• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jvirolPermissionsJournals.ASM.orgJournalJV ArticleJournal InfoAuthorsReviewers
J Virol. Mar 1998; 72(3): 2456–2462.
PMCID: PMC109546

Mutations in a Conserved Residue in the Influenza Virus Neuraminidase Active Site Decreases Sensitivity to Neu5Ac2en-Derived Inhibitors


The influenza virus neuraminidase (NA)-specific inhibitor zanamivir (4-guanidino-Neu5Ac2en) is effective in humans when administered topically within the respiratory tract. The search for compounds with altered pharmacological properties has led to the identification of a novel series of influenza virus NA inhibitors in which the triol group of zanamivir has been replaced by a hydrophobic group linked by a carboxamide at the 6 position (6-carboxamide). NWS/G70C variants generated in vitro, with decreased sensitivity to 6-carboxamide, contained hemagglutinin (HA) and/or NA mutations. HA mutants bound with a decreased efficiency to the cellular receptor and were cross-resistant to all the NA inhibitors tested. The NA mutation, an Arg-to-Lys mutation, was in a previously conserved site, Arg292, which forms part of a triarginyl cluster in the catalytic site. In enzyme assays, the NA was equally resistant to zanamivir and 4-amino-Neu5Ac2en but showed greater resistance to 6-carboxamide and was most resistant to a new carbocyclic NA inhibitor, GS4071, which also has a hydrophobic side chain at the 6 position. Consistent with enzyme assays, the lowest resistance in cell culture was seen to zanamivir, more resistance was seen to 6-carboxamide, and the greatest resistance was seen to GS4071. Substrate binding and enzyme activity were also decreased in the mutant, and consequently, virus replication in both plaque assays and liquid culture was compromised. Altered binding of the hydrophobic side chain at the 6 position or the triol group could account for the decreased binding of both the NA inhibitors and substrate.

Influenza virus possesses two surface glycoproteins, hemagglutinin (HA) and neuraminidase (NA). HA is responsible for recognition of the cell surface receptor, and NA is thought to be responsible for the elution of progeny virions from infected cells, and from each other by cleavage of terminal sialic acid residues (Neu5Ac). The potential of NA as a target for antiviral therapy was investigated many years ago, when Meindl and Tuppy (13) first synthesized the unsaturated sialic acid analog Neu5Ac2en, which inhibited influenza virus replication in vitro but not in vivo (16, 17). Based on the knowledge of the three-dimensional structure of NA complexed with Neu5Ac (23), a derivative of Neu5Ac2en with a substitution of a guanidinium group at the 4 position, 4-guanidino-Neu5Ac2en (zanamivir), has been synthesized and has been shown to have potent antiviral activity both in vitro and in vivo when administered topically within the respiratory tract (7, 25, 27). The search for compounds with altered pharmacological properties has led to the identification of a novel series of influenza virus NA inhibitors in which the triol group of zanamivir was replaced with a hydrophobic group linked by a carboxamide at the 6 position (21). An essential aspect of drug development is determining if and how resistant variants may arise after prolonged exposure to the inhibitor. We and others have reported the generation of variants with decreased sensitivity to zanamivir as a result of mutations in either NA (1, 3, 4, 12, 22) or HA (3, 11). We were interested in determining whether we could also isolate variants to the 6-carboxamide derivative of zanamivir by in vitro passaging in the presence of the inhibitor.



The NWS/G70C virus was originally obtained from Robert Webster (St. Jude Children’s Medical Research Center, Memphis, Tenn.). The reassortant contains the NA from the A/tern/Australia/G70C/75 avian virus, and the rest of the genes are thought to derive from the NWS parent.

Cells and media.

MDCK cells were grown in Dulbecco’s modified Eagle’s medium/Ham’s F12 (Trace Biosciences) supplemented with 2% fetal calf serum (Trace Biosciences), 1% Ultroser G (Sepracorp), glutamine, penicillin-streptomycin (Trace Biosciences), and amphotericin B (Fungizone; Squibb).


5-N-Acetyl-4-guanidino-6-methyl(propyl)carboxamide-4,5-dihydro-2H-pyran-2-carboxylic acid (6-carboxamide [21]), 4-guanidino-Neu5Ac2en (zanamivir), 4-amino-Neu5Ac2en, and the 6-carbocyclic NA inhibitor, GS4071 (8), were synthesized by Glaxo Wellcome Research and Development Ltd. (Stevenage, United Kingdom). Neu5Ac2en was obtained from Sigma (St. Louis, Mo.). The structures of the inhibitors are shown in Fig. Fig.1.1.

FIG. 1
Structure of NA substrate and inhibitors used in the study. 1, Neu5Ac; 2, Neu5Ac2en; 3, 4-amino-Neu5Ac2en; 4, zanamivir (4-guanidino-Neu5Ac2en); 5, 6-carboxamide; 6, GS4071.

Generation of resistant virus.

The NWS/G70C virus was passaged by limiting-dilution passaging as described previously (11). Initial passaging took place in 0.2 μg of 6-carboxamide per ml, increasing slowly to 3.2 μg/ml by passage 15. From passage 15, the drug concentration was increased to 10 to 20 μg/ml. Relative resistance was determined in a plaque assay at various passages, and potential variants were selected for drug sensitivity testing in both cell culture and enzyme-based assays.

Sequencing of influenza virus NA and HA genes.

Influenza virus RNA was extracted directly from tissue culture supernatants as previously described (1). The PCR products were sequenced with the PRISM Ready Reaction Dyedeoxy Terminator cycle-sequencing kit (Applied Biosystems, Foster City, Calif.).

NA enzyme inhibition assay.

NA enzyme activity and drug inhibition assays were based on the method of Potier et al. (18) with methylumbelliferone N-acetylneuraminic acid (MUNANA) as the substrate, as described by Blick et al. (1). The sensitivity of NA to Neu5Ac2en, 4-amino-Neu5Ac2en, zanamivir, 6-carboxamide, and GS4071 was tested by using dilutions of inhibitors ranging from 0.0001 to 10,000 μM.

Enzyme kinetics.

NA was solubilized from whole virus by the addition of Nonidet P-40 (final concentration, 0.1% [vol/vol]) to a sucrose gradient-purified virus suspension. Hydrolysis of MUNANA and the rate of binding of 6-carboxamide, zanamivir, 4-amino-Neu5Ac2en, and Neu5Ac2en were determined as previously described (6, 11). A simplified kinetics of binding of the GS4071 was monitored by pretreating the samples with a 50% inhibitory concentration (IC50) of inhibitor for 2, 4, 6, 8, 10, 15, and 20 min before the addition of the MUNANA substrate. The percent inhibition was compared to that resulting from preincubation with inhibitor for 60 min.

Purification of NA.

Embryonated eggs (10 days old) were inoculated with the ju2 virus since this gave the best yield of virus for NA purification. The eggs were chilled after 48 h, and the allantoic fluids were harvested. Virus was pelleted by centrifugation in a type 19 rotor at 18,000 × g for 2.5 h and resuspended, and NA was cleaved with pronase (Calbiochem, La Jolla, Calif.) at 1 mg/ml and purified on a Superose 12 column, as described previously (10).

Specific activity of NA.

The relative specific activity was determined for purified NA heads, redissolved crystals of the NA, and intact virions by quantitating the amount of native NA protein in an NC-10 antibody capture enzyme-linked immunosorbent assay (ELISA) and comparing this to the amount of NA activity in a MUNANA enzyme assay (1) with substrate at 100 μM. The relative specific activity of the NA on the surface of infected cells was determined by fixing cells in 96-well plates with 1% formalin in normal saline at 4°C. MUNANA reaction mix was added to the wells, and after 1 h at 37°C the reactions were stopped and the supernatants were transferred to an Optiplate (Canberra Packard) for reading in the fluorimeter (Perkin-Elmer LS50B). NA protein was quantified on the same cells with the NC-10 monoclonal antibody.

Cell culture drug sensitivity assays.

For plaque selection and to determine the relative sensitivity of the passaged viruses, approximately 100 PFU of virus was tested against concentrations of 6-carboxamide ranging from 0.001 to 10 μg/ml, zanamivir ranging from 0.0003 to 3 μg/ml, 4-amino-Neu5Ac2en ranging from 0.01 to 100 μg/ml, and GS4071 ranging from 0.0003 to 3 μg/ml.

For a yield reduction assay, MDCK cells in a 24-well cluster dish were infected with virus at a multiplicity of infection (MOI) of 0.1. Inhibitor concentrations ranged from 0.01 to 100 μg/ml for zanamivir, from 0.0001 to 100 μg/ml for 6-carboxamide, and from 0.0001 to 100 μg/ml for GS4071. Samples were harvested at 24 h postinfection. The yields were determined in a 50% tissue culture infective dose (TCID50) assay by fixing the cells with formalin after 4 days and staining them with neutral red, or, if the cytopathic effects were not distinct, the virus replication was detected by a MUNANA enzyme assay on the formalin-fixed cells.

Stability of the resistance phenotype.

To determine the stability of the resistant phenotype, the mutant viruses were serially passaged by limiting dilution in the absence of drug for 10 passages, as described previously (1).

Effect of exogenous NA.

To determine whether a deficiency in NA activity contributed to the poor growth of the ir2 virus, variant and parent stocks were subjected to plaque formation in the presence of Clostridium perfringens NA (Sigma) at 1 mU/ml. The plaque sizes were compared for the NA-supplemented assay and the control cultures.

Kinetics of replication.

To determine whether the mutations affected the ability of the mutants to replicate in cell culture, MDCK cells were infected at a MOI of 0.1 PFU/cell. Samples were taken from individual wells at 8, 16, 24, and 32 h postinfection. Supernatants were assayed by the TCID50 assay. The cells were fixed with formalin after 4 days and were stained with neutral red or were assayed for enzyme activity in a MUNANA-based assay on the fixed cells. Kinetics of replication of the ir2 and parent viruses were also determined in the presence of 1 mU of exogenous NA per ml.

Thermal stability of NA and HA.

Dilutions of variant and control viruses were chosen to give a reading of around 50 fluorescence units in a MUNANA assay for 1 h. Aliquots were then incubated at 4 and 37°C, samples were taken daily for 14 days, and the residual NA activities were compared.

Replicate samples of virus were incubated at 37, 45, 50, 55, and 60°C for 1 h. HA activity was assayed by binding samples to an enzyme-linked immunosorbent assay plate coated with 10 μg of fetuin (Sigma) per ml. Bound virus was then detected with a monoclonal anti-HA antibody (kindly provided by E. M. Anders, Melbourne University), followed by detection with a sheep anti-mouse horseradish peroxidase conjugate (Silenus). The residual activity was calculated compared to that of virus incubated on ice.


To determine whether HA mutations affected the affinity for the cell receptor, the efficiency of adsorption to cells was monitored. Approximately 100 PFU of virus was adsorbed to cells at room temperature for 15, 30, or 60 min before removal of inoculum and overlaying with agarose. The control was the same inoculum adsorbed for 60 min at room temperature and then for 60 min at 37°C. The plates were incubated at 37°C for 4 days, when they were fixed with formalin and stained with neutral red. The numbers of plaques at each time point were expressed as a percentage of the control values.


Generation of resistant virus.

There have now been several reports of variants which have been selected with decreased sensitivity to zanamivir (1, 3, 4, 11, 22). To determine whether similar mutations would arise in response to the modified compound, the NWS/G70C virus was passaged by limiting dilution in the presence of 6-carboxamide.

A variety of resistant viruses, containing either NA and/or HA mutations, which possessed different phenotypic properties were isolated at different passages (Table (Table1).1). The same NA mutation, Arg292Lys, was found in three variants, both as the single mutation and in conjunction with additional HA mutations. This NA mutation was in a site which had been previously conserved and which forms part of a catalytic triad of arginines (23) in the active site of the enzyme. Although the NA mutation dramatically reduced the plaque size, once it was acquired, it was maintained through all subsequent passages.

Passage history and properties of 6-carboxamide-derived mutants

Of the three HA mutations, one arose before the NA mutation but was lost on further passaging and two arose after the NA mutation. The HA mutation in the ju2 variant partially rescued the minute-plaque phenotype of the ir2 NA mutant to produce medium plaques, and the HA mutation in the ty4 variant fully rescued the large-plaque phenotype. A reassortant, xy22, which contained only the latter HA mutation was generated to determine the relative contribution of the HA mutation to resistance in the ty4 variant. HA variants with decreased sensitivity to the NA inhibitors have previously been isolated (12). These variants had mutations in or around the receptor binding sites, which we suggested altered the affinity of the HA for the cellular receptor. Structural analysis of these HA mutants revealed that the Gly143Glu mutation in ju2 was close to the 134 to 138 residues which form the right edge of the receptor binding site (15); hence, this mutation could have some effect on receptor binding. Although they were not in the receptor binding site, the other two HA mutations introduced potential glycosylation sites, Asn199Ser (glycosylation on Asn197) (in ty4/xy22) and Ser165Asn (glycosylation on Asn 165) (in tg2). The Asn165 residue is glycosylated in other influenza virus strains (26); hence, it is likely that glycosylation, which could affect receptor binding, also occurs here. In ty4/xy22, a new carbohydrate side chain at the Asn197 site could mask the binding of residues 190, 194, and 195 to the receptor (15), thus decreasing the efficiency of receptor binding.

NA enzyme inhibition assay.

The sensitivity of the mutant NA to the inhibitors was determined in a MUNANA-based enzyme assay. The results are presented in Fig. Fig.2.2. In contrast to the previously described Glu119Gly NA mutant (1), the Arg292Lys NA mutant was equally cross-resistant to zanamivir and 4-amino-Neu5Ac2en and slightly less resistant to Neu5Ac2en. It was significantly more resistant to 6-carboxamide (2.9 log10 more resistant), against which the mutant was raised, with the greatest resistance (3.9 log10 more resistant) seen to GS4071. Unlike the situation for the Glu119Gly mutant, this data suggests that the substitution at the 6 position and not the 4 position is a key determinant in the resistance of the Arg292Lys NA mutant.

FIG. 2
Inhibition of the NWS/G70C parent and ir2 (Arg292Lys) NA variant enzyme activity in a MUNANA assay. (a) Neu5Ac2en; (b) 4-amino-Neu5Ac2en; (c) zanamivir; (d) 6-carboxamide; (e) GS4071 analog.

Kinetic analysis of zanamivir and 6-carboxamide binding.

The altered binding of all Neu5Ac2en-based inhibitors was further confirmed by determination of binding constants from detergent extracts of the virus NA (Table (Table2).2). Progress plots for the Arg292Lys mutant NA-catalyzed hydrolysis of MUNANA in the presence of several concentrations of zanamivir or 6-carboxamide demonstrated that in contrast to the slow binding (14) of zanamivir by the parent enzyme (11), neither inhibitor was bound slowly by the Arg292Lys mutant NA. The kinetics of binding of the GS4071 showed that this also had lost the slow binding seen with the parent virus.

Determination of binding constants for the NWS/G70C parent and the Arg292Lys NA mutant

Specific activity of the NA.

We previously observed that a mutation at another conserved site, Glu119Gly, had not altered either the Km or the relative specific activity of the enzyme when the amount of native NA protein, rather than just total protein, was quantitated (1). NA protein, determined by NC-10 antibody reactivity, was titrated in parallel with the enzyme activity for both purified NA and whole virions. The results for purified NA are presented in Fig. Fig.3.3. Unlike the previous mutation, the relative specific activity of the Arg292Lys enzyme was only 20% of that of the parent enzyme. When NA protein levels and NA activity were assayed on the cell surface, the NA also demonstrated around 20% of the activity of the parent NA. Cleavage of fetuin in an ELISA also was reduced to around 30% of the activity of the parent NA (results not shown). In addition, unlike the Glu119Gly mutant, for which we could show no difference in the binding of substrate (1), a decrease in substrate binding was demonstrated by an altered Km for MUNANA (Table (Table2).2).

FIG. 3
Relative specific activity of the NWS/G70C parent and Arg292Lys NAs. Crystals of both were redissolved, and the activity was assessed in a 96-well MUNANA-based enzyme activity assay. The same dilutions were titrated in parallel in a capture ELISA, using ...

Cell culture drug sensitivity assays.

To determine the magnitude of resistance in cell culture, variants were assayed in a plaque inhibition assay against 4-amino-Neu5Ac2en, zanamivir, 6-carboxamide, and GS4071. NA inhibitors decrease the size of plaques in a plaque inhibition assay before they affect the number of plaques (11). It is thus often difficult to determine an IC50, depending upon whether it is based on plaque size or number. Resistance based on a decrease in plaque size is presented in Table Table3.3. It was impossible to directly determine the contribution of the NA mutation to resistance in a plaque assay, based on a decrease in the plaque size, since the ir2 virus plaques were minute even in the absence of inhibitor. Sensitivity could only be inferred from the difference in the sensitivities of the ty4 HA/NA mutant and the xy22 HA mutant.

Sensitivity of the NWS/G70C parent virus and 6-carboxamide-derived variants to NA inhibitors in cell culture assays

As we have previously observed (1, 11), NA and HA mutations each contributed to resistance in cell culture, either alone or in association with each other. In a plaque assay, the HA variants tg2 and xy22 demonstrated a similar reduction in sensitivity to each of the inhibitors. Since the mechanism of resistance by HA mutations is independent of which inhibitor is tested (11), this is not surprising.

Resistance to 4-amino-Neu5Ac2en and zanamivir was similar for ty4, with both NA and HA mutations, and was only three times greater than the resistance seen with the xy22 single HA mutation. Thus, the additional contribution of the NA to resistance to these two inhibitors would appear to be very small. For both NA-HA mutants, there was a greater decrease in sensitivity to 6-carboxamide, with the greatest decrease in sensitivity seen to GS4071, which correlates with the relative resistance seen to these 6-substituted inhibitors in enzyme inhibition assays. Since the ju2 and ty4 variants contained the same NA mutation, the results also suggested that the HA mutation in the ty4 variant conferred greater resistance than did the HA mutation in the ju2 virus.

To more directly quantify the contribution of the NA mutation to resistance in cell culture, the sensitivities of the mutants to zanamivir, 6-carboxamide, and GS4071 were determined in yield reduction assays. The results are presented in Table Table3.3. Although the Arg292Lys NA mutation conferred 1.5 log10 resistance to zanamivir in an enzyme assay, the resistance of the ir2 mutant to zanamivir in the yield reduction assay was only 2-fold. This supports the small effect of the NA mutation inferred from the results with the ty4 double mutant in the plaque assay. In contrast, there was approximately a 2 log10 reduction in the sensitivity of the ir2 mutant to 6-carboxamide and approximately a 3 to 4 log10 reduction in its sensitivity to GS4071. This further supports the greater resistance to these inhibitors in both enzyme and plaque assays. As observed in the plaque assay, mutants with both HA and NA mutations were more resistant than was the mutant with the NA single mutation. Thus, it is clear that in cell culture, the HA mutations conferred cross-resistance to all the NA inhibitors while the Arg292Lys NA mutation predominantly decreased the binding of the 6-substituted analogs.

Stability of the resistance phenotype.

The ir2 NA mutant and the ju2 mutant were passaged through 10 passages in cell culture in the absence of inhibitor to determine the stability of their mutations. Even though their plaque-forming ability was compromised, there was no reversion of either the NA or HA mutations.

Effect of exogenous NA.

Since the ir2 mutant produced minute plaques, we investigated whether the lack of virus spread was due to the low activity of the mutant NA by supplementing the plaque assay with exogenous C. perfringens NA. The results are shown in Fig. Fig.4.4. Although the wild-type plaques were also slightly larger, the plaque size of the ir2 mutant was greatly enhanced, and the small plaque size was thus clearly due to a deficiency in NA activity.

FIG. 4
Effect of exogenous C. perfringens NA on the plaque morphology of the NWS/G70C parent and ir2 Arg292Lys NA mutant.

Kinetics of replication.

Since the NA mutation clearly affected the ability of the virus to produce plaques, replication of the virus in the absence and presence of external NA was investigated. Cells were infected with an MOI of 0.1, and the yields were assayed as described in Materials and Methods (Fig. (Fig.5).5). The yields of the ir2 and ju2 mutants were lower at all times than that of the parent virus, with the final yield being approximately 1 log10 unit lower. When the ir2 mutant was supplemented with exogenous NA, the replication kinetics were similar to those of the wild type in the absence of external NA. Interestingly, replication of the wild type was also enhanced, as also seen in the plaque assay. This confirms the negative effect of the NA mutation on virus replication. In contrast, the ty4/xy22 HA mutation was able to compensate for the Arg292Lys mutation, resulting in growth kinetics similar to those of the wild-type virus (Fig. (Fig.5).5). Growth of the NA mutant viruses in eggs (for NA purification) was, however, poorer for all the mutants than for the parent virus. In fact, growth of the ty4 mutant was poorer than growth of the ju2 virus. Thus, NA deficiency does affect both the replication and spread of the virus in the plaque assay, as well as replication in liquid culture. While the ty4 HA mutation was able to overcome the deficiencies in replication in vitro, it was not able to do so in vivo.

FIG. 5
Kinetics of replication of the NWS/G70C virus and the mutants. Cells were infected with an MOI of 0.1, and the yields were assayed by a TCID50 assay. The effect of exogenous C. perfringens NA on the replication kinetics was also monitored for the ir2 ...

Thermal stability of NA and HA.

We had previously shown that another mutation in the active site, Glu119Gly, rendered the NA more unstable than the parent NA (12). We had also shown that two of the HA mutations rendered these HAs more thermolabile than the parent HA (11). We therefore investigated the thermostability of the Arg292Lys mutant NA and the mutant HAs. The Arg292Lys NA was no more unstable than the parent NA, in agreement with the recent observations of Gubareva et al. (4), who isolated the same mutation in a turkey H4/N2 virus after passaging in zanamivir. Furthermore, none of the three HA mutations was any more unstable than the parent HA.


We have previously proposed that the HA mutations may confer resistance to NA inhibitors due to a decreased affinity for the cellular receptor, so that there is less need for NA activity for elution of progeny virions (11, 19). To investigate whether drug resistance could be related to an alteration in affinity, adsorption kinetics for the tg2, ju2, and ty4 variants were investigated. Using a previously developed immunofluorescence assay (19), we were able to show a significant decrease in affinity only for the ty4 HA. However, with the more quantitative plaque assay used here, we were able to demonstrate that all the HA mutations altered the efficiency of adsorption of the variants to MDCK cells (Fig. (Fig.6),6), with the lowest efficiency corresponding to the most resistant ty4 variant.

FIG. 6
Kinetics of adsorption of the NWS/G70C virus and mutants. Approximately 100 PFU of virus was adsorbed onto MDCK cells for 15, 30, or 60 min before the inoculum was removed and overlay was added. The percent efficiency of adsorption was calculated by comparing ...


The influenza virus NA inhibitor zanamivir has been shown to be effective in humans when administered topically within the respiratory tract (25). The search for compounds with altered pharmacological properties has led to the identification of a novel series of influenza virus NA inhibitors in which the triol side chain at the 6 position of zanamivir was replaced with a hydrophobic group via a carboxamide linkage, designated 6-carboxamide (21). There has been a recent report of an orally bioavailable prodrug of a transition state analog of sialic acid (9), synthesized with a carbocyclic template in the place of the dihydropyran ring of the Neu5Ac2en system (8). This analog, GS4071, also has a hydrophobic side chain in the position equivalent to the hydrophobic group in the 6-carboxamide. We passaged the NWS/G70C (H1N9) virus in vitro in the presence of 6-carboxamide and tested the sensitivity of the mutants we generated to 4-amino-Neu5Ac2en and zanamivir as well as to 6-carboxamide and GS4071.

As previously observed with zanamivir-derived mutants (1, 3, 4, 11, 12), several passages were required before resistant variants were isolated. The variants had NA and HA mutations, either alone or in conjunction. Three different HA mutations arose during the passaging, with an NA mutation arising after eight passages. An HA mutation does not seem to be a prerequisite for selection of an NA variant, since we initially isolated the NA variant with no concomitant HA variant. The HA mutations conferred cross-resistance to all the NA inhibitors in cell culture, which we have also previously observed (11).

The NA mutation, Arg292Lys, was in a residue which forms part of a triarginyl cluster involved in substrate binding and catalysis (23). This residue was previously conserved in all influenza virus and bacterial NAs (2). In contrast to the framework mutant Glu119Gly (12), the Arg292Lys NA was no more unstable than the parent NA.

The NA mutation decreased the sensitivity of the virus to all inhibitors in enzyme inhibition assays, with a similar decrease in sensitivity to the 4-amino-Neu5Ac2en and zanamivir inhibitors but greater resistance to the 6-substituted inhibitors, 6-carboxamide and GS4071. However, in cell culture, the NA mutation conferred only a low level of resistance to zanamivir in either a plaque assay or a yield reduction assay. In contrast, significant resistance to 6-carboxamide and GS4071 was seen in both assays. The greatest resistance in all assays was to GS4071. This is the first report of variants with resistance to this new GS4071 inhibitor, and it suggests that the mechanism of resistance will also be through both HA and NA mutations.

Detailed structural analysis of the mutant NA (24) reveals that there are three main mechanisms which would correlate with the relative decrease in binding of the substrate and inhibitors. There is altered binding of both the triol group and the carboxylate group, which would affect the binding of substrate and all sialic acid analogs. This would correlate with a similar decrease in the binding of the 4-amino-Neu5Ac2en, zanamivir, and Neu5Ac2en. Altered binding of the 6-substituted hydrophobic group further increases resistance. The carboxamide group is normally able to bind due to the creation of a hydrophobic pocket formed by the movement of Arg224 to form a salt link with Glu276 (20). In the mutant NA, the Lys292 stabilizes the Glu276, resulting in decreased binding of the 6-carboxamide. The greatest resistance to GS4071 correlates with the lack of formation of the hydrophobic pocket.

Gubareva et al. (4) have also recently described the isolation of the same Arg292Lys mutant in an A/turkey/Minnesota/833/80 (H4N2) virus, but after passaging in zanamivir. Our results initially appear to be in contrast to theirs, since they observed significant resistance to zanamivir in cell culture. However, closer inspection of their data reveals that when wild-type reassortants were generated with NWS HA and turkey NA, the new reassortant parent virus was significantly less sensitive to zanamivir than was the original turkey HA/NA parent virus. If sensitivity was based on reduction in size in a plaque assay, the difference in the sensitivity of their NA-only mutant with respect to their new reassorted parent control, rather than to the original parent, was only twofold, which is similar to our data. Furthermore, they also discussed the lack of correlation between drug sensitivity in enzyme inhibition assays and in the plaque assay. We have clearly shown that the sensitivity in a plaque assay reflects the total contribution of both NA and HA mutations, whereas the in vitro enzyme assays represent only the contribution of the NA mutation. Thus, it would appear that the HA mutations contributed significantly to resistance with both their viruses and our viruses. In fact, although the Arg292Lys mutation was in a highly conserved residue, which conferred significant resistance to the 6-substituted compounds, its contribution to resistance to zanamivir in cell culture was small. It is therefore interesting that this mutant was selected by Gubareva et al. (4) by passaging in zanamivir.

The mutant enzyme no longer demonstrated slow binding of the zanamivir, 6-carboxamide, or GS4071 inhibitors, and substrate binding was decreased as demonstrated by an increased Km for substrate and an 80% decrease in the relative specific activity. Gubareva et al. (4) demonstrated a possible decrease of 50% in the specific activity of their Arg292Lys turkey enzyme, but since they only quantified total virus protein and not native NA protein (12), these numbers cannot be compared.

The low enzyme activity of the NA impaired the replication of virus in MDCK cells, since supplementation with bacterial NA rescued the growth potential. HA mutations complemented the NA mutation, producing greater resistance to the inhibitors and rescuing the poor-growth phenotype. Despite the deleterious effect on growth, the NA mutation was stable through 10 passages in the absence of inhibitor. Gubareva et al. (4) showed no difference in replication of their mutant virus. However, they did not specify whether they used the NA-only reassortant, and our data shows that HA mutations can clearly rescue the poorer growth potential of the NA-only mutant.

Previous HA mutations associated with decreased sensitivity to the inhibitors were in residues directly associated with receptor binding (11). We suggested that the HA mutations decreased the efficiency of binding to the cell receptor so that there was less dependence on NA activity for release of progeny virions. All three HA mutations studied here decreased the efficiency of virus binding to MDCK cells. Although the altered residues in these mutants were not directly involved in receptor binding, the ju2 Gly143Glu mutation is close to residues 134 to 138, which form the right edge of the receptor binding site (15), and it is possible that a mutation here could interfere with receptor binding. Molecular modelling of the tg2 Ser165Asn HA suggests that a new carbohydrate chain at Asn165 could impair the accessibility of adjacent HAs in the trimer to the receptor. In xy22/ty4, we observed the greatest decrease in adsorption efficiency and drug resistance, and it is likely that a new carbohydrate side chain at Asn197 masks the binding of residues 190, 194, and 195 to the receptor (15).

We have therefore isolated a panel of variants with NA and/or HA mutations with altered sensitivity to all NA inhibitors based on Neu5Ac2en and a new analog, GS4071. The NA mutation, in a previously conserved residue in the enzyme active site, affected the binding of both substrate and inhibitors, and the replication of the mutant virus was also compromised. Despite HA mutations being able to rescue the growth defect in cell culture, replication of the HA and NA mutants was poorer in eggs than was replication of the parent virus, and preliminary studies suggest that replication in mice is also affected for all mutants. Comprehensive in vivo infectivity and drug sensitivity experiments are under way for this panel of mutants. There is only preliminary evidence for zanamivir resistance in the clinical situation, and this is a single case from a severely immunocompromised child infected with influenza B virus (5). While this variant had both HA and NA mutations and altered sensitivity in an enzyme inhibition assay, there was no difference in sensitivity in cell culture-based assays. It is therefore important to determine as many phenotypic characteristics of these resistant variants as possible, to know which phenotypes may be consistent with reduced drug sensitivity in vivo.


We thank Derek Evans, Enzyme Medicinal Chemistry, Glaxo Wellcome Research and Development U.K., for synthesis of GS4071.

This work was supported in part by funding from Glaxo Wellcome Australia, Pty. Ltd.


1. Blick T J, Tiong T, Sahasrabudhe A, Varghese J N, Colman P N, Hart G J, Bethell R C, McKimm-Breschkin J L. Generation and characterization of an influenza virus neuraminidase variant with decreased sensitivity to the neuraminidase-specific inhibitor 4-guanidino-Neu5Ac2en. Virology. 1995;214:475–484. [PubMed]
2. Colman P M, Hoyne P A, Lawrence M C. Sequence and structure alignment of paramyxovirus hemagglutinin-neuraminidase with influenza virus neuraminidase. J Virol. 1993;67:2972–2980. [PMC free article] [PubMed]
3. Gubareva L V, Bethell R, Hart G J, Murti K G, Penn C R, Webster R G. Characterization of mutants of influenza A virus selected with the neuraminidase inhibitor 4-guanidino-Neu5Ac2en. J Virol. 1996;70:1818–1827. [PMC free article] [PubMed]
4. Gubareva L V, Robinson M J, Bethell R C, Webster R G. Catalytic and framework mutations in the neuraminidase active site of influenza viruses that are resistant to 4-guanidino-Neu5Ac2en. J Virol. 1997;71:3385–3390. [PMC free article] [PubMed]
5. Gubareva L V, Brenner M K, Webster R G. Abstracts of the 16th Annual Meeting of the American Society of Virology. 1997. Molecular characterization of influenza B virus from a patient after treatment with a neuraminidase targeted antiviral, abstr. W6-9.
6. Hart G J, Bethell R C. 2,3-Didehydro-2,4-dideoxy-4-guanidino-N-acetyl-d-neuraminic acid (4-guanidino-Neu5Ac2en) is a slow binding inhibitor of sialidase from both influenza A virus and influenza B virus. Biochem Mol Biol Int. 1995;36:695–703. [PubMed]
7. Hayden F G, Treanor J J, Betts R F, Lobo M, Eisenhart J D, Hussey E K. Safety and efficacy of the neuraminidase inhibitor GG167 in experimental human influenza. JAMA. 1996;275:295–299. [PubMed]
8. Kim C U, Lew W, Williams M A, Liu H, Zhang L, Swaminathan S, Bischofberger N, Chen M S, Mendel D B, Tain C Y, Laver W G, Stevens R C. Influenza neuraminidase inhibitors possessing a novel hydrophobic interaction in the enzyme active site: design, synthesis, and structural analysis of carbocyclic sialic acid analogues with potent anti-influenza activity. J Am Chem Soc. 1997;119:681–690. [PubMed]
9. Lee W-X, Escarpe P A, Chen M S, Lew W, Williams M, Zhang L, Kim C U, Bischofberger N, Cundy K C, Eisenberg E J, Mendel D B. Abstracts of the 10th International Conference on Antiviral Research, 1997. 1997. Gs4104 is a highly bioavailable prodrug of the novel potent influenza neuraminidase inhibitor GS 4071; p. A74.
10. McKimm-Breschkin J L, Caldwell J, Guthrie R E, Kortt A A. A new method for the purification of the influenza A virus neuraminidase. J Virol Methods. 1991;32:121–124. [PubMed]
11. McKimm-Breschkin J L, Blick T J, Sahasrabudhe A V, Tiong T, Marshall D, Hart G J, Bethell R C, Penn C R. Generation and characterization of variants of the NWS/G70C influenza virus after in vitro passage in 4-amino-Neu5Ac2en and 4-guanidino-Neu5Ac2en. Antimicrob Agents Chemother. 1996;40:40–46. [PMC free article] [PubMed]
12. McKimm-Breschkin J L, McDonald M, Blick T J, Colman P M. Mutation in the influenza virus neuraminidase gene resulting in decreased sensitivity to the neuraminidase inhibitor 4-guanidino-Neu5Ac2en leads to instability of the enzyme. Virology. 1996;225:240–242. [PubMed]
13. Meindl P, Tuppy H. 2-Deoxy-2,3-dehydrosialic acids. I. Synthesis and properties of 2-deoxy-2,3-dehydro-N-acylneuraminic acids and their methyl esters. Monatsh Chem. 1969;100:1295–1306.
14. Morrison J F, Walsh C T. The behaviour and significance of slow binding enzyme inhibitors. Adv Enzymol Relat Areas Mol Biol. 1988;61:201–301. [PubMed]
15. Nobusawa E, Aoyama T, Kato H, Suzuki Y, Tateno Y, Nakajima K. Comparison of complete amino acid sequences and receptor binding properties among 13 serotypes of hemagglutinins of influenza virus. Virology. 1991;182:475–485. [PubMed]
16. Palese P, Compans R W. Inhibition of influenza virus replication in tissue culture by 2-deoxy-2,3-dehydro-N-trifluoroacetylneuraminic acid (FANA): mechanism of action. J Gen Virol. 1976;33:159–163. [PubMed]
17. Palese P, Schulman J L. Inhibitors of viral neuraminidase as potential antiviral drugs. In: Oxford J S, editor. Chemoprophylaxis and viral infections of the respiratory tract. Cleveland, Ohio: CRC Press, Inc.; 1977. pp. 189–205.
18. Potier M, Mameli L, Belislem M, Dallaire L, Melanxon S B. Fluorometric assay of neuraminidase with a sodium (4-methylumbelliferyl-a-d-N-acetylneuraminate) substrate. Anal Biochem. 1979;94:287–296. [PubMed]
19. Sahasrabudhe A, Blick T, McKimm-Breschkin J L. Influenza virus variants resistant to GG167 with mutations in the haemagglutinin. In: Brown L E, Hamspon A W, Webster R G, editors. Options for the control of influenza III. Amsterdam, The Netherlands: Excerpta Medica. Elsevier Biomedical Press; 1996. pp. 748–757.
20. Smith P W, Sollis S L, Howes P D, Cherry P C, Cobley K N, Taylor H, Whittington A R, Scicinski J, Bethell R C, Taylor N, Skarzynski T, Cleasby A, Singh O, Wonacott A, Varghese J, Colman P. Novel inhibitors of influenza sialidases related to GG167. Structure-activity, crystallographic and molecular dynamics studies with 4H-pyran-2-carboxylic acid 6-carboxamides. Bioorg Med Chem Lett. 1996;6:2931–2936.
21. Sollis S, Smith P W, Howes P D, Cherry P C, Bethell R C. Novel inhibitors of influenza sialidase related to GG167. Synthesis of 4-amino and guanidino-4H-pyran-2-carboxylic acid-6-propylamides: selective inhibitors of influenza virus sialidase. Bioorg Med Chem Lett. 1996;6:1805–1808.
22. Staschke K A, Colacino J M, Baxter A J, Air G M, Bansal A, Hornback W J, Munroe J E, Laver W G. Molecular basis for resistance of influenza viruses to 4-guanidino-Neu5Ac2en. Virology. 1995;214:642–646. [PubMed]
23. Varghese J N, McKimm-Breschkin J L, Caldwell J B, Kortt A A, Colman P M. The structure of the complex between influenza virus neuraminidase and sialic acid, the viral receptor. Proteins. 1992;14:327–332. [PubMed]
24. Varghese, J. N., P. M. Colman, P. W. Smith, S. L. Sollis, T. J. Blick, A. Sahasrabudhe, and J. L. McKimm-Breschkin. A structural basis for resistance to potent neuraminidase inhibitors in a variant of influenza virus neuraminidase. Submitted for publication. [PubMed]
25. von Itzstein M, Wu W Y, Kok G B, Pegg M S, Dyason J C, Jin B, Phan T V, Smythe M L, White H F, Oliver S W, Colman P M, Varghese J N, Ryan D M, Woods J M, Bethell R C, Hotham V J, Cameron J M, Penn C R. Rational design of potent sialidase-based inhibitors of influenza virus replication. Nature (London) 1993;363:418–423. [PubMed]
26. Ward C W, Dopheide T A. The Hong Kong (H3) hemagglutinin. Complete amino acid sequence and oligosaccharide distribution for the heavy chain of A/Memphis/102/72. In: Laver G W, Air G, editors. Structure and variation in influenza virus. New York, N.Y: Elsevier/North-Holland Publishing Co.; 1980. pp. 27–38.
27. Woods J M, Bethell R C, Coates J A V, Healy N, Hiscox S A, Pearson B A, Ryan M, Ticehurst J, Tilling J, Walcott S M, Penn C R. 4-Guanidino-2,4-dideoxy-2,3-dehydro-N-acetylneuraminic acid is a highly effective inhibitor both of the sialidase (neuraminidase) and of growth of a wide range of influenza A and B viruses in vitro. Antimicrob Agents Chemother. 1993;37:1473–1479. [PMC free article] [PubMed]

Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...