High genetic compatibility and increased pathogenicity of reassortants derived from avian H9N2 and pandemic H1N1/2009 influenza viruses.

Abstract

H9N2 influenza viruses have been circulating worldwide in multiple avian species and repeatedly infecting mammals, including pigs and humans, posing a significant threat to public health. The coexistence of H9N2 and pandemic influenza H1N1/2009 viruses in pigs and humans provides an opportunity for these viruses to reassort. To evaluate the potential public risk of the reassortant viruses derived from these viruses, we used reverse genetics to generate 127 H9 reassortants derived from an avian H9N2 and a pandemic H1N1 virus, and evaluated their compatibility, replication ability, and virulence in mice. These hybrid viruses showed high genetic compatibility and more than half replicated to a high titer in vitro. In vivo studies of 73 of 127 reassortants revealed that all viruses were able to infect mice without prior adaptation and 8 reassortants exhibited higher pathogenicity than both parental viruses. All reassortants with higher virulence than parental viruses contained the PA gene from the 2009 pandemic virus, revealing the important role of the PA gene from the H1N1/2009 virus in generating a reassortant virus with high public health risk. Analyses of the polymerase activity of the 16 ribonucleoprotein combinations in vitro suggested that the PA of H1N1/2009 origin also enhanced polymerase activity. Our results indicate that some avian H9-pandemic reassortants could emerge with a potentially higher threat for humans and also highlight the importance of monitoring the H9-pandemic reassortant viruses that may arise, especially those that possess the PA gene of H1N1/2009 origin.

Full text

Page 1

High genetic compatibility and increased pathogenicity
of reassortants derived from avian H9N2 and
pandemic H1N1/2009 influenza viruses
Yipeng Suna,1, Kun Qinb,1, Jingjing Wanga, Juan Pua, Qingdong Tanga, Yanxin Hua, Yuhai Bia,c, Xueli Zhaoa,
Hanchun Yanga, Yuelong Shub, and Jinhua Liua,d,2
aKey Laboratory of Zoonosis of Ministry of Agriculture, College of Veterinary Medicine, China Agricultural University, Beijing 100193, China;bChinese
National Influenza Center, State Key Laboratory for Molecular Virology and Genetic Engineering, National Institute for Viral Disease Control and Prevention,
Chinese Center for Disease Control and Prevention, Beijing 100052, China;dThe Shandong Animal Disease Control Center, Jinan, Shandong 250022, China;
andcCAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China
Edited by Peter Palese, Mount Sinai School of Medicine, New York, NY, and approved January 28, 2011 (received for review December 21, 2010)
H9N2influenzaviruseshavebeencirculatingworldwideinmultiple
avian species and repeatedly infecting mammals, including pigs
and humans, posing a significant threat to public health. The
coexistence of H9N2 and pandemic influenza H1N1/2009 viruses in
pigs and humans provides an opportunity for these viruses to
reassort. To evaluate the potential public risk of the reassortant
viruses derived from these viruses, we used reverse genetics to
generate 127 H9 reassortants derived from an avian H9N2 and a
pandemic H1N1virus,andevaluatedtheircompatibility,replication
ability, and virulence in mice. These hybrid viruses showed high
geneticcompatibilityandmorethanhalfreplicatedtoahightiterin
vitro. In vivo studies of 73 of 127 reassortants revealed that all
viruses were able to infect mice without prior adaptation and 8
reassortants exhibited higher pathogenicity than both parental
viruses.All reassortants with higher virulencethan parental viruses
contained the PA gene from the 2009 pandemic virus, revealing the
important role of the PA gene from the H1N1/2009 virus in gen-
erating a reassortant virus with high public health risk. Analyses of
the polymerase activity of the 16 ribonucleoprotein combinations
in vitro suggested that the PA of H1N1/2009 origin also enhanced
polymerase activity. Our results indicate that some avian H9-
pandemic reassortants could emerge with a potentially higher
threat for humans and also highlight the importance of monitoring
the H9-pandemic reassortant viruses that may arise, especially
those that possess the PA gene of H1N1/2009 origin.
H
noteworthy that H9N2 influenza viruses in poultry have occa-
sionally been transmitted to mammalian species, including
humans and pigs (5–9). Human H9N2 infections produce a typ-
ical human flu-like illness that can easily be overlooked (6, 10),
so they may have a greater opportunity to adapt to humans and
acquire the ability of human-to-human transmission. In fact,
several serological surveys revealed that a large number of
people in China, ranging from 13.7% to 37.2%, might have ev-
idence of prior infections of the H9N2 virus (11, 12). In addition,
previous studies demonstrated that a significant proportion of
H9N2 field isolates have acquired preference for a human virus-
like receptor (10, 13). Thus, H9N2 influenza virus, along with
H5N1 virus, is high on the list of candidates that could poten-
tially cause another human influenza pandemic.
Pandemic H1N1/2009 influenza virus has spread by human-to-
human transmission across the globe at an unprecedented rate
since it was first isolated from humans in Mexico in 2009 (14, 15,
16). Recently, it has been announced by the World Health Or-
ganization (WHO) that the pandemic H1N1 influenza virus is
now in the postpandemic period but is expected to become a
recurrent seasonal influenza virus and circulate for some years
(WHO; http://www.who.int/csr/disease/swineflu/en/). Addition-
ally, pandemic H1N1/2009 influenza viruses were also frequently
9N2 influenza viruses circulate worldwide and are endemic
in multiple terrestrial avian species in Asia (1–4). It is
isolated from pigs (17–19) that were proposed to be “mixing
vessels” for the reassortment of influenza viruses.
Coinfection with H9N2 and pandemic H1N1/2009 influenza
viruses in the same host (e.g., pigs and humans) provides the
opportunity for reassortment between these viruses. Reassort-
mentisanimportantmechanismforthegenerationofapandemic
influenza strain (20, 21). For example, the pandemic influenza
viruses of 1957 and 1968 emerged through genetic reassortment
ofavian viruses withthe prevailing human viruses topossessnovel
antigenicity and efficient human-to-human transmissibility (22,
23). These facts remind us that avian H9N2 influenza viruses
could acquire some functions critical for pandemic strains by
reassortment with pandemic H1N1 influenza viruses when
infecting the same host. Furthermore, previous studies revealed
that frequent reassortment is an important evolution mechanism
of H9N2 influenza viruses (4, 5) and that the pandemic H1N1/
2009 influenza virus had already reassorted with the H1N1 swine
influenza virus (24). Therefore, the concern is that if H9N2 in-
fluenza virus reassorts with a pandemic (H1N1) virus, another
pandemic strain will emerge.
Reverse genetics provides a tool to predict the potential public
health risk of novel influenza viruses. Using such an approach,
previous studies generated various avian H5N1–human H3N2
reassortant viruses to evaluate their biological properties (25,
26). In the current study, we generated a panel of reassortants
derived from contemporary avian H9N2 and pandemic H1N1/
2009 viruses by reverse genetics to study their genetic compati-
bility and biological characteristics. Theoretically, 254 (256 mi-
nus 2 parental viruses) genotypes of reassortants can be
generated from two distinct influenza A viruses. In the present
study, we attempted to generate all of the 127 (128 minus 1
H9N2 parental virus) reassortants derived from an avian H9N2
influenza virus (A/chicken/Hebei/LC/2008, HB08) and a pan-
demic H1N1 influenza virus (A/Beijing/16/2009, BJ09). These
Author contributions: Y. Sun, J.P., and J.L. designed research; Y. Sun, K.Q., J.W., Q.T., Y.H.,
Y.B., and X.Z. performed research; K.Q. and J.L. contributed new reagents/analytic tools;
Y. Sun, K.Q., J.P., Q.T., Y.H., H.Y., Y. Shu, and J.L. analyzed data; and Y. Sun, K.Q., J.W.,
and J.L. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
The sequences reported in this paper have been deposited in the GenBank database
[accession nos. GQ373026 (HB08, PB2 gene), GQ373043 (HB08, PB1 gene), GQ373060
(HB08, PA gene), GQ373077 (HB08, HA gene), GQ373094 (HB08, NP gene), GQ373111
(HB08, NA gene), GQ373128 (HB08, M gene), GQ373144 (HB08, NS gene), HQ698624
(BJ09, PB2 gene), HQ698625 (BJ09, PB1 gene), HQ698626 (BJ09, PA gene), HQ698627
(BJ09, HA gene), HQ698628 (BJ09, NP gene), HQ698629 (BJ09, NA gene), HQ698630
(BJ09, M gene), and (BJ09, NS gene)].
1Y. Sun and K.Q. contributed equally to this work.
2To whom correspondence should be addressed. E-mail: ljh@cau.edu.cn.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1019109108/-/DCSupplemental.
4164–4169
| PNAS
| March 8, 2011
| vol. 108
| no. 10www.pnas.org/cgi/doi/10.1073/pnas.1019109108

Page 2

reassortants possessed an HA gene from the avian H9N2 in-
fluenza virus and the other seven genes from either parental
virus. Our in vitro and in vivo analyses indicated that these two
influenza viruses possessed high genetic compatibility and some
reassortants showed enhanced pathogenicity in mice compared
with parental viruses. Therefore, the possibility of novel pan-
demic strains being generated from reassortment between avian
H9N2 and H1N1/2009 influenza viruses exists.
Results
Replication Abilities of the Reassortant Viruses Between HB08 and
BJ09 in Vitro. To investigate the properties of H9 subtype avian–
pandemic reassortant viruses, we first established an eight-plasmid
reversegenetics(rg)systemofthetwoparentalviruses,avianH9N2
influenza virus (HB08) and 2009 pandemic H1N1 influenza virus
(BJ09). Transfections of all of the 127 combinations of plasmids
between these two parental viruses with the HA gene from avian
H9N2 virus were performed. At 72 h posttransfection, culture su-
pernatant was inoculated into 10-d-old specific-pathogen-free
(SPF) embryonated chicken eggs to produce virus stocks. The titer
of each virus was determined by plaque assay on Madin–Darby
canine kidney (MDCK) cells. The rgHB08 and rgBJ09 viruses
replicated efficiently in vitro with titers of 8.9 log10pfu/mL and 7.5
log10pfu/mL, respectively. Given the titers achieved for each reas-
sortant, the 127 reassortants could be categorized into four groups:
i) Reassortants with a high replication ability (replication group
I). A total of 73 reassortants (∼58% of the 127 reassortants)
possessed replication ability comparable to that of HB08
and BJ09 viruses with titers ≥7.0 log10pfu/mL (Table S1).
ii) Reassortants with a moderate replication ability (replication
group II). This group contained 36 reassortants (∼28% of
reassortants) with titers ranging from 4 to 6 log10pfu/mL
(Table S2).
iii) Reassortants with a low replication ability (replication group
III). Five influenza viruses (∼4% of reassortants) grew
poorly in vitro with titers between 1 and 3 log10pfu/mL,
suggesting that the replication of these viruses was im-
paired (Table S3).
iv) Nonviable reassortants (replication group IV). A total of 13
reassortants (∼10% of reassortants) had severe replication
defects characterized by the absence of any observable cy-
topathic effects (CPE) or visible plaques in MDCK cells
(Table S4).
To determine the effect of a single genomic segment on virus
replication in vitro, two viruses with only one gene segment from
different origins were paired and analyzed by the paired-samples
t test. Statistical analyses suggested that the introduction of BJ09
PB2 (P < 0.01) or NS (P < 0.05) could enhance the replication of
reassortants, and the introduction of HB08 PB1 (P < 0.01), NP
(P < 0.01), NA (P < 0.01), or M (P < 0.01) usually enhanced
reassortant virus replication (Table S5). No significant difference
in viral titers was observed between the pairs of viruses with HB08
PA or BJ09 PA in the same genetic background (P > 0.05).
The interaction between two segments of all combinations was
also analyzed (Table S6). We noted that some gene segments
could coordinate better with heterologous genes than with ho-
mologous genes. For example, BJ09 PB2 usually enhanced the
replication of the reassortants with HB08 M (P < 0.05) in the
same genetic background; and BJ09 NS increased the replication
ability of the reassortants with HB08 PB2 (P < 0.05) or NA (P <
0.01). HB08 PB1 enhanced the reassortants with BJ09 PB2 in the
same genetic background (P < 0.01); HB08 NP enhanced the
reassortants with BJ09 PB1 (P < 0.01), PA (P < 0.01), NA (P <
0.01), or NS (P < 0.01); HB08 NA enhanced the reassortants
with BJ09 PB2 (P < 0.01), PA (P < 0.01), NP (P < 0.01), M (P <
0.05), or NS (P < 0.01); and HB08 M enhanced the reassortants
with BJ09 NA (P < 0.01).
Taken together, our data show that the gene segments from
avian H9N2 and pandemic H1N1 viruses could readily coexist
and functionally coordinate, indicating that these two viruses had
high genetic compatibility.
Pathogenicity of the Reassortants in Mice. The 50% mouse in-
fectious dose (MID50), 50% mouse lethal dose (LD50), and viral
replication in tissues, including lung, nasal turbinate, brain, liver,
spleen, kidney, and colon, of two parental viruses in BALB/c
mice were evaluated. Both of the parental viruses readily infec-
ted mice (MID50< 101pfu), and the LD50of HB08 and BJ09
viruses was 105.5and 104.8pfu, respectively. These viruses rep-
licated efficiently in the lung and nasal turbinate (>107pfu/mL)
on day 4 following an intranasal inoculation of 104pfu (Table S7),
whereas no virus was detected in other tissues. With the 105pfu
inoculation dose, HB08 virus caused 8.1% weight loss and did not
cause any deaths, and BJ09 virus caused 15.1% weight loss with
one mouse death.
We next evaluated the pathogenicity in mice of all 73 reassor-
tants in replication group I in vitro (with titer ≥7.0 log10pfu/mL).
All tested reassortants readily infected the mice (MID50< 101.5
pfu) and could replicate in the nasal turbinate (Table S7), but
were not detected in the other tissues. These reassortants were
sorted into three groups according to their pathogenicity in mice:
i) Reassortants of higher pathogenicity than both parental viruses
(pathogenicity group I). A total of 8 reassortants (∼11% of
the 73 tested reassortants) with LD50 102.8to 104.5pfu
belonged to this group (Table 1). The mean survival time
(MST) following infection with 105pfu virus was 5–6.7 d.
Obvious clinical signs, such as ruffled coat, lethargy, and
dyspnea could be observed. Hematoxylin and eosin (H&E)
staining showed that viruses of this group caused severe
interstitial pneumonia and bronchopneumonia, character-
ized by edema, hemorrhages, serious dropout of epithelial
cells, and extensive infiltration of inflammatory cells includ-
ing lymphocytes, neutrophils, and plasma cells (Fig. 1 A and
B). To determine if there was any mutation contributing to
the high-pathogenicity phenotype, the viruses of this group
were fully sequenced. No mutations were found, indicating
that the high-pathogenicity phenotype is an intrinsic prop-
erty of these reassortants.
ii) Reassortantswithpathogenicitysimilartothatof theBJ09virus
(pathogenicity group II). A total of 11 reassortants (∼15% of
the tested reassortants) with pathogenicity comparable to
wild-type BJ09 virus were tested in vivo. Typically, one to
three of the four mice inoculated died after infection with
104pfu of these viruses. The LD50was determined to be
between 104.8and 105.3pfu and the MST 7.7–12.3 d (Table
2). These viruses caused moderate bronchopneumonia sim-
ilar to both parental viruses (Fig. 1 C and D).
iii) Reassortants with pathogenicity similar to or lower than that
of the HB08 virus (pathogenicity group III). This group con-
tained eight reassortants (∼11% of the tested reassortants).
Similar to wild-type HB08 virus, they did not cause any
death, but resulted in >5% body weight loss at a 105-pfu
inoculation dose (Table 3). Mild to moderate broncho-
pneumonia was observed in the H&E-stained lung tissues
of mice inoculated with these viruses (Fig. 1E).
iv)Reassortantsoflowerpathogenicitythanthatofparentalviruses
(pathogenicity group IV). A total of 46 reassortants (∼63% of
thetestedreassortants)belongedtothisgroup(Table4).All
mice inoculated with viruses of this group survived with no
obvious weight loss or other clinical signs. H&E staining
showed that lesions were mild or absent in the lung tissues
of mice infected with this group of viruses (Fig. 1F).
Sun et al. PNAS
| March 8, 2011
| vol. 108
| no. 10
| 4165
MICROBIOLOGY

Page 3

It is noteworthy that all eight reassortant viruses belonging to
pathogenicity group I possessed the PA gene derived from the
pandemic BJ09 virus, indicating that BJ09 PA is a prerequisite
for the emergence of highly virulent reassortant viruses. To
evaluate the contributions to pathogenicity made by the H1N1/
2009 virus genes of the reassortant viruses, we analyzed patho-
genicity data of single-gene reassortant viruses, each containing
one gene segment from the H1N1/2009 virus. We found that only
reassortants with BJ09 PA (r3 virus in Table 1) had much higher
pathogenicity compared with parental viruses. The single reas-
sortant with BJ09 NA (r6 virus in Table 2) exhibited similar
pathogenicity to those of the HB08 virus, whereas the pathoge-
nicity of the other five single reassortants (r1, r2, r5, r7, and r8
viruses in Table 4) was lower than the parental viruses. Analysis
of the paired viruses with or without BJ09 PA in the same genetic
background included a total of 24 pairs of viruses. Among these
viruses, 11 reassortants possessing BJ09 PA had higher patho-
genicity than the corresponding viruses with a PA gene derived
from the HB08 virus, whereas only two viruses with BJ09 PA had
lower pathogenicity than the corresponding reassortants without
BJ09 PA (Table S8). The other 11 pairs of viruses had no
pathogenic difference with or without the BJ09 PA gene. These
results emphasize the important effect of BJ09 PA on the
pathogenicity of reassortants in mice.
By comparing the 24 pairs of viruses with or without BJ09 PB1
in the same genetic background, we found that 12 viruses with
BJ09 PB1 had increased pathogenicity after introducing HB08
PB1, and only one had reduced pathogenicity (Table S9). The
other 11 pairs of viruses had no pathogenic difference with or
without the BJ09 PB1 gene. These results suggested that the
BJ09 PB1 gene usually attenuated the pathogenicity of reassor-
tants in mice. By analyzing other segments in the same way, we
found that BJ09 NA typically increased the pathogenicity of
reassortants in mice (Table S10), whereas the influence of BJ09
PB2, NP, M, or NS was not obvious.
Taken together, the pathogenicity analyses indicated that the
PA gene of H1N1/2009 virus origin is indispensable for reassor-
tant viruses to be more virulent in mice. The NA gene from the
pandemic H1N1/2009 virus also typically increased the pathoge-
nicity of reassortant viruses in mice. In contrast, the PB1 gene
from H1N1/2009 virus attenuated viruses. Other pandemic virus
Table 1. Reassortants of higher pathogenicity than both parental viruses (pathogenicity group I)
“r” in virus names denotes reassortant. The numbers in the virus names indicate gene segments derived from the BJ09 virus as follows:
1, PB2; 2, PB1; 3, PA; 4, HA; 5, NP; 6, NA; 7, M; and 8, NS. The virus segments derived from the HB08 virus were not assigned numbers.
Segments derived from BJ09 virus are in yellow, and those derived from HB08 virus are in blue. The pathogenicity of the reassortant viruses
wasdeterminedasdescribedinMaterialsandMethods.Maximummeanweightlosswasdeterminedfromthreemice(percentageweightloss
relative to day 0 p.i.) following infection with 105pfu of virus. MST, mean survival time of mice infected with 105pfu.
Fig. 1.
sues. (A) r3/5/6/8 virus and (B) r3/6/7/8 virus showing severe bronchopneu-
monia and interstitial pneumonia; edema and bronchial epithelial cell
desquamation (thick solid arrow); massive immune cell infiltrates around
bronchi and blood vessels including lymphocytes, neutrophils, and plasma
cells (thick white arrow); alveolar wall thickening (▲); and alveolar epithelial
cells, erythrocytes, and inflammatory cells within alveolar spaces (△). (C)
WTBJ09 virus and (D) r1/2/3/5/6/8 virus showing moderate bronchopneu-
monia; edema, severe desquamation, and lesion of bronchial epithelial cells
(thick solid arrow), and presence of inflammatory cells around bronchi and
blood vessels (thick open arrow). (E) WTHB08 virus. Mild bronchopneumo-
nia. Mild desquamation of bronchial epithelial cells (thick solid arrow). (F) r1
virus. Except for congestion, no obvious lesions. (Scale bar: 50 μm.)
Representative histopathological changes in H&E-stained lung tis-
Table 2.
(pathogenicity group II)
Reassortants with pathogenicity similar to BJ09 virus
Symbols and nomenclature of the viruses are as described in the legend in
Table 1.
4166
| www.pnas.org/cgi/doi/10.1073/pnas.1019109108Sun et al.

Page 4

gene segments, including the PB2, NP, M, and NS genes, did not
significantly influence the pathogenicity of reassortant viruses.
In Vitro Viral Polymerase Activity. Previous studies demonstrated
that the viral ribonucleoprotein (RNP) complex hadan important
correlation with viral generation, replication, and pathogenicity
(27, 28). To study the mechanisms underlying the differences in
thereplicationandpathogenicity phenotypes oftheH9-pandemic
reassortants, the activity of 16 RNP combinations of PB2, PB1,
PA, and NP from either HB08 or BJ09 viruses was determined by
measuring the activity of luciferase at 33 °C or 37 °C, which
mimics the temperature of the upper and lower respiratory tract
as previously described (25, 26). As shown in Fig. 2, the avian
HB08 RNP (HPB2HPB1HPAHNP; “H” stands for HB08 virus) ac-
tivity at 33 °C was ∼40% lower than that at 37 °C. In contrast, the
RNP (BPB2BPB1BPABNP; “B” stands for BJ09 virus) activity of
pandemic BJ09 at 33 °C was comparable to that at 37 °C. All 13
viruses in replication group IV (nonviable reassortants) possessed
RNP complexes with low activity (<20%) in the minigenome
assay, with only two exceptions (r2/3/6/7 and r3/5/6/7 viruses),
indicating that RNP activity plays an important role in the
emergence and replication ability of reassortants. Although RNP
activity for most reassortant viruses in the high replication group
was high or moderate, some reassortants with a low RNP complex
were alsoinclude in this group, suggesting thatother factors could
influence viral replication, such as HA, NA, M, and NS genes that
were not included in this assay. The RNP constituted by PA from
BJ09 virus resulted in high polymerase activities at 37 °C. For
example, six of seven hybrid RNP combinations containing BJ09
PA protein showed higher activity than the BPB2BPB1BPABNP.
Consistently, all of the viruses in pathogenicity group I contained
the BJ09 PA gene. Furthermore, all three RNP combinations
(HPB2HPB1BPAHNP, HPB2HPB1BPABNP, and BPB2BPB1BPABNP)
included in pathogenicity group I also showed high activities
at 33 °C.
Discussion
Reassortment is an important mechanism for the evolution of
influenza viruses that could lead to antigenic shift and the gen-
eration of pandemic strains (16). The coexistence of pandemic
H1N1/2009 influenza virus and some avian influenza viruses (such
as avian H5N1 and H9N2 viruses) in humans and pigs provides
an opportunity for the incorporation of avian virus genes into
mammalian-adapted viruses that results in the emergence of novel
viruses with considerable potential threat to public health. Several
studies focused on the reassortment between highly pathogenic
avian H5N1 and pandemic H1N1 viruses (29, 30). However, all
previous influenza pandemic strains evolved from viruses with low
pathogenicity (16). To evaluate the potential public risk of avian
H9N2–pandemic H1N1 reassortants, we systematically studied the
127 combinations of reassortant viruses derived from an avian
H9N2 influenza virus and a 2009 pandemic H1N1 influenza virus.
All these reassortant viruses had the HA gene of H9 origin. High
genetic compatibility was observed between avian H9N2 and
pandemic H1N1 influenza viruses, and eight reassortants pos-
sessed enhanced virulence in mice compared with parental viruses.
Using reverse genetics, previous studies systematically gener-
ated avian H5N1–human H3N2 influenza reassortants and
evaluated their genetic compatibility (25, 26). Chen et al. (26)
used reverse genetics to generate 63 reassortant viruses derived
from A/Thailand/16/04 (H5N1) and A/Wyoming/3/03 (H3N2)
viruses, containing the H5N1 surface protein genes and one extra
single-gene reassortant virus, r6, possessing the A/Wyoming/3/03
NA in the background of the A/Thailand/16/04 virus. Among the
64 reassortants, ∼44% of the reassortants yielded wild-type or
near-wild-type replication efficiency in MDCK cell culture, and
∼20% of the reassortants were nonviable or marginally viable. Li
et al. (25) attempted to generate all 254 combinations of reas-
sortant viruses between A/chicken/South Kalimantan/UT6028/06
(H5N1) and A/Tokyo/Ut-Sk-1/07 (H3N2) influenza viruses.
Among the 127 reassortants with HA genes of H5 origin, ∼48%
viruses grew comparable to their parental viruses and ∼21%
reassortants produced no visible CPE. In the present study,
among the 127 H9 reassortants between avian H9N2 and 2009
pandemic H1N1 viruses, up to ∼57.5% of these viruses had a
high replication ability similar to their parental viruses, and only
∼10.2% reassortants had no visible CPE. The higher percentage
of H9 reassortants with efficient replication ability and the lower
percentage of nonviable viruses compared with reassortants de-
rived from avian H5N1 and seasonal human H3N2 influenza
viruses suggested that avian H9N2 and pandemic H1N1 in-
fluenza viruses have a high genetic compatibility.
Table 3.
(pathogenicity group III)
Reassortants with pathogenicity similar to HB08 virus
Symbols and nomenclature of the viruses are as described in the legend in
Table 1. None of these viruses killed mice at 105pfu. Therefore, the LD50was
assigned as ≥105.5pfu and the MST was >14 d.
Table 4.
viruses (pathogenicity group IV)
Reassortants of lower pathogenicity than parental
Symbols and nomenclature of the viruses are as described in the legend in
Table 1. None of these viruses killed mice at 105pfu. Therefore, the LD50was
assigned as ≥105.5pfu and the MST was >14 d.
Sun et al. PNAS
| March 8, 2011
| vol. 108
| no. 10
| 4167
MICROBIOLOGY

Page 5

The pathogenicity of all 73 reassortants with high replication
ability (replication group I) was evaluated in a BALB/c mouse
model. All viruses tested readily infected mice and could be
detected in nasal turbinate. Importantly, eight of these viruses
possessed higher pathogenicity in mice than their parental viru-
ses. These results indicated that the reassortment between avian
H9N2 and pandemic H1N1 influenza viruses could generate
reassortants that were more virulent for mammals than their
parental viruses. Additionally, the other less virulent reassortants
also readily infected mice, suggesting that these viruses were able
to circulate in a mammalian host that could easily be ignored and
evolve by adaptive mutation or further reassortment capable of
posing a potential threat for humans. Thus, the present study
highlights the potential risk of such reassortants derived from
avian H9N2 and pandemic H1N1 viruses to public health.
Our findings demonstrated that the PA gene from the BJ09
virus may significantly contribute to the pathogenicity of avian–
human reassortants. H9N2 virus that was replaced only by
a single PA gene from pandemic H1N1 virus could substantially
increase its pathogenicity in mice. Although the introduction of
the PA gene from pandemic H1N1/2009 virus into avian H9N2
virus did not necessarily result in a more pathogenic phenotype,
our results revealed that the PA gene of H1N1/2009 origin is
a prerequisite for the emergence of reassortants with increased
pathogenicity. It is known that the PA gene of pandemic H1N1
(2009) virus was of North American avian origin (16, 31). Thus,
the pandemic PA gene may be compatible with the other gene
segments from avian H9N2 viruses as a prerequisite for the in-
crease of viral pathogenicity. Furthermore, a previous study
found that the introduction of a human influenza PA gene into an
avian polymerase gene could overcome restriction of avian
polymerase in human cells (32). Similarly, in the present study,
the PA gene of pandemic H1N1/2009 origin could increase the
polymerase activity of avian-pandemic RNP, especially for the
HPB2HPB1BPAHNP combination. Therefore, the PA gene of
pandemic H1N1/2009 origin that was adapted to humans might
contribute to the high polymerase activity leading to enhanced
virulence of the reassortants in mice.
Pathogenicity studies in mice demonstrated that the effect of
a single gene segment on pathogenicity was influenced by the
other seven segments. For example, although the single PA gene
replaced by those of 2009 pandemic virus origin in the back-
ground of H9N2 influenza virus could increase virulence in mice,
not all of the H9 reassortants possessing the PA gene from the
pandemic H1N1/2009 virus had a higher virulence than their
parental viruses, suggesting that other gene segments influenced
the effect of the pandemic-origin PA gene. Indeed, the patho-
genesis of influenza virus includes polygenic virulence factors
(33). Li et al. (25) reported that reassortment between two in-
fluenza viruses results in the emergence of reassortant viruses
with higher pathogenicity than parental viruses. Their study
evaluated the pathogenicity of 75 reassortant viruses derived
from avian H5N1 and human H3N2 influenza viruses and found
that 10 reassortants were more virulent than parental viruses.
Although some H9-pandemic reassortant viruses (pathoge-
nicity group I viruses) possessed higher virulence than parental
viruses, none resulted in systemic spread in mice, as observed for
the highly pathogenic H5N1 influenza virus. These reassortants
were restricted to the respiratory system. However, we found
that all reassortants replicated efficiently in the lungs of mice and
showed high polymerase activity. These results suggested that
a high virus load and high polymerase activity were important
factors for the virulence of the reassortants in mice. Additionally,
we observed that inflammatory cell infiltration in the lungs of
infected mice by pathogenicity group I viruses was significantly
greater than that of less virulent reassortant viruses, suggesting
that a severe inflammatory response is likely to be induced by the
reassortant viruses with higher pathogenicity.
In summary, we found that the contemporary avian H9N2
influenza virus and pandemic H1N1 virus showed high genetic
compatibility and could generate reassortants with higher path-
ogenicity than parental viruses, indicating that H9N2 influenza
virus could contribute to the evolution of viruses with an in-
creased threat for public health by reassorting with the pandemic
H1N1/2009 virus. Our findings also demonstrate that the PA
gene of pandemic origin plays an important role in determining
the high pathogenicity of H9 reassortants and high RNP activity,
which could potentially serve as an indicator for recognizing
reassortants with high risk to public health.
Materials and Methods
Viruses and Cells. A/chicken/Hebei/LC/2008 (H9N2) (HB08) virus was isolated
from diseased chicken in Hebei, China, in January 2008 and propagated in
10-d-old SPF embryonated chicken eggs. The phylogenetic and genetic
properties ofHB08virushavebeendescribedpreviously(4). A/Beijing/16/2009
(H1N1) (BJ09) virus was isolated from a patient with flu-like symptoms in
Beijing, China, in November 2009 and grown in MDCK cells. Human em-
bryonic kidney cells (293T) and MDCK cells were grown in DMEM (Invi-
trogen) containing 10% FBS (Invitrogen). Viral infectivity was determined by
plaque assay on MDCK cells, as previously described (34).
Generation of Reassortant Viruses by Reverse Genetics. Reverse transcription-
PCR (RT-PCR) amplicons of the eight viral genes from HB08 and BJ09 viruses
were cloned into a dual-promoter plasmid, PHW2000. MDCK and 293T cells
were cocultured and transfected with 0.5 μg of each of the eight plasmids
and 10 μL lipofectamine 2000 (Invitrogen) in a total volume of 1 mL of Opti-
MEM (Invitrogen). After incubation at 37 °C for 6 h, the transfection mixture
was removed from the cells and 2 mL of Opti-MEM containing 1 μg/mL of
tosylphenylalanine chloromethyl ketone treated trypsin (TPCK-trpsin, Wor-
thington Biochemical Corporation) was added. After 72 h, the supernatant
was inoculated in 10-d-old SPF embryonated chicken eggs to produce stock
viruses. The titers of the stock viruses were determined by plaque assay on
MDCK cells. Viral RNA was extracted and analyzed by RT-PCR, and each viral
segment was sequenced to confirm the identity of the virus. All experiments
with live viruses and transfectants generated by reverse genetics were per-
formed in a biosafety level 3 containment laboratory approved by the
Ministry of Agriculture of the People’s Republic of China.
Fig. 2.
BJ09 viruses. The 293T cells were transfected in duplicate with luciferase re-
porter plasmid p-Luci and internal control plasmid Renilla, together with plas-
mids expressing PB2, PB1, PA, and NP from either HB08 or BJ09 viruses.
Segments derived from the BJ09 virus are in yellow, and those derived from
HB08 virus are in blue. Virus number corresponding to the RNP combination of
virusesislisted.Cellswereincubatedat33°C(stippledbars)orat37°C(solidbars)
for 24 h, and cell lysates were analyzed to measure Firefly and Renilla lucif-
erase activities. Values shown are the mean ± SD of the three independent
experiments and are standardized to those ofBJ09 measured at 37 °C (100%).
Polymerase activity of 16 RNP combinations between the HB08 and
4168
| www.pnas.org/cgi/doi/10.1073/pnas.1019109108 Sun et al.

Page 6

Pathogenicity Test in Mice. Groups of 6- to 8-wk-old female BALB/c mice (Vital
River Laboratory) were anesthetized with Zoletil (tiletamine-zolazepam;
Virbac; 20 μg/g) and intranasally inoculated with 50 μL of serial 10-fold
dilutions of infectious virus in PBS. For reassortant viruses, 105pfu virus was
the highest dose used for inoculation. Three mice from each group were
euthanized on day 4 postinfection (p.i.) and tissues, including nasal turbi-
nate, lung, brain, liver, spleen, kidney, and colon, were harvested for plaque
assay in MDCK. Additionally, a portion of each lung tissue was immersed in
10% neutral buffered formalin solution, routinely processed, and embedded
in paraffin. Five-micrometer sections were stained with H&E. The remaining
three mice in the 105and 104pfu groups were monitored daily for 14 d for
weight loss and mortality. Mice that lost >20% of their body weight were
euthanized. MID50and LD50were calculated and expressed as the pfu value
corresponding to 1 MID50or LD50.
RNP Minigenome Luciferase Assay. Four expression plasmids of PB2, PB1, PA,
and NP (50ng each) from either HB08 or BJ09 viruses were cotransfected into
293T cells together with the luciferase reporter plasmid p-Luci (10 ng)
and internal control plasmid Renilla (5 ng). The assay was performed at both
33 and 37 °C. At 24 h posttransfection, cell lysate was prepared with Dual-
Luciferase Reporter Assay System (Promega) and luciferase activity was
measured using GloMax 96 microplate luminometer (Promega).
Statistical Analysis. To determine the effect of a single genomic segment on
virus replication in vitro, viruses with a different origin from this segment but
in the same background were paired. For each gene segment, the statistical
significance of a group of 64 virus pairs was determined by paired-sampled
t test. If the P value (two-tailed) of the t test was <0.05, the mean titers of
viruses with or without this segment were compared in order to evaluate
the effect of this segment for virus replication. If P > 0.05, we concluded
there was no significant effect of this segment on virus replication.
To evaluate the influence of one segment on another in virus replication,
viruses with one segment and with or without another segment in the same
background were paired. The statistical significance of a group of 32 virus
pairs was determined by paired-sampled t test. If the P value (two-tailed) of
the t test was <0.05, the mean titers of these viruses with or without another
segment were compared in order to evaluate the effect of another segment
in virus replication. If P > 0.05, we concluded that there was no significant
effect of another segment on this segment in virus replication.
ACKNOWLEDGMENTS. We thank Huijie Gao, Honglei Sun, and Yuan Ma for
excellent technical assistance. This work was supported by the National
Science Fund for Distinguished Young Scholars (31025029), the National
Basic Research Program (973 Program, 2011CB504702), the National Natu-
ral Science Foundation of China (30901072), the Fundamental Research
Funds for the Central Universities (2009JC02), the Specialized Research Fund
for the Doctoral Program of Higher Education (20090008120007), and
the Program for Cheung Kong Scholars and Innovative Research Teams in
Chinese Universities (IRT0866). J.L. was funded by the Taishan Scholar
Foundation.
1. Xu KM, et al. (2007) The genesis and evolution of H9N2 influenza viruses in poultry
from southern China, 2000 to 2005. J Virol 81:10389–10401.
2. Panshin A, et al. (2010) Variability of NS1 proteins among H9N2 avian influenza
viruses isolated in Israel during 2000-2009. Virus Genes 41:396–405.
3. Abolnik C, et al. (2010) Phylogenetic analysis of influenza A viruses (H6N8, H1N8,
H4N2, H9N2, H10N7) isolated from wild birds, ducks, and ostriches in South Africa
from 2007 to 2009. Avian Dis 54 (1 Suppl)313–322.
4. Sun Y, et al. (2010) Genotypic evolution and antigenic drift of H9N2 influenza viruses
in China from 1994 to 2008. Vet Microbiol 146:215–225.
5. Cong YL, et al. (2007) Antigenic and genetic characterization of H9N2 swine influenza
viruses in China. J Gen Virol 88:2035–2041.
6. Butt KM, et al. (2005) Human infection with an avian H9N2 influenza A virus in Hong
Kong in 2003. J Clin Microbiol 43:5760–5767.
7. Yu H, et al. (2010) Genetic diversity of H9N2 influenza viruses from pigs in China: A
potential threat to human health? Vet Microbiol, 10.1016/j.vetmic.2010.11.008.
8. Butt AM, Siddique S, Idrees M, Tong Y (2010) Avian influenza A (H9N2):
Computational molecular analysis and phylogenetic characterization of viral surface
proteins isolated between 1997 and 2009 from the human population. Virol J 7:319.
9. Peiris JS, et al. (2001) Cocirculation of avian H9N2 and contemporary “human” H3N2
influenza A viruses in pigs in southeastern China: Potential for genetic reassortment?
J Virol 75:9679–9686.
10. Lin YP, et al. (2000) Avian-to-human transmission of H9N2 subtype influenza A
viruses: Relationship between H9N2 and H5N1 human isolates. Proc Natl Acad Sci
USA 97:9654–9658.
11. Chen YM, Ge WY, Huang C, Lai ZP, Fan XH (2008) Serological survey of antibody to H9
and H6 subtypes of bird flu virus in healthy youths in Guanxi. China Trop Med 8:
985–986.
12. Liang Q, et al. (2003) Seroepidemiologic survey for antibody of healthy youth to H9,
H6 and H5 subtypes of influenza A virus in Chaoshan Area. J. Shantou University
Medical College 16:107–112.
13. Matrosovich MN, Krauss S, Webster RG (2001) H9N2 influenza A viruses from poultry
in Asia have human virus-like receptor specificity. Virology 281:156–162.
14. Trifonov V, Khiabanian H, Rabadan R (2009) Geographic dependence, surveillance,
and origins of the 2009 influenza A (H1N1) virus. N Engl J Med 361:115–119.
15. Smith GJ, et al. (2009) Origins and evolutionary genomics of the 2009 swine-origin
H1N1 influenza A epidemic. Nature 459:1122–1125.
16. Neumann G, Noda T, Kawaoka Y (2009) Emergence and pandemic potential of swine-
origin H1N1 influenza virus. Nature 459:931–939.
17. Sreta D, et al. (2010) Pandemic (H1N1) 2009 virus on commercial swine farm, Thailand.
Emerg Infect Dis 16:1587–1590.
18. Weingartl HM, et al. (2010) Genetic and pathobiologic characterization of pandemic
H1N1 2009 influenzaviruses froma naturally infected swine herd. J Virol 84:2245–2256.
19. Pasma T, Joseph T (2010) Pandemic (H1N1) 2009 infection in swine herds, Manitoba,
Canada. Emerg Infect Dis 16:706–708.
20. Kaverin N (2010) Postreassortment amino acid substitutions in influenza A viruses.
Future Microbiol 5:705–715.
21. Taubenberger JK, Kash JC (2010) Influenza virus evolution, host adaptation, and
pandemic formation. Cell Host Microbe 7:440–451.
22. Rajagopal S, Treanor J (2007) Pandemic (avian) influenza. Semin Respir Crit Care Med
28:159–170.
23. Hsieh YC, et al. (2006) Influenza pandemics: Past, present and future. J Formos Med
Assoc 105:1–6.
24. Vijaykrishna D, et al. (2010) Reassortment of pandemic H1N1/2009 influenza A virus in
swine. Science 328:1529.
25. Li C, et al. (2010) Reassortment between avian H5N1 and human H3N2 influenza viruses
creates hybrid viruses with substantial virulence. Proc Natl Acad Sci USA 107:4687–4692.
26. Chen LM, Davis CT, Zhou H, Cox NJ, Donis RO (2008) Genetic compatibility and
virulence of reassortants derived from contemporary avian H5N1 and human H3N2
influenza A viruses. PLoS Pathog 4:e1000072.
27. Li C, Hatta M, Watanabe S, Neumann G, Kawaoka Y (2008) Compatibility among
polymerase subunit proteins is a restricting factor in reassortment between equine
H7N7 and human H3N2 influenza viruses. J Virol 82:11880–11888.
28. Leung BW, Chen H, Brownlee GG (2010) Correlation between polymerase activity and
pathogenicity in two duck H5N1 influenza viruses suggests that the polymerase
contributes to pathogenicity. Virology 401:96–106.
29. Zhang Y, et al. (2010) Neuraminidase and hemagglutinin matching patterns of a highly
pathogenic avian and two pandemic H1N1 influenza A viruses. PLoS ONE 5:e9167.
30. Octaviani CP, Ozawa M, Yamada S, Goto H, Kawaoka Y (2010) High level of genetic
compatibility between swine-origin H1N1 and highly pathogenic avian H5N1
influenza viruses. J Virol 84:10918–10922.
31. Garten RJ, et al. (2009) Antigenic and genetic characteristics of swine-origin 2009 A
(H1N1) influenza viruses circulating in humans. Science 325:197–201.
32. Mehle A, Doudna JA (2009) Adaptive strategies of the influenza virus polymerase for
replication in humans. Proc Natl Acad Sci USA 106:21312–21316.
33. Chen H, et al. (2007) Polygenic virulence factors involved in pathogenesis of 1997
Hong Kong H5N1 influenza viruses in mice. Virus Res 128:159–163.
34. Kawaoka Y, Naeve CW, Webster RG (1984) Is virulence of H5N2 influenza viruses in
chickens associated with loss of carbohydrate from the hemagglutinin? Virology 139:
303–316.
Sun et al.PNAS
| March 8, 2011
| vol. 108
| no. 10
| 4169
MICROBIOLOGY