The single Asian elephant used in this study was a seven-year-old captive-born female housed at a zoological collection within the United Kingdom. The animal was clinically healthy at the time of sample collection. Blood samples were collected under trained behaviour into ethylenediaminetetraacetic acid (EDTA) blood tubes for the purpose of routine EEHV monitoring and/or diagnostics according to routine veterinary practices. Veterinary interventions such as blood sample collection for the purpose of clinical health screening fall under the Veterinary Surgeons Act 1966 and do not require further ethical approval, as they are intended solely to maintain animal welfare. Samples were sent at ambient temperature overnight to the laboratory and refrigerated at 4-8°C until being processed within 2 hours.

Extraction of gDNA from 140 µL aliquots of whole EDTA blood was performed using the QIAamp Viral RNA Mini Kit (Qiagen, Manchester, UK) according to the manufacturer’s instructions and samples were eluted into a final volume of 60 μL of buffer AVE.

The published sequences from six Asian elephant genomes [NCBI SRA database https://www.ncbi.nlm.nih.gov/sra; accession numbers SRS927124, SRS927123, SRS927126, SRP065915, SRS1158889, ERS365881 and ERS365882; (51, 52)] were used in combination with the online NCBI BLAST tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi) for the assembly of predicted Asian elephant IFNα and IFNβ genes. The published corresponding gene sequences of the horse (Equus caballus; accession numbers NM_001099441, NM_001099440), and predicted corresponding gene sequences of the African elephant (Loxodonta africana; accession numbers XM_003407658.2, XM_003407655.2, XM_023545333.1, XM_003407652.2, XM_003407651.2, XM_003407333.2, XM_023545332.1, XM_010587902.2, XM_003407336.2) were used as respective templates in predicting IFNs of Asian elephants. The 5’ and 3’ untranslated regions (UTRs) of the IFN genes were also included in their assemblies. Primers were designed within the UTRs either side of the 5’ and the 3’ ends of the predicted IFN genes to verify entire coding sequences through polymerase chain reaction (PCR) and Sanger sequencing.

Primers used in this study were synthesised by Eurofins Genomics (Ebersburg, Germany; Table 1 ). PCRs were performed in a Veriti Thermal Cycler (Life Technologies, Paisley, UK) and, as target genes were predicted to be intronless, gDNA was used as a template. Each PCR contained 10 μl of Fast Cycling PCR Master Mix (Qiagen), 2 pmol each of forward and reverse primer, 2 μl of gDNA and nuclease-free water to a final reaction volume of 20 μl. Thermocycling was performed for five minutes at 95°C, followed by 40 cycles of 95°C for 15 seconds, 55°C for 30 seconds and 72°C for one minute.

Aliquots of each PCR product were mixed at a ratio of 5:1 with 6x Blue/Orange Loading Dye (Promega UK Ltd.), separated through electrophoresis in a 2% agarose gel containing SYBR^® Safe DNA Gel stain (Life Technologies) and visualised under ultraviolet light using a Universal Hood II transilluminator (Bio-Rad, Watford, UK). DNA bands of the predicted size were cut from the gel and purified using the MinElute^® Gel Extraction Kit (Qiagen) according to the manufacturer’s instructions.

Purified DNA amplicons were subjected to a single run of Sanger sequencing at the APHA Central Sequencing Unit. The nucleotide (NT) and deduced amino acid (AA) sequences were analysed via NCBI BLAST and the Protein Homology/Analogy Recognition Engine V 2.0 [Phyre²; http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index; (53)] respectively. All NT and AA sequence alignments were performed using ClustalW in MegAlign 15.0 (DNASTAR).

Amino acid sequences of the IFNα (EleIFNα1) and IFNβ (EleIFNβ) genes were analysed for signalling peptides using the SignalP v 4.0 online tool [http://www.cbs.dtu.dk/services/SignalP/index.php; (54)]. The signalling peptide NT sequences and the stop codon were then excluded from each gene sequence. The IFNα and IFNβ sequences were each codon-optimised for expression in Escherichia coli (E.coli) using Eurofins Genomics online software (https://www.eurofinsgenomics.eu/en/home), synthesised, and sub-cloned into Invitrogen™ Champion™ pET303CT-His plasmid vectors (Fischer Scientific, Loughborough, UK).

The pET303CT-His plasmid vectors were transformed into BL21(DE3) Competent Cells (New England Biolabs, Hitchin, UK) according to the manufacturer’s instructions. Up to 100 µL of each transformation reaction was spread over LB agar plates infused with 100 μg/mL ampicillin, and single colonies were picked the following day to inoculate a 10 mL starter culture of selective LB broth (containing 100 μg/mL ampicillin).

Starter cultures were screened after 24 hours for gene inserts using T7 sequencing primers ( Table 1 ) in a PCR as described above but only running for 35 cycles. Amplicons were verified by visualisation of the predicted size DNA band following agarose gel electrophoresis. Aliquots of each starter culture were then added to an expression culture of selective LB broth. Expression cultures were incubated at 37°C and approximately 250 rpm until the optical density, at a wavelength of 600 nm, reached 0.5 to 0.8. Protein expression was then induced with isopropyl β-D-1-thiogalactopyranoside (IPTG) at a final concentration of 400 µM for between three and five hours at 37°C. Bacteria were harvested by centrifugation at 5,000 x g for 15 minutes at 4°C, the supernatant was discarded, and individual cell pellets were weighed before storage at -20°C.

Bacterial cell pellets were thawed at room temperature before 10 mL B-PER™ Complete Bacterial Protein Extraction Reagent and 100 μL 10x Halt™ Protease Inhibitor Cocktail (both Fisher Scientific) were added per gram of cell pellet. Cell pellets were resuspended by vortexing and incubated for 20 minutes at room temperature in a bench-top shaker at 150 rpm.

Solubilisation of bacterial inclusion bodies was performed using a combination of sodium lauroyl sarcosinate (sarkosyl), Triton X-100 and 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) detergents similar to previous reports (55, 56). Sarkosyl (Sigma Aldrich, Gillingham, UK) was added to a final concentration of 2% and incubated at room temperature in a bench-top shaker at 150 rpm for at least two hours. This was followed by centrifugation at 20,000 x g for 20 minutes at 4°C and the supernatant was transferred to a new tube and stored. The process was repeated with the insoluble pellet but sarkosyl was added to a final concentration of 10%. The supernatants were combined, and Triton X-100 and CHAPS (both Sigma Aldrich) were added to final concentrations of 4% and 40 mM, respectively.

Purification of the His-tagged Asian elephant IFN proteins was performed using 1 mL tubes of the Thermo Scientific™ HisPur™ Ni-NTA Purification Kit (Fisher Scientific) according to the manufacturer’s protocol. Purified protein preparations were concentrated using Amicon^® Ultra-15 10K Centrifugal Filter Devices (Millipore Corporation, Watford, UK) at 4,000 x g for 20 minutes.

Residual imidazole was removed from the soluble protein preparations using Zeba™ Spin Desalting Columns (5 mL size) (Fisher Scientific) according to the manufacturer’s instructions. Final protein preparations in sterile PBS were divided into 50 µL aliquots and stored until further use at -80°C.

Dilutions of these protein preparations were made in sterile PBS and quantified using the Thermo Scientific™ Coomassie Plus (Bradford) Assay Kit (Fisher Scientific) according to the manufacturer’s “microplate procedures” protocol. All standards and samples were measured in triplicate. The absorbance of each sample at 595 nm was measured using an Infinite^® M Plex plate reader (Tecan UK Ltd., Reading, UK).

Aliquots of protein preparations were added to 5 µL of 4x Bolt™ Lithium Dodecyl Sulphate (LDS) Sample Buffer, 2 µL of 10x Bolt ™ Reducing Agent and sterile deionised water to a final volume of 20 µL. The samples were heated for 10 minutes at 70°C and loaded onto a Bis-Tris Plus 4-12% gradient polyacrylamide gel in 2-(N-morpholino) ethanesulfonic acid (MES) SDS buffer containing antioxidant (all Fisher Scientific). Invitrogen™ iBright™ Prestained Protein Ladder (11-250 kDa; Fisher Scientific) and BLUeye™ Prestained Protein Ladder (10-245 kDa; Geneflow Ltd., Lichfield, UK) were also loaded onto the gel to monitor protein separation. Alternatively, SeeBlue^® Pre-Stained Protein Standard (3-198 kDa; Fisher Scientific) replaced the BLUeye™ Protein Ladder. Electrophoresis was carried out at 200 volts for 25 minutes.

For Coomassie staining, gels were washed in sterile deionised water and immersed in Coomassie Brilliant Blue R-250 staining solution (Bio-Rad) for one hour. The staining solution was replaced with de-staining solution (40% methanol, 10% glacial acetic acid, 50% distilled water) until background staining had been removed and the bands of the protein ladder and/or sample proteins were clearly visible. The gels were visualised using an Azure c280 imaging system (Cambridge Bioscience, Cambridge, UK).

For Western blots, protein bands were transferred to nitrocellulose membranes using the Invitrogen™ iBlot™ 2 Dry blotting system (Fisher Scientific) according to the manufacturer’s instructions. The membrane was then washed in sterile deionised water and immersed in 5 mL of 1x iBind™ solution (Fisher Scientific). Membrane blocking and antibody binding were performed using the iBind™ Western System (Fisher Scientific) according to the manufacturer’s instructions. A primary monoclonal mouse anti-His tag antibody (Sigma Aldrich; Clone 6AT18) was diluted 1:1,000 and the secondary polyclonal rabbit anti-mouse IgG – horseradish peroxidase antibody (Sigma Aldrich) was diluted 1:16,000. After incubations, the membrane was washed twice in 20 mL of sterile distilled water for two minutes and then immersed in 7.5 mL of Thermo Scientific™ SuperSignal™ West Pico PLUS Chemiluminescent Substrate working solution (Fisher Scientific) at room temperature for five minutes. The membrane was placed in a clear plastic wrap and imaged using an Azure c280 imaging system.

Primary Asian elephant preputial fibroblast (EF) cells were cultured in a 75 cm² cell culture flask containing EF medium consisting of: Gibco™ Minimum essential medium (Fisher Scientific), 10% FBS, Penicillin-Streptomycin (10,000 units penicillin/mL, 10,000 µg streptomycin/mL) and Nystatin (10,000 units/mL; Fisher Scientific). Upon reaching 75 to 80% confluency, cells were passaged or seeded into 24-well plates at 50% confluency per well.

As attempts to culture EEHVs have so far been unsuccessful (57–59), infections of primary Asian elephant fibroblasts were trialled with two other herpesviruses, to ascertain whether either was capable of inducing a measurable cytopathic effect (CPE) in this cell type. Serial tenfold dilutions (10^-1 to 10^-8) were made in Dulbecco’s Modified Eagle Medium (DMEM) of BoHV-1 Oxford strain with a titre of 10^7.5 PfU/mL and Equine alphaherpesvirus 1 (EHV-1) strain V592. These viruses were chosen in part for their availability during this proof-of-concept study, but also based on practical considerations during infection assays, compared to the phylogenetically more closely related human cytomegalovirus (HCMV) or human herpesvirus 6 (HHV-6). A 100 µL aliquot of each virus preparation or 100 µL of DMEM for no-virus-controls was added to each well. Plates were sealed and incubated at 37°C and 5% CO₂ for one hour. Each well was subsequently washed three times with 1.5 mL DMEM medium before 1.5 mL EF medium was applied, the plate was re-sealed, and incubated for up to seven days.

EF cells were fixed with neutral buffered 10% formalin (Sigma Aldrich) for at least 40 minutes. Formalin was then replaced with 250 µL of 2.3% Crystal Violet solution (Sigma Aldrich). Plates were incubated at room temperature for five minutes, washed and air-dried. Assessment of cell monolayers was performed using a Leica DM IL LED microscope (Leica Microsystems UK Ltd., Milton Keynes, UK). Over the course of the study, BoHV-1 infection of Asian elephant fibroblasts (without the addition of IFNs) was investigated in five experimental replicates.

Initial investigations of the bioactivity of human and Asian elephant IFNs were carried out using 1:100 dilutions of unquantified Asian elephant recombinant IFNα1 and β preparations (rEleIFNα1, rEleIFNβ) alongside a control preparation from the insert-less pET303CT-His plasmid vector. Similarly, dilutions were made of recombinant human interferon alpha 2a (hIFNα2a) and human interferon beta 1a (hIFNβ1a; both Immunotools, Friesoythe, Germany) to produce preparations of 1,000 international units (IU)/mL. All preparations were produced in DMEM medium.

In three experimental replicates, EF medium was removed from each well and 1 mL aliquots of unquantified IFN solutions were added 8 hours prior to infection with BoHV-1.

For a subsequent investigation of the time point at which IFNs no longer protect EF cells from BoHV-1 infection, serial 1:10 dilutions were made of quantified rEleIFNα1, rEleIFNβ, hIFNα2a, and recombinant human interferon beta 1b (hIFNβ1b; Miltenyi Biotec., Bisley, UK). All dilutions were produced in DMEM medium and were made with equal masses of rEleIFNs and the corresponding human protein (IFNα: 500 ng/mL - 5 pg/mL; IFNβ: 1.67 µg/mL and 333.33 ng/mL - 30 pg/mL). Based on IU/ng data available for the hIFN preparations, these dilutions equated to 100,000-1 IU/ml. In two experimental replicates, EF cells were infected with BoHV-1 and at 0-, 8-, 12-, 16-, 24-, 48- and 72-hours post-infection (hpi), the EF medium was removed and replaced with 1 mL of the respective IFN solutions. Incubation in IFNs was continued for 24 hours before preparations were replaced once more with 1.5 mL EF medium. Plates were fixed, stained and assessed as described above.

Using predicted sequences from the African elephant and published sequences from the horse as templates, 13 distinct intronless IFNα-like sequences and a single IFNβ-like sequence were retrieved from the Asian elephant sequence assemblies available at the time. PCR-based amplification and sequencing identified four putatively functional Asian elephant IFNα genes, annotated in order of identification (EleIFNα1, IFNα2, IFNα3, IFNα4; GenBank accession numbers OR450024, OR450025, OR450026, OR450027). The EleIFNα1, IFNα2, IFNα3 and IFNα4 genes were each 591 base pairs in length and translated to proteins of 196 AA with predicted molecular weights of approximately 22.5 kDa. Upon alignment, the IFNα were 94.4-97.5% identical at the AA level. The fifth IFNα gene amplified was 600 base pairs in length, with a premature stop codon at position 107 of the AA sequence, demonstrating the identification of an Asian elephant IFNα pseudogene (GenBank accession number OR450028; EleIFNαPG; Figure 1 ). The 564-bp EleIFNβ gene identified encoded a putatively functional protein of 187 AA with predicted molecular weight of 21.6 kDa (GenBank accession number OR450029).

Using NCBI BLAST, the coding regions of the Asian elephant IFNα sequences had the highest homology with that of the African elephant with 91-99% and 86-98% identity at the NT and AA levels, respectively. Of similar immunogenetic relation were the Florida manatee (Trichechus manatus latirostris) sequences with 86-94% NT and 82-90% AA identity, whilst other mammalian species, including humans, were of significantly lower homology (53-78% identity at the AA level). Although the Asian and African elephant IFNβ AA sequences were 98.9% identical, the identity level dropped to 60% for human IFNβ ( Table 2 ).

rEleIFNα1 and rEleIFNβ were successfully expressed, demonstrated by the presence of strong protein bands corresponding to the predicted sizes of 20.1 kDa and 19.2 kDa ( Figure 2 ). Although co-purified contaminants were present, rEleIFNα1 and rEleIFNβ represented the substantial majority of proteins in each preparation, judged by the visible density of the corresponding protein bands in the Coomassie stained gels. Western blot analysis similarly confirmed expression of His-tagged rEleIFNα1 and rEleIFNβ. Whilst only a single protein band was visualised for rEleIFNβ, additional bands were observed in the rEleIFNα1 preparations, but at approximately double and triple the predicted molecular weight of the monomeric protein, possibly representing oligomerization. No chemiluminescent signal was recorded for control preparations.

A surrogate system was required to investigate the antiviral activity of recombinant IFNs as no in-vitro EEHV infection model has yet been established (57–59). To do so, monolayers of EF cells were inoculated with either BoHV-1 or EHV-1 and the cells were fixed and stained 7 dpi to visualise CPE. Monolayers of EF cells were consistently and effectively destroyed by a 10^-3 dilution of BoHV-1, but no CPE was observed in cells inoculated with a 10^-1 dilution of EHV-1 (data not shown). To determine the kinetics of BoHV-1-in EF cells, quadruplicate wells were fixed and stained at 24-hour intervals from 2 to 7 dpi ( Figure 3 ). Viral plaques were consistently visible at 3 dpi and destruction of the EF cell monolayer was evident within five days. Accordingly, a 10^-3 dilution of BoHV-1 was used in all subsequent experiments to determine the antiviral bioactivity of IFNs in this study.

Bioactivity of recombinant human and Asian elephant IFNs was initially demonstrated in an experiment using 1,000 IU/mL of rhIFNα2a or rhIFNβ1a or 1:100 dilutions of unquantified preparations of rEleIFNα1 or rEleIFNβ. EF cells were incubated with the IFNs for 8 hours prior to infection with BoHV-1. rEleIFNα1 and rEleIFNβ both protected cells from BoHV-1 induced CPE, whereas monolayers in control preparations were destroyed ( Figure 4 ). Monolayers also remained intact following pre-exposure to 1,000 IU of hIFNα2a but were destroyed in wells pre-exposed to 1,000 IU of hIFNβ1a. No CPE was evident in uninfected cells. Considering the outcome of this experiment, the hIFNβ supplier was changed.

IFNs require a certain timeframe to exert their antiviral effects, and for rapidly growing viruses, pre-incubation of cells with IFN is essential for anti-viral protection. A further experiment was conducted to ascertain the timepoint at which IFN no longer protects EF cell monolayers and to compare relative bioactivities of recombinant human and Asian elephant IFNs. To do so, serial dilutions of IFNα (500 ng/mL - 5 pg/mL) and IFNβ (1.67 µg/mL and 333.33 ng/mL - 30 pg/mL) were applied at 0, 8, 12, 16, 24, 48 and 72 hpi. Based on data available for the rhIFN preparations, the dilutions tested ranged between 100,000 IU to 1 IU of IFN. Pre-exposure to at least 1,000 IU (5 ng) of hIFNα2a provided EF cells with full protection from BoHV-1 infection when administered up to 48 hpi ( Figure 5 ). In contrast, 500 ng of rEleIFNα1 was required to consistently prevent CPE and protection was achievable only when applied up to 24 hpi. Cell monolayers were protected from BoHV-1 infection with only 33.3 ng of rEleIFNβ but the protective effect was also largely lost if applied beyond 24 hpi. No significant protective effect was observed using hIFNβ1b, indicating a difference in cross-species reactivity between human IFN alpha and beta. No protection from BoHV-1 infection was observed when IFNs were applied beyond 48 hpi.