Viral cloning and mutagenesis.
EHV-1 strain Ab4 (GB80_1_2 isolated from a quadriplegic mare [27
]) was cloned as a BAC by cotransfection of a mini-F plasmid with wild-type genomic DNA in rabbit kidney RK13 cells, as described previously [28
]. Briefly, the mini-F bacterial origin of replication sequence pHA2 containing a chloramphenicol resistance gene (cat
) and the egfp
gene were cloned into a transfer plasmid and introduced in lieu of nonessential gene 71 (encoding glycoprotein gp2 of EHV-1) by homologous recombination in RK13 cells [29
]. The BAC was maintained in E. coli
EL250 cells, which harbor the recombination system of phage λ under the control of a temperature-sensitive repressor [30
]. The BAC was mutated to the opposite Pol genotype (N752) by converting nucleotide number 2254 of pol
) from a guanine to an adenine (see Table S1
for primers) using two-step Red-mediated en passant recombination [13
]. Following confirmation of the respective genotypes by nucleotide sequencing (see Table S1
for primers), repaired virus was produced by cotransfection of a plasmid encoding EHV-1 Ab4 gene 71 with DNA of the respective mutant BAC. Resultant GFP-negative viruses, lacking the BAC cassette, were fully restored in gp2 expression, and exhibited restriction enzyme patterns identical to those of the parental Ab4 virus ().
High-titer stocks of each virus were produced by passaging the transfection product once on equine NBL-6 cells in Eagle's minimal essential medium (EMEM) supplemented with 20% fetal bovine serum (FBS). Infected cells were frozen/thawed twice (−80 °C/37 °C), centrifuged for 5 min at 4,500g
, supernatants were collected and stored at −80 °C. Titration of the virus suspension was performed on RK13 cells [29
DNA polymerase assays.
The pTM1-ORF30 plasmid, which expresses Pol (D752 variant) of EHV-1 Ab4 under a T7 promoter, was previously described [20
]. The pTM1-ORF30 N752 plasmid harboring the pol N752
mutant was created using the Quick-Change Mutagenesis kit (Stratagene, La Jolla, CA), amplifying the pTM1-ORF30 plasmid with primers listed in Table S1
pTM1-ORF30_D752N_F (forward) and pTM1-ORF30_D752N_R (reverse). The mutated pol N752
gene was completely sequenced and shown to contain only wild-type sequences except for the engineered point mutation.
In vitro transcription-translation of the pol D752
and pol N752
genes was performed from plasmid pTM1-ORF30 or pTM1-ORF30 N752 using the TNT T7 coupled reticulocyte lysate system from Promega (Madison, WI) according to the manufacturer's guidelines. To examine protein expression levels, the translation products were labeled with [35
S]methionine (Perkin Elmer, Waltham, MA) and analyzed by sodium dodecyl sulfate-7.5% polyacrylamide gel electrophoresis (SDS-PAGE) and autoradiography. Purified baculovirus-expressed pORF18, the accessory subunit of EHV-1 DNA polymerase, was prepared as described previously [20
Basal DNA polymerase activity of Pol D752 and of Pol N752 and stimulation of their activity by pORF18 were assayed by measuring the incorporation of [3
H]dTTP (Amersham Bioscience-GE Healthcare, Milan, Italy) into a poly(dA)-oligo(dT) template (Amersham Bioscience-GE Healthcare) as previously reported [20
], using 12 μl of in vitro transcribed-translated Pol D752 or Pol N752 in the absence or in the presence of 600 fmol of purified baculovirus-expressed pORF18 in a 60-μl reaction volume.
The effect of aphidicolin on Pol activity was tested in similar assays, with 4 μl of in vitro-transcribed and -translated Pol D752 or Pol N752 plus 200 fmol of pORF18 in the presence or absence of various amounts of drug in a 20-μl reaction volume. Aphidicolin (Sigma-Aldrich, St. Louis, MO) was dissolved at a 300 μM concentration in 10% DMSO. In these assays, the final concentration of compound-derived DMSO was maintained at 0.5% (vol/vol) in all samples. In control samples with no drug added, a corresponding volume of pure DMSO was added to reach a final concentration of 0.5%.
Viral drug sensitivity assays.
Aphidicolin (Calbiochem, San Diego, CA) was dissolved as a 1 mg/ml (2.95 mM) stock in DMSO, then serially diluted in EMEM supplemented with 0.5% FBS. For virus yield titration and qPCR assays, RK13 cells were plated at 2 × 105 cells/ml in 24- and 96-well plates (BD Falcon, San Jose, CA), allowed to adhere overnight, and infected with N752 mutant or D752 revertant virus at a multiplicity of infection (MOI) of 0.01. After virus adsorption for 2 h at 37 °C, cells were washed twice with PBS and incubated with 1 ml (for virus yield titration assays) or 200 μl (for qPCR assays) of fresh media containing aphidicolin at various concentrations. Each final drug concentration was tested in three independent wells. The final concentration of the DMSO vehicle was less than 0.04 % (vol/vol). Plates were incubated for 2 d at 37 °C and then subjected to two freeze-thaw cycles (−80 °C and 37 °C). Virus yield titers were determined by transferring 100 μl aliquots from each of the wells to a fresh 24-well monolayer culture of RK13 cells followed by 1:10 serial dilution across the plate. Cultures were incubated for 3 d, then cells were stained, and the numbers of plaques were counted. Viral DNA genome copies were quantitated by qPCR as described below, using the cellular 18S rRNA gene for normalization.
In vitro PBMC infection assay.
PBMC were isolated by density gradient centrifugation over Histopaque 1077 (Sigma-Aldrich) from buffy coats of 30 ml of heparinized blood collected from healthy horses (approximate ages 10–29 y) with no history of ongoing or previous EHV-1 infection. Four experiments were conducted, each using blood from a different donor horse. Cells were infected immediately after isolation at an MOI of 1 using the GFP-expressing, reconstituted BAC clones of the Pol mutant (N752) and revertant (D752) viruses, and incubated for 48 h in conical polypropylene tubes (BD Falcon). Cell populations were characterized with a Becton Dickinson FACScalibur after immunofluorescent staining using mouse monoclonal antibodies recognizing either equine CD4 (clone CVS4, Abd-Serotec, Raleigh, NC), CD8 (clone 73/6.9.1, VMRD, Pullman, WA), or CD14 (clone 105, kindly provided by Dr. Bettina Wagner, Cornell University, Ithaca, NY), and a secondary Cy5-labeled anti-mouse antibody (Jackson ImmunoResearch, West Grove, PA), or a Cy5-conjugated goat-anti-horse B cell antiserum (kindly provided by Dr. Bettina Wagner).
The mouse and horse experiments were performed with approval of the Cornell University Institutional Animal Care and Use Committee. The pony experiment performed in Newmarket (UK) was approved by the UK Home Office. All treatment administration, sample and data collection, clinical examinations, and statistical analyses for these studies were performed with the identity of the treatment concealed from the investigators.
BALB/c mice (n = 16 per group) were anesthetized with intraperitoneal xylazine/ketamine and inoculated intranasally with 1 × 105 PFU of virus suspended in 25 μl of EMEM, or 25 μl of uninfected NBL-6 cell supernatant for the placebo group. Body weights were recorded daily and mean daily percentage body weight loss was compared between groups. Approximately 100 μl of blood was collected from ten mice per group at days 2, 6, and 9 postinfection by submandibular venipuncture, diluted to 200 μl with sterile 1× PBS containing 100 mM EDTA, and frozen at −80 °C. On days 3 and 5, three mice from each group were humanely killed by CO2 inhalation. Lungs, spleen, brain, and the upper 3–4 cervical vertebrae were collected using sterile procedures. Half of each organ was fixed in formalin for histopathology and immunohistochemistry; the other half was immediately frozen. DNA was purified in parallel with frozen blood samples using the DNeasy 96 Blood/Tissue kit (Qiagen, Valencia, CA).
Two independent equine infection experiments were performed. Both experiments tested the same BAC-derived mutant (N752) and revertant (D752) viruses. All animals enrolled had paired serum virus neutralizing antibody titers <1:32 (horses), or paired complement fixing antibody titers <1:10 (ponies) and were randomized to treatment groups. Starting 2 d prior to infection and for the duration of the experiment, animals were monitored daily for rectal temperature, nasal discharge, coughing, SMLN swelling, and neurologic signs. Cumulative clinical scores were calculated as follows: 1 point for serous nasal discharge, or 2 points for mucopurulent or heavy, discolored discharge; 1 point for infrequent coughing, or 2 points for frequent coughing; 1 point for SMLN swelling; 1 point for respiratory rate > 30, or 2 points for respiratory rate > 50.
Blood was collected daily in heparinized Vacutainers (Becton-Dickinson, San Jose, CA). Each buffy coat from two heparinized blood tubes was layered onto a gradient of Histopaque 1077 and 1119 for isolation of viable PBMCs as recommended by the supplier (Sigma procedure 1119). Serum from coagulated blood drawn at weekly intervals was assayed for viral neutralization by a veterinary diagnostic laboratory. Nasal swabs were also collected daily, by simultaneously inserting two 15-cm polyester-tipped swabs (Fisher Scientific, Pittsburgh, PA) into the ventral meatus of one nostril. The samples were immediately placed in 2 ml of viral transport media: 10% neonatal calf serum in phosphate-buffered saline, containing 3× antibiotic-antimycotic (Gemini BioProducts, Woodland, CA): 300 U/ml penicillin, 300 μg/ml streptomycin, 0.75 μg/ml fungizone; and 68 μg/ml enrofloxacin (Bayer Animal Health, Shawnee Mission, KS). Swabs were then incubated on ice for 2–4 h. For viral isolation, 200 μl of the nasal swab solution was diluted in EMEM (supplemented with 10% FCS and 3× antibiotic-antimycotic) and titrated on confluent monolayers of RK13 cells. After 2 h, the medium was replaced with 0.8% methylcellulose dissolved in the same growth medium. Plates were read 3 d postinoculation after acetone fixation and crystal violet staining. Results were recorded as titers of PFU per milliliter of inoculate. Another 200 μl of nasal swab solution was frozen at −80 °C and later thawed for DNA extraction as described below.
In the first equine experiment, two groups of four Welsh Mountain ponies, all 2-y-old females, were infected while housed in a BSL-3 facility in Newmarket, UK. Each group was in a separate room under negative air pressure with its own air supply. Each pony was given 1 × 107
(approximately 7 × 106
PFU) of aerosolized virus, as described previously [31
]. A second equine challenge experiment was performed at Cornell University, Ithaca, NY. Two groups of seven adult, mixed-breed horses, aged 3–16 y, were infected with 1 × 107
PFU of aerosolized virus while housed in individual BSL-2 rooms as described previously [7
]. Ages were normally distributed within each group, with a mean of 8 y in the D752 group and 10 y in the N752 group. Each group had five females and two castrated males. Horses were kept inside the rooms until virus could no longer be cultured from nasal swabs, then were led outside daily for neurologic exams. Under sedation with detomidine (Pfizer Animal Health, Exton, PA) at 0.01 mg/kg, CSF was collected aseptically as deemed necessary using a 5-in 18 gauge spinal needle in the lumbosacral space. CSF was also collected postmortem in the atlanto-occipital joint space from humanely killed horses. At necropsy, tissues from all major organs including brain and spinal cord were collected. Sterile tissue samples for qPCR were taken, immediately frozen at −80 °C, and processed as described below. CNS tissues were fixed in neutral buffered formalin as a whole and gross examination was completed after the tissues were fully permeated by the preservative. After gross examination was completed, representative sections of CNS tissue were processed for histopathology.
Real-time quantitative PCR.
Aliquots of 5 × 106
PBMC, 200 μl of nasal swab, and postmortem tissues were processed with the QIAamp96 DNA blood/tissue kit (Qiagen), with a final DNA elution volume of 200 μl. qPCR was performed using the 7500-FAST real-time PCR system (Applied Biosystems, Foster City, CA) with reaction mixtures containing TaqMan Fast Universal PCR Master Mix, 900 nM primers, 250 nM probe, and 5 μl of DNA sample, in a 20 μl total volume. The thermal cycling program was: 20 s at 95 °C, followed by 40 cycles of 95 °C for 3 s and 60 °C for 30 s. Each sample was run in triplicate. All normalized viral genome copy numbers for animal samples were calculated by the relative standard curve method, with the BAC clone of Ab4 used as a viral standard, a BAC clone of equine chromosome ECA1 for horse genome copy quantification (kindly provided by Drs. R. Tallmadge and D. Antczak [32
]), and a partial clone of the murine iNOS
(inducible nitric oxide synthase) gene for mouse genome quantification. A Taqman assay for the EHV-1 IR6
gene was used to quantify viral genome copies [7
]. Taqman genomic normalization assays were designed for the mouse iNOS
gene or the horse β2-microglobulin gene, β2-M
). Normalization of viral DNA yield from nasal swabs was performed by addition of 5 × 103
copies of Marek's disease virus (MDV) BAC clone pRB-1B [33
] to the lysis buffer during DNA purification, and detecting copies of MDV gD [34
]. Purification efficiency of nasal swabs was controlled by dividing each copy number result by the mean number of copies in all samples. Relative quantification of EHV-1 genome copies in the drug sensitivity experiment was performed using the ΔΔCT
], with the 18S rRNA
gene as the endogenous control, and the untreated, infected well as the calibrator.
Statistical analyses were performed using SAS v 9.1 (SAS Institute, Cary, NC). The significance level for all experiments was set at α = 0.05. Animal experiment data were fitted to linear models using PROC MIXED including appropriate two-way interactions, after verifying that residuals were normally distributed. Repeated ANOVA tests were performed for all experiments involving repeated sampling or repeated measurements. Significance of results was assessed at the pair-wise level using Bonferroni t-tests with α = 0.05. Data that did not pass a Shapiro-Wilks test for normal distribution of data are represented in figures in the form of medians or individual points.
Summary values were analyzed by Kruskal-Wallis testing. These values included: neurologic status (assessed as the highest score per horse), duration of fever (number of consecutive days of rectal temperature above 38.5 °C as measured each morning), and duration of infectious nasal shedding (number of days from the first to the last day of positive viral culture). Kaplan-Meier scores and p values for nasal shedding duration were calculated using PROC LIFETEST.
In vitro FACS raw data analysis was performed with FlowJo v. 7.2 (Tree Star, Ashland, OR), using the histogram function to measure the percent of infected cells expressing each cellular marker. Differences between virus treatments for PBMCs from different horses were tested for normality using the Shapiro-Wilks test, then compared to the null hypothesis of being equal to zero with a t-test. Differences between values in the in vitro aphidicolin sensitivity experiments were tested with a one-way ANOVA. IC50 values for the drug sensitivity assay were calculated based on linear regression model fit equations.
Structure predictions were made based on the HSV-1 Pol crystal structure of Liu et al. as a template ([23
], Protein Data Bank number 2GV9). The SwissProt accession numbers Q6S6P1 and P28858 were used for modeling Pol N752 and D752, respectively. The software program SWISS-MODEL (http://swissmodel.expasy.org/SWISS-MODEL.html
) was used to make the diagrams [36
]. In the surface rendering diagram, the solvent-accessible surface is blue and surface-exposed residues are red. All diagrams show the protein in the same orientation (annotated in E).