Outbreak Histories

From May through September 2007, cases of severe respiratory disease occurred in animals at 3 veterinary clinics located 10–30 km apart in Kyunggi Province and 1 kennel located in Jeolla Province (southern part of South Korea). The first case, which occurred in May, was identified in a 5-year-old miniature schnauzer that had nasal discharge for 3 days and sneezing for 2 days, after which the signs subsided and the dog recovered. In August, another case was identified in a 3-year-old cocker spaniel that had fever, cough, nasal discharge, and anorexia and died after the onset of clinical signs. In September, severe respiratory disease was identified in 2 Jindo dogs (a Korean breed of hunting dog that originated on Jindo Island) and a 3-year-old Yorkshire terrier. These animals had severe cough, fever, and nasal discharge and died 2 days after visiting the same animal hospital. Finally, an outbreak of canine influenza occurred in an animal clinic in which all 13 dogs housed in a shelter facility were found to be infected with the same virus; their clinical signs were nasal discharge, cough, and high fever. Of the dogs in the affected kennel in Jeolla Province, paired serum samples showed that 47 (90%) of 52 were seropositive for canine influenza virus (H3N2) at the first sampling and that 100% were seropositive by the second sampling.

Nasal swabs from the miniature schnauzer, cocker spaniel, and Yorkshire terrier were submitted to Animal Genetics, Inc. (Suwon, South Korea) for reverse transcription–PCR (RT-PCR) and testing with a commercial rapid influenza virus antigen detection kit (Animal Genetics, Inc.). Hemagglutinin inhibition (HI) tests were performed according to the World Organization for Animal Health recommendations; commercial nucleocapsid protein (NP)–based ELISA (Animal Genetics, Inc.) was used for serologic testing.

RT-PCR and Sequencing

Nasal swabs from the above-mentioned 3 dogs were also used to isolate the influenza A virus by inoculation into 11-day-old chicken eggs. After 3–4 days of incubation, allantoic fluids were clarified by low speed centrifugation, and these fluids were shown to agglutinate chicken erythrocytes. Virus RNA was extracted from allantoic fluids by using Trizol LS (Molecular Research Center, Inc., Cincinnati, OH, USA) according to the manufacturer’s instructions. RT-PCR was performed under standard conditions with random hexamer primers. Isolated influenza virus was subtyped by RT-PCR analysis by using primers specific for canine, swine, and avian hemagglutinin 3 (H3) genes. Primers for the detection of viral genes H3, neuraminidase 2 (N2), polymerase basic protein (PB) 1, PB2, polymerase acidic protein (PA), NP, matrix protein (M), and nonstructural protein (NS) were designed by using the Primer 3 program with modifications (Whitehead Institute, Massachusetts Institute for Technology Center for Genome Research, Boston, MA, USA).

For PCR, pairs of primers were used to detect target genes. cDNA (2 μL) was mixed with a reaction mixture containing 2.5 μL of 10× Taq DNA polymerase buffer, 1.5 mmol/L MgCl₂, 2.0 μL of dNTPs (2.5 mmol/L/μL), 1 μL of each specific primer (10 pmol/L each), and 1 μL of Taq DNA polymerase (Promega, Madison, WI, USA). Distilled water was added to make a final volume of 25 μL. PCR was performed by reaction initiation at 94°C for 10 min, amplification for 32 cycles at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s, and by final extension at 72°C for 10 min. The reaction was held at 4°C until further use. PCR products were analyzed by electrophoresis in 1.5% agarose gel containing ethidium bromide. Sequences of the isolated virus genes were edited and analyzed by using Bioedit software (www.mbio.ncsu.edu/BioEdit/bioedit.html). Phylogenetic trees were generated by using the MEGALIGN program (DNASTAR, Madison, WI, USA) with the ClustalX alignment algorithm (www.megasoftware.net).

Experimental Infection with Isolated Virus

We experimentally reproduced viral infection in 10-week-old conventional beagle puppies that had been divided into inoculated (I) and noninoculated (NI) groups. Group I puppies (n = 9) were inoculated intranasally with 2 mL of virus isolate with a titer of 10^6.9 50% egg infectious dose (EID₅₀)/0.1 mL; group NI puppies (n = 6) were inoculated intranasally with 2 mL of sterile phosphate buffered saline. Before they were inoculated, the animals were sedated by intramuscular injection of 0.1 mg/kg acepromazine malate (Bayer, Seoul, South Korea). Clinical signs of infection were monitored for 7 days after inoculation, and feces and nasal discharge were examined for virus shedding by RT-PCR for 10 days after inoculation. To detect antibodies against nucleoprotein and HI for hemagglutinin, we analyzed convalescent-phase serum samples from 3 puppies in each group for virus-specific antibodies by ELISA (Animal Genetics, Inc.). HI tests were performed according to World Organization for Animal Health–recommended procedures (9). At 3, 6, and 9 days postinoculation (dpi), 3 group I puppies and 2 group NI puppies were humanely euthanized for gross and histopathologic examination. All necropsy procedures were performed by veterinary pathologists. All organs from dogs and pigs (positive control) were rapidly immersed in 10% neutral formalin buffer to prevent autolysis and stored overnight. To detect influenza A virus antigens in group I or group NI tissues, we performed immunohistochemical examination by using goat anti–influenza A virus antibody (1:100; Chemicon, Temecula, CA, USA). To determine the presence or absence of SA α2,3-gal linkages comprising avian influenza virus receptors and SA α2,6-gal linkages comprising human influenza receptors in the respiratory tracts of noninfected puppies, lectin-based staining was performed as previously reported (10). Porcine tissue served as a positive control. All experimental procedures were approved by an independent animal care and use committee, and the guidelines of National Veterinary Research and Quarantine Service for the reproduction of pathogenesis in dogs were respected.

Isolation of Virus

Nasal swabs from the miniature schnauzer, cocker spaniel, and Yorkshire terrier were positive for influenza virus and negative for other pathogens, including canine distemper virus, canine parainfluenza-2 virus, and Bordetella bronchiseptica. The isolated viruses were designated A/canine/Korea/01/2007 (H3N2), A/canine/Korea/02/2007 (H3N2), and A/canine/Korea/03/2007 (H3N2).

Nucleotide Sequences

Eight gene segments (H3, N2, PB1, PB2, PA, NP, M and NS) of each isolated canine influenza virus were sequenced (EU127500, H gene; EU127501, N gene), and homologous sequences were sought in GenBank (Table). Sequences from avian influenza viruses that displayed homologies from 95.5% to 98.9% were identified for all 8 gene segments from 1 of the 3 subtype H3N2 canine isolates (A/canine/Korea/01/2007). The HA and NA genes of this isolate showed greatest identity with those of Korean avian influenza virus isolate S11, and the NS gene showed greatest identity to that of avian influenza virus (A/chicken/Nanchang/7-010/2000 [H3N6]) isolated from Chinese chickens. All the other genes, including PB1, PB2, PA, NP and M, were closely related to those of avian influenza virus isolated from ducks in Hong Kong, Japan, and China.

Phylogenetic Relationships

Figure 1

Figure 1. Phylogenetic relationship among hemagglutinin genes of canine influenza virus isolates. Tree of hemagglutinin genes from representative isolates from dog, human, bird, pig, and horse. Scale bar represents a difference of 5%....

Phylogenetic analysis indicated that the canine influenza virus isolates from South Korea belonged to a different cluster than those of equine and canine influenza subtype H3N8 viruses. The HA and NA genes of the canine isolate (A/canine/Korea/01/2007 [H3N2]) were closely related to those of avian influenza virus (H3N2) from South Korea (Figure 1).

Serologic Responses to Inoculation

All group I puppies had negative serologic assay results before inoculation. Group NI control puppies remained negative throughout the experiment.

Figure 2

Figure 2. Body temperature, virus shedding, and antibody seroconversion after challenge with canine influenza virus. Body temperature was increased from 1 day postinoculation (dpi) and slowly decreased to normal temperature by 7 dpi....

In nucleoprotein-specific ELISA, the percent inhibition values for group I at 6 dpi were substantially higher than those for group NI (Figure 2); and the HI antibody titers of group I (HI titer 80) were induced at 8 dpi.

Clinical Responses to Challenge

Clinical signs, including sneezing and nasal discharge in group I, were observed at 2–7 dpi. The rectal temperatures of group NI animals remained below 39°C throughout the experiment. At 24 h after inoculation, fever developed in group I puppies (mean rectal temperature 40.14°C) (Figure 2) and lasted through 6 dpi.

Virus Shedding

Influenza virus was not detected in feces. However, for group I puppies, virus shedding in nasal discharge began at 1 dpi and continued to 6 dpi; the highest titers, 10^6.1(EID₅₀/0.1 mL), were reached by 4 dpi. RT-PCR products generated from shed viruses were sequenced and identified as identical to the inoculated virus.

Histopathologic Findings

Figure 3

Figure 3. Histopathologic lesions in the trachea and lungs of control (A and C) or experimentally infected (B, D–F) dogs (A/canine/Korea/01/2007 [H3N2]) at different days postinoculation (dpi). A) Control dog at 9 dpi,...

Gross lesions were limited to the lungs and were characterized by multifocal to coalescing reddish consolidation. In tissues collected on 3, 6 and 9 dpi, histopathologic lesions were observed in the trachea and lungs, and extrapulmonary lesions were absent in puppies infected with the isolate (A/canine/Korea/01/2007 [H3N2]). Severe virus-induced necrosis and inflammation of the upper (trachea and bronchi) and lower (bronchiole and alveoli) respiratory tracts were noted on histologic examination. Although minor differences in the severity of the histologic findings were observed among the 9 infected dogs, all infected dogs shared the following histopathologic features regardless how long after inoculation tissues were collected: 1) moderate to severe multilobular or diffuse necrotizing tracheobronchitis with suppurative inflammation in the lumina and squamous metaplasia of the tracheobronchial epithelium (Figure 3, panel B); 2) moderate to severe multilobular or diffuse necrotizing bronchiolitis and alveolitis (i.e., bronchioalveolitis, occasionally accompanied by chronic peribronchiolar and perivascular inflammation) (Figure 3, panels D and E); and 3) mild to moderate multilobular or diffuse thickening of alveolar septa by infiltrates of inflammatory cells, such as interstitial pulmonary macrophages. At 3, 6, and 9 dpi, large amounts of influenza A virus antigen were found in bronchial and bronchiolar epithelium and lumens (Figure 3, panel F).

Receptor Binding Assay

Figure 4

Figure 4. Lectin staining (red stain) for SAα 2,3-gal (avian influenza virus receptors) and SAα 2,6-gal (human influenza virus receptors) in canine trachea, bronchus, and bronchioles, together with porcine tissues as a positive...

Consistent with the histologic lung lesions, large amounts of SAα 2,3-gal were found on the surface of bronchial and bronchiolar epithelial cells of group NI puppies and were rarely found on tracheal epithelial cells (Figure 4). In contrast, SAα 2,6-gal was not detected in tracheal, bronchial, or bronchiolar epithelial cells, which suggests that canine species may have a lesser role as intermediate hosts for transmission of human influenza viruses to dogs than for avian influenza viruses.