Virological Characterization of Duck Avian Influenza H5N2 Viruses Isolated in a Live-Bird Market in Taiwan, 2005-2006

• Main Authors: Yao-Tsun Li, 李曜存
• Other Authors: 金傳春
• Format: Others
• Language: en_US
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/52618914560168012170

碩士 === 國立臺灣大學 === 流行病學與預防醫學研究所 === 101 === Human infection with avian influenza viruses (AIV) has always raised global health concern of emerging novel inter-species transmissible virus that might have potential to cause pandemic. Waterfowls, as the predominant nature reservoir of AIV, harbor precursor genetic lineages contributing to viral evolution on past pandemic strains. Routine virological surveillance at live-bird markets (LBMs) can acquire essential information on viral evolution and host range through AIV isolates. In recent years, several outbreaks of H5N2 in chickens occurred in Taiwan. Additionally, previous study identified 22 duck H5N2 viruses from a LBM with different gene origins in eight segments, but their virological properties remained unclear. Therefore, the specific aims of this study were: (1) to investigate the replication efficiencies of those H5N2 isolates with differ ent gene constellations, (2) to compare the growth properties of H5N2 isolates with different viral phenotypes in both avian and mammalian cells, and (3) to elucidate the possible molecular factors contributing to such phenotypic variations. The four duck H5N2 influenza viruses, DV30, DV192, DV413 and DV518, isolated during 2005-2006 representing distinct gene constellations were selected for their phylogenetic differences in PB2 and NA genes (>5% nucleotide identities). The Madin-Darby Canine Kidney cell (MDCK) and chicken embryo fibroblast cell (DF1) were firstly infected with these viruses to monitor their replication by measuring the levels of viral RNA using real-time polymerase chain reaction (qPCR) targeting viral M gene. Second, viral phenotypes were characterized by the size of plaques in MDCK cells infected with various strains. Besides, the strains possessing different plaque morphologies were further plaque-purified, and 50% tissue culture infectious dose (TCID50) method was performed to evaluate the growth kinetics of viruses in MDCK and DF1 cells. Third, all of the viral strains plaque-purified were sequenced to compare the differences in amino acids. And subsequently, cell binding test with human lung carcinoma cell (A549), MDCK and DF1 cells were investigated. In addition, minigenome assays in 293T and DF1 cells were conducted to clarify possible roles of amino acid changes in viral polymerase complex proteins conferring differences in viral growths. Results from the growth properties of the four duck H5N2 viruses showed similar replication in M gene of vRNA; however, DV518 had relatively higher levels of the same transcripts in both MDCK and DF1 cell lines among all the four isolates. DV413, which had only three amino acids different from DV518, presented significant lower virus yields than DV518, implying possible underlying determinants of viral growth. To further investigate the residues between DV413 and DV518, two strains of DV518 were plaque-purified in MDCK cells, and their plaque morphologies together with that of DV413 in MDCK were characterized. Plaques generated by DV413 were smaller than both 518 L, which had large plaque size, and 518 S. Higher virus titers were observed in MDCK cells infected with 518 L strain, than those infected with either 518 S or DV413. However, both of these two strains of 518 showed similar growth kinetics in DF1 cells, whereas DV413 displayed significant lower replication efficiency. Furthermore, viral yield after in ovo inoculation in SPF embryonated chicken eggs also correlated with the results in DF1 cells. The results of binding assay to assess the receptor specificity of these H5N2 viruses presented no apparent pattern in mammalian and avian cells. However, DV518 isolates, possessing 170D in HA, showed enhanced binding to MDCK cells based on higher levels of attached viral gene (both 518 S v.s. DV413 and 518 L v.s. DV413, p<0.05). In addition, the 518 L, possessing the amino acid substitutions of E72D in PB2 and P224S in PA, demonstrated significant higher polymerase activities by luciferase signals in minigenome assay only in 293T cells but not in DF1 cells than those infected with 518 S virus. These results support that possible molecular determinants may be involved in differences in viral replication in mammalian cells, implying a potential marker associated with host-specific factors. In summary, this study tried to elucidate the relationship between amino acid changes and phenotypic variations among the three Taiwan duck influenza H5N2 viruses isolated in the field. Our data and supports from previous studies provides important insights on molecular determinants that may also be present in the heterogeneous populations of duck low pathogenic avian influenza viruses, which would be capable to adapt and to replicate better in mammalian cells. Future studies need to use reverse genetics to verify the role of each point mutation identified from this study affecting viral entry, replication, and interaction with host.