Identification of a New Equid Herpesvirus 1 DNA Polymerase (ORF30) Genotype with the Isolation of a C2254/H752 Strain in French Horses Showing no Major Impact on the Strain Behaviour

Equid herpesvirus 1 is one of the most common viral pathogens in the horse population and is associated with respiratory disease, abortion and still-birth, neonatal death and neurological disease. A single point mutation in the DNA polymerase gene (ORF30: A2254G, N752D) has been widely associated with neuropathogenicity of strains, although this association has not been exclusive. This study describes the fortuitous isolation of a strain carrying a new genotype C2254 (H752) from an outbreak in France that lasted several weeks in 2018 and involved 82 horses, two of which showed neurological signs of disease. The strain was characterised as UL clade 10 using the equid herpesvirus 1 (EHV-1) multi-locus sequence typing (MLST) classification but has not been identified or isolated since 2018. The retrospective screening of EHV-1 strains collected between 2016 and 2018 did not reveal the presence of the C2254 mutation. When cultured in vitro, the C2254 EHV-1 strain induced a typical EHV-1 syncytium and cytopathic effect but no significant difference was observed when compared with A2254 and G2254 EHV-1 strains. An experimental infection was carried out on four Welsh mountain ponies to confirm the infectious nature of the C2254 strain. A rapid onset of marked respiratory disease lasting at least 2 weeks, with significant virus shedding and cell-associated viraemia, was observed. Finally, an in vitro antiviral assay using impedance measurement and viral load quantification was performed with three antiviral molecules (ganciclovir (GCV), aciclovir (ACV) and aphidicolin (APD)) on the newly isolated C2254 strain and two other A/G2254 field strains. The three strains showed similar sensitivity to ganciclovir and aphidicolin but both C2254 and A2254 strains were more sensitive to aciclovir than the G2254 strain, based on viral load measurement.


Introduction
Equid herpesvirus 1 (EHV-1) is an alphaherpesvirus, classified among the Herpesviridae family and the Varicellovirus genus [1]. EHV-1 is one of the most important viral pathogens infecting horses. Some cases have also been reported in other equid and non-equid mammals [2,3].
Transmission of the EHV-1 occurs through direct contact between horses; inhalation of infectious aerosols; direct contact with infected fomites, placenta and aborted foetus; or indirect contact with humans. EHV-1 first infects and replicates in the epithelial cells of the upper respiratory tract [4,5]. At this stage of EHV-1 infection, horses may present a respiratory form of disease associated with nasal discharge, cough and pyrexia. EHV-1 quickly translocates to the draining lymph nodes and associated lymphoid tissues, where it infects leucocytes [6]. Circulation of infected leucocytes during cell-associated viraemia disseminates EHV-1 through the organism to secondary sites such as the central nervous system or the reproductive tract. In these cases, EHV-1 can induce equine herpesvirus encephalomyelitis (EHM; from a mild limb ataxia to quadriplegia) or abortion, stillbirth and neonatal death [7][8][9][10]. Finally, like all herpesviruses, EHV-1 has the ability to establish latency after primo-infection and to reactivate under environmental stress conditions, immune weakness or after corticoid-based treatments [11][12][13][14][15].
EHV-1 genome is 150 kbp and carries 76 open reading frames (ORFs) encoding for proteins involved at different stages of the viral replication cycle. Among them, the 3663-kb-long open reading frame 30 (ORF30) encodes for the DNA polymerase [16]. The mechanisms conditioning one type of disease from another are not well-known, but a correlation between a non-synonymous mutation in position 2254 of ORF30 and the pathogenicity of EHV-1 strains was presented in 2006 by Nugent et al. [17]. Strains carrying a guanine (G) in position 2254 (aspartic acid (D) in position 752 of the protein), were identified as neuropathogenic, whereas those carrying an adenine (A) in the same position (asparagine (N752)) were identified as non-neuropathogenic [17][18][19]. Several studies based on field isolates have since suggested that abortion is largely associated with A 2254 EHV-1 strains, whereas in the case of EHM, the genotype/pathotype association is less obvious [20][21][22]. It has been reported that G 2254 EHV-1 strains replicate more efficiently in the horse and produce significantly higher viral loads [23].
Due to the diseases it causes, EHV-1 has a significant impact on equine health and the equine industry. As a preventive strategy, thorough vaccination against EHV-1 reduces the intensity of the clinical signs, virus shedding and therefore the extent of an outbreak [24]. It is generally accepted that implementation of EHV-1 vaccination in the late 1960s, alongside biosafety measures and herd management, have reduced the abortion storm occurrence [10]. To date, no vaccine exists against EHM among the various vaccines against EHV-1, possibly due to the difficulty of experimentally reproducing this form of disease [20,25,26]. Concerning the therapeutic strategy, no EHV-1 specific treatment has been commercialised. However, some studies have shown the in vitro efficacy of antiviral compounds such as ganciclovir (GCV) and aciclovir (ACV) [27,28]. The ORF30 position 2254 mutation has no effect on these molecules' efficacy in vitro but has been shown to modify EHV-1 s sensitivity to aphidicolin (APD) in vitro. Indeed, APD had a better antiviral effect against A 2254 (N752) EHV-1 strains than against G 2254 (D752) EHV-1 strains [18]. However, in vivo experimental infections showed heterogeneous data after aciclovir treatment [28,29].
This study describes a new ORF30 genetic profile, with the identification of a strain carrying a cytosine (C) in position 2254, resulting in a histidine (H) in amino acid position 752. The strain was isolated at several occasions over 6 months in 2018 from one single and localised EHV-1 outbreak, during which 82 vaccinated horses were affected and showed clinical signs of respiratory diseases, associated with hyperthermia and hind limb oedema [30]. Two horses showed neurological signs of disease and one of them was euthanised. Virus isolation, complete ORF30 sequencing and multi-locus sequence typing (MLST) allowed the identification of a U L clade 10 EHV-1 strain. A new PCR genotyping test was developed to identify specifically the C 2254 strain. Experimental infection on four Welsh mountain ponies confirmed the infectious nature of the newly isolated strain. In vitro antiviral

PCR Assays
Detection/quantitative PCR (PCR 1), a quantitative polymerase chain reaction (qPCR) assay, was performed on purified nucleic acids for EHV-1 glycoprotein B (gB) gene detection and quantification using Diallo et al. (2006) primers and probe. All qPCR details are summarised in Supplementary Table S2 [31]. TaqMan Universal PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) was used for this assay. A standard curve based on a cloned sequence was used for the calculation of the number of copies per sample, as described previously [27,32].
Typing PCR (PCR 2), a second real-time PCR assay, was performed using Allen et al. (2007) primers and probes for the ORF30 A/G 2254 typing [33]. TaqMan Universal Master Mix was also used for this PCR and all details are summarised in Supplementary Table S2. A new PCR (PCR 3) assay was designed with a specific probe for the detection of a cytosine at the ORF30 2254 position (see Results Section 3.2).  Table S2). Amplification PCR (PCR 4) was performed using Phusion Hot Start II DNA Polymerase (2 U/µL) chemistry (Thermo Fisher Scientific). Amplicons were purified using a QIAquick PCR Purification Kit ® (Qiagen). Sequencing PCR (PCR 5) was performed using a BigDye™ Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific). All details are described in Supplementary  Table S2. Sequencing products were purified using a ZR DNA Sequencing Clean-Up ® Kit (Zymo Research, Irvine, CA, USA) and electrophoresis was performed using the 3500 Genetic Analyser ® (Thermo Fisher Scientific). Complete ORF30 sequencing was performed by BIOFIDAL (Vaulx-en-Velin, France), as previously described [22] (Supplementary Table S3 and Figure S3). Sequences were analysed  [35].

Sequencing Assay, MLST, Phylogenetic Analysis and 3D Modeling
The MLST analysis was performed as previously described [22,36]. Sequences were also analysed using BioEdit, and a neighbour joining network was built using Splitstree 4 [37].
Peripheral blood mononucleated cells (PBMCs) were isolated from heparinised blood samples using the Hue et al. method [39] and were co-cultured with a RK13 cell monolayer for up to 6 days. Nasal swabs were sampled in 4 mL, from which 500 µL were inoculated on E. Derm and RK13 cells and cultured for up to 7 days. Co-culture was monitored using IncuCyte ® Live Cell Analysis technology (Essen BioScience, Ann Arbor, MI, USA). The newly isolated EHV-1 strain (FR-56628) was amplified on both E. Derm and RK13 cells. Second passage virus titer was determined on RK13, E. Derm and EEK cells lines using the Karber method [40].
EEK cells were inoculated (1.2 × 10 4 cells per well) in three 96-well plates for xCELLigence acquisition (E-plate VIEW 96 PET, ACEA Biosciences, San Diego, CA, USA), IncuCyte ® acquisition and viral load quantification, respectively. Cells were cultured for 24 h prior to infection with four EHV-1 strains (the newly isolated C 2254 EHV-1 strain, strains FR-6815 (A 2254 ) and FR-38991 (G 2254 ) isolated from the field, and the reference EHV-1 strain KyD (VR700 TM , ATCC ® )) at a multiplicity of infection (MOI) of 0.05. The time corresponding to 50% of the cell index decrease (CIT 50 ) was measured [41].

Neutralisation Assay
A neutralisation assay was performed following the World Organisation for Animal Health (OIE) method. E. Derm cells (1.2 × 10 4 cells per well) and the EHV-1 KyD strain (VR-700 TM , ATCC ® ) were used for this assay [42].

Experimental Infection with EHV-1, Clinical Signs of Diseases, Virus Shedding and Cell-Associated Viraemia
Four 10-month-old male Welsh mountain ponies, seronegative for EHV-1 and EHV-4 (confirmed by seroneutralisaiton (SN) and complement fixing (CF) assays) and with no history of EHV infection (EHV-1 PCR was negative 3 days before infection), were experimentally infected by individual nebulisation (Flexineb ® nebuliser; VLC Europe, Bazoches sur Guyonne, France) [43] with the C 2254 EHV-1 strain (third passage on RK13 cells; 2 mL dose per pony, containing a total of 5 × 10 7 tissue culture infectious dose 50 (TCID 50 )). The day of experimental infection was defined as Day 0. Rectal temperature (RT) was recorded twice a day from Day −5 to Day +21 post-infection (pi). Pyrexia was defined as a rectal temperature exceeding 38.8 • C. Clinical scores were recorded twice a day from Day −5 to Day +21 for cough, attitude, nasal discharge, ocular discharge, mandibular lymph node enlargement, ataxia and other signs of neurological disease. The cumulative clinical score after experimental infection with EHV-1 was calculated using the score for each clinical sign, according to the formula used and reported previously for equine influenza virus infection in Welsh mountain ponies, with some minor modifications [45]: Cumulative clinical score = score RT + score nasal discharge + 2*(score cough) + score lymph node enlargement + score ocular discharge.
Nasopharyngeal swabs were collected on Day −3, from Day +1 to Day +10 and on Days +12, +14, +16, +18, +20 and +21. The swabs were placed directly into 3 mL of EMEM + 1% antibiotics. Virus DNA extraction was performed from nasopharyngeal swab extracts to measure virus shedding using the QIAamp ® Viral RNA Mini Kit (Qiagen) and according to the manufacturer's instructions. Quantitative PCR was performed as described in Section 2.3. Blood samples were collected by jugular venepuncture into EDTA and dry tubes each day from Day 0 to Day +21. Virus DNA extraction from 2 mL of EDTA blood was performed to determine cell-associated viraemia, using the Nucleospin ® Blood L kit (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions. Quantitative PCR (PCR 1) was performed as described in Section 2.3.

In Vitro Antiviral Assay
Antiviral assay was performed using xCELLigence Real-Time Cell Analysis (ACEA, Bioscience) and Incucyte ® technologies. Cells were infected with three EHV-1 strains (the newly isolated C 2254 EHV-1 strain (FR-56628), and two EHV-1 strains FR-6815 (A 2254 ) and FR-38991 (G 2254 ) isolated from the field), as described in Section 2.5, and treated simultaneously with three antiviral drugs in a dose-response manner, as presented in Table 1. All compounds were previously dissolved at 20 mM in dimethyl sulfoxide (DMSO) and stored at −20 • C. One plate was frozen 48 h post-infection (hpi) for viral load quantification. From this plate, nucleic acids were extracted as described in Section 2.2, and viral load was quantified as described in Section 2.3. The two other plates were kept in culture for up to 140 hpi. Infected or non-infected cells treated with 0.5% DMSO were used in each plate as controls. The half maximal effective concentration (EC 50 ) was calculated as previously described [27]. Experiments were carried out at least three times.  [46], 2 Matthews and Boehme 1988 [47], 3 Sheaff et al. 1991 [48].

Statistical Analysis
For the antiviral assay results, statistical analysis was performed on StatGraphics ® Centurion XVI Version 16.1.12 for Windows (StatPoint Technologies, Inc., Warrenton, VA, USA) [49]. A one-way analysis of variance (ANOVA) was used to compare CIT 50 . EC 50 were compared using Tukey's multiple comparison test. Data were considered significant for p values < 0.05.

EHV-1 Outbreak Description
In 2018, 82 horses located in a single site in France and kept in different stable blocks showed clinical signs of disease associated with EHV-1 infection (i.e., cough, nasal discharge, anorexia, pyrexia, depression and limb oedema) over a 6-month period from March 2018 to November 2018. The origin of this specific outbreak is unknown.
Due to the increased number of EHV infection cases reported to RESPE in February and March 2018 (12 outbreaks between 1 January 2018 and 31 March 2018), and rapidly after the first clinical signs appeared in a first stable block (5 March), the 64 horses kept in this stable block were monitored for a set of clinical signs of respiratory disease (e.g., cough, nasal discharge, anorexia, defined as "respiratory signs" in this report); temperature; depression and limb oedema from 5 March 2018 to the 22 April 2018. There are no data available for the first block after this date.
As illustrated in Figure 1A, an increased frequency of clinical signs of disease was observed around 22 days (26 March 2018) after the first clinical signs were reported in this group. Sixty-three horses out of 64 showed clinical signs of disease during the monitoring period. The available data showed the first peak on the 25th day (29 March 2018) after the beginning of the monitoring, and this was followed by a second and major peak on the 33rd day (6 April 2018). Overall, 62 (96.9%) horses displayed respiratory signs with a duration ranging from 1 to 13 days (average: 3.2 ± 2.4 days), 45 (70.3%) horses showed pyrexia (> 38.8 • C) lasting 1 to 7 days (average: 1.8 ± 1.4 days), 38 (59.4%) horses showed limb oedema lasting 1 to 6 days (average: 2.5 ± 1.5 days) and finally 18 (28.1%) horses presented signs of depression lasting 1 to 4 days (average: 1.5 ± 1 days) ( Figure 1B).
The clinical picture was variable from horse to horse, as shown in Figure 1B. Clinical signs appeared alone or in combination, with the most represented combinations being respiratory signs + pyrexia + leg oedema + depression (29.7%); respiratory signs + pyrexia (20.3%); and respiratory signs + pyrexia + leg oedema (14.1%).
Among this population, two horses showed neurological signs of disease (9 April 2018), which required euthanasia for one of them. There is no information available concerning treatments administered during the outbreak or concerning the clinical signs' evolution after 22 April 2018.
During this period, nasal swabs were taken from four horses from the first stable block (Horses 1 to 4) and tested positive for EHV-1 using the PCR 1 assay (5 April 2018). Soon after this first episode, three other horses from a second stable block in the premises (Horses 5 to 7) were sampled and also tested positive for EHV-1 using the same PCR assay (19 April and 26 April 2018). Finally, a new episode was reported from October to November 2018, during which 14 horses from a third stable block in the premises (Horses 8 to 21) were sampled and 11 of them tested positive for EHV-1 using the PCR 1 assay (samples from Horses 12, 15 and 20 were negative; Supplementary Table S4).

New (ORF30) C 2254 Mutation Identification
As described above, EHV-1 was detected in 18 nasal swab samples out of 21, all taken from different horses from the premises described previously, using a detection PCR assay (PCR 1) that amplifies a 63-bp fragment of the EHV-1 gB gene (Supplementary Table S4). As the A/G typing of the ORF30 mutation in position 2254 (PCR 2) was unsuccessful, 145 bp ORF30 fragments from the samples taken from Horses 2, 6 and 7 were sequenced, which revealed a cytosine at position 2254.

New (ORF30) C2254 Mutation Identification
As described above, EHV-1 was detected in 18 nasal swab samples out of 21, all taken from different horses from the premises described previously, using a detection PCR assay (PCR 1) that amplifies a 63-bp fragment of the EHV-1 gB gene (Supplementary Table S4). As the A/G typing of the ORF30 mutation in position 2254 (PCR 2) was unsuccessful, 145 bp ORF30 fragments from the samples taken from Horses 2, 6 and 7 were sequenced, which revealed a cytosine at position 2254.
In order to identify new C2254 strains using a PCR assay (PCR 3), a modified MGB probe was designed based on the existing probe sequences of Allen et al. (2007) ( Table 2). Though A2254 and G2254 probes were used in a multiplex PCR (PCR 2), a C2254 probe was used in a simplex PCR (PCR 3), and amplification curves are shown in Figure 2. The specificity of the C2254 probe was validated with the EHV-1 isolates FR-6815 (A2254) and FR-38991 (G2254) and 160 DNA EHV-1 positive field samples (including the 21 samples collected in this study) obtained from different biological matrices and collected between 2016 and 2018 (Supplementary Table S1). Seventeen of the 18 EHV-1 positive (PCR 1) nasal swabs collected in 2018 from the premises described in Section 3.1 were tested and confirmed to contain C2254 EHV-1 strains only. These samples were the only ones to contain the C2254 subtype. No further samples from these premises were available (in 2019 and 2020). Thirty out of the 160 EHV-1 positive samples could not be typed due to low EHV-1 viral load. Fifty-six samples contained the A2254 subtype and 57 samples contained the G2254 subtype.  In order to identify new C 2254 strains using a PCR assay (PCR 3), a modified MGB probe was designed based on the existing probe sequences of Allen et al. (2007) ( Table 2). Though A 2254 and G 2254 probes were used in a multiplex PCR (PCR 2), a C 2254 probe was used in a simplex PCR (PCR 3), and amplification curves are shown in Figure 2. The specificity of the C 2254 probe was validated with the EHV-1 isolates FR-6815 (A 2254 ) and FR-38991 (G 2254 ) and 160 DNA EHV-1 positive field samples (including the 21 samples collected in this study) obtained from different biological matrices and collected between 2016 and 2018 (Supplementary Table S1). Seventeen of the 18 EHV-1 positive (PCR 1) nasal swabs collected in 2018 from the premises described in Section 3.1 were tested and confirmed to contain C 2254 EHV-1 strains only. These samples were the only ones to contain the C 2254 subtype. No further samples from these premises were available (in 2019 and 2020). Thirty out of the 160 EHV-1 positive samples could not be typed due to low EHV-1 viral load. Fifty-six samples contained the A 2254 subtype and 57 samples contained the G 2254 subtype.

Virus Isolation
Virus isolation from three nasal swabs (Horses 2, 6 and 7) was attempted without success. Consequently, PBMCs were isolated from heparinised blood taken from six other horses (Horses 10 to 15) and co-cultured on RK13 cells. Cytopathic effects (CPEs) characterised by syncytia formation were observed for two samples (from Horses 10 and 13) after 1 and 4 days of incubation, respectively ( Figure 3A). The new C 2254 probe (PCR 3) confirmed the presence of C 2254 EHV-1 in the culture supernatants collected after 5 days (Horse 13) and 6 days (Horse 10) of incubation ( Figure 3B). The strain isolated from Horse 13 was identified as FR-56628 (MT968035). Virus isolation on an E. Derm cell monolayer from a nasal swab collected from Horse 13, at the same time, was also successful. The presence of mucosal neutralising antibodies was investigated in five nasal swabs (when sample quantity allowed it) isolated from Horses 16 to 19 and from Horse 21 in a neutralisation assay using the EHV-1 strain. No neutralisation was measured (data not shown). In vitro growth of the strain FR-56628 (C 2254 ) on EEK cells was compared to three other EHV-1 strains FR-6815 (A 2254 ), FR-38991 (G 2254 ) and reference strain KyD VR700 TM (G 2254 ), showing a similar decrease of the cell index for the four strains ( Figure 3C). The viral load was measured and calculated in Log 10 for each strain 48 h post-infection (hpi) on EEK cells (MOI of 0.05) (Supplementary Table S5). There was no significant variation (p = 0.77) of viral load 48 hpi between the EHV-1 strains FR-56628 (C 2254 /H 752 ), FR-6815 (A 2254 /N 752 ) and FR-38991 (G 2254 /D 752 ) and KyD VR700 TM . The viral load values in Log 10 for each strain are as follows: 8.99 ± 0.46 (FR-56628); 9.21 ± 0.20 (FR-6815); 9.14 ± 0.18 (FR-38991); and 9.04 ± 0.07 (KyD VR700 TM ). However, there was a significant difference in the CIT 50 mean ± standard between the EHV-1 strain FR-6815 and the EHV-1 strains FR-38991 and FR-56628 (p = 0.033). The CIT 50 values for each strain are as follows: 1.11 ± 0.10 (FR-56628); 1.00 ± 0.06 (FR-6815); 1.12 ± 0.13 (FR-38991); and 1.00 ± 0.05 (KyD VR700 TM ).

Phylogeny
The full ORF30 sequence was obtained from the newly isolated EHV-1 strain FR-56628 and compared with ORF30 from the EHV-1 reference strains Ab4 and V592. Three single nucleotide polymorphisms (SNPs) were identified between EHV-1 strains FR-56628 and Ab4 (positions 96, 2254 and 2968) and three SNPs were also identified between EHV-1 strains FR-56628 and V592 (positions 96, 924 and 2254), as shown in Figure 4 (complete ORF30 alignment in Supplementary Figures S5 and S6). The 2254 and 2968 SNPs are non-synonymous (D752H or N752H and E990K, respectively).

Phylogeny
The full ORF30 sequence was obtained from the newly isolated EHV-1 strain FR-56628 and compared with ORF30 from the EHV-1 reference strains Ab4 and V592. Three single nucleotide polymorphisms (SNPs) were identified between EHV-1 strains FR-56628 and Ab4 (positions 96, 2254 and 2968) and three SNPs were also identified between EHV-1 strains FR-56628 and V592 (positions 96, 924 and 2254), as shown in Figure 4 (complete ORF30 alignment in Supplementary Figures S5 and  S6). The 2254 and 2968 SNPs are non-synonymous (D752H or N752H and E990K, respectively). Three dimensional models of the DNA polymerase were obtained based on the homology of the ORF30 amino acid sequences from EHV-1 strains Ab4 (G2254), V592 (A2254) and FR-56628 (C2254) with herpes simplex virus 1 (HSV-1) ( Figure 5). The amino acid residue in position 752 is located in the palm domain and remains close to the fingers domain. After analysis of the sequence by Phyre2 Investigator, the web server predicted that this residue change was unlikely to affect the protein function (data not shown).  Three dimensional models of the DNA polymerase were obtained based on the homology of the ORF30 amino acid sequences from EHV-1 strains Ab4 (G 2254 ), V592 (A 2254 ) and FR-56628 (C 2254 ) with herpes simplex virus 1 (HSV-1) ( Figure 5). The amino acid residue in position 752 is located in the palm domain and remains close to the fingers domain. After analysis of the sequence by Phyre2 Investigator, the web server predicted that this residue change was unlikely to affect the protein function (data not shown).

Phylogeny
The full ORF30 sequence was obtained from the newly isolated EHV-1 strain FR-56628 and compared with ORF30 from the EHV-1 reference strains Ab4 and V592. Three single nucleotide polymorphisms (SNPs) were identified between EHV-1 strains FR-56628 and Ab4 (positions 96, 2254 and 2968) and three SNPs were also identified between EHV-1 strains FR-56628 and V592 (positions 96, 924 and 2254), as shown in Figure 4 (complete ORF30 alignment in Supplementary Figures S5 and  S6). The 2254 and 2968 SNPs are non-synonymous (D752H or N752H and E990K, respectively). Three dimensional models of the DNA polymerase were obtained based on the homology of the ORF30 amino acid sequences from EHV-1 strains Ab4 (G2254), V592 (A2254) and FR-56628 (C2254) with herpes simplex virus 1 (HSV-1) ( Figure 5). The amino acid residue in position 752 is located in the palm domain and remains close to the fingers domain. After analysis of the sequence by Phyre2 Investigator, the web server predicted that this residue change was unlikely to affect the protein function (data not shown).  The complete ORF30 gene sequence was compared with previously published sequences from 14 EHV-1 strains from France and Belgium, 67 EHV-1 strains mostly from the United Kingdom and reference strains Ab4 and V592 ( Figure 6) [22,50]. As previously described, the ORF30 median joining network discriminates these strains in two groups, the second one being a group of three clusters. The EHV-1 strain FR-56628 is located in cluster 2 with one French strain isolated in 2010, three other French strains isolated between 2015 and 2018, a Belgium strain isolated in 2017 and six strains from the UK isolated between 1993 and 2013 (ORF30 maximum likelihood tree shown in Supplementary Figure S7). The EHV-1 strain FR-56628 showed one to four nucleotide differences when compared with the other EHV-1 strains from cluster 2 (including the mutation in position 2254), and from one to five nucleotide differences when compared with all 83 EHV-1 strains of the network (including the mutation in position 2254). The complete ORF30 gene sequence was compared with previously published sequences from 14 EHV-1 strains from France and Belgium, 67 EHV-1 strains mostly from the United Kingdom and reference strains Ab4 and V592 ( Figure 6) [22,50]. As previously described, the ORF30 median joining network discriminates these strains in two groups, the second one being a group of three clusters. The EHV-1 strain FR-56628 is located in cluster 2 with one French strain isolated in 2010, three other French strains isolated between 2015 and 2018, a Belgium strain isolated in 2017 and six strains from the UK isolated between 1993 and 2013 (ORF30 maximum likelihood tree shown in Supplementary Figure S7). The EHV-1 strain FR-56628 showed one to four nucleotide differences when compared with the other EHV-1 strains from cluster 2 (including the mutation in position 2254), and from one to five nucleotide differences when compared with all 83 EHV-1 strains of the network (including the mutation in position 2254). Figure 6. Median joining network representing ORF30 nucleotide sequences of Ab4 and V592 reference strains, 67 EHV-1 strains previously described [50] and 15 EHV-1 strains from France and Belgium [22], including the newly isolated EHV-1 strain FR-56628 (highlighted in yellow). The ORF30 position 2254 nucleotide is annotated. Figure 6. Median joining network representing ORF30 nucleotide sequences of Ab4 and V592 reference strains, 67 EHV-1 strains previously described [50] and 15 EHV-1 strains from France and Belgium [22], including the newly isolated EHV-1 strain FR-56628 (highlighted in yellow). The ORF30 position 2254 nucleotide is annotated.
The EHV-1 MLST, as described by Garvey et al. (2019) [36], was also performed, and the newly isolated strain was classified as U L clade 10 (Supplementary Table S8). The EHV-1 strain FR-56628 MLST amino acid sequence was compared to 79 EHV-1 strains from the UK, Ireland, China, Australia, the United States, Belgium and France and was grouped with the EHV-1 French strains NORM/2/2010, NORM18/2018, the EHV-1 Belgium strain BELG/12/2017, the EHV-1 UK strain SUFF/87/2009 and the EHV-1 Irish strain IRL/348/2002, all from U L clade 10 ( Figure 7). The amino acid change in ORF30 position 752 was the only feature discriminating the EHV-1 strain FR-56628 from the five other U L clade 10 strains shown in this report (Supplementary Table S8 The EHV-1 MLST, as described by Garvey et al. (2019) [36], was also performed, and the newly isolated strain was classified as UL clade 10 (Supplementary Table S8). The EHV-1 strain FR-56628 MLST amino acid sequence was compared to 79 EHV-1 strains from the UK, Ireland, China, Australia, the United States, Belgium and France and was grouped with the EHV-1 French strains NORM/2/2010, NORM18/2018, the EHV-1 Belgium strain BELG/12/2017, the EHV-1 UK strain SUFF/87/2009 and the EHV-1 Irish strain IRL/348/2002, all from UL clade 10 ( Figure 7). The amino acid change in ORF30 position 752 was the only feature discriminating the EHV-1 strain FR-56628 from the five other UL clade 10 strains shown in this report (Supplementary Table S8   Two different ORF37 profiles were identified and confirmed for the EHV-1 strain FR-56628 isolated from Horse 13 (A 795 and G 795 ), which did not induce an amino acid change but which suggests that two variants of the strain were isolated (V1 and V2, Table 3). V2 was the only variant isolated from the Horse 10 sample.

Experimental Infection
Four 10 month-old Welsh ponies were experimentally infected by nebulisation with EHV-1 strain FR-56628 (C 2254 ). Daily clinical scores were evaluated twice a day. As illustrated in Figure 8A and Supplementary Figure S9, all ponies showed a significant increase (p < 0.0001, multifactiorial ANOVA with days as main effect and pony ID as covariates) of the cumulative clinical score from day 1 post-experimental infection, from the second checking point of the day. Overall, the cumulative clinical score started to decrease 10 to 14 days post-infection, depending on the individual, but remained above pre-infection level until the end of the study (i.e., 21 days after experimental infection). Pyrexia (i.e., body temperature >38.8 • C) was measured in all ponies on day 1.5 post-infection, on the second checking point of the day ( Figure 8B). Abnormal body temperatures were mostly recorded in the afternoon and lasted from 4 to 8 days, with peaks of temperature up to 40.6 • C. Nasal discharge was first observed on day 1 and was recorded until the end of the study. Cough was recorded in all ponies on day 2 post-infection and intermittently thereafter. Mandibular lymph node enlargement was observed for one pony on day 2, and all four ponies from day 3 to day 11. All ponies showed signs of lethargy on day 2, and an overall reduction in food consumption was also noted during the first week post-infection. Finally, tail hypotonia was observed for all ponies, from day 2 post-infection and lasted from 4 to 5 days.
Viral load quantification showed that virus shedding started on day 1, with a peak between days 2 and 6, depending on the pony ( Figure 8C). From day 7, nasal shedding started to decrease but EHV-1 remained detectable in nasopharyngeal swabs from all ponies on day 16 post-infection, and was still detected in two out of four ponies on day 20. EHV-1 was detected in the blood sample collected from one pony on day 3, from three out of four ponies on day 5 and in blood samples from all four ponies from day 6 to the end of sampling (day 20). The peak was measured on day 9 ( Figure 8D).

Discussion
Equid herpesvirus 1, which was first designated as equine abortion virus [51] is a major cause of infectious abortion and can also cause sporadic EHM cases, in addition to usually mild respiratory infection. In 2006, Nugent et al. suggested that ORF30 position 2254 could discriminate a neuropathogenic genotype (G 2254 /D 752 ) from a non-neuropathogenic genotype (A 2254 /N 752 ), although the mechanisms involved are not well defined [17,20]. The current study reports the discovery of a third genotype with the isolation of a C 2254 /H 752 strain in one single site in France. The EHV-1 strain affecting these premises was isolated in 2018 and a new PCR assay was designed for the detection of the C 2254 mutation and successfully discriminated A/G and C 2254 strains, in combination with a PCR assay designed by Allen et al. [33]. The C 2254 mutation was not detected in strains from other premises in France between 2016 and 2020. Due to the absence of breeding activity on the affected premises, and the absence of further C 2254 /H 752 EHV-1 outbreaks since 2018, it is not possible to evaluate the effect of the C 2254 /H 752 EHV-1 strain in pregnant mares at the current time. In addition, the site from which the strain was isolated is very restrictive on horse mixing and circulation, which could explain the absence of identification of any other C 2254 strains since the 2018 outbreak. It should also be taken into account that 2018 was a year with a large number of EHV outbreaks and consequently with a large number of samples taken. The number of EHV outbreaks in France and associated sampling is usually much lower, limiting opportunities for virus isolation. Despite anecdotal reports of frequent vaccination in these premises (every 6 months), there was no information available on the immune status of the horses at the time of infection. Vaccine efficacy against this strain would need to be further investigated. Isolation from nasal swabs was not as successful as from co-culture with PBMCs, which could be explained by the possible presence of interferon alpha (INF-α) in infected horses' nasal secretions, as shown in 2010 by Gryspeerdt et al. [52]. When cultured in vitro, the strain induced cytopathic effects (CPEs) associated with polykaryocytes and syncytia formation, as usually observed by microscopy in cell culture after infection with EHV-1 [53]. The new EHV-1 strain, named FR-5628, questions the effect of ORF30 position 2254 mutation on the strain activity.
Based on the amino acid in position 752 (or equivalent) of DNA polymerase gene comparison between several herpesviruses, including the eight known equid herpesviruses (EHV-1, 2, 3, 4, 5, 7, 8 and 9), as already pointed out by Nugent et al. (2006), D 752 (or an equivalent position in other herpesviruses) appears to be a highly conserved amino acid amongst the Herpesviridae family (Supplementary Figure S11) [17]. It has been suggested that D 752 strains are more likely to be ancestral strains from which N 752 would have evolved, as shown in Figure 9, in which the G 2254 strain KENT/64/1994 (see Figure 6) was proposed to have evolved from an A 2254 strain from group 2 cluster 2. All of the sequences compared carried an aspartic acid (D) at this position, with the exception of numerous EHV-1 strains (with V592 as a representative N 752 EHV-1 strain) and EHV-1 FR-56628, the only H 752 strain described to date [30]. To our knowledge, asparagine (N) and histidine (H) are the only two other amino acids identified in equid herpesviruses' DNA polymerase gene in position 752. In addition, a study showed that N 752 strains (isolated from abortion) could infect equally monocytic CD172a+ cells and T lymphocytes, whereas D 752 strains (isolated from neurological cases) could infect CD172a+ cells in a larger proportion than T lymphocytes. In accordance with other studies, the authors suggested that N 752 had adapted its immune evasion strategy after co-evolving from D 752 strains with the host immune system [52,54,55]. Concerning the genotype/pathotype association suggested by Nugent et al. (2006), it was followed by a statistical report by Perkins et al. performed on 176 EHV-1 isolates, showing a strong but not strict association between G 2254 /D 752 strains and neurological disease, and A 2254 strains and abortion disease. According to the authors, the probability of neurological disease being associated with G 2254 /D 752 strains was 162 times greater than the probability of it being associated with A 2254 /N 752 strains [56]. Several studies then showed that this genotype/pathotype association is not exclusive [21,22,[57][58][59]. The discovery of a third genotype C 2254 /H 752 and its potential association with neurological signs makes the understanding of the genotype/pathotype association even more complex. EHV-1 is the only equid herpesvirus known to have possibly evolved from D 752 strains to N 752 and H 752 strains. Along with EHV-8 and EHV-9, these three viruses are also the only equid herpesviruses to our knowledge that have been reported as causing neurological diseases. This observation is in line with other statements suggesting that the amino acid in position 2254 is not the only determinant for neurovirulence and could also depend on the environment (air quality, stress, training management, etc.), the immune status of the horse and its genetic background [9,56,60].

Conclusions
This study describes the isolation of a new 2254 mutant strain (C2254/H752) from a unique site in France in 2018 which was associated with a large number of affected animals, despite frequent vaccination. There was no other identification or isolation of C2254 EHV-1 strains before 2018, nor in 2019 and 2020. The pathogenicity of this strain was confirmed in the context of an in vivo experimental infection that highlighted the rapid onset of disease and significant excretion and cellassociated viraemia. The impact of the mutation on ganciclovir, aciclovir and aphidicolin susceptibility was limited. The molecular characterisation of the strain, the predictive 3D model of the ORF30 and the in vitro culture taken altogether did not show evidence of a major effect of this new C2254 mutation on the strain's behaviour. Most equine herpesviruses have the D752 mutation. EHV-1 is the only equine herpesvirus to carry the N752 or H752 mutations. This is why ORF30 and MLST phylogeny tend to indicate that this C2254/H752 has evolved from N752 strains.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Table S1: ORF30 position 2254 typing using the three different probes (A2254, G2254 and C2254) in order to check the probe C2254 specificity. Table S2: PCR 1, 2, 4 and 5 primers and probe sequences, thermoprofiles and mixtures. Table and Figure S3: complete ORF30 sequencing primers (A) and their distribution on ORF30 (B). Table S4: PCR results, viral load and ORF30 position 2254 subtype in the 14 nasal swabs collected on the premises described in Section 3.1. Figure S5: ORF30 nucleotide alignment of the EHV-1 strains Ab4 (AY665713), FR-56628 (MT968035) and V592 (AY464052). Figure S6: ORF30 amino acid alignment of the EHV-1 strains Ab4 (AY665713), FR-56628 (MT968035) and V592 (AY464052). Figure   An antiviral assay with the newly isolated FR-56628 EHV-1 strain and two other A/G 2254 strains was performed with ganciclovir, aciclovir and aphidicolin. Ganciclovir and aciclovir are two nucleoside analogues (analogues to deoxyguanosine). Indeed, triphosphorylated ganciclovir and aciclovir act as competitive nucleotide substrates for DNA polymerase [46,47]. When phosphorylated, first by the viral thymidine kinase (TK) and then by the cellular TK, ganciclovir and aciclovir are incorporated in the synthesised DNA strand. Replication stops either immediately after treatment (aciclovir), or after the incorporation of several nucleotides (ganciclovir) [61]. Aphidicolin is a tetracyclic diterpen that has been described to inhibit the activity of B-family DNA polymerase (to which EHV-1 DNA polymerase belongs) as a competitive dNTP binding site molecule [62]. Some studies have shown a good efficacy of this molecule against human herpesviruses, especially in strains resistant to nucleoside analogues [63][64][65]. The herpesvirus DNA polymerase crystal structure has only been described for HSV-1. However, as explained in 2007 by Goodman et al., the EHV-1 amino acid D 752 (G 2254 ) residue is homologous to HSV-1 amino acid D751 residue, which is located in the palm domain, a region close to the dNTP binging domain (fingers domain) [18,66,67]. In addition, in their in vivo study, Goodman et al. (2007) suggest that amino acid change from D 752 (G 2254 ) to N 752 (A 2254 ) could alter DNA polymerase activity, after they highlighted thst aphidicolin was less effective against DNA polymerase N 752 (A 2254 ) mutant and N 752 (A 2254 ) strains than against DNA polymerase D 752 (G 2254 ) mutant and D 752 (G 2254 ) strains [18]. After binding with dNTP, the DNA polymerase subdomain rotates and changes conformation to allow dNTP incorporation [66,68]. Aphidicolin may change the fingers and palm domain interaction, as it is associated with Human Pol α by Baranosvki et al. [62]. The effect of an amino acid change in the palm domain on the interaction of aphidicolin (but also possibly of ganciclovir and aciclovir) with the DNA polymerase dNTP binding site was therefore questioned. In this study, aphidicolin and ganciclovir showed a good in vitro effect against the three EHV-1 strains (FR-56628, FR-6815 and FR-38991), whereas a higher concentration of aciclovir was required to inhibit these three strains' replication. As both ganciclovir and aciclovir require the viral TK for their first phosphorylation, it is unlikely that aciclovir's lower efficacy was due to strain TK resistances. The 3D models of the EHV-1 DNA polymerase, based on the HSV-1 DNA polymerase model, did not predict an effect of an amino acid change in position 752 on the protein function. In addition, antiviral assays did not show an evident influence of position 752 amino acid residue on the efficacy of either ganciclovir or aphidicolin in vitro (no significant EC 50 difference was found between FR-56628 and FR-6815, and between FR-56628 and FR-38991). EC 50 , measured by impedance measurements, showed that aphidicolin was significantly more efficient against FR-38991 (G 2254 /D 752 ) than against FR-6815 (A 2254 /N 752 ), but this was not confirmed by the qPCR assay. Even if a good correlation between RTCA and qPCR EC 50 values has been demonstrated, the impedance method depends on parameters that are more variable than the qPCR assay, which can explain the differences [69]. Aciclovir showed differences in efficacy against the three strains which were not reproduced with the two methods. As nucleotide position 2254 is not the only site where non-synonymous mutations occur in ORF30, these could also imply changes of DNA polymerase conformation, different from the strains used by Goodman et al. (2007) and constitute a possible explanation for the aphidicolin effects observed in the two studies.
When compared to 83 other EHV-1 strains, the complete ORF30 nucleotide sequence of the newly isolated EHV-1 strain FR-56628 showed that, in addition to the mutation in position 2254, this part of the genome had from zero to three other mutations, when compared to the other EHV-1 strains in group 2 cluster 2. This cluster is mainly represented by A 2254 strains, with the exception of one G 2254 strain and the FR-56628 C 2254 strain. As all other G 2254 strains are clustered in the ORF30 group 1, the·possibility is not excluded that some EHV-1 A 2254 strains from this cluster (group 2) reversed to EHV-1 G 2254 strains or evolved to a C 2254 strain (Figure 9). Considering the amino acid sequence, the EHV-1 MLST placed the EHV-1 strain FR-56628 among U L clade 10, alongside five EHV-1 strains also classified into ORF30 group 2 cluster 2. The amino acid change in ORF30 position 752 differentiates the strain FR-56628 MLST sequence from the other members of the U L clade 10 used in this study. However, the nucleotide MLST sequence not only showed another silent variation site in ORF37, but also suggested that two different FR-56628 variants have been isolated from Horse 13. ORF37 is a conserved gene among alphaherpesviruses and encodes the HSV-1 UL24 homolog [16]. On the contrary to ORF30, this gene has been classified among the non-essential genes for growth on cultured cells [70,71]. HSV-1 UL24 has been shown to play a role in pathogenicity and to affect in vivo and in vitro replication and reactivation from latency [72,73]. This protein seems to carry numerous functions, such as viral gene expression regulation, dispersion of cellular proteins in the nucleus and glycoproteins distribution in the cytoplasm [74,75]. It was also shown to be required for EHV-1 pathogenicity in a mouse model [70]. Numerous transcription initiation sites and therefore transcripts have been identified for HSV-1 UL24. The impact of the silent mutation detected in this study on different transcripts from different transcription initiation sites is unknown, although all the ATG codons from EHV-1 ORF37 seemed to be in the same reading frame. This silent mutation was not identified in the EHV-1 strain isolated from Horse 10, but no information is available concerning other horses, making it difficult to date the mutation event. Indeed, the strain could either have mutated after infecting Horse 13, or this could also be the result of a double infection with the two variants, after physical contact with different horses.
Experimental infection of four 10-month-old Welsh ponies with EHV-1 strain FR-56628 (C 2254 ) induced characteristic clinical signs of EHV-1 infection in all ponies starting from Day 1 post-infection.

Conclusions
This study describes the isolation of a new 2254 mutant strain (C 2254 /H 752 ) from a unique site in France in 2018 which was associated with a large number of affected animals, despite frequent vaccination. There was no other identification or isolation of C 2254 EHV-1 strains before 2018, nor in 2019 and 2020. The pathogenicity of this strain was confirmed in the context of an in vivo experimental infection that highlighted the rapid onset of disease and significant excretion and cell-associated viraemia. The impact of the mutation on ganciclovir, aciclovir and aphidicolin susceptibility was limited. The molecular characterisation of the strain, the predictive 3D model of the ORF30 and the in vitro culture taken altogether did not show evidence of a major effect of this new C 2254 mutation on the strain's behaviour. Most equine herpesviruses have the D 752 mutation. EHV-1 is the only equine herpesvirus to carry the N 752 or H 752 mutations. This is why ORF30 and MLST phylogeny tend to indicate that this C 2254 /H 752 has evolved from N 752 strains.