Immobilization of Acetylcholinesterase in Lipid Membranes Deposited

Nov 18, 2010 - Signaling membrane proteins, such as tyrosine kinase receptors, cytokine receptors, neurotransmitters, and G protein-coupled receptors,...
0 downloads 16 Views 4MB Size
pubs.acs.org/Langmuir © 2010 American Chemical Society

Immobilization of Acetylcholinesterase in Lipid Membranes Deposited on Self-Assembled Monolayers )

Eftim Milkani,† Aung M. Khaing,†,‡ Fei Huang,§ Daniel G. Gibson, Scott Gridley,^ Norman Garceau,^ Christopher R. Lambert,*,† and W. Grant McGimpsey*,†,§ Bioengineering Institute, ‡Department of Biomedical Engineering, §Department of Chemistry and Biochemistry, and Department of Biology and Biotechnology, Worcester Polytechnic Institute, 100 Institute Road, Worcester, Massachusetts, United States 01609, and ^Blue Sky Biotech, 60 Prescott Street, Worcester, Massachusetts, United States 01605 )



Received May 13, 2010. Revised Manuscript Received October 15, 2010 Human red blood cell acetylcholinesterase was incorporated into planar lipid membranes deposited on alkanethiol self-assembled monolayers (SAMs) on gold substrates. Activity of the protein in the membrane was detected with a standard photometric assay and was determined to be similar to the protein in detergent solution or incorporated in lipid vesicles. Monolayer and bilayer lipid membranes were generated by fusing liposomes to hydrophobic and hydrophilic SAMs, respectively. Liposomes were formed by the injection method using the lipid dimyristoylphosphatidylcholine (DMPC). The formation of alkanethiol SAMs and lipid monolayers on SAMs was confirmed by sessile drop goniometry, ellipsometry, and electrochemical impedance spectroscopy. In this work, we report acetylcholinesterase immobilization in lipid membranes deposited on SAMs formed on the gold surface and compare its activity to enzyme in solution.

Introduction Signaling membrane proteins, such as tyrosine kinase receptors, cytokine receptors, neurotransmitters, and G protein-coupled receptors, are targets of roughly half of the 100 best-selling therapeutic drugs on the market.1 The discovery and screening of pharmaceuticals that interact with these types of membrane proteins is challenging due to their hydrophobic nature and limited solubility and loss of function in aqueous solutions. In other words, if the protein is removed from the membrane it may no longer respond to drug targets in the same way as when it is bound to the membrane. Phospholipids are the most abundant lipids in cell membranes. Due to their amphiphilic nature, they form spherical lipid bilayers known as liposomes or vesicles. Phospholipid vesicles are known to fuse with the surface and form planar membranes.2-5 This occurs when lipid vesicles are heated to a temperature above that of the phase transition of the phospholipid, resulting in the formation of a lipid monolayer (one leaflet) or bilayer (both leaflets) on the surface.6 The lipid monolayer is formed when vesicles are exposed to a hydrophobic surface. It has been demonstrated that exposing lipid vesicles to a hydrophobic selfassembled monolayer (SAM) on gold causes vesicles to rupture and fuse with the surface, resulting in the formation of a lipid monolayer.7 Lipid bilayer membranes are formed when vesicles are exposed to hydrophilic surfaces. Silicon oxide and mica surfaces, which are hydrophilic, have been shown to allow the formation of a *Authors to whom correspondence should be addressed.

(1) Terstappen, G. C.; Reggiani, A. Trends Pharmacol. Sci. 2001, 22, 23–26. (2) Mueller, P.; Rudin, D. O.; Tien, H. T.; Wescott, W. C. Nature 1962, 194, 979–980. (3) McConnell, H. M.; Watts, T. H.; Weis, R. M.; Brian, A. A. Biochim. Biophys. Acta 1986, 864, 95–106. (4) Plant, A. L. Langmuir 1993, 9, 2764–2767. (5) Reimhult, E.; Hoeoek, F.; Kasemo, B. Langmuir 2003, 19, 1681–1691. (6) Tero, R.; Takizawa, M.; Li, Y.-J.; Yamazaki, M.; Urisu, T. Langmuir 2004, 20, 7526–7531. (7) Plant, A. L. Langmuir 1999, 15, 5128–5135. (8) Egawa, H.; Furusawa, K. Langmuir 1999, 15, 1660–1666. (9) Keller, C. A.; Glasmastar, K.; Zhdanov, V. P.; Kasemo, B. Phys. Rev. Lett. 2000, 84, 5443–5446.

18884 DOI: 10.1021/la103333c

lipid bilayer on their surface, after exposure to a suspension of lipid vesicles.8,9 A hydrophilic SAM would be expected to provide a good substrate for vesicle fusion and formation of a lipid bilayer. Lipid monolayers are expected to accommodate membrane proteins that span only to one leaflet of the cell membrane via their hydrophobic domain. The ability of lipid monolayers on hydrophobic SAMs to accommodate integral proteins that span both leaflets is debatable, as SAMs on gold are expected to be more strongly packed (therefore less fluid) than the hydrophobic tails of the lipid monolayer. Lipid bilayers are analogous to a cellular membrane permanently fixed to a surface. They are predicted to accommodate integral membrane proteins. Incorporation of the whole transmembrane protein could be challenging if it contains domains on both sides of the membrane. However, if a protein construct contains the hydrophobic domain and only one outer domain (the ligand binding or the cytoplasmic domain), once inserted in the bilayer membrane, it will interact with the surroundings and fold so that it has a conformation and function similar to that of the whole protein which resides in the cell membrane. In addition, distinct domains which reside on one side of the membrane can be engineered to be lipidated so they will act like lipid-anchored membrane proteins. Providing that the surface-bound protein maintains its normal function, the lipid monolayer or bilayer membrane could be implemented in a surface-based high-throughput screening system, which would employ standard assays and surface characterization techniques to test activating or inhibiting agents against the protein of interest. In this work, lipid monolayer and bilayer membranes were formed using unilamellar lipid vesicles of DMPC (1,2-dimyristoyl-sn-glycero-3-phosphocholine). Phosphatidyl choline lipids are among the most common lipids in mammalian cell membranes; therefore, they would provide a favorable template for the incorporation of membrane proteins and their proper conformation and function. Lipid monolayers were formed on SAMs of hydrophobic alkanethiols of three different lengths, and the formation of the membrane was monitored in situ using electrochemical

Published on Web 11/18/2010

Langmuir 2010, 26(24), 18884–18892

Milkani et al.

Article

impedance spectroscopy. The same method was used to confirm the formation of lipid bilayer on hydrophilic SAMs prepared on gold substrates. Some of the hydrophilic SAMs presented alkyl chains to anchor the bilayer membrane. An enzyme containing a lipid tail, red blood cell acetylcholiesterase (AChE), was immobilized on the lipid layer, and its activity was tested with a standard photometric assay. AChE has been immobilized on substrates in the past to be used as a biosensor.10-13 In addition, red blood cells have been fused on hydrophobic SAMs carrying with them AChE and showing its activity on the surface.14 In this study, we show the immobilization of AChE after the generation of the lipid membrane. This work demonstrates the activity of amphiphilic AChE on lipid layers generated on hydrophobic and hydrophilic SAMs on gold surfaces, and compares that activity to enzyme in solution. Incorporation of membrane proteins on planar phospholipid monolayers or bilayers mimicking cell membranes will generate a platform for studying membrane protein functions or activities such as ligand binding, protein folding, and phosphorylation, using a number of surface characterization techniques, including surface plasmon resonance, electrochemical impedance spectroscopy, and quartz crystal gravimetry, in addition to the standard protein activity assays. The combination of a homogeneous solution-based activity assay and surface-based physical association measurements would be a powerful combination for high-throughput screening of new pharmaceuticals.

Experimental Details Materials. All chemicals and solvents were reagent grade or better and used as received. 11-Amino-1-undecanethiol hydrochloride (AUT) was purchased from Dojindo Molecular Technologies (Rockville, MD). Acetylcholinesterase (AChE), amphiphilic form, from human erythrocytes (EC 3.1.1.7, type C0663, 0.22 mg/ mL, MW 80 000 Da), 11-mercaptoundecanoic acid (MUA), and 1,2-dimyristol-rac-glycero-3-phosphocholine (DMPC) were purchased from Sigma (St. Louis, MO). All other chemicals and reagents were obtained from Alfa Aesar (Ward Hill, MA). Trisbuffered saline (TBS) was prepared as a 20 mM Tris and 150 mM NaCl solution, pH 7.0 or 7.5. Deionized water was obtained from a Millipore Synergy UV system (Billerica, MA). Gold-coated glass slides were obtained commercially from Evaporated Metal Films (Ithaca, NY). The float glass slides (25 mm  75 mm  1 mm) are coated with 50 A˚ of a chromium adhesion layer followed by 1000 A˚ of gold. Phosphotungstic acid, Formvar-coated 150 mesh copper grids, and Kodak electron microscope film were purchased from Electron Microscopy Sciences (Hatfield, PA). Alkanethiol Monolayer Preparation. Prior to monolayer formation, the gold slides were cut to the desired size and cleaned by immersion in a piranha solution (70% concentrated sulfuric acid, 30% concentrated hydrogen peroxide) at 90 C for 10 min (Caution: piranha reacts violently with organic compounds and should not be stored in closed containers). The slides were washed thoroughly with distilled water, followed by absolute ethanol, and dried in a stream of nitrogen. The slides were further cleaned in oxygen plasma for 45 s using Plasma Prep II from SPI Supplies (West Chester, PA) and were used immediately. Carboxyl- and amino-terminated alkanethiol SAMs were prepared as described by Wang et al.15 The carboxyl-terminated (10) Nikolelis, D. P.; Simantiraki, M. G.; Siontorou, C. G.; Toth, K. Anal. Chim. 2005, 537, 169–177. (11) Rehak, M.; Snejdarkova, M.; Hianik, T. Electroanalysis 1997, 9, 1072–1077. (12) Puu, G.; Gustafson, I.; Artursson, E.; Ohlsson, P. A. Biosens. Bioelectron. 1995, 10, 463–476. (13) Rodriguez-Mozaz, S.; Lopez de Alda, M. J.; Marco, M.-P.; Barcelo, D. Talanta 2005, 65, 291–297. (14) Rao, N. M.; Plant, A. L.; Silin, V.; Wight, S.; Hui, S. W. Biophys. J. 1997, 73, 3066–3077. (15) Wang, H.; Chen, S.; Li, L.; Jiang, S. Langmuir 2005, 21, 2633–2636.

Langmuir 2010, 26(24), 18884–18892

surfaces were further modified by immersing in an aqueous solution of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (100 mM) and N-hydroxysuccinimide (20 mM) for 15 min at room temperature followed by overnight incubation in 1 mM dodecylamine solution in anhydrous ethanol at room temperature. Amino-terminated surfaces were modified by incubating overnight in a 1 mM dodecanoic acid and 5 mM dicyclohexyl carbodiimide solution in anhydrous ethanol at room temperature. Preparation of Lipid Vesicles. Lipid vesicles were formed by the injection method,16 which has been reported to provide similar results to the extrusion method.4 Briefly, a 10 mL volume of TBS was transferred in a 2-cm-diameter glass tube, and it was incubated in a temperature-controlled water bath at 37 C, above the phase transition temperature (23 C) of the phospholipid. While stirring the buffer in the glass tube, a solution of DMPC in absolute ethanol (12.5 mg in 0.5 mL) was slowly injected with a 22 gauge syringe needle into the buffer solution at a 17 μL/min flow rate using a syringe pump. The injection took place approximately 2 cm below the surface of TBS. The lipid vesicle suspension (at a final DMPC concentration of 2 mM) was used immediately after formation or stored at 4 C for later use. The vesicle suspension used for AChE immobilization experiments was spun at 100 000 g at room temperature for 15 min to remove any vesicle aggregates and then diluted four times with TBS. Transmission Electron Microscopy (TEM) Imaging. Vesicle size distribution was determined by transmission electron microscopy using negative staining, in which the subject is surrounded and outlined by an electron-dense stain. Samples were prepared by mixing 0.25 mL of vesicle suspension in TBS with 0.75 mL of 1.0% phosphotungstic acid (PTA) in distilled water. After 5 min, the vesicles were spun down at 11 000 g for 5 min to concentrate them, and 0.75 mL of the supernatant was discarded. The remaining 0.25 mL was vortexed to resuspend the concentrated vesicles. Small drops (10 μL) of the suspension were placed on Formvar (polyvinyl formal)-coated copper grids (150 mesh) and allowed to dry. The grids were then observed at 80 kV accelerating voltage in a JEOL 100CX electron microscope, where the electron-dense PTA surrounding the vesicles outlined their shapes. Vesicles were photographed on Kodak 4489 electron microscope film. After development by conventional methods, film was scanned at 2400 dpi with a backlit scanner and the digitized negatives were converted to positive and adjusted for contrast with graphics software. Contact Angle Goniometry. Sessile drop contact angle measurements were made using a Rame-Hart model 300 goniometer (Netcong, NJ). Measurements were obtained using 1 μL drops of deionized water deposited on the substrates using the Automated Dispensing System accessory coupled to the goniometer. Images were obtained by an integrated digital camera, and the entire system was under computer control using Rame-Hart’s DROPimage standard software package. The software automatically provides contact angle measurements once the liquid is dispensed. Ellipsometry. Ellipsometric measurements were obtained with a Manual Photoelectric Rudolph 439L633P ellipsometer (Rudolph Instruments, Fairfield, NJ). The measurements were taken at a 70 angle of incidence using a He/Ne laser (632.8 nm wavelength). The thickness calculations were obtained with the software package which determines the thickness based on optical constants of the substrate. A bare gold substrate was used to determine the optical constants of gold, and the results were compared to values previously reported in the literature.17,18 Values for the extinction coefficient and refractive index of the samples were assumed to be (16) Batzri, S.; Korn, E. D. Biochim. Biophys. Acta, Biomembr. 1973, 298, 1015– 1019. (17) Chechik, V.; Stirling, C. J. M. The chemistry of organic derivatives of gold and silver; John Wiley & Sons: New York, 1999; pp 551-649. (18) Douglass, E. F., Jr.; Driscoll, P. F.; Liu, D.; Burnham, N. A.; Lambert, C. R.; McGimpsey, W. G. Anal. Chem. 2008, 80, 7670–7677.

DOI: 10.1021/la103333c

18885

Article 0 and 1.47, respectively.17,18 Five measurements were taken per slide for five different samples, and the results were averaged. Electrochemical Impedance Spectroscopy. Electrochemical measurements were taken with a Gamry Instruments Reference 600 Potentiostat/Galvanostat/ZRA (Warminster, PA). Modified gold slides were used as the working electrodes, an Accumet saturated calomel electrode (SCE) from Fisher Scientific (Pittsburgh, PA) as the reference electrode, and an Accumet platinum wire as the counter electrode. The gold slide was contacted with an alligator clip and an area of 1 cm2 was immersed in the electrolyte solution. Nonfaradaic impedance measurements were carried out at a fixed potential of 0 V vs SCE and a 10 mV AC perturbation from 100 kHz to 1 Hz. The impedance spectra were recorded in the form of Nyquist plots, and they were fitted to an equivalent circuit based on the Randles model,19-21 using nonlinear least-squares fitting to extrapolate capacitance values. The electrochemical cell was immersed in a temperature-controlled water bath. Impedance measurements were conducted every 5-6 min. The capacitance was monitored sequentially under different temperatures or solution composition for each type of SAM. For each change in the system, the measurements were taken until a stable capacitance value was obtained for at least 15 min (or at least 3 consecutive measurements), meaning a capacitance change of less than 0.005 μF/cm2 between two consecutive measurements. The impedance measurements were first carried out at room temperature in TBS, pH 7.0, and then at 37 C. This was followed by replacing 10% of the volume of TBS serving as the electrochemical cell solution with TBS, pH 7.0, containing 5% (v/v) ethanol, resulting in a final ethanol concentration of 0.5%, in order to see if the ethanol addition would affect the capacitance of the monolayer, since the vesicle solution contains ethanol. The next change was the addition of vesicles by replacing 10% volume of the electrochemical cell solution with DMPC vesicle suspension prepared in TBS, pH 7.0, for a final phospholipid concentration of 0.2 mM. Following the addition of vesicles, the impedance was measured over the course of 120 min in order to ensure the formation of the lipid monolayer or bilayer membrane.22 Finally, the system was cooled down and the impedance was measured at room temperature. The measurements for each SAM were done in triplicate, using three different gold slides.

Formation of Lipid Membranes and AChE Immobilization. Lipid membranes were formed by incubating 1 cm2 alkanethiol modified gold slides in vesicle solution prepared in TBS, pH 7.5. The gold slides were placed individually in 12-well polystyrene culture plates and 2.0 mL of vesicle solution was added on each well to ensure that the slides were completely covered with solution. The surfaces were incubated overnight at 37 C to allow formation of the lipid monolayer and bilayer membranes. At the end of the incubation period, the slides were washed, while not allowing the surface to dry or be exposed to air, which is known to destroy the lipid monolayer.4 This was done by adding 1.0 mL of TBS, pH 7.5, in each well, mixing and then removing 1.0 mL. These steps were repeated twelve times, finally leaving 2.0 mL of buffer. Following the rinsing process, AChE was added to each well and the 12-well plate was agitated on an orbital shaker for 4 h at room temperature to allow the immobilization of the protein in the membrane. The surfaces were rinsed again 12 times with TBS, pH 7.5, by exchanging 1.0 mL volumes as described above. The 1.0 mL solution removed for the first, sixth, and 12th rinse were stored to test the effectiveness of successive rinses, by checking for the presence of the enzyme in the rinse samples using the AChE (19) Randles, J. E. B. Discuss. Faraday Soc. 1947, 1, 11–19. (20) Plant, A. L.; Gueguetchkeri, M.; Yap, W. Biophys. J. 1994, 67, 1126–1133. (21) Park, S.-M.; Yoo, J.-S. Anal. Chem. 2003, 75, 455A–461A. (22) Plant, A. L.; Brigham-Burke, M.; Petrella, E. C.; Oshannessy, D. J. Anal. Biochem. 1995, 226, 342–348.

18886 DOI: 10.1021/la103333c

Milkani et al. activity assay. The surfaces were kept submerged in the 2.0 mL TBS volume at 4 C until further use. Assessment of Protein Activity. AChE activity in solution was calibrated using a standard photometric technique.23 First, stock solutions of acetylthiocholine iodide (100 mM) and 5,50 dithiobis-(2-nitrobenzoic acid) (DTNB) (500 μM) were prepared in TBS, pH 7.5. The DTNB solution was used to make a 0.1 pmol/ μL AChE solution. Different amounts of AChE enzyme (0 to 100 pmol) and 30 μL of acetylthiocholine iodide solution were added to the DTNB solution for a final volume of 3.0 mL in a 3 mL cuvette. Using an Evolution 300 UV-visible double beam spectrophotometer from Thermo Scientific (Waltham, MA) facilitated with VISIONpro software, the absorbance measurements were recorded at a 410 nm every minute for up to 5 min at room temperature. In order to determine the Michaelis constant (Km) and maximum velocity (Vmax) of AChE in solution, 30 μL of varying acetylthiocholine concentrations (resulting in 0.010, 0.025, 0.050, 0.10, 0.25, 0.50, and 1.0 mM final concentrations) were mixed with a constant amount of enzyme (10 pmol) in 500 μM DTNB to a final volume of 3.0 mL. For the determination of Km and Vmax of AChE incorporated in DMPC vesicles, a 0.1 pmol/μL solution of AChE was prepared in a vesicle suspension prepared in 500 μM solution of DTNB in TBS, pH 7.5. The mixture was agitated for 4 h on an orbital shaker at room temperature to allow for the incorporation of AChE in vesicles, and it was then used for the kinetic assay. Km and Vmax of the enzyme were obtained from the Lineweaver-Burke plot. AChE activity on the surface was determined with a modified method based on the solution assay. The remaining 2.0 mL of TBS in each well of the 12-well plate was replaced with 3.0 mL of 500 μM DTNB solution. This was done by removing 0.50 mL of TBS and adding 1.5 mL of 1.0 mM DTNB solution in TBS, pH 7.5. Then, 30 μL of acetylthiocholine iodide solution was added into each well (after removing 30 μL of DTNB solution). The 12well plate was immediately agitated at 70 rpm on an orbital shaker at room temperature. After 4 min incubation, 1.0 mL was removed from each well and the absorbance was measured at 410 nm in a 1 mL cuvette. For the determination of Km and Vmax of AChE on the surface, different acetylthiocholine concentrations were added to the wells. The absorbance of the solution in each well was assayed for 20 at 2 min intervals while agitating the plate on the orbital shaker at room temperature. This was done by taking a 1 mL sample from the well at each time point, measuring the absorbance at 410 nm, and then immediately replacing the sample into the well.

Results and Discussion Characterization of Alkanethiol SAMs. Two of the hydrophobic alkanethiols were selected, since the phase transition temperature of the acyl group equivalents of dodecanethiol (DDT) and octadecanethiol (ODT) alkyl chains are, respectively, well below or above the temperature range (21-37 C) where the formation of lipid monolayer was monitored.24 The acyl chains of the DMPC lipid have a phase transition temperature of 23 C and have the same number of carbons as the alkyl chain of tetradecanethiol (TDT). The hydrophilic alkanethiols used were 11-mercaptoundecanoic acid (MUA) and 11-amino-1-undecanethiol (AUT). Self-assembled monolayers (SAMs) of alkanethiol on gold were characterized by contact angle goniometry, ellipsometry, and impedance spectroscopy. Contact angle measurements (Table 1) confirmed the hydrophobic and hydrophilic nature of the methyl-terminated and carboxyl- or amino-terminated SAMs, respectively, and (23) Ellman, G. L.; Courtney, K. D.; Andres, V., Jr.; Featherstone, R. M. Biochem. Pharmacol. 1961, 7, 88–95. (24) Silvius, D. R. Thermotropic Phase Transitions of Pure Lipids in Model Membranes and Their Modifications by Membrane Proteins; John Wiley & Sons, Inc.: New York, 1982.

Langmuir 2010, 26(24), 18884–18892

Milkani et al. Table 1. Characterization of Dodecanethiol (DDT), Tetradecanethiol (TDT), Octadecanethiol (ODT), 11-Mercaptoundecanoic acid (MUA), and 11-Amino-1-undecanethiol (AUT) SAMs by Goniometry and Ellipsometry measured thickness SAM (carbon contact theoretical (nm) ( S.D. chain length) angle ( S.D.a thicknessb (nm) DDT (C12) 102.2 ( 1.9 1.6 1.4 ( 0.2 TDT (C14) 110.9 ( 0.8 1.8 1.8 ( 0.1 ODT (C18) 111.0 ( 1.1 2.2 2.4 ( 0.2 MUA (C11) 8.1 ( 1.5 1.6 1.0 ( 0.2 AUT (C11) 42.0 ( 1.8 1.6 0.9 ( 0.2 a Standard deviation, determined after making five measurements per slide for three different samples. b Based on the molecular length of the alkanethiol and the 30 tilt of the SAM.25

they are consistent with previously published results.25 Ellipsometric measurements (Table 1) on the SAM determined a film thickness values that were consistent with published reports and also comparable with the theoretical thickness based on the molecular length of the alkanethiol.25-29 Electrochemical impedance spectroscopy was used to further confirm the presence of a well-ordered SAM. Impedance measurements and subsequent equivalent circuit fitting determined the monolayer capacitance at room temperature to be 0.71 ( 0.13, 0.67 ( 0.06, and 0.52 ( 0.06 μF/cm2 for DDT, TDT, and ODT, respectively, as shown in Table 2. The decrease in capacitance between the DDT and ODT SAMs appears to be consistent with the increase in the SAM thickness. At 37 C, the capacitance values dropped slightly to 0.70 ( 0.11, 0.66 ( 0.04, and 0.49 ( 0.09 μF/cm2 for DDT, TDT, and ODT, respectively. These results indicate that the packing of the alkyl SAM is not affected by temperature, including TDT whose alkyl chains are the same length as the fatty acid tails of the DMPC phospholipid, which have a phase transition temperature of 23 C. The elevated temperature might be expected to increase capacitance by reorganizing the alkanethiol molecules and promoting diffusion of electrolyte ions into the film. The hydrophobic nature of these alkyl chains varying from 12 to 18 carbons, indicated by the water contact angles, ensures a stable SAM packing in the presence of water, which apparently is not affected by increasing the temperature to 37 C. This observation is important because the formation of a lipid monolayer depends on the interaction between the hydrophobic tails of the phospholipid and the hydrophobic alkyl chains of the SAM, and furthermore, the alkanethiol monolayer packing could affect the formation of the lipid monolayer membrane. Lipid Vesicles. Liposome formation and size was confirmed by transmission electron microscopy (TEM), as shown in Figure 1. Vesicles were measured against the scale, and the average diameter of 30 vesicles was determined to be 37 nm, ranging from 27 to 100 nm. Formation of Lipid Membrane on SAMs. Lipid monolayer membranes formed on hydrophobic SAMs may be ruptured by surface tension of the liquid if removed from solution.20 Thus, characterization of the membrane was limited to experiments that could be performed while the substrate remained in solution. Deposition of the lipid monolayer was monitored in situ by electrochemical impedance spectroscopy as described in the (25) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321–335. (26) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559–3568. (27) Peterlinz, K. A.; Georgiadis, R. Langmuir 1996, 12, 4731–4740. (28) Mark, S. S.; Sandhyarani, N.; Zhu, C.; Campagnolo, C.; Batt, C. A. Langmuir 2004, 20, 6808–6817. (29) Han, X.; Achalkumar, A. S.; Bushby, R. J.; Evans, S. D. Chem.;Eur. J. 2009, 15, 6363–6370.

Langmuir 2010, 26(24), 18884–18892

Article Table 2. Measured Capacitance Values of the Alkanethiol SAM (CSAM), Alkanethiol SAM with Lipid Monolayer (CSAMþLL) at Room Temperature, and the Calculated Capacitance (CLL) and Thickness of the Lipid Monolayera SAM

CSAM ( μF/cm2)

CSAMþLL ( μF/cm2)

CLL ( μF/cm2)

lipid layer thickness (nm)

DDT 0.71 0.52 1.94 1.2 TDT 0.67 0.52 2.40 1.0 ODT 0.52 0.40 1.75 1.4 MUA 3.07 2.13 ; ; a The measured values represent the average of three measurements on three different slides.

Figure 1. TEM image of the DMPC lipid vesicles. The vesicles were prepared by injecting an ethanol solution of DMPC (12.5 mg in 0.5 mL) in 10 mL of TBS at 37 C while stirring the mixture.

Experimental Details. Since the DMPC lipids have a phase transition temperature of 23 C, the alkanethiol modified gold substrates and the electrochemical cell were heated to 37 C to allow for vesicle fusion. The monolayer capacitance was monitored over time, as shown in Figure 2. The results showed a significant decrease in capacitance after the addition of vesicles consistent with the formation of a lipid monolayer onto the alkanethiol modified gold surface. This graph indicates that a complete lipid monolayer is formed on the surface after at least 2 h as indicated by a relatively constant film capacitance value after this time. The capacitance of the film decreases as the lipid membrane is formed, as the capacitance of a parallel plate capacitor is inversely proportional to the distance between the plates, given by the equation C ¼

εεο Α d

ð1Þ

where C is capacitance of the film or monolayer on the electrode, ε is the relative permittivity of the capacitor (or the dielectric constant of the film on the electrode surface), εo is the permittivity of vacuum, A is the surface area of the capacitor (area of the gold surface dipped in solution), and d is the distance between the plates (or film thickness, separating the gold surface from the electrolyte solution). The decrease in capacitance indicates an increase in the film thickness, d, on the working electrode as also reported by others.20,30-32 Decreasing the temperature from (30) Diao, P.; Jiang, D.; Cui, X.; Gu, D.; Tong, R.; Zhong, B. Bioelectrochem. Bioenerg. 1999, 48, 469–475. (31) Lingler, S.; Rubinstein, I.; Knoll, W.; Offenhaeusser, A. Langmuir 1997, 13, 7085–7091. (32) Ding, L.; Li, J.; Dong, S.; Wang, E. J. Electroanal. Chem. 1996, 416, 105– 112.

DOI: 10.1021/la103333c

18887

Article

Milkani et al.

be considered to be capacitors in series. The capacitance of the lipid monolayer can be determined as follows: 1 1 1 ¼ CLL CSAM þ LL CSAM

Figure 2. Capacitance of an octadecanethiol (ODT) and 11-mercaptoundecanoic acid (MUA) SAM on gold monitored sequentially over time: at room temperature (RT), at 37 C, in 0.5% (v/v) ethanol, in DMPC vesicle solution at 37 C, and in DMPC vesicle solution at room temperature (RT). Impedance measurements were conducted in TBS (150 mM NaCl and 20 mM Tris), pH 7.0. The spectra were obtained in the frequency range from 100 kHz to 1 Hz at a fixed DC potential of 0 V vs SCE, and the modeling was done with Randles circuit.

37 C to the initial room temperature (∼22 C) following the formation of the lipid monolayer did not affect the capacitance values. Since the area of the electrode is fixed at 1 cm2, the only parameters that can affect the capacitance upon changes in temperature are the dielectric constant and thickness of the film on the gold surface. Any changes in dielectric constant would be attributed to changes in the packing of the alkanethiol SAM and/ or lipid monolayer, whereas changes in lipid monolayer membrane thickness would occur if the depth of interdigitation between the alkyl chains of the two monolayers was affected by temperature. A reported study on lipid monolayers on hydrophobic SAMs using surface-enhanced Raman spectroscopy and reflection-absorption infrared spectroscopy has indicated that changes in the alkanethiol layer are minute upon the formation of the lipid monolayer.7 According to these reports, any changes in capacitance upon cooling of the system would indicate a change in the packing of the lipid or alkanethiol monolayer. Although the phase transition temperature of DMPC is 23 C, the packing of either monolayer was not affected by the change in temperature. A decrease in capacitance was observed after incubation in vesicle solution for all three alkanethiol SAMs, DDT, TDT, and ODT, as shown in Table 2. The change in capacitance upon the formation of the lipid monolayer is similar, while there is a steady decrease in capacitance compared to the alkanethiol chain length. Since the alkanethiol monolayer does not change significantly with the addition of the lipid monolayer, the two monolayers can 18888 DOI: 10.1021/la103333c

ð2Þ

where CLL, CSAMþLL, and CSAM are the capacitances of the lipid layer, alkanethiol SAM with lipid layer, and alkanethiol SAM, respectively. Using the capacitance values at room temperature, the capacitance and the thickness of the lipid layer were calculated for each alkanethiol SAM (Table 2). The dielectric constant of the lipid was assumed to be 2.7 as reported previously.4 The lipid monolayer capacitance values are in accordance with other published results, which vary from 1.48 to 2.9 μF/cm2.4,20,32-34 The thickness values are also consistent with the reported thickness of acyl chains in DMPC bilayers35 and lipid monolayers formed with DMPC lipids.36 These values represent only the thickness of the hydrophobic acyl group of the phospholipids, as the polar headgroup is fully hydrated during the electrochemical impedance measurements. We note that the largest relative standard deviation observed was 16%. Fusion of DMPC vesicles was also done in the presence of a COOH-terminated SAM, and changes in capacitance were monitored by electrochemical impedance spectroscopy, as shown in Figure 2. The MUA SAM is very hydrophilic, indicated by the water contact angle (Table 1). Thus, vesicle fusion is not expected to happen through one leaflet as with the hydrophobic alkanethiol SAMs, but instead, vesicle fusion is expected to result in the formation of a lipid bilayer. The capacitance of the COOHterminated SAM is larger than those obtained for the hydrophobic SAMs and the value is similar to previous reports.37,38 This result indicates that the carbon chains of MUA are not as well packed as those of the methyl-terminated alkanethiols due to the presence of the COOH head groups. The capacitance of the MUA SAM decreases after the addition of vesicles, similar to the hydrophobic alkanethiol SAMs, as reported previously.39 The average drop in capacitance after the addition of vesicles was 31% of the MUA SAM capacitance. This drop is larger than the drop in capacitance observed for the hydrophobic alkanethiol SAMs upon formation of the lipid monolayer membrane. The drop in capacitance upon formation of the lipid layer cannot be used to determine the capacitance and thickness of the lipid bilayer, since this represents a more complex system and the dielectric constant of the MUA SAM is most likely higher than the dielectric constant of the hydrophobic SAMs, as indicated by the larger capacitance of the MUA SAM. Immobilization of Acetylcholinesterase into the Lipid Monolayer Membrane. Many eukaryotic lipid-anchored proteins are linked to a molecule of glycosyl-phosphatidylinositol, also known as a GPI anchor, which contains a 1,2-diacyl-glycerol linked to a glycan as the membrane anchor. The protein used for these experiments was an acetylcholinesterase (AChE) which contains a GPI anchor. The AChE presence can be determined with a standard photometric assay by measuring the absorbance at 410 nm of a yellow product resulting from the hydrolysis of (33) Terrettaz, S.; Stora, T.; Duschl, C.; Vogel, H. Langmuir 1993, 9, 1361–1369. (34) Rao, N. M.; Plant, A. L.; Silin, V.; Wight, S.; Hui, S. W. Biophys. J. 1997, 73, 3066–3077. (35) Lewis, B. A.; Engelman, D. M. J. Mol. Biol. 1983, 166, 211–217. (36) Florin, E. L.; Gaub, H. E. Biophys. J. 1993, 64, 375–383. (37) Wang, J.; Profitt, J. A.; Pugia, M. J.; Suni, I. I. Anal. Chem. 2006, 78, 1769– 1773. (38) Sanders, W.; Vargas, R.; Anderson, M. R. Langmuir 2008, 24, 6133–6139. (39) Stelzle, M.; Weissmueller, G.; Sackmann, E. J. Phys. Chem. 1993, 97, 2974– 2981.

Langmuir 2010, 26(24), 18884–18892

Milkani et al.

Figure 3. AChE activity assay in solution. (A) Hydrolization of the acetylthiocholine substrate in the presence of AChE. (B) Reduction of 5,50 -dithiobis-(2-nitrobenzoic acid) (DTNB) in the presence of thiocholine results in the formation of a yellow-colored product (2-nitro-5-thiobenzoic acid), whose concentration is equivalent to the concentration of the AChE reaction product, thiocholine. (C) Calibration of AChE activity in solution with or without 0.1% (v/v) Triton X-100 detergent. Absorbance was measured at 410 nm after 4 min incubation for different amounts of amphiphilic AChE in the enzyme assay solution (1 mM acetylthiocholine iodide and 500 μM DTNB in TBS, pH 7.5).

acetylthiocholine by AChE (Figure 3A,B).23 Calibration curves of enzyme activity were obtained by varying the concentration of AChE and measuring the absorbance of the yellow product. The activity was measured either in the presence or lack of detergent in the activity assay buffer (Figure 3C). The activity of AChE in detergent is twice as high compared to the enzyme activity in detergent-free buffer. This result indicates that the enzyme is more active in the presence of detergent, due to the presence of the amphiphilic GPI anchor, since the lipid tail is inserted into the detergent micelles, stabilizing the enzyme. This suggests that when the amphiphilic AChE is introduced in a detergent-free buffer it is expected to spontaneously anchor itself onto the lipid monolayer membrane by inserting its lipid tail into the lipid layer once it comes into contact with it. Immobilization of the AChE enzyme on the surface would allow for determination of enzyme activity on the surface, thus indirectly confirming the presence of a planar membrane on the surface. ODT SAMs prepared on 1 cm2 gold substrates were used for the formation of lipid monolayer membranes (Figure 4A). The SAMs were incubated in vesicle solution overnight at 37 C to ensure complete formation of the lipid monolayer. After thorough rinsing (see Experimental Details), the surfaces were exposed to AChE solution in detergent-free buffer, and agitated for 4 h at room temperature to allow the enzyme to interact with and anchor to the membrane. After thoroughly rinsing the surfaces Langmuir 2010, 26(24), 18884–18892

Article

again to remove all enzyme present in solution, they were subjected to the enzyme assay. The lipid monolayer surfaces exposed to AChE solution contained the enzyme, indicated by the formation of yellow product (Figure 4B,C). The control samples, consisting of ODT SAMs and lipid monolayer surfaces not exposed to AChE solution, and ODT SAMs exposed to AChE, did not turn the solution yellow even after 1 h incubation in the activity assay solution. The lipid monolayer samples exposed to AChE turned the solution yellow within 10 min. This indicates the immobilization onto the membrane of a lipid-anchored enzyme in its active form. The rinse samples obtained during the rinsing process after incubation in enzyme solution (see Experimental Details) showed that the first rinse removed the majority of the enzyme present in solution, whereas the 6th and 12th rinses did not show yellow coloration for at least 1 h. This control experiment demonstrats that the enzyme activity observed on the lipid monolayer surfaces was not due to any residual enzyme left behind in solution. Determination of AChE Coverage and Activity in the Lipid Monolayer Membrane. Lipid-anchored proteins are expected to have comparable activities on artificial planar membrane compared to their solution counterparts inserted in detergent micelles or vesicles, if they are incorporated properly on the membrane. In addition, all proteins immobilized on membranes are expected to be active, since the lipid anchor ensures their correct orientation on the substrate surface. Thus, it can be assumed that the activity of the AChE enzyme on the surface is similar to that of the enzyme in solution. In order to quantify the amount of enzyme that is immobilized on the surface, ODTcoated gold slides that were exposed to a vesicle solution to form lipid monolayer membranes were incubated in different solution concentrations of AChE in detergent-free buffer. Assuming similar activity of the enzyme on the surface and in solution and using the calibration curve obtained in the activity assay with detergent (Figure 3C), the amount of enzyme on the surface can be calculated and correlated with the concentration of AChE in the incubating solution. As the graph in Figure 5 shows, in the 1-10 pmol/mL concentration range, AChE immobilization on the membrane appears to have a linear dependence with the solution concentration of AChE. This suggests that the amount of lipid-anchored protein on the surface can be controlled by setting the solution concentration in the linear range of the surface density dependence on solution concentration. At 50 pmol/mL solution concentration, the surface coverage appears to be at or near the saturation point or maximum coverage. The enzyme activity was also compared in different media, and the results showed that the Michaelis constant (Km) and maximum velocity (Vmax) were similar for the enzyme in detergent and DMPC vesicles (Table 3). More importantly, the Km for the lipid membrane-bound enzyme was similar to the Km values in detergent and vesicles. The Vmax value for the surface-immobilized enzyme was also similar to the values obtained in detergent solution and vesicles (in proportion with the enzyme amount immobilized on the surface). These results indicate that immobilized AChE has the same activity as enzyme in detergent solution, and more importantly the same activity as enzyme in lipid vesicles. This is in agreement with the previous assumption that enzyme activity on the surface and in solution are similar. Thus, the values obtained for the amount of enzyme immobilized on the surface are reliable. Immobilization of Acetylcholinesterase into the Lipid Bilayer Membranes. Lipid bilayer membranes were formed on hydrophilic SAMs with the same procedure used to generate DOI: 10.1021/la103333c

18889

Article

Milkani et al.

Figure 4. Immobilization of the AChE enzyme on the lipid monolayer membrane. (A) Diagram of the steps used to immobilize AChE on the lipid layer formed on an octadecanethiol (ODT) SAM. (B) Testing of AChE activity on the surface. AChE was immobilized by adding 2 μL of AChE from the 0.22 mg/mL stock solution into each the well (2.0 mL) and agitating the plate for 4 h at room temperature. The lipid monolayer surfaces exposed to enzyme (ODTþVþAChE) turned the solution yellow within 10 min. Lipid monolayer samples (ODTþV) and samples (ODT) which were not exposed to AChE did not turn the solution yellow. Most importantly, no yellow color was observed on ODT surfaces exposed to enzyme (ODT þAChE). (C) Digital image of the assay taken after 10 min incubation. Table 3. Km and Vmax Values Obtained for AChE Kinetic Assays with Acetylthiocholine Iodide as Substrate in Three Different Media: Detergent Solution, DMPC Vesicles, and Lipid Monolayer Membranea enzyme medium detergent solution DMPC vesicles lipid monolayer membrane

Figure 5. AChE activity assay on the surface. ODT SAMs prepared on gold substrates (1 cm2) were incubated overnight in a 12-well plate submerged in a DMPC vesicle solution to allow the formation of the lipid monolayer, and after thorough rinsing, they were incubated in 1, 2, 5, 10, 20, and 50 pmol/mL AChE solution. The slides were thoroughly rinsed again and subsequently incubated in AChE activity solution with agitation on an orbital shaker at room temperature. Following 4 min incubation, 1.0 mL volume samples were taken from each well and the absorbance was measured at 410 nm. The absorbance values given at each concentration represent the average obtained from two samples. The amount of enzyme on the surface was calculated based on the calibration curve shown in Figure 3.

lipid monolayer membranes (Figure 6A). Two different hydrophilic SAMs were used; a COOH-terminated SAM (MUA) and an NH2-terminated SAM (AUT). In aqueous solutions of neutral pH, the MUA and AUT SAMs have a negatively or positively charged surface, respectively. As can be seen from the chemical structure of DMPC, the polar head of the lipid is a zwitterion, therefore DMPC vesicles were expected to fuse with both surfaces. Some of the MUA and AUT surfaces were modified by surface chemistry so they contained alkyl chains protruding out of the charged surface (Figure 6A). Since some membrane-associated proteins are anchored on membranes with one fatty acyl chain linked to an amino acid residue by an amide bond, these alkyl chains are predicted to anchor the bilayer membrane once it forms 18890 DOI: 10.1021/la103333c

Km Vmax enzyme amount (10-4 M) (10-5 M/min) in assay (pmol) 2.1 ( 0.3 2.2 ( 0.1 1.8 ( 0.3

2.0 ( 0.3 2.1 ( 0.1 0.47 ( 0.08

10 10 2.8

a AChE was immobilized on the surface by exposing the lipid monolayer membrane samples formed on ODT SAMs to a 10 pmol/mL solution of AChE. The amount of AChE on the 1 cm2 gold substrates was estimated from the graph in Figure 5. The kinetic assays were performed as described in the Experimental Details. Km and Vmax values were obtained from Lineweaver-Burke plots. The error was derived from regression analysis of the Lineweaver-Burke plots.

on the hydrophilic surface. The surface modification of the MUA and AUT SAMs was performed by reacting the terminal group of the SAM with dodecylamine and dodecanoic acid, respectively, resulting in the respective alkyl-terminated surface, MUA þ Dod and AUT þ Dod, via the formation of amide bonds. These surface reactions are known to result in approximately 50% surface coverage at 1 M concentration of reactants;40 therefore, at 1 mM concentration fewer alkyl chains are expected to bind to the surface. When tested in the AChE activity assay, all surfaces exposed to lipid vesicles generated yellow product within 10 min, whereas the negative controls (surfaces that were not exposed to vesicles) not did not generate yellow product during the same time frame (Figure 6B,C). Both hydrophilic surfaces tested showed the presence of active AChE on the surface, regardless of the charge on the terminal group. The alkyl-modified MUA and AUT SAMs also showed the presence of active enzyme. In future experiments, when truncated proteins containing hydrophobic domains will be inserted into lipid bilayer membranes, the effect of the alkyl groups from the underlying SAM will be assayed with respect to the bilayer membrane stability in either wet or dry conditions. (40) Yan, L.; Marzolin, C.; Terfort, A.; Whitesides, G. M. Langmuir 1997, 13, 6704–6712.

Langmuir 2010, 26(24), 18884–18892

Milkani et al.

Article

Figure 6. Immobilization of AChE on the lipid bilayer membrane. (A) Diagram of the steps used to immobilize AChE on the lipid layer formed on hydrophilic SAMs. (B) Testing AChE activity on the surface. AChE activity was observed for all SAMs exposed to vesicle solution (SAMþVþAChE), but not observed for SAMs that were not exposed to vesicles (SAMþAChE). Enzyme was immobilized by adding 2 μL of AChE from the 0.22 mg/mL stock solution into each well (2.0 mL) and agitating the plate for 4 h at room temperature. (C) Digital image of the assay taken after 10 min incubation. The surfaces are as follows: 11-mercapto undecanoic (MUA), MUA reacted with 1 mM dodecyl amine (MUA þ Dod), 11-amino-1-undecanethiol (AUT), and AUT reacted with 1 mM dodecanoic acid (AUT þ Dod).

These experiments also demonstrate that using amphiphilic AChE and its activity assay provides a practical and rapid test for verification of lipid membrane formation. This method could be used to confirm the presence of the membrane on other substrates such as glass and polymers that are difficult to characterize with standard surface characterization techniques.

Conclusions The formation of lipid monolayer and bilayer membranes on alkanethiol-coated gold surfaces was monitored electrochemically by impedance spectroscopy and the observed decrease in capacitance indicated that lipid membranes were formed. Testing SAMs of hydrophobic alkanethiols with different alkyl chain lengths showed that SAM thickness did not affect the formation of the lipid monolayer membrane or the ability to detect its formation. The electrochemistry results suggest that the packing of the SAM is denser than the packing of the lipid monolayer or bilayer and neither was affected by changes in temperature in the range between room temperature (22 C) and 37 C. A lipid-anchored enzyme, AChE, was successfully incorporated into both types of planar membranes, and it was shown to exhibit the similar biochemical activity as enzyme reconstituted in detergent or lipid vesicles. The lipid monolayer can be used for the surface attachment of other lipid-anchored proteins or water-soluble proteins of interest that have been lipidated. The lipid bilayer is expected to be capable of hosting integral membrane proteins or truncated Langmuir 2010, 26(24), 18884–18892

membrane proteins that still contain the hydrophobic domain. The results described in this work indicate that (1) these artificial planar membranes can be employed for the surface attachment of lipidanchored proteins; (2) the amount of immobilized protein can be preset by controlling the incubating solution concentration and other parameters such as incubation time and temperature; (3) AChE and its activity assay can be used for rapid testing and conformation of the presence of artificial lipid membranes on other substrates; and (4) the system can be easily scaled up for highthroughput screening applications by using a low-cost and commercially available substrate. Many therapeutic drugs target membrane proteins that require a lipid membrane environment to function properly. Solutionbased assays are more reliable than cell-based assays for running large-scale high-throughput screening assays, and they offer the use of homogeneous and chemically defined medium. However, hydrophilic domains of membrane proteins expressed without their respective transmembrane (hydrophobic) domains are expected to have different biochemical activities and pharmacological effects from the full-length proteins inserted in membranes. Insertion of extra or intracellular domains via hydrophobic domains in planar membranes would be a closer model to the fulllength protein observed in normal physiological settings, while allowing for a homogeneous and controlled composition of the membrane medium. Artificial planar membranes provide a practical platform for the incorporation of truncated membrane proteins that combines the advantages of the solution-based assay DOI: 10.1021/la103333c

18891

Article

and the cell membrane environment. More importantly, solutionbased assays can be paired with surface-based detection technologies to determine protein interactions and activity providing a powerful dual-approach for drug discovery applications. In addition, by controlling the surface chemistry, artificial membranes can potentially be fabricated on any type of substrate. The planar membranes described in this work could be fabricated in multiwell arrays and combined with the potential multiple modes of detection that are compatible with this technology, leading to the faster discovery of novel pharmaceuticals.

18892 DOI: 10.1021/la103333c

Milkani et al.

Acknowledgment. We are grateful to Dr. Boquan Li of WPI’s Materials Science and Engineering Program for the use of the JEOL 100CX Transmission Electron Microscope. The authors gratefully acknowledge partial financial support from the Telemedicine and Advanced Technology Research Center (TATRC) at the U.S. Army Medical Research and Materiel Command (USAMRMC) under Award No. W81XWH-07-2-0106. The views, opinions, and/or findings contained in this report are those of the author(s) and should not be construed as an official Department of the Army position, policy, or decision unless so designated by other documentation.

Langmuir 2010, 26(24), 18884–18892