31P nuclear magnetic resonance studies of the phospholipid-protein interface in cell membranes.

Both native and recombined membrane systems from the human erythrocyte membrane and the rabbit sarcoplasmic reticulum have been studied with 31P Nuclear Magnetic Resonance (NMR). We compare intensities of the anisotropic 31P resonance exhibited by these membranes with the intensity expected from the known phospholipid content of the membranous sample. In a recombinant with human erythrocyte glycophorin, a component of the phospholipid is "missing" from the 31P NMR resonance, apparently due to a severe broadening of the resonance of that component. Approximately 29 phospholipid molecules were found immobilized per glycophorin molecule in the membrane, regardless of the phospholipid:protein ratio. Cholesterol may inhibit the immobilization of phospholipids by glycophorin. Recombinants with band three from the human erythrocyte membrane contain an immobilized phospholipid component, analogous to the results with glycophorin. 31P NMR data from the native sarcoplasmic reticulum membrane also revealed an immobilized phospholipid component whose magnitude is independent of temperature between 30 degrees C and 45 degrees C. Extensive papain proteolysis of the membrane completely digests the Ca++ Mg++ ATPase and removes the immobilization of phospholipids noted in the intact membrane. Limited trypsin cleavage, however, does not completely remove the immobilized component; salt reduces the immobilized component.     

Perturbations of phospholipid head groups by membrane proteins in biological membranes and recombinants.

P-31 nuclear magnetic resonance (NMR) spin-lattice relaxation times (T1) have been used to probe the behavior of phospholipid head groups in the presence of membrane proteins. Measurements have been made on rabbit muscle sarcoplasmic reticulum and recombinants of the Ca2+ Mg2+ ATPase, rod outer segment disk membranes and recombinants of rhodopsin, and human erythrocyte ghosts and recombinants of human erythrocyte glycophorin. Recombined membranes with lipid/protein ratios greater than or equal to that found in biological membranes showed T1 behavior similar to the biological membranes and pure phosphatidylcholine. However, recombined membranes with a low lipid/protein ratio exhibited a T1 that was dramatically shorter than any of the other systems. Analysis of the relaxation mechanism and the factors contributing to it implicate a phospholipid head group conformation change at high protein content. It is suggested that this is due to trapping of phospholipid between proteins and is not the same phenomenon as motional restriction at the lipid-protein interface at higher lipid contents.     

31P nuclear magnetic resonance evidence for polyphosphoinositide associated with the hydrophobic segment of glycophorin A.

Glycophorin A, the major human erythrocyte sialoglycoprotein, contains a significant amount of phosphorus when isolated by the lithium diiodosalicylate-phenol procedure. Only a small percentage (approximately 1%) of this phosphorus is phosphoprotein. 31P nuclear magnetic resonance (NMR) analysis of glycophorin A has identified the remaining phosphorus content as phospholipid in origin. From the 31P chemical shifts, the phospholipid has been identified as diphosphoinositide. 31P NMR spectra of the peptides produced by trypsin hydrolysis of glycophorin A reveal that all the diphosphoinositide is closely associated with the hydrophobic region of the protein, suggesting that there is a specific affinity between this phospholipid and the intramembranous portion of glycophorin A.     

Incorporation of the human erythrocyte sialoglycoprotein into recombined membranes containing cholesterol

Summary 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 tested 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 Å 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.     

Coisolation of glycophorin A and polyphosphoinositides from human erythrocyte membranes.

The lipid composition of purified erythrocyte membrane glycophorin was measured. Diphosphoinositide, triphosphoinositide, and phosphatidylserine are the major phospholipids in glycophorin preparation. Nearly all of the radioactive diphosphoinositide and triphosphoinositide extracted from erythrocyte membranes by lithium d?odosalicylate are recoverd in purified glycophorin. There appeared to be no significant enrichment of other acidic membrane phospholipids in the protein. The results do not permit a firm conclusion as to whether the polyphosphoinositides are associated specifically with the membrane protein or whether fortuitous binding has occurred during purification.     

Interaction between glycophorin and phospholipids in recombined systems

Both the MN-glycoprotein from human erythrocytes and the hydrophobic fragment from the protein isolated with trypsin treatment, T(is), have been recombined with egg phosphatidylcholine in bilayers at various phospholipid/protein ratios. In order to investigate the effect of the protein on the phospholipid headgroups, ³¹P nuclear magnetic resonance spectra were obtained with the MN-glycoprotein recombined with egg phosphtidylcholine, which revealed two classes of phospholipid enviroments, one immobilized and one not immobilized. Electron spin resonance (ESR) of fatty acid methyl ester spin labels provided supporting evidence. Computer analysis of the ESR spectra indicate that 4–5 moles of phospholipid are immobilized per mole of protein over a wide range of lipid-to-protein ratios. The immobilization of the phospholipids appears mediated by both the polar headgroups and the hydrocarbon tails of the phospholipid.     

Phosphorylation in membranes of intact human erythrocytes.

Studies of phosphorylation in membranes of intact human erythrocytes were performed by incubating erythrocytes in inorganic [32P]phosphate. Analysis of membrane proteins by polyacrylamide gel electrophoresis showed a pattern of phosphorylation similar to that observed when ghost membranes were incubated with [gamma-32P]ATP. Membrane lipid phosphorylation was also similar in intact cells and ghosts. The most heavily phosphorylated lipid, polyphosphoinositide, was closely associated with glycophorin A, the major erythrocyte membrane sialoglycoprotein obtained when the sialoglycoprotein fraction was isolated by the lithium diiodosalicylate-phenol partition procedure. Only 1 molecule of glycophorin A out of every 100 was found to be phosphorylated, and the phosphate exchange occurred specifically in the COOH-terminal intracellular portion of glycophorin A. These studies show that the human erythrocyte can be used as a model for membrane phosphorylation in an intact cell system.     

Phospholipid-protein interactions in human low density lipoprotein detected by 31P nuclear magnetic resonance.

31P nuclear magnetic resonance (NMR) spectra of human low density lipoprotein (LDL) has been obtained and the major phospholipid components identified. Analysis of the spectra revealed two phospholipid environments: one occupied by 4/5 of the phospholipid with high resolution resonances possessing properties similar to phospholipids in vesicles, and a second occupied by 1/5 of the phospholipid with broad lines indicative of immobilization. Limited trypsin treatment of the particle cleaved all of the B peptide into smaller molecular weight peptides which remained with the particle. Trypsin-treated LDL eluted from a Sepharose CL-6B column similarly to native LDL so that the modified particle remained intact. 31P NMR spectra of trypsin-treated LDL showed little or no immobilized phospholipid. The immobilization in the native LDL particle is attributed to lipid-protein interactions between 1/5 of the phospholipid and the B peptide.     

Association of human erythrocyte membrane glycoproteins with blood-group Cad specificity.

Sodium dodecyl sulphate/polyacrylamide-gel electrophoresis of erythrocyte membranes from a blood-group-B individual with the rare Cad phenotype indicates a lower-than-normal mobility of the main sialoglycoproteins, suggesting an increase in apparent molecular mass of 3kDa and 2kDa respectively for glycoprotein alpha (synonym glycophorin A) and glycoprotein delta (synonym glycophorin B). Since the chief structural determinant of Cad specificity is N-acetylgalactosamine, the membrane receptors have been isolated by affinity binding on immobilized Dolichos biflorus (horse gram) lectin. The predominant species eluted from the gel was the abnormal glycoprotein alpha, whereas in control experiments no material could be recovered from the adsorbent incubated with group-B Cad-negative erythrocyte membranes. After partition of the membranes with organic solvents, the blood-group-Cad activity was found in aqueous phases containing the sialoglycoproteins, but not in the organic phases containing simple or complex glycolipids, which, however, retained the blood-group-B activity. The carbohydrate composition of highly purified lipid-free glycoprotein alpha molecules prepared from Cad and control erythrocytes was determined. Interestingly the molar ratio of N-acetylneuraminic acid to N-acetylgalactosamine was equal to 2:1 in the case of controls and equal to 1:1 in the case of Cad erythrocytes. Taken together these results suggest that Cad specificity is defined by N-acetylgalactosamine residues carried by the alkali-labile oligosaccharide chains attached to the erythrocyte membrane sialo-glycoproteins.     

Effect of glycophorin on lipid polymorphism: A 31P-NMR study

(1) The effect of glycophorin, a major intrinsic glycoprotein of the human erythrocyte membrane, on lipid polymorphism has been investigated by ³¹P-NMR (at 36.4 MHz) and by freeze-fracture electron microscopy. (2) Incorporation of glycophorin into vesicles of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) results in the formation of unilamellar vesicles (1000–5000 Å diameter) which exhibit ³¹P-NMR bilayer spectra over a wide range of temperature. A reduction in the chemical shift anisotropy (Δσ_csa^eff) and an increase in spectral linewidth in comparison to dioleoylphosphatidylcholine liposomes may suggest a decrease in phospholipid headgroup order. (3) 1,2-Dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), in the presence of excess water, undergoes a bilayer to hexagonal (H_II) phospholipid arrangement as the temperature is increased above 0°C. Incorporation of glycophorin into this system stabilizes the bilayer configuration, prohibiting the formation of the H_II phase. (4) Cosonication of glycophorin with DOPE in aqueous solution (pH 7.4) produces small, stable unilamellar vesicles (300–1000 Å diameter), unlike DOPE alone which is unstable and precipitates from solution. (5) The current study demonstrates the bilayer stabilizing capacity of an intrinsic membrane protein, glycophorin, most likely by means of a strong hydrophobic interaction between the membrane spanning portion of glycophorin and the hydrophobic region of the phospholipid.     

Effect of lipid composition on HVJ-mediated fusion of glycophorin liposomes to erythrocytes

We previously reported that liposomes containing glycophorin or gangliosides, both of which were isolated from human erythrocytes, are efficiently fused to erythrocyte membranes in the presence of HVJ (Umeda, M. et al., J. Biochem. 94, 1955-1966 (1983), and Virology 133, 172-182 (1984]. In the present work, the effect of lipid composition in glycophorin liposomes on their sensitivity to fusion with erythrocytes was studied. Very little fusion occurred when glycophorin liposomes composed of dipalmitoylphosphatidylcholine-dicetylphosphate (9:1), dimyristoylphosphatidylcholine-dicetylphosphate (9:1), or egg yolk phosphatidylcholine-dicetylphosphate (9:1) were incubated with human erythrocytes in the presence of HVJ at 37 degrees C. Addition of cholesterol into these liposomal membranes greatly enhanced the sensitivity of the liposomes to fusion. The presence of phosphatidic acid and phosphatidylethanolamine in liposomes also enhanced the sensitivity, whereas the presence of lysophosphatidylcholine had no significant effect on the ability of the liposomes to fuse. The fusion efficiency of liposomes was also enhanced by the presence of glucosylceramide. Change of lipid composition in liposomes had, however, no appreciable influence on the HVJ-mediated binding of liposomes to erythrocytes, suggesting that the interaction between HANA protein of HVJ and glycophorin in liposomes was not affected by the lipid composition of the liposomes.     

Relation between various phospholipase actions on human red cell membranes and the interfacial phospholipid pressure in monolayers.

The action of purified phospholipases on monomolecular films of various interfacial pressures is compared with the action on erythrocyte membranes. The phospholipases which cannot hyorolyse phospholipids of the intact erythrocyte membrane, phospholipase C from Bacillus cereus, phospholipase A2 from pig pancreas and Crotalus adamanteus and phospholipase D from cabbage, can hydrolyse phospholipid monolayers at pressure below 31 dynes/cm only. The phospholipases which can hydrolyse phospholipids of the intact erythrocyte membrane, phospholipase C from Clostridium welchii phospholipase A2 from Naja naja and bee venom and sphingomyelinase from Staphylococcus aureus, can hydrolyse phospholipid monolayers at pressure above 31 dynes/cm. It is concluded that the lipid packing in the outer monolayer of the erythrocyte membrane is comparable with a lateral surface pressure between 31 and 34.8 dynes/cm.     

Interaction of sialoglycoproteins with wheat germ agglutinin-sepharose of varying ratio of lectin to Sepharose. Use for the purification of mucin glycoproteins from membrane extracts

The influence of varying the amount of wheat germ agglutinin immobilized on Sepharose beads on the binding of glycoproteins to these beads was investigated. A series of wheat germ agglutinin-Sepharose gels containing between 0.10 and 10.0 mg of lectin/ml of gel was prepared, and the actual lectin content was established by acid hydrolysis of the gel followed by analysis of glycine, a major amino acid in wheat germ agglutinin. Affinity chromatography of labeled glycoproteins indicated that glycophorin bound to all the wheat germ agglutinin-Sepharose preparations. Fetuin, ovomucoid, and alpha 1-acid glycoprotein bound not at all or very poorly to gels with a low content of wheat germ agglutinin (less than 0.95 mg/ml). The specific binding of these glycoproteins increased with increasing lectin content on the gels, and on gels of high content (greater than 3 mg/ml) the binding was virtually quantitative. On chromatographing a mixture of glycophorin, alpha 1-acid glycoprotein, fetuin, and ovomucoid on wheat germ agglutinin-Sepharose, containing 0.08 mg of lectin/ml of gel, glycophorin was selectively retained on the gel. It was possible to purify glycophorin from an extract of human erythrocyte membranes in one step by chromatography on the above gel. By using the series of gels, it was demonstrated that Morris hepatoma 7777 membranes contained at least 4-fold more sialoglycoproteins which bound to low density wheat germ agglutinin-Sepharose compared to rat liver membranes. These hepatoma sialoglycoproteins were isolated, purified, and partially characterized as having a high proportion of O-linked sialyloligosaccharides. Our studies illustrate the use of low density wheat germ agglutinin-Sepharose gels both for the detection and for easy isolation of mucin-type glycoproteins from crude extracts of cells or membranes.     

Phosphorus nuclear magnetic resonance studies of lipid-protein interactions: human erythrocyte glycophorin and phospholipids

Human erythrocyte glycophorin containing four molecules of phospholipid tightly bound to the protein was isolated from human red cell ghosts. This protein preparation was reconstituted into a digalactosyl diglyceride bilayer. The 31P NMR spectrum of this reconstituted membrane produced an axially symmetric powder pattern arising exclusively from the phospholipids bound to glycophorin. The width of the powder pattern, about 90 ppm, is about twice as broad as that normally exhibited by a phospholipid bilayer. The chemical shift tensor is perturbed relative to phospholipids in a bilayer. The spin-lattice relaxation rate of these protein-bound phospholipids is found to be nearly an order of magnitude faster than phospholipids in a bilayer. The results are consistent with phospholipids tightly bound to the membrane protein and undergoing rotational diffusion, perhaps as a complex of phospholipid and protein.     

Bilayer structure in phospholipid-cytochrome c model membranes.

Lipid protein interactions in biological membranes differ markedly depending on whether the protein is intrinsic or extrinsic. These interactions are studied using lipid spin labels diffused into model systems consisting of phospholipid bilayers and a specific protein. Recently, an intrinsic protein complex, cytochrome oxidase, was examined and the data suggest there is a boundary layer of immobilized lipid between the hydrophobic protein surfaces and adjacent fluid bilayer regions. In the present study, a typical extrinsic protein, cytochrome c, was complexed with a cardiolipin/lecithin (1:4 by weight) mixture. The phospholipids in the presence and absence of cytochrome c exhibit typical bilayer behavior as jedged by four spin-labeling criteria: fluidity gradient, spectral anisotropy of oriented bilayers, response to hydration and the polarity profile. Any effects of cytochrome c on the ESR spectra of lipid spin labels are small, in contrast to the effects of intrinsic proteins. These data are consistent with electrostatic binding of cytochrome c to the charged groups of the phospholipids, and indicate that the presence of extrinsic proteins will not interfere with measurements of boundary lipid in intact biological membranes.     

Static and magic angle spinning (31)P NMR spectroscopy of two natural plasma membranes

Static and magic angle spinning (31)P NMR spectroscopy was used for the first time in natural plasma membranes from erythrocytes and skeletal muscle to study phospholipid arrangement and composition. Typical static powder-like spectra were obtained showing that phospholipids were in a bilayer arrangement. Magic angle spinning narrowed spectra into two components. The first one corresponded to phosphatidylcholine and the second one to the other phospholipids with intensities in agreement with the known phospholipid composition. These findings show that NMR data previously acquired using model membranes can be transposed to studies on phospholipids in their natural environment.     

Organization of lipids in sarcoplasmic reticulum membrane and Ca2+-dependent ATPase activity.

The organization of lipids in sarcoplasmic reticulum membrane was studied with a variety of stearic spin labels and a phosphatidylcholine spin label. The ESR spectra of the spin-labeled membranes consisted of two components, one due to labels in lipid bilayer structure and the other due to more immobilized labels. The relative intensity of the immobilized component increased when the lipid content of the membrane was decreased by treatment with phospholipase A [EC 3.1.1.4] and subsequent washing with bovine serum albumin. Membrane containing 30% of the intact phospholipid, i.e.0.15 mg of phospholipid per mg of protein, showed a spectrum consisting only of the immobilized component (the overall splitting ranged from 58.5 G to 60.5 G). The immobilized component was ascribed to lipids complexed with protein. The fraction of lipids in the two different organizations was determined from the ESR spectrum. The activity of the Ca2+-Mg2+ dependent ATPase [ATP phosphohydrolase, EC 3.6.1.3] was found to increase almost linearly with the lipid bilayer content in the membrane, whereas phosphoenzyme formation was almost independent of the bilayer content. This indicated that the bilayer structure is necessary for the ATPase to attain its full transport activity.     

Spin-label studies of the oligomeric structure of band 3 protein in erythrocyte membranes and in reconstituted systems

A spin-labeled fatty acid (16-doxylstearic acid), linked by an ester bond to a maleimide or a nitrene residue, was covalently attached to band 3 of erythrocyte membranes. The electron spin resonance spectrum of the spin-labeled protein was examined at different temperatures in: (a) whole erythrocyte ghosts; (b) ghosts depleted of spectrin and actin; (c) alkaline-treated ghosts; (d) vesicles made with purified band 3 reassociated with dimyristoylphosphatidylcholine. Most spectra are composite with a major component corresponding to a large overall splitting. The determination of the percentage of the immobilized component was carried out by pairwise subtraction. At low temperatures (1–7°C), the highest fraction of immobilized component was found in dimyristoylphosphatidylcholine vesicles (approx. 100%); alkaline-treated membranes had approx. 75% of the immobilized component at the same temperature; whole erythrocyte, spectrin/actin-depleted and spectrin/actin/ankyrin-depleted ghosts gave identical results (approx. 60% of immobilized component). The immobilized fraction decreased in all samples with increasing temperature or addition of a nonsolubilizing concentration of dodecyl octaethylene glycol monoether. In dimyristoylphosphatidylcholine vesicles, however, the modification in the ratio of the two components was obtained only above the lipid transition temperature (23°C). The strong immobilization of the spin-labeled lipid chain at all temperatures suggested trapping of the lipid chain between proteins. At low temperature, in dimyristoylphosphatidylcholine vesicles or in alkaline-treated ghosts, lipid-protein segregation is likely to take place. In whole erythrocyte ghosts, on the other hand, the large contribution of the motionally restricted component at physiological temperature indicates the oligomeric nature of band 3. Partial dissociation of the oligomers occurs as the temperature is increased, but the presence or absence of cytoskeletal proteins has no influence on the state of oligomerization of band 3.     

Optical-rotatory-dispersion studies of compounds related to cholesterol in liposomes and the membranes of erythrocyte `ghosts''

1. Steroid molecules containing the alpha,beta-unsaturated oxo group in various positions were incorporated with egg phosphatidylcholine into liposomes and into human erythrocyte membranes. 2. The liposomes formed contained 0.3-0.94mol of steroid/mol of phospholipid and the steroids replaced 19-76% of the erythrocyte membrane sterol. 3. The optical rotatory dispersion (o.r.d.) spectra of the steroids in these structures were compared with those obtained in solvents of different polarity. 4. The o.r.d. spectra of cholesta-4,6-dien-3-one and 3-hydroxycholest-3-en-2-one in liposomes resembled those obtained with polar solvents such as ethanol or triethyl phosphate-water (1:1, v/v). 5. The o.r.d. spectra of 3-hydroxycholest-7-en-6-one and 3-hydroxycholest-5-en-7-one in liposomes resembled those obtained with moderately polar solvents such as dioxan. 6. The o.r.d. spectrum of 3-hydroxycholest-8(14)-en-15-one in liposomes resembled those obtained with non-polar solvents such as cyclohexane. 7. 3-Hydroxycholest-3-en-2-one did not exchange with erythrocyte membrane cholesterol, but the other steroids did do so and the o.r.d. spectra of the membranes containing them closely resembled those obtained with liposomes. 8. From the results, the position of sterol molecules with respect to the phospholipid molecules in liposomes and membranes of human erythrocyte ;ghosts' can be deduced.     

Reincorporation of adenosine 5'-diphosphate/adenosine 5'-triphosphate carrier into phospholipid membranes. Phospholipid-protein interaction as studied by phosphorus-31 nuclear magnetic resonance and electron microscopy

Combined phosphorus-31 nuclear magnetic resonance (31P NMR) and electron microscopic studies were performed on the ADP/ATP carrier protein from beef heart mitochondria. The protein was incorporated into phospholipids by addition of Triton-protein micelles to a lipid suspension or to the dry lipid. All of the phospholipid (egg phosphatidylcholine or mixtures of egg phosphatidylcholine and egg phosphatidylethanolamine) that contributed to the observed 31P NMR signal under these conditions appeared to be in a bilayer configuration. Freeze-fracturing and negative-staining electron microscopy showed unilamellar vesicles and multilayers. An isotropic signal could be attributed to vesicle rotation, judging from its sensitivity to increasing viscosity. The presence of small vesicles was also noticeable in the 31P NMR spectra of planar oriented membranes. In the presence of phosphatidylethanolamine, aggregation of protein particles was observed. Gel chromatography of the protein-Triton-phospholipid mixture revealed that, before Triton removal, large amounts of protein are associated with multibilayers. Separation of loaded and unloaded membranes by centrifugation in D2O showed that, upon stepwise addition, protein incorporates preferentially into unloaded liposomes. From these findings a mechanism of protein reincorporation was deduced.