Membrane fusion of enveloped viruses: especially a matter of proteins

Temperature-induced vesicle aggregation in catanionic surfactant systems: the effects of the headgroup and counterion

The peculiar nature of temperature-induced vesicle aggregation (TIVA) in some catanionic surfactant systems is systematically investigated. On the basis of a general analysis of the intervesicular interactions, the main driving force for this phenomenon is considered to be the intervesicular hydrophobic interaction among the exposed hydrophobic part of the surfactant headgroups. The addition of an oppositely charged hydrophobic salt to the catanionic vesicle systems is also found to promote the occurrence of TIVA. In fact, TIVA can be induced in ordinary catanionic vesicle systems by the addition of an oppositely charged hydrophobic counterion.

A critical evaluation of the conformational requirements of fusogenic peptides in membranes

It is generally assumed that fusogenic peptides would require a certain conformation, which triggers or participates in the rate-determining step of membrane fusion. Previous structure analyses of the viral fusion peptide from gp41 of HIV-1 have yielded contradictory results, showing either an alpha-helical or a beta-stranded conformation under different conditions. To find out whether either of these conformations is relevant in the actual fusion process, we have placed sterically demanding substitutions into the fusion peptide FP23 to prevent or partially inhibit folding and self-assembly. A single substitution of either D- or L-CF(3)-phenylglycine was introduced in different positions of the sequence, and the capability of these peptide analogues to fuse large unilamellar vesicles was monitored by lipid mixing and dynamic light scattering. If fusion proceeds via a beta-stranded oligomer, then the D- and L-epimers are expected to differ systematically in their activity, since the D-epimers should be unable to form beta-structures due to sterical hindrance. If an alpha-helical conformation is relevant for fusion, then the D-epimers would be slightly disfavoured compared to the L-forms, hence a small systematic difference in fusion activity should be observed. Interestingly, we find that (1) all D- and L-epimers are fusogenically active, though to different extents compared to the wild type, and--most importantly--(ii) there is no systematic preference for either the D- or L-forms. We therefore suggest that a well-structured alpha-helical peptide conformation or a beta-stranded oligomeric assembly can be excluded as the rate-determining state. Instead, fusion appears to involve conformationally disordered peptides with a pronounced structural plasticity.

Effects of temperature on viral glycoprotein mobility and a possible role of internal "viroskeleton" proteins in Sendai virus fusion

The effect of temperature on fusion of Sendai virus with target membranes and mobility of the viral glycoproteins was studied with fluorescence methods. When intact virus was used, the fusion threshold temperature (20-22 degrees C) was not altered regardless of the different types of target membranes. Viral glycoprotein mobility in the intact virus increased with temperature, particularly sharply at the fusion threshold temperature. This effect was suppressed by the presence of erythrocyte ghosts and/or dextran sulfate in the virus suspension. In these cases also, no change in the fusion threshold temperature was observed. On the other hand, reconstituted viral envelopes (virosomes) bearing viral glycoproteins but lacking matrix proteins were capable of fusing with erythrocyte ghosts even at temperatures lower than the fusion threshold temperature and no fusion threshold temperature was observed over the range of 10-40 degrees C. The mobility of viral glycoproteins on virosomes was much greater and virtually temperature-independent. The intact virus treated with an actin-affector, jasplakinolide, reduced the extent of fusion with erythrocyte ghosts and the mobility of viral glycoproteins, while the treatment of virosomes with the same drug did not affect the extent of fusion of virosomes with erythrocyte ghosts and the mobility of the glycoproteins. These results suggest that viral matrix proteins including actins affect viral glycoprotein mobility and may be responsible for the temperature threshold phenomenon observed in Sendai virus fusion.

The compound DATEM inhibits respiratory syncytial virus fusion activity with epithelial cells

The effect of diacetyltartaric acid esters of mono and diglycerides (DATEM) on fusion of respiratory syncytial virus (RSV) with HEp-2 cells was studied using the R18 fluorescence dequenching fusion assay. At DATEM concentrations less than 2.0 microg/ml, the inhibition of fusion increased with the concentration of DATEM. At 2 microg/ml of DATEM, the fusion was suppressed by 80-90%. Studies examining possible mechanism of fusion-inhibition indicated that DATEM was likely adsorbed onto lipid membranes of both viral envelope and target cell membranes. Quantitative measurements of DATEM adsorption onto membranes were also performed using lipid monolayers and vesicles. The surface pressure of lipid monolayer formed at the air/aqueous interface increased as the concentration of DATEM in the monolayer subphase increased, suggesting that DATEM was inserted into the monolayer. As the concentration of DATEM in vesicle suspensions increased, electrophoretic mobility of initially uncharged lipid vesicles also increased, reflective of increased negative charge at vesicle surfaces. These results strongly suggest that the insertion of DATEM onto membranes inhibited viral fusion. DATEM may prove to be effective in limiting the infectivity of RSV by interference with the fusion of the viral envelope with target cell membranes.

Membrane fusion by an RGD-containing sequence from the core protein VP3 of hepatitis A virus and the RGA-analogue: implications for viral infection

The interaction of an RGD-containing epitope from the hepatitis A virus VP3 capsid protein and its RGA-analogue with lipid membranes was studied by biophysical methods. Two types of model membrane were used: vesicles and monolayers spread at the air/water interface, with a composition that closely resembles the lipid moiety of hepatocyte membranes: PC/SM/PE/PC (40:33:12:15; PC: 1-palmitoyl-2-oleoylglycero-sn-3-phosphocholine; SM: sphingomyelin from chicken egg yolk; PE, 1,2-dipalmitoyl-phosphatidylethanolamine; PS: L-alpha-phosphatidyl-L-serine from bovine brain). In addition, zwitterionic PC/SM/PE (47:39:14) and cationic PC/SM/PE/DOTAP (40:33:12:15; DOTAP: 1,2-dioleoyl-3-trimethylammonium-propane) membranes were also prepared in order to dissect the electrostatic and hydrophobic components in the interaction. Changes in tryptophan fluorescence, acrylamide quenching, and resonance energy transfer experiments in the presence of vesicles, as well as the kinetics of insertion in monolayers, indicate that both peptides bind to the three types of membrane at neutral and acidic pH; however, binding is irreversible only at low pH. Membrane-destabilizing and fusogenic activities are triggered by acidification at pH 4-6, characteristic of the endosome. Fluorescence experiments show that VP3-RGD and VP3-RGA induce mixing of lipids and leakage or mixing of aqueous contents in anionic and cationic vesicles at pH 4-6, indicating leaky fusion. Interaction with zwitterionic vesicles (PC/SM/PE) results in leakage without lipid mixing, indicating pore formation. Replacement of aspartic acid in the RGD motif by alanine maintains the membrane-destabilizing properties of the peptide at low pH, but not its antigenicity. Since the RGD tripeptide is related to receptor-mediated cell adhesion and antigenicity, results suggest that receptor binding is not a molecular requirement for fusion. The possible involvement of peptide-induced membrane destabilization in the mechanism of hepatitis A virus infection of hepatocytes by the endosomal route is discussed.

A new class of fusion-associated small transmembrane (FAST) proteins encoded by the non-enveloped fusogenic reoviruses

The non-enveloped fusogenic avian and Nelson Bay reoviruses encode homologous 10 kDa non-structural transmembrane proteins. The p10 proteins localize to the cell surface of transfected cells in a type I orientation and induce efficient cell-cell fusion. Mutagenic studies revealed the importance of conserved sequence-predicted structural motifs in the membrane association and fusogenic properties of p10. These motifs included a centrally located transmembrane domain, a conserved cytoplasmic basic region, a small hydrophobic motif in the N-terminal domain and four conserved cysteine residues. Functional analysis indicated that the extreme C-terminus of p10 functions in a sequence-independent manner to effect p10 membrane localization, while the N-terminal domain displays a sequence-dependent effect on the fusogenic property of p10. The small size, unusual arrangement of structural motifs and lack of any homologues in previously described membrane fusion proteins suggest that the fusion-associated small transmembrane (FAST) proteins of reovirus represent a new class of membrane fusion proteins.

Direct evidence that the N-terminal heptad repeat of Sendai virus fusion protein participates in membrane fusion

Recent studies have demonstrated the importance of heptad repeat regions within envelope proteins of viruses in mediating conformational changes at various stages of viral infection. However, it is not clear if heptad repeats have a direct role in the actual fusion event. Here we have synthesized, fluorescently labeled and functionally and structurally characterized a wild-type 70 residue peptide (SV-117) composed of both the fusion peptide and the N-terminal heptad repeat of Sendai virus fusion protein, two of its mutants, as well as the fusion peptide and heptad repeat separately. One mutation was introduced in the fusion peptide (G119K) and another in the heptad repeat region (I154K). Similar mutations have been shown to drastically reduce the fusogenic ability of the homologous fusion protein of Newcastle disease virus. We found that only SV-117 was active in inducing lipid mixing of egg phosphatidylcholine/phosphatidyiglycerol (PC/PG) large unilamellar vesicles (LUV), and not the mutants nor the mixture of the fusion peptide and the heptad repeat. Functional characterization revealed that SV-117, and to a lesser extent its two mutants, were potent inhibitors of Sendai virus-mediated hemolysis of red blood cells, while the fusion peptide and SV-150 were negligibly active alone or in a mixture. Hemagglutinin assays revealed that none of the peptides disturb the binding of virions to red blood cells. Further studies revealed that SV-117 and its mutants oligomerize similarly in solution and in membrane, and have similar potency in inducing vesicle aggregation. Circular dichroism and FTIR spectroscopy revealed a higher helical content for SV-117 compared to its mutants in 40 % tifluorethanol and in PC/PG multibilayer membranes, respectively, ATR-FTIR studies indicated that SV-117 lies more parallel with the surface of the membrane than its mutants. These observations suggest a direct role for the N-terminal heptad repeat in assisting the fusion peptide in mediating membrane fusion.

To fuse or not to fuse: the effects of electrostatic interactions, hydrophobic forces, and structural amphiphilicity on protein-mediated membrane destabilization

The development of lipid-based delivery vehicles for therapeutic molecules has become a topic of intense research. Recently, much of this effort has been directed towards mimicking the characteristics of viruses that give them an advantage for the delivery of genetic medicines. One of the most desirable properties of viral-based vectors is the ability to promote the destabilization of the host cell membrane to allow the entry of the genetic medicine into the target cell. This has been found to be largely controlled by the coat proteins on the surface of enveloped viruses. Although the exact mechanism by which proteins involved in the fusion process are able to promote the destabilization of membranes has yet to be elucidated, much understanding based upon information gained from a wide variety of studies is advancing the state of knowledge in this area. Parameters such as hydrophobic and electrostatic interactions as well as structural amphiphilicity, control to a large extent, the nature of the interaction of proteins with membranes. Thus, membrane fusion is mediated primarily by these forces acting in concert with one another. Ultimately, the knowledge gained from these studies will help to develop the ideal delivery system for the next generation of therapeutics.

Membrane fusion promoters and inhibitors have contrasting effects on lipid bilayer structure and undulations

It has been established that the fusion of both biological membranes and phospholipid bilayers can be modulated by altering their lipid composition (Chernomordik et al., 1995 .J. Membr. Biol. 146:3). In particular, when added exogenously between apposing membranes, monomyristoylphosphatidylcholine (MMPC) inhibits membrane fusion, whereas glycerol monoleate (GMO), oleic acid (OA), and arachidonic acid (AA) promote fusion. This present study uses x-ray diffraction to investigate the effects of MMPC, GMO, OA, and AA on the bending and stability of lipid bilayers when bilayers are forced together with applied osmotic pressure. The addition of 10 and 30 mol% MMPC to egg phosphatidylcholine (EPC) bilayers maintains the bilayer structure, even when the interbilayer fluid spacing is reduced to approximately 3 A, and increases the repulsive pressure between bilayers so that the fluid spacing in excess water increases by 5 and 15 A, respectively. Thus MMPC increases the undulation pressure, implying that the addition of MMPC promotes out-of-plane bending and decreases the adhesion energy between bilayers. In contrast, the addition of GMO has minor effects on the undulation pressure; 10 and 50 mol% GMO increase the fluid spacing of EPC in excess water by 0 and 2 A, respectively. However, x-ray diffraction indicates that, at small interbilayer separations, GMO, OA, or AA converts the bilayer to a structure containing hexagonally packed scattering units approximately 50 A in diameter. Thus GMO, OA, or AA destabilizes bilayer structure as apposing bilayers are brought into contact, which could contribute to their role in promoting membrane fusion.

Evidence for the extended phospholipid conformation in membrane fusion and hemifusion

Molecular-level mechanisms of fusion and hemifusion of large unilamellar dioleoyl phosphatidic acid/phosphocholine (DOPA/DOPC, 1:1 molar ratio) vesicles induced by millimolar Ca2+ and Mg2+, respectively, were investigated using fluorescence spectroscopy. In keeping with reduction of membrane free volume Vf, both divalent cations increased the emission polarization for 1,6-diphenyl-1,3, 5-hexatriene (DPH). An important finding was a decrease in excimer/monomer emission intensity ratio (Ie/Im) for the intramolecular excimer-forming probe 1, 2-bis[(pyren-1-)yl]decanoyl-sn-glycero-3-phosphocholine (bis-PDPC) in the course of fusion and hemifusion. Comparison with another intramolecular excimer-forming probe, namely, 1-[(pyren-1)-yl]decanoyl-2-[(pyren-1)-yl]tetradecanoyl-sn-gl ycero-3-p hosphocholine (PDPTPC), allowed us to exclude changes in acyl chain alignment to be causing the decrement in Ie/Im. As a decrease in Vf should increase Ie/Im for bis-PDPC and because contact site between adhering liposomes was required we conclude the most feasible explanation to be the adoption of the extended conformation (P.K.J., Chem. Phys. Lipids 63:251-258) by bis-PDPC. In this conformation the two acyl chains are splaying so as to become embedded in the opposing leaflets of the two adhered bilayers, with the headgroup remaining between the adjacent surfaces. Our data provide evidence for a novel mechanism of fusion of the lipid bilayers.

Lipid headgroup spacing and peptide penetration, but not peptide oligomerization, modulate peptide-induced fusion

In this study, the mechanism by which an amphipathic negatively charged peptide consisting of 11 amino acids (WAE) induces fusion of liposomal phosphatidylcholine membranes is investigated. WAE-induced fusion, which only occurs when the peptide is covalently attached to the bilayer, shows a highly remarkable dependence on naturally occurring phosphatidylcholine species. The initial rate of fusion increased in the order 1-palmitoyl 2-arachidonoyl PC (PAPC) > 1-palmitoyl 2-oleoyl PC (POPC) > 1-stearoyl 2-oleoyl PC (SOPC) > dioleoyl PC (DOPC) > egg yolk PC. Interestingly, the susceptibility of the various PC species toward WAE-induced fusion matched a similar order of increase in intrinsic lipid headgroup spacing of the target membrane. The degree of spacing, in turn, was found to be related to the extent by which the fluorescence quantum yield of the Trp residue increased, which occurred upon the interaction of WAE with target membranes. Therefore, these results demonstrate an enhanced ability for WAE to engage in hydrophobic interactions when headgroup spacing increases. Thus, this latter parameter most likely regulates the degree of penetration of WAE into the target membrane. Apart from penetrating, WAE oligomerizes at the site of fusion as revealed by monitoring the self-quenching of the fluorescently derivatized lipid anchor to which WAE is attached. Clustering appears specifically related to the process of membrane fusion and not membrane aggregation. This is indicated by the fact that fusion and clustering, but not aggregation, display the same strict temperature dependence. However, evidence is presented indicating that clustering is an accompanying event rather than a prerequisite for fusion. The notion that various biologically relevant fusion phenomena are accompanied by protein clustering and the specific PC-species-dependent regulation of membrane fusion emphasize the biological significance of the peptide in serving as a model for investigating mechanisms of protein-induced fusion.

Fusion between uninfected cells in retrovirus-induced fusion-from-within

We previously examined Moloney murine leukemia virus-induced fusion-from-within (FFWI) and fusion-from-without (FFWO) of SC-1 mouse cells. FFWI and FFWO can be distinguished by their stimulation by ionophores and polycations, respectively. FFWI is caused by infected cells. Normally, fusion between an infected cell and uninfected cells (heterofusions) is described, but we have surprisingly found that the infected cells also caused homofusion between uninfected cells in their vicinity (named neighbor homofusions). It was shown that neighbor homofusions were not induced by free virus particles (by FFWO). Transfectants expressing envelope proteins only induced heterofusions, indicating that virus production is necessary for the formation of neighbor homofusions. Both plasma membrane fragments and easily removable material from the surface of infected cells were able to induce fusion with the same stimulation pattern as FFWI and neighbor homofusion. These materials, especially the latter, have properties in common with virions, and it is discussed whether immature virions are involved in the formation of the neighbor homofusions.

Calcium-triggered selective intermembrane exchange of phospholipids by the lung surfactant protein SP-A

It is shown that human lung surfactant protein (SP-A) mediates selective exchange of phospholipid probes with unlabeled phospholipid in excess vesicles in the presence of calcium and NaCl. The exchange occurs without leakage of vesicle contents, or transbilayer movement (flip-flop) of the phospholipid probes, or fusion of vesicles. Individual steps preceding the exchange are dissected by a combination of protocols, and the results are operationally interpreted in terms of a model where a calcium-dependent change in SP-A triggers aggregation of vesicles followed by probe exchange between the vesicles in contact through SP-A. The contacts remain stable in the presence of calcium; i.e., the vesicles in contact do not change their partners on the time scale of several minutes. The binding of SP-A to vesicles and the aggregation of vesicles are rapid, and the aggregation is rapidly reversed by EGTA; i.e., both the forward and reverse aggregation reactions are complete in about 1 min. The exchange rate of the various probes between aggregated vesicles below 1 mM calcium in the presence of NaCl shows selectivity, i.e., a modest dependence on the net anionic charge on vesicles and for the headgroup of the probe. Exchange with lower selectivity is seen at >2 mM Ca in the absence of NaCl. SP-A binding to vesicles does not show an absolute specificity for the phospholipid structure, but the time course of the subsequent changes does. The results suggest that SP-A contacts between phospholipid interfaces could mediate the exchange of phospholipid species (trafficking and sorting) between lung surfactant pools in the hypophase and all accessible phospholipid interfaces of the alveolar space.

Membrane fusion is induced by a distinct peptide sequence of the sea urchin fertilization protein bindin

Fertilization in the sea urchin is mediated by the membrane-associated acrosomal protein bindin, which plays a key role in the adhesion and fusion between sperm and egg. We have investigated the structure/function relationship of an 18-amino acid peptide fragment "B18," which represents the minimal membrane binding motif of the protein and resembles a putative fusion peptide. The peptide was found to mimic the behavior of its parent protein bindin with respect to (a) its high affinity for lipid bilayers, (b) the ability to aggregate and fuse vesicles, (c) the binding of Zn2+ by a histidine-rich motif, (d) the tendency to self-assemble, and (e), as indicated earlier, the adhesion to cell surface polysaccharides. Fluorescence and light scattering assays were used here to monitor peptide-induced lipid mixing, leakage, and aggregation of large unilamellar sphingomyelin/cholesterol vesicles. For these activities, B18 requires the presence of Zn2+ ions, with which it forms oligomeric complexes and assumes a partially alpha-helical conformation, as observed by circular dichroism. We conclude that aggregation and fusion involves a "trans-complex" between peptides on apposing vesicles that are connected by Zn2+ bridges.

Probe transfer with and without membrane fusion in a fluorescence fusion assay

An analysis of the R18 fusion assay was made during the fusion of the Sendai virus with erythrocyte ghosts. The increase in R18 fluorescence, reflecting the interaction process, was evaluated in terms of the different processes that in principle may contribute to this increase, that is, monomeric probe transfer, hemifusion, and complete fusion. To this end, the kinetics of the R18-labeled lipid mixing were compared to those obtained with an assay in which the fusion-monitoring probe, eosin-maleimide, was attached to the viral surface proteins. The experiments relied on the use of native and fusion-inactive viruses and studies involving viral and target membranes that were modified by the incorporation of the lysophospholipid. The total dequenching signal detected in the R18 assay consists of components from probe transferred without fusion and from fusion itself. At 37 degrees C, the initial rate of dequenching (within two minutes) was predominately from the probe diluted by fusion with little contribution from transfer. The dequenching signal due to the probe transfer without fusion occurred at temperatures as low as 10 degrees C and increased linearly with time. Complete fusion started at about 20-25 degrees C and increased sharply at 30 degrees C. The extent of hemifusion was deduced from the total R18 dequenching data and those of the eosin-maleimide labeled protein dilution method for the limiting cases; the analysis indicates that hemifusion started at about 15 degrees C and increased over the range 20-25 degrees C. The initial rate of dequenching of the R18 assay measured within 2 min gives an accurate measure of membrane fusion above 30 degrees C.

Structural plasticity of the feline leukaemia virus fusion peptide: a circular dichroism study

The secondary structure of the feline leukaemia virus (FeLV) fusion peptide was investigated using circular dichroism (CD). Our results show that this peptide can readily flip between random, alpha-helical and beta-sheet conformations, depending upon its environment. The CD spectrum changes from one characteristic of random coil to predominantly beta-sheet type, and finally to that showing the characteristics of alpha-helical structure on moving from an aqueous solvent, through several increasingly hydrophobic systems, to a highly hydrophobic solvent. Electron microscopy confirmed the presence of beta structure. We propose that the structural plasticity demonstrated here is crucial to the ability of the fusion peptide to perturb lipid bilayers, and thus promote membrane fusion.

K+ ionophores stimulate retrovirus-induced fusion-from-within

Fusion-from-without (FFWO) and fusion-from-within (FFWI) for Moloney MLV in SC-1 cells were selectively stimulated by the polycation polybrene and the ionophore amphotericin B, which can be used to discriminate between the two fusion types. FFWI was stimulated by a number of different K+ ionophores. The stimulation occurred within few hours, did not require protein synthesis, and depended on the uninfected cell type, which suggests that FFWI is stimulated by a permutation of the K+ gradient of the uninfected cell. Plasma membrane vesicles from infected cells could also stimulate fusion with the same stimulation pattern as FFWI.

Membrane fusion induced by 11-mer anionic and cationic peptides: a structure-function study

We recently demonstrated that an amphipathic net-negatively charged peptide consisting of 11 amino acids (WAE 11) strongly promotes fusion of large unilamellar liposomes (LUV) when anchored to a liposomal membrane [Pecheur, E. I., Hoekstra, D., Sainte-Marie, J., Maurin, L., Bienvenue, A., and Philippot, J. R. (1997) Biochemistry 36, 3773-3781]. To elucidate a potential relationship between peptide structure and its fusogenic properties and to test the hypothesis that specific structural motifs are a prerequisite for WAE-induced fusion, three 11-mer WAE-peptide analogues (WAK, WAEPro, and WAS) were synthesized and investigated for their structure and fusion activity. Structural analysis of the synthetic peptides by infrared attenuated total reflection spectroscopy reveals a distinct propensity of each peptide toward a helical structure after their anchorage to a liposomal surface, emphasizing the importance of anchorage on conveying a secondary structure, thereby conferring fusogenicity to these peptides. However, whereas WAE and WAK peptides displayed an essentially nonleaky fusion process, WAS- and WAEPro-induced fusion was accompanied by substantial leakage. It appears that peptide helicity as such is not a sufficient condition to convey optimal fusion properties to these 11-mer peptides. Studies of changes in the intrinsic Trp fluorescence and iodide quenching experiments were carried out and revealed the absence of migration of the Trp residue of WAS and WAEPro to a hydrophobic environment, upon their interaction with the target membranes. These results do not support the penetration of both peptides as their mode of membrane interaction and destabilization but rather suggest their folding along the vesicle surface, posing them as surface-seeking helixes. This is in striking contrast to the behavior observed for WAE and WAK, for which at least partial penetration of the Trp residue was demonstrated. These results indicate that subtle differences in the primary sequence of a fusogenic peptide could induce dramatic changes in the way the peptide interacts with a bilayer, culminating in equally drastic changes in their functional properties. The data also reveal a certain degree of sequence specificity in WAE-induced fusion.

Interaction of the influenza hemagglutinin fusion peptide with lipid bilayers: area expansion and permeation

Fusion is a crucial event in the infection of animal cells by enveloped viruses (e.g., HIV or influenza). Viral fusion is mediated by glycoproteins, spanning the viral envelope, which attach to a membrane surface and induce fusion of the viral envelope to the cellular membrane. Influenza fusion protein (hemagglutinin) contains an amino-terminal segment critical to fusion, referred to as the fusion peptide. We show here that the native fusion peptide (wt-20) of hemagglutinin destabilizes membranes formed of 99% 1 -stearoyl-2-oleoylphosphatidylcholine (SOPC). The first step in destabilization is rapid insertion of the peptide into the membrane, in which membrane area increases by as much as 11% in just seconds. We visualized and quantified the area expansion by using optical video microscopy combined with micropipette aspiration. This rapid membrane area expansion is followed by the formation of membrane defects in the size range of 0.5 nm, and results in membrane rupture. Both the rate of area increase and maximum area increase are significantly higher at a pH near 5.0 compared to pH 7.0. These results suggest that enhanced membrane insertion of wt-20 and accompanying area expansion at pH 5.0 are responsible for the relatively greater lytic activity at this pH. We show that a deletion of the N-terminal glycine of wt-20 results in a lack of area expansion or membrane perturbation at pH 5.0.

Protein involvement in the fusion between the equatorial segment of acrosome-reacted human spermatozoa and liposomes

Artificial membranes (liposomes) can interact with the equatorial segment (ES) of human spermatozoa, provided that the acrosome reaction (AR) has occurred [Arts, Kuiken, Jager and Hoekstra (1993) Eur. J. Biochem. 217, 1001-1009]. Using fluorescently labelled liposomes, this interaction can be seen as either punctate fluorescence in the ES (lip-ESp), reflecting only bound liposomes, or as diffuse fluorescence in this region (lip-ESd), indicating that the liposomes have fused with the ES membrane. Only equatorial segments that still contain constituents of the acrosomal matrix have the capacity to bind liposomes and eventually to fuse with them. Since the exposure of such intact equatorial segments is the exclusive result of induction of the AR under physiological conditions, these results imply that liposomes can be used for the rapid detection of acrosome-reacted spermatozoa. The lip-ESp and lip-ESd patterns were shown to be reflections of two distinct properties of the ES. Proteolytic treatment after AR completely inhibited the formation of a lip-ESd pattern, whereas formation of the lip-ESp pattern was only marginally inhibited by the proteolytic treatment. The same results were obtained using anti-sperm antibodies, which did not react with acrosome-intact spermatozoa. Proteolytic treatment of spermatozoa before AR induction had no effect on the fusion capacity of the ES after subsequent AR, which implies that the putative fusion protein is not accessible before AR. Thus fusion of liposomes with the ES of human spermatozoa is mediated by a sperm protein(s), whereas the lip-ESp pattern is not likely to represent the liposome-binding stage that precedes the fusion step.

Fusion peptides derived from the HIV type 1 glycoprotein 41 associate within phospholipid membranes and inhibit cell-cell Fusion. Structure-function study

The fusion domain of human immunodeficiency virus (HIV-1) envelope glycoprotein (gp120-gp41) is a conserved hydrophobic region located at the N terminus of the transmembrane glycoprotein (gp41). A V2E mutant has been shown to dominantly interfere with wild-type envelope-mediated syncytium formation and virus infectivity. To understand this phenomenon, a 33-residue peptide (wild type, WT) identical to the N-terminal segment of gp41 and its V2E mutant were synthesized, fluorescently labeled, and characterized. Both peptides inhibited HIV-1 envelope-mediated cell-cell fusion and had similar alpha-helical content in membrane mimetic environments. Studies with fluorescently labeled peptide analogues revealed that both peptides have high affinity for phospholipid membranes, are susceptible to digestion by proteinase-K in their membrane-bound state, and tend to self- and coassemble in the membranes. In SDS-polyacrylamide gel electrophoresis the WT peptide formed dimers as well as higher order oligomers, whereas the V2E mutant only formed dimers. The WT, but not the V2E mutant, induced liposome aggregation, destabilization, and fusion. Moreover, the V2E mutant inhibited vesicle fusion induced by the WT peptide, probably by forming inactive heteroaggregates. These data form the basis for an explanation of the mechanism by which the gp41 V2E mutant inhibits HIV-1 infectivity in cells when co-expressed with WT gp41.

Membrane anchorage brings about fusogenic properties in a short synthetic peptide

The fusogenic properties of an amphipathic net-negative peptide (wae 11), consisting of 11 amino acid residues, were studied. We demonstrate that, whereas the free peptide displays no significant fusion activity, membrane fusion is strongly promoted when the peptide is anchored to a liposomal membrane. The fusion activity of the peptide appears to be independent of pH, and membrane merging is an essentially nonleaky process. Thus, the extents of lipid mixing and contents mixing were virtually indistinguishable. Vesicle aggregation is a prerequisite for fusion. For this process to take place, the target membranes required a positive charge which was provided by incorporating lysine-coupled phosphatidylethanolamine (PElys). The coupled peptide, present in one population, could thus cause vesicle aggregation via nonspecific electrostatic interaction with PElys. However, the free peptide failed to induce aggregation of PElys vesicles, suggesting that the spatial orientation of the coupled peptide codetermined its ability to bring about vesicle aggregation and fusion. With the monitoring of changes in the intrinsic Trp fluorescence, in conjunction with KI-quenching studies, it would appear that hydrophobic interactions facilitate the fusion event, possibly involving (partial) peptide penetration. Such a penetration may be needed to trigger formation of a transient, nonbilayer structure. Since lysophosphatidylcholine inhibited while monoolein strongly stimulated peptide-induced fusion, our data indicate that wae 11-induced fusion proceeds according to a model consistent with the stalk-pore hypothesis for membrane fusion.

Salt-triggered intermembrane exchange of phospholipids and hemifusion by myelin basic protein

Intervesicle phospholipid exchange through molecular contacts induced by the C1 molecular species of myelin basic protein (MBP) are characterized by using methods that amplify the effect of MBP-membrane interaction. The effect of salt concentration (KCl) on the vesicle-vesicle interaction of anionic sonicated covesicles of 30% 1-palmitoyl-2-oleoylglycero-sn-3-phosphocholine and 70% 1,2-dimyristoylglycero-sn-3-phosphomethanol (POPC/DMPM) by MBP is dissected by a combination of protocols into individual steps: aggregation of vesicles, apposition and contact formation, and hemifusion. Scattering and resonance energy transfer measurements reveal that, in the absence of KCl, MBP promotes rapid aggregation of the vesicles without lipid mixing. At >40 mM KCl, the extent of aggregation is larger and time-dependent. Fluorescence dequenching due to dilution of labeled phospholipids indicates that on a somewhat slower time scale, hemifusion of vesicles is triggered by salt, with mixing of the outer monolayer lipids but without flip-flop of phospholipids and without mixing or leakage of the aqueous contents. The exchange and hemifusion are seen with anionic vesicles; the effect of the structure of phospholipid, composition of vesicles, and the protein/lipid ratio is primarily on the kinetics of these and other competing processes. Thus, at 0.022 mol % of MBP and less than 100 mM KCl, it is possible to uncouple three sequential steps: (1) aggregation of vesicles by MBP; (2) apposition of bilayers and selective lipid exchange through vesicle-vesicle contacts established by MBP, i.e., anionic and zwitterionic phospholipids exchange, but cationic probes are excluded; and (3) hemifusion and lipid mixing of contacting monolayers of vesicles.

Detection and quantification of asymmetric lipid vesicle fusion using deuterium NMR

It is demonstrated that deuterium nuclear magnetic resonance (2H NMR) spectroscopy can be used to detect and to quantify fusion between anionic giant unilamellar vesicles (GUVs) and cationic small unilamellar vesicles (SUVs). The sensitivity to fusion relies on the conformational response of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) to changes in membrane surface electrostatic charge due to lipid mixing upon fusion. This conformational change is reported in the 2H NMR spectrum as a change in the quadrupolar splitting from choline-deuterated POPC. GUVs were composed of varying molar ratios of the anionic lipid 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPG), plus cholesterol (CHOL), plus POPC. SUVs were composed of the cationic lipid 1,2-dioleoyloxy-3-(dimethylammonio)-propane (DODAP), plus POPC with or without 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE). Using a quantitative model that relates the 2H NMR quadrupolar splitting to the mole fractions of cationic, anionic, and neutral lipids in the vesicle membrane, it was possible to deduce the extent of fusion between the two oppositely-charged vesicle populations directly from the quadrupolar splitting. SUVs composed of DODAP + POPC + POPE (40/40/20) fused 100% with GUVs composed of POPC + CHOL + POPG (60/30/10). Removing POPE from the SUVs reduced the extent of fusion, as did reducing the POPG content of the GUVs.

Lysis of Large Unilamellar Vesicles Induced by Analogs of the Fusion Peptide of Influenza Virus Hemagglutinin

We have developed a micropipette aspiration assay to observe the lysis of large (20-30 &mgr;m diameter) vesicles aspirated using micropipettes. Single membrane lysis events can be seen with the light microscope and are followed using fluorescence assays and video microscopy. In this study, we have examined the ability of two analogs of the fusion peptide from influenza virus hemagglutinin to induce the lysis of large unilamellar egg phosphatidylcholine vesicles, as a function of peptide concentration and pH. X31 is a wild-type peptide from one strain of Influenza A, and E5 is an analogue which has several residues replaced by glutamate residues. Both peptides were found to induce lysis of large vesicles in a pH-dependent manner. Both peptides exhibited maximal activity at pH 5, measured in terms of both rate and extent of lysis. E5 was active at much lower concentrations than X31. Our results with both peptides are compared to results published from other laboratories.

The fusion pore and mechanisms of biological membrane fusion

Membrane fusion occurs as part of processes as different as synaptic neurotransmitter transmission and infection with influenza virus. Recent evidence paints a picture in which the organization of proteins into a macromolecular scaffold brings the two fusing membranes together and induces hemifusion, that is, the fusion of the apposing leaflets of the two membranes to form a common bilayer. A small dynamic fusion pore forms in the common bilayer and usually expands to allow complete membrane merging. The mechanisms of fusion appear to be remarkably similar in exocytosis and virus-induced fusion. During exocytotic fusion, there is an additional twist to the mechanism, as sometimes the fusion pores close after release of small non-quantal amounts of secretory products.

Specificity for the exchange of phospholipids through polymyxin B mediated intermembrane molecular contacts

Structural specificity for the direct vesicle-vesicle exchange of phospholipids through stable molecular contacts formed by the antibiotic polymyxin B (PxB) is characterized by kinetic and spectroscopic methods. As shown elsewhere [Cajal, Y., Rogers, J., Berg, O.G., & Jain, M.K. (1996) Biochemistry 35, 299-308], intermembrane molecular contacts between anionic vesicles are formed by a small number of PxB molecules, which suggests that a stoichiometric complex may be responsible for the exchange of phospholipids. Larger clusters containing several vesicles are formed where each vesicle can make multiple contacts if sterically allowed. In this paper we show that the overall process can be dissected into three functional steps: binding of PxB to vesicles, formation of stable vesicle-vesicle contacts, and exchange of phospholipids. Polycationic PxB binds to anionic vesicles. Formation of molecular contacts and exchange of monoanionic phospholipids through PxB contacts does not depend on the chain length of the phospholipid. Only monoanionic phospholipids (with methanol, serine, glycol, butanol, or phosphatidylglycerol as the second phosphodiester substituent in the head group) exchange through these contacts, whereas dianionic phosphatidic acid does not. Selectivity for the exchange was also determined with covesicles of phosphatidylmethanol and other phospholipids. PxB does not bind to vesicles of zwitterionic phosphatidylcholine, and its exchange of covesicles is not mediated by PxB. Vesicles of dianionic phospholipids, like phosphatidic acid, bind PxB; however, this phospholipid does not exchange. The structural features of the contacts are characterized by the spectroscopic and chemical properties of PxB at the interface. PxB in intermembrane contacts is readily accessible from the aqueous phase to quenchers and reagents that modify amino groups. Results show that PxB at the interface can exist in two forms depending on the lipid/PxB ratio. Additional studies show that the stable PxB-mediated vesicle-vesicle contacts may be structurally and functionally distinct from "stalks", the putative transient intermediate for membrane fusion. The phenomenon of selective exchange of phospholipids through peptide-mediated contacts could serve as a prototype for intermembrane targeting and sorting of phospholipids during their biosynthesis trafficking in different compartments of a cell. The protocols and results described here also extend the syllogistic foundation in interfacial equilibria and catalysis.

Intermembrane molecular contacts by polymyxin B mediate exchange of phospholipids

Direct intermembrane exchange of dimyristoylphosphatidylmethanol is mediated by polymyxin B (PxB), a cationic amphipathic cyclic decapeptide. The possibility that the phospholipid exchange is mediated by solubilization of phospholipids or by fusion of vesicles is ruled out. By kinetic and spectroscopic methods it is shown that the exchange occurs directly through vesicle-vesicle contacts formed by a few PxB molecules. The contact is stable on the time scale of several minutes such that neither PxB nor the vesicles in the pair forming a contact exchange with excess vesicles. Several contacts may be formed on a vesicle, which leads to the formation of a cluster of vesicles, and the lipid molecules on the outer monolayers of vesicles exchange throughout the cluster. Kinetics of substrate replenishment during processive interfacial catalysis suggests that the exchange of anionic lipids over the contact occurs at a rate considerably faster than 300 s-1. The exchange through the contact is specific for certain lipids, and phospholipids with a modified head group or phospholipase A2 bound to a vesicle are not transferred to the other vesicle in contact. Since this phenomenon has not been described before, possible implications of direct vesicle-vesicle exchange of phospholipids through peptide-mediated molecular contacts are discussed. Such a mechanism for intermembrane transfer of phospholipids could be responsible for intracellular trafficking and sorting of phospholipids; it could be a necessary first step for the sequence of events leading to budding, vesiculation, and secretion; and PxB-mediated transfer between the inner and outer membranes of Gram-negative bacteria could also account for its antibiotic action.

Rotavirus-induced fusion from without in tissue culture cells

We present the first evidence of fusion from without induced in tissue culture cells by a nonenveloped virus. Electron micrographs of two strains of rotavirus, bovine rotavirus C486 and rhesus rotavirus, show that virally mediated cell-cell fusion occurs within 1 h postinfection. Trypsin activation is necessary for rotavirus to mediate cell-cell fusion. The extent of fusion is relative to the amount of virus used, and maximum fusion occurs between pHs 6.5 and 7.5. Fusion does not require virus-induced protein synthesis, as virus from both an empty capsid preparation and from an EDTA-treated preparation, which is noninfectious, can induce fusion. Incubation of rotavirus with neutralizing and nonneutralizing monoclonal antibodies before addition to cells indicates that viral protein 4 (VP4; in the form of VP5* and VP8*) and VP7 are involved in fusion. Light and electron micrographs document this fusion, including the formation of pores or channels between adjacent fused cells. These data support direct membrane penetration as a possible route of infection. Moreover, the assay should be useful in determining the mechanisms of cell entry by rotavirus.

Characterization of retrovirus-induced SC-1 cell fusion

Virus-mediated cell-cell fusion with Moloney MLV and SC-1 cells was characterized. The level of fusion was highly dependent on the cell line used for propagation of the virus. Efficient fusion appeared to be very sensitive to negative charges on the cell surface and surroundings. Addition of polycations, removal of serum, and treatment with neuraminidase or hyaluronidase all stimulated fusion. Conversely, fusion was inhibited by fibronectin. Kinetic results and the time of action of inhibitors indicated that virus particles (or virus material) on the cell surface lead directly to fusion. The fusion then proceeded rapidly and required actin movement as shown by cytochalasin inhibition.

Lateral inhomogeneous lipid membranes: theoretical aspects

The physical concepts underlying the lateral distribution of the components forming a lamellar assembly of amphiphiles are discussed in this review. The role of amphiphiles' molecular structure and/or aqueous environment (ionic strength, water soluble substances) on formation and stability of lateral patterns is investigated. A considerable effort is devoted to the analysis of the properties of patterned structure which can be different from those of randomly mixed multi-component lamellae. Examples include adhesion and fusion among laterally inhomogeneous bilayers, enhanced interfacial adsorption of ions and polymers, enhanced transport across the bilayer, modified mechanical properties, local stabilization of non-planar geometries (pores, edges) and related phenomena (electroporation, budding transition and so on). Furthermore, an analysis of chemical reactivity within or at the water interface of a laterally inhomogeneous bilayer is briefly discussed. A link between these concepts and experimental findings taken from the biological literature is attempted throughout the review.

Micropipette manipulation technique for the monitoring of pH-dependent membrane lysis as induced by the fusion peptide of influenza virus

We have assembled a micropipette aspiration assay to measure membrane destabilization events in which large (20-30 microns diameter) unilamellar vesicles are manipulated and exposed to membrane destabilizing agents. Single events can be seen with a light microscope and are recorded using both a video camera and a photomultiplier tube. We have performed experiments with a wild-type fusion peptide from influenza virus (X31) and found that it induces pH-dependent, stochastic lysis of large unilamellar vesicles. The rate and extent of lysis are both maximum at pH 5; the maximum rate of lysis is 0.018 s-1 at pH 5. An analysis of our data indicates that the lysis is not correlated either to the size of the vesicles or to the tension created in the vesicle membranes by aspiration.

Design of a short membrane-destabilizing peptide covalently bound to liposomes

We characterized the physical and biological properties of a 14-residue amphipathic sequence called SFP (for short fusogenic peptide). At acidic pH, this short synthetic peptide interacts with various phospholipidic monolayers. These interactions were correlated with a pH-dependent conformational transition of SFP resulting in a hydrophobic alpha-helical structure. The hemolysis assay showed a pH-dependent weak membrane destabilizing activity of SFP. However, membrane anchoring of SFP through a covalently bound myristic acid enhanced by 1000-fold its membrane-destabilizing power. Moreover, SFP covalently bound to fluorescent-labeled liposomes induced a pH-dependent mixing of both membranes. SFP, a small synthetic peptide, is thus able to mimick many aspects of viral protein-induced membrane fusion: conformational change, membrane destabilization, membrane anchoring and finally pH-dependent lipid mixing.

Induction of cell fusion in cultured fibroblasts and epithelial cells by microinjection of EGTA, GTP gamma S and antifodrin antibodies

CaCl2, EGTA, GTP gamma S and anti-alpha-fodrin antibodies were injected into fibroblast-like IMR-33 cells and Madin-Darby bovine kidney (MDBK) epithelial cells cultured both in the presence and absence of cycloheximide and fetal calf serum. EGTA, GTP gamma S antifodrin antibody induced fusion of MDBK cells within one hour after injection. The cells formed polykaryons with up to 15 nuclei, reaching an average fusion index of 20%. IMR-33 cells fused at a slower kinetics and only upon injection of GTP gamma S or antifodrin antibodies. No fusions were seen in serum-deprived, quiescent cells. On the other hand, cycloheximide treatment did not prevent the fusions. The results show that cells can be induced to fuse by using agents that interfere with the regulation of the G-proteins, intracellular calcium level or membrane skeleton. We suggest that the putative fusogens are resident proteins of the plasma membrane which become exposed upon destabilization of the membrane skeleton.

Nucleic acid transfer through cell membranes: towards the underlying mechanisms

Various cases of DNA (RNA) transfer through membranes of living cells are reviewed. They are classified into two major categories: those which occur in Nature (natural transfer) and those imposed by various physical and chemical treatments of cells (induced transfer). Among the examples of natural transfer surveyed are the transfer during bacterial conjugation, genetic transformation, viral infection of bacteria, and nuclear membrane trafficking. Consideration of the induced transfer is focused on the two methods most widely used at present to introduce foreign genetic information into pro- and eukaryotic cells: Ca2+ (and some other divalent cations)-induced and calcium phosphate-induced transfer, and transfer during electroporation of cells. Emphasis is made on the underlying mechanisms of transfer, or rather on what is currently known about them. Energetic aspects of transfer are also discussed and different tentative models of transfer are presented.

Fusion of artificial membranes with mammalian spermatozoa. Specific involvement of the equatorial segment after acrosome reaction

The fusogenic properties of bovine and human spermatozoa membranes were investigated, using phospholipid bilayers (liposomes) as target membranes. Fusion was monitored by following lipid mixing, as revealed by an assay based on resonance-energy transfer. In addition, fusion was visualized by fluorescence microscopy, using fluorescent lipid vesicles. Cryopreserved bovine sperm fused with liposomes before induction of the acrosome reaction, fluorescence being located in essentially all spermatozoa membrane domains. Fresh bovine and human spermatozoa fused with liposomes only after the induction of the acrosome reaction, as triggered by calcium ionophore A23187 or zonae pellucidae (proteins), while the fluorescence distribution was mainly restricted to the equatorial segment (ES). However, with spermatozoa that had undergone a freeze/thawing cycle, domains other than ES also became labeled. Hence, the redistribution of the lipid probes over the entire membrane occurring during lipid mixing with cryopreserved bovine sperm is probably related to membrane perturbations caused by long-term cryopreservation. Fusion with liposomes was governed by spermatozoa factors and required the presence of acidic phospholipids like cardiolipin and phosphatidylserine in the liposomal bilayer. Incorporation of the zwitterionic lipid phosphatidylcholine in the vesicles inhibited the fusion reaction. Fusion was pH dependent. The results indicate that the ES is the primary domain of spermatozoa membranes that harbours the fusogenic capacity of sperm. Liposomes appear a valuable tool in further characterizing the properties of this domain, which has been claimed [Yanagimachi, R. (1988) in The physiology of reproduction (Knobil, E. & Neill, J., eds) pp. 135-185, Raven Press, New York] to represent the putative, initial fusion site for the oocyte.

Orientation of fusion-active synthetic peptides in phospholipid bilayers: determination by Fourier transform infrared spectroscopy

A group of synthetic peptides having an amino acid sequence related to the N-terminal region of the influenza virus hemagglutinin HA-2 chain can induce phospholipid membrane fusion in a pH-dependent manner. These peptides bind to membranes to form alpha-helices even at pH's where no fusion activity is seen. We determined the orientation of these alpha-helical peptides in lipid multibilayers using attenuated total reflection infrared spectroscopy and found that the peptide alpha-helices took a preferential orientation, the helix axis being about 70 degrees from the normal of the membrane plane, or in other words rather parallel to the membrane plane. The orientation was almost independent of pH and a modification of the N-terminal amino group which reduced the fusion activity of the peptides. The determination was carried out for peptides in lipid multibilayers in dry or hydrated (membranes equilibrated with D2O vapor) conditions. Although a slight decrease in the helix orientation angle from the membrane normal was noticed for a hydrated system, the difference between the results for dry and hydrated conditions was small.

Defensins promote fusion and lysis of negatively charged membranes

Defensins, a family of cationic peptides isolated from mammalian granulocytes and believed to permeabilize membranes, were tested for their ability to cause fusion and lysis of liposomes. Unlike alpha-helical peptides whose lytic effects have been extensively studied, the defensins consist primarily of beta-sheet. Defensins fuse and lyse negatively charged liposomes but display reduced activity with neutral liposomes. These and other experiments suggest that fusion and lysis is mediated primarily by electrostatic forces and to a lesser extent, by hydrophobic interactions. Circular dichroism and fluorescence spectroscopy of native defensins indicate that the amphiphilic beta-sheet structure is maintained throughout the fusion process. Taken together, these results support the idea that protein-mediated membrane fusion depends not only on hydrophobic and electrostatic forces but also on the spatial arrangement of the amino acid residues to form a three-dimensional amphiphilic structure, which promotes the efficient mixing of the lipids between membranes. A molecular model for membrane fusion by defensins is presented, which takes into account the contributions of electrostatic forces, hydrophobic interactions, and structural amphiphilicity.

Interaction of myelin basic protein with artificial membranes. Parameters governing binding, aggregation and dissociation

The interaction of myelin basic protein (MBP) with large unilamellar vesicles, composed of phosphatidylserine (PtdSer), phosphatidylserine/phosphatidylcholine (PtdSer/Ole2GroPCho) and phosphatidylcholine/cholesterol (Ole2GroPCho/cholesterol) was examined. Binding of MBP to the bilayers as well as the kinetics of this process were determined by a resonance energy transfer procedure. The ability of the protein to aggregate the vesicles subsequently was monitored continuously by absorbance measurements. The interaction was further characterized by determining the ability of MBP to induce membrane perturbations, as reflected by release of aqueous vesicle contents, and lipid mixing. The results demonstrate that Ole2GroPCho inhibits, while PtdSer and cholesterol strongly facilitate MBP-induced membrane aggregation. Furthermore, binding of MBP to vesicles and the subsequent aggregation event are separate processes, i.e. the extent of binding does not necessarily reflect the aggregation susceptibility. Overall, aggregation appears to be the rate-limiting step. Interaction of MBP with PtdSer bilayers results in a limited degree of lipid mixing, which is accompanied by extensive release of vesicle contents. For all other compositions, no lipid mixing occurs, while cholesterol effectively prevents release of vesicle contents. pH-dependent experiments indicate distinct mechanisms to be operative in MBP-induced aggregation of PtdSer and Ole2GroPCho/cholesterol bilayers. At neutral pH, protein-protein interactions appear relevant, while at acidic pH intervesicular bridges, established by monomers that may cause aggregation of PtdSer vesicles, but not of Ole2GroPCho/cholesterol vesicles. The observation that divalent cations reverse MBP-induced vesicle aggregation may have physiological relevance.