Effect of the Electrostatic Charge on the Mechanism Inducing

3074

Langmuir 2004, 20, 3074-3079

Effect of the Electrostatic Charge on the Mechanism Inducing Liposome Solubilization: A Kinetic Study by Synchrotron Radiation SAXS M. Co´cera,† O. Lo´pez,*,† R. Pons,† H. Amenitsch,‡ and A. de la Maza† Departamento de Tecnologı´a de Tensioactivos, Instituto de Investigaciones Quı´micas y Ambientales de Barcelona (I.I.Q.A.B.), Consejo Superior de Investigaciones Cientı´ficas (C.S.I.C.), Calle Jorge Girona 18-26, 08034 Barcelona, Spain and ELETTRA, Sincrotrone Trieste, S.S. 14 Km 163.5, Bassovizza, 34012 Trieste, Italy Received October 22, 2003. In Final Form: February 10, 2004 The anionic surfactant sodium dodecyl sulfate (SDS) was used to induce the initial steps of the solubilization of liposomes. The structural transformations as well as the kinetics associated with this initial period were studied by means of time-resolved small-angle X-ray scattering (SAXS) using a synchrotron radiation source. Neutral and electrically charged (anionic and cationic) liposomes were used to investigate the effect of the electrostatic charges on the kinetics of these initial steps. The mechanism that induces the solubilization process consisted of adsorption of surfactant on the bilayers and desorption of mixed micelles from the liposomes surface to the aqueous medium. In all cases the time needed for desorption of the first mixed micelles was shorter than that for complete adsorption of the surfactant on the liposomes surface. The present work demonstrates that adsorption of the SDS molecules on negatively charged liposomes was slower and release of mixed micelles from the surface of these liposomes was faster than for neutral liposomes. In contrast, in the case of positively charged liposomes, the adsorption and release processes were, respectively, faster and slower than those for neutral vesicles.

Introduction Surfactants are indispensable reagents in the solubilization of membrane proteins. In this sense, study of the interaction of surfactants with liposomes is currently attracting considerable interest.1-3 Given the ionic character of biological membranes, study of the charge effect on liposome solubilization is essential to the understanding of processes in which charges are involved, such as the insertion of proteins, transport of ions and proteins, and cell signaling.4-6 The electrostatic charge does not seem to have a relevant influence at the concentration level at which solubilization occurs. In fact, regardless of the charge, this process has been described as a three-stage model: vesicle saturation, formation of mixed micelles, and complete solubilization.7,8 However, the mechanism that leads to the saturation, i.e., the initial subsolubilizing steps in which a very fast adsorption of surfactant is involved, is not still clear. At this initial level, the effect of the charge has not been evaluated owing to the lack of techniques with sufficiently short experimental time scales. The same problem exists in the study of the * To whom correspondence should be addressed. Phone: 34-93 400 61 61. Fax: 34-93 204 59 04. † Instituto de Investigaciones Quı´micas y Ambientales de Barcelona. ‡ ELETTRA. (1) Deo, N.; Somasundaran, P. Langmuir 2003, 19, 7271. (2) Deo, N.; Somasundaran, P.; Subramanyan, K.; Ananthapadmanabhan, K. J. Colloid Interface Sci. 2002, 256 (1), 100. (3) Tan, A.; Ziegler, A.; Steinbauer, B.; Seelig, J. Biophys. J. 2002, 83 (3), 1547. (4) Miller, C. R.; Bondurant, B.; McLean, S. D.; McGovern, K. A.; O’Brien, F. Biochemistry 1998, 37, 1287. (5) Bordi, F.; Cametti, C.; Motta, A. J. Phys. Chem. B 2000, 104 (22), 5318. (6) Trkanjec, Z.; Demarin, V. Med. Hypotheses 2001, 56, 540. (7) Urbaneja, M. A.; Alonso, A.; Gonza´lez-Man˜as, J. M.; Gon˜i, F. M.; Partearroyo, M. A.; Tribout, M.; Paredes, S. Biochem. J. 1990, 270, 305. (8) De la Maza, A.; Parra, J. L.; Sa´nchez-Leal, J. Langmuir 1992, 8, 2422.

liposome solubilization kinetics. Although a number of works have reported decisive aspects of the dynamics of this process as a whole,9,10 data on the kinetics associated with the fast initial steps of the solubilization are lacking. Thus, the development and application of techniques using synchrotron radiation, which provides high time resolution, are increasingly important. In earlier papers we investigated the solubilization of liposomes by surfactants from a structural viewpoint.11,12 This raised a number of questions about the kinetics. Dynamic light scattering (DLS) and freeze fracture electron microscopy (FFEM) techniques were performed to investigate this kinetic aspect,13,14 and more recent papers have attempted to study this topic in greater depth.15,16 However, the initial steps of solubilization were too rapid to be properly detected. In this work, we seek to describe the initial steps of the liposome solubilization induced by the sodium dodecyl sulfate surfactant (SDS) from mechanistic and kinetic viewpoints. The effect of the electrostatic charge on these initial steps was also investigated. To this end, small-angle X-ray scattering (SAXS) using a stopped-flow cell and synchrotron radiation was used. This technique has proved to be a very good method to study the phase behavior of systems containing (9) Ramaldes, G. A.; Fattal, E.; Puisieux, F.; Ollivon, M. Colloid Surf. B 1996, 6, 363. (10) Sa´ez-Cirio´n, A.; Alonso, A.; Gon˜i, F. M.; McMullen, T. P.; McElhaney, R. N.; Rivas, E. A. Langmuir 2000, 16, 1960. (11) Lo´pez, O.; de la Maza, A.; Coderch, L.; Lo´pez-Iglesias, C.; Wehrli, E.; Parra, J. L. FEBS Lett. 1998, 426, 314. (12) Lo´pez, O.; Co´cera, M.; Wehrli, E.; Parra, J. L.; de la Maza, A. Arch. Biochem. Biophys. 1999, 367, 153. (13) Lo´pez, O.; Co´cera, M.; Pons, R.; Azemar, N.; de la Maza, A. Langmuir 1998, 14, 4671. (14) Lo´pez, O.; Co´cera, M.; Pons, R.; Azemar, N.; Lo´pez-Iglesias, C.; Wehrli, E.; Parra, J. L.; de la Maza, A. Langmuir 1999, 15, 4678. (15) Co´cera, M.; Lo´pez, O.; Estelrich, J.; Parra, J. L.; de la Maza, A. Langmuir 2000, 16, 4068. (16) Lo´pez, O.; Co´cera, M.; Coderch, L.; Parra, J. L.; Barsukov, L.; de la Maza, A. J. Phys. Chem. B 2001, 105, 9879.

10.1021/la035972+ CCC: $27.50 © 2004 American Chemical Society Published on Web 03/16/2004

Solubilization of Charged Liposomes

lipids and surfactants and to monitor very fast biological processes.17-19 Thus, the use of this sensitive methodology opens up new possibilities for the control of processes containing surfactants and lipids from both biological and physicochemical perspectives. Materials and Methods Lipoid S-100, whose main component (>94%) is soybean phosphatidylcholine (PC), was obtained from Lipoid GmbH (Ludwigshafen, Germany). The anionic lipid phosphatidic acid (PA) and the cationic lipid stearylamine (SA) were purchased from Sigma Chemicals Co. (St. Louis, MO). The SA has been used extensively to afford a positive charge in liposomes.20 Tris(hydroxymethyl)aminomethane (TRIS) was obtained from Merck (Darmstad, Germany). TRIS buffer was prepared as 5.0 mM TRIS adjusted to pH 7.4 with HCl, containing 100 mM NaCl. The anionic surfactant sodium dodecyl sulfate (SDS) was obtained from Sigma Chemicals Co. and further purified by a column chromatographic method.21 The surface tensions of buffered solutions containing increasing amounts of SDS were measured by the ring method using a Kru¨ss tensiometer. The surfactant critical micelle concentration (CMC) was determined from the abrupt change in the slope of the surface tension values versus surfactant concentration22 showing a value of 0.75 mM. Reagentgrade organic solvents were purchased from Merck. Polycarbonate membranes and membrane holders were purchased from Nucleopore (Pleasanton, CA). Liposome Preparation. A lipidic film was formed by removing the organic solvent by rotary evaporation from a chloroform-containing Lipoid S-100 solution. Phosphatidylcholine liposomes were obtained by hydration of the lipidic film with TRIS buffer. Thereafter, the liposomes were extruded through 800-200 nm polycarbonate membranes in order to obtain vesicles with a defined size of 200 nm in diameter.23 Neutral liposomes were formed only by Lipoid S-100, whose main component was PC. To obtain liposomes with anionic or cationic character, 5% (w/w) of PA and SA with respect to the total lipid amount were added. Solubilization Parameters. Different systems containing SDS (concentration ranging from 100 to 300 mM) and neutral, anionic, or cationic liposomes (total lipid concentration 30 mM) were studied at the equilibrium. We consider that a system is equilibrated when it reaches a stable state. This state should remain stable until altered by other factors. The solubilization of phospholipid bilayers induced by the SDS surfactant results in changes in the turbidity of the systems, which were monitored by measuring the variations in the static light scattering (SLS).24 The curve obtained allowed us to determine empirically the SDS concentration needed for the complete solubilization of the liposomes, named as SSOL.25 All the SLS measurements were made at 25 °C with a Shimadzu RF spectrofluorophotometer equipped with a thermoregulated cell compartment, with both monochromators adjusted to 500 nm. The assays were carried out in triplicate, and the results given are the average of those obtained. Stopped-Flow Experiments and SAXS Analyses. A volume of 0.5 mL of buffered solutions containing a starting SDS concentration of 480 mM were added to an equal volume (0.5 mL) of neutral, anionic, or cationic liposomes (total lipid (17) Hirai, M.; Kawai-Hirai, R.; Sanafdo, M.; Iwase, H.; Mitsuya, S. J. Phys. Chem. B 1999, 103 (44), 9658. (18) Wang, X.; Semmler, K.; Richter, W.; Quinn, P. J. Arch. Biochem. Biophys. 2000, 377, 304. (19) Woenckhaus, J.; Ko¨hling, R.; Thiyagarajan, P.; Littrell, K. C.; Seifert, S.; Royer, C. A.; Winter, R. Biophys. J. 2001, 80, 1518. (20) Valcarcel, C. A.; Dalla Serra, M.; Potrich, C.; Bernhart, I.; Tejuca, M.; Martı´nez, D.; Pazos, F.; Lanio, M. E.; Menestrina, G. Biophys. J. 2001, 80, 2761. (21) Rosen, M. J. J. Colloid Interface Sci. 1981, 79, 217. (22) Lunkenheimer, K.; Wantke, D. Colloid Polym. Sci. 1981, 259, 354. (23) Dorovska-Taran, V.; Wich, R.; Walde, P. Anal. Biochem. 1996, 240, 37. (24) De la Maza, A.; Parra, J. L Langmuir 1995, 11, 2435. (25) Lichtenberg, D.; Robson, J.; Dennis, E. A. Biochim. Biophys. Acta 1985, 821, 470.

Langmuir, Vol. 20, No. 8, 2004 3075 concentration 60 mM). The mixture led to a final concentration of 240 mM SDS and 30 mM lipid. This dilution did not provide micellar disintegration given that the final concentration was clearly higher than the critical micellar concentration of the surfactant (0.75 mM in the working medium). The interaction between the liposomes and the SDS was monitored by obtaining X-ray scattering curves every 1 s for 300 s. On the basis of an earlier study13-14 we assume that the first steps of the solubilization take place during the first minutes of interaction of liposomes and surfactant. For this reason, the systems were only studied for the first 5 min. The experiments were performed on a stopped-flow cell on a Bio-Logic Stopped Flow Module (SMF), which was fully computer-controlled. The stopped-flow portion of the apparatus comprised four pneumatically driven feed syringes, which were thermostated using a recirculating water bath to 25.0 ( 0.1 °C. The solutions were driven from these syringes through a mixing cell to a quartz capillary observation cell. The total sample wasted per mixing cycle was 100 µL, and the “dead volume” was 550 µL. The “dead time” required to transfer the fresh feed mixture to the observation cell was 1 ms. The time-resolved X-ray diffraction data were collected at the SAXS beam line at the Synchrotron radiation source Elettra (Trieste, Italy) containing 1024 channels and 2 GeV electron storage ring. The SAXS measurements were performed with a resolution of 1-140 nm in real space, with acquisition times of 1 s in time-resolved measurements and 100 s for static measurements. Experiments on the mixtures SDS/liposome were performed in triplicate, and the SAXS experiments were ran 10 times separately with each of the triplicates. The runs were averaged to increase the signal-to-noise ratio. The accepted runs, 5-7 per experiment, were averaged to increase statistics. The point statistic for 1 s curves were around 3% standard error between runs and 1-2% for 10 s acquisition time. The SAXS results concerning the adsorption and desorption times (tad and tde) as well as the q value and intensity of the peaks were the same for each of the triplicates. The data were collected on a linear position-sensitive gas detector Gabriel type, window size 8 × 100 mm, active length of 86.1 mm with a resolution of 0.159 mm/channel,26 which enabled simultaneous detection of the whole resolution range. The scattering intensities were plotted as a function of the scattering vector q defined as q ) (4π sin(θ/2))/λ, in which θ is the scattering angle and λ the wavelength. The radiation wavelength was 1.542 Å. The positions of the diffraction peaks are directly related to the repeat distance of the molecular structure as described by Braggs law 2d sin(θ/2) ) nλ, in which n and d are the order of the diffraction peak and the repeat distance, respectively. In the case of a lamellar structure, the various peaks are located at equidistant positions, qn ) 2πn/d, with qn being the position of the nth order reflection. Concerning the peaks in the liposome scattering curves, the maximum of each peak was assigned to a Bragg’s repetition distance. In addition, as for the pure SDS micelles and mixed micelles, the bands were fitted to a two-shell model.27

Results Static Light Scattering Experiments. Three types of unilamellar liposomes were used: neutral (PC liposomes), anionic (PC/PA liposomes), and cationic liposomes (PC/SA liposomes). The solubilization of these liposomes by SDS was studied by monitoring the variation in the static light scattered (SLS) of each system as a function of surfactant concentration. Measurement of changes in the SLS was determined 20 h after addition of the surfactant to liposomes. This time was chosen as the optimum period to reach the equilibrium.12 Figure 1 shows the variations in SLS for the solubilization of neutral liposome (lipid concentration 30 mM) arising from the addition of different concentrations of SDS. At low surfactant concentration, an initial increase in the SLS was observed until a maximum was reached. This behavior (26) Gabriel, A.; Dauvergne, F. Nucl. Instrum. Methods 1982, 201, 223. (27) Hayter, J. B.; Penfold, J. Colloid. Polym. Sci. 1983, 261, 1022,

3076

Langmuir, Vol. 20, No. 8, 2004

Figure 1. Variation in static light scattering for the solubilization of PC liposomes (lipid concentration 30 mM) induced by addition of increasing concentrations of SDS surfactant. Arrow A corresponds to the composition in which the liposomes are saturated of surfactant, and arrow B indicates the system at which the solubilization of liposomes is completed.

could be associated with a slight variation in the size of the vesicles possibly due to the incorporation of surfactant molecules in the liposome surface;28 the maximum SLS was related to the bilayer saturation by surfactant. Increasing amounts of surfactant led to a decrease in the scattered intensity until a low constant value was reached. This fact is associated with the formation of mixed micelles and consequent progressive liposome solubilization.12 We consider that complete solubilization occurred when the SLS reached the minimum value. Thus, from these curves, the total surfactant concentration producing the saturation (maximum in the SLS, (SSAT)) and the total surfactant concentration producing the complete solubilization of liposomes (minimum in the SLS, (SSOL)) were obtained by graphical methods.28 Points A and B of Figure 1 correspond to these parameters (SSAT ) 50 mM and SSOL ) 200 mM, respectively). The solubilization curves for anionic and cationic liposomes showed similar values for these parameters. This result confirms those previously published, demonstrating that the main forces involved in the incorporation of surfactant molecules, either charged or neutral, into liposomes were hydrophobic in nature. The electrostatic interactions only modulated the permeability alterations of liposomes.8,29 SAXS Experiments. First of all, PC, PC/PA, and PC/ SA liposomes (lipid concentration 30 mM) and pure SDS micellar solutions (240 mM) were analyzed by SAXS. Likewise, mixtures of SDS and neutral, anionic, or cationic liposomes at the same concentrations as the kinetic studies (30 mM PC and 240 mM SDS) were studied 20 h after mixing, when the solubilization had been completed and the mixed micelles were stable.13 Given that the total concentration needed for complete solubilization of liposomes (PC concentration 30 mM) is 200 mM (results obtained by SLS measurements), we assume that a SDS concentration of 240 mM ensures the complete solubilization of the vesicles at the experimental conditions. The recorded X-ray scattering patterns for the three types of liposomes are shown in Figure 2. Mainly one diffraction peak on each scattering curve is seen at q values of 0.094, 0.094, and 0.087 Å-1 for the PC, PC/PA, and PC/SA liposomes, respectively. The peaks correspond to repeating distances of 67, 67, and 72 Å, which can be associated with the thickness bilayer of the PC, PC/PA, and PC/SA liposomes, respectively. These data are in agreement with (28) Partearroyo, M. A.; Urbaneja, M. A.; Gon˜i, F. M. FEBS Lett. 1992, 302, 138. (29) Inoue, T.; Miyakawa, K.; Shimozawa, R. Chem. Phys. Lipids 1986, 42, 261.

Co´ cera et al.

Figure 2. X-ray scattering curves for the neutral (PC), anionic (PC/PA), and cationic (PC/SA) liposomes.

Figure 3. X-ray scattering curves for the pure SDS micelles and for the mixed micelles 20 h after mixing SDS and three different liposomes (PC-SDS, PC/PA-SDS, and PC/SA-SDS).

the usual width described previously for phospholipid bilayers and specifically with the results reported by other authors, in which diffraction techniques were used for measuring this parameter.30 The fact that this value was similar for the three types of liposomes indicates that the electrostatic charge did not affect the thickness of the bilayers; only in the case of the bilayers containing SA did a slightly larger thickness result. Figure 3 shows the scattering curves for the SDS pure micellar solution and those for the three different mixed micelles. The X-ray pattern for the SDS pure micelles present a band located at q ) 0.145 Å-1, for a spacing of 43 Å. Because of the high electrolyte content of these solutions (50 mM TRIS buffer containing 100 mM NaCl), the interparticle correlation was hindered. Thus, the origin of this peak is the electron density difference between the center and the shell of the micelles and the medium. Data obtained in earlier papers on the diameter of the SDS micelles (40 Å) allowed us to attribute this spacing to the diameter of the pure SDS micelles.12,13 The scattering curves shown in Figure 3 for the three liposome-surfactant mixtures, PC-SDS, PC/ PA-SDS, and PC/SA-SDS, 20 h after mixing showed bands at q values of 0.118, 0.118, and 0.110 Å-1, respectively (spacings of 53, 53, and 57 Å), which were associated with the diameter of the mixed micelles. The bands obtained for SDS micelles and for the mixed micelles were fitted to a two-shell model; the absolute intensity of the model was always within 30% of the experimental level.27 The structure factor was considered unity because of the high electrolyte concentration. For the SDS micelles the core radius is 1.79 nm and the particle radius is 2.77 (30) Skalko, N.; Bouwstra, J.; Spies, F.; Stuart, M.; Frederik, P. M.; Gregoriadis, G. Biochim. Biophys. Acta 1998, 1370, 151.

Solubilization of Charged Liposomes

Figure 4. X-ray scattering patterns for the system neutral liposome (PC) and SDS after different times of mixing; these times represent the most relevant stages at which changes in number and position of the diffraction peaks were detected.

nm with electronic densities of 0.293 and 0.361 e/Å3 for the core and shell. The radii compared favorably with the literature values (1.67 and 2.85 nm for the core and particle radii, respectively) in similar conditions, 0.4 M SDS + 0.1 M NaCl.26 For the mixed micelles in the system PC-SDS, the fitted parameters for the shell model were 2.94 nm for the particle radius, 1.99 nm for the hydrocarbon-like core with electronic densities of 0.365 e/Å3 for the outer shell and 0.288 e/Å3 for the core. With these values both the form and absolute scattering intensity were fitted. As for the PC/PA-SDS system, a radius of 3.04 nm for the total particle and 2.02 nm for the core with electronic densities of 0.365 and 0.288 e/Å3 for the shell and core were obtained. Concerning the PC/SA-SDS system, a radius of 2.95 nm for the total particle, 1.99 nm for the core with electronic densities of 0.358 and 0.293 e/Å3 for the shell and core were obtained. In this case, the quality of the fit was clearly worse than for the PC/PA-SDS and PC-SDS systems. This could be due to a small change in the morphology of the micelle. The mixed micelles had slightly bigger core radii, about 0.2 nm, and similar shell thickness. Because of the relative amounts of material, 69 g/L SDS and 21 g/L lipid, the results were consistent with the solubilization of the lipid inside the micelles. On geometric grounds a volume fraction increase from 0.081 (volume of pure SDS micelles) to 0.107 (volume of SDS-lipid mixed micelles) would correspond to a radius increase of 10%, keeping constant the number of micelles, which compared well with the experimental increase of 12%. The differences between the mixed micelles are small, and bigger differences are obtained for the PC/SA-SDS system where the fit was worst. However, the tendency of decreased electronic density for these mixed micelles could be real due to some ion pairing between the SDS molecules and the SA, leading to a lower degree of counterion binding to the micelle. Figure 4 plots the X-ray scattering curves for the system PC-SDS (neutral liposomes and anionic surfactant) after different mixing times. These times represent the most relevant stages at which changes in the number and position of the diffraction peaks and bands were detected. Thus, the scattering curve 1 s after mixing shows two peaks at q values of 0.094 (for a spacing of 67 Å) and 0.145 Å-1 (for a spacing of 43 Å). The first peak was associated with the thickness of the bilayer and coincides with the q value of Figure 2 (PC). The second peak was attributed to the diameter of pure SDS micelles, since its position is very similar to that obtained for the pure micelles (Figure 3, SDS). A third band was detected 2 s after mixing at q

Langmuir, Vol. 20, No. 8, 2004 3077

Figure 5. X-ray scattering patterns for the system anionic liposome (PC/PA) and SDS after different times of mixing; these times represent the most relevant stages at which changes in number and position of the diffraction peaks were detected. Table 1. Values Corresponding to the tad and tde for the Three Systems Studied PC-SDS PC/PA-SDS PC/SA-SDS

tad(s)

tde (s)

12 30 10

2 0-1 10

) 0.118 Å-1 (spacing 53 Å). We attributed this band to the presence of mixed micelles. These micelles were almost spherical given the good fit with model previously described and the scattering curve shown in Figure 3 (PCSDS). Accordingly, the size of the mixed micelles detected by DLS in an earlier work12 yielded a similar value: 50 nm hydrodynamic diameter. The three described peaks remained in the system until 12 s after mixing liposomes and SDS. From this time only two peaks were present in the spectra. It was also noted that the intensity of the peak for the spacing of 67 Å (liposome bilayer) decreased slightly and that for the pure micelles drastically decreased along the experiment. As for the mixed micelles, the intensity of their band increased through the experiment. All these variations point to solubilization of the liposomes by adsorption of pure micelles and formation of mixed micelles. Different periods of time associated with the first steps of solubilization may be established. The adsorption time (tad) was the time needed for the complete adsorption of surfactant on the liposome surface. The desorption time (tde) was that needed for the detachment of the first mixed micelles from the liposome surface. The tad and tde for the three systems are shown in Table 1. Considering these descriptions the time needed for complete adsorption of surfactant on the liposome surface (tad) was 12 s given that around this period of time the band for the pure micelles tended to disappear. As stated before, this band is mainly due to the electronic density contrast and not to the interparticle correlation. The band associated with the mixed micelles was detected 2 s after mixing. As a consequence, we consider this time as that needed for the first mixed micelles starting to desorb (tde, Table 1). Figure 5 plots the scattering curves for different times after mixing PC/PA liposomes and SDS (anionic liposome and anionic surfactant). It can be seen that until 30 s three peaks were observed in the spectra corresponding to the thickness of liposome bilayer at q ) 0.094 Å-1 (spacing 67 Å), pure SDS micelles at q ) 0.145 Å-1 (spacing 43 Å), and mixed micelles at q ) 0.118 Å-1 (spacing 53 Å). From 31 s until the end of the experiment (300 s) only two peaks were present, those associated with the liposome bilayer and those with the mixed micelle diameter. Thus, it should be pointed out that tad was in this case 30 s (Table 1). The

3078

Langmuir, Vol. 20, No. 8, 2004

Figure 6. X-ray scattering patterns for the system cationic liposome (PC/SA) and SDS after different times of mixing; these times represent the most relevant stages at which changes in number and position of the diffraction peaks were detected.

mixed micelles were detected from the first second, that is, tde was less than 1 s (Table 1). As in the case of the interaction between neutral liposome and SDS, the peaks for the liposome bilayers and for the pure micelles relatively decreased in intensity with respect to the mixed micelles band, indicating the progressive solubilization of the bilayer by action of the SDS. As for the experiments performed with the cationic liposomes and the SDS (system PC/SA-SDS), the different scattering curves are depicted in Figure 6. Here, it can be seen that for times lower than 10 s, only two peaks were detected in the scattering curves, those associated with the liposome (q ) 0.087 Å-1, spacing 72 Å) and with the pure surfactant micelles (q ) 0.145 Å-1, spacing 43 Å). X-ray scattering data collected at times above 10 s also showed two peaks, that for the liposomes and that corresponding to the mixed micelles (q ) 0.114 Å-1, spacing 55 Å-1). Thus, all data indicate that for this system both tad and tde were 10 s (Table 1). These results indicate that although tad and tde were different depending on the charges present in each system, the mechanism involved in the saturation of vesicles by SDS was similar in all cases. This mechanism consists of a surfactant adsorption on the liposome surface and a mixed micelles desorption from the liposome surface to the medium. Although it is reasonable to assume that the adsorption process is prior to the desorption one, tad and tde values indicate that these processes were simultaneous at least for the system PC/PA-SDS and to a minor degree for the PC-SDS system (tde < tad). The PC/SA-SDS system was the only that required complete adsorption of the SDS molecules on the liposomes to start the desorption process (tde ) tad). Discussion The feasibility of the time-resolved SAXS studies using a stopped-flow device has been described in recent works. Thus, Schmolzer et al.31 published a time resolution in the millisecond range. This methodology was used by these authors for describing the transformation of mixtures of cationic and anionic micelles to vesicles. In fact, these systems are very similar to those studied in our work. Another interesting article is that published by Zanini et al.,32 in which structural transitions in the submillisecond (31) Schmolzer, S.; Grabner, D.; Gradzielski, M.; Narayanan, T. Phys. Rev. Lett. 2002, 24, 258301. (32) Zanini, F.; Lausi, A.; Savoia, A. Genetica 1999, 106, 171.

Co´ cera et al.

time region are described. This research was carried out at Elettra facilities as with the experiments performed in our work. It should be noted that the peak corresponding to the bilayer remained 5 min after mixing liposomes and SDS in all systems. This is confirmed by SAXS, turbidity measurements based on SLS (5 min after mixing, data not shown), and DLS results, which demonstrated that the SDS surfactant needed 10 h for the complete solubilization of liposomes.13 The adsorption of surfactant on the liposomes involves a breaking of the micellar structure by release of the surfactant molecules from micelles to monomers (associated with the micellar relaxation time) and the incorporation of these monomers into the bilayers. The relaxation time (described as the time needed for the release of the surfactant molecules from micelles to monomers) for 240 mM SDS in water is about 3 s, and it seems that in the present work, in which salt is present, the value expected could be even higher.33 This value is lower than the adsorption times (tad) determined in this work. In any case, to attribute the relaxation time described by Patist et al.33 to our systems is risky since this parameter has been described for systems in which no liposomes were present. In this work, we determined the time for the surfactant adsorption, which includes both the release of the surfactant molecules from micelles to monomers (relaxation time) and the incorporation of these monomers into the bilayers. Table 1 compares the results obtained from the scattering curves of the three systems studied. It should be noted that both tad and tde depended on the electrostatic charge of the species involved in the systems, i.e., the electrostatic charge associated with the surfactant and the lipids. Thus, our results indicate that the shortest tad corresponds to the system composed of PC/SA liposomes and SDS. In other words, the adsorption process is faster when the electrostatic charges of the lipid and surfactant are opposite. In this regard, for the system PC/PA-SDS (lipid and surfactant with the same type of charge), the process of surfactant adsorption proved to be the slowest (tad ) 30 s). As for the system in which neutral lipids were used to form the liposomes, tad yielded an intermediate value (12 s). These results highlight an important role of the electrostatic charge in the kinetics of surfactant adsorption. Similar results have been obtained in different processes also related to the adsorption of molecules on ionic liposomes.34,35 Given that the adsorption of surfactant involves the proximity of the liposomes and pure SDS micelles, it seems that this approximation could be hindered or enhanced by the electrostatic potential gradient. In general, for noncharged vesicles, the driving force involved in the adsorption of surfactant (and other agents) is based on the hydrophobic/hydrophilic nature of the species. However, when electrostatic charges are present, in both liposomes and surfactant, the adsorption also depends on the nature and probably on the amount of net charge. In fact, in some systems the hydrophobic and electrostatic factors could have counteracting effects.36 A similar behavior was reported by Ridder et al.,37 who found that the presence of anionic lipids induced the adsorption and insertion of proteins in lipid bilayers. In (33) Patist, A.; Chhabra, V.; Pagidipati, R.; Shah, R.; Shah, D. O. Langmuir 1997, 13, 432. (34) Campanha, M. T. N.; Mamizuka, E. M.; Carmona-Ribeiro, A. M. J. Phys. Chem. B 2001, 105 (34), 8230. (35) Matsiu, H.; Pan, S. Langmuir 2001, 17, 571. (36) Philippova, O. E.; Andreeva, A. S.; Khokhlou, A. R.; Islamov, A. Kh.; Kuklin, A. I.; Gordeliy, V. I. Langmuir 2003, 19, 7240. (37) Ridder, A. N. J. A.; Kuhn, A.; Killian, J. A.; de Kruijff, B. EMBO Rep. 2001, 2, 403.

Solubilization of Charged Liposomes

Langmuir, Vol. 20, No. 8, 2004 3079

the systems studied the surfactant adsorption always took place even when the electrostatic factor ran counter to adsorption (system PC/PA-SDS anionic charges in both liposome and surfactant). Thus, we assume that the hydrophobic/hydrophilic factor is more important than the electrostatic factor, at least for the intensity of the charge used in our experiments (5% of ionic lipids). However, we could hypothesize that the presence of higher charge values could modulate the adsorption process, even blocking it. Considering that the proportion of electrostatic charge present in biological membranes is significantly higher (the inner membrane of Escherichia coli contains 75% of zwitterionic lipid and 25% of negatively charged lipid37), our hypothesis could help to explain some questions about the higher or lower ability of the biomembranes to adsorb ionic molecules as a function of the electrostatic charge of these membranes. As a consequence, some aspects of different biological processes related to the adsorption, such as fusion, endocytosis, viral infection, etc., could be clarified. Nevertheless, we are aware that further studies are warranted. As for desorption of the mixed micelles from the liposome surface, it may be seen that regardless of the electrostatic charges, the desorption process was generally faster than the adsorption one, i.e., tde e tad. This could be related to the thermodynamics associated with the liposome solubilization. In this regard, other authors have described a strong dependence of the thermodynamics on processes involving lipid/surfactant mixtures.38 The liposomes (formed only by phospholipids) are fairly stable structures, and adsorption of surfactant induces the formation of unstable mixed liposomes formed by lipids and surfactant.16 However, desorption of mixed micelles from the mixed liposome surface involves formation of very stable structures, the mixed micelles. Thus, given that the formation of more stable structures is in general thermodynamically favored, the different stability of the structures resulting from surfactant adsorption and mixed micelles desorption could play a key role in determining the different kinetics of these two processes. Regarding the effect of electrostatic charges on the desorption of mixed micelles from the liposome surface, Table 1 shows that this process is faster when the liposomes and the surfactant have the same type of charge (negative charge). However, for mixtures in which the surfactant and lipids are oppositely charged, the desorption process became slower. For the system with neutral liposomes (formed by PC-SDS), tde bore a greater resemblance to that with equally charged species (PC/PASDS) than to that with oppositely charged species (PC/SA-SDS). Thus, we can assume that the presence of opposite charges slows down the process of mixed micelle desorption. Moreover, our results suggest that the importance of the charge on the lipids lies in the electrostatic potential gradient induced by these lipids versus other charged substances. This could explain the role of ionic lipids in complex biological processes such as exocytosis of synaptic vesicles, secretion of membranous vesicles, etc., in which a detachment of vesicles takes place.39 Our

SAXS results show the overlap in the time of the two phenomena (the SDS adsorption and the mixed micelles desorption). On one hand, the pure SDS micelles are detected only up to 10-30 s after mixing and consequently the SDS adsorption on PC, PC/PA, and PC/SA liposomes is very fast and complete. On the other hand, the first mixed micelles are detected in a shorter or equal period of time than the time needed for finishing the adsorption. However, the liposome solubilization is not completed for the experiment (300 s). The solubilization is finished when all the liposome has been transformed in mixed micelles (about 8 h). In summary, the SDS adsorption and the mixed micelles desorption are very fast (seconds) in comparison with the complete liposome solubilization (hours). This means that a lag period occurs between surfactant adsorption and complete solubilization from the lamellar to micellar phase. Some authors have related the slow kinetics of the liposome solubilization by the SDS to this lag period as follows: the marked hydrophilic character of its head polar group would hinder the SDS translocation to the inner leaflet of the liposomes.3,40 In connection, the solubilization mechanism proposed by Lo´pez et al.11 that included the formation of mixed micelles “in situ” on the bilayer, before their release to the bulk of the solution, could explain these facts. The mixed micelle formation “in situ” would require the translocation of the SDS molecules to the inner monolayer, and this slow movement40 would affect the solubilization kinetics. We are aware that a number of parameters, such as proportion of ionic lipids, type of surfactant, ionic force, and membrane model, affect the kinetics of solubilization. Some of these parameters have been previously studied,13 whereas others need to be explored in future work. The findings reported in the present work underline the importance of the electrostatic charge on the kinetics of surfactant adsorption and on the mixed micelle desorption processes. Although the kinetic differences are only a matter of few seconds, it should be bore in mind that a number of biological processes associated with membranes are really dependent on short periods of time. This and the fact that the electrostatic charges can either accelerate or slow these processes could be the key to modulating the organization and functional properties of membranes using charged molecules as control tools.

(38) Opatowski, E.; Lichtenberg, D.; Kozlov, M. Biophys. J. 1997, 73, 1458. (39) Stefani, G.; Onofri, F.; Valtorta, F.; Vaccaro, P.; Greengard, P.; Benfenati, F. J. Physiol. 1997, 504, 501.

LA035972+

Abbreviations CMC, critical micellar concentration; SLS, static light scattering; PA, phosphatidic acid; PC, phosphatidylcholine; SA, stearylamine; SAXS, small-angle X-ray scattering; SDS, sodium dodecyl sulfate; TRIS, tris(hydroximethyl)aminomethane. Acknowledgment. The authors wish to thank the staff of the Small-Angle Scattering beamline, synchrotron ELETTRA, Italy, for their skillful assistance. We are indebted to the Company ORTEVE, Barcelona, Spain, for the provision of the lipids (Lipoid S-100). This work was supported by funds from the European Community (Project HPRI-CT-1999-00033) and from CICYT (MAT2001-1188-CO2-02).

(40) Kragh-Hansen, U.; le Marie, M.; Moller, J. Biophys J. 1998, 75, 2932.