An osmotic model for the fusion of biological membranes

Effect of Osmotic Pressure on the Stability of Whole Inactivated Influenza Vaccine for Coating on Microneedles

Enveloped virus vaccines can be damaged by high osmotic strength solutions, such as those used to protect the vaccine antigen during drying, which contain high concentrations of sugars. We therefore studied shrinkage and activity loss of whole inactivated influenza virus in hyperosmotic solutions and used those findings to improve vaccine coating of microneedle patches for influenza vaccination. Using stopped-flow light scattering analysis, we found that the virus underwent an initial shrinkage on the order of 10% by volume within 5 s upon exposure to a hyperosmotic stress difference of 217 milliosmolarity. During this shrinkage, the virus envelope had very low osmotic water permeability (1 - 6×10-4 cm s-1) and high Arrhenius activation energy (Ea = 15.0 kcal mol-1), indicating that the water molecules diffused through the viral lipid membranes. After a quasi-stable state of approximately 20 s to 2 min, depending on the species and hypertonic osmotic strength difference of disaccharides, there was a second phase of viral shrinkage. At the highest osmotic strengths, this led to an undulating light scattering profile that appeared to be related to perturbation of the viral envelope resulting in loss of virus activity, as determined by in vitro hemagglutination measurements and in vivo immunogenicity studies in mice. Addition of carboxymethyl cellulose effectively prevented vaccine activity loss in vitro and in vivo, believed to be due to increasing the viscosity of concentrated sugar solution and thereby reducing osmotic stress during coating of microneedles. These results suggest that hyperosmotic solutions can cause biphasic shrinkage of whole inactivated influenza virus which can damage vaccine activity at high osmotic strength and that addition of a viscosity enhancer to the vaccine coating solution can prevent osmotically driven damage and thereby enable preparation of stable microneedle coating formulations for vaccination.

Effects of newcastle disease virus glycoproteins on the structural and thermal behaviour of 1,2-dihexadecyl-sn-glycero-3-phosphatidylcholine lipid membranes under osmotic stress conditions

The interaction of hem agglutininneuraminidase (HN) and fusion (F) glycoproteins with swollen vesicles of 1,2-dihexadecyl-sn-glycero-3-phosphatidylcholine (DHPC) was investigated under transition from gel to fluid phase. X-ray studies of the structure of lipid/HN-F mixtures in normal and swollen vesicles have shown that the lamellar bilayer structure predominate in the gel and liquid crystalline phases. A swollen lipid phase, in which the mean repeat distance of lipid bilayers is larger than in the other phases was found. The nature of this phase is similar to the anomalous bilayer swelling reported in literature. The presence of HN and F in the vesicles led to the coexistence of structures with low and high lamellar order, showing larger repeat distance in comparison with the pure lipid. This finding was attributed to the increase in the lipid bilayer thickness due to the HN-F included in the free water layer. The thermal behaviour of the system was not affected by the vesicle swelling. The data showed the existence of gel and liquid crystalline lamellar phases and changes in lipid/HN-F specific heats, mainly due to the concentration effect of the HN-F and its location in the free water layer.

Staurosporine-inhibitable protein kinase activity associated with secretory granule membranes isolated from rat submandibular gland cells

Protein kinases, such as protein kinase C, have been shown to be associated with secretory granules and to regulate the event of exocytosis in various tissues including parotid salivary acinar cells. However, in submandibular acinar cells that play an important role in the secretion of proteins into the oral cavity, kinase activity on the granule membrane has not been explored. Therefore, in the present study, we isolated the secretory granules from rat submandibular acinar cells and investigated the localisation of protein kinases on the granule membrane. Initially, we isolated and purified secretory granules from rat submandibular acinar cells. Addition of [gamma-32P] ATP to granule-membrane lysate phosphorylated the granule-membrane-associated 26, 32, 55 and 58kDa proteins, suggesting the presence of endogenous kinase activity on the membrane. Moreover, the phosphorylation of 26 and 32kDa proteins was inhibited by staurosporine and K252a, both non-specific protein kinase C inhibitors. However, the phosphorylation of 26 and 32kDa proteins was not inhibited by other protein kinase C inhibitors, such as calphostin C, GF109203X and chelerythrine, indicating that protein kinase C was not responsible for the phosphorylation. In addition, H-89, ML-9, KN-62 and genistein did not appear to inhibit this phosphorylation, indicating that protein kinase A, myosin light chain kinase (MLCK), Ca2+/calmodulin-dependent protein kinase II (CAMKII) and tyrosine kinase were not involved in the phosphorylation of 26 and 32kDa proteins. Moreover, Ca2+ had no effect on the kinase activity. Therefore, our results suggest that an unidentified, staurosporine-inhibitable protein kinase activity is associated with the secretory granule membrane of rat submandibular acinar cells.

Staurosporine mobilizes Ca(2+) from secretory granules by inhibiting protein kinase C in rat submandibular acinar cells

Staurosporine was previously shown to mobilize Ca(2+) from the thapsigargin-insensitive Ca(2+) store in rat submandibular acinar cells. However, the nature of the store is not yet known. Therefore, in the present study, the staurosporine-releasable intracellular Ca(2+) store was characterized. Staurosporine increased the cytosolic Ca(2+) concentration ([Ca(2+)](c)) after the inositol 1,4,5-trisphosphate (IP(3))-sensitive Ca(2+) store was depleted. Ionomycin caused only small increases in [Ca(2+)](c) after the depletion of the IP(3)-sensitive Ca(2+) store, whereas ionomycin+monensin caused large increases. However, ionomycin+monensin did not increase [Ca(2+)](c) when added after [Ca(2+)](c) was increased by staurosporine, indicating that the acidic Ca(2+) store was the main source of Ca(2+). The acidic Ca(2+) store appeared to be associated with secretory granules, since ionomycin+monensin- and staurosporine-induced [Ca(2+)](c) increases were significantly reduced when the acinar cells were degranulated. The effect of staurosporine on [Ca(2+)](c) was mimicked by other protein kinase C inhibitors. Therefore, we conclude that staurosporine mobilizes Ca(2+) from secretory granules, probably through the inhibition of protein kinase C in rat submandibular acinar cells.

The transmembrane domain of influenza hemagglutinin exhibits a stringent length requirement to support the hemifusion to fusion transition

Glycosylphosphatidylinositol-anchored influenza hemagglutinin (GPI-HA) mediates hemifusion, whereas chimeras with foreign transmembrane (TM) domains mediate full fusion. A possible explanation for these observations is that the TM domain must be a critical length in order for HA to promote full fusion. To test this hypothesis, we analyzed biochemical properties and fusion phenotypes of HA with alterations in its 27-amino acid TM domain. Our mutants included sequential 2-amino acid (Delta2-Delta14) and an 11-amino acid deletion from the COOH-terminal end, deletions of 6 or 8 amino acids from the NH(2)-terminal and middle regions, and a deletion of 12 amino acids from the NH(2)-terminal end of the TM domain. We also made several point mutations in the TM domain. All of the mutants except Delta14 were expressed at the cell surface and displayed biochemical properties virtually identical to wild-type HA. All the mutants that were expressed at the cell surface promoted full fusion, with the notable exception of deletions of >10 amino acids. A mutant in which 11 amino acids were deleted was severely impaired in promoting full fusion. Mutants in which 12 amino acids were deleted (from either end) mediated only hemifusion. Hence, a TM domain of 17 amino acids is needed to efficiently promote full fusion. Addition of either the hydrophilic HA cytoplasmic tail sequence or a single arginine to Delta12 HA, the hemifusion mutant that terminates with 15 (hydrophobic) amino acids of the HA TM domain, restored full fusion activity. Our data support a model in which the TM domain must span the bilayer to promote full fusion.

Monensin and hypo-osmolar medium cause calcium-independent beta-endorphin secretion from melanotropes

Monensin has been shown to cause nonexocytotic release of catecholamines from adrenal medullary and PC12 cells. We examined the effect of monensin on peptide secretion with cultured melanotropes from the rat pituitary as a model. 1 microM monensin caused an immediate, transient increase in beta-endorphin secretion. The effect was still seen in a calcium-free medium, but was totally abolished in a sodium-free medium. Intracellular calcium concentration was measured with Fura 2: no increase was observed during monensin stimulation. Hypo-osmolar medium mimicked the effect of monensin, causing a 12-fold transient increase in beta-endorphin secretion. This effect was not abolished in either calcium-free or sodium-free medium. No increase in the number of exocytotic figures captured by tannic acid incubation was observed during 5 min of incubation with 1 microM monensin or hypo-somolar medium. We thus show that monensin causes beta-endorphin secretion from the melanotrope and that this effect is due to sodium influx and resultant cell swelling. The calcium independency and lack of increase of exocytotic figures suggest that swelling-induced secretion is nonexocytotic, possibly via transient exocytotic pore opening.

Osmotic and pH transmembrane gradients control the lytic power of melittin

Transmembrane osmotic gradients applied on large unilamellar 1-palmitoyl-2-oleoyl-phosphatidylcholine vesicles were used to modulate the potency of melittin to induce leakage. Melittin, an amphipathic peptide, changes the permeability of vesicles, as studied using the release of entrapped calcein, a fluorescent marker. A promotion of the ability of melittin to induce leakage was observed when a hyposomotic gradient (i.e., internal salt concentration higher than the external one) was imposed on the vesicles. It is proposed that structural perturbations caused by the osmotic pressure loosen the compactness of the outer leaflet, which facilitates the melittin-induced change in membrane permeability. Additionally, we have shown that this phenomenon is not due to enhanced binding of melittin to the vesicles using intrinsic fluorescence of the melittin tryptophan. Furthermore, we investigated the possibility of using a transmembrane pH gradient to control the lytic activity of melittin. The potency of melittin in inducing release is known to be inhibited by increased negative surface charge density. A transmembrane pH gradient causing an asymmetric distribution of unprotonated palmitic acid in the bilayer is shown to be an efficient way to modulate the lytic activity of melittin, without changing the overall lipid composition of the membrane. We demonstrate that the protective effect of negatively charged lipids is preserved for asymmetric membranes.

GPI-anchored influenza hemagglutinin induces hemifusion to both red blood cell and planar bilayer membranes

Under fusogenic conditions, fluorescent dye redistributed from the outer monolayer leaflet of red blood cells (RBCs) to cells expressing glycophosphatidylinositol-anchored influenza virus hemagglutinin (GPI-HA) without transfer of aqueous dye. This suggests that hemifusion, but not full fusion, occurred (Kemble, G. W., T. Danieli, and J. M. White. 1994. Cell. 76:383-391). We extended the evidence for hemifusion by labeling the inner monolayer leaflets of RBCs with FM4-64 and observing that these inner leaflets did not become continuous with GPI-HA-expressing cells. The region of hemifusion-separated aqueous contents, the hemifusion diaphragm, appeared to be extended and was long-lived. But when RBCs hemifused to GPI-HA-expressing cells were osmotically swollen, some diaphragms were disrupted, and spread of both inner leaflet and aqueous dyes was observed. This was characteristic of full fusion: inner leaflet and aqueous probes spread to cells expressing wild-type HA (wt-HA). By simultaneous video fluorescence microscopy and time-resolved electrical admittance measurements, we rigorously demonstrated that GPI-HA-expressing cells hemifuse to planar bilayer membranes: lipid continuity was established without formation of fusion pores. The hemifusion area became large. In contrast, for cells expressing wt-HA, before lipid dye spread, fusion pores were always observed, establishing that full fusion occurred. We present an elastic coupling model in which the ectodomain of wt-HA induces hemifusion and the transmembrane domain, absent in the GPI-HA-expressing cells, mediates full fusion.

Granule swelling in stimulated bovine adrenal chromaffin cells: regulation by internal granule pH

Adrenal medullary chromaffin cells secrete catecholamines through exocytosis of their intracellular chromaffin granules. Osmotic granule swelling has been implicated to play a role in the generation of membrane stress associated with the fusion of the granule membrane. However, controversy exists as to whether swelling occurs before or after the actual fusion event. Using morphometric methods we have determined the granule diameter distributions in rapidly frozen, freeze-substituted chromaffin cells. Our measurements show that intracellular chromaffin granules increase in size from an average of 234 nm to 274 nm or 277 nm in cells stimulated to secrete with nicotine or high external K+, respectively. Granule swelling occurs before the formation of membrane contact. Ammonium chloride, an agent which inhibits stimulated catecholamine secretion by approximately 50% by altering the intragranular pH, also inhibits granule swelling. In addition, ammonium chloride-treated secreting cells show more granule-plasma membrane contacts than untreated secreting cells. Sodium propionate induces granule swelling in the absence of secretagogue and has been shown to enhance nicotine- and high K(+)- induced catecholamine release. These results indicate that in adrenal chromaffin cells granule swelling is an essential step in exocytosis before fusion pore formation, and is related to the pH of the granule environment.

Intermembrane lipid transfer during Trypanosoma cruzi-induced erythrocyte membrane destabilization

The ability of Trypanosoma cruzi to induce erythrocyte membrane destabilization in vitro was studied. Epimastigote forms adhered to human erythrocytes and caused fusion or lysis of the red cells, depending on the conditions of the interaction. Red cells were fused in the presence of calcium, while haemolysis was induced in the absence of the cation. Dextran 60 C facilitated fusion but delayed lysis. Optimum pH and temperature for fusion were 7.4 and 37 degrees C, respectively. Lipid alterations were produced in the plasma membrane of the red cell during the interaction with the parasite. A Ca(2+)-independent increase of lysophospholipids and free fatty acids was common to both the lysis and fusion processes. A relative increase of 1,2-diacylglycerides was unique to the fusion process and these changes were dependent on Ca2+. The transfer of free fatty acids and lysophospholipids from T. cruzi to erythrocyte membranes was demonstrated using parasites pre-labelled with radioactive phospholipids. Pre-treatment of parasites with exogenous phospholipase A2 abolished the fusogenicity, while lysis was increased. Neither fusion nor haemolysis occurred when the parasites were pre-treated with fatty acid free albumin, phospholipase A2 inhibitors or when these compounds were present in the medium during the parasite-erythrocyte interaction. Our results suggest that T. cruzi induces erythrocyte membrane destabilization in vitro by transfer of lipid material in a calcium independent manner and that this ion is necessary for other membrane alterations that lead to erythrocyte fusion.

Kinetics of pore size during irreversible electrical breakdown of lipid bilayer membranes

The kinetics of pore formation followed by mechanical rupture of lipid bilayer membranes were investigated in detail by using the charge-pulse method. Membranes of various compositions were charged to a sufficiently high voltage to induce mechanical breakdown. The subsequent decrease of membrane voltage was used to calculate the conductance. During mechanical breakdown, which was probably caused by the widening of one single pore, the membrane conductance was a linear and not exponential function of time after the initial starting process. In a large number of experiments using various lipids and electrolytes, the characteristic opening process of the pore turned out to be independent of the actual membrane potential and electrolyte concentration. Our theoretical analysis of the pore formation suggested that the voltage-induced irreversible breakdown is due to a decrease in edge energy when the pore had formed. After initiation of the pore, the electrical contribution to surface tension is negligible. The time course of the increase of pore size shows that our model of the irreversible breakdown is in good agreement with mechanical properties of membranes reported elsewhere.

Fusion of lipid bilayers: a model involving mechanistic connection to HII phase forming lipids

A model for the molecular mechanism of the fusion of lipid bilayers is described. A crucial feature of this model and related to the lamellar-->hexagonal phase HII transition is a novel, hypothetical lipid conformation, tentatively referred to here as extended. During fusion this conformation could manifest itself in the contact site between two vesicles in close proximity and involves the extension of the acyl chains of a phospholipid molecule in opposite directions, i.e. embedded into the two opposing bilayers while maintaining the headgroup in the interface. Although evidence for the occurrence of the extended conformation for phospholipids is sparse this conformation appears to be compatible with currently available experimental data. Of importance also is that the extended conformation allows for the fusion of two bilayer membranes to proceed with minimal exposure of the lipid hydrocarbon chains to water. It can also account for other features of membrane fusion such as lipid mixing in the intermediate state without mixing of the vesicle contents as well as for the molecular basis of the action of fusogenic lipids.

Changes in membrane permeability during Semliki Forest virus induced cell fusion

The infection of Aedes albopictus cells by Semliki Forest virus (SFV) is a non lytic event. Exposure of infected cells to mildly acidic pH (less than 6.2) leads to syncytium formation. This polykaryon formation is accompanied by an influx of protons into the cells (Kempf et al. Biosci. Rep. 7, 761-769, 1987). We have further investigated this permeability change using various fluorescent or radiolabeled compounds. A significant, pH dependent increase of the membrane permeability to low molecular weight compounds (M(r) less than 1000) was observed when infected cells were exposed to a pH less than 6.2. The pH dependence of the permeability change was very similar to the pH dependence of cell-cell fusion. The permeability change was sensitive to divalent cations, protons and anionic antiviral drugs such as trypan blue. The nature of this virus induced, pH dependent permeability change is discussed.

Calcium uptake in rat parotid gland secretory granules

45Ca2+ uptake in isolated rat parotid secretory granules was examined in the presence of oxalate. Uptake of calcium was dependent on time, with the maximum occurring at 15 min. The uptake of calcium was dependent on adenosine-5'-triphosphate (ATP), and substitution of ATP with beta, gamma-methylene-ATP did not stimulate calcium uptake. Enzyme marker analysis indicated that mitochondria accounted for no greater than 3.0 +/- 0.2% of the observed ATP-dependent calcium uptake. Calcium uptake was blocked by the ATPase inhibitors tributyltin, IC50 = 12.2 +/- 0.6 nmol/L and 4-acetamido-4'-isothiocyano-2,2'-stilbene disulphonic acid (SITS), IC50 = 3.0 +/- 0.3 mumol/L. These results indicate that in the parotid secretory granule there is a calcium uptake mechanism that is dependent on the hydrolysis of ATP and is suppressed by two inhibitors of granule ATPase.

Evidence that the spectrin network and a nonosmotic force control the fusion product morphology in electrofused erythrocyte ghosts

The conversion of the membrane area in the "contact zones" shared by erythrocyte ghosts held in contact by dielectrophoresis into a fusion product by electrofusion was studied by both light and electron microscopy. Fusion products fell into two categories: (a) those with a freely expanding open lumen which ended in the "giant cell morphology" and with considerable internal vesicle membrane fragments, and (b) linear chains of polyghosts with long term stability but having planar diaphragms at the ghost-ghost junctions. Thin section electron microscopy showed each of these planar diaphragms to be a double membrane septum multiply-perforated with fusion pores. Heat and low ionic strength treatments known to denature or detach spectrin caused the stable planar diaphragms to dissolve, thereby quickly converting the polyghost chains to the giant cell morphology, thereby suggesting that spectrin restricts fusion zone diameter expansion if it is intact. Other indications suggest that the expansion of the open lumens appears to take place as a result of one or more membrane-specific forces with a nonosmotic origin but this tendency to expansion can be overcome if the spectrin network on only one side of a contact zone is intact.

Membrane fusion without cytoplasmic fusion (hemi-fusion) in erythrocytes that are subjected to electrical breakdown

There are many reports of hemi-fusion in phospholipid vesicles but few published studies on hemi-fusion in cells. We report evidence from both fluorescence microscopy and freeze-fracture electron microscopy for hemi-fusion in the electrofusion of human erythrocytes. We have also characterised the conditions that favour hemi-fusion as opposed to complete fusion, and discuss the possibility that hemi-fusion might precede complete electrically-induced cell fusion. A membrane probe (DiIC16) and a cytoplasmic probe (6-carboxyfluorescein) were used to investigate the behaviour of doubly-labelled human erythrocytes which were aligned in chains by dielectrophoresis and then exposed to high voltage breakdown pulses. Some of the cells were fused by the pulses, as shown by diffusion of both membrane and cytoplasmic probes from labelled to unlabelled cells. With other cells, the membrane probe diffused into unlabelled cells after the breakdown pulses, without the cytoplasmic probe diffusing into unlabelled cells or leaking into the medium. Membrane fusion (hemi-fusion) thus occurred without cytoplasmic fusion in these erythrocytes. Such cells were irreversibly, but fragilely, attached to their neighbours by the breakdown pulses. There was an inverse relationship between conditions that permit complete fusion and those that favour hemi-fusion, with respect to breakdown pulse length, breakdown voltage and, in particular, osmolarity and temperature. The incidence of hemi-fusion in 250 mM erythritol was twice that in 150 mM erythritol, and hemi-fusion was 5-fold greater at 25 degrees C than at 20 degrees C. Hemi-fused erythrocytes occasionally fused completely on heating to 50 degrees C, demonstrating that hemi-fusion can proceed to complete cell fusion. Freeze fracture electron micrographs of preparations of hemi-fused cells revealed long-lived, complementary depressions and protrusions on the E- and P-fracture faces, respectively, of tightly apposed cells that may mediate hemi-fusion. The possibility that the fusion of closely adjacent human erythrocytes by electrical breakdown pulses may involve an intermediate, shared bilayer structure, which is stable in certain conditions but which can be ruptured by osmotic swelling of the permeabilised cells, is discussed.

Selection of hybrid plants obtained by electrofusion of vacuolated x evacuolated plant protoplasts in hypo-osmolar solution

Vacuolated and evacuolated tobacco mesophyll protoplasts were electrically fused in hypo-osmolar media by using an alternating field of modulated amplitude for alignment. The vacuolated fusion partner was isolated from Nicotiana tabaccum L. cv Xanthi and the evacuolated one from the streptomycin-resistant strain Nicotiana tabaccum L. cv Petit Havana SR1. The field and osmolarity conditions used ensured relatively high yields of heterologous fusion products despite the differences in density and size of the parental cells. After removal of the evacuolated, streptomycin-resistant fused and unfused protoplasts by flotation of vacuole-containing cells on iso-osmolar sucrose medium, the cybrids and hybrids were cultured in 25 microliters drops of agarose. During the first 5 weeks the non-fused Xanthi-protoplasts were used as a nurse culture. After addition of streptomycin to the growth media, cybrids and hybrids were successfully selected whereas fused and unfused vacuole-containing protoplasts died within 6 days. Only the streptomycin-resistant cybrids and hybrids developed into whole plants. On average a yield of 0.025% of streptomycin-resistant plants (referred to the total number of parental cells) was obtained. Polyacrylamide gel electrophoresis of leaf extracts of these plants showed that at least 50% of the streptomycin-resistant plants had a hybrid-esterase isoenzyme pattern. The protocol can be generalised by fusion of iodoacetamide-inactivated vacuolated protoplasts with meristematic (or evacuolized) protoplasts carrying no genetic marker. Use of evacolated protoplasts for electrofusion with vacuole-containing protoplasts therefore offers a way of overcoming the lack of suitable genetic markers for hybrid selection.

Tension in secretory granule membranes causes extensive membrane transfer through the exocytotic fusion pore

For fusion to occur the repulsive forces between two interacting phospholipid bilayers must be reduced. In model systems, this can be achieved by increasing the surface tension of at least one of the membranes. However, there has so far been no evidence that the secretory granule membrane is under tension. We have been studying exocytosis by using the patch-clamp technique to measure the surface area of the plasma membrane of degranulating mast cells. When a secretory granule fuses with the plasma membrane there is a step increase in the cell surface area. Some fusion events are reversible, in which case we have found that the backstep is larger than the initial step, indicating that there is a net decrease in the area of the plasma membrane. The decrease has the following properties: (i) the magnitude is strongly dependent on the lifetime of the fusion event and can be extensive, representing as much as 40% of the initial granule surface area; (ii) the rate of decrease is independent of granule size; and (iii) the decrease is not dependent on swelling of the secretory granule matrix. We conclude that the granule membrane is under tension and that this tension causes a net transfer of membrane from the plasma membrane to the secretory granule, while they are connected by the fusion pore. The high membrane tension in the secretory granule may be the critical stress necessary for bringing about exocytotic fusion.

Surface exposure of phosphatidylserine is associated with the swelling and osmotically-induced fusion of human erythrocytes in the presence of Ca2+

An assay for procoagulant activity has been used to investigate the Ca2(+)-dependent exposure of phosphatidylserine at the surface of human erythrocytes that were induced to swell and to fuse osmotically. Since the phosphatidylserine of human erythrocytes is located in the inner leaflet of the plasma membrane, it is inaccessible in intact cells which therefore had no procoagulant activity in an isotonic solution of sucrose. The procoagulant activity of erythrocytes incubated in increasingly hypotonic, sucrose solutions containing Mg2+ paralleled the percentage haemolysis, reflecting the accessibility of phosphatidylserine in an increasing number of lysed cells. However, cells in mildly hypotonic sucrose solutions containing Ca2+ had an abnormally high procoagulant activity indicating that phosphatidylserine was exposed in intact cells under these conditions. Erythrocytes that were subjected to continuous swelling at 37 degrees C, which was induced by entry of the permeant molecule poly(ethylene glycol) 400 (PEG 400) developed procoagulant activity in the presence of Ca2+ prior to extensive lysis. Cells treated in this way also fused. With Mg2+, PEG 400-treated erythrocytes lysed without fusing, and the development of procoagulant activity paralleled the rate of lysis. Erythrocytes incubated with ionophore A23187, subtilisin, and Ca2+ developed procoagulant activity (with less than 20% lysis), and they fused on subsequent exposure to a hypotonic medium. The procoagulant activity reached its maximum before fusion could be induced in the hypotonic medium. It is concluded that the entry of Ca2+ facilitates a translocation of phosphatidylserine to the outer leaflet of the erythrocyte plasma membrane that plays an important role in fusion protocols that involve cell swelling. It is also suggested that transbilayer movements of phosphatidylserine could be an important control factor in the cell biology of membrane fusion phenomena.

Membrane fusion

The factors involved in the regulation of biological membrane fusion and models proposed for the molecular mechanism of biomembrane fusion are reviewed. The results obtained in model systems are critically discussed in the light of the known properties of biomembranes and characteristics of biomembrane fusion. Biological membrane fusion is a local-point event; extremely fast, non-leaky, and under strict control. Fusion follows on a local and most probably protein-modulated destabilization, and a transition of the interacting membranes from a bilayer to a non-bilayer lipid structure. The potential role of type II non-bilayer preferring lipids and of proteins in the local destabilization of the membranes is evaluated. Proteins are not only responsible for the mutual recognition of the fusion partners, but are most likely also to be involved in the initiation of biomembrane fusion, by locally producing or activating fusogens, or by acting as fusogens.

A mechanism of formation of protein-free regions in the red cell membrane: the rupture of the membrane skeleton

The process of rupture and redistribution of the red cell membrane skeleton is analyzed theoretically. Following the emergence of the rupture the spectrin-actin network is redistributed on the cytoplasmic surface of the membrane bilayer. Due to the interaction of the membrane skeleton and integral proteins the redistribution of the spectrin-actin network leads to the release of purely lipid regions of the membrane. The scale of the protein redistribution caused by the rupture of the membrane skeleton and the size of the lipid domains produced depend on the shape of the membrane and the value of the electrical interaction of the membrane proteins. The lipid domains occurring as a result of the rupture and relaxation of the spectrinactin network can spontaneously increase or decrease its area. The criteria determining the conditions which result in the system's evolutions leading to the domain growth have been obtained. The character of the evolution is determined by the shape of the membrane region in which the rupture occurs as well as the relation between the effective linear tension of the rupture boundary and the modulus of elasticity of the spectrin-actin network.

Evaluation of the annexins as potential mediators of membrane fusion in exocytosis

Membrane fusion is a central event in the process of exocytosis. It occurs between secretory vesicle membranes and the plasma membrane and also among secretory vesicle membranes themselves during compound exocytosis. In many cells the fusion event is regulated by calcium. Since the relevant membranes do not undergo fusion in vitro when highly purified, much attention has been paid to possible protein mediators of these calcium-dependent fusion events. The annexins comprise a group of calcium-dependent membrane-aggregating proteins, of which synexin is the prototype, which can initiate contacts between secretory vesicle membranes which will then fuse if the membranes are further perturbed by the addition of exogenous free fatty acids. This review discusses the secretory pathway and the evidence obtained from in vitro studies that suggests the annexins may be mediators or regulators of membrane fusion in exocytosis.

Surfactant enhancement of polyethyleneglycol-induced cell fusion

B16 mouse melanoma cells in monolayers may be satisfactorily fused with 50% PEG 1500. However, pre-treatment with detergents in solution at low concentrations significantly increases PEG fusion, up to 8-fold in some instances, without impairing cell viability. The practical and mechanistical implications of this finding are discussed.

Effect of transbilayer phospholipid distribution on erythrocyte fusion

Phospholipid packing has been suggested as a relevant variable in the control of membrane fusion events. To test this possibility in a model system, a comparison was made of the fusability of erythrocytes with a normal asymmetric transbilayer distribution of plasma membrane phospholipids (tightly packed exterior lipids) and erythrocytes with a symmetric transbilayer distribution of phospholipids (more loosely packed exterior lipids), using polyethylene glycol as fusogen. Not only were lipid-symmetric cells more readily fused, but fusions of mixtures of lipid-symmetric and lipid-asymmetric cells indicated that both fusing partners must have a symmetric distribution for fusion to be enhanced. Lipid-symmetric cells may fuse more readily because loose packing of the exterior lipids enhances hydrophobic interactions between cells. Alternatively, enhanced membrane fluidity may facilitate intramembranous particle clustering, previously implicated as a potentiator of fusion. Finally, exposure of phosphatidylserine on the surface of lipid-symmetric erythrocytes may be responsible for their enhanced fusion.

Enhanced hybridoma production by electrofusion in strongly hypo-osmolar solutions

Electrofusion of mammalian cells in strongly hypo-osmolar media containing sorbitol, small amounts of divalent cations and albumin resulted in high yields of hybrids. The number of viable hybrids was higher than any value for chemically- or electrically-mediated fusion reported in the literature. Optimum clone numbers were obtained for fusion of osmotically-stable subclones of murine myeloma cells with DNP-Hy-stimulated lymphocytes provided that the osmolarity of the fusion medium was as low as 75 mosmol/l. Similar results were obtained for fusion of osmotically stable subclones of myeloma cells with the murine hybridoma cell line G8. Due to the dramatic increase in volume the field strength of the breakdown pulse (leading to fusion of the dielectrophoretically aligned cells) has to be reduced, as predicted by theory. The efficacy of hypo-osmolar electrofusion allowed the use of very few cells (about 10(5) lymphocytes or G8 cells per fusion chamber). This figure is considerably smaller than that reported in the literature for iso-osmolar electrofusion. It is significant that, in contrast to iso-osmolar conditions, the fusion yield in hypo-osmolar electrofusion was reproducible over long periods of time and less dependent of variations between cultures. At suspension densities of about 10(6) cells per fusion chamber (normally used in iso-osmolar electrofusion) hypo-osmolar electrofusion of homogeneous cell suspensions resulted in the formation of many giant cells when the appropriate field conditions were applied. Similar high or, at some field strengths, even higher numbers of clones at low cell suspension density were obtained when G8 and myeloma cells were first exposed during the washing procedure to strongly hypo-osmolar media, but then transferred to iso-osmolar solutions for electrofusion. Similar experiments with lymphocytes and myeloma cells failed because of destruction of many lymphocytes by the two osmotic shock steps in rapid succession. Volume distribution measurements of G8 and myeloma cells showed that after re-incubation of the osmotically pre-stressed cells the original volume distribution is largely, but not completely re-established. This and other results indicate that osmotic pressure gradients and associated tensions in the membrane do not play a primary role in the initiation of the electrofusion process. The experiments suggest that due to the osmotic (pre-) stress the membrane permeability is slightly and uniformly increased presumably due to the dissolution of membrane- and cell-skeleton proteins. Obviously, this facilitates electrofusion in hypo-osmolar or subsequently in iso-osmolar solutions.

Human platelet electrorotation change induced by activation: inducer specificity and correlation to serotonin release

Electrorotation of single platelets was compared with [14C]serotonin release, aggregation and electron microscopy. Activation of washed and degranulated platelets was induced by thrombin, arachidonic acid, collagen, adrenaline, platelet activation factor (PAF), ADP and A23187. A strong correlation between electrorotation decrease and serotonin release was found. Electrorotation did not correlate with aggregation. It was concluded that an increase of the specific conductivity of the platelet membrane by three orders of magnitude (approx. 1.0.10(-7) S.m-1 to 1.0.10(-4) S.m-1) upon activation was responsible for the observed decrease of anti-field rotation and the shift of the first characteristic frequency towards higher values. Electrorotation allowed for time-dependent measurements of activation. Characteristic activation times in the order of minutes were found. There was the following sequence of activators classified by increasing activation time constants: A23187 was the fastest followed by thrombin, collagen, PAF, arachidonic acid, adrenaline, and ADP.

Human platelet electrorotation change induced by activation: inducer specificity and correlation to serotonin release

Electrorotation of single platelets was compared with [14C]serotonin release, aggregation and electron microscopy. Activation of washed and degranulated platelets was induced by thrombin, arachidonic acid, collagen, adrenaline, platelet activation factor (PAF), ADP and A23187. A strong correlation between electrorotation decrease and serotonin release was found. Electrorotation did not correlate with aggregation. It was concluded that an increase of the specific conductivity of the platelet membrane by three orders of magnitude (approx. 1.0.10(-7) S.m-1 to 1.0.10(-4) S.m-1) upon activation was responsible for the observed decrease of anti-field rotation and the shift of the first characteristic frequency towards higher values. Electrorotation allowed for time-dependent measurements of activation. Characteristic activation times in the order of minutes were found. There was the following sequence of activators classified by increasing activation time constants: A23187 was the fastest followed by thrombin, collagen, PAF, arachidonic acid, adrenaline, and ADP.

Milk lipid globules: control of their size distribution

Micro lipid droplets fuse with each other in vivo to form larger precursors of milk lipid globules. The extent of fusion dictates the size range of lipid globules in milk. A cell-free system in which micro lipid droplet fusion can be induced has been developed. Fusion was promoted by calcium and a protein complex from cytosol. Exogenously supplied gangliosides were potent fusion-promoting agents. Endogenous gangliosides were detected on micro lipid droplet surfaces by immunocytochemical localization. Monoclonal antibody to disialyllactosylceramide (GD3) inhibited micro lipid droplet fusion. These observations are suggestive of a role for gangliosides in micro lipid droplet fusion in situ.

The influenza virus-induced fusion of erythrocyte ghosts does not depend on osmotic forces

The role of osmotic forces and cell swelling in the influenza virus-induced fusion of unsealed or resealed ghosts of human erythrocytes was investigated under isotonic and hypotonic conditions using a recently developed fluorescence assay (Hoekstra, D., De Boer, T., Klappe, K., Wilschut, J. (1984) Biochemistry 23, 5675-5681). The method is based on the relief of fluorescence selfquenching of the fluorescent amphiphile octadecyl rhodamine B chloride (R18) incorporated into the ghost membrane as occurs when labeled membranes fuse with unlabeled membranes. No effect neither of the external osmotic pressure nor of cell swelling on virally mediated ghost fusion was established. Influenza virus fused unsealed ghosts as effectively as resealed ghosts. It is concluded that neither osmotic forces nor osmotic swelling of cells is necessary for virus-induced cell fusion. This is supported by microscopic observations of virus-induced fusion of intact erythrocytes in hypotonic and hypertonic media. A disruption of the spectrin-actin network did not cause an enhanced cell fusion at acidic pH of about 5 or any fusion at pH 7.4.

Are osmotic forces involved in influenza virus-cell fusion?

The kinetics of the fusion process of unsealed and resealed erythrocyte ghosts with influenza virus (A/PR8/34, A/Chile 1/83) were measured under hypotonic, isotonic and hypertonic conditions using a recently developed fluorescence assay (Hoekstra et al. (1984) Biochemistry 23:5675-5681]. No correlation between the external osmotic pressure and kinetics and extent of fusion was observed. Influenza viruses fuse as effectively with unsealed ghosts as with resealed ghosts. It is concluded that osmotic forces as well as osmotic swelling of cells are not necessary for virus-cell membrane fusion.

Cell surface effects of human immunodeficiency virus

Cell killing by human immunodeficiency virus (HIV) is thought to contribute to many of the defects of the acquired immunodeficiency syndrome (AIDS). Two types of cytopathology are observed in HIV-infected cultured cells: cell-cell fusion and killing of single cells. Both killing processes appear to involve cell surface effects of HIV. A model is proposed for the HIV-mediated cell surface processes which could result in cell-cell fusion and single cell killing. The purpose of this model is to define the potential roles of individual viral envelope and cell surface molecules in cell killing processes and to identify alternative routes to the establishment of persistently-infected cells. Elucidation of HIV-induced cell surface effects may provide the basis for a rational approach to the design of antiviral agents which are selective for HIV-infected cells.

Membrane fusion: overcoming of the hydration barrier and local restructuring

The early stages of membrane fusion have been investigated theoretically. It has been shown that the hydration repulsion, operating between apposed membranes, is overcome locally under the action of out-of-plane thermal fluctuations of the bilayers. The fluctuations lead to the formation of close (less than 0.5 nm) contact between the membranes within a small area (approximately 10 nm2). Increasing hydration repulsion between apposed polar heads of lipid molecules in this area causes the rupture of interacting monolayers. The rupture results in monolayer fusion of the membranes, i.e. in the formation of a bridge connecting the monolayers, which is usually named the monolayer stalk. The influence of degree of hydration of the monolayers and their spontaneous curvature on conditions of monolayer fusion have been analysed. The proposed mechanism of early stages of fusion process can proceed without preliminary formation of tight dehydrated contact between the membranes and even without any dehydration.

Fusion of negatively charged liposomes with clathrin-uncoated vesicles

The interaction of lipid vesicles with uncoated vesicles from bovine brain has been studied by fluorescence energy transfer between fluorescent lipid analogs (NBD-PE, Rh-DOPE), by loss of fluorescence self-quenching (NBD-PE, carboxyfluorescein) and by freeze-fracture electron microscopy. The fluorescence techniques monitor the mixing of membranous lipids and the induced release of encapsulated material. The results demonstrate a mixing of the negatively charged lipid (PA, PS) vesicles with the uncoated vesicles. In parallel with the lipid mixing a release of intravesicularly encapsulated material takes place. Lipid vesicles composed of zwitterionic lipids (PC, DOPC, PC:PE) do not specifically interact with uncoated vesicles. The electron micrographs reveal single fusion events. Studies on the kinetics are consistent with a fusional mechanism of the negatively charged lipid vesicles with uncoated vesicles.