ARTICLE IN PRESS O. Roualdes et al. Biomaterials xxx(2009)1-12 Spray dried granules were pressed uniaxially (150 MPa)and isostatically(300 MPa) wells were recovered and stored at -70C for further experiments(see below )and in order to obtain medium and 100 uL of MIT (5 mg/mL in PBS, Gibco) Cells (Hot Isostatic Pressing) at 1520.C for 1 h, with a press p with shaking. In order to eliminate the particles that can They have been fully characterized in interfere with the optical density measurement, each well was centrifuged (5 min, to aging in previous works [10, 16, 17 Finally samples were mirror-polished using 1600 rpm)and the optical density was measured at 570 nm. diamond suspensions(Buhler, Lyon, France), either down to 1 um or only down to 30 um to give a rougher surface state 2.4. Extra-cellular matrix assay uman fibronectin(FN)and human type I collagen(CIH)contents in the culture 3. Cell culture medium were assessed by competition ELISA following standard procedures. Briefly nicro-well plates(96-well, Costar, Coming inc) were previously coated with 200 HL 2.3.1.cels of FN or CiH(respectively diluted at 1 300 and 1 /160 from a 1 mg/mL solution in PBS, man fibroblasts, obtained from healthy new born arm skin explants Biomerieux, Marcy I'Etoile, France) for 24 h at 4C then rinsed in PBS. For FN by Dr M. Cabot, Cent oakville, MD)were cultured in 25 cm" tissues flasks( Costar, Corning Inc )in the Lyon, France). After 1h30 at room temperature wells were rinsed in PBS and 200 uL ence of Dulbecco's modified eagle,'s medium(DMEM, Gibco) supplemented of peroxydase labeled anti-rabbit antibody diluted at 1/1000 were added for 1h30 at vere incubated at 37C, with 5% COz. saturated with humidity until confluent Inc onolayer and then sub-cultured after trypsination MG-63 osteoblast like cells are with 100 uL of CIH antibody diluted at 1/750 in PBS BS ALP activity in ere rinsed in pBs xtra-cellular matrix containing in particular collagen and nd 200 uL of peroxydase labeled anti-rabbit antibody d the expression of classical cell adhesion molecules such integrins 149, 50]. For for 1h30 at room temperature both cell types, experiments were carried out on cells from passage 4 through 30. After the second antibody incubation for both FN and CH assays, micro-well Using von Kossa staining v (named after"osteoblasts")to mineralize(data not shown). 202 were added f ength of 450 nm using a multiscan reader (Multiskan EX Thermo Scientific Electron Corporation). Intra ell culture(dry air, 180C for 4 h). After were lower than 10% for both FN and CIH ELISA. sterilization, ceramic substrates were deposited into a 24-well plate (co 2.5. Cell morphology (7C, 5% COz and wet atmosphere)for 2 h in order to pn hen 1.9 mL of culture medium was added into each well. the medium was ashed 2 times quickly with PBS(Gibco)in controls, cells were seeded directly on the plastic culture surfaces and paraformaldehyde in PBS for 1 h at room e the ceramic substrates massie Blue Brilliant(Sigma Aldrich) and observed using light microscopy. 2二 used for osteoblasts and fibroblasts experiments. 2.6. Animal experiment led onto 24-well culture plates(Costar, Corning Inc )and incubated using standard culture conditi ur male Sprag Dale sighing ading. After 24 (mean+ SEM) were used in this study. The animals were randomly assigned to 3 00 or 1000 ug/mL particles of each powder sterilized prior cell culture(dry groups. The first group received an intra-articular injection of sterile saline at day t 80. for 4 h). This time represented day O. The medium was changed at day 3, 6 The second and the third groups were injected three times with alur ely. Animals during the 10. The cell culture controls followed the same procedure but without particle addition into the medium. riod at the Institut Claude-Bourgelat, Lyon. All experiments were conducted in full ompliance with the National Veterinary School of Lyon(ENVL)Ethical Committee Guideline (p agreement n 0826)according to current European and french Legislations for Animal Protection. At 3, 6 and 10 days of culture, fibroblast and os by MIT assay (3-14,5-dimethylthiazol-2-yll-2, 5-dip um bromide, sigma Analgesia was provided before and after the operation with morphine (3 mg/kg drich) This reagent is transformed by mitochondrial dehydrogenases of active cells The animals were anesthetized with a mixture of isoflurane in oxygen. The right into formazan, providing a measure of cell activation and viability. The medium of al red for injection. Intra-articular injections were nade with 29G needle by a transpatellar approach We injected 50 uL of either saline for control, alumina or zirconia(0. 1% in volume )at 0, 2 and 4 weeks. A arlan re euthanasied two weeks after the last injection(6 weeks after the first injec- tion) by intracardiac injection of Dolethal euthanasia solution after anesthesic amine 90 mg/kg and xylazine 10 mg/kg, intramuscular injection Synovium and articular surface of femur, tibia and patella were removed and examined for gross evidence of abnormalities. 2.7. Histological The whole sy membrane and patella were removed from the joint and fixed saline for 72 h. The hen decalcified with EDTA 10% for 7 days before paraffin-embedding. 5 um sections from the saggital cut and stained with Goldner trichrome toluidine blue or were scored independently by two observers. A semi-quantitative three points scale tero represented the normal or the lowest state, one represented a slight Fig. 2. Scanning electron microscopy image of the composite after polishing and moderate increase and two the highest level of increase. The highest level of the thermal etching showing micron-size alumina grains(in grey)and zirconia sub-micron present score was much lower than that observed in pathologies such as osteoar particles(in white thritis or rheumatoid arthritis. Indeed, we adjusted our scale to the maximum Please cite this article in press as: Roualdes 0, et al, In vitro and in vivo evaluation of an, Biomaterials(2009). doi: 10.1016/ j biomaterials 2009. 11.107Spray dried granules were pressed uniaxially (150 MPa) and isostatically (300 MPa) in order to obtain small cylinders of 14 mm diameter and 2 mm thickness. Most of the samples were then sintered in air at 1520 C for 2 h, and then post treated by HIP (Hot Isostatic Pressing) at 1520 C for 1 h, with a pressure of 200 MPa. The final microstructure is shown in Fig. 2. All the processed samples reached full density. They have been fully characterized in terms of mechanical properties and re´ sistance to aging in previous works [10,16,17]. Finally samples were mirror-polished using diamond suspensions (Buhler, Lyon, France), either down to 1 mm or only down to 30 mm to give a rougher surface state. 2.3. Cell culture 2.3.1. Cells Human fibroblasts, obtained from healthy new born arm skin explants (provided by Dr MT. Zabot, Centre de Biotechnologie cellulaire, Lyon, France), and osteoblasts (human osteoblast like cell MG-63, American Type Culture Collection, Rockville, MD) were cultured in 25 cm2 tissues flasks (Costar, Corning Inc.) in the presence of Dulbecco’s modified eagle’s medium (DMEM, Gibco) supplemented with 10% foetal calf serum (Gibco) and 1% penicillin/streptomycin (Bayer). Cultures were incubated at 37 C, with 5% CO2, saturated with humidity until confluent monolayer and then sub-cultured after trypsination. MG-63 osteoblast like cells are characterized by an increased ALP activity in response to 1,25-(OH)2D3, the synthesis of an extra-cellular matrix containing in particular collagen and fibronectin [47,48], and the expression of classical cell adhesion molecules such integrins [49,50]. For both cell types, experiments were carried out on cells from passage 4 through 30. Using von Kossa staining we have checked the capacity of these osteoblasts-like cells (named after ‘‘osteoblasts’’) to mineralize (data not shown). 2.3.2. Cell culture on ceramics discs The ceramic discs were sterilized prior cell culture (dry air, 180 C for 4 h). After sterilization, ceramic substrates were deposited into a 24-well plate (Costar, Corning Inc.). 100 mL of culture medium containing 5 104 cells (fibroblasts or osteoblasts) were seeded upon each ceramic substrate. Plates were incubated in standard culture conditions (37 C, 5% CO2 and wet atmosphere) for 2 h in order to promote cell adhesion then 1.9 mL of culture medium was added into each well. The medium was changed at 3, 6 and 10 days. For the controls, cells were seeded directly on the plastic culture surfaces and treated like the ceramic substrates. 2.3.3. Cell culture in the presence of particles The same procedure was used for osteoblasts and fibroblasts experiments. 2.5 104 cells were seeded onto 24-well culture plates (Costar, Corning Inc.) and incubated using standard culture conditions in order to promote cell adhesion and spreading. After 24 h, the medium was replaced by 2 mL of fresh medium containing 100 or 1000 mg/mL particles of each powder sterilized prior cell culture (dry air, 180 C for 4 h). This time represented day 0. The medium was changed at day 3, 6 and 10. The cell culture controls followed the same procedure but without particles addition into the medium. 2.3.4. Cell proliferation At 3, 6 and 10 days of culture, fibroblast and osteoblast proliferation was assessed by MTT assay (3-[4,5-dimethylthiazol-2-yl]-2, 5-diphenyl tetrazolium bromide, Sigma Aldrich). This reagent is transformed by mitochondrial dehydrogenases of active cells into formazan, providing a measure of cell activation and viability. The medium of all wells were recovered and stored at 70 C for further experiments (see below) and replaced by 900 mL of fresh medium and 100 mL of MTT (5 mg/mL in PBS, Gibco). Cells culture plates were incubated in 5% (v/v) CO2 in air at 37 C for 3 h. The medium was removed and replaced by 500 mL of ethanol/DMSO solution (vol/vol) and incubated for 5 min at room temperature with shaking. In order to eliminate the particles that can interfere with the optical density measurement, each well was centrifuged (5 min, 1600 rpm) and the optical density was measured at 570 nm. 2.4. Extra-cellular matrix assay Human fibronectin (FN) and human type I collagen (CIH) contents in the culture medium were assessed by competition ELISA following standard procedures. Briefly micro-well plates (96-well, Costar, Corning inc) were previously coated with 200 mL of FN or CIH (respectively diluted at 1/300 and 1/160 from a 1 mg/mL solution in PBS, Biomerieux, Marcy l’Etoile, France) for 24 h at 4 C then rinsed in PBS. For FN assessment, 100 mL of standard point or medium sample were incubated with 100 mL of human FN antibody diluted at 1/1600 in PBS BSA 3% (24 911 antibody, Novotec, Lyon, France). After 1h30 at room temperature wells were rinsed in PBS and 200 mL of peroxydase labeled anti-rabbit antibody diluted at 1/1000 were added for 1h30 at room temperature. For CIH assessment, 100 mL of standard point or medium sample were incubated with 100 mL of CIH antibody diluted at 1/750 in PBS BSA 3% (Human type I collagen antibody 20111, Novotec). After 3 h at room temperature wells were rinsed in PBS and 200 mL of peroxydase labeled anti-rabbit antibody diluted at 1/800 were added for 1h30 at room temperature. After the second antibody incubation for both FN and CIH assays, micro-well plates were washed in PBS and 200 mL of o-Phenylenediamine dihydrochloride in H2O2 were added for 30 min on the dark at room temperature. Optical densities were obtained at the wavelength of 450 nm using a multiscan reader (Multiskan EX, Thermo Scientific Electron Corporation). Intra and inter-assay reproductibilities were lower than 10% for both FN and CIH ELISA. 2.5. Cell morphology Cells were cultured in the presence of particles at 100 mg/mL in DMEM as described previously (paragraph 2.3.3). At 2, 6, 24 h and 3, 6 10 and 15 days of culture, cells were washed 2 times quickly with PBS (Gibco) in order to remove particles then fixedwith 4% paraformaldehyde in PBS for 1 h at room temperature. Cells were stained with Coo￾massie Blue Brilliant (Sigma Aldrich) and observed using light microscopy. 2.6. Animal experiment Thirty-four male Sprage Dawley rats 4 months old and weighing 479  38 g (mean  SEM) were used in this study. The animals were randomly assigned to 3 groups. The first group received an intra-articular injection of sterile saline at day 0. The second and the third groups were injected three times with alumina and zirconia particles respectively. Animals were housed in groups during the same period at the Institut Claude-Bourgelat, Lyon. All experiments were conducted in full compliance with the National Veterinary School of Lyon (ENVL) Ethical Committee Guideline (protocol agreement n 0826) according to current European and French Legislations for Animal Protection. Analgesia was provided before and after the operation with morphine (3 mg/kg). The animals were anesthetized with a mixture of isoflurane in oxygen. The right stifle joint was clipped and prepared for injection. Intra-articular injections were made with 29G needle by a transpatellar approach. We injected 50 mL of either saline for control, alumina or zirconia (0.1% in volume) at 0, 2 and 4 weeks. A veterinarian closely monitored the animals for infections and other complications. The animals were euthanasied two weeks after the last injection (6 weeks after the first injec￾tion) by intracardiac injection of Dolethal euthanasia solution after anesthesic protocol (ketamine 90 mg/kg and xylazine 10 mg/kg, intramuscular injection). Synovium and articular surface of femur, tibia and patella were removed and examined for gross evidence of abnormalities. 2.7. Histological preparation and examination The whole synovial membrane and patella were removed from the joint and fixed in neutral-buffered formal saline for 72 h. They were then decalcified with EDTA 10% for 7 days before paraffin-embedding. 5 mm sections from the saggital plane of the synovium was cut and stained with Goldner trichrome, toluidine blue or hematoxilin eosine safranin. Morphological studies were performed using a light microscope (Zeiss Axioskop 50 Germany, Oberkochen, software Explora Nova Morpho Expert, La Rochelle, France). All areas of each biopsy section were examined and histological features were scored independently by two observers. A semi-quantitative three points scale was adapted from previously used scoring systems for synovial tissue [44,45,51–55]. Zero represented the normal or the lowest state, one represented a slight or moderate increase and two the highest level of increase. The highest level of the present score was much lower than that observed in pathologies such as osteoar￾thritis or rheumatoid arthritis. Indeed, we adjusted our scale to the maximum Fig. 2. Scanning electron microscopy image of the composite after polishing and thermal etching showing micron-size alumina grains (in grey) and zirconia sub-micron particles (in white). O. Roualdes et al. / Biomaterials xxx (2009) 1–12 3 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107