Isolation and characterization of equine influenza virus (H3N8) from an equine influenza outbreak in Malaysia in 2015

Summary Equine influenza is a major cause of respiratory infections in horses and can spread rapidly despite the availability of commercial vaccines. In this study, we carried out molecular characterization of Equine Influenza Virus (EIV) isolated from the Malaysian outbreak in 2015 by sequencing of the HA and NA gene segments using Sanger sequencing. The nucleotide and amino acid sequences of HA and NA were compared with representative Florida clade 1 and clade 2 strains using phylogenetic analysis. The Florida clade 1 viruses identified in this outbreak revealed numerous amino acid substitutions in the HA protein as compared to the current OIE vaccine strain recommendations and representative strains of circulating Florida sub‐lineage clade 1 and clade 2. Differences in HA included amino acids located within antigenic sites which could lead to reduced immune recognition of the outbreak strain and alter the effectiveness of vaccination against the outbreak strain. Detailed surveillance and genetic information sharing could allow genetic drift of equine influenza viruses to be monitored more effectively on a global basis and aid in refinement of vaccine strain selection for EIV.


Equine influenza virus (EIV), an influenza A virus belonging to the
Orthomyxoviridae family, is a major cause of respiratory diseases in horses and can spread rapidly between naïve animals. Influenza A viruses are subtyped according to their surface glycoprotein haemagglutinin (HA) and neuramindase (NA). The HA mediates virus entry into the host cell by binding to the sialic acid receptors and mediating fusion of viral and host membranes (Dimitrov, 2004). On the other hand, the NA is involved in cleavage of sialic acid to aid virus release from the infected cells (Shtyrya, Mochalova, & Bovin, 2009).

Isolation and characterization of equine influenza virus (H3N8) from an equine influenza outbreak in Malaysia in 2015
Eurasian, and since then, the American lineage has further diverged into the Kentucky, South American and Florida sub-lineage clades 1 and 2 (Bryant et al., 2009;Murcia, Wood, & Holmes, 2011). Antigenic and genetic information suggest that evolution of EIV is mainly based on the mutations found on the HA surface glycoprotein. While the NA gene sequences of clade 1 and clade 2 viruses were clearly distinguishable, the majority of published sequences are for HA, making it difficult to study evolution of the NA gene segment.
Between 2006 and 2009, Florida sub-lineage clade 2 viruses were predominantly isolated from the equine influenza outbreaks in Europe while Florida sub-lineage clade 1 viruses circulated in North America (Bryant et al., 2009). In recent years, both clades of EIV have caused massive outbreaks of equine influenza in various geographic locations. For example, clade 1 viruses were responsible for the outbreaks in Japan and Australia in 2007 (Cowled, Ward, Hamilton, & Garner, 2009;Yamanaka, Niwa, Tsujimura, Kondo, & Matsumura, 2008) whereas clade 2 viruses were involved in the Mongolia outbreak in 2008 and the India outbreak in 2009 (Virmani et al., 2010;Yondon et al., 2013). Furthermore, in the 2014 equine influenza outbreaks, clade 1 viruses were detected in the USA (OIE, 2015;Sreenivasan et al., 2018) while clade 2 viruses were detected in France, Germany, Ireland, Sweden and the UK (Fougerolle et al., 2017;Gildea et al., 2018;Rash et al., 2017).
Vaccination plays an important role in the prophylactic treatment against equine influenza. Vaccines offer protection by the induction of antibodies to viral surface glycoproteins, in particular HA (Dimitrov, 2004). Similar to other influenza viruses, EIV may undergo antigenic drift to evade antibody responses. Therefore, the effectiveness of vaccines is dependent on the antigenic differences between the vaccine strain and the outbreak strain (Laver et al., 1979;Park et al., 2009). As such, vaccine strains for equine influenza need to be up to date to offer optimal protection. With this in mind, a comprehensive surveillance programme and active collaborative efforts between the World Organization for Animal Health In August-September 2015, 25 Thoroughbred horses transiting in Singapore from a racing club in Malaysia were found to have influenza-like clinical signs. All 25 horses had been vaccinated against EIV, although full details of vaccination were unavailable. Nasopharyngeal swabs were obtained from the horses and submitted to the Animal Health Laboratory (AHL) of the Agri-Food & Veterinary Authority (AVA) for disease diagnosis. The affected horses were kept in isolation and empirically treated pending outcome of laboratory diagnostics. Since Singapore is free from equine influenza, the import of horses from Malaysia was suspended.
In this study, virus isolated from the Malaysia outbreak in 2015 were genetically characterized by sequencing of the HA and NA gene segments. The Florida clade 1 viruses identified in this outbreak have further diverged as compared to the current OIE vaccine strain recommendations, pointing to a potential vaccine breakdown. Furthermore, sequence differences found between representative strains of two circulating clades of the Florida sub-lineage were also highlighted.
Information from this report could allow genetic drift of equine influenza viruses to be monitored more effectively on a global basis.

| MATERIAL S AND ME THODS
Nasopharyngeal swabs were collected from 25 horses, exhibiting signs of acute respiratory illness. Each nasopharyngeal swab was collected using a swab with synthetic tip (e.g. Dacron) and a plastic shaft in 3 ml of Virus Transport Medium (Puritan UniTranz-RT) and mixed vigorously using a vortex mixer to give the swab suspension before RNA extraction. RNA was isolated from 50 µl swab suspen- Suspensions from nasopharyngeal swabs that were tested positive by the rRT-PCR assay were cultivated in specific pathogenfree (SPF) chicken-embryonated eggs. Briefly, 200 μl of each swab suspension (neat) was inoculated into the allantoic cavity of five 9-11-day-old embryonated chicken eggs and incubated at 37°C.
Three to five days later, eggs were chilled and the allantoic fluid was harvested and tested for EIV using the haemagglutination assay (Killian, 2014). If the allantoic fluid were positive in the haemagglutination assay, the virus was used for further characterization after EIV rRT-PCR confirmation. In order to rule out the presence of EIV in the samples that were negative in the haemagglutination assay, the samples were subjected to a maximum of three passages in the eggs. Subsequently, PCR using EIV-H3-and EIV-H7-specific primers were used to determine the HA subtype (Alves Beuttemmuller et al., 2016). The PCR products of the HA were subjected to Sanger sequencing (AITBiotech) for additional confirmation.
The full-length sequences of H3 and N8 were mapped out using subtype-specific primers elongated at the 5′ end by adding either M13 forward or M13 reverse primer sequences (Table 1) (Rash, Woodward, Bryant, McCauley, & Elton, 2014). Each segment was amplified in a 50-µl PCR reaction consisting of 5 µl RNA using the Qiagen OneStep RT-PCR kit. The one-step RT-PCR cycling protocol comprises of a reverse transcription step at 50°C for 30 min, an initial denaturation at 95°C for 15 min, followed by 35 cycles of denaturation at 94°C for 30 s, primer annealing at 50°C for 30 s, elongation at 72°C for 1 min and a final elongation at 72°C for 10 min.
PCR reactions were analysed on a 1% agarose gel containing ethidium bromide stain according to manufacturer's instructions. PCR products were sequenced using Sanger sequencing.
Nucleotide sequences were aligned to representative sequences for HA1 or NA sequences obtained from GenBank and Global Initiative on Sharing All Influenza Data (GISAID) (Shu & McCauley, 2017)   TA B L E 1 Primer sequences used to sequence the genome of H3N8 EIV  TA B L E 2 Equine influenza viruses included in phylogenetic analysis. We gratefully acknowledge the originating and submitting laboratories of the sequences from GISAID's EpiFlu TM Database on which this research is based

TA B L E 2 (Continued)
F I G U R E 1 Phylogenetic analysis of HA sequences encoded by EIV subtype H3N8. The phylogenetic tree depicts five major clusters of global EIVs as indicated by the bars on the right. The Malaysia isolates (marked by an asterisk [*]) are found in the Florida clade 1 cluster. GenBank and GISAID accession numbers for the sequences are listed in Table 2 F I G U R E 2 HA amino acid differences between the Malaysia isolates and representative   (Woodward, Rash, Medcalf, Bryant, & Elton, 2015). It is unknown how many substitutions can make up a genetic F I G U R E 3 Phylogenetic analysis of NA sequences encoded by EIV subtype H3N8. The phylogenetic tree depicts five major clusters of global EIVs as indicated by the bars on the right. The Malaysia isolates (marked by an asterisk [*]) are found in the Florida clade 1 cluster. GenBank and GISAID accession numbers for the sequences are listed in Table 2 drift in EIV to cause a vaccine breakdown in the field but it is important to emphasize that continued surveillance is necessary to monitor the virus strains which are in circulation.

| D ISCUSS I ON
Laboratory testing plays a critical role in confirming a clinical diagnosis for EIV infection as well as in routine surveillance programmes. The rRT-PCR probe-based assay was used as the primary method of diagnosis during this study and subtyping/clade differentiation was complemented by sequencing methods. Through the use of specific primers in rRT-PCR, the PCR assay allows the detection and identification of EIV from nasopharyngeal swabs in a rapid and highly sensitive manner, allowing quick clinical diagnosis to be made.
Though both conventional RT-PCR and rRT-PCR are increasingly being used for EIV detection due to its high sensitivity and specificity, it remains important to examine the genetic sequence of emerging EIV isolates and evaluate genetic drift in the field.
In conclusion, the outbreak of respiratory disease reported in the Malaysian horses was caused by a Florida clade 1 virus. There was no indication that the virus was particularly virulent as majority of horses recovered clinically in a relatively short period of time.
This raises the importance of mandatory routine surveillance systems for emerging EIV strains through advances in diagnosis and this can serve as an early warning system to facilitate implementation of proper prophylactic and control measures.

ACK N OWLED G EM ENTS
We thank the staff of the racing club for assistance in sample collection; staff at the Animal Health Laboratory Department, Agri-Food and Veterinary Authority of Singapore for their help in sample preparation and laboratory testing. This study is supported by the Agri-Food and Veterinary Authority of Singapore.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.