Replication of avian, human and swine influenza viruses in porcine respiratory explants and association with sialic acid distribution

Abstract

Background: Throughout the history of human influenza pandemics, pigs have been considered the most likely "mixing vessel" for reassortment between human and avian influenza viruses (AIVs). However, the replication efficiencies of influenza viruses from various hosts, as well as the expression of sialic acid (Sia) receptor variants in the entire porcine respiratory tract have never been studied in detail. Therefore, we established porcine nasal, tracheal, bronchial and lung explants, which cover the entire porcine respiratory tract with maximal similarity to the in vivo situation. Subsequently, we assessed virus yields of three porcine, two human and six AIVs in these explants. Since our results on virus replication were in disagreement with the previously reported presence of putative avian virus receptors in the trachea, we additionally studied the distribution of sialic acid receptors by means of lectin histochemistry. Human (Sia alpha2-6Gal) and avian virus receptors (Sia alpha2-3Gal) were identified with Sambucus Nigra and Maackia amurensis lectins respectively.

Results: Compared to swine and human influenza viruses, replication of the AIVs was limited in all cultures but most strikingly in nasal and tracheal explants. Results of virus titrations were confirmed by quantification of infected cells using immunohistochemistry. By lectin histochemistry we found moderate to abundant expression of the human-like virus receptors in all explant systems but minimal binding of the lectins that identify avian-like receptors, especially in the nasal, tracheal and bronchial epithelium.

Conclusions: The species barrier that restricts the transmission of influenza viruses from one host to another remains preserved in our porcine respiratory explants. Therefore this system offers a valuable alternative to study virus and/or host properties required for adaptation or reassortment of influenza viruses. Our results indicate that, based on the expression of Sia receptors alone, the pig is unlikely to be a more appropriate mixing vessel for influenza viruses than humans. We conclude that too little is known on the exact mechanism and on predisposing factors for reassortment to assess the true role of the pig in the emergence of novel influenza viruses.

Figure 1

Virus yields, expressed as log TCID₅₀/ml, in the supernatant of the explants. Virus titers were determined at 1, 24 and 48 hpi. Each row shows the results per explant system, from NE down to LE. Each column represents the host from which the different virus subtypes were isolated: pigs, humans and birds. Each value is the mean of three experiments, bars show the S.D. NE: nasal explants, TE: tracheal explants, BE: bronchial explants, LE: lung explants

Figure 2

Dose response curves for Sw/Gent/7625/99, Duck/Belgium/06936/05 and Chicken/Belgium/150/99. Three different inoculation doses were applied: 10⁶, 10⁵ and 10⁴ log EID₅₀. Each row represents one explant system, each column one influenza virus. The values are the mean of two experiments, bars show the S.D. NE: nasal explants, TE: tracheal explants, BE: bronchial explants, LE: lung explants

Figure 3

Immunohistochemical analysis of infected cells. Nasal (A, a), tracheal (B, b), bronchial (C, c) and lung (D→F, d→f) explants at 48 hpi inoculated with Swine/Gent/7625/99 (H1N2) (A→F) and Duck/Belgium/06936/06 (H4N6) (a→f) were analyzed. In the nasal (A: black arrow) and tracheal (B: orange arrow) explants, single swine influenza virus positive cells were diffusely spread while no avian influenza virus positive cells were present (a, b). Swine influenza virus positive cells were also found as a continuous line in bronchial epithelium (C), as multiple foci in the bronchioles (D: red arrows, E) and as single alveolar cells (F: green arrows) in lung explants. Avian influenza viral antigen-positive cells were limited to bronchiolar epithelium in lung explants (d: red arrows, e). Symbols underneath the pictures give the results for the semi-quantitative analysis of influenza virus positive cells by IF. -: no virus positive epithelial cells, +/-: single positive cells covering <10% of the epithelium, +: between 11 and 40% of the epithelium is positive, ++: between 41 and 70% of the epithelium is positive, +++: between 71 and 100% of the epithelium is positive.

Figure 4

Tissue binding of Sambucus nigra agglutinin (SNA), Maackia amurensis agglutinin I (MAL-I) and Maackia amurensis agglutinin II (MAL-II) in the different explant systems. SNA binding (first column) was abundant in the epithelium of nasal (NE), tracheal (TE) and bronchial explants (BE) and in the epithelium of bronchioles (Bronch.), but moderate at the level of the alveolae (Alv.). MAL-I binding to epithelial cells was absent to rare in all explants systems (second column). MAL-II binding (third column) was rare in the epithelium of NE, TE and BE. At the level of the bronchioles and the alveolar tissue, it became moderate to abundant (as indicated by the black arrows).

Figure 5

Comparison of binding with digoxigenin-conjugated MAA in paraffin sections (A) and cryosections (B) of the porcine trachea. Only the cryosections showed clear positivity in the glands (black arrows) and the small blood vessels (blue arrows), while paraffin sections were completely negative.

Figure 6

Influence of the conjugation method of MAL-I and -II lectins on the staining intensities in duck small intestines. Biotinylated MAL-I (A) and MAL-II (a) both resulted in epithelial cell binding (black arrows), but MAL-II (a) was additionally staining the goblet cells (red arrow). For both lectins binding was abolished by sialidase treatment of the sections (B, b). Digoxigenin labelled MAL-I (C) and MAL-II (c) failed to bind to the same tissues.