Interleukin-21 Receptor Deficiency Increases the Initial Toll-like Receptor 2 Response but Protects Against Joint Pathology by Reducing Th1 and Th17 Cells During Streptococcal Cell Wall Arthritis




The cytokine interleukin-21 (IL-21) can have both proinflammatory and immunosuppressive effects. The purpose of this study was to investigate the potential dual role of IL-21 in experimental arthritis in relation to Th17 cells.


Antigen-induced arthritis (AIA) and chronic streptococcal cell wall (SCW) arthritis were induced in IL-21 receptor–deficient (IL-21R−/−) and wild-type mice. Knee joints, synovial tissue, and serum were analyzed for arthritis pathology and inflammatory markers.


During AIA and chronic SCW arthritis, IL-21R deficiency protected against severe inflammation and joint destruction. This was accompanied by suppressed serum IgG1 levels and antigen-specific T cell responses. Levels of IL-17 were reduced during AIA, and synovial lymphocytes isolated during SCW arthritis for flow cytometry demonstrated that mainly IL-17+ interferon-γ (IFNγ)–positive T cells were reduced in IL-21R−/− mice. However, during the acute phases of SCW arthritis, significantly higher joint swelling scores were observed, consistent with enhanced tumor necrosis factor and IL-6 expression. Interestingly, IL-21R−/− mice were significantly less capable of up-regulating suppressor of cytokine signaling 1 (SOCS-1) and SOCS-3 messenger RNA. IL-21 stimulation also affected the Toll-like receptor 2 (TLR-2)/caspase recruitment domain 15 response to SCW fragments in vitro, indicating that impaired SOCS regulation in the absence of IL-21 signaling might contribute to the increased local activation during SCW arthritis.


In contrast to the proinflammatory role of IL-21 in adaptive immunity, which drives IL-17+IFN+ cells and joint pathology during chronic experimental arthritis, IL-21 also has an important immunosuppressive role, presumably by inhibiting TLR signaling via SOCS-1 and SOCS-3. If this dual role of IL-21 in various immune processes is present in human disease, it could make IL-21 a difficult therapeutic target in rheumatoid arthritis.

Rheumatoid arthritis (RA) is a systemic autoimmune disease of unknown etiology, characterized by chronic joint inflammation and destruction of articular cartilage and bone. Hyperplasia of the synovial lining and influx of proinflammatory cells leads to thickening of the synovium, and production of soluble mediators like cytokines, enzymes, and oxygen radicals by both resident and infiltrating cells contributes to the final destruction of the joint matrix.

Tumor necrosis factor α (TNFα) and interleukin-1 (IL-1) are the most extensively studied cytokines in the arthritis process, and biologic treatments neutralizing TNFα have shown good clinical effects in many RA patients. However, for RA patients who do not respond well to current available therapies, the search for new therapeutic strategies continues. Since nonresponders to anti-TNF treatment show an increased percentage of Th17 cells in the circulation ([1]), additional new therapies targeting the Th17 cell pathway might provide good candidates. Previous studies of the potential of the major Th17 cell cytokine IL-17 in various mouse models demonstrated arthritis suppression by anti–IL-17 treatment, especially when IL-17 blocking was combined with anti-TNF treatment in protecting against cartilage damage ([2]), or when human RA synovial tissue was confirmed to be rich in CD3+ T cells before treatment in the RA synovium SCID mouse model ([3]).

Another Th17 cell–derived candidate is IL-21, a pleiotropic cytokine that binds to the IL-21 receptor (IL-21R), consisting of the common cytokine receptor γ chain and the specific IL-21R subunit ([4]). IL-21 is not produced solely by Th17 cells, but is also produced by follicular helper T (Tfh) cells and natural killer (NK) T cells. The IL-21 receptor is expressed on a variety of hematopoietic immune cells, including T and B cells, dendritic cells (DCs), and NK cells, as well as on tissue cells including endothelial cells and fibroblasts. IL-21 acts on B cells and plays an important role in normal Ig production and B cell differentiation into plasma cells ([5-7]). In addition, IL-21 contributes to antibody production indirectly via its effect on Tfh cell generation and germinal center formation ([8]). More interestingly, IL-21 also acts as an autocrine cytokine of activated T cells, since it induces Th17 cell differentiation of both human and murine naive CD4+ T cells in the presence of transforming growth factor β ([9-11]). In vivo studies also demonstrated that Th17 cell differentiation and responses are disturbed in the absence of IL-21 or its receptor ([9, 10, 12]).

Increased levels of IL-21 are found in serum and synovial fluid of RA patients compared to osteoarthritis patients and healthy controls ([1, 13]), suggesting a role for IL-21 in the pathogenesis of RA. Enhanced IL-21R expression was found in RA synovial fibroblasts and macrophages, although regulation of IL-21R could not be induced by the most common proinflammatory cytokines, IL-1 and TNF ([14]). IL-21 on T cells induces the secretion of proinflammatory cytokines like TNFα and interferon-γ (IFNγ) and enhances the expression of RANKL ([13, 15]). Synovial fibroblasts show increased expression of various matrix metalloproteinases as well as RANKL upon IL-21 stimulation ([13]).

The involvement of IL-21 in autoimmune disease was previously shown in various disease models. IL-21 deficiency protected against experimental autoimmune encephalomyelitis ([9]), and blocking IL-21 in MRL/lpr mice reduced signs of experimental systemic lupus erythematosus ([12]). Neutralizing IL-21 in experimental arthritis already showed a role for IL-21 in rat adjuvant-induced arthritis and murine collagen-induced arthritis (CIA), with reduced clinical signs of disease and improved histologic scores ([16]). However, the link between IL-21 and Th17 cells was not made in these studies. The objective of this study was to investigate the effect of IL-21R deficiency on joint pathology in relation to Th17 cells during chronic experimental arthritis.



IL-21R–deficient mice were provided by Pfizer. C57BL/6J mice as wild-type (WT) controls were obtained from Elevage Janvier. All mice were housed in filter-top cages under specific pathogen–free conditions, and a standard diet and water were provided ad libitum. The mice were used between ages 10 and 12 weeks. All animal procedures were approved by the institutional ethics committee.


Methylated bovine serum albumin (mBSA) was purchased from Sigma-Aldrich. RPMI 1640 medium, TRIzol reagent, oligo(dT) primers, and Moloney murine leukemia virus (MMLV) reverse transcriptase (RT) were obtained from Life Technologies. Primers were purchased from Biolegio. SYBR Green Master Mix was purchased from Applied Biosystems. Cytokine kits for the Luminex multianalyte system (Milliplex) were obtained from Millipore.

Induction of antigen-induced arthritis (AIA).

Mice were immunized with 100 μg of mBSA emulsified in 100 μl Freund's complete adjuvant (CFA; Difco). Injections were divided over both flanks. Heat-killed Bordetella pertussis (RIVM) was administered intraperitoneally as an additional adjuvant. Two subcutaneous booster injections with a total of 50 μg mBSA in CFA were given in the neck region 1 week after the initial immunization. Three weeks after these injections, primary AIA was induced by injecting 60 μg of mBSA in 6 μl phosphate buffered saline (PBS) into the right knee joint, resulting in chronic arthritis.

Preparation of streptococcal cell wall (SCW) fragments and induction of SCW arthritis

Streptococcus pyogenes

T12 organisms were cultured overnight in Todd-Hewitt broth. Cell wall fragments were prepared as described previously ([17]). The supernatant resulting from centrifugation at 10,000g was used throughout the experiments. These preparations contained 11% muramic acid. Unilateral arthritis was induced by intraarticular injection of 25 μg of SCW fragments (rhamnose content) in 6 μl pyrogen-free saline into the right knee joints of naive mice.

Measurement of 99mTc uptake

Joint inflammation was measured by 99mTc-pertechnetate uptake in the knee joint. Mice were sedated with chloral hydrate and injected intraperitoneally with 20 μCi of 99mTc. Thirty minutes afterward, γ-particle radiation was assessed by use of a collimated sodium iodide scintillation crystal with the knee in a fixed position. Arthritis was scored as the ratio of the 99mTc uptake in the arthritic right knee joint to that in the control left knee joint. Right:left knee joint 99mTc uptake ratios >1.1 were taken to indicate inflammation of the right knee joint.


For standard histology of the knee joints, tissue was fixed for 4 days in 10% formalin, decalcified in 5% formic acid, and subsequently dehydrated and embedded in paraffin. Standard frontal sections of 7 μm were mounted on SuperFrost slides (Menzel-Gläser). Hematoxylin and eosin staining was performed to study joint inflammation. The severity of inflammation in the knee joints was scored as the amount of infiltrating cells on a scale of 0–3 (0 = no cells, 1 = mild cellularity, 2 = moderate cellularity, and 3 = maximal cellularity). To study proteoglycan (PG) depletion from the cartilage matrix, sections were stained with Safranin O followed by counterstaining with fast green. PG depletion was determined using an arbitrary scale of 0–3, ranging from normal, fully stained cartilage to destained cartilage completely depleted of PGs. The degree of chondrocyte death was scored on a scale of 0–3, ranging from no empty lacunae to complete loss of chondrocytes in the cartilage layer. Histopathologic changes in the knee joints were scored in the patella and femur/tibia regions on 5 semiserial sections of the joint spaced 70 μm apart. Scoring was performed in a blinded manner by 2 independent observers (RJM, MIK).

Measurement of anti-mBSA and anti-SCW antibodies

Levels of anti-mBSA or anti-SCW antibodies in sera of arthritic mice were analyzed by enzyme-linked immunosorbent assay. Briefly, 10 ng of mBSA or SCW fragments were coated onto 96-well plates overnight. Plates were washed, and nonspecific binding sites were blocked with 1% BSA in PBS–Tween (0.05%). Serial 2× dilutions of sera, starting with an initial dilution of 20×, were incubated in the plates for 1 hour. The plates were then washed, and isotype-specific horseradish peroxidase–labeled goat anti-mouse Ig (1:1,000 dilution) was added for 1 hour at room temperature; 5-aminosalicylic acid was used as substrate. Absorbance was measured at 450 nm.

Lymphocyte stimulation test

Spleens were isolated and disrupted, and erythrocytes were lysed in 0.16M NH4Cl (pH 7.2). The cell suspension was washed with saline and enriched for lymphocytes by allowing antigen-presenting cells to adhere to plastic culture flasks for 45 minutes. Cells (2 × 105/well) were cultured for 72 hours at 37°C in an atmosphere of 5% CO2, in RPMI 1640 (Gibco Invitrogen) supplemented with 5% fetal calf serum (FCS), 1 mM pyruvate, and 50 mg/liter of gentamicin in the presence of 2-fold serial dilutions of antigen (mBSA or SCW fragments). For the last 16 hours, the cells were labeled with 0.25 μCi 3H-thymidine. After harvesting, incorporation of 3H-thymidine was measured and the increase in counts per minute caused by stimulation with specific antigen was determined.

Synovial washouts

To determine levels of several cytokines and chemokines in patellae washouts, patellae with surrounding soft tissue consisting of the tendon and synovium were dissected in a standardized manner. Patellae were cultured in RPMI 1640 medium containing 0.1% BSA (200 μl/patella) for 1 hour at room temperature. Afterward, supernatant was harvested and centrifuged for 5 minutes at 1,000g. Cytokine levels were determined using the Luminex multianalyte technology.

Luminex analysis

To determine levels of the cytokines and chemokines in serum samples and synovial washouts, we used Luminex multianalyte technology in combination with multiplex cytokine kits (Milliplex; Millipore). Cytokines were measured in 25 μl of synovial washout or in serum diluted 1:3 in assay buffer. The sensitivity of the multiplex kit was <1 pg/ml.

Isolation of synovial cells

The protocol was kindly provided by Prof. Shimon Sakaguchi (Osaka University, Osaka, Japan) and adjusted. After mice were killed, the joint synovium was dissected. Briefly, synovial biopsy samples were incubated with enzymatic digestion buffers (Liberase; Roche) for 30 minutes at 37°C. Next, a 70-μm nylon cell strainer (BD Falcon) was used to process the digested tissue. The cell preparation was collected in RPMI with 10% FCS. To isolate the single mononuclear cells, Lympholyte-M (Cedarlane) was used according to the manufacturer's protocol. Subsequently, the isolated cells were prepared for intracellular flow cytometry.

Intracellular flow cytometry

Cells were stimulated for 4–6 hours with phorbol myristate acetate (50 ng/ml; Sigma-Aldrich), ionomycin (1 μg/ml; Sigma-Aldrich), and GolgiPlug (1 μl/ml; BD Biosciences). Next, for intracellular detection of IFNγ and IL-17, cells were washed, fixed, and permeabilized according to the manufacturer's protocol (eBioscience) and incubated with anti-IFNγ (BD Biosciences) and anti–IL-17 (BD Biosciences) antibodies.

RNA isolation and RT reaction

Mice were killed by cervical dislocation, immediately followed by dissection of the patella with adjacent synovium. Two biopsy samples with a diameter of 3 mm (1 from the lateral side and 1 from the medial side of the synovial tissue) were punched out using a biopsy punch (Stiefel). The synovium samples were immediately frozen in liquid nitrogen. Total RNA was extracted in 1 ml TRIzol reagent. Afterward, RNA was precipitated with isopropanol, washed with 70% ethanol, and redissolved in water. Isolated RNA was treated with DNase before being reverse-transcribed into complementary DNA using oligo(dT) primers and MMLV RT.

Quantitative polymerase chain reaction (PCR).

Quantitative real-time PCR was performed using the ABI/Prism 7000 Sequence Detection System for quantification with SYBR Green and melting curve analysis (Applied Biosystems). PCR conditions were as follows: 2 minutes at 50°C and 10 minutes at 95°C, followed by 40 cycles of 15 seconds at 95°C and 1 minute at 60°C, with data collection in the last 30 seconds. For all PCRs, SYBR Green Master Mix was used in the reaction. Primer concentrations were 300 nmoles/liter. All PCRs were performed in a total volume of 25 μl. Relative quantification of the PCR signals was performed by comparing the cycle threshold (Ct) value, in duplicate, of the gene of interest for each sample with the Ct values of the reference gene GAPDH.

Statistical analysis

Differences between experimental groups were tested using the Mann-Whitney U test or Kruskal-Wallis test, unless stated otherwise. P values less than 0.05 were considered significant. Results are expressed as the mean ± SEM unless stated otherwise.


IL-21R deficiency protects against severe inflammation and bone destruction during AIA

The role of IL-21 in experimental arthritis was first studied in the mouse model of chronic AIA using IL-21R–deficient mice and their WT controls. Measurements of 99mTc after intraarticular injection with the antigen mBSA showed similar joint swelling in both groups in the first 2 days of arthritis. However, on day 4, a more rapid decline was observed in IL-21R–deficient mice compared to WT controls (Figure 1A). More importantly, histologic analysis on day 7 of arthritis demonstrated significantly reduced joint inflammation, cartilage destruction, and bone erosion in the absence of the IL-21 receptor (Figure 2). This reduced joint pathology in IL-21R−/− mice was accompanied by suppressed adaptive immune responses; specific subsets of anti-mBSA antibodies and also antigen-specific lymphocyte proliferation were significantly decreased by IL-21R deficiency (Figures 1B and C).

Figure 1.

Interleukin-21 receptor (IL-21R) deficiency results in a decrease in arthritis with less destruction paired with reduced adaptive immune response. A, IL-21R deficiency during the induction of primary antigen-induced arthritis results in a more rapid decrease in joint swelling (ratio of 99mTc uptake in the arthritic right knee joint [R] to that in the control left knee joint [L]) (n = 6–8 mice per group). B, Serum levels of methylated bovine serum albumin (mBSA)–specific IgG1 antibodies are decreased on day 7 in IL-21R−/− mice (n = 8 mice per group). C, Proliferation of mBSA-specific splenic lymphocytes is reduced in IL-21R−/− mice. Nonadherent spleen cells were stimulated with various concentrations of mBSA or 1 μg/ml concanavalin A (Con A) for 72 hours, followed by 18 hours of incubation with 0.25 μCi 3H-thymidine (n = 7 mice per group). Results are shown as netto counts (cpm stimulation – cpm background). D, IL-21R deficiency reduces levels of IL-6, CXCL1 (keratinocyte chemoattractant), and IL-17 in synovial washouts from arthritic mice (n = 6 mice per group). Values are the mean ± SEM. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001 by Student's t-test. WT = wild-type; OD/50 = optical density at 50% transmittance; TNFα = tumor necrosis factor α; IFNγ = interferon-γ.

Figure 2.

Histopathologic effects of IL-21R deficiency during antigen-induced arthritis (AIA). Total knee joints were isolated for histopathologic analysis (n = 8 mice per group). A, IL-21R deficiency during AIA results in suppressed joint inflammation, cartilage proteoglycan (PG) depletion, chondrocyte death, and bone erosion on day 28. Values are the mean ± SEM. ∗ = P < 0.05; ∗∗ = P < 0.01 by Student's t-test. B and C, Histologic sections show infiltrating cells (hematoxylin and eosin staining) (B) and cartilage PG depletion (Safranin O staining) (C) in frontal knee sections. Images are representative of 8 mice per group. Original magnification × 50. See Figure 1 for other definitions.

Interestingly, basal serum IgG1 and IgG2b levels were already reduced by IL-21R deficiency (IgG1 was reduced from a mean ± SEM of 0.41 ± 0.06 mg/ml to 0.29 ± 0.07 mg/ml; IgG2b was reduced from 1.05 ± 0.21 mg/ml to 0.38 ± 0.05 mg/ml), as has also been demonstrated by others ([5]). However, during experimental arthritis, only the antigen-specific IgG1 anti-mBSA antibodies were reduced in the absence of IL-21 signaling, while total IgG and IgG2b were not affected (Figure 1B). In addition, the cytokines IL-6 and IL-17 and the chemokine CXCL1 were significantly reduced in synovial washouts from IL-21R−/− mice on day 7 of AIA (Figure 1D). These data indicate an important proinflammatory role for IL-21 in driving adaptive immunity and chronic destructive arthritis during murine AIA.

Enhanced local activation in IL-21R–deficient mice during SCW arthritis

In addition to the AIA model, the role of IL-21/IL-21R signaling was investigated in SCW arthritis. In this model, a single injection of SCW fragments into the knee joint results in local acute inflammation, and subsequent repeated weekly injections turn this acute inflammation into a chronic destructive arthritis. In contrast to our findings in the AIA model, IL-21R–deficient mice showed enhanced joint swelling at various time points during SCW arthritis, most significantly as soon as day 1 after arthritis induction (Figure 3A). This enhanced swelling was accompanied by significantly enhanced synovial TNF and IL-6 messenger RNA (mRNA) expression on day 1 and a trend toward higher IL-6 and CXCL1 levels in synovial washouts on day 22 (Figure 3B), indicating enhanced local activation in the absence of IL-21R. However, systemic cytokine levels at this time point did not show any difference between IL-21R−/− and WT mice. More remarkably, at the time that mice were killed (day 28), cytokine levels in IL-21R−/− mice were significantly reduced compared to those in WT controls (Figure 3C), in contrast to the previous enhanced local activation.

Figure 3.

IL-21R deficiency in a streptococcal cell wall (SCW) arthritis model results in enhanced local activation. A, IL-21R deficiency during chronic SCW arthritis resulted in increased joint swelling. Joint swelling was measured by 99mTc uptake (n = 8–9 mice per group). B, IL-21R deficiency during chronic SCW arthritis enhanced synovial IL-6 and TNF mRNA expression (n = 4 mice per group) and did not significantly affect proinflammatory cytokine levels in synovial washouts from arthritic mice on day 22 (n = 6–8 mice per group). C, In contrast to findings on day 22, serum cytokine levels of IL-1β, IL-6, and IL-10 were reduced by day 28 (n = 6–8 mice per group). Values are the mean ± SEM. ∗ = P < 0.05; ∗∗∗ = P < 0.001 by Student's t-test. NS = not significant (see Figure 1 for other definitions).

Reduced joint pathology and IL-17+IFNγ+ Th cells in IL-21R−/− mice during SCW arthritis

To study the absolute effect of IL-21R deficiency on joint pathology during this SCW arthritis model, knee joints were isolated on day 28 for histologic analysis. As shown in Figures 4A and B, the arthritic joints of IL-21R−/− mice showed a significant reduction in joint inflammation, in contrast to the enhanced joint swelling during the course of the arthritis. In addition, the protective effect of IL-21R deficiency against cartilage damage was also visible, as both PG depletion and chondrocyte death were clearly suppressed compared to that in WT controls (Figures 4A and C). Consistent with the results during AIA, antigen-specific B and T cell responses were reduced in IL-21R−/− mice (Figures 5A and B).

Figure 4.

Histopathologic effects of IL-21R deficiency during the chronic phase of streptococcal cell wall (SCW) arthritis. Total knee joints were isolated for histopathologic analysis (n = 8 mice per group). A, IL-21R deficiency during chronic SCW arthritis suppresses the influx of infiltrating cells, cartilage proteoglycan (PG) depletion, and chondrocyte death on day 28. Values are the mean ± SEM. ∗ = P < 0.05; ∗∗ = P < 0.01 by Student's t-test. B and C, Histologic sections stained with hematoxylin and eosin show infiltrating cells (B) and chondrocyte death (C) in frontal knee sections. Images are representative of 8 mice per group. Original magnification × 50 in B; × 200 in C. See Figure 1 for other definitions. Color figure can be viewed in the online issue, which is available at

Figure 5.

Decreased adaptive immune responses in IL-21R–deficient mice during chronic streptococcal cell wall (SCW) arthritis. A, Serum levels of SCW-specific IgG1 and IgG3 antibodies were decreased on day 28 in IL-21R−/− mice (n = 6–8 mice per group). B, Proliferation of SCW-specific splenic lymphocytes was reduced in IL-21R−/− mice. Nonadherent spleen cells were stimulated with various concentrations of SCW fragments or 1 μg/ml Con A for 72 hours, followed by 18 hours of incubation with 0.25 μCi 3H-thymidine (n = 7–8 mice per group). Results are shown as netto counts (cpm stimulation – cpm background). C and D, Representative dot plots of isolated cells gated on CD4 (C) and bar graphs depicting counts from 6 mice per group (D) show that the percentage of IL-17+/IFNγ+ cells was decreased in IL-21R−/− mice, whereas the percentages of IL-17+/IFNγ− and IL-17−IFNγ+ cells were not (n = 6 mice per group). Values are the mean ± SEM. ∗ = P < 0.05 by Student's t-test. See Figure 1 for other definitions.

The reduced IL-17 levels during AIA in IL-21R−/− mice suggested that T cells and, in particular, Th17 cells might be affected by the absence of IL-21R signaling. Therefore, during the chronic SCW arthritis model we dissected synovial tissue of the arthritic knee joints and isolated the infiltrating cells for flow cytometric analysis. As demonstrated in Figure 5C, the percentages of CD4+ Th17 cells (IL-17+IFNγ–) and Th1 cells (IL-17–IFNγ+) were not changed in IL-21R−/− mice. However, double-positive IL-17+IFNγ+ (Th17/Th1) T cells were greatly reduced by IL-21R deficiency (Figures 5C and D), suggesting a potential role for IL-21 in driving these double-positive Th17/Th1 cells and mediating joint pathology.

IL-21 induces suppressor of cytokine signaling (SOCS) expression and reduces the Toll-like receptor (TLR)/caspase recruitment domain (CARD) response to SCW fragments

To further investigate enhanced local activation during SCW arthritis in IL-21R−/− mice, we studied the expression of TLR-2 and CARD-15, which are crucial for the response to SCW fragments. Basal synovial levels of TLR-2 and CARD-15 mRNA expression were equal in IL-21R−/− and WT mice, and the up-regulation of these receptors 4 days after the injection of SCW fragments was also comparable (data not shown). Since triggering of TLR-2/CARD-15 not only results in induction of proinflammatory cytokines but also in negative feedback via SOCS, we also studied the regulation of SOCS expression. Quantitative PCR analysis of synovial tissue demonstrated that besides a nonsignificant reduction in the basal level of SOCS-1, IL-21R−/− mice had clearly impaired up-regulation of SOCS-1 and SOCS-3 during arthritis (Figure 6A). The decreased expression of SOCS-3 protein on day 4 was confirmed by immunohistochemical staining of sections of knee joints from IL-21R−/− mice (Figure 6B).

Figure 6.

IL-21R–deficient mice show impaired up-regulation of suppressor of cytokine signaling 1 (SOCS-1) and SOCS-3 during arthritis. A, IL-21R deficiency results in suppressed up-regulation of SOCS-1 and SOCS-3 during streptococcal cell wall (SCW) arthritis on day 4 (n = 6 mice per group). B, Shown are representative histologic sections of immunohistochemical staining of SOCS-3 on day 4. For each mouse strain, original magnification × 50 in left panels; × 200 in right panels. C, Stimulation of spleen cells from WT mice with IL-21 up-regulates SOCS-1 and SOCS-3 (n = 6 mice per group). D, In spleen cells from WT mice, increases in IL-6 and CXCL1 (keratinocyte chemoattractant [KC]) levels are suppressed after 2 hours of exposure to IL-21 followed by 24 hours of stimulation with SCW (n = 6 mice per group), while spleen cells from IL-21R–deficient mice fail to show this response. Values are the mean ± SEM. ∗ = P < 0.05; ∗∗∗ = P < 0.001 by Student's t-test. NS = not significant (see Figure 1 for other definitions). Color figure can be viewed in the online issue, whic is available at

This IL-21–dependent up-regulation of SOCS-1 and SOCS-3 was confirmed in subsequent in vitro experiments in which we cultured splenocytes from IL-21R−/− and WT mice with or without recombinant IL-21 for 24 hours. As demonstrated in Figure 6C, cells from IL-21R−/− mice again showed a nonsignificant reduction in basal levels of SOCS-1 and SOCS-3. Furthermore, cells from WT mice had enhanced SOCS-1 and SOCS-3 mRNA levels upon IL-21 stimulation, whereas cells from IL-21R−/− mice did not respond. Interestingly, cells stimulated with SCW fragments after pre-exposure to IL-21 produced less IL-6 and CXCL1 than did cells not exposed to IL-21 (Figure 6D), suggesting that IL-21–mediated SOCS up-regulation can functionally inhibit responses to TLR-2/CARD-15 triggering. This suggests that lack of SOCS up-regulation in the absence of IL-21 signaling may contribute to enhanced local activation in IL-21R−/− mice during SCW arthritis.


In the present study, we demonstrated that IL-21R deficiency during AIA causes clearly reduced inflammation and PG depletion as well as protection against severe destruction of cartilage and bone. However, IL-21R−/− mice show enhanced local activation early after the induction of SCW arthritis, accompanied by impaired up-regulation of SOCS-1 and SOCS-3. Therefore, this is the first study to describe a dual role of IL-21 in experimental arthritis, on the one hand driving adaptive immunity via proinflammatory IFNγ+IL-17+ Th cells, while on the other hand suppressing TLR responses with a potential role for inhibitory SOCS signaling.

CD4+ T cells play an important role in the arthritis process. Where genetic studies revealed an important contribution of genes related to T cell activation, such as the HLA–DR4 allele, PTPN22, and CTLA-4 ([18]), clinical evidence of the T cell contribution was obtained by the successful treatment of RA patients with CTLA-4Ig (abatacept) ([19]). A new subset of T cells, the Th17 cell, has recently emerged as a major pathogenic T cell involved in RA pathogenesis. While the role of IL-17, the main cytokine produced by Th17 cells, has already been studied extensively in experimental arthritis (for review, see ref.[20]) and other autoimmune and inflammation models (for review, see ref.[21]), the role of the other Th17 cell cytokines IL-21 and IL-22 remains to be further elucidated.

To investigate the role of IL-21 in experimental arthritis, we used IL-21R−/− mice during AIA. In this chronic destructive arthritis model, IL-17 was previously demonstrated to be expressed in the synovial tissue ([22]), and recombination-activating gene 1–deficient and CD4-deficient mice showed clear protection from arthritis pathology ([23]). In our current study, joint swelling and histologic inflammation during AIA were significantly reduced in IL-21R−/− mice. In addition, lack of IL-21R expression protected against PG depletion and chondrocyte death as parameters of cartilage destruction, and bone erosions were also less prominent in IL-21R−/− mice, indicating a clear proinflammatory role for IL-21 in arthritis. This was consistent with previous studies of IL-21 blocking in CIA ([16]) and with the observation that IL-21R deficiency completely blocks arthritis development in K/BxN mice ([24]).

IL-21 has a broad spectrum of responsive cell types, including cells from the innate as well as the adaptive immune system ([4]). The role of IL-21 in B cell responses has been intensively investigated ([5-7]), and in our study, we confirmed that IL-21 is an important player in antigen-specific antibody responses. IL-21R deficiency resulted in significant inhibition of Th2 cell–related IgG1 and Th1 cell–related IgG3 antibody levels in the serum of arthritic mice. Since the SCW-specific IgG2b subclass was not affected by IL-21R deficiency, and our study and those of others showed a general suppression of IgG1 and IgG2b basal levels in IL-21R−/− mice ([5]), we conclude that the suppressed IgG1 levels are not merely due to a generic difference between WT and knockout mice, but that this reduced antibody response is a specific IL-21–dependent effect. While antigen-specific T cell responses were highly suppressed in the absence of IL-21 signaling, an inhibitory effect on T cell cytokines in the arthritic joint could only be found for IL-17 levels in synovial washouts, while no differences were observed for the Th1 cell cytokine IFNγ.

We attempted to confirm our findings of IL-21 as a proinflammatory cytokine in experimental arthritis by inducing chronic SCW arthritis in IL-21R−/− mice. By local intraarticular injection of TLR-2/CARD-15 ligands ([25]), this acute macrophage-driven arthritis becomes T cell dependent after repeated weekly exposure to the antigen ([26]). Consistent with our findings in the AIA model, IL-21R deficiency during SCW arthritis resulted in reduced antigen-specific T and B cell responses. To study the effect of IL-21 on Th cells in closer detail, flow cytometry was performed on lymphocytes isolated from synovial tissue. This analysis showed comparable percentages of CD4+IL-17+ Th17 cells in IL-21R−/− mice and WT controls, but it also showed a remarkable effect of IL-21R deficiency on CD4+ T cells that produce both IFNγ and IL-17.

From this additional arthritis model we might conclude that IL-21R deficiency indeed suppresses adaptive immunity and, in particular, double-positive Th1/Th17 cells, although the protective effects on histologic scores of inflammation and destruction were less pronounced than in the chronic AIA model. However, it is of great interest that throughout the SCW arthritis model, continuously enhanced joint swelling was observed in the IL-21R−/− mice as compared to the WT controls. Also, in the synovial washouts from the arthritic joint, no suppression of cytokines and chemokines could be found, as was previously reported for the AIA model. On the contrary, enhanced local activation was observed with significantly elevated IL-6 and TNF mRNA on day 1 and a trend toward increased levels of IL-6 and CXCL1 on day 22. The aggravated response to SCW fragments in IL-21R−/− mice could not be explained by a difference in expression levels of TLR-2 and CARD-15, the receptors responsible for a reaction to the SCW fragments ([25]), since equal mRNA levels were found in synovial biopsy samples of both mouse strains. Since IL-21R deficiency does not interfere with K/BxN serum–induced arthritis ([24]), acting through immune complexes, Fcγ receptors, and complement activation, the aggravated response of IL-21R−/− mice to SCW fragments cannot be generalized to all innate immune responses.

In vitro stimulation assays showed that IL-21 can induce SOCS expression. These suppressor of cytokine signaling proteins act as negative regulators and inhibit cytokine pathways as well as TLR signaling ([27]). In our IL-21R−/− mice, the basal expression level of SOCS-1 was already slightly reduced compared to those in WT controls; remarkably, however, IL-21/IL-21R signaling seems to be essential for SOCS up-regulation during experimental arthritis. While control mice demonstrated a massive increase in SOCS expression, the up-regulation of SOCS-1 and SOCS-3 was clearly impaired in IL-21R−/− mice. We assume that in vivo, soon after the injection of SCW fragments, infiltrating lymphocytes (e.g., γ/δ T cells and NK T cells) contribute to increased IL-21 levels, as ∼40 pg/ml of IL-21 could be detected in synovial washouts on day 1 of SCW arthritis (data not shown). In addition to the in vivo SOCS differences, in vitro responses to SCW fragments were significantly suppressed in the presence of IL-21, suggesting that IL-21–induced SOCS expression affects this innate TLR response. IL-21–driven SOCS-dependent inhibition of lipopolysaccharide signaling was previously described for monocyte-derived DCs in vitro ([28]), and our findings suggest that this mechanism might well account for the enhanced joint swelling and proinflammatory cytokine expression in IL-21R−/− mice during SCW arthritis.

In summary, studying the role of IL-21 in arthritis, we have seen 2 independent mechanisms that determine the final outcome of joint inflammation and destruction: 1) IL-21 as a proinflammatory cytokine driving adaptive immunity, with a specific effect on double-positive IFNγ+IL-17+ Th cells, and 2) IL-21 as a negative regulator of TLR responses, presumably via induction of SOCS-1 and SOCS-3 expression as demonstrated in TLR-2–dependent SCW arthritis. The net effects of inhibiting IL-21 will depend on the environment. IL-21 will act in a more proinflammatory capacity in T cell–driven arthritis like our AIA model, in which the memory response is independent of TLR pathways (Abdollahi-Roodsaz, et al: unpublished observations), whereas in inflammation that is still partly driven through the innate system with TLR involvement, such as during SCW arthritis, IL-21 can also demonstrate its antiinflammatory capacity.

IL-21 seems to have a dual role in immune responses during arthritis, turning off an initial innate response mediated through TLRs and up-regulated in adaptive immune responses to continue to drive a chronic inflammatory response. If this dual role of IL-21 in various immune processes is present in human disease, it could make IL-21 a difficult therapeutic target in 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. Koenders 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. Marijnissen, Nickerson-Nutter, van de Loo, Boots, van den Berg, Koenders.

Acquisition of data. Marijnissen, Roeleveld, van de Loo, van Spriel, Koenders.

Analysis and interpretation of data. Marijnissen, Roeleveld, Young, Abdollahi-Roodsaz, de Aquino, van Spriel, Boots, Koenders.


Pfizer Inc. facilitated the study by grant support but had no role in data collection, interpretation of the results, or reviewing and approving the manuscript prior to submission.


Author Boots was an employee of NV Organon at the time of the study.


We thank the personnel of the animal facility for taking good care of our mice, and we thank the research technicians Birgitte Walgreen, Monique Helsen, and Liduine van den Bersselaar for their excellent technical assistance throughout the experimental arthritis studies.