Characterization of H9N2 influenza viruses isolated
from vaccinated flocks in an integrated broiler
chicken operation in eastern China during a
5 year period (1998–2002)
Pinghu Zhang,1,2Yinghua Tang,1Xiaowen Liu,1Daxin Peng,1Wenbo Liu,1
Hongqi Liu,1Shan Lu3,4and Xiufan Liu1
1Animal Infectious Disease Laboratory, College of Veterinary Medicine, Yangzhou University,
Yangzhou 225009, PR China
2Jiangsu Center for Drug Screening, China Pharmaceutical University, Nanjing 210038, PR China
3China–US Vaccine Research Center, the First Affiliated Hospital of Nanjing Medical University,
Nanjing 210029, PR China
4Laboratory of Nucleic Acid Vaccines, Department of Medicine, University of Massachusetts
Medical School, Worcester, MA 01605, USA
Received 16 July 2008
Accepted 26 August 2008
In the current study, we characterized H9N2 influenza viruses isolated from vaccinated flocks in an
integrated broiler chicken operation during a 5 year period (1998–2002). Phylogenetic analysis of
the 8 genes of 11 representative viruses showed that they all shared high similarity to that of the
first isolate, A/Chicken/Shanghai/F/1998 (Ck/SH/F/98), and clustered to the same lineages.
Furthermore,all11viruseshad a9 ntdeletionbetweenpositions206and214oftheneuraminidase
gene. These genetic characteristics strongly suggest that these viruses are descendants of the first
isolate. In addition, our study also showed that the H9N2 viruses circulating in the operation during
this 5 year period were evolving, as shown by antigenic variations between viruses manifested
by reactivity with polyclonal antisera and monoclonal antibodies, by haemagglutination with
erythrocytes from different animals, by amino acid differences in haemagglutinin and neuraminidase
proteins, and by variation in their ability to replicate in the respiratory and intestinal tract and to be
transmitted by aerosol. Phylogenetic analysis revealed that the internal genes from some H5N1
viruses of duck origin clustered together with those from H9N2 virus and that the RNP genes of
these H5N1 viruses isolated after 2001 are more closely related to the genes of the Ck/SH/F/
98-like H9N2 viruses, indicating more recent reassortment events between these two subtypes of
viruses. Continuous surveillance of influenza virus in poultry and waterfowl is critical for monitoring
the genesis and emergence of potentially pandemic strains in this region.
Wild birds, including waterfowls, gulls and shorebirds, are
the natural reservoirs for influenza A viruses, in which these
viruses are thought to be in evolutionary stasis (Alexander,
2000; Webster et al., 1992). However, when avian influenza
viruses are transmitted to new hosts, such as terrestrial
poultry or mammals, they evolve rapidly and may cause
occasional severe systemic infection with high morbidity
influenza virus infection occurs commonly in chickens, it is
unable to persist for a long period of time due to control
efforts and/or a failure of the virus to adapt to new hosts
(Suarez, 2000). However,increasing numbers ofoutbreaksin
poultry have occurred in the past 20 years, suggesting that
avian influenza virus can infect and spread in aberrant hosts
Webby et al., 2003; Choi et al., 2004; Lee et al., 2006). In
Hong Kong and mainland China, the H9N2 subtype of
influenza virus has been isolated from pigs and humans with
influenza-like illness during the past decade (Peiris et al.,
1999, 2001; Saito et al., 2001; Xu et al., 2004). These findings
confirm that the H9N2 subtype virus is ubiquitous in China
and poses a greatthreattobotheconomics and public health.
Infections with the H9N2 subtype of the avian influenza
virus (AIV) in domestic poultry, especially in chickens,
The GenBank/EMBL/DDBJ accession numbers for the nucleotide
sequence data obtained in this study are EU753271–EU753350.
Two supplementary tables are available with the online version of this
Journal of General Virology (2008), 89, 3102–3112
3102 2008/005652G2008 SGMPrinted in Great Britain
have been reported frequently in China and other Asian
countries since the late 1990s (Naeem et al., 1999; Guo et al.,
2000; Liu et al., 2003; Choi et al., 2004; Perk et al., 2006; Lee
et al., 2006; Xu et al., 2007). In 1994, for the first time, H9N2
virus was isolated from diseased chickens in Guangdong
province, China. Since then, this subtype of influenza virus
has been found in domestic poultry in other provinces in
China (Chen et al., 1994; Guo et al., 2000; Liu et al., 2003;
Choi et al., 2004; Li et al., 2005; Xu et al., 2007). In late
October 1998, a severe outbreak of H9N2 avian influenza
occurred in 3–5-week-old broiler chickens in a large
integrated broiler chicken operation in eastern China; a
virus designated A/chicken/Shanghai/F/98 (Ck/SH/F/98) was
isolated and characterized (Liu et al., 2003; Lu et al., 2005).
To controlinfectionand transmission of H9N2viruses in the
operation, an intensive vaccination program was implemen-
ted in the following years with a killed oil-emulsion vaccine
based on Ck/SH/F/98. Despite the vaccination effort and the
isolated from chickens during the next 5 years.
In this study, our goal was to determine whether the
circulating H9N2 viruses isolated from the operation
during a 5 year period were from a single introduction
or from repeated reintroduction, by using genetic analysis,
and to explore the extent of antigenic variation among
Virus isolation and identification. Tracheal swabs and faecal
samples were collected and put into 1 ml freezing medium (50%
glycerol in PBS) containing antibiotics, as described previously (Guan
et al., 2000). Supernatants from processed samples were inoculated
into the allantoic and the amniotic cavities of 9–10-day-old
embryonated chicken eggs and incubated for 48–72 h at 35 uC
(Shortridge et al., 1998). The allantoic fluids were harvested and
tested for haemagglutinin (HA) activity, then frozen in aliquots at
270 uC for further analysis. The HA of viral isolates was subtyped by
the haemagglutination inhibition (HI) test using the chickens’
positive antiserum for AIV Ck/SH/F/98 (H9N2). The neuraminidase
(NA) subtypes of the viruses were determined by conducting a BLAST
search on viral nucleotide sequences available in the GenBank
database from the National Center for Biotechnology Information.
The viruses used in this study are listed in Table 1. The 50% egg
infectious dose (EID50) was determined by serial dilutions of virus in
eggs and calculated by using the method described by Reed & Muench
Chicken antisera. Six-week-old, specific pathogen-free (SPF)
chickens were immunized by intramuscular injection of 1 ml oil
emulsion of inactivated whole virus vaccines of representative viruses
for 4 weeks and boosted intravenously with 1 ml allantoic fluids
(containing more than 107EID50virus ml21). Chickens were bled
10 days after the booster injections. Sera were treated with receptor-
destroying enzyme (RDE) to remove non-specific haemagglutination
inhibitors, as described previously (Ninomiya et al., 2002), before an
HI test was performed (Edwards, 2006).
Table 1. H9N2 influenza A viruses isolated from an integrated chicken operation during a 5 year period
All virus strains were isolated from sick chickens. ND, Not determined.
Virus Abbreviation Date of isolation Titre (log10EID50
per 0.2 ml)
*Pathogenicity for 6-week-old SPF chickens.
Characterization of H9N2 avian influenza viruses
Monoclonal antibodies (mAbs). A panel of mAbs to the HA of two
H9N2 strains were used: two mAbs (2A1 and 5D3) to Ck/SH/F/98
and four mAbs (2B12, 5B1, 4A2 and 8E6) to A/Chicken/Shanghai/14/
2001 (Ck/SH/14/01) which were prepared as described by Kaverin
et al. (2004).
Transmission experiments in chickens. For the transmission
experiments, 10 or 11 5-week-old SPF white Leghorn chickens for each
contact group (three to four chickens) and (iii) aerosol contact group
(four chickens) (Table 3). The physical contact group was raised in the
same cage as chickens from the infected group; they were put in the
cage about 30 min after inoculation of the infected group. The aerosol
contact group was placed in a cage directly adjacent to the infected
group with a distance of 100 cm between cages. The infected group was
inoculated orally, intranasally or intratracheally with 107EID50ml21
virus (about 100 chicken infectious doses) (Reed & Muench, 1938).
Tracheal and cloacal swabs were collected 3, 5 and 7 days post-
inoculation, respectively, and AIVs were titrated for infectivity in
embryonated chicken eggs. Meanwhile, birds were observed daily for
signs of disease within 21 days.
RT-PCR and sequence analyses. Viral RNA was extracted directly
from the allantoic fluid of the first passage virus with the Trizol LS kit
(Invitrogen) to reduce the effects of passage history in embryonated
chicken eggs on additional mutations. Reverse transcription was
conducted with Uni-12 primer and PCR was performed with primers
specific for each of the eight RNA segments (primer sequences are
available on request). PCR products were purified with a gel DNA
purification kit (TaKaRa) and sequenced using an ABI Prism dye
terminator cycle sequencing kit on an ABI 377 DNA Sequencer
The nucleotide sequences were analysed with the Seqman module and
the nucleotide and deduced amino acid sequences were aligned and
analysed using the MEGALIGN module of the Lasergene sequence
analysis package (DNASTAR) and MEGA version 3.1 (Kumar et al., 2004).
Evolutionary trees were constructed using the neighbour-joining
method (MEGA version 3.1) on the basis of the following gene
sequences: nt 88–1048 (960 bp) of HA1; 88–1369 or 1378 (1282 or
1291 bp) of NA; 589–1785 (1197 bp) of PB2; 736–1888 (1153 bp) of
PB1; 784–2016 (1233 bp) of PA; 76–1483 (1408 bp) of NP; 53–984
(932 bp) of M; and 66–838 (773 bp) of NS. Estimates of the
phylogenies were calculated by performing 1000 neighbour-joining
bootstrap replicates. The nucleotide sequence data obtained in this
studyare available fromGenBank
In 1998, a disease outbreak was reported in chicken flocks
in a large integrated broiler chicken operation in eastern
China (60 million broiler chickens each year), resulting in
the isolation of an H9N2 influenza virus, Ck/SH/F/98 (Liu
et al., 2003). The disease outbreak was associated with
coughing and respiratory distress in 80% of the birds, with
a 30–50% mortality rate within 2 weeks of the first
appearance of clinical signs. Microbial assay showed that
Escherichia coli was the most common bacterial pathogen,
suggesting that it contributed to the severity of the disease.
The outbreak started in a 5-week-old broiler flock and
spread quickly to many flocks in different locations of the
operation, including breeder flocks. To control the disease
in the operation, an intensive vaccination programme was
implemented with the inactivated whole virus vaccines
prepared from Ck/SH/F/98; all chickens were immunized.
Additionally, biosecurity measures around the broiler
operation were heightened. The antibody titres induced
by this vaccine in the chickens vaccinated in this operation
were approximately 27–9. In order to assess the protective
efficacy of the inactivated vaccine, tracheal and cloacal
swabs were collected from flocks in the operation during a
5 year surveillance period (1998–2002). A total of 22 H9N2
influenza viruses were isolated (Table 1).
Phylogenetic analysis of HA, NA, matrix protein
(M) and non-structural protein (NS) genes
To determine the genetic characterization of these viruses,
all eight gene segments of three viral isolates from 1998,
1999 and 2000, six viral isolates from 2001 and two viral
isolates from 2002 in six different chicken flocks were
partially sequenced. The phylogenetic relationship of the
HA, NA, M and NS genes are shown in Fig. 1. The HA and
NA genes of these 11 viruses shared 94.6–96.9 and 94.7–
96.6% nucleotide identity, respectively, with those of the
A/Chicken/Beijing/1/1994 (Ck/BJ/1/94) H9N2 virus, an
earlier H9N2 isolate from southern China (Fig. 1a and b).
These results indicated that the HA and NA genes of these
viruses belonged to the Ck/BJ/1/94-like lineage of H9N2
viruses. Furthermore, the NA genes of all 11 viral isolates
had a deletion of 9 nt between positions 206 and 214,
suggesting that they were of the same origin.
Similar to analysis of the HA and NA genes, phylogenetic
analysis of the M and NS genes revealed that they also
belonged to the Ck/BJ/1/94-like lineage (Fig. 1c and d),
sharing 96.6–100 and 96.7–97.9% nucleotide identity,
respectively, with those of the Ck/BJ/1/94 H9N2 virus.
The M and NS genes of these viruses were different from
the Gs/GD/1/96-like H5N1 viruses, indicating that they
belonged to different lineage. However, it is notable that
the M gene of H5N1 virus Dk/GD/01/01 and M and NS
genes of H5N1 virus Dk/FJ/19/00, which were isolated
from domestic duck in 2001 and 2000, respectively, were
clustered together with those of the H9N2 viruses included
in this study, suggesting that reassortment events might
have occurred between these two virus subtypes.
Taken together, our findings clearly indicated that the HA,
NA, M and NS genes of these H9N2 variants isolated from
this integrated operation belonged to Ck/BJ/1/94-like
lineage and that they are different from those of Qa/HK/
G1/97-like H9N2 viruses. However, phylogenetic analysis
of the M and NS genes showed some of the duck H5N1
isolates were closely related to these H9N2 variants,
suggesting that a reassortment event may have occurred
in this region. Our findings further supported the view that
interspecies transmission between different types of poultry
exists in southern China.
P. Zhang and others
3104Journal of General Virology 89
Phylogenetic analysis of the PB2 and PB1, PA and
The results of a parallel phylogenetic analysis of the PB2,
PB1, PA and NP genes which code for the proteins of the
RNP complex are very similar (Fig. 2a–d). The four RNP
complex genes of the H9N2 viruses shared more than
96.4% nucleotide identity. The topologies of the phylo-
genetic trees of the four RNP complex genes were also very
similar. The four RNP genes of the 11 H9N2 viral isolates
had evolved in a sequential fashion from 1998 to 2002 and
Fig. 1. Phylogenetic trees for the HA1 (a), NA (b), M (c) and NS (d) genes of the H9N2 influenza A viruses that were analysed.
The HA1 and M phylogenetic trees are rooted to A/Duck/Alberta/60/1976 (Dk/Alberta/60/76) (H12N5) (GenBank accession
no. AB288334) and A/Equine/Prague/1/1956 (Eq/Prague/1/56) (H7N7) (GenBank accession nos M73519 and CY005803),
respectively. The length of the horizontal lines is proportional to the minimum number of nucleotide differences required to join
nodes. Vertical lines are for spacing and labelling. The viruses obtained in this study are underlined and their names can be
found in Table 1. Abbreviations: Qa, quail; Gs, Geese; G1, A/quail/HongKong/G1/97; Ck, chicken; Dk, duck; Ty, Turkey; CA,
California; WI, Wisconsin; HK, Hong Kong; Kor, Korea; BJ, Beijing; SD, Shandong; GD, Guangdong; SH, Shanghai; NC,
Nanchang; YN, Yunnan; HB, Hubei; HN, Hunan; FJ, Fujian; ZJ, Zhengjiang.
Characterization of H9N2 avian influenza viruses
formed a unique lineage, a Ck/SH/F/98-like lineage (Li et
al., 2005), which was distinguishable from the Ck/BJ/1/94-
like (H9N2) lineage and Gs/GD/1/96-like (H5N1) lineage.
We noted that the RNP complex genes of a few H5N1
viruses derived from domestic duck also clustered into the
Ck/SH/F/98-like (H9N2) lineages, indicating that reassort-
ment events had occurred between these two subtypes of
viruses. For example, the PB2 gene of H5N1 viruses Dk/FJ/
19/00 and Dk/SH/35/02 (Fig. 2a), the PB1 gene of Dk/SH/
35/02 (Fig. 2b), the PA gene of H5N1 viruses WDk/HN/
211/05, Dk/SH/38/01 and Dk/SH/35/02 (Fig. 2c) and the
NP gene of eight H5N1 viruses, including WDk/HN/211/
Fig. 2. Phylogenetic trees for the PB2 (a), PB1 (b), PA (c) and NP (d) genes of the H9N2 influenza A viruses that were
analysed. The phylogenetic trees of PB2, PB1, PA and NP are rooted to Eq/Prague/1/56 (H7N7) (GenBank accession nos
M73519 and CY005803) and A/Equine/London/1416/1973 (Eq/London/16/73) (H7N7) (GenBank accession nos M25928
and M26087), respectively. The length of the horizontal lines is proportional to the minimum number of nucleotide differences
required to join nodes. Vertical lines are for spacing and labelling. The viruses obtained in this study are underlined, and the
name of the viruses and abbreviations are as described in Fig. 1.
P. Zhang and others
3106 Journal of General Virology 89
05 and Dk/SH/35/02 (Fig. 2d), all clustered into a
corresponding Ck/SH/F/98-like lineage. It is interesting to
note that many H5N1 viruses isolated from domestic duck
since 2001 in mainland China contain an NP gene segment
with high identity to that of Ck/SH/F/98. Furthermore, the
four RNP genes of the H5N1 virus Dk/SH/35/02 isolated in
Shanghai in 2002 shared very high identity with those of
Ck/SH/F/98-like variants and also clustered into the same
lineage. Taken together, these findings indicate that the Ck/
SH/F/98-like viruses in this study may have originated
from a single introduction and become a possible donor
that provided internal genes for H5N1 viruses since 2001.
Antigenic analysis of H9N2 isolates
The H9N2 viruses included in the current study were
compared antigenically with representative H9N2 virus
from 1994 (Ck/GD/SS/94) and the first isolate from this
operation, Ck/SH/F/98. Chicken polyclonal hyper-immune
sera against the representative H9N2 virus Ck/GD/SS/94
and Ck/SH/F/98 reacted equally well with all the viruses
isolated from this operation (Table 2). Although the HI
test via hyper-immune sera against Ck/GD/SS/94 and
Ck/SH/F/98 did not allow for differentiation between these
two viruses, variation in titres between viruses existed,
suggesting a close relationship in antigenicity among these
viruses. The reactivity to a panel of mAbs did not
discriminate between the earlier H9N2 and these H9N2
variants. However, one of these H9N2 variants, Ck/SH/11/
01, demonstrated a unique reactivity pattern; it reacted
with only two of six mAbs. We sequenced the HA gene of
this virus and revealed the presence of a potential
carbohydrate site due to a mutation from Ser to Asn at
aa 127 in the HA1 region (data not shown). Based on the
presence or absence of this mutational change, we found
that all the viruses isolated from eastern China during the
past 10 years could be separated into two groups which
correlated with the reactivity to the panel of mAbs (data
not shown). A mutagenesis study to confirm whether this
amino acid change directly contributes to the reactivity
profile of mAbs is now under way in our laboratory.
Pathogenicity of the H9N2 influenza viruses in
We examined the pathogenicity of these viruses by
inoculating eight 6-week-old SPF white Leghorn chickens
intravenously with 100 ml allantoic fluid containing the test
Table 2. Cross-reactivity of representative H9 viruses in the HI test
,, Titre less than 1:10; ., titre more than 1:10240.
mAbs to H9 virusesPolyclonal sera*
2A15D3 2B125B14A2 8E6
*Antisera were treated with RDE.
DCk/GD/SS/94 is a variant of Ck/BJ/1/94 H9N2 virus.
Characterization of H9N2 avian influenza viruses
virus. In contrast with the high morbidity and mortality
observed for chickens living in the operation, the SPF
chickens challenged with most of these viral isolates did not
develop severe clinical signs or die during the 3 week
observation period. At 21 days post-infection, all of the
challenged chickens were seroconverted. However, Ck/SH/
7/01 and Ck/SH/1/02 seemed to be more pathogenic than
other viruses, as chickens infected with these viruses showed
clinical signs of disease, including diarrhoea and facial
oedema, with decreased food and water consumption.
Replication and transmission of H9N2 influenza
isolates in chickens
In order to explore the biological features that enabled the
H9N2 viruses to circulate for an extended period of time in
an integrated chicken operation, we examined their
replication and transmission properties in chickens. Table
3 summarizes the experiment to examine replication and
transmission of the viruses in chickens. Most of these
variants were shed mainly from the respiratory tract with
titres in the range of 4–7 log10EID50ml21on day 5 post-
inoculation. The maximum virus shedding was observed
between days 3 and 5 post-inoculation. Viruses were
detected in tracheal swabs of directly inoculated birds and
physical contact birds as early as 24 h post-contact. It was
notable that the ability of these viruses to replicate and be
transmitted among the birds was correlated with their
pathogenicity. Most of the viruses replicated in the
inoculated chickens, but failed to transmit efficiently to
the aerosol contact group. However, the more pathogenic
Ck/SH/7/01 and Ck/SH/1/02 viruses replicated efficiently
in the infected chickens and were transmitted efficiently to
the physical contact group and the aerosol contact
chickens. Ck/SH/7/01 and Ck/SH/1/02 viruses could be
detected in the tracheal and cloacal swabs from the aerosol
contact birds by 24 h post-contact; however, lower virus
titres were detected in the cloacal swabs compared with the
tracheal swabs from both infected and contact birds. Ck/
SH/7/01 and Ck/SH/1/02 virus titres were present in the
tracheal and cloacal swabs of the birds on day 7 post-
infection. The mean antibody titres in the aerosol contact
group of Ck/SH/7/01 and Ck/SH/1/02 virus at days 10 and
20 post-inoculation were significantly higher than those of
other viruses (Table 3).
Haemagglutination activity with erythrocytes from
Receptor specificity of the HA protein has been shown to
be a determinant of host range for influenza viruses
(Connor et al., 1994). Previous work has demonstrated that
the receptor specificity of influenza viruses is related to the
agglutination of erythrocytes from different animal species
Table 3. Transmission of H9N2 influenza viruses in chickens
The number of chickens shedding/total number of chickens, and antibody titre are given at the indicated times post-infection.
Virus Method of
No. of chickens shedding/total no. of chickensAntibody titre*
3 days 5 days 7 days
TracheaCloaca TracheaCloacaTracheaCloaca 10 days 20 days
Ck/SH/F/98 Direct inoculation
11, 11, 11
3, 4, 0
2, 0, 0, 0 (Bbc)D
3, 7, 6
3, 0, 0
0, 0, 0, 0 (Bc)
6, 7, 4
0, 0, 0
0, 0, 0, 0 (Bc)
9, 10, 9, 11
6, 6, 6, 11 (Aa)
8, 5, 6
0, 0, 6, 6
0, 0, 0, 0 (Bc)
10, 10, 11
9, 10, 9
2, 2, 2, 2 (Bb)
11, 11, 9
4, 11, 11
3, 0, 0, 0 (Bb)
4, 7, 8
0, 1, 0
0, 0, 0, 0 (Bb)
4, 7, 6
0, 0, 5
0, 0, 0, 0 (Bb)
7, 7, 6
0, 0, 7, 8
0, 0, 0, 0 (Bb)
11, 11, 11
8, 9, 11
8, 9, 11,11 (Aa)
*Antibody titre was calculated by 2N.
DLetters indicate a significant difference between results in the same column (upper case, P,0.01 and lower case, P,0.05).
P. Zhang and others
3108 Journal of General Virology 89
(Ito et al., 1997). In order to explore the receptor binding
specificity of the H9N2 viruses, the agglutination patterns
of these H9N2 variants were systematically analysed with
erythrocytes from different animal species (Supplementary
Table S1, available in JGV Online). All of these H9N2
variants agglutinate chicken, duck, goose, pigeon, quail,
guinea pig, dog and human (type O) erythrocytes, whereas
almost all of them did not agglutinate pig erythrocytes.
Furthermore, compared with the agglutination patterns
with chicken, duck and goose erythrocytes, part of these
viruses agglutinate pigeon and quail erythrocytes with
much lower titre. Based on the difference in haemagglu-
tination patterns from those of
erythrocytes, most of these H9N2 variants could be
discriminated into two further groups. In addition, two
2001 isolates (Ck/SH/11/01 and Ck/SH/12/01) could
agglutinate donkey erythrocytes. These results clearly
showed that the receptor binding site of these H9N2
variants was, in fact, different. Therefore, direct analysis of
receptor specificity of these H9N2 variants using defined
glycoconjugates is needed to understand the differences in
agglutination patterns of these H9N2 variants.
buffalo and goat
Amino acid differences in the HA and NA proteins
As shown in Supplementary Table S2 (available in JGV
Online), amino acid changes in the HA glycoprotein
accumulated in a sequential fashion over time. Although
some amino acid changes occurred in previously proposed
antigenic sites (Ha et al., 2002), the majority of the
changes were located in the outer surface of the globular
head of HA1. Compared with Ck/SH/F/98, 11 aa
differences in the HA protein were found in 2000, 2001
and 2002 isolates. In addition, 11 unique amino acid
substitutions in the HA protein from 2002 isolates were
distinguishable from those of viruses isolated before 2001.
However, the majority of these substitutions occurred in
the distal globular head of HA1 rather than the previously
proposed antigenic sites and only one change, from Asn to
Ser at position 148 (158 in H3 numbering), was located in
a previously proposed antigenic site of the HA protein.
The other six critical amino acid residues involved in
forming the receptor binding site are highly conserved, as
previously reported (Liu et al., 2003). A residue change at
position 180 (190 in H3 numbering) of the receptor
binding domain of the HA protein from 1998 and 1999
isolates was found to be changed from Ala to Thr in some
of the 2000–2001 isolates. The change from Glu to Leu at
position 216 (226 in H3 numbering) of H9N2 viruses has
been reported to have receptor specificity similar to that
of human H3N2 viruses (Saito et al., 2001). A change at
position 216 (Gln to Leu) of the receptor binding domain
of the HA protein from 2001 isolates was maintained in
2002 isolates. However, whether these H9N2 variants
possessing Leu substitution at position 216 have the
ability to infect humans and other mammals is yet to be
All of the isolates possessed the same amino acid sequence
(Pro-Ala-Arg-Ser-Ser-Arg) at the cleavage site between
HA1 and HA2, which is characteristic of the low
pathogenic AIVs. Seven potential glycosylation sites, five
in HA1 (11, 123, 200, 280 and 287) and two in HA2 (154
and 213), were highly conserved, as is found in most of the
isolates from mainland China (Liu et al., 2003; Li et al.,
2005). In addition, Ck/SH/11/01 has an additional
potential glycosylation site at residue 127, due to the
change of Ser to Asn, which was not found in the HA of
other viruses in this study.
Analysis of the substitutions of the entire NA gene revealed
that amino acid substitutions in NA of the H9N2 viruses
during a 5 year period were erratic, while amino acid
residues located in the substrate-binding pocket were
highly conserved. A highly variable region, where 15
substitutions occurred between residues 19 and 85, was
observed. Furthermore, two regions with multiple sub-
stitutions were also found. One is a 110 aa region from
residues 344 to 455, which are located in the head of NA,
where eight amino acid substitutions have occurred.
Another multiple mutation region is between residues
125 and 221, where eight substitutions occurred during this
5 year period. Comparison of the substitution rate of
amino acids in the NA stalk with that from the entire NA
of H9N2 influenza viruses revealed that the rate of amino
acid substitutions in the stalk region of the NA is higher
than that in the entire coding region or in the head alone,
which supports the observations by Xu et al. (2004). The
NA proteins from viruses of 2002 differed from those of
1998–2001 isolates by 10 aa. Only one of the four
characteristic amino acids, Met-20, from viruses from
2000 was maintained in isolates from 2001. Six potential
NA protein glycosylation sites (at positions 69, 86, 146,
200, 234 and 404) from H9N2 viruses in this study were
highly conserved. Two isolates of 2002 had an additional
potential glycosylation site at residue 246 due to the change
of His to Asn, which was not observed in other viruses.
Furthermore, all of the NA proteins from the H9N2 viruses
examined had a ‘marking’ deletion of 9 nt at positions
206–214 (at aa 62–64) in the stalk of the protein, as
previously described (Lu et al., 2005).
In the present study, we isolated and characterized 22
H9N2 viruses from chicken flocks in a large integrated
broiler operation over a 5 year period. One question arises
from this: do these viruses originate from a single
introduction and maintain a presence in the operation or
do they originate from repeated reintroduction? Genetic
characterization of 11 representative isolates of these
viruses revealed that their HA, NA, M and NS genes all
belonged to the Ck/BJ/1/94-like lineage with high identity,
while their four RNP genes (PB2, PB1, PA and NP) all
clustered together to form a specific Ck/SH/F/98-like
lineage. Furthermore, the NA gene of all 11 viruses had a
Characterization of H9N2 avian influenza viruses
9 nt deletion between positions 206 and 214. This genetic
evidence strongly suggested that all of these viruses are
descendants of the first isolates, Ck/SH/F/98, and that they
originate from a single introduction.
Interestingly, phylogenetic analysis revealed that the RNP
genes of some H5N1 viruses isolated from domestic ducks
after 2001 are more closely related to the genes of the Ck/
SH/F/98-like H9N2 viruses. For example, one H5N1
isolate, Dk/SH/35/02, a novel genotype in domestic ducks,
Phylogenetic analysis of the internal genes of Dk/SH/35/
02 indicated that the four RNP genes are closely related to
those of the Ck/SH/F/98-like H9N2 viruses, while the other
genes (M and NS) belong to an H5N1-like virus. In
addition, phylogenetic analysis also revealed that a few
H5N1 viruses isolated from domestic duck since 2001
contained a Ck/SH/F/98-like RNP gene complex, while the
M and NS genes of these H5N1 viruses were similar to the
Ck/BJ/1/94-like H9N2 lineage. These findings strongly
suggest that interspecies transmission between chickens
and domestic ducks may have occurred in this region
during the past 10 years.
the suburbof Shanghai.
Chickens are generally considered to be aberrant hosts of
influenza viruses because the mutation rates of many
chicken viruses are higher than those for viruses isolated
from aquatic birds (Suarez, 2000). Furthermore, when an
influenza virus infects a new species, infection typically
lasts for a short period of time and rarely transmits well
enough in the new species to cause an epidemic, due to
control efforts or failure of the virus to adapt to the new
host (Suarez, 2000; Webster et al., 1992). However, several
studies showed that H9N2 influenza viruses have circulated
widely in chicken flocks for more than 10 years since their
first detection in mainland China in 1994 (Chen et al.,
1994), indicating that H9N2 influenza viruses have the
ability to circulate in chickens for a long period of time.
The results presented in this study also indicate that H9N2
influenza viruses have been circulating in chicken flocks
from the same integrated chicken operation for at least
Most of the avian influenza viruses are restricted to
infecting aquatic birds. However, more and more studies
are reporting that H9N2 viruses can replicate and circulate
in terrestrial poultry, such as chicken, quail and pigeon.
Previous studies have reported that G1-like H9N2 viruses
can infect humans and can replicate in the respiratory
system of mice (Choi et al., 2004; Guo et al., 2000). The
present study demonstrates that H9N2 viruses isolated in
eastern China have gradually acquired the ability to
replicate efficiently in the respiratory system of chickens
and can effectively transmit throughout chicken flocks by
aerosol. It is possible that these variants emerged as a result
of selective pressure due to vaccination or adaptation in a
single host for a long period of time.
The molecular basis of adaptation of influenza A viruses to
a new host species is poorly understood. Previous studies
have demonstrated that mutation of the polymerase
complex is critical for adaptation to the new environment
once the virus has been transmitted to a new host (Gabriel
et al., 2005). The four RNP genes of these H9N2 viruses are
highly related to those from the subtype H5N1 that prevails
in poultry in southern China, indicating that they probably
shared a recent common progenitor virus. Whether these
RNP complexes are necessary for mediating adaptation of
the H9N2 virus to chicken should be investigated in future
studies. Previous studies reported that the deletion of
amino acids in the NA gene of influenza A virus is related
to the adaptation of influenza A virus to its host
(Alexander, 2000). However, further studies are needed
to explore whether a 3 aa deletion at aa 62–65 of the NA
gene of these H9N2 variants is related to the adaptation to
A major determinant of host range is the affinity of the
viral HA protein to the sialic acid (SA) receptor of the host
cell. A change of preference for the avian influenza virus for
a(2, 3)-linked Gal to a(2, 6)-linked Gal for the SA receptor
is highly related to host specificity. Stephenson et al. (2003)
demonstrated that erythrocytes from different hosts can be
used to rapidly define the receptor specificity of influenza
viruses because of the differences in the oligosaccharide
composition of glycoproteins and glycolipids of erythro-
cytes. In this study, we found that the chicken H9N2
viruses isolated during the 5 year period differed from each
other in their ability to agglutinate erythrocytes from
different hosts, suggesting differences in receptor specifi-
cities in the H9N2 influenza viruses. Ito et al. (1997)
demonstrated that the a(2, 3) linkage and NeuGc, not
NeuAc, recognition appear to be essential for agglutination
of bovine and equine erythrocytes. Whether the difference
in the agglutination pattern of these H9N2 variants to goat
and buffalo erythrocytes is related to NeuGc recognition
still needs to be determined. Wan & Perez (2007) also
demonstrated that the Glu to Leu substitution at amino
acid position 226 in HA allows H9N2 viruses to replicate
more efficiently in human airway epithelial cells cultured in
vitro. The substitution of Gln for Leu at residue 226 and the
change of Ala to Thr at position 190 occurred in the
receptor binding site of those H9N2 variants that have
previously been reported to be involved in the binding
specificity to receptors in host cells (Perez et al., 2003).
Therefore, further studies are needed to understand the
role of these mutations in the receptor binding site in
restricting the host range of these H9N2 variants.
In contrast, it has been reported that avian influenza
viruses have not been under constant threat by vaccines
and also are not very well matched, antigenically, with the
chicken vaccines that are efficacious in poultry (Lipatov et
al., 2004; Webster et al., 1992). However, recent studies
have demonstrated that the commercial vaccine was not
able to completely prevent virus shedding when chickens
were challenged with antigenically different isolates (Lee
et al., 2004). Our study confirms the previous finding that
chickens vaccinated with homologous vaccine may be
P. Zhang and others
3110 Journal of General Virology 89
protected against clinical disease resulting from AIV
challenge, but it is very difficult to completely stop
infection in the field.
In this study, we have demonstrated that recent variants
isolated during a specific 5 year period might be derived
from the early isolate, Ck/SH/F/98. The H9N2 viruses are
poorly pathogenic for chickens under the experimental
conditions; however, they often infect birds and compete
with other potentiallyharmful
Staphylococcus, Haemophilus and E. coli, pathogens that
have caused considerable economic loss to the poultry
industry in China (Guo et al., 2000; Kishida et al., 2004).
The available evidence indicates that Ck/BJ/1/94-like
viruses have become established in chickens as a stable
lineage in mainland China (Liu et al., 2003; Li et al., 2005).
Furthermore, with their long time adaptation to chickens,
the symptoms associated with infection with these viruses
are less noticeable but could gradually acquire the ability to
spread efficiently in chicken flocks. Once these viruses
acquire the ability to replicate and spread efficiently in a
human population, H9N2 influenza virus could be likely to
become a novel influenza subtype capable of causing
human influenza pandemics in the future. Therefore, it is
imperative that particular attention is paid to H9N2 viruses
of avian origin to avert any future pandemic in humans.
This study was supported by National Key Technologies R & D
Program of China (grant No. 2006BAD06A01), by National High-
Tech Research and Development Program of China (grant no.
2006AA10A205) to X. Liu and by Jiangsu Nature & Science Fund,
BK2007581(to Jiangsu Center
Pharmaceutical University) and BK2006728 (to China–US Vaccine
Research Center, The First Affiliated Hospital of Nanjing Medical
University). We are grateful to Professor Xinan Jiao for kindly
providing Ck/SH/F/98-specific mAbs and to Guijie Hao, Haiyan
Wang, Min Gu, Xusheng Qiu and Feng Xue for their kind help with
animal experiments. We would like to thank Dr Jill M. Grimes
Serrano for critical reading of the manuscript.
for Drug Screening,China
Alexander, D. J. (2000). A review of avian influenza in different bird
species. Vet Microbiol 74, 3–13.
Chen, B. L., Zhang, Z. J. & Chen, W. B. (1994). Isolation and
preliminary serological characterization of type A influenza viruses
from chickens. Chin J Vet Med 22, 3–5.
Choi, Y. K., Ozaki, H., Webby, R. J., Webster, R. G., Peiris, J. S.,
Poon, L., Butt, C., Leung, Y. H. C. & Guan, Y. (2004). Continuing
evolution of H9N2 influenza viruses in southeastern China. J Virol 78,
Connor, R. J., Kawaoka, Y., Webster, R. G. & Paulson, J. C. (1994).
Receptor specificity in human, avian, and equine H2 and H3
influenza virus isolates. Virology 205, 17–23.
Edwards, S. (2006). OIE laboratory standards for avian influenza.
Dev Biol (Basel) 124, 159–162.
Gabriel, G., Dauber, B., Wolff, T., Planz, O., Klenk, H. D. & Stech, J.
(2005). The viral polymerase mediates adaptation of an avian
influenza virus to a mammalian host. Proc Natl Acad Sci U S A
Guan, Y., Shortridge, K. F., Krauss, S., Chin, P. S., Dyrting, K. C., Ellis,
T. M., Webster, R. G. & Peiris, M. (2000). H9N2 influenza viruses
possessing H5N1-like internal genomes continue to circulate in
poultry in southeastern China. J Virol 74, 9372–9380.
Guo, Y. J., Krauss, S., Senne, D. A., Mo, I. P., Lo, K. S., Xiong, X. P.,
Norwood, M., Shortridge, K. F., Webster, R. G. & Guan, Y. (2000).
Characterization of the pathogenicity of members of the newly
established H9N2influenza virus lineages in Asia. Virology 267, 279–288.
Ha, Y., Stevens, D. J., Skehel, J. J. & Wiley, D. C. (2002). H5 avian and
H9 swine influenza virus haemagglutinin structures: possible origin of
influenza subtypes. EMBO J 21, 865–875.
Ito, T., Suzuki, Y., Mitanaul, L., Vines, A., Kida, H. & Kawaoka, Y.
(1997). Receptor specificity of influenza A viruses correlates with the
agglutination of erythrocytes from different animal species. Virology
Kaverin, N. V., Rudneva, I. A., Ilyushina, N. A., Lipatov, A. S., Krauss, S.
& Webster, R. G. (2004). Structural differences among hemagglutinins
of influenza A virus subtypes are reflected in their antigenic
architecture: analysis of H9 escape mutants. J Virol 78, 240–249.
Kishida, N., Sakoda, Y., Eto, M., Sunaga, Y. & Kida, H. (2004). Co-
infection of Staphylococcus aureus or Haemophilus paragallinarum
exacerbates H9N2 influenza A virus infection in chickens. Arch Virol
Kumar, S., Tamura, K. & Nei, M. (2004). MEGA3: integrated software
for Molecular Evolutionary Genetics Analysis and sequence align-
ment. Brief Bioinform 5, 150–163.
Lee, C. W., Senne, D. A. & Suarez, D. L. (2004). Effect of vaccine use
in the evolution of Mexican lineage H5N2 avian influenza virus.
J Virol 78, 8372–8381.
Lee, Y. J., Shin, J. Y., Song, M. S., Lee, Y. M., Choi, J. G., Lee, E. K.,
Jeong, O. M., Sung, H. W., Kim, J. H. & other authors (2006).
Continuing evolution of H9 influenza viruses in Korean poultry.
Virology 359, 313–323.
Li, C., Yu, K., Tian, G., Yu, D., Liu, L., Jing, B., Ping, J. & Chen, H.
(2005). Evolution of H9N2 influenza viruses from domestic poultry
in Mainland China. Virology 340, 70–83.
Lipatov, A. S., Govorkova, E. A., Webby, R. J., Ozaki, H., Peiris, M.,
Guan, Y., Poon, L. & Webster, R. G. (2004). Influenza: emergence and
control. J Virol 78, 8951–8959.
Liu, H., Liu, X., Cheng, J., Peng, D., Jia, L. & Huang, Y. (2003).
Phylogenetic analysis of the hemagglutinin genes of twenty-six avian
influenza viruses of subtype H9N2 isolated from chickens in China
during 1996–2001. Avian Dis 47, 116–127.
Lu, J. H., Liu, X. F., Shao, W. X., Liu, Y. L., Wei, D. P. & Liu, H. Q. (2005).
Phylogenetic analysis of eight genes of H9N2 subtype influenza virus:
a mainland China strain possessing early isolates’ genes that have been
circulating. Virus Genes 31, 163–169.
Ludwig, S., Stitz, L., Planz, O., Van, H., Fitch, W. M. & Scholtissek, C.
(1995). European swine virus as a possible source for the next
influenza pandemic? Virology 212, 555–561.
Naeem, K., Ullah, A., Manvell, R. J. & Alexander, D. J. (1999). Avian
influenza A subtype H9N2 in poultry in Pakistan. Vet Rec 145, 560.
Ninomiya, A., Takada, A., Okazaki, K., Shortridge, K. F. & Kida, H.
(2002). Seroepidemiological evidence of avian H4, H5, and H9
influenza A virus transmission to pigs in southeastern China. Vet
Microbiol 88, 107–114.
Peiris, M., Yuen, K. Y., Leung, C. W., Chan, K. H., Ip, P. L., Lai, R. W.,
Orr, W. K. & Shortridge, K. F. (1999). Human infection with influenza
H9N2. Lancet 354, 916–917.
Characterization of H9N2 avian influenza viruses
Peiris, J. S., Guan, Y., Markwell, D., Ghose, P., Webster, R. G. &
Shortridge, K. F. (2001). Cocirculation of avian H9N2 and
contemporary ‘‘human’’ H3N2 influenza A viruses in pigs in
southeastern China: potential for genetic reassortment? J Virol 75,
Perez, D. R., Lim, W., Seiler, J. P., Yi, G., Peiris, M., Shortridge,
K. F. & Webster, R. G. (2003). Role of quail in the interspecies
transmission of H9 influenza A viruses: molecular changes on HA
that correspond to adaptation from ducks to chickens. J Virol 77,
Perk, S., Panshin, A., Shihmanter, E., Gissin, I., Pokamunski, S.,
Pirak, M. & Lipkind, M. (2006). Ecology and molecular epidemiology
of H9N2 avian influenza viruses isolated in Israel during 2000–2004
epizootic. Dev Biol (Basel) 124, 201–209.
Reed, L. J. & Muench, H. (1938). A simple method for estimating fifty
percent endpoints. Am J Hyg 27, 493–497.
Saito, T., Lim, W., Suzuki, T., Suzuki, Y., Kida, H., Nishimura, S. I. &
Tashiro, M. (2001). Characterization of a human H9N2 influenza
virus isolated in Hong Kong. Vaccine 20, 125–133.
Shortridge, K. F., Zhou, N. N., Guan, Y., Gao, P., Ito, T., Kawaoka, Y.,
Kodihalli, S., Krauss, S., Markwell, D. & other authors (1998).
Characterization of avian H5N1 influenza viruses from poultry in
Hong Kong. Virology 252, 331–342.
Stephenson, I., Wood, J. M., Nicholson, K. G. & Zambon, M. C.
(2003). Sialic acid receptor specificity on erythrocytes affects
detection of antibody to avian influenza haemagglutinin. J Med
Virol 70, 391–398.
Suarez, D. L. (2000). Evolution of avian influenza viruses. Vet
Microbiol 74, 15–27.
Wan, H. & Perez, D. R. (2007). Amino acid 226 in the hemagglutinin
of H9N2 influenza viruses determines cell tropism and replication in
human airway epithelial cells. J Virol 81, 5181–5191.
Webby, R. J., Woolcock, P. R., Krauss, S. L., Walker, D. B., Chin, P. S.,
Shortridge, K. F. & Webster, R. G. (2003). Multiple genotypes of
nonpathogenic H6N2 influenza viruses isolated from chickens in
California. Avian Dis 47, 905–910.
Webster, R. G., Bean, W. J., Gorman, O. T., Chambers, T. M. &
Kawaoka, Y. (1992). Evolution and ecology of influenza A viruses.
Microbiol Rev 56, 152–179.
Xu, C., Fan, W., Wei, R. & Zhao, H. (2004). Isolation and identification
of swine influenza recombinant A/Swine/Shandong/1/2003(H9N2)
virus. Microbes Infect 6, 919–925.
Xu, K. M., Li, K. S., Smith, G. J. D., Li, J. W., Tai, H., Zhang, J. X.,
Webster, R. G., Peiris, J. S. M., Chen, H. & Guan, Y. (2007). Evolution
and molecular epidemiology of H9N2 influenza A viruses from quail
in southern China, 2000 to 2005. J Virol 81, 2635–2645.
P. Zhang and others
3112 Journal of General Virology 89