Research Project: Intervention Strategies to Control Influenza A Virus Infection in Swine

Location: Virus and Prion Research

2021 Annual Report

Objectives
Objective 1. Identify mechanisms of influenza A virus (IAV) pathogenesis and host adaptation to swine. This includes investigating host-pathogen interactions at cellular or molecular levels, identifying determinants of swine IAV infection and shedding from respiratory mucosa, and investigating host range restriction to identify mechanisms by which non-swine adapted viruses infect and adapt to swine. Objective 2. Evaluate emerging IAV at the genetic and antigenic levels as a risk to swine or other host species. This includes identifying emerging IAV and monitoring genetic and antigenic evolution in swine, and identifying genetic changes important for antigenic drift or pathogenicity in swine or other hosts. Objective 3. Identify novel influenza vaccine platforms and improve vaccination strategies. This includes characterizing humoral and cellular immune responses to wild-type and attenuated viruses compared to inactivated vaccines to identify correlates of protection, investigating adjuvants or immune-modulatory agents that result in robust immune responses (mucosal delivered, long lived, broadly cross-protective and/or reduce the number of vaccine boosters), and investigating technologies to override IAV vaccine interference from passively acquired immunity.

Approach
Influenza A virus (IAV) will be investigated in swine or relevant in vitro models to 1) understand the genetic predictors of host range and virulence in swine; 2) understand the genetic and antigenic variability of endemic viruses and how this affects vaccine strain selection and efficacy; and 3) develop new vaccines that can override maternally-derived antibody interference and provide broader cross-protection. Disease pathogenesis, transmission, and vaccine efficacy studies will be conducted in the natural swine host. Knowledge obtained will be applied to break the cycle of transmission through development of better vaccines or other novel intervention strategies. Computational biology methods will be used to evaluate virus evolution in the natural host to enable predictions to be made on virulence and/or antigenic factors. These predictions will be tested in the lab and in animal studies with wild type viruses and through the use of reverse engineering and mutational studies to identify virulence components of IAV. Experimentally mutated viruses will be evaluated by test parameters that measure both virus and host properties. Development of vaccines that provide better cross-protective immunity than what is currently available with today's vaccines will be approached through understanding correlates of protection, the impact of prior exposure or passive immunity, and through vaccine vector platform development, attenuated strains for vaccines, and other novel vaccine technologies.

Progress Report
This is the final report for the project 5030-32000-120-00D terminating October 18, 2021. In support of Objective 1, Subobjective 1.1, to investigate host-pathogen interactions at cellular or molecular levels, RNA was extracted from porcine alveolar macrophages and lung tissue from vaccinated or unvaccinated pigs infected with IAV and compared to uninfected pigs. Host gene expression profiles were examined using transcriptomic messenger RNA sequencing techniques and using PCR arrays targeting 168 genes associated with swine antiviral response and cytokine and chemokine pathways. Differential gene expression patterns were observed by both techniques. In support of Objective 1, Subobjective 1.2, to identify determinants of swine IAV infection and shedding at respiratory mucosa, infectivity and replication in a swine cell line with swine IAV isolates correlated with in vivo susceptibility in pigs. Additional IAV were tested to assess the potential use of this cell line to predict susceptibility and virus phenotype in pigs. Four conserved hemagglutinin (HA) amino acid mutations in 2010.1 swine field strains were identified that were distinct from human seasonal H3 IAV. Using site-directed mutagenesis in various in vitro assays, these four HA amino acids were evaluated in the adaptation of human H3 IAV to swine. Assays included receptor binding avidity, surfactant protein D neutralization, virion stability in a thermal stability HA assay, and replication in primary porcine respiratory cell cultures. A wildtype swine IAV was compared to isogenic engineered viruses with mutations in the HA gene altering glycosylation sites compared against two human seasonal strains in in vitro and in vivo experiments. Subtle differences in transmission were detected. A swine strain identified as a recent human seasonal IAV introduction replicated to higher titers compared to its human seasonal precursor viruses in cell culture and transmitted to contact pigs, but the human seasonal H3N2 did not. In support of Objective 1, Subobjective 1.3, to investigate host range restriction to identify mechanisms by which non-swine adapted viruses infect and adapt to swine, a recently established 2010.1 H3 virus lineage in swine was shown to be antigenically distinct from human seasonal precursor H3. Four mutations in the HA of the swine adapted H3 compared to the precursor human seasonal H3 were shown to be important for swine adaptation of human seasonal H3 using in vitro methods, and a pig pathogenesis and transmission study was conducted. Representative human and swine viruses were used to perform virus histochemistry on swine tissue and in vitro replication assays. A pathogenesis and transmission study was completed with a North American 2017 H7N9 low pathogenic avian influenza virus, showing limited infection and no transmission. Swine to ferret transmission studies were completed to assess the risk of dominant swine H1 and H3 IAV to humans since ferrets are the gold standard animal model for human influenza research. All swine IAV strains transmitted to ferrets via aerosol, although differences in transmission efficiency were observed. A machine learning modeling study was trained on sequence data collected from swine and humans from 2010 to 2020 and used to predict host origin of genetic sequence data to determine the role of H1N1pdm09 burden in human populations and transmission to swine. In support of Objective 2, Subobjective 2.1a, to identify emerging IAV and monitor genetic and antigenic evolution in swine, genetic patterns were monitored to identify changing patterns or emerging viruses. IAV with molecular signatures suggesting antigenic changes were identified and virus isolates obtained from the USDA IAV-S surveillance repository for characterization. Phylogenetic methods were used to infer molecular evolutionary rates of HA gene sequences. The most dominant viruses based upon surveillance data demonstrated increased evolutionary rates, and structural modeling of the HA suggested antigenic drift. Comprehensive phylogenetic analyses were conducted on publicly available swine IAV genes. A one letter code for each internal gene was used to designate the genetic constellation and paired with HA and NA phylogenetic clade combinations, partitioned by state, and clustered by distance metrics to identify temporal and spatial patterns. An evolutionary distance-based method was developed to identify gene reassortment, and an algorithm that merged the evolutionary history of individual genes into a larger phylogenetic network describing the evolution of viruses was developed. An in vivo pig study was completed to evaluate pathogenesis and transmission of IAV from swine with differing whole genome gene constellations. In support of Subobjective 2.1b, a phylogenetic based method for classifying H1 IAV was developed and validated on a large global dataset. The highly accurate automated tool was implemented on the Influenza Research Database (IRD; fludb.org). An automated clade tool for H3 with standardized global nomenclature and statistical criteria was developed to assign lineage categories to sequence data. Criteria were included to classify co-circulating swine H3 lineages, human, avian, canine, and equine H3 sequences. This tool will be implemented on IRD. An automated pipeline was created to assign lineage and genetic clade to query gene segments (octoFLU) and a graphical web interface supported by a novel network NoSQL database is in development to visualize phylogenetic categorization was developed (octoFLU-SHOW) on a website hosted on Amazon Web Services as part of SCINet (https://flu-crew.org). In support of Objective 2, Subobjective 2.2, to identify genetic changes important for antigenic drift or pathogenicity in swine or other hosts, H1 and H3 with unique antigenic motifs, predicted to be antigenically distinct, were obtained from the USDA IAV-S surveillance system and collaborators to test by hemagglutination inhibition (HI) assays. Vaccines developed to control IAV infection focus on the HA protein, and the decision to update vaccine components includes sequencing and analyzing HA genes. New antigenic motif patterns in H3 were shown to be distinct and changed in frequency of detection over time. Additional contemporary H3N2 viruses and H1N1/N2 viruses were also tested. Antigenic evolution of N2 was evaluated in neuraminidase inhibition (NI) assays and quantified by antigenic cartography. Machine learning models were parameterized by HI data and demonstrated that H3 sequence identity and mutations at ten amino acid positions predicted antigenic distance. An H3 phylogenetic clade possessing an N156H amino acid substitution and different N2 and nucleoprotein genes by reassortment was detected with increasing frequency. The mutation in the HA gene did not result in antigenic drift, suggesting that increased detection was driven by the reassorted genome. A Nextstrain web application to integrate these data was deployed on Amazon Web Services as part of SCINet (https://flu-crew.org). In support of Objective 3, Subobjective 3.1, to characterize humoral and cellular immune responses to wild-type and attenuated viruses compared to inactivated or vectored vaccines to identify correlates of protection, whole inactivated virus (WIV), live attenuated influenza virus (LAIV), and an RNA vectored vaccine platforms were compared against IAV with H3 that differed in only a few key amino acid positions. The LAIV and RNA vectored vaccines demonstrated superior protection from heterologous challenge. An in vivo study was conducted to evaluate vectored HA and NA vaccines. Pigs were vaccinated with replicon particle vaccines expressing matched or mismatched HA in combination with matched NA and challenged with H1N1 and compared to WIV. A study was conducted in pigs to assess sequential vaccination with H1 viruses that varied at key antigenic sites and the impact on VAERD. Human seasonal H1N1 strains prior to 2009 were used to vaccinate the pigs, followed by vaccination with H1N1pdm09 and challenge with an H1N1pdm09 drifted strain. Sequential vaccination skewed neutralizing antibody responses toward shared epitopes but did not appear to influence protection from challenge. In support of Objective 3, Subobjective 3.2, to investigate adjuvants or immune-modulatory agents that result in robust immune responses, a study was conducted to test the effect of sequential heterologous infection in imprinting the humoral immune response. The order of infection significantly impacted the humoral immune response to each of the viruses and certain exposure patterns led to increased lung pathology. To test LAIV engineered to carry immunomodulatory genes to improve heterologous protection, a vaccination study was conducted. In collaboration with University of Georgia, a bivalent reverse engineered LAIV with H1N1 and H3N2 strains expressing a synthetic swine IgA-inducing protein (IGIP) was generated and tested in pigs. The LAIV with or without IGIP were compared to each other and to non-vaccinated pigs. A vaccine study to compare cell mediated immunity between different vaccine platforms was also completed. An in vivo study was conducted in pigs to assess the impact of a nonreplicating human adenovirus type 5 vector expressing the porcine interferon-alpha (Ad5-pINF-a) on viral replication and disease.

Accomplishments
1. Evolution of N2 neuraminidase genes from swine IAV in the United States impacted immunity. The influenza A virus (IAV) neuraminidase (NA) gene encodes for one of two major viral surface glycoproteins important for antibody immunity against the virus. NA in swine are either N1 or N2 subtypes. The N2 subtype currently accounts for approximately two-thirds of the NA detections in IAV from U.S. domestic swine populations. These N2 genes evolved in swine following two separate introductions of human seasonal IAV N2 into swine in 1998 and 2002. Increased genetic diversity of the N2 gene suggested an increase in antigenic evolution of the virus surface glycoprotein and may allow escape from natural and/or vaccine induced immunity. To assess the potential loss in immune recognition among N2 from swine field strains, ARS scientists in Ames, Iowa, characterized the antigenic evolution of swine IAV by generating a panel of antisera against representative N2 from different years after N2 introduction and measured how antibody recognition changed over time. We found in the 20+ years following the introductions of IAV N2 from humans, the genetic diversity of N2 genes in swine increased. This corresponded with an increase in antigenic diversity of the N2 glycoproteins measured by antibody recognition. Antibodies generated against representative N2 of the 1998 N2-lineage did not inhibit N2 of the 2002 lineage and vice versa. Further, both the 1998 and 2002 N2 lineages displayed substantial amount of antigenic diversity shown by loss of antibody recognition. Knowledge on the antigenic diversity of N2 subtypes in swine over time will aid vaccine design and prevention strategies for swine veterinarians and pork producers.

2. Antigenic distances between North American swine H3N2 and human seasonal influenza A H3N2 viruses are an indicator of zoonotic risk to humans. Human H3N2 influenza A viruses (IAV) spread to pigs in North America in the 1990s and more recently in the 2010s. These cross-species events led to sustained circulation of H3N2 in swine and increased IAV diversity in pig populations. The evolution in swine H3N2 led to a reduced similarity with human seasonal H3N2 and the vaccine strains used to protect human populations. ARS scientists in Ames, Iowa, quantified the antigenic properties and found that North American swine H3N2 lineages retained more antigenic similarity to historical human vaccine strains from the decade of incursion but had substantial differences compared with more recent human H3N2 vaccine strains. Additionally, pandemic preparedness H3N2 vaccine strains developed for public health also demonstrated a loss in similarity with these contemporary swine strains. Lastly, post-exposure and post-vaccination human sera revealed that although these adults had antibodies against human H3N2 strains, many had limited immunity to swine H3N2 This was particularly evident for the swine H3N2 from the 1990s, and the limited immunity was greater in older adults born before 1970. These antigenic assessments of swine H3N2 provide critical information for pandemic preparedness and candidate vaccine development for human populations.

3. Development of a novel live attenuated influenza A virus vaccine encoding the IgA-inducing protein (IGIP) indicates potential for incorporating additional genes to improve immunity. Seasonal influenza A virus (IAV) infections are among the most important respiratory diseases worldwide. Vaccination is considered the first line of defense against IAV, but virus evolution through genetic change makes vaccines less effective after a single season or against pandemic strains. Live attenuated influenza virus (LAIV) vaccines induce a combination of immunity in the bloodstream and respiratory tract by mimicking a natural infection. To further enhance protective respiratory mucosal responses upon intranasal LAIV administration, ARS scientists in Ames, Iowa, collaborated with scientists at University of Georgia to incorporate into the LAIV a gene called IgA-inducing protein (IGIP), which is reported to enhance the production of IgA antibody secreted by the respiratory tract mucosa. Mice were vaccinated with the modified IGIP-LAIV and antibodies from immunized mice were tested for binding to a panel of IAV. Serum and respiratory tract fluid from IGIP-LAIV vaccinated mice had a trend for higher antibody responses to IAV proteins. These results show that IAVs are amenable to the introduction of genes encoding immunomodulatory functions that can serve to improve LAIV. Such types of modifications could improve overall vaccine safety and efficacy and these studies are significant in the context of developing more efficacious vaccine approaches against influenza.

4. Antigenic drift of an endemic H3 lineage of IAV in swine was linked to a single amino acid mutation in the HA protein. Influenza A virus (IAV) is an important pathogen in birds, humans, pigs and other mammals. Spread of IAV from humans to pigs plays an important role in the diversity of IAV found in pig populations, and the diversity of IAV in pigs complicates control by swine vaccines. During the 2010-11 human influenza season, a human H3N2 IAV infected pigs in the United States. A portion of the virus genome, the hemagglutinin (HA) gene, was maintained in swine, and these viruses with this H3 continue to be frequently detected in swine herds. ARS scientists in Ames, Iowa, investigated how these H3N2 IAV changed over time by assessing recognition by the swine antibody response and by infection properties in pigs. This swine H3N2 IAV also previously infected people at multiple agricultural fairs and therefore is a public health concern. We showed that these swine H3N2 IAV evolved in recent years to acquire genetic changes related to evading antibody immunity through a single amino acid mutation in the HA, and each of the strains we tested were fully capable of infecting pigs, causing disease and spreading among pigs. This understanding will indicate the risk to swine populations and aid in improving influenza vaccines used in the swine industry. These contemporary swine H3N2 viruses should be considered for inclusion in swine influenza vaccines and as a potential risk to humans.

5. OFFLU Animal Influenza Report to the World Health Organization: Swine detections from July 2020 to December 2020. Regular characterization of influenza A viruses (IAV) currently circulating in swine provides rational criteria for vaccine strain selection, control strategies, and may identify viruses with pandemic potential. IAV frequently move between people and pigs. To understand how bidirectional transmission of IAV between humans and swine impacts agricultural production and human health, ARS scientists in Ames, Iowa, in collaboration with the joint World Organization for Animal Health (OIE) and Food and Agriculture Organization of the United Nations (FAO) scientific network on animal influenza, OFFLU, quantified the global genetic diversity of swine IAV circulating from 2016 to present, and the genetic diversity of swine IAV in the U.S. over the past six months. The similarity of circulating swine IAV diversity was compared to human IAV vaccine and current candidate human vaccine viruses (CVV) that are used for pandemic preparedness. Using a local knowledgebase called octoFLUdb, representative strains were characterized with a panel of anti-sera derived from human vaccine strains or CVV strains to understand the risk of swine IAV to human populations. These data demonstrated that only 11 of the 32 distinct IAV types detected in swine globally were recognized by human sera from CVV or previous human vaccine strains; and the degree to which those vaccines provide protection to swine IAV strains is therefore dubious. This work identified swine IAV that should be considered in vaccine strain selection for pandemic preparedness.

6. OFFLU Animal Influenza Report to the World Health Organization: Swine detections from February 2020 to September 2020. Regular characterization of influenza A viruses (IAV) provide rational criteria for vaccine strain selection, control strategies, and may identify viruses with pandemic potential. From the 1918 human influenza pandemic to date, human-to-swine interspecies IAV transmission events have repeatedly occurred, some leading to sustained transmission and increased IAV diversity in pig populations. These swine IAV have the potential to be reintroduced back into the human population if they are substantially different from current human seasonal strains. ARS scientists in Ames, Iowa, in collaboration with the joint World Organization for Animal Health (OIE) and Food and Agriculture Organization of the United Nations (FAO) scientific network on animal influenza, OFFLU, quantified the global genetic diversity of swine IAV circulating from 2018 to present, and the genetic diversity of swine IAV in the USA over the past six months. We determined the similarity between circulating swine IAV diversity and human IAV vaccine strains and current candidate vaccine viruses (CVV) that are used in human pandemic preparedness efforts. Using a local knowledgebase called octoFLUdb (https://github.com/flu-crew/octofludb), and novel sampling algorithms (https://github.com/flu-crew/flutile and https://github.com/flu-crew/smot), representative strains were antigenically characterized with a panel of anti-sera derived from human vaccine strains or CVV strains to understand the risk of swine IAV to human populations. These data demonstrated that the tested swine IAV was significantly different to the current H1 and H3 components of human IAV vaccines. Additionally, only 7 strains of the 24 distinct genetic clades detected in swine globally were by recognized by CVV antibodies; and the degree to which those CVVs may provide protection was dubious given observed genetic differences. This work demonstrated that monitoring IAV in swine populations is a critical aspect in pandemic preparedness.

7. A novel algorithm for inferring reassortment and evolution in segmented RNA viruses indicates new strains of concern. The identification of genetically novel influenza A viruses (IAV) that are a mix of genes derived from human-, swine-, or avian-origin IAV is critical for controlling infection in swine. These novel viruses may be undergoing rapid changes in genetic diversity that reduce the efficacy of vaccine control methods and may also pose a greater risk to humans for zoonotic infection. In this study, ARS scientists in Ames, Iowa, in collaboration with scientists at Iowa State University, developed an algorithm that merged the evolutionary history of individual genes into a larger phylogenetic network describing the evolution of the virus genome. The accuracy of the algorithm was validated using swine IAV genome sequencing data with a known evolutionary history that included transmission of human IAV into swine and subsequent mixing of IAV genes. The algorithm detected known gene mixing events, along with additional events between genetically different circulating swine IAV strains. The development of this algorithm provides computational support for swine IAV surveillance as it can objectively rank the novelty of swine IAV strains. These data may aid vaccine development through the objective targeting of novel IAV strains at risk for spread among swine and reduce the risk of interspecies transmission by identifying new emerging viruses that may be of greater risk to humans.

8. Detection of a swine-origin influenza A(H1N1)pdm09 virus infecting a human and development of sequence based methods to distinguish species of origin. Influenza virus infects a wide range of hosts, resulting in illnesses that vary from asymptomatic cases to severe pneumonia and death. Viral transfer can occur between human and non-human hosts resulting in circulation in novel hosts. In this work, ARS scientists in Ames, Iowa, in collaboration with scientists at the CDC, identified the first case of a swine-origin influenza A(H1N1)pdm09 virus resulting in a human infection. This shows that A(H1N1)pdm09 viruses not only infect swine hosts, but continued to evolve and distinguish themselves from previously circulating human-origin influenza viruses. The development of techniques for distinguishing human-origin and swine-origin viruses are necessary for the continued surveillance of influenza viruses. We showed that unique genetic signatures in the hemagglutinin gene differentiated circulating swine-associated strains from circulating human-associated strains of influenza A(H1N1)pdm09, and these signatures can be used to enhance surveillance of swine-origin influenza to inform public health of interspecies transmission between humans and swine.

9. A machine learning model was developed to predict antigenic drift from genetic sequence data. Vaccines developed to control influenza A virus infection focus on the hemagglutinin (HA) protein, with custom vaccine components often determined through sequencing of HA genes. As the HA evolves, prior antibody immunity can fail to recognize mutated strains and cause vaccines to be less protective, a process known as antigenic drift. Though sequencing of HA provides evidence of change, there are no rapid and accurate predictive approaches that adequately link HA sequence data to vaccine protection. ARS scientists in Ames, Iowa, evaluated a combination of machine learning models to estimate virus antigenic features from HA genetic sequence data. The model was used to identify and rank the importance of mutations in the HA gene that caused changes in the HA protein and predicted whether these would affect antibody recognition. Four previously untested IAV strains were selected to experimentally validate model predictions. Errors between predicted and measured distances of uncharacterized strains were 0.35, 0.61, 1.69, and 0.13 antigenic units. Linking changes in the HA protein to recognition by antibody has important implications for effective vaccine design. The findings from this study are critical to help inform vaccine manufacturers and swine producers on how to develop and implement vaccines to control influenza A virus infection more effectively.

Review Publications
Zeller, M.A., Gauger, P.C., Arendsee, Z.W., Souza, C.K., Vincent, A.L., Anderson, T.K. 2021. Machine learning prediction and experimental validation of antigenic drift in H3 influenza A viruses in swine. mSphere. 6(2). https://doi.org/10.1128/mSphere.00920-20.
Cook, P.W., Stark, T., Jones, J., Kondor, R., Zanders, N., Benfer, J., Scott, S., Jang, Y., Janas-Martindale, A., Lindstrom, S., Blanton, L., Schiltz, J., Tell, R., Griesser, R., Shult, P., Reisdorf, E., Danz, T., Fry, A., Barnes, J., Vincent, A.L., Wentworth, D.E., Davis, T. 2020. Detection and characterization of swine-origin Influenza A(H1N1) pandemic 2009 viruses in humans following zoonotic transmission. Journal of Virology. 95(2). https://doi.org/10.1128/JVI.01066-20.
Kaplan, B.S., Falkenberg, S.M., Dassanayake, R.P., Neill, J.D., Velayudhan, B., Li, F., Vincent, A.L. 2020. Virus strain influenced the interspecies transmission of influenza D virus between calves and pigs. Transboundary and Emerging Diseases. https://doi.org/10.1111/tbed.13943.
Powell, J.D., Abente, E.J., Chang, J., Souza, C.K., Rajao, D.S., Anderson, T.K., Zeller, M.A., Gauger, P.C., Lewis, N.S., Vincent, A.L. 2021. Characterization of contemporary 2010.1 H3N2 swine influenza A viruses circulating in United States pigs. Virology. 553:94-101. https://doi.org/10.1016/j.virol.2020.11.006.
Vandoorn, E., Leroux-Roels, I., Leroux-Roels, G., Parys, A., Vincent, A.L., Van Reeth, K. 2020. Detection of H1 Swine Influenza A Virus in Human Serum Samples by Age Group. Emerging Infectious Diseases. 26(9):2118-2128. https://doi.org/10.3201/eid2609.191796.
Caceres, J.C., Cardenas-Garcia, S., Jain, A., Gay, C.J., Carnaccini, S., Seibert, B., Ferreri, L.M., Geiger, G., Nakajiima, R., Perez, D., Rajao, D.S., Isakova-Sivak, I., Rudenko, L., Vincent, A.L., Davies, H. 2021. Development of a novel Live Attenuated Influenza A Virus vaccine encoding the IgA-inducing protein. Vaccines. 9(7). Article 703. https://doi.org/10.3390/vaccines9070703.