Xerogel

Jul 16, 2004 - Kevin C. Weng,†,| Johan J. R. Stålgren,‡,| David J. Duval,§. Subhash H. Risbud,§,| and Curtis W. Frank*,‡,|. Department of Mat...
13 downloads 0 Views 348KB Size
Download PDF
7232

Langmuir 2004, 20, 7232-7239

Fluid Biomembranes Supported on Nanoporous Aerogel/ Xerogel Substrates Kevin C. Weng,†,| Johan J. R. Stålgren,‡,| David J. Duval,§ Subhash H. Risbud,§,| and Curtis W. Frank*,‡,| Department of Materials Science and Engineering, Stanford University, Stanford, California 94305, Department of Chemical Engineering, Stanford University, Stanford, California 94305, Department of Chemical Engineering and Materials Science, University of California at Davis, Davis, California 95616, and Center on Polymer Interfaces and Macromolecular Assemblies (CPIMA), Stanford, California 94305 Received January 7, 2004. In Final Form: March 15, 2004 Planar supported lipid bilayers have attracted immense interest for their properties as model cell membranes and for potential applications in biosensors and lab-on-a-chip devices. We report the formation of fluid planar biomembranes on hydrophilic silica aerogels and xerogels. Scanning electron microscopy results showed the presence of interconnected silica beads of approximately 10-25 nm in diameter and nanoscale open pores of comparable size for the aerogel and grain size of ∼36-104 nm with ∼9-24 nm diameter pores for the xerogel. When the aerogel/xerogel was prehydrated and then allowed to incubate in L-R-phosphatidylcholine (egg yolk PC) unilamellar vesicle (∼30 nm diameter) solution, lipid bilayers were formed due to the favorable interaction of vesicles with the hydroxyl-abundant silica surface. Lateral mobility of labeled lipid N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine was retained in the membranes. A diffusion coefficient of 0.61 ( 0.22 µm2/s was determined from fluorescence recovery after photobleaching analysis for membranes on aerogels, compared to 2.46 ( 0.35 µm2/s on flat glass. Quartz crystal microbalance-dissipation was utilized to monitor the kinetics of the irreversible adsorption and fusion of vesicles into bilayers on xerogel thin films.

Introduction Membranes comprising lipid bilayers are crucial to many biological functions such as cellular transport, signal transduction, molecular recognition, and compartmentalization. To study the properties of such lipid assemblies by various surface-sensitive techniques as well as to develop technological applications based on such architecture, significant effort has been devoted to creating model lipid bilayers supported on planar substrates.1 The incorporation of membrane proteins into these bilayers is also a high priority for many scientific and technological purposes.2-4 Recently, analytical multiplexing was achieved through development of the capability of partitioning the supported membranes and spatially addressing individual elements.5 To mimic cell membranes, certain structural and dynamic features are required for the artificially constructed lipid bilayers. An important aspect is that * To whom correspondence should be addressed. E-mail: [email protected]. † Department of Materials Science and Engineering, Stanford University. ‡ Department of Chemical Engineering, Stanford University. § Department of Chemical Engineering and Materials Science, University of California at Davis. | Center on Polymer Interfaces and Macromolecular Assemblies (CPIMA). (1) Sackmann, E. Supported Membranes: Scientific and Practical Applications. Science 1996, 271, 43-48. (2) Cornell, B. A.; Braach-Maksvytis, V. L. B.; King, L. G.; Osman, P. D. J.; Raguse, B.; Wieczorek, L.; Pace, R. J. A Biosensor that Uses Ion-channel Switches. Nature 1997, 385, 580-583. (3) Stora, T.; Lakey, J. H.; Vogel, H. Ion-Channel Gating in Transmembrane Receptor Proteins: Functional Activity in Tethered Lipid Membranes. Angew. Chem., Int. Ed. 1999, 38, 389-392. (4) Graneli, A.; Rydstrom, J.; Kasemo, B.; Hook, F. Formation of Supported Lipid Bilayer Membranes on SiO2 from Proteoliposomes Containing Transmembrane Proteins. Langmuir 2003, 19, 842-850. (5) Groves, J. T.; Boxer, S. G. Micropattern Formation in Supported Lipid Membranes. Acc. Chem. Res. 2002, 35, 149-157.

the two-dimensional membranes suspended on the surface should display lateral mobility because many membrane activities in living cells involve transport, recruitment, or assembly of specific components. Fluidity is also critical for membranes to respond to environmental stress such as that resulting from temperature and pressure changes.6 Various methods have been utilized to achieve such configurations. Proteoliposome adsorption and subsequent membrane formation on a mica surface was first demonstrated by Brian and McConnell in 1984.7 Because of its simplicity and reproducibility, the technique has been one of the most commonly adopted approaches to prepare supported membranes. There are several important driving forces in the formation of supported membranes from vesicles. In particular, the interaction of vesicles and resulting bilayers with surfaces involves a subtle balance among van der Waals, electrostatic, hydration, and steric forces.8 The widely held perspective is that the interaction of vesicles and the surface plays the determining role in the successful fusion and spreading of lipid bilayers.9 Generally speaking, hydrophilic surfaces that have strong attractive interactions with lipids promote bilayer formation. The process has been studied by atomic force microscopy (AFM)10,11 and single vesicle fluorescence (6) Shinitzky, M. Physiology of Membrane Fluidity; Shinitzky, M., Ed.; CRC Press: Boca Raton, FL, 1984. (7) Brian, A. A.; McConnell, H. M. Allogeneic Stimulation of Cytotoxic T Cells by Supported Planar Membranes. Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 6159-6163. (8) Israelachvili, J. N. Intermolecular & Surface Forces; Academic Press: San Diego, 1992. (9) Keller, C. A.; Kasemo, B. Surface Specific Kinetics of Lipid Vesicles Adsorption Measured with a Quartz Crystal Microbalance. Biophys. J. 1998, 75, 1397-1402. (10) Leonenko, Z. V.; Carnini, A.; Cramb, D. T. Supported Planar Bilayer Formation by Vesicle Fusion: The Interaction of Phospholipid Vesicles with Surfaces and the Effect of Gramicidin on Bilayer Properties Using Atomic Force Microscopy. Biochim. Biophys. Acta 2000, 1509, 131-147.

10.1021/la049940d CCC: $27.50 © 2004 American Chemical Society Published on Web 07/16/2004

Fluid Biomembranes Supported on Gel Substrates

assays,12 which have revealed the mechanistic steps in vesicle-to-bilayer transformation. Researchers have explored many kinds of surfaces for bilayer construction through vesicle fusion.11,13 Due to intrinsic material properties and their interactions with lipids in an aqueous environment, different substrates offer particular advantages and disadvantages. For example, silicon oxide, both in the planar and microspherical formats, has been recognized as one of the most appropriate surfaces for vesicle fusion.14-16 The membrane is supported by a thin lubricating layer of water, approximately 1 nm in thickness, as shown by proton NMR, neutron reflectivity, and fluorescence interference-contrast microscopy results.17-20 According to previous studies on membrane protein insertion, the water layer typically does not provide enough separation between the membranes and the substrates, and the integral membrane proteins encounter nonphysiological interactions with the solid supports.21 To resolve this difficulty, a hydrophilic, swollen polymer is being actively investigated as the cushion material for supported membranes.22-25 In some cases, however, the fully tethered or semitethered polymers hinder lipid diffusion in the membranes.26 This approach also complicates the design and chemistry involved in the synthesis. In reality, the coupling of a supported lipid bilayer with suitable substrates to create a fully (11) Reviakine, I.; Brisson, A. Formation of Supported Phospholipid Bilayers from Unilamellar Vesicles Investigated by Atomic Force Microscopy. Langmuir 2000, 16, 1806-1815. (12) Johnson, J. M.; Ha, T.; Chu, S.; Boxer, S. G. Early Steps of Supported Bilayer Formation Probed by Single Vesicle Fluorescence Assays. Biophys. J. 2002, 83, 3371-3379. (13) Groves, J. T.; Ulman, N.; Cremer, P. S.; Boxer, S. G. SubstrateMembrane Interactions: Mechanisms for Imposing Patterns on a Fluid Bilayer Membrane. Langmuir 1998, 14, 3347-3350. (14) Groves, J. T.; Ulman, N.; Boxer, S. G. Micropatterning Fluid Lipid Bilayers on Solid Supports. Science 1997, 275, 651-653. (15) Loidl-Stahlhofen, A.; Schmitt, J.; Noller, J.; Hartmann, T.; Brodowsky, H.; Schmitt, W.; Keldenich, J. Solid-Supported Biomolecules on Modified Silica Surfaces: A Tool for Fast Physicochemical Characterization and High-Throughput Screening. Adv. Mater. 2001, 13, 18291834. (16) Buranda, T.; Huang, J.; Ramarao, G. V.; Ista, L. K.; Larson, R. S.; Ward, T. L.; Sklar, L. A.; Lopez, G. P. Biomimetic Molecular Assemblies on Glass and Mesoporous Silica Microbeads for Biotechnology. Langmuir 2003, 19, 1654-1663. (17) Bayerl, T. M.; Bloom, M. Physical Properties of Single Phospholipid Bilayers Adsorbed to Micro Glass Beads. Biophys. J. 1990, 58, 357-362. (18) Johnson, S. J.; Bayerl, T. M.; McDermott, D. C.; Adam, G. W.; Rennie, A. R.; Thomas, R. K.; Sackmann, E. Structure of an Adsorbed Dimyristoylphosphatidylcholine Bilayer Measured with Specular Reflection of Neutrons. Biophys. J. 1991, 59, 289-294. (19) Koenig, B. W.; Krueger, S.; Orts, W. J.; Majkrzak, C. F.; Berk, N. F.; Silverton, J. V.; Gawrisch, K. Neutron Reflectivity and Atomic Force Microscopy Studies of a Lipid Bilayer in Water Adsorbed to the Surface of a Silicon Single Crystal. Langmuir 1996, 12, 1343-1350. (20) Fromherz, P.; Kiessling, V.; Kottig, K.; Zeck, G. Membrane Transistor with Giant Lipid Vesicle Touching a Silicon Chip. Appl. Phys. A: Mater. Sci. Process. 1999, 69, 571-576. (21) Salafsky, J.; Groves, J. T.; Boxer, S. G. Architecture and Function of Membrane Proteins in Planar Supported Bilayers: A Study with Photosynthetic Reaction Centers. Biochemistry 1996, 35, 14773-14781. (22) Knoll, W.; Frank, C. W.; Heibel, C.; Naumann, R.; Offenhausser, A.; Ruhe, J.; Schmidt, E. K.; Shen, W. W.; Sinner, A. Functional Tethered Lipid Bilayers. Rev. Mol. Biotechnol. 2000, 74, 137-158. (23) Sackmann, E.; Tanaka, M. Supported Membranes on Soft Polymer Cushions: Fabrication, Characterization and Applications. Trends Biotechnol. 2000, 18, 58-64. (24) Wagner, M. L.; Tamm, L. K. Tethered Polymer-Supported Planar Bilayers for Reconstitution of Integral Membrane Proteins: SilanePolyethyleneglycol-Lipid as a Cushion and Covalent Linker. Biophys. J. 2000, 79, 1400-1414. (25) Munro, J. C.; Frank, C. W. Adsorption of Lipid-Functionalized Poly(ethylene glycol) to Gold Surfaces as a Cushion for PolymerSupported Lipid Bilayers. Langmuir 2004, 20, 3339-3349. (26) Shen, W. W.; Boxer, S. G.; Knoll, W.; Frank, C. W. PolymerSupported Lipid Bilayers on Benzophenone-Modified Substrates. Biomacromolecules 2001, 2, 70-79.

Langmuir, Vol. 20, No. 17, 2004 7233

functional and biomimetic system still presents a major engineering challenge. Since their discovery by Kistler in the early 1930s,27 aerogels have found a wide range of applications because of their unique physical and chemical properties.28,29 Silica aerogels are typically prepared by the sol-gel process, and the solvent is removed after gelation under supercritical conditions, which preserves the tenuous solid network. An aerogel dried by supercritical carbon dioxide is hydrophilic, similar to silicon oxide, but also permeable with extraordinarily high porosity (up to 99.8%), unlike glass.30 The preparation of a xerogel differs only in the solvent removal step in which the liquid part of the alcogel is removed by evaporation.31 Because the solvent is not removed under supercritical conditions, substantial shrinkage occurs in the gel volume. Hydrophilic silica aerogels and xerogels typically have the same chemical composition but different grain sizes and porosity (in our samples, the average silica grains in xerogels are ∼3.5 times larger than in aerogels). Bilayer suspension on a porous substrate has been attempted previously with an anodically etched aluminum foil.32 The surface was modified with a 10 nm layer of gold followed by self-assembly of 3-mercaptopropionic acid (MPA) for fusion of positively charged N,N-dimethyl-N,Ndioctadecylammonium bromide (DODAB) vesicles. The pores achieved were 60 ( 10 nm in diameter with a surface porosity of 16 ( 2%. Aerogels and xerogels typically present much higher controlled porosities and are essentially different from aluminum in their chemical compositions. Moreover, no measurements on lateral mobility of the membranes deposited on etched aluminum were reported. Our objective is to study the interaction of unilamellar vesicles with aerogel/xerogel surfaces, which possess a rich landscape of nanoporous cavities separated by a skeleton of solid silica. In the following, we demonstrate phospholipid bilayer membrane formation on the aerogel/ xerogel surface, as depicted schematically in Figure 1, and discuss the potential impact of this nanostructural supramolecular configuration. Results Scanning Electron Microscopy (SEM). Figure 2 shows a 100 000× SEM image and corresponding binary contrast enhancement that display highly interconnected open pores on the surface. Silica aerogels are made of low-density silicon oxide with formula SiOx(OH)y. In the SEM image, the “pearl string” structure of the aerogel is prominent, where ∼10-25 nm size silica particles are connected to form a three-dimensional network. Larger silica grains of ∼36-104 nm and smaller pores of ∼9-24 nm in diameter were observed in the SEM micrographs of xerogel surfaces (not shown). Silica aerogels typically exhibit surface topography of ca. 20-150 nm in height (27) Kistler, S. S. Coherent Expanded Aerogels. J. Phys. Chem. 1932, 36, 52-64. (28) Hrubesh, L. W. Aerogel Applications. J. Non-Cryst. Solids 1998, 225, 335-342. (29) Schmidt, M.; Schwertfeger, F. Applications for Silica Aerogel Products. J. Non-Cryst. Solids 1998, 225, 364-368. (30) Baumann, T. F.; Gash, A. E.; Fox, G. A.; Satcher, J. H.; Hrubesh, L. W. Aerogels. In Handbook of Porous Solids; Schuth, F., Sing, K. S. W., Weitkamp, J., Eds.; Wiley-VCH: Weinheim, 2002. (31) Brinker, C. J. Sol-Gel Processing of Silica. In The Colloid Chemistry of Silica; Advances in Chemistry Series, Vol. 234; American Chemical Society: Washington, DC, 1994; pp 361-402. (32) Hennesthal, C.; Steinem, C. Pore-Spanning Lipid Bilayers Visualized by Scanning Force Microscopy. J. Am. Chem. Soc. 2000, 122, 8085-8086.

7234

Langmuir, Vol. 20, No. 17, 2004

Weng et al.

Figure 1. A schematic of a lipid membrane supported on a nanoporous aerogel substrate. The curvature of the bilayer and the relative sizes of lipid molecules and silica nanostructures are not based on experimental observation. However, they reflect the approximate scale of the configuration as typical bilayers are ∼5 nm in thickness and the silica beads we observed are around 10-25 nm in diameter.

differences, as illustrated by atomic force microscopy.33 Our AFM results were consistent with this, yielding up to ∼70 nm variation in surface elevation (data not shown). Quartz Crystal Microbalance-Dissipation (QCMD). The quartz crystal microbalance with dissipation (QCM-D, Q-Sense AB, model D300) technique was used to monitor the adsorption and transformation of vesicles on the surfaces. Interactions on xerogel thin films were compared with those on the silicon oxide surface exposed to vesicle solutions of the same concentration (1 mg/mL). With the change of mass and viscoelastic properties of the adsorbed materials, the resonance frequency and the energy damping loss are modulated correspondingly, thus accounting for the shift in frequency (∆f) and dissipation (∆D) captured by the real-time curves.34 A decrease in resonance frequency is interpreted as a gain in mass on the crystal surface. For rigid thin films, a linear Sauerbrey relationship between the mass and frequency shift is observed.35 The increase in dissipation energy monitored simultaneously reflects the increase in the viscoelasticity of the thin films, with “softer” films displaying higher dissipation. Vesicle-vesicle and vesicle-surface interactions lead to vesicle deformation, vesicle fusion, and finally bilayer spreading. Since bilayers weigh less and are more rigid than a system consisting of an equivalent coverage of viscous, water-encapsulating vesicles, the resulting curves for ∆f and ∆D are distinct, thus making QCM-D an effective tool to differentiate the configurations of lipid assemblies on the surface.9 During vesicle fusion, the release of trapped water inside and between vesicles causes a decrease in adsorbed mass on the crystal. (33) Borne, A.; Chevalier, B.; Chevalier, J. L.; Quenard, D.; Elaloui, E.; Lambard, J. Characterization of Silica Aerogel with the Atomic Force Microscope and SAXS. J. Non-Cryst. Solids 1995, 188, 235-242. (34) Rodahl, M.; Hook, F.; Kasemo, B. QCM Operation in Liquids: An Explanation of Measured Variations in Frequency and Q Factor with Liquid Conductivity. Anal. Chem. 1996, 68, 2219-2227. (35) Munro, J. C.; Frank, C. W. Adsorption of Disulfide-Modified Polyacrylamides to Gold and Silver Surfaces as Cushions for PolymerSupported Lipid Bilayers. Polymer 2003, 44, 6335-6344.

Figure 2. SEM of a thin gold-palladium coated silica aerogel surface. (a) The original 100 000× image exhibits the threedimensional pearl bead-string structure of the aerogel with particle sizes at the nanometer scale. (b) A binary enhanced contrast image of the same image shows the interconnected particles and pores. The scale bar is 200 nm.

We have used the convention by Cady36 and Salt37 for designating the overtone numbers of the resonant frequencies of the QCM crystal. Thus, 5 MHz is referred to as the fundamental frequency, 15 MHz as the “first” overtone, 25 MHz as the “second” overtone, and so on. The interpretations of QCM-D results are typically based on comparison of shifts in 15 MHz frequency divided by a scaling factor of 3. Figure 3 shows the temporal variations of frequency (∆f) and dissipation (∆D) for crystal resonance at ∼15 MHz during vesicle adsorption and fusion, using absolute frequency and dissipation of buffer at equilibrium for initial baseline values. Because lower pH has been reported to facilitate membrane formation on silicon oxide surfaces,38 we used pH ∼ 4 to accelerate the vesicle fusion and bilayer spreading process in the xerogel experiment. Panels a and b of Figure 3 show vesicles adsorbed on thermally evaporated silicon oxide and xerogel thin film, respectively. The final frequency shift (∆f) for the silicon oxide surface was ∼88 Hz at 15 MHz resonance (first overtone) and ∼29 Hz at 5 MHz (fundamental), roughly (36) Cady, W. G. Piezoelectricity; Dover: New York, 1964. (37) Salt, D. Hy-Q Handbook of Quartz Crystal Devices; van Nostrand Reinhold: Cornwall, U.K., 1987. (38) Cremer, P. S.; Boxer, S. G. Formation and Spreading of Lipid Bilayers on Planar Glass Supports. J. Phys. Chem. B 1999, 103, 25542559.

Fluid Biomembranes Supported on Gel Substrates

Langmuir, Vol. 20, No. 17, 2004 7235

Figure 3. QCM-D data of egg PC vesicles on (a) thermally evaporated silicon oxide and (b) xerogel thin film. The left axis is frequency shift and the right axis is dissipation change for quartz crystal resonance at ∼15 MHz. The inset in panel a shows the expanded graph from 9.5 to 11.5 min where the minimum in ∆f and maximum in ∆D took place. Egg PC vesicles adsorb and transform into lipid bilayers on the solid silicon oxide surface faster than on the xerogel surface. Recovery after the first dip in ∆f reflects the transition of weight loss due to vesicle rupture overcoming continuous adsorption of intact vesicles. The second dip (at 40 min in panel a and 70 min in panel b) results from the rinsing of buffer to confirm the stabilizing of the lipid bilayers formed.

in agreement with data for a supported bilayer reported by Keller and Kasemo.9 Qualitatively, each ∆f curve exhibits an initial decrease followed by a minimum and ultimately an increase toward an asymptote. However, the kinetics and absolute amount of adsorbed lipid are different for each case. The change of df/dt occurred in less than 2 min for the silicon oxide surface, while the transition took ∼30 min for vesicles on a xerogel surface. In addition, greater magnitudes in both |∆f| and |∆D| at the extremum points for the xerogel surface signify higher increase in both mass and dissipation. This indicates that the surface coverage of vesicles on the xerogel is larger than on thermally evaporated silicon oxide

when bilayer formation starts to dominate the process. The spikes at 40 min in Figure 3a and at 70 min in Figure 3b for both ∆f and ∆D curves result from the pressure and temperature disturbances generated by the impinging buffer. ∆f and ∆D showed no significant increase or decrease after rinsing, indicating that the bilayers on the surface were stable and no gain or loss of mass (further adsorption or desorption) took place after the rinsing. Epifluorescence Microscopy and Fluorescence Recovery after Photobleaching (FRAP). To inspect the uniformity and long-range mobility of the lipid bilayers supported on aerogel substrates, epifluorescence microscopy was utilized to collect micron-scale images. FRAP

7236

Langmuir, Vol. 20, No. 17, 2004

Weng et al.

Figure 4. Epifluorescence micrographs and FRAP of L-R-phosphatidylcholine (egg PC) mixed with 3 mol % fluorescently labeled lipid N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (NBD-PE) lipid bilayers supported on aerogels. The mixing of mobile lipids caused the recovery of fluorescence intensity in the bleached region. The scale bar is 50 µm. (a) Before bleaching; (b) bleaching; (c) recovery begins at t ) 0 s; (d) t ) 2 min; (e) t ) 16 min; (f) t ) 96 min.

analysis was performed to extract diffusion coefficients from the temporal fluorescence intensity profiles of selected regions during recovery. The diffusion coefficients were calculated based on the Axelrod method39 by the following equation.

r2 D ) 0.224 t1/2 where r is the radius of the bleached area and t1/2 is the half recovery time defined by recovery fraction f(t1/2) ) [F(t1/2) - F(0)]/[F(∞) - F(0)] ) 0.5. F(t) is the fluorescence intensity at time t, fitted to a single-exponential function. The immobile fraction in the membrane was determined by [1 - F(tf∞)] × 100%. The fluorescence micrographs revealed that deep and irregular cracks containing aggregated fluorescently labeled lipids had formed on the surface of the aerogel (data not shown). Upon absorption of fluids, the tenuous silica network of the aerogel can undergo partial collapse due to induced capillary forces at the liquid-solid-air interface.40 Nevertheless, it was still possible to identify and characterize surface-spanning lipid membranes supported on the crack-free regions. Aerogels with stronger mechanical strength and appropriate surface modification can possibly be used to preserve the structural integrity in the hydrated state and eliminate the surface cracking.40 We are exploring this point. Figure 4a-f shows crack-free regions of the surface before bleaching and during recovery. No defects in the bilayers were apparent, except those replicating the imperfect finish of the aerogel surface created by the casting technique. Figure 5 presents the linescans of fluorescence intensity across the bleached spot for different times. Figure 6 shows the recovery curves, normalized by the intensity prior to bleaching, for supported lipid bilayers on microscope cover glass and aerogels. A diffusion coefficient of ca. 0.61 ( 0.22 µm2/s was obtained from the half-life of recovery for the membranes on aerogels, and ca. 2.46 ( 0.35 µm2/s for glass at room temperature. (39) 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. (40) Leventis, N.; Sotiriou-Leventis, C.; Zhang, G.; Rawashdeh, A.M. M. Nanoengineering Strong Silica Aerogels. Nano Lett. 2002, 2, 957-960.

Figure 5. The intensity profiles (arbitrary units) of linescans across the bleached spot in the egg PC lipid bilayers supported on the aerogel surface. The trend of increase and averaging in fluorescence counts with time was observed. The average intensity of the bleached spot recovered to ∼75% of the original (prior to bleaching) intensity.

The asymptotes of FRAP curves can be interpreted by a Brownian motion model, in which the molecular populations participating in diffusion are defined by mobile versus immobile fractions. By this notion, the lack of full recovery to unity is attributed to an immobile subpopulation that is trapped inside the bleached region. Alternatively, the partial recovery can be modeled as anomalous subdiffusion, in which the diffusion of some components in the membranes is time-dependent or constrained.41 The first model is chosen for its simplicity and applicability for our experimental setup. On the aerogel surface, a lower mobile fraction of the lipids was evident compared to the glass surface. The fluorescence intensity of the bleached membrane on the aerogel surface recovered to ∼75% relative to the membrane on glass with 100% recovery, indicating that nearly one-fourth of the lipid population was immobilized by the interactions among lipids and with the aerogel surface. (41) Feder, T. J.; Brust-Mascher, I.; Slattery, J. P.; Baird, B.; Webb, W. W. Constrained Diffusion or Immobile Fraction on Cell Surfaces: A New Interpretation. Biophys. J. 1996, 70, 2767-2773.

Fluid Biomembranes Supported on Gel Substrates

Langmuir, Vol. 20, No. 17, 2004 7237

To interpret the QCM-D results, a linear relationship between ∆f and adsorbed mass (∆m) derived from the classical Sauerbrey equation is employed:

C ∆m ) ∆f n

Figure 6. Fluorescence intensities of photobleached regions recovered in the egg PC with 3 mol % NBD-PC lipid bilayers supported on glass and aerogel surfaces. The curves were fitted to a single-exponential function for FRAP analysis. Membranes on the aerogel surface showed lower lateral fluidity and smaller mobile fraction than on glass. Both membranes were immersed in 10 mM Tris of 100 mM ionic strength and inspected by the 40× objective in an epifluorescence microscope.

Discussion An enhanced interaction between the porous surfaces and the vesicles is anticipated based on two aspects. First, the porous and microscopically rough terrain of the aerogel/ xerogel provides a large surface area for effective contact that greatly exceeds the surface area for typical glass substrates. Second, the use of hydrophilic aerogels/xerogels encourages the attractive interactions to deform the vesicles and induce fusion. One thus would expect a facilitated fusion and bilayer formation on the porous surfaces. However, our QCM-D data, as shown in Figure 3, reveal a slower vesicle fusion process on the xerogel surface. To understand these results, we note that adequate vesicle-vesicle interactions are required for the adhesive forces to promote coalescence. Moreover, fusion of vesicles without immediate rupture is considered to be an intermediate step in the bilayer formation.12 In the case of the aerogel/xerogel, the adsorbed vesicles are likely to be unevenly supported on the surface since the silica grains and pore sizes are of the same dimensions as the vesicles. The larger surface area of aerogels/xerogels also requires more adsorbed vesicles to reach a critical coverage to initiate fusion compared to on a flat surface. As a result, the aggregation of vesicles prior to fusion and rupture is less efficient, making bilayer formation slower than on the flat surface. The spreading of lipid bilayers might also be hindered by the rugged platform, as indicated by the greater length of time required to reach steady state in the QCM-D curves. Thus, the roughness exerts an opposing influence against the favorable attraction between vesicles and the silica surface and leads to the slower kinetics of the vesicle fusion process. Another possible scenario is the formation of patchy membranes as a result of inhomogeneous distribution of accessible silanol groups on the surface. Although the silanols may be distributed uniformly on the silica particles comprising the xerogel surface, the landscape is tortuous and does not uniformly contact the adsorbed vesicles with equal distances at the same height. Since the surface presents an abundance of pores, it is less effective in facilitating the vesiclemembrane transformation compared to a smooth glass surface.

where C is the mass-sensitivity constant with the value of 17.8 ng/cm2/Hz for our QCM-D crystal at 5 MHz, and n is the overtone number (1 for the fundamental and 3, 5, 7 for the overtones). First, we look at the maximum values in |∆f| and |∆D| at the extremum points, where two different phases of the vesicle fusion process are distinguished. The amount of lipids adsorbed at the transitional cusp was ∼32% greater on the xerogel surface than on the silicon oxide surface, and the dissipation was nearly 230% higher. After the steady state is reached, the asymptotic frequency shift of 29 Hz for vesicle fusion on the silicon oxide surface, assuming an average molecular weight of 768 Da for egg phosphatidylcholine (egg PC),42 corresponds to a bilayer packing at projected surface area of 50 Å2 per lipid molecule, slightly lower than the previously reported 55 Å2.9 The difference could be attributed to the additional mass from the thin hydration layer of water coupled to the surface, and possibly to a small amount of still bound, unfused vesicles. From the final frequency offset of ∼32 Hz, we estimate that approximately 10% more mass adsorbs on the xerogel surface compared to the silicon oxide surface. However, this is likely to be an underestimation of the actual mass due to neglect of the viscoelastic component of intact vesicles, which deviates from the Sauerbrey relation. The asymptotic dissipation for the membranes on a xerogel is higher (∼29%) than on glass, while both dissipations are close to zero and ensure that the loss-less film assumption for lipid bilayers is valid. For the aerogel surface, there is no clear distinction between domains of crevices and flat plateaus. Rather, the open pore structure of the interconnected silica spheres results in a surface that is better viewed as a continuous, rough, and corrugated platform. We expect the supported bilayers to follow the roughness of the surface to some extent, as a result of the balance between the bilayersurface interaction and the unfavorable cost of the bilayer bending energy. The slower diffusion observed for membranes supported on aerogel surfaces can also be addressed by the geometry of the bilayers. Presumably, the contours on the surface created unequal tensions in opposed leaflets of the bilayers due to asymmetric bending. As a reference for the extreme case, abrading scratches on a solid glass surface were shown to be effective barriers to mixing due to the costly bending energy and the membrane’s selflimiting expansion.43 Spreading of lipids on a mechanically roughened surface is much less favorable than on a smooth surface.38 This implies that the aerogel/xerogel surface does not present enough trapping force to arrest the membrane, although it does decrease the mobility as a result of nanoscale surface contours. In our analysis, the lateral diffusion in egg PC membranes supported on an aerogel is about one-fourth of their counterpart on flat glass. Again, the highly undulating surface is suspected to be the main reason for the relatively slow lateral (42) Huang, C.; Mason, J. T. Geometric Packing Constraints in Egg Phosphatidylcholine Vesicles. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 308-310. (43) Cremer, P. S.; Groves, J. T.; Kung, L. A.; Boxer, S. G. Writing and Erasing Barriers to Lateral Mobility into Fluid Phospholipid Bilayers. Langmuir 1999, 15, 3893-3896.

7238

Langmuir, Vol. 20, No. 17, 2004

mobility compared to 1-5 µm2/s for typical lipid bilayers supported on a silicon oxide surface.44 Normally, the fluidity of the lipids within the planar bilayers also depends on the hydration state of the surface, as facile realignment of water molecules accompanies the lipid motion. The structure of the water film has been studied by second harmonic generation and vibrational sum frequency spectroscopy,45-47 and pH has been shown to change the water structure in the lubricating layer on the surface. The hydrophilic aerogels made by supercritical CO2 extraction of the residual solvent are estimated to have an average of ∼5 polar hydroxy (-OH)/nm2 on the surface, a value consistent with that for other forms of silica.48,49 The hydration state of the aerogel surface in the FRAP measurement is thus expected to be similar to that of other types of silica substrates. In the fluorescence recovery series of Figure 4, some bright spots or blemishes appeared on the surface and were immobile. While the planar membrane was bleached, the intensities of these spots were diminished to a smaller extent and generally recovered to nearly the original level. QCM-D results also indicate that a higher mass was adsorbed on the surface of xerogels in the final lipid assemblies. This suggests that a combination of intact vesicle aggregates and suspended bilayers exists on the rough surface. The recovery is mainly attributed to the lipids in the bilayers, and the vesicles do not participate in the lateral exchange, similar to the case where they are adsorbed on a glass surface from highly dilute solution.12 This illustrates the effect of surface geometry on the configuration of resulting lipids. Although some vesicles were trapped on the nanoporous surface, why they do not follow the same recovery pattern and how they interact with the lipid bilayers and affect the membrane properties are still unknown. Based on the QCM-D and FRAP results, it is reasonable to assume that the porosity and tortuosity of the aerogel/ xerogel surface can compromise the attractive interactions of vesicles with silica, leading to the observed slowingdown of the fusion process (as much as 30 times) as well as of the lateral mobility (to approximately one-fourth of the solid glass value) once the membranes are formed. However, in recognition of the drawbacks of currently employed solid supports, the formation of lipid bilayers on nanoporous aerogels/xerogels immediately suggests a number of potential advantages. Most notably, the immense porosity and interconnected channels of the substrate accommodate a large volume of aqueous medium. Thus, the porous scaffold could be used as a water reservoir for many lipid- or protein-mediated membrane functions. Many membrane-associated proteins utilize the space both inside and outside the membrane to carry out biomolecular transport, and their activities have to be monitored asymmetrically across the bilayers. The unique (44) Stelzle, M.; Miehlich, R.; Sackmann, E. Two-Dimensional Microelectrophoresis in Supported Lipid Bilayers. Biophys. J. 1992, 63, 1346-1354. (45) Ong, S.; Zhao, X.; Eisenthal, K. B. Polarization of Water Molecules at Charged Interface: Second Harmonic Studies of the Silica/Water Interface. Chem. Phys. Lett. 1992, 191, 327-335. (46) Kim, J.; Kim, G.; Cremer, P. S. Investigations of Water Structure at the Solid/Liquid Interface in the Presence of Supported Lipid Bilayers by Vibrational Sum Frequency Spectroscopy. Langmuir 2001, 17, 72557260. (47) Du, Q.; Freysz, E.; Shen, Y. R. Vibrational Spectra of Water Molecules at Quartz/Water Interfaces. Phys. Rev. Lett. 1994, 72, 238241. (48) Iler, R. K. The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties, and Biochemistry; John Wiley & Sons: New York, 1979. (49) Zhuravlev, L. T. Concentration of Hydroxyl Groups on the Surface of Amorphous Silicas. Langmuir 1987, 3, 316-318.

Weng et al.

structure of aerogels/xerogels would allow manipulations on both sides of the membrane to better mimic the cytosolic and extracellular environments necessary for the functions of membrane-associated components. For example, bidirectional water transport by aquaporins could be studied with hydraulic or osmotic gradient controls.50 Since the aerogel/xerogel is permeable to aqueous media, it would serve both as the plenum and as the channel for membrane transport. For membranes supported on hard substrates such as glass, maneuvering and probing of events on both sides of the membranes is technically limited. In addition, the interactions between the membrane components and the supporting surface could also be enhanced by modifying the aerogel/xerogel composition with cytosolic-borne species so the underlying foundation for the membranes could be made protein-friendly or cytosol-mimetic. The compositions may be controlled such that organic, inorganic, or biologically active components can be embedded in bulk or modified on the surface.30,51-54 This is one of the most important characteristics to explore in order to fully realize the value of an aerogel/xerogel as a membrane support. Conclusions In conclusion, planar lipid bilayers were formed on the surface of a nanoporous silica substrate. The micron-scale homogeneity was examined by epifluorescence microscopy, and the lateral mobility was analyzed by fluorescence recovery after photobleaching. Quartz crystal microbalance-dissipation results confirmed the transformation from vesicles to membranes, while slower kinetics and higher adsorption mass compared to silicon oxide surfaces were also observed. The aerogel/xerogel has emerged as a new support for lipid membranes that could potentially offer many advantages over other solid substrates, as the porosity provides accessible space under the lipid bilayers and may allow protein insertion. This novel combination of nanoporous interface with planar lipid bilayers may provide new possibilities for designing and fabricating versatile platforms of functional supported membranes. Methods Preparation of Aerogels. Two vessels are prepared for the aerogel synthesis: beaker A contains 20 g of tetramethoxysilane (TMOS) and 22 g of methanol; beaker B contains 7.2 g of water, 18.6 g of methanol, and 100 µL of concentrated ammonium hydroxide (6 M solution, CAS 1336-21-6). The contents of beaker B are poured into beaker A with continuous stirring. Gelation usually occurs in less than an hour, after which the soft “gels” are transferred into cylindrical molds containing glass plates on the top and bottom of the cylinder that ensure a smooth surface finish. The wet gels in the mold are processed in a Polaron supercritical point drier in which the methanol solvent in the pores of the wet gel is exchanged for liquid carbon dioxide at 10-15 °C. The chamber was brought to supercritical conditions of carbon dioxide above 31 °C and 7.4 MPa to dry the solvent. (50) Agre, P.; King, L. S.; Yasui, M.; Guggino, W. B.; Ottersen, O. P.; Fujiyoshi, Y.; Engel, A.; Nielsen, S. Aquaporin Water Channels: From Atomic Structure to Clinical Medicine. J. Physiol. 2002, 524, 3-16. (51) Morris, C. A.; Anderson, M. L.; Stroud, R. M.; Merzbacher, C. I.; Rolison, D. R. Silica Sol as a Nanoglue: Flexible Synthesis of Composite Aerogels. Science 1999, 284, 622-624. (52) Buisson, P.; Hernandez, C.; Pierre, M.; Pierre, A. C. Encapsulation of Lipases in Aerogels. J. Non-Cryst. Solids 2001, 285, 295-302. (53) Power, M.; Hosticka, B.; Black, E.; Daitch, C.; Norris, P. Aerogels as Biosensors: Viral Particle Detection by Bacteria Immobilized on Large Pore Aerogel. J. Non-Cryst. Solids 2001, 285, 303-308. (54) Wallace, J. M.; Rice, J. K.; Pietron, J. J.; Stroud, R. M.; Long, J. W.; Rolison, D. R. Silica Nanoarchitectures Incorporating SelfOrganized Protein Superstructures with Gas-Phase Bioactivity. Nano Lett. 2003, 3, 1463-1467.

Fluid Biomembranes Supported on Gel Substrates After processing under these conditions for 4-5 h, aerogel samples were retrieved. Preparation of Xerogel Thin Films on QCM Crystals. The xerogel synthesis is initiated by the preparation of two solutions: (a) a solution of 10 mL of tetraethoxysilane (TEOS) + 8 mL of ethanol giving a water (MilliQ)/TEOS ratio suitable for spin coating and (b) a solution of 2 mL of water + 8 mL of ethanol + 2-3 drops of HCl as an acid catalyst. The two solutions were first mixed independently with a magnetic stirrer, and then solution b was added to solution a while stirring continuously. The final vessel was closed tightly and stirred for about 24 h in a bath maintained at 27 °C. Spin coating onto the QCM crystal was achieved by introducing the mixed solutions dropwise on top of disks that were subsequently spun at preset speeds varying from 200 to 4500 rpm. Complete coverage could be obtained within 10 s, and the thickness was varied by applying one or multiple coatings. The thinnest films of ∼0.2 µm were obtained by spin coating at 40004500 rpm for 10 s. Scanning Electron Microscopy. SEM micrographs were taken by a field-emission scanning electron microscope (Sirion, FEI Co., Portland, OR). Low accelerating voltage (3-5 kV), high scan speed, and a through-lens detector were used for ultrahighresolution image acquisition. For most inspections, the aerogels were coated with an ultrathin layer of gold-palladium to minimize charging. The images were taken at a working distance of approximately 5 mm using magnifications from 20 000× to 100 000×. Unilamellar Vesicle Preparation. L-R-Phosphatidylcholine (egg yolk PC) was purchased from Avanti Polar Lipids (Alabaster, AL). N-(7-Nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoylsn-glycero-3-phosphoethanolamine, triethylammonium salt (NBDPE) was purchased from Molecular Probes (Eugene, OR). Lipids were first mixed in chloroform, and the solvent was evaporated to form a lipid film on the wall of a round-bottomed flask. For FRAP measurements, a molar ratio of 97:3 for egg yolk PC to NBD-PE was prepared. MilliQ water (18.2 MΩ cm) was added to 5 mg/mL overall concentration. Small unilamellar vesicles were prepared by extruding the lipid mixture suspended in MilliQ water through a 30 nm polycarbonate membrane (Whatman Inc., Kent, U.K.) at least 11 times. Vesicle Fusion and Membrane Formation. The aerogel was carefully cut into a ∼20 × 20 × 10 mm block that fits into a plastic chamber for subsequent vesicle incubation. Prior to hydration, the surface was subjected to oxygen plasma cleaning (∼100 W for 4 min). Surface alkoxy groups resulting from incomplete hydrolysis of precursors and organic contaminants were removed by this treatment. Due to their immense capacity for imbibing water to nearly the same extent as the apparent volume, the aerogels need to be prehydrated prior to exposure to a vesicle suspension; otherwise, contact of vesicles with the dry porous surface would partially or fully dehydrate the lipids in the vesicles. Tris buffer (10 mM Tris, pH ∼ 8, ionic strength

Langmuir, Vol. 20, No. 17, 2004 7239 of 100 mM assisted by sodium chloride) was slowly added to fill up the container and soak the aerogel. One hour was allowed for the air trapped in the aerogel to vent. The vesicle solution was then added on top of the submerged aerogel, and 90 min was allowed for the vesicles to interact with the surface. After the incubation, thorough rinsing with equal amounts of buffers was performed at least three times to remove the free vesicles. The aerogel with its supported membrane was kept in buffer at all times. Quartz Crystal Microbalance with Dissipation. A QSense D300 (Q-Sense AB, Gothenburg, Sweden) equipped with a QAFC 301 axial flow chamber was used to conduct QCM-D measurements. AT-cut crystals (Q-Sense) of 14 mm in diameter with silicon oxide or gold surfaces were used for reference vesicle interactions and adsorption on thin xerogel films, respectively. The crystal was driven at its resonance frequency of 5 MHz, and the frequency and dissipation changes for the first three overtones at 15, 25, and 35 MHz were also monitored. To capture the characteristic dissipation, the drive circuit was short-circuited and the exponential decay of the crystal oscillation was recorded. The temperature of the Q-Sense cell was accurately controlled by a Peltier element in the cell with fluctuation smaller than ×b10.1 °C. Each QCM crystal was treated with oxygen plasma at ∼80 W for ∼4 min prior to measurements. The treatment removed organic contaminants on the surface and generally produced a more hydrophilic surface with apparent increase in wettability. All measured curves were compared against the baseline for the buffer at equilibrium. Epifluorescence Microscopy and FRAP. A Nikon Eclipse E800 upright microscope with an epifluorescence package was used to observe the membranes. 10× and 40× water-immersion objectives were used to acquire images of the membrane submerged in Tris buffer. Bleaching was achieved using a contracted field diaphragm through the 40× objective with a 100 W high-pressure mercury light source. A high-resolution aircooled CCD camera (Photometrics CoolSNAP, Roper Scientific) was used to capture the images, and MetaMorph software (Universal Imaging, CA) was used to collect image stacks and analyze the digitized fluorescence counts of the specified regions.

Acknowledgment. This work was supported primarily by the MRSEC program of the National Science Foundation under Award Number DMR-0213618 through the Center on Polymer Interfaces and Macromolecular Assemblies (CPIMA). J.J.R.S. acknowledges the Swedish Foundation for International Cooperation in Research and Higher Education for a postdoctoral fellowship. The authors declare that they have no competing financial interests. LA049940D