JOURNAL TRANSCRIPT

---

Osteoarthritis and Cartilage (2002) 10, 799–807 © 2002 OsteoArthritis Research Society International. Published by Elsevier Science Ltd. All rights reserved. doi:10.1053/joca.2002.0829, available online at http://www.idealibrary.com on

1063–4584/02/$35.00/0

International Cartilage Repair Society

Contribution of interleukin 17 to human cartilage degradation and synovial inflammation in osteoarthritis M. C. Honorati*†, M. Bovara*, L. Cattini*, A. Piacentini* and A. Facchini*† *Laboratorio di Immunologia e Genetica, Istituti Ortopedici Rizzoli, Bologna, Italy †Dipartimento di Medicina Interna e Gastroenterologia, Universita` degli Studi di Bologna, Bologna, Italy Summary Objective: To compare the effect of interleukin (IL)-17, IL-1
and TNF- on chemokine production by human chondrocytes and synovial fibroblasts isolated from patients with osteoarthritis (OA). The expression of IL-1
mRNA by OA chondrocytes was also assessed, as well as the presence and expression of IL-17 receptor (IL-17R) in OA chondrocytes and synovial fibroblasts after stimulation with IL-17, IL-1
and TNF-. Design: Synovial fibroblasts and chondrocytes isolated from patients with OA were stimulated in vitro with IL-17, IL-1
or TNF-. Supernatants were collected and immunoassayed for the presence of IL-8, GRO- (CXC chemokines) and MCP-1, RANTES (CC chemokines). The cells were used to detect the presence of IL-17R and the expression of IL-17R mRNA. Stimulated chondrocytes were also used to detect IL-1
production and mRNA expression. Results: IL-17 upregulated the release of IL-8 and GRO- both by synovial fibroblasts and chondrocytes, and the release of MCP-1 only by chondrocytes. IL-17 was a weaker stimulator than IL-1
and TNF-, except for GRO- release which was maximally upregulated by IL-1
, less by IL-17 and minimally by TNF-. When compared to IL-1
, IL-17 was more active on chondrocytes than on fibroblasts. In chondrocytes the expression of IL-1
mRNA was enhanced by IL-17 and TNF-, with a maximum level reached by IL-1
. IL-17 and TNF- stimulated IL-1
release in few subjects. Neither IL-17, IL-1
nor TNF- modulated the presence of IL-17R and the expression of IL-17R mRNA. Conclusions: These data suggest that IL-17 could contribute to cartilage breakdown and synovial infiltration in OA by inducing both the release of chemokines by chondrocytes and synovial fibroblasts and, in a less extent, the synthesis of IL-1
by chondrocytes. © 2002 OsteoArthritis Research Society International. Published by Elsevier Science Ltd. All rights reserved. Key words: IL-17, Osteoarthritis, Cartilage destruction, Chemokines.

causing inflammation. Fibroblast-like synoviocytes B (synovial fibroblasts) and chondrocytes can also secrete chemokines9–13. Chemokines can be classified on the basis of their primary structure. Two main chemokine families have been classified as CXC and CC according to the arrangement of the first two cysteine residues which are separated by one amino acid and adjacent, respectively14. Their chemoattractant effect on the target cells is known to depend on the difference in structure15–17. It has been recently demonstrated that chemokines induced by IL-1
and TNF- in chondrocytes can stimulate release of soluble factors just in chondrocytes, such as inducible nitric oxide synthase (iNOS) and matrix metalloproteases (MMPs), which are both responsible for cartilage degradation7,8,18. Human IL-17 is a 20–30 kDa glycosylated homodimeric polypeptide produced by activated T lymphocytes19–21. This cytokine shows properties similar to those of IL-1
and TNF-. They induce in vitro the synthesis of nitric oxide22–24 and metalloproteases23,25 by chondrocytes and the production of IL-6 and IL-8 by fibroblasts21,26. IL-17 has therefore been defined as a proinflammatory cytokine. Such evidence indicates that IL-17 is a cytokine able to induce the release of other cytokines, which can, in turn, contribute to tissue damage. Detectable levels of IL-17 have been reported in the synovial fluid from patients with rheumatic diseases27, although only a few cells producing IL-17 have been detected by immunohistochemistry and only in the synovial tissue from rheumatoid arthritis (RA)

Introduction Osteoarthritis (OA) is a degenerative articular disease of unknown origin. The available evidence suggests that alterations in cartilage are the events leading to the onset of this pathology, and are followed by synovial and subchondral bone damage1. Moreover, OA is often associated with an inflamed synovial membrane with evident vascularization, infiltration of T lymphocytes and mononuclear cells, and proliferation of synovial cells2–5. Among the several features characterizing the onset and progression of this pathology, two appear to be relevant, i.e. the disappearance of a balanced cytokine network, including proinflammatory (i.e. interleukin (IL)-1
and tumour necrosis factor (TNF)-) and antinflammatory (i.e. IL-4, IL-10, IL-1 receptor (IL-1R) antagonist) soluble factors at the articular tissue level6, and the changes in the expression of chemokine receptors in chondrocytes7,8. Among the soluble factors released in OA, chemokines have been described as important mediators at articular level, which attract and activate T cells, monocytes and granulocytes Received 19 October 2001; revision accepted 20 June 2002. Grants: MURST-Ricerca Fondamentale Orientata and Ricerca Corrente I.O.R. Address correspondence and reprint requests to: Maria Cristina Honorati, Laboratorio di Immunologia e Genetica, I.O.R., Via di Barbiano 1/10, 40136 Bologna, Italy. Tel.: +39 051 6366803; Fax: +39 051 6366807; E-mail: [email protected]

799

800 patients28,29. Conversely, we have recently reported the presence of higher amounts of IL-17R in cartilage from OA patients than in that from RA patients29. The role played by IL-17 in cartilage degradation and synovial inflammation in OA is still unclear. Further investigation is needed to assess whether it induces the release of degrading and proinflammatory soluble factors at articular level, such as CXC and CC chemokines and IL-1
which is an important degrading factor in OA6–8. In addition, whether IL-17R expression in chondrocytes and synovial fibroblats can also be stimulated by soluble factors acting mainly in OA needs clarification, too. In this study we evaluated the ability of IL-17 to induce the production of chemokines and regulate the presence of IL-17R on synovial fibroblasts and chondrocytes. Additionally, its ability to regulate the expression of IL-1
mRNA by chondrocytes alone was also assessed. Finally, the ability of IL-17 was compared with that of IL-1
and TNF-.

Materials and methods TISSUE SAMPLES AND PATIENTS

The study was performed on patients with OA undergoing knee or, in some cases, hip replacement surgery. Synovial samples were obtained from 12 patients (mean age 68 years, range 51 to 80 years) (11 knee and one hip) and cartilage samples from 19 patients (mean age 67 years, range 39 to 80 years) (16 knee and three hip). Diagnosis of OA was based on clinical, radiological and laboratory parameters30. All patients involved in the study showed grade III–IV radiological features according to the Kellgren and Lawrence grading system31. The study was approved by the ethical committee of Istituti Ortopedici Rizzoli and informed consent was obtained from the patients. ISOLATION OF SYNOVIAL FIBROBLASTS

Specimens of synovial tissue were sometimes vascularized and generally hypertrophic. Vascularized areas were avoided and specimens were cut into small pieces (no more than 1 mm2) and washed in phosphate buffered saline (PBS) (Gibco BRL, Grand Island, NY, U.S.A.). They were then seeded in 25 cm2 culture flasks in Optimem-1 (Gibco BRL) supplemented with 200 g/ml gentamycin (Flow Laboratories, Biaggio, Switzerland), 15% heat inactivated fetal calf serum (FCS) (Gibco BRL) (complete Optimem) at 37°C in 5% CO2 humidified atmosphere. After 48 h, the culture medium was changed. After another 5 days, when adherent cells were evident, synovium specimens were removed and the culture medium was changed. Adherent cells were split at confluence by 0.05% trypsin/ 0.5 mM EDTA (Sigma, St Louis, MO, U.S.A.) and passaged into a 75 cm2 flask. Fibroblasts were used between passages 3 and 7. After brief trypsinization and washing with PBS, flow cytometry analysis was performed using anti-CD3, CD4, CD8, CD14, CD68 (Becton Dickinson, Mountain View, CA, U.S.A.) and anti-human fibroblast (Dako, Glostrup, Denmark) mouse monoclonal antibodies (MoAb), in order to assess the homogeneity of synovial fibroblast population. It was then followed by incubation with rabbit antimouse immunoglobulins conjugated to phycoerythrin (PE) (Becton Dickinson). Analysis was performed with a FACStar Plus (Becton Dickinson, St Diego, CA, U.S.A.).

M. C. Honorati et al.: IL-17 in osteoarthritis Synovial fibroblasts appeared as a homogeneous population and showed positive staining only for anti-human fibroblast MoAb.

CHONDROCYTE ISOLATION

Most of the articular cartilage tissue was removed from the medial condyles or tibial plateau. As expected, the cartilage samples were macroscopically different and with erosion areas. During tissue sampling regions with signs of erosion were avoided. The remaining fragments of cartilage tissue from each patient were pooled together before digestion. Chondrocytes were isolated from the cartilage by sequential digestion of enzymes from Sigma (0.1% hyaluronidase, 0.5% pronase and 0.2% collagenase for 30 min, 1 h and 1 h, respectively) at 37°C in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco BRL) with 25 mM HEPES (Sigma), 100 U/ml penicillin (Biological Industries, Israel), 100 g/ml streptomycin (Biological Industries), 50 g/ml gentamycin (Flow Laboratories) and 2.5 g/ml amphotericin B (Biological Industries). After filtration through 100 m and 70 m nylon meshes, chondrocytes were washed and seeded at a density of 2×105 cells/cm2 in DMEM supplemented with 4 mM glutamine (Sigma), 200 g/ml gentamycin and 10% foetal calf serum (FCS) (complete DMEM) at 37°C in 5% CO2 humidified atmosphere. All experiments were performed with chondrocytes, either immediately after isolation or at 2 weeks, and before the first splitting.

CELL STIMULATION FOR CHEMOKINE RELEASE

Cells were cultured in a 96-well flat bottom microplate in 0.2 ml of appropriate complete culture medium. Fibroblasts (2.8×104 cells/cm2) or chondrocytes (2×105 cells/cm2) were cultured overnight. Then, 0.1 ml was substituted with 0.1 ml of complete medium with or without stimuli: 50 ng/ml recombinant human (rh) IL-17 (IL-17) (R&D Systems, Minneapolis, MN, U.S.A.), 100 U/ml rhIL-1
(IL-1
; specific activity 5×107 U/mg) (Boehringer, Mannheim, Germany), 100 U/ml rhTNF- (TNF- specific activity 1×108 U/mg) (Boehringer) (final concentration). Previous experiments had shown that these concentrations are optimal for detection of IL-8 release, being the highest of the linear zone of a dose-effect curve. After an incubation of 72 h, supernatants were collected and maintained at −80°C until their use.

CELL STIMULATION FOR RT-PCR

Previous experiments performed at different incubation times (0, 3, 6, 9 and 24 h) had shown that 6 h was the optimal time for IL-8 mRNA detection. Cells from each patient were therefore cultured in appropriate medium alone or with stimuli, for 6 h in six-well tissue culture plates. Fibroblasts (3×104 cells/cm2) were cultured overnight in 4 ml of complete Optimem. Subsequently, 2 ml were substituted with 2 ml of complete Optimem, either alone or with the following stimuli: 50 ng/ml IL-17, 100 U/ml IL-1
, and 100 U/ml TNF (final concentration). Isolated chondrocytes (2×104 cells/cm2) were cultured in complete DMEM. When the wells were full, that is to say after about 15 days of incubation, the culture medium was substituted with complete DMEM, either with or without stimuli. At the end of the

Osteoarthritis and Cartilage, Vol. 10, No. 10 6-h incubation in medium alone or with stimuli, the supernatant was removed and the total cellular RNA was isolated by a single step method with RNAwiz (Ambion, Austin, TX, U.S.A.) following the manufacturer’s instructions, and maintained at –80°C until its use.

801 camera (Kodak, Rochester, NY, U.S.A.) and 1D Image Analysis Software (Kodak). A 100 bp ladder (Promega, Madison, WI, U.S.A.) was used as a molecular weight marker. Optical density values recorded from each sample were normalized to the data obtained in a parallel amplification for GAPDH in the same sample. Therefore, results were expressed as the optical density ratio.

ELISA FOR CHEMOKINES AND IL-1


An enzyme-linked immunosorbant assay (ELISA) for chemokines was standardized with PharMingen (San Diego, CA, U.S.A.) reagents in our laboratory to determine the concentrations of IL-8, GRO- and MCP-1 in cell supernatants32. RANTES concentrations were evaluated with Endogen (Woburn, MA, U.S.A.) reagents, using 3 g/ml of capture antibody and 0.5 g/ml of detecting antibody, following the manufacturer’s recommended protocol. IL-1
concentrations were evaluated with a DuoSet® ELISA (R&D Systems), using 4 g/ml of capture antibody and 0.1 g/ml of detecting antibody, following the manufacturer’s recommended protocol. The detection limit was 15 pg/ml for all of the chemokines evaluated and 4 pg/ml for IL-1
. Streptavidin-HRP (PharMingen), diluted 1:1000, was used for the assay described, and the substrate was orthophenylene diamine (SIGMA). Absorbance was measured at 492 nm by an ELISA reader (Labsystems, Helsinki, Finland).

RT-PCR

For RT-PCR analysis, 1 g of total RNA per sample was converted to complementary DNA using Moloney murine leukemia virus reverse transcriptase (RT) (Perkin Elmer, Norwalk, CT, U.S.A.) for 1 h at 42°C. Specific amplification of IL-17R, IL-1
or GAPDH was then performed in a Gene Amp PCR System 9600 thermal cycler (Perkin Elmer), using recombinant Taq DNA polymerase (Perkin Elmer) and the following specific primer pairs: IL-17R extracellular domain sense, 5′-CTAAACTGCACGGTCAAGAAT-3′; IL-17R extracellular domain antisense, 5′-ATGAACCAGTA CACCCAC-3′; IL-1
sense, 5′-AAACAGATGAAGTGCTC CTTCAGG-3′; IL-1
anti-sense, 5′-TGGAGAACACCACTT GTTGCTCCA-3′; internal quality control glyceraldehyde-3phosphate dehydrogenase (GAPDH) sense, 5′-TGGTATC GTGGAAGGACTCATGAC-3′ and GAPDH anti-sense 5′-ATG CCAGTGAGCTTCCCGTTCAGC-3′, fragments sized 833, 391 and 190 bp (M-Medical Genenco, Florence, Italy). The Raji B and phorbol myristate acetate-activated monocytic U937 cell lines were used as positive controls for IL-17R33 and IL-1
, respectively. The reaction to detect IL-17R amplicons began with a 10 min incubation at room temperature with Taq Start antibody (Sigma), followed by 5 min denaturation at 94°C, 32 amplification cycles (denaturation at 94°C for 15 s; annealing-extension at 60°C for 30 s), and 7 min final extension at 72°C. The protocols for PCR amplification with primers for IL-1
and GAPDH were similar, except for the number of cycles (35 and 23 cycles, respectively). Semiquantitative analysis of samples was performed on serial three-fold dilutions of cDNA. The appropriate number of PCR cycles was previously determined for each primer set, in order to evaluate the linear amplification range and allow comparative analysis of signal strength. Reaction products were electrophoresed on a 2% agarose gel stained with 1 g/ml ethidium bromide. Resulting bands were analysed by densitometry of gel photographs, using a Kodak Digital Science DC 120

FLOW CYTOMETRY FOR IL-17R

The expression of IL-17R by chondrocytes and fibroblasts was evaluated by flow cytometry analysis. After washing trypsinized cells with PBS, chondrocytes or fibroblasts were fixed with 4% paraformaldehyde (SIGMA) in PBS for 15′ at 4°C. After washing cells with 0.1% saponin in PBS supplemented with 2% heat inactivated FCS (saponin-PBS) for a better membrane permeabilization, 105 cells were incubated with 20 g/ml anti-IL-17R mouse MoAb M202 (kindly provided by Immunex Corporation, Seattle, WA) for 30 min at 4°C in saponin-PBS. Then, after washing cells in saponin-PBS twice, they were incubated in 20 g/ml anti-IL-17 MoAb (R&D). This procedure was used to detect any binding of IL-17R to IL-17 which was used in the cell culture. This procedure was followed for all the samples tested. Finally, samples were washed and incubated as previously described with rabbit anti-mouse immunoglobulins conjugated to PE. Analysis was performed with a FACStar. All specimens were incubated in 0.1% saponin.

STATISTICAL ANALYSIS

Non-parametric tests were used for statistical analysis of the results: differences between groups were analysed by the Mann–Witney U test and when the Friedman ANOVA test for multiple comparison was significant, paired data were analysed by the Wilcoxon matched pairs test; correlations were analysed by Spearman test.

Results EFFECT OF IL-17 ON CHEMOKINE RELEASE BY SYNOVIAL FIBROBLASTS

Most of the synovial tissue (11/12) was obtained during knee surgery. Synovial fibroblasts maintained in culture with complete medium alone and for 72 h, produced small amounts of IL-8 and GRO-. The amounts of MCP-1 were significantly higher than those of IL-8 and GRO- (P