Abstract
Nanostructured materials play a fundamental role in orthopedic research owing to their outstanding properties and excellent biocompatibility. Titania nanotube (TNT) arrays engineered by electrochemical anodization process have been extensively explored and used as effective carriers for controlled drug delivery. In this study, we proposed a drug delivery system based on coordination bond. Iron (III), Fe3+, on the nanotube surface can effectively bind to alendronate sodium (NaAL), a drug for the treatment of osteoporosis, through coordination bonds, which can be formed or broken through the change of pH, and thus can be controlled by pH. The pH-responsive system was prepared by three-step procedure: (i) fabrication of TNTs by electrochemical anodization, (ii) modification of amino groups on the surface of nanotubes by hydrothermal method, and (iii) amino-functionalized nanotubes by Fe3+ solution soak. The Fe-modified TNTs not only allowed alendronate-loading content of up to 50.2% by weight, which is significantly higher than most drug delivery systems previously reported, but also delayed and prolonged drug release. Moreover, in vitro drug release experiments demonstrated that coordination bond-based TNT system may have great potential applications in clinical use.
1. Introduction
With the continuous developments of material and bone sciences, biomedical materials have been widely used in orthopedic field. Bone plays a key role in many important functions of human physiology. Surgical implants are usually produced from titanium and its alloys, which are designed to link and connect broken bones [1]. However, bone infection, a serious condition found in the orthopedics, is one of the most common clinical complications associated with bone implants. Such infection may also lead to more serious complications, such as implant failure and rejection, which can cause great pain and even death of patient [1]. Systemic drug delivery approaches have been used traditionally to combat these clinical complications and treat bone infections or poor implant integrations. These approaches are, however, inefficient in reaching to the bone site, and drug overdose can cause toxicity issues and side effects in patients. Therefore, localized drug delivery, a method carrying out direct and localized delivery of drugs from implants, has become a potential alternative to overcome these limitations. TiO2 nanotubes can be prepared by anodic oxidation of titanium substrates in dielectric solution, which are further aligned with the matrix metal. Not only do TiO2 nanotubes have good biocompatibility, their nanotubular structures also facilitate the drug loading and sustainable release. Thus, they have been used to achieve various drug delivery goals.
These works ultimately are aimed at developing drug delivery implants that have favorable release kinetics and optimized rate and time of release for a broad application in clinical therapies. In order to release drugs at uniform and constant rates, the following improvements were made to the implants: (a) structural modifications of TNT dimensions (in terms of diameter and length), (b) surface modifications, (c) biodegradable polymer coatings (to reduce pore openings), (d) polymeric micelle-encapsulated drug carriers, and (e) external source to stimulate drug release [2].
Stimuli-responsive drug delivery systems have been widely used to reduce premature release [3–5]. Stimuli can consist of external stimuli (such as temperature, light, electric field, and magnetic field) and internal stimuli (such as pH, redox, and biological ions) [6–8]. Among these, the pH-responsive delivery systems have been extensively studied [9–13]. Until recently, a novel coordination bond-based pH-responsive drug delivery system, which enabled the release of drug molecules by small changes of pH, was developed [14–16]. The formation and cleavage of metal ions and ligand coordination bonds are sensitive to pH because metal ions and protons (Lewis acids) compete with each other to bind to Lewis base ligands. Therefore, the coordination bond-based pH-responsive drug delivery system is a biomimetic process.
In this study, we reported a TNT-based pH-triggered drug delivery system. Vertically aligned nanotube arrays of titanium oxide were first fabricated on the surface of titanium substrate by direct anode oxidation with NH4F (the supporting electrolyte). The top-opened TiO2 nanotubes were then aligned and organized into high-density uniform arrays with diameters ranging from nm to nm.
2. Materials and Methods
2.1. Fabrication of Titania Nanotube (TNT) Arrays on Ti
To prepare TNT layer on Ti foil using anodic oxidation method, Ti foils were cleaned by ultrasonication in deionized water, ethanol, and acetone prior to anodization, in which titanium (Ti) foil acted as the working anode, while platinum foil was the cathode. The titanium oxide layers were fabricated in an optimized electrolyte containing NH4F (0.50 wt%), deionized water (10 vol%), and glycerol at a constant voltage of 60 V.
2.2. Synthesis of TNT-Fe Coordination System
In order to increase the amount of hydroxyl on the surface of TNTs, a base NaOH was used in heat treatment. TNTs were first placed in a hydrothermal reactor. Following the addition of 1 mg/ml ε-pl/HEPES solution, the samples were subjected to heat treatment at 70°C for 6 h. The resulting aminated nanotubes were then immersed in 0.1 M FeCl3 ethanol solution and stirred for 6 h to obtain the TNT-Fe coordination system. In this step, amino groups on the surface of TNTs should form coordination bonds with empty orbitals of Fe3+ ions.
2.3. Drug Loading
pH-responsive systems rely on the formation and cleavage of the metal-ligand complexes in response to external pH, which causes protonation/deprotonization of ligands. In this pH-responsive drug delivery system, alendronate sodium (NaAL) was used as coordination drugs. Under physiological condition (pH 7.4), strong coordinative affinity of –NH2 towards metal ions is established.
Two-step methods, including vacuum drying and hydrothermal methods, were used for drug loading. (1) Vacuum drying method: an appropriate amount of NaAL solution (40 mg/ml) was dropped (by pipetting) and spread evenly onto the surface of TNTs, and was then allowed to dry under vacuum at room temperature for 2 h. This step was repeated 5 times. (2) In hydrothermal method: the resulting samples from (1) were soaked in 4 mg/ml NaAL solution and incubated at 70°C for 6 h.
2.4. In Vitro Drug Release Study
The in vitro drug release was designed to study the pH-responsive properties of Fe-TNT and was compared with unmodified TNTs and free NaAL (1 mg, 5 mg) [12, 13]. Drug release was performed at a constant temperature of 37°C with a vibration of 100 rpm. The as-prepared Fe-TNT was immersed in 10 ml of PBS at pH 4.6, 6.0, and 7.4, respectively, while unmodified TNTs and NaAL were immersed in PBS at pH 7.4, respectively. 500 μl of PBS was collected from the supernatant predetermined time intervals for each sample. At the same time, 500 μl of corresponding fresh PBS was added. The amount of released drug was measured via UV-Vis spectroscopy.
2.5. Characterization
Structural characterizations of the TNT substrates before and after drug loading were performed using field emission scanning electron microscopy (FESEM). Fourier Transform Infrared Spectroscopy (FTIR) was used to study vibrations of the samples. Elemental composition and chemical state of sample surface were analyzed by X-ray photoelectron spectroscopy (XPS).
3. Results and Discussion
3.1. Structure and Morphology of Fe-Modified TNTs
Structure and morphology of the sample substrates were characterized by SEM and are summarized in Figure 1. SEM images of unmodified TNTs and ε-polylysine-modified TNTs are presented in Figures 1(a)–1(d). The images demonstrated the top surface of ε-pl-TNT, which confirmed that pores became smaller, about nm on average. In addition, the corresponding cross-sectional view shown in Figure 1(d) confirmed the grafting of polylysine, as indicated by the thick layers filling the materials.
(a)
(b)
(c)
(d)
(e)
(f)
Figures 1(e) and 1(f) show the morphology of Fe-modified nanotubes, demonstrating that the tube morphology did not change. However, the average diameter decreased from nm to nm after modification with Fe3+. This was the preliminarily proof that Fe3+ was modified onto nanotubes.
3.2. Structure and Morphology of TNTs before and after Loading
Figure 2 shows the morphology of TNTs loaded with alendronate at . It could be seen that the wall of the nanotubes became thicker with average diameters decreased from nm to nm after drug loading. The wall side was also significantly thickened, indicating that drugs were loaded into the nanotubes.
(a)
(b)
(c)
(d)
3.3. FTIR Analysis of Drug-Loaded Samples
To further investigate whether alendronate was successfully loaded, the samples were tested by Fourier Transform Infrared Spectroscopy. Figure 3 shows the infrared spectra of Fe-modified TiO2 nanotube arrays loaded with alendronate. The FTIR spectra were characterized by numbers of characteristic bands observed at 1643 and 1637 cm-1. Bands observed between 3500 and 3200 cm-1 were due to hydroxyl stretching vibrations. While a band at 1600 cm-1 was associated with –NH2 bond stretching, that at 1156 cm-1 was the absorption band of P=O [17]. These results demonstrated that NaAL was successfully conjugated with and attached to the surface of modified TiO2 nanotube arrays.
3.4. XPS Analyses of Fe-TNT and Drug-Fe-TNT
The XPS analyses of TNT-ε-pl-Fe and TNT-ε-pl-Fe-NaAL were carried out to study their elemental compositions and coordination states. Full XPS spectrum shown in Figures 4(a) and 4(b) showed that the samples contained Ti, O, N, and Fe elements. The corresponding higher resolution image of TNT-ε-pl-Fe sample (Figure 4(c)) showed two peaks at 710.1 and 724.5 eV, which indicated the two binding energy peaks of Fe2p. This is in agreement with that of Fe3+, indicating that TNTs on the surface of Fe were in two coordination states [18]. Moreover, the higher resolution of N1s in the same sample, as shown in Figure 4(e), showed two peaks at 399.5 and 401.8 eV corresponding to –NH2. This further demonstrated that polylysine was successfully grafted onto the surface of TNTs.
(a)
(b)
(c)
(d)
(e)
(f)
Figure 4(d) shows the higher resolution of Fe2p in TNT-ε-pl-Fe-NaAL sample. Peaks at 710.2 and 724.0 eV were the results of Fe2p3/2 and Fe2p1/2 after drug loading. This indicated that Fe3+ was in two coordination states [19]. Figure 4(f) further shows the higher resolution of N1s in the same sample; two peaks at 399.8 and 401.5 eV were corresponded to the Fe-NH3 and C-N+ bonds, respectively [20]. These results showed that drug and polylysine were successfully loaded into nanotubes by coordination bonds with Fe3+. Coordination bonds can be formed and broken by changes of pH so that pH-responsive drug release can be achieved.
3.5. Drug Release Study
To study the effects of different drug loading methods and pH to drug release, two samples were placed in 10 ml PBS buffers of different pH. Unmodified TNTs and different doses of free NaAL (low dose: 1 mg, high dose: 5 mg) were used as controls [12, 13]. As shown in Figures 5 and 6, five distinct drug release profiles were observed: (i) Since alendronate is easily soluble in water, as expected, NaAL was almost completely dissolved after 12 hours; similarly, the unmodified nanotube-loaded drug was full released within the first 24 hours. This is due to the physical adsorption between the drug and the nanotube, and the binding force is weak, which makes the drug easily fall off the nanotubes. (ii) In drug release directly from Fe-TNT, the release was as quick as in the first 24 hours. What was released at this time was the drug that dissociated on the surface of the nanotubes. (iii) In PBS buffer , the coordination bond was unstable and the drug release rate was relatively high. (iv) In PBS buffer and 7.4, while the amount of burst release and sustainable release was small, the drug release rates at both pH were similar. These results demonstrated that coordination bonds between Fe3+ and -NH2 was stable under neutral condition, and the sustainable release effect was obtained. So even under inflammatory conditions (), the drug can be released slowly to achieve the goal of long-term treatment.
In order to study the kinetics of drug release in the slow release phase, the release curve was fitted with the zero- and first-order kinetic equations and Higuchi equation. The results showed that the release of the drug from “amino-Fe-drug” loading system best-fitted with zero-order kinetic equation, which could be expressed as: , , where is the amount released during time, is the total amount of complete release, and is the release constant. Figure 7 is a linear fitting map of the drug release phase at PBS .
4. Conclusions
Coordination bond-based pH-responsive drug delivery systems can be fabricated onto the surface of TNTs. Guest molecules are incorporated by the coordination bond into the “host-metal-drug” architecture, in which host, metal, and drug represent functional groups of carriers, metal ions, and drug molecules, respectively. This work demonstrated that alendronate drug release time could be prolonged to more than 15 days by using Fe-TNT coordination system as a carrier and was achieved in pH-responsive manner. Moreover, the alendronate (at ) could be loaded into the nanotubes by vacuum drying and hydrothermal methods.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper.
Acknowledgments
This work was financially supported by Natural Science Foundation of Fujian Province of China (2016Y0025 and 2017J01685) and Fuzhou Science Technology Project (2017-G-61 and 2018-S-wq20).