Equine Influenza Virus Gabriele A. Landolt,
Dr med vet, MS, PhD
KEYWORDS Influenza A virus Virus evolution Epidemiology Diagnosis Control KEY POINTS Despite extensive use of vaccines, equine influenza virus continues to be one of the most important equine viral respiratory pathogens. Control of equine influenza virus is hampered by continued genetic evolution of the virus (antigenic drift). Antigenic drift negatively affects the degree of immunoprotection evoked by inactivated vaccines. Isolation and genetic characterization of currently circulating equine influenza viruses remains a priority, and is essential for vaccine strain selection.
INTRODUCTION
Influenza is a well-known and ancient disease. In fact, a disease resembling influenza was described by Hippocrates in 412 BC. Yet hardly a month goes by without a new headline about influenza in the media. Although millions of dollars have been spent on research, influenza virus continues to challenge our understanding of its ecology and our ability to control its spread. Two key reasons why influenza virus has remained one of the most important causes of viral respiratory disease are its potential for establishing genetic and antigenic diversity and its ability to occasionally transmit between different host species.1 For decades the horse has been viewed as an isolated or “dead-end” host for influenza A viruses, with equine influenza virus being considered as relatively stable genetically. Although equine influenza viruses are genetically more stable than their human-lineage counterparts, they are by no means in evolutionary stasis. Moreover, recent transmission of equine-lineage influenza viruses to dogs also challenges the horse’s status as a dead-end host. This article reviews recent developments in the epidemiology and evolution of equine influenza virus. In addition, the clinical presentation of equine influenza infection, diagnostic techniques, and vaccine recommendations are briefly summarized.
Funding Sources: Merck Animal Health. Conflict of Interest: None. Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, 300 West Drake Road, Fort Collins, CO 80523, USA E-mail address: [email protected] Vet Clin Equine 30 (2014) 507–522 http://dx.doi.org/10.1016/j.cveq.2014.08.003 vetequine.theclinics.com 0749-0739/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
508
Landolt
ETIOLOGY
Equine influenza virus is a member of the influenza A viruses, which belong to the Orthomyxoviridae family. The Orthomyxoviridae family comprises 5 genera (Fig. 1), all containing enveloped viruses with segmented, single-stranded, negative-sense RNA genomes. In contrast to the rather narrow host ranges of influenza B and C viruses, influenza A viruses can infect a wide variety of species (see Fig. 1). Influenza A viruses possess a host-cell–derived lipid envelope, which contains 3 envelope glycoproteins: the hemagglutinin (HA), neuraminidase (NA), and M2 ion channel protein (Fig. 2). Both the HA and NA are major surface antigens of the virus, and antibodies to these proteins are associated with resistance to infection. Based on antigenic properties of the HA and the NA, influenza A viruses are divided into subtypes. Cross-protection between subtypes (heterotypic immunity) is weak. Eighteen HA subtypes (H1–H18) and 11 NA subtypes (N1–N11) have been found. Equine influenza has been caused by viruses of H7N7 (A/equine/1) and H3N8 (A/equine/2) subtypes.
In addition to the envelope glycoproteins, the influenza A virus genome encodes for an additional 5 structural (M1 matrix protein, nucleoprotein [NP], and polymerase complex [PA, PB1, PB2]) and 3 “nonstructural” proteins (NS1, NS2 [also referred to as the NEP “nuclear export protein”], and PB1-F2) (see Fig. 2). During infection antibodies are also produced to the internal proteins, but these antibodies are not protective. Similarly, the cytotoxic T-cell (CTL) response, primarily
Fig. 1. The 5 Orthomyxoviridae genera. In contrast to the wide host range of influenza A viruses, influenza B and C viruses have a more narrow host range. Whereas influenza A and B contain 8 separate segments of single-stranded RNA, influenza C viruses possess only 7.
Equine Influenza Virus
Fig. 2. The influenza A virus virion. The 3 envelope glycoproteins hemagglutinin (HA), neuraminidase (NA), and the M2 ion channel protein are embedded in a host-cell–derived lipid envelope. In addition, the influenza A virus genome encodes for 5 structural (M1, NP, PA, PB1, PB2) and 3 nonstructural (NS1, NEP, PB1-F2) proteins.
directed against M, NP, and PB2, does not prevent infection but seems to play a role in recovery and virus clearance.2 In contrast to the humoral response that provides only limited heterotypic immunity, the CTL response is cross-reactive between influenza A virus subtypes.3,4 GENETIC DIVERSITY OF INFLUENZA A VIRUS
Both the RNA-based genome and its segmented configuration confer influenza A viruses with the capacity for vast genetic diversity. As viral RNA polymerases lack proofreading functions, RNA viruses generally demonstrate high mutation rates, with resultant potential for rapid evolution. This genetic diversity allows adaptation to a new environment (eg, a new host species) or immune escape. Mechanisms of generating antigenic diversity by influenza A viruses: Antigenic drift: Selection, driven by the host immune system, of viruses with mutations affecting the antigenic sites in the HA (and NA) protein Antigenic shift: Exchange of gene segments (genetic reassortment) between 2 influenza viruses during replication resulting in the switch of the progeny virus’ HA (or NA) subtype
Although there is no evidence to date that equine viruses have participated in genetic reassortment events, it is clear that they undergo some degree of antigenic drift. Despite the rate of genetic diversion of equine influenza viruses being small compared with human influenza viruses,1 the sustained genetic and antigenic evolution has an impact in terms of immunization. It has long been recognized that the antigenic distance between 2 influenza virus strains affects the extent of cross-protection provided by antibodies raised to 1 strain. Therefore, the level of protection provided by
509
510
Landolt
vaccines is often critically dependent on how closely the vaccine strain matches the virus encountered by the horse.5–8 Over the years, several outbreaks have been recorded that affected vaccinated and unvaccinated horses alike,9–12 highlighting the importance of updating vaccine strains periodically. Evolution of Equine Influenza Virus and Vaccine Composition
In the 1980s, genetic evolution of the equine-lineage H3N8 viruses resulted in the formation of 2 distinct evolutionary lineages, Eurasian and American. Continued genetic divergence of the American lineage viruses subsequently resulted in the formation of 3 American-like sublineages (South American, Kentucky, and Florida sublineages).13 Recently, the Florida sublineage evolved into 2 antigenically distinct groups, referred to as Florida sublineage clades 1 and 2. Although clade 1 viruses predominate in America, they have also spread to and caused outbreaks in Europe,14–16 Australia,17 Africa,18 and Asia.19 Clade 2 viruses have been isolated in Europe,20,21 parts of Africa,22 and Asia.23–25 As the H7N7 viruses have not been isolated from horses since the late 1970s,26 representatives of these subtypes are no longer included in contemporary equine influenza vaccines. Similarly, viruses of the Eurasian H3N8 lineage have not been isolated since 2005.11 Thus, the OIE Expert Panel for equine influenza advised that there was no longer a need to include these viruses in equine influenza vaccines. By contrast, with the divergence of the Florida sublineage viruses into 2 antigenically distinct clades, the panel recommended that killed vaccines should include strains representative of both Florida clade 1 and clade 2. As the sustained genetic evolution of the equine H3N8 viruses will likely continue to result in suboptimal vaccine protection in the future, inclusion of viral antigens representative of contemporary circulating viruses has to remain a priority. Cross-Species Transmission of Equine Influenza
It has long been recognized that influenza A viruses exhibit partial restriction of their host range, meaning that viruses from one host species occasionally can transmit to infect another host.1 It is generally accepted that the HA protein has an important role in determining the species specificity of influenza A viruses.27 Serving as the viral receptor-binding protein, the HA mediates the fusion of the virus envelope with the host-cell membrane, with the subsequent release of the virus into the cytoplasm.28 Binding of the HA protein to sialic acid (SA) receptor on the host cell is determined by the SA species (N-acetylneuraminic acid [Neu5Ac] or N-glycolneuraminic acid [Neu5Gc]) and its linkage to galactose residues (a2,6 linkage or a2,3 linkage).27 Human influenza viruses prefer binding to SAa2,6-gal in NeuAc form, matched by SAa2,6-gal expression on the human upper respiratory tract epithelium. Avian, equine, and canine viruses prefer binding to SAa2,3-gal (NeuGc and NeuAc [avian]; NeuGc [equine]), mirrored by a predominance of SAa2,3-gal expression on the epithelial cells of horse trachea and duck intestine.
As already stated, the horse has been viewed as an isolated host for influenza A viruses, meaning that exchange of viruses or their gene segments between horses and other species is limited.1 However, the transmission H3N8 equine influenza viruses to dogs in the United States,29 the United Kingdom,30,31 and Australia32 clearly challenges this notion. In addition to these naturally occurring cross-species
Equine Influenza Virus
transmission events, equine influenza virus has also been found to spread from an experimentally infected horse to a dog that was housed in the same stall,33 and experimental inoculation of dogs with equine influenza virus resulted in nasal virus shedding and subsequent seroconversion.34 In this regard it is interesting that recent studies showed that dogs predominantly express SAa2,3-linked residues throughout the respiratory tract,30,34 which is matched by the preferential binding of both canine and equine influenza viruses to this type of receptor.34,35 On first glance, these findings provide an attractive explanation for the apparent susceptibility of dogs to infection with equine H3N8 viruses. Despite this fact, there are subtle differences in binding specificity among these viruses. Although equine and canine isolates prefer binding to SAa2,3-gal, Yamanaka and colleagues36 found that whereas equine influenza viruses display a clear binding preference for the Neu5Gc receptor moiety, canine H3N8 isolates did not appear to have a preference for either the Neu5Gc or the Neu5Ac sialic acid analogue.36 In light of these similarities in receptor-binding preference, one could expect influenza viruses to spread also in the reverse direction, meaning from dog to horse. Intriguingly recent studies found that 2 distinct isolates of H3N8 canine influenza viruses were unable to infect, replicate, and spread among influenza-naı¨ve equids.36,37 Moreover, inoculation of horses with these canine isolates did not result in clinical disease in either study. These results suggest that factors other than receptor-binding preference likely contribute to species specificity of canine and equine H3N8 influenza viruses. EPIDEMIOLOGY
Although horses of all ages are susceptible to infection, disease incidence is lower in young foals38–41 and is thought to be due to the presence of maternally derived antibodies. Influenza-specific serum antibody concentration is often used as a correlate for protection against infection and disease, and animals with high concentrations of homologous antibody are almost always protected against experimental challenge.42–46 Following experimental inoculation: Horses begin to shed virus in nasal secretions within 24 to 48 hours Nasal virus shedding typically lasts between 6 and 7 days Partially immune animals might shed less virus for shorter durations Morbidity is often high with low mortality
Outbreaks of disease occur most often when susceptible animals are congregated and housed in close contact with each other (eg, horse shows, racetracks, sale barns).47 Anecdotal evidence suggests that disease can spread rapidly through a group of immunologically naı¨ve animals (ie, in a matter of hours to days). In partially immune animals the spread of disease is often considerably slower, and outbreaks may last as long as 3 to 4 weeks.48 In contrast to the seasonal pattern of human influenza epidemics, equine influenza may occur at any time of the year. Spread of virus among susceptible animals may occur through 3 modes: direct contact with infected animals or fomites, droplet transmission (droplets >10 mm and capable of being projected over moderate distances by coughing and sneezing), and aerosol transmission (droplet nuclei
Dr med vet, MS, PhD
KEYWORDS Influenza A virus Virus evolution Epidemiology Diagnosis Control KEY POINTS Despite extensive use of vaccines, equine influenza virus continues to be one of the most important equine viral respiratory pathogens. Control of equine influenza virus is hampered by continued genetic evolution of the virus (antigenic drift). Antigenic drift negatively affects the degree of immunoprotection evoked by inactivated vaccines. Isolation and genetic characterization of currently circulating equine influenza viruses remains a priority, and is essential for vaccine strain selection.
INTRODUCTION
Influenza is a well-known and ancient disease. In fact, a disease resembling influenza was described by Hippocrates in 412 BC. Yet hardly a month goes by without a new headline about influenza in the media. Although millions of dollars have been spent on research, influenza virus continues to challenge our understanding of its ecology and our ability to control its spread. Two key reasons why influenza virus has remained one of the most important causes of viral respiratory disease are its potential for establishing genetic and antigenic diversity and its ability to occasionally transmit between different host species.1 For decades the horse has been viewed as an isolated or “dead-end” host for influenza A viruses, with equine influenza virus being considered as relatively stable genetically. Although equine influenza viruses are genetically more stable than their human-lineage counterparts, they are by no means in evolutionary stasis. Moreover, recent transmission of equine-lineage influenza viruses to dogs also challenges the horse’s status as a dead-end host. This article reviews recent developments in the epidemiology and evolution of equine influenza virus. In addition, the clinical presentation of equine influenza infection, diagnostic techniques, and vaccine recommendations are briefly summarized.
Funding Sources: Merck Animal Health. Conflict of Interest: None. Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, 300 West Drake Road, Fort Collins, CO 80523, USA E-mail address: [email protected] Vet Clin Equine 30 (2014) 507–522 http://dx.doi.org/10.1016/j.cveq.2014.08.003 vetequine.theclinics.com 0749-0739/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
508
Landolt
ETIOLOGY
Equine influenza virus is a member of the influenza A viruses, which belong to the Orthomyxoviridae family. The Orthomyxoviridae family comprises 5 genera (Fig. 1), all containing enveloped viruses with segmented, single-stranded, negative-sense RNA genomes. In contrast to the rather narrow host ranges of influenza B and C viruses, influenza A viruses can infect a wide variety of species (see Fig. 1). Influenza A viruses possess a host-cell–derived lipid envelope, which contains 3 envelope glycoproteins: the hemagglutinin (HA), neuraminidase (NA), and M2 ion channel protein (Fig. 2). Both the HA and NA are major surface antigens of the virus, and antibodies to these proteins are associated with resistance to infection. Based on antigenic properties of the HA and the NA, influenza A viruses are divided into subtypes. Cross-protection between subtypes (heterotypic immunity) is weak. Eighteen HA subtypes (H1–H18) and 11 NA subtypes (N1–N11) have been found. Equine influenza has been caused by viruses of H7N7 (A/equine/1) and H3N8 (A/equine/2) subtypes.
In addition to the envelope glycoproteins, the influenza A virus genome encodes for an additional 5 structural (M1 matrix protein, nucleoprotein [NP], and polymerase complex [PA, PB1, PB2]) and 3 “nonstructural” proteins (NS1, NS2 [also referred to as the NEP “nuclear export protein”], and PB1-F2) (see Fig. 2). During infection antibodies are also produced to the internal proteins, but these antibodies are not protective. Similarly, the cytotoxic T-cell (CTL) response, primarily
Fig. 1. The 5 Orthomyxoviridae genera. In contrast to the wide host range of influenza A viruses, influenza B and C viruses have a more narrow host range. Whereas influenza A and B contain 8 separate segments of single-stranded RNA, influenza C viruses possess only 7.
Equine Influenza Virus
Fig. 2. The influenza A virus virion. The 3 envelope glycoproteins hemagglutinin (HA), neuraminidase (NA), and the M2 ion channel protein are embedded in a host-cell–derived lipid envelope. In addition, the influenza A virus genome encodes for 5 structural (M1, NP, PA, PB1, PB2) and 3 nonstructural (NS1, NEP, PB1-F2) proteins.
directed against M, NP, and PB2, does not prevent infection but seems to play a role in recovery and virus clearance.2 In contrast to the humoral response that provides only limited heterotypic immunity, the CTL response is cross-reactive between influenza A virus subtypes.3,4 GENETIC DIVERSITY OF INFLUENZA A VIRUS
Both the RNA-based genome and its segmented configuration confer influenza A viruses with the capacity for vast genetic diversity. As viral RNA polymerases lack proofreading functions, RNA viruses generally demonstrate high mutation rates, with resultant potential for rapid evolution. This genetic diversity allows adaptation to a new environment (eg, a new host species) or immune escape. Mechanisms of generating antigenic diversity by influenza A viruses: Antigenic drift: Selection, driven by the host immune system, of viruses with mutations affecting the antigenic sites in the HA (and NA) protein Antigenic shift: Exchange of gene segments (genetic reassortment) between 2 influenza viruses during replication resulting in the switch of the progeny virus’ HA (or NA) subtype
Although there is no evidence to date that equine viruses have participated in genetic reassortment events, it is clear that they undergo some degree of antigenic drift. Despite the rate of genetic diversion of equine influenza viruses being small compared with human influenza viruses,1 the sustained genetic and antigenic evolution has an impact in terms of immunization. It has long been recognized that the antigenic distance between 2 influenza virus strains affects the extent of cross-protection provided by antibodies raised to 1 strain. Therefore, the level of protection provided by
509
510
Landolt
vaccines is often critically dependent on how closely the vaccine strain matches the virus encountered by the horse.5–8 Over the years, several outbreaks have been recorded that affected vaccinated and unvaccinated horses alike,9–12 highlighting the importance of updating vaccine strains periodically. Evolution of Equine Influenza Virus and Vaccine Composition
In the 1980s, genetic evolution of the equine-lineage H3N8 viruses resulted in the formation of 2 distinct evolutionary lineages, Eurasian and American. Continued genetic divergence of the American lineage viruses subsequently resulted in the formation of 3 American-like sublineages (South American, Kentucky, and Florida sublineages).13 Recently, the Florida sublineage evolved into 2 antigenically distinct groups, referred to as Florida sublineage clades 1 and 2. Although clade 1 viruses predominate in America, they have also spread to and caused outbreaks in Europe,14–16 Australia,17 Africa,18 and Asia.19 Clade 2 viruses have been isolated in Europe,20,21 parts of Africa,22 and Asia.23–25 As the H7N7 viruses have not been isolated from horses since the late 1970s,26 representatives of these subtypes are no longer included in contemporary equine influenza vaccines. Similarly, viruses of the Eurasian H3N8 lineage have not been isolated since 2005.11 Thus, the OIE Expert Panel for equine influenza advised that there was no longer a need to include these viruses in equine influenza vaccines. By contrast, with the divergence of the Florida sublineage viruses into 2 antigenically distinct clades, the panel recommended that killed vaccines should include strains representative of both Florida clade 1 and clade 2. As the sustained genetic evolution of the equine H3N8 viruses will likely continue to result in suboptimal vaccine protection in the future, inclusion of viral antigens representative of contemporary circulating viruses has to remain a priority. Cross-Species Transmission of Equine Influenza
It has long been recognized that influenza A viruses exhibit partial restriction of their host range, meaning that viruses from one host species occasionally can transmit to infect another host.1 It is generally accepted that the HA protein has an important role in determining the species specificity of influenza A viruses.27 Serving as the viral receptor-binding protein, the HA mediates the fusion of the virus envelope with the host-cell membrane, with the subsequent release of the virus into the cytoplasm.28 Binding of the HA protein to sialic acid (SA) receptor on the host cell is determined by the SA species (N-acetylneuraminic acid [Neu5Ac] or N-glycolneuraminic acid [Neu5Gc]) and its linkage to galactose residues (a2,6 linkage or a2,3 linkage).27 Human influenza viruses prefer binding to SAa2,6-gal in NeuAc form, matched by SAa2,6-gal expression on the human upper respiratory tract epithelium. Avian, equine, and canine viruses prefer binding to SAa2,3-gal (NeuGc and NeuAc [avian]; NeuGc [equine]), mirrored by a predominance of SAa2,3-gal expression on the epithelial cells of horse trachea and duck intestine.
As already stated, the horse has been viewed as an isolated host for influenza A viruses, meaning that exchange of viruses or their gene segments between horses and other species is limited.1 However, the transmission H3N8 equine influenza viruses to dogs in the United States,29 the United Kingdom,30,31 and Australia32 clearly challenges this notion. In addition to these naturally occurring cross-species
Equine Influenza Virus
transmission events, equine influenza virus has also been found to spread from an experimentally infected horse to a dog that was housed in the same stall,33 and experimental inoculation of dogs with equine influenza virus resulted in nasal virus shedding and subsequent seroconversion.34 In this regard it is interesting that recent studies showed that dogs predominantly express SAa2,3-linked residues throughout the respiratory tract,30,34 which is matched by the preferential binding of both canine and equine influenza viruses to this type of receptor.34,35 On first glance, these findings provide an attractive explanation for the apparent susceptibility of dogs to infection with equine H3N8 viruses. Despite this fact, there are subtle differences in binding specificity among these viruses. Although equine and canine isolates prefer binding to SAa2,3-gal, Yamanaka and colleagues36 found that whereas equine influenza viruses display a clear binding preference for the Neu5Gc receptor moiety, canine H3N8 isolates did not appear to have a preference for either the Neu5Gc or the Neu5Ac sialic acid analogue.36 In light of these similarities in receptor-binding preference, one could expect influenza viruses to spread also in the reverse direction, meaning from dog to horse. Intriguingly recent studies found that 2 distinct isolates of H3N8 canine influenza viruses were unable to infect, replicate, and spread among influenza-naı¨ve equids.36,37 Moreover, inoculation of horses with these canine isolates did not result in clinical disease in either study. These results suggest that factors other than receptor-binding preference likely contribute to species specificity of canine and equine H3N8 influenza viruses. EPIDEMIOLOGY
Although horses of all ages are susceptible to infection, disease incidence is lower in young foals38–41 and is thought to be due to the presence of maternally derived antibodies. Influenza-specific serum antibody concentration is often used as a correlate for protection against infection and disease, and animals with high concentrations of homologous antibody are almost always protected against experimental challenge.42–46 Following experimental inoculation: Horses begin to shed virus in nasal secretions within 24 to 48 hours Nasal virus shedding typically lasts between 6 and 7 days Partially immune animals might shed less virus for shorter durations Morbidity is often high with low mortality
Outbreaks of disease occur most often when susceptible animals are congregated and housed in close contact with each other (eg, horse shows, racetracks, sale barns).47 Anecdotal evidence suggests that disease can spread rapidly through a group of immunologically naı¨ve animals (ie, in a matter of hours to days). In partially immune animals the spread of disease is often considerably slower, and outbreaks may last as long as 3 to 4 weeks.48 In contrast to the seasonal pattern of human influenza epidemics, equine influenza may occur at any time of the year. Spread of virus among susceptible animals may occur through 3 modes: direct contact with infected animals or fomites, droplet transmission (droplets >10 mm and capable of being projected over moderate distances by coughing and sneezing), and aerosol transmission (droplet nuclei
