Background. The viral fitness of neuraminidase inhibitor (NAI)-resistant influenza viruses is believed to be impaired. Unexpectedly, an oseltamivir-resistant A(H1N1) variant containing the H274Y neuraminidase (NA) mutation recently disseminated worldwide, suggesting that the replication and virulence properties of this mutant virus were not compromised.
Methods. In vitro replicative capacities were determined for old (A/WSN/33, A/Mississipi/3/01, A/New Caledonia/ 20/99, and A/Solomon Islands/03/06) and recent (A/Brisbane/59/2007-like) influenza A(H1N1) viruses either harboring or not harboring the H274Y NA mutation. Ferrets were infected with the A/Brisbane/59/2007-like wild-type (WT) isolate and its H274Y NA variant.
Results. Old A(H1N1) WT viruses grew at higher titers than did the A/Brisbane/59/2007-like viruses in vitro. The H274Y mutation was associated with reduced viral plaque areas in cells infected with A/WSN/33 and A/ Mississippi/3/01, whereas the 2 A/Brisbane/59/2007-like isolates showed similar plaque sizes. In ferrets, the pyrexic response induced by the A/Brisbane/59/2007-like H274Y mutant was significantly higher than that induced by the WT isolate. Nasal wash viral titers were significantly greater for the mutant isolate on day 2 after inoculation, whereas the 2 viruses showed similar titers between days 3 and 7 after inoculation.
Conclusions. The viral fitness of the recent A/Brisbane/59/2007-like H274Y variant is not impaired, consistent with its global dissemination. These results reinforce the need for new antiviral strategies.
Besides immunization strategies, antiviral agents can play a major role in the control of seasonal influenza outbreaks and are also expected to confer significant prophylactic and therapeutic benefits during an influenza pandemic. The new class of anti-influenza agents known as neuraminidase inhibitors (NAIs) includes 2 commercially available compounds (inhaled zanamivir and oral oseltamivir), which have been demonstrated to have broad in vitro activity as well as clinical benefits for the treatment and prevention of seasonal influenza virus infections [1]. In addition, the World Health Organization recommends the use of oseltamivir for the treatment of confirmed cases of infection with influenza A(H5N1) viruses as well as for postexposure prophylaxis of contacts. The recent swine-origin A(H1N1) virus that has been circulating globally since April 2009 is currently also susceptible to NAIs [2]. These compounds inhibit the activity of the viral neuraminidase (NA) enzyme, which is essential to cleave terminal sialic acids attached to glycoproteins and glycolipids, thus allowing viral propagation. As for other antivirals, however, the development and dissemination of drug resistance may significantly compromise the clinical benefits of these agents. Indeed, subtype-specific NA mutations that are part of the enzyme catalytic site (termed functional residues) or that surround it (termed framework residues) have been reported in viruses selected in vitro and in vivo [1, 3, 4].
Influenza viruses that are resistant to NAIs were not detected before the availability of zanamivir and oseltamivir. During the first 3 years (1999–2001) after the introduction of these compounds, low rates of resistance (0.22% to 0.41%) were seen in community-isolated influenza A(H1N1) and A(H3N2) viruses [3]. Also, rates of oseltamivir resistance were <4% and between 4% and 8% in treated outpatient adults and children, respectively, enrolled in clinical trials [4]. However, resistance to oseltamivir has been reported to be more important and/or more clinically relevant in some clinical therapeutic settings, such as in young hospitalized children (up to 18%), immunocompromised patients, and human cases of influenza A(H5N1) infections (reviewed in [4]). During the 2007–2008 season, unexpectedly high rates of natural resistance to oseltamivir were reported worldwide among influenza A(H1N1) viruses resulting from a mutation at a framework residue (H274Y; N2 numbering) within the NA gene. Moreover, in the subsequent season (2008–2009), almost all characterized influenza A/Brisbane/59/ 2007(H1N1)-like strains from North America and Europe have been reported to be resistant to oseltamivir [5]. There was no evidence of an association between the development of resistance in recent seasonal A(H1N1) viruses and oseltamivir use as reported in epidemiological studies [6–8]. The rapid dissemination of the oseltamivir-resistant A/Brisbane/59/2007-like variants suggests that the H274Y NA mutation does not have a compromising effect on viral fitness and the transmissibility of recent influenza viruses in this new, naturally emerged sequence context. This contrasts with previous animal studies that used different A(H1N1) strains, such as the A/Texas/36/ 91 [9] and A/New Caledonia/99/01-like [10] strains that were derived from a drug-induced resistance selection process.
To verify the hypothesis that the newly emerging A/Brisbane/ 59/2007(H1N1) viral isolates provide a sequence context that allowed the introduction of H274Y without compromising fitness, we assessed the in vitro replicative capacities of different influenza A(H1N1) viruses either harboring or not harboring the H274Y mutation. We also evaluated the virulence of recent A/Brisbane/59/2007-like isolates, including wild-type (WT) and H274Y mutant strains, in ferrets.
The recombinant WT and NA mutant (H274Y) influenza A/WSN/33(H1N1) viruses, which are mouse-adapted strains, were previously generated by reverse genetics and site-directed mutagenesis [11]. Influenza A/Solomon Islands/03/06(H1N1) and A/New Caledonia/20/99(H1N1) reference strains were provided by GlaxoSmithKline Biologicals. The influenza A/Mississippi/3/01(H1N1) WT and H274Y mutant viruses were plaque-purified strains originating from a clinical isolate and were provided by Dr Aeron Hurt (World Health Organization Influenza Centre, Victoria, Australia). This oseltamivir-resistant clinical isolate has been previously described in a surveillance study performed by the Neuraminidase Inhibitor Susceptibility Network (NISN) [3]. The influenza A/Brisbane/59/2007(H1N1)-like isolates, including A/Quebec/15230/08 (oseltamivir susceptible, or WT) and A/Quebec/15349/08 (oseltamivir resistant, or H274Y) viruses were isolated from untreated patients hospitalized for flulike illnesses in Quebec City, Canada, during the 2007–2008 influenza season. The latter viruses originated from nasopharyngeal aspirates that were passaged once on Madin-Darby canine kidney (MDCK) cells and then a second time on ST6GalI MDCK cells overexpressing the α2,6 sialic acid receptors (provided by Dr Y. Kawaoka, Department of Pathological Sciences, School of Veterinary Medicine, University of Wisconsin, Madison) [12].
Susceptibility profiles of A(H1N1) viruses against zanamivir (GlaxoSmithKline), oseltamivir carboxylate (Hoffmann-La Roche), peramivir (BioCryst), and A-315675 (Abbott Laboratories) were determined using NA inhibition assays with methylumbelliferyl-N-acetylneuraminic acid (Munana; Sigma) as the fluorescent substrate [13].
To investigate the replicative capacities of different A(H1N1) viruses, MDCK cells (for the recombinant mouse-adapted A/WSN/33 strains) or ST6GalI-expressing MDCK cells (for the A/Brisbane/59/2007-like, A/New Caledonia/20/99, A/Solomon Islands/03/06, and A/Mississippi/3/01 strains) were infected at a multiplicity of infection of 0.001 plaque-forming units (PFUs)/cell. Supernatants were collected at 12, 24, 36, 48, and 72 h after infection and assayed for numbers of PFUs by means of standard plaque assays. The mean viral plaque area for the different viruses was determined at 48 h after infection from a minimum of 20 plaques, using the ImageJ software (version 1.41) developed byWayne Rasband from the National Institutes of Health.
Groups of 4 ferrets (900–1500 g) were lightly anesthetized by means of isoflurane and received by intranasal instillation 125 µL of phosphate-buffered saline (PBS) containing 12.5×104 PFUs of A/Brisbane/59/2007-likeWT or H274Y isolates. Telemetric transmitters (Dataquest; Data Sciences) were subcutaneously implanted, and the temperature profiles of ferrets were recorded every 15 min starting 1 day before and up to 7 days after inoculation. Ferrets were weighed daily, and nasal wash samples were collected from animals on a daily basis during the 7 days after inoculation by instillation of 5 mL of PBS into the external nares of the animals. Viral titers were determined by standard plaque assays using ST6GalI-expressing MDCK cells. These experiments were conducted in accordance with the animal experimentation guidelines of the Centre Hospitalier Universitaire de Québec.
All data were expressed as mean values from 3 or 4 replicates with standard errors of the mean (SEM). Nasal wash viral titers and areas under the curve of temperature values over 7 days were compared using the unpaired t test with Prism software (version 5; GraphPad). One-way analysis of variance was done to compare in vitro viral titers. Viral plaque areas were compared using the Mann-Whitney rank-sum test.
Sequence analysis of the viral hemagglutinin (HA) genes demonstrated that 2 clinical Canadian influenza A(H1N1) viruses isolated in 2008, including an oseltamivir-resistant virus and an oseltamivir-susceptible (WT) strain, shared 99.7% and 99.1% amino acid sequence identities, respectively, with the A/Brisbane/59/07 reference strain (accession no. ISDN282676). Sequence analysis of the NA genes showed the presence of the I52V, S153N, H274Y, and D357G changes (N2 numbering) in the sequence of the resistant isolate, whereas the NA sequence of the WT isolate was unchanged relative to that of the A/Brisbane/59/07 reference strain (accession no., ISDN285099). Two additional pairs of related influenza A(H1N1) viruses were used in this study. The A/WSN/33 virus pair was generated by site-directed mutagenesis and thus differs in amino acid sequence only at NA position 274 (H or Y). The A/Mississippi/3/01-derived virus pair was obtained as part of the NISN reference panel. This pair was obtained by plaque purification from a single clinical isolate and thus also differs in NA sequence only at position 274 (H or Y).
In the NA inhibition assays, the A/Brisbane/59/2007-like H274Y mutant isolate exhibited high levels of resistance to oseltamivir and peramivir (>700-fold increases in median inhibitory concentration [IC50] values) but remained susceptible to zanamivir and A-315675 (2- and 3-fold increases in IC50 values, respectively), compared with the WT isolate. The drug phenotypes of the A/Mississippi/3/01 and the A/WSN/33 mutant viruses were comparable to those of the clinical A/Brisbane/ 59/2007-like H274Y mutant (Table 1).
Viral titers in the supernatant of infected ST6GalI-expressing MDCK cells were significantly lower (P=.05) for the A/Brisbane/59/2007-like WT isolate than for previous A(H1N1) reference strains (such as A/New Caledonia/20/99) at the 36-, 48-, and 72-h time points and for A/Solomon Islands/3/06 at the 24-, 36-, and 48-h time points (Figure 1A). On the other hand, there was no significant difference in viral titers at all serial time points (12–72 h) between the A/Brisbane/59/2007-likeWT strain and its H274Y mutant variant (Figure 1A). In sharp contrast, the recombinant A/WSN/33 H274Y mutant virus had impaired viral kinetics, with significantly lower viral titers at the 12–24-h time points (P=.05) than the recombinant A/ WSN/33 WT virus (Figure 1B). The two A/Brisbane/59/2007-like viruses (WT and H274Y mutant) generated plaques of similar size after 2 days of growth in ST6GalI-expressing MDCK cells (Figure 2A). On the other hand, the recombinant A/WSN/33 H274Y mutant virus exhibited plaques that were significantly smaller than those of the recombinant WT virus (Figure 2B). Similarly, the clinical A/Mississippi/3/01 isolate with the H274Y mutation exhibited plaques of lower size than those of the corresponding WT virus (Figure 2C).
In vitro replicative capacities of wild-type (WT) and H274Y neuraminidase mutant influenza A(H1N1) viruses. A, Viral titers determined at the indicated time points from supernatants of ST6GalI-expressing Madin-Darby canine kidney (MDCK) cells infected with influenza A/New Caledonia/2001/99, A/Solomon Islands/3/06, A/Brisbane/59/ 2007-like WT virus, and the A/Brisbane/59/2007-like H274Y variant at a multiplicity of infection (MOI) of 0.001. B, Viral titers determined at the indicated time points from supernatants of MDCK cells infected with the recombinant influenza A/WSN/33 WT virus and its H274Y variant at a MOI of 0.001. For both panels, viral titers were determined using standard plaque assays; data are presented as mean values for triplicate experiments with standard errors of the mean. *P<.05 and **P<.001 for the comparison between the WT and H274Y viral titers.
Viral plaque characteristics of wild-type (WT) and H274Y neuraminidase mutant influenza A(H1N1) viruses. At least 20 plaques were visualized after 2 days of incubation at 37°C for determination of the mean plaque area. Inserts show an enlarged magnification (×5.33). A, Viral plaque area of A/Brisbane/59/2007-like WT virus and its H274Y variant (mean ± standard error of the mean [SEM], 0.32±0.03 vs 0.44±0.07 mm2; P=.32). B, Viral plaque area of the recombinant influenza A/WSN/33 WT virus and its H274Y variant (mean ± SEM, 1.2±0.12 vs 0.25±0.02 mm2; P<.001). C, Viral plaque area of A/Mississippi/3/01 WT and its H274Y variant (mean ± SEM, 0.3±0.01 vs 0.06±0.03 mm2; P<.001).
Intranasal inoculation of ferrets with either of the 2 A/Brisbane/ 59/2007-like isolates resulted in a pyrexic response between days 2 and 5 after inoculation (Figure 3). The mean area under the curve of subcutaneous temperature over a period of 7 days was significantly higher in animals infected with the H274Y mutant than in ferrets infected with the WT virus (mean ± SEM, 307.8±0.17 vs 305.8±0.46; P=.006). Viral titers in nasal wash samples collected on day 2 after inoculation were significantly higher for the mutant isolate than for WT virus (mean, 9.7×104 vs 1.6×103 PFU/mL; P=.014), whereas similar viral titers were obtained for the 2 isolates between days 3 and 7 after inoculation (Figure 4). Also, the mean percentage of body weight loss on day 3 after infection was significantly higher in animals infected with the H274Y mutant than with the WT virus (4% vs 1%; P=.004).
Body temperature of ferrets infected with wild-type (WT) and H274Y neuraminidase mutant influenza A(H1N1) viruses. Body temperatures were recorded by implanted thermometers for 7 days after inoculation in groups of 4 ferrets infected with the influenza A/Brisbane/59/2007-like WT isolate (A) and its H274Y variant (B). Each line shows data for 1 ferret.
Viral titers in nasal wash samples from ferrets infected with wild-type (WT) and H274Y neuraminidase mutant influenza A(H1N1) viruses. Viral titers were determined on the indicated days after inoculation from nasal washes of ferrets infected with the influenza A/Brisbane/59/ 2007-like WT virus and its H274Y variant. Viral titers were determined using standard plaque assays; data are presented as mean values for triplicate experiments with standard errors of the mean. *P<.05 for the comparison between the WT and H274Y viral titers.
The development of drug resistance is a key factor in determining the long-term usefulness of antiviral therapy. The clinical utility of the first class of anti-influenza agents, the adamantanes (amantadine and rimantadine), has been dramatically decreased because of high levels (>90%) of natural resistance observed among recent epidemic A(H3N2) viruses as well as among clade 1 A(H5N1) avian strains from Southeast Asia resulting from the S31N mutation in the M2 gene [14, 15]. Natural resistance to oseltamivir, the most frequently used NAI, is also on the rise in recent A(H1N1) viruses, specifically in A/Brisbane/59/2007-like isolates [5]. In fact, the rate of naturally occurring resistance to oseltamivir in epidemic A(H1N1) viruses isolated during the 2008–2009 influenza season is comparable to that of amantadine resistance in A(H3N2) viruses reported after 2005 [14]. Although it has been shown that amantadine-resistant influenza A viruses retain their in vitro and in vivo fitness and transmissibility [16, 17], influenza viruses harboring NA mutations that confer resistance to oseltamivir (including H274Y mutants) have been generally attenuated in animal models [9, 10]. However, A/Brisbane/59/2007(H1N1)-like viruses containing the H274Y mutation have been efficiently transmitted between humans in the absence of antiviral pressure and were shown to cause typical flulike illnesses [6]. Differences in viral sequence backgrounds and changes in the HA/NA balance could explain such discrepant results.
In the present study, we have shown that a recent A/Brisbane/ 59/2007-like virus isolated in Quebec, Canada, in 2008 carried the H274Y mutation and had at least comparable (if not increased) viral fitness both in vitro and in ferrets relative to that of a closely related WT strain that was also isolated in Quebec in 2008. In sharp contrast, the same mutation present in different viral sequence backgrounds (ie, the recombinant A/WSN/33 and the clinical A/Mississippi/3/01 viruses) resulted in reduced replicative capacity in vitro, compared with their corresponding WT counterparts. Moreover, the pyrexic response and loss of body weight induced in ferrets by the A/Brisbane/59/2007 H274Y mutant were more accentuated than those of the related WT virus, in line with recent information suggesting that the new A/Brisbane/59/2007-like H274Y mutant viruses are pathogenic in humans in the absence of treatment [6, 7, 18] and had in fact replaced the related A/Brisbane/59/ 2007-like WT viruses in 2008–2009.
The 2 A/Brisbane/59/2007-like isolates (WT and H274Y mutant) are genetically closely related on the basis of surface gene sequences. There is only 1 HA change—G189V (H3 numbering), which is not part of the receptor-binding site—in the resistant mutant compared with the WT virus. Recent oseltamivir-resistant and oseltamivir-susceptible A(H1N1) viruses are also antigenically related on the basis of HA inhibition assays and belong to clade 2B, represented by A/Brisbane/59/2007 [6, 19]. As for the NA gene, apart from the H274Y resistance mutation there are 3 other changes (I52V, S153N, and D357G; N2 numbering) between the 2 viruses. The I52V and S153N mutations have not been found in other oseltamivir-resistant variants, in contrast to the D357G mutation, which has been detected in most but not all recent H274Y variants worldwide [19]. Thus, it is unlikely that these additional NA mutations could exert a compensatory effect for the substitution at codon 274, which is a framework residue of the NA enzyme that interacts with E276 in the active site. Rameix-Welti et al [19] have demonstrated that recent susceptible A/Brisbane/59/2007-like viruses had increased NA activity compared with older A(H1N1) viruses (such as A/New Caledonia/2001/99 and A/Solomon Islands/03/06) because of a combination of specific amino acids in the vicinity of the NA substrate binding site that emerged in 2007. It can thus be hypothesized that the H274Y mutation somewhat reduced the affinity for the substrate compared with the WT virus and that such mutant viruses may in fact have a more appropriate balance of HA/NA activities, which would explain their better dissemination. A more appropriate HA/NA balance could also explain the better replicative capacities of older A(H1N1) strains (such as A/NewCaledonia/20/99 and A/Solomon Islands/03/06) over both the WT and the mutant A/Brisbane/59/07 strain, as suggested by our results (Figure 1A)—although we cannot exclude the possibility that the reference strains grew more efficiently because of a greater number of cell passages. The H274Y mutation may have altered the HA/NA balance in the older genetic background of A/Mississippi/3/01 (a A/New Caledonia/2001/99-like isolate), resulting in impaired viral fitness that prevented the wider dissemination of the A/Mississippi/3/01 H274Y variant. In addition, transmission and replication may also be modulated by mutations in other internal viral genes.
As expected, the A/Brisbane/59/2007-like H274Y mutant virus showed cross-resistance to oseltamivir and peramivir, whereas this variant remained susceptible to zanamivir and A-375315. Similar phenotypes of resistance were obtained for the A/Mississippi/3/01 and the A/WSN/33 H274Y variants (Table 1). Of note, the level of resistance to peramivir was higher for the A/Brisbane/59/2007-like and A/Mississippi/3/01 H274Y mutants than for the A/WSN/33 H274Y mutant. Conversely, there is an interesting potential for inhaled zanamivir and the investigational orally available NAI agent A-322278 (the prodrug of A-315675 [20]) for the treatment of infection with oseltamivir-resistant A(H1N1) and potentially A(H5N1) viruses harboring theH274Y mutation.
Susceptibility Profiles of Different Influenza A(H1N1) H274Y Mutants with Respect to Neuraminidase Inhibitors
In conclusion, the H274Y mutation confers a high level of resistance to oseltamivir without any apparent detrimental cost in terms of viral fitness and virulence in recent A(H1N1) viruses. Our in vitro and in vivo results are in agreement with recent epidemiological data showing that, during the 2007–2008 influenza season in Norway and elsewhere, infections with A(H1N1) strains harboring the H274Y NA mutation were associated with similar viral shedding, primary symptoms, and influenza complications compared with infections with oseltamivir-susceptible A(H1N1) viruses [6, 7, 18]. The selection process that led to the emergence of the A/Brisbane/59/2007(H1N1)-like virus strains remains unknown, and thus future sequence evolution is difficult to predict. Overall, such data highlight the need for new and improved treatment strategies for influenza virus infections, including drug combinations and identification of new antiviral targets.
Potential conflicts of interest: Partial study support was provided to G.B. from Hoffmann-La Roche. Of note, the company did not participate in the design of the study or in the writing of the manuscript.
Financial support: Canadian Institutes of Health Research (grant OPP-86937); Hoffmann-La Roche (grant to G.B.).
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy