2.1

Substrate design and surface analysis

Commercially pure titanium and yttria-stabilized tetragonal zirconia polycrystal (Y-TZP) disks (20 mm in diameter and 1.5 mm in thickness) and miniature cylindrical implants (2 mm in length, 1 mm in diameter) were fabricated for this study. The titanium surface was either turned by machining (Ti-m) or roughened by electrochemical anodization (TiUnite ^® , Nobel Biocare, Gothenburg, Sweden). The zirconia surface was either turned by machining (TZP-A-m) or roughened by sandblasting with Al ₂ O ₃ (grain size 30–130 μm) at 4–6 bar and acid-etching with hydrofluoric acid, nitric acid, and sulfuric acid with subsequent heat treatment (1200–1400 °C) in order to smoothen the sharp edges as a result of the etching procedure (TZP-proc, Patent pending, Patent application number: 20,100,178,636, VITA Zahnfabrik, Bad Säckingen, Germany). All materials were sterilized using low-temperature hydrogen peroxide gas plasma technology (STERRAD 100/100S, Advanced Sterilization Products (A.S.P), Johnson & Johnson Medical, Irvine, USA).

The surface of the substrates was examined by scanning electron microscopy (SEM) (Zeiss Leo 32, Zeiss, Oberkochen, Germany) after being gold-palladium-sputtered, and by 3D laser scanning (3D Laser Microscope VK-9700K, Keyence Corp., Osaka, Japan). Average roughness (S _a ), the values of peak-to-valley (S _z ), average mean spacing of profile peaks in the mean plane as expressed in the x direction (S _cx ), and the developed surface area ratio (S _dr ) were calculated.

Raman spectroscopy (inVia Raman Microscope, Renishaw, Gloucestershire, UK) was used to detect the effect of the surface treatment on the Al ₂ O ₃ content and potential tetragonal to monoclinic phase transformations in the zirconia-based ceramic.

2.2

Cell culture

The human osteoblast cell line hFOB 1.19 (ATCC, LGC Promochem, Wesel, Germany) was used for the cell culture experiment. These cells show minimal chromosome abnormalities and synthesize a normal spectrum of matrix proteins .

The cells were cultured according to the ATCC recommendations at 33.5 °C in a 1:1 mixture of Dulbecco’s Minimal Essential Medium (high glucose) and Hams F12 (Lonza, Verviers, Belgium) with 15 mM Hepes (Biochrom, Berlin, Germany), 0.5 mM Na Pyruvat, 1.2 g/l Na bicarbonate, 0.3 mg/ml G418, 10% fetal calf serum (FCS; Biochrom, Berlin, Germany), 100 U/ml penicillin and 10 mg/ml streptomycin (Biochrom, Berlin, Germany). Cells from passage 5–10 were used for all experiments. For cell proliferation h.FOB 1.19 cells were cultured at 33.5 °C in a 5% CO ₂ humified atmosphere. When cells reached confluence of approximately 80%, they were trypsinated and finally seeded at a density of 1.3 × 10 ⁴ cells onto the test substrates in 12-well plates. The medium was exchanged every third day for up to 7 days. From day 7 until day 28, the medium was exchanged every second day. Since the hFOB 1.19 cell line is immortalized with a temperature-sensitive SV40 large T antigen, which can be inactivated at 39.5 °C, cell division is slowly ceased at 39.5 °C and cell differentiation is induced. In our experimental setup cell differentiation was induced in confluent cultures by raising the temperature to 39.5 °C.

2.3

Cell proliferation

Cell proliferation was measured with the EZ4U kit (EASY FOR YOU, Biozol diagnostica Corp., Eching, Germany). The assay is based on the mitochondrial reduction of tetrazolium in living cells and the subsequent release of the reduction product formazan into the culture medium. Formazan can be then quantified in the supernatant by spectrophotometrical measurement at 450 nm in a microplate reader. The assay was performed according to the manufacturer’s protocol. Cell growth on polystyrene was used as control. All experiments were performed in triplicate after 1, 3, 7, 14, 21, and 28 days.

2.4

Confocal laser scanning microscopy

Cell surface contact and cytoskeletal arrangement of the osteoblasts on the different substrates were examined by indirect immunofluorescence of the focal adhesion protein vinculin and the cytoskeleton protein actin with a commercial staining kit (Millipore, Billerica, USA). For this purpose, cells were cultured for 1 and 28 days on the different substrates and fixed for 15 min in 4% paraformaldehyde. After fixation, the hFOB 1.19 cells were rinsed with PBS, permeabilized for 5 min with 0.1% TritonX-100 (Sigma Diagnostics, St. Louis, USA) in PBS and incubated overnight with anti-vinculin (mouse anti-human antibody) in a dilution of 1:500. After 24 h incubation with the primary antibody, probes were rinsed 3 times with wash buffer (PBS, containing 0.05% Tween-20) and incubated with a FITC-conjugated anti-mouse antibody (green) in a dilution of 1:1000. Actin was detected with TRIC-coupled phalloidin (red) in a dilution of 1:500 after an incubation time of 60 min. The nucleoli of the cells were counterstained with DAPI (blue) for 5 min. Images were acquired using a confocal laser scanning microscope (Leica TCS SP2 AOBS, Leica Microsystems Bensheim, Germany).

2.5

Gene expression analysis

Gene expression of several bone-specific markers ( Table 1 ) was analyzed by semi-quantitative real-time RT-PCR. For this purpose total mRNA from hFOB 1.19 cells was isolated after 3, 7, 14, 21, and 28 days of culture on the different materials using the RNeasy MiniKit (Qiagen, Hilden, Germany). Integrity and concentration of the RNA were determined using a lab on a chip RNA 6000 Nano Series II kit (Agilent Technologies, Waldbronn, Germany) that was run on an Agilent 2100 Bioanalyzer instrument (Agilent Technologies). Reverse transcription was performed with 5 μg total RNA.

The cells cultured on the discs were harvested by trypsinization and washed with PBS. The primer sequences of selected genes were determined with the “Universal Probe Library” software from Roche ( http://qpcr.probefinder.com/organism.jsp ). Only intron-spanning sequences were chosen in order to exclude contamination with genomic DNA. Real-time RT-PCR was performed in a Light Cycler 489 (Roche, Basel, Switzerland) in a 384 plate. A 10 μl volume with 5 μl SYBR Green I Mastermix (Roche), 50 ng cDNA as template and 0.5 μM of the primer-pairs were used respectively.

Light cycling conditions were as follows: activation (95 °C for 10 s), 40 amplification cycles (95 °C for 10 s, 52 °C for 5 s, and 72 °C for 12 s). Melting curve analysis was used to ensure that all transcripts under investigation were represented by a single peak, indicating specificity. Gene expression was calculated from the real-time RT-PCR efficiency in relation to the mean of five housekeeping genes (LMNA, GAPDH, ACTB, RPS18, and RPLP0). Three independent cultures were used for real-time RT-PCR measurements.

2.6

Placement of experimental implants in the rat femur

Fifty-six 8-week-old male Sprague-Dawley rats were anesthetized with 1–2% isoflurane inhalation. After the legs were shaved and disinfected with 0.2% chlorhexidine, the distal parts of the femurs were exposed. One zirconia implant with a sandblasted and acid-etched surface (TZP-proc) and one titanium implant with an electrochemical anodized surface (TiUnite ^® ) were randomly placed into the left and right femurs, respectively, of 28 rats. The remaining 28 rats received one zirconia implant with a turned surface (TZP-A-m) and one titanium implant with a turned surface (Ti-m) which were also randomly placed into left and right femurs, respectively. The implant site was prepared at 7–9 mm from the distal edge of the femur using a 0.8 mm round burr and reamers (ISO 90). Irrigation with sterile saline solution was used for cooling and cleaning. The implants were installed into the osteotomy and pushed into place until they were even with the femoral bone. The tissues were then closed with resorbable sutures (Vicryl ^® , Ethicon GmbH, Norderstedt, Germany). Animals with TiUnite ^® and TZP-proc implants were randomly divided into four groups of seven each. Two groups were killed at week 2 and two groups were killed at week 4 of the healing period. Animals with Ti-m and TZP-A-m implants were also divided into four groups of seven each. Again, two groups were killed at week 2 and two groups were killed at week 4 of the healing period. One of the two groups killed at week 2 and of the two groups killed at week 4 was used for histologic evaluation. The remaining groups were used for the implant push-in test.

The study protocol was approved by the University of Freiburg Animal Research Committee (Study No. G-04/60) and was conducted according to the German Federal Guidelines for Animal Research.

2.7

Histologic procedure and histomorphometric analysis

Bone segments containing the implant were harvested and the specimens were thoroughly rinsed with saline and immersed in 10% buffered formalin for 2 weeks at 4 °C. The specimens were then dehydrated in alcohol and finally embedded in photocuring resin (Technovit 7200 VLC, Heraeus Kulzer, Wehrheim, Germany). The non-decalcified embedded specimens were cut using a diamond saw and successively ground to a thickness of approximately 80–100 μm with a grinding system (Exakt Apparatebau, Norderstedt, Germany) . The histologic specimens were then stained with basic fuchsine and examined by light microscopy (Zeiss Axioskop, Zeiss, Oberkochen, Germany).

Histologic examination and computer-assisted histomorphometric analysis were performed at 20× and 40× magnification using a light microscope (Zeiss Axioskop), equipped with a video camera (Color View III, Olympus) and the software program cell* (Olympus). Histomorphometric analysis comprised the evaluation of the fraction of the implant in contact to the mineralized bone tissue over the entire implant length.

2.8

Implant push-in test

A push-in test was performed to assess the biomechanical strength of bone-implant integration .

Bone segments containing the implant were harvested, immediately embedded in autopolymerizing resin (Technovit 4071, Heraeus Kulzer, Wehrheim, Germany), and loaded axially in a universal testing machine (Zwick, Ulm, Germany). For the loading a 0.8 mm diameter stainless steel pushing rod was used in a 2000 N load cell and a cross-head speed of 1 mm/min. The applied load and the displacement of the implant were monitored at a sampling rate of 4 Hz. The maximum load prior to a rapid decrease in the load-displacement curve was regarded as the push-in value.

2.9

Statistics

A two-way analysis of variance (ANOVA) followed by one-way ANOVA at p < 0.05 level of significance was applied to evaluate the effects of substrate types and culture times on cell proliferation and gene expression. Bonferroni multiple comparison was used for post hoc testing.

Data of the in vivo experiments (histological evaluation and the push-in test) were expressed as mean values ± standard deviations. For statistical inference, each animal was considered as a cluster because data were collected on two femurs per animal. A repeated measures analysis of variance (ANOVA) has been used to take the within-animal dependence into account. The model assumption, i.e. normal distribution of the residuals, has been checked by looking at the histograms and normal probability plots. Non-normality could not be detected in the data (Kolmogorov–Smirnov test: p -value >0.15). The group effects and differences of least-square means were calculated with their 95% confidence intervals. The p -values for the pair wise comparison were adjusted by the Tukey–Kramer method. All calculations have been performed using PROC MIXED and others of the statistical software SAS 9.1.2.

2.1

Substrate design and surface analysis

Commercially pure titanium and yttria-stabilized tetragonal zirconia polycrystal (Y-TZP) disks (20 mm in diameter and 1.5 mm in thickness) and miniature cylindrical implants (2 mm in length, 1 mm in diameter) were fabricated for this study. The titanium surface was either turned by machining (Ti-m) or roughened by electrochemical anodization (TiUnite ^® , Nobel Biocare, Gothenburg, Sweden). The zirconia surface was either turned by machining (TZP-A-m) or roughened by sandblasting with Al ₂ O ₃ (grain size 30–130 μm) at 4–6 bar and acid-etching with hydrofluoric acid, nitric acid, and sulfuric acid with subsequent heat treatment (1200–1400 °C) in order to smoothen the sharp edges as a result of the etching procedure (TZP-proc, Patent pending, Patent application number: 20,100,178,636, VITA Zahnfabrik, Bad Säckingen, Germany). All materials were sterilized using low-temperature hydrogen peroxide gas plasma technology (STERRAD 100/100S, Advanced Sterilization Products (A.S.P), Johnson & Johnson Medical, Irvine, USA).

The surface of the substrates was examined by scanning electron microscopy (SEM) (Zeiss Leo 32, Zeiss, Oberkochen, Germany) after being gold-palladium-sputtered, and by 3D laser scanning (3D Laser Microscope VK-9700K, Keyence Corp., Osaka, Japan). Average roughness (S _a ), the values of peak-to-valley (S _z ), average mean spacing of profile peaks in the mean plane as expressed in the x direction (S _cx ), and the developed surface area ratio (S _dr ) were calculated.

Raman spectroscopy (inVia Raman Microscope, Renishaw, Gloucestershire, UK) was used to detect the effect of the surface treatment on the Al ₂ O ₃ content and potential tetragonal to monoclinic phase transformations in the zirconia-based ceramic.

2.2

Cell culture

The human osteoblast cell line hFOB 1.19 (ATCC, LGC Promochem, Wesel, Germany) was used for the cell culture experiment. These cells show minimal chromosome abnormalities and synthesize a normal spectrum of matrix proteins .

The cells were cultured according to the ATCC recommendations at 33.5 °C in a 1:1 mixture of Dulbecco’s Minimal Essential Medium (high glucose) and Hams F12 (Lonza, Verviers, Belgium) with 15 mM Hepes (Biochrom, Berlin, Germany), 0.5 mM Na Pyruvat, 1.2 g/l Na bicarbonate, 0.3 mg/ml G418, 10% fetal calf serum (FCS; Biochrom, Berlin, Germany), 100 U/ml penicillin and 10 mg/ml streptomycin (Biochrom, Berlin, Germany). Cells from passage 5–10 were used for all experiments. For cell proliferation h.FOB 1.19 cells were cultured at 33.5 °C in a 5% CO ₂ humified atmosphere. When cells reached confluence of approximately 80%, they were trypsinated and finally seeded at a density of 1.3 × 10 ⁴ cells onto the test substrates in 12-well plates. The medium was exchanged every third day for up to 7 days. From day 7 until day 28, the medium was exchanged every second day. Since the hFOB 1.19 cell line is immortalized with a temperature-sensitive SV40 large T antigen, which can be inactivated at 39.5 °C, cell division is slowly ceased at 39.5 °C and cell differentiation is induced. In our experimental setup cell differentiation was induced in confluent cultures by raising the temperature to 39.5 °C.

2.3

Cell proliferation

Cell proliferation was measured with the EZ4U kit (EASY FOR YOU, Biozol diagnostica Corp., Eching, Germany). The assay is based on the mitochondrial reduction of tetrazolium in living cells and the subsequent release of the reduction product formazan into the culture medium. Formazan can be then quantified in the supernatant by spectrophotometrical measurement at 450 nm in a microplate reader. The assay was performed according to the manufacturer’s protocol. Cell growth on polystyrene was used as control. All experiments were performed in triplicate after 1, 3, 7, 14, 21, and 28 days.

2.4

Confocal laser scanning microscopy

Cell surface contact and cytoskeletal arrangement of the osteoblasts on the different substrates were examined by indirect immunofluorescence of the focal adhesion protein vinculin and the cytoskeleton protein actin with a commercial staining kit (Millipore, Billerica, USA). For this purpose, cells were cultured for 1 and 28 days on the different substrates and fixed for 15 min in 4% paraformaldehyde. After fixation, the hFOB 1.19 cells were rinsed with PBS, permeabilized for 5 min with 0.1% TritonX-100 (Sigma Diagnostics, St. Louis, USA) in PBS and incubated overnight with anti-vinculin (mouse anti-human antibody) in a dilution of 1:500. After 24 h incubation with the primary antibody, probes were rinsed 3 times with wash buffer (PBS, containing 0.05% Tween-20) and incubated with a FITC-conjugated anti-mouse antibody (green) in a dilution of 1:1000. Actin was detected with TRIC-coupled phalloidin (red) in a dilution of 1:500 after an incubation time of 60 min. The nucleoli of the cells were counterstained with DAPI (blue) for 5 min. Images were acquired using a confocal laser scanning microscope (Leica TCS SP2 AOBS, Leica Microsystems Bensheim, Germany).

2.5

Gene expression analysis

Gene expression of several bone-specific markers ( Table 1 ) was analyzed by semi-quantitative real-time RT-PCR. For this purpose total mRNA from hFOB 1.19 cells was isolated after 3, 7, 14, 21, and 28 days of culture on the different materials using the RNeasy MiniKit (Qiagen, Hilden, Germany). Integrity and concentration of the RNA were determined using a lab on a chip RNA 6000 Nano Series II kit (Agilent Technologies, Waldbronn, Germany) that was run on an Agilent 2100 Bioanalyzer instrument (Agilent Technologies). Reverse transcription was performed with 5 μg total RNA.

The cells cultured on the discs were harvested by trypsinization and washed with PBS. The primer sequences of selected genes were determined with the “Universal Probe Library” software from Roche ( http://qpcr.probefinder.com/organism.jsp ). Only intron-spanning sequences were chosen in order to exclude contamination with genomic DNA. Real-time RT-PCR was performed in a Light Cycler 489 (Roche, Basel, Switzerland) in a 384 plate. A 10 μl volume with 5 μl SYBR Green I Mastermix (Roche), 50 ng cDNA as template and 0.5 μM of the primer-pairs were used respectively.

Light cycling conditions were as follows: activation (95 °C for 10 s), 40 amplification cycles (95 °C for 10 s, 52 °C for 5 s, and 72 °C for 12 s). Melting curve analysis was used to ensure that all transcripts under investigation were represented by a single peak, indicating specificity. Gene expression was calculated from the real-time RT-PCR efficiency in relation to the mean of five housekeeping genes (LMNA, GAPDH, ACTB, RPS18, and RPLP0). Three independent cultures were used for real-time RT-PCR measurements.

2.6

Placement of experimental implants in the rat femur

Fifty-six 8-week-old male Sprague-Dawley rats were anesthetized with 1–2% isoflurane inhalation. After the legs were shaved and disinfected with 0.2% chlorhexidine, the distal parts of the femurs were exposed. One zirconia implant with a sandblasted and acid-etched surface (TZP-proc) and one titanium implant with an electrochemical anodized surface (TiUnite ^® ) were randomly placed into the left and right femurs, respectively, of 28 rats. The remaining 28 rats received one zirconia implant with a turned surface (TZP-A-m) and one titanium implant with a turned surface (Ti-m) which were also randomly placed into left and right femurs, respectively. The implant site was prepared at 7–9 mm from the distal edge of the femur using a 0.8 mm round burr and reamers (ISO 90). Irrigation with sterile saline solution was used for cooling and cleaning. The implants were installed into the osteotomy and pushed into place until they were even with the femoral bone. The tissues were then closed with resorbable sutures (Vicryl ^® , Ethicon GmbH, Norderstedt, Germany). Animals with TiUnite ^® and TZP-proc implants were randomly divided into four groups of seven each. Two groups were killed at week 2 and two groups were killed at week 4 of the healing period. Animals with Ti-m and TZP-A-m implants were also divided into four groups of seven each. Again, two groups were killed at week 2 and two groups were killed at week 4 of the healing period. One of the two groups killed at week 2 and of the two groups killed at week 4 was used for histologic evaluation. The remaining groups were used for the implant push-in test.

The study protocol was approved by the University of Freiburg Animal Research Committee (Study No. G-04/60) and was conducted according to the German Federal Guidelines for Animal Research.

2.7

Histologic procedure and histomorphometric analysis

Bone segments containing the implant were harvested and the specimens were thoroughly rinsed with saline and immersed in 10% buffered formalin for 2 weeks at 4 °C. The specimens were then dehydrated in alcohol and finally embedded in photocuring resin (Technovit 7200 VLC, Heraeus Kulzer, Wehrheim, Germany). The non-decalcified embedded specimens were cut using a diamond saw and successively ground to a thickness of approximately 80–100 μm with a grinding system (Exakt Apparatebau, Norderstedt, Germany) . The histologic specimens were then stained with basic fuchsine and examined by light microscopy (Zeiss Axioskop, Zeiss, Oberkochen, Germany).

Histologic examination and computer-assisted histomorphometric analysis were performed at 20× and 40× magnification using a light microscope (Zeiss Axioskop), equipped with a video camera (Color View III, Olympus) and the software program cell* (Olympus). Histomorphometric analysis comprised the evaluation of the fraction of the implant in contact to the mineralized bone tissue over the entire implant length.

2.8

Implant push-in test

A push-in test was performed to assess the biomechanical strength of bone-implant integration .

Bone segments containing the implant were harvested, immediately embedded in autopolymerizing resin (Technovit 4071, Heraeus Kulzer, Wehrheim, Germany), and loaded axially in a universal testing machine (Zwick, Ulm, Germany). For the loading a 0.8 mm diameter stainless steel pushing rod was used in a 2000 N load cell and a cross-head speed of 1 mm/min. The applied load and the displacement of the implant were monitored at a sampling rate of 4 Hz. The maximum load prior to a rapid decrease in the load-displacement curve was regarded as the push-in value.

2.9

Statistics

A two-way analysis of variance (ANOVA) followed by one-way ANOVA at p < 0.05 level of significance was applied to evaluate the effects of substrate types and culture times on cell proliferation and gene expression. Bonferroni multiple comparison was used for post hoc testing.

Data of the in vivo experiments (histological evaluation and the push-in test) were expressed as mean values ± standard deviations. For statistical inference, each animal was considered as a cluster because data were collected on two femurs per animal. A repeated measures analysis of variance (ANOVA) has been used to take the within-animal dependence into account. The model assumption, i.e. normal distribution of the residuals, has been checked by looking at the histograms and normal probability plots. Non-normality could not be detected in the data (Kolmogorov–Smirnov test: p -value >0.15). The group effects and differences of least-square means were calculated with their 95% confidence intervals. The p -values for the pair wise comparison were adjusted by the Tukey–Kramer method. All calculations have been performed using PROC MIXED and others of the statistical software SAS 9.1.2.