD Falcon) was subsequently used to process the digested tissue. Joint cells (1 × 106/well) were cultured onto 24-well plates in RPMI 1640 medium containing 10% fetal bovine serum (FBS) (Gibco) and incubated for 12 hours in the presence of various concentrations of IL-21, with or without TNFα or IL-17.
Preparation of mouse CD4+ T cells.
Mouse spleens were sieved through a mesh, and red blood cells were lysed with 0.83% ammonium chloride. The remaining spleen cells were maintained in RPMI 1640 medium containing 10% FBS. Anti-CD4 microbeads were used as recommended by the manufacturer (Miltenyi Biotec). The cell suspension (1 × 106 cells/well) was dispensed into 24-well plates (Nunc) and incubated at 37°C in an atmosphere of 5% CO2.
Mouse in vitro osteoclastogenesis.
Bone marrow–derived monocyte/macrophages were isolated as described previously (17). Osteoclast precursor cells (1 × 105/well) were further cultured onto 48-well plates in the presence of 10 ng/ml recombinant human M-CSF, 100 ng/ml soluble RANKL (sRANKL; PeproTech), or various concentrations of IL-21 for 4 days to generate osteoclasts.
Preparation of human CD4+ T cells and synovial fibroblast isolation.
This study was approved by the Institutional Review Committee of the Catholic Medical Center, Catholic University of Korea. Informed consent for the use of human mononuclear cells was obtained from all study subjects. Peripheral blood was obtained using a heparin-treated syringe. Peripheral blood mononuclear cells (PBMCs) were isolated by density-gradient centrifugation using Ficoll-Hypaque (Pharmacia LKB). To purify CD4+ T cells, anti-CD4 microbeads were used as recommended by the manufacturer (Miltenyi Biotec). FLS from patients with RA were obtained as described previously (18), and FLS from passages 4–8 were used for the experiments.
Mouse joint tissue and RA synovium were fixed in 4% paraformaldehyde, decalcified in EDTA bone decalcifier, embedded in paraffin, and sectioned as described previously (17). The sections were stained with hematoxylin and eosin, Safranin O, and toluidine blue to detect proteoglycans. Immunostaining for IL-21 (1 μg/ml), RANKL (1 μg/ml), CD4 (0.5 μg/ml), IL-17 (2 μg/ml), and TNFα antibodies (1 μg/ml) (all from Santa Cruz Biotechnology) was performed in mouse joints. Anti–IL-21 (1 μg/ml), RANKL antibody (1 μg/ml) (both from Santa Cruz Biotechnology), and anti–IL-21 receptor (anti–IL-21R) antibody (0.5 μg/ml) (R&D Systems) were immunostained on synovial samples from patients with RA (the group with mild disease and the group with severe disease) and patients with osteoarthritis (OA). The sections were counterstained with hematoxylin. Samples were photographed with an Olympus photomicroscope.
Human in vitro osteoclastogenesis.
PBMCs obtained from healthy human volunteers were separated from buffy coats using Ficoll-Hypaque. Red blood cell–free cells were seeded onto 24-well plates at 5 × 105 cells/well and incubated at 37°C for 2 hours to separate the floating and adherent cells. The adherent cells were washed with sterile phosphate buffered saline (PBS) (Gibco) and cultured with 100 ng/ml M-CSF for 3 days. After 3 days, the preosteoclast cells were further cultured in the presence of 25 ng/ml M-CSF, 30 ng/ml RANKL, and various concentrations of IL-21 for 11 days, to generate osteoclasts. On day 3, the medium was replaced with fresh medium containing M-CSF, RANKL, and IL-21.
Cocultures with human osteoclast precursor cells and FLS.
RA FLS from passages 4–8 were seeded on 60-mm dishes at 1 × 105 cells/dish with Dulbecco's modified Eagle's medium containing 10% FBS overnight to attach on the well, and the cells were serum-starved with 1× insulin–transferrin–selenium A (Invitrogen) for 12 hours. After a wash with PBS, cells were stimulated with various concentrations of IL-21 for 2 days and then detached using trypsin–EDTA, and the medium was changed to α-minimum essential medium containing 10% FBS. Next, FLS (5 × 102 cells/well) were cocultured with human osteoclast precursor cells, which were prepared on 12-well plates (Nalge Nunc International) as described above, with 25 ng/ml M-CSF and 10 ng/ml RANKL.
A commercial TRAP kit (Sigma-Aldrich) was used according to the manufacturer's instructions, omitting counterstaining with hematoxylin. TRAP-positive multinucleated cells were counted 3 times, without knowledge of the previously counted numbers of osteoclasts.
Bone resorption analysis.
Mouse bone marrow–derived monocyte/macrophages and human PBMCs prepared using the method described above were cultured in 96-well dentine discs (Immunodiagnostic Systems) according to the manufacturer's instructions. Cells on dentin were removed, and the dentin slices were immersed in Mayer's hematoxylin (Muto Glass) to stain resorption pits formed by the mature osteoclasts. The erosive areas were identified using the analySIS TS Lite program (Olympus).
Gene expression analysis using real-time polymerase chain reaction (PCR).
PCR amplification and analysis were achieved using a LightCycler 2.0 instrument (Roche Diagnostics) with version 4.0 software. All reactions were performed using LightCycler FastStart DNA Master SYBR green I (Takara), according to the manufacturer's instructions. The level of messenger RNA (mRNA) expression was normalized to β-actin expression levels.
Enzyme-linked immunosorbent assay (ELISA) of IL-21 and sRANKL.
Levels of IL-21 or sRANKL were measured with ELISA kits for human IL-21 (eBioscience), mouse IL-21 (eBioscience), and mouse sRANKL (R&D Systems), according to the manufacturers' instructions.
Cells were harvested and lysed with lysis buffer. Protein concentrations were determined using the Bradford method (Bio-Rad). Protein samples were separated using 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes (Amersham Pharmacia Biotech). For Western hybridization, the membrane was preincubated with blocking buffer for 2 hours, followed by incubation with primary antibodies to RANKL (1 μg/ml) and β-actin for 1 hour. After washing, horseradish peroxidase–conjugated secondary antibodies were added, and the membranes were incubated for 1 hour at room temperature. After washing, the hybridized bands were detected using enhanced chemiluminescence detection kits (Pierce) and Hyperfilm.
Data are presented as the mean ± SD. Statistical significance was determined using Student's t-tests with SPSS version 10.0 software. P values less than 0.05 were considered significant.
High expression of IL-21 in the synovium of mice with CIA.
We first performed immunohistochemical staining to investigate the distribution and localization of IL-21 in joint tissue from mice with CIA and wild-type (WT) mice. Figure 1A shows that osteoclast formation as well as marked destruction of cartilage and bone were observed in mice with CIA. As shown in Figure 1B, increased expression of IL-21 was observed in all synovial samples from mice with CIA; positive staining was mainly observed in the pannus of the hyperplastic synovium. However, IL-21 was rarely expressed in the joints of WT mice. Immunohistochemical staining for RANKL, TNFα, and IL-17 was also performed on joint tissue specimens from mice with CIA. As was observed for IL-21, the expression of RANKL, IL-17, and TNFα was also increased in the inflamed synovium of mice with CIA (Figure 1B). We also measured serum levels of IL-21 and sRANKL in mice with CIA (n = 7) and WT mice (n = 5). As shown in Figure 1C, the expression of both IL-21 and sRANKL was significantly higher in mice with CIA than in WT mice (for IL-21, mean ± SD 165 ± 56 pg/ml versus 23 ± 11 pg/ml [P < 0.01]; for sRANKL, 245 ± 87 pg/ml versus 17 ± 2 pg/ml [P < 0.01]).
IL-21–enhanced RANKL expression in cultured mixed joint cells and CD4+ T cells from mice with CIA.
Because RANKL is known to be the prototype mediator of osteoclastogenesis (16), we investigated whether IL-21 could induce the expression of RANKL in vitro. We isolated mixed joint cells from mice with CIA and then added various proinflammatory cytokines, including IL-21, to the mixed joint cell cultures and analyzed the expression of RANKL mRNA, by real-time PCR. As seen in Figure 2A, IL-21 dose-dependently increased the expression of RANKL. Treatment with IL-21 in combination with other proinflammatory cytokines, such as TNFα and IL-17, had an additive effect on RANKL expression in mixed joint cells (Figure 2B).
It is well known that CD4+ T cells express RANKL and can promote osteoclastogenesis (19). Therefore, we isolated splenic CD4+ T cells from mice with CIA and cultured them with various concentrations of IL-21 and/or TNFα. As shown in Figure 2C, IL-21 dose-dependently increased the expression of RANKL, as determined by real-time PCR. Treatment with IL-21 plus TNFα had an additive effect on RANKL expression in CD4+ T cells (Figure 2D). We also demonstrated that IL-21 increased the expression of RANKL in the splenocytes of mice with CIA, as determined by Western blot analysis (Figure 2E).
Requirement of STAT-3 signaling for IL-21–stimulated RANKL expression.
STAT-3 appears to be required for IL-21 signaling (10). Based on this finding, we investigated the signaling pathway that mediates IL-21–induced RANKL expression. Splenic CD4+ T cells isolated from mice with CIA were pretreated for 30 minutes with various signaling-pathway inhibitors (AG490 [JAK-2/STAT-3 inhibitor; 5 μM], LY294002 [phosphatidylinositol 3-kinase/Akt inhibitor; 20 μM], SP600125 [activator protein 1 inhibitor; 1 μM], and PD98059 [ERK inhibitor; 10 μM]) and then cultured with IL-21 (10 ng/ml) for 12 hours. The expression of RANKL was determined by real-time PCR. As expected, the STAT-3 inhibitor AG490 suppressed IL-21–induced RANKL expression in CD4+ T cells (Figure 2F), while the other signaling pathway inhibitors had no significant effect (data not shown). Similar results were observed when the experiments were performed using mixed joint cells (Figure 2G). Thus, these findings demonstrated that IL-21–stimulated RANKL expression is mediated by STAT-3.
IL-21–induced stimulation of osteoclastogenesis in mice with CIA.
Activated CD4+ T cells can express RANKL and have the capacity to induce osteoclastogenesis (17, 19). It is conceivable that IL-21 has osteoclastogenic potential, because IL-21 induced RANKL expression by CD4+ T cells in mice with CIA (Figures 2C and D). Therefore, to investigate the functional potency of IL-21–induced RANKL expression by CD4+ T cells to induce bone marrow cells to differentiate into mature osteoclasts, we added CD4+ T cells and IL-21–stimulated CD4+ T cells to bone marrow cell cultures. After 7 days of culture, multinucleated mature osteoclasts had differentiated from the monocytes in the wells that contained either CD4+ T cells or IL-21–stimulated CD4+ T cells, as determined by the TRAP staining assay. As expected, the osteoclastogenic effect of IL-21–stimulated CD4+ T cells was more potent than that of CD4+ T cells (Figure 3A). In addition, compared with CD4+ T cells, IL-21–stimulated CD4+ T cells significantly increased the expression of osteoclastogenic markers such as calcitonin receptor and cathepsin K (Figure 3B).
Our next experiment was conducted to examine whether IL-21 has an osteoclastogenic impact without stimulation by RANKL-providing cells such as CD4+ T cells. Bone marrow cells were prepared from mice with CIA and then stimulated with IL-21, M-CSF, and RANKL to induce osteoclastogenesis. IL-21 alone failed to induce osteoclastogenesis, as did M-CSF (data not shown). However, to our surprise, treatment with IL-21 increased the number of differentiated osteoclasts as much as did M-CSF and high-dose RANKL (100 ng/ml in mice). These findings were evaluated by counting the number of TRAP-positive cells per well (Figure 3C).
We also performed bone resorption assays to confirm the osteoclastogenic activity of IL-21. As shown in Figure 3D, the bone resorption assay demonstrated that IL-21 promoted osteoclastogenesis in the presence of M-CSF and low-dose RANKL (10 ng/ml). The expression levels of various markers of osteoclastogenesis such as calcitonin receptor, cathepsin K, and matrix metalloproteinase 9 (MMP-9) were also measured by real-time PCR (Figure 3E). IL-21 increased the expression of various osteoclastogenic markers dose-dependently in the presence of M-CSF and low-dose RANKL (10 ng/ml). Collectively, these findings suggest that IL-21 promotes osteoclastogenesis both in the absence of RANKL-providing cells and via up-regulation of RANKL expression by CD4+ T cells.
IL-21–enhanced RANKL expression in CD4+ T cells and FLS from patients with RA.
We examined whether IL-21 could induce RANKL expression in human CD4+ T cells. As was observed with mouse cells, IL-21 dose-dependently increased the expression of RANKL in CD4+ T cells from patients with RA and patients with OA (Figure 4A). Interestingly, the expression of RANKL in CD4+ T cells from patients with RA was markedly higher than that in cells from patients with OA.
Rheumatoid synoviocytes are also known to express RANKL and thereby promote osteoclastogenesis (20, 21). Therefore, we next investigated whether IL-21 could induce RANKL expression in synoviocytes from patients with RA. As expected, IL-21 dose-dependently increased RANKL expression in cultured RA FLS (n = 4) (Figure 4B). IL-21 also induced the production of diverse MMPs (MMP-1, MMP-3, MMP-9, and MMP-13). Moreover, treatment with IL-21 combined with other proinflammatory cytokines such as TNFα, IL-17, and IL-22 showed an additive or synergistic effect on RANKL expression in cultured RA FLS (Figure 4C).
Impact of IL-21–stimulated osteoclastogenesis on human RA.
We investigated whether IL-21 could promote osteoclastogenesis in patients with RA. IL-21 stimulated osteoclastogenesis indirectly, possibly through up-regulation of RANKL in FLS, because the osteoclastogenic potential of IL-21–stimulated FLS was more potent than that of unstimulated FLS (Figure 5A). In addition, compared with unstimulated FLS, IL-21–stimulated FLS more significantly increased the expression of osteoclastogenic markers such as calcitonin receptor and cathepsin K (Figure 5B). Moreover, in the presence of low-dose RANKL and M-CSF, IL-21 had an osteoclastogenic potential to cause monocytes from patients with RA to differentiate into mature osteoclasts in the absence of RANKL-providing cells such as CD4+ T cells and FLS (Figures 5C and D). Furthermore, IL-21 dose-dependently increased the expression of osteoclastogenic markers such as calcitonin receptor and cathepsin K in the presence of low doses of RANKL and M-CSF (Figure 5E).
High expression of IL-21 in the synovium, synovial fluid, and serum of patients with RA.
Finally, we performed immunohistochemical staining of the synovium of 8 patients with RA (4 with mild disease and 4 with severe disease), using antibodies to IL-21, RANKL, and IL-21R (Figure 6A). As was observed with RANKL, IL-21 was highly expressed in all synovial tissue sections from patients with RA. The expression of IL-21, RANKL, and IL-21R was higher in the synovium of patients with severe RA compared with patients with mild RA and patients with OA. Moreover, as shown in Figure 6B, IL-21 concentrations were significantly higher in the synovial fluid of patients with RA (n = 30) than in the synovial fluid of patients with OA (n = 29) (mean ± SD 964.8 ± 560.1 pg/ml and 584.7 ± 309.8 pg/ml, respectively; P < 0.001).
One of the major cell types comprising the RA synovium is the synoviocyte. Therefore, we used an ELISA to measure IL-21 production in cultured FLS. However, similar to what was observed in a previous study (22), IL-21 was not expressed by cultured RA FLS (data not shown). We also examined serum IL-21 concentrations, which were shown to be significantly higher in patients with RA (n = 30) than in patients with OA (n = 29) (mean ± SD 332.1 ± 121 pg/ml and 131.5 ± 56 pg/ml, respectively; P < 0.01).
IL-21 belongs to a family of cytokines, all of which bind to a compound receptor that includes the common cytokine receptor γ chain. IL-21 has diverse biologic effects on different target cells. One of its essential roles is the promotion of B cell activation, differentiation to plasma cells, and immunoglobulin production (23–25). However, in addition to its role in humoral immunity, IL-21 has a variety of immunomodulatory effects. It enhances the proliferation of lymphoid cells, increasing the cytotoxicity of CD8+ T cells and NK cells. Conversely, IL-21 inhibits the function of dendritic cells and can induce apoptosis in B cells and NK cells (10). IL-21 is also produced by Th17 cells and critically regulates Th17 cell development (11, 12). Given these pleiotropic effects on diverse immune cells, the net immunomodulatory activity of IL-21 would be determined by the differentiation state of target cells as well as by other cytokines or costimulatory molecules.
Several studies have shown that IL-21 production is associated with the development of certain autoimmune diseases such as systemic lupus erythematosus (26–28), RA (29–31), and primary Sjögren's syndrome (32). Regarding the role of IL-21 in inflammatory forms of arthritis such as RA, studies in both mice and humans suggest that IL-21 might be involved in disease progression (29–31). Young et al demonstrated that blockade of IL-21 with the IL-21R–Fc fusion protein ameliorated clinical disease activity in animal models of inflammatory arthritis, suggesting that IL-21 contributes to the pathologic process of RA (30). Jang et al reported that IL-21 signaling is essential for the development of inflammatory arthritis in the autoimmune K/BxN mouse, a model of RA that is thought to arise mainly from autoantibody-mediated inflammatory responses (31). Those investigators demonstrated that IL-21R–deficient K/BxN mice were “completely refractory to the development of spontaneous arthritis.” They also showed that IL-21 up-regulated the expression of RANKL in splenocytes from K/BxN mice, and few RANKL-expressing infiltrates were observed in the synovium of IL-21R–deficient K/BxN mice, which suggests that IL-21 might regulate osteoclastogenesis. However, the exact role of IL-21 in osteoclastogenesis remains to be clarified.
In the current study, we demonstrated that IL-21 up-regulated the expression of RANKL—the key osteoclastogenic molecule expressed by osteoclastogenesis-supporting cells—in cultured mixed joint cells and CD4+ T cells from mice with CIA (Figures 2A and C) and in CD4+ T cells and FLS from patients with RA (Figures 4A and B). To our surprise, IL-21 more profoundly up-regulated RANKL expression in CD4+ T cells from patients with RA than in those from patients with OA. A previous study showed that the CD4+ T cells of patients with RA expressed significantly higher levels of IL-21R than did those from patients with OA (29), and we observed the same result (data not shown). This differential expression level of IL-21R in CD4+ T cells between patients with RA and patients with OA might explain the different degree of RANKL up-regulation in CD4+ T cells stimulated by IL-21.
Li et al reported that blood and synovial fluid T cells from patients with RA produced higher levels of TNFα and interferon-γ after stimulation with anti-CD3 and IL-21 than did T cells from patients with OA and healthy controls, suggesting that IL-21 contributes to the progression of RA by up-regulating the expression of proinflammatory cytokines (29). Therefore, it is conceivable that the IL-21–induced RANKL expression of CD4+ T cells might be partially mediated by the TNFα induced by IL-21. However, in our experiments, treatment with IL-21 alone (i.e., without anti-CD3 antibody) did not induce significant TNFα production from RA CD4+ T cells, and neutralizing anti-TNFα antibodies were not able to reverse IL-21–induced RANKL expression in CD4+ T cells (data not shown). Thus, it appears that IL-21 induces RANKL expression directly.
We also demonstrated that IL-21 was highly expressed in inflammatory arthritic synovial samples from both mice with CIA and patients with RA (Figures 1B and 6A). The distribution of IL-21 expression was similar to that of RANKL in both mouse and human samples, which supports the in vitro data showing that IL-21 induces RANKL expression in CD4+ T cells. Therefore, we speculate that IL-21–stimulated CD4+ T cells and FLS can have osteoclastogenic potential, because the results of our experiments proved that IL-21 induces RANKL expression in both CD4+ T cells and FLS. As expected, the osteoclastogenic potential of IL-21–stimulated CD4+ T cells and FLS was more potent than that of CD4+ T cells and FLS that were not stimulated with IL-21 (Figures 3A and 5A).
One intriguing point pertaining to our results is that IL-21 showed osteoclastogenic impact in both mice and humans in the absence of RANKL-providing cells such as CD4+ T cells and FLS. Although IL-21 alone did not promote osteoclastogenesis, it markedly potentiated it in the presence of RANKL and M-CSF, in both mouse and human samples, as determined by TRAP staining (Figures 3C and 5C) and by bone resorption assay (Figures 3D and 5D). Surprisingly, in the presence of M-CSF and low-dose RANKL stimulation (10 ng/ml in mice), treatment with IL-21 increased the number of differentiated osteoclasts as much as did M-CSF and high-dose RANKL (100 ng/ml in mice) (Figure 3C). These findings suggest that IL-21 has osteoclastogenic potential, although it is not as potent as that of RANKL.
The potential involvement of the IL-21/STAT-3 signaling pathway in osteoclastogenesis requires more exploration. A previous study demonstrated that the STAT-3 pathway might be implicated in RANKL-induced osteoclastogenesis by showing that the addition of two STAT-3 inhibitors completely abolished RANKL-induced osteoclastogenesis (33). Further research will be needed to elucidate the exact signaling pathway of IL-21/STAT-3 and the interrelationship with RANKL or M-CSF in osteoclastogenesis.
Bone destruction is of utmost importance in the clinical course of human RA, because it is closely associated with functional disability. Many drugs have been developed for the treatment of RA, most of which modulate immune reactions and reduce both inflammation and pain. However, despite continuous treatment with conventional disease-modifying antirheumatic drugs, disease progression does not stop (34, 35), and some patients still have to undergo joint replacement surgery because of progressive bone destruction (36). Since the introduction of biologic agents for treating patients with RA, especially the TNFα inhibitors, the therapeutic strategy has changed. This is because treatment with a combination of methotrexate and TNFα inhibitors such as infliximab, etanercept, and adalimumab has proved to be significantly superior to treatment with methotrexate alone for improving the signs and symptoms of disease but also by inhibiting radiographic progression of RA (37–39). Therefore, IL-21, which was shown to have osteoclastogenic activity in this study, could be a good therapeutic target for the treatment of patients with RA, especially in terms of preventing bone destruction.
In conclusion, we demonstrated that IL-21 promotes osteoclastogenesis in patients with RA as well as in mouse models of arthritis, in the absence of RANKL-providing cells as well as indirectly by inducing RANKL expression in CD4+ T cells and synoviocytes, thereby contributing to bone destruction in inflamed joints (Figure 6C). These findings suggest that a novel therapy targeting IL-21 might be effective for preventing bone destruction in patients with RA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. S.-H. Park 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 conception and design. Kwok, Cho, M.-K. Park, Oh, J.-S. Park, Her, Lee, Youn, Ju, K. S. Park, S.-I. Kim, H.-Y. Kim, S.-H. Park.
Acquisition of data. Kwok, Cho, M.-K. Park, Oh, J.-S. Park, Her, Lee, Youn, Ju, K. S. Park, S.-I. Kim, H.-Y. Kim, S.-H. Park.
Analysis and interpretation of data. Kwok, Cho, M.-K. Park, Oh, J.-S. Park, Her, Lee, Youn, Ju, K. S. Park, S.-I. Kim, H.-Y. Kim, S.-H. Park.
- 25IL-21 acts directly on B cells to regulate Bcl-6 expression and germinal center responses. J Exp Med 2007: 353–63., , , , , , et al.
- 33Interleukin-6 inhibits receptor activator of nuclear factor κB ligand-induced osteoclastogenesis by diverting cells into the macrophage lineage: key role of Serine727 phosphorylation of signal transducer and activator of transcription 3. Endocrinology 2008; 149: 3688–97., , , , , , et al.
- 39The PREMIER study: a multicenter, randomized, double-blind clinical trial of combination therapy with adalimumab plus methotrexate versus methotrexate alone or adalimumab alone in patients with early, aggressive rheumatoid arthritis who had not had previous methotrexate treatment. Arthritis Rheum 2006; 54: 26–37., , , , , , et al.