SNo

O*CE

R

cwedtfcsg1mtcmscwfig©

dcm

mms

2

0CA

Archives of Biochemistry and BiophysicsVol. 367, No. 2, July 15, pp. 153–160, 1999Article ID abbi.1999.1267, available online at http://www.idealibrary.com on

olubilization of Liposomes by Sodium Dodecyl Sulfate:ew Mechanism Based on the Direct Formationf Mixed Micelles

lga Lopez,* Merce Cocera,* Ernst Wehrli,† Jose Luis Parra,* and Alfonso de la Maza*,1

Departamento de Tensioactivos, Centro de Investigacion y Desarrollo (C.I.D.), Consejo Superior de Investigacionesientıficas (C.S.I.C.), C/ Jordi Girona 18-26, 08034 Barcelona, Spain; and †Laboratory for Electron Microscopy I,TH-Zentrum, Universitatsstrasse 16, CH-8092 Zurich, Switzerland

eceived February 12, 1999, and in revised form April 27, 1999

snsTpmf1fpfamts(aso

tcttnfna

fPlr

The vesicle-to-micelle structural transitions that oc-urred in the interaction of sodium dodecyl sulfateith phosphatidylcholine vesicles were studied at the

quilibrium by means of dynamic light scattering (atifferent scattering angles) and freeze–fracture elec-ron microscopy techniques. The incorporation of sur-actant monomers in the bilayers resulted in an initialontraction of the mixed vesicles formed up to theiraturation (size reduction of about 10%). Then, a pro-ressive relaxation of these structures (growth from70 to 225 nm) and a simultaneous formation of mixedicelles (particles of about 6 nm) occurred. Hence, in

his interval “relaxed mixed vesicles” and mixed mi-elles coexisted in different proportions without for-ation of intermediate complex aggregates (bimodal

ize distribution curves). Freeze–fracture electron mi-roscopy showed a direct formation of mixed micellesithin the bilayer and their subsequent separation

rom the vesicle surface without formation of complexntermediate aggregates. This simple process pro-ressed up to the complete vesicle solubilization.1999 Academic Press

Key Words: phosphatidylcholine liposomes; sodiumodecyl sulfate; liposome solubilization; vesicle to mi-elle structural transitions; freeze–fracture electronicroscopy; dynamic light scattering.

The solubilization and reconstitution of biologicalembranes using surfactants is currently attractinguch interest (1–3). In this sense, the interaction of

urfactant with liposomes represents a good model for

1

To whom correspondence should be addressed. Fax: (34-93)04.59.04.

tm

003-9861/99 $30.00opyright © 1999 by Academic Pressll rights of reproduction in any form reserved.

tudying the solubilization of cell membranes (4–7). Aumber of complex lipid/surfactant aggregates are as-ociated to the vesicle to micelle transformations.hus, open bilayer fragments in coexistence with com-lex micellar structures have been reported as inter-ediate aggregates in the interaction of various sur-

actants with phosphatidylcholine (PC)2 liposomes (8–0). However, recent studies using vesicles preparedrom nonionic and oppositely charged surfactants pro-osed rapid and simpler mechanisms for these trans-ormations as single-step processes (11–15). The inter-ction of sodium dodecyl sulfate (SDS) with simplifiedembrane models such as PC liposomes has been ex-

ensively studied given the frequent use of this anionicurfactant as irritant agent in biological membranes16–18). Despite the fact that the kinetics of this inter-ction have recently been published (19), a detailedtudy of the vesicle to micelle transformations thatccurred at the equilibrium is still lacking.We first studied the interaction of different surfac-

ants with PC liposomes (20–24). Here, we seek toharacterize in detail the vesicle to micelle structuralransitions involved in the interaction of SDS surfac-ant with PC liposomes using a combination of dy-amic light scattering (DLS) (Ar laser source) andreeze–fracture electron microscopy (FFEM) tech-iques. The use of these two techniques for measuringnd visualizing small and large particles in the same

2 Abbreviations used: DLS, dynamic light scattering; FFEM,reeze–fracture electron microscopy; HD, hydrodynamic diameter;C, phosphatidylcholine; PI, polydispersity index; ReSAT, surfactant/

ipid molar ratio for bilayer saturation; ReSOL, surfactant/lipid molaratio for bilayer solubilization; SDS, sodium dodecyl sulfate; TEM,

ransmission electron microscopy; TX-100, Triton X-100; CMC, criticalicelle concentration.

153

soa

M

bt(Twgtmcs(

duttzhtt

aflpectmorldtMwfwfiav

2pwgnaTsok

R

D

m

fbasasntscrStvibibtvcnlccI(l(tm

m

154 LOPEZ ET AL.

ample may shed light on the transformations thatccurred in this interaction that are so important frombiological viewpoint.

ATERIALS AND METHODS

PC was purified from egg lecithin (Merck, Darmstadt, Germany)y the method of Singleton (25) and was shown to be pure byhin-layer chromatography. Tris(hydroxymethyl)aminomethaneTris) was obtained from Merck. Tris buffer was prepared as 5.0 mMris adjusted to pH 7.4 with HCl, containing 100 mM of NaCl. SDSas obtained from Merck and further purified by a column chromato-raphic method (26). The surface tensions of buffered solutions con-aining increasing amounts of SDS were measured by the plateethod using a Kruss tensiometer. The surfactant critical micelle

oncentration (CMC) was determined from the abrupt change in thelope of the surface tension values versus surfactant concentration27) showing a value of 0.75 mM.

Liposome preparation and solubilization by SDS. Liposomes of aefined size (about 200 nm) were prepared by extrusion of largenilamellar vesicles obtained by means of a reverse phase evapora-ion technique (23) followed by a total of 30-fold passage through 800-o 200-nm polycarbonate membranes. In order to study the solubili-ation of PC liposomes by SDS surfactant concentrations lower andigher than its CMC in Tris buffer (from 0.5 to 10 mM) were addedo the liposomes (PC concentration 0.5 mM) and the resulting mix-ures were left to equilibrate for 24 h at 20°C (24).

Dynamic light scattering measurements. The hydrodynamic di-meter (HD) of pure PC vesicles, pure SDS micelles, and particlesormed 24 h after mixing different concentrations of SDS with PCiposomes were determined by means of a DLS technique using ahoton correlator spectrometer (Malvern Autosizer 4700c PS/MV)quipped with an Ar laser source (wavelength, 488 nm). Quartzuvettes were filled with the samples and all the experiments werehermostatically controlled at 25°C. All the DLS measurements wereade with a scattering angle of 90° and some of them were carried

ut using other angles (60° and 120°). In order to acquire the fullange of decay times necessary to determine the signal from both thearge and the small particles, a low sample time value (2 ms) and ailatation of 3 with parallel subcorrelators was used. The analysis ofhe data was performed using CONTIN software provided byalvern Instruments (England). The validity of the CONTIN resultsas tested by fitting a single or a biexponential to the correlation

unction. If a biexponential had to be fitted, first a single exponentialas fitted to a long time range and the second exponential was thentted to the residual. Both methods agreed fairly well. The resultsre given as diameters and the percentages correspond to intensityalues.Freeze–fracture electron microscopy. A FFEM study was done

4 h after mixing PC liposomes and surfactant according to therocedure described by Egelhaaf et al. (28). About 1 ml of suspensionas sandwiched between two copper platelets using a 400-mesh goldrid as the spacer. Then, the samples were cryofixed by dipping intoitrogen-cooled liquid propane at 2180°C and fractured at 2110°Cnd 5 3 1027 mbar in a BAF-060 freeze–etching apparatus (BAL-EC, Liechtenstein). The replicas were obtained by unidirectionalhadowing with 2 nm of Pt/C and 20 nm of C, and they were floatedn distilled water and examined in a Hitachi H-600AB TEM at 75V.

ESULTS AND DISCUSSION

ynamic Light Scattering Experiments

The DLS curves (at a scattering angle of 90°) oficellar SDS solutions (SDS concentration ranging

ml

rom 1 to 10 mM) always showed a monomodal distri-ution with a peak at 4 nm. The DLS data (scatteringngle of 90°) for pure PC vesicles and for some repre-entative surfactant/PC systems (PC concn, 0.5 mM)re given in Table I. Pure PC liposomes (sample 1)howed a monomodal distribution with an HD of 190m (PI 0.134). The addition of low SDS concentrationso liposomes first led to a progressive decrease in theize of surfactant–lipid vesicles formed (formation ofontracted mixed vesicles) until a minimun HD waseached for sample 4 (peak at 170.8 nm for 1.2 mMDS). This minimun HD corresponded to the surfac-ant/lipid molar ratio for bilayer saturation (ReSAT) pre-iously reported for this system by static light scatter-ng measurements (21). This vesicle contraction maye attributable to the electrostatic effect due to thencorporation of charged surfactant monomers into PCilayers, in line with the findings reported for the in-eraction of ionic surfactants with dimeric amphiphileesicles (13). However, this contraction process sharplyontrasts with the growth of PC vesicles reported foronionic and anionic surfactants using DLS (He–Ne

aser source) and cryo-transmission electron micros-opy (cryo-TEM) techniques (8, 9). When the SDS con-entration achieved a value of 1.5 mM (sample 5, Table), a new peak in the size distribution curve appeared6.0 nm) corresponding to the formation of surfactant–ipid mixed micelles. Increasing surfactant amountsfrom sample 5 to sample 9) led to a progressive rise inhe proportion of mixed micelles and to a fall in that forixed vesicles.It is noteworthy that in this interval the HD of mixedicelles remained almost unaltered, whereas that for

TABLE I

DLS Data for a Scattering Angle of 90° for Pure PCVesicles (Sample 1) and Mixed Vesicles Corresponding tothe Interaction of SDS with PC Liposomes Varying theSDS Concentration from 0.5 to 10 mM (Samples 2–11)

Sample

SDSconcentration

(mM)

Mixedmicelles Vesicles

nm % nm %

1 0 — — 190.0 1002 0.5 — — 178.6 1003 0.9 — — 176.1 1004 1.2 — — 170.8 1005 1.5 6.0 13 174.6 876 2 6.0 16 200.1 847 2.5 6.2 40 206.4 608 3 6.3 65 215.3 359 3.5 6.2 91 224.2 9

10 4 6.2 100 — —11 10 5.1 100 — —

ixed vesicles rose to the complete solubilization ofiposomes (formation of “relaxed mixed vesicles”). As a

cmomaHtbtdfsdi(pp

swccpToeus

saTtotatmg

sacsssntdttdmam1H“rsawmsitt(t2

F

stidFps

ttttnmt1it

T

155SOLUBILIZATION OF LIPOSOMES BY SODIUM DODECYL SULFATE

onsequence, in this interval “relaxed vesicles” andixed micelles coexisted in different proportions with-

ut formation of intermediate complex aggregates (bi-odal size distribution curves). These findings are in

greement with those published recently by Kragh-ansen et al. (29), who did not detect variegated struc-

ural morphology in the study of liposome and mem-rane solubilization by surfactants. The SDS concen-ration of 4 mM showed again a monomodalistribution curve for mixed micelles. In the rangerom 4 to 10 mM (samples 10 and 11) a slight fall in theize of mixed micelles occurred (HD from 6.2 to 5.1 nm)ue to the progressive enrichment of these aggregatesn SDS. It may be noted that from 1.5 to 4 mM SDSTable I) a progressive fall and rise in the respectiveroportions of mixed vesicles and mixed micelles tooklace.The total SDS concentration reported for liposome

olubilization and determined by static light scatteringas 1.4 mM (9, 21). It is interesting to note that this

oncentration was clearly less than that needed for theomplete solubilization of the relaxed vesicles (com-lete formation of mixed micelles), which was 4 mM.he strong dependence of the scattered light intensityn the size of particles may explain in part this differ-nce and emphasizes the suitability of the DLS methodsed to discriminate the simultaneous presence ofmall and large particles (28).In order to study the shape of the particles formed, a

eries of DLS experiments was performed at scatteringngles of 60° and 120° for pure PC liposomes (sample 1,able I) and some representative surfactant/PC sys-ems (samples 2, 4, 6, 8, and 9, Table I). The resultsbtained are given in Table II. These data confirmedhe presence of two types of particles for samples 6, 8,nd 9 and only one for samples 1, 2, and 4, as well ashe aforementioned contraction and relaxation of the

TABLE II

DLS Data for Scattering Angles of 60 and 120°Corresponding to Samples 1, 2, 4, 6, 8, and 9 of Table I

Sample

Scattering angle of 60° Scattering angle of 120°

Mixedmicelles Vesicles

Mixedmicelles Vesicles

nm % nm % nm % nm %

1 — — 192.0 100 — — 191.0 1002 — — 202.2 100 — — 170.2 1004 — — 211.3 100 — — 158.0 1006 10.4 17 213.1 83 8.4 14 195.6 868 6.1 84 240.9 16 6.4 87 229.2 139 6.5 93 233.0 7 7.1 94 210.0 6

ixed vesicles formed (specially using a scattering an-le of 120°). The fact that pure PC vesicles exhibited a

nT

imilar size (190–192 nm) regardless of the scatteringngle indicated the spherical structure of these vesi-les. However, the increased variation in the particleize with the scattering angle that occurred at theaturation of vesicles (highest vesicle contraction forample 4) indicated that these contracted vesicles wereot spheroidal exhibiting an elongated shape. It is in-eresting to note that at a scattering angle of 60° theecrease in the HD of the mixed vesicles was not de-ected, but rather an increase. Thus, although the con-raction of mixed vesicles occurred, this effect was notetected at 60° due to the irregular shape of theseixed vesicles. In fact, this large HD could be associ-

ted with the longest axis of the elongated shape ofixed vesicles. Increasing surfactant amounts (above

.2 mM) resulted in a decrease in the variation of theD of large particles. Thus, the elongated shape of the

contracted mixed vesicles” became spheroidal for theelaxed mixed vesicles. The size of mixed micelles forample 6 showed a slight variation with the scatteringngle, whereas for samples 8 and 9 the size variationas negligible. These data indicated that the shape ofixed micelles was first slightly elongated, becoming

pheroidal when the proportion of SDS in the systemncreased. It is noteworthy that the formation of con-racted and relaxed mixed vesicles was not reported inhe kinetic study of the initial period of this interaction19). Hence, these aggregates are the result of struc-ural transformations that occurred at the equilibrium4 h after mixing.

reeze–Fracture Electron Microscopy Observations

FFEM has been shown to be useful for topologicaltudies of liposomes and aggregates formed in the in-eraction of surfactant with lipid membranes withoutntroducing artifacts due to changes in temperature orehydration (7, 24, 30, 31). Figure 1 shows sevenFEM micrographs for the most representative sam-les of Table I and one micrograph for a micellar SDSolution.Particle sizes measured in the micrographs were

ested and shown to belong to the log normal distribu-ions obtained by DLS (Table I) using Student’s t sta-istics. On the assumption that no differences in par-icle size were induced by DLS and FFEM techniques,o significant differences between particle size in theicrographs and in the distribution curves were de-

ected, with a probability level always higher than0%. This percentage means that no evidence of signif-cant differences in particle size obtained with thesewo techniques can be considered statistically.

The TEM micrograph for pure liposomes (sample 1,able I) shows vesicles with a diameter of about 190

m, in agreement with the DLS data of Tables I and II.he micrograph of sample 2 (0.5 mM SDS) shows the

pwocl

mvI

F5 ellam

156 LOPEZ ET AL.

resence of slightly contracted vesicles (lower HD), inhich a slight elongation is observed. The micrographf sample 4 shows vesicles with the highest level of

IG. 1. FFEM micrographs of pure PC liposomes (sample 1 of Tab, 8, 9, and 10 of Table I. The last micrograph corresponds to a micixed micelles with arrows and mixed vesicles with arrowheads.

ontraction. The shape variation observed in the bi-ayer surface is due to the “in situ” formation of mixed w

icelles in the lamellar phase of bilayers. This shapeariation is in agreement with the DLS data of Tablesand II (highest elongation shape).

) and six surfactant/PC systems corresponding to the samples 2, 4,r solution of pure SDS (10 mM). Structures are marked as follows:

le I

The micrograph for sample 5 shows a vesicle alsoith structural enlarging, clear signs of disintegration,

aispmp(

tcmT4

Co

157SOLUBILIZATION OF LIPOSOMES BY SODIUM DODECYL SULFATE

nd the presence of small particles (arrow) correspond-ng to mixed micelles. The micrograph of sample 8hows the progressive vesicle disintegration and theresence of mixed micelles without formation of inter-ediate complex aggregates. The micrograph of sam-

FIG. 1—

le 9 also shows the coexistence of mixed micellesarrows) and relaxed vesicles (arrowheads) without in-

sm

ermediate aggregates. These three last micrographsorroborate the coexistence of mixed vesicles andixed micelles, in agreement with the data of Table I.he image for the system with a SDS concentration ofmM (sample 10, Table I) shows only the presence of

ntinued

mall particles (mixed micelles, see arrow). The lasticrograph corresponds to a SDS micellar solution (10

mdi

it

ipam

Co

158 LOPEZ ET AL.

M), which has been included in order to compare theifferences in the backgrounds of the pictures contain-ng increasing SDS concentrations.

From these results the following mechanism for ves-

FIG. 1—

cle into micelle transformations is proposed based onhe direct formation of mixed micelles. First, a initial

gr

ncorporation of surfactant molecules in the lamellarhase of vesicles occurs with a progressive contractionnd morphological variations of these structures (for-ation of “contracted vesicles”). Second, a subsequent

ntinued

rowth of saturated vesicles takes place (formation ofelaxed vesicles) together with a progressive liberation

oa

pTst

sactu

Co

159SOLUBILIZATION OF LIPOSOMES BY SODIUM DODECYL SULFATE

f the mixed micelles from the lamellar phase to thequeous medium.Comparison of the present results with those re-

orted for the interaction of the nonionic surfactantriton X-100 (TX-100) with PC liposomes (24) shows a

FIG. 1—

imilar solubilization mechanism. However, significa-ive differences in the size of the particles formed

rt

hould be noted. First, the contraction of mixed vesiclest the saturation point observed for electronegativellyharged mixed vesicles (incorporation of SDS) did notake place using TX-100. This fact may be mainly attrib-ted to the electrostatic effect caused by the incorpo-

ntinued

ation of charged SDS molecules. Second, the libera-ion of the mixed micelles from the micellar phase of

lpwwfttimlomtar

A

rTfn

R

1

1

1

1

1

1

1

1

1

1

2

22

2

2

2

22

2

2

3

3

160 LOPEZ ET AL.

iposomes led in the case of the TX-100 to the formation oferforated mixed vesicles with a decrease in their size,hereas in the case of the SDS this process occurredith an increase in the size of the relaxed vesicles

ormed and with changes in the shape of these struc-ures. This opposite effect may be also due to the facthat the liberation of charged SDS mixed micelles maynduce an electrostatic relaxation in the structure of

ixed vesicles. In addition, in the case of TX-100, theiberation of mixed micelles larger (17 nm) than thosef SDS (6 nm) may favor the size decrease of the re-aining vesicles. The different alterations produced by

hese two surfactants in biological membranes may bessociated with these different structural effects,ather than with the mechanisms of solubilization.

CKNOWLEDGMENTS

The FFEM analysis was performed at Labor fur Elektronenmik-oskopie 1, ETH-Zentrum, Zurich, and at the Serveis Cientıfico-ecnicos, Barcelona University. This work was supported by funds

rom DGICYT (Direccion General de Investigacion Cientıfica y Tec-ica) (Prog. No. PB94-0043), Spain.

EFERENCES

1. Aveldano, M. I. (1995) Arch. Biochem. Biophys. 324, 331–343.2. Hanna, R., Brummell, D. A., Camirand, A., Hensel, A., Russell,

E. F., and Maclachlan, G. A. (1991) Arch. Biochem. Biophys. 290,7–13.

3. Jiang, S. S., Fan, L. L., Yang, S. J., Kuo, S. Y., and Pan, R. L.(1997) Arch. Biochem. Biophys. 346, 105–112.

4. Paternostre, M., Meyer, O., Grabielle-Madelmont, C., Lesieur, S.,Ghanam, S., and Ollivon, M. (1995) Biophys. J. 69, 2476–2488.

5. Tsilou, E., Hamel, C. P., Yu, S., and Redmond, T. M. (1997) Arch.Biochem. Biophys. 346, 21–27.

6. Wenk, M. R., and Seelig J. (1997) Biophys. J. 73, 2565–2574.7. Polozova, A. I., Dubachev, G. E., Simonova, T. N., and Barsukov

L. I. (1995) FEBS Lett. 358, 17–22.8. Edwards, K., and Almgren, M. (1992) Langmuir 8, 824–832.9. Silvander, M., Karlsson, G., and Edwards, K. (1996) J. Colloid

Interface Sci. 179, 104–113.

0. Gustafsson, J., Oradd, G., Lindblom, G., Olsson, U., and

Almgren, M. (1997) Langmuir 13, 852–860.

1. Soderman, O., Herrington, K. L., Kaler, E. W., and Miller, D. D.(1997) Langmuir 13, 5531–5538.

2. Seras, M., Edwards, K., Almgren, M., Carlson, G., Ollivon, M.,and Lesieur, S. (1996) Langmuir 12, 330–336.

3. Danino, D., Talmon, Y., and Zana, R. (1997) J. Colloid InterfaceSci. 185, 84–93.

4. O’Connor, A. J., Hatton, T. A., and Bose, A. (1997) Langmuir 13,6931–6940.

5. Friberg S. E., Campbell, S., Fei, L., Yang, H., Patel, R., andAikens, P. A. (1997) Colloid Surfaces A, 129–130, 167–173.

6. Wertz, P. W. (1992) in Liposome Dermatics: Chemical Aspects ofthe Skin Lipid Approach: Griesbach Conference (Braun-Falco,O., Korting, H. C., and Maibach, H., Eds.), pp. 38–43, Springer,Heidelberg.

7. Downing, D. T., Abraham, W., Wegner, B. K., Willman, K. W.,and Marshall, J. L. (1993) Arch. Dermatol. Res. 285, 151–157.

8. Inoue, T. (1996) in Vesicles (Rosoff, M., Ed.), pp. 151–195, Dek-ker, New York.

9. Lopez, O., Cocera, M., Pons, R., Azemar, N., and de la Maza, A.(1998) Langmuir 14, 4671–4674.

0. de la Maza, A., and Parra, J. L. (1995) Arch. Biochem. Biophys.322, 167–173.

1. de la Maza, A., and Parra, J. L. (1995) Langmuir 11, 2435–2441.2. de la Maza, A., and Parra, J. L. (1996) Arch. Biochem. Biophys.

329, 1–8.3. de la Maza, A., and Parra, J. L. (1997) Biophys. J. 72, 1668–

1675.4. Lopez, O., de la Maza, A., Coderch, L., Lopez-Iglesias, C., Wehrli,

E., and Parra, J. L. (1998) FEBS Lett. 426, 314–318.5. Singleton, W. S., Gray, M. S., Brown, M. L., and White, J. L.

(1965) J. Am. Oil Chem. Soc. 42, 53–57.6. Rosen, M. J. (1981) J. Colloid Interface Sci. 79, 587–588.7. Lunkenheimer, K., and Wantke, D. (1981) Colloid Polym. Sci.

259, 354–366.8. Egelhaaf, S. U., Wehrli, E., Muller, M., Adrian, M., and Schur-

tenberger, P. (1996) J. Microsc. 184, 214–228.9. Kragh-Hansen U., le Maire M., and Moller J. V. (1998) Biophys.

J. 75, 2932–2946.0. Spector, M. S., and Zasadzinski, J. A. (1996) Langmuir 12,

4704–4708.1. Warriner, H. E., Keller., S. L., Idziak, S. H. J., Slack, N. L.,

Davidson, P., Zasadzinski, J. A., and Safinya, C. R. (1998) Bio-phys. J. 75, 272–293.