Acquisition of a Polybasic Hemagglutinin Cleavage Site by a Low-Pathogenic Avian Influenza Virus Is Not Sufficient for Immediate Transformation into a Highly Pathogenic Strain

Recombinant viruses.The cloning of all eight viral gene segments of strain A/Duck/Ukraine/1/1963 (H3N8) into the vector pHW2000 (12) has been described previously (29). Modification of the HACS regions was performed by site-directed QuikChange mutagenesis; the primer sequences are available on request. Recombinant viruses were rescued essentially as described previously (10) and propagated in 11-day-old embryonated chicken eggs. After virus rescue, the gene composition and presence of the expected cleavage sites of the mutants were verified by sequencing following reverse transcription-PCR from isolated viral RNA (data not shown).

Plaque assays and growth curves.Plaque assays were performed as described previously (30) either in the presence of 2 μg/ml N-tosyl-l-phenylalanine chloromethyl ketone (TPCK)-treated trypsin (Sigma, Taufkirchen, Germany) or in the absence of any exogenous protease. The volume of inoculum was 333 μl. Growth curves on Madin-Darby canine kidney (MDCK) cells and quail fibrosarcoma (QT05 6) cells in the presence of 2 or 1 μg/ml TPCK-treated trypsin, respectively, and in the absence of any exogenous protease in the supernatant were determined by titration of cell culture supernatants at 0, 8, 48, and 96 h pastinoculation at a multiplicity of infection of 10⁻² by plaque assay on MDCK cells without trypsin in case of the HACS mutants and with trypsin in case of the wild-type A/Duck/Ukraine/1/1963 (H3N8). The reported plaque titers are the averages from two independent experiments.

Western blotting.MDCK cell cultures were infected with virus in the presence of either 2 μg/ml TPCK-treated trypsin or no protease for 16 h in minimal essential medium containing 0.2% bovine serum albumin. Proteins from supernatant were separated on sodium dodecyl sulfate-8% polyacrylamide gels and electrotransferred to nitrocellulose membranes. For detection of HA, a monoclonal mouse antibody to A/Shangdong/9/93 (H3N2) (3HG3, 1:100,000, incubated overnight at room temperature; HyTest Ltd., Turku, Finland) and as a secondary antibody a mouse-specific goat immunoglobulin G Fab fragment conjugated with horseradish peroxidase (1:20,000, 1 h at room temperature; Dianova, Hamburg, Germany) were used, followed by chemiluminescence (Supersignal West Pico chemiluminescent substrate kit from Pierce).

Animal experiments.Ten 4-week-old White Leghorn specific-pathogen-free chickens were infected oculonasally with each virus. Each bird was observed daily for clinical signs and classified as healthy (0), sick (1) (exhibiting one of the following: respiratory symptoms, depression, diarrhea, cyanosis, edema, or nervous symptoms), severely sick (2) (severe or more than one of the previously mentioned symptoms), or dead (3), as described previously (1). In addition, chickens exhibiting slight depression were given a classification of 0.5. For virus isolation, cloacal swabs were taken and diluted in 1 ml phosphate-buffered saline (PBS) prior to plaque titration.

Generation of recombinant viruses.In order to obtain different HACS mutants of an avirulent avian strain, we replaced the HACS and adjacent amino acid residues of influenza virus A/Duck/Ukraine/1/1963 (H3N8) with those of A/Chicken/ Italy/8/98 (H5N2), A/Chicken/Hong Kong/220/97 (H5N1), or A/Chicken/Germany/R28/03 (H7N7) by site-directed mutagenesis. To address possible structural constraints within the HA of the parent virus, we chose three polybasic HACSs of various lengths and with different types of adjacent amino acids (Table 1). The wild type (DkUkr63-Wt) or the HACS mutants (DkUkr63-It89_HACS, DkUkr63-Hk220_HACS, or DkUkr63-R28_HACS) were rescued by cotransfection of plasmids encoding the wild-type or mutated HA genes together with plasmids encoding the PB2, PB1, PA, NP, NA, M, and NS genes of A/Duck/Ukraine/1/1963 (H3N8).

Introduction of a polybasic cleavage site leads to trypsin-independent replication in vitro similar to that of HPAIV.In order to investigate whether the cleavage site mutants are dependent on trypsin for multicycle replication, plaque assays were performed with DkUkr63-Wt and mutants DkUkr63-It89_HACS, DkUkr63-Hk220_HACS, and DkUkr63-R28_HACS on MDCK cells in the presence and in the absence of trypsin. Whereas DkUkr63-Wt proved to be dependent on trypsin for plaque formation, as expected for a low-pathogenic avian influenza virus, all three HACS mutants formed plaques independent of exogenous trypsin (Fig. 1A). For demonstration of activation cleavage of the HA precursor HA0, Western blot analyses of supernatants from virus-infected MDCK cells were performed. In accordance with the plaque assays, the HA0 of DkUkr63-Wt remained uncleaved in the absence of trypsin, whereas the HA0 precursors of the HACS mutants were cleaved into the HA1 and HA2 fragments to various extents. DkUkr63-It89_HACS displayed efficient HA0 cleavage, whereas in the case of DkUkr63-Hk220_HACS and DkUkr63-R28_HACS, the cleavage was incomplete (Fig. 1B). Moreover, the latter two mutants displayed an additional peptide with a molecular weight slightly higher than that of the HA1 fragment, in particular in the presence of trypsin, suggesting aberrant HA maturation. Analysis of multicycle growth kinetics corresponded to the plaque formation and proteolytic activation pattern of the HA. In contrast to DkUkr63-Wt, which required trypsin for productive replication in both MDCK and QT6 cells, all cleavage site mutants replicated independently of the addition of trypsin (Fig. 2). However, mutant DkUkr63-R28_HACS grew considerably less efficiently in QT6 cells irrespective of the presence or absence of trypsin, correlating with the formation of very small plaques and inefficient proteolytic HA activation (Fig. 1 and 2).

• Open in new tab
• Download powerpoint

FIG. 1.

In vitro replication of DkUkr63-Wt in comparison with the HACS mutants DkUkr63-It98_HACS, DkUkr63-Hk220_HACS, and DkUkr63-R28_HACS. (A) Plaque assays on MDCK cells in the presence and in the absence of trypsin. (B) Western blots of supernatants from infected MDCK cells incubated in the presence (T) and in the absence (−) of trypsin, using a monoclonal antibody specific to an HA of the H3 serotype.

• Open in new tab
• Download powerpoint

FIG. 2.

Multicycle growth curves of DkUkr63-Wt (blue), DkUkr63-It98_HACS (red), DkUkr63-Hk220_HACS (orange), and DkUkr63-R28_HACS (brown) on MDCK and QT6 cells in the presence (diamonds) and in the absence (circles) of trypsin.

Nevertheless, in contrast to the wild type, all three HACS mutants undergo multicycle replication in vitro, and their HA0 is processed in the absence of an exogenous protease. The independence of viral replication from exogenous trypsin is indicative of proteolytic activation by furin (32). These are features attributed only to HPAIV.

Pathogenicity in chicken.To investigate the virulence of the cleavage site mutants in vivo, five groups of 10 chickens each were infected oculonasally with either DkUkr63-It89_HACS (1.7 × 10⁶ PFU/animal), DkUkr63-Hk220_HACS (1.7 × 10⁶ PFU/animal), DkUkr63-R28_HACS (7.2 × 10⁴ PFU/animal), or DkUkr63-Wt (1.7 × 10⁶ PFU/animal) or mock infected by the administration of PBS and were observed for 10 days. (Due to its inefficient growth, DkUkr63-R28_HACS had to be administered at a lower dosage.) All animals inoculated with PBS, DkUkr63-Wt, and DkUkr63-Hk220_HACS exhibited no symptoms. Eight DkUkr63-It89_HACS-infected animals displayed slight transient depression; two animals (no. 3 and 9) further developed central nervous symptoms (Fig. 3). Chicken no. 3 exhibited a head inclination on day 9 after infection, whereas chicken no. 9 showed a slight head inclination on day 6, which decreased during the following days. We then determined the titers of shed virus in cloacal samples from day 5 by plaque assay beginning with undiluted inoculum from four animals of each group, including those with central nervous symptoms. Animals which had been inoculated with PBS, DkUkr63-Wt, or DkUkr63-Hk220_HACS did not shed virus, whereas in the cloacal samples from four animals inoculated with DkUkr63-It89_HACS and from one animal inoculated with DkUkr63-R28_HACS, virus could be detected at a titer of up to 10^3.7 PFU per swab. Remarkably, virus could be reisolated only from animals which already had developed or proceeded to show signs of disease (Table 2). In contrast to trypsin-independent replication in vitro, the insertion of a polybasic cleavage site into the HA of the avirulent DkUkr63-Wt did not cause immediate transformation into an HPAIV; however, cloacal shedding of virus in accordance with clinical symptoms suggests enhanced replication in chicken.

• Open in new tab
• Download powerpoint

FIG. 3.

Virulence in chickens. Survival and disease after oculonasal inoculation with PBS (mock), DkUkr63-Wt, DkUkr63-It98_HACS, DkUkr63-Hk220_HACS, or DkUkr63-R28_HACS. The birds were observed for 10 days for clinical signs and classified as healthy (0), slightly ill (0.5), or ill (1); the daily clinical index (DCI) was calculated from the sum of individual clinical scores from all birds divided by the number of animals per group (10 chickens). Chickens 3 and 9 from the group infected with DkUkr63-It98_HACS developed central nervous symptoms.