Sequence and Phylogenetic Analysis of Three Isolates of Avian Influenza H9N2 from Chickens in Southern China

Three isolates of H9N2 inﬂuenza virus were isolated from chickens in Guangxi province. In this study, eight full-length genes of each of the H9N2 isolates were obtained and sequenced. Sequence analysis and phylogenetic studies were conducted by comparing eight genes of each isolate with those of the available H9N2 strains at GenBank. Results showed a high degree of homology between the Guangxi isolates and isolates from Guangdong and Jiangsu provinces, suggesting that Guangxi isolates originated from the same source. However, the eight genes of these three isolates from Guangxi were not in the same sublineages in the phylogenic trees, which suggest that they were products of natural reassortment between H9N2 avian inﬂuenza viruses from di ﬀ erent sublineages. The 9 nucleotides ACAGAGATA encoding amino acids T, E, I were absent between nucleotide 205 and 214 in the open reading frame of NA genes in the Guangxi isolates. AIV strains that infect human have, in their HA proteins, leucine at position 226. The analysis of deduced amino acid sequence of HA proteins showed that the position 226 of these isolates was glycine instead of leucine, suggesting that these three isolates di ﬀ er from H9N2 AIV strains isolated from human infections.


Introduction
Avian influenza virus (AIV) belongs to the Orthomyxoviridae family of RNA viruses. It is an enveloped virus with a helical nucleocapsid and eight segments of single-stranded negative-sense RNA. The envelope contains the haemagglutinin (HA) and neuraminidase (NA) proteins, of which there are currently 16 HA (H1-H16) and 9 NA (N1-N9) subtypes [1,2]. The H9N2 subtype that is worldwide spread in poultry could infect humans and is prevalent in China [3][4][5][6][7][8]. The first outbreak of disease attributed to H9N2 in China happened in Guangdong province in 1994. Other strains of this subtype were subsequently isolated [6,7,9,10].
Studies to characterize these strains were focused on the HA genes for one of the surface viral proteins that plays a crucial role in the pathogenicity [10]. Recent studies characterized the nonstructural [11] and neuraminidase genes [12] of H9N2 viruses in China. Others showed that some novel reassortant H9N2 influenza viruses may possess certain genes from H5 subtype viruses [7,13,14]. In this study, we have characterized three strains of H9N2 influenza viruses that were isolated from chicken farms in Guangxi province. All eight full-length genes of these isolates were obtained individually by means of RT-PCR. Sequence analysis and phylogenetic study were conducted by comparing the eight genes of each isolate with sequences available in Genbank (H9N2 isolates from 1966-2005).

RNA Extraction and RT-PCR.
Viral RNA was obtained from allantoic fluid by TRIzol extraction according to the method described by Xie et al. [16]. The concentrations of RNA were determined by spectrophotometry using the UV2501PC (Shimadzu Cooperation, Tokyo, Japan) and then stored at −20 • C. Reverse transcription was done by using random primers with RNA PCR (AMV) V3.1 kit (TaKaRa Biotechnology, Dalian, China). Amplification of the eight full-length genes was carried out by PCR as described previously by Xie et al. [16] using pairs of specific primers as described in Table 1. The reaction was carried out at 30 • C for 5 minutes followed by 60 minutes incubation at 37 • C. PCR was performed in a reaction mixture of 50 μL containing master mix buffer with dNTPs, 50-time Advantage 2 polymerase (Clontech Mountain View, California, USA), 10 μM of each specific primer, and 200 ng cDNA template. The PCR condition for the amplification of Pb2, Pb1, and HA was 95 • C for 2 minutes denaturation, and 35 cycles of 95 • C for 30 seconds, 62 • C for 45 seconds (annealing), and 68 • C for 3 minutes, followed by 70 • C for 10 minutes final extension. The PCR condition for the amplification of NP, NA, M, and NS genes was the same as above, except that the annealing temperature was reduced to 60 • C for NP and NA genes and 58 • C for M and NS genes.

Sequence Analysis.
All PCR products were subjected to electrophoresis in a 1% (w/v) agarose gel. DNA fragments of the expected length were extracted and purified with DNA Glass-milk Rapid Purification Kit (MK001-2, BioDev, Beijing, China). The purified DNA fragments were cloned into pMD18-T easy vector according to the manufacturer protocol (TaKaRa Biotechnology Co. Ltd., Dalian, China).
Three clones of each of the 8 fragments were sequenced by the Takara Biotechnology Co. Ltd., Dalian, China. DNA sequences of eight genes were compared with the GenBank database. DNA sequences of each cloned gene were repeated twice to confirm the similarity of sequence data. The nucleotide sequences obtained in this study are available in the GenBank database under accession numbers: DQ485205 to DQ485228.

Phylogenetic Analysis.
The DNA sequences of the eight genes of three Guangxi H9N2 subtypes were compared initially with the Megalign program of the package with the Clustal alignment algorithm (DNASTAR Inc., Madison, Wis, USA) against the twelve H9N2 virus sequences as listed in Table 3. Pairwise sequence alignments were also performed with the Clustal alignment algorithm in the Megalign program to determine sequence similarity and phylogenetic relationship of different H9N2 subtype viruses. The ancestral relationships within eight genes among the fifteen H9N2 subtypes are presented in a phylogenetic tree created by the Megalign program. . This sequence comparison also shows that 9 nucleotides between nucleotide positions 205 and 214, lying between position 187 to 195 open reading (ORF) in the NA genes, are absent in these three Guangxi isolates (see Table 3) analyzed here. These 9 nucleotides encode for amino acids T, E, I. The deduced amino acid sequences of HA genes of C/GX/1/00, C/GX/14/00, and C/GX/17/00 at the cleavage site of HA contain a PARSSR/GL motif, which denotes the sequence found in low pathogenicity avian influenza (LPAI) viruses, as described previously [4,10,17].

Discussion
AIV has a segmented negative-strand genome that includes HA, NA, M, NS, NP, PA, Pb1, Pb2 genes. The results of sequencing and comparisons with the other H9N2 virus sequences showed that we have obtained each of eight genes from three Chinese isolates, C/GX/1/00, C/GX/14/00, and C/GX/17/00 successfully.
Previous studies have defined two distinct lineages of H9N2 influenza viruses: North American and Eurasian lineages. The Eurasian lineage consists of at least three sublineages [7,13,14]. Phylogenetic analysis in our study showed similar patterns (Figures 1-8). Viral genomes of H9N2 viruses analyzed in previous studies had shown a common source of origin from southern China [6,9,10,18]. Sequence comparison and phylogenetic analysis illustrated that NP and Pb1 genes of C/GX/14/00, and NP, Pb1, and Pb2 genes of C/GX/17/00 were incorporated into the sublineage represented by A/Quail/Hongkong/G1/97. HA, M, NP, Pb1, and Pb2 genes of C/GX/1/00, HA and M genes of C/GX/14/00, and HA and M genes of C/GX/17/00 were incorporated into the sublineage represented by A/duck/Hongkong/Y280/97. None of the genes from these three isolates was incorporated into the sublineage represented by A/duck/Hongkong/Y439/97. PA, NA, and NS genes of C/GX/1/00, NA, PA, NS, and Pb2 genes of C/GX/14/00, and the NA, NS, and PA genes of C/GX/17/00 were not incorporated into any of the sublineages. So, it is possible that AIV H9N2 subtype strains C/GX/1/00, C/GX/14/00, and C/GX/17/00 were products of natural reassortment of avian influenza viruses, suggesting that there may be a specific gene pool in China, as described by Lu et al. [18].