- Email: [email protected]
Extrusion of electroformed giant unilamellar vesicles through track-etched membranes
Chemistry and Physics of Lipids 165 (2012) 475–481
Contents lists available at SciVerse ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
Extrusion of electroformed giant unilamellar vesicles through track-etched membranes Yogita P. Patil a , Mrunmayi D. Kumbhalkar b , Sameer Jadhav a,∗ a b
Department of Chemical Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India Department of Chemical Engineering, Birla Institute of Technology and Science, Vidya Vihar Campus, Pilani, Rajasthan 333 031, India
a r t i c l e
i n f o
Article history: Available online 3 December 2011 Keywords: Phospholipid Bilayer Lamellarity Lysis tension Encapsulation efficiency Dynamic light scattering
a b s t r a c t Unilamellar vesicle populations having a narrow size distribution and mean radius below 100 nm are preferred for drug delivery applications. In the present work, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) was used to prepare giant unilamellar vesicles (GUVs) by electroformation and multilamellar vesicles (MLVs) by thin film hydration. Our experiments show that in contrast to MLVs, a single-pass extrusion of GUVs through track-etched polycarbonate membranes at moderate pressure differences is sufficient to produce small liposomes having low polydispersity index. Moreover, we observe that the drug encapsulating potential of extruded liposomes obtained from GUVs is significantly higher compared to liposomes prepared by extrusion of MLVs. Furthermore, our experiments carried out for varying membrane pore diameters and extrusion pressures suggest that the size of extruded liposomes is a function of the velocity of GUV suspensions in the membrane pore. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Liposomes are spherical capsules that comprise of one or more phospholipid bilayers. Liposomes may be unilamellar vesicles with a single bilayer encapsulating an aqueous core or multi-lamellar vesicles (MLVs) with several concentric bilayer shells. Based on size, unilamellar vesicles are classified as small unilamellar vesicles (0.025–0.05 m, SUVs), large unilamellar vesicles (0.05–0.5 m, LUVs), and giant unilamellar vesicles (greater than 1 m, GUVs) (Vemuri and Rhodes, 1995; Vuillemard, 1991). To increase the efficacy of therapeutic agents, liposomes are increasingly used as immune evasive carriers for targeted delivery and controlled release of drugs (Lasic, 1993; Storm et al., 1995). Since drug-loaded MLVs are known to exhibit erratic pharmacokinetics and undergo rapid clearance by the reticulo-endothelial system (Hunt et al., 1979), mono-disperse unilamellar vesicles having mean diameter less than 200 nm are preferred for drug delivery applications (Harashima et al., 1994; Woodle, 1995). It is well established that processes including thin film hydration, reverse phase evaporation and ethanol injection lead to the formation of MLVs that exhibit heterogeneity in lamellarity and size (Bangham et al., 1965; Deamer and Bangham, 1976; Szoka and Papahadjopoulos, 1978). A severe limitation of MLVs is their poor ability to encapsulate hydrophilic drugs when compared to unilamellar vesicles of equivalent lipid content. Even though
∗ Corresponding author. Tel.: +91 22 2576 7219; fax: +91 22 25726895. E-mail address: [email protected] (S. Jadhav). 0009-3084/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.chemphyslip.2011.11.013
drug encapsulation efficiency of MLVs was increased using methods such as freeze-drying and hydrating lipids in presence of organic solvents, additional steps such as detergent removal were introduced (Gruner et al., 1985; Ohsawa et al., 1984). Previous studies have shown that subjecting MLV suspensions to sonication, freeze–thaw cycles or multiple membrane extrusions produces vesicles with reduced size and lamellarity (Hope et al., 1985, 1986; Johnson et al., 1971). Extrusion of MLV suspensions through well-defined cylindrical pores of track-etched polycarbonate membranes has been the focus of several studies (Frisken et al., 2000; Hunter and Frisken, 1998; Mayer et al., 1986; Patty and Frisken, 2003; Popa et al., 2009). These studies reveal that many extrusion cycles at high trans-membrane pressures are required to obtain smaller liposomes with narrower size distributions. For example, more than ten extrusion cycles at pressures greater than 300 psig (∼2 MPa) are typically required to realize the minimum achievable liposome size by extrusion of MLV suspension with a membrane having mean pore radius of 50 nm (Mayer et al., 1986; Patty and Frisken, 2003). However, even after subjecting MLV suspensions to several extrusion cycles through membranes having nominal pore radii of 100 and 200 nm, it was observed that ∼30% of extruded vesicles remained oligo-lamellar (Mayer et al., 1986). Moreover, it was reported that subjecting MLVs to multiple extrusions led to further loss of encapsulated drug as well as lipid (Jousma et al., 1987; Perkins et al., 1993). While attempts have been made towards theoretical prediction of extruded liposome size, the exact mechanism by which MLVs are converted to smaller liposomes remains unknown. Clerc and Thompson (1994) have proposed a model based on flow rate
476
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
dependent Rayleigh instability, where MLVs are squeezed through membrane pores as long tubular structures that become unstable and break into smaller cylinders either inside or at the exit of the pore. The maximum size of the extruded liposomes was predicted to be 0.5 times the membrane pore radius, though the relation between fluid velocity and mean liposome size was not worked out. In contrast, Patty and Frisken (2003) proposed a simple bubble model where liposome formation occurs via MLV lysis at the entrance of the pore and the resulting liposome size depends on the applied pressure difference and not on fluid velocity within the pore. One of the limitations of the aforementioned study is that membrane lysis tension and extruded liposome size are estimated from data obtained from different extrusion cycles. Moreover, subsequent theoretical and experimental studies have shown that the membrane lysis tension at a given temperature is not a singlevalued material property, but is a function of the rate at which the membrane is mechanically stressed (Boucher et al., 2007; Evans et al., 2003). In contrast to MLVs, GUVs have been shown to enclose larger aqueous volumes for encapsulation of hydrophilic drugs (Kim and Martin, 1981). Therefore, in the present work we propose that, compared to MLVs, extrusion of GUVs requires lower trans-membrane pressures and fewer extrusion cycles to obtain unilamellar liposomes with larger drug encapsulating capacity. To this end, we carried out single-pass extrusion of electroformed GUVs and compared the mean size and fluorophore encapsulating efficiency of the resulting liposomes to those obtained after multiple extrusion cycles of MLVs. Moreover, we systematically investigated the effect of applied pressure difference on the mean size and polydispersity of liposomes formed by extrusion of GUV suspension through track-etched polycarbonate membranes having nominal pore radii of 25, 50 and 100 nm. Furthermore, we propose a simple functional dependence of extruded liposome size on the average velocity of the vesicle suspension inside the membrane pore. 2. Materials and methods 2.1. Materials The phospholipids 1,2-dipalmitoyl-sn-glycero-3phosphocholine (DPPC) and 1-palmitoyl-2-oleoyl-sn-glycero3-phosphocholine (POPC) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). Triton X-100 was procured from Merck (Darmstadt, Germany) while the fluorophore 5(6) carboxyfluorescein was from Fluka (Buchs, Switzerland). Indium tin oxide (ITO) coated glass slides (25 mm × 75 mm × 1.1 mm) were obtained from Delta Technologies (Stillwater, MN, USA). Nuclepore track-etched polycarbonate membranes (25 mm diameter and 0.006 mm thickness) with nominal pore radii of 25, 50 and 100 nm as well as polyester drain disc supports were purchased from Whatman Inc (Clifton, NJ, USA) while dialysis membrane (14.3 mm diameter, 12–15 kDa MW cut off, capacity 1.61 ml/cm) was from Himedia (Mumbai, India). Ethanol, methanol and chloroform were procured from SD Fine Chemicals (Mumbai, India). 2.2. Preparation of GUV and MLV suspensions GUVs were prepared by the electroformation method using a previously published protocol (Estes and Mayer, 2005). Briefly, 50 l of 2 mg/ml DPPC lipid solution in chloroform was spin-coated at 1000 rpm for 1 min onto ITO slides mounted on a spin coater (Photoresist Spinner PRS-6K, Ducom Instruments, Bangalore, India) and the residual chloroform was removed by overnight application of vacuum. Lipid coated faces of two ITO slides were separated by a silicone rubber spacer and held together with clips forming a
35 mm × 15 mm × 2 mm chamber. Next, 1 ml deionized water was injected with a syringe through the silicone rubber spacer into the chamber and the set up was kept in an oven maintained at 55 ◦ C. An AC voltage (10 Hz, 3 Vpp, sinusoidal) was applied for 2 h across the ITO slides using a function generator (Model 33220A, Agilent Technologies, CA, USA). Finally, a square wave was applied for 1 min to detach the vesicles from the ITO surface. The resulting GUV suspension was immediately used for extrusion studies. For preparation of POPC GUVs, electroformation was carried out at 25 ◦ C. DPPC MLVs were prepared by the well-established thin film hydration method (Elzainy et al., 2005; Mayer et al., 1986). Briefly, 0.3 ml of 2 mg/ml DPPC lipid solution in chloroform was evaporated in a round-bottom flask of a rotary evaporator (Roteva Equitron, Medica Instruments, Mumbai, India) at a speed of 100 rpm for 10 min and maintained at 55 ◦ C. The thin lipid film in the round-bottom flask was thoroughly dried in a vacuum desiccator overnight to remove residual chloroform and subsequently hydrated using 3 ml of deionized water in the rotary evaporator at 100 rpm for 60 min maintained at 55 ◦ C. The resulting MLV suspension was taken through five freeze–thaw cycles which involved freezing the suspension in liquid nitrogen and subsequently thawing at 55 ◦ C in a water bath. The MLV suspension was immediately used for extrusion studies. 2.3. Extrusion of GUVs and MLVs An extruder was fabricated in-house that comprised of a cylindrical stainless steel chamber with a holder for the polycarbonate membrane and polyester membrane support. The chamber was water-jacketed to carry out extrusions at defined temperatures and pressurized using a pressure regulated nitrogen gas cylinder. The DPPC GUV suspensions were subjected to single-pass extrusion through track-etched polycarbonate membranes with nominal pore radii of 25, 50 and 100 nm at 55 ◦ C and pressures ranging from 1 to 400 psi (0.007–2.75 MPa). The DPPC MLV suspensions were subjected to double-pass extrusion through membranes with nominal pore radius 50 nm at 55 ◦ C and pressures ranging from 5 to 400 psi (0.035–2.75 MPa). The flow rate of the resulting liposome suspension from the extruder was measured using a measuring cylinder and stop watch. The membrane permeability for GUV suspensions (as well as for water and MLV suspensions) was calculated as the slope of the linear fit to the permeate flux vs. extrusion pressure plot. For the phospholipid POPC, GUV suspensions were extruded at 25 ◦ C for pressures ranging from 1 to 100 psi (0.007–0.7 MPa) through polycarbonate membranes having nominal pore radius of 100 nm. For experiments related to estimation of encapsulation efficiency, single-pass extrusion of fluorophoreloaded GUV suspensions was carried out while MLV suspensions were subjected to eleven cycles of extrusion through polycarbonate membranes having nominal pore radius of 100 nm at 20 psi (0.14 MPa) and 55 ◦ C. 2.4. Estimation of mean radius and polydispersity of extruded liposomes The mean radius and polydispersity of the extruded liposomes were measured by Dynamic Light Scattering (DLS, Zetasizer Nano ZS, Malvern Instruments, Worcestershire, UK) at 25 ◦ C. The vesicle suspension of lipid concentration 0.2 mg/ml was placed in a glass cuvette. The vesicle suspension was then illuminated by He–Ne red laser at 633 nm, the scattered light intensity was measured by the detector positioned at 173◦ . Using the digital correlator and software provided with the instrument, the temporal auto-correlation function of scattered light intensity was calculated. The method of cumulants was used to obtain the average decay rate and polydispersity index from the correlogram (Koppel, 1972). The intensity
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
477
weighted diffusion coefficient was calculated from the average decay rate and the Stokes–Einstein equation was used to obtain mean radius of liposome from the diffusion coefficient (Selser et al., 1976). Three measurements were made per sample and the experiments were repeated at least three times. The mean radius of extruded liposomes was also reported as a function of the average velocity of suspensions in the membrane pore. The average velocity of the GUV suspension in the membrane pores was estimated from the product of average velocity of water (calculated using Hagen–Poisuelle equation) at a given pressure difference and the ratio of permeability of the suspension to that of water. 2.5. Estimation of fluorophore encapsulating efficiency of liposomes The capacity to encapsulate a hydrophilic fluorophore by GUVs, MLVs and the extruded liposomes was estimated using a previously published protocol (Zhang et al., 1997). Briefly, hydration step in the preparation of GUVs was carried out using a 12 mM aqueous solution of 5(6) carboxyfluorescein (CF). One ml of the GUV suspension was introduced into a dialysis bag, placed in a beaker containing 200 ml water that was magnetically stirred at 25 ◦ C for 36 h to remove unencapsulated CF. The resulting CF encapsulated GUV suspension (10 l) from the dialysis bag was diluted with water to a final volume of 1000 ml. Subsequently, Triton X-100 (10%, v/v) was added to break the liposomes and fluorescence emission spectrum of the released CF was recorded from 490 to 650 nm using a fluorescence spectrophotometer (LS 55, Perkin-Elmer, MA, USA). The fluorescence of an aqueous solution of CF (10 l of 12 mM CF dissolved in water to a final volume of 1000 ml) was also recorded. Based on total CF present in the hydrating solution, the percentage of CF detected in the GUVs was reported as encapsulation efficiency. Efficiency of CF encapsulation by MLVs was estimated using a procedure similar to that employed for GUVs. CF loaded GUVs and MLVs were extruded through polycarbonate membrane of nominal pore radius 100 nm at 20 psi (0.14 MPa) and the encapsulation efficiency of resulting liposomes was also estimated. 3. Results In the present study we carried out a comparison of DPPC liposome formation by membrane extrusion of (a) MLV suspension, obtained by thin film hydration, and (b) electroformed GUV suspension. To this end, we systematically characterized the effect of extrusion pressure on the size and polydispersity of liposomes formed by the two methods. Moreover, we also quantified the capacity of GUVs and MLVs to encapsulate the fluorescent marker CF, as well as fluorophore loss due to extrusion of the vesicular suspensions. Furthermore, we examined the effect of nominal pore radius of the membranes on the size of liposomes formed by extrusion of GUV suspensions as a function of the applied pressure difference. Finally, we propose a simple model relating the mean radius of extruded liposomes to the average velocity of the vesicular suspension in the membrane pore. 3.1. Effect of extrusion pressure on GUV and MLV suspensions As a first step we compared the permeate flux during extrusion of GUV and MLV suspensions as a function of extrusion pressure. To this end, extrusions of GUVs and MLVs of DPPC were carried out through track-etched polycarbonate membranes having nominal pore radius of 50 nm at 55 ◦ C for extrusion pressures ranging from 0.007 to 0.7 MPa. A linear increase in permeate flux with respect to extrusion pressure was observed for both, GUV and MLV suspensions (Fig. 1). However, for a given pressure, the flow of GUV
Fig. 1. Effect of extrusion pressure on permeate flux for GUV and MLV suspensions. MLV suspensions were prepared by thin film hydration method while GUV suspensions were obtained by electroformation method. The permeate flux for DPPC GUV and MLV suspensions as well as for water were measured at 55 ◦ C as a function of extrusion pressure during first-pass through polycarbonate track-etched membrane with nominal pore radius of 50 nm. Data represented as mean ± SEM for n ≥ 3 independent experiments.
suspensions through the membrane was significantly faster compared to that of MLV suspensions. Moreover, when compared to water, permeate flux values observed for GUV suspensions were substantially lower to those of water at similar pressures (Fig. 1). Furthermore, the membrane permeability for water and vesicle suspensions was derived from the relation between permeate flux and extrusion pressure. The respective permeability values for water, GUV and MLV suspensions were 1.76 × 10−8 , 4.41 × 10−9 and 2.65 × 10−9 m/(s Pa). Subsequent experiments were aimed at comparing the size distribution of GUVs and MLVs after single-pass extrusion through a membrane with nominal pore radius of 50 nm as a function of extrusion pressure. We observed that a single pass of MLV suspensions through the membrane at low extrusion pressures resulted in large liposomes that were highly polydisperse for the mean size or size distribution to be reliably measured using DLS (data not shown). Even at higher pressure differences, MLV suspensions were required to be extruded at least twice to obtain good fits to the scattering data for estimation of mean radius and polydispersity index. For GUV suspensions subjected to single-pass extrusion, the mean radius of liposomes asymptotically decreased with increasing extrusion pressure to ∼70 nm at 100 psi (0.7 MPa), after which no further reduction in size was observed (Fig. 2A). In comparison, double-pass extrusion of MLV suspensions required substantially higher pressure differences to generate liposomes of mean radii similar to those generated from single-pass GUV extrusion. In fact, MLV suspensions subjected to an extrusion pressure of 400 psi (2.75 MPa) resulted in liposomes having a mean radius of 85 nm (Fig. 2A). For both, GUV and MLV suspensions, the extruded liposome populations became progressively monodisperse with increase in the applied pressure difference across the membrane. However, similar to observations related to mean radius, considerably larger extrusion pressures were required for MLV suspensions to generate extruded liposome populations having polydispersity index comparable to GUV derived liposomes (Fig. 2B). Taken together, our results demonstrate that relatively small and monodisperse population of liposomes are obtained by single-pass extrusion of GUVs at substantially lower
478
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
Fig. 3. Encapsulation efficiency of MLV and GUV suspensions before and after extrusion. Aqueous solution of 5(6) carboxyfluorescein (12 mM) was encapsulated during thin film hydration of MLVs and electroformation of GUVs at 55 ◦ C. GUV suspensions were subjected to single-pass extrusion while MLV suspensions were subjected to eleven extrusion cycles at 0.14 MPa through membrane with nominal pore radius of 100 nm at 55 ◦ C. The amount of encapsulated CF was estimated after dialysis of extra vesicular fluorophore. Data represented as mean ± SEM for n ≥ 3 experiments. NS: not significant with respect to MLVs before extrusions, *p < 0.05 with respect to GUVs before extrusion.
adequate for obtaining small liposomes with a narrow size distribution. The GUV derived lipsomes were found to enclose 45% of CF present in the hydrating solution used to prepare GUVs (Fig. 3), indicating that liposomes obtained from single-pass extrusion of GUVs exhibit almost a 4-fold larger drug-holding capacity compared to those prepared by the extrusion of MLV suspensions. 3.3. Effect of membrane pore size on extruded liposomes
Fig. 2. Effect of extrusion pressure on mean radius and polydispersity index of liposomes obtained by extrusion of GUV and MLV suspensions. GUV suspensions were subjected to single-pass extrusion while MLV suspensions were subjected to double pass extrusion through membrane with nominal pore radius of 50 nm at 55 ◦ C. The mean radius (A), and polydispersity index (B), of the extruded liposome populations are reported as a function of extrusion pressure. Data represented as mean ± SEM for n ≥ 3 experiments.
trans-membrane pressures when compared to liposomes derived from MLVs. 3.2. Encapsulation efficiency of GUVs, MLVs and extruded liposomes To compare the drug encapsulating potential of GUVs to that of MLVs, we quantified the fraction of the fluorophore 5(6) carboxyfluorescein (CF) present in the hydrating solution that was enclosed into vesicles. Only ∼15% of CF was encapsulated by MLVs, whereas a 4-fold higher percentage of CF was enclosed in GUVs (Fig. 3). To assess whether membrane extrusion resulted in CF loss from the liposomes, CF loaded MLVs suspension was subjected to eleven cycles of extrusion through membrane with nominal pore size of 100 nm at 20 psi (0.14 MPa). The resulting lipsomes were found to contain 12% of the CF present in the hydrating solution used for MLV preparation (Fig. 3). The CF loaded GUVs suspension was subjected to single-pass extrusion through membrane with nominal pore size of 100 nm at 20 psi (0.14 MPa) since this was found
Ensuing experiments were aimed at investigating the effect of average pore radius of the membrane on the size and polydispersity of liposomes obtained from single-pass extrusion of GUVs as a function of the applied pressure difference. To this end, applied pressure difference was varied from 0.007 to 2.75 MPa during extrusion of suspensions of DPPC GUVs at 55 ◦ C through polycarbonate membranes having nominal pore radii of 25, 50 and 100 nm. The permeate rate for pure water as well as for GUV suspensions was found to increase linearly with the trans-membrane pressure difference in the low pressure regime (Fig. 4). For membranes having nominal pore size of 25, 50 and 100 nm, respective permeability values of water were found to be 4.41 × 10−9 , 1.76 × 10−8 and 4.41 × 10−8 m/(s Pa), while those for GUV suspensions were 1.76 × 10−9 , 4.41 × 10−9 and 3.09 × 10−8 m/(s Pa). To examine the effect of membrane pore radius on the size of liposomes obtained from single-pass extrusion of GUV suspensions, we estimated ratio of the mean radius of the extruded liposomes (Rv ) to nominal pore radius of the membranes (Rp ). Our data show that extrusion of GUV suspensions at low extrusion pressures (less than 0.014 MPa) through each of the membranes (25, 50 and 100 nm) resulted in liposomes having mean radius larger than the average membrane pore radius (Fig. 5A). Moreover, increase in trans-membrane pressure difference resulted in an asymptotic decrease in the mean radius of the extruded liposomes to a critical value beyond which their size remained unchanged with further increase in extrusion pressure. Interestingly, at high extrusion pressures mean liposome radius became smaller than the pore radius only for membrane with nominal pore radius of 100 nm (Fig. 5A).
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
479
Fig. 4. Effect of membrane pore radius on permeate flux for GUV suspension as a function of extrusion pressure. Flow rates of GUV suspension and water, perfused through membranes having nominal pore radii of 25, 50 and 100 nm at 55 ◦ C, were measured as a function of applied trans-membrane pressure difference. Solid symbols used for water while open symbols are for GUV suspensions. Data represented as mean ± SEM for n ≥ 3 independent experiments.
In fact, the smallest attainable values of (Rv /Rp ) were found to be 1.4 and 1.8 for membranes with pore radii of 50 and 25 nm, respectively (Fig. 5A). Also larger extrusion pressures were required for membranes with smaller pores to reach the smallest attainable values of (Rv /Rp ). Since DPPC is a saturated phospholipid with a glass transition temperature (Tg ) of 41 ◦ C, the extrusion experiments for DPPC GUV suspension was carried out at 55 ◦ C. In order to examine whether type of phospholipid or extrusion temperature affects the liposome size, we carried out the extrusion of POPC GUVs (Tg = −3 ◦ C) through membrane with nominal pore size of 100 nm at a temperature of 25 ◦ C over a range of trans-membrane pressure differences. Our data shows that the mean radius of the extruded POPC liposomes was not substantially different than that observed for DPPC liposomes extruded at 55 ◦ C over the pressure range examined (Fig. 5A). The polydispersity index of the DPPC liposome populations was found to decrease with increasing extrusion pressure and this decrease became steeper with increasing membrane pore size (Fig. 5B). However, smaller values of polydispersity index were attained with membranes having a smaller nominal pore size. Altogether, our data shows that smaller and relatively monodisperse liposomes are obtained when GUV suspensions are subjected to single-pass extrusion through membranes having smaller pores. In a previous work, it was suggested that the size of extruded liposomes depends on membrane pore size as well as on fluid velocity within the pore (Clerc and Thompson, 1994). Since our data showed that the mean radius of extruded liposome Rv did not converge to the membrane pore radius Rp (Fig. 5A), we expressed it in terms of the minimum attainable radius Rf . The value (Rv /Rf − 1) was plotted as a function of the average velocity of the GUV suspension in the membrane pore. As can be seen from Fig. 6, GUV extrusion data from all three membranes (25, 50 and 100 nm) collapses on to a single curve suggesting that (Rv /Rf − 1) is a function of the suspension velocity in the membrane. Recent theoretical and experimental studies on drop generation by flow focusing in micro-fluidic devices show that the drop size is inversely proportional to the capillary number (Ca = U/) where is the surface tension of the drop while and U represent the viscosity and the characteristic speed of the continuous phase, respectively (Lee et al., 2009; Stan et al., 2009). Therefore, we may
Fig. 5. Effect of membrane pore size on radius and polydispersity index of extruded liposomes. GUV suspensions were extruded through membranes having nominal pore radii of 25, 50 and 100 nm at 55 ◦ C. The average liposomes radius normalized by the membrane pore radius (Rv /Rp ) (A), and polydispersity index (B), of extruded liposomes was measured as a function of extrusion pressure. POPC GUV suspensions were extruded through membrane having nominal pore radius of 100 nm at 25 ◦ C. Data represented as mean ± SEM for n ≥ 3 independent experiments.
write (Rv /Rf − 1) = (k1 /U), where k1 is a constant so that Rv → Rf when U → ∞. Moreover, theoretical and experimental works have also shown that the vesicle lysis tension is a dynamic property related to the rate at which the membrane is stressed or loading rate LR (Boucher et al., 2007; Evans et al., 2003). In fact, in the limit of high loading rates the vesicle lysis tension is given by / ı ≈ Loge (LR /oı ı ) where oı is the spontaneous rate of defect formation in the vesicle bilayer and ı = kB T/rı2 is the tension scale of rate exponentiation as set by the defect area rı2 (Evans et al., 2003). Assuming the loading rate acting on the vesicle during extrusion to be proportional to the average velocity of the suspension (LR = k2 U) in the polycarbonate membrane pore, we have / ı ≈ Loge (k2 U/oı ı ). Substituting for in the expression for (Rv /Rf − 1) we get (Rv /Rf − 1) = (k1 ı Loge (k2 U/oı ı )/U), which may be written in terms of constants c1 and c2 as (Rv /Rf − 1) = c1 (Loge (c2 U))/U. Using the Levenberg–Marquardt nonlinear least squares algorithm, a fit of the aforementioned equation to (Rv /Rf − 1) vs U data (Fig. 6) was carried out and the constants c1 and c2 were found to be 0.00039 and 9700, respectively.
480
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
Fig. 6. Effect of average suspension velocity in membrane pore on normalized radius of extruded liposomes. GUV suspensions were extruded through membranes having nominal pore radii of 25, 50 and 100 nm at 55 ◦ C. The extruded liposome size, expressed as (Rv /Rf − 1) was plotted as a function of average velocity of the GUV suspension in the membrane pores. The model (Rv /Rf − 1) = c1 (Loge (c2 U))/U was fit to the data for c1 = 9700 and c2 = 0.0039.
detected for both MLV and GUV suspensions (Figs. 1 and 4) probably due to the higher temperature (55 ◦ C) and the lower lipid concentrations (0.2 mg/ml) at which extrusion was carried out. Our observations reveal that compared to MLV suspensions, single-pass extrusion at substantially lower trans-membrane pressure differences was sufficient to obtain small and mono-disperse liposomes from GUV suspensions (Fig. 2). Our data showed that the minimum attainable value for mean radius of extruded liposomes was smaller than the pore radius for membranes having nominal pore radius 100 nm or larger (Fig. 5) which is in agreement with a previous study (Frisken et al., 2000). A possible reason for this might be that a significant fraction of vesicles generated during MLV formation by thin film hydration or electroformation of GUVs may have radii below 100 nm. Moreover, our cumulative data from extrusion of GUV suspensions through polycarbonate membranes with nominal pore radii of 25, 50 and 100 nm showed that the ratio Rv /Rf was a unique function of the average velocity of the suspension within the membrane pore. Using the insights gained from recent results on droplet formation in microfluidic devices (Lee et al., 2009) and dynamic tension spectroscopy of phospholipid vesicles (Evans et al., 2003) we have proposed a functional relationship between extruded vesicle size and velocity of vesicular suspension within membrane pores. The proposed model in this work is limited to the high loading rate regime where membrane lysis tension is proportional to the logarithm of the membrane loading rate.
4. Discussion Several studies have quantified the encapsulation of markers by MLVs or GUVs and the effect of membrane extrusion on their encapsulation efficiency. Using confocal single-molecule detection, Sun and Chiu (2005) determined the CF encapsulation efficiencies of individual vesicles that consisted of 90% DPPC. The vesicles were classified as oligo-lamellar or multi-lamellar with the oligolamellar vesicles exhibiting higher mean encapsulation efficiency value of 36.3% compared to MLVs encapsulating only 17.5% CF (Sun and Chiu, 2005). Zhang et al. (2004) encapsulated the marker homocarnosine in DPPC MLVs which were subsequently extruded through membranes with nominal pore radii of 25, 50 and 500 nm yielding encapsulating efficiency values of 6, 14.5 and 22 percent, respectively. Our estimated CF encapsulation efficiency values for DPPC MLVs before and after extrusion are well within the range of values reported in the aforementioned studies. It is interesting that subjecting the MLV suspension to several cycles of extrusion did not have any significant effect on the CF encapsulation efficiency (Fig. 3). Our data suggest that extrusion of MLVs may primarily lead to the removal of outer lamellae from the vesicles, thereby leaving intact, the innermost lamellae enveloping the CF-bearing aqueous phase. In contrast, GUVs which exhibit much higher capacity for drug encapsulation compared to MLVs, are prone to lysis during extrusion leading to fluorophore release. As can be seen in Fig. 3, even single-pass extrusion of GUV suspensions results in significant reduction in the encapsulation efficiency. Previous work by Frisken and co-workers showed that a threshold trans-membrane pressure was required to initiate permeate flow during extrusion of MLV suspensions and this pressure was related to the lysis tension of the liposome bilayer (Frisken et al., 2000; Hunter and Frisken, 1998; Patty and Frisken, 2003). Based on their experimental findings, they proposed a model predicting the radius of the extruded liposome as a function of the ratio of transmembrane pressure difference and lysis tension of the MLVs (Patty and Frisken, 2003). Although they tried to minimize the effect of multi-lamellarity by subjecting the MLV suspensions to multiple freeze–thaw cycles, several high pressure extrusion cycles were required to obtain small liposomes with low polydispersity index. In our study the lysis tension related threshold pressure was not
5. Conclusions To our knowledge this is the first study to systematically characterize liposomes obtained by the extrusion of electroformed giant unilamellar vesicles through track-etched polycarbonate membranes as a function of the trans-membrane pressure difference. The major findings of this work are: (a) a single-pass extrusion of GUVs at moderate trans-membrane pressure difference is sufficient to generate monodisperse liposomes of small size comparable to those prepared by subjecting MLVs to multiple extrusion cycles at high applied pressure differences, (b) the GUV-derived liposomes have a significantly higher drug encapsulating potential compared to those generated from MLVs and (c) the size of extruded liposomes depends primarily on the suspension velocity within the membrane pore. Cumulatively, our data demonstrates that membrane extrusion of GUVs is an efficient method for preparing drug-loaded liposomes of controlled size for therapeutic applications. Acknowledgment The authors acknowledge Department of Science and Technology, India for financial support. References Bangham, A.D., Standish, M.M., Watkins, J.C., 1965. Diffusion of univalent ions across the lamellae of swollen phospholipids. Journal of Molecular Biology 13, 238–252. Boucher, P.A., Joos, B., Zuckermann, M.J., Fournier, L., 2007. Pore formation in a lipid bilayer under a tension ramp: modeling the distribution of rupture tensions. Biophysical Journal 92, 4344–4355. Clerc, S.G., Thompson, T.E., 1994. A possible mechanism for vesicle formation by extrusion. Biophysical Journal 67, 475–477. Deamer, D., Bangham, A.D., 1976. Large volume liposomes by an ether vaporization method. Biochimica et Biophysica Acta 443, 629–634. Elzainy, A.A.W., Gu, X., Simons, F.E.R., Simons, K.J., 2005. Hydroxyzine- and cetirizine-loaded liposomes: effect of duration of thin film hydration, freezethawing, and changing buffer pH on encapsulation and stability. Drug Development and Industrial Pharmacy 31, 281–291. Estes, D.J., Mayer, M., 2005. Electroformation of giant liposomes from spin-coated films of lipids. Colloids and Surfaces B: Biointerfaces 42, 115–123.
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481 Evans, E., Heinrich, V., Ludwig, F., Rawicz, W., 2003. Dynamic tension spectroscopy and strength of biomembranes. Biophysical Journal 85, 2342–2350. Frisken, B.J., Asman, C., Patty, P.J., 2000. Studies of vesicle extrusion. Langmuir 16, 928–933. Gruner, S.M., Lenk, R.P., Janoff, A.S., Ostro, M.J., 1985. Novel multilayered lipid vesicles: comparison of physical characteristics of multilamellar liposomes and stable plurilamellar vesicles. Biochemistry 24, 2833–2842. Harashima, H., Sakata, K., Funato, K., Kiwada, H., 1994. Enhanced hepatic uptake of liposomes through complement activation depending on the size of liposomes. Pharmaceutical Research 11, 402–406. Hope, M.J., Bally, M.B., Mayer, L.D., Janoff, A.S., Cullis, P.R., 1986. Generation of multilamellar and unilamellar phospholipid vesicles. Chemistry and Physics of Lipids 40, 89–107. Hope, M.J., Bally, M.B., Webb, G., Cullis, P.R., 1985. Production of large unilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution, trapped volume and ability to maintain a membrane potential. Biochimica et Biophysica Acta—Biomembranes 812, 55–65. Hunt, C.A., Rustum, Y.M., Mayhew, E., Papahadjopoulos, D., 1979. Retention of cytosine arabinoside in mouse lung following intravenous administration in liposomes of different size. Drug Metabolism and Disposition 7, 124–128. Hunter, D.G., Frisken, B.J., 1998. Effect of extrusion pressure and lipid properties on the size and polydispersity of lipid vesicles. Biophysical Journal 74, 2996–3002. Johnson, S.M., Bangham, A.D., Hill, M.W., Korn, E.D., 1971. Single bilayer liposomes. Biochimica et Biophysica Acta 233, 820–826. Jousma, H., Talsma, H., Spies, F., Joosten, J.G.H., Junginger, H.E., Crommelin, D.J.A., 1987. Characterization of liposomes. The influence of extrusion of multilamellar vesicles through polycarbonate membranes on particle size, particle size distribution and number of bilayers. International Journal of Pharmaceutics 35, 214–263. Kim, S., Martin, G.M., 1981. Preparation of cell-size unilamellar liposomes with high captured volume and defined size distribution. Biochimica et Biophysica Acta—Biomembranes 646, 1–9. Koppel, D.E., 1972. Analysis of macromolecular polydispersity in intensity correlation spectroscopy—method of cumulants. Journal of Chemical Physics 57, 4814. Lasic, D.D., 1993. Liposomes: From Physics to Application. Elsevier Science B.V., Amsterdam. Lee, W., Walker, L.M., Anna, S.L., 2009. Role of geometry and fluid properties in droplet and thread formation processes in planar flow focusing. Physics of Fluids, 21.
481
Mayer, L.D., Hope, M.J., Cullis, P.R., 1986. Vesicles of variable sizes produced by a rapid extrusion procedure. Biochimica et Biophysica Acta—Biomembranes 858, 161–168. Ohsawa, T., Miura, H., Harada, K., 1984. A novel method for preparing liposome with a high capacity to encapsulate proteinous drugs: freeze-drying method. Chemical and Pharmaceutical Bulletin 32, 2442–2445. Patty, P.J., Frisken, B.J., 2003. The pressure-dependence of the size of extruded vesicles. Biophysical Journal 85, 996–1004. Perkins, W.R., Minchey, S.R., Ahl, P.L., Janoff, A.S., 1993. The determination of liposome captured volume. Chemistry and Physics of Lipids 64, 197–217. Popa, I., Trulsson, M., Papastavrou, G., Borkovec, M., Jonsson, B., 2009. Long-ranged attractive forces induced by adsorbed dendrimers: direct force measurements and computer simulations. Langmuir 25, 12435–12438. Selser, J.C., Yeh, Y., Baskin, R.J., 1976. Light-scattering characterization of membranevesicles. Biophysical Journal 16, 337–356. Stan, C.A., Tang, S.K.Y., Whitesides, G.M., 2009. Independent control of drop size and velocity in microfluidic flow-focusing generators using variable temperature and flow rate. Analytical Chemistry 81, 2399–2402. Storm, G., Koppenhagen, F., Heeremans, A., Vingerhoeds, M., Woodle, M.C., Crommelin, D.J.A., 1995. Novel developments in liposomal delivery of peptides and proteins. Journal of Controlled Release 36, 19–24. Sun, B., Chiu, D.T., 2005. Determination of the encapsulation efficiency of individual vesicles using single-vesicle photolysis and confocal single-molecule detection. Analytical Chemistry 77, 2770–2776. Szoka Jr., F., Papahadjopoulos, D., 1978. Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proceedings of the National Academy of Sciences of the United States of America 75, 4194–4198. Vemuri, S., Rhodes, C.T., 1995. Preparation and characterization of liposomes as therapeutic delivery systems: a review. Pharmaceutica Acta Helvetiae 70, 95–111. Vuillemard, J.C., 1991. Recent advances in the large-scale production of lipid vesicles for use in food products: microfluidization. Journal of Microencapsulation 8, 547–562. Woodle, M.C., 1995. Sterically stabilized liposome therapeutics. Advanced Drug Delivery Reviews 16, 249–265. Zhang, L., Hu, J., Lu, Z., 1997. Preparation of liposomes with a controlled assembly procedure. Journal of Colloid and Interface Science 190, 76–80. Zhang, X.M., Patel, A.B., de Graaf, R.A., Behar, K.L., 2004. Determination of liposomal encapsulation efficiency using proton NMR spectroscopy. Chemistry and Physics of Lipids 127, 113–120.
Contents lists available at SciVerse ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
Extrusion of electroformed giant unilamellar vesicles through track-etched membranes Yogita P. Patil a , Mrunmayi D. Kumbhalkar b , Sameer Jadhav a,∗ a b
Department of Chemical Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India Department of Chemical Engineering, Birla Institute of Technology and Science, Vidya Vihar Campus, Pilani, Rajasthan 333 031, India
a r t i c l e
i n f o
Article history: Available online 3 December 2011 Keywords: Phospholipid Bilayer Lamellarity Lysis tension Encapsulation efficiency Dynamic light scattering
a b s t r a c t Unilamellar vesicle populations having a narrow size distribution and mean radius below 100 nm are preferred for drug delivery applications. In the present work, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) was used to prepare giant unilamellar vesicles (GUVs) by electroformation and multilamellar vesicles (MLVs) by thin film hydration. Our experiments show that in contrast to MLVs, a single-pass extrusion of GUVs through track-etched polycarbonate membranes at moderate pressure differences is sufficient to produce small liposomes having low polydispersity index. Moreover, we observe that the drug encapsulating potential of extruded liposomes obtained from GUVs is significantly higher compared to liposomes prepared by extrusion of MLVs. Furthermore, our experiments carried out for varying membrane pore diameters and extrusion pressures suggest that the size of extruded liposomes is a function of the velocity of GUV suspensions in the membrane pore. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Liposomes are spherical capsules that comprise of one or more phospholipid bilayers. Liposomes may be unilamellar vesicles with a single bilayer encapsulating an aqueous core or multi-lamellar vesicles (MLVs) with several concentric bilayer shells. Based on size, unilamellar vesicles are classified as small unilamellar vesicles (0.025–0.05 m, SUVs), large unilamellar vesicles (0.05–0.5 m, LUVs), and giant unilamellar vesicles (greater than 1 m, GUVs) (Vemuri and Rhodes, 1995; Vuillemard, 1991). To increase the efficacy of therapeutic agents, liposomes are increasingly used as immune evasive carriers for targeted delivery and controlled release of drugs (Lasic, 1993; Storm et al., 1995). Since drug-loaded MLVs are known to exhibit erratic pharmacokinetics and undergo rapid clearance by the reticulo-endothelial system (Hunt et al., 1979), mono-disperse unilamellar vesicles having mean diameter less than 200 nm are preferred for drug delivery applications (Harashima et al., 1994; Woodle, 1995). It is well established that processes including thin film hydration, reverse phase evaporation and ethanol injection lead to the formation of MLVs that exhibit heterogeneity in lamellarity and size (Bangham et al., 1965; Deamer and Bangham, 1976; Szoka and Papahadjopoulos, 1978). A severe limitation of MLVs is their poor ability to encapsulate hydrophilic drugs when compared to unilamellar vesicles of equivalent lipid content. Even though
∗ Corresponding author. Tel.: +91 22 2576 7219; fax: +91 22 25726895. E-mail address: [email protected] (S. Jadhav). 0009-3084/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.chemphyslip.2011.11.013
drug encapsulation efficiency of MLVs was increased using methods such as freeze-drying and hydrating lipids in presence of organic solvents, additional steps such as detergent removal were introduced (Gruner et al., 1985; Ohsawa et al., 1984). Previous studies have shown that subjecting MLV suspensions to sonication, freeze–thaw cycles or multiple membrane extrusions produces vesicles with reduced size and lamellarity (Hope et al., 1985, 1986; Johnson et al., 1971). Extrusion of MLV suspensions through well-defined cylindrical pores of track-etched polycarbonate membranes has been the focus of several studies (Frisken et al., 2000; Hunter and Frisken, 1998; Mayer et al., 1986; Patty and Frisken, 2003; Popa et al., 2009). These studies reveal that many extrusion cycles at high trans-membrane pressures are required to obtain smaller liposomes with narrower size distributions. For example, more than ten extrusion cycles at pressures greater than 300 psig (∼2 MPa) are typically required to realize the minimum achievable liposome size by extrusion of MLV suspension with a membrane having mean pore radius of 50 nm (Mayer et al., 1986; Patty and Frisken, 2003). However, even after subjecting MLV suspensions to several extrusion cycles through membranes having nominal pore radii of 100 and 200 nm, it was observed that ∼30% of extruded vesicles remained oligo-lamellar (Mayer et al., 1986). Moreover, it was reported that subjecting MLVs to multiple extrusions led to further loss of encapsulated drug as well as lipid (Jousma et al., 1987; Perkins et al., 1993). While attempts have been made towards theoretical prediction of extruded liposome size, the exact mechanism by which MLVs are converted to smaller liposomes remains unknown. Clerc and Thompson (1994) have proposed a model based on flow rate
476
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
dependent Rayleigh instability, where MLVs are squeezed through membrane pores as long tubular structures that become unstable and break into smaller cylinders either inside or at the exit of the pore. The maximum size of the extruded liposomes was predicted to be 0.5 times the membrane pore radius, though the relation between fluid velocity and mean liposome size was not worked out. In contrast, Patty and Frisken (2003) proposed a simple bubble model where liposome formation occurs via MLV lysis at the entrance of the pore and the resulting liposome size depends on the applied pressure difference and not on fluid velocity within the pore. One of the limitations of the aforementioned study is that membrane lysis tension and extruded liposome size are estimated from data obtained from different extrusion cycles. Moreover, subsequent theoretical and experimental studies have shown that the membrane lysis tension at a given temperature is not a singlevalued material property, but is a function of the rate at which the membrane is mechanically stressed (Boucher et al., 2007; Evans et al., 2003). In contrast to MLVs, GUVs have been shown to enclose larger aqueous volumes for encapsulation of hydrophilic drugs (Kim and Martin, 1981). Therefore, in the present work we propose that, compared to MLVs, extrusion of GUVs requires lower trans-membrane pressures and fewer extrusion cycles to obtain unilamellar liposomes with larger drug encapsulating capacity. To this end, we carried out single-pass extrusion of electroformed GUVs and compared the mean size and fluorophore encapsulating efficiency of the resulting liposomes to those obtained after multiple extrusion cycles of MLVs. Moreover, we systematically investigated the effect of applied pressure difference on the mean size and polydispersity of liposomes formed by extrusion of GUV suspension through track-etched polycarbonate membranes having nominal pore radii of 25, 50 and 100 nm. Furthermore, we propose a simple functional dependence of extruded liposome size on the average velocity of the vesicle suspension inside the membrane pore. 2. Materials and methods 2.1. Materials The phospholipids 1,2-dipalmitoyl-sn-glycero-3phosphocholine (DPPC) and 1-palmitoyl-2-oleoyl-sn-glycero3-phosphocholine (POPC) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). Triton X-100 was procured from Merck (Darmstadt, Germany) while the fluorophore 5(6) carboxyfluorescein was from Fluka (Buchs, Switzerland). Indium tin oxide (ITO) coated glass slides (25 mm × 75 mm × 1.1 mm) were obtained from Delta Technologies (Stillwater, MN, USA). Nuclepore track-etched polycarbonate membranes (25 mm diameter and 0.006 mm thickness) with nominal pore radii of 25, 50 and 100 nm as well as polyester drain disc supports were purchased from Whatman Inc (Clifton, NJ, USA) while dialysis membrane (14.3 mm diameter, 12–15 kDa MW cut off, capacity 1.61 ml/cm) was from Himedia (Mumbai, India). Ethanol, methanol and chloroform were procured from SD Fine Chemicals (Mumbai, India). 2.2. Preparation of GUV and MLV suspensions GUVs were prepared by the electroformation method using a previously published protocol (Estes and Mayer, 2005). Briefly, 50 l of 2 mg/ml DPPC lipid solution in chloroform was spin-coated at 1000 rpm for 1 min onto ITO slides mounted on a spin coater (Photoresist Spinner PRS-6K, Ducom Instruments, Bangalore, India) and the residual chloroform was removed by overnight application of vacuum. Lipid coated faces of two ITO slides were separated by a silicone rubber spacer and held together with clips forming a
35 mm × 15 mm × 2 mm chamber. Next, 1 ml deionized water was injected with a syringe through the silicone rubber spacer into the chamber and the set up was kept in an oven maintained at 55 ◦ C. An AC voltage (10 Hz, 3 Vpp, sinusoidal) was applied for 2 h across the ITO slides using a function generator (Model 33220A, Agilent Technologies, CA, USA). Finally, a square wave was applied for 1 min to detach the vesicles from the ITO surface. The resulting GUV suspension was immediately used for extrusion studies. For preparation of POPC GUVs, electroformation was carried out at 25 ◦ C. DPPC MLVs were prepared by the well-established thin film hydration method (Elzainy et al., 2005; Mayer et al., 1986). Briefly, 0.3 ml of 2 mg/ml DPPC lipid solution in chloroform was evaporated in a round-bottom flask of a rotary evaporator (Roteva Equitron, Medica Instruments, Mumbai, India) at a speed of 100 rpm for 10 min and maintained at 55 ◦ C. The thin lipid film in the round-bottom flask was thoroughly dried in a vacuum desiccator overnight to remove residual chloroform and subsequently hydrated using 3 ml of deionized water in the rotary evaporator at 100 rpm for 60 min maintained at 55 ◦ C. The resulting MLV suspension was taken through five freeze–thaw cycles which involved freezing the suspension in liquid nitrogen and subsequently thawing at 55 ◦ C in a water bath. The MLV suspension was immediately used for extrusion studies. 2.3. Extrusion of GUVs and MLVs An extruder was fabricated in-house that comprised of a cylindrical stainless steel chamber with a holder for the polycarbonate membrane and polyester membrane support. The chamber was water-jacketed to carry out extrusions at defined temperatures and pressurized using a pressure regulated nitrogen gas cylinder. The DPPC GUV suspensions were subjected to single-pass extrusion through track-etched polycarbonate membranes with nominal pore radii of 25, 50 and 100 nm at 55 ◦ C and pressures ranging from 1 to 400 psi (0.007–2.75 MPa). The DPPC MLV suspensions were subjected to double-pass extrusion through membranes with nominal pore radius 50 nm at 55 ◦ C and pressures ranging from 5 to 400 psi (0.035–2.75 MPa). The flow rate of the resulting liposome suspension from the extruder was measured using a measuring cylinder and stop watch. The membrane permeability for GUV suspensions (as well as for water and MLV suspensions) was calculated as the slope of the linear fit to the permeate flux vs. extrusion pressure plot. For the phospholipid POPC, GUV suspensions were extruded at 25 ◦ C for pressures ranging from 1 to 100 psi (0.007–0.7 MPa) through polycarbonate membranes having nominal pore radius of 100 nm. For experiments related to estimation of encapsulation efficiency, single-pass extrusion of fluorophoreloaded GUV suspensions was carried out while MLV suspensions were subjected to eleven cycles of extrusion through polycarbonate membranes having nominal pore radius of 100 nm at 20 psi (0.14 MPa) and 55 ◦ C. 2.4. Estimation of mean radius and polydispersity of extruded liposomes The mean radius and polydispersity of the extruded liposomes were measured by Dynamic Light Scattering (DLS, Zetasizer Nano ZS, Malvern Instruments, Worcestershire, UK) at 25 ◦ C. The vesicle suspension of lipid concentration 0.2 mg/ml was placed in a glass cuvette. The vesicle suspension was then illuminated by He–Ne red laser at 633 nm, the scattered light intensity was measured by the detector positioned at 173◦ . Using the digital correlator and software provided with the instrument, the temporal auto-correlation function of scattered light intensity was calculated. The method of cumulants was used to obtain the average decay rate and polydispersity index from the correlogram (Koppel, 1972). The intensity
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
477
weighted diffusion coefficient was calculated from the average decay rate and the Stokes–Einstein equation was used to obtain mean radius of liposome from the diffusion coefficient (Selser et al., 1976). Three measurements were made per sample and the experiments were repeated at least three times. The mean radius of extruded liposomes was also reported as a function of the average velocity of suspensions in the membrane pore. The average velocity of the GUV suspension in the membrane pores was estimated from the product of average velocity of water (calculated using Hagen–Poisuelle equation) at a given pressure difference and the ratio of permeability of the suspension to that of water. 2.5. Estimation of fluorophore encapsulating efficiency of liposomes The capacity to encapsulate a hydrophilic fluorophore by GUVs, MLVs and the extruded liposomes was estimated using a previously published protocol (Zhang et al., 1997). Briefly, hydration step in the preparation of GUVs was carried out using a 12 mM aqueous solution of 5(6) carboxyfluorescein (CF). One ml of the GUV suspension was introduced into a dialysis bag, placed in a beaker containing 200 ml water that was magnetically stirred at 25 ◦ C for 36 h to remove unencapsulated CF. The resulting CF encapsulated GUV suspension (10 l) from the dialysis bag was diluted with water to a final volume of 1000 ml. Subsequently, Triton X-100 (10%, v/v) was added to break the liposomes and fluorescence emission spectrum of the released CF was recorded from 490 to 650 nm using a fluorescence spectrophotometer (LS 55, Perkin-Elmer, MA, USA). The fluorescence of an aqueous solution of CF (10 l of 12 mM CF dissolved in water to a final volume of 1000 ml) was also recorded. Based on total CF present in the hydrating solution, the percentage of CF detected in the GUVs was reported as encapsulation efficiency. Efficiency of CF encapsulation by MLVs was estimated using a procedure similar to that employed for GUVs. CF loaded GUVs and MLVs were extruded through polycarbonate membrane of nominal pore radius 100 nm at 20 psi (0.14 MPa) and the encapsulation efficiency of resulting liposomes was also estimated. 3. Results In the present study we carried out a comparison of DPPC liposome formation by membrane extrusion of (a) MLV suspension, obtained by thin film hydration, and (b) electroformed GUV suspension. To this end, we systematically characterized the effect of extrusion pressure on the size and polydispersity of liposomes formed by the two methods. Moreover, we also quantified the capacity of GUVs and MLVs to encapsulate the fluorescent marker CF, as well as fluorophore loss due to extrusion of the vesicular suspensions. Furthermore, we examined the effect of nominal pore radius of the membranes on the size of liposomes formed by extrusion of GUV suspensions as a function of the applied pressure difference. Finally, we propose a simple model relating the mean radius of extruded liposomes to the average velocity of the vesicular suspension in the membrane pore. 3.1. Effect of extrusion pressure on GUV and MLV suspensions As a first step we compared the permeate flux during extrusion of GUV and MLV suspensions as a function of extrusion pressure. To this end, extrusions of GUVs and MLVs of DPPC were carried out through track-etched polycarbonate membranes having nominal pore radius of 50 nm at 55 ◦ C for extrusion pressures ranging from 0.007 to 0.7 MPa. A linear increase in permeate flux with respect to extrusion pressure was observed for both, GUV and MLV suspensions (Fig. 1). However, for a given pressure, the flow of GUV
Fig. 1. Effect of extrusion pressure on permeate flux for GUV and MLV suspensions. MLV suspensions were prepared by thin film hydration method while GUV suspensions were obtained by electroformation method. The permeate flux for DPPC GUV and MLV suspensions as well as for water were measured at 55 ◦ C as a function of extrusion pressure during first-pass through polycarbonate track-etched membrane with nominal pore radius of 50 nm. Data represented as mean ± SEM for n ≥ 3 independent experiments.
suspensions through the membrane was significantly faster compared to that of MLV suspensions. Moreover, when compared to water, permeate flux values observed for GUV suspensions were substantially lower to those of water at similar pressures (Fig. 1). Furthermore, the membrane permeability for water and vesicle suspensions was derived from the relation between permeate flux and extrusion pressure. The respective permeability values for water, GUV and MLV suspensions were 1.76 × 10−8 , 4.41 × 10−9 and 2.65 × 10−9 m/(s Pa). Subsequent experiments were aimed at comparing the size distribution of GUVs and MLVs after single-pass extrusion through a membrane with nominal pore radius of 50 nm as a function of extrusion pressure. We observed that a single pass of MLV suspensions through the membrane at low extrusion pressures resulted in large liposomes that were highly polydisperse for the mean size or size distribution to be reliably measured using DLS (data not shown). Even at higher pressure differences, MLV suspensions were required to be extruded at least twice to obtain good fits to the scattering data for estimation of mean radius and polydispersity index. For GUV suspensions subjected to single-pass extrusion, the mean radius of liposomes asymptotically decreased with increasing extrusion pressure to ∼70 nm at 100 psi (0.7 MPa), after which no further reduction in size was observed (Fig. 2A). In comparison, double-pass extrusion of MLV suspensions required substantially higher pressure differences to generate liposomes of mean radii similar to those generated from single-pass GUV extrusion. In fact, MLV suspensions subjected to an extrusion pressure of 400 psi (2.75 MPa) resulted in liposomes having a mean radius of 85 nm (Fig. 2A). For both, GUV and MLV suspensions, the extruded liposome populations became progressively monodisperse with increase in the applied pressure difference across the membrane. However, similar to observations related to mean radius, considerably larger extrusion pressures were required for MLV suspensions to generate extruded liposome populations having polydispersity index comparable to GUV derived liposomes (Fig. 2B). Taken together, our results demonstrate that relatively small and monodisperse population of liposomes are obtained by single-pass extrusion of GUVs at substantially lower
478
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
Fig. 3. Encapsulation efficiency of MLV and GUV suspensions before and after extrusion. Aqueous solution of 5(6) carboxyfluorescein (12 mM) was encapsulated during thin film hydration of MLVs and electroformation of GUVs at 55 ◦ C. GUV suspensions were subjected to single-pass extrusion while MLV suspensions were subjected to eleven extrusion cycles at 0.14 MPa through membrane with nominal pore radius of 100 nm at 55 ◦ C. The amount of encapsulated CF was estimated after dialysis of extra vesicular fluorophore. Data represented as mean ± SEM for n ≥ 3 experiments. NS: not significant with respect to MLVs before extrusions, *p < 0.05 with respect to GUVs before extrusion.
adequate for obtaining small liposomes with a narrow size distribution. The GUV derived lipsomes were found to enclose 45% of CF present in the hydrating solution used to prepare GUVs (Fig. 3), indicating that liposomes obtained from single-pass extrusion of GUVs exhibit almost a 4-fold larger drug-holding capacity compared to those prepared by the extrusion of MLV suspensions. 3.3. Effect of membrane pore size on extruded liposomes
Fig. 2. Effect of extrusion pressure on mean radius and polydispersity index of liposomes obtained by extrusion of GUV and MLV suspensions. GUV suspensions were subjected to single-pass extrusion while MLV suspensions were subjected to double pass extrusion through membrane with nominal pore radius of 50 nm at 55 ◦ C. The mean radius (A), and polydispersity index (B), of the extruded liposome populations are reported as a function of extrusion pressure. Data represented as mean ± SEM for n ≥ 3 experiments.
trans-membrane pressures when compared to liposomes derived from MLVs. 3.2. Encapsulation efficiency of GUVs, MLVs and extruded liposomes To compare the drug encapsulating potential of GUVs to that of MLVs, we quantified the fraction of the fluorophore 5(6) carboxyfluorescein (CF) present in the hydrating solution that was enclosed into vesicles. Only ∼15% of CF was encapsulated by MLVs, whereas a 4-fold higher percentage of CF was enclosed in GUVs (Fig. 3). To assess whether membrane extrusion resulted in CF loss from the liposomes, CF loaded MLVs suspension was subjected to eleven cycles of extrusion through membrane with nominal pore size of 100 nm at 20 psi (0.14 MPa). The resulting lipsomes were found to contain 12% of the CF present in the hydrating solution used for MLV preparation (Fig. 3). The CF loaded GUVs suspension was subjected to single-pass extrusion through membrane with nominal pore size of 100 nm at 20 psi (0.14 MPa) since this was found
Ensuing experiments were aimed at investigating the effect of average pore radius of the membrane on the size and polydispersity of liposomes obtained from single-pass extrusion of GUVs as a function of the applied pressure difference. To this end, applied pressure difference was varied from 0.007 to 2.75 MPa during extrusion of suspensions of DPPC GUVs at 55 ◦ C through polycarbonate membranes having nominal pore radii of 25, 50 and 100 nm. The permeate rate for pure water as well as for GUV suspensions was found to increase linearly with the trans-membrane pressure difference in the low pressure regime (Fig. 4). For membranes having nominal pore size of 25, 50 and 100 nm, respective permeability values of water were found to be 4.41 × 10−9 , 1.76 × 10−8 and 4.41 × 10−8 m/(s Pa), while those for GUV suspensions were 1.76 × 10−9 , 4.41 × 10−9 and 3.09 × 10−8 m/(s Pa). To examine the effect of membrane pore radius on the size of liposomes obtained from single-pass extrusion of GUV suspensions, we estimated ratio of the mean radius of the extruded liposomes (Rv ) to nominal pore radius of the membranes (Rp ). Our data show that extrusion of GUV suspensions at low extrusion pressures (less than 0.014 MPa) through each of the membranes (25, 50 and 100 nm) resulted in liposomes having mean radius larger than the average membrane pore radius (Fig. 5A). Moreover, increase in trans-membrane pressure difference resulted in an asymptotic decrease in the mean radius of the extruded liposomes to a critical value beyond which their size remained unchanged with further increase in extrusion pressure. Interestingly, at high extrusion pressures mean liposome radius became smaller than the pore radius only for membrane with nominal pore radius of 100 nm (Fig. 5A).
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
479
Fig. 4. Effect of membrane pore radius on permeate flux for GUV suspension as a function of extrusion pressure. Flow rates of GUV suspension and water, perfused through membranes having nominal pore radii of 25, 50 and 100 nm at 55 ◦ C, were measured as a function of applied trans-membrane pressure difference. Solid symbols used for water while open symbols are for GUV suspensions. Data represented as mean ± SEM for n ≥ 3 independent experiments.
In fact, the smallest attainable values of (Rv /Rp ) were found to be 1.4 and 1.8 for membranes with pore radii of 50 and 25 nm, respectively (Fig. 5A). Also larger extrusion pressures were required for membranes with smaller pores to reach the smallest attainable values of (Rv /Rp ). Since DPPC is a saturated phospholipid with a glass transition temperature (Tg ) of 41 ◦ C, the extrusion experiments for DPPC GUV suspension was carried out at 55 ◦ C. In order to examine whether type of phospholipid or extrusion temperature affects the liposome size, we carried out the extrusion of POPC GUVs (Tg = −3 ◦ C) through membrane with nominal pore size of 100 nm at a temperature of 25 ◦ C over a range of trans-membrane pressure differences. Our data shows that the mean radius of the extruded POPC liposomes was not substantially different than that observed for DPPC liposomes extruded at 55 ◦ C over the pressure range examined (Fig. 5A). The polydispersity index of the DPPC liposome populations was found to decrease with increasing extrusion pressure and this decrease became steeper with increasing membrane pore size (Fig. 5B). However, smaller values of polydispersity index were attained with membranes having a smaller nominal pore size. Altogether, our data shows that smaller and relatively monodisperse liposomes are obtained when GUV suspensions are subjected to single-pass extrusion through membranes having smaller pores. In a previous work, it was suggested that the size of extruded liposomes depends on membrane pore size as well as on fluid velocity within the pore (Clerc and Thompson, 1994). Since our data showed that the mean radius of extruded liposome Rv did not converge to the membrane pore radius Rp (Fig. 5A), we expressed it in terms of the minimum attainable radius Rf . The value (Rv /Rf − 1) was plotted as a function of the average velocity of the GUV suspension in the membrane pore. As can be seen from Fig. 6, GUV extrusion data from all three membranes (25, 50 and 100 nm) collapses on to a single curve suggesting that (Rv /Rf − 1) is a function of the suspension velocity in the membrane. Recent theoretical and experimental studies on drop generation by flow focusing in micro-fluidic devices show that the drop size is inversely proportional to the capillary number (Ca = U/) where is the surface tension of the drop while and U represent the viscosity and the characteristic speed of the continuous phase, respectively (Lee et al., 2009; Stan et al., 2009). Therefore, we may
Fig. 5. Effect of membrane pore size on radius and polydispersity index of extruded liposomes. GUV suspensions were extruded through membranes having nominal pore radii of 25, 50 and 100 nm at 55 ◦ C. The average liposomes radius normalized by the membrane pore radius (Rv /Rp ) (A), and polydispersity index (B), of extruded liposomes was measured as a function of extrusion pressure. POPC GUV suspensions were extruded through membrane having nominal pore radius of 100 nm at 25 ◦ C. Data represented as mean ± SEM for n ≥ 3 independent experiments.
write (Rv /Rf − 1) = (k1 /U), where k1 is a constant so that Rv → Rf when U → ∞. Moreover, theoretical and experimental works have also shown that the vesicle lysis tension is a dynamic property related to the rate at which the membrane is stressed or loading rate LR (Boucher et al., 2007; Evans et al., 2003). In fact, in the limit of high loading rates the vesicle lysis tension is given by / ı ≈ Loge (LR /oı ı ) where oı is the spontaneous rate of defect formation in the vesicle bilayer and ı = kB T/rı2 is the tension scale of rate exponentiation as set by the defect area rı2 (Evans et al., 2003). Assuming the loading rate acting on the vesicle during extrusion to be proportional to the average velocity of the suspension (LR = k2 U) in the polycarbonate membrane pore, we have / ı ≈ Loge (k2 U/oı ı ). Substituting for in the expression for (Rv /Rf − 1) we get (Rv /Rf − 1) = (k1 ı Loge (k2 U/oı ı )/U), which may be written in terms of constants c1 and c2 as (Rv /Rf − 1) = c1 (Loge (c2 U))/U. Using the Levenberg–Marquardt nonlinear least squares algorithm, a fit of the aforementioned equation to (Rv /Rf − 1) vs U data (Fig. 6) was carried out and the constants c1 and c2 were found to be 0.00039 and 9700, respectively.
480
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481
Fig. 6. Effect of average suspension velocity in membrane pore on normalized radius of extruded liposomes. GUV suspensions were extruded through membranes having nominal pore radii of 25, 50 and 100 nm at 55 ◦ C. The extruded liposome size, expressed as (Rv /Rf − 1) was plotted as a function of average velocity of the GUV suspension in the membrane pores. The model (Rv /Rf − 1) = c1 (Loge (c2 U))/U was fit to the data for c1 = 9700 and c2 = 0.0039.
detected for both MLV and GUV suspensions (Figs. 1 and 4) probably due to the higher temperature (55 ◦ C) and the lower lipid concentrations (0.2 mg/ml) at which extrusion was carried out. Our observations reveal that compared to MLV suspensions, single-pass extrusion at substantially lower trans-membrane pressure differences was sufficient to obtain small and mono-disperse liposomes from GUV suspensions (Fig. 2). Our data showed that the minimum attainable value for mean radius of extruded liposomes was smaller than the pore radius for membranes having nominal pore radius 100 nm or larger (Fig. 5) which is in agreement with a previous study (Frisken et al., 2000). A possible reason for this might be that a significant fraction of vesicles generated during MLV formation by thin film hydration or electroformation of GUVs may have radii below 100 nm. Moreover, our cumulative data from extrusion of GUV suspensions through polycarbonate membranes with nominal pore radii of 25, 50 and 100 nm showed that the ratio Rv /Rf was a unique function of the average velocity of the suspension within the membrane pore. Using the insights gained from recent results on droplet formation in microfluidic devices (Lee et al., 2009) and dynamic tension spectroscopy of phospholipid vesicles (Evans et al., 2003) we have proposed a functional relationship between extruded vesicle size and velocity of vesicular suspension within membrane pores. The proposed model in this work is limited to the high loading rate regime where membrane lysis tension is proportional to the logarithm of the membrane loading rate.
4. Discussion Several studies have quantified the encapsulation of markers by MLVs or GUVs and the effect of membrane extrusion on their encapsulation efficiency. Using confocal single-molecule detection, Sun and Chiu (2005) determined the CF encapsulation efficiencies of individual vesicles that consisted of 90% DPPC. The vesicles were classified as oligo-lamellar or multi-lamellar with the oligolamellar vesicles exhibiting higher mean encapsulation efficiency value of 36.3% compared to MLVs encapsulating only 17.5% CF (Sun and Chiu, 2005). Zhang et al. (2004) encapsulated the marker homocarnosine in DPPC MLVs which were subsequently extruded through membranes with nominal pore radii of 25, 50 and 500 nm yielding encapsulating efficiency values of 6, 14.5 and 22 percent, respectively. Our estimated CF encapsulation efficiency values for DPPC MLVs before and after extrusion are well within the range of values reported in the aforementioned studies. It is interesting that subjecting the MLV suspension to several cycles of extrusion did not have any significant effect on the CF encapsulation efficiency (Fig. 3). Our data suggest that extrusion of MLVs may primarily lead to the removal of outer lamellae from the vesicles, thereby leaving intact, the innermost lamellae enveloping the CF-bearing aqueous phase. In contrast, GUVs which exhibit much higher capacity for drug encapsulation compared to MLVs, are prone to lysis during extrusion leading to fluorophore release. As can be seen in Fig. 3, even single-pass extrusion of GUV suspensions results in significant reduction in the encapsulation efficiency. Previous work by Frisken and co-workers showed that a threshold trans-membrane pressure was required to initiate permeate flow during extrusion of MLV suspensions and this pressure was related to the lysis tension of the liposome bilayer (Frisken et al., 2000; Hunter and Frisken, 1998; Patty and Frisken, 2003). Based on their experimental findings, they proposed a model predicting the radius of the extruded liposome as a function of the ratio of transmembrane pressure difference and lysis tension of the MLVs (Patty and Frisken, 2003). Although they tried to minimize the effect of multi-lamellarity by subjecting the MLV suspensions to multiple freeze–thaw cycles, several high pressure extrusion cycles were required to obtain small liposomes with low polydispersity index. In our study the lysis tension related threshold pressure was not
5. Conclusions To our knowledge this is the first study to systematically characterize liposomes obtained by the extrusion of electroformed giant unilamellar vesicles through track-etched polycarbonate membranes as a function of the trans-membrane pressure difference. The major findings of this work are: (a) a single-pass extrusion of GUVs at moderate trans-membrane pressure difference is sufficient to generate monodisperse liposomes of small size comparable to those prepared by subjecting MLVs to multiple extrusion cycles at high applied pressure differences, (b) the GUV-derived liposomes have a significantly higher drug encapsulating potential compared to those generated from MLVs and (c) the size of extruded liposomes depends primarily on the suspension velocity within the membrane pore. Cumulatively, our data demonstrates that membrane extrusion of GUVs is an efficient method for preparing drug-loaded liposomes of controlled size for therapeutic applications. Acknowledgment The authors acknowledge Department of Science and Technology, India for financial support. References Bangham, A.D., Standish, M.M., Watkins, J.C., 1965. Diffusion of univalent ions across the lamellae of swollen phospholipids. Journal of Molecular Biology 13, 238–252. Boucher, P.A., Joos, B., Zuckermann, M.J., Fournier, L., 2007. Pore formation in a lipid bilayer under a tension ramp: modeling the distribution of rupture tensions. Biophysical Journal 92, 4344–4355. Clerc, S.G., Thompson, T.E., 1994. A possible mechanism for vesicle formation by extrusion. Biophysical Journal 67, 475–477. Deamer, D., Bangham, A.D., 1976. Large volume liposomes by an ether vaporization method. Biochimica et Biophysica Acta 443, 629–634. Elzainy, A.A.W., Gu, X., Simons, F.E.R., Simons, K.J., 2005. Hydroxyzine- and cetirizine-loaded liposomes: effect of duration of thin film hydration, freezethawing, and changing buffer pH on encapsulation and stability. Drug Development and Industrial Pharmacy 31, 281–291. Estes, D.J., Mayer, M., 2005. Electroformation of giant liposomes from spin-coated films of lipids. Colloids and Surfaces B: Biointerfaces 42, 115–123.
Y.P. Patil et al. / Chemistry and Physics of Lipids 165 (2012) 475–481 Evans, E., Heinrich, V., Ludwig, F., Rawicz, W., 2003. Dynamic tension spectroscopy and strength of biomembranes. Biophysical Journal 85, 2342–2350. Frisken, B.J., Asman, C., Patty, P.J., 2000. Studies of vesicle extrusion. Langmuir 16, 928–933. Gruner, S.M., Lenk, R.P., Janoff, A.S., Ostro, M.J., 1985. Novel multilayered lipid vesicles: comparison of physical characteristics of multilamellar liposomes and stable plurilamellar vesicles. Biochemistry 24, 2833–2842. Harashima, H., Sakata, K., Funato, K., Kiwada, H., 1994. Enhanced hepatic uptake of liposomes through complement activation depending on the size of liposomes. Pharmaceutical Research 11, 402–406. Hope, M.J., Bally, M.B., Mayer, L.D., Janoff, A.S., Cullis, P.R., 1986. Generation of multilamellar and unilamellar phospholipid vesicles. Chemistry and Physics of Lipids 40, 89–107. Hope, M.J., Bally, M.B., Webb, G., Cullis, P.R., 1985. Production of large unilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution, trapped volume and ability to maintain a membrane potential. Biochimica et Biophysica Acta—Biomembranes 812, 55–65. Hunt, C.A., Rustum, Y.M., Mayhew, E., Papahadjopoulos, D., 1979. Retention of cytosine arabinoside in mouse lung following intravenous administration in liposomes of different size. Drug Metabolism and Disposition 7, 124–128. Hunter, D.G., Frisken, B.J., 1998. Effect of extrusion pressure and lipid properties on the size and polydispersity of lipid vesicles. Biophysical Journal 74, 2996–3002. Johnson, S.M., Bangham, A.D., Hill, M.W., Korn, E.D., 1971. Single bilayer liposomes. Biochimica et Biophysica Acta 233, 820–826. Jousma, H., Talsma, H., Spies, F., Joosten, J.G.H., Junginger, H.E., Crommelin, D.J.A., 1987. Characterization of liposomes. The influence of extrusion of multilamellar vesicles through polycarbonate membranes on particle size, particle size distribution and number of bilayers. International Journal of Pharmaceutics 35, 214–263. Kim, S., Martin, G.M., 1981. Preparation of cell-size unilamellar liposomes with high captured volume and defined size distribution. Biochimica et Biophysica Acta—Biomembranes 646, 1–9. Koppel, D.E., 1972. Analysis of macromolecular polydispersity in intensity correlation spectroscopy—method of cumulants. Journal of Chemical Physics 57, 4814. Lasic, D.D., 1993. Liposomes: From Physics to Application. Elsevier Science B.V., Amsterdam. Lee, W., Walker, L.M., Anna, S.L., 2009. Role of geometry and fluid properties in droplet and thread formation processes in planar flow focusing. Physics of Fluids, 21.
481
Mayer, L.D., Hope, M.J., Cullis, P.R., 1986. Vesicles of variable sizes produced by a rapid extrusion procedure. Biochimica et Biophysica Acta—Biomembranes 858, 161–168. Ohsawa, T., Miura, H., Harada, K., 1984. A novel method for preparing liposome with a high capacity to encapsulate proteinous drugs: freeze-drying method. Chemical and Pharmaceutical Bulletin 32, 2442–2445. Patty, P.J., Frisken, B.J., 2003. The pressure-dependence of the size of extruded vesicles. Biophysical Journal 85, 996–1004. Perkins, W.R., Minchey, S.R., Ahl, P.L., Janoff, A.S., 1993. The determination of liposome captured volume. Chemistry and Physics of Lipids 64, 197–217. Popa, I., Trulsson, M., Papastavrou, G., Borkovec, M., Jonsson, B., 2009. Long-ranged attractive forces induced by adsorbed dendrimers: direct force measurements and computer simulations. Langmuir 25, 12435–12438. Selser, J.C., Yeh, Y., Baskin, R.J., 1976. Light-scattering characterization of membranevesicles. Biophysical Journal 16, 337–356. Stan, C.A., Tang, S.K.Y., Whitesides, G.M., 2009. Independent control of drop size and velocity in microfluidic flow-focusing generators using variable temperature and flow rate. Analytical Chemistry 81, 2399–2402. Storm, G., Koppenhagen, F., Heeremans, A., Vingerhoeds, M., Woodle, M.C., Crommelin, D.J.A., 1995. Novel developments in liposomal delivery of peptides and proteins. Journal of Controlled Release 36, 19–24. Sun, B., Chiu, D.T., 2005. Determination of the encapsulation efficiency of individual vesicles using single-vesicle photolysis and confocal single-molecule detection. Analytical Chemistry 77, 2770–2776. Szoka Jr., F., Papahadjopoulos, D., 1978. Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proceedings of the National Academy of Sciences of the United States of America 75, 4194–4198. Vemuri, S., Rhodes, C.T., 1995. Preparation and characterization of liposomes as therapeutic delivery systems: a review. Pharmaceutica Acta Helvetiae 70, 95–111. Vuillemard, J.C., 1991. Recent advances in the large-scale production of lipid vesicles for use in food products: microfluidization. Journal of Microencapsulation 8, 547–562. Woodle, M.C., 1995. Sterically stabilized liposome therapeutics. Advanced Drug Delivery Reviews 16, 249–265. Zhang, L., Hu, J., Lu, Z., 1997. Preparation of liposomes with a controlled assembly procedure. Journal of Colloid and Interface Science 190, 76–80. Zhang, X.M., Patel, A.B., de Graaf, R.A., Behar, K.L., 2004. Determination of liposomal encapsulation efficiency using proton NMR spectroscopy. Chemistry and Physics of Lipids 127, 113–120.