TiO2 nanotube composite layers as delivery system for ZnO and Ag nanoparticles — An unexpected overdose effect decreasing their antibacterial efficacy

Materials Science and Engineering C 51 (2015) 158–166

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

TiO2 nanotube composite layers as delivery system for ZnO and Ag nanoparticles — An unexpected overdose effect decreasing their antibacterial efficacy A. Roguska a,⁎, A. Belcarz b, M. Pisarek a, G. Ginalska b, M. Lewandowska c a b c

Institute of Physical Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw, Poland Chair Department of Biochemistry and Biotechnology, Medical University of Lublin, Chodzki 1, 20-093 Lublin, Poland Faculty of Material Science and Engineering, Warsaw University and Technology, Wołoska 141, 02-507 Warsaw, Poland

a r t i c l e

i n f o

Article history: Received 14 October 2014 Received in revised form 16 January 2015 Accepted 24 February 2015 Available online 26 February 2015 Keywords: TiO2 nanotubes Ag nanoparticles ZnO nanoparticles Staphylococcus epidermidis Antibacterial properties Overdose

a b s t r a c t Enhancement of biocompatibility and antibacterial properties of implant materials is potentially beneficial for their practical value. Therefore, the use of metallic and metallic oxide nanoparticles as antimicrobial coatings components which induce minimized antibacterial resistance receives currently particular attention. In this work, TiO2 nanotubes layers loaded with ZnO and Ag nanoparticles were designed for biomedical coatings and delivery systems and evaluated for antimicrobial activity. TiO2 nanotubes themselves exhibited considerable and diameter-dependent antibacterial activity against planktonic Staphylococcus epidermidis cells but favored bacterial adhesion. Loading of nanotubes with moderate amount of ZnO nanoparticles significantly diminished S. epidermidis cell adhesion and viability just after 1.5 h contact with modified surfaces. However, an increase of loaded ZnO amount unexpectedly altered the structure of nanoparticle–nanolayer, caused partial closure of nanotube interior and significantly reduced ZnO solubility and antibacterial efficacy. Co-deposition of Ag nanoparticles enhanced the antibacterial properties of synthesized coatings. However, the increase of ZnO quantity on Ag nanoparticles co-deposited surfaces favored the adhesion of bacterial cells. Thus, ZnO/Ag/TiO2 nanotube composite layers may be promising delivery systems for combating post-operative infections in hard tissue replacement procedures. However, the amount of loaded antibacterial agents must be carefully balanced to avoid the overdose and reduced efficacy. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Extensive efforts have been undertaken in past decades to combat postoperative implant infections. Despite perioperative care, infectionassociated complications appear in 3–4% of all cases in this field. The consequences are usually severe because of frequent formation of bacterial biofilm on the implant surface. Therefore, different strategies are applied to overcome these difficulties. Systemic antibiotic treatment is often combined with local delivery of antibacterial agents (releasekilling mechanism) since 1970s [1]. Another possible strategy is the modification of implanted surface to lower the degree of bacterial cell adhesion [2,3]. Titanium dioxide (TiO2) has been widely studied as a coating material for orthopedic implants due to its excellent corrosion resistance and high adhesion strength to different substrates [4]. Recently, TiO2 in a form of nanotubes (NTs) has been reported to offer numerous benefits as a possible coating for biomedical devices. It was demonstrated that TiO2 nanotubes of different diameters can control the rate of cell ⁎ Corresponding author. E-mail address: [email protected] (A. Roguska).

http://dx.doi.org/10.1016/j.msec.2015.02.046 0928-4931/© 2015 Elsevier B.V. All rights reserved.

adhesion, growth and differentiation [5]. TiO2 NT layer showed the advantage over TiO2 layer naturally occurring on Ti surface at inhibition of bacterial adhesion and biofilm formation [6]. TiO2 NTs also offer the possibility to be loaded with antibiotics, therefore it can be considered as a local drug delivery system [7]. Antibiotics are the most popular antibacterial agents used in modern medicine. However, bacteria show an increasing resistance to these drugs, thus limiting their efficacy. In search for other effective bactericidal agents, inducing no or limited bacterial resistance, metals and metal oxide nanoparticles (NPs) were proposed. Among them, Ag, Cu, Au, CuO, Cu2O, ZnO and TiO2 are the most cited [8]. According to a common opinion, metallic NPs rarely induce bacterial resistance, although some reports on Ag-resistance are available [9–11]. Despite this, nanosilver is one of the most studied and well described nanometals of antibacterial activity [8]. It was confirmed that nano-Ag exhibits higher bactericidal potency than its bulk counterpart, mainly because of its larger specific surface area [12]. Zinc oxide NPs are another potential tool for combating of bacterial infections because zinc, although an essential metal ion, in higher concentration shows toxicity against bacteria (inhibits their respiratory electron transport) [13]. ZnO-implanted titanium not only inhibited growth of Gram-positive and Gram-negative

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

bacteria [14] but also significantly stimulated proliferation and ALP activity, collagen secretion and extracellular matrix mineralization of osteoblastic cells [15]. Although it was postulated that bacteria may develop the resistance to toxic zinc level due to numerous mechanisms (as summarized by Choudhury & Srivastava [16]), the occurrence of zinc resistance in bacteria is limited, as shown for periodontal species Prevotella intermedia and Porphyromonas gingivalis [17]. Also, toxicity of zinc, in comparison with other metals like Hg, Cd, Pb, and Ni, is quite low [18]. Therefore, the application of ZnO NPs may be considered as a beneficial alternative to Ag NPs in antibacterial prophylaxis. Modern antibacterial strategies propose the solutions based on synergistic action of two or more antibacterial agents, e.g. micropatterned surface plus antibiotic treatment, for enhanced bactericidal effect [19]. Combinations of different metallic agents were also reported to exhibit high antibacterial activity. For example, TiO2/ZnO powders [20] and ZnO/TiO2 composite nanofibers [21] were found to be highly effective against Escherichia coli cells, especially after induction by UV light. ZnO nanorods with attached Ag NPs were investigated for antibacterial activity and showed an enhanced and synergistic effect [22]. However, beneficial effect of combination of two antibacterial agents is not always obvious. For silver-doped ZnO NPs prepared by solution route spincoating process, the minimal inhibitory concentration (MIC) remained constant for all concentrations of Ag [23]. Importance of both high specific surface area of metallic NPs and their moderate solubility was postulated to be crucial for efficacy of metallic NP-based antibacterial materials [24]. In this work, we combined all the strategies mentioned above by the deposition of two types of NPs, Ag and ZnO, on TiO2 NT layers fabricated on Ti surface. Samples with different amounts of loaded nanoparticles were designed to enable the evaluation of dosage effect on composite layers antibacterial activity. NPs penetrated deeply the interior and exterior of TiO2 tubes. Therefore such obtained surfaces were expected to release Ag+ and Zn2 + ions in a sustained manner. We evaluated the antibacterial properties of the composite coatings taking into consideration both release-killing and contact-killing mechanisms. Staphylococcus epidermidis bacterial strain, a pathogen responsible for over 90% of implant-associated osteomyelitis [25], was selected for this pilot study as a model microorganism. 2. Methods 2.1. Preparation of composite layers The TiO2 NT layers were fabricated by the electrochemical anodization of Ti samples (Ti foil, 0.25 mm-thick, 99.5% purity) in an optimized electrolyte: NH 4 F (0.86 wt.%) + deionized (DI) water (47.14 wt.%) + glycerol (52 wt.%) under a constant voltage of 20 and 30 V. After anodization, the samples were rinsed with DI water and dried in air. Subsequently, thermal annealing was performed at 650 °C for 3 h to transform the TiO2 NT structure from amorphous to crystalline (mixture of anatase and rutile) and to obtain mechanically stable nanotubes well integrated with Ti substrate. Silver (15 nm thick layer) was deposited using the magnetron sputtering technique (Leica EM MED020) in a vacuum chamber with a base pressure of about 2 × 10−5 Pa. The thickness value is a nominal value, as measured on the surface of a quartz crystal micro balance. ZnO was electrodeposited from 0.005 M Zn(NO3)2 aqueous solution at a potential of − 2.0 V vs SCE at 60 °C. The duration of the deposition was 1, 3 or 5 min. After the deposition, the samples were rinsed with DI water and annealed at 300 °C for 15 min.

159

5 kV. For chemical characterization a Thermo Noran X-ray energy dispersive spectrometer (EDS) coupled with SEM was used. Analysis was performed at the accelerating voltage of 8 kV. The release of silver and zinc from modified titanium samples was measured by inductively coupled plasma mass spectrometry (ICP-MS, Elan 9000 Perkin-Elmer). The samples were incubated in distilled water (15 ml) for 1.5, 3 and 24 h at room temperature without stirring. The amount of metal released was determined in resulting solution. 2.3. Bacterial strain and medium Reference bacterial strain from American Type Culture Collection, S. epidermidis ATCC 12228, was maintained as a stock in sterile Microbanks (Technical Service Consultants Limited, UK) at −20 °C. Before the experiment, bacterial cells were transplanted onto fresh slant culture medium (Mueller-Hinton Agar) and cultured for 20–24 h at 37 °C. Afterwards, the bacteria were scraped out from agar medium into liquid medium (Mueller-Hinton Broth) and cultured for 20–24 h at 37 °C. Resulting suspension was diluted to appropriate density directly before the experiment. 2.4. Release-killing antimicrobial activity evaluation Evaluation of antimicrobial activity of tested titanium-based samples was performed according to standard method JIS Z 2801:2000 [26]. Survival of bacterial cells was evaluated after 1.5 h, 3 h and 24 h incubation at 37 °C, with separate controls for each incubation period. Results of three experiments were calculated as means ± SD. 2.5. Contact-killing antimicrobial activity evaluation (bacterial adhesion test) Samples, defatted and cleaned in acetone bath (10 min, twice) followed by chloroform (10 min, twice) and ethanol (10 min, twice), were placed in 24-well plates (Costar, Corning Inc., USA) and sterilized by ethylene oxide method in paper/plastic peel pouch (1 h at 55 °C, followed by 20 h aeration). 1 ml of sterile suspension of bacterial cells (approx. 1.0 × 108 cells/ml) in Mueller-Hinton broth, standardized to the McFarland Equivalence Standards using PhoenixSpec nephelometer (Becton Dickinson, USA) was added to each well. After 1.5 h incubation of plates at 37 °C, without shaking, non-adhered bacteria were gently washed away with 0.9% NaCl (50 ml, 3 times) and samples were subjected to evaluation of adhered bacterial cells. For this purpose, wet samples were incubated with Viability/ Cytotoxicity Assay Kit for Bacteria Live & Dead Cells (Biotium, USA) in 0.9% NaCl (according to manufacturer instructions). Dyes used in this kit show different properties. DMAO preferentially binds to dsDNA of both damaged and intact cells and stains live and dead cells green. Ethidium homodimer III overlaps green fluorescence produced by DMAO absorption in cells with damaged cell membranes and stains only dead cells (red fluorescence). Thus, cells with greenish-yellow fluorescence can be considered as severely damaged or dying. After staining at R/T, 15 min, in darkness, plates were washed twice in 0.9% NaCl to remove non-absorbed dye. The bacteria were visualized by fluorescence microscopy (Olympus BX41 with UC 12 Soft Imaging System camera). The live, dying and dead bacterial cells were counted using CellSens Dimension program (Olympus, Japan). The results are the means ± SD of two independent experiments using 10 randomly chosen samples from each repeat. 2.6. SEM observation of bacteria adhered to Ti surfaces

2.2. Physicochemical characterization of composite layers For the morphological characterization of the composite layers fabricated microscopic observations were carried out with a scanning electron microscope (SEM, Hitachi S-5500) at the accelerating voltages of

In one experiment, titanium samples subjected to bacterial adhesion were stained for SEM procedure in 2.5% glutaraldehyde in ultrapure water (1 h), washed 3 times in PBS (phosphate buffered saline; 15 min) and dehydrated in series of 30%, 50%, 70%, 80% and absolute

160

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

ethanol (10 min each). Dried samples were subjected to SEM observation. 3. Results 3.1. Sample characterization The optimized anodization conditions resulted in the formation of TiO2 nanotubes perpendicular to the substrate and separated from each other. The nanotubes, with an average diameter of ~100 nm for anodization potential of 20 V and ~150 nm for 30 V, were open at the top and closed at the bottom. The height of the nanotube layers was about 1 μm. Fig. 1 shows SEM images of typical TiO2 NT layers fabricated at 20 V (Fig. 1a–c) and 30 V (Fig. 1d–f) after electrodeposition of ZnO NPs for different times. After 1 min of electrodeposition, only some single spherical nanoparticles could be found on the nanotube edges (Fig. 1a and d). Electrodeposition of ZnO NPs for 3 min led to the formation of spherical nanoparticles tightly covering the nanotube edges and uniformly distributed over the nanotubular layer (Fig. 1b and e). Size of these nanoparticles is about 10 nm. TiO2 NT loaded with ZnO NPs for 5 min exhibited quite different morphology. The increase of electrodeposition time led to the formation of heterogeneous coatings with some areas of the TiO2 nanotube surface completely covered by the agglomerated ZnO deposits (Fig. 1c and f). Fig. 2 shows a typical SEM image of TiO2 NT covered with 15 nm of Ag without and with ZnO NPs. Detailed inspection of Fig. 2a revealed that silver did not form a continuous layer but an inhomogeneous film with island-like agglomerations of nanoparticles. These structures appeared due to the highly developed specific surface area of the titanium oxide nanotubes and a very low amount of a metal deposited. The island-like agglomerations were located mainly on the top edges of the nanotubes. Some single spherical Ag nanoparticles with diameter below 50 nm were also found on the inner and outer side walls of the tubes. Our previous results revealed, however, that in contrast to ZnO nanoparticles which were present on the whole length of the TiO2 nanotubes, the distribution of Ag nanoparticles exhibited the in depth gradient. The highest amount of Ag was located at the nanotube top edges and decreased along the side wall of the nanotube to reach the lowest value at the bottom of the tube. The detailed description of the distribution of both types of nanoparticles in the TiO2 nanotube layers is given elsewhere [27]. Fig. 2b–d shows the ZnO NPs electrodeposited for different times on the TiO2 NT layer previously covered with 15 nm Ag. As for samples without Ag NPs, the deposition of ZnO NPs for 1 min led to the formation of single particles on the nanotube edges (Fig. 2b). Electrodeposition of ZnO for 3 min led to the formation of elongated and pointed particles,

similar in shape to a rice bean (Fig. 2c). The length of these particles was below 100 nm. They were located at the nanotube top edges as well as on the nanotube inner and outer side walls. Increasing the time of electrodeposition to 5 min resulted in the higher population of ZnO NPs and a complete coverage of the TiO2 nanotubes surface (Fig. 2d). The chemical composition of the modified surfaces was analyzed using EDS technique. In general, increasing the time of ZnO electrodeposition led to the higher concentration of Zn at the TiO2 NT surface. However, for wider nanotubes (150 nm; 30 V) the amount of Zn was higher (for the same time of electrodeposition) than for smaller nanotubes (100 nm, 20 V). The highest concentration of Zn was observed for nanotube layer previously covered with 15 nm of Ag. For instance, 3 min of ZnO electrodeposition resulted in 3.8 wt.% Zn content for nanotubes fabricated at 20 V (100 nm), 5.2 wt.% for 30 V (150 nm) nanotubes and 9.6 wt.% for 30 V nanotubes with Ag nanoparticles. The concentration of Ag for sample without ZnO was about 25 wt.% and it has decreased for samples with electrodeposited ZnO to about 18–21 wt.%. This may be the effect of silver dilution during the ZnO electrodeposition process. Full characteristics of fabricated samples were summarized in Table 1. 3.2. Zinc and silver release Fig. 3a shows the profile of zinc release from tested samples. The concentration of released metal increased with time for all surfaces. Surprisingly, higher content of loaded ZnO did not correlate with higher amount of released Zn. The highest levels of Zn in medium were found for surfaces deposited for 3 min. This observation appeared for samples modified with TiO2 NT of 100 nm in diameter (C–E), of 150 nm in diameter (G–I) and for samples co-modified with Ag NPs (K–M). As expected, higher release of Zn was observed for samples modified with 150 nm TiO2 NT (G–I), with larger quantities of loaded ZnO NPs, than for respective samples modified with 100 nm TiO2 NT (C–E) (Table 1; Fig. 3a). Profile of Ag release was presented in Fig. 3b. The amount of solubilized silver increased with time. However, samples with additionally electrodeposited ZnO NPs (samples K–M) released Ag on a significantly lower level. Moreover, amounts of Ag were similar for all Ag/ZnO NPsdeposited surfaces (samples K–M). It seems that external layer of electrodeposited ZnO particles acted as a barrier inhibiting the release of silver sublayer. 3.3. Release-killing antimicrobial activity evaluation In this test, we determined the survival of S. epidermidis cells in suspension incubated with titanium samples modified with bactericidal

Fig. 1. SEM images of ZnO nanoparticles electrodeposited on TiO2 nanotube-modified titanium surfaces. Nanotubes were fabricated at 20 V (a–c) and 30 V (d–f) and electrodeposited ZnO NPs for 1 min (a, d), 3 min (b, e) or 5 min (c, f).

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

161

Table 1 Composition of tested samples. Zn and Ag atomic content was estimated by EDS technique. Sample code

TiO2 nanotubes (ø)

Antibacterial nano-agent (application time or layer thickness)

Zn content (wt.%)

Ag content (wt.%)

A B C D E F G H I J K L M

– TiO2 NT 20 V (100 nm) TiO2 NT 20 V (100 nm) TiO2 NT 20 V (100 nm) TiO2 NT 20 V (100 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm) TiO2 NT 30 V (150 nm)

– – ZnO (1 min) ZnO (3 min) ZnO (5 min) – ZnO (1 min) ZnO (3 min) ZnO (5 min) Ag (15 nm) Ag (15 nm) + ZnO (1 min) Ag (15 nm) + ZnO (3 min) Ag (15 nm) + ZnO (5 min)

– – 1.9 3.8 16.4 – 3.1 5.2 22.1 – 2.4 9.6 42.9

– – – – – – – – – 25.5 17.9 20.5 21.3

agents: nanoZnO and nanoAg particles. The amount of surviving cells was evaluated and calculated as a percent of that in control bacterial suspension. Results, presented in Fig. 4, were compared with reference samples: pure titanium (sample A) and titanium modified with 100 nm and 150 nm nanotubes (samples B and F, respectively). The bactericidal effect of the Ti samples observed after 24 h incubation was significant and reached 100% (complete lack of surviving bacteria) for all nanoZnO-modified samples. However, no differences between the bactericidal activities of tested samples could be distinguished. Therefore, the time of incubation with bacterial suspension was reduced, initially to 3 h and finally to 1.5 h. 1.5 h incubation time revealed the differences between the tested samples. The following observations were made (Fig. 4): (1) For samples with nanoZnO particles only, a medium (3 min) amount of loaded ZnO was the most effective at S. epidermidis killing. A higher amount of loaded nanoZnO resulted in decreased bactericidal activity. (2) For samples with both nanoZnO and nanoAg particles (samples K, L and M), the amount of loaded nanoZnO did not affect the level of bactericidal activity in any particular way. The survival

of S. epidermidis was low (3.5–6.1%) for all amounts of nanoZnO loaded samples. Moreover, sample J, loaded solely with nanoAg (without nanoZnO) also exhibited high bacterial-killing force (only 16.7% survival). (3) TiO2 NTs themselves exhibited a significant bactericidal activity. This effect was stronger for nanotubes of higher diameter (150 nm) — only 43.8% survival of bacterial cells in comparison with 73.3% survival observed for 100 nm TiO2 NT. Similarly, nanoZnO-modified samples showed higher bactericidal activity when NPs were loaded on the samples modified with 150 nm TiO2 NT than with 100 nm TiO2 NT.

3.4. Contact-killing antimicrobial activity evaluation (bacterial adhesion tests) In this experiment, the amounts of S. epidermidis cells adhered to tested samples were evaluated. 1.5 h incubation time was selected based on the results of previous experiment. Longer time of experiment was expected to reduce completely the survivability of bacteria. The results were summarized in Figs. 5 and 6 and in Table 2.

Fig. 2. SEM images of TiO2 nanotube-modified titanium surfaces (30 V) covered with 15 nm of Ag without ZnO NPs (a), with electrodeposited for 1 min ZnO NPs (b), 3 min ZnO NPs (c) and 5 min ZnO NPs (d).

162

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

Fig. 3. The release of zinc (a) and silver (b) from modified titanium samples after 1.5 h, 3 h and 24 h incubation in distilled water. Sample symbols as described in Table 1.

It was found that bacteria, mainly live ones, adhered abundantly (≈ 147,000–199,000 cells/mm2) not only to titanium modified with 100 nm and 150 nm nanotubes (samples B and E, respectively) but also to control titanium (≈63,000 cells/mm2; sample A). Numerous dividing cells were observed on these surfaces (Figs. 5, 6). Much less adhered cells were detected on titanium/TiO2 NT surfaces loaded with nanoAg and nanoZnO particles. The only exception was sample G (titanium/ 150 nm TiO2 NT/nanoZnO 1 min), with ≈ 200,000 adhered cells/mm2 of surface; however, the cells were mainly dead or dying (Table 2). SEM observations (Fig. 6) suggested that significant morphological changes occurred in few bacterial cells adhered to modified surfaces (samples D and E). Between samples modified with TiO2 NT and nanoZnO, the following regularity was observed: medium amount of ZnO (3 min) caused the highest reduction of number of adhered bacterial cells among all tested samples (Table 2). This phenomenon was observed both for 100 nm NT (samples C–E) and 150 nm NT (samples G–I). In the first group, the number of adhered cells was only ≈3700 cells/mm2 (sample D), whereas in the second group, the amount of adhered cells was ≈14,000 cells/mm2 (sample H). In these considered cases, live adhered bacteria dominated over dead adhered bacteria, except of sample G. For samples modified with nanoAg (samples J–M), the lowest amount of adhered cells was observed on surfaces without additional deposition of ZnO NPs (sample J) (Table 2). As the amount of loaded nanoZnO NPs increased, the increasing number of adhered bacterial cells was observed. This suggests that the co-presence of nanoZnO and nanoAg layers suppressed the anti-adhesive properties of nanoAg particles, possibly by the formation of structural complexes between Ag and ZnO NPs (Fig. 2). Bacteria adhered to the surfaces of these samples were

Fig. 4. Survival of S. epidermidis cells after incubation of bacterial cell suspension with tested surfaces for 1.5 h, 3 h and 24 h. Sample symbols as described in Table 1.

mainly dead and dying, except for sample K (Table 2). The most significant changes in morphology of adhered cells were observed on samples with medium (3 min) content of nanoZnO (sample L) (Fig. 6). 4. Discussion The primary aim of this study was the estimation of antibacterial properties of ZnO nanoparticles deposited on titanium surface modified with TiO2 nanotubes layer. ZnO NPs were already reported as a potent tool for the inhibition of bacterial growth [8,14]. Moreover, Zn ion plays an important role in control of osteoblast cell function [28]. It was therefore expected that surfaces modified with ZnO NPs could show the combined activities: antibacterial and bone-forming ones. Such surfaces would be probably highly beneficial for the modification of titanium-based bone implants. TiO2 nanotubes used for the modification of titanium surfaces showed significant bactericidal activity even without deposition of nanometal particles. This observation is in agreement with our previous results [6]. The mechanism of TiO2 toxicity towards bacterial cells is based on the release of reactive oxygen species (ROS) which appears even without photooxidation [29,30]. More significant bactericidal effect (lower survival rate of bacterial cells) was observed for surfaces modified with TiO2 nanotubes of larger diameter. This can be explained by faster diffusion of liquids throughout the interior of wider nanotubes and faster release of active ROS. The same phenomenon was observed for ZnO-deposited surfaces: lower rate of bacterial survival appeared for ZnO-deposited samples coated with TiO2 NT of higher diameter. This suggests the cumulative effect of both TiO2 nanotubes and ZnO nanoparticles as antibacterial agents. We also observed different rates of bacterial cell adhesion to surfaces modified with TiO2 nanotubes of different diameters. Bacteria adhered more abundantly to samples coated with wider nanotubes. According to a common opinion, cells (eukaryotic and bacterial) prefer roughened than smooth surfaces. Brammer et al. [5,31] reported that osteoblast adhesion and differentiation depended on the diameter of TiO2 NT. Our observations are therefore in agreement with these data. However, one should note that the bacteria adhered to TiO2 NT modified surfaces were mainly live. This suggests that TiO2 nanotubes-coated surfaces may — to the some extent — promote the adhesion of bacterial cells and formation of bacterial biofilm. Therefore, the protection of such surfaces with additional antibacterial agents is highly recommended. All proposed systems containing Ag and ZnO NPs showed very high bactericidal activity against S. epidermidis model strain. The mechanism of their action is based on Ag and ZnO NP solubility (which was presented in Fig. 3) and subsequent effect of resulting metal ions. Zn2 + ions were proved to have an inhibitory effect (in excessive concentration) on bacterial processes. Namely, they bind to bacterial membrane and prolong the lag phase of the growth cycle [32], modulate membraneassociated enzymes, impair calcium uptake and change membrane fluidity [33]. Mechanism of Ag+ action on bacteria is still not fully clear but

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

163

Fig. 5. Fluorescence microscopy images of S. epidermidis cells adhered to tested surfaces after 1.5 h incubation with bacterial cell suspension. Live cells were stained green, dead cells were stained red. Sample symbols as described in Table 1.

two of the most possible mechanisms include: (i) disruption of ATP synthesis apparatus by Ag+ binding to thiol groups of respiratory and transport proteins [34,35] and (ii) disruption of DNA replication due to Ag+dependent increase of membrane permeability [36]. In our study, highest antibacterial activity (lowest bacterial survival rate) of combined Ag + ZnO NP system suggests the cumulative antibacterial effect of Ag+ and Zn2+ ions released from dissolving nanoparticles. However,

lowest bacterial survival rate observed for Ag + ZnO NPs system is not correlated with lowest bacterial adhesion to tested surfaces. Bacteria (although mainly dead) adhered more abundantly to Ag + ZnO NPs modified surfaces when compared to ZnO NPs modified surfaces (especially for 3 and 5 min-deposited samples). This phenomenon results probably from more developed surface topography of Ag + ZnO NP system. Therefore, to evaluate in vitro antibacterial efficacy of surfaces

164

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

Fig. 6. SEM images of glutaraldehyde-fixed S. epidermidis cells adhered to tested surfaces after 1.5 h incubation with bacterial cell suspension. Sample symbols as described in Table 1.

designed for biomedical applications, both bacterial-killing and adhesion-inhibiting properties should be considered because these two tests may produce contradictory results. Vargas-Reus et al. [37] evidenced that Ag NPs showed higher antimicrobial capacity than Ag-ZnO NPs. In contrast, we showed that the codeposition of nanoAg and nanoZnO particles on tested surfaces resulted in an enhanced antibacterial properties when compared to solely nanoAg-deposited samples. Moreover, as one should note, samples loaded solely with nanoAg particles exhibited a significant bactericidal

effect (over 80% killed bacterial cells after 1.5 h incubation). Surprisingly, we have previously reported that nanoAg-loaded TiO2 NT showed only slight bactericidal activity after 24 h incubation with S. epidermidis suspension [6]. One of the most possible explanations of these differences may lie in various Ag NP sizes because Cao et al. reported that small and large Ag NPs acted distinctively on bacteria: large ones induced serious toxicity while small ones did not [38]. However, this explanation is not relevant in our case because Ag NPs in both our studies were similar and smaller than 50 nm. More probably, these differences come out

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166 Table 2 Amount of S. epidermidis cells adhered to tested samples after 1.5 h incubation with bacterial suspension (105 CFU). Results were calculated as means ± SD from 5 measurements. Adhered bacterial cells (cells/mm2)

A B C D E F G H I J K L M

Live

Dead or dying

Total

27 763 ± 821 107 643 ± 16,767 29 344 ± 7460 3 244 ± 1102 55 168 ± 4609 4 348 ± 2228 192 431 ± 16,036 12 486 ± 1424 20 523 ± 4974 1 840 ± 247 56 786 ± 127 2 058 ± 802 4 324 ± 303

34 995 ± 14,304 39 386 ± 3102 3 583 ± 413 471 ± 207 8 656 ± 2990 191 932 ± 29365 6 582 ± 2661 1 566 ± 878 1 758 ± 1068 19 259 ± 4958 9 297 ± 483 71 651 ± 16,701 92 624 ± 3586

62 758 ± 15,126 147 029 ± 16,706 32 927 ± 7722 3 716 ± 1245 63 825 ± 6910 196 280 ± 30,890 199 013 ± 16,970 14 052 ± 1311 22 281 ± 5590 21 100 ± 4711 66 084 ± 611 73 710 ± 17118 96 948 ± 3849

from the details of nanosilver deposition methods used in both experimental series: magnetron sputter deposition used in this work and evaporation deposition used previously. This may suggest that not only the particle size but also the method selected for nanoAg deposition is crucial for antibacterial activity of modified samples and may allow for precise control of bactericidal properties of titanium-based surfaces. Samples modified with different amounts of ZnO NPs showed nonlinear antibacterial efficacy. Surprisingly, we observed that samples with medium ZnO content (deposition for 3 min) exhibited higher bacterial-killing capacity and higher anti-adhesive effect than samples with the highest (deposition for 5 min) ZnO content. This phenomenon is accompanied by a highest rate of Zn2+ release from 3 min deposited surfaces in all systems (ZnO + 100 nm NT, ZnO + 150 nm NT and Ag + ZnO + 150 nm NT). This correlation suggests that — most probably — bacterial-killing and anti-adhesive capacity of nanoparticle-modified systems depend of solubility of metal NPs. Reduced solubility of metal NPs deposited on 5 min-treated surfaces may be associated with their altered topography. Microscopic observations revealed that, for 5 min-deposited samples, the layer of ZnO NPs was thick, nonhomogeneous and formed island-like aggregates. They closed the interior of some TiO2 nanotubes and probably reduced the specific surface area of modified titanium samples. Thus, ZnO NPs created a pseudo-bulk metallic multilayer of reduced (as confirmed by Zn release profile) solubility in comparison with samples with ZnO NPs deposited only for 3 min. Recent literature data showed that metal nanoparticles (as evidenced for nanoAg) exhibit higher antibacterial efficacy than its bulk counterpart [12]. It was also proved that antibacterial capacity of ZnO nanoparticles increases with the reduction of particles size [39]. Therefore, based on our results and available literature we conclude that the amount of antibacterial metal NPs loaded on surfaces designed for biomedical implants should be carefully selected. It should be neither too small (because bacteria are capable to metabolize Zn2+ as an oligoelement, as shown by Brayner et al. [40]) nor too large (to avoid the transformation to bulk-like state). Most probably, the quantity of loaded antibacterial agents may also have an impact on the development of microorganism resistance to zinc. As summarized by Goudouri et al. [41], zinc resistance of bacteria can be developed due to several mechanisms including sequestration, efflux or extracellular accumulation. Thus, loading of biomaterial surfaces with too large quantity of ZnO NPs could result in extracellular Zn2 + accumulation and induce zinc-resistance of bacteria. We want to emphasize that, in our opinion, complete and reliable evaluation of antibacterial efficacy of ZnO NP-loaded titanium surfaces cannot be fulfilled based on the performed experiments. We used stationary conditions, as described in the standard for the evaluation of antimicrobial activity. However, in vivo conditions are different. Circulating tissue liquids will probably gradually dissolve and remove the

165

solubilized antimicrobial agents deposited on the titanium implant. Thus, to evaluate the potential of antimicrobial implant accurately, the complex study should be performed in future in a flow-through model, mimicking in vivo system. The results will allow predicting more reliably the most beneficial dose of nanometals deposited on titanium implants. 5. Conclusions In design of titanium implants for the prevention of post-operative bacterial infections, a lot of factors were found essential. TiO2 nanotubes showed significant bactericidal power, related to nanotube diameter, but promoted the adhesion of bacterial cells. Ag and ZnO nanoparticles, deposited on TiO2 NT-coated Ti, killed bacteria very efficiently and reduced the adhesion of bacterial cells. This confirmed the high potential of this local delivery system in the prevention of bone implant-related infections. However, the quantity of loaded nanoparticles may be a limiting factor (overdose effect) in overall antibacterial activity of manufactured implants. Complex study of their bactericidal efficacy should probably include both stationary and flow-through models to select the optimal dose of loaded metal nanoparticles. Co-deposition of two metal nanoparticles may be considered a tool for enhanced control of post-operative infections related to Ti-based bone implants. Acknowledgments This work was financially supported by the National Science Center (DEC 2011/03/N/ST5/04388) and developed using the equipment purchased within the agreement No. POPW.01.03.00-06-010/09-00 Operational Program Development of Eastern Poland 2007–2013, Priority Axis I, Modern Economy, Operations 1.3. Innovations Promotion. References [1] J.S. Gogia, J.P. Meehan, P.E. Di Cesare, A.A. Jamali, Local antibiotic therapy in osteomyelitis, Semin. Plast. Surg. 23 (2009) 100. [2] M. Yoshinari, Y. Oda, T. Kato, K. Okuda, Influence of surface modifications to titanium on antibacterial activity in vitro, Biomaterials 22 (2001) 2043. [3] G. Cheng, Z. Zhang, S. Chen, J.D. Bryers, S. Jiang, Inhibition of bacterial adhesion and biofilm formation on zwitterionic surfaces, Biomaterials 28 (2007) 4192. [4] X. Liu, P.K. Chu, Ch. Ding, Surface modification of titanium, titanium alloys, and related materials for biomedical applications, Mater. Sci. Eng. R 47 (2004) 49. [5] K.S. Brammer, S. Oh, Ch.J. Frandsen, S. Jin, TiO2 nanotube structures for enhanced cell and biological functionality, JOM 62 (2010) 50. [6] A. Roguska, A. Belcarz, T. Piersiak, M. Pisarek, G. Ginalska, M. Lewandowska, Evaluation of the antibacterial activity of Ag-loaded TiO2 nanotubes, Eur. J. Inorg. Chem. 32 (2013) 5199. [7] K.C. Popat, M. Eltgroth, T.J. La Tempa, C.A. Grimes, T.A. Desai, Decreased Staphylococcus epidermidis adhesion and increased osteoblasts functionality on antibiotic-loaded titania nanotubes, Biomaterials 28 (2007) 4880. [8] R.P. Allaker, K. Memarzadeh, Nanoparticles and the control of oral infections, Int. J. Antimicrob. Agents 43 (2014) 95. [9] X.-Z. Li, H. Nikaido, K.E. Williams, Silver-resistant mutants of Escherichia coli display active efflux of Ag+ and are deficient in porins, J. Bacteriol. 179 (1997) 6127–6132. [10] A. Gupta, M. Maynes, S. Silver, Effect of halides and plasmid-mediated silver resistance in Escherichia coli, Appl. Environ. Microbiol. 64 (1998) 5042–5045. [11] S. Silver, T. Phung Le, G. Silver, Silver as biocides in burn and wound dressings and bacterial resistance to silver compounds, J. Ind. Microbiol. Biotechnol. 33 (2006) 627–634. [12] M. Rai, A. Yadav, A. Gade, Silver nanoparticles as a new generation of antimicrobials, Biotechnol. Adv. 27 (2009) 76. [13] S.J. Beard, M.N. Hughes, R.K. Poole, Inhibition of the cytochrome bd-terminated NADH oxidase system in Escherichia coli K-12 by divalent metal ions, FEMS Microbiol. Lett. 131 (1995) 205. [14] M. Premanathan, K. Karthikeyan, K. Jeyasubramanian, G. Manivannan, Selective toxicity of ZnO nanoparticles toward Gram-positive bacteria and cancer cells by apoptosis through lipid peroxidation, Nanomed. Nanotechnol. 7 (2011) 184. [15] G. Jin, H. Cao, Y. Qiao, F. Meng, H. Zhu, X. Liu, Osteogenic activity and antibacterial effect of zinc ion implanted titanium, Colloids Surf. B 117 (2014) 158. [16] R. Choudhury, S. Srivastava, Zinc resistance mechanisms in bacteria, Curr. Sci. 81 (2001) 768. [17] E.A. Pfau, M.J. Avila-Campos, Prevotella intermedia and Porphyromonas gingivalis isolated from osseointegrated dental implants: colonization and antimicrobial susceptibility, Brazil, J. Microbiol. 36 (2005) 281. [18] T. Duxbury, Toxicity of heavy metals to soil bacteria, FEMS Microbiol. Lett. 11 (1981) 217.

166

A. Roguska et al. / Materials Science and Engineering C 51 (2015) 158–166

[19] C. Díaz, A. Miñán, P.L. Schilardi, M. Fernández Lorenzo de Mele, Synergistic antimicrobial effect against early biofilm formation: micropatterned surface plus antibiotic treatment, Int. J. Antimicrob. Agents 40 (2012) 221. [20] A. Stoyanova, H. Hitkova, A. Bachvarova-Nedelcheva, J Synthesis and antibacterial activity of TiO2/ZnO nanocomposites prepared via nonhydrolytic route, Chem. Technol. Metall. 48 (2013) 154. [21] S.H. Hwang, J. Song, Y. Jung, O.Y. Kweon, H. Song, J. Jang, Electrospun ZnO/TiO2 composite nanofibers as a bactericidal agent, Chem. Commun. 47 (2011) 9164. [22] A.H. Shah, E. Manikandan, M. Basheer Ahmed, V. Ganesan, Enhanced bioactivity of Ag/ZnO nanorods — a comparative antibacterial study, J. Nanomed. Nanotechnol. 4 (2013) 2. [23] N. Sharma, J. Kumar, S. Thakur, S. Sharma, Antibacterial study of silver doped zinc oxide nanoparticles against Staphylococcus aureus and Bacillus subtilis, Drug Invent. Today 5 (2013) 50. [24] O. Akhavan, E. Ghaderi, Bactericidal effects of Ag nanoparticles immobilized on surface of SiO2 thin film with high concentration, Curr. Appl. Phys. 9 (2009) 1381. [25] J. Ciampolini, K.G. Harding, Pathophysiology of chronic bacterial osteomyelitis. Why do antibiotics fail so often? Postgrad. Med. J. 76 (2000) 479. [26] JIS Z 2801, Antimicrobial products — test for antimicrobial activity and efficacy, Japanese Industrial Standard, Japanese Standard Association, 2000. [27] A. Roguska, M. Pisarek, M. Andrzejczuk, M. Lewandowska, Synthesis and characterization of ZnO and Ag nanoparticle-loaded TiO2 nanotube composite layers intended for antibacterial coatings, Thin Solid Films 553 (2014) 173–178. [28] S.L. Hall, H.P. Dimai, J.R. Farley, Effects of zinc on human skeletal alkaline phosphatase activity in vitro, Calcif. Tissue Int. 64 (1999) 163. [29] J.-R. Gurr, A.S.S. Wang, Ch-H. Chen, K.-J. Jan, Ultrafine titanium dioxide particles in the absence of photoactivation can induce oxidative damage to human bronchial epithelial cells, Toxicology 213 (2005) 66. [30] L. Braydich-Stolle, N.M. Schaeublin, R.C. Murdock, J. Jiang, P. Biswas, J.J. Schlager, S.M. Hussain, Crystal structure mediates mode of cell death in TiO2 nanotoxicity, J. Nanopart. Res. 11 (2009) 1361.

[31] K.S. Brammer, S. Oh, Ch.J. Cobb, L.M. Bjursten, H. van der Heyde, S. Jin, Improved bone-forming functionality on diameter-controlled TiO2 nanotube surface, Acta Biomater. 5 (2009) 3215. [32] L.L. Radke, B.L. Hahn, D.K. Wagner, P.G. Sohnle, Effect of abscess fluid supernatants on the kinetics of Candida albicans growth, Clin. Immunol. Immunopathol. 73 (1994) 344. [33] T.A. Södenberg, B. Sunzel, S. Holm, T. Elmros, G. Hallmans, S. Sjöberg, Antibacterial effect of zinc oxide in vitro, Scand. J. Plast. Reconstr. Surg. Hand Surg. 24 (1990) 193. [34] P. Dibrov, J. Dzioba, K.K. Gosink, C.C. Häse, Chemiosmotic mechanism of antimicrobial activity of Ag+ in Vibrio cholerae, Antimicrob. Agents Chemother. 46 (2002) 2668. [35] C.-N. Lok, C.-M. Ho, R. Chen, Q.-Y. He, W.-Y. Yu, H. Sun, et al., Proteomic analysis of the mode of antibacterial action of silver nanoparticles, J. Proteome Res. 5 (2006) 916. [36] C. Marambio-Jones, E.M.V. Hoek, A review of the antibacterial effects of silver nanomaterials and potential applications for human health and the environment, J. Nanopart. Res. 12 (2010) 1531. [37] M.A. Vargas-Reus, K. Memarzadeh, J. Huang, G.G. Ren, R.P. Allaker, Antimicrobial activity of nanoparticulate metal oxides against peri-implantitis pathogens, Int. J. Antimicrob. Agents 40 (2012) 135. [38] H. Cao, Y. Qiao, X. Liu, T. Lu, T. Cui, F. Meng, P.K. Chu, Electron storage mediated dark antibacterial action of bound silver nanoparticles: smaller is not always better, Acta Biomater. 9 (2013) 5100. [39] S. Nair, A. Sasidharan, R.V. Divya, D. Menon, K. Manzoor, S. Raina, Role of size scale of ZnO nanoparticles and microparticles on toxicity towards bacteria and osteoblast cancer cells, J. Mater. Sci. Mater. Med. 20 (2009) S253. [40] R. Brayner, R. Ferrari-Illiou, N. Brivois, S. Djediat, M. Benedetti, F. Fiévet, Toxicological impact studies based on Escherichia coli bacteria in ultrafine ZnO nanoparticles colloidal medium, Nano Lett. 6 (2006) 866. [41] O.-M. Goudouri, E. Kontonasaki, U. Lohbauer, A.R. Boccaccini, Antibacterial properties of metal and metalloid ions in chronic periodontitis and peri-implantitis therapy, Acta Biomater. 10 (2014) 3795.