Tenascin-C induces inflammatory mediators and matrix degradation in osteoarthritic cartilage

Specimens of non-OA cartilage obtained from human donors with consent had no history of joint disorders; they were received within 36 hours following autopsy (National Disease Research Interchange, Philadelphia, PA), examined macroscopically and smooth intact surfaces without OA-like lesions were harvested for protein extraction and RNA preparation. The average age of the 7 non-OA donors was 43 years with an age range of 38-58 years. Specimens of OA cartilage with visible lesions were obtained with consent from patients undergoing knee replacement surgery at New England Baptist Hospital, and harvested within a few hours of surgery. The average age of the 7 OA cartilage donors was 68 years with an age range of 50-82 years. This study was performed under the approval of Pfizer's Institutional Human Ethics Committee. Cartilage slices harvested under sterile conditions were cut into explants (6-8 mm width), rinsed three times in PBS, and flash frozen. Cartilage (1 g each) was pulverized in a Spex Certiprep freezer mill Model 6750 (15 Hz; 2 × 1 min with 1 min pause between cycles) under liquid nitrogen for protein extraction and RNA preparation. RNA was prepared from pulverized cartilage as described [21]. For protein extraction, the powdered cartilage was immediately suspended in 10 ml of 4 M guanidine HCl, 50 mM sodium acetate pH 5.8 containing protease inhibitor cocktail (Calbiochem, Gibbstown, NJ) and extracted for 48 hours at 4°C on a rotator. The mixture was then centrifuged at 3,000 rpm for 10 min and the supernatant dialyzed against 20 mM Tris-HCl, pH 8.2 (with protease inhibitors) overnight at 4°C. OA and non-OA cartilage extracts (5 μg protein equivalent) were deglycosylated with 0.15 U/ml chondroitinase ABC (Sigma, St Louis, MO), 0.15 U/ml Keratanase I and 0.0075 U/ml Keratanase II (Seikagaku America, Falmouth, MA) at 37°C for 3 hours. The samples were separated on a 3-8% Tris-Acetate gel (Invitrogen, Carlsbad, CA), transferred to nitrocellulose membrane and probed with anti-human Tenascin-C (EGF-like domain) antibody 4F10TT (Immuno-Biological Laboratories Co., Ltd. Japan) at 1:100 dilution followed by incubation in anti-mouse IgG (H&L) conjugated to alkaline phosphatase (Promega, Madison, WI) at 1:3000 dilution. Detection of reactive bands was performed with NBT/BCIP substrate (Promega). Purified human TN-C protein (Millipore, Billerica, MA) was used as a positive control in the Western blot analysis. The blots were also probed with secondary antibody alone to confirm specificity of detection.

Bovine and human primary chondrocytes were prepared under sterile conditions by pronase and collagenase treatments followed by filtration and centrifugation as previously described [22]. Cells were washed, resuspended in DMEM-F12 (Gibco, Carlsbad, CA), 10% FBS (Sigma, St. Louis, MO), 1% antimycotic-antibiotic solution (Sigma), and counted on a hemocytometer. Cell viability was determined by trypan blue dye exclusion, cell viability was found to be >95%. Cells were plated at 1 million/well in a 24 well tissue culture plate (1 ml/well) and maintained at 37°C. The cells were serum starved overnight once they were confluent (in 3-4 days), and washed with serum free media before induction. LPS from E. coli R515 (Re) (Axxora, San Diego, CA) at 0 to 1000 ng/ml or TN-C protein at 0 to 10 μg/ml was added and incubated for 48 hours at 37°C to study dose-dependent induction of primary chondrocytes. Heat killed TN-C that was heated at 100°C for 30 min, and LPS preincubated for 1 hour with polymyxin-B (PMB; Sigma, St. Louis, MO) served as negative controls for TN-C and LPS treatment, respectively. TN-C at 10 μg/ml preincubated with 3 μg/ml PMB was also tested to confirm that the induction effects observed with TN-C were not endotoxin related. TAK242, a specific TLR4 inhibitor [23], was synthesized at Pfizer. For TAK242 treatment, the cells were pretreated with inhibitor alone for 2 hours prior to induction with 1000 ng/ml LPS or 10 μg/ml TN-C in the presence of inhibitor. The media was removed after 48 hours of induction and analyzed in IL-6, nitrate, and PGE₂ assays. Cells treated with 1000 ng/ml LPS, 10 μg/ml TN-C or 5 ng/ml IL-1β with or without TAK242 for 48 hours were washed in PBS, and lysed in lysis buffer for RNA preparation using RNAeasy kit (Qiagen, Valencia, CA) following the manufacturer's protocol.

Articular cartilage explant discs were harvested under sterile conditions from young bovine metacarpal phalangeal joints (Research 87, Hopkinton, MA). Briefly, full-thickness plugs were punched using a 8 mm cork borer and cartilage discs were generated by slicing 1 mm thick sections from the articular surface of the plugs. Discs were rinsed in PBS and subsequently cultured in medium. The medium consisted of Dulbecco's Modified Eagle's medium (JRH Biosciences, Lenexa, KS), 50 μg/ml ascorbic acid (Wako, Osaka, Japan), 10 mM HEPES (Mediatech, Herndon, VA), 2 mM L-glutamine (Mediatech), antibiotic-antimycotic solution (Sigma). Discs were cultured for 5 days with one media change in a 37°C and 5% CO₂ environment to equilibrate the tissue prior to treatment. Following equilibration, 3 discs (~100 mg/well) were weighed and placed in 24-well tissue culture plate in 1 ml medium with or without 1 or 10 ng/ml of IL-1α (Sigma, St Louis, MO) for 48 hours for the first study. The media was tested for TN-C levels, and RNA prepared from cartilage discs for TN-C taqman analysis. For the second study, explants were treated with 5 ng/ml IL-1α, 10 μg/ml TN-C, or 1000 ng/ml LPS with or without TAK242 (0.01, 0.1, or 1 μM). For TAK242 effects, explants were pre-treated with the inhibitor for 2 hours prior to induction in the presence of inhibitor. The media was removed for the analysis of proteoglycan release after 48 hours of induction.

Neat human knee joint synovial fluids from patients with end stage osteoarthritis (OA; n = 8: 4 males and 4 females; 50-83 years old) were obtained from NEBH, and synovial fluids from knee-healthy reference subjects (Ref; n = 8: 5 males and 3 females; 45-56 years old) were from NDRI or Northland labs with patient consent. The OA group included 7 synovial fluids of the same donors from whom cartilage samples were used for TN-C protein and mRNA expression. Representative OA and reference synovial fluids (15 μl each) from the above set were treated with 10 U of hyaluronidase (Sigma) at RT overnight and subjected to Western blot analysis with anti-human Tenascin-C (EGF-like domain) antibody 4F10TT as described above for cartilage extracts. The blots were probed with secondary antibody alone to confirm specificity of detection.

Male Lewis rats weighing approximately 300 grams were obtained from Charles River Laboratories (Wilmington, MA). The rats underwent medial meniscal surgery in the right knee to induce joint instability leading to cartilage degeneration as described [24]. The animals were euthanized at different times after surgery (4 days, 1 wk, 2 wks, and 3 wks; n = 10 animals/time point). Synovial fluid lavages and serum were collected. Five naïve animals per time point were also included. Serum and synovial fluid lavage urea levels in each rat were used to correct TN-C, proteoglycan, and ARG-aggrecan values for dilution. This study was performed under the approval of Pfizer's Institutional Animal Care and Use Committee.

TN-C was measured in cartilage extracts, conditioned media, and synovial fluid samples using the TN-C Large ELISA kit (Immuno-Biological Laboratories Co., Ltd. Japan). The ELISA uses anti-TN-C 19C4MS monoclonal antibody against the FNIII-C domain for capture, and HRP conjugated anti-TN-C 4F10TT mouse monoclonal antibody against the EGF domain for detection. 4F10TT binds an epitope from the EGF domain and recognizes both the small and large TN-C variants. 19C4MS binds an epitope of the FNIII-C domain and recognizes large variants. The characteristics of these antibodies have been described elsewhere [15, 25]. TN-C standard in the kit was run at 0-24 ng/ml for a standard curve. Samples were appropriately diluted in PBS and assayed in the TN-C ELISA using manufacturer's protocol. TN-C standard or human synovial fluid samples incubated in PBS- or mouse IgG-coated wells were included as controls. To confirm specificity of TN-C detection, aggrecan, a major cartilage proteoglycan purified from human cartilage as described [26] was tested at various concentrations in the assay as a negative control. To further confirm specificity of detection in synovial fluid, two human synovial fluids were immunodepleted of TN-C using anti-TN-C 4C8MS monoclonal antibody (Immuno-Biological Laboratories Co.) against the FNIII-B domain [27], or anti-human TN-C BC-24 (Abcam, Cambridge, MA) against the EGF domain, and then analyzed in the ELISA. Protein-G Dynabeads (Invitrogen) were used following manufacturer's protocol for immunoprecipitation (20 μg antibody; 200 μl of synovial fluid), Mouse IgG was used as a negative control in immunodepletion experiments. In order to determine spike-in recovery of TN-C, two human synovial fluids diluted to 1:100, 1:200, or 1:400 were spiked in with TN-C standard at a final concentration of 5 or 10 ng/ml and analyzed in the ELISA.

Protein was quantified using the microplate Bradford protein assay (Thermo Scientific, Rockford, IL). Cell toxicity was determined in primary cell and explant cultures by measuring lactate in the conditioned media using a lactate assay (Trinity biotech, St. Louis, MO). Prostaglandin E₂ (PGE₂) release was measured using a PGE₂ ELISA (R&D Systems, Minneapolis MN). Measurement of nitrate concentrations was performed using a nitrate/nitrite colorimetric assay kit (Cayman Chemical Company, Ann Arbor, MI). Human chondrocyte conditioned media were screened using a human proinflammatory 7-plex MSD tissue culture kit (Meso Scale Discovery, Gaithersburg, Maryland). Human IL-6 and IL-8 were measured individually using MSD human cytokine assay tissue culture kits (Meso Scale Discovery). The proteoglycan content in bovine explant conditioned media was measured as sulfated glycosaminoglycan (sGAG) by a colorimetric assay with dimethylmethylene blue (Serva, Heidelberg, Germany) [28]. Proteoglycan levels in human synovial fluids were determined by the sGAG assay (Kamiya Biomedical Company, Seattle, WA) [29]. ARG-aggrecan fragments in synovial fluids were measured in an ELISA developed at Pfizer [26].

Taqman gene expression assays were done using one-step RT-PCR reagents and Assay on Demand primer-probe sets (Applied BioSystems, Foster City, CA) following manufacturer's protocol. For analyzing bovine samples, GAPDH (Bt03210913_g1), and ADAMTS4 (Bt03224693_m1) primer/probe sets were used. For the human samples, GAPDH (Hs99999905_m1), ADAMTS4 (Hs00192708_m1), ADAMTS5 (Hs00199841_m1), and TN-C (Hs01115665_m1) primer/probe sets were used. 100 ng RNA per sample was tested in duplicates and results averaged.