doi:10.1006/scdb.2000.0231, available online at http://www.idealibrary.com onseminars in CELL & DEVELOPMENTAL BIOLOGY , Vol. 12, 2001: pp. 139–148

Lipid distribution and transport across cellular membranes

Thomas Pomorski a, Sigrún Hrafnsdóttir a, Philippe F. Devaux b and Gerrit vanMeer a,∗

In eukaryotic cells, the membranes of different intracellularorganelles have different lipid composition, and variousbiomembranes show an asymmetric distribution of lipid typesacross the membrane bilayer. Membrane lipid organizationreflects a dynamic equilibrium of lipids moving across thebilayer in both directions. In this review, we summarizedata supporting the role of specific membrane proteins incatalyzing transbilayer lipid movement, thereby controllingand regulating the distribution of lipids over the leaflets ofbiomembranes.

Key words: flip–flop / lipid asymmetry / translocase /flippases / ABC transporter

c© 2001 Academic Press

Introduction

The basic structure of biological membranes isthe lipid bilayer. This structure is mainly formedfrom three different classes of lipids (glycerolipids,sterols and sphingolipids). In eukaryotic cells, themembranes of different intracellular organelles havedifferent lipid compositions. For example, plasmamembranes are typically enriched in sphingolipids,phosphatidylserine (PS) and cholesterol, while theendoplasmic reticulum (ER) is depleted in theselipids. In addition to the heterogeneous distributionof lipids between membranes, there are also strikingdifferences in the distribution of lipids across themembrane bilayer. For instance, while the ER is

From the aDepartment of Cell Biology and Histology, Academic MedicalCenter, University of Amsterdam, P.O. Box 22700, 1100 DE Amsterdam,The Netherlands and bInstitut de Biologie Physico-Chimique 13, rue Pierreet Marie Curie, 75005 Paris, France. *Corresponding author.

c©2001 Academic Press1084–9521/01/020139+ 10/$35.00/0

assumed to have a symmetric lipid distribution, lipidsin the plasma membrane of eukaryotic cells showa clear asymmetric arrangement, with the majorityof the glycosphingolipids and phosphatidylcholine(PC) in the exoplasmic leaflet and the aminophos-pholipids, PS and phosphatidylethanolamine (PE),on the cytoplasmic face (reviewed in Reference 1).

The lipid distribution across membranes resultsfrom a continuous inward and outward movementof lipids between the two monolayers (Figure 1).Although neutral lipids like cholesterol and chargedlipids in a protonated form, such as free fatty acid,phosphatidic acid or phosphatidylglycerol, canmove fast between leaflets, spontaneous transbilayermovement of most lipids with a polar (or charged)lipid headgroup is a very slow process (t1/2=hoursto days in model membranes; see Reference 1).However, the assembly of certain cellular membranesrelies on a rapid transbilayer movement of polarlipids. Examples are the ER of eukaryotes or in thecytoplasmic membrane of prokaryotes, where lipidbiosynthesis occurs predominately in the cytoplasmicleaflet. Furthermore, the compositional asymmetryof the plasma membrane does not correspond to theasymmetry of lipid synthesis or hydrolysis. Hence,lipid asymmetry must be formed and afterwardsmaintained by specific mechanisms that control lipidmovement across the bilayer and counterbalancethe lipid randomization caused by spontaneoustransbilayer movement.

Techniques to assess transbilayer lipidmovement

The asymmetric arrangement of phospholipids inplasma membranes was originally established for nat-ural lipids in erythrocytes using chemical labeling,2

hydrolysis by phospholipases,3 and exchange by lipidtransfer proteins,4,5 and by the same techniques

139

T. Pomorski et al.

Figure 1. Lipids can cross biological membranes by vari-ous mechanisms. Spontaneous diffusion refers to the non-specific lipid movement occurring between the membraneleaflets; its rate is determined by the biophysical proper-ties of both the lipid and the membrane (note that trans-fer of a lipid from one monolayer to the other is not nec-essarily coupled with movement of a lipid towards the op-posite direction). Flippases facilitate an ATP-independent,bi-directional movement of lipids but are unable to accumu-late a given lipid in one leaflet. Translocases directly use theenergy released by ATP-hydrolysis to drive unidirectionallipid movement against a gradient in the membrane.

in plasma membrane-derived viruses.6–8 Whereasinitially lipid translocation across biomembraneswas also measured by these techniques, evidence forthe involvement of proteins is based on techniquesusing lipid analogs, where in general one fatty acidchain has been replaced by a short chain carryinga radiolabel, a spin-label or a fluorescent moiety9

(Figure 2). Clearly, in any case where a protein hasbeen identified as being involved in translocatinglipid analogs across a biomembrane, the activityof the protein towards naturally occurring lipidswill have to be established. Also the kinetics oftransbilayer movement of natural lipids can only bedetermined by studying the lipids themselves, butdue to technical complications this has been achievedin only a few cases. It is, therefore, a challenge tothe field to develop methods to establish whetherthe putative lipid translocators use natural lipids assubstrates, and by which the transbilayer movementof natural lipids can be measured in living systems.The use of fluorescent annexin V is a new methodthat has proved extremely sensitive for the detection

of small amounts of endogeneous PS on the surfaceof apoptotic and aged cells.10

Lipid movement in the ER and bacterialmembranes: evidence for flippases

The ER is the principal site of membrane assemblyin eukaryotes and in that aspect similar to the cy-toplasmic membrane in prokaryotes. Phospholipidbiosynthesis in these membranes is an asymmetricprocess, resulting in the insertion of a newly syn-thesized lipid in the cytoplasmic leaflet. In order tocreate a bilayer, a fraction of the phospholipids hasto be translocated to the other leaflet (Figure 3).Indeed, rapid phospholipid movement has beenreported in microsomal membranes (t1/2= secondsto minutes).11–17 Although the measured rate oftransport varied between assays, probably due todifferences in kinetic resolution, all observations todate indicate that phospholipid movement in the ERis bi-directional, ATP-independent and non-specifictowards the phospholipid headgroup. Protein-modifying reagents partially inhibited phospholipidmovement suggesting that proteins are directly orindirectly involved.13,15–17 This notion was furthersubstantiated when membrane proteins from rat livermicrosomes were reconstituted into proteoliposomesand found to facilitate lipid translocation, whileprotein-free liposomes or proteoliposomes contain-ing proteins from human erythrocyte membranewere inactive.14 Recently, the transbilayer movementof short-chain (diC4) water-soluble PC was assayedin proteoliposomes reconstituted from Triton X-100soluble fractions of rat liver microsomes.17 Thetransport activity was recovered in liposomes contain-ing a protein fraction of a low sedimentation rateand was sensitive to proteolysis, whereas similarlyprepared liposomes, containing solely the ER lipids,were inactive. Fractionation of the detergent extractresulted in proteoliposomes with different specific ac-tivity, indicating that specific microsomal membraneproteins were responsible for the transport.

Besides the phospholipids, the glycolipidsmannosyl-phosphodolichol, glucosyl-phosphodolicholand an oligosaccharide-diphosphodolichol have toundergo transbilayer movement in the ER for thesynthesis of the full length lipid-linked oligosaccha-ride that is subsequently transferred to a lumenaldomain of proteins. Rapid flip–flop of water-solubleanalogs of these glycolipids has been reported.18

140

Transbilayer lipid distribution and movement

Figure 2. Schematic representation of the techniques used for monitoring transbilayer movement of lipid. Most techniquesare based on the use of lipid analogs with either one or two short fatty acid chains. Dibutyroylphosphatidylcholine (diC4PC)is rather water-soluble and will be released into the lumen when it redistributes from the outer to the inner leaflet ofthe vesicle. Thus, rapid separation of the vesicles from the incubation medium by filtration allows determination of lipidtranslocation by measuring the amount of radioactivity associated with the vesicles. Acyl-chain-labeled lipid analogs containa radiolabel, a spin-label or fluorescent group on a short fatty acyl chain at the C2 position of the glycerolipid or onthe amine of the sphingosine. These lipids can be incorporated into synthetic donor vesicles. Upon addition to cells, thelipid analogs (but not the normal long-chain lipids) transfer spontaneously and very rapidly into the outer leaflet of thebiological membrane. Their transbilayer distribution can be monitored as a function of time by either back extraction of theprobes still present in the exposed leaflet onto albumin, or by chemical modification of the non-translocated analogs withmembrane impermeant reagents such as ascorbate (for spin-label) or dithionite (for the fluorescent NBD group). Whenusing fluorescent analogs, vesicles can be prepared containing the non-exchangeable, headgroup labeled N-rhodaminePE and the NBD lipid of interest. During exchange and translocation of the NBD-lipid, a decrease of the energy transferbetween the NBD lipid and the N-Rh-PE occurs which results in an increase in the NBD fluorescence and allows continuousmeasurement of lipid translocation. Modification of lipids present in the outer leaflet by chemical reagents or action ofphospholipases can be applied to endogenous, long-chain lipids but has a limited temporal resolution.

Similarly, the synthesis of the glycosylphosphatidyli-nositol anchor of GPI-proteins is initiated on thecytosolic surface of the ER and, at some stage of itsbiosynthesis, must translocate to the ER lumen foraddition to proteins.

As in the ER, fast transbilayer movement of phos-pholipids is needed to propagate the bilayer of thebacterial cytoplasmic membrane, at a rate sufficientfor a rapidly growing bacterium. Phospholipidflip–flop in bacterial cytoplasmic membranes hasaccordingly been found to be fast.19–23 Interestingly,

phospholipid flip–flop in bacteria has similar char-acteristics as in the ER and is bi-directional, energyindependent and non-specific towards the phospho-lipid headgroup. In support for the involvementof membrane proteins in the accelerated flip–flop,phospholipid movement in Baccillus membrane vesi-cles was found to be protease-sensitive.22 However,protein modification of E. coli inner membraneproteins had no detectable effect on phospholipidtransport,21,23 probably because the assay used wouldnot have allowed detection of partial inhibition.

141

T. Pomorski et al.

Figure 3. Various proteins control lipid sidedness across cellular membranes. Synthesis of phosphatidylcholine (PC),phosphatidylserine (PS) and phosphatidylethanolamine (PE) occurs on the cytoplasmic leaflet of the endoplasmicreticulum (ER). PC and PE are synthesized by addition of the appropriate phosphoryl base to diacylglycerol (DAG), while PSis derived by base exchange of serine for the ethanolamine moiety of PE. Ceramide can be converted to galactosylceramide(GalCer) on the lumenal side of the ER. A non-specific flippase allows the rapid redistribution of the newly synthesizedphospholipids11–17 and GalCer.74 Glucosylceramide (GlcCer), the precursor of higher glycosphingolipids, is synthesizedat the cytosolic face of Golgi membranes. It can be translocated across the Golgi membrane and used in the biosynthesisof lactosylceramide (LacCer) and other glycosphingolipids, but it is unclear whether translocation is bi-directional. Thecomplex glycolipids and sphingomyelin (SM), which are produced at the lumenal face, do not translocate towards thecytosolic face.73,74 A candidate aminophospholipid translocase has been localized in the late Golgi.35,39 In the plasmamembrane, an ATP-dependent aminophospholipid translocase transports PS and PE towards the cytoplasmic leaflet andmaintains a permanent lipid asymmetry.9,25,27 Members of the ABC transporter family can translocate specific lipids fromthe cytoplasmic to the exoplasmic leaflet of the plasma membrane, e.g. MDR3 P-pg mediates PC translocation across thecanalicular domain of the hepatocyte membrane.43,45–47 Scramblase action, which depends on activation, results in rapidtransbilayer movement of all phospholipids and loss of lipid asymmetry in the plasma membrane.65–67

Reconstitution of flippase activity from bacteria hasbeen reported.24 The transport of a short-chainphospholipid analog and a long-chain phospholipidwas associated with the detergent extract of themembrane but not the lipid extract. Transport wasprotease-sensitive and was enriched in a fractionsedimenting at ∼4S on a glycerol gradient. Recoveryof activity in other gradient fractions was low despitethe presence of a complex mixture of membraneproteins. These data suggest that bacteria containspecific proteins capable of facilitating phospholipidflip–flop. Whether the bacterial and the ER flippasesturn out to be related proteins remains to be seen.

Lipid movement across the plasma membraneof eukaryotic cells

Maintenance and regulation of the asymmetric lipiddistribution across the plasma membrane is governedby the concerted action of specific membrane pro-teins controlling lipid movement across the bilayer.The inward movement of PC and sphingomyelin(SM) from the exoplasmic to the inner plasmamembrane is, under normal conditions for mostcells, a slow, non-mediated process. The presenceof cholesterol in the plasma membrane contributeslargely to the stability of SM transbilayer distribution.

142

Transbilayer lipid distribution and movement

In contrast, the aminophospholipids, PS and PE,are rapidly transported from the outer to the innerleaflet by an active ATP dependent and protein-mediated process, thus maintaining lipid asymmetry.The involvement of proteins and ATP dependencehas also been reported for the outward movementof lipids. The asymmetrical lipid distribution of theplasma membrane can be scrambled rapidly. ThisATP-independent process triggered by cytosolic cal-cium, would involve the action of a scramblase. Themaintenance of the nonrandom lipid distributionis important for cellular functions. Any change inthis distribution generally triggers a physiologicalevent. Exposure of PS at the surface of activated orinjured blood cells or endothelium serves to promoteblood coagulation (reviewed in Reference 25).Surface exposure of plasma membrane PS and PEsignals the removal of injured apoptotic cells by thereticuloendothelia system.26 Besides these functionsof an asymmetric lipid distribution in specific cells,the transfer of lipids from one leaflet to the other incellular membranes may be of general significancefor the functioning of a single cell. For example, itcould be involved in the regulation of membrane cur-vature, e.g. during endocytosis, or in the modulationof the activity of membrane proteins.27,28

Active inward translocation: the aminophospholipidtranslocase

The inward movement of aminophospholipidshas been shown to depend upon an ATP-drivenaminophospholipid translocase which transportsPS and PE selectively from the exoplasmic to thecytosolic leaflet of mammalian plasma membranes(t1/2=minutes for PS). First described in humanerythrocyte membrane,9 this aminophospholipidtranslocating activity has now been demonstratedin various plasma membranes and in chromaffingranules (reviewed in References 1,25,27). A fastATP-dependent inward movement of aminophos-pholipids analogs was also found for the yeast plasmamembrane.29,30 Here, aminophospholipid translo-cation was accompanied by rapid internalization ofPC analogs suggesting the presence of either anadditional PC-specific translocase or a new type oftranslocase, translocating aminophospholipids aswell as PC towards the cytoplasmic leaflet. Such arapid inward movement of PC was also found forthe basolateral plasma membrane of hepatocytes31

and in both the basolateral and apical membrane ofkidney epithelial cells.32

The aminophospholipid translocase has not yetbeen unequivocally identified. Tentative purificationsfrom erythrocytes33,34 suggested a Mg2+ ATPase witha molecular mass of 115–120 kDa. However purifica-tion and cloning of the gene encoding the ATPase IIfrom bovine chromaffin granules, another candidateprotein, revealed a slightly bigger protein belongingto a novel subfamily of P-type ATPases.35 Disruptionof the homologous gene in yeast, the DRS2 gene,abolished the internalization of a fluorescent PSanalog (C6-NBD-PS) at low temperature.35 This ob-servation was interpreted as evidence that the DRS2pand the ATPase from bovine chromaffin granulesare aminophospholipid translocases. Interestingly,Axelsen and Palmgren36 have claimed to have foundan analog of the chromaffin ATPase II in plant cellsthat is responsible for PS translocation. In laterstudies, but under somewhat different experimentalconditions, deletion of the DRS2 gene did neitherspecifically abolish the translocation of fluorescent PSand PE analogs, nor affect the preferential orienta-tion of endogenous aminophospholipids towards thecytoplasmic leaflet of yeasts.37,38 These observationsand the recent localization of the DRS2p in the Golgicomplex39 argue against the idea that this proteinacts as an exclusive or major aminophospholipidtranslocase in the plasma membrane of yeast.

Active outward translocation: involvement of ABCtransporters

Evidence for an ATP-dependent and protein-mediated outward movement of lipids towards thecell surface were found first for erythrocyte mem-brane.40,41 This transport activity has been observedfor analogs of the aminophospholipids as well as ofPC with rather slow rate (t1/2≈1.5 hours). However,a role in the transport of endogenous lipids has notbeen well established.

A different outward-directed translocase activitylocated in the plasma membrane was identified instudies originally related to multidrug resistance(MDR) in cancer cells. One form of MDR resultsfrom overexpression of the MDR1 P-glycoprotein (P-gp). This membrane protein belongs to ATP-bindingcassette (ABC) transporter family and extrudes awide variety of amphipathic drugs from cells. Theclosely related ABC transporter MDR3 P-gp is highlyexpressed in the bile canalicular membrane of hepa-tocytes. Mice with a disruption of the mdr2 gene (themouse homolog of MDR3), were found unable totransport PC into bile.42 Studies on secretory vesicles

143

T. Pomorski et al.

from yeast transfected with mdr2 P-gp showed thatthis protein transports the PC analog C6-NBD-PCfrom the cytoplasmic to the exoplasmic leaflet ofthe membrane bilayer43 (however, cf. Angeletti andNichols44 for a methodological note). Nies et al.45 confirmed this ATP-dependent PC translocationin isolated canalicular plasma membrane vesicles.Specificity for PC was demonstrated on polarizedepithelial cells transfected with human MDR3 P-gp.Newly synthesized C6-NBD-PC could reach theapical plasma membrane in the absence of vesiculartransport, while NBD analogs of PE, SM and glucosyl-ceramide (GlcCer) were not translocated.46 Evidencefor translocation of endogenous PC was provided infibroblasts from mdr2 knock out mice expressingMDR3 Pgp.47 Besides MDR3 Pgp, evidence has beenpresented for an ATP-independent PC flippase incanalicular plasma membrane.48

Unexpectedly, the drug transporter MDR1 P-gp wasfound to be responsible for transporting a variety ofshort-chain lipid analogs from the inner to the outerleaflet of the plasma membrane.46,49,50 Amongstthese were short-chain analogs of PC and of GlcCer,two lipids synthesized on the cytosolic surface of ERand Golgi. MDR1 P-gp was found unable to rescue PCtransport into the bile in the mdr2 knockout mice,42

suggesting that natural long-chain PC is not a naturalsubstrate. However, evidence has been obtained thatsecretion of the short-chain PC platelet activatingfactor (PAF) is greatly enhanced in cells transfectedwith MDR1 P-gp and is sensitive to inhibitors andsubstrates of MDR1 P-gp (Reference 51; R. Raggers,I. Vogels and G. van Meer, submitted). In addition,expression of MDR1 P-gp has been correlated withan increase in cellular GlcCer and higher glycol-ipids.52,53 Using an enzymatic assay for GlcCerappearance on the cell surface, it has now beenfound that MDR1 P-gp rescues GlcCer from hydroly-sis by a non-lysosomal enzyme through removal fromthe cytosolic leaflet of the plasma membrane andtranslocation towards the outer leaflet (R. Raggerset al., manuscript in preparation). All cells expressMDR1 P-gp, but expression is particularly high inthe apical plasma membrane domain of epithelialcells. The observation that an MDR1 P-gp knockoutmouse does not display lipid-related physiologicaldefects suggests that its activity as a lipid translocatoris physiologically irrelevant or that an alternativesystem(s) can compensate for the defect.54

The multidrug resistance protein MRP1, anotherdrug-pumping ABC-transporter, was also found totranslocate C6-NBD-analogs. In contrast to MDR1

P-gp, it did not translocate sphingolipids withoutthe NBD moiety nor C6-NBD-PC.55 In studies withMRP1 knockout mice, an MRP1-mediated transportof C6-NBD-PS (and C6-NBD-PC) was reported inerythrocytes. However, no changes in the distributionof endogenous PS were detected.56,57 Therefore, itis still unclear whether MRP1 is involved in thetranslocation of natural lipids.

Finally, mutations of the ABC transporter ABC1gene have been shown to be the cause of familialhigh-density lipoprotein (HDL) deficiency and Tang-ier disease, an autosomal recessive disorder of lipidmetabolism, resulting in very low plasma HDL levelsand increased amounts of cholesterol-ester storagein cells.58–60 Tangier fibroblasts manifest a decreasein the HDL- or apolipoprotein A1-induced efflux ofboth cholesterol and phospholipids (radiolabeled PCand SM)61,62 suggesting a possible role as cholesterolefflux transporter. ABC1 was reported to be involvedin the engulfment of apoptotic cells by increasing thetransbilayer movement of PS.63 Recent studies withABC1 knockout mice revealed that calcium triggeredspin-labeled PS redistrubution is indeed acceleratedby ABC1.64

Bi-directional movement: the scramblase

Calcium-influx accompanying cellular activationcauses a loss of phospholipid asymmetry in theplasma membrane. The scrambling process is bi-directional and involves all major phospholipidclasses, moving at comparable rates (t1/2≈10 to20 minutes), with slower mobility for SM.65–67 Onereport has suggested that, at least in platelets, PS andPE are preferentially externalized.68 Inhibition ofthe aminophospholipid translocase alone does notdirectly lead to the exposure of PS. Since phospho-lipid scrambling is inhibited by protein-modifyingreagents67 and is ATP-independent, the activationof a specific calcium-dependent flippase, calledscramblase, has been suggested. A rare and verysevere bleeding disorder called Scott syndrome wasthought to correspond to the lack of the scramblase;at least it is associated with an impairment of cal-cium triggered lipid redistribution in platelets.69 A37 kDa membrane protein from erythrocytes wascapable of mediating calcium-dependent transbi-layer movement of phospholipids in reconstitutedliposomes.70,71 However, the apparent rate of phos-pholipid scrambling was rather low (t1/2≈2 hours)

144

Transbilayer lipid distribution and movement

and this protein is present in blood cells of the Scottpatients.72 Hence, the identity of the scramblase isstill obscure.

Lipid movement across other intracellularmembranes

In the Golgi, GlcCer, the precursor for the higherglycosphingolipids, is synthesized on the cytoplasmicleaflet. All higher glycosphingolipids are synthesizedon the lumenal leaflet, resulting in the presenceof numerous different glycosphingolipids on thecell surface. Therefore, GlcCer has to translocateacross the Golgi membrane for synthesis of higherglycolipids. This has been demonstrated with a short-chain analog (C6-NBD-GlcCer) on isolated Golgimembranes,73,74 suggesting that a GlcCer flippaseis present in the Golgi membrane. The flippasedid not require exogenous energy, allowed passageof galactosylceramide as well as GlcCer and didnot allow the metabolic products lactosylceramide,(galactosyl)2ceramide or sulfatide to translocate backto the cytosolic surface. The localization of a putativeaminophospholipid translocator in the late Golgi,39

raises the possibility that lipid asymmetry might beestablished in the Golgi complex.

Evidence for a novel ATP-dependent phospholipidtranslocase has been recently reported for gastricvesicles which translocates PS, PE and PC fromthe cytosolic to the lumenal leaflet.75 Transportwas dependent on ATP on the cytosolic side andabolished by an inhibitor of the gastric H+, K+-ATPase but not by an inhibitor or substrates of themultidrug resistance proteins. A similar translocaseactivity was reported for synaptic vesicles but assumedto depend on lumenal ATP.76

Summarizing remarks

Although the last decades have brought us firstinsights into the transversal orientation of mem-brane lipids and the mechanisms responsible forgenerating and maintaining this organization, manyquestions remain. With respect to the transbilayerdistribution of the lipids, little is known concerningthe orientation of cholesterol, one of the majorlipids in the plasma membrane, and other moreminor lipids like phosphatidylinositol and its phos-phorylated derivatives. An additional complexity is

the fact that the lateral distribution of the variouslipids in the plane of the membrane is probably nothomogeneous. Apart from difficulties in definingthe relevant parameters of lipid asymmetry, thispredicts that not all molecules of a certain lipid classpresent in one membrane leaflet will translocatewith the same kinetics. The flippases or translocasesmay be preferentially localized in a specific lipidenvironment.

A number of proteins have now been identifiedas being involved in the translocation of lipids.In essentially all cases it remains to be establishedwhether these proteins can translocate a certain lipidacross a lipid bilayer on their own, or whether they arepart of an oligomeric complex. To resolve this issue,reconstitution of purified proteins into liposomeswould seem the only suitable method. Only thenwill we be ready to start addressing the problem ofhow the activities of the various lipid transporters areregulated and coordinated to satisfy the physiologicaldemands of a dynamic cell.

Acknowledgements

We thank J.C.M. Holthuis for helpful discussionsand critically reading the manuscript; A. Herrmannand D. Wustner for communicating results priorto publication. T.P. acknowledges support from anEMBO fellowship. P.F.D. contributed to this reviewduring a stay in the Netherlands made possible bythe Huygens–Descartes award of the year 2000 fromthe Netherland Academy of Sciences. A researchtraining grant from the EC, HPRN-CT-2000-00077, isacknowledged to S. H., P. F. D., and G. V. M.

References

1. Zachowski A (1993) Phospholipids in animal eukaryoticmembranes: transverse asymmetry and movement. BiochemJ 294:1–14

2. Bretscher MS (1972) Phosphatidylethanolamine: differentiallabelling in intact cells and cell ghosts of human erythrocytesby a membrane-impermeable reagent. J Mol Biol 71:523–528

3. Verkleij AJ, Zwaal RFA, Roelofsen B, Comfurius P, KastelijnD, van Deenen LLM (1973) The asymmetric distribution ofphospholipids in the human red cell membrane. A com-bined study using phospholipases and freeze-etch electron mi-croscopy. Biochim Biophys Acta 323:178–193

4. Crain RC, Zilversmit DB (1980) Two nonspecific phospho-lipid exchange proteins from beef liver. 2. Use in studyingthe asymmetry and transbilayer movement of phosphatidyl-choline, phosphatidylethanolamine, and sphingomyelin in in-

145

T. Pomorski et al.

tact rat erythrocytes. Biochemistry 19:1440–14475. van Meer G, Poorthuis BJHM, Wirtz KWA, Op den Kamp JAF,

van Deenen LLM (1980) Transbilayer distribution and mobil-ity of phosphatidylcholine in intact erythrocyte membranes.A study with phosphatidylcholine exchange protein. Eur JBiochem 103:283–288

6. Fong BS, Hunt RC, Brown JC (1976) Asymmetric distributionof phosphatidylethanolamine in the membrane of vesicularstomatitis virus. J Virol 20:658–663

7. Lenard J, Rothman JE (1976) Transbilayer distribution andmovement of cholesterol and phospholipid in the membraneof influenza virus. Proc Natl Acad Sci USA 73:391–395

8. Rothman JE, Tsai DK, Dawidowicz EA, Lenard J (1976)Transbilayer phospholipid asymmetry and its maintenance inthe membrane of influenza virus. Biochemistry 15:2361–2370

9. Seigneuret M, Devaux PF (1984) ATP-dependent asymmetricdistribution of spin-labeled phospholipids in the erythrocytemembrane: relation to shape changes. Proc Nat Acad Sci USA81:3751–3755

10. Tait JF, Gibson D (1994) Measurement of membrane phos-pholipid asymmetry in normal and sickle-cell erythrocytes bymeans of annexin V binding. J Lab & Clin Med 123:741–748

11. Zilversmit DB, Hughes ME (1977) Extensive exchange ofrat liver microsomal phospholipids. Biochim Biophys Acta469:99–110

12. van den Besselaar AMHP, de Kruijff B, van den Bosch H, vanDeenen LLM (1978) Phosphatidylcholine mobility in livermicrosomal membranes. Biochim Biophys Acta 510:242–255

13. Bishop WR, Bell RM (1985) Assembly of the endoplas-mic reticulum phospholipid bilayer: the phosphatidylcholinetransporter. Cell 42:51–60

14. Backer JM, Dawidowicz EA (1987) Reconstitution of a phos-pholipid flippase from rat liver microsomes. Nature 327:341–343

15. Herrmann A, Zachowski A, Devaux PF (1990) Protein-mediated phospholipid translocation in the endoplasmicreticulum with a low lipid specificity. Biochemistry 29:2023–2027

16. Buton X, Morrot G, Fellmann P, Seigneuret M (1996) Ultra-fast glycerophospholipid-selective transbilayer motion medi-ated by a protein in the endoplasmic reticulum membrane. JBiol Chem 271:6651–6657

17. Menon AK, Watkins WE 3rd, Hrafnsdóttir S (2000) Specificproteins are required to translocate phosphatidylcholinebidirectionally across the endoplasmic reticulum. Curr Biol10:241–252

18. Rush JS, van Leyen K, Ouerfelli O, Wolucka B, WaechterCJ (1998) Transbilayer movement of Glc-P-dolichol and itsfunction as a glucosyl donor: protein-mediated transport ofa water-soluble analog into sealed ER vesicles from pig brain.Glycobiology 8:1195–1205

19. Rothman JE, Kennedy EP (1977) Rapid transmembranemovement of newly synthesized phospholipids during mem-brane assembly. Proc Natl Acad Sci USA 74:1821–1825

20. Langley KE, Kennedy EP (1979) Energetics of rapid trans-membrane movement and of compositional asymmetry ofphosphatidylethanolamine in membranes of Bacillus mega-terium. Proc Natl Acad Sci USA 76:6245–6249

21. Huijbregts RP, de Kroon AI, de Kruijff B (1996) Rapidtransmembrane movement of C6-NBD-labeled phospholipidsacross the inner membrane of Escherichia coli. Biochim Bio-phys Acta 1280:41–50

22. Hrafnsdottir S, Nichols JW, Menon AK (1997) Transbilayermovement of fluorescent phospholipids in Bacillus megateriummembrane vesicles. Biochemistry 36:4969–4978

23. Huijbregts RP, de Kroon AI, de Kruijff B (1998) Rapidtransmembrane movement of newly synthesized phos-phatidylethanolamine across the inner membrane of Es-cherichia coli. J Biol Chem 273:18936–18942

24. Hrafnsdottir S, Menon AK (2000) Reconstitution and partialcharacterization of phospholipid flippase activity from deter-gent extracts of the Bacillus subtilis cell membrane. J Bacteriol182:4198–4206

25. Bevers EM, Comfurius P, Dekkers DW, Zwaal RF (1999) Lipidtranslocation across the plasma membrane of mammaliancells. Biochim Biophys Acta 1439:317–330

26. Fadok VA, Bratton DL, Rose DM, Pearson A, Ezekewitz RA,Henson PM (2000) A receptor for phosphatidylserine-specificclearance of apoptotic cells. Nature 405:85–90

27. Devaux PF (1991) Static and dynamic lipid asymmetry in cellmembranes. Biochemistry 30:1163–1173

28. Devaux PF (2000) Is lipid translocation involved during endo-and exocytosis? Biochimie 82:497–509

29. Kean LS, Fuller RS, Nichols JW (1993) Retrograde lipid trafficin yeast: identification of two distinct pathways for internal-ization of fluorescent-labeled phosphatidylcholine from theplasma membrane. J Cell Biol 123:1403–1419

30. Kean LS, Grant AM, Angeletti C, Mahé Y, Kuchler K, FullerRS, Nichols JW (1997) Plasma membrane translocation offluorescent-labeled phosphatidylethanolamine is controlledby transcription regulators, PDR1 and PDR3. J Cell Biol138:255–270

31. Muller P, Pomorski T, Porwoli S, Tauber R, Herrmann A(1996) Transverse movement of spin-labeled phospholipidsin the plasma membrane of a hepatocytic cell line (HepG2):implications for biliary lipid secretion. Hepatology 24:1497–1450

32. Pomorski T, Herrmann A, Muller P, van Meer G, Burger K(1999) Protein-mediated inward translocation of phospho-lipids occurs in both the apical and basolateral plasma mem-brane domains of epithelial cells. Biochemistry 38:142–150

33. Morrot G, Hervé P, Zachowski A, Fellmann P, Devaux PF(1989) Aminophospholipid translocase of human erythro-cytes: phospholipid substrate specificity and effect of choles-terol. Biochemistry 28:3456–3462

34. Zimmerman ML, Daleke DL (1993) Regulation of a candidateaminophospholipid-transporting ATPase by lipid. Biochem-istry 32:12257–12263

35. Tang X, Halleck MS, Schlegel RA, Williamson P (1996) A sub-family of P-type ATPases with aminophospholipid transport-ing activity. Science 272:1495–1497

36. Axelsen KB, Palmgren MG (1998) Evolution of substratespecificities in the P-type ATPase superfamily. J Mol Evol46:84–101

37. Siegmund A, Grant A, Angeletti C, Malone L, NicholsJW, Rudolph HK (1998) Loss of Drs2p does not abolishtransfer of fluorescence-labeled phospholipids across theplasma membrane of Saccharomyces cerevisiae. J Biol Chem273:34399–34405

38. Marx U, Polakowski T, Pomorski T, Lang C, Nelson H, NelsonN, Herrmann A (1999) Rapid transbilayer movement offluorescent phospholipid analogues in the plasma membraneof endocytosis-deficient yeast cells does not require the drs2protein. Eur J Biochem 263:254–264

39. Chen CY, Ingram MF, Rosal PH, Graham TR (1999) Rolefor Drs2p, a P-type ATPase and potential aminophospholipidtranslocase, in yeast late Golgi function. J Cell Biol 147:1223–1236

40. Bitbol M, Devaux PF (1988) Measurement of outward translo-cation of phospholipids across human erythrocyte mem-

146

Transbilayer lipid distribution and movement

brane. Proc Natl Acad Sci USA 85:6783–678741. Connor J, Pak CH, Zwaal RFA, Schroit AJ (1992) Bidirectional

transbilayer movement of phospholipid analogs in human redblood cells. J Biol Chem 267:19412–19417

42. Smit JJM et al. (1993) Homozygous disruption of the murinemdr2 P-glycoprotein gene leads to a complete absence ofphospholipid from bile and to liver disease. Cell 75:451–462

43. Ruetz S, Gros P (1994) Phosphatidylcholine translocase: aphysiological role for the mdr2 gene. Cell 77:1–20

44. Angeletti C, Nichols JW (1998) Dithionite quenching ratemeasurement of the inside-outside membrane bilayer dis-tribution of 7-nitrobenz-2-oxa-1,3-diazol-4-yl-labeled phospho-lipids. Biochemistry 37:15114–15119

45. Nies AT, Gatmaitan Z, Arias IM (1996) ATP-dependentphosphatidylcholine translocation in rat liver canalicularplasma membrane vesicles. J Lipid Res 37:1125–1136

46. van Helvoort A, Smith AJ, Sprong H, Fritzsche I, Schinkel AH,Borst P, van Meer G (1996) MDR1 P-glycoprotein is a lipidtranslocase of broad specificity, while MDR3 P-glycoproteinspecifically translocates phosphatidylcholine. Cell 87:507–517

47. Smith AJ, Timmermans-Hereijgers JLPM, Roelofsen B, WirtzKWA, van Blitterswijk WJ, Smit JJM, Schinkel AH, Borst P(1994) The human MDR3 P-glycoprotein promotes translo-cation of phosphatidylcholine through the plasma membraneof fibroblasts from transgenic mice. FEBS Lett 354:263–266

48. Berr F, Meier PJ, Stieger B (1993) Evidence for the presenceof a phosphatidylcholine translocator in isolated rat livercanalicular plasma membrane vesicles. J Biol Chem 268:3976–3979

49. Bosch I, Dunussi-Joannopoulos K, Wu R-L, Furlong ST, CroopJ (1997) Phosphatidylcholine and phosphatidylethanolaminebehave as substrates of the human MDR1 P-glycoprotein.Biochemistry 36:5685–5694

50. Abulrob AG, Gumbleton M (1999) Transport of phosphatidyl-choline in MDR3-negative epithelial cell lines via drug-induced MDR1 P-glycoprotein. Biochem Biophys Res Com-mun 262:121–126

51. Ernest S, Bello-Reuss E (1999) Secretion of platelet-activatingfactor is mediated by MDR1 P-glycoprotein in culturedhuman mesangial cells. J Am Soc Nephrol 10:2306–2313

52. Lavie Y, Cao HT, Bursten SL, Giuliano AE, Cabot MC(1996) Accumulation of glucosylceramides in multidrug-resistant cancer cells. J Biol Chem 271:19530–19536

53. Lala P, Ito S, Lingwood CA (2000) Retroviral transfection ofMadin–Darby canine kidney cells with human MDR1 resultsin a major increase in globotriaosylceramide and 10(5)- to10(6)-fold increased cell sensitivity to verocytotoxin — role ofP-glycoprotein in glycolipid synthesis. J Biol Chem 275:6246–6625

54. Schinkel AH, Wagenaar E, van Deemter L, Mol CAAM,Borst P (1995) Absence of the mdr1a P-glycoprotein inmice affects tissue distribution and pharmacokinetics ofdexamethasone, digoxin, and cyclosporin A. J Clin Invest96:1698–1705

55. Raggers RJ, van Helvoort A, Evers R, van Meer G (1999) Thehuman multidrug resistance protein MRP1 translocates sph-ingolipid analogs across the plasma membrane. J Cell Sci112:415–422

56. Kamp D, Haest CWM (1998) Evidence for a role of the mul-tidrug resistance protein (MRP) in the outward translocationof NBD-phospholipids in the erythrocyte membrane. BiochimBiophys Acta 1372:91–101

57. Dekkers DWC, Comfurius P, Schroit AJ, Bevers EM, ZwaalRFA (1998) Transbilayer movement of NBD-labeled phospho-lipids in red blood cell membranes: Outward directed trans-

port by the Multidrug Resistance Protein 1 (MRP1). Biochem-istry 37:14833–14837

58. Bodzioch M et al. (1999) The gene encoding ATP-bindingcassette transporter 1 is mutated in Tangier disease. Nat Genet22:347–351

59. Rust S et al. (1999) Tangier disease is caused by mutations inthe gene encoding ATP-binding cassette transporter 1. NatGenet 22:352–355

60. Brooks-Wilson A et al. (1999) Mutations in ABC1 in Tangierdisease and familial high-density lipoprotein deficiency. NatGenet 22:336–345

61. Francis GA, Knopp RH, Oram JF (1995) Defective removal ofcellular cholesterol and phospholipids by apolipoprotein A-Iin Tangier Disease. J Clin Invest 96:78–87

62. Rogler G, Trumbach B, Klima B, Lackner KJ, Schmitz G(1995) HDL-mediated efflux of intracellular cholesterol is im-paired in fibroblasts from Tangier disease patients. Arte-rioscler Thromb Vasc Biol 15:683–690

63. Marguet D, Luciani MF, Moynault A, Williamson P, Chimini G(1999) Engulfment of apoptotic cells involves the redistribu-tion of membrane phosphatidylserine on phagocyte and prey.Nature Cell Biol 1:454–456

64. Hamon Y et al. (2000) ABC1 promotes engulfment of apop-totic cells and transbilayer redistribution of phosphatidylser-ine. Nature Cell Biol 2:399–406

65. Williamson P, Kulick A, Zachowski A, Schlegel RA, DevauxPF (1992) Ca2+ induces transbilayer redistribution of allmajor phospholipids in human erythrocytes. Biochemistry31:6355–6360

66. Smeets EF, Comfurius P, Bevers EM, Zwaal RF (1994)Calcium-induced transbilayer scrambling of fluorescent phos-pholipid analogs in platelets and erythrocytes. Biochim Bio-phys Acta 1195:281–286

67. Williamson P, Bevers EM, Smeets EF, Comfurius P, SchlegelRA, Zwaal RF (1995) Continuous analysis of the mechanism ofactivated transbilayer lipid movement in platelets. Biochem-istry 34:10448–10455

68. Basse F, Gaffet P, Rendu F, Bienvenue A (1993) Translocationof spin-labeled phospholipids through plasma membraneduring thrombin- and ionophore A23187-induced plateletactivation. Biochemistry 32:2337–2344

69. Bevers EM, Wiedmer T, Comfurius P, Shattil SJ, Weiss HJ,Zwaal RF, Sims PJ (1992) Defective Ca(2+)-induced mi-crovesiculation and deficient expression of procoagulant ac-tivity in erythrocytes from a patient with a bleeding disorder: astudy of the red blood cells of Scott syndrome. Blood 79:380–388

70. Bassé F, Stout JG, Sims PJ, Wiedmer T (1996) Isolationof an erythrocyte membrane protein that mediates Ca2+-dependent transbilayer movement of phospholipid. J BiolChem 271:17205–17210

71. Zhou Q, Zhao J, Stout JG, Luhm RA, Wiedmer T, SimsPJ (1997) Molecular cloning of human plasma membranephospholipid scramblase. A protein mediating transbilayermovement of plasma membrane phospholipids. J Biol Chem272:18240–18244

72. Stout JG, Basse F, Luhm RA, Weiss HJ, Wiedmer T, SimsPJ (1997) Scott syndrome erythrocytes contain a membraneprotein capable of mediating Ca2+-dependent transbilayermigration of membrane phospholipids. J Clin Invest 99:2232–2238

73. Lannert H, Bünning C, Jeckel D, Wieland FT (1994) Lactosyl-ceramide is synthesized in the lumen of the Golgi apparatus.FEBS Lett 342:91–96

74. Burger KNJ, van der Bijl P, van Meer G (1996) Topology of

147

T. Pomorski et al.

sphingolipid galactosyltransferases in ER and Golgi: transbi-layer movement of monohexosyl sphingolipids is required forhigher glycosphingolipid biosynthesis. J Cell Biol 133:15–28

75. Suzuki H, Kamakura M, Morii M, Takeguchi N (1997) Thephospholipid flippase activity of gastric vesicles. J Biol Chem272:10429–10434

76. Lee DS, Anzai K, Hirashima N, Kirino Y (1998) Phospholipidtranslocation from the outer to the inner leaflet of synap-tic vesicle membranes isolated from the electric organ ofJapanese electric ray Narke japonica. J Biochem 124:798–803

148