How Equine Influenza viruses mutate

  How equine influenza viruses mutate   Debra Elton and Adam Rash      Overview      Equine influenza virus (EIV) causes equine influenza in horses, characterised by a raised temperature and harsh dry cough and rapid transmission amongst unprotected horses. It is a major threat to the thoroughbred racing industry as it has the potential to spread so quickly and can cause the cancellation of events and restriction of horse movement. The last major outbreak in Europe occurred in 2003, when over 1000 vaccinated horses in Newmarket became infected. The virus spread throughout the UK and outbreaks were also reported in Ireland and Italy. More recently, more than 50,000 horses were infected during the 2007 outbreak in Australia, large-scale outbreaks occurred in India during 2008 and 2009 and multiple countries were affected by widespread outbreaks in South America in 2012. At the time of writing, another widespread outbreak has been  affecting South America, with reports from Chile, Argentina, Uruguay and Colombia to date. International transport of horses for events and breeding purposes means that equine influenza can spread readily from one country to another. Infected horses can shed the virus before they show any clinical signs of infection and vaccinated animals can be infectious without showing any obvious signs, adding to the risk.     Regular vaccination against equine influenza offers the best protection against infection. Three major vaccine manufacturers make products for the European market, each differing in the virus strains that are included in the vaccine. Sophisticated adjuvants are included in these vaccines, which help boost the horse’s immune response. However, EIV, like other influenza viruses, can mutate to change its surface proteins and can thereby escape from immunity generated by vaccination. It is important that vaccines contain relevant vaccine strains, to give them the best chance of working against current EIVs.     EIV belongs to the influenza A group of viruses, which infect a variety of other animals including humans, birds, pigs and dogs. The natural reservoir for most influenza A viruses is wild aquatic birds, from this pool some viruses go on to infect new hosts and adapt to spread in them. Influenza A viruses are subtyped according to two proteins found on the surface of the virus, haemagglutinin (HA) and neuraminidase (NA). Sixteen HA subtypes and 9 NA subtypes are found in aquatic birds, however only two subtypes are known to have become adapted to horses, H3N8 and H7N7. Equine H7N7 viruses were first isolated in 1956 but have not been isolated since the late 1970s and are now thought to be extinct. Equine H3N8 viruses were first isolated in 1963 when they caused an influenza pandemic in horses and continue to circulate today.      Antigenic drift and shift      The HA and NA proteins on the surface of the influenza virus particle induce antibodies in the host when the virus infects it. For EIV, these antibodies protect the horse against further infection provided the horse encounters similar viruses. A similar process occurs when horses are immunised with a vaccine, most vaccines contain virus proteins that induce the horse’s immune system to make protective antibodies. However, the response to the vaccine is not as good as to virus infection, so horses need to be vaccinated regularly to maintain a protective immune response.     To overcome the horse’s immune response and enable the virus to survive in the equine population, EIV gradually makes changes to its surface proteins. This process is called  antigenic drift . The result is that eventually the horse’s antibodies no longer recognise the virus, which is then able to infect the animal. The two proteins that are important for antigenic drift are HA and NA. HA is involved in virus entry into target cells of the respiratory tract. Antibodies against HA block virus infection, either by preventing the virus from binding to the cell surface, or by preventing a later stage of the infectious cycle that occurs within the infected cell. Antibodies against HA are described as ‘neutralising’ because they prevent virus infection. By changing the HA protein, equine influenza can avoid recognition by these neutralising antibodies. NA is also involved in virus entry, it is thought to help break through the mucus layer that protects the respiratory tract. It also plays a part in virus release, enabling newly formed virus particles to escape from the surface of the cell that made them. Antibodies against NA are thought to block this process, preventing the virus from spreading to new cells. By changing the NA protein, the virus can avoid inhibition by these antibodies and go on to infect new cells.     Equine influenza virus belongs to a family of viruses that have RNA as their genetic material rather than DNA. RNA viruses tend to mutate more rapidly than DNA viruses. The virus has an enzyme called RNA-dependent RNA polymerase that is responsible for making new RNA copies of the virus genetic material for packaging into new virus particles. This is an essential step during the virus life cycle. Compared to the polymerase enzymes found in DNA viruses, the influenza polymerase makes more mistakes when it is copying the virus RNA and this is how changes are made in the genes that code for HA and NA.     As well as undergoing antigenic drift, influenza viruses including equine influenza virus can change their genes by a process called  antigenic shift.  This is a much bigger rapid change, brought about by the virus-swapping sections of its genome with another influenza virus. This process is called  reassortment  and is possible because the virus genome is made up from eight separate segments of RNA, each individually packaged in a set of proteins. If a horse is infected with two different equine influenza viruses at the same time, the eight segments from each virus can be mixed up, generating progeny viruses with new combinations of segments compared to the two parent viruses. This can lead to new combinations of HA and NA that haven’t been seen before, meaning there is no immunity to the new virus. This has happened during the evolution of human influenza viruses and resulted in the influenza pandemics of 1957, 1968 and 2009. In two of these examples, human influenza viruses swapped genes with avian viruses, leading to viruses that replicated well in humans but had a new HA gene from an avian virus. In the 2009 pandemic, a new reassortant virus was generated in pigs then transmitted to humans. Reassortment has also happened with equine influenza viruses. The two different subtypes of equine influenza viruses, H7N7 and H3N8, underwent reassortment resulting in viruses that had most of the internal components of the H3N8 virus but with the HA and NA surface proteins from the H7N7 virus. Eventually these viruses died out and the only equine influenza viruses now in circulation are H3N8. There has been reassortment amongst the different sublineages of equine H3N8 viruses too, for example several of the viruses isolated in the UK during 2009 had a mixture of Florida clade 1 and Florida clade 2 HA and NA. Fortunately these reassortant viruses do not contain a novel HA or NA that has not been seen in horses before, so have not resulted in a major epidemic threat to horses.     In addition to antigenic drift and antigenic shift, the other source of potential new influenza viruses is an animal reservoir, such as birds. We know that horses can be infected by viruses belonging to the H3N8 and H7N7 subtypes and both of these are found in wild aquatic birds. It is thought that the 1963 H3N8 equine pandemic probably arose as a result of cross-species transmission from birds to horses in South America. Such an event happened in China in 1989, when an avian H3N8 infected horses with a much higher mortality rate than is usual for equine influenza. This virus spread amongst horses within China but died out after a relatively short time. It is possible that further avian-equine cross species transmission events could take place, however the virus must then adapt to its new host in order to become established in horses and be able to transmit efficiently from horse to horse. This will require mutations in various virus genes that help the virus attach to and replicate in cells lining the horse’s respiratory tract and spread via droplet infection to other horses.      Evolution of equine influenza      Equine H3N8 virus was first isolated from infected horses in Florida in 1963. This virus spread across continents in the following two years and has been circulating in horses ever since. During this time the virus has evolved into different lineages and sublineages. At first, it was thought that equine influenza did not undergo antigenic drift as it didn't seem to change very much. However, it has since become clear that the virus does change, just more slowly than human influenza viruses. To start with, EIV evolved in a linear fashion, gradually collecting sequential mutations in HA. In the 1980s this single lineage divided into two, the Eurasian lineage and the American lineage, based on the geographical location of the different strains. The Eurasian lineage caused serious outbreaks in Europe, including the 1989 outbreak in the UK during which vaccinated horses succumbed to infection. At this time, the virus strains present in vaccines were older, some contained viruses from 1979 and 1963. Vaccines broke down and failed to protect against infection and laboratory investigations revealed that the virus had undergone antigenic drift. As a result of this, it was realized that a programme of surveillance for equine influenza was essential, in order to monitor the viruses that were in circulation and to have a process to select suitable strains for use in vaccines. The American sublineage has divided into three further sublineages, Kentucky, Argentinian and Florida; only viruses belonging to the Florida sublineage appear to circulate today and these are divided between two groups, Florida clade 1 and Florida clade 2. Florida clade 1 predominates in the Americas, Florida clade 2 in Europe and Asia. Florida clade 1 was responsible for the 2007 outbreak in Australia, and more recently the widespread outbreaks in South America during 2012 and 2018. Florida clade 2 caused an extensive outbreak in vaccinated horses in the UK in 2003 and was also responsible for the outbreaks reported in India during 2008-2009 and Mongolia in 2011. At the moment there are at least three different variants amongst the Florida clade 2 viruses, one of these was first isolated in the UK and has a mutation in HA at position 144, hence viruses belonging to this group are known as ‘144 viruses’. On mainland Europe, viruses with a mutation at position 179 in HA were predominant and in Asia there are viruses with a different mutation at position 144. The 144 viruses react differently to the other viruses in a laboratory assay called a virus neutralisation assay, they are not recognised by equine antibodies raised to an old vaccine strain belonging to the Argentinian sublineage. However, using another antibody-based assay called a haemagglutination-inhibition assay, these viruses are indistinguishable from the others. It is likely that the Florida clade 2 viruses will continue to mutate and diverge from each other so it is important to maintain surveillance to monitor how much they change.      Consequences of antigenic drift: vaccine strains      Once an equine influenza virus has acquired enough mutations in the surface proteins to undergo antigenic drift away from commercial vaccine strains, existing vaccines will no longer offer adequate protection against infection and they will need to be updated. Suitable vaccine strains are recommended by the World Organisation for Animal Health (OIE), based on an international programme of surveillance for equine influenza that is carried out by a network of OIE reference laboratories and several collaborating laboratories. Data is collected by the participating laboratories, then reviewed annually by the OIE Expert Surveillance Panel (ESP). This panel includes representatives from the OIE reference laboratories, the OIE and the World Health Organisation. The conclusions of the ESP are published by the OIE in the second quarterly bulletin each year and the recommendations for vaccine strains are used by vaccine manufacturers worldwide.     The first vaccines licensed for equine influenza contained early virus strains belonging to the pre-divergence lineage. After the outbreak of 1979, which shut down horseracing in the UK, vaccination became mandatory for racing thoroughbreds and vaccines incorporated strains from 1979. These stood until the outbreaks caused by the Eurasian viruses in 1989, which also affected vaccinated horses. In 1995 the first recommendations of the ESP were made, to include a representative of the American (Kentucky) lineage and the Eurasian lineage for H3N8 equine influenza, but also a representative of the H7N7 subtype. These recommendations remained in place until 2004, following the outbreaks in the UK and South Africa in 2003, marking the emergence of the Florida sublineage. Initially the OIE recommendations were updated to include a member of the Florida sublineage such as A/equine/South Africa/4/03 and to keep the Eurasian strain, but the H7N7 strain was no longer needed. However, surveillance data accumulated over the next few years showed that the Florida sublineage had undergone antigenic drift and in 2010 the ESP recommended that representatives of both Florida clade 1 and Florida clade 2 should be included in vaccines. It was also indicated that there was no need for inclusion of a Eurasian strain as these had not been isolated since 2005. The 2010 recommendations still stand today, viruses isolated around the world between 2010 and 2017 belong to either Florida clade 1 or clade 2 and appear to be antigenically closely related to their respective recommended vaccine strains. For Florida clade 1 this remains as A/equine/South Africa/4/03 and for Florida clade 2 this remains as A/equine/Richmond/07. It is important that any strains that are incorporated into vaccines are antigenically representative of these two isolates and the OIE reference laboratories make suitable strains available to vaccine manufacturers so that they can update their products. At the time of writing, in Europe there is only one vaccine available that complies with the OIE recommendations, in the USA there are two manufacturers with products that comply.      Surveillance      Surveillance for equine influenza in the UK has been funded by the Horserace Betting Levy Board for many years. Based at the Animal Health Trust, veterinary practices that have registered with the surveillance scheme receive free testing of nasopharyngeal swabs taken from suspected cases of equine influenza. Veterinary surgeons are also encouraged to submit paired blood samples, with the first sample taken as early as possible and the second samples taken 14-21 days later. Samples are sent to the AHT diagnostic lab and any nasopharyngeal swab samples that test positive for equine influenza using a very sensitive quantitative PCR assay are sent to the influenza research team, where attempts are made to isolate live virus from the sample. These virus isolates play a crucial part in the process of vaccine strain selection. The AHT laboratory sequences the HA and NA genes of flu positive samples, in some cases it is possible to generate data directly from the swab sample without having to grow the virus first. However, to carry out antigenic characterisation of the virus to determine whether it has undergone antigenic drift or not requires a stock of live virus to be grown up. This virus is then tested in the laboratory against a panel of different antibodies to measure how well they recognise the virus. If the antibodies still bind to the virus, this means that the virus has not changed its surface HA to the point where we need to worry. However, if the antibodies no longer recognise the virus, this means that antigenic drift has taken place and it is time for vaccine strains to be updated. To help interpret the data from the antibody binding experiments, the OIE reference laboratories for equine influenza share their data with collaborators based at the University of Cambridge and the Royal Veterinary College, who use a technique called antigenic cartography to map the antigenic distances between new virus strains isolated from horses and existing OIE-recommended vaccine strains. The further the map distance between the new strains and the vaccine strains, the greater the degree of antigenic drift. This technique allows the differences between viruses to be visualized on three-dimensional maps and helps to determine whether recent virus strains have undergone significant antigenic drift or not.     Surveillance for equine influenza and the selection of suitable vaccine strains is dependent upon the submission of samples from infected horses by equine veterinarians to a laboratory that takes part in the process. In turn, this is dependent upon the veterinarian being called out to take those samples and to do so whilst the horse is still shedding live virus. The antigenic work required to characterise the viruses can only be carried out on live virus as it requires large quantities to be grown up. We therefore urge horse owners, trainers and yard staff to call their vet as soon as they suspect a respiratory infection, clinical signs can be very mild in vaccinated animals and are easy to miss.     Research on human influenza viruses identified antigenic sites within the HA protein, regions that varied more than other parts of the protein. These sites are recognised by antibodies and mutations within them can stop them binding to HA. The antigenic sites for equine H3 have not been mapped directly, although some research has shown that some specific changes in these sites seem to have a greater antigenic effect than others.      Cross species transmission and receptor binding      In 2004, an outbreak of severe respiratory disease in racing greyhounds was reported in Florida, USA, caused by equine influenza virus. The virus went on to spread to most states in the USA and adapted to transmit readily between dogs. It became established in the canine population, particularly in large rehoming centres, and was named H3N8 canine influenza virus. It now looks as though it may have died out, replaced by an avian H3N2 virus that was imported with dogs from Korea. The equine H3N8 virus adapted to transmission in dogs by acquiring multiple mutations, however there was a key change in the HA that affected part of the receptor binding site. This is the region of HA that is important for it to bind to the surface of host cells, an early step in the infectious cycle. Changes in the receptor binding site can affect which species a given influenza virus will infect and some of these changes have been studied in depth for human and avian influenza viruses. Humans and birds have different receptors on the surface of their cells, making it difficult for avian viruses to infect humans. For equine influenza viruses, the preferred receptor is more like that found in birds, making it unlikely that human viruses will infect horses or equine flu will infect humans. Some natural isolates of EIV have a change in the receptor binding site that look similar to that found in canine H3N8 and it is possible that the virus that first crossed the species barrier into dogs had this mutation. Further mutations that gave a growth advantage in the new canine host are likely to have been selected as the virus adapted. Experiments carried out with canine influenza showed that it did not readily infect horses, so once the virus had adapted to dogs it lost the ability to infect horses. In the UK, there have been two outbreaks of equine influenza in foxhounds, which can come into close contact with horses during transport. Similarly, there were reports that yard dogs developed antibodies to equine influenza after the outbreak of equine influenza in Australia. In all three cases, the virus did not then become established in dogs. Cross species transmission is a rare event and it is thought that multiple mutations are required before an influenza virus becomes fully adapted to enable it to transmit in a new host.      Acknowledgements   The authors are grateful to the HBLB for funding surveillance work for equine influenza for many years. This work is a collaborative effort on the part of the OIE Reference Laboratories for EIV and collaborating laboratories around the world and would not be possible without horse owners and veterinarians, who contribute by sending samples in from influenza-infected horses.

By Debra Elton and Adam Rash

Overview

Equine influenza virus (EIV) causes equine influenza in horses, characterised by a raised temperature and harsh dry cough and rapid transmission amongst unprotected horses. It is a major threat to the thoroughbred racing industry as it has the potential to spread so quickly and can cause the cancellation of events and restriction of horse movement. The last major outbreak in Europe occurred in 2003, when over 1000 vaccinated horses in Newmarket became infected. The virus spread throughout the UK and outbreaks were also reported in Ireland and Italy. More recently, more than 50,000 horses were infected during the 2007 outbreak in Australia, large-scale outbreaks occurred in India during 2008 and 2009 and multiple countries were affected by widespread outbreaks in South America in 2012. At the time of writing, another widespread outbreak has been affecting South America, with reports from Chile, Argentina, Uruguay and Colombia to date. International transport of horses for events and breeding purposes means that equine influenza can spread readily from one country to another. Infected horses can shed the virus before they show any clinical signs of infection and vaccinated animals can be infectious without showing any obvious signs, adding to the risk.

Regular vaccination against equine influenza offers the best protection against infection. Three major vaccine manufacturers make products for the European market, each differing in the virus strains that are included in the vaccine. Sophisticated adjuvants are included in these vaccines, which help boost the horse’s immune response. However, EIV, like other influenza viruses, can mutate to change its surface proteins and can thereby escape from immunity generated by vaccination. It is important that vaccines contain relevant vaccine strains, to give them the best chance of working against current EIVs.

EIV belongs to the influenza A group of viruses, which infect a variety of other animals including humans, birds, pigs and dogs. The natural reservoir for most influenza A viruses is wild aquatic birds, from this pool some viruses go on to infect new hosts and adapt to spread in them. Influenza A viruses are subtyped according to two proteins found on the surface of the virus, haemagglutinin (HA) and neuraminidase (NA). Sixteen HA subtypes and 9 NA subtypes are found in aquatic birds, however only two subtypes are known to have become adapted to horses, H3N8 and H7N7. Equine H7N7 viruses were first isolated in 1956 but have not been isolated since the late 1970s and are now thought to be extinct. Equine H3N8 viruses were first isolated in 1963 when they caused an influenza pandemic in horses and continue to circulate today.

Antigenic drift and shift

International travel of horses means the virus can spread readily from one country to another.

The HA and NA proteins on the surface of the influenza virus particle induce antibodies in the host when the virus infects it. For EIV, these antibodies protect the horse against further infection provided the horse encounters similar viruses. A similar process occurs when horses are immunised with a vaccine, most vaccines contain virus proteins that induce the horse’s immune system to make protective antibodies. However, the response to the vaccine is not as good as to virus infection, so horses need to be vaccinated regularly to maintain a protective immune response.

To overcome the horse’s immune response and enable the virus to survive in the equine population, EIV gradually makes changes to its surface proteins. This process is called antigenic drift. The result is that eventually the horse’s antibodies no longer recognise the virus, which is then able to infect the animal. The two proteins that are important for antigenic drift are HA and NA. HA is involved in virus entry into target cells of the respiratory tract. Antibodies against HA block virus infection, either by preventing the virus from binding to the cell surface, or by preventing a later stage of the infectious cycle that occurs within the infected cell. Antibodies against HA are described as ‘neutralising’ because they prevent virus infection.

By changing the HA protein, equine influenza can avoid recognition by these neutralising antibodies. NA is also involved in virus entry, it is thought to help break through the mucus layer that protects the respiratory tract. It also plays a part in virus release, enabling newly formed virus particles to escape from the surface of the cell that made them. Antibodies against NA are thought to block this process, preventing the virus from spreading to new cells. By changing the NA protein, the virus can avoid inhibition by these antibodies and go on to infect new cells.

Equine influenza virus belongs to a family of viruses that have RNA as their genetic material rather than DNA. RNA viruses tend to mutate more rapidly than DNA viruses. The virus has an enzyme called RNA-dependent RNA polymerase that is responsible for making new RNA copies of the virus genetic material for packaging into new virus particles. This is an essential step during the virus life cycle. Compared to the polymerase enzymes found in DNA viruses, the influenza polymerase makes more mistakes when it is copying the virus RNA and this is how changes are made in the genes that code for HA and NA.

Figure 1.jpg

As well as undergoing antigenic drift, influenza viruses including equine influenza virus can change their genes by a process called antigenic shift. This is a much bigger rapid change, brought about by the virus-swapping sections of its genome with another influenza virus. This process is called reassortment and is possible because the virus genome is made up from eight separate segments of RNA, each individually packaged in a set of proteins. If a horse is infected with two different equine influenza viruses at the same time, the eight segments from each virus can be mixed up, generating progeny viruses with new combinations of segments compared to the two parent viruses. This can lead to new combinations of HA and NA that haven’t been seen before, meaning there is no immunity to the new virus. This has happened during the evolution of human influenza viruses and resulted in the influenza pandemics of 1957, 1968 and 2009. In two of these examples, human influenza viruses swapped genes with avian viruses, leading to viruses that replicated well in humans but had a new HA gene from an avian virus.

In the 2009 pandemic, a new reassortant virus was generated in pigs then transmitted to humans. Reassortment has also happened with equine influenza viruses. The two different subtypes of equine influenza viruses, H7N7 and H3N8, underwent reassortment resulting in viruses that had most of the internal components of the H3N8 virus but with the HA and NA surface proteins from the H7N7 virus. Eventually these viruses died out and the only equine influenza viruses now in circulation are H3N8. There has been reassortment amongst the different sublineages of equine H3N8 viruses too, for example several of the viruses isolated in the UK during 2009 had a mixture of Florida clade 1 and Florida clade 2 HA and NA. Fortunately these reassortant viruses do not contain a novel HA or NA that has not been seen in horses before, so have not resulted in a major epidemic threat to horses.

In addition to antigenic drift and antigenic shift, the other source of potential new influenza viruses is an animal reservoir, such as birds. We know that horses can be infected by viruses belonging to the H3N8 and H7N7 subtypes and both of these are found in wild aquatic birds. It is thought that the 1963 H3N8 equine pandemic probably arose as a result of cross-species transmission from birds to horses in South America. Such an event happened in China in 1989, when an avian H3N8 infected horses with a much higher mortality rate than is usual for equine influenza. This virus spread amongst horses within China but died out after a relatively short time. It is possible that further avian-equine cross species transmission events could take place, however the virus must then adapt to its new host in order to become established in horses and be able to transmit efficiently from horse to horse. This will require mutations in various virus genes that help the virus attach to and replicate in cells lining the horse’s respiratory tract and spread via droplet infection to other horses.

TO READ MORE --

BUY THIS ISSUE IN PRINT OR DOWNLOAD -

October - December 2018, issue 63 (PRINT)
6.95
Quantity:
Purchase in print

WHY NOT SUBSCRIBE?

DON'T MISS OUT AND SUBSCRIBE TO RECEIVE THE NEXT FOUR ISSUES!

Print & Online subscription
24.95 every 12 months
Add to Cart

Travelling Stateside - The trainer's perspective

Combating contamination in the racing yard

0