Lipid Bilayer Formation by Contacting Monolayers in a Microfluidic

Publication Date (Web): November 10, 2006. Copyright © 2006 ..... Toru Ide , Yuko Takeuchi , Hiroyuki Noji and Kazuhito V. Tabata. Langmuir 2010 26 (...
0 downloads 0 Views 499KB Size
Anal. Chem. 2006, 78, 8169-8174

Accelerated Articles

Lipid Bilayer Formation by Contacting Monolayers in a Microfluidic Device for Membrane Protein Analysis Kei Funakoshi,†,‡ Hiroaki Suzuki,† and Shoji Takeuchi*,†,‡

Center for International Research on Micromechatronics (CIRMM), Institute of Industrial Science (IIS), The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan, and PRESTO, JST, 4-1-8 Honcho, Kawaguchi, Saitama, Japan

Artificial planar lipid bilayers are a powerful tool for the functional study of membrane proteins, yet they have not been widely used due to their low stability and reproducibility. This paper describes an accessible method to form a planar lipid bilayer, simply by contacting two monolayers assembled at the interface between water and organic solvent in a microfluidic chip. The membrane of an organic solvent containing phospholipids at the interface was confirmed to be a bilayer by the capacitance measurement and by measuring the ion channel signal from reconstituted antibiotic peptides. We present two different designs for bilayer formation. One equips two circular wells connected, in which the water/solvent/water interface was formed by simply injecting a water droplet into each well. Another equips the cross-shaped microfluidic channel. In the latter design, formation of the interface at the sectional area was controlled by external syringe pumps. Both methods are extremely simple and reproducible, especially in microdevices, and will lead to automation and multiple bilayer formation for the highthroughput screening of membrane transport in physiological and pharmaceutical studies. Lipid bilayers and membrane proteins are essential components of cell membranes. Lipid bilayers function as a barrier to the external environment, whereas membrane proteins transduce chemical signals or transport ions/molecules selectively in and out of cells. In pharmaceutical studies, membrane proteins constitute the most important research targets because drugs are * To whom correspondence should be addressed. E-mail: takeuchi@ iis.u-tokyo.ac.jp. † The University of Tokyo. ‡ PRESTO, JST. 10.1021/ac0613479 CCC: $33.50 Published on Web 11/10/2006

© 2006 American Chemical Society

recognized by the membrane proteins to trigger a cascade of intracellular events that leads to a change in the state of the cell. However, functions of the greater part of membrane proteins are still unclear compared to water-soluble proteins. Membrane proteins are more difficult to study compared to their water-soluble counterparts as they are insoluble in water and change their original conformations and activities when isolated from cell membranes. In conventional biological research, membrane transport phenomena are mostly studied in living cells using molecular probes or electrophysiological methods (e.g., radio isotope technique or patch clamping methods1-4). Although these methods make it possible to detect the physiological properties and molecular interactions of cell membranes, quantitative measurements of the functions of single membrane proteins are still difficult due to the presence of heterogeneous biomolecules in the membranes. Artificial lipid bilayers, a simplified model of the membrane consisting of purified or synthesized phospholipids, have been used as platforms to analyze single-species-specific membrane proteins for almost five decades. There are two major platformss lipid vesicles and planar bilayers. Lipid vesicles (liposomes)5,6 are spherical bilayer structures containing an aqueous compartment; these vesicles are commonly used in biology and biochemistry due to their ease of preparation and handling. On the other hand, the planar bilayer, also known as black lipid membrane (BLM), is usually formed across a tiny aperture opened in a solid support.7-9 Advantages of BLM methods over lipid vesicles (1) (2) (3) (4) (5) (6)

Neher, E.; Sakmann, B. Nature 1976, 260, 799-802. Neher, E. Science 1992, 256, 498-502. Levitan, E. S.; Kramer, R. K. Nature 1990, 348, 545-547. Magin, R. L.; Morse, P. D. Biochim. Biophys. Acta 1983, 760, 357-362. Edwards, K. A.; Baeumner, A. J. Talanta 2006, 68, 1421-1431. Lasic, D. D. Liposomes: From Physics to Applications; Elsevier: New York, 1993; p 201.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 8169

methods are that the electrodes for electrophysiological study are readily accessible, and the exchange of reagents is possible. BLMs are often formed with either the painting method, in which a lipid solution (organic solvent containing phospholipid) is applied across a tiny aperture that separates two aqueous compartments,10,11 or the Langmuir-Blodgett method, in which the lipid monolayers at the water/air interface are brought together by raising them above the aperture.12,13 With planar bilayers, we can perform membrane studies in a precisely defined environment, e.g., composition of buffer, lipids, and membrane potential.14 However, the BLM method suffers from fragility, unsteadiness, and low reproducibility, making it difficult to be applied in highthroughput systems for pharmaceutical screenings. Recently, miniaturized analytical systems, such as microfluidic or microarray systems, have greatly facilitated high-throughput and high-sensitivity biomolecular analyses.15-17 These technologies allow precise fluidic control, fine patterning of biomolecules, and process automation. For membrane protein analysis, there have been attempts to reconstitute artificial bilayers in micromachined apertures with precisely controlled diameters.18-23 Although the microfabricated apertures make membranes more stable than the conventional handmade apertures, the formation of the membrane is still a skilled and laborious operation (e.g., painting methods). We believe that there are more suitable and accessible methods to form membranes in microdevices and have reported several methods using microfluidic techniques.24-26 In this paper, we describe an alternative approach to form a bilayer without apertures using simple fluidic control. The principle of this approach is shown in Figure 1. At the interface of water and organic solvent containing amphipathic molecules (phospholipids), the monolayer assembles spontaneously.27 When two interfaces are driven to come in contact, two monolayers form a bilayer, having hydrophobic acyl chains face each other. This principle is suitable to perform in microchannels because the (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27)

Tien, H. T.; Ottova, A. L. J. Membr. Sci. 2001, 189, 83-117. Walde, P.; Ichikawa, S. Biomol. Eng. 2001, 18, 143-177. Ide, T.; Ichikawa, T. Biosens. Bioelectron. 2005, 21, 672-677. Miller, C. Ion Channel Reconstitution; Plenum Press: New York, 1986. Mueller, P.; Rudin, D. O.; Tien, H. T.; Wescott, W. C. Nature 1962, 194, 979-980. Montal, M.; Mueller, P. Proc. Natl. Acad. Sci. U.S.A. 1972, 69, 35613566. Dambacher, K. H.; Fromherz, P. Biochim. Biophys. Acta 1986, 861, 331336. Oiki, S.; Danho, D. O.; Madison, V.; Montal, M. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 2393-2397. Haab, B. B.; Dunham, M. J.; Brown, P. O. Genome Biol. 2001, 2, 1-13. MacBeath, G.; Schreiber, S. L. Science 2000, 289, 1760-1763. Dufva, M. Biomol. Eng. 2005, 22, 173-184. Schmidt, C.; Mayer, M.; Vogel, H. Angew. Chem., Int. Ed. 2000, 39, 313733140. Neubert, H. Anal. Chem. 2004, 76, 327A-330A. Sinclair, J.; Pihl, J.; Olofsson, J.; Karlsson, M.; Jardemark, K.; Chiu, D. T.; Orwar, O. Anal. Chem. 2002, 74, 6133-6138. Fertig, N.; Tilke, A.; Blick, R. H.; Kotthaus, J. P. Appl. Phys. Lett. 2000, 77, 1218-1220. Li, X.; Klemic, K. G.; Reed, M. A.; Sigworth, F. J. Nano Lett. 2006, 6, 815819. Washizu, M.; Kurosawa, O.; Kurahashi, H.; Katoh, A. IEICE, Trans. Electron. 1995, E78-C (2), 157-161. Suzuki, H.; Tabata, K. V.; Kato-Yamada, K.; Noji, H.; Takeuchi, S. Lab Chip 2004, l4, 502-505. Suzuki, H.; Tabata, K. V.; Noji, H.; Takeuchi, S. Langmuir 2006, 22, 19371942. Suzuki, H.; Tabata, K. V.; Noji, H.; Takeuchi, S. Biosens. Bioelectron. In press. Langmuir, I. J. Am. Chem. Soc. 1917, 39, 1848-1906.

8170 Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

Figure 1. Conceptual diagram of lipid bilayer formation by contacting two lipid monolayers. Two aqueous droplets in organic solution containing lipid molecules are brought together, and a bilayer forms.

contact procedure can be easily achieved by microfluidic control. Also, when performed in planar chip configuration, a bilayer is formed in vertical direction, allowing direct observation of membrane transport under the microscope. Here, we demonstrate two configurations to realize lipid bilayer formation in a microfluidic device using this principle. The first configuration is the “double well chip” that equips two overalpping wells. By simply injecting two water droplets in each well that is already filled with lipid solution, their interfaces come in contact to form a bilayer. The second configuration is the “cross-channel chip,” where the two interfaces controlled by syringe pumps are brought together at the cross section of two microchannels. In both cases, we confirmed that the layer at the interface is a bilayer based on capacitance measurements and managed to record ion channel signals through peptide channels reconstituted into the bilayer. EXPERIMENTAL SECTION Reagents. 1,2-Diphytanoyl-sn-glycero-3-phosphocholine (DphPC) was purchased from Avanti Polar Lipids, Inc. Sodium chloride, n-decane, R-hemolysin (RHL), and gramicidin A (gA) were all obtained from Sigma-Aldrich (Tokyo, Japan). A 100 mM solution of KCl in Milli-Q water was used as an electrolyte, and 25 mg of DphPC in 1 mL of n-decane was used as a lipid solution. Chip Design, Procedure, and Fabrication. All the microfluidic chips were fabricated by machining poly(methyl methacrylate) (PMMA) plastic plates using an automated CAD/CAM modeling machine (Modia, MM-100).26 Design and bilayer formation procedure of the double well chip is shown in Figure 2. Two wells, 4 mm in diameter and 2 mm in depth, are curved on a 5-mm-thick PMMA plastic plate. They have an overlapped area, where the width of the intersectional plane is 2 mm. The volume of each chamber is 22 µL. To prevent swelling of PMMA due to absorption of organic solvent, a 5-µm-thick Parylene layer was deposited on the whole chip. The procedure is illustrated in Figure 2b and c. First, 15 µL of lipid solution was introduced into the chambers. Then, two droplets of electrolyte, 15 µL in volume, were injected into each chamber with a glass capillary pipet. Thus, interfaces of two droplets made contact at the intersection of two wells, and a bilayer formed spontaneously. This design provides the simplest operation for the lipid bilayer formation. The schematic and image of the cross-channel are shown in Figure 3. A cross-shaped fluidic channel of 0.5 mm in both width and depth was fabricated in a 2-mm-thick PMMA plate. Another PMMA plate was bonded to close the channel. Two opposing inlet

Figure 2. (a) Overview of the double well chip, consisting of two wells connected. (b) Procedure of the bilayer formation. Two droplets are injected in each well filled with lipid solution, forming a water/ lipid solution/water interface at the section. (c) Schematic diagram of the experimental setup. Electrodes are inserted in each droplet, and membrane current is measured using a patch clamp amplifier.

ports were connected to syringes containing electrolyte, and one port in the crossing channel was connected to that with lipid solution. In this design, the position of the aqueous solution interfaces can be controlled precisely using syringe pumps, enabling foramtion and removal of the water/lipid solution/water interfaces in a controlled manner. The bilayer formation procedure is illustrated in Figure 4. First, electrolyte was injected from two opposing inlets of one channel, and their fronts were stopped at the cross section. Next, lipid solution was injected into the crossing channel, displacing the air that divided the electrolyte at the intersection (Figure 4a). The interfaces between the water and lipid solution were then contacted by pushing opposing electrolytes to form a bilayer (Figure 4b). An important advantage of this approach is its ease of redoing the process. When the bilayer breaks and two electrolytes are fused, we can repeat the process immediately by simply injecting the lipid solution into the crossing channel again (Figure 4c). Optical and Electrical Monitoring. A digital microscope (Keyence, VH-5000C) with zoom lens and inverted microscope (Olympus, IX-71) systems were used for image recording. For the electrical measurement of the membrane capacitance and electric current, a patch clamp amplifier (Nihon koden, CEZ-2400) and a digital data acquisition system (Axon Instruments, Digidata 1322A and pCLAMP 9.2) were employed. In both configurations, the Ag/AgCl electrodes were inserted into two aqueous compartments as depicted in Figures 2 and 3. The whole setup was placed in a Faraday cage to minimize the electrical noise. RESULTS AND DISCUSSION Bilayer Formation in a Double Well Chip. Photographs of the droplets in contact in a double well chip are shown in Figure 5. When n-decane without lipid molecules were used as an organic solution, we observed that the water droplets fused immediately after injection. With organic solution containing lipid molecules,

Figure 3. (a) Schematic of the cross-channel chip. Two aqueous solutions with lipid solution in between make contact at the cross section to form a lipid bilayer. (b) Photograph of the cross channel chip.

the interface did not rupture as shown in Figure 5a, where the colored and noncolored droplets stayed unmixed for a long period of time (more than 1 h). This system was found to be stable, and formation of multiple membranes is readily achieved. The top view of the interface is shown in Figure 5b. This layer is composed of three layers: lipid monolayer, n-decane, and lipid monolayer.28 Figure 5c shows the image of direct observation of the membrane from its normal direction. This image was taken from the side of the chip using the digital microscope. The chip was cut at the plane parallel to the membrane and intersecting the center of one well. A circular edge was observed at the interface; the edge was similar to the Plateau-Gibbs border that comes from the refraction due to the sudden change of the membrane thickness.29 This image indicates that the area inside the border is much thinner than outside. Since the lipid bilayer works as a capacitor in an electrical circuit, measurement of the specific capacitance gives a rough estimate of the membrane thickness. The current across the (28) Benz, R.; Frohlich, O.; Lauger, O.; Montal, M. Biochim. Biophys. Acta 1975, 374, 323-334. (29) Mingins, J.; Pethica, B. A. Langmuir 2004, 20, 7493-7498.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

8171

Figure 6. Transient current across the bilayer membrane when 10 mVp-p square signals were applied. (a) Global time response. Magnified section showing the (b) initial state and (c) stable state during the bilayer formation process. Figure 4. Bilayer formation process in the cross-channel chip. (a) Two aqueous solutions are injected into opposite inlets of one channel, and their interfaces are stopped at the entrance of the intersection. Lipid solution is injected into the crossing channel to fill the space in between. (b) The interfaces contacted are made by infusing the aqueous solusions using syringe pumps. (c) The process can be repeated by flushing the section with lipid solution. (d) Schematic representation of the regulation of opening/closing gramicidin A channels by infusing/withdrawing the aqueous solutions.

Figure 7. Recording of single channel current through R-hemolysin nanopores incorporated into the lipid bilayer formed in the double well chip.

Figure 8. Microscopic images of the interfaces of two aqueous solutions detached/contacted in the cross-channel device. The volume outside aqueous solution is filled with lipid solution.

Figure 5. (a) Photo of the double well chip with two different color droplets. They do not mix as the lipid bilayer forms at the interface of two droplets. Microscopic images of the water/lipid solution/water interface viewed from the top (b) and normal direction (c).

membrane when a (10 mV peak-peak square signal at 100 Hz was applied is shown in Figure 6a. After contacting the droplets at t ) 0, the capacitive transient current started to increase and reached constant at t ) 150 s. The short time plot at the beginning of membrane formation and at the stable state is shown is Figure 6b. Membrane capacitance was calculated by integrating those 8172 Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

signals in a half-period and was 25 nF at the stable state. Although we have to determine the area of the bilayer precisely to achieve the specific capacitance (capacitance per unit area), the measurements in Figures 5 and 6 were obtained in different experiments, and we did not monitor the current and bilayer area simultaneously. However, from many observations such as Figure 5c, the bilayer area was estimated in between 3 and 5 mm2 (a bilayer area larger than the cross-sectional area of two wells can be possible when the bilayer bulged out due to a slight difference in the hydrostatic pressure between the right and left wells). Assuming the area within this range, specific capacitance was 0.50.8 µF cm-2. This value agrees well with the capacitance of reconstituted bilayers of 10-nm thickness reported in the literature (typically 0.6 µF cm-2).30

Figure 9. Recording of the membrane current when gramicidin A ion channels were incorporated into the bilayer. Bilayer formation and accompanying channel opening events can be controlled by the infusing and withdrawing aqueous solutions.

Ion Channel Recording in Double Well Chip. In the previous section, it was shown that the membrane capacitance was of the same order as that of reconstituted bilayers, indicating the thickness was thin enough. The direct proof of being a functional lipid bilayer is to observe functions of the membrane proteins incorporated. For this purpose, we attempted reconstitution of the peptide ion channel, RHL, from Staphylococcus aureus, which passes ions and small molecules. To insert them into the bilayer, 25 µg/mL RHL was mixed in one aqueous droplet. The ion current across the membrane was recorded under the voltage clamp at 50 mV. A few minutes after contacting droplets, the current steps that corresponded to the incorporation of single RHL pores into the lipid bilayer were detected as 80-100 pS conductance as shown in Figure 7.31 This result indicates that the functional lipid bilayer was formed at the interface of two droplets, and this membrane can be used for the functional study of reconstituted membrane proteins. Bilayer Formation in the Cross-Channel Chip. Figure 8 shows the interface of two electrolytes separated by the organic solvent containing phospholipids; this interface is formed in the cross-shaped fluidic channel with the method described in Figure 4. The membrane capacitance was measured by the same method described in Figure 6. When the interfaces of monolayers contacted, the peak value of the transient current increased dramatically and the capacitance at the steady state was 130 pF. To achieve the specific capacitance, we estimated the contacting area to be ∼5 × 10-2 mm2 (320 µm × 160 µm; vertical length was assumed from the ratio of channel width and height) from the picture in Figure 8. Using this value, the specific capacitance was 0.25 µF cm-2, which is smaller but on the same order as the capacitance derived in the double well chip. Considering the fact that the contacting area in this device is 2 orders of magnitude (30) Fujiwara, H.; Fujihara, M.; Ishiwata, T. J. Chem. Phys. 2003, 119, 67686775. (31) Futaki, S.; Zhang, Y.; Kiwada, T.; Nakase, I.; Yagami, T.; Oiki, S.; Sugiura, Y. Bioorg. Med. Chem. 2004, 12, 1343-1350.

smaller than the double well device, the actual size of the bilayer is expected to be smaller than the contacting area due to the curvature of the droplets. Using this fluidic device, we measured the ion channel currents of gA; this is a hydrophobic peptide that forms a nanopore in the lipid bilayer along with dimerization and passes monovalent ions. The ion currents can be generated by the opening and closing events of the peptide. In the experiment, electrolyte containing gA molecules was used as two aqueous solutions. When the voltage over 70 mV was applied, the membrane repeatedly broke and two electrolytes fused together; the membrane did not break when the voltage was less than 70 mV. Figure 9 shows the time series of current across the bilayer with gA. After contacting the interfaces of lipid monolayers, the current started to flow across the membrane without breaking at 63-mV clamping voltage. It corresponds to the formation of gA nanopores along with expansion of the bilayer. When interfaces were separated by withdrawing the aqueous phases with the syringe pump, the current dropped to zero. This result proves the existence of the bilayer, since gA lets the current flow through only when it is in a bilayer. The total conductance of gA channels at the steady state is 280 nS. Since the conductance of a single gA channel is 20 pS,32 the number of gA forming nanopores is ∼104 in this bilayer. In the experiment, forming and disassembling a lipid bilayer can be controlled by pushing and withdrawing the aqueous phase; thus it is possible to determine the start and end of the membrane transport phenomena for quantitative analysis. Also, this technique allows the cross-sectional observation of both compartments separated with the lipid bilayer. CONCLUSIONS We have developed the simple and facile methods for lipid bilayer formation by contacting two lipid monolayers. In the double (32) Al-Momani, L.; Reiss, P.; Koert, U. Biochem. Biophys. Res. Commun. 2005, 328, 342-347.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

8173

well chip, the lipid bilayer was formed by contacting two water droplets separated by organic solvent containing lipid molecules. It was confirmed as a functional bilayer by measuring the membrane capacitance and by reconstituting the nanopores of RHL. We also applied this principle to the microfluidic device for the timing control of bilayer formation. And this method allowed the continuous formation of lipid bilayers. These methods will be a useful technology for the high-throughput membrane protein analysis chip using multiarray lipid bilayer system. Furthermore, since this method allows the direct imaging of cis and trans

8174

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

electrolyte compartments, it will be useful for observation of the membrane transports in physiological and pharmaceutical studies. ACKNOWLEDGMENT We thank Hiroyuki Noji at Osaka University for helpful discussions. This research was supported by Precursory Research for Embryonic Science and Technology (PRESTO), JST. Received for review July 24, 2006. Accepted October 20, 2006. AC0613479