Identification of Hendra virus G glycoprotein residues that are critical for receptor binding

Abstract

Hendra virus (HeV) is an emerging paramyxovirus capable of infecting and causing disease in a variety of mammalian species, including humans. The virus infects its host cells through the coordinated functions of its fusion (F) and attachment (G) glycoproteins, the latter of which is responsible for binding the virus receptors ephrinB2 and ephrinB3. In order to identify the receptor binding site, a panel of G glycoprotein constructs containing mutations was generated using an alanine-scanning mutagenesis strategy. Based on a predicted G structure, charged amino acids residing in regions that could be homologous to those in the measles virus H attachment glycoprotein known to be involved in its protein receptor interaction were targeted. Using a coprecipitation-based assay, seven single-amino-acid substitutions in HeV G were identified as having significantly impaired binding to both the ephrinB2 and ephrinB3 viral receptors: D257A, D260A, G439A, K443A, G449A, K465A, and D468A. The impairment of receptor interaction conferred a concomitant diminution in their abilities to promote membrane fusion when coexpressed with F. The G glycoprotein mutants were also recognized by three or more conformation-dependent monoclonal antibodies of a panel of five, were expressed on the cell surface, and retained their abilities to bind and coprecipitate F. Interestingly, some of these mutant G glycoproteins coprecipitated with F more efficiently than wild-type G. Taken together, these data provide strong biochemical and functional evidence that some of these residues could be part of a conformation-dependent, discontinuous, and overlapping ephrinB2 and -B3 binding domain within the HeV G glycoprotein.

FIG. 1.

Expression and receptor binding of HeV G mutants. (A) Plasmids encoding alanine substitution mutants of HeV G were transfected into HeLa-USU cells and metabolically labeled overnight. After a 1.5-h chase, cell lysates were prepared, and equivalent amounts of each lysate were immunoprecipitated with a HeV G-specific rabbit polyclonal antiserum (bottom row) or FC-tagged human (hu.) ephrinB1, murine (mu.) ephrinB2, or human ephrinB3. The HeV F protein was also subjected to the same conditions and served as a negative control for the specificity of protein binding. In the case of the HeV F protein, F-specific antiserum was used as a control (bottom row). The precipitated, metabolically labeled proteins were resolved by SDS-PAGE under reducing conditions and detected by autoradiography. The black line below the lanes containing G439A, N446A, and G449A indicates that they are mutants that were made as a second panel to further explore the beta sheet 4 region. (B) The percentage of WT ephrinB2 or -B3 binding activity for each HeV G mutant is shown. The results were calculated using values obtained from densitometric measurements of autorad bands in comparison to the values obtained for WT HeV G.

FIG. 2.

A schematic model of the HeV G glycoprotein based on the model of Yu et al. (40). Putative disulfide bonds are represented as bridges above the schematic, and beta sheets 1 to 6 of the globular-head domain are depicted beneath the schematic.

FIG. 3.

Binding of HeV G substitution mutants to human (hu.) ephrinB2. (A) Mutants defective in murine ephrinB2 and human ephrinB3 binding were expressed in HeLa-USU cells, and the lysates were precipitated with either soluble, S-epitope-tagged human ephrinB2 plus S-agarose beads (top row) or polyclonal G-specific antiserum plus protein G beads (bottom row) as a control. (B) The percentage of WT human ephrinB2 binding activity for each HeV G mutant is shown. The results were calculated using values obtained from densitometric measurements of autorad bands in comparison to the values obtained for WT HeV G.

FIG. 4.

Cell fusion promotion activities of HeV G mutants. The alanine substitution mutants of HeV G were tested for the ability to promote cell fusion when coexpressed with HeV F using a quantitative reporter gene cell fusion assay. (A) The data shown are the average percentage of WT fusion levels measured for each mutant calculated from two or three separate experiments, using 293T cells as the target population. The bars represent the range from multiple experiments. (B) Aliquots of each effector cell population coexpressing HeV F, along with the various mutants or WT G, were subjected to flow cytometric analysis using a FITC-conjugated antibody specific for the myc epitope tag on G. The relative expression levels of the various mutants compared to WT G were then used to normalize the β-Gal readings from this reporter gene assay. The data are presented as the predicted percentage of WT activity each mutant would demonstrate if the mutant were expressed on the surfaces of cells to the same extent as the WT. Each mutant effector population was tested for cell surface expression and fusion with several target cell lines on at least two separate occasions; a representative experiment with 293T target cells is shown.

FIG. 5.

Interaction of HeV G substitution mutants with HeV F. (A) HeV G substitution mutants were coexpressed with HeV F in HeLa-USU cells. Lysates were immunoprecipitated (IP) with rabbit polyclonal F-specific antiserum and then blotted with mouse polyclonal G-specific antiserum as a test for the abilities of the various G mutants to interact and coprecipitate with HeV F (top row) or immunoprecipitated with rabbit polyclonal G-specific antiserum and then probed with mouse polyclonal G-specific antiserum as a control for the relative expression level of each mutant (bottom row). HeV F (F control) and WT HeV G (G control) were also expressed singly in the absence of the partner glycoprotein and subjected to the same immunoprecipitation and blotting conditions to illustrate the specificity of the coprecipitation interaction. (B) The relative F-binding ability of each HeV G mutant is shown in comparison to that of WT HeV G. The results were calculated using values obtained from densitometric measurements of autorad bands in comparison to the values obtained for WT HeV G.