Large-scale preparation of asymmetrically labeled fluorescent lipid vesicles

Large-scale preparation of asymmetrically labeled fluorescent lipid vesicles

ANALYTICAL BIOCI-IRMISTRY 196,46-53 (1991) Large-Scale Preparation of Asymmetrically Fluorescent Lipid Vesicles Hong-Ngoc Thi Dao,l Jonathan C. M...

3MB Sizes 6 Downloads 33 Views

ANALYTICAL

BIOCI-IRMISTRY

196,46-53 (1991)

Large-Scale Preparation of Asymmetrically Fluorescent Lipid Vesicles Hong-Ngoc

Thi Dao,l Jonathan

C. McIntyre,2 and Richard G. Sleight3

Department of Molecular Genetics, Biochemistry and Microbiology, University College of Medicine, 231 Bethesda Avenue, Cincinnati, Ohio 45267-0524

Received

December

of Cincinnati

10, 1990

A method for producing lipid vesicles containing fluorescent phospholipid analogues localized to the inner leaflet of their membrane was developed. Incubation of a 460-fold molar excess of serum albumin with lipid vesicles symmetrically labeled with 1 mol % l-palmitoyl-2-(N-4-nitrobenzo-2-oxa-1,3-diazolyl)amino-caproyl phosphatidylcholine resulted in the removal of 99% of the fluorescent lipid from the outer leaflet. Asymmetrically labeled vesicles were separated from albumin/lipid complexes by gel filtration chromatography. Vesicles prepared in this manner were unable to transfer fluorescent lipid to cells during liposome-cell incubations. Liposomes asymmetrically labeled with other 4-nitrobenzo-2-oxa-1,3-diazole (NBD)-phospholipid analogues were also prepared. Removal of aminododecanoyl-NBD-labeled lipids from the outer leaflet of liposomes required three times more bovine serum albumin, and 48 h of incubation. This method can be used to produce large amounts of asymmetrically labeled liposomes suitable for use in investigating a variety of membrane phenomena. o is01 Academic PWSS. I~C.

Fluorescent lipid analogues containing 4-nitrobenzo2-oxa-1,3diazolylaminocaproic acid (C,-NBD-FA)4 i Present address Department of Medicine, Division of Infectious Disease, Stanford University School of Medicine, Stanford University Medical Center, Stanford, CA 943055107. * J.C.M. was supported by Molecular and Cellular Training Grant HL07527. s This work was supported in part by NIH Grant GM39035 to R.G.S. ’ Abbreviations used: NBD, I-nitrobenzo-2-oxa-1,3diazole; (acyl, C,-NBD)-PE, 1-acyl-2-C,-NBD-phosphatidyl-ethanolamine; (acyl, C,,-NBD)-PC, I-acyl-2-C,,-NBD-phosphatidylcholine; BSA, bovine serum albumin; Cs-NBD-FA, N- [7- (4-nitrobenzo-2-oxa1,3-diazolyl)]aminocaproic acid; C,,-NBD-FA, N-[7-(4-nitrobenzo-2-oxa1,3diazolyl)]aminododecanoic acid; CHO, Chinese hamster ovary; DOPC, dioleoyl phosphatidylcholine; HMEM, 10 rnr.4 4-(2-hydroxy46

Labeled

mimic their natural counterparts in all aspects except solubility (1,2). These analogues have been used in a variety of biochemical and biophysical studies examining cellular lipid trafficking (3,4), membrane phase transitions (5), lipid monomer transfer (6,7), membrane fusion (8), and phospholipid flippase activity (4,9). When C,-NBD-labeled lipids are included in lipid vesicle preparations, they become symmetrically distributed between inner and outer leaflets of the liposomal membrane (10). Selective incorporation of these probes into the outer leaflet can be accomplished by addition of the C,-NBD-phospholipid in ethanol to preformed vesicles (10). Only a few examples of selective incorporation to the inner leaflet have been reported (9-11). Many important regulatory events occur at the inner leaflet of the plasma membrane including the PI cycle, the action of phospholipid exchange proteins, and phospholipid flippase activity. In principle, fusing liposomes labeled exclusively at their inner leaflet with cells would allow these events to be studied with fluorescence techniques. Current methods for producing asymmetrically labeled vesicles are not practical for many applications. These procedures are limited by the amount of vesicles that can be produced in a reasonable time (9,ll). In addition, some of the previously described methods produce vesicles that are not completely asymmetric in their labeling. Herein, we present a method for producing asymmetrically labeled vesicles in large quantities. The chemical and physical properties of C,-NBDphospholipids are the predominant factors that make this method possible. A thorough description of these ethyl)-1-piperazineethanesulfonic acid-buffered Eagle’s minimum essential medium, pH 7.4, without indicator; N-Rh-PE, N-(lissamine rhodamine B sulfonylhlioleoyl phosphatidylethanolamine; (palmitoyl, C, - NBD) - PC, 1 - palmitoyl - 2 - C, - NBD - phosphatidylcholine; (palmitoyl, C,-NBD)-PA, 1-palmitoyl-2-C,-NBD-phosphatidic acid; (palmitoyl, C!, - NBD) - PS, 1 - palmitoyl - 2 - C, - NBD -phosphatidylserine. 0003-2697/91$3.00

Copyright 0 1991 All rights of reproduction

by

Academic Press, Inc. in any form reserved.

PREPARATION

OF

ASYMMETRICALLY

characteristics has been previously reported (1,2). C,NBD-phospholipid derivatives spontaneously transfer between lipid vesicles (6,7). The half times of equilibration of C,-NBD-PC and C,,-NBD-PC between liposomes are 0.51 and 150 min, respectively (7). This rate is directly influenced by the membrane dissociation rates of C,-NBD-lipids, which are controlled by the structure of the analogue (i.e., chain length, degree of saturation, etc.) (12). Although these analogues can move in and out of membranes, they do not spontaneously move across membrane bilayers at a significant rate (1,2). Therefore, analogues in the outer leaflet will equilibrate with the outer surroundings, while the inner leaflet label remains in place. It has been demonstrated that serum albumin binds dispersed monomers of phosphatidylcholine, but does not extract native phosphatidylcholine from liposomes (13,14). Thus, incubating liposomes,. containing C!,NBD-phospholipid, with an excess of bovine serum albumin (BSA) should shift the equilibrium distribution of the labeled analogs from the outer leaflet of the liposomes to BSA. The removal of a small amount of fluorescent lipid can be accomplished without the loss of membrane integrity. This strategy has been employed previously, transferring the label from donor to acceptor liposomes (10,15). Although liposome-based methods are somewhat successful, the use of multiple centrifugations to separate the two populations of vesicles is time consuming and produces a low yield. In this paper, we describe a method for producing asymmetrically-labeled liposomes that is rapid and reproducible. The method produces much larger quantities of labeled liposomes than possible with previous techniques. The ability to rapidly produce large quantities of asymmetrically labeled liposomes should provide investigators with a powerful tool for examining lipid transport. MATERIALS

AND

METHODS

Chemicals Essentially fatty acid-free bovine serum albumin, Sepharose 6B-100, and Triton X-100 were obtained from Sigma Chemical (St. Louis, MO). Dye-binding protein assays were performed using a kit purchased from BioRad (Richmond, CA). Materials for tissue culture were obtained from Gibco (Grand Island, NY). Aquacide II (sodium salt of carboxymethylcellulose, high viscosity) was purchased from Calbiochem (La Jolla, CA). Cell Culture Chinese hamster ovary cells (line Kl) were obtained from the American Type Culture Collection (Rockville, MD). Cells were grown in Eagle’s minimal essential me-

LABELED

47

LIPOSOMES

dium supplemented with 10% fetal calf serum in a water-saturated atmosphere of 5% CO, in air. For microscopy, the cells were grown on glass coverslips. Lipid and Lipid

Vesicles

Dioleoyl phosphatidylcholine (DOPC), (palmitoyl, C,-NBD)-PC, (acyl, C,-NBD)-PE, (acyl, C,,-NBD)-PC, and N-Rh-PE were obtained from Avanti Polar Lipids (Pelham, AL). (Palmitoyl, C,-NBD)-PS and (palmitoyl, C,-NBD)-PA were synthesized from (palmitoyl, C,NBD)-PC by transphosphatidylation using phospholipase D (Boehringer Mannheim, Indianapolis, IN) and a modified version of the procedure of Comfurius (16). Radiolabeled cholesteryl oleate, okate-1-14C, was purchased from New England Nuclear (Wilmington, DE), and used as a nonexchangeable marker for liposomes (10). All lipids were stored in chloroform:methanol (1:l) at -20°C. Phospholipid concentrations were determined by lipid phosphate analysis (17). Lipid vesicles were prepared by ethanol injection using 13.33 pmole/ ml of lipid in ethanol, producing a 7.5% ethanolic solution (18). Vesicles were dialyzed against 10 mM Hepes, 0.9% NaCl, pH 7.4, for 8 h to remove ethanol. Procedures Vesicles

for Preparing

Asymmetrically

Labeled Lipid

Freshly prepared lipid vesicles having a symmetric distribution of C, - NBD - phospholipid and trace amounts of cholesteryl [14C]oleate were incubated at room temperature with fatty acid-free BSA. Incubation times and the ratio of C,-NBD-lipid/BSA used during incubation are detailed under Results and Discussion. The BSA was added from a 100 mg/ml stock in 10 mM Hepes, 0.9% NaCl, pH 7.4. After the incubation, liposomes and BSA/C,-NBD-lipid complexes were separated by Sepharose 6B-100 column chromatography. Radioactivity in the lipid vesicles, which eluted in the void volume, was monitored by scintillation counting. BSA was monitored either by absorbance at 280 nm or by the BioRad dye binding assay. When necessary, vesicles were concentrated in dialysis membranes using Aquacide II. Vesicles with C,-NBD-lipid localized exclusively to the outer leaflet were prepared by adding the C,-NBD-lipid in ethanol to preformed liposomes (10). Energy

Transfer

Assay

The percentage of C,-NBD-phospholipid removed from vesicles by BSA treatment was determined by a modified version of a resonance energy transfer assay originally designed to measure the movement of C,NBD-lipids between liposomes (6,7,11,15). Vesicles were prepared containing 1 mol % C,-NBD-phospholipid, 1 mol % N-Rh-PE, 98 mol % DOPC, and trace amounts of cholesteryl [14C]oleate. The N-Rh-PE was

48

DAO,

MCINTYRE,

used as a nonexchangeable (between vesicles) fluorescence resonance energy transfer acceptor of NBD (excitation, 450 nm). When a loo-fold excess of unlabeled liposomes (acceptor vesicles) is incubated with fluorescent vesicles (donor vesicles), C,-NBD-lipids transfer spontaneously from the outer leaflet of the donors to the acceptors (6,7,11,15). C,-NBD-phospholipid present in acceptor liposomes is not quenched by energy transfer as these liposomes do not contain N-Rh-PE. Thus an increase in NBD fluorescence is observed. The incubation time of acceptors and donors was 10 min for C,-NBD-phospholipids and 6 h for C,,-NBD-phospholipids. At these times, exchange between donors and acceptors had reached equilibrium. When this measurement is made using vesicles before and after BSA treatment, the percentage of exchangeable lipid removed due to BSA treatment can be determined by the equations

FTr = Feq - Fi - FAc

WI

% removed from outer leaflet

= @‘do - V’T~)BSAM’T~L PI In Eq. [ 11,FTr is the NBD fluorescence transferred from donor to acceptor vesicles. This value is obtained by subtracting the apparent fluorescence of the acceptor vesicles alone, FAc, and the fluorescence of the donor vesicles alone, Fi, from the fluorescence at equilibrium of the donor and acceptor vesicles together, F,. In Eq. [2], the subscript o designates measurements using liposomes not treated with BSA and the subscript BSA stands for treated vesicles. Equal amounts of liposomes were used in both treated and untreated measurements as determined by the amount of cholesteryl [14C]oleate present in the vesicles.

Vesicle-Cell

Incubation

Monolayer cultures grown on glass coverslips were washed three times with cold HMEM and then allowed to cool to 2°C for 5 min in an ice-water bath. Vesicles were then added to make a lipid concentration of 50 pM in 1.50 ml. After incubating for 30 min in an ice-water bath, the cells were washed three times with HMEM and examined microscopically.

AND

SLEIGHT

25

$ 8 d x

15 10

5 0 0

10

20

30

Fraction

FIG.

1. Separation of liposomes and BSA by Sepharose 6B-100 column chromatography. A 2.0 X 48.0 cm column of Sepharose 6B100 was first equilibrated in 10 mM Hepes, 0.9% NaCl, pH 7.4. Five ml of a solution containing 70 mg/ml BSA and 0.3 mM DOPC liposomes was run through the column to saturate nonspecific binding sites. The sample load volume varied for different experiments but never exceeded 2.5 ml. In this example a 2.0-ml sample containing 0.25 mM (palmitoyl, C,-NBD)-PC-labeled liposomes (1 mol W) and 75 mg/ml BSA was loaded onto the column after a 20 min incubation. The flow rate was approximately 0.4 ml/min and 1.8-ml fractions were collected. Measurements are presented as the fraction of the total sample loaded and therefore represent the percent recovery in each fraction. (0) Cholesteryl [iqC]oleate (liposome marker); (A) BSA; (0) (palmitoyl, C,-NBD)-PC.

100X planapochromat objective and appropriate filters for visualizing NBD fluorescence. RESULTS

AND

DISCUSSION

Separation of BSA and Liposomes Chromatography

by Column

Asymmetrically labeled liposomes and BSA/C,NBD-phospholipid complexes were separated by column chromatography using Sepharose 6B-100. Columns were preequilibrated with both BSA and liposomes to saturate nonspecific binding sites. A typical column profile for the separation of BSA/(palmitoyl, C,-NBD)-PC complexes and (palmitoyl, C,-NBD)PC-labeled liposomes is shown in Fig. 1. The liposomes, labeled with both cholesteryl [14C]oleate and NBDphospholipid, eluted in the void volume and were recovered quickly. BSA/(palmitoyl, C,-NBD)-PC complexes eluted from the column at a volume far removed from the liposomes.

Instrumentation

Assaying for Asymmetry

Fluorescence measurements were obtained on either a Perkin-Elmer LS-5 fluorescence spectrophotometer or a Specs 1681 0.22 spectrometer. Absorbance was measured using a DU-65 spectrophotometer, Beckman Instruments. Vesicle to cell transfer experiments were observed with a Zeiss IM 35 microscope equipped with a

The raw data obtained from a typical asymmetry assay is presented in Fig. 2. The increase in fluorescence over time is due to the transfer of (palmitoyl, C,-NBD)PC from quenched donor vesicles to unquenched acceptor vesicles (6). The solid line indicates the fluorescence increase over time for donor vesicles that had not been

PREPARATION

c> 20 40

r 1

r---

_----

_

_

if I

OF

ASYMMETRICALLY

_

--

----

1

I

0 0 ii

200

400 Time

600

(seconds)

FIG. 2. Measurement of C,-NBD-lipid transfer from symmetrically labeled and asymmetrically labeled liposomes to nonfluorescent acceptors. Donor vesicles containing 1 mol % (palmitoyl, C,-NBD)PC, 1 mol % N-Rh-PE, and 98 moi % DOPC (final concentration 5 pM) were incubated with nonfluorescent acceptor liposomes (final concentration, 500 @M) at 24°C (solid line). The acceptors were added at t = 30 s as indicated by the arrow A to a total volume of 2 ml. The increase in NBD fluorescence over time is due to the spontaneous transfer of C,-NBD-lipid monomers from donor vesicles containing N-Rh-PE to the nonquenching environment of the acceptors. The experiment was repeated (dashed lines) with liposomes that were pretreated with a 450 molar excess of BSA for 20 min. In this experiment, the acceptor liposomes were added at arrow B. Arrow C indicates the amount of fluorescence due to acceptor liposomes alone. The fluorescence measurements were obtained using a Specs 16810.22 spectrometer (excitation, 470 nm; emission, 530 nm; 2-nm slit widths).

treated with BSA. When acceptor vesicles were added (arrow A), a rapid increase in fluorescence was observed, partially due to the light scattering of the acceptors. The fluorescence continued to increase for approximately 5 min and then plateaued. When the same assay was performed using liposomes treated with BSA (dashed line, arrow B), almost no increase in fluorescence above that of acceptors alone (arrow C) occurred. By subtracting the fluorescence due to the acceptors alone and the initial fluorescence of the donors, the fluorescence due to transfer was determined (F&. The degree of asymmetry can be determined by comparing the values of FTr for BSA-treated and untreated liposomes (see Materials and Methods). To check the validity of these measurements the following control experiment was performed. Two sets of liposomes were prepared each containing 1 mol % (palmitoyl, C,-NBD)-PC. The first set was made with the fluorescent lipid included in the initial preparation, making these vesicles symmetric with respect to the fluorescent label. The second set of liposomes was made by adding the (palmitoyl, C,-NBD)-PC in ethanol to preformed vesicles, which localizes the label exclusively to the outer leaflet of the membrane (10). Asymmetry assays were performed on both sets of vesicles with and without BSA treatment. Whereas all the fluorescence of the all outside labeled vesicles transferred to the accep-

LABELED

LIPOSOMES

49

tors, only 70% of the (palmitoyl, C,-NBD)-PC in the symmetrically labeled vesicles was transferred (Table 1). Since the spontaneous movement of C,-NBD-phospholipids between liposomes occurs by the transfer of monomers in solution, and since C,-NBD-phospholipids do not undergo spontaneous transmembrane movement, only C!,-NBD-lipid from the outer leaflet of the donors was removed (15). Treatment of the donor vesicles with BSA prior to assaying for asymmetry removed 98-99% of the label from the outer leaflet. The fraction of lipid located in the outer leaflet of small unilamellar vesicles has been reported to be in the range of 60-70% (19). Our finding of a (palmitoyl, C,-NBD)-PC distribution of 70% outside/30% inside is consistent with the predicted size of vesicles prepared by ethanol injection (17). When the fraction of lipid located in the outer leaflet was determined by an assay described by Denkins and Schroit (11) values of 62-70% fluorescent lipid in the outer leaflet were obtained (data not shown). This control experiment validates the use of the asymmetry assay for assessing the amount of label localized to the outer leaflet. Ratio of BSAlC,-NBD-phospholipid Required to Produce Asymmetrically Labeled Vesicles To determine the optimal ratio of BSAQ-NBDphospholipid required to produce asymmetrically labeled vesicles, the fraction of (palmitoyl, C,-NBD)-PC removed from the outer leaflet of symmetrically labeled vesicles was determined after incubations with increas-

TABLE Lipid Transfer and without

Liposome

labeling

All outside Symmetric All outside Symmetric

BSA treatment + +

1

from BSA

Vesicles Treatment”

FTr (RFU) 95 67 2 1

with

% NBD-lipid

transferred 100 70 2 1

o Both sets of vesicles were originally prepared containing 1 mol % (palmitoyl, C,-NBD)-PC, 1 mol % N-Rh-PE and 98 mol % DOPC. NBD-labeled lipid was added to the “all outside” vesicles after formation by ethanol injection (10). BSA-treated vesicles were incubated with a 450/l ratio of BSA/(palmitoyl, C,-NBD)-PC for 20 min at room temperature. The fraction of (palmitoyl, C,-NBD)-PC located on the outer leaflet of the vesicles was determined by transport to acceptor vesicles composed of DOPC as described under Materials and Methods. To determine the total amount of (palmitoyl, CsNBD)-PC present in each vesicle preparation, the fluorescence of the assay mixtures was determined after the addition of Triton X-100 (15). The values of Fm were then adjusted relative to the fluorescence measured after the addition of Triton. * Fr, is the relative units of NBD fluorescence transferred from the outer leaflet of the donors to the acceptors.

50

DAO,

MCINTYRE,

ing amounts of BSA (Fig. 3). At low molar ratios of BSA/(palmitoyl, C,-NBD)-PC, a considerable amount of the outer leaflet label was removed. Much of this is probably due to transfer from the vesicles to the column which is pretreated with liposomes and BSA. However, a much larger molar excess was required to obtain complete transfer. Above a BSA/(palmitoyl, C,-NBD)-PC ratio of 30011, 98% of the outer leaflet label was removed after a 20-min incubation. We considered these vesicles to be completely asymmetric with respect to the fluorescent PC, because our assay has a 2-3% margin of error. A BSA/(palmitoyl, C,-NBD)-PC ratio of 450/l was used for the remainder of our studies, since this ratio caused complete removal and permitted prudent use of materials. C,-NBD-phospholipids have a high partition coefficient for DOPC membranes (6,7,12). A large excess of BSA relative to (palmitoyl, C,-NBD)-PC was required to drive the equilibrium distribution toward BSA. BSA has three to four high affinity hydrophobic binding sites for phosphatidylcholine and the capacity to bind as many as 20 phosphatidylcholine monomers (13,14). The great excess of BSA used did not appear to effect the integrity of the membrane, since fluorescent lipid located in the inner leaflet of the membrane could not be removed by the protein. Once liposomes having an asymmetric distribution of fluorescent PC were prepared (i.e., inner leaflet labeled), repeated incubations with BSA and chromatography did not result in the removal of more fluorescent lipid. In addition, soluble contents markers such as rhodamine-labeled dextran are not released from the vesicles after treatment with BSA (data not shown). Liposomes separated from BSA by column chromatography did not have BSA asso-

g

100

8 -I b 2

90

0

E 0 .k -0 F g B w

80

1

70

;

200 Mole

BSA /

400 Mole

(polmiotyl.

600 Ce-NBD)-PC

FIG. 3. Effect of increasing BSA concentrations on the removal of (palmitoyl, C,-NBD)-PC from symmetrically labeled liposomes. Liposomes containing 1 mol % (palmitoyl, C,-NBD)-PC, 1 mol % NRh-PE, and 98 mol % DOPC were incubated with up to a 600-fold molar excess of BSA for 20 min at room temperature, then separated by column chromatography. The liposomes isolated from the column were assayed for lipid asymmetry and compared to untreated liposomes as described under Materials and Methods. The data is presented as the average of at least three determinations &SD. In some instances the error bars are smaller than the size of the symbol.

AND

SLEIGHT

0

10

20 Time

30

(min)

FIG. 4. Effect of incubation time on the removal of (palmitoyl, CsNBD)-PC from liposomes. Liposomes containing 1 mol % (palmitoyl, Cs-NBD)-PC, 1 mol % N-Rh-PE, and 98 mol % DOPC were incubated for various times with a &O-fold molar excess of BSA at room temperature. The liposomes were isolated by column chromatography and assayed for asymmetry as described under Materials and Methods. The data is presented as the average of at least three determinations *SD. In some instances the error bars are smaller than the size of the symbol.

ciated with them (Fig. 1). Since commercially obtained fatty acid-free BSA may contain a small amount of fatty acid, it is possible that some nonfluorescent fatty acid transfers from the protein to lipid vesicles. However, we have not detected free fatty acids in vesicles (1 pmol lipid) treated with BSA, suggesting that the amount transferred is extremely small. Time course of Fluorescent Acid-Free BSA

Lipid

Exchange

to Fatty

To determine the minimum time necessary for producing asymmetrically labeled vesicles, a time course of BSA-liposome incubations was performed. The published half time for the equilibration of (palmitoyl, C,NBD)-PC between lipid vesicles is 0.51 min (7), suggesting that the required incubation time for equilibration of C,-NBD-labeled derivatives with BSA is on the order of minutes. After a 1-min incubation with a 450-fold excess of BSA, 78% of the outer leaflet label was removed from (palmitoyl, C,-NBD)-PC-labeledvesicles (Fig. 4). At least 10 min of incubation with (palmitoyl, C!,-NBD)-PC-labeled vesicles was required for complete transfer. Incubation times of 20 min were used throughout the remainder of our work to ensure that equilibrium had been established prior to separation of liposomes. from BSA/C,-NBD-lipid complexes. Liposome to Cell Transfer C,-NBD)-PC

of Outer Leaflet (Palmitoyl,

We and others have shown that the plasma membrane of cells can be labeled with fluorescent (palmitoyl, C,-NBD)-PC by incubation with symmetrically labeled fluorescent liposomes (20). This occurs by spontaneous

PREPARATION

OF

ASYMMETRICALLY

LABELED

51

LIPOSOMES

FIG. 5. Transfer of (palmitoyl, C,-NBD)-PC from liposomes to cells. CHO-Kl cells were incubated with liposomes at 2°C for 30 min then washed and photographed. Capital letters indicate fluorescence photomicrographs and lowercase letters indicate the phase-contrast photomicrographs of the same cells. Photographic exposure times and printing times were identical for all fluorescence photomicrographs. (A,a) Cells incubated with 50 pM symmetrically labeled vesicles containing 10 mol % (palmitoyl, C,-NBD)-PC; (B,b) cells incubated with 50 pM vesicles labeled asymmetrically (inside only) with 10 mol % (palmitoyl, C,-NBD)-PC before BSA treatment; (C,c) cells incubated with 250 FM of the same vesicles in (B,b).

transfer of lipid monomers from liposomes to cells and is not the result of liposome-cell fusion (3,4,20). As a control for assessing the degree of asymmetry obtained by our method, the transfer of outer leaflet label from (palmitoyl, C,-NBD)-PC labeled vesicles to CHO-Kl cells was observed. Intense labeling of the plasma membrane was seen when symmetrically labeled liposomes were incubated with cells (Fig. 5A). However, when the same amount of BSA-treated lipid vesicles were incubated with cells, no plasma membrane labeling could be photographed (Fig. 5B). Incubating cells with 5 times more of the BSA treated liposomes still produced no labeling (Fig. 5C). We observed a small amount of patchy fluorescence at the plasma membrane while examining cells that had been incubated with asymmetrically labeled vesicles. This fluorescence was too dim to photograph. To determine if this fluorescence was due to lipid-monomer transfer or to liposome sticking, a control experiment with liposomes asymmetrically labeled with (palmitoyl, C!,-NBD)-PC and also labeled with N-Rh-PE was performed. These liposomes were incubated with cells as

described under Materials and Methods. The cells were then harvested and both NBD and Rh fluorescence were determined. Since N-Rh-PE cannot spontaneously transfer between membranes (1,2), any rhodamine fluorescence observed is due to liposome sticking. Therefore, the ratio of NBD/Rh fluorescence of cells after vesicle-cell incubations is an indicator of transfer vs sticking. We found no difference in the ratio of NBD/ Rh fluorescence in cells incubated with liposomes labeled only at the inner leaflet as compared to these liposomes alone (data not shown). This suggests that the very dim fluorescence associated with cells after incubation with asymmetrically labeled vesicles is due to the sticking of a small number of liposomes, rather than lipid transfer. Producing Asymmetrically Labeled NBD-phospholipid Analogues

Vesicles with

Other

Our method for producing asymmetrically labeledvesicles was applied to liposomes prepared with a variety of C,-NBD-phospholipids (PS, PE, and PA). Only C,-

52

DAO,

MCINTYRE,

NBD-PS showed a difference in percentage removed from the outer leaflet (Table 2) after a ZO-min incubation with a 450/l ratio of BSA/C,-NBD-phospholipid. Incubation with BSA removed 95% of the outer leaflet label (palmitoyl, C,-NBD)-PS under these conditions. The difference in asymmetry seen with the fluorescent PS may be due to differences in the dissociation rate of the PS analogue from the membrane or a lower affinity of BSA for the fluorescent PS. The optimal ratio of BSA/C,,-NBD-phospholipid and the minimum time of incubation with BSA was determined for vesicles symmetrically labeled with (acyl, C,,-NBD)-PC. The kinetics of transfer for (acyl, C,,NBD)-PC were comparatively slow relative to C,-NBDphospholipids. After a 48-h incubation, with a BSAI (acyl, C,,-NBD)-PC ratio of 450/l, the transfer of (palmitoyl, C,,-NBD)-PC to the BSA was only 91% complete (Fig. 6). Increasing the molar excess of BSA to 1350 during a 48-h incubation resulted in removing 97% of the outer leaflet label. The longer incubation time needed to remove (acyl, C,,-NBD)-PC, compared to C!,NBD-lipids, is probably due to the longer half time of between vesitransfer (t,,Z is 150 min for equilibration cles) (7). This slow transfer is a result of a slower dissociation rate from donor membranes (12). Three times more BSA was required to remove all of the C,,-NBDlabeled analogues. This may be due to a lower affinity of C,,-NBD-phospholipids for BSA and/or a higher affinity of C,,-NBD derivatives for membranes, relative to C,-NBD-phospholipids. Conclusion

An improved method labeled fluorescent lipid been presented. Other asymmetrically labeled (lOJ5). These methods

for producing asymmetrically vesicles (all inside label) has methods for the production of vesicles have been reported are not feasible for producing

TABLE

Comparison a Variety Cs-NBD-phospholipid (Palmitoyl, C,-NBD)-PC (Palmitoyl, C,-NBD)-PA (Acyl, C,-NBD)-PE (Palmitoyl, C,-NBD)-PS

2

of Labeling Asymmetry Using of Cs-NBD-Labeled Lipids” % removed

from

outer

leaflet

99 98 98 95

“Vesicles were prepared containing 1 mol % C,-NBD-phospholipid, 1 mol % N-Rh-PE, and 98 mol % DOPC. Liposomes were incubated with a 459fold molar excess of BSA for 20 min at room temperature then isolated by chromatography. Percentage C,NBD-phospholipid removed from the outer membrane leaflet of the liposomes was determined as described under Materials and Methods.

AND

SLEIGHT

a

0

16

24 Time

32

40

48

(h)

FIG. 6. Effect of incubation time the removal of (acyl, C,,-NBD)PC from liposomes. Liposomes containing 1 mol % (acyl, C,,-NBD)PC, 1 mol % N-Rh-PE, and 98% DOPC were incubated for various times with a 450- (0) or 1350-fold (A) molar excess of BSA at room temperature. The samples were then separated by column chromatography and assayed for asymmetry as described under Materials and Methods. The data is presented as the average of at least three determinations *SD. In some instances the error bars are smaller than the size of the symbol.

large amounts of vesicles, and are very time consuming as several long centrifugation steps are required. Our method utilizes the spontaneous movement of C,NBD-phospholipids (6,7,15) from the outer leaflet of liposomes to acceptors, combined with the affinity of BSA for phospholipid monomers (13,14). A 450-fold molar excess of BSA incubated with lipid vesicles symmetrically labeled with (palmitoyl, C,-NBD)-PC resulted in the complete transfer of the outer leaflet label to BSA and the concomitant formation of asymmetrically labeled vesicles. The use of Sepharose 6B column chromatography to separate lipid vesicles from BSA after incubation provided the means to scale up the process for the production of vesicle quantities necessary for cell biological studies. The only limitation on the amount of liposomes that can be prepared is column size. We routinely produce asymmetrically labeled vesicles starting with 25 pmol lipid and using a 5 X 48 cm column. Liposome to cell transfer experiments verified the complete loss of the outer leaflet label from vesicles prepared by our method. This method was successful in producing asymmetrically labeled liposomes with other NBD-phospholipids. However, for C,,-NBD-phospholipids very long (48 h) incubation times and large BSA/ C,,-NBD-lipid ratios (1350/l) were required to obtain similar results. REFERENCES 1. Sleight, 2. Pagano,

R. G. (1987) Annu. R. E., and Sleight,

Reu. Physid. R. G. (1985)

49,193-208. Science 229,1051-1057.

3. Sleight,

R. G., and Abanto,

M. N. (1989)

J. Cell Sci. 93,363-374.

PREPARATION

4. Sleight, 1154.

R. G., andpagano,

5. Ellens, H., Bentz, 4141-4147. 6. Nichols, 2789. 7. Nichols, 1726.

R. E. (1985)

J., and

Szoka,

OF

J. Biol.

F. C. (1986)

ASYMMETRICALLY

Chem.

260,1146-

Biochemistry

26,

J. W., and Pagano,

R. E. (1981)

Biochemistry

20,2783-

J. W., and Pagano,

R. E. (1982)

Biochemistry

21,1720-

8. Struck, D. K., Hoekstra, try 20,4093-4099. 9. Conner, J., and Schroit,

D., and Pagano, A. J. (1988)

R. E. (1981)

Biochemistry

Biochemis-

27,848-851.

10. Conner, J., and Schroit, A. J. (1987) Biochemistry 26,5099-5105. 11. Denkins, Y. M., and Schroit, A. J. (1986) Biochim. Biophys. A& 862,343-351. 12. Nichols, J. W. (1985) Biochemistry 24, 6390-6398.

LABELED

53

LIPOSOMES

13. Jonas,

A. (1975)

Biochem.

Biophys.

Res. Commun.

14. Jonas,

A. (1976)

Biochim.

Biophys.

Actu

15. Pagano, R. E., Martin, 0. C., Schroit, (1981) Biochemistry 20,4920-4927. 16. Comfurius, P., and Zwaal, R. F. A. (1977) 488,36-42. 17. Rouser, G., Siakotos, A. N., and Fleisher, 86. 18. Kremer, Wiersema,

64,1003-1008.

427,325-336. A. J., and

Struck,

Biochim.

Biophys.

S. (1966)

Lipids

J. M. H., Esker, M. W. J., Pathmamanohan, P. H. (1977) Biochemistry 16,3932-3935.

D. K., and Pagano,

R. E. (1980)

Biochemistry

Acta

1,85K.,

19. Szoka, F., and Papahadjopoulos, D. (1981) in Liposomes: Physical Structure to Therapeutic Applications (Knight, Ed.), pp. 51-82, Elsevier, New York. 20. Struck, 5411.

D. K.

and

From C. G.,

11, 5404-