Search

The structural and functional consequences of a mismatch between the hydrophobic thickness d(P) of a transmembrane protein and that d(L) of the supporting lipid bilayer were investigated using melibiose permease (MelB) from Escherichia coli reconstituted in a set of bis saturated and monounsaturated phosphatidylcholine species differing in acyl-chain length. Influence of MelB on the midpoint gel-to-liquid-phase transition temperature, T(m), of the saturated lipids was investigated through fluorescence polarization experiments, with 1,6-diphenyl-1,3,5-hexatriene as the probe, for varying protein/lipid molar ratio. Diagrams in temperature versus MelB concentration showed positive or negative shifts in T(m) with the short-chain lipids DiC12:0-PC and DiC14:0-PC or the long-chain lipids DiC16:0-PC and DiC18:0-PC, respectively. Theoretical analysis of the data yielded a d(L) value of 3.0 +/- 0.1 nm for the protein, similar to the 3.02 nm estimated from hydropathy profiles. Influence of the acyl chain length on the carrier activity of MelB was investigated in the liquid phase, using the monounsaturated PCs. Binding of the sugar to the transporter showed no dependence on the acyl chain length. In contrast, counterflow and Deltapsi-driven experiments revealed strong dependence of melibiose transport on the lipid acyl chain length. Similar bell-shaped transport versus acyl chain length profiles were obtained, optimal activity being supported by diC16:1-PC. On account of a d(P) value of 2.65 nm for the lipid and of various local constraints which would all tend to elongate the acyl chains in contact with the protein, one can conclude that maximal activity was obtained when the hydrophobic thickness of the bilayer matched that of the protein.

A combined experimental and theoretical study is performed on binary dilauroylphosphatidylcholine/distearoylphosphatidylcholine (DLPC/DSPC) lipid bilayer membranes incorporating bacteriorhodopsin (BR). The system is designed to investigate the possibility that BR, via a hydrophobic matching principle related to the difference in lipid bilayer hydrophobic thickness and protein hydrophobic length, can perform molecular sorting of the lipids at the lipid-protein interface, leading to lipid specificity/selectivity that is controlled solely by physical factors. The study takes advantage of the strongly nonideal mixing behavior of the DLPC/DSPC mixture and the fact that the average lipid acyl-chain length is strongly dependent on temperature, particularly in the main phase transition region. The experiments are based on fluorescence energy transfer techniques using specifically designed lipid analogs that can probe the lipid-protein interface. The theoretical calculations exploit a microscopic molecular interaction model that embodies the hydrophobic matching as a key parameter. At low temperatures, in the gel-gel coexistence region, experimental and theoretical data consistently indicate that BR is associated with the short-chain lipid DLPC. At moderate temperatures, in the fluid-gel coexistence region, BR remains in the fluid phase, which is mainly composed of short-chain lipid DLPC, but is enriched at the interface between the fluid and gel domains. At high temperatures, in the fluid phase, BR stays in the mixed lipid phase, and the theoretical data suggest a preference of the protein for the long-chain DSPC molecules at the expense of the short-chain DLPC molecules. The combined results of the experiments and the calculations provide evidence that a molecular sorting principle is active because of hydrophobic matching and that BR exhibits physical lipid selectivity. The results are discussed in the general context of membrane organization and compartmentalization and in terms of nanometer-scale lipid-domain formation.

The tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA) is a potent mitogenic factor which can replace the growth promoting activity of basic fibroblast growth factor (bFGF) on bovine aortic endothelial cells. However, TPA-treated cells lose their strict contact inhibition at confluence, which is a characteristic of cells grown in the presence of bFGF. We have examined whether these changes could be related to modifications of the transbilayer and lateral motions of fluorescent lipids, namely 1-acyl-2-[6-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]caproyl]-p hosphatidylcholine (C6-NBD-PC), -phosphatidylserine (C6-NBD-PS), and -phosphatidylethanolamine (C6-NBD-PE) inserted in the outer leaflet of the cell plasma membrane. In TPA-treated cells, the three fluorescent phospholipids remained located in the outer leaflet for at least 1 h at 20 degrees C after their insertion, indicating a blockade of the aminophospholipid translocase activity which is normally present in the plasma membrane of bFGF-treated cells. TPA also induced a large increase in the percentage of C6-NBD-PC and C6-NBD-PE probes which were free to diffuse laterally. The mobile fractions M reached values of approximately 100% for the two lipids, while for bFGF-treated cells they were found around 85 and 75%, respectively. For the C6-NBD-PS probe, M remained unchanged in bFGF and TPA-treated cells, at around 85%. TPA treatment also induced a twofold increase in the lateral diffusion coefficients of C6-NBD-PC and C6-NBD-PE, while that of C6-NBD-PS remained nearly unchanged. These effects of TPA may be related to the observed loss of differentiated properties of vascular endothelial cells and not to its mitogenic properties.

Steady-state and time-resolved fluorescence properties of the 7-nitrobenz-2-oxa-1, 3-diazole-4-yl (NBD) fluorophore attached either to the sn-2 acyl chain of various phospholipids (phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and phosphatidic acid) or to the polar headgroup of phosphatidylethanolamine were studied after insertion of these NBD-labeled lipid probes into unilamellar vesicles of phosphatidylcholine, phosphatidylglycerol, phosphatidic acid, and phosphatidylserine. The fluorescence response of the NBD group was observed to strongly depend on the chemical structure and physical state of the host phospholipids and on the chemical structure of the lipid probe itself. Among the various fluorescence parameters studied, i.e., Stokes' shifts, lifetimes, and quantum yields, the quantum yields were by far the most affected by these structural and environmental factors, whereas the Stokes' shifts were practically unaffected. Thus, depending on the phospholipid probe and the host phospholipid, the fluorescence emission of the NBD group was found to vary by a factor of up to 5. Careful analysis of the data shows that for the various couples of probe and host lipid molecules studied, deexcitation of the fluorophore was dominated by nonradiative deactivation processes. This great sensitivity of the NBD group to environmental factors originates from its well-known solvatochromic properties, and comparison of these knr values with those obtained for n-propylamino-NBD in a set of organic solvents covering a large scale of polarity indicates that in phospholipids, the NBD fluorophore experiences a dielectric constant of around 27–41, corresponding to a medium of relatively high polarity. From these epsilon values and on the basis of models of the dielectric transition that characterizes any water-phospholipid interface, it can be inferred that for all of the phospholipid probes and host phospholipids tested, the NBD group is located in the region of the polar headgroups, near the phosphoglycerol moiety of the lipids.

Recently, we have developed a photocrosslinking approach which uses anthracene as a photoactivatable group and which allows us to determine the lateral distribution of lipids in membranes quantitatively. In synchronous cultures of the gram-positive bacterium Micrococcus luteus, this approach shows that the spatial distribution of phosphatidylglycerol and diamnnosyldiacylglycerol, the two major lipids in the bacterial membrane, varies greatly during the cell cycle. Minimum heterogeneity was observed during cell growth while maximum heterogeneity was detected during cell division.

Vascular endothelial cells form the inner nonthrombogenic lining of the large blood vessel. Through back-exchange and fluorescence recovery after photobleaching experiments and using the two fluorescent lipids 1-acyl-2-[6-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino] hexanoyl]glycerophosphocholine and 1-acyl-2[6-N-(7-nitrobenz-2-oxa-1, 3-diazol-4-yl)amino]hexanoyl]glycerophosphoethanolamine, we have recently shown that an energy-dependent and protein-dependent aminophospholipid translocase activity is present in the plasma membrane of cultured bovine aortic endothelial cells, which specifically transports phosphatidylethanolamine from the outer leaflet toward the inner leaflet of the membrane lipid bilayer. In the present study, using the same approach and 1-acyl-2-[6-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]hexanoyl] glycerophosphoserine as the probe, it is shown that this conclusion is also valid for phosphatidylserine. Furthermore, evidence is presented indicating that this aminophospholipid translocase activity can be maintained, suppressed, and restored at will, depending on the conditions of cell incubation. Thus, the translocase activity is detected for cells maintained in their normal culture medium or in a serum-free incubation medium [Dulbecco's modified Eagle's medium (DMEM)] supplemented with the basic fibroblast growth factor, whereas inhibition is observed for cells exposed for at least 2 h to DMEM. The translocase activity is restored when these pretreated cells are further incubated at least for 1 h in the presence of serum or of basic fibroblast growth factor. In view of the importance of basic fibroblast growth factor as a mitogenic and differentiating agent for vascular endothelial cells, various growth factors were tested (acidic fibroblast growth factor, epidermal growth factor, platelet-derived growth factor, transforming growth factors alpha and beta, vascular endothelial growth factor, interferon gamma, tumor-necrosis factor, insulin, and interleukin 4). Only basic fibroblast growth factor was active in the maintenance and restoration of the translocase activity. With respect to the effects of serum, evidence is presented showing that high-density lipoproteins might play a role in the control of the translocase activity. However, the positive effects of basic fibroblast growth factor, serum and high-density lipoproteins on the translocase activity were suppressed when experiments were carried out in the presence of an anti-(basic fibroblast growth factor) IgG, thus indicating that in all cases, basic fibroblast growth factor was directly involved in the modulation of the aminophospholipid translocase activity present in the plasma membrane of bovine aortic endothelial cells.

Electrophoretic mobilities of negatively charged lipid vesicles were measured after incubation with cationic drugs (adriamycin, celiptium, ethidium bromide, tetracaine) at different drug concentrations. We show here that comparison of the measured electrophoretic mobility with adsorption isotherms allows calculation of the drug-binding constants over a large range of drug concentrations, whatever the structure of the drug, its topology into the lipid bilayer, or its aggregation state

Glycopeptidolipids (GPLs) are specific constituents of mycobacteria known as opportunistic pathogens. The influence of the carbohydrate moiety on GPL-induced membrane alterations was examined with GPLs bearing 1-5 sugar residues (GPL-1 to GPL-5) and a sulfated GPL (S-GPL-2). GPLs decreased the ADP/O ratio and increased controlled respiration of isolated mitochondria. The more polar GPLs were the less active, with the following order of efficiency: GPL-1 > GPL-2 > S-GPL-2 = GPL-3 = GPL-5. GPL-1 and GPL-2 increased passive permeability of liposomes to carboxyfluorescein (GPL-1 > GPL-2), while GPL-3 and GPL-5 were inactive. GPL-2 and GPL-3 decreased the transmembrane electrical potential (delta psi) in isolated mitochondria (GPL-2 > GPL-3). These results suggest that GPLs uncouple oxidative phosphorylation by increasing the passive permeability of the mitochondrial membrane to protons. Compression isotherms of GPL-2 monolayers showed that, at low surface pressure, the area per GPL-2 molecule was about 5 times that of an acyl chain: it is likely that the peptide moiety was at the air/water interface. With an increase in the surface pressure, its area decreased, down to that of a tightly packed acyl chain. It is postulated that the glycopeptidic moiety can be either at in the interface or dipping into the water.(ABSTRACT TRUNCATED AT 250 WORDS)

In the plasma membrane of various eucaryotic cell types, in particular blood platelets and erythrocytes, it is known that phospholipids are asymmetrically distributed between the two leaflets of the lipid bilayer and that this transverse asymmetry is controlled by an aminophospholipid translocase activity. In this respect, it was of interest to check whether there are differential transbilayer movements between amino- and neutral phospholipids in the apical plasma membrane of vascular endothelial cells which form the inner nonthrombogenic lining of the large blood vessel. In the first step we compared the transbilayer localization and also the rate of lateral motion of two fluorescent analogs of phosphatidylcholine and phosphatidylethanolamine, namely C6-NBD-PC and C6-NBD-PE, inserted into the apical plasma membrane of bovine aortic endothelial cells, in vitro. By the use of back-exchange experiments we have found that C6-NBD-PC could be removed from the cell membrane toward the culture medium regardless of the incubation conditions used, i.e., just after cell labeling at 0 degrees C or even after further cell incubation for 1 h at 0 or 20 degrees C. In contrast, C6-NBD-PE could be removed only when the cells were maintained at 0 degrees C. After incubation for 1 h at 20 degrees C, 85% of the probe molecules remained nonexchangeable, indicating probe translocation from the outer to the inner leaflet of the lipid bilayer. This "flip" process, which occurred at 20 degrees C, was abolished when the endothelial cells were preincubated with N-ethylmaleimide, diamide, vanadate (VO4(3-)) and vanadyl (VO2+) ions, a set of substances which inhibit aminophospholipid translocase activity in various systems, and with a combination of sodium azide and 2-deoxyglucose which led to nearly complete ATP depletion in the cells. Fluorescence recovery after photobleaching experiments were also carried out to specify more precisely the localization and dynamics of the probes in the two leaflets of the plasma membrane lipid bilayer. They produced lateral diffusion coefficients D of 1.2 +/- 0.05 x 10(-9) cm2/s for C6-NBD-PC and 2.8 +/- 0.3 x 10(-9) cm2/s for C6-NBD-PE, when the two probes were located in the outer leaflet of the plasma membrane, just after cell labeling at 0 degree C.(ABSTRACT TRUNCATED AT 400 WORDS)

Two cell lines of Eucalyptus gunnii have been shown to keep their differential frost tolerance at the cellular level after long-term culture. They have been used to investigate the fluidity of specific cell membranes in relation with frost tolerance. Protoplasts and isolated vacuoles were obtained from both cell lines. In addition, purified plasma membrane and tonoplast (the vacuolar membrane) were separated from a crude microsomal fraction through free-flow electrophoresis. The lateral and rotational mobilities of lipids in these different membranes were studied by two biophysical techniques: fluorescence recovery after photobleaching (FRAP) and fluorescence polarization. After labeling the vacuoles isolated from the frost-sensitive cells with 1-oleoyl-2-(7-nitro-2,1,3-benz-oxadiazol-4-yl)aminocaproyl phosphatidylcholine, a single mobile component was observed with a diffusion coefficient of 2.4 x 10(-9) cm(2) s(-1) and a mobile fraction close to 100% at a temperature of 23 degrees C. When using isolated vacuoles from the frost tolerant line, a higher lateral diffusion of tonoplast lipids was found with a diffusion coefficient of 3.2 x 10(-9) cm(2) s(-1), still with a mobile fraction close to 100%. No convincing data were obtained when performing fluorescence recovery after photobleaching experiments on protoplasts. Fluorescence polarization experiments confirmed the differential behavior of the two cell lines for tonoplast and also for plasma membrane. In addition, they showed that intrinsically tonoplast exhibited a higher fluidity than plasma membrane. Our results provide the first information on the fluidity of tonoplast and on the compared properties of two important plant membranes-tonoplast and plasma membrane-through the use of two complementary biophysical approaches. In addition, they suggest there is a correlation between membrane fluidity and cold tolerance. The potential interest of plant vacuole as a natural model system in membrane studies is emphasized.

Micrococcus luteus bacterium has been used as a tool for the preparation of a fluorescent (anthracene labeled) cardiolipin by a method involving two steps. First, 9-(2-anthryl)nonanoic acid is incorporated into the lipids of the bacterium by usual metabolic pathways; then phosphatidylglycerol, a major lipid in this bacterium, is converted into cardiolipin by endogeneous cardiolipin synthetase. It is shown that cardiolipin synthetase transforms anthracene–phosphatidylglycerol into anthracene–cardiolipin at a high rate. The method can be extended to the synthesis of cardiolipin molecules bearing other hydrophobic marker groups.Key words: cardiolipin, fluorescence, biosynthesis, Micrococcus luteus.

International audience; We have used a published method of membrane preparation based on the precoating of the apical membrane of aortic endothelial cells with cationic silica microbeads (with or without polyacrylic acid) in combination with an osmotic shock and mechanical shearing to isolate the apical from the basal plasma membranes of these cells, in vitro. After labeling of the plasma membrane of adherent endothelial cells with a fluorescent derivative of phosphatidylcholine and by using laser confocal fluorescence scanning microscopy, we found that this method of membrane isolation rapidly induced invaginations of the basal plasma membrane to an extent which makes this method unsuitable for further membrane lipid analysis. Morphological analysis of the cells and fluorescence recovery after photobleaching experiments on the plasma membranes were performed at each step of the purification procedure and showed that only hypotonic shock and mechanical shearing of the cells enabled the basal plasma membranes to be purified without significant morphological changes. ß

The consequences of the binding of annexin V on its lateral mobility and that of lipids were investigated by means of experimental and simulated FRAP experiments. Experiments were carried out on planar supported bilayers (PC/PS 9:1 mol/mol mixtures) in the presence of 1 mM CaCl2 in the subphase. The probes C12-NBD-PS and fluorescein-labeled annexin V were used and the data compared with that previously obtained for C12-NBD-PC [Saurel, O., Cézanne, L., Milon, A., Tocanne, J. F.,Demange, P. (1998) Biochemistry 37, 1403-1410]. At complete coverage of the lipid bilayer by the protein (Cannexin = 80 nM), the lateral mobility of C12-NBD-PC was reduced by 40% while C12-NBD-PS and bound annexin V molecules were nearly immobilized (D10(-)11 cm2/s). At moderate protein concentration (20 nMCannexin80 nM), best fitting of the lipid and protein probe recoveries was achieved with one single diffusion coefficient and a mobile fraction close to 100%, indicating homogeneous lipid and protein populations. In contrast, at low protein concentration (Cannexin20 nM), C12-NBD-PS showed a two-component diffusion. The slow PS population at Cannexin20 nM and the single PS population at Cannexin20 nM moved at the same rate that bound annexin V (mobile fraction close to 100%), indicating strong PS/protein interactions. With the aid of computer simulations of the lateral motion of PC molecules, based on the 2-D crystalline networks formed by annexin V in contact with the lipid bilayer, these FRAP results may be accounted for by considering a rather simple model of a proteolipidic complex consisting of an extended 2-D crystalline protein network facing the lipid bilayer and stabilized by strong interactions between annexin V and PS molecules. In this model, immobilization of annexin V and PS molecules originates from their mutual interactions. The slowing down of PC molecules is due to various obstacles to their lateral diffusion which can be described as: the four PS molecules bound to the protein, the tryptophan 187 which presumably interacts with the lipids at the level of their polar headgroups and probably the three other hydrophobic amino acid residues located on the AB calcium-binding loops of the protein.

In this paper we show that FRAP experiments at variable beam radii provide an experimental approach for investigating membrane organization and dynamics, with great potential for identifying micrometer-sized domains and determining their size and the diffusion coefficient of the lipid and protein molecules they contain. Monte Carlo simulations of FRAP experiments at variable beam radii R on models of compartmentalized membranes have allowed us to establish the relationships (i) between the mobile fraction M of a diffusing particle and the size r of the domains, and (ii) between the apparent diffusion coefficient Dapp and the real diffusion coefficient DO of this particle inside the domains. Furthermore, in its present stage of development, this approach allows us to specify whether these domains are strictly closed or not. This approach was first validated on an experimental model of a strictly compartmentalized membrane consisting of a monolayer of apposed spherical phospholipid bilayers supported by silica beads of known radius (0.83 micron). To prevent fusion between the spherical bilayers 5 mol% of a polymer-grafted phospholipid was added to the lipids. Analysis of the M versus R data yielded a radius r of 0.92 +/- 0.09 microns for the spherical bilayers, close to that of the supporting silica beads. When applied to the experimental data available for lipids and proteins in the plasma membrane of living cells, this approach suggests the existence of domains within these membranes with a radius of about 0.4-0.7 microns for the lipids and 0.25 micron for the proteins. These domains are not strictly closed and they are believed to be delineated by fluctuating barriers which are more or less permeable to lipid and protein molecules.

The consequences of the binding of annexin V on the structure and dynamics of PC/PS bilayers were studied by means of fluorescence polarization, 31P NMR, 2H NMR, and fluorescence recovery after photobleaching (FRAP). Even at complete coverage of the lipid bilayers by the protein, annexin V showed no influence on the lipid molecular packing and the acyl chain flexibility of both PC and PS. The fluorescence polarization of the probe DPH, the 31P NMR spectra, and deuterium quadrupolar splittings of P(d31)OPS remained unchanged. However, upon binding of annexin V, two distinct populations of PC were visible in 2H NMR, which were in slow exchange on the deuterium NMR time scale (microseconds). One component in the spectrum was identical to the protein-free sample, while a second, broad, component appeared. The presence of the protein induced a decrease in the transverse relaxation times (T2e), indicative of the appearance of slow motions (milliseconds to microseconds), in the P(d31)-OPS spectrum and in the P(d31)OPC broad component. FRAP experiments were carried out with the probes C12-NBD-PC and C12-NBD-PS: at saturation, annexin V reduced the lateral diffusion rate of PC by 40% and nearly blocked the diffusion of PS. These combined experiments are consistent with a model in which annexin V enters a proteolipidic complex in the form of an extended 2D network, stabilized by specific interactions with PS. As seen from the lateral diffusion rates and the acyl chains NMR spectral parameters, two separate lipid populations appear, presumably corresponding to those interacting with annexinV (PC and PS) and protein free domains (mainly PC).

To test the hypothesis that differences in lipid bilayer and membrane protein hydrophobic thicknesses can brings about a mechanism of lipid molecular sorting by proteins, computer simulations and Fluorescence Resonance Energy Transfer experiments (between NBD-labeled lipids as donor and retinal as acceptor) were carried out on Bacteriorhodopsin reconstituted in DLPC/DSPC mixtures. This study took advantage of the non-ideal mixing behavior of DLPC and DSPC and the fact that the average lipid acyl-chain length depends on temperature. At low temperature, in the gel-gel coexistence region, BR is found associated with the short-chain lipid DLPC. At moderate temperature, in the fluid-gel coexistence region, BR still shows preference for DLPC but now stands at the fluid-gel boundary. At high temperature, in the fluid-fluid phase, the theoretical data shows preference of BR for the longchain DSPC at the expense of the short-chain DLPC molecules.

Mycobacterial glycopeptidolipid (GPL) interactions with membranes were analysed with monolayer experiment, using GPLs bearing 3, 1, or 0 carbohydrate residues (GPL3, GPL1, GPL0). Compression isotherms and surface potential determinations suggested that the glycopeptidic moiety of GPL3 permanently dipped in water, while those of GPL1 and GPL0 can lay in the interface. Insertion of GPL molecules into a preformed phospholipid monolayer was observed using GPL3 or GPL1 dispersions, but not from GPL0. It is postulated that the activity of GPL0 is low due to its failure to become inserted into membranes, as is that of GPL3 owing to its insertion only by its acyl chain. GPL1 is likely to disturb membranes by inserting its glycopeptidic moiety into the interface.

Fluorescence Recovery After Photobleaching experiments were simulated using a computer approach in which a membrane lipid leaflet was mimicked using a triangular lattice obstructed with randomly distributed immobile and non-overlapping circular obstacles. Influence of the radius r and area fraction c of these obstacles and of the radius R of the observation area on the relative diffusion coefficient D* (Eq. (1)) and mobile fraction M was analyzed. A phenomenological equation relating D* to r and c was established. Fitting this equation to the FRAP data we obtained with the probe NBD-PC embedded in bacteriorhodopsin/egg-PC multilayers suggests that this transmembrane protein rigidifies the surrounding lipid phase over a distance of about 18 A (approximately equal to two lipid layers) from the protein surface. In contrast, analysis of published diffusion constants obtained for lipids in the presence of gramicidin suggests that in terms of lateral diffusion, this relatively small polypeptide does not significantly affect the surrounding lipid phase. With respect to the mobile fraction M, and for point obstacles above the percolation threshold, an increase in R led to a decrease in M which can be associated with the existence of closed domains whose average size and diffusion properties can be determined. Adaptation of this model to the re-interpretation of the FRAP data obtained by Yechiel and Edidin (J Cell Biol (1987) 115:755-760) for the plasma membrane of human fibroblasts consistently leads to the suggestion that the lateral organization of this membrane would be of the confined type, with closed lipid domains of approximately equal to 0.5 microns 2 in area.

Mismatch between the hydrophobic thicknesses of transmembrane proteins and the supporting lipid bilayer and its consequences on the lateral organization of lipids have been investigated with bacteriorhodopsin and phosphatidylcholine species with a variety of acyl-chain lengths. The purple membrane, from the bacterium Halobacterium halobium, was used and reconstituted with dilauroyl-(Lau2GroPCho), dimyristoyl- (Myr2GroPCho), dipalmitoyl- (Pam2GroPCho) and distearoyl- (Ste2-GroPCho) glycerophosphocholine. The phase behaviour of the lipids was investigated at different temperatures and different protein/lipid molar ratios, by analyzing the fluorescence excitation spectra of the 1-acyl-2-[8-(2-anthroyl)-octanoyl]-sn-glycero-3-phosphocholine probe, and by measuring the fluorescence depolarization of the 1,6-diphenyl-1,3,5-hexatriene probe. Data obtained with 1-acyl-2-[8-(2-anthroyl)-octanoyl]-sn-glycero-3-phosphocholine shows that bacteriorhodopsin produced positive or negative shifts in the phase transition temperature of the host lipids depending on the strength and sign of the mismatch between the lipid and protein hydrophobic thicknesses and also on the protein concentration and aggregation state in the lipid bilayer. In the region of high protein concentration (bacteriorhodopsin/phosphatidylcholine molar ratios ∼ 1:50) and despite the presence of the endogenous lipids, bacteriorhodopsin (hydrophobic length dp∼ 3.0–3.1 nm) brought about a large upward shift in the phase-transition temperature of Lau2GroPCho (ΔT∼ 40 K, mean hydrophobic thickness d∼ 2.4 nm), and to a lesser extent of Myr2GroPCho (ΔT∼ 23 K, d∼ 2.8 nm), accounting for a strong rigidifying effect of the protein on these short-chain lipids. Bacteriorhodopsin had no influence on the phase properties of Pam2GroPCho (ΔT∼ 0 K, d∼ 3.2 nm), a lipid whose mean hydrophobic thickness is similar to that of the protein. In contrast, the transition temperature of Ste2GroPCho was decreased (ΔT∼−13 K, d∼ 3.7 nm), indicating a fluidifying effect of the protein on this long-chain lipid. Similar effects on the lipid acyl-chain order were observed in the region of high-protein dilution (bacteriorhodopsin/phosphatidylcholine molar ratios

When studying lipid-lipid or lipid-protein interaction in membranes, the correct interpretation of data obtained when using fluorescent phospholipid probes requires the best possible knowledge of probe behaviour in phosopholipid membranes. Analysis of the translational dynamics and photochemical properties of the anthracene-labeled phosphatidylcholine (EAPC) shows that a self-association process occurs with this probe in the membrane at the ground state. This anthracene self-association is characterized and leads to a hypochromic effect which has been studied by means of ultraviolet absorption spectroscopy in unilamellar egg-yolk phosphatidylcholine (EggPC) vesicles. A model with indefinite linear self-association, in which each step has the same equilibrium constant, best describes the data. The equilibrium constant was found to be in the 300–500 M−1 range and the complex lateral distribution pattern of EAPC in model membranes, which results from this self-association process, is characterized and seems to be mainly controlled by the amount of EAPC incorporated into the lipid bilayer.

Two methods are described enabling the plasma membrane from Chinese hamster ovary (CHO) cells to be obtained rapidly, relatively pure and with a good yield. In both cases, cells were disrupted by nitrogen cavitation in an isoosmotic buffer either at pH 5.4 or at pH 7.4. In the first approach, cells were lysed at pH 7.4 and the plasma membrane and cell organelles were isolated on a self-generated gradient of Percoll, at neutral pH. Mitochondria and endoplasmic reticulum were recovered in the denser fractions, plasma membrane fragments were found in the lighter fractions, but always contaminated by lysosomes. Because lysosomes were found to sediment in acidic conditions, cells were lysed at pH 5.4 and presedimentation (1500 x g) of the cell homogenate at the same pH enabled more than 80% of the lysosomes to be removed. Then, ultracentrifugation of the supernatant over a Percoll gradient at neutral pH yielded plasma membrane fractions practically free of lysosomes with an enrichment ratio of 3 and fractions of mitochondria and endoplasmic reticulum with enrichment ratios of 17 and 6, respectively. A major problem was encountered in the final step of elimination of Percoll from the purified plasma membrane fractions. Whatever the technique used for eliminating Percoll, plasma membranes were observed to be contaminated by a Percoll constituent which prevented further purification and biochemical identification of the lipids extracted from these membrane fractions to be carried out. A second method of plasma membrane preparation was tested consisting first in the coating of the cell surface with positive colloidal silica which was stabilized by an anionic polymer. Then, and through differential centrifugations, plasma membrane fractions were easily obtained within less than 1 h, with a yield of 65% and an enrichment ratio of 7. The coating pellicle was quantitatively removed thus enabling any biochemical manipulation of the plasma membrane to be carried out. The lipids present in the plasma membrane of CHO cells were analyzed and are described, both in terms of headgroup and acyl chain composition.

The potent antiarrhythmic drug, amiodarone (AMIO) exhibits phototoxicity, which is thought to be related to its interaction with biological membranes. We report here a spectroscopic study of the interactions of this drug with phosphatidylglycerol (PG) and phosphatidylcholine (PC) liposomes used as membrane model systems. A linear increase in absorbance at 300 nm was observed with increasing addition of AMIO to dimyristoyl- dl -PC (DMPC) liposomes over all the drugs-lipid molar ratio (Ri)s tested. In contrast, in the dimyristoyl-DL-PG (DMPG) liposomes, there was a dramatic increase in absorbance at values of Ri above unity. Light scattering by DMPG liposomes at 350 nm increased with increasing AMIO concentration up to a Ri = 1, and then decreased with increasing drug concentration. Such changes were not observed with the DMPC liposomes. Moreover, addition of AMIO changed the fluorescence polarization rate of 1,6-diphenyl 1,3,5-hexatriene embedded in these liposomes. It reduced the rate below the phase transition temperature (Tt) of the lipid, but increased it above this temperature. These effects on the lipidic phases observed at low Ri were more pronounced on the DMPG than on the DMPC liposomes. the strong interactions of AMIO with phospholipids, especially the acidic ones, were confirmed by liposome size determinations. All these data strongly suggest that the drug was incorporated in the core of the lipid bilayers. Such a penetration would favor a drug-photoinduced peroxidation of lipids. Indeed, UV irradiation of AMIO-DOPG mixtures led to the disappearance of the unsaturated fatty acids of phospholipids, checked by gas chromatography measurements, which was correlated with the amount of oxygen consumed. This showed that AMIO did photosensitize phospholipid peroxidation.

Analyses of interactions between substance P (SP) and phospholipids were performed by combined surface pressure and surface potential measurements in monolayers and by 13 C-NMR experiments on liposomes. This study was carried out using synthetic SP molecules: [1- 13 C-Gly 9 ]SP and [1- 13 C-Gly 2 ]SP. Injection of SP into the aqueous subphase led to an expansion of phosphatidylcholine (PtdCho) or phosphatidylserine (PtdSer) monolayer surface area. An apparent association constant of SP for PtdSer was estimated to be around 10 6 –10 7 M 1 . The surface potential ΔV / n varied linearly with the molecular area whereas the variation of surface pressure was biphasic, suggesting that at least two binding states contributed to the monolayer expansion. These two states Si (SP is inserted into the bilayer) and Ss (SP is stuck on the surface) were observed on vesicular membranes by 13 C-NMR. The kinetic of interconversion between these two states can be estimated by NMR, the Ss state being the stablest one. No perpendicular insertion of SP into these vesicular preparations seemed to occur, as previously postulated. However, SP might form aggregates in contact with these model systems, leading to a loss of permeability of the lipid vesicles.

Fluorescence excitation and emission spectra, relative fluorescence quantum yield φ r and fluorescence lifetime τ of methyl 8-(2-anthroyl)-octanoate have been studied in a set of organic solvents covering a large scale of polarity and in the presence of water. In this probe, the 2-anthroyl chromophore exhibits quite remarkable and unique fluorescence properties. Thus, when going from n -hexane to methanol, the maximum emission wavelength λ em max shifts from 404 nm to 492 nm while φ r and τ increase from 1 to 17.7 and from 0.91 ns to 13.5 ns, respectively. These increments are still more accentuated in the presence of water with estimated values of 526 nm for λ em max , 27 for φ r and 20 ns for τ in this solvent. Because of the presence of a keto group which is a hydrogen bond acceptor and which can conjugate with the aromatic ring so as to provide the chromophore with a high dipole moment, the fluorescence properties of the probe strongly depend on the polarity of the surrounding medium. They can be accounted for in terms of general solvent effects (dipolar solute/solvent interactions) in the presence of aprotic solvents and in terms of specific solvent effects (hydrogen bonding) in protic solvents. Such properties of solvatochromism make the 2-anthroyl chromophore, after 8-(2-anthroyl)octanoic acid has been attached to phospholipids (E. Perochon and J.F. Tocanne (1991) Chem. Phys. Lipids 58, 7–17) a potential tool for studying microenvironmental polarity in biological membranes.

It is shown that investigating the lateral motion of lipids in biological membranes can provide useful information on membrane lateral organization. After labeling membranes with extrinsic or intrinsic lipophilic fluorescent probes, fluorescence recovery after photobleaching experiments strongly suggests that specialized cells like spermatozoa, eggs and epithelia exhibit surface membrane regionalization or macrocompartmentation and that lateral microheterogeneities or lipid microdomains exist in the plasma membrane of many cellular systems.Phospholipid; Photobleaching; Lateral diffusion; Membrane organization; Macrodomain; Microdomain

The interaction with membranes of three anti-cancer drugs of the ellipticine family was studied by fluo- rescence quenching of membrane probes. The fluorescence of three probes, located at different levels in mem- branes, was quenched by addition of two types of ellipticine derivatives, one amphiphilic drug (9-methoxy- ellipticine) and two dipolar molecules (9-aminoellipticine and 9-hydroxyellipticine). By comparing the quenching curves obtained, the following can be proposed. a) 9-Methoxyellipticine can penetrate deeper in the lipid layers than 9-aminoellipticine and 9-hydroxy- ellipticine can. b) The three drugs are able to penetrate at least between the first methylene groups of the acyl chains of lipids in liposomes. c) In an isolated bacterial membrane, only 9-methoxyellipticine can be located in the region of the first methylene groups of lipids, the two dipolar drugs being adsorbed on the membrane surface. It was also shown that cholesterol hindered the penetration of 9-methoxyellipticine in the bilayer of liposomes.

Ellipticine and some of its derivatives constitute an interesting class of substances displaying an antitumoral activity against both experimental and human tumors. We have recently shown that some of these molecules were able to interact rapidly and strongly with acidic phospholipids. In the present report, amphiphilic ellipticine and 9-methoxyellipticine were tested against synthetic dilauroyl phosphatidylglycerol and dipalmitoylphosphatidylglycerol as well as a sample of phosphatidyl-glycerol with methyl-branched fatty acids, to investigate the stoichiometry of these interactions and their consequences on the organization of the lipid phase. Through surface pressure and surface potential measurements, carried out at the air-water interface, it is shown that the two drugs strongly interact with phosphatidylglycerols, whatever the film surface pressure or the nature and physical state of the lipid acyl chains may be. Such interactions bring about large film expansions and a complete modification of the phase properties of the lipids. These interactions achieve, at saturation, drug/lipid associations near the 1/1 molar ratio, stabilized by both electrostatic and hydrophobic forces. The drug molecules can be viewed as inserted between the lipid molecules, the—NH+ group interacting with the phosphate function and the aromatic moiety being in contact with the lipid acyl chains. From compression isotherms of 1/1 drug/lipid mixtures, it can be inferred that the two drugs cause a large condensing effect on both the dilauroyl and dipalmitoyl derivatives of phosphatidyl-glycerol, as compared to the behaviour of the lipids alone when ionized. In the mixed films, both lipids would exist in a very similar and rather condensed state, intermediate between the liquid-expanded and gel states in which these two lipids can respectively exist on pure water. The possibility of membranes being one of the sites of action of ellipticines is discussed.

9-(2-Anthryl)-nonanoic acid, a newly synthesized photoactivable molecule, is shown to be incorporated into the membrane lipids of the bacterium Micrococcus luteus, through the regular metabolic pathway. This incorporation, which occurs at the sn-1 position exclusively and without any degradation or elongation of the anthracene fatty acid, is accompanied by an upward shift of the chain length of the other fatty acids.

Interactions between the antibiotic polymyxin B and monolayers of dipalmitoylglycerophosphoglycerol have been reinvestigated through a study of the structure and dynamics of the complexes by means of an interface fluorimeter of our fabrication. A fluorescence technique has been developed where the use of linearly polarized incident beams gives the simultaneous determination of the orientation and the lateral diffusion rate of a fluorescent probe inserted in the film. The present investigation was carried out with 12-(9-anthroyloxy)-stearic acid, a fluorescent compound which forms non-fluorescent photodimers upon illumination. Orientation of the probe was studied by computing the ratio of the two dimerization constants KD and the ratio of the fluorescence intensities obtained with crossed linearly polarized incident lights. The lateral diffusion rate of the probe was obtained by measuring fluorescence recovery after photobleaching (photodimerization) of the probe. Control experiments, carried out with dimyristoylglycerophosphocholine, a lipid which does not interact with polymyxin B, show that the antibiotic does not significantly modify the behaviour of the probe. Both in terms of orientation and dynamics, with respect to dipalmitoylglycerophosphoglycerol, when the antibiotic is present in the subphase (1 microM, saturating conditions), data indicate that the lipid remains in a liquid-expanded state. This is true even at a high surface pressure (pi approximately equal to 37 mN X m-1), above the apparent 'transition' which can be observed at 30-35 mN X m-1 on its compression isotherm. Computation of the contribution of polymyxin B to the film expansion to the conclusion that this 'transition' would be a structural transition between two models of interaction: one, below the 'transition', where the polypeptide ring penetrates between the film-forming lipid molecules and another one, above the 'transition', were the antibiotic is adsorbed at the lipid-water interface with only its hydrocarbon chain penetrating the film.

C-6 esters of methyl alpha-D-glucoside and C-6, C-6' 'diesters of alpha, alpha'-D-trehalose with C18 and C32 threo and erythro mycolic acids (from chemical source) and of C80-erythro-mycolic acid (from natural source) have been synthesized. Esters of a C32 deoxy analogue were prepared as well. Throughout a monolayer study at the air-water interface, these glycolipids are shown to form well organized phases in which the two hydrocarbon chains of mycoloyl residues must be in interaction. Compression isotherms of C32 esters suggested a transition between liquid-expanded and liquid-condensed states. Latent heats Qc and entropy changes delta S associated with these phase transitions as well as the critical temperature at which they occur have been measured. Within the monolayer, the molecular packing of these glycolipids depends on the presence of the hydroxyl group of mycoloyl residues and on its stereochemistry. In particular intermolecular hydrogen bonds between these groups are postulated in the case of the bis(C32-erythro-mycoloyl)-trehalose. On the other hand, short chain C18 esters form fluid phases (t greater than 10 degrees C) whereas very long chain C80 mycoloyl esters of trehalose exist in a condensed state (t = 20 degrees C). These glycolipids were found to interact strongly with dipalmitoylphosphatidylcholine and egg yolk lecithins (3-sn-phosphatidylcholine). Their phase behaviours are discussed in connection with hypotheses concerning the way they can interact with mitochondrial membranes.

Fast lateral proton conduction was observed along the lipid/water interface using a fluorescence technique. This conduction can be detected for a large number of lipids, both phospholipids and glycolipids. The efficiency of the proton transfer is dependent on the molecular packing of the host lipid at a given surface pressure. The proton conduction which is present in the liquid expanded state is abolished by the transition to the liquid condensed state. The proton transfer is affected slightly by the ionic content of the aqueous subphase except in the case of calcium which can inhibit the conduction along phosphatidylglyceroethanolamine. We suggest that the transfer of the protons occurs along a bidimensional hydrogen-bond network formed from the polar head groups, their water molecules of hydration and the water molecules which are intercalated between the lipid molecules.

A fluorescent analogue of cord factor, a glycolipid toxin of mycobacteria, has been synthesized and its interactions with liposomes and isolated mitochondria have been studied. This compound, methyl alpha-D-6[12-(9-anthroyl)stearoyl]glucoside, is shown to be active against oxidative phosphorylation. When spread as a monolayer at the air-water interface, it forms a well organized phase and it strongly interacts with phosphatidylcholine. Addition of phosphatidylcholine liposomes or of isolated mitochondria to a water disperson of this fluorescent cord factor analogue results in a large increase of the fluorescence intensity. Moreover, the glycolipid probes for the temperature-dependent phase transition of the added suspensions. It is thus suggested that this cord factor analogue penetrates within mitochondrial membranes, a result which is discussed with respect to our previous conclusions concerning the way natural cord factors can interact with these organelles.

Synthesis and physical properties of a new anthracene fatty acid, 9-(2-anthryl)nonanoic acid, and the corresponding anthracene-phosphatidylcholines which were obtained by condensing the acid with sn -1-palmitoyl-lysophosphatidylcholine (PAPC) and with egg lysophosphatidylcholine (EAPC) are described. Differential scanning calorimetry experiments show that these lipids can undergo a liquid-crystal to gel phase transition at temperatures of 15°C and 18°C for EAPC and PAPC, respectively. In monolayers, PAPC exhibits a compression curve nearly superimposable to that of dipalmitoylphosphatidylcholine (DPPC), with a molecular area of 0.48 nm 2 at π = 30 mN m −1 . The data indicate that in these lipids, the anthracene group is only slightly more bulky than a normal acyl chain and that it does not significantly affect the regular phospholipid molecular packing. In ethanol solutions or when incorporated into egg phosphatidylcholine liposomes in a molar ratio of 1%, these lipids display UV absorption spectra and fluorescence emission spectra similar to those of 2-methyl anthracene. For EAPC liposomes, a broad and structureless fluorescence emission spectrum centered at around 450 nm, was recorded, suggesting the occurrence of anthracene excimers. As ascertained by UV spectrophotometry, differential scanning calorimetry, fluorescence polarization and anthracene photodimerization experiments, EAPC displays good miscibility properties with lipids in the liquid state (egg phosphatidylcholine) or in the gel state (distearoylphosphatidylcholine (DSPC)). The potential of these anthracene derivatives for studying the dynamics and the topological distribution of lipids in biomembranes is discussed.

The interaction of methyl-lysine, 1,2-diaminoethane, 1,3-diaminopropane, and 1,5-diaminopentane with films of synthetic didodecanoyl- and dihexadecanoyl-phosphatidylglycerol, spread at the air-water interface, has been studied through surface pressure, potential, and fluorescence measurements. It is shown that methyl-lysine and diamine molecules can adsorb to the phosphatidylglycerol-water interface to form a complex stabilized by both electrostatic and hydrophobic interactions, in the molar ratio: one diamine molecule/two lipid molecules. Together with recent arguments concerning the orientation of phospholipid polar heads in membranes, the interpretation of these data gives rise to a model where methyl-lysine or diamine molecules, in an extended conformation, lie parallel to the interface plane, with the two amino groups interacting with two phosphate groups and the hydrophobic moiety parallel to and interacting with the glycerol residues. From comparisons between these data and those obtained for both synthetic and naturally occurring lysylphosphatidylglycerols and for their 1 1 mixtures with phosphatidylglycerols, subjected to similar monolayer experiments, it is suggested that this model still applies to the polar head of lysylphosphatidylglycerol. In other words, this polar head is postulated in a folded conformation such that the lysyl and glycerol residues interact, parallel to each other and to the interface.

The interaction of a number of positively charged anti-tumor drugs with cardiolipin-containing model membranes has been investigated using 31P nuclear magnetic resonance (31P-NMR), differential scanning calorimetry (DSC) and monolayer techniques. It appeared that the ellipticines used (i.e., celiptium and 2-N-methylellipticinium), and also ethidium bromide, completely blocked Ca2+-induced HII phase formation in pure cardiolipin liposomes at molar ratios of drug-to-lipid of approx. 1:1. For the anthracyclines adriamycin and 4'-epi-adriamycin, a similar effect was observed, but now a 2:1 ratio was required. 31P-NMR experiments on dioleoylphosphatidylethanolamine/cardiolipin mixed liposomes indicated that the two anthracyclines, but not the other three drugs, were capable of inducing macroscopic phase separation into domains enriched in drug-cardiolipin complexes and domains enriched in the zwitterionic phospholipid species. DSC experiments on dipalmitoylphosphatidylcholine/cardiolipin mixtures led, with the exception of 2-N-methylellipticinium, to the same conclusion. Measurements of surface pressure and surface potential of cardiolipin monolayers at the air/water interface as well as conformational analysis of the various drug-cardiolipin recombinants showed that the elliptciines are deeply embedded in the acyl chain region of the bilayer, while the anthracyclines and ethidium bromide are preferentially localized in the interface. All drugs share an important electrostatic interaction with the negatively charged phosphates of cardiolipin.

Monolayers, fluorescence polarization, differential scanning calorimetry and X-ray diffraction experiments have been carried out to examine the effect of the polypeptide antibiotic polymyxin B on the phase behaviour of dipalmitoylphosphatidylglycerol (DPPG) either pure or mixed with dimyristoylphosphatidylcholine (DMPC) and dipalmitoylphosphatidylcholine (DPPC). It is shown that in both phosphatidylglycerol alone and phosphatidylglycerol/phosphatidylcholine mixtures, polymyxin B can induce either phase separation between lipid domains of various compositions or interdigitation of the acyl chains in the solid state, without segregation of the two lipids. Phase separation was observed by fluorescence and differential scanning calorimetry after addition of the antibiotic to vesicles composed of mixtures of DMPC and DPPG in conditions where polymyxin B did not saturate phosphatidylglycerol (DPPG to polymyxin B molar ratio, R i , higher than 15). Phase separation was also observed in mixed monolayers of DPPC and of the 5:1 DPPG/polymyxin B complex, at high surface pressure. Acyl chain interdigitation was observed by X-ray diffraction in both 5:1 DPPG/polymyxin B mixtures and preformed 5:5:1 DMPC/DPPG/polymyxin B mixture, in which the antibiotic saturates phosphatidylglycerol ( R i ). In both cases, raising the temperature gave rise to a complex double-peaked phase transition by differential scanning calorimetry, from the interdigitating phase to a normal L α lamellar phase. As it is known that polymyxin B does not interact with phosphatidylcholine, the data presented show that, when phosphatidylcholine and phosphatidylglycerol are mixed together, a phase perturbation such as acyl chain interdigitation, which normally affects only phosphatidylglycerol, is also felt by phosphatidylcholine.

The membrane of the Gram-positive bacterium Micrococcus luteus was shown to contain at least four proteins which can be acylated by palmitate. Experiments were carried out by incubating the cells in the presence of [14C]- or [3H]palmitic acid and by studying their distribution between membrane lipids and proteins. One of the acylated proteins of molecular weight around 38 kDa was clearly identified to be acylated by palmitate using a multi-step procedure. This included culture of the bacterium in the presence of [14C]palmitic acid, unlabeled acetate and unlabeled amino acids, followed by preparative polyacrylamide gel electrophoresis of membrane proteins after a careful delipidation step and, for given proteins, electroelution, acid hydrolysis and analysis of the fatty acids by gas chromatography. An overall analysis of the four acylated proteins extracted from bacteria cultured in the absence of exogeneous fatty acids reveals the presence, in situ, of palmitate as the main fatty acid, and of myristic, stearic and ararchidonic acids.

Ionization of the acidic phospholipid phosphatidylglycerol has been studied by measuring the surface potential of monomolecular films of the lipid as a function of the aqueous subphase pH and the concentration of monovalent cations (Li, Na, Cs). It is shown that the experimental data can be interpreted by means of the Gouy-Chapman theory in its simplest formulation, provided an adsorption of cations at the membrane surface is accounted for. This allows us to predict the ionization state of the lipid for given ionic conditions in the subphase. Above pH 4, for subphase ion concentration higher than 10 mM, or for ion concentrations above 0.1 mM at pH 5.6, phosphatidylglycerol is fully deprotonated. Within the limits of our theoretical approach, association constants of the cations to the lipid lie around 0.1-0.6 M-1.

A simple fluorescence recovery after photobleaching (FRAP) apparatus using a fluorescence microscope with a conventional mercury arc lamp, working under conditions of "uniform disk illumination" is described. This set-up was designed essentially for the use of anthracene as fluorescent probe, which is bleached (photodimerization reaction) by illumination in the near ultraviolet range (360 nm). It is shown that the lateral diffusion coefficients D can be readily calculated from fluorescence recovery curves using a finite differentiate method in combination with statistical analysis of the data. In contrast to the analytical solutions so far described, this numerical approach is particularly versatile. With a minimization algorithm, D and the probe mobile fraction can be readily calculated for any recovery time under various experimental conditions. These include different probe concentration profiles in the illuminated area after the bleaching step, and situations of infinite or noninfinite reservoir in the diffusion area outside the illuminated area.

Biological membranes are complex and well-organized multimolecular assemblies composed of a wide variety of protein and lipid molecular species. If such a diversity in protein and lipid polar headgroup structures may easily be related to a large panel of functions, the wide dispersion in acyl chain length and structure which the lipids display is more difficult to understand. It is not required for maintaining bilayer assembly and fluidity. Direct information on the lateral distribution of these various molecular species, on their potential specificity for interaction between themselves and with proteins and on the functional implications of these interactions is also still lacking. Because hydrophobic interactions play a major role in stabilizing membrane structures, we suggest considering the problem from the point of view of the matching of the hydrophobic surface of proteins by the acyl chains of the lipids. After a brief introduction to the hydrophobic matching principle, we will present experimental results which demonstrate the predictive power of the current theories and then, we will introduce the new and important concept of protein/lipid sorting in membranes. Finally, we will show how the hydrophobic matching condition may play a key role in the membrane organization and function.

Influence of the transmembrane protein bacterioopsin (the retinal-free form of bacteriorhodopsin) on the polarity of egg-phosphatidylcholine bilayers was studied by means of a steady-state and time-resolved fluorescence approach exploiting the solvatochromic properties of the 2-anthroyl fluorophore. Introduced in phosphatidylcholine molecules in the form of 8-(2-anthroyl)octanoic acid, this fluorophore probed the hydrocarbon core of the lipid bilayer. As previously shown (E. Perochon et al., Biochemistry 31 (1992) 7672–7682), water molecules were detected in this region of the terminal part of the lipid acyl chains. Their number was considerably reduced upon addition of bacterioopsin to the lipids. This was assessed by a blue shift in the fluorescence emission spectra of the probe and a marked decrease in the fractional population of fluorophores interacting with water, to the benefit of those experiencing a hydrophobic environment. In agreement with current theories, this decrease in the hydration of the bilayer may be linked to an increase in the acyl chain order and a decrease in the lateral diffusion coefficient of lipids near the protein. The data obtained at high protein concentration accounts for a protein/lipid interface which is much less hydrated than the hydrophobic core of a protein-free lipid bilayer.

The driving force in energy-transducing membranes is now recognized to be linked to a flux of protons between membrane-bound donors and acceptors. The question of whether the proton pathway is delocalized within the two bulk aqueous phases on each side of the membrane (delocalized chemiosmotic theory1), or is localized at its surface (semi-localized hypothesis2) still remains open to discussion3,4. Using an original fluorescence monolayer technique, we have recently been able to provide direct experimental evidence that a phospholipid–water interface can act as an efficient proton conductor5,6. This observation strongly supports the semi-localized hypothesis2. In the present study, comparisons between surface potential and fluorescence measurements in monolayers provide additional evidence of a facilitated proton conduction along phospholipid–water interfaces. They suggest the existence of an induced steep surface pH gradient from a more acidic surface towards the bulk. They also show that lateral proton transfer along the surface of a biological membrane would alter the surface potential of that membrane.

Movements of H+ along the polar heads of phospholipids spread in monolayers were compared to movements of H+ in the aqueous subphase. The probe for detecting H+ movement along the monolayer was a pH-sensitive fluorescein chromophore covalently bound to the head group of phosphatidylethanolamine. The behavior of this probe was not affected by the electrical properties of the lipid/water interface. Lateral diffusion of H+ along the phospholipid/water interface was then studied by acid-jump experiments in which advantage was taken of the large size of the monolayer. H+ was injected a few centimeters away from the probe observation area. The time needed for H+ diffusion to the probe was monitored by the change in the fluorescence signal, fluorescein being nonfluorescent in an acid medium. Diffusion of H+ in the bulk phase was monitored by the fluorescence change of water-soluble fluorescein isothiocyanate. Diffusion along the lipid monolayer was found to be 20 times faster than in the bulk water phase and required a structured monolayer in order to occur, as revealed by variation of the molecular area occupied by the lipid molecules. The molecular basis of rapid H+ transfer along the lipid monolayer may be the existence of a hydrogen-bond network along the polar heads, capable of supporting a rapid "hop and turn" of H+.

It is shown that photo-induced cross-linking reaction between anthracene-labelled phospholipids can be used for studying, at a molecular level, their lateral distribution in bilayer structures. A simple and versatile method is proposed. It is based on the property of anthracene to form covalently bound dimers upon irradiation in the near ultraviolet (360 nm) and on the possibility of separating the lipid photo-dimers from the lipid monomers by thin-layer chromatography. Identification of the photo-dimers is easily achieved since, upon illumination at a shorter wavelength (250–280 nm), they partially dissociate to the native monomer molecules. The feasibility of the method was tested by checking the effects of cations (sodium, calcium) on the homogeneity of 1/1 mixtures of anthracene-phosphatidylcholine, i.e. 1-acyl-2-[9-(2-anthryl)-nonanoyl]-sn-glycero-3-phosphocholine (Anthr-PC) with anthracene-phosphatidic acid (Anthr-PA) and with anthracene-phosphatidylglycerol (Anthr-PG) in the form of liposomes. These lipids were anthracene-labelled by acylation of their glycerol backbone at the sn-2 position with the synthetic 9-(2-anthryl)-nonanoic acid. Data presented indicate a good miscibility of these lipids in the presence of sodium. For each lipid mixture, the lipid heterodimers were clearly identified and, quantitatively, they dominated the lipid homodimers, as expected for a regular distribution of the lipids in the 1/1 mixture. Addition of calcium ions to the lipid suspensions did not alter the miscibility properties of Anthr-PC and Anthr-PG. In contrast, calcium triggered a clear-cut phase separation in the Anthr-PC/Anthr-PA mixture as, in this case, only traces of the heterodimer form of the lipids remained observable on the chromatogram. The three anthracene-phospholipids, pure or mixed together, exhibit a clear-cut gel-to-liquid phase transition which was detectable by fluorescence intensity measurements. The analysis of, the corresponding phase-transition temperatures confirms, at a ‘macroscopic’ level, the effects of sodium and calcium on the mixing properties of the anthracene phospholipids which were revealed at a ‘microscopic’ level by the dimerization procedure.

Bovine vascular endothelial cells can be maintained in a highly differentiated state in vitro, either by the addition of fibroblast growth factor (FGF) to the culture medium or by plating the cells on extracellular matrix (ECM)-coated dishes. Under these conditions the cells proliferate actively and at confluence form a tightly packed monolayer composed of nonoverlapping polarized cells. A fluorescence recovery after photobleaching method was used to determine the lateral mobility coefficient D of the lipophilic fluorescent probe, 5N-(hexadecanoyl)-aminofluorescein (HEDAF), in the basal and apical plasma membranes of endothelial cells under various culture conditions (cells on glass coverslips in the presence or absence of FGF, or cells plated on ECM in the exponential growth phase or at confluence). A heterogeneous distribution of lateral diffusion coefficients D was found in a given cell population. Nevertheless, for the basal membrane, a “mean” D value close to 2.0 × 10−9 cm2/s was found for all the culture conditions. The “mean” D value of HEDAF in the apical pole was slightly higher when sparse cells were exposed to FGF (D = 2.2 × 10−9cm2/s) and was further enhanced when cells were growing or confluent on ECM-coated coverslips (D = 2.7 × 10−9cm2/s). On the other hand, when the cells were maintained in the absence of FGF on glass coverslips, similar “mean” D values were found in both cell poles (D = 2.0 × 10−9cm2/s). These results show that lateral mobility of lipids in endothelial plasmalemma varies in response to external factors such as FGF and the ECM.

Through a monolayer investigation (pi, delta V), it is shown that the cationic antibiotic polymyxin B (or E) strongly interacts with films of acidic lipids, namely the didodecanoyl- and dihexadecanoylphosphatidylglycerol. The zwitterionic dihexadecanoylphosphatidylcholine was an unsuitable substrate. Interactions occurred at and above a polymyxin B concentration in the subphase of 2.5 . 10(-7) M, bringing about a considerable increase of both pi and delta V. These interactions proceeded in two steps, as revealed by a biphasic change of delta V with time. They were independent of the film molecular packing (fluid or gel states) and of the initial film pressure. Since it was possible to monitor the relative number of polymyxin B and didodecanoyl- or dihexadecanoylphosphatidylglycerol molecules in the monolayer, it is demonstrated that, at saturation, one polymyxin B molecule is bound to five phosphatidylglycerol molecules, a result which accounts for an exact neutralization of the charges. From competition experiments, it is shown that Na+ is ineffective in removing polymyxin B from the interface. Ca2+ appeared to be a stronger competitor but no complete antibiotic desorption was observed even at a Ca2+ concentration of 100 mM. As a working hypothesis, the antibiotic/lipid (1/5) system was assumed to constitute by itself one molecular species. The mixing of the polymyxin B/didodecanoylphosphatidylglycerol (1/5) system with an excess of lipid molecules in the monolayer was found to be ideal both in terms of pi and delta V. With dihexadecanoylphosphatidylglycerol, a small condensing effect could be detected only at intermediate surface pressures, in a region where the lipid phase transition occurred. The molecular area of polymyxin B interacting with didodecanoylphosphatidylglycerol can be calculated to be 1.23 +/- 0.05 nm2. It is proposed that the whole antibiotic molecule penetrates the film, the five bound lipid molecules being distributed aroung the peptide structure, at given positions imposed by the five 2,4-diaminobutyric acid residues.

Using a fluorescence method applied to phospholipids spread at the air-water interface, we showed the occurrence of a localized proton pathway along the phospholipid polar heads. From the systematic investigation of this phenomenon, we conclude that the protons diffuse through a hydrogen bond network which is a matrix of hydrated polar head groups.