Synthesis of nanorod apatites with templates at critical micelle concentrations and in vitro evaluation of cytotoxicity and antimicrobial activity

ABSTRACT This study aimed to investigate the synthesis, cytotoxicity, and antimicrobial activity of nanorod apatites obtained using different surfactants at their critical micelle concentrations via hydrothermal method. Nanoscale apatite was obtained from ionic solutions without a template (nHA) compared with synthesized nanorod apatites of T-nHA, S-nHA, F-nHA and P-nHA with four different templates, i.e. the cationic/cetyltrimethylammonium bromide (CTAB), anionic/sodium dodecyl sulfate (SDS), nonionic/Pluronic F-127, and zwitterionic/cocamidopropyl betaine (CAPB) surfactants. Results showed that all of the synthesized apatites have a nanoscale rod-shaped morphology with bacteriostatic properties on day 1. However, only the nanorod apatite of T-nHA demonstrated long-term antibacterial activity up to day 14 due to the combined nanoscale-sized effects and surface phenomena. Among the nanorod apatites produced by the surfactant molecular geometry and solution conditions, the synthesized nanorod apatite of P-nHA possessed the smallest homogeneous crystals. Cytotoxicity results revealed that the nanorod apatites of nHA and F-nHA present insignificant cytotoxicity. Given its acceptable bacteriostatic effect and biocompatibility, the F-nHA may be considered better than nHA. Compared with the conventional-sized apatites, surfactant template-assisted nanorod apatite of T-nHA with high antimicrobial activity may be used as composite grafts for reconstructive surgery to improve inflammation that may be caused by bacteria.

Apatite is generally deposited with intrinsic collagen fibers perpendicular to the surfaces of osteogenesis. Fully developed bundles of nanorod apatite were found between or within collagen fibers and organized in nematic symmetry, with rough cross-sectional widths of 33-65 nm and lengths of 100-1000 nm [11]. Synthetic methods to obtain nanorod apatites are in great demand because these materials mimic natural bone architecture with numerous hydroxyl groups [12]. Nanorod apatites could be synthesized by various strategies, such as hydrothermal method [13], surfactant-mediated crystalline growth [14], solgel synthesis through phase transition of crystalline β-Ca 3 (PO 4 ) 2 [15], and ultrasonication with precipitation [16]. All of these strategies allow good control of the size and length-to-width aspect ratio of the resultant nanorod apatites. As precedent studies revealed the cytotoxic and antiproliferative effects of nanorod apatite [8,9,[13][14][15][16][17], the biocompatibility of nanorod apatite obtained through various synthetic procedures appears to be strongly related to their surface conditions. However, the synthetic mechanisms through which these nanorods exert their cytotoxic effects remain unknown and must be verified by experiments. Apatite nanoparticles could achieve the substitution of Ca 2+ ions with other metal ions, such as Cu 2+ , Zn 2+ , and Ag 2+ , or the substitution of PO 4 3− ions by CO 3 2− ions without changing its initial structure and properties [18,19]. Another feature is the drug-grafting nanorod apatite that could be taken up by the cells and transported to the intracellular compartments [20,21]. Accordingly, even though these nanoparticles are still controversial in terms of biocompatibility, one of their substantial application nowadays is binding to many antibacterial agents, organic or inorganic, toward developing new compounds with high biocompatibility and antimicrobial properties [18][19][20][21]. The synthetic processes of apatite could be roughly classified as dry methods [22], wet processes [23][24][25], synthetic surfactant templates [26], and hightemperature processes [27]. The shape (e.g. sphere, cylinder, or lamella) and size of apatite differ in accordance with the synthesis method employed. The surfactant-based process could inhibit excessive agglomeration of nanoparticles because the morphology control of nanorod apatite could be achieved by restricting crystal growth. Although many effective surfactant template-assisted methods to synthesize apatite have been widely reported under hydrothermal conditions [28][29][30], the method used to prepare apatite in the present study was not on the same synthesis as that of surfactants at critical micelle concentrations (CMCs). In the present experiment, four types of surfactants were used as follows. First, a cationic surfactant cetyltrimethylammonium bromide (CTAB) was ionized in an aqueous solution to produce a positively charged monomer. This cationic surfactant could self-assemble into micelles and after reaching the concentration higher than CMC, it could be used as a surface nucleation site for PO 4 3− and Ca 2+ ions when its micelle size reaches 0.7-2.5 nm. The nucleation reaction could lead to the growth of nanoscale HA (nHA) crystals with uniform morphology and size via low-temperature hydrothermal method and precipitation [31,32]. Second, the anionic surfactant sodium dodecyl sulfate (SDS) is commonly used in detergents, foaming agents, and skincare products, and it could be used to synthesize nanorod apatite [33]. Third, Pluronic F-127 is the trade name of poloxamer 407 with a specific triblock copolymer. Poloxamer 407 is a hydrophilic nonionic surfactant of the general class of copolymers known as poloxamers. The hydrophobic polypropylene oxide (PPO) group in the middle of the surfactant molecule links two hydrophilic polyethylene oxide (PEO) groups. Pluronic-based micelles with an estimated diameter of 30-50 nm are spontaneously formed at a concentration equal to or higher than the CMC [34]. Fourth, the zwitterionic surfactant cocamidopropyl betaine (CAPB) is an amphoteric synthetic detergent that is increasingly used in cosmetics and personal hygiene products [35]. The use of nontoxic reagents during manufacturing is necessary to decrease the toxicity and increase the cell viability of the resultant products in vivo [36].
As apatite nanoparticles synthesized without the use of surfactants showed a higher tendency to aggregate than those synthesized with the use of surfactants [37], four types of surfactants, including cationic, anionic, nonionic, and zwitterionic surfactants, were used as templates in the present study. The addition of controlled CMCs of the surfactants in nanorod apatite synthesized as templates through wet-chemical process under hydrothermal conditions was investigated. The bacteriostatic properties and biocompatibility of the synthesized nanorod apatite were also characterized and compared.

Materials
CTAB was purchased from Ferak (01739, Berlin, Germany). SDS (≥ 99.0% purity) and Pluronic F-127 (P2443, M w = 12.6 kg mol −1 ) with the triblock copolymer PEO-PPO-PEO was purchased from Sigma-Aldrich. CAPB was obtained from the First Chemical Group (20,130,312,042). The different chemical structures of the surfactants are shown in Figure 1a, and the different groups of nanorod apatite synthesized using various surfactant-templated precipitation and their surfactant concentrations at the CMCs are listed in Table 1.

Synthesis of nanorod apatite
Five groups of nanorod apatite were synthesized by dissolving 0.215 g of calcium nitrate (PanReac, C0150) and a specific amount of each surfactant (Table 1) in 90 mL of deionized water. Then, 0.713 g of diammonium phosphate (HSE Pure Chemicals, KHA1050) and 2.647 g of sodium citrate (PanReac, 0001113173) were added to 70 mL of deionized water. The ionic solutions were vigorously stirred with a magnetic bar for 30 min. The pH of each system was carefully adjusted to 7.0 within 5 min, and the different solutions of surfactants, as indicated in Table 1, at their CMCs were reacted under hydrothermal conditions at 180°C for 12 h. The resulting water vapor pressure was increased to approximately 190 psi (lb/in 2 ). The fine precipitates formed were centrifuged, washed with deionized water and ethanol, dried at 50°C for 12 h, and then stored in a dry box to avoid the effects of humidity before further characterization. The yield ratio of nanorod apatite obtained from hydrothermal synthesis relative to the theoretical yield (w/w) was approximately 30% in all groups.

Fourier-transform infrared spectroscopy (FTIR)
The synthesized nanorod apatite was evaluated via FTIR (Nicolet 6700, Thermo Fisher Scientific, Waltham, MA, USA). Scanning was conducted in the wavenumber range of 650-4000 cm −1 to determine the functional groups present in the products.

X-ray diffraction (XRD) analysis
The crystal phases of the obtained apatite were evaluated using XRD (XRD-6000, Shimadzu, Japan) with CuKα radiation at 30 kV and 20 mA. The scan rate was 2°/min, and scanning was conducted in the 2θ range of 20°-60°.

Antibacterial abilities
Staphylococcus aureus (ATCC No.: 25,923) and Escherichia coli (ATCC No.: 10,798) were cultured in Luria-Bertani broth. The bacterial suspensions were diluted to achieve an optical density at 595 nm (OD 595 ) between cell numbers of 0.2-1 × 10 7 cells/mL of bacteria, which was confirmed with an ELISA reader. The bacterial suspensions of S. aureus and E. coli were subsequently diluted to achieve an OD 595 value of 0.2. Then, on the basis of the effectiveness and difference of multiple pre-trial attempts, 2 mL of each suspension was placed in a sterilized tube containing 0.03 g of the as-prepared  nanorod apatite, and changes in bacterial growth were observed during incubation at 37°C for 1-14 days.
A 100 μL culture medium suspension was prepared, and the number of bacteria was determined using an ELISA reader by measuring the OD 595 value (n = 3).

Cell viability
The L929 cell line was derived from newborn mouse fibroblasts and provided by the National Institute of Health in Taiwan. This cell line was used to test the cytotoxicity of the as-synthesized nanorod apatite. The solute samples were prepared at a sample-to-medium ratio of 1 g/5 mL; here, the apatite samples were immersed in the culture medium for 24 h. In accordance with the cytocompatibility indicated in the ISO 10,993-5:2009 guideline, the cell viability was determined using the extracts of different groups to culture the L929 cells. Assay was carried out in triplicate (n = 3). The cells were seeded in a 96-well culture plate at a density of 1 × 10 4 cells per well in minimum essential medium alpha (Gibco, Invitrogen Taiwan, Ltd., MD) containing 10% horse serum culture medium (Biological Industries, Haemek, Israel) and allowed to attach overnight. The culture medium was removed, and the nanorod apatite extracts were added to the wells. Culture was performed for another 24 h. The presence of metabolically active cells was measured by reducing tetrazolium salt (Cell Proliferation Assay, Biological Industries, Israel) to formazan. Thereafter, the viability of the L929 cells was determined at OD 492 by using an ELISA plate reader (EZ Read 400, Biochrom, Cambridge, UK). The morphological characteristics of the cells were observed under an optical microscope (IX71, Olympus, Japan).

Statistical analysis
The mean length, diameter, and aspect ratio of the apatite samples (n = 40) were analyzed using a two-sample t-test and ANOVA for statistical analysis (IBM SPSS Statistics 20 software).

Preparation mechanism for nanorod apatites
The chemical structures of the various surfactants are shown in Figure 1a, and schematics of the relevant hexagonal growth mechanisms of the different nanorod apatites synthesized from different surfactants at their CMCs are shown in Figure 1b. The apatite in nucleated embryos specifically and preferentially grows into hexagonal crystals, especially under hydrothermal conditions [23][24][25][28][29][30]38]. The micellar space limitations and ionic effects in the solution appeared to affect the shapes (a(b)-plane faces) and growth directions (c-axis) of the resulting crystals [31][32][33][34][35]. Nanorod apatite is preferentially oriented toward the c-axis, thus leading to the development of a(b)planes featuring positively charged Ca 2+ ions and c-planes featuring negatively charged PO 4 3− and OH − [38,39]. Accordingly, nanorod apatite surfaces exhibit anisotropic charged characteristics, leading to anisotropic adsorption profiles for template molecules (Figure 1). Micelle-templated precipitations were conducted in the presence of surfactant micelles as a nanostructured template. The shape, size, and charges of a micelle could be utilized for preparing nanorod apatites with regulated length-to-width aspect ratio. Therefore, when anionic SDS was used as the surfactant, excess PO 4 3− ions could not attach to the micelles easily due to the increased electrostatic repulsion charges, leading to a large length-to-width aspect ratio [37][38][39][40][41].

Nanorod apatite growth and FTIR spectral analysis
The functional groups of the synthesized nanorod apatite are shown in Figure 2. peaks of the functional groups of zwitterionic CAPB and anionic SDS were not detected in the corresponding spectra. By contrast, the characteristic bands of the cation CATB template-assisted T-nHA and anion SDS template-assisted S-nHA groups, including peaks at 2853 and 2927 cm −1 , were due to the symmetric and asymmetric stretching vibrations of -CH 2 , respectively.   These peaks indicated that the methylene groups of CATB and SDS clearly adhered to the nanorod apatite surfaces [39]. The peak at 1384 cm −1 was attributed to NH 4 + [40], thus reflecting the presence of (NH 4 ) 2 HPO 4 residues in nHA/control, S-nHA, F-nHA, and P-nHA.

XRD phase identification
The phase analyses of the nanorod apatite are shown in Figure 3.  [42]. The FTIR and XRD results indicated that a large amount of apatite and small amounts of DCPD and CaP 2 O 6 were formed when synthesis was carried out using (NH 4 ) 2 HPO 4 . The characteristic peaks of DCPD and CaP 2 O 6 were metastable, and they were produced by the incomplete solid-state reaction during hydrothermal preparation.

TEM morphological observation
The precipitate dimensions in Figure 4 illustrate the effect of the surfactants at their CMCs on the size of the nanorod apatite. All of the images showed that the nanorods have a large length-to-width aspect ratio and the shapes of the products are strongly associated with the surfactant used for their synthesis. In particular, the surfactant effectively restricted and unified the nucleus growth of the synthesized apatite during precipitation. The XRD, FTIR, and TEM results (Figures 2-4) confirmed that nanorod apatite was formed. Growth appeared to occur via a layering mechanism, and lengthened growth times resulted in the development of planar crystal faces and rod-shaped morphologies at the initial growth stage [13][14][15][16].
The average dimensional scales and length-to-width aspect ratios of the nanorod apatite are summarized in Table 2. The length of the nanorod apatite in each group ranged from 50 nm to 80 nm, and their widths ranged from 12 nm to 21 nm in diameter. The nanorod apatites synthesized with the T-nHA and F-nHA templates were more uniform than those obtained from other surfactants, and the nanorods obtained from P-nHA were relatively smaller than the other products. An important factor influencing the properties of CAPB in an aqueous solution is its strong interaction with anions [43]. Zwitterionic surfactants, comprising positive and negative groups in their headgroups, are essentially electro-neutral as monomers, but their micelles preferentially uptake anions similar to cationic surfactant micelles [44].
Nanoparticles are tiny microscopic particles with at least one dimension in the nanometer scale (usually 100 nm or less) [45]. In emulsion with micelles formed when the surfactant concentration was over the CMCs, the controlled precipitation in a restricted space of nanostructured templates, a significant refined nanorod apatite could be achieved by restricting crystal growth.

Quantitative test of antibacterial ability
The quantitative results of the bacteriostatic effect of each surfactant against S. aureus and E. coli are illustrated in Figure 5 to achieve an enhanced understanding of the effect of surfactant residues on the antibacterial properties of the resultant nanorod apatite. The sterilization of the groups was confirmed by comparing their performance with that of the negative control group. The OD values of CTAB, SDS, and CAPB were similar to that of the positive group of dimethyl sulfoxide. Thus, these surfactants at their CMCs demonstrated obvious antibacterial activity against S. aureus and E. coli. The OD value of Pluronic F-127 was higher than that of the other groups but remained lower than that of the negative control group. Thus, this surfactant may be concluded to have moderate antibacterial activity compared with the other surfactants.
The antibacterial properties of the synthesized nanorod apatite against S. aureus and E. coli are shown in Figure 6. Each group of nanorod apatite showed antibacterial activity against S. aureus and E. coli on day 1, and quantitative measurements of S. aureus indicated that groups synthesized with surfactant templates have greater antibacterial properties than those synthesized without templates (Figures 2  and 6a). However, except for the T-nHA group, all of the apatite samples demonstrated relatively shortlived antibacterial ability. Indeed, most of the groups lost their antibacterial ability as the incubation time increased. The antibacterial ability of the nanorod apatite against E. coli was measured (Figure 6b), and the T-nHA and P-nHA groups showed obvious antibacterial effects compared with the nHA group (p < 0.05) on day 1. However, on days 2 and 3, only the T-nHA group continued to demonstrate antibacterial activity against S. aureus and E. coli. The antibacterial ability of the T-nHA group was observed until day 14 of culture ( Figure 6c). T-nHA revealed stable antibacterial effects against S. aureus up to day 14. By comparison, the antibacterial effect of this group against E. coli was Table 2. Average length, width, and length-to-width aspect ratio of nanorod apatite prepared using different surfactants under hydrothermal conditions (n = 40). initially strong and then declined with increasing contact time until day 14 (n = 3). This finding illustrated that the inhibitory ability of the nanorod apatite may be lost with increasing contact time. Comparison of Figure 6c with Figure 6a and 6b indicated that the nanorods may only temporarily inhibit the proliferation of bacteria. The surface effects of the nanorods may increasingly influence their antibacterial effects over time. The mechanisms through which the different samples exert their antibacterial activity are complex and may include the interplay of the nanosize, shape, length-to-width aspect ratio, and the ions or molecules attracted or bonded to the surface charge characteristics of apatites [46]. Nanorod apatite contains cations of Ca 2+ site and anions of PO 4 3in its structure, and it has high affinity for organic molecules (Figure 1b). When considering the obvious residual effects of cationic/CTAB and NH 4 + cations on the surfaces of nanorod apatites (Figure 2), the appearance of the antibacterial activity of cationic/CTAB in T-nHA was higher than the cation of NH 4 + in S-nHA, F-nHA and P-nHA due to that CTAB has the ability to inducing superoxide stress in bacteria and led bacterial cells to a generation state of superoxide and hydrogen peroxide and therefore play a key action of antibacterial surfactant [38,47,48].

Cytotoxicity quantitative and qualitative testing
The control group exhibited 100% cell viability after 24 h of exposure to the apatite extract (Figure 7a). In accordance with ISO 10,993-5: 2009(E), cytotoxicity was considered a cytotoxic effect due the reduction in cell viability by more than 30% of L929 cells after 24 h in comparison with the control group [49]. Accordingly, no significant difference was observed among the negative group of high-density polyethylene (HDPE), control, and nHA/control groups (p > 0.05). Furthermore, only the F-nHA group showed no toxic effect on L929 cells. The cell viability of this group was significantly different compared with that of the T-nHA, S-nHA, and P-nHA groups (p < 0.05). This result indicated that F-nHA does not remarkably inhibit cell growth. As shown in Figure 5, Pluronic F-127 showed weaker antibacterial effects and less cytotoxicity than the other surfactants. Figure 7b consistently indicates that L929 cells have a spindle-like morphology. Thus, the cells attached to the samples of the negative control, HDPE, nHA/control, and F-nHA groups showed normal proliferation and no signs of cytotoxicity. The specific influence of various apatite shapes on biocompatibility is unclear. Precedent studies demonstrated the shape-dependent effects of nHA (needles, plates, spheres, and rods) on cytotoxicity [50]. They revealed that needle-and plate-shaped nHA induce the most significant cell-specific cytotoxicity because these nanorod apatites could enhance the production of reactive oxygen species and promote DNA damage measured by γ-H2AX phosphorylation [17]. In the present study, the cell morphology of the F-nHA group was similar to that of the nHA/control group; thus, nanorods synthesized with F-nHA as the surfactant template were not cytotoxic to L929 cells. The results generally indicated that residual surfactants may remain on the nanorod apatite despite rinsing with a large amount of deionized water following synthesis. This factor, in combination with the nanorod apatite morphology, may contribute to the cytotoxicity of the synthesized materials.

Conclusion
Nanorod apatites successfully prepared using a simple technique through precipitation using surfactants at their CMCs as synthetic templates under hydrothermal conditions were investigated. The use of different template groups resulted in differences in length-to-width Figure 7. Viability of L929 cells exposed to different extracts of apatite prepared using different surfactants via hydrothermal synthesis for 24 h: (a) Quantitative (n = 3) and (b) qualitative results. The extract of high-density polyethylene (HDPE) was added for negative control, and 10% dimethyl sulfoxide (DMSO) was added for positive control. aspect ratio, antibacterial ability, and biocompatibility. The P-nHA group exhibited the smallest average length and the diameter of homogeneous nanorod apatite, suggesting that the binding of large amounts of residual NH 4 + on the apatite surface to the sites of negatively charged PO 4 3− ions restricts nanorod apatite growth.
Quantitative measurements of the bacteriostatic test demonstrated the combined antibacterial effects of the surfactants and residual cations against S. aureus and E. coli. Only the nHA/control and F-nHA groups showed no cytotoxicity toward L929 cells. Among the samples synthesized, the nanorod apatite of F-nHA obtained with nonionic surfactant as the template showed a short-term antibacterial effect on S. aureus on day 1 and no cytotoxicity. Taken together the results demonstrate a high antimicrobial activity of T-nHA, however, it could be used alone in medical treatment due to the induced cytotoxicity. Therefore, if the focus is on nanorod apatite with good antibacterial effect of T-nHA synthesized in the presence of cationic/CTAB, it must be compounded with other bone repair materials to control the release and thus reduce its cytotoxicity. These samples may also represent a candidate material that could be used in bone restorative applications in the future.