Effects of glycosylation on the properties and functions of influenza virus hemagglutinin.

The influenza virus A hemagglutinin (HA) is a trimeric glycoprotein that contains 3-9 N-linked glycosylation sequons per subunit, depending on the strain. The location of these sites is determined by the nucleotide sequence of the HA gene, and, since the viral genome is replicated by an error-prone RNA polymerase, mutations, which add or remove glycosylation sites, occur at a high frequency. Mutations that are not lethal to the virus add to the structural diversity of the virus population. Factors that determine the glycosylation of the HA are reviewed herein, as are the effects of host-specific glycosylation on receptor binding, fusion activity, and antigenic properties of the virus. Effects of host-specific glycosylation and selection on virulence and on vaccine efficacy and surveillance are discussed. In addition, inadequacies in our understanding of HA glycosylation and its effects on host range are emphasized.

The glycosylation of the influenza A virus hemagglutinin by mammalian cells. A site-specific study.

We have characterized the glycans at individual sites on the hemagglutinin of three influenza A variants to obtain information on the role of cell-specific glycosylation in determining the receptor binding properties of this virus. The variants differ in whether they have a glycosylation site at residue 129 on the tip of the hemagglutinin and whether amino acid 184 (near to the receptor binding site) is His or Asn. We found that all sites on each variant are glycosylated in Madin-Darby bovine kidney cells, that the glycosylation is site-specific, and that the glycans at the same site in each variant are highly similar. One site that is buried in the hemagglutinin trimer contains only oligomannose glycans. The remaining sites carry complex glycans of increasing size as the distance of the site from the viral membrane decreases. Most of these complex glycans are terminated with alpha-galactose residues, a consequence in bovine cells of the removal of terminal sialic acids by the viral neuraminidase. Although the glycans at residue 129 are among the smallest on the molecule, they are large enough to reach the receptor binding pocket on their own and adjacent monomers. The results suggest that the reduction in receptor binding observed with Madin-Darby bovine kidney cell-grown virus is due to the combined effect of large complex glycans at the tip of the hemagglutinin and a His to Asn substitution close to the receptor binding pocket.

The hemagglutinins of duck and human H1 influenza viruses differ in sequence conservation and in glycosylation.

We determined the deduced amino acid sequences of two H1 duck influenza A virus hemagglutinins (HAs) and found that the consensus sequence of the HA, determined directly from virus recovered from the intestinal tract, remains unchanged through many generations of growth in MDCK cells and chicken embryos. These two duck viruses differ from each other by 5 amino acids and from A/Dk/Alberta/35/1976 (F. J. Austin, Y. Kawaoka, and R. G. Webster, J. Gen. Virol. 71:2471-2474, 1990) by 9 and 12 amino acids, most of which are in the HA1 subunit. They are antigenically similar to each other but different from the Alberta virus. We compared these H1 duck HAs with the HAs of human isolates to identify structural properties of this viral glycoprotein that are associated with host range. By comparison to the human H1 HAs, the duck virus HA sequences are highly conserved as judged by the small fraction of nucleotide differences between strains which result in amino acid substitutions. However, the most striking difference between these duck and human HAs is in the number and distribution of glycosylation sites. Whereas duck and swine viruses have four and five conserved glycosylation sites per HA1 subunit, none of which are on the tip of the HA, all human viruses have at least four additional sites, two or more of which are on the tip of the HA. These findings stress the role of glycosylation in the control of host range and suggest that oligosaccharides on the tip of the HA are important to the survival of H1 viruses in humans but not in ducks or swine.

Viral glycoprotein heterogeneity-enhancement of functional diversity.

Variations in the amino acid sequence of RNA virus envelope glycoproteins can cause changes in their antigenicity and can alter the host-cell tropism of the virus and the degree of virulence which it exhibits. Such changes may alter the course and outcome of viral diseases, either directly because of changes in the biological properties of the glycoproteins or indirectly through effects on immune surveillance and vaccine efficacy. The nature and extent of glycosylation of the surface glycoproteins of RNA viruses have also been implicated in such phenotypic alterations. It follows therefore that the 'plasticity' of the viral genome and the host-encoded glycosylation machinery combine to create populations of highly diverse viruses. This diversity is considered to be responsible for survival of these viruses in a variety of biological niches and for their ability to overcome the inhibitory effects of neutralizing antibodies and antiviral agents. In this article we discuss the implications of the inter-relationship between these two mechanisms for the generation of diversity.

Single amino acid substitutions in the hemagglutinin can alter the host range and receptor binding properties of H1 strains of influenza A virus.

We have previously characterized an influenza A (H1N1) virus which has host-dependent growth and receptor binding properties and have shown that a mutation which removes an oligosaccharide from the tip of the hemagglutinin (HA) by changing Asn-129 to Asp permits this virus to grow to high titer in MDBK cells, (C. M. Deom, A. J. Caton, and I. T. Schulze, Proc. Natl. Acad. Sci. USA 83:3771-3775, 1986). We have now isolated monoclonal antibodies specific for the mutant HA and have used escape mutants to identify alterations in HA sequence which reduce virus yields from MDBK cells without reducing those from chicken embryo fibroblasts. Two types of escape mutants which grow equally well in chicken embryo fibroblasts were obtained. Those with the parent phenotype contain Asn at residue 129 and are glycosylated at that site. Those with the mutant phenotype are unchanged at residue 129 but have a Gly to Glu substitution at residue 158, which is close to residue 129 on the HA1 subunit. Binding assays with neoglycoproteins containing N-acetylneuraminic acid in either alpha 2,3 or alpha 2,6 linkage to galactose showed that the MDBK-synthesized oligosaccharides at Asn-129 reduce binding to both of these receptors, leaving the HA's preference for alpha 2,6 linkages unchanged. Glu at residue 158 greatly reduces binding to both receptors without reducing virus yields from MDBK cells. We conclude that changes in the receptor binding properties of the HA can result either from direct alteration of the HA protein by host cell glycosylation or from mutations in the HA gene and that these changes generate heterogeneity that can contribute to the survival of influenza A virus populations in nature.

Sequence of an influenza virus hemagglutinin determined directly from a clinical sample.

The sequence of the HA1 region of the hemagglutinin gene of an influenza virus has been determined without growing the virus in eggs or in cultured cells. The virus used was an H1 strain of influenza A from a clinical specimen taken from a patient in 1987. RNA was extracted directly from virus that had been sedimented out of the transport medium in which the sample had been stored. DNA copies of the hemagglutinin gene, obtained by reverse transcription, were then amplified by the polymerase chain reaction and were sequenced by the dideoxy termination method. The deduced amino acid sequence is highly similar to that of other H1 viruses that had been isolated at about the same time and cultured for a limited number of passages in eggs. Furthermore, the HA1 sequence of progeny virus from this isolate obtained after one passage in chicken embryos is identical to that of the virus obtained directly from the nasopharynx. The results suggest that H1 isolates that have been grown for a limited number of passages in embryonated eggs have HA1 subunits that faithfully represent the virus population in the clinical samples from which they were derived.

Effect of influenza B virus on nutrient transport in cultured epithelial cells.

The possibility that influenza virus could induce changes in membrane permeability to nutrients ordinarily concentrated within the cell was examined. Madin-Darby canine kidney cells were infected with egg-grown influenza B virus at 37 degrees C and pH 7.4 (a condition in which influenza virus enters cells by endocytosis). Control cells were mock-infected with allantoic fluid from chick embryos. Transport of phosphate, 2-deoxyglucose, and alpha-aminoisobutyric acid was measured at various intervals, 0 to 10 hours after infection. Uptake of alpha-aminoisobutyric acid and phosphate by infected cells was inhibited at 2 hours as compared with controls, whereas at 6 to 10 hours, the uptake of all nutrients was higher in infected cells. Infected cells preloaded with phosphate or 2-deoxyglucose did not demonstrate increased release of these nutrients. Thus, the virally induced inhibition of uptake early in infection is not a consequence of loss of membrane integrity. Transport studies were also performed in cells with prebound virus exposed to pH 5.0 for 60 seconds at 37 degrees C and then incubated at pH 7.4, at 37 degrees C. Under these conditions, influenza A viruses are known to enter the cell membrane by fusing directly with it and to initiate cell to cell fusion as well. We demonstrated that influenza B virus also caused cell fusion under these conditions. In contradistinction to studies described above at pH 7.4, fused, infected cells demonstrated both marked release and diminished uptake of nutrients as compared with controls. We conclude that influenza B virus does have an effect on host cell membrane permeability; the type of effect seen is markedly influenced by factors known to determine mode of virus entry into the cell.

Host cell-mediated selection of a mutant influenza A virus that has lost a complex oligosaccharide from the tip of the hemagglutinin.

During serial passage in Madin-Darby bovine kidney (MDBK) cells, a substrain of influenza virus A/WSN is lost from the population and is replaced by a mutant virus with altered host cell binding properties. This selection does not occur during growth in chicken embryo fibroblasts (CEF). It occurs during growth in MDBK cells because the parental virus produced by these cells has a dramatically reduced affinity for cellular receptors [Crecelius, D.M., Deom, C. M. & Schulze, I. T. (1984) Virology 139, 164-177]. We have now compared the hemagglutinin (HA) subunits, HA1 and HA2, of the parent and mutant viruses by NaDodSO4/PAGE and have found that when the viruses are grown in either host cell the HA1 subunit of the mutant is smaller than that of the parent virus. The nonglycosylated HAs, made in the presence of tunicamycin, have the same apparent molecular weight, indicating that the HA1 subunit of the mutant virus contains less carbohydrate than that of the parent. This reduction in carbohydrate content was observed with 11 independently derived mutants that had been selected by growth in MDBK cells. The nucleotide sequence of the HA gene of the parent and mutant viruses indicates that there are five potential glycosylation sites on the parent HA1 subunit and four on the mutant and that the mutation responsible for this difference is a single base change that eliminates the glycosylation site at amino acid 125 of the parent HA1 subunit. Treatment of the parent and mutant HAs from both cell sources with endo-beta-N-acetylglucosaminidases F and H showed that the HA1 of the parent virus has four complex and one high-mannose oligosaccharides, whereas that of the mutant virus has three complex and one high-mannose oligosaccharides. Thus, all of the potential sites on both HA1 subunits are glycosylated. We conclude that the oligosaccharide attached to amino acid 125 of the parent HA by MDBK cells can reduce the affinity of the virus for cellular receptors and that the mutant virus has a higher affinity than the parent because the mutant HA is not glycosylated at that site. Since amino acid 125 of the parent HA is glycosylated by both CEF and MDBK cells, we further conclude that the host-determined structure of the oligosaccharide at that site affects the affinity of the parent virus for cellular receptors and, thereby, determines whether the mutant virus will have a growth advantage.

Oligosaccharide composition of an influenza virus hemagglutinin with host-determined binding properties.

We have previously reported that the binding properties of the hemagglutinin (HA) of the WSN-F strain of influenza A are affected by the cells in which the virus is grown (Crecelius, D. M., Deom, C. M., and Schulze, I.T. (1984) Virology 139, 164-177); at 37 degrees C chick embryo fibroblast-grown F virus has a greater affinity for host cells than does the same virus grown in Madin-Darby bovine kidney (MDBK) cells. In an attempt to explain this host-determined property, we have characterized the carbohydrate put onto the viral HA by these two cells. Experiments using tunicamycin indicate that the HA made by MDBK cells contains about 4000 daltons of carbohydrate in excess of that on the HA from chick embryo fibroblast. Serial lectin affinity chromatography of the asparagine-linked oligosaccharides on the HA subunits, HA1 and HA2, detected a number of host-dependent differences in the complex oligosaccharides. Both HA1 and HA2 from MDBK cells contained more highly branched (i.e. tri- and tetraantennary) complex oligosaccharides than did the subunits from chick embryo fibroblasts. In addition, the HA subunits from the two sources differed in the amount of galactose-containing "bisected" complex oligosaccharides and in the presence of certain fucosylated triantennary oligosaccharides. Profiles of the asparagine-linked oligosaccharides from the host cells did not show these differences, indicating that the HA subunit profiles were not necessarily representative of the structures found on the cellular glycoproteins. The data support the conclusion that bulky oligosaccharides on the MDBK-HA subunits of WSN-F reduce the affinity of the virus for cellular receptors.

Characterization of host cell binding variants of influenza virus by monoclonal antibodies.

We have previously shown that a plaque-type mutant of influenza virus A/WSN has a growth advantage in MDBK cells because its hemagglutinin (HA) has a greater affinity for host cell receptors than does the HA of the parent virus. We show here that the mutant is also less sensitive than the parent to neutralization by antibodies to epitopes in at least two regions on the HA. WSN-specific monoclonal antibodies which had higher radioimmunoassay (RIA) titers against the parent than the mutant virus also had higher plaque inhibition (PI) and hemagglutination inhibition (HI) titers. In contrast, cross-reacting antibodies bound equally well to the parent and mutant viruses as judged by RIA but those which bound to the Cb region of the HA exhibited higher PI and HI titers against the parent virus. The results suggest that preferential neutralization of the parental virus by antibodies can contribute to the selective advantage of mutants which have increased affinity for cellular receptors.

Biological properties of a hemagglutinin mutant of influenza virus selected by host cells.

Chick embryo fibroblast (CEF)-grown stocks of the WSN strain of influenza A(HINI) contain two variants which were designated F and C for fuzzy and clear plaque morphology on Madin-Darby bovine kidney (MDBK) cells. During growth in MDBK cells plaque-isolated F virus was completely replaced by C virus (L. Noronha-Blob and I.T. Schulze (1976), Virology 69, 314-322). The parental (F) and the mutant (C) viruses contain hemagglutinins which differ in their ability to bind to host cells. In addition, the host cells from which the purified viruses are obtained affect their binding properties. Thus, as compared to MDBK-grown F virus (FBK), MDBK-grown C virus (CBK) produced high amounts of mRNA and high virus yields in MDBK cells. CBK had greater affinity for SA alpha 2,3Gal and SA alpha 2,6Gal linkages on derivatized human erythrocytes than did FBK, independent of whether neuraminidase was present on the virions. CBK was also resistant to components of calf serum which inhibited FBK hemagglutination at 37 degrees. As compared to FBK, CBK had increased ability to bind to both MDBK cells and CEF at 37 degrees in the presence or absence of an inhibitor of neuraminidase. In addition, when cells with virus bound at 0 degrees were transferred to 37 degrees, CBK remained cell associated whereas about 80% of FBK dissociated from both cells. Thus, mutation from F to C increased the ability of the virus to associate with MDBK cell receptors. Studies carried out with F and C viruses from both cells indicated that the expression of the mutation depended in part on the host cells in which the virus was grown and in part on the cells used to measure the binding properties. A model relating these observations to selection of HA variants in nature is presented.