Roles of conserved residues in the receptor binding sites of human parainfluenza virus type 3 HN protein

Human parainfluenza virus type 3 (hPIV-3) entry and intrahost spread through membrane fusion are initiated by two envelope glycoproteins, hemagglutinin-neuraminidase (HN) and fusion (F) protein. Binding of HN protein to the cellular receptor via its receptor-binding sites triggers conformational changes in the F protein leading to virus-cell fusion. However, little is known about the roles of individual amino acids that comprise the receptor-binding sites in the fusion process. Here, residues R192, D216, E409, R424, R502, Y530 and E549 located within the receptor-binding site Ⅰ, and residues N551 and H552 at the putative site Ⅱ were replaced by alanine with site-directed mutagenesis. All mutants except N551A displayed statistically lower hemadsorption activities ranging from 16.4% to 80.2% of the wild-type (wt) level. With standardization of the number of bound erythrocytes, similarly, other than N551A, all mutants showed reduced fusogenic activity at three successive stages: lipid mixing (hemifusion), content mixing (full fusion) and syncytium development. Kinetic measurements of the hemifusion process showed that the initial hemifusion extent for R192A, D216A, E409A, R424A, R502A, Y530A, E549A and H552A was decreased to 69.9%, 80.6%, 71.3%, 67.3%, 50.6%, 87.4%, 84.9% and 25.1%, respectively, relative to the wt, while the initial rate of hemifusion for the E409A, R424A, R502A and H552A mutants was reduced to 69.0%, 35.4%, 62.3%, 37.0%, respectively. In addition, four mutants with reduced initial hemifusion rates also showed decreased percentages of F protein cleavage from 43.4% to 56.3% of the wt. Taken together, Mutants R192A, D216A, E409A, R424A, R502A, Y530A, E549A and H552A may lead to damage on the fusion activity at initial stage of hemifusion, of which decreased extent and rate may be associated with impaired receptor binding activity resulting in the increased activation barrier of F protein and the cleavage of it, respectively.

Role of the head-stalk linker region (aa115-122) of Newcastle disease virus haemagglutinin-neuraminidase in regulating fusion triggering

To gain insights into the role of the head-stalk linker region in the fusion triggering, we constructed mutants by deleting or substituting the linker region (115-NGAANNSG-122) of Newcastle disease virus (NDV) haemagglutinin-neuraminidase (HN) with the corresponding sequences of other paramyxoviruses. The results showed that these HN mutants exhibited different levels of fusion-triggering activity, but most of them maintained comparable levels with wide-type HN in both receptor recognition and neuraminidase activity. We tried to figure out reasons for fusion alteration through assessing the expression and the oligomeric state of HN mutants. Moreover, four mutants with significant fusion changes were introduced into the headless HN stem (HN1-123) to intensively investigate the role of the linker region in fusion triggering. Consequently, the stability of HN oligomers and the structural integrity of the 4 helical-bundle of stalk have complicated influences on the alteration of fusion-triggering activities for different mutants. These data suggested that the head-stalk linker could regulate the fusion triggering at both full-length and headless HN levels.

Role of N-linked glycosylation of the human parainfluenza virus type 3 hemagglutinin-neuraminidase protein

Human parainfluenza virus type 3 (hPIV-3) is a major respiratory tract pathogen that affects infants and young children. The hPIV-3 hemagglutinin-neuraminidase (HN) protein is a multifunctional protein mediating hemadsorption (HAD), neuraminidase (NA), and fusion promotion activities, each of which affects the ability of HN to promote viral fusion and entry. The hPIV-3 HN protein contains four potential sites (N308, N351, N485 and N523) for N-linked glycosylation. Electrophoretic mobility analysis of mutated HN proteins indicated that N308, N351 and N523 sites, but not the N485 site in HN protein, were targeted for the addition of glycans in BHK-21 cells. These functional glycosylation sites were systematically eliminated in various combinations from HN to form a panel of mutants in which the roles of individual carbohydrate chains and groups of carbohydrate chains could be analyzed. Removal of individual or multiple N-glycans on the hPIV-3 HN protein had no effects on transport to the cell surface, expression and NA activity. Single glycosylation site mutants (G1, G2 and G4) not only impaired fusion promotion activity but also reduced HAD activity of HN protein, which was even more obvious for all three double mutants (G12, G14 and G24) and the triple mutant (G124). In addition, every mutant protein retained F-interactive capability that was equal to the wild-type protein capability. Interestingly, the F protein that could be co-immunoprecipitated with the G12 mutated protein or immunoprecipitated with anti-F antibody was not efficiently cleaved. For G14, G24 and G124, little cleaved F protein was detected in co-immuoprecipitation F protein assay and its total amounts where in the cell lysates. The mechanism underlying hPIV-3 HN and F protein remained associated before and after receptor engagement and the strength of the HN-receptor interaction modulated the activation of F the protein which could determine the extent of fusion. Finally, we demonstrated that single or multiple N-glycosylation site mutations inhibited fusion at the earliest stages. Taken together, these results indicated that N-glycosylation of hPIV-3 HN is critical to its receptor recognition activity, cleavage of the F protein, and fusion promotion activity, but had no influence on its interaction with the homologous F protein and NA activity.