Materials and Methods

Experimental animals—The procedure for antemortem detection of EHV-1 latency and subsequent identification of latent virus pathotype was performed in 2 categories of adult horses. The first group was composed of 24 young mixedbreed mares that had been experimentally infected as weanlings 4 to 5 years previously with neuropathogenic (n = 12 horses) or non-neuropathogenic (12) strains of EHV-1. The second group of horses was composed of 12 aged Thoroughbred mares donated by commercial horse farms to the University of Kentucky because of chronic problems with infertility, laminitis, or other forms of lameness. The 24 experimentally infected horses were kept together as an isolated group in a 10-acre grass field. None of the horses had clinical signs or laboratory evidence of EHV-1 infection at the time of MLN biopsy.

MLN biopsy procedure—Mares were sedated by IV administration of 10 mg of butorphanol tartrate solution and 500 mg of xylazine hydrochloride solution. After 5 minutes, horses were anesthetized via IV administration of 1.0 g of ketamine hydrochloride solution. With a sterile scalpel blade, a 1.5- inch incision was made through the skin overlying palpable MLNs. An aggregate of lymph nodes weighing approximately 1.0 g was aseptically freed from surrounding connective tissue by means of blunt dissection and placed in sterile saline (0.9% NaCl) solution for transport on ice to the laboratory.

Preparation of DNA from MLN tissue—For isolation of cellular DNA from MLN tissues, 750 mg of tissue was finely minced with a sterile disposable scalpel blade, homogenized in 5.0 mL of lysis buffer solution (10mM Tris-HCl [pH, 8.0]; 100mM EDTA; 0.5% SDS), and incubated for approximately 24 hours at 50°C with proteinase K^a (concentration, 100 μg/mL) and ribonuclease A^b (concentration, 20 μg/mL). Purification of DNA from the digested cell lysate proceeded via 2 extractions with phenol:chloroform:isoamyl alcohol^a (25:24:1) solution and precipitation of DNA from the aqueous phase with ammonium acetate and isopropanol. The DNA precipitate was collected by use of centrifugation, rinsed twice with 70% ethanol, dissolved in 2.0 mL of sterile water, and quantitated by measuring spectrometric absorbance at 260 nm.

Detection of EHV-1 DNA by PCR assay—The DNA purified from MLN tissue was tested for EHV-1 DNA by use of the magnetic bead, sequence-capture, nested PCR method.^15–19 The procedure involves oligonucleotidehybridization enrichment and biotin-streptavidin magnetic bead technology for capture of EHV-1 DNA and was developed for detection of rare, low-abundance sequences below the detection threshold of conventional nested PCR assays. Two thousand four hundred micrograms of each MLN DNA preparation was analyzed for EHV-1 DNA. Mandibular lymph node DNA was digested overnight at 37°C with 7,500 units of Bgl-II restriction endonuclease,^c denatured by boiling for 10 minutes, and hybridized for 24 hours at 60°C with 24 pmol of a biotinylated capture oligonucleotide (Appendix) complementary to the EHV-1 gene sequence encoding the viral DNA polymerase (ie, ORF30). Six hundred micrograms of streptavidin-coated paramagnetic beads^e was added to the hybridization mixture, and the mixture was incubated for 24 hours at 43°C. After washing, beads were dispensed into six 100-μL PCR reaction mixtures containing 200μM each of deoxynucleoside triphosphates, 2.5mM MgCl₂, 5 units of Taq DNA polymerase,^f and 0.25μM ORF30 primers designed to amplify a 256-bp gene fragment that encompasses the site of the mutation associated with neuropathotype strains of EHV-1 (ORF30 G₂₂₅₄).²⁰ The biotinylated capture oligonucleotide and the amplification primers used are specific for EHV-1 and do not cross-amplify DNA from EHV-4. Amplification via PCR was performed according to the following specifications: initial denaturation at 94°C for 3 minutes; 40 cycles of 94°C for 30 seconds, 60°C for 30 seconds, and 65°C for 1 minute; and a final extension at 65°C for 10 minutes. The second round of PCR consisted of amplification of 2 μL of a 1:10 dilution of the first PCR product with a nested set of primers and the same thermocycler specifications. After the second round of PCR amplification, reaction mixtures were analyzed for a 256-bp DNA fragment by use of electrophoresis through 1.5% agarose gel.

Peripheral blood mononuclear cells collected at the time of MLN biopsy from each of the 36 horses were also assayed for EHV-1 DNA with an identical amount (2.4 mg) of PBMC DNA and the same sequence-capture, nested PCR procedure.

To minimize the possibility of contamination of reaction tubes with previously amplified PCR products or from another test sample, sample DNA preparation, PCR assay setup and amplification, and postamplification analysis were performed in separate rooms. Also, room-dedicated reagents, micropipets, and filtered pipet tips were used. Finally, a comparable mass of MLN tissue in which there was no EHV-1 DNA (collected from horses submitted for necropsy) was processed alongside each set of test samples.

To compare detection sensitivity of the magnetic bead, sequence-capture PCR method with that of conventional nested and real-time PCR assays, DNA preparations from the MLN of each of the 36 horses were also tested via the other 2 formats.^21,22

Reverse transcription-PCR assay—To rule out the possibility of active (rather than latent) EHV-1 infection in the horses with positive PCR assay results, RNA was isolated from MLN tissue of 6 of the mares with positive results by means of extraction with total RNA isolation reagent^g and subjected to magnetic bead–based, sequence-capture reverse transcriptase–nested PCR²³ by use of the same capture probe and PCR primers as described.

Pathotype identification of latent EHV-1—For MLNs with positive results for EHV-1 DNA, the 256-bp fragments amplified by sequence-capture, nested PCR were sequenced to identify the base at the polymorphic site (ie, position 2254 of ORF30) that allows determination of the pathotype of the latent herpesvirus strain (ie, neuropathotype [G₂₂₅₄] or nonneuropathotype [A₂₂₅₄]). The 256-bp amplicons were purified by use of a DNA purification resin.^b Nucleotide sequence determination of the amplicons was performed at the University of Kentucky core sequencing facility by use of an automated fluorescence sequencer^h and cycle sequencing^h technology.

Results

Biopsy specimens of MLNs collected from each of the 36 mares were tested for EHV-1 DNA. Results of tests for detection of EHV-1 latency and identification of the pathotype of each latent virus strain were summarized (Table 1). Mandibular lymph node tissue from 26 of the 36 (72%) study mares contained latent EHV-1 DNA. Equine herpesvirus-1 DNA was detected in 18 of the 24 (75%) experimental mares that had been inoculated 4 to 5 years previously with known neuropathogenic or non-neuropathogenic EHV-1 strains. Both pathotypes were identified in the latently infected mares, and in each instance, the pathotype of latent EHV-1 detected in the MLN was identical to the pathotype with which the horse had been inoculated intranasally 4 to 5 years earlier. Latent viral DNA was detected in 10 of the 12 mares inoculated with neuropathogenic strains of EHV-1 as weanlings. The results yielded no evidence for differences between the 2 EHV-1 pathotypes in their capacity for establishing postinfection latency. Eight of the 12 donated mares (67%) had positive test results for latent EHV-1 in MLN tissue, and only 1 of those 8 mares had a neuropathogenic strain of EHV-1. Sequence data at the polymorphic site (2254 A/G) of ORF30 revealed no evidence of coinfection with both EHV-1 pathotypes in any of the mares latently infected with the virus.

None of the PBMC samples collected from the 26 horses with latent EHV-1 DNA in the MLN contained EHV-1 DNA. Only 13 of the 26 (50%) horses with positive results for EHV-1 DNA via sequence-capture PCR assay also had positive results according to conventional nested or real-time PCR assays. The specificity of the antemortem detection procedure for detection of latent EHV-1 was validated by the absence of PCR amplification in EHV-1–negative MLN DNA samples to which EHV-4 DNA had been added.

Results of reverse transcription–PCR assay yielded no evidence of EHV-1 RNA transcribed from the ORF30 gene in MLN tissue from any of 6 tested mares that had positive results for EHV-1 DNA.

The A₂₂₅₄-to-G₂₂₅₄ mutation unique to the ORF30 gene in neuropathogenic strains of EHV-1 creates a Sal I cleavage site and a consequent restriction fragment length polymorphism in the 256-bp DNA fragment of ORF30 amplified by PCR. Agarose gel electrophoretic analysis of DNA fragments generated by Sal I digestion of the 256-bp amplicon provided an alternative and simple approach (compared with DNA sequencing) for differentiation of neuropathogenic from non-neuropathogenic strains of EHV-1 (Figure 1) The results also revealed that the quantitative burden of latent viral DNA in MLNs varied among horses (Figure 2). In some horses, EHV-1 DNA was amplified from all 6 sample replicates, whereas in others, as few as 1 or 2 replicates contained detectable viral DNA.

View Full Size

Results of agarose gel electrophoresis of DNA fragments after Sal I digestion of the PCR-amplified, 256-bp EHV-1 ORF30 gene fragment that encompasses the site of the A₂₂₅₄-to-G₂₂₅₄ mutation associated with neuropathogenic potential of the virus. Lanes 1 to 3 and 5 to 7 were loaded with Sal I digests of the 256-bp amplicons generated from neuropathogenic and non-neuropathogenic isolates of EHV-1, respectively. Lane 4 was loaded with a SalI digest of a negative-control PCR amplification containing MLN DNA free of EHV-1 sequences. A 100-bp DNA ladder was electrophoresed in lane 8.

Citation: American Journal of Veterinary Research 67, 8; 10.2460/ajvr.67.8.1401

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Results of agarose gel electrophoresis of PCR products in 6 replicate amplification reactions, with input template DNA derived from biotinylated oligonucleotide-hybridization and streptavidin-coated magnetic bead capture of EHV-1 DNA sequences in 750 mg of MLN tissue collected via biopsy from each of 2 horses (horses A and B). In the far right lane of each gel is a 100-bp DNA ladder for molecular sizing of the PCR-amplified products. Notice that in MLN tissue from horse A, all 6 replicate PCR amplification reactions contained an EHV-1–specific, 256-bp amplicon, whereas in tissue from horse B, only 2 PCR reactions contained the amplified EHV-1 ORF30 gene fragment.

Citation: American Journal of Veterinary Research 67, 8; 10.2460/ajvr.67.8.1401

• Download Figure
• Download figure as PowerPoint slide

Discussion

The technique described in the present study permitted determination of whether certain horses carry DNA of neuropathogenic strains of EHV-1 latently sequestered in cells of the respiratory tract lymph nodes. A single-point mutation (A-to-G transition) in EHV-1 ORF30 that encodes the catalytic subunit of the viral DNA polymerase is highly associated with the viral attribute of neuropathogenicity.²⁰ The recent increase in outbreaks of neurologic disease caused by such EHV-1 mutants has generated a practical need for identifying latent carriers of mutant neuropathogenic strains of the virus.

Detection of rare herpesviral DNA sequences among large quantities of host cellular DNA is of key importance in investigations of EHV-1 latency. Both conventional nested PCR and real-time PCR techniques are limited by the sample mass that can be tested (microgram quantities), reducing their detection of ultra-low–frequency target molecules. The present report describes the development and validation of an alternative enhanced-sensitivity PCR methodology (sequence-capture, nested PCR) and its usefulness in antemortem detection of latent EHV-1 DNA in biopsy specimens of MLN tissue. The 3 steps of the technique are enrichment of EHV-1 DNA from a large sample mass of cellular DNA by means of hybridization to a biotinylated EHV-1 allele-specific oligonucleotide, capture of the oligo–EHV-1 hybrids on streptavidin-coated paramagnetic beads, and amplification of the captured EHV-1 DNA by use of nested PCR. A final step of nucleotide sequencing or restriction fragment length polymorphism of the amplified DNA permits pathotype identification of the latent virus. The settings for temperature, time, and concentration during the procedure were optimized for detection of latent EHV-1 DNA in MLN tissue. Results of preliminary studies designed to assess the detection threshold of the sequence-capture PCR method by testing for low numbers of EHV-1 DNA molecules in a dilution series indicated that the threshold for detection was limited only by the volume of sample available for analysis. The technique used in the study enables detection of EHV-1 DNA molecules at very low concentrations in large amounts (ie, in milligram quantities) of cellular DNA. A procedure for antemortem screening of groups of horses for carriers of latent neuropathogenic strains of EHV-1 has not previously been available.

To ascertain whether the technique could be used for screening horses for latent EHV-1 infections and for determining the pathotype of detected latent virus, MLNs of 24 experimental horses with previous exposure to known pathotypes of EHV-1 and of 12 horses with unknown EHV-1 infection histories were tested. Results of those experiments indicate that the procedure is a minimally invasive, practical, and low-threshold technique for detecting latent neuropathotype EHV-1 in live horses. Furthermore, the studies confirm that survivors of EHV-1 neurologic disease may become latently infected carriers of the hypervirulent neuropathotype strains.

Two unexpected findings resulted from the study. First, in 50% of the horses in which latent EHV-1 infection was detected in the MLN via sequence-capture nested PCR, the latent virus load was below the threshold of detection for conventional nested and real-time PCR assays. Second, EHV-1 DNA was not detected in the PBMCs of horses whose MLNs contained latent virus. Because only 6 μg of MLN DNA was screened by the conventional and real-time PCR assays (ie, six 1-μg replicates), 400 individual reactions would be required to achieve a level of detection similar to that of the sequence-capture PCR technique. Thus, results of studies that use only conventional PCR techniques or test only peripheral blood leukocytes as means of detecting latent EHV-1 infections would underestimate the true prevalence of both herpesviral pathotypes in the study sample. Regarding the nature of the cell type in which latent EHV-1 infection is established, data suggest that a sessile, noncirculating cell immobilized within lymph nodes may harbor the latent viral DNA molecules. Another question raised by results of the present study is whether EHV-1 DNA detected in PBMCs in previous investigations^24–26 on latency represented a true latent state or active persistent infection by the virus.

The present study also revealed that the quantitative burden of latent viral DNA in MLNs varied among horses. The MLNs of some horses contained latent EHV-1 DNA molecules in numbers low enough to result in a Poisson sampling distribution (ie, a substantial fraction of replicate samples without detectable EHV-1 DNA). It was this category of horses in which latency was not detected by conventional formats of PCR testing.

Critical to the success of the detection methodology was the large-mass sampling and testing (eg, 2.4 mg of cellular DNA purified from 750 mg of MLN tissue from each horse and tested as six 400-μg replicates). With conventional nested or real-time PCR assays, only microgram quantities of target DNA can be tested without encountering target-excess inhibition. The benefits of biotinylated oligonucleotide–magnetic bead enrichment of EHV-1 DNA and the ensuing removal of inhibitory substances associated with such large starting masses of cellular DNA are essential features of the enhanced sensitivity of sequence-capture PCR methodologies. Therefore, it is likely that the prevalence of latent EHV-1 even in MLN tissues would also be underestimated in studies in which only single, small-volume test samples of the tissues were evaluated.

Importantly, the restriction fragment length polymorphism associated with the sequence difference between the 2 pathotypes of EHV-1, apparent after Sal I digestion of the 256-bp PCR amplicon of ORF30, offers a simple, rapid, and less-expensive alternative to DNA sequencing for pathotype identification of latent EHV-1.