- Email: [email protected]
The susceptibility of membrane vesicles to interaction with ethanol
Chemistry and Physics of Lipids, 52 (1990) 179-187 Elsevier Scientific Publishers Ireland Ltd.
179
The susceptibility of membrane vesicles to interaction with ethanol Chang-Hwei Chen and Stuart G. Engel Wadsworth Center[or Laboratories and Research, New York State Department of Health, Albany, New York 12201 and Department of Biomedical Sciences, State University of New York, Albany, New York (U.S.A.) (Received April 12th, 1989: revised and accepted May 30th, 1989)
The susceptibility of membranes to interaction with ethanol is an important consideration in the further understanding of the ethanol-membrane interaction. Interactions of membrane vesicles, including passive diffusion of ethanol across membranes, leakage of internal molecules out of membranes and membrane-membrane interaction, were examined systematically using two populations of fluorescent probe-encapsulated phospholipid bilayer vesicles, each prepared with 1,2-dimyristoyl phosphatidylcholine, cholesterol and a fluorescent probe. Fluorescence quenching experiments with these vesicles were performed in a medium containing a wide range of ethanol concentrations (0.30--3.5 M). In the presence of a lower concentration of ethanol in the external medium, passive diffusion of ethanol across membrane vesicles occurred. This was demonstrated by an interaction of ethanol with the encapsulated fluorescence probe molecules inside the vesicles, resulting in an increase in the fluorescence intensity and a shift of the fluorescence emission spectrum to a shorter wavelength. While, in the presence of a higher concentration of ethanol in the external medium, a strong perturbation of lipid bilayers by ethanol was found, leading to an over expansion of membranes and consequently causing the membrane leakage. As a result of this, the inititally encapsulated probe molecules leaked out of the vesicles so as to interact with the other probe molecules in the external medium. Consequently, fluorescence quenching was observed. Moreover, studies of the mixture of two populations of fluorescence probe-encapsulated membrane vesicles revealed that ethanol acted on individual membranes and did not promote membrane-membrane interactions. The implication of the present results to the alcohol-mediated expansion of membranes is discussed.
Keywords: ethanol-membrane interaction.
Introduction
Over the last fifteen years, studies of the action of alcohols on membranes have been emphasized essentially on their ability to perturb membrane lipids. Alcohols are believed to increase the fluidity or the disorder of natural and model membranes. This has been demonstrated by physical methods including nuclear magnetic resonance [1-3], electron paramagnetic reasonance [4,5], fluorescence probe technique [6~7] and differential scanning microcalorimetry [8-10]. A favorable hypothesis in the ethanol-membrane interaction Correspondence to: C.-H. Chen.
is that ethanol acts at either the hydrophobic regions of the lipids or the proteins in the membranes [11,12]. In a systematic study employing a series of alcohols containing various alkyl chain lengths [9], we have found that the order of alcohols in increasing membrane fluidity and affecting phasetransition parameters is similar to that of the ability of the alcohols for hydrophobic interaction with a protein [9,13], supporting an interaction of alcohols preferentially with the hydrophobic region. The anesthetic action of alcohols with higher alkyl chains has also been given considerable attention. It has been proposed that the action of
0009-3084/90/$03.50 © 1990 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland
180
alcohols may be mediated by membrane expansion [14]. An important hypothesis in this area of study is that alcohols induce an anisotropic expansion of the membrane, such that membrane thickness decreases while its area increases [15-17]. To address the question of the alcoholmediated expansion of membranes, it is essential to examine the susceptibility of membranes to interaction with ethanol. However, there is a lack of information in this respect. In continuation of our previous work on the effects of ethanol on membrane fluidity [9] and the alcohol-protein interaction [13], the present work has concentrated on the studies of the susceptibility of membranes to interaction with ethanol. Since lipid bilayer vesicles are a good model membrane system [18-22], we have used fluorescent probe-encapsulated phospholipid bilayer vesicles prepared with 1,2-dimyristoyl phosphatidylcholine, cholesterol and a fluorescent probe for such an investigation. Fluorescence quenching experiments were performed to examine the susceptibility of membrane vesicles to interaction with ethanol, including the passive diffusion of ethanol across membranes, the leakage of internal molecules out of membranes and the membrane-membrane interaction. The implication of the present results to the alcohol-mediated expansion of membranes was discussed.
Experimental Chemicals 1,2-Dimyristoyl phosphatidylcholine (DMPC) and cholesterol were obtained from Sigma Chemical Co. Fluorescence probes, 8-aminonaphthalene-l,3,6-trisulfonic acid, disodium salt (ANTS) and p-xylene-bis-pyridinium bromide (DPX), were purchased from Molecular Probes, Inc. Sephadex G-75 was obtained from Pharmacia Fine Chemicals. All other chemicals, including ethanol, were of reagent grade and obtained from commercial sources. Fluorescence probe-encapsulated membrane vesicles Fluorescence-probe encapsulated membrane vesicles were prepared by the reverse phase evap-
oration method [23]. Large unilamellar vesicles were formed from water-in-oil emulsion of lipids and buffer in an excess organic phase, and followed by the removal of organic phase under reduced pressure. DMPC exhibits a phase transition temperature at 24.6°C. The effect of ethanol on its phase transition properties has been characterized [8-10]. DMPC and cholesterol (10:1) were dissolved initially in a flask containing a mixture of chloroform/methanol (9:1). By purging the mixture with N2, the lipids were dried and spread on the glass wall. Fluorescence probe (ANTS or DPX) in tris buffer solution (pH 7.5; 10 mM) and then organic solvents (isopropyl ether/chloroform (1:1)) were added into the flask. The ratio of aqueous phase to organic solvent was 1:6. The suspension was sonicated with a sonicator (Laboratory Supplies Company) for a brief period (3-5 min), forming a homogeneous emulsion. Organic solvents in the emulsion were then removed by a two-step procedure: a reduced pressure was applied until an emulsion became a gel and then continuous evaporation at a lower pressure until a homogeneous dispersion of liposomes was obtained. The dispersions were then passed through a Sephadex G-75 column to separate the probe-encapsulated vesicles from the free probe. Column elution solution contained 100 mM NaC1, 2 mM Tes, 2 mM L-histidine and 1 mM EDTA (pH 7.4). Finally, the fluorescent probe-encapsulated membrane vesicles obtained were examined by fluorescence spectrophotometry. This type of lipid vesicles has been reported to have an average size of about 0.16 ~m [23]. Fluorescence spectrophotometry A Perkin-Elmer MPF-44A fluorescence spectrophotometer equipped with cut-off filters was used to perform fluorescence quenching experiments. The sample cell holder could be made thermostatic by permitting circulation of water at a constant temperature. Its instrumentation has been described in our previous work on 1,6-diphenyl-l,3,5-hexatriene-(DPH) labelled phospholipid bilayers [24]. In the experiments, samples containing ANTS (as the fluorophore) were excited at 370 nm and emission spectra recorded from 420 to 630 nm. DPX itself did not
181 exhibit fluorescence emission, but was used as the quencher [23]. Direct contact of ANTS with DPX resulted in fast quenching of ANTS fluorescence intensity. Fluorescence experiments were performed at 25°C.
Sonicated vesicles A sonication technique was employed to monitor the entrapped probe molecules as well as the closed structure of the prepared vesicles. The vesicles after sonication were denoted as broken vesicles. The entrapped ANTS or DPX molecules were completely released from vesicles after 10 min sonication. The released ANTS or DPX molecules from the vesicles were monitored by fluorescence quenching experiments, where the outer medium contained DPX or ANTS-buffer solution, respectively. That is, the fluorescence intensity of the released ANTS molecules was quenched by DPX in the outer medium or that of the released DPX molecules quenched the ANTS fluorescence intensity in the outer medium. The concentration of ANTS or DPX in the vesicles was determined by comparison of fluorescence intensities in broken vesicles with those in a series of known concentrations of ANTS-DPX mixing solutions.
Ethanol experiments Fluorescence quenching experiments were carried out to examine the effects of ethanol on the susceptibility of membrane vesicles to interaction with ethanol, including the passive diffusion of ethanol, the leakage of internal molecules out of membranes and the membrane-membrane intraction. Ethanol was added to three sets of lipid vesicles: (a) ANTS-encapsulated vesicles; (b) DPX-encapsulated vesicles; and (c) a mixture of both vesicles. To comprehend the observed phenomena and to assure that the data were correctly interpreted, a wide range of ethanol concentrations (0.30-3.5 M) was employed in the experiments. Results
Probe-encapsulated vesicles In using the prepared probe-encapsulated vesicles to investigate the susceptibility of membrane
0.8~__
a
a
M. 0.6-z
W Iz Ill 0.4 ~ o z
IM t9
0.2
re-
o
ii.-I
o
I 450
I
I 510
r
WAVELENGTH,
I 570
r
r 630
I
nm
Fig. 1. Typicalfluorescenceemission spectra of ANTS-encapsulated vesicles: (a) in the absence of DPX; (b) in the presence of 45 mM DPX in the outside medium; and (c) with sonication performed in case (b). The concentration of lipid in the vesicles = 4.9 mg/ml.
vesicles to interaction with ethanol, it is important to make sure that these vesicles exhibit a closed structure. To examine this issue, fluorescence emission spectra of ANTS-encapsulated vesicles were measured in the absence and presence of DPX in the outer medium. A typical measurement is presented in Fig. 1, where the spectrum of ANTS-encapsulated vesicles mixing with the buffer medium containing DPX (Fig. lb) almost overlaps with that containing no DPX (Fig. la). The maximum fluorescence intensity of ANTS around 510 nm is essentially unchanged. Results show that the fluorescence signal associated with ANTS in the vesicles is not quenched by DPX in the outer medium. This observation indicates that (a) ANTS is essentially encapsulated inside the vesicles, (b) the vesicles have a closed structure that exhibits no leakage and (c) the entrapped ANTS molecules do not diffuse across the vesicles. This finding is supported by the sonication experiment as presented in Fig. lc. In the broken vesicles (after sonication), the entrapped ANTS molecules were released from the vesicles. The fluorescence spectrum of the released ANTS is then quenched by D P X in the outside medium. In a parallel experiment, the closed structure
182
of the prepared DPX-encapsulated vesicles was also examined. This was performed by mixing DPX-encapsulated vesicles with the buffer medium containing ANTS and the fluorescence emission spectrum of ANTS in the outer buffer solution was then measured (Fig. 2). The fluorescence intensity of ANTS in the outer medium (Fig. 2a) is reduced by only about 15% as the ANTS solution was mixed with DPX-encapsulated vesicles (Fig. 2b). This observed minor quenching of ANTS is due to the presence of a small amount of DPX molecules attached to the vesicle surface, rather than the leakage of the DPX-encapsulated vesicles. This claim is supported by the following observations: (a) no significant further reduction in the observed fluorescence was noted when monitored for a period of 1 h or more; (b) a similar result was obtained if these vesicles were re-passed through a Sephadex G-75 column and an identical experiment was then performed; and (c) in the sonication experiment, the fluorescence intensity of ANTS in the outer medium is substantially quenched by the broken DPX vesicles (Fig. 2c). These results indicate that the majority of DPX molecules is entrapped inside the vesicles
Q8
which exhibit a closed structure. Comparison of Figs. 2c and lc demonstrates that fluorescence quenching in Fig. 2c is not as complete as that in Fig. lc. This is because the concentration of quencher (DPX) in Fig. 2c (where DPX is initially encapsulated) is much lower than that in Fig. lc (where DPX is present in the outer medium). The nature of the closed structure of these two types of probe-encapsulated vesicles is not affected as they are mixed together. In other words, upon mixing, the entrapped ANTS molecules in one population of vesicles do not mix with the entrapped DPX molecules in another population of vesicles. This is demonstrated in Fig. 3, which presents the fluorescence emission spectra of the mixture of the two populations of ANTS-encapsulated vesicles and DPX-encapsulated vesicles. The mixed vesicles also exhibit a high ANTS fluorescence intensity (Fig. 3a). However, in the sonication experiment where both ANTS and DPX molecules were released from their vesicles, the fluorescence intensity around 510 nm is largely reduced (Fig. 3b), resulting from the direct contact between the released ANTS molecules and the released DPX mol-
m
LL
0.8 ¢/) Z w I,Z O.6
0.6 I,Z W O
Z
0.4
M .
¢1
UJ
1,1.1
O
Z I.I.I
W
~e o.2
o
0,4
1.1.1 ~e
=,J I,I,.
O
I
450
I
[
510
WAVELENGTH,
I
570
i
I
630
.=.1
O.2
I
nm 450
Fig. 2. Typical fluorescence emission spectra of ANTS-buffer solution: (a) in the absence of DPX-encapsulated vesicles: (b) in the presence of DPX-encapsulated vesicles: and (c) with sonication performed in case (b). The concentration of lipid in the vesicles and that of ANTS in ANTS-buffer solution are 5.0 mg/ml and 2.5 m M , respectively.
510
570
WAVELENGTH,
nm
630
Fig. 3. Typical fluorescence emission spectra in a mixture of ANTS-encapsulated vesicles and DPX-encapsulated vesicles: (a) intact vesicles and (b) broken vesicles (after sonication). See Figs. I and 2 for the concentrations of lipid in the vesicles.
183
0.I
O
AFdl
-O.I
-0.z
I 0
I
I
20
I
I
]
40
I
60
I 80
A M O U N T OF BUFFER A D D E D , ~Jl Fig. 4. Plot of the difference (Al~dit = Fx -- F~) in the normalized fluorescence intensity (f') in ANTS-encapsulated vesicles versus the a m o u n t of buffer solution added to the vesicles, where F~ and F. denote F at 510 nm after and before the dilution, respectively. The initial volume of the vesicles = 130 ~1.
ecules and the consequent quenching of the ANTS fluorescence signal by DPX. Furthermore, the closed structure of probeencapsulated vesicles in the absence of cholesterol was also examined. The absence of the incorporation of cholesterol into encapsulated vesicles does not seem to affect the closed stucture
0.8
•
of membranes, since probe-encapsulated vesicles which are prepared in the absence of cholesterol also exhibit a closed structure. Control measurements" for the ethanol action In the elucidation of the susceptibility of membrane vesicles to interaction with ethanol, two
•
0.6
AF. 0.4
0.2
0
I
2
.3
4
[A],M Fig. 5. Plot of the difference ( A F , = F , - F , ~ ) in the normalized fluorescence intensity in ANTS-buffer solution versus the concentration of ethanol [A], where F,, and F. denote F at 510 nm in the presence and absence of ethanol, respectively. The concentration of A N T S = 2.5 mM.
184
control experiments were performed: (a) the fluorescence intensity of ANTS-encapsulated vesicles upon the addition of buffer solution (to examine the dilution effect); and (b) the fluorescence intensity of ANTS-buffer solution in the presence of ethanol (to examine the direct mixing of ANTS molecules with ethanol). The dilution effects are presented in Fig. 4, which plots the difference in the normalized fluorescence intensity ~F,~ ( F , - - Fo) versus the amount of buffer solution added to the vesicles, where F~ and F~ denote the fluorescence intensities after and before the dilution, respectively. Results show that, upon buffer dilution, the fluorescence emission intensity of the vesicles is gradually reduced. The effect of mixing ethanol with ANTSmolecules on the fluorescence intensity of ANTSbuffer solution is demonstrated in Fig. 5. The difference in the normalized fluorescence intensity AF, ( F , - Fo) is plotted versus the concentration of added alcohol ([A] in mol/l), where F, and Fn denote the F of ANTS in the presence and absence of ethanol. Figure 5 reveals that, upon ANTS-ethanol mixing, ANTS fluorescence emission intensity is increased significantly as [A] increases.
The susceptibility of membrane vesicles to ethanol Experiments on the susceptibility o f membrane vesicles to interaction with ethanol were performed in three sets of membrane vesicles: (a) ANTS-encapsulated vesicles; (b) DPX-encapsulated vesicles; and (c) a mixture of these two populations of vesicles. This enabled us to check the internal consistency of the results obtained. In the examination of the action of ethanol on ANTS-encapsulated vesicles, ethanol was added to the suspension of ANTS-encapsulated vesicles mixing with the buffer medium which contains DPX. Here, the fluorescence emission of the suspension was attributed to ANTS molecules inside the vesicles. The plot of AF,, versus [A] is shown in Fig. 6. The magnitude of AF, increases slowly as [A] increases up to [A] = 1.5 M, and then starts decreasing as [A] > 1.5 M. The slow increase in AF, at [A] < 1.5 M is due to the presence of ethanol inside the vesicles through passive diffusion. This is demonstrated by the presence
0.2
AFa -0.2
-0.4
0
I
2
3
Fig. 6. Plot of ,.~F,, of ANTS-encapsulated vesicles mixing with DPX-buffer solution versus [A], where the concentration of lipid in the vesicles = 4.9 mg/ml.
of ethanol causing a decrease in the polarity of the medium inside the vesicles and consequently an increase in the fluorescence quantum yield (see below). Membrane vesicles are obviously able to resist a high concentration of ethanol (up to 1.5 M) without exhibiting membrane leakage. However, as [A] > 1.5 M, membrane leakage occurs. The entrapped ANTS molecules begin to leak out of membrane vesicles and they then mix with DPX molecules in the outside medium. As a result of this, the quenching of ANTS fluorescence intensity by DPX is observed. The degree of quenching is enhanced as [A] increases, due to an increase in the extent of the membrane leakage. Such an observation is believed to be a consequence of the expansion of membrane volume in response to an increase in [A] [15,16] (see Discussion). In a parallel experiment to examine the effect of ethanol on DPX-encapsulated vesicles, ethanol was added to the suspension of DPX-encapsulated vesicles mixing with the buffer medium that contains ANTS. Here, the fluorescence emission intensity of the suspension was attributed to ANTS molecules in the outer medium rather than inside the vesicles. The plot of AF, versus [A] is
185
O.l
0.05
AFa
~Fa ° O,
-0.1
-0.05 0
I
I
I
I
2
3
[A] ,M
Fig. 7. Plot of AF,, of ANTS-buffer solution (2.5 raM) mixing with DPX-encapsulated vesicles versus [A], where the concentration of lipid in the vesicles = 5.0 mg/ml.
presented in Fig. 7. A similar observation to that shown in Fig. 6 is found here. At [A] > 1.5 M, the presence of ethanol causes a gradual increase in the fluorescence intensity of ANTS, due to the direct mixing of ethanol with ANTS in the outer medium. However, at [A] < 1.5 M, a decrease in the fluorescence intensity (AFa) is observed. The magnitude of decrease in AFa is enhanced as [A] increases. Such a decrease in AFa resulted from the membrane leakage which leads to the release of DPX molecules from the vesicles to the outer medium. The released DPX molecules then quench the ANTS fluorescence intensity in the outer medium. For the same reason as indicated in the comparison of Figs. 2c and lc, 5F, in Fig. 7 appears to level off around zero whereas it continues to decrease in Fig. 6 with increasing [A]. A mixture of two populations of ANTS-encapsulated vesicles and DPX-encapsulated vesicles was also used to further elucidate the susceptibility of membrane vesicles to interaction with ethanol. The ANTS fluorescence signal in the mixed populations of vesicles was monitored in the absence and presence of ethanol. Results of the action of ethanol are presented in Fig. 8. At [A] < 1.5 M, the value of AF,, also increases as [A] increases, due to the passive diffusion of ethanol across the mixed vesicles. While at [A] > 1.5 M, ethanol causes a decrease in the fluorescence
-0.15 0
I
2
[A] ,M
3
4
Fig. 8. Plot of AF~ of a mixture of ANTS-encapsulated vesicles and DPX-encapsulated vesicles versus [A]. See Figs. 1 and 2 for the concentrations of lipid in these vesicles. The experiment was performed in tris buffer (pH 7.5; 10 mM).
intensity of ANTS. The magnitude of decrease in AFa is enhanced as [A] increases. Such a decrease in AFa is caused by the leakage of both populations of vesicles, resulting in the released ANTS molecules mixing with the released DPX molecules in the outer medium. This finding is consistent with those observed with their individual vesicles as described above. Therefore, the results reveal that ethanol does not promote the membrane-membrane interactions, since no quenching of ANTS fluorescence by DPX molecules inside the vesicles was observed. Discussion
Using three sets of membrane vesicles: (a) ANTS-encapsulated vesicles; (b) DPX-encapsulated vesicles; and (c) a mixture of these two populations of vesicles, systematic and self-consistent results are obtained in conjunction with the passive diffusion of ethanol across the vesicles, the leakage of internal molecules out of membranes and the membrane-membrane interaction. At [A] < 1.5 M, passive diffusion of ethanol across membrane vesicles is demonstrated by the observed slow increase in the ANTS fluorescence intensity (AFt) in the encapsulated vesicles (Figs. 6 and 8). Since ethanol and water have dielectric constants of 24.3 and 78.5 at 25°C,
186
respectively, the presence of ethanol inside the vesicles will cause a lowering of the polarity of the internal medium. As the polarity of the internal medium decreases, the quantum yield and the fluorescence intensity increase [26,27]. An increase in the fluorescence emission intensity of ANTS molecules is shown in Fig. 5. Furthermore, a decrease in the polarity of the internal medium generally results in a shift of the emission spectrum to a shorter wavelength [27]. That is, if passive diffusion of ethanol across membrane vesicles occurs, then the wavelength of the maximum fluorescence emission of ANTS in the vesicles should shift to a shorter wavelength. Result of the effect of ethanol on the maximum wavelength (am) of the ANTS fluorescence emission intensity in the mixture of two populations of ANTS-encapsulated vesicles and DPX-encapsulated vesicles is presented in Fig. 9. The figure indeed shows that the value of )~m decreases from 509 nm to 500 nm as [A] increases from 0 to 1.3 M. In the mixed ANTS-encapsulated and DPXencapsulated vesicles and at the range of ethanol concentration < 1.5 M, the value of AF,, increases slowly rather than decreases as [A] increases. The observation obviously does not suggest a facilitation of membrane-membrane interaction by
52O
510
m
50C
I O
I I
I 2
[A], M
I 3
Fig. 9. Plot of the wavelength (~,,,,) at the m a x i m u m fluorescence intensity in a mixture of ANTS-encapsulated vesicles and DPX-encapsulated vesicles versus [A]. See Fig. 8 for the concentrations of lipid in these vesicles.
ethanol. Should such an interaction occur, the ANTS fluorescence signal would be quenched by DPX molecules inside the vesicles and a decrease rather than an increase in AFt, would be observed. The penetration of ethanol into membrane bilayers is expected from the investigations of alcohol-membrane interactions such as membrane fluidity [2-4,9,11,12], ethanol transport across the membrane [28] and the action of ethanol on membrane enzymatic activities [29,30]. Miller [15,17] and Trudell [16] have proposed that alcohols induce an anisotropic expansion of the membrane, such that membrane thickness decreases while its area increases. Our observation is consistent with this membrane expansion hypothesis, presuming that the membrane expansion eventually leads to the membrane leakage at a higher concentration range of ethanol. It should be noted that membrane expansion should also occur at a lower [A] except that the degree of membrane expansion has not yet reached the stage that causes membrane leakage to occur. In the present study, membrane expansion increases as [A] increases and consequently causes membrane leakage to occur at [A] > 1.5 M. This is demonstrated in Figs. 6 and 8. The physiologically significant concentration of ethanol is in the range of 50-100 mM. The concentration of lipid in cell membranes is also low, for instance 3.9 p,mol/ml in red cells [31]. In the present biochemical study, the concentrations of ethanol ranging from 300 mM to 3.5 M were used, which are much higher than the physiological concentration range. To compensate for this factor and to meet a ratio of ethanol to lipid more relevant to a biological system, a higher concentration of vesicles (2-5 mg of lipid/ml) was used here. In addition, at this lipid concentration range, the stability of the encapsulated vesicles was higher than that of the more diluted ones. It has been proposed that, in the presence of a higher concentration of ethanol ( > 0.54 M for distearoyl phosphatidylcholine (DSPC) o r > 1.0 M for dipalmitoyl phosphatidylcholine (DPPC)), lipid suspensions form an interdigitated phase where the lipid hydrocarbon chains from opposing monolayers fully interpenetrate or interdigitate [32]. At this stage, we do not have enough
187
information to speculate if the formation of an interdigitized phase would cause the leakage of the encapsulated DPX or ANTS out of lipid vesicles. The present work has provided a systematic study leading to an understanding of the susceptibility of membrane vesicles to interaction with ethanol. The obtained information is valuable in the biochemical studies of the ethanol-membrane interaction. These model membrane studies could be used as a basis for future investigations in biological membranes.
Acknowledgement This work was supported in part by a research grant (AA07048) awarded by the National Institute on Alcohol Abuse and Alcoholism.
References 1 J.M. Pope and D.W. Dubro (1986) Biochim. Biophys. Acta 858, 243-253. 2 R.J. Hitzeman, H.E. Schueler, C. Graham-Britain and G.P. Kreishman (1986) Biochim. Biophys. Acta 859, 189-197. 3 G.L. Turner and E. Oldfield (1979) Nature 277,669-670. 4 J.H. Chin and D.B. Goldstein (1981) Mol. Pharmacol. 19, 425-431. 5 K.-Y. Pang, L.M. Braswell, L. Chang, T.J. Sommer and K.W. Miller (1980) Mol. Pharmacol. 18, 84-90. 6 A.G. Lee (1976) Biochemistry 15, 2448-2454. 7 G.B. Zavoico and H. Kutchai (1980) Biochim. Biophys. Acta 600, 263-269. 8 M.K. Jain and N.M. Wu (1977) J. Membrane Biol. 34, 157-201. 9 C.-H. Chen (1981) J. Phys. Chem. 85, 603-608.
10 A.W. Eliasz, D. Chapman and D.F. Ewing (1976) Biochim. Biophys. Acta 448, 220-230. 11 G.Y. Sun and A.Y. Sun (1985) Alcohol Clin. Exp. Res. 9, 164-180. 12 R.J. Hitzmann, R.A. Haris and H.H. Loh (1984) in: M. Shinitsky (Ed.), Physiology of Membrane Fluidity, CRC Press, Florida, pp. 109-126 13 C.-H. Chen (1980) J. Phys. Chem. 84, 2050-2053. 14 P. Seeman (1972) Pharmacol. Rev. 24, 583-655. 15 K.W. Miller, W.D.M. Paton, R.A. Smith and E.B. Smith (1973) Mol. Pharmacol. 9, 131-143. 16 J.R. Trudell (1977) Biochim. Biophys. Acta 470, 509-510. 17 Y. Kita and K.W. Miller (1982) Biochemistry 21, 2840-2847. 18 D. Chapman (1975) Quart. Rev. Biophys. 8, 185-235. 19 J. Seelig and A. Seelig (1980) Quart. Rev. Biophys. 13, 19-61. 20 S. Mabrey and J.M. Sturtevant (1978) Methods Membr. Biol. 9, 237-274. 21 J.T. Mason, C.-H. Huang and R.L. Biltonen (1983) Biochemistry, 2013-2018. 22 J.T. Mason and C.-H. Huang (1981) Lipids 16,604-608, 23 F. Szoka and D. Papahadjopoulos (1980) Annu. Rev. Biophys. Bioenerg. 9,467-508. 24 D. Guard-Friar, C.-H. Chen and A.S. Engel, (1985) J. Phys. Chem. 89, 1810--1813. 25 N.P. Franks and W.R. Lieb (1981) Nature 292,248-251. 26 J.R. Lakowicz (1983) in: Principles of Fluorescence Spectroscopy, Plenum Press, New York, pp. 187-214. 27 L.S. Stryer (1965) J. Mol. Biol. 13,482-495. 28 P. Naccache and R.I. Shaafi (1973) J. Gen. Physiol. 62, 714-736. 29 A.Y. Sun and R.N. Seaman (1980) Neurochem. Res. 5, 537-545. 30 A.C. Collins, A. Smolen, A.L. Wayman and M.J. Marks (1984) Alcohol 1,237-246. 31 J.T. Dodge and G.B. Phill!ps (1967) J. Lipid Res. 8, 667-675. 32 S.A. Simon and T.J. McIntosh (1984) Biochim. Biophys. Acta 773, 169-172.
179
The susceptibility of membrane vesicles to interaction with ethanol Chang-Hwei Chen and Stuart G. Engel Wadsworth Center[or Laboratories and Research, New York State Department of Health, Albany, New York 12201 and Department of Biomedical Sciences, State University of New York, Albany, New York (U.S.A.) (Received April 12th, 1989: revised and accepted May 30th, 1989)
The susceptibility of membranes to interaction with ethanol is an important consideration in the further understanding of the ethanol-membrane interaction. Interactions of membrane vesicles, including passive diffusion of ethanol across membranes, leakage of internal molecules out of membranes and membrane-membrane interaction, were examined systematically using two populations of fluorescent probe-encapsulated phospholipid bilayer vesicles, each prepared with 1,2-dimyristoyl phosphatidylcholine, cholesterol and a fluorescent probe. Fluorescence quenching experiments with these vesicles were performed in a medium containing a wide range of ethanol concentrations (0.30--3.5 M). In the presence of a lower concentration of ethanol in the external medium, passive diffusion of ethanol across membrane vesicles occurred. This was demonstrated by an interaction of ethanol with the encapsulated fluorescence probe molecules inside the vesicles, resulting in an increase in the fluorescence intensity and a shift of the fluorescence emission spectrum to a shorter wavelength. While, in the presence of a higher concentration of ethanol in the external medium, a strong perturbation of lipid bilayers by ethanol was found, leading to an over expansion of membranes and consequently causing the membrane leakage. As a result of this, the inititally encapsulated probe molecules leaked out of the vesicles so as to interact with the other probe molecules in the external medium. Consequently, fluorescence quenching was observed. Moreover, studies of the mixture of two populations of fluorescence probe-encapsulated membrane vesicles revealed that ethanol acted on individual membranes and did not promote membrane-membrane interactions. The implication of the present results to the alcohol-mediated expansion of membranes is discussed.
Keywords: ethanol-membrane interaction.
Introduction
Over the last fifteen years, studies of the action of alcohols on membranes have been emphasized essentially on their ability to perturb membrane lipids. Alcohols are believed to increase the fluidity or the disorder of natural and model membranes. This has been demonstrated by physical methods including nuclear magnetic resonance [1-3], electron paramagnetic reasonance [4,5], fluorescence probe technique [6~7] and differential scanning microcalorimetry [8-10]. A favorable hypothesis in the ethanol-membrane interaction Correspondence to: C.-H. Chen.
is that ethanol acts at either the hydrophobic regions of the lipids or the proteins in the membranes [11,12]. In a systematic study employing a series of alcohols containing various alkyl chain lengths [9], we have found that the order of alcohols in increasing membrane fluidity and affecting phasetransition parameters is similar to that of the ability of the alcohols for hydrophobic interaction with a protein [9,13], supporting an interaction of alcohols preferentially with the hydrophobic region. The anesthetic action of alcohols with higher alkyl chains has also been given considerable attention. It has been proposed that the action of
0009-3084/90/$03.50 © 1990 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland
180
alcohols may be mediated by membrane expansion [14]. An important hypothesis in this area of study is that alcohols induce an anisotropic expansion of the membrane, such that membrane thickness decreases while its area increases [15-17]. To address the question of the alcoholmediated expansion of membranes, it is essential to examine the susceptibility of membranes to interaction with ethanol. However, there is a lack of information in this respect. In continuation of our previous work on the effects of ethanol on membrane fluidity [9] and the alcohol-protein interaction [13], the present work has concentrated on the studies of the susceptibility of membranes to interaction with ethanol. Since lipid bilayer vesicles are a good model membrane system [18-22], we have used fluorescent probe-encapsulated phospholipid bilayer vesicles prepared with 1,2-dimyristoyl phosphatidylcholine, cholesterol and a fluorescent probe for such an investigation. Fluorescence quenching experiments were performed to examine the susceptibility of membrane vesicles to interaction with ethanol, including the passive diffusion of ethanol across membranes, the leakage of internal molecules out of membranes and the membrane-membrane interaction. The implication of the present results to the alcohol-mediated expansion of membranes was discussed.
Experimental Chemicals 1,2-Dimyristoyl phosphatidylcholine (DMPC) and cholesterol were obtained from Sigma Chemical Co. Fluorescence probes, 8-aminonaphthalene-l,3,6-trisulfonic acid, disodium salt (ANTS) and p-xylene-bis-pyridinium bromide (DPX), were purchased from Molecular Probes, Inc. Sephadex G-75 was obtained from Pharmacia Fine Chemicals. All other chemicals, including ethanol, were of reagent grade and obtained from commercial sources. Fluorescence probe-encapsulated membrane vesicles Fluorescence-probe encapsulated membrane vesicles were prepared by the reverse phase evap-
oration method [23]. Large unilamellar vesicles were formed from water-in-oil emulsion of lipids and buffer in an excess organic phase, and followed by the removal of organic phase under reduced pressure. DMPC exhibits a phase transition temperature at 24.6°C. The effect of ethanol on its phase transition properties has been characterized [8-10]. DMPC and cholesterol (10:1) were dissolved initially in a flask containing a mixture of chloroform/methanol (9:1). By purging the mixture with N2, the lipids were dried and spread on the glass wall. Fluorescence probe (ANTS or DPX) in tris buffer solution (pH 7.5; 10 mM) and then organic solvents (isopropyl ether/chloroform (1:1)) were added into the flask. The ratio of aqueous phase to organic solvent was 1:6. The suspension was sonicated with a sonicator (Laboratory Supplies Company) for a brief period (3-5 min), forming a homogeneous emulsion. Organic solvents in the emulsion were then removed by a two-step procedure: a reduced pressure was applied until an emulsion became a gel and then continuous evaporation at a lower pressure until a homogeneous dispersion of liposomes was obtained. The dispersions were then passed through a Sephadex G-75 column to separate the probe-encapsulated vesicles from the free probe. Column elution solution contained 100 mM NaC1, 2 mM Tes, 2 mM L-histidine and 1 mM EDTA (pH 7.4). Finally, the fluorescent probe-encapsulated membrane vesicles obtained were examined by fluorescence spectrophotometry. This type of lipid vesicles has been reported to have an average size of about 0.16 ~m [23]. Fluorescence spectrophotometry A Perkin-Elmer MPF-44A fluorescence spectrophotometer equipped with cut-off filters was used to perform fluorescence quenching experiments. The sample cell holder could be made thermostatic by permitting circulation of water at a constant temperature. Its instrumentation has been described in our previous work on 1,6-diphenyl-l,3,5-hexatriene-(DPH) labelled phospholipid bilayers [24]. In the experiments, samples containing ANTS (as the fluorophore) were excited at 370 nm and emission spectra recorded from 420 to 630 nm. DPX itself did not
181 exhibit fluorescence emission, but was used as the quencher [23]. Direct contact of ANTS with DPX resulted in fast quenching of ANTS fluorescence intensity. Fluorescence experiments were performed at 25°C.
Sonicated vesicles A sonication technique was employed to monitor the entrapped probe molecules as well as the closed structure of the prepared vesicles. The vesicles after sonication were denoted as broken vesicles. The entrapped ANTS or DPX molecules were completely released from vesicles after 10 min sonication. The released ANTS or DPX molecules from the vesicles were monitored by fluorescence quenching experiments, where the outer medium contained DPX or ANTS-buffer solution, respectively. That is, the fluorescence intensity of the released ANTS molecules was quenched by DPX in the outer medium or that of the released DPX molecules quenched the ANTS fluorescence intensity in the outer medium. The concentration of ANTS or DPX in the vesicles was determined by comparison of fluorescence intensities in broken vesicles with those in a series of known concentrations of ANTS-DPX mixing solutions.
Ethanol experiments Fluorescence quenching experiments were carried out to examine the effects of ethanol on the susceptibility of membrane vesicles to interaction with ethanol, including the passive diffusion of ethanol, the leakage of internal molecules out of membranes and the membrane-membrane intraction. Ethanol was added to three sets of lipid vesicles: (a) ANTS-encapsulated vesicles; (b) DPX-encapsulated vesicles; and (c) a mixture of both vesicles. To comprehend the observed phenomena and to assure that the data were correctly interpreted, a wide range of ethanol concentrations (0.30-3.5 M) was employed in the experiments. Results
Probe-encapsulated vesicles In using the prepared probe-encapsulated vesicles to investigate the susceptibility of membrane
0.8~__
a
a
M. 0.6-z
W Iz Ill 0.4 ~ o z
IM t9
0.2
re-
o
ii.-I
o
I 450
I
I 510
r
WAVELENGTH,
I 570
r
r 630
I
nm
Fig. 1. Typicalfluorescenceemission spectra of ANTS-encapsulated vesicles: (a) in the absence of DPX; (b) in the presence of 45 mM DPX in the outside medium; and (c) with sonication performed in case (b). The concentration of lipid in the vesicles = 4.9 mg/ml.
vesicles to interaction with ethanol, it is important to make sure that these vesicles exhibit a closed structure. To examine this issue, fluorescence emission spectra of ANTS-encapsulated vesicles were measured in the absence and presence of DPX in the outer medium. A typical measurement is presented in Fig. 1, where the spectrum of ANTS-encapsulated vesicles mixing with the buffer medium containing DPX (Fig. lb) almost overlaps with that containing no DPX (Fig. la). The maximum fluorescence intensity of ANTS around 510 nm is essentially unchanged. Results show that the fluorescence signal associated with ANTS in the vesicles is not quenched by DPX in the outer medium. This observation indicates that (a) ANTS is essentially encapsulated inside the vesicles, (b) the vesicles have a closed structure that exhibits no leakage and (c) the entrapped ANTS molecules do not diffuse across the vesicles. This finding is supported by the sonication experiment as presented in Fig. lc. In the broken vesicles (after sonication), the entrapped ANTS molecules were released from the vesicles. The fluorescence spectrum of the released ANTS is then quenched by D P X in the outside medium. In a parallel experiment, the closed structure
182
of the prepared DPX-encapsulated vesicles was also examined. This was performed by mixing DPX-encapsulated vesicles with the buffer medium containing ANTS and the fluorescence emission spectrum of ANTS in the outer buffer solution was then measured (Fig. 2). The fluorescence intensity of ANTS in the outer medium (Fig. 2a) is reduced by only about 15% as the ANTS solution was mixed with DPX-encapsulated vesicles (Fig. 2b). This observed minor quenching of ANTS is due to the presence of a small amount of DPX molecules attached to the vesicle surface, rather than the leakage of the DPX-encapsulated vesicles. This claim is supported by the following observations: (a) no significant further reduction in the observed fluorescence was noted when monitored for a period of 1 h or more; (b) a similar result was obtained if these vesicles were re-passed through a Sephadex G-75 column and an identical experiment was then performed; and (c) in the sonication experiment, the fluorescence intensity of ANTS in the outer medium is substantially quenched by the broken DPX vesicles (Fig. 2c). These results indicate that the majority of DPX molecules is entrapped inside the vesicles
Q8
which exhibit a closed structure. Comparison of Figs. 2c and lc demonstrates that fluorescence quenching in Fig. 2c is not as complete as that in Fig. lc. This is because the concentration of quencher (DPX) in Fig. 2c (where DPX is initially encapsulated) is much lower than that in Fig. lc (where DPX is present in the outer medium). The nature of the closed structure of these two types of probe-encapsulated vesicles is not affected as they are mixed together. In other words, upon mixing, the entrapped ANTS molecules in one population of vesicles do not mix with the entrapped DPX molecules in another population of vesicles. This is demonstrated in Fig. 3, which presents the fluorescence emission spectra of the mixture of the two populations of ANTS-encapsulated vesicles and DPX-encapsulated vesicles. The mixed vesicles also exhibit a high ANTS fluorescence intensity (Fig. 3a). However, in the sonication experiment where both ANTS and DPX molecules were released from their vesicles, the fluorescence intensity around 510 nm is largely reduced (Fig. 3b), resulting from the direct contact between the released ANTS molecules and the released DPX mol-
m
LL
0.8 ¢/) Z w I,Z O.6
0.6 I,Z W O
Z
0.4
M .
¢1
UJ
1,1.1
O
Z I.I.I
W
~e o.2
o
0,4
1.1.1 ~e
=,J I,I,.
O
I
450
I
[
510
WAVELENGTH,
I
570
i
I
630
.=.1
O.2
I
nm 450
Fig. 2. Typical fluorescence emission spectra of ANTS-buffer solution: (a) in the absence of DPX-encapsulated vesicles: (b) in the presence of DPX-encapsulated vesicles: and (c) with sonication performed in case (b). The concentration of lipid in the vesicles and that of ANTS in ANTS-buffer solution are 5.0 mg/ml and 2.5 m M , respectively.
510
570
WAVELENGTH,
nm
630
Fig. 3. Typical fluorescence emission spectra in a mixture of ANTS-encapsulated vesicles and DPX-encapsulated vesicles: (a) intact vesicles and (b) broken vesicles (after sonication). See Figs. I and 2 for the concentrations of lipid in the vesicles.
183
0.I
O
AFdl
-O.I
-0.z
I 0
I
I
20
I
I
]
40
I
60
I 80
A M O U N T OF BUFFER A D D E D , ~Jl Fig. 4. Plot of the difference (Al~dit = Fx -- F~) in the normalized fluorescence intensity (f') in ANTS-encapsulated vesicles versus the a m o u n t of buffer solution added to the vesicles, where F~ and F. denote F at 510 nm after and before the dilution, respectively. The initial volume of the vesicles = 130 ~1.
ecules and the consequent quenching of the ANTS fluorescence signal by DPX. Furthermore, the closed structure of probeencapsulated vesicles in the absence of cholesterol was also examined. The absence of the incorporation of cholesterol into encapsulated vesicles does not seem to affect the closed stucture
0.8
•
of membranes, since probe-encapsulated vesicles which are prepared in the absence of cholesterol also exhibit a closed structure. Control measurements" for the ethanol action In the elucidation of the susceptibility of membrane vesicles to interaction with ethanol, two
•
0.6
AF. 0.4
0.2
0
I
2
.3
4
[A],M Fig. 5. Plot of the difference ( A F , = F , - F , ~ ) in the normalized fluorescence intensity in ANTS-buffer solution versus the concentration of ethanol [A], where F,, and F. denote F at 510 nm in the presence and absence of ethanol, respectively. The concentration of A N T S = 2.5 mM.
184
control experiments were performed: (a) the fluorescence intensity of ANTS-encapsulated vesicles upon the addition of buffer solution (to examine the dilution effect); and (b) the fluorescence intensity of ANTS-buffer solution in the presence of ethanol (to examine the direct mixing of ANTS molecules with ethanol). The dilution effects are presented in Fig. 4, which plots the difference in the normalized fluorescence intensity ~F,~ ( F , - - Fo) versus the amount of buffer solution added to the vesicles, where F~ and F~ denote the fluorescence intensities after and before the dilution, respectively. Results show that, upon buffer dilution, the fluorescence emission intensity of the vesicles is gradually reduced. The effect of mixing ethanol with ANTSmolecules on the fluorescence intensity of ANTSbuffer solution is demonstrated in Fig. 5. The difference in the normalized fluorescence intensity AF, ( F , - Fo) is plotted versus the concentration of added alcohol ([A] in mol/l), where F, and Fn denote the F of ANTS in the presence and absence of ethanol. Figure 5 reveals that, upon ANTS-ethanol mixing, ANTS fluorescence emission intensity is increased significantly as [A] increases.
The susceptibility of membrane vesicles to ethanol Experiments on the susceptibility o f membrane vesicles to interaction with ethanol were performed in three sets of membrane vesicles: (a) ANTS-encapsulated vesicles; (b) DPX-encapsulated vesicles; and (c) a mixture of these two populations of vesicles. This enabled us to check the internal consistency of the results obtained. In the examination of the action of ethanol on ANTS-encapsulated vesicles, ethanol was added to the suspension of ANTS-encapsulated vesicles mixing with the buffer medium which contains DPX. Here, the fluorescence emission of the suspension was attributed to ANTS molecules inside the vesicles. The plot of AF,, versus [A] is shown in Fig. 6. The magnitude of AF, increases slowly as [A] increases up to [A] = 1.5 M, and then starts decreasing as [A] > 1.5 M. The slow increase in AF, at [A] < 1.5 M is due to the presence of ethanol inside the vesicles through passive diffusion. This is demonstrated by the presence
0.2
AFa -0.2
-0.4
0
I
2
3
Fig. 6. Plot of ,.~F,, of ANTS-encapsulated vesicles mixing with DPX-buffer solution versus [A], where the concentration of lipid in the vesicles = 4.9 mg/ml.
of ethanol causing a decrease in the polarity of the medium inside the vesicles and consequently an increase in the fluorescence quantum yield (see below). Membrane vesicles are obviously able to resist a high concentration of ethanol (up to 1.5 M) without exhibiting membrane leakage. However, as [A] > 1.5 M, membrane leakage occurs. The entrapped ANTS molecules begin to leak out of membrane vesicles and they then mix with DPX molecules in the outside medium. As a result of this, the quenching of ANTS fluorescence intensity by DPX is observed. The degree of quenching is enhanced as [A] increases, due to an increase in the extent of the membrane leakage. Such an observation is believed to be a consequence of the expansion of membrane volume in response to an increase in [A] [15,16] (see Discussion). In a parallel experiment to examine the effect of ethanol on DPX-encapsulated vesicles, ethanol was added to the suspension of DPX-encapsulated vesicles mixing with the buffer medium that contains ANTS. Here, the fluorescence emission intensity of the suspension was attributed to ANTS molecules in the outer medium rather than inside the vesicles. The plot of AF, versus [A] is
185
O.l
0.05
AFa
~Fa ° O,
-0.1
-0.05 0
I
I
I
I
2
3
[A] ,M
Fig. 7. Plot of AF,, of ANTS-buffer solution (2.5 raM) mixing with DPX-encapsulated vesicles versus [A], where the concentration of lipid in the vesicles = 5.0 mg/ml.
presented in Fig. 7. A similar observation to that shown in Fig. 6 is found here. At [A] > 1.5 M, the presence of ethanol causes a gradual increase in the fluorescence intensity of ANTS, due to the direct mixing of ethanol with ANTS in the outer medium. However, at [A] < 1.5 M, a decrease in the fluorescence intensity (AFa) is observed. The magnitude of decrease in AFa is enhanced as [A] increases. Such a decrease in AFa resulted from the membrane leakage which leads to the release of DPX molecules from the vesicles to the outer medium. The released DPX molecules then quench the ANTS fluorescence intensity in the outer medium. For the same reason as indicated in the comparison of Figs. 2c and lc, 5F, in Fig. 7 appears to level off around zero whereas it continues to decrease in Fig. 6 with increasing [A]. A mixture of two populations of ANTS-encapsulated vesicles and DPX-encapsulated vesicles was also used to further elucidate the susceptibility of membrane vesicles to interaction with ethanol. The ANTS fluorescence signal in the mixed populations of vesicles was monitored in the absence and presence of ethanol. Results of the action of ethanol are presented in Fig. 8. At [A] < 1.5 M, the value of AF,, also increases as [A] increases, due to the passive diffusion of ethanol across the mixed vesicles. While at [A] > 1.5 M, ethanol causes a decrease in the fluorescence
-0.15 0
I
2
[A] ,M
3
4
Fig. 8. Plot of AF~ of a mixture of ANTS-encapsulated vesicles and DPX-encapsulated vesicles versus [A]. See Figs. 1 and 2 for the concentrations of lipid in these vesicles. The experiment was performed in tris buffer (pH 7.5; 10 mM).
intensity of ANTS. The magnitude of decrease in AFa is enhanced as [A] increases. Such a decrease in AFa is caused by the leakage of both populations of vesicles, resulting in the released ANTS molecules mixing with the released DPX molecules in the outer medium. This finding is consistent with those observed with their individual vesicles as described above. Therefore, the results reveal that ethanol does not promote the membrane-membrane interactions, since no quenching of ANTS fluorescence by DPX molecules inside the vesicles was observed. Discussion
Using three sets of membrane vesicles: (a) ANTS-encapsulated vesicles; (b) DPX-encapsulated vesicles; and (c) a mixture of these two populations of vesicles, systematic and self-consistent results are obtained in conjunction with the passive diffusion of ethanol across the vesicles, the leakage of internal molecules out of membranes and the membrane-membrane interaction. At [A] < 1.5 M, passive diffusion of ethanol across membrane vesicles is demonstrated by the observed slow increase in the ANTS fluorescence intensity (AFt) in the encapsulated vesicles (Figs. 6 and 8). Since ethanol and water have dielectric constants of 24.3 and 78.5 at 25°C,
186
respectively, the presence of ethanol inside the vesicles will cause a lowering of the polarity of the internal medium. As the polarity of the internal medium decreases, the quantum yield and the fluorescence intensity increase [26,27]. An increase in the fluorescence emission intensity of ANTS molecules is shown in Fig. 5. Furthermore, a decrease in the polarity of the internal medium generally results in a shift of the emission spectrum to a shorter wavelength [27]. That is, if passive diffusion of ethanol across membrane vesicles occurs, then the wavelength of the maximum fluorescence emission of ANTS in the vesicles should shift to a shorter wavelength. Result of the effect of ethanol on the maximum wavelength (am) of the ANTS fluorescence emission intensity in the mixture of two populations of ANTS-encapsulated vesicles and DPX-encapsulated vesicles is presented in Fig. 9. The figure indeed shows that the value of )~m decreases from 509 nm to 500 nm as [A] increases from 0 to 1.3 M. In the mixed ANTS-encapsulated and DPXencapsulated vesicles and at the range of ethanol concentration < 1.5 M, the value of AF,, increases slowly rather than decreases as [A] increases. The observation obviously does not suggest a facilitation of membrane-membrane interaction by
52O
510
m
50C
I O
I I
I 2
[A], M
I 3
Fig. 9. Plot of the wavelength (~,,,,) at the m a x i m u m fluorescence intensity in a mixture of ANTS-encapsulated vesicles and DPX-encapsulated vesicles versus [A]. See Fig. 8 for the concentrations of lipid in these vesicles.
ethanol. Should such an interaction occur, the ANTS fluorescence signal would be quenched by DPX molecules inside the vesicles and a decrease rather than an increase in AFt, would be observed. The penetration of ethanol into membrane bilayers is expected from the investigations of alcohol-membrane interactions such as membrane fluidity [2-4,9,11,12], ethanol transport across the membrane [28] and the action of ethanol on membrane enzymatic activities [29,30]. Miller [15,17] and Trudell [16] have proposed that alcohols induce an anisotropic expansion of the membrane, such that membrane thickness decreases while its area increases. Our observation is consistent with this membrane expansion hypothesis, presuming that the membrane expansion eventually leads to the membrane leakage at a higher concentration range of ethanol. It should be noted that membrane expansion should also occur at a lower [A] except that the degree of membrane expansion has not yet reached the stage that causes membrane leakage to occur. In the present study, membrane expansion increases as [A] increases and consequently causes membrane leakage to occur at [A] > 1.5 M. This is demonstrated in Figs. 6 and 8. The physiologically significant concentration of ethanol is in the range of 50-100 mM. The concentration of lipid in cell membranes is also low, for instance 3.9 p,mol/ml in red cells [31]. In the present biochemical study, the concentrations of ethanol ranging from 300 mM to 3.5 M were used, which are much higher than the physiological concentration range. To compensate for this factor and to meet a ratio of ethanol to lipid more relevant to a biological system, a higher concentration of vesicles (2-5 mg of lipid/ml) was used here. In addition, at this lipid concentration range, the stability of the encapsulated vesicles was higher than that of the more diluted ones. It has been proposed that, in the presence of a higher concentration of ethanol ( > 0.54 M for distearoyl phosphatidylcholine (DSPC) o r > 1.0 M for dipalmitoyl phosphatidylcholine (DPPC)), lipid suspensions form an interdigitated phase where the lipid hydrocarbon chains from opposing monolayers fully interpenetrate or interdigitate [32]. At this stage, we do not have enough
187
information to speculate if the formation of an interdigitized phase would cause the leakage of the encapsulated DPX or ANTS out of lipid vesicles. The present work has provided a systematic study leading to an understanding of the susceptibility of membrane vesicles to interaction with ethanol. The obtained information is valuable in the biochemical studies of the ethanol-membrane interaction. These model membrane studies could be used as a basis for future investigations in biological membranes.
Acknowledgement This work was supported in part by a research grant (AA07048) awarded by the National Institute on Alcohol Abuse and Alcoholism.
References 1 J.M. Pope and D.W. Dubro (1986) Biochim. Biophys. Acta 858, 243-253. 2 R.J. Hitzeman, H.E. Schueler, C. Graham-Britain and G.P. Kreishman (1986) Biochim. Biophys. Acta 859, 189-197. 3 G.L. Turner and E. Oldfield (1979) Nature 277,669-670. 4 J.H. Chin and D.B. Goldstein (1981) Mol. Pharmacol. 19, 425-431. 5 K.-Y. Pang, L.M. Braswell, L. Chang, T.J. Sommer and K.W. Miller (1980) Mol. Pharmacol. 18, 84-90. 6 A.G. Lee (1976) Biochemistry 15, 2448-2454. 7 G.B. Zavoico and H. Kutchai (1980) Biochim. Biophys. Acta 600, 263-269. 8 M.K. Jain and N.M. Wu (1977) J. Membrane Biol. 34, 157-201. 9 C.-H. Chen (1981) J. Phys. Chem. 85, 603-608.
10 A.W. Eliasz, D. Chapman and D.F. Ewing (1976) Biochim. Biophys. Acta 448, 220-230. 11 G.Y. Sun and A.Y. Sun (1985) Alcohol Clin. Exp. Res. 9, 164-180. 12 R.J. Hitzmann, R.A. Haris and H.H. Loh (1984) in: M. Shinitsky (Ed.), Physiology of Membrane Fluidity, CRC Press, Florida, pp. 109-126 13 C.-H. Chen (1980) J. Phys. Chem. 84, 2050-2053. 14 P. Seeman (1972) Pharmacol. Rev. 24, 583-655. 15 K.W. Miller, W.D.M. Paton, R.A. Smith and E.B. Smith (1973) Mol. Pharmacol. 9, 131-143. 16 J.R. Trudell (1977) Biochim. Biophys. Acta 470, 509-510. 17 Y. Kita and K.W. Miller (1982) Biochemistry 21, 2840-2847. 18 D. Chapman (1975) Quart. Rev. Biophys. 8, 185-235. 19 J. Seelig and A. Seelig (1980) Quart. Rev. Biophys. 13, 19-61. 20 S. Mabrey and J.M. Sturtevant (1978) Methods Membr. Biol. 9, 237-274. 21 J.T. Mason, C.-H. Huang and R.L. Biltonen (1983) Biochemistry, 2013-2018. 22 J.T. Mason and C.-H. Huang (1981) Lipids 16,604-608, 23 F. Szoka and D. Papahadjopoulos (1980) Annu. Rev. Biophys. Bioenerg. 9,467-508. 24 D. Guard-Friar, C.-H. Chen and A.S. Engel, (1985) J. Phys. Chem. 89, 1810--1813. 25 N.P. Franks and W.R. Lieb (1981) Nature 292,248-251. 26 J.R. Lakowicz (1983) in: Principles of Fluorescence Spectroscopy, Plenum Press, New York, pp. 187-214. 27 L.S. Stryer (1965) J. Mol. Biol. 13,482-495. 28 P. Naccache and R.I. Shaafi (1973) J. Gen. Physiol. 62, 714-736. 29 A.Y. Sun and R.N. Seaman (1980) Neurochem. Res. 5, 537-545. 30 A.C. Collins, A. Smolen, A.L. Wayman and M.J. Marks (1984) Alcohol 1,237-246. 31 J.T. Dodge and G.B. Phill!ps (1967) J. Lipid Res. 8, 667-675. 32 S.A. Simon and T.J. McIntosh (1984) Biochim. Biophys. Acta 773, 169-172.