Interleukin-6 modulates production of T lymphocytederived cytokines in antigen-induced arthritis and drives inflammation-induced osteoclastogenesis.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 54, No. 1, January 2006, pp 158–168 DOI 10.1002/art.21537 © 2006, American College of Rheumatology Interleukin-6 Modulates Production of T Lymphocyte–Derived Cytokines in Antigen-Induced Arthritis and Drives Inflammation-Induced Osteoclastogenesis Peter K. K. Wong,1 Julian M. W. Quinn,2 Natalie A. Sims,3 Annemarie van Nieuwenhuijze,1 Ian K. Campbell,1 and Ian P. Wicks1 Objective. To determine the cellular mediators of antigen-induced arthritis (AIA) and the relative contribution of members of the interleukin-6 (IL-6) family and tumor necrosis factor (TNF) in AIA. Methods. AIA was induced in mice deficient in T and B lymphocytes, IL-6 (IL-6ⴚ/ⴚ), TNF (TNFⴚ/ⴚ), IL-11 receptor, and oncostatin M receptor, by immunization with methylated bovine serum albumin (mBSA) followed 7 days later by intraarticular injection of mBSA. Arthritis severity was assessed histologically, and T lymphocyte responses were assessed in vitro. Anti-TNF neutralizing antibody was administered to wild-type mice during AIA. Bone marrow osteoclasts were generated in vitro via culture with RANKL and macrophage colony-stimulating factor. Results. AIA was dependent on CD4ⴙ T lymphocytes, but not CD8ⴙ T lymphocytes or B cells. IL-6ⴚ/ⴚ mice had reduced AIA severity and fewer osteoclasts at sites of bone erosion. This protective effect was not seen with a deficiency of other IL-6 family members and was similar to that in TNFⴚ/ⴚ mice or wild-type mice receiving TNF blockade treatment. IL-6ⴚ/ⴚ CD4ⴙ T lymphocytes from draining lymph nodes had reduced antigen-induced proliferation and produced less IL-17 and less RANKL, relative to osteoprotegerin, than cells from wild-type mice. Bone marrow from IL-6ⴚ/ⴚ mice generated fewer osteoclasts in vitro than bone marrow from either wild-type or TNFⴚ/ⴚ mice. Conclusion. AIA is driven by CD4ⴙ T lymphocytes. IL-6 is an important mediator of bone destruction in AIA because it regulates T lymphocyte production of key osteoclastogenic cytokines and inflammationinduced bone marrow osteoclast differentiation. These findings have implications for reducing bone and joint damage in rheumatoid arthritis. Rheumatoid arthritis (RA) is an autoimmune inflammatory disease characterized by sustained local overproduction of cytokines such as tumor necrosis factor (TNF) (1) and interleukin-1 (IL-1) (2). While targeted blockade of these cytokines has been a major therapeutic advance, many patients fail to respond to either TNF (3) or IL-1 blockade treatment (4). It is possible that other inflammatory mediators, such as IL-6 (5), may also be important in driving joint inflammation. Results from early clinical trials of targeted IL-6 blockade treatment in RA appear promising (6,7). RA patients have elevated levels of IL-6 in synovial fluid and serum compared with patients with osteoarthritis or healthy controls (8,9). Production of IL-6 by RA synoviocytes is up-regulated by TNF and IL-1 (10). Serum levels of IL-6 and IL-6 receptor (IL-6R) correlate with radiographic joint destruction in RA (8). Matrix metalloproteinases (MMPs) are important mediators of joint destruction in RA (11). IL-6 has been found to enhance IL-1–induced production of proMMP-1 and proMMP-3 and production of tissue inhibitor of metalloproteinases by RA synovial fibroblasts (12). Bone loss is a prominent feature of RA and may Supported by the Reid Charitable Trusts, the Arthritis Foundation of Australia, and the National Health and Medical Research Council of Australia. 1 Peter K. K. Wong, MBBS, FRACP, PhD, Annemarie van Nieuwenhuijze, MSc, Ian K. Campbell, PhD, Ian P. Wicks, MBBS, FRACP, PhD: The Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia; 2Julian M. W. Quinn, DPhil: St. Vincent’s Institute of Medical Research, Fitzroy, Victoria, Australia; 3Natalie A. Sims, PhD: St. Vincent’s Hospital, University of Melbourne, Fitzroy, Victoria, Australia. Address correspondence and reprint requests to Ian P. Wicks, MBBS, FRACP, PhD, Reid Rheumatology Laboratory, Division of Autoimmunity and Transplantation, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3050, Australia. E-mail: email@example.com. Submitted for publication April 7, 2005; accepted in revised form September 29, 2005. 158 ROLE OF IL-6 IN AIA occur focally as periarticular erosions or more generally throughout the skeleton, resulting in osteoporosis. Bone loss is due to enhanced osteoclast activity modulated by RANKL. RANKL is expressed by synovial fibroblasts (13), osteoblasts (14), and T lymphocytes (15). The action of RANKL is inhibited by the decoy receptor osteoprotegerin (OPG), which is produced by RANKLexpressing cells (16). Osteoclasts are found at sites of bony erosion in RA (17). Proinflammatory cytokines, such as TNF (18), IL-1 (19), IL-6/soluble IL-6R (20), and IL-17 (21), act on stromal/osteoblastic cells to up-regulate RANKL expression and promote osteoclastogenesis. T lymphocyte production of RANKL (22) and IL-17 (23) is therefore likely to be involved in the pathogenesis of bone destruction in RA. Antigen-induced arthritis (AIA) is a widely used murine model of inflammatory arthritis that is induced by intradermal (ID) immunization with methylated bovine serum albumin (mBSA) in the presence of Freund’s complete adjuvant (CFA), followed by intraarticular (IA) challenge with mBSA 7 or more days later. AIA has many histologic features of RA, such as a leukocyte infiltration, synovitis, and cartilage and bone destruction. AIA is dependent on T lymphocyte activity (24,25), specifically CD4⫹ T lymphocytes (26–28). The presence of anti-mBSA antibodies in murine AIA has been reported, but the functional contribution to the disease was not defined (29). Similarly, the role of CD8⫹ T lymphocytes and B cells in AIA has not been directly addressed. Mice deficient in IL-6 were protected against severe joint inflammation and destruction in AIA, even though TNF and IL-1 were produced in the synovium (29,30). Cells from draining lymph nodes (LNs) of arthritic IL-6⫺/⫺ mice mounted a weaker antigenspecific proliferative response, but whether this defect was due to defective CD4⫹ T lymphocytes or to antigenpresenting cells (APCs) was not examined (29,30). This is important because APCs in IL-6⫺/⫺ mice have been reported to be less efficient stimulators of T lymphocyte activation (31,32). In another study of AIA in IL-6⫺/⫺ mice, transient joint inflammation developed, but failed to progress to long-term infiltration (33), suggesting that IL-6 was important in converting acute inflammation to a chronic phase. Expression of IL-6 was found in the bone marrow before and during the course of AIA, and AIA could be adoptively transferred into wild-type mice, but not into IL-6⫺/⫺ mice, by bone marrow cells from arthritic wild-type mice (34). We began this study by undertaking a detailed characterization of the cellular mediators of AIA using a 159 variety of mice deficient in T and B lymphocytes, because the contributions of these cell types have not been fully defined. We confirmed that AIA is dependent on CD4⫹ T lymphocytes, but found that it does not require CD8⫹ T lymphocytes or B cells. We next examined the role of IL-6 family cytokines in AIA, including a comparison of the relative importance of IL-6 and TNF in this disease model. Finally, we examined the effect of IL-6 on T lymphocyte production of pro-osteoclastogenic cytokines and inflammationinduced osteoclastogenesis during AIA. We found that IL-6 may be at least as important as TNF in AIA, and exerts major effects on T lymphocyte production of IL-17 and RANKL/OPG. Finally, IL-6 was implicated in regulating bone marrow osteoclastogenesis during peripheral joint inflammation in AIA. MATERIALS AND METHODS Mice. IL-6⫺/⫺ mice (35) were obtained from W. Alexander (Walter and Eliza Hall Institute of Medical Research [WEHI]). Mice deficient in TNF (36) were obtained from J. Sedgwick (Centenary Institute, Sydney, New South Wales, Australia). Recombination-activating gene 1 (RAG-1)– deficient mice (37) were provided by L. Corcoran (WEHI). GK mice, which are deficient in CD4⫹ T lymphocytes, were provided by Y. Zhan (WEHI) (38). Mice deficient in ␤2microglubulin (␤2m) (39) were provided by L. Harrison (WEHI). B lymphocyte–deficient MT/MT mice (40) were provided by D. Tarlinton (WEHI). IL-11 receptor–deficient (IL-11R1⫺/⫺) mice (41) and mice deficient in both IL-6 and IL-11R (IL-6⫺/⫺/IL-11R1⫺/⫺) were obtained from L. Robb (WEHI). Mice deficient in the oncostatin M receptor (OSMR) (42) were provided by A. Miyajima (Tokyo, Japan). All gene-knockout mice were either derived from C57BL/6 (B6) embryonic stem cells or were backcrossed onto B6 mice for ⱖ10 generations. All mice were ⱖ8 weeks of age at the time of experimentation. B6 mice, obtained from WEHI Animal Services (Kew, Victoria, Australia), were used as wild-type controls in all experiments. All animal procedures were approved by the institutional ethics committee. Antigen-induced arthritis. Methylated BSA (2 mg/ml; Sigma, St. Louis, MO) was emulsified in an equal volume of CFA containing 1 mg/ml heat-killed Mycobacterium tuberculosis (strain H37Ra; Difco, Detroit, MI). On day 0, mice were primed by ID injection of 100 l mBSA/CFA emulsion into the base of the tail. On day 7, mice were anesthetized and injected IA with 10 l of a 20-mg/ml solution of mBSA (in sterile saline) into the knee joints. Control joints received the same volume of vehicle (normal saline). Mice were killed 7 days after the IA injections (day 14), and the rear limbs were removed and processed, as previously described (43), for histologic assessment of hematoxylin and eosin (H&E)– stained frontal knee joint sections. Sections were assessed by 2 independent observers with no knowledge of the experimental groups (PKKW, IKC) and graded in severity from 0 (normal) to 5 (severe) for the following components: joint exudate, 160 synovitis, pannus, cartilage destruction, and bone destruction. The overall mean total histologic severity score (of a maximum possible score of 25) of an experimental group was calculated by averaging the sum of the 5 histologic features, with each individual feature scored at 4 section depths per joint. Osteoclast numbers were measured using standard histomorphometric procedures (44) on H&E-stained frontal knee joint sections. Measurements were carried out from the chondro– osseous junctions of both the proximal and distal articular cartilage with the lateral and medial periosteal surfaces. Each measurement followed the periosteal surface for a maximum of 300 m. A minimum of 3 sites was measured per animal. TNF neutralization in AIA. Wild-type arthritic mice were administered an anti-TNF neutralizing antibody (XT22; 0.5 mg/day) or an isotype control antibody on days 6, 9, and 11 of AIA. The dose of XT22 was previously shown to be effective in collagen-induced arthritis (45). Mice were killed on day 14, and joints were processed and scored as described above. Draining LN T lymphocyte proliferation assay. Popliteal and inguinal LN cells (4 ⫻ 105 cells/well) from arthritic mice were cultured for 72 hours with mBSA (0–50 g/ml) or concanavalin A (Con A; 2 g/ml), as previously described (43). Incorporation of tritiated thymidine was determined on a microplate scintillation counter (Canberra Packard, Melbourne, Victoria, Australia) as a measure of T lymphocyte proliferation. Antigen-specific proliferation of CD4ⴙ T lymphocytes. CD4⫹ T lymphocytes, isolated from draining LNs by staining with fluorescein isothiocyanate (FITC)–conjugated CD4 and anti-FITC microbeads (Miltenyi Biotec, Auburn, CA), were positively selected using an AutoMACS according to the instructions of the manufacturer (Miltenyi Biotec) and resuspended in 100 l T lymphocyte medium (RPMI 1640 supplemented with 10% fetal calf serum [FCS] and 50 M 2-mercaptoethanol with 1 mM sodium pyruvate) at 1 ⫻ 105 cells/100 l in 96-well plates. To determine purity, cells were analyzed by flow cytometry on a FACScan using CellQuest software (Becton Dickinson, San Jose, CA). Wild-type B6 splenocytes were irradiated (2,000 rads) after overnight incubation at 37°C with 1 mg/ml mBSA in RPMI 1640 supplemented with 10% FCS in the presence of 5% CO2. Splenocytes were then resuspended in 100 l T lymphocyte medium to give a final ratio of 1–10 CD4⫹ T lymphocytes per APC. CD4⫹ T lymphocytes and APCs were cocultured for 72 hours at 37°C in the presence of 5% CO2. Tritiated thymidine incorporation was used to measure cellular proliferation. Antigen presentation by dendritic cells (DCs). Sevenweek-old naive wild-type and IL-6⫺/⫺ mice were injected intravenously with whole ovalbumin (OVA; 3 mg per mouse) (Sigma) or phosphate buffered saline vehicle. Splenic DCs were isolated 16 hours later (purity 50–70% CD11c⫹) and DC phenotypic markers were assessed by flow cytometry, as previously described (45,46). Irradiated (2,000 rads) DCs were cultured (0–4 ⫻ 104 DCs/well) with OVA-specific OVAtransgenic I (OT-I) CD8⫹ T and OT-II CD4⫹ cells (1 ⫻ 104 cells/well). OT-I and OT-II T lymphocyte proliferation was measured after 3–4 days by tritiated thymidine incorporation for 18 hours. Cytokine enzyme-linked immunosorbent assays (ELISAs). IL-17 and interferon-␥ (IFN␥) ELISAs were performed using the appropriate matched antibodies (BD Phar- WONG ET AL Mingen, San Jose, CA). The optical density at 450 nm was read using a Multiskan Ascent photometer (Thermo Electron, Beverly, MA). Semiquantitative polymerase chain reaction (PCR). Total RNA was extracted from draining LN T lymphocytes cultured with mBSA using RNAzol B reagent (Tel-Test, Friendswood, TX). Complementary DNA was generated with Moloney murine leukemia virus reverse transcriptase, using 0.5 g total RNA and oligo primers diluted to a final volume of 20 l. To determine the expression of RANKL and OPG relative to ␤-actin, semiquantitative PCR was performed using a LightCycler FastStart DNA Master SYBR Green I kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s instructions, and a LightCycler (Roche Diagnostics), with primers specific for RANKL (forward 5⬘-GGTCGGGCAATTCTGAATT-3⬘, reverse 5⬘-GGGAATTACAAAGTGCACCAG-3⬘); OPG (forward 5⬘-ACCTCACCACAGAGCAGCTT-3⬘, reverse 5⬘-GTGCAGGAACCTCATGGTCT-3⬘); and ␤-actin (forward 5⬘-ATGGATGACGATATCGCTG-3⬘, reverse 5⬘-ATGAGGTAGTCTGTCAGGT-3⬘). Cycling conditions (40 cycles) for RANKL relative to ␤-actin and for OPG relative to ␤-actin were 95°C for an initial 12 minutes, then 95°C for 5 seconds, 55°C for 15 seconds, and 72°C for 60 seconds (RANKL) or 45 seconds (OPG). Osteoclast cultures. The osteoclastogenic potential of bone marrow and bone marrow macrophage precursors from wild-type, TNF⫺/⫺, and IL-6⫺/⫺ mice with and without AIA was determined as previously described (47). Briefly, bone marrow preparations flushed from femora and tibiae of mice were plated at a density of 105 cells/well and cultured in the presence of soluble recombinant RANKL and macrophage colony-stimulating factor (M-CSF; R&D Systems, Minneapolis, MN), or in coculture with a stromal cell line (Kusa O) in the presence of 1,25-dihydroxyvitamin D3 (1,25[OH]2D3; 10 nM) with prostaglandin E2 (100 nM) (47,48). Tartrate-resistant acid phosphatase (TRAP)–positive multinucleated cells were counted on days 7 and 11 (49). Statistical analysis. Results are expressed as the mean ⫾ SEM. The Mann-Whitney U test was used to compare mean histologic severity scores of test and control groups. Student’s 2-tailed t-test was used to analyze differences in tritiated thymidine incorporation and TRAP⫹ cell numbers. P values less than 0.05 were considered significant. RESULTS Necessity of CD4ⴙ T lymphocytes for AIA. The relative importance of T lymphocyte subsets and B cells in AIA has not been fully evaluated, but is important for understanding the effect of specific cytokines. Consistent with findings of previous studies (37), RAG-1⫺/⫺ mice were completely resistant to AIA (Table 1, experiment 1). The importance of CD4⫹ T lymphocytes in AIA was investigated next. GK mice, which were rendered CD4⫹ T lymphocyte–deficient by transgenic expression of an anti-CD4⫹ antibody (38), failed to develop significant AIA (Table 1, experiment 2). The role ROLE OF IL-6 IN AIA 161 Table 1. Dependence of AIA on CD4⫹ T lymphocytes* Cellular defect of mutant Mean ⫾ SEM total histologic score WT RAG-1⫺/⫺ No mature T or B lymphocytes 17.2 ⫾ 0.8 1.3 ⫾ 0.4† WT GK mice No CD4⫹ T lymphocytes 17.1 ⫾ 0.8 4.0 ⫾ 0.9† WT ␤2m⫺/⫺ No CD8⫹ T lymphocytes 18.1 ⫾ 0.6 16.2 ⫾ 1.1 WT MT/MT No B lymphocytes or IgM or IgG antibodies 17.7 ⫾ 0.5 15.7 ⫾ 1.0 Experiment, genotype 1 2 3 4 * Antigen-induced arthritis (AIA) was induced as described in Materials and Methods. Frontal sections from arthritic knee joints (at least 10 joints and 6 mice per group) were assessed by an investigator with no knowledge of the experimental groups, and graded in severity from 0 (normal) to 5 (severe) for the following components: joint exudate, synovitis, pannus, cartilage destruction, and bone destruction. The mean total histologic severity score in each experimental group (lymphocyte-deficient and wild-type [WT]) was calculated by averaging the sum of the score for these 5 histologic features (maximum possible score ⫽ 25). All mutant mice were either derived on or backcrossed onto a C57BL/6 background for at least 10 generations and were age- and sex-matched with WT controls. RAG-1⫺/⫺ ⫽ recombination-activating gene 1 deficient; ␤2m⫺/⫺ ⫽ ␤2 microglobulin deficient. † P ⬍ 0.001 versus WT mice. of CD8⫹ lymphocytes in AIA was assessed by using ␤2m⫺/⫺ mice, which lack CD8⫹ T lymphocytes (39). There was no reduction in AIA in ␤2m⫺/⫺ mice compared with wild-type mice (Table 1, experiment 3). To evaluate the role of B lymphocytes in AIA, MT/MT mice, which lack mature B lymphocytes and do not produce IgM or IgG antibodies (40), were examined. AIA in these mice was similar to that in control animals (Table 1, experiment 4). These data confirm that CD4⫹ T lymphocytes are essential for disease in the acute phase of AIA and that neither CD8⫹ T lymphocytes nor B cells are involved. Greater protective effect of IL-6 deficiency than TNF deficiency. AIA was induced in wild-type, IL-6⫺/⫺, and TNF⫺/⫺ mice. The overall severity of AIA was reduced by ⬃30% in IL-6⫺/⫺ mice compared with wild-type controls (Figures 1A and B). In particular, there was a marked reduction in osteoclast recruitment to inflamed joints, as measured by the number of osteoclasts per millimeter of bone perimeter (mean ⫾ SEM 2.7 ⫾ 0.5 in wild-type mice versus 1.1 ⫾ 0.5 in IL-6⫺/⫺ mice; P ⬍ 0.01) (Figure 1C). Since TNF is important in the pathogenesis of RA (1), we compared the protective benefits of IL-6 and TNF deficiency in Figure 1. Reduction in antigen-induced arthritis and osteoclast recruitment to the arthritic joint in interleukin-6–deficient (IL-6⫺/⫺) mice. Mice were killed 7 days after intraarticular injection of methylated bovine serum albumin (day 14 from initial treatment) and the knee joints were removed and processed as described in Materials and Methods. A, Frontal sections of knee joints from a naive wild-type (WT) C57BL/6 mouse (i), an arthritic WT mouse (ii), and an IL-6⫺/⫺ mouse (iii) (hematoxylin and eosin [H&E] stained; original magnification ⫻ 40). Magnified views of pannus (PN) adjacent to the patella (P) in an arthritic WT mouse (iv) and an IL-6⫺/⫺ mouse (v), respectively, are shown (H&E stained; original magnification ⫻ 100). Representative section of a knee joint of an arthritic WT mouse shows multinucleated osteoclasts (arrows) resorbing bone (Bn) (H&E stained; original magnification ⫻ 200). Severity of the arthritic changes was graded from 0 (normal) to 5 (severe) for exudate, synovitis (S), pannus, cartilage (C) destruction, and bone destruction. E ⫽ exudate; F ⫽ femur. B, Total histologic score (mean and SEM; n ⫽ 15 joints per group, n ⫽ 10 mice per group pooled from 2 independent experiments). C, Number of osteoclasts relative to bone surface, measured using standard histomorphometric procedures as described in Materials and Methods (mean and SEM; n ⫽ 15 joints per group, n ⫽ 10 mice per group pooled from 2 independent experiments). BPm ⫽ bone perimeter. ⴱ ⫽ P ⬍ 0.01 versus WT mice. 162 inflammatory arthritis by inducing AIA in IL-6⫺/⫺ and TNF⫺/⫺ mice. Both types of mutant mice had less disease compared with wild-type mice (mean ⫾ SEM total histologic score 21.5 ⫾ 0.5 in wild-type mice [n ⫽ 14 joints]; P ⬍ 0.001 versus TNF⫺/⫺ and IL-6⫺/⫺ mice) (Figure 2A). IL-6⫺/⫺ mice had reduced disease compared with TNF⫺/⫺ mice (mean ⫾ SEM total histologic scores 17.4 ⫾ 0.5 in TNF⫺/⫺ mice and 13.8 ⫾ 0.8 in IL-6⫺/⫺ mice [n ⫽ 16 joints per group]; P ⬍ 0.01) (Figures 2A and B). To confirm our finding of reduced disease in TNF⫺/⫺ mice and IL-6⫺/⫺ mice, we induced AIA in wild-type mice and administered a neutralizing antibody to TNF (XT22) or isotype control antibody (GL113). IL-6⫺/⫺ mice had a greater reduction in disease compared with wild-type mice that received TNF blockade treatment (total histologic score 19.4 ⫾ 0.4 in wild-type mice, 17.4 ⫾ 1.3 in GL113-treated wild-type mice, 14.5 ⫾ 0.5 in XT22-treated wild-type mice, and 12.9 ⫾ 0.5 in IL-6⫺/⫺ mice [n ⫽ 10 joints per group]; P ⬍ 0.05 for IL-6⫺/⫺ versus XT22-treated wild-type mice and P ⬍ 0.01 for IL-6⫺/⫺ and XT22-treated wild-type mice versus wild-type or GL113-treated wild-type mice) (Figure 2C). These findings show that IL-6 deficiency confers equal or greater protection against inflammation in AIA compared with TNF deficiency or blockade treatment. Independence of AIA from IL-11 and OSM signaling. To determine whether the protective effect of IL-6 deficiency was specific, we induced AIA in the absence of other IL-6 family members. There was no difference in arthritis severity between wild-type and IL-11R1⫺/⫺ mice (total histologic score 18.1 ⫾ 0.4 in wild-type mice versus 18.6 ⫾ 0.6 in IL-11R1⫺/⫺ mice [n ⫽ 10 joints per group]) (Figure 3A). The absence of both IL-6 and IL-11 signaling in IL-6⫺/⫺/IL-11R1⫺/⫺ mice did not confer additional protection relative to that seen in IL-6⫺/⫺ mice (total histologic score 8.9 ⫾ 0.4 in IL-6⫺/⫺/IL-11R1⫺/⫺ mice versus 9.8 ⫾ 0.4 in IL-6⫺/⫺ mice [n ⫽ 10 joints per group]) (Figure 3A). There was only a minor difference in disease severity between wild-type mice and OSMR⫺/⫺ mice (total histologic score 21 ⫾ 0.4 in wild-type mice versus 18 ⫾ 0.5 in OSMR⫺/⫺ mice [n ⫽ 14 joints per group]) (Figure 3B). These findings indicate that the protective effect of IL-6 deficiency in AIA is specific for IL-6, and that AIA is probably independent of IL-11 and OSM signaling. IL-6ⴚ/ⴚ T lymphocytes show reduced antigenspecific proliferation. Having established that AIA is dependent on CD4⫹ T lymphocytes but not CD8⫹ T lymphocytes or B cells, and that IL-6 deficiency confers significant protection against AIA, we next investigated WONG ET AL Figure 2. IL-6 deficiency is more protective than tumor necrosis factor deficiency (TNF⫺/⫺) in antigen-induced arthritis (AIA). Mice were killed on day 14 of AIA, and the knee joints were removed and processed as described in Materials and Methods. Severity of the arthritic changes in frontal sections of knee joints was graded from 0 (normal) to 5 (severe) for joint exudate, synovitis, pannus, cartilage destruction, and bone destruction. The mean total histologic severity score was calculated by averaging the sum of these 5 features. A, IL-6⫺/⫺ mice developed less severe AIA than TNF⫺/⫺ mice (values are the mean and SEM total histologic score; n ⫽ 14–16 joints per group, n ⫽ 9–10 mice per group pooled from 2 independent experiments). ⴱ ⫽ P ⬍ 0.001 versus WT mice; # ⫽ P ⬍ 0.01 versus TNF⫺/⫺ mice. B, Representative frontal sections of knee joints from an arthritic TNF⫺/⫺ mouse (i) and an IL-6⫺/⫺ mouse (ii) (H&E stained; original magnification ⫻ 100). C, IL-6⫺/⫺ mice had a greater reduction in disease compared with WT mice who were administered a TNF-neutralizing antibody (0.5 mg/day XT22 on days 6, 9, and 11 of AIA) (values are the mean and SEM total histologic score; n ⫽ 10 joints per group, n ⫽ 7 mice per group). GL113 ⫽ isotype control antibody (n ⫽ 10 joints per group, n ⴝ 7 mice per group). ⴱ ⫽ P ⬍ 0.01 versus WT and GL113-treated mice; # ⫽ P ⬍ 0.05 versus XT22-treated mice. See Figure 1 for other definitions. antigen-specific T lymphocyte proliferation in IL-6⫺/⫺ mice. T lymphocytes from draining LNs of wild-type or IL-6⫺/⫺ mice with AIA were cultured in the presence of ROLE OF IL-6 IN AIA Figure 3. Independence of antigen-induced arthritis (AIA) from IL-11 and oncostatin M (OSM) signaling. Mice were killed on day 14 of AIA, and the knee joints were removed and processed as described in Materials and Methods. A, WT and IL-11 receptor antagonist 1–deficient (IL-11Ra1⫺/⫺) mice developed similar levels of disease (values are the mean and SEM; n ⫽ 10 joints per group, n ⫽ 7 mice per group). The absence of both IL-6 and IL-11 signaling in the IL-6⫺/⫺/ IL-11Ra1⫺/⫺ mice did not confer additional protection relative to that seen in IL-6⫺/⫺ mice. ⴱ ⫽ P ⬍ 0.01 versus WT and IL-11Ra1⫺/⫺ mice. B, WT mice and mice deficient in OSM receptor (OSMR) developed similar levels of disease (values are the mean and SEM; n ⫽ 14 joints per group, n ⫽ 9 mice per group). See Figure 1 for other definitions. mBSA for 72 hours. T lymphocytes from IL-6⫺/⫺ mice exhibited ⬃30% reduction in antigen-specific proliferation compared with cells from wild-type mice (Figure 4A), with no difference in mitogenic response following Con A stimulation (data not shown). IL-6 has been reported to regulate DC differentiation (32), and DCs of IL-6⫺/⫺ mice may induce less antigen-specific T lymphocyte activation (31). However, we found that reduced proliferation in purified CD4⫹ T lymphocytes of IL6⫺/⫺ mice persisted following culture with mBSA-coated APCs from wild-type mice (Figure 4B). There was no difference in proliferation following culture with APCs not coated with mBSA (Figure 4B) or following Con A stimulation (data not shown). DC antigen presentation function was directly assessed by comparing the ability of DCs from IL-6⫺/⫺ and wild-type mice to present OVA to OVA-specific T lymphocytes. OVA-pulsed DCs from IL-6⫺/⫺ mice produced equivalent proliferation of OT-I CD8⫹ T lymphocytes (Figure 4C) and OT-II CD4⫹ T lymphocytes (Figure 4D) compared with OVA-pulsed DCs from wild-type mice. These findings suggest that the reduced antigen-specific proliferation of T lymphocytes in IL-6⫺/⫺ mice is not due to defective antigen presentation function by DCs. T lymphocytes in IL-6ⴚ/ⴚ mice produce fewer pro-osteoclastogenic cytokines. IL-17 is produced by T lymphocytes and has been shown to promote osteoclastogenesis (15). Because IL-6⫺/⫺ mice showed reduced 163 osteoclast recruitment to erosive sites in arthritic joints (Figure 1C), we investigated T lymphocyte production of IL-17. T lymphocytes in IL-6⫺/⫺ mice produced 6-fold less IL-17 compared with wild-type controls following antigen-specific stimulation in vitro (with 50 g/ml mBSA, mean ⫾ SEM 1.2 ⫾ 0.5 ng/ml in wild-type mice versus 0.2 ⫾ 0.1 in IL-6⫺/⫺ mice; P ⬍ 0.01) (Figure 5A). There was no difference in IL-17 production following Con A stimulation (data not shown), which indicates that T lymphocytes in IL-6⫺/⫺ mice can respond to mitogenic stimuli and that the reduced IL-17 production was from antigen-specific T lymphocytes. Figure 4. IL-6⫺/⫺ lymphocytes have reduced antigen-specific proliferation, but not due to defective antigen presentation by dendritic cells (DCs). A, Lymphocytes from draining lymph nodes (LNs) of arthritic WT mice and IL-6⫺/⫺ mice were cultured in the presence of methylated bovine serum albumin (mBSA) for 72 hours and pulsed with tritiated thymidine for the last 8 hours. There was no difference in mitogenic response following concanavalin A stimulation (data not shown). LNs were pooled from 10 mice per group. ⴱ ⫽ P ⬍ 0.05 versus WT mice. B, Purified CD4⫹ lymphocytes from draining LNs of arthritic WT mice and IL-6⫺/⫺ mice were cultured for 72 hours in the presence of irradiated WT antigen-presenting cells (APCs) coated with mBSA. Tritiated thymidine incorporation was used to measure cellular proliferation. Wells containing APCs not coated with mBSA or having no APCs were used as negative controls. LNs were pooled from 8 mice per group. ⴱ ⫽ P ⬍ 0.05 versus WT mice. C and D, Proliferation of C, ovalbumin (OVA)–specific OVA-transgenic I (OT-I) CD8⫹ and D, OT-II CD4⫹ T lymphocytes in response to OVA-pulsed DCs from WT and IL-6⫺/⫺ mice. After 3–4 days, cellular proliferation was measured by tritiated thymidine incorporation for the last 18 hours. Values are the mean and SEM (n ⫽ 3 mice per group). PBS ⫽ phosphate buffered saline (see Figure 1 for other definitions). 164 Figure 5. Reduced production of pro-osteoclastogenic cytokines by IL-6⫺/⫺ T lymphocytes. A, IL-17 in supernatants from the same cultures used in the studies shown in Figure 4A, harvested at 72 hours, was measured by enzyme-linked immunosorbent assay (ELISA). There was no difference in IL-17 levels following concanavalin A stimulation (data not shown). ⴱ ⫽ P ⬍ 0.01 versus WT mice. B, IL-17 in supernatants from the same cultures used in the studies shown in Figure 4B, harvested at 72 hours, was measured by ELISA. CD4⫹ T lymphocytes from IL-6⫺/⫺ mice produced no detectable IL-17 (ⴱ ⫽ P ⬍ 0.05 versus WT mice). Values are the mean and SEM. C, Total RNA was extracted from T lymphocyte cultures used in the studies shown in Figure 4A, and cDNA was generated via reverse transcriptase–polymerase chain reaction. To determine expression of RANKL and osteoprotegerin (OPG) relative to ␤-actin, semiquantitative polymerase chain reaction was performed with primers specific for RANKL, OPG, and ␤-actin as described in Materials and Methods. Following antigen-specific stimulation in vitro, the ratio of RANKL mRNA to OPG mRNA was up to 10-fold higher in WT compared with IL-6⫺/⫺ mouse lymphocytes from draining lymph nodes. Values are the mean. mBSA ⫽ methylated bovine serum albumin; APC ⫽ antigen-presenting cells (see Figure 1 for other definitions). To further clarify the cellular origin of the defective IL-17 production, ELISA was used to measure supernatants from purified CD4⫹ T lymphocytes from draining LNs in coculture with mBSA-coated APCs in wild-type mice. CD4⫹ T lymphocytes in wild-type mice had a dose-dependent increase in IL-17 production following stimulation with mBSA-coated APCs (Figure 5B). In comparison, IL-17 was undetectable in supernatants from antigen-stimulated CD4⫹ T lymphocytes in IL-6⫺/⫺ mice (Figure 5B). One of the most important determinants of osteoclastogenesis is the ratio of RANKL to OPG (50). Following antigen-specific stimulation in vitro, the ratio of RANKL to OPG messenger RNA was up to 10-fold higher in lymphocytes of wild- WONG ET AL type mice compared with T lymphocytes of IL-6⫺/⫺ mice (when normalized to ␤-actin) (Figure 5C). IFN␥ has been reported to have an inhibitory effect on osteoclastogenesis (51). Hence, a possible explanation for the observed reduction in osteoclast recruitment to arthritic joints in IL-6⫺/⫺ mice may have been elevated local production of IFN␥. However, there was no difference in T lymphocyte production of IFN␥ from antigen-stimulated cultures (data not shown). This finding demonstrates that reduced IL-17 production by antigen-specific CD4⫹ T lymphocytes in IL-6⫺/⫺ mice is not due to a generalized defect in cytokine production. Taken together, these data suggest that IL-6 may modulate production of the pro-osteoclastogenic cytokines IL-17 and RANKL (relative to OPG) by antigen-specific T lymphocytes, which might help explain reduced bone destruction in IL-6⫺/⫺ mice with AIA. Osteoclast generation in bone marrow of arthritic IL-6ⴚ/ⴚ mice. Naive IL-6⫺/⫺ mice have been shown to have normal osteoclast numbers in vivo (52) and in vitro (49). We hypothesized that the reduction in osteoclast recruitment to inflamed joints in AIA may also relate to reduced expansion of osteoclast precursors in bone marrow of arthritic IL-6⫺/⫺ mice. Therefore, we examined osteoclast generation in bone marrow harvested from arthritic mice. Following RANKL and M-CSF stimulation for 11 days, bone marrow from IL-6⫺/⫺ mice generated ⬃50% fewer osteoclasts than bone marrow from wild-type mice or TNF⫺/⫺ arthritic mice (mean ⫾ SEM TRAP⫹ cells per well 135 ⫾ 24 in bone marrow of wild-type mice, 175 ⫾ 24 in bone marrow of TNF⫺/⫺ mice, and 78 ⫾ 13 in bone marrow of IL-6⫺/⫺ mice; P ⬍ 0.01, IL-6⫺/⫺ mice versus either wild-type or TNF⫺/⫺ mice) (Figure 6A). To determine if this difference was present in naive mice, we cultured bone marrow from nonarthritic wild-type, TNF⫺/⫺, and IL-6⫺/⫺ mice in the presence of RANKL and M-CSF. Bone marrow from naive IL-6⫺/⫺ mice generated ⬃50% fewer osteoclasts than that from wild-type or TNF⫺/⫺ mice (data not shown). Although TNF⫺/⫺ mice developed less severe AIA than wild-type mice (Figure 2A), bone marrow from arthritic TNF⫺/⫺ and wild-type mice generated the same number of osteoclasts in vitro (Figure 6A). These findings were confirmed when bone marrow from arthritic wild-type, TNF⫺/⫺, and IL-6⫺/⫺ mice was cultured for 7 days in the presence of Kusa O stromal cells and 1,25(OH)2D3 (TRAP⫹ cells per well 504 ⫾ 84 in bone marrow of wild-type mice, 498 ⫾ 71 in bone marrow of TNF⫺/⫺ mice, and 312 ⫾ 34 in bone marrow of IL-6⫺/⫺ mice; P ⬍ 0.01, IL-6⫺/⫺ mice versus either wild-type mice or ROLE OF IL-6 IN AIA Figure 6. Reduced generation of osteoclasts in vitro by bone marrow from arthritic IL-6⫺/⫺ mice. Bone marrow preparations flushed from the femora and tibiae of arthritic WT mice, tumor necrosis factor (TNF)–deficient mice, and IL-6⫺/⫺ mice (day 14) were plated at a density of 105 cells/well A, in the presence of soluble recombinant glutathione S-transferase–RANKL and macrophage colonystimulating factor or B, in coculture with a stromal cell line in the presence of 1,25-hydroxyvitamin D 3 . Tartrate-resistant acid phosphatase–positive multinucleated cells were counted on days 7 and 11. Values are the mean and SEM from 2 independent experiments; bone marrow was pooled from a total of 10 mice per group. ⴱ ⫽ P ⬍ 0.01 versus WT mice and TNF⫺/⫺ mice. See Figure 1 for other definitions. TNF⫺/⫺ mice) (Figure 6B). Our findings suggest that IL-6 (but not TNF) may be implicated in the regulation of osteoclast precursors in the bone marrow, before and during inflammatory arthritis. DISCUSSION In these studies, we used a murine model of acute, destructive inflammatory arthritis to further investigate the effects of endogenous IL-6. While AIA is a widely used model of inflammatory arthritis, the cellular mediators have not been completely defined. We used gene-knockout mice to demonstrate that AIA, at least in 165 the acute phase, is completely dependent on CD4⫹ T lymphocytes, but not CD8⫹ or B lymphocytes. These findings do not, however, exclude a role for B cells or CD8⫹ T lymphocytes in the chronic phase of AIA. Having characterized acute AIA as a CD4⫹ T lymphocyte–dependent model of disease, we investigated the role of IL-6 in detail. IL-6⫺/⫺ mice have previously been shown to have reduced AIA, with increased production of Th2 cytokines from draining LNs (30,33). We confirmed that IL-6⫺/⫺ mice had significantly reduced levels of AIA. Furthermore, we found that IL-6 deficiency conferred at least as much protection against AIA as the absence of TNF. These findings strengthen the rationale for IL-6 blockade in the treatment of inflammatory joint disease and encouraged us to undertake mechanistic studies. All cytokines in the IL-6 family use the common transmembrane receptor, gp130, as part of a multimeric complex (for review, see ref. 53). Both IL-11 (54) and OSM (55) have been implicated in the pathogenesis of inflammatory arthritis. However, we found that AIA was independent of both IL-11 and OSM signaling. Furthermore, the combined absence of both IL-6 and IL-11 conferred no additional protection relative to that seen in the absence of IL-6 signaling alone. These findings suggest that the protective effect of IL-6 deficiency in AIA is quite specific. Our finding that IL-6⫺/⫺ mice had markedly reduced osteoclast recruitment to sites of joint disease in AIA led us to investigate the effect of IL-6 on the RANKL:OPG ratio and T lymphocyte production of IL-17. IL-17 has been reported to induce IL-6 and IL-8 production by fibroblasts (56), enhance IL-1 and TNF production by monocytes (57), and up-regulate nitric oxide and prostaglandin production by chondrocytes (58). Elevated levels of IL-17 have been found in RA synovial fluid (21). RA synovial fluid induced osteoclastogenesis in cocultures of bone marrow cells with osteoblasts (21). IL-17 also promoted bone erosion in murine CIA through modulation of the RANKL:OPG ratio (23). RANKL can directly induce peripheral blood monocyte differentiation into osteoclasts (15). Synovial macrophages can undergo osteoclast differentiation in the presence of both RANKL and M-CSF (59). Activated T lymphocytes from RA patients have been shown to express RANKL (13). We found that CD4⫹ T lymphocytes from draining LNs of arthritic IL-6⫺/⫺ mice had reduced antigenspecific proliferative responses, produced less IL-17, and had a lower ratio of RANKL to OPG. Early studies showed that IL-6⫺/⫺ mice had defective T helper– 166 dependent responses to viral infection (35). Subsequent work demonstrated that T lymphocytes in IL-6⫺/⫺ mice had impaired antigen-specific proliferation, probably due to reduced IL-2 receptor expression (60) rather than decreased IL-2 production (61). Given that IL-6 regulates DC differentiation (32) and function (31), we were careful to exclude defective DC function as a possible explanation for the reduced antigen-specific proliferation of T lymphocytes in IL-6⫺/⫺ mice. IFN␥ has been reported to inhibit osteoclastogenesis by enhancing degradation of TNF receptor–associated factor 6 (51). However, we found that the absence of IL-6 had no effect on T lymphocyte production of IFN␥. Our findings suggest that IL-6 selectively modulates activated T lymphocyte production of IL-17 and RANKL/OPG. Targeted treatment with IL-6 blockade in RA may thus have a particularly favorable effect on T lymphocyte– mediated osteoclast recruitment and activation, thereby reducing joint destruction. Naive IL-6⫺/⫺ mice have been reported to have normal osteoclast numbers in vivo (52) and in vitro (49). We found that bone marrow from naive IL-6⫺/⫺ mice generated fewer osteoclasts than bone marrow from wild-type mice, possibly because the bone marrow was cultured in the presence of RANKL and M-CSF, rather than 1,25(OH)2D3, as in a previous study (49). Although our data show a defect in basal osteoclastogenesis in naive IL-6⫺/⫺ mice, this does not appear to be functionally important, since basal histomorphometric parameters were normal in naive IL-6⫺/⫺ mice (49). Perhaps of greater interest is that IL-6⫺/⫺ mice were protected against bone loss following estrogen depletion (52) and ethanol ingestion (49). We found that bone marrow from arthritic IL-6⫺/⫺ mice generated fewer osteoclasts than that from wild-type controls after culture in the presence of pro-osteoclastogenic stimuli. IL-6 may therefore be more important in regulating bone marrow osteoclastogenesis and systemic osteoporosis in disease states, such as inflammatory arthritis. In contrast, bone marrow from arthritic TNF⫺/⫺ mice generated the same number of osteoclasts as bone marrow from arthritic wild-type mice. Therefore, although both TNF and IL-6 are important in the pathogenesis of inflammatory arthritis, IL-6 may selectively mediate inflammation-induced expansion of osteoclast precursors in the bone marrow. These observations suggest that IL-6 may mediate bone destruction during AIA by 3 possible mechanisms. The reduced joint destruction observed in IL-6⫺/⫺ mice during AIA may have been, at least partially, a consequence of generally reduced joint in- WONG ET AL flammation. IL-6 may specifically modulate local T lymphocyte production of IL-17, RANKL, and OPG during inflammatory arthritis. Unopposed RANKL and synergy between IL-17 and other proinflammatory cytokines, such as IL-1 and TNF, may then induce local synovial monocyte/macrophage differentiation into osteoclasts. Our findings also suggest that IL-6 deficiency leads to reduced osteoclastogenesis from bone marrow precursors during inflammatory disease, such as AIA. Systemic osteoporosis is a poorly understood feature of RA. These findings raise the possibility that IL-6 plays a major role in systemic bone loss, as well as local osteoporosis, during inflammatory joint disease. Our findings strengthen the rationale for targeting IL-6 in inflammatory arthritis. Therapeutic inhibition of IL-6 may reduce generalized osteoporosis as well as joint inflammation and local joint destruction in RA. ACKNOWLEDGMENTS We thank S. Mihajlovic and E. Tsui for histologic processing; W. Alexander, L. Corcoran, Y. Zhan, L. Harrison, D. Tarlinton, L. Robb, and A. Miyajima for provision of mice; and J. Morrow and J. Merryfull for animal husbandry. REFERENCES 1. Feldmann M. Development of anti-TNF therapy for rheumatoid arthritis. Nat Rev Immunol 2002;2:364–71. 2. Dayer JM, Bresnihan B. Targeting interleukin-1 in the treatment of rheumatoid arthritis [editorial]. Arthritis Rheum 2002;46: 574–8. 3. Moreland LW, Baumgartner SW, Schiff MH, Tindall EA, Fleischmann RM, Weaver AL, et al. Treatment of rheumatoid arthritis with a recombinant human tumor necrosis factor receptor (p75)-Fc fusion protein. N Engl J Med 1997;337:141–7. 4. Jiang Y, Genant HK, Watt I, Cobby M, Bresnihan B, Aitchison R, et al. A multicenter, double-blind, dose-ranging, randomized, placebo-controlled study of recombinant human interleukin-1 receptor antagonist in patients with rheumatoid arthritis: radiologic progression and correlation of Genant and Larsen scores. Arthritis Rheum 2000;43:1001–9. 5. Kishimoto T. Interleukin-6: from basic science to medicine: 40 years in immunology [review]. Annu Rev Immunol 2005;23:1–21. 6. Wendling D, Racadot E, Wijdenes J. Treatment of severe rheumatoid arthritis by anti-interleukin 6 monoclonal antibody. J Rheumatol 1993;20:259–62. 7. Nishimoto N, Yoshizaki K, Miyasaka N, Yamamoto K, Kawai S, Takeuchi T, et al. Treatment of rheumatoid arthritis with humanized anti–interleukin-6 receptor antibody: a multicenter, doubleblind, placebo-controlled trial. Arthritis Rheum 2004;50:1761–9. 8. Kotake S, Sato K, Kim KJ, Takahashi N, Udagawa N, Nakamura I, et al. Interleukin-6 and soluble interleukin-6 receptors in the synovial fluids from rheumatoid arthritis patients are responsible for osteoclast-like cell formation. J Bone Miner Res 1996;11: 88–95. 9. Okamoto H, Yamamura M, Morita Y, Harada S, Makino H, Ota Z. The synovial expression and serum levels of interleukin-6, ROLE OF IL-6 IN AIA 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. interleukin-11, leukemia inhibitory factor, and oncostatin M in rheumatoid arthritis. Arthritis Rheum 1997;40:1096–105. Guerne PA, Zuraw BL, Vaughan JH, Carson DA, Lotz M. Synovium as a source of interleukin 6 in vitro. J Clin Invest 1989;83:585–92. Brinckerhoff CE, Matrisian LM. Matrix metalloproteinases: a tail of a frog that became a prince [review]. Nat Rev Mol Cell Biol 2002;3:207–14. Ito A, Itoh Y, Sasaguri Y, Morimatsu M, Mori Y. Effects of interleukin-6 on the metabolism of connective tissue components in rheumatoid synovial fibroblasts. Arthritis Rheum 1992;35: 1197–201. Gravallese EM, Manning C, Tsay A, Naito A, Pan C, Amento E, et al. Synovial tissue in rheumatoid arthritis is a source of osteoclast differentiation factor. Arthritis Rheum 2000;43:250–8. Kong YY, Feige U, Sarosi I, Bolon B, Tafuri A, Morony S, et al. Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 1999; 402:304–9. Kotake S, Udagawa N, Hakoda M, Mogi M, Yano K, Tsuda E, et al. Activated human T cells directly induce osteoclastogenesis from human monocytes: possible role of T cells in bone destruction in rheumatoid arthritis patients. Arthritis Rheum 2001;44: 1003–12. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, Luthy R, et al. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 1997;89:309–19. Gravallese EM, Harada Y, Wang JT, Gorn AH, Thornhill TS, Goldring SR. Identification of cell types responsible for bone resorption in rheumatoid arthritis and juvenile rheumatoid arthritis. Am J Pathol 1998;152:943–51. Thomson BM, Mundy GR, Chambers TJ. Tumor necrosis factors ␣ and ␤ induce osteoblastic cells to stimulate osteoclastic bone resorption. J Immunol 1987;138:775–9. Quinn JM, Horwood NJ, Elliott J, Gillespie MT, Martin TJ. Fibroblastic stromal cells express receptor activator of NF-B ligand and support osteoclast differentiation. J Bone Miner Res 2000;15:1459–66. Palmqvist P, Persson E, Conaway HH, Lerner UH. IL-6, leukemia inhibitory factor, and oncostatin M stimulate bone resorption and regulate the expression of receptor activator of NF-B ligand, osteoprotegerin, and receptor activator of NF-B in mouse calvariae. J Immunol 2002;169:3353–62. Kotake S, Udagawa N, Takahashi N, Matsuzaki K, Itoh K, Ishiyama S, et al. IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J Clin Invest 1999;103:1345–52. Romas E, Gillespie MT, Martin TJ. Involvement of receptor activator of NFB ligand and tumor necrosis factor-␣ in bone destruction in rheumatoid arthritis [review]. Bone 2002;30:340–6. Lubberts E, van den Bersselaar L, Oppers-Walgreen B, Schwarzenberger P, Coenen-de Roo CJ, Kolls JK, et al. IL-17 promotes bone erosion in murine collagen-induced arthritis through loss of the receptor activator of NF-B ligand/osteoprotegerin balance. J Immunol 2003;170:2655–62. Brackertz D, Mitchell GF, Mackay IR. Antigen-induced arthritis in mice. I. Induction of arthritis in various strains of mice. Arthritis Rheum 1977;20:841–50. Brackertz D, Mitchell GF, Vadas MA, Mackay IR. Studies on antigen-induced arthritis in mice. III. Cell and serum transfer experiments. J Immunol 1977;118:1645–8. Petrow PK, Thoss K, Katenkamp D, Brauer R. Adoptive transfer of susceptibility to antigen-induced arthritis into severe combined immunodeficient (SCID) mice: role of CD4⫹ and CD8⫹ T cells. Immunol Invest 1996;25:341–53. Nissler K, Pohlers D, Huckel M, Simon J, Brauer R, Kinne RW. Anti-CD4 monoclonal antibody treatment in acute and early 167 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. chronic antigen induced arthritis: influence on macrophage activation. Ann Rheum Dis 2004;63:1470–7. Pohlers D, Nissler K, Frey O, Simon J, Petrow PK, Kinne RW, et al. Anti-CD4 monoclonal antibody treatment in acute and early chronic antigen-induced arthritis: influence on T helper cell activation. Clin Exp Immunol 2004;135:409–15. Boe A, Baiocchi M, Carbonato M, Papoian R, Serlupi-Crescenzi O. Interleukin 6 knock-out mice are resistant to antigen-induced experimental arthritis. Cytokine 1999;11:1057–64. Ohshima S, Saeki Y, Mima T, Sasai M, Nishioka K, Nomura S, et al. Interleukin 6 plays a key role in the development of antigeninduced arthritis. Proc Natl Acad Sci U S A 1998;95:8222–6. Bleier JI, Pillarisetty VG, Shah AB, DeMatteo RP. Increased and long-term generation of dendritic cells with reduced function from IL-6-deficient bone marrow. J Immunol 2004;172:7408–16. Park SJ, Nakagawa T, Kitamura H, Atsumi T, Kamon H, Sawa S, et al. IL-6 regulates in vivo dendritic cell differentiation through STAT3 activation. J Immunol 2004;173:3844–54. De Hooge AS, van de Loo FA, Arntz OJ, van den Berg WB. Involvement of IL-6, apart from its role in immunity, in mediating a chronic response during experimental arthritis [published erratum appears in Am J Pathol 2001;158:1545]. Am J Pathol 2000; 157:2081–91. Kobayashi H, Ohshima S, Nishioka K, Yamaguchi N, UmeshitaSasai M, Ishii T, et al. Antigen induced arthritis (AIA) can be transferred by bone marrow transplantation: evidence that interleukin 6 is essential for induction of AIA. J Rheumatol 2002;29: 1176–82. Kopf M, Baumann H, Freer G, Freudenberg M, Lamers M, Kishimoto T, et al. Impaired immune and acute-phase responses in interleukin-6-deficient mice. Nature 1994;368:339–42. Campbell IK, O’Donnell K, Lawlor KE, Wicks IP. Severe inflammatory arthritis and lymphadenopathy in the absence of TNF. J Clin Invest 2001;107:1519–27. Mombaerts P, Iacomini J, Johnson RS, Herrup K, Tonegawa S, Papaioannou VE. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 1992;68:869–77. Zhan Y, Corbett AJ, Brady JL, Sutherland RM, Lew AM. Delayed rejection of fetal pig pancreas in CD4 cell deficient mice was correlated with residual helper activity. Xenotransplantation 2000; 7:267–74. Serreze DV, Leiter EH, Christianson GJ, Greiner D, Roopenian DC. Major histocompatibility complex class I-deficient NOD-␤2m null mice are diabetes and insulitis resistant. Diabetes 1994;43: 505–9. Kitamura D, Roes J, Kuhn R, Rajewsky K. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin chain gene. Nature 1991;350:423–6. Nandurkar HH, Robb L, Tarlinton D, Barnett L, Kontgen F, Begley CG. Adult mice with targeted mutation of the interleukin-11 receptor (IL11Ra) display normal hematopoiesis. Blood 1997;90:2148–59. Tanaka M, Hirabayashi Y, Sekiguchi T, Inoue T, Katsuki M, Miyajima A. Targeted disruption of oncostatin M receptor results in altered hematopoiesis. Blood 2003;102:3154–62. Lawlor KE, Campbell IK, O’Donnell K, Wu L, Wicks IP. Molecular and cellular mediators of interleukin-1–dependent acute inflammatory arthritis. Arthritis Rheum 2001;44:442–50. Sims NA, Green JR, Glatt M, Schlict S, Martin TJ, Gillespie MT, et al. Targeting osteoclasts with zoledronic acid prevents bone destruction in collagen-induced arthritis. Arthritis Rheum 2004; 50:2338–46. Lawlor KE, Campbell IK, Metcalf D, O’Donnell K, van Nieuwenhuijze A, Roberts AW, et al. Critical role for granulocyte colonystimulating factor in inflammatory arthritis. Proc Natl Acad Sci U S A 2004;101:11398–403. Pooley JL, Heath WR, Shortman K. Cutting edge: intravenous 168 47. 48. 49. 50. 51. 52. 53. 54. 55. WONG ET AL soluble antigen is presented to CD4 T cells by CD8⫺ dendritic cells, but cross-presented to CD8 T cells by CD8⫹ dendritic cells. J Immunol 2001;166:5327–30. Quinn JM, Whitty GA, Byrne RJ, Gillespie MT, Hamilton JA. The generation of highly enriched osteoclast-lineage cell populations. Bone 2002;30:164–70. Allan EH, Ho PW, Umezawa A, Hata J, Makishima F, Gillespie MT, et al. Differentiation potential of a mouse bone marrow stromal cell line. J Cell Biochem 2003;90:158–69. Dai J, Lin D, Zhang J, Habib P, Smith P, Murtha J, et al. Chronic alcohol ingestion induces osteoclastogenesis and bone loss through IL-6 in mice. J Clin Invest 2000;106:887–95. Teitelbaum SL. Bone resorption by osteoclasts [review]. Science 2000;289:1504–8. Takayanagi H, Ogasawara K, Hida S, Chiba T, Murata S, Sato K, et al. T-cell-mediated regulation of osteoclastogenesis by signalling cross-talk between RANKL and IFN-␥. Nature 2000;408:600–5. Poli V, Balena R, Fattori E, Markatos A, Yamamoto M, Tanaka H, et al. Interleukin-6 deficient mice are protected from bone loss caused by estrogen depletion. EMBO J 1994;13:1189–6. Wong PK, Campbell IK, Egan PJ, Ernst M, Wicks IP. The role of the interleukin-6 family of cytokines in inflammatory arthritis and bone turnover [review]. Arthritis Rheum 2003;48:1177–89. Wong PK, Campbell IK, Robb L, Wicks IP. Endogenous IL-11 is pro-inflammatory in acute methylated bovine serum albumin/ interleukin-1-induced (mBSA/IL-1) arthritis. Cytokine 2005;29: 72–6. Plater-Zyberk C, Buckton J, Thompson S, Spaull J, Zanders E, Papworth J, et al. Amelioration of arthritis in two murine models 56. 57. 58. 59. 60. 61. using antibodies to oncostatin M. Arthritis Rheum 2001;44: 2697–702. Fossiez F, Djossou O, Chomarat P, Flores-Romo L, Ait-Yahia S, Maat C, et al. T cell interleukin-17 induces stromal cells to produce proinflammatory and hematopoietic cytokines. J Exp Med 1996;183:2593–603. Jovanovic DV, Di Battista JA, Martel-Pelletier J, Jolicoeur FC, He Y, Zhang M, et al. IL-17 stimulates the production and expression of proinflammatory cytokines, IL-␤ and TNF-␣, by human macrophages. J Immunol 1998;160:3513–21. LeGrand A, Fermor B, Fink C, Pisetsky DS, Weinberg JB, Vail TP, et al. Interleukin-1, tumor necrosis factor ␣, and interleukin-17 synergistically up-regulate nitric oxide and prostaglandin E2 production in explants of human osteoarthritic knee menisci. Arthritis Rheum 2001;44:2078–83. Itonaga I, Fujikawa Y, Sabokbar A, Murray DW, Athanasou NA. Rheumatoid arthritis synovial macrophage-osteoclast differentiation is osteoprotegerin ligand-dependent. J Pathol 2000;192: 97–104. La Flamme AC, Pearce EJ. The absence of IL-6 does not affect Th2 cell development in vivo, but does lead to impaired proliferation, IL-2 receptor expression, and B cell responses. J Immunol 1999;162:5829–37. Takeda K, Kaisho T, Yoshida N, Takeda J, Kishimoto T, Akira S. Stat3 activation is responsible for IL-6-dependent T cell proliferation through preventing apoptosis: generation and characterization of T cell-specific Stat3-deficient mice. J Immunol 1998;161: 4652–60.