Characterization of adenosine receptors in bovine chondrocytes and fibroblast-like synoviocytes exposed to low frequency low energy pulsed electromagnetic fields

Osteoarthritis and Cartilage (2008) 16, 292e304 ª 2007 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.joca.2007.07.004

International Cartilage Repair Society

Characterization of adenosine receptors in bovine chondrocytes and fibroblast-like synoviocytes exposed to low frequency low energy pulsed electromagnetic fields

K. Varani Ph.D.ya, M. De Mattei Ph.D.za, F. Vincenzi Sc.D.y, S. Gessi Ph.D.y, S. Merighi Ph.D.y, A. Pellati Sc.D.z, A. Ongaro Ph.D.z, A. Caruso Ph.D.z, R. Cadossi Ph.D.x and P. A. Borea Ph.D.y* y Department of Clinical and Experimental Medicine, University of Ferrara, Italy z Department of Morphology and Hystology, University of Ferrara, Italy x Igea Biophysics Laboratory, Carpi, Italy Summary

Objective: The present study describes the presence and binding parameters of the A1, A2A, A2B and A3 adenosine receptors in bovine chondrocytes and fibroblast-like synoviocytes. The effect of low frequency low energy pulsed electromagnetic fields (PEMFs) on the adenosine receptor affinity and density was studied. Methods: Saturation, competition binding experiments and Western blotting assays in the absence and in the presence of PEMFs on the adenosine receptors in bovine chondrocytes or fibroblast-like synoviocytes were performed. Thermodynamic analysis of the A2A or A3 binding was studied to investigate the forces driving drugereceptor coupling. In the adenylyl cyclase and proliferation assays the potency of typical high-affinity A2A or A3 agonists in the absence and in the presence of PEMFs was evaluated. Results: Bovine chondrocytes and fibroblast-like synoviocytes expressed all adenosine receptors. PEMFs evoked an up-regulation of A2A and A3 receptors and thermodynamic parameters indicate that adenosine binding is enthalpy and entropy driven. In PEMF-treated cells the potency of typical A2A or A3 agonists on cyclic AMP assays was significantly increased when compared with the untreated cells. PEMFs potentiated the effect of A2A or A3 agonists on cell proliferation in both cell types. Conclusions: PEMFs mediate an up-regulation of A2A and A3 receptors related to an increase of their functional activities in bovine chondrocytes and fibroblast-like synoviocytes. No differences are present in adenosine affinity and in the drugereceptor interactions. Our data could be used as a trigger to future studies addressed to PEMFs and adenosine therapeutic intervention in inflammatory joint diseases. ª 2007 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved. Key words: Adenosine receptors, Chondrocytes, Fibroblast-like synoviocytes, Electromagnetic field, Binding assays, cAMP assays.

stimulate cAMP (cyclic AMP) accumulation8. A role of adenosine in modulating chondrocytes and fibroblast-like synoviocytes is documented by previous studies. In chondrocytes it has been reported that adenosine and its analogs were capable of altering cAMP and nitric oxide (NO) production1,2. In addition, it seems likely that for normal cartilage function, extracellular adenosine levels must be quite tightly regulated since depletion leads to increased glycosaminoglycan (GAG) release and the production of matrix metalloproteinases (MMPs) such as MMP-3 and MMP-13, prostaglandin E2 (PGE2) and NO, whilst its increase may trigger chondrocyte death2,9,10. In fibroblast-like synoviocytes adenosine receptor stimulation is involved in the regulation of MMPs11,12. Recently, in human synoviocytes it has been reported the selective involvement of the A2A receptor subtype in the immunomodulatory actions of methotrexate13. Further, in vivo adenosine A2A receptor agonists inhibit cartilage damage when used in the treatment of septic arthritis by diminishing interleukin-8 expression and GAG loss and reduce rat adjuvant induced arthritis14e16. Although A2A and A2B adenosine receptor transcripts have been described in articular chondrocytes17, however, the binding parameters such as affinity and density of adenosine subtypes have not

Introduction Chronic inflammation is a significant factor in the pathophysiology of many forms of joint disease. Adenosine has been reported to reduce inflammation in several in vivo models suggesting a potential value of this purine nucleoside as a therapeutic mediator of inflammatory joint disease able to limit articular cartilage degeneration1,2. In the peripheral system, adenosine has been shown to limit systemic inflammatory responses through receptor-mediated regulation of a wide variety of cell types3e7. Adenosine interacts with four cell surface receptor subtypes named as A1, A2A, A2B and A3, which are coupled to different G-proteins8. A1 and A3 receptors mediate inhibition of the adenylate cyclase activity, in contrast both A2A and A2B subtypes a

These authors contributed equally to this work *Address correspondence and reprint requests to: Prof. Dr Pier Andrea Borea, Chair of Pharmacology, Faculty of Medicine, University of Ferrara, Department of Clinical and Experimental Medicine, Pharmacology Unit, via Fossato di Mortara 17-19, 44100 Ferrara, Italy. Tel: 39-0532-291214; Fax: 39-0532-291205; E-mail: [email protected] Received 1 February 2007; revision accepted 3 July 2007.

292

293

Osteoarthritis and Cartilage Vol. 16, No. 3 been investigated in chondrocytes and fibroblast-like synoviocytes. Low frequency electromagnetic fields (PEMFs) have important effects on cartilage. In particular, in vitro cultured chondrocytes increase their proliferation as a function of PEMF exposure length18,19. In cartilage explants PEMFs increase proteoglycan synthesis preventing the catabolic effect of the pro-inflammatory cytokine IL-1 and acting in synergy with insulin-like growth factor I20e22. In vivo, it has been reported that PEMFs preserve the morphology of articular cartilage, retard the development of osteoarthritic lesions in guinea pig and actually articular cartilage immunohistochemistry demonstrated that pro-inflammatory cytokines interleukin-1 (IL-1) were down-regulated while anabolic cytokines were up-regulated transforming growth factor beta (TGF-b)23e27. Further, clinical studies in humans show that PEMF stimulation has beneficial effects in patients after arthroscopic surgery for cartilage lesions28. To explain the biological effects of PEMFs it has been proposed that they may influence ligand binding with cell membrane receptors affecting also the membrane protein distribution29. It has been also demonstrated that PEMFs evoke an up-regulation of the A2A and A3 adenosine receptors and alter the response of these receptor subtypes in human neutrophils30,31. Thus, the chondroprotective effects of PEMF might, at least partially, result from an anti-inflammatory mechanism of PEMFs together with adenosine receptor agonists23. From this background, the aim of this study was to investigate the expression and binding parameters of the A1, A2A, A2B and A3 adenosine receptors in bovine chondrocyte and fibroblast-like synoviocyte membranes. In the same cell types we have also studied the effect of PEMF treatment on the affinity and density of the A1, A2A, A2B and A3 adenosine receptors. A thermodynamic analysis of the A2A and A3 adenosine receptor binding was performed with the aim to investigate new insights into the effect of PEMFs on the forces driving drugereceptor coupling. Competition experiments, cAMP levels and cell proliferation assays of A2A or A3 receptor agonists in the absence and in the presence of PEMFs have been performed. Materials and methods CHONDROCYTE MONOLAYER CULTURES Bovine articular cartilage derived from the metacarpophalangeal joints of 14e18-month-old animals (Limousine breed). Chondrocytes were isolated from cartilage fragments obtained from the weight-bearing region of the articular surface32. Briefly, the cartilage was dissected out and cut into small pieces. Pieces were subjected to a sequential digestion in Dulbecco’s modified Eagle’s/Ham’s F12 (1:1) medium (DMEM/F12) (Gibco-Invitrogen, Paisley, UK) with pronase from Streptomyces griseus (Calbiochem, Darmstadt, Germany) for 90 min and collagenase P from Clostridium histolyticum (Roche, Indianapolis, USA) for about 12 h. The resulting cell suspension was filtered to remove undigested cartilage. Chondrocytes were recovered by centrifugation, counted and plated at high density (150,000/cm2) in 75 cm2 flasks (Falcon, Becton Dikinson and Company, Franklin Lakes, NJ, USA) and in multiwells (Nunc, Denmark, 6.6  6.6 cm, 1.6 cm the diameter of each well). Chondrocytes were cultured in DMEM/F12 supplemented with 10% fetal bovine serum (FBS) and antibiotics (penicillin 100 U/ml, streptomycin 0.1 mg/ml) (Gibco-Invitrogen, Paisley, UK) (complete medium). Only chondrocytes without subculturing and maintained in culture for 1 week were used in the binding and functional experiments.

CHONDROCYTES AND FIBROBLAST-LIKE SYNOVIOCYTES’ CHARACTERIZATION The expression of aggrecan, type II and type I collagens in the same chondrocyte monolayer cultures used for binding experiments was evaluated by immunohistochemistry as previously described22. Incubations with the primary mouse monoclonal antibody [6-B-4] to aggrecan (1:50) and the rabbit polyclonal to collagen type II (1:200) (Abcam, Cambridge Science Park, Cambridge, UK) were performed for 1 h at 37 C. Fibroblast-like synoviocytes were characterized by immunofluorescence staining with vimentin, a specific cellular marker for mesenchymal cells33. Cells were fixed with cold methanol, washed with phosphate buffer saline (PBS) and incubated with the primary mouse antibody specific for the human vimentin at 1:200 dilution for 1 h at 37 C. Washed slides were then incubated with a secondary fluorescein isothiocyanate-conjugated goat anti-mouse antibody for 1 h at 37 C in the dark. Antibodies were from SigmaeAldrich S.r.l. (Milan, Italy). Fluorescence was visualized using the Nikon Eclipse TE 2000-E microscope (Nikon Instruments Spa, Sesto Fiorentino, Firenze, Italy) equipped with a digital camera (DXM 1200F, Nikon Instruments Spa, Sesto Fiorentino, Firenze, Italy). To confirm that fibroblast-like synoviocyte cultures were not contaminated by macrophages, CD14 expression was evaluated by reverse transcription polymerase chain reaction (RT-PCR). Total RNA extraction was performed using a commercial kit. RNA conversion to cDNA was performed using the commercial kit Superscript First-strand Synthesis System for RT-PCR (Invitrogen Life Technologies, Carlsbad, CA, USA). Two microliters of cDNA were amplified by specific oligonucleotide primers for CD 14 (dp50 -CTG GAA GCC GGC G-30 ; rp50 -AGC TGA GCA GGA ACC TGT GC-30 ). Primer sequences were selected to amplify both human and bovine genome and were from separate exons to exclude a possibly genomic DNA contamination of the RNA samples. PCR reactions were performed in a total volume of 25 ml containing 1 U Taq DNA polymerase (Roche Molecular Biochemicals, Indiana, USA), 25 pmol of each primer, 200 mM deoxynucleotide triphosphates (dNTPs) in 1 PCR buffer (10 mM Tris, pH 8.3, 50 mM KCl, 1.5 mM MgCl2). Cycling parameters were as follows: 1 min at 94 C, 1 min at 55 C, 1 min at 72 C for CD14. The size of the amplified sequence was 403 bp. mRNA from human macrophages was used as a positive control for CD14 expression. PCR products were analyzed on 1.5% agarose gel, stained with ethidium bromide.

PREPARATION OF BOVINE CHONDROCYTE AND FIBROBLAST-LIKE SYNOVIOCYTE MEMBRANES For membrane preparation, the culture medium was removed the cells were washed with PBS and scraped off T75 flasks in ice-cold hypotonic buffer (5 mM TriseHCl, 2 mM ethylendiamine tetraacetic acid (EDTA) pH 7.4). The cell suspension was homogenized by using a Polytron and was centrifuged for 30 min at 100,000g. The membrane pellet was resuspended in the same buffer solution used in the binding experiments, incubated with 2 IU/ml of adenosine deaminase for 30 min at 37 C and centrifuged for 30 min at 100,000g. Finally the suspension was used in saturation and competition binding experiments. The protein concentration was determined according to a Bio-Rad method with bovine albumin as reference standard34.

FIELD CHARACTERISTICS The chondrocyte and fibroblast-like synoviocyte membranes or monolayer cultures were exposed to a PEMF generated by a pair of rectangular horizontal coils (14  23 cm) each made of 1400 turns of copper wire; coils were powered by a pulse generator (IGEA, Italy) used in previous works21,22,30,31. The characteristics of the field were as follows: the pulse duration of the signal was 1.3 ms and the repetition rate was 75 Hz, yielding a duty cycle of 1/10. The intensity peak of the magnetic field was 1.5 mT and the induced electric field, as detected with a standard coil probe (50 turns, 0.5 cm internal diameter of the coil probe, 0.2 cm copper diameter), was 0.07 mV/cm. The temperature, continuously monitored by a thermoresistor within the incubator, was constant through the exposure time and exactly maintained during the binding and functional experiments.

WESTERN BLOTTING OF ADENOSINE RECEPTORS FIBROBLAST-LIKE SYNOVIOCYTE CULTURES Fibroblast-like synoviocytes were obtained by culture of the bovine synovial fluid. Synovial fluid was aspirated from the metacarpophalangeal joints of 14e18-month-old animals by a syringe. Then, fresh synovial fluid was diluted 1:4 with complete medium and plated in 25 cm2 culture flasks (Falcon, Becton Dikinson and Company, Franklin Lakes, NJ, USA). After 3 h, medium was removed and fresh complete medium was added to the flasks. Cells were maintained in culture and passaged when reaching confluence. Synovial cells at the third and fourth passages were used for the analysis of adenosine receptors.

Chondrocytes and fibroblast-like synoviocytes untreated or PEMF-treated were harvested and washed with ice-cold PBS containing 1 mM sodium orthovanadate, 104 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, 0.08 mM aprotinin, 2 mM leupeptin, 4 mM bestatin, 1.5 mM pepstatin A, and 1.4 mM E-64. Then cells were lysed in Triton lysis buffer and the protein concentration was determined using bicinchoninic acid (BCA) protein assay kit (Pierce, Illinois, USA). Aliquots of total protein sample (50 mg) were analyzed using antibodies specific for human A1, A2A, A2B and A3 adenosine receptors (1 mg/ml dilution)35. Filters were washed and incubated for 1 h at room temperature with peroxidase-conjugated secondary antibodies (1:2000 dilution). Specific reactions were revealed

294

K. Varani et al.: Adenosine receptors in chondrocytes and synoviocytes

with Enhanced Chemiluminescence Western blotting detection reagent (Amersham Biosciences, New York, USA).

SATURATION AND COMPETITION BINDING EXPERIMENTS TO ADENOSINE RECEPTORS The chondrocyte and synoviocyte membranes were PEMF treated for the specific incubation times relative to the A1, A2A, A2B and A3 binding experiments30,31. Saturation binding experiments to A1 adenosine receptors were performed according to the method described previously using [3H]-1,3-dipropyl-8-cyclopentyl-xanthine ([3H]-DPCPX, specific activity 120 Ci/mmol; NEN-Perkin Elmer Life and Analytical Sciences, USA) as radioligand36. The membranes derived from PEMFs-treated or untreated chondrocytes or fibroblast-like synoviocytes (100 mg of protein/assay) with 8e10 concentrations of the radioligand [3H]-DPCPX (0.01e20 nM) were incubated in TriseHCl 50 mM, pH 7.4, for 90 min at 4 C. Nonspecific binding was determined in the presence of DPCPX 1 mM. Saturation binding experiments to A2A adenosine receptors were performed according to the method described previously using [3H]-4-(2-[7-amino-2-(2-furyl)[1,2,4] triazolo [2,3-a] [1,3,5] triazin-5-yl-amino]ethyl ([3H]-ZM 241385, specific activity 27.4 Ci/mmol; American Radiolabeled Chemicals Inc, Saint Louis, MO, USA) as radioligand5. The membranes derived from PEMFs-treated or untreated chondrocytes or fibroblast-like synoviocytes (100 mg of protein/assay) were incubated for 60 min at 4 C with 8e10 concentrations of the radioligand [3H]-ZM 241385 (0.01e20 nM) and TriseHCl 50 mM, MgCl2 10 mM, pH 7.4. Nonspecific binding was determined in the presence of ZM 241385 1 mM. Saturation binding experiments to A2B adenosine receptors were performed using [3H]-N-benzo [1,3[dioxol-5-yl-2-[5-(2,6-dioxo1,3-dipropyl-2,3,6,7-tetrahydro-1H-purin-8-yl)-1-methyl-1H-pyrazol-3-yloxy]-acetamide ([3H]-MRE 2029F20, specific activity 123 Ci/mmol; Amersham International Chemical Laboratories, Buckinghamshire, UK) as radioligand37. The membranes obtained as previously described (100 mg of protein/assay) with 8e10 concentrations of [3H]-MRE 2029F20 in the range 0.01e20 nM were incubated in TriseHCl 50 mM, MgCl2 10 mM, EDTA 1 mM, pH 7.4 at 4 C for 60 min. Nonspecific binding was determined in the presence of MRE 2029F20 1 mM. Saturation binding experiments to A3 adenosine receptors were performed using [3H]-5N-(4-methoxyphenylcarbamoyl) amino-8-propyl-2-(2-furyl) pyrazolo [4,3-e]-1,2,4-triazolo [1,5-c]pyrimidine ([3H]-MRE 3008F20, specific activity 67 Ci/mmol; Amersham International Chemical Laboratories, Buckinghamshire, UK) as radioligand38. The membranes treated as above mentioned (100 mg of protein/assay) with 8e10 concentrations in the range 0.01e50 nM of [3H]-MRE 3008F20 were incubated in TriseHCl 50 mM, MgCl2 10 mM, EDTA 1 mM, pH 7.4, at 4 C for 150 min. Nonspecific binding was determined in the presence of MRE 3008F20 1 mM. In competition experiments, carried out to determine the A2A or A3 affinity values, 1 nM of [3H]-ZM 241385 or 2 nM of [3H]-MRE 3008F20, bovine chondrocyte or fibroblast-like synoviocyte membranes (100 mg of protein per assay) and at least 6e8 different concentrations of 2-[p-(2-carboxyethyl)phenetyl-amino]-50 -N-ethyl-carboxamido adenosine (CGS 21680, Sigma-RBI, St. Louis, MO, USA) or N 6-(3-iodobenzyl)2-chloroadenosine-50 -N-methyluronamide (Cl-IB-MECA, Sigma-RBI, St. Louis, MO, USA) as typical A2A or A3 adenosine agonists were incubated at 4 C for 60 or 150 min, respectively. In saturation or competition binding experiments, at the end of the incubation time, bound and free radioactivity was separated by filtering the assay mixture through Whatman GF/B glass fiber filters by using a Brandel cell harvester. The filter bound radioactivity was counted by Scintillation Counter Packard Tri Carb 2500 TR with an efficiency of 58%.

THERMODYNAMIC ANALYSIS For a generic binding equilibrium L þ R ¼ LR (L ¼ ligand, R ¼ receptor) the affinity association constant KA ¼ 1/KD is directly related to the standard free energy DG (DG ¼ RT ln KA) which can be separated in its enthalpic and entropic contributions according to the Gibbs equation: DG ¼ DH   TDS . The standard free energy was calculated as DG ¼ RT ln KA at 298.15 K, the standard enthalpy, DH  , from the van’t Hoff plot ln KA vs (1/T ) (the slope is DH  /R) and the standard entropy as DS ¼ (DH   DG )/T with T ¼ 298.15 K and R ¼ 8.314 J/K/mol39. KA values were obtained from saturation experiments of [3H]-ZM 241385 or [3H]-MRE 3008F20 binding to the human PEMF-treated or untreated bovine chondrocyte and fibroblast-like synoviocyte membranes carried out at 0, 10, 15, 20, 25 and 30 C in a thermostatic bath assuring a temperature of 0.1 C.

MEASUREMENT OF cAMP LEVELS IN BOVINE CHONDROCYTES OR FIBROBLAST-LIKE SYNOVIOCYTES PEMF-treated or untreated bovine chondrocytes or fibroblast-like synoviocytes (106 cells/ml) were suspended in 0.5 ml incubation mixture Krebs Ringer phosphate buffer, containing 1.0 IU/ml adenosine deaminase and 0.5 mM 4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone (Ro 20-1724) as

phosphodiesterase inhibitor and preincubated for 10 min in a shaking bath at 37 C. Then the effect of a typical A2A adenosine agonist was studied by using CGS 21680 at different concentrations (1 nMe1 mM) that was added to the mixture for a further 5 min. In similar experimental conditions, the effect of N-ethylcarboxamidoadenosine (NECA) on adenosine nonselective agonist was studied. To evaluate the effect of a typical A3 adenosine agonist, forskolin 1 mM and ClIB-MECA at different concentrations (0.1 nMe100 nM) were added to the mixture and the incubation continued for a further 5 min. The effect of a selective A2A or A3 antagonist such as 7-(2-phenylethyl)-2-furyl)pyrazolo [4,3-e]-1,2,4-triazolo-[1,5-c] pyrimidine (SCH 58261) (1 mM) or MRE 3008F20 (1 mM) on CGS 21680 (1 mM) or Cl-IB-MECA (100 nM) was evaluated, respectively. The reaction was terminated by the addition of cold 6% trichloroacetic acid (TCA). The cells were also incubated with forskolin (1 mM) and/or Ro 20-1724 (0.5 mM) to evaluate the adenylyl cyclase activity. The TCA suspension was centrifuged at 2000g for 10 min at 4 C and the supernatant was extracted four times with water saturated diethyl ether. The final aqueous solution was tested for cAMP levels through a competition protein binding assay by using [3H]-cAMP as radioligand (specific activity 21 Ci/mmol, NEN Research Products, Boston, MA, USA)30. Samples of cAMP standards (0e10 pmol) were added to each test tube containing trizma base 0.1 M, aminophylline 8.0 mM, mercaptoethanol 6.0 mM, pH 7.4 and [3H]-cAMP (at the final concentration of 1 nM). The binding protein, previously prepared from beef adrenals, was added to the samples and incubated at 4 C for 150 min. At the end of the incubation time and after the addition of charcoal the samples were centrifuged at 2000g for 10 min. The clear supernatant was mixed with 4 ml of Atomlight and counted in a Scintillation Counter Packard Tri Carb 2500 TR. PEMF RESPONSE SPECIFICITY Binding to a2 adrenergic receptors was carried out with [3H]-UK 14304 (0.2e10 nM) on bovine chondrocytes or fibroblast-like synoviocytes in a 50 mM TriseHCl buffer, pH 7.4 containing MgCl2 10 mM for 60 min at 25 C. Nonspecific binding was determined with 1 mM of UK 14304. Binding to b2 adrenergic receptors was carried out with [3H]-CGP 12177 (0.1e10 nM) on bovine chondrocytes or fibroblast-like synoviocytes in a 50 mM TriseHCl buffer, pH 7.4 containing MgCl2 10 mM for 60 min at 25 C. Nonspecific binding was determined with 1 mM of CGP 12177. The effects of isoproterenol (1 nMe10 mM) on stimulation of adenylate cyclase activity was evaluated to verify if PEMFs involve the functionality of another Gprotein coupled receptors30. [3H] THYMIDINE INCORPORATION Chondrocytes or fibroblast-like synoviocytes were treated with CGS 21680 and Cl-IB-MECA (10 mM) in complete medium containing adenosine deaminase (ADA) and 1 mCi/ml [3H] thymidine. Parallel cultures were exposed to PEMF. Cells cultured in the absence of adenosine agonists and PEMF exposure were used as controls. After 24 h of labeling cells were trypsinized, and [3H] thymidine was evaluated as previously described35. In all cultures, cell viability was evaluated by the Trypan blue exclusion test40. STATISTICAL ANALYSIS A weighted nonlinear least-squares curve fitting program Ligand41 was used for computer analysis of saturation and competition binding experiments. Functional experiments were calculated by nonlinear regression analysis using the equation for a sigmoid concentrationeresponse curve (GraphPAD Prism, San Diego, CA, USA). Analysis of data was done with Student’s t test (unpaired analysis). Differences were considered significant at a value of P < 0.01. All data are reported as mean  S.E.M. of independent experiments (n ¼ 3e6 as indicated in Results).

Results PHENOTYPE CHARACTERIZATION OF BOVINE CHONDROCYTES AND FIBROBLAST-LIKE SYNOVIOCYTES

Chondrocytes, cultured in monolayer, show phenotypic instability and can dedifferentiate toward the fibroblast phenotype shifting from the synthesis of type II to type I collagen22. When analyzed by immunohistochemistry, under the same experimental conditions of binding experiments, chondrocytes stained positive for aggrecan and type II collagen, specific markers of the chondrocytic phenotype, and resulted negative for type I collagen, which is expressed in dedifferentiated chondrocytes (Fig. 1). Negative controls, obtained without primary antibodies, did not show any staining. Fibroblast-like synoviocytes used in our experiments

295

Osteoarthritis and Cartilage Vol. 16, No. 3

A

M

B

MC

SF

SF

SF

400 bp

28s 18s

Fig. 2. Expression of vimentin in synovial fibroblasts (A) (original magnification 20). Analysis of CD14 mRNA expression (B upper panel) in macrophages (MC) and in bovine fibroblast-like synoviocytes (FS). M is 100 bp DNA ladder (Biolabs). One microgram of total RNA was loaded for lane and stained with ethidium bromide to confirm equal RNA quantity used for RT-PCR (B lower panel). WESTERN BLOTTING

Fig. 1. Immunohistochemistry for aggrecan (A), type II (B) and type I (C) collagens of bovine chondrocytes cultured in monolayers. Nuclei were counterstained by hematoxylin.

showed the expression of vimentin, the main intermediate filament protein in mesenchymal cells and synovial fibroblast [Fig. 2(A)]. Results obtained by RT-PCR showed the absence of CD14, thus indicating the absence of contaminating macrophages in our cultures [Fig. 2(B)].

Figure 3(A) shows the immunoblot signals of A1, A2A and A2B adenosine receptors in bovine chondrocytes and fibroblast-like synoviocytes in the absence and in the presence of PEMF. The intensity of each band in immunoblot assay was quantified using molecular analyst/PC densitometry software (Bio-Rad). Mean densitometry data from three independent experiments were normalized to control that was set to 100% [Fig. 3(B,C)]. Interestingly, in both the cells examined only A2A adenosine receptors was altered in the presence of PEMFs showing a fold of increase statistically significant from 1.45 to 1.65 (*, P < 0.02). Unfortunately Western blot of A3 adenosine receptors was undetectable probably due to the low degree of homology between bovines and humans.

SATURATION BINDING EXPERIMENTS IN BOVINE CHONDROCYTES AND FIBROBLAST-LIKE SYNOVIOCYTES

A series of experiments were carried out to determine the effect of PEMFs on the binding parameters of A1, A2A, A2B and A3 adenosine receptors in bovine chondrocytes and fibroblast-like synoviocytes. All experiments were performed

296

K. Varani et al.: Adenosine receptors in chondrocytes and synoviocytes A1 R

A -

A2A R + -

+

A2B R + -

Bovine Chondrocytes

Bovine fibroblast-like Synoviocytes - Control + PEMF exposure

Bovine Chondrocytes

B 200

Control

Protein level (% of control)

PEMF exposure

*

150

*, P<0.02

100

50

0 A1 R

A2A R

A2B R

Bovine fibroblast-like Synoviocytes

C 200

Protein level (% of control)

*

Control PEMF exposure

*, P<0.02

150

100

50

0 A1 R

A2A R

A2B R

Fig. 3. Western blotting analysis of bovine chondrocytes and fibroblast-like synoviocytes in the absence and in the presence of PEMFs for A1, A2A and A2B adenosine receptors (A). Each band was quantified by using molecular analyst/PC densitometry software (Bio-Rad). Mean densitometry data from three independent experiments were normalized to control from chondrocytes (B) and fibroblast-like synoviocytes (C).

with an intensity of PEMFs of 1.5 mT according to the previous studies30,31 where no effect of PEMFs on adenosine receptors in human neutrophils was observed under 0.5 mT and the maximum effect appeared from 1 to 3.5 mT reaching a stable plateau. Table I reports the affinity (KD, nM) and density (Bmax, fmol/mg protein) of A1, A2A, A2B and A3 adenosine receptors in bovine chondrocyte and in fibroblast-like synoviocyte membranes untreated or exposed to PEMF. The affinity values were strictly similar in both PEMF-treated or untreated cells (n ¼ 4). Moreover also the Bmax values (n ¼ 4) of A1 and A2B adenosine receptors did not change in the presence of PEMFs (Table I). On the contrary the Bmax

values of A2A and A3 adenosine receptors were altered after the PEMF treatment. Figures 4 and 5 report a saturation curve of [3H]-ZM 241385 binding to A2A adenosine receptors in bovine chondrocyte [Fig. 4(A)] and in fibroblast-like synoviocyte [Fig. 5(A)] membranes untreated or treated with PEMFs. Figures 4 and 5 show also a saturation curve of [3H]-MRE 3008F20 binding to A3 adenosine receptors in bovine chondrocyte [Fig. 4(C)] and in fibroblast-like synoviocyte [Fig. 5(C)] membranes untreated or treated with PEMFs, respectively. The Scatchard plot analysis of A2A adenosine receptors indicated the presence of a single class of binding sites with a KD value of 1.71  0.32 nM and a Bmax value of

297

Osteoarthritis and Cartilage Vol. 16, No. 3 Table I Saturation binding experiments on adenosine receptors in bovine chondrocyte and in fibroblast-like synoviocyte membranes Chondrocyte membranes KD (nM)

Fibroblast-like synoviocyte membranes

Bmax (fmol/mg protein)

KD (nM)

Bmax (fmol/mg protein)

A1 adenosine receptors Controls 2.18  0.22 PEMF 2.17  0.16

41  2 40  3

0.67  0.01 0.68  0.03

30  2 31  3

A2A adenosine receptors Controls 1.71  0.32 PEMF 2.25  0.24

53  5 100  8*

2.05  0.17 2.52  0.20

76  6 181  11*

A2B adenosine receptors Controls 2.19  0.24 PEMF 2.02  0.14

51  3 55  3

1.21  0.13 1.18  0.08

84  5 78  6

A3 adenosine receptors Controls 4.61  0.35 PEMF 4.48  0.15

79  5 183  8*

1.86  0.22 2.08  0.09

83  4 185  6*

Data are expressed as the mean  S.E.M. (n ¼ 4 or 6 independent experiments as indicated in Results); *, P < 0.01 vs controls.

53  5 fmol/mg protein in untreated bovine chondrocyte [n ¼ 6; Fig. 4(B)]. In bovine chondrocyte membranes treated with PEMFs the KD value was 2.25  0.24 nM and the Bmax value was 100  8* fmol/mg protein [*, P < 0.01 vs controls, n ¼ 6; Fig. 4(B)]. The Scatchard plot analysis of A3 adenosine receptors indicated the presence of a single class of binding sites with a KD value of 4.61  0.35 nM and a Bmax value of 79  5 fmol/mg protein in untreated bovine chondrocyte [n ¼ 6; Fig. 4(D)]. In bovine chondrocyte treated with PEMFs the KD value was 4.48  0.15 nM and the Bmax value was

183  8* fmol/mg protein [*, P < 0.01 vs controls, n ¼ 6; Fig. 4(D)]. The Scatchard plot analysis of A2A and A3 adenosine receptors in bovine fibroblast-like synoviocyte membranes indicated the presence of a single class of binding sites with a KD value of 2.05  0.17 nM and 1.86  0.22 nM, respectively [n ¼ 6; Fig. 5(B,D)]. The receptor density, in the same cell line, expressed as Bmax value was 76  6 and 83  4 fmol/mg protein, respectively [n ¼ 6; Fig. 5(B,D)]. In bovine fibroblast-like synoviocyte membranes treated with PEMFs the KD and Bmax values of A2A adenosine receptors

120

[3H]-ZM 241385 bound (fmol/mg protein)

Bovine chondrocytes

A

100

B

A2A R

Bound/Free

80 60 40 20

60

A2A R control KD =1.71±0.32 nM Bmax =53±5 fmol/mg protein

40

PEMF exposure KD =2.25± 0.24 nM Bmax =100±8* fmol/mg protein *, P<0.01

20

control PEMF exposure

0 0

5

10

15

0 0

20

50

[3H]-ZM 241385 free (nM) 200

D

A3 R

60

100

50

150

A3 R control KD =4.61±0.35 nM Bmax =79±5 fmol/mg protein

150

Bound/Free

[3H]MRE 3008F20 bound (fmol/mg protein)

C

100

Bound

40

PEMF exposure KD =4.48±0.15 nM Bmax =183±8* fmol/mg protein *, P<0.01

20

control PEMF exposure

0

0 0

10

20

30

40

[3H]MRE 3008F20 free (nM)

50

0

50

100

150

200

250

Bound

Fig. 4. Saturation of [3H]-ZM 241385 binding to A2A adenosine receptors (A) and of [3H]-MRE 3008F20 binding to A3 adenosine receptors (C) on untreated or PEMF-treated bovine chondrocyte membranes. Experiments were performed as described in Materials and methods. Values are the means and vertical lines S.E.M. of six separate experiments performed in triplicate. The Scatchard plots of the same data are shown in panels B and D.

298

K. Varani et al.: Adenosine receptors in chondrocytes and synoviocytes

Bovine fibroblast-like synoviocytes 200

B

A2A R

150

100

A2A R

control KD=2.05±0.17 nM Bmax=76±6 fmol/mg protein

75

Bound/Free

[3H]-ZM 241385 bound (fmol/mg protein)

A

100

50

control

PEMF exposure KD=2.52±0.20 nM Bmax=181±11* fmol/mg protein *, P<0.01

50

25

PEMF exposure 0

0 0

5

10

15

20

0

50

[3H]-ZM 241385 free (nM) 200

150

200

250

Bound

D

A3 R

150

Bound/Free

[3H]MRE 3008F20 bound (fmol/mg protein)

C

100

100

50

100

A3 R

control KD=1.86±0.22 nM Bmax=83±4 fmol/mg protein

75

PEMF exposure KD=2.08±0.09 nM Bmax=185±6* fmol/mg protein *, P<0.01

50 25

control PEMF exposure

0 0

5

10

15

0 20

[3H]MRE 3008F20 free (nM)

0

50

100

150

200

250

Bound

Fig. 5. Saturation of [3H]-ZM 241385 binding to A2A adenosine receptors (A,B) and of [3H]-MRE 3008F20 binding to A3 adenosine receptors (C,D) on untreated or PEMF-treated bovine fibroblast-like synoviocyte membranes. Experiments were performed as described in Materials and methods. Values are the means and vertical lines S.E.M. of six separate experiments performed in triplicate. The Scatchard plots of the same data are shown in panels B and D.

were 2.52  0.20 nM and 181  11* fmol/mg protein, respectively [*, P < 0.01 vs controls, n ¼ 6; Fig. 5(B)]. The Scatchard plot analysis of A3 adenosine receptors in the same cells treated with PEMFs indicated the presence of a single class of binding sites with a KD value of 2.08  0.09 nM and a Bmax value of 185  6* fmol/mg protein [*, P < 0.01 vs controls, n ¼ 6; Fig. 5(D)]. THERMODYNAMIC STUDIES

Saturation experiments were performed at 0, 10, 15, 20, 25 and 30 C and the Scatchard plots resulted to be linear in the concentration range investigated (n ¼ 3). In untreated bovine chondrocyte or in fibroblast-like synoviocyte membranes KD values of [3H]-ZM 241385 or [3H]-MRE 3008F20 binding to A2A or A3 adenosine receptors increased with the increase of investigated temperatures. While the dissociation constant (KD) changed with temperature, Bmax values of A2A or A3 adenosine receptors appeared to be largely independent of it. Figure 6 shows the van’t Hoff of ln KA vs 1/T for [3H]-ZM 241385 or [3H]-MRE 3008F20 binding to the A2A or A3 adenosine receptors in untreated or PEMF-treated bovine chondrocyte [Fig. 6(A,B)] or in fibroblast-like synoviocyte membranes [Fig. 6(C,D)]. The final equilibrium thermodynamic parameters of A2A adenosine receptors (expressed as mean values  S.E.M. of three independent experiments) in untreated bovine chondrocyte membranes were as follows: DG ¼ 47.27  0.14 kJ/mol; DH  ¼ 38.92  3.53 kJ/mol and DS ¼ 28.78  2.84 J/mol/K. The thermodynamic parameters of A3 adenosine receptors in untreated bovine chondrocyte

membranes were as follows: DG ¼ 47.80  0.16 kJ/mol; DH  ¼ 33.93  3.42 kJ/mol and DS ¼ 44.48  4.25 J/mol/K. The final equilibrium thermodynamic parameters of A2A adenosine receptors in untreated bovine fibroblast-like synoviocyte membranes were as follows: DG ¼ 47.22  0.11 kJ/mol; DH  ¼ 35.22  3.77 kJ/mol and DS ¼ 30.71  3.03 J/mol/K. The thermodynamic parameters of A3 adenosine receptors in untreated bovine fibroblast-like synoviocyte membranes were as follows: DG ¼ 48.12  0.14 kJ/mol; DH  ¼ 35.59  3.67 kJ/mol and DS ¼ 42.08  4.15 J/mol/K. Similar results were also obtained in bovine chondrocyte or fibroblast-like synoviocyte membranes treated with PEMFs suggesting that the final equilibrium thermodynamic parameters in the examined substrates were not influenced by the PEMF treatment and indicating that the binding to A2A or A3 adenosine receptors was enthalpy and entropy driven. COMPETITION BINDING EXPERIMENTS IN BOVINE CHONDROCYTES AND FIBROBLAST-LIKE SYNOVIOCYTES

Figure 7 shows the competition curves of CGS 21680 (A,C) and Cl-IB-MECA (B,D) in untreated or PEMF-treated bovine chondrocyte or in fibroblast-like synoviocyte membranes, respectively. The affinity (Ki) of CGS 21680 or Cl-IBMECA in untreated bovine chondrocyte membranes was 28  3 or 2.6  0.3 nM, respectively (n ¼ 4). In parallel studies, performed in the same cell membranes treated with PEMFs, CGS 21680 or Cl-IB-MECA had a Ki value of 25  2 or 2.9  0.3 nM, respectively (n ¼ 4). In untreated

299

Osteoarthritis and Cartilage Vol. 16, No. 3

Bovine chondrocytes control ΔH°=-38.92±3.53 kJmol-1 ΔS°= 28.78±2.84 Jmol-1 K-1

A2A R

-1

20

ΔG°=-47.27±0.14 kJmol

19

PEMF exposure ΔH°=-38.08±3.72 kJmol-1 ΔS°= 30.44±2.97 Jmol-1 K-1 ΔG°=-47.13±0.15 kJmol

18 3.2

3.4

3.6

B

21

control

A3 R

ΔH°= -33.93±3.42 kJmol-1 ΔS°= 44.48±4.25 Jmol-1 K-1

20

ΔG°=-47.80±0.16 kJmol-1

ln KA

21

ln KA

A

PEMF exposure ΔH°= -32.85±3.27 kJmol-1 ΔS°=46.15 ±4.32 Jmol-1 K-1

19

-1

ΔG°=-47.69±0.13 kJmol-1 18 3.2

3.8

T-1 (Kx1000)

3.4

3.6

3.8

T-1 (Kx1000)

Bovine fibroblast-like synoviocytes 21

control ΔH°=-35.22±3.77 kJmol-1 ΔS°=30.71±3.03 Jmol-1 K-1

A2A R

D

ln KA

PEMF exposure ΔH°=-37.24±3.68 kJmol-1 ΔS°= 33.14±3.85 Jmol-1 K-1

19

control

A3 R

ΔG°=-47.22±0.11 kJmol-1

20

21

ΔH°=-35.59±3.67 kJmol-1 ΔS°=42.08±4.15 Jmol-1 K-1

20

ΔG°=-48.12±0.14 kJmol-1

ln KA

C

PEMF exposure ΔH°=-34.76±3.46 kJmol-1 ΔS°=44.45±4.23 Jmol-1 K-1

19

ΔG°= -47.35±0.13 kJmol-1 18

ΔG°=-48.02±0.16 kJmol-1 18

3.2

3.4

3.6

3.8

T-1 (Kx1000)

3.2

3.4

3.6

3.8

T-1 (Kx1000)

Fig. 6. Van’t Hoff plot showing the effect of temperature on the equilibrium binding association constant, KA ¼ 1/KD of [3H]-ZM 241385 to A2A adenosine receptors (A,C) or [3H]-MRE 3008F20 to A3 adenosine receptors (B,D) in untreated or PEMF-treated bovine fibroblast-like synoviocyte membranes, respectively. The plots are essentially linear in the temperature range investigated (4e30 C). Binding experiments were performed as described in Materials and methods (n ¼ 3).

fibroblast-like synoviocyte membranes CGS 21680 or Cl-IBMECA showed a Ki value of 30  3 or 3.3  0.3 nM, respectively (n ¼ 4). In PEMF-treated fibroblast-like synoviocyte membranes CGS 21680 or Cl-IB-MECA revealed a Ki value of 27  2 or 3.5  0.3 nM, respectively (n ¼ 4). The competition curves of both the agonists examined, performed in the absence and in the presence of PEMFs, exhibited Hill coefficients near the unity and were best described by the existence of one high-affinity binding site.

fmol/mg protein; b2 adrenergic receptors: KD ¼ 2.13  0.16 nM; Bmax ¼ 21  2 fmol/mg protein) or in fibroblast-like synoviocytes (a2 adrenergic receptors: KD ¼ 2.18  0.20 nM; Bmax ¼ 24  3 fmol/mg protein; b2 adrenergic receptors: KD ¼ 2.11  0.16 nM; Bmax ¼ 28  3 fmol/mg protein). In addition, in bovine chondrocytes and in fibroblast-like synoviocytes, isoproterenol stimulated cAMP levels showing EC50 values of 250  30 and 282  26 nM, respectively. PEMFs in the same cell types did not modify the potency of isoproterenol (EC50 ¼ 237  24 and 267  23 nM, respectively).

PEMF RESPONSE SPECIFICITY

To verify that PEMF activity is specific to A2A and A3 adenosine receptors, we studied other types of membrane receptors coupled to G-proteins. The expression of a2 and b2 adrenergic receptors in PEMF-treated and untreated bovine chondrocytes or in fibroblast-like synoviocytes was determined by performing saturation binding experiments using [3H]-UK 14304 or [3H]-CGP 12177, respectively. A single saturation binding site was detected for adrenergic receptors in untreated or PEMF-treated bovine chondrocytes or in fibroblast-like synoviocytes. In bovine chondrocytes or in fibroblast-like synoviocytes, saturation of [3H]-UK 14304 binding shows a KD value of 2.18  0.15 or 2.06  0.11 nM and a Bmax value of 18  2 or 23  3 fmol/ mg protein, respectively. In the same cells, [3H]-CGP 12177 exhibits high affinity for b2 adrenergic receptors with a KD value of 2.05  0.12 or 2.07  0.10 nM and a Bmax value of 20  2 or 26  3 fmol/mg protein, respectively. None of these binding were significantly affected by PEMFs in bovine chondrocytes (a2 adrenergic receptors: KD ¼ 2.25  0.05 nM; Bmax ¼ 20  2

cAMP ASSAYS IN BOVINE CHONDROCYTES AND FIBROBLAST-LIKE SYNOVIOCYTES

The A2A adenosine receptors are coupled to stimulation of adenylate cyclase via Gs stimulatory proteins, which leads to an increase of cAMP formation. Untreated or PEMF-treated bovine chondrocyte or fibroblast-like synovial cells did not reveal change of basal enzyme activity and of the response of adenylate cyclase activator forskolin. No change in cAMP production was also observed in the absence or in the presence of the cAMP-dependent phosphodiesterase inhibitor, Ro 20-1724 (Table II). We have also evaluated the effect of a typical A2A adenosine agonist such as CGS 21680 on the adenylate cyclase activity. When CGS 21680 was incubated with PEMF-treated or untreated examined cells an amplification of adenylate cyclase response was detected revealing a significant increase of cAMP production in a concentration-dependent manner. As shown in Fig. 8 the log dose response curve for CGS 21680 in control and in PEMF-treated bovine

300

K. Varani et al.: Adenosine receptors in chondrocytes and synoviocytes

Bovine chondrocytes 100

B

A2AR

80

Control Ki = 28±3 nM

60

PEMF exposure Ki = 25±2 nM

40 20

[3H]MRE 3008F20 bound (% of specific binding)

[3H]ZM 241385 bound (% of specific binding)

A

100

A3 R

75

Control Ki = 2.6±0.3 nM

50

PEMF exposure Ki = 2.9±0.3 nM

25

0

0 -10

-9

-8

-7

-6

-11

-10

log [CGS 21680] (M)

-9

-8

-7

-6

log [Cl-IB-MECA] (M)

Bovine fibroblast-like synoviocytes 100

D

A2A R

80

Control Ki = 30±3 nM

60

PEMF exposure Ki = 27±2 nM

40 20

[3H]MRE 3008F20 bound (% of specific binding)

[3H]ZM 241385 bound (% of specific binding)

C

100

A3 R

75

Control Ki = 3.3±0.3 nM

50

PEMF exposure Ki = 3.5±0.3 nM

25

0

0 -10

-9

-8

-7

-6

log [CGS 21680] (M)

-11

-10

-9

-8

-7

-6

log [Cl-IB-MECA] (M)

Fig. 7. Competition experiments of specific [3H]-ZM 241385 to A2A adenosine receptors (A,C) or [3H]-MRE 3008F20 binding to A3 adenosine receptors (B,D) of CGS 21680 or Cl-IB-MECA in untreated and PEMF-treated bovine chondrocytes (A,B) or fibroblast-like synoviocyte (C,D) membranes. Competition experiments were performed as described in Materials and methods (n ¼ 4).

chondrocytes or in fibroblast-like synoviocytes revealed a shift vs left of the curve suggesting an increase in cAMP production after PEMF treatment. In particular CGS 21680 elicited a stimulation of cAMP levels in untreated or PEMF-treated bovine chondrocytes with an EC50 of 82  7 and 46  5* nM, respectively [*, P < 0.01, Fig. 8(A)]. Moreover, CGS 21680 elicited a stimulation of cAMP levels in untreated or PEMF treated in fibroblastlike synoviocytes with an EC50 of 91  8 and 54  6* nM, respectively [*, P < 0.01, Fig. 8(C)]. The selective A2A antagonist SCH 58261 (1 mM) totally inhibited the rise in cAMP levels induced by CGS 21680 (1 mM) suggesting that the stimulatory effect was essentially A2A mediated (n ¼ 4, Table II). On the contrary, the A3 adenosine receptors are coupled to inhibition of adenylyl cyclase via Gi proteins, which leads to decreases of cAMP formation. We have also evaluated the inhibitory effect of a typical A3 adenosine agonist like Cl-IBMECA on the adenylyl cyclase activity. When forskolin (1 mM) and the agonist were incubated with PEMF-treated or untreated neutrophils, a reduction of adenylyl cyclase response was detected revealing a significant decrease of cAMP production in a concentration-dependent manner (n ¼ 4, Fig. 8). Cl-IB-MECA determined a decrease of cAMP levels in untreated or PEMF-treated bovine chondrocytes with an IC50 values of 6.33  0.64 and 3.37  0.31* nM, respectively [*, P < 0.01, n ¼ 4, Fig. 8(B)]. Moreover Cl-IB-MECA induced a decrease of cAMP levels in untreated or PEMF-treated fibroblast-like synoviocytes with an IC50 values of 7.90  0.71 and 4.03  0.42* nM,

respectively [*, P < 0.01, n ¼ 4, Fig. 8(D)]. The selective A3 antagonist MRE 3008F20 (1 mM) antagonized Cl-IB-MECA (100 nM) mediated cAMP inhibition suggesting that the inhibitory effect was essentially A3 mediated (n ¼ 4, Table II). NECA, a typical adenosine receptor agonist, determines a stimulation of cAMP levels in untreated or PEMF-treated bovine chondrocytes with an EC50 value of 40  3 or 66  5* nM, respectively (*, P < 0.01). Analogous results were obtained in fibroblast-like synoviocytes with an EC50 value of 37  3 and 59  5* nM, respectively (*, P < 0.01). This stimulatory effect due to the interaction of NECA with A2A and A2B adenosine receptors predominates on the inhibitory effect mediated by A1 and A3 adenosine receptors.

CELL PROLIFERATION AND VIABILITY

Chondrocytes or fibroblast-like synoviocytes treated with CGS 21680 or Cl-IB-MECA (10 mM) showed no statistically significant modulation of cell proliferation. However, in fibroblast-like synoviocytes, Cl-IB-MECA induced a slight reduction of cell proliferation. PEMF exposure did not modify cell proliferation in the absence of adenosine agonists with respect to controls. Interestingly, in the presence of PEMFs, CGS 21680 induced an increase on cell proliferation in both cells which was significant in fibroblast-like synoviocytes. Analogously, the inhibitory effect of Cl-IB-MECA was increased by PEMFs (Fig. 9). In all tested conditions cell viability was not changed with respect to control conditions (99% viable cells).

301

Osteoarthritis and Cartilage Vol. 16, No. 3 Table II Basal and stimulated cAMP production in bovine chondrocytes and in fibroblast-like synoviocytes Chondrocytes (pmol cAMP  106 cells)

Basal levels þ Ro 20-1724 (0.5 mM) þ CGS 21680 (1 mM) A2A stimulation þ CGS 21680 (1 mM) þ SCH 58261 (1 mM) Forskolin (1 mM) Forskolin (1 mM) Ro 20-1724 (0.5 mM) þ Cl-IB-MECA (100 nM) A3 inhibition þ Cl-IB-MECA (100 nM) þMRE 3008F20 (1 mM)

Fibroblast-like synoviocytes (pmol cAMP  106 cells)

Control

PEMF treatment

Control

PEMF treatment

10  1 25  2 51  4 27  2

11  1 27  2 80  6* 28  2

11  1 23  2 50  4 25  2

12  1 24  2 78  6* 26  2

98  10 120  12 82  6 116  11

95  9 115  13 45  5* 110  12

96  9 112  11 84  6 108  11

97  10 118  12 46  5* 114  10

Data are expressed as the mean  S.E.M. (n ¼ 4 independent experiments); *, P < 0.01 vs controls.

Discussion The purpose of the current investigation was to document the presence and the binding parameters of adenosine receptor subtypes in bovine chondrocytes or in fibroblastlike synoviocytes and to evaluate the effect of PEMFs on adenosine receptors in these cell types. Preliminary characterization of cells used in our experiments showed the expression of chondrocyte and synoviocyte phenotypic markers, as well as the absence of contaminating cells in fibroblast-like synoviocyte cultures22. Our pharmacological data in unexposed cells report that A1, A2A, A2B and A3 receptors are expressed and similarly distributed in both cell types showing high-affinity values

in the nanomolar range and a receptor density from 30 to 83 fmol/mg of protein. After PEMF treatment A1 and A2B receptors show similar binding parameters with respect to untreated cells. A single class of A2A and A3 binding sites with a similar affinity in untreated or PEMF-treated examined cells was found according to saturation binding experiments performed with [3H]-ZM241385 or [3H]-MRE 3008F20, respectively. On the contrary, the number of binding sites in PEMF-treated bovine chondrocytes or in fibroblast-like synoviocytes was increased significantly (P < 0.01) vs the control conditions. These data revealed that the predominant effect of PEMFs mediated a significant increase of A2A or A3 adenosine receptor density. The presence of adenosine

Bovine chondrocytes 100

B

A2A R

80 60

control EC50 = 82±7 nM

40

PEMF exposure EC50 = 46±5* nM * P<0.01

20

% cAMP accumulation

% cAMP accumulation

A

100

A3 R

control IC50 = 6.33±0.64 nM PEMF exposure IC50 = 3.37±0.31* nM * P<0.01

80

60

40

0 -10

-9

-8

-7

-6

-11

-5

-10

log [CGS 21680] (M)

-9

-8

-7

-6

log [Cl-IB-MECA] (M)

Bovine fibroblast-like synoviocytes 100

D

A2A R

80 60

control EC50 = 91±8 nM

40

PEMF exposure EC50 = 54±6* nM * P<0.01

20 0 -10

% cAMP accumulation

% cAMP accumulation

C

100

A3 R

control IC50 = 7.90±0.71 nM PEMF exposure IC50 = 4.03±0.42* nM * P<0.01

80

60

40 -9

-8

-7

log [CGS 21680] (M)

-6

-5

-11

-10

-9

-8

-7

-6

log [Cl-IB-MECA] (M)

Fig. 8. cAMP experiments by using a typical A2A agonist, CGS 21680 (A,C) or a typical A3 agonist, Cl-IB-MECA (B,D) in untreated and PEMF-treated bovine chondrocytes (A,B) or fibroblast-like synoviocytes (C,D). cAMP experiments were performed as described in Materials and methods (n ¼ 4).

302

K. Varani et al.: Adenosine receptors in chondrocytes and synoviocytes

Bovine Chondrocytes

A Cell proliferation (% of basal)

160

Control PEMF exposure

120

80

40

0 Basal

Cl-IB-MECA

Bovine fibroblast-like Synoviocytes

B

*

160

Cell proliferation (% of basal)

CGS 21680

Control PEMF exposure

120

*, P<0.05

80

* 40

0 Basal

CGS 21680

Cl-IB-MECA

Fig. 9. Proliferation activity measured by [3H] thymidine incorporation assay. Bovine chondrocytes (A) or fibroblast-like synoviocytes (B) were treated with CGS 21680 and Cl-IB-MECA at 10 mM concentration.

receptors and the effect of PEMFs was also investigated through Western blotting analysis that confirmed binding data described above showing no differences for A1 and A2B receptors and a statistically significant increase for A2A receptors in PEMF-treated cells compared to controls. As a consequence of the saturation binding experiments we have selectively studied A2A or A3 adenosine receptor binding site and the mechanism of interaction with typical adenosine ligands. Thermodynamic parameters obtained from the van’t Hoff plot indicate that [3H]-ZM 241385 or [3H]-MRE 3008F20 binding to A2A and A3 adenosine receptors is enthalpy and entropy driven, with a major contribution of the enthalpic component in agreement with data obtained in other cells expressing human adenosine receptors or by using typical adenosine antagonists6,7,31,37,38. Further, the final thermodynamic parameters in the examined substrates were not influenced by PEMF treatment suggesting that no alteration in the drugereceptor interaction was found. The results obtained in competition binding experiments show that the affinity of CGS 21680 or Cl-IB-MECA in the PEMF-treated bovine chondrocytes or in fibroblast-like synoviocytes were strictly similar to those obtained in untreated cells confirming that the treatment did not modify the druge receptor interaction and affinity values of these agonists. In addition, the saturation binding experiments performed on other G-protein coupled receptors such as a2 and b2

adrenergic subtypes suggest that PEMFs did not modify the binding parameters. The capability of isoproterenol to stimulate adenylate cyclase activity and to determine an increase of cAMP levels is not modified by the presence of PEMFs suggesting that the response to PEMF is strictly linked to A2A and A3 adenosine receptors. A similar PEMF specificity was also reported in human neutrophils30. Another purpose of the present study was to investigate if the increase in A2A or A3 adenosine receptor binding sites induced by PEMF treatment might be related to a modulation of receptor functional activities. To this aim, we analyzed the effects of A2A and A3 agonists on adenylyl cyclase activity and cell proliferation in the absence and in the presence of PEMF treatment. Our results did not show any change of basal enzyme activity and of the response of adenylyl cyclase to the direct activator forskolin used in the absence or in the presence of cAMP-dependent phosphodiesterase inhibitor (Ro 201724). Forskolin, which directly activated adenylate cyclase was utilized in this study as a positive control for cAMP production. Ro 20-1724, a type IV phosphodiesterase inhibitor prevented the rapid degradation of cAMP allowing the accurate detection of the cAMP levels produced. Moreover, we have evaluated the capability of typical A2A or A3 adenosine agonists such as CGS 21680 or Cl-IB-MECA to modulate cAMP levels. These compounds showed potency values in the nanomolar range, in agreement with their affinity in binding experiments. Interestingly, in the PEMF-treated bovine chondrocytes or in fibroblast-like synoviocytes the potency of CGS 21680 or Cl-IB-MECA, respectively, in stimulating and inhibiting cAMP production, were significantly increased when compared with the untreated cells. To further confirm that the effects induced by the agonists on cAMP formation were due to the modulation of adenosine receptors by PEMFs, we performed experiments in the presence of typical selective A2A or A3 adenosine antagonists such as SCH 58261 or MRE 3008F20. These antagonists in the presence or in the absence of PEMFs were able to prevent the effect of cAMP induced by CGS 21680 or Cl-IB-MECA through a selective modulation of the adenylyl cyclase via the A2A or A3 receptors, respectively. In addition, the levels of cAMP increased in response to NECA, a nonselective agonist, suggesting that in physiological condition (presence of endogenous adenosine) the activation of A2A may prevail over the A3 mediated inhibitory action. This is not surprising because a similar behavior has already been reported in other substrates6,7. In line with cAMP data, PEMFs did not modify cell proliferation in the absence of adenosine agonists and in the presence of ADA. On the other hand, previous studies report that PEMFs, in the absence of ADA, can increase cell proliferation18,19. On the contrary, in the presence of CGS 21680, PEMF exposure was able to stimulate cell proliferation. Further, Cl-IB-MECA inhibited slightly cell proliferation in fibroblast-like synoviocytes and this effect was potentiated by PEMF exposure. Collectively, our results on cAMP and cell proliferation assays show that the increase in adenosine binding sites induced by PEMFs is correlated to an increase of A2A or A3 activity in the presence of adenosine receptor agonists. In conclusion, the novel findings of this study in bovine chondrocytes and fibroblastlike synoviocytes are as follows: (1) identification and binding characterization of adenosine receptors; (2) upregulation of A2A and A3 adenosine receptors by PEMFs but not of A1 and A2B; (3) alteration of cAMP levels and cell proliferation by PEMFs related to the increase of A2A and A3 adenosine receptor density; (4) similar affinity of

Osteoarthritis and Cartilage Vol. 16, No. 3 typical A2A and A3 agonists in the absence or in the presence of PEMFs; (5) no changes by PEMFs in the drugereceptor interaction through thermodynamic analysis. These results identify for the first time a molecular target for PEMFs and provide a rational base to suggest a potential anti-inflammatory role of PEMFs mediated by adenosine receptors in chondrocytes and fibroblast-like synoviocytes. The modulation of PEMFs on A2A and A3 receptors may help to explain previous data showing that in vitro PEMF exposure protects bovine full-thickness cartilage explants from the catabolic effect of pro-inflammatory cytokine IL-1 and in vivo prevents osteoarthritis (OA) progression in DunkineHartley guinea pigs23,27. It is to note that the effects of PEMFs may be different in pathologic conditions as age differences in response to physical forces have been reported42. On the other hand PEMF exposure has a chondroprotective effect on OA progression in DunkineHartley guinea pigs of different age and OA severity23,24. Further clinical studies have demonstrated a positive effect of PEMF stimulation in patients after knee arthroscopy or suffering from knee OA28,43. Although additional studies are needed to fully understand the relationship between adenosine-mediated anti-inflammatory effects44,45 and PEMF exposure, our results open interesting perspectives to develop new antiinflammatory approaches in local joint disease treatment. References 1. Tesch AM, MacDonald MH, Kollas-Baker C, Benton HP. Chondrocytes respond to adenosine via A2 receptors and activity is potentiated by an adenosine deaminase inhibitor and a phosphodiesterase inhibitor. Osteoarthritis Cartilage 2002;10:34e43. 2. Tesch AM, MacDonald MH, Kollas-Baker C, Benton HP. Endogenously produced adenosine regulates articular cartilage matrix homeostasis: enzymatic depletion of adenosine stimulates matrix degradation. Osteoarthritis Cartilage 2004;12:349e59. 3. Varani K, Gessi S, Dalpiaz A, Ongini E, Borea PA. Characterization of A2A adenosine receptors in human lymphocyte membranes by 3HSCH 58261 binding. Br J Pharmacol 1997;122:386e92. 4. Varani K, Gessi S, Dionisotti S, Ongini E, Borea PA. 3H-SCH 58261 labelling of functional A2A adenosine receptors in human neutrophil membranes. Br J Pharmacol 1998;123:1723e31. 5. Varani K, Laghi-Pasini F, Camurri A, Capecchi PA, Maccherini M, Di Ciolla F, et al. Changes of peripheral A2A adenosine receptors in chronic heart failure and cardiac transplantation. FASEB J 2003;17: 280e2. 6. Gessi S, Varani K, Merighi S, Cattabriga E, Iannotta V, Leung E, et al. A3 adenosine receptors in human neutrophils and promyelocytic HL60 cells: a pharmacological and biochemical study. Mol Pharmacol 2002;61: 415e24. 7. Gessi S, Varani K, Merighi S, Cattabriga E, Avitabile A, Gavioli R, et al. Expression of A3 adenosine receptors in human lymphocytes: upregulation in T cell activation. Mol Pharmacol 2004;65:711e9. 8. Fredholm BB, Ijzerman AP, Jacobson KA, Klotz KN, Linden J. International Union of Pharmacology. XXV. Nomenclature and classification of adenosine receptors. Pharmacol Rev 2001;53:527e52. 9. Benton HP, MacDonald MH, Tesch AM. Effects of adenosine on bacterial lipopolysaccharide and interleukin induced nitric oxide release from equine articular chondrocytes. Am J Vet Res 2002;63:204e10. 10. Mistry D, Chambers MG, Mason RM. The role of adenosine in chondrocyte death in murine osteoarthritis and in murine chondrocyte cell line. Osteoarthritis Cartilage 2006;14:486e95. 11. Boyle DL, Sajjadi FG, Firestein GS. Inhibition of synoviocyte collagenase gene expression by adenosine receptor stimulation. Arthritis Rheum 1996;39:923e30. 12. Boyle DL, Zuoning H, Rutter JL, Brinckerhoff E, Firestein GS. Posttranscriptional regulation of collagenase-1 gene expression in synoviocytes by adenosine receptor stimulation. Arthritis Rheum 1997;40:1772e9. 13. Ralph JA, McEvoy AN, Kane D, Bresnihan B, FitzGerald O, Murphy EP. Modulation of orphan nuclear receptor NURR1 expression by methotrexate in human inflammatory joint disease involves adenosine A2A receptor-mediated responses. J Immunol 2005;175:555e65. 14. Boyle DL, Moore J, Yang L, Sorkin LS, Firestein GS. Spinal adenosine receptor activation inhibits inflammation and joint destruction in rat adjuvant induced arthritis. Arthritis Rheum 2002;46:3076e82.

303 15. Cohen SB, Gill SS, Baer GS, Leo BM, Scheld WM, Diduch DR. Reducing joint destruction due to septic arthrosis using an adenosine A2A receptor agonist. J Orthop Res 2004;22:427e35. 16. Cohen SB, Leo BM, Baer GS, Turner MA, Beck G, Diduch DR. An adenosine A2A receptor agonist reduces interleukin-8 expression and glycosaminoglycan loss following septic arthrosis. J Orthop Res 2005; 23:1172e8. 17. Koolpe M, Pearson D, Benton HP. Expression of both P1 and P2 purine receptor genes by human articular chondrocytes and profile of ligand mediated prostaglandin E2 release. Arthritis Rheum 1999;42:258e67. 18. De Mattei M, Caruso A, Traina GC, Pezzetti F, Baroni T, Sollazzo V. Correlation between pulsed electromagnetic fields exposure time and cell proliferation increase in human osteosarcoma cell lines and human normal osteoblast cells in vitro. Bioelectromagnetics 1999; 20:177e82. 19. De Mattei M, Caruso A, Pezzetti F, Pellati A, Stabellini G, Traina GC. Effects of pulsed electromagnetic fields on human articular chondrocyte proliferation. Connect Tissue Res 2001;42:269e79. 20. De Mattei M, Pasello M, Pellati A, Stabellini G, Massari L, Gemmati D, et al. Effects of electromagnetic fields on proteoglycan metabolism of bovine articular cartilage explants. Connect Tissue Res 2003;44: 154e9. 21. De Mattei M, Fini M, Setti S, Ongaro A, Gemmati D, Stabellini G, et al. Proteoglycan synthesis in bovine articular cartilage explants exposed to different low-frequency low energy pulsed electromagnetic fields. Osteoarthritis Cartilage 2007;15:163e8. 22. De Mattei M, Pellati A, Pasello M, Ongaro A, Setti S, Massari L, et al. Effects of physical stimulation with electromagnetic field and insulin growth factor-I treatment on proteoglycan synthesis of bovine articular cartilage. Osteoarthritis Cartilage 2004;12:793e800. 23. Fini M, Giavaresi G, Torricelli P, Cavani F, Setti S, Cane` V, et al. Pulsed electromagnetic fields reduce knee osteoarthritic lesion progression in the aged Dunkin Hartley guinea pig. J Orthop Res 2005;23:899e908. 24. Fini M, Torricelli P, Giavaresi G, Nicoli Aldini N, Cavani F, Setti S, et al. Effect of pulsed electromagnetic field stimulation on knee cartilage, subchondral and epyphiseal trabecular bone of aged Dunkin Hartley guinea pigs. Biomed Pharmacother 2007 Apr 3 [Epub ahead of print]. 25. Aaron RK, Boyan BD, Ciombor DM, Schwartz Z, Simon BJ. Stimulation of growth factor synthesis by electric and electromagnetic fields. Clin Orthop Relat Res 2004;419:30e7. 26. Aaron RK, Ciombor DM, Wang S, Simon B. Clinical biophysics: the promotion of skeletal repair by physical forces. Ann N Y Acad Sci 2006; 1068:513e31. 27. Ciombor DM, Aaron RK, Wang S, Simon B. Modification of osteoarthritis by pulsed electromagnetic field e a morphological study. Osteoarthritis Cartilage 2003;11:455e62. 28. Zorzi C, Dall’oca C, Cadossi R, Setti S. Effects of pulsed electromagnetic fields on patients’ recovery after arthroscopic surgery: prospective, randomized and double-blind study. Knee Surg Sports Traumatol Arthrosc 2007 Feb 28 [Epub ahead of print]. 29. Chiabrera A, Bianco B, Moggia E, Kaufman JJ. ZeamaneStark modeling of the RF EMF interaction with ligand binding. Bioelectromagnetics 2000;21:312e24. 30. Varani K, Gessi S, Merighi S, Iannotta V, Cattabriga E, Spisani S, et al. Effect of low frequency electromagnetic fields on A2A adenosine receptors in human neutrophils. Br J Pharmacol 2002;136:57e66. 31. Varani K, Gessi S, Merighi S, Iannotta V, Cattabriga E, Pancaldi C, et al. Alteration of A3 adenosine receptors in human neutrophils and low frequency electromagnetic fields. Biochem Pharmacol 2003;66: 1897e906. 32. Pezzetti F, De Mattei M, Caruso A, Cadossi R, Zucchini P, Carinci F, et al. Effects of pulsed electromagnetic fields on human chondrocytes: an in vitro study. Calcif Tissue Int 1999;65:396e401. 33. Upragarin NA, van Asten PC, Topten W, Landman JM, Gruys E. Serum amyloid a production by chicken fibroblast-like synoviocytes. Vet Immunol Immunopathol 2005;106:39e51. 34. Bradford M. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein dye-binding. Anal Biochem 1976;72:248e54. 35. Merighi S, Mirandola P, Dilani D, Varani K, Gessi S, Klotz KN, et al. Adenosine receptors as mediators of both cell proliferation and cell death of cultured human melanoma cells. J Invest Dermatol 2002; 119:923e33. 36. Borea PA, Varani K, Dalpiaz A, Capuzzo A, Fabbri E, Ijzerman AP. Full and partial agonists and thermodynamic binding parameters of adenosine A1 receptor ligands. Eur J Pharmacol 1994;267:55e61. 37. Gessi S, Varani K, Merighi S, Cattabriga E, Pancaldi C, Szabadkai Y, et al. Expression, pharmacological profile and functional coupling of A2B receptors in a recombinant system and in peripheral blood cells by using a novel selective antagonist radioligand, [3H]-MRE 2029F20. Mol Pharmacol 2005;67:1e11. 38. Varani K, Merighi S, Gessi S, Klotz KN, Leung E, Baraldi PG, et al. [3H]MRE 3008-F20: a novel antagonist radioligand for the pharmacologi-

304

K. Varani et al.: Adenosine receptors in chondrocytes and synoviocytes

cal and biochemical characterization of human A3 adenosine receptors. Mol Pharmacol 2000;57:968e75. 39. Borea PA, Dalpiaz A, Varani K, Gessi S, Gilli G. Binding thermodynamics at A1 and A2A adenosine receptors. Life Sci 1996;59: 1373e88. 40. Merighi S, Benini A, Mirandola P, Gessi S, Varani K, Leung E, et al. A3 adenosine receptor activation inhibits cell proliferation via phosphatidylinositol 3-kinase/Akt-dependent inhibition of the extracellular signalregulated kinase phosphorylation in A375 human melanoma cells. J Biol Chem 2005;20:19516e26. 41. Munson PJ, Rodbard D. Ligand: a versatile computerized approach for the characterization of ligand binding systems. Anal Biochem 1980; 107:220e39.

42. Bobacz K, Graninger WB, Amoyo L, Smolen JS. Effect of pulsed electromagnetic fields on proteoglycan biosynthesis of articular cartilage is age dependent. Ann Rheum Dis 2006;65(7): 949e51. 43. Trock DH, Bollet AJ, Markoll R. The effect of pulsed electromagnetic fields in the treatment of osteoarthritis of the knee and cervical spine. Report of randomized, double blind, placebo controlled trials. J Rheumatol 1994;21(10):1903e11. 44. Picher M, Graff RD, Lee GM. Extracellular nucleotide metabolism and signaling in the pathophysiology of articular cartilage. Arthritis Rheum 2003;48:2722e36. 45. Burnstock G. Purinergic signalling. An overview. Novartis Found Symp 2006;276:26e48.