Evaluation of the Variability of the ORF34, ORF68, and MLST Genes in EHV-1 from South Korea

Equine herpesvirus-1 (EHV-1) is an important pathogen in horses. It affects horses worldwide and causes substantial economic losses. In this study, for the first time, we characterized EHV-1 isolates from South Korea at the molecular level. We then aimed to determine the genetic divergences of these isolates by comparing them to sequences in databases. In total, 338 horse samples were collected, and 12 EHV-1 were isolated. We performed ORF30, ORF33, ORF68, and ORF34 genetic analysis and carried out multi-locus sequence typing (MLST) of 12 isolated EHV-1. All isolated viruses were confirmed as non-neuropathogenic type, showing N752 of ORF30 and highly conserved ORF33 (99.7–100%). Isolates were unclassified using ORF68 analysis because of a 118 bp deletion in nucleotide sequence 701–818. Seven EHV-1 isolates (16Q4, 19R166-1, 19R166-6, 19/10/15-2, 19/10/15-4, 19/10/18-2, 19/10/22-1) belonged to group 1, clade 10, based on ORF34 and MLST analysis. The remaining 5 EHV-1 isolates (15Q25-1, 15D59, 16Q5, 16Q40, 18D99) belonged to group 7, clade 6, based on ORF34 and MLST analysis.


Introduction
Equine herpesvirus-1 (EHV-1), a major pathogen infecting horses, can have devastating effects, causing severe economic burden in the horse industry worldwide [1]. EHV-1 belongs to the subfamily Alphaherpesvirinae and family Herpesviridae. It has a linear, doublestranded DNA genome of approximately 150 kbp, containing 80 open reading frames (ORFs), four of which are duplicated, and consists of long and short unique regions (UL and US, respectively), and the former is flanked by a small inverted repeat (TRL/IRL) and the latter by a large inverted repeat (TRs/IRs) [2]. Several EHV-1 infections are sub-clinical, but the virus can also cause respiratory diseases of varying severity, abortion, neonatal death, or neurological disease, referred to as equine herpesvirus myeloencephalopathy (EHM) [3]. Although the neurological form of the disease is less common than abortion or respiratory disease, EHM appears to have increased in some parts of the world over the past 10-15 years, causing concerns among horse owners and veterinarians because it can result in fatalities [4]. EHV-1 is spread by saliva and nasal discharge as well as aborted fetuses, placentas, or placental fluids [5][6][7][8].
Previous studies have primarily analyzed ORF30 and ORF33 (gB) genes related to pathogenicity [9][10][11][12]. The ORF30 gene, which encodes the DNA polymerase gene, is considered a marker of pathogenicity because its potential to cause neuropathogenicity is significantly higher in EHV-1 strains that carry a single nucleotide polymorphism (SNP) at

EHV-1 Identification and Virus Isolation
EHV-1 was detected in 12 equine samples (3.55%, 12/338). The 134 horses with neurological symptoms did not have EHV-1 detected, and 51 of these samples had several parasites detected, such as strongyle and Halicephalobus gingivalis [21]. There were two aborted fetuses in 2015, three in 2016, one in 2018, and four lung samples and two nasal swabs from cases with respiratory symptoms in 2019. Two strains were isolated from the same farm: 19R66-1 and 19R66-6. EHV-1 positive samples were isolated from RK-13 cells. The 12 samples were confirmed by ORF33 specific real-time PCR [9] ( Table 1). The virus line had a visible CPE (rounding, clustering and lysis) 3 days post-inoculation. EHV-1 viruses were isolated from positive samples using RK-13 cells.

ORF68 Sequence Analysis
Comparison of the V592 sequence with that of Ab4 resulted in the identification of a polymorphic region of ORF68 that was found to be particularly useful for grouping isolates as this locus displays several SNPs within a relatively short region. The nucleotide sequence of the Ab4 strain as a member of EHV-1 group 1 served as a basis for comparing nucleotide changes [10]. PCR products from the 12 EHV-1 isolates were obtained from the ORF68 region of the genome, including the approximately 600 bp-long polymorphic segments, which were sequenced and aligned to distinguish SNPs. The sequences were deposited in GenBank (accession numbers MT940243-MT940254, Table S1), and these SNPs are presented in Table 2.

Phylogeny and Multi-Locus Sequence Analysis
Phylogenetic analysis was performed using an artificial peptide consisting of concatenated amino acids of UL and US based on 31 non-synonymous substitutions between Ab4 and V592 [1]. MLST analysis for the 12 EHV-1 strains is shown in Figures 2 and 3

Discussion
EHV-1 is a World Organization for Animal Health (OIE) listed disease that must be notified to OIE to ensure safe international trade in horses [22]. Considering the potential economic and emotional impact of EHV-1 infections, it would be beneficial to understand the molecular evolution of these viruses, facilitated by tracking the sources of EHV-1 during an outbreak and developing effective control and prevention strategies. Unfortunately, limited genetic data are available in Korea. To date, only ORF30 and ORF33 have been analyzed [19,20,23]; however, these sequences are not present in databases.
In this study, various clinical equine samples were collected in Korea and analyzed for molecular characterization of 12 EHV-1 isolates. ORF30, ORF33, ORF34, and ORF68 of all isolates were sequenced, and the genetic information was deposited in GenBank. Unfortunately, we could not detect EHV-1 in 134 samples from horses with neurological symptoms; however, several reports demonstrated that nucleotide substitution N752 of ORF30 is not the only determinant of neurological disease [24]. EHV-1 N752 genotype viruses are more commonly associated with abortion, responsible for 15~26% of the EHM outbreak [25]. All isolated viruses belonged to the ORF30 N752 genotype, indicating non-neuropathogenic EHV-1 in Korea [19,23,26]. Although the ORF30 D752 genotypes (3/55, 5.5%) were reported in 2014 in Korea [18], this neurological genotype may not be common in Korea. Recently, similar studies were reported in other countries, and the prevalence of the neuropathogenic genotype was extremely low in Japan (2.7%), Australia (3%) and Argentina (7.4%) [23]. In particular, neuropathogenic cases were extremely rare in Korea [16]. Furthermore, this study found an additional SNP at G2968A (E990K) in seven samples by comparing with the reference strain Ab4. Although this SNP was previously reported as not associated with pathogenicity, data are insufficient to predict the impact of the substitutions on protein activity [22].
The homology of ORF33 in this study was 99.7-100% with a few SNPs in two isolated strains (19/10/15-2, 16Q5), while the other 10 EHV-1 isolates showed sequence identity with the ORF33 sequence of Ab4. Our data suggest that this sequence of the ORF33 gene is generally highly conserved and is a good target for diagnostic methods, although some SNPs were observed. A few solitary changes in sequence usually do not affect diagnostic sensitivity, and continuously monitoring these changes is important to avoid false-negative results of EHV-1 diagnosis.
Since 2006, the ORF68 polymorphic region has been used as a putative molecular marker in epidemiological studies. It has commonly been used for 6 groups of EHV-1 isolates in different countries, including Australia, Japan, and Poland [3,27,28]. Group 4 and group 5 genotypes were reported to be predominant in Europe, Africa, and North America, and the group 2 genotype was reported in Japan [15]. Following the original proposal of the 6 groups of ORF68, more SNPs have been described, and new groups have been proposed [1,3,28]. ORF68 sequence analysis of the 12 EHV-1 isolates revealed that they were unclassified according to the groups originally described by Nugent et al. [10] because these isolates showed a 118 bp deletion of the nucleotide sequence 701-818, resulting in a shorter amino acid sequence. Indeed, it was the same deletion found in KyA (MF975655) and Racl11 (MF975656), isolated in the USA, which were unclassified [24]. In the present study, all isolates had an infrequent deletion of 118 (701-818) bp in the ORF68 gene, which was unclassified. The same results were obtained for the unassigned group in KyA, RacL11, and Italian isolates [16,29]. Although ORF68 sequences of the 12 EHV-1 isolates were very similar to KyA and Racl11, these have SNPs at A236C, G689T, and C690T compared to KyA and RacL11 ( Table 2). The 09m142 (MN226987), which was isolated from Italy, also showed the A629 (H210) variation that was present in the 12 EHV-1 isolates. In particular, 7 strains (16Q5, 19R166-1, 19R166-6, 19/10/15-2, 19/10/15-4, 19/10/18-2, and 19/10/22-1) demonstrated 100% homology with 09m142 (MN226987) in nucleotide sequence 236-825 of ORF68. Similar to the results of the studies performed in Hungary and Poland, this suggests that ORF68 is not a suitable global marker [1,3]. However, this type of strain variation has been demonstrated to be a useful adjunct to epidemiological data when investigating disease outbreaks on multiple premises [10,26,30,31]. Furthermore, the presence of the 118 (701-818) bp deletion in EHV-1 strains from other geographical areas and the pathogenic properties of isolates with this deletion should be thoroughly evaluated [16,27].
Studies have confirmed that ORF34 exhibits the highest sequence variability, which should be used to determine whether ORF34 can be a useful marker [17]. Via ORF34 analysis, the 12 groups were identified and named 1 to 12, in which group 1 includes the reference strain Ab4 and group 12 includes isolated strains from zebra, onager, and Thomson's gazelle [28]. Here, a complete analysis of the ORF34 sequences available in GenBank or reported in bibliographies showed that some SNPs were repeated in strain groups [15,17,32]. The 12 EHV-1 isolates in Korea were located in groups 1 and 7. Group 1 (16Q4, 19R166-1, 19R166-6, 19/10/15-2, 19/10/15-4, 19/10/18-2, and 19/10/22-1) showed an ORF34 sequence identical to the sequence of the reference strain Ab4. Group 7 (15Q25-1, 15D59, 16Q5, 16Q40, and 18D99) revealed new SNPs that have not been reported (Table 3). ORF34 studies suggest that the ORF34 protein is required for optimal EHV-1 replication in cultured cells during early infection [33]. The impact of different ORF34 gene mutations on viral replication is unknown. As limited investigations have been carried out thus far, we can speculate that more SNPs will be found in the ORF34 gene, new groups will be described, and the function of the molecular maker for ORF34 will be discovered [16].
Genotyping studies on EHV-1 are extremely limited in Korea; however, our results suggested a possible co-circulation of two types of EHV-1 (groups 1 and 7 for ORF34 analysis, clades 6 and 10 for MLST analysis). EHV-1 in Korea may independently circulate to Asian countries such as Japan, closely related to Europe, the USA, and Oceania. There was, however, no information on the circulation of EHV-1 strains before 2015 and little information on the genetic diversity of EHV-1 in Korea and Asia. Therefore, there is a need for more information on several EHV-1 isolates in Korea through retrospective studies, and more data will have to be obtained via a collection of clinical samples. Additionally, there is a need for more genetic information on EHV-1 strains in Asian countries such as China.
Further studies of genetic diversity of EHV-1 will corroborate some premises as the source of virus will assist in implementation of targeted movement restriction, quarantine and other control measures. The contribution of genetic characterization to our understanding of viral pathogenesis, the development of diagnostics, and the predictions of the likely outcome of disease spread will increase in the future.

Sample Collection
A total of 273 samples were obtained from whole equine bodies from the Animal and Plant Quarantine Agency (APQA) to diagnose horse diseases from 2015 to 2019 throughout all seasons. The samples were collected across all Korean provinces geographically and included mainly clinical signs of EHV-1 such as abortion and respiratory symptoms [36]. Following autopsies of most horse samples, the tissue types submitted for investigation varied but typically included brain, spleen, liver, lung, heart, kidney, placenta, blood samples, nasal swabs, and genital swabs. From the 273 samples collected, it was found that 134 horses had neurological symptoms, 22 had respiratory symptoms, 15 were aborted fetuses, and 102 had other clinical symptoms such as loss of vigor, decreased appetite, gonarthritis, and emaciation. Additionally, for the detection of EHV-1, 65 samples were randomly collected in 2019 (9 nasal swabs with respiratory symptoms and 7 genital swabs with infertility from farms, 2 whole blood samples from farms, and 47 lung tissue samples from an abattoir in Jeju). Data on sex, age, breed, and clinical signs were recorded for blood samples, nasal swabs, and genital swabs, whereas the lung tissue samples represented individual horses with no additional data recorded. Eighty-two horse samples were submitted in 2015, 75 in 2016, 13 in 2017, 5 (2 tissue samples and 3 blood samples) in 2018, and 163 in 2019.

DNA Extraction and EHV-1 Identification
For DNA extraction from swabs and tissue samples, approximately 25 mg of tissue was ground in 2 mL of a serum-free minimum essential medium α (Gibco, UK) solution by using a homogenizer, and prepared by centrifuging the whole sample at 3500 rpm for 10 min [10].
DNA was extracted from homogenized clinical samples using the Intron ® Patho Gene-spin™ DNA/RNA Extraction Kit (iNtRON Biotechnology, Seongnam, Korea) according to the manufacturer's instructions. For detection of EHV-1, ORF33 specific real-time PCR was performed [9]. Two microliters of the DNA extract were applied in the PCRs.
Cells were observed daily and microscopically for the appearance of the virus. They were incubated at 37 • C in 5% (v/v) CO 2 until 70% of cells showed cytopathic effect (CPE) [17,27].