Enhanced Cytocompatibility and Antibacterial Property of Zinc Phosphate Coating on Biodegradable Zinc Materials

Abstract

Zinc (Zn) has recently emerged as a promising biodegradable metal thanks to its critical physiological roles and promising degradation behavior. However, cytocompatibility and antibacterial property of Zn is still suboptimal, in part, due to the excessive Zn ions released during degradation. Inspired by the calcium phosphate-based minerals in natural bone tissue, zinc phosphate (ZnP) coatings were prepared on pure Zn using a chemical conversion method in this study. The coating morphology was then optimized through controlling the pH of coating solution, resulting in a homogeneous micro-/nano-ZnP coating structure. The ZnP coating significantly increased the cell viability, adhesion, and differentiation of pre-osteoblasts and vascular endothelial cells, while significantly reduced the adhesion of the platelets and E. coli. Additionally, ZnP coating significantly reduced the Zn ion release from the bulk material during degradation process, resulting in a much lower Zn²⁺ concentration and pH change in the surrounding environment. The improved hemocompatibility, cytocompatibility and antibacterial performance of ZnP coated Zn biomaterials could be mainly attributed to the controlled Zn ion release and micro-/nano-scaled coating structure. Taken together, ZnP coating on Zn-based biomaterial appears to be a viable approach to enhance its biocompatibility and antibacterial property as well as to control its degradation rate.

Graphical Abstract

1. Introduction

Compared to the conventional metallic implant materials, biodegradable metals as temporary implants have been developed to avoid a secondary surgery, thereby accelerating the entire healing process while simultaneously reducing health risks, costs and scarring [1, 2]. Up to now, magnesium (Mg), iron (Fe) and zinc (Zn) are the three main classes of biodegradable metals as functional but temporary implants [1-6]. Zn is considered a promising biodegradable metal thanks to its essential role in many enzymes and in cell metabolic activity and functions [7-9]. In addition, it has a probably more suitable degradation rate, which is more likely in line with the clinical demand [10, 11].

Orthopedic and cardiovascular implants are their top two targeted applications [4, 12, 13]. Zn ion plays critical roles in bone growth through promoting osteoblast differentiation and inhibiting osteoclast differentiation [14], and maintaining cardiac function through redox signaling pathway [9, 15]. In vivo implantation of pure Zn wires and stents showed no severe inflammation, platelet aggregation, neointima or thrombosis formation in the abdominal aortas [16-18]. However, one of the significant concerns about Zn as degradable metals is its local and systemic toxicity as the recommended dietary allowance (RDA) for Zn is only 15-40 mg/day, much lower than that of Mg (300-400 mg/day) [19]. Moreover, notable cytotoxicity of Zn has been reported in different cells including human bone cells [20-22] and vascular cells [21, 23].

Thus, we hypothesize that controlling burst Zn²⁺ release from the implant is the key to minimize the toxicity of Zn implants. To achieve this goal, we proposed to surface engineering a zinc phosphate (ZnP) coating, inspired by calcium phosphate (CaP). CaP owns the inherent bone tissue compatibility due to their similar composition to carbonated apatite in natural bone tissue [24]. Therefore, they are proverbially applied in the orthopedic applications as ceramic substrates, reinforcement in composites, bone cement, or surface biofunctional coating [25-32]. Recently hydroxyapatite, a typical CaP, was incorporated into pure Zn resulting in a novel orthopedic implant with tunable degradation rates, enhanced bone formation ability, and effective antibacterial properties [33]. However, CaP may be not suitable for vascular stent application because of its potential adverse effect on vascular calcification [34-37].

Similar to CaP for orthopedic application, as a natural phosphate of Zn-based metals, ZnP with stable chemical property and biocompatibility showed the promising potential for biomedical applications [38, 39]. In addition, the feasibility of ZnP coating has been explored on several biomedical metallic substrates, including Ti, Fe and Mg alloys but not Zn-based ones [40-42]. Its stable chemical property could help to modify the degradation rate of the biodegradable metals and also promoted the fibroblast cell’s adhesion on the Ti [40]. On the other hand, Zn ion released from the Zn-based materials could potentially interact with the bacteria surface to induce cell deformation and bacteriolysis [43]. The ZnP coating could potentially have the antibacterial ability to reduce the risk of bacterial contamination and postoperative infection. Therefore, ZnP is expected to be a promising coating on Zn to improve the biocompatibility for both orthopedic and vascular applications.

In the present study, a ZnP coating was prepared on pure Zn using a chemical conversion method for the first time. The coating’s micro-/nano-structure, in vitro degradation behaviors, hemocompatibility, cell adhesion, viability, and differentiation of pre-osteoblasts and vascular endothelial cells, together with antibacterial performance have been studied to evaluate the performance of ZnP coating on pure Zn intended for orthopedic and vascular applications.

2. Materials and methods

2.1. Coating preparation

Pure Zn discs with dimensions of 10 mm × 4 mm were cut from rods (99.99+% purity, Goodfellow Ltd. Pennsylvania) and used as a starting material. Zn samples were polished using #1500 sandpaper, cleaned by sonication in acetone for 5 min, then immersed in 50 mL coating solution at room temperature for 5 min followed by rinsing with deionized water and drying in air before characterization. The coating solution was composed of 0.07 M Zn(NO₃)₂ and 0.15 M H₃PO₄ and the pH value was adjusted to 2, 2.5, and 3, respectively. The collagen (Col) coating was also prepared for comparison through incubating Zn samples in the 1 mg/ml rat tail type I Col solution (Corning, USA) for 30 min at room temperature [44, 45].

2.2. Surface characterizations

Surface and cross-sectional coating morphology and composition of the coating were identified using scanning electron microscopy (SEM, FEI QUANTA 200, USA) equipped with an energy-dispersive X-ray spectroscopy (EDS, Oxford Swift ED 3000, UK). The phases of different coatings were characterized by X-ray diffraction (XRD, Rigaku Dymax, Japan) using a monochromator at 40 kV and 44 mA with a scan rate and step size of 4°/min and 0.02°, respectively.

2.3. In vitro degradation behavior

In vitro degradation studies, including the electrochemical and immersion tests, were performed in a modified Hanks’ solution in naturally aerated condition at 37 ± 0.5 °C. Briefly, the solution was prepared with 9.7 g Hanks’ balanced salts (H1387, Sigma-Aldrich) together with 4.169 g of HEPES acid (Sigma-Aldrich), 16.65 g of HEPES sodium salt (Sigma-Aldrich) and 3.3 g of sodium bicarbonate (Sigma-Aldrich) dissolved in 1.4 L deionized water [31, 46].

The electrochemical tests, including electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization tests, were performed in an electrochemical station (Princeton Versa STAT3, USA) [47, 48]. A three-electrode cell was applied with Ag/AgCl saturated KCl as the reference electrode, platinum plate as the counter electrode, and the sample as the working electrode (exposed surface area = 0.2826 cm²). After immersion in the solution for 10 min to reach a steady open circuit potential, the EIS test was performed from 10⁵ Hz down to 10⁻² Hz with a potential amplitude of 10 mV. Afterward, the potentiodynamic polarization test was performed at a potential range of ±0.5 V relative to the open circuit potential with a scanning rate of 1 mV/s, which was routinely used in many previous studies [11, 22, 33, 49]. The polarization data was analyzed by CorrView software. The corrosion rate (CR) was calculated according to the following equation [47, 50]:

$$CR=3.27\times {10}^{-3}\cdot \frac{i_{\text{corr}}}{\rho }EW=14.98\times {10}^{-3}{i}_{\text{corr}}$$

where i_corr is corrosion current density (μA/cm²), ρ is the density of corroding materials (g/cm³), EW is the corresponding equivalent weight (g).

The immersion degradation behaviors were tested for 2 months according to the previous study [31]. Briefly, each sample was socked in the modified Hanks’ solution at 37 °C. The pH value of the solution was monitored during the immersion tests. The surface morphologies and phase composition of degraded samples after 2 months of immersion were observed with SEM, EDS, and XRD.

2.4. Hemocompatibility evaluation

The hemolysis tests and platelet adhesion tests were performed according to the method described previously [33, 51]. In brief, healthy human blood (anticoagulant with 3.8% citric acid sodium, Zen-Bio, US) was diluted by 0.9% sodium chloride solution with a volume ratio of 4:5. All samples were pre-cultured with 9.8 ml 0.9% sodium chloride solution at 37 °C for 30 min and 0.2 mL diluted blood was then added to each tube and incubated at 37 °C for 60 min. Deionized water and 0.9% sodium chloride solution were incubated with 0.2 mL diluted blood as the positive and negative control, respectively. After centrifuging at 3,000 rpm for 5 min, the supernatants were collected in 96-well plates and the absorbance (A) was measured by a plate reader (Cytation 5, Biotek, US) at 545 nm. The hemolysis rate (HR) was calculated by the following equation: Hemolysis = (A_sample – A_negative)/(A_positive – A_negative).

Platelet rich plasma (PRP) with 10⁸/μl platelets, purchased from Zen-Bio, Inc., was used for platelet adhesion test. 80 μl PRP was overlaid on each sample surface and incubated at 37 °C for 1 hour. After gently rinsed by PBS for 3 times to remove the non-adherent platelets, adherent platelets on samples were fixed with 4% paraformaldehyde (PFA, Affymetrix, US) and 2% glutaraldehyde solution (Fisher Chemical, US) at room temperature for 2 h, then dehydrated with gradient alcohol solution (30%, 50%, 70%, 90%, and 100%) and hexamethyldisilazane (HMDS) for 10 min, respectively, and finally dried in desiccator. The samples were coated by gold and observed by SEM. The number of adherent platelets was counted by Image J software on at least five different SEM images for each sample.

2.5. Cell Viability

Murine calvarial pre-osteoblasts (MC3T3-E1, ATCC CRL-2593, US) and human endothelial cells (EA.hy926, ATCC CRL-2922, US) were adopted to evaluate the cytocompatibility of uncoated and coated samples. The cell culture procedure can be found in previous publications [51-53]. For the indirect assays, cells were exposed to extract media from degraded samples according to ISO 10993-5 and −12. Briefly, Minimum Essential Media Alpha (MEM-α, Gibco, US) and Dulbecco’s Modified Eagle Medium (DMEM, ATCC, US) with 10% serum were used as the culture media for pre-osteoblasts and endothelial cells, respectively. Extract media was prepared by incubating samples in the corresponding cell culture media at a ratio of 1.25 mL/cm² for 3 days. Afterward, the collected extract solution was diluted with culture media to specific concentrations of 50% and 25%. The Zn ion concentration of different extract media was measured with a Zn colorimetric assay kit (Bio-Vision, US). The cell viability was measured with the MTT assay (Thermo Fisher Scientific, US) [6, 54]. Briefly, cells were seeded in 96-well plates at a density of 2,000/well and cultured for 24 h. Afterward, the original media was replaced with different concentrations of diluted extracts (100%, 50%, and 25%), and the cells were further cultured for 1, 3, and 5 days. Culture media without extract media was used as the control. The absorbance (A) was measured by a plate reader (Cytation 5, Biotek, US) at 562 nm.

2.6. Fluorescence staining and cell morphology

Cells were seeded on glass slides in 24-well plates at a density of 3,000/well and cultured for 24 h. Afterward, the original media was replaced with the 50% diluted extract media and the cells were further cultured for 3 days. Culture media without extract media was used as the control. Cells were rinsed gently by PBS and fixed with 4% PFA solution for 15 min at room temperature. After that, cells were permeabilized and blocked for 1 h at room temperature by using the PBST solution (PBS with 0.2 % Triton X-100) containing 0.03 g/mL bovine serum albumin (BSA, Sigma-Aldrich, US) and 0.1% goat serum (Sigma-Aldrich, US). The samples were then incubated with anti-paxillin rabbit monoclonal antibody with 1:200 dilution in PBST (Abcam, US) for 2 h, followed by Alexa Fluor 555 goat anti-rabbit secondary antibody with 1:200 dilution in PBST (Life Technologies, US) labeling for 1 h. F-actin was stained with Alexa Fluor 488 phalloidin with 1:200 dilution in PBST (Life Technologies, US). Slides were mounted using ProLong Gold Antifade Reagent with 4,6-diamidino-2-phenylindole (DAPI, Life Technologies, US). Images were taken by a Nikon Ti eclipse fluorescence microscope.

2.7. Cell adhesion and morphology

Cells were seeded onto the coated and uncoated surfaces in 24-well plate at a density of 1×10⁴/well and cultured at 37°C for 3 days. The cell morphology was observed by SEM after being fixed and dehydration in the same method with the platelet adhesion test as described above.

2.8. Cell differentiation

The differentiation of pre-osteoblasts was measured by alkaline phosphatase (ALP) staining and Activity Assay Kit (Abcam, US) as described previously [55]. In brief, pre-osteoblasts were precultured in a 6-well plate at a seeding density of 5×10⁴ cells/well for 24 h. The cell culture media was replaced with the 50% diluted extract media and refreshed every 3 days. Culture media without extract media was used as the control. Cells were further cultured for 14 days, rinsed by PBS, and fixed in 4% PFA for 30 min. After fixing, cells were washed by PBS for 3 times and incubated with NBT (nitro-blue tetrazolium chloride) and BCIP (5-bromo-4-chloro-3'-indolyphosphate p-toluidine salt) solution (Thermo Scientific, US) for 30 min at room temperature. Afterward, cells were rinsed by PBS with 0.2% Triton X-100 for 3 times. Images were taken by a Zeiss microscope.

ALP activity was measured according to the manufacturer's protocol. Briefly, the cells were lysed by Cell Lysis Buffer and incubated for 10 min at 4°C. The cell lysate was collected and centrifuged at 12,000 × g for 10 min. The supernatant was collected, and a bicinchoninic acid assay was performed to determine the total protein in the supernatant. The supernatant was mixed with equal volume of ALP Activity Assay Substrate and incubated for 30 min. The absorbance was read at 405 nm and the ALP activity was calculated and normalized by the total protein concentration.

The osteogenic differentiation of pre-osteoblasts was also assessed by Alizarin Red Staining (Sigma, US). Cells were precultured in a 6-well plate at a seeding density of 5×10⁴ cells/well for 24 h. The cell culture media was replaced with the 50% diluted extract media and refreshed every 3 days. Culture media without extract media was used as the control. Cells were further cultured for 21 days, rinsed by PBS, and fixed in 4% PFA for 30 min. After fixing, cells were washed by deionized water. The alizarin red solution (40 mM) was added and cells were kept at room temperature for 30 min and then rinsed with deionized water. Images were taken by a Zeiss microscope.

2.9. Antibacterial property

E. coli (ATCC 25922, US) was cultured according to the procedures in the previous study [56]. Briefly, the bacteria in frozen stock were cultured in Lysogeny broth (LB) media at 37 °C and 220 rpm to reach the optical density of 0.5-0.6 at 600 nm, corresponding to the bacteria density of approximately 4 × 10⁸ colony forming units (CFU) mL⁻¹. Then, 2 ml of the diluted bacterial suspension with a concentration of 1 × 10⁷ CFU mL⁻¹ in LB media was incubated with samples for 24 h at 37 °C and 120 rpm. Plain suspension was used as the blank control. The absorbance of the collected bacteria suspension was read at 600 nm. Antibacterial rates for E. coli in the media were calculated with the following equation: Antibacterial rates = (A_negative – A_sample) / A_negative. The Zn ion concentration of the collected bacteria suspension was measured with a Zn colorimetric assay kit (Bio-Vision, US). Samples incubated with bacteria were observed by SEM after being fixed and dehydration using the same method with the platelet adhesion test as described above. The number of adherent bacterial was counted by Image J software on at least five different SEM images for each sample.

2.10. Statistical analysis

Data were expressed as mean ± standard deviation. Statistical analysis was performed with SPSS 18.0 software package (SPSS Inc. Chicago. USA) between different groups by one-way analysis of variance (ANOVA) followed by post hoc Turkey’s multiple comparison tests. Statistical significance was considered at p < 0.05.

3. Results

3.1. Coating optimization and characterization

Fig. 1 showed the typical surface morphologies and the corresponding EDS and XRD results of the ZnP coatings prepared at various pH values with the collagen (Col) coating on pure Zn serving as a comparative benchmark control. It could be observed that the Col coating was uniform on the Zn surface with a coating thickness of around 2.5 μm (Fig. 1 a, b), corresponding to the carbon (C) appearance and higher oxygen (O) content in its EDS result (Fig. 1 i). ZnP coatings owned a similar coating thickness of 5-6 μm, but the coating morphology evolved with pH value. The coating formed at pH=2 had a finer flake-like structure with uniformly distributed pores (Fig. 1 c, d). Increasing pH to 2.5 resulted in a complete and uniform coating coverage with a combined micro-/nano-sized coating structure (Fig. 1 e, f). A few relatively large coating plates appeared sparsely at pH=3 (Fig. 1 g, h). There were no significant differences for the O and P content among these three ZnP coatings as revealed by EDS and XRD data (Fig. 1 i, j), indicating the pH values could only influence the coating morphology and thickness, while the coating composition did not change, mainly Zn₃(PO₄)₂·4H₂O (hopeite, PDF No. 37-0465). The uniform coating formed at pH=2.5 was selected as ZnP coating for the following degradation and biological tests.

Fig. 1.

Surface and cross-sectional coating morphology and phase composition: (a, b) Col coating, (c-h) ZnP coatings formed at (c, d) pH=2, (e, f) pH=2.5 and (g, h) pH=3, (i) elemental compositions (EDS), and (j) XRD patterns. Inserts are the cross-sectional coating morphology.

3.2. Electrochemical corrosion property

Fig. 2 showed the typical polarization and EIS curves for the uncoated samples and coated samples in Hanks’ solution at 37 °C. The electrochemical parameters obtained from the polarization curves were summarized in Table 1. The ZnP coating and Col coating showed different effects on the corrosion behaviors of pure Zn. When compared to that of pure Zn, the corrosion potential (E_corr) showed a shift toward the positive direction for ZnP coated sample but shifted negatively for the Col coated sample (Fig. 2a). This indicated the retarded and hastened polarization reaction of ZnP and Col coated samples, respectively, which was possibly due to the different solubility of two different coating materials, consistent with previous studies [57, 58]. The polarization current density (i_corr) and corrosion rate (CR) decreased, and the polarization resistance (R_p) increased for both coated samples (Table 1). Especially, i_corr and CR of ZnP coated sample were more than two orders of magnitude lower as that of pure Zn. Different shapes of Nyquist plots obtained from the EIS tests described the different types of electrochemical reactions taking place on the electrode surface (Fig. 2b). The Nyquist plot of pure Zn only contained one capacitance loop, which referred to the simple process at the solution/metal interface, indicating there was no corrosion product film formed after a short immersion period in Hanks’ solution, and this obviously was different from the long-term degradation process in the immersion test. It is noteworthy that the thin oxide film on the pure Zn surface formed in atmospheric environment disappeared or was transformed during the initial 10 min of immersion test, and it reached a steady open circuit potential before the subsequent EIS test and thus would not affect the EIS results significantly [59, 60]. The Nyquist plot showed two time constants for both coated samples. The higher-frequency time constant was contributed to the solution/coating interface while the lower-frequency one was related to the coating/substrate interface. The diameters of the two semicircles for the ZnP coated sample was greatly larger than those of pure Zn and Col coated sample (146 and 20 times, respectively), indicating its significantly higher corrosion protection.

Fig. 2.

Electrochemical corrosion behaviors of the pure Zn and different coatings on pure Zn: (a) Potentiodynamic polarization, (b) Electrochemical impedance spectroscopy (EIS).

Table 1.

Electrochemical corrosion parameters of different samples in modified Hank’s solution

Samples	Ecorr (mV)	icorr (μA/cm2)	Rp (Ω·cm2)	CR (mm/year)
Pure Zn	−1159 ± 13	28 ± 1	442 ± 13	0.42 ± 0.02
Col coated	−1243 ± 51	5.7 ± 0.6	3125 ± 343	0.085 ± 0.009
ZnP coated	−973 ± 22	0.24 ± 0.07	64766 ± 4078	0.004 ± 0.001

3.3. Immersion tests

The immersion degradation behavior of uncoated and coated Zn samples in the Hanks’ solution for 2 months was shown in Fig. 3. From the SEM images of the degraded surface, uncoated and Col coated samples showed localized corrosion pits on the surface (Fig. 3 a, e). Although their corroded surface morphology was not same, the corrosion products were similar and made of Zn(OH)₂ (sweetite, PDF No. 38-0356) and CaZn₂(PO₄)₂·2H₂O (CaZnP, scholzite, PDF No. 27-0095) and a small amount of ZnP (Fig. 3 j). There were three different typical morphologies on the degraded surface of the uncoated Zn, as shown in Fig. 3 b, c, and d, respectively, evolving from the slightly corroded flat area to the corrosion pit. Only some white spherical precipitates appeared on the slight degraded surface in the flat area (Fig. 3 b), corresponding to the CaZnP and ZnP. A dense and fine-structured layer made of nanorods clusters formed in the middle area (Fig. 3 c), which was mainly composed of Zn(OH)₂, as deduced from the EDS and XRD results. On the surface of the pitting area, there was a porous and coarse layer made of micro-sized particles and nanorods (Fig. 3 d) and composed of CaZnP and ZnP (Fig. 3 i, j).

Fig. 3.

Degraded surface morphology and composition after 2 months of immersion test in the Hanks’ solution: (a) general view and (b-d) magnified view of different areas of pure Zn surface, (e, f) Col coated surface, (g, h) ZnP coated surface, and the corresponding (i) EDS results and (j) XRD patterns; (k) evolution of pH values with immersion time.

The Col coating was corroded severely with cracks and pits on the surface (Fig. 3 e). The Col coating before corrosion tests did not show cracks on the surface (Fig. 1a-b), so the cracks were formed during the corrosion tests, indicating Col coated sample could not uniformly degrade in the Hanks’ solution for 2 months. The large amount of surface corrosion products (Fig. 3e-f) resulted in significantly increased O content and decreased C content as revealed by EDS (Fig. 3i) after the degradation test. The uniform degradation was dominant in ZnP coated sample. The ZnP coated sample retained its flake-like coating morphology, and the new formed coating flakes as shown in Fig. 3 h are corresponding to the formation of CaZnP, as indicated in Fig. 3 j. This was consistent with the appearance of Ca in the in vivo degradation products [17, 61].

The slight pH change for all the samples was presented in Fig. 3 k, and all samples possessed a similar pH evolution trend in the whole immersion period. The highest pH value appeared at 2 months for Col coated sample, which was also less than 8.0, indicating the stable pH environment with all the samples.

3.4. Hemocompatibility

The morphology and number of human platelets adhered on the surfaces were presented in Fig. 4 a-f and g, respectively. Numerous platelets adhered on the uncoated Zn surface with spreading long pseudopodia (Fig. 4 a), and some of them clumped together (Fig. 4 d). The number of the adhered platelets decreased significantly on the Col coated surface (Fig. 4 g), although there were also some spreading pseudopodia and aggregation (Fig. 4 e). There were cracks formed on the surfaces of pure Zn and Col coated samples during the degradation process in PRP (Fig. 4 a-b). This is consistent with many previous studies on Zn material surfaces [21, 33]. The quantity of adhered platelets on the ZnP coated surface significantly decreased and most of the platelets showed round morphology (Fig. 4 c, f, g). The hemolysis rates of pure Zn and both coated samples were far below 5% (Fig. 4 h), indicating the pure Zn and the surface coating were not hemolytic according to ASTM F 756-08 [62].

Fig. 4.

Hemocompatibility of different samples: (a-f) Platelets adhesion morphology on (a, d) pure Zn, (b, e) Col coating, (c, f) ZnP coating, and (g) the corresponding number of adhered platelets, (h) hemolysis percentage. *p < 0.05, **p < 0.005, compared between groups.

3.5. Cell Viability

Fig. 5 illustrated the cell viability of pre-osteoblasts and endothelial cells after 1, 3 and 5 days of incubation in different concentration of extract media with different samples and the Zn ion concentration of the corresponding extract media. The viability of both cells increased with the dilution of the extract media. The cell viability of pre-osteoblasts cultured in 100% extract media was ~20% on the first day and continuously decreased to ~7% on the fifth day (Fig. 5 a). The ZnP coated group significantly improved cell viability when compared to other groups in 50% extract media (Fig. 5 b). All the groups had good cytocompatibility (higher than 75%) in 25% extract media, while the ZnP coated groups showed even better viability than the control groups (Fig. 5 c). The trend of Zn²⁺ concentrations in the extract media was consistent with the immersion degradation test (Fig. 5 d, h). The two coatings decreased the degradation rate of Zn, and the released Zn²⁺ concentrations from ZnP coated sample was almost a half of that from pure Zn.

Fig. 5.

MTT assay for cell viability of (a-c) pre-osteoblasts and (e-g) endothelial cells cultured with different concentrations of extract media prepared by incubation with samples for 72h, and (d, h) Zn ion concentration of the corresponding extract media. #p < 0.05, compared with control group; *p < 0.05, **p < 0.005, compared between groups.

The endothelial cell viability showed similar trends but generally better results than pre-osteoblasts. The viability of all groups was similar and higher than 90% on the first day but decreased over time (~25% on the fifth day) in 100% extract media (Fig. 5 d). The ZnP coated groups also showed significantly higher viability than other groups in 50% and 25% extract media (Fig. 5 e, f).

3.6. Cell adhesion

Fig. 6 showed the SEM images of pre-osteoblasts and endothelial cells in direct contact with different sample surfaces. The pre-osteoblasts showed round morphology with limited spreading on the pure Zn surface. The cell spread more with filopodia on the Col coated surface, although the cells still displayed a round shape and poorly spreading morphology (Fig. 6 a, b, d, e). The endothelial cells showed a much more elongated morphology on both pure Zn and Col coated surfaces than pre-osteoblasts (Fig. 6 g, h, j, k). Both cells attached on the ZnP coated surface presented highly spreading morphology and interconnected to form a cell layer on the porous coating surface, which were significantly improved when compared to the pure Zn and Col coated surface (Fig. 6 c, f, i, l).

Fig. 6.

Cell adhesion morphology of (a-f) pre-osteoblasts and (g-l) endothelial cells with different samples for 72h: (a, d, g, j) pure Zn, (b, e, h, k) Collagen coating, (c, f, i, l) ZnP coating.

3.7. Fluorescence staining and cell morphology

Fig. 7 showed the fluorescence staining images of pre-osteoblasts and endothelial cells with different sample extract media. Both cells in the control group exhibited highly spreading, multi-polar morphology with reorganized actin fibers (Fig. 7 a, e). Consistent to SEM results above, both cells displayed round shape in pure Zn group (Fig. 7 b, f). The cells in the Col coated group showed the elongated cell morphology but less spreading and cell area than the control group (Fig. 7 c, g). The ZnP coated group significantly enhanced the expression of actin filaments and focal adhesions when compared to the pure Zn and Col coated groups (Fig. 7 d, h).

Fig. 7.

Fluorescent imaging of (a-d) pre-osteoblasts and (e-h) endothelial cells cultured with different samples with F-actin (green), paxillin (red) and DAPI (blue) staining, respectively; (a, e) control, (b, f) pure Zn, (c, g) Col coating, (d, h) ZnP coating.

3.8. Cell differentiation

The effects of uncoated and coated Zn on the cell differentiation of pre-osteoblasts toward osteoblasts were evaluated by the ALP tests and Alizarin red staining (Fig. 8). When compared to the cell control group, the pure Zn and Col coated groups showed a negative role on the ALP activity, while the ZnP coated group induced significantly improved ALP expression and activity of pre-osteoblasts (Fig. 8 a-e). Similarly, the ZnP coated group induced significantly improved calcific deposition, as indicated by its higher Alizarin red staining level in Fig. 8 i, when compared to the cell control group and other test groups. The results indicated the ZnP coating enhanced the differentiation of pre-osteoblasts.

Fig. 8.

Cell differentiation behavior of pre-osteoblasts cultured with different sample extracts. (a-d) ALP staining and (e) ALP activity and (f-i) alkaline red staining: (a, f) negative control, (b, g) pure Zn, (c, h) Col coating, (d, i) ZnP coating. *p < 0.05, **p < 0.005, compared between groups.

3.9. Antibacterial property

The antibacterial performance of different samples was shown in Fig. 9. The morphology and amount of bacterial on material surface indicated there was no formation of biofilm (Fig. 9 a-d). As compared to the pure Zn, there was much less bacterial adhesion and growth on the Col coating and ZnP coated surfaces, especially the ZnP coating showed a significant anti-adhesion performance to the bacterial cells, with more than 30 times less bacterial cells visible on the flake-like surface than pure Zn surface (Fig. 9 c). Antibacterial rates in the LB media were around 80% for all the groups but showed a different trend with the surface adhered bacterial number (Fig. 9 d). The pure Zn group showed slightly better performance than both coated groups, while the antibacterial rate of ZnP coated group was around 75%. The Zn ion concentration in the LB media had the same trend to the antibacterial rates in the media (Fig. 9 e). The Zn ion concentration significantly decreased in the media cultured with ZnP coated sample, which was consistent with the degradation behavior in Hanks solution and cell culture media (Fig. 3 and Fig. 5 d, h).

Fig. 9.

Antibacterial performance of different samples cultured with E. coli for 24 h. Bacterial adhesion on surfaces of (a) pure Zn, (b) Col coating, and (c) ZnP coating, and (d) the corresponding number of adhered bacterial cells; (e) Antibacterial rates in the LB medium; (f) Zn ion concentrations in LB medium after bacterial culture. #p < 0.05, compared with control group. *p < 0.05, **p < 0.005, compared between groups.

4. Discussion

As a recently emerged biodegradable metal, Zn has been indicated to own the moderate degradation rate as a potential biomedical implant material, especially for the cardiovascular stent application. In addition, the Zn ion plays significant roles in many enzymes and cell functions [7, 14, 15]. However, the RDA value for Zn is much lower than that of another biodegradable metal Mg [19], so its overdose toxicity is always a concern. Moreover, it has been found previously that the Zn ion induced biphasic cellular responses in vascular smooth muscle cells and endothelial cells. Zn ion is beneficial to cellular functions until it reaches limit of 60 – 80 μM (3.9 – 5.2 μg/ml) [54, 63]. Similarly, the viability of cells increased with the increased dilution of the extract media in this study, which was highly related to the Zn ion concentration in the extract media (Fig. 5). The cell viability could be higher than 80% when the Zn ion concentration was lower than ~15 and 7 μg/ml for preosteoclasts and endothelial cells, respectively. There were over 10 μg/ml/day Zn ions released in both cell media cultured with pure Zn (Fig. 5 d, h), which could explain why the cells in the direct assay had a round morphology and hindered adhesion (Figs. 6 d, j and 7 b, f), consistent with the previous studies [23, 64]. However, this data was based on static conditions in vitro, and it could be a different story for in vivo scenarios.

Surface modification is one of the most effective ways to control the ion release from the bulk material during the degradation and improve the surface biocompatibility for implants [65-67]. ZnP seems to have stable chemical property, good biocompatibility and is of easy synthesis, making it a promising coating material on the Zn-based biomaterials. In fact, ZnP cement has been applied at the tooth-prosthesis interface in dentistry for many years [68]. ZnP coating could be simply formed on biomedical metallic substrates and have been proven to be a good coating material to control the corrosion behavior of Fe and Mg alloys [40-42].

Here, ZnP coating on Zn substrate showed a good stability and degradation controlling ability, as indicated by the degraded coating morphology and decreased corrosion rate in Figs. 2 and 3. It is noteworthy that the 100% extract media of ZnP coated sample showed serious cytotoxicity for pre-osteoblasts in indirect assay, while its 50% extract media showed significantly improved cytocompatibility, and the improved and spreading cell morphology could be observed on the surface of ZnP coated sample in direct assay as well. A possible explanation is that the extract media in indirect assay was prepared via the incubation with the sample for 3 days resulting in a much higher accumulated Zn ion concentration in the exacts than the local ion concentration on the surface in the direct assay. This is also why it has been suggested by many previous studies to use diluted extracts for a more rationalized cell viability test of biodegradable metals [69, 70]. In addition, the cell adhesion and coverage on the sample surface in the direct assay, especially a complete cell coverage and cell layer formation on the ZnP coated sample surface as shown in this study, will affect the degradation process and ion release. The Zn ion release from ZnP coated sample in the cell culture media decreased significantly when compared to other groups (Fig. 5 d, h). In addition, the low concentration of extracellular Zn ion is shown to be beneficial to cell viability, adhesion, and spreading through enhancing the vascular and bone related gene expression profiles in our previous studies [54, 55, 63]. Therefore, the ZnP coated sample significantly improved the hemocompatibility (Fig. 4) and enhanced the adhesion and viability of pre-osteoblasts and vascular endothelial cells, as well as pre-osteoblasts differentiation (Figs. 5, 6, 7 and 8).

Zn is well-known for its inherent antimicrobial property and thus has been applied extensively as antibacterial materials, including ZnO films, Zn-incorporated coatings, cement, and nanocomposite scaffolds [71-73]. Zn ion released from the Zn-based implant materials could potentially interact with the bacteria surface to induce cell deformation and bacteriolysis and thus reduce the risk of bacterial infection and implant failure. Therefore, the ZnP coating should have the advantage of retaining the antibacterial ability in addition to controlling the degradation of the substrates compared to other types of surface coatings. In this study, ZnP coated samples showed an antibacterial rate of ~78% when incubated with E. coli in the LB media (Fig. 9 e, f).

In addition to the Zn ion, surface roughness and pattern could also affect the adhesion and morphology of the mammalian and bacterial cells [74, 75]. Data showed that the rough surface of the ZnP coated sample induced a significant cell adhesion and cell clusters formation (Fig. 6 c, f, i, l). It has been found previously that the increased surface roughness at microscale could lead to fewer cell contact areas and increased cellular stress, but the combined micro-/nano-scaled structure of ZnP coating in the present study could act as micro-spikes to facilitate the vascular and bone cell adhesion and spreading, consistent with previous studies [76, 77].

In contrast to the pre-osteoblasts and endothelial cells, the number of E. coli adhered on the ZnP coated surface decreased significantly, when compared to the pure Zn group. This was consistent with other studies indicating that the bacteria prefer the surface of microbial dimensions [75, 78]. The pure Zn and Col coated samples with flat surfaces showed good antibacterial adhesion performances and thus there was no biofilm formation on these surfaces. However, during the bacterial culture period in the LB media, the released Zn ion concentration was significantly higher than that in the cell culture media (Fig. 5 d, h, and Fig. 9 f). This means these samples degraded faster in the LB media, which was possibly due to lack of pH buffering capacity of LB media. Also, there was no biofilm formation on all the sample surfaces. Moreover, it has been found that the surface topography plays a critical role on the bacterial adhesion, but there is no linear dependence [79-81]. The surfaces before and after the degradation of the pure Zn and Col coated samples were flatter than that of ZnP coated samples, as shown in Fig. 1 a, b and Fig. 9 a, b. The surface morphology of pure Zn and Col coated samples showed similar topography (Fig. 9 a, b) and thus induce the similar bacterial adhesion behaviors. However, the surface morphology of the ZnP coated sample possessed a coarser and larger surface feature than the bacterial and thus inhibited the bacterial adhesion, consistent with a previous study [81]. The bacterial antiadhesion property is more important than the antibacterial rate in the media for the implant material to prevent the bacterial contamination and postoperative infection.

5. Conclusions

ZnP coating was prepared on the pure Zn by a simple chemical conversion coating method. The coating morphology could be controlled by changing the pH value of the coating solution. The ZnP coating controlled degradation rate and significantly improved hemocompatibility, cytocompatibility and antibacterial performance for the Zn biomaterials. The amount of released Zn ion accumulated in the local environment and the surface morphology were shown to be pivotal for the cytocompatibility and antibacterial performance of Zn material. In sum, the ZnP coating on Zn-based material could be an easy but effective way to enhance its biocompatibility and antibacterial property as well as to control its degradation rate.

Statement of Significance.

Zn and its alloys are promising biodegradable implant materials for orthopedic and cardiovascular applications. However, notable cytotoxicity has been reported due to degradation products accumulated in the local environment, largely overdosed Zn²⁺. Thus, controlling burst Zn²⁺ release is the key to minimize the toxicity of Zn implants. To achieve this goal, we prepared a homogenous ZnP coating on Zn metals thanks to its easy synthesis, stable chemical property, and good biocompatibility. Results showed that ZnP not only improved the cell viability, adhesion and proliferation, but also significantly reduced the attachment of platelet and bacterial. Therefore, ZnP could be a promising approach to improve the functional performance of Zn-based implants, and potentially be applied to many other medical implants.