To analyze the contribution of neuraminidase (NA) toward protection against avian influenza virus (AIV) infection, three different recombinant Newcastle disease viruses (NDVs) expressing hemagglutinin (HA) or NA, or both, of highly pathogenic avian influenza virus (HPAIV) were generated. The lentogenic NDV Clone 30 was used as backbone for the insertion of HA of HPAIV strain A/chicken/Vietnam/P41/05 (H5N1) and NA of HPAIV strain A/duck/Vietnam/TG24-01/05 (H5N1). The HA was inserted between the genes encoding NDV phosphoprotein (P) and matrixprotein (M), and the NA was inserted between the fusion (F) and hemagglutinin-neuraminidase protein (HN) genes, resulting in NDVH5VmPMN1FHN. Two additional recombinants were constructed carrying the HA gene between the NDV P and M genes (NDVH5VmPM) or the NA between F and HN (NDVN1FHN). All recombinants replicated well and stably expressed the HA gene, the NA gene, or both. Chickens immunized with NDVH5VmPMN1FHN or NDVH5VmPM were protected against two different HPAIV H5N1 and also against HPAIV H5N2. In contrast, immunization of chickens with NDVN1FHN induced NDV- and AIV N1-specific antibodies but did not protect the animals against a lethal dose of HPAIV H5N1. Furthermore, expression of AIV N1, in addition to AIV H5 by NDV, did not increase protection against HPAIV H5N1.

ABSTRACT Highly pathogenic avian influenza viruses (HPAIV) differ from all other strains by a polybasic cleavage site in their hemagglutinin. All these HPAIV share the H5 or H7 subtype. In order to investigate whether the acquisition of a polybasic cleavage site by an avirulent avian influenza virus strain with a hemagglutinin other than H5 or H7 is sufficient for immediate transformation into an HPAIV, we adapted the hemagglutinin cleavage site of A/Duck/Ukraine/1/1963 (H3N8) to that of the HPAIV A/Chicken/Italy/8/98 (H5N2), A/Chicken/HongKong/220/97 (H5N1), or A/Chicken/Germany/R28/03 (H7N7) and generated the recombinant wild-type and cleavage site mutants. In contrast to the wild type, multicycle replication of these mutants in tissue culture was demonstrated by positive plaque assays and viral multiplication in the absence of exogenous trypsin. Therefore, in vitro all cleavage site mutants resemble an HPAIV. However, in chicken they did not exhibit high pathogenicity, although they could be reisolated from cloacal swabs to some extent, indicating enhanced replication in vivo. These results demonstrate that beyond the polybasic hemagglutinin cleavage site, the virulence of HPAIV in chicken is based on additional pathogenicity determinants within the hemagglutinin itself or in the other viral proteins. Taken together, these observations support the notion that acquisition of a polybasic hemagglutinin cleavage site by an avirulent strain with a non-H5/H7 subtype is only one among several alterations necessary for evolution into an HPAIV.