Influenza virus surface glycoprotein neuraminidase (NA) is the primary target of influenza for structurebased drug design (von Itzstein M, et al., 1993). To date, the most effective anti-influenza NA drugs (NAIs) include oseltamivir (Tamiflu) and zanamivir (Relenza) that are specific for influenza A and B viruses (Gubareva L V, et al., 1995; Hata K, et al., 2008; Ryan D M, et al., 1995; Woods J M, et al., 1993). However, virus variants associated with drug-resistance have arisen from treatment with current NAIs, which include drugresistant seasonal H1N1 and H3N2 human influenza viruses as well as drug-resistant highly pathogenic avian influenza (HPAI) viruses of the H5N1 subtype. Emergence of drug-resistant influenza viruses undermines our current drug therapies.
Oseltamivir treatment has led to influenza A virus drug-resistant variants within N1, N2 and N9 subtypes as well as influenza B virus variants. Clinical human influenza A/H3N2 and A/H1N1 isolates recovered from oseltamivir-treated individuals contain R292K, E119V and N294S substitutions in NA N2 and H274Y and N294S substitutions in NA N1. Furthermore, studies of drug-treated clinical H5N1 isolates and recombinant H5N1 viruses have demonstrated reduced sensitivity to oseltamivir (de Jong M D, et al., 2005; Le Q M, et al., 2005; Yen H L, et al., 2007). In this report, we have adapted a cell-based assay to investigate NA susceptibility to oseltamivir. Several catalytic residue substitutions have been implicated in oseltamivir resistance and we have found that a handful of these residues display varying phenotypes in an NA N1 background shedding light into subtype specific effects on particular mutations of the enzyme active site associated with altered NAI susceptibility. Using a lentiviral-based pseudotyping system, we have demonstrated inhibition of HIV/HA pseudoviral particles from producer cells in the presence of oseltamivir permitting a resourceful screen to assess altered NAI susceptibility of various NA mutants. The utility of this system may provide a platform to screen novel therapeutics in influenza viral infection.
Resistance to neuraminidase inhibitors (NAIs) continues to increase as the influenza virus acquires new mutations, and a quick screening system is crucial to assess possible pandemic threatening mutations that confer resistance to globally stockpiled antivirals. To meet this goal, we have employed a lentivirus-based pseudotyping system to study the effects of various substitutions in NA on oseltamivir sensitivity.
Lentiviral-based reporter viruses pseudotyped with an avian influenza hemagglutinin (HA) glycoprotein were generated in human 293T producer cells by transient co-transfection of an HA-expression plasmid along with an env-deficient HIV-1 vector (pNL4-3-Luc-R--E-) carrying a luciferase reporter gene. Susceptible target cells were challenged with the producer cell culture supernatants collected 48 h post-transfection and transduction was determined by luciferase activity in the target cells. Inclusion of an NA gene from a mouse-adapted human virus strain (PR8), heretofore known as NAH, resulted in efficient transduction, 1.7 × 107 RLU (Fig. 1A).
Figure 1. Inhibition of parental NA activity by oseltamivir. HIV/HA pseudovirions were produced by 293T cells in the absence or presence of oseltamivir and release was measured through ability of producer cell supernatants to (A) infect 293T target cells as measured via luciferase activity in the target cell lysates. Data shown is representative data from several experiments, each done in triplicate. Background luciferase levels were measured from lysates of target cells transduced with supernatant from producer cells lacking NA in their transfection. B: As another measure of pseudovirion release, a hemagglutination assay was done using equal volumes of chicken red blood cells and purified pseudovirus collected from producer cells. Virus was serial diluted out until 1:128. PBS was used as a negative control. C: NA enzymatic activity was directly measured in the absence or presence of oseltamivir by ability of producer cell lysate to cleave 2'-(4-methylumbelliferyl)-N-acetylneuarminic acid substrate and measured. Lysates of untransfected cells were used as a control. Error bars represent standard deviation. Data shown is representative data from several experiments, each done in triplicate.
HA/NAH-mediated luciferase levels exhibit a dosedependent decrease with increasing concentrations of oseltamivir carboxylate (OC) added to the culture medium of 293T producer cells during HIV/HA pseudovirion production (Fig. 1A). In the absence of inhibitor, resulting luciferase activity was more than 1 × 107 RLUs for 293T cells. A marginal decrease in luciferase levels was observed in the presence of 1 nmol/L OC. Luciferase levels started to decrease more dramatically in the presence of 10 nmol/L or higher OC compared to no drug. In the presence of 100 nmol/ LOC, luciferase activity dramatically decreased more than 100-fold, and the levels further decreased in the presence of 1 μmol/L OC comparable to background levels in this system (Fig. 1A). Thus it is clear that presence of OC in the producer 293T cells reduced the transduction ability of the HIV pseudovirions, consistent with the notion that OC inhibits viral particle release of HIV/HA pseudovirions.
To directly measure viral particle release, the HIV/HA pseudovirions in the presence or absence of OC were collected, concentrated and used in a hemagglutination assay. In the absence of OC, even at 1:64 viral dilution, hemagglutination activity was observed (Fig. 1B, top row). In contrast, in the presence of 1 000 nmol/L of OC, no hemagglutination activity was observed even without any viral dilution (Fig. 1B, bottom row). Furthermore, a reverse OC dose-dependent hemagglutination activity was observed (Fig. 1B, comparing rows with 1 nmol/L to 1 000 nmol/L). These results demonstrate that OC directly inhibited viral particle release of HIV/HA pseudovirions from the 293T cells.
The NA enzymatic activity and its inhibition by OC were assayed by the ability of NA to cleave a 2'-(4-methylumbelliferyl)-N-acetylneuraminic acid substrate (4-MUNANA). Cell lysates from the 293T producer cells untreated with OC were cleared and incubated with 4-MUNANA substrate and either with or without OC, and cleavage was measured fluorometrically. Without OC, the enzymatic activity of NA was approximately 30 000 arbitrary units (a.u.), and the level dropped nearly 50% with the 1nM of OC (Fig. 1C). At the higher concentrations (100 nmol/L or 1 000 nmol/L), the enzymatic activity of NA was completely abolished. Therefore, together with the aforementioned results, these results demonstrate that the HIV-based pseudotyping assay system can be used as a sensitive surrogate assay system for NA to oseltamivir.
The most commonly reported NA mutation that confers oseltamivir resistance in subtype N1 is H274 substitution, which is a structural residue in the active site (Fig. 2). To further validate the HIV-based pseudotyping system as a surrogate assay for NA resistance to OC, first we used site-directed mutagenesis to incorporate the H274Y substitution in the parental NA and performed the same assays as described above. Mutant H274Y mutant behaved like the parental NA at 0, 1, and 10nM concentrations of OC; in contrast, at 100 nmol/L and 1 000 nmol/L of OC, the HIV/HA pseudovirions with mutant H274Y gave at least 100-fold more transduction (measured by luciferase levels) compared with the parental NA (Fig. 3A), consistent with the notion that H274Y is resistant to OC.
Figure 2. NAH catalytic residues within the active site. A: NA amino acid sequence. Targeted residues E119, I222, D198, H274, and N294 are colored blue. Additional catalytic residues are shaded black. B: Left panel, ribbon diagram of NA active site with mutation sites labeled blue and catalytic residues R118 and R371 colored red. Right panel, ribbon and surface representation of NA active site. Oseltamivir carboxylate is displayed in yellow.
Figure 3. NA substitution H274Y confers resistance to oseltamivir carboxylate inhibition. A: Transduction of HIV/HA pseudovirions expressing NA H274Y. Relative infectivity is determined by luciferase activity (relative light units RLUs) from infected 293T target cell lysates. Values depict the average of triplicate samples. B: Hemagglutination activity of HIV/HA pseudovirions expressing NA H274Y Cell culture supernatants were harvested 48 h post-transfection. Mixtures of concentrated pseudovirions and chicken red blood cells were incubated at 4 ℃ and HA titers recorded 1 h later. PBS served as a negative control. C: Analysis of NA H274Y enzymatic activity measured by the release of 4-MUNANA.
Next, a hemagglutination assay was performed to evaluate how mutant H274Y behaved in the presence of OC (Fig. 3B). It is clear that mutant H274Y was very resistant to OC in its ability to release the HIV/HA virions in stark contrast to the parental NA (comparing Fig. 3B with Fig. 1B). For mutant H274Y, even at the OC concentration of 1, 000 nM, hemagglutination was observed when the viruses were diluted at 1:4 (the bottom row of Fig. 3B). These results demonstrate that mutant H274Y can function more efficiently than the parental NA in mediating HIV/HA particle release.
Interestingly, the enzymatic activity of mutant H274Y, as measured by substrate cleavage, was notably lower (about one third of the parental NA) than the parental NA in the absence of OC (Fig. 3C, and Table 1). Also it was observed that mutant H274Y displayed resistance to OC since its enzymatic activity did not decrease in a dose-dependent manner like the parental NA (Fig. 3C). These results are totally consistent with the behavior of mutant H274Y in response to oseltamivir reported by others.
NA Construct Oseltamivir
Parental 0 1.8(±0.01)×107 32 3.1(±0.2)×104 1 1.8(±0.06)×107 4 1.4(±0.1)×104 10 1.1(±0.2)×107 2 5.6(±0.4)×103 100 1.6(±0.09)×105 0 1.0(±0.04)×103 1000 2.2(±1.3)×103 0 4.9(±0.3)×102 H274Y 0 1.8(±0.09)×107 8 1.0(±0.1)×104 1 1.6(±0.05)×107 16 6.4(±0.4)×103 10 1.5(±0.1)×107 16 6.0(±0.4)×103 100 1.8(±0.02)×107 8 5.3(±0.1)×103 1000 1.1(±0.3)×106 4 1.9(±0.2)×103 N294S 0 1.8(±0.05)×107 4 1.1(±0.06)×104 1 1.8(±0.03)×107 4 1.2(±0.07)×104 10 1.8(±0.03)×107 4 1.1(±0.1)×104 100 4.0(±1.2)×106 4 5.7(±0.2)×103 1000 2.2(±0.7)×105 0 1.3(±0.06)×103 D198N 0 5.7(±2.1)×105 2 7.0(±0.5)×102 1 2.6(±0.4)×105 2 6.6(±0.2)×102 10 4.0(±1.3)×104 0 5.1(±0.2)×102 100 1.6(±0.2)×104 0 3.3(±0.2)×102 1000 1.1(±0.3)×104 0 2.5(±0.08)×102 I222L 0 1.5(±0.2)×107 8 5.1(±0.4)×103 1 1.5(±0.2)×107 8 4.4(±0.2)×103 10 1.1(±0.1)×107 2 3.6(±0.2)×103 100 1.4(±0.3)×105 0 1.1(±0.2)×103 1000 4.1(±1.5)×104 0 3.1(±0.08)×102 E119V 0 2.5(±1.8)×104 0 5.7(±0.7)×102 1 3.3(±2.9)×104 0 4.0(±0.2)×102 10 7.6(±4.3)×104 0 4.3(±0.1)×102 100 2.0(±2.6)×104 0 3.5(±0.3)×102 1000 7.4(±0.3)×103 0 2.7(±0.5)×102 R292K 0 4.9(±1.4)×103 0 6.4(±0.3)×102 1 1.1(±0.1)×104 0 4.1(±0.2)×102 10 8.1(±3.8)×103 0 4.3(±0.5)×102 100 4.7(±0.53)×103 0 4.1(±0.7)×102 1000 1.4(±0.32)×104 0 4.6(±0.1)×102
Table 1. Summary of Characteristics of NA mutants
Therefore, we conclude that the HIV-based pseudotyping assay is a reliable surrogate system to study oseltamivir resistance of influenza viruses.
In addition to mutant H274Y of NA, other NA mutants have previously been reported to show oseltamivir resistance (Fig. 2). N294S has been reported to show resistance in subtype N1; R292K, E119V, and I222L have been reported as showing resistance in subtype N2; D198N has been shown to demonstrate resistance in influenza B. These NA mutants were evaluated in their oseltamivir resistance profiles using the HIV-based pseudotyping assay described above.
R292K and E119V. These substitutions were among the earliest reported as showing oseltamivir resistance to subtype N2 NA. In this study, using a subtype N1 NA, the pseudovirion release as measured by luciferase was equivalent to the background level and the thus oseltamivir had no effect at any concentration (Fig. 4A). There was also no observable hemagglutination activity (Table 1). Furthermore, the neuraminidase activity as assayed in the cleavage assay was equivalent to the background level, and oseltamivir had no effect at any concentration (Fig. 4B). Taken together, these results suggest that these mutations render the subtype N1 NA enzymatically inactive.
Figure 4. NA substitutions displaying varying sensitivities to oseltamivir carboxylate inhibition. A: Transduction potential of HIV/HA pseudovirions expressing NA substitution mutants. Relative infectivity is determined by luciferase activity (relative light units RLUs) from infected 293T target cell lysates. Values depict the average of triplicate samples. Error bars indicate standard deviations. B: Analysis of NA substitution mutant enzymatic activity measured by the release of 4-MUNANA. C: Relative NA enzymatic activities of NA substitutions in response to different concentrations of oseltamivir carboxylate based on the data from (B).
N294S. When this NA mutant was evaluated by the transduction with the HIV/HA pseudovirions produced in the 293T producer cells, mutant N294S behaved like mutant H274Y (Fig. 4A), consistent with the reports that both mutants are resistant to OC in the N1 subtype. The hemagglutination assay showed lower HA titers of mutant N294S in the absence of OC, but these levels were maintained at 100 nmol/L of OC, contrasting to that of the parental NA which displayed no hemagglutination activity at such a OC concentration (Table 1). The enzymatic activity of mutant N294S mimicked that of mutant H274Y (Fig. 4B and 4C). Therefore we conclude that mutant N294S, like mutant H274Y, is resistant to oseltamivir.
D198N. This mutant was found to display an intermediate phenotype when it was evaluated in the HIV/HA pseuodtyping transduction assay without OC treatment, while it showed a dose-dependent sensitivity to OC treatment (Fig. 4A). Very weak or no hemagglutination activity was detected for this mutant (Table 1). The enzymatic activity of this mutant was really low (Fig. 4B and Table 1) compared to the parental NA. These results indicate that D198N substitution renders NA enzymatically inactive.
I222L. This mutant behaved like the parental NA in its OC profile except at 1 000 nmol/L, measured by transduction, suggesting that this mutant is somewhat resistant to OC treatment (Fig. 4A). The HA activity of this mutant was lower compared to the parental NA (Table 1), and the enzymatic activity of this mutant was approximately 1/6 that of the parental NA, and showed some resistance to OC treatment (Fig. 4B and 4C). These results indicate that this mutant is somewhat deficient in the enzymatic activity, and displays a slight oseltamivir-resistant phenotype.