This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Porotto, M.
Right arrow Articles by Moscona, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Porotto, M.
Right arrow Articles by Moscona, A.

 Previous Article  |  Next Article 

Journal of Virology, April 2007, p. 3216-3228, Vol. 81, No. 7
0022-538X/07/$08.00+0     doi:10.1128/JVI.02617-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

A Second Receptor Binding Site on Human Parainfluenza Virus Type 3 Hemagglutinin-Neuraminidase Contributes to Activation of the Fusion Mechanism{triangledown}

Matteo Porotto,1 Micaela Fornabaio,2 Glen E. Kellogg,2 and Anne Moscona1*

Departments of Pediatrics and of Microbiology and Immunology, Weill Medical College of Cornell University, 515 East 71st St., Box 309, New York, New York 10021,1 Department of Medicinal Chemistry and Institute for Structural Biology and Drug Discovery, Box 980540, Virginia Commonwealth University, Richmond, Virginia 232982

Received 27 November 2006/ Accepted 10 January 2007


arrow
ABSTRACT
 
The hemagglutinin-neuraminidase (HN) protein of paramyxoviruses carries out three discrete activities that each affect the ability of HN to promote viral fusion and entry: receptor binding, receptor cleaving (neuraminidase), and triggering of the fusion protein. The interrelationship between the receptor binding and fusion-triggering functions of HN has not been clear. For human parainfluenza type 3 (HPIV3), one bifunctional site on HN can carry out both receptor binding and neuraminidase activities, and this site's receptor binding can be inhibited by the small receptor analog zanamivir. We now report experimental evidence, complemented by computational data, for a second receptor binding site near the HPIV3 HN dimer interface. This second binding site can mediate receptor binding even in the presence of zanamivir, and it differs from the second receptor binding site of the paramyxovirus Newcastle disease virus in its function and its relationship to the primary binding site. This second binding site of HPIV3 HN is involved in triggering F. We suggest that the two receptor binding sites on HPIV3 HN each contribute in distinct ways to virus-cell interaction; one is the multifunctional site that contains both binding and neuraminidase activities, and the other contains binding activity and also is involved in fusion promotion.


arrow
INTRODUCTION
 
Paramyxoviruses, including the human parainfluenza viruses, possess an envelope protein hemagglutinin-neuraminidase (HN) that has receptor cleaving as well as receptor binding activity. HN is also essential for activating the fusion protein (F) to mediate the merger of the viral envelope with the host cell membrane. For the parainfluenza viruses as well as other HN-containing paramyxoviruses, this one molecule carries out three different but critical activities at specific points in the process of viral entry. A clear mechanistic understanding of how these activities are regulated is key for the design of strategies to block viral entry (21). The specific role of HN in triggering F is a key question and has eluded simple explanations (3); we propose that a second binding site on HN exists and serves an important function in the triggering mechanism.

Crystallographic studies of the human parainfluenza virus type 3 (HPIV3) HN (17), the Newcastle disease virus (NDV) HN (6, 33), and the simian virus 5 HN (32) indicated that a single active/catalytic site on the globular head of the HN molecule can indeed provide both receptor binding and neuraminidase activities. For NDV, experimental evidence for the importance of this site ("site I") to receptor binding includes the finding that mutations in site I alter receptor binding (2, 18). For HPIV3, mutations at the putative receptor binding site also affect receptor binding avidity (22, 23, 25). However, the effects of the receptor analog zanamivir on activities of the two HN molecules were different: for HPIV3 HN, both receptor binding and neuraminidase activity were blocked by zanamivir, while for NDV HN, receptor binding was partially resistant to the effect of zanamivir but its neuraminidase activity was as sensitive to inhibitor as that of HPIV3 (27). For NDV HN, these data were consistent with the presence of two active sites on NDV HN: "site I," exhibiting both neuraminidase and receptor binding activities, and "site II," possessing only receptor binding activity and insensitive to zanamivir (23). Indeed, two sites were identified in the X-ray crystal structure of NDV HN (33). We investigated the operation of the NDV HN site II and its functional relationship to the bifunctional site I and found that engagement of site I may lead to activation of site II and that the activated site II has higher receptor avidity than site I. By activating the avidly binding site II upon attachment of site I to a target cell, NDV entry could be promoted at the start of the viral life cycle. Neither site has been directly implicated in triggering of the NDV F. However, a mutation at the NDV HN dimer interface that led to decreased receptor binding also impaired fusion promotion (4).

For HPIV3, the findings that zanamivir inhibited both receptor binding and neuraminidase activities (10) and that a mutation in the putative active/catalytic site could cause resistance to both inhibitory effects (23) suggested the presence of a bifunctional site at T193 (23). This residue was subsequently identified crystallographically at the active site (17). Zanamivir has been a useful tool for the study of HN active-site functions. For HN variants with increased avidity for receptor, resistance to zanamivir can be explained by competition for receptor binding. When the threonine at residue 193 in the active site is replaced by isoleucine, resistance ensues because steric hindrance excludes the zanamivir molecule from this site (27). Surprisingly, however, a mutation at a distant site, H552Q, also led to an HN with high receptor avidity and partial resistance to zanamivir (23).

The distant residue affecting receptor avidity (H552) lies at the C terminus of the second strand of the sixth ß-sheet within the HN monomer (27); the loop between this ß-sheet and the next strand forms part of the dimeric interface in the HPIV3 HN structure (17). The HPIV3 HN structure in any of its complexed forms did not suggest involvement of this residue in the formation of the ligand binding site around T193 (17), consistent with the fact that the mutation at H552 did not affect zanamivir binding affinity or neuraminidase activity. We speculated that the increased receptor binding avidity of the H552Q variant (22, 23) could indicate the existence of a putative second receptor binding site. The second receptor binding site in NDV HN (33) is formed in the groove of the dimer interface (33); however, we noted that this interface is different in HPIV3 HN (27) in terms of the orientation of the constituent monomers with respect to each other (17). Thus, we did not anticipate that the second receptor binding site in HPIV3 HN, if indeed one exists, would be directly analogous to the site II of NDV.

HPIV3 HN-receptor interaction is critical for F activation, and HN in its receptor-bound state triggers F to undergo the final conformational changes that lead to fusion readiness (25, 29). The relationship between HN's receptor binding and subsequent triggering of F protein and the nature of the signal that is transmitted from HN to F during this process are unknown and are of the utmost interest. In the present study, we show experimental evidence for a second receptor binding site on HPIV3 HN that plays a role in triggering F. The studies take advantage of three variant viruses: two with mutations at the dimer interface—the proposed location of the second binding site—and one with a mutation in the primary binding site. The primary binding site mutation (T193A) confers enhanced fusion promotion as a result of higher relative avidity for receptor (22, 25). The dimer interface variants are the HN mentioned above with a mutation at H552 and a newly characterized variant with a mutation at the adjacent residue, N551, which confers enhanced fusion properties without affecting receptor avidity. The properties conferred by mutations at N551 suggest direct involvement of this second sialic acid binding site in triggering F. These experimental data are placed in the context of the three-dimensional structural information for HPIV3 HN (17), using computational molecular modeling. We show that HPIV3 HN has two sites that each contribute in distinct ways to receptor binding; one (site I) is the multifunctional site that performs both binding and neuraminidase activities, and the other (site II) possesses binding activity and also is involved in fusion promotion.


arrow
MATERIALS AND METHODS
 
Cells. 293T (human kidney epithelial), Vero, HeLa, and CV1 cells were grown in Dulbecco's modification of Eagle's medium (Mediatech Cellgro) supplemented with 10% fetal bovine serum and antibiotics.

Chemicals. 4-Guanidino-DANA (zanamivir) was a generous gift from Glaxo Smith Kline. A 50 mM stock solution was prepared in serum-free medium, and stock solutions were stored at –20°C. DANA was obtained from Sigma.

Assay of neuraminidase activity. Neuraminidase assays were performed in transiently transfected 293T cell monolayers as previously described (23).

HN and F Constructs. Mutagenized HPIV3 HN cDNAs were digested with SacI or EcoRI and BamHI and ligated into digested pCAGGS and pEGFP mammalian expression vectors as previously described (23).

Transient expression of HN and F genes. Transfections were performed according to the Lipofectamine Plus or the Lipofectamine 2000 manufacturer's protocols (Invitrogen).

Quantification of cell surface expression of HN and F by ELISA. To quantify the amount of HN or F expressed on the surface of 293T cells, an enzyme-linked immunosorbent assay (ELISA) was performed as described previously (23). Transfected 293T cells were washed with phosphate-buffered saline (PBS) after incubation at 37°C, fixed for 10 min with 4% formaldehyde-PBS, and reacted with a mixture of anti-HPIV3 HN monoclonal antibodies or F monoclonal antibodies supplied by Judy Beeler from the WHO repository. The cells were left at room temperature for 30 min and then washed three times with PBS. Horseradish peroxidase-conjugated anti-mouse immunoglobulin G (Bio-Rad) was then added to the cells in PBS-1% bovine serum albumin (1:10,000 dilution), which were then incubated for 30 min at room temperature. The cells were washed three times with PBS before incubation with 3,3',5,5'-tetramethylbenzidine substrate (Pierce) as per the manufacturer's protocol. Absorbance measurements were taken at 450 nm with a Spectramax M5 ELISA reader.

HAD assays. Hemadsorption (HAD) was performed and quantified as previously described (24). Following aspiration of the medium from transfected 293T cell monolayers in 24- or 48-well Biocoat plates (Becton Dickson Labware, Bedford, MA), the medium was replaced with 300 or 150 µl of 1% red blood cells (RBCs) in serum-free, CO2-independent medium (pH 7.5) and placed at 4°C for 30 min. The wells were then washed three times with 300 or 150 µl cold CO2-independent medium (Gibco, Gaithersburg, MD; catalog no. 18045-088). The bound RBCs were lysed with 200 µl RBC lysis solution (0.145 M NH4Cl, 17 mM Tris HCl), and the absorbance was read at 410 nm with a Spectramax M5 ELISA reader.

Partial removal of sialic acid receptors from RBCs. Partial receptor depletion of red blood cells was achieved by treatment of 2 ml of a 10% RBC solution in serum-free medium for 2 h at 37°C with 0 to 200 milliunits of Clostridium perfringens neuraminidase (type X from C. perfringens) (catalog no. N-2876; Sigma Scientific, St. Louis, MO) as described previously (23). Neuraminidase was then removed by washing the RBCs three times with serum-free medium. Each set of RBCs was then resuspended in serum-free, CO2-independent medium to make final RBC stocks of 2% RBCs.

Use of receptor-depleted RBCs to assess HN receptor binding avidity. RBCs partially depleted of their surface sialic acid receptors (described above) were used to determine the relative receptor binding avidities of variant HN molecules as described previously (23). In each experiment, all the RBCs were from the same preparation of depleted stocks (as above). The RBCs were overlaid onto 48-well plates of 293T cell monolayers transiently transfected 24 h prior with wild-type (wt) or variant HN expressed as described above. The plates were placed at 4°C for 30 min to allow RBC binding. The cell monolayers were then washed at 4°C with cold CO2-independent medium to remove unbound RBCs, the bound RBCs were lysed (0.145 M NH4Cl, 17 mM Tris HCl), and the absorbance was read at 410 nm on the Spectramax ELISA reader. Results are presented as percent retention of RBCs relative to the control (undepleted RBCs) versus degree of depletion (expressed as mU of bacterial neuraminidase). For pretreatment of HN-expressing cells with neuraminidase, cells in monolayers were treated with 25 mU per well in 48-well plates (5 x 105 cells) for 3 h at 37°C, transferred to 4°C, and washed.

Assays for F activation, receptor retention, and receptor release. Monolayers of 293T cells transiently expressing F and wt or mutant HNs (or HNs alone as a control) were washed three times and incubated with 1% RBC suspensions at pH 7.5 for 30 min at 4°C. After rinsing to remove unbound RBCs, the plates were placed at the indicated temperature for 1 h, the plates were rocked, and the liquid phase was collected in V-bottom tubes for measurement of released RBCs. The monolayers were then incubated at 4°C with 200 µl of RBC lysis solution; lysis of unfused RBCs with NH4Cl removes the RBCs whose membranes have not fused with HN/F-coexpressing cells. The liquid phase was collected in V-bottom 96-well plates for measurement of reversibly bound RBCs. The cells were then lysed in 200 µl 0.2% Triton X-100-PBS and were transferred to flat-bottom 96-well plates for quantification of the pool of fused RBCs. The amount of RBCs in each of the three compartments described above was determined by measurement of absorption at 410 nm.

Luminescence fusion assay. A previously described luciferase reporter gene assay for cell fusion was used for quantifying cell fusion promoted by HPIV3 envelope proteins (26). Twenty-four-well plates of Vero cells were transfected with human immunodeficiency virus type 1 (HIV-1) long terminal repeat (LTR) luciferase plus HPIV3 F and wt or variant HN. Vero cells transfected with pSV2-TAT (AIDS Reference Reagent Program) were added to the plates of Vero cells and incubated at 37°C for 12 h. Luciferase expression occurs only when cells containing Tat fuse with those transfected with luciferase. The cells were then washed, lysed (using luciferase lysis buffer [Promega]), and luciferase activity resulting from fusion of the two cell types was quantified using luciferase assay substrate (Promega) and a Spectramax M5 (Molecular Devices) luminescence microplate reader.

Beta-galactosidase complementation-based fusion assay. We adapted an assay (19) that detects early stages of fusion activation, since the readout does not depend upon downstream transactivation events. We used this assay for experiments in which the greater range of detection of this assay was necessary. The assay is based on alpha complementation of beta-galactosidase; the beta-galactosidase protein lacking the N-terminal 85 residues (omega peptide) is expressed from one plasmid, and the N-terminal 85 residues (alpha peptide) is expressed from a second plasmid. Cell fusion leads to complementation, and beta-galactosidase is quantitated using the Galacto-Star (Applied Biosystems) chemiluminescent reporter gene assay system. Receptor-bearing cells expressing the omega peptide are mixed with HPIV3 HN/F-coexpressing cells that also express the alpha peptide, at various temperatures and at specific time points. Fusion is stopped by lysing the cells with lysis buffer.

Molecular models. Molecular models for the ligand-bound active sites were built from available crystallographic structures using tools within the Sybyl 7.1 suite (Tripos, Inc., St. Louis, MO). In particular, models of ligands bound at site I were built from the structure of HPIV3 HN with sialic acid (PDB code 1v3c; resolution, 2.30 Å [17]), DANA (PDB code 1v3d; resolution, 2.28 Å [17]), and zanamivir (PDB code 1v3e; resolution, 1.89 Å [17]) cocrystallized at site I. Hydrogen atoms, which are generally not present in the PDB files, were added and energy minimized with the Tripos force field to an energy gradient of 0.005 kcal mol–1Å–1 to reduce steric hindrance. Variants of HN (T193A, N551D, and H552Q) were built from the wild-type protein by using the residue replacement tools in the Sybyl Biopolymer module. The regions around the single mutated residues (6-Å "hot" radius) were relaxed with a molecular force field energy minimization to an energy gradient of 0.005 kcal mol–1Å–1, using standard parameters and procedures. In order to distinguish between the effect of the mutation and the effect of the energy optimization that follows the mutation (in particular for the site I T193A variant compared to the wt), parallel results were obtained for the wt HN subjected to the same relaxation protocol and parameters as used for the variants.

Molecular models of the putative site II were built as a result of docking experiments performed with the genetic algorithm docking program GOLD, version 3.0 (12). The ligand molecules sialic acid, DANA, and zanamivir were "docked" in a broad 10-Å-radius region of HN (wild type and variants N551D and H552Q) around residues 551 and 552. All the parameters of the docking runs were kept as default, while the number of generated poses for each docked ligand was set to 50. All the poses generated for each ligand were manually examined and then, for cases where true docking was observed, analyzed with the HINT (hydropathic interactions) (Tripos, Inc., St. Louis, MO) scoring function. A large number of poses were found to be surface bound and/or mostly solvent exposed and thus not true binding sites. The best docked pose of each small molecule determined by HINT was then energy minimized in its site to allow for site relaxation concomitant with the ligand binding. When the ligands docked were more solvent accessible, the distance-dependent dielectric was changed to 4. In the Discussion, we define "fit" for DANA and zanamivir to mean that the modeled ligand is not surface (nonspecifically) bound, i.e., that more than half of the ligand is in contact with the protein, with a protein-ligand HINT score suggestive of {Delta}G value of ≤0. For sialic acid, which simulates target cell binding to HN, a looser definition of "fit" is used, in which only the energetic requirement is enforced.

Random docking with GOLD was then followed by manual docking experiments. The best sialic acid poses docked at site II of the HN complex structure with sialic acid cocrystallized at site I were used as reference for manual docking. Ligand molecules were superimposed over the sialic acid references by matching chemical moieties. Then, a 6-Å "hot" radius region around the manually placed ligands was allowed to relax to an energy gradient of 0.005 kcal mol–1Å–1.

Hydropathic scoring. The relative energies of the bound ligands in the molecular models were evaluated with the HINT program (13, 15, 31). This program uses a non-Newtonian force field derived from experimental measurements of solvent partitioning to calculate free energy scores that can be converted to binding free energies (1, 5, 8, 9). In this work, HINT 3.10S, which includes a number of local modifications (14, 30), was used to calculate binding scores. The hydrogen treatment option was set to "essential," which explicitly considers all polar hydrogens but incorporates nonpolar hydrogens as part of their heavy atom (i.e., the hydrogens of a methyl group are subsumed into the carbon-united "CH3" atom). In addition, the solvent-accessible surface area for protein backbone nitrogens was corrected with the "+20" option. HINT scores were converted to relative free energies of binding by using the previously determined relationship that –1 kcal mol–1 {approx} 500 HINT score units (8). We obtain {Delta}{Delta}G values (relative binding energies with respect to sialic acid bound at site I) as an indication of binding efficiency, but the absence of crystallographically bound template for binding at site II leads to a fairly high degree of uncertainty. The modeling approach chose the best scoring pose from the many generated by the GOLD docking program and used that to calculate binding energy. This is subject to an uncertainty as to whether we have obtained the overall best pose.

Statistical analysis. The Student-Newman-Keuls multiple-comparison test was applied for statistical analyses, using the InStat 3.0 software (GraphPad Software Inc., San Diego, CA).


arrow
RESULTS
 
Fusion promotion by HNs with mutations at the dimer interface. In order to investigate the possibility of a second receptor binding site on HPIV3 HN, we characterized a newly identified variant virus with an HN mutated at the HN dimer interface region (wt F). This variant, similarly to other specific HN variants discussed here and in previous publications (11, 20, 24), emerged under the selective pressure of growth in neuraminidase-treated cells. The N551D mutation in HN confers high fusogenicity on the variant virus. The following series of experiments are based upon comparison of this new variant (N551D) with two previously characterized fusogenic variant viruses with wt F and single mutations in HN: one with a mutation in the adjacent residue in the dimer interface region(H552Q [22]) and one with a mutation in the primary HN binding site (T193A [22]). We first assessed fusion promotion activity of the expressed mutant HNs, coexpressed at equal levels with wt F on the cell surface. The H552Q HN and T193A HN have previously been shown to be more fusion promoting than wt HN and are included here for comparison with the new variant. Figure 1 demonstrates that the N551D HN is fusion promoting (approximately twofold compared to wt HN) when coexpressed with wt F. The enhanced fusion promotion of each of the three variant HNs (T193A HN, H552Q HN, or N551D HN) compared to wt HN is evident at each time point studied. The fusion properties of these variant HNs and viruses, along with other properties of the viruses and/or HNs that are explored in the subsequent figures, are presented in summary form in Table 1.


Figure 1
View larger version (27K):
[in this window]
[in a new window]

  		FIG. 1. Time course of fusion mediated by wt F coexpressed with wt HN, T193A HN, H552Q HN, or N551D HN. HPIV3 HN/F-coexpressing cells were allowed to fuse with receptor-bearing cells (Vero) at 37°C. Fusion was stopped at the indicated time points by lysing the cells and was quantitated using a beta-galactosidase complementation assay as described in Materials and Methods. Fusion (y axis) in the presence of each coexpressed HN at the indicated time point (x axis) is expressed in relative luminescence units (RLU).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Features of HPIV3 HN variants

Relative neuraminidase activities and receptor binding avidities of wt and mutated HNs. The balance between neuraminidase activity and receptor avidity modulates the fusion-promoting potential of HN. We therefore determined neuraminidase activity and receptor avidity for the mutant HNs in comparison to wt HN. T193 is a residue of the bifunctional (binding/catalytic) active site, and we have shown previously that mutations in this site may increase (in the case of T193I [23]) or decrease (in the case of D216N [11, 29]) sialic acid cleavage. The T193A and H552Q mutations, on the other hand, caused no change in neuraminidase activity (23). We have now found this to be true also for the mutation (N551D) of a residue that, like H552, is at or near the dimer interface of the HN structure (data not shown). The neuraminidase activity of this HN is identical to that of wt HN.

To determine the receptor binding avidity of the HNs, we measured the avidity of cell surface-expressed HN both with and without pretreatment with neuraminidase. The pretreatment was designed to remove the receptors on the expressing cells' surfaces and eliminate potential cell surface interactions between HN and receptor (see Discussion and Fig. 9). In this avidity experiment, aliquots of RBCs with different degrees of receptor depletion (see Materials and Methods) were used to quantify RBC binding on monolayers of cell transiently expressing wt or HPIV3 variant HNs (23, 29). HN molecules with higher avidity can bind RBCs that have lower receptor density. Figure 2 shows that the avidities of wt and N551D HNs are similar, both with and without neuraminidase pretreatment. In the absence of neuraminidase pretreatment, both T193A and H552Q HNs have higher avidity than wt or N551D HN. A large increase in avidity after neuraminidase pretreatment is shown by T193A HN. The neuraminidase pretreatment thus reveals the higher potential avidity of T193A HN compared to the other HNs.


Figure 9
View larger version (48K):
[in this window]
[in a new window]

  		FIG. 9. Schematic illustration of HN receptor avidity measurement: identifying the maximal potential receptor avidity of each HN by eliminating potential same-cell surface interactions. (A) The previously described neuraminidase-deficient HN showed that same-cell interaction between HN and sialic acid-containing molecules can affect HN's availability to bind with target cell receptors. Neuraminidase pretreatment of the expressing cell allows this HN to demonstrate wt receptor avidity. (B) Pretreatment of expressing cells (shown in the right panels, compared to cells without pretreatment shown in the left panels) permits differentiation between HNs of different avidities by using the RBC receptor depletion avidity assay. As the RBCs are progressively depleted (top to bottom), pretreatment allows each HN maximal opportunity to bind to receptor-depleted RBCs without competition from sialic acid moieties on the same cell surface.


Figure 2
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 2. Receptor avidity of wt HN, N551D HN, T193A HN, and H552Q HN, with and without pretreatment with neuraminidase, to remove receptors on surfaces of the HN-expressing cells. A panel of RBCs with different degrees of receptor depletion (prepared as described in Materials and Methods) were used to quantify HAD on cell monolayers expressing the indicated HN, at 4°C, in the absence (black lines) or presence (red lines) of neuraminidase (n'ase) pretreatment of the HN-expressing cells. At 4°C, these HPIV3 neuraminidases are not active. The extent of binding of each depleted RBC preparation (y axis) is expressed as a percentage of that of the control (i.e., of the amount of untreated, nondepleted RBCs bound to cells expressing the corresponding HN).

In the experiment shown in Fig. 2, RBC binding is plotted against the degree of RBC depletion; the less steep the decline of the curve, the higher the HN's avidity. Without neuraminidase pretreatment of the HN-expressing cells, the avidity of N551D does not differ significantly from wt (Fig. 2). T193A HN exhibited the highest avidity, and the avidity of H552Q HN was also higher than that of the wt or N551D HNs. Pretreatment of HN-expressing cells before RBC addition had no appreciable effect on wt HN or N551D HN but raised the avidity of H552Q HN and, most strikingly, raised the avidity of T193A HN (Fig. 2). Key messages from this experiment are that wt and N551D HNs have similar avidity, an avidity lower than that of either the variant H552Q or T193A (the HN with a mutated primary binding site). Thus, the enhanced fusion promotion property of N551D HN cannot be attributed to either altered neuraminidase activity or increased receptor avidity and must be due to a third property conferred by this mutation.

Resistance of the binding properties of HN to zanamivir or DANA: comparison reveals the presence of a second receptor binding site. We have shown that the sensitivity of wt and variant HPIV3 HNs to zanamivir (by the measures of viral infectivity or fusion of HN/F-expressing cells) depends on the extent to which zanamivir inhibits HN receptor binding activity. Our previous findings suggested that HN's avidity alone does not entirely explain zanamivir resistance (23, 27). The experiment shown in Fig. 3 compares the contributions of receptor avidity to inhibitor sensitivity for the site I variant HN (T193A) and the proposed site II variant HNs under conditions of maximal potential avidity, i.e., in neuraminidase-pretreated HN-expressing cells. Zanamivir is the tool used to distinguish between effects of mutations at the two locations. We determined the HNs' sensitivity to inhibition of binding by either zanamivir or the smaller inhibitor Neu5Ac2en (DANA), which lacks the guanidinium group of zanamivir. HAD was carried out on monolayers of cells expressing approximately equivalent levels of the wt and variant HNs in the presence or absence of zanamivir (Fig. 3A) or DANA (Fig. 3B). Figure 3A shows that the H552Q HN is more resistant than wt HN to the effects of zanamivir on receptor binding, while T193A HN is as sensitive as wt HN. This result is notable in light of the markedly lower relative avidity of H552Q HN than T193A HN shown in Fig. 2. N551D HN, with wt avidity, also has wt sensitivity to the effects of zanamivir on receptor binding (see also Discussion). Figure 3B shows that all three variants are equally sensitive to the smaller molecule DANA under these conditions, suggesting that DANA can fit into either binding site I or the putative site II. The zanamivir results shown in Fig. 3A indicate that the newly identified dimer interface mutation, N551D, neither alters avidity nor confers resistance to the effects of zanamivir on receptor binding. However for the HN variant mutated in the adjacent residue, H552Q (intermediate avidity, dimer interface mutation), resistance to zanamivir cannot be attributed to receptor avidity mediated by the receptor binding site at T193 and thus may be mediated by a second site.


Figure 3
View larger version (29K):
[in this window]
[in a new window]

  		FIG. 3. Sensitivity of receptor binding activity of wt HN, N551D HN, T193A HN, and H552Q HN to zanamivir or DANA. Cell monolayers transiently expressing the indicated HNs were assayed by HAD (at 4°C) as described in Materials and Methods. The binding (y axis) at the indicated concentrations of zanamivir or DANA (x axis) is expressed as the percentage of binding obtained in the absence of inhibitor. The results are means of results for triplicate wells from representative experiments, with bars denoting standard deviations.

Fusion promoted by variant HNs: resistance to inhibition by zanamivir is not mediated by the primary binding site. If indeed the H552Q variant HN's resistance to zanamivir's effect on receptor binding is mediated by avidity at a site distinct from the primary binding site, fusion promoted by this variant HN should also be resistant to zanamivir, more so than that of the high-avidity primary binding site variant T193A. If the adjacent N551D mutation has no effect on either property, its fusion promotion should have wt sensitivity to zanamivir. To test the sensitivity of wt and variant HN fusion promotion activity to inhibition by zanamivir, we used a fusion assay based on luciferase expression (26) (see Materials and Methods). Fusion in the presence or absence of zanamivir, added together with the overlay cells, was measured after incubation at 37°C for 12 h. Figure 4 shows the sensitivity to zanamivir inhibition of fusion mediated by F coexpressed with wt HN, N551D HN, T193A HN, and H552Q HN. Despite the higher receptor avidity of T193A HN, only the H552Q HN is resistant to the effects of zanamivir on fusion. For H552Q HN, 50% inhibition was not achieved until a concentration of 2.5 mM was used; however, for wt HN, T193A HN, and N551D HN, fusion mediated by the coexpressed wt F was decreased by zanamivir in a dose-dependent manner, with 100% inhibition at 0.1 mM. This result suggests that the resistance of H552Q HN to the effects of zanamivir on fusion promotion cannot be due to receptor avidity mediated by the receptor binding site at T193, since despite possessing lower avidity than the T193A variant HN, the H552Q HN is the only variant of this set that is resistant to zanamivir.


Figure 4
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 4. Zanamivir inhibition of cell fusion mediated by wt F coexpressed with wt HN, T193A HN, H552Q HN, or N551D HN. The reporter gene assay quantitates fusion between target Vero cells that express HIV Tat (which activates HIV LTR luciferase production) and effector HeLa cells transfected with luciferase and wt F with one of the indicated HNs. The target and effector cells were combined along with different concentrations of zanamivir (x axis). Luminescence was measured after 12 h at 37°C. The results are expressed as percentage of fusion compared to cells that were not treated with zanamivir (y axis). The values are means (± standard deviations) of results from at least three experiments.

HPIV3 HN receptor binding site II is zanamivir resistant and mediates receptor binding when site I is occupied by zanamivir. One interpretation of the data shown in Fig. 4 is that the resistance of H552Q HN to the effect of zanamivir on receptor binding and fusion is due to enhanced binding conferred by the mutation at the putative second receptor binding site. To test this hypothesis, we modified a previously described strategy to separate the functions of the two sites (28). The assay uses the extent to which target cells remain bound over time to cells that express HN alone in the presence of zanamivir as a measure of the combined impact of avidity, neuraminidase activity, and zanamivir resistance of the different HNs (23, 25). We now used this approach to distinguish effects resulting from mutations at the first and putative second sites. Target cells (RBCs) were bound to HN-expressing cells at 4°C (without zanamivir), and the release of these RBCs after transfer to 37°C (mediated by the neuraminidase activity of HN) was measured with and without added zanamivir. Zanamivir added at the time of transfer to 37°C should promote RBC release by interfering with the receptor binding activity of site I, but if site II is zanamivir resistant, then this site should permit binding even in the presence of zanamivir (10, 23).

In the experiment shown in Fig. 5, monolayers of cells expressing wt or mutant HNs were incubated with 1% RBC suspensions for 30 min at 4°C and rinsed to remove unbound RBCs, and then 0.5 mM zanamivir was added to the experimental plates at the time of transfer to 37°C. At the indicated time points, the liquid phase was collected for measurement of released RBCs. The monolayers were then incubated at 4°C with RBC lysis solution (see Materials and Methods for details) for quantification of the pool of bound RBCs. If receptor attachment for the H552Q HN is mediated in part by a second, zanamivir-resistant site, the zanamivir added at the time of transfer to 37°C should not promote RBC release at 37°C as extensively as for the other HNs. If the neuraminidase inhibitory effect on the bifunctional site I has an impact, then this would serve to slow RBC release for all the HNs. If, on the other hand, receptor attachment for H552Q HN is mediated by site I, then zanamivir added at 37°C would compete for this binding and promote receptor dissociation and RBC release (as we have shown previously [10, 23, 28]), and this effect would depend only on the relative avidity of site I.


Figure 5
View larger version (23K):
[in this window]
[in a new window]

  		FIG. 5. Release of target RBCs that were prebound to wt HN-, T193A HN-, H552Q HN-, and N551D HN-expressing cells in the absence (A) or presence (B) of zanamivir. (A) HAD at 4°C was carried out on cells expressing the indicated HN. The cells were then rinsed and transferred to 37°C, and the release of RBCs was determined at each time point indicated. (B) Results of identical experiments in the presence of 0.5 mM zanamivir during the 37°C incubation. The amount of RBCs released at the indicated times (x axis) are expressed as the percentage of RBCs that were bound at the end of the HAD period (i.e., at time point zero).

To compare the features of site I and putative site II for wt HN, T193A HN, H552Q HN, and N551D HN, at several time points after transfer of HN-expressing cells with target RBCs to 37°C, the amount of bound RBCs (retained by HN-receptor interaction) was compared with the amount of released RBCs (released by HN's neuraminidase). The graphs in Fig. 5A show the amount of RBCs in each pool (bound versus released) for the cells expressing the four different HNs at each time after transfer to 37°C, in the absence of zanamivir. Figure 5B shows the same wt and HN variants but in the presence of zanamivir, which was added to enhance the release of all bound RBCs from those HNs sensitive to these effects, throughout the 37°C incubation.

Receptor detachment from wt HN and N551D HN increased during the entire 180-min incubation at 37°C in the absence of zanamivir, as reflected by the released RBC pool in Fig. 5A. In contrast, the slower receptor detachment from the T193A HN in the absence of zanamivir reflects the higher receptor avidity of this variant HN, and the lower-than-wt rate of receptor detachment from the H552Q HN reflects its higher receptor avidity. Zanamivir addition (Fig. 5B) reverses this order. The T193A HN allows for rapid receptor detachment since zanamivir interferes with receptor interaction at site I, and this HN's higher avidity does not overcome this effect. However, the H552Q HN remains attached to receptor even in the presence of zanamivir, indicating that the zanamivir-resistant site II plays a significant role in receptor attachment. Binding is no longer mediated by site I (both wt HN and T193A HN have completely released the attached receptors) yet at the same zanamivir concentration the variant with a higher-avidity site II can effectively bind to sialic acid receptor. This experiment shows that while receptor binding site I is zanamivir sensitive, site II is zanamivir resistant and mediates receptor binding when site I is occupied by zanamivir.

Fusion promotion efficiency is enhanced by mutations at HN site II. We hypothesized that the fusion advantage conferred by the mutations at site II results from a more efficiently fusion-promoting HN. HNs more efficient at F triggering should lower the activation energy required to activate F and should promote fusion at lower temperatures. The experiment shown in Fig. 6 measures fusion, quantitated by the cell fusion assay shown in Fig. 1, after incubation for 6 h of cells coexpressing wt F with wt HN, T193A HN, H552Q HN, or N551D HN. In addition, as a control for measurement of triggering efficiency we include a previously characterized variant, P111S/D216N HN, whose mutation at P111 confers a specific defect in F triggering; it was this HN variant that provided the first proof that a paramyxovirus HN triggers F (29). Figure 6A shows that at each temperature, the HN variants H552Q and N551D have a fusion advantage. Figure 6B expresses these results for each HN variant as the percentage of fusion promoted by wt HN. The H552Q mutation confers fourfold-higher fusogenicity than wt HN at 15°C (P < 0.001) and threefold higher fusogenicity than the wt at 20°C (P < 0.001); at 37°C, the fusion advantage is only 50%. The difference in fusion promotion is revealed most clearly at the lower temperatures by the differences in bar height at 15 and 20°C. Note that at 30°C (and at 37°C) the impact of the higher affinity of the T193A HN is manifested by this HN's effective fusion promotion. The advantage of the putative second-site HN variants (H552Q and N551D HN) in fusion promotion at lower temperatures suggests that the mutations in HN (in the case of N551D, a mutation that does not increase receptor avidity) increase the efficiency of F activation or decrease the activation energy required for triggering of F, and this question was addressed in the next experiment.


Figure 6
View larger version (26K):
[in this window]
[in a new window]

  		FIG. 6. Temperature dependence of fusion promoted by wt HN, T193A HN, H552Q HN, or N551D HN compared to F-triggering-deficient HN P111S/D216N HN. HN/F-coexpressing cells were allowed to fuse with receptor-bearing cells (Vero) for 6 h at the indicated temperatures. Fusion was stopped at the indicated time points and quantitated using a beta-galactosidase complementation assay as described in Materials and Methods. (A) The fusion (y axis) after 6 h in the presence of each coexpressed HN at the indicated temperature (x axis) is expressed in relative luminescence units (RLU). The results are means of results for triplicate wells from representative experiments ± standard deviations. (B) Results for each HN variant at each temperature are expressed as the percentage of fusion promoted by wt HN ± standard deviation.

The second receptor binding site on HN modulates the activation energy required for HN to trigger F: N551D HN and H552Q HN are enhanced in F triggering. HN must be receptor bound in order to trigger F, but the F-triggering efficiency of HN is an independent variable, affecting the fusion promotion capacity of a receptor-bound HN (25). Our assay for comparing the F-activating potentials of wt and variant HNs makes use of the fact that RBCs can fuse with HN/F-coexpressing cells after F has been activated but not before (29). We modified this assay to compare the temperature dependence of the F activation activity of each HN. We reasoned that since HN serves to decrease the temperature dependence of fusion, evidence for F triggering at lower temperatures would suggest a more efficient triggering process.

To compare the F activation functions of wt HN, T193A HN, H552Q HN, and N551D HN at a range of temperatures, we determined the amounts of target RBCs that were (i) released by HN's neuraminidase, (ii) bound by HN-receptor interaction, and (iii) fused, after 60 min of incubation of HN/F-coexpressing cells with target RBCs at a range of temperatures. For the experiment shown in Fig. 7, cells coexpressing wt F and HN were allowed to bind RBCs at 4°C. The cells were then transferred to a range of temperatures: 15, 20, 24, 30, or 37°C. These temperatures were chosen because preliminary studies of the temperature dependence of F activation by wt HN revealed that at 20°C F activation was readily detectable but that at 37°C activation of F is more rapid and effective than at 20°C (25). At 60 min, the amount of released RBCs (released by HN's neuraminidase) was compared with the amounts of bound RBCs (retained by HN-receptor interaction) and of fused RBCs (see Materials and Methods). The graphs in Fig. 7 show the amount of RBCs in each pool (released, bound, and fused) for the cells coexpressing F and HN at each temperature. Fusion by all HNs that are coexpressed with wt F increased with increasing temperature. RBC release remained minimal at all temperatures, and the decrease in bound RBCs is accounted for entirely by the increase in fused RBCs. Comparing the wt and variant HNs coexpressed with F reveals the enhanced F-triggering potential of the two HN variants with mutations at site II: H552Q and N551D. For H552Q HN and N551D HN, at 15°C and 20°C, more activation of F is accomplished than for the wt at these temperatures. It can be seen that H552Q HN is the most effective at these lower temperatures, with 50% of F activated at 15°C, compared to 40% for N551D HN (P < 0.001) but only 20% for both wt HN (P < 0.001) and the high-avidity T193A HN (P < 0.001). At the temperature of 20°C the N551D variant is as effective as H552Q HN; both have activated over 60% of F. At a temperature of 24°C the differences between the variants and wt HN disappear, and wt HN, T193A HN, H552Q HN, and N551D HN all activate around 80% of F. Of particular note, the HN variant with highest avidity at site I, T193A, demonstrates temperature dependence for F activation that is virtually identical to that of the wt HN. The comparison among these four HNs shows that mutations at the putative site II enhance F triggering at lower temperatures and further implicates this proposed site II in activation of F.


Figure 7
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 7. Dependence of F activation by expressed wt and variant HNs on temperature. Quantitation of the amounts of RBCs that are released, bound, or fused via F triggering at different temperatures is shown. For the F activation assay, monolayers of cells coexpressing wt F and wt HN, T193A HN, H552Q HN, or N551D HN with RBCs bound to them at 4°C were transferred to the indicated temperatures. After 60 min of incubation, RBCs that were (i) released by the HNs' neuraminidase (dashed red lines and red symbols), (ii) bound by HN-receptor interaction (dashed black lines and open symbols), and (iii) fused (solid black lines and yellow symbols) were quantified. The temperatures at which the level of fusion is 50% for cells expressing wt, T193A, and H552Q HNs are indicated. The ordinate values are means of results of three experiments ± standard deviations.

F activation mediated by HN site II plays a role in viral entry. Infection with virus may be affected by different variables than fusion between cells. It was therefore important to determine whether the F triggering advantage conferred by mutations at HN site II (Fig. 7) was relevant to infection with virus. Evidence for F triggering in intact viruses at lower temperatures would suggest a more efficient triggering process that is important for infectivity and is biologically meaningful.

To compare the F activation function of wt HN and N551D HN in viruses at a range of temperatures, we determined the amounts of viral particles that proceeded to infect cells after 30 min of incubation on cell monolayers at a range of temperatures from 15 to 37°C. For the experiment shown in Fig. 8, viruses with either wt HN or variant N551D HN were allowed to bind Vero cell monolayers at 4°C for 60 min. The cells were then transferred to a range of temperatures: 15, 18, 23, or 37°C. At 30 min, 10 mM zanamivir was added. The extent to which virus particles can infect after the addition of zanamivir provides a measure of the F-activating potentials of the different HNs, since we have shown that zanamivir interferes with HN-receptor interaction (29), releasing virus that is bound via HN alone (24). After addition of zanamivir, the cells were transferred to 37°C and washed, an overlay of agarose was added, and viral plaques were counted 24 h later. At each temperature below 37°C, more of the N551D variant viral particles produced plaques. This result indicates that viral particles bearing the N551D HN are more efficient at initiating entry into target cells, doing so at a lower temperature, than particles bearing wt HN. We interpret these data to mean that the site II mutation allows for more efficient F triggering and F insertion by virus.


Figure 8
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 8. F activation on virus particles: effect of temperature on viral entry promoted by wt HN and N551D HN. Viruses with wt HN or N551D HN were incubated with cell monolayers at 4°C for 60 min and then transferred to the indicated temperatures (x axis). At 30 min 10 mM zanamivir was added to release virus particles bound via HN alone; the cells were transferred to 37°C, washed, and overlaid with agarose; and viral plaques were counted 24 h later (y axis).


arrow
DISCUSSION
 
For paramyxoviruses, the role of the receptor binding protein HN in triggering the fusion protein is a major unanswered question. For NDV we provided experimental evidence for a second receptor binding site on HN (27), which was corroborated by the revised NDV HN crystal structure (33). We showed that engagement of the primary receptor binding site on NDV HN can lead to activation of the second binding site and that this second site has higher avidity for receptor, thus promoting NDV fusion and entry into the target cell (28). For HPIV3, however, a second receptor binding site has not been identified in the extant crystal structures (17, 32). Despite the evidence that mutations at the dimer interface, distant from the primary HPIV3 HN binding site, modulate receptor avidity (22, 27) and the implication of a second functional receptor binding site on this molecule (27), it has been suggested that HPIV3 HN lacks a second binding site and therefore that such a site is unlikely to be involved in triggering F (3). In this paper we have experimentally confirmed the existence of a second receptor binding site on HPIV3 HN, and here we provide a number of possible computational modeling explanations for the failure to identify this site in the HN crystal structure. Unlike the second site on NDV HN, this HPIV3 site has lower avidity for receptor, and we show that this site is involved in triggering F.

Key tools that allowed for identification of the presence of a second receptor binding site are several variant HN molecules with single-residue alterations in HN. In order to separate the F-triggering properties of these variant HN molecules from their alterations in receptor avidity, it was key to measure receptor avidity in a way that reflected the binding potential of each HN molecule without interference from other molecules on the same cell surface. We hypothesize that the high receptor avidity of T193A HN results in its interaction with sialic acid-containing receptors on the expressing cell surface, rendering this HN unavailable to bind with target cell receptors. The nature of this "cis" interaction may also be affected by the neuraminidase activity of this HN. This hypothesis is supported by the previous demonstration that a cell surface-expressed neuraminidase-deficient HN fails to bind target cells at all unless pretreated with neuraminidase, but has wt-level avidity for target cell receptors, once freed from adjacent sialic acid residues (24). The model for this neuraminidase-deficient variant is diagrammed in Fig. 9A. The impact of this phenomenon on the avidity measurements for HN variants is diagrammed in Fig. 9B. Progressive depletion of sialic acid receptors on the target RBC surface reveals the higher binding avidity of specific HNs (right panels); however, without neuraminidase pretreatment (left panels), high avidity, as assessed by binding to highly receptor-depleted RBCs, is masked by interaction with same-surface receptors (bottom panels). Thus, the use of neuraminidase pretreatment in our avidity assays allowed for comparison of the maximal potential receptor avidity of each HN without the need to compensate for interactions with same-surface molecules. This advance was important in permitting experiments that separate the binding versus F-triggering functions of first- and second-site HN variants.

To distinguish the two functions of the second receptor binding site, the variant HN with a mutation at residue N551 was important. The study of this variant virus containing the N551D mutation in HN was a key to discovering that the F-triggering activity of site II is functionally distinct from this site's impact on receptor avidity. The N551D mutation in HN does not alter receptor avidity (Fig. 2) but enhances fusion (Fig. 1) and F activation (Fig. 7 and 8). The variant virus with the H552Q mutation, altered in a location distant from the known active site but possessing increased receptor avidity, had provided the first indication that receptor binding activity could reside at a second site (27). The finding in the present study that zanamivir resistance was conferred by the H552Q mutation but not by the higher-avidity T193A mutation in the primary site further confirmed the presence of a second site that could mediate binding. However, the proof that this site indeed activates F required analysis of the F-triggering enhancement seen in the case of N551D HN.

To seek structural corroboration for the experimental data suggesting a second binding site, the interactions of sialic acid, DANA and zanamivir with the putative site II were analyzed using computational modeling based on the crystallographic structures of HPIV3 HN (17). In order to complete the picture of wt and variant interactions with both HN sites, we also examined ligand interactions with site I for both wt HN and T193A HN. Our computational analysis on site I of HPIV3 HN suggests that sialic acid binds somewhat more strongly (–0.3 kcal mol–1) to the T193A variant HN than to the wt HN, as predicted by the receptor avidity data shown in Fig. 2. In contrast, both DANA and zanamivir, which bind to the wt protein more strongly than sialic acid, with DANA binding slightly more tightly than zanamivir, may have more affinity (ca. 1 kcal mol–1) for the wild-type protein than for the variant.

To evaluate the putative site II, docking experiments (12) were performed on the HPIV3 HN protein in the vicinity of residues N551 and H552 as well as on molecular models of the mutations at these sites present in the variant HNs (N551D and H552Q). A broad multifaceted binding site was located in the N551/H552 region (site II) (Fig. 10a, inset 2), where sialic acid and DANA appear to bind with multiple, possibly noncompetitive, orientations. Site II is defined by the residues 551 and 552; the portion of site II that is predominantly defined by residue 551 is more solvent accessible and is in close proximity to site I (Fig. 10b). The other portion of site II is more involved in the HN dimer interface crevice and is predominantly defined by residue 552 (Fig. 10c). The close proximity of sites I and II in HPIV3 HN is different than what is observed in NDV HN (33), where the two sites are separated by more than 10 Å.


Figure 10
View larger version (123K):
[in this window]
[in a new window]

  		FIG. 10. Computational model of two sites of HN complexed with ligand. (a) Crystallographic structure of HPIV3 HN complexed with DANA (purple sticks) bound at site I (described in Materials and Methods). Connolly surfaces for HN monomers A and B, created with MOLCAD, are shown with cavity depth property coding (monomer A) and local curvature property coding (monomer B). Residues T193, N551, and H552 from both monomers are rendered in ball-and-stick form, colored by atom type, behind red translucent Connolly surfaces. Inset 1 shows the ligand DANA and T193 at site I of monomer A. Inset 2 shows the N551 and H552 residues of monomer A (left) and monomer B (right). Inset 1' shows DANA and T193 at site I of monomer B. (b) Close-up view of site II with DANA (red ball-and-stick model) docked near N551. A portion of the deeper site I cavity is seen in the upper left; T193 is visible at the edge of this pocket. The N551 residue is at the lower right edge of the site II pocket. (c) Close-up view of site II with DANA (red ball-and-stick model) docked at the dimer interface, i.e., on the H552 side of the site. As shown in panel a, this site is occluded; thus, the foreground surface contours have been removed for clarity. The two monomers, A and B, are shown in orange and green, respectively.

The three available crystal structures for HPIV3 HN have sialic acid, DANA, and zanamivir bound at site I, so docking was performed with all three structure models and with N551D and H552Q mutant models of each. There is no crystallographically bound ligand at the region of residues N551 and H552 of HPIV3 HN, nor are there unambiguously definable binding pockets in this region in the (low-temperature) crystal structure; the site II observed in our modeling results from minor side chain rearrangements that would be reasonable under the biological conditions, i.e., elevated temperature, of receptor binding as opposed to those of crystallographic structure determination. The in silico mutation H552Q led to this rearrangement of side chains and revealed the potential pocket at this site. These perturbations, simulated by the computational structure relaxation steps that we applied in the docking experiments, then allowed for visualization of the second-site binding pocket in wt HN as well. Since the H552Q HN exists in a viable virus, we propose that these side chain rearrangements are consistent with functional HN structure.

Somewhat different morphologies are seen at site II depending on the ligand bound at site I. When either sialic acid or DANA is bound to site I of HPIV3 HN, either molecule can also fit into site II of wt HN or the HN variants N551D and H552Q (data not shown). However, zanamivir cannot fit into site II of HN for the wt or the H552Q variant HN when zanamivir is bound at site I. There are two likely contributing factors for this. First, the presence of the somewhat bulkier zanamivir in site I may cause rearrangement of the residues in the vicinity of site II. Second, the larger zanamivir cannot fit in site II, as the guanidinium group is usually solvent exposed. With zanamivir bound at site I, sialic acid appears to bind weakly at site II for wt HN but more favorably at site II for H552Q. The {Delta}{Delta}G values (relative binding energies with respect to sialic acid bound at site I) for the interaction of sialic acid at site II when zanamivir is at site I are >5.0 kcal mol–1 for wt HN and for N551D but 4.3 kcal mol–1 for H552Q. The N551D mutant, in contrast, shows a different behavior: zanamivir can bind in site II of this mutant HN. While sialic acid binds weakly at site II for this HN variant, zanamivir may be computationally docked in site II such that its positively charged guanidinium group interacts with the negatively charged D551 carboxylate group of the N551D HN variant (the {Delta}{Delta}G value for zanamivir binding at site II is 2.9 kcal mol–1 for the N551D HN, while the ligand is nonspecifically bound, i.e., solvent exposed, for the wt and H552Q HNs). These modeling results are consistent with the experimental findings regarding the effect of zanamivir on the avidity of HN for cellular receptors. The wt HN and the T193A and H552Q HN mutants are resistant to zanamivir's inhibitory effect on receptor binding because of the presence of site II, where zanamivir cannot bind but sialic acid can bind (weakly except for the H552Q mutant). However, the N551D variant is apparently not resistant to zanamivir because zanamivir can bind at site II, partially due to a change in the site II morphology but also because the carboxylic acid moiety of the aspartate can interact strongly with the zanamivir guanidinium. The net result is that zanamivir can block site II in the N551D mutant from sialic acid binding, as shown in Fig. 3 and diagrammed schematically in Fig. 11.


Figure 11
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 11. Model for the roles of sites I and II in receptor binding.

Figure 11 shows our proposed model of how the two HN sites function in receptor binding, taking into account the experimental results. We suggest (Fig. 11A) that both HN sites may bind to receptors on target cells. The small receptor analog DANA can fit into both sites and completely abolish binding (Fig. 11B). Figure 11 C, D, and E show the wt and the two second-site variant HNs in the presence of the bulkier zanamivir. Zanamivir fits efficiently into site I (Fig. 11C), and interestingly, our computational analysis suggests that for wt HN, while zanamivir cannot bind site II, zanamivir's occupation of site I reduces receptor binding by site II. We speculate that zanamivir interaction with site I results in an alteration at site II that inhibits receptor binding. For the variant HN H552Q (Fig. 11D), site II is different; zanamivir binding with site I does not prevent receptor binding at site II, as we have shown experimentally (Fig. 3) and corroborated computationally. Figure 11D also schematically represents the higher site II receptor avidity after the H552Q mutation, by a deeper pocket. Figure 11E depicts that the mutation of glutamine at residue 551 to aspartic acid allows for interaction between zanamivir and the aspartic acid side chain; zanamivir can now dock at site II and inhibit receptor binding at site II as well as at site I.

It is of interest to speculate why the specific HN mutations at site II discussed here emerged under selective pressure. The variant viruses arose under the selective pressure of growth in neuraminidase-treated cells. Viruses successful at growing under conditions of receptor scarcity have been characterized and include several types of highly fusogenic variants with single-residue alterations in HN (11, 22). HN variants that have higher avidity for sialic acid receptors emerged from this selection process, and others arose that have lower neuraminidase activity and thus maximize the available sialic acid receptors. With the analysis of the N551D HN variant, a fusogenic variant with neither affinity nor neuraminidase alterations, it is clear that enhanced triggering efficiency provided an advantage to the virus bearing this HN. Under conditions of decreased receptor availability, the virus containing the second-site mutation conferring enhanced F triggering would more successfully infect cells via fusion with the target cell membrane. When the selection process was carried out using HPIV3 viral neuraminidase rather than the less-specific bacterial neuraminidase, instead of a panel of HN variants, only one HN variant emerged: H552Q HN (20). We can now speculate that under these particularly stringent conditions where receptor molecules specific for the viral HN were preferentially depleted, only the variant with two advantages—higher receptor avidity and enhanced F triggering—was successful.

Mechanisms for regulation of receptor binding sites I and II on HPIV3 HN, and for the role of site II in F triggering, can now be proposed. Each of the two HPIV3 sites contributes differently to receptor avidity. While for NDV we have shown that once activated, site II has higher binding avidity than site I, in the case of HPIV3 site II has a smaller contribution to receptor binding than site I. We propose here that the HPIV3 site II is involved in promotion of fusion. Thus, the enhanced fusion promotion properties conferred by site II are not simply due to the increased time in contact with receptor allowing for more F activation (as for site I) but rather are due to a direct decrease in the energy of activation for F triggering. The diagrams in Fig. 12 depict the contributions of the individual properties of HN sites I and II to fusion promotion. Compared to the level of wt HN fusion promotion, T193A HN has a higher level of fusion promotion resulting from higher avidity at site I. H552Q HN, with wt-like site I but enhancements in both avidity and F-triggering activity at site II, has the highest fusion promotion. The N551D HN has only the advantage of enhanced F triggering at site II and thus has enhanced fusion promotion activity compared to that of wt HN but less than that of H552Q HN.


Figure 12
View larger version (23K):
[in this window]
[in a new window]

  		FIG. 12. Model for the contributions of the individual properties of sites I and II to HN's fusion promotion activity. The sizes of the arrows represent the relative magnitude of each property, i.e., binding, neuraminidase, and F triggering, for each of the two sites in the wt, T193A, H552Q, and N551D HNs. The upward arrows (binding avidity and F triggering) represent favorable contributions to fusion.

The difference in the roles of the two sites for HPIV3 and NDV can be explained in the light of the balance between binding activity and neuraminidase activity. For NDV, high neuraminidase activity necessitates a correspondingly high avidity to allow for entry, and site II of NDV HN, once activated, provides this function. For HPIV3, however, the lower neuraminidase activity and the (relative to NDV) high avidity of site I meant that the binding activity of the second site was not revealed until a site II variant emerged under selective pressure of exogenous neuraminidase. Only under pressure to enter cells in the presence of limiting receptor availability, mutations that enhanced site II binding and F activation emerged: H552Q and N551D. Given the strength of binding site I, the more important role of binding site II is its contribution to F activation upon receptor binding. Molecular modeling experiments based upon the three-dimensional structure of HPIV3 HN show that both residues form part of a potential second receptor binding site at or near the HN dimer interface. The H552Q HN fusion promotion properties could initially seem to be based upon the high receptor avidity, as for other high-avidity HN variants; however, the N551D HN variant clearly delineates the function of site II. N551D HN does not have demonstrably higher receptor avidity than the wt in our assays yet has enhanced F-triggering capacity, and it is this variant that most clearly reveals the fusion-triggering function of site II.

We can only speculate how receptor binding by sites I and II is translated into the F activation activity of site II. The modeling evidence suggests that a portion of site II is close to site I in the 3D structure and therefore offers the possibility that receptor interaction at site I can modulate site II; however, we have no experimental evidence to support this possibility. It has been proposed by several authors (7, 16, 22) that the interaction of HN with receptor alters the interaction with F and thus activates F; however, little is yet known about how this might take place. We now suggest that site II is involved in this interaction with or activation of F, and we anticipate that structural analysis of HN molecules with the mutations discussed here, in the presence of ligand and of F, will elucidate this mechanism.


arrow
ACKNOWLEDGMENTS
 
This work was supported by Public Health Service grants AI 31971 to A.M. and GM 71894 to G.E.K. from the National Institutes of Health and by March of Dimes research grant 06-322 to A.M.

We thank J. Beeler for providing needed antibodies; Olga Greengard, Richard Peluso, and Paolo Carta for helpful discussions and critical reviews of the manuscript; S. Rusconi for the beta-galactosidase complementation reagents; and Lynn Marks, Margaret Tisdale, and Judith Ng-Cashin at GSK for kindly providing zanamivir as a laboratory reagent.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Departments of Pediatrics and of Microbiology and Immunology, Weill Medical College of Cornell University, 515 East 71st St., Box 309, New York, NY 10021. Phone: (212) 746-4801. Fax: (212) 746-8261. E-mail: anm2047{at}med.cornell.edu. Back

{triangledown} Published ahead of print on 17 January 2007. Back


arrow
REFERENCES
 
    1
  1. Burnett, J. C., P. Botti, D. Abraham, and G. Kellogg.2001 . Computationally accessible method for estimating free energy changes resulting from site-specific mutations of biomolecules: systematic model building and structural/hydropathic analysis of deoxy and oxy hemoglobins. Proteins Struct. Funct. Genet. 42:355-377.[CrossRef][Medline]
    2. 2
  2. Connaris, H., T. Takimoto, R. Russell, S. Crennell, I. Moustafa, A. Portner, and G. Taylor. 2002. Probing the sialic acid binding site of the hemagglutinin-neuraminidase of Newcastle disease virus: identification of key amino acids involved in cell binding, catalysis, and fusion. J. Virol. 76:1816-1824.[Abstract/Free Full Text]
    3. 3
  3. Connolly, S. A., G. P. Leser, H. S. Yin, T. S. Jardetzky, and R. A. Lamb.2006 . Refolding of a paramyxovirus F protein from prefusion to postfusion conformations observed by liposome binding and electron microscopy. Proc. Natl. Acad. Sci. USA 103:17903-17908.[Abstract/Free Full Text]
    4. 4
  4. Corey, E. A., A. M. Mirza, E. Levandowsky, and R. M. Iorio. 2003. Fusion deficiency induced by mutations at the dimer interface in the Newcastle disease virus hemagglutinin-neuraminidase is due to a temperature-dependent defect in receptor binding. J. Virol. 77:6913-6922.[Abstract/Free Full Text]
    5. 5
  5. Cozzini, P., M. Fornabaio, A. Marabotti, D. J. Abraham, G. E. Kellogg, and A. Mozzarelli. 2002. Simple, intuitive calculations of free energy of binding for protein-ligand complexes. 1. Models without explicit constrained water. J. Med. Chem. 45:2469-2483.[CrossRef][Medline]
    6. 6
  6. Crennell, S., T. Takimoto, A. Portner, and G. Taylor. 2000. Crystal structure of the multifunctional paramyxovirus hemagglutinin-neuraminidase. Nat. Struct. Biol. 7:1068-1074.[CrossRef][Medline]
    7. 7
  7. Deng, R., Z. Wang, P. J. Mahon, M. Marinello, A. Mirza, and R. M. Iorio. 1999. Mutations in the Newcastle disease virus hemagglutinin-neuraminidase protein that interfere with its ability to interact with the homologous F protein in the promotion of fusion. Virology 253:43-54.[CrossRef][Medline]
    8. 8
  8. Fornabaio, M., P. Cozzini, A. Mozzarelli, D. J. Abraham, and G. E. Kellogg. 2003. Simple, intuitive calculations of free energy of binding for protein-ligand complexes. 2. Computational titration and pH effects in molecular models of neuraminidase-inhibitor complexes. J. Med. Chem. 46:4487-4500.[CrossRef][Medline]
    9. 9
  9. Fornabaio, M., F. Spyrakis, A. Mozzarelli, P. Cozzini, D. J. Abraham, and G. E. Kellogg. 2004. Simple, intuitive calculations of free energy of binding for protein-ligand complexes. 3. The free energy contribution of structural water molecules in HIV-1 protease complexes. J. Med. Chem. 47:4507-4516.[CrossRef][Medline]
    10. 10
  10. Greengard, O., N. Poltoratskaia, E. Leikina, J. Zimmerberg, and A. Moscona.2000 . The anti-influenza virus agent 4-GU-DANA (Zanamivir) inhibits cell fusion mediated by human parainfluenza virus and influenza virus HA. J. Virol. 74:11108-11114.[Abstract/Free Full Text]
    11. 11
  11. Huberman, K., R. Peluso, and A. Moscona. 1995. The hemagglutinin-neuraminidase of human parainfluenza virus type 3: role of the neuraminidase in the viral life cycle.Virology. 214:294-300.[CrossRef][Medline]
    12. 12
  12. Jones, G., P. Willett, R. Glen, A. Leach, and R. Taylor.1997 . Development and validation of a genetic algorithm for flexible docking. J. Mol. Biol. 267:727-748.[CrossRef][Medline]
    13. 13
  13. Kellogg, G., and D. Abraham. 2000. Hydrophobicity. Is logPo/w more than the sum of its parts?. Eur. J. Med. Chem. 35:651-661.[CrossRef][Medline]
    14. 14
  14. Kellogg, G., and D. Chen. 2004. The importance of being exhaustive. Optimization of bridging structural water molecules and water networks in models of biological systems. Chem. Biodiv. 1:98-105.
    15. 15
  15. Kellogg, G. E., M. Fornabaio, F. Spyrakis, A. Lodola, P. Cozzini, A. Mozzarelli, and D. J. Abraham.2004 . Getting it right: modeling of pH, solvent and "nearly" everything else in virtual screening of biological targets. J. Mol. Graph Model. 22:479-486.[CrossRef][Medline]
    16. 16
  16. Lamb, R. 1993. Paramyxovirus fusion: a hypothesis for changes. Virology 197:1-11.[CrossRef][Medline]
    17. 17
  17. Lawrence, M. C., N. A. Borg, V. A. Streltsov, P. A. Pilling, V. C. Epa, J. N. Varghese, J. L. McKimm-Breschkin, and P. M. Colman.2004 . Structure of the haemagglutinin-neuraminidase from human parainfluenza virus type III. J. Mol. Biol. 335:1343-1357.[CrossRef][Medline]
    18. 18
  18. Li, J., E. Quinlan, A. Mirza, and R. M. Iorio.2004 . Mutated form of the Newcastle disease virus hemagglutinin-neuraminidase interacts with the homologous fusion protein despite deficiencies in both receptor recognition and fusion promotion. J. Virol. 78:5299-5310.[Abstract/Free Full Text]
    19. 19
  19. Moosmann, P., and S. Rusconi. 1996. Alpha complementation of LacZ in mammalian cells. Nucleic Acids Res. 24:1171-1172.[Free Full Text]
    20. 20
  20. Moscona, A., and R. W. Peluso. 1996. Analysis of human parainfluenza virus 3 receptor binding variants: evidence for the use of a specific sialic acid-containing receptor. Microb. Pathog. 20:179-184.[CrossRef][Medline]
    21. 21
  21. Moscona, A. 2005. Entry of parainfluenza virus into cells as a target for interrupting childhood respiratory disease.J. Clin. Investig. 115:1688-1698.[CrossRef][Medline]
    22. 22
  22. Moscona, A., and R. W. Peluso. 1993. Relative affinity of the human parainfluenza virus 3 hemagglutinin-neuraminidase for sialic acid correlates with virus-induced fusion activity.J. Virol. 67:6463-6468.[Abstract/Free Full Text]
    23. 23
  23. Murrell, M., M. Porotto, T. Weber, O. Greengard, and A. Moscona.2003 . Mutations in human parainfluenza virus type 3 HN causing increased receptor binding activity and resistance to the transition state sialic acid analog 4-GU-DANA (zanamivir).J. Virol. 77:309-317.
    24. 24
  24. Porotto, M., O. Greengard, N. Poltoratskaia, M.-A. Horga, and A. Moscona.2001 . Human parainfluenza virus type 3 HN-receptor interaction: the effect of 4-GU-DANA on a neuraminidase-deficient variant. J. Virol. 76:7481-7488.
    25. 25
  25. Porotto, M., M. Murrell, O. Greengard, L. Doctor, and A. Moscona.2005 . Influence of the human parainfluenza virus 3 attachment protein's neuraminidase activity on its capacity to activate the fusion protein. J. Virol. 79:2383-2392.[Abstract/Free Full Text]
    26. 26
  26. Porotto, M., L. Doctor, P. Carta, M. Fornabaio, O. Greengard, G. Kellogg, and A. Moscona. 2006. Inhibition of Hendra virus membrane fusion. J. Virol. 80:9837-9849.[Abstract/Free Full Text]
    27. 27
  27. Porotto, M., M. Murrell, O. Greengard, M. Lawrence, J. McKimm-Breschkin, and A. Moscona. 2004. Inhibition of parainfluenza virus type 3 and Newcastle disease virus hemagglutinin-neuraminidase receptor binding: effect of receptor avidity and steric hindrance at the inhibitor binding sites.J. Virol. 78:13911-13919.[Abstract/Free Full Text]
    28. 28
  28. Porotto, M., M. Fornabaio, O. Greengard, M. T. Murrell, G. E. Kellogg, and A. Moscona. 2006. Paramyxovirus receptor-binding molecules: engagement of one site on the hemagglutinin-neuraminidase protein modulates activity at the second site. J. Virol. 80:1204-1213.[Abstract/Free Full Text]
    29. 29
  29. Porotto, M., M. Murrell, O. Greengard, and A. Moscona. 2003. Triggering of human parainfluenza virus 3 fusion protein (F) by the hemagglutinin-neuraminidase (HN): an HN mutation diminishing the rate of F activation and fusion. J. Virol. 77:3647-3654.[Abstract/Free Full Text]
    30. 30
  30. Spyrakis, F., M. Fornabaio, P. Cozzini, A. Mozzarelli, D. Abraham, and G. Kellogg. 2004. Computational titration analysis of a multiprotic HIV-1 protease-ligand complex. J. Am. Chem. Soc. 126:11764-11765.[CrossRef][Medline]
    31. 31
  31. Wireko, F. C., G. E. Kellogg, and D. J. Abraham. 1991. Allosteric modifiers of hemoglobin. 2. Crystallographically determined binding sites and hydrophobic binding/interaction analysis of novel hemoglobin oxygen effectors.J. Med. Chem. 34:758-767.[CrossRef][Medline]
    32. 32
  32. Yuan, P., T. B. Thompson, B. A. Wurzburg, R. G. Paterson, R. A. Lamb, and T. S. Jardetzky.2005 . Structural studies of the parainfluenza virus 5 hemagglutinin-neuraminidase tetramer in complex with its receptor, sialyllactose. Structure 13:803-815.[Medline]
    33. 33
  33. Zaitsev, V., M. von Itzstein, D. Groves, M. Kiefel, T. Takimoto, A. Portner, and G. Taylor. 2004. Second sialic acid binding site in Newcastle disease virus hemagglutinin-neuraminidase: implications for fusion. J. Virol. 78:3733-3741.[Abstract/Free Full Text]

Journal of Virology, April 2007, p. 3216-3228, Vol. 81, No. 7
0022-538X/07/$08.00+0     doi:10.1128/JVI.02617-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Porotto, M., Yokoyama, C. C., Orefice, G., Kim, H.-S., Aljofan, M., Mungall, B. A., Moscona, A. (2009). Kinetic Dependence of Paramyxovirus Entry Inhibition. J. Virol. 83: 6947-6951 [Abstract] [Full Text]  
  • Palermo, L. M., Porotto, M., Yokoyama, C. C., Palmer, S. G., Mungall, B. A., Greengard, O., Niewiesk, S., Moscona, A. (2009). Human Parainfluenza Virus Infection of the Airway Epithelium: Viral Hemagglutinin-Neuraminidase Regulates Fusion Protein Activation and Modulates Infectivity. J. Virol. 83: 6900-6908 [Abstract] [Full Text]  
  • Krishnan, A., Verma, S. K., Mani, P., Gupta, R., Kundu, S., Sarkar, D. P. (2009). A Histidine Switch in Hemagglutinin-Neuraminidase Triggers Paramyxovirus-Cell Membrane Fusion. J. Virol. 83: 1727-1741 [Abstract] [Full Text]  
  • Aguilar, H. C., Ataman, Z. A., Aspericueta, V., Fang, A. Q., Stroud, M., Negrete, O. A., Kammerer, R. A., Lee, B. (2009). A Novel Receptor-induced Activation Site in the Nipah Virus Attachment Glycoprotein (G) Involved in Triggering the Fusion Glycoprotein (F). J. Biol. Chem. 284: 1628-1635 [Abstract] [Full Text]  
  • Mahon, P. J., Mirza, A. M., Musich, T. A., Iorio, R. M. (2008). Engineered Intermonomeric Disulfide Bonds in the Globular Domain of Newcastle Disease Virus Hemagglutinin-Neuraminidase Protein: Implications for the Mechanism of Fusion Promotion. J. Virol. 82: 10386-10396 [Abstract] [Full Text]  
  • Alymova, I. V., Taylor, G., Mishin, V. P., Watanabe, M., Murti, K. G., Boyd, K., Chand, P., Babu, Y. S., Portner, A. (2008). Loss of the N-Linked Glycan at Residue 173 of Human Parainfluenza Virus Type 1 Hemagglutinin-Neuraminidase Exposes a Second Receptor-Binding Site. J. Virol. 82: 8400-8410 [Abstract] [Full Text]  
  • Nichols, W. G., Peck Campbell, A. J., Boeckh, M. (2008). Respiratory Viruses Other than Influenza Virus: Impact and Therapeutic Advances. Clin. Microbiol. Rev. 21: 274-290 [Abstract] [Full Text]  
  • Porotto, M., Carta, P., Deng, Y., Kellogg, G. E., Whitt, M., Lu, M., Mungall, B. A., Moscona, A. (2007). Molecular Determinants of Antiviral Potency of Paramyxovirus Entry Inhibitors. J. Virol. 81: 10567-10574 [Abstract] [Full Text]  
  • Palermo, L. M., Porotto, M., Greengard, O., Moscona, A. (2007). Fusion Promotion by a Paramyxovirus Hemagglutinin-Neuraminidase Protein: pH Modulation of Receptor Avidity of Binding Sites I and II. J. Virol. 81: 9152-9161 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Porotto, M.
Right arrow Articles by Moscona, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Porotto, M.
Right arrow Articles by Moscona, A.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»