Kirsten M. Pondmana, A. Wouter Maijenburgb, F. Burcu Celikkolc, Ansar A. Pathand, Uday Kishored, Bennie ten Hakena and Johan E. ten Elshof*b
aNeuro Imaging, MIRA Institute, University of Twente, Enschede, The Netherlands
bInorganic Material Science, MESA+ Institute for nanotechnology, University of Twente, Enschede, The Netherlands. E-mail: j.e.tenelshof@utwente.nl
cNanobiophysics, MIRA Institute, University of Twente, Enschede, The Netherlands
dCentre for Infection, Immunity and Disease Mechanisms, Biosciences, Brunel University, London, UK
First published on 19th September 2013
Due to their shape anisotropy, high aspect ratio magnetic nanoparticles offer many advantages in biomedical applications. For biocompatibility, it is essential to have full control over the dimensions and surface chemistry of the particles. The aim of this study was to synthesize biocompatible nanowires with tuneable dimensions. This was achieved by electrodeposition of Ni in polycarbonate membranes. To ensure biocompatibility, a continuous gold coating was deposited onto the Ni wires by a newly developed electroless deposition method. The coating was analysed using electron microscopy and X-ray diffraction. Magnetic properties, anisotropy and Au film thickness were studied using vibrating sample magnetometry. After biofunctionalization, no significant cytotoxic effects were found in studies involving a diverse range of primary and tumour cells exposed to increasing concentrations of nanowires for up to 7 days. These nanowires may thus be used for in vivo applications such as magnetic drug delivery.
Most research on elongated nanoparticles was performed with carbon nanotubes. This research showed that elongated particles are capable of entering cells by penetration, allowing for direct access to the cytosol without the need for endosome escape strategies.7,14,15
Several chemical methods to synthesize elongated nanoparticles can be found in the literature (Fig. 1).16
![]() | ||
Fig. 1 (A) Size reduction of a 1D-microstructure; (B) (self)assembly of spherical nanostructures; (C) spontaneous synthesis by dictation by the anisotropic crystallographic structure of a solid; (D) kinetic control by a capping reagent; (E) confinement by a liquid droplet in a vapor–liquid–solid procedure; (F) direction through the use of a template. |
Templated electrodeposition has several advantages in this application, including ease of fabrication, reproducibility, and low cost. Nanowires made by this method are always continuous since any breakage would halt the growth of the nanowires. Depending on the template and deposition time used, nanowires can be made with nanometer to micrometer dimensions, spanning several relevant biological length scales. Various magnetic metals such as iron (Fe), cobalt (Co) and nickel (Ni), and their alloys have been used to form nanowires inside the pores of silica, anodized alumina, or polycarbonate membranes.17–20
Having full control over the dimensions of nanowires is essential for biomedical applications. Previous research on asbestos and carbon nanotubes has suggested that cytotoxicity is highly dependent on aspect ratio. Long nanowires have been associated with inflammation, fibrosis and malignant mesothelioma and are likely to generate reactive oxygen species that can cause DNA damage and cell death.21 Unfortunately, most proposed nanowires8,9,22 are larger than cells, leaving the wires to protrude from the cells, thereby reducing the viability of the cells. Particles longer than 5 μm cannot be considered as “safe” in terms of uptake and clearance by macrophages and possibly induce inflammation and carcinogenic effects.23 A length below 5 μm will also make it possible for the nanowires to flow through the curvy capillaries allowing for local drug delivery.
In addition to dimensional restrictions, the materials and surface chemistry of the particles should be biocompatible and stable during use. Fe is often considered the most advantageous option, but due to the ease of oxidation of Fe, Ni is often preferred as a magnetic phase for nanowires. Although Ni is a naturally occurring transition material and an essential nutrient,24 there are indications that Ni may be a carcinogen and induce apoptosis.25–27 Studies on the toxicity of Ni wires are divided; some studies indicate no toxicity of Ni nanowires at low concentrations, and the need for long term exposure to act as carcinogens.4,28 But at higher concentration, even short incubation times have been shown to be lethal to cells.8 Some authors indicate the benefits of induced apoptosis by Ni nanowires, e.g. reactive oxygen species mediated apoptosis caused by Ni nanowires in human pancreatic adenocarcinoma cells.29 Unfortunately, these apoptotic effects are not specific to carcinoma cells and would therefore also occur in other healthy bystander cells. Shielding of the potentially toxic Ni from contact with the body was therefore considered essential in this study. The coating is also important to protect the Ni from oxidation. For this purpose, various polymers, proteins and phospholipids have been proposed.22,30,31 These coatings are biocompatible but cannot shield the particles from oxygen. Therefore, a solid Au shell is considered most advantageous as Au is known for its excellent biocompatibility and it cannot be removed by factors within the body.32 Furthermore, Au provides a platform for functionalization, such as conjugation of biomolecules, and targeting moieties and drugs.33,34
This work aims at the development of a magnetic nanowire for application in magnetic drug delivery and similar biomedical applications. The proposed electrodeposited Ni nanowire has a tuneable length below 5 μm with narrow size distribution. To increase the biocompatibility, the nanowires were coated with an Au shell. To limit toxic effects, the use of cyanide, which is commonly used in Au coatings,32 was avoided. Therefore, a new coating procedure was developed, inspired by a method to synthesize Au nanotubes by electroless deposition.35 The nanowires were characterized using a variety of techniques in order to determine their dimensions, magnetic properties, crystallographic structure and biocompatibility.
After deposition, the membranes were dissolved in dichloromethane and washed at least 3 times with dichloromethane and 3 times with MilliQ water by centrifugation.
![]() | ||
Fig. 2 Scheme showing stepwise Au coating on Ni nanowires. (1) Amination, (2) Sn2+ deposition, (3) Ag deposition, (4) replacement of Ag by Au and (5) growth of the Au layer. |
Clean Ni nanowires were dispersed by sonication in an aqueous solution of 10 mM 11-aminoundecanoic acid for 2 h, to form an aminated monolayer (step 1 in Fig. 2). The nanowires were then washed at least 3 times with H2O.
A fresh aqueous solution containing 54 mM SnCl2 and 70 mM trifluoroacetic acid was added to the nanowires for 45 min. Sn(II) ions interacted with the electronegative amine groups forming a layer of electrostatically bound Sn2+ on the surface (step 2 in Fig. 2). After incubation, the nanowires were washed repeatedly with H2O.
Sn2+ was oxidized by Ag+via Sn2+ + 2Ag+ → Sn4+ + 2Ag0, by adding an aqueous solution containing 29 mM AgNO3 and 100 mM NH4OH (step 3 in Fig. 2). After repeated washing with H2O, the Ag phase can be replaced by Au.
A solution of 1 g Na2SO3, 0.6 g NaHCO3, 6 mL Na3Au(SO3)2 (Oromerse B) and 480 mL H2O was prepared. The nanowires were placed in 10 mL aliquots of this solution and sonicated at 3.6 °C for 30 min.
To allow for the formation of a closed layer of Au on the Ni nanowires, 25 μL mL−1 formaldehyde was added to the solution. This destabilized the Au+–sulfite complex, and induced the galvanic replacement reaction Ag0 + Au+ → Ag+ + Au0 (steps 4 and 5 in Fig. 2), and further electroless growth of the Au layer. The nanowires were sonicated for 12 h at 3.6 °C, and washed at least 5 times with H2O to remove free ions. Nanowires prepared using this method are named Au–Ni nanowires.
Magnetic characterization was performed using a vibrating sample magnetometer (Model 10 Mark II VSM, Microsense). Samples were dried in a glass container before measurement; a blank measurement was subtracted from the sample data. Dispersed samples were measured in a glass container closed with Parafilm, and a blank containing the liquid in which the dispersion was made was subtracted from the data.
Human monocytes and lymphocytes, and PBMC were isolated from blood samples from 2 healthy subjects using Ficol-Paque plus (GE Healthcare) gradient according to the manufacturer's instructions. PBMC obtained from the interface of plasma and Ficoll-Paque were washed twice with RPMI and suspended in complete RPMI. 5 × 106 PBMC in complete RPMI were plated out in each well of a 6 well plate (SPL Life Sciences), and pre-coated with human AB serum overnight. After 2 h of incubation at 37 °C in 5% CO2, the non-adherent PBMC (containing T, B and NK cells) were gently removed from adherent cells (monocytes).
After 48 h the cell medium was carefully removed and the cells were washed with PBS, thereafter 100 μL of cell medium was added with 10 μL of CellTiter-Blue and incubated for 4 h. The fluorescence was measured on a spectrophotometer (Victor, Perkin Elmer) with an excitation wavelength of 560 nm and an emission wavelength of 590 nm. The fluorescence values were normalized by the controls and expressed as percent viability. Validity of the assay was assessed by cell counting and visual inspection of the cells.
![]() | ||
Fig. 3 HR-SEM images of Ni nanowires and Au–Ni nanowires. (a) Original Ni nanowires, (b) Oromerse-B Ni nanowires, (c) Sn–Ag–Au–Ni nanowires, and (d) 11-aminoundecanoic acid–Sn–Ag–Au nanowires, hereafter indicated as Au–Ni nanowires. |
Employing only the Au deposition reaction with Oromerse B (steps 4 and 5 in Fig. 2) did not result in Au nucleation on the Ni nanowire surface as can be seen in Fig. 3b. Therefore, an activation step was introduced in which Sn nuclei were deposited, followed by galvanic replacement by Ag, and finally growth of Au on the Ag nuclei. Unfortunately, the Au coating was still discontinuous (Fig. 3c), most likely due to an insufficient concentration of Sn nuclei on the Ni surface. An amination step, using 11-aminoundecanoic acid, was introduced to provide more homogeneous Sn deposition, allowing a continuous Au coating to form over the entire nanowire surface (Fig. 3d), with an approximate thickness of 40 nm as estimated from SEM. Longer Au deposition will result in the formation of a thicker Au layer.
The phase and grain size of the Ni nanowires and Au coating were investigated by XRD (Fig. 4). Ni nanowires showed diffraction peaks corresponding to the (111) and (200) planes of cubic Ni. The peak at 2θ 38° can be assigned to the Au (111) plane of the back electrode. After the complete Au coating procedure diffraction peaks were detected, which can be assigned to Au (110), (111), (200), (220) and (311). The Ni (111) and Au (200) peaks overlap, therefore it is not possible to distinguish between them. The broader peaks in the Au spectrum compared to the Ni spectrum indicate a smaller crystallite size for the Au layer. Using the Debye–Scherrer equation τ = Kλ/βcosθ, the crystallite sizes of Ni and Au can be estimated. Here, τ is the mean size of the crystalline domain, K the shape factor, β the full width at half maximum, θ the Bragg angle and λ the X-ray wavelength. The estimated crystallite sizes are 54 nm for Ni and 20 nm for Au. By comparing the crystallite sizes from XRD with SEM imaging, we can conclude that the crystalline domain size in the central Ni nanowire is about 1/3 of its diameter, and it is surrounded by an Au layer with an equivalent thickness of about 2 grains.
![]() | ||
Fig. 4 XRD patterns of randomly oriented Ni nanowires and Au–Ni nanowires and a pure Ni reference sample. |
Fig. 5 shows the room temperature magnetization curves for uncoated Ni nanowires inside the template with magnetic field (H) applied in two directions, i.e. parallel and perpendicular to the wire direction. The squareness ratio Mr/Ms and coercivity are known to increase significantly with aspect ratio.36 The obtained hysteresis loops show magnetic anisotropy where the magnetic easy direction is along the axis, as can be seen from the squareness ratio Mr/Ms = 0.48 in the parallel direction. From VSM measurements on randomly orientated nanowires as depicted in Fig. 8 the saturation magnetization of the Ni nanowires was found to be 34 emu g−1, which is significantly less than the value for bulk Ni (55.4 emu g−1). This can be explained by the surface anisotropy which is significant for a wire with small diameter. After Au coating of the Ni nanowires the saturation magnetization decreased to 10 emu g−1 (Fig. 6). From this value, the diameter and length of the Ni nanowire and the densities of Ni (8.9 × 106 g m−3) and Au (19.3 × 106 g m−3), the shell thickness can be calculated. Neglecting the influence of the other coating layers the Au layer thickness is approximately 34 nm, which is in good agreement with the thickness estimated from HR-SEM images.
![]() | ||
Fig. 5 Magnetic hysteresis curves of Ni nanowires measured inside the membrane under an external field parallel and perpendicular to the nanowire long axis. Parallel coercivity value of 302 Oe and Mr/Ms = 0.48 and perpendicular coercivity of 378 Oe and Mr/Ms = 0.17. |
![]() | ||
Fig. 6 Hysteresis curves of Ni nanowires and Au–Ni nanowires in random orientation under an external field. Ms = 34 emu g−1 for uncoated and Ms = 10 emu g−1 for Au coated Ni nanowires, suggesting a coating thickness of 34 nm on a 2.6 μm long Ni nanowire with 150 nm diameter. |
After biofunctionalisation using SH-PEG and glutathione, a suspension of Au–Ni nanowires was obtained (Fig. 7). The nanowires could be easily collected using a permanent magnet and dispersed by vortexing. Due to the possibility to move freely in the liquid, the magnetization curves of suspended nanowires showed superparamagnetic behaviour, with no remnant magnetisation (Fig. 8). Attachment of GSH and mPEG-SH was examined by FTIR spectroscopy (Fig. 9). GSH attachment can be identified by two bands around 1600 cm−1 for both cysteine and glutamic acid-carbonyl. The anchoring is via the thiol group as the band at 2500 cm−1 for SH stretching is not visible on the Au–Ni nanowires. Binding of mPEG-SH can be seen by a strong band at 1380 cm−1 showing C–H stretching, and the band at 1080 cm−1 shows C–O–C stretching from mPEG-SH.
![]() | ||
Fig. 7 (a) Suspension of Au–Ni nanowires stabilised with GSH and mPEG-SH. (b and c) The Au–Ni nanowires can be collected using a permanent magnet. |
![]() | ||
Fig. 8 Magnetisation loop M vs. applied field H for suspended Au–Ni nanowires showing superparamagnetic behaviour. The inset shows that the coercivity of the Au–Ni nanowires is approximately 1 kA m−1 for dispersed samples. |
![]() | ||
Fig. 9 FTIR spectra of GSH, mPEG-SH, and uncoated Au–Ni nanowires, and the attachment of GSH and mPEG-SH to the Au–Ni nanowires. |
The uptake of Au–Ni nanowires was studied by confocal microscopy (Fig. 10). 3D imaging (not shown here) ascertained that the particles were inside the cells and not simply adhered to the membrane. After 24 h of incubation, RAW 264.7 cells had taken up many nanowires and clusters of nanowires. As these cells are macrophages, the uptake pathway is most likely phagocytosis. Confocal microscopy analysis showed that the nanowires were completely taken up by these cells without signs of frustrated phagocytosis. This indicates that macrophages will be able to remove the particles from the bloodstream. Jurkat cells are not capable of phagocytosis and only take up particles by receptor mediated endocytosis. As expected, these cells did not take up any nanowires, and appeared as healthy as control untreated cells (see ESI†).
![]() | ||
Fig. 10 Interaction of Au–Ni nanowires with (a) RAW 264.7 cells and (b) Jurkat cells. |
The effects of Au–Ni nanowires on cell viability were evaluated using a CTB assay. Au–Ni nanowires did not show a decrease in cell viability of both HeLa and RAW cells over 48 h (Fig. 11). In a study between phagocytic and non-phagocytic cells, a slightly higher cytotoxicity was observed for phagocytic (RAW 264.7) cells compared to non-phagocytic (HeLa) cells, which might indicate the importance of uptake on cytotoxicity. It should be noted that the high concentration of 1000 nanowires per cell is much higher than targeted in potential applications in biomedicine, and is only used here to investigate the biocompatibility of the nanowires. In RAW cells Ni nanowires showed increased cytotoxicity compared to the Au coated Ni nanowires, indicating the necessity of the Au coating. The used concentration of nanowires is much higher than the lethal concentration of Ni nanowires with a length of 20 μm found by Byrne et al.,8 which indicates and illustrates the high biocompatibility of the much shorter Ni and Au–Ni nanowires.
![]() | ||
Fig. 11 Short term viability compared to untreated control of RAW and HeLa cells assessed by CTB assay after 48 h incubation with Ni and Au–Ni nanowires. The bars indicate standard deviation. |
A range of cells – primary and cancerous, phagocytic and non-phagocytic, adherent and non-adherent – were used in order to systematically investigate cytotoxicity over a long term (Table 1). All 9 different cell types used in this study did not show significant cytotoxicity of the Au–Ni nanowires for periods up to 7 days.
Cell type | 12 h | 24 h | 48 h | 5 days | 7 days | |
---|---|---|---|---|---|---|
U937 (non-adherent) | Control | 97.6 ± 1.0 | 96.3 ± 0.6 | 96.9 ± 0.4 | 98.1 ± 0.4 | 97.2 ± 0.1 |
Au–Ni wires | 96.5 ± 0.7 | 96.5 ± 0.7 | 96.7 ± 0.6 | 98.7 ± 1.4 | 98.2 ± 1.2 | |
THP1 (non-adherent) | Control | 99.7 ± 0.5 | 95.7 ± 0.1 | 98.0 ± 1.5 | 99.0 ± 0.3 | 98.5 ± 0.4 |
Au–Ni wires | 98.5 ± 0.5 | 93.7 ± 0.1 | 95.5 ± 1.8 | 99.0 ± 0.4 | 99.3 ± 0.3 | |
A549 (adherent) | Control | 95.3 ± 1.4 | 98.1 ± 0.6 | 92.9 ± 0.3 | 98.5 ± 1.4 | 97.0 ± 0.9 |
Au–Ni wires | 98.8 ± 0.6 | 98.6 ± 0.2 | 94.2 ± 1.4 | 98.5 ± 1.3 | 96.6 ± 0.7 | |
Raji (non-adherent) | Control | 97.7 ± 0.3 | 98.2 ± 0.4 | 98.7 ± 0.1 | 99.5 ± 0.1 | 96.6 ± 0.6 |
Au–Ni wires | 97.9 ± 0.1 | 97.7 ± 0.4 | 98.6 ± 0.7 | 98.4 ± 0.4 | 95.4 ± 2.9 | |
GBM (adherent) | Control | 98.7 ± 0.1 | 97.8 ± 3.1 | 99.1 ± 0.5 | 96.0 ± 0.7 | 96.3 ± 2.5 |
Au–Ni wires | 98.9 ± 0.2 | 98.4 ± 0.0 | 98.7 ± 0.7 | 96.1 ± 0.7 | 97.1 ± 0.9 | |
Jurkat (non-adherent) | Control | 98.4 ± 0.3 | 96.9 ± 0.6 | 97.8 ± 0.7 | 94.6 ± 0.8 | 93.6 ± 1.7 |
Au–Ni wires | 97.8 ± 0.5 | 97.9 ± 0.2 | 98.1 ± 1.0 | 94.6 ± 0.1 | 92.0 ± 1.0 | |
PBMC (adherent) | Control | 99.5 ± 0.2 | 98.4 ± 0.6 | 99.5 ± 0.4 | 98.4 ± 1.0 | 98.4 ± 0.5 |
Au–Ni wires | 99.8 ± 0.2 | 99.3 ± 0.5 | 99.5 ± 0.4 | 98.4 ± 0.9 | 98.5 ± 0.4 | |
PBMC (non-adherent) | Control | 99.5 ± 0.4 | 97.8 ± 0.9 | 99.8 ± 0.3 | 96.3 ± 1.3 | 98.2 ± 0.9 |
Au–Ni wires | 99.6 ± 0.4 | 98.5 ± 0.8 | 99.7 ± 0.2 | 96.1 ± 1.4 | 97.7 ± 0.6 | |
A431 (adherent) | Control | 98.7 ± 0.1 | 98.6 ± 0.1 | 96.2 ± 1.9 | 96.1 ± 3.8 | 96.1 ± 3.8 |
Au–Ni wires | 98.9 ± 0.2 | 98.2 ± 1.1 | 95.6 ± 0.2 | 95.2 ± 1.7 | 95.2 ± 1.7 |
Footnote |
† Electronic supplementary information (ESI) available: Control cell microscopy images of RAW 264.7 and Jurkat cells. See DOI: 10.1039/c3tb20808g |
This journal is © The Royal Society of Chemistry 2013 |