Preparation of well-distributed titania nanopillar arrays on Ti6Al4V surface by induction heating for enhancing osteogenic differentiation of stem cells

Rolled Ti6Al4V plates with dimensions of 10 × 10 × 2 mm and cylinders with a diameter of 12 mm and a length of 10 mm were successively abraded with #200, #400, #600 and #1000 SiC papers, and then washed with acetone, ethanol and deionized water in an ultrasonic cleaner for 15 min, respectively. The ground specimens were subjected to supersonic frequency induction heating treatment with a frequency of 50 kHz and a power level of 60 kW, meaning it was possible to obtain an induction coil with a heating duration controllable to within 1 s. The IHT period of the Ti6Al4V plates lasted for 20–35 s, and they were then cooled to room temperature (25 °C) in air. However, the IHT times of the cylinders lasted individually for 14, 16, 18 and 20 s with a different induction coil. Finally, the heated specimens were all ultrasonically cleaned in acetone, deionized water and alcohol for 10 min, respectively, and then dried at 50 °C for 24 h in air.

The surface morphology of the Ti6Al4V specimens before and after IHT oxidation was assessed using a scanning electron microscope (SEM, HATA-CHI SU-70, Japan) at 15 kV. Prior to imaging, the specimens were coated with a gold target for 30 s under vacuum to improve electrical conductivity. Energy dispersive x-ray spectroscopy (EDS) was also performed to investigate the element composition of the oxide layer at specific areas on the surface. In addition, the cross-sectional view of the oxide layers was achieved using other SEM equipment (Helios G3CX, FEI, USA) with a 2 kV, 1.6 nA electron beam. Samples with an inclination angle of 52° was rotated in order to obtain the SEM cross-sectional images. Atomic force microscopy (AFM) was used to evaluate the surface topography and roughness of the oxide coatings, which was performed with a stand-alone instrument (Dimension Icon, Veeco Instruments Inc., USA) operating in tapping mode in air. A SiNi tip with a nominal spring constant of 42 N m⁻¹, was tapped under a scan rate of 1 Hz and at a resolution of 512 samples per line at 5 × 5 μm² scan area. The values of nanoroughness were calculated by NanoScope Analysis software (version 1.5), included in the AFM. The root mean square (RMS) roughness values were chosen as representative of the average surface roughness.

The surface phase compositions of the IHT-treated plants were examined by x-ray diffraction (XRD, Rigaku, Tokyo, Japan) using Cu Kα radiation with a wavelength of 0.154 nm. The x-rays were generated from a copper source operating at 40 kV and 100 mA. The goniometer was set at a scan rate of 4° min^–1 with a 2θ scan from 10–90° using step-scan intervals of 0.02°/2θ. The chemical composition and binding state of the external layer of the specimens oxidized by IHT were analyzed by x-ray photoelectron spectroscopy (XPS, ESCALAB 250, Thermo Fisher Scientific, USA). The XPS spectra were calibrated with respect to the C 1s peak at an energy of 284.8 eV, and recorded using monochromatic Al Kα radiation with a 90° take-off angle. In addition, the Al-Kα energies were 200 eV for the survey and 30 eV for the high resolution scans. High-resolution narrow scanning was also recorded to examine the chemical binding state of each element (e.g. O 1s, Ti 2p, and Al 2p).

The oxide layer on the Ti6Al4V surface was sliced into a parallel-sided thin foil using a focused ion beam equipped with scanning electron microscopy (Helios G3CX, FEI, USA), both gallium ion (Ga⁺) beam irradiation and subsequent platinum (Pt) electrode deposition under vacuum. The 2 kV, 1.6 nA electron beam was used for navigating and positioning the interest region of the sample surface. After the selection of a suitable region, the area was covered with electron-beam-deposited Pt (∼300 nm thick). The final oxide lamella with a size of 10 × 6 × 2 μm was formed using a 30 kV, 1 pA ion beam. Then, the thickness of the lamella was thinned to 60 ± 10 nm by the ion beam with decreasing energy (precisely controlling the thickness variations) in order to allow the e-beam to penetrate easily during subsequent tests. TEM analysis was utilized to obtain information about the morphology, crystal structure and lattice spacing of the FIB-thinned sample. The bright-field TEM-images and selected-area electron diffraction (SAED) patterns of the oxide were obtained on a transmission electron microscope (JEOL JEM-2010, Tokyo, Japan) at 120 kV. In addition, high-resolution TEM (HRTEM) images were obtained with a high-resolution transmission electron microscope operating at 200 kV with a maximum sample tilt angle of ±20° and a 1.9 Å point-to-point resolution.

The microstructure evolution in the Ti6Al4V plates and cylinders before and after IHT oxidation was investigated by EBSD observation using a Hitachi S-3400 N SEM equipped at 20 kV with an HKL-EBSD system. To minimize strain during the sample preparation, the specimens were mechanically polished and finally electropolished in a solution of 35% butarol, 60% carbinol and 6% perchloric acid with a voltage of 30 V at room temperature. The grip and gauge regions of the specimens were chosen for EBSD analysis at a step size of 1.5 μm, respectively. The dimension of the scan area was approximately 480 × 650 μm. The EBSD maps and grain orientations for microstructural characterization were evaluated by Channel 5 software (HKL Technology).

In order to determine the change of surface wettability of the oxidized Ti6Al4V, the wetting angle (θ) was evaluated according to the sessile drop method. A contact angle instrument (JC2000D3X, Powereach, China) was used to measure the static angle formed by dropping a deionized water droplet on the specimen surface at room temperature. To obtain a stable value for each sample, the measurement was repeated five times for specimens with the same treatment (n = 5). Adhesive strength between the oxide coating and substrate was investigated by scratch test using an automatic scratch tester (WS-2005, Lanzhou Institute of Chemical Physics, CAS, China) with a 200 μm radius diamond tip. The dynamic load increased from 0.2 to 30 N at a spend of 15 N min⁻¹ and was designated as the critical load while the coating was totally peeled off from the Ti6Al4V substrate. Three single scratches were performed for each sample in order to ensure a reliable result. Otherwise, the surface microhardness (HV) test of the Ti6Al4V plants before and after IHT heating was performed using a Micro Vickers Hardness Tester (DHV-1000, China) with a diamond indenter in ambient conditions. A static load of 0.2 kg for 10 s was applied to each sample surface. The averages of five indentations from five different samples were used in the statistical analysis in order to ensure the acquisition of a reasonably representative value.

To evaluate the in vitro formation ability of HA, 1.5 times SBF (1.5 × SBF) with various ion concentrations was applied in this work. It was prepared by dissolving appropriate amounts of reagent-grade chemicals of NaCl, NaHCO₃, KCl, K₂HPO₄, MgCl₂ · 6H₂O, CaCl₂ · 2H₂O and Na₂SO₄ in distilled water, buffered at pH 7.40, with (HOCH₂)₃CNH₂ (tris) and 1 M HCl at 37 °C. The mass of every agent in 1.5 × SBF was increased 1.5 times according to Kokubo's recipe [40], which similarly multiplied the ion concentrations in the solution, as shown in table S1 available online at stacks.iop.org/NANO/29/045101/mmedia. The IHT-35 s specimens were randomly chosen to identify the HA deposition ability of the as-prepared nanoscale surface in plastic containers with about 50 ml 1.5 × SBF. After immersion in SBF for 7–14 days, refreshed every two days, the specimens were cleaned and dried in air.

The BMSCs were purchased from ATCC (PCS-500-012, USA) and cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (Invitrogen, USA), 100 IU ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin (Gibco, USA), 2 mM l-glutamine (Gibco, USA) at 37 °C in a humidified incubator with 5% CO₂ in air. The medium was refreshed every two days. The representative specimens oxidized by IHT for 20 and 30 s were selected and prewashed before cell seeding in the following manner: all samples were ultrasonically washed in acetone first for 5 min and then in dilute water for 10 min to remove any contamination; finally, they were dried in a nitrogen stream. After these washing steps, the specimens were sterilized by autoclave and placed in 24-well tissue culture plates. The second passage of the BMSCs was trypsinized from culture flasks and the total cell number was counted using a hemacytometer to calculate the required volume of the medium for resuspension. After centrifugation, the cell pellet was resuspended with DMEM. Next, 500 μl of suspension—about 10 000 cells—was dropped into each well. The cell/sample constructs were incubated for 24 h to allow complete cell attachment, then the medium was transferred to an osteogenic induction medium containing 50 mg ml⁻¹ ascorbic acid and 10 mM β-glycerophosphate sodium (Sigma-Aldrich, USA). The osteogenic induction medium was changed every 2 days during the experimental periods.

Cell viability was assessed at 24 h post-seeding using live/dead staining (Live/Dead Staining Kit, Invitrogen, USA). The Ti6Al4V specimens were rinsed with phosphate-buffered saline (PBS) and then incubated for 30 min in live/dead stain, calcein AM and ethidium homodimer-1, diluted in PBS according to the manufacturer's instructions. After incubation, the specimens were again washed with PBS and then imaged, and counted using fluorescence microscopy (Olympus CKX41, Olympus DP25 Microscope Camera, Olympus America Inc.). For microscopy, the immunofluorescence staining assays were performed 24 h post-seeding. After 24 h in the culture, the BMSCs were fixed with 4% paraformaldehyde (PFA) for 20 min, treated in 0.1% Triton-X 100 for 10 min, and blocked in 3% bovine serum for 20 min. Mouse monoclonal antihuman F-actin antibody and rat monoclonal antihuman vinculin (VCL) antibody (Abcam, Cambridge, MA, USA) were used at a concentration of 1/200. Sections were inoculated overnight at 4 °C, followed by the application of green fluorescence-labeled donkey antimouse secondary antibody and red fluorescence-labeled goat anti-rat secondary antibody (Invitrogen, Carlsbad, CA, USA) for 2 h at room temperature. After culturing for 1, 4 and 7 days, BMSC proliferation was assessed using the Cell Counting Kit-8 (CCK-8) assay. In brief, at the prescribed time points, the specimens were rinsed with PBS and transferred to new 24-well plates. Then, 300 ml of DMEM and 30 μl of the CCK-8 solution (Beyotime, USA) were added to each specimen and incubated at 37 °C for 2 h. The absorbance was measured at 450 nm.

The cells were fixed with 4% paraformaldehyde for 20 min, permeabilized in 0.1% Triton-X 100 for 5 min and blocked with 3% bovine serum for 20 min. Mouse monoclonal antihuman runt-related transcription factor 2 (Runx2) primary antibody and rabbit monoclonal antihuman alkaline phosphatase (ALP) antibody (Abcam, Cambridge, MA, USA) were used with 1:200 dilution, as recommended by the manufacturer's datasheet. Sections were inoculated at 4 °C overnight. Red fluorescence-labeled donkey antimouse secondary antibody and green fluorescence-labeled goat antirabbit secondary antibody (Invitrogen, Carlsbad, CA, USA) were used at 1:300 dilution for 2 h. In addition, immunofluorescence stained cells were imaged on an Olympus BX61 inverted microscope system using filters. The expression of several osteogenic genes, the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH), Runx2, ALP, collagen I (COL-1), osterix (OSX) and osteocalcin (OCN), was analyzed by a reverse-transcriptase polymerase chain reaction (RT-PCR). The extraction of the messenger RNA (mRNA) and the reverse-transcription into complementary DNA (cDNA) were performed according to a previous procedure. Briefly, RNA extraction was performed using an RNeasy mini kit (Qiagen, USA) according to the manufacturer's instructions. The total RNA was reverse-transcribed using SuperScript III transcriptase according to the manufacturer's protocols (Invitrogen, USA). The PCR reactions were performed using Master Mix (Promega, USA). The PCR amplification cycles included denaturation for 30 s at 95 °C, annealing for 90 s and extension for 2 min at 72 °C for 30 cycles. The primer sequences are listed in detail in table 1. The gene expressions were normalized to the internal GAPDH expression, and the relative fold change was expressed by comparing it to that of the control group at each week. A comparative 2-^ΔΔCT method was used to quantify the relative mRNA expression.

One hour after seeding, the BMSCs were collected, rinsed and lysed with RIPA buffer (Boston Bioproducts, Worcester, MA, USA). Protein concentrations were determined by a bicinchoninic acid (BCA) kit (Invitrogen, USA). About 20 μl solution for each sample was run. Protein transfer was performed using a semi-dry system (Bio-rad, USA). A polyvinylidene fluoride (PVDF) membrane was used for the transfer after methanol activation and blocked in bovine serum albumin (BSA)-tween solution for 1 h. The ERK, p-ERK and GAPDH antibodies (Abcam, Cambridge, MA, USA) were used for antigen detection (1/1000). The secondary antibody was HRP-conjugated (1/10000). After antibody incubation, the PVDF membrane was thoroughly washed and then scanned using an LI-COR system (LI-COR Biosciences, Lincoln, NE, USA) and the images were analyzed by Quantity One software (Bio-Rad, USA). For the western blot assay of OCN, Runx2 and COL-1 protein, BMSCs were cultured on the Ti6Al4V samples for 1 week, and the same procedure was followed. The tests were repeated three times, and gray levels of the bands were quantified by ImageJ software (WS Rasband, NIH, Bethesda, MD, USA).

Statistical analysis was performed with GraphPad Prism software (La Jolla, USA). All experimental data is expressed as mean ± standard deviation (SD). A two-tailed paired t test for the independent sample was applied to evaluate the differences between the two groups, which were oxidized for different periods, with the statistical significance set at the 5% level (p-value < 0.05).

Figure 1 shows the XRD analysis of the Ti6Al4V plates treated by IHT oxidation for 20–35 s. In contrast to the Ti peaks generated by the bare specimen, the position of the diffraction peaks at 2θ with 27.42°, 35.97°, 41.10°, 54.23° and 56.83° detected on the surface revealed the crystalline phase of rutile TiO₂. In addition, weak peaks of both Al₂O₃ at 2θ with 21.26° and anatase at 2θ with 37.88° were obviously observed from the XRD spectra. This showed that the diffraction peaks of TiO₂ and Al₂O₃ were enhanced and the Ti peaks were consequently weakened along with the increasing IHT period from 20–35 s.

Zoom In Zoom Out Reset image size

To assess the binding state of the main chemical composition of the oxide layer, the specimens were investigated by x-ray photoelectron spectroscopy. The XPS spectra of the survey, O 1s, Ti 2p and Al 2p of the Ti6Al4V specimen treated by IHT for 25 and 35 s are shown in figure 2. From the XPS survey spectra, the O, Ti, Al and C were all detected (figure 2(a)). In addition, the values of the measured binding energy were calibrated using the C 1s peak (284.84 eV) as the internal standard. Both the O 1s peaks at 529.41, 530.05, 530.91 eV and 529.14, 529.76, 530.80 eV confirmed the presence of TiO₂ on the Ti6Al4V surface oxidized by IHT for 25 and 35 s, respectively. Similarly, the presence of Al₂O₃ in the oxide layer was also proved by the O 1s peaks at 528.90 and 531.97 eV obtained after IHT for 25 s and the peaks at 528.62 and 532.22 eV collected on the sample surface treated by IHT for 35 s. In line with the binding energies of the O 1s peaks, the binding energies of Ti 2p at 458.06, 459.03 eV; 457.85, 458.74 eV and 463.78 eV; 463.61—produced after oxidation for 25 and 35 s—separately corresponded to the Ti 2p_3/2 and Ti 2p_1/2 of Ti⁴⁺, which were ascribed to the Ti-O bond. The Al–O bonds in the oxide layers for 25 and 35 s were also confirmed by the bonding energies of the Al 2p peaks at 74.11 and 73.99 eV, respectively (figures 2(b) and (c)). Table 2 records the bonding energy and atom concentration (%) of the detected elements on the IHT-oxidized surface based on the XPS spectra. This shows that the oxide layers were mainly composed of O, Ti and Al, and the element content increased in increments of IHT time from 25 to 35 s. Otherwise, the value of the Al content was about twice as much as the Ti when heated for both 25 and 35 s, which was in contrast to the XRD results.

Zoom In Zoom Out Reset image size

Figure 3 shows the SEM images of the typical surface morphologies of the oxide layer and element content calculated from the EDS spectra on Ti6Al4V plates such as oxygen, titanium and aluminum, while being machined and then heated by IHT for 20–35 s. A relatively smooth surface texture with distinct abrasive scratches was obviously observed on the machined matrix. However, after IHT heating for 20–35 s, a well-distributed nanoscale pillar-like oxide grew vertically on the Ti6Al4V surfaces, and the nanopillar size successively increased both in width and particularly in length as the heating periods were upregulated (figure 3(a)). To investigate the change in chemical composition of the oxide layer surfaces during IHT, the element content was counted based on EDS measurements. The results showed that the detected O had the highest concentration, followed by Ti, Al and minor amounts of V on the surfaces treated by IHT for both 20 and 35 s, whereas only the substrate elements (Ti, Al and V) were detected on the machined surface (figure 3(b)). When heated for 20 s, it appeared that the oxygen content (at%) increased sharply to more than 60% from the 0 of the bare sample (0 s), and the growth rate slowed down as the IHT period further increased to 35 s. Meanwhile, this resulted in the lowering of the relative content of Ti and Al (figure 3(c)).

Zoom In Zoom Out Reset image size

The SEM cross-sectional view of the oxide layers was achieved by FIB filling to characterize the thickness of the layers and the growth patterns of the TiO₂ nanopillars, and the results are shown in figures 4(a)–(b). In particular, at oxide layers of IHT-30 and 35 s, the oriented TiO₂ pillars with increasing thickness (between green parallels) show vertical growth on the Ti6Al4V substrate after IHT oxidation (figure 4(a)). Finer oxide crystallites equivalently show up at the bottom of the oxide layers, revealing that the original fine TiO₂ crystallites had combined with large-sized nanopillars during the constant heating process. Otherwise, the surface morphologies of the oxide layers exhibited different levels of damage from the Ga⁺ ions during FIB milling. Figure 4(b) showed the calculated thickness values of the oxide layers after precise angle correction. The real values were successively increased from 571.2 ± 25.9 nm to 1467 ± 61.9 nm due to upregulation of the IHT period to 35 s.

Zoom In Zoom Out Reset image size

In order to determine the characteristic topographical features and surface roughness of the oxide layer, AFM was performed within a scanned area of 5 × 5 μm² and the corresponding results are presented in figures 4(c) and (d). The AFM topographical 3D images demonstrate that the nanomorphological crystallites appeared on the IHT-treated surface and continuously grew as the IHT time length increased (from 20–35 s). In contrast, parallel abrasive scratches with fewer protrusions were only observed on the machined Ti6Al4V surface (figure 4(c)). Moreover, the roughness (Ra) of the IHT-treated surfaces exhibited decreased values compared with that of the untreated surface (42.6 ± 4.5 nm), which was 26.3 ± 2.0 nm after IHT for 20 s, and increased to 31.3 ± 2.7 nm for 35 s, respectively (figure 4(d)).

The TEM sample was prepared by means of FIB milling technology to investigate the crystal structure of the oxide and the representative results after IHT for 25 s are shown in figure 5. The slender crystallites in the outermost layer grew outwardly, perpendicular to the surface, which was in agreement with the cross-sectional morphologies (figure 4(a)), and the grain boundaries (GBs) of the oxide were obviously indicated in the cross-sectional TEM image (figure 5(a)). Figures 5(b) and (c) showed the HRTEM images and the corresponding SAED patterns taken from the oxidation layer at different regions, illustrating two different crystal structures in the oxide layer. The measured out-of-plane d-spacings were 0.227 and 0.228 nm, and 0.237 nm, respectively. With the inserted SAED patterns, there were good tetragonal (t) and hexagonal close-packed (hcp) crystal arrangements of the diffraction planes, which are attributed to the rutile TiO₂ and α-Al₂O₃ crystallites separately. The hcp arrangement of the α-Al₂O₃ was located in the narrow region between the grains with different crystallographic orientations, which illustrate that a small amount of α-Al₂O₃ crystallites had formed at the GBs of the TiO₂ (within the dashed line in figure 5(c)).

Zoom In Zoom Out Reset image size

In order to evaluate the microstructure evolution in the Ti6Al4V plates before and after IHT heating, EBSD was applied to characterize the grain size and microrientation angle of the GBs, as shown in figure 6. Phase identification results suggest that the missive α-Ti (more than 99%) and a minor amount of β-Ti phases were found in the untreated Ti6Al4V (α + β Ti alloy), and the α-Ti slightly transformed to β phase after IHT-35 s. According to the grain boundary maps (figure 6(a)), the initial microstructure, which consisted of almost small-angle GBs (θ < 15°, green curves) was translated into large-angle ones (θ > 15°, black curves) when oxidized by IHT for 35 s. The quantitative statistics of the boundary angles shown in figure 6(b) illustrate that the fraction of small-angle GBs in the untreated substrate was estimated to be no less than 73.4%, which mainly resulted from the massive lattice defects existing in the rolled Ti6Al4V plate, such as dislocations. However, the transforming of small-angle GBs into large ones (about 60° and 90°) was caused by twin GBs [41, 42], and the proportion of small-angle ones dropped to 26.5% after IHT for 35 s. Figure 6(c) shows the grain size in the IHT-heated Ti6Al4V compared with that of the bare plate. This suggests that the IHT-treated grain diameters focus on low values (3 ± 2 μm) with a mean size of 4.61 μm, and the initial grains have higher values (4 ± 2 μm) with a mean size of 5.47 μm, contributed by recrystallization during induction heating.

Zoom In Zoom Out Reset image size

Figure 7 shows the EBSD orientation maps and textures with different colors of Ti6Al4V matrices before and after IHT oxidation. In the bare sample, most of the α-Ti grains with special orientations followed $\{0001\}$ (red color on IPF map) $\{\bar{1}2\bar{1}0\}$ (green color on IPF map); in contrast, the grains after IHT oxidation for 35 s possessed their prismatic planes. The martensitic phase with lamellar microstructures formed in the Ti6Al4V matrix due to the high temperature gradients during the rapid IHT process (figure 7(a)) [43]. The $\{0001\},$ $\{11\bar{2}0\},$ $\{10\bar{1}0\}$ pole figures in figure 7(b) appeared as a symmetrical distribution in the ND/TD section, which suggest the intense preferred orientation of α-Ti grains, and was due to the deformation texture during the rolling process of the Ti6Al4V plate. After IHT oxidation for 35 s, the intensities of the $\{0001\},$ $\{11\bar{2}0\},$ $\{10\bar{1}0\}$ planes were evenly distributed around the basal pores in the Ti6Al4V alloy, which indicated that the grain orientations varied toward multiple directions and the texture was highly reduced. Furthermore, the grain refinement and deformation texture reduction caused the enhancement of the mechanical properties of the alloy matrix, in terms of strength, plasticity and toughness [44].

Zoom In Zoom Out Reset image size

The sessile drop method was applied to measure the contact angles of Ti6Al4V surface oxidized by the IHT, which aimed to identify the evolution of wettability towards water when varying the IHT periods, as shown in figure 8. The θ of a liquid droplet on a solid surface is given by Young's equation [45]:

$$\begin{eqnarray}&&\cos \,\theta =\displaystyle \frac{{\gamma }_{{\rm{s}}{\rm{g}}}-{\gamma }_{{\rm{s}}{\rm{l}}}}{{\gamma }_{\mathrm{lg}}}\end{eqnarray} \tag{ 1 }$$

where ${\gamma }_{{\rm{s}}{\rm{g}}},$ ${\gamma }_{{\rm{s}}{\rm{l}}},$ and ${\gamma }_{\mathrm{lg}}$ are the interfacial free energy per unit area of the solid–gas, solid–liquid and liquid–gas interfaces, respectively. According to equation (1), θ was determined by the included angle between the liquid–solid interface and the liquid–gas interface (figures 8(a) and (b)) [46]. Compared with the value of the machined surface (83.5 ± 2.2°), the θ decreased successively to 63.9 ± 1.7° upon increasing the IHT time to 35 s, which indicates that IHT oxidation plays an important role in both reducing the hydrophobicity and increasing the wettability of the surface of the specimen (figure 8(c)).

Zoom In Zoom Out Reset image size

Figure 9 shows the results of the adhesion and proliferation of BMSCs on IHT-treated Ti6Al4V. The effect of specimens treated by IHT for 20 and 30 s on the viability of the BMSCs was assessed at 24 h time points and the LIVE/DEAD assay results are shown in figure 9(a). Green cells were considered live and red cells were considered dead (cleaned out). As evident from the representative LIVE/DEAD images in figure 9(b), the viable cells in the two groups displayed similar staining patterns and there was no significant difference between IHT-20 s and machined group, which indicates that the IHT-20 s treatment had no significant effect on stem cell adhesion for 24 h. However, the cell number on IHT-30 s group was slightly lower than the machined group.

Zoom In Zoom Out Reset image size

In order to confirm whether the IHT of Ti6Al4V affected the state of F-actin in the BMSC lineage, we induced F-actin depolymerization in BMSCs over the 24 h time period, as shown in figure 9(c). In line with the LIVE/DEAD assay results, the results of immunofluorescence staining revealed nuclear F-actin and vinculin in the focal adhesion plaques, and there was no significant difference between the machined group and the IHT groups, which indicated that the IHT-treated Ti6Al4V did not make a particularly notable improvement in cell adhesion with the machined surface.

Figure 9(d) shows the results of cell proliferation on the specimens before and after IHT oxidation measured by the CCK-8 assay. At day 1, the BMSC numbers on the IHT-treated Ti6Al4V surface were a little smaller than that on the control group. Moreover, at day 4 and day 7 the cell numbers for the IHT-20 s and IHT-30 s groups increased rapidly. Although the cell numbers on the IHT-20 s group were still no more than that of the machined group for up to 7 days, it possessed the highest values of IHT-30 s specimens after culturing for 4 and 7 days, while there was no statistical significance between the values in each group (p ≥ 0.05).

The differentiation of BMSCs and osteoblasts on the IHT-treated specimens was further confirmed by the immunostaining of specific proteins expressed at 7 days, as shown by figure 10(a). Early marker expression was evaluated by ALP expression in 7 days, as per the pattern observed in the above study. Significant expression of the ALP was clearly observed in the cell culture on the IHT groups when compared to those on the machined Ti6Al4V surface. When increasing the IHT period, the expression of ALP in BMSCs was upregulated in the IHT-20 s and IHT-30 s groups. However, the Runx2 expression was also detected in different conditions, with a similar evaluation of the ALP expression pattern in the BMSCs.

Zoom In Zoom Out Reset image size

The bone-related gene expression profile (e.g. ALP, OCN, Runx2, OSX and COL-1) for the BMSC-seeded specimens is shown in figure 10(b). The expression levels of genes, especially ALP and OSX, increased significantly among the IHT-20 s and IHT-30 s group as compared with the control group in the first week, having higher values of gene expressions than the IHT-20 s group when increasing the IHT period to 30 s, except for CON and OSX. In the second week, the expressions of OCN, Runx2 and COL-1 further increased in the IHT groups and all were higher than the machined group. However, at three weeks the OCN expression dramatically reached its highest value in the IHT-30 s group, compared with the IHT-20 s and control group.

In addition, the western blot for p-ERK 1/2 and ERK 1/2 was performed 1 h after seeding the BMSCs to the IHT and control samples, while that for OCN, Runx2 and COL-1 was performed at 1 week. While the BMSCs were cultured on the IHT and machined samples, osteogenic differentiation was indicated by the expression of osteogenesis-related molecules, as shown in figures 10(c) and (d). The results reveal that IHT treatment obviously increases the p-ERK 1/2 level and further raises the Runx2, OCN and COL-1 levels in the IHT-20 s group. Furthermore, the molecule expression was enhanced a little, accompanied by an increase in the IHT time to 30 s.