ARTICLE pubs.acs.org/Langmuir
Block Copolymer Vesicle Permeability Measured by Osmotic Swelling and Shrinking Autumn Carlsen,†,‡ Nicolas Glaser,† Jean-Franc-ois Le Meins,*,†,‡ and Sebastien Lecommandoux*,†,‡ † ‡
Universite de Bordeaux, ENSCPB, 16 avenue Pey Berland, 33607 Pessac Cedex, France CNRS, Laboratoire de Chimie des Polymeres Organiques, UMR5629, Pessac, France
bS Supporting Information ABSTRACT: Vesicle response to osmotic shock provides insight into membrane permeability, a highly relevant value for applications ranging from nanoreactor experimentation to drug delivery. The osmotic shock approach has been employed extensively to elucidate the properties of phospholipid vesicles (liposomes) and of varieties of polymer vesicles (polymersomes). This study seeks to compare the membrane response for two varieties of polymersomes, a comb-type siloxane surfactant, poly(dimethylsiloxane)-g-poly(ethylene oxide) (PDMS-g-PEO), and a diblock copolymer, polybutadiene-b-poly(ethylene oxide) (PBut-b-PEO). Despite similar molecular weights and the same hydrophilic block (PEO), the two copolymers possess different hydrophobic blocks (PBut and PDMS) and corresponding glass transition temperatures (-31 and 123 °C, respectively). Dramatic variations in membrane response are observed during exposure to osmotic pressure differences, and values for polymer membrane permeability to water are extracted. We propose an explanation for the observed phenomena based on the respective properties of the PBut-b-PEO and PDMS-g-PEO membranes in terms of cohesion, thickness, and fluidity.
1. INTRODUCTION Closed-membrane structures are known to self-assemble from both phospholipid and amphiphilic copolymer materials through solution-based processes, including simple rehydration and electroformation. The resulting fluid-filled sacs, or vesicles, resemble biologically occurring structures in both size and behavior. In addition to offering an easily accessible, simplified model for cellular behavior studies, such synthetic vesicles also provide a durable, biologically compatible container with predictable release mechanisms (pH, temperature, etc.). These and other advantageous properties have pushed synthetic vesicles to the forefront of drug delivery research. In order for a vesicle to serve as an effective transport vehicle for drug delivery (or as a nanoreactor), its permeability to various substances must be well-understood. To that end, characterization and control of vesicle permeability are significant issues. Water permeation through liposomal membranes has been investigated by techniques that are based either on the measurement of the size parameters of the vesicles (from scattering or microscopy) or on the environment of the solutes probed by spectroscopy. Those include selfquenching of entrapped fluorophore,1,2 anti-Stokes Raman scattering,3 optical microscopy,4-6 and dynamic light scattering.7,8 In most cases, the driving force for permeation is osmotic shock. However, the effect of a few other types of physical stress applied on membrane properties have also been studied, such as the shear in a Couette’s cell9 or the increase of surface tension by adhesion on a substrate. As for polymersome membranes, osmotic shock9-12 and nuclear magnetic resonance13 have been r 2011 American Chemical Society
used to study their permeability to water. Diffusion of specific hydrophilic molecules through the polymersome membrane has also been studied by dynamic light scattering and absorbance measurements in UV, which revealed Fickian transport behavior, with the rate of diffusion being inversely proportional to the membrane thickness.14 Further, fluorescent probes have been employed to study the hydration of the polymer membrane itself15 and to determine permeability to nonaqueous compounds.16 The osmotic shock approach presents a simple method to obtain information on membrane permeability. In studies of giant vesicles (with diameters in the tens of micrometers), optical microscopy is conveniently employed to observe the response of a fluid-filled vesicle to an external solution of differing osmolarity from the vesicle’s interior solution.4 The dynamics of the vesicle’s changing volume is governed by the semi-permeable nature of its membrane. More precisely, only solvent molecules can cross the hydrophobic barrier, but not hydrophilic solutes unless they are small and of low polarity (e.g., ethanol or glycerine). The molar flux density, j, of water through the membrane can be written as j = -PΔc, where Δc is the difference in molar concentrations of all the solutes between the internal and external solutions and P is the permeability of the membrane to the solvent (usually several orders of magnitude larger than to the solutes). The balance of the volume flow rate through the entire membrane of surface area Received: December 21, 2010 Revised: February 18, 2011 Published: March 15, 2011 4884
dx.doi.org/10.1021/la105045m | Langmuir 2011, 27, 4884–4890
Langmuir
ARTICLE
4πR2 with the derivative of volume 4πR2 dR/dt leads to: P ¼ -
1 dR vw Δc dt
polymersomes properties, which may also vary according to the different architectural characteristics of these copolymers. ð1Þ
where vw represents the molar volume of water (roughly equivalent to 18 mL/mol) and R is the radius of the vesicle at a given time t. Osmotic shock studies, both theoretical17,18 and experimental,4-7,10-12,19,20 have been carried out on a number of closed-membrane systems. With regard to measurements performed on giant-sized vesicles, the primary challenge for osmotic shock studies is the need for prolonged observation of a free-floating object within a restricted field of view. This challenge is heightened when the intent is to follow a specific structure before and after addition of a secondary solution. A variety of techniques have been employed to address this difficulty. Some studies5 report formation of the vesicles in a solution containing larger sugar molecules (such as sucrose) followed by dilution into a solution containing smaller molecules (such as glucose), as a means to enhance the image quality obtained by phase contrast microscopy. The refractive index asymmetry of the internal and external solutions increases the viewing contrast, while the negative buoyancy of the vesicles in the less dense external solution causes them to sink to the bottom of the chamber (where observation is optimal with an inverted microscope). Caution is required, however, since gravity’s effects on heavier vesicles (produced from greater sugar concentrations) could alter observed characteristics, such as their sphericity.21 In addition, budding or fluctuation of the membrane may be dampened by interaction with the floor of the observation chamber.22 Another approach employed in the literature6,23 is the use of a fine-pored intermediary membrane between the initial vesicle-containing solution and the secondary sugar solution. The membrane minimizes convection, and therefore vesicle motion, within the viewing chamber. However, the uncertainty of the initial time of contact between vesicles and secondary solution may introduce additional error in the perceived permeability values.6,23A third reported method consists of using a micropipet to grip a vesicle by suction and then transferring it abruptly into a second solution.11 The immediate placement of the vesicle into the secondary solution allows a more straightforward calculation of exposure time. However, damage could occur to the vesicle membrane (at least in the case of liposomes) during aspiration with the micropipet,8 potentially altering perceived permeability values. Further, the aspiration of excess membrane during osmotic shock prevents observation of naturally occurring phenomena, such as budding and membrane fluctuations.6 In this study, use of a syringe pump enabled delivery of the secondary solution into the observation chamber at a known rate. The flow rate and concentrations were optimized to allow observation of the same vesicle before and after osmotic shock. The effects of osmotic shock were investigated for two different types of vesicles, one type composed of the diblock copolymer polybutadiene-b-poly(ethylene oxide) (PBut-b-PEO) and the other type composed of the copolymer surfactant Dow Corning 5329 (PDMS-g-PEO), which can be viewed as a grafted copolymer (see section 2.1). While each copolymer material has comparable molecular weight and the same stabilizing hydrophilic block, the higher glass transition temperature of PBut in comparison to that of PDMS (-31 and -123 °C, respectively) hints at potentially significant differences in the resulting
2. EXPERIMENTAL SECTION 2.1. Materials and Sample Preparation. Dow Corning 5329 is composed of a PDMS chain decorated with two arms of PEO, on average, each measuring about 12 monomers in length. The weight fraction of ethylene oxide is 47% according to the furnisher’s data and has been checked by 1H NMR (Supporting Information). The average viscometric molecular weight is equal to 3000 g/mol according to the literature and furnisher’s data24,25 and has been verified by viscosity measurements (Supporting Information). Previous studies reported a membrane core thickness of 5 nm measured by Cryo-TEM.26 This has been verified in the lab using the same technique (Supporting Information). The PBut46-b-PEO30 (supplied by Polymer Source) is a well-defined block copolymer with a polydispersity index of 1.05 and a numberaverage molecular weight of 3800 g/mol. The weight fraction of ethylene oxide is 35%. A chromatogram as well as NMR spectrum can be founded in the Supporting Information. Related literature reported a 9.0 ( 1 nm membrane core thickness27 for vesicles formed from a similar PBut-bPEO polymer (PBut46-b-PEO26). Polymer vesicles were obtained through a well-known electroformation approach.28 For PBut-b-PEO, 1 mg of the polymer was dissolved per mL of chloroform and then the solution was drip-deposited on a glass slide coated with indium tin oxide (resistivity 15-25 Ω/square). For PDMS-g-PEO, a solution of 10 mg/mL in chloroform was used. The fluorescent dye Nile Red was also sometimes added to this initial chloroform solution of either type of copolymer at a concentration of 0.5 mM, in order to facilitate imaging in later steps. Deposition of the copolymer chloroform solution (with or without dye) onto an ITOcoated glass was immediately followed by spin-coating for 60 s at 300 rpm. A film of copolymer was formed on the ITO-coated slide as the solvent evaporated. The edge of the copolymer film was outlined with a Vitrex paste spacer to create a cavity of approximately 1 mm between the ITO-coated glass slide bearing the copolymer film and a second ITO-coated glass slide. An aqueous solution of given osmolarity was injected into this cavity. Optical microscopy was employed to monitor the growth, size, and morphology of the vesicles. Variations in the absorption of the aqueous solution produce buckling and blistering of the film, followed by formation of vesicles due to energy minimization constraints. In the electroformation approach, a voltage difference is applied between the upper and lower slides to induce the electro-osmotic flow of water. The electro-hydrodynamic instability resulting from the net accumulation of conduction charges onto the film and the periodic motion of water by electro-osmosis is believed to explain29 the production of more homogeneous and uniform vesicles.30 After cessation of voltage application, the system was tipped back and forth. Air bubbles traveled over the film surface, freeing tethered vesicles into solution. Extraction of vesicles was performed using an Eppendorf pipet equipped with a 1 mL tip, sufficiently large to prevent vesicle damage by shear stress. A precise volume of the solution was then deposited with a micropipet into a homemade observation chamber (Figure 1) consisting of a single glass microscope slide separated from a coverslip by a step of two coverslips glued in place (approximate thickness: 350 μm). A narrow tube connected to a syringe pump was inserted at the edge of the small deposited droplet. During each experiment, the syringe pump (model 33, Harvard Apparatus) delivered the solution of differing glucose concentration at a known rate. In a typical experiment, the 3-6 μL of secondary solution necessary to adjust the external osmolarity was delivered to the original 6 μL solution within 1 min. The process was observed with a 40/0.7 numerical aperture objective lens working in bright-field Nomarsky’s contrast (differential 4885
dx.doi.org/10.1021/la105045m |Langmuir 2011, 27, 4884–4890
Langmuir
ARTICLE
Figure 1. Schematic diagram of the setup for osmotic shock experiments. interference contrast or DIC) using an Axiovert 40 inverted microscope equipped with a Canon Powershot A640 digital camera for recording images. Error for vesicle size readings was estimated at (0.5 μm on the basis of the standard variation of the vesicle radius measurements from digital images.
3. RESULTS AND DISCUSSION 3.1. Shrinking. When a closed-membrane structure formed in pure water is then exposed to an external solution containing a significant glucose concentration, the osmotic pressure pushes water out of the structure, thereby producing a visible overall shrinking. Besides a simple shrinking of the vesicle, it is known that excess membrane surface area resulting from volume reduction can produce budding in the interior of the original vesicle. For example, Bernard et al.23 observed that an external hypertonic glucose solution produced raspberry-like clusters inside vesicles composed of natural egg phosphatidylcholine (EPC) with or without cholesterol. This phenomenon was attributed to a nonvanishing spontaneous curvature of the bilayers.4 Menager et al.6 used trisodium citrate to produce the same raspberry effect in vesicles of the purified zwitterionic phospholipid 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC). The reversibility of the phenomena—swelling of the raspberry vesicle into a larger spherical vesicle under hypotonic conditions—was offered as evidence of a neck or connection remaining between the daughter vesicles and the parent membrane. An external hypertonic glucose solution also produced daughter vesicles inside vesicles of the negatively charged phospholipid dioleoylphosphatidylglycerol (DOPG), as reported by Claessens et al.,5 who proposed that the size of the daughter vesicles would be controlled by the mean bending modulus of the bilayer. Both polymer vesicle systems (PDMS-g-PEO and PBut-bPEO) were exposed to positive gradients of external glucose concentration, and observation was made to determine similarities and differences. Under hypertonic conditions, giant unilamellar PDMS-g-PEO vesicles tended to shrink uniformly, maintaining a spherical shape (Figure 2). Response to osmotic pressure difference was rapid, occurring within seconds of exposure time. The rate of vesicle shrinking gradually increased until ∼110 s had passed. After that point, a linear dependence of the size evolution with time was clearly observed, as well as the formation of internal vesicles. Regardless of the osmotic pressure
Figure 2. Evolution of water-filled PDMS-g-PEO vesicle in 250 mM glucose solution. Digital images A-C show DIC microscopy observation of the same vesicle over the course of time; white scale bar represents 20 μm. (D) Graph of the vesicle radius with respect to time; the shaded area signifies a rapid growth period of internal vesicles.
Figure 3. Evolution of water-filled PBut-b-PEO vesicles in glucose solution observed by DIC microscopy. Vesicles A and B were observed in 250 mM glucose, and the white scale bar represents 20 μm. Vesicle C was observed in 100 mM glucose.
difference between the internal and external solutions, the internal vesicles presented a homogeneous size distribution with radii near 3 μm. In comparison to vesicles formed from PDMS-g-PEO, the response of PBut-b-PEO vesicles to increased external glucose concentration was noticeably delayed and more complex. Different behaviors were observed, as illustrated in Figures 3 and 4. In some instances, the vesicle appeared to be indented, producing 4886
dx.doi.org/10.1021/la105045m |Langmuir 2011, 27, 4884–4890
Langmuir
Figure 4. Evolution of water-filled PBut-b-PEO vesicle in 100 mM glucose solution. Digital images A-C show DIC microscopy observation of the same vesicle over the course of time; white scale bar represents 20 μm. (D) Graph of the vesicle radius with respect to time; the shaded area signifies a rapid growth period of internal vesicles.
a discoid shape (Figure 3, vesicle A). In other instances, the overall size of the PBut-b-PEO vesicles remained constant for a longer period of time, although displaying signs of membrane reorganization (such as the faceting visible in vesicle B at 3 min, 20 s, in Figure 3). This period of no apparent vesicle size change was followed by dramatic, random deformation (vesicle B at 15 min, 31 s, in Figure 3) or occasionally, by vesicle bursting and creation of several smaller vesicles (vesicle C in Figure 3). In those few cases where the PBut-b-PEO vesicles did maintain the spherical shape commonly observed for PDMS-g-PEO vesicles during osmotic shock, the rate of vesicle shrinking was not smooth but showed periods of marked increase interspersed with little to no change (Figure 4). Besides summarizing the morphological effects on vesicle appearance, an effort was made to calculate the permeability from the rate of vesicle shrinking in those cases where the vesicle remained quasispherical for both PDMS-g-PEO and PBut-bPEO. In either case, a steep decline in vesicle radius was observed near the onset of internal vesiculation. A similar decline in vesicle radius was noted in a DOPC vesicle system.6 Deviation from linear behavior in terms of a decrease in vesicle radius was also observed at the onset of internal vesiculation in that system. A correction term was reported4,6 to account for the appearance of internal vesicles when determining the permeability P. The adjusted equation takes into account the radius r of the internal vesicles, as follows: dR 2 r 1ð2Þ ¼ - vw ΔcP dt 3R In the case of the shrinking PDMS-g-PEO vesicle, no internal vesiculation was visible for the initial phase when the concentrated glucose solution was mixing with the native vesicle solution. A couple of minutes later during the shrinking process, internal vesicles can be seen forming within the parent PDMS-g-
ARTICLE
PEO vesicle. During the period of internal vesicle formation (and using eq 2 to account for the change in parent vesicle surface area), the rate of radius decrease leads to a water permeability value of 28 μm/s. This experimentally determined permeability for PDMS-g-PEO vesicles is consistent with the range for liposomal membranes, which generally exhibit water permeability of 25-150 μm/s.31 The diffusion rate of small solutes through polymersome membranes has been described as inversely proportional to the thickness of the hydrophobic part of the membrane.14 However, membrane thickness alone cannot account for the experimentally determined value of 28 μm/s for PDMS-g-PEO vesicles. Consider the water permeability value of 0.8 μm/s recently reported for vesicles obtained from the triblock copolymer PMOXA-bPDMS-b-PMOXA [PMOXA being poly(2-methyloxazoline)]. While the hydrophobic block is PDMS, its molecular weight (MW) of ∼8000 g/mol is roughly 4 times greater than the molecular weight of the PDMS portion in the polymer under investigation here, PDMS-g-PEO. Since the thickness of the hydrophobic layer scales as MW0.532 (and since permeability is inversely proportional to that thickness), one would expect the permeability of PDMS-g-PEO to be just twice that of PMOXAb-PDMS-b-PMOXA. This is not the case, as a factor close to 35 is found between the permeability of PMOXA-b-PDMS-bPMOXA and PDMS-g-PEO vesicles. This larger-than-predicted gap suggests that the copolymer architecture must certainly play an important role in polymersome membrane permeability. Since the rate of PBut-b-PEO vesicle shrinking was not smooth, it was not possible to identify an appropriate period for determination of permeability. PBut-b-PEO vesicle size declined significantly more rapidly at the onset of internal vesicle formation, similar to the behavior displayed by PDMS-g-PEO vesicles. It is noteworthy that, while prior phospholipid studies reported an increase in internal vesicle size with increasing mean bending modulus of the bilayer,5,23 no apparent size difference was observed for the internal vesicles produced in PBut-b-PEO and PDMS-g-PEO vesicles under hypertonic conditions, despite the presumed difference in the mean bending moduli of the two bilayer membranes (based on their differing thickness and fluidity.) 3.2. Swelling. While vesicles exposed to a greater external glucose solution show an overall decrease in size, an increase in size typically occurs for vesicles exposed to a lower external glucose solution. Water is pulled into the vesicle to equilibrate the internal and external glucose concentrations, producing an overall expansion in vesicle diameter that generally leads to bursting of the vesicles. To enhance viewing facility, a hydrophobic fluorescent dye (Nile red) was incorporated into the copolymer films of PBut-bPEO and PDMS-g-PEO prior to electroformation. A droplet of the subsequently formed glucose-filled vesicles was then diluted with a water solution, and vesicle response was monitored by fluorescence microscopy. As in the case of shrinking vesicles, a range of behaviors were observed, and we describe those most frequently encountered. PBut-b-PEO vesicles were often observed to swell steadily and then abruptly burst (Figure 5). Protrusions in the outer surface of the parent PBut-b-PEO vesicle were sometimes observed (Figure 6), although an overall spherical shape was generally maintained. Observation of three PButb-PEO vesicles which maintained a spherical shape when exposed to an environment of lower glucose concentration 4887
dx.doi.org/10.1021/la105045m |Langmuir 2011, 27, 4884–4890
Langmuir
ARTICLE
Figure 7. Evolution of 1000 mM glucose filled PDMS-g-PEO vesiclein hypotonic solution with external concentration decreasing stepwise from 1000 mM glucose at 5 min (A) to 667 mM glucose (B) to 600 mM glucose (C) to 460 mM glucose at 18 min (D). Ther white arrow (C) indicates material lost during the bursting/resealing phase. Digital images A-D show selected frames from a video recorded during fluorescence microscopy observation of the same vesicle over the course of time; the white scale bar represents 20 μm.
Figure 5. Evolution of 500 mM glucose filled PBut-b-PEO vesicle in hypotonic solution (external concentration decreasing steadily from 500 mM glucose at 0 s to 205 mM at 405 s, when the vesicle burst). Digital images A-C show fluorescence microscopy observation of the same vesicle over the course of time; white scale bar represents 20 μm. (D) Graph of the vesicle radius with respect to time, with the vertical dashed line marking the time at which the vesicle burst.
Figure 6. Evolution of 500 mM glucose filled PBut-b-PEO vesicle in 250 mM glucose solution observed by DIC microscopy. The white scale bar represents 20 μm.
provided a value for permeability of 3.1 ( 1.6 μm/s, using the standard permeability equation (eq 1). Interestingly, this value for water permeability of the PBut-b-PEO membrane is in good agreement with a value of 2.5 ( 1.2 μm/s, which was previously reported for giant polymersomes made from the self-assembly of poly(ethylene oxide)-b-poly(ethylethylene) PEE-b-PEO with similar molecular weight (3900 g/mol) during osmotic shock with hypertonic conditions.11 It should be noted that permeability measurements in hypotonic conditions (swelling) are not ideal because the overpressure created inside the vesicle imposes a surface tension on the membrane that could lead to bursting of the vesicle, should this surface tension exceed the membrane’s lysis tension. In our osmotic pressure conditions, the surface tension σ estimated via Laplace’s law (Δp = 2σ/R) is around 6 N/m, far above the reported lysis tension of polymersomes.11,33 However, the bursting of the polymersomes does not occur immediately. In the example illustrated in Figure 5, the radius of the vesicle increases from ∼20 to 23.5 μm in 300s, corresponding to an areal strain at rupture (ΔA/A0) of 0.38, in agreement with the critical areal strain values reported for polymersomes.11 It seems reasonable to think that the membrane does not undergo
Figure 8. Evolution of 500 mM sucrose filled PDMS-g-PEO vesicle in 250 mM glucose solution after 0 min (A) and 10 min (B) observed by optical microscopy. Image C was taken of the same vesicle after an additional 11 min in 167 mM glucose solution. The white scale bar represents 20 μm.
the sudden onset of a 6 N/m surface tension but that instead the swelling induced by the entry of water into the vesicle progressively increases the surface tension until the point of lysis. Pores are then generated, leading to vesicle bursting. Again, although the molecular weights of the PDMS-g-PEO and PBut-b-PEO used in this study are rather similar, the measured permeability values differ by 1 order of magnitude (28 and 3.1 μm/s, respectively), showing the strong influence of the copolymers’ architecture and the nature of the hydrophobic block on the membrane’s permeability. A qualitative determination of the permeability for PDMS-gPEO vesicles in the hypotonic regime is less direct. As observed for certain liposomes,7,34 a swell-burst cycle sometimes occurs upon introduction of PDMS-g-PEO vesicles into a solution of reduced osmolarity. Vesicles first swell as they absorb water to match the external concentration and then burst, reseal, and repeat the process (Figure 7). Material may be lost during the bursting/resealing phase of the cycle (Figure 7C), resulting in a reduced vesicle diameter prior to the subsequent swelling phase. Due to the high speed at which the process occurs, it is difficult to pinpoint every burst event and therefore to determine a permeability value with certainty. In other instances, no change in PDMS-g-PEO vesicle size was observed, despite large glucose gradients (Figure 8). This may be attributed to a swell-burst cycle too fast and subtle to be recorded with a simple camera. It could also be due to defects in the PDMS-g-PEO membrane allowing free exchange of the internal and external solution.
4. DISCUSSION AND CONCLUSION The objective of this study was to compare the osmotic shock response of two specific polymer membranes made of PDMS-g4888
dx.doi.org/10.1021/la105045m |Langmuir 2011, 27, 4884–4890
Langmuir PEO and PBut-b-PEO using an experimental method limiting the uncertainty of the initial time of contact between vesicles and secondary solution and preserving the vesicle unstressed (e.g., no mechanical fixation of the vesicle). PDMS-g-PEO vesicles form swiftly from a dried film in aqueous solution at room temperature. Such ease in vesicle formation is associated with a low bending elasticity, meaning large thermal undulations of the bilayers.35 The PDMS-g-PEO membrane permeability is found to be in the range of 28 μm/s under hypertonic conditions, similar to values reported for phospholipid vesicles. An internal vesiculation is also frequently observed as in liposomes.23 Also like phospholipid vesicles, PDMS-g-PEO vesicles respond to hypotonic conditions with a swell-burst cycle, with the vesicles swelling to accommodate additional water and then bursting and releasing part of their content to relieve the surface tension.36 In the hypertonic case, PBut-b-PEO vesicles sometimes shrink and internal vesicles can occasionally be observed. Since the shrinking process does not occur at a steady rate, the accurate determination of water permeability on the basis of the hypertonic experiments alone is not possible. More often than simply shrinking, the PBut-b-PEO vesicles show signs of reorganization at the membrane surface and then deform dramatically, as was theoretically predicted by the area difference elasticity model (ADE) and confirmed with results obtained in a systematic experimental study on phospholipids.37 However, the shapes obtained in the present study do not correspond to any shape predicted by the ADE model, which is restricted to axisymmetric shapes. Hypotonic conditions produced simpler PBut-b-PEO vesicle behavior. Although vesicles occasionally bursted or developed protrusions, a permeability value of 3.1 ( 1.6 μm/s was obtained when vesicles simply swelled. This value, 1 order of magnitude lower than for lipidic vesicles or PDMS-g-PEO vesicles, is in good agreement with a previously reported permeability value for block PEE-b-PEO copolymer vesicles with similar molecular weight (3900 g/mol), differing only by the hydrophobic block (PEE being hydrogenated PBut).11 This is the first example to our knowledge in which the permeability of polymer vesicles has been measured in hypotonic conditions. The differences observed between PBut-b-PEO and PDMS-gPEO in their permeabilities, 28 and 3 μm/s, respectively, and in their overall behavior under hypertonic and hypotonic conditions cannot be explained by their differing membrane thicknesses alone (PDMS-g-PEO ∼5 nm24 and PBut-b-PEO ∼9 nm27). Dissimilarities in the copolymers’ membrane cohesion and fluidity must also be taken into account. The stiffer PBut-bPEO vesicles were much more likely to burst—breaking apart completely—than their more flexible PDMS-g-PEO counterparts, which would burst then reseal and swell again. Such breakage of polymersomes during osmotic shock has previously been observed on poly(ethylene glycol)-b-poly(lactic acid) polymersomes and was attributed to the semicrystalline character of poly(lactic acid).38 Since PDMS-g-PEO films form vesicles more easily than PBut-b-PEO (which requires additional heating and/ or time to produce the same size and quantity of vesicles), it is reasonable to expect that a PDMS-g-PEO vesicle in a damaged or nonequilibrium state might also re-form a closed membrane more quickly. Indeed, in hypotonic conditions PDMS-g-PEO vesicles appear much more compliant. They generally maintain a spherical shape, as would be expected for the equilibrium state of inflated vesicles that can release their surface tension rapidly by
ARTICLE
solvent permeation and membrane adjustment (e.g., the formation of transient pores in the PDMS-g-PEO bilayer due to a lower line tension than that of the stiffer PBut-b-PEO bilayer). In hypotonic conditions, PBut-b-PEO vesicles also exhibit a slower rate of osmotic shock response, maintaining a nonequilibrium state for a longer period of time, which might be ascribed to their larger stiffness and higher resistance before rupture. Finally, the comparison of the permeability values obtained in this study for PBut-b-PEO and PDMS-g-PEO with values reported in literature for similar block copolymers (PEE-b-PEO and PMOXA-bPDMS-b-PMOXA) shows that, in addition to the well-known effects of membrane thickness, the architecture of the copolymer constituting the membrane plays a non-negligible role in the permeability. Osmotic shock response proved to be a facile and effective method to obtain quantitative measurements of the membrane permeability to water molecules of PBut-b-PEO and PDMS-g-PEO bilayers through observation of the kinetics of radius change in swelling and shrinking experiments. While the permeability values for one polymer were obtained only from swelling and the other value obtained only from shrinking, we believe the order of magnitude difference supports the conclusion that molecular weight alone is not sufficient for explaining the permeability differences. Behavioral differences in membrane response also provided insight into the effects of hydrophobic core thickness and fluidity of the hydrophobic block. The investigation of the osmotic shock behavior of these large polymersomes (diameter >10 μm) provides a platform for the analysis of their nanoscale counterparts (diameter ∼100 nm) by light and small angle neutron scattering that will be described in a future publication. This fundamental understanding is of major relevance for the use of polymersomes in biomedical devices, where they have a promising future.39
’ ASSOCIATED CONTENT
bS
Supporting Information. Full characterization data of the two copolymers under investigation (NMR, GPC, CryoTEM) and movies showing the internal vesiculation of PDMS-gPEO and the random deformation of PBut-b-PEO vesicles. This material is available free of charge via the Internet at http://pubs. acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected] (J.-F.L.M.) and
[email protected] (S.L.).
’ ACKNOWLEDGMENT The authors are grateful for funding from the Agence Nationale de la Recherche (ANR, Projet -07-NANO-061 MONOPOLY) and to Olivier Sandre for fruitful comments on the manuscript. ’ REFERENCES (1) Chen, P. Y.; Pearce, D.; Verkman, A. S. Biochemistry 1988, 27 (15), 5713–8. (2) Mathai, J. C.; Sprott, G. D.; Zeidel, M. L. J. Biol. Chem. 2001, 276 (29), 27266–71. (3) Potma, E.; de Boeij, W. P.; van Haastert, P. J.; Wiersma, D. A. Proc. Natl. Acad. Sci. U. S. A. 2001, 98 (4), 1577–82. 4889
dx.doi.org/10.1021/la105045m |Langmuir 2011, 27, 4884–4890
Langmuir (4) Boroske, E.; Elwenspoek, M.; Helfrich, W. Biophys. J. 1981, 34 (1), 95–109. (5) Claessens, M. M.; Leermakers, F. A.; Hoekstra, F. A.; Stuart, M. A. Biochim. Biophys. Acta 2008, 1778 (4), 890–5. (6) Menager, C.; Cabuil, V. J. Phys. Chem. B 2002, 106 (32), 7913–7918. (7) Li, W.; Aurora, S. T.; Haines, H. T.; Cummins, H. Z. Biochemistry 1986, 25, 8220–8229. (8) Miyamoto, S.; Maeda, T.; Fujime, S. Biophys. J. 1988, 53 (4), 505–12. (9) Bernard, A. L.; Guedeau-Boudeville, M. A.; Marchi-Artzner, V.; Gulik-Krzywicki, T.; di Meglio, J. M.; Jullien, L. J. Colloid Interface Sci. 2005, 287 (1), 298–306. (10) Discher, B. M.; Bermudez, H.; Hammer, D. A.; Discher, D. E.; Won, Y. Y.; Bates, F. S. J. Phys. Chem. B 2002, 106 (11), 2848–2854. (11) Discher, B. M.; Won, Y. Y.; Ege, D. S.; Lee, J. C. M.; Bates, F. S.; Discher, D. E.; Hammer, D. A. Science 1999, 284 (5417), 1143–1146. (12) Lorenceau, E.; Utada, A. S.; Link, D. R.; Cristobal, G.; Joanicot, M.; Weitz, D. A. Langmuir 2005, 21 (20), 9183–9186. (13) Bauer, A.; Kopschutz, C.; Stolzenburg, M.; Forster, S.; Mayer, C. J. Membr. Sci. 2006, 284 (1-2), 1–4. (14) Battaglia, G.; Ryan, A. J.; Tomas, S. Langmuir 2006, 22 (11), 4910–4913. (15) Lee, J. C. M.; Law, R. J.; Discher, D. E. Langmuir 2001, 17 (12), 3592–3597. (16) Kim, K. T.; Cornelissen, J. J. L. M.; Nolte, R. J. M.; Van Hest, J. C. M. Adv. Mater. 2009, 21 (27), 2787–2791. (17) Blok, M. C.; van der Neut-Kok, E. C.; van Deenen, L. L.; de Gier, J. Biochim. Biophys. Acta 1975, 406 (2), 187–96. (18) Dammann, B.; Fogedby, H. C.; Ipsen, J. H.; Jeppesen, C. J. Phys. I 1994, 4 (8), 1139–1149. (19) Mui, B. L.; Cullis, P. R.; Evans, E. A.; Madden, T. D. Biophys. J. 1993, 64 (2), 443–53. (20) Shoemaker, S. D.; Vanderlick, T. K. Ind. Eng. Chem. Res. 2002, 41 (3), 324–329. (21) Kraus, M.; Seifert, U.; Lipowsky, R. Europhys. Lett. 1995, 32 (5), 431–436. (22) Dimova, R.; Aranda, S.; Bezlyepkina, N.; Nikolov, V.; Riske, K. A.; Lipowsky, R. J. Phys.: Condens. Matter 2006, 18 (28), S1151– S1176. (23) Bernard, A. L.; Guedeau-Boudeville, M. A.; Jullien, L.; di Meglio, J. M. Biochim. Biophys. Acta 2002, 1567 (1-2), 1–5. (24) Hill, R. M. Langmuir 1993, 9, 2789–2798. (25) Nam, J.; Santore, M. M. Langmuir 2007, 23, 7216–7224. (26) Lin, Z.; Hill, R. M.; Davis, H. T.; Scriven, L. E.; Talmon, Y. Langmuir 1994, 10 (4), 1008–1011. (27) Lee James, C. M.; Bermudez, H.; Discher, B. M.; Sheehan, M. A.; Won, Y. Y.; Bates, F. S.; Discher, D. E. Biotechnol. Bioeng. 2001, 73 (2), 135–145. (28) Angelova, M. I.; Dimitrov, D. S. Faraday Discuss. Chem. Soc. 1986, 81, 303–311. (29) Sens, P.; Isambert, H. Phys. Rev. Lett. 2002, 88 (12), 1281021–1281024. (30) Angelova, M. I.; Dimitrov, D. S. Mol. Cryst. Liq. Cryst. Incorporating Nonlinear Opt. 1987, 152, 89–104. (31) Derek, M. Handbook of Lipid Bilayers; CRC: Boca Raton, FL, 2010. (32) Srinivas, G.; Discher, D. E.; Klein, M. L. Nat. Mater. 2004, 3 (9), 638–644. (33) Bermudez, H.; Brannan, A. K.; Hammer, D. A.; Bates, F. S.; Discher, D. E. Macromolecules 2002, 35 (21), 8203–8208. (34) Peterlin, P.; Arrigler, V. Colloids Surf. B 2008, 64 (1), 77–87. (35) Antonietti, M.; F€orster, S. Adv. Mater. 2003, 15 (16), 1323–1333. (36) Sandre, O.; Moreaux, L.; Brochard-Wyart, F. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 10591–10596. (37) D€obereiner, H. G.; Evans, E.; Kraus, M.; Seifert, U.; Wortis, M. Phys. Rev. E 1997, 55 (4), 4458–4474.
ARTICLE
(38) Shum, H. C.; Kim, J. W.; Weitz, D. A. J. Am. Chem. Soc. 2008, 130 (29), 9543–9. (39) Upadhyay, K. K.; Agrawal, H. G.; Upadhyay, C.; Schatz, C.; Le Meins, J. F.; Misra, A.; Lecommandoux, S. Crit. Rev. Ther. Drug Carrier Syst. 2009, 26 (2), 157–205.
4890
dx.doi.org/10.1021/la105045m |Langmuir 2011, 27, 4884–4890
