Volume 302, number 2. 138-140 FEBS 11031 Cc 1992 Federation of European Biochcmicul Societies 00145793/921%5.00

May 1992

Effective detergent/lipid ratios in the solubilization of phosphatidylcholine vesicles by Triton X- 100

M. Arkzazu Partearroyo, M. Angeles Urbaneja and FClix M. Gofii

Dqm~rtot I q/’ Biocketw.stt~r*. Uttiwrsii~~ OJ ihe Busqtte Cmtrrry, PO Box 644. 48080 Bilhuo, Spin

Received 9 March 1992: revised version received 30 March 1992

Efkctivc detcrgcnt:lipid ratios (i.e. molar r;~tios in the mixed aggregates, vcsiclcs or miccllcs) have been cstimatcd for the solubilization of phosphatidylcholinr vesicles by Triton X-100. Effective molar ratios are given for both the onset nnd the completion of bilayer solubilization; small unilamcllar. large unilamcllar and multilamrllar vesicles have been used. Effective detcrgent:lipid r;\tios urc independent of phosphnlipid concen-

tration. and their USC ullows a deeper understanding of nrembranc-surf~Ict;Inl interactions.

Surlhctant: Detergent: Membrane solubilization: Liposomc: Triton X.100

I. INTRODUCTION

Detergents are widely used tools in membrane bio- chemistry. Consequently, the interaction between lipid bilayers and surfactants has been the object of consider- able attention [l-3]. Triton X-100 has found widespread application in biological research as a non-ionic deter- gent. Solubilization of phospholipid bilayers by Triton X-100 was first characterized by Dennis et al. (see [2] for a review), who found that a detergent/lipid ratio above 2 was required for solubilization to occur. Previous work from our laboratory has also been devoted to the sub-lylic and lytic effects of Triton X-100 on model membranes [4-71. Lichtenberg and co-workers [8-lo] have established the guidelines for improved quantita- tive studies on membrane-surfactant interaction by de- fining the so-called ‘effective detergent/lipid ratios’ and proposing convenient procedures for their estimation. The effective detergent/lipid ratio. R,, is defined as the detergent/lipid molar ratio in the mixed aggregates, ves- icles or micelles [IO]. R, can be easily calculated by performing solubilization experiments at different lipid concentrations, L, and noting the total detergent con- centration producing solubilization, L+. A relationship exists [IO]:

Ahbrcrbri0rl.s: SW. small unilamcllar vesicles: LUV, large unilemcl- lar vesicles; MLV. multilamellnr vesicles; R,, effective ratio, molar rutio 01’ dctcrgcntilipid in the mixed aggregates (vesicles or micelles): RDSfir. R, value at which the vesicles arc saturated with the detergent; RcSoL, R, value at which the solubilizlltion of the lipid is completed: /I,, detergent concentration in the aqueous phase.

Cmwpotrrkncr address: F.M. Godi, Dcpartmcnt of Biochemistry. Uni*versi~.y of the Basque Country. PO Box 644.45080 Bilbao, Spain. Fax: (34) (4) 464 8500.

DT = R, {L -I- I / [(R, + I)]] (1)

where K is the distribution coefficient of Triton X-100 between the vesicles and the aqueous medium. The equation means that the total detergent concentration, D.,., required for obtaining any effective ratio A,.. has a linear dependence on L, R, is given by the slope of the straight line defined by the equation, which should in- tercept with tile s-axis at -!I[K(R, + l)] . In addition, the intercept with the paxis corresponds to the concen- tration of free detergent in water, D,., which in turn represents the apparent critical micellar concentration of the surfactant in the presence of lipid.

Following these ideas, we have calculated effective Triton X-loo/PC ratios for the solubilization of three commonly used liposome preparations, namely mul- tilamellar vesicles (MLV), small unilamellar vesicles (SUV) and large unilamellar vesicles (LUV). DT values corresponding to both the onset and the completion of solubilization were recorded, from which saturation and solubilization effective ratios (respectively, ReSAT and RcsoL) could be computed.

2. MATERIALS AND METHODS

Triton X-100 (regular, batch No. 63F~0419) was purchased from Sigma (Sl. Louis, MO) and used without further purification, This detergent exhibits some hctcrogcneity in the length of the poly oxyeth- ylenc moiety, which is supposed to increase its solubilization proper- tics [I]. Egg-yolk phosphatidylcholinc was grade 1 from Lipid Prod- ucts (SoA N utficld. England). Lipcsorues were formed in a SO mM Tris-HCI. pH 7.0 buffer. WV were prepared by sonication [4] and LUV by extrusion through 0.1 pm Nucleopore filters [I I]. Liposome suspensions wcrc mixed with the same volumes of the appropriute detcrgcnt solutions (in the same buffer). The samples were left to equilibrate for 24 h at room tcmpcrature, and solubilization was as- scsscd from hc clu~~gcs in light scattering [IZ]. Light scattering was

138

Volume 302, number 2 FEBS LETTERS May 1992

log [Triton x-loo] (Id)

Fig. 1, Dcurmination of D,,, and D,,,, . The percent change in light scnttcring of thrceegg-yolk phosphatidylcholine liposomc preparations (3.5 mM) in the presence of various concentrations of Triton X-100. Arrows 1 and 2 correspond. respectively. to Dan and D,.. Data points are the average

of three indcpcndcnt measurements.

measured at 90°C in a Perkin Elmer LS-50 spcctrofluoromctcr. with both monochromutors adjusted at 500 nm, Total dctergcnt conccntra- tions producing the onset and the completion of solubilization (Da,,, and D,,,,. rcspectivcly) were determined graphically as shown in Fig, I for each kind of liposomc prepuration.

3. RESULTS AND DISCUSSION

When D,,, and Dloo are plotted as a function of lipid concentration, the straight lines predicted by equation 1 arc found (Fig. 2). From those, the corresponding f?, values are readily calculated (Table I). RcsoL, i.e. R,

12

8

4

0 0 6

Fig:. 2. Determination ol’ R, values. D,,, (:;) and D,,,, (0) arc plotted as a function of lipid (EY L) concentration. for each type of liposome preparation; R, values are estimated from the slop+z of the resulting

straight lines.

valuesatwhich thesolubilizationofthelipidiscompleted, are found of around 34 Triton X-100 molecules per phospholipid molecule, independently of the type of iiposome preparation.

A ‘*’ values. i.e. R, at which the vesicles are saturated witl; Triton X-100, are higher for uniiameliar than for multiiamellar vesicles (0.7-0.8 vs. 0. IS). AtSAT values are derived from measurements at the onset of solubiiiza- tion (Figs. 1 and 2): the low RcsAT value for MLV may be explained by assuming that the outermost bilayers are solubilized before the system has reached equiiib- rium. It has been suggested [I 31 that. in the early stages of MLV solubilization, the outer layers are ‘peeled or, giving rise to transient structures in the form of small unilameilar vesicles. Since in our calculations we con- sider the total amount of lipid, this ICC& t-o an appar- ently low RpSAT value for MLV. Independent measure- ments have shown that, in fact, Triton X-100 takes several hours to equilibrate across the various bilayers of MLV, while solubilization of the outer lamellae starts only seconds after surfactant addition [14]. For SUV. the RDSAT value is in good agreement with the slope of the ‘maximum turbidity vs. lipid concentration’ plot published by Alonso et al. [15], who found a value of about O&7; this was expected since the point of maxi-

Table I

Effective dcicrgcnt ‘lipid molar r;ltios and rebated parameters in the solubiliracion of phosphatidylcholinc vesicles by Triton X-100

Type of vesicles

WV LUV MLV

H SAl

0: (mM) 0.18 0.71 0.15 0.46 0.37 0.29

K (mW’) 1.7 1.9 0.52 RSDL

D:. (mlY1) 3.7 3.0 3.6 0.53 0.45 0.23

-

Data arc derived from plots as shown in Fig. 2.

139

Volume 302. number 2 FEBS LETTERS May 1992

mum turbidity of a detergent-treated SUV suspensicn is also the point at which solubilization starts. Also our A SAT for SUV is reasonably close to the value found eiperimentally by Lasch et al. [ 161. i.e. 1.13. In general, our values for both RcSAT and RzoL are in good agree- ment with the estimation of Dennis [17], who, after chromatographic and sedimentation studies of Triton X-100:phosphatidylcholine mixtures, found that egg phosphatidylcholine bilayers are able to incorporate the surfactant at molar ratios of Triton to phospholipid below about 1: I, whereas above a molar ratio of about 2:1 all of the phospholipid is converted into mixed micelles.

Values for D,, free detergent concentrations in water (= apparent critical micellar concentrations of Triton X-100 in the presence of lipid) are also shown in Table I. Figures of the same order of magnitude as the critical micellar concentration in pure water (0.24 mM [ 181) are found in all cases, in accordance with the theoretical predictions.

Schurtenberger et al. [8] proposed that, for ideal mix- ing of lipid and detergent, in dilute aqueous media, the distribution of detergent between lipid bilayers and aqueous media obeys a partition coefficient, K, that for low R, values can be estimated as K = RJD,. The cor- responding data are also included in Table I, taking A, as the value at the onset of solubilization,

Effective detergent/lipid ratios for surfactant-induced leakage of phospholipid vesicles have been calculated recently [ 16,191. Values of A, for leakage are lower than A SAf (in this paper), as expected from previous studies [2co].

The above results are of practical importance in view of tile significant deviations that are often found be- tween total and effective detergent:lipid ratios, particu- larly at low phospholipid concentrations. Previous re- sults in the literature may require a re-appraisal in the light of the present data.

Ackr~&&nn~ This work was supported in part by Grants UPVO42.310-029/90 from University of the Basque Country and PGV9002 from the Basque Government. M.A.P. thanks the Basque Government for a postdoctoral grant.

REFERENCES

[II Helenius, A. and Simons, K, (I 975) Biochim, Biophys. Acta 415, 29-79.

[2] Lichtenberg, D.. Robson. R.J. and Dennis, E.A. (1983) Biochim. Biophys. Acta 737, 285-304.

131 Lichtcnberg, D. (1992) in: Handbook of Biological Membranes (Shinitaki, M., cd.) in press.

[4] Alonso, A,, Villcaa, A. and Ciofli, F.M. (1981) FEBS Lett, 123, ?OO-2@4.

[5] Goni, FM., Urbancja. M-A., Arrondo, J.L., Alonso, A., Dur- rani, A.A. and Chapman, D. (198G) Eur. J. Biocbcm. 160. 659- 665.

[6] Urbaneja, M.A., Go?& F,M. and Alonso, A. (1988) Eur. J. Bio- chcm, 173, 585-588.

171

PI

PI WI

[I 11

WI

Il.31

iI41

[ISI

[I61

Urbaneja, M.A., Nieva, J,L., Got%. F,M. and Alonso, A. (1987) Biochim. Biophys. Acta 904. 337-345, Schurtenbcrger. P.. Maaar, N.A. and Kanzig, W. (1985) J. Phys. Chem. 89, 1042-1049. Lichtcnbsrg, D. (1985) Biochim. Biophys. Acta 821, 470-478. Almog, S,, Litman, B.J,, Wimlcy, W., Cohen, J., Wachtcl, E.J., Burcnhole, Y.. Ben-Shaul, A, and Lichtenberg, D, (1990) Bio- chemistry 29. 4582-4592. Mayer, L,D.. Hope, MJ. and Cullis, P.R. (1986) Biochim. Bio. phys. Acta 858, 161-168. Urbaneju, M.A., Alonsti, A., Gonzalezas, J.M., GoAi, F.M., Par- tearroyo, M,A,. Tribout, M. and Paredes, S. (1990) Biochem. J, 270, 305-308. Urbaneja, M.A., Arrondo, J.L.R.. Alonso, A. and Golti, F.M. (1986) in: Surfactants in Solution, vol. 5 (Mittal, K.L. and Both borcl, P., eds.) pp, 759-771, Plenum. New York. Alamo, A., Urbnneja, M-A.. Gorli, F.M., Carmona, F.G., Cano vas, F.C. and Gbmez-FernBndez. J.C. (1987) Biochim. Biophys, Acta 902, 237-246. Alonso. A.. Saner. R., Villcna, A. and GOAL. F.M. (1982) J. Membrane Biol. 67, 55-62. Lasch. J., Hoffmann, J.. Omelyanenko, W.G., Klibanov, A.A., Torchilin, V.P., Binder, H. and Gawrisch, K. (1990) Biochim. Biophys. Acta 1022, 171-180. Dennis, E.A, (1974) Arch, Biachem. Biophys. 165, 764-773. Helcnius, A.. McCaslin. D.R., Fries, E. and Tanford, C, (1979) Methods Enzymol. 56 734-749.

[I91 De la Maza. A.. Stinchcz Lenl, J., Parra, J.L., Garcia, M.T. and Ribosa, 1. (1991) J. Am. Oil Chem. Sot. 68. 315-319.

[20] Ruiz, J., Goni, F,M, and Alonso, A, (1988) Biochim. Biophys. Acta 937, 127-i34.

140