﻿Characterization of H9N2 influenza A viruses isolated from
chicken products imported into Japan from China
M. MASE1
*, M. ETO2
, K. IMAI3
, K. TSUKAMOTO1
AND S. YAMAGUCHI1
1
Department of Infectious Diseases, National Institute of Animal Health, Kannondai, Tsukuba, Ibaraki, Japan
2
Animal Quarantine Service, Haramachi, Yokohama, Japan
3
Obihiro University of Agriculture and Veterinary Medicine, Inada, Obihiro, Hokkaido, Japan
(Accepted 10 April 2006; first published online 7 July 2006)
SUMMARY
We characterized eleven H9N2 influenza A viruses isolated from chicken products imported
from China. Genetically they were classified into six distinct genotypes, including five already
known genotypes and one novel genotype. This suggested that such multiple genotypes of the
H9N2 virus have possibly already become widespread and endemic in China. Two isolates have
amino-acid substitutions that confer resistance to amantadine in the M2 region, and this
supported the evidence that this mutation might be a result of the wide application of amantadine
for avian influenza treatment in China. These findings emphasize the importance of surveillance
for avian influenza virus in this region, and of quarantining imported chicken products as
potential sources for the introduction of influenza virus.
INTRODUCTION
H9N2 influenza A viruses have been distributing
widely from East Asia to points in the Middle East,
Europe, and Southern Africa [1­3]. H9N2 influenza
A viruses were also isolated from humans in 1999
and 2003 [4, 5]. The H9N2 viruses isolated in
1999 shared common internal gene components
with the H5N1 influenza A viruses from 1997 that
were lethal to humans in Hong Kong [6]. These H5N1
viruses were considered to be genetic reassortants
generated from two or three viruses, including
the H9N2 virus [7]. Of importance is the fact that
recognition of human-like receptor specificity by
recent H9N2 viruses in Asia could have facilitated
direct transmission to humans [8]. Thus, H9N2
viruses along with H5N1 viruses are high on the list
of candidates that could potentially cause a future
human influenza pandemic.
However, since its first detection in southern China
(i.e. A/goose/Guandong/1/96), the H5N1 influenza
A virus has spread in Asian countries, where it is now
enzootic and has caused multiple outbreaks in poultry
[9, 10]. Among them, the isolation of an H5N1
virus from duck meat imported from China to Korea
in 2001 raises the possibility of the introduction
of viruses and transmission to humans from novel
sources [11]. Given this possibility, the Japanese
government decided in 2001 to halt the import of
duck meat and to survey for the presence of the
viruses in imported chicken products from China at
the Animal Quarantine Service (AQS). During this
process, H9N2 viruses and Newcastle disease viruses
were isolated from chicken products such as meat and
bone marrow.
In this study, we performed a genetic, antigenic,
and pathogenic analysis of H9N2 viruses isolated
* Author for correspondence: Dr M. Mase, Department of
Infectious Diseases, National Institute of Animal Health, 3-1-5
Kannondai, Tsukuba, Ibaraki 305-0856, Japan.
(Email: masema@affrc.go.jp)
Epidemiol. Infect. (2007), 135, 386­391. f 2006 Cambridge University Press
doi:10.1017/S0950268806006728 Printed in the United Kingdom
from chicken products at the AQS in Japan in order
to compare them with other influenza A viruses.
METHODS
Sampling and virus isolation
At the AQS in Japan, a total of 473 samples of frozen
chicken meat and bone marrow collected between
June 2001 and July 2002 were used for virus isolation.
Briefly, applications were submitted to the AQS to
import lots of chicken products (500­3000 cartons/lot
under a single import application). One carton from
each imported lot was randomly selected and tested
for the presence of the influenza virus by sampling 1 g
meat from each of 10 packages per carton, which was
pooled and made into a 10% homogenate in phos-
phate buffered saline (PBS). The homogenates were
then centrifuged at low speed (3000 g for 10 min) and
the supernatants filtered through a sterile 0.45 mm
membrane filter before inoculation into the allantoic
cavity of embryonated specific pathogen-free (SPF)
eggs. All isolates were identified as influenza A viruses
of the H9N2 subtype by conventional haemaggluti-
nation inhibition (HI) and neuraminidase activity in-
hibition (NI) tests. Additionally, the H9N2 isolate,
which was isolated in surveillance for imported
chicken products at the AQS in 1997 (A/chicken/
Osaka/aq48/97), was also used in this study (Table 1).
The H9 prototype A/turkey/Wisconcin/66 strain (Ty/
WI/66) and our previous isolates from imported
parakeets, A/parakeet/Chiba/1/97 (Pa/Chiba/1/97)
and A/parakeet/Narita/92A/98 (Pa/Narita/92A/98)
[12] were also used in this study.
Antigenic characterization
All chicken antisera against Ty/WI/66, Pa/Chiba/1/
97, A/chicken/Osaka/aq48/97 (Ck/Os/aq48/97) and
A/chicken/Yokohama/aq55/2001 (Ck/Yo/aq55/01)
were prepared in our laboratory according to a pre-
viously described method [7]. Briefly, 3-week-old
SPF chickens were inoculated intranasally and orally
with 0.1 ml allantoic fluid [y106
50% egg infectious
dose (EID50)]. The chickens were bled 3 weeks post-
infection and boosted intravenously with 0.1 ml
of infectious virus. HI and NI tests were performed
according to standard procedures [13].
Genetic and phylogenetic analysis
RNA extraction, RT­PCR, and sequencing of the
PCR products were carried out as described previously
[12]. PCR amplification of the coding regions of the
viral gene segments was performed with gene-specific
primer sets (sequences of the primers available upon
request). The nucleotide sequences were analysed
using version 12.0 of the sequence analysis software
package GENETYX-MAC (Software Development,
Tokyo, Japan). The nucleotide sequences obtained
from this study are available from GenBank under
accession numbers AB256663­AB256750. Multiple
alignments were conducted by the Cluster X program
[14], and phylogenetic trees were constructed by
the neighbour-joining (NJ) method as previously
described [15]. Based on each phylogenetic analysis
of each gene segment, the genotypes of the H9N2
viruses isolated from chicken products were compared
with previously reported genotypes [2, 7, 16­18].
Table 1. H9N2 influenza A viruses isolated from chicken products imported from China at AQS
Virus Abbreviation Origin
Amino acid residue*
Genotype
Corresponding
genotype
183 190 226
A/chicken/Osaka/aq48/97 Ck/Os/aq48/97 Shanghai N A Q I A [18]
A/chicken/Osaka/aq19/2001 Ck/Os/aq19/01 Shenzhen N V L II Novel
A/chicken/Kobe/aq26/2001 Ck/Ko/aq26/01 Shandong N T L III Dk/ST/1605/01 [2]
A/chicken/Yokohama/aq55/2001 Ck/Yo/aq55/01 Jilin N V L II Novel
A/chicken/Osaka/aq58/2001 Ck/Os/aq58/01 Liaoning N A L IV Dk/ST/1796/00 [2]
A/chicken/Osaka/aq69/2001 Ck/Os/aq69/01 Shanghai N A L V H [18]
A/chicken/Yokohama/aq120/2001 Ck/Yo/aq120/01 Jilin N A L VI F [18]
A/chicken/Yokohama/aq135/2001 Ck/Yo/aq135/01 Shenzhen N T L II Novel
A/chicken/Yokohama/aq144/2001 Ck/Yo/aq144/01 Jilin N V L VI F
A/chicken/Yokohama/aq45/2002 Ck/Yo/aq45/02 Shandong N V L I A
A/chicken/Yokohama/aq134/2002 Ck/Yo/aq134/02 Shandong N A L III Dk/ST/1605/01 [2]
* H3 numbering.
H9N2 isolates from chicken products 387
Pathogenicity to chickens
Three isolates, Ck/Os/aq19/01, Ck/Ko/aq26/01, and
Ck/Yo/aq55/01, were chosen for the pathogenicity
test. For pathogenicity testing according to the
guidelines established by the World Organization for
Animal Health (Office International des Epizooties;
OIE) [19], 6-week-old SPF chickens were used in this
study. Eight chickens were inoculated intravenously
with 0.2 ml of a 1:10 dilution of infected allantoic
fluid, and clinical signs were observed daily for 10
days. In addition, we performed an intracerebral
inoculation test according to that for the Newcastle
disease virus based on the guidelines established
by the OIE [19]. Ten chickens were inoculated
intracerebrally with 0.05 ml of a 1:10 dilution of in-
fected allantoic fluid and clinical signs were observed
daily for 8 days.
RESULTS
Genetic and phylogenetic analysis
The HA genes of the H9N2 viruses isolated from
chicken products belonged to the A/chicken/Beijing/
1/94 (Ck/BJ/1/94) sub-lineage [Fig. (a); phylogenetic
trees of genes of influenza A viruses can be seen in the
Figure on the Journal's website] as described by Li
et al. [18]. The amino-acid sequence at the HA cleav-
age site of H9N2 isolates from chicken products
possessed a PARSSR-GLF motif, which corre-
sponded to low-pathogenicity viruses [1, 16, 20]. The
amino acids at the receptor binding site of HA
proteins are associated with differences in the receptor
binding specificity [21]. Table 1 shows the amino-acid
positions 183, 190, and 226 (numbering according to
H3 HA). All H9N2 viruses isolated from chicken
products possessed N at amino-acid position 183.
Except for one strain (Ck/Os/aq48/97), all H9N2
viruses isolated from chicken products possessed L at
amino-acid position 226. Five potential glycosylation
sites in HA1 (11, 123, 200, 280, and 287) were con-
served in all H9N2 viruses isolated from chicken
products (data not shown).
The M genes of the H9N2 viruses isolated from
chicken products also belonged to the Ck/BJ/1/94
sub-lineage [Fig. (b)] [18]. Resistance to the two types
of influenza antiviral compounds (M2 ion channel
blockers, e.g. amantadine and rimantadine, and NA
inhibitors, e.g. oseltamivir and zanamivir [22]) is as-
sociated with particular mutations. Viruses become
resistant to amantadine through a single amino-acid
substitution at positions 26, 27, 30, 31, or 34 in the
transmembrane region of the M2 protein [23]. Two
isolates, Ck/Os/19/01 and Ck/Yo/45/02, have the
amino-acid substitution at position 30 (A to T) and 31
(S to N), and these substitutions confer resistance to
amantadine in the M2 protein, respectively.
The NA genes of the H9N2 viruses isolated from the
chicken products also divided into two sub-lineages
(Ck/BJ/1/94-like and Ck/HK/G9/97-like) [18] [Fig.
(c)]. Six isolates clustered with Ck/BJ/1/94-like
viruses, which contain a three (at amino-acid pos-
itions 63­65) amino-acid deletion in the stalk region.
The other five isolates clustered with Ck/HK/G9/97-
like viruses, which do not contain a stalk deletion.
Through a single amino-acid substitution at positions
119, 152, 274, 292 or 294 in the NA active centre
[24, 25], viruses can become resistant to oseltamivir.
None of the amino-acid substitutions that were
resistant to oseltamivir in the NA protein were
found in the H9N2 viruses isolated from chicken
products.
Phylogenetic analysis of the remaining gene seg-
ments of the H9N2 isolates from chicken products
suggested that they were divided into six genotypes
(temporarily designated I­VI) [Fig. (d­h), Table 1].
These sequences were compared with representa-
tive H9N2 viruses, including Ck/BJ/1/94, A/quail/
Hong Kong/G1/97 (Qa/HK/G1/97), A/chicken/Hong
Kong/G9/97 (Ck/HK/G9/97), A/duck/Hong Kong/
Y280/97 (Dk/HK/Y280/97), A/chicken/Shanghai/
F/98 (Ck/SH/F/98), and some duck viruses isolated
in Shantou, Guangdong Province in mainland China,
reported previously by others [2, 7, 16­18]. Compared
with the H9N2 genotypes that have been described
(Table 1), genotype I seemed to correspond to geno-
type A defined by Li et al. [18]. Genotypes III and VI
seemed to correspond to genotypes Dk/ST/1605/01
and Dk/ST/1796/00 defined by Li et al. [2] respect-
ively. Genotype V seemed to correspond to genotype
H defined by Li et al. [18]. Genotype VI seemed to
correspond to genotype F defined by Li et al. [18]. The
remaining genotype II seemed to be a novel genotype
in which the PB1 belonged to the Ck/BJ/1/94 lineage,
with the other seven segments corresponding to
genotype I as defined by Li et al. [18].
Antigenic analysis
Antisera against the H9N2 isolates from chicken
products imported from China reacted with all
tested strains within the threefold dilution of the
388 M. Mase and others
homologous HI titres (Table 2). The reactivity of the
H9N2 isolates from chicken products to antisera
against the H9 prototype Ty/WI/66 was diverse.
However, antisera against Pa/Chiba/1/97, which be-
longs to the Qa/HK/G1/97 lineage [12], reacted with
most tested strains except Ck/Yo/aq55/01, with the
fourfold dilution lower than the homologous HI titres.
NI tests revealed that the antisera against the H9N2
isolates from the chicken products imported from
China also reacted with most tested strains within
the threefold dilution of the homologous NI titres
(Table 3). Differing from the results of the HI tests,
the antisera against Pa/Chiba/1/97 reacted with all
tested strains within the threefold dilution of the
homologous NI titres. The antisera against the H9
prototype Ty/WI/66 reacted with most H9N2 isolates
from chicken products, with the fourfold dilution
being lower than the homologous NI titres.
Pathogenicity tests
The three H9N2 isolates from chicken products were
not judged as highly pathogenic for chickens by the
OIE guidelines, since none of the chickens died and
Table 2. Cross haemaggulutination inhibition test among H9N2 viruses
Viruses
HI titres of chickens' antisera
Ty/WI/66
Ck/Os/
aq48/97
Pa/Chiba/
1/97
Ck/Yo/
aq55/01
Ty/WI/66 1280 320 160 640
Ck/Os/aq48/97 1280 5120 80 1280
Pa/Chiba/1/97 320 640 2560 1280
Pa/Narita/92A/98 160 640 160 640
Ck/Os/aq19/01 160 1280 160 2560
Ck/Ko/aq26/01 160 2560 160 2560
Ck/Yo/aq55/01 320 2560 320 2560
Ck/Os/aq58/01 160 5120 160 1280
Ck/Os/aq69/01 80 2560 80 2560
Ck/Yo/aq120/01 80 2560 80 2560
Ck/Yo/aq135/01 40 1280 80 1280
Ck/Yo/aq144/01 40 1280 40 640
Ck/Yo/aq45/02 160 10240 160 5120
Ck/Yo/aq134/02 80 5120 80 2560
Table 3. Cross neuraminidase activity inhibition test among H9N2 viruses
Viruses
NI titres of chickens' antisera
Ty/WI/66
Ck/Os/
aq48/97
Pa/Chiba/
1/97
Ck/Yo/
aq55/01
Ty/WI/66 2560 80 160 160
Ck/Os/aq48/97 80 320 80 80
Pa/Chiba/1/97 640 320 320 80
Pa/Narita/92A/98 320 80 640 40
Ck/Os/aq19/01 80 80 40 80
Ck/Ko/aq26/01 160 160 160 80
Ck/Yo/aq55/01 160 80 160 320
Ck/Os/aq58/01 80 320 160 320
Ck/Os/aq69/01 80 160 80 80
Ck/Yo/aq120/01 40 40 40 40
Ck/Yo/aq135/01 160 160 80 320
Ck/Yo/aq144/01 80 80 80 320
Ck/Yo/aq45/02 160 320 160 160
Ck/Yo/aq134/02 160 160 80 40
H9N2 isolates from chicken products 389
no clinical signs were observed in the inoculated
chickens. In the intracerebral inoculation tests, only
one chicken inoculated with Ck/Ko/aq26/01 died
during the observation period. The range of values of
the intracerebral pathogenicity index of the tested
viruses was 0.00­0.2, which showed low pathogenicity
to chickens.
DISCUSSION
We characterized the H9N2 isolates from chicken
products imported from China. The HA of these
H9N2 viruses seemed to be antigenically diverse, but
the NA seemed to have a close relationship with one
another. The amino acids of the receptor binding sites
in HA are similar to those affecting humans. Almost
all viruses possess L at position 226, which is identical
to the human H2 and H3 strains [8]. L at position 226
binds to the Neu Aca2,6 Gal linkage, whereas that
possessing Q 226 binds to the Neu Aca2,3 Gal linkage
[26]. These results suggested that the recent H9N2
viruses might be candidates for a pandemic influenza
outbreak in humans.
This surveillance of chicken products imported
from China was conducted from 2001 to 2002, and
the H9N2 viruses were isolated from 17 (2.7%) of
621 samples at a comparatively high frequency. This
suggests that such contaminated chicken products
could be a potential source for the introduction of
avian influenza viruses into other countries. The
pathogenicity index indicates that the H9N2 viruses
isolated from chicken products were of low patho-
genicity to chickens, and the reason for the frequent
isolation of low-pathogenic H9N2 viruses from
chicken products in this study was not completely
understood. In outbreaks of the H9N2 virus in Iran, a
mixed infection caused severe clinical features with
bacteria or mycoplasma [27]. In addition, Kishida
et al. revealed that co-infection with Streptococcus
or Haemophilus increased the virulence of H9N2
viruses in chickens [28]. Hence, the same secondary
pathogens may be related to severe pathogenicity and
isolation of the virus from novel sources such as
chicken products.
So far, the prevalence of multiple H9N2 genotypes
in mainland China has been reported [2, 16­18].
Previously, Liu et al. showed that all eight segments
of H9N2 viruses isolated in China between 1995
and 1999 genetically belonged to Ck/HK/G9/97
lineages [20]. Thereafter, multiple genotypes of the
H9N2 viruses have been generated from complicated
reassortment with various viruses [18]. In this study,
the multiple genotypes of H9N2 viruses, which
include the genotype isolated from only duck [2], were
detected in chicken products imported from China.
Interestingly, most genotypes of the poultry products
originated from only one province; genotype III only
in Shandong, IV from Liaoning, V only in Shanghai,
and VI only in Jilin. In contrast, genotype I originated
from Shanghai and Shandong and genotype II from
Shenzhen or Jilin. This may suggest that there is a
prevalent genotype in each province, and that it may
be related to the emergence of novel genotype H5N1
viruses [2, 6].
The amino-acid substitutions, which confer re-
sistance to amantadine in the M2 region [23], were
detected in two isolates (Ck/Os/aq19/01 and Ck/Yo/
aq45/02 isolated in Shenzhen and Shandong respect-
ively). These isolates were obtained from chicken
products with an unknown history of amantadine
antiviral therapy. As previously suggested [18], this
mutation may be the result of the application of
amantadine for avian influenza treatment, and there
may be a wide distribution of this resistant strain
to these drugs in mainland China. Continuing sur-
veillance of avian influenza viruses in this region is
needed.
ACKNOWLEDGEMENTS
This work was partially supported by a Grant-in-Aid
from the Zoonoses Control Project of the Ministry of
Agriculture, Forestry and Fisheries, Japan.
DECLARATION OF INTEREST
None.
NOTE
Supplementary information accompanies this paper
on the Journal's website (http://journals.cambridge.
org).
REFERENCES
1. Banks J, et al. Phylogenetic analysis of influenza A
viruses of H9 haemagglutinin subtype. Avian Pathology
2000; 29: 353­360.
2. Li KS, et al. Characterization of H9 subtype influenza
viruses from the ducks of southern China: a candidate
for the next influenza pandemic in humans? Journal of
Virology 2003; 77: 6988­6994.
390 M. Mase and others
3. Naeem K, et al. Avian influenza A subtype H9N2 in
poultry in Pakistan. Veterinary Record 1999; 145: 560.
4. Peiris M, et al. Human infection with influenza H9N2.
Lancet 1999; 354: 916­917.
5. Butt KM, et al. Human infection with an avian H9N2
influenza A virus in Hong Kong in 2003. Journal of
Clinical Microbiology 2005; 43: 5760­5767.
6. Lin YP, et al. Avian-to-human transmission of H9N2
subtype influenza A viruses: relationship between
H9N2 and H5N1 human isolates. Proceedings of the
National Academy of Sciences USA 2000; 97: 9654­
9658.
7. Guan Y, et al. Molecular characterization of H9N2
influenza viruses: were they donors of the `internal'
genes of H5N1 viruses in Hongkong? Proceedings of
the National Academy of Sciences USA 1999; 96:
9363­9367.
8. Matrosovich MN, Krauss S, Webster RG. H9N2 influ-
enza A viruses from poultry in Asia have human virus-
like receptor specificity. Virology 2001; 281: 156­162.
9. Li K, et al. Genesis of a highly pathogenic and poten-
tially pandemic H5N1 influenza virus in eastern Asia.
Nature 2004; 430: 209­213.
10. World Health Organization Global Influenza Program
Surveillance Network. Evolution of H5N1 avian influ-
enza viruses in Asia. Emerging Infectious Diseases 2005;
11: 1515­1521.
11. Tumpey TM, et al. Characterization of a highly patho-
genic H5N1 avian influenza A virus isolated from duck
meat. Journal of Virology 2002; 76: 6344­6355.
12. Mase M, et al. Imported parakeets harbor H9N2 influ-
enza A viruses that are genetically closely related to
those transmitted to humans in Hong Kong. Journal of
Virology 2001; 75: 3490­3494.
13. Madeley CR, Allan WH, Kendal AP. Studies with avian
influenza A viruses: serological relations of the hae-
magglutinin and neuraminidase antigens of ten virus
isolates. Journal of General Virology 1971; 12: 69­78.
14. Thompson JD, et al. The CLUSTAL-X windows inter-
face: flexible strategies for multiple sequence alignment
aided by quality analysis tools. Nucleic Acids Research
1997; 25: 4876­4882.
15. Saitou N, Nei M. The neighbor-joining method: a new
method for reconstructing phylogenetic trees. Molecu-
lar Biology and Evolution 1987; 4: 406­425.
16. Guo YJ, et al. Characterization of the pathogenicity of
members of the newly established H9N2 influenza virus
lineages in Asia. Virology 2000; 267: 279­288.
17. Choi YK, et al. Continuing evolution of H9N2 influenza
viruses in Southeastern China. Journal of Virology
2004; 78: 8609­8614.
18. Li C, et al. Evolution of H9N2 influenza viruses from
domestic poultry in Mainland China. Virology 2005;
340: 70­83.
19. Office International des Epizooties. Newcastle Disease,
chapter 2.1.15. In: Manual of Diagnostic Tests and
Vaccines for Terrestrial Animals, 5th edn. Paris, France:
Office International des Epizooties, 2004 (http://
www.oie.int).
20. Liu J, et al. H9N2 influenza viruses prevalent in
poultry in China are phylogenetically distinct from
A/quail/Hong Kong/G1/97 presumed to be the donor
of the internal protein genes of the H5N1 Hong Kong/
97 virus. Avian Pathology 2003; 32: 551­560.
21. Rogers GN, et al. Single amino acid substitutions in in-
fluenza haemagglutinin change receptor binding speci-
ficity. Nature 1983; 304: 76­78.
22. Monto AS. The role of antivirals in the control of in-
fluenza. Vaccine 2003; 21: 1796­1800.
23. Crumpacker C. Antiviral therapy. In: Knipe DM,
Howley PM, eds. Fields Virology, 4th edn. Philadelphia:
Lippincott Williams & Wilkins, 2001, pp. 393­433.
24. Gubareva LV, Webster RG, Hayden FG. Detection of
influenza virus resistance to neuraminidase inhibitors
by an enzyme inhibition assay. Antiviral Research 2002;
53: 47­61.
25. Kiso M, et al. Resistant influenza A viruses in children
treated with oseltamivir: descriptive study. Lancet
2004; 364: 759­765.
26. Weis W, et al. Structure of the influenza virus haem-
agglutinin complexed with its receptor, sialic acid.
Nature 1988; 333: 426­431.
27. Nill H, Asasi K. Natural cases and an experimental
study of H9N2 avian influenza in commercial
broiler chickens of Iran. Avian Pathology 2002; 31:
247­252.
28. Kishida N, et al. Co-infection of Staphylococcus aureus
or Haemophilus paragallinarum exacerbates H9N2
influenza A virus infection in chickens. Archives of
Virology 2004; 149: 2095­2104.
H9N2 isolates from chicken products 391
