NANO LETTERS
The Formation of Supported Lipid Bilayers on Silica Nanoparticles Revealed by Cryoelectron Microscopy
2005 Vol. 5, No. 2 281-285
Ste´phane Mornet,† Olivier Lambert,† Etienne Duguet,‡ and Alain Brisson*,† Molecular Imaging and Nano-Bio-Technology, IECB, UMR-CNRS 5471, UniVersity of Bordeaux 1, 2 rue Robert Escarpit, F-33607 Pessac, France, and Hybrid Materials, ICMCB-CNRS 9048, UniVersity of Bordeaux 1, 87 aVenue Dr A. Schweitzer, 33608 Pessac, France Received November 8, 2004; Revised Manuscript Received November 19, 2004
ABSTRACT The controlled fabrication of biocompatible devices made of lipid bilayers deposited onto flat solid supports presents interest as models of cell membranes as well as for their biotechnological applications. We report here on the formation of supported lipid bilayers on silica nanoparticles (nanoSLBs). The successive steps of the adsorption of lipid vesicles on nanoparticles and the formation of nanoSLBs are revealed in detail by cryotransmission electron microscopy (cryo-EM). The formation of nanoSLBs was achieved for liposomes with positive, neutral, and low net negative charge, while liposomes with a high net negative charge adsorbed to silica nanoparticles but did not rupture. The nanoSLBs were found to follow faithfully the surface contours of the particles, information yet unavailable for SLB formation on planar solid substrates.
The use of nanoparticles in biotechnology presents a growing interest for the numerous possibilities offered by combining the world of materials, with its advanced technologies and their diverse properties, and the biological world, with its elaborate molecular architectures, properties, and functions. Silica nanoparticles are attractive for their intrinsic properties of optical transparency, controllable porosity, chemical inertness, and biocompatibility, as well as for their property of self-assembling into two-dimensional (2D) or three-dimensional (3D) colloidal crystals used in the development of photonic crystals,1,2 of separation membranes2 or in the conception of patterned surfaces.3 Various methods of chemical modification are available for coupling polymers or (bio)organic molecules to silica surfaces,4 and modified silica particles have already been used in biological and biotechnological applications.5 However, the control of surface biofunctionalization still constitutes the main issue in the conception of nanobiomaterials with controlled properties. A widely used approach to render a surface biocompatible consists of creating a lipid bilayer onto the solid surface via the spontaneous deposition of lipid vesicles.6,7 Supported lipid bilayers (SLBs) present interest for their use as models of cell membranes as well as for their potential biotechnological applications. Recent studies have provided a coherent picture * Corresponding author. E-mail:
[email protected]; Tel: +33 5 40003458; Fax +33 5 40003484. † IECB, University of Bordeaux. ‡ ICMCB, University of Bordeaux. 10.1021/nl048153y CCC: $30.25 Published on Web 12/24/2004
© 2005 American Chemical Society
of the processes of deposition of lipid vesicles on solid surfaces and of formation of continuous, defect-free, fluid lipid membranes separated from the substrate by a thin (∼1-3 nm) water layer.8,9 The formation of SLBs has been mainly studied at the level of cm2-size planar substrates, e.g. glass, silica, or mica,10-13 and has also been reported with µm-size silica beads.9,14-17 With the aim of designing nanovectors made of functionalized nanoparticles, we investigated whether silica nanoparticles could be covered, in a controllable and stable manner, by SLBs (referred to hereafter as nanoSLBs). Except for a few reports,18,19 the formation of such nanoSLBs has yet received little attention. This study stresses the critical importance of the methods of characterization in the development of nanomaterials. As demonstrated in this paper, cryotransmission electron microscopy20 (cryo-EM) allows to follow the adsorption of lipid vesicles on silica nanoparticles and the formation of nanoSLBs at the nm-scale, extending the panoply of characterization methods already available in nanotechnology of hybrid materials. Colloidal silica nanoparticles with controlled characteritics of size, dispersity, and morphology were prepared using the well-known method developed by Sto¨ber et al.21 Briefly, after hydrolysis and condensation of tetraethoxysilane in an ethanol/ammonia/water mixture (50:1:2, v:v:v) for 12 h at ambient temperature, amorphous silica nanoparticles with a mean diameter of 110 nm ((30 nm) were obtained. They were dialyzed against ultrapure water (10 MΩ) and stored
at a final concentration of 65 g/L, which corresponds to a specific surface area of 1700 m2/L. For the experiments reported here, the nanoparticles were diluted 10× in a buffer made of 150 mM NaCl, 10 mM Hepes, 2 mM NaN3, pH 7.4 (buffer A). As metal oxide particles are often unstable when dispersed in a saline buffer, we must stress the remarkable stability of silica sols in this buffer of high ionic strength, which is necessary for the subsequent experiments involving biological molecules. The stability of the silica nanoparticles is certainly related to the low Hamaker constant of silica, as compared to other metal oxides. The deposition of small unilamellar vesicles (SUVs) onto silica nanoparticles was characterized by cryo-EM (Tecnai F20 (FEI), operated at 200 kV). Cryo-EM of specimens embedded in thin films of vitreous water or buffer was preferred over classical methods of EM, like negative staining, for its main advantage of maintaining specimens in their native, hydrated state.20 Cryo-EM images showed silica particles presenting an overall spherical shape, with some surface roughness, and a heterogeneous electron-scattering density due to their microporosity (Figure 1A). SUVs of desired lipid mixture in buffer A were prepared by sonication, according to a procedure already described.13 Nanoparticles and SUVs (100 µg/mL and 45 µg/mL, respectively) were mixed for 1 h at room temperature. In these conditions, the lipid bilayer surface is expected to be 15× larger than the surface of the particles. The deposition of SUVs onto silica nanoparticles was first investigated with a lipid mixture made of zwitterionic dioleoylphosphatidylcholine (DOPC) and negatively charged dioleoylphosphatidylserine (DOPS) (weight ratio 4:1), which is known to form continuous SLBs on planar substrates of both silica13 and mica.10 By cryo-EM, the silica nanoparticles were found to be surrounded by a continuous ring of electron-dense material, separated from the particle edge by about 4-5 nm (Figures 1B,E), a distance that corresponds to the known thickness of a lipid bilayer (Figure 1C).22 The electron-dense ring observed at the particle periphery corresponds to the outer, distal, lipid leaflet of the nanoSLB. The inner, proximal, lipid leaflet is not resolved here due to its closeness to the particle edge, the resolution of the cryo-EM images being around 2 nm. The cryo-EM images shown here provide the first direct views, to our knowledge, of SLBs adhering to a substrate. They demonstrate that the SLBs follow faithfully the surface roughness of silica particles (arrow in Figure 1E). In addition, they suggest that the thickness of the water layer separating the inner lipid layer from the support9,23,24 is at most 1 nm, as a distance of only 4-5 nm separates the substrate from the outer lipid layer. To get insight into the mechanism of nanoSLB formation, images of intermediate steps were recorded by freezing samples at different incubation times after mixing DOPC/ DOPS (4:1) SUVs and silica nanoparticles. After 1-2 min incubation, the shortest time that can be analyzed experimentally, various steps could be resolved (Figure 2). A majority of nanoparticles exhibited at their surface one or few patches of electron dense material, corresponding to 282
Figure 1. Formation of nanoSLBs around silica nanoparticles, by cryo-EM. (A) Silica nanoparticles (mean diameter: 110 nm) appear as spheres of uneven density with a rough surface. (Scale bar: 50 nm. B and C at the same magnification as A). (B) After addition of DOPC/DOPS (4:1) SUVs, the particles appear surrounded by a ring of electron-dense material, which corresponds to the outer lipid layer of a SLB covering the particle surface, as schematically drawn in (D). The inset presents an image obtained by the classical negative staining method (scale bar: 50 nm). As expected, dehydration damage prevents revealing the structural details of the lipid material associated with the particles. (C) Typical cryo-EM image of a liposome suspended in a thin droplet of frozen buffer. The two concentric rings correspond to the polar headgroups from the two lipid layers, which scatter electrons more strongly than the hydrocarbon chains. (D) Scheme of a nanoparticle surrounded by a nanoSLB. The dashed and continuous circular lines represent the inner and outer lipid leaflets, respectively. The inner leaflet is not resolved in the cryo-EM images. (E) This image stresses the close contact between the nanoSLB and the substrate (scale bar: 20 nm).
nanoparticles covered incompletely by nanoSLBs (Figure 2A). The maximal size of these lipid bilayer patches was about 50-100 nm, which corresponds to the size of bilayer disks formed by rupture of vesicles of about 30-50 nm diameter. Few particles were covered by a continuous nanoSLB (Figure 2C). In addition, rare silica particles presented one or few liposomes adsorbed at their surface, unruptured (Figure 2B,C,D). The low number of unruptured vesicles, as compared to the large number of lipid bilayer patches, indicates that the kinetics of vesicle rupture is fast. After 5 min incubation, most silica nanoparticles presented continuous nanoSLBs. These observations are in overall agreement with results obtained on large and flat silica substrates, which indicate that the formation of SLBs occurs within few minutes in conditions similar to those used here.12,13 The sols made of nanoSLBs and SUVs were stable for long periods (> 1 month), at a temperature of 4 °C. Nano Lett., Vol. 5, No. 2, 2005
Figure 2. Gallery of images from mixtures of silica particles and DOPC/DOPS (4:1) SUVs, quickly frozen after 1 min incubation time. Most particles present one or two patches of lipid bilayer at their surface (black arrows in A). Few particles present vesicles adsorbed, unruptured, at their surface (white asterisks in B,C). At the contact areas (white arrows in B,C), vesicles present one single electron-dense line, while the two lipid leaflets are resolved in the nonadhesive areas. In (D), the contact areas between the vesicles (white arrowheads) and the particle are not visible, as they are “hidden” by the superposed silica mass. Few particles present a complete nanoSLB (D, black asterisk). Scale bar: 50 nm.
The observation of these distinct intermediate steps allows us to propose a straightforward sequence of events, from the adsorption of single SUVs to the formation of complete nanoSLBs. After adsorption of a SUV to a silica particle, the contact area between the lipid surface and the substrate increases, due to adhesion forces. As bilayer patches seem to originate from the rupture of single vesicles (from size considerations), it is logical to postulate that there is a limit in the deformation of SUVs above which rupture occurs. This leads to the formation of bilayer patches covering incompletely the silica nanoparticles. The process proceeds via the subsequent adsorption, deformation, and rupture of additional SUVs, either independently or dependently from the presence of preformed bilayer patches. What is meant by “dependently” is that the edges of bilayer patches are likely to activate the decomposition of adsorbed vesicles, as evidenced by AFM on flat silica.13 A simple calculation shows that about ten SUVs of 30 nm diameter are necessary to cover entirely a spherical nanoparticle of 110 nm diameter. The completion of the formation of nanoSLBs is likely to involve the coalescence of neighboring bilayer patches when they come into contact, via active edge effects.10 Such a mechanism is consistent with previous experimental10,13,25 and theoretical studies.26,27 However, while a local critical density of DOPC/DOPS (4:1) vesicles is required for initiating the formation of SLBs on large silica supports,13 vesicles seem to rupture one by one on silica nanoparticles. The curvature of the particles and/or the different properties, Nano Lett., Vol. 5, No. 2, 2005
e.g., charge, surface density of silanol groups, roughness, between the two silica substrates may explain this difference. To further characterize the nature of the interactions leading to the formation of nanoSLBs, we studied the influence of the lipid composition of SUVs on the formation of nanoSLBs. As electrostatic interactions play a determining role in the formation of SLBs on planar silica substrates,13,14 we selected a variety of lipids, including positively charged lipids (1,2-dioleoyl-3-trimethylammonium-propane (DOTAP)), neutral zwitterionic lipids (DOPC; 1-palmitoyl-2oleoylphosphatidylcholine (POPC); dimyristoylphosphatidylcholine (DMPC)), and negatively charged lipids (DOPS). The formation of nanoSLBs was achieved with DOTAP, with DOPC, as well as with (DOPC/DOPS) mixtures of low net negative charge (4:1 and 3:1). On the other hand, for (DOPC/ DOPS) mixtures with a higher net negative charge (1:1), complete nanoSLBs were observed only on a few particles, while a large number of nanoparticles exhibited unruptured vesicles bound at their surface (Figure 3). These results correlate well with results previously reported on flat silica supports,13,14 establishing that the formation of nanoSLBs on silica nanoparticles is also highly dependent on electrostatic interactions. Interestingly, the extent of deformation of unruptured SUVs varies with their net charge and can be quantitatively determined by cryo-EM (compare for example the deformation of unruptured vesicles in Figures 2B,C and 3B). This constitutes novel experimental information that should help refine the theoretical description of processes 283
Figure 3. Adsorption of DOPC/DOPS (1:1) SUVs to silica nanoparticles. (A, B, C) The silica particles appear fully covered with adsorbed vesicles. Several images provide detailed information on the deformation of vesicles (white arrows in A,B; double dark arrowhead in A). (B) Note that the two lipid leaflets of the adsorbed vesicles are visible in the nonadhesive areas (black asterisk), while only one leaflet is visible in the adhesive areas (white asterisk). (C) Rare example of a particle with a lipid patch (black arrow). Scale bar: 50 nm.
Figure 4. NanoSLBs formed with SUVs made of (A) DOPC; (B) POPC; (C) DMPC. Scale bar 50 nm.
of adhesion and rupture of lipid vesicles on solid substrates.26,27 In addition, the formation of nanoSLBs was achieved with DOPC, POPC, and DMPC lipids, which present various degrees of unsaturation yet are all in fluid phase in the experimental conditions used here (Figure 4). In conclusion, this study provides several interesting findings concerning the deposition of lipid vesicles on solid supports and the formation of nanoSLBs. It stresses the unique possibilities offered by cryo-EM to characterize nanomaterials such as nanoparticles in situ and to study in detail, at 1-2 nm resolution and almost in real time, processes of nanoparticle surface modification. Cryo-EM may become a routine method for characterizing the properties as well as the potential hazards28 of nanoparticle-based materials. The images of adsorbed liposomes and of nanoSLBs presented here provide information that is highly complementary to information obtained either by quantitative physicochemical methods such as QCM-D,12,13 SPR,29 or ellipsometry30 or by imaging methods such as AFM10,11 or fluorescence.31-33 This study gives a direct and convincing demonstration that lipid bilayers follow faithfully the topography of the silica substrate. The long-debated question of 284
whether an SLB does or does not follow the roughness of a surface finds here an unambiguous answer, at least for silica particles. Irregularities in the silica surface do not prevent the spreading of SUVs nor the formation of SLBs, contrary to previous speculations.8,14,34 Nanoparticles covered with SLBs combine the intrinsic properties of metal oxide particles and of lipid bilayers and constitute objects with potential interest both in basic and in applied science. By uncoupling the inorganic surface from the surrounding aqueous phase, nanoSLBs provide a natural environment for biomolecules, eliminating or reducing possible problems of nonspecific adsorption or protein denaturation, known to be critical in the development of biomaterials. SLB-coated nanoparticles open numerous possibilities of functionalization for the development of nanovectors with specific molecular targets as well as for carrying and delivering molecules to a selected site. In addition, strategies developed for silica nanoparticles are directly applicable to core-shell silica nanoparticles, such as superparamagnetic ferrite nanoparticles35 or fluorescent quantum dots,36 extending the range of physical properties that could be used for addressing, activating, or detecting. Nano Lett., Vol. 5, No. 2, 2005
Acknowledgment. S.M. is recipient of a postdoctoral fellowship from the Conseil Re´gional d’Aquitaine. This work has been supported in part by EC grant “NMP4-CT2003505868 -Nanocues”. References (1) Liz-Marza´n, L. M.; Mulvaney, P. J. Phys. Chem. B 2003, 107, 73127326. (2) Wang, D.; Caruso, R. A.; Caruso, F. Chem. Mater. 2001, 13, 364371. (3) Xia, D.; Brueck, S. R. J. Nano Lett. 2004, 4, 1295-1299. (4) Wong, S. S. Chemistry of protein conjugation and cross-linking; CRC Press: Boca Raton, 1991. (5) Zhao, X.; Hilliard, L. R.; Wang, K.; Tan, W. Encyclopedia of Nanoscience and Nanotechnology; American Scientific Publishers: Stevenson Ranch, 2004; Vol. 1, pp 255-268. (6) McConnell, H. M.; Watts, T. H.; Weis, R. M.; Brian, A. A. Biochim. Biophys. Acta 1986, 864, 95-106. (7) Sackmann, E. Science 1996, 271, 43-48. (8) Ra¨dler, J.; Strey, H.; Sackmann, E. Langmuir 1995, 11, 4539-4548. (9) Bayerl, T. M.; Bloom, M. Biophys. J. 1990, 58, 357-362. (10) Reviakine, I.; Brisson, A. Langmuir 2000, 16, 1806-1815. (11) Jass, J.; Tja¨rnhage, T.; Puu, G. Biophys. J. 2000, 79, 3153-3163. (12) Reimhult, E.; Ho¨o¨k, F.; Kasemo, B. Langmuir 2003, 19, 1681-1691. (13) Richter, R.; Mukhopadhyay, A.; Brisson, A. Biophys. J. 2003, 85, 3035-3047. (14) Cremer, P.; Boxer, S. G. J. Phys. Chem. B 1999, 103, 2554-2559. (15) Linseisen, F. M.; Hetzer, M.; Brumm, T.; Bayerl, T. M. Biophys. J. 1997, 72, 1659-1667. (16) Scott, M. J.; Jones, M. N. Colloids Surf. A 2002, 207, 69-79. (17) Baksh, M. M.; Jaros, M.; Groves, J. T. Nature 2004, 427, 139-141. (18) Rapuano, R.; Carmona-Ribeiro, E. M. J. Colloid Interface Sci. 1997, 193, 104-111.
Nano Lett., Vol. 5, No. 2, 2005
(19) Rapuano, R.; Carmona-Ribeiro, E. M. J. Colloid Interface Sci. 2000, 226, 299-307. (20) Dubochet, J.; Adrian, M.; Chang, J.-J.; Homo, J.-C.; Lepault, J.; McDowall, A. W.; Schultz, P. Q. ReV. Biophys. 1988, 21, 129-228. (21) Sto¨ber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62. (22) Tahara, Y.; Fujiyoshi, Y. Micron. 1994, 25, 141-149. (23) Johnson, J.; Bayerl, T. M.; McDermott, D. C.; Adam, G. W.; Rennie, A. R.; Thomas, R. K.; Sackmann, E. Biophys. J. 1991, 59, 289294. (24) Kiessling, V.; Tamm, L. K. Biophys. J. 2003, 84, 408-418. (25) Keller, C. A.; Kasemo, B. Biophys. J. 1998, 75, 1397-1402. (26) Seifert, U.; Lipowsky, R. Phys. ReV. A 1990, 42, 4768-4771. (27) Lipowsky, R.; Seifert, U. R. Mol. Cryst. Liq. Cryst. 1991, 202, 1725. (28) Report from the UK Royal Society and Royal Academy of Engineering 2004 http://www.nanotec.org.uk/finalReport.htm (accessed September 2004). (29) Knoll, W. Annu. ReV. Phys. Chem. 1998, 49, 569-638. (30) Benesˇ, M.; Billy, D.; Hermens, W. T.; Hof, M. Biol. Chem. 2002, 383, 337-341. (31) Benda, A.; Benesˇ, M.; Marecˇek, V.; Lhotsky´, A.; Hermens, W. T.; Hof, M. Langmuir 2003, 19, 4130-4126. (32) Johnson, J.; Ha, T.; Chu, S.; Boxer, S. G. Biophys. J. 2002, 83, 33713379. (33) Groves, J. T.; Ulman, K.; Boxer, S. G. Science 1997, 275, 651653. (34) Muresan, A. S.; Lee, K. Y. C. J. Phys. Chem. B 2001, 105, 852855. (35) Mornet, S.; Grasset, F.; Portier, J.; Duguet, E. Eur. Cells Mater. 2002, 3, 110-113. (36) Rogach, A. L.; Nagesha, D.; Ostrander, J.; Giersig, M.; Kotov, N. A. Chem. Mater. 2000, 12, 2676-2685.
NL048153Y
285
