www.acsnano.org
Remote Control of Multimodal Nanoscale Ligand Oscillations Regulates Stem Cell Adhesion and Differentiation Heemin Kang,†,⊥ Dexter Siu Hong Wong,†,⊥ Xiaohui Yan,‡ Hee Joon Jung,§,#,∥ Sungkyu Kim,§,#,∥ Sien Lin,¶ Kongchang Wei,† Gang Li,¶ Vinayak P. Dravid,*,§,#,∥ and Liming Bian*,†,Δ,▽,◊,□ †
Department of Biomedical Engineering, ‡Department of Physics, ¶Department of Orthopaedics and Traumatology, Faculty of Medicine, ΔShun Hing Institute of Advanced Engineering, and ▽Shenzhen Research Institute, □ Centre for Novel Biomaterials, The Chinese University of Hong Kong, Hong Kong, China § Department of Materials Science and Engineering and #NUANCE Center, Northwestern University, Evanston, Illinois 60208, United States ∥ International Institute for Nanotechnology, Evanston, Illinois 60208, United States ◊ China Orthopedic Regenerative Medicine Group (CORMed), Hangzhou 310000, China S Supporting Information *
ABSTRACT: Cellular adhesion is regulated by the dynamic ligation process of surface receptors, such as integrin, to adhesive motifs, such as Arg-Gly-Asp (RGD). Remote control of adhesive ligand presentation using external stimuli is an appealing strategy for the temporal regulation of cell−implant interactions in vivo and was recently demonstrated using photochemical reaction. However, the limited tissue penetration of light potentially hampers the widespread applications of this method in vivo. Here, we present a strategy for modulating the nanoscale oscillations of an integrin ligand simply and solely by adjusting the frequency of an oscillating magnetic field to regulate the adhesion and differentiation of stem cells. A superparamagnetic iron oxide nanoparticle (SPION) was conjugated with the RGD ligand and anchored to a glass substrate by a long flexible poly(ethylene glycol) linker to allow the oscillatory motion of the ligand to be magnetically tuned. In situ magnetic scanning transmission electron microscopy and atomic force microscopy imaging confirmed the nanoscale motion of the substrate-tethered RGD-grafted SPION. Our findings show that ligand oscillations under a low oscillation frequency (0.1 Hz) of the magnetic field promoted integrin−ligand binding and the formation and maturation of focal adhesions and therefore the substrate adhesion of stem cells, while ligands oscillating under high frequency (2 Hz) inhibited integrin ligation and stem cell adhesion, both in vitro and in vivo. Temporal switching of the multimodal ligand oscillations between low- and high-frequency modes reversibly regulated stem cell adhesion. The ligand oscillations further induced the stem cell differentiation and mechanosensing in the same frequencydependent manner. Our study demonstrates a noninvasive, penetrative, and tunable approach to regulate cellular responses to biomaterials in vivo. Our work not only provides additional insight into the design considerations of biomaterials to control cellular adhesion in vivo but also offers a platform to elucidate the fundamental understanding of the dynamic integrin−ligand binding that regulates the adhesion, differentiation, and mechanotransduction of stem cells. KEYWORDS: integrin ligand oscillations, SPION, multimodal control, mesenchymal stem cells, in vivo cell adhesion, stem cell differentiation fibronectin, vitronectin, and collagen.3,4 Integrins are heterodimeric proteins with two membrane-spanning subunits, which regulate intracellular mechanotransduction through dynamic links to
M
any cell behaviors and functions, such as adhesion, are dynamically regulated by the interactions between cells and the extracellular matrix (ECM) in vivo.1,2 Such cell−ECM interactions are mediated by the dynamic binding of cell surface receptors, such as integrin, to their recognition ligands, such as the Arg-Gly-Asp (RGD) tripeptide sequences, present in cell-adhesive ECM components including © 2017 American Chemical Society
Received: April 25, 2017 Accepted: August 25, 2017 Published: August 25, 2017 9636
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
Cite This: ACS Nano 2017, 11, 9636-9649
Article
ACS Nano
Scheme 1. Summary of experimental procedures used in this study. Remote control of multimodal ligand oscillations was used to regulate the adhesion and differentiation of stem cells by altering the oscillation frequency of the magnetic field. A superparamagnetic iron oxide nanoparticle (SPION), functionalized with an amino-silica coating, was conjugated to a glass substrate with a long flexible polyethylene glycol linker (Mn = 5000 Da) to which the RGD ligand was grafted. Such a platform was used to control stem cell adhesion both in vitro and in vivo and subsequent differentiation.
adhesive ligands.5−7 The ligation of integrins to their adhesive ligands triggers the formation of cell focal adhesions, which link to the actin cytoskeleton and mediate mechanotransduction signaling. The decoration of integrin ligands such as RGD peptide on biomaterials, particularly synthetic materials lacking such integrin recognition motifs, is essential for conferring the required bioactivity and eliciting the desirable host responses of biomaterial implants. Integrins undergo free diffusion and immobilization cycles in regulating focal adhesions through their dynamic nanoscale organizations.8 The focal adhesion-mediated cell traction force is highly dynamic and undergoes temporal changes that are attributed to the dynamic integrin−ligand coupling.7,9 It was recently shown that the magnitude of cell adhesion forces on fibronectin or an anti-integrin β1-coated surface displayed oscillatory cycles at a specific frequency.10 Such prior reports imply that dynamic materials with nanoscale oscillations of integrin ligands could be harnessed to regulate cellular adhesion. Dynamic presentation of cell-adhesive ligands offers great potential utility in modulating multifaceted cell−biomaterial interactions.11−14 A recent study showed that the mobile RGD ligand tethered to biomaterials can be harnessed to modulate integrin−ligand binding and therefore cell adhesion in a dynamic fashion.15 Furthermore, the remote control of the presentation of integrin ligand can offer versatile opportunities for the temporal regulation of the cell-adhesive microenvironment.11,16 Such potential has been predominantly realized with in vitro culture platforms by the application of various external stimuli,11 such as electrical potential,17,18 heat,19 enzymes,20 and light.21−27
Among these different stimuli, light, with its potential in vivo applicability, has been most extensively exploited to modulate cellular adhesion and detachment via photochemical reactions, such as the photocleavage of caging molecules21−23 or linker moieties,24,25 or photoinduced isomerization.26,27 Lee et al. recently reported the utilization of light for the temporal regulation of cellular adhesion in vivo.23 However, light that can trigger photochemical reactions is readily absorbed by living tissues, thus resulting in limited penetration depth and potential harmful effects on living cells in vivo. Alternatively, a magnetic field has been utilized to remotely control the motion of magnetic nanoparticles and can penetrate living tissues with significantly reduced absorption, thus allowing widespread and prolonged (longer than hourly scale) applications in vivo.16,28−30 A high-intensity magnetic field (3 T) has been applied for the magnetic resonance imaging of patients, and no harmful effects have been reported.31 Our own study also showed that a static magnetic field can be used to tune the tether compliance of integrin ligands on a substrate to modulate cell adhesion for in vitro culture.32 In addition, oscillating magnetic fields have been applied to manipulate the oscillatory responses of magnetic nanoparticles.30 Thus, harnessing the advantages of oscillating magnetic fields and nanotechnology to modulate the dynamic presentation of integrin ligands can offer a versatile strategy to regulate cell−material interactions in vivo. In this study, we developed a system in which the nanoscale motion of an integrin ligand can be remotely controlled simply by tuning the frequency of an oscillating magnetic field, as graphically illustrated in Scheme 1. Specifically, an RGD ligand-tethered superparamagnetic iron oxide nanoparticle (SPION) was coupled to a 9637
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
Figure 1. Tunable oscillation speeds of an RGD ligand-tethered SPION. (A) X-ray diffraction spectra of a superparamagnetic iron oxide nanoparticle (SPION). Crystalline plane indices were assigned from the diffraction peaks of magnetite (Fe3O4). (B) Transmission electron micrograph, (C) dynamic light scattering analysis, and (D) vibrating sample magnetometer measurement of an RGD ligand-tethered SPION. Scale bar indicates 50 nm. The magnetization was normalized to the dry weight of the SPION. (E) Time-lapse snapshots of the tunable oscillatory motions of an RGD-bearing SPION under an oscillating magnetic field at various frequencies (2.5, 5, or 12.5 mHz). Blue dotted circles highlight a red-fluorescent SPION. Blue arrow indicates the direction of the moving SPION. Scale bar represents 20 μm. (F) Quantified average motion speeds of the SPION under oscillatory magnetic fields of various frequencies. Data are displayed as mean ± standard errors (n = 20).
substrate by a long flexible poly(ethylene glycol) (PEG) linker (Mn = 5000 Da) to allow its controllable motion. We demonstrate that changing the frequency of the applied oscillating magnetic field led to significant changes in the motion speed of the ligandgrafted SPION, and this suggests that a magnetic field can be used to regulate the nanoscale motion of a substrate-tethered ligandgrafted SPION in a frequency-dependent manner. Our results show that stem cell adhesion was regulated solely by altering the frequency of the oscillating magnetic field, both in vitro and in vivo. We also show the temporal regulation of stem cell adhesion and differentiation by switching the oscillation frequency of the magnetic field during the cellular adhesion processes.
mediated by the oscillating movement of a SPION conjugated with the RGD ligand. X-ray diffraction spectra revealed that the diffraction peaks of the SPION corresponded to those of magnetite (Fe3O4) (Figure 1A), as reported previously.33 Transmission electron micrographs (TEM) of the SPIONs showed that the size of the SPIONs was 15 ± 3 nm (Supplementary Figure S1A), which increased to 40 ± 4 nm upon amino-functionalized silica coating (Figure 1B). Further, dynamic light scattering (DLS) measurement revealed that the hydrodynamic sizes of the bare and silica-coated SPION were 14 ± 2 and 46 ± 5 nm, respectively, which is consistent with the TEM observations (Figure 1B and Supplementary Figure S1B). We then characterized the magnetic properties of the bare and silica-coated SPION. Vibrating sample magnetometer hysteresis measurement showed that both the amino-functionalized silicacoated SPIONs (Figure 1D) and bare SPIONs (Supplementary Figure S1C) displayed superparamagnetic characteristics, consistent
RESULTS AND DISCUSSION Tunable Oscillatory Motion Speeds of an AdhesiveLigand-Bearing SPION. Bioadhesive ligand oscillations were 9638
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano with a previous finding33 with saturation magnetization (Ms) of 21 and 51 emu/g, respectively. This finding suggests that the magnetic responses of the SPIONs can be solely controlled by an external magnetic field, with complete reversibility. We next characterized the conjugation of the long flexible linker molecule and the subsequent grafting of the RGD ligands onto the silica-coated SPIONs. Zeta potential measurements revealed sequential changes in the surface charges in each step of the molecular tethering process. The amino-silica coating conferred a positive surface charge of 20.6 ± 6.1 mV, which differed from that of the bare SPION of −1.3 ± 1.1 mV (Supplementary Figure S2A). We further grafted the flexible linker of maleimide-polyethylene glycol-N-hydroxysuccinimide ester, with a large molecular weight (Mn = 5000 Da), to the amino-functionalized surface of the SPIONs, resulting in a negative surface charge of −17.5 ± 4.7 mV. We subsequently tethered the RGD peptide onto the PEGylated surface of the SPIONs, yielding a surface charge of −7.7 ± 4.8 mV. In addition to the analysis of the changes in the surface charges, Fourier transform infrared spectroscopy measurements revealed Fe−O bonds of the bare SPIONs detected at 582 cm−1 (Supplementary Figure S2B), consistent with a previous report.34 The RGD ligand-tethered SPIONs exhibited absorption peaks at 795 and 1065 cm−1, indicative of Si−O bonds, and peaks at 1539 and 1648 cm−1, suggestive of amide II and amide I bonds, respectively, and these collectively confirmed the successful tethering of the RGD peptide to the SPIONs. Following the characterization of the bioadhesive ligandbearing SPION, we characterized the controllable motion of the SPION under an oscillating magnetic field. We fluorescently labeled the SPION with rhodamine isothiocyanate. We externally applied an oscillating magnetic field by oscillating a permanent magnet at various frequencies ranging from 0 to 12.5 mHz and monitored the motion trajectory of the SPION. The red-fluorescent SPION displayed an oscillating motion under the oscillating magnetic field, and the motion speeds of the SPION increased as the oscillation frequency (i.e., oscillation speed) of the magnetic field increased (Figure 1E,F). The realtime motion of the SPION under an oscillating magnetic field is shown in the movies, on which bright field images were overlaid as references (Supplementary Movies S1−S4). Such trackable motion of the nanoparticle was previously demonstrated using an optical microscope.35 Such controlled motion speeds of the SPION, achieved by changing the frequency of the applied oscillating magnetic field, could be extrapolated to tuning the nanoscale motion speeds of the SPION tethered to a substrate via a long flexible linker, in a frequency-dependent manner. In Situ Magnetic Imaging of Nanoscale Motion and Integrin Binding of an Adhesive-Ligand-Bearing SPION Tethered onto a Substrate. Having verified the tunable motion speeds of the RGD ligand−SPION by modulating the oscillation frequencies of the applied magnetic field, we further conjugated the ligand−SPION to a substrate and examined whether the controllable motions of the substrate-tethered SPION under various magnetic field oscillations are achievable. To allow the nanoscale motion of the substrate-tethered SPION under a magnetic field, we utilized a long flexible linker, PEG (Mn = 5000 Da). Tethering of the RGD ligand to a substrate via a PEG molecule of such a high molecular weight is known to inhibit cell adhesion.36 We pretreated the glass substrate to present thiol groups for the subsequent serial conjugations of the PEGylated SPIONs and RGD ligand (Supplementary Figure S3). We measured the water contact angles to analyze the changes in the surface chemistry
of the substrates (Supplementary Figure S4A,B). The contact angle measurements revealed that the thiolization of the glass substrate with mercaptopropylsilatrane (MPS) molecules reduced the surface hydrophilicity with increasing the water contact angle from 34.7 ± 2.4° to 61.7 ± 2.3°. Following the grafting of a PEGylated SPION onto the substrate, surface hydrophilicity increased with a decreased contact angle of 42.0 ± 2.8° owing to the hydrophilic nature of the PEG molecule. We further tethered the RGD peptide to the PEGylated SPION and observed a contact angle of 39.6 ± 1.4°. We further assessed the distribution and topography of tethered RGD ligand−SPION on the substrate. Scanning electron microscopy (SEM) images show a homogeneous distribution of tethered ligand−SPION in a monolayer onto the substrate (Figure 2A). The density of the ligand-tethered SPION was measured as 12 ± 2 particles/μm2. Atomic force microscopy (AFM) analysis showed the substratetethered SPION with lateral and vertical diameters of around 40 nm (Figure 2B), consistent with prior TEM analysis. Thus, we suggest that such tethered SPIONs present RGD ligand clusters37 that have been shown to regulate cellular adhesion and differentiation depending on various parameters, such as the coupling strength,38 spacing,39 and order/disorder.40 We next characterized the in situ nanoscale motion of the ligand−SPION under an oscillating magnetic field by in situ AFM imaging. We coconjugated gold nanorods onto a thiolated substrate through a Au−S bond, and the gold nanorods possess an aspect ratio of 4, in stark contrast to the spherical shape of the SPION, and serve as a nonmagnetic reference. TEM imaging revealed the gold nanorod exhibited dimensions of approximately 78 nm × 18 nm (Supplementary Figure S5A). Furthermore, we used tapping mode AFM imaging to minimize the nonmagnetic motion of the nanoparticles that could be caused by the scanning cantilever probe, and the nonmagnetic movement was negligible within 1 nm scale (Supplementary Figure S5B). We subsequently characterized the magnetic motion of the ligand−SPION against the reference nanorod under a magnetic field applied in different locations and found that the center of the ligand−SPION relative to the gold nanorod under an oscillating magnetic field was shifted (Figure 2C) by approximately 9.4 ± 2.1 nm through quantification of 10 movements. Interestingly, such nanoscale ligand displacement is comparable to the dimension of integrin, which is approximately 10 nm.39,41,42 Furthermore, we characterized the nanoscale motion of the ligand−SPION by in situ magnetic scanning transmission electron microscopy (STEM) imaging. We constructed the setup by attaching the magnet to the in situ electrical biasing holder to manipulate the magnetic field inside the STEM chamber during in situ STEM imaging (Supplementary Figure S6A,B). We characterized the magnetic motion of the ligand−SPION by changes in their interparticle spacing under a magnetic field applied in three different locations, as revealed by relative positional changes of upper ligand−SPIONs to lower ligand− SPION through the movement of both upper and lower ligand− SPIONs (Figure 2D and Supplementary Figure S7). The interparticle spacing between the centers of ligand−SPIONs was found to change approximately by 11.2 ± 3.3 nm through quantification of 10 movements. We demonstrated that the motion speed of the non-substratetethered ligand−SPION was closely dependent on the oscillation frequency of the magnetic field (Figure 1E,F and Supplementary Movies S1−4). Therefore, we postulate that the frequency of the magnetic field oscillation will influence the oscillation speed of the tethered RGD-bearing SPION, and this may affect the 9639
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
Figure 2. In situ magnetic imaging of the nanoscale motion and integrin binding of the RGD ligand tethered onto the substrate. (A) Scanning electron micrograph and (B) 3D atomic force micrograpy (AFM) image for a monolayer of RGD ligand−SPION-coated substrate. The scale bars represent 1 μm. (C) In situ AFM images scanned on the identical area with external magnetic field applied at two opposite outer sides of the scan area. Spherical particle represents a SPION, while utilizing a Au nanorod (aspect ratio of 4) as a nonmagnetic reference. Black dotted lines are drawn along the long axis of the nanorod to characterize the movement of the SPION relative to the nanorod. Red dotted lines are drawn across the centers of the SPIONs to show the movement of the SPIONs. White arrow indicates the direction of the moving SPION from the left to right image. The scale bars indicate 50 nm. (D) In situ magnetic high-angle annular dark-field scanning transmission electron microscopy (HAADFSTEM) images scanned on the identical area with the localized magnetic field manipulated by a magnetic tip relocated at three different positions with an approximately10 min time interval near SPIONs inside the STEM chamber. Gray dotted lines are drawn across the center of lower SPIONs to show the relative positional changes of upper SPIONs through the movement of both upper and lower SPIONs. Red dotted lines are drawn across the centers of the upper SPIONs to show the relative movement of the SPIONs. Red arrows indicate the direction of positional changes of upper SPIONs relative to lower SPIONs from the left to middle to right image. The scale bars indicate 50 nm. (E) Immunofluorescent staining images of integrin β1 binding to the substrate-tethered “Stationary”, “Slow”, “Moderate”, and “Fast” ligand−SPION oscillations under various oscillation frequencies (0, 0.1, 0.5, and 2 Hz) of applied magnetic field, respectively, and (F) corresponding quantification of fluorescence intensities. Red arrows indicate ligand-bound integrin clusters. Scale bars represent 50 μm. Data are shown as mean ± standard errors (n = 6). Different letters indicate statistical significances (p < 0.05) as determined by one-way ANOVA with Tukey−Kramer post hoc test.
mediate focal adhesion and mechanosensing. We applied a magnetic field of various oscillation frequencies (0, 0.1, 0.5, or 2 Hz; hereafter referred to as “Stationary”, “Slow”, “Moderate”, and “Fast”, respectively) to the substrate conjugated with RGD-bearing SPIONs and monitored the binding of
integrin-RGD binding and therefore cell adhesion and spreading on the substrate. To test our hypothesis, we examined whether the oscillation speed of the tethered ligand−SPION under various oscillation frequencies of the applied magnetic field would modulate integrin−ligand binding, which is known to 9640
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
Figure 3. Ligand oscillations regulate stem cell adhesion with oscillation frequency dependency. (A) Time-dependent immnuofluorescent staining micrographs of vinculin (green), actin (red), and nuclei (blue) after 12 and 48 h of hMSC culture under “Stationary”, “Slow”, “Moderate”, and “Fast” ligand−SPION oscillations by applying various oscillation frequencies (0, 0.1, 0.5, and 2 Hz) of magnetic field, respectively, with “No RGD” and “No magnet” controls. Scale bars represent 100 μm. (B) Corresponding quantifications of adhered cell density and area. Data are shown as mean ± standard errors (n = 30). Different letters represent statistical significances (p < 0.05) at the same culture time as determined by one-way ANOVA with the Tukey−Kramer post hoc test.
integrin β1 to the RGD ligand. Immunofluorescent staining of integrin β1 revealed that “Slow” ligand oscillation significantly promoted integrin binding with the ligand, evident as more speckles of fluorescent staining, whereas “Fast” ligand oscillation significantly inhibited integrin ligation, compared with those in the “Stationary” and “Moderate” groups (Figure 2E,F). Altering Integrin-Ligand Oscillation Speeds Regulates Adhesion of Stem Cells. We next evaluated whether the application of a magnetic field at various frequencies can regulate the adhesion of human mesenchymal stem cells (hMSCs) on the substrate coupled with RGD-bearing SPIONs. We coated the RGD-bearing substrate with bovine serum albumin (BSA) to minimize nonspecific cellular adhesion. We also included controls of RGD-bearing substrate in the absence of applied magnetic field and substrate devoid of coupled RGD−SPION (hereafter named “No magnet” and “No RGD”, respectively).
After 12 h of culture, altering RGD ligand oscillation speed by adjusting the oscillation frequencies of the magnetic field yielded strikingly disparate levels of cellular adhesion, analogous to the integrin−ligand binding results described above. Compared to the “Stationary” control under a nonoscillating static magnetic field, “Slow” ligand oscillation significantly promoted cellular adhesion, whereas “Fast” ligand oscillation inhibited cell adhesion (Figure 3A). Quantification of density of cellular adhesion corroborated the observations described above showing 50% more and 49% less adherent cells in the “Slow” and “Fast” groups, respectively, compared to the “Stationary” group (Figure 3B). We also applied the magnetic field of more extreme oscillation frequencies (0.02 and 5 Hz) and observed that the density of cellular adhesion was comparable to the “Slow” group (0.1 Hz) and the “Fast” group (2 Hz), respectively (Supplementary Figure S8A−D). Furthermore, we observed no significant difference in the number of adherent cells among “No magnet”, “Stationary”, and “Moderate” 9641
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
adhesion, thus exhibiting minimal vinculin expression and highly elongated cell shape. Such observation could be attributed to poor integrin binding with a highly oscillatory ligand. Collectively, ligand oscillations at various frequencies (“Slow”, “Moderate”, and “Fast”) mediated by a remotely and continuously applied magnetic field offer a powerful tool with which to regulate cell adhesion and spreading in a physical and noncontact manner. Temporal Switching of Ligand Oscillations Modulates Focal Adhesion of Stem Cells. We next evaluated whether the temporal switching of the oscillation frequency of the magnetic field can regulate stem cell adhesion. We observed the distinct cell adhesion under “Slow” and “Fast” RGD ligand oscillations induced by the oscillating magnetic field and superparamagnetic characteristics of the SPIONs. Therefore, we switched the oscillation frequencies of the magnetic field from “Slow” to “Fast” ligand oscillations (“Slow−Fast”) or from “Fast” to “Slow” ligand oscillations (“Fast−Slow”) after 12 h of culture and further cultured for another 36 h, until 48 h of total culture time. We included nonswitched continuous “Slow” or “Fast” ligand oscillations for 48 h of culture as control groups (Figure 4A). After 12 h of culture, nonswitched continuous “Slow” and “Fast” ligand oscillations resulted in significantly enhanced or inhibited adhesion of stem cells, respectively, and this distinct trend persisted until 48 h of culture (Figure 4A,B). Upon temporal switching from “Slow” ligand oscillation to “Fast” ligand oscillation after 12 h of culture (“Slow−Fast” group), the spreading area and the number of focal adhesions of stem cells significantly decreased (by 46% and 68%, respectively) at 48 h of culture, compared to 12 h of culture, and cells switched their morphology from a spread to an elongated shape (with 208% increase in aspect ratio). The temporally switched “Fast−Slow” group exhibited a reverse trend as compared to the “Slow−Fast” group, with an increase in the spreading area and the number of focal adhesions as well as a switch from elongated to spread morphology. Such findings indicate that the temporal control over oscillation speeds of cell adhesive ligands enabled reversible cell adhesion and spreading. Integrin−ligand binding is highly dynamic during cellular adhesion,7−9 and reversible cellular adhesion was previously demonstrated via photochemical reactions.26,27 Compared to the light-based approach, the utility of a magnetic field to regulate cell adhesion and spreading has the advantage of longer working distance without being obstructed by physical barriers such as biological tissues, thereby allowing potentially widespread in vivo applications. Furthermore, the magnetic control over ligand oscillation speeds also affords finetuning on the extent of cell adhesion and spreading by adjusting the magnetic field oscillation frequency. This allows investigation on the behaviors of cells with different levels of adhesion free from other confounding factors, such as ligand density and substrate stiffness. In addition, switching of the magnetic field oscillation frequency at selected time points provides the opportunity to study the effect of temporal changes in cell spreading on cell behaviors. Tuning Ligand Oscillations Modulates Stem Cell Differentiation. Having examined the effect of RGD ligand oscillation speeds on cellular adhesion, we evaluated whether such dynamic cell−ligand interactions could influence the differentiation of hMSCs. The directed differentiation of stem cells is essential to the success of regenerative therapies and is dictated by the responses of cells to microenvironmental cues, such as the physicochemical properties of the matrix12,13,44−47
groups, but all these three groups displayed more adherent cells than the “No RGD” control, suggesting the efficacy of the RGD ligand to mediate cell adhesion in the presence of BSA blocking agent. As a control, we used PEG as a blocking agent and found that “Slow” and “Fast” ligand oscillation also significantly promoted and inhibited cellular adhesion, respectively, compared with the “Stationary” control, similarly to the results obtained on the BSA-blocked substrates (Supplementary Figure S8E,F). To further evaluate such ligand oscillation-mediated stimulation and suppression of stem cell adhesion, we performed a series of control experiments by using SPION-tethered substrate, but without the conjugation of the RGD ligand (Supplementary Figure S9A) or with the conjugation of scrambled RGD sequences, such as RAD and RGE. We found no significant differences in cellular adhesion among all groups regardless of the oscillating magnetic field frequency, in the absence of RGD conjugation (Supplementary Figure S9B,C) or the conjugation of RAD or RGE (Supplementary Figure S10A−D). This finding suggests that RGD ligand tethering to SPION is required to modulate stem cell adhesion through the oscillating motion of SPION. To examine whether the oscillating magnetic field affects cellular adhesion independent of the magnetic-field-mediated ligand motion, we prepared RGD-bearing pure silica nanoparticles that do not respond to the magnetic field. We found that pure silica nanoparticles exhibited a size (approximately 40 nm) comparable to that of the silica-coated SPION used in this study (Supplementary Figure S11A), and the density of the conjugated pure silica nanoparticles was comparable to that of the silicacoated SPION (Supplementary Figure S11B). In the presence of substrate-tethered RGD-bearing silica nanoparticles, we observed no significant difference in the number of adherent cells among “No magnet”, “Stationary”, “Slow”, and “Fast” groups, suggesting that the oscillating magnetic field itself does not affect the cellular adhesion (Supplementary Figure S11C,D). Because integrin−ligand binding mediates focal adhesion (FA) formation by recruiting FA adaptor molecules, such as vinculin, we examined the cell spreading area and focal adhesion development after 12 and 48 h of culture. The disparate adherent cell density correlated to different extents of cell spreading. “Slow” and “Fast” ligand oscillations resulted in significantly larger (approximately by 168%) or smaller (approximately by 30%) cell spread area, respectively, as compared to the “Stationary” and “Moderate” oscillation groups (Figure 3A,B). The number of focal adhesions per cell quantified by vinculin staining43 correlated positively with the cell spread area (Supplementary Figure S12A). Quantification of the aspect ratio (length/width) of adherent cells revealed the inverse relationship between cell aspect ratio and cell spread area/number of focal adhesions (Supplementary Figure S12B). “Slow” and “Fast” ligand oscillations resulted in significantly smaller and larger aspect ratios, respectively. The focal adhesion of cells is dynamically regulated by integrin ligation events. The “Slow” (0.1 Hz) nanoscale oscillation motion of the tethered RGD ligand yielded the optimal integrin binding and cell adhesion to the substrate. A recent report showed that the magnitude of cell adhesion forces to fibronectin or anti-integrin β1-coated surface exhibits oscillatory cycles at 0.01−0.02 Hz.10 In correlation with the dynamic nature of integrin ligation processes, we suggest that “Slow” ligand oscillation increases the likelihood that integrin will contact the ligand and thus stimulate cell adhesion, analogous to a previous report.15 In stark contrast, “Fast” ligand oscillation (2 Hz) was found to significantly inhibit integrin ligation and cellular 9642
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
Figure 4. Time-regulated switching of the ligand oscillations allows reversible focal adhesion of stem cells. (A) Time-dependent immnuofluorescent staining micrographs of vinculin (green), actin (red), and nuclei (blue) after 12 and 48 h of hMSC culture under ligand−SPION oscillations with reversible switching for oscillation frequencies between “Slow” oscillation (0.1 Hz) and “Fast” oscillation (2 Hz) of the magnetic field after 12 h of culture for “Slow−Fast” and “Fast−Slow” groups. Nonswitched “Slow” and “Fast” oscillation groups are included as a control. Scale bars indicate 50 μm. (B) Corresponding quantification of adhered cell area, focal adhesion number, and cell shape factor. Data are shown as mean ± standard errors (n = 30). Different letters indicate statistical significances (p < 0.05) at the same culture time as determined by one-way ANOVA with the Tukey−Kramer post hoc test.
with the “Slow” group 222% higher than the “Stationary” group (Figure 5B). In addition, switching ligand oscillation from “Slow” to “Fast” mode after 12 h (“Slow−Fast” group) resulted in lower RUNX2 expression at day 7 (by approximately 54%) compared to the continuous “Slow” oscillation group. Interestingly, switching from “Fast” to “Slow” ligand oscillation (“Fast−Slow” group) led to higher RUNX2 expression comparable to that of the “Slow” group. These findings indicate that “Slow” ligand oscillation helped to recover the adhesion, spreading, and differentiation potential of stem cells following the initial temporal suppression of cellular adhesion under “Fast” ligand oscillation. We further performed alkaline phosphatase (ALP) staining to characterize the osteogenesis of stem cells under altered ligand oscillations. The degree of ALP staining was found to correlate to the level of RUNX2 nuclear localization, further confirming the effect of tuning ligand oscillations on stem cell differentiation.
and soluble factors.48−51 The adhesion and morphology of stem cells are known to play an important role in regulating mechanotransduction and differentiation of stem cells.52−54 hMSCs were cultured in basal growth medium for 12 h to allow cellular adhesion and subsequently cultured in osteogenic induction medium for 7 d to examine their osteogenesis. We included the groups with temporal switching of ligand oscillations, “Slow−Fast” and “Fast−Slow” groups, at 12 h of culture to analyze differentiations at 7 d, and the continuous “Fast” ligand oscillation group was excluded owing to significantly inhibited cellular adhesion. Immunofluorescent staining for RUNX2, a master regulator in osteogenic differentiation, revealed that “Slow” RGD ligand oscillation led to the highest level of RUNX2 expression in hMSCs (Figure 5A). Quantification of nuclear to cytoplasmic fluorescence ratio corroborated such a pattern of expression level 9643
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
Figure 5. Modulating ligand oscillations regulates stem cell differentiation through involvement of mechanosensing. (A) Immunofluorescent staining against RUNX2 (green), actin (red), and nuclei (blue) and ALP staining as well as YAP (green), actin (red), and nuclei (blue) after 7 and 2 d of hMSC culture, respectively, under “Stationary”, “Slow”, and “Moderate” oscillations by applying various oscillation frequencies (0, 0.1, and 0.5 Hz) of a magnetic field, respectively, with “No magnet” control. The groups include time-regulated switched groups “Slow−Fast” and “Fast− Slow” at 12 h of culture. Scale bars indicate 100 μm. (B) Corresponding ratiometric analyses of nuclear to cytoplasmic fluorescence intensities for RUNX2 and YAP. Data are presented as mean ± standard errors (n = 30). Different letters indicate statistical significances (p < 0.05) as determined by one-way ANOVA with the Tukey−Kramer post hoc test.
Remotely Tuning RGD Ligand Oscillation Speeds Modulates in Vivo Substrate Adhesion of Stem Cells. We further evaluated whether the remote control of RGD ligand oscillation speeds could regulate stem cell adhesion in vivo. We implanted a substrate bearing the RGD−SPION into the subcutaneous pocket of nude mice and injected hMSCs onto the substrate. We externally applied a magnetic field of varying oscillation frequencies to the implantation area, while keeping the mice under anesthesia for 4 h before extracting the implanted substrate to examine cell adhesion. Significantly more adhered cells were found in the “Slow” ligand oscillation, “Stationary”, and “No magnet” groups (approximately 1400−1550 cells/cm2), compared to the minimal cellular adhesion observed in the “Fast” oscillation and “No RGD” groups (approximately 550−650 cells/cm2) (Figure 6A,B). We next identified cell species by immunofluorescent staining of human-specific nuclear antigen and found that all in vivo adhered cells were human donor cells (Figure 6A). This could be attributable to the insufficient
We then investigated the role of mechanotransduction in regulating such disparate differentiation potentials because focal adhesion is known to mediate mechanosensing of stem cells. Immunostaining of Yes-associated protein (YAP), a key regulator of mechanotransduction, at day 2 revealed more pronounced nuclear localization in the “Slow” and “Fast− Slow” group compared to other control groups (Figure 5A,B), confirming the enhanced mechanotransduction signaling in hMSCs under “Slow” ligand oscillations. Immunofluorescent staining of transcriptional coactivator with PDZ-binding motif (TAZ) showed higher nuclear localization in the “Slow” group than the “No Magnet”, “Stationary”, and “Moderate” groups, further confirming that the “Slow” ligand oscillation promoted the mechanotransduction (Supplementary Figure S13A,B). Therefore, we collectively suggest that the remote control of ligand oscillation speeds via magnetic field regulates the adhesion, spreading, mechanosensing, and consequently differentiation of stem cells. 9644
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
regulation of cellular adhesion in vivo by adjusting the oscillation frequency of an external magnetic field has not been demonstrated before. The noninvasive, penetrative, and tunable nature of the external magnetic field potentially allows temporal manipulation of cell adhesion to implants in deep interior anatomic sites in vivo. Having verified the effective ligand oscillation-mediated control over stem cell adhesion in vivo, we evaluated whether pronounced spreading and focal adhesion of the in vivo adhered stem cells under “Slow” ligand oscillation stimulate their mechanosensing in vivo. The nuclear to cytoplasmic ratio of YAP in the “Slow” group was found to be significantly higher than that in the other groups showing minimal nuclear staining, thereby indicating the possible involvement of focal-adhesion-associated mechanotransduction signaling in vivo (Supplementary Figure S14A,B).
CONCLUSION In summary, this study demonstrates an approach to controlling the nanoscale motion of the integrin-binding ligands by adjusting the oscillation frequency of an applied magnetic field and shows that the motion speeds of the ligands significantly influence the adhesion and differentiation of stem cells. Specifically, RGD ligand-coupled SPION was bound to a substrate with a long and flexible polymer linker. The controllable motion speeds of the RGD-ligand-coupled SPION in an oscillation-frequency-dependent manner (demonstrated by fluorescent imaging) led to their nanoscale motion upon substrate tethering (demonstrated by in situ magnetic STEM and AFM imaging). Remotely inducing slow or fast ligand oscillations by tuning the oscillation frequency of the magnetic field significantly promoted or inhibited integrin ligation and the adhesion of stem cells, respectively, both in vitro and in vivo, thereby significantly regulating the osteogenic differentiation of the stem cells. Temporal switching of the multimodal ligand oscillation speed reversibly regulated cellular adhesion. The speeds of the ligand oscillations significantly affected the differentiation and mechanosensing of the stem cells. Such controlled nanoscale ligand oscillations can be harnessed to shed light on the dynamic processes of integrin ligation to an extracellular matrix. Furthermore, such a noninvasive, tissue penetrative, cytocompatible, and tunable approach by the application of a magnetic field holds great potential utility to manipulate and study the cellular adhesion and responses to implant biomaterials in the host microenvironment in vivo.
Figure 6. Remote control of ligand oscillations regulates stem cell adhesion in vivo. (A) Immnuofluorescent staining micrographs of human-specific nuclear antigen (HuNu; green) or vinculin (green), actin (red), and nuclei (blue) after 4 h of in vivo hMSC injection onto the implanted substrate under “Stationary”, “Slow”, and “Fast” ligand oscillations by applying various oscillation frequencies (0, 0.1, and 2 Hz) of a magnetic field, respectively, with “No RGD” and “No magnet” controls. Scale bars indicate 100 μm. (B) Corresponding quantification of adhered cell density, area, focal adhesion, and aspect ratio. Data are presented as mean ± standard errors (n = 30). Different letters indicate statistical significances (p < 0.05) as determined by one-way ANOVA with the Tukey−Kramer post hoc test.
METHODS
time (4 h) for recruitment and subsequent adhesion of host cells. The adhesion of recruited immune cells including neutrophils and macrophages onto RGD-bearing biomaterials was observed following 24 h of implantation.23 Furthermore, “Slow” ligand oscillation significantly increased the number of focal adhesion sites as evidenced by vinculin staining (by approximately 236−433% higher) and spread size of adherent cells (by approximately 291−445% higher), compared to all other groups (Figure 6A,B). In contrast, the “Fast” group exhibited 65% less focal adhesions and 33% smaller cell size than those of the “Stationary” group. These findings demonstrate that remote control of ligand oscillations via magnetic field to regulate stem cell adhesion remains effective under the dynamic and complex in vivo environment. To the best of our knowledge, such multimodal (“Slow” and “Fast” ligand oscillations) remote
Preparation of Superparamagnetic Iron Oxide Nanoparticles. To remotely control integrin ligand-tethered nanoparticles, a SPION was prepared by following a previously reported procedure.55 Briefly, 36.5 g (120 mmol) of sodium oleate and 10.8 g (40 mmol) of iron(III) chloride hexahydrate were dissolved in a mixture of 80 mL of ethanol, 60 mL of deionized (DI) water, and 140 mL of heptane at 70 °C for 4 h under an inert atmosphere. The iron-oleate-containing heptane layer was separated and washed with 40 mL of DI water followed by the evaporation of heptane. Then 36 g (40 mmol) of the dried iron oleate was mixed with 5.7 g (20 mmol) of oleic acid and 200 g of 1-octadecene at 100 °C for 5 min. The mixture was further heated to 320 °C and kept for 30 min. The mixture was then cooled to 25 °C, and 500 mL of ethanol was added. The obtained SPION was washed with ethanol and dispersed in heptane. Functionalization of SPIONs. To couple a long flexible polymer chain to a SPION, the SPION was functionalized through amino-silica coating (SPION@SiO2). A 30 mg amount of SPION was dispersed in 9645
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano 100 mL of cyclohexane, 20 mL of Triton-X, 20 mL of 1-hexanol, 2 mL of NH4OH, and 4 mL of DI water and stirred for 30 min. To the mixture, 100 μL of tetraethyl orthosilicate (TEOS) was added and stirred for 10 min. (3-Aminopropyl)triethoxysilane (APTES) was further added to the mixture and stirred for 6 h. As a control, this protocol was also used to prepare pure silica nanoparticles in the absence of a SPION. A 100 mL amount of acetone was added to the mixture, which was then washed with acetone and dimethylformamide (DMF). The resultant aminosilica-coated SPION was collected by a permanent magnet. Conjugation of a Flexible Polymer Linker to a Functionalized SPION. To tether a functionalized SPION (SPION@SiO2) onto a substrate and subsequently maneuver the motion, a long flexible polymer linker was conjugated. A 20 mg portion of amino-silica-coated SPIONs was suspended in 1 mL of DMF to which 10 mg of Maleimide−polyethylene glycol (Mn = 5000 Da)−NHS ester (JenKem Technology) and 0.2% N,N-diisopropylethylamine (DIPEA) were added. The suspension was stirred for 16 h under dark conditions, collected by a permanent magnet, and then washed with DMF and DMSO. The resultant PEGylated SPION was later used for conjugation onto a substrate. For characterization of the integrin−ligand-tethered PEGylated SPION, the suspension in DMSO was mixed with 0.14 mM thiolated RGD peptides (GCGYGRGDSPG; Mn = 1025.06 Da, GenScript), 0.2% DIPEA, and 10 mM tris(2-carboxyethyl)phosphine hydrochloride (TCEP) for 16 h under gentle stirring and dark conditions and washed with DMSO and DI water. As a control, we conjugated scrambled RGD peptides (GCGYGRADSPG or GCGYGRGESPG, GenScript). X-ray Diffraction (XRD). To characterize the crystalline phase of the synthesized magnetic nanoparticles, XRD spectra were measured. The magnetic nanoparticles were subject to flash-freezing and lyophilization. The dried particles were mounted and scanned by using a Smartlab (Rigaku) diffractometer with a Cu target (40 kV, 80 mA) at a diffraction angle ranging from 20 to 70 degrees at 10 degrees/min. Transmission Electron Microscopy. To characterize the size and morphology of the SPION before and after amino-silica coating, TEM imaging was performed by using a CM-200 (Philips, Oregon, USA) at an accelerating voltage of 100 kV. The size of the SPION was quantified by ImageJ from 20 different particles in three different images and presented as mean ± standard errors. Dynamic Light Scattering and Zeta Potential. To determine the size distribution profile and changes in surface charges of the SPION before and after amino-silica coating, DLS and zeta potential measurements were performed by using photon correlation spectroscopy (Brookhaven Instruments Corporation). The SPION was suspended in DI water and subject to the measurements. Vibrating Sample Magnetometer (VSM) Measurement. To evaluate the magnetic property of the SPION, VSM measurement (PPMS model 6000 Quantum Design) was conducted. The measurement was conducted at 300 K under an applied magnetic field from −6000 to 6000 Oe. The magnetic moment (emu/g) of the hysteresis loop was presented after normalization of the dry weight of the SPION. Fourier Transform Infrared Spectroscopy (FTIR). To evaluate changes in chemical bond characteristics of the SPION before and after RGD ligand tethering, FTIR was conducted by using an IRTracer-100 (Shimadzu). The SPION was subject to flash-freezing and lyophilization prior to the measurement. Synthesis of Fluorescent SPION. To fluorescently track the motion of the SPION, we conjugated a fluorescent molecule to the SPION. A 10 mg sample of rhodamine isothiocyanate (RITC) was added into a suspension of 20 mg of amino-silica-coated SPIONs in 1 mL of DMF. The mixture was stirred for 2 d in the dark and washed with DMF. Maleimide−-polyethylene glycol−NHS ester coupled with RGD peptide was added to the RITC-conjugated amino-silica-coated SPION with 0.2% DIPEA and kept for 16 h. The resulting fluorescent RGD-ligand-bearing SPION was washed with DMF and DI water for fluorescent imaging. Confocal Microscopy Imaging of an Oscillating Fluorescent SPION. To examine the tunable oscillatory motion of the SPION dependent upon the frequencies of the oscillating magnetic field,
confocal microscopy imaging was performed. The RGD-bearing fluorescent SPION was suspended in DI water. The fluorescent image with the bright field image as a reference was taken on the same area every 10 s while the external magnetic field was applied by oscillating a permanent magnet (40 mT) at frequencies of 0, 2.5, 5, and 12.5 mHz. Twenty time-lapse snapshots of the oscillating SPION were used to quantify the frequency-dependent average motion speed of the SPION and generate movies demonstrating the oscillatory motion of the SPION at various frequencies. Coupling of a Flexible-Chain-Tethered SPION to a Substrate and Further Grafting of an Integrin Ligand. To conjugate a flexiblechain-tethered (PEGylated) SPION to a substrate, culture-grade glass coverslips (22 mm × 22 mm) were used. The glass substrates were immersed in the mixture of HCl and MeOH (1:1) for 30 min to remove any organic impurities from the surface of the substrates and subsequently washed with DI water. For thiolization, the substrates were immersed in 1 mM MPS in methanol for 30 min under dark conditions and then washed with methanol. The thiolized substrate was soaked in 1 mL of PEGylated SPION in DMSO containing 0.2% DIPEA in the dark for 16 h and washed with DMSO. To further conjugate the RGD ligand to the PEGylated SPIONcoated substrate, the substrate was immersed in 0.14 mM RGD peptides in 1 mL of DMSO containing 0.2% DIPEA and 10 mM TCEP for 16 h under gentle rotation and dark conditions and washed with DMSO and DI water. Contact Angle Measurement. To evaluate changes in the surface chemistry of the substrate upon surface modifications, water contact angle measurements were carried out. The substrates were dried at 37 °C for 3 h. Approximately 2 μL of water droplets was applied onto the substrates, and the surfaces of the substrates were imaged at fixed position by using a contact angle analyzer (JC2000C1, Powerrach Ltd.). The contact angle was quantified from four different images and presented as mean ± standard errors. Scanning Electron Microscopy. To visualize the conjugation of RGD-ligand−SPION (RGD-PEG-SPION@SiO2) onto the glass substrate, SEM imaging (SU8000 Series, Hitachi High-Tech Ltd., UK) was carried out. The substrate was rinsed with DI water and dried at 37 °C for 3 h. The dried substrate was gold-coated using a sputter coater and imaged at an accelerating voltage of 15 kV. The density of the substrate-coupled RGD-ligand−SPION was quantified by ImageJ from 10 different images and presented as mean ± standard errors. In Situ Atomic Force Microscopy. To visualize the 3D distribution of tethered integrin−ligand−SPION (RGD-PEG-SPION@SiO2) onto the glass substrate as well as characterize movement of the tethered ligand−SPION, AFM imaging (Bruker) was performed in tapping mode at 25 °C. The spring constant of the AFM cantilever (Bruker) was 0.4 N/m. The 3D height image was obtained at a scan size of 5 × 5 μm. To visualize and quantify the movement of the tethered ligand− SPION in situ, a gold nanorod (aspect ratio of 4) was coconjugated to the substrate that served as a nonmagnetic reference by following a previously reported protocol.56 Briefly, 0.7 g of hexadecyltrimethylammonium bromide (CTAB) and 0.1234 g of sodium oleate (TCI America) were dissolved in 25 mL of DI water at 30 °C, and 1.8 mL of AgNO3 was added to the mixture. The mixture was stirred for 15 min, and 25 mL of 1 mM Au3+ from gold(III) chloride trihydrate (HAuCl4· 3H2O, Sigma) was injected into the solution. The mixture was stirred for 90 min, and 210 μL of 37% HCl was added. After 15 min of stirring, 125 μL of 64 mM ascorbic acid and seed solution (the mixture of 2.5 mL of 200 mM CTAB, 2.5 mL of 0.5 mM Au3+, and 0.3 mL of 10 mM chilled sodium borohydride that was aged for 2 h) were added to the mixture and kept for 12 h. Thiolated glass substrates (SPION conjugated) were immersed in a 0.5 nM gold nanorod solution for 2 h and washed with DI water thoroughly and dried under nitrogen. In situ AFM imaging was performed on the same scan area with an external magnetic field applied by a permanent magnet (40 mT) placed at two opposite outer sides of the scan area to visualize and quantify the movement of tethered ligand−SPION from the reference gold nanorod. Approximately 10 movements were quantified between the centers of the ligand−SPION and presented as mean ± standard errors. 9646
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano In Situ Magnetic High-Angle Annular Dark-Field Scanning Transmission Electron Microscopy (HAADF-STEM). To characterize nanoscale motion of the RGD-ligand−SPION conjugated to the substrate, in situ HAADF-STEM imaging was performed by using a JEOL 2100F with a 1 nm probe size, a condenser aperture of 20 μm, and a collection angle of 80−150 mrad for Z contrast. We constructed a setup by tethering a flexible curved rubber magnet (sized to 1 mm thick and 3 mm long) securely to a mobile tungsten STM tip of an in situ electrical biasing holder (Nanofactory ST-1000) that can move controllably inside the STEM chamber during in situ TEM imaging. A 50-nm-thick silicon oxide TEM grid (Dune Sciences, cat. no. NG02011A), used as a substrate for conjugation of the RGD-ligand−SPION, was fixed at the other electrode side. The crooked tip part of the rubber magnet in contact with the TEM grid was relocated at three different positions to manipulate the local magnetic field near individual SPIONs and monitored by HAADF-STEM to characterize interparticle movement. Note that each STEM image has at least a 5−10 min interval after relocation of the magnetic tip to correct the ronchigram (beam astigmatism) because local magnetic field variation distorts the electron beam condition significantly. Approximately 10 movements were quantified between the centers of ligand−SPIONs and presented as mean ± standard errors. Integrin β1 Binding to Oscillating RGD Ligands. To evaluate integrin β1 binding onto oscillating RGD ligands at various frequencies, an RGD-ligand−SPION tethered substrate was incubated in 50 μg/mL of integrin β1 in phosphate-buffered saline (PBS) at 37 °C for 30 min under the applied magnetic field by a permanent magnet (40 mT) at oscillation frequencies of 0, 0.1, 0.5, and 2 Hz by electrical setup. The incubated substrate was fixed and subject to immunofluorescent staining of integrin β1 to examine the bound integrin β1 onto the oscillating RGD ligand. In Vitro Culture. To examine cellular adhesion on oscillating RGD ligands in varying frequencies, an RGD-ligand−SPION coupled (culture-grade) substrate was sterilized and blocked with 1% BSA (Sigma-Aldrich) at 37 °C for 1 h or 0.1% maleimide−polyethylene glycol (Mn = 5000 Da)−OH (JenKem Technology) for 1 h as a control. The blocked substrate was immediately subject to seeding of hMSCs (passage 4, Lonza) at plating density of 5000 cells/cm2 and cultured in basal growth medium (high glucose DMEM, 10% [v/v] fetal bovine serum, 4 mM L-glutamine, and 50 U/mL penicillin/streptomycin; Thermo Scientific) at 37 °C and 5% CO2. The adhesion of stem cells onto oscillating RGD ligands for 12 and 48 h under the applied magnetic field was done by oscillating a permanent magnet (40 mT) at frequencies of 0, 0.1, 0.5, and 2 Hz by electrical setup. Temporal switching of the oscillation frequency was performed after 12 h of cell plating between 0.1 and 2 Hz. The differentiation of stem cells following their focal adhesion was examined by changing the basal growth medium to osteogenic-inducing medium [basal growth medium containing 10 mM β-glycerophosphate (Sigma-Aldrich), 50 μM ascorbic acid-2-phosphate (Sigma-Aldrich), and 100 nM dexamethasone (Sigma-Aldrich)] after 12 h of cell adhesion. The osteogenic-inducing medium was replenished every 2 days, and differentiation was analyzed after 7 days of culture. Immunofluorescent Staining. The cell cultures were fixed by using 4% (w/v) paraformaldehyde at 25 °C for 15 min and washed with PBS. The fixed cells were permeabilized with 0.25% (v/v) Triton-X (Sigma-Aldrich) in PBS at 25 °C for 10 min. The treated cells were blocked with 1% (w/v) BSA at 37 °C for 1 h. The blocked cells were incubated with primary antibody against integrin β1 (1:100, mouse, Santa Cruz Biotechnology), vinculin (1:400, mouse, Sigma-Aldrich), RUNX2 (1:100, mouse, Santa Cruz Biotechnology), YAP (1:100, mouse, Santa Cruz Biotechnology), or TAZ (1:100, mouse, Santa Cruz Biotechnology) in blocking solution at 4 °C for 16 h and washed with PBS containing 0.5% Tween 20 three times. The cells were incubated with secondary antibody (1:250; Thermo Scientific) and phalloidin (1:100, Molecular Probes) in the blocking solution at 25 °C for 1 h and washed with PBS containing 0.5% Tween 20. The nuclei were stained by using DAPI (1:1000, Molecular Probes) in PBS at 25 °C for 10 min and washed with PBS. The stained cells were mounted onto glass slides and visualized by using a confocal microscope (Nikon). The images were acquired using the same exposure conditions for all groups.
The background of images was identically subtracted for all images by using ImageJ software with a rolling ball radius of 750 pixels. Fluorescent Staining Analyses. To quantify integrin binding as well as cell adhesion and differentiation, immunofluorescently stained images were utilized for the analyses by ImageJ. To quantify the integrin−ligand binding, the fluorescent intensities of integrin β1 staining from six different images were calculated by using the histogram function in ImageJ. To calculate adhered cell density, the number of cell nuclei from 10 different images was counted. To quantify the cell area, F-actin staining from each cell of approximately 30 cells from five different images was used. To determine the number of focal adhesions, vinculin staining greater than 1 μm2 from each cell of approximately 30 cells from five different images was counted by following a previously established procedure.43 To calculate the cell shape factor, F-actin staining from each cell of approximately 30 cells from five different images was used. To determine nuclear localization of the osteogenic marker (RUNX2) and mechanotransduction regulator (YAP), ratiometric analyses of nuclear fluorescence intensity to cytoplasmic fluorescence intensity (excluding nuclei area) were performed from each cell of approximately 30 cells from five different images. The data are presented as mean ± standard errors. Alkaline Phosphatase Staining. The hMSCs were analyzed for RGD-ligand oscillation-frequency-dependent osteogenic differentiation by ALP staining after 7 days of culture under osteogenic induction medium. The hMSCs were fixed in 4% (w/v) paraformaldehyde for 2 min at room temperature, which preserves ALP activity, and washed with PBS. The fixed cells were stained with a mixture of 4% (v/v) naphthol AS-MX phosphate (Sigma-Aldrich) and 0.06% (w/v) Fast Blue RR salt solution (Sigma-Aldrich) in DI water for 45 min at 37 °C under dark conditions and washed with PBS. The stained cells were visualized using a microscope under an A filter in color mode. Subcutaneous Implantation. To examine stem cell adhesion onto oscillating RGD ligands at various frequencies in vivo, 10 3-month-old nude mice were used with the approval of the Institutional Animal Care and Use Committee at the Chinese University of Hong Kong. The nude mice were subject to inhalation anesthesia with 3% isoflurane, and a 2-cm-long incision was made in the back of the mice. The ligand−SPION grafted substrate following BSA coating was subcutaneously implanted, and skin was carefully closed. hMSCs were subcutaneously injected onto the substrate at approximately 20 k/cm2 under an oscillating magnetic field at various frequencies. Following 4 h of stem cell injection, the implanted substrate was collected for immunostaining analyses of cell adhesion and mechanosensing. The in vivo adhered cells were analyzed by immunofluorescent staining (including human nuclei antigen, 1:100, Millipore for identifying human donor cells) through the same procedures described in analyzing in vitro cultured cells. Statistical Analyses. All experiments were repeated at least two times. Using Graphpad Prism 5, statistical significances were assigned for p-values less than 0.05. One-way analysis of variance (ANOVA) with the Tukey−Kramer post hoc test was used to compare multiple groups at the same time point. The two-tailed Student’s t test was used to compare two groups at the same time point.
ASSOCIATED CONTENT S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.7b02857. Characterizations of the bare SPIONs and RGD ligand tethering to the SPIONs, substrate tethering of the RGDligand-bearing SPIONs, characterization setup of nanoscale motion of the substrate-tethered RGD-bearing SPION by in situ AFM or STEM, cellular adhesion control experiments with different oscillation frequencies of the magnetic field, PEG blocking method, the absence of conjugated RGD ligand, conjugated RAD or RGE, or pure silica nanoparticle, focal adhesion analysis, immunofluorescent staining for TAZ and YAP (PDF) 9647
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano
and Focal Adhesion Assembly: A Close Relationship Studied Using Elastic Micropatterned Substrates. Nat. Cell Biol. 2001, 3, 466−472. (10) Zhu, Y.; Qiu, H.; Trzeciakowski, J. P.; Sun, Z.; Li, Z.; Hong, Z.; Hill, M. A.; Hunter, W. C.; Vatner, D. E.; Vatner, S. F. Temporal Analysis of Vascular Smooth Muscle Cell Elasticity and Adhesion Reveals Oscillation Waveforms That Differ with Aging. Aging Cell 2012, 11, 741−750. (11) Robertus, J.; Browne, W. R.; Feringa, B. L. Dynamic Control Over Cell Adhesive Properties Using Molecular-Based Surface Engineering Strategies. Chem. Soc. Rev. 2010, 39, 354−378. (12) Mager, M. D.; LaPointe, V.; Stevens, M. M. Exploring and Exploiting Chemistry at the Cell Surface. Nat. Chem. 2011, 3, 582−589. (13) Stevens, M. M.; George, J. H. Exploring and Engineering the Cell Surface Interface. Science 2005, 310, 1135−1138. (14) Pan, G.; Guo, Q.; Ma, Y.; Yang, H.; Li, B. Thermo-Responsive Hydrogel Layers -Imprinted with RGDS Peptide: A System for Harvesting Cell Sheets. Angew. Chem., Int. Ed. 2013, 52, 6907−6911. (15) Seo, J.-H.; Kakinoki, S.; Inoue, Y.; Yamaoka, T.; Ishihara, K.; Yui, N. Inducing Rapid Cellular Response on RGD-Binding Threaded Macromolecular Surfaces. J. Am. Chem. Soc. 2013, 135, 5513−5516. (16) Dobson, J. Remote Control of Cellular Behaviour with Magnetic Nanoparticles. Nat. Nanotechnol. 2008, 3, 139−143. (17) Ng, C. C. A.; Magenau, A.; Ngalim, S. H.; Ciampi, S.; Chockalingham, M.; Harper, J. B.; Gaus, K.; Gooding, J. J. Using An Electrical Potential to Reversibly Switch Surfaces Between Two States for Dynamically Controlling Cell Adhesion. Angew. Chem., Int. Ed. 2012, 51, 7706−7710. (18) Yeo, W.-S.; Yousaf, M. N.; Mrksich, M. Dynamic Interfaces Between Cells and Surfaces: Electroactive Substrates That Sequentially Release and Attach Cells. J. Am. Chem. Soc. 2003, 125, 14994−14995. (19) Li, L.; Wu, J.; Gao, C. Gradient Immobilization of A Cell Adhesion RGD Peptide on Thermal Responsive Surface for Regulating Cell Adhesion and Detachment. Colloids Surf., B 2011, 85, 12−18. (20) Roberts, J. N.; Sahoo, J. K.; McNamara, L. E.; Burgess, K. V.; Yang, J.; Alakpa, E. V.; Anderson, H. J.; Hay, J.; Turner, L.-A.; Yarwood, S. J. Dynamic Surfaces for the Study of Mesenchymal Stem Cell Growth Through Adhesion Regulation. ACS Nano 2016, 10, 6667−6679. (21) Ohmuro-Matsuyama, Y.; Tatsu, Y. Photocontrolled Cell Adhesion on a Surface Functionalized with a Caged Arginine-GlycineAspartate Peptide. Angew. Chem., Int. Ed. 2008, 47, 7527−7529. (22) Petersen, S.; Alonso, J. M.; Specht, A.; Duodu, P.; Goeldner, M.; del Campo, A. Phototriggering of Cell Adhesion by Caged Cyclic RGD Peptides. Angew. Chem., Int. Ed. 2008, 47, 3192−3195. (23) Lee, T. T.; García, J. R.; Paez, J. I.; Singh, A.; Phelps, E. A.; Weis, S.; Shafiq, Z.; Shekaran, A.; Del Campo, A.; García, A. J. Light-Triggered In Vivo Activation of Adhesive Peptides Regulates Cell Adhesion, Inflammation and Vascularization of Biomaterials. Nat. Mater. 2015, 14, 352−360. (24) Li, W.; Wang, J.; Ren, J.; Qu, X. Near-Infrared Upconversion Controls Photocaged Cell Adhesion. J. Am. Chem. Soc. 2014, 136, 2248−2251. (25) Wirkner, M.; Alonso, J. M.; Maus, V.; Salierno, M.; Lee, T. T.; García, A. J.; Del Campo, A. Triggered Cell Release from Materials Using Bioadhesive Photocleavable Linkers. Adv. Mater. 2011, 23, 3907− 3910. (26) Li, W.; Chen, Z.; Zhou, L.; Li, Z.; Ren, J.; Qu, X. Noninvasive and Reversible Cell Adhesion and Detachment Via Single-Wavelength NearInfrared Laser Mediated Photoisomerization. J. Am. Chem. Soc. 2015, 137, 8199−8205. (27) Liu, D.; Xie, Y.; Shao, H.; Jiang, X. Using Azobenzene-Embedded Self-Assembled Monolayers To Photochemically Control Cell Adhesion Reversibly. Angew. Chem., Int. Ed. 2009, 48, 4406−4408. (28) Colombo, M.; Carregal-Romero, S.; Casula, M. F.; Gutiérrez, L.; Morales, M. P.; Böhm, I. B.; Heverhagen, J. T.; Prosperi, D.; Parak, W. J. Biological Applications of Magnetic Nanoparticles. Chem. Soc. Rev. 2012, 41, 4306−4334. (29) Carregal-Romero, S.; Guardia, P.; Yu, X.; Hartmann, R.; Pellegrino, T.; Parak, W. J. Magnetically Triggered Release of Molecular
Movies for the oscillatory motion of the RGD-bearing SPIONs (AVI) (AVI) (AVI) (AVI)
AUTHOR INFORMATION Corresponding Authors
*E-mail: v-dravid@northwestern.edu. *E-mail: lbian@mae.cuhk.edu.hk. ORCID
Kongchang Wei: 0000-0002-6555-2768 Vinayak P. Dravid: 0000-0002-6007-3063 Liming Bian: 0000-0003-4739-0918 Author Contributions ⊥
H. Kang and D. S. H. Wong contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS This work was supported by a General Research Fund grant from the Research Grants Council of Hong Kong (Project No. 14202215), the National Natural Science Foundation of China (Project No. 31570979), and the Health and Medical Research Fund, the Food and Health Bureau, the Government of the Hong Kong Special Administrative Region (Reference No. 03140056). This research was also supported by project BME-p3-15 of the Shun Hing Institute of Advanced Engineering and the Chow Yuk Ho Technology Centre for Innovative Medicine, The Chinese University of Hong Kong. This work was performed at the EPIC facility of Northwestern University’s NUANCE Center with the support from the Soft and Hybrid Nanotechnology Experimental (SHyNE) Resource (NSF ECCS-1542205); the MRSEC program (NSF DMR-1121262) at the Materials Research Center; the International Institute for Nanotechnology (IIN); the Keck Foundation; and the State of Illinois, through the IIN. REFERENCES (1) Wolfenson, H.; Lavelin, I.; Geiger, B. Dynamic Regulation of the Structure and Functions of Integrin Adhesions. Dev. Cell 2013, 24, 447− 458. (2) Rozario, T.; DeSimone, D. W. The Extracellular Matrix in Development and Morphogenesis: A Dynamic View. Dev. Biol. 2010, 341, 126−140. (3) Evans, E. A.; Calderwood, D. A. Forces and Bond Dynamics in Cell Adhesion. Science 2007, 316, 1148−1153. (4) Ruoslahti, E.; Pierschbacher, M. D. New Perspectives in Cell Adhesion: RGD and Integrins. Science 1987, 238, 491−497. (5) Jalali, S.; del Pozo, M. A.; Chen, K.-D.; Miao, H.; Li, Y.-S.; Schwartz, M. A.; Shyy, J. Y.-J.; Chien, S. Integrin-Mediated Mechanotransduction Requires Its Dynamic Interaction with Specific Extracellular Matrix (ECM) Ligands. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 1042−1046. (6) Kim, S.-H.; Turnbull, J.; Guimond, S. Extracellular Matrix and Cell Signalling: The Dynamic Cooperation of Integrin, Proteoglycan and Growth Factor Receptor. J. Endocrinol. 2011, 209, 139−151. (7) Zhang, Y.; Ge, C.; Zhu, C.; Salaita, K. DNA-Based Digital Tension Probes Reveal Integrin Forces During Early Cell Adhesion. Nat. Commun. 2014, 5.516710.1038/ncomms6167 (8) Rossier, O.; Octeau, V.; Sibarita, J.-B.; Leduc, C.; Tessier, B.; Nair, D.; Gatterdam, V.; Destaing, O.; Albiges-Rizo, C.; Tampé, R. Integrins β1 and β3 Exhibit Distinct Dynamic Nanoscale Organizations Inside Focal Adhesions. Nat. Cell Biol. 2012, 14, 1057−1067. (9) Balaban, N. Q.; Schwarz, U. S.; Riveline, D.; Goichberg, P.; Tzur, G.; Sabanay, I.; Mahalu, D.; Safran, S.; Bershadsky, A.; Addadi, L. Force 9648
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649
Article
ACS Nano Cargo from Iron Oxide Nanoparticle Loaded Microcapsules. Nanoscale 2015, 7, 570−576. (30) Corten, C. C.; Urban, M. W. Repairing Polymers Using Oscillating Magnetic Field. Adv. Mater. 2009, 21, 5011−5015. (31) Rössler, K.; Donat, M.; Lanzenberger, R.; Novak, K.; Geissler, A.; Gartus, A.; Tahamtan, A.; Milakara, D.; Czech, T.; Barth, M. Evaluation of Preoperative High Magnetic Field Motor Functional MRI (3 T) in Glioma Patients by Navigated Electrocortical Stimulation and Postoperative Outcome. J. Neurol., Neurosurg. Psychiatry 2005, 76, 1152− 1157. (32) Wong, D. S.; Li, J.; Yan, X.; Wang, B.; Li, R.; Zhang, L.; Bian, L. Magnetically Tuning Tether Mobility of Integrin Ligand Regulates Adhesion, Spreading, and Differentiation of Stem Cells. Nano Lett. 2017, 17, 1685−1695. (33) Yan, X.; Zhou, Q.; Yu, J.; Xu, T.; Deng, Y.; Tang, T.; Feng, Q.; Bian, L.; Zhang, Y.; Ferreira, A. Magnetite Nanostructured Porous Hollow Helical Microswimmers for Targeted Delivery. Adv. Funct. Mater. 2015, 25, 5333−5342. (34) Larsen, E. K.; Nielsen, T.; Wittenborn, T.; Birkedal, H.; VorupJensen, T.; Jakobsen, M. H.; Østergaard, L.; Horsman, M. R.; Besenbacher, F.; Howard, K. A. Size-Dependent Accumulation of PEGylated Silane-Coated Magnetic Iron Oxide Nanoparticles in Murine Tumors. ACS Nano 2009, 3, 1947−1951. (35) Ma, X.; Hahn, K.; Sanchez, S. Catalytic Mesoporous Janus Nanomotors for Active Cargo Delivery. J. Am. Chem. Soc. 2015, 137, 4976−4979. (36) Attwood, S. J.; Cortes, E.; Haining, A. W. M.; Robinson, B.; Li, D.; Gautrot, J.; del Río Hernández, A. Adhesive Ligand Tether Length Affects the Size and Length of Focal Adhesions and Influences Cell Spreading and Attachment. Sci. Rep. 2016, 610.1038/srep34334. (37) Maheshwari, G.; Brown, G.; Lauffenburger, D. A.; Wells, A.; Griffith, L. G. Cell Adhesion and Motility Depend on Nanoscale RGD Clustering. J. Cell Sci. 2000, 113, 1677−1686. (38) Choi, C. K. K.; Xu, Y. J.; Wang, B.; Zhu, M.; Zhang, L.; Bian, L. Substrate Coupling Strength of Integrin-Binding Ligands Modulates Adhesion, Spreading, and Differentiation of Human Mesenchymal Stem Cells. Nano Lett. 2015, 15, 6592−6600. (39) Ye, K.; Wang, X.; Cao, L.; Li, S.; Li, Z.; Yu, L.; Ding, J. Matrix Stiffness and Nanoscale Spatial Organization of Cell-Adhesive Ligands Direct Stem Cell Fate. Nano Lett. 2015, 15, 4720−4729. (40) Huang, J.; Grater, S. V.; Corbellini, F.; Rinck, S.; Bock, E.; Kemkemer, R.; Kessler, H.; Ding, J.; Spatz, J. P. Impact of Order and Disorder in RGD Nanopatterns on Cell Adhesion. Nano Lett. 2009, 9, 1111−1116. (41) Xiong, J.-P.; Stehle, T.; Diefenbach, B.; Zhang, R.; Dunker, R.; Scott, D. L.; Joachimiak, A.; Goodman, S. L.; Arnaout, M. A. Crystal Structure of the Extracellular Segment of Integrin αVβ3. Science 2001, 294, 339−345. (42) Xiong, J.-P.; Stehle, T.; Zhang, R.; Joachimiak, A.; Frech, M.; Goodman, S. L.; Arnaout, M. A. Crystal Structure of the Extracellular Segment of Integrin αVβ3 in Complex with An Arg-Gly-Asp Ligand. Science 2002, 296, 151−155. (43) Deeg, J. A.; Louban, I.; Aydin, D.; Selhuber-Unkel, C.; Kessler, H.; Spatz, J. P. Impact of Local Versus Global Ligand Density on Cellular Adhesion. Nano Lett. 2011, 11, 1469−1476. (44) Bian, L.; Guvendiren, M.; Mauck, R. L.; Burdick, J. A. Hydrogels That Mimic Developmentally Relevant Matrix and N-cadherin Interactions Enhance MSC Chondrogenesis. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 10117−10122. (45) Cha, C.; Liechty, W. B.; Khademhosseini, A.; Peppas, N. A. Designing Biomaterials To Direct Stem Cell Fate. ACS Nano 2012, 6, 9353−9358. (46) Gaharwar, A. K.; Mihaila, S. M.; Swami, A.; Patel, A.; Sant, S.; Reis, R. L.; Marques, A. P.; Gomes, M. E.; Khademhosseini, A. Bioactive Silicate Nanoplatelets for Osteogenic Differentiation of Human Mesenchymal Stem Cells. Adv. Mater. 2013, 25, 3329. (47) Xavier, J. R.; Thakur, T.; Desai, P.; Jaiswal, M. K.; Sears, N.; Cosgriff-Hernandez, E.; Kaunas, R.; Gaharwar, A. K. Bioactive
nanoengineered hydrogels for bone tissue engineering: a growthfactor-free approach. ACS Nano 2015, 9, 3109−3118. (48) Kang, H.; Shih, Y.-R. V.; Nakasaki, M.; Kabra, H.; Varghese, S. Small Molecule−Driven Direct Conversion of Human Pluripotent Stem Cells into Functional Osteoblasts. Sci. Adv. 2016, 2, e1600691. (49) Xu, J.; Li, J.; Lin, S.; Wu, T.; Huang, H.; Zhang, K.; Sun, Y.; Yeung, K. W.; Li, G.; Bian, L. Nanocarrier-Mediated Codelivery of Small Molecular Drugs and siRNA To Enhance Chondrogenic Differentiation and Suppress Hypertrophy of Human Mesenchymal Stem Cells. Adv. Funct. Mater. 2016, 26, 2463−2472. (50) Min, J.; Choi, K. Y.; Dreaden, E. C.; Padera, R. F.; Braatz, R. D.; Spector, M.; Hammond, P. T. Designer Dual Therapy Nanolayered Implant Coatings Eradicate Biofilms and Accelerate Bone Tissue Repair. ACS Nano 2016, 10, 4441−4450. (51) Choi, C. K. K.; Li, J.; Wei, K.; Xu, Y. J.; Ho, L. W. C.; Zhu, M.; To, K. K.; Choi, C. H. J.; Bian, L. A Gold@Polydopamine Core−Shell Nanoprobe for Long-Term Intracellular Detection of MicroRNAs in Differentiating Stem Cells. J. Am. Chem. Soc. 2015, 137, 7337−7346. (52) McBeath, R.; Pirone, D. M.; Nelson, C. M.; Bhadriraju, K.; Chen, C. S. Cell Shape, Cytoskeletal Tension, and RhoA Regulate Stem Cell Lineage Commitment. Dev. Cell 2004, 6, 483−495. (53) Mahmoudi, M.; Bonakdar, S.; Shokrgozar, M. A.; Aghaverdi, H.; Hartmann, R.; Pick, A.; Witte, G.; Parak, W. J. Cell-Imprinted Substrates Direct the Fate of Stem Cells. ACS Nano 2013, 7, 8379−8384. (54) Connelly, J. T.; Gautrot, J. E.; Trappmann, B.; Tan, D. W.-M.; Donati, G.; Huck, W. T.; Watt, F. M. Actin and Serum Response Factor Transduce Physical Cues from the Microenvironment To Regulate Epidermal Stem Cell Fate Decisions. Nat. Cell Biol. 2010, 12, 711−718. (55) Bloemen, M.; Brullot, W.; Luong, T. T.; Geukens, N.; Gils, A.; Verbiest, T. Improved Functionalization of Oleic Acid-Coated Iron Oxide Nanoparticles for Biomedical Applications. J. Nanopart. Res. 2012, 14, 1100. (56) Yang, H.; Chen, Z.; Zhang, L.; Yung, W. Y.; Leung, K. C. F.; Chan, H. Y. E.; Choi, C. H. J. Mechanism for the Cellular Uptake of Targeted Gold Nanorods of Defined Aspect Ratios. Small 2016, 12, 5178−5189.
9649
DOI: 10.1021/acsnano.7b02857 ACS Nano 2017, 11, 9636−9649