2.1

Sample preparation

This study employed coin shaped samples with a diameter of 6.25 mm and a height of 2 mm. The samples were made of titanium zirconium (TiZr) with a zirconium content of 13–17% zirconium; they were grit blasted and acid etched in hydrochloric and sulphuric acid (SBAE). After production, handling and packaging was done under nitrogen cover gas and the samples were stored in 0.9% NaCl, making the surfaces comparable to the commercially available SLActive ^® surface (Institut Straumann AG, Basel, Switzerland) .

Before coating, the test coins were unpacked under laminar flow and washed for 5 min in an ultrasound bath with deionized reverse osmosis water. The hydridation buffer was a 2 M acetate buffer solution at a pH of 3 mixed from acetic acid (Fluka, Oslo, Norway) and sodium acetate (Sigma–Aldrich, Oslo, Norway) and applied at a volume of 200 ml to the test beakers. This buffer was used directly for production of the hydrided coins from TiZr SBAE base material. For the biomolecule coated samples, 200 mg of Doxy were dissolved in the buffer to obtain a final concentration of 1 mg/ml. Doxy Hyclate was delivered by Yangzhou Pharmaceutical (Yangzhou Pharmaceutical Co., Ltd., Jiangsu, China).

The coating was performed on a custom made setup, controlled by the software LabView, providing a constant direct current of 0.65 mA on individual channels for each sample. The process runtime was 75 min and the coating was done at room temperature. Thereby the sample coins were connected to the cathode of the custom made coating setup, whereas a platinum anode served as the counter electrode. The TiZr coins were mounted on individual titanium holders, Teflon caps protected the interface between the holder and the coin from the electrolyte. After the process, the Doxy coated coins were dried under nitrogen gas flow and packed in containers filled with nitrogen prior to further analysis. Control samples were washed with ultrasound and dried under nitrogen in the same fashion as the Doxy coated samples. As a reference for the FTIR analysis four coins were spiked with 10 μl of water dissolved Doxy to obtain a defined Doxy coverage on the reference sample. The amount of applied Doxy was 11.5 μg per coin. The solution covered the whole top coin surface and samples were dried under nitrogen flow. All samples were kept under cover gas before analysis, as the exposure to air results in the attraction of environmental carbon, which can impair surface chemistry analysis . Storage conditions were cold (<10 °C), dry and dark.

2.2

Fourier transform infrared (FTIR) spectroscopy

A PerkinElmer Spectrum 400 FT-IR/FT-NIR spectrometer (PerkinElmer, Waltham, MA, USA) was used to analyze the chemical composition of the surface. For analysis of the Doxy powder the ATR accessory was used, while the Doxy coated coins were measured with a diffuse reflectance (DR) accessory in order to compensate for the rough surfaces. The DR accessory is designed with an adjustable mirror to collect the scattered beams by the rough surface. After recording of a background spectrum from an untreated TiZr SBAE coin, the samples were scanned on three different spots each. The scans were executed on a wavelength between 4000 cm ⁻¹ and 1000 cm ⁻¹ in the mid infrared region. Every measurement was constituted of 8 individual measurements with a resolution of 8 cm ⁻¹ . After scanning all samples were background corrected, normalized and the Kubelka–Munk algorithm was applied to the datasets.

2.3

X-ray photoelectron spectroscopy (XPS)

The XPS analysis was carried out on an Axis Ultra ^DLD XP spectrometer (Kratos Analytical, Manchester, UK) using monochromatic Al Kα radiation ( hν = 1486.6 eV). Survey spectra were recorded in a range between 1100 eV and 0 eV binding energy (BE). Detail spectra were acquired in the energy regions of C 1s, N 1s and Ti 2p. The instrument resolution was 1.1 eV for the survey scans and 0.55 eV for the detail scans. The analysis area was 300 μm × 700 μm.

The samples used for the XPS analysis were not stored under nitrogen cover gas after coating and washing. Samples were mounted on a sample bar with conductive carbon tape. The energy shift due to surface charging was below 1 eV based on the C 1s peak position relative to established BEs, therefore the experiment was performed without charge compensation. All BEs were referenced to the C 1s peak 284.5 eV according to Molder et al. .

2.4

UV–vis spectroscopy (UV–vis)

A release study was executed to assess the total amount of Doxy present on the coins. The release was performed in 0.5 ml of a solution of 60% acetonitrile and 3% trifluoroacetic acid (ACN–TFA). Two time points with n = 3 Doxy coated samples per group were analyzed after release in ACN–TFA. The release fluids were analyzed after 6 h and after 7 days with a Nano-Drop ND 1000 spectrometer (Nanodrop Technologies, Wilmington, DE, USA) in UV–vis mode at 263 nm. A calibration curve was recorded at Doxy concentrations between 0.005 mg/ml and 0.25 mg/ml.

2.5

Secondary ion mass spectroscopy (SIMS)

Depth profiles of carbon and hydrogen were measured by SIMS (Cameca IMS 7f, Paris, France) on the test coins. The analysis was carried out for ¹ H and ¹² C ions. One control sample was included as background and a hydrided sample served as a control for carbon contamination. A 50 nA primary beam of 15 keV Cs ⁺ was used for rastering over a surface area of 200 μm × 200 μm whereas negative secondary ions were collected from central part (67 μm × 67 μm) of the crater at room temperature. Doxy inclusion in the surface was detected by analysis of the depth profile of the ¹² C isotope.

The depth of each sputtering crater was analyzed by a blue light profilometer (SensofarPLμ 2300, Sensofar-Tech S.L., Terrassa, Spain), and the aspect heights calculated using an advanced software package (Sensomap, Sensofar-Tech S.L., Terrassa, Spain). The crater depth was measured with a 50× objective (50× PI, Nikon, Tokyo, Japan) over an extended topography of 2 × 2 images, each with a viewing area of 253 μm × 190 μm at 20% overlapping. The result data from the SIMS measurement was calibrated according to the obtained crater depth.

2.6

Field emission scanning electron microscope (FE-SEM)

The micro- and nano structural analysis of the surface topography and morphology was performed with a FEI Quanta 200 FE-SEM (FEI Hillsboro, Oregon, USA). The coins were platinum sputtered prior to analysis and mounted on a 45° aluminum rack with conductive carbon tape. Images were acquired with a working distance between 5 mm and 7 mm. Acceleration voltages of 3–5 kV were applied to achieve magnifications of 10.000× and 30.000×. High vacuum (HV) operation mode was used for the examination of all coins.

2.7

In vitro study

The mouse osteoblastic cell line MC3T3-E1 was obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). MC3T3-E1 cells were routinely cultured at 37 °C in a humidified atmosphere of 5% CO ₂ , and maintained in α-MEM supplemented with 10% fetal calf serum (FCS) and antibiotics (50 IU penicillin/ml and 50 μg streptomycin/ml). Cells were subcultured 1:50 before reaching confluency using PBS and trypsin/EDTA. All experiments were performed in the same passage of the MC3T3-E1 cells (passage 16).

The coins were placed in a well plate and 7 × 10 ³ cells were seeded on each well to study cell differentiation and cell toxicity. The same number of cells was cultured in parallel in plastic tissue culture dishes (TCP) in all experiments for visual control. Trypan blue stain was used to determine total and viable cell number. For the experiments, MC3T3-E1 cells were maintained for 14 days on the implants in α-MEM supplemented with 10% FCS and antibiotics. Culture media was changed every other day. Culture media was collected after 24 h to test the toxicity of the treatments and of the different implant surfaces (LDH activity). To study cell differentiation, cells were harvested at 14 days and collagen 1 (Coll-1), alkaline phosphatase (ALP) and osteocalcin (OC) gene expression were analyzed using real-time RT-PCR.

The presence of LDH activity in the culture media was used as an index of cell death. LDH activity was determined spectrophotometrically after 30 min incubation at 25 °C using 50 μl of culture media and adding 50 μl of the reaction mixture. Positive control (100%) was cell culture media from cells seeded on TCP dishes and incubated with Triton X-100 at 1%. Negative control (0%) was cell culture media from cells seeded on TCP dishes without any treatment.

Total RNA was isolated with Tripure (Roche Diagnostics), following the instructions of the manufacturer. RNA was quantified using a spectrophotometer set at 260 nm (Nanodrop). Real-time PCR was performed for two housekeeping genes , 18S ribosomal RNA (18S rRNA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and three target genes , ALP, collagen-I and osteocalcin. The primer sequences have been previously reported . The same amount of total RNA (0.25 μg) from each sample was reverse transcribed to cDNA at 37 °C for 60 min in a final volume of 20 μl, using High Capacity RNA to cDNA kit (Applied Biosystems). Each cDNA was diluted 1/4. A negative control without cDNA template was run in each assay. The concentration of PCR reaction components, amplification program, and quantification were done as documented earlier .

2.8

In vivo rabbit study

An animal study for assessment of the in vivo performance of Doxy was executed with 10 Grey Bastard Chinchillas. The animals were 6 months old and had a weight of 2.86 kg in average (Crl:CHB, Charles River Laboratories International, Inc., Research Models and Services, Sulzfeld, Germany). During the experimental period, the animals were kept in cages 2 days post-surgery, and 1 day prior to euthanization. Room temperature, humidity and diet were standardized. Sedation and anesthesia were done according to the Norwegian School of Veterinary Science, Laboratory Animal Unit; SOP on anesthesia of rabbits. The experiments had been approved by the Norwegian Animal Research Authority (NARA) and registered by this authority. The procedures have thus been conducted in accordance with the Animal Welfare Act of June 1, 2010, No. 94 and Regulation on Animal Experimentation of January 15, 1996.

Sedation of the rabbits as well as euthanasia and all surgical procedures were done as described by Rønold and Ellingsen . The only difference to this process was the integration of a central defect, drilled into the bone marrow region, with a diameter of 3.5 mm as described by Haugen et al. .

After implant detachment, wound fluid sampling and the analysis of LDH analysis, ALP activity and total protein were carried out analogous to Haugen et al. and Monjo et al. . Total RNA was isolated from peri-implant bone tissue as described already elsewhere . An amount of 280 ng of extracted RNA was reverse transcribed to cDNA as described above for the in vitro experiment.

Real-time PCR was performed with a Lightcycler-480 ^® (Roche Diagnostics, Mannheim, Germany) using SYBR green detection. Real-time PCR was done for three housekeeping genes: 18S ribosomal RNA (18S rRNA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and beta-actin, and nine target genes: bone morphogenetic protein 2 (BMP2), collagen type I (Coll-I), osteocalcin (OC), vacuolar type proton ATPase (H ⁺ -ATPase), tartrate-resistant acid phosphatase (TRAP), interleukin-6 (IL-6) and interleukin-10 (IL-10). The primer sequences, concentration of PCR reaction components and amplification program were analogous to Monjo et al. .

All samples were normalized by the geometric mean of the expression levels of β-actin and GADPH and fold changes were related to 4 weeks of implant placement using the mathematical model described by Pfaffl et al. , as follows: <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='ratio=EtargetΔCp target(mean control4weeks−sample)/EreferenceΔCp target(mean control4weeks−sample)’>ratio=EΔCp target(mean control4weeks−sample)target/EΔCp target(mean control4weeks−sample)referenceratio=EtargetΔCp target(mean control4weeks−sample)/EreferenceΔCp target(mean control4weeks−sample)
ratio = E target Δ Cp target ( mean control 4 weeks − sample ) / E reference Δ Cp target ( mean control 4 weeks − sample )
, where E is the amplification efficiency for the real-time PCR reaction and Cp is the crossing point (cycle number at detection threshold) of the reaction amplification curve as determined by the Lightcycler-480 ^® software, for the target and reference gene, respectively. Stability of reference genes was calculated using the BestKeeper tool . Thus, values were expressed as a percentage of TiZr (reference implant) at 4 weeks, which were set to 100%.

After extraction of the coins and wound fluid sampling, the tibia samples were prepared for analysis and scanned with μ-CT as described elsewhere . Furthermore reconstruction and bone mineral density (BMD) calibration was done according to the same description. The volume of interest (VOI) in each bone sample was chosen as a cylinder with a diameter of 3.5 mm and a height of 2.5 mm starting from the outer cortical bone defect (which was in contact with the implant surface) and toward the marrow space. This VOI was identical to the defect created in the cortical bone prior to placement of the samples. Furthermore analysis of the BMD devolution dependent on the distance to the implant surface (2D BMD) was executed with a diameter of 3.5 mm and a height of 2.5 mm. The threshold level was set from 65 to 255.

2.9

Statistics

Analysis of the raw data was carried out with Excel 2007 (Microsoft, Redmond, USA) and SigmaPlot 11.0 (Systat Software, Inc., Chicago, USA). All datasets were examined for parametric or non-parametric distributions (Shapiro–Wilk test). All parametric data are presented as mean values ± standard error of the mean (SEM) and non-parametric data as median ± interquartile range (IQR). For the in vitro study, ANOVA on ranks with Kruskal–Wallis multiple comparison was used for the assessment of OC. One way ANOVA was used for cytotoxicity, and gene expression of Coll-I and osteocalcin. ALP gene expression was analyzed with paired t -test. The statistical analysis of the in vivo results was done with Student’s t -test and Doxy coated samples were compared to TiZr SBAE and hydrided at each time point. Differences were considered significant (*) for p ≤ 0.05 and highly significant (**) for p ≤ 0.01, and labeled respectively. For differences compared to the hydrided group ( ^# ) for p ≤ 0.05 and highly significant ( ^## ) for p ≤ 0.01 were used, and labeled, respectively.

Statistical analysis of the μ-CT results was done with one way ANOVA or ANOVA on ranks for non-parametric datasets. (*) Indicates p < 0.05 compared to TiZr SBAE at each time point.