Cross-Species Infectivity of H3N8 Influenza Virus in an Experimental Infection in Swine

Abstract

Avian influenza A viruses have gained increasing attention due to their ability to cross the species barrier and cause severe disease in humans and other mammal species as pigs. H3 and particularly H3N8 viruses, are highly adaptive since they are found in multiple avian and mammal hosts. H3N8 viruses have not been isolated yet from humans; however, a recent report showed that equine influenza A viruses (IAVs) can be isolated from pigs, although an established infection has not been observed thus far in this host. To gain insight into the possibility of H3N8 avian IAVs to cross the species barrier into pigs, in vitro experiments and an experimental infection in pigs with four H3N8 viruses from different origins (equine, canine, avian, and seal) were performed. As a positive control, an H3N2 swine influenza virus A was used. Although equine and canine viruses hardly replicated in the respiratory systems of pigs, avian and seal viruses replicated substantially and caused detectable lesions in inoculated pigs without previous adaptation. Interestingly, antibodies against hemagglutinin could not be detected after infection by hemagglutination inhibition (HAI) test with avian and seal viruses. This phenomenon was observed not only in pigs but also in mice immunized with the same virus strains. Our data indicated that H3N8 IAVs from wild aquatic birds have the potential to cross the species barrier and establish successful infections in pigs that might spread unnoticed using the HAI test as diagnostic tool.

Importance: Although natural infection of humans with an avian H3N8 influenza A virus has not yet been reported, this influenza A virus subtype has already crossed the species barrier. Therefore, we have examined the potential of H3N8 from canine, equine, avian, and seal origin to productively infect pigs. Our results demonstrated that avian and seal viruses replicated substantially and caused detectable lesions in inoculated pigs without previous adaptation. Surprisingly, we could not detect specific antibodies against hemagglutinin in any H3N8-infected pigs. Therefore, special attention should be focused toward viruses of the H3N8 subtype since they could behave as stealth viruses in pigs.

FIG 1

(A) MDCK viral growth in vitro of swine (Sw), avian (Av), and seal (Se) IAVs. (B) Plaque sizes of Sw, Av, and Se IAVs on MDCK cells.

FIG 2

Percentage of animals with fever (>40°C) after IAV inoculation. Twelve pigs by group were inoculated with avian, seal, equine, and canine origin H3N8 IAVs. Six or twelve animals were mock infected or infected with H3N2 IAVs, respectively. The temperature was recorded daily. The data shown are averages from all animals per group.

FIG 3

Macroscopic pulmonary lesions observed at 3 dpi in pigs inoculated with different IAVs. (A) Control. Note the absence of lesions. (B) Swine (Sw) IAV. There are extended areas of pneumonia (arrows) affecting 50% of both cardiac lobes and a small portion of the right diaphragmatic lobe. (C) Equine (Eq) IAV. Note the absence of lesions. (D) Canine (Ca) IAV. Again, note the absence of lesions. (E) Avian (Av) IAV. There are multifocal areas of pneumonia affecting 25% of both cardiac lobes (arrows). (F) Seal (Se) IAV. There are multifocal areas of pneumonia affecting 25% of both cardiac lobes (arrows).

FIG 4

Immunohistochemical detection of IAV antigen at 3 dpi in the lungs of pigs infected with different IAVs. (A) Negative control. (B) Swine (Sw) IAV. There is positive nuclear staining in a few epithelial and desquamated cells of the bronchioli. (C) Equine (Eq) IAV. Note the negative staining. (D) Canine (Ca) IAV. Note the negative staining. (E) Avian (Av) IAV. There is positive nuclear staining in a few epithelial cells of the bronchioli. (F) Seal (Se) IAV. There is positive nuclear staining in a few epithelial cells of the bronchioli.

FIG 5

Neuraminidase (NA) inhibition by sera from pigs infected with each of the indicated IAVs. The data are means ± the standard deviations (SD) for four sera collected 21 days p.i. from animals inoculated with swine H3N2 (Sw), avian H3N8 (Av), seal H3N8 (Se), equine H3N8 (Eq), and canine H3N8 (Ca) IAVs and two sera from mock-infected pigs.

FIG 6

ELISAs were performed using purified A/PR/8/34 IAVs and sera of mice immunized intraperitoneally with 2.7 × 10⁷ EID₅₀ per mouse of swine H3N2 (Sw), avian H3N8 (Av), and seal H3N8 (Se) IAVs inactivated by UV. (A) ELISAs were performed using sera collected 9 days after the first immunization. (B) ELISA were performed using sera collected 14 days after the second immunization with the same amount of immunogen. The data are presented as the median ± the SD.

FIG 7

Neighbor-joining trees of eight gene segments of viruses examined in the present study (●) and H3N8 IAVs from equine (Eq), canine (Ca), and avian (Av) origins retrieved from GenBank.