Membrane fusion proteins of enveloped animal viruses

Membrane fusion of mumps virus with ghost erythrocytes and CV-1 cells

The octadecyl rhodamine (R18) fluorescent dequenching assay was used to examine membrane fusion between mumps virus and mammalian cells. Rapid fluorescent dequenching, indicative of membrane fusion, was observed when labeled mumps virus was mixed with either ghost erythrocytes or CV-1 cells. After 15 min a saturation limit of 18 virus per erythrocyte ghost and 6400 virus per CV-1 cell was observed. Fetuin was found to inhibit virus fusion, suggesting a role for sialic acid in virus binding to the cells. Two dequenching processes were observed of which the faster process is thought to be membrane fusion and the second process is thought to be probe proximal transfer.

The influence of dextran sulfate on influenza A virus fusion with erythrocyte membranes

Dextran sulfate suppresses the low pH-induced fusion of influenza virus A/Brazil 11/78 with erythrocyte membranes, as shown by fluorescence dequenching assay, using the fluorophore octadecylrhodamine B chloride (R18). Inhibition of fusion was maximal at pH 5.0, while at higher pH values (> 5.6) fusion was not affected. Hemolysis of intact red blood cells by influenza A virus at low pH values is also prevented by dextran sulfate. The inhibiting effect of the polymer is mainly ascribed to repression of virus attachment. Evidence is given that the conformational change of the virus envelope protein hemagglutinin (HA) responsible for triggering fusion is not affected by the polymer.

ph-dependent hydrophobicity profile of hemagglutinin of influenza virus and its possible relevance in virus fusion

The hydropathy profile of hemagglutinin (HA) subunits HA1 and HA2 of influenza virus X31 and A/PR 8/34 is analyzed at different pH. At neutral pH (7.4) pronounced hydrophobic sequences of HA correspond to the N-terminus and the transmembrane spanning sequence of HA2. At pH 5.0 where influenza virus is known to fuse with biological membranes several hydrophobic sequences in the ectodomain exist which are comparable in both the hydrophobicity and length of the N-terminus of HA2. It is suggested that these hydrophobic stretches are important for the fusion complex, in addition to the N-terminal site of HA2.

Bacteriophage phi 6 envelope elucidated by chemical cross-linking, immunodetection, and cryoelectron microscopy

Bacteriophage phi 6 is an enveloped dsRNA virus which infects the plant pathogenic Pseudomonas syringae bacterium. Using low dose cryoelectron microscopy we show that the nucleocapsid, spikeless virion, and intact virion have radii of 29, 35, and 43 nm, respectively. Thus, the membrane is 6 nm thick and the surface spikes of the receptor binding protein P3 extend 8 nm from the membrane surface. Cross-linking, immunological, and complementation evidence suggest that the spikes are formed of multimeric P3 molecules and that P3 is associated with membrane-bound protein P6. We observe that the envelope can accommodate up to 400 molecules of P3 but that the average virion contains less than one-fourth of this amount. Assembly of a very small number of P3 or truncated P3 molecules onto inactive virions restores infectivity, showing that only a few spikes are necessary for receptor binding and membrane fusion.

Recovery of penetration ability in protease-treated zona-free mouse eggs occurs coincident with recovery of a cell surface 94 kD protein

Previous studies have demonstrated that protease treatment of zona-free mouse eggs impairs sperm-egg interaction (Boldt et al.: Biol Reprod 39:19-27, 1988) and causes modification of a 94 kD egg plasma membrane protein (Boldt et al., Gamete Res 23:91-101, 1989). In this report, the ability of eggs to recover penetration ability following protease treatment was examined. Zona-free mouse eggs were isolated and treated with either trypsin or chymotrypsin (1 mg/ml, 20 min), then cultured for 0, 3, or 6 hr before insemination. Eggs cultured for 3 or 6 hr displayed significantly higher penetration levels than eggs inseminated immediately after protease treatment, indicating a recovery of penetration ability during the 3 or 6 hr incubation period. The recovery of penetration ability was not blocked by inclusion of cyclohexamide (50 micrograms/ml) during the 3 or 6 hr culture period, indicating that protein synthesis was not required for recovery of fusion ability. Cell surface radiolabeling studies with 125I revealed that a 94 kD cell surface protein was lost immediately following trypsin or chymotrypsin treatment but was found on the egg surface after the 3 or 6 hr recovery period. Recovery of the 94 kD egg surface protein occurred in the presence of cyclohexamide, and metabolic radiolabeling studies with 35S-methionine confirmed that synthesis of a 94 kD protein was blocked by cyclohexamide. These results suggest that the recovery of penetration ability after protease treatment of zona-free eggs is due to recovery of the 94 kD cell surface protein, providing further evidence for the involvement of the 94 kD protein in sperm-egg interaction.

Computer detection of the rapid diffusion of fluorescent membrane fusion markers in images observed with video microscopy

We have developed an algorithm for automated detection of the dynamic pattern characterizing flashes of fluorescence in video images of membrane fusion. The algorithm detects the spatially localized, transient increases and decreases in brightness that result from the dequenching of fluorescent dye in phospholipid vesicles or lipid-enveloped virions fusing with a planar membrane. The flash is identified in video images by its nonzero time derivative and the symmetry of its spatial profile. Differentiation is implemented by forward and backward subtractions of video frames. The algorithm groups spatially connected pixels brighter than a user-specified threshold into distinct objects in forward- and backward-differentiated images. Objects are classified as either flashes or noise particles by comparing the symmetries of matched forward and backward difference profiles and then by tracking each profile in successive difference images. The number of flashes identified depends on the brightness threshold, the size of the convolution kernel used to filter the image, and the time difference between the subtracted video frames. When these parameters are changed so that the algorithm identifies an increasing percentage of the flashes recognized by eye, an increasing number of noise objects are mistakenly identified as flashes. These mistaken flashes can be eliminated by a human observer. The algorithm considerably shortens the time needed to analyze video data. Tested extensively with phospholipid vesicle and virion fusion with planar membranes, our implementation of the algorithm accurately determined the rate of fusion of influenza virions labeled with the lipophilic dye octadecylrhodamine (R18).

Mammalian sperm-egg fusion: the rat egg has complementary sites for a sperm protein that mediates gamete fusion

Rat epididymal protein DE is localized on the fusogenic region of the acrosome-reacted spermatozoa and has a potential role in sperm-egg fusion. We investigated the presence of DE binding sites on the egg surface by co-incubating zona-free eggs and capacitated sperm in different concentrations of pure DE. Results indicate that DE produced a concentration-dependent decrease in egg penetration by sperm (fusion), with almost complete inhibition at 200 micrograms/ml. This inhibition was not due to an effect of DE on initial sperm binding to the egg membrane, since the presence of this protein did not affect the percentage of oocytes with bound sperm nor the number of bound sperm per egg. Those sperm that failed to penetrate the egg in the presence of DE became able to do so after transfer of the eggs to protein- and sperm-free medium, indicating a role for DE in an event subsequent to binding and leading to fusion. Indirect immunofluorescence using a polyclonal antibody against DE revealed a patchy labeling over the entire egg surface, with the exception of the area overlying the second metaphase spindle. This conclusion was supported by the disappearance of the DE-negative area on the fertilized egg. Zona-free eggs, incubated with DE at 4 degrees C or fixed before exposure to DE, displayed a uniform staining, suggesting that the patchy labeling resulted from aggregation of DE binding sites by the purified protein. The aggregation of these egg components may represent a necessary step of the fusion process. To our knowledge, this is the first study reporting the existence and localization of complementary sites to a specific sperm protein on the plasma membrane of the mammalian egg.

The amino-terminal peptide of HIV-1 glycoprotein 41 interacts with human erythrocyte membranes: peptide conformation, orientation and aggregation

Structural studies assessed interactions between the amino-terminal peptide (FP-I; 23 residues 519-541) of the glycoprotein 41,000 (gp41) of Human Immunodeficiency Virus Type-1 (HIV-1) and human erythrocyte membranes and simulated membrane environments. Peptide binding was examined at sub-hemolytic (approx. less than 5 microM) and hemolytic (greater than or equal to 5 microM) doses (Mobley et al. (1992) Biochem. Biophys. Acta 1139, 251-256), using circular dichroism (CD) and Fourier-transform infrared (FTIR) measurements with FP-I, and electron spin resonance (ESR) studies employing FP-I spin-labeled at either the amino-terminal alanine (FP-II; residue 519) or methionine (FP-III; position 537). In the sub-lytic regime, FP-I binds to both erythrocyte lipids and dispersions of SDS with high alpha-helicity. Further, ESR spectra of FP-II labeled erythrocyte ghosts indicated peptide binding to both lipid and protein. In ghost lipids, FP-II was monomeric and exhibited low polarity and rapid, anisotropic motion about its long molecular axis (i.e., alpha-helical axis), with restricted motion away from this axis. The spin-label at the amino-terminal residue (Ala-519) is insensitive to the aqueous broadening agent chromium oxalate and buried within the hydrophobic core of the membrane; the angle that the alpha-helix (residues 519-536) makes to the normal of the bilayer plane is either 0 degree or 40 degrees. Contrarily, ESR spectra of ghost lipids labeled with sub-lytic doses of FP-III indicated high mobility and polarity for the reporter group (Met-537) at the aqueous-membrane interface, as well as extreme sensitivity to chromium oxalate. At lytic FP-I doses, CD and FTIR showed both alpha-helix and beta-structure for peptide in ghost lipids or detergent, while ESR spectra of high-loaded FP-II in ghost membranes indicated peptide aggregates. Membrane aggregates of FP-I may be involved in hemolysis, and models are suggested for N-terminal gp41 peptide participation in HIV-induced fusion and cytolysis.

Membrane fusion of influenza virus with phosphatidylcholine liposomes containing viral receptors

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

Biological activity of paramyxovirus fusion proteins: factors influencing formation of syncytia

The fusion (F) and hemagglutinin-neuraminidase (HN) glycoproteins of the paramyxovirus simian virus 5 (SV5) were expressed individually or coexpressed in CV-1 cells by using SV40-based vectors and recombinant vaccinia viruses. The extent of detectable fusion in a syncytium formation assay was found to be affected by the expression system used. In addition, when HN was coexpressed with F, it was found that the expression vector system influenced the contribution of HN in forming syncytia. The abilities of the SV5, human parainfluenza virus type 3, and Newcastle disease virus F glycoproteins to cause fusion, when expressed alone or coexpressed with HN, were directly compared by using the SV40-based vector system in CV-1 cells. The F proteins exhibited various degrees of fusion activity independent of HN expression, but the formation of syncytia could be enhanced to different extents by the coexpression of the homotypic HN protein.

Interaction of rotavirus particles with liposomes

We have studied the interactions of purified viral particles with liposomes as a model to understand the mechanism of entry of rotavirus into the cell. Liposomes, made from pure as well as mixed lipids, that contained encapsulated self-quenching concentrations of the fluorophore carboxyfluorescein (CF) were used. Rotavirus-liposome interactions were studied from the fluorescence dequenching of CF resulting from its release to the bulk solution. Purified infectious double-shelled virus particles induced a concentration- and temperature-dependent release of CF. The rate and extent of CF release was maximum between pH 7.3 and 7.6. The removal of outer structural proteins VP4 and VP7 from virus, which results in the formation of single-shelled particles, prevented virus interaction with liposomes. Rotavirus particles with uncleaved VP4 did not interact with liposomes, but treatment in situ of these particles with trypsin restored the interaction with the liposomes and resulted in CF dequenching. Our data support the view that rotavirus enters the cell through direct penetration of the plasma membrane. In contrast, adenovirus, the only other nonenveloped virus studied by this method, shows the optimum rate of marker release from liposomes at around pH 6 (R. Blumenthal, P. S. Seth, M. C. Willingham, and I. Pastan, Biochemistry 25:2231-2237, 1986). The interaction between rotavirus and liposomes is sensitive to specific divalent metal ions, unlike the adenovirus-liposome interaction, which is independent of them.

Proteins on exocytic vesicles mediate calcium-triggered fusion

In many exocytic systems, micromolar concentrations of intracellular Ca2+ trigger fusion. We find that aggregates of secretory granules isolated from sea urchin eggs fuse together when perfused with greater than or equal to 10 microM free Ca2+. Mixing of membrane components was demonstrated by transfer of fluorescent lipophilic dye, and melding of granule contents was seen with differential interference microscopy. A technique based upon light scattering was developed to conveniently detect fusion. Two protein modifiers, trypsin and N-ethylmaleimide, inhibit granule-granule fusion at concentrations similar to those that inhibit granule-plasma membrane fusion. We suggest that molecular machinery sufficient for Ca(2+)-triggered fusion resides on secretory granules as purified and that at least some of these essential components are proteinaceous.

Membrane destabilization by N-terminal peptides of viral envelope proteins

The fusion of lipid enveloped viruses with cellular membranes is thought to be mediated by the insertion into the target membrane of the N-terminal polypeptides of viral spike glycoproteins. Since membrane destabilization is a necessary step in membrane fusion, we investigated whether synthetic peptides with amino acid sequences corresponding to the N-termini of influenza virus hemagglutinin (HA2), vesicular stomatitis virus G-protein and Sendai virus F-protein, induce the destabilization and fusion of phospholipid vesicles. Membrane destabilization by the peptides was monitored by the release of aqueous contents of large unilamellar phospholipid vesicles. Aggregation was detected by a resonance energy transfer assay. Membrane fusion was followed by means of assays for the intermixing of phospholipids and of aqueous contents. The 17-amino acid HA2 peptide (HA2.17) destabilized phosphatidylcholine (PC) vesicles even at neutral pH, but the rate and extent of destabilization increased at lower pH. This peptide did not mediate appreciable release of contents from phosphatidylserine (PS) vesicles. HA2.17 induced neither aggregation nor fusion of PC or PS vesicles. In contrast, the 7-amino acid N-terminal peptide of G-protein (G.7) destabilized PS-containing membranes and not pure PC vesicles. Although G.7 caused aggregation of and lipid mixing between PS vesicles, it did not mediate any detectable intermixing of aqueous contents. The presence of cholesterol in PC membranes did not affect the destabilization caused by the N-terminal peptide of Sendai virus F-protein (F1.7), suggesting that cholesterol is not necessary for the effective interaction of this peptide with membranes, contrary to earlier proposals. Our results support the hypothesis that the hydrophobic N-terminal region of certain viral envelope proteins insert into and destabilize target membranes.

Studies on the fusion peptide of a paramyxovirus fusion glycoprotein: roles of conserved residues in cell fusion

The role of residues in the conserved hydrophobic N-terminal fusion peptide of the paramyxovirus fusion (F) protein in causing cell-cell fusion was examined. Mutations were introduced into the cDNA encoding the simian virus 5 (SV5) F protein, the altered F proteins were expressed by using an eukaryotic vector, and their ability to mediate syncytium formation was determined. The mutant F proteins contained both single- and multiple-amino-acid substitutions, and they exhibited a variety of intracellular transport properties and fusion phenotypes. The data indicate that many substitutions in the conserved amino acids of the simian virus 5 F fusion peptide can be tolerated without loss of biological activity. Mutant F proteins which were not transported to the cell surface did not cause cell-cell fusion, but all of the mutants which were transported to the cell surface were fusion competent, exhibiting fusion properties similar to or better than those of the wild-type F protein. Mutant F proteins containing glycine-to-alanine substitutions had altered intracellular transport characteristics, yet they exhibited a great increase in fusion activity. The potential structural implications of this substitution and the possible importance of these glycine residues in maintaining appropriate levels of fusion activity are discussed.

Cell fusion caused by herpes simplex virus-1 (HSV-1) strains tsB5 and MP is inhibited at pH 6.7 and pH 7.0

We investigated the effect of different pH conditions on Vero cell cultures infected with herpes simplex virus-1 (HSV-1) wild-type strain KOS, and syncytial mutants HSV-1 HFEM (tsB5) and HSV-1 mp (MP). Cell fusion was inhibited when infected cells were continuously incubated with culture media adjusted to pH 6.7 or pH 7.0. Inhibition of cell fusion was rapidly reversible when infected cell cultures were returned to pH 7.5. The rate of synthesis and cell-surface expression of virus-specified glycoproteins gB, gC, gD, and gH were not affected during continuous incubation at pH 7.0, but they were reduced at pH 6.7 in comparison to pH 7.5. At later hours p.i. however, these glycoproteins continued to accumulate at all tested pH levels. Accumulation of infectious virions was substantially reduced for MP, KOS, and tsB5 at pH 6.7. At pH 7.0, KOS and tsB5 titers were greatly reduced but MP titers were not affected.

Mechanism of entry of a xenotropic MMuLV-derived recombinant retrovirus into porcine cells using the expression of the reporter nlslacZ gene

A xenotropic Moloney murine leukemia virus-derived recombinant retrovirus (MMuLVSVnlslacZ) has been utilized to study the mechanism of virus entry into endothelial and epithelial porcine cells. In the genome of this recombinant retrovirus, the nlslacZ reporter gene is under the transcriptional control of both LTR and SV40 early promoter. The entry of the retrovirus has been determined from the expression of this transduced reporter gene after its integration into the infected cells. This allows the detection of a very low level of viral infection and hence entry of the virus. Exposure of the virus-cell mixture to acidic pH (less than 6) during the early phase of interaction reduces the level of internalization. Cellular infection in presence of weak bases, ammonium chloride and amantadine and an ionophore monensin at concentrations sufficient to neutralize the endosomal pH does not modify the extent of viral entry into the cells. The results indicate that the entry of the recombinant retrovirus into porcine cells takes place by a pH-independent viral membrane-cell plasma membrane fusion mechanism.

Significance of basolateral domain of polarized MDCK cells for Sendai virus-induced cell fusion

Fusion (fusion from within) of polarized MDCK monolayer cells grown on porous membranes was examined after infection with Sendai viruses. Wild-type virus, that buds at the apical membrane domain, did not induce cell fusion even when the F glycoprotein expressed at the apical domain was activated with trypsin. On the other hand, a protease activation mutant, F1-R, with F protein in the activated form and that buds bipolarly at the apical and basolateral domains, caused syncytia formation in the absence of exogenous protease. Anti-Sendai virus antibodies added to the basolateral side, but not at the apical side, inhibited cell fusion induced by F1-R. In addition, T-9, a mutant with bipolar budding phenotype of F1-R but with an uncleavable F protein phenotype like wild-type virus, induced cell fusion exclusively when trypsin was added to the basolateral medium. By electron microscopy, cell-to-cell fusion was shown to occur at the lateral domain of the plasma membrane. These results indicate that in addition to proteolytic activation of the F protein, basolateral expression of Sendai virus envelope glycoproteins is required to induce cell fusion.

Introduction of intersubunit disulfide bonds in the membrane-distal region of the influenza hemagglutinin abolishes membrane fusion activity

Influenza virus hemagglutinin (HA) mediates viral entry into cells by a low pH-induced membrane fusion event in endosomes. A number of structural changes occur throughout the length of HA at the pH of fusion. To probe their significance and their necessity for fusion activity, we have prepared a site-directed mutant HA containing novel intersubunit disulfide bonds designed to cross-link covalently the membrane-distal domains of the trimer. These mutations inhibited the low pH-induced conformational changes and prevented HA-mediated membrane fusion; conditions that reduced the novel disulfide bonds restored membrane fusion activity. We conclude that structural rearrangements in the membrane distal region of the HA are required for membrane fusion activity.

New aspects of influenza viruses

Influenza virus infections continue to cause substantial morbidity and mortality with a worldwide social and economic impact. The past five years have seen dramatic advances in our understanding of viral replication, evolution, and antigenic variation. Genetic analyses have clarified relationships between human and animal influenza virus strains, demonstrating the potential for the appearance of new pandemic reassortants as hemagglutinin and neuraminidase genes are exchanged in an intermediate host. Clinical trials of candidate live attenuated influenza virus vaccines have shown the cold-adapted reassortants to be a promising alternative to the currently available inactivated virus preparations. Modern molecular techniques have allowed serious consideration of new approaches to the development of antiviral agents and vaccines as the functions of the viral genes and proteins are further elucidated. The development of techniques whereby the genes of influenza viruses can be specifically altered to investigate those functions will undoubtedly accelerate the pace at which our knowledge expands.

Investigation of human immunodeficiency virus fusion peptides. Analysis of interrelations between their structure and function

The N-terminal region of the human immunodeficiency virus type 1 (HIV-1) gp41 appears to be involved in virus-cell membrane fusion. To study the influence of fusion domain structure on gp41 interaction with artificial lipid membranes, two families of peptides were synthesized. The peptides of the first family starting from the C-terminal Gly-532 of gp160 (BRU isolate) were assembled in a stepwise manner to N-terminus of gp41(Ala-517). These hydrophobic peptides, containing 10-16 amino acid residues (a.a.), were able to form channel-like current fluctuation through planar lipid membranes, and the longest 15-16 a.a. peptides lysed the liposomes. Peptides of the second family beginning from the C-terminal Arg-538 and continuing to Val-510 contained several hydrophilic amino acid residues. These 15-22 a.a. peptides also increased the conductance of planar lipid bilayers and lysed liposomes. The degree of liposome lysis depended upon peptide length and concentration. The attachment of gp120 C-terminal amino acid or peptides to N-terminus of 517-538 peptide resulted in complete loss of activity. The effects of the second family of peptides on membranes were reduced to a great extent at acidic pH. The conjugation of 22 a.a. Lys peptide with bovine serum albumin decreased its lytic activity. The circular dichroism study of these peptides revealed alpha-helix configuration in hydrophobic and aqueous media only for deca- and longer peptides. The electron microscopy of 22 a.a. peptide performed in the aqueous medium showed large spherical aggregates about 0.5-0.7 micron in diameter consisting of long filaments approximately 5 nm in diameter. Other tested peptides could generate only short strings. Thus, the effects of fusion peptides on lipid membranes depends on their sequence and length, secondary and tertiary structures, and freedom of their N-terminus.

Fusion of Sendai virus with human HL-60 and CEM cells: different kinetics of fusion for two isolates

The kinetics of fusion of Sendai virus (Z strain) with the human promyelocytic leukemia cell line HL-60, and the human T lymphocytic leukemia cell line CEM was investigated. Fusion was monitored by fluorescence dequenching of octadecylrhodamine (R-18) incorporated in the viral membrane. For one virus isolate (Z/G), the overall rate of fusion (at 37 degrees C) increased as the pH was lowered, reaching a maximum at about pH 5, the lowest pH tested. For another isolate (Z/SF) the rate and extent of fusion were lower at pH 5 than at neutral pH. Lowering the pH from neutral to 5 after several minutes of incubation of either isolate with HL-60 cells resulted in an enhanced rate of fluorescence dequenching. Nevertheless, experiments utilizing NH4Cl indicated that fusion of the virus with cells was not enhanced by the mildly acidic pH of the endosome lumen. Analysis of the kinetics of fusion by means of a mass action model resulted in good simulation and predictions for the time-course of fusion. For the isolate which showed maximal fusogenic activity at pH 5, the rate constant of fusion (approx. 0.1 s-1) at neutral pH was in the range found previously for virus-liposome fusion, whereas the rate constant of adhesion was close to the upper limit for diffusion-controlled processes (1.4.10(10) M-1 s-1). However, for the other isolate (Z/SF) the rate constant of fusion at neutral pH was very small (less than 0.01 s-1), whereas the rate constant of adhesion was larger (greater than or equal to 2.10(10) M-1 s-1). Lowering the temperature decreased the fusion rate. Experiments involving competition with excess unlabeled virions indicated that not all binding sites for Sendai virus on HL-60 cells are fusion sites. The virus fusion activity towards HL-60 cells at neutral pH was not altered significantly by pre-incubation of the virus at pH 5 or 9, in contrast to earlier observations with liposomes and erythrocyte ghosts, or results based on erythrocyte hemolysis or cell-cell fusion.

Identification of the fusion domain in the visna virus transmembrane protein

Visna virus, a lentivirus of sheep, causes fusion of susceptible cells. Fusion has previously been shown to be mediated by the viral envelope glycoprotein. The transmembrane protein of visna virus contains a hydrophobic region at its amino terminus. This region is similar to the fusion epitopes of the orthomyxoviruses and paramyxoviruses. This region is located in a position similar to that of the fusion epitopes in the transmembrane proteins of HIV-1 and SIV. To determine the role of this hydrophobic region in visna virus-induced cell fusion, a peptide of 24 amino acids corresponding to this region was synthesized. The peptide alone induces fusion of goat cells. Antibodies to this peptide inhibit both viral-induced cell fusion and peptide fusion in goat cells. Further, the direct fusion of cells by this peptide is a unique observation and may be useful for studying the fusion epitopes of other lentiviruses. Thus, this hydrophobic region appears to be one epitope responsible for visna virus-induced cell fusion.

Interactions of antisense DNA oligonucleotide analogs with phospholipid membranes (liposomes)

Antisense oligonucleotides have the ability to inhibit individual gene expression in the potential treatment of cancer and viral diseases. However, the mechanism by which many oligonucleotide analogs enter cells to exert the desired effects is unknown. In this study, we have used phospholipid model membranes (liposomes) to examine further the mechanisms by which oligonucleotide analogs cross biological membranes. Permeation characteristics of 32P or fluorescent labelled methylphosphonate (MP-oligo), phosphorothioate (S-oligo), alternating methylphosphonate-phosphodiester (Alt-MP) and unmodified phosphodiester (D-oligo) oligodeoxynucleotides were studied using liposomal membranes. Efflux rates (t1/2 values) at 37 degrees C for oligonucleotides entrapped within liposomes ranged from 7-10 days for D-, S- and Alt-MP-oligos to about 4 days for MP-oligos. This suggests that cellular uptake of oligonucleotides by passive diffusion may be an unlikely mechanism, even for the more hydrophobic MP-oligos, as biological effects are observed over much shorter time periods. We also present data that suggest oligonucleotides are unlikely to traverse phospholipid bilayers by membrane destabilization. We show further that MP-oligos exhibit saturable binding (adsorption) to liposomal membranes with a dissociation constant (Kd) of around 20nM. Binding appears to be a simple interaction in which one molecule of oligonucleotide attaches to a single lipid site. In addition, we present water-octanol partition coefficient data which shows that uncharged 12-15 mer MP-oligos are 20-40 times more soluble in water than octanol; the low organic solubility is consistent with the slow permeation of MP-oligos across liposome membranes. These results are thought to have important implications for both the cellular transport and liposomal delivery of modified oligonucleotides.

Membrane fusion activity of the influenza virus hemagglutinin: interaction of HA2 N-terminal peptides with phospholipid vesicles

We have investigated the interaction of a number of synthetic 20-residue peptides, corresponding to the HA2 N-terminus of the influenza virus hemagglutinin (X31 strain), with phospholipid vesicles and monolayers. Besides the wild-type sequence, two peptides were studied with mutations corresponding to those previously studied in entire HA's expressed in transfected cells [Gething et al., (1986) J. Cell. Biol. 102, 11-23]. These mutations comprised a single Glu replacement for Gly at the N-terminus ("El" mutant) or at position 4 ("E4") of the HA2 subunit and were shown to produce striking alterations in virus-induced hemolysis and syncytia formation, especially for E1. The X31 "wild-type" (wt) peptide and its E4 variant are shown here to have the capacity to insert into phosphatidylcholine (POPC) large unilamellar vesicle (LUV) membranes in a strictly pH-dependent manner, penetration being marginal at pH 7.4 and significant at pH 5.0. Bilayer insertion was evident from a shift in the intrinsic Trp fluorescence of the wt and E4 peptides and from the induction of calcein leakage from POPC LUV and correlated well with the peptides' ability at pH 5.0 to penetrate into POPC monolayers at initial surface pressures higher than 30 mN/m. By contrast, the E1 peptide was found, at pH 5.0, to bind less tightly to vesicles (assessed by a physical separation method) and to cause much less leakage of POPC LUV than the wt, even under conditions where the peptides were bound to approximately the same extent. Consistent with the correlation between leakage and penetration observed for the wt peptide at pH 5 versus 7, the E1 peptide, even at low pH, showed much less lipid-vesicle-induced shift of its Trp fluorescence than wt, caused a much slower rate of leakage of vesicle contents, and did not insert into POPC monolayers at surface pressures beyond 28.5 mN/m. Circular dichroism spectroscopy measurements of peptides in POPC SUV showed that the conformations of all three peptides are sensitive to pH, but only the wt and E4 peptides became predominantly alpha-helical at acid pH.

Semliki Forest virus envelope proteins function as proton channels

It has been shown that isolated nucleocapsids of Semliki Forest virus (SFV) contract upon low pH exposure (Soederlund et al., 1972). This contraction of the nucleocapsids has been used as an indicator to demonstrate that the spike proteins of SFV can translocate protons into the interior of the virus particle upon low pH (5.8) exposure. Spikeless virus particles obtained after bromelain digestion, which were used as a control, did not translocate protons. This implies that the ectodomain of the spike plays a crucial role for the proton translocation.

Modification of the N-terminus of membrane fusion-active peptides blocks the fusion activity

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

Liposomes for the transformation of eukaryotic cells

Gene therapy of human disease is a method of treatment under active development. DNA-loaded liposomes exhibit great promise for use in this field. Liposome-based transfection vectors have many inherent advantages that will likely lead to their wide in vivo use. Vectors with low toxicity and a high degree of targetability can now be easily prepared. These vectors are also free of the length constraints governing retroviral vectors. In this review we discuss recent developments in the use of liposomes for transfection of eukaryotic cells.

Identification of the envelope V3 loop as the primary determinant of cell tropism in HIV-1

Cells of the monocyte-macrophage lineage are targets for human immunodeficiency virus-1 (HIV-1) infection in vivo. However, many laboratory strains of HIV-1 that efficiently infect transformed T cell lines replicate poorly in macrophages. A 20-amino acid sequence from the macrophage-tropic BaL isolate of HIV-1 was sufficient to confer macrophage tropism on HTLV-IIIB, a T cell line--tropic isolate. This small sequence element is in the V3 loop, the envelope domain that is the principal neutralizing determinant of HIV-1. Thus, the V3 loop not only serves as a target of the host immune response but is also pivotal in determining HIV-1 tissue tropism.

Defective acidification of intracellular organelles in cystic fibrosis

The phenotype of cystic fibrosis (CF) includes abnormalities in transepithelial transport of Cl- (refs 1-5), decreased sialylation and increased sulphation and fucosylation of glycoproteins, and lung colonization with Pseudomonas. It is not apparent how these abnormalities are interrelated, nor how they result from loss of function of the CF gene-encoded transmembrane regulator (CFTR). We have previously shown that that the pH of a secretory granule is regulated by the vesicular conductance for Cl- (ref. 11). Here we find defective acidification in CF cells of the trans-Golgi/trans-Golgi network, of prelysosomes and of endosomes as a result of diminished Cl- conductance. Sialytation of proteins and lipids is reduced and ligand traffic altered. These abnormalities can result from defective acidification because vacuolar pH regulates glycoprotein processing and ligand transport. The CF phenotype is similar to that of alkalinized cells and acidification-defective mutatants.

The interaction of synthetic analogs of the N-terminal fusion sequence of influenza virus with a lipid monolayer. Comparison of fusion-active and fusion-defective analogs

The amino terminus of subunit-2 of influenza virus hemagglutinin (NHA2) plays a crucial role in the induction of fusion between viral and endosomal membranes leading to the infection of a cell. Three synthetic analogs with an amino acid sequence corresponding to NHA2 of variant hemagglutinins were studied in a monolayer set up. Comparison of the interaction of a fusion-active and two fusion-defective analogs with a lipid monolayer revealed a greater surface activity of the fusion-active analog. Pronounced differences were found if the pure peptides were spread at the air/water interface; the fusion-active analog showed a higher collapse pressure and a greater limiting molecular area. Circular dichroism measurements on collected lipid monolayers indicated a high content of alpha-helical structure for the fusion-active and one of the fusion-defective analogs. A simple relation between alpha-helical content and fusogenicity does not seem to exist. Instead, the extent of penetration, a defined tertiary structure or orientation of the alpha-helical peptide may be essential for its membrane perturbing activity.

Interaction of non-enveloped plant viruses and their viral coat proteins with phospholipid vesicles

The interaction of the non-enveloped plant viruses TMV (rod-shaped) and CCMV (spherical) and of their coat proteins in several well-defined aggregation states, with artificial membranes was investigated to study the early stages of the cellular infection process. Information about the separate steps in the interaction mechanisms was obtained by employing three assays, performed as a function of vesicle size, net membrane charge, pH and ionic strength. The assays allow to discriminate between aggregation of vesicles (turbidity assay) and membrane destabilization (vesicle leakage assay and lipid mixing assay). The aggregation of the vesicles is a result of electrostatic interactions between the viral material and vesicles surface (cross-linking), while the destabilization of the membrane is a result of penetration or bilayer disruption by hydrophobic protein domains. TMV virions and its coat protein, and CCMV virions, due to their net negative charge, predominantly interact with positively charged membranes. The coat protein of CCMV was found to interact with negatively charged membranes, an interaction that can be assigned to its basical N-terminal sequence. Changing the aggregational state of the viral coat proteins yielded most significant interactions in case of TMV coat protein aggregated in the disk form and CCMV coat protein aggregated in empty capsids with oppositely charged membranes. These protein aggregates are found to be the best compromise between efficiency (capacity of the protein to bridge vesicles and destabilize their membranes) and concentration of protein aggregates. The results are discussed with respect to previously proposed biological models of the early stages of plant virus infection.

A model for the study of the mechanism of a low pH-induced interaction of the virus fusion proteins and cell membranes

A model is proposed for the study of molecular mechanisms of a low pH-induced interaction of fusion proteins of enveloped viruses and cell membranes. The model consists of large monolamellar liposomes containing ionophore nigericin in their membranes and ectodomains of fusion protein in their inner space. The process of interaction of the protein with the lipid bilayer is triggered by acidification of the liposomal constituents to the pH of fusion with the help of nigericin by adding citric acid to the outer medium. To visualize the protein structural reorganization, the tritium planigraphy was used. Comparison of the values of specific labelling of the proteins and distribution of radioactivity in individual amino acids in control (at neutral pH) and experimental liposome samples (at the pH of fusion) permits to realise the character of protein-membrane interaction. We have obtained the first results in the study of interaction of the bromelain-released soluble ectodomain of the HAXX molecule (BHA)--with the lipid membrane. The observed increase in the protein specific activity and selective increase in the specific activity of hydrophobic amino acids Ile, Phe and Tyr in experimental liposome samples as compared with the controls did not contradict to the conventional concept, that a hydrophobic N-terminus of HA2 subunit of hemagglutinin is responsible for its interaction with lipid membranes.

The role of N-acetylneuraminic (sialic) acid in the pH dependence of influenza virion fusion with planar phospholipid membranes

It is known that fusion of influenza virus to host cell membranes is strongly promoted by acidic pH. We have determined conditions required to obtain pH-dependent fusion of influenza virus to planar bilayer membranes. The rate of viral fusion was determined from the flash rate of R18-labeled virions delivered to the surface of the planar membrane by pressure-ejection from a pipette. For a bilayer formed only of phospholipids and cholesterol, the fusion rate was independent of pH and unaffected by the phospholipid composition. When the gangliosides GD1a + GT1b were included in the planar membrane, however, the fusion rate varied steeply with pH. The rate at pH 7.4 in the presence of the gangliosides was about an order of magnitude less than in their absence. At pH less than approximately 5.5, the rate was about an order of magnitude greater in the presence of gangliosides than in their absence. The fusion rate with planar membranes containing globoside, a ceramide-backboned glycolipid, was also independent of pH, indicating that the pH dependence required sialic acid on the carbohydrate moiety of the glycolipid. The gangliosides GM1a and GM3, both of which possess sialic acid, produced the same pH-dependent fusion rate as seen with GD1a + GT1b, indicating that the presence, but not the location, of terminal sialic acids is critical. Incubating virus with soluble sialyllactose blocked fusion to both ganglioside-free and ganglioside-containing planar membranes. These results show that the pH dependence of influenza virion fusion arises from the interaction of the sialic acid receptor with the influenza hemagglutinin. A model for sialic acid-hemagglutinin interactions accounting for pH-dependent fusion is presented.

Fusion of influenza virions with a planar lipid membrane detected by video fluorescence microscopy

The fusion of individual influenza virions with a planar phospholipid membrane was detected by fluorescence video microscopy. Virion envelopes were loaded with the lipophilic fluorescent marker octadecylrhodamine B (R18) to a density at which the fluorescence of the probe was self-quenched. Labeled virions were ejected toward the planar membrane from a micropipette in a custom-built video fluorescence microscope. Once a virion fused with the planar membrane, the marker was free to diffuse, and its fluorescence became dequenched, producing a flash of light. This flash was detected as a transient spot of light which increased and then diminished in brightness. The diffusion constants calculated from the brightness profiles for the flashes are consistent with fusion of virus to the membrane with consequent free diffusion of probe within the planar membrane. Under conditions known to be fusigenic for influenza virus (low pH and 37 degrees C), flashes appeared at a high rate and the planar membrane quickly became fluorescent. To further establish that these flashes were due to fusion, we showed that red blood cells, which normally do not attach to planar membranes, were able to bind to membranes that had been exposed to virus under fusigenic conditions. The amount of binding correlated with the amount of flashing. This indicates that flashes signaled the reconstitution of the hemagglutinin glycoprotein (HA) of influenza virus, a well-known erythrocyte receptor, into the planar membrane, as would be expected in a fusion process. The flash rate on ganglioside-containing asolectin membranes increased as the pH was lowered. This is also consistent with the known fusion behavior of influenza virus with cell membranes and with phospholipid vesicles. We conclude that the flashes result from the fusion of individual virions to the planar membrane.

Delay time for influenza virus hemagglutinin-induced membrane fusion depends on hemagglutinin surface density

We have studied the kinetics of low-pH-induced fusion between erythrocyte membranes and membranes containing influenza virus hemagglutinin by using assays based on the fluorescence dequenching of the lipophilic dye octadecylrhodamine. Stopped-flow mixing and fast data acquisition have been used to monitor the early stages of influenza virus fusion. We have compared this with the kinetics observed for fusion of an NIH 3T3 cell line, transformed with bovine papillomavirus, which constitutively expresses influenza virus hemagglutinin (GP4f cells). Virus and GP4f cells both display a pH-dependent time lag before the onset of fluorescence dequenching, but of an order of magnitude difference, ca. 2 s versus ca. 20 s. We have adopted two strategies to investigate whether the difference in lag time reflects the surface density of acid-activated hemagglutinin, able to undergo productive conformational change. (i) Hemagglutinin expressed on the cell surface requires proteolytic cleavage with trypsin from an inactive HAO form; we have limited the extent of proteolysis. (ii) We have used infection of CV-1 cells with a recombinant simian virus 40 bearing the influenza virus hemagglutinin gene. The surface expression of hemagglutinin is a function of time postinfection. For low-pH-induced fusion of both types of cell with erythrocytes, the lag time decreases with increasing hemagglutinin densities. Our results do not indicate a cooperative phenomenon at the level of the principal rate-determining step. We also show in the instance of virus fusion, that the magnitude of the delay time is a function of the target membrane transbilayer lipid distribution. We conclude that for a given amount of pH-activated hemagglutinin per unit area of membrane, the kinetics of fusion is determined by nonspecific physical properties of the membranes involved.

Specific cell surface requirements for the infection of CD4-positive cells by human immunodeficiency virus types 1 and 2 and by Simian immunodeficiency virus

Human CD4 was expressed on a range of mammalian cell lines. CD4+ non-primate cells, derived from rat, hamster, mink, cat, and rabbit, bind recombinant gp120 of human immunodeficiency virus type 1 (HIV-1) but are resistant to HIV-1 infection. CD4 expression on various human, rhesus, and African green monkey cell lines confers differential susceptibilities for HIV-1, HIV-2, and simian immunodeficiency (SIV) strains. For example, CD4+ TE671 rhabdomyosarcoma cells are sensitive to HIV-1 and HIV-2 but resistant to SIV, whereas CD4+ U87 glioma cells are resistant to HIV-1 infection but sensitive to HIV-2 and SIV. HIV-1 infection was not dependent on human major histocompatibility class I expression. Studies of cell fusion and of infection by vesicular stomatitis virus pseudotypes bearing HIV-1 and HIV-2 envelopes showed that the differential cell tropisms of HIV-1, HIV-2, and SIV are determined at the cell surface.

Genetic analysis of the Rous sarcoma virus subgroup D env gene: mammal tropism correlates with temperature sensitivity of gp85

Subgroup D avian sarcoma and leukosis viruses can penetrate a variety of mammalian cells in addition to cells from their natural host, chickens. Sequences derived from the gp85-coding domain within the env gene of a mammal-tropic subgroup D virus (Schmidt-Ruppin D strain of Rous sarcoma virus [SR-D RSV]) and a non-mammal-tropic subgroup B virus (Rous-associated virus type 2) were recombined to map genetic determinants that allow penetration of mammalian cells. The following conclusions were based on host range analysis of the recombinant viruses. (i) The determinants of gp85 that result in the mammal tropism phenotype of SR-D RSV are encoded within the 160 codons that lie 3' of codon 121 from the corresponding amino terminus of the gp85 protein. (ii) Small linear domains of the SR-D RSV gp85-coding domain placed in the subgroup B background did not yield viruses with titers equal to that of the subgroup D virus in a human cell line. (iii) Recombinant viruses that contained subgroup D sequences within the hr1 variable domain of gp85 showed modest-to-significant increases in infectivity on human cells relative to chicken cells. A recombinant virus that contained three fortuitous amino acid substitutions in the gp85-coding domain was found to penetrate the human cell line and give a titer similar to that of the subgroup D virus. In addition, we found that the subgroup D virus, the mutant virus, and recombinant viruses with an increased mammal tropism phenotype were unstable at 42 degrees C. These results suggest that the mammal tropism of the SR-D strain is not related to altered receptor specificity but rather to an unstable and fusogenic viral glycoprotein. A temperature sensitivity phenotype for infectivity of mammalian cells was also observed for another mammal-tropic avian retrovirus, the Bratislava 77 strain of RSV, a subgroup C virus, but was not seen for any other avian retrovirus tested, strengthening the correlation between mammal tropism and temperature sensitivity.

Fusogenic activity of SIV (simian immunodeficiency virus) peptides located in the GP32 NH2 terminal domain

Peptides of 12, 16 and 24 amino acids length corresponding to the NH2 terminal sequence of SIV gp32 were synthesized. Fluorescence energy transfer studies have shown that those peptides can induce lipid mixing of SUV (Small Unilamellar Vesicles) of various compositions at pH 7.4 and 37 degrees C. LUV (Large Unilamellar Vesicles) were shown to undergo fusion, provided they contained PE in their lipid composition. This work is an attempt to determine how the fusogenic activity depends on the structure of the peptide inserted into a lipidic environment. The peptides secondary structure and orientation in the lipid bilayer were determined using Fourier Transform infrared spectroscopy (FTIR). They adopt mainly a beta-sheet conformation in the absence of lipids. After interaction with DOPC SUV, the beta-sheet is partly converted into alpha-helix oriented obliquely with respect to the membrane interface. We bring here evidence that this oblique orientation is a prerequisite to the fusion process.

Fusion activity of influenza virus PR8/34 correlates with a temperature-induced conformational change within the hemagglutinin ectodomain detected by photochemical labeling

Fusion of influenza viruses with membranes is catalyzed by the viral spike protein hemagglutinin (HA). Under mildly acidic conditions (approximately pH 5) this protein undergoes a conformational change that triggers the exposure of the "fusion peptide", the hydrophobic N-terminal segment of the HA2 polypeptide chain. Insertion of this segment into the target membrane (or viral membrane?) is likely to represent a key step along the fusion pathway, but the details are far from being clear. The photoreactive phospholipid 1-palmitoyl-2-[11-[4-[3-(trifluoromethyl)diazirinyl]phenyl] [2-3H]undecanoyl]-sn-glycero-3-phosphocholine ([3H]PTPC/11), inserted into the bilayer of large unilamellar vesicles (LUVs), allowed us to investigate both the interaction of viruses with the vesicles under "prefusion" conditions (pH 5; 0 degrees C) and the fusion process itself occurring at elevated temperatures (greater than 15-20 degrees C) only. Despite the observed binding of viruses to LUVs at pH 5 and 0 degrees C, labeling of HA2 was very weak (less than 0.002% of the radioactivity originally present). In contrast, fusion could be readily monitored by the covalent labeling of that polypeptide chain. We have studied also the effect of temperature on the acid-induced (pH 5) interaction of bromelain-solubilized HA (BHA) with vesicles. Labeling of the BHA2 polypeptide chain was found to show a remarkable correlation with the temperature dependence of the fusion activity of whole viruses. A temperature-induced structural change appears to be critical for both the interaction of BHA with membranes and the expression of fusion activity of intact viruses.

Cholesterol is required for infection by Semliki Forest virus

Semliki Forest virus (SFV) and many other enveloped animal viruses enter cells by a membrane fusion reaction triggered by the low pH within the endocytic pathway. In vitro, SFV fusion requires cholesterol in the target membrane, but the role of cholesterol in vivo is unknown. In this paper, the infection pathway of SFV was studied in mammalian and inset cells substantially depleted of sterol. Cholesterol-depleted cells were unaltered in their ability to bind, internalize, and acidify virus, but were blocked in SFV fusion and subsequent virus replication. Depleted cells could be infected by the cholesterol-independent vesicular stomatitis virus, which also enters cells via endocytosis and low pH-mediated fusion. The block in SFV infection was specifically reversed by cholesterol but not by cholestenone, which lacks the critical 3 beta-hydroxyl group. Cholesterol thus is central in the infection pathway of SFV, and may act in vivo to modulate infection by SFV and other pathogens.

Potassium dependence for sperm-egg fusion in mice

In this study, we examined the potassium requirements for sperm-egg fusion in mouse. Zona-free mouse eggs prepared by the method described by Boldt and Wolf were inseminated with capacitated sperm in culture media containing 0-6 mM extracellular K+, and scored for penetration. Penetration of zona-free eggs was dependent on extracellular K+, with no penetration observed under K(+)-free conditions. Media transfer experiments indicated that the lack of penetration observed was due to effects on fusion, and not on postpenetration events such as sperm head decondensation. To analyze whether the K+ effect was attributable to an effect on the sperm (i.e., occurrence of acrosome reactions), sperm were treated with the Ca2+ ionophore A23187 before insemination. Less than 5% of zona-free eggs were penetrated with ionophore-treated sperm under K(+)-free conditions, suggesting that K+ is required for fusion per se. Addition of ionophore to insemination cultures similarly did not overcome the block to fusion observed under K(+)-free conditions. The potassium channel blockers 4-aminopyridine (0.1-5 mM) and tetraethyl ammonium chloride (5-50 mM) had no inhibitory effect on fusion. These data indicate that extracellular K+ is required for sperm-egg fusion and that this requirement may not involve membrane K+ channels.