- Email: [email protected]
calcium silicate hydrate hierarchical coating on titanium
Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
A novel titania/calcium silicate hydrate hierarchical coating on titanium Qianli Huang a , Xujie Liu a , Tarek A. Elkhooly a,c , Ranran Zhang a , Zhijian Shen a , Qingling Feng a,b,∗ a
State key laboratory of new ceramics and fine processing, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China Key Laboratory of Advanced Materials of Ministry of Education of China, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China c Biomaterials Department, Inorganic Chemical Industries Division, National Research Centre, Dokki, 12622 Cairo, Egypt b
a r t i c l e
i n f o
Article history: Received 29 March 2015 Received in revised form 28 June 2015 Accepted 1 July 2015 Available online 3 July 2015 Keywords: Micro-arc oxidation Hydrothermal treatment Titanium implant Hierarchical surface topography
a b s t r a c t Recently, surface micron/nano-topographical modifications have attracted a great deal of attention because it is capable of mimicking the hierarchical characteristics of bone. In the current work, a novel titania/calcium silicate hydrate (CSH) bi-layer coating with hierarchical surface topography was successfully prepared on titanium substrate through micro-arc oxidation (MAO) and subsequent hydrothermal treatment (HT). MAO treatment could lead to a micron-scale topographical surface with numerous crater-like protuberances. The subsequent HT process enables the in situ nucleation and growth of CSH nanoplates on MAO-fabricated titania surface. The nucleation of CSH nanoplates is considered to follow a dissolution–precipitation mechanism. Compared to MAO-fabricated coating with single-scale surface topography, MAO–HT-fabricated coating with hierarchical surface topography exhibits enhanced hydrophilicity, fibronectin adsorption and initial MG-63 cell attachment. The process of cell-material interactions is considered to be triggered by surface properties of the coated layer and indirectly mediated by protein adsorption on coating surface. These results suggest that MAO–HT treatment is an efficient way to prepare coatings with hierarchical surface topography on titanium surface, which is essential for altering protein adsorption and initial cell attachment. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The establishment of integration between the implant and the surrounding bone is a key factor for successful implants fixation in orthopedic and dental applications [1]. The generally accepted process of osseointegration involves a complex chain of biological events between the implant surface and the physiological environment. These events are likely initiated by protein adsorption and blood clotting at the implant surface, followed by mesenchymal stem cells and osteoblasts adhesion to the surface and finally result in bone matrix formation by those differentiated cells at the interface [2,3]. The surface characteristics of titanium-based implants can significantly influence these interactions that emerge at the interface [4]. Unfavorable surface properties lead to the formation of fibrous layer around the implant and finally result in failure
∗ Corresponding author at: School of Materials Science and Engineering, Tsinghua University, Beijin 100084, China. Fax: +86 10 62771160. E-mail address: [email protected] (Q. Feng). http://dx.doi.org/10.1016/j.colsurfb.2015.07.002 0927-7765/© 2015 Elsevier B.V. All rights reserved.
of implantation [5]. Among all the surface properties of titaniumbased implants, surface topography and chemistry are considered to be the two critical factors that influence bone response [6]. In order to enhance these surface features of titanium-based implants for successful osseointegration, numerous modification strategies involving chemical, physical and biological tools have been established in the last decades [7–10]. The surface modification of titanium implants with micron/nano-topography has been proven useful to mimic the hierarchical characteristics of the bone [11]. Bone undergoes remodeling consisting of sequential osteoclastic resorption and osteoblastic formation to repair the microdamages and replace old bone with new one [12,13]. The resorpted lacunae created by osteoclastic activity have micron-scale features (pits with a diameter of 30–100 m) and nanostructure (collagen fibers left on the surface), exhibiting unique hierarchical structure [14,15]. This kind of hierarchical structure could be the signal for osteoblast attachment, proliferation and differentiation [2]. Many dental and orthopedic implants have been improved for better cellular response by modifying their surface topography at micron- and nanoscale [11,16]. Kubo et al.
170
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Fig. 1. SEM micrographs of MAOed (a and b) and MAO–HTed coatings (d and e) at different magnifications; 3D toughness profile of MAOed (c) and MAO–HTed (f) coatings.
reported that micron- and nano-hybrid topographies enhanced both osteoblast proliferation and differentiation compared to only micron-scale topography [17]. An investigation conducted by Gittens et al. reported that the surface with both micron-scale roughness and nano-scale features promoted osteoblast differentiation as well as local factor production more significantly than surfaces with only micron-scale roughness or nano-scale features [18]. Zhou et al. developed a hierarchical structure by patterning Srdoped hydroxyapatite (HA) nanorods on micro-arc oxidized TiO2 layer and found that the inter-rod spacing played an important
role in mediating cell adhesion [19]. These findings suggest that the modification of surface topography of titanium-based implants in both micron/nano-scale might have a great potential for successful osseointegration with host bone. The modification process of titanium surface with micron/nanotopography is usually achieved by two steps. It starts with the development of micron-scale topography on titanium-based substrate, followed by the superposition of nanostructure to the former modified surface. To create a surface with micron-scale topography on titanium, several techniques are commonly used, such as acid
Fig. 2. SEM–BSI of the cross-sectional morphology of MAO–HTed coating (a); TF-XRD patterns of MAO and MAO–HTed coatings (b); HRTEM image (c) and SAED pattern of scratched crystal plates of MAO–HTed coating.
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
171
bial materials [29]. Dou et al. investigated the cytocompatibility of poly(l-lactic acid)/xonotlite (Ca6 (Si6 O17 )(OH)2 ) and found that the incorporation of xonotlite enhanced the attachment and proliferation of bone marrow stromal cells [31]. Moreover, silicate-based materials are well known to exhibit good biocompatibility and have been proved to promote osteoblast proliferation and differentiation [32,33]. The favorable biological performance of CSH and the possibility to synthesize CSH through HT method have inspired us to develop a novel TiO2 /CSH bi-layer coating through MAO–HT treatment on titanium. The objectives of the current work are twofold. First, we aim to develop a novel TiO2 /CSH bi-layer coating on titanium through MAO–HT treatment in order to take advantages of the micron/nano-topographical structure and the benefits of CSH. Second, to investigate the physicochemical properties, the protein adsorption and the initial cell attachment of such modified hierarchical coating. 2. Experimental procedure 2.1. Coating preparation Grade 2 commercially pure titanium plates were used as substrate material (10 × 10 × 1 mm). The surfaces of substrates were abraded with abrasive paper up to grid 2000 and then successively cleaned with acetone, ethyl alcohol and distilled water. Such treated titanium substrate was referred as s-Ti. A MAO system was composed of the specimen as anode, a stainless steel container as cathode, electrolyte and a pulse power supply system (WHD-20, Haerbin, China). The electrolyte was prepared as follows: firstly, 0.10 mol/L Na2 (EDTA), 0.10 mol/L Ca(CH3 COO)2 ·H2 O and 0.25 mol/L NaOH were in turn dissolved in distilled water with continuous stirring. After that, the solution was stirred for 24 h in order to completely transform Ca2+ into Ca-chelate (Ca(EDTA)2− ). Finally, 0.02 mol/L Na2 SiO3 ·9H2 O was added into the solution. During the MAO process, the applied voltage, pulse frequency, duty cycle and the duration time were set at 250 V, 50 Hz, 50% and 7 min, respectively. After MAO treatment, the samples (referred as MAOed coating) were washed with distilled water and then dried in air. For HT treatment of MAOed coating, Teflon-lined autoclaves with a volume of 20 mL were employed and 4 mL ammonia aqueous solution (pH 11) for each sample was filled into Teflon-lined vessels. The autoclaves were heated at 200 ◦ C for 24 h. After HT treatment, MAO–HT-fabricated samples (referred as MAO–HTed coating) were washed with distilled water and then dried in air.
Fig. 3. C 1s (a), Ca 2p (b) and Si 2p (c) XPS spectra of MAOed and MAO–HTed coatings.
etching, sand blasting and micro-arc oxidation (MAO) [20,21]. To develop surface nanostructure, some other techniques are always employed, such as plasma spray, hydrothermal treatment (HT) and sol–gel method [11,22,23]. Among these techniques, the combination use of MAO and HT treatment (MAO–HT treatment) has received a lot of attention because it is able to produce a controllable surface hierarchical structure regardless the geometrical complexity of the titanium-based substrate [24,25]. Calcium silicate hydrate (CSH) along with series of crystalline minerals and poor ordered phases, is the major hydrate product in Portland cement [26,27]. Recently, several phases of CSH have been investigated on their potential for biomedical applications [28–31]. Tobermorite (Ca5 (Si6 O16 )(OH)2 ·4H2 O) synthesized by HT method was reported to be bioactive and biodegradable by Lin et al. [28]. Another study on tobermorite reported that Ag+ - and Zn+ exchanged tobermorite has a potential to be applied as antimicro-
2.2. Physicochemical properties analysis The specimen surface topography was evaluated by field emission scanning electron microscopy (FESEM, JSM-7001F; JEOL, Japan). 3D roughness profiling of MAOed and MAO–HTed coatings were examined using 3D profiling system (MicroXAM-3D Phase Shift; ADE Co., USA). The microstructure of the CSH crystallites, which were carefully scratched from the surface of MAO–HTed coating, was investigated using a transmission electron microscopy (TEM, JEM-2100; JEOL, Japan) operated at 200 kV. The phase composition of the coatings was characterized by thin-film X-ray diffraction with an incidence angle of 1◦ using an X-ray diffractometer (TF-XRD, Rigaku, Tokyo, Japan). The chemical state of the relevant elements was measured with an X-ray photoelectron spectrometer (XPS, Escalab 250Xi; Thermo Scientific, UK). The XPS test was conducted using an AlK␣ X-ray source and the obtained XPS spectra were calibrated against the binding energy of C 1s in the CH2 group at 284.8 eV. Static water contact angle of the coatings was measured by sessile drop using an optical tensiometer (JC2000C1;
172
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Fig. 4. Schematics showing the mechanism of the nanoplate nucleation on titania surface during HT process.
Powereach, China). At least three samples were used for each group to obtain the mean data. 2.3. Protein adsorption Fluorescein isothiocyanate labelled bovine serum albumin (FITC-BSA, ZKCY-BIO, China) and rhodamin labelled fibronectin (Rhodamin-FN) from bovine plasma (Cytoskeleton, USA) were used as protein models to investigate the protein adsorption. For single protein adsorption, a solution of 20 g/mL Rhodamin-FN diluted in PBS was used. For competitive protein adsorption, a mixture solution containing 2 mg/mL FITC-BSA and 20 g/mL Rhodamin-FN was used. The 24-well plate was kept at 37 ◦ C for 4 h. Subsequently, the samples were washed with PBS for three times and photographed under a fluorescence microscope (Leica, Germany). The amount of adsorbed protein on coating surface was semiquantitatively determined by using imageJ2x analysis software. All the images were converted to 8-bit black/white files and then inverted to a binary image. The OD value was defined with gray level 255 (white) as a value of 0 and gray level 0 (black) as a value of 2.708. In this way, the mean OD value of the edited images is proportional to the fluorescence intensity of corresponding original images.
1.0 mL cell suspension with cell density of 1 × 104 cell/mL was added into each well. LG-DMEM medium supplemented with FBS (10%) and without FBS were used in the cellular suspensions. The seeded cells were cultured at 37 ◦ C in an atmosphere of 5% CO2 and 100% humidity. After 24 h, the samples were gently washed twice with PBS to remove non-adherent cells and transferred to new 24-well plates. To investigate the interactions of cells with material surfaces, the cells were fixed with 2.5% glutaraldehyde in PBS overnight and dehydrated in a series of graded alcohols. After exchanging the alcohol within cells with a series of graded tertiary butanol (25%, 50%, 75% and 100% in ethanol), the samples were freeze dried, sputtercoated with gold and observed using SEM. Cell spreading and attachment were evaluated by immunostaining for vinculin and staining for actin and nucleus. The samples were fixed in 4% paraformaldehyde, permeablized with 0.3% triton X-100 in PBS, and then blocked with 10% goat serum in PBS. The primary antibodies used were rabbit anti-vinculin antibody diluted
2.4. Cell attachment Human osteoblastic MG63 cells (supplied by China Infrastructure of Cell Line Sources, China) were used in the present study because these cells are considered as an appropriate cell model to study initial cell attachment [34]. The cells were cultured in low glucose Dulbecco’s modified Eagle’s medium (LG-DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 ◦ C in an atmosphere of 5% CO2 and 100% humidity. The media were refreshed every 48 h. To investigate the cell attachment, s-Ti, MAOed and MAO–HTed specimens were put into the well center of 24-well plates. Then
Fig. 5. Static contact angle of MAOed and MAO–HTed coatings.
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
173
Fig. 6. Fluorescent images and the calculated fluorescence intensities of the adsorbed FN from a single solution of FN (a–c) and a mixture solution of BSA and FN (e and f) on MAOed and MAO–HTed coatings.
Fig. 7. Fluorescence images of DAPI-stained nuclei of MG-63 cells grown for 24 h on s-Ti (a), MAOed (b) and MAO–HTed (c) coatings and average cell density adhering 24 h after seeding (*p < 0.05, **p < 0.01) (d) 89.
1:100 (Sigma, USA) in tris-PBS (TPBS). The secondary antibodies used were TRITC labelled goat anti-rabbit antibody (Bioss, China) at 1:300. Actin was stained with 5 g/mL phalloidin-TRITC (Sigma, USA) and nucleus was stained with 4 -6-diamidino-2-phenylindole (DAPI). Thereafter, the samples were photographed by fluorescence (Leica, Germany) and confocal microscopy (Zeiss 710, Germany). For cell number quantification in Fig. 8, ImageJ2x analysis software was used. The raw image was converted to an 8-bit file and then converted to a binary image by setting a threshold. Threshold values were determined empirically, giving the most accurate image for a subset of randomly selected photomicrographs. Cell numbers were calculated in randomly selected 1000 m squares by “analyze particles” in ImageJ2x. Five samples per condition were
analyzed and this procedure was repeated five times (n = 5), and the statistical significance of differences in means were determined by one-way analysis of variance (ANOVA) followed by post hoc comparisons with least significant difference (LSD) method. A value of p < 0.05 was considered as statistically significant. 3. Results and discussion 3.1. Physicochemical properties analysis Fig. 1 shows the surface morphology and roughness profile of MAOed and MAO–HTed coatings. MAOed coating has numerous micron-sized and crater-shaped protuberances on its surface,
174
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Fig. 8. Actin (red), cell nucleus (blue) and Vinculin fluorescence and SEM images of the cells after 24 h of culture on s-Ti, MAOed and MAO–HTed coatings. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
exhibiting a three-dimensional morphology with micron-scale topography (Fig. 1a). The porous surface morphology of MAOed coating, with a pore size ranging from 0.4 to 2 m, is considered to be beneficial for bone in growth and the consequent mechanical interlocking between the implant and bone [6]. The wall of the pores of MAOed coating appears to be smooth at low magnification (Fig. 1a). But at higher magnification, the external surface of the protuberance is noticed to be covered by numerous nano-sized crystals (Fig. 1b). These crystals are titania formed during MAO treatment [35]. MAO–HTed coating is a bi-layer coating which consists of titania layer formed firstly by MAO treatment and second layer formed during the subsequent HT process (Fig. 1d and e). The second layer of the coating is composed of large scale of nanoplates and exhibit nano-scale features (Fig. 1e). These nanoplates, with an average thickness of 10 nm, grow in situ on the surface of the titania layer and distribute homogeneously following the surface curve of the crater-shaped protuberances. The network formed by nanoplates is similar to the nano-scale topography associated with collagen fibrils left by the osteoclasts after bone resorption [36]. As the second layer is very thin, the first layer formed by MAO treatment can preserve its micron-scale surface topography after the HT process (Fig. 1d). Both MAOed and MAO–HTed coatings exhibit submicron-scale surface roughness (Fig. 1c and f). The arithmetic mean surface roughness Ra of the coating decreases slightly from 535 ± 3 nm to 471 ± 16 nm through the superposition of nano-sized structure to the micron-scale topographical surface. These results suggest that
MAO–HT treatment is an efficient way to fabricate a hierarchical surface structure on titanium surface. The cross-sectional morphology of MAO–HTed coating is shown in Fig. 2a. Only the first titania layer is observed because the second layer is very thin and easy to be destroyed during the cross-sectional sample preparation. It is easy to distinguish the coating from the substrate from the SEM back scattered image (BSI) because of their different chemical compositions. The thickness of the coating is uniform with an average value of 3.5 m. There is no discontinuity at the interface of the substrate and the coating, indicating that the coating firmly binds to the substrate. Fig. 2b shows the TF-XRD patterns of MAOed and MAO–HTed coatings. The Ti peaks (JCPDS # 44-1294) are attributed to the substrate. For MAOed coating, both the anatase peak (JCPDS # 21-1272) and the rutile peak (JCPDS # 21-1276) are detected and the former one has stronger intensity than the latter one. It indicates that MAOed coating consists of rutile and predominant anatase. Different from that of MAOed coating, the TF-XRD pattern of MAO–HTed coating shows an additional peak at 29.2◦ , which is the typical peak of CSH [37]. It indicates that the second layer formed by HT is one kind of CSH. The XRD results are consistent with the observation of nano-sized particles after MAO and silicate nanoplates after HT treatment in SEM micrographs in Fig. 1. To determine the exact crystalline phase of the nanoplates formed after HT process, TEM is employed. The high-resolution TEM (HRTEM) image (Fig. 2c) reveals that the CSH nanoplate is an incompletely crystallized layer. The nanoplate formed after HT treatment contains predominant crystalline phase and small amor-
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
phous phase. The selected area electron diffraction (SAED) pattern shows distinct spots corresponding to 6CaO·3SiO2 ·H2 O (Fig. 2d). The current work reveals that CSH coating can be prepared in situ on titanium surface through the incorporation of Ca and Si elements into MAO coating and subsequent HT process. Thus, it introduces an interesting modification to existing MAO technique [38–40]. It is worth mentioning that peaks corresponding to crystalline phases composed of Ca and Si are not observed in the TF-XRD patterns of MAOed coating (Fig. 2b). It may be because these elements exist in the formed compounds as amorphous phases in MAOed coating. Therefore, the chemical states of Ca and Si on the surface of MAOed and MAO–HTed coatings are analyzed by XPS in detail. The high-resolution XPS spectra of TiO2 and B–TiO2 films and their deconvoluted peaks fitted using the Gaussian fitting method are shown in Fig. 3. The C1s spectrum of MAOed coating can be deconvoluted into four components: C C/C H (284.5 eV), C N (286.4 eV), O C O (288.6 eV) and p–p* stacking (290.9 eV) [41,42], indicating that Ca-EDTA complex exists in MAOed coating (Fig. 3a). The presence of Ca–EDTA complex is due to the movement of negatively charged ions (Ca(EDTA)2− ) to the anode during MAO treatment [43]. The C1s spectrum of MAO–HTed coating is composed of three main peaks: C C/C H (284.5 eV), C O (286.1 eV) and O C O (288.6 eV) (Fig. 3a). All these complexes are classically observed and attributed to surface contamination [44]. As shown in Fig. 3b, the deconvoluted peaks of the Ca 2p spectrum of MAOed coating indicate the presence of three Ca-containing compounds: CaO (Ca 2p3/2 at 346.6 eV and Ca 2p1/2 at 349.9 eV) [45], CaSiO3 (Ca 2p3/2 at 347.3 eV and Ca 2p1/2 at 351.5 eV) [46] and Ca-EDTA complex (Ca 2p3/2 at 349.7 eV and Ca 2p1/2 at 353.7 eV Fig. 3b). However, for MAO–HTed coating, the Ca 2p peaks are located at 346.6 eV (2p3/2 ) and 350.2 eV (2p1/2 ), corresponding to Ca in 6CaO·3SiO2 ·H2 O (Fig. 3b). The Si 2p XPS spectrum of MAOed coatings could be divided into two components with binding energy of 102.1 eV and 104.3 eV, which are, respectively, ascribed to CaSiO3 and SiO2 (Fig. 3c) [45]. While for MAO–HTed coating, the Si 2p peak located at 102.8 eV is ascribed to Si in 6CaO·3SiO2 ·H2 O (Fig. 3c). The XPS results indicate that the Ca- and Si-containing compounds in MAOed coating could transform to 6CaO·3SiO2 ·H2 O nanoplates in MAO-HTed coating after HT process. The nucleation mechanism of CSH nanoplates is discussed in Section 3.2.
3.2. The nucleation and growth of CSH nanoplates It is observed that CSH nanoplates can nucleate in situ on the titania surface during HT process. The process of nucleation can be described as a dissolution–precipitation mechanism (schematically shown in Fig. 4). Under HT condition, the Ca- and Si-containing compounds, such as CaO, CaSiO3 , Ca–EDTA complex and SiO2 , can be dissociated into the aqueous environment and thus increase the ion concentrations of Ca2+ and SiO3 2− (Fig. 5a). Meanwhile, the titania surface is attacked by OH− ions in the alkaline environment and transforms to Ti O− surface according to the following reaction: Ti O Ti + OH− → 2 Ti O− + H2 O Therefore, the whole interface between the coating surface and the solution environment becomes negatively charged. Positively charged Ca2+ ions can be attracted to the interface (Fig. 4b). Consequently, a positively charged surface is created through the accumulation of Ca2+ ions. The positively charged interface subsequent attracts negatively charged ions, such as OH− and SiO3 2− to the interface (Fig. 4c). Finally, the enrichment of Ca2+ , SiO3 2− and OH− ions at the interface can induce the heterogeneous nucleation of CSH on the surface of the titania (Fig. 4d). In this way, the CSH nanoplates can form in situ on titania surface during HT process.
175
3.3. Contact angle measurement To evaluate the hydrophilicity of MAOed and MAO–HTed coatings and its impact on protein adsorption and cell adhesion, static water contact angle was measured. As shown in Fig. 5, the contact angles of MAOed and MAO-HTed coatings are 21.3 ± 2.6◦ and 5.7 ± 1.1◦ , respectively. The contact angle is reduced by 15.6◦ after HT process, indicating the enhanced hydrophilicity of MAO–HTed coating. The hydrophilicity of a solid surface is considered to be affected by two main factors: the surface energy and surface topography [47]. In the current case, both surface energy and surface topography might contribute to the enhanced hydrophilicity of MAO–HTed coating, but the later one is considered to be predominant. The surface properties, such as surface topography, surface chemistry and hydrophilicity, influence the protein adsorption process under physiological conditions and thus mediate the cellmaterial interactions [48,49]. In the current study, MAO–HTed coating exhibits hierarchical surface topography, different surface chemistry, enhanced hydrophilicity compared to MAOed coating. The effects of these surface properties on protein adsorption and initial cell attachment are discussed in the following sections. 3.4. Protein adsorption In this work, fluorescence microscopy was used to semiquantitatively quantify the adsorption amounts of fibronectin (FN) with and without the competitive counterpart [50]. Fig. 6 shows the amount of adsorbed fibronectin on both coatings in different solutions. For single protein adsorption, MAO–HTed coating adsorbs roughly 35% more FN than MAOed coating (Fig. 6c). It is considered to be caused by the increase surface area of MAO–HTed coating [48]. However, for competitive protein adsorption, MAO–HTed coating adsorbs roughly 70% more compared to MAOed coating (Fig. 6f). It is found that FN adsorption on MAO–HTed coating is further enhanced during competitive protein adsorption, indicating that MAO–HTed coating has a preference for FN adsorption compared to bovine serum albumin (BSA). This phenomenon may be caused by two mechanisms. One is the charge repulse effect of the excessive negative charge of SiO4 4− [51]. As BSA is negatively charged in PBS [49], the repulsive force between the negatively charged MAO–HTed coating and BSA lead to the inhibition of BSA adsorption. Even though FN is also negatively charged in PBS, it has been proven that FN can be adsorbed to negatively charged surfaces in a more extended or relaxed conformation compared with less charged surfaces [52]. The other mechanism is the hydrophobic interactions. It has been proven that the hydrophobicity is the key determinant for BSA adsorption [53], while FN is known to adsorb successfully in an active conformation on hydrophilic surface [54]. Thus, due to the charge repulse effect and the hydrophobic interactions, MAO–HTed coating might have an effect on inhibiting BSA adsorption without significantly influencing FN adsorption. 3.5. Initial cell attachment To evaluate the initial cell attachment, MG-63 osteoblast-like cells were cultured on s-Ti, MAOed and MAO–HTed coatings for 24 h. The density of attached MG-63 cells on different surfaces was visualized by DAPI-stained nuclei, as shown in Fig 7a–c. It was further quantified using a computer-assisted cell counting program (shown in Fig. 7d). The total number of cells per cm2 is increased by 70% and 130% on MAO and MAO–HT coatings compared to sTi, respectively. However, compared to MAOed coating, the cell density on MAO–HT coating is increased by 30%. Fig. 8 shows the spreading and attachment of MG-63 cells cultured on different coating surfaces after 24 h of culture in the
176
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
presence and absence of FBS. Compared to the relatively roundish cells cultured in the absence of FBS (Fig. 8j–l), cells exhibit a more spread morphology on coating surfaces in the presence of FBS (Fig. 8a–c). It indicates that the process of cell-material interactions is mediated by protein adsorption. Cells poorly spread on the s-Ti surface and the formation stress fiber is not visible (Fig. 8a). On MAOed coating, cells spread widely with cytoskeleton extensions (Fig. 8b). The micro-scale protuberance could serve as cell-anchoring site for cell attachment (Fig. 8h). For cells attached on MAO–HTed coating, the extension of long filopodia and the formation of lamellipodia on CSH nanoplates are noticed (Fig. 8i). It is reported that filopodia could probe the microenvironment and trigger the extension of lamellipodia through signal transduction [55]. Meanwhile, the formation of actin stress fibers is noticed on both MAOed and MAO–HTed coating (Fig. 8b and c). It has been accepted that the formation of focal adhesion points (FAPs) is an important aspect of osteointegration [56]. FAPs are clusters of integrins linking the cells to ECM [57]. In the current work, focal adhesions are observed to be more in individual cells on MAO–HTed coating, less on the MAOed coating and not obvious on s-Ti (Fig. 8d–f, shown by white arrows). The results indicate that hierarchically topographical surface may contain more signals that mediate MG-63 cell attachment than the surface with single-scale topography. The results show that the initial cell attachment on MAO–HTed coating is enhanced compared to MAOed coating. It is widely accepted that surface properties, such as surface topography, chemistry and energy, affect the type, amount and conformation of adsorbed proteins on material surface and subsequently influence cellular response [19]. In the current work, MAO–HTed coating exhibits significantly different surface properties, such as surface topography and chemistry, compared to MAOed coating. These properties could mediate protein adsorption and subsequently trigger cell response. Due to the hierarchy surface topography, different surface chemistry and enhanced hydrophilicity, MAO–HTed coating exhibits a preferential trend on FN adsorption compared to MAOed coating. As FN adsorption on biomaterial interface has been reported to promote cell adhesion by forming focal adhesion with integrin receptors in fibroblasts [58,59], it is reasonable to assume that the enhanced cell attachment on MAO–HTed coating is indirectly mediated by protein adsorption. The results show that the superposition of CSH nanoplates to TiO2 coating promotes both FN adsorption and the subsequent cell attachment. It indicates that cell-material interactions, such as cell spread and attachment, can be promoted through appropriate surface modifications of materials on the chemical composition and surface topography. The current work suggests that MAO–HT treatment is an effective way to prepare coatings with both micron- and nanohierarchical surface topography on titanium surface, which is essential for altering protein adsorption and initial cell attachment.
4. Conclusions In this study, a novel titania/CSH bi-layer coating with hierarchical surface topography was successfully prepared on titanium substrate through MAO–HT treatment. MAO treatment leads to a micron-scale topographical surface with numerous crater-like protuberances. The subsequent HT process enables the in situ growth of CSH nanoplates on titania surface. Compared to MAOed coating with single-scale surface topography, MAO–HTed coating with hierarchical surface topography exhibits enhanced hydrophilicity, fibronectin adsorption and initial MG-63 cell attachment. These results suggest that MAO–HT treatment is an efficient way to prepare coatings with hierarchical surface topography on titanium surface and the establishment of hierarchical surface topography is essential for altering protein adsorption and initial cell attachment.
Acknowledgments The authors are grateful for the financial support from National Natural Science Foundation of China (51361130032, 51472139) and Doctor Subject Foundation of the Ministry of Education of China (20120002130002). References [1] D. Lakstein, W. Kopelovitch, Z. Barkay, M. Bahaa, D. Hendel, N. Eliaz, Acta Biomater. 5 (2009) 2258–2269. [2] R.A. Gittens, R. Olivares-Navarrete, Z. Schwartz, B.D. Boyan, Acta Biomater. 10 (8) (2014) 3363–3371. [3] A.A.P. Mansur, H.S. Mansur, Mater. Sci. Eng. C-Mater. Biol. Appl. 30 (2010) 288–294. [4] A. Lugovskoy, S. Lugovskoy, Mater. Sci. Eng. C-Mater. Biol. Appl. 43 (2014) 527–532. [5] N. Athanasou, J. Quinn, C. Bulstrode, J. Bone Joint Surg. Br. 74 (1992) 57–62. [6] Y.-T. Sul, C. Johansson, E. Byon, T. Albrektsson, Biomaterials 26 (2005) 6720–6730. [7] L.-H. Li, Y.-M. Kong, H.-W. Kim, Y.-W. Kim, H.-E. Kim, S.-J. Heo, J.-Y. Koak, Biomaterials 25 (2004) 2867–2875. [8] S. He, P. Zhou, L. Wang, X. Xiong, Y. Zhang, Y. Deng, S. Wei, J. R. Soc. Interface 11 (2014) 20140169. [9] G. Orsini, B. Assenza, A. Scarano, M. Piattelli, A. Piattelli, Int. J. Oral Maxillofac. Implants 15 (1999) 779–784. [10] J. Li, X. Liu, Y. Qiao, H. Zhu, J. Li, T. Cui, C. Ding, RSC Adv. 3 (2013) 11214–11225. [11] W. Zhang, G. Wang, Y. Liu, X. Zhao, D. Zou, C. Zhu, Y. Jin, Q. Huang, J. Sun, X. Liu, X. Jiang, H. Zreiqat, Biomaterials 34 (2013) 3184–3195. [12] M. Okamoto, J. Murai, Y. Imai, D. Ikegami, N. Kamiya, S. Kato, Y. Mishina, H. Yoshikawa, N. Tsumaki, J. Bone Miner. Res. 26 (2011) 2511–2522. [13] E.F. Eriksen, Rev. Endocr. Metab. Disord. 11 (2010) 219–227. [14] R.A. Gittens, T. McLachlan, R. Olivares-Navarrete, Y. Cai, S. Berner, R. Tannenbaum, Z. Schwartz, K.H. Sandhage, B.D. Boyan, Biomaterials 32 (2011) 3395–3403. [15] J.-Y. Rho, L. Kuhn-Spearing, P. Zioupos, Med. Eng. Phys. 20 (1998) 92–102. [16] R.A. Gittens, R. Olivares-Navarrete, A. Cheng, D.M. Anderson, T. McLachlan, I. Stephan, J. Geis-Gerstorfer, K.H. Sandhage, A.G. Fedorov, F. Rupp, B.D. Boyan, R. Tannenbaum, Z. Schwartz, Acta Biomater. 9 (2013) 6268–6277. [17] K. Kubo, N. Tsukimura, F. Iwasa, T. Ueno, L. Saruwatari, H. Aita, W.-A. Chiou, T. Ogawa, Biomaterials 30 (2009) 5319–5329. [18] R.A. Gittens, R. Olivares-Navarrete, T. McLachlan, Y. Cai, S.L. Hyzy, J.M. Schneider, Z. Schwartz, K.H. Sandhage, B.D. Boyan, Biomaterials 33 (2012) 8986–8994. [19] J.H. Zhou, Y. Han, S.M. Lu, Int. J. Nanomed. 9 (2014) 1243–1260. [20] M. Herrero-Climent, P. Lázaro, J.V. Rios, S. Lluch, M. Marqués, J. Guillem-Martí, F. Gil, J. Mater. Sci. – Mater. Med. 24 (2013) 2047–2055. [21] R. Zhou, D. Wei, S. Cheng, W. Feng, Q. Du, H. Yang, B. Li, Y. Wang, D. Jia, Y. Zhou, ACS Appl. Mater. Interfaces 6 (2014) 4797–4811. [22] J. Zhou, J. Shao, Y. Han, Appl. Surf. Sci. 303 (2014) 367–372. [23] Y. Zhang, K. Bai, Z. Fu, C. Zhang, H. Zhou, L. Wang, S. Zhu, S. Guan, D. Li, J. Hu, Appl. Surf. Sci. 258 (2012) 2939–2943. [24] Y. Han, J.H. Zhou, L. Zhang, K.W. Xu, Nanotechnology 22 (2011) 11. [25] J. Takebe, S. Ito, S. Miura, K. Miyata, K. Ishibashi, Mater. Sci. Eng. C – Mater. Biol. Appl. 32 (2012) 55–60. [26] I.G. Richardson, Cem. Concr. Res. 38 (2008) 137–158. [27] P. Yu, R.J. Kirkpatrick, B. Poe, P.F. McMillan, X. Cong, J. Am. Ceram. Soc. 82 (1999) 742–748. [28] K. Lin, J. Chang, R. Cheng, Acta Biomater. 3 (2007) 271–276. [29] N.J. Coleman, A.H. Bishop, S.E. Booth, J.W. Nicholson, J. Eur. Ceram. Soc. 29 (2009) 1109–1117. [30] A.P. Hurt, G. Getti, N.J. Coleman, Int. J. Biol. Macromol. 64 (2014) 11–16. [31] Y. Dou, C. Wu, J. Chang, Acta Biomater. 8 (2012) 4139–4150. [32] C.-H. Liu, C. Hung Jr., T.-H. Huang, C.-C. Lin, C.-T. Kao, M.-Y. Shie, Mater. Sci. Eng. C – Mater. Biol. Appl. 43 (2014) 359–366. [33] H. Hu, Y. Qiao, F. Meng, X. Liu, C. Ding, Colloid Surf. B – Biointerfaces 101 (2013) 83–90. [34] E. Czekanska, M. Stoddart, R. Richards, J. Hayes, Eur. Cells Mater. 24 (2012) 1l17. [35] J. Yan, J.F. Sun, P.K. Chu, Y. Han, Y.M. Zhang, J. Biomed. Mater. Res. Part A 101 (2013) 2465–2480. [36] J.E. Davies, Biomaterials 28 (2007) 5058–5067. [37] L.H. Long, L.D. Chen, J. Chang, Ceram. Int. 32 (2006) 457–460. [38] T.-C. Yang, H.-Y. Shu, H.-T. Chen, C.-J. Chung, J.-L. He, Surf. Coat. Technol. 259 (2014) 185–192. [39] Q.-M. Zhao, L. Cheng, H.-L. Yang, Z.-T. Liu, D.-H. Feng, Radiat. Eff. Defects Solids 169 (2014) 1056–1063. [40] S.-D. Wu, H. Zhang, X.-D. Dong, C.-Y. Ning, A.S. Fok, Y. Wang, Appl. Surf. Sci. 329 (2015) 347–355. [41] N.V. Qui, P. Scholz, T.F. Keller, K. Pollok, B. Ondruschka, Chem. Eng. Technol. 36 (2013) 300–306. [42] A. Mesnage, M.A. Magied, P. Simon, N. Herlin-Boime, P. Jégou, G. Deniau, S. Palacin, J. Mater. Sci. 46 (2011) 6332–6338.
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177 [43] W. Daqing, Z. Yu, J. Dechang, W. Yarning, Acta Biomater. 3 (2007) 817–827. [44] A. Izena, M. Sakuraba, T. Matsuura, J. Murota, J. Cryst. Growth 188 (1998) 131–136. [45] W. Zhang, K. Du, C. Yan, F. Wang, Appl. Surf. Sci. 254 (2008) 5216–5223. [46] Q. Meng, J. Lin, L. Fu, H. Zhang, S. Wang, Y. Zhou, J. Mater. Chem. 11 (2001) 3382–3386. [47] B. He, N.A. Patankar, J. Lee, Langmuir 19 (2003) 4999–5003. [48] V.V. Divya Rani, L. Vinoth-Kumar, V.C. Anitha, K. Manzoor, M. Deepthy, V.N. Shantikumar, Acta Biomater. 8 (2012) 1976–1989. [49] S. Patil, A. Sandberg, E. Heckert, W. Self, S. Seal, Biomaterials 28 (2007) 4600–4607. [50] M. Sun, J. Deng, Z. Tang, J. Wu, D. Li, H. Chen, C. Gao, Colloid Surf. B – Biointerfaces 122 (2014) 134–142.
[51] [52] [53] [54] [55] [56] [57] [58] [59]
177
X. Chen, T. Wu, Q. Wang, J.-W. Shen, Biomaterials 29 (2008) 2423–2432. B.E. Rapuano, J.J. Lee, D.E. MacDonald, Eur. J. Oral Sci. 120 (2012) 185–194. P.P. Roach, D. Farrar, C.C. Perry, J. Am. Chem. Soc 127 (2005) 8168–8173. J. Lee, B.H. Chu, K.-H. Chen, F. Ren, T.P. Lele, Biomaterials 30 (2009) 4488–4493. R. You, X. Li, Y. Liu, G. Liu, S. Lu, M. Li, J. Biomed. Mater. Res. Part A 102 (2014) 4206–4212. V.D. Rani, K. Manzoor, D. Menon, N. Selvamurugan, S.V. Nair, Nanotechnology 20 (2009) 195101. M.J.P. Biggs, R.G. Richards, M.J. Dalby, Nanomed. – Nanotechnol. Biol. Med. 6 (2010) 619–633. A.J. García, Biomaterials 26 (2005) 7525–7529. M.D. Mager, V. LaPointe, M.M. Stevens, Nat. Chem. 3 (2011) 582–589.
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
A novel titania/calcium silicate hydrate hierarchical coating on titanium Qianli Huang a , Xujie Liu a , Tarek A. Elkhooly a,c , Ranran Zhang a , Zhijian Shen a , Qingling Feng a,b,∗ a
State key laboratory of new ceramics and fine processing, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China Key Laboratory of Advanced Materials of Ministry of Education of China, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China c Biomaterials Department, Inorganic Chemical Industries Division, National Research Centre, Dokki, 12622 Cairo, Egypt b
a r t i c l e
i n f o
Article history: Received 29 March 2015 Received in revised form 28 June 2015 Accepted 1 July 2015 Available online 3 July 2015 Keywords: Micro-arc oxidation Hydrothermal treatment Titanium implant Hierarchical surface topography
a b s t r a c t Recently, surface micron/nano-topographical modifications have attracted a great deal of attention because it is capable of mimicking the hierarchical characteristics of bone. In the current work, a novel titania/calcium silicate hydrate (CSH) bi-layer coating with hierarchical surface topography was successfully prepared on titanium substrate through micro-arc oxidation (MAO) and subsequent hydrothermal treatment (HT). MAO treatment could lead to a micron-scale topographical surface with numerous crater-like protuberances. The subsequent HT process enables the in situ nucleation and growth of CSH nanoplates on MAO-fabricated titania surface. The nucleation of CSH nanoplates is considered to follow a dissolution–precipitation mechanism. Compared to MAO-fabricated coating with single-scale surface topography, MAO–HT-fabricated coating with hierarchical surface topography exhibits enhanced hydrophilicity, fibronectin adsorption and initial MG-63 cell attachment. The process of cell-material interactions is considered to be triggered by surface properties of the coated layer and indirectly mediated by protein adsorption on coating surface. These results suggest that MAO–HT treatment is an efficient way to prepare coatings with hierarchical surface topography on titanium surface, which is essential for altering protein adsorption and initial cell attachment. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The establishment of integration between the implant and the surrounding bone is a key factor for successful implants fixation in orthopedic and dental applications [1]. The generally accepted process of osseointegration involves a complex chain of biological events between the implant surface and the physiological environment. These events are likely initiated by protein adsorption and blood clotting at the implant surface, followed by mesenchymal stem cells and osteoblasts adhesion to the surface and finally result in bone matrix formation by those differentiated cells at the interface [2,3]. The surface characteristics of titanium-based implants can significantly influence these interactions that emerge at the interface [4]. Unfavorable surface properties lead to the formation of fibrous layer around the implant and finally result in failure
∗ Corresponding author at: School of Materials Science and Engineering, Tsinghua University, Beijin 100084, China. Fax: +86 10 62771160. E-mail address: [email protected] (Q. Feng). http://dx.doi.org/10.1016/j.colsurfb.2015.07.002 0927-7765/© 2015 Elsevier B.V. All rights reserved.
of implantation [5]. Among all the surface properties of titaniumbased implants, surface topography and chemistry are considered to be the two critical factors that influence bone response [6]. In order to enhance these surface features of titanium-based implants for successful osseointegration, numerous modification strategies involving chemical, physical and biological tools have been established in the last decades [7–10]. The surface modification of titanium implants with micron/nano-topography has been proven useful to mimic the hierarchical characteristics of the bone [11]. Bone undergoes remodeling consisting of sequential osteoclastic resorption and osteoblastic formation to repair the microdamages and replace old bone with new one [12,13]. The resorpted lacunae created by osteoclastic activity have micron-scale features (pits with a diameter of 30–100 m) and nanostructure (collagen fibers left on the surface), exhibiting unique hierarchical structure [14,15]. This kind of hierarchical structure could be the signal for osteoblast attachment, proliferation and differentiation [2]. Many dental and orthopedic implants have been improved for better cellular response by modifying their surface topography at micron- and nanoscale [11,16]. Kubo et al.
170
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Fig. 1. SEM micrographs of MAOed (a and b) and MAO–HTed coatings (d and e) at different magnifications; 3D toughness profile of MAOed (c) and MAO–HTed (f) coatings.
reported that micron- and nano-hybrid topographies enhanced both osteoblast proliferation and differentiation compared to only micron-scale topography [17]. An investigation conducted by Gittens et al. reported that the surface with both micron-scale roughness and nano-scale features promoted osteoblast differentiation as well as local factor production more significantly than surfaces with only micron-scale roughness or nano-scale features [18]. Zhou et al. developed a hierarchical structure by patterning Srdoped hydroxyapatite (HA) nanorods on micro-arc oxidized TiO2 layer and found that the inter-rod spacing played an important
role in mediating cell adhesion [19]. These findings suggest that the modification of surface topography of titanium-based implants in both micron/nano-scale might have a great potential for successful osseointegration with host bone. The modification process of titanium surface with micron/nanotopography is usually achieved by two steps. It starts with the development of micron-scale topography on titanium-based substrate, followed by the superposition of nanostructure to the former modified surface. To create a surface with micron-scale topography on titanium, several techniques are commonly used, such as acid
Fig. 2. SEM–BSI of the cross-sectional morphology of MAO–HTed coating (a); TF-XRD patterns of MAO and MAO–HTed coatings (b); HRTEM image (c) and SAED pattern of scratched crystal plates of MAO–HTed coating.
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
171
bial materials [29]. Dou et al. investigated the cytocompatibility of poly(l-lactic acid)/xonotlite (Ca6 (Si6 O17 )(OH)2 ) and found that the incorporation of xonotlite enhanced the attachment and proliferation of bone marrow stromal cells [31]. Moreover, silicate-based materials are well known to exhibit good biocompatibility and have been proved to promote osteoblast proliferation and differentiation [32,33]. The favorable biological performance of CSH and the possibility to synthesize CSH through HT method have inspired us to develop a novel TiO2 /CSH bi-layer coating through MAO–HT treatment on titanium. The objectives of the current work are twofold. First, we aim to develop a novel TiO2 /CSH bi-layer coating on titanium through MAO–HT treatment in order to take advantages of the micron/nano-topographical structure and the benefits of CSH. Second, to investigate the physicochemical properties, the protein adsorption and the initial cell attachment of such modified hierarchical coating. 2. Experimental procedure 2.1. Coating preparation Grade 2 commercially pure titanium plates were used as substrate material (10 × 10 × 1 mm). The surfaces of substrates were abraded with abrasive paper up to grid 2000 and then successively cleaned with acetone, ethyl alcohol and distilled water. Such treated titanium substrate was referred as s-Ti. A MAO system was composed of the specimen as anode, a stainless steel container as cathode, electrolyte and a pulse power supply system (WHD-20, Haerbin, China). The electrolyte was prepared as follows: firstly, 0.10 mol/L Na2 (EDTA), 0.10 mol/L Ca(CH3 COO)2 ·H2 O and 0.25 mol/L NaOH were in turn dissolved in distilled water with continuous stirring. After that, the solution was stirred for 24 h in order to completely transform Ca2+ into Ca-chelate (Ca(EDTA)2− ). Finally, 0.02 mol/L Na2 SiO3 ·9H2 O was added into the solution. During the MAO process, the applied voltage, pulse frequency, duty cycle and the duration time were set at 250 V, 50 Hz, 50% and 7 min, respectively. After MAO treatment, the samples (referred as MAOed coating) were washed with distilled water and then dried in air. For HT treatment of MAOed coating, Teflon-lined autoclaves with a volume of 20 mL were employed and 4 mL ammonia aqueous solution (pH 11) for each sample was filled into Teflon-lined vessels. The autoclaves were heated at 200 ◦ C for 24 h. After HT treatment, MAO–HT-fabricated samples (referred as MAO–HTed coating) were washed with distilled water and then dried in air.
Fig. 3. C 1s (a), Ca 2p (b) and Si 2p (c) XPS spectra of MAOed and MAO–HTed coatings.
etching, sand blasting and micro-arc oxidation (MAO) [20,21]. To develop surface nanostructure, some other techniques are always employed, such as plasma spray, hydrothermal treatment (HT) and sol–gel method [11,22,23]. Among these techniques, the combination use of MAO and HT treatment (MAO–HT treatment) has received a lot of attention because it is able to produce a controllable surface hierarchical structure regardless the geometrical complexity of the titanium-based substrate [24,25]. Calcium silicate hydrate (CSH) along with series of crystalline minerals and poor ordered phases, is the major hydrate product in Portland cement [26,27]. Recently, several phases of CSH have been investigated on their potential for biomedical applications [28–31]. Tobermorite (Ca5 (Si6 O16 )(OH)2 ·4H2 O) synthesized by HT method was reported to be bioactive and biodegradable by Lin et al. [28]. Another study on tobermorite reported that Ag+ - and Zn+ exchanged tobermorite has a potential to be applied as antimicro-
2.2. Physicochemical properties analysis The specimen surface topography was evaluated by field emission scanning electron microscopy (FESEM, JSM-7001F; JEOL, Japan). 3D roughness profiling of MAOed and MAO–HTed coatings were examined using 3D profiling system (MicroXAM-3D Phase Shift; ADE Co., USA). The microstructure of the CSH crystallites, which were carefully scratched from the surface of MAO–HTed coating, was investigated using a transmission electron microscopy (TEM, JEM-2100; JEOL, Japan) operated at 200 kV. The phase composition of the coatings was characterized by thin-film X-ray diffraction with an incidence angle of 1◦ using an X-ray diffractometer (TF-XRD, Rigaku, Tokyo, Japan). The chemical state of the relevant elements was measured with an X-ray photoelectron spectrometer (XPS, Escalab 250Xi; Thermo Scientific, UK). The XPS test was conducted using an AlK␣ X-ray source and the obtained XPS spectra were calibrated against the binding energy of C 1s in the CH2 group at 284.8 eV. Static water contact angle of the coatings was measured by sessile drop using an optical tensiometer (JC2000C1;
172
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Fig. 4. Schematics showing the mechanism of the nanoplate nucleation on titania surface during HT process.
Powereach, China). At least three samples were used for each group to obtain the mean data. 2.3. Protein adsorption Fluorescein isothiocyanate labelled bovine serum albumin (FITC-BSA, ZKCY-BIO, China) and rhodamin labelled fibronectin (Rhodamin-FN) from bovine plasma (Cytoskeleton, USA) were used as protein models to investigate the protein adsorption. For single protein adsorption, a solution of 20 g/mL Rhodamin-FN diluted in PBS was used. For competitive protein adsorption, a mixture solution containing 2 mg/mL FITC-BSA and 20 g/mL Rhodamin-FN was used. The 24-well plate was kept at 37 ◦ C for 4 h. Subsequently, the samples were washed with PBS for three times and photographed under a fluorescence microscope (Leica, Germany). The amount of adsorbed protein on coating surface was semiquantitatively determined by using imageJ2x analysis software. All the images were converted to 8-bit black/white files and then inverted to a binary image. The OD value was defined with gray level 255 (white) as a value of 0 and gray level 0 (black) as a value of 2.708. In this way, the mean OD value of the edited images is proportional to the fluorescence intensity of corresponding original images.
1.0 mL cell suspension with cell density of 1 × 104 cell/mL was added into each well. LG-DMEM medium supplemented with FBS (10%) and without FBS were used in the cellular suspensions. The seeded cells were cultured at 37 ◦ C in an atmosphere of 5% CO2 and 100% humidity. After 24 h, the samples were gently washed twice with PBS to remove non-adherent cells and transferred to new 24-well plates. To investigate the interactions of cells with material surfaces, the cells were fixed with 2.5% glutaraldehyde in PBS overnight and dehydrated in a series of graded alcohols. After exchanging the alcohol within cells with a series of graded tertiary butanol (25%, 50%, 75% and 100% in ethanol), the samples were freeze dried, sputtercoated with gold and observed using SEM. Cell spreading and attachment were evaluated by immunostaining for vinculin and staining for actin and nucleus. The samples were fixed in 4% paraformaldehyde, permeablized with 0.3% triton X-100 in PBS, and then blocked with 10% goat serum in PBS. The primary antibodies used were rabbit anti-vinculin antibody diluted
2.4. Cell attachment Human osteoblastic MG63 cells (supplied by China Infrastructure of Cell Line Sources, China) were used in the present study because these cells are considered as an appropriate cell model to study initial cell attachment [34]. The cells were cultured in low glucose Dulbecco’s modified Eagle’s medium (LG-DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 ◦ C in an atmosphere of 5% CO2 and 100% humidity. The media were refreshed every 48 h. To investigate the cell attachment, s-Ti, MAOed and MAO–HTed specimens were put into the well center of 24-well plates. Then
Fig. 5. Static contact angle of MAOed and MAO–HTed coatings.
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
173
Fig. 6. Fluorescent images and the calculated fluorescence intensities of the adsorbed FN from a single solution of FN (a–c) and a mixture solution of BSA and FN (e and f) on MAOed and MAO–HTed coatings.
Fig. 7. Fluorescence images of DAPI-stained nuclei of MG-63 cells grown for 24 h on s-Ti (a), MAOed (b) and MAO–HTed (c) coatings and average cell density adhering 24 h after seeding (*p < 0.05, **p < 0.01) (d) 89.
1:100 (Sigma, USA) in tris-PBS (TPBS). The secondary antibodies used were TRITC labelled goat anti-rabbit antibody (Bioss, China) at 1:300. Actin was stained with 5 g/mL phalloidin-TRITC (Sigma, USA) and nucleus was stained with 4 -6-diamidino-2-phenylindole (DAPI). Thereafter, the samples were photographed by fluorescence (Leica, Germany) and confocal microscopy (Zeiss 710, Germany). For cell number quantification in Fig. 8, ImageJ2x analysis software was used. The raw image was converted to an 8-bit file and then converted to a binary image by setting a threshold. Threshold values were determined empirically, giving the most accurate image for a subset of randomly selected photomicrographs. Cell numbers were calculated in randomly selected 1000 m squares by “analyze particles” in ImageJ2x. Five samples per condition were
analyzed and this procedure was repeated five times (n = 5), and the statistical significance of differences in means were determined by one-way analysis of variance (ANOVA) followed by post hoc comparisons with least significant difference (LSD) method. A value of p < 0.05 was considered as statistically significant. 3. Results and discussion 3.1. Physicochemical properties analysis Fig. 1 shows the surface morphology and roughness profile of MAOed and MAO–HTed coatings. MAOed coating has numerous micron-sized and crater-shaped protuberances on its surface,
174
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
Fig. 8. Actin (red), cell nucleus (blue) and Vinculin fluorescence and SEM images of the cells after 24 h of culture on s-Ti, MAOed and MAO–HTed coatings. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
exhibiting a three-dimensional morphology with micron-scale topography (Fig. 1a). The porous surface morphology of MAOed coating, with a pore size ranging from 0.4 to 2 m, is considered to be beneficial for bone in growth and the consequent mechanical interlocking between the implant and bone [6]. The wall of the pores of MAOed coating appears to be smooth at low magnification (Fig. 1a). But at higher magnification, the external surface of the protuberance is noticed to be covered by numerous nano-sized crystals (Fig. 1b). These crystals are titania formed during MAO treatment [35]. MAO–HTed coating is a bi-layer coating which consists of titania layer formed firstly by MAO treatment and second layer formed during the subsequent HT process (Fig. 1d and e). The second layer of the coating is composed of large scale of nanoplates and exhibit nano-scale features (Fig. 1e). These nanoplates, with an average thickness of 10 nm, grow in situ on the surface of the titania layer and distribute homogeneously following the surface curve of the crater-shaped protuberances. The network formed by nanoplates is similar to the nano-scale topography associated with collagen fibrils left by the osteoclasts after bone resorption [36]. As the second layer is very thin, the first layer formed by MAO treatment can preserve its micron-scale surface topography after the HT process (Fig. 1d). Both MAOed and MAO–HTed coatings exhibit submicron-scale surface roughness (Fig. 1c and f). The arithmetic mean surface roughness Ra of the coating decreases slightly from 535 ± 3 nm to 471 ± 16 nm through the superposition of nano-sized structure to the micron-scale topographical surface. These results suggest that
MAO–HT treatment is an efficient way to fabricate a hierarchical surface structure on titanium surface. The cross-sectional morphology of MAO–HTed coating is shown in Fig. 2a. Only the first titania layer is observed because the second layer is very thin and easy to be destroyed during the cross-sectional sample preparation. It is easy to distinguish the coating from the substrate from the SEM back scattered image (BSI) because of their different chemical compositions. The thickness of the coating is uniform with an average value of 3.5 m. There is no discontinuity at the interface of the substrate and the coating, indicating that the coating firmly binds to the substrate. Fig. 2b shows the TF-XRD patterns of MAOed and MAO–HTed coatings. The Ti peaks (JCPDS # 44-1294) are attributed to the substrate. For MAOed coating, both the anatase peak (JCPDS # 21-1272) and the rutile peak (JCPDS # 21-1276) are detected and the former one has stronger intensity than the latter one. It indicates that MAOed coating consists of rutile and predominant anatase. Different from that of MAOed coating, the TF-XRD pattern of MAO–HTed coating shows an additional peak at 29.2◦ , which is the typical peak of CSH [37]. It indicates that the second layer formed by HT is one kind of CSH. The XRD results are consistent with the observation of nano-sized particles after MAO and silicate nanoplates after HT treatment in SEM micrographs in Fig. 1. To determine the exact crystalline phase of the nanoplates formed after HT process, TEM is employed. The high-resolution TEM (HRTEM) image (Fig. 2c) reveals that the CSH nanoplate is an incompletely crystallized layer. The nanoplate formed after HT treatment contains predominant crystalline phase and small amor-
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
phous phase. The selected area electron diffraction (SAED) pattern shows distinct spots corresponding to 6CaO·3SiO2 ·H2 O (Fig. 2d). The current work reveals that CSH coating can be prepared in situ on titanium surface through the incorporation of Ca and Si elements into MAO coating and subsequent HT process. Thus, it introduces an interesting modification to existing MAO technique [38–40]. It is worth mentioning that peaks corresponding to crystalline phases composed of Ca and Si are not observed in the TF-XRD patterns of MAOed coating (Fig. 2b). It may be because these elements exist in the formed compounds as amorphous phases in MAOed coating. Therefore, the chemical states of Ca and Si on the surface of MAOed and MAO–HTed coatings are analyzed by XPS in detail. The high-resolution XPS spectra of TiO2 and B–TiO2 films and their deconvoluted peaks fitted using the Gaussian fitting method are shown in Fig. 3. The C1s spectrum of MAOed coating can be deconvoluted into four components: C C/C H (284.5 eV), C N (286.4 eV), O C O (288.6 eV) and p–p* stacking (290.9 eV) [41,42], indicating that Ca-EDTA complex exists in MAOed coating (Fig. 3a). The presence of Ca–EDTA complex is due to the movement of negatively charged ions (Ca(EDTA)2− ) to the anode during MAO treatment [43]. The C1s spectrum of MAO–HTed coating is composed of three main peaks: C C/C H (284.5 eV), C O (286.1 eV) and O C O (288.6 eV) (Fig. 3a). All these complexes are classically observed and attributed to surface contamination [44]. As shown in Fig. 3b, the deconvoluted peaks of the Ca 2p spectrum of MAOed coating indicate the presence of three Ca-containing compounds: CaO (Ca 2p3/2 at 346.6 eV and Ca 2p1/2 at 349.9 eV) [45], CaSiO3 (Ca 2p3/2 at 347.3 eV and Ca 2p1/2 at 351.5 eV) [46] and Ca-EDTA complex (Ca 2p3/2 at 349.7 eV and Ca 2p1/2 at 353.7 eV Fig. 3b). However, for MAO–HTed coating, the Ca 2p peaks are located at 346.6 eV (2p3/2 ) and 350.2 eV (2p1/2 ), corresponding to Ca in 6CaO·3SiO2 ·H2 O (Fig. 3b). The Si 2p XPS spectrum of MAOed coatings could be divided into two components with binding energy of 102.1 eV and 104.3 eV, which are, respectively, ascribed to CaSiO3 and SiO2 (Fig. 3c) [45]. While for MAO–HTed coating, the Si 2p peak located at 102.8 eV is ascribed to Si in 6CaO·3SiO2 ·H2 O (Fig. 3c). The XPS results indicate that the Ca- and Si-containing compounds in MAOed coating could transform to 6CaO·3SiO2 ·H2 O nanoplates in MAO-HTed coating after HT process. The nucleation mechanism of CSH nanoplates is discussed in Section 3.2.
3.2. The nucleation and growth of CSH nanoplates It is observed that CSH nanoplates can nucleate in situ on the titania surface during HT process. The process of nucleation can be described as a dissolution–precipitation mechanism (schematically shown in Fig. 4). Under HT condition, the Ca- and Si-containing compounds, such as CaO, CaSiO3 , Ca–EDTA complex and SiO2 , can be dissociated into the aqueous environment and thus increase the ion concentrations of Ca2+ and SiO3 2− (Fig. 5a). Meanwhile, the titania surface is attacked by OH− ions in the alkaline environment and transforms to Ti O− surface according to the following reaction: Ti O Ti + OH− → 2 Ti O− + H2 O Therefore, the whole interface between the coating surface and the solution environment becomes negatively charged. Positively charged Ca2+ ions can be attracted to the interface (Fig. 4b). Consequently, a positively charged surface is created through the accumulation of Ca2+ ions. The positively charged interface subsequent attracts negatively charged ions, such as OH− and SiO3 2− to the interface (Fig. 4c). Finally, the enrichment of Ca2+ , SiO3 2− and OH− ions at the interface can induce the heterogeneous nucleation of CSH on the surface of the titania (Fig. 4d). In this way, the CSH nanoplates can form in situ on titania surface during HT process.
175
3.3. Contact angle measurement To evaluate the hydrophilicity of MAOed and MAO–HTed coatings and its impact on protein adsorption and cell adhesion, static water contact angle was measured. As shown in Fig. 5, the contact angles of MAOed and MAO-HTed coatings are 21.3 ± 2.6◦ and 5.7 ± 1.1◦ , respectively. The contact angle is reduced by 15.6◦ after HT process, indicating the enhanced hydrophilicity of MAO–HTed coating. The hydrophilicity of a solid surface is considered to be affected by two main factors: the surface energy and surface topography [47]. In the current case, both surface energy and surface topography might contribute to the enhanced hydrophilicity of MAO–HTed coating, but the later one is considered to be predominant. The surface properties, such as surface topography, surface chemistry and hydrophilicity, influence the protein adsorption process under physiological conditions and thus mediate the cellmaterial interactions [48,49]. In the current study, MAO–HTed coating exhibits hierarchical surface topography, different surface chemistry, enhanced hydrophilicity compared to MAOed coating. The effects of these surface properties on protein adsorption and initial cell attachment are discussed in the following sections. 3.4. Protein adsorption In this work, fluorescence microscopy was used to semiquantitatively quantify the adsorption amounts of fibronectin (FN) with and without the competitive counterpart [50]. Fig. 6 shows the amount of adsorbed fibronectin on both coatings in different solutions. For single protein adsorption, MAO–HTed coating adsorbs roughly 35% more FN than MAOed coating (Fig. 6c). It is considered to be caused by the increase surface area of MAO–HTed coating [48]. However, for competitive protein adsorption, MAO–HTed coating adsorbs roughly 70% more compared to MAOed coating (Fig. 6f). It is found that FN adsorption on MAO–HTed coating is further enhanced during competitive protein adsorption, indicating that MAO–HTed coating has a preference for FN adsorption compared to bovine serum albumin (BSA). This phenomenon may be caused by two mechanisms. One is the charge repulse effect of the excessive negative charge of SiO4 4− [51]. As BSA is negatively charged in PBS [49], the repulsive force between the negatively charged MAO–HTed coating and BSA lead to the inhibition of BSA adsorption. Even though FN is also negatively charged in PBS, it has been proven that FN can be adsorbed to negatively charged surfaces in a more extended or relaxed conformation compared with less charged surfaces [52]. The other mechanism is the hydrophobic interactions. It has been proven that the hydrophobicity is the key determinant for BSA adsorption [53], while FN is known to adsorb successfully in an active conformation on hydrophilic surface [54]. Thus, due to the charge repulse effect and the hydrophobic interactions, MAO–HTed coating might have an effect on inhibiting BSA adsorption without significantly influencing FN adsorption. 3.5. Initial cell attachment To evaluate the initial cell attachment, MG-63 osteoblast-like cells were cultured on s-Ti, MAOed and MAO–HTed coatings for 24 h. The density of attached MG-63 cells on different surfaces was visualized by DAPI-stained nuclei, as shown in Fig 7a–c. It was further quantified using a computer-assisted cell counting program (shown in Fig. 7d). The total number of cells per cm2 is increased by 70% and 130% on MAO and MAO–HT coatings compared to sTi, respectively. However, compared to MAOed coating, the cell density on MAO–HT coating is increased by 30%. Fig. 8 shows the spreading and attachment of MG-63 cells cultured on different coating surfaces after 24 h of culture in the
176
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177
presence and absence of FBS. Compared to the relatively roundish cells cultured in the absence of FBS (Fig. 8j–l), cells exhibit a more spread morphology on coating surfaces in the presence of FBS (Fig. 8a–c). It indicates that the process of cell-material interactions is mediated by protein adsorption. Cells poorly spread on the s-Ti surface and the formation stress fiber is not visible (Fig. 8a). On MAOed coating, cells spread widely with cytoskeleton extensions (Fig. 8b). The micro-scale protuberance could serve as cell-anchoring site for cell attachment (Fig. 8h). For cells attached on MAO–HTed coating, the extension of long filopodia and the formation of lamellipodia on CSH nanoplates are noticed (Fig. 8i). It is reported that filopodia could probe the microenvironment and trigger the extension of lamellipodia through signal transduction [55]. Meanwhile, the formation of actin stress fibers is noticed on both MAOed and MAO–HTed coating (Fig. 8b and c). It has been accepted that the formation of focal adhesion points (FAPs) is an important aspect of osteointegration [56]. FAPs are clusters of integrins linking the cells to ECM [57]. In the current work, focal adhesions are observed to be more in individual cells on MAO–HTed coating, less on the MAOed coating and not obvious on s-Ti (Fig. 8d–f, shown by white arrows). The results indicate that hierarchically topographical surface may contain more signals that mediate MG-63 cell attachment than the surface with single-scale topography. The results show that the initial cell attachment on MAO–HTed coating is enhanced compared to MAOed coating. It is widely accepted that surface properties, such as surface topography, chemistry and energy, affect the type, amount and conformation of adsorbed proteins on material surface and subsequently influence cellular response [19]. In the current work, MAO–HTed coating exhibits significantly different surface properties, such as surface topography and chemistry, compared to MAOed coating. These properties could mediate protein adsorption and subsequently trigger cell response. Due to the hierarchy surface topography, different surface chemistry and enhanced hydrophilicity, MAO–HTed coating exhibits a preferential trend on FN adsorption compared to MAOed coating. As FN adsorption on biomaterial interface has been reported to promote cell adhesion by forming focal adhesion with integrin receptors in fibroblasts [58,59], it is reasonable to assume that the enhanced cell attachment on MAO–HTed coating is indirectly mediated by protein adsorption. The results show that the superposition of CSH nanoplates to TiO2 coating promotes both FN adsorption and the subsequent cell attachment. It indicates that cell-material interactions, such as cell spread and attachment, can be promoted through appropriate surface modifications of materials on the chemical composition and surface topography. The current work suggests that MAO–HT treatment is an effective way to prepare coatings with both micron- and nanohierarchical surface topography on titanium surface, which is essential for altering protein adsorption and initial cell attachment.
4. Conclusions In this study, a novel titania/CSH bi-layer coating with hierarchical surface topography was successfully prepared on titanium substrate through MAO–HT treatment. MAO treatment leads to a micron-scale topographical surface with numerous crater-like protuberances. The subsequent HT process enables the in situ growth of CSH nanoplates on titania surface. Compared to MAOed coating with single-scale surface topography, MAO–HTed coating with hierarchical surface topography exhibits enhanced hydrophilicity, fibronectin adsorption and initial MG-63 cell attachment. These results suggest that MAO–HT treatment is an efficient way to prepare coatings with hierarchical surface topography on titanium surface and the establishment of hierarchical surface topography is essential for altering protein adsorption and initial cell attachment.
Acknowledgments The authors are grateful for the financial support from National Natural Science Foundation of China (51361130032, 51472139) and Doctor Subject Foundation of the Ministry of Education of China (20120002130002). References [1] D. Lakstein, W. Kopelovitch, Z. Barkay, M. Bahaa, D. Hendel, N. Eliaz, Acta Biomater. 5 (2009) 2258–2269. [2] R.A. Gittens, R. Olivares-Navarrete, Z. Schwartz, B.D. Boyan, Acta Biomater. 10 (8) (2014) 3363–3371. [3] A.A.P. Mansur, H.S. Mansur, Mater. Sci. Eng. C-Mater. Biol. Appl. 30 (2010) 288–294. [4] A. Lugovskoy, S. Lugovskoy, Mater. Sci. Eng. C-Mater. Biol. Appl. 43 (2014) 527–532. [5] N. Athanasou, J. Quinn, C. Bulstrode, J. Bone Joint Surg. Br. 74 (1992) 57–62. [6] Y.-T. Sul, C. Johansson, E. Byon, T. Albrektsson, Biomaterials 26 (2005) 6720–6730. [7] L.-H. Li, Y.-M. Kong, H.-W. Kim, Y.-W. Kim, H.-E. Kim, S.-J. Heo, J.-Y. Koak, Biomaterials 25 (2004) 2867–2875. [8] S. He, P. Zhou, L. Wang, X. Xiong, Y. Zhang, Y. Deng, S. Wei, J. R. Soc. Interface 11 (2014) 20140169. [9] G. Orsini, B. Assenza, A. Scarano, M. Piattelli, A. Piattelli, Int. J. Oral Maxillofac. Implants 15 (1999) 779–784. [10] J. Li, X. Liu, Y. Qiao, H. Zhu, J. Li, T. Cui, C. Ding, RSC Adv. 3 (2013) 11214–11225. [11] W. Zhang, G. Wang, Y. Liu, X. Zhao, D. Zou, C. Zhu, Y. Jin, Q. Huang, J. Sun, X. Liu, X. Jiang, H. Zreiqat, Biomaterials 34 (2013) 3184–3195. [12] M. Okamoto, J. Murai, Y. Imai, D. Ikegami, N. Kamiya, S. Kato, Y. Mishina, H. Yoshikawa, N. Tsumaki, J. Bone Miner. Res. 26 (2011) 2511–2522. [13] E.F. Eriksen, Rev. Endocr. Metab. Disord. 11 (2010) 219–227. [14] R.A. Gittens, T. McLachlan, R. Olivares-Navarrete, Y. Cai, S. Berner, R. Tannenbaum, Z. Schwartz, K.H. Sandhage, B.D. Boyan, Biomaterials 32 (2011) 3395–3403. [15] J.-Y. Rho, L. Kuhn-Spearing, P. Zioupos, Med. Eng. Phys. 20 (1998) 92–102. [16] R.A. Gittens, R. Olivares-Navarrete, A. Cheng, D.M. Anderson, T. McLachlan, I. Stephan, J. Geis-Gerstorfer, K.H. Sandhage, A.G. Fedorov, F. Rupp, B.D. Boyan, R. Tannenbaum, Z. Schwartz, Acta Biomater. 9 (2013) 6268–6277. [17] K. Kubo, N. Tsukimura, F. Iwasa, T. Ueno, L. Saruwatari, H. Aita, W.-A. Chiou, T. Ogawa, Biomaterials 30 (2009) 5319–5329. [18] R.A. Gittens, R. Olivares-Navarrete, T. McLachlan, Y. Cai, S.L. Hyzy, J.M. Schneider, Z. Schwartz, K.H. Sandhage, B.D. Boyan, Biomaterials 33 (2012) 8986–8994. [19] J.H. Zhou, Y. Han, S.M. Lu, Int. J. Nanomed. 9 (2014) 1243–1260. [20] M. Herrero-Climent, P. Lázaro, J.V. Rios, S. Lluch, M. Marqués, J. Guillem-Martí, F. Gil, J. Mater. Sci. – Mater. Med. 24 (2013) 2047–2055. [21] R. Zhou, D. Wei, S. Cheng, W. Feng, Q. Du, H. Yang, B. Li, Y. Wang, D. Jia, Y. Zhou, ACS Appl. Mater. Interfaces 6 (2014) 4797–4811. [22] J. Zhou, J. Shao, Y. Han, Appl. Surf. Sci. 303 (2014) 367–372. [23] Y. Zhang, K. Bai, Z. Fu, C. Zhang, H. Zhou, L. Wang, S. Zhu, S. Guan, D. Li, J. Hu, Appl. Surf. Sci. 258 (2012) 2939–2943. [24] Y. Han, J.H. Zhou, L. Zhang, K.W. Xu, Nanotechnology 22 (2011) 11. [25] J. Takebe, S. Ito, S. Miura, K. Miyata, K. Ishibashi, Mater. Sci. Eng. C – Mater. Biol. Appl. 32 (2012) 55–60. [26] I.G. Richardson, Cem. Concr. Res. 38 (2008) 137–158. [27] P. Yu, R.J. Kirkpatrick, B. Poe, P.F. McMillan, X. Cong, J. Am. Ceram. Soc. 82 (1999) 742–748. [28] K. Lin, J. Chang, R. Cheng, Acta Biomater. 3 (2007) 271–276. [29] N.J. Coleman, A.H. Bishop, S.E. Booth, J.W. Nicholson, J. Eur. Ceram. Soc. 29 (2009) 1109–1117. [30] A.P. Hurt, G. Getti, N.J. Coleman, Int. J. Biol. Macromol. 64 (2014) 11–16. [31] Y. Dou, C. Wu, J. Chang, Acta Biomater. 8 (2012) 4139–4150. [32] C.-H. Liu, C. Hung Jr., T.-H. Huang, C.-C. Lin, C.-T. Kao, M.-Y. Shie, Mater. Sci. Eng. C – Mater. Biol. Appl. 43 (2014) 359–366. [33] H. Hu, Y. Qiao, F. Meng, X. Liu, C. Ding, Colloid Surf. B – Biointerfaces 101 (2013) 83–90. [34] E. Czekanska, M. Stoddart, R. Richards, J. Hayes, Eur. Cells Mater. 24 (2012) 1l17. [35] J. Yan, J.F. Sun, P.K. Chu, Y. Han, Y.M. Zhang, J. Biomed. Mater. Res. Part A 101 (2013) 2465–2480. [36] J.E. Davies, Biomaterials 28 (2007) 5058–5067. [37] L.H. Long, L.D. Chen, J. Chang, Ceram. Int. 32 (2006) 457–460. [38] T.-C. Yang, H.-Y. Shu, H.-T. Chen, C.-J. Chung, J.-L. He, Surf. Coat. Technol. 259 (2014) 185–192. [39] Q.-M. Zhao, L. Cheng, H.-L. Yang, Z.-T. Liu, D.-H. Feng, Radiat. Eff. Defects Solids 169 (2014) 1056–1063. [40] S.-D. Wu, H. Zhang, X.-D. Dong, C.-Y. Ning, A.S. Fok, Y. Wang, Appl. Surf. Sci. 329 (2015) 347–355. [41] N.V. Qui, P. Scholz, T.F. Keller, K. Pollok, B. Ondruschka, Chem. Eng. Technol. 36 (2013) 300–306. [42] A. Mesnage, M.A. Magied, P. Simon, N. Herlin-Boime, P. Jégou, G. Deniau, S. Palacin, J. Mater. Sci. 46 (2011) 6332–6338.
Q. Huang et al. / Colloids and Surfaces B: Biointerfaces 134 (2015) 169–177 [43] W. Daqing, Z. Yu, J. Dechang, W. Yarning, Acta Biomater. 3 (2007) 817–827. [44] A. Izena, M. Sakuraba, T. Matsuura, J. Murota, J. Cryst. Growth 188 (1998) 131–136. [45] W. Zhang, K. Du, C. Yan, F. Wang, Appl. Surf. Sci. 254 (2008) 5216–5223. [46] Q. Meng, J. Lin, L. Fu, H. Zhang, S. Wang, Y. Zhou, J. Mater. Chem. 11 (2001) 3382–3386. [47] B. He, N.A. Patankar, J. Lee, Langmuir 19 (2003) 4999–5003. [48] V.V. Divya Rani, L. Vinoth-Kumar, V.C. Anitha, K. Manzoor, M. Deepthy, V.N. Shantikumar, Acta Biomater. 8 (2012) 1976–1989. [49] S. Patil, A. Sandberg, E. Heckert, W. Self, S. Seal, Biomaterials 28 (2007) 4600–4607. [50] M. Sun, J. Deng, Z. Tang, J. Wu, D. Li, H. Chen, C. Gao, Colloid Surf. B – Biointerfaces 122 (2014) 134–142.
[51] [52] [53] [54] [55] [56] [57] [58] [59]
177
X. Chen, T. Wu, Q. Wang, J.-W. Shen, Biomaterials 29 (2008) 2423–2432. B.E. Rapuano, J.J. Lee, D.E. MacDonald, Eur. J. Oral Sci. 120 (2012) 185–194. P.P. Roach, D. Farrar, C.C. Perry, J. Am. Chem. Soc 127 (2005) 8168–8173. J. Lee, B.H. Chu, K.-H. Chen, F. Ren, T.P. Lele, Biomaterials 30 (2009) 4488–4493. R. You, X. Li, Y. Liu, G. Liu, S. Lu, M. Li, J. Biomed. Mater. Res. Part A 102 (2014) 4206–4212. V.D. Rani, K. Manzoor, D. Menon, N. Selvamurugan, S.V. Nair, Nanotechnology 20 (2009) 195101. M.J.P. Biggs, R.G. Richards, M.J. Dalby, Nanomed. – Nanotechnol. Biol. Med. 6 (2010) 619–633. A.J. García, Biomaterials 26 (2005) 7525–7529. M.D. Mager, V. LaPointe, M.M. Stevens, Nat. Chem. 3 (2011) 582–589.