﻿JOURNAL OF VIROLOGY, Oct. 2007, p. 10389­10401 Vol. 81, No. 19
0022-538X/07/$08.000 doi:10.1128/JVI.00979-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
The Genesis and Evolution of H9N2 Influenza Viruses in Poultry from
Southern China, 2000 to 2005
K. M. Xu,1,2
G. J. D. Smith,1,2
J. Bahl,1,2
L. Duan,1,2
H. Tai,2
D. Vijaykrishna,1,2
J. Wang,1,2
J. X. Zhang,1,2
K. S. Li,1
X. H. Fan,2
R. G. Webster,3
H. Chen,1,2
J. S. M. Peiris,2
and Y. Guan1,2
*
International Institute of Infection and Immunity, Shantou University, Shantou, Guangdong 515031, China1
;
State Key Laboratory of Emerging Infectious Diseases, Department of Microbiology, Li Ka Shing Faculty of
Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong, SAR, China2
; and
Virology Division, Department of Infectious Diseases, St. Jude Children's Research Hospital,
Memphis, Tennessee 381053
Received 7 May 2007/Accepted 11 July 2007
H9N2 influenza viruses have become established in terrestrial poultry in different Asian countries over the
last 2 decades. Our previous study demonstrated that quail harbor increasingly diverse novel H9N2 reassor-
tants, including both Chicken/Beijing/1/94 (Ck/Bei-like) and Quail/Hong Kong/G1/97 (G1-like) viruses. How-
ever, since 1999, the genesis and evolution of H9N2 viruses in different types of poultry have not been
investigated systematically. In the present study, H9N2 viruses isolated from chickens, ducks, and other minor
poultry species were characterized genetically and antigenically. Our findings demonstrate that Ck/Bei-like
H9N2 viruses have been introduced into many different types of poultry in southern China, including quail,
partridges, chukar, pheasant, guinea fowl, and domestic ducks, while G1-like viruses were commonly detected
in quail, less frequently detected in other minor poultry species, and not detected in chickens and ducks.
Genetic analysis revealed 35 genotypes of H9N2 viruses, including 14 novel genotypes that have not been
recognized before. Our results also suggested that two-way interspecies transmission exists between different
types of poultry. Our study demonstrates that the long-term cocirculation of multiple virus lineages (e.g., H5N1
and H9N2 viruses) in different types of poultry has facilitated the frequent reassortment events that are mostly
responsible for the current great genetic diversity in H9N2 and H5N1 influenza viruses in this region. This
situation favors the emergence of influenza viruses with pandemic potential.
While highly pathogenic H5N1 influenza viruses have spread
widely throughout Eurasia and Africa, another subtype of in-
fluenza viruses, H9N2, has also become endemic in different
types of terrestrial poultry in multiple countries on the Eur-
asian continent. Epidemiological and genetic studies revealed
that three distinct lineages of H9N2 influenza viruses have
been responsible for outbreak events. These included viruses
represented by Chicken/Beijing/1/94 (Ck/Bei-like), Quail/
Hong Kong/G1/97 (G1-like), and Duck/Hong Kong/Y439/97
(Y439-like or Korean-like). Ck/Bei-like and G1-like viruses
have been prevalent mainly in China since the mid-1990s, while
the G1-like viruses have also been recorded as causing out-
breaks in chickens in the Middle East and Germany (1, 2, 4).
While Y439-like viruses were isolated from domestic ducks in
Hong Kong in 1997, similar viruses have also been identified
from disease outbreaks in chickens in South Korea since 1996
(8, 17).
H9N2 influenza viruses were detected only in domestic
ducks during influenza virus surveillance in southern China
from 1976 to 1980 (28). Since the late 1990s, Ck/Bei-like and
G1-like H9N2 viruses from southern China have become pre-
dominant in chickens and quail, respectively (11, 12). In 2000,
Ck/Bei-like viruses may have been transmitted reversely back
to domestic ducks, wherein multiple reassortant variants of
H9N2, or genotypes, were recognized in this region (19). Mo-
lecular epidemiological studies of H9N2 viruses from quail
isolated from 2000 to 2005 revealed that G1-like viruses are
still predominant in quail and are frequently reassorted with
either Ck/Bei-like or H5N1/01-like viruses to generate novel
reassortants. While only four G1-like reassortants were de-
tected in quail, there were 16 different genotypes of Ck/Bei-
like H9N2 viruses. Those H9N2 variants from quail, particu-
larly Ck/Bei-like viruses, contained gene segments from
multiple sources, including those closely related to highly
pathogenic H5N1 influenza viruses. This updated information
suggested that two-way transmission between quail and chick-
ens occurred frequently, as the genotypes of Ck/Bei-like vi-
ruses were transient and did not become established in quail
(31).
Although the quail is considered a possible intermediate
host for the introduction of influenza viruses from aquatic
birds to terrestrial poultry species (26), this species is only one
type of minor poultry species and chickens still account for
approximately 70% of the total poultry population in China. It
is noted that in the past 2 decades, a variety of birds, collec-
tively named minor poultry species, including pheasant,
chukar, partridges, guinea fowl, and pigeons, have also been
farmed. These operations dramatically increased the complex-
* Corresponding author. Mailing address: State Key Laboratory of
Emerging Infectious Disease, Department of Microbiology, The Uni-
versity of Hong Kong, Li Ka Shing Faculty of Medicine Building, 21
Sassoon Road, Pokfulam, Hong Kong, SAR, China. Phone: (852)
28199830. Fax: (852) 28171958. E-mail: yguan@hkucc.hku.hk.

Published ahead of print on 25 July 2007.
10389
ity of the influenza virus ecosystem in this region. Recent
studies suggested that most influenza A virus subtypes could
replicate asymptomatically in similar kinds of birds under lab-
oratory conditions (16, 21). However, the impact of such
changes in the poultry industry on influenza virus ecology has
not been investigated.
In addition to the quail isolates, H9N2 influenza viruses have
also been isolated regularly from both chickens and other
minor poultry species in our surveillance program in southern
China, but their genetic diversity and mechanism of genesis
have not been determined since 1999. The interrelationships of
H9N2 influenza viruses from different types of poultry are still
not determined. This situation highlights the necessity of ex-
ploring the possible role of different types of poultry in the
ecology of influenza virus in southern China, while the broad
distribution of H9N2 influenza viruses in those birds provides
such an opportunity.
Previous studies revealed that H9N2 influenza viruses from
poultry could occasionally be transmitted from poultry to
mammalian species, including humans and pigs (3, 20, 24, 25,
30). The Ck/Bei-like and G1-like viruses were initially recog-
nized from both a human and pigs in the late 1990s and were
also observed in 2003 and recently in Hong Kong (5). Genetic
analyses demonstrated that the human H9N2 influenza virus
isolate in 2003 was a novel reassortant and most likely origi-
nated directly from local live poultry markets (3). These recent
interspecies transmission events suggest that current H9N2
influenza virus variants are still potentially infectious for hu-
mans.
Our long-term influenza virus surveillance program in south-
ern China focused mainly on major poultry, including chickens,
ducks, and geese. In addition, we also sampled a variety of
other minor poultry species. In the present study, H9N2 influ-
enza viruses isolated from chickens, domestic ducks, and other
minor poultry species from 2000 to 2005 were characterized
genetically and antigenically. The findings of the present study
demonstrate that Ck/Bei-like H9N2 viruses have been intro-
duced into many different types of poultry in this region, in-
cluding quail, partridges, chukar, pheasant, guinea fowl, and
domestic ducks, while G1-like viruses were commonly detected
in quail, less frequently detected in other minor poultry spe-
cies, and not detected in chickens and ducks. Genetic studies
revealed that two-way interspecies transmission exists between
different types of poultry. Phylogenetic analysis suggests that
the long-term cocirculation of multiple virus lineages (e.g.,
H5N1 and H9N2 viruses) in different types of poultry facili-
tated frequent reassortment events that were mostly responsi-
ble for the current great genetic diversity in H9N2 and H5N1
variants in this region (6, 10, 11, 18). The present study pro-
vides insight into the genesis and evolution of H9N2 influenza
viruses in southern China. The current influenza virus ecosys-
tem in southern China favors the emergence of influenza vi-
ruses with pandemic potential.
MATERIALS AND METHODS
Sampling and virus isolation. Our previous studies characterized H9N2 vi-
ruses from quail, which was considered a "minor poultry species." Therefore,
minor poultry species in the present study, namely, chukar, guinea fowl, par-
tridges, and pheasant, are referred to as "other minor poultry species." A total of
47,255 chickens and 6,925 birds of other minor poultry species were sampled
from six provinces in southern China between July 2000 and December 2005.
During the same period, a total of 49,150 ducks were also sampled. Of those
samples, 20,535 chicken, 3,008 other minor poultry species, and 5,381 duck
specimens were paired tracheal and cloacal swabs, while the remaining speci-
mens were either cloacal or fecal swabs. Our influenza virus surveillance program
was carried out as previously reported (30). In brief, surveillance was initiated in
Guangdong in July 2000. Since 2002, this program was gradually expanded to five
other provinces in this region, including Fujian, Guangxi, Guizhou, Hunan, and
Yunnan. Viruses were isolated in 9- to 11-day-old embryonated chicken eggs as
previously described (18, 19).
Antigenic analysis. All virus isolates were subtyped by standard hemaggluti-
nation inhibition (HI) and neuraminidase (NA) inhibition tests, using a panel of
World Health Organization reference antisera as previously described (3). An-
tigenic analysis was performed using three different panels of monoclonal anti-
bodies (MAbs), against Qa/HK/G1/97, Dk/HK/Y280/97, and Ck/HK/G9/97, as
previously described (31). All MAbs were produced at the Department of In-
fectious Diseases, St. Jude Children's Research Hospital, TN (7). Numerical
analysis of HI titers was conducted using PRIMER, version 5.2.9 (PRIMER-E,
Plymouth, United Kingdom), also as previously described (31).
Phylogenetic and molecular analyses. One or two virus isolates from each
positive sampling occasion were selected for characterization. Viral RNA extrac-
tion, cDNA synthesis, PCR, and sequencing were carried out as previously
described (12). All eight gene segments sequenced from these viruses were
characterized and phylogenetically analyzed with available virus sequence data
from GenBank. The program MrModeltest 2.2 (23) was used to determine the
appropriate DNA substitution model and -rate heterogeneity. The generated
model was used in all subsequent analyses. Neighbor-joining trees were con-
structed using PAUP* 4.0 (29), and Bayesian analysis was conducted with Mr-
Bayes 3.1 (15) by using two replicates of 1 million generations with six chains,
sampling every 100 generations. Estimates of the phylogenies were calculated by
performing 1,000 neighbor-joining bootstrap replicates, and Bayesian posterior
probabilities were calculated from the consensus of 18,000 trees after excluding
the first 2,000 trees as burn-in.
Genotype definition. Virus genotypes were defined by gene phylogeny. A
distinct phylogenetic lineage with bootstrap support of 80% indicated a com-
mon origin. Viruses with G1-like and Ck/Bei-like hemagglutinin (HA) genes
were designated genotype A and B series, respectively, as previously described
(31).
Nucleotide sequence accession numbers. The nucleotide sequences obtained
in this study are available from GenBank under accession numbers CY023090 to
CY024737.
RESULTS
Prevalence of H9N2 influenza viruses in chickens, ducks,
and other minor poultry species. A total of 103,300 samples
were collected from chickens, ducks, and other minor poultry
species from 2000 to 2005. In general, H9N2 viruses were
isolated from apparently healthy birds year-round in live poul-
try markets in southern China, but the isolation rate was usu-
ally higher in the winter season than in the summer season
(Fig. 1; Table 1). The majority of viruses were isolated from
chickens and other minor poultry species, while ducks tested
positive on very few sampling occasions, except for those in
2000 and 2001 (Fig. 1).
There were 1,189 strains of H9N2 influenza viruses isolated
from 47,225 chicken samples (overall isolation rate, 2.5%)
(Table 1). Of these viruses, 944 were isolated from 20,535
tracheal swabs (isolation rate, 4.6%), while 245 strains were
isolated from 26,690 cloacal or fecal swabs (isolation rate,
0.9%) (Table 2). It is noteworthy that the rate of isolation of
H9N2 viruses from fecal material and cloacal samples declined
from 5.4% in 2000 to 0.5% in 2005, while there was no dra-
matic change in the isolation rate for tracheal samples from
chickens during the study period (Table 2). This suggests that
H9N2 influenza viruses have gradually adapted to replicate in
the respiratory tract of chickens.
10390 XU ET AL. J. VIROL.
A total of 89 H9N2 viruses were isolated from 49,150 duck
samples (overall isolation rate, 0.18%) (Table 1). Fifty-one of
these H9N2 viruses were isolated from 2,959 specimens (iso-
lation rate, 1.7%) during 2000 to 2001, while only 38 viruses
were isolated from 46,191 cloacal or fecal swabs (isolation rate,
0.08%) from 2002 to 2005 (Fig. 1; Table 1). While tracheal
swabs were collected only in 2005, only a single H9N2 virus was
isolated from 5,381 samples, indicating that the major site of
H9N2 influenza virus replication in the duck is the intestine.
The isolation rates of H9N2 influenza viruses from duck have
declined since 2002, with ducks testing positive infrequently
(Fig. 1; Table 1).
Three hundred twenty-eight H9N2 influenza viruses were
isolated from a total of 6,925 samples collected from other
minor poultry species (overall isolation rate, 4.7%) (Table 1).
Of these viruses, 292 were isolated from tracheal swabs, while
only 36 strains were isolated from cloacal or fecal swabs (Table
2). There is a clear replication pattern where H9N2 viruses
replicate mainly in the respiratory tract, not the intestine, in
other minor poultry species.
Antigenic analysis. The antigenic properties of representa-
tive H9N2 influenza viruses were investigated with a panel of
MAbs raised against Qa/HK/G1/97, Dk/HK/Y280/97, and Ck/
HK/G9/97 by HI assay (Table 3). Numerical analysis of HI
titers was conducted to visualize the antigenic variation and
revealed three distinct groups that were in agreement with the
results of phylogenetic analyses (Fig. 2; see below). One group
included two viruses from other minor poultry species, Gf/ST/
2076/01 and Cu/ST/22116/05, which reacted well with two Qa/
HK/G1/97 MAbs and both Ck/HK/G9/97 MAbs, a reaction
pattern similar to that of Qa/HK/G1/97. The remaining two
groups contained Ck/Bei-like viruses (subgroups 1 and 2). Ck/
Bei-like subgroup 2 viruses were isolated from both chickens
and other minor poultry species and reacted well with all tested
MAbs, except G1-9 and 1073-9, a pattern similar to that of the
prototype virus, Dk/HK/Y280/97 (Fig. 2; Table 3). Ck/Bei-like
FIG. 1. Comparison of H9N2 influenza virus isolation rates from chickens (a), other minor poultry species (b), and ducks (c) from southern
China, July 2000 to December 2005. Surveillance was conducted in live poultry markets in Fujian, Guangdong, Guangxi, Guiyang, Hunan, and
Yunnan Provinces. *, positive sampling occasions with low isolation rates of 0.3%.
VOL. 81, 2007 GENESIS AND EVOLUTION OF H9N2 INFLUENZA VIRUS 10391
subgroup 1 viruses were mostly isolated from other minor
poultry species and had high HI titers against MAb Y280-8C4
and only moderate reactivity to Y280-18B10, a reactivity pat-
tern similar to that of the recent human H9N2 isolate (HK/
2108/03) and a wild duck isolate (WDk/ST/4108/01) (Fig. 2;
Table 3).
Phylogenetic analysis of surface genes. To better understand
the evolutionary pathway of H9N2 viruses in southern China,
112 of 1,189 (9.4%) viruses from chickens and 79 of 328 (9.4%)
viruses from other minor poultry species were sequenced. A
further 15 of 79 (19%) H9N2 viruses from ducks, isolated from
2001 to 2005, were also sequenced. At least one isolate from
each positive sampling occasion was sequenced. Those se-
quence data were analyzed phylogenetically together with data
available in public databases. Phylogenetic analysis of the H9
HA gene showed that the majority of isolates tested belonged
to the Ck/Bei-like lineage. Three isolates from other minor
poultry species (Gf/ST/2076/01, Pa/ST/2875/01, and Cu/ST/
22116/05) clustered within the G1-like lineage, and two isolates
from ducks (Dk/ST/163/04 and Dk/ST/7448/04) had a Korean-
like HA gene (Fig. 3a). The Ck/Bei-like lineage contained two
subgroups, including subgroup 1, represented by Qa/ST/243/
00, and subgroup 2, represented by Dk/HK/Y280/97, as de-
scribed in our previous study (31). All H9N2 viruses from
chickens belonged to subgroup 2, with the exception of three
viruses (SCk/ST/473/04, SCk/ST/999/04, and Ck/ST/6786/04)
that clustered in subgroup 1 (Fig. 3a). Similarly, the HA genes
of all Ck/Bei-like duck viruses fell into subgroup 2, except for
a single virus (Dk/ST/3658/03) that belonged to subgroup 1. In
comparison, 47 H9N2 viruses from other minor poultry species
clustered into subgroup 1, while 29 viruses belonged to sub-
group 2 (Fig. 3a).
Phylogenetic analysis of the NA gene showed a similar evo-
lutionary pattern to that of the HA gene, wherein all viruses
clustered within the Ck/Bei-like lineage, except for three vi-
ruses from other minor poultry species and two viruses from
ducks, which clustered in the G1-like and Korean-like lineages,
respectively (Fig. 3b). These results show that Ck/Bei-like vi-
ruses are predominant in chickens, ducks, and other minor
poultry species, in comparison to previous results that indi-
cated that G1-like viruses were maintained mainly in quail
(31).
Phylogenetic analysis of internal genes. Phylogenetic anal-
ysis of the ribonucleoprotein complex genes (PB2, PB1, PA,
and NP genes) revealed that these genes had more diversified
TABLE
1.
Prevalence
of
H9N2
viruses
from
chickens,
other
minor
poultry
species,
and
ducks
in
southern
China
during
2000
to
2005
Month
No.
of
H9N2
isolates/total
sample
no.
Chickens
Other
minor
poultry
species
Ducks
2000
2001
2002
2003
2004
2005
2000
2001
2002
2003
2004
2005
2000
2001
2002
2003
2004
2005
January
1/161
5/70
10/844
13/817
110/2,056
4/43
33/79
1/42
29/224
9/254
0/502
1/200
0/769
3/587
2/1,380
February
15/215
4/8
2/440
10/666
46/1,432
0/11
9/36
2/52
6/176
1/150
0/191
0/162
0/428
0/361
1/1,035
March
3/182
0/35
2/689
11/834
26/1,651
0/15
3/39
1/45
10/308
0/162
2/149
0/102
0/681
0/696
0/1,396
April
13/178
0/49
1/603
11/732
84/2,204
0/5
0/47
0/47
0/206
0/138
2/118
0/108
0/561
0/695
1/1,972
May
0/176
0/48
2/652
1/757
67/2,023
1/6
0/36
0/19
0/181
0/150
14/121
0/114
0/576
0/635
0/2,153
June
1/162
11/256
0/604
0/733
31/1,931
3/8
0/19
0/49
0/220
0/204
5/115
0/125
0/533
0/663
1/2,089
July
14/203
2/181
0/123
0/646
0/843
22/1,881
8/60
1/11
0/18
0/40
1/207
2/174
0/119
1/135
0/341
0/567
0/818
0/2,426
August
0/115
1/154
0/115
3/584
6/1,071
27/1,803
0/35
0/19
0/49
3/85
0/266
1/188
0/100
1/168
0/152
0/494
0/994
2/3,034
September
5/172
2/139
0/132
1/707
13/922
75/1,725
0/89
5/28
0/43
0/99
16/241
0/160
4/131
8/211
0/190
0/605
0/952
6/2,214
October
2/242
6/177
2/537
3/725
8/994
119/1,755
0/46
5/27
0/79
4/68
39/244
3/256
0/58
3/205
1/339
6/595
0/1,051
6/2,574
November
26/199
10/154
28/860
10/699
45/1,360
41/1,679
8/36
6/27
6/36
14/134
21/274
0/164
2/131
3/210
4/889
2/627
1/1,241
0/2,670
December
10/106
21/181
12/685
32/899
37/1,738
126/1,702
5/38
5/71
3/33
33/227
15/212
12/170
4/120
4/175
1/625
0/765
1/1,097
0/2,910
Total
57/1,037
75/2,060
62/2,918
66/8,092
155/11,467
774/21,842
21/304
30/271
54/514
58/907
137/2,759
28/2,170
10/659
41/2,300
6/3,347
8/7,201
5/9,790
19/25,853
a
The
isolation
rates
for
chickens
in
2000
to
2005
were
5.5%,
3.6%,
2.1%,
0.8%,
1.4%,
and
3.5%,
respectively.
The
isolation
rates
for
other
minor
poultry
species
in
2000
to
2005
were
6.9%,
11.1%,
10.5%,
6.4%,
5.0%,
and
1.3%,
respectively.
The
isolation
rates
for
ducks
in
2000
to
2005
were
1.8%,
1.8%,
0.17%,
0.12%,
0.06%,
and
0.07%,
respectively.
TABLE 2. Comparison of replication sites of H9N2 viruses from
chickens and other minor poultry species
Year
No. of H9N2 isolates/total sample no. (isolation rate %)
Chickens Other minor poultry species
Tracheal swabs
Fecal and
cloacal swabs
Tracheal
swabs
Fecal and
cloacal swabs
2000 4/64 (6.3) 53/973 (5.4) 20/60 (33.3) 1/244 (0.4)
2001 4/41 (9.8) 71/2,019 (3.5) 29/126 (23) 1/145 (0.7)
2002 28/839 (3.3) 34/2,079 (1.7) 48/237 (20.3) 6/277 (2.2)
2003 37/3,520 (1.0) 29/4,572 (0.7) 34/116 (29.3) 24/791 (3.0)
2004 147/4,985 (3.0) 8/6,482 (0.1) 134/1,366 (9.8) 3/1,393 (0.2)
2005 724/11,086 (6.5) 50/10,756 (0.5) 27/1,103 (2.4) 1/1,067 (0.1)
Total 944/20,535 (4.6) 245/26,690 (0.9) 292/3,008 (9.7) 36/3,917 (0.9)
10392 XU ET AL. J. VIROL.
sources than the surface genes and that H9N2 viruses circu-
lating in chickens, ducks, and other minor poultry species had
undergone extensive reassortment to generate multiple novel
reassortants or genotypes (Fig. 4a to d; Table 4). Analysis of
the PB2 gene revealed seven distinct evolutionary lineages,
including G1-like (n  66), Ck/Bei-like (n  3), H5N1/01-like
(n  1), and Gs/Gd-like (n  1) lineages, an unknown avian
source (n  77), and two duck lineages (for Dk1, n  2; and for
Dk2, n  56) (Fig. 4a).
The PB2 genes of the majority of chicken isolates fell into
two groups, namely, unknown avian, for which we could not
identify a source, and Dk2 (Fig. 4a). The PB2 genes of the Dk2
lineage are most closely related to an H9N2 virus, Dk/Hok-
kaido/9/99, an isolate possibly obtained from a migratory duck.
Interestingly, one chicken isolate (Ck/GX/521/05) was closely
related to an H5N1 genotype G virus (Gs/GX/345/05), with a
Gs/GD-like PB2 gene (Fig. 4a). The majority of the H9N2
viruses from other minor poultry species had G1-like PB2
genes, while duck H9N2 isolates had PB2 genes from many
different lineages, including G1-like, Dk2, and unknown avian
(Fig. 4a; data not shown). However, two of them (Dk/ST/
163/03 and Dk/ST/7488/04) contained a PB2 gene from the
domestic duck gene pool (Dk1) in this region (Fig. 4a).
Phylogenetic analysis of the PB1 gene showed that the H9N2
viruses formed seven different lineages, including G1-like (n 
66), Ck/Bei-like (n  61), and H5N1/01-like (n  4) lineages,
three duck lineages (for Dk1, n  1; for Dk2, n  53; and for
Dk3, n  1), and an unknown avian lineage (n  20) (Fig. 4b).
The PB1 genes of the majority of chicken isolates fell into the
Ck/Bei-like and Dk2 lineages, while those of other minor poul-
try species isolates were mostly G1-like and those of duck
viruses were from diverse sources. The Dk2 PB1 genes were
closely related to those of viruses isolated from migratory
ducks or sentry ducks in southern China and Japan (e.g., Dk/
ST/2030/00 [H9N1] and Dk/Hokkaido/120/01 [H6N2]). It was
noted that the PB1 genes of Dk/ST/163/03 and Dk/ST/7488/04,
both of which have Korean-like HA and NA genes, clustered
into different lineages, namely, Dk1 and Dk3 (Fig. 4b).
Six different PA lineages for the H9N2 influenza viruses
tested were recognized, including G1-like (n  4), Ck/Bei-like
(n  10), and H5N1/01-like (n  123) lineages and three duck
lineages (for Dk1, n  11; for Dk2, n  1; and for Dk3, n  57)
(Fig. 4c). Most H9N2 viruses isolated from chickens, ducks,
and minor poultry species since 2002 and 2003 had H5N1/01-
like PA genes related to that of a contemporary duck virus
(Dk/HK/Y439/97 [H9N2]). It is interesting that Gf/HK/38/02
(H5N1; genotype X), together with 10 H9N2 chicken isolates
and 1 duck isolate, clustered with the Dk1 lineage (Fig. 4c).
TABLE 3. HI titers from antigenic analysis of influenza A H9N2 viruses
Virus Genotype
HI titera
Qa/HK/G1/97 Dk/HK/Y280/97 Ck/HK/G9/97
1073-9 26 29 7B10 8C4 15F1 18G4 3D11 4G3 19A10 18B10 2F4 18B1 G9-6 G9-25
Qa/HK/G1/97 A0 200   100 100 100 100 100 100 100 100 100 100  
HK/1073/99 A0 100   100 100 100 100 100 100 100 100 100 100  
Dk/HK/Y280/97 B0 100 400         3,200    
Ck/HK/G9/97 Bn 400 800         6,400    
HK/2108/03 B7 100 100 100 200  100 100 100 100 100 1,600 100 100 100 100
WDk/ST/4808/01 B7 100 100 100 200  100 100 100 100 100 3,200 100 100 100 100
Gf/ST/2076/01 A0 100   100 100 100 100 100 100 100 200 100 100  
Cu/ST/22116/05 A4 3,200   100 100 100 100 100 100 100 100 100 100  
Pa/ST/5692/00 B2 800 400       12,800  12,800    
Pa/ST/4525/02 B3 100 800         6,400    
Ph/ST/4340/03 B23 800 1,600             
Ph/ST/5588/04 B14 200 800             
Ph/ST/111/05 B3 200 800             
Ck/ST/94/00 B3 200 800             
Ck/ST/1126/01 B0 100 800             
Ck/ST/1890/01 B3 800 1,600             
Ck/HN/774/02 B17 100 800         6,400    
Ck/YN/401/02 B18 200 800         6,400    
Ck/ST/2204/02 B3 800 1,600         6,400    
SCk/ST/1404/03 B19 800 1,600         6,400    
SCk/ST/4104/03 B22 100 1,600         6,400    
Ck/GX/1857/04 B25 100 800             
Ck/ST/19465/05 B29 800 800             
Ck/GX/521/05 B26 100 800             
Ck/HN/4444/05 B28 800 800             
Ck/FJ/3080/05 B14 1,600 1,600         12,800    
Pa/ST/1800/00 B1 800 100 100 800  100 100 100 100 100 12,800 100 100 100 100
Pa/ST/4541/01 B7 100 100 100 400  100 100 100 100 100 3,200 100 100 100 100
Cu/ST/338/02 B8 100 100 100 400  100 100 100 100 100 3,200 100 100 100 100
Ph/ST/3530/03 B7 100 100 100 400  100 100 100 100 100 3,200 100 100 100 100
Cu/ST/290/04 B7 3,200 100 100 400  100 100 100 100 100 6,400 100 100 100 100
Ph/ST/17033/05 B16 200 100 100 400  100 100 200 100 100 6,400 100 100 100 100
SCk/ST/473/04 B7 1,600 100 100 400  100 100 100 100 100 6,400 100 100 100 100
a
, HI titer of 12,800. The HI assay was started at a 1:100 dilution.
VOL. 81, 2007 GENESIS AND EVOLUTION OF H9N2 INFLUENZA VIRUS 10393
The nucleoprotein (NP) genes of these H9N2 viruses
formed five lineages, including G1-like (n  14), Ck/Bei-like
(n  50), H5N1/01-like (n  85), duck (n  2), and aquatic bird
(n  55) lineages (Fig. 4d). It was noted that the NP genes of
many H9N2 viruses tested apparently were derived from an
H6N9 virus (Gs/HK/W217/97), and therefore, this lineage was
assigned as an aquatic bird lineage. The majority of chicken
and duck H9N2 isolates had NP genes belonging to either the
H5N1/01-like or aquatic bird lineage, while viruses from other
minor poultry species were from the Ck/Bei-like or H5N1/01-
like lineage. It is interesting that one H5N1 virus isolated in
1997 from central China (Ck/Hubei/wm/97) also grouped with
the aquatic bird lineage.
The matrix (M) and nonstructural (NS) protein genes
showed much less diversity than the other genes and be-
longed to either the G1-like or Ck/Bei-like lineage (Fig. 4e
and f). Thirty-two chicken, 2 duck, and 55 minor poultry
species isolates had G1-like M genes, while 80 chicken, 11
FIG. 2. Numerical analysis of HI titers (Table 3) by nonmetric multidimensional ordination in two dimensions (a) and by using hierarchical
agglomerative clustering (b).
10394 XU ET AL. J. VIROL.
duck, and 24 minor poultry species isolates contained Ck/
Bei-like M genes. For the NS gene, all of these H9N2 viruses
had a Ck/Bei-like gene segment, except for three viruses
(Gf/ST/2076/01, Pa/ST/2875/01, and Cu/ST/22116/05) whose
genes belonged to the G1-like lineage, similar to the results
for the HA gene. It is interesting that two duck H9N2
viruses (Dk/ST/163/04 and Dk/ST/7488/04) had both M and
NS genes that always clustered with those of viruses from
the gene pool in this region, including the Korean-like
H9N2 virus.
Genotyping. All H9N2 viruses from chickens and 76 of 79
viruses from minor poultry species belonged to the Ck/Bei-like
lineage (genotype B series), while only 3 viruses from other
minor poultry species belonged to the G1-like lineage (geno-
type A series). Except for two duck isolates that were closely
related to the Korean-like H9N2 virus and were of pure duck
origin for each gene segment, the remaining duck H9N2 iso-
lates were all genotype B series (Table 4). Thus, a total of 35
genotypes of H9N2 influenza viruses were identified from dif-
ferent types of poultry under our surveillance, including 14
novel genotypes that were not recognized in our previous study
(31) and were designated genotypes B17 to B29 and genotype
A4 (Fig. 5; Table 4). Nineteen genotypes were identified from
chickens, including the progenitor genotype B0 in 2001, and 16
additional genotypes were detected from other minor poultry
species (Table 4). These novel genotypes were all triple or even
FIG. 3. Phylogenetic relationships of HA (a) and NA (b) genes of representative influenza A viruses isolated in Asia. Trees were generated by
the neighbor-joining method in the PAUP* program. Numbers above and below branches indicate neighbor-joining bootstrap values and Bayesian
posterior probabilities, respectively. Analysis was based on nucleotides 129 to 1042 of the HA gene and 231 to 1297 of the NA gene. The HA and
NA trees were rooted to Qa/Arkansas/29209-1/93 (H9N2) and Ck/Pennsylvania/8125/83 (H5N2), respectively. Viruses characterized in this study
are highlighted in green. Genotypes characterized in this study are shown in parentheses and defined in Table 4. Abbreviations: BJ and Bei, Beijing;
Ck, chicken; Dk, duck; FJ, Fujian; GD, Guangdong; Gf, guinea fowl; GX, Guangxi; HK, Hong Kong; HLJ, Heilongjiang; HN, Hunan; NC,
Nanchang; Pg, pigeon; Ph, pheasant; Qa, quail; SCk, silky chicken; SD, Shandong; SH, Shanghai; ST, Shantou; Ty, turkey; WDk, wild duck; YN,
Yunnan. Bar, 0.01 substitution per site.
VOL. 81, 2007 GENESIS AND EVOLUTION OF H9N2 INFLUENZA VIRUS 10395
FIG. 4. Phylogenetic relationships of the PB2 (a), PB1 (b), PA (c), NP (d), M (e), and NS (f) genes of representative influenza A viruses isolated
in Asia. Trees were generated by the neighbor-joining method in PAUP*. Numbers above and below branches indicate neighbor-joining bootstrap
values and Bayesian posterior probabilities, respectively. Analysis was based on the following nucleotides: PB2, 1079 to 2138; PB1, 42 to 1217; PA,
1429 to 2127; NP, 31 to 917; M, 49 to 864; and NS, 88 to 815. The PB2, PA, NP, and M trees were rooted to A/equine/Prague/1/56 (H7N7), the
PB1 tree was rooted to Qa/Arkansas/29209-1/93 (H9N2), and the NS tree was rooted to A/swine/Hong Kong/168/93 (H1N1). Viruses characterized
in this study are highlighted in green. Aq, aquatic bird. Other virus names and abbreviations can be found in the legend to Fig. 3. Bar, 0.01
substitution per site.
10396 XU ET AL. J. VIROL.
quadruple reassortants, with gene segments from Ck/Bei-like,
G1-like, aquatic bird, duck, and H5N1/01-like viruses (Fig. 4
and 5; Table 4).
It is noted that genotype B3 viruses have been detected in
chickens and other minor poultry species every year since they
appeared in 2000, suggesting that this subtype of viruses has
become established in these birds. Genotype B7 viruses were
initially detected in other minor poultry species and quail from
2001 onwards and failed to be detected in quail in 2005 (31),
being detected only in chickens in 2004­2005, suggesting that
the interspecies transmission direction of this genotype was
from quail or other minor poultry species to chickens. The
remaining genotypes, genotype B17 to B29, were only occa-
sionally and transiently detected in chickens or minor poultry
species. It has been noted that the genotype number being
recognized in both chickens and other minor poultry species
has increased since 2003, suggesting that H9N2 influenza
viruses have become more and more diversified in this re-
gion (see Table 6). However, the number of genotypes de-
tected in ducks has decreased since 2003, which correlates
with the decreased number of isolates detected from ducks,
suggesting that the Ck/Bei-like variants are not adaptive in
domestic ducks.
Of the three G1-like viruses, two were nonreassortant vi-
ruses (genotype A0), while the other was a novel reassortant
with a PB1 gene of unknown source that was designated ge-
notype A4, in contrast to genotype A3 viruses. This novel PB1
gene had also been detected in Ck/Bei-like H9N2 influenza
viruses from minor poultry species but not from quail, suggest-
ing that genotype A4 viruses might be generated directly within
other minor poultry species (Table 4).
Molecular characterization. The deduced amino acid se-
quences of the viruses were aligned and compared with
those of other representative H9N2 viruses in this region.
Except for six viruses from other minor poultry species, the
HAs of all viruses tested had 226Leu at receptor binding
sites (H3 numbering), as recognized in our previous studies,
while those of the other six viruses had Glu at position 226
(Table 5) (22). Other substitutions related to receptor bind-
ing sites have not been recognized. Most H9N2 viruses an-
alyzed maintained an Arg-Ser-Ser-Arg (R-S-S-R) motif at
the connecting peptide of their HA, but a few substitutions
were observed at each site of the connecting peptide (Table
5). However, no additional basic amino acids were found in
all tested viruses. Thirty-one of the Ck/Bei-like H9N2 vi-
ruses had the same three-amino-acid deletion (positions 62
TABLE 4. Gene constellations of different genotypes of H9N2 influenza viruses and their host distribution in southern China
Genotype Host (no. of viruses)a
Lineage of gene segment
PB2 PB1 PA HA NP NA M NS
A0 Qa (9), MP (2) G1 G1 G1 G1 G1 G1 G1 G1
A1 Qa (4) G1 G1 G1 G1 G1 G1 G1 Ck/Bei
A2 Qa (1) G1 G1 G1 G1 G1 G1 Ck/Bei Ck/Bei
A3 Qa (19) G1 G1 H5N1/01 G1 G1 G1 G1 G1
A4 MP (1) G1 ? H5N1/01 G1 G1 G1 G1 G1
B0 Ck (3) Ck/Bei Ck/Bei Ck/Bei Ck/Bei Ck/Bei Ck/Bei Ck/Bei Ck/Bei
B1 Qa (1), MP (2) G1 G1 G1 Ck/Bei Ck/Bei Ck/Bei G1 Ck/Bei
B2 Qa (1), MP (3) Dk Dk Dk Ck/Bei Aq Ck/Bei G1 Ck/Bei
B3 Qa (3), Ck (33), MP (8), Dk (7) Dk Dk Dk Ck/Bei Aq Ck/Bei Ck/Bei Ck/Bei
B4 Qa (1), Ck (2), MP (1) Dk Ck/Bei Dk Ck/Bei Aq Ck/Bei Ck/Bei Ck/Bei
B5 Qa (3), MP (2) G1 G1 Dk Ck/Bei Ck/Bei Ck/Bei G1 Ck/Bei
B6 Qa (1) G1 G1 G1 Ck/Bei G1 G1 G1 Ck/Bei
B7 Qa (9), Ck (5), MP (36), Dk (1) G1 G1 H5N1/01 Ck/Bei Ck/Bei Ck/Bei G1 Ck/Bei
B8 Qa (6), MP (3) G1 G1 H5N1/01 Ck/Bei H5N1/01 Ck/Bei G1 Ck/Bei
B9 Qa (1) H5N1/01 G1 H5N1/01 Ck/Bei H5N1/01 Ck/Bei G1 Ck/Bei
B10 Qa (1) Dk Dk H5N1/01 Ck/Bei Aq Ck/Bei G1 Ck/Bei
B11 Qa (1) G1 G1 H5N1/01 Ck/Bei Ck/Bei Ck/Bei Ck/Bei Ck/Bei
B12 Qa (1), Ck (1), MP (1), Dk (1) G1 G1 H5N1/01 Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B13 Qa (1) G1 G1 H5N1/01 Ck/Bei G1 Ck/Bei Ck/Bei Ck/Bei
B14 Qa (3), Ck (19), MP (9) ? Ck/Bei H5N1/01 Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B15 Qa (5), Ck (22), MP (3), Dk (1) ? Ck/Bei H5N1/01 Ck/Bei H5N1/01 Ck/Bei G1 Ck/Bei
B16 Qa (2), Ck (1), MP (3) G1 G1 H5N1/01 Ck/Bei G1 Ck/Bei G1 Ck/Bei
B17 Ck (1) ? ? Dk Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B18 Ck (7) G1 G1 Ck/Bei Ck/Bei G1 Ck/Bei Ck/Bei Ck/Bei
B19 Ck (1) ? Ck/Bei Dk Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B20 Ck (2) Dk Dk Dk Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B21 Ck (5), MP (3), Dk (3), ? ? H5N1/01 Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B22 Ck (1) ? ? H5N1/01 Ck/Bei Aq Ck/Bei Ck/Bei Ck/Bei
B23 MP (1) Dk ? H5N1/01 Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B24 MP (1) H5N1/01 G1 H5N1/01 Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B25 Ck (3) ? H5N1/01 Dk Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B26 Ck (1) Gs/GD H5N1/01 Dk Ck/Bei H5N1/01 Ck/Bei Ck/Bei Ck/Bei
B27 Ck (1) ? ? Dk Ck/Bei Ck/Bei Ck/Bei Ck/Bei Ck/Bei
B28 Ck (3) ? ? Dk Ck/Bei H5N1/01 Ck/Bei G1 Ck/Bei
B29 Ck (1) ? Dk H5N1/01 Ck/Bei H5N1/01 Ck/Bei G1 Ck/Bei
a
Number of viruses characterized in this study and in reference 31. Abbreviations: ?, unknown avian host or lineage; Aq, aquatic bird; Ck, chicken; Ck/Bei,
Ck/Beijing/1/94-like; Dk, duck; G1, Qa/HK/G1/97-like; Gs/GD, Gs/GD-like; H5N1/01, H5N1/01-like; MP, minor poultry species, except for quail; Qa, quail.
VOL. 81, 2007 GENESIS AND EVOLUTION OF H9N2 INFLUENZA VIRUS 10397
to 64) at the NA stalk region, as previously recognized (11,
13), while one G1-like virus had a two-amino-acid deletion
(positions 38 and 39) which was also previously recognized
in G1-like viruses from quail (11, 12, 31).
An Arg292Lys mutation in the NA gene that has been as-
sociated with oseltamivir resistance was not detected in any of
the H9N2 viruses tested (14). However, 11 viruses had Asn at
residue 31 of the M2 protein, which is responsible for aman-
FIG. 5. Genotypes of H9N2 influenza viruses of chickens and other minor poultry species in southern China. The figure shows progenitors of
H9N2 influenza virus genotype A and B series and internal gene donors (a) and genotypes from other minor poultry species (b) and chickens (c)
in southern China. Dashed lines represent transient and short-lived genotypes. Details of transient genotypes are given in Table 6. The eight gene
segments (horizontal bars starting from the top) are PB2, PB1, PA, HA, NP, NA, M, and NS. Each color represents a virus lineage. Genotype
definitions are described in Materials and Methods. Abbreviations are listed in Table 4.
TABLE 5. Comparison of amino acid sequences of HA, NA, and M2 genes of representative viruses from southern China
Virus Genotype
Residue at RBSa
NA
deletion (aa)
Connecting
peptidec
M2 residue at
amantadine
resistance
mutation position
226 228 27 31
Qa/HK/G1/97 A0 L G 38­39 R-S-S-R V S
Dk/HK/Y280/97 B0 L G 62­64 R-S-S-R V S
Dk/HK/Y439/97 Korean Q G A-S-N-R V S
Dk/HK/289/78 Q G A-S-N-R
Dk/ST/163/04 Korean Q G A-S-D-R V S
Dk/ST/7448/04 Korean Q G A-S-G-R V S
Cu/ST/22116/05 A4 Q G 38­39 R-S-S-R V S
Pa/ST/2063/00 B1 Qb
G R-S-S-R V N
Pa/ST/24/00 B5 L G R-S-S-R V S
Gf/ST/1677/00 B3 L G 62­64 R-S-S-R V S
Cu/ST/338/02 B8 L G R-S-S-R V S
Pa/ST/4525/02 B3 L G R-S-S-R V S
Ph/ST/443/03 B7 L G K-S-S-R V S
Ck/ST/94/00 B3 L G 62­64 R-S-S-R V S
Ck/ST/1579/00 B3 L G 62­64 R-S-S-R V N
Ck/ST/4608/02 B3 L G R-S-S-R A S
Ck/HN/774/02 B17 L G 62­64 R-L-S-R V G
Ck/ST/3040/03 B21 L G R-S-I-R V S
Ck/GX/1857/04 B25 L G 62­64 R-A-S-R V S
Ck/GX/187/05 B25 L G 62­64 R-A-S-K V S
Ck/ST/22504/05 B14 L G R-S-S-R V N
a
RBS, receptor binding site.
b
Representative of six viruses from other minor poultry species.
c
Connecting peptide from positions 4 to 1 of HA1. Italics represent newly identified motifs.
10398 XU ET AL. J. VIROL.
tadine resistance of influenza viruses (1). Four of 112 H9N2
influenza viruses from chickens had a Val27Ala mutation,
which is also associated with amantadine resistance of influ-
enza virus (1). One virus, Ck/HN/774/02, had a Ser31Gly mu-
tation which had not been observed before. Three G1-like
viruses from minor poultry species had Glu at position 92 of
the NS1 protein, a mutation related to the pathogenicity of
H5N1 influenza virus in pigs, but the remaining 203 viruses had
Asp, a residue typically observed in avian influenza viruses, at
this position (27).
DISCUSSION
Since the late 1990s, two distinct H9N2 virus lineages have
become established in chickens and quail in southern China
(11­13). In the present study, we genetically and antigenically
characterized H9N2 viruses isolated from chickens, ducks, and
other minor poultry species in our surveillance from 2000 to
2005. All H9N2 viruses from chickens and most of the viruses
from ducks and other minor poultry species belonged to the
Ck/Bei-like lineage, while G1-like viruses prevailed mainly in
quail and were rarely detected in other minor poultry species.
Phylogenetic studies revealed that two-way interspecies trans-
mission occurred between different types of poultry and that
reassortment events among established virus lineages (e.g.,
H5N1 and H9N2 viruses) were mostly responsible for the cur-
rent great genetic diversity in H9N2 and H5N1 variants in this
region.
Even though many novel reassortants have been generated,
relative host restriction is present for H9N2 viruses. Quail are
able to harbor both Ck/Bei-like and G1-like viruses and may
play a central role in the current ecosystem of southern China
(Fig. 6), while chickens could support only Ck/Bei-like viruses,
but other minor poultry species seem to be in between quail
and chickens, as most of their isolates were Ck/Bei-like, with a
few G1-like viruses. It was also noted that only a few Ck/Bei-
like viruses have been recognized in domestic ducks since 2002
(Fig. 1 and Table 6). These findings suggest that the current
direction of gene flow of Ck/Bei-like H9N2 viruses is from
chickens to quail, other minor poultry species, and ducks. G1-
like viruses and those Ck/Bei-like viruses with G1-like gene
segments could be introduced from quail to other minor poul-
try species.
It is noted that G1-like viruses were not detected in chickens
in the present study but did cause outbreaks in chickens in
Middle Eastern and European countries, including Iran, Saudi
Arabia, United Arab Emirates, and Germany (1, 2, 4). Even
though those G1-like viruses underwent further reassortment
with "local" influenza viruses and generated novel reassortants
(1), the reasons that G1-like viruses have not been detected in
other types of terrestrial poultry in our surveillance are still
unknown. Obviously, the G1-like viruses detected in southern
China and Middle Eastern countries shared the same progen-
itor and evolutionary pathway. Thus, current G1-like viruses in
southern China might have originally been introduced from
Middle Eastern countries, or it is also likely that the virus
spread the other way around, similar to the transmission of
FIG. 6. Ecology of H9N2 influenza viruses in southern China.
"A" and "B" shown in black boxes indicate Qa/HK/G1/97-like and
Ck/Bei/1/94-like virus lineages, respectively. Solid lines represent
confirmed gene flow directions. Dashed lines indicate indirect evi-
dence of gene flow.
TABLE 6. Distribution of different H9N2 virus genotypes among different types of poultry from southern China
Year
Genotype (no. of isolates)a
Chickens Other minor poultry species Quail Ducksb
2000 B3 (8) B1 (2), B2 (1), B3 (1), B5 (2) A0 (3), A1 (1), B1 (1), B2 (1),
B3 (1)
B1 (2), B3 (5), B4 (1),
B5 (1), B-n8 (1)
2001 B0 (3), B3 (8), B4 (2) A0 (2), B3 (2), B4 (1), B7 (2) A0 (4), A1 (3), A2 (1), B3 (1),
B4 (1), B5 (3), B6 (1)
B3 (4), B7 (1), B-n6
(1)
2002 B3 (7), B17 (1), B18 (3) B3 (2), B7 (5), B8 (2) A0 (2), A3 (3), B3 (1), B7 (2),
B8 (3)
B3 (2), B12 (1)
2003 B3 (3), B18 (2),B19 (1), B20
(1), B21 (5), B22 (1)
B2 (2), B3 (1), B7 (11), B12 (1),
B21 (3), B23 (1)
A3 (7), B7 (6), B9 (1), B10 (1),
B11 (1), B12 (1), B13 (1)
B3 (1), B7 (1), B21 (2)
2004 B3 (5), B18 (2), B21 (1), B7 (4),
B12 (1), B14 (5), B15 (1),
B25 (2)
B3 (1), B7 (16), B8 (1), B14 (8),
B15 (2), B24 (1)
A3 (5), B7 (1), B14 (2), B15
(3)
B21 (1)
2005 B3 (2), B7 (1), B14 (14), B15
(21), B25 (1), B16 (1), B26
(1), B27 (1), B28 (3), B29 (1)
A4 (1), B3 (1), B7 (2), B14 (1),
B15 (1), B16 (3)
A3(4), B8 (3), B14 (1), B15
(2), B16 (2)
B15 (1)
a
Number of isolates recognized among representative strains tested.
b
Viruses characterized in reference 19 were also included.
VOL. 81, 2007 GENESIS AND EVOLUTION OF H9N2 INFLUENZA VIRUS 10399
Qinghai-like H5N1 virus from the east transmitting in a west-
ern or northwestern direction (6). It has been understood that
there is influenza virus gene exchange between the extremities
of Europe and Asia (9).
The present study revealed that prototypes of Ck/Bei-like
and G1-like viruses (genotypes B0 and A0) have been replaced
by their descendant reassortants since 2002 and 2003, respec-
tively (Table 6). Genotypes B3 and B7 were persistent in either
chickens or other minor poultry species from 2001 to 2005, but
these two genotypes still have not become predominant in their
host, as novel reassortants continued to emerge every year
(Fig. 5 and Table 6). This suggests that the Ck/Bei-like viruses
are of genetically unstable and transient gene constellations.
This situation could give rise to great uncertainty for the cur-
rent ecosystem in southern China. First, novel reassortants
could continue to have further interspecies transmission and
further reassort with other viruses to cause new outbreaks; and
second, novel reassortants keep challenging the species barrier
between birds and mammals.
The findings of the present investigation show a dynamic
ecosystem with multiple interspecies transmissions of H9N2
influenza virus (Fig. 6). Two-way transmissions of H9N2 be-
tween different types of poultry in southern China promote the
development of various genotypes, not only for H9N2 but also
for H5N1 and other influenza viruses. The combination of a
dynamic H9N2 ecosystem and the presence of multiple novel
genotypes increases the risk of H9N2 viruses entering the hu-
man population themselves or, like the Hong Kong H5N1 bird
flu incident (11), indirectly contributing their internal gene
complex to promote the introduction of other subtypes to
humans. This situation has posed a persistent and significant
pandemic threat in the past 10 years. However, to reduce this
kind of risk, these interspecies transmissions must be dis-
rupted. This study provides clear clues about how to interrupt
these two-way transmissions between different poultry species,
e.g., with modified market systems and farming practices in the
affected regions.
H5N1 viruses have usually caused infections with high mor-
tality in humans, which resulted in early detection and preven-
tion of further adaptation and reassortment to develop human-
to-human transmission; however, H9N2 human infections
manifest with a typical human flu-like illness that can easily be
overlooked (3, 5). At this point, H9N2 viruses have a greater
chance and time to develop the ability of human-to-human
transmission. Continuing influenza virus surveillance of both
animal and human aspects seems to be the best option for
detecting and interrupting this kind of development.
ACKNOWLEDGMENTS
This study was supported by the Li Ka Shing Foundation, the Research
Fund for Control of Infectious Diseases and Research Grants Council
(HKU1/05C) of the Hong Kong SAR Government, and the National
Institutes of Health (NIAID contract HHSN266200700005C).
REFERENCES
1. Aamir, U. B., U. Wernery, N. Ilyushina, and R. G. Webster. 2007. Charac-
terization of avian H9N2 influenza viruses from United Arab Emirates 2000
to 2003. Virology 361:45­55.
2. Banks, J., E. C. Speidel, P. A. Harris, and D. J. Alexander. 2000. Phyloge-
netic analysis of influenza A viruses of H9 haemagglutinin subtype. Avian
Pathol. 29:353­360.
3. Butt, K. M., G. J. D. Smith, H. Chen, L. J. Zhang, Y. H. Leung, K. M. Xu,
W. Lim, R. G. Webster, K. Y. Yuen, J. S. M. Peiris, and Y. Guan. 2005.
Human infection with an avian H9N2 influenza A virus in Hong Kong in
2003. J. Clin. Microbiol. 43:5760­5767.
4. Cameron, K. R., V. Gregory, J. Banks, I. H. Brown, D. J. Alexander, A. J.
Hay, and Y. P. Lin. 2000. H9N2 subtype influenza A viruses in poultry in
Pakistan are closely related to the H9N2 viruses responsible for human
infection in Hong Kong. Virology 278:36­41.
5. Centre for Health Protection. 20 March 2007. Girl recovered from influenza
A virus (H9N2). Centre for Health Protection, Department of Health, Hong
Kong, SAR, China. http://www.chp.gov.hk/content.asp?langen&info_id
9158&id116.
6. Chen, H., G. J. D. Smith, K. S. Li, J. Wang, X. H. Fan, J. M. Rayner, D.
Vijaykrishna, J. X. Zhang, L. J. Zhang, C. T. Guo, C. L. Cheung, K. M. Xu,
L. Duan, K. Huang, K. Qin, Y. H. Leung, W. L. Wu, H. R. Lu, Y. Chen, N. S.
Xia, T. S. Naipospos, K. Y. Yuen, S. S. Hassan, S. Bahri, T. D. Nguyen, R. G.
Webster, J. S. Peiris, and Y. Guan. 2006. Establishment of multiple sublin-
eages of H5N1 influenza virus in Asia: implications for pandemic control.
Proc. Natl. Acad. Sci. USA 103:2845­2850.
7. Choi, Y. K., H. Ozaki, R. J. Webby, R. G. Webster, J. S. Peiris, L. Poon, C.
Butt, Y. H. Leung, and Y. Guan. 2004. Continuing evolution of H9N2 influ-
enza viruses in southeastern China. J. Virol. 78:8609­8614.
8. Choi, Y. K., S. H. Seo, J. A. Kim, R. J. Webby, and R. G. Webster. 2005.
Avian influenza viruses in Korean live poultry markets and their pathogenic
potential. Virology 332:529­537.
9. Duan, L., L. Campitelli, X. H. Fan, C. Y. H. Leung, D. Vijaykrishna, J. X.
Zhang, I. Donatelli, M. Delogu, K. S. Li, E. Foni, C. Chiapponi, W. L. Wu,
H. Kai, R. G. Webster, K. F. Shortridge, J. S. M. Peiris, G. J. D. Smith, H.
Chen, and Y. Guan. 2007. Characterization of low-pathogenic H5 subtype
influenza viruses from Eurasia: implications for the origin of highly patho-
genic H5N1 viruses. J. Virol. 81:7529­7539.
10. Guan, Y., J. S. Peiris, A. S. Lipatov, T. M. Ellis, K. C. Dyrting, S. Krauss,
L. J. Zhang, R. G. Webster, and K. F. Shortridge. 2002. Emergence of
multiple genotypes of H5N1 avian influenza viruses in Hong Kong SAR.
Proc. Natl. Acad. Sci. USA 99:8950­8955.
11. Guan, Y., K. F. Shortridge, S. Krauss, and R. G. Webster. 1999. Molecular
characterization of H9N2 influenza viruses: were they the donors of the
"internal" genes of H5N1 viruses in Hong Kong? Proc. Natl. Acad. Sci. USA
16:9363­9367.
12. Guan, Y., K. F. Shortridge, S. Krauss, P. S. Chin, K. C. Dyrting, T. M. Ellis,
R. G. Webster, and M. Peiris. 2000. H9N2 influenza viruses possessing
H5N1-like internal genomes continue to circulate in poultry in southeastern
China. J. Virol. 74:9372­9380.
13. Guo, Y. J., S. Krauss, D. A. Senne, I. P. Mo, K. S. Lo, X. P. Xiong, M.
Norwood, K. F. Shortridge, R. G. Webster, and Y. Guan. 2000. Character-
ization of the pathogenicity of members of the newly established H9N2
influenza virus lineages in Asia. Virology 267:279­288.
14. Herlocher, M. L., R. Truscon, S. Elias, H. L. Yen, N. A. Roberts, S. E. Ohmit,
and A. S. Monto. 2004. Influenza viruses resistant to the antiviral drug
oseltamivir: transmission studies in ferrets. J. Infect. Dis. 190:1627­1630.
15. Huelsenbeck, J. P., and F. R. Ronquist. 2001. MrBayes: Bayesian inference
of phylogenetic trees. Bioinformatics 17:754­755.
16. Humberd, J., Y. Guan, and R. G. Webster. 2006. Comparison of the repli-
cation of influenza A viruses in Chinese ring-necked pheasants and chukar
partridges. J. Virol. 80:2151­2161.
17. Lee, Y. J., J. Y. Shin, M. S. Song, Y. M. Lee, J. G. Choi, E. K. Lee, O. M.
Jeong, H. W. Sung, J. H. Kim, Y. K. Kwon, J. H. Kwon, C. J. Kim, R. J.
Webby, R. G. Webster, and Y. K. Choi. 2007. Continuing evolution of H9
influenza viruses in Korean poultry. Virology 359:313­323.
18. Li, K. S., Y. Guan, J. Wang, G. J. D. Smith, K. M. Xu, L. Duan, A. P.
Rahardjo, P. Puthavathana, C. Buranathai, T. D. Nguyen, A. T. Estoepang-
estie, A. Chaisingh, P. Auewarakul, H. T. Long, N. T. Hanh, R. J. Webby,
L. L. Poon, H. Chen, K. F. Shortridge, K. Y. Yuen, R. G. Webster, and
J. S. M. Peiris. 2004. Genesis of a highly pathogenic and potentially pan-
demic H5N1 influenza virus in eastern Asia. Nature 430:209­213.
19. Li, K. S., K. M. Xu, J. S. Peiris, L. L. Poon, K. Z. Yu, K. Y. Yuen, K. F.
Shortridge, R. G. Webster, and Y. Guan. 2003. Characterization of H9
subtype influenza viruses from the ducks of southern China: a candidate for
the next influenza pandemic in humans? J. Virol. 77:6988­6994.
20. Lin, Y. P., M. Shaw, V. Gregory, K. Cameron, W. Lim, A. Klimov, K.
Subbarao, Y. Guan, S. Krauss, K. Shortridge, R. Webster, N. Cox, and A.
Hay. 2000. Avian-to-human transmission of H9N2 subtype influenza A vi-
ruses: relationship between H9N2 and H5N1 human isolates. Proc. Natl.
Acad. Sci. USA 97:9654­9658.
21. Makarova, N. V., H. Ozaki, H. Kida, R. G. Webster, and D. R. Perez. 2003.
Replication and transmission of influenza viruses in Japanese quail. Virology
310:8­15.
22. Matrosovich, M. N., S. Krauss, and R. G. Webster. 2001. H9N2 influenza A
viruses from poultry in Asia have human virus-like receptor specificity. Vi-
rology 281:156­162.
10400 XU ET AL. J. VIROL.
23. Nylander, J. A. A. 2004. MRMODELTEST 2. Evolutionary Biology Centre,
Uppsala University, Uppsala, Sweden.
24. Peiris, J. S., Y. Guan, D. Markwell, P. Ghose, R. G. Webster, and K. F.
Shortridge. 2001. Cocirculation of avian H9N2 and contemporary "human"
H3N2 influenza A viruses in pigs in southeastern China: potential for genetic
reassortment? J. Virol. 75:9679­9686.
25. Peiris, M., K. Y. Yuen, C. W. Leung, K. H. Chan, P. L. Ip, R. W. Lai, W. K.
Orr, and K. F. Shortridge. 1999. Human infection with influenza H9N2.
Lancet 354:916­917.
26. Perez, D. R., W. Lim, J. P. Seiler, G. Yi, M. Peiris, K. F. Shortridge, and
R. G. Webster. 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. 77:3148­3156.
27. Seo, S. H., E. Hoffmann, and R. G. Webster. 2002. Lethal H5N1 influenza
viruses escape host anti-viral cytokine responses. Nat. Med. 8:950­954.
28. Shortridge, K. F. 1992. Pandemic influenza: a zoonosis? Semin. Respir.
Infect. 7:11­25.
29. Swofford, D. L. 2001. PAUP*: phylogenetic analysis using parsimony (and
other methods) 4.0 beta. Sinauer Associates, Sunderland, MA.
30. Xu, C., W. Fan, R. Wei, and H. Zhao. 2004. Isolation and identification of
swine influenza recombinant A/Swine/Shandong/1/2003 (H9N2) virus. Mi-
crobes Infect. 10:919­925.
31. Xu, K. M., K. S. Li, G. J. Smith, J. W. Li, H. Tai, J. X. Zhang, R. G. Webster,
J. S. Peiris, H. Chen, and Y. Guan. 2007. Evolution and molecular epide-
miology of H9N2 influenza A viruses from quail in southern China, 2000 to
2005. J. Virol. 81:2635­2645.
VOL. 81, 2007 GENESIS AND EVOLUTION OF H9N2 INFLUENZA VIRUS 10401
