An avian influenza virus A(H7N9) reassortant that recently emerged in the United States with low pathogenic phenotype does not efficiently infect swine

In 2017, outbreaks of low and highly pathogenic avian H7N9 viruses were reported in four States in the United States. In total, over 270 000 birds died or were culled, causing significant economic loss. The potential for avian‐to‐swine transmission of the U.S. avian H7N9 was unknown. In an experimental challenge in swine using a representative low pathogenic H7N9 (A/chicken/Tennessee/17‐007431‐3/2017; LPAI TN/17) isolated from these events, no infectious virus in the upper and minimal virus in the lower respiratory tract was detected, nor was lung pathology or evidence of transmission in pigs observed, indicating that the virus cannot efficiently infect swine.

In 2017, outbreaks of low and highly pathogenic avian H7N9 viruses were reported in four States in the United States. In total, over 270 000 birds died or were culled, causing significant economic loss. The potential for avian-to-swine transmission of the U.S. avian H7N9 was unknown. In an experimental challenge in swine using a

| Animal study
Sixty-five crossbred pigs were obtained from a healthy herd for the challenge study and split into naïve (Table 1) and pre-immune groups ( Table 2). For the naïve groups, two distinct age cohorts were used to evaluate whether age affected susceptibility to infection. On the day of challenge, fifteen pigs were 4 weeks old and twenty pigs were 14 weeks old. Pigs were confirmed to be seronegative to influenza A virus (IAV) antibodies against the nucleoprotein (NP) as measured by ELISA (Swine Influenza Virus Antibody Test, IDEXX, Westbrook, ME) prior to the study. Pigs from both age cohorts were divided into two groups: non-challenged and challenged. Additionally, contact pigs were included in the 4-week-old cohort to examine transmission. For pre-immune experiments, pigs were inoculated with 2 mL of 10 5 TCID50/ naïve contact pigs were placed in a separate raised deck in the same room as the ten 4-week-old pigs in the naive group that had been challenged, but with separate food and water and approximately 2 feet (0.6 m) away from the inoculated group to evaluate indirect contact transmission.

| Sample collection and pathological examination of lungs
Nasal swab samples were collected daily 1-5 dpi for challenged and negative control pigs, and 1-5, 7, and 9 days post-contact (dpc) for contact pigs. Negative control and challenged pigs were humanely 0/10 0/10 0/10 a Pigs exposed to 2 mL of 10 5 TCID 50 /mL delivered intranasally at 4 weeks of age. b Samples collected at 1-5 dpi. c HI assay performed using LPAI TN17 as the antigen.
of the surface of the entire lung affected by pneumonia was calculated on the basis of the weighted proportions of each lobe to the total lung volume. Five contact pigs were humanely euthanized at 16 dpc, and sera were collected to assess seroconversion.  (Table 1).

RNA was extracted from NS and BALF samples with the MagMAX
Viral RNA Isolation Kit (Thermo Fisher Scientific Inc, Waltham, MA).

RT-PCR was performed with the VetMAX-Gold SIV Detection Kit
following the manufacturer's instructions (Thermo Fisher Scientific Inc) with a Ct value of less than 35 deemed as definitively positive for influenza.

| Serology
Sera were treated with receptor-destroying enzyme (Denka Seiken, Japan), heat inactivated at 56°C for 30 minutes, and adsorbed with 50% turkey red blood cells (RBCs) to remove non-specific hemagglutinin inhibitors and natural serum agglutinins. Hemagglutination inhibition assays were performed with TN/17 as the antigen and 0.5% turkey RBCs using standard techniques. All of the sera samples were negative for HI activity and NP reactivity. Influenza RNA was detected by RT-PCR in six of twenty inoculated pigs in the BALF samples from 5 dpi, but in none of the nasal swabs. Consistent with the absence of viral infection in challenged animals, contact pigs were also negative for virus and IAV-specific antibodies. Samples from non-challenged pigs (n = 10 for each age group) were negative for virus isolation, and NP and HI titers.

| D ISCUSS I ON
Experimental challenge of pigs with a LPAI H7N9 isolate resulted in minimal replication in the lung with no evidence of transmission to contact pigs, suggesting a low risk to the swine population. Due to logistical challenges of studying the HPAI H7N9, we tested the LPAI H7N9 TN/17 isolate due to its genetic relatedness to the HPAI that was detected at the same time. The HA of the LPAI TN/17 isolate used in this study shares high homology with the HPAI H7N9 across the genome (see Figure S1, Table S1). H7N9 of other phylogenetic clades have caused five epidemic waves in humans in China since 2013 and remain a public health concern. 8 Notably, Asian LPAI H7N9 has caused the majority of human infections with most human isolates having acquired G186V and Q226L/I substitutions in the HA and E627K substitution in the PB2, 8,9 although HPAI H7N9 remains a risk to the human population. 10 The LPAI TN/17 tested in this study is of North American wild bird lineage and does not possess these previously identified adaptive substitutions.
Three pig challenge studies found that early human isolates from the first wave of the H7N9 epidemic in China, A/Shanghai/02/2013 and A/Anhui/1/2013, caused mild disease although infectious virus was isolated from nasal swab samples. [11][12][13] Both of these human isolates encode mammalian adaptive substitutions in the HA ( Figure S1). In one study, a control avian LPAI A/chicken/Zhejiang/ DTID-ZU01/2013 H7N9 was used that has the G186V substitution although it encodes 226Q in the HA and virus shedding was observed. Furthermore, serially passaged A/Anhui/1/2013 acquired a L226Q change in pigs, suggesting that G186V may be more important for infection in pigs. 12 Belser et al 2 found that ferret infections with both LPAI and HPAI were restricted to the upper respiratory tract, but nasal titers were positive in only one of three contact ferrets for LPAI TN/17 while there was no contact transmission with the HPAI. In BALB/c mice, both LPAI and HPAI infections resulted in mild illness and no deaths.
In contrast to these animal studies, we found that the LPAI TN/17 was restricted from efficiently replicating in pigs, even in the upper respiratory tract where ferrets were susceptible. We also examined swine susceptibility to LPAI TN/17 in pre-immune pigs exposed to H3 and H1 viruses and did not detect any differences in protection or susceptibility similar to the naïve pigs.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.