Inhibition of interleukin-1 -stimulated production of matrix metalloproteinases by hyaluronan via CD44 in human articular cartilage.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 50, No. 2, February 2004, pp 516–525 DOI 10.1002/art.20004 © 2004, American College of Rheumatology Inhibition of Interleukin-1␤–Stimulated Production of Matrix Metalloproteinases by Hyaluronan via CD44 in Human Articular Cartilage Sohel M. Julovi, Tadashi Yasuda, Makoto Shimizu, Teruko Hiramitsu, and Takashi Nakamura Objective. To investigate the mechanism of the inhibitory action of hyaluronan (HA) on interleukin-1␤ (IL-1␤)–stimulated production of matrix metalloproteinases (MMPs) in human articular cartilage. Methods. IL-1␤ was added to normal and osteoarthritic (OA) human articular cartilage in explant culture to stimulate MMP production. Articular cartilage was incubated or preincubated with a clinically used form of 800-kd HA to assess its effect on IL-1␤– induced MMPs. Levels of secreted MMPs 1, 3, and 13 in conditioned media were detected by immunoblotting; intracellular MMP synthesis in chondrocytes was evaluated by immunofluorescence microscopy. Penetration of HA into cartilage tissue and its binding to CD44 were analyzed by fluorescence microscopy using fluoresceinated HA. Blocking experiments with anti-CD44 antibody were performed to investigate the mechanism of action of HA. Results. Treatment and pretreatment with 800-kd HA at 1 mg/ml resulted in significant suppression of IL-1␤–stimulated production of MMPs 1, 3, and 13 in normal and OA cartilage explant culture. Fluorescence histocytochemistry revealed that HA penetrated cartilage tissue and localized in the pericellular matrix around chondrocytes. HA-binding blocking experiments using anti-CD44 antibody demonstrated that the association of HA with chondrocytes was mediated by CD44. Preincubation with anti-CD44 antibody, which suppressed IL-1␤–stimulated MMPs, reversed the inhibi- tory effect of HA on MMP production that was induced by IL-1␤ in normal and OA cartilage. Conclusion. This study demonstrates that HA effectively inhibits IL-1␤–stimulated production of MMP-1, MMP-3, and MMP-13, which supports the clinical use of HA in the treatment of OA. The action of HA on IL-1␤ may involve direct interaction between HA and CD44 on chondrocytes. Osteoarthritis (OA) is the most prevalent disease of articular joints and is the major cause of disability in the elderly. Pathophysiologic changes occur in OA cartilage due to the excessive expression of cartilagedegrading proteinases, the resultant progressive breakdown of collagen fibers, and the degradation of proteoglycan, mainly aggrecan (1). Matrix metalloproteinases (MMPs) are zinccontaining, calcium-dependent proteinases, which collectively degrade all components of the extracellular matrix. MMPs are considered to be important in the chondrolytic processes that contribute to the degenerative changes in OA cartilage (2–4). Recent studies have identified the messenger RNA (mRNA) for some MMPs, such as MMP-1, MMP-3, MMP-9, and MMP-13, in human OA cartilage (4,5), and other investigators have reported specific MMP proteins and collagenasemediated type II collagen degradation products (6,7). There is a consensus that these enzymes play a critical role in intrinsic chondrocyte-mediated degenerative changes of the cartilage matrix in OA. Proinflammatory cytokines such as interleukin-1 (IL-1) strongly stimulate the expression of MMPs by chondrocytes in arthritis (8). Hyaluronan (HA) is a major component of synovial fluid and cartilage matrix, and it plays a central role in joint lubrication. HA is now widely used in the treatment of OA by intraarticular administration into affected joints. Although a clinical benefit of HA has been demonstrated with respect to pain relief in patients Sohel M. Julovi, MD, Tadashi Yasuda, MD, PhD, Makoto Shimizu, MD, Teruko Hiramitsu, MD, Takashi Nakamura, MD, PhD: Kyoto University Graduate School of Medicine, Kyoto, Japan. Address correspondence and reprint requests to Tadashi Yasuda, MD, PhD, Department of Orthopaedic Surgery, Kyoto University Graduate School of Medicine, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan. E-mail: firstname.lastname@example.org. Submitted for publication April 27, 2003; accepted in revised form October 13, 2003. 516 HA INHIBITION OF IL-1␤–STIMULATED MMP PRODUCTION IN ARTICULAR CARTILAGE with OA (9), the level at which the drug acts remains unclear. Recent studies have also shown that accumulation of HA at the cartilage surface blocks the penetration of a fibronectin fragment, which is a stimulator of chondrolysis and MMP production in pathologic joints (10,11), into cartilage tissue, resulting in decreased expression of MMP-3 by the fibronectin fragment in normal human articular cartilage explant culture for the first week, with no detectable blocking of MMP-3 for the second or third week (12). From these findings, the mechanism of action of HA may be a barrier effect at the cartilage surface. However, HA can penetrate into cartilage after IL-1 treatment (13). Thus, in addition to the barrier effect at the cartilage surface, different mechanisms of HA action may be at work in degenerative cartilage. Articular chondrocytes express CD44 (14), the principal cell surface receptor for HA (15). CD44 expression in chondrocytes is up-regulated by proinflammatory cytokines such as IL-1 (16,17). Anti-CD44 treatment using monoclonal antibodies has been reported to suppress joint swelling in a murine model of proteoglycan-induced arthritis (18) and inhibit cartilage destruction by rheumatoid arthritis (RA) synovial fibroblasts in vitro (19). These data suggest that CD44 may mediate inflammatory processes and joint destruction in both OA and RA. HA can also bind another cell surface receptor, intercellular adhesion molecule 1 (ICAM-1) (20), which is constitutively expressed in chondrocytes (21). Similar to CD44, proinflammatory cytokines enhance ICAM-1 expression in chondrocytes (22). Involvement of such receptors in the action of HA on articular cartilage remains to be elucidated. In this study, we attempted to identify the mechanism of HA action on the IL-1␤–stimulated production of MMP-1, MMP-3, and MMP-13 in human articular cartilage explant culture. We show herein that suppression of IL-1␤ action by HA involved the binding of HA to CD44 on chondrocytes in articular cartilage. MATERIALS AND METHODS Materials. HA of 800 kd, the form clinically used for the treatment of OA in Japan, was a gift from Seikagaku (Tokyo, Japan). Recombinant human IL-1␤ was purchased from R&D Systems (Minneapolis, MN). Anti-human MMP-1 that reacts with 53-kd and 51-kd proenzyme (M4177), antihuman MMP-3 that recognizes the 59-kd and 57-kd proenzyme (M4802), and anti-human MMP-13 that recognizes the latent proenzyme 60 kd (M4052) were obtained from Sigma (Tokyo, Japan). Alkaline phosphatase–conjugated goat antirabbit IgG was purchased from Southern Biotechnology (Bir- 517 mingham, AL). Anti-CD44 antibody (IM7) was obtained from Fujisawa Pharmaceutical (Tokyo, Japan). HA of 720 kd labeled with 5-aminofluorescein and fluorescein isothiocyanate (FITC)–conjugated OS/37 anti-CD44 antibody were obtained from Seikagaku. FITC-conjugated anti-rabbit IgG and nonspecific rat IgG2b were obtained from Sigma. Cartilage explant culture. Human femoral head cartilage without macroscopic and microscopic signs of articular degeneration such as fibrillation was obtained at the time of replacement surgery from 10 patients with femoral neck fracture. OA cartilage was obtained from the distal femur and the proximal tibia from 5 patients undergoing total knee replacement surgery. Patients were diagnosed as having OA based on the criteria developed by the American College of Rheumatology (23). Cartilage samples were placed in the wells of a 24-well Corning plate (Corning, NY) (50–60 mg/well) and maintained in 1.5 ml of Dulbecco’s modified Eagle’s medium containing 10 mM HEPES buffer, 100 units/ml of penicillin, 100 units/ml of streptomycin (Gibco BRL, Grand Island, NY), and 3.7 gm/liter of NaHCO3 (DMEM). The cartilage was precultured for 2 days at 37°C in a humidified atmosphere of 5% CO2, 95% air. At medium change on day 0, the cartilage was incubated with or without 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of 800-kd HA. Cartilage explants and conditioned media were harvested on day 4 and stored at –80°C prior to analysis. Control cultures had no additives. In another set of experiments, cartilage was treated with 2 ng/ml of IL-1␤ for 12 days with or without 1 mg/ml of HA. IL-1␤ and HA were freshly added at medium changes on days 0, 4, and 8. Cartilage explants and conditioned media were harvested on day 12. In some experiments, cartilage was preincubated with 1 mg/ml of 800-kd HA in DMEM for 48 hours. Thereafter, DMEM was discarded, the cartilage washed extensively with fresh DMEM, and then the cartilage explant was placed into another well of a culture plate in DMEM containing 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of HA. In other experiments, cartilage was placed in a 48-well plate (Corning) and preincubated with anti-CD44 or nonspecific control IgG for 24 hours before treatment with HA and/or IL-1␤. Cartilage explants and conditioned media were collected on day 2 or day 4 and stored at –80°C prior to analysis. DNA assay. After collection of conditioned media, cartilage explants were digested overnight at 56°C with 0.5 mg/ml of proteinase K in 50 mM Tris (pH 7.5). DNA content was measured in proteinase K digests of articular cartilage explants as described previously (24). Immunoblot analysis. Conditioned media were heated with sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer at 80°C for 20 minutes. Proteins were separated by SDS-PAGE under reducing conditions and then transferred onto nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany). Gel loading was standardized according to the DNA content of the cartilage explants. Membranes were blocked in phosphate buffered saline (PBS), pH 7.4, containing 5% nonfat dry milk, and incubated with the first antibody (1:1,000 in PBS) overnight at 4°C. After incubation with alkaline phosphatase–conjugated second antibody (1:2,000 in PBS) for 3 hours at room temperature, immunoreactive bands were visualized using BCIP and 518 nitroblue tetrazolium. The presence of HA in the conditioned media had no significant effect on the results of immunoblotting for MMP-1, MMP-3, or MMP-13 (data not shown). Protein band intensity was evaluated by densitometry using NIH Image Analysis software (online at http://rsb.info.nih.gov/ nih-image/). Immunolocalization. After preculture for 2 days, normal cartilage slices were stimulated for 48 hours with 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of HA in a 48-well culture plate in DMEM. The ionophore monensin (5 M; Sigma) was added to the cultures for the last 6 hours to prevent the secretion of newly synthesized proteins. Cartilage pieces were embedded in TissueTek OCT compound and frozen in liquid nitrogen. Cryostat sections at a thickness of 6 m were fixed with freshly prepared 4% paraformaldehyde in PBS (10 minutes at room temperature). The fixative was then removed by washing the sections in PBS (3 times for 5 minutes each). Sections were permeabilized for 10 minutes at room temperature with 0.15% Triton X-100 in PBS, washed as described above, and blocked with 1% bovine serum albumin (BSA) in PBS for 30 minutes at room temperature. The sections were then stained by indirect immunofluorescence using anti-human MMP-1, MMP-3, and MMP-13 (20 g/ml in PBS containing 1% BSA). Following extensive washing, the sections were incubated with FITC-conjugated anti-rabbit IgG (1:100 in PBS containing 1% BSA) and counterstained with propidium iodide (KPL, Gaithersburg, MD) at a 1:1,000 dilution in PBS for 3 minutes. Sections were mounted in Dako Glycergel mounting medium (Dako, Kyoto, Japan) on coverslides and then evaluated by confocal microscopy (Fluoview; Olympus, Tokyo, Japan). Evaluation of HA penetration into cartilage and HA binding to CD44 by fluorescence microscopy. After preculture for 2 days, articular cartilage slices were incubated for 48 hours at 37°C with or without 5-aminofluoresceinated 720-kd HA at 125 g/ml in a 48-well culture plate containing DMEM. In another set of experiments, cartilage explants were preincubated for 24 hours at 37°C with anti-CD44 antibody IM7 or control IgG (5 g/ml), followed by incubation for 48 hours at 37°C with 125 g/ml of 5-aminofluoresceinated HA in DMEM. After blocking with 1% BSA for 24 hours, articular cartilage slices were incubated with 5 g/ml of FITCconjugated antibody OS/37 or 5 g/ml of subclass-matched FITC-conjugated mouse IgG1 (R&D Systems) for 24 hours at 37°C to investigate the expression of CD44 on chondrocytes. Cartilage slices were recovered, sectioned with a cryostat (6 m), and slides were fixed with 4% paraformaldehyde in PBS for 20 minutes. After extensive washing with PBS, slides were counterstained with propidium iodide. All sections were mounted in Dako Glycergel on coverslides and subjected to confocal microscopy. Measurement of HA concentration in conditioned media. Cartilage slices were cultured for 4 days in the presence or absence of 2 ng/ml of IL-1␤. Each 1 ml of the conditioned media was chromatographed on PD-10 columns (Pharmacia, Uppsala, Sweden). The void volume (2.5 ml) was discarded, and the fraction that eluted with 3.5 ml of distilled water was collected. This fraction was concentrated at 0.2 ml with a Centrifugal Evaporator centrifuge model EC-57C (Sakuma Seisakusyo, Tokyo, Japan). Each 0.15 ml of the concentrated samples was treated with 2.5 turbidity-reducing units of Strep- JULOVI ET AL tomyces hyaluronidase (Seikagaku) in 200 l of 0.025M sodium acetate buffer (pH 6.0) at 37°C for 16 hours, and then the mixture was ultrafiltered using an Ultrafree C3GC system (molecular size cutoff 10,000; Japan Millipore, Tokyo, Japan). High-performance liquid chromatography (HPLC) of the unsaturated tetrasaccharide of HA (⌬tetra-HA) and the unsaturated hexasaccharide of HA (⌬hexa-HA) was performed according to the method described by Takazono and Tanaka (25) and Shinmei et al (26). The HPLC system used in this study was constructed from 2 pumps (model 880-PU; Japan Spectroscopic, Tokyo, Japan), an autosampling injector (model 851-AS; Japan Spectroscopic), a stainless steel column packed with polyamine-bound silica (YMC gel PA-120; YMC, Kyoto, Japan), a dry reaction bath (DB-3; Shimamura Instrument Company, Tokyo, Japan), a fluoro-monitor (model FP920; Japan Spectroscopic), and an integrator (model 807-IT; Japan Spectroscopic). The ⌬tetra-HA and ⌬hexa-HA in each sample were eluted with a gradient of 0–100 mM sodium sulfate for 45 minutes at a flow rate of 0.5 ml/minute. To the eluent from the column was added 100 mM sodium tetraborate buffer (pH 9.0) containing 1% 2-cyanoacetamide at a flow rate of 0.5 ml/minute. The mixture passed through polyetheretherketone tubing (0.5 mm inside diameter ⫻ 10 meters) set in a dry reaction bath that was thermostated at 137°C, and the effluent was monitored by the fluoro-monitor (excitation 331 nm, emission 383 nm). The area of each peak corresponding to ⌬tetra-HA and ⌬hexa-HA was calculated by the integrator and converted to an amount of hyaluronan against the area of standard ⌬tetra-HA and ⌬hexa-HA (Seikagaku). Statistical analysis. Comparisons between 2 groups were performed by Student’s t-test. P values less than 0.05 were considered significant. RESULTS Suppression of MMP production in IL-1␤– stimulated articular cartilage by HA. Incubation of normal human articular cartilage with 2 ng/ml of IL-1␤ under serum-free conditions for 4 days resulted in enhanced secretion of MMP-1, MMP-13, and MMP-3 into conditioned media, as determined by immunoblot analysis (Figure 1). Control cultures without IL-1␤ treatment secreted basal levels of MMP-1 and MMP-3 and barely detectable levels of MMP-13 into conditioned media. Levels of HA secreted into conditioned media from normal cartilage explant cultures in the presence and absence of 2 ng/ml of IL-1␤ on day 4 were 0.92 ⫾ 0.73 g/ml and 0.42 ⫾ 0.13 g/ml, respectively (mean ⫾ SD; n ⫽ 5). Thus, compared with the concentration in normal synovial fluid (2–4 mg/ml) (27) and with the amount of exogenous HA used in the present study (1 mg/ml), intrinsic HA levels were considered to be significantly low. We found that 800-kd HA at a concentration of ⱕ1 g/ml produced no effect on IL-1␤–stimulated MMPs (data not shown). In contrast, 1 mg/ml of HA HA INHIBITION OF IL-1␤–STIMULATED MMP PRODUCTION IN ARTICULAR CARTILAGE Figure 1. Inhibitory effects of hyaluronan (HA) on interleukin-1␤ (IL-1␤)–stimulated matrix metalloproteinase (MMP) production in normal human articular cartilage explant culture. Articular cartilage was incubated with or without 2 ng/ml of IL-1␤ in the presence or absence of 1.0 mg/ml of 800-kd HA for 4 or 12 days. Secreted levels of MMP-1, MMP-3, and MMP-13 in conditioned media during days 0–4 and 8–12 were detected by Western blotting using specific antibodies. The amount of sample applied was determined based on DNA content of the explant cartilage. Control cultures contained no additives. Molecular standards (in kd) are indicated at the right of the day 4 blots. (within the range of concentrations in normal synovial fluid) reduced the IL-1␤–induced production of MMPs (Figure 1). Incubation of cartilage with 1 mg/ml of HA alone resulted in no clear increase in MMP levels. Compared with IL-1␤–induced levels in the absence of HA (calculated as 100%), the levels of MMP-1, MMP-3, and MMP-13 in IL-1-␤–treated cultures in the presence of 1 mg/ml of HA were, respectively, 39.7 ⫾ 10.2% (P ⬍ 0.05), 57.7 ⫾ 10.4% (P ⬍ 0.05), and 15.9 ⫾ 8.9% (P ⬍ 0.05) (mean ⫾ SD; n ⫽ 10). After longer exposure of cartilage explants to 2 ng/ml of IL-1␤ with 1 mg/ml of HA from day 0 to day 12, 519 immunoblot analysis revealed that HA was still protective against IL-1␤–induced MMP elevation on day 12 (Figure 1). Compared with IL-1␤–induced levels in the absence of HA during days 8–12 (calculated as 100%), the levels of MMP-1, MMP-3, and MMP-13 from IL-1␤–treated cultures in the presence of 1 mg/ml of HA during days 8–12 were, respectively, 39.4 ⫾ 16.8% (P ⬍ 0.05), 65.8 ⫾ 2.4% (P ⬍ 0.05), and 28.3 ⫾ 20.3% (P ⬍ 0.05) (mean ⫾ SD; n ⫽ 4). The decreased levels of IL-1␤–induced MMPs in conditioned media in the presence of HA treatment could reflect decreased enzyme synthesis or decreased release of the enzymes from matrix-bound stores. To determine whether HA could suppress the IL-1␤– stimulated synthesis of MMPs 1, 3, and 13, articular cartilage was incubated with 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of 800-kd HA for 48 hours and the ionophore monensin was added for the final 6 hours. Since monensin caused the intracellular accumulation of newly synthesized MMPs, IL-1␤ treatment resulted in a marked increase in MMPs 1, 3, and 13 compared with cultures without monensin treatment. When articular cartilage was coincubated with 1 mg/ml of HA in the presence of IL-1␤, intracellular staining of MMPs was decreased after monensin treatment (data not shown). Thus, HA inhibited IL-1␤–stimulated synthesis of MMPs in articular chondrocytes, consistent with the results of the immunoblot studies (Figure 1). Penetration of HA into articular cartilage explants. We next attempted to clarify whether exogenously added HA could penetrate into articular cartilage explants. OA and normal articular cartilage slices were incubated with 5-aminofluoresceinated 720-kd HA for 48 hours, washed extensively, and then analyzed by fluorescence microscopy. Consistent with the findings of previous studies using degenerated cartilage produced by treatment with IL-1␤ (13), 5-aminofluoresceinated HA penetrated into the OA cartilage, and fluorescent signals were localized around chondrocytes (Figure 2A). As was seen in OA cartilage, 5-aminofluoresceinated HA penetrated into normal human cartilage slices and was localized around chondrocytes, with a thin layer of accumulation of HA at the articular surface and with little accumulation at the cut surfaces of cartilage slices (Figure 2B and insets). These findings are consistent with those of previous studies using normal human (12) and bovine (28) articular cartilage slices. Suppression of IL-1␤–induced MMP by single pretreatment with HA. Because HA was still associated with chondrocytes 48 hours after treatment, as shown in Figure 2, we examined the effects of a single HA 520 Figure 2. CD44-mediated association of hyaluronan (HA) with chondrocytes, as demonstrated by fluorescence microscopy. In osteoarthritic (A) and normal (B) cartilage samples incubated with 125 g/ml of 5-aminofluoresceinated 720-kd HA for 48 hours, there is penetration of HA into both cartilage slices, with localization around chondrocytes. In the normal cartilage sample, there is a thin layer of accumulated HA (left inset) at the articular surface (top), but little accumulation (right inset) at the cut surface of the cartilage (top). In normal cartilage samples incubated with fluorescein isothiocyanate (FITC)–conjugated anti-CD44 antibody OS/37 (C) and FITCconjugated nonspecific IgG (D) at 5 g/ml for 48 hours, there is intense fluorescence around chondrocytes in the sample incubated with anti-CD44 antibody, but sparse signals in the sample incubated with nonspecific IgG. In normal cartilage samples preincubated with anti-CD44 antibody IM7 (E) and nonspecific IgG (F) at 5 g/ml for 24 hours and then incubated with 125 g/ml of 5-aminofluoresceinated 720-kd HA for a further 48 hours, there is decreased signal in the sample incubated with the anti-CD44 antibody, but little effect on the sample incubated with nonspecific IgG. Results are representative of 3 separate experiments, all of which yielded similar results. Bars ⫽ 50 m in A and B and 100 m in C–F. pretreatment on IL-1␤–induced MMPs. After preincubation with 1 mg/ml of 800-kd HA for 48 hours followed by extensive washing, articular cartilage was relocated to another well of the culture plate and then incubated with 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of HA for 4 days. Immunoblot analysis showed that this single pretreatment with HA was also effective in suppressing the induction of MMPs by IL-1␤ (Figure 3). Compared with JULOVI ET AL IL-1␤–induced levels (calculated as 100%), the levels of MMP-1, MMP-3, and MMP-13 in IL-1␤–treated cultures in the absence of HA but with HA pretreatment were, respectively, 28 ⫾ 16.0% (P ⬍ 0.05), 50.2 ⫾ 8.5% (P ⬍ 0.05), and 31.9 ⫾ 15.3% (P ⬍ 0.05) (mean ⫾ SD; n ⫽ 3). The simultaneous addition of HA with IL-1␤ caused similar inhibitory effects on the IL-1␤–induced MMP levels. Together with the observation that HA localized around chondrocytes with little accumulation at the cartilage surface, it was unlikely that the protective effect of HA on the action of IL-1␤ was the result of either the formation of nonpenetrating HA–IL-1␤ complexes through their interaction in the medium or blocking of IL-1␤ penetration into cartilage tissue at the cartilage surface. Comparison of the effects of HA on IL-1␤– stimulated MMPs in normal and OA cartilage. Because HA is used in the clinical treatment of OA, the effects of HA on IL-1␤–stimulated MMPs were compared in normal and OA cartilage samples. After preincubation with 1 mg/ml of 800-kd HA for 48 hours, normal and OA cartilage samples were stimulated with 2 ng/ml of IL-1␤ for 48 hours in the presence and absence of HA. Samples from normal and OA cartilage explant cultures were then transferred onto the same membrane. Immunoblot analysis showed that MMP-1 and MMP-3 se- Figure 3. Effects of a single pretreatment with HA on IL-1␤– stimulated MMP production. Normal human cartilage was preincubated with or without 1 mg/ml of 800-kd HA for 48 hours and then with 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of HA for 4 days. Secreted levels of MMP-1, MMP-3, and MMP-13 in conditioned media were detected by Western blotting using specific antibodies. The amount of sample applied was determined based on DNA content of the explant cartilage. Control cultures contained no additives. A single pretreatment with HA was effective in suppressing the MMP induction by IL-1␤, as shown in the far right lane. HA treatment alone had no effect on MMP levels, as shown in Figure 1. See Figure 1 for definitions. HA INHIBITION OF IL-1␤–STIMULATED MMP PRODUCTION IN ARTICULAR CARTILAGE 521 Association of HA with chondrocytes via CD44 in cartilage explants. Chondrocytes in normal (Figure 2C) and OA (data not shown) articular cartilage explants expressed CD44. When normal articular cartilage was incubated with 125 g/ml of 5-aminofluoresceinated HA for 48 hours, intense fluorescent signals were seen around chondrocytes (Figure 2B). When articular cartilage was incubated with 125 g/ml of 5-aminofluoresceinated HA for 48 hours after 24 hours of preincubation with 5 g/ml of anti-CD44 antibody IM7, decreased signals were detected (Figure 2E). In contrast, preincubation with nonspecific IgG caused Figure 4. Comparison of the effects of HA on IL-1␤–stimulated MMP production in normal and osteoarthritic (OA) cartilage. A, Cartilage was preincubated with or without 1 mg/ml of 800-kd HA for 48 hours and then stimulated with 2 ng/ml of IL-1␤ in the presence or absence of HA for 48 hours. Levels of secreted MMP-1 and MMP-3 in conditioned media were detected by Western blotting using specific antibodies. The amount of sample applied was determined based on DNA content of the explant cartilage. Control cultures contained no additives. HA treatment alone had no effect on MMPs, as shown in Figure 1. B, Densitometric analysis of 3 normal and 3 OA cartilage samples. The band intensity of protein from IL-1␤–stimulated normal cartilage is defined as 100%. Values are the mean and SD of 3 separate experiments. ⴱ ⫽ P ⬍ 0.05 versus IL-1␤–stimulated normal cartilage; ⴱⴱ ⫽ P ⬍ 0.05 versus IL-1␤–stimulated OA cartilage; # ⫽ P ⬍ 0.05 versus IL-1␤–stimulated normal cartilage, by Student’s t-test. See Figure 1 for other definitions. creted from OA cartilage in response to IL-1␤ showed stronger bands than those secreted from normal cartilage (Figure 4A). Densitometric analysis revealed that pretreatment with HA suppressed IL-1␤–stimulated secretion of MMP-1 and MMP-3 to a similar extent in normal and OA cartilage (Figure 4B). MMP-13 stimulated by IL-1␤ for 48 hours was below the level of detection in both OA and normal cartilage (data not shown). Figure 5. Comparison of the effects of the anti-CD44 antibody IM7 on the action of HA in IL-1␤–stimulated MMPs in normal cartilage. A, After pretreatment with 25 g/ml of IM7 for 24 hours, cartilage was incubated with or without 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of 800-kd HA for 4 days. Levels of secreted MMP-1, MMP-3, and MMP-13 in conditioned media were detected by immunoblotting. The amount of sample applied was determined based on DNA content of the explant cartilage. HA treatment alone had no effect on MMPs, as shown in Figure 1. B, Densitometric analysis of 4 normal cartilage samples. The band intensity of protein from IL-1␤– stimulated normal cartilage is defined as 100%. Values are the mean and SD of 3 separate experiments. ⴱ ⫽ P ⬍ 0.05 versus IL-1␤– stimulated normal cartilage; # ⫽ P ⬍ 0.05, by Student’s t-test. See Figure 1 for definitions. 522 JULOVI ET AL IM7 was required to block MMP-1 and MMP-13 induced by IL-1␤ (Figure 6). Compared with normal cartilage, higher levels (100 g/ml) of IM7 were required to block the action of IL-1␤ on MMPs in OA cartilage (Figure 7A). Pretreatment with IM7 at 100 g/ml for 24 hours reversed the inhibitory effect of 1 mg/ml of HA on the IL-1␤– Figure 6. Dose-dependent effect of the anti-CD44 antibody IM7 on IL-1␤–stimulated MMPs in normal cartilage. After pretreatment with 1, 5, and 25 g/ml of IM7 for 24 hours, cartilage was incubated with 2 ng/ml of IL-1␤ for 4 days. Secreted levels of MMP-1, MMP-3, and MMP-13 in conditioned media were detected by immunoblotting. The amount of sample applied was determined based on DNA content of the explant cartilage. See Figure 1 for definitions. little effect on the binding of HA to chondrocytes (Figure 2F). Thus, HA binding to chondrocytes in human articular cartilage explant culture involved CD44. Role of HA ligation with CD44 in the inhibition of IL-1␤–stimulated MMP production in articular cartilage. In order to investigate whether the mechanism of HA is biologically mediated by CD44 on chondrocytes, the anti-CD44 antibody IM7 was used to block HA binding to chondrocytes in normal and OA articular cartilage explant cultures. Pretreatment with 25 g/ml of IM7 for 24 hours partially blocked the inhibitory effect of 1 mg/ml of HA on IL-1␤–stimulated secretion of MMPs 1, 3, and 13 (Figure 5A). Densitometric analysis confirmed that IM7 significantly reversed the effects of HA on MMPs 1, 3, and 13 induced by IL-1␤ (Figure 5B). Compared with 1 mg/ml of HA, 25 g/ml of IM7 itself caused weaker, but significant, suppression of IL-1␤– stimulated secretion of MMP-1 and MMP-13 in normal cartilage (Figure 5B). Preincubation with nonspecific IgG caused no significant effect on IL-1␤–stimulated secretion of MMPs (data not shown). When normal cartilage was incubated with 2 ng/ml of IL-1␤ after preincubation with IM7 at increasing concentrations (1, 5, and 25 g/ml) for 24 hours, we found that 25 g/ml of Figure 7. Effects of the anti-CD44 antibody IM7 on the action of HA in IL-1␤–stimulated MMPs in osteoarthritic (OA) cartilage. A, Cartilage was preincubated with or without 25 or 100 g/ml of IM7 for 24 hours and then incubated with 2 ng/ml of IL-1␤ for 4 days. B, After pretreatment with or without 100 g/ml of IM7 for 24 hours, cartilage was incubated with 2 ng/ml of IL-1␤ in the presence or absence of 1 mg/ml of 800-kd HA for 4 days. Secreted levels of MMP-1, MMP-3, and MMP-13 in conditioned media were detected by immunoblotting. The amount of sample applied was determined based on DNA content of the explant cartilage. Control cultures contained no additives. HA treatment alone had no effect on MMPs, as shown in Figure 1. C, Densitometric analysis of 4 OA cartilage samples. The band intensity of protein from IL-1␤–stimulated OA cartilage is defined as 100%. Values are the mean and SD of 3 separate experiments. ⴱ ⫽ P ⬍ 0.05 versus IL-1␤–stimulated OA cartilage; # ⫽ P ⬍ 0.05, by Student’s t-test. See Figure 1 for other definitions. HA INHIBITION OF IL-1␤–STIMULATED MMP PRODUCTION IN ARTICULAR CARTILAGE stimulated secretion of MMPs 1, 3, and 13 in OA cartilage (Figure 7B). Densitometric analysis showed that preincubation with IM7 significantly blocked the inhibitory effects of HA against IL-1␤–induced MMPs (Figure 7C). DISCUSSION IL-1 is considered to play an important role in the pathogenesis of arthritis, including OA, mainly because it can induce the resorption of proteoglycan (29) and type II collagen (30). In OA cartilage, proteoglycan loss that may involve MMP-3 results in a reduction of cartilage stiffness (31,32), whereas degradation and loss of type II collagen that involves collagenase (MMP-1 and MMP-13) result in an irreversible loss of tensile properties and structural integrity (32). Although HA is used in the treatment of OA, whether HA treatment of OA joints can alter the rate of disease progression is still a subject of controversy. HA enhances proteoglycan synthesis in articular cartilage upon treatment with fibronectin fragment (12) and IL-1 (33). Such anabolic actions by HA may therefore suppress cartilage damage by catabolic stimulators. In addition, HA has been shown to block proteoglycan release from cartilage (34). This study is the first to clearly demonstrate that 1 mg/ml of HA, which is within the range of physiologic concentrations in synovial fluids (2–4 mg/ml) (27), was able to block IL-1␤–stimulated production of MMP-1, MMP-3, and MMP-13 in human OA articular cartilage as well as normal cartilage. Therefore, HA treatment may prevent the IL-1␤–induced breakdown of type II collagen and proteoglycan in OA cartilage by blocking collagenases (MMP-1 and MMP-13) and stromelysin 1 (MMP-3), respectively, leading to the deceleration of OA progression. Since in vitro responses may be different in cartilage specimens obtained from different joints (12,35), further studies using articular cartilage from hip or knee joints may be required to confirm the present findings. The mechanism of action of HA on chondrocytes in articular cartilage is not entirely clear. Exogenously added HA has been demonstrated to accumulate at the intact cartilage surface without penetrating into the cartilage (13). Once the articular cartilage is degraded, however, HA can penetrate the cartilage tissue and localize in the pericellular matrix around chondrocytes, as shown in the present study (Figure 2). Although recent studies suggest that HA does not decrease the penetration of IL-1␤ into cartilage tissue (28), the possibility has not been completely excluded that when 523 injected intraarticularly, HA may act as a barrier against catabolic substances, such as IL-1␤, at the surface of the OA cartilage. At present, there is no clear explanation of how high molecular weight HA penetrates articular cartilage. HA oligosaccharides could stimulate not only proteoglycan degradation, with exhibition of the NITEGE epitope and increased gelatinase activity, but also proteoglycan synthesis in chondrocytes via CD44, whereas there is no available evidence that confirms the presence of such HA oligosaccharides within the cartilage tissue or the synovial fluid (35). Although the lack of increase in the production of MMPs with HA treatment alone (Figure 1) may contradict the generation of HA oligosaccharides, it is possible that degraded HA may penetrate cartilage. The involvement of degraded HA in the present studies remains to be determined. IM7, the monoclonal anti-CD44 antibody, blocked the binding of HA to chondrocytes (Figure 2E), indicating that HA binds to chondrocytes via CD44 in articular cartilage explants. In order to investigate whether the mechanism of HA action is biologically mediated by CD44 on chondrocytes, we performed HA-binding inhibition studies using the anti-CD44 antibody. Treatment with IM7 resulted in a significant reduction of the inhibitory effects of HA on MMP production in IL-1␤–stimulated cartilage (Figures 5 and 7). The interaction between HA and CD44 has been shown to reduce anti-Fas–induced apoptosis of chondrocytes (36), which is further evidence of the direct involvement of CD44 in the mechanism of HA action. Antisense inhibition of chondrocyte CD44 expression results in proteoglycan degradation with NITEGE expression in cartilage (37), indicating that CD44 may be a key player in the maintenance of the cartilage structure. Of interest, monovalent ligation of the anti-CD44 antibody IM7 with CD44 also caused significant inhibitory effects on IL-1␤–stimulated MMPs (Figures 5, 6, and 7), which reflects a minor role of the barrier effect of HA through pericellular accumulation around chondrocytes. Higher doses of IM7 were required to suppress IL-1␤– stimulated MMPs in OA cartilage (Figure 7) compared with normal cartilage (Figure 6). Because CD44 is highly expressed in inflammatory conditions, up-regulation of CD44 in OA cartilage may be involved in this. It has been demonstrated that CD44 functions as a signaling receptor in various types of cells. Cell stimulation by anti-CD44 antibodies or natural CD44 ligands transmits the signal into the cells, leading to the activation of T cells and the release of cytokines or chemokines from monocytes and RA synovial fibroblasts (38,39). In addition, there is evidence that the IM7 524 JULOVI ET AL antibody induces some cellular responses via CD44 (40). Thus, anti-CD44 treatment using the IM7 antibody and the natural ligand HA could activate some intracellular signaling pathways that block the action of IL-1␤ on MMPs; these intracellular signaling pathways remain to be determined. Alternatively, since ligands for CD44 include collagens, fibronectin, laminin, chondroitin sulfate, and osteopontin as well as HA (41), HA interference in the ligation of CD44 with other ligands could induce the suppression of IL-1␤–induced MMPs in chondrocytes. Because chondrocytes express other HA receptors, such as ICAM-1 (21), which is up-regulated by proinflammatory cytokines in chondrocytes (22), such receptors may also contribute to the action of HA in articular cartilage. Overall, the present study highlights the specific role of CD44 on chondrocytes in the inhibitory action of HA on IL-1␤–induced MMP production. HA has been shown to be a potent inhibitor of proteoglycan depletion in rabbit articular cartilage (34). In addition, HA can reduce anti-Fas–induced apoptosis of OA chondrocytes (36). Indeed, arthroscopic evaluations of articular cartilage have demonstrated a potential structure-modifying effect of HA in patients with knee OA (42,43). Such findings indicate that HA may slow cartilage breakdown and restore chondrocyte density in OA cartilage. Therefore, the forms of HA used clinically in the intraarticular treatment of OA may provide not only viscoelastic supplementation, but also chondroprotective effects on the OA cartilage by ligation of HA with CD44, its principal receptor, on chondrocytes. REFERENCES 1. Poole AR. Cartilage in health and disease. In: Koopman WJ, editor. Arthritis and allied conditions: a textbook of rheumatology. 14th ed. Baltimore: Lippincott Williams & Wilkins; 2001. p. 226–84. 2. Woessner JF Jr, Gunja-Smith Z. Role of metalloproteinases in human osteoarthritis. J Rheumatol 1991;18:99–101. 3. Shlopov BV, Lie W-R, Mainardi CL, Cole AA, Chubinskaya S, Hasty KA. Osteoarthritic lesions: involvement of three different collagenases. Arthritis Rheum 1997;40:2065–74. 4. Freemont AJ, Hampson V, Tilman R, Goupille P, Taiwo Y, Hoyland JA. Gene expression of matrix metalloproteinase 1, 3, and 9 by chondrocytes in osteoarthritic human knee articular cartilage is zone and grade specific. Ann Rheum Dis 1997;56: 542–9. 5. Shlopov BV, Gumanovskaya ML, Hasty KA. Autocrine regulation of collagenase 3 (matrix metalloproteinase 13) during osteoarthritis. Arthritis Rheum 2000;43:195–205. 6. Freemont AJ, Byers RJ, Taiwo YO, Hoyland JA. In situ zymographic localisation of type II collagen degrading activity in osteoarthritic human cartilage. Ann Rheum Dis 1999;58:357–65. 7. Billinghurst RC, Dahlberg L, Ionescu M, Reiner A, Bourne R, Rorabeck C, et al. Enhanced cleavage of type II collagen by 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. collagenases in osteoarthritic cartilage. J Clin Invest 1997;99: 1534–45. Tetlow LC, Adlam DJ, Woolley DE. Matrix metalloproteinase and proinflammatory cytokine production by chondrocytes of human osteoarthritic cartilage: associations with degenerative changes. Arthritis Rheum 2001;44:585–94. Peyron JG. Intraarticular hyaluronan injections in the treatment of osteoarthritis: state-of-the-art review. J Rheumatol 1993;20 Suppl 39:10–5. Homandberg GA, Meyers R, Xie DL. Fibronectin fragments cause chondrolysis of bovine articular cartilage slices in culture. J Biol Chem 1992;267:3597–604. Yasuda T, Poole AR. A fibronectin fragment induces type II collagen degradation by collagenase through an interleukin1–mediated pathway. Arthritis Rheum 2002;46:138–48. Kang Y, Eger W, Koepp H, Williams J, Kuettner K, Homandberg GA. Hyaluronan suppresses fibronectin fragment-mediated damage to human cartilage explant cultures by enhancing proteoglycan synthesis. J Orthop Res 1999;17:858–69. Fukuda K, Dan H, Takayama M, Kumano F, Saitoh M, Tanaka S. Hyaluronic acid increases proteoglycan synthesis in bovine articular cartilage in the presence of interleukin-1. J Pharmacol Exp Ther 1996;277:1672–5. Salter DM, Godolphin JL, Gourlay MS, Lawson MF, Hughes DE, Dunne E. Analysis of human articular chondrocyte CD44 isoform expression and function in health and disease. J Pathol 1996;179: 396–402. Aruffo A, Stamenkovic I, Melnick M, Underhill CB, Seed B. CD44 is the principal cell surface receptor for hyaluronate. Cell 1990;61: 1303–13. Toba T, Mizusawa N, Tajima G, Horiuchu S. Upregulation of CD44 mRNA expression by interleukin-1␤ in cultured rabbit articular chondrocytes. J Bone Miner Metab 1997;15:84–93. Nishida Y, D’Souza AL, Thonar EJ-MA, Knudson W. Stimulation of hyaluronan metabolism by interleukin-1␣ in human articular cartilage. Arthritis Rheum 2000;43:1315–26. Mikecz K, Dennis K, Shi M, Kim JH. Modulation of hyaluronan receptor (CD44) function in vivo in a murine model of rheumatoid arthritis. Arthritis Rheum 1999;42:659–68. Neidhart M, Gay RE, Gay S. Anti–interleukin-1 and anti-CD44 interventions producing significant inhibition of cartilage destruction in an in vitro model of cartilage invasion by rheumatoid arthritis synovial fibroblasts. Arthritis Rheum 2000;43:1719–28. McCourt PA, Ek B, Forsberg N, Gustafson S. Intercellular adhesion molecule-1 is a cell surface receptor for hyaluronan. J Biol Chem 1994;269:30081–4. Davies ME, Dingle JT, Pigott R, Power C, Sharma H. Expression of intercellular adhesion molecule 1 (ICAM-1) on human articular cartilage chondrocytes. Connect Tissue Res 1991;26:207–16. Kienzle G, von Kempis J. Vascular cell adhesion molecule 1 (CD106) on primary human articular chondrocytes: functional regulation by cytokines and comparison with intercellular adhesion molecule 1 (CD54) and very late activation antigen 2. Arthritis Rheum 1998;41:1296–305. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039–49. Labarca C, Paigen K. A simple, rapid, and sensitive DNA assay procedure. Anal Biochem 1980;102:344–52. Takazono I, Tanaka Y. Quantitative analysis of hyaluronic acid by high-performance liquid chromatography of Streptomyces hyaluronidase digests. J Chromatogr 1984;288:167–76. Shinmei M, Miyauchi S, Machida A, Miyazaki K. Quantitation of chondroitin 4-sulfate and chondroitin 6-sulfate in pathologic joint fluid. Arthritis Rheum 1992;35:1304–8. Dahl LB, Dahl IM, Laurent AE, Granath K. Concentration and HA INHIBITION OF IL-1␤–STIMULATED MMP PRODUCTION IN ARTICULAR CARTILAGE 28. 29. 30. 31. 32. 33. 34. molecular weight of sodium hyaluronate in synovial fluid from patients with rheumatoid arthritis and other arthropathies. Ann Rheum Dis 1985;44:817–22. Homandberg GA, Ummadi V, Kang H. High molecular weight hyaluronan promotes repair of IL-1␤-damaged cartilage explants from both young and old bovines. Osteoarthritis Cartilage 2003; 11:177–86. Saklatvala J, Pilsworth LMC, Sarsfield ST, Gavrilovic J, Heath JK. Pig catabolin is a form of interleukin-1: cartilage and bone resorb, fibroblasts make prostaglandin and collagenase, and thymocyte proliferation is augmented in response to one protein. Biochem J 1984;224:461–6. Billinghurst RC, Wu W, Ionescu M, Reiner A, Dahlberg L, Chen J, et al. Comparison of the degradation of type II collagen and proteoglycan in nasal and articular cartilages induced by interleukin-1 and the selective inhibition of type II collagen cleavage by collagenase. Arthritis Rheum 2000;43:664–72. Bonassar LJ, Frank EH, Murray JC, Paguio CG, Moore VL, Lark MW, et al. Changes in cartilage composition and physical properties due to stromelysin degradation. Arthritis Rheum 1995;38: 173–83. Kempson GE. Mechanical properties of articular cartilage. In: Freeman MAR, editor. Adult articular cartilage. Tunbridge Wells: Pitman Medical; 1979, p. 333–414. Yasui T, Akatsuka M, Tobetto K, Umemoto J, Ando T, Yamashita K, et al. Effects of hyaluronan on the production of stromelysin and tissue inhibitor of metalloproteinase-1 in bovine articular chondrocytes. Biomedical Res 1992;13:343–8. Shimazu A, Jikko A, Iwamoto M, Koike T, Yan W, Okada Y, et al. Effects of hyaluronic acid on the release of proteoglycan from the cell matrix in rabbit chondrocyte cultures in the presence and absence of cytokines. Arthritis Rheum 1993;36:247–53. 525 35. Knudson W, Casey B, Nishida Y, Eger W, Kuettner KE, Knudson CB. Hyaluronan oligosaccharides perturb cartilage matrix homeostasis and induce chondrocytic chondrolysis. Arthritis Rheum 2000;43:1165–74. 36. Lisignoli G, Grassi F, Zini N, Toneguzzi S, Piacentini A, Guidolin D, et al. Anti-Fas–induced apoptosis in chondrocytes reduced by hyaluronan: evidence for CD44 and CD54 (intercellular adhesion molecule 1) involvement. Arthritis Rheum 2001;44:1800–7. 37. Chow G, Nietfeld JJ, Knudson CB, Knudson W. Antisense inhibition of chondrocyte CD44 expression leading to cartilage chondrolysis. Arthritis Rheum 1998;41:1411–9. 38. McKee CM, Penno MB, Cowman M, Burdick MD, Strieter RM, Bao C, et al. Hyaluronan fragments induce chemokine gene expression in alveolar macrophages. J Clin Invest 1996;98: 2403–13. 39. Fujii K, Tanaka Y, Hübscher S, Saito K, Ota T, Eto S. Crosslinking of CD44 on rheumatoid synovial cells augments interleukin 6 production. Lab Invest 1999;79:1439–46. 40. Henke C, Bitterman P, Roongta U, Ingbar D, Polunovsky V. Induction of fibroblast apoptosis by anti-CD44 antibody: implications for the treatment of fibroproliferative lung disease. Am J Pathol 1996;149:1639–50. 41. Naor D, Nedvetzki S. CD44 in rheumatoid arthritis. Arthritis Res Ther 2003;5:105–15. 42. Listrat V, Ayral X, Patarnello F, Bonvarlet JP, Simonnet J, Amor B, et al. Arthroscopic evaluation of potential structure modifying activity of hyaluronan (Hyalgan) in osteoarthritis of the knee. Osteoarthritis Cartilage 1997;5:153–60. 43. Frizziero L, Govoni E, Bacchini P. Intra-articular hyaluronic acid in the treatment of osteoarthritis of the knee: clinical and morphological study. Clin Exp Rheumatol 1998;16:441–9.