Tantalum coating on TiO₂ nanotubes induces superior rate of matrix mineralization and osteofunctionality in human osteoblasts

Abstract

Nanostructured surface geometries have been the focus of a multitude of recent biomaterials research, and exciting findings have been published. However, only a few publications have directly compared nanostructures of various surface chemistries. The work herein directly compares the response of human osteoblast cells to surfaces of identical nanotube geometries with two well-known orthopedic biomaterials: titanium oxide (TiO₂) and tantalum (Ta). The results reveal that the Ta surface chemistry on the nanotube architecture enhances alkaline phosphatase activity, and promotes a ~30% faster rate of matrix mineralization and bone-nodule formation when compared to results on bare TiO₂ nanotubes. This study implies that unique combinations of surface chemistry and nanostructure may influence cell behavior due to distinctive physico-chemical properties. These findings are of paramount importance to the orthopedics field for understanding cell behavior in response to subtle alterations in nanostructure and surface chemistry, and will enable further insight into the complex manipulation of biomaterial surfaces. With increased focus in the field of orthopedic materials research on nanostructured surfaces, this study emphasizes the need for careful and systematic review of variations in surface chemistry in concurrence with nanotopographical changes.

1. INTRODUCTION

Current research in the orthopedics industry to a large extent has been concentrated on advanced surface technologies that would instigate biological fixation, enhancing patient healing time and reducing chance for aseptic loosening. An increasing number of studies have shown that cells respond to minute changes in surface characteristics such as wettability [1], surface roughness [2-3], surface energy [4], nanotopography [5-8], and surface chemistry [9]. The ideal orthopedic material surface would be composed of a unique combination of these surface properties, promoting the optimum environment for bone ingrowth. A common trend in biomaterials research has been focused especially on the investigation of nanostructured material surfaces, since topographical cues on the nanometer scale have been linked to promoting cellular function and response [10]. However, while nanostructured surface geometries have provided exciting findings in the field of biomaterials research, only a few publications have directly compared nanostructures of various surface chemistries [11]. The history of orthopedic implant materials has made it obvious that the body tissues respond differently to surfaces depending on the type of foreign material [12]. In fact, it has been proposed that there are two major mechanisms of osseointegration: mechanical interlocking through bone growth in pores (i.e. topography), and biochemical bonding (i.e. surface chemistry) [13]. There remains to be insufficient understanding on how these factors interact in producing biological responses [14]. This is largely due to the difficulty in varying surface chemistry and topography independently. Additionally, in vitro and in vivo studies involving surface chemistry are typically multifaceted and complex, due to the multitude of properties that can affect biochemical reactions at the surface (i.e. surface charge, isoelectric point, fluid flow, pH, ionic release from the surface, precipitation of biomolecules from the culture media/biological fluids) [13]. Nevertheless, we believe that a unique combination of surface chemistry and nanostructured geometry may provide a balance of defined characteristics towards an optimal orthopedic implant. Though the majority of related nanotopographical studies compare only nano-textured with non-textured surfaces of the same material, an important addition to this research would be the direct comparison of the same nanostructure with different surface chemistries.

Metallic tantalum (Ta) has been a biomaterial of more recent interest for orthopedic applications, as it has been found to be highly corrosion resistant and bioinert [15], as well as bioactive in vivo, forming a bone-like apatite layer in simulated body fluid that biologically bonds to bone [16]. Tantalum has regained interest in the biomaterials field mainly due to a new porous (trabecular) tantalum material of micro-porosity approved by the FDA in 1997, which has been shown to possess excellent osseointegrative properties [17]. Since then, studies have compared the biocompatibility, bacterial adherence [18] and osteoconductivity [19] of Ta with that of other common implant materials, such as titanium (Ti), cobalt chromium (CoCr), and hydroxyapatite coatings [20]. In particular, a clinical review was published in 2010 demonstrating that a higher degree of implant fixation was obtained in patients who received porous tantalum acetabular cups when compared to those with hydroxyapatite-coated titanium cups (industry gold standard) [21]. Additionally, a recent in vitro study by Sagomonyants, et al. demonstrated that porous Ta even stimulates the proliferation and osteogenesis of osteoblasts from elderly female patients with compromised bone-forming abilities when compared with titanium fiber mesh [22]. However, despite the promising results to-date, the relatively expensive manufacturing cost, as well as the inability to produce a modular all-Ta implant has prevented its widespread acceptance [17]. A simple solution that has been suggested previously [20, 23] is to coat a Ti implant with a Ta film, thus incorporating the Ta surface chemistry while maintaining the mechanical advantages of a Ti implant (i.e. relatively low elastic modulus).

Few studies have been published to-date investigating nanostructured tantalum as a biomaterial [24-25]. One study in 2009 by Ruckh, et al. shows evidence that anodized tantala nanotubes provide a substrate for enhanced osseointegration when compared to flat Ta [24]. However, the study only compares the non-textured surface with the nanotextured surface of the same surface chemistry. Additionally, the nanotubes are of relatively great length (2-11 μm), which has been found in our laboratory to cause a tendency of the nanotube layer to delaminate easily. The relatively unstable nature of this structure is of great concern for an orthopedic implant surface.

Titanium oxide (TiO₂) nanotubes introduced on Ti implant surfaces have proven to be an effective substrate for significantly enhanced osteoblast adhesion and growth [26] as well as noticeably enhanced osseointegration with several times stronger bone-implant adhesion [27] as compared to flat or sandblasted Ti implants. Specifically, vertically aligned TiO₂ nanotubes with 100nm diameter have shown improved stem cell elongation and differentiation [30]. The present work deals with significantly further improved bone growth capability of Ta-modified TiO₂ nanotube surface.

Since our recent work in which we examined the effect of a carbon-coated TiO₂ nanotube surface on osteoblast and osteo-progenitor cells [11], we have been interested in other surface chemistries which may enhance the osteofunctionality on the nanotube surface. In light of the promising findings regarding a Ta biomaterial of microtopography (~500-700 μm pore size), as well as the results of Ruckh, et al., we chose to study the effects of the Ta surface chemistry in direct comparison with TiO₂ on the same nanotopography (i.e. vertically aligned, laterally spaced 100 nm diameter nanotubes), as well as flat surface controls. The objective of the work herein is to compare the response of human osteoblast cells to bare TiO₂ nanotube surface and a Ta-coated nanotube surface in terms of osteofunctionality and bone-forming ability.

2. MATERIALS AND METHODS

2.1. Nanotube substrate fabrication

TiO₂ nanotube surfaces were created using a two electrode set-up anodization process as described previously [28]. A 0.25 mm thick commercially pure Ti sheet (99.5 % metal basis, Alfa-Aesar, USA) was used for this process, which was first cleaned successively in acetone and isopropyl alcohol with ultrasonication followed by deionized (DI) water rinse. The nanotubes were prepared in a 1:7 volumetric ratio of acetic acid (≥99.99 % purity, Sigma-Aldrich, USA) to a weight percent fraction of 0.5 % hydrofluoric acid in water (48 % w/v, EM Science, USA) at 20 V for 30 min. A platinum electrode (99.9 %, Alfa-Aesar, USA) served as the cathode. The samples were then washed with deionized water, dried at 80° C, and heat treated at 500 °C for 2 h in order to crystallize the as-fabricated amorphous structured TiO₂ nanotubes to anatase structure. Tantalum (Ta) films (20 nm-thick) were vacuum-deposited onto TiO₂ nanotube and flat Ti control substrates from a tantalum target in a Denton Discovery 18 sputter system. To ensure preferential coating of the TiO₂ nanotube surface, the deposition angle used was ~30° off the vertical axis with substrate rotation during deposition. 200 W plasma was applied when Ar pressure reached 3 mTorr after base pressure reached 10⁻⁶ torr. The as-deposited Ta film is expected to be of amorphous nature.

2.2. Contact angle measurement

The measurement of contact angle for each experimental flat and nanotube surfaces was carried out using a video contact angle measurement system model VSA 2500 XE (AST Products Inc.). A small deionized water droplet (~3 mg) was placed on the respective surfaces to measure the static contact angle. Measurement of the contact angle is a simple method for analyzing the surface energy and hydrophilic nature of a surface. In this case we also wanted to verify that the tantalum (more hydrophilic in nature compared with TiO₂) had been deposited and to observe the changes in surface energy after deposition.

2.3. Energy dispersive x-ray analysis

The presence of the tantalum thin film coating on the flat Ti and TiO₂ nanotube surfaces was confirmed via energy dispersive X-ray analysis (EDX). The Oxford EDX attachment and Inca Software were used to determine elemental composition of the surface. In addition, matrix mineralization was analyzed by EDX analysis after 7, 14, and 21 d of culture. After the experimental culture time was complete, the cells were trypsinized by trypsin EDTA 0.25 % (Invitrogen, USA) and allowed to air dry for SEM and EDX analysis. Identical samples were incubated in cell-free media as a control. In these studies, EDX was used to determine the elemental composition of deposited bone matrix. The reported atomic percent values were normalized by the relative amounts of Ca and P present on the surfaces incubated under identical conditions in cell-free media.

2.4. Osteoblast cell culture

For these studies, human osteoblast (HOb) cells isolated from a single-donor normal adult human bone (406-05a, Cell Applications, Inc., USA) were used. Each 1 ml of cryo-conserved stock was mixed with 10 ml of alpha minimum essential medium (αMEM; Invitrogen, USA) in the presence of a volume fraction of 10 % fetal bovine serum (FBS; Invitrogen, USA) and a volume fraction of 1 % penicillin-streptomycin (PS; Invitrogen, USA). The cell suspension was plated in a polystyrene cell culture dish and incubated at 37° C in a volume fraction of 5 % CO₂ environment. Each 1.27×1.27 cm² experimental substrate (TiO₂ nanotubes, Ta-coated nanotubes, and flat control Ti and Ta) was placed into individual wells of a 12-well polystyrene plate. The polystyrene (PS) culture dish was used as a control. When the cells reached confluency, the HOb osteoblast cells were seeded at a concentration of 2.5×10⁴ cells per well onto the experimental substrates and stored in a CO₂ incubator for the experimental time durations. For experimental time points beyond 7 d of culture, cell media was changed at 7 d to osteogenic induction media in order to induce osteoblast maturation and mineralization [29]. Osteogenic induction media includes αMEM containing a volume fraction of 10 % FBS, a volume fraction of 1 % PS, 10 nmol/L dexamethasone (Sigma, USA), 150 μg/ml ascorbic acid (Sigma, USA) and 10 mmol/L β-glycerol phosphate (Sigma, USA) [30]. All experimental substrates were moved to a new 12-well dish before cell assays were performed in order to isolate the cells on the substrate of interest from cells on the surrounding polystyrene dish.

2.5. MTT assay

To estimate the metabolic activity of the cells, an MTT (3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyl tetrazolium bromide) assay was employed. After the selected incubation periods, the samples were washed by PBS and transferred to a new 12-well polystyrene culture plate. Fresh cell culture media was added to each well, and the MTT dye agent was added in an amount equal to 10 % of the culture media volume, according to manufacturer's instructions (MTT kit, Sigma, USA). After 2 h of incubation in a 5 % CO₂ incubator, 1 ml of solubilizing solution was added to each well and the polystyrene plate was shaken for 30 s to dissolve the formazan crystals. The absorbance of converted dye of each solution was measured at a wavelength of 570 nm with the subtraction of the 650 nm background by ultraviolet-visible (UV-vis) spectrophotometer (Biomate™ 3, Thermo Electron Co., USA).

2.6. Scanning electron microscopy (SEM) for cell and substrate examination

After 24 h and 21 d of culture, the cells on the substrates were washed with PBS and fixed with a mass fraction of 2.5 % glutaraldehyde (Sigma, USA) in PBS for 1 h. After fixation, they were washed three times with PBS for 15 min each wash. The cells were then dehydrated in a graded series of ethanol (volume fractions of 50, 75, 90 and 100 %) for 30 min each and left in 100 % ethanol to be dried by a critical point dryer (EMS 850, Electron Microscopy Science Co., USA). Next, the dried samples were sputter-coated with palladium metallization for examination by scanning electron microscopy (SEM). The morphology of the samples as well as that of the adhered cells was observed using a Phillips XL30 field emission environmental scanning electron microscope (FEI Co., USA).

2.7. Fluorescence microscopy of cytoskeletal actin

After 24 h of culture, the cells were fixed in 4 % paraformaldehyde in PBS for 15 min at room temperature. Once fixed, the cells were washed twice with wash buffer (PBS containing a volume fraction of 0.05 % Tween-20). To permeabilize the cells, 0.1 % Triton X-100 in PBS was added for 10 min, followed by washing twice with wash buffer. TRITC-conjugated phalloidin (1:1000 Chemicom International) in PBS was added and incubated for 1 h at room temperature, after which the cells were washed three times with wash buffer for 5 min each wash. Samples were then inverted onto coverslips with a dab of Fluoromount-G (Electron Microscopy Sciences, USA), visualized and photographed using a Rhodamine (536 nm excitation) filter by a fluorescence Leica, Co. DM IRB microscope.

2.8. Alkaline phosphatase activity test

In order to measure the bone forming ability of cells on the experimental surfaces, the alkaline phosphatase (ALP) activity was detected using a colorimetric assay kit. Briefly, after the selected incubation periods, the samples were washed by PBS and transferred to a new 12-well polystyrene culture plate. Cells were then gently washed twice with phosphate buffer provided by the kit supplier, followed by the addition of a volume fraction of 0.2 % Triton X-100 in phosphate buffer. The adherent cells were further scraped off of the sample substrate and collected in a microcentrifuge tube. Cell suspension was incubated on ice for 10 min under agitation, and then centrifuged at 2500 ×g for 10 min at 4°C. The supernatants were stored at -60°C until further analysis by AnaSpec SensoLyte pNPP Alkaline Phosphatase Assay Kit colorimetric assay (AnaSpec, Inc., USA) following the manufacturer's instructions. The ALP values were normalized by protein content obtained by a BCA kit (Sigma, USA).

2.9. Alizarin red S staining

Alizarin red staining was employed to detect osteoblast mineralization (calcium deposits). After 7, 14, and 21 d of culture, the cells were fixed for 20 min with 4 % paraformaldehyde (Sigma, USA) and stained overnight at 4 °C with a mass fraction of 2 % alizarin red S (Sigma, USA) in distilled water, and the pH was adjusted to 4.1—4.3 using a volume fraction of 10 % ammonium hydroxide. The samples were then washed with distilled water with gentle rocking three times for 10 min each. The samples were visualized using a Leica DM IRB fluorescence microscope. In order to quantify the alizarin red area on the experimental surface, the microscope fields were analyzed using Image J software.

2.10. Statistical analysis and error bars on graphs

Sigma Plot 11.0 software (2008) was utilized to demonstrate the statistical significance of the assays. The graphs show the average ± standard error bars associated, with p-values listed in the figure captions.

3. RESULTS AND DISCUSSION

Fig. 1 presents SEM images of the flat Ti, flat Ta, and both as-made TiO₂ nanotube and Ta-coated nanotube surfaces. The nanotube images reveal near identical structures, with an outer diameter of ~100 nm, ~10 nm wall thickness, ~10 nm spacing, and ~300 nm height, as previously described [28, 31]. The Ta coating performed by vacuum sputter deposition allows for deposition of a conformal layer with high control of the Ta thickness (20 nm). Ta is reported to be a biocompatible material, its corrosion resistance equivalent to Ti, and both Ta and Ta oxide possess low solubility and toxicity [32]. Titanium and tantalum are similar in that they both form a natural oxide layer when exposed to air, which has been attributed to the excellent biocompatibility of these materials [33-34]. It is thus assumed that the amorphous Ta coatings in this study possess a thin natural oxide layer, and are not entirely metallic. The water contact angles are displayed in yellow in the upper right corner of the SEM images in Fig. 1, demonstrating that the flat Ta surface is more hydrophilic in nature than the flat Ti surface (31° and 54°, respectively). In addition, the Ta coating induced a very slight increase in hydrophilicity from ~4° to ~0° on the TiO₂ and Ta coated nanotube surfaces, respectively.

Figure 1.

SEM images of the flat Ti, flat Ta, TiO₂ nanotube (NT), and Ta-coated nanotube (NT) substrates. The images depict preservation of the nanotube geometry and structure after tantalum coating. The contact angle for each surface is shown in yellow, indicating an increase in hydrophilicity on both tantalum-coated surfaces: from 54° to 31° (flat), and 4° to 0° (nanotube).

The presence of the Ta coating was confirmed viaEDX, which shows the presence of peaks corresponding to Ta as shown in Fig. 2. In addition, the Ta coating resulted in a redish metallic color on the nanotube substrates (in contrast to a greenish/blue color on the as-made nanotube samples). The flat Ta sample showed similar evidence of the Ta coating with EDX analysis and a slight color change to yellow/gold (in contrast to silver metallic color of the as-received Ti foil).

Figure 2.

EDX spectrum illustrating the presence of Ta (red arrows) on the surface of the TiO₂ nanotube surface. The area highlighted by the red box is enlarged in the inset.

An MTT assay was utilized in order to measure the metabolic activity of the cells and to indirectly estimate the number of adhered, viable cells. Results of the MTT analysis in Fig. 3(a) reveal a trend of increasing cell viability until day 14, at which the proliferation begins to decrease. This trend is consistent with published literature, which contributes a peak in osteoblast proliferation (usually at around 12-14 days) as the end of the proliferative period, and the beginning of the matrix maturation phase [35-36]. The MTT data shows no significant difference between the flat control samples and the TiO₂ and Ta-coated nanotube surfaces. This trend is rather unexpected, since prior findings in our laboratory have shown significant acceleration of MC3T3-E1 osteoblast cell growth on TiO₂ nanotube surfaces when compared to flat Ti controls [28]. However, the differing results observed in this study could be due to the use of primary osteoblast cells isolated from normal human bone in contrast to the highly proliferative murine cell line used in prior studies. The literature also contains contrasting trends of cell adhesion and proliferation on nanostructured surfaces, which seem to vary based on differing cell type, species, source and maturation [37-38]. Moreover, many of such experiments are performed using very robust and proliferative cell lines, which may bias the findings [37].

Figure 3.

(a) MTT assay data showing the optical density (OD) of the reaction product of the MTT working solution of HOb cells cultured on the nanotube surfaces as a function of incubation time (n = 3). The line graph shows the mean ± standard error bars. (b) SEM micrographs of HOb cells after 24 h incubation, showing extensive filopodia activity on both nanosurfaces (yellow arrows). Scale bars = 5 μm.

SEM morphological examination shown in Fig. 3(b) after 24 h of culture reveals extensive filopodial activity on both TiO₂ and Ta surfaces, but not on the flat control surfaces (as indicated by the yellow arrows). A common speculation is that finger-like filopodia are a cell-sensing mechanism which are used to detect both chemical and nanotopographical cues [39]. An increase in filopodial activity has been demonstrated previously on both TiO₂ [31] and ZrO₂ [40] nanotube architectures when compared to respective flat controls surfaces. The presence of many filopodia on both nanotube surfaces indicates that the HOb cells are relatively equally activated by the nanotube architecture, independent of surface chemistry.

In addition, cytoskeletal actin organization and cell morphology was also unaffected, as depicted in Fig. 4. The figure shows immunofluorescent staining of the cytoskeletal actin of HObs after 24 h of culture on the flat surfaces and both TiO₂ and Ta nanotube substrates. It is apparent that the cells on the nanotube surfaces possess a criss-cross pattern within the cell body, as signified by the yellow arrows. The unique cytoskeletal arrangement observed on the nanotube substrates is likely to be a result of an altered placement of proteins adsorbed onto the nanotopographical features from the culture media [31], as well as extracellular protein such as fibronectin and vitronectin [41]. Nanotopography has also been shown to cause distinctive integrin and focal adhesion plaque placement on the surfaces [42], which in turn affects cytoskeletal arrangement [43]. The actin organization by visual comparison appears to be minimally effected by the two surface chemistries compared in this work. These results are perhaps in agreement with prior results which showed no significant difference between proliferation, attachment, or cytoskeletal arrangement between human osteoblast cells cultured on Ta and Ti substrates [19].

Figure 4.

Immunofluorescent images of cytoskeletal actin (red) of HOb cells on flat and nanotube surfaces after 24 h of culture incubation, showing a criss-cross pattern on both TiO₂ and Ta surfaces (yellow arrows). Scale bars = 50 μm.

ALP activity was measured as a function of incubation time to estimate the bone-forming ability of osteoblast cells on the experimental substrates (Fig. 5). As mentioned earlier, the peak in cell growth at day 14 (Fig. 3) is consistent with the progressive development of osteoblasts demonstrated in the literature, which indicates the end of the proliferation stage, and the onset of extracellular matrix maturation. In concurrence with a decrease in cell proliferation, an up-regulation in ALP activity in osteoblasts occurs during the matrix maturation phase [35-36]. This reciprocal relationship is evident in comparing Fig. 3 and 5: the onset of increasing ALP activity (at ~day 10-14) with decreasing cell proliferation (at day 14) indicates proper osteoblast matrix maturation [36]. It is also apparent that the ALP activity is increased on both nanotube surfaces when compared to their respective flat surfaces (Ti/TiO₂ NT, and Ta/Ta-NT). This observation is expected, since prior studies have demonstrated enhanced bone function on nanotube architectures when compared to flat controls for TiO₂ [31], ZrO₂ [40], and Ta₂O₅ [24].

Figure 5.

ALP activity of HOb cells cultured on the nanotube surfaces vs. incubation time (n = 3). The graph points show the mean ± standard error bars. The p-values after performing an ANOVA test reaching statistical significance (p < 0.001) are marked on the graph (*).

However, when comparing the two nanotube surfaces in this study, it is interesting to note that no difference was observed at shorter time points; conversely at 10 days and beyond, a significantly higher ALP activity was detected on the Ta-coated nanotube surfaces when compared to all of the other samples. The TiO₂ nanotube ALP activity at days 10-21 do not follow the same trend as observed on the Ta-coated nanotube surface. In fact, it appears that the osteoblasts on the TiO₂ nanotube surface exhibit a slightly delayed response at day 14 in comparison to the Ta-coated nanotube surface. Since the nanotube surfaces are of identical topography, but with differing surface chemistry, one could speculate that the Ta chemistry may enhance the osteogenic functionality of the HOb cells on the nanotube surface. However, the ALP levels of the flat Ta compared to the flat Ti do not clearly support this claim (flat Ta is lower than Ti at day 3, 10, and 14, although the difference is not statistically significant). Additionally, the literature comparing in vitro osteogenic behavior on Ti and Ta surfaces is not entirely consistent. An increased ALP activity and osteogenesis-compatibility was demonstrated by multiple studies on tantalum substrates when compared to titanium substrates with similar topographies [32, 44-45]. On the other hand, other studies have demonstrated that the osteogenic behavior on Ta and Ti are relatively the same, although the topography of the comparative surfaces was not identical [19]. The fact that we only saw a significant upregulation on the nanostructured tantalum surface could indicate that there is a unique combination of physico-chemical properties on this surface that is not present on the others. For example, while the TiO₂ nanotube surface is hydrophilic (contact angle of 4°), the Ta-coated nanotube surface is superhydrophilic, with a contact angle of 0°. This is in great contrast to the flat Ti and Ta rather hydrophobic contact angles (54° and 31°). Furthermore, it is known that surface hydrophilicity is directly related to the energy at the surface of a biomaterial, which is defined by the general charge density and net polarity of the charge [47].

In understanding the relative charge of a surface, a discussion of isoelectric points is necessary. It is well known that the surface of a metal oxide film terminates in an outermost layer of hydroxyl groups [48]. In an aqueous solution with a pH equal to the isoelectric point, the surface hydroxyl groups will remain undissociated [48]. Although the isoelectric points of metal oxides vary based on factors such as temperature and crystal structure as well as method of measurement [49], the estimated isoelectric point of anatase phase titanium oxide (6.1 [50]) is much higher than that of tantalum oxide (2.7-3.0 [48]). However, the optimal pH for human osteoblast cell culture in vitro is ~7.2 [51] (most culture media have pH=7.4). Since the pH of the culture media (7.4) is greater than the isoelectric point of both TiO₂ (6.1) and Ta₂O₅ (2.7-3.0), the surface will acquire a negative charge by either of the following reactions [48]:

$$-{\mathrm{MOH}}_{\mathrm{surf}}+{\mathrm{OH}}^{-}\leftrightarrows -{{\mathrm{MO}}_{\mathrm{surf}}}^{-}+{\mathrm{H}}_2\mathrm{O}$$ (1)

or

$$-{\mathrm{MOH}}_{\mathrm{surf}}\leftrightarrows -{{\mathrm{MO}}_{\mathrm{surf}}}^{-}+{{\mathrm{H}}_{\left(\mathrm{aq}\right)}}^{+}.$$ (2)

However, for a lower isoelectric point, more negative charges will accumulate on the surface [52]. This means that it is probable that the Ta-coated surfaces are more negatively charged than the TiO₂ surfaces, leading towards greater affinity of positively charged ions and molecules to accumulate on the surface (i.e. Ca²⁺, Mg²⁺, Na⁺, which are present in cell culture media). The presence of such ions on a surface has been shown to increase osteoblast function in vitro [53-55], as well as enables the surface to bind electrostatically to a variety of proteins [56]. Furthermore, a study by Moller, et al. revealed that a higher degree of osteoblast differentiation was achieved on a biocomposite with the lowest isoelectric point [57].

Another factor that may affect the relative charge of the surfaces is that the TiO₂ nanotubes were heat treated in order to transform the as-made amorphous structure to anatase (for osteoblast preference [28]). Heat treatment has been shown to diminish the amount of OH groups on a surface [58], and is likely to affect the overall surface charge. In contrast, the Ta-coated nanotube surface was heat treated, then deposited with a thin layer of Ta, which spontaneously forms hydroxyl groups on the surface. Lastly, the nanostructured surface presents a much larger surface area than a flat surface. The larger surface area would introduce more active sites and active OH groups on the nanotube surfaces than on the corresponding flat substrates [58]. As a result, the combination of the nanostructure and more negatively charged Ta surface coating of the Ta-coated nanotube surface may be more highly active for ion and biomolecule accumulation on the surface, thus enhancing the osteoblast cells ability to form mature extracellular matrix (which has been linked to upregulation in ALP activity [36]). The interplay between these physico-chemical differences and osteoblast behavior is not well-understood, and their possible effects in this study are only speculation. However, the Ta-coated nanotube surface demonstrates a significantly higher ALP activity than the TiO₂ nanotube surface as well as flat controls, which may signify a greater extent of matrix maturation at this stage [36].

The third phase of osteoblast development is the onset of matrix mineralization, which is essential for expression of osteoblast phenotypical genes [35]. In order to evaluate the degree of matrix mineralization of the bone cells on each experimental surface, the osteoblasts were analyzed for bone nodule formation via various analytical techniques. After 21 d of culture, the osteoblast morphology on the nanotube surfaces was assessed by SEM, which revealed the presence of large bone nodules [59] on both nanotube surfaces, with very little nodule-like formation on the flat controls (Fig. 6(a)). The presence of bone nodules was also visible on the positive control tissue culture plastic (data not shown). In addition, EDX analysis revealed significantly higher amounts of both phosphorus and calcium on the Ta-coated nanotube samples than was found on all other samples, as shown in the graph in Fig. 6(b). This indicates that although bone nodule formation readily occurred on both nanotube surfaces, the Ta-coating appears to have had the effect of inducing increased deposition of bone matrix minerals. The formation of bone-like apatite on tantalum metal in simulated body fluid has been previously reported [55], as well as on Ta treated with sodium hydroxide (NaOH) [60]. Furthermore, a prior study by Oh, et al. demonstrated the formation of nano-scale hydroxyapaptite on the surface of TiO₂ nanotubes after treatment with NaOH [61]. It would be interesting to compare the apatite-formation behavior of the TiO₂ and Ta-coated nanotube array in simulated body fluid. Osteoblast cells have been shown to preferentially differentiate to form mineralized extracellular matrix on apatite layers [60]. The speculation that a Ta nanostructure has apatite-inducing properties supports the hypothesis that it also encourages the production of mineralized matrix by HOb cells.

Figure 6.

Bone nodule formation by HOb cells cultured for 3 weeks. (a) SEM micrographs at 1000x showing larger bone nodule formation on the Ta-coated NT surface. Scale bar = 20 μm. (b) EDX analysis of the atomic percent of calcium and phosphorous mineral elements on the surfaces (n = 5). The bar graph shows the mean ± standard error bars. The p-values after performing an ANOVA test reached statistical significance (p ≤ 0.001), as indicated by (*).

The kinetics of matrix mineralization was also examined in order to determine whether the rate of mineralization was affected by the nanostructure or surface chemistry. EDX analysis estimating the atomic percent of phosphorus (Fig. 7(a)) and calcium (Fig. 7(b)) on each substrate after 7, 14, and 21 d of culture revealed that the highest rate of mineralization occurred on the Ta-coated nanotube surface. A linear trendline was estimated for the mineral atomic percent as a function of incubation time for each sample type, and the slope of each line was determined and recorded as the rate of phosphorus or calcium deposition (Fig. 7(c)). The rates of both phosphorus and calcium deposition are estimated to be roughly 30% faster on the Ta-coated nanotube substrate than on the TiO₂ nanotube substrate.

Figure 7.

Mineralization kinetics study. EDX analysis of the atomic percent of phosphorus (a), and calcium (b) mineral elements on the surfaces as a function of time (n = 5), with linear trendlines overlaid and correlation coefficients labeled. (c) A table of the corresponding rates (slopes of the linear trendlines) of phosphorus and calcium deposition for each substrate. The line graphs show the mean ± standard error bars. The p-values after performing an ANOVA test reached statistical significance (p ≤ 0.001) for all comparisons between samples except for Ti vs. Ta.

The results of matrix mineralization kinetics were verified by alizarin red S staining, a simple and convenient method for detecting calcium mineral deposition. The fluorescent images show the stained area (bright red) on each experimental surface after 1, 2, and 3 weeks of culture in Fig. 8(a) (from left to right). After 1 week, only small amounts of mineral were detected on the TiO₂ and Ta-coated nanotube surfaces, while nothing was visible on the flat substrates. After 2 weeks, more highly concentrated areas of calcium mineral deposits (indicated by arrows) were visible on the nanotube surfaces, with a few on the flat surfaces. After 3 weeks, large areas of several bone nodules were present on the nanotube surfaces; the nodules on the flat surfaces were fewer and less pronounced. The fluorescent staining was semi-quantified using ImageJ analysis in order to validate the relative surface areas covered in bone-like mineral shown in Fig. 8(a). This data is presented in Fig. 8(b), and reveals a higher stained surface area on the Ta-coated nanotubes than all other surfaces. These results verify the observations by EDX analysis. In addition, increased alizarin red staining was reported by Stiehler, et al. of MSCs on Ta when compared to Ti, which supports our findings [32].

Figure 8.

Alizarin red staining for mineral deposition by HObs cultured for 1, 2, and 3 weeks. Fluorescent images show alizarin red staining (bright red) for mineral deposition. Arrows indicate bone nodule formation. Scale bar = 50 μm.

Both the TiO₂ and Ta-coated nanotube surfaces enhanced osteogenic function over that of flat controls of smooth Ti and Ta-coated smooth Ti. However, the Ta-coated nanotube surfaces had superior osteofunctionality in terms of ALP activity, bone nodule formation, and the rate of matrix mineralization. These results indicate that HOb filopodial activity and cytoskeletal arrangement may be influenced primarily by nanotopographical cues (they were similar on both nanotube surfaces, but different on the flat surfaces). However, the osteogenesis may be more highly influenced by surface chemistry/material properties than nanotopography. This hypothesis coincides with our prior findings on carbon-coated nanotube surfaces [11]. Furthermore, some studies have shown that osteoblast differentiation in vitro is more highly influenced by differences in surface chemistry irrespective of surface microtopography [62-63]. Another experiment by Cyster, et al. demonstrated that fibroblasts preferentially attached to titanium nitride surfaces of a defined surface chemistry at early time points; however at later time points, preference was shown towards a nanostructured surface, with no preference for surface chemistry [64]. While the interplay of the surface chemistry and nanotopography are not completely understood, it is apparent that unique combinations can have substantial results, and that cell preference for certain properties may change over time.

4. CONCLUSION

Herein we have compared the behavior of human osteoblast cells on TiO₂ and Ta-coated nanotube surfaces of nearly identical nanotopography (and the respective flat controls surfaces), in order to assess the effect of changes in surface characteristics due to a Ta coating alone. It was determined that both nanotube surfaces instigate similar levels of cell adhesion, proliferation, and morphology. However, at advanced culture times, the osteofunctionality was enhanced on the Ta surface in terms of alkaline phosphatase activity, bone nodule formation, and matrix mineral deposition. In fact, the Ta surface promoted a ~30% faster rate of matrix mineralization and bone-nodule formation when compared to results on bare TiO₂ nanotubes, indicating that a surface chemistry modification superimposed onto nanotopography provides an additional benefit to the bone growth behavior. These findings are of paramount importance to the orthopedics field for understanding cell behavior in response to subtle alterations in nanostructure and surface chemistry, and will enable further insight into the complex manipulation of biomaterial surfaces.

We believe that a unique combination of surface chemistry and nanostructured geometry may provide a balance of defined characteristics towards an optimal orthopedic implant, and this study indicates that osteoblasts may prefer nanostructured Ta to nanostructured TiO₂ in vitro. An explanation for the observed behavior is not straightforward due to lack of studies comparing nanostructured Ti and Ta surface behavior in vitro. Further studies are required to gain insight to this phenomenon, such as comparative analysis of protein adsorption, hydroxyapatite formation, and biomolecule interaction with the TiO₂ and Ta-coated nanotube surfaces. Future studies should also compare an amorphous Ta coating with that of crystalline Ta on the nanotube surface, since differences have been observed in osteoblast response to amorphous and crystalline TiO₂ nanotubes [28]. Additionally, it has been found that oxide layer thickness can affect protein adsorption on tantalum [34]. Hence it would be of interest to assess the effects of Ta surface oxide content (i.e. natural oxide layer versus a fully oxidized Ta layer) on osteogenic function. Lastly, it has been demonstrated that different cell types (i.e. osteoblasts versus osteoprogenitor cells) have different preferences for surface chemistry [11], thus forthcoming research should compare the response of different cell types on varied surface chemical/nanotopography combinations.

The work presented herein demonstrates the highly sensitive nature of osteoblast cells to seemingly minute and simple modifications to a surface. With increased focus in the field of orthopedic materials research on nanostructured surfaces, this study emphasizes the need for careful and systematic review of variations in surface chemistry in concurrence with nanotopographical changes.