Virologica Sinica: Research Article
Virologica Sinica: Research Article
Virologica Sinica
journal homepage: www.keaipublishing.com/en/journals/virologica-sinica
www.virosin.org
Research Article
A R T I C L E I N F O A B S T R A C T
Keywords: Since mid-2016, the low pathogenic H7N9 influenza virus has evolved into a highly pathogenic (HP) phenotype in
H7N9 China, raising many concerns about public health and poultry industry. The insertion of a “KRTA” motif at
Highly pathogenic hemagglutinin cleavage site (HACS) occurred in the early stage of HP H7N9 variants. During the co-circulation,
Cleavage site
the HACS of HP-H7N9 variants were more polymorphic in birds and humans. Although HP-H7N9 variants, unlike
Basic and non-basic amino acids
Pathogenicity
the H5 subtype virus, exhibited the insertions of basic and non-basic amino acids, the underlying function of those
insertions and substitutions remains unclear. The results of bioinformatics analysis indicated that the
PEVPKRKRTAR/G motif of HACS had become the dominant motif in China. Then, we generated six H7N9 viruses
bearing the PEIPKGR/G, PEVPKGR/G, PEVPKRKRTAR/G, PEVPKGKRTAR/G, PEVPKGKRIAR/G, and
PEVPKRKRR/G motifs. Interestingly, after the deletion of threonine and alanine (TA) at HACS, the H7N9 viruses
manifested decreased thermostability and virulence in mice, and the PEVPKRKRTAR/G-motif virus is prevalent in
birds and humans probably due to its increased transmissibility and moderate virulence. By contrast, the insertion
of non-basic amino acid isoleucine and alanine (IA) decreased the transmissibility in chickens and virulence in
mice. Remarkably, the I335V substitution of H7N9 virus enhanced infectivity and transmission in chickens,
suggesting that the combination of mutations and insertions of amino acids at the HACS promoted replication and
pathogenicity in chickens and mice. The ongoing evolution of H7N9 increasingly threatens public health and
poultry industry, so, its comprehensive surveillance and prevention of H7N9 viruses should be pursued.
1. Introduction deaths had been reported. Previous studies indicated that the 2013 H7N9
influenza viruses isolated from birds were low pathogenic (LP) for
Since its emergence in 2013, the novel H7N9 influenza virus has chickens (Liu et al., 2013); however, by mid-2016, a novel H7N9 variant
continued to circulate in Chinese mainland and has caused six waves of possessing combined insertions of basic and non-basic amino acids at the
human infection (Zhu et al., 2018), while the first four of which causes hemagglutinin cleavage site (HACS) was first isolated from Guangdong
773 cases of human infection with a fatality rate of 42% (Wang et al., Province of China and was shown to be highly pathogenic (HP) in
2017). As of May 2021, a total of 1568 human cases and 616 related chickens (Jiang et al., 2017; Zhang et al., 2017). In the fifth wave that
* Corresponding authors. College of Veterinary Medicine, South China Agricultural University, Guangzhou, 510642, China.
E-mail addresses: [email protected] (M. Liao), [email protected] (W. Qi).
1
Aobaixue Zhou and Jiahao Zhang contributed equally to this work.
https://doi.org/10.1016/j.virs.2022.01.001
Received 2 April 2021; Accepted 22 July 2021
Available online 13 January 2022
1995-820X/© 2022 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
followed, the co-circulation of highly pathogenic avian influenza (HPAI) before May 2021. The distributions of influenza A (H7N9) viruses
H7N9 viruses caused high genetic diversity and host adaption, posing available from GenBank (http://ncbi.nlm.nih.gov/genbank/) and Global
serious public health concerns (Quan et al., 2018). Other studies have Initiative on Sharing All Influenza Data (https://www.gisaid.org) before
also indicated that those H7N9 variants are not only HP to mice but also May 2021 were calculated in GraphPad Prism 7 (https://www.graphp
highly transmissible in ferrets, posing the threats of pandemic potential ad.com/) and ArcGIS 10.4 for Desktop (http://www.esri.com/softwar
in humans (Imai et al., 2017; Shi et al., 2018; Yang et al., 2018). After the e/arcgis/arcgis-for-desktop/).
H5/H7 bivalent vaccine for poultry was used in September 2017, the
prevalence of H7N9 viruses in birds and humans decreased dramatically 2.2. Cells and viruses
(Shi et al., 2018). However, these HP-H7N9 viruses re-emerged during
2019 and several immune escape mutations were observed (Zhang et al., Human embryonic kidney (HEK293T) cells, Madin-Darby canine
2020), indicative of the antigenic drift of H7N9 variants. kidney (MDCK), chicken embryo fibroblast (CEF), and human lung
Hemagglutinin (HA) is particularly important to the infectivity and adenocarcinoma (A549) cells were maintained in Dulbecco's modified
pathogenicity of the viral particles of influenza viruses (Webster and Eagle's medium (DMEM, Gibco, 11054020, Paisley, Scotland, UK) with
Rott, 1987). An essential step in the infection of influenza virus is the 10% fetal bovine serum (FBS, Biological Industries, 04-011-1A, Kibbutz,
cleavage of the HA precursor, HA0, into subunits HA1 and HA2 through Israel) and 1% penicillin/streptomycin (Gibco, 15140122) at 37 C in an
intracellular proteases (Rott, 1992). The insertion of multiple basic atmosphere of 5% CO2. Both A/chicken/Guangdong/G1/2013 [H7N9]
amino acids at the HACS had been described in the H5 subtype avian (G1) and A/chicken/Guangdong/HZ-3/2016 [H7N9] (HZ-3) of H7N9
influenza virus (AIV) via repeated passaging in chickens (Ito et al., 2001). were obtained from the National and Local Joint Engineering Laboratory
Unlike HP-H5 subtype viruses, HP-H7 subtype viruses showed the di- for the Medicine of Zoonosis Prevention and Control at South China
versity of HACS, including the insertion of basic and non-basic amino Agricultural University's College of Veterinary Medicine. Viruses were
acids at HACS. Previous studies have demonstrated that the insertion of propagated in 9- to 11-day-old specific-pathogen-free (SPF) embryonated
basic amino acids at HACS in H7 subtype viruses plays an important role chicken eggs and stored at 80 C. All experiments with H7N9 influenza
in HA cleavage, pH fusion, and viral pathogenesis (Abdelwhab et al., viruses were conducted in an animal biosecurity level 3 (ABSL-3)
2016; Seekings et al., 2020; Sun et al., 2019). Additionally, increased laboratory.
number of basic amino acids in combination with mutations adjacent to
the HACS is important for viral virulence (Lee et al., 2006). These find- 2.3. Generation of recombinant viruses by reverse genetics
ings suggested that the insertions of polybasic amino acids and other
amino acid substitutions of HA protein affected the virulence of H7 The viral RNA of the HA gene of HZ-3 was amplified, and the muta-
subtype viruses. tions were introduced into the cleavage sites by inducing site-directed
The HACS of emerging HP-H7N9 viruses in China were more poly- mutagenesis using PCR. The six H7N9 recombinants were rescued by
morphic than their predecessors. Earlier reports confirmed four motifs at reverse genetics after cloning the different HA genes and the seven
the HACS of HP-H7N9 viruses, including the motifs PEVPKRKRTAR/G, remaining genes of the amplified G1 into plasmid pHW2000 (Hoffmann
PEVPKGKRTAR/G, PEVPKGKRIAR/G, and PEVPKRKRAAR/G. The et al., 2000). Human embryonic kidney cell monolayers in 6-well plates
emphasis sign represents the mutation of amino acid, and the underline were transfected at 80%–90% confluency with 4 μg of the eight plasmids
represents the insertion of amino acid. The PEVPKRKRTAR/G and (i.e., 500 ng of each plasmid) by using Lipofectamine 3000 (Invitrogen,
PEVPKGKRTAR/G motifs were observed to circulate widely in poultry, L3000015, Carlsbad, CA, USA) according to the instructions of manu-
whereas the PEVPKGKRIAR/G and PEVPKRKRAAR/G motifs were facturer. DNA and transfection reagent were mixed, incubated at room
observed only in humans (Quan et al., 2018). Later, the PEVPKRKR- temperature for 5 min, and added to the cells. After 4 h, the mixture was
TAR/G motif of the H7N9 virus became the dominant cleavage site motif replaced with Opti-MEM (Gibco, 11058021) containing 0.2% bovine
in China (Fig. 1A). In our previous research, we observed that the four serum albumin (BSA) and 1 μg/mL tolylsulfonyl-phenylalanyl chlor-
novel HPAI H7N9 variants carrying unique HACS motifs showed differ- omethyl ketone (TPCK)-treated trypsin (Sigma-Aldrich, T4049, Saint
ences in virulence in chickens, mice, and ferrets (Bao et al., 2019; Qi Louis, USA). The supernatant collected after transfection was inoculated
et al., 2018), and we also found that isoleucine at site 335 of the most into SPF embryonated chicken eggs, and the full-length genomes of all of
HP-H7N9 strains was mutated to valine. However, the contribution of the H7N9 viruses were confirmed by sequencing.
basic and non-basic amino acids to the motifs remains unclear. In order to
elucidate the function of those combined insertions and the substitution 2.4. Replication kinetics
of adjacent residues, reverse genetics were conducted to generate six
H7N9 recombinants containing different HACS motifs (i.e., PEIPKGR/G, Replication kinetics was determined by inoculating CEF or A549 cells
PEVPKGR/G, PEVPKRKRTAR/G, PEVPKGKRTAR/G, PEVPKGKRIAR/G, in the presence or absence of 1 μg/mL TPCK-treated trypsin. Next, 90%-
and PEVPKRKRR/G) with the backbone of the A/chick- confluent CEF and A549 cells were infected at a multiplicity of infection
en/Guangdong/G1/2013 [H7N9] (G1) virus and the HA gene of the (MOI) of 0.01 and 0.1, respectively, washed twice with phosphate-
A/chicken/Guangdong/HZ-3/2016 [H7N9] (HZ-3) virus (PEIPKGR/G buffered saline (PBS) after 1 h of incubation, and incubated with 1 mL
cleavage site motif). We characterized the reconstituted viruses in terms of DMEM containing 0.2% BSA at 37 C with 5% CO2. Supernatants were
of viral replication in avian and mammalian cells, thermostability and sampled at 12, 24, 48, and 72 h after inoculation, and viral titers were
acid stability, cleavage efficiency, the virulence in mice, and pathoge- determined by 50% tissue culture infective dose (TCID50) assay in MDCK
nicity and transmissibility in chickens. cells.
2.1. Data collection CEF and A549 cells were inoculated with the wild-type and HA
mutant constructs at an MOI of 1 in the presence of either 1 μg/mL TPCK-
All data regarding cases of influenza A (H7N9) were collected from treated trypsin or no protease for 6 h in minimal essential medium
the World Health Organization (http://www.who.int/en/), the EMPRES- containing 0.2% BSA. Proteins from the cell cultures were separated on
I datasets of the Food and Agriculture Organization of the United Nations 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels and
(http://empres-i.fao.org/eipws3g/#h¼1), and the National Health were electro-transferred onto nitrocellulose (NC) membranes. Polyclonal
Commission of the People's Republic of China (http://en.nhfpc.gov.cn/) rabbit anti-HA antibody from A/Shanghai/1/2013H7N9 (Sino
39
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
Fig. 1. The prevalence and evolution of HPAIV H7N9 in China. A Distribution of HPAIV H7N9 in China, 2016–2020. The bar chart represents HPAIV H7N9 in China,
and the line graph represents human isolates. B Evolutionary history of HPAIV H7N9 in China. C The schematic diagram of cleavage site motifs of six recombinant
viruses. HPAIV, highly pathogenic avian influenza virus; LP, Low pathogenic; HP, Highly pathogenic; HA, Hemagglutinin.
Biological, 40104-RP02, Beijing, China) (1:1000 dilution for 1 h at room (SuperSignal West Pico Chemiluminescent Substrate Kit; Pierce, Rock-
temperature) was used to detect HA protein, whereas goat anti-rabbit IgG ford, IL, USA). Last, the NC film was scanned with an Odyssey bicolor
conjugated with horseradish peroxidase (LICOR, 926-32211, Lincoln, infrared laser imaging system (LI-COR Odyssey) to obtain a sample strip
USA) (1:10,000 dilution for 1 h at room temperature) was used as a image. Grayscale analysis of individual bands was determined by ImageJ
secondary antibody, followed by chemiluminescence detection v1.8.0 (National Institution of Health, USA), and relative cleavage
40
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
efficiency of H7N9 viruses was calculated according to the equation (Supplementary Table S1). Molecular characterization showed that the
HA1/(HA0 þ HA1). Three independent replicates of HA cleavability low pathogenic H7N9 viruses bearing the PEIPKGR/G cleavage site motif
were conducted in the experiment. firstly emerged in China in February 2013. After mid-2016, H7N9 viruses
containing PEVPKGKRTAR/G and PEVPKRKRTAR/G motifs, first iso-
2.6. Viral thermostability and pH stability lated in Guangdong Province of China, had spread to more than 10
provinces (Supplementary Table S2). However, a steep increase in the
Viral thermostability was measured by determining the loss of the PEVPKRKRTAR/G motif of HACS arose after January 2017 and had been
virus titer after incubating the viruses at 50 C for 0–5 h. Afterward, 1 mL stably maintained ever since (Fig. 1A and B). Additionally, approximately
(106 TCID50/mL) virus-containing allantoic fluids were incubated at 50 95% of HACS in human-origin HP-H7N9 variants exhibited the
PEVPKRKRTAR/G motif, while the HACS of two HP-H7N9 strains
C for 0, 1, 2, 3, 4, and 5 h, respectively, and the infectivity of H7N9
viruses was assessed by TCID50 assay in MDCK cells. Acid stability was showed IA and AA alleles of non-basic amino acids (Fig. 1B, Supple-
determined by treating each virus at low pH. Different pH levels (4.0, 4.5, mentary Table S1). The amino acid residue at 335 site in the HA protein
5.0, 5.5, 6.0, 7.0, and 7.4) of PBS were mixed with 105 TCID50/mL of of the LP-H7N9 virus was more polymorphic, including changes with
H7N9 viruses, after which H7N9 viruses were incubated at 37 C for 1 h. isoleucine (I) (99.6%), leucine (L) (0.16%), threonine (T) (0.12%), and
The viral titers of acid-treated H7N9 viruses were determined by TCID50 valine (V) (0.16%). Interestingly, V residue at 335 site in HA protein
assay in MDCK cells. appeared in 99.2% HP-H7N9 viruses (Supplementary Table S3). Taken
together, our finding indicated that PEVPKRKRTAR/G-motif of bird- and
2.7. Animal experiments human-origin H7N9 viruses had become the dominant motif in China.
In the first chicken experiment, the intravenous pathogenicity index 3.2. Generation of H7N9 viruses
(IVPI) test was performed according to the standard procedure of the
World Organisation for Animal Health (https://www.oie.int/en/home/). Eight-plasmid reverse genetic systems with the HA plasmid of HZ-3
Ten 6-week-old SPF chickens (Guangdong Dahuanong Animal Health and the seven plasmids of G1 were established. Six mutants differing in
Products Co., Ltd., Guangdong Province, China) were injected intrave- the HACS were generated by site-directed mutagenesis. The HACS of HZ-
nously in the ulnar vein with 0.1 mL of a 1/10 dilution of H7N9 viruses. 3 was mutated to generate the PEVPKGKRTAR/G and PEVPKGKRIAR/G
All chickens were observed for clinical symptoms for 10 days; at each motifs present in human cases. In order to assess the effects of non-basic
observation, each chicken was classified as healthy (score of 0), sick amino acids TA of the HACS, the HACS of the PEVPKRKRR/G motif was
(score of 1), severely sick (score of 2), or dead (score of 3). The IVPI index also generated. In order to determine the function of I335V substitution,
is the average score per chicken per day over 10 days. we replaced V with I in the 335 site of HA protein. The six recombinant
In the second chicken experiment, nine chickens from each group H7N9 viruses were designated PEIPKGR/G, PEVPKGR/G, PEVPKRKR-
were intranasally inoculated with 104 TCID50/200 μL of six recombinant TAR/G, PEVPKGKRTAR/G, PEVPKGKRIAR/G, and PEVPKRKRR/G,
viruses, whereas nine chickens inoculated with 200 μL of PBS served as respectively (Fig. 1C).
negative controls. After 24 h, nine naive chickens were introduced into a
clean sealed containers with 60 26 27 cm3 with infected chickens to 3.3. Cleavage efficiency of H7N9 viruses in CEF and A549 cells
assess transmission. Cloacal and throat swabs were collected at 3, 5, and
7 days post-infection (dpi) from the six infected and naive chickens. The cleavage efficiency of the HA protein is a major determinant of
Three infected chickens in each group were euthanized at 3 dpi, and the the virulence of the AIV. In order to determine how the H7N9 viruses
lung and brain samples were collected for virus titration. In the same containing different cleavage sites influenced the activation of the
way, the contacted groups were euthanized at 4 dpi to test for the virus in cleavage of HA0, a Western blot analysis of CEF and A549 cells inocu-
the lungs and brains. Tissue and swab samples were collected for virus lated with the wild-type and HA mutant constructs at an MOI of 1 in the
titration by TCID50 assay. After 14 days, antibodies were detected in the presence of either 1 μg/mL TPCK-treated trypsin or no protease for 6 h
serum samples collected from all surviving chickens. was performed. As expected, in the absence of TPCK trypsin, the HA
In a separate experiment, nine 5-week-old SPF female BALB/c mice protein of all mutants with monobasic HACS was not cleaved into HA1
(Vital River Laboratory, Beijing, China) in six groups were anesthetized and HA2; however, the HA0 carrying polybasic HACS was cleaved to
with CO2 and inoculated intranasally with 106 TCID50/50 μL H7N9 vi- HA1 and HA2 to varying degrees, whether in CEF or A549 cells (Fig. 2A
ruses. Nine mice were inoculated PBS as mock group. For 14 dpi, the mice and B). The HA protein of monobasic H7N9 viruses (PEVPKGR/G and
in each group were monitored daily for clinical symptoms, weight loss, PEIPKGR/G) can be cleaved into HA1 and HA2 in the presence of TPCK
and mortality, and the mice were euthanized if they lost more than 25% trypsin. Of the four H7N9 viruses with polybasic cleavage sites in the
of the initial body weight. Three mice in each group were euthanized at 4 absence of TPCK-trypsin, PEVPKRKRTAR/G showed the most efficient
dpi, after which the lung and brain samples were also collected. Viral cleavage efficiency and PEVPKGKRTAR/G the least, whereas the cleav-
titers were determined by TCID50 assay on MDCK cells. age efficiency of PEVPKRKRR/G and PEVPKGKRIAR/G was moderate in
CEF and A549 cells. Importantly, the cleavage efficiency of PEVPKRKRR/
2.8. Statistical analyses G (P ¼ 0.0085) and PEVPKRKRTAR/G (P ¼ 0.0065) were both higher
than that of the PEVPKGKRTAR/G in CEF cells, whereas PEVPKGKR-
Data were presented as mean standard deviation and were analyzed TAR/G showed more limited cleavage efficiency in A549 cells (Fig. 2B).
with GraphPad Prism 5.0. The independent samples t-test was used for
analysis, and P values less than 0.05 were considered to indicate statis- 3.4. Impact of H7N9 viruses on viral thermostability and acid stability
tical significance.
The pH stability of HA protein has repeatedly been shown to be an
3. Results important factor in the adaption of influenza viruses to efficiently
transmit among chickens and ferrets (Russell et al., 2018). In order to
3.1. Bioinformatic analysis of the H7N9 cleavage site investigate the effects of the acidic conditions of the viruses, six H7N9
recombinant viruses were incubated at pH levels ranging from 5.0 to 7.2
As of August 2021, 242 HA sequences of HP-H7N9 viruses were for 30 min. Viral titer assay showed that all H7N9 viruses displayed a
collected by the Global Initiative on Sharing All Influenza Data (htt decreased viral activity at pH 5.5, and PEIPKGR/G exhibited significantly
ps://www.gisaid.org), including 62 human-infecting H7N9 strains reduced infectivity (P < 0.001) compared to the PEVPKGR/G (Fig. 2C).
41
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
Fig. 2. HA cleavage efficiency and stability of H7N9 viruses. A Western blots of cell cultures from CEFs and A549 cell cultures inoculated with H7N9 viruses at a
multiplicity of infection (MOI) of 1 for 6 h, treated with (þ) or without () trypsin. B Grayscale analysis of H7N9 recombinant viruses. Cleavage efficiencies were
calculated according to the equation HA1/(HA0 þ HA1). Each point on the curve is the mean standard deviation from three independent experiments. The in-
dependent samples t-test was used for analysis. C Stability after incubation of viruses at pH 5, 5.5, 6, 6.5, or 7.2. D Heat stability at 50 C for 1, 2, 3, 4, and 5 h.
Statistical significance was analyzed using an unpaired t-test. *P < 0.05, **P < 0.01, ***P < 0.001.
These results suggested that the I335V substitution increased the resis- mutants exhibited differences in viral titer at different time points,
tance against the acid of the H7N9 viruses. In addition, we investigated although the viral titers of PEVPKGR/G were higher than those of
the thermostability of the H7N9 viruses containing different HACS. Our PEIPKGR/G, especially in the prophase of replication (Fig. 3A).
result showed that the monobasic PEIPKGR/G and PEVPKGR/G viruses In A549 cells, the growth rate of PEVPKRKRR/G was significant
lost all infectivity after incubation for 3 h. Conversely, PEVPKRKRR/G higher (P < 0.05) than that of PEVPKGR/G at 24 h in the presence of
and PEVPKGKRIAR/G were inactivated after 4 h of incubation, sugges- TPCK-treated trypsin (Fig. 3D). In the absence of TPCK-treated trypsin,
tive of considerable tolerance to high temperatures. PEVPKRKRTAR/G however, the replicative ability of H7N9 viruses containing different
and PEVPKGKRTAR/G, inactivated after 5 h of incubation, were the most MBCSs was not significantly different, nor was the monobasic H7N9 vi-
resistant (Fig. 2D). Overall, these results indicated that the insertion of ruses in A549 cells (Fig. 3C). These findings indicated that the insertion
basic and non-basic amino acids enhanced the thermostability of the of amino acids KRTA promoted the replicative ability in CEFs and A549
H7N9 viruses. cells, and the I335V substitution promoted the early replication of the
H7N9 virus in CEF cells.
42
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
Fig. 3. The growth kinetics of H7N9 viruses. A CEFs were inoculated with viruses at a multiplicity of infection (MOI) of 0.01 in the absence of trypsin. B CEFs were
inoculated with viruses at an MOI of 0.01 in the presence of trypsin. C A549 cells were infected with the indicated viruses at an MOI of 0.1 in the absence of trypsin. D
A549 cells were infected with the indicated viruses at an MOI of 0.1 in the presence of trypsin. Samples were collected at 12, 24, 48, and 72 h post-inoculation (hpi).
Viral titers were determined via 50% tissue culture infectious dose (TCID50) assay. Each point on the curve is the mean standard deviation from three independent
experiments. Statistical analysis of PEVPKGR/G was conducted compared to the PEIPKGR/G. Statistical analysis of PEVPKGKRTAR/G, PEVPKRKRTAR/G, and
PEVPKRKRR/G were conducted compared to the PEVPKGKRIAR/G, respectively. Statistical significance was analyzed using an unpaired t-test. *P < 0.05, **P < 0.01,
***P < 0.001.
In order to determine which mutations at cleavage sites abolished PEVPKGKRIAR/G contact, showed virus shedding at 5 dpi and serocon-
virulence and transmissibility in chickens, six-week-old SPF chickens version on 14 dpi. Hemagglutinin inhibition assay showed that four
were intranasally inoculated with 104 TCID50/200 μL of H7N9 viruses. contacted chickens in the PEVPKGR/G group were infected, whereas
After 24 h, nine uninfected chickens were placed in a clean sealed only one was infected in the PEIPKGR/G group (Table 1), which indi-
container and housed with the inoculated chickens. Four groups of cated that I335V substitution enhanced the susceptibility, virulence, and
chickens with recombinant viruses containing MBCSs showed clinical transmissibility of H7N9 virus in chickens.
symptoms and tissue lesions and died within 5–7 dpi (Fig. 4B). However,
chickens infected with PEVPKGR/G and PEIPKGR/G did not exhibit any 3.7. Pathogenicity of H7N9 viruses in mice
clinical symptoms. Three inoculated and exposed chickens in each group
were dissected in order to evaluate viral titers in lung and brain at 3 dpi. In order to investigate the effects of H7N9 viruses with different
PEVPKGKRTAR/G, PEVPKRKRR/G, PEVPKGKRIAR/G, and PEVPKRKR- HACS in mammals, nine five-week-old female BALB/c mice weighing
TAR/G achieved relatively high but not significantly different viral titers 14g–16 g were intranasally inoculated with 106 TCID50/50 μL of each
in lungs compared to the monobasic mutants. By contrast, in brain, H7N9 viruses. The body weight of the mice infected with PEVPKGR/G
PEVPKRKRTAR/G was detected only in one of the three chickens. virus increased during the observation period. Mice infected with
Importantly, PEVPKGR/G replicated more efficiently than PEIPKGR/G in PEVPKRKRR/G and PEVPKGKRIAR/G showed slight weight loss (< 5%)
lungs (Fig. 4D). at about 6 dpi, and one mouse infected with PEVPKGKRIAR/G died at 8
Contacted chickens of the PEVPKGKRTAR/G, PEVPKRKRR/G, and dpi (Fig. 5B). However, mice infected with PEVPKRKRTAR/G showed
PEVPKRKRTAR/G viruses died within 7 days; however, 50% of the significant weight loss at 6 dpi compared with the PEVPKGR/G group (P
chickens directly exposed to PEVPKGKRIAR/G survived, which under- ¼ 0.0003), and mice infected with PEVPKGKRTAR/G showed significant
scored the difference of transmission in chickens (Fig. 4C). Beyond the weight loss at 7 dpi compared with PEVPKGR/G (P ¼ 0.0001) (Fig. 5A).
contacted group, the viral titer of lung samples in the PEVPKGKRTAR/G Surprisingly, the monobasic mutants (i.e., PEIPKGR/G) showed clear
group was significantly higher than that of the PEVPKRKRTAR/G group symptoms of infection, including shaggy hair and abdominal breathing,
(P ¼ 0.015). Only PEVPKRKRR/G, with an average titer of 2.73 after the second day of infection, followed by 15% of weight loss
lgTCID50/100 μL (Fig. 4E), was detected in the brains of the contacted compared to the initial weight and death, with a mortality rate of 35%. In
chickens. Furthermore, virus shedding was detected at 3 and 5 dpi both addition, each H7N9 viruses can replicate efficiency in lungs of the
in throat and cloacal swabs. The chickens that survived from infected mice (Fig. 5C). The PEVPKGKRTAR/G (P ¼ 0.0081) and
43
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
Fig. 4. Pathogenicity of the H7N9 viruses in chickens. A Survival rate of IVPI chickens. B Survival rate of inoculated chickens. Six groups of nine 6-week-old SPF
chickens were infected intranasally with 104 TCID50/200 μL of each virus. C Survival rate of SPF chickens in direct contact with infected chickens. D Viral titers in
lungs and brains of three infected chickens. E Viral titers in lungs and brains of three contacted chickens. Each point on the curve is the mean standard deviation
from three independent experiments. Statistical significance was analyzed using an unpaired t-test: *P < 0.05. Viral titers were determined via TCID50 assay. IVPI,
intravenous pathogenicity index; TCID50, 50% tissue culture infective dose.
Table 1
Virus titers in chicken cloacal and oropharyngeal swabsa.
Virus Oropharyngeal swabs, n Cloacal swabs, n Seroconversion HI antibody titers (14 dpi), n (range) IVPI Cleavage site sequence
PEVPKRKRTAR/G (P ¼ 0.0109) significantly increased viral titers in the investigated the biological function of the unique HACS motifs of H7N9
lungs of mice compared to that of the PEVPKGR/G group, indicating that viruses by focusing on the effects of the combination of basic and non-
the insertion of KRTA increased the pathogenicity of mice. Moreover, basic amino acids in H7N9 viruses. Related to the findings proposing
none of each H7N9 viruses were detected in brain of mice (Fig. 5C). All that the cleavability closely related to virulence (Horimoto and Kawaoka,
surviving mice recovered at 14 dpi. 1997), our results revealed that the insertion of MBCSs significantly
increased the cleavage efficiency of HA protein when compared to that of
4. Discussion the monobasic H7N9 viruses. PEVPKRKRTAR/G showed significantly
higher cleavage efficiency than that of PEVPKGKRTAR/G, likely because
The HP-H7N9 viruses emerged in Guangdong Province since mid- the presence of arginine at the 338 site increased the cleavage efficiency
2016 and widely spread in poultry populations. In our study, we of HA protein. At the same time, PEVPKRKRR/G manifested high
44
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
Fig. 5. Pathogenicity of the H7N9 viruses in mice. Six groups of nine female BALB/c mice were intranasally inoculated with 106 TCID50/50 μL of the viruses or with
PBS as negative controls. Mice that lost more than 25% of their initial weight were found dead or euthanized. A Relative weight changes in the mice compared to the
initial body weight. B Survival rate of mouse. C Viral titers in lungs and brains of mice at necropsy 4 days post-infection. Each point on the curve is the mean
standard deviation from three independent experiments. Statistical significance was analyzed using an unpaired t-test: *P < 0.05, **P < 0.01, ***P < 0.001.
cleavage efficiency in avian cells, which suggested the broader tissue exhibited significant differences in contacted groups. PEVPKGKRIAR/G
tropism of different MBCSs in H7N9 viruses. showed limited transmission in chickens via direct contact. In addition,
HA stability was regarded as an important contributor to the infec- we found that the replicative ability of PEVPKGKRIAR/G virus was lower
tivity, transmissibility, and pathogenicity of influenza viruses in different than that of the other H7N9 viruses in CEF cell and SPF chickens, which
hosts. After the virus internalized by endocytosis, the acidic pH in the indicated that it had not adapted in poultry. Interestingly, the viral titers
endosomes triggers an irreversible conformational change in HA protein, in lung samples of infected and contacted chickens infected with
that creates a viral envelope with an endosomal membrane (Bullough PEVPKGR/G were higher than that of PEIPKGR/G virus, indicating that
et al., 1994). High temperatures at a neutral pH can also induce I335V substitution increased the transmission in chickens. The viral titers
conformational changes in HA that prompt protein inactivation (Carr in lung samples of chickens infected with PEVPKGKRTAR/G virus were
et al., 1997; Scholtissek, 1985). In this study, the environmental stability higher than that of the other groups, whereas PEVPKRKRTAR/G showed
of the H7N9 viruses at different pH levels was consistent; however, the mild virulence in chickens than other H7N9 viruses containing MBCS.
PEVPKRKRTAR/G and PEVPKGKRTAR/G became inactive after 5 h of In our assessment, PEVPKRKRTAR/G, a moderately virulent strain,
incubation, which suggested that the insertion of non-basic amino acids though exhibited the high pathogenicity and strong transmission in
(TA) increased the resistance to higher temperature. As reported, influ- chickens, might leave individuals of the species with greater fitness to
enza viruses with high stability were more infectious in ferrets (Cotter circulate in poultry.
et al., 2014). Compared to TA, amino acid IA exhibited low virulence and In previous studies, HP-H7N9 was not lethal in mice (Shi et al., 2017)
transmissibility in contacted chickens, which might have been respon- or else killed certain mice (Liu et al., 2018) in absence of mutations at the
sible for the lower viral thermostability. Although the insertion of amino PB2 protein, which was corroborated in our results. The phenomenon
acid KRTA in HACS did not influence the pH stability of the H7N9 virus, that we observed, namely, increased virulence when infected mice and
the I335V substitutions in HA protein enhanced the pH stability. Addi- varying virulence between different MBCSs, suggested that the MBCSs
tionally, other studies have also demonstrated that a lysine to glutamic represented an important factor in the virulence in mice. Another study
amino acid mutations at the HA2-64 position increases the threshold pH reported that the single amino acid of the HACS affected the pathoge-
for HA activation in HP-H7N9 viruses (Sun et al., 2019), indicating that nicity and neurotoxicity of the H5N1 influenza virus in mice (Zhang
the mutations in the HA protein of H7N9 viruses affect the pH fusion. et al., 2012). Notably, we observed that the insertion of non-basic amino
In in vivo experiment, H7N9 viruses containing MBCS were high acids TA and IA in H7N9 viruses caused death in mice despite having the
pathogenicity in chickens. However, all of the HP-H7N9 viruses same viral titer, which suggested that HPAIV H7N9 inserted with a
45
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
non-basic amino acid posed a potential threat to humans. PEVPKRKRR/G conceptualization, data curation, formal analysis, funding acquisition,
does not contain non-basic amino acids, despite being highly pathogenic project administration, writing-review & editing. All the authors
in chickens. However, the virulence in mice was not more significant reviewed and revised the first and final drafts of this manuscript.
than that of PEVPKRKRTAR/G, which was consistent with the cleavage
efficiency in A549 cells. We also observed that mice infected with the
PEIPKGR/G strain showed a sharp drop in body weight, followed by Conflict of interest
death. In our previous study, we also found polymorphic mutations in the
PB2 protein of dead mice infected with HP-H7N9 viruses by The authors declare that they have no conflict of interest.
next-generation sequencing (Qi et al., 2018). Consequently, the death of
mice might be caused by dynamic mutations of other genes. Acknowledgments
There remains an important question to answer regarding the
mechanism of the insertion of amino acids at the HACS in H7N9 viruses. This work was supported by National Natural Science Foundation of
Researchers have reported that influenza viruses cause the insertion of China (31672586 and 31830097), the Key Research and Development
exogenous nucleotide sequences into the HACS by nonhomologous Program of Guangdong Province (2019B020218004), Earmarked Found
recombination (Khatchikian et al., 1989; Orlich et al., 1994; Suarez et al., for China Agriculture Research System (CARS-41-G16), Guangdong
2004), and that all intersegmental recombination involving HA described Province Universities and Colleges Pearl River Scholar Funded Scheme
thus far has only involved the H7 subtype (Khatchikian et al., 1989). (2018, Wenbao Qi), and Young Scholars of Yangtze River Scholar Pro-
Therefore, we hypothesized that the amino acid KRTA at the initial HACS fessor Program (2019, Wenbao Qi). We acknowledge the authors who
of the HP-H7N9 virus could be obtained by inserting exogenous nucle- submitted the sequences from GISAID's EpiFlu Database on which this
otides. The BLAST search for that nucleotide sequence in the influenza research is based.
database revealed that the most closely related sequence was the PB1
genome of many viruses of the H6 subtype, with 12 nucleotide identities
Appendix A. Supplementary data
at positions 2226–2251. We also found entirely identical nucleotides in
chromosome 1 of gallus, located in the range of dihydrolipoamide de-
Supplementary data to this article can be found online at https://do
hydrogenase, and found that the two ends contained two nucleotides that
i.org/10.1016/j.virs.2022.01.001.
matched the viral sequence, that was, the matching of 16 nucleotides.
The results above suggest that the occurrence of the insertion of amino
acids at the HACS is probably due to recombination events. References
Abdelwhab, E.M., Veits, J., Ulrich, R., Kasbohm, E., Teifke, J.P., Mettenleiter, T.C., 2016.
5. Conclusions Composition of the hemagglutinin polybasic proteolytic cleavage motif mediates
variable virulence of H7N7 avian influenza viruses. Sci. Rep. 6, 39505.
In conclusion, we found that the combination of basic and non-basic Bao, L., Bi, Y., Wong, G., Qi, W., Li, F., Lv, Q., Wang, L., Liu, F., Yang, Y., Zhang, C.,
Liu, W.J., Quan, C., Jia, W., Liu, Y., Liu, W., Liao, M., Gao, G.F., Qin, C., 2019. Diverse
amino acids improved the cleavage efficiency and stability of H7N9 vi- biological characteristics and varied virulence of H7N9 from Wave 5. Emerg. Microb.
ruses while reducing their virulence in chickens. Notably, H7N9 viruses Infect. 8, 94–102.
with non-basic amino acid insertion (TA) caused the death of mice. The Bullough, P.A., Hughson, F.M., Skehel, J.J., Wiley, D.C., 1994. Structure of influenza
haemagglutinin at the pH of membrane fusion. Nature 371, 37–43.
PEVPKRKRTAR/G motif has been widespread in poultry and human Carr, C.M., Chaudhry, C., Kim, P.S., 1997. Influenza hemagglutinin is spring-loaded by a
isolates. Therefore, the comprehensive surveillance and evaluation of the metastable native conformation. Proc. Natl. Acad. Sci. U. S. A. 94, 14306–14313.
mutations identified should be urgently undertaken in order to assess the Cotter, C.R., Jin, H., Chen, Z., 2014. A single amino acid in the stalk region of the
H1N1pdm influenza virus HA protein affects viral fusion, stability and infectivity.
potential of the influenza virus epidemic to morph into a pandemic. PLoS Pathog. 10, e1003831.
Hoffmann, E., Neumann, G., Kawaoka, Y., Hobom, G., Webster, R.G., 2000. A DNA
Data availability transfection system for generation of influenza A virus from eight plasmids. Proc.
Natl. Acad. Sci. U. S. A. 97, 6108–6113.
Horimoto, T., Kawaoka, Y., 1997. Biologic effects of introducing additional basic amino
All the data generated during the current study are included in the acid residues into the hemagglutinin cleavage site of a virulent avian influenza virus.
manuscript and/or the Supplementary data. Additional data related to Virus Res. 50, 35–40.
this article may be requested from the authors. Imai, M., Watanabe, T., Kiso, M., Nakajima, N., Yamayoshi, S., Iwatsuki-Horimoto, K.,
Hatta, M., Yamada, S., Ito, M., Sakai-Tagawa, Y., Shirakura, M., Takashita, E.,
Fujisaki, S., McBride, R., Thompson, A.J., Takahashi, K., Maemura, T., Mitake, H.,
Ethics statement Chiba, S., Zhong, G., Fan, S., Oishi, K., Yasuhara, A., Takada, K., Nakao, T.,
Fukuyama, S., Yamashita, M., Lopes, T.J.S., Neumann, G., Odagiri, T., Watanabe, S.,
Shu, Y., Paulson, J.C., Hasegawa, H., Kawaoka, Y., 2017. A highly pathogenic avian
All animal experiments were carried out in ABSL-3 facilities in H7N9 influenza virus isolated from A human is lethal in some ferrets infected via
compliance with the biosafety committee of ABSL-3 Laboratory of South respiratory droplets. Cell Host Microbe 22, 615–626.
China Agriculture University (CNAS BL0011) protocols. All animal ex- Ito, T., Goto, H., Yamamoto, E., Tanaka, H., Takeuchi, M., Kuwayama, M., Kawaoka, Y.,
Otsuki, K., 2001. Generation of a highly pathogenic avian influenza A virus from an
periments were reviewed and approved by the Institutional Animal Care avirulent field isolate by passaging in chickens. J. Virol. 75, 4439–4443.
and Use Committee at SCAU and were carried out in accordance with the Jiang, H., Wu, P., Uyeki, T.M., He, J., Deng, Z., Xu, W., Lv, Q., Zhang, J., Wu, Y.,
approved guidelines (2017A002). Tsang, T.K., Kang, M., Zheng, J., Wang, L., Yang, B., Qin, Y., Feng, L., Fang, V.J.,
Gao, G.F., Leung, G.M., Yu, H., Cowling, B.J., 2017. Preliminary epidemiologic
assessment of human infections with highly pathogenic avian influenza A(H5N6)
Author contributions virus, China. Clin. Infect. Dis. 65, 383–388.
Khatchikian, D., Orlich, M., Rott, R., 1989. Increased viral pathogenicity after insertion of
a 28S ribosomal RNA sequence into the haemagglutinin gene of an influenza virus.
Zhou Aobaixue: conceptualization, investigation, data curation, Nature 340, 156–157.
formal analysis, writing-original draft. Zhang Jiahao: conceptualization, Lee, C.W., Lee, Y.J., Senne, D.A., Suarez, D.L., 2006. Pathogenic potential of North
investigation, data curation, formal analysis, writing-original draft, American H7N2 avian influenza virus: a mutagenesis study using reverse genetics.
Virology 353, 388–395.
writing-review & editing. Li Huanan: investigation, formal analysis. Xu
Liu, D., Shi, W., Shi, Y., Wang, D., Xiao, H., Li, W., Bi, Y., Wu, Y., Li, X., Yan, J., Liu, W.,
Qiang: investigation. Chen Yiqun: investigation. Li Bo: investigation. Liu Zhao, G., Yang, W., Wang, Y., Ma, J., Shu, Y., Lei, F., Gao, G.F., 2013. Origin and
Wanying: investigation. Su Guanming: investigation. Ren Xingxing: diversity of novel avian influenza A H7N9 viruses causing human infection:
investigation. Lao Guangjie: investigation. Luo Baozheng: writing-review phylogenetic, structural, and coalescent analyses. Lancet 381, 1926–1932.
Liu, D., Zhang, Z., He, L., Gao, Z., Li, J., Gu, M., Hu, J., Wang, X., Liu, X., Liu, X., 2018.
& editing. Liao Ming: conceptualization, data curation, formal analysis, Characteristics of the emerging chicken-origin highly pathogenic H7N9 viruses: a
funding acquisition, project administration. Qi Wenbao: new threat to public health and poultry industry. J. Infect. 76, 217–220.
46
A. Zhou et al. Virologica Sinica 37 (2022) 38–47
Orlich, M., Gottwald, H., Rott, R., 1994. Nonhomologous recombination between the Suarez, D.L., Senne, D.A., Banks, J., Brown, I.H., Essen, S.C., Lee, C.W., Manvell, R.J.,
hemagglutinin gene and the nucleoprotein gene of an influenza virus. Virology 204, Mathieu-Benson, C., Moreno, V., Pedersen, J.C., Panigrahy, B., Rojas, H.,
462–465. Spackman, E., Alexander, D.J., 2004. Recombination resulting in virulence shift in
Qi, W., Jia, W., Liu, D., Li, J., Bi, Y., Xie, S., Li, B., Hu, T., Du, Y., Xing, L., Zhang, J., avian influenza outbreak, Chile. Emerg. Infect. Dis. 10, 693–699.
Zhang, F., Wei, X., Eden, J.S., Li, H., Tian, H., Li, W., Su, G., Lao, G., Xu, C., Xu, B., Sun, X., Belser, J.A., Yang, H., Pulit-Penaloza, J.A., Pappas, C., Brock, N., Zeng, H.,
Liu, W., Zhang, G., Ren, T., Holmes, E.C., Cui, J., Shi, W., Gao, G.F., Liao, M., 2018. Creager, H.M., Stevens, J., Maines, T.R., 2019. Identification of key hemagglutinin
Emergence and adaptation of a novel highly pathogenic H7N9 influenza virus in residues responsible for cleavage, acid stability, and virulence of fifth-wave highly
birds and humans from a 2013 human-infecting low-pathogenic ancestor. J. Virol. 92 pathogenic avian influenza A(H7N9) viruses. Virology 535, 232–240.
e00921–17. Wang, X., Jiang, H., Wu, P., Uyeki, T.M., Feng, L., Lai, S., Wang, L., Huo, X., Xu, K.,
Quan, C., Shi, W., Yang, Y., Yang, Y., Liu, X., Xu, W., Li, H., Li, J., Wang, Q., Tong, Z., Chen, E., Wang, X., He, J., Kang, M., Zhang, R., Zhang, J., Wu, J., Hu, S., Zhang, H.,
Wong, G., Zhang, C., Ma, S., Ma, Z., Fu, G., Zhang, Z., Huang, Y., Song, H., Yang, L., Liu, X., Fu, W., Ou, J., Wu, S., Qin, Y., Zhang, Z., Shi, Y., Zhang, J., Artois, J.,
Liu, W.J., Liu, Y., Liu, W., Gao, G.F., Bi, Y., 2018. New threats from H7N9 influenza Fang, V.J., Zhu, H., Guan, Y., Gilbert, M., Horby, P.W., Leung, G.M., Gao, G.F.,
virus: spread and evolution of high- and low-pathogenicity variants with high Cowling, B.J., Yu, H., 2017. Epidemiology of avian influenza A H7N9 virus in human
genomic diversity in wave five. J. Virol. 92, e00301–e00318. beings across five epidemics in mainland China, 2013-17: an epidemiological study
Rott, R., 1992. The pathogenic determinant of influenza virus. Vet. Microbiol. 33, of laboratory-confirmed case series. Lancet Infect. Dis. 17, 822–832.
303–310. Webster, R.G., Rott, R., 1987. Influenza virus A pathogenicity: the pivotal role of
Russell, C.J., Hu, M., Okda, F.A., 2018. Influenza hemagglutinin protein stability, hemagglutinin. Cell 50, 665–666.
activation, and pandemic risk. Trends Microbiol. 26, 841–853. Yang, W., Yin, X., Guan, L., Li, M., Ma, S., Shi, J., Deng, G., Suzuki, Y., Chen, H., 2018.
Scholtissek, C., 1985. Stability of infectious influenza A viruses at low pH and at elevated A live attenuated vaccine prevents replication and transmission of H7N9 highly
temperature. Vaccine 3, 215–218. pathogenic influenza viruses in mammals. Emerg. Microb. Infect. 7, 153.
Seekings, A.H., Howard, W.A., Nu~ nez, A., Slomka, M.J., Banyard, A.C., Hicks, D., Zhang, F., Bi, Y., Wang, J., Wong, G., Shi, W., Hu, F., Yang, Y., Yang, L., Deng, X.,
nez-García, J., Hartgroves, L.C., Barclay, W.S., Banks, J., Brown, I.H.,
Ellis, R.J., Nu~ Jiang, S., He, X., Liu, Y., Yin, C., Zhong, N., Gao, G.F., 2017. Human infections with
2020. The emergence of H7N7 highly pathogenic avian influenza virus from low recently-emerging highly pathogenic H7N9 avian influenza virus in China. J. Infect.
pathogenicity avian influenza virus using an in ovo embryo culture model. Viruses 75, 71–75.
12, 920. Zhang, J., Ye, H., Li, H., Ma, K., Qiu, W., Chen, Y., Qiu, Z., Li, B., Jia, W., Liang, Z.,
Shi, J., Deng, G., Kong, H., Gu, C., Ma, S., Yin, X., Zeng, X., Cui, P., Chen, Y., Yang, H., Liao, M., Qi, W., 2020. Evolution and antigenic drift of influenza A (H7N9) viruses,
Wan, X., Wang, X., Liu, L., Chen, P., Jiang, Y., Liu, J., Guan, Y., Suzuki, Y., Li, M., China, 2017-2019. Emerg. Infect. Dis. 26, 1906–1911.
Qu, Z., Guan, L., Zang, J., Gu, W., Han, S., Song, Y., Hu, Y., Wang, Z., Gu, L., Yang, W., Zhang, Y., Sun, Y., Sun, H., Pu, J., Bi, Y., Shi, Y., Lu, X., Li, J., Zhu, Q., Gao, G.F., Yang, H.,
Liang, L., Bao, H., Tian, G., Li, Y., Qiao, C., Jiang, L., Li, C., Bu, Z., Chen, H., 2017. Liu, J., 2012. A single amino acid at the hemagglutinin cleavage site contributes to
H7N9 virulent mutants detected in chickens in China pose an increased threat to the pathogenicity and neurovirulence of H5N1 influenza virus in mice. J. Virol. 86,
humans. Cell Res. 27, 1409–1421. 6924–6931.
Shi, J., Deng, G., Ma, S., Zeng, X., Yin, X., Li, M., Zhang, B., Cui, P., Chen, Y., Yang, H., Zhu, W., Dong, J., Zhang, Y., Yang, L., Li, X., Chen, T., Zhao, X., Wei, H., Bo, H., Zeng, X.,
Wan, X., Liu, L., Chen, P., Jiang, Y., Guan, Y., Liu, J., Gu, W., Han, S., Song, Y., Huang, W., Li, Z., Tang, J., Zhou, J., Gao, R., Xin, L., Yang, J., Zou, S., Chen, W.,
Liang, L., Qu, Z., Hou, Y., Wang, X., Bao, H., Tian, G., Li, Y., Jiang, L., Li, C., Chen, H., Liu, J., Shu, Y., Wang, D., 2018. A gene constellation in avian influenza A (H7N9)
2018. Rapid evolution of H7N9 highly pathogenic viruses that emerged in China in viruses may have facilitated the fifth wave outbreak in China. Cell Rep. 23, 909–917.
2017. Cell Host Microbe 24, 558–568.
47