Postreassortment changes in influenza A virus hemagglutinin restoring HA-NA functional match.

An important function of influenza virus neuraminidase (NA) is the removal of sialic acid residues from virion components in order to prevent the aggregation of virus particles. In previous communications we have reported that reassortant viruses containing the NA gene of A/USSR/90/77 (H1N1) virus and HA genes of H3, H4, H10, or H13 subtypes had a tendency to virion aggregation at 4 degrees C and that the virion clusters irreversibly dissociated after the treatment with bacterial neuraminidase. It was concluded that in such reassortants the removal of sialic acid residues is inefficient. Nonaggregating variants of the reassortants were selected in the course of serial passages in embryonated chicken eggs. In the present paper a reassortant virus, R2, having the HA gene of A/Duck/Ukraine/1/63 (H3N8) virus and the other genes of A/USSR/90/77 (H1N1) virus, as well as its non-aggregating passage variants and both parent viruses, have been studied in order to reveal the presence of unremoved sialic acid residues in the virions. An assay of sialic acid content by high-performance liquid chromatography with fluorescent detection has revealed the presence of sialic acid in the purified virus preparations of A/USSR/90/77 (H1N1) virus and the R2 reassortant and its nonaggregating variants, whereas only trace amounts of sialic acid have been detected in the A/Duck/Ukraine/1/63 (H3N8) parent virus. The data obtained with the use of the labeled "indicator" virus suggest that the unremoved sialic acid residues are present at the virion surface. The nonaggregating variants have been shown to possess a lower affinity toward high-molecular-weight sialic acid-containing substrates compared to the initial reassortant R2. Sequencing of HA genes has revealed amino acid changes in the nonaggregating variants compared to the initial reassortant. One substitution, N248D in HA1, is the same in two independently selected nonaggregating variants. The presented data suggest that the complete removal of sialic acid residues by viral NA from the virion components is not obligatory for the absence of virus particle aggregation: the latter may be achieved (in the reassortants and, presumably, in the wild-type virus) through a balance between the degree of HA affinity toward the sialic acid-containing receptors and the extent of the removal of sialic acid residues by NA.