Freezing or Wrapping: The Role of Particle Size in the Mechanism of

Purchase temporary access to this content. ACS Members purchase additional access options · Ask your library to provide you and your colleagues site-w...
1 downloads 4 Views 4MB Size
Article pubs.acs.org/Langmuir

Freezing or Wrapping: The Role of Particle Size in the Mechanism of Nanoparticle−Biomembrane Interaction Shengwen Zhang, Andrew Nelson,* and Paul A. Beales* Centre for Molecular Nanoscience, School of Chemistry, University of Leeds, Leeds, U.K. S Supporting Information *

ABSTRACT: Understanding the interactions between nanoparticles (NPs) and biological matter is a high-priority research area because of the importance of elucidating the physical mechanisms underlying the interactions leading to NP potential toxicity as well as NP viability as therapeutic vectors in nanomedicine. Here, we use two model membrane systems, giant unilamellar vesicles (GUVs) and supported monolayers, to demonstrate the competition between adhesion and elastic energy at the nanobio interface, leading to different mechanisms of NP− membrane interaction relating to NP size. Small NPs (18 nm) cause a “freeze effect” of otherwise fluid phospholipids, significantly decreasing the phospholipid lateral mobility. The release of tension through stress-induced fracture mechanics results in a single microsize hole in the GUVs after interaction. Large particles (>78 nm) promote membrane wrapping, which leads to increased lipid lateral mobility and the eventual collapse of the vesicles. Electrochemical impedance spectroscopy on the supported monolayer model confirms that differently sized NPs interact differently with the phospholipids in close proximity to the electrode during the lipid desorption process. The time scale of these processes is in accordance with the proposed NP/GUV interaction mechanism.

1. INTRODUCTION With the advance of nanoparticle (NP) engineering, intensive investigations into their potential industrial and biomedical applications have been initiated.1−3 Various studies using NPs to assist the intracellular delivery and controlled release of drug molecules have shown encouraging results.4,5 However, significant safety concerns have arisen, with supporting findings from cytotoxicity studies.3,6,7 Thus, the issue addressing how NPs interact with biomembranes has become a pressing concern.1,3 Whether NPs are seen as the therapeutic solution to enhanced drug delivery or a potential health threat, the mechanisms behind the NP−biomembrane interactions are yet to be fully established. The complexities of cell membranes and the diverse properties of NPs make it extremely challenging to study the interactions between the plasma membrane and NPs. On the cellular level, NPs can interfere with cell signaling, affect gene expression, induce oxidative stress, and cause structural damage to membranes or organelles. Because of the complex properties of NPs and diverse cell responses, cytotoxicity studies of NPs sometimes lead to contradictory results.8 For example, a study of Au NPs has identified 50 nm as the most efficient size for cellular uptake with HeLa cells9 whereas another study of Au NPs using connective tissue fibroblasts, epithelial cells, macrophages, and melanoma cells has suggested 1.4 nm to be the most effective size.10 Shape is another intrinsic parameters of NPs, which sparks debate. Rodlike particles have been reported to internalize faster into HeLa cells compared to spherical particles of the same volume.11 Another study, also using HeLa cells, has stated that © 2012 American Chemical Society

the uptake of rod-shaped Au NPs was much less than that of spherical ones.12 Nevertheless, it is generally agreed that the local geometry of a particle at the point of contact with the cell membrane determines the rate of internalization.13 In addition, surface charge and surface modifications of NPs also play significant roles in determining the interaction of NPs with cell membranes.14 Asymmetry in the particle structure or surface texture can also be important: it has been reported that 6 nm gold NPs with alternating anionic and hydrophobic ligand coatings can penetrate cells at both 37 and 4 °C without disrupting the membrane,15 which has been attributed to the lowering of the energy barrier for translocation across the membrane.16 This work takes a reductionist approach to studying the interactions between spherical SiO2 NPs with narrow size distributions and dioleoyl phosphatidylcholine (DOPC) model membranes at physiological pH. It has been reported that SiO2 NPs adsorb on the biomembrane surface17 and SiO2 NPs across a broad range of particle sizes perturb the permeability barrier of biomembranes, even at very low concentrations.18 Two complementary model membrane systems, giant unilamellar vesicles (GUVs) and supported monolayers, are used in this work. Free-standing bilayer GUVs are of a similar size and morphology to native cells and facilitate microscale imaging and analysis,19−21 whereas supported monolayers allow detailed electrochemical investigations on the molecular level.22 DOPC is Received: April 30, 2012 Revised: June 20, 2012 Published: June 20, 2012 12831

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837

Langmuir

Article

Figure 1. Confocal laser scanning microscopy of DOPC GUVs after NP interactions: (a−d) GUVs after 20 min of interaction with 18 nm SiO2. (a, b) Reconstituted 3D images of GUVs with unusual curvature and stabilized holes; (c) a helmet-shaped GUV; (d) 2D confocal image of deformed GUVs and dextran fluorescence leakage after interaction with 18 nm 25 μg mL−1 SiO2 NPs; (e) 3D reconstruction of a GUV interacting with 182 nm SiO2 NPs; (f) schematic view of our interpretation of the effect of size on the NP−membrane interactions, where small NPs adsorb to the membrane (top) and the membrane wraps the large NPs (bottom). Depth profiles are shown in a−c and e. Insets of c: 2D and 3D view of the same GUV. Insets of d, additional deformed vesicles from the same experiment. Inset of e: 2D view of a GUV undergoing a wrapping event.

DOPC monolayers have been reported to undergo a series of well-defined and reversible electric-field-induced reorientations (Figure SI-1).22,24−26 Monitoring how these structural transitions are influenced by extraneous stimuli gives insight into the interactions at this nanobio interface.

used as the sole component of the biomembranes, eliminating possible complexities arising from compositionally distinct domains in multicomponent mixtures. DOPC has a melting temperature23 of ∼−20 °C and hence is in a fluid, physiologically relevant state at room temperature. Moreover, Hg-supported 12832

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837

Langmuir

Article

transformed with Excel (Microsoft) to the complex capacitance plane by plotting −C′′sp versus C′sp in units of μF cm−2. −C′′sp versus C′sp are the imaginary and real specific capacitances, respectively, where the specific capacitance (Csp) is the capacitance (C) divided by the electrode area. All of the electrochemical measurements were conducted in a Faraday cage. At the desorption potential, the RC time constants were obtained from the impedance data: RC = 1/(2πf1), with f1 being the characteristic frequency for this process where the imaginary capacitance reaches a maximum value. The zero-frequency capacitance (ZFC) is the measured double-layer capacitance of the electrode−electrolyte interface extrapolated to zero ac frequency. The ZFC can be calculated from C = RC/Ru, where Ru is the uncompensated solution resistance. The low-frequency relaxation processes of DOPC desorption with and without the presence of NPs were analyzed with an impedance model developed for this system,26,30,31 from which the relaxation time constant τ was calculated.

2. EXPERIMENTAL SECTION 2.1. Materials. DOPC and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (18:1 Liss Rhod PE) were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). Dextran (Alexa Fluor 647, 10 000 MW) was obtained from Invitrogen Molecular Probes. Colloidal SiO2 NPs (Ludox SM-30, 18 nm diameter, 30 wt % suspension in water) were purchased from Sigma-Aldrich Ltd. Colloidal SiO2 microspheres (78 nm) were purchased from Polyscience, Inc. as 5 wt % aqueous dispersion. AngstromSphere SiO2 particles (180 nm, dry form) were purchased from Fiber Optic Center Inc. and dispersed in water according to the technique provided by the supplier. Particle sizes were measured by dynamic light scattering (DLS) in both 100 mM KCl + 10 mM PBS and 165 mM NaCl + 10 mM HEPES electrolytes at a particle concentration of 500 μg mL−1 (Zetasizer Nano S, Malvern Instruments Ltd.). The SiO2 colloidal systems are stable for at least 3 days. The DLS results are presented in Supporting Information Figure SI-2. All experiments were carried out at solution pH 7.4 where the SiO2 particles have a negative surface charge of ∼2 μC cm2.27 2.2. Methods. Confocal laser scanning microscopy and fluorescence recovery after photobleaching (FRAP) measurements were conducted on DOPC GUVs. GUVs were prepared by the electroformation technique in a 300 mM sucrose solution described previously from DOPC with 0.5 mol % fluorophore (Liss Rhod PE).21 The experiment was conducted at room temperature. A Zeiss LSM T-PMT/LSM700 confocal laser scanning inverted microscope and Zeiss ZEN software were used for the experiments. Glass-bottom culture dishes (MatTek), pretreated with a 10% BSA solution to prevent GUVs adhering to the glass, were used for microscopy. A pH 7.4 microscope buffer (10 mM 4(2-hydroxyethyl)-1-piperazineethanesulfonic acid +165 mM NaCl) was added such that GUVs sedimented to the bottom of sample and sat on the coverslip. The GUVs observed in this study were in the range of ∼4− 20 μm (diameter). No variation in biomembrane responses was observed across this size range. FRAP data fitting was performed with IGOR Pro (Wavemetrics, Inc.). Electrochemical studies were performed on DOPC monolayers supported on mercury film electrodes (MFEs). The fabrication of the wafer-based MFEs had been described previously.26,28,29 All potentials in this article were quoted versus the potential of a Ag/AgCl/3.5 mol dm−3 KCl reference electrode separated from the electrolyte by a porous glass frit and a Pt counter electrode (Metrohm U.K. Ltd.). Supported DOPC monolayers on MFEs were developed by initially spreading 13 μL of a 2.54 mmol dm−3 solution of DOPC in pentane (HPLC grade, Fisher Scientific Chemicals Ltd.) at the argon−electrolyte interface in the electrochemical cell, followed by slowly lowering the MFE through the phospholipid on the electrolyte interface.26,30,31 As a result, the DOPC layer is transferred to the electrode when it is present in >2-fold excess surface coverage at the gas−water interface. It is well established that a phospholipid monolayer is always transferred to the electrode irrespective of the lipid coverage at the gas−water interface.32 The DOPC coverage at the gas−water interface will therefore have a collapse pressure of ∼45 mN m−1.33 For a deposited DOPC monolayer on Hg, the surface pressure at the mercury−electrolyte interface is 52 mN m−1 at the potential of zero charge (∼ −0.375 V).22 This surface pressure is potential-dependent and approaches zero at a lipid desorption potential of −1.3 to −1.5 V.22 Accordingly, the lipid desorption potential can be defined as when the surface pressure, or adhesion energy, of DOPC on Hg decreases to zero22 and becomes insufficient to retain the lipids on the mercury surface. The integrity of the monolayer on Hg was confirmed with rapid cyclic voltammetry between −0.2 and −1.6 V at 40 V s−1.26 The KCl (0.1 mol dm−3) electrolyte was prepared from Analar KCl (Fisher Chemicals Ltd.) calcined at 600 °C for 5 h and dissolved in 18.2 MΩ Milli-Q water with added 0.01 mol dm−3 phosphate buffer (pH 7.4). The impedance spectroscopy measurements were carried out by applying logarithmically distributed frequencies of between 65 000 and 0.1 Hz at −0.35 V, the position of zero charge (PZC) for Hg, and −1.32 V (the DOPC desorption potential) using Autolab (Ecochemie, Utrecht, Netherlands).26,30,31 An ac amplitude of 0.002 V was applied during the measurements. The impedance data obtained were then

3. RESULTS AND DISCUSSION 3.1. NP−Biomembrane Interactions on GUVs. The morphology of the GUVs is studied by confocal microscopy. The morphology changes of GUVs after interacting with SiO2 NPs are remarkably dependent on the size of the NPs. Significantly, 18 nm NPs create unusual crinkles and permanent holes in GUVs (Figure 1a−d), transforming the previously smooth, spherical GUVs into crumpled “paper bags” with microscale openings. These unusual holes and curvatures on the GUVs are also shown in 2D (Figure 1d). Fluorescent dextran ( 10 kDa) was added to these samples and was observed to enter deformed GUVs immediately, confirming that these are indeed holes, not dye-excluding lipid domains. In rare events, the spherical GUV buckles into a “helmet” morphology without forming a micropore (Figure 1c). Nonetheless, crumpled “pot”shaped vesicles with a single micropore are more commonly observed. In direct contrast, the outcome of the interaction between GUVs and 182 nm particles is dramatically different. During the initial interaction, wrapping events can be clearly observed on the surfaces of GUVs (Figure 1e). These wrapping events deplete the lipids and eventually lead to GUV breakdown. This striking size dependence of the NP−biomembrane interaction can also be observed in their distinctly different effects on lipid lateral mobility. Fluorescence recovery after photobleaching (FRAP) is employed to investigate the modifications in lipid lateral mobility.34 Diffusion coefficients, D, are calculated on the basis of the mobility of lipid-based fluorescent probes within the bilayer of GUVs. Normalized fluorescence recovery curves are presented in Figure 2. The FRAP data are fitted to a classic fluorescence recovery model35,36 ⎧ ⎛ 2τ ⎞⎡ ⎛ 2τ ⎞ ⎛ 2τ ⎞⎤⎫ f (t ) = A ⎨exp⎜ − D ⎟⎢I0⎜ D ⎟ + I1⎜ D ⎟⎥⎬ ⎝ t ⎠⎦⎭ ⎩ ⎝ t ⎠⎣ ⎝ t ⎠

(1)

in which t is time, A is the recovery level, τD is characteristic recovery time, and I0 and I1 are modified Bessel functions of the first kind. Diffusion coefficient is calculated from D = w2/4τD (w is the radius of the circular bleached area).36 The good fits of this model to our data in Figure 2 suggest that the intramembrane molecular dynamics of DOPC with and without the presence of NPs are well described by a single characteristic diffusion coefficient. Attempts to fit the fluorescence recovery using two coexisting diffusion coefficients within the membrane did not improve the fits. This confirms that, on microscopic length scales, the GUV bilayers display uniform lipid dynamics, suggesting that the membrane consists of a single homogeneous phase on these length scales. Mean values of the diffusion coefficient are listed in Table 1. The typical 12833

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837

Langmuir

Article

3.2. NP−Biomembrane Interactions on Supported Monolayers. Electrochemical impedance spectroscopy (EIS) is a noninvasive yet extremely sensitive method of studying interactions at the lipid interface.39 In this work, EIS is applied to corroborate the interactions of the DOPC assemblies with SiO2 particles and to gain further insight into the mechanisms on the molecular level. In this classical model of a DOPC monolayer supported on a Hg electrode, the configuration of the assembled phospholipids can be varied by altering the applied electric field (Figure SI-1). Close to the PZC of Hg,22 an intact DOPC monolayer exists with its apolar tail groups physically associated with the apolar Hg surface25,26,29,30 resembling the outer leaflet of phospholipid bilayer vesicles. Such a supported monolayer can be electrically represented by a simple RC circuit of solution resistance Ru and monolayer capacitance, C, in series. Following an EIS measurement, this circuit gives rise to a single RC semicircle in the complex capacitance plot,31 which has a specific ZFC value of ∼1.8 μF cm−2 (Figure 3 a). Some impedance terminologies are illustrated in Figure SI-4. Primary EIS results suggest that SiO2 particles of all sizes adsorb on supported monolayers but do not alter the dielectric constant or thickness of the DOPC monolayer. The monolayer capacitance is not significantly modified after interaction with each of the three SiO2 NP samples (Figure 3a), as is also confirmed by ac voltammetry results (Figure SI-5). However, additional and identical capacitive elements arise at low frequencies after DOPC interaction with all sizes of NPs, showing a similar structural interference of adsorbed NPs with a well-oriented phospholipid monolayer.17,30,31 At the PZC of Hg, DOPC has a surface pressure, equivalent to the adhesion energy of DOPC on Hg, of 52 × 10−3 J m−2.22 This value is much greater than the adhesion energy between SiO2 and phosphatidylcholine, which has been measured to be 0.5 to 1 × 10−3 J m−2.40 Thus, particles of all sizes adsorb on the DOPC layer on Hg at this potential. Crucially, the size dependence of the NP−biomembrane interaction becomes evident near the lipid desorption potential, which is initiated at −1.3 V and is complete at ∼ −1.5 V for supported DOPC on Hg.22 The initiation of lipid desorption at −1.3 V where the DOPC adhesion energy on Hg is ∼4 × 10−3 J m−2 s22 allows the lipids to interact freely with the NPs yet still remain physically in the vicinity of the electrode.25 In addition, at −1.3 V the negative potential on the Hg surface is screened by the counterions in the electrolyte,24 facilitating any interaction between the negatively charged SiO2 NP-coated27,41 bilayer patches and the electrolyte-coated mercury surface. At high applied ac frequencies, the lipids (and NPs) are outside the double layer of the electrode, giving a specific ZFC of ∼20 μF cm−2, which is characteristic of an electrolyte-coated Hg electrode,22,26 as shown in Figure 3b.4. At low applied ac frequencies, the movement of the lipids becomes coupled with the fluctuating electric field and is shown as a low-frequency relaxation adjacent to the RC semicircle. It is significant that the adsorbed NPs substantially alter this low-frequency relaxation process and the modification is dependent on the NP diameter (Figure 3b). Relaxation time constants, τ, at low frequency are presented in Figure 3b.5. The τ for the DOPC relaxation is about 2.1 ms. The presence of NPs significantly modifies this relaxation time constant to 0.31, 2.5, and 24.4 ms as estimated in the presence of 18, 78, and 182 nm NPs, respectively. The lipid desorption potential is not altered in the presence of particles adsorbed on the DOPC shown in the ac voltammetry results (Supporting Information Figure SI-5). This indicates that

Figure 2. Representative FRAP recovery curves of DOPC in the presence of 18 nm SiO2, 78 nm SiO2, and 182 nm SiO2 with a particle concentration of 500 μg mL−1. Recovery curves are fit to eq 1 for each data set presented (solid lines). The inset presents the initial recoveries and the fittings.

recovery level is between 90 and 95%, which we attribute to bleaching of the fluorophores. Table 1. Mean Diffusion Coefficients of DOPC in GUVs before and after the Interactions with SiO2 NPs (Data Sets Ranging from 6 to 12 GUVs per Measurement) sample (nm)

NP concentration (μg mL−1)

D (μm2 s−1)

SD ±

GUV 18

0 5 25 50 500 50 500 50 500

3.10 3.12 0.61 0.18 0.25 3.01 5.64 2.66 7.14

0.34 0.64 0.72 0.12 0.14 0.37 1.80 0.36 2.40

78 182

We find that small NPs substantially decrease the lipid mobility, whereas larger NPs lead to a discernible increase in lateral diffusion in the membrane. The average lateral diffusion coefficient (D) in DOPC GUVs is 3.1 ± 0.34 μm2 s−1 (Table 1) and is consistent with literature values.37,38 We observe a negligible change in D when the GUVs are exposed to SiO2 NPs at low concentrations (e.g., 5 μg mL−1 for 18 nm NPs and 50 μg mL−1 for 78 and 182 nm particles). Significantly, D decreases when increasing the concentration of 18 nm SiO2 NPs. This decrease is also dependent on the incubation time, suggesting a progressive interaction, and is ∼45 times lower after 60 min of interaction with 18 nm NPs (Figure SI-3). This progressive decrease in lipid mobility suggests that the freeze effect does not proceed by a sharp first-order phase transition as would be expected for a thermodynamic fluid-to-gel phase transformation in a membrane crossing its Tm. Notably, D increases after DOPC membranes are exposed to high doses of 182 and 78 nm NPs. The increased mobility is attributed to defects in the GUV membrane created by the wrapping mechanism. Moreover, the substantially larger SD suggests that these disorganized lipids are far from equilibrium. 12834

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837

Langmuir

Article

Figure 3. Impedance spectroscopy presented in the complex capacitance plane of a supported DOPC monolayer on Hg in 0.1 mol dm−3 KCl with 0.01 mol dm−3 PBS buffer at pH 7.4 before and after 50 min of interaction with 500 μg mL−1 SiO2 NPs at (a) −0.35 V and (b) −1.32 V for (b.1) DOPC, (b.2) DOPC with 18 nm SiO2, (b.3) DOPC with 78 nm SiO2, and (b.4) DOPC with 78 nm SiO2. The insets display the fit to the impedance model (red line), from which the low-frequency relaxation time constants are obtained. The summary of relaxation time constants (black triangle) and extrapolated specific ZFCs (blue dots) are presented in b.5. Each error bar marks the SD from a set of three experimental results.

results are consistent with our findings from the GUV experiments. In general, the particle size, shape, charge density, surface chemistry, and crystallinity determine how particles interact with biomembranes.2 In this current system, we interpret our observations to arise from the balance between adhesion and elastic energies at the membrane surface.42 Quantitative analysis shows that the bending energy barrier for DOPC membranes around 18 nm NPs is 2 orders of magnitude larger than that for 182 nm NPs (Supporting Information SI-5). Thus, the energy required for membrane bending is the dominant barrier when particles are small,42,43 suppressing membrane wrapping.44 Adhesion arises from van der Waals forces and electrostatic interaction between the negative charges (ca. 2 μC cm−2)27 on the SiO2 surface and the positively charged region of the DOPC P−−N+ dipole, which is believed to alter the tilt angle of the headgroup, condensing the surface area per lipid.45 This

the SiO2 nanoparticles do not stabilize the lipid on the Hg surface. However, the NP affects the kinetics and mechanism of the adsorption/desorption process, which is clearly indicated in the complex capacitance plots. In these plots, the time constant, τ, of the low-frequency relaxation process is associated with the movement of desorbed DOPC assemblies formed when the DOPC bilayer structures leave the electrode. Notably, τ is about 7 times smaller in the presence of 18 nm NPs. This decrease in τ for lipid desorption indicates that the 18 nm particles stabilize the transition state of desorbing bilayer patches within the electrode Debye length. τ is 2 orders of magnitude larger in the presence of 182 nm particles. This can be interpreted as showing that the larger NPs are unable to stabilize the transition state until the lipid has wrapped around the particle. Thus, the extended mechanism of bilayer wrapping hinders the lipid adsorption/ desorption process in the presence of the electric field. These 12835

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837

Langmuir



increased lipid packing density decreases the lateral mobility, resulting in a rigid membrane with a high lateral tension (i.e., a freeze effect). This tension is eventually released through stressinduced fracture mechanics, which creates a single microsized hole in the membrane. The formation of a single opening in GUVs under external stress has been observed previously.46,47 However, such openings are usually transient: the formation of a pore results in the relaxation of tension isotropically throughout the GUV membrane, and the pore is then closed by the line tension at the edge of the opening.46 In this study, however, the microsized openings were stabilized by the 18 nm SiO2 NPs adsorbed on the GUVs because they significantly decreased the lateral mobility of the individual phospholipids in the GUVs. In a recent study with binary GUVs, the solidification of membranes has been shown to cause pores in the presence of surfactants that minimize the line tension created at the rim of the pore.48 Therefore, we suggest that small SiO2 NPs may also function as a “line-actant” to stabilize holes within biomembranes. Because the adhesion energy between SiO2 and phosphatidylcholine membranes is in the range of 0.5 to 1 × 10−3 J m−2,40 we can estimate that bending and adhesive energies equate for a particle size in the range of 28−40 nm (Supporting Information SI-5). For 182 nm particles (clearly outside this range), wrapping is energetically more favorable, as observed by fluorescence microscopy and supported by electrochemical results. NPs of 78 nm diameter interact with DOPC GUVs in a similar but less aggressive manner. A moderate increase in lipid lateral mobility gradually followed by GUV breakdown is observed, which is also attributed to the defects in the GUVs created by the wrapping mechanism. A phospholipid matrix forms the major part of the plasma membrane, implying that NP−phospholipid interactions alone can compromise its barrier function.18 Indeed, recent work with lung alveolar carcinoma cells suggests the nonendocytotic uptake of the very same 18 nm SiO2 NPs at 4 °C, a temperature at which active membrane trafficking through ATP-driven endocytosis is inhibited.49 Therefore, the fundamental NP−biomembrane interaction mechanisms we report here are still applicable to complex biomembrane systems and biological cells and will facilitate further work in both nanomedicine and nanotoxicology.

AUTHOR INFORMATION

Corresponding Author

*(A.N.) Tel: +44 113 343 6409. Fax: +44 113 343 6452. E-mail: [email protected]. (P.A.B.) Tel: +44 113 343 9101. Fax: +44 113 343 6452. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS A.N. and S.Z. are grateful for the Brian Mercer Award for Innovation from the Royal Society (U.K.) for funding this work. P.A.B. thanks the Biomedical and Health Research Centre (BHRC) in Leeds for funding and support. This work is also supported by the ENNSATOX programme funded by EU FP7 under grant agreement no. NMP-229244.



ABBREVIATIONS DOPC, dioleoyl phosphatidylcholine; EIS, electrochemical impedance spectroscopy; FRAP, fluorescence recovery after photobleaching; DLS, dynamic light scattering; GUV, giant unilamellar vesicle; MFE, mercury film electrode; NP, nanoparticle; PZC, position of zero charge; ZFC, zero frequency capacitance



REFERENCES

(1) Nel, A.; Xia, T.; Madler, L.; Li, N. Toxic potential of materials at the nanolevel. Science 2006, 311, 622−627. (2) Nel, A. E.; Madler, L.; Velegol, D.; Xia, T.; Hoek, E. M. V.; Somasundaran, P.; Klaessig, F.; Castranova, V.; Thompson, M. Understanding biophysicochemical interactions at the nano-bio interface. Nat. Mater. 2009, 8, 543−557. (3) Yamashita, K.; Yoshioka, Y.; Higashisaka, K.; Mimura, K.; Morishita, Y.; Nozaki, M.; Yoshida, T.; Ogura, T.; Nabeshi, H.; Nagano, K.; Abe, Y.; Kamada, H.; Monobe, Y.; Imazawa, T.; Aoshima, H.; Shishido, K.; Kawai, Y.; Mayumi, T.; Tsunoda, S.-i.; Itoh, N.; Yoshikawa, T.; Yanagihara, I.; Saito, S.; Tsutsumi, Y. Silica and titanium dioxide nanoparticles cause pregnancy complications in mice. Nat Nano 2011, 6, 321−328. (4) Bernardos, A.; Mondragon, L.; Aznar, E.; Marcos, M. D.; MartinezManez, R.; Sancenon, F.; Soto, J.; Barat, J. M.; Perez-Paya, E.; Guillem, C.; Amoros, P. Enzyme-responsive intracellular controlled release using nanometric silica mesoporous supports capped with “saccharides. ACS Nano 2010, 4, 6353−6368. (5) Vivero-Escoto, J. L.; Slowing, I. I.; Wu, C.-W.; Lin, V. S. Y. Photoinduced intracellular controlled release drug delivery in human cells by gold-capped mesoporous silica nanosphere. J. Am. Chem. Soc. 2009, 131, 3462−3463. (6) Napierska, D.; Thomassen, L. C. J.; Rabolli, V.; Lison, D.; Gonzalez, L.; Kirsch-Volders, M.; Martens, J. A.; Hoet, P. H. Sizedependent cytotoxicity of monodisperse silica nanoparticles in human endothelial cells. Small 2009, 5, 846−853. (7) Tao, Z.; Toms, B. B.; Goodisman, J.; Asefa, T. Mesoporosity and functional group dependent endocytosis and cytotoxicity of silica nanomaterials. Chem. Res. Toxicol. 2009, 22, 1869−1880. (8) Horie, M.; Kato, H.; Fujita, K.; Endoh, S.; Iwahashi, H. In vitro evaluation of cellular response induced by manufactured nanoparticles. Chem. Res. Toxicol. 2011, 25, 605−619. (9) Chithrani, B. D.; Ghazani, A. A.; Chan, W. C. W. Determining the size and shape dependence of gold nanoparticle uptake into mammalian cells. Nano Lett. 2006, 6, 662−668. (10) Pan, Y.; Neuss, S.; Leifert, A.; Fischler, M.; Wen, F.; Simon, U.; Schmid, G.; Brandau, W.; Jahnen-Dechent, W. Size-dependent cytotoxicity of gold nanoparticles. Small 2007, 3, 1941−1949. (11) Gratton, S. E. A.; Ropp, P. A.; Pohlhaus, P. D.; Luft, J. C.; Madden, V. J.; Napier, M. E.; DeSimone, J. M. The effect of particle design on

4. CONCLUSIONS With highly consistent observations on microscopic and molecular length scales, we demonstrate the size-dependent interaction mechanisms between NPs and biomembrane models. Small NPs cause a freeze effect of otherwise fluid phospholipids, significantly decreasing the phospholipid lateral mobility. The tension created is released through stress-induced fracture mechanics, resulting in microsized opening in GUVs. In contrast, large particles promote membrane wrapping, which leads to moderate increases in lipid lateral mobility and the eventual collapse of the vesicles. The size-dependent interaction mechanisms are confirmed electrochemically using the same phospholipids in close proximity to the electrode during the lipid desorption process. The size-dependent mechanisms of the NP− membrane interaction are attributed to the competition between adhesion and elastic energy at the nanobio interface.



Article

ASSOCIATED CONTENT

S Supporting Information *

Details of electrochemical data analysis and model membrane systems. This material is available free of charge via the Internet at http://pubs.acs.org 12836

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837

Langmuir

Article

cellular internalization pathways. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 11613−11618. (12) Chithrani, B. D.; Chan, W. C. W. Elucidating the mechanism of cellular uptake and removal of protein-coated gold nanoparticles of different sizes and shapes. Nano Lett. 2007, 7, 1542−1550. (13) Mitragotri, S.; Lahann, J. Physical approaches to biomaterial design. Nat. Mater. 2009, 8, 15−23. (14) Verma, A.; Stellacci, F. Effect of surface properties on nanoparticle−cell interactions. Small 2010, 6, 12−21. (15) Verma, A.; Uzun, O.; Hu, Y.; Hu, Y.; Han, H.-S.; Watson, N.; Chen, S.; Irvine, D. J.; Stellacci, F. Surface-structure-regulated cellmembrane penetration by monolayer-protected nanoparticles. Nat. Mater. 2008, 7, 588−595. (16) Li, Y.; Li, X.; Li, Z.; Gao, H. Surface-structure-regulated penetration of nanoparticles across a cell membrane. Nanoscale 2012, 4, 3768−3775. (17) Vakourov, A.; Brydson, R.; Nelson, A. Electrochemical modeling of silica nanoparticle−biomembrane interaction. Langmuir 2012, 28, 1246−1255. (18) de Planque, M. R. R.; Aghdaei, S.; Roose, T.; Morgan, H. Electrophysiological characterization of membrane disruption by nanoparticles. ACS Nano 2011, 5, 3599−3606. (19) Beales, P. A.; Bergstrom, C. L.; Geerts, N.; Groves, J. T.; Vanderlick, T. K. Single vesicle observations of the cardiolipincytochrome c interaction: induction of membrane morphology changes. Langmuir 2011, 27, 6107−6115. (20) Beales, P. A.; Vanderlick, T. K. Specific binding of different vesicle populations by the hybridization of membrane-anchored DNA. J. Phys. Chem. A 2007, 111, 12372−12380. (21) Beales, P. A.; Vanderlick, T. K. Partitioning of membraneanchored DNA between coexisting lipid phases. J. Phys. Chem. B 2009, 113, 13678−13686. (22) Bizzotto, D.; Nelson, A. Continuing electrochemical studies of phospholipid monolayers of dioleoyl phosphatidylcholine at the mercury−electrolyte interface. Langmuir 1998, 14, 6269−6273. (23) Wittmaack, K. Novel dose metric for apparent cytotoxicity effects generated by in vitro cell exposure to silica nanoparticles. Chem. Res. Toxicol. 2010, 24, 150−158. (24) Brukhno, A. V.; Akinshina, A.; Coldrick, Z.; Nelson, A.; Auer, S. Phase phenomena in supported lipid films under varying electric potential. Soft Matter 2011, 7, 1006−1017. (25) Nelson, A. Electrochemical analysis of a phospholipid phase transition. J. Electroanal. Chem. 2007, 601, 83−93. (26) Zhang, S.; Nelson, A.; Coldrick, Z.; Chen, R. The effects of substituent grafting on the interaction of pH-responsive polymers with phospholipid monolayers. Langmuir 2011, 27, 8530−8539. (27) Sonnefeld, J. Determination of surface charge density constants for spherical silica particles using a linear transformation. J. Colloid Interface Sci. 1996, 183, 932−938. (28) Coldrick, Z.; Penezic, A.; Gasparovic, B.; Steenson, P.; Merrifield, J.; Nelson, A. High throughput systems for screening biomembrane interactions on fabricated mercury film electrodes. J. Appl. Electrochem. 2011, 41, 939−949. (29) Coldrick, Z.; Steenson, P.; Millner, P.; Davies, M.; Nelson, A. Phospholipid monolayer coated microfabricated electrodes to model the interaction of molecules with biomembranes. Electrochim. Acta 2009, 54, 4954−4962. (30) Protopapa, E.; Maude, S.; Aggeli, A.; Nelson, A. Interaction of selfassembling B-sheet peptides with phospholipid monolayers: the role of aggregation state, polarity, charge and applied field. Langmuir 2009, 25, 3289−3296. (31) Whitehouse, C.; O’Flanagan, R.; Lindholm-Sethson, B.; Movaghar, B.; Nelson, A. Application of electrochemical impedance spectroscopy to the study of dioleoyl phosphatidylcholine monolayers on mercury. Langmuir 2003, 20, 136−144. (32) Nelson, A.; Benton, A. Phospholipid monolayers at the mercury/ water interface. J. Electroanal. Chem. Interfacial Electrochem. 1986, 202, 253−270.

(33) Yuan, C.; Furlong, J.; Burgos, P.; Johnston, L. J. The size of lipid rafts: an atomic force microscopy study of ganlioside GM1 domains in sphingomyelin/DOPC/cholesterol membranes. Biophys. J. 2002, 82, 2526−2535. (34) Nam, J.; Beales, P. A.; Vanderlick, T. K. Giant phospholipid/block copolymer hybrid vesicles: mixing behavior and domain formation. Langmuir 2010, 27, 1−6. (35) Axelrod, D.; Koppel, D. E.; Schlessinger, J.; Elson, E.; Webb, W. W. Mobility measurement by analysis of fluorescence photobleaching recovery kinetics. Biophys. J. 1976, 16, 1055−1069. (36) Soumpasis, D. M. Theoretical analysis of fluorescence photobleaching recovery experiments. J. Biophys. Soc. 1983, 41, 95−97. (37) Baksh, M. M.; Jaros, M.; Groves, J. T. Detection of molecular interactions at membrane surfaces through colloid phase transitions. Nature 2004, 427, 139−141. (38) Fahey, P. F.; Webb, W. W. Lateral diffusion in phospholipid bilayer membranes and multilamellar liquid crystals. Biochemistry 1978, 17, 3046−3053. (39) Weiss, S.; Millner, P.; Nelson, A. Monitoring protein binding to phospholipid monolayers using electrochemical impedance spectroscopy. Electrochim. Acta 2005, 50, 4248−4256. (40) Anderson, T. H.; Min, Y.; Weirich, K. L.; Zeng, H.; Fygenson, D.; Israelachvili, J. N. Formation of supported bilayers on silica substrates. Langmuir 2009, 25, 6997−7005. (41) Tadros, T. F.; Lyklema, J. Adsorption of potential-determining ions at the silica-aqueous electrolyte interface and the role of some cations. J. Electroanal. Chem. Interfacial Electrochem. 1968, 17, 267−275. (42) Deserno, M. Elastic deformation of a fluid membrane upon colloid binding. Phys. Rev. E 2004, 69, 031903. (43) Ginzburg, V. V.; Balijepalli, S. Modeling the thermodynamics of the interaction of nanoparticles with cell membranes. Nano Lett. 2007, 7, 3716−3722. (44) Roiter, Y.; Ornatska, M.; Rammohan, A. R.; Balakrishnan, J.; Heine, D. R.; Minko, S. Interaction of nanoparticles with lipid membrane. Nano Lett. 2008, 8, 941−944. (45) Wang, B.; Zhang, L.; Bae, S. C.; Granick, S. Nanoparticle-induced surface reconstruction of phospholipid membranes. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 18171−18175. (46) Karatekin, E.; Sandre, O.; Guitouni, H.; Borghi, N.; Puech, P.; Bochard-Wyart, B. Cascades of transient pores in giant vesicles: line tension and transport. Biophys. J. 2003, 84, 1734−1749. (47) Rodriguez, N.; Cribier, S.; Pincet, F. Transition from long- to short-lived transient pores in giant vesicles in an aqueous medium. Phys. Rev. E 2006, 74, 061902−061912. (48) Sakuma, Y.; Taniguchi, T.; Imai, M. Pore formation in a binary giant vesicle induced by cone-shaped lipids. Biophys. J. 2010, 99, 472− 479. (49) Mu, Q.; Hondow, N.; Krzeminski, L.; Brown, A.; Jeuken, L.; Routledge, M. Mechanism of cellular uptake of genotoxic silica nanoparticles. Part. Fibre Toxicol., in press.

12837

dx.doi.org/10.1021/la301771b | Langmuir 2012, 28, 12831−12837