Customized 3D printed porous titanium scaffolds with nanotubes loading antibacterial drugs for bone tissue engineering

2.1 Design and 3D printing of porous titanium alloy scaffolds

The porous titanium alloy scaffolds were designed by means of Solidworks. Literature suggests that the ideal pore size of a bone tissue engineering scaffold is 300–900 μm and ideal porosity ranges within 60–95% [28]. Thereafter, as shown in Figure 2, a diamond structure with pores of 660 μm and struts diameter of 330 μm was used as the basic pore unit of the scaffold, and the scaffold’s shape was a cylinder with a height of 5 mm and diameter of 10 mm (Φ10 × 5 mm). This structure has good load bearing capacity and can smoothly conduct stress to maintain the integrity of the structure and function [29]. The scaffolds were made by SLM system (M2, Concept laser, German). The selected Ti6Al4V powder with diameter of 17–50 μm was used as the printing raw materials. The printer had a laser power of 90 W and the laser scanning speed was set to 500 mm·s⁻¹.

Figure 2

Modeling of scaffold: (a) porous diamond unit, which was used for construction of the main body of porous scaffolds and (b) 3D modeling of the bone tissue engineering scaffold.

2.2 Surface modification and characterization of the porous titanium alloy scaffolds

The residual printing powder on the porous scaffolds was first removed by chemical corrosion (pickling) [30]. The corrosive solution was hydrofluoric acid (HF) solution consisting of 1 mL HF (40.0 wt%, Chengdu Chron Chemical Reagent Factory, Sichuan, China) and 50 mL deionized water. The scaffolds were put into the corrosive solution for 15 min of reaction at room temperature. Then the corroded samples were put into deionized water and treated by ultrasonic waves for 5 min. This process was repeated three times. After being washed with deionized water, the scaffolds were put into absolute alcohol and then treated by ultrasonic waves for 5 min. This process was also repeated three times. The samples were dried in an oven at a constant temperature of 60°C. Nanotubes were prepared on the 3D-printed titanium alloy scaffolds by anodic oxidation [31]. The printed porous titanium alloy scaffold was used as the anode while a platinum plate (0.5 mm × 0.5 mm × 0.1 mm) was used as the cathode, and the distance between the two electrodes was set to 20 mm. Ethylene glycol solution containing 0.5 wt% NH₄F and 10 vol% deionized water was selected as the electrolyte. In order to observe the effect of anodic oxidation voltage and oxidation time on the nanotube morphology, we obtained three groups of data by setting the voltage to 50, 60, and 70 V, respectively, and the anodic oxidation time to 1 h at room temperature and collected another three groups of data by maintaining the voltage at 60 V and setting the anodic oxidation time to 15 min, 30 min, and 1 h, respectively. Three scaffolds of each different type are used in this research to mitigate errors. Finally, all the prepared porous scaffolds were put into a muffle furnace for 2 h of heat treatment to eliminate internal stress and enhance the bonding force between nanotubes and their substrate.

Due to the influence of liquid flow on the diffusion of ions, the anodization within the pores of a porous scaffold may be slower than the external contact surface of the scaffold. Therefore, in this article, the surface inside the scaffold pores is referred as the “inner surface,” while the externally accessible surface is referred as the “outer surface.” The microscopic structure of the inner and outer surface and the crystallinity of these scaffolds were characterized. The experiments were performed using a SEM (JSE-5900LV, Japan) and the voltage was set to 5.0 kV. The diameter of nanotubes was measured and recorded by the software ImageJ. Then the main components on the sample surface were determined by XRD after 1 h of anodic oxidation at the voltage of 60 V [32].

2.3 Drug loading treatment of the porous titanium alloy scaffolds

After 1 h of anodic oxidation at the voltage of 60 V under an aseptic condition away from light, the heat-treated scaffolds were immersed into 1 mg·mL⁻¹ MH solution and kept in the solution for 2 h. These scaffolds were taken out of the solution and the coating surfaces were quickly washed with phosphate-buffered saline (PBS) solution to remove the residual drug outside the nanotubes. PLGA was then used to control the drug release rate. PLGA solution of 10 μL with a concentration of 10 mg·mL⁻¹ was added to the surface of the scaffold. After the solution was immersed in the hole of the scaffold, it was put into a drying oven to dry. The process was repeated four times. The scaffolds were put into a hot air oven at the interval between the coatings and then dried. Four specimens of each type of scaffold were used to mitigate test errors. Spectrophotometry was adopted to measure the drug release rate of these scaffolds. After the completion of the preparation experiment, each sample was immersed in 5 mL of PBS solution (0.01 M, pH 7.4) at 37°C. About 3 mL of the solution was replaced after every 5 h and supplemented with same volume of PBS. Three repeated samples were recorded at each time point to learn about the experiment results more accurately. The amount of MH released from the nanotubes was measured by UV–vis spectrophotometer at 360 nm.

2.4 Cytotoxicity test

The experiments were performed using MC3T3-E1 preosteoblastic cells, and the experiments were divided into three groups, each group includes five samples to mitigate errors, which is named as: scaffolds only treated by anodic oxidation (blank), scaffolds loaded with MH, and scaffolds loaded with MH and coated with PLGA (MH/PLGA). MC3T3-E1 cells were seeded in 24-well plates at a density of 4 × 10⁴ cells/scaffold in α-MEM medium (Hyclone, USA) containing 1% penicillin/streptomycin (Hyclone, USA). The cells were cultured under 5% CO₂ at 37°C, and the culture medium was changed every other day. After 1, 4, and 7 days of culture, cells were visualized by laser scanning confocal microscopy (Zeiss, SM 800, Germany) using fluorescein diacetate (Sigma), rhodamine phalloidin (Sigma) and 4ʹ,6-diamino-2-phenylindole (Sigma) as fluorescent dyes. Meanwhile, cell counting kit-8 (CCK-8; Beyotime, Shanghai) was used for cell counting.

2.5 Anti-infection function test

The inhibitory effect of different samples on gram-positive microorganisms (Staphylococcus aureus) was detected by zone of inhibition (ZI). First, the samples were divided into two groups: scaffolds loaded with drug (MH) and scaffolds loaded with MH and then coated with PLGA (MH/PLGA). Three scaffolds from each group were placed in normal saline and immersed for 3 days. Three scaffolds in each group were without immersion. The immersed scaffolds were rinsed with PBS once before the experiment and dried in a vacuum drying oven. A total of 12 scaffolds were used for subsequent experiments. These 12 samples were placed on LB solid agar Petri dishes containing approximately 10⁴ CFU S. aureus per plate so that the coated side of the material was in contact with the inoculated agar. Then, incubation was performed at 37°C for 24 h, and finally a zone was formed as an indicator of anti-infection activity. We conducted ZI statistical analysis using the software ImageJ.

2.6 Statistical analysis

All experimental data expressed as mean ± standard were calculated based on at least three independent experiments. Statistical analyses were carried out using OriginPro 2022 (OriginLab Corporation, USA). Data were analyzed using GraphPad Prism software (GraphPad Software Inc.) by Student’s t-test (unpaired and two-tailed), one-way or two-way analysis of variance (ANOVA), followed by Tukey post-hoc test. Statistical differences were shown with three significance levels: *p < 0.05, **p < 0.01, and *** p < 0.001.

3.1 Microstructures and the surface characteristics of the porous scaffolds

Refer Figure 3(a) for the particle size of titanium powder used for printing and see Figure 3(b) for the appearance and dimensions of the finished scaffold. We observed a high degree of sphericity from the spherical powder, with the particle size roughly ranging from 15 to 45 μm. SEM images of lower magnification before and after pickling are shown in Figure 4. It can be observed that scaffold surface is smooth after pickling because the pickling process has removed the residual titanium powder particles that were left by the printing process and had not been fully melted. The residual titanium powder particles may cause harm to the human body, and the pickling process helps improve the scaffold surface quality [33].

Figure 3

Raw materials and printing scaffold: (a) SEM image of titanium powders used for 3D printed and (b) appearance and dimensions of the obtained 3D printed scaffold.

Figure 4

SEM images of scaffolds before and after pickling: (a) appearance of the surface microstructures of the 3D printed scaffold before acid pickling, (b) appearance of the surface microstructures of the 3D printed scaffold after acid pickling, which shows a smooth surface and no residual powder remained on the surface of the scaffold.

When conducting experiments at different anodic oxidation voltages, we observed significant differences in the results, as shown in Figure 5. When the voltage is 50, 60, and 70 V, the diameter of nanotubes is 102, 128, and 152 nm after anodic oxidation for 1 h, and the diameter of nanotubes increases with the rising anodic oxidation voltage. The corresponding semi-quantitative results showed significant differences between the groups. The above results indicate that the diameter of nanotubes generated by anodic oxidation is positively correlated with the applied voltage, and can be precisely controlled by regulating the anodic oxidation voltage [34–36]. As shown in Figure 6, we can observe that holes begin to appear on the scaffold surface after the scaffold is anodized for 15 min at the voltage of 60 V. A dense non-conductive oxide layer is generated on the scaffold surface due to oxidation reactions after anodic oxidation starts. With the continuous oxidation reaction, the oxide layer on the metal surface is gradually thickened, thus blocking current flow and slowing down the thickening process of oxide layer. Since Al and V exist in the alloy, the local electrical field density is not uniform and fluorine ions were gathered at the location with large electric field, which is helpful for the chemical dissolution of TiO₂. Then holes are formed on the oxide layer through dissolution. After 30 min of anodic oxidation, we can observe that the outline of nanotubes is gradually formed. After all, the barrier layer at the pit bottom becomes thinner, the electric field density increases and more fluorine ions are accumulated. The accumulation further accelerates the dissolution rate of TiO₂, and small holes gradually grow into small pores. Unoxidized conductive titanium covered with a thick non-conductive oxide layer can be observed from the pore wall region. As the pores grow, these regions protrude, and the electric field density increases greatly due to the underlying conductive titanium. This effect accelerates the chemical dissolution. Since the non-conductive layer near the holes is thicker than that in the relatively distant regions, the dissolution rate is higher at the more distant locations. After 1 h of anodic oxidation, we finally observed the nanotubes with a good morphology, indicating a balance between the chemical dissolution at the bottom of nanotubes and the chemical solution at the mouth of nanotubes. Under such circumstance, the length of nanotubes stops increasing. These experimental results show that anodic oxidation voltage and anodic oxidation time have a significant effect on the growth and morphology of nanotubes, which provides a good condition for the subsequent drug loading of these scaffolds. Due to the formation of nanotubes, the scaffold surface becomes rougher, which significantly improves the surface roughness of the scaffolds. The surface with higher roughness shows better cell adhesions compared to smoother surface [37]. Moreover, rough surfaces can reduce the relative movement between the material and surrounding tissue, thereby relieving friction and irritation, reducing the inflammatory response caused by the material, improving the wetting properties of the material and making it easier to contact with biological fluids. A rough surface is important for such applications as medical devices, biosensors, and drug delivery systems because it can increase the interfacial area between materials and biological fluids to increase the efficiency of adhesion [36,38].

Figure 5

SEM images of nanotubes generated after 1 h of anodization at different voltage levels: (a) 50 V, (b) 60 V, and (c) 70 V. (d) Semi-quantitative results of the diameter of nanotubes. (Two-tailed t-test or one-way ANOVA with Tukey, *p < 0.05, **p < 0.01, and ***p < 0.001, error bars represent SD, n ≥ 3 independent experiments).

Figure 6

Morphology of nanotubes formed after anodization at a voltage of 60 V for varying durations: (a) low-magnification appearance of unanodized sample, (b) low-magnification appearance after 15 min of anodic oxidation, (c) low-magnification appearance after 30 min of anodic oxidation, (d) low-magnification appearance after 1 h of anodic oxidation, (e) micro-surface morphology of unanodized sample, (f) micro-surface morphology after 15 min of anodic oxidation, (g) micro-surface morphology after 30 min of anodic oxidation, and (h) micro-surface morphology after 1 h of anodic oxidation.

The scaffold after 1 h of anodic oxidation at 60 V was taken as a sample and the surface properties of this scaffold were tested. The XRD pattern shows that differently from the scaffolds not treated through anodic oxidation, the surface of the scaffolds after anodic oxidation has generated TiO₂ nanotubes with anatase structure, as shown in Figure 7. Titanium dioxide is a naturally occurring oxide of titanium and is commonly found in three main crystal structures: rutile, anatase, and brookite. The anatase structure is one of these crystal structures, and it has distinct physical and chemical properties. Anatase is characterized by tetragonal crystal symmetry, meaning its unit cell has a fourfold rotational symmetry axis. The crystal lattice arrangement in anatase is different from that of rutile and brookite. The nanotubes with anatase structure are proved to have better physical and chemical properties and contribute more to the osseointegration of orthopedic scaffolds [39,40].

Figure 7

XRD pattern of the scaffold before and after anodic oxidation.

3.2 Drug loading results of the scaffolds

After the scaffolds were finished and loaded with MH, the top view of the nanotubes and the morphological changes before and after drug loading were observed by means of SEM. As shown in Figure 8(a), the nanotubes before drug loading are composed of densely arranged tubular structures, and the top of them is in an open state, presenting a highly ordered layered structure. The morphology of the nanotubes after immersed in MH solution is shown in Figure 8(b) and the gaps between nanotubes are filled with MH. In order to slow down the drug release, the nanotubes loaded with MH are further covered with PLGA. The top morphology of the composite coating (nanotube/MH/PLGA) is shown in Figure 8(c). The surface is smooth and glossy and we cannot observe any tubular structure. Further studies have verified the successful deposition of MH in the nanotubes [41].

Figure 8

SEM images of the scaffold surface morphologies: (a) scaffolds are not loaded with drug, (b) scaffolds are loaded with MH, and (c) scaffold are loaded with MH and coated with PLGA.

Drug release kinetics is one of the important factors to control the therapeutic effects and side effects of drugs. The in vitro drug release behavior of the scaffolds only loaded with MH and of the scaffolds loaded with MH and then coated with PLGA was evaluated during the experiments. As shown in Figure 9, the experimental results show that the two types of scaffolds have different behaviors in terms of drug release. For the scaffolds only loaded with MH, the burst drug release behavior appeared within 25 h, and the drug release rate increased rapidly and then gradually stabilized after the burst drug release. For the scaffolds loaded with MH and then coated with PLGA, the initial release rate was lower, the release amount in 24 h was 48% of that from the scaffolds not coated with PLGA and only a small amount of drug was released. There was a significant difference in drug release between the two scaffolds at each time point. Then the release was steadily accelerated and tended to be consistent at last, with no significant difference in drug release between the two scaffolds at each time point. These results indicate that the scaffolds coated with PLGA on the surface can effectively reduce drug release rate.

Figure 9

Drug release curve of the porous titanium alloy scaffolds. (Two-tailed t-test or one-way ANOVA with Tukey, *p < 0.05, **p < 0.01, and ***p < 0.001, error bars represent SD, n ≥ 3 independent experiments.).

3.3 Cytotoxicity test of the porous titanium alloy scaffolds

CCK-8 experimental results are shown in Figure 10. In the experiment, the growth and survival of MC3T3-E1 preosteoblastic cells were affected by different scaffold materials. First of all, in the blank scaffolds, the number of cells showed a stable growth trend, indicating that the scaffold material has no significant effect on cell growth and survival. However, massive cell deaths occurred on the first day and cells increased steadily on the fourth and seventh day on the scaffolds only loaded with MH. However, they were significantly different from the blank scaffolds at the same period. Cytotoxicity was caused by the burst release of MH, resulting in massive cell death especially in the initial culture phase. The cell growth on the scaffolds coated with PLGA (MH/PLGA) was slight on the first day and many dead cells were detected on the fourth day. This situation may be caused by the harmful substances released by the degradation of PLGA and the substances produce an adverse effect to cells. However, the number of cells increased steadily on the seventh day, indicating that the substances from the degradation of PLGA had been fully released or the cells had gradually adapted to the environment and begun to grow again. A similar conclusion was obtained from the results of a cell survival experiment, as shown in Figure 11. In the experiment with blank scaffolds, the number of cells increased gradually, and almost no dead cells appeared. In the experiment on the scaffolds only loaded with MH, the fact that there were fewer viable cells than the other two groups were observed on the first day. This fact resulted from the toxicity caused by MH in the scaffolds. In the subsequent experiments, we observed an increasing number of cells since the fourth day, indicating that cells could still continue to proliferate on the scaffolds only loaded with drug. Although fewer viable cells were observed on the first day, the death rate of cells decreased gradually and the number of cells was gradually restored to a normal level in the subsequent experiments. For the scaffolds coated with PLGA, fewer cells died on the first day than for the scaffolds only loaded with drug. This is because the release of PLGA is a slow process and will not cause severe damage to cells. However, an increasing number of dead cells was observed on the fourth day, indicating that the degradation of PLGA released some loaded drugs and produced some adverse effects on the cells. These effects are within an acceptable range. According to the relevant results, the scaffolds only loaded with MH can significantly inhibit cell proliferation at beginning, but the cells are still viable and will proliferate later. Differently from the scaffolds only loaded with MH, the scaffolds loaded with MH and coated with PLGA can slow down the drug release to a certain extent, thus achieving sustained drug release and reducing the number of cell deaths in the early phase of cell culture. Certainly, the presence of PLGA on the surface of the scaffold also alters the mechanical properties of the scaffold’s surface and has a certain impact on cell adhesion. These findings confirmed some previous literature [42]. Considering the above, PLGA has an excellent effect in slowing down drug release. It is expected to apply PLGA to the design of drug delivery systems.

Figure 10

Cell CCK-8 results of different porous titanium alloy scaffolds. (Two-tailed t-test or one-way ANOVA with Tukey, *p < 0.05, **p < 0.01, and ***p < 0.001, error bars represent SD, n ≥ 3 independent experiments).

Figure 11

Test of viable/dead cells on the scaffolds loaded with drug: (a)–(c) viable/dead cell staining on the first day, (d)–(f) viable/dead cell staining on the fourth day, and (g)–(i) viable/dead cell staining on the seventh day. The red dots are the dead cells and the green dots are the live cells.

3.4 Anti-infection function test of the scaffolds

The anti-infection test results are shown in Figure 12. According to the results, both drug-loaded scaffolds have anti-infection effects with obvious ZI before immersion and the ZI diameter of MH scaffolds is greater than that of MH/PLGA scaffolds. The corresponding semi-quantitative results showed significant differences between the two groups. ZI still appears on MH scaffolds after immersion. However, the ZI is smaller and shows a lighter color than before. It indicates that the release of MH is affected by immersion. The loaded drug is released quickly, resulting in a rapid decline in its anti-infection effect. Only a small amount of drug solution still exists on the scaffold, showing a very weak anti-infection effect. The corresponding semi-quantitative results showed no significant difference between the two groups. Before immersion, the MH/PLGA scaffolds have an excellent anti-infection effect. After immersion, the anti-infection effect can be preserved, indicating that the MH/PLGA scaffolds have a certain sustained good anti-infection activity and proving that the coating of PLGA can delay the release of MH and maintain its anti-infection effect.

Figure 12

Anti-infection test results: (a) and (b) ZI of the two types of drug-loaded scaffolds before immersion, (c) and (d) ZI of the two types of drug-loaded scaffolds after immersion for 3 days, and (e) semi-quantitative results of ZI diameters. (Two-tailed t-test or one-way ANOVA with Tukey, *p < 0.05, **p < 0.01, and ***p < 0.001, error bars represent SD, n ≥ 3 independent experiments).