Effect of glycophorin incorporation on the physico-chemical properties of phospholipid bilayers

Differential Scanning Calorimetry of Protein-Lipid Interactions

Differential scanning calorimetry (DSC) is a highly sensitive nonperturbing technique used for studying the thermodynamic properties of thermally induced transitions. Since these properties might be affected by ligand binding, DSC is particularly useful for the characterization of protein interactions with biomimetic membranes. The advantages of this technique over other methods consist in the direct measurement of intrinsic thermal properties of the samples, requiring no chemical modifications or extrinsic probes. This chapter describes the basic theory of DSC and provides the reader with an understanding of the capabilities of DSC instrumentation and the type of information that can be achieved from DSC studies of lipid-protein interactions. In particular, the chapter provides a detailed analysis of DSC data to assess the effects of proteins on biomimetic membranes.

Differential scanning calorimetry of protein-lipid interactions

Differential scanning calorimetry (DSC) is a highly sensitive non-perturbing technique for measuring the thermodynamic properties of thermally induced transitions. This technique is particularly useful for the characterization of lipid/protein interactions. This chapter presents an introduction to DSC instrumentation, basic theory, and methods and describes DSC applications for characterizing protein effects on model lipid membranes. Examples of the use of DSC for the evaluation of protein effects on modulation of membrane domains and membrane stability are given.

Lipid-protein interaction in the glycophorin-dipalmitoylphosphatidylcholine system: Raman spectroscopic investigation

Raman spectra have been recorded as a function of temperature for lipid-protein complexes of glycophorin isolated from erythrocyte membranes reconstituted with dipalmitoylphosphatidylcholine (DPPC) and its chain perdeuterated analogue ([(2)H(62)]DPPC). The conformation of the phospholipid hydrocarbon chains in the vicinity of protein is drastically altered from that in pure lipid dispersions. Analysis of the chain C-(2)H stretching vibrations for complexes of [(2)H(62)]-DPPC-glycophorin shows that at lipid:protein mole ratios of 125:1, a broad melting event occurs that is not observable by calorimetric techniques. The midpoint occurs at temperatures about 15 degrees C below that of the gel/liquid crystal phase transition for [(2)H(62)]DPPC in multilamellar dispersions. The same number of gauche rotamers form in the phospholipid hydrocarbon chains during the melting process as in the phase transition of the unperturbed molecule. Analysis of the C-H stretching region of the Raman spectrum in DPPC-glycophorin complexes indicates that lateral interactions between phospholipid chains in the complex are reduced so that interchain vibrational coupling is minimized. The observed differences between the Raman melting curves and the calorimetric endothermic transitions arise because different populations of phospholipid molecules are sampled in the two experiments. The advantages of Raman spectroscopy for the study of lipid-protein interaction are demonstrated in the current work. Implications for the structure of the lipid in the immediate vicinity of membrane protein are discussed.

Liposomes for gene transfer and expression in vivo

A recombinant plasmid encoding rat preproinsulin I was encapsulated in large liposomes and injected intravenously into rats. Glycaemia and blood, splenic and hepatic insulin were assayed from 6 h after inoculation. Control animals received (i) empty liposomes, (ii) liposomes carrying the Escherichia coli pBR 322 plasmid, (iii) the free rat insulin I gene, or (iv) no injection. All controls showed unchanged glucose and insulin levels. Six hours after inoculation the treated rats had 72 +/- 5 mg glucose/100 ml of blood, compared with 107 +/- 2 mg/ml for controls. Radioimmunoassay of blood insulin gave 61 +/- 8 microunits/ml (43 +/- 5 microunits/ml for controls). Spleen and liver values were 242 +/- 22 and 204 +/- 20 microunits/g of tissue, respectively (112 +/- 20 and 87 +/- 15 microunits/g in controls). The kinetics and extent of uptake of liposomes by spleen and liver were studied by external gamma-camera imaging after injection of 111In-labelled liposomes. The results paralleled insulin synthesis in the two organs. The insulin gene was localized in liver cells after injection of liposomes containing the plasmid encoding the gene. Livers were processed 4 h after inoculation for isolation of hepatocytes, Kupffer cells and endothelial cells. DNA was purified and exogenous DNA detected by Southern blotting. Kupffer cells were the primary target for gene incorporation with liposomes consisting of phospholipids and cholesterol. Targeting of liposomes to other liver cells was attempted by including lactosylceramide in the liposomes. This increased the amount of the exogenous gene in hepatocytes and particularly in endothelial cells. The efficiency of liposome-mediated gene transfer in vivo is high, since a few per cent of the transferred DNA is taken up by the liver cells and detected.

Differential hydrolysis of erythrocyte and mitochondrial membrane phospholipids by two phospholipase A2 isoenzymes (NK-PLA2-I and NK-PLA2-II) from the venom of the Indian monocled cobra Naja kaouthia

We previously demonstrated that venom from the Indian monocled cobra Naja kaouthia is a rich source of phospholipase A2 enzymes, and we purified and characterized a major PLA2 isoenzyme (NK-PLA2-I) from N. kaouthia venom. In the present study, we report the purification and biochemical characterization of a second PLA2 isoenzyme (NK-PLA2-II) from the same venom. A comparison of the membrane phospholipid hydrolysis patterns by these two PLA2s has revealed that they cause significantly more damage to mitochondrial membranes (NK-PLA2-I > NK-PLA2-II) as compared to erythrocyte membranes due to more efficient binding of the enzymes to mitochondrial membranes. Fatty acid release patterns by these PLA2s from the membrane phospholipid PC-pools indicate that NK-PLA2-I does not discriminate between saturated and unsaturated fatty acids whereas NK-PLA2-II shows a preference for unsaturated fatty acids during the initial phase of attack. The current investigation provides new insight into the molecular arrangement of NK-PLA2-sensitive domains in erythrocyte and mitochondrial membranes and highlights the contribution of polar, but uncharged, amino acids such as serine and cysteine in NK-PLA2 induced membrane damage.

Reconstitution of the glycosylphosphatidylinositol-anchored protein Thy-1: interaction with membrane phospholipids and galactosylceramide

Glycosylphosphatidylinositol (GPI)-anchored membrane proteins are proposed to interact preferentially with glycosphingolipids and cholesterol to form microdomains, which may play an important role in apical targeting and signal transduction. The objective of the present study was to investigate the interaction of the GPI-anchored protein Thy-1 with phospholipids and a glycosphingolipid. Purified Thy-1 was reconstituted into lipid bilayer vesicles of dimyristoyl-phosphatidylcholine (DMPC) alone or in combination with galactosylceramide (GC). The ability of Thy-1 to perturb the gel to a liquid-crystalline phase transition of DMPC was examined by differential scanning calorimetry. As the mole fraction of Thy-1 increased, the phase transition enthalpy, deltaH, declined. Analysis indicated that each molecule of Thy-1 perturbed over 50 phospholipids, suggesting that, in addition to the anchor insertion into the bilayer, the protein itself may interact with the membrane surface. Inclusion of 5% w/w GC in the bilayer resulted in a striking change in the interaction of Thy-1 with phospholipids. At low Thy-1 content, there was a reduction in the phase transition temperature and an increase in phospholipid cooperativity, suggesting the formation of Thy-1/GC-enriched domains. deltaH initially decreased with increasing Thy-1 content of the bilayer; however, at higher Thy-1 mole ratios, deltaH rose again. These results are interpreted in terms of a model whereby, at low protein:lipid mole ratios, Thy-1 preferentially sequesters GC to form enriched microdomains. At high protein:lipid mole ratios, Thy-1 may alter its conformation in response to steric crowding within these domains such that its interaction with the bilayer surface is reduced.Key words: glycosylphosphatidylinositol anchor, Thy-1 antigen, reconstitution, lipid bilayer, glycosphingolipid, differential scanning calorimetry, dynamic light scattering.

Insertion of Escherichia coli alpha-haemolysin in lipid bilayers as a non-transmembrane integral protein: prediction and experiment

alpha-Haemolysin is an extracellular protein toxin (approximately 107 kDa) secreted by Escherichia coli that acts at the level of the plasma membranes of target eukaryotic cells. The nature of the toxin interaction with the membrane is not known at present, although it has been established that receptor-mediated binding is not essential. In this work, we have studied the perturbation produced by purified alpha-haemolysin on pure phosphatidylcholine bilayers in the form of large unilamellar vesicles, under conditions in which the toxin has been shown to induce vesicle leakage. The bilayer systems containing bound protein have been examined by differential scanning calorimetry, fluorescence spectroscopy, differential solubilization by Triton X-114, and freeze-fracture electron microscopy. All the data concur in indicating that alpha-haemolysin, under conditions leading to cell lysis, becomes inserted in the target membrane in the way of intrinsic or integral proteins. In addition, the experimental results support the idea that inserted alpha-haemolysin occupies only one of the membrane phospholipid monolayers, i.e. it is not a transmembrane protein. The experimental data are complemented by structure prediction studies according to which as many as ten amphipathic alpha-helices, appropriate for protein-lipid interaction, but no hydrophobic transmembrane helices are predicted in alpha-haemolysin. These observations and predictions have important consequences for the mechanism of cell lysis by alpha-haemolysin; in particular, a non-transmembrane arrangement of the toxin in the target membrane is not compatible with the concept of alpha-haemolysin as a pore-forming toxin.

The differential miscibility of lipids as the basis for the formation of functional membrane rafts

The formation of sphingolipid-cholesterol microdomains in cellular membranes has been proposed to function in sorting and transport of lipids and proteins as well as in signal transduction. An increasing number of cell biological and biochemical studies now supports this concept. Here we discuss the structural properties of lipids in a cell biological context. The sphingolipid-cholesterol microdomains or rafts are described as dispersed liquid ordered phase domains. These domains are dynamic assemblies to which specific proteins are selectively sequestered while others are excluded. The proteins associated to rafts can act as organizers and can modulate raft size and function.

A biophysical study of integral membrane protein folding

In order to characterize the thermodynamic constraints on the process of integral membrane protein folding and assembly, we have conducted a biophysical dissection of the structure of bacteriorhodopsin (BR), a prototypical alpha-helical integral membrane protein. Seven polypeptides were synthesized, corresponding to each of the seven transmembrane alpha-helices in BR, and the structure of each individual polypeptide was characterized in reconstituted phospholipid vesicles. Five of the seven polypeptides form stable transmembrane alpha-helices in isolation from the remainder of the tertiary structure of BR. However, using our reconstitution protocols, the polypeptide corresponding to the F helix in BR does not form any stable secondary structure in reconstituted vesicles, and the polypeptide corresponding to the G helix forms a hyperstable beta-sheet structure with its strands oriented perpendicular to the plane of the membrane. [The polypeptide corresponding to the C helix spontaneously equilibrates in a pH-dependent manner between a transmembrane alpha-helical conformation, a peripherally bound nonhelical conformation, and a fully water soluble conformation; the conformational properties of this polypeptide are the subject of the accompanying paper: Hunt et al. (1997) Biochemistry 36, 15177-15192.] Our observations suggest that the folding of alpha-helical integral membrane proteins may proceed spontaneously. However, the preference for a non-native conformation exhibited by two of the polypeptides suggests that the formation of some transmembrane substructures could require external constraints such as the links between the helices, interactions with the rest of the protein, or the involvement of cellular chaperones or translocases. Our results also suggest a strategy for improving the thermodynamic stability of alpha-helical integral membrane proteins, a goal that could facilitate attempts to overexpress and/or refold them.

Protein location in liposomes, a drug carrier: a prediction by differential scanning calorimetry

Location of protein drugs in lipid carriers often determines the stability, loading efficiency, and release rate of these drugs from the carriers following administration. On the basis of conventional differential scanning calorimetry (DSC) measurements, Papahadjopoulos et al. (Biochim. Biphys. Acta 1975, 401, 317-335) proposed that proteins can be classified into three categories depending on their effects on the thermotropic behavior of the lipids, e.g., transition temperature and enthalpy. Interactions are usually electrostatic, hydrophobic, or their combination. The nature of these interactions are reflected by changes in various thermotropic parameters. Our study aims to test the validity of Papahadjopoulos' classification. Hydrophilic ribonuclease A, cytochrome c, and superoxide dismutase (SOD), as well as hydrophobic cyclosporin A, are used as model proteins. Neutral lipids, e.g., dipalmitoylphosphatidylcholine, and/or negatively charged lipids, e.g., dipalmitoylphosphatidylglycerol (DPPG), are used to prepare liposomes. Results from conventional and high-sensitivity DSC are compared. High-sensitivity DSC gives significant, more reproducible results. We find that the classification of Papahadjopoulos et al. needs to be modified. No hydrophilic proteins bind to liposomes exclusively on the surface by electrostatic interactions, and some degree of penetration is observed in most cases. An unexpected binding between SOD and DPPG liposomes is observed. The binding of SOD to negatively charged lipids may account, at least in part, for its ability to protect lipid membranes against oxygen-mediated injury.

Calorimetry of tetraether lipids from Thermoplasma acidophilum: incorporation of alamethicin, melittin, valinomycin, and nonactin

The development and application of model membrane systems on the basis of tetraether lipids from Thermoplasma acidophilum has been proposed. In this respect incorporation of membrane proteins and ionophores is indispensable and is demonstrated in the case of alamethicin, melittin, nonactin, and valinomycin by calorimetry. Dipalmitoylphosphatidylcholine (DPPC) and dihexadecylmaltosylglycerol (DHMG) were chosen for comparison. Melittin and alamethicin prove to broaden the lipid phase transition and to reduce the melting temperature Tm and enthalpy change (delta H) of the main phospholipid from T. acidophilum (MPL) and DPPC. The decrease in Tm, however, is more pronounced in DPPC than in MPL. Valinomycin shows only a marginal effect on the temperature and width of the transition; delta H is reduced in MPL and remains constant in DPPC and DHMG. With nonactin the phase transition of DPPC is quenched, and delta H and the half-height width are increased. DHMG is affected to a lesser extent and MPL only marginally. The four ionophores exhibit different modulation of the phase transition behavior of the various lipids as expected from their varying molecular structures. Thus, the integral membrane protein alamethicin, the peripheral protein melittin, valinomycin, and nonactin interact primarily with lipid head groups and are readily incorporated into the tetraether lipid structures.

Incorporation of synthetic peptide helices in membranes of tetraether lipids from Thermoplasma acidophilum. A calorimetric study

Four analogues of the membrane-modifying, alpha-helical polypeptide antibiotic alamethicin were synthesized. the alpha-helical deca-, undeca-, heptadeca-, and icosapeptides were mixed with the main tetraether lipid of the Archaebacterium Thermoplasma acidophilum (MPL), dipalmitoylphosphatidylcholine (DPPC) and dihexadecylmaltosylglycerol (DHMG) in various ratios and the modification of the lipid phase transition was determined by differential thermal analysis (DTA). The polypeptides form mixed phases with MPL and DPPC, however, not with DHMG. Heptadeca- and icosapeptide exert a much stronger reduction of enthalpy (delta H) than deca- and undecapeptide and bind about 0.5 molecule of MPL (or one molecule of DPPC) per peptide molecule. delta H of the DPPC pretransition is reduced by the deca- and the undecapeptides and completely disappears with heptadeca- and icosapeptides (at 0.2 mole of peptide/mole of lipid). The modulation of the melting point Tm by the incorporation of peptides is more pronounced with MPL than with DPPC, the heptadecapeptide exhibiting the strongest reduction (with MPL) and the strongest broadening of the transition peak (with DPPC). Helix length, amphiphilicity and charge of the polypeptides can be correlated with the observed modifications of the lipid phase transitions.

The effect of surface-bound protein on the permeability of proteoliposomes

Proteoliposomes have been prepared from mixtures of dipalmitoylphosphatidylcholine and phosphatidylinositol by sonication (SUV) and reverse phase evaporation (REV) and conjugated with succinyl concanavalin A (sConA). The proteoliposomes were characterised in terms of size and composition and covered a range of size (weight-average diameter) from approx. 80 to 300 nm and surface-bound sConA (weight-average number of protein molecules per liposome) from approx. 200 to 1800. The permeabilities of the proteoliposomes to encapsulated D-glucose have been measured and found to increase linearly with protein conjugation. The D-glucose permeability also increases with temperature and passes through a maximum in the region of the gel to liquid-crystalline phase transition temperature. Conjugation has no effect on the chain-melting temperature but slightly decreases the enthalpy of the transition consistent with the withdrawal of some phospholipid participation in chain-melting. The D-glucose permeabilities and thermotropic properties of the proteoliposomes are discussed in terms of the dislocation of the bilayer by the possible off-axis motion of the lipid which anchors the protein to the liposomal surface.

Deuterium nuclear magnetic resonance studies on the interaction of glycophorin with 1,2-dimyristoylamido-1,2-deoxyphosphatidylcholine

Membrane dynamics of dimyristoylphosphatidylcholine (DMPC) lipid bilayer which contains glycophorin with an artificial boundary lipid, 1,2-dimyristoylamido-1,2-deoxyphosphatidylcholine (DDPC), was investigated by 2H-NMR technique. For this purpose, both DMPC and DDPC were deuterated at the position of the 8th carbon atom of their acyl chains. Comparing, with DMPC bilayers, DDPC bilayers showed larger deuterium quadrupole splitting (delta nu rho) by approx. 2 kHz. This was explicable in terms of the stabilization of the membrane due to the formation of a strong hydrogen bonding in bilayers. Addition of glycophorin to the DMPC or DDPC single bilayers caused an increase in the delta nu rho value. The delta nu rho value of DMPC/DDPC mixed lipid bilayer was smaller than that of each single lipid bilayer DDPC in the DDPC/DMPC mixed bilayer was not phase-separated but homogeneously distributed. In glycophorin-reconstituted DMPC-d4/DDPC mixed bilayers, the delta nu rho of DMPC-d4 was almost identical to that of the simple DMPC-d4 bilayer. On the other hand, the delta nu rho of DDPC-d4 in the DMPC/DDPC-d4 mixed bilayer increased significantly upon the reconstitution of glycophorin. Judging from these data, we concluded that, in the DDPC/DMPC mixed bilayer which contains glycophorin, DMPC simply behaves as the matrix lipid, while DDPC surrounds glycophorin and certainly plays a role of the boundary lipid.

Low-pH association of proteins with the membranes of intact red blood cells. II. Studies of the mechanism

The low-pH interaction of proteins with erythrocyte membranes has been found to be correlated with pH-induced changes in the erythrocyte membrane. Using a 90 degree lightscattering method it was shown that red blood cell hemolysis was slow between pH 5.8 and 5 (t1/2 above 1 h) but became fast at and below pH 4.7 (t1/2 less than 20 min). At pH 4.7, the presence of glycophorin in the incubation medium inhibited the hemolysis of erythrocytes and this protective effect was found to be dependent on the glycophorin concentration. Electron microscope experiments showed the presence of membrane defects after 10 s incubation at pH 4.6 in the absence of glycophorin in the incubation medium. These defects could further develop into openings with average widths of 14 nm after 1.5 min incubation under the acidic conditions. Fluorescence and flow cytometry studies showed that at pH 4.7, but not at pH 7.4, glycophorin tightly associates with phosphatidylcholine liposomes, and that liposome associated glycophorin molecules are recognized by anti-glycophorin monoclonal antibodies.

Protein and lipid structural transitions in cytochrome c oxidase-dimyristoylphosphatidylcholine reconstitutions

The thermotropic behavior of the mitochondrial enzyme cytochrome c oxidase (EC 1.9.3.1) reconstituted in dimyristoylphosphatidylcholine (DMPC) vesicles has been studied by using high-sensitivity differential scanning calorimetry and fluorescence spectroscopy. The incorporation of cytochrome c oxidase into the phospholipid bilayer perturbs the thermodynamic parameters associated with the lipid phase transition in a manner analogous to other integral membrane proteins: it reduces the enthalpy change, lowers the transition temperature, and reduces the cooperative behavior of the phospholipid molecules. Analysis of the dependence of the enthalpy change on the protein:lipid molar ratio indicates that cytochrome c oxidase prevents 99 +/- 5 lipid molecules from participating in the main gel-liquid-crystalline transition. These phospholipid molecules presumably remain in the same physical state below and above the transition temperature of the bulk lipid, thus providing a more or less constant microenvironment to the protein molecule. The effect of the phospholipid bilayer matrix on the thermodynamic stability of the cytochrome c oxidase complex was examined by high-sensitivity differential scanning calorimetry. Detergent (Tween 80)-solubilized cytochrome c oxidase undergoes a complex, irreversible thermal denaturation process centered at 56 degrees C and characterized by an enthalpy change of 550 +/- 50 kcal/mol of enzyme complex. Reconstitution of the cytochrome c oxidase complex into DMPC vesicles shifts the transition temperature upward to 63 degrees C, indicating that the phospholipid bilayer moiety stabilizes the native conformation of the enzyme. The lipid bilayer environment contributes approximately 10 kcal/mol to the free energy of stabilization of the enzyme complex. The thermal unfolding of cytochrome c oxidase is not a two-state process.(ABSTRACT TRUNCATED AT 250 WORDS)

Role of peptide structure in lipid-peptide interactions: high-sensitivity differential scanning calorimetry and electron spin resonance studies of the structural properties of dimyristoylphosphatidylcholine membranes interacting with pentagastrin-related pentapeptides

The effects of amino acid substitutions in the pentapeptide pentagastrin on the nature of its interactions with dimyristoylphosphatidylcholine (DMPC) are assessed by differential scanning calorimetry and electron spin resonance. In two peptide analogues, the Asp at position 4 in pentagastrin (N-t-Boc-beta-Ala-Trp-Met-Asp-Phe-NH2) is replaced by Gly or Phe. These uncharged, more hydrophobic peptides have little effect on the transition temperature of DMPC, but they broaden the transition and lower the transition enthalpy as do integral membrane proteins. These peptides also mimic the behavior of integral membrane proteins in decreasing the order of a 5-doxylstearic acid spin probe below the transition temperature and in exhibiting a second immobilized lipid component using a 16-doxylstearic acid spin probe in DMPC. Three charged peptides were studied: pentagastrin, an analogue with positions 4 and 5 reversed (i.e., ending in Phe-Asp-NH2), and one with Asp replaced by Arg at position 4. All three of these charged peptides altered the phase transition behavior of DMPC to give two components, one above and one below the transition temperature of the pure lipid. With increasing peptide concentration, the higher melting transition became more prominent. The arginine-containing peptide produced the largest shifts in melting temperature followed by pentagastrin and then the "reversed" peptide. The arginine-containing peptide also increased the enthalpy of the transition. These peptides also increased the ordering of DMPC below the phase transition as measured with both 5- and 16-doxylstearic acid. The ordering effect was most pronounced with the arginine-containing peptide using the 5-doxylstearic acid probe. The results demonstrate that even the zwitterionic DMPC can interact more strongly with positively charged peptides than with negatively charged ones. In addition, peptide sequence as well as composition is important in determining the nature of peptide-lipid interactions. The markedly different effects of these pentagastrin peptides on the phase transition and motional properties of DMPC occur despite the similar depth of burial of these peptides with DMPC.

The influence of lipid composition and lectin-glycophorin interaction on the rotational diffusion of glycophorin in vesicles, as measured by time-resolved phosphorescence depolarization

The rotational mobility of glycophorin in various lipid vesicles was studied, using time-resolved measurements of the depolarization of laser flash excited phosphorescence of glycophorin labelled with the triplet probe erythrosin. With the exception of dimyristoylphosphatidylcholine at the phase transition no phosphorescence depolarization decays were observed in the 1-300 microseconds time interval following the laser flash. Instead, a constant anisotropy level was observed, with two distinct values depending on the experimental system. In liquid-crystalline bilayers of dioleoylphosphatidylcholine, bovine brain phosphatidylserine and dimyristolyphosphatidylcholine, the anisotropy was 0.01. This was increased to 0.03 upon addition of wheat germ agglutinin which aggregates glycophorin. In the case of gel state dimyristoylphosphatidylcholine and liquid-crystalline dioleoylphosphatidylethanolamine the anisotropy also amounted to 0.03. Experiments with glycerol to vary the viscosity of the medium, and theoretical considerations, exclude the possibility that these different anisotropy levels are related to differences in motional properties of the entire protein/lipid vesicles. These results strongly suggest that the anisotropy level of 0.03 corresponds to slowly rotating glycophorin (rotational relaxation time greater than 0.3 ms) while the anisotropy level of 0.01 corresponds to fast rotating glycophorin (rotational correlation time less than 1 microseconds). The difference in glycophorin mobility is discussed in terms of aggregation state of the protein, lipid composition of the vesicle bilayer and membrane viscosity. The observed differences in rotational mobility of glycophorin in glycophorin/dioleoylphosphatidylcholine vesicles, glycophorin/bovine heart phosphatidylserine vesicles as compared to glycophorin/dioleoylphosphatidylethanolamine vesicles are not in quantitative agreement with the relative size of the intramembrane particles in these systems as revealed by freeze-fracture electron microscopy.

Phase behavior of membranes reconstituted from dipentadecanoylphosphatidylcholine and the Mg2+-dependent, Ca2+-stimulated adenosinetriphosphatase of sarcoplasmic reticulum: evidence for a disrupted lipid domain surrounding protein

A new method was used for reconstituting active sodium deoxycholate solubilized Ca2+-ATPase of rabbit skeletal muscle sarcoplasmic reticulum. Removal of the detergent by dialysis at the pretransition temperature of the pure lipid (22 degrees C) favored the formation of sheet-like structures with a lipid and protein content close to that of the detergent-solubilized sample. Freeze-fracture electron micrographs revealed the Ca2+-ATPase to be organized in rows corresponding to the typical banded pattern seen in low-temperature freeze-fracture micrographs of pure lipid bilayers. Incubation of the sheetlike structures at a temperature (38 degrees C) above the pure lipid main phase transition (33.5 degrees C) caused closure of the sheets into vesicles displaying homogeneous intramembranous particle distributions, at least for membranes containing less than 150 lipids per Ca2+-ATPase. However, in membranes of higher lipid content, free lipid patches were seen both above and below the lipid phase transition. By use of high-sensitivity differential scanning calorimetry, three classes of excess heat capacity peaks were observed in the vesiculated samples. A broadened "free lipid" peak occurred for samples containing between 550 and 200 lipids per protein (Tm = 33.5 degrees C, as for the order-disorder transition in pure lipid vesicles). Between 200 and 150 lipids per Ca2+-ATPase, a broad shoulder became apparent in the range of 29-32 degrees C. Below 150 lipids per Ca2+-ATPase, a peak at 26-28 degrees C became increasingly prominent with lower lipid content. At a lipid to protein ratio of about 30, no peaks in heat capacity were observed. The temperature dependence of diphenylhexatriene fluorescence anisotropy revealed a similar pattern of membrane phase behavior, except that a phase transition was detected at 33.5 degrees C in all membranes studied. On the basis of these observations, we propose that the Ca2+-ATPase is surrounded by a "lipid annulus" of motionally inhibited lipid molecules that do not contribute to a calorimetrically detectable phase transition. Beyond the annulus, "secondary domains" of disrupted lipid packing account for the peak at 26-28 degrees C and the 29-32 degrees C shoulders. At high lipid to protein ratios, the secondary domains coexist with protein-free, lipid-bilayer patches, which account for the peak at 33.5 degrees C.

Mattress model of lipid-protein interactions in membranes

A thermodynamic model is proposed for describing phase diagrams of mixtures of lipid bilayers and amphiphilic proteins or polypeptides in water solution. The basic geometrical variables of the model are the thickness of the hydrophobic region of the lipid bilayer and the length of the hydrophobic region of the proteins. The model incorporates the elastic properties of the lipid bilayer and the proteins, as well as indirect and direct lipid-protein interactions expressed in terms of the geometrical variables. The concept of mismatch of the hydrophobic regions of the lipids and proteins is an important ingredient of the model. The general phase behavior is calculated using simple real solution theory. The phase behavior turns out to be quite rich and is used to discuss previous experiments on planar aggregations of proteins in phospholipid bilayers and to propose a systematic study of synthetic amphiphilic polypeptides in bilayers of different thicknesses. The model is used to interpret the influence of the lipid-protein interaction on calorimetric measurements and on local orientational order as determined by deuterium nuclear magnetic resonance.

Lipid-protein interactions of the human erythrocyte concanavalin A receptor in phospholipid bilayers

The interaction of the human erythrocyte concanavalin A receptor (a subpopulation of Band 3) with phospholipids has been investigated using differential scanning microcalorimetry of reconstituted vesicles prepared by detergent dialysis. The mean diameter of dialyzed phospholipid vesicles jumps dramatically on inclusion of the concanavalin A receptor and then increases linearly with the fraction of protein in the bilayer. The glycoprotein has a dramatic effect on the phospholipid gel to liquid-crystalline phase transition, and delta H decreases linearly with increasing mole fraction of protein up to a protein/lipid mole ratio of around 1:1160. Extrapolation of this data indicates that each concanavalin A receptor is able to perturb about 685 molecules of dimyristoylphosphatidylcholine, withdrawing them from the main phase transition. The cooperativity of phospholipid melting is profoundly disrupted by small amounts of glycoprotein, with the cooperative unit dropping to less than half its initial values at a protein/lipid mole ratio of 1:3800. A break occurs in the delta H curve as the protein/lipid mole ratio is increased above 1:1160, and delta H then increases linearly with increasing amounts of concanavalin A receptor in the bilayer. This phenomenon may be interpreted in terms of protein-protein aggregation which occurs in the phospholipid bilayer above a certain critical mole fraction of concanavalin A receptor, resulting in perturbed phospholipids being returned to the phase transition. In addition, the hydrophilic domains of the glycoprotein may exist in two different conformations depending on the protein concentration in the bilayer, and these may differ in their ability to interact with phospholipid headgroups at the membrane surface.

The influence of lipid composition on glycophorin-induced bilayer permeability

Glycophorin was incorporated into large unilamellar vesicles and the bilayer permeability was measured as a function of the lipid composition. In agreement with previous data (Van der Steen, A.T.M., De Kruijff, B. and De Gier, J. (1982) Biochim. Biophys. Acta 691, 13-23) it was found that glycophorin greatly increased the bilayer permeability of DOPC vesicles. This effect was observed for a large variety of phosphatidylcholines, differing in their fatty acid composition and homogeneity. In sharp contrast, it was observed that variations in the polar headgroups by incorporation of DOPE, DOPS and, to a lesser extent, cholesterol, into the DOPC/glycophorin vesicles restored the barrier function. These results are compared to the size of the particles, revealed by freeze-fracture electron microscopy on the glycophorin-containing bilayer and are discussed in the light of various types of lipid-protein interactions and protein aggregation state.

Interaction of glycophorin with phosphatidylserine: a Fourier transform infrared investigation

Glycophorin, from the human erythrocyte membrane, has been isolated in pure form and reconstituted into unilamellar vesicles with bovine brain phosphatidylserine (PS). Fourier transform infrared spectroscopy has been used to monitor the protein conformation as well as the effect of protein on lipid order and melting. Glycophorin, at levels of 1 mol %, nearly abolishes the gel to liquid-crystal phase transition seen in pure PS vesicles between 8 and 16 degrees C by inducing significant disorder into the lipid gel phase. A transition of reduced magnitude remains between 14 and 22 degrees C in the lipid/protein complexes. Evidence is presented for specific interaction of glycophorin with the interfacial region of PS. In general, the effects on lipid melting produced by protein at the 1 mol % level are more pronounced than those noted in a previous study of glycophorin/phosphatidylcholine interactions [ Mendelsohn , R., Dluhy , R. A., Taraschi , T., Cameron, D., & Mantsch , H.H. (1981) Biochemistry 20, 6699-6706]. Two bands are observed for the protein amide I (C = O stretching) mode. A main feature at 1653 cm-1 indicates that the bulk of the secondary structure is random coil or alpha-helical. A weaker shoulder at 1675 cm-1 suggests the occurrence of a small proportion of the beta-sheet form. The results confirm circular dichroism studies of Schulte & Marchesi (1979) [ Schulte , T.H., & Marchesi , V.T. (1979) Biochemistry 18, 275-280]. Fourier transform infrared (FT-IR) studies of a ternary complex of PS/dipalmitoyl-phosphatidylcholine- d62 (DPPC- d62 )/glycophorin indicate that the glycophorin preferentially interacts with the PS component.(ABSTRACT TRUNCATED AT 250 WORDS)

Calorimetric studies of cytochrome oxidase-phospholipid interactions

Thermotropic phase transitions in phospholipid vesicles reconstituted with mitochondrial cytochrome oxidase (EC 1.9.3.1) were studied using differential scanning calorimetry. Both dimyristoylphosphatidylcholine (DMPC) and mixtures of DMPC and cardiolipin were used at different lipid-to-protein ratios. The incorporated protein reduces the energy absorbed during phase transitions of DMPC vesicles, and causes a small decrease in the transition temperature (tm). delta H depends on the amount of protein in the vesicles. This dependence indicates that about 72 DMPC molecules are influenced per cytochrome alpha alpha 3 monomer. The transition parameters remain unaffected by changes in ionic strength or by reduction of the enzyme. Incorporation of cytochrome oxidase depleted of subunit III into DMPC liposomes resulted in a larger decrease of tm, but the amount of perturbed phospholipids remains similar to that in the case of the intact enzyme. Incorporation of cytochrome oxidase into DMPC/cardiolipin vesicles counteracts the effect of cardiolipin in decreasing the enthalpy of the DMPC transition. Thus cytochrome oxidase segregates the phospholipids by attracting cardiolipin from the bulk lipid. Cytochrome c does not significantly affect this apparent cardiolipin 'shell' around membranous cytochrome oxidase.

Incorporation of the human erythrocyte sialoglycoprotein into recombined membranes containing cholesterol

Glycophorin, the major sialoglycoprotein from the human erythrocyte membrane, has been isolated and recombined with phosphatidylcholine and cholesterol. Sucrose density gradient analysis of the recombinants shows that it is possible not only to recombine this protein with phospholipid, but also with phospholipid-cholesterol mixtures. Surprisingly, by the same analysis, it was possible to make a recombinant with cholesterol and glycophorin, only, in the absence of added phospholipid. The accessibility of the protein to trypsin was ested in each of these recombinants. In all the recombinants which contained either phospholipid, or phospholipid and cholesterol, the protein was protected from extensive hydrolysis. This is consistent with closed vesicles and incorporation of the protein into the recombinant membrane. Extensive hydrolysis of the protein occurred in the cholesterol-glycophorin recombinant indicating some differences in structure. Freeze-fracture electron microscopy of the phospholipid and the phospholipid-cholesterol recombinants showed mostly unilamellar vesicles, 1000 to 5000 A in diameter. Intramembranous particles were observed on both fracture faces, and the fracture planes were those expected for phospholipid bilayers. The glycophorin-cholesterol recombinants also showed fracture planes consistent with bilayers, and revealed intramembranous particles. Pieces of membrane-like structures as well as apparent vesicular structures were observed. Finally in the recombinants of glycophorin with phospholipid and cholesterol, cholesterol is shown to reduce the population of the motionally restricted phospholipid headgroup environment, in proportion to the mole percent cholesterol content.

The anion permeability of vesicles reconstituted with intrinsic proteins from the human erythrocyte membrane

Band 3 protein was reconstituted with lipid vesicles consisting of 94:6 (molar ratio) egg phosphatidylcholine-bovine heart phosphatidylserine in a 2500:1 phospholipid:protein molar ratio by means of a Triton X-100/beads method. The SO2-4 permeability of the resulting vesicles was measured using an influx assay procedure in which the vesicles were sampled and subsequently eluted over Sephadex columns at appropriate time intervals. The accuracy of the assay was greatly increased by using an internal standard in order to correct for vesicle recovery. In agreement with previous work, it could be demonstrated that incorporation of band 3 in the vesicles caused an increase in SO2-4 permeability, which could be (partially) inhibited by high concentrations of DIDS or a competitive anion such as thiocyanate. However, the magnitude of the increased SO2-4 permeability was highly variable, even when vesicles were reconstituted using band 3 isolated from one batch of ghosts. In addition, the SO2-4 influx curves showed complex kinetics. These results are related to the existence of vesicle heterogeneity with respect to protein content and vesicle size as revealed by stractan density gradient centrifugation and freeze-fracture electron microscopy. Band 3 incorporation also increased the L-glucose permeability of the vesicles which could also be inhibited by DIDS. Glycophorin, which has no known transport function, reconstituted with lipid vesicles consisting of 94:6 (molar ratio) egg phosphatidylcholine-bovine heart phosphatidylserine in a 400:1 phospholipid:protein molar ration increased the bilayer permeability towards SO2-4 as well as towards L-glucose. Surprisingly, the SO2-4 permeability in the vesicles could also be inhibited by DIDS and thiocyanate. It is concluded that the use of DIDS and a competitive anion, thiocyanate, in order to prove that band 3 is functionally reconstituted, is highly questionable. The increased SO2-4 and L-glucose permeability of band 3-lipid as well as glycophorin-lipid vesicles and the inhibitory action of DIDS are discussed in the light of the presence of defects at the lipid/protein interface and protein aggregation, which may induce the formation of pores. Since the band 3-lipid vesicles are more permeable for SO2-4 than for L-glucose, in contrast to the glycophorin-containing vesicles, it is suggested that some anion specificity of the increased bilayer permeability in the band 3-lipid vesicles is still preserved.

The effect of an integral membrane protein on lipid polymorphism in the cardiolipin-Ca2+ system

The addition of Ca2+ to aqueous dispersions of cardiolipin triggers complete hexagonal (HII) phase formation at Ca2+/cardiolipin molar ratios greater than or equal to 1.0 as detected by 31PNMR and freeze-fracture electron microscopy. Incorporation of the integral membrane protein glycophorin prevents the bilayer leads to hexagonal (HII) phase transition at Ca2+/cardiolipin ratios as high as 15:1. Removal of the outwardly oriented, negatively charged sialic-acid-containing sugar groups of glycophorin with trypsin had little effect on the bilayer-stabilizing capacity of the protein. As the Ca2+ binding was found to be similar in both the cardiolipin and the cardiolipin-glycophorin systems, it can be concluded that the protein exerts a bilayer-stabilizing effect on the cardiolipin. In addition, the possibility that glycophorin may prevent vesicle fusion is also discussed.

Low-frequency motion in membranes. The effect of cholesterol and proteins

Nuclear magnetic resonance (NMR) relaxation techniques have been used to study the effect of lipid-protein interactions on the dynamics of membrane lipids. Proton enhanced (PE) 13C-NMR measurements are reported for the methylene chain resonances in red blood cell membranes and their lipid extracts. For comparison similar measurements have been made of phospholipid dispersions containing cholesterol and the polypeptide gramicidin A+. It is found that the spin-lattice relaxation time in the rotating reference frame (T1 rho) is far more sensitive to protein, gramicidin A+ or cholesterol content than is the laboratory frame relaxation time (T1). Based on this data it is concluded that the addition of the second component to a lipid bilayer produces a low-frequency motion in the region of 10(5) to 10(7) Hz within the membrane lipid. The T1 rho for the superimposed resonance peaks derived from all parts of the phospholipid chain are all influenced in the same manner suggesting that the low frequency motion involves collective movements of large segments of the hydrocarbon chain. Because of the molecular co-operativity implied in this type of motion and the greater sensitivity of T1 rho to the effects of lipid-protein interactions generally, it is proposed that these low-frequency perturbations are felt at a greater distance from the protein than those at higher frequencies which dominate T1.

Lectin-mediated agglutination of liposomes containing glycophorin. Effect of acyl chain length

Glycophorin from human erythrocytes has been incorporated into liposomes of dimyristoylphosphatidylcholine (DMPC), dipalmitoylphosphatidylcholine (DPPC) and distearoylphosphatidylcholine (DSPC). The thermal properties of unsonicated liposomes with glycophorin/lipid molar ratios up to 4.10(-3) have been studied by differential scanning calorimetry and the numbers of lipids withdrawn from participation in the gel-to-lamellar phase transition were found to be 42 +/- 22 (DMPC), 197 +/- 28 (DPPC) and 240 +/- 64 (DSPC). The initial rates of agglutination of sonicated liposomes with glycophorin/lipid molar ratios up to 4.10(-3) by wheat germ agglutinin in the concentration range 0-7 microM have been measured over a range of temperature. Below the gel-to-lamellar phase transition (Tc) the rates of agglutination increase with acyl chain length in the sequence DMPC less than DPPC less than DSPC. Agglutination is found to be second order in liposome concentration and is completely reversed on saturation of the wheat germ agglutinin-binding sites by N-acetylglucosamine. Agglutination rates decrease with increasing temperature below Tc and are largely independent of temperature above Tc. The results are discussed in relation to the clustering of glycophorin in the phospholipid bilayers and its effect on binding and subsequent interliposomal bridge formation by wheat germ agglutinin.

Phase separation induced by melittin in negatively-charged phospholipid bilayers as detected by fluorescence polarization and differential scanning calorimetry

Interactions between melittin and a variety of negatively-charged lipid bilayers have been investigated by intrinsic fluorescence, fluorescence polarization of 1,6-diphenylhexatriene and differential scanning calorimetry. (1) Intrinsic fluorescence of the single tryptophan residue of melittin shows that binding of this peptide to negatively-charged phospholipids is directly related to the surface charge density, but is unaffected by the physical rate of lipids, fluid or gel, single-shell vesicles or unsonicated dispersions. (2) Changes in the thermotropic properties of negatively-charged lipids upon melittin binding allow to differentiate two groups of lipids: (i) A progressive disappearance of the transition, without any shift in temperature, is observed with monoacid C14 lipids such as dimyristoylphosphatidylglycerol and -serine (group 1). (ii) With a second group of lipids (group 2), a transition occurs even at melittin saturation, and two transitions are detected at intermediate melittin content, one corresponding to remaining unperturbed lipids, the other shifted downward by 10-20 degrees C. This second group of lipids is constituted by monoacid C16 lipids, dipalmitoylphosphatidylglycerol and -serine. Phosphatidic acids also enter this classification, but it is the net charge of the phosphate group which allows to discriminate: singly charged phosphatidic acids belong to group 2, whereas totally ionized ones behave like group 1 lipids, whatever the chain length. (3) It is concluded that melittin induces phase separations between unperturbed lipid regions which give a transition at the same temperature as pure lipid, and peptide rich domains in which the stoichiometry is 1 toxin per 8 phospholipids. The properties of such domains depend on the bilayer stability: in the case of C16 aliphatic chains and singly charged polar heads, the lipid-peptide domains have a transition at a lower temperature than the pure lipid. With shorter C14 chains or with two net charges by polar group, the bilayer structure is probably totally disrupted, and the new resulting phase can no longer lead to a cooperative transition.

The use of differential scanning calorimetry and differential thermal analysis in studies of model and biological membranes

Differential scanning calorimetry (DSC), and to a lesser extent differential thermal analysis (DTA), are powerful yet relatively rapid and inexpensive thermodynamic techniques for studying the thermotropic phase behavior of lipids in model and biological membranes, without the introduction of exogenous probe molecules. In this review the principles as well as the scope and limitations of DSC and DTA are discussed first. The application of these techniques to the study of the thermotropic phase behavior of aqueous dispersions of various single synthetic phospholipids are then summarized, and the effects of cholesterol, free fatty acids, lysophospholipids, drugs, anesthetics and proteins on the gel to liquid-crystalline phase transitions exhibited by these model systems are discussed. The phase mixing properties of model membranes consisting of mixtures of two or more synthetic or natural phospholipids are considered next. Finally, the thermotropic phase behavior of prokaryotic plasma membranes and of the plasma, microsomal and mitochondrial membranes of eukaryotic cells are reviewed, and the applications of DSC and DTA to study the thermal behavior of specific membrane proteins, as well as the physical properties of the membrane lipid phase, are summarized.

Lateral diffusion of gramicidin S, M-13 coat protein and glycophorin in bilayers of saturated phospholipids. Mean field and Monte Carlo studies

We have developed a general model that relates the lateral diffusion coefficient of one isolated large intrinsic molecule (mol. wt. greater than or approximately 1000) in a phosphatidylcholine bilayer to the static lipid hydrocarbon chain order. We have studied how protein lateral diffusion can depend upon protein-lipid interactions but have not investigated possible non-specific contributions from gel-state lattice defects. The model has been used in Monte Carlo simulations or in mean-field approximations to study the lateral diffusion coefficients of Gramicidin S, the M-13 coat protein and glycophorin in dimyristoyl- and dipalmitoylphosphatidylcholine (DMPC and DPPC) bilayers as functions of temperature. Our calculated lateral diffusion coefficients for Gramicidin S and the M-13 coat protein are in good agreement with what has been observed and suggest that Gramicidin S is in a dimeric form in DMPC bilayers. In the case of glycophorin we find that the 'ice breaker' effect can be understood as a consequence of perturbation of the lipid polar region around the protein. In order to understand this effect is necessary that the protein hydrophilic section perturb the polar regions of at least approx. 24 lipid molecules, in good agreement with the numbers of 29-30 measured using 31P-NMR. Because of lipid-lipid interactions this effect extends itself out to four or five lipid layers away from the protein so that the hydrocarbon chains of between approx. 74 and approx. 108 lipid molecules are more disordered in the gel phase, so contributing less to the transition enthalpy, in agreement with the numbers of 80-100 deduced from differential scanning calorimetry (DSC). An understanding of the abrupt change in the diffusion coefficient at a temperature below the main bilayer transition temperature requires an additional mechanism. We propose that this change may be a consequence of a 'coupling-uncoupling' transition involving the protein hydrophilic section and the lipid polar regions, which may be triggered by the lipid bilayer pretransition. Our calculation of the average number of gauche bonds per lipid chain as a function of temperature and distance away from an isolated polypeptide or integral protein shows the extent of statically disordered lipid around such molecules. The range of this disorder depends upon temperature, particularly near the main transition.