The structure and receptor binding properties of the 1918 influenza hemagglutinin.

The 1918 influenza pandemic resulted in about 20 million deaths. This enormous impact, coupled with renewed interest in emerging infections, makes characterization of the virus involved a priority. Receptor binding, the initial event in virus infection, is a major determinant of virus transmissibility that, for influenza viruses, is mediated by the hemagglutinin (HA) membrane glycoprotein. We have determined the crystal structures of the HA from the 1918 virus and two closely related HAs in complex with receptor analogs. They explain how the 1918 HA, while retaining receptor binding site amino acids characteristic of an avian precursor HA, is able to bind human receptors and how, as a consequence, the virus was able to spread in the human population.

Influenza fusion peptides.

The 'fusion peptides' of a group of enveloped viruses that includes influenza, paramyxo-, retro-and filo-viruses are N-terminal regions of their membrane fusion proteins generated by cleavage of non-functional precursors. For the influenza membrane fusion protein, haemagglutinin (HA), the three-dimensional structures of precursor HA, cleaved HA and fusion-activated HA show that the fusion peptides are located in different positions in all three forms and adopt different structures. Analyses of mutant HAs with changes in fusion peptide sequence indicate the importance of specific residues for membrane-fusion activity and suggest a structure for the fusion peptide in a fusion-active molecule.

Studies on influenza haemagglutinin fusion peptide mutants generated by reverse genetics.

Influenza haemagglutinin (HA) is responsible for fusing viral and endosomal membranes during virus entry. In this process, conformational changes in the HA relocate the HA(2) N-terminal 'fusion peptide' to interact with the target membrane. The highly conserved HA fusion peptide shares composition and sequence features with functionally analogous regions of other viral fusion proteins, including the presence and distribution of glycines and large side-chain hydrophobic residues. HAs with mutations in the fusion peptide were expressed using vaccinia virus recombinants to examine the requirement for fusion of specific hydrophobic residues and the significance of glycine spacing. Mutant HAs were also incorporated into infectious influenza viruses for analysis of their effects on infectivity and replication. In most cases alanine, but not glycine substitutions for the large hydrophobic residues, yielded fusion-competent HAs and infectious viruses, suggesting that the conserved spacing of glycines may be structurally significant. When viruses containing alanine substitutions for large hydrophobic residues were passaged, pseudoreversion to valine was observed, indicating a preference for large hydrophobic residues at specific positions. Viruses were also obtained with serine, leucine or phenylalanine as the N-terminal residue, but these replicated to significantly lower levels than wild-type virus with glycine at this position.

Mechanisms of cell entry by influenza virus.

A wide range of viruses, including many human and animal pathogens representing various taxonomic groups, contain genomes that are enclosed in lipid envelopes. These envelopes are generally acquired in the final stages of assembly, as viruses bud from regions of the membrane of the infected cell at which virally encoded membrane proteins have accumulated. The viruses procure their membranes during this process and mature particles 'pinch off' from the cellular membranes. Under most circumstances, initiation of another round of infection is dependent on two critical functions supplied by the envelope proteins. The virus must bind to cell-surface receptors of a new host cell, and fusion of the viral and cellular membranes must occur to transfer the viral genome into the cell. Enveloped viruses have evolved a variety of mechanisms to execute these two basic functions. Owing to their relative simplicity, studies of binding and fusion using enveloped viruses and their components have contributed significantly to the overall understanding of receptor-ligand interactions and membrane fusion processes - fundamental activities involved in a plethora of biological functions.

Interaction of mutant influenza virus hemagglutinin fusion peptides with lipid bilayers: probing the role of hydrophobic residue size in the central region of the fusion peptide.

The amino-terminal region of the membrane-anchored subunit of influenza virus hemagglutinin, the fusion peptide, is crucial for membrane fusion of this virus. The peptide is extruded from the interior of the protein and inserted into the lipid bilayer of the target membrane upon induction of a conformational change in the protein by low pH. Although the effects of several mutations in this region on the fusion behavior and the biophysical properties of the corresponding peptides have been studied, the structural requirements for an active fusion peptide have still not been defined. To probe the sensitivity of the fusion peptide structure and function to small hydrophobic perturbations in the middle of the hydrophobic region, we have individually replaced the alanine residues in positions 5 and 7 with smaller (glycine) or bulkier (valine) hydrophobic residues and measured the extent of fusion mediated by these hemagglutinin constructs as well as some biophysical properties of the corresponding synthetic peptides in lipid bilayers. We find that position 5 tolerates a smaller and position 7 a larger hydrophobic side chain. All peptides contained segments of alpha-helical (33-45%) and beta-strand (13-16%) conformation as determined by CD and ATR-FTIR spectroscopy. The order parameters of the peptide helices and the lipid hydrocarbon chains were determined from measurements of the dichroism of the respective infrared absorption bands. Order parameters in the range of 0.0-0.6 were found for the helices of these peptides, which indicate that these peptides are most likely aligned with their alpha-helices at oblique angles to the membrane normal. Some (mostly fusogenic) peptides induced significant increases of the order parameter of the lipid hydrocarbon chains, suggesting that the lipid bilayer becomes more ordered in the presence of these peptides, possibly as a result of dehydration at the membrane surface.

Role of hemagglutinin cleavage for the pathogenicity of influenza virus.

Although human epidemics of influenza occur on nearly an annual basis and result in a significant number of "excess deaths," the viruses responsible are not generally considered highly pathogenic. On occasion, however, an outbreak occurs that demonstrates the potential lethality of influenza viruses. The human pandemic of 1918 spread worldwide and killed millions, and the limited human outbreak of highly pathogenic avian viruses in Hong Kong at the end of 1997 is a warning that this could happen again. In avian species such as chickens and turkeys, several outbreaks of highly pathogenic influenza viruses have been documented. Although the reason for the lethality of the human 1918 viruses remains unclear, the pathogenicity of the avian viruses, including those that caused the human 1997 outbreak, relates primarily to properties of the hemagglutinin glycoprotein (HA). Cleavage of the HA precursor molecule HA0 is required to activate virus infectivity, and the distribution of activating proteases in the host is one of the determinants of tropism and, as such, pathogenicity. The HAs of mammalian and nonpathogenic avian viruses are cleaved extracellularly, which limits their spread in hosts to tissues where the appropriate proteases are encountered. On the other hand, the HAs of pathogenic viruses are cleaved intracellularly by ubiquitously occurring proteases and therefore have the capacity to infect various cell types and cause systemic infections. The x-ray crystal structure of HA0 has been solved recently and shows that the cleavage site forms a loop that extends from the surface of the molecule, and it is the composition and structure of the cleavage loop region that dictate the range of proteases that can potentially activate infectivity. Here influenza virus pathogenicity is discussed, with an emphasis on the role of HA0 cleavage as a determining factor.

Structure of the hemagglutinin precursor cleavage site, a determinant of influenza pathogenicity and the origin of the labile conformation.

The membrane fusion potential of influenza HA, like many viral membrane-fusion glycoproteins, is generated by proteolytic cleavage of a biosynthetic precursor. The three-dimensional structure of ectodomain of the precursor HA0 has been determined and compared with that of cleaved HA. The cleavage site is a prominent surface loop adjacent to a novel cavity; cleavage results in structural rearrangements in which the nonpolar amino acids near the new amino terminus bury ionizable residues in the cavity that are implicated in the low-pH-induced conformational change. Amino acid insertions at the cleavage site in HAs of virulent avian viruses and those of viruses isolated from the recent severe outbreak of influenza in humans in Hong Kong would extend this surface loop, facilitating intracellular cleavage.

Studies of the binding properties of influenza hemagglutinin receptor-site mutants.

Site-specific mutations have been made in the influenza hemagglutinin (HA) receptor binding site to assess the contribution of individual amino acid residues to receptor recognition. Screening of mutant HAs, expressed using recombinant vaccinia virus-infected cells, for their abilities to bind human erythrocytes indicated that substitutions involving conserved residues Y98F, H183F, and L194A severely restricted binding and that the substitution W153A prevented cell surface expression of HA. Mutation of residues E190 and S228 that are in positions to form hydrogen bonds with the 9-OH of sialic acid appeared to increase erythrocyte binding slightly, as did the substitution G225R. Substitutions of other residues that are directly or indirectly involved in receptor binding, S136T, S136A, Y195F, G225D, and L226P, had intermediate effects on binding between these two extremes. Estimates of changes in receptor binding specificity based on inhibition of binding to erythrocytes by nonimmune horse sera indicated that mutants G225R and L226P, unlike wild-type HA, were not inhibited; Y195F and G225D mutants were, like wild type, inhibited; and erythrocyte binding by mutants S136A, S136T, E190A, and S228G was only partially inhibited. Viruses containing mutant HAs Y98F, S136T, G225D, and S228G that cover the range of erythrocyte binding properties observed were also constructed by transfection. All four transfectant viruses replicated in MDCK cells and embryonated hens' eggs as efficiently as wild-type X-31 virus, although the Y98F mutant virus was unable to agglutinate erythrocytes. Mutant MDCK cells that have reduced levels of cell surface sialic acids were susceptible to infection by S136T, G225D, and S228G transfectant viruses and by wild type but not by the Y98F transfectant virus.

Adaptation of egg-grown and transfectant influenza viruses for growth in mammalian cells: selection of hemagglutinin mutants with elevated pH of membrane fusion.

A series of eight transfectant influenza viruses was generated by reverse genetics for studies of the palmitylated cysteine residues in the cytoplasmic tail of the hemagglutinin glycoprotein (HA). Following amplification of these viruses in MDCK cells we found that all had developed an elevated pH of membrane fusion--an unexpected result since previous mutant HA expression studies had shown that substitutions of the cysteine residues had no effect on fusion properties. Sequence analyses revealed that each of the viruses had at least one additional mutation in the ectodomain of HA which was responsible for the increase in fusion pH. Similarly, when we passaged egg-grown wild-type X-31 virus in three different lines of MDCK cells or in MDBK cells, high pH fusion mutants were selected within a few passages in every case. The locations of the substitutions in the HA structure are in or near the "fusion peptide" or at subunit interfaces throughout the length of the trimer--reminiscent of the changes selected in earlier studies on amantadine resistance. The observation that passage of certain viruses in mammalian cells can result in the selection of mutants with elevated fusion pH has potential implications both for reverse genetic experiments and, perhaps more importantly, for the choice of substrates for propagation of vaccine viruses.

Feline immunodeficiency virus (FIV) infection in cats at necropsy: a serological study.

Sera collected post mortem during a 6-month period from cats were tested for feline immunodeficiency virus (FIV)-specific antibodies by (1) an enzyme-linked immunosorbent assay (ELISA), (2) an indirect peroxidase-based immunocytological test (IP), (3) a Western immunoblotting (WB) method with FIV-infected cell lysates, and (4) a WB method with purified viral antigen. All four methods were capable of detecting FIV-specific antibodies in haemolysed sera. However, the ELISA showed the lowest "positive predictive value" (PVpos = 22%) followed by the IP (PVpos 50-60%). Serum was FIV antibody-positive in 6% (15/255) of all cats examined. The mean age of seropositive cats was 9 years (4 years among seronegative cases) and the male-to-female ratio in such cats was 1.8 to 1 (overall ratio 0.8 to 1). Forty per cent of the seropositive cats were in the final phase of acquired immune deficiency syndrome. Feline leukaemia virus (FeLV) predominated among viral co-infections. It was concluded that (1) a combination of the IP and WB reliably detected FIV-specific antibodies in sera collected post mortem, and (2) at post-mortem examination, cats from high-risk groups (male, > 5 years old, hypercellular bone marrow) were frequently infected with FIV.

Studies using double mutants of the conformational transitions in influenza hemagglutinin required for its membrane fusion activity.

Amino acid substitutions widely distributed throughout the influenza hemagglutinin (HA) influence the pH of its membrane fusion activity. We have combined a number of these substitutions in double mutants and determined the effects on the pH of fusion and on the pH at which the refolding of HA required for fusion occurs. By analyzing combinations of mutations in three regions of the metastable neutral-pH HA that are rearranged at fusion pH we obtain evidence for both additive and nonadditive effects and for an apparent order of dominance in the effects of amino acid substitutions in particular regions on the pH of fusion. We conclude that there are at least three components in the structural transition required for membrane fusion activity and consider possible pathways for the transition in relation to the known differences between neutral and fusion pH HA structures.

Equine adenocarcinomas of the large intestine with osseous metaplasia.

Large intestinal adenocarcinoma with osseous metaplasia was diagnosed in two horses, a 15-year-old standard bred gelding and a 9-year-old Haflinger mare. Clinically, both animals had displayed weight loss and anaemia. A presumptive diagnosis of abdominal neoplasia was made and the horses were humanely killed. At necropsy, the gelding and the mare were found to have ulcerated tumours growing into the lumen of the caecum and colon, respectively. In the mare, the mass extended through the mesocolon and was evident in the left dorsal and ventral colon. Histopathologically, the tumours consisted of well-differentiated cords of single-layered columnar to cuboidal epithelial cells. Mitotic figures were very uncommon. In both lesions, well-formed bony spicules and osteoid were present in the fibrovascular stroma. The tumours were well-demarcated from surrounding mucosal tissue but had invaded the intestinal wall. Metastases were not observed.

Studies of the membrane fusion activities of fusion peptide mutants of influenza virus hemagglutinin.

Influenza virus hemagglutinin (HA) fuses membranes at endosomal pH by a process which involves extrusion of the NH2-terminal region of HA2, the fusion peptide, from its buried location in the native trimer. We have examined the amino acid sequence requirements for a functional fusion peptide by determining the fusion capacities of site-specific mutant HAs expressed by using vaccinia virus recombinants and of synthetic peptide analogs of the mutant fusion peptides. The results indicate that for efficient fusion, alanine can to some extent substitute for the NH2-terminal glycine of the wild-type fusion peptide but that serine, histidine, leucine, isoleucine, or phenylalanine cannot. In addition, mutants containing shorter fusion peptides as a result of single amino acid deletions are inactive, as is a mutant containing an alanine instead of a glycine at HA2 residue 8. Substitution of the glycine at HA2 residue 4 with an alanine increases the pH of fusion, and valine-for-glutamate substitutions at HA2 residues 11 and 15 are without effect. We confirm previous reports on the need for specific HAo cleavage to generate functional HAs, and we show that both inappropriately cleaved HA and mutant HAs, irrespective of their fusion capacities, upon incubation at low pH undergo the structural transition required for fusion.

Electron microscopy of antibody complexes of influenza virus haemagglutinin in the fusion pH conformation.

Activation of the membrane fusion potential of influenza haemagglutinin (HA) at endosomal pH requires changes in its structure. X-ray analysis of TBHA2, a proteolytic fragment of HA in the fusion pH conformation, indicates that at the pH of fusion the 'fusion peptide' is displaced by > 10 nm from its location in the native structure to the tip of an 11 nm triple-stranded coiled coil, and that the formation of this structure involves extensive re-folding or reorganization of HA. Here we examine the structure of TBHA2 with the electron microscope and compare it with the fusion pH structure of HA2 in virosomes, HA2 in aggregates formed at fusion pH by the soluble, bromelain-released ectodomain BHA and HA2 in liposomes with which BHA associates at fusion pH. We have oriented each HA2 preparation for comparison, using site-specific monoclonal antibodies. We conclude that the structural changes in membrane-anchored and soluble HA preparations at the pH of fusion appear to be the same; that in the absence of a target membrane, the 'fusion peptide' of HA in virosomes associates with the virosome membrane so that HA2 is membrane bound at both N- and C-termini, which implies that inversion of the re-folded HA can occur; and that the structural changes observed by X-ray analysis do not result from the proteolytic digestions used in the preparation of TBHA2.

Lack of evidence for proofreading mechanisms associated with an RNA virus polymerase.

The in vitro fidelity of the virion-associated RNA polymerase of vesicular stomatitis virus was quantitated for a single conserved viral RNA site and the usual high in vitro base misincorporation error frequencies (approx. 10(-3)) were observed at this (guanine) site. We sought evidence for RNA 3'-->5' exonuclease proofreading mechanisms by varying the concentrations of the next nucleoside triphosphate, by incorporation of nucleoside[1-thio]triphosphate analogues of the four natural RNA nucleosides, and by varying the concentrations of pyrophosphate in the in vitro polymerase reaction. None of these perturbations greatly affected viral RNA polymerase fidelity at the site studied. These results fail to show evidence for proofreading exonuclease activity associated with the virion replicase of an RNA virus. They suggest that RNA virus replication might generally be error-prone, because RNA replicase base misincorporations are proofread very inefficiently or not at all.

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.

RNA virus populations as quasispecies.

RNA virus mutation frequencies generally approach maximum tolerable levels, and create complex indeterminate quasispecies populations in infected hosts. This usually favors extreme rates of evolution, although periods of relative stasis or equilibrium, punctuated by rapid change may also occur (as for other life forms). Because complex quasispecies populations of RNA viruses arise probabilistically and differentially in every host, their compositions and exact roles in disease pathogenesis are indeterminate and their directions of evolution, and the nature and timing of "new" virus outbreaks are unpredictable.

Amantadine selection of a mutant influenza virus containing an acid-stable hemagglutinin glycoprotein: evidence for virus-specific regulation of the pH of glycoprotein transport vesicles.

Mutants of influenza Rostock virus (H7N1 subtype) were selected for resistance to amantadine hydrochloride at concentrations of the antiviral drug known to affect the function of the virus M2 transmembrane protein. Sequence analysis revealed that three mutants had no changes in M2 but contained a lysine to isoleucine substitution in the hemagglutinin (HA) membrane glycoprotein at position 58 of HA2. The mutant viruses were found to fuse membranes at a pH value 0.7 lower than wild type and to exhibit changes in the conformation of their HAs specifically at the lower pH. The homologous lysine to isoleucine substitution was introduced by site-specific mutagenesis into the HA of X-31 influenza virus (H3 subtype), which was expressed by using vaccinia virus recombinants. The expressed HA also mediated membrane fusion and changed in conformation at a pH value 0.7 lower than wild type. These results indicate that increased acid stability of the HA obviates the consequences of the inhibition of M2 function by amantadine and provide further evidence for the role of M2 in regulating the pH of vesicles involved in glycoprotein transport to the cell surface.

Deacylation of the hemagglutinin of influenza A/Aichi/2/68 has no effect on membrane fusion properties.

The effects of deacylating the H3 influenza hemagglutinin (HA) on its membrane fusion activity were investigated. Chemical deacylation caused no change in the ability of HA to fuse liposomes in vitro. Site-specific mutagenesis of the three cysteine residues in the cytoplasmic tail singly, or in combination, showed that all three were palmitoylated. Substitution of one, two, or all three cysteines with serine and subsequent lack of palmitoylation at mutated sites had no effect on the pH of the conformational change in HA required for fusion activity or the extent of fusion activity.