Incorporation of Na+,K+-ATP-ase into the Thiolipid Biomimetic

Incorporation of Na+,K+-ATP-ase into the Thiolipid Biomimetic Assemblies via the Fusion of Proteoliposomes. Agnieszka Żebrowska, and Paweł Krysińsk...
3 downloads 10 Views 251KB Size
Download PDF
Langmuir 2004, 20, 11127-11133

11127

Incorporation of Na+,K+-ATP-ase into the Thiolipid Biomimetic Assemblies via the Fusion of Proteoliposomes Agnieszka Z˙ ebrowska and Paweł Krysin´ski* Laboratory of Electrochemistry, Department of Chemistry, University of Warsaw, 02-093 Warsaw, Pasteura 1, Poland Received May 28, 2004. In Final Form: September 13, 2004 The black lipid membranes (BLMs) are artificial membrane systems that have been widely used in the study of different biological processes. In this paper the planar bilayer lipid membranes have been used to study the behavior of thiolipid moleculessdipalmitoyl-phosphatidyl-ethanolamine-mercaptopropionamide (DPPE-MPA) and cholesteryl 3-mercaptopropionate (Chs-MPA)sas compared to classical BLM made of natural lipids. We present our experiments on black thiolipid bilayer (BTM) formation from a thiolipid solution and basic results of pump currents generated by sodium-potassium pumpsNa+,K+ATP-asesintroduced to such bilayer systems via proteoliposome adsorption with subsequent fusion. Our results imply that no substantial difference exists between BLMs formed from classical lipids and those made from thiolipids used in this study. The same thiolipid molecules were subsequently used for the formation of covalently bound, tethered bilayer lipid membranes (t-BLMs) on polycrystalline gold electrodes. Similarly, as in the case of BLMs, we took advantage of proteoliposome adsorption/fusion to obtain a t-BLM system with reconstituted enzyme. The vesicle fusion on hydrophobic or hydrophilic substrates is one of the main ways to obtain a bilayer system with incorporated biological species. In this paper we present also our preliminary results of electrochemical experiments using rapid solution exchange technique on such t-BLMs systems and their comparison with painted solid supported membranes (SSMs) and BLMs. We have also followed the process of vesicles fusion onto thiolipid monolayer by means of in situ atomic force microscopy in tapping mode (TM-AFM). On the basis of these experiments, we conclude that DPPEMPA and Chs-MPA molecules used in our experiments preserve lipid properties, allowing for at least partial reconstitution of Na+,K+-ATP-ase into such t-BLMs. On the other hand, the relatively compact organization on polycrystalline gold and the hydrophobic nature of the first monolayer of tethered thiolipids slows down the proteoliposome fusion onto such monolayers and consequently hinders the protein insertion. However, this effect can be overcome by mechanical stimulus that facilitates proteoliposome delamination onto the self-assembled monolayer.

Introduction Fabrication of functional interfaces between biomolecules (e.g., proteins and cells) and electrodes that can be addressed electronically (e.g., metal and semiconductors) is attractive both from the scientific and technological points of view. The primary limitation of this approach is the interaction between biological entities and conducting substrates that typically alters the biological activity of the material. Therefore, to create a biologically amenable interface, electrochemical connection with the adsorbate needs to be carefully devised.1-5 To overcome these limitations, layered molecular structures tethered onto metallic or semiconducting surfaces (ITO, silicon-based), mimicking the cell membranes, have been devised.1,4-13 Great interest has been focused on selfassembled films that are attached to the substrate by formation of a covalent linkage between the self-assembled molecules and the solid support, yielding structures of (1) Willner, I.; Katz, E. Angew. Chem., Int. Ed. 2000, 39, 1180-1218. (2) Guidelli, R.; Aloisi, G.; Becucci, L.; Dolfi, A.; Moncelli, M. R.; Buoninsegni, F. T. J. Electroanal. Chem. 2001, 504, 1-28. (3) Sackmann, E. Science 1996, 271, 43-48 (4) Sinner, E.; Knoll, W. Curr. Opin. Chem. Biol. 2001, 5, 705-711 (5) Krysin´ski, P.; Tien, H. T.; Ottova, A. Biotechnol. Prog. 1999, 15, 974-990. (6) Plant, A. L. Langmuir 1999, 15, 5128-5135 (7) Knoll, W.; Frank, C. W.; Heibel, C.; Naumann, R.; Offenhauser, A.; Ruhe, J.; Schmidt, E. K.; Shen, W. W.; Sinner, A. Rev. Mol. Biotechnol. 2000, 74, 137. (8) Laibnis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152-7167. (9) Peng, Z.; Tang, J.; Han, X.; Wang, E.; Dong, S. Langmuir 2002, 18, 4834-4839.

high long-term stability. Monolayers of alkanethiols on gold are probably the most widely used and best characterized of all self-assembled films to date.6,14,15 The deposition of a second phospholipid monolayer on top of the first chemisorbed alkanethiol monolayer can be accomplished by several different preparation techniques, including liposome fusion,14,16-20 which exploit the attractive hydrophobic interactions between the alkanethiol and phospholipid hydrocarbon tails, resulting in the placement of the polar heads of the phospholipids in the direction of the aqueous solution. In more recent research, the alkanethiol monolayer has been replaced by a monolayer of phospholipid molecules whose polar heads are covalently linked to a hydrophilic “spacer” terminated with an anchoring thiol group (thiolipids), responsible for the (10) Naumann R.; Schiller, S. M.; Giess, F.; Grohe, B.; Hartman, K. B.; Kaercher, I.; Koeper, I.; Luebben, J.; Vasilev, K.; Knoll, V. Langmuir 2003, 19, 5435-5443. (11) Silin, V. I.; Wieder, H.; Woodward, J. T.; Valincius, G.; Offenhauser, A.; Plant, A. L. J. Am. Chem. Soc. 2002, 124, 14676-14683. (12) Krysin´ski, P.; Blanchard, G. J. Langmuir 2003, 19, 3875-3882. (13) Kelepouris, L.; Krysin´ski, P.; Blanchard, G. J. J. Phys. Chem. B 2003, 107, 4100-4106. (14) Twardowski, M.; Nuzzo, R. G. Langmuir 2003, 19, 9781-9791. (15) Finklea, H. O. Electroanal. Chem. 1996, 19, 109-335. (16) Lang, H.; Duschl, C.; Vogel, H. Langmuir 1994, 10, 197-210. (17) Cornell, B. A.; Braach-Maksvytis, V. L. B.; King, L. C.; Osman, P. D. J.; Raguse, B.; Wieczorek, L.; Pace, R. J. Nature 1997, 387, 580583. (18) Williams, L. M.; Evans, S. D.; Flynn, T. M.; Marsh, A.; Knowles, P. F.; Bushby, R. J.; Boden, N. Langmuir 1997, 13, 751-757. (19) Becucci, L.; Guidelli, R.; Liu, Q.; Bushby, R, J.; Evans, S. D. J. Phys. Chem. B 2002, 106, 10410-10416. (20) Krysinski, P.; Z˙ ebrowska, A.; Palys, B.; Lotowski, Z. J. Electrochem. Soc. 2002, 149, 189-194.

10.1021/la048675t CCC: $27.50 © 2004 American Chemical Society Published on Web 11/03/2004

11128

Langmuir, Vol. 20, No. 25, 2004

chemisorption on gold.10,16-20 The role of these hydrophilic spacer moieties is to decouple the bilayer lipid membranes from the surface of the metal, providing hydrophilic ionic reservoirs on both sides of black lipid membranes (BLMs), meeting more closely the behavior of interfacial regions of native biomembranes. We report here the behavior of new thiolipid molecules that can be used both for the formation of classical bilayer lipid membranes (BLMs) and for the formation of bilayers covalently tethered to the surface of gold (t-BLMs). Comparing to commonly used polyethyleneoxythiol derivatives of various length, peptides or polymers serving as spacer moieties anchoring lipid monolayer to the surface of gold, our thiolipid molecules contain only relatively short mercaptopropionic acid molecule covalently bound via either amide or ester bond to dipalmitoylphosphatidylethanolamine and cholesterol, respectively. The advantage of using a relatively short spacer chain was that such derivatized lipid molecules were still capable of forming free-standing bilayer lipid membranes. After electrochemical characterization of these molecules, we demonstrate the activity of Na+,K+-ATP-ase reconstituted into such host bilayers via proteoliposome adsorption with subsequent fusion. Na+,K+-ATP-ase is an ion translocating membrane protein that is found in the plasma membrane of almost all animal cells. It is also one of the most thoroughly investigated enzymes of charge translocating membrane proteins. Under physiological conditions Na+,K+-ATP-ase transports three Na+ ions out of the cell and two K+ ions into the cell at the expense of hydrolysis of one molecule of ATP. Due to its stoichiometry the overall reaction cycle is electrogenic, resulting in a charge separation across the membrane. During its ion translocating cycle, the pump can assume two main conformations, E1 and E2, with ion binding sites facing the cytoplasm and the extracellular medium, respectively. Binding the Na+ ions takes place during the formation of phosphoenzyme E2P in the presence of ATP, while K+ transport to the cytoplasmic side is correlated with the decay of E2P and the formation of the ground-state E1. This sequential ion binding/transport mechanism, generally described by the Post-Albers model,21 is supported by a large amount of experimental evidence22 and can be monitored electrochemically by time-resolved solution exchange or concentration jump techniques.24-27 We demonstrate also that the introduction of relatively short, mercaptopropionic functionalities into lipid molecules does not affect significantly their capabilities to form unsupported planar bilayer (thio-)lipid membraness BTMssyet rendering these molecules suitable to form t-BLMs on solid support. The Na+,K+-ATP-ase activity is compared for both. Experimental Section Chemicals. All reagents were of the highest purity commercially available without further purification: CHCl3 (Sigma), NaCl (Sigma), KCl (Sigma), MgCl2 (Sigma), imidazole (Sigma), dithiothereitiol (DTT, 99.5%, Roth, Karlsruhe, Germany), HCl (21) Post, R. L.; Kume, S.; Tobin, T.; Orcutt, B.; Sen, A. K. J. Gen. Physiol. 1969, 54, 306-326. (22) La¨uger, P. Electrogenic Ion Pumps; Sinclair Associates: Sunderland, MA, 1991; pp 168-224 (23) Krysin´ski, P.; Z˙ ebrowska, A.; Michota, A.; Bukowska, J.; Becucci, L.; Moncelli, M. R. Langmuir 2001, 17, 3852-3857. (24) Gropp, T.; Cornelius, F.; Fendler, K. Biochim. Biophys. Acta 1998, 1368, 184-200. (25) Fendler, K.; Grell, E.; Haubs, M.; Bamberg E. EMBO J. 1985, 4, 3079-3085. (26) Pintschovius, J.; Fendler, K. Biophys. J 1999, 76, 814-826. (27) Pintschovius, J.; Fendler, K.; Bamberg, E. Biophys. J 1999, 76, 827-836.

Z˙ ebrowska and Krysin´ ski (Sigma), ATP (Fluka, Neu-Ulm, Germany), octadecylamine (98%, Riedel-DeHaen AG, Seelze-Hannover, Germany), adenosine-5′triphosphate magnesium salt (95%, Sigma-Aldrich Chemicals, Deisenhofen, Germany), Na2 salt of NPE caged ATP (Na2 salt of P3-1-(2-nitro)phenylethyladenosine 5′-triphosphate, Calbiochem), orthovanadate trisodium salt (99%, Aldrich), n-decane (Sigma), and Hellmanex detergent (Sigma). Aqueous solutions were prepared from water of high purity (Seral water, Milli-Q). The two thiolipids, dipalmitoyl-phosphatidyl-ethanolamine-mercaptopropionamide (DPPE-MPA) and cholesteryl 3-mercaptopropionate (Chs-MPA), were synthesized according to the procedure described previously.23 Instrumentation. In the case of BTM experiments, all signals were amplified with a current amplifier (gain, 107 V/A) followed by a voltage amplifier (gain,102; bandwith, 1 kHz). The amplified signal was then filtered with a 500 Hz filter and recorded with a digital oscilloscope as described in detail previously.24,25 To photolyze the caged ATP, light pulses of an excimer laser (Lambda Physik, EMG 100, 308 nm wavelength, 10 ns pulse duration) were focused on the membrane. The energy density at the membrane was about 150 mJ/cm2. A detailed description of the setup used in rapid solution exchange experiments on t-BLMs is given elsewhere.26 The typical procedure consisted of three steps: (a) washing the cuvette with the nonactivating (ATP-free) buffer for ca. 5 s; (b) injecting the activating buffer for 2 s; (c) removing the activating substrate from the cuvette with ATP-free buffer for ca. 5 s. The changes of solution were controlled by specific configuration of the electrical 3 × 3 valve (Teflon Valve 360T05, NResearch Inc., Maplewood, NJ). To avoid additional disturbance, the whole setup was enclosed in a Faradaic cage. Similarly, as in the case of BTM, the signal was amplified with current amplifier, filtered, and recorded with a digital oscilloscope assisted with a personal computer as described in prior publications.26,27 Commercially available Nanoscope IIIa from Digital Instruments was used in the tapping mode to collect in situ AFM images (atomic force microscopy) presented in this paper. The AFM cantilevers used in these studies had a nominal spring constant of 0.6 N/m.

Materials and Methods Electrode Preparation. Gold electrodes for rapid solution exchange technique were prepared as described before.24,26 Briefly, glass microscopic plates were cleaned with alcohol and subsequently purged with nitrogen to remove any remaining debris. Next, they were immersed in a cleaning solution of 2% Helmanex in water, ultrasound washed, and sonicated twice for 10 min. After that they were dried in a dryer at 80 °C. Next the plates were covered with a 5 nm layer of chromium and a 150 nm layer of gold by evaporation at a vacuum pressure smaller than 5 × 10-6 mbar. The thickness of metal layers was measured using a quartz crystal microbalance. Such freshly evaporated electrodes were used to adsorb thiolipid monolayers. The gold electrodes of 99.99% purity used for AFM studies were commercially available polycrystalline gold slides evaporated on a glass. They were used after flame annealing in a gasair flame directly after purging off the dust with an argon stream. Formation of BTMs from Thiolipids: Electrochemical Characterization. Black thiolipid membranes were formed in a septum hole with an area of ca. 1 mm2 of a thermostated (24 °C) Teflon cuvette, filled with approximately 1.3 mL (each side) of supporting buffer. The membrane bathing solution contained 25 mM imidazole, 3 mM MgCl2, and 0.2 mM DTT, with a pH adjusted to 7.0 by 1 M HClaq. Additionally, it contained NaCl in concentrations varying from 130 to 100 mM and KCl of 0-30 mM, respectively, yielding a final salt concentration of 130 mM. Teflon cuvette was placed in a Faradaic cage. The hole in the Teflon cell septum was pretreated with 0.5% DPPE-MPA solution in hexane. Next the membrane forming thiolipid solution containing 13.5 mg/mL DPPE-MPA, 1.5 mg of Chs-MPA, and 0.025% (w/v) octadecylamine in decane was spread over the septum of the Teflon cell. BTM formation was observed with an optical microscope. In comparison to the traditional planar membranes prepared from lipids, such black thiolipid membranes were much thicker and unstable at the beginning of the formation process. Moreover, they often broke easily. However, it was

Thiolipid Biomimetic Assemblies Scheme 1. Diagram Showing the “Ideal” Organization of All Components of the Systems Presented in This Papera

a For the sake of simplicity, all spacer molecules attached to the phospholipid headgroups in BTM are shown stretched, facing the aqueous solutions.

observed that thiolipid-based membranes turned to BTM relatively quickly. Their stability significantly increased with the time needed for reorganization and thinning. Measurements of two electrical parametersscapacitance Cm and resistance Rmsmonitored the electrical properties of black thiolipid membranes. Planar lipid bilayerssBLMs, formed from a typical phospholipid and interposed between two identical aqueous solutionssusually have a thickness of about 6 nm and are characterized by two electrical parameters: Cm equal to about 0.5 µF/cm2 and Rm greater than 108 Ω‚cm2.29-31 In comparison to traditional planar bilayers, BTMs formed from forming solution containing lipids with covalently bound short thiol molecules were characterized by a lower resistance value Rm of ca. 107 Ω‚cm2 and two times smaller capacitance value Cm ≈ 0.25 µF/ cm2. These somehow worse electrical parameters never approached the values expected for standard BLMs. The difference in these parameters could be explained by a change in total thickness of the membrane that was caused by the presence of a mercaptopropionic moiety and/or by worse organization due to the presence of organic solvent from the forming solution (decane). Despite this, usually within 2 h after the addition of membrane forming solution to the cuvette, the values of capacitance and resistance of the membrane appeared to be stable and BTMs were ready for use in subsequent experiments. Thus, on the basis of microscopic and electrical measurements, we can expect that the thiolipid molecules in BTMs formed a structure with hydrocarbon tails directed toward the interior of the bilayer and hydrophilic spacers facing the two aqueous solutions bathing the membrane. Part of thiolipid molecules is confined to the border of the Teflon septum. This is shown in Scheme 1. The membrane was connected to an external measuring circuit by Ag/AgCl and platinum electrodes via poliacrylamide gel salt bridges filled with the same buffer. The obtained signals were amplified, filtered, and recorded on a digital oscilloscope as described before.24 Preparation of Proteoliposomes Containing Na+, K+-ATP-ase. Proteoliposomes containing Na+,K+-ATPase were prepared according to an earlier procedure.25 Briefly, the proteoliposomes were prepared as follows: 3.85 mg of pig kidney (28) Seifert, K.; Fendler, K.; Bamberg, E. Biophys. J. 1993, 64, 384391. (29) Tien, H. T. J. Theor. Biol. 1967, 16, 97. (30) Tien, H. T., Ed. Membrane Biophysics, 1st ed.; Elsevier Science B. V.: Amsterdam, 2000; p 639. (31) Anzai, K.; Ogawa, K.; Ozawa T.; Yamamoto H. J. Biochem. Biophys. Methods 2001, 48, 283-291.

Langmuir, Vol. 20, No. 25, 2004 11129 Na+,K+-ATP-ase (generous gift form Prof. Klaus Fendler, Max Planck Institute for Biophysics, Frankfurt am Main, Germany) was added to the mixture of 1.2 mL of histidine (150 mM, Fluka), 0.03 mL of EDTA (100 mM), and 32 mL of distilled water, to obtain 2.372 mg/mL final concentration of the enzyme. To this mixture, cooled to 15 °C, 0.6 mL of 45 mM CHAPS buffer solution was added with stirring and then the mixture was placed in the ice bath and left for 10 min. Then the resulting mixture was centrifuged at 50 000 rpm and the supernatant collected and placed for 1 h in a 4 °C bath while mixing at 1000 rpm. Next, 0.5 mL of enzyme solution was added to 2 mL of previously prepared liposome suspension (“dry” method, phosphatidylcholine) with 57 µL of CHAPS buffer (180 mM) at 0 °C and incubated while stirring at 4 °C for 1 h with biobeads. After removal of biobeads, the mixture was left overnight at 4 °C while stirring. Such prepared proteoliposomes were ready for subsequent experiments. The activity of the enzyme was assayed directly after defrosting and sonication of obtained sample by means of the ATP-ase activity test commonly used and described in the literature.32 Usually, the enzyme activity was approximately 10 µM Pi/(mg of protein × min). Adsorption/Fusion of Proteoliposomes Containing Na+,K+-ATP-ase into the BTMs Formed from Synthesized Thiolipids. The process of reconstitution of Na+,K+-ATP-ase into BTMs formed from thiolipids was the same as that described previously for the traditionally obtained BLMs.24 This process relied on bringing into contact the thiolipid bilayer with suspension of proteoliposomes containing reconstituted Na+,K+-ATPase (protein concentration, 2-3 mg/mL), prepared as described above. A 15 µL aliquot of proteoliposome suspension was added under stirring to one compartment of a Teflon cell. After leaving the suspension in contact with BTM for 20 min, the stock solution of caged ATP was added to a final concentration of 100 µM. Inhibition experiments were carried out using 30 mM sodium orthovanadate stock solution, which was added to the membrane bathing solution to a final concentration of 1 mM Monolayer Preparation. Spontaneously assembled thiolipid monolayers were typically prepared at room temperature by immersing freshly cleaned gold-covered glass slides in an approximately 1.2 × 10-3 M coating solution of a given thiolipid in chloroform, for ca. 3-5 days. The electrodes were then thoroughly rinsed, first with chloroform, ethanol, and then distilled water to remove the unbound material. In the case of experiments with rapid solution exchange technique after incubation with thiolipids, a part of evaporated gold electrode not designed to work as membrane-modified electrode but only as an electrical connection was insulated by covering it with nail polish. The working area of such prepared electrodes was approximately 7.85 × 10-3 cm2. The reference electrode was Ag/AgCl electrode separated from the streaming solution by a polyacrylamide gel bridge. Before the addition of vesicle suspension, similarly as in the case of BTM experiments, we determined the electrical parameters of such formed monolayers. We found that capacitance values of DPPE-MPA modified electrodes were approximately two times higher than in the case of solid supported membranes (SSMs, 300-500 nF/cm2 28), widely used in such experiments and varied around Cm ≈ 1 µF/cm2. This capacitance was in agreement with the expected value for lipid monolayers.30 Moreover, the second characteristic parametersthe conductance in the case of DPPE-MPA modified electrodeswas much higher than in the case of SSM and had the range Gm ≈ 110-200 nS/ cm2. This corresponds to the resistance values of Rm ≈ 5-9 MΩ cm2. Despite these somehow smaller resistance values, we decided to use our thiolipid monolayers in future experiments with Na+,K+-ATP-ase. Adsorption/Fusion of Proteoliposomes Containing Na+,K+-ATP-ase onto the Thiolipid Monolayers. In this case the second layer was formed according to a previously described procedure26 by the addition of 30 µL of sonicated suspension of proteoliposomes in an appropriate buffer to the Teflon cell with monolayer-covered electrode. The suspension was left in contact (32) Esmann, M.; Skou, J. C. Biochim. Biophys. Acta 1988, 944, 344350.

11130

Langmuir, Vol. 20, No. 25, 2004

with modified electrode in the cell for a required time (0.5-2.5 h), depending on the experiment. After that, the suspension was replaced with proteoliposome-free buffer solution. This should lead to the adsorption and fusion of proteoliposomes onto the monolayer.9-11,14,33-35 Even though the liposome and proteoliposome fusion with various surfaces to form supported bilayers is frequently reported in the literature, the details of such interaction are poorly understood, particularly for the case of hydrophobic surfaces.11,36-40 Vesicle adsorption was studied by Keller and Kasemo36 with use of quartz crystal microbalance on different surfaces. They found that while lipid bilayers were usually formed on hydrophilic surfaces of glass, lipid monolayers were formed on hydrophobic surfaces. They also reported intact vesicles adsorbed on an oxidized gold surface. Evans et al.37 in their studies of interaction of egg phosphatidylcholine (PC) liposomes with thioderivatized cholesterol monolayers mixed with thiooligoethyleneglycol reported that a single layer of PC forms on hydrophobic film, yielding a hybrid bilayer structure tethered to gold. Glazier et al.38 have used vesicle fusion onto various monolayers of alkane thiol derivatives (with or without hydrophilic spacer between the gold and the hydrocarbon moiety) to reconstitute the pore-forming membrane proteinsR-hemolysin. They reported at least partial reconstitution of this protein into the hybrid bilayer structure, particularly if such a bilayer was separated from the surface of gold by hydrophilic spacer molecules. Rao et al.39,40 formed the cell membrane hybrid bilayers that were continuous over large areas by spontaneous reorganization of vesicles either from red blood cell ghosts or monkey kidney COS-1 cells. Here, we found the process of adsorption and fusion of vesicles containing Na+,K+-ATP-ase on thiolipid monolayer much slower than that reported for the case of solid supported, hydrophilic membranes.26 On the basis of our experiments that will be discussed later in this paper, we estimated this time to be ca. 2-3 h, a result comparable with literature data on similar systems.41,42

Results and Discussion According to the generally accepted mechanism of the pump cycle,21,22,24 a typical current response generated by Na+,K+-ATP-ase activated by ATP concentration jump, in the absence of K+, consists of three phases: a rapid rise (τ1 ≈ 5-10 ms), a slower decay (τ2 ≈ 15-30 ms), and a very slow component with opposite amplitude (τ3 ≈ 100300 ms). After addition of K+ ions to the bathing solution, the slow phase turns into a phase with the same amplitude as fast components. Upon further increase of the K+ concentration, the stationary pumping increases the amplitude of the slowest decaying phase. Figure 1 shows the capacitive current response for BTMs with reconstituted protein in the absence of K+ ions that is attributed to E1PfE2P conformational changes and ATP/ caged ATP binding and exchange reactions.24,37 When the process was initiated with the release of 100 µM caged ATP, the transport activity of Na+,K+-ATP-ase is exhibited as a transient ionic current flow. Since K+ ions are absent, only the Na+ binding/transport branch of the pump cycle is operative and the protein is blocked in (33) Kalb E.; Frey S.; Tamm, L. K. Biochim. Biophys. Acta 1992, 1103, 307-316. (34) Puu G.; Gustafson I. Biochim. Biophys. Acta 1997, 1327, 149161. (35) Jass J.; Tja¨rnhage T.; Puu, G. Biophys. J. 2000, 79, 3153-3163 (36) Keller, C. A.; Kasemo, B. Biophys. J. 1998, 7, 1397-1402. (37) Evans, S. D.; Sharma, R.; Ulman, A. Langmuir 1991, 7, 146161. (38) Glazier, S. A.; Vanderah, D. J.; Plant, A. l.; Bayley, H.; Valincius, G.; Kasianowicz, J. J. Langmuir 2000, 16, 10428-10435. (39) Rao, N. M.; Plant, A, L.; Silin, V.; Wight, S.; Hui, S. W. Biophys. J. 1977, 73, 3066-3077. (40) Rao, N. M.; Silin, V.; Ridge, K. D.; Woodward, J. T. Anal. Biochem. 2002, 307, 117-130. (41) Twardowski, M.; Nuzzo, R. G. Langmuir 2003, 19, 97819791. (42) Twardowski, M.; Nuzzo, R. G. Langmuir 2004, 20, 175-180.

Z˙ ebrowska and Krysin´ ski

Figure 1. Current transients generated by Na+,K+-ATP-ase reconstituted into the bilayer thiolipid membranes (BTMs) following ATP concentration jump in the absence of K+ ions ([Na+] ) 130 mM) (a) and after the addition of 1 mM VO43- (b); background current (c).

Figure 2. Current transients generated by Na+,K+-ATP-ase reconstituted into the bilayer thiolipid membranes (BTMs) following ATP concentration jump in the presence of 30 mM K+ ions ([Na+] ) 100 mM) (a) and after the addition of 1 mM VO43(b); background current (c). For comparison, the inset shows current transients both in the absence (curve a) and in the presence of K+ (curve b).

E2P conformation. Its charge transporting behavior is terminated, reaching the background level with no steadystate current observed. After the addition of a well-known pump blocking agent, VO43- (up to 1 mM final concentration), to the same side as caged ATP, the magnitude of the observed signal diminishes, as expected. However, the current signal could not be eliminated in the presence of even 1 mM VO43-, a behavior observed also for the classical BLMs.43 Because orthovanadate inhibits the pump from the cytoplasmic side, the inhibition experiments demonstrate that the enzyme population with the cytoplasmic side facing the aqueous solution is responsible for the observed current transient. Different system responses were obtained when potassium ions were present in the buffer solution, allowing for the full cycle of conformational changes and ion transfer of Na+,K+-ATP-ase. This is shown in Figure 2, where the observed current response is characteristic for ion transporting behavior of reconstituted protein.44 (43) Fendler, K.; Drose, S.; Altendorf, K.; Bamberg, E. Biochemistry 1996, 35, 8009-8017. (44) Fendler, K.; Grell, E.; Bamberg, E. FEBS Lett. 1987, 224, 8388. (45) Fendler K.; Jaruschewski. S.; Hobbs A.; Albers A.; Froehlich J. J. Gen. Physiol. 1993, 102, 631-666. (46) Vansteenkiste, S. O.; Davies, M. C.; Roberts, C. J.; Tendler, S. J. B.; Williams, M. Prog. Surf. Sci. 1998, 57, 95-136. (47) Ohlsson, P.-A.; Tja¨rnhage, T.; Herbai, E.; Lo¨fas, S.; Puu, G. Bioelectrochem. Bioenerg. 1995, 38, 137-148. (48) Shibata-Seki, T.; Masai, J.; Tagawa, T.; Sorin, T.; Kondo, S. Thin Solid Films 1996, 273, 297-303. (49) Diociaiuti, M.; Molinari, A.; Ruspantini, I.; Gaudiano, M. C.; Ippoliti, R.; Lendaro, E.; Bordi, F.; Chistolini, P.; Arancia, G. Biochim. Biophys. Acta 2002, 1559, 21-31.

Thiolipid Biomimetic Assemblies

Figure 3. Current transients generated by Na+,K+-ATP-ase reconstituted into the tethered bilayer lipid membranes (tBLMs) following ATP concentration jump in the absence of K+ ions (a) and in the presence of K+ ions (b); background current - (c).

While the current transient is still dominated by the E1PfE2P conformational transition and Na+ binding, there is a very pronounced difference in the decreasing part of the signal. This signal now decays slowly in time, showing a tendency to achieve a stationary value, a behavior characteristic for the full cycle of pumping action of reconstituted protein and denoting the occurrence of a number of pump turnovers. Fitting our data presented in Figures 1 and 2 to the mechanism described above44,55 yielded the values of τ1 ≈ 14, τ2 ≈ 16, and τ3 ≈ 270 ms for the absence of K+ ions. For the presence of K+ ions these values were 14, 17, and 1300 ms, respectively. The observed signals were qualitatively analogous to those reported for classical BLMs.25,44,45 The reason that the signal ultimately vanishes in time is the depletion of ATP that was released by 10 ns laser pulse from its caged form and subsequently consumed by Na+,K+-ATP-ase. The protein is inhibited by addition of 1 mM Na3VO4 and exhibits similar response as in the absence of potassium ions. Thus, according to our data, the mercaptopropionic terminal functionalities that modify DPPE and Chs molecules do not affect significantly their capability to form planar, unsupported BTMs. Therefore, we decided next to use the DPPE-MPA molecules to modify the electrodes with self-assembled monolayer in further experiments with rapid solution exchange technique to determine the biomimetic properties of membranes formed. Figure 3 presents the electrical signals recorded after ATP concentration jump at a given concentration of (a) [Na+] ) 130 mM and [K+] ) 0 and (b) [Na+] ) 100 mM and [K+] ) 30 mM (Figure 3csbackground current). As in the case of BTM we verified the origin of the signal by addition of sodium orthovanadate to the activating solution containing ATP (not shown). Almost complete inhibition of the signal confirmed its source. Even though our spacer molecules decoupling the BLM from the surface of metal were relatively short, we assumed the presence of at least a hydrophilic region (50) Pignataro, B.; Steinem, C.; Galla, H. J.; Fuchs, H.; Janshoff, A. Biophys. J. 2000, 78, 487-498. (51) Albersdo¨rfer, A.; Feder, T.; Sackmann, E. Biophys. J. 1997, 73, 245-257. (52) Albersdo¨rfer, A.; Bruinsma, R.; Sackmann, E. Europhys. Lett. 1998, 42, 227-231. (53) Leonenko, Z. V.; Carnini, A.; Cramb, D. T. Biochim. Biophys. Acta 2000, 1509, 131-147. (54) Kumar, S.; Hoh, J. H. Langmuir 2000, 16, 9936-9940. (55) Tadini-Buoninsegni, F.; Nassi, P.; Nediani, C.; Dolfi, A.; Guidelli, R. Biochim. Biophys. Acta 2003, 1211, 70-80.

Langmuir, Vol. 20, No. 25, 2004 11131

Figure 4. Current transients as in Figure 3b, after the mechanical stimulation of delamination adsorbed proteoliposomes by the stream of air bubbles (trace a) and after the addition of 1 mM VO43- (trace b); background current (c).

accessible to K+ ions, large enough to allow for the full cycle of protein function after its reconstitution into the bilayer. Therefore, we expected to observe similar dependence of the signal shape triggered by ATP concentration jump upon the addition of K+ ions as in the case of BTMs (cf. Figure 2). In particular, we expected to observe much slower decay of the signal within the same time window, due to the appearance of a stationary current superimposing the capacitive transients. Yet initially, after a traditional way of addition of vesicles containing Na+,K+ATP-ase to the cell containing thiolipid-modified electrode, we could not observe the stationary current that could be assigned to the full cycle of both: the conformational changes and ion transfer of the pump. Only typical, fastdecaying currents were observed, as is shown in Figure 3a,b. This response is qualitatively analogous to those reported in the literature for the case of solid supported membranes.26,27,55 In the absence of K+, the transient signal in Figure 3a is characterized by a fast rise with a time constant of 9 ms and a decay with a time constant of 14 s. One can also notice a very slow component with a reverse amplitude, and a time constant of ca. 600 ms. As expected, these values are similar to those obtained in analogous experiments with intact, Na+,K+-ATP-ase containing proteoliposomes.26,55 Thus, the observed response can be related to the conformational transition and Na+ binding/ translocation reaction due to the presence of ATP. Apparently, K+ ions from bathing solution cannot reach their binding site on that part of Na+,K+-ATP-ase that is directed toward the proteoliposome interior (proteoliposome internal solution is potassium-free) and the pump is virtually stopped after the Na+-dependent steps (Figure 3b). However, by passing the air bubbles through the pipet tip placed close to the modified electrode, we could trigger the system response that could be assigned to the pumping action of reconstituted Na+,K+-ATP-ase. This is shown in Figure 4. Upon the addition of K+ to the bathing solution the recorded current transients now show a slow decaying component (time constant τ ≈ 2000 ms) with a tendency to reach a stationary value. This type of signal can be attributed to a number of Na+,K+-ATP-ase turnovers.26,27,55 We explain this observation by a mechanical stimulation of delamination of proteoliposomes that were adsorbed on a hydrophobic surface. The delamination process may be a consequence of a shear force applied by a fast flow of multiple, very small air bubbles in direct proximity to the adsorbed proteoliposomes. This suggests again that initially there is only an adsorption of hydrophilic proteoliposomes on hydrophobic DPPE-MPA monolayer without any significant fusion and protein reconstitution.

11132

Langmuir, Vol. 20, No. 25, 2004

Z˙ ebrowska and Krysin´ ski

Figure 5. Tapping-mode AFM 3-D topographic images of (a) bare gold surface after flame annealing (x, 500 nm/div; z, 250 nm/div) and (b) bare gold surface covered with thiolipid (DPPE-MPA) monolayer (x, 500 nm/div; z, 250 nm/div).

Figure 6. Tapping-mode AFM 3D sequential topographic images of bilayer formation via the liposome fusion (a-c) or proteoliposome (d) on DPPE-MPA monolayer: (a) a few minutes after the liposome suspension was replaced by a buffer; (b) 30 min later; (c) 1 h later; (d) 2 h after proteoliposome fusion [(a, d) x, 1000 nm/div; z, 100 nm/div; (b, c) x, 500 nm/div; z, 100 nm/div).

Apparently, the mechanical stimulus facilitates the rupture of proteoliposomes and their fusion onto the underlying thiolipid monolayer, so that the K+ ions can move toward their appropriate binding side of Na+,K+ATP-ase. However, the intensity of the observed signal is much smaller as compared with that in Figure 3. Also, in the absence of K+ ions the intensity is much smaller (not shown). We attribute this result to the partial removal of adsorbed proteoliposomes by the stream of air bubbles. However, the protein denaturation is also possible. To verify the above conclusions, we monitored the process of adsorption/fusion of liposomes and proteoliposomes containing Na+,K+-ATP-ase onto the thiolipid monolayers by means of atomic force microscopy. It is well-known that the tapping-mode atomic force microscopy (TM-AFM) is an excellent tool for the in situ characterization of biological membranes and associated processes.46 This technique enables one to visualize the processes in their natural aqueous environment.34,35,47-50 Parts a and b of Figure 5 show the TM-AFM threedimensional topographic images of bare gold surface after flame annealing and the same electrode after adsorption of thiolipid monolayer. For both cases the electrode was immersed in a buffer solution. As expected, the bare gold surface seems to be sufficiently smooth in a desirable range to study the adsorption of liposomes. The same electrode with a deposited self-assembled DPPE-MPA monolayer was also quite flat, without visible regions of multilayer

structures or domains of DPPE-MPA (thickness change expected for a monolayer formation). Sequential imaging of bilayer formation by liposome fusion is shown in Figure 6a-c. It was initiated by scanning the surface a few minutes after the liposome suspension was replaced by a pure buffer. As we can see in Figure 6a, the surface is covered by dome-shaped structures adsorbed on the thiolipid monolayer. It is also apparent that the whole surface is covered with biological material. Parts b and c of Figure 6 present roughly the same region a half-hour and 1 h later, respectively. As can be seen from the 3-D TM-AFM images, the surface becomes much smoother and some of the domeshaped structures disappear in time, suggesting their fusion with the monolayer. However, after more than 1 h we can still observe some domelike structures. Our result is in agreement with the previous studies of liposome adsorption on surfaces modified with Langmuir-Blodgett films.47 The presence of domelike structures is even more pronounced when similar experiments were carried out for the case of proteoliposomes adsorption/fusion onto thiolipid-modified gold electrode, Figure 6d. These results confirm our prediction that after the adsorption of proteoliposomes we do not observe the disappearance of liposome-like structures within the time scale of our experiments (2-3 h). They seem to be quite stable on the monolayer surface after the initial adsorption. From the obtained images we can calculate the size of adsorbed

Thiolipid Biomimetic Assemblies

lipid structures. Thus, the average diameter of such an adsorbed structure was ca. 200 nm, and the smallest of adsorbed structures were about two times smaller. All liposome diameters varied within the range of 100-500 nm, which is in good agreement with literature values for large unilamellar liposome structures (LUVs) with uncontrolled size.30,50-52 Moreover, despite of the fact that, due to the tip penetration into the sample, the LUV contour recognized by the cantilever tip is imperfect, we were able to determine a general shape of such structures adsorbed on modified electrode in an aqueous buffer. As can be seen from the 3-D images shown in Figure 6, one can observe vesicles that are ruptured, as well as spread to a different extent and even the intact structures. These images show that, after good coverage of the electrode surface with a blocking, well-ordered monolayer that meets the expectations and requirements of a good biomimetic lipid layer, the process of proteoliposome splitting/fusion is very slow and inefficient. Our results are in agreement with previously published studies of Puu et al.34,35,47 as well as Diociaiuti et al.49 In these works it was found that, in the case of proteoliposome adsorption on the solid surface modified with hydrophobic monolayer, the presence of the large protein molecule and hydrophilicity of liposomes hinders the complete spreading over the hydrophobic monolayer and the process of vesicles disintegration and fusion proposed by Kalb at al.33 is very slow. According to Leonenko et al.,53 this is consistent with a protein-induced stiffening of the vesicle membrane. As we show here, vesicle fusion can be stimulated mechanically (e.g., air bubbles), finally leading to a bilayer structure. However, our AFM experiments show that even after elongated time, we can still observe domelike structures that can be assigned to relatively intact liposomes and proteoliposomes adsorbed on the thiolipid monolayer.54 Conclusions On the basis of our results we found the process of proteoliposome adsorption and fusion to be a useful way

Langmuir, Vol. 20, No. 25, 2004 11133

of preparation of t-BLMs based on the thiolipid used in our studies. We showed that membrane pump protein reconstituted into such a system via proteoliposome fusion retains its activity. We visualized the process of adsorption of liposomes and proteoliposomes and found their tendency to collapse and fuse on hydrophobic, DPPE-MPA modified gold surface, especially after mechanical stress. Moreover, the TM-AFM imaging process confirmed our electrochemical results, which showed that the process of proteoliposome fusion on hydrophobic monolayer takes much longer than for the case of hydrophilic, painted solid supported membranes reported in the literature. The present work may contribute to the understanding of mechanisms of such system formation via proteoliposome adsorption and fusion onto the self-assembled monolayers. We found that the method of proteoliposome adsorption on the thiolipid monolayer can be a useful tool for protein reconstitution into a bilayer system, but relatively good packing of the monolayer that leads to its highly hydrophobic surface prevents the facile formation of uniform bilayers. We think that by increasing the length of the hydrophilic spacer, less densely packed monolayers can be formed, ensuring a larger aqueous reservoir between the electrode surface and the tethered bilayers. Thus, such a structure could accommodate both hydrophilic and hydrophobic parts of the protein being introduced by proteoliposome fusion.

Acknowledgment. We wish to express our gratitude to Prof. Klaus Fendler, Max Planck Institute for Biophysics, Frankfurt am Main, Germany, for providing resources and scientific help within his group to A.Z˙ . in doing proteoliposome fusion experiments. LA048675T