Search

Peptide E5 is an analogue of the fusion peptide of influenza virus hemagglutinin and K5 is a cationic peptide which has an arrangement of electric charges complementary to that of E5. We reported that a stoichiometric mixture of E5 and K5 caused fusion of large unilamellar vesicles (LUV) of neutral phospholipids (Murata, M., Kagiwada, S., Takahashi, S. and Ohnishi, S. (1991) J. Biol. Chem. 266, 14353-14358). K5 caused fusion of LUV composed of dioleoylphosphatidylcholine (DOPC) at pH > 10, but not at neutral pH. In the presence of oligophosphates, such as 1 mM ATP, GTP, or polyphosphate, K5 caused rapid and efficient fusion of DOPC LUV at neutral pH without hydrolysis of oligophosphate groups, but another anions such as citrate, acetate, AMP, phosphate, or EDTA were ineffective. The peptide/oligophosphate-induced fusion behaviors have been investigated by a fluorescence resonance energy transfer assay for lipid mixing of LUV and negative staining electron microscopy. At higher ionic strengths ( > 0.3 M KCl) or in the presence of 5.0 mM MgCl2, the fusion was inhibited. Even at the inhibitory conditions, the association of K5 with lipid vesicles at neutral pH was directly confirmed by the Ficoll gradient assay method and by blue shifts of the tryptophan fluorescence of the peptide. A nonhydrolyzable GTP analogue, GTP gamma S, also induced fusion. These observations suggested that the electrostatic interactions between the positive and negative charges of K5 and oligophosphate, respectively, induced complex formation, triggering membrane fusion.

To investigate the interaction between a signal peptide and the lipid bilayer, two kinds of peptides, L8-M5 (L8 = MRL8PLAALG, M5 = KVFER) and L14-M5 (L14 = MRL14PLAALG), were examined in membranes composed of dioleoylphosphatidylcholine (DOPC). Peptides L8 and L14 are artificially designed signal sequences, and M5 is the N-terminal five residues of human lysozyme; L8 mediated effective secretion of human lysozyme in yeast, while L14 did not [Yamamoto, Y., et al. (1987) Biochem. Biophys. Res. Commun. 149, 431-436]. DOPC liposomes incorporating L8-M5 or L14-M5 were observed by electron cryomicroscopy as pairs of concentric circles, and the separation of the bilayer was measured along the membrane. Peptide L8-M5 was found to reduce the bilayer thickness, but L14-M5 did not. CD measurements revealed that L8-M5 adopted an alpha-helical conformation with random coil in the liposome membranes and that L14-M5 adopted a more helical and less random conformation than L8-M5. Fluorescence spectroscopy using both aqueous and membranous probes revealed that L8-M5 destabilized the lipid bilayer more strongly than L14-M5. These results suggest that functional L8-M5 reduces the bilayer thickness and destabilizes the lipid bilayer and that these activities are important for signal peptide function.

pH-dependent membrane fusion of influenza virus with liposomes made of phosphatidylcholine was studied by the spin-labelling method. Efficiency of viral fusion with liposomes composed of dimyristoyl phosphatidylcholine or dipalmitoyl phosphatidylcholine was considerably lower compared to dioleoyl phosphatidyl choline or egg yolk phosphatidylcholine, suggesting importance of unsaturation of acyl chains of lipid bilayers. Reconstitution of specific viral receptors such as Glycophorin or sialylparagloboside strongly enhanced fusion with liposomes composed of dimyristoyl phosphatidylcholine. A direct comparison between the activities of the receptors showed that Glycophorin was about 50 times more effective than sialyparagloboside at the same receptor/phosphatidylcholine molar ratio.

The amphiphilic anionic peptides E5 and E5L can mimic the fusogenic activity of influenza hemagglutinin(HA). These peptides induced fusion of egg yolk phosphatidylcholine small or large unilamellar vesicles only at acidic pH in a similar manner to viral HA. Acetylation or acetimidylation of the N-terminus of the peptides drastically reduced the fusion activity of the intact peptides, while C-terminal amidation left the activity unchanged. The binding assay suggested that the interaction of the modified peptides with lipid membranes was almost unchanged in comparison with those of the parent peptides, and the CD spectra showed that these peptides were alpha-helical. The results showed the importance of the N-terminus of the peptides on the membrane fusion activity, although why the N-terminal modifications affect the activity is still unclear.

An anionic amphiphilic peptide and the charge-reversed cationic peptide are synthesized. They contain 20 amino acids with the same sequence except for 5 Glu residues for the anionic versus 5 Lys residues for the cationic peptides. Fusion of egg phosphatidylcholine large unilamellar vesicles is assayed with the fluorescent probes by the lipid mixing and the internal content mixing at neutral pH. The peptide mixture causes a rapid and efficient membrane fusion, in spite of no fusions with each peptide by itself. Each peptide takes nearly random coils with a small amount of helix, but the peptide mixture has an ordered helical structure. The equimolar peptide mixture forms a much more hydrophobic complex than those of different molar ratios of peptides and also that of each peptide itself. The equimolar peptide mixture causes the most efficient fusion. Preincubations of two peptides before addition to vesicles cause the slower rates of fusion. The fusion is greatly reduced at higher ionic strength and nearly zero at 800 mM NaCl and 40 mM sodium phosphate. Each peptide and the peptide mixture show the same alpha-helical structure, interact with vesicles, but do not induce fusion at higher ionic strengths. These results suggest that the two peptides interact mutually through the electrostatic Coulombic interaction between the charged groups. The electrically neutralized hydrophobic complex aggregates the separate vesicles together and interacts with the hydrocarbon region of lipid bilayers to cause fusion.

Ca2+-induced phase separation in phosphatidylserine/phosphatidylethanolamine and phosphatidylserine/phosphatidylethanolamine/phosphatidylcholine model membranes was studied using spin-labeled phosphatidylethanolamine and phosphatidylcholine and compared with that in phosphatidylserine/phosphatidylcholine model membranes studied previously. The phosphatidylethanolamine-containing membranes behaved in qualitatively the same way as did phosphatidylserine/phosphatidylcholine model membranes. There were some quantitative differences between them. The degree of phase separation was higher in the phosphatidylethanolamine-containing membranes. For example, the degree of phase separation in phosphatidylserine/phosphatidylethanolamine membranes containing various mole fractions of phosphatidylserine was 94--100% at 23 degrees C and 84--88% at 40 degrees C, while the corresponding value for phosphatidylserine/phosphatidylcholine membranes was 74--85% at 23 degrees C and 61--79% at 40 degrees C. Ca2+ concentration required for the phase separation was lower for phosphatidylserine/phosphatidylethanolamine than that for phosphatidylserine/phosphatidylcholine membranes; concentration to cause a half-maximal phase separation was 1.4 . 10(-7) M for phosphatidylserine-phosphatidylethanolamine and 1.2 . 10(-6) M for phosphatidylserine/phosphatidylcholine membranes. The phase diagram of phosphatidylserine/phosphatidylethanolamine membranes in the presence of Ca2+ was also qualitatively the same as that of phosphatidylserine/phosphatidylcholine except for the different phase transition temperatures of phosphatidylethanolamine (17 degrees C) and phosphatidylcholine (-15 degrees C). These differences were explained in terms of a greater tendency for phosphatidylethanolamine, compared to phosphatidylcholine, to form its own fluid phase separated from the Ca2+-chelated solid-phase phosphatidylserine domain.

Poly(ethylene glycol)- (PEG-) induced change in membrane fluidity and aggregation of phospholipid vesicles were studied. A threshold concentration of PEG was required to induce the aggregation. This concentration increased with a decrease in the molecular weight of PEG, e.g., from 5% (w/w) with PEG 6000 (PEG with an average molecular weight of 7500) to more than 30% (w/w) with PEG 200. The aggregation was reversible upon dilution of PEG if the initial PEG concentration was smaller than a certain value, e.g., 22% (w/w) for PEG 6000. Addition of PEG caused a decrease in membrane fluidity of the vesicles detected by fluorescence anisotropy of diphenylhexatriene and by electron spin resonance of a spin-labeled fatty acid. The anisotropy change of diphenylhexatriene fluidity change had an inflection point at approximately 5% (w/w) of PEG 6000, which might suggest that the aggregation would make the decrease of membrane fluidity smaller. Transfer of lipid molecules between phospholipid vesicles was enhanced by the PEG-induced aggregation. The enhancement occurred not only upon direct addition of PEG to the suspending medium, but also upon dialysis of the vesicle suspension against a high concentration of PEG. All these features are consistent with osmoelastic coupling in the phospholipid membranes and the subsequent osmophobic association of the vesicles. The imbalance of osmolarity between the region adjacent to the vesicle surface (exclusion layer) and the bulk aqueous phase, which results from the preferential exclusion of PEG from the exclusion layer in the case of direct addition of PEG, exerts an osmotic stress on the vesicles.(ABSTRACT TRUNCATED AT 250 WORDS)

Interaction of filipin and amphotericin B with sterols in phosphatidylcholine membranes has been studied using various spin probes; epiandrosterone, cholestanone, phosphatidylcholine with 12-nitroxide or 5-nitroxide stearate attached to 2 position and also with tempocholine at the head group. Filipin caused increase in the fluidity of cholesterol-containing phosphatidylcholine membranes near the center, while it rather decreased the fluidity near the polar surface. On the other hand, amphotericin B did not apparently affect the fluidity. In the electron spin resonance spectrum of steriod spin probes in the antibiotic-containing membranes, both bound and free signals were observed and the association constant was calculated from the siganal intensity. In the binding of steriods with filipin, both 3 and 17 positions were involved, while the 17 positions was less involved in the binding with amphotericin B. Phase change in the host membrane markedly affected the interaction of filipin with epiandrosterone probe. The bound fraction jumped from 0.4 to 0.8 on going to the crystalline state and increased further with decrease in temperature. The overall splitting of the bound signal also increased on lowering the temperature below phase transition. This change was attributed to aggregate formation of filipin-steriod complexes in the crystalline state. On the other hand, effect of phase transition was much smaller on the interaction of amphotericin B with the steriod probe.

Phosphatidylcholine (PC) transfer activity was found in human erythrocyte hemolysate. The transfer activity was assayed by the ESR peak height increase when spin-labeled PC vesicles were incubated with egg yolk PC vesicles. The transfer activity was isolated from hemoglobin by an ion exchange chromatography followed by gel filtration. The partial purification resulted in a 405-fold increase in the specific transfer activity compared with that of the hemolysate. The molecular weight of the PC transfer protein was estimated to be 23,000 by gel filtration. The transfer activity was inactivated by heat-treatment at 75 degrees C for 10 min. Phosphatidylserine vesicles strongly inhibited the activity. Half-maximal inhibition occurred on addition of 0.24 mol% of phosphatidylserine vesicles to the incubation mixture. Ca2+ restored the activity. The transfer protein was quite similar to the PC transfer protein obtained from bovine liver cytosol.

Transfer of phospholipid from the envelope of hemagglutinating virus of Japan (HVJ) to erythrocyte (RBC) membrane and the virus-induced transfer of phospholipid between RBC membranes were studied using spin-labeled phosphatidylcholine (PC). The transfer of PC from membranes labeled densely with PC to unlabeled membranes was followed by the peak height increase in the electron spin resonance spectrum. The two kinds of transfer reactions took place very rapidly as reported previously. To obtain further details, the transfer reactions were studied with HVJ, HVJ inactivated by trypsin, HVJ harvested early, HVJ grown in fibroblast cells, the fibroblast HVJ activated by trypsin, influenza virus, and glutaraldehyde-treated RBCs. The results demonstrated that the viral F glycoprotein played a crucial role in the transmembrane phospholipid movements as well as in the fusion and hemolysis of RBCs. The transfer from HVJ to RBC's occurred partially through an exchange mechanism not accompanying the envelope fusion. This was shown by a decrease in the exchange broadening of the electron spin resonance spectrum of released spin-labeled HVJ (HVJ) and also by an increase in the ratio of PC to viral proteins incorporated into RBC membranes. HVJ modified RBC membrane so as to be able to exchange its phospholipids with those of inactive membranes such as fibroblast HVJ, influenza virus, glutaraldehyde-treated RBC'S, and phosphatidylcholine vesicles. HVJ affected the fluidity of RBC membranes markedly, the environments around PC being much fluidized. The virus-induced fusion was discussed based on close apposition of the membranes by HANA proteins and on the destabilization and fluidization of RBC membranes by F glycoproteins.

Effect of bilayer membrane curvature of substrate phosphatidylcholine and inhibitor phosphatidylserine on the activity of phosphatidylcholine exchange protein has been studied by measuring transfer of spin-labeled phosphatidylcholine between vesicles, vesicles and liposomes, and between liposomes. The transfer rate between vesicles was more than 100 times larger than that between vesicles and liposomes. The transfer rate between liposomes was still smaller than that between vesicles and liposomes and nearly the same as that in the absence of exchange protein. The markedly enhanced exchange with vesicles was ascribed to the asymmetric packing of phospholipid molecules in the outer layer of the highly curved bilayer membrane. The inhibitory effect of phosphatidylserine was also greatly dependent on the membrane curvature. The vesicles with diameter of 17 nm showed more than 20 times larger inhibitory activity than those with diameter of 22 nm. The inhibitory effect of liposomes was very small. The size dependence was ascribed to stronger binding of the exchange protein to membranes with higher curvatures. The protein-mediated transfer from vesicles to spiculated erythrocyte ghosts was about four times faster than that to cup-shaped ghosts. This was ascribed to enhanced transfer to the highly curved spiculated membrane sites rather than greater mobility of phosphatidylcholine in the spiculated ghost membrane.

Disappearance of Ca2+-induced phase separation in phosphatidylserine-phosphatidylcholine membrane has been studied under several conditions by monitoring electron spin resonance spectrum of spin-labeled phosphatidylcholine. The membranes were prepared in Millipore filters. Electron micrographs of the pre parations showed formation of multilayered structures lined on the pore surface. The phase separation was disappeared when the membrane was soaked in non-buffered salt solution (100 ml KCl, pH 5.5). It was markedly contrasting that when the bathing salt solution was buffered no disappearance was observed. Disappearance of the phase separation was also observed when the Ca2+-treated membrane was transferred to acidic salt solutions (less than or equal to pH 2.5) or to low ionic strength media (less than or equal to mM) buffered at pH 5.5, and then to the buffered salt solution (100 mM KCl, pH 5.5). These are due to replacement of Ca2+ by proton, proton-induced separation, followed by disappearance of the phase separation in the buffered salt solution. Biological significance of the competition between Ca2+ and proton for the phase separation or domain formation in the membranes was emphasized.

Effects of ph and ionic strength on phosphatidylserine/phosphatidylcholine mixed membranes prepared on Millipore filter pore surfaces have been studied using spin-labeled phosphatidylcholine. Lowering pH at constant ionic strength and lowering ionic strength at constant pH caused a lateral reorganization of the membrane. The trigger was protonation of the serine carboxyl group which caused solidification of phosphatidylserine molecules in the membrane, leaving a fluid phase consisting mainly of phosphatidylcholine. The appearent pK for the proton-induced phase separation was measured in a wide range of salt concentrations. The ionic strength dependence was satisfactorily explained based on the electrostatic free energy of proton in the field of membrane surface potential. The Gouy-Chapman theory gave a good approximation for the surface potential. The surface pK of phosphatidylserine and phosphatidic acid vesicles was directly measured in various salt concentrations by 31P-NMR and the results confirmed validity of the Gouy-Chapman-type analysis. The lateral reorganization was triggered by electrostatic interaction but the bulk of the stabilization energy for the structural changes would be the gains in intermolecular van der Waals energy due to closer packing of phosphatidylserine on solidification.

Reconstituted vesicles of hemagglutinin glycoproteins into egg yolk phosphatidylcholine/spin-labeled phosphatidylcholine/cholesterol (molar ratio 1.6:0.4:1) were prepared by dialysis. Preparations at appropriate protein-to-lipid ratios (1:44 and 1:105 mol/mol) contained vesicles with a diameter of 100-300 nm and a high density of spikes on the surface. These vesicles showed low pH-induced membrane fusion activity. At pH 5.2 and 37 degrees C, fusion with erythrocyte membranes took place very rapidly within 1-2 min and reached a plateau at 63-66% fusion. The fusion was negligibly small at neutral pH and was induced to occur at pH values lower than 6.0. The reconstituted vesicles caused hemolysis and fusion of human erythrocyte cells in the same pH range as that of the fusion with erythrocyte membranes. The low pH-induced fusion activity of the reconstituted vesicles is essentially the same as that of the parent virus. These vesicles can be used to deliver some reagents or drugs into target cell cytoplasm via fusion at lysosomes.