Langmuir 2000, 16, 1801-1805
1801
Formation of Lipid Bilayers on a New Amphiphilic Polymer Support Patrick The´ato and Rudolf Zentel* Department of Chemistry and Institute of Materials Science, University of Wuppertal, Gaussstrasse 20, D.42097 Wuppertal, Germany Received March 12, 1999. In Final Form: October 6, 1999 In this paper we present a way to build up tether supported lipid bilayers. First we synthesized novel amphiphilic polymers with an adjustable hydrophilicity. This was achieved by introducing ionic groups together with disulfide and lipid anchor groups in polyacrylamides or N,N-dimethylacrylamides. The self-assembly of these polymers on a gold surface was monitored by surface plasmon spectroscopy and was characterized by contact angle measurements and swelling experiments. The contact angle decreased linearly with the percent of ionic groups in the polymer, resulting in a more hydrophilic surface. Swelling experiments demonstrated the greater hydrophilicity of the charged polymers. To build up tether supported lipid bilayers, we fused phospholipid vesicles onto the prepared thin polymer films. The charged polymers initiated a vesicle fusion on the thin polymer films and resulted in a supported lipid bilayer.
Introduction Surface modification through polymer adsorption has become important. Many scientists focus their attention on thin organic films, since these provide ways to control surface properties such as wetting, adhesion, and biocompatibility for future technological applications.1,2 Lipid bilayers have the natural behavior to form defined assemblies of nanoarchitectures. Preparation of supported lipid bilayers on solid substrates may allow incorporation of specific enzymes to prepare biosensors.3 Supported lipid bilayers may be prepared by fusion of small unilamellar vesicles onto the surface. In this respect several successful attempts have been reported.4 Some research groups have fused phospholipid vesicles onto hydrophobic self-assembled monolayers (alkylthiols) to prepare supported lipid bilayers.18 In these systems the lower (proximal) leaflet is, however, fully attached to the (1) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press: San Diego, CA, 1991. (2) Ulman, A. Chem. Rev. 1996, 96, 1533. (3) Sackmann, E. Science 1996, 271, 43. (4) Heibel, C.; Maus, S.; Knoll, W.; Ru¨he, J. ACS Book Series 1998, 104, 695. (5) Hausch, M.; Zentel, R.; Knoll, W. Macomol. Chem. Phys. 1999, 200, 174. (6) Hausch, M.; Beyer, D.; Knoll, W.; Zentel, R. Langmuir 1998, 14, 7213. (7) Ringsdorf, H.; Beyer, D.; Elender, G.; Knoll, W.; Ku¨hner, M.; Maus, S.; Sackmann, E. Angew. Chem. 1996, 108, 1791. Angew. Chem., Int. Ed. Engl. 1996, 35, 1682. (8) Ferruti, P.; Bettelli, A.; Fere´, A. Polymer 1972, 13, 462. (9) Hausch, M. Ph.D. Thesis, University Mainz, Germany, 1999. (10) Kretschmann, E. Z. Phys. 1971, 241, 313. (11) Kretschmann, E. Opt. Commun. 1983, 6, 185. (12) Raether, H. Surface Plasmons; Springer Tracts in Modern Physics: Springer: Berlin, 1998. (13) Stro¨m, G.; Fredriksson, M.; Stenius, P. J. Colloid Interface Sci. 1987, 119, 352. (14) Knoll, W. MRS Bull. 1991, 16, 29. (15) Bunjes, N. ; Schmidt, E. K.; Jonczyk, A.; Rippmann, F.; Beyer, D.; Ringsdorf, H.; Gra¨ber, P.; Knoll, W.; Naumann, R. Langmuir 1997, 13, 6188. (16) Williams, L. M.; Evans, S. D.; Flynn, T. M.; Marsh, A.; Knowles, P. F.; Bushby, R. J.; Boden, N. Langmuir 1997, 13, 751. (17) Israelachvili, J. Intermolecular & Surface Forces, 2nd ed.; Academic Press: San Diego, CA, 1991. (18) Spinke, J.; Yang, J.; Wolf, H.; Liley, M.; Ringsdorf, H.; Knoll, W. Biophys. J. 1992, 63, 1667.
substrate via a hydrophilic spacer (either oligomer or polymer). To increase the mobility of both leaflets only a part of the proximal leaflet should be bound to the substrate. Such systems have been prepared by a sequence of adsorption of mixed vesicles, their desorption, and a second adsorption of vesicles.16 It is our aim to prepare and characterize fully functionalized polymer systems, which chemisorb to a surface to form a hydrophilic polymer film with small amounts of lipid anchor groups (see Figure 1).5,6 Our amphiphilic polyacrylamide polymers P1 - P16 fulfill three requirements.7 They consist of disulfide anchor groups that chemisorb the polymer onto gold surfaces, they provide water soluble groups for a swellable polymer cushion, which decouples the lipid bilayer from the solid support, and they contain a polymer bound natural lipid (here DMPE) to induce vesicle fusion and to anchor the lipid bilayer onto the polymer support. Our synthesis strategy provides a very flexible way to obtain these multifunctional polymers through a polymer analogous reaction. We use a reactive ester polymer 1,8 which can react with a huge variety of amines, as shown in Figure 1. Alternatively, unreacted ester can be used to link the polymer to an aminofunctionalized surface.9 Our main goal is to build up a lipid bilayer through vesicle fusion onto the self-assembled polymer support. Vesicles prepared by extrusion (diameter approximately 80 nm) do, however, not adsorb on most of the hydrophilic polymer supports we have synthesized. Only some of the supports proved to be useful9 and it seemed that ionic forces played an important role. Therefore the idea of using charged polymers arose. The charges should increase the attraction of the polymer cushion to the lipid vesicles, forcing them to fuse onto the modified surface. Experimental Section Polymer Synthesis. We have published a detailed description of the synthesis elsewhere.6 In brief, the reactive ester polymer 1 reacts in a polymer analogous reaction in four steps with different amines in a DMF/pyridine mixture, first, with cysteaminemethyl disulfide 2, which is responsible for the chemisorption of the polymer onto a gold surface, second, with DMPE 3 (L-R-phosphatidylethanolamine, dimyristol), which acts as the lipid anchor, third, with (2-aminoethyl)-trimethylammonium-
10.1021/la990292l CCC: $19.00 © 2000 American Chemical Society Published on Web 12/29/1999
1802
Langmuir, Vol. 16, No. 4, 2000
The´ ato and Zentel
Figure 1. Polymer analogous reaction of the hydrophilic polymer spacer with a lipid anchor and a disulfide anchor. Table 1. Compositions of the Polymers P1-P16a
polymer
-NH2 (mol %)
anchor groups -NH(CH3)2 (mol %)
amine 4 or 5b (mol %)
P1, P9 P2, P10 P3, P11 P4, P12
88 83 78 73
0 0 0 0
5 10 15 20
P5, P13 P6, P14 P7, P15 P8, P16
0 0 0 0
88 83 78 73
5 10 15 20
a All polymers contained 5 mol % DMPE 3 and 2 mol % disulfide 2. b Amine 4 in the series P1-P8, amine 5 in the series P9-P16.
chloride 4, and fourth, with an excess of either ammonia or dimethylamine. The reaction scheme is summarized in Figure 1. To compare the influence of the positive charges, polymers with 1-amino-2(dimethylamino)-ethane 5 instead of 4 are prepared. Table 1 shows the composition of the prepared polymers with different contents of 4 and 5, respectively. TLC control showed the quantitative reaction of the different amines 2-5 with the reactive ester polymer 1. The polymer composition was also verified by 1H NMR, while a determination of the molecular weight by GPC was not possible. Gold Slides. Gold surfaces were prepared by evaporation. The glass slides (BK-270, Berliner Glas, Germany) were cleaned by ultrasonication in methanol, in detergent (Hellmanex, Hellma, Mu¨hlheim, Germany) and five times in Milli-Q water. Chromium (for better gold-adhesion) and gold (purity > 99.99%) films (thickness 2 nm and 48 nm, respectively) were deposited through evaporation on the slides in a vacuum chamber at 5 × 10-6 mbar with a Blazer vapor deposition apparatus (Bae250). The slides were stored under argon. Surface Plasmon Spectroscopy. In a standard Kretschmann configuration9 surface plasmon spectroscopy (SPS) was used to characterize the self-assembled polymer films on gold. The gold coated glass slides were refractive index matched with a 90° glass prism, used as the surface plasmon coupler. During the experiment the resonant excitation of the surface mode, which is very sensitive to the actual interfacial architecture and,
therefore, to the thickness of the adsorbed polymer, was monitored by recording the total internally reflected light from a HeNeLaser (λ ) 633 nm) as a function of the angle of incidence.11,12 All measurements were performed at 30 °C. The thickness of the thin polymer was calculated from the SPS curves based on a Fresnel fitting. The refractive index of the polymer was assumed to be n ) 1.5. Polymer Adsorption. The polymers were adsorbed on gold surfaces. Fresh gold slides were immersed for 60 min into a solution of 1 mg polymer in 1 mL pure water (Milli-Q) or 1 mL ethanol (p. a.), respectively. The slides were then washed thoroughly with pure water or ethanol and dried in an argon stream, to be used later for contact angle measurements by use of a Kru¨ss G1 system (Kru¨ss GmbH, Germany).13 Swelling Behavior. A thickness increase should arise from the swelling of the self-assembled polymers. We performed this in a homemade swelling chamber. The polymer film was exposed to the chamber, in which the relative humidity was adjusted to 0%, 43%, or 98%, using solid KOH, saturated K2CO3 solution, or saturated KNO3 solution.14 The swelling was monitored by surface plasmon spectroscopy. Vesicle Preparation. Vesicles were prepared by extrusion. DMPC was dissolved in chloroform and under an argon stream the chloroform was evaporated yielding a thin DMPC film on the glass wall. The film was vacuum-dried for 3 h and then suspended with 0.1 mM NaCl solution at 30 °C, to obtain a 1 mg/mL solution. The suspension was allowed to swell for 1 h at 30 °C. The milky vesicle suspension was then pushed 30 times through a polycarbonate filter with a pore size of 50 nm (Avestin, Ottawa, Canada). The extruded vesicles were stored at 30 °C (above the phase transition temperature of DMPC) and used within 5 days. Vesicle Fusion. The freshly adsorbed polymer surfaces were immersed for 60 min into a suspension of the prepared vesicles. The fusion process was monitored by kinetic SPS measurements. Finally an SPS curve was recorded to calculate the final film thickness increase. After flushing the cuvette with 0.1 mM NaCl solution another SPS curve was recorded to check if a stable bilayer was achieved.
Results and Discussion Self-Assembly. The self-assembly process is a useful method to prepare thin polymer films on gold surfaces.15,16 We followed it by surface plasmon spectroscopy. A typical
Lipid Bilayers on a New Amphiphilic Polymer Support
Langmuir, Vol. 16, No. 4, 2000 1803
Figure 2. P4 self-assembly from ethanol solution onto gold; recorded by surface plasmon spectroscopy. Table 2. Contact Angles θa and θr of the Polymers P1-P16a polymer θa (°) [EtOH] θa (°) [H2O] θr (°) [EtOH] θr (°) [H2O] P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 a
60 58 56 57 66 61 63 66 69 63 69 72 68
45 42 38 51 47 46 42 -
24 20 24 24 20 33 27 35 43 37 40 43 38
17 13 21 20 16 15 16 -
Figure 4. (a) Dependence between the contact angle θa and the amount of charge on the polymer. (b) Dependence between the contact angle θa and the amount of uncharged amine 5 on the polymer.
The standard deviation was about 1° for all samples.
Figure 5. Swelling of P1 starting (a) at 0% humidity, (b) from 0% to 43%, and (c) from 43% to 98% humidity, respectively. Figure 3. Schematic sketch of the conformation of amphiphilic polymers in (a) water as micelles and in (b) ethanol.
kinetic curve is shown in Figure 2. All polymers produced a film of about 25 Å thickness. As can be seen in Figure 2, the self-assembly process of the polymer was finished after 40 min. Contact Angles. The hydrophilic behavior of all polymers on the surface was studied by measuring the advancing contact angles.17 All charged polymers, P1P8, were self-assembled from water solutions. All uncharged polymers, P9-P16, were only soluble in ethanol and therefore the self-assembly was done from that solution. The measured contact angles are the mean values of five independent measurements and are collected in Table 2. As can be seen in the case of polymers P1-P5 the contact angles of polymers self-assembled from water
were approximately 16° smaller than the ones from ethanol. This indicates that the solvent has a direct influence not only on the conformation of the polymer in solution, but also affects the adsorbed polymer film. It seems that the hydrophobic parts of the polymer adsorbed from ethanol solution are closer to the interface with the solution. In contrast to that, the micelle formed in water chemisorbs directly on the surface and the polymer coil does not have the possibility of rearranging to the same conformation afterward. This is shown schematically in Figure 3. Looking in more detail at the contact angles, one can see the huge difference between the contact angles θa and θr. They should theoretically have the same value, if the position of hydrophilic and hydrophobic parts of the polymer is rigidly fixed. But the change of the contact angle between the expanding and retreating water droplet
1804
Langmuir, Vol. 16, No. 4, 2000
The´ ato and Zentel
Figure 6. Schematic sketch of the vesicle fusion on the polymer coated surface.
reflects a rearrangement of the lipid chains with regard to the surface. Apparently a larger rearrangement is possible. Nevertheless the difference between the films adsorbed from different solvents remains. The dependence of the contact angle on the content of charged groups in the polymer can be better seen in a plot of θa against percentage of charge. This is shown in Figure 4a and 4b for the amines 4 and 5, respectively. There was a linear dependence between the contact angle θa and the amount of charge on the polymer. The more charged the polymer is, the smaller the contact angle (see Figure 4a). It is interesting to note that there exists also a linear relation between the contact angle θa and the amount of amine 5 bound to the polymer (see Figure 4b). In this case the amine 5 acts, however, as a hydrophobic part increasing the contact angle. Swelling. Another possibility to examine the hydrophilicity of the self-assembled polymer films offers the swelling of the polymers presented. Hydrophilic films should be able to take up water and swell, resulting in a thicker film. A typical swelling curve is shown in Figure 5. Before measuring the swelling, the polymer film was dried for 3 h above KOH. By changing the salt solutions one after another the humidity was changed and the corresponding swelling was measured. The swelling was not linear with humidity. The polymer film swelled only by one-third when the humidity was increased from 0% to 43%, compared to the swelling resulting from a humidity change from 43% to 98%. This was observed for all the measured polymer films, charged or uncharged. The thickening of the polymer films during swelling from 0% to 98% humidity can be summarized as follows. The polymers P1-P4 all swelled approximately by 23%, P5P8: 16%, P9-P12: 8%, and P13-P16: 6%, see Table 3. However, the accuracy is not very high, since the resolution of the surface plasmon spectroscopy is (1 Å. The reduced swelling of P1-P4 as compared to P5-P8 seems to be due to the less hydrophilic N,N-dimethylacrylamide main chain spacer. But it is obvious that the charged polymers P1-P8 all swell better than the uncharged polymers P9P16. This demonstrates that the incorporation of charges increases the hydrophilicity of the polymer supports. Vesicle Fusion. One way to prepare tethered supported lipid bilayers is the fusion of vesicles on self-assembled polymer films.18,19 Our model has two major steps. The first one is the vesicle adsorption onto the surface and the second is the “coasting” or fusion of the vesicles to lipid bilayers, as shown schematically in Figure 6. Especially in the first step the attraction of the polymer surface to vesicles in the solution plays a major role. If our model is true, more hydrophilic charged polymers, such as P1P8, should have a higher attraction to vesicles than (19) Sohling, U.; Schouten, A. J. Langmuir 1996, 12, 3912.
Table 3. Swelling of the Polymers P1-P16 from 0% to 98% Humidity polymer P1 P2 swelling (%) [(4 %] 22 25
P3 21
P4 25
P5 12
P6 21
P7 11
P8 21
polymer P9 P10 P11 P12 P13 P14 P15 P16 swelling (%) [(4 %] 0 13 13 7 8 4 5 8
Figure 7. Monitored kinetics of the vesicle fusion on P2.
uncharged polymers, like P9-P16. The surface plasmon measurements show that a fusion of vesicles of diameter approximately 80 nm20 takes place only on the charged polymer surfaces prepared from P1-P8. On the contrary, no increase in thickness can be measured upon vesicle interaction with the uncharged polymer surfaces prepared from P9-P16. A typical kinetic curve of a vesicle fusion onto a charged polymer support is shown in Figure 7. It is evident that most of the increase takes place within the first minute. Thereafter only a small thickness increase is observed which reaches a plateau. After flushing the cuvette with 0.1 mM NaCl solution, no decrease in thickness could be observed. It seems that the adsorption of vesicles takes place rapidly. On the other hand the rupture of vesicles is a second process which takes place over a longer time period. Within that process a lipid bilayer is formed. It is not yet clear if the two time constants correspond to the two steps in the model of the vesicle fusion or not. Further measurements to answer this question are in progress. The thickness increase after vesicle fusion is about 30 Å (P1-P8). This is less than the theoretical value of 50 Å for a lipid bilayer, because the amount of lipid already bound to the polymer has to be added to the measured value. This amount of lipid (see Figure 6, left side) was already measured with the thickness of the lipopolymer film of about 25 Å. So, the measured thickness of 30 Å seems to be reasonable and is within the range of thickness measured by others.16,18 (20) Nollert, P.; Kiefer, H.; Ja¨hnig, F. Biophys. J. 1995, 69, 1447.
Lipid Bilayers on a New Amphiphilic Polymer Support
Conclusions and Future Aspects The idea to increase the attraction of polymer surfaces to lipid vesicles in solution by charges might be seen as another step toward the controlled assembly of biomaterials at nanosize dimensions. It could be shown that the hydrophilic behavior of polymer surfaces can be controlled by using charged groups. Contact angles depend linearly on the amount of charges bound to the polymers. Com-
Langmuir, Vol. 16, No. 4, 2000 1805
parison of the charged polymers with their uncharged equivalents showed a simple result. Only the charged polymer surfaces were able to initiate the adsorption of vesicles onto the surface. It is not yet clear if the vesicle fusion process can be described by a two step model. Further measurements have to be done before a detailed description can be established. LA990292L
