Article pubs.acs.org/Langmuir
Magnetically Self-Assembled Colloidal Three-Dimensional Structures as Cell Growth Scaffold Gašper Kokot,*,†,‡ Špela Zemljič Jokhadar,§ Urška Batista,§ and Dušan Babič∥ †
Jožef Stefan Institute, Jamova cesta 39, SI-1000 Ljubljana, Slovenia Faculty of Medicine, Institute of Biophysics, University of Ljubljana, Vrazov trg 2, SI-1000 Ljubljana, Slovenia ∥ Faculty of Mathematics and Physics, University of Ljubljana, Jadranska 19, SI-1000 Ljubljana, Slovenia §
S Supporting Information *
ABSTRACT: Understanding the chemical and physical conditions for cell growth is important from biological and medical aspects. Many tissues and cell types (e.g., epithelial cells and neurons) naturally grow on surfaces that span in three-dimensions and offer structural or mechanical support. The scaffold surface has to promote adhesion and cell proliferation as well as support their weight and retain its structural integrity. Here, we present a flexible method that uses self-assembly of micrometer superparamagnetic particles to produce appropriate scaffold surfaces with controllable general appearance in three dimensions, such as oriented membranes, branched structure, or void network. As a proof of principle, the Chinese hamster ovary epithelial cell line was successfully grown for several days on inclined membranes. Robustness of the oriented membrane architecture was probed with optical tweezers. We measured the magnetic force holding one particle in a self-assembled upright hexagonal sheet and modeled it as a sum of pair interaction forces between spatially arrested static dipoles.
■
to the applied field when crossing ϑo = 54.7°. Remarkably, precessing the magnetic field back and forth at this magic angle can lead to spontaneous assembly of robust self-repairing membranes, as a result of many body effects in dense particle solutions.16 We used the same magnetic field actuation principles to obtain a variety of stable architectures. They can be enlarged during the experiment and are appropriate for 3D cell growth scaffolds, as we show by growing a Chinese hamster ovary (CHO) epithelial cell line on them for several days without any influence on the integrity of the colloidal scaffold. Tunable forces holding the scaffolds together were probed with optical tweezers on vertical sheets. The principle of optical trapping of a dielectric particle with a focused laser beam is widely used in fields of microrheology17 and molecular biology.18 Forces that can be exerted can reach up to 100 pN and can be measured with an appropriate selection of the particle and calibration method.19 Comparing our force measurements to a calculated peak force on a sample bead removed from a static hexagonal sheet yielded expected dependence upon the magnetic field.
INTRODUCTION Organ or tissue failure is a common clinical condition. Even when organ donation is a possibility, tissue matching a donor and a recipient is a toilsome service. A way to avoid this bottleneck of modern medicine capabilities is to engineer tissues and organs for implantation. Techniques that aim to permanently or temporarily replace the extracellular matrix include xenogeneic extracellular matrix,1 polymers in the form of electrospun nanofibers2−4 or hydrogel,5 carbon nanotubes,6 and three-dimensional (3D) printing.7−9 The method of choice depends upon the natural function of the tissue in question and the level of degeneration.10 When rigid scaffolds are fabricated, control over the architecture parameters, such as porosity or density, is possible; they can be coated with various molecules to address cell adhesion or biological response, and different mechanical properties can be achieved by the choice of biocompatible material. The method that we present here includes the above considerations and offers additional dynamical control over the forces holding the scaffold surfaces together and, thus, changing its mechanical properties, which are an important factor influencing cell growth11 and, in some cases, even drug treatment response.12 The scaffolds are composed of micrometer superparamagnetic beads. In the presence of an external magnetic field, a magnetic dipole is induced in the beads.13 If the external magnetic field is precessing with an opening angle ϑo, particle dipoles dynamically assemble into rotating pairs14,15 and clusters,15 which show phase transitions in their response © 2015 American Chemical Society
■
EXPERIMENTAL SECTION
Imaging was performed with either a complementary metal oxide semiconductor (CMOS) camera (Pixelink, PL-A741) or electron Received: June 14, 2015 Revised: July 30, 2015 Published: August 13, 2015 9576
DOI: 10.1021/acs.langmuir.5b02176 Langmuir 2015, 31, 9576−9581
Article
Langmuir multiplying charge-coupled device (EM CCD) camera (Hamamatsu Photonics, C9100-13) for fluorescence visualization. The wall of the growth chamber for experiments in an external magnetic field was made from biocompatible polymer polydimethylsiloxane (PDMS, Sylgard 184 silicone elastomer kit, Dow Corning Corporation). The polymer was molded to a desired shape. Two glass plates coated with indium tin oxide (ITO) served as heating plates, and a resistor was placed inside of the chamber to probe the temperature. They were all connected to a temperature controller (Oxford Instruments ITC 503S, 0.1 °C precision) to keep the temperature constant at 37 °C. The scaffolding was composed of superparamagnetic beads (1.05 μm diameter, Dynabeads MyOne carboxylic acid, Life Technologies) in a solution. The self-assembly of robust planar configurations spanning in three dimensions was provoked by external magnetic field modulation. It was cycled clockwise and counterclockwise in a tilted inverted cone fashion ⃗ = B0 R y(ϑt)(sin ϑo cos(ωt ), β sin ϑo sin(ωt ), cos ϑo)T Bext
F⃗ =
i
(1)
(2)
with β = 0.34 and ω = 600π s−1 fixed. Brownian fluctuation of the probe silica sphere (9 μm diameter, Bangs Laboratories) in the optical trap was first used to establish the trapping constant by observing the Brownian motion. Then, it was pushed into the self-assembled sheet until a single superparamagnetic bead adhered to its surface. The probe bead was pulled away in discrete steps, and force was measured each time by time-average displacement from the center of the trapping potential. The peak force right before the superparamagnetic bead was pulled out when compared to a simple model. For each magnetic field, the measurement was repeated on different samples at least 10 times in total; the mean values and standard deviations were obtained by fitting a Gaussian distribution over the corresponding histogram. In our calculations, we modeled the sheet as an infinite half-plane with point dipoles fixed in a hexagonal arrangement. Magnetic moments m⃗ were considered to be induced by an external magnetic field
m⃗ =
χV ⃗ Bext μ0
3μ0 ⎛ 5(mr⃗ i ⃗)(m⃗ i ri )⃗ ⎞ ⎜(m⃗ i ri )⃗ m⃗ + (mr⃗ i )⃗ m⃗ i + (m⃗ im⃗ ) ri ⃗ − ri ⃗⎟ 5 4πri ⎝ ri 2 ⎠ (4)
The resulting force was pointing perpendicularly toward the lattice, and the distance dependence yielded a peak as observed in the experiment (see the Supporting Information). To test the suitability of superparamagnetic scaffolds, Chinese hamster fibroblast cell line (V79) and CHO epithelial cell line were cultivated in Advanced minimum essential medium (MEM) (Gibco) supplemented with antibiotics (penicillin−streptomycin, Gibco), 5% fetal bovine serum (FBS) (Gibco), and L-glutamine (2 mmol/L) at 37 °C in a CO2 incubator (5% CO2 and 95% air, with 95% relative humidity). The cells were detached with TrypLE Express (Gibco). We performed two control experiments for cell growth: first in the presence of the modulated external magnetic field without superparamagnetic beads and second on the dried densely packed colloidal surface without a magnetic field. For both cases, V79 cells were used. After the cells were detached, they were seeded in cultivation chambers in Leibovitz’s L-15 medium (Gibco) supplemented with 10% FBS. To test the cell viability after a given time, calcein AM (Invitrogen) was added to the cells. Photoluminescent calcein activates when inside of the living cells, and they start to fluoresce. Fibronectin Coating. The beads in the manufacturer solution were disinfected with 10 min microwave treatment at a maximum power (Candy, CMW, 1770 M; rated output, 700 W; and frequency, 2450 MHz). The storage solution of the beads was subtracted, and a fibronectin (Sigma) solution (2 μg/mL) was added to the beads. Incubation lasted 45 min with alternation between shaking and sonication every 5 min. When finished, we exchanged the fibronectin solution with water and added the beads in the growth chamber.
where B0 is the amplitude of the magnetic field, Ry(ϑt) is the rotation matrix around the y axis for a tilt angle ϑt, ϑo is the semi-cone opening angle, ω is the angular frequency, and β is the eccentricity in the xy plane. The parameters used in experiments were ϑt = 20°, ϑo = 0° and 53°, ω = 600π s−1, B0 = 10 mT, and β = 0.43. For force measurements (Figure 3B), we used optical tweezers (Aresis Tweez 250si system). The magnetic field modulation was in a simpler form ⃗ = B0 (0, β sin(ωt ), 1) Bext
∑
■
RESULTS Scaffold Architectures and Force Measurements. The external magnetic field was created with six pairwise perpendicular semi-Helmholtz coils. The coil assembly allows for arbitrary time modulation of orthogonal magnetic field vector components. For our purpose, we used oscillating fields in the xy plane parallel to the plane of observation and a constant field in the z direction perpendicular to it (eq 1). To obtain a time-average interaction that leads to selfassembly of stable colloidal membranes, the magnetic field was cycled clockwise and counterclockwise in a semi-cone fashion (Figure 1) with a magic opening angle of ϑo = 54.7°.16 With
(3)
where V is the volume of the bead determined by its radius, μ0 is the vacuum permeability, and χ is the shape-dependent magnetic susceptibility of a sphere [3(χm − χs)/(3 + χm + 2χs), with χm and χs being the magnetic susceptibilities of the material and the solvent, respectively20] in water (χs = −9 × 10−6 21). The magnetization curve to determine χm at a given B0 was obtained from the manufacturer. In the model, the vertical sheet was considered to be a flat half-plane and the dipoles at positions corresponding to centers of 1 μm spheres in a close-packed hexagonal arrangement (panels A and B of Figure 4). One magnetic dipole from the tenth row was displaced in the direction of the vertical sheet normal, while the others stayed fixed at their initial positions. All of the dipole magnetic moments were pointing in the vertical direction at all times, because they are induced by an oscillating external field, which, after time averaging, has only the vertical component. For each position of the displaced dipole m⃗ , the force acting on it was the summed pair dipolar force over all of the other m⃗ i particles at a distance ri⃗ from m⃗ .
Figure 1. Sketch of magnetic field modulation. The xy plane is in parallel with the microscopy glass. The parameters describing the movement of the magnetic field direction are the angular velocity ω = dφ/dt, the semi-cone opening angle ϑo, and the tilt angle ϑt. 9577
DOI: 10.1021/acs.langmuir.5b02176 Langmuir 2015, 31, 9576−9581
Article
Langmuir
Figure 2. Structures with different general appearances may be achieved with an appropriate approach. In all cases, ϑo was about 53°, ϑt was 20°, and the focal plane z ≈ 50 μm above the glass surface. The differences in preparation of scaffolds were (A) void network, initial distribution of rotating vortices; (B) branched structure, adding the particles while the magnetic field is running; and (C) sheets, homogeneous initial distribution, offset in one horizontal direction.
Figure 3. (A) Comparison of measured peak forces (black dots; error bars are standard deviations obtained from Gaussian distribution fits to the histograms) and calculations from the model (dashed line). Quadratic magnetic field dependence of force was confirmed by a fit of AB0δ (solid line) that yielded δ = 1.9 ± 0.5 (error is the standard deviation). The value of A = 63 ± 5 μN/T2 (at δ = 2) translates to χ = 0.42 ± 0.2 for our beads. In the model, the magnetic moment linearly proportional to the magnetic field with a uniform value of susceptibility was corrected for magnetization nonlinearity by taking the value of the susceptibility from the magnetization curve at each magnetic field (dashed line). (B) Example of the force measurement. There was always some drift of the sheet present as we pulled; therefore, the step coordinate is used just to mark consecutive measurements. The optical trap was moved in discrete steps of 0.1 μm, each time that the force was measured. A force peak value right before the superparamagnetic bead was completely pulled out was compared to the model. (C) Series of images representative for a measurement of the force needed to pull one superparamagnetic bead out of the colloidal sheet taken at three steps corresponding to coordinates 2, 50, and 56 in panel B. A much larger silica bead (d = 9 μm, white circle with a dark halo from light scattering at the edges) serves as a force probe. It is first pushed into the vertical sheet (in the x−y projection, it is a black line as imaged on the far left of the snapshots) and then pulled away until one superparamagnetic bead (d = 1 μm) becomes pulled out and stays adhered to the probe bead by van der Waals forces (seen as a “pimple” on the probe bead and indicated by the arrow in the last snapshot).
We could vary ϑo for several degrees without any apparent influence on the shape or integrity of the structure as long as it was kept smaller than the magic angle. The tilt angle was a more influential parameter especially in the case of sheets. If it was too small, the sheets erected completely vertical, and when too large, they merged horizontally, leading to colloidal grounding instead of a 3D configuration. When measuring forces that hold the membrane sheet together, we employed the architecture that enabled us to use optical tweezers, namely, well-defined vertical sheets (eq 2). In this configuration, the optical path of the confining laser beam was unobstructed to ensure unperturbed measurements. The waist diameter at the focus was 1 μm and, therefore, deep inside the silica probe bead (d = 9 μm). By keeping close to the surface, z = 10 μm, the beam waist at the bottom glass had only 3 μm in diameter for our infrared light wavelength and, thus, did not pass through the vertical sheet. In cases where the sheet was accidentally placed in the laser beam path, it was immediately pulled to the focus and the constituting paramagnetic particles heated as a result of strong absorption. This created a series of consecutive explosions, which rendered the
modifications to this time modulation (see the Experimental Section) of the external magnetic field and different initial conditions, we were able to achieve a void network, branched structure, and oriented sheets (Figure 2). The void network (Figure 2A) was established with an opening angle of ϑo = 53°. The beads were added to the solution while the magnetic field was rotating. The resulting initial distribution was an array of rotating clusters, which were arrested when we started to rock the modulation back and forth. For the branched structure (Figure 2B), a tilt angle of ϑt = 20° and eccentricity β = 0.43 were introduced to the magnetic field time modulation; therefore, the cross-section of the semicone was a tilted ellipse instead of a circle. The beads were added to the solution, while the magnetic field was rocking back and forth. Oriented membranes (Figure 2C) had the same magnetic field modulation. The sole difference was a homogeneous distribution of beads settled on the bottom of the growth chamber before turning on the magnetic field. The result was sheets with a gradual inclination. 9578
DOI: 10.1021/acs.langmuir.5b02176 Langmuir 2015, 31, 9576−9581
Article
Langmuir measurement impossible. For this reason, we chose the vertical sheet configuration over the scaffold architectures used in the cell growth experiments, where it was important to assure a large x−y projection area to catch the cells sedimenting from top, and the vertical sheet is only a line in the x−y plane (Figure 3C).
Figure 5. Illustration of the growth chamber cross-section. The PDMS wall was sealed with two cover glasses and enclosed the growth medium with the cells, superparamagnetic beads, and temperature sensor inside. The temperature was controlled with the heating plates on top and bottom.
only (Figure 6A) and superparamagnetic beads only (Figure 6B) were carried out with V79 cells. A comparison of snapshots
Figure 4. (A) Micrograph of self-assembled sheets in the horizontal plane to allow for the inspection of the packing. It is clearly seen that the superparamagnetic beads form a two-dimensional hexagonal lattice. (B) Sketch of the pulling configuration from the simulation. The hexagonal lattice and a displaced magnetic dipole are shown. The proximity of the glass surface is taken into account by creating a halfplane and pulling a test magnetic dipole from the tenth row. (C) Example of pulling from the simulation. The force on the displaced point magnetic dipole is calculated as a sum of pair interaction forces at several distances from the sheet (eq 4). The peak is taken for comparison with the experiment.
Figure 6. Two snapshots at different times (t = 4 and 24 h) of V79 cells growing (A) in a magnetic field without colloids and (B) on a dried layer of colloids in the absence of a magnetic field. Both control experiments showed unaffected growth of the cells.
4 h and 1 day after the start of the experiment shows that neither influences the cell growth. For scaffold experiments, we used CHO cells. After adding them to the growth medium, we found them randomly distributed on the colloidal scaffold and at different heights depending upon the vertical range of the support (up to 300 μm in our experiments). In some cases, rows of cells spanned between two sections of the structure (Figure 7A) without any support for the cells in the middle. Clusters of cells sitting on top of the colloidal structure (Figure 7B) have also been noted. Collapse of the scaffold has never been observed, and the
We measured the force needed to pull one superparamagnetic bead out of the formed vertical sheet in relation to the interaction strength, which yielded the expected quadratic dependence upon magnetic field B0 (solid line in Figure 3A) and χ = 0.42 ± 0.2. This is approximately half of the value that we used in the simulations for shape-dependent χ calculated from material susceptibility. This force was compared to a calculated force from a pair interaction model of static hexagonal arrangement of point dipoles (eq 4). The dashed line in Figure 3A shows the calculation with no fit parameters. In many cases, we noticed local bending of the sheet before the pulled superparamagnetic bead detaches. The bending was less pronounced at high magnetic fields. These observations demonstrate that mechanical properties of the scaffold also change with B0. In some experiments, we were able to observe one or several force drops before the final detachment (Figure 3B), indicating structural rearrangements of the scaffold. Growing the CHO Epithelial Cell Line. The cells were grown in a custom-made PDMS chamber with temperature control (Figure 5). Control experiments with magnetic field
Figure 7. (A) Bridge of CHO cells spans between two sections of the structure without any support for the middle cells (z ≈ 50 μm). (B) Cluster of cells sitting on top of vertical sheets (z ≈ 170 μm). No collapse or deformation of the structure was observed. 9579
DOI: 10.1021/acs.langmuir.5b02176 Langmuir 2015, 31, 9576−9581
Article
Langmuir
batch differences in magnetic susceptibility of the beads. We could not perform a magnetization curve measurement as a result of sample quantity limitation. Additionally, Brownian motion provokes wobbling of the membranes at low magnetic fields, which makes the assumption of a fixed lattice unsuited. Other contributions come from the assumptions of the simulation, most significantly the pairwise summation of the dipole−dipole interaction energy, which does not take many body effects into account,16,20,22−24 and the error of this can be seen in Figure 4C. The interaction between a colloidal sheet and a single particle is repulsive in the direction of the sheet normal,16,20 when the particle is not an integral part of the sheet, while in our case, it only drops to 0 for large separations. This explains the overestimation of the peak force from our model. Nevertheless, the B02 scaling of the interaction energy remains true also for the case of dominating quadrupolar terms in balanced triaixal fields (magic angle spinning). Although dipolar and quadrupolar contributions have a different distance falloff, they are comparable when the beads are in contact. The ratio of the dipolar versus quadrupolar term in the timeaveraged magnetic energy for two dipoles in an oscillating external magnetic field ≈8πd3/3χV ≈ 40, for χ = 0.42 and separation d equal to the diameter of the bead, ≈1 μm. This ratio is even smaller in reality because the beads do not have a point dipole in the center but have maghemite nanoparticles dispersed throughout the volume.13 In our experiments, we saw cells up to 300 μm above the glass surface, which is in accordance with the maximum hydrostatic height hmax predicted from theory.20 For sheets of typical size of 20 μm (estimated from Figure 8) in water at 37 °C (ρH2O = 0.993 g/cm3) made from beads with ρ = 1.7 g/ cm3 13 and diameter of 1.05 μm, the volume fraction f V = 0.15. At B0 = 10 mT, χ = 0.42, and g = 9.81 m/s2, we obtain hmax = f Vχ2B02/(ρ − ρH2O)gμ0 ≈ 320 μm. This height can be overcome by adding more superparamagnetic beads with higher χ and employing stronger magnetic fields (using the same particles and B0 = 200 mT and hmax ≈ 13 cm). Growing macroscopic scaffolds should thus be possible. With the result of the pulling experiment, we can estimate how much weight the proposed scaffold is able to carry. In the growth experiments, the magnetic field was 10 mT. That gives force of 6 ± 2 pN to pull out a bead, which in water at 37 °C is already itself enough to carry a 10 μm diameter sphere of density ρ = 2.2 ± 1.4 g/cm3. Because the cells have density close to water25,26 and attach to patches of colloids rather than a single colloid, the scaffold is clearly capable of supporting the cells after they adhere. When B0 is increased, this force may be further enhanced. This does not prove that the scaffold will be able to support large numbers of cells. They may bend it due to traction forces, which are in the nanonewton range;27 therefore, further studies with longer growth times and larger scaffolds should address this questions. Observations of scaffold stiffening with increased B0 indicate that this method may be used for exploration of cell proliferation dependent upon mechanical properties of the growth substrate. Further advances may include coating of the magnetic spheres with different interaction proteins: such as matrix proteins, antibodies, or growth factors. The variety of accessible scaffolds can be expanded by other interaction potentials and self-assembly approaches28 or by particles with different shapes or sizes29 as well as mixtures of various particles.
general appearance of the shapes does not differ from experiments without the cells. The adhesion of the cells to the structure was tested with optical tweezers (see Video S1 of the Supporting Information). Immediately after the onset of the experiment, a sample cell attached only to colloids 40 μm above the glass surface was pulled with an optical force of several tens of piconewtons. The structure deformed as the cell was moved around and was returned to its initial configuration after the cell escaped the optical trap or the optical tweezers were turned off. There was no rupturing of the colloidal structure or detachment of the cells from the scaffold. We have first grown CHO cells on sheets made of untreated beads, but their survival was limited to 1 day after the seeding; therefore, we coated the beads with fibronectin, and the cells survived up to 5 days. Their viability was proven with a fluorescent dye calcein AM, which activates after entering a living cell. Figure 8 shows an example snapshot of cells attached
Figure 8. (A) Bright field image and (B) fluorescent image (z ≈ 30 μm) of the same area. Cells after 5 days still grow in the chamber. The viability was checked with calcein AM, which starts fluorescing when it enters the metabolically active cell. The images show that the cells are viable.
to just the scaffold 30 μm above the bottom of the growth chamber. A comparison of the bright field image and fluorescence microscopy image after 5 days of growing reveals that cells are still metabolically active.
■
DISCUSSION Before a scaffold may be proclaimed as suitable for cell growth, several criteria have to be met: the ability of the structure to support the cells and retain its structural integrity, successful attachment of the cells to the structure, and long-term survival of the cells. We have shown here that self-assembled magnetic colloidal scaffolds satisfy these criteria. It should be taken into consideration that we have not yet grown a tissue as a result of limitations of our growth chamber, most significantly, disability to exchange the growth medium once the experiment is running. Metabolic activity of the cells after 5 days proves that the superparamagnetic beads are biocompatible and appropriate for tissue engineering. Nevertheless, to transfer our method to medical applications like face prostethics, long-term biocompatibility of superparamagnetic beads should also be addressed. Superparamagnetism of the beads that we used comes from iron oxide (mostly maghemite) nanoparticles monodispersed and enclosed in a polystyrene matrix. The scaffold cannot be removed after the experiment, and it is not biodegradable; therefore, effects of long exposure to a tissue with embedded superparamagnetic beads should be determined. There is an overestimation of the force obtained from model calculation compared to optical tweezer measurement (dashed and solid lines in Figure 3A). It partially arises from batch to 9580
DOI: 10.1021/acs.langmuir.5b02176 Langmuir 2015, 31, 9576−9581
Article
Langmuir
■
(8) Hollister, S. J. Porous scaffold design for tissue engineering. Nat. Mater. 2005, 4, 518−24. (9) Yoo, D. New paradigms in hierarchical porous scaffold design for tissue engineering. Mater. Sci. Eng., C 2013, 33, 1759−72. (10) Chan, B. P.; Leong, K. W. Hydrogels for tissue engineering: scaffold design variables and applications. Eur. Spine. J. 2008, 17, 467− 79. (11) Wang, J. H.; Lin, J. S. Cell traction force and measurement methods. Biomech. Model. Mechanobiol. 2007, 6, 361−71. (12) Tokuda, E. Y.; Leight, J. L.; Anseth, K. S. Modulation of matrix elasticity with PEG hydrogels to study melanoma drug responsiveness. Biomaterials 2014, 35, 4310−4318. (13) Fonnum, G.; Johansson, C.; Molteberg, A.; Mørup, S.; Aksnes, E. Characterisation of Dynabeads® by magnetization measurements and Mössbauer spectroscopy. J. Magn. Magn. Mater. 2005, 293, 41−47. (14) Helgesen, G.; Pieranski, P.; Skjeltorp, A. T. Dynamic behavior of simple magnetic hole systems. Phys. Rev. A: At., Mol., Opt. Phys. 1990, 42, 7271−7280. (15) Tierno, P.; Muruganathan, R.; Fischer, T. M. Viscoelasticity of Dynamically Self-Assembled Paramagnetic Colloidal Clusters. Phys. Rev. Lett. 2007, 98, 028301. (16) Osterman, N.; Poberaj, I.; Dobnikar, J.; Frenkel, D.; Ziherl, P.; Babić, D. Field-Induced Self-Assembly Of Suspended Colloidal Membranes. Phys. Rev. Lett. 2009, 103, 228301. (17) Furst, E. M. Applications of laser tweezers in complex fluid rheology. Curr. Opin. Colloid Interface Sci. 2005, 10, 79−86. (18) Moffitt, J. R.; Chemla, Y. R.; Smith, S. B.; Bustamante, C. Recent Advances in Optical Tweezers. Annu. Rev. Biochem. 2008, 77, 205− 228. (19) Neuman, K. C.; Nagy, A. Single-molecule force spectroscopy: optical tweezers, magnetic tweezers and atomic force microscopy. Nat. Methods 2008, 5, 491−505. (20) Kulić, I. M.; Kulić, M. L. Theory of coherent van der Waals matter. Phys. Rev. E Stat. Nonlin. Soft Matter Phys. 2014, 90, 062313. (21) Shevkoplyas, S. S.; Siegel, A. C.; Westervelt, R. M.; Prentiss, M. G.; Whitesides, G. M. The force acting on a superparamagnetic bead due to an applied magnetic field. Lab Chip 2007, 7, 1294−302. (22) Martin, J. E.; Anderson, R. A.; Williamson, R. L. Generating strange magnetic and dielectric interactions: Classical molecules and particle foams. J. Chem. Phys. 2003, 118, 1557−1570. (23) Martin, J. E.; Snezhko, A. Driving self-assembly and emergent dynamics in colloidal suspensions by time-dependent magnetic fields. Rep. Prog. Phys. 2013, 76, 126601. (24) Kulić, I. M.; Kulić, M. L. Self-Assembly of Colloidal Superstructures in Coherently Fluctuating Fields. Phys. Rev. Lett. 2013, 111, 198301. (25) Grover, W. H.; Bryan, A. K.; Diez-Silva, M.; Suresh, S.; Higgins, J. M.; Manalis, S. R. Measuring single-cell density. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 10992−6. (26) Bryan, A. K.; Hecht, V. C.; Shen, W.; Payer, K.; Grover, W. H.; Manalis, S. R. Measuring single cell mass, volume, and density with dual suspended microchannel resonators. Lab Chip 2014, 14, 569−76. (27) Discher, D. E.; Janmey, P.; Wang, Y. L. Tissue cells feel and respond to the stiffness of their substrate. Science 2005, 310, 1139−43. (28) Yan, J.; Bloom, M.; Bae, S. C.; Luijten, E.; Granick, S. Linking synchronization to self-assembly using magnetic Janus colloids. Nature 2012, 491, 578−81. (29) Sacanna, S.; Pine, D. J.; Yi, G.-R. Engineering shape: the novel geometries of colloidal self-assembly. Soft Matter 2013, 9, 8096−8106.
CONCLUSION We have successfully demonstrated the ability of the selfassembled magnetic scaffolds to support cell growth. The cells survived up to 5 days in our setup, thus showing that tissue growth on appropriately coated structures is possible. The cells firmly attach to the colloidal scaffold and are affected by neither magnetic field nor colloids. The method offers a variety of possible shapes accessible by other ways of manipulation of the external magnetic field or using different particles with modified geometry or surface interaction. The scaffold can be enlarged during a running experiment by adding more colloids to the growth medium, and its mechanical properties can be controlled by an external magnetic field. We believe that this method will find use in experiments with cell cultures not accessible with existing approaches, such as response of cell growth to substrate stiffness.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b02176. Supplementary video S1 caption (PDF) Video showing a cell attached to the scaffold while being pulled by optical tweezers (AVI)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Present Address ‡
Gašper Kokot: Department of Chemistry, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank Mojca Vilfan and Andrej Vilfan for discussions and advice. The authors acknowledge funding by the Slovenian Research Agency, Grants P1-0192, P1-0099, P10055, J1-2209, and J1-2200.
■
REFERENCES
(1) Badylak, S. F. Xenogeneic extracellular matrix as a scaffold for tissue reconstruction. Transplant Immunol. 2004, 12, 367−77. (2) Li, W. J.; Laurencin, C. T.; Caterson, E. J.; Tuan, R. S.; Ko, F. K. Electrospun nanofibrous structure: a novel scaffold for tissue engineering. J. Biomed. Mater. Res. 2002, 60, 613−21. (3) Huang, Z.-M.; Zhang, Y.-Z.; Kotaki, M.; Ramakrishna, S. A review on polymer nanofibers by electrospinning and their applications in nanocomposites. Compos. Sci. Technol. 2003, 63, 2223−2253. (4) Bhardwaj, N.; Kundu, S. C. Electrospinning: a fascinating fiber fabrication technique. Biotechnol. Adv. 2010, 28, 325−47. (5) Drury, J. L.; Mooney, D. J. Hydrogels for tissue engineering: scaffold design variables and applications. Biomaterials 2003, 24, 4337−5. (6) Correa-Duarte, M. A.; Wagner, N.; Rojas-Chapana, J.; Morsczeck, C.; Thie, M.; Giersig, M. Fabrication and Biocompatibility of Carbon Nanotube-Based 3D Networks as Scaffolds for Cell Seeding and Growth. Nano Lett. 2004, 4, 2233−2236. (7) Lee, M.; Wu, B. M. Recent advances in 3D printing of tissue engineering scaffolds. Methods Mol. Biol. 2012, 868, 257−67. 9581
DOI: 10.1021/acs.langmuir.5b02176 Langmuir 2015, 31, 9576−9581
