Hemagglutinin specificity and neuraminidase coding capacity of neuraminidase-deficient influenza viruses

Neuraminidase (NA)-deficient mutant virus stocks have been obtained by passaging A/NWS/33(HA)-tern/Australia/G70c/ 75(NA) (H1N9) influenza virus in medium containing neuraminidase from Micromonospora viridifaciens and antiserum against the influenza NA. Growth of the resulting mutants is dependent on addition of bacterial neuraminidase to the medium. Nucleotide sequence analysis showed large single deletions in the NA genes, with both ends of the NA gene segments conserved. These RNA fragments all have the capacity to code for a peptide that contains the N-terminal ''tail'' and membrane-anchoring region of the NA, but the presence of this peptide has not been demonstrated in virions or infected cells. In contrast to the ease of selection of NA-deficient mutants from the H1N9 virus, no mutants were selected from three other viruses. The HA-coding segments of parental H1N9 and mutant NWSc-Mvi predict a change of Pro to His at residue 227 (H3 numbering), close to the receptor-binding site of H3 HA, compared to the HA of an H1N2 reassortant that contains the NWS/33 HA gene. This change may contribute to an altered HA specificity that allows selection of mutants that can infect cells in the presence of high levels of NA activity. It appears that the role of NA in influenza infection is to remove sialic acid from the HA rather than to destroy receptors on cells. (C) 1997 Academic Press.