The recovery of recombinant influenza A virus entirely from cDNA was recently described (9, 19). We adapted the technique for engineering influenza B virus and generated a mutant bearing an amino acid change E116G in the viral neuraminidase which was resistant in vitro to the neuraminidase inhibitor zanamivir. The method also facilitates rapid isolation of single-gene reassortants suitable as vaccine seeds and will aid further investigations of unique features of influenza B virus.

Zanamivir is a highly selective neuraminidase (NA) inhibitor with demonstrated clinical efficacy against influenza A and B virus infections. In phase II clinical efficacy trials (NAIB2005 and NAIB2008), virological substudies showed mean reductions in virus shedding after 24 h of treatment of 1.5 to 2.0 log 10 50% tissue culture infective doses compared to a placebo, with no reemergence of virus after the completion of therapy. Paired isolates (n ‫؍‬ 41) obtained before and during therapy with zanamivir demonstrated no shifts in susceptibility to zanamivir when measured by NA assays, although for a few isolates NA activity was too low to evaluate. In plaque reduction assays in MDCK cells, the susceptibility of isolates to zanamivir was extremely variable even at baseline and did not correlate with the speed of resolution of virus shedding. Isolates with apparent limited susceptibility to zanamivir by plaque reduction proved highly susceptible in vivo in the ferret model. Further sequence analysis of paired isolates revealed no changes in the hemagglutinin and NA genes in the majority of isolates. The few changes observed were all natural variants. No amino acid changes that had previously been identified in vitro as being involved with reduced susceptibility to zanamivir were observed. These studies highlighted problems associated with monitoring susceptibility to NA inhibitors in the clinic, in that no reliable cell-based assay is available. At present the NA assay is the best available predictor of susceptibility to NA inhibitors in vivo, as measured in the validated ferret model of infection.

We describe the in vitro selection and characterisation of virus derived from B/Beijing/1/87 passaged in the presence of zanamivir. During zanamivir passage, the phenotype of virus isolates was either drug dependent or drug resistant in plaque reduction assays. The susceptibility of the neuraminidase of the drug-dependent isolates was unchanged from that of the wild-type enzyme. The drug-dependent isolates contained two mutations in the viral haemagglutinin: V90A, close to the proposed secondary sialic acid-binding site, and L240Q, close to the primary sialic acid-binding site. Virus isolates that were drug resistant contained the same mutations in the haemagglutinin but also contained the mutation E116G in the neuraminidase. For the drug-dependent viruses, zanamivir susceptibility could not be measured because plaque numbers increased with increasing drug concentration. The in vitro zanamivir susceptibility of drug-resistant viruses was lower than that of the wild-type virus by a factor of 275- to >2532-fold. Neuraminidase containing the E116G mutation has a 33-fold lower affinity for zanamivir than the wild-type enzyme. The finding that the same haemagglutinin mutations are found in both drug-dependent and drug-resistant viruses confirms that the same changes to the receptor binding function can contribute to both phenotypes. This observation demonstrates the interplay between the influenza virus haemagglutinin and neuraminidase in escape from zanamivir inhibition in vitro.