Monovalent cation-induced fusion of acidic phospholipid vesicles

Chemistry and Physics of Lipids, 52 (1990) 199-206 Elsevier Scientific Publishers Ireland Ltd.

199

Monovalent cation-induced fusion of acidic phospholipid vesicles K a r i K. E k l u n d Department of Medical Chemistry, University of Helsinki, Siltavuorenpenger 10, 00170, Helsinki (Finland) (Received April 13th, 1989; revision received June 7th, 1989; accepted June 9th, 1989) Fusion of small unilamellar vesicles (SUV) consisting of dimyristoylphosphatidylcholine (DMPC), dimyristoylphosphatidylglycerol (DMPG) and phosphatidylglycerol (PG) from egg yolk, dipalmitoylphosphatidylserine (DPPS) and phosphatidylserine (PS) from bovine brain was studied as a function of monovalent cation concentration. Fusion was detected by measuring the changes in the excimer to monomer fluorescence intensity ratio (/E/U) of pyrene-labeled phospholipid analogues upon fusion of the pyrenelabeled and unlabeled vesicles. No fusion was observed for vesicles consisting of DMPC, PS from bovine brain or PG from egg yolk upon addition of NaC1 (up to 1 M). However, considerable fusion was evident for vesicles consisting of DMPG or DPPS upon addition of monovalent cations (300 mM to 1 M). Fusion kinetics were fast reaching a plateau after 5 min of addition of cations. The order of efficiency of different monovalent cations to induce the fusion of DMPG vesicles as judged by the changes of the IE/M ratio was L i + > N a + > K + > Cs ÷. DSC-scan of sonicated DMPG vesicles showed, in the absence of salt, a phase transition at 19.2°C with enthalpy of 1.1 kcal • mo1-1. After incubation in the presence of 600 mM NaC1 the DSC scan showed a narrow phase transition at 24.1°C with enthalpy of 6.9 kcal. mole -1 and a pronounced pretransition, both supporting that the fusion of the vesicles had occurred in the presence of NaCI. The results indicate that sonicated vesicles consisting of acidic phospholipids with fully saturated fatty acids fuse in the presence of monovalent cations, whereas those containing unsaturated fatty acids do not.

Keywords: acidic phospholipids; vesicle fusion; monovalent cations; unsaturation.

Introduction

Membrane fusion is of fundamental importance in a wide range of cellular functions such as intracellular membrane transport, endocytosis, exocytosis and neurotransmission. To gain insight into the molecular mechanisms of fusion, phospholipid vesicles have been commonly used as a model system. Perhaps most thoroughly elucidated is the Ca2+-induced fusion of small unilamellar vesicles (SUV) consisting of phosphatid-

Abbreviations: DMPG, 1,2-dimyristoyl-sn-glycero-3-phospho-ras-glycerol; DPPG, 1,2-dipalmitoyl-sn-glycero3-phospho-ras-glycerol; DPPS, 1,2-dipalmitoyl-snglycero-3-phospho-serine: DSC, differential scanning calorimetry; PG, phosphatidylglycerol; PS, phosphatidylserine; Pyr-PC, 1-2-pyrenedecanoyl-sn-glycero-3-phosphocholine; Pyr-PG, 1-palmitoyl-2-pyrenedecanoyl-sn-glycero-3-phospho-sn- 1'-glycerol.

ylserine (PS) [1-3] and refs. cited therein). Addition of Ca 2÷ concentrations higher than 1 mM leads to extensive aggregation [4] and fusion of sonicated PS vesicles [5]. In contrast to Ca 2÷, addition of Mg 2÷ has been shown to induce only limited fusion of PS SUV [6,7] and no fusion of large unilamellar PS vesicles [6,8]. The binding of monovalent cations to PS membrane induces a massive aggregation of small vesicles [9,10], but this aggregation has been shown to be fully reversible and does not lead to fusion of the vesicles [11]. In most of the previous studies on membrane ftlsion the fatty acid composition of the phospholipids used has been heterogeneous with a high degree of unsaturated fatty acids. In only a few earlier studies phospholipids with fully saturated fatty acids have been employed [12-14]. Recently it was shown that the affinity of binding of Li ÷ to PS membranes containing only saturated fatty

0009-3084/90/$03.50 © 1990 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland

200 acids is considerably higher than binding to PS m e m b r a n e s with unsaturated fatty acids [15,16]. It was concluded that binding of Li + depends on the cross-sectional area of the phospholipids in the m e m b r a n e [16]. It is likely that differences in the affinity of cation binding to saturated vs. unsaturated phospholipids are also reflected in cation-induced m e m b r a n e processes such as aggregation and fusion. The excimer to m o n o m e r fluorescence emission intensity ratio (/E/M) of pyrene and its lipid analogues is directly proportional to the local concentration of pyrene in the m e m b r a n e [17-19]. Therefore the IE/M ratio of pyrene-labeled lipids has been used to study various m e m b r a n e processes such as phospholipid phase transitions [19,20], lateral phase separations [21,22] and fusion of virions with phospholipid vesicles [23,24]. We have used pyrene-labeled phospholipid analogues and a probe dilution-type fusion assay to study the cation-induced fusion of acidic phospholipid vesicles. It will be shown, that in contrast to sonicated phospholipid vesicles with unsaturated fatty acids, vesicles consisting of acidic phospholipids with fully saturated fatty acids can be induced to fuse also by monovalent cations. Materials and methods A m m o n i u m salt of 1,2 - dimyristoyl - s n - glycero - 3 - phosphatidyl - r a s - glycerol ( D M P G ) , 1,2 - dipalmitoyl - s n - glycero - 3 - phosphatidyl ras - glycerol ( D P P G ) , 1,2 - dipalmitoyl - s n glycero - 3 - phosphatidylserine (DPPS), 1,2 dimyristoyl - s n - glycero - 3 - phosphatidylcholine ( D M P C ) and phosphatidylserine (PS) from bovine brain and phosphatidylglycerol (PG) prepared from egg yolk were purchased from Sigma Chemicals. The a m m o n i u m salts of I - palmitoyl 2 - pyrenedecanoyl - s n - glycero - 3 - phosphatidyl - s n - 1 - glycerol (Pyr-PG) and 1 - palmitoyl 2 - pyrenedecanoyl - s n - glycero - 3 - phosphatidylcholine (Pyr-PC) were from KSV chemicals (Helsinki, Finland). No impurities were detected in any of these lipids on thin layer chromatography on silica gel (Merck, Darmstatd, F . R . G . ) developed with chloroform/methanol/water/ammonia (65:20:2:2). Lipid concentrations were de-

termined by phosphorus analysis [25]. Sodium and potassium chloride were of ACS grade and other chemicals used were of reagent grade and were obtained from Sigma Chemicals. The water used in experiments was freshly deionized in a Milli-RO/milli Q (Millipore) filtering system. Stock solutions of lipids were made in chloroform. The lipids were pipetted in this solvent which was thereafter removed under a stream of nitrogen and the samples were kept under reduced pressure overnight. Phospholipids were suspended in 10 mM Tris-HC1 p H 7.4 buffer with 0.1 mM E D T A at a t e m p e r a t u r e of 40-50°C for 5 h. Thereafter the liposomes were sonicated for 1 h with a bath type sonicator under argon atmosphere at 30°C (at 50°C for D P P G and DPPS). Optically clear preparations were obtained and any possible remaining large vesicles were removed by centrifugation of the samples for i h at 35,000 x g at 25°C. Final lipid concentration in the fusion assay was 25/zM and experiments were conducted at 30°C. The

fusion

assay

Two vesicle populations were prepared. PyrPG (5 m o l % ) was included into the "labeled" D M P G and egg P G vesicles. Pyr-PC (5-10 m o l % ) was included in labeled vesicles consisting of either D M P C , DPPS or PS from bovine brain. These labeled vesicles were mixed with ten-fold excess of "unlabeled" vesicles consisting of the corresponding phospholipids without pyrene-labeled lipids. The fusion was started by adding concentrated cation solution (3 M) to produce the indicated concentration. U p o n fusion of the labeled vesicles with vesicles devoid of pyrene, the concentration of pyrene-labeled lipids in the m e m b r a n e decreased which led to decreased excimer and increased m o n o m e r fluoresence emission intensities [19]. The ratio of the excimer to m o n o m e r fluorescence intensity (/E/M) was monitored continuously with a SLM-4800 fluorometer with a thermostated cuvette compartment. The experiments were carried out under an argon atmosphere and the samples and cation solutions were saturated thoroughly with argon before measurements. The m o n o m e r fluorescence signal was selected by a m o n o c h r o m a t o r set at 377 nm

201

and the excimer fluorescence with a 470-nm long pass filter. The excitation wavelength was 343 nm. The extent of the fusion was determined by making vesicles (in the presence of cations) which contained D M P G and Pyr-PG in the ratio corresponding to the maximum possible fusion. The IE/M value of these vesicles was then compared with the IE/M value observed after completion of the fusion. However, since the morphology of these vesicles and those produced by fusion may be different, some caution should be taken when considering the absolute extent of the fusion.

DSC experiments The samples for DSC experiments were prepared as described for those for fluorescence measurements. Phospholipid concentration was 0 . 5 - 0 . 7 m g - m 1 - 1 . Before scans, the sonicated liposomes were incubated at 30°C for 15 min in the absence of salt or in the presence of 600 mM NaC1 (added just before incubation). DSC-scans were then performed with a Privalov DASM 4M differential scanning calorimeter at a heating rate of 0.5°C • min -I. Transition temperatures were taken as the maximum deviations from the baseline. Calorimetric enthalpies were determined by calculating the area under the transition peak by paper weighing. Results

No significant change could be observed in IWM of sonicated DMPC liposomes containing 10 mol% of Pyr-PC and in the presence of tenfold excess of unlabeled DMPC vesicles upon addition of NaCI (data not shown). This indicates that no dilution of fluorescent Pyr-PC occurred and thus that no fusion of DMPC vesicles took place upon increasing NaCI concentration. It is also evident that the IEZM ratio of pyrene is not significantly influenced by the increased ionic strength. Figure 1 shows the IE/M VS. time plots for sonicated D M P G vesicles as a function of [NaC1] in a similar fusion assay. No change in the IE/M ratio was observed upon addition of buffer. In the presence of 100 mM NaC1 no fusion was observed but instead an increase in the IE/M ratio was evi-

1.2

100 mM

0.9 0.6

0.3

1M,-"

0 0

2

4

6

8

10

t (min)

Fig. i. The excimer to monomer fluorescence intensity ratio (/E/M) VS. time plots for sonicated D M P G liposomes containing 5 mol% of Pyr-PG and in the presence of ten-fold excess of unlabeled liposomes. At time zero, concentrated NaC1 was added to produce the indicated NaCI concentration. Phospholipid concentration is 25/~M and temperature is 30°C.

dent. However, at 300 mM and higher [NaCI] a clear decrease in the IE/M ratio was observed, indicating the fusion of D M P G vesicles. The decrease of the IE/M ratio was dependent on the concentration of NaC1, and thus the decrease was most pronounced in the presence of 1 M NaC1. The fusion of the vesicles was completed after only 2-5 min. Identical results were obtained using Pyr-PG concentrations of 2.5, 5 and 10 mol% of the labeled vesicles (data not shown), indicating that the observed fusion was not likely to have been induced by perturbing effects of the pyrene label. The comparison of the IE/M values after completion of the fusion to the theoretical maximum values (see Materials and methods) indicated vesicle fusion of approximately 87% from the theoretical maximum in the presence of 600 mM NaC1. Several control experiments were conducted. No change in the IE/M ratio was observed if vesicles after completion of fusion by 600 mM NaC1 were diluted with buffer to 100 mM NaC1 in order to reverse the vesicle aggregation [11]. Figure 2 shows the IE/MVS. time plot for sonicated vesicles consisting of phosphatidylserine from bovine brain in a similar fusion assay. NaC1 was added at time zero to a concentration of 600 mM NaC1. In accordance with all previous studies, no change in the IE/M ratio and thus no indication of fusion

202

1.2

brain

2 shows the 1ElM ratio of sonicated dipalmitoylphosphatidylserine (DPPS) vesicles in the fusion assay upon addition of NaC1. This time a clear decrease of the IE/M ratio was evident indicating that fusion of DPPS vesicles takes place in the presence of 6 0 0 m M NaCi. The fusion rates of brain PS and DPPS vesicles in the presence of 5 - 1 0 r a M Ca 2+ were comparable (data not shown). The fusion of sonicated D M P G and D P P G vesicles by NaCI was studied further using differential scanning calorimetry (DSC). Vesicles were prepared similarly to those used in the fluorescence studies but without pyrene lipids. Figure 3 shows the DSC-scans of sonicated D M P G (panel A) and D P P G (panel B) vesicles after 15 min of incubation at 30°C in the absence of cations (a) and in the presence of 600 mM NaC1 (b,c). The phase transition temperature of sonicated D M P G vesicles was observed at 19. I°C with a very small enthalpy of 1.1 kcal. mol 1 . The DSC-scan of sonicated D P P G vesicles indicated the presence of a more heterogeneous vesicle population after

PS

0.9

0.6

DPPS

0.3

0 0

2

4

6

8

10

t (rain)

Fig. 2. The IE/M vs. time plots for sonicated vesicles of PS from bovine brain and DPPS, containing 5 mol% of fluorescent Pyr-PC and in the presence of ten-fold excess of unlabeled liposomes. At time zero NaC1 was added to produce 600 m M NaCI concentration. Phospholipid concentration is 25 # m and temperature is 30°C.

could be observed for PS vesicles in the presence of NaC1. F u r t h e r m o r e , no fusion was indicated by the assay for vesicles consisting of PG obtained from egg yolk (data not shown). However, Fig.

A

T

"d e"e.. 0 ltl

o

...,1 U. l-ill "r

b

10

20

30

30

4o

50

T,°C Fig. 3. DSC scans for sonicated liposomes of DMPG (panel A) and DPPG (panel B) after incubation of 15 rain at 3(FC in the absence (a) or in the presence (b, c)of 600 mM NaCI. The scanning rate is 0.5°C. min ~and the calibration bar is 50 × 10 ~'W. See text for details.

203

sonication and the DSC trace is very similar to that observed previously for sonicated D P P G [13] and D P P C vesicles [26]. The DSC scan of D P P G immediately after incubation with 600 mM NaCI (Panel B, scan b) shows a narrow main transition at 42.2°C and a pronounced pretransition at 36.1°C, which are both consistent with the presence of only large D P P G vesicles and thus indicate complete fusion of the vesicles by 600 mM NaCI. In the case of D M P G , the first scan after incubation with NaCI indicates that the majority of vesicles are large, with main transition temperature (Tin) at 24. I°C, but also that some smaller vesicles are present. The second scan (c) is consistent with the presence of only large vesicles with Tm at 24.1°C and transition enthalpy of 6 . 9 k c a l . mol -~. The results also indicate that fusion of PG vesicles can occur above and below the phospholipid phase transition temperature, as the incubation temperature (30°C) was above the T,,, of D M P G and below the Tm of D P P G . In addition, the narrow main transitions and pronounced pretransitions observed for both D M P G and D P P G indicate that no significant hydrolysis of the phospholipids had occurred during the sonication of the liposomes. The capability of different monovalent cations to induce the fusion of D M P G vesicles was compared at 600 mM cation concentration. Figure 4 shows the IE/M VS. time plots for D M P G vesicles in the fusion assay upon addition of LiC1, NaCI, KC1 and CsC1 to produce 6 0 0 m M cation concentration. The order of efficiency of these cations to induce the fusion of D M P G vesicles as judged by the IE/M ratio was Li + > Na + > K + > Cs +. This is the same order as their increasing size and the reported binding affinities towards PG membrane [27]. However, the comparison of the extent of fusion produced by different cations was somewhat complicated by the different ability of the cations to induce the condensation of the D M P G membrane. This is illustrated in Fig. 5, which shows the IE/MVS. time plots for D M P G vesicles upon addition to cations to 6 0 0 m M concentration. However, this time 10mol% of Pyr-PG was included in all of the vesicles and no unlabeled recipient vesicles were

1.2

0.9 14.1

0.6

c$÷ K÷,N \

0.3

0 0

2

4

6

8

10

t (min)

Fig. 4. The IE/MVS. time plots for D M P G liposomes containing 5 mol% of Pyr-PG and in the presence of ten-fold excess of unlabeled liposomes. At time zero concentrated solutions of indicated cations were added to produce 600 mM cation concentration. Phospholipid concentration is 25/xM and temperature is 30°C.

present and thus no dilution of the pyrene-labeled phospholipids could occur upon fusion. In this case only an increase of the IE/M ratio could be observed. Furthermore, the order of efficiency of the cations to increase the IE/M ratio was Cs ÷ > K + > Na + > Li +. In order to determine the extent of the fusion produced by different cations, vesicles containing Pyr-PG and D M P G in the ratios corresponding to the maximum possible fusion were prepared and hydrated in the pres-

Cs+,,~ K+~ Nr,z+~,

1.2

Li+7

0.9

0.6

0.3

0 0

2

4

6

8

10

t (min)

Fig. 5. The IE/MVS. time plots for D M P G liposomes containing 10 mol% of Pyr-PG. No unlabeled vesicles are present. Indicated cations were added at time zero to produce 600 mM cation concentration. Lipid concentration is 25/xM and temperature is 30°C.

204 ence of 600 mM concentration of a particular cation. Comparing these IE/M values to those observed after fusion, the extent of fusion was estimated to be 87% for NaC1 and 71%, 64% and 57%, for LiC1, CsC1 and KCI respectively, of the theoretical maximum. Finally, in the preliminary experiments no significant difference was observed in the fusion rates of the different polar headgroup stereoisomers of D M P G by monovalent cations (data not shown). This is in contrast to the differences observed in the aggregation of I ' - D M P G and 3 ' - D M P G by NaC1 in the previous study [28]. However, the aggregation studies were conducted below the phase transition temperature of D M P G and the fusion studies in the present study were performed above the T,, of D M P G . Therefore it seems likely that the differences which have been observed between the two stereoisomers of D M P G are evident only in the gel phase [29,30]. Discussion

Recently, when different fusion assays were compared it was concluded that assays based on the dilution of membrane-associated probes are usually sensitive for the fusion and do not indicate the mere aggregation of the vesicles [31]. In the present study no indication of fusion was observed for DMPC vesicles, vesicles consisting of PS from bovine brain or PG from egg yolk upon addition of monovalent cations. This indicates that the decrease of the IE/M ratio observed for saturated D M P G and DPPS upon addition of monovalent cations is not due to the effect of increased ionic strength on the fluorescence properties of pyrene. Identical results were obtained with D M P G vesicles containing 2.5, 5 and 10 mol% of Pyr-PG (data not shown). Therefore it is unlikely that the vesicle fusion observed had been induced by the membrane perturbing effects of the pyrene residue. In addition, evidence for fusion of D M P G and D P P G vesicles in the presence of monovalent cations was observed also in DSC experiments where no fluorescent probes were used. The exchange of pyrene-labeled probes by spontaneous transfer between vesicles is also possible. However, we consider this very

unlikely as the observed fusion process was rather fast and again no decrease of IE/M occurred with vesicles consisting of brain PS. Furthermore, increasing the ionic strength is likely to inhibit spontaneous lipid transfer [32]. Vesicles were exposed to rather high osmotic stress and thus the breakdown of the vesicles and consequent lipid intermixing cannot be excluded. However, this is considered somewhat unlikely as no lipid intermixing was observed for DMPC (disaturated) or brain PS and egg PG vesicles under similar osmotic stress. Finally, whether the observed fusion represents " t r u e " fusion of the vesicles or so called "semifusion" [33,34] is beyond the scope of the present study. However, the DSC results suggest that complete fusion of the vesicles occurred. In most of the previous fusion studies, the phospholipids used have contained unsaturated fatty acids with only a few exceptions. Using DSC, Papahadjopoulos et al. [13] studied the fusion of D M P G and D P P G vesicles in the presence of 100 mM NaC1. Some evidence for fusion was observed. However, as increasing the concentration of NaC1 inhibited the fusion, results were interpreted to indicate the exchange of phospholipids by diffusion and not a true fusion of the vesicles. The main difference compared to the present study was that vesicles were prepared and sonicated in the presence of NaC1 and thus Na + ions were present at both sides of the membrane before incubation of the vesicles. Stabilization of sonicated vesicles by the presence of Na+-ions inside the vesicles is consistent with the recent suggestion that protonation of the internal bilayer surface of small sonicated PS vesicles would be responsible for the thermodynamic stability of these vesicles [35]. The results of the present study indicate that vesicles consisting of negatively charged phospholipids with fully saturated fatty acids fuse upon addition of monovatent cations, whereas those also containing unsaturated fatty acids do not. Several factors could be involved. First, it is possible that sonicated iiposomes consisting of phospholipids .with only saturated fatty acids are smaller and more unstable than those containing unsaturated fatty acids. Therefore it will be interesting to study whether large unilamellar vesicles

205

(LUV) also can be induced to fuse by addition of monovalent cations. Secondly, it was recently shown that the affinity of binding of Li + to PS membranes is highest when PS is fully saturated and that it decreases upon increasing the amount of unsaturation [15,16]. This could also explain the observed differences in the fusion of vesicles consisting of saturated vs. unsaturated phospholipids in the present study. Thirdly, it has been suggested that in Ca2+-induced fusion, the Ca 2+gradient between the outside and inside of the vesicles is important [14]. This was supported by the observation that increased CaZ+concentration was needed to induce fusion of PS vesicles when CaZ+-ionophore was included in the vesicles [14]. As the vesicles consisting of saturated phospholipids are less permeable to cations compared to unsaturated phospholipids, larger cation gradients are probably present in vesicles with saturated phospholipids which possibly renders them more fusogenic. This would also be consistent with the lack of fusion in the presence of NaC1 when sodium ions were present at both sides of the membranes [13]. In the absence of unlabeled vesicles, when no dilution of pyrene probe could occur upon fusion, an increase in the IE/M ratio was evident for D M P G upon addition of cations. This further indicates that the decreases observed in the IE/M ratio are not due to mere ionic effects or aggregation of the vesicles. This increase of IE/M was even more pronounced upon addition of CaCI and if mono- or divalent cations were added at a temperature below the phase transition temperature of D M P G (data not shown). The reason for this increase of IE/M is not clear. It occurs also using neutral Pyr-PC as a probe and thus it is not likely to be caused by the probe-cation interactions. We consider it more likely to result from the condensation of the acidic phospholipid membrane upon cation binding. Furthermore, the increase of IE/M upon addition of monovalent cations was evident for vesicles of DPPS and D M P G but not for those consisting of brain PS or DMPC (data not shown), indicating probably again different interactions between cations and fully saturated phospholipids [16]. The possibility of changes in IE/M ratio upon binding of cations to

membranes should perhaps be also considered when pyrene-labeled phospholipids are used as probes in studies on lateral phase separation or membrane fusion. To conclude, in addition to polar headgroup structure the phospholipid fatty acid composition seems to be important in the cation induced vesicle fusion. If membranes containing high degree of fully saturated fatty acids are in general more susceptible to fusion this may have important implications in the regulation of the degree of unsaturation of natural membranes. It is also possible that only small sonicated vesicles consisting of fully saturated acidic phospholipids are more labile and fusogenic compared to those containing unsaturated fatty acids. Therefore their use as a model system for studying membrane fusion can be questioned. Finally, the use of liposomes with fully saturated phospholipids in various applications of liposomes, such as delivery of macromolecules into cells or bacteria, could be worth considering.

Acknowledgments Many thanks are due to Drs. P. Kinnunen, J. Virtanen and P. Somerharju for critically reading the manuscript and Maru/kkermann for technical assistance. Financial support was provided by the Emil Aaltonen Foundation and by the Finnish Medical Foundation.

References 1 D. Papahadjopoulos, G. Poste, E.B. Schaeffer and W.J. Vail (1974) Biochim. Biophys. Acta 352, 10-28. 2 J. Bentz, N. Dtizg/ines and S. Nit (1985) Biochemistry 24, 1064-1072. 3 S. Nit, J. Bentz, J. Wilschut and N. Dtizgtines (1983) Progr. Surf. Sci. 13, 1-124. 4 J. Lansman and D.H. Haynes (1975) Biochim. Biophys. Acta 394, 335-347. 5 D. Papahadjopoulos, W. J. Vail, K. Jacobson and G. Poste (1975) Biochim. Biophys. Acta 394,483~,91. 6 J. Wilschut, N. Dtizg0nes and D. Papahadjopoulos (1981) Biochemistry 20, 3126-3133. 7 D. Hoekstra (1982) Biochemistry 21, 2833-2840. 8 J. Bentz and N. Diizgtines (1985) Biochemistry 24, 5436-5443. 9 S. Ohki, N. Dtizgiines and K. Leonards (1982) Biochemistry 21, 2127-2133.

206 10 S. Ohki, S. Roy, H. Ohshima and K. Leonards (1984) Biochemistry 23, 612~6132. 11 E.P. Day, A.Y.W. Kwok, S.K. Hark, J.T. Ho, W.J. Vail, J. Bentz and S. Nir (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 4026--4029. 12 D. Papahadjopoulos, W.J. Vail, W.A. Pangborn and G. Poste (1976a) Biochim. Biophys. Acta 448, 265-283. 13 D. Papahadjopoulos, S. Hui, J. Vail and G. Poste (1976b) Biochim. Biophys. Acta 448, 245-264. 14 D. Papahadjopoulos, W.J. Vail, C. Newton, S. Nir, K.Jacobson, G. Poste and Lazo, R. (1977) Biochim. Biophys. Acta 465, 579-598. 15 H.L. Casal, H.H. Mantsch and H. Hauser (1987a) Biochemistry 26, 4408-4416. 16 H.L. Casal, A. Martin, H.H. Mantsch, F. Paltauf and H. Hauser (1987b) Biochemistry 26, 7395-7401. 17 J.B. Birks, D.J. Dyson and J.H. Munro (1963) Proc. R. Soc. London A 275,575-588. 18 Th. F6rster (1969) Angew. Chem. 81,364-374. 19 H.J. Galla and W. Hartman (1980) Chem. Phys. Lipids 27, 199-219. 20 P.J. Somerharju, J.A. Virtanen, K.K. Eklund, P. Vainio and P.K.J. Kinnunen (1985) Biochemistry 24, 2773-2781. 21 M.E. Jones and B.R. Lentz (1986) Biochemistry 25, 567-574. 22 K.K. Eklund, J. Vuorinen, J. Mikkola, J.A. Virtanen and P.K.J. Kinnunen (1988) Biochemistry 27, 3433-3437. 23 S. Amselem, Y. Barenholz, A. Loyter, S. Nir and D.

24 25 26 27 28 29 30 31 32 33

34 35

Lichtenberg (1986) Biochim. Biophys. Biophys. Acta 860, 301-313. R. Pal, Y. Barenholz and R.R. Wagner (1988) Biochemistry 27, 30-36. P.J. Chen, Jr., T.Y. Toribara and H. Warner (1956) Anal. Chem. 28, 1756-1758. J. Suurkusk, B.R. Lentz, Y. Barenholz, R.L. Biltonen and T.E. Thompson (1976) Biochemistry 15, 1393-1401. M. Eisenberg, T. Gresalfi, T. Riccio and S. McLaughlin (1979) Biochemistry 18, 5213-5223. K.K. Eklund, J.A. Virtanen, K. Vuori, J. Patrikainen and P.K.J. Kinnunen (1987) Biochemistry 26, 7542-7545. T. Lotta, I.S. Salonen, J.A. Virtanen, K.K. Eklund and P.K.J. Kinnunen (1988) Biochemistry 27, 8158-8169. I. Salonen, K.K. Eklund, J.A. Virtanen and P.K.J. Kinnunen (1989) Biochim. Biophys. Acta 982, 205-215. N. Diizgtines, T.M. Allen, J. Fedor and D. Papahadjopoulos (1987) Biochemistry 26, 8435-8442. S.C. Charlton and L.C. Smith (1982) Biochemistry 21, 4023-4030. N. Diizgiines, R.M. Straubinger, P.A. Baldwin, D.S. Friend and D. Papahadjopoulos (1985) Biochemistry 24, 3091-3098. J. Rosenberg, N. Diizgiines and C. Kayalar (1983) Biochim. Biophys. Acta 735, 173-180. R.A. Demel, F. Paltauf and H. Hauser (1987) Biochemistry 26, 8659-8665.