journal of biological vol. no. 12, 1987 in transbilayer ... · the journal of biological chemistry...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc. Vol. 262, No. 12, Issue of April 25, pp. 5890-5898, 1987

Printed in U.S.A.

Transbilayer Movement of Fluorescent Analogs of Phosphatidylserine and Phosphatidylethanolamine at the Plasma Membrane of Cultured Cells EVIDENCE FOR A PROTEIN-MEDIATED AND ATP-DEPENDENT PROCESS(ES)*

(Received for publication, December 17,1986)

Ona C. Martin and Richard E. Paganot From the Department of Embryology, Carnegie Institution of Washington, Baltimore, Maryland 21210-3301

The internalization of fluorescent analogs of phosphatidylserine and phosphatidylethanolamine following their insertion into the plasma membrane of cultured Chinese hamster fibroblasts was examined. When liposomes containing the fluorescent lipid 1,2- (palmitoyl-N-4-nitrobenzo-2-oxa-1,3-diazole-aminocaproyl) phosphatidylserine ((palmitoyl-C6-NBD)-PS), were incubated with monolayer cell cultures at 2 “C, spontaneous transfer of the fluorescent lipid from the liposomes to the cells occurred, resulting in prominent labeling of the plasma membrane. However, if the cells were washed and warmed to 7 “C for 30 min, the (palmitoyl-C6-NBD)-PS also labeled numerous intracellular membranes. Evidence is presented suggesting that this internalization was not due to endocytosis, but was the result of transmembrane movement of the (palmitoyl-C6-NBD)-PS art the plasma membrane followed by translocation of lipid monomers from the plasma membrane to internal membranes. This transmembrane movement was reversibly inhibited by depletion of cellular ATP levels and was blocked by treatment with structural analogs of the lipid or by pretreatment of cells with glutaraldehyde or N-ethylmaleimide. A fluorescent analog of phosphatidylethanolamine ((palmitoyl-C6-NBD)-PE), which also exhibits transmembrane movement at the plasma membrane at 7 “C (Sleight, R. G., and Pagano, R. E. (1985) J. Biol. Chem. 260, 1146-1154), was further studied. Its transmembrane movement was also inhibited by depletion of cellular ATP levels, or by pretreatment of cells with N-ethylmaleimide. The transmembrane movement of the fluorescent phosphatidylserine and phosphatidylethanolamine analogs was inhibited when the unnatural D-isomers of these lipids were used, further suggesting that this process was stereospecific and therefore likely to have been protein-mediated.

While partial asymmetry of lipids has been reported in plasma membrane preparations from nucleated cells (Ete- madi, 1980; Houslay and Stanley, 1982; Storch and Kleinfeld, 1985), little information on the maintenance or dynamics of that asymmetry is available. In this report, we examine the behavior of fluorescent analogs of phosphatidylserine ((pal-

* This work was supported by United States Public Health Service Grant GM-22942. The costs of publication of this article were de- frayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed Carnegie Insti- tution, 115 W. University Parkway, Baltimore, MD 21210-3301.

mitoyl-Cfi-NBD)-PS)’ and phosphatidylethanolamine in liv- ing cultured fibroblasts and show that these lipids were ca- pable of rapid transbilayer movement at the plasma membrane. This process occurred at low temperatures where endocytosis is blocked, and was both ATP-dependent and stereospecific, leading us to speculate that a protein(s), perhaps analogous to those recently described in erythrocytes (Seig- neuret and Devaux, 1984; Daleke and Huestis, 1985; Tilley et d., 1986) or rat liver microsomes (Bishop and Bell, 1985; Backer and Dawidowicz, 1985) may be involved in the maintenance of plasma membrane lipid asymmetry in these nucleated cells.

EXPERIMENTAL PROCEDURES

Materials-The following chemicals were purchased from the indicated sources: DOPC and (palmitoyl-Cs-NBD)-PC (Avanti Bio- chemicals, Birmingham, AL); L-DPPC and D-DPPC (Serdary Re- search Laboratories, Port Huron, MI); carbonyldiimidazole (Aldrich); NBD-hexanoic acid (Molecular Probes, Eugene, OR); phospholipase D (cabbage) (Boehringer Mannheim); [l-”Clpalmitic acid (New Eng- land Nuclear). Solvents were from Burdick and Jackson (Muskegon, MI), and all other reagents were from Sigma. SRh-SAD was synthesized as described (Wolf et al., 1977) using a -71,000 molecular weight dextran as the starting material.

Fluorescent Phospholipids-Fluorescent PS and PE were synthesized from (palmitoyl-C6-NBD)-PC by transphosphatidylation (Com- furius and Zwaal, 1977) using phospholipase D and L-serine or ethanolamine, respectively, and purified by preparative TLC using chloroform/methanol/acetic acid/water (90:4012:2) (Solvent 1) as the developing solvent ( R F for PS = 0.26; RF for PE = 0.58).

For the experiments described in Table I11 and Figs. 5 and 6, L- and D-(palmitoyl-Ce-NBD)-PS were synthesized as follows from L- DPPC and D-DPPC, respectively. L-DPPC and D-DPPC were hydro- lyzed separately in the presence of base (Brockerhoff and Yurkowski, 1965), and 0.05 mmol of each was dried separately in uucuo. Each glycerophosphorylcholine was combined with a 1.5-fold excess of NBD-aminocaproylimidaazole and a 7.5-fold excess of palmitoylimidazole (Longmuir et al., 1985), stirred at 62 “C for 18 h in 0.5 ml distilled benzene, and cooled to room temperature. An additional 7.5-

The abbreviations used are: (palmitoyl-Cs-NBD)-Pc, 1,2-(palmitoyl-NBD-aminocaproy1)phosphatidylcholine; (palmitoyl-Cfi- NBD)-PS, 1,2-(palmitoyl-NBD-aminocaproyl)phosphatidylserine; (palmitoyl-C6-NBD)-PE, 1,2-(palmitoyl-NBD-aminocaproyl)phosphatidylethanolamine; (C6-NBD)-PS, 1,2-(acyl-NBD-aminoca- proy1)phosphatidylserine; DOPC, dioleoyl phosphatidylcholine; DPPC, dipalmitoyl phosphatidylcholine; HCMF, 10 mM 4-(2-hydrox- yethy1)-l-piperazineethane sulfonic acid-buffered Puck’s saline without calcium and magnesium; HMEM, 10 mM 4-(2-hydroxyethyl)-l- piperazineethane sulfonic acid-buffered, Eagle’s minimal essential medium, pH 7.4, without indicator; NBD, 4-nitrobenzo-2-oxa-1,3- diazole; NEM, N-ethylmaleimide; N-Rh-PE, N-(lissamine rhodamine B sulfonyl)-dioleoylphosphatidylethanolamine; PIPES, piperazine- N,N’-bis(2-ethanesulfonic acid); PE, phosphatidylethanolamine; PS, phosphatidylserine; RF, mobility relative to solvent front; SRh-SAD, sulforhodamine-labeled stearoyl amino dextran; TLC, thin-layer chromatography.

5890

Transbilayer Movement of Fluorescent Aminophospholipids 5891

fold excess of palmitoylimidazole in 0.5 ml distilled benzene was added to each reaction and stirred for 24 h at 62 "C. The reaction mixtures containing the crude L- or D-isomers of (palmitoyl-C6- NBD)-PC were cooled to room temperature, dried under nitrogen, extracted (Bligh and Dyer, 1959), and purified by preparative TLC using chloroform/methanol/ammonium hydroxide (6535:5) (Solvent 2) as the developing solvent (RF = 0.16). The corresponding L- and ~-(palmitoyl-C~-NBD)-Ps and -PE derivatives were prepared by transphosphatidylation using phospholipase D and L-serine or ethanolamine, respectively (Comfurius and Zwaal, 1977), and purified by preparative TLC using Solvent 1 as the developing solvent. The mobilities of both isomers of each lipid were identical on TLC using either Solvent 1 or 2 as the developing system. In high pressure liquid chromatography analysis (Martin and Pagano, 1986), both isomers had identical retention volumes. The D-isomers were resistant to hydrolysis by phospholipase Az (Kates, 1972).

containing (palmitoyl-C6-NBD)-PS and DOPC (20/80 mol %) were Lipid Vesicles-For experiments using fluorescent PS, lipid vesicles

prepared by ethanol injection (Kremer et al., 1977) in HCMF and dialyzed overnight against HMEM, or glucose-free HMEM as indicated. In some experiments the lipid vesicles also contained 1 mol % N-Rh-PE, a nonexchangeable lipid (Struck and Pagano, 1980). For the experiments described in Fig. 5 and Table 111, vesicles were prepared by the same method but contained 20 mol % either L- or D- (palmitoyl-C6-NBD)-PS and 80 mol % DOPC. For experiments using fluorescent PE, vesicles containing 50 mol % (palmitoyl-C6-NBD)- PE and 50 mol % DOPC were made as described (Sleight and Pagano, 1985).

Cell Culture-Monolayer cultures of Chinese hamster V79 lung fibroblasts (Ford and Yerganian, 1958) were grown for biochemical experiments and for microscopy as described (Sleight and Pagano, 1984).

Vesicle-Cell Incubations-For microscopy, monolayer cultures grown on glass coverslips were washed three times with HMEM and chilled for 15 min on an ice-water bath. The medium was removed, and the cells were incubated for 30 min on ice-water with a vesicle suspension containing fluorescent PS or PE (100 p~ total lipid) in HMEM. The cells were then washed three times with HMEM. In some experiments, the washed cells were further incubated at 7 "C for 30 min (fluorescent PS) or 2 h (fluorescent PE), washed, and photographed. (Relatively little fluorescent PE was internalized after 30 min at 7 "C, and a longer incubation time was therefore used so that internalization could be easily photographed.) For the experiments in Tables I1 and 111, the cells were then chilled and incubated at 2 "C (three times, 10 min) with HMEM containing 250 p~ DOPC vesicles (back-exchange) or with HMEM alone (control). For the experiment described in Fig. 3, cells were washed three times with HMEM, incubated for 30 min at 2 "C with liposomes containing (palmitoyl-C6-NBD)-PS, washed with HMEM, and further incubated for 10 min at 2 "C with 100 pg of SRh-SAD/ml of HMEM. The cells were then washed, warmed for 30 minutes at 7 "C, washed, and photographed.

For biochemical experiments, incubation conditions were identical to those described above, except that monolayer cultures grown on plastic culture dishes were used. After the appropriate incubation, cells were harvested using a rubber policeman, and aliquots were removed for determination of DNA content (Leyva and Kelley, 1974) and for lipid extraction (Bligh and Dyer, 1959). The relative fluorescence of the lipid extracts was determined using an Aminco-Bowman spectrophotofluorometer and normalized to the DNA content. The amounts of fluorescent NBD ( L = 470 nm; Lrn = 530 nm) and rhodamine ( L = 560 nm; L, = 575 nm) lipids present in the cell extracts were determined by reference to standard curves generated from known amounts of (palmitoyl-C6-NBD)-lipids or N-Rh-PE in CHCl$CH30H (2:1, v/v). Lipid extracts were analyzed qualitatively by TLC on Silica Gel 60 thin-layer plates (Merck) using Solvent 2 as the developing solvent.

Determination of Quantum Yield in Cells-Cell suspensions were prepared from monolayer cultures by treatment with 0.05% trypsin in HCMF for 12 min at 37 "C. 5 X 106 cells/ml were suspended in HMEM containing 50 p~ lipid vesicles composed of (palmitoyl-C6- NBD)-PS/DOPC (2080, mol/mol). After incubation for 30 min at 2 "C, the cells were washed three times in cold HCMF with the last wash performed in a new tube. The cells were then incubated at 7 "C for various times, washed three times with cold HCMF, and the NBD fluorescence determined ( L = 470 nm; Lrn = 530 nm) in the spectrophotofluorometer, first in the absence, and then in the presence of 1% Triton X-100.

Deacylation/Reacylation Study-Monolayer cultures were grown for 3 days in complete medium containing 0.5 pCi/ml [1-"Clpalmitic acid complexed to defatted bovine serum albumin (Chen, 1967). The cells were washed with HMEM, chilled, and incubated for 30 min at 2 "C with vesicles containing (palmitoyl-C~-NBD)-PS, washed, and warmed for 30 min at 7 "C. The cells were then washed, harvested with a rubber policeman, extracted, and the extract subjected to two- dimensional TLC using Solvent 2 in the first dimension and chloroform/methanol/petroleum ether/acetic acid/boric acid (4020:3&101.8; Gilfillan et al., 1983) in the second dimension, followed by autoradiography.

Measurement of ATP Levels in Control and ATP-depleted Cells- Samples of control and ATP-depleted cells were obtained as follows. Triplicate monolayer cultures (100-mm dishes) were washed three times with HMEM and incubated 30 min at 37 "C in HMEM (control cells). Alternatively, cultures were washed three times with HMEM lacking glucose and incubated 30 min at 37 'C in HMEM lacking glucose but containing 5 mM sodium azide and 50 mM 2-deoxyglucose (ATP-depleted cells). All dishes were washed with HMEM lacking glucose, treated with a 0.25% trypsin solution at 37 "C for the mini- mum time required for dissociation (approximately 3 min), centrifuged, and resuspended in 5 ml of HMEM lacking glucose. Triplicate 0.6-ml aliquots were quickly removed and digested immediately using perchloric acid (see below). Another triplicate set of 0.6-ml aliquots was removed and frozen at -20 "C for assay of DNA content (Leyva and Kelly, 1974).

Cell suspensions (0.6 ml) containing approximately 1 X lo6 cells were prepared for ATP measurement essentially as described (Weigul and Englund, 1975) using 0.25 ml of ice-cold 10% perchloric acid. After 15 min on ice, the samples were centrifuged at 8800 X g for 5 min. 0.25 ml of the supernatant was transferred to a new tube, neutralized with 0.065 ml of 1.47 N KOH in 25 mM Tris, pH 8.0, and centrifuged at 8800 X g for 5 min. 0.25 ml of the supernatant was diluted 10-fold with 0.05 M glycine diluent containing 0.15 mM luciferin (Sigma F-4755). 100 pl of this diluted supernatant was injected into 100 pl of glycine diluent containing 10 pg of luciferase (Sigma L-5256), and the peak flash height emission was detected by a photomultiplier and recorded. ATP levels in the cell digests were determined by comparison to a standard curve prepared using known amounts of ATP. The standard curve was linear over a range of to lo-' mol of ATP, and ATP levels of the cell digests were within this range.

Microinjection of (Palmitoyl-C6-NBD)-PS-Microinjection of sin- gle cells with the L- or D-isomer of (palmitoyl-C6-NBD)-PS was performed as described (Pagano, 1983; Pagano and Sleight, 1985) using 125 mM KC1,2 mM PIPES, pH 7.0, as the microinjection buffer. Self-quenched vesicles made from (palmitoyl-C6-NBD)-PS alone (Tanaka and Schroit, 1983), or (palmitoyl-C6-NBD)-PS/defatted bovine serum albumin were microinjected. Results using either carrier were identical. Cells were observed in the fluorescence microscope immediately after microinjection.

Miscellaneous Procedure-Cells were treated with inhibitors as described in Table I. The size of the exchangeable pool of (palmitoyl- C6-NBD)-PS in unilamellar vesicles and its ability to undergo transmembrane movement in unilamellar vesicles was studied using previously described assays based on resonance energy transfer between the NBD lipid and N-Rh-PE (Pagano et al., 1981b; Pagano and Longmuir, 1985) Fluorescence microscopy was performed with a Zeiss IM-35 inverted microscope equipped with a Planapo 100 X (1.3 n.a.) objective and barrier filters which allowed no crossover of NBD and rhodamine fluorescence. Photomicrographs were taken using Kodak Tri-X film which was processed at ASA 1600 with Diafhe developer. Lipid concentrations were determined by phosphorus analysis (Rouser et al., 1966).

RESULTS

Uptake and Internalization of (Palmitoyl-C6-NBD)-Ps- When cells were incubated with DOPC vesicles containing N- Rh-PE and (palmitoyl-CG-NBD)-PS for 30 min at 2 "C, a significant amount (-50 pmollpg DNA) of the NBD lipid was transferred to cells. Less than 2% of the cell-associated NBD lipid was due to the presence of intact vesicles, as calculated by the presence of N-Rh-PE, a nonexchangeable lipid (data not shown). This indicated that most of the uptake of (palmitoyl-C6-NBD)-PS was due to its preferential transfer and not to association of intact vesicles with cells (Struck and

5892 Transbiluyer Movement of Fluorescent Aminophospholipids

Pagano, 1980; Pagano et al., 1981c; Struck et al., 1981; Pagano et al., 1983). TLC analysis indicated that the only fluorescent lipid present in the cells or medium after the 2 "C treatment was (palmitoyl-C6-NBD)-PS. The uptake of (palmitoyl-C6- NBD)-PS at 2 "C was not affected by the inclusion of up to 50 mol % cholesterol in the lipid vesicles, or by the omission of divalent cations from the vesicle-cell incubation medium (not shown). The amount of uptake of (palmitoyl-C6-NBD)- PS by cells during the 2 "C incubation was independent of the stereochemistry of the lipid and the ATP content of the cells (data not shown).

When cells were incubated with liposomes containing (palmitoyl-C6-NBD)-PS for 30 min at 2 "C, washed, and examined immediately in the fluorescence microscope, the plasma membranes of the cells were highly fluorescent, but little intracellular fluorescence was apparent (Fig. lA). However, if the cells which were incubated at 2 "C with (palmitoyl-C,-NBD)- PS were washed and warmed for 30 min at 7 "C and then examined, the fluorescence appeared both at the plasma mem-

I FIG. 1. Insertion and internalization of fluorescent PS in

Chinese hamster fibroblasts. Cultures were incubated for 30 min at 2 "C with vesicles (100 pM total lipid) containing 20 mol % (palmitoyl-C6-NBD)-PS. The cells were washed and warmed to 7 'C for 0 min ( A ) or 30 min ( B ) prior to photomicroscopy. Bar represents 10 pm.

brane and in internal structures (Fig. 1B). The pattern of intracellular fluorescence was virtually identical to that previously observed using a fluorescent analog of phosphatidylethanolamine (Sleight and Pagano, 1985) in which the nuclear envelope and mitochondria were prominently labeled. A similar pattern of intracellular fluorescence also developed if the washed cells were held at 2 "C for long periods of time (>2-4 h; not shown).

Self-quenching at high concentrations is a common property of fluorescent probes (Weinstein et al., 1977; Nichols and Pagano, 1981) and (C6-NBD)-PS also exhibits this property (Tanaka and Schroit, 1983). To determine if a large amount of self-quenching occurred in the cells containing fluorescent PS, the following experiment was performed. Cells in suspension were incubated at 2 "C with vesicles containing fluorescent PS, washed, warmed to 7 "C for various times, washed, and the amount of fluorescence in the cells determined in the absence or presence of 1% Triton X-100. The results of this experiment are shown in Fig. 2. The presence of Triton disrupts membranes and puts the lipid in a non-self-quenching environment (Pagano et al., 1981~). The ratio of fluorescence in the absence or presence of Triton can then be calculated, and, if quenching occurs, this ratio should change as the (palmitoyl-C6-NBD)-PS moves into or out of a quenched compartment. As seen in Fig. 2 (dashed line), this ratio increased slightly with time at 7 "C, suggesting that the fluorescent PS was partially quenched after its insertion into the plasma membrane at 2 "C, and was dequenched as inter-

] 0.30 - a - 8 s!

3 U

0.20

0.10

0.w

T

I I I I I I 0 20 40 60

time at PC (min)

2.5

2.0

1.5

1 .o

0.5

0.0

FIG. 2. Relief of (palmitoyl-Cs-NBD)-PS quenching during internalization. Cell suspensions were incubated with lipid vesicles containing (palmitoyl-C6-NBD)-PS for 30 min at 2 "C, washed, incubated at 7 "C for various times, washed with HCMF, and the fluorescence determined in the absence (closed circles) and presence (open circles) of 1% Triton X-100. The ratio of fluorescence in the absence or presence of Triton X-100 (triangles) is also shown. Data points represent the mean f S.D. of five determinations.


nalization proceeded. The basis for quenching at the plasma membrane is unknown but might be due to the presence of microscopic domains (Edidin, 1984) containing self-quenching concentrations of fluorescent PS or to interaction with a membrane protein(s).

Tests for DeacylatwnlReacylation and Endocytosis-To rule out the possibility that (palmitoyl-C6-NBD)-PS entered cells as a result of deacylation/reacylation reactions, cells were grown in the presence of [l-14C]palmitic acid and then incubated with the fluorescent lipid (see "Experimental Proce- dures''). The cellular lipids were extracted and subjected to two-dimensional TLC and autoradiography. The (palmitoyl- C6-NBD)-PS, which was completely separated from endogenous cellular lipids by this method, was not isotopically labeled.

To determine whether internalization of the fluorescent lipid at 7 "C was due to endocytosis, cells were incubated at 2 "C with (palmitoyl-C6-NBD)-Ps followed by SRh-SAD prior to warming to 7 "C for 30 min. As shown in Fig. 3, the fluorescent stearoylated dextran uniformly labeled the plasma membrane of treated cells, while the (palmitoyl-C6-NBD)-PS was internalized and labeled intracellular membranes. In control experiments in which cells labeled with SRh-SAD were deliberately warmed to 37 "C to allow endocytosis to proceed, the SRh-SAD accumulated in punctate intracellular vesicles scattered throughout the cytoplasm (not shown).

Effect of Inhibitors, Structural Analogs, Proteases, and Pro- tein Modification Reagents on Internalization of (Palmitoyl- C6-NBD)-PS"Ce1ls were treated with various reagents to test for their possible effects on the internalization of (palmitoyl-C6-NBD)-PS (Table I). Of the various treatments examined, several inhibited internalization of the fluorescent lipid as seen in control cells (Fig. 4A). (i) When cells were incubated in glucose-free HMEM containing 5 mM sodium azide and 50 mM 2-deoxyglucose prior to incubation with (palmitoyl-C6-NBD)-Ps, cellular ATP levels were reduced to 17% of control values, and internalization of the fluorescent lipid was arrested (Fig. 4B). Fluorescence was restricted to and remained at the plasma membrane, even if the cells were incubated in the presence of the inhibitors an additional 30 min at room temperature or at 37 "C (not shown). However, if these depleted cells were washed and incubated with HMEM lacking the inhibitors but containing glucose for either 30 min at 7 "C or 15 min at room temperature, internalization of fluorescence was seen (not shown). Under these incubation conditions, no internalization of the endocytic marker, SRh-SAD, was detected. (ii) In cells fixed with glutaraldehyde prior to incubation with (palmitoyl-C6-NBD)-Ps, fluorescence was present only at the plasma membrane (as in Fig. 4B). (iii) Pretreatment of cells with 0.5 mM NEM for 30 min at 2 "C prior to incubation with (palmitoyl-C6-NBD)-Ps inhibited internalization of the fluorescent lipid (as in Fig. 4B). When liposomes containing (palmitoyl-C6-NBD)-PS were incubated with 0.5 mM NEM under these same conditions and extracted, no additional fluorescent products were observed by TLC (not shown), suggesting that the effect of NEM on lipid internalization was not due to its covalent reaction with the fluorescent lipid. If cells were treated with (palmitoyl-C,-NBD)-PS for 30 min at 2 'C prior to treatment with NEM, internalization of fluorescent PS was not blocked (not shown), suggesting that the presence of fluorescent PS at the plasma membrane interfered with the reaction of NEM with a sulfhydryl group critical to the internalization process. (iv) When cells were incubated with (palmitoyl-C6-NBD)-PS and glycerophosphorylserine or glycerophosphorylethanolamine at 2 "C, washed, and warmed to 7 "C in the presence of

FIG. 3. Internalization of fluorescent PS is not due to endocytosis. Cells were incubated for 30 min at 2 "C with (palmitoyl- Cs-NBD)-PS, washed, and further incubated for 10 min at 2 "C with a fluorescent dextran (SRh-SAD) which binds to the cell surface. The cells were then washed, warmed to 7 'C for 30 min, washed, and photographed. A , NBD fluorescence; B, SRh-fluorescence.

these structural analogs, internalization of the fluorescent lipid was blocked (as in Fig. 4B). This inhibition was not seen when the structural analogs were omitted during warming to 7 "C, or when glycerophosphorylcholine, L-serine, phosphoserine, phosphorylethanolamine, or phosphorylcholine were used (Table I).

Removal of (Palmitoyl-C6-NBD)-PS from Cells by Back- exchange to Nonflwrescent Vesicles-The amount of fluorescent PS which could be removed from cells by incubation with nonfluorescent vesicles depended on the ATP content of the cells and whether the cells had been warmed to 7 "C prior to back-exchange (Table 11). After the initial incubation with fluorescent PS at 2 "C, most of the fluorescent lipid could be removed by back-exchange. The amount of (palmitoyl-C6- NBD)-PS which could be removed by back-exchange was not increased by performing all incubations in calcium- and magnesium-free medium or by further increasing the number of


TABLE I Inhibition of (palmitoyl-C6-NBD)-PS internalization

Cell cultures on glass cover slips were washed 3 times with HMEM and incubated with each reagent at the indicated concentration, temperature, and time. The cells were then washed, chilled for 15 min on an ice-water bath, and incubated with liposomes containing (palmitoyl-C6-NBD)-PS for 30 min at 2 "C. The cells were then washed, briefly warmed, and examined for intracellular fluorescence. Where indicated, the reagents were also present throughout the experiment (i.e. during incubation with liposomes, during warming, and during the microscopic observation). Parallel control cultures showed intracellular fluorescence similar to that in Fig. 1B. The abbreviations used are: GPS, glycerophosphorylserine; GPE, glycerophosphorylethanolamine; GPC, glycerophosphorylcholine; TNBS, trinitrobenzenesulfonic acid; DIDS, 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid; SITS, 4-acetamido-4'-isothiocyanostilbene-2.2'-disu~fonic acid.

Treatment

- Pretreatment conditions Present

Conc throughout Internalization inhibited Temperature Time exD.

Inhibitors & ionophores

2-Deoxyglucose" Sodium azide + Cycloheximide Ouabain Nigericin Valinomycin

Structural analogs GPS GPE GPC L-Serine Phosphoserine Phosphorylethanolamine Phosphorylcholine Choline chloride

NEM' Glutaraldehyde' Iodoacetate Iodoacetamide TNBS DIDS SITS

Proteases Neuraminidase Trypsin Trypsin Pronase Proteinase K

Protein modification reagents

5 mM 50 mM 10 d m 1

5 ag/ml 0.5 mM

5 ng/ml

10 mM 10 mM 10 mM 40 mM 40 mM 40 mM 40 mM 40 mM

0.5 mM 0.5%

1&10 mM 1&10 mM 1, 10,100 mM

1 mM 1 mM

2.5 pg/ml 25 pg/ml

250 pg/ml 25 d m 1

"C

37 37 37 37 37

2 2 2 2 2 2 2 2

2 22 2 2 2 2 2

2 2

37 2

min

30 24 h 30 30 30

30 30 30 30 30 30 30 30

30 30 30 30 30 30 30

30 30 12 30

25 &ml 2 30 - -

a ATP-depletion was performed by pretreating cells with 5 mM NaN3 and 50 mM 2-deoxyglucose in glucose-free HMEM for 30 min at 37 "C and carrying out all subsequent incubations and washes in glucose-free HMEM containing these inhibitors.

NEM treatment was followed by a 5-min treatment on ice with 0.5 mM dithiothreitol in HMEM and washing with HMEM Drior to incubation with vesicles containing (palmitoyl-C6-NBD)-PS. Identical results were obtained at pH 7.2 an68.5.

~~

Fixative contained 0.5% glutaraldehyde, 5% sucrose (w/v), and 0.1 M PIPES, pH 7.0.

back-exchanges. Much less fluorescent PS could be removed if the cells were warmed to 7 "C to allow internalization to proceed prior to back-exchange. However, when internalization of the fluorescent lipid was blocked by ATP-depletion, almost all of the NBD lipid was available for back-exchange in both the 2 "C and 7 "C experiments. Similar results were obtained in cells pretreated with NEM (-80% removal of the fluorescent lipid by back-exchange in both the 2 "C and 7 "C experiments). These findings are consistent with our observations that internalization of the fluorescent lipid was slow at 2 "C (Fig. lA) and blocked by ATP-depletion (Fig. 4B) or NEM treatment and suggest that back-exchange to nonfluorescent vesicles preferentially removes (palmitoyl-C6-NBD)- PS from the plasma membrane.

Incubation of the L- and D-Isomers of (Palmitoyl-Cfi-NBD)- PS with CeZ&"ells were incubated for 30 min at 2 "C with liposomes containing L- or o-(palmitoyl-C,-NBD)-PS, washed, and warmed for 30 min at 7 "C. Prominent intracel-

lular labeling was seen with the L-isomer (Fig. predominantly surface labeling occurred when the was used (Fig. 5B).

The amount of (palmitoyl-Cfi-NBD)-PS which

5A), but D-isomer

could be removed from the plasma membrane by back-exchange depended on the ATP-content of the cells, the isomer used, and whether the cells had been warmed to 7 "C prior to back- exchange (Table 111). When ATP-depleted cells were incubated with liposomes containing fluorescent PS for 30 min at 2 "C, washed, and warmed to 7 "C for 0 or 30 min, almost all of either the L- or D-isomer of the fluorescent lipid could be removed. In nondepleted cells, less of the L-isomer relative to the D-isomer was available for back-exchange after warming to 7 "C for either 0 or 30 min, suggesting that only a small amount of internalization of the D-iSOmer had occurred. These quantitative results are consistent with our observations that the L- and D-isomers of fluorescent Ps were internalized to much different extents (Fig. 5, A and B).

Transbilayer Movement of Fluorescent Amimphospholipids 5895

I

FIG. 4. Inhibition of internalization of fluorescent PS. Cells were washed and incubated in HMEM at 2 "C (A), or pretreated with 5 mM NaN3 + 50 mM 2-deoxyglucose in glucose-free HMEM to lower cellular ATP levels (see Table I) (B) . The cells were then incubated with vesicles containing (palmitoyl-C6-NBD)-PS for 30 min at 2 "C, washed, warmed for 30 min at 7 "C, and photographed.

TABLE I1 Removal of fluorescent PS from the plusmu membrane by

back-excharwe

Time at 7 'C

% NBD-PS removed by back-exchange"

Nondepleted ATP-depletedb min

0 68.6 f 8.1 88.2 * 2.1 16.2 & 12.6 87.9 f 3.9 30

a Cells were incubated for 30 min at 2 "C with vesicles containing (palmitoyl-C6-NBD)-Ps, washed, and warmed to 7 "C for 0 or 30 min. The cells were then incubated (three times, 10 min) with either HMEM alone (control cells) or with HMEM containing 250 pM DOPC vesicles (back-exchanged cells) to remove (palmitoyl-C6- NBD)-PS at the cell surface. The cell-associated lipids were extracted and the NBD-fluorescence quantified and ,normalized to cellular DNA. % NBD-PS removed by back-exchange was calculated by

[l - (RFU back-exchanged cells/pg of DNA)/

(RFU control cells/pg of DNA)] X 100

Values represent the mean f S.D. of three determinations. RFU, relative fluorescence unit.

* ATP depletion was performed as described in Table I.

Internalization of (Palmitoyl-C6-NBD)-PE-When cells were incubated at 2 "C with vesicles containing fluorescent PE, washed, and subsequently warmed to 7 "C, internalization of the (palmitoyl-C6-NBD)-PE to the nuclear envelope and

mitochondria occurred, as previously described (Sleight and Pagano, 1985). This internalization was blocked by pretreatment of cells with NEM as described above for (palmitoyl- Cs-NBD)-PS (not shown). The internalization of fluorescent PE was also blocked when cellular ATP levels were depleted to 17% of control values using the protocol described in Table I (not shown). However, as previously reported (Sleight and Pagano, 1985), when cells were incubated with (palmitoyl-C6- NBD)-PE in the presence of the metabolic inhibitors, but without preincubation at 37 "C in glucose-free medium containing the inhibitors, the internalization of (palmitoyl-C6- NBD)-PE did not decrease noticeably. Under these conditions, cellular ATP levels were 81% of control values.

Incubation of the L- and 0-Isomers of (Palmitoyl-C6-NBD)- PE with Cells-Internalization of the L- and D-isomers of (palmitoyl-C6-NBD)-PE by cells was also examined. Cells were incubated at 2 "C with vesicles containing fluorescent PE, washed, and warmed to 7 "C (see "Experimental Proce- dures"). Prominent intracellular labeling was seen when the L-isomer was used (Fig. 5C), but little or no intracellular labeling was seen when the D-isomer was used (Fig. 50).

Microinjection of Cells with Fluorescent Phosphatidylser- ine-When cells were microinjected with the D-isomer of (palrnitoyl-C,-NBD)-PS and immediately observed in the fluorescence microscope, intense fluorescent labeling of the mitochondria, nuclear envelope, and possibly other organelles was seen (Fig. 6). This pattern of fluorescence was the same when the L-isomer of this lipid was used (not shown) and was also seen in cells which had been subjected to ATP depletion (see Table I) followed by microinjection with (palmitoyl-C6- NBD)-PS (not shown).

DISCUSSION

The asymmetric distribution of the major phospholipid classes in the two leaflets of the erythrocyte membrane is well established (for reviews, see Op den Kamp, 1979; Etemadi, 1980; Houslay and Stanley, 1982; Storch and Kleinfeld, 1985). In this system the choline-containing lipids, sphingomyelin and phosphatidylcholine, are enriched on the extracyto- plasmic face of the membrane while the amino-containing lipids, phosphatidylserine and phosphatidylethanolamine, re- side predominantly on the cytoplasmic half of the membrane. Recently, several laboratories have shown that exogenously supplied phosphatidylserine and phosphatidylethanolamine (or their spin-labeled analogs) can be incorporated into the erythrocyte membrane. These lipids are rapidly translocated to the inner leaflet of the membrane bilayer by a process which is ATP-dependent and protein-mediated (Seigneuret and Devaux, 1984; Daleke and Heustis, 1985; Tilley et al., 1986; Zachowski et al., 1985; Zachowski et al., 1986). However, when similar experiments are performed using exogenous phosphatidylcholine or sphingomyelin, no rapid transbilayer movement is seen.

In this paper we have studied the behavior of fluorescent analogs of PS and PE inserted into the plasma membranes of cultured fibroblasts and show that they behave in a manner similar to that described for exogenously supplied PS and PE given to erythrocytes.

Internalization of Fluorescent PS and PE from the Plasma Membrane at 7°C Is Due to Transbilayer Movement-We explored various mechanisms which might have been responsible for internalization of the fluorescent aminophospholipids at low temperature. Two pieces of evidence suggested that the internalization of fluorescent PS at 7 "C was not due to endocytosis. (i) Under conditions which permitted internalization of the fluorescent lipid, no internalization

5896 Transbihyer Movement of Fluorescent Aminophospholipids

FIG. 5. Internalization of fluorescent PS and PE is stereospecific. Cells were incubated for 30 min at 2 "C with vesicles containing 20 mol % of either the L- (A) or D-isomer of (palmitoyl-C~-NBD)-Ps (B). The cells were then washed, warmed for 30 min at 7 'C, washed, and photographed. Alternatively, cells were incubated for 30 min at 2 "C with vesicles containing 50 mol % of either the L- ( C ) or D-isomer of (palmitoyl-C6-NBD)-PE (D). The cells were then washed, warmed for 2 h at 7 "C, washed, and photographed. Bar represents 10 pm.

TABLE I11 Removal of stereoisomers of fluorescent PS from the plasma

membrane by back-exchange % NBD-PS removed by

Stereoisomer T:!yt back-exchange" Nondepleted ATP-depletedb

min L 0 64.7 k 14 89.9 f 2.1 D 0 86.5 f 3.5 92.5 f 4.4 L 30 21.7 f 11.4 85.1 f 3.0 D 30 72.2 f 1.0 86.2 f 4.1

All incubations were carried out as described in Table 11, except that vesicles containing either the L- or D-isomer of (palmitoyl-Cp- NBD)-PS were used. Data were calculated as described in Table 11.

ATP-depletion was performed as described in Table I.

of a fluorescent endocytic marker, SRh-SAD (Lippincott- Schwartz and Fambrough, 1986), was detected (Fig. 3). (ii) If the fluorescent lipids were internalized as a result of endocytosis, any fluorescent lipid inserted into the plasma membrane lipid bilayer should have been internalized. Although both the L- and D-isomers of (palmitoyl-CG-NBD)-PS and -PE could be inserted into the plasma membrane of cells, the D-isomers of these fluorescent lipids were not internalized under conditions which permitted pronounced internalization of the L- isomers (Fig. 5).

Internalization of fluorescent PS was not dependent on the presence of calcium or magnesium in the medium or to lipid

remodeling via deacylation/reacylation reactions. Internali- zation of the fluorescent PS was not the result of membrane fusion, since a non-exchangeable vesicle marker did not be- come cell-associated. The internalization of fluorescent PS was not affected when cells were incubated in the presence of valinomycin (Table I) or in medium containing a high level of potassium,2 suggesting that membrane potential did not play a primary role in this process.

Our results indicated that the first step in the internalization of (palmitoyl-C6-NBD)-PS from the plasma membrane involved its transbilayer movement, followed by diffusion of the fluorescent lipid through the cytosol to label intracellular membranes. This conclusion was supported by the following. First, when cells were microinjected with fluorescent PS, pronounced intracellular labeling occurred immediately after microinjection (Fig. 6). This pattern of intracellular fluorescence was virtually identical to that seen with exogenously- supplied fluorescent PS after warming to 7 "C. However, unlike the exogenous PS, the distribution of the microinjected lipid was unaffected by ATP depletion. Second, in microinjection studies of L- and D-(palmitoyl-C6-NBD)-PS, no differ- ences were seen between their intracellular distributions. These results indicated that once inside the cell, both isomers of fluorescent PS rapidly transferred to intracellular membranes, probably by a monomeric diffusion process (Nichols and Pagano, 1981) as seen with (C6-NBD)-PS in liposomes

* 0. C. Martin and R. E..Pagano, unpublished observations.


I FIG. 6. Microinjection of fluorescent PS. Fluorescence (A) and

phase-contrast photomicrographs ( B ) of cells obtained immediately after microinjection of the D-isomer of (palmitoyl-C6-NBD)-PS. Sim- ilar results were obtained using the L-isomer of this lipid and with ATP-depleted cells. Bar represents 10 pm.

(Tanaka and Schroit, 1986). We conclude that the transbilayer movement of (palmitoyl-C6-NBD)-Ps, and not its transfer to intracellular membranes, was stereospecific.

Evidence for the transbilayer movement of fluorescent PE at the plasma membrane at 7 "C has been summarized elsewhere (Sleight and Pagano, 1985).

Transbilayer Movement of Fluorescent PS and PE is En- ergy-dependent and Protein-mediated-When cellular ATP levels were depleted to 17% of control values, transbilayer movement of the fluorescent analogs of PS and PE was inhibited. Although we previously reported that internalization of fluorescent PE at 7 "C was energy-independent (Sleight and Pagano, 1985), we found in the present study that the previously used incubation conditions decreased ATP levels to only 81% of control values. Although this reduction was sufficient to inhibit endocytosis (Sleight and Pagano, 1985), it was apparently inadequate to inhibit transmembrane movement.

Four pieces of evidence indicate that the transbilayer movement of fluorescent PE and PS was protein-mediated. First, no transmembrane movement was detected in lipid vesicles containing (palmitoyl-CG-NBD)-PE (Pagano and Sleight, 1985) or -PS (data not shown), suggesting that transmembrane movement was not due to a passive, physical property of these lipids, in contrast to fluorescent analogs of diaoyl- glycerol (Pagano and Longmuir, 1985) and ceramide (Pagano and Sleight, 1985). Second, pretreatment of cells with NEM, a sulfhydryl-derivatizing reagent, inhibited the transbilayer movement of these lipids. Third, the presence of the structural analogs, glycerophosphorylserine or glycerophosphorylethanolamine, inhibited transbilayer movement as might be ex- pected for a facilitated process (Bishop and Bell, 1985).

Fourth, the internalization of both fluorescent PE and PS was stereospecific. We are presently exploring strategies for specifically labeling the protein(s) responsible for (fluorescent) aminophospholipid translocation at the plasma membrane.

Although the study of lipid asymmetry in the plasma membrane of cells containing subcellular organelles is considerably more difficult than in the red blood cell, the existing data (reviewed in Etemadi, 1980; Houslay and Stanley, 1982) indicate that the endogenous membrane lipids of nucleated cells may be asymmetrically distributed across the plasma membrane bilayer in a manner similar to that seen in the erythrocyte. Elsewhere we have shown that NBD analogs of phosphatidylcholine and sphingomyelin can be inserted into the plasma membrane bilayer of cultured fibroblasts at 2 "C and that these lipids do not undergo transmembrane movement (Sleight and Pagano, 1984; Lipsky and Pagano, 1984, Pagano and Sleight, 1985). These findings and the present study with the NBD analogs of the aminophospholipids, PE and PS, lead us to speculate that the exogenously supplied NBD lipids may assume a transbilayer distribution similar to that of the endogenous membrane lipids, and that maintenance of lipid asymmetry is a fundamental property of cells. In this regard, the finding that (C,-NBD)-PS inserted into the erythrocyte membrane can serve as a signal for triggering their in vivo recognition and clearance from the circulation is particularly exciting (Tanaka and Schroit, 1983; Schroit et al., 1984, Schroit et al., 1985).

AcknowZedgments-We thank Dr. W. Bigley, Dr. H. Seliger, and Mr. A. Ting for assistance in measuring cellular ATP levels, M. Koval for synthesizing SRh-SAD, and acknowledge Dr. D. Hoekstra for preliminary observations made in our laboratory on the behavior of (Cs-NBD)-PS in cells.

REFERENCES

Backer, J. M., and Dawidowicz, E. A. (1985) J. CeU Bwl. 101, 261

Bishop, W. R., and Bell. R. M. (1985) Cell 42,5140 Bligh, E. G., and Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37,

Brockerhoff, H., and Yurkowski, M. (1965) Can. J. Biochm. 43,

Chen, F. R. (1967) J. Bwl. Chem. 242,173-181 Comfurius, P., and Zwaal, R. F. A. (1977) Biochim. Bwphys. Acta

Daleke, D. L., and Huestis, W. H. (1985) Biochemistry 24,5406-5416 Edidin, M. (1984) Comments Mol, Cell. Bwphys. 2,285-293 Etemadi, A-H. (1980) Biochim. Bwphys. Acta 604,423-475 Ford, D. K., and Yerganian, G. (1958) J. Natl. Cancer Znst. 21,393-

Gilfillan, A. M., Chu, A. J., Smart, D. A., and Rooney, S. A. (1983) J.

Houslay, M. D., and Stanley, K. K. (1982) in Dynamics of Bwbgical

Kates, M. (1972) in Techniques of Lipidobgy (Work, T. S., and Work,

Kremer, J. M. H., v. d. Esker, M. W. J., Pathmamanoharan, C., and

Leyva, A., Jr., and Kelly, W. N. (1974) Anal. Biochem. 62,173-179 Lippincott-Schwartz, J. and Fambrough, D. M. (1986) J. Cell BWL

Lipsky, N. G., and Pagano, R. E. (1984) Ann. N. Y. Acud. Sci. 435,

Longmuir, K. J., Martin, 0. C., and Pagano, R. E. (1985) Chem. Phys.

Martin, 0. C., and Pagano, R. E. (1986) A d Biochem. 159, 101-

Nichols, J. W., and Pagano, R. E. (1981) Biochemistry 20,2783-2789 Op den Kamp, J. A. F. (1979) Annu. Rev. Biochem. 48,47-71 Pagano, R. E. (1983) in Liposome Letters (Bangham, A. D., d) pp.

Pagano, R. E., and Longmuir, K. J. (1985) J. Bwl. Chem. 260,1909-

(abstr.)

911-917

1777

488,36-42

425

Lipid Res. 24,1651-1656

Membranes, pp. 196-199, J. Wiley and Sons, New York

E., eds) p. 569, Elsevier Scientific Publishing Co., Amsterdam

Wiersema, P. H. (1977) Biochemistry 16,3932-3935

102,1593-1605

306-308

Lipids 36,197-207

108

83-96, Academic Press, New York


1916 Pagano, R. E., Longmuir, K., Martin, 0. C., and Struck, D. K. (1981a)

J. Cell Biol. 91,872-877 Pagano, R. E., Martin, 0. C., Schroit, A. J., and Struck, D. K. (1981b)

Biochemistry 20,4920-4927 Pagano, R. E., Schroit, A. J., and Struck, D. K. (1981~) in Liposomes:

From Physical Structure to Therapeutic Applications (Knight, C. G., ed) pp. 323-348, Elsevier-Scientific Publishing Co., Amsterdam

Pagano, R. E., Longmuir, K. J., and Martin, 0. C. (1983) J. Biol. Chem. 258,2034-2040

Pagano, R. E., and Sleight, R. G. (1985) Science 229,1051-1057 Rouser, G., Siakotos, A. N., and Fleisher, S. (1966) Lipids 1,8546 Schroit, A. J., Madsen, J. W., and Tanaka, Y. (1985) J. Biol. Chem.

Schroit, A. J., Tanaka, Y., Madsen, J., and Fidler, I. J. (1984) Biol.

Seigneuret, M., and Devaux, P. F. (1984) Proc. Natl. Acad. Sci. U. S.

Sleight, R. G., and Pagano, R. E. (1984) J. Cell Biol. 99, 742-751 Sleight, R. G., and Pagano, R. E. (1985) J. Biol. Chem. 260, 1146-

260,5131-5138

Cell 51,227-238

A. 81,3751-3755

1154

Storch, J., and Kleinfeld, A. M. (1985) Trends Biochem. Sci. 10,418-

Struck, D. K., and Pagano, R. E. (1980) J. Biol. Chem. 255, 5404-

Struck, D. K., Hoekstra, D., and Pagano, R. E. (1981) Biochemistry

Tanaka, Y., and Schroit, A. J. (1983) J. Biol. Chem. 258, 11335-

Tanaka, Y., and Schroit, A. J. (1986) Biochemistry 25,2141-2148 Tilley, L., Cribier, S., Roelofsen, B., Op den Kamp, J. A. F., and van

Weigel, P. H., and Englund, P. T. (1975) J. Biol. Chem. 250, 8536-

Weinstein, J. N., Yoshikami, S., Henkart, P., Blumenthal, R., and

Wolf, D. E., Schlessinger, J., Elson, E. L., Webb, W. W., Blumenthal,

Zachowski, A., Favre, E., Cribier, S., Herve, P., and Devaux, P. F.

Zachowski, A., Fellman, P., and Devaux, P. F. (1985) Biochim. Bio-

421

5410

20,4093-4099

11343

Deenen, L. L. M. (1986) FEBS Lett. 194,21-27

8542

Hagins, W. A. (1977) Science 195,489-492

R., and Henkart, P. (1977) Biochemistry 16,3476-3483

(1986) Biochemistry 26,2585-2590

phys. Acta 815, 510-514

journal of biological vol. no. 12, 1987 in transbilayer ... · the journal of biological chemistry...

Documents