pubs.acs.org/Langmuir © 2009 American Chemical Society
)
Rate of Permeabilization of Giant Vesicles by Amphiphilic Polyacrylates Compared to the Adsorption of These Polymers onto Large Vesicles and Tethered Lipid Bilayers F. Vial,† F. Cousin,‡ L. Bouteiller,§, and C. Tribet*,†
)
† Laboratoire de Physico-chimie des Polym eres et des Milieux Dispers es, UPMC and CNRS UMR 7615, ESPCI, 10 rue Vauquelin, F-75005 Paris, France, ‡Laboratoire L eon Brillouin, CEA Saclay, 91191 Gif sur Yvette Cedex, France, §UPMC Univ Paris 06, UMR 7610, Chimie des Polym eres, F-75005 Paris, France, and CNRS, UMR 7610, Chimie des Polym eres, F-75005 Paris, France
Received January 21, 2009. Revised Manuscript Received March 9, 2009 We examined by fluorescence microscopy the permeabilization of giant vesicles by hydrophobically modified polyacrylates (called amphipols). Amphipols trigger permeabilization to FITC-dextran of egg-PC/DPPA vesicles with no breakage of the lipid bilayers. The polyanionic amphipols were passing through bilayers as shown by permeabilization of multilamellar vesicles. Remarkably, the vesicles were not simultaneously permeable but became leaky one after the other. Altogether, our observations suggest a random formation of pores having diameters above a few nanometers. Decreasing pH and increasing ionic strength and polymer concentration were increasing the rate of permeabilization. The rate and efficiency of permeabilization was compared to the rate and density of adsorption of amphipols onto lipid membranes (as estimated by titration calorimetry onto large unilamellar vesicles and neutron reflectivity measurements on tethered bilayers). The polymer adsorption layer is built up in a few minutes. We conclude that the rate-limiting step for permeabilization is not the adsorption from the bulk solution but relates to slow intramembrane reorganizations.
Introduction The development of novel agents for permeabilization of lipid membranes recovers fundamental aspects of the stability of lipid and cell membranes, and has practical impacts in galenic pharmacy, or cosmetics. Polymers attached to lipids have recently emerged as interesting alternatives to small molecular agents. Pharmaceutical prospects include, for instance, antimicrobial compounds and cost-effective and robust mimics of natural polypeptides capable of permeabilizing the outer membrane of bacteria or fungus.1,2 Substantial challenges are also identified to appropriately design mixed lipid/polymer “drug cargoes”3-5 and polymer-coated liposomal formulations in the currently approved drug delivery systems.6 It was shown that macromolecular coronae can provide liposomes with the ability to respond to stimuli, and to trigger endocytosis,7 formation of lipid domains,8 or transmembrane delivery on demand.9 Amphihilic polymers appear especially suited at forming hydrophobic noncovalent, although tight associations with lipids which in turn couple stimuli-responsiveness of the polymers with bilayer properties (e.g., variation of the hydrophobicity of polymers with pH, *Corresponding author. E-mail:
[email protected]. (1) Zasloff, M. Nature (London) 2002, 415, 389. (2) Sisson, A. L.; Shah, M. R.; Bhosale, S.; Matile, S. Chem. Soc. Rev. 2006, 35, 1269. (3) Lukyanov, A. N.; Torchilin, V. P. Adv. Drug Delivery Rev. 2004, 56, 1273. (4) Malmsten, M. Soft Matter 2006, 2, 760. (5) Ruysschaert, T.; Sonnen, A. F. P.; Haefele, T.; Meier, W.; Winterhaltert, M.; Fournier, D. J. Am. Chem. Soc. 2005, 127, 6242. (6) Allen, T. M.; Cullis, P. Science 2004, 303, 1818. (7) Gerasimov, O. V.; Boomer, J. A.; Qualls, M. M.; Thompson, D. H. Adv. Drug Delivery Rev. 1999, 38, 317. (8) Binder, W. H.; Barragan, V.; Menger, F. M. Angew. Chem., Int. Ed. 2003, 42, 5802. (9) Yessine, M. A.; Leroux, J. C. Adv. Drug Delivery Rev. 2004, 56, 999. (10) Thomas, J. L.; Barton, S. W.; Tirrell, D. A. Biophys. J. 1994, 67, 1101. (11) Linhardt, J. G.; Tirrell, D. A. Langmuir 2000, 16, 122. (12) Tribet, C.; Vial, F. Soft Matter 2008, 4, 68.
7506 DOI: 10.1021/la900261s
temperature, etc.).10-12 A particular group of these macromolecules, namely, short amphiphilic copolymers called amphipols, has been extensively considered by biologists as an alternative to detergents for the solubilization and handling of integral membrane proteins.13 The hydrophobicity of polymers plays the central role in their detergency, as shown by the correlation found between water solubility of polymer chains and destabilization of polymercoated vesicles. Controlling the hydrophobic interactions between polymers and lipid bilayers, or biomembranes, certainly ensures practical achievement of permeabilization. Most studies have accordingly focused on the design of hydrophobe-containing polymers to achieving control of membrane breakage in biologically relevant conditions (pH, temperature, ionic strength) and control of intracytosolic delivery.9,14-16 Of importance to the handling of cell membranes (e.g., for mild cytosolic infusion of probes and/or bioactive proteins in the cell culture medium), some amphipols with balanced amphiphilicity permeabilize the membranes with no solubilization of the lipids and in the absence of membrane disruption.17 The kinetics of membrane response to perturbation by macromolecules remains however surprisingly slow. Minutes to hours long lag times are sometimes needed between the addition of polymers and the onset of leakage of the membranes.18 The corresponding organization of polymers on/in (13) Popot, J. L.; Berry, E. A.; Charvolin, D.; Creuzenet, C.; Ebel, C.; :: Engelman, D. M.; Flotenmeyer, M.; Giusti, F.; Gohon, Y.; Herve, P.; Hong, Q.; Lakey, J. H.; Leonard, K.; Shuman, H. A.; Timmins, P.; Warschawski, D. E.; Zito, F.; Zoonens, M.; Pucci, B.; Tribet, C. Cell. Mol. Life Sci. 2003, 60, 1. (14) Yessine, M. A.; Meier, C.; Petereit, H. U.; Leroux, J. C. Eur. J. Pharm. Biopharm. 2006, 63, 1. (15) Roux, E.; Francis, M.; Winnik, F. M.; Leroux, J. C. Int. J. Pharm. 2002, 242, 25. (16) Chen, T.; McIntosh, D.; He, Y. H.; Kim, J.; Tirrell, D. A.; Scherrer, P.; Fenske, D. B.; Sandhu, A. P.; Cullis, P. R. Mol. Membr. Biol. 2004, 21, 385. (17) Binder, W. H. Angew. Chem., Int. Ed. 2008, 47, 3092. (18) Vial, F.; Rabhi, S.; Tribet, C. Langmuir 2005, 21, 853.
Published on Web 04/16/2009
Langmuir 2009, 25(13), 7506–7513
Vial et al.
Article
Scheme 1. Structures of the Random Copolymers 5-25C8 (25 ( 2 mol % Octyl Groups) and 5-25C8-40C3 (x = 28 ( 3 mol % Octyl and 41 ( 4 mol % Isopropyl Side Groups)
the lipid layers at incipient destabilization of the membranes has not yet been examined. At least two stages of membrane reorganization can be suspected to be rate-limiting in the process of permeabilization: (i) the formation of the mixed polymer/lipid layer by slow adsorption from solution18,19 and (ii) “intramembrane” steps such as variation/fluctuations of the polymer surface concentration, or spontaneous opening of transmembrane channels, whose lifetime depends on the presence/absence of adsorbed polymer. Here, we examined the permeabilization of giant vesicles and adsorption, density, and thickness of polymer layers on model lipid bilayers with the aim of determining whether the ratelimiting step is controlled by adsorption or by “intramembrane” reorganizations. Short octyl-modified polyacrylates (also called amphipols13) were studied as representatives of the amphiphilic polyacids often developed to permeabilize liposomes.9,14,15 These weak polyacids belong to the class of chains whose effective hydrophobicity increases with decreasing pH and increasing ionic strength. Accordingly, optical microscopy on giant vesicles loaded with a soluble fluorescent probe provided evidence for pH and ionic strength dependent leakage with no membrane breakage. Isothermal titration calorimetry (ITC) was carried out on solutions of large vesicles (prepared by extrusion) and enabled one to value the rate of polymer adsorption. The thickness and density of adsorbed polymer layers were characterized by neutron reflectivity experiments on tethered lipid membranes. We show that the polymer layer forms rapidly as compared to the time needed for the onset of membrane leakage, pointing to the predominance of a slow intramembrane reorganization as the rate-limiting step. These results should help the understanding of the formation of stable pores in membranes, and the practical purposes of mild permeabilization by synthetic channels.
Materials and Methods Polymers. The polymers, 5-25C8 and 5-25C8-40C3 (Scheme 1), were obtained by coupling in N-methylpyrrolidone as solvent, octylamine, and isopropylamine, respectively, to a sample of poly(acrylic acid) (Aldrich, molar mass = 5000 g/mol under its acid form, Ip ∼ 1.6) following a protocol described elsewehere.20 With the aim of enabling measurements of polymer concentrations by spectrophotometry, 5-25C8 also contains 4 mol % naphtylamino side groups that were grafted in the chain with the same procedure as for the octylamine groups. In addition, previous studies by fluorescence have provided clear evidence for the binding on small vesicles of this naphtyl-labeled polymer. 18 Stock solutions (1-10 g/L) were prepared by dissolving freeze-dried samples in deionized water under gentle stirring for at least 2 h at room temperature, prior to mixing with concentrated buffer/glucose solution in order to prepare (19) Ferri, J. K.; Miller, R.; Makievski, A. V. Colloids Surf, A 2005, 261, 39. (20) Ladaviere, C.; Tribet, C.; Cribier, S. Langmuir 2002, 18, 7320.
Langmuir 2009, 25(13), 7506–7513
polymer solutions in the same buffer as the buffer used for vesicles (mixing procedures described in the following).
Giant Vesicles: Preparation and Fluorescence Microscopy. Stock solutions of giant vesicles having a diameter in the
range 5-50 μm were prepared by spontaneous swelling of dried films of lipids in 300 mM sucrose, 10 mM BisTRIS pH 6.8 buffers containing varying concentrations of NaCl, and the soluble fluorescent probe, FITC-dextran (FD-20S, Mw = 17 500 g/mol and hydrodynamic radius of 3.2 nm, 21 Sigma Chemical, MO; FITC stands for conjugates with fluorescein isothiocyanate, here a fluoresceine/monomer ratio of 0.01 mol/ mol). Droplets of about 0.5 μL of 10 g/L lipid solution in chloroform (egg-phosphatidylcholine, dipalmitoyl-phosphatidic acid, and rhodamine-modified phosphatidylethanolamine 88:10:2 mol % from Avanti Polar lipids Incs, AL) were spotted on a flat Teflon sheet that had been scratched with a razor blade in order to avoid dewetting. Prior to deposition of lipids, the Teflon was cleaned in a 50% v/v mixture of 96% sulfuric acid and 35 wt % hydrogen peroxide and extensively washed in water. Stains of lipid films formed on the Teflon supports by evaporation in air of chloroform and were dried under vacuum for 4 h at room temperature. The spontaneous swelling was achieved with no agitation at room temperature. To observe vesicles by fluorescent microscopy (LEICA DM-IRE2), the stock vesicle sample was diluted by 20-fold in a buffer having the same composition and pH as the swelling buffer but with no fluorescent probe, and 300 mM glucose instead of sucrose (quoted below “dilution buffer”). The same osmolarity was reached in the inner and outer compartments of the vesicle, and the presence sucrose inside made the vesicle slightly denser than the outer phase, which enables their sedimentation onto the bottom plate of the microscope for easier observation. A droplet (25-50 μL) of the diluted sample was deposited on a bovine serum albumin (BSA)-passivated glass plate (i.e., plate that has been coated with BSA to avoid irreversible adhesion of the lipids on raw silica; the plates were soaked in a water solution of ∼50 g/ L BSA for 10 min and rinsed with water). During a brief incubation (5-10 min), the vesicles sedimented on the bottom of the glass plate. To mix vesicles with polymer, an aliquot of 25-50 μL of polymer stock solution (in the same dilution buffer as the vesicles) was added at time quoted “zero” by simple pipetting (Gilson P100 pipetman) and homogenized with gentle agitation with the tip of the pipet. The fluorescence of giant vesicles’ membrane was excited at 515-560 nm, while the fluorescence of the internal compartments was excited at 460500 nm. The fluorescence was recorded on a CoolSnap cf monochrome camera (Roper Scientific, Germany) under fixed exposure conditions (typical exposure time 200 ms, binning 2, which corresponded to a resolution of ∼0.5 μm) and analyzed with MetaView 6.3 software to measure fluorescence intensities. Large Vesicles (LVs) of Egg-PC or DMPC. Lipid dispersions were obtained by hydration in pure D2O (DMPC for neutron reflectivity) or in BisTris-HCl buffers (egg-PC/DPPA for ITC titrations) of vacuum-dried lipid films (formed by (21) Ladokhin, A. S.; Selsted, M. E.; White, S. H. Biophys. J. 1997, 72, 1762.
DOI: 10.1021/la900261s
7507
Article evaporation of 1 g/L lipid solution in chloroform). The dispersions were filtered on polycarbonate membranes (magna PCTE pore diameter 0.45 μm, three filtrations) and subjected to five extrusions through membranes of pore diameter 0.22 μm (Osmonics). This procedure has been shown to form predominantly unilamellar vesicles (LUVs).22 The final concentration of lipids after filtration (measured by 31P NMR) indicated that more than 95% of the lipids films was recovered in solution.20 In order to facilitate adsorption and formation of polymer-tethered bilayers in neutron reflectivity experiments, an aliquot of ∼2 mol % cetrimide (hexadecyl trimethylammonium bromide, final concentration 0.013 mM) was added in the LUV solutions (∼0.5 g/L lipid, i.e., ∼0.7 mM). Isothermal Titration Calorimetry (ITC). Measurements were carried out with a VP-ITC instrument from Microcal Inc. (Northampon, MA). Prior to each experiment, the sample cell (volume = 1.43 mL) was filled with large unilamellar vesicles (LUV, 1 g/L) in buffer. LUVs were prepared from egg-PC/ DPPA 9:1 g/g dispersions in 10 mM BisTris pH 6.8 with 40-600 mM NaCl (vide supra). The LUV solution and polymer solution (3 g/L) prepared in the same buffer were in addition dialyzed overnight against the pH buffer (both samples plunged in the same external bath) to ensure equilibration of pH and NaCl. The heats evolved by lipid/polymer mixing were obtained from titrations of polymer solutions (0-5.0 g/L) into the LUV solutions. The polymer solutions were placed in a 300 μL syringe continuously stirred (300 rpm) during the titration, allowing rapid mixing, and aliquots (2-10 μL) were injected into the sample cell in intervals of 300 s. All measurements were carried out at 25 °C. Each titration was performed twice to ensure reproducibility of the results. Data were analyzed using the Microcal ORIGIN software by integration of the raw data signal to give the experimental enthalpy change hi resulting from the ith polymer solution injection. Tethered Bilayers and Neutron Reflectivity. Silicon wafers of diameter 50.8 mm with one side polished were purchased from ACM (Villiers Saint Frederic, France). They were cleaned by treatment in hot piranha solution for 45 min (30% vol H2O2/ H2SO4 30:70 wt/wt) and then thoroughly washed with Milli-Q water. A silica layer formed on the top of the wafer, and its thickness (of 2.1 ( 0.2 nm) was measured by ellipsometry (Sentech SE 400, Germany). In order to provide an hydrophilic and soft interface for the deposition of lipid bilayers, a “cushion” of polyelectrolytes with controlled thickness was deposited on the wafer using the layer-by-layer technique. All polyelectrolytes (poly(ethylene imine), Mw = 750 000 g/mol; poly(styrene sulfonate), PSS, Mw = 70 000 g/mol, and poly(allylamonium chloride), PAH, Mw = 65 000 g/mol) were purchased from Aldrich and dissolved in Milli-Q water under mild agitation for at least 24 h. All solutions were filtered though a 0.45 μm syringe filter. First, a cationic layer was adsorbed by dipping the wafers for 15 min in 2.4 g/L poly(ethylene imine) solution in water. The surface was rinsed in Milli-Q water for 5 min, prior to adsorption of six layers of poly(styrene sulfonate)/poly(allylammonium). Each layer was obtained by a cycle of immersions of the wafer including (i) 10 min in 4 g/L PSS, 1 M NaCl, (ii) 5 min in Milli-Q water, (iii) 10 min in 4 g/L PAH, 1 M NaCl, and (iv) 5 min in Milli-Q water. A top layer of PSS was applied by ending with steps (i) and (ii). The specular neutron reflectivity experiments were carried out on the time-of-flight reflectometer EROS in Laboratoire Leon Brillouin (LLB). The measurements were performed at the silicon/D2O interface in a closed cell (the neutron beam entering through the silicon), enabling one to flush aqueous solutions or air on the wafer. Pure D2O or 150 mM NaCl in D2O was initially introduced in the cell for a first measurement with only hydrated polyelectrolyte layer on the top of the wafer. Other measurements (22) Traikia, M.; Warchawski, D. E.; Recouvreur, M.; Cartaud, J.; Devaux, P. Eur. Biophys. J. 2000, 29, 184.
7508 DOI: 10.1021/la900261s
Vial et al.
Figure 1. Fluorescences of dextran-loaded giant vesicles with rhodamine-coupled lipid in the membrane (in the absence of polymer in dilution buffer). (A,B) Epifluorescence with excitations at, respectively, 515-560 nm and 460-500 nm. (C) Internal fluorescence intensity (FITC-dextran) as a function of vesicle diameter. (D) Histogram of the distribution of lipid fluorescence (Rho-lipid) in a random sampling of 44 giant vesicles showing the fractions of GUV, bilamellar, and other multilamellar objects. were performed at 25 °C, after replacement of this top liquid by either pH buffer, LUVs, or polymer solutions in the same buffer as described in Results section. Removal of a solution for air and filling the cell with a fresh solution took less than 1 min. Data were collected at a fixed angle θ of 1.34° (Δθ ∼ 0.05°), with a neutron white beam covering wavelengths from 4 to 25 A˚ (Δλ/λ = 0.07) and momentum of transfer range 0.012 A˚-1 < Q < 0.07 A˚-1 (with Q = 4π sin θ/λ). During the experiments, both the cell for measurements and the D2O can provide significant incoherent scattering which acts as a background for our samples. This intensity of incoherent scattering has been measured in an off-specular geometry, far away from the specular reflection, and has then been subtracted from the reflectivity curves. The scattering length density Nb profiles obtained from the reflectivity measurements were analyzed using a ,box. model consisting of dividing the film at the Si/D2O interface into a series of layers. Each layer is characterized by a finite thickness, scattering length density Nb, and interfacial roughness with the neighboring layer. The reflectivity curves were calculated using the optical matrix method. The calculation takes into account the Q-resolution of the spectrometer.
Results Permeabilization of Giant Vesicles. We examined the leakage of giant vesicles loaded with fluorescent probes in the membrane(s) (by incorporating a rhodamine-coupled lipid in lipid formulation) and in the internal compartment (by encapsulating a fluorescent dextran, see Materials and Methods section). Figure 1A,B shows representative images of GUVs with membrane fluorescence excited at 515-560 nm, while internal fluorescence was excited at 460-500 nm. In the absence of polymer in the dilution buffer, the internal fluorescence was homogeneously distributed among the vesicles and varied in proportion to their diameter (Figure 1C, and see the Supporting Information for details on determination of intensities), as Langmuir 2009, 25(13), 7506–7513
Vial et al.
Article Table 1. Fraction of Emptied Giant Vesicles (% within a Random Sampling of 50 Vesicles, Unilamellar or Multilamellar) Following Addition at Time Zero of 5-25C8 up to 0.5 g/L in the External NaCl, 10 mM Bis-Tris-HCl pH 6.8 Buffer no NaCl incubation time
GUV
GMV
150 mM NaCl GUV
GMV
1 M NaCl GUV
GMV
b
Figure 2. Example of dextran-loaded vesicles coexisting with a leaky vesicle at time 24 h after addition of 0.5 g/L 5-25C8: (left) observation of Rho-lipid fluorescence in the membranes and (right) observation of FITC-dextran. The diameter of the largest vesicle (upper one) was here 22 μm.
0 10 0 5 min 0 0 15 (0) 10 50 0 30 min 0 0 50 (5)b 100 25 3h 0 0 100 (20)b 50 a a 24 h 5 0 100 (75)b 90 a No vesicle left on the bottom of the observation chamber, or a density of vesicles that was significantly reduced after incubation. b Values in parentheses are measurements at 0.05 g/L polymer.
Table 2. Fraction of Emptied Giant Vesicles (% within a Random Sampling of 50 Vesicles, Unilamellar or Multilamellar) Following Addition of 5-25C8-40C3 up to 0.5 g/L in the External NaCl, 10 mM Bis-Tris-HCl pH 6.8 Buffer
expected for a constant internal concentration of the probe. Concerning the fluorescence of rhodamine in the membrane (see Supporting Information Figure S2 for details on its measurements), no correlation can be found with vesicle diameters (not shown). The histogram of intensities in Figure 1D shows two main peaks in fluorescence centered at 600 and 1200 au. We concluded that the membranes were distributed mostly in a dominant population of unilamellar vesicles (GUVs) and a population of multilamellar vesicles (GMVs) displaying intensities twice as high as those of GUVs. Other multilamellar vesicles with higher intensities of rhodamine fluorescence where also present to a lesser extent (intensities beyond 1200 au in Figure 1D). In what follows, we discriminate between GUVs and GMVs in order to examine whether the polymer affects the permeability of a stack a membranes, that is, translocates through several bilayers. Following the addition of polymer, quoted time zero, 5 min was necessary to let the flow of liquid stop in the droplet (see Materials and Methods section). At time 5 min, the fluorescence of the membranes was most typically the same as that in the absence of polymers. The fluctuations of membrane shape were however frozen upon supplementation with 5-25C8, and all vesicles turned to a spherical shape. In contrast, in samples with 5-25C8-40C3, fluctuations of shapes and transient protrusion of tubes could be observed (prior to addition of polymer, GUVs are not under tension and their shape is also slightly fluctuating). These typical effects of variations of tension and elastic modulus brought on by adsorption of a polyelectrolyte on bilayers have been discussed elsewhere and are beyond the purpose of the present study.20,23-25 Here, we used membrane fluorescence to report on the homogeneity of the lipid dispersion in GUVs and to possibly reveal disruption and large holes in the bilayers. In the present experimental conditions, the membrane homogeneity and apparent continuity were always preserved at the scale of optical resolution. The internal fluorescence of encapsulated FITC-dextran was in contrast markedly affected by the presence of polymer. Following addition of polymer, a fraction of vesicles lost their internal fluorescence and were coexisting with vesicles having kept their initial fluorescence (e.g., one empty and one loaded giant vesicle in Figure 2, right; N.B. a group of smaller emptied vesicles also appears as a white domain in Figure 2, left). About 50 different vesicles were counted in each sample, and the fraction of emptied ones in this population could be determined within an uncertainty of ∼10% (due to fluctuation from drop to drop). As listed in Tables 1 and 2, the fraction of
empty GUVs increases with increasing incubation time and reaches 100% at long time. Superscript footnote “a” in Tables 1 and 2 points to samples where the vesicles vanished after incubation beyond a few hours, which presumably corresponds to the solubilization of most membranes (e.g., in Table 2, 0.5 g/L 5-25C8-40C3 at time 24 h irrespective of NaCl). We cannot totally rule out that the vesicles once they were permeable to glucose would not detach from the bottom of the chamber, but our observations in the bulk of the drop did not reveal a significant amount of vesicle left in these cases. Triggering the capacity of polymers to solubilize lipids is not new, and studies have reported that small or large vesicles (SUV, LUV) can be dissolved by amphiphilic macromolecules provided that the hydrophobicity of the chain is above a critical value.10-12 Parameters such as the density of alkyl side groups, pH, and NaCl concentration affect the polymer charge density and interchain repulsion/attraction and finally control the association to lipid bilayer. In LUVs and SUVs, the tension of membranes facilitates the disruption of vesicles.11,16,26 For that reason, polymer-triggered leakage in the absence of membrane disruption has not often been reported for intact LUVs.18,27,28
(23) Safran, S. A.; Kuhl, T. L.; Israelachvili, J. N. Biophys. J. 2001, 81, 659. (24) Tsapis, N.; Ober, R.; Chaffotte, A.; Warschawski, D. E.; Everett, J.; Kauffman, J.; Kahn, P.; Waks, M.; Urbach, W. Langmuir 2002, 18, 4384. (25) Angelova, M. I.; Hristova, N.; Tsoneva, I. Eur. Biophys. J. 1999, 28, 142.
(26) Asokan, A.; Cho, M. J. Biochim. Biophys. Acta, Biomembr. 2003, 1611, 151. (27) Zignani, M.; Drummond, D. C.; Meyer, O.; Hong, K.; Leroux, J. C. Biochim. Biophys. Acta, Biomembr. 2000, 1463, 383. (28) Thomas, J. L.; Tirrell, D. A. J. Controlled Release 2000, 67, 203.
Langmuir 2009, 25(13), 7506–7513
no NaCl incubation time
GUV
GMV
150 mM NaCl GUV
GMV
1 M NaCl GUV
GMV
b
5 min 50 25 50 (0) 25 20 5 50 10 30 min 100 100 100 (35)b 100 b a a 100 3h 100 100 100 (90) a a a a a 24 h (100)a,b a No vesicle left on the bottom of the observation chamber, or a density of vesicles that was significantly reduced after incubation. b Values in parentheses are measurements at 0.05 g/L polymer.
Table 3. Influence of pH on the Fraction (%) of Empty Giant Vesicles Following Addition up to 0.5 g/L 5-25C8 in the External Buffer (If Not Specified, Then 150 mM NaCl, 10 mM Bis-Tris-HCl) pH 6.8 GUV
GMV
pH 7.5a
pH 7.2 GUV
GMV
GUV
5 min 15 0 0 0 0 30 min 50 10 0 0 0 3h 100 50 25 5 10 24 h 100 90 75 10 50 a In 100 mM NaCl, 50 mM Tris-HCl pH 7.5 buffer.
DOI: 10.1021/la900261s
GMV 0 0 0 10
7509
Article
Vial et al.
Figure 3. Images of FITC-dextran fluorescence in a GUV that was homogeneously fluorescent at time zero (A). Same GUV observed by epifluorescence (B) and phase contrast (C) at time 3 h after addition of 0.5 g/L 5-25C8. The fluorescence in the inner vesicles vanished at time 24 h. The three photographs are displayed at the same scale and diameter of the outer vesicle is 23 μm.
The present observations point in contrast to the absence of visible membrane breakage and absence of rip at the micrometer scale, while Tables 1-3 point to obvious polymer-triggered permeabilization. The release of the water-soluble probe (dextran) from these GUVs indicates that interlipid cohesion in the bilayer was nevertheless significantly weakened and allows for pores of diameters above a few nanometers to be formed. Several parameters modulate the rate of leakage, including polymer concentration (data in parentheses for 0.05 g/L in Tables 1 and 2), the structure of the polymer (5-25C8-40C3 was more efficient than 5-25C8), and NaCl concentration. Varying the pH has also a critical influence, and upon increasing the pH above 7 the leakage markedly slows down (Table 3). The conditions for rapid leakage are accordingly low pH (
