ABSTRACT
Influenza virus neuraminidase (NA) cleaves off sialic acid from cellular receptors of hemagglutinin (HA) to enable progeny escape from infected cells. However, NA variants (D151G) of recent human H3N2 viruses have also been reported to bind receptors on red blood cells, but the nature of these receptors and the effect of the mutation on NA activity were not established. Here, we compare the functional and structural properties of a human H3N2 NA from A/Tanzania/205/2010 and its D151G mutant, which supports HA-independent receptor binding. While the wild-type NA efficiently cleaves sialic acid from both α2-6- and α2-3-linked glycans, the mutant exhibits much reduced enzymatic activity toward both types of sialosides. Conversely, while wild-type NA shows no detectable binding to sialosides, the D151G NA exhibits avid binding with broad specificity toward α2-3 sialosides. D151G NA binds the 3′ sialyllactosamine (3′-SLN) and 6′-SLN sialosides with equilibrium dissociation constant (KD ) values of 30.0 μM and 645 μM, respectively, which correspond to much higher affinities than the corresponding affinities (low mM) of HA to these glycans. Crystal structures of wild-type and mutant NAs reveal the structural basis for glycan binding in the active site by exclusively impairing the glycosidic bond hydrolysis step. The general significance of D151 among influenza virus NAs was further explored by introducing the D151G mutation into three N1 NAs and one N2 NA, which all exhibited reduced enzymatic activity and preferential binding to α2-3 sialosides. Since the enzymatic and binding activities of NAs are not routinely assessed, the potential for NA receptor binding to contribute to influenza virus biology may be underappreciated.
INTRODUCTION
Influenza A viruses are the major cause of annual flu epidemics and occasional pandemics. Their genome consists of eight RNA segments containing genes that encode 11 viral proteins, including two viral surface glycoproteins, hemagglutinin (HA) and neuraminidase (NA). The well-documented influenza virus infection cycle starts with the attachment of virus through its HA to sialic acid-containing glycan receptors on the host cell surface. The virus is then internalized by endocytosis, and the viral genome is released to the cytoplasm through HA-mediated membrane fusion and virus uncoating. After entry, the viral RNA is transported into the cell nucleus, where viral genes direct the production of new viral components to form progeny virions. The newly assembled virions bud from the apical cell membrane with HA and NA on the surface. The NA cleaves off sialic acid from glycans on the host cell and the emerging virions, thus allowing release of progeny viruses from the host cell and preventing virus aggregation. Due to their opposite functions, the balance of HA receptor-binding and NA receptor-destroying activity is critical for influenza virus replication and transmission (18, 32).
HA mediates agglutination of red blood cells (RBCs) by influenza virus via its binding to surface α2-3-linked “avian-type” and/or α2-6-linked “human-type” sialoside receptors (hemagglutination activity) (26, 31). Recently, however, several NAs from human H3N2 viruses isolated in MDCK cells were reported to have acquired the capacity to bind receptors on RBCs (15). It was found that postinfection ferret antisera to many A/Wisconsin/67/2005-like H3N2 viruses from 2005 to 2009 that were propagated on MDCK cell cultures showed poor inhibition of hemagglutination by antibodies to HA in assays using RBCs from various species, but this antibody-resistant hemagglutination was sensitive to the influenza virus neuraminidase inhibitor oseltamivir carboxylate (the active metabolite of Tamiflu from Roche) (15). It was further determined that this activity did not arise from any mutations in the HA, but from a single mutation in aspartic acid 151 of NA to glycine, asparagine, or alanine. Furthermore, relatively low levels of Asp-to-Gly, -Asn, or -Ala substitutions at NA position 151 have also been reported in clinical specimens of seasonal H3N2 viruses (∼13%) (15). Although it was proposed that these receptors were refractory to NA hydrolytic activity, as well as inhibition by an anti-HA antibody, the nature of the receptors and the precise binding site on the NA were not identified (15).
One of these viruses had a neuraminidase that was postulated to enable binding to RBCs in a hemagglutination assay and was a variant of the influenza H3N2 virus A/Tanzania/205/2010 (TZ205), a human isolate of the 2010 influenza season from the Centers for Disease Control and Prevention. The TZ205 variant with a D151G mutant NA emerged during isolation of the virus in MDCK cells. To provide insights into the nature of the receptors and the biological significance of the novel hemagglutination activity, we expressed wild-type NA from TZ205 (TZ205 D151 NA) and its D151G mutant (TZ205 G151 NA) in a baculovirus system. Although it was originally reported that the NA catalytic activity of the D151G mutant was normal (15), activity assays of the recombinant NA with the commonly used fluorescent substrate 2′-(4-methylumbelliferyl)-α-d-N-acetylneuraminic acid (4-MU-NANA) or a glycan array-based neuraminidase assay clearly showed drastically reduced enzymatic activity. Surprisingly, the D151G mutant demonstrated strong and unequivocal binding to most α2-3 sialosides and to some α2-6 sialosides present on the glycan microarray. In contrast, the wild-type NA possessed no apparent binding to any of the tested α2-3 and α2-6 sialosides. Isothermal titration calorimetry (ITC) analysis confirmed that the D151G mutant NA binds both 3′ sialyllactosamine (3′-SLN) and 6′-SLN. Crystal structures of the wild-type NA in complex with sialic acid, oseltamivir carboxylate, or HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], and the D151G mutant in complex with 3′-SLN, 6′-SLN, oseltamivir carboxylate, or HEPES were determined at resolutions ranging from 2.2 to 2.6 Å, and the precise binding modes of the mutant NA with avian and human receptors were fully elucidated. The D151G mutant appears to mediate the weak hydrolytic and enhanced hemadsorption (sialic acid binding) activities within its active site, thus explaining the postulated receptor-binding properties of the H3N2 G151 NA that were previously reported in hemagglutination inhibition experiments with antibodies to the HA (15).
The importance of D151 for influenza virus NA activity was confirmed by introducing the D151G mutation into three N1 NAs and one N2 NA. The three N1 viruses were A/New York/06/2009 (H1N1) (NY06), A/California/04/2009 (H1N1) (Cali09), and A/Swine/England/WVL7/1992 (H1N1) (sw/England92), and the N2 virus was A/Hong Kong/68 (H3N2) (HK68). All four D151G NA mutants had significant decreases in both kcat and Km values against 4-MU-NANA from wild-type NAs, showed much reduced enzymatic activity toward human and avian sialoside receptors, and exhibited HA-like binding to avian-type receptors.
MATERIALS AND METHODS
Cloning, expression, and purification of the neuraminidases.The ectodomain (positions 82 to 469) and ectodomain plus stalk region (positions 37 to 469) of NA from the influenza virus A/Tanzania/205/2010 (H3N2) bearing the D151 NA (TZ205 D151 NA, GISAID [Global Initiative on Sharing Avian Influenza Data] database accession number EPI342198) and its D151G NA mutant (TZ205 G151 NA, GISAID accession number EPI279969) were expressed in a baculovirus system for structural and functional analyses. The cDNAs corresponding to the NA ectodomain and ectodomain plus stalk region of TZ205 D151 and TZ205 G151 were inserted into a baculovirus transfer vector, pFastbacHT-A (Invitrogen) with an N-terminal gp67 signal peptide, a thrombin cleavage site, a His6 tag, and an N-terminal tetramerization domain, essentially as previously described (35, 37). The constructed plasmids were used to transform DH10bac competent bacterial cells by site-specific transposition (Tn7 mediated) to form a recombinant bacmid with β-galactosidase blue-white receptor selection. The purified recombinant bacmids were used to transfect Sf9 insect cells for overexpression. NA proteins were produced by infecting suspension cultures of Hi5 cells with recombinant baculovirus at a multiplicity of infection (MOI) of 5 to 10 and incubated at 28°C shaking at 110 rpm. After 72 h, Hi5 cells were removed by centrifugation, and supernatants containing secreted, soluble NAs were concentrated and buffer exchanged into 20 mM Tris (pH 8.0), 300 mM NaCl, and 2.5 mM CaCl2, further purified by metal affinity chromatography using Ni-nitrilotriacetic acid (NTA) resin (Qiagen). For crystal structure determination, the ectodomain NAs were digested with thrombin to remove the tetramerization domain and His6 tag. The cleaved ectodomain NAs were purified further by size exclusion chromatography on a Hiload 16/90 Superdex 200 column (GE Healthcare) in 20 mM Tris (pH 8.0), 100 mM NaCl, 10 mM CaCl2, and 0.02% NaN3. For the NA solution-based activity assay, NA glycan array-based receptor-binding and cleavage assay, as well as the ITC assay, the uncleaved NA ectodomain plus stalk region of TZ205 D151 NA and its D151G mutant with tetramerization domain and His6 tag attached were concentrated after Ni-NTA purification in 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, and 0.02% NaN3.
The ectodomain plus stalk region (positions 37 to 469 [N2 numbering]) of four D151G NAs from A/Hong Kong/68 (H3N2) (HK68), A/California/04/2009 (H1N1) (Cali09), A/New York/06/2009 (H1N1) (NY06), and A/Swine/England/WVL7/1992 (H1N1) (sw/England92) were also expressed in our baculovirus system for glycan array binding and cleavage analyses, as well as solution-based NA substrate cleavage analyses.
Neuraminidase activity assay with substrate 4-MU-NANA.We measured NA enzymatic activities in 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, 0.02% NaN3 buffer by using fluorescent substrate 2′-(4-methylumbelliferyl)-α-d-N-acetylneuraminic acid (4-MU-NANA) (21) with excitation and emission wavelengths of 365 nm and 450 nm, respectively. The reaction was conducted for 10 min at 37°C in a total volume of 80 μl for NA proteins with ectodomain plus stalk region. The reactions were all performed in triplicate and were stopped by adding 80 μl of 1 M Na2CO3. To compare NA cleavage activities at 12 different NA concentrations, with a fixed substrate 4-MU-NANA concentration of 0.025 mM, the NA starting solutions were serially diluted 1:2. From the sigmoidal curve produced for each NA, the concentration that corresponds to the midpoint of the linear section of the curve was selected and used for a kinetics assay in which the enzyme reaction velocity was measured as a function of substrate concentration. NA concentrations of 0.0205 μg/ml and 2.56 μg/ml were used in kinetics assays for TZ205 D151 NA and TZ205 G151 NA, respectively. The Km, Vmax, and kcat were calculated by fitting the data to the appropriate Michaelis-Menten equations by using nonlinear regression in the GraphPad Prism software (GraphPad Software, La Jolla, CA). Dissociation constants for NA complexes with inhibitors oseltamivir carboxylate and zanamivir (Relenza) were determined by measuring the reduction in the rate of 4-MU-NANA hydrolysis observed in the presence of different concentrations of inhibitor. A fixed substrate 4-MU-NANA concentration of 0.025 mM was applied, and NA concentrations of 0.0205 μg/ml and 1.28 μg/ml were used in inhibition assays for TZ205 D151 NA and TZ205 G151 NA, respectively. The Ki values were calculated by using the “Ki from IC50” function in the GraphPad Prism software.
The kinetics assays of four D151G NA mutants from HK68, NY06, Cali09, and sw/England92 were carried out by using the same method as described above. NA concentrations of 0.125 μg/ml, 0.125 μg/ml, 0.25 μg/ml, and 0.125 μg/ml were used in kinetics assays for HK68 G151 NA, NY06 G151 NA, Cali09 G151 NA, and sw/England92 G151 NA, respectively.
Glycan array neuraminidase cleavage activity and specificity assay.NA proteins with ectodomain plus stalk region were diluted to 40 μg/ml in buffer containing 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, and 0.02% NaN3 in glycan arrays separated by silicon superstructures to form 100-μl incubation wells (see Fig. S7 in the supplemental material for a list of the glycans) (26). NA incubation was carried out for 1 h, and then each array was washed by removal of the NA solution by pipetting and exchanging buffer in the wells 3 times with 1× phosphate-buffered saline (PBS) containing 0.05% Tween (pH 7.4) at room temperature (∼22°C). Following the third wash, 100 μl of 10 U/ml protease from Streptococcus griseus (Sigma) plus 10% SDS in 1× PBS was added to each well, and the arrays were allowed to incubate at 37°C for 2 h to degrade any protein bound to the array surface. After 2 h, each array was washed as described above, and 100 μl of a precomplexed solution of biotinylated Erythrina cristagalli lectin (ECA) (10 μg/ml; VectorLabs) plus streptavidin-Alexa Fluor 555 (2 μg/ml; Invitrogen) was applied directly to the array surface and incubated for 2 h. Following incubation, ECA-streptavidin solution was removed by pipetting, washed 3 times with 1× PBS plus 0.05% Tween by exchanging buffer in the wells and, subsequently, by dipping the slides 3 times in 1× PBS and then 3 times in distilled H2O. Washed slides were dried by centrifugation and scanned on a ProScanArray Express HT (PerkinElmer) confocal slide scanner at the Alexa Fluor 555 setting. Image data were stored as a TIFF image, and signal data were collected using Imagene (BioDiscovery) imaging software. Collected signal data were processed to determine the averaged (mean signal minus mean background) values of 4 replicate spots on the array for each unique printed glycan.
NA glycan array receptor-binding assay.NA proteins with ectodomain plus stalk region at 16 μg/ml were precomplexed at a 4:2:1 molar ratio with mouse anti-His antibody conjugated to Alexa Fluor 488 (anti-His-Alexa Fluor 488) (Qiagen) and goat anti-mouse IgG-Alexa Fluor 488 (Invitrogen), respectively, in buffer containing 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, and 0.02% NaN3. Protein-antibody complexes were allowed to form for 15 min on ice prior to application on the array (see Fig. S7 in the supplemental material for a list of the glycans). One hundred microliters of protein-antibody complex in solution was added directly to the surface of the array and allowed to incubate in a humidified chamber, protected from the light, for 1 h at room temperature (∼22°C). Following the initial incubation, protein-antibody solution was removed by pipetting the solution and washing 3 times with 100 μl of 1× PBS supplemented with 0.05% Tween (pH 7.4) and subsequently by dipping the slides 3 times in 1× PBS and then 3 times in distilled H2O. Washed slides were dried by centrifugation and scanned on a ProScanArray Express HT (PerkinElmer) confocal slide scanner at the Alexa Fluor 488 setting. Image data were stored as a TIFF image, and signal data were collected using Imagene (BioDiscovery) imaging software. The signal data were processed to determine averaged (mean signal minus mean background) values of 4 replicate spots on the array for each unique printed glycan.
Neuraminidase binding assay with isothermal titration calorimetry (ITC).Calorimetric assays were carried out with an ITC200 system instrument (MicroCal Inc., part of GE Healthcare, Piscataway, NJ). Reaction cells (200 μl) were filled with protein solutions and equilibrated at the indicated temperatures. The stirring speed was 1,000 rpm, and the reference power was 5 μcal/s. The volume of injection and the concentrations of protein and ligand were optimized to obtain precise KD and enthalpy change (ΔH) values. For TZ205 G151 NA binding with 3′-SLN, a typical ITC experiment was carried out by adding 2.5-μl aliquots of 3.2 mM 3′-SLN into a solution containing 0.107 μM TZ205 G151 NA with ectodomain plus stalk region in 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, and 0.02% NaN3 at 25°C. Fit of the binding isotherm to a 0.75:1 stoichiometry one-site binding model was performed by Origin 7 (OriginLab Corporation, Northampton, MA). For TZ205 G151 NA binding with 6′-SLN, a typical ITC experiment was carried out by adding 3.3-μl aliquots of 6.18 mM 6′-SLN into a solution containing 0.206 μM TZ205 G151 NA with ectodomain plus stalk region in 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, and 0.02% NaN3 at 30°C. Fit of the binding isotherm was carried out using a 1:1 stoichiometry one-site binding model. For TZ205 D151 NA, a typical ITC experiment consisted of adding 3.3-μl aliquots of 0.95 mM 3′-SLN into a solution containing 0.032 μM TZ205 D151 NA in 100 mM imidazole-malate (pH 6.15), 150 mM NaCl, 10 mM CaCl2, and 0.02% NaN3 at 30°C.
Crystal structure determination.Crystallization experiments were set up using the sitting drop vapor diffusion method. Diffraction quality crystals for TZ205 D151 NA ectodomain in complex with bound HEPES were obtained by mixing 0.6 μl of the concentrated protein at 9.2 mg/ml in 20 mM Tris (pH 8.0), 100 mM NaCl, 10 mM CaCl2, and 0.02% NaN3 with 0.6 μl of the well solution in 0.1 M HEPES (pH 7.5) and 17 to 25% polyethylene glycol 8000 at 22°C. The TZ205 G151 NA ectodomain at 7.2 mg/ml was crystallized in the similar conditions as wild-type NA. For cocrystallization of TZ205 D151 NA ectodomain and its mutant NA ectodomain with glycans, the protein sample was prepared in 20 mM Tris (pH 8.0), 100 mM NaCl, 10 mM CaCl2, and 0.02% NaN3 to which 3′-SLN and 6′-SLN were added in 50-fold molar excess at 4°C at least 4 h before the crystallization experiment was set up at 22°C. For the cocrystal of TZ205 D151 NA and its mutant NA with oseltamivir carboxylate, 15-fold molar excess of oseltamivir carboxylate was added to the protein sample at 4°C at least 4 h before the crystallization experiment was set up at 22°C. The crystallization conditions are similar to those for the HEPES complexes. All the crystals were cryoprotected in mother liquor with addition of 25% glycerol before being flash cooled at 100 K. Diffraction data were collected at beamlines 11-1 and 12-2 at the Stanford Synchrotron Radiation Lightsource (SSRL) or beamline 23-ID-B at the Advanced Photon Source (APS) (Table 1). Data for all crystals were integrated and scaled with HKL2000 (20). Data collection statistics are outlined in Table 1.
Data collection and refinement statistics of TZ205 D151 NA and TZ205 G151 NA crystals
Both TZ205 D151 and D151G NA mutant structures were determined by molecular replacement using the program Phaser (17). The TZ205 D151 NA complex structure with bound HEPES was first determined using the A/Memphis/31/98 (H3N2) N2 NA structure (PDB accession number 2AEP) as a model. All other NA structures were subsequently determined using this refined NA structure as an input model. The models were initially refined by rigid body refinement in Refmac5 (19), using all data from 50- to 3-Å resolution. Model building was carried out with the program Coot (5). The last cycles of Refmac-restrained refinement included data for the highest-resolution shell using isotropic B values. The final statistics for both structures are represented in Table 1. The quality of the structures was using the JCSG validation suite (www.jcsg.org). All figures were generated with Bobscript (6) except for Fig. S4, which was generated with PyMol (www.pymol.org).
RESULTS
Effects of the D151G mutation on NA cleavage activity with the substrate 4-MU-NANA.TZ205 D151 NA and its D151G mutant were overexpressed in a baculovirus expression system as previously described (for details, see Materials and Methods) (35). With the commonly used, linkage-indiscriminate substrate 4-MU-NANA (for catalytic reaction, see Fig. S1 in the supplemental material), the NA activities of the wild-type TZ205 NA were similar to other human N2 NAs (34), but its D151G mutant had drastically reduced activity (Fig. 1). With substrate 4-MU-NANA at 37°C, wild-type TZ205 NA and its D151G mutant had kcat values of 26.8 ± 0.5 s−1 and 0.167 ± 0.002 s−1, respectively, and Km values of 29.0 ± 1.9 μM and 0.824 ± 0.058 μM, respectively (Table 2; see Fig. S2A and S2B in the supplemental material) that represented large decreases in kcat (∼160-fold) and Km (∼35-fold) upon D151G substitution. The D151G substitution of TZ205 did slightly reduce binding affinities to NA inhibitors, oseltamivir carboxylate, and zanamivir, with Ki values increased from 14.3 nM to 17.0 nM for oseltamivir carboxylate and from 0.16 nM to 0.34 nM for zanamivir (Table 2; Fig. S2C to S2F). The weak enzymatic activity of the recombinant D151G NA contrasts with the studies of Lin et al. (15), who studied the NA activity of a related H3N2 virus (A/Hong Kong/4443/2005) but found no marked changes in the Km or Vmax of the D151N mutant with 4-MU-NANA and fetuin as the substrates. However, as noted by these authors, the D151G mutant spontaneously reverts to the wild type during amplification of the virus in eggs or MDCK cells; thus, these studies may reflect the presence of trace amounts of wild-type NA in the virus preparations.
The NA cleavage activity was quantified by a solution assay using a fluorescent substrate 4-MU-NANA concentration of 0.025 mM and a reaction temperature of 37°C. TZ205 D151 NA is more active than TZ205 G151 NA (see Fig. S1 for catalytic reaction). Kinetics measurements are presented in Table 2 and in Fig. S2 in the supplemental material.
Kinetic parameters for TZ205 D151 NA and TZ205 G151 NA against 4-MU-NANA at 37°C and inhibition by oseltamivir carboxylate and zanamivir
Effects of the D151G mutation on NA substrate specificity using a glycan array cleavage assay.To assess the possibility for substrate specificity changes upon the D151G mutation, both TZ205 D151 NA and its D151G mutant were screened for activity against diverse sialosides on a glycan microarray cleavage assay. Briefly, a sample of NA is incubated with the glycan array for 1 h before washing with 1× PBS buffer. Sialosides that are cleaved on the array have a terminal galactose, which can be detected by incubation for 2 h with a precomplexed solution of biotinylated Erythrina cristagalli lectin (ECA), which is specific for the terminal Galβ1-4GlcNAc sequence formed after NA cleavage of the sialic acid for the majority of the glycans on the array (see Fig. S7 in the supplemental material). The specificity of the ECA lectin excludes 19 glycans from analysis, including those that would yield a product with terminal Galβ1-3GlcNAc (glycans 13 to 15), Galβ1-3GalNAc (glycans 16 to 18 and 41), Galβ1-4Glc (glycans 27 and 28), GalNAcβ1-4Gal (glycans 26 and 27), or Galβ1-4GlcNAc when the Gal is substituted with sulfate (glycans 5 and 6), or GlcNAc is substituted with fucose (glycans 4, 6, and 29 to 32).
The wild-type NA showed greater activity, with preference for α2-3 sialosides over α2-6 sialosides (Fig. 2A), compared to the mutant NA with drastically reduced activities for α2-3 and α2-6 sialosides without obvious preference (Fig. 2B). The wild-type NA cleaves α2-3 linear sialosides (glycans 9 to 12, 21 and 22, and 33) and α2-6 linear sialosides (glycans 36 to 38, 40, 45, and 51), as shown in Fig. 2A. It can also cleave two α2-3 sulfated sialosides (glycans 3 and 7), a α2-6 sulfated sialoside (glycan 34), two branched α2-3 sialosides (glycans 19 and 20), and four branched α2-6 sialosides (glycans 39, 42 and 43, and 48), as well as α2-3 biantennary sialosides (glycans 23 to 25) or α2-6 biantennary sialosides (glycans 46 and 47 and glycans 49 and 50); the wild-type NA also cleaves α2-3/6 biantennary sialosides (glycans 52 and 53). However, the D151G mutant displayed much reduced activities against α2-3 sialosides and α2-6 sialosides that wild-type NA hydrolyzes (glycans 3, 7, 9 to 11, 20, 22 to 25, 34, 38 and 39, 49, and 52 and 53) (Fig. 2B). Except for the much weaker activities, the mutant NA shows substrate specificities toward α2-3 sialosides that are similar to those of the wild-type NA. The major differences in cleavage activities of the wild-type and mutant NAs indicate that D151 is indeed essential for optimal catalysis, consistent with previous structural and mechanistic studies that D151 plays a critical role in stabilizing the transition state intermediate in the catalytic reaction (3, 8, 11).
NA cleavage activity and specificity for glycan structures on a glycan microarray. (A and B) TZ205 D151 NA (A) and TZ205 G151 NA (B) were measured against printed glycans on a microarray with the same starting NA concentration of 40 μg/ml. Cleavage signals are shown by filled bars with the color of the bar indicating the sialic acid linkages: control glycans that terminated in galactose are shown in black (glycans 1 and 2), α2-3 linkage glycans are shown in blue (glycans 3 to 33), α2-6 linkage glycans in red (glycans 34 to 51), and α2-3/6 mixed glycans in cyan (glycans 52 and 53). A list of printed glycans and their structures is available in Fig. S7 in the supplemental material. TZ205 D151 NA and mutant TZ205 G151 NA cleave both α2-3- and α2-6-linked sialylated glycans, but TZ205 G151 shows much reduced activities. In both NA experiments, glycans 1 and 2 with terminal galactose were used as positive controls. The cleaved glycans on the array have a terminal galactose, which can be detected by ECA. The specificity of the ECA lectin excludes certain glycans from analysis, including those that would yield a product with terminal Galβ1-3GlcNAc (glycans 13 and 15), Galβ1-3GalNAc (glycans 16 to 18 and 41), Galβ1-4Glc (glycans 27 and 28), GalNAcβ1-4Gal (glycans 26 and 27), or the Galβ1-4GlcNAc when the Gal is substituted with sulfate (glycans 5 and 6) or the GlcNAc is substituted with fucose (glycans 4, 6, and 29 to 32). (C and D) NA binding affinity and specificity were also investigated against the same printed glycans (Fig. S7) for TZ205 D151 NA (C) and TZ205 G151 NA (D) with the same starting concentration of 16 μg/ml. Binding signals are shown in filled bars by sialic acid linkages using the same color scheme as in panels A and B. TZ205 D151 NA does not exhibit residual binding to array glycans, but TZ205 G151 NA shows binding to α2-3- and α2-6-linked glycans with preference for α2-3-linked glycans.
Effects on binding capacity and specificity of the D151G mutation from glycan array binding assay and isothermal titration calorimetry.To provide a structural basis for the hemagglutination role of the D151G mutant, the binding affinity and specificity of both purified TZ205 D151 NA and its D151G mutant at 15 μg/ml were studied by glycan microarray binding analysis, as applied for influenza virus HA proteins (1, 25). Analyses of glycan array binding data revealed dramatic differences in apparent receptor-binding capacities between the two NAs. The conventional role of the influenza virus NA is to destroy the receptors used by HA by removing the terminal sialic acid moieties, while the HA is responsible for the attachment of virus to sialic acid-containing glycoconjugates on susceptible cells. As expected, the wild-type NA exhibited no detectable binding to any glycans on the array due to rapid cleavage and release (Fig. 2C; see Fig. S7 in the supplemental material for a list of the glycans). However, consistent with the observed HA antibody-resistant activity of the mutant NA in hemagglutination assays with whole virus (15), the recombinant mutant NA, which has substantially reduced enzymatic activity, showed strong binding to most α2-3 sialosides and moderate binding to a few α2-6 sialosides (Fig. 2D). The mutant NA recognizes α2-3 sulfated sialosides (glycans 3 to 7), α2-3 linear sialosides (glycans 8 to 11, 13 to 17, and 21 and 22), α2-3 branched sialosides (glycans 18 to 20), and α2-3 biantennary sialosides (glycans 23 and 24), as well as α2-3 fucosylated sialosides (glycans 29 to 32). In addition to α2-3 sialosides, the mutant NA displayed moderate binding to one α2-6 sulfated sialoside (glycan 34), one α2-6 linear sialoside (glycan 37), and one α2-6 internal sialoside (glycan 41). It is also interesting that the mutant NA binds strongly to one α2-3/6 biantennary N-linked glycan (glycan 53) but very weakly with the other (glycan 52), indicating branch selectivity of sialoside presentation on N-glycans.
In order to directly assess the binding affinity of the D151G mutant with sialosides, 3′-SLN (glycan 9) and 6′-SLN (glycan 36) were studied by ITC analysis. The mutant NA exhibited equilibrium dissociation constants (KDs) of 30.0 ± 5.5 μM and 645 ± 55 μM, respectively (Fig. 3). For comparison, the wild-type NA was studied by ITC analysis with 3′-SLN, and as expected, no binding was detected (see Fig. S3 in the supplemental material).
ITC analysis of TZ205 G151 NA binding with 3′-SLN (A) and 6′-SLN (B). (Top) Raw data for a typical ITC binding experiment; (bottom) integrated data after correcting for the heat of dilution. TZ205 G151 NA shows preferential binding to 3′-SLN, rather than 6′-SLN, with equilibrium dissociation constants (KDs) of 30.0 ± 5.5 μM and 645 ± 55 μM for 3′-SLN and 6′-SLN, respectively, which are all significantly better than HA-glycan binding affinities with KD values in the millimolar range.
Structural basis of sialoside binding and cleavage specificity of wild-type and mutant neuraminidases.TZ205 D151 NA and the TZ205 D151G mutant NA were crystallized in 0.1 M HEPES (pH 7.5) and 17 to 25% polyethylene glycol 8000, with oseltamivir carboxylate, 3′-SLN, or 6′-SLN added as the substrates/ligands. Three crystal structures were obtained for TZ205 D151 NA in complex with HEPES (Fig. 4B), oseltamivir carboxylate (Fig. 4D), or sialic acid (Fig. 4F). For the mutant TZ205 G151 NA, four crystal structures were obtained in complex with HEPES (Fig. 4C), oseltamivir carboxylate (Fig. 4E), 3′-SLN (Fig. 4G), or 6′-SLN (Fig. 4H). The seven crystal structures were determined at 2.20- to 2.60-Å resolution by molecular replacement (see Materials and Methods), and crystallographic statistics are shown in Table 1. The overall structures of the monomers in the N2 NA tetramer are very similar to each other (Fig. 4A) and are also similar to previously published N2 NA structures with root mean square deviations (RMSDs) of only 0.4 Å between the wild-type NA and N2 from A/Tokyo/3/67 (H2N2) (PDB accession number 1NN2) or N2 from A/Memphis/31/98 (H3N2) (PDB accession number 2AEP). The RMSD values between the five wild-type NA and mutant NA structures are all within 0.2 to 0.3 Å for all the Cα atoms, indicating no large conformational changes upon the D151G mutation.
Crystal structures of TZ205 D151 and TZ205 G151 NAs. The NA is shown in ribbon representation, and the ligands are shown by yellow carbons. (A) Tetramer of TZ205 D151 NA in complex with sialic acid in the active site. One NA monomer is shown with colored ribbons. N-linked glycosylation sites are indicated as sticks with brown carbon atoms. (B) Active site of TZ205 D151 NA in complex with HEPES. (C) Active site of TZ205 G151 NA in complex with HEPES. (D) Active site of TZ205 D151 NA in complex with oseltamivir carboxylate. (E) Active site of TZ205 G151 NA in complex with oseltamivir carboxylate. (F) Active site of TZ205 D151 NA in complex with sialic acid. (G) Active site of TZ205 G151 NA in complex with 3′-SLN. (H) Active site of TZ205 G151 NA in complex with 6′-SLN (only Sia-1 and Gal-2 are ordered). For ease of comparison, panels B, C, D, and E are shown in the same orientation, whereas panels F, G, and H are shown in another orientation for clarity. The corresponding ligands for each structure are shown in different colors as follows: carbon atoms are shown in yellow, nitrogen atoms in blue, oxygen atoms in red, and sulfur atoms in green. The hydrogen bonds between ligands and NA proteins are indicated by blue dotted lines. The electron densities for all ligands are shown in Fig. S4 in the supplemental material. See text for details.
(i) Wild-type TZ205 D151 NA and its D151G mutant in complex with HEPES.The wild-type NA and its D151G mutant were crystallized from a precipitant solution containing 0.1 M HEPES. Clear, interpretable electron density was found for HEPES in the active site (see Fig. S4A and S4B in the supplemental material), albeit with partial occupancy as demonstrated by higher B values of the HEPES ligands compared to surrounding NA residues (Table 1). The D151G amino acid substitution results in only very localized changes in the NA around the mutation site compared to wild-type NA (Fig. 4B and C and Fig. S5A). Both crystal structures are superimposable, including most of the side chains of the active site residues, such as the first-shell Arg118, Arg371, Arg292, Tyr406, Glu277, Arg224, Arg152, and Glu119 (Fig. 4B and C and Fig. S5A). The sulfate moiety of HEPES forms electrostatic interactions with three arginines at positions 118, 292, and 371 and appears to play a key role in the binding of HEPES in the NA active site; other parts of the HEPES molecule form only van der Waals interactions with the NA (Fig. 4B and C). In wild-type NA, the Asp151 carboxylate forms no interactions with NA residues but does form van der Waals contacts with the bound HEPES (Fig. 4B). Upon D151G mutation, no obvious conformational changes are found in the residues in the loop at position 150 even for the main chain of G151 (Fig. S5A). In the mutant NA structure, due to the loss of the Asp151 side chain, the HEPES molecule shifts about 0.6 Å toward the loop at position 150 to form van der Waals interactions with the Cα of G151.
(ii) Wild-type TZ205 D151 NA and its D151G mutant in complex with oseltamivir carboxylate.The crystal structures of both TZ205 D151 NA and its D151G mutant in complex with oseltamivir carboxylate are also very similar except for minor local conformational change due to the D151G mutation (Fig. 4D and E). When complexed with TZ205 D151 NA and its mutant, oseltamivir carboxylate is positioned almost identically in the active site pockets (see Fig. S5B, S4C, and S4D in the supplemental material), and also similarly to other available NA-oseltamivir carboxylate complex structures from N1 and N8 neuraminidases, although the interactions between oseltamivir carboxylate and NAs are slightly different due to small NA structural variations in the different subtypes (23). Most of the interactions, including hydrogen bond and van der Waals interactions are similar in both structures. However, upon D151G mutation, moderate conformational changes occur in the main chain of residue 151 with the Cα of G151 pushed 0.8 Å toward the active site pocket. In the wild type, the side chain of D151 forms a hydrogen bond and van der Waals interaction with oseltamivir carboxylate, while in the mutant, G151 lacks a side chain and is too distant from oseltamivir carboxylate to form any interactions. Together with other subtle changes in the interactions between the drug and NA proteins, the lack of interaction between the side chain on residue 151 and the drug might explain why the D151G mutation has reduced binding affinity for oseltamivir carboxylate (Table 2), which was also observed in other N2 NA D151G mutants (15).
(iii) Wild-type TZ205 D151 NA in complex with sialic acid.In the crystal structure of the wild-type NA in complex with avian receptor 3′-SLN, as expected, only the product sialic acid was retained in the active site after NA enzymatic cleavage (Fig. 4F; see Fig. S4E in the supplemental material). The sialic acid interacts with conserved active site residues similar to another N2 NA sialic acid complex structure from influenza virus A/Tokyo/3/1967 (H2N2) (PDB accession number 2BAT), which was obtained by soaking sialic acid into the apo-form N2 crystals (30). The carboxylate group, which forms hydrogen bonds with three guanidinium groups of arginine residues 118, 292, and 371, is held equatorial to the distorted sugar structure (Fig. 4F and Fig. S5C). The pyranose ring was refined in a boat configuration instead of a chair configuration as in the other N2 NA sialic acid complex (30). The oxygen of the N-acetyl group is hydrogen bonded to Arg152, the hydroxyl groups of the glycerol side chain hydrogen bond to Glu276, and the 4-hydroxyl is in close proximity to Glu119 and Asp151. The glycosidic oxygen attached to C-2 is axial and directed away from the binding site pocket and is hydrogen bonded to the side chain of Asp151, consistent with the proposed role of Asp151 as a key residue in stabilizing the transition state of the sialic acid cleavage reaction (8).
A second sialic acid binding site (hemadsorption site) was previously revealed in avian N9 NA when the wild-type NA crystals were soaked with sialic acid at 4°C, but not at 18°C (29), and was proposed to account for its hemadsorption activity at 4°C. It was further found that all avian virus NAs possess a high level of hemadsorption activity, whereas human N1 and N2 NAs display much weaker activity (12, 28). A minimal signature for hemadsorption activity in avian NAs of all nine antigenic subtypes was proposed as the triple-serine SXXSXS loop of residues 367 to 372, Asn400, and Trp403 (28, 29). In wild-type and mutant TZ205 NAs, residue 367 is an Asn (see Table S1 in the supplemental material), which is glycosylated as seen in the electron density maps during structure refinement (data not shown). Furthermore, residues 372, 400, and 403 are Leu, Arg, and Arg, instead of Ser, Asn, and Trp, in TZ205 D151 NA, eliminating the possibility here for the second sialic acid binding site that is observed in avian NAs.
(iv) TZ205 D151G mutant in complex with avian receptor 3′-SLN.The crystal structure of the D151G mutant (TZ205 G151 NA) in complex with the substrate trisaccharide, 3′-SLN, revealed remarkably intact and well-ordered features for all three saccharides, Sia-1, Gal-2, and GlcNAc-3, in the NA active site (Fig. 4G; see Fig. S4F in the supplemental material). The terminal Sia-1 moiety is positioned similarly in the active site as sialic acid in complex with wild-type NA (Fig. 4F and Fig. S5C). The 3′-SLN binds in an extended conformation with the N-acetyllactosamine (Gal-2 and GlcNAc-3) being almost perpendicular to Sia-1 and directed vertically out of the active site.
The Sia-1 moiety forms the majority of hydrogen bonds and contacts with NA active site residues. Similar to the hydrogen bond scheme between the wild-type NA and sialic acid (Fig. 4F), Sia-1 hydrogen bonds with three arginine residues at positions 118, 371, and 292, as well as Arg152 and Glu277 (Fig. 4G). Due to the D151G mutation, no hydrogen bond can form between residue 151 and Sia-1; instead, the 4-hydroxyl of Sia-1 shifts about 0.5 Å toward the base of the active site and hydrogen bonds with the side chain of Glu119 (Fig. 4G). A water-mediated hydrogen bond between the carbonyl oxygen of Val149 and O4 of Gal-2 is the only other hydrogen bond observed between the NA and 3′-SLN. Interestingly, although interpretable electron density was observed for GlcNAc-3 (see Fig. S4F in the supplemental material), no direct hydrogen bonds or van der Waals interactions were formed between NA and GlcNAc-3. Similarly, in published HA–3′-SLN structures, it was also found that only the first two saccharides, Sia-1 and Gal-2, form hydrogen bonds with HA, but not with GlcNAc-3 (7). Avian-like glycans adopt a trans conformation of the α2-3 sialoside linkages (4), and it seems that only the first two saccharides, Sia-1 and Gal-2, are required for binding to NA, as found in HA-receptor complexes (reviewed in reference 7) and some recent structures (14, 33, 36).
As predicted from molecular modeling studies for N2 NA (27) and N1 NA (22), the pyranose ring of Sia-1 adopts an energetically unfavorable, distorted boat configuration in the mutant NA–3′-SLN complex structure (Fig. 4G; see Fig. S5C in the supplemental material), providing insights for the NA catalytic mechanism. In contrast, all the Sia-1 moieties in the reported HA-sialic acid receptor complexes adopt the energetically favorable chair conformation (7).
(v) TZ205 D151G mutant in complex with human receptor 6′-SLN.The structure of the D151G mutant in complex with human receptor 6′-SLN revealed again the substrate rather than its cleavage product, but only Sia-1 and Gal-2 are ordered (Fig. 4H; see Fig. S4G in the supplemental material). As in the mutant NA–3′-SLN complex, the 6′-SLN binds in an extended conformation with Gal-2 almost perpendicular to Sia-1 (Fig. 4H and Fig. S5D). Sia-1 remains in the same position with a similar hydrogen bonding scheme, but the Gal-2 position is elevated about 1.5 Å in the binding site (Fig. S5D). Sia-1 with a distorted boat pyranose ring interacts with the mutant NA in a way similar to that seen in the mutant NA 3′-SLN complex, but Gal-2 only forms an intramolecular, saccharide-saccharide interaction with Sia-1 (Fig. 4H). Consistent with the results from glycan array binding assay and ITC binding analysis, the poor electron density map and fewer interactions with NA residues suggest a reduced binding affinity for 6′-SLN trisaccharide due to the cis conformation of α2-6 sialoside linkages compared to the trans 3′-SLN.
The D151G mutation reduces NA activity and enhances sialoside binding by N1 and N2 NAs from other human and swine viruses.As the TZ205 D151G NA mutation was found to reduce enzymatic activity while retaining the ability to bind sialic acid receptors in its active site, we then wished to determine whether it reflected a general effect of the D151G mutation in influenza virus neuraminidases. To this end, three NAs from human pandemic virus (A/Hong Kong/68 [H3N2] [HK68], A/California/04/2009 [H1N1] [Cali09], and A/New York/06/2009 [H1N1] [NY06]) and one NA from swine virus (A/Swine/England/WVL7/1992 [H1N1] [sw/England92]) were engineered with the D151G mutation and expressed in a baculovirus expression system.
The NA activities of all four D151G NA mutants were assessed by a solution-based assay. With the 4-MU-NANA substrate at room temperature (∼22°C), all the D151G NA mutants (Table 3) had significant decreases in both kcat and Km values from wild-type NAs (34). For pandemic HK68 N2 NA, the D151G mutation decreased kcat by ∼37-fold and Km by ∼10-fold. In addition to N2 NAs, such as TZ205 NA and HK68 NA, the D151G mutation can also decrease kcat and Km values in N1 NAs. Upon D151G mutation, the kcat values decreased about 10-, 16-, and 13-fold, and the Km values decreased about 28-, 25-, and 1.8-fold for Cali09 N1, NY06 N1, and sw/England92 N1, respectively (Table 3).
Summary of NA activities of NA D151G mutants in the viruses tested
To assess the possibility for substrate specificity changes upon the D151G mutation, all four D151G mutant NAs were screened for activity against diverse sialoside sequences on a glycan microarray cleavage analysis. The D151G mutant NAs of HK68 N2, NY06 N1, and Cali09 N1 also showed much reduced activities for α2-3 and α2-6 sialosides compared to wild-type NAs with a slight preference for α2-3 sialosides (Fig. 5A, C, and E). Interestingly, although the sw/England92 G151 NA showed much reduced activity against α2-3 sialosides, it had significant activity against some α2-6 sialosides (Fig. 5G).
NA cleavage activity and specificity for glycan structures on a glycan microarray. (A, C, E, and G) HK68 G151 NA (A), NY06 G151 NA (C), Cali09 G151 NA (E), and sw/England92 G151 NA (G) were measured against printed glycans on a microarray (see Fig. S7 in the supplemental material) with the same starting NA concentration of 40 μg/ml. Cleavage signals are shown by filled bars with the color of the bar indicating the sialic acid linkages: control glycans that terminated in galactose are shown in black (glycans 1 and 2), α2-3 linkage glycans are shown in blue (glycans 3 to 33), α2-6 linkage glycans in red (glycans 34 to 51), and α2-3/6 mixed glycans in cyan (glycans 52 and 53). (B, D, F, and H) NA binding affinity and specificity were also investigated against the same printed glycans for HK68 G151 NA (B), NY06 G151 NA (D), Cali09 G151 NA (F), and sw/England92 G151 NA (H) with the same starting concentration of 16 μg/ml. Binding signals are shown by filled bars by sialic acid linkages using the same color scheme as described above for panels A, C, E, and G (see text for details).
Significantly, similar to TZ205 G151 NA, glycan microarray binding analysis with the four purified D151G mutant NAs at 16 μg/ml revealed strong binding activity for most α2-3 sialosides and variable binding to some α2-6 sialosides (Fig. 5B, D, F, and H). In contrast, none of the wild-type NAs from these four viruses showed strong binding to any of the tested α2-3 and α2-6 sialosides (see Fig. S6 in the supplemental material).
To exclude the possibility that this sialic acid binding was from the second binding site in NA that is present in avian viruses (12, 13), sequence comparisons found that the minimal signature sequences consisting of the triple-serine SXXSXS loop of residues 367 to 372, Asn400, and Trp403 (28, 29) were not conserved in the four NAs (see Table S1 in the supplemental material), which likely abolishes any possible hemadsorption activity at this site. Taken together, the four D151G NA mutants appear to have sialic acid binding and cleavage activities in their active sites with conserved residues as in TZ205 G151 NA (Table S2).
DISCUSSION
HA and NA, the two surface glycoproteins of the influenza virus membrane are both required to mediate interactions of the virus with sialic acid-containing receptors. HA binds to sialic acid-containing glycans of host cell receptors to mediate virus entry, while NA removes the sialic acid from the same receptors to ensure efficient progeny virus release. In addition, HA binds to sialic acids on virion surface glycoproteins HA and NA, leading to virus clumping in the absence of NA activity. Recently, Lin et al. reported mutant D151G NAs that appeared to deviate from this paradigm, in that the NA supported binding of the virus to cellular receptors (15). In this study, we investigate the structural basis for the binding activity of NA with this D151G mutation. Results from NA structure analysis and a variety of assays, including glycan array cleavage and binding assays, ITC binding experiments, and NA activity assays demonstrate conclusively that the D151G NA mutants, as exemplified by the human seasonal influenza H3N2 virus neuraminidase variant TZ205 and analogous mutants reported here, have sialic acid binding activity for avian α2-3 and human α2-6 sialic acid glycans in their active sites.
A previous study concluded that, in some recently circulating human H3N2 viruses, replacement of Asp151 to Gly, Asn, or Ala in the N2 NA enables attachment to sialic acid receptors on RBCs (15). It was further suggested that the D151G mutation of the NA might induce a change in the NA specificity so that it possessed the capacity to bind to receptors that were refractory to NA enzymatic cleavage, without altering its ability to remove receptors for HA. However, the results of our studies clearly indicate that the D151G mutant possesses much reduced NA cleavage activities yet retains binding to the same sialosides that it can cleave (Fig. 2). Although the early report concluded that the activity of NA with the D151G mutation was normal, whole virus was used for the analysis. Therefore, because the mutant NA has negligible activity under normal assay conditions, we believe it is likely that the apparent discrepancy from the earlier report could have resulted from a minor population of virus with the wild-type (D151) NA, which exhibited normal kinetics.
The catalytic reaction of NA can be divided into four major steps (9, 11, 27). The first step is binding of the sialic acid-containing receptors. As illustrated in the crystal structures of TZ205 G151 NA in complex with substrates 3′-SLN and 6′-SLN (Fig. 4E and F), the carboxylate group moves and rotates from an axial position to a pseudoequatorial position due to the charge interactions with the N2 NA arginine residues at positions 118, 292, and 371 and steric constraints from residues at the base of the active site, such as Tyr406. The D151G NA mutant shows tight binding to both 3′-SLN and 6′-SLN with ITC-measured apparent KD values of 30.0 ± 5.5 μM and 645 ± 55 μM, respectively (Fig. 3), which are much stronger than the corresponding affinities of HA with its sialylated receptors; a direct nuclear magnetic resonance (NMR) estimate for the sialoside affinity of a human H3 HA indicated KD values of 2.1 mM for α2,6-sialyllactose (6′-SL) and 3.2 mM for α2,3-sialyllactose (3′-SL) (24). The second step of the catalytic reaction involves proton donation from the solvent and formation of transition state intermediates. Active site residues D151, R152, and E277 are believed to stabilize the transition state intermediate (27), and this role for D151 is consistent with the results of this study of the D151G mutant showing drastically reduced enzymatic activity. In the final two steps of catalytic reaction, the product sialic acid is formed and released. As shown in the structure of TZ205 D151 NA in complex with sialic acid, the pyranose ring was refined to a boat configuration (Fig. 4F), as in most other neuraminidase-sialic acid structures (11). It was proposed that the sialic acid is expelled from the active site by changing its conformation from the initial β-anomer to the thermodynamically more stable α-anomer in solution (27).
For an enzyme-catalyzed reaction, the catalytic efficiency (kcat/Km) is defined by two kinetics parameters, kcat and Km. kcat, the turnover number of the enzyme, defines the maximum number of substrates converted to product per unit of time and enzyme under optimal conditions, while Km is the concentration of substrate needed to reach half the maximum velocity, which is a measure of substrate affinity with smaller Km values reflecting tighter binding. With the 4-MU-NANA substrate at 37°C, the kcat and Km values are 26.8 s−1 and 29.0 μM for TZ205 D151 NA, respectively, and 0.167 s−1 and 0.824 μM for its D151G mutant, respectively. Therefore, the D151G mutant has a much reduced kcat (160×) and Km (35×) against 4-MU-NANA compared to the wild type, which may explain why the mutant shows apparent binding to sialic acid receptors with tight substrate binding (low Km) and low turnover (low kcat). This explanation was further confirmed by the survey of D151G mutation in three N1 NAs and one N2 NA, which revealed much reduced kcat and Km values against 4-MU-NANA for all four mutant NAs compared to wild-type NAs (Table 3) and binding activity toward α2-3 and α2-6 sialosides.
Although our results shed light on the molecular mechanism for receptor binding by the NA with the D151G mutation, the biological significance is not entirely clear. Both the HA and NA engage glycan receptors on host cells, and their interplay has important implications for viral replication and transmission (18, 32). During infections, it has been well established that the NA is important in release of virus from infected cells, and it is widely believed that preventing virus release is the basis for the efficacy of oseltamivir carboxylate and zanamivir. However, there is some evidence that the NA also plays important roles in the initiation of influenza virus infection in the human airway epithelium (16), where it can play a role in removal of decoy receptors on mucins, cilia, and cellular glycocalyx that would impede viral access to the receptors on the target cell surfaces (16). Furthermore, it has been known for some time that, in addition to the catalytic site, the NAs of some avian influenza viruses have a separate hemadsorption site that is not present in human viruses (10, 12, 13). The biological significance of this NA hemadsorption activity was suggested to be enhancement of the catalytic efficiency of avian NAs by binding natural multivalent substrates without an obvious impact on the initiation of infection (28).
Recently, a D151G mutation was noticed in MDCK cultures of recent H3N2 viruses that grow poorly in these cells (15). Given the characteristics of the mutant NA, the mutation could create a replication advantage at the binding step by reduced cleavage of cell surface receptors or increased binding to receptors present on MDCK cells. This activity would represent a nonclassical role for the NA. Given the importance of the balance between the HA and NA, it is possible that similar situations occur in nature, and a systematic study of NAs to identify those with low activity may reveal examples that occur in natural settings.
ACKNOWLEDGMENTS
We thank the National Influenza Laboratory in Dar es Salaam, United Republic of Tanzania, and the U.S. Centers for Disease Control and Prevention for virus and sequence information. We gratefully acknowledge the authors and originating and submitting laboratories of the sequences in GISAID's EpiFlu Database on which this research is based. The database accession numbers are detailed in Materials and Methods. All submitters of sequence data may be contacted directly via the GISAID website (www.gisaid.org). We thank Ruben Donis of the Centers for Disease Control and Prevention for cDNAs of TZ205 NAs and for helpful advice and comments on the manuscript. We also thank Henry Tien of the Robotics Core at the Joint Center for Structural Genomics for automated crystal screening, Jean-Philippe Julien for excellent assistance in ITC experiments, and Rui Xu for helpful discussions. X-ray diffraction data sets were collected at the Stanford Synchrotron Radiation Lightsource (SSRL) beamlines 12-2 and 11-1 and the Advanced Photon Source (APS) beamline 23-ID-B.
This study was supported in part by the governments of Egypt and the United States. The work was supported in part by NIH grant AI058113 (I.A.W.), the Skaggs Institute for Chemical Biology, and the Scripps Microarray Core Facility. The glycans used in the plate binding assay were provided by the Consortium for Functional Glycomics (http://www.functionalglycomics.org/) supported by NIH grant GM62116 (J.C.P.). Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy (DOE) Office of Basic Energy Sciences. The Stanford Synchrotron Radiation Lightsource Structural Molecular Biology Program is supported by the U.S. DOE Office of Biological and Environmental Research and by the NIH National Center for Research Resources Biomedical Technology Program (P41RR001209) and the National Institute of General Medical Sciences (NIGMS). The GM/CA CAT 23-ID-B beamline has been funded in whole or in part with federal funds from the National Cancer Institute (Y1-CO-1020) and NIGMS (Y1-GM-1104). Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science, under contract no. DE-AC02-06CH11357.
The content is solely the responsibility of the authors and does not necessarily represent the official views of NIGMS or the NIH.
FOOTNOTES
- Received 13 June 2012.
- Accepted 20 September 2012.
- Accepted manuscript posted online 26 September 2012.
This article is publication 21792 from The Scripps Research Institute.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/JVI.01426-12.
- Copyright © 2012, American Society for Microbiology. All Rights Reserved.