Population Analysis of Structural Properties of Giant Liposomes by

Aug 11, 2009 - Department of Bioinformatic Engineering, Graduate School of Information Science and Technology, Osaka University, 1-5 Yamadaoka, Suita,...
1 downloads 9 Views 3MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

Population Analysis of Structural Properties of Giant Liposomes by Flow Cytometry )

Kazuya Nishimura,† Tomohiro Hosoi,‡ Takeshi Sunami,§ Taro Toyota,‡ Masanori Fujinami,‡ Koichi Oguma,‡ Tomoaki Matsuura,† Hiroaki Suzuki,† and Tetsuya Yomo*,†,§, †

)

Department of Bioinformatic Engineering, Graduate School of Information Science and Technology, Osaka University, 1-5 Yamadaoka, Suita, Osaka 565-0871, Japan, ‡Department of Applied Chemistry and Biotechnology, Graduate School of Engineering, Chiba University, 1-33 Yayoi-cho, Inage, Chiba 263-8522, Japan, §ERATO Complex Systems Biology Project, Japan Science and Technology Agency (JST ), 3-8-1 Komaba, Meguro-ku, Tokyo 153-8902, Japan, and Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan Received June 22, 2009. Revised Manuscript Received July 29, 2009 We used fluorescence flow cytometry to analyze the structural properties of populations of giant liposomes formed by different preparation methods. The inner aqueous volumes and nominal membrane surface areas of a large number of individual liposomes were measured simultaneously by using fluorescent markers. We compared these properties of liposomes prepared by the natural swelling method, the freeze-dried empty liposomes method, and the water-in-oil (W/O) emulsion method. A two-dimensional contour distribution map of the inner volume and the nominal surface area was used to elucidate the structural properties of liposomes over a wide range of liposome sizes. Lamellarity of liposomes was evaluated as the ratio of the nominal surface area to the theoretical surface area calculated from the liposome inner volume. This population analysis revealed the dependency of lamellarity on liposome volume: while the nominal surface areas of populations of liposomes prepared by the natural swelling and the freeze-dried empty liposome methods were widely distributed, those prepared by the W/O emulsion method had a narrower distribution within small values. Furthermore, with the latter method, the nominal surface area varied in proportion to the two-thirds power of the inner volume ranging for several orders of magnitude, indicating the liposomes had a thin membrane, which was constant for the wide volume range. The results as well as the methodology presented here would be useful in designing giant liposomes with desired properties.

1. Introduction Liposomes have been studied extensively as carriers in drug targeting systems, as cell-sized bioreactors, and as model cell membranes. The structural properties of liposome populations, however, may vary widely depending on the materials and the preparation methods, because multiple free energy minima exist in self-assembling of amphiphilic molecules.1 To date, numerous liposome preparation methods have been proposed, and the physical properties of resultant liposomes have been extensively investigated with regard to material encapsulation,2-5 stability,6 membrane permeability,7,8 ease of preparation,9 and specific affinity for targeting.10 The most fundamental physical properties of liposomes are their size and lamellarity, and liposomes are thus generally categorized as unilamellar or multilamellar and as either small *To whom correspondence should be addressed. Tel: þ81-6-6879-4171. Fax: þ81-6-6879-7433. E-mail: [email protected].

(1) Laughlin, R. G. Colloids Surf., A 1997, 128, 27–38. (2) Szoka, F.; Papahadjopoulos, D. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 4194–4198. (3) Kirby, C.; Gregoriadis, G. BioTechnology 1984, 2, 979–984. (4) Monnard, P. A.; Oberholzer, T.; Luisi, P. Biochim. Biophys. Acta 1997, 1329, 39–50. (5) Walde, P.; Ichikawa, S. Biomol. Eng. 2001, 18, 143–77. (6) van Winden, E. C. A.; Crommelin, D. J. A. J. Controlled Release 1999, 58, 69–86. (7) Liburdy, R. P.; Tenforde, T. S.; Magin, R. L.; Niesman, M. Biophys. J. 1986, 49, A515–A515. (8) Tamba, Y.; Yamazaki, M. Biochemistry 2005, 44, 15823–15833. (9) Moscho, A.; Orwar, O.; Chiu, D. T.; Modi, B. P.; Zare, R. N. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 11443–11447. (10) Chonn, A.; Cullis, P. R. Adv. Drug Delivery Rev. 1998, 30, 73–83.

Langmuir 2009, 25(18), 10439–10443

(diameter 1 μm). Small liposomes are mainly used in drug delivery systems, but giant liposomes are of particular interest as artificial cell models because their sizes are comparable to those of living cells. Several research groups have demonstrated that complex biochemical reactions can occur in liposomes prepared by the natural swelling method.11-13 This is the most straightforward way of preparing giant liposomes, in which thin dry lamellae of phospholipids gradually swell into giant lipid structures during incubation for several hours in an aqueous environment. Liposomes prepared this way, however, vary widely in size and lamellarity depending on their lipid composition, the ionic strength of the buffer, and the materials encapsulated when the liposomes form. The freezedried empty liposomes (FDEL) method3,14 features efficient encapsulation of complex and dense suspensions containing even large biomolecules,15,16 but the liposomes formed this way are highly multilamellar.17,18 The water-in-oil (W/O) emulsion (11) Kaneko, T.; Itoh, T. J.; Hotani, H. J. Mol. Biol. 1998, 284, 1671–1681. (12) Nomura, S.; Tsumoto, K.; Hamada, T.; Akiyoshi, K.; Nakatani, Y.; Yoshikawa, K. ChemBioChem 2003, 4, 1172–1175. (13) Tsumoto, K.; Nomura, S. M.; Nakatani, Y.; Yoshikawa, K. Langmuir 2001, 17, 7225–7228. (14) Kikuchi, H.; Suzuki, N.; Ebihara, K.; Morita, H.; Ishii, Y.; Kikuchi, A.; Sugaya, S.; Serikawa, T.; Tanaka, K. J. Controlled Release 1999, 62, 269–277. (15) Ishikawa, K.; Sato, K.; Shima, Y.; Urabe, I.; Yomo, T. FEBS Lett. 2004, 576, 387–390. (16) Sunami, T.; Sato, K.; Matsuura, T.; Tsukada, K.; Urabe, I.; Yomo, T. Anal. Biochem. 2006, 357, 128–136. (17) Sato, K.; Obinata, K.; Sugawara, T.; Urabe, I.; Yomo, T. J. Biosci. Bioeng. 2006, 102, 171–178. (18) Hosoda, K.; Sunami, T.; Kazuta, Y.; Matsuura, T.; Suzuki, H.; Yomo, T. Langmuir 2008, 24, 13540–13548.

Published on Web 08/11/2009

DOI: 10.1021/la902237y

10439

Letter

Nishimura et al.

method19,20 has recently attracted attention because it enables suspensions of materials to be encapsulated in unilamellar liposomes without changing the concentration of the suspended material. Liposome preparation protocols are usually selected empirically because the properties of the resultant liposomes are often unpredictable. The engineering of liposomes with specific properties would thus be facilitated by a facile method for characterizing liposome structure. The size distribution of liposomes less than a micrometer in diameter is usually evaluated by dynamic light scattering measurement,21-23 but this method cannot reveal the lamellarity. The size and lamellarity of giant liposomes can be estimated from the images obtained by optical or electron microscopes,22,24 but a statistical understanding of wide size distributions is not readily obtained from such images. Flow cytometry has been proven to be a powerful tool for evaluating various properties of liposomes because it can obtain multiple scattering and fluorescence signals from a large number of individual liposomes simultaneously.25-29 In this study, we examined the structural properties of giant liposomes prepared using different methods. In order to gain statistical and high-dimensional information about liposome size and lamellarity, we used fluorescence flow cytometry to measure the liposome surface area and volume, which were respectively evaluated from the intensities of fluorescent marker molecules embedded in the lipid membrane and the enclosed solution. Their correlation was analyzed based on the method we reported previously.17 We found that each liposome population formed by a different preparation method showed a characteristic volume-area distribution. We also found that the W/O emulsion method yielded liposomes that had a thin membrane down to volumes as small as 1 fL, whereas the natural swelling method, which had been reported to produce giant unilamellar liposomes, yielded those with a thin membrane only when their volume was greater than 10 fL.

2. Experimental Section Materials. 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1glycerol)] (POPG) were purchased from Avanti Polar Lipids (Alabaster, AL), cholesterol was purchased from Nacalai Tesque (Japan), and distearoylphosphatidylethanolamine-poly(ethylene glycol) 5000 (DSPE-PEG5000) was kindly supplied by NOF Corporation (Japan). Red fluorescent protein, allophycocyanin (APC), and green fluorescent lipid, 2-(4,4-difluoro-5,7dimethyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoyl)-1-hexadecanoyl-sn-glycero-3-phosphocholine (BODIPY-HPC) were purchased from Invitrogen (Carlsbad, CA). Protease from Streptomyces griseus was purchased from Sigma-Aldrich Corporation (19) Pautot, S.; Frisken, B. J.; Weitz, D. A. Langmuir 2003, 19, 2870–2879. (20) Noireaux, V.; Libchaber, A. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 17669– 17674. (21) Maulucci, G.; De Spirito, M.; Arcovito, G.; Boffi, F.; Castellano, A. C.; Briganti, G. Biophys. J. 2005, 88, 3545–3550. (22) Egelhaaf, S. U.; Wehrli, E.; Muller, M.; Adrian, M.; Schurtenberger, P. J. Microscopy 1996, 184, 214–228. (23) Woodbury, D. J.; Richardson, E. S.; Grigg, A. W.; Welling, R. D.; Knudson, B. H. J. Liposome Res. 2006, 16, 57–80. (24) Akashi, K.; Miyata, H.; Itoh, H.; Kinosita, K. Biophys. J. 1996, 71, 3242– 3250. (25) Vorauer-Uhl, K.; Wagner, A.; Borth, N.; Katinger, H. Cytometry 2000, 39, 166–171. (26) Zazueta, C.; Ramirez, J.; Garcia, N.; Baeza, I. J. Membr. Biol. 2003, 191, 113–122. (27) Pozharski, E. V.; MacDonald, R. C. Anal. Biochem. 2005, 341, 230–240. (28) Kageyama, Y.; Toyota, T.; Murata, S.; Sugawara, T. Soft Matter 2007, 3, 699–702. (29) Toyota, T.; Takakura, K.; Kageyama, Y.; Kurihara, K.; Maru, N.; Ohnuma, K.; Kaneko, K.; Sugawara, T. Langmuir 2008, 24, 3037–3044.

10440 DOI: 10.1021/la902237y

(St. Louis, MO). The aqueous buffer we used had a pH of 7.6 and, unless otherwise noted, consisted of 50 mM Hepes-KOH, 13 mM Mg(OAc)2, and 100 mM potassium glutamate. Preparation of Giant Liposomes. We prepared giant liposomes consisting of almost identical lipid and buffer compositions using three different liposome preparation methods. In each, the lipid mixture contained 0.01% BODIPY-HPC (molar ratio), and the aqueous buffer to be encapsulated contained 500 nM APC. Natural Swelling Method. After POPC, POPG, and cholesterol were mixed in chloroform in a round flask at molar proportions of 83:0:17 or 75:8:17, a rotary evaporator was used to remove the chloroform and form a film of 0.13 μmol lipids. Then 250 μL of the buffer containing APC was added gently. As it is well known that, with this method, giant unilamellar liposomes are likely to form at the low ionic concentration,30 we used 5-fold diluted buffer (10 mM Hepes-KOH, 2.6 mM Mg(OAc)2, and 20 mM potassium glutamate). Giant liposomes were made by incubating the suspension overnight at 37 °C,24 and multilamellar liposomes were made by vortexing the suspension for 10 s 10 min after adding the buffer. FDEL Method. Freeze-dried liposomes were prepared according to the protocol used in our previous reports.16-18 Briefly, a mixture of POPC, cholesterol, and DSPE-PEG5000 dissolved in chloroform at molar proportions of 58:39:3 (1.2 μmol total) was subjected to rotary evaporation in a round flask under reduced pressure to yield a thin lipid film. DSPE-PEG5000 was included to obtain a highly porous structure of freeze-dried liposomes for realizing high encapsulation efficiency. Fifteen minutes after 1 mL of deionized water was added under argon gas, the lipid film was vortexed in order to disperse the liposomes. The liposome dispersion was then homogenized on ice by sonication with an ultrasonic disrupter (Tomy Seiko, Japan), aliquoted in tubes (40 μL each), and lyophilized in a freeze-dryer (Labconco, Kansas City, MO). Finally, liposomes were obtained by adding 10 μL of buffer containing APC. W/O Emulsion Method. We used a modified version of the protocol presented by Noireaux et al.20 The mixture of POPC, POPG, and cholesterol dissolved in 100 μL of chloroform at molar proportions of 83:0:17 or 75:8:17 was mixed with 2 mL of liquid paraffin and heated at 80 °C for 30 min in order to completely dissolve the lipids and evaporate the chloroform. Five hundred microliters of this solution was then transferred to a glass tube to which was added 50 μL of the buffer containing 150 mM sucrose, 350 mM glucose, and APC (inner solution). This mixture was vortexed for 10 s to form a W/O emulsion that was then equilibrated on ice for 10 min. Four hundred microliters of this emulsion was gently placed on top of 400 μL of buffer containing 500 mM glucose (outer solution) in a new tube. When this tube was centrifuged at 18 000 g (14 000 rpm) for 30 min at 4 °C, emulsions passed through the oil/water interface saturated by lipids to form a bilayer structure, since there is a difference in the relative density between the inner and outer solutions. The precipitated liposomes were collected through a hole opened at the bottom of the tube. In each preparation procedure, background fluorescence of APC in the solution not encapsulated was completely eliminated by adding 10 μL of buffer containing protease (1 mg/mL) to 90 μL of liposome suspension and incubating the suspension at 37 °C for 1 h.

Fluorescence-Activated Cell Sorting (FACS) Analysis. The fluorescence intensities due to APC and BODIPY-HPC (respectively FIR and FIG) in individual liposomes were measured simultaneously by using a FACSAria flow cytometer (BD Biosciences, San Jose, CA). Before the measurement, the liposome dispersion was further diluted with an isotonic buffer (FACS Flow, BD Biosciences) in order to reduce the data sampling rate to less than 20 000 events/s. In each measurement, we obtained (30) Mueller, P.; Chien, T. F.; Rudy, B. Biophys. J. 1983, 44, 375–381.

Langmuir 2009, 25(18), 10439–10443

Nishimura et al.

Letter

50 000 data points and used them for the analysis. APC was excited with a HeNe laser (633 nm), and the emission was detected through a filter passing wavelengths between 650 and 670 nm. The number of APC molecules NAPC was calculated from FIR by using the calibration curve relating the intensities from the beads carrying a known amount of another fluorescent protein R-phycoerythrin (R-PE) (QuantiBRITE PE Quantitation kit; BD Biosciences Clontech, Palo Alto, CA) and APC; i.e., NAPC = FIR/(c1 3 c2), where c1 and c2 are the intensity per single R-PE molecule (1.16 intensity/molecule) and the intensity ratio of APC to R-PE (1.40), respectively. The liposome aqueous volume Vw was then calculated as Vw = NAPC /(NA 3 CAPC 3 10-15) = 0.0020FIR fL, where NA is Avogadro’s number and CAPC is the APC concentration (500 nM). BODIPY-HPC, the fluorescent phospholipid, was excited with a 488-nm semiconductor laser, and its emission was detected through a filter passing wavelengths between 515 and 545 nm. The number of BODIPY-HPC molecules NHPC was calculated from FIG by using the calibration curve relating the intensities from the beads carrying GFP (BD Living Colors EGFP Calibration Beads; BD Biosciences Clontech) and BOPIPY-HPC; i.e., NHPC =FIG/(c3 3 c4), where c3 and c4 are the intensity per single GFP molecule (4.2  10-2 intensity/molecule) and the intensity ratio of BODIPY-HPC to GFP (0.98), respectively. We then estimated the nominal surface area Sn (μm2) of the lipid membrane. By using the reported values of the surface area of single POPC (or POPG) and cholesterol molecules (respectively 0.65 and 0.28 nm2),31,32 we have   1 NHPC rchol Sn ¼   0:65 þ NHPC   0:28  10 -6 ð1Þ 2 rHPC rHPC where rHPC and rchol are respectively the molar ratios of BODIPYHPC and cholesterol to the phospholipid. The coefficient 1/2 corresponds to the fact that lipid layers are assembled in the bilayer form. The first and second terms in parentheses are respectively the surface areas of the phospholipids and the cholesterol. Linearity between the amounts of fluorescent markers and the intensities in the measurement range was confirmed by introducing these markers in different concentrations (Supporting Information, Figure S1). We also confirmed that the shape of the volume and area distributions did not change for different marker concentrations, which proves that incorporation of the marker molecules did not affect the assembly of liposomes. That is, it did not affect their sizes and lamellarities. Microscope Observation. Microscope images were obtained using an inverted light microscope (IX70; Olympus, Japan) with 40 or 100 oil-immersion objectives and a digital color chargecoupled device (CCD) camera (VB-7000, Keyence, Japan). Bright-field images were obtained through the differential interference contrast observation. Fluorescence images of the red and green marker molecules APC and BODIPY-HPC were obtained through corresponding filter and dichroic mirror units (respectively Cy5-4040A, excitation 608-648 nm/emission 672-712 nm, Semrock, USA, and NIBA, excitation 470-490 nm/emission 510-550 nm, Olympus, Japan).

3. Results and Discussion We measured the nominal membrane surface area and inner aqueous volume of liposomes formed by three methods. In the natural swelling method, a dried lipid film is hydrated to swell into liposomes. This usually produces multilamellar liposomes, but modified versions of this method are reported to be able to produce giant unilamellar liposomes with diameters up to tens of micrometers.24,33,34 Figure 1a-c shows the two-dimensional (31) Lantzsch, G. J. Fluoresc. 1994, 4, 339–343. (32) Hofsass, C.; Lindahl, E.; Edholm, O. Biophys. J. 2003, 84, 2192–2206. (33) Yamada, N. L.; Hishida, M.; Seto, H.; Tsumoto, K.; Yoshimura, T. Europhys. Lett. 2007, 80, 48002. (34) Tsumoto, K.; Matsuo, H.; Tomita, M.; Yoshimura, T. Colloids Surf., B: Biointerfaces 2009, 68, 98–105.

Langmuir 2009, 25(18), 10439–10443

(2D) contour maps (plots of inner aqueous volume Vw versus nominal membrane surface area Sn, both on a logarithmic scale) of the frequency distribution of 50 000-liposome populations formed by the natural swelling method under different conditions. Figure 1a shows the data for multilamellar liposomes formed by vortexing (POPC/cholesterol=83:17). In this liposome population, the most frequent value (mode) of Vw was 3-4 fL, which would correspond to about 2 μm in diameter (assuming the shape was spherical). The mode of Sn was 100-200 μm2. To understand the relation between these values, in each of these maps we drew a theoretical line relating the volume and surface area of a sphere with a unilamellar bilayer shell. The equation of this line is Vw = (1/6) 3 π-1/2 3 Sn3/2 or, on a logarithmic scale, (log10 Vw) = 1.5  (log10 Sn) - 1.027. We refer to this line as the unilamellar vesicle line. In each map we also drew a dotted line with a slope of 1.5 that passes through the most frequent population. The horizontal distance between the solid and dotted lines represents the nominal lamellarity for the most frequent population. In Figure 1a, for example, the x intercept of the dotted line is larger than that of the unilamellar vesicle line by a factor of approximately 10, indicating the nominal number of the lamellae is 10 on average. Note that this nominal lamellarity does not necessarily mean that the membrane consisted of 10 lamellae. A nominal lamellarity greater than 1 can also be due to aggregation or to a complex internal structure. The corresponding bright-field and green fluorescence micrographs of a representative liposome show the thick and complex membrane structure. Figure 1b shows the data for a population formed when a lipid film (POPC/cholesterol=83:17) was gently hydrated and incubated overnight. Long-time incubation promotes the formation of giant liposomes with thinner membranes.30 In this case, relatively large liposomes (Vw>10 μm3) were more frequent, but the mode Vw was unchanged. The Sn of the most frequent liposomes, however, decreased slightly to ∼70 μm2, resulting in the corresponding nominal lamellarity of 7. Moreover, at Vw>5 fL, the distribution of the contour map was extended parallel to the line with a slope of 1.5, indicating that the nominal lamellarity was constant in this range of the liposome size. As shown in Figure 1c for POPC/POPG/cholesterol = 75:8:17, adding the negatively charged phospholipid POPG to the lipid mixture increased the most frequent Vw to ∼10 μm3 and decreased the lamellarity slightly to ∼6. Adding a small amount of charged phospholipid has been reported to facilitate the formation of giant unilamellar liposomes, even when there is salt in the solution.24 In the present result, although the nominal lamellarity of the most frequent population was not unity, a ridge closer to the unilamellar vesicle line appeared in the range of Vw larger than 10 fL (dashed line). The nominal lamellarity of the population along this line is estimated to be ∼4, and the population distributed above this line has smaller nominal lamellarity. Thus, production of giant liposomes in the simple hydration method with the negatively charged lipids was also confirmed in the present measurement, but we found that most of the population, especially smaller than Vw =10 fL, has multiple lamellae. We next measured liposomes prepared by the FDEL method (Figure 1d), which are known to have highly complex and heterogeneous membrane structures. The contour map showed the expected broad distribution of Sn values, and the Vw and nominal lamellarity of the most frequent liposomes were respectively ∼1 fL and 20-30. Micrographs of the liposomes also showed that the membrane was thick. Lastly, we measured the liposomes formed by the W/O emulsion method, in which liposomes are formed by passing the W/O emulsion template through an oil-and-water interface that was equilibrated with the phospholipids. In the contour DOI: 10.1021/la902237y

10441

Letter

Nishimura et al.

Figure 1. Contours maps of the relative frequency of liposome populations produced using various methods: (a) natural swelling followed by vortexing (POPC/cholesterol=83:17), (b) natural swelling with overnight incubation (POPC/cholesterol=83:17), (c) natural swelling with overnight incubation (POPC/POPG/cholesterol =75:8:17), (d) FDEL method (POPC/cholesterol/DSPE-PEG5000=58:39:3), (e) W/O emulsion method (POPC/cholesterol = 83:17), and (f) W/O emulsion method (POPC/POPG/cholesterol = 75:8:17). The vertical and horizontal axes are respectively the aqueous volume Vw and the nominal membrane surface area Sn. Each of the solid lines is the theoretical line for unilamellar vesicles with a bilayer membrane (eq 1). The dashed lines represent the line with a slope of 1.5 that passes through the most frequent population. Microscope images on the right-hand side of each map show corresponding representative differential-interferencecontrast, green-fluorescence (lipid membrane), and red-fluorescence (inner volume) images (scale bar=5 μm).

map shown in Figure 1e, for liposomes with a membrane consisting of PC and cholesterol, it is clear that the distribution of Sn is of very narrow width and is parallel to the unilamellar vesicle line over a wide range of Vw (1-100 fL). Fluorescence micrographs confirmed that the liposomes had a thin membrane, and the inner volume was filled with the red fluorescent protein, showing that there was no complex lipid structure inside the shell. When approximately 10% of PG was added to the lipid component, the liposome size increased slightly, but the distribution was still close to the unilamellar vesicle line. The nominal lamellarities of the most frequent population in these preparations were statistically between 2 and 3. However, liposomes formed by the W/O emulsion method should be, in principle, and is reported to be, unilamellar.19 We think the overestimation of the lamellarity was due to the adherence of the small aggregate or micelle-like lipid structures onto the surface (Supporting Information, Figure S2). If small lipid micelles are distributed uniformly on average over the liposome surface, Sn is still proportional to Vw2/3. To directly compare the lamellarity between the preparation methods, we calculated the nominal lamellarity averaged over the entire range of Vw. This parameter is defined as the ratio of Sn to the theoretical surface area calculated from Vw, assuming the liposomes to be unilamellar spheres. That is, it is defined as Sn/Sth, where Sth = π 1/3 3 (6Vw)2/3. None of the distributions shown in Figure 2 have nonzero relative frequencies at Sn/Sth values smaller than 1, which is in accordance with the physical fact that the membrane cannot be thinner than a single bilayer. Most of the liposomes formed by the W/O emulsion method have Sn/Sth values between 2 and 3. Liposomes formed by the natural swelling method consisting of PC and PG also have a population with thin 10442 DOI: 10.1021/la902237y

Figure 2. Relative frequency distributions of the averaged nominal lamellarities of various liposome populations.

membrane (Sn/Sth