JVI Accepts, published online ahead of print on 1 May 2013 J. Virol. doi:10.1128/JVI.00804-13 Copyright © 2013, American Society for Microbiology. All Rights Reserved.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Antigenic characterization of H3N2 influenza A viruses from Ohio agricultural fairs Running title: Antigenic profile of swine-origin H3N2 IAVs Zhixin Feng1,2$, Janet Gomez1$, Andrew S. Bowman3, Jianqiang Ye1, Li-Ping Long1, Sarah W. Nelson3, Jialiang Yang1, Brigitte Martin1, Kun Jia1, Jacqueline M. Nolting3, Fred Cunningham4, Carol Cardona5, Jianqiang Zhang6, Kyoung-Jin Yoon6, Richard D. Slemons3, and Xiu-Feng Wan1* 1
Department of Basic Sciences, College of Veterinary Medicine, Mississippi State University, Mississippi State, MS 39762, the United States; 2Institute of Veterinary Medicine, Jiangsu Academy of Agricultural Sciences, Nanjing 210014, People's Republic of China; 3Department of Veterinary Preventive Medicine, College of Veterinary Medicine, The Ohio State University, Columbus, OH 43210, the United States; 4USDA/APHIS/WS, National Wildlife Research Center, 5 Mississippi Field Station, Mississippi State, MS 39762, the United States; College of Veterinary Medicine, University of Minnesota, St. Paul, MN 55108, the United States; 6Department of Veterinary Diagnostic and Production Animal Medicine, College of Veterinary Medicine, Iowa State University, Ames, IA 50011, the United States $
Equal contribution to this work.
*Correspondence:
[email protected] or
[email protected].
27 28 29
Abstract
30
viruses (IAVs) and swine exposure at agricultural fairs has raised concerns about the human
31
health risk posed by IAV-infected swine. Understanding the antigenic profiles of IAVs
32
circulating in pigs at agricultural fairs is critical to developing effective prevention and control
33
strategies. Here 68 H3N2 IAV isolates recovered from pigs at Ohio fairs (2009 to 2011) were
34
antigenically characterized. These isolates were compared with other H3 IAVs recovered from
35
commercial swine, wild birds, and canines, along with human seasonal and variant H3N2 IAVs.
36
Antigenic cartography demonstrated that H3N2 IAV isolates from Ohio fairs could be divided
37
into two antigenic groups: 1) the 2009 fair isolates and 2) the 2010 and 2011 fair isolates. These
38
same two antigenic clusters have also been observed in commercial swine populations in recent
39
years. Human H3N2v isolates from 2010 and 2011 are antigenically clustered with swine-origin
40
IAVs from the same time period. The isolates recovered from pigs at fairs did not cross react
41
with ferret antisera produced against the human seasonal H3N2 IAVs circulating during the past
42
decade, raising the question of the degree of immunity the human population has to swine-origin
43
H3N2 IAVs. Our results demonstrate that H3N2 IAVs infecting pigs at fairs and H3N2v isolates
44
were antigenically similar to the IAVs circulating in commercial swine, demonstrating that
45
exhibition swine can function as a bridge between commercial swine and the human population.
46 47
The demonstrated link between the emergence of H3N2 variant (H3N2v) influenza A
48 49 50
Introduction
51
bi-directional interspecies transmission of influenza A viruses (IAVs). Numerous cases of variant
52
influenza A have been linked to swine exposure at agricultural fairs across the United States and
53
the first documented transmissions of the 2009 pandemic (pdm09) H1N1 IAV from humans to
54
pigs also occurred at agricultural fairs (1-5). Since July 2011, more than 325 confirmed human
55
cases of H3N2 variant (H3N2v) influenza A have occurred in 14 states (6). Many of these cases
56
reported direct or indirect contact to exhibited swine with limited subsequent human-to-human
57
transmission of H3N2v IAVs (7, 8).
The swine-human interface at agricultural fairs has proven to be an important site for the
58
Influenza A viruses can cause sporadic, epidemic, and pandemic respiratory disease in
59
humans and from mild to severe respiratory diseases in horses, pigs, domestic poultry, and sea
60
mammals (9-12). Because the tracheal epithelial cells of swine contain both avian-like alpha 2,3-
61
linked sialic acid receptors and human-like alpha 2,6-linked sialic acid receptors (13, 14), pigs
62
can be infected by avian- and mammalian-origin IAVs which may facilitate reassortment events
63
and result in the subsequent generation of novel IAV capable of infecting humans (13, 14).
64
Currently, multiple antigenically and genetically divergent IAVs co-circulate in swine
65
populations and antigenic recycling of IAVs can occur in modern swine management systems
66
where there is a continuous supply of naive pigs (15). Antigenic characterization of the IAV
67
(H1N1)pdm09 demonstrated that the H1 had been circulating among swine for decades prior to
68
the onset of the 2009 pandemic (16) confirming the potential for the emergence of pandemic
69
IAV from pigs.
70 71
Influenza A virus was first isolated from swine in 1930 (17). This H1N1 strain, likely derived from the 1918 H1N1 pandemic IAV (18), was the only major IAV circulating among US
72
swine for more than 60 years and is now known as “classical” swine influenza virus. A triple
73
reassortant H3N2 (trH3N2) IAV composed of three genes (HA, NA, and PB1) derived from
74
human H3N2 seasonal influenza viruses (huH3N2), two genes (PB2 and PA) from avian-origin
75
IAVs, and three genes (NP, MP, and NS) from “classical” swine influenza virus (19, 20)
76
emerged in swine populations in the mid-1990s and has since become well-established in North
77
America. Phylogenetic analyses of HA genes from H3N2 IAVs isolated from North American
78
pigs after 1998 demonstrates that these trH3N2 IAV can be classified into at least four genetic
79
groups (clusters I, II, III, and IV) (21) with one lineage (cluster IV) predominating since 2005.
80
Neutralization assays with swine antisera demonstrated that these four genetic clusters are also
81
antigenically distinct (22). Several reassortment events between trH3N2 IAVs, human-origin
82
IAVs, and classical swine influenza virus have occurred since 1998 resulting in multiple gene
83
constellations with various HA and NA genes in combination with the Triple Reassortant
84
Internal Gene (TRIG) cassette (which includes PB2, PB1, PA, NP, NS, and M segments) (23,
85
24). Human cases of infection with swine-origin IAVs reported by the Centers for Disease
86
Control and Prevention (CDC) after 1998 were all characterized as containing the TRIG cassette
87
and were linked antigenically or genetically, to IAVs concurrently circulating in pigs. Most of
88
these cases were isolated infections in which patients reported indirect or direct exposure to pigs
89
(25, 26).
90
The degree of immunity the human population has to swine-origin H3N2 IAVs is open to
91
question (6). Because agricultural fairs bring people into close contact with swine, defining the
92
antigenic profiles of the IAVs circulating in exhibited pigs at fairs is critical to developing
93
effective strategies to prevent and/or control human infections of swine-origin IAVs.
94
Hemagglutination inhibition (HI) tests were used to compare the antigenic profiles of IAV
95
isolates recovered from pigs at Ohio agricultural fairs during 2009-2011 with H3 IAV isolates of
96
human, swine, canine and avian origins. The antigenic profiles from HI tests were validated by
97
microneutralization (MN) assays. The genomic sequences of the HA and NA segments of
98
selected swine-origin H3N2 IAV isolates were also analyzed to investigate the genomic changes
99
contributing to antigenic profiles.
100 101
Materials and methods
102
Viruses and Sera. From 2009 to 2011, a systematically designed surveillance effort was
103
performed across 53 agricultural fairs in Ohio, USA (27). From the 1,073 pigs sampled, 155
104
were positive for IAV, with 128 H3N2 viruses isolated. Sixty-eight representative (from the 128
105
total) H3N2 IAV (TABLE 1) were selected for the present study. The HA, NA, and M genes of
106
these isolates were characterized as described previously (27). Additionally, four H3N2v IAV
107
isolates (TABLE 1) were kindly provided by CDC, Atlanta, GA. Also included in the study, and
108
listed in TABLE 2 and 3, were twelve H3N2 IAV isolates from commercial pigs, two H3N2 IAV
109
isolates from wild birds, two H3 IAV isolates from dogs (28, 29), and 12 human seasonal H3N2
110
isolates. After being propagated in Madin-Darby canine kidney (MDCK) cells (ATCC,
111
Manassas, VA, USA), all IAVs were aliquoted and stored at -70˚C until use.
112
Ferret antisera were generated in 6 to 8-week-old ferrets, which had baseline HI titers less
113
than 1:10 against A/Brisbane/10/2007(H3N2), A/Brisbane/59/2007(H1N1), and
114
A/California/4/2009(H1N1). Each ferret was inoculated intranasally with 106 TCID50 for each
115
IAV isolate (TABLE 1). Sera were collected from the ferrets at three weeks post inoculation if
116
the HI titer in a ferret was at least 1:160 at 2 weeks post inoculation. Otherwise, a second dose of
117
106 TCID50 of the corresponding virus was given and serum was collected five weeks post first
118
inoculation. All sera were aliquoted and stored at -70˚C until use.
119
Eight swine H3N2 IAV isolates representing six fair events were used to generate ferret
120
antisera. The virus designation is summarized in TABLE 4. In addition, ferret antisera were
121
produced against 12 human seasonal H3N2 IAV isolates, two H3N2 IAV isolates from wild
122
birds, and two canine H3 IAV isolates for comparison. All these isolates are contemporary H3
123
IAVs (TABLE 1, 2, and 3).
124 125
Genomic sequencing. Viral RNA was extracted using the RNeasy Mini Kit according to the
126
manufacturer’s instructions (QIAGEN Inc., Valencia, CA, USA). Reverse transcription and PCR
127
were performed using influenza-specific primers (30). PCR products were separated using
128
agorose gel electrophoresis and purified using the QIAquik Gel Extraction Kit (QIAGEN Inc,
129
Valencia, CA, USA). Sequencing was performed by the Life Sciences Core Laboratories Center
130
at Cornell University using the Applied Biosystems Automated 3730 DNA Analyzer, with Big
131
Dye Terminator chemistry and AmpliTaq-FS DNA Polymerase. The HA genes of nine
132
agricultural fair isolates are fully sequenced, and the HA genes of 12 commercial swine IAV
133
isolates were partially sequenced and at least covered the entire HA1. The GenBank accession
134
numbers of these sequences are xxxxxx-xxxxxx (pending).
135 136
Hemagglutination inhibition assay. Before conducting HI tests, ferret antisera were treated
137
with receptor-destroying enzyme (RDE) (Denka Seiken Co., Tokyo, Japan) at 37˚C for 18 hours,
138
and then heat inactivated at 55˚C for 30 minutes. The sera were then diluted with phosphate-
139
buffered saline (for a final dilution of 1:10), and tested by HI assay with 0.5% turkey red blood
140
cells according to the WHO manual on animal influenza diagnosis and surveillance
141
(http://www.who.int/vaccine_research/diseases/influenza/WHO_manual_on_animal-
142
diagnosis_and_surveillance_2002_5.pdf).
143 144
Microneutralization assay. The sera were tested by performing a microneutralization (MN)
145
assay in MDCK cells. Neutralizing titers were expressed as the reciprocal of the serum dilution
146
that inhibited 50% of viral growth of 100 tissue culture infectious dose (TCID50) of virus. The
147
MN titers were determined by HA assay using 0.5% turkey red blood cells (31).
148 149
Antigenic cartography and antigenic cluster identification. Antigenic cartography was
150
generated based on HI data using AntigenMap (32, 33) by setting 1:10 as the low reactor. Data
151
normalization was performed as previously reported (34-36). Influenza A viruses were classified
152
into antigenic clusters using k-mean clustering (37). Student’s t-test was performed to assess the
153
significance of antigenic distances.
154 155
Molecular characterization and phylogenic analysis. The multiple sequence alignments were
156
conducted by using the MUSCLE software package (38). The phylogenetic analyses were
157
performed using maximum likelihood by GARLI version (39), and bootstrap resampling
158
analyses were conducted using PAUP* 4.0 Beta (40) with a neighborhood joining method, as
159
previously described (41).
160
161 162
Results
163
Genetic characterization of H3N2 influenza A viruses. Phylogenetic analysis of HA genes
164
showed the HA genes of four human H3N2v isolates and all H3N2 IAVs isolates from both
165
agricultural fairs and commercial swine farms belonged to cluster IV (FIG. 1A) (22). A high
166
degree of genetic diversity was observed among the viruses. The amino acid sequence identity
167
between the HA amino acid sequence of A/swine/Ohio/10SW130/2010(H3N2) and that of
168
A/swine/Ohio/11SW111/2011(H3N2) was 99.3%; the HA protein sequence identity between
169
A/swine/Ohio/09SW64/2009(H3N2) and A/swine/Ohio/10SW130/2010(H3N2) and that
170
between A/swine/Ohio/09SW64/2009(H3N2) and A/swine/Ohio/11SW111/2011(H3N2) was
171
95.8% and 95.1%, respectively and the human H3N2v isolate A/Iowa/07/2011(H3N2) was
172
99.5% identical to A/swine/Ohio/11SW111/2011(H3N2) (TABLE 5).
173
The phylogenetic trees of influenza NA genes showed that the NA genes of these
174
agricultural fair isolates were phylogenetically close to the NA gene of
175
A/swine/Ontario/33853/2005 and to those NA genes of other IAVs in the same cluster, as similar
176
to the HA genes (FIG. 1B).
177 178
Antigenic diversity of H3N2 IAVs from pigs at agricultural fairs. The antisera generated
179
against selected IAV isolates from pigs at Ohio fairs during 2010 and 2011 cross-reacted with
180
the isolates from 2010 and 2011 fairs, but were less reactive to IAV recovered from pigs at fairs
181
during 2009. Similarly, the antisera generated against IAV isolates from the 2009 fair season
182
reacted poorly with 2010 and 2011 swine isolates (TABLE 1). The IAV isolates from pigs at
183
agricultural fairs could be divided into two antigenic groups: 1) the 2009 fair isolates (antigenic
184
cluster H3N2-alpha), and 2) the 2010 and 2011 fair isolates (antigenic cluster H3N2-beta).
185
The k-mean clustering also grouped 84 isolates (68 exhibition swine isolates, 4 human
186
H3N2v isolates, and 12 commercial swine isolates) into one of the two identified antigenic
187
clusters (35 in cluster alpha and 49 in cluster beta) as summarized in TABLE 1. The average
188
distance between the viruses within cluster alpha and cluster beta was 1.45 and 1.07 units,
189
respectively. The minimum distance between antigenic clusters was 2.32 units (e.g.,
190
A/Pennsylvania/14/2010 and A/swine/Iowa/11333/2006) and the maximum distance between
191
clusters was 6.53 units (e.g., A/swine/North Carolina/23592/2009 and
192
A/swine/Ohio/11SW214/2011). The average distance between two clusters from their centers
193
was 4.18 units. Each unit corresponds to a 2-fold change in HI assay. Statistically, the antigenic
194
distances between clusters were significantly different from those within each antigenic cluster
195
(p