Human articular cartilage was obtained with ethical approval from the Lothian Research Ethics Committee and patients' consent at operation from knee joint arthroplasty specimens and amputations for peripheral vascular disease. Cartilage was assessed macroscopically for the presence or absence of osteoarthritic changes and graded for OA using the Collins/McElligott system. Chondrocytes were isolated by sequential enzyme digestion, and cells were seeded in Iscove's modified Dulbecco's medium (Gibco, now part of Invitrogen Corporation, Carlsbad, CA, USA) supplemented with 10% fetal calf serum (Sigma-Aldrich, St. Louis, MO, USA), 100 IU/mL penicillin (Invitrogen Corporation), and 100 μg/mL streptomycin (Invitrogen Corporation) to a final density of 5 × 10⁵/mL. Primary, non-confluent, 1- to 2-week cultures of chondrocytes were used in all experiments. The day before mechanical stimulation, culture media containing serum was replaced by serum-free media. In some experiments, cell lines were used. OV10 cells, an ovarian carcinoma cell line lacking CD47/IAP, were stably transfected with human CD47/IAP form 2 cDNA in the absence (315 – CD47-2) or presence (315/164 – β₃/CD47-2) of β₃ cDNA or with CD47's extracellular IgV domain linked to a glycosylphosphatidylinositol anchor (OV10 148A10T – CD47IgV-GPI) 22. 3657L cells are polyclonal CD47⁺ fibroblasts derived from B6 mice. 3656L cells are lung fibroblasts derived from CD47/IAP-null mice 23 and transfected with mouse (1/171) or human (1/315) CD47/IAP form 2 cDNA or vector control (1/Sra).

The following primary antibodies were used in electrophysiology experiments, immunohistology, and immunoprecipitation/Western blotting as indicated. Anti-CD47/IAP – Bric 126 (International Blood Group Reference Laboratory, Bristol, UK), CC2C6 (Hans-Jörg Büring, University of Tübingen, Germany), B6H12, miap 400.1, and 2B7 (Washington University School of Medicine, St. Louis, MO, USA), goat polyclonal anti-CD47 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), anti-CD49e/α5 integrin – Sam-1 (Serotec Ltd., Oxford, UK), anti-CD29/β1 integrin clone JB1A (Chemicon International, Temecula, CA, USA), anti-TSP clone P10 (Chemicon International), anti-SIRPα (rabbit polyclonal antibody; ABR, Affinity BioReagents, Inc., Golden, CO, USA), anti-HRP (phosphotyrosine-horseradish peroxidase) conjugated (Amersham Life Sciences, now part of GE Healthcare, Little Chalfont, Buckinghamshire, UK), and anti-focal adhesion kinase (Santa Cruz Biotechnology, Inc.).

Samples of normal articular cartilage (Collins grade 0) obtained from 1 female (age 67 years) and 7 males (median age 71 years, range 53 to 88 years) and osteoarthritic cartilage (Collins grade 1 to 3) obtained from 13 males (median age 71 years, range 53 to 88 years) were snap-frozen in liquid nitrogen. Sections (4 μm) were cut with a Bright cryostat (Bright Instrument Co. Ltd., Huntingdon, UK), mounted on poly-L-lysine-coated glass slides, allowed to come to room temperature, and fixed with acetone for 10 minutes. Sections were stained with Bric 126 by an avidin-biotin-immunoperoxidase technique. For negative control, the primary antibody was replaced by non-immune mouse immunoglobulin.

The methods for protein extraction, immunoprecipitation, and Western blotting used have been described previously 7. Cells at rest or following mechanical stimulation were washed with ice-cold phosphate-buffered saline containing 100 μM Na₃VO₄ (Sigma-Aldrich) and lysed in situ with ice-cold lysis buffer containing 1% Igepal (Sigma-Aldrich), 100 μM Na₃VO₄, and protease inhibitor cocktail tablet (Boehringer Ingelheim GmbH, Ingelheim, Germany) at 4°C for 15 minutes. Supernatants were collected after centrifugation at 13,000 rpm for 15 minutes. For CD47/IAP immunoprecipitation, a 1-mL aliquot of protein at a concentration of 500 μg/mL was incubated at 4°C with either Bric 126 or CC2C6 for 1 hour and then with protein A-Sepharose (Pharmacia-LKB Biotechnology, Uppsala, Sweden) for 1 hour. Whole-cell extracts or immunoprecipitated proteins were separated on a 7.5% SDS-PAGE under reducing conditions. Following electrophoresis, immunoprecipitated proteins or whole-cell lysates were transferred onto polyvinylidene fluoride membranes (Immobilon-P; Millipore Corporation, Billerica, MA, USA, and Sigma-Aldrich). Membranes were blocked overnight at 4°C with 2% bovine serum albumin (BSA) in TBST (12.5 mM Tris/HCl, pH 7.6, 137 mM NaCl, 0.1% Tween 20). After washing with TBST, blots were incubated for 1 hour at room temperature with primary antibodies and then HRP-labelled secondary antibodies. Membranes were rewashed extensively and bands were detected using Enhanced Chemiluminescense Plus (GE Healthcare). For immunoprecipitation of phosphotyrosine, monoclonal antiphosphotyrosine agarose beads (Sigma-Aldrich) were used.

Membrane potentials of cells were recorded using a single electrode bridge circuit and calibrator 489. Membrane potentials of cells were measured and results were accepted if, on cell impalement, any rapid changes in voltage to the membrane potential level remained constant for at least 30 seconds. The membrane potentials of 5 to 10 cells were measured prior to and following addition of the reagent to be tested and/or mechanical stimulation. Each experiment was undertaken at least three times on cells from different donors. The technique and apparatus used for mechanical stimulation of primary human articular chondrocytes have been described in detail 4. Plastic tissue culture dishes (diameter 55 mm) (NUNC A/S, Roskilde, Denmark) containing sparse primary monolayer cultures of human articular chondrocyte monolayer were placed in a sealed pressure chamber with inlet and outlet ports. The chamber was pressurised using helium gas from a cylinder at a frequency determined by an electronic timer that controlled the inlet and outlet valves. The standard stimulation regime used was a frequency of 0.33 Hz (2 seconds on, 1 second off) for 20 minutes at 37°C at a pressure of 16 kPa above atmospheric pressure. This system produces 3700 microstrain on the base of the culture dish.

Total RNA was extracted from cultured chondrocytes using a denaturing buffer of 4 M guanidine thiocyanate, 0.75 M sodium citrate, 10% (wt/vol) lauroyl sarcosine, and 7.2 μL/mL β-mercaptoethanol. Prior to cDNA synthesis, all RNA samples were incubated with DNAse I (Life Technologies, now part of Invitrogen Corporation) for 15 minutes in the presence of an RNAse inhibitor (Pharmacia, now part of GE Healthcare). Template cDNA was synthesised using 0.5 μg of RNA, Superscript II, and oligo dT (12–18) (Invitrogen Corporation) according to the manufacturer's instructions. The primers used for the polymerase chain reactions (PCRs) were (upstream/downstream) CD47/IAP, 5'-GTTGGACTGAGTCTCTGTATTGCGGCGTGT-3' and 5'-CACAAGTGTATTCCTTTCACGTCTTACTACTC-3'; glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5'-CCACCCATGGCAAATTCCATGGCA-3' and 5'-TCTAGACGGCAGGTCAGGTCCACC-3'; and aggrecan, 5'-TGAGGAGGGCTGGAACAAGTACC-3' and 5'-GGAGGTGGTAATTGCAGGGAACA-3'.

A typical 20-μL PCR contained 20 mM ammonium sulphate, 75 mM Tris-HCl, pH 8.8, 0.01% (vol/vol) Tween-20, 1 μM each primer, 2 μL of cDNA, 100 μM dNTPs, 0.1% (wt/vol) BSA, and 0.25 U Taq polymerase (BioGene Ltd., Kimbolton, UK). The magnesium chloride concentrations for each primer pair were 2.5 mM (GAPDH) and 1.25 mM (aggrecan), and the following programme was used for aggrecan reactions: 94°C for 3 minutes, 24 cycles of 94°C for 1 minute, 60°C for 1 minute, and 72°C for 1 minute 30 seconds. PCR products were analysed by electrophoresis using a 1% (wt/vol) agarose gel stained with ethidium bromide, and the intensity of each band was measured under UV fluorescence using EASY Image analysis software (Scotlab Ltd., Coatbridge, UK). The ratio of intensities of the bands for the aggrecan product compared with the housekeeping gene GAPDH was calculated and compared. Each donor was tested in duplicate, and at least three donors were used for each experiment.

The mean, standard deviation, and standard error of the mean were calculated for each time point. The Student t test was used to evaluate whether there was statistical significance between each of the points calculated.

The vast majority of chondrocytes in both normal (n = 8) and osteoarthritic (n = 13) cartilage were strongly stained with antibodies to CD47/IAP (Figure 1a,b). There was no apparent difference in immunoreactivity between chondrocytes in normal and osteoarthritic cartilage. Similarly, no difference in expression of CD47/IAP was seen when protein extracts from primary monolayer cultures of chondrocytes derived from normal and osteoarthritic cartilage were immunoblotted with a panel of three different anti-CD47/IAP antibodies. In both normal and OA chondrocytes, CD47/IAP was identified as a molecule of approximately 50 kDa molecular weight (Figure 2a). Reverse transcription-polymerase chain reaction from extracted RNA resulted in the production of two products, with a predominant product consistent with expression of the type 2 isoform of CD47/IAP (Figure 2b). The cell line OV10-315 is transfected with CD47/IAP type 2 DNA and therefore expresses only this isoform.

CD47/IAP has been shown to be associated with a number of different extracellular and membrane proteins, including integrins, TSPs, and SIRPα, in a variety of cell types. To establish whether similar associations were occurring in human articular chondrocytes, a series of experiments were undertaken to show the presence of potential ligands and to investigate possible interactions. Protein extracts from primary cultures of human articular chondrocytes were first immunoblotted for TSP-1 and SIRPα, and expression of these molecules by human articular chondrocytes was confirmed (Figure 3a). Protein extracts from human articular chondrocytes were then immunoprecipitated with anti-CD47/IAP, and the precipitated proteins were subjected to gel electrophoresis and Western blotting with antibodies specific for TSP-1, SIRPα, and α5 integrin. TSP-1 and α5 integrin, but not SIRPα, were shown to immunoprecipitate with CD47/IAP (Figure 3b). Immunoprecipitation of α5 integrin and β1 integrin and subsequent Western blotting showed co-precipitation of CD47/IAP with α5 integrin precipitates only (results not shown).

To investigate potential roles for CD47/IAP in chondrocyte mechanotransduction, chondrocytes were subjected to cyclical mechanical stimulation and the responses of the cells assessed by measuring either changes in membrane potential, changes in protein tyrosine phosphorylation, or changes in the relative levels of aggrecan mRNA within cells.

Twenty minutes of mechanical stimulation of normal human articular chondrocytes at 0.33 Hz results in a significant and reproducible membrane hyperpolarisation. In contrast, chondrocytes from osteoarthritic cartilage show a membrane depolarisation response. Previous studies have shown that these responses are the result of activation of a mechanotransduction pathway that is α5β1-integrin-dependent 38. In normal chondrocytes, mechanical signal results in activation of SK (small conductance) calcium-activated potassium channels, with potassium efflux and cell membrane hyperpolarisation. In osteoarthritic chondrocytes, the mechanotransduction pathway leads to activation of tetrodotoxin-sensitive channels, sodium influx, and cell membrane depolarisation. To establish whether CD47/IAP has roles in the response of articular chondrocytes to 0.33-Hz mechanical stimulation, chondrocytes were mechanically stimulated in the presence of the function-modifying anti-CD47/IAP antibodies Bric 126 and B6H12 (Table 1). The addition of Bric 126 and B6H12 to articular chondrocytes 10 minutes prior to the period of mechanical stimulation had no significant effect on the resting cell membrane potential. In the presence of Bric 126, both the membrane hyperpolarisation response of normal chondrocytes and the membrane depolarisation response of osteoarthritic chondrocytes to 0.33-Hz mechanical stimulation were blocked. When normal chondrocytes were mechanically stimulated in the presence of B6H12, an anti-CD47/IAP antibody that has partial agonist activity 2425, a membrane depolarisation response was seen.

Similar experiments were undertaken with chondrocytes from osteoarthritic cartilage and antibodies to TSP-1 and SIRPα (Table 1). Neither the anti-TSP-1 antibody nor the anti-SIRPα antibody SEAA5 had significant effects on the resting membrane potential of chondrocytes. When OA chondrocytes were mechanically stimulated at 0.33 Hz in the presence of the anti-TSP-1 antibody, no change in cell membrane potential was seen, indicating blockade of the mechanotransduction pathway that results in membrane depolarisation. In contrast, in the presence of SE5A5, an antibody against SIRPα, a statistically significant change in membrane potential – membrane depolarisation – was seen following mechanical stimulation. The degree to which the membrane potential changed following mechanical stimulation in the presence of SE5A5, however, was less than that seen in the absence of antibody, raising the possibility of a partial blocking effect.

Mechanical stimulation of normal human articular chondrocytes at 0.33 Hz results in increased tyrosine phosphorylation of a protein of approximately 125 kDa within 1 minute. This response is inhibited by the presence of anti-CD47/IAP monoclonal antibody Bric 126 (Figure 4a). Bric 126 has no direct effect on tyrosine phosphorylation in the absence of mechanical stimulation (results not shown).

Following mechanical stimulation at 0.33 Hz for 20 minutes, relative levels of aggrecan mRNA are elevated in normal articular chondrocytes at 1 and 3 hours after stimulation (Figure 4b; p < 0.05). In the presence of the anti-CD47/IAP monoclonal antibody Bric 126, no changes in the aggrecan mRNA ratio relative to the housekeeping gene GAPDH are seen following mechanical stimulation (Figure 4b; p > 0.05). Bric 126, in the absence of mechanical stimulation, had no effect on relative levels of aggrecan mRNA.

To ascertain whether CD47/IAP is required for mechanical signalling at different frequencies of stimulation, CD47/IAP-null cells (OV10, human carcinoma cell lines, and mouse CD47/IAP^-/- lung fibroblasts) and cells transfected with CD47/IAP type 2 DNA or vector controls were used and the electrophysiological membrane response was assessed. The results are shown in Table 2. 3657L cells, lung fibroblasts derived from wild-type CD47/IAP-expressing mice, showed a membrane depolarisation response when mechanically stimulated at 0.33 Hz. The same cells, however, showed a membrane hyperpolarisation response when mechanically stimulated at 0.083 Hz. 3656 1/Sra cells, CD47/IAP^-/- mouse lung fibroblasts derived from CD47/IAP knockout mice transfected with vector control, showed no change in cell membrane potential when mechanically stimulated at both 0.33 and 0.083 Hz. CD47/IAP^-/- mouse lung fibroblasts transfected with either mouse (3656 1/171) or human (3656 1/315) CD47/IAP type 2 DNA showed a membrane hyperpolarisation response following stimulation at 0.083 Hz and a membrane depolarisation response following mechanical stimulation at 0.33 Hz.

OV10 cells, derived from a CD47/IAP-negative human carcinoma, showed no significant change in membrane potential when stimulated at 0.083 Hz. Mechanical stimulation (0.083 Hz) of OV10 cells that had been transfected with human CD47/IAP alone or transfected with CD47/IAP in association with β3 integrin resulted in a significant membrane hyperpolarisation of these cells. In contrast, OV10 cells transfected with CD47/IAP extracellular IgV domain linked to a glycosylphosphatidylinositol anchor showed no significant change in cell membrane potential following 0.083-Hz stimulation. These cells, although expressing the extracellular IgV domain of CD47/IAP, do not express the transmembrane domain or the intracytoplasmic tail of the CD47/IAP, indicating that the extracellular domain by itself is insufficient for cellular recognition and response to the mechanical stimulus.