Molecular Epidemiology of Avian Influenza Viruses in A Live Bird Market in Taiwan during 2005-2006

Interface areas between avian species and humans including live-bird markets (LBMs) had been documented their epidemiological importance since the highly pathogenic avian influenza (HPAI) virus H5N1 was isolated from human cases visited LBM in Hong Kong in 1997. The molecular epidemiology studies proved that the high similarity between the HPAI H5N1 isolates from Qinhai Lake in China and those from the subsequent spread to Europe and Africa during 2005-2006, suggesting the role of migratory birds. Taiwan, is located amid the flying route of migratory birds, thus elevates the possibility of imported avian influenza viruses. To avoid public health threat, we monitored the molecular changes of the 26 low pathogenic avian influenza (LPAI) H5 viruses isolated from healthy ducks through a virological surveillance system established in a large LBM in Northern Taiwan during Sept. 2005- Oct. 2006 and investigated their public health implications.he aims of the study are to understand the molecular epidemiology and biological characteristics of the 26 LPAI H5 viruses and to examine any molecular differences between single H5 subtype and mixed subtype AI viruses. To separate the mixed viruses, we used one generation plaque purification on [Madin Darby canine kidney (MDCK) cells] and employed reverse transcriptase-polymerase chain reaction (RT-PCR) from the single plaque to get the full-length sequences, particularly each of the eight segments of all the 11 H5N2 viruses. Phylogenic analysis of our LPAI H5 isolates compared with different HPAI and LPAI H5 viruses in Asia and characterization of amino acids (AAs) involved in receptor binding sites, pathogenesis related sites, and anti-viral drug resistant sites were further investigated.hylogenic analysis showed that all the 26 duck LPAI H5 viruses were strikingly different from the 2003 Taiwan’s chicken LPAI H5N2 viruses (HA similarity:76.2%-79.4%; N2 similarity:84.3%-85.5%). Additionally, 15.4% (4/26) of the M gene of the studied duck LPAI H5 viruses might have the co-evolution with the 1999 Taiwan’s chicken H6N1 virus [A/chicken/Taiwan/165/99 (H6N1)] (similarity: 96.5%-99.3%). In NS gene, 11.6% (3/26) of our duck LPAI H5 viruses were grouped much closer to A/duck/Hokkaido/447/2000(H5N3) (similarity:94.8%-99.8%). Interestingly, A/duck/Taiwan/DV647/2006 (H5Ny) isolated from Tainan in October of 2006 when migrating birds flew into Taiwan, had its seven segments other than NA gene were very different from our other H5 isolates (similarity:70.1%-93.2%) implied that it might be an imported H5 virus. xamining the AAs of the 26 duck LPAI H5 viruses revealed four major findings. First, the HA gene showed the lowest similarity (88.6%-98.9%) while the N2 gene showed the greatest one (94%-100%) among the eight segments of the 11 H5N2 viruses. Second, all of them were LPAI viruses demonstrated by the AAs at the cleavage site of HA protein (---R/), however, A/duck/Taiwan/DV647/2006 (H5Ny) virus contained more than one basic aa “REKR/” at such a cleavage site. In evaluating the possible changes of the receptor binding site (RBS) of HA due to passage of the single plaque-derived AIV (PP-AIV) on unnatural host cells, we found that the patterns of AAs of single vs multiple subtyped LPAI H5 duck viruses were quite similar implying that one passage of the PP-AIV on MDCK cells was insufficient to change RBS. In addition, 71.4% (10/14) of these RBS had the same sites as LPAI A/duck/ Singapore/ Q-F119-3/1997 (H5N3) virus. Furthermore, the sequences of the antigenic epitopes of HA protein of our LPAI H5 isolates were very much different from those currently available HPAI H5 viruses; thus we can’t map the antigenic sites of our LPAI H5 viruses to the known structure of HPAI H5 viruses. Third, all the known AAs at the sties related to viral pathogenicity (PB1-F2, PB2-627, NS1-92, NA Stalk) maintained all the molecular traits of LPAI and no mutation occurred at NS1-92 for IFN. Lastly, the analysis of antiviral important AAs of NA and M2 proteins of our LPAI H5 isolates found that all of them were sensitive to the neuraminidase inhibitor, Tamiflu and adamantane derivatives, respectively. All above findings further supported that our LPAI H5 viruses are very unique to Taiwan and persistence of these LPAI H5 viruses over several months may imply the potential endemicity of these LPAI H5 viruses in certain parts of Taiwan.n summary, Taiwan’s 2005-2006 LPAI H5 viruses isolated from healthy ducks of one LBM had no public threat to humans and provided us the first time to learn biologic characteristics of LPAI H5 viruses. However, many questions still remained. Whether may the quasi-species of H5 be different in using less suitable cell lines to culture? How does the glycosylation of viral HA proteins affect receptor binding sites and their subsequent interactions with antibodies? Most importantly, would changes of LPAI H5 viral AAs and nucleotides have special patterns while time goes by? If it does, are their patterns helpful to understand the transition from LPAI to HPAI so that the trends towards to HPAI become more predictable and avoidable? We believe that thorough understanding the underlying mechanisms involved in stepwise micro-changes at each of the eight segments from LPAI to HPAI will definitely help to prevent future public health threat of pandemic influenza.