Phospholipid Bilayers as Biomembrane-like Barriers in Layer-by

Jun 22, 2007 - Estelle Sorrenti , Vincent Ball , Doriane Del Frari , Claire Arnoult , Valérie ... Ryan J. Nichols , Slavica Praporski , Isabelle L. d...
0 downloads 0 Views 451KB Size
8236

Langmuir 2007, 23, 8236-8242

Phospholipid Bilayers as Biomembrane-like Barriers in Layer-by-Layer Polyelectrolyte Films Ana-Maria Pilbat,† Zsolt Szegletes,† Zolta´n Ko´ta,† Vincent Ball,‡ Pierre Schaaf,§ Jean-Claude Voegel,‡ and Bala´zs Szalontai*,† Institute of Biophysics, Biological Research Center of the Hungarian Academy of Sciences, Szeged, TemesVa´ ri krt. 62, H-6701 P. O. Box 521, Hungary, Faculte´ de Chirurgie Dentaire, UniVersite´ Louis Pasteur, UMR 595, Institut National de la Sante´ et de la Recherche Me´ dicale, 11 Rue Humann, 67085 Strasbourg Cedex, France, and Institut Charles Shadron, Unite´ Propre 22, Centre National de la Recherche Scientifique, 6 Rue Boussingault, 67083 Strasbourg Cedex, France ReceiVed March 22, 2007. In Final Form: May 11, 2007 Dipalmitoylphosphatidylcholine (DPPC) bilayer was created on the surface of an exponentially growing poly(glutamic acid)/poly(lysine) (PGA/PLL) layer-by-layer polyelectrolyte film. The lipid bilayer decreased the surface roughness of the polyelectrolyte film. The layer-by-layer construction of the polyelectrolyte film could be continued on the top of the DPPC layer. The lipid bilayer, however, formed a barrier in the interior of the polyelectrolyte film, which blocked the diffusion (a prerequisite for exponential growth) of the polyelectrolytes. Thus, a new growth regime started in the upper part of the polyelectrolyte film, which was added to embed the DPPC bilayer. The structure and the dynamics of the DPPC bilayer on the polyelectrolyte film surface remained similar to that of its hydrated multibilayers, except that the phase transition became wider. In the case of embedded DPPC bilayers, in addition, the phase-transition temperature also decreased. This is the result of interactions with the nonconcerted movements of the barrier-separated lower and higher parts of the polyelectrolyte film. Gramicidin A (GRA) as a model of lipid-soluble peptides and proteins was successfully incorporated into such DPPC films. The DPPC films, either with or without GRA, were remarkably stable; as many heating-cooling cycles to measure phase transition could be carried out without visible alterations as wanted.

Introduction Polyelectrolyte multilayers built by layer-by-layer (LBL) adsorption offer a simple and versatile tool to have surfaces with adjustable properties.1 It has been shown that by varying the electrolytes and/or the build-up conditions, the properties of the films, such as thickness,2 cell adhesion,3 or protein adsorption4 capacity, can be altered at will. In attempts for practical applications, such films were functionalized with features ranging, e.g., from antifungal activity,5 through anti-inflammatory properties,6 to electrooptical devices.7 Depending on the components used, polyelectrolyte films may exhibit linear or exponential growth regimes.4,8 Exponential growth requires the free diffusion of at least one of the film components in the interior of the film.9 Therefore these films have less ordered structures than linearly growing films, which have been shown to be partially stratified with some interpen* To whom correspondence should be addressed. Phone: 0036 62 599 605. Fax: 0036 62 433 133. E-mail: [email protected]. † Hungarian Academy of Sciences. ‡ Institut National de la Sante ´ et de la Recherche Me´dicale. § Centre National de la Recherche Scientifique. (1) Decher, G. Science 1997, 277, 1232-1237. (2) Shiratori, S. S.; Rubner, M. F. Macromolecules 2000, 33, 4213-4219. (3) Richert, L.; Lavalle, P.; Vautier, D.; Senger, B.; Stoltz, J. F.; Schaaf, P.; Voegel, J. C.; Picart, C. Biomacromolecules 2002, 3, 1170-1178. (4) Gergely, C.; Bahi, S.; Szalontai, B.; Flores, H.; Schaaf, P.; Voegel, J. C.; Cuisinier, F. J. G. Langmuir 2004, 20, 5575-5582. (5) Etienne, O.; Voegel, J. C.; Bolcato-Bellemin, A. L.; Egles, C.; Taddei, C.; Gasnier, C.; Aunis, D.; Metz-Boutigue, M. H.; Schaaf, P. Biomaterials 2005, 26, 6704-6712. (6) Schultz, P.; Debry, C.; Vautier, D.; Richert, L.; Jessel, N.; Haikel, Y.; Voegel, J. C.; Ogier, J.; Schaaf, P. Biomaterials 2005, 26, 2621-2630. (7) Eckle, M.; Decher, G. Nano Lett. 2001, 1, 45-49. (8) Lavalle, P.; Gergely, C.; Cuisinier, F. J. G.; Decher, G.; Schaaf, P.; Voegel, J. C.; Picart, C. Macromolecules 2002, 35, 4458-4465. (9) Picart, C.; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.; Schaaf, P.; Voegel, J. C.; Lavalle, P. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12531-12535.

etration between neighboring layers.1 The films made from poly(L-glutamic acid) (PGA) and poly(L-lysine) (PLL) used in the present study are of the exponential type.8 Considering the extensively charged nature of the polyelectrolytes in these films, there is no chance for direct incorporation of nonpolar, hydrophobic compounds into them. From the point of view of practical applications, however, it could be very useful if such compounds, e.g., different proteins, peptides, and drugs, could be incorporated. For the incorporation of such protein molecules, at least lipid bilayers are needed. If once such bilayers were formed in the interior of polyelectrolyte films, they might be utilized as controllable internal barriers as well. The present paper describes the first experiments toward this goal. Recently, Delajon et al.10 have shown that it was possible to form a lipidic bilayer from dimyristoylphosphatidylcholine (DMPC) on the surface of a film made from positively charged poly(allylamine hydrochloride) (PAH) and negatively charged poly(sodium 4-styrenesulfonate) (PSS). In addition, these authors showed that the DMPC bilayer could only be incorporated between two PSS layers and that the lipidic bilayer embedding could not be achieved between two PAH layers.10 This observation was explained by the fact that the cationic choline group (the positive partner in the zwitterionic headgroup of DMPC) could directly interact with the negative groups of PSS. The applied technique, neutron reflectometry, could prove the presence of a solid, continuous lipid bilayer, but could not show the actual physical state (gel or liquid crystalline phase), or the dynamical properties (e.g., gel to liquid crystalline phase transition) of the adsorbed phospholipid bilayer. To our knowledge, there are no investigations considering the possibilities of using such lipid bilayers as internal barriers, (10) Delajon, C.; Gutberlet, T.; Steitz, R.; Mo¨hwald, H.; Krastev, R. Langmuir 2005, 21, 8509-8514.

10.1021/la700839p CCC: $37.00 © 2007 American Chemical Society Published on Web 06/22/2007

Lipid Bilayers in Polyelectrolyte Films

separating two different compartments of the exponentially growing polyelectrolyte multilayer films. With a different approach, however, by alternating linearly and exponentially growing film regimes using different compounds, certain barriers were constructed in the interior of a polyelectrolyte film.11,12 The linearly growing films made from PSS and PAH represented a barrier against diffusion of the components of the exponentially growing regimes. One of the aims of this article is to demonstrate the feasibility of separating two exponentially growing multilayer film regimes by a lipid bilayer and to isolate these polyelectrolyte compartments one from another. To follow structural changes in polyelectrolyte films, Fourier transform infrared (FTIR) spectroscopy has become an established and very sensitive method; especially its attenuated total reflection (ATR) mode is well-suited for structural studies, since there the spectra are taken from the direction of the surface (an infrared transmitting, high-refractive-index crystal), on which the polyelectrolyte film is adsorbed. This arrangement is mimicking real situations, which polyelectrolyte films may face upon applications. With FTIR-ATR, the secondary structures of exponentially growing PGA/PLL films have been studied in details.13,14 We have also shown earlier that polyelectrolyte films can modify and stabilize the structure of adsorbed proteins.15,16 FTIR spectroscopy has also been extensively used to study the structure of lipid bilayers both in model systems17-19 and in biological membranes.20,21 Thus, it can offer methods required to answer the questions raised by lipid bilayer-polyelectrolyte film interaction. In the present work, we first built a PGA/PLL polyelectrolyte film, and then on the surface of this film a dipalmitoylphosphatidylcholine (DPPC) bilayer was created. The film construction could be continued; thus, the lipid bilayer was embedded into the polyelectrolyte film. According to atomic force microscopy (AFM) data, the surface roughness of the PGA/PLL film was considerably decreased by the lipid bilayer. In addition, the lipid surface completely closed the underlying PGA/PLL film since a new growth regime started when further PGA/PLL layers were adsorbed onto the lipid-covered underlayers. Should the lipid bilayer be permeable for the polyelectrolytes, the exponential growth of the film could continue. The thermotropic response of the νsym(CH2) stretching mode revealed that DPPC preserved its dynamics on the surface and in the interior of the PGA/PLL architecture. To separate the lipid-related temperature-dependent structural changes from the complex behavior of the (polyelectrolyte + lipid) architecture, the infrared spectra were analyzed by singular value decomposition (SVD).22 As a protein model, gramicidin A (GRA) could be stably incorporated into the lipid (11) Garza, J. M.; Schaaf, P.; Muller, S.; Ball, V.; Stoltz, J. F.; Voegel, J. C.; Lavalle, P. Langmuir 2004, 20, 7298-7302. (12) Garza, J. M.; Jessel, N.; Ladam, G.; Dupray, V.; Muller, S.; Stoltz, J. F.; Schaaf, P.; Voegel, J. C.; Lavalle, P. Langmuir 2005, 21, 12372-12377. (13) Debreczeny, M.; Ball, V.; Boulmedais, F.; Szalontai, B.; Voegel, J. C.; Schaaf, P. J. Phys. Chem. B 2003, 107, 12734-12739. (14) Pilbat, A. M.; Ball, V.; Schaaf, P.; Voegel, J. C.; Szalontai, B.; Allakhverdiev, S. I.; Tsvetkova, N.; Mohanty, P.; Szalontai, B.; Byoung, Y. M.; Debreczeny, M.; Murata, N. Langmuir 2006, 22, 5753-5759. (15) Schwinte´, P.; Voegel, J. C.; Picart, C.; Haikel, Y.; Schaaf, P.; Szalontai, B. J. Phys. Chem. B 2001, 105, 11906-11916. (16) Schwinte, P.; Ball, V.; Szalontai, B.; Haikel, Y.; Voegel, J. C.; Schaaf, P. Biomacromolecules 2002, 3, 1135-1143. (17) Casal, H. L.; Mantsch, H. H. Biochim. Biophys. Acta 1984, 779, 381401. (18) Mendelsohn, R.; Moore, D. J. Chem. Phys. Lipids 1998, 96, 141-157. (19) Kota, Z.; Debreczeny, M.; Szalontai, B. Biospectroscopy 1999, 5, 169178. (20) Szalontai, B.; Nishiyama, Y.; Gombos, Z.; Murata, N. Biochim. Biophys. Acta 2000, 1509, 409-419. (21) Szalontai, B.; Kota, Z.; Nonaka, H.; Murata, N. Biochemistry 2003, 42, 4292-4299. (22) Henry, E. R.; Hofrichter, J. Methods Enzymol. 1992, 129-192.

Langmuir, Vol. 23, No. 15, 2007 8237

bilayer. This is promising considering the later incorporation of larger, real membrane proteins, and their functioning in this environment. Such embedded membrane proteins could include for example channel-forming molecules, which then could allow the exchange of specific molecules between the separated polyelectrolyte compartments. Materials and Methods Construction of the Polyelectrolyte Multilayers. For infrared spectroscopy (Bruker IFS66, ATR mode, 1024 scans, 2 cm-1 resolution, MCT detector, Opus 3.1 software), the polyelectrolyte films were constructed on the surface of a ZnSe internal reflection element (ATR crystal). The ATR cell (Circle Cell, Specac, U.K.) was modified; the sample volume of the cell was decreased to about 700 µL by making a new, stainless steel cylinder around the crystal. This cylinder was thermostated by circulating water with the precision of about 0.1 °C. All experiments were performed in tris(hydroxymethyl)aminomethane (Tris, 10 mM, Gibco BRL) and sodium chloride (NaCl, 0.15 M) buffer at pD 6.4 in D2O (99.9% D, Aldrich). The build-up of the polyelectrolyte film started with circulating poly(ethyleneimine) (PEI, Aldrich, MW ) 750 000 g‚mol-1) for 6 min in the form of a 5 mg‚mL-1 solution around the ZnSe crystal. This and all the later adsorption steps were followed with rinsing with pure buffer for 6 min. After PEI, poly(L-glutamic acid) (PGA) (Sigma, P-4886, MW ) 17 000 g‚mol-1), was allowed to adsorb from a 1 mg‚mL-1 solution. After another rinsing, for the same period of time the polycation solution containing 1 mg‚mL-1 poly(L-lysine) (PLL, Sigma, MW ) 32 600 g‚mol-1) was circulated. This procedure was repeated until the PEI-(PGA-PLL)5-PGA architecture was achieved. Creation of a Phospholipid Bilayer on the Surface of the Polyelectrolyte Film. A 2 mg amount of DPPC was dissolved in chloroform:methanol (2:1) in an appropriate test tube. Afterward the solvent was evaporated by N2 flow. Then 8 mL of Tris buffer was added to the dried DPPC, and liposomes were formed by sonication. Sonication was applied in several steps until turbidity was observed in the test tube. No special care was taken to have unilamellar liposomes. For being adsorbed, this liposome solution was circulated first at 25 °C above the (PGA-PLL)5-PGA film. [This architecture is an analogue with that used by Delajon et al.,10 who also incorporated DMPC between negatively charged polyelectrolyte surfaces.] During this period, only a very small lipid signal could be observed in the 2800-3000 cm-1 region of the infrared spectrum, as monitored continuously during the adsorption procedure. Then, the temperature was raised to 45 °C, i.e. above the phase-transition temperature of DPPC (41.6 °C), and the liposomes remained circulating overnight. Then the temperature was decreased to ambient. This heating-cooling cycle resulted in a large increase (up to 10fold) of the lipid signal. Then, the Circle cell was washed with pure buffer. Washing did not affect the amount of the adsorbed DPPC considerably. The intensity of the lipid infrared spectrum remained fairly constant afterward, and the lipid bilayer on the top of the polyelectrolyte film could survive as many phase-transition measurements as wanted. Incorporation of Polypeptides into the Lipid Bilayer. Gramicidin A (GRA, Fluka, MW ) 1884 g‚mol-1) was dissolved together with DPPC in a 1:10 molar ratio in 2,2,2-trifluoroethanol (TFE). After 2 h of incubation at room temperature, the sample was dried first with N2 and then under vacuum overnight. Afterward, all treatments were the same as those described above for pure DPPC. GRA incorporation was evidenced by the appearance of its amide I band in the infrared spectrum of the lipid bilayer. Atomic Force Microscopy. Freshly cleaved mica (SPI-Chem Mica Sheets, Structure Probe, Inc., West Chester, PA) surfaces were used. The mica surface was covered with PEI-(PGA-PLL)5-PGA and finally with DPPC in the same way as described above for infrared spectroscopy. AFM measurements were carried out with an Asylum MFP-3D head and Molecular Force Probe 3D controller (Asylum Research, Santa Barbara, CA). The driver program MFP-3D Xop was written

8238 Langmuir, Vol. 23, No. 15, 2007

Pilbat et al.

Figure 1. C-H stretching, amide I and II regions of the Fourier transform infrared spectra of a polyelectrolyte film PEI-(PGA/PLL)5-PGA (curve a), of a dipalmitoylphosphatidylcholine bilayer built onto a polyelectrolyte film (DPPC) (curve b), and of DPPC bilayer containing gramicidin A also on the top of a polyelectrolyte film (DPPC + GRA) (curve c). For details, see the text.

in IGOR Pro software (version 5.04b, Wavemetrics, Lake Oswego, OR). Silicon nitride cantilevers (Bio-Lever BL-RC150VB-C1, Olympus Optical Co., Ltd., Tokyo, Japan) were used for the experiments. The calibration of the spring constant of each cantilever was performed individually by thermal fluctuation technique.23-25 Measurements were made both in alternative contact (AC) and in contact modes in buffer. Typically 512 × 512 points scans were taken at 1 Hz scan rate. Singular Value Decomposition Analysis of the Infrared Spectra. Singular value decomposition was applied as described in details by Henry and Hofrichter,22 and citatia therein. Briefly, if a system as a function of an external parameter (temperature, concentration, and time, etc.) contains si species, which are spectrally distinguishable, at any moment the measured spectrum can be described as a linear combination of the spectra of these si species. The output of the SVD analysis is a reduced representation of the data matrix in terms of a set of si basis spectra, an associated set of temperature-dependent (or other variable-dependent) amplitude vectors (vi), and a diagonal matrix with the weights of the si species (wi). Since the set of output components is ordered by decreasing size, each subset consisting the first j-components provides the best j-component approximation of the data matrix. Thus, a set of sj output components can be selected, which describe the data matrix within experimental precision. When spectra taken as a function of the temperature are to be analyzed, the s1 basis spectrum shows a temperature average, s2 shows the largest changes, which have to be combined with the s1 spectrum to get the actual spectrum at a given temperature. The v2 amplitude vector gives the temperature dependence of the largest change, manifested in the s2 spectrum. In our case, not only the lipid bilayer but also the underlying/ embedding polyelectrolyte film was changing upon increasing temperature (having different temperature dependence as compared to the lipids). We hoped to separate these contributions by the use of SVD analysis. It turned out that the s2 spectra were very similar for pure DPPC layers, for DPPC bilayers on the surface of the PGA/ PLL film and for DPPC bilayers embedded into the PGA/PLL film. Therefore, their corresponding v2 vectors could be compared and considered as characteristic of the lipid structural changes (given in the s2 spectra) as a function of temperature in the different architectures. (23) Hutter, J. L.; Bechhoefer, J. ReV. Sci. Instrum. 1993, 64, 1868-1873. (24) Butt, H. J.; Jaschke, M. Nanotechnology 1995, 6, 1-7. (25) Florin, E. L.; Rief, M.; Lehmann, H.; Ludwig, M.; Dornmair, C.; Moy, V. T.; Gaub, H. E. Biosens. Bioelectron. 1995, 10, 895-901.

All data processing on the spectra, including the SVD analysis, was performed with the SPSERV spectrum handling software (Dr. Bagyinka, Cs., BRC, Szeged, Hungary).

Results DPPC Bilayer on the Surface of a PGA/PLL Polyelectrolyte Film. The infrared absorption spectrum of a pure PGA/PLL film was calculated from single-beam spectra of PEI and PEI-(PGA/ PLL)5-PGA by taking them as background (I0) and sample (I), respectively (OD ) -lg(I/I0)). In Figure 1, the C-H stretching region of the infrared spectrum of the PGA/PLL film has less intensive, broad features, while the 1350-1800 cm-1 region is showing the strong, characteristic bands of the PGA/PLL complexes, which adopt mostly β-structures. For detailed analysis of the secondary structure of PGA/PLL films, see refs 13, 14, and 26. The infrared absorption spectrum of the DPPC bilayer created on the surface the polyelectrolyte film was obtained by taking the single-beam spectra of PEI-(PGA/ PLL)5-PGA as background and that of PEI-(PGA/PLL)5-PGA-DPPC as sample in the calculation. Upon circulating DPPC liposomes above the polyelectrolyte film, and going above the gel f liquid crystalline phase-transition temperature, the liposomes “melt” into a continuous bilayer; thus, more lipid molecules were in the depth accessible for the evanescent light. Therefore, the bands related to DPPC molecules became more intensive in the infrared absorption spectrum. In Figure 1, the presence of DPPC on the film surface is evident from the characteristic νsym(CH2) band at around 2850 cm-1, the νas(CH2) band at around 2920 cm-1, and the corresponding νsym(CH3), νas(CH3) modes at around 2872 and 2958 cm-1, respectively. The low νsym(CH2) frequency (≈2850 cm-1) is indicative of ordered fatty-acyl chains being in gel phase.17,19 In the infrared spectrum of the DPPC bilayer we measure not only the contribution of the DPPC molecules but also any additional changes, which might have appeared upon the formation of the lipid bilayer. Therefore, the weak features corresponding to characteristic PGA/PLL bands in the DPPC spectrum in Figure 1 indicate small rearrangements in the underlying PGA/PLL film. (26) Boulmedais, F.; Schwinte´, P.; Gergely, C.; Voegel, J. C.; Schaaf, P. Langmuir 2002, 18, 4523-4525.

Lipid Bilayers in Polyelectrolyte Films

Langmuir, Vol. 23, No. 15, 2007 8239

Figure 2. Atomic force microscopic images of pure PEI-(PGA/PLL)5-PGA polyelectrolyte film surfaces (panels A and B), and the same films covered with a DPPC bilayer (panels C and D). The left panels show pictures obtained in AC mode; the right panels were obtained in contact mode. Note that upon covering the polyelectrolyte films with DPPC, surface roughness decreased considerably, and the shape of the extruding features also became smoother. Root mean squared (RMS) roughness for each sample is given in the panels.

Atomic force microscopy has shown (Figure 2) that the surface roughness of the PGA/PLL films decreased considerably when they were covered with a DPPC bilayer. In addition, the extruding features became smoother, more “round” on the DPPC-covered films. It seems that the forces, which keep the DPPC bilayer together, in order to optimize the bilayer structure, can compress the extruding parts of the PGA/PLL films. Barrier Properties of the Embedded DPPC Bilayer. It has been shown earlier with optical waveguide light-mode spectroscopy (OWLS) that PGA/PLL films exhibit exponential growth.4 This exponential growth can be followed with infrared spectroscopy as well, by recording the infrared spectra after the adsorption of each polyelectrolyte layer, as is shown in Figure 3. Here the increasing absorption of the lower set of infrared spectra shows that, indeed, the PGA/PLL film is growing exponentially. Reaching the PEI-(PGA/PLL)5-PGA architecture, a DPPC bilayer was created on the film surface. The upper set of infrared spectra in Figure 3 shows the growth of the additional PGA/PLL layers deposited onto the DPPC-covered film up to the (PGA/PLL)6 architecture. As it can be seen, a new growth regime starts on the top of the DPPC bilayer. (The growth rates were also analyzed; the detailed analysis is given in the Supporting Information.) This means that the DPPC bilayer represents such a barrier in the PGA/PLL film, which does not allow the diffusion of any polyelectrolyte from one compartment to the other. In contrast to Delajon et al.10 who could deposit only a very thin poly(sodium 4-styrenesulphonate)/poly(allylamine hydrochloride) (PSS/PAH) film on top of their DMPC bilayer, we could deposit six PGA/PLL double layers on top of the lipid bilayer without any sign of problem concerning the adsorption of further

Figure 3. Growth of the PGA/PLL film under and above the DPPC bilayer as shown by the infrared absorption spectra recorded layerby-layer during the film build-up. The lower set of spectra shows the six PGA layers up to the PEI-(PGA/PLL)5-PGA architecture. Above the schematically depicted DPPC bilayer, the infrared spectra of the other six PGA layers in the (PGA/PLL)6 architecture, built upon the lipid-covered surface, are shown. These spectra were displaced upward for clarity. A detailed, quantitative analysis of the growth of the film under and above the incorporated lipid bilayer is given in the Supporting Information.

layers. The difference between the PSS/PAH and PGA/PLL films may well be related to their different structures and their different hydrations.

8240 Langmuir, Vol. 23, No. 15, 2007

Pilbat et al.

Figure 4. Singular value decomposition (SVD) analysis of DPPC bilayers in different architectures: (first row) DPPC bilayer on bare ZnSe crystal; (second row) DPPC bilayer on the surface of a PEI-(PGA/PLL)5-PGA film; (third row) DPPC bilayer embedded in a PEI-(PGA/ PLL)5-PGA-DPPC-PGA-(PLL/PGA)5 film. For the preparation of DPPC bilayers from adsorbed liposomes, and for the details of the SVD analysis, see Materials and Methods. For the meaning of “expjump”, see the text.

Dynamics and Phase Properties of the Adsorbed/Embedded DPPC Bilayer. First, as a reference, the DPPC bilayer was prepared on bare ZnSe crystal in the same way as on the surface of PGA/PLL films, from circulating liposomes. Its SVD analysis is shown in the upper row of Figure 4. DPPC-s1 shows the C-H stretching region of the DPPC infrared spectrum averaged over the whole temperature range studied. DPPC-s2 shows the largest change, which, in first approximation, has to be combined with the DPPC-s1 to get the originally measured spectra. (In this approximation we neglect the other changes, whose weights are decreasing rapidly. An illustrative figure taking into account further s spectra, v vectors, and w weights are given in the Supporting Information.) The shape of DPPC-s2 (second panel in the first row of Figure 4) agrees with that obtained for thick hydrated DPPC multilayers measured with transmission infrared spectroscopy.19 We have also shown earlier that the v2 vector belonging to the s2 basis spectrum describes the gel-to-liquid crystalline phase transition of DPPC in a perfect agreement with the conventional method based on the temperature-induced frequency shift of the νsym(CH2) band.19 The third panel of the first row in Figure 4 shows the temperature dependence of DPPC-v2. Note, that in agreement with the average nature of DPPC-s1, the intensity of DPPC-v2 is going from negative to positive having zero intensity around

the middle of the change. By interpolating 500 points between the first and the last measurements using a shifting polynomial of second degree, a smooth curve could be obtained. The first derivative of this curve exhibited one sharp maximum at 42.3 °C, the value of which can be considered as the phasetransition temperature for a pure DPPC bilayer adsorbed onto a bare ZnSe crystal in the presence of D2O. When DPPC was layered onto PEI-(PGA/PLL)5-PGA, both its s1 and s2 basis spectra (Figure 4, second row) were very similar to those of DPPC on the bare ZnSe (Figure 4, first row). This is not surprising since the infrared absorption spectra of the film-adsorbed DPPC bilayer were calculated from single-beam spectra recorded immediately before and after the adsorption of DPPC onto the film. Nevertheless, if there are changes in the polyelectrolyte film accompanying the DPPC bilayer formation, those will also be present in the s1, s2 basis spectra, and in the v2 vector. Not so much the actual structure, but the temperature dependence of the DPPC bilayer has changed on the surface of the PEI-(PGA/PLL)5-PGA film. This is reflected by the altered temperature dependence of the film + DPPC-v2 vector in the last panel of the second row in Figure 4. For this curve, the abovementioned interpolation-derivation method did not provide a narrow, single peak from which the phase-transition temperature could be determined. Therefore, to give at least a phenomeno-

Lipid Bilayers in Polyelectrolyte Films

logical description of the phase transition, we used the function, y ) A/(1 + exp((x - x0)/Γ)) (where A is the amplitude and Γ is the width of the transition), which is generally used to characterize transitions from one state to the other; we call it expjump here. Thus, the 40.3 °C middle temperature and the 6 °C width were obtained for the transition of the film + DPPCv2 vector, which is at somewhat lower temperature and is definitely wider than that found for the DPPC bilayer on bare ZnSe. The most challenging task was to determine the lipid-related changes in the case where DPPC was embedded into a PEI(PGA/PLL)5-PGA-DPPC-PGA-(PLL/PGA)5 architecture. Here, the difference spectra, which included the DPPC bilayer, contained also the six PGA layers above the lipid. Therefore, the bands, characteristic of the lipid C-H stretching region, appeared only as shoulders in the film-DPPC-film-s1 spectrum (Figure 4, third row). Nevertheless, the largest change in this architecture upon increasing temperatures is in the lipid structure, as it can be seen from the large similarity of the film-DPPC-film-s2 spectrum to the DPPC-s2 and film-DPPC-s2 spectra. This result shows the power of the SVD analysis to separate individual phenomena from complex processes. The temperature dependence of the phase transition of the embedded DPPC bilayer is shown by film-DPPC-film-v2 in the third row of Figure 4. This is also a wide transition as in the case of film-DPPC-v2 above. Determined with the same two-state function, we obtained Tm ) 34.1 °C and width ) 10.3 °C for the temperature and the width of the transition, respectively. Gramicidin A (GRA) Incorporation into the DPPC Bilayer. The interaction of GRA with different lipids has already been investigated in detail.27 Here, we first checked the amount of the incorporated peptide among our conditions. Figure 5A shows the relative intensity of the lipid-related ester CdO and peptiderelated amide I bands in a thick sample (dashed line). This thick sample was obtained by drying from organic DPPC + GRA solution first onto the ZnSe crystal, and then the dried layer was hydrated; thus, lipid multi-bilayers (dashed line) could be formed. This relative intensity of the GRA amide I band versus the DPPC ester CdO band is similar (precisely: somewhat lower) to that of found for myristoyl-containing lipids;27 i.e., the intensity of the amide I band is considerably higher than that of the ester CdO band. In contrast, in the case of our method, when liposomes were circulating above the crystal surface and only those molecules remained on the surface, which were able to adsorb by themselves, the relative intensity of the amide I band as compared to the ester CdO is much lower (Figure 5A, continuous line). As a consequence, GRA being in a high proportion in the dried-hydrated sample was able to affect considerably the phase transition of DPPC, in contrast to samples obtained directly from circulating liposomes, where the GRA concentration was much lower (Figure 5B). These latter ones had a very similar phase transition to that of the pure DPPC (shown by DPPC-v2 in Figure 4). This similarity between the behavior of the DPPC and (DPPD + GRA)-containing cases is persistent in all architectures; therefore, we do not show the analysis of the latter ones here; a detailed figure is given in the Supporting Information. The characteristic band of GRA is its amide I band at around 1630 cm-1, but due to the large changes of the intensive bands of the PGA/PLL film in the same region, we could not separate the GRA-related alterations even by SVD analysis as a function of the temperature. Nevertheless, the data show that it is possible to incorporate GRA into the DPPC bilayers, and the GRA molecules remain stably embedded in the lipid environment. (27) Ko´ta, Z.; Pa´li, T.; Marsh, D. Biophys. J. 2004, 86, 1521-1531.

Langmuir, Vol. 23, No. 15, 2007 8241

Figure 5. (A, continuous line) Ester CdO and amide I bands in the infrared absorption spectra of gramicidin A (GRA) containing DPPC bilayers made from liposomes adsorbed onto the ATR crystal from a circulating suspension. This spectrum is five-times magnified for better visibility. (Dashed line) DPPC + GRA dried onto the ATR crystal by evaporating the organic solvent and then hydrated. Note the higher relative intensity of the GRA-related amide I band in the dried-hydrated sample. (B, squares) v2 SVD vector calculated from the infrared spectra of GRA-containing DPPC bilayers obtained from adsorbed liposomes. Its phase-transition temperature is very similar to that of the pure DPPC bilayer shown in Figure 4, the low GRA content little disturbed the DPPC bilayer; (triangles) v2 SVD vector of DPPC + GRA multi-bilayers from dried-hydrated samples. Note that here the phase-transition temperature is considerably downshifted.

Discussion We have shown that DPPC bilayers can be created on the surface of PGA/PLL films terminated with a PGA layer. The DPPC bilayer was smoothing the surface of the polyelectrolyte film. Onto this DPPC bilayer, first, a PGA layer could be adsorbed and then an unlimited number of PLL/PGA layer pairs. The DPPC bilayer forms a barrier, through which the large polyelectrolyte molecules cannot pass. Due to this barrier behavior when the construction of the polyelectrolyte film is continued on the top of the lipid bilayer, the newly added polyelectrolyte layers start a new growth regime, with small adsorption at the beginning, which then grows layer-by-layer. There are several considerations behind speaking consequently about DPPC bilayers among our conditions. One is the above-mentioned experimental proof of having a single DMPC bilayer on the surface of a PAH/PSS film.10 The other is that we think, in our case, there is no energetic reason for having DPPC multi-bilayers. The situation, when the lipid is dried onto a surface by evaporating

8242 Langmuir, Vol. 23, No. 15, 2007

the organic solvents and then is being hydrated, is very different from ours. This difference is especially important when considering the amount of the incorporated peptides and proteins into such systems. In the dried-hydrated case, the lipid-soluble compounds cannot easily leave the lipid multi-bilayers, and depending on the degree up to which they might be “forced” to coexist with the lipids, the observed structural distortions in any of the components of such a system might be “more than natural”. In our approach, the circulation of the liposomes in the presence of excess water, and the cooling-melting cycle provides more dynamics when the lipid layer is created. The components not fitting optimally into such a layer may partition into liposomes or micelles, which are circulating above the forming lipid bilayer. Thus, the lipid bilayer may obtain an energetically optimal structure (i.e., which is free from disturbing peptides and proteins). For more conclusive results, however, further studies, involving other lipids and proteins as well, are needed. Nevertheless, we think that this approach can lead to more realistic lipid-protein ratios and interactions than those used so far in the model systems.

Pilbat et al.

The dynamic properties of the DPPC bilayer, which remained largely similar to that of a pure lipid multi-bilayer system, allow the planning of further studies, in which such DPPC bilayers could be used as membrane models, whose barrier properties can be tuned by selecting proper lipids, by incorporating transporter or pore-forming proteins. Even regulated transport through such lipid bilayers can be imagined. Acknowledgment. This work was supported by a FrancoHungarian bilateral research project (“Balaton”, Grant F-05/34), and by the Hungarian Research Foundation (OTKA, Grant T043425). Supporting Information Available: Figure S1, showing the SVD analysis with s3 and v3 components, Figure S2, showing detailed SVD analysis of gramicidin A-containing DPPC bilayers, and Figure S3, showing analysis of the growth rates in different polyelectrolyte film + DPPC architectures. This information is available free of charge via the Internet at http://pubs.acs.org. LA700839P